diff options
| author | Yoav Steinberg <yoav@monfort.co.il> | 2021-10-10 18:26:48 +0300 |
|---|---|---|
| committer | Yoav Steinberg <yoav@monfort.co.il> | 2021-10-10 18:26:48 +0300 |
| commit | 4d5911b4e47e1ac45fa53c1e80a735127bf8860e (patch) | |
| tree | 459fc082ecfd0ff2bfeb4bbf843cf6f93fed9459 /deps | |
| parent | 4a884343f5935f7d470ab0ce013a421f119cfb3a (diff) | |
| parent | 220a0f0880419450c9409202aac1fab4b8be0719 (diff) | |
| download | redis-4d5911b4e47e1ac45fa53c1e80a735127bf8860e.tar.gz | |
Merge commit '220a0f0880419450c9409202aac1fab4b8be0719' as 'deps/jemalloc'
Diffstat (limited to 'deps')
322 files changed, 80506 insertions, 0 deletions
diff --git a/deps/jemalloc/.appveyor.yml b/deps/jemalloc/.appveyor.yml new file mode 100644 index 000000000..90b03688d --- /dev/null +++ b/deps/jemalloc/.appveyor.yml @@ -0,0 +1,42 @@ +version: '{build}' + +environment: + matrix: + - MSYSTEM: MINGW64 + CPU: x86_64 + MSVC: amd64 + CONFIG_FLAGS: --enable-debug + - MSYSTEM: MINGW64 + CPU: x86_64 + CONFIG_FLAGS: --enable-debug + - MSYSTEM: MINGW32 + CPU: i686 + MSVC: x86 + CONFIG_FLAGS: --enable-debug + - MSYSTEM: MINGW32 + CPU: i686 + CONFIG_FLAGS: --enable-debug + - MSYSTEM: MINGW64 + CPU: x86_64 + MSVC: amd64 + - MSYSTEM: MINGW64 + CPU: x86_64 + - MSYSTEM: MINGW32 + CPU: i686 + MSVC: x86 + - MSYSTEM: MINGW32 + CPU: i686 + +install: + - set PATH=c:\msys64\%MSYSTEM%\bin;c:\msys64\usr\bin;%PATH% + - if defined MSVC call "c:\Program Files (x86)\Microsoft Visual Studio 14.0\VC\vcvarsall.bat" %MSVC% + - if defined MSVC pacman --noconfirm -Rsc mingw-w64-%CPU%-gcc gcc + - pacman --noconfirm -Suy mingw-w64-%CPU%-make + +build_script: + - bash -c "autoconf" + - bash -c "./configure $CONFIG_FLAGS" + - mingw32-make + - file lib/jemalloc.dll + - mingw32-make tests + - mingw32-make -k check diff --git a/deps/jemalloc/.autom4te.cfg b/deps/jemalloc/.autom4te.cfg new file mode 100644 index 000000000..fe2424db5 --- /dev/null +++ b/deps/jemalloc/.autom4te.cfg @@ -0,0 +1,3 @@ +begin-language: "Autoconf-without-aclocal-m4" +args: --no-cache +end-language: "Autoconf-without-aclocal-m4" diff --git a/deps/jemalloc/.cirrus.yml b/deps/jemalloc/.cirrus.yml new file mode 100644 index 000000000..019d2c38c --- /dev/null +++ b/deps/jemalloc/.cirrus.yml @@ -0,0 +1,21 @@ +env: + CIRRUS_CLONE_DEPTH: 1 + ARCH: amd64 + +task: + freebsd_instance: + matrix: + image: freebsd-12-0-release-amd64 + image: freebsd-11-2-release-amd64 + install_script: + - sed -i.bak -e 's,pkg+http://pkg.FreeBSD.org/\${ABI}/quarterly,pkg+http://pkg.FreeBSD.org/\${ABI}/latest,' /etc/pkg/FreeBSD.conf + - pkg upgrade -y + - pkg install -y autoconf gmake + script: + - autoconf + #- ./configure ${COMPILER_FLAGS:+ CC="$CC $COMPILER_FLAGS" CXX="$CXX $COMPILER_FLAGS" } $CONFIGURE_FLAGS + - ./configure + - export JFLAG=`sysctl -n kern.smp.cpus` + - gmake -j${JFLAG} + - gmake -j${JFLAG} tests + - gmake check diff --git a/deps/jemalloc/.gitattributes b/deps/jemalloc/.gitattributes new file mode 100644 index 000000000..6313b56c5 --- /dev/null +++ b/deps/jemalloc/.gitattributes @@ -0,0 +1 @@ +* text=auto eol=lf diff --git a/deps/jemalloc/.gitignore b/deps/jemalloc/.gitignore new file mode 100644 index 000000000..5ca0ad1da --- /dev/null +++ b/deps/jemalloc/.gitignore @@ -0,0 +1,92 @@ +/bin/jemalloc-config +/bin/jemalloc.sh +/bin/jeprof + +/config.stamp +/config.log +/config.status +/configure + +/doc/html.xsl +/doc/manpages.xsl +/doc/jemalloc.xml +/doc/jemalloc.html +/doc/jemalloc.3 + +/jemalloc.pc + +/lib/ + +/Makefile + +/include/jemalloc/internal/jemalloc_preamble.h +/include/jemalloc/internal/jemalloc_internal_defs.h +/include/jemalloc/internal/private_namespace.gen.h +/include/jemalloc/internal/private_namespace.h +/include/jemalloc/internal/private_namespace_jet.gen.h +/include/jemalloc/internal/private_namespace_jet.h +/include/jemalloc/internal/private_symbols.awk +/include/jemalloc/internal/private_symbols_jet.awk +/include/jemalloc/internal/public_namespace.h +/include/jemalloc/internal/public_symbols.txt +/include/jemalloc/internal/public_unnamespace.h +/include/jemalloc/jemalloc.h +/include/jemalloc/jemalloc_defs.h +/include/jemalloc/jemalloc_macros.h +/include/jemalloc/jemalloc_mangle.h +/include/jemalloc/jemalloc_mangle_jet.h +/include/jemalloc/jemalloc_protos.h +/include/jemalloc/jemalloc_protos_jet.h +/include/jemalloc/jemalloc_rename.h +/include/jemalloc/jemalloc_typedefs.h + +/src/*.[od] +/src/*.sym + +/run_tests.out/ + +/test/test.sh +test/include/test/jemalloc_test.h +test/include/test/jemalloc_test_defs.h + +/test/integration/[A-Za-z]* +!/test/integration/[A-Za-z]*.* +/test/integration/*.[od] +/test/integration/*.out + +/test/integration/cpp/[A-Za-z]* +!/test/integration/cpp/[A-Za-z]*.* +/test/integration/cpp/*.[od] +/test/integration/cpp/*.out + +/test/src/*.[od] + +/test/stress/[A-Za-z]* +!/test/stress/[A-Za-z]*.* +/test/stress/*.[od] +/test/stress/*.out + +/test/unit/[A-Za-z]* +!/test/unit/[A-Za-z]*.* +/test/unit/*.[od] +/test/unit/*.out + +/VERSION + +*.pdb +*.sdf +*.opendb +*.VC.db +*.opensdf +*.cachefile +*.suo +*.user +*.sln.docstates +*.tmp +.vs/ +/msvc/Win32/ +/msvc/x64/ +/msvc/projects/*/*/Debug*/ +/msvc/projects/*/*/Release*/ +/msvc/projects/*/*/Win32/ +/msvc/projects/*/*/x64/ diff --git a/deps/jemalloc/.travis.yml b/deps/jemalloc/.travis.yml new file mode 100644 index 000000000..2da5da8eb --- /dev/null +++ b/deps/jemalloc/.travis.yml @@ -0,0 +1,195 @@ +language: generic +dist: precise + +matrix: + include: + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: osx + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=clang CXX=clang++ COMPILER_FLAGS="" CONFIGURE_FLAGS="" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="-m32" CONFIGURE_FLAGS="" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + addons: &gcc_multilib + apt: + packages: + - gcc-multilib + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-debug" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-prof" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--disable-stats" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--disable-libdl" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-opt-safety-checks" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--with-malloc-conf=tcache:false" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--with-malloc-conf=dss:primary" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--with-malloc-conf=percpu_arena:percpu" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--with-malloc-conf=background_thread:true" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: osx + env: CC=clang CXX=clang++ COMPILER_FLAGS="" CONFIGURE_FLAGS="" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: osx + env: CC=gcc CXX=g++ COMPILER_FLAGS="-m32" CONFIGURE_FLAGS="" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: osx + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-debug" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: osx + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--disable-stats" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: osx + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--disable-libdl" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: osx + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-opt-safety-checks" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: osx + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--with-malloc-conf=tcache:false" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=clang CXX=clang++ COMPILER_FLAGS="-m32" CONFIGURE_FLAGS="" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + addons: *gcc_multilib + - os: linux + env: CC=clang CXX=clang++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-debug" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=clang CXX=clang++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-prof" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=clang CXX=clang++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--disable-stats" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=clang CXX=clang++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--disable-libdl" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=clang CXX=clang++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-opt-safety-checks" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=clang CXX=clang++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--with-malloc-conf=tcache:false" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=clang CXX=clang++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--with-malloc-conf=dss:primary" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=clang CXX=clang++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--with-malloc-conf=percpu_arena:percpu" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=clang CXX=clang++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--with-malloc-conf=background_thread:true" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="-m32" CONFIGURE_FLAGS="--enable-debug" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + addons: *gcc_multilib + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="-m32" CONFIGURE_FLAGS="--enable-prof" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + addons: *gcc_multilib + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="-m32" CONFIGURE_FLAGS="--disable-stats" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + addons: *gcc_multilib + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="-m32" CONFIGURE_FLAGS="--disable-libdl" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + addons: *gcc_multilib + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="-m32" CONFIGURE_FLAGS="--enable-opt-safety-checks" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + addons: *gcc_multilib + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="-m32" CONFIGURE_FLAGS="--with-malloc-conf=tcache:false" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + addons: *gcc_multilib + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="-m32" CONFIGURE_FLAGS="--with-malloc-conf=dss:primary" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + addons: *gcc_multilib + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="-m32" CONFIGURE_FLAGS="--with-malloc-conf=percpu_arena:percpu" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + addons: *gcc_multilib + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="-m32" CONFIGURE_FLAGS="--with-malloc-conf=background_thread:true" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + addons: *gcc_multilib + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-debug --enable-prof" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-debug --disable-stats" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-debug --disable-libdl" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-debug --enable-opt-safety-checks" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-debug --with-malloc-conf=tcache:false" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-debug --with-malloc-conf=dss:primary" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-debug --with-malloc-conf=percpu_arena:percpu" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-debug --with-malloc-conf=background_thread:true" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-prof --disable-stats" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-prof --disable-libdl" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-prof --enable-opt-safety-checks" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-prof --with-malloc-conf=tcache:false" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-prof --with-malloc-conf=dss:primary" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-prof --with-malloc-conf=percpu_arena:percpu" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-prof --with-malloc-conf=background_thread:true" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--disable-stats --disable-libdl" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--disable-stats --enable-opt-safety-checks" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--disable-stats --with-malloc-conf=tcache:false" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--disable-stats --with-malloc-conf=dss:primary" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--disable-stats --with-malloc-conf=percpu_arena:percpu" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--disable-stats --with-malloc-conf=background_thread:true" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--disable-libdl --enable-opt-safety-checks" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--disable-libdl --with-malloc-conf=tcache:false" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--disable-libdl --with-malloc-conf=dss:primary" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--disable-libdl --with-malloc-conf=percpu_arena:percpu" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--disable-libdl --with-malloc-conf=background_thread:true" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-opt-safety-checks --with-malloc-conf=tcache:false" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-opt-safety-checks --with-malloc-conf=dss:primary" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-opt-safety-checks --with-malloc-conf=percpu_arena:percpu" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-opt-safety-checks --with-malloc-conf=background_thread:true" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--with-malloc-conf=tcache:false,dss:primary" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--with-malloc-conf=tcache:false,percpu_arena:percpu" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--with-malloc-conf=tcache:false,background_thread:true" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--with-malloc-conf=dss:primary,percpu_arena:percpu" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--with-malloc-conf=dss:primary,background_thread:true" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--with-malloc-conf=percpu_arena:percpu,background_thread:true" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + # Development build + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-debug --disable-cache-oblivious --enable-stats --enable-log --enable-prof" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + # --enable-expermental-smallocx: + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-debug --enable-experimental-smallocx --enable-stats --enable-prof" EXTRA_CFLAGS="-Werror -Wno-array-bounds" + + # Valgrind + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="" EXTRA_CFLAGS="-Werror -Wno-array-bounds" JEMALLOC_TEST_PREFIX="valgrind" + addons: + apt: + packages: + - valgrind + + +before_script: + - autoconf + - scripts/gen_travis.py > travis_script && diff .travis.yml travis_script + - ./configure ${COMPILER_FLAGS:+ CC="$CC $COMPILER_FLAGS" CXX="$CXX $COMPILER_FLAGS" } $CONFIGURE_FLAGS + - make -j3 + - make -j3 tests + +script: + - make check + diff --git a/deps/jemalloc/COPYING b/deps/jemalloc/COPYING new file mode 100644 index 000000000..3b7fd3585 --- /dev/null +++ b/deps/jemalloc/COPYING @@ -0,0 +1,27 @@ +Unless otherwise specified, files in the jemalloc source distribution are +subject to the following license: +-------------------------------------------------------------------------------- +Copyright (C) 2002-present Jason Evans <jasone@canonware.com>. +All rights reserved. +Copyright (C) 2007-2012 Mozilla Foundation. All rights reserved. +Copyright (C) 2009-present Facebook, Inc. All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are met: +1. Redistributions of source code must retain the above copyright notice(s), + this list of conditions and the following disclaimer. +2. Redistributions in binary form must reproduce the above copyright notice(s), + this list of conditions and the following disclaimer in the documentation + and/or other materials provided with the distribution. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDER(S) ``AS IS'' AND ANY EXPRESS +OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF +MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO +EVENT SHALL THE COPYRIGHT HOLDER(S) BE LIABLE FOR ANY DIRECT, INDIRECT, +INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR +PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF +LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE +OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF +ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. +-------------------------------------------------------------------------------- diff --git a/deps/jemalloc/ChangeLog b/deps/jemalloc/ChangeLog new file mode 100644 index 000000000..e55813b7b --- /dev/null +++ b/deps/jemalloc/ChangeLog @@ -0,0 +1,1518 @@ +Following are change highlights associated with official releases. Important +bug fixes are all mentioned, but some internal enhancements are omitted here for +brevity. Much more detail can be found in the git revision history: + + https://github.com/jemalloc/jemalloc + +* 5.2.1 (August 5, 2019) + + This release is primarily about Windows. A critical virtual memory leak is + resolved on all Windows platforms. The regression was present in all releases + since 5.0.0. + + Bug fixes: + - Fix a severe virtual memory leak on Windows. This regression was first + released in 5.0.0. (@Ignition, @j0t, @frederik-h, @davidtgoldblatt, + @interwq) + - Fix size 0 handling in posix_memalign(). This regression was first released + in 5.2.0. (@interwq) + - Fix the prof_log unit test which may observe unexpected backtraces from + compiler optimizations. The test was first added in 5.2.0. (@marxin, + @gnzlbg, @interwq) + - Fix the declaration of the extent_avail tree. This regression was first + released in 5.1.0. (@zoulasc) + - Fix an incorrect reference in jeprof. This functionality was first released + in 3.0.0. (@prehistoric-penguin) + - Fix an assertion on the deallocation fast-path. This regression was first + released in 5.2.0. (@yinan1048576) + - Fix the TLS_MODEL attribute in headers. This regression was first released + in 5.0.0. (@zoulasc, @interwq) + + Optimizations and refactors: + - Implement opt.retain on Windows and enable by default on 64-bit. (@interwq, + @davidtgoldblatt) + - Optimize away a branch on the operator delete[] path. (@mgrice) + - Add format annotation to the format generator function. (@zoulasc) + - Refactor and improve the size class header generation. (@yinan1048576) + - Remove best fit. (@djwatson) + - Avoid blocking on background thread locks for stats. (@oranagra, @interwq) + +* 5.2.0 (April 2, 2019) + + This release includes a few notable improvements, which are summarized below: + 1) improved fast-path performance from the optimizations by @djwatson; 2) + reduced virtual memory fragmentation and metadata usage; and 3) bug fixes on + setting the number of background threads. In addition, peak / spike memory + usage is improved with certain allocation patterns. As usual, the release and + prior dev versions have gone through large-scale production testing. + + New features: + - Implement oversize_threshold, which uses a dedicated arena for allocations + crossing the specified threshold to reduce fragmentation. (@interwq) + - Add extents usage information to stats. (@tyleretzel) + - Log time information for sampled allocations. (@tyleretzel) + - Support 0 size in sdallocx. (@djwatson) + - Output rate for certain counters in malloc_stats. (@zinoale) + - Add configure option --enable-readlinkat, which allows the use of readlinkat + over readlink. (@davidtgoldblatt) + - Add configure options --{enable,disable}-{static,shared} to allow not + building unwanted libraries. (@Ericson2314) + - Add configure option --disable-libdl to enable fully static builds. + (@interwq) + - Add mallctl interfaces: + + opt.oversize_threshold (@interwq) + + stats.arenas.<i>.extent_avail (@tyleretzel) + + stats.arenas.<i>.extents.<j>.n{dirty,muzzy,retained} (@tyleretzel) + + stats.arenas.<i>.extents.<j>.{dirty,muzzy,retained}_bytes + (@tyleretzel) + + Portability improvements: + - Update MSVC builds. (@maksqwe, @rustyx) + - Workaround a compiler optimizer bug on s390x. (@rkmisra) + - Make use of pthread_set_name_np(3) on FreeBSD. (@trasz) + - Implement malloc_getcpu() to enable percpu_arena for windows. (@santagada) + - Link against -pthread instead of -lpthread. (@paravoid) + - Make background_thread not dependent on libdl. (@interwq) + - Add stringify to fix a linker directive issue on MSVC. (@daverigby) + - Detect and fall back when 8-bit atomics are unavailable. (@interwq) + - Fall back to the default pthread_create if dlsym(3) fails. (@interwq) + + Optimizations and refactors: + - Refactor the TSD module. (@davidtgoldblatt) + - Avoid taking extents_muzzy mutex when muzzy is disabled. (@interwq) + - Avoid taking large_mtx for auto arenas on the tcache flush path. (@interwq) + - Optimize ixalloc by avoiding a size lookup. (@interwq) + - Implement opt.oversize_threshold which uses a dedicated arena for requests + crossing the threshold, also eagerly purges the oversize extents. Default + the threshold to 8 MiB. (@interwq) + - Clean compilation with -Wextra. (@gnzlbg, @jasone) + - Refactor the size class module. (@davidtgoldblatt) + - Refactor the stats emitter. (@tyleretzel) + - Optimize pow2_ceil. (@rkmisra) + - Avoid runtime detection of lazy purging on FreeBSD. (@trasz) + - Optimize mmap(2) alignment handling on FreeBSD. (@trasz) + - Improve error handling for THP state initialization. (@jsteemann) + - Rework the malloc() fast path. (@djwatson) + - Rework the free() fast path. (@djwatson) + - Refactor and optimize the tcache fill / flush paths. (@djwatson) + - Optimize sync / lwsync on PowerPC. (@chmeeedalf) + - Bypass extent_dalloc() when retain is enabled. (@interwq) + - Optimize the locking on large deallocation. (@interwq) + - Reduce the number of pages committed from sanity checking in debug build. + (@trasz, @interwq) + - Deprecate OSSpinLock. (@interwq) + - Lower the default number of background threads to 4 (when the feature + is enabled). (@interwq) + - Optimize the trylock spin wait. (@djwatson) + - Use arena index for arena-matching checks. (@interwq) + - Avoid forced decay on thread termination when using background threads. + (@interwq) + - Disable muzzy decay by default. (@djwatson, @interwq) + - Only initialize libgcc unwinder when profiling is enabled. (@paravoid, + @interwq) + + Bug fixes (all only relevant to jemalloc 5.x): + - Fix background thread index issues with max_background_threads. (@djwatson, + @interwq) + - Fix stats output for opt.lg_extent_max_active_fit. (@interwq) + - Fix opt.prof_prefix initialization. (@davidtgoldblatt) + - Properly trigger decay on tcache destroy. (@interwq, @amosbird) + - Fix tcache.flush. (@interwq) + - Detect whether explicit extent zero out is necessary with huge pages or + custom extent hooks, which may change the purge semantics. (@interwq) + - Fix a side effect caused by extent_max_active_fit combined with decay-based + purging, where freed extents can accumulate and not be reused for an + extended period of time. (@interwq, @mpghf) + - Fix a missing unlock on extent register error handling. (@zoulasc) + + Testing: + - Simplify the Travis script output. (@gnzlbg) + - Update the test scripts for FreeBSD. (@devnexen) + - Add unit tests for the producer-consumer pattern. (@interwq) + - Add Cirrus-CI config for FreeBSD builds. (@jasone) + - Add size-matching sanity checks on tcache flush. (@davidtgoldblatt, + @interwq) + + Incompatible changes: + - Remove --with-lg-page-sizes. (@davidtgoldblatt) + + Documentation: + - Attempt to build docs by default, however skip doc building when xsltproc + is missing. (@interwq, @cmuellner) + +* 5.1.0 (May 4, 2018) + + This release is primarily about fine-tuning, ranging from several new features + to numerous notable performance and portability enhancements. The release and + prior dev versions have been running in multiple large scale applications for + months, and the cumulative improvements are substantial in many cases. + + Given the long and successful production runs, this release is likely a good + candidate for applications to upgrade, from both jemalloc 5.0 and before. For + performance-critical applications, the newly added TUNING.md provides + guidelines on jemalloc tuning. + + New features: + - Implement transparent huge page support for internal metadata. (@interwq) + - Add opt.thp to allow enabling / disabling transparent huge pages for all + mappings. (@interwq) + - Add maximum background thread count option. (@djwatson) + - Allow prof_active to control opt.lg_prof_interval and prof.gdump. + (@interwq) + - Allow arena index lookup based on allocation addresses via mallctl. + (@lionkov) + - Allow disabling initial-exec TLS model. (@davidtgoldblatt, @KenMacD) + - Add opt.lg_extent_max_active_fit to set the max ratio between the size of + the active extent selected (to split off from) and the size of the requested + allocation. (@interwq, @davidtgoldblatt) + - Add retain_grow_limit to set the max size when growing virtual address + space. (@interwq) + - Add mallctl interfaces: + + arena.<i>.retain_grow_limit (@interwq) + + arenas.lookup (@lionkov) + + max_background_threads (@djwatson) + + opt.lg_extent_max_active_fit (@interwq) + + opt.max_background_threads (@djwatson) + + opt.metadata_thp (@interwq) + + opt.thp (@interwq) + + stats.metadata_thp (@interwq) + + Portability improvements: + - Support GNU/kFreeBSD configuration. (@paravoid) + - Support m68k, nios2 and SH3 architectures. (@paravoid) + - Fall back to FD_CLOEXEC when O_CLOEXEC is unavailable. (@zonyitoo) + - Fix symbol listing for cross-compiling. (@tamird) + - Fix high bits computation on ARM. (@davidtgoldblatt, @paravoid) + - Disable the CPU_SPINWAIT macro for Power. (@davidtgoldblatt, @marxin) + - Fix MSVC 2015 & 2017 builds. (@rustyx) + - Improve RISC-V support. (@EdSchouten) + - Set name mangling script in strict mode. (@nicolov) + - Avoid MADV_HUGEPAGE on ARM. (@marxin) + - Modify configure to determine return value of strerror_r. + (@davidtgoldblatt, @cferris1000) + - Make sure CXXFLAGS is tested with CPP compiler. (@nehaljwani) + - Fix 32-bit build on MSVC. (@rustyx) + - Fix external symbol on MSVC. (@maksqwe) + - Avoid a printf format specifier warning. (@jasone) + - Add configure option --disable-initial-exec-tls which can allow jemalloc to + be dynamically loaded after program startup. (@davidtgoldblatt, @KenMacD) + - AArch64: Add ILP32 support. (@cmuellner) + - Add --with-lg-vaddr configure option to support cross compiling. + (@cmuellner, @davidtgoldblatt) + + Optimizations and refactors: + - Improve active extent fit with extent_max_active_fit. This considerably + reduces fragmentation over time and improves virtual memory and metadata + usage. (@davidtgoldblatt, @interwq) + - Eagerly coalesce large extents to reduce fragmentation. (@interwq) + - sdallocx: only read size info when page aligned (i.e. possibly sampled), + which speeds up the sized deallocation path significantly. (@interwq) + - Avoid attempting new mappings for in place expansion with retain, since + it rarely succeeds in practice and causes high overhead. (@interwq) + - Refactor OOM handling in newImpl. (@wqfish) + - Add internal fine-grained logging functionality for debugging use. + (@davidtgoldblatt) + - Refactor arena / tcache interactions. (@davidtgoldblatt) + - Refactor extent management with dumpable flag. (@davidtgoldblatt) + - Add runtime detection of lazy purging. (@interwq) + - Use pairing heap instead of red-black tree for extents_avail. (@djwatson) + - Use sysctl on startup in FreeBSD. (@trasz) + - Use thread local prng state instead of atomic. (@djwatson) + - Make decay to always purge one more extent than before, because in + practice large extents are usually the ones that cross the decay threshold. + Purging the additional extent helps save memory as well as reduce VM + fragmentation. (@interwq) + - Fast division by dynamic values. (@davidtgoldblatt) + - Improve the fit for aligned allocation. (@interwq, @edwinsmith) + - Refactor extent_t bitpacking. (@rkmisra) + - Optimize the generated assembly for ticker operations. (@davidtgoldblatt) + - Convert stats printing to use a structured text emitter. (@davidtgoldblatt) + - Remove preserve_lru feature for extents management. (@djwatson) + - Consolidate two memory loads into one on the fast deallocation path. + (@davidtgoldblatt, @interwq) + + Bug fixes (most of the issues are only relevant to jemalloc 5.0): + - Fix deadlock with multithreaded fork in OS X. (@davidtgoldblatt) + - Validate returned file descriptor before use. (@zonyitoo) + - Fix a few background thread initialization and shutdown issues. (@interwq) + - Fix an extent coalesce + decay race by taking both coalescing extents off + the LRU list. (@interwq) + - Fix potentially unbound increase during decay, caused by one thread keep + stashing memory to purge while other threads generating new pages. The + number of pages to purge is checked to prevent this. (@interwq) + - Fix a FreeBSD bootstrap assertion. (@strejda, @interwq) + - Handle 32 bit mutex counters. (@rkmisra) + - Fix a indexing bug when creating background threads. (@davidtgoldblatt, + @binliu19) + - Fix arguments passed to extent_init. (@yuleniwo, @interwq) + - Fix addresses used for ordering mutexes. (@rkmisra) + - Fix abort_conf processing during bootstrap. (@interwq) + - Fix include path order for out-of-tree builds. (@cmuellner) + + Incompatible changes: + - Remove --disable-thp. (@interwq) + - Remove mallctl interfaces: + + config.thp (@interwq) + + Documentation: + - Add TUNING.md. (@interwq, @davidtgoldblatt, @djwatson) + +* 5.0.1 (July 1, 2017) + + This bugfix release fixes several issues, most of which are obscure enough + that typical applications are not impacted. + + Bug fixes: + - Update decay->nunpurged before purging, in order to avoid potential update + races and subsequent incorrect purging volume. (@interwq) + - Only abort on dlsym(3) error if the failure impacts an enabled feature (lazy + locking and/or background threads). This mitigates an initialization + failure bug for which we still do not have a clear reproduction test case. + (@interwq) + - Modify tsd management so that it neither crashes nor leaks if a thread's + only allocation activity is to call free() after TLS destructors have been + executed. This behavior was observed when operating with GNU libc, and is + unlikely to be an issue with other libc implementations. (@interwq) + - Mask signals during background thread creation. This prevents signals from + being inadvertently delivered to background threads. (@jasone, + @davidtgoldblatt, @interwq) + - Avoid inactivity checks within background threads, in order to prevent + recursive mutex acquisition. (@interwq) + - Fix extent_grow_retained() to use the specified hooks when the + arena.<i>.extent_hooks mallctl is used to override the default hooks. + (@interwq) + - Add missing reentrancy support for custom extent hooks which allocate. + (@interwq) + - Post-fork(2), re-initialize the list of tcaches associated with each arena + to contain no tcaches except the forking thread's. (@interwq) + - Add missing post-fork(2) mutex reinitialization for extent_grow_mtx. This + fixes potential deadlocks after fork(2). (@interwq) + - Enforce minimum autoconf version (currently 2.68), since 2.63 is known to + generate corrupt configure scripts. (@jasone) + - Ensure that the configured page size (--with-lg-page) is no larger than the + configured huge page size (--with-lg-hugepage). (@jasone) + +* 5.0.0 (June 13, 2017) + + Unlike all previous jemalloc releases, this release does not use naturally + aligned "chunks" for virtual memory management, and instead uses page-aligned + "extents". This change has few externally visible effects, but the internal + impacts are... extensive. Many other internal changes combine to make this + the most cohesively designed version of jemalloc so far, with ample + opportunity for further enhancements. + + Continuous integration is now an integral aspect of development thanks to the + efforts of @davidtgoldblatt, and the dev branch tends to remain reasonably + stable on the tested platforms (Linux, FreeBSD, macOS, and Windows). As a + side effect the official release frequency may decrease over time. + + New features: + - Implement optional per-CPU arena support; threads choose which arena to use + based on current CPU rather than on fixed thread-->arena associations. + (@interwq) + - Implement two-phase decay of unused dirty pages. Pages transition from + dirty-->muzzy-->clean, where the first phase transition relies on + madvise(... MADV_FREE) semantics, and the second phase transition discards + pages such that they are replaced with demand-zeroed pages on next access. + (@jasone) + - Increase decay time resolution from seconds to milliseconds. (@jasone) + - Implement opt-in per CPU background threads, and use them for asynchronous + decay-driven unused dirty page purging. (@interwq) + - Add mutex profiling, which collects a variety of statistics useful for + diagnosing overhead/contention issues. (@interwq) + - Add C++ new/delete operator bindings. (@djwatson) + - Support manually created arena destruction, such that all data and metadata + are discarded. Add MALLCTL_ARENAS_DESTROYED for accessing merged stats + associated with destroyed arenas. (@jasone) + - Add MALLCTL_ARENAS_ALL as a fixed index for use in accessing + merged/destroyed arena statistics via mallctl. (@jasone) + - Add opt.abort_conf to optionally abort if invalid configuration options are + detected during initialization. (@interwq) + - Add opt.stats_print_opts, so that e.g. JSON output can be selected for the + stats dumped during exit if opt.stats_print is true. (@jasone) + - Add --with-version=VERSION for use when embedding jemalloc into another + project's git repository. (@jasone) + - Add --disable-thp to support cross compiling. (@jasone) + - Add --with-lg-hugepage to support cross compiling. (@jasone) + - Add mallctl interfaces (various authors): + + background_thread + + opt.abort_conf + + opt.retain + + opt.percpu_arena + + opt.background_thread + + opt.{dirty,muzzy}_decay_ms + + opt.stats_print_opts + + arena.<i>.initialized + + arena.<i>.destroy + + arena.<i>.{dirty,muzzy}_decay_ms + + arena.<i>.extent_hooks + + arenas.{dirty,muzzy}_decay_ms + + arenas.bin.<i>.slab_size + + arenas.nlextents + + arenas.lextent.<i>.size + + arenas.create + + stats.background_thread.{num_threads,num_runs,run_interval} + + stats.mutexes.{ctl,background_thread,prof,reset}. + {num_ops,num_spin_acq,num_wait,max_wait_time,total_wait_time,max_num_thds, + num_owner_switch} + + stats.arenas.<i>.{dirty,muzzy}_decay_ms + + stats.arenas.<i>.uptime + + stats.arenas.<i>.{pmuzzy,base,internal,resident} + + stats.arenas.<i>.{dirty,muzzy}_{npurge,nmadvise,purged} + + stats.arenas.<i>.bins.<j>.{nslabs,reslabs,curslabs} + + stats.arenas.<i>.bins.<j>.mutex. + {num_ops,num_spin_acq,num_wait,max_wait_time,total_wait_time,max_num_thds, + num_owner_switch} + + stats.arenas.<i>.lextents.<j>.{nmalloc,ndalloc,nrequests,curlextents} + + stats.arenas.i.mutexes.{large,extent_avail,extents_dirty,extents_muzzy, + extents_retained,decay_dirty,decay_muzzy,base,tcache_list}. + {num_ops,num_spin_acq,num_wait,max_wait_time,total_wait_time,max_num_thds, + num_owner_switch} + + Portability improvements: + - Improve reentrant allocation support, such that deadlock is less likely if + e.g. a system library call in turn allocates memory. (@davidtgoldblatt, + @interwq) + - Support static linking of jemalloc with glibc. (@djwatson) + + Optimizations and refactors: + - Organize virtual memory as "extents" of virtual memory pages, rather than as + naturally aligned "chunks", and store all metadata in arbitrarily distant + locations. This reduces virtual memory external fragmentation, and will + interact better with huge pages (not yet explicitly supported). (@jasone) + - Fold large and huge size classes together; only small and large size classes + remain. (@jasone) + - Unify the allocation paths, and merge most fast-path branching decisions. + (@davidtgoldblatt, @interwq) + - Embed per thread automatic tcache into thread-specific data, which reduces + conditional branches and dereferences. Also reorganize tcache to increase + fast-path data locality. (@interwq) + - Rewrite atomics to closely model the C11 API, convert various + synchronization from mutex-based to atomic, and use the explicit memory + ordering control to resolve various hypothetical races without increasing + synchronization overhead. (@davidtgoldblatt) + - Extensively optimize rtree via various methods: + + Add multiple layers of rtree lookup caching, since rtree lookups are now + part of fast-path deallocation. (@interwq) + + Determine rtree layout at compile time. (@jasone) + + Make the tree shallower for common configurations. (@jasone) + + Embed the root node in the top-level rtree data structure, thus avoiding + one level of indirection. (@jasone) + + Further specialize leaf elements as compared to internal node elements, + and directly embed extent metadata needed for fast-path deallocation. + (@jasone) + + Ignore leading always-zero address bits (architecture-specific). + (@jasone) + - Reorganize headers (ongoing work) to make them hermetic, and disentangle + various module dependencies. (@davidtgoldblatt) + - Convert various internal data structures such as size class metadata from + boot-time-initialized to compile-time-initialized. Propagate resulting data + structure simplifications, such as making arena metadata fixed-size. + (@jasone) + - Simplify size class lookups when constrained to size classes that are + multiples of the page size. This speeds lookups, but the primary benefit is + complexity reduction in code that was the source of numerous regressions. + (@jasone) + - Lock individual extents when possible for localized extent operations, + rather than relying on a top-level arena lock. (@davidtgoldblatt, @jasone) + - Use first fit layout policy instead of best fit, in order to improve + packing. (@jasone) + - If munmap(2) is not in use, use an exponential series to grow each arena's + virtual memory, so that the number of disjoint virtual memory mappings + remains low. (@jasone) + - Implement per arena base allocators, so that arenas never share any virtual + memory pages. (@jasone) + - Automatically generate private symbol name mangling macros. (@jasone) + + Incompatible changes: + - Replace chunk hooks with an expanded/normalized set of extent hooks. + (@jasone) + - Remove ratio-based purging. (@jasone) + - Remove --disable-tcache. (@jasone) + - Remove --disable-tls. (@jasone) + - Remove --enable-ivsalloc. (@jasone) + - Remove --with-lg-size-class-group. (@jasone) + - Remove --with-lg-tiny-min. (@jasone) + - Remove --disable-cc-silence. (@jasone) + - Remove --enable-code-coverage. (@jasone) + - Remove --disable-munmap (replaced by opt.retain). (@jasone) + - Remove Valgrind support. (@jasone) + - Remove quarantine support. (@jasone) + - Remove redzone support. (@jasone) + - Remove mallctl interfaces (various authors): + + config.munmap + + config.tcache + + config.tls + + config.valgrind + + opt.lg_chunk + + opt.purge + + opt.lg_dirty_mult + + opt.decay_time + + opt.quarantine + + opt.redzone + + opt.thp + + arena.<i>.lg_dirty_mult + + arena.<i>.decay_time + + arena.<i>.chunk_hooks + + arenas.initialized + + arenas.lg_dirty_mult + + arenas.decay_time + + arenas.bin.<i>.run_size + + arenas.nlruns + + arenas.lrun.<i>.size + + arenas.nhchunks + + arenas.hchunk.<i>.size + + arenas.extend + + stats.cactive + + stats.arenas.<i>.lg_dirty_mult + + stats.arenas.<i>.decay_time + + stats.arenas.<i>.metadata.{mapped,allocated} + + stats.arenas.<i>.{npurge,nmadvise,purged} + + stats.arenas.<i>.huge.{allocated,nmalloc,ndalloc,nrequests} + + stats.arenas.<i>.bins.<j>.{nruns,reruns,curruns} + + stats.arenas.<i>.lruns.<j>.{nmalloc,ndalloc,nrequests,curruns} + + stats.arenas.<i>.hchunks.<j>.{nmalloc,ndalloc,nrequests,curhchunks} + + Bug fixes: + - Improve interval-based profile dump triggering to dump only one profile when + a single allocation's size exceeds the interval. (@jasone) + - Use prefixed function names (as controlled by --with-jemalloc-prefix) when + pruning backtrace frames in jeprof. (@jasone) + +* 4.5.0 (February 28, 2017) + + This is the first release to benefit from much broader continuous integration + testing, thanks to @davidtgoldblatt. Had we had this testing infrastructure + in place for prior releases, it would have caught all of the most serious + regressions fixed by this release. + + New features: + - Add --disable-thp and the opt.thp mallctl to provide opt-out mechanisms for + transparent huge page integration. (@jasone) + - Update zone allocator integration to work with macOS 10.12. (@glandium) + - Restructure *CFLAGS configuration, so that CFLAGS behaves typically, and + EXTRA_CFLAGS provides a way to specify e.g. -Werror during building, but not + during configuration. (@jasone, @ronawho) + + Bug fixes: + - Fix DSS (sbrk(2)-based) allocation. This regression was first released in + 4.3.0. (@jasone) + - Handle race in per size class utilization computation. This functionality + was first released in 4.0.0. (@interwq) + - Fix lock order reversal during gdump. (@jasone) + - Fix/refactor tcache synchronization. This regression was first released in + 4.0.0. (@jasone) + - Fix various JSON-formatted malloc_stats_print() bugs. This functionality + was first released in 4.3.0. (@jasone) + - Fix huge-aligned allocation. This regression was first released in 4.4.0. + (@jasone) + - When transparent huge page integration is enabled, detect what state pages + start in according to the kernel's current operating mode, and only convert + arena chunks to non-huge during purging if that is not their initial state. + This functionality was first released in 4.4.0. (@jasone) + - Fix lg_chunk clamping for the --enable-cache-oblivious --disable-fill case. + This regression was first released in 4.0.0. (@jasone, @428desmo) + - Properly detect sparc64 when building for Linux. (@glaubitz) + +* 4.4.0 (December 3, 2016) + + New features: + - Add configure support for *-*-linux-android. (@cferris1000, @jasone) + - Add the --disable-syscall configure option, for use on systems that place + security-motivated limitations on syscall(2). (@jasone) + - Add support for Debian GNU/kFreeBSD. (@thesam) + + Optimizations: + - Add extent serial numbers and use them where appropriate as a sort key that + is higher priority than address, so that the allocation policy prefers older + extents. This tends to improve locality (decrease fragmentation) when + memory grows downward. (@jasone) + - Refactor madvise(2) configuration so that MADV_FREE is detected and utilized + on Linux 4.5 and newer. (@jasone) + - Mark partially purged arena chunks as non-huge-page. This improves + interaction with Linux's transparent huge page functionality. (@jasone) + + Bug fixes: + - Fix size class computations for edge conditions involving extremely large + allocations. This regression was first released in 4.0.0. (@jasone, + @ingvarha) + - Remove overly restrictive assertions related to the cactive statistic. This + regression was first released in 4.1.0. (@jasone) + - Implement a more reliable detection scheme for os_unfair_lock on macOS. + (@jszakmeister) + +* 4.3.1 (November 7, 2016) + + Bug fixes: + - Fix a severe virtual memory leak. This regression was first released in + 4.3.0. (@interwq, @jasone) + - Refactor atomic and prng APIs to restore support for 32-bit platforms that + use pre-C11 toolchains, e.g. FreeBSD's mips. (@jasone) + +* 4.3.0 (November 4, 2016) + + This is the first release that passes the test suite for multiple Windows + configurations, thanks in large part to @glandium setting up continuous + integration via AppVeyor (and Travis CI for Linux and OS X). + + New features: + - Add "J" (JSON) support to malloc_stats_print(). (@jasone) + - Add Cray compiler support. (@ronawho) + + Optimizations: + - Add/use adaptive spinning for bootstrapping and radix tree node + initialization. (@jasone) + + Bug fixes: + - Fix large allocation to search starting in the optimal size class heap, + which can substantially reduce virtual memory churn and fragmentation. This + regression was first released in 4.0.0. (@mjp41, @jasone) + - Fix stats.arenas.<i>.nthreads accounting. (@interwq) + - Fix and simplify decay-based purging. (@jasone) + - Make DSS (sbrk(2)-related) operations lockless, which resolves potential + deadlocks during thread exit. (@jasone) + - Fix over-sized allocation of radix tree leaf nodes. (@mjp41, @ogaun, + @jasone) + - Fix over-sized allocation of arena_t (plus associated stats) data + structures. (@jasone, @interwq) + - Fix EXTRA_CFLAGS to not affect configuration. (@jasone) + - Fix a Valgrind integration bug. (@ronawho) + - Disallow 0x5a junk filling when running in Valgrind. (@jasone) + - Fix a file descriptor leak on Linux. This regression was first released in + 4.2.0. (@vsarunas, @jasone) + - Fix static linking of jemalloc with glibc. (@djwatson) + - Use syscall(2) rather than {open,read,close}(2) during boot on Linux. This + works around other libraries' system call wrappers performing reentrant + allocation. (@kspinka, @Whissi, @jasone) + - Fix OS X default zone replacement to work with OS X 10.12. (@glandium, + @jasone) + - Fix cached memory management to avoid needless commit/decommit operations + during purging, which resolves permanent virtual memory map fragmentation + issues on Windows. (@mjp41, @jasone) + - Fix TSD fetches to avoid (recursive) allocation. This is relevant to + non-TLS and Windows configurations. (@jasone) + - Fix malloc_conf overriding to work on Windows. (@jasone) + - Forcibly disable lazy-lock on Windows (was forcibly *enabled*). (@jasone) + +* 4.2.1 (June 8, 2016) + + Bug fixes: + - Fix bootstrapping issues for configurations that require allocation during + tsd initialization (e.g. --disable-tls). (@cferris1000, @jasone) + - Fix gettimeofday() version of nstime_update(). (@ronawho) + - Fix Valgrind regressions in calloc() and chunk_alloc_wrapper(). (@ronawho) + - Fix potential VM map fragmentation regression. (@jasone) + - Fix opt_zero-triggered in-place huge reallocation zeroing. (@jasone) + - Fix heap profiling context leaks in reallocation edge cases. (@jasone) + +* 4.2.0 (May 12, 2016) + + New features: + - Add the arena.<i>.reset mallctl, which makes it possible to discard all of + an arena's allocations in a single operation. (@jasone) + - Add the stats.retained and stats.arenas.<i>.retained statistics. (@jasone) + - Add the --with-version configure option. (@jasone) + - Support --with-lg-page values larger than actual page size. (@jasone) + + Optimizations: + - Use pairing heaps rather than red-black trees for various hot data + structures. (@djwatson, @jasone) + - Streamline fast paths of rtree operations. (@jasone) + - Optimize the fast paths of calloc() and [m,d,sd]allocx(). (@jasone) + - Decommit unused virtual memory if the OS does not overcommit. (@jasone) + - Specify MAP_NORESERVE on Linux if [heuristic] overcommit is active, in order + to avoid unfortunate interactions during fork(2). (@jasone) + + Bug fixes: + - Fix chunk accounting related to triggering gdump profiles. (@jasone) + - Link against librt for clock_gettime(2) if glibc < 2.17. (@jasone) + - Scale leak report summary according to sampling probability. (@jasone) + +* 4.1.1 (May 3, 2016) + + This bugfix release resolves a variety of mostly minor issues, though the + bitmap fix is critical for 64-bit Windows. + + Bug fixes: + - Fix the linear scan version of bitmap_sfu() to shift by the proper amount + even when sizeof(long) is not the same as sizeof(void *), as on 64-bit + Windows. (@jasone) + - Fix hashing functions to avoid unaligned memory accesses (and resulting + crashes). This is relevant at least to some ARM-based platforms. + (@rkmisra) + - Fix fork()-related lock rank ordering reversals. These reversals were + unlikely to cause deadlocks in practice except when heap profiling was + enabled and active. (@jasone) + - Fix various chunk leaks in OOM code paths. (@jasone) + - Fix malloc_stats_print() to print opt.narenas correctly. (@jasone) + - Fix MSVC-specific build/test issues. (@rustyx, @yuslepukhin) + - Fix a variety of test failures that were due to test fragility rather than + core bugs. (@jasone) + +* 4.1.0 (February 28, 2016) + + This release is primarily about optimizations, but it also incorporates a lot + of portability-motivated refactoring and enhancements. Many people worked on + this release, to an extent that even with the omission here of minor changes + (see git revision history), and of the people who reported and diagnosed + issues, so much of the work was contributed that starting with this release, + changes are annotated with author credits to help reflect the collaborative + effort involved. + + New features: + - Implement decay-based unused dirty page purging, a major optimization with + mallctl API impact. This is an alternative to the existing ratio-based + unused dirty page purging, and is intended to eventually become the sole + purging mechanism. New mallctls: + + opt.purge + + opt.decay_time + + arena.<i>.decay + + arena.<i>.decay_time + + arenas.decay_time + + stats.arenas.<i>.decay_time + (@jasone, @cevans87) + - Add --with-malloc-conf, which makes it possible to embed a default + options string during configuration. This was motivated by the desire to + specify --with-malloc-conf=purge:decay , since the default must remain + purge:ratio until the 5.0.0 release. (@jasone) + - Add MS Visual Studio 2015 support. (@rustyx, @yuslepukhin) + - Make *allocx() size class overflow behavior defined. The maximum + size class is now less than PTRDIFF_MAX to protect applications against + numerical overflow, and all allocation functions are guaranteed to indicate + errors rather than potentially crashing if the request size exceeds the + maximum size class. (@jasone) + - jeprof: + + Add raw heap profile support. (@jasone) + + Add --retain and --exclude for backtrace symbol filtering. (@jasone) + + Optimizations: + - Optimize the fast path to combine various bootstrapping and configuration + checks and execute more streamlined code in the common case. (@interwq) + - Use linear scan for small bitmaps (used for small object tracking). In + addition to speeding up bitmap operations on 64-bit systems, this reduces + allocator metadata overhead by approximately 0.2%. (@djwatson) + - Separate arena_avail trees, which substantially speeds up run tree + operations. (@djwatson) + - Use memoization (boot-time-computed table) for run quantization. Separate + arena_avail trees reduced the importance of this optimization. (@jasone) + - Attempt mmap-based in-place huge reallocation. This can dramatically speed + up incremental huge reallocation. (@jasone) + + Incompatible changes: + - Make opt.narenas unsigned rather than size_t. (@jasone) + + Bug fixes: + - Fix stats.cactive accounting regression. (@rustyx, @jasone) + - Handle unaligned keys in hash(). This caused problems for some ARM systems. + (@jasone, @cferris1000) + - Refactor arenas array. In addition to fixing a fork-related deadlock, this + makes arena lookups faster and simpler. (@jasone) + - Move retained memory allocation out of the default chunk allocation + function, to a location that gets executed even if the application installs + a custom chunk allocation function. This resolves a virtual memory leak. + (@buchgr) + - Fix a potential tsd cleanup leak. (@cferris1000, @jasone) + - Fix run quantization. In practice this bug had no impact unless + applications requested memory with alignment exceeding one page. + (@jasone, @djwatson) + - Fix LinuxThreads-specific bootstrapping deadlock. (Cosmin Paraschiv) + - jeprof: + + Don't discard curl options if timeout is not defined. (@djwatson) + + Detect failed profile fetches. (@djwatson) + - Fix stats.arenas.<i>.{dss,lg_dirty_mult,decay_time,pactive,pdirty} for + --disable-stats case. (@jasone) + +* 4.0.4 (October 24, 2015) + + This bugfix release fixes another xallocx() regression. No other regressions + have come to light in over a month, so this is likely a good starting point + for people who prefer to wait for "dot one" releases with all the major issues + shaken out. + + Bug fixes: + - Fix xallocx(..., MALLOCX_ZERO to zero the last full trailing page of large + allocations that have been randomly assigned an offset of 0 when + --enable-cache-oblivious configure option is enabled. + +* 4.0.3 (September 24, 2015) + + This bugfix release continues the trend of xallocx() and heap profiling fixes. + + Bug fixes: + - Fix xallocx(..., MALLOCX_ZERO) to zero all trailing bytes of large + allocations when --enable-cache-oblivious configure option is enabled. + - Fix xallocx(..., MALLOCX_ZERO) to zero trailing bytes of huge allocations + when resizing from/to a size class that is not a multiple of the chunk size. + - Fix prof_tctx_dump_iter() to filter out nodes that were created after heap + profile dumping started. + - Work around a potentially bad thread-specific data initialization + interaction with NPTL (glibc's pthreads implementation). + +* 4.0.2 (September 21, 2015) + + This bugfix release addresses a few bugs specific to heap profiling. + + Bug fixes: + - Fix ixallocx_prof_sample() to never modify nor create sampled small + allocations. xallocx() is in general incapable of moving small allocations, + so this fix removes buggy code without loss of generality. + - Fix irallocx_prof_sample() to always allocate large regions, even when + alignment is non-zero. + - Fix prof_alloc_rollback() to read tdata from thread-specific data rather + than dereferencing a potentially invalid tctx. + +* 4.0.1 (September 15, 2015) + + This is a bugfix release that is somewhat high risk due to the amount of + refactoring required to address deep xallocx() problems. As a side effect of + these fixes, xallocx() now tries harder to partially fulfill requests for + optional extra space. Note that a couple of minor heap profiling + optimizations are included, but these are better thought of as performance + fixes that were integral to discovering most of the other bugs. + + Optimizations: + - Avoid a chunk metadata read in arena_prof_tctx_set(), since it is in the + fast path when heap profiling is enabled. Additionally, split a special + case out into arena_prof_tctx_reset(), which also avoids chunk metadata + reads. + - Optimize irallocx_prof() to optimistically update the sampler state. The + prior implementation appears to have been a holdover from when + rallocx()/xallocx() functionality was combined as rallocm(). + + Bug fixes: + - Fix TLS configuration such that it is enabled by default for platforms on + which it works correctly. + - Fix arenas_cache_cleanup() and arena_get_hard() to handle + allocation/deallocation within the application's thread-specific data + cleanup functions even after arenas_cache is torn down. + - Fix xallocx() bugs related to size+extra exceeding HUGE_MAXCLASS. + - Fix chunk purge hook calls for in-place huge shrinking reallocation to + specify the old chunk size rather than the new chunk size. This bug caused + no correctness issues for the default chunk purge function, but was + visible to custom functions set via the "arena.<i>.chunk_hooks" mallctl. + - Fix heap profiling bugs: + + Fix heap profiling to distinguish among otherwise identical sample sites + with interposed resets (triggered via the "prof.reset" mallctl). This bug + could cause data structure corruption that would most likely result in a + segfault. + + Fix irealloc_prof() to prof_alloc_rollback() on OOM. + + Make one call to prof_active_get_unlocked() per allocation event, and use + the result throughout the relevant functions that handle an allocation + event. Also add a missing check in prof_realloc(). These fixes protect + allocation events against concurrent prof_active changes. + + Fix ixallocx_prof() to pass usize_max and zero to ixallocx_prof_sample() + in the correct order. + + Fix prof_realloc() to call prof_free_sampled_object() after calling + prof_malloc_sample_object(). Prior to this fix, if tctx and old_tctx were + the same, the tctx could have been prematurely destroyed. + - Fix portability bugs: + + Don't bitshift by negative amounts when encoding/decoding run sizes in + chunk header maps. This affected systems with page sizes greater than 8 + KiB. + + Rename index_t to szind_t to avoid an existing type on Solaris. + + Add JEMALLOC_CXX_THROW to the memalign() function prototype, in order to + match glibc and avoid compilation errors when including both + jemalloc/jemalloc.h and malloc.h in C++ code. + + Don't assume that /bin/sh is appropriate when running size_classes.sh + during configuration. + + Consider __sparcv9 a synonym for __sparc64__ when defining LG_QUANTUM. + + Link tests to librt if it contains clock_gettime(2). + +* 4.0.0 (August 17, 2015) + + This version contains many speed and space optimizations, both minor and + major. The major themes are generalization, unification, and simplification. + Although many of these optimizations cause no visible behavior change, their + cumulative effect is substantial. + + New features: + - Normalize size class spacing to be consistent across the complete size + range. By default there are four size classes per size doubling, but this + is now configurable via the --with-lg-size-class-group option. Also add the + --with-lg-page, --with-lg-page-sizes, --with-lg-quantum, and + --with-lg-tiny-min options, which can be used to tweak page and size class + settings. Impacts: + + Worst case performance for incrementally growing/shrinking reallocation + is improved because there are far fewer size classes, and therefore + copying happens less often. + + Internal fragmentation is limited to 20% for all but the smallest size + classes (those less than four times the quantum). (1B + 4 KiB) + and (1B + 4 MiB) previously suffered nearly 50% internal fragmentation. + + Chunk fragmentation tends to be lower because there are fewer distinct run + sizes to pack. + - Add support for explicit tcaches. The "tcache.create", "tcache.flush", and + "tcache.destroy" mallctls control tcache lifetime and flushing, and the + MALLOCX_TCACHE(tc) and MALLOCX_TCACHE_NONE flags to the *allocx() API + control which tcache is used for each operation. + - Implement per thread heap profiling, as well as the ability to + enable/disable heap profiling on a per thread basis. Add the "prof.reset", + "prof.lg_sample", "thread.prof.name", "thread.prof.active", + "opt.prof_thread_active_init", "prof.thread_active_init", and + "thread.prof.active" mallctls. + - Add support for per arena application-specified chunk allocators, configured + via the "arena.<i>.chunk_hooks" mallctl. + - Refactor huge allocation to be managed by arenas, so that arenas now + function as general purpose independent allocators. This is important in + the context of user-specified chunk allocators, aside from the scalability + benefits. Related new statistics: + + The "stats.arenas.<i>.huge.allocated", "stats.arenas.<i>.huge.nmalloc", + "stats.arenas.<i>.huge.ndalloc", and "stats.arenas.<i>.huge.nrequests" + mallctls provide high level per arena huge allocation statistics. + + The "arenas.nhchunks", "arenas.hchunk.<i>.size", + "stats.arenas.<i>.hchunks.<j>.nmalloc", + "stats.arenas.<i>.hchunks.<j>.ndalloc", + "stats.arenas.<i>.hchunks.<j>.nrequests", and + "stats.arenas.<i>.hchunks.<j>.curhchunks" mallctls provide per size class + statistics. + - Add the 'util' column to malloc_stats_print() output, which reports the + proportion of available regions that are currently in use for each small + size class. + - Add "alloc" and "free" modes for for junk filling (see the "opt.junk" + mallctl), so that it is possible to separately enable junk filling for + allocation versus deallocation. + - Add the jemalloc-config script, which provides information about how + jemalloc was configured, and how to integrate it into application builds. + - Add metadata statistics, which are accessible via the "stats.metadata", + "stats.arenas.<i>.metadata.mapped", and + "stats.arenas.<i>.metadata.allocated" mallctls. + - Add the "stats.resident" mallctl, which reports the upper limit of + physically resident memory mapped by the allocator. + - Add per arena control over unused dirty page purging, via the + "arenas.lg_dirty_mult", "arena.<i>.lg_dirty_mult", and + "stats.arenas.<i>.lg_dirty_mult" mallctls. + - Add the "prof.gdump" mallctl, which makes it possible to toggle the gdump + feature on/off during program execution. + - Add sdallocx(), which implements sized deallocation. The primary + optimization over dallocx() is the removal of a metadata read, which often + suffers an L1 cache miss. + - Add missing header includes in jemalloc/jemalloc.h, so that applications + only have to #include <jemalloc/jemalloc.h>. + - Add support for additional platforms: + + Bitrig + + Cygwin + + DragonFlyBSD + + iOS + + OpenBSD + + OpenRISC/or1k + + Optimizations: + - Maintain dirty runs in per arena LRUs rather than in per arena trees of + dirty-run-containing chunks. In practice this change significantly reduces + dirty page purging volume. + - Integrate whole chunks into the unused dirty page purging machinery. This + reduces the cost of repeated huge allocation/deallocation, because it + effectively introduces a cache of chunks. + - Split the arena chunk map into two separate arrays, in order to increase + cache locality for the frequently accessed bits. + - Move small run metadata out of runs, into arena chunk headers. This reduces + run fragmentation, smaller runs reduce external fragmentation for small size + classes, and packed (less uniformly aligned) metadata layout improves CPU + cache set distribution. + - Randomly distribute large allocation base pointer alignment relative to page + boundaries in order to more uniformly utilize CPU cache sets. This can be + disabled via the --disable-cache-oblivious configure option, and queried via + the "config.cache_oblivious" mallctl. + - Micro-optimize the fast paths for the public API functions. + - Refactor thread-specific data to reside in a single structure. This assures + that only a single TLS read is necessary per call into the public API. + - Implement in-place huge allocation growing and shrinking. + - Refactor rtree (radix tree for chunk lookups) to be lock-free, and make + additional optimizations that reduce maximum lookup depth to one or two + levels. This resolves what was a concurrency bottleneck for per arena huge + allocation, because a global data structure is critical for determining + which arenas own which huge allocations. + + Incompatible changes: + - Replace --enable-cc-silence with --disable-cc-silence to suppress spurious + warnings by default. + - Assure that the constness of malloc_usable_size()'s return type matches that + of the system implementation. + - Change the heap profile dump format to support per thread heap profiling, + rename pprof to jeprof, and enhance it with the --thread=<n> option. As a + result, the bundled jeprof must now be used rather than the upstream + (gperftools) pprof. + - Disable "opt.prof_final" by default, in order to avoid atexit(3), which can + internally deadlock on some platforms. + - Change the "arenas.nlruns" mallctl type from size_t to unsigned. + - Replace the "stats.arenas.<i>.bins.<j>.allocated" mallctl with + "stats.arenas.<i>.bins.<j>.curregs". + - Ignore MALLOC_CONF in set{uid,gid,cap} binaries. + - Ignore MALLOCX_ARENA(a) in dallocx(), in favor of using the + MALLOCX_TCACHE(tc) and MALLOCX_TCACHE_NONE flags to control tcache usage. + + Removed features: + - Remove the *allocm() API, which is superseded by the *allocx() API. + - Remove the --enable-dss options, and make dss non-optional on all platforms + which support sbrk(2). + - Remove the "arenas.purge" mallctl, which was obsoleted by the + "arena.<i>.purge" mallctl in 3.1.0. + - Remove the unnecessary "opt.valgrind" mallctl; jemalloc automatically + detects whether it is running inside Valgrind. + - Remove the "stats.huge.allocated", "stats.huge.nmalloc", and + "stats.huge.ndalloc" mallctls. + - Remove the --enable-mremap option. + - Remove the "stats.chunks.current", "stats.chunks.total", and + "stats.chunks.high" mallctls. + + Bug fixes: + - Fix the cactive statistic to decrease (rather than increase) when active + memory decreases. This regression was first released in 3.5.0. + - Fix OOM handling in memalign() and valloc(). A variant of this bug existed + in all releases since 2.0.0, which introduced these functions. + - Fix an OOM-related regression in arena_tcache_fill_small(), which could + cause cache corruption on OOM. This regression was present in all releases + from 2.2.0 through 3.6.0. + - Fix size class overflow handling for malloc(), posix_memalign(), memalign(), + calloc(), and realloc() when profiling is enabled. + - Fix the "arena.<i>.dss" mallctl to return an error if "primary" or + "secondary" precedence is specified, but sbrk(2) is not supported. + - Fix fallback lg_floor() implementations to handle extremely large inputs. + - Ensure the default purgeable zone is after the default zone on OS X. + - Fix latent bugs in atomic_*(). + - Fix the "arena.<i>.dss" mallctl to handle read-only calls. + - Fix tls_model configuration to enable the initial-exec model when possible. + - Mark malloc_conf as a weak symbol so that the application can override it. + - Correctly detect glibc's adaptive pthread mutexes. + - Fix the --without-export configure option. + +* 3.6.0 (March 31, 2014) + + This version contains a critical bug fix for a regression present in 3.5.0 and + 3.5.1. + + Bug fixes: + - Fix a regression in arena_chunk_alloc() that caused crashes during + small/large allocation if chunk allocation failed. In the absence of this + bug, chunk allocation failure would result in allocation failure, e.g. NULL + return from malloc(). This regression was introduced in 3.5.0. + - Fix backtracing for gcc intrinsics-based backtracing by specifying + -fno-omit-frame-pointer to gcc. Note that the application (and all the + libraries it links to) must also be compiled with this option for + backtracing to be reliable. + - Use dss allocation precedence for huge allocations as well as small/large + allocations. + - Fix test assertion failure message formatting. This bug did not manifest on + x86_64 systems because of implementation subtleties in va_list. + - Fix inconsequential test failures for hash and SFMT code. + + New features: + - Support heap profiling on FreeBSD. This feature depends on the proc + filesystem being mounted during heap profile dumping. + +* 3.5.1 (February 25, 2014) + + This version primarily addresses minor bugs in test code. + + Bug fixes: + - Configure Solaris/Illumos to use MADV_FREE. + - Fix junk filling for mremap(2)-based huge reallocation. This is only + relevant if configuring with the --enable-mremap option specified. + - Avoid compilation failure if 'restrict' C99 keyword is not supported by the + compiler. + - Add a configure test for SSE2 rather than assuming it is usable on i686 + systems. This fixes test compilation errors, especially on 32-bit Linux + systems. + - Fix mallctl argument size mismatches (size_t vs. uint64_t) in the stats unit + test. + - Fix/remove flawed alignment-related overflow tests. + - Prevent compiler optimizations that could change backtraces in the + prof_accum unit test. + +* 3.5.0 (January 22, 2014) + + This version focuses on refactoring and automated testing, though it also + includes some non-trivial heap profiling optimizations not mentioned below. + + New features: + - Add the *allocx() API, which is a successor to the experimental *allocm() + API. The *allocx() functions are slightly simpler to use because they have + fewer parameters, they directly return the results of primary interest, and + mallocx()/rallocx() avoid the strict aliasing pitfall that + allocm()/rallocm() share with posix_memalign(). Note that *allocm() is + slated for removal in the next non-bugfix release. + - Add support for LinuxThreads. + + Bug fixes: + - Unless heap profiling is enabled, disable floating point code and don't link + with libm. This, in combination with e.g. EXTRA_CFLAGS=-mno-sse on x64 + systems, makes it possible to completely disable floating point register + use. Some versions of glibc neglect to save/restore caller-saved floating + point registers during dynamic lazy symbol loading, and the symbol loading + code uses whatever malloc the application happens to have linked/loaded + with, the result being potential floating point register corruption. + - Report ENOMEM rather than EINVAL if an OOM occurs during heap profiling + backtrace creation in imemalign(). This bug impacted posix_memalign() and + aligned_alloc(). + - Fix a file descriptor leak in a prof_dump_maps() error path. + - Fix prof_dump() to close the dump file descriptor for all relevant error + paths. + - Fix rallocm() to use the arena specified by the ALLOCM_ARENA(s) flag for + allocation, not just deallocation. + - Fix a data race for large allocation stats counters. + - Fix a potential infinite loop during thread exit. This bug occurred on + Solaris, and could affect other platforms with similar pthreads TSD + implementations. + - Don't junk-fill reallocations unless usable size changes. This fixes a + violation of the *allocx()/*allocm() semantics. + - Fix growing large reallocation to junk fill new space. + - Fix huge deallocation to junk fill when munmap is disabled. + - Change the default private namespace prefix from empty to je_, and change + --with-private-namespace-prefix so that it prepends an additional prefix + rather than replacing je_. This reduces the likelihood of applications + which statically link jemalloc experiencing symbol name collisions. + - Add missing private namespace mangling (relevant when + --with-private-namespace is specified). + - Add and use JEMALLOC_INLINE_C so that static inline functions are marked as + static even for debug builds. + - Add a missing mutex unlock in a malloc_init_hard() error path. In practice + this error path is never executed. + - Fix numerous bugs in malloc_strotumax() error handling/reporting. These + bugs had no impact except for malformed inputs. + - Fix numerous bugs in malloc_snprintf(). These bugs were not exercised by + existing calls, so they had no impact. + +* 3.4.1 (October 20, 2013) + + Bug fixes: + - Fix a race in the "arenas.extend" mallctl that could cause memory corruption + of internal data structures and subsequent crashes. + - Fix Valgrind integration flaws that caused Valgrind warnings about reads of + uninitialized memory in: + + arena chunk headers + + internal zero-initialized data structures (relevant to tcache and prof + code) + - Preserve errno during the first allocation. A readlink(2) call during + initialization fails unless /etc/malloc.conf exists, so errno was typically + set during the first allocation prior to this fix. + - Fix compilation warnings reported by gcc 4.8.1. + +* 3.4.0 (June 2, 2013) + + This version is essentially a small bugfix release, but the addition of + aarch64 support requires that the minor version be incremented. + + Bug fixes: + - Fix race-triggered deadlocks in chunk_record(). These deadlocks were + typically triggered by multiple threads concurrently deallocating huge + objects. + + New features: + - Add support for the aarch64 architecture. + +* 3.3.1 (March 6, 2013) + + This version fixes bugs that are typically encountered only when utilizing + custom run-time options. + + Bug fixes: + - Fix a locking order bug that could cause deadlock during fork if heap + profiling were enabled. + - Fix a chunk recycling bug that could cause the allocator to lose track of + whether a chunk was zeroed. On FreeBSD, NetBSD, and OS X, it could cause + corruption if allocating via sbrk(2) (unlikely unless running with the + "dss:primary" option specified). This was completely harmless on Linux + unless using mlockall(2) (and unlikely even then, unless the + --disable-munmap configure option or the "dss:primary" option was + specified). This regression was introduced in 3.1.0 by the + mlockall(2)/madvise(2) interaction fix. + - Fix TLS-related memory corruption that could occur during thread exit if the + thread never allocated memory. Only the quarantine and prof facilities were + susceptible. + - Fix two quarantine bugs: + + Internal reallocation of the quarantined object array leaked the old + array. + + Reallocation failure for internal reallocation of the quarantined object + array (very unlikely) resulted in memory corruption. + - Fix Valgrind integration to annotate all internally allocated memory in a + way that keeps Valgrind happy about internal data structure access. + - Fix building for s390 systems. + +* 3.3.0 (January 23, 2013) + + This version includes a few minor performance improvements in addition to the + listed new features and bug fixes. + + New features: + - Add clipping support to lg_chunk option processing. + - Add the --enable-ivsalloc option. + - Add the --without-export option. + - Add the --disable-zone-allocator option. + + Bug fixes: + - Fix "arenas.extend" mallctl to output the number of arenas. + - Fix chunk_recycle() to unconditionally inform Valgrind that returned memory + is undefined. + - Fix build break on FreeBSD related to alloca.h. + +* 3.2.0 (November 9, 2012) + + In addition to a couple of bug fixes, this version modifies page run + allocation and dirty page purging algorithms in order to better control + page-level virtual memory fragmentation. + + Incompatible changes: + - Change the "opt.lg_dirty_mult" default from 5 to 3 (32:1 to 8:1). + + Bug fixes: + - Fix dss/mmap allocation precedence code to use recyclable mmap memory only + after primary dss allocation fails. + - Fix deadlock in the "arenas.purge" mallctl. This regression was introduced + in 3.1.0 by the addition of the "arena.<i>.purge" mallctl. + +* 3.1.0 (October 16, 2012) + + New features: + - Auto-detect whether running inside Valgrind, thus removing the need to + manually specify MALLOC_CONF=valgrind:true. + - Add the "arenas.extend" mallctl, which allows applications to create + manually managed arenas. + - Add the ALLOCM_ARENA() flag for {,r,d}allocm(). + - Add the "opt.dss", "arena.<i>.dss", and "stats.arenas.<i>.dss" mallctls, + which provide control over dss/mmap precedence. + - Add the "arena.<i>.purge" mallctl, which obsoletes "arenas.purge". + - Define LG_QUANTUM for hppa. + + Incompatible changes: + - Disable tcache by default if running inside Valgrind, in order to avoid + making unallocated objects appear reachable to Valgrind. + - Drop const from malloc_usable_size() argument on Linux. + + Bug fixes: + - Fix heap profiling crash if sampled object is freed via realloc(p, 0). + - Remove const from __*_hook variable declarations, so that glibc can modify + them during process forking. + - Fix mlockall(2)/madvise(2) interaction. + - Fix fork(2)-related deadlocks. + - Fix error return value for "thread.tcache.enabled" mallctl. + +* 3.0.0 (May 11, 2012) + + Although this version adds some major new features, the primary focus is on + internal code cleanup that facilitates maintainability and portability, most + of which is not reflected in the ChangeLog. This is the first release to + incorporate substantial contributions from numerous other developers, and the + result is a more broadly useful allocator (see the git revision history for + contribution details). Note that the license has been unified, thanks to + Facebook granting a license under the same terms as the other copyright + holders (see COPYING). + + New features: + - Implement Valgrind support, redzones, and quarantine. + - Add support for additional platforms: + + FreeBSD + + Mac OS X Lion + + MinGW + + Windows (no support yet for replacing the system malloc) + - Add support for additional architectures: + + MIPS + + SH4 + + Tilera + - Add support for cross compiling. + - Add nallocm(), which rounds a request size up to the nearest size class + without actually allocating. + - Implement aligned_alloc() (blame C11). + - Add the "thread.tcache.enabled" mallctl. + - Add the "opt.prof_final" mallctl. + - Update pprof (from gperftools 2.0). + - Add the --with-mangling option. + - Add the --disable-experimental option. + - Add the --disable-munmap option, and make it the default on Linux. + - Add the --enable-mremap option, which disables use of mremap(2) by default. + + Incompatible changes: + - Enable stats by default. + - Enable fill by default. + - Disable lazy locking by default. + - Rename the "tcache.flush" mallctl to "thread.tcache.flush". + - Rename the "arenas.pagesize" mallctl to "arenas.page". + - Change the "opt.lg_prof_sample" default from 0 to 19 (1 B to 512 KiB). + - Change the "opt.prof_accum" default from true to false. + + Removed features: + - Remove the swap feature, including the "config.swap", "swap.avail", + "swap.prezeroed", "swap.nfds", and "swap.fds" mallctls. + - Remove highruns statistics, including the + "stats.arenas.<i>.bins.<j>.highruns" and + "stats.arenas.<i>.lruns.<j>.highruns" mallctls. + - As part of small size class refactoring, remove the "opt.lg_[qc]space_max", + "arenas.cacheline", "arenas.subpage", "arenas.[tqcs]space_{min,max}", and + "arenas.[tqcs]bins" mallctls. + - Remove the "arenas.chunksize" mallctl. + - Remove the "opt.lg_prof_tcmax" option. + - Remove the "opt.lg_prof_bt_max" option. + - Remove the "opt.lg_tcache_gc_sweep" option. + - Remove the --disable-tiny option, including the "config.tiny" mallctl. + - Remove the --enable-dynamic-page-shift configure option. + - Remove the --enable-sysv configure option. + + Bug fixes: + - Fix a statistics-related bug in the "thread.arena" mallctl that could cause + invalid statistics and crashes. + - Work around TLS deallocation via free() on Linux. This bug could cause + write-after-free memory corruption. + - Fix a potential deadlock that could occur during interval- and + growth-triggered heap profile dumps. + - Fix large calloc() zeroing bugs due to dropping chunk map unzeroed flags. + - Fix chunk_alloc_dss() to stop claiming memory is zeroed. This bug could + cause memory corruption and crashes with --enable-dss specified. + - Fix fork-related bugs that could cause deadlock in children between fork + and exec. + - Fix malloc_stats_print() to honor 'b' and 'l' in the opts parameter. + - Fix realloc(p, 0) to act like free(p). + - Do not enforce minimum alignment in memalign(). + - Check for NULL pointer in malloc_usable_size(). + - Fix an off-by-one heap profile statistics bug that could be observed in + interval- and growth-triggered heap profiles. + - Fix the "epoch" mallctl to update cached stats even if the passed in epoch + is 0. + - Fix bin->runcur management to fix a layout policy bug. This bug did not + affect correctness. + - Fix a bug in choose_arena_hard() that potentially caused more arenas to be + initialized than necessary. + - Add missing "opt.lg_tcache_max" mallctl implementation. + - Use glibc allocator hooks to make mixed allocator usage less likely. + - Fix build issues for --disable-tcache. + - Don't mangle pthread_create() when --with-private-namespace is specified. + +* 2.2.5 (November 14, 2011) + + Bug fixes: + - Fix huge_ralloc() race when using mremap(2). This is a serious bug that + could cause memory corruption and/or crashes. + - Fix huge_ralloc() to maintain chunk statistics. + - Fix malloc_stats_print(..., "a") output. + +* 2.2.4 (November 5, 2011) + + Bug fixes: + - Initialize arenas_tsd before using it. This bug existed for 2.2.[0-3], as + well as for --disable-tls builds in earlier releases. + - Do not assume a 4 KiB page size in test/rallocm.c. + +* 2.2.3 (August 31, 2011) + + This version fixes numerous bugs related to heap profiling. + + Bug fixes: + - Fix a prof-related race condition. This bug could cause memory corruption, + but only occurred in non-default configurations (prof_accum:false). + - Fix off-by-one backtracing issues (make sure that prof_alloc_prep() is + excluded from backtraces). + - Fix a prof-related bug in realloc() (only triggered by OOM errors). + - Fix prof-related bugs in allocm() and rallocm(). + - Fix prof_tdata_cleanup() for --disable-tls builds. + - Fix a relative include path, to fix objdir builds. + +* 2.2.2 (July 30, 2011) + + Bug fixes: + - Fix a build error for --disable-tcache. + - Fix assertions in arena_purge() (for real this time). + - Add the --with-private-namespace option. This is a workaround for symbol + conflicts that can inadvertently arise when using static libraries. + +* 2.2.1 (March 30, 2011) + + Bug fixes: + - Implement atomic operations for x86/x64. This fixes compilation failures + for versions of gcc that are still in wide use. + - Fix an assertion in arena_purge(). + +* 2.2.0 (March 22, 2011) + + This version incorporates several improvements to algorithms and data + structures that tend to reduce fragmentation and increase speed. + + New features: + - Add the "stats.cactive" mallctl. + - Update pprof (from google-perftools 1.7). + - Improve backtracing-related configuration logic, and add the + --disable-prof-libgcc option. + + Bug fixes: + - Change default symbol visibility from "internal", to "hidden", which + decreases the overhead of library-internal function calls. + - Fix symbol visibility so that it is also set on OS X. + - Fix a build dependency regression caused by the introduction of the .pic.o + suffix for PIC object files. + - Add missing checks for mutex initialization failures. + - Don't use libgcc-based backtracing except on x64, where it is known to work. + - Fix deadlocks on OS X that were due to memory allocation in + pthread_mutex_lock(). + - Heap profiling-specific fixes: + + Fix memory corruption due to integer overflow in small region index + computation, when using a small enough sample interval that profiling + context pointers are stored in small run headers. + + Fix a bootstrap ordering bug that only occurred with TLS disabled. + + Fix a rallocm() rsize bug. + + Fix error detection bugs for aligned memory allocation. + +* 2.1.3 (March 14, 2011) + + Bug fixes: + - Fix a cpp logic regression (due to the "thread.{de,}allocatedp" mallctl fix + for OS X in 2.1.2). + - Fix a "thread.arena" mallctl bug. + - Fix a thread cache stats merging bug. + +* 2.1.2 (March 2, 2011) + + Bug fixes: + - Fix "thread.{de,}allocatedp" mallctl for OS X. + - Add missing jemalloc.a to build system. + +* 2.1.1 (January 31, 2011) + + Bug fixes: + - Fix aligned huge reallocation (affected allocm()). + - Fix the ALLOCM_LG_ALIGN macro definition. + - Fix a heap dumping deadlock. + - Fix a "thread.arena" mallctl bug. + +* 2.1.0 (December 3, 2010) + + This version incorporates some optimizations that can't quite be considered + bug fixes. + + New features: + - Use Linux's mremap(2) for huge object reallocation when possible. + - Avoid locking in mallctl*() when possible. + - Add the "thread.[de]allocatedp" mallctl's. + - Convert the manual page source from roff to DocBook, and generate both roff + and HTML manuals. + + Bug fixes: + - Fix a crash due to incorrect bootstrap ordering. This only impacted + --enable-debug --enable-dss configurations. + - Fix a minor statistics bug for mallctl("swap.avail", ...). + +* 2.0.1 (October 29, 2010) + + Bug fixes: + - Fix a race condition in heap profiling that could cause undefined behavior + if "opt.prof_accum" were disabled. + - Add missing mutex unlocks for some OOM error paths in the heap profiling + code. + - Fix a compilation error for non-C99 builds. + +* 2.0.0 (October 24, 2010) + + This version focuses on the experimental *allocm() API, and on improved + run-time configuration/introspection. Nonetheless, numerous performance + improvements are also included. + + New features: + - Implement the experimental {,r,s,d}allocm() API, which provides a superset + of the functionality available via malloc(), calloc(), posix_memalign(), + realloc(), malloc_usable_size(), and free(). These functions can be used to + allocate/reallocate aligned zeroed memory, ask for optional extra memory + during reallocation, prevent object movement during reallocation, etc. + - Replace JEMALLOC_OPTIONS/JEMALLOC_PROF_PREFIX with MALLOC_CONF, which is + more human-readable, and more flexible. For example: + JEMALLOC_OPTIONS=AJP + is now: + MALLOC_CONF=abort:true,fill:true,stats_print:true + - Port to Apple OS X. Sponsored by Mozilla. + - Make it possible for the application to control thread-->arena mappings via + the "thread.arena" mallctl. + - Add compile-time support for all TLS-related functionality via pthreads TSD. + This is mainly of interest for OS X, which does not support TLS, but has a + TSD implementation with similar performance. + - Override memalign() and valloc() if they are provided by the system. + - Add the "arenas.purge" mallctl, which can be used to synchronously purge all + dirty unused pages. + - Make cumulative heap profiling data optional, so that it is possible to + limit the amount of memory consumed by heap profiling data structures. + - Add per thread allocation counters that can be accessed via the + "thread.allocated" and "thread.deallocated" mallctls. + + Incompatible changes: + - Remove JEMALLOC_OPTIONS and malloc_options (see MALLOC_CONF above). + - Increase default backtrace depth from 4 to 128 for heap profiling. + - Disable interval-based profile dumps by default. + + Bug fixes: + - Remove bad assertions in fork handler functions. These assertions could + cause aborts for some combinations of configure settings. + - Fix strerror_r() usage to deal with non-standard semantics in GNU libc. + - Fix leak context reporting. This bug tended to cause the number of contexts + to be underreported (though the reported number of objects and bytes were + correct). + - Fix a realloc() bug for large in-place growing reallocation. This bug could + cause memory corruption, but it was hard to trigger. + - Fix an allocation bug for small allocations that could be triggered if + multiple threads raced to create a new run of backing pages. + - Enhance the heap profiler to trigger samples based on usable size, rather + than request size. + - Fix a heap profiling bug due to sometimes losing track of requested object + size for sampled objects. + +* 1.0.3 (August 12, 2010) + + Bug fixes: + - Fix the libunwind-based implementation of stack backtracing (used for heap + profiling). This bug could cause zero-length backtraces to be reported. + - Add a missing mutex unlock in library initialization code. If multiple + threads raced to initialize malloc, some of them could end up permanently + blocked. + +* 1.0.2 (May 11, 2010) + + Bug fixes: + - Fix junk filling of large objects, which could cause memory corruption. + - Add MAP_NORESERVE support for chunk mapping, because otherwise virtual + memory limits could cause swap file configuration to fail. Contributed by + Jordan DeLong. + +* 1.0.1 (April 14, 2010) + + Bug fixes: + - Fix compilation when --enable-fill is specified. + - Fix threads-related profiling bugs that affected accuracy and caused memory + to be leaked during thread exit. + - Fix dirty page purging race conditions that could cause crashes. + - Fix crash in tcache flushing code during thread destruction. + +* 1.0.0 (April 11, 2010) + + This release focuses on speed and run-time introspection. Numerous + algorithmic improvements make this release substantially faster than its + predecessors. + + New features: + - Implement autoconf-based configuration system. + - Add mallctl*(), for the purposes of introspection and run-time + configuration. + - Make it possible for the application to manually flush a thread's cache, via + the "tcache.flush" mallctl. + - Base maximum dirty page count on proportion of active memory. + - Compute various additional run-time statistics, including per size class + statistics for large objects. + - Expose malloc_stats_print(), which can be called repeatedly by the + application. + - Simplify the malloc_message() signature to only take one string argument, + and incorporate an opaque data pointer argument for use by the application + in combination with malloc_stats_print(). + - Add support for allocation backed by one or more swap files, and allow the + application to disable over-commit if swap files are in use. + - Implement allocation profiling and leak checking. + + Removed features: + - Remove the dynamic arena rebalancing code, since thread-specific caching + reduces its utility. + + Bug fixes: + - Modify chunk allocation to work when address space layout randomization + (ASLR) is in use. + - Fix thread cleanup bugs related to TLS destruction. + - Handle 0-size allocation requests in posix_memalign(). + - Fix a chunk leak. The leaked chunks were never touched, so this impacted + virtual memory usage, but not physical memory usage. + +* linux_2008082[78]a (August 27/28, 2008) + + These snapshot releases are the simple result of incorporating Linux-specific + support into the FreeBSD malloc sources. + +-------------------------------------------------------------------------------- +vim:filetype=text:textwidth=80 diff --git a/deps/jemalloc/INSTALL.md b/deps/jemalloc/INSTALL.md new file mode 100644 index 000000000..b8f729b0d --- /dev/null +++ b/deps/jemalloc/INSTALL.md @@ -0,0 +1,421 @@ +Building and installing a packaged release of jemalloc can be as simple as +typing the following while in the root directory of the source tree: + + ./configure + make + make install + +If building from unpackaged developer sources, the simplest command sequence +that might work is: + + ./autogen.sh + make dist + make + make install + +Note that documentation is not built by the default target because doing so +would create a dependency on xsltproc in packaged releases, hence the +requirement to either run 'make dist' or avoid installing docs via the various +install_* targets documented below. + + +## Advanced configuration + +The 'configure' script supports numerous options that allow control of which +functionality is enabled, where jemalloc is installed, etc. Optionally, pass +any of the following arguments (not a definitive list) to 'configure': + +* `--help` + + Print a definitive list of options. + +* `--prefix=<install-root-dir>` + + Set the base directory in which to install. For example: + + ./configure --prefix=/usr/local + + will cause files to be installed into /usr/local/include, /usr/local/lib, + and /usr/local/man. + +* `--with-version=(<major>.<minor>.<bugfix>-<nrev>-g<gid>|VERSION)` + + The VERSION file is mandatory for successful configuration, and the + following steps are taken to assure its presence: + 1) If --with-version=<major>.<minor>.<bugfix>-<nrev>-g<gid> is specified, + generate VERSION using the specified value. + 2) If --with-version is not specified in either form and the source + directory is inside a git repository, try to generate VERSION via 'git + describe' invocations that pattern-match release tags. + 3) If VERSION is missing, generate it with a bogus version: + 0.0.0-0-g0000000000000000000000000000000000000000 + + Note that --with-version=VERSION bypasses (1) and (2), which simplifies + VERSION configuration when embedding a jemalloc release into another + project's git repository. + +* `--with-rpath=<colon-separated-rpath>` + + Embed one or more library paths, so that libjemalloc can find the libraries + it is linked to. This works only on ELF-based systems. + +* `--with-mangling=<map>` + + Mangle public symbols specified in <map> which is a comma-separated list of + name:mangled pairs. + + For example, to use ld's --wrap option as an alternative method for + overriding libc's malloc implementation, specify something like: + + --with-mangling=malloc:__wrap_malloc,free:__wrap_free[...] + + Note that mangling happens prior to application of the prefix specified by + --with-jemalloc-prefix, and mangled symbols are then ignored when applying + the prefix. + +* `--with-jemalloc-prefix=<prefix>` + + Prefix all public APIs with <prefix>. For example, if <prefix> is + "prefix_", API changes like the following occur: + + malloc() --> prefix_malloc() + malloc_conf --> prefix_malloc_conf + /etc/malloc.conf --> /etc/prefix_malloc.conf + MALLOC_CONF --> PREFIX_MALLOC_CONF + + This makes it possible to use jemalloc at the same time as the system + allocator, or even to use multiple copies of jemalloc simultaneously. + + By default, the prefix is "", except on OS X, where it is "je_". On OS X, + jemalloc overlays the default malloc zone, but makes no attempt to actually + replace the "malloc", "calloc", etc. symbols. + +* `--without-export` + + Don't export public APIs. This can be useful when building jemalloc as a + static library, or to avoid exporting public APIs when using the zone + allocator on OSX. + +* `--with-private-namespace=<prefix>` + + Prefix all library-private APIs with <prefix>je_. For shared libraries, + symbol visibility mechanisms prevent these symbols from being exported, but + for static libraries, naming collisions are a real possibility. By + default, <prefix> is empty, which results in a symbol prefix of je_ . + +* `--with-install-suffix=<suffix>` + + Append <suffix> to the base name of all installed files, such that multiple + versions of jemalloc can coexist in the same installation directory. For + example, libjemalloc.so.0 becomes libjemalloc<suffix>.so.0. + +* `--with-malloc-conf=<malloc_conf>` + + Embed `<malloc_conf>` as a run-time options string that is processed prior to + the malloc_conf global variable, the /etc/malloc.conf symlink, and the + MALLOC_CONF environment variable. For example, to change the default decay + time to 30 seconds: + + --with-malloc-conf=decay_ms:30000 + +* `--enable-debug` + + Enable assertions and validation code. This incurs a substantial + performance hit, but is very useful during application development. + +* `--disable-stats` + + Disable statistics gathering functionality. See the "opt.stats_print" + option documentation for usage details. + +* `--enable-prof` + + Enable heap profiling and leak detection functionality. See the "opt.prof" + option documentation for usage details. When enabled, there are several + approaches to backtracing, and the configure script chooses the first one + in the following list that appears to function correctly: + + + libunwind (requires --enable-prof-libunwind) + + libgcc (unless --disable-prof-libgcc) + + gcc intrinsics (unless --disable-prof-gcc) + +* `--enable-prof-libunwind` + + Use the libunwind library (http://www.nongnu.org/libunwind/) for stack + backtracing. + +* `--disable-prof-libgcc` + + Disable the use of libgcc's backtracing functionality. + +* `--disable-prof-gcc` + + Disable the use of gcc intrinsics for backtracing. + +* `--with-static-libunwind=<libunwind.a>` + + Statically link against the specified libunwind.a rather than dynamically + linking with -lunwind. + +* `--disable-fill` + + Disable support for junk/zero filling of memory. See the "opt.junk" and + "opt.zero" option documentation for usage details. + +* `--disable-zone-allocator` + + Disable zone allocator for Darwin. This means jemalloc won't be hooked as + the default allocator on OSX/iOS. + +* `--enable-utrace` + + Enable utrace(2)-based allocation tracing. This feature is not broadly + portable (FreeBSD has it, but Linux and OS X do not). + +* `--enable-xmalloc` + + Enable support for optional immediate termination due to out-of-memory + errors, as is commonly implemented by "xmalloc" wrapper function for malloc. + See the "opt.xmalloc" option documentation for usage details. + +* `--enable-lazy-lock` + + Enable code that wraps pthread_create() to detect when an application + switches from single-threaded to multi-threaded mode, so that it can avoid + mutex locking/unlocking operations while in single-threaded mode. In + practice, this feature usually has little impact on performance unless + thread-specific caching is disabled. + +* `--disable-cache-oblivious` + + Disable cache-oblivious large allocation alignment for large allocation + requests with no alignment constraints. If this feature is disabled, all + large allocations are page-aligned as an implementation artifact, which can + severely harm CPU cache utilization. However, the cache-oblivious layout + comes at the cost of one extra page per large allocation, which in the + most extreme case increases physical memory usage for the 16 KiB size class + to 20 KiB. + +* `--disable-syscall` + + Disable use of syscall(2) rather than {open,read,write,close}(2). This is + intended as a workaround for systems that place security limitations on + syscall(2). + +* `--disable-cxx` + + Disable C++ integration. This will cause new and delete operator + implementations to be omitted. + +* `--with-xslroot=<path>` + + Specify where to find DocBook XSL stylesheets when building the + documentation. + +* `--with-lg-page=<lg-page>` + + Specify the base 2 log of the allocator page size, which must in turn be at + least as large as the system page size. By default the configure script + determines the host's page size and sets the allocator page size equal to + the system page size, so this option need not be specified unless the + system page size may change between configuration and execution, e.g. when + cross compiling. + +* `--with-lg-hugepage=<lg-hugepage>` + + Specify the base 2 log of the system huge page size. This option is useful + when cross compiling, or when overriding the default for systems that do + not explicitly support huge pages. + +* `--with-lg-quantum=<lg-quantum>` + + Specify the base 2 log of the minimum allocation alignment. jemalloc needs + to know the minimum alignment that meets the following C standard + requirement (quoted from the April 12, 2011 draft of the C11 standard): + + > The pointer returned if the allocation succeeds is suitably aligned so + that it may be assigned to a pointer to any type of object with a + fundamental alignment requirement and then used to access such an object + or an array of such objects in the space allocated [...] + + This setting is architecture-specific, and although jemalloc includes known + safe values for the most commonly used modern architectures, there is a + wrinkle related to GNU libc (glibc) that may impact your choice of + <lg-quantum>. On most modern architectures, this mandates 16-byte + alignment (<lg-quantum>=4), but the glibc developers chose not to meet this + requirement for performance reasons. An old discussion can be found at + <https://sourceware.org/bugzilla/show_bug.cgi?id=206> . Unlike glibc, + jemalloc does follow the C standard by default (caveat: jemalloc + technically cheats for size classes smaller than the quantum), but the fact + that Linux systems already work around this allocator noncompliance means + that it is generally safe in practice to let jemalloc's minimum alignment + follow glibc's lead. If you specify `--with-lg-quantum=3` during + configuration, jemalloc will provide additional size classes that are not + 16-byte-aligned (24, 40, and 56). + +* `--with-lg-vaddr=<lg-vaddr>` + + Specify the number of significant virtual address bits. By default, the + configure script attempts to detect virtual address size on those platforms + where it knows how, and picks a default otherwise. This option may be + useful when cross-compiling. + +* `--disable-initial-exec-tls` + + Disable the initial-exec TLS model for jemalloc's internal thread-local + storage (on those platforms that support explicit settings). This can allow + jemalloc to be dynamically loaded after program startup (e.g. using dlopen). + Note that in this case, there will be two malloc implementations operating + in the same process, which will almost certainly result in confusing runtime + crashes if pointers leak from one implementation to the other. + +* `--disable-libdl` + + Disable the usage of libdl, namely dlsym(3) which is required by the lazy + lock option. This can allow building static binaries. + +The following environment variables (not a definitive list) impact configure's +behavior: + +* `CFLAGS="?"` +* `CXXFLAGS="?"` + + Pass these flags to the C/C++ compiler. Any flags set by the configure + script are prepended, which means explicitly set flags generally take + precedence. Take care when specifying flags such as -Werror, because + configure tests may be affected in undesirable ways. + +* `EXTRA_CFLAGS="?"` +* `EXTRA_CXXFLAGS="?"` + + Append these flags to CFLAGS/CXXFLAGS, without passing them to the + compiler(s) during configuration. This makes it possible to add flags such + as -Werror, while allowing the configure script to determine what other + flags are appropriate for the specified configuration. + +* `CPPFLAGS="?"` + + Pass these flags to the C preprocessor. Note that CFLAGS is not passed to + 'cpp' when 'configure' is looking for include files, so you must use + CPPFLAGS instead if you need to help 'configure' find header files. + +* `LD_LIBRARY_PATH="?"` + + 'ld' uses this colon-separated list to find libraries. + +* `LDFLAGS="?"` + + Pass these flags when linking. + +* `PATH="?"` + + 'configure' uses this to find programs. + +In some cases it may be necessary to work around configuration results that do +not match reality. For example, Linux 4.5 added support for the MADV_FREE flag +to madvise(2), which can cause problems if building on a host with MADV_FREE +support and deploying to a target without. To work around this, use a cache +file to override the relevant configuration variable defined in configure.ac, +e.g.: + + echo "je_cv_madv_free=no" > config.cache && ./configure -C + + +## Advanced compilation + +To build only parts of jemalloc, use the following targets: + + build_lib_shared + build_lib_static + build_lib + build_doc_html + build_doc_man + build_doc + +To install only parts of jemalloc, use the following targets: + + install_bin + install_include + install_lib_shared + install_lib_static + install_lib_pc + install_lib + install_doc_html + install_doc_man + install_doc + +To clean up build results to varying degrees, use the following make targets: + + clean + distclean + relclean + + +## Advanced installation + +Optionally, define make variables when invoking make, including (not +exclusively): + +* `INCLUDEDIR="?"` + + Use this as the installation prefix for header files. + +* `LIBDIR="?"` + + Use this as the installation prefix for libraries. + +* `MANDIR="?"` + + Use this as the installation prefix for man pages. + +* `DESTDIR="?"` + + Prepend DESTDIR to INCLUDEDIR, LIBDIR, DATADIR, and MANDIR. This is useful + when installing to a different path than was specified via --prefix. + +* `CC="?"` + + Use this to invoke the C compiler. + +* `CFLAGS="?"` + + Pass these flags to the compiler. + +* `CPPFLAGS="?"` + + Pass these flags to the C preprocessor. + +* `LDFLAGS="?"` + + Pass these flags when linking. + +* `PATH="?"` + + Use this to search for programs used during configuration and building. + + +## Development + +If you intend to make non-trivial changes to jemalloc, use the 'autogen.sh' +script rather than 'configure'. This re-generates 'configure', enables +configuration dependency rules, and enables re-generation of automatically +generated source files. + +The build system supports using an object directory separate from the source +tree. For example, you can create an 'obj' directory, and from within that +directory, issue configuration and build commands: + + autoconf + mkdir obj + cd obj + ../configure --enable-autogen + make + + +## Documentation + +The manual page is generated in both html and roff formats. Any web browser +can be used to view the html manual. The roff manual page can be formatted +prior to installation via the following command: + + nroff -man -t doc/jemalloc.3 diff --git a/deps/jemalloc/Makefile.in b/deps/jemalloc/Makefile.in new file mode 100644 index 000000000..7128b007e --- /dev/null +++ b/deps/jemalloc/Makefile.in @@ -0,0 +1,623 @@ +# Clear out all vpaths, then set just one (default vpath) for the main build +# directory. +vpath +vpath % . + +# Clear the default suffixes, so that built-in rules are not used. +.SUFFIXES : + +SHELL := /bin/sh + +CC := @CC@ +CXX := @CXX@ + +# Configuration parameters. +DESTDIR = +BINDIR := $(DESTDIR)@BINDIR@ +INCLUDEDIR := $(DESTDIR)@INCLUDEDIR@ +LIBDIR := $(DESTDIR)@LIBDIR@ +DATADIR := $(DESTDIR)@DATADIR@ +MANDIR := $(DESTDIR)@MANDIR@ +srcroot := @srcroot@ +objroot := @objroot@ +abs_srcroot := @abs_srcroot@ +abs_objroot := @abs_objroot@ + +# Build parameters. +CPPFLAGS := @CPPFLAGS@ -I$(objroot)include -I$(srcroot)include +CONFIGURE_CFLAGS := @CONFIGURE_CFLAGS@ +SPECIFIED_CFLAGS := @SPECIFIED_CFLAGS@ +EXTRA_CFLAGS := @EXTRA_CFLAGS@ +CFLAGS := $(strip $(CONFIGURE_CFLAGS) $(SPECIFIED_CFLAGS) $(EXTRA_CFLAGS)) +CONFIGURE_CXXFLAGS := @CONFIGURE_CXXFLAGS@ +SPECIFIED_CXXFLAGS := @SPECIFIED_CXXFLAGS@ +EXTRA_CXXFLAGS := @EXTRA_CXXFLAGS@ +CXXFLAGS := $(strip $(CONFIGURE_CXXFLAGS) $(SPECIFIED_CXXFLAGS) $(EXTRA_CXXFLAGS)) +LDFLAGS := @LDFLAGS@ +EXTRA_LDFLAGS := @EXTRA_LDFLAGS@ +LIBS := @LIBS@ +RPATH_EXTRA := @RPATH_EXTRA@ +SO := @so@ +IMPORTLIB := @importlib@ +O := @o@ +A := @a@ +EXE := @exe@ +LIBPREFIX := @libprefix@ +REV := @rev@ +install_suffix := @install_suffix@ +ABI := @abi@ +XSLTPROC := @XSLTPROC@ +XSLROOT := @XSLROOT@ +AUTOCONF := @AUTOCONF@ +_RPATH = @RPATH@ +RPATH = $(if $(1),$(call _RPATH,$(1))) +cfghdrs_in := $(addprefix $(srcroot),@cfghdrs_in@) +cfghdrs_out := @cfghdrs_out@ +cfgoutputs_in := $(addprefix $(srcroot),@cfgoutputs_in@) +cfgoutputs_out := @cfgoutputs_out@ +enable_autogen := @enable_autogen@ +enable_doc := @enable_doc@ +enable_shared := @enable_shared@ +enable_static := @enable_static@ +enable_prof := @enable_prof@ +enable_zone_allocator := @enable_zone_allocator@ +enable_experimental_smallocx := @enable_experimental_smallocx@ +MALLOC_CONF := @JEMALLOC_CPREFIX@MALLOC_CONF +link_whole_archive := @link_whole_archive@ +DSO_LDFLAGS = @DSO_LDFLAGS@ +SOREV = @SOREV@ +PIC_CFLAGS = @PIC_CFLAGS@ +CTARGET = @CTARGET@ +LDTARGET = @LDTARGET@ +TEST_LD_MODE = @TEST_LD_MODE@ +MKLIB = @MKLIB@ +AR = @AR@ +ARFLAGS = @ARFLAGS@ +DUMP_SYMS = @DUMP_SYMS@ +AWK := @AWK@ +CC_MM = @CC_MM@ +LM := @LM@ +INSTALL = @INSTALL@ + +ifeq (macho, $(ABI)) +TEST_LIBRARY_PATH := DYLD_FALLBACK_LIBRARY_PATH="$(objroot)lib" +else +ifeq (pecoff, $(ABI)) +TEST_LIBRARY_PATH := PATH="$(PATH):$(objroot)lib" +else +TEST_LIBRARY_PATH := +endif +endif + +LIBJEMALLOC := $(LIBPREFIX)jemalloc$(install_suffix) + +# Lists of files. +BINS := $(objroot)bin/jemalloc-config $(objroot)bin/jemalloc.sh $(objroot)bin/jeprof +C_HDRS := $(objroot)include/jemalloc/jemalloc$(install_suffix).h +C_SRCS := $(srcroot)src/jemalloc.c \ + $(srcroot)src/arena.c \ + $(srcroot)src/background_thread.c \ + $(srcroot)src/base.c \ + $(srcroot)src/bin.c \ + $(srcroot)src/bitmap.c \ + $(srcroot)src/ckh.c \ + $(srcroot)src/ctl.c \ + $(srcroot)src/div.c \ + $(srcroot)src/extent.c \ + $(srcroot)src/extent_dss.c \ + $(srcroot)src/extent_mmap.c \ + $(srcroot)src/hash.c \ + $(srcroot)src/hook.c \ + $(srcroot)src/large.c \ + $(srcroot)src/log.c \ + $(srcroot)src/malloc_io.c \ + $(srcroot)src/mutex.c \ + $(srcroot)src/mutex_pool.c \ + $(srcroot)src/nstime.c \ + $(srcroot)src/pages.c \ + $(srcroot)src/prng.c \ + $(srcroot)src/prof.c \ + $(srcroot)src/rtree.c \ + $(srcroot)src/safety_check.c \ + $(srcroot)src/stats.c \ + $(srcroot)src/sc.c \ + $(srcroot)src/sz.c \ + $(srcroot)src/tcache.c \ + $(srcroot)src/test_hooks.c \ + $(srcroot)src/ticker.c \ + $(srcroot)src/tsd.c \ + $(srcroot)src/witness.c +ifeq ($(enable_zone_allocator), 1) +C_SRCS += $(srcroot)src/zone.c +endif +ifeq ($(IMPORTLIB),$(SO)) +STATIC_LIBS := $(objroot)lib/$(LIBJEMALLOC).$(A) +endif +ifdef PIC_CFLAGS +STATIC_LIBS += $(objroot)lib/$(LIBJEMALLOC)_pic.$(A) +else +STATIC_LIBS += $(objroot)lib/$(LIBJEMALLOC)_s.$(A) +endif +DSOS := $(objroot)lib/$(LIBJEMALLOC).$(SOREV) +ifneq ($(SOREV),$(SO)) +DSOS += $(objroot)lib/$(LIBJEMALLOC).$(SO) +endif +ifeq (1, $(link_whole_archive)) +LJEMALLOC := -Wl,--whole-archive -L$(objroot)lib -l$(LIBJEMALLOC) -Wl,--no-whole-archive +else +LJEMALLOC := $(objroot)lib/$(LIBJEMALLOC).$(IMPORTLIB) +endif +PC := $(objroot)jemalloc.pc +MAN3 := $(objroot)doc/jemalloc$(install_suffix).3 +DOCS_XML := $(objroot)doc/jemalloc$(install_suffix).xml +DOCS_HTML := $(DOCS_XML:$(objroot)%.xml=$(objroot)%.html) +DOCS_MAN3 := $(DOCS_XML:$(objroot)%.xml=$(objroot)%.3) +DOCS := $(DOCS_HTML) $(DOCS_MAN3) +C_TESTLIB_SRCS := $(srcroot)test/src/btalloc.c $(srcroot)test/src/btalloc_0.c \ + $(srcroot)test/src/btalloc_1.c $(srcroot)test/src/math.c \ + $(srcroot)test/src/mtx.c $(srcroot)test/src/mq.c \ + $(srcroot)test/src/SFMT.c $(srcroot)test/src/test.c \ + $(srcroot)test/src/thd.c $(srcroot)test/src/timer.c +ifeq (1, $(link_whole_archive)) +C_UTIL_INTEGRATION_SRCS := +C_UTIL_CPP_SRCS := +else +C_UTIL_INTEGRATION_SRCS := $(srcroot)src/nstime.c $(srcroot)src/malloc_io.c +C_UTIL_CPP_SRCS := $(srcroot)src/nstime.c $(srcroot)src/malloc_io.c +endif +TESTS_UNIT := \ + $(srcroot)test/unit/a0.c \ + $(srcroot)test/unit/arena_reset.c \ + $(srcroot)test/unit/atomic.c \ + $(srcroot)test/unit/background_thread.c \ + $(srcroot)test/unit/background_thread_enable.c \ + $(srcroot)test/unit/base.c \ + $(srcroot)test/unit/bitmap.c \ + $(srcroot)test/unit/bit_util.c \ + $(srcroot)test/unit/binshard.c \ + $(srcroot)test/unit/ckh.c \ + $(srcroot)test/unit/decay.c \ + $(srcroot)test/unit/div.c \ + $(srcroot)test/unit/emitter.c \ + $(srcroot)test/unit/extent_quantize.c \ + $(srcroot)test/unit/extent_util.c \ + $(srcroot)test/unit/fork.c \ + $(srcroot)test/unit/hash.c \ + $(srcroot)test/unit/hook.c \ + $(srcroot)test/unit/huge.c \ + $(srcroot)test/unit/junk.c \ + $(srcroot)test/unit/junk_alloc.c \ + $(srcroot)test/unit/junk_free.c \ + $(srcroot)test/unit/log.c \ + $(srcroot)test/unit/mallctl.c \ + $(srcroot)test/unit/malloc_io.c \ + $(srcroot)test/unit/math.c \ + $(srcroot)test/unit/mq.c \ + $(srcroot)test/unit/mtx.c \ + $(srcroot)test/unit/pack.c \ + $(srcroot)test/unit/pages.c \ + $(srcroot)test/unit/ph.c \ + $(srcroot)test/unit/prng.c \ + $(srcroot)test/unit/prof_accum.c \ + $(srcroot)test/unit/prof_active.c \ + $(srcroot)test/unit/prof_gdump.c \ + $(srcroot)test/unit/prof_idump.c \ + $(srcroot)test/unit/prof_log.c \ + $(srcroot)test/unit/prof_reset.c \ + $(srcroot)test/unit/prof_tctx.c \ + $(srcroot)test/unit/prof_thread_name.c \ + $(srcroot)test/unit/ql.c \ + $(srcroot)test/unit/qr.c \ + $(srcroot)test/unit/rb.c \ + $(srcroot)test/unit/retained.c \ + $(srcroot)test/unit/rtree.c \ + $(srcroot)test/unit/safety_check.c \ + $(srcroot)test/unit/seq.c \ + $(srcroot)test/unit/SFMT.c \ + $(srcroot)test/unit/sc.c \ + $(srcroot)test/unit/size_classes.c \ + $(srcroot)test/unit/slab.c \ + $(srcroot)test/unit/smoothstep.c \ + $(srcroot)test/unit/spin.c \ + $(srcroot)test/unit/stats.c \ + $(srcroot)test/unit/stats_print.c \ + $(srcroot)test/unit/test_hooks.c \ + $(srcroot)test/unit/ticker.c \ + $(srcroot)test/unit/nstime.c \ + $(srcroot)test/unit/tsd.c \ + $(srcroot)test/unit/witness.c \ + $(srcroot)test/unit/zero.c +ifeq (@enable_prof@, 1) +TESTS_UNIT += \ + $(srcroot)test/unit/arena_reset_prof.c +endif +TESTS_INTEGRATION := $(srcroot)test/integration/aligned_alloc.c \ + $(srcroot)test/integration/allocated.c \ + $(srcroot)test/integration/extent.c \ + $(srcroot)test/integration/malloc.c \ + $(srcroot)test/integration/mallocx.c \ + $(srcroot)test/integration/MALLOCX_ARENA.c \ + $(srcroot)test/integration/overflow.c \ + $(srcroot)test/integration/posix_memalign.c \ + $(srcroot)test/integration/rallocx.c \ + $(srcroot)test/integration/sdallocx.c \ + $(srcroot)test/integration/slab_sizes.c \ + $(srcroot)test/integration/thread_arena.c \ + $(srcroot)test/integration/thread_tcache_enabled.c \ + $(srcroot)test/integration/xallocx.c +ifeq (@enable_experimental_smallocx@, 1) +TESTS_INTEGRATION += \ + $(srcroot)test/integration/smallocx.c +endif +ifeq (@enable_cxx@, 1) +CPP_SRCS := $(srcroot)src/jemalloc_cpp.cpp +TESTS_INTEGRATION_CPP := $(srcroot)test/integration/cpp/basic.cpp +else +CPP_SRCS := +TESTS_INTEGRATION_CPP := +endif +TESTS_STRESS := $(srcroot)test/stress/microbench.c \ + $(srcroot)test/stress/hookbench.c + + +TESTS := $(TESTS_UNIT) $(TESTS_INTEGRATION) $(TESTS_INTEGRATION_CPP) $(TESTS_STRESS) + +PRIVATE_NAMESPACE_HDRS := $(objroot)include/jemalloc/internal/private_namespace.h $(objroot)include/jemalloc/internal/private_namespace_jet.h +PRIVATE_NAMESPACE_GEN_HDRS := $(PRIVATE_NAMESPACE_HDRS:%.h=%.gen.h) +C_SYM_OBJS := $(C_SRCS:$(srcroot)%.c=$(objroot)%.sym.$(O)) +C_SYMS := $(C_SRCS:$(srcroot)%.c=$(objroot)%.sym) +C_OBJS := $(C_SRCS:$(srcroot)%.c=$(objroot)%.$(O)) +CPP_OBJS := $(CPP_SRCS:$(srcroot)%.cpp=$(objroot)%.$(O)) +C_PIC_OBJS := $(C_SRCS:$(srcroot)%.c=$(objroot)%.pic.$(O)) +CPP_PIC_OBJS := $(CPP_SRCS:$(srcroot)%.cpp=$(objroot)%.pic.$(O)) +C_JET_SYM_OBJS := $(C_SRCS:$(srcroot)%.c=$(objroot)%.jet.sym.$(O)) +C_JET_SYMS := $(C_SRCS:$(srcroot)%.c=$(objroot)%.jet.sym) +C_JET_OBJS := $(C_SRCS:$(srcroot)%.c=$(objroot)%.jet.$(O)) +C_TESTLIB_UNIT_OBJS := $(C_TESTLIB_SRCS:$(srcroot)%.c=$(objroot)%.unit.$(O)) +C_TESTLIB_INTEGRATION_OBJS := $(C_TESTLIB_SRCS:$(srcroot)%.c=$(objroot)%.integration.$(O)) +C_UTIL_INTEGRATION_OBJS := $(C_UTIL_INTEGRATION_SRCS:$(srcroot)%.c=$(objroot)%.integration.$(O)) +C_TESTLIB_STRESS_OBJS := $(C_TESTLIB_SRCS:$(srcroot)%.c=$(objroot)%.stress.$(O)) +C_TESTLIB_OBJS := $(C_TESTLIB_UNIT_OBJS) $(C_TESTLIB_INTEGRATION_OBJS) $(C_UTIL_INTEGRATION_OBJS) $(C_TESTLIB_STRESS_OBJS) + +TESTS_UNIT_OBJS := $(TESTS_UNIT:$(srcroot)%.c=$(objroot)%.$(O)) +TESTS_INTEGRATION_OBJS := $(TESTS_INTEGRATION:$(srcroot)%.c=$(objroot)%.$(O)) +TESTS_INTEGRATION_CPP_OBJS := $(TESTS_INTEGRATION_CPP:$(srcroot)%.cpp=$(objroot)%.$(O)) +TESTS_STRESS_OBJS := $(TESTS_STRESS:$(srcroot)%.c=$(objroot)%.$(O)) +TESTS_OBJS := $(TESTS_UNIT_OBJS) $(TESTS_INTEGRATION_OBJS) $(TESTS_STRESS_OBJS) +TESTS_CPP_OBJS := $(TESTS_INTEGRATION_CPP_OBJS) + +.PHONY: all dist build_doc_html build_doc_man build_doc +.PHONY: install_bin install_include install_lib +.PHONY: install_doc_html install_doc_man install_doc install +.PHONY: tests check clean distclean relclean + +.SECONDARY : $(PRIVATE_NAMESPACE_GEN_HDRS) $(TESTS_OBJS) $(TESTS_CPP_OBJS) + +# Default target. +all: build_lib + +dist: build_doc + +$(objroot)doc/%.html : $(objroot)doc/%.xml $(srcroot)doc/stylesheet.xsl $(objroot)doc/html.xsl +ifneq ($(XSLROOT),) + $(XSLTPROC) -o $@ $(objroot)doc/html.xsl $< +else +ifeq ($(wildcard $(DOCS_HTML)),) + @echo "<p>Missing xsltproc. Doc not built.</p>" > $@ +endif + @echo "Missing xsltproc. "$@" not (re)built." +endif + +$(objroot)doc/%.3 : $(objroot)doc/%.xml $(srcroot)doc/stylesheet.xsl $(objroot)doc/manpages.xsl +ifneq ($(XSLROOT),) + $(XSLTPROC) -o $@ $(objroot)doc/manpages.xsl $< +else +ifeq ($(wildcard $(DOCS_MAN3)),) + @echo "Missing xsltproc. Doc not built." > $@ +endif + @echo "Missing xsltproc. "$@" not (re)built." +endif + +build_doc_html: $(DOCS_HTML) +build_doc_man: $(DOCS_MAN3) +build_doc: $(DOCS) + +# +# Include generated dependency files. +# +ifdef CC_MM +-include $(C_SYM_OBJS:%.$(O)=%.d) +-include $(C_OBJS:%.$(O)=%.d) +-include $(CPP_OBJS:%.$(O)=%.d) +-include $(C_PIC_OBJS:%.$(O)=%.d) +-include $(CPP_PIC_OBJS:%.$(O)=%.d) +-include $(C_JET_SYM_OBJS:%.$(O)=%.d) +-include $(C_JET_OBJS:%.$(O)=%.d) +-include $(C_TESTLIB_OBJS:%.$(O)=%.d) +-include $(TESTS_OBJS:%.$(O)=%.d) +-include $(TESTS_CPP_OBJS:%.$(O)=%.d) +endif + +$(C_SYM_OBJS): $(objroot)src/%.sym.$(O): $(srcroot)src/%.c +$(C_SYM_OBJS): CPPFLAGS += -DJEMALLOC_NO_PRIVATE_NAMESPACE +$(C_SYMS): $(objroot)src/%.sym: $(objroot)src/%.sym.$(O) +$(C_OBJS): $(objroot)src/%.$(O): $(srcroot)src/%.c +$(CPP_OBJS): $(objroot)src/%.$(O): $(srcroot)src/%.cpp +$(C_PIC_OBJS): $(objroot)src/%.pic.$(O): $(srcroot)src/%.c +$(C_PIC_OBJS): CFLAGS += $(PIC_CFLAGS) +$(CPP_PIC_OBJS): $(objroot)src/%.pic.$(O): $(srcroot)src/%.cpp +$(CPP_PIC_OBJS): CXXFLAGS += $(PIC_CFLAGS) +$(C_JET_SYM_OBJS): $(objroot)src/%.jet.sym.$(O): $(srcroot)src/%.c +$(C_JET_SYM_OBJS): CPPFLAGS += -DJEMALLOC_JET -DJEMALLOC_NO_PRIVATE_NAMESPACE +$(C_JET_SYMS): $(objroot)src/%.jet.sym: $(objroot)src/%.jet.sym.$(O) +$(C_JET_OBJS): $(objroot)src/%.jet.$(O): $(srcroot)src/%.c +$(C_JET_OBJS): CPPFLAGS += -DJEMALLOC_JET +$(C_TESTLIB_UNIT_OBJS): $(objroot)test/src/%.unit.$(O): $(srcroot)test/src/%.c +$(C_TESTLIB_UNIT_OBJS): CPPFLAGS += -DJEMALLOC_UNIT_TEST +$(C_TESTLIB_INTEGRATION_OBJS): $(objroot)test/src/%.integration.$(O): $(srcroot)test/src/%.c +$(C_TESTLIB_INTEGRATION_OBJS): CPPFLAGS += -DJEMALLOC_INTEGRATION_TEST +$(C_UTIL_INTEGRATION_OBJS): $(objroot)src/%.integration.$(O): $(srcroot)src/%.c +$(C_TESTLIB_STRESS_OBJS): $(objroot)test/src/%.stress.$(O): $(srcroot)test/src/%.c +$(C_TESTLIB_STRESS_OBJS): CPPFLAGS += -DJEMALLOC_STRESS_TEST -DJEMALLOC_STRESS_TESTLIB +$(C_TESTLIB_OBJS): CPPFLAGS += -I$(srcroot)test/include -I$(objroot)test/include +$(TESTS_UNIT_OBJS): CPPFLAGS += -DJEMALLOC_UNIT_TEST +$(TESTS_INTEGRATION_OBJS): CPPFLAGS += -DJEMALLOC_INTEGRATION_TEST +$(TESTS_INTEGRATION_CPP_OBJS): CPPFLAGS += -DJEMALLOC_INTEGRATION_CPP_TEST +$(TESTS_STRESS_OBJS): CPPFLAGS += -DJEMALLOC_STRESS_TEST +$(TESTS_OBJS): $(objroot)test/%.$(O): $(srcroot)test/%.c +$(TESTS_CPP_OBJS): $(objroot)test/%.$(O): $(srcroot)test/%.cpp +$(TESTS_OBJS): CPPFLAGS += -I$(srcroot)test/include -I$(objroot)test/include +$(TESTS_CPP_OBJS): CPPFLAGS += -I$(srcroot)test/include -I$(objroot)test/include +ifneq ($(IMPORTLIB),$(SO)) +$(CPP_OBJS) $(C_SYM_OBJS) $(C_OBJS) $(C_JET_SYM_OBJS) $(C_JET_OBJS): CPPFLAGS += -DDLLEXPORT +endif + +# Dependencies. +ifndef CC_MM +HEADER_DIRS = $(srcroot)include/jemalloc/internal \ + $(objroot)include/jemalloc $(objroot)include/jemalloc/internal +HEADERS = $(filter-out $(PRIVATE_NAMESPACE_HDRS),$(wildcard $(foreach dir,$(HEADER_DIRS),$(dir)/*.h))) +$(C_SYM_OBJS) $(C_OBJS) $(CPP_OBJS) $(C_PIC_OBJS) $(CPP_PIC_OBJS) $(C_JET_SYM_OBJS) $(C_JET_OBJS) $(C_TESTLIB_OBJS) $(TESTS_OBJS) $(TESTS_CPP_OBJS): $(HEADERS) +$(TESTS_OBJS) $(TESTS_CPP_OBJS): $(objroot)test/include/test/jemalloc_test.h +endif + +$(C_OBJS) $(CPP_OBJS) $(C_PIC_OBJS) $(CPP_PIC_OBJS) $(C_TESTLIB_INTEGRATION_OBJS) $(C_UTIL_INTEGRATION_OBJS) $(TESTS_INTEGRATION_OBJS) $(TESTS_INTEGRATION_CPP_OBJS): $(objroot)include/jemalloc/internal/private_namespace.h +$(C_JET_OBJS) $(C_TESTLIB_UNIT_OBJS) $(C_TESTLIB_STRESS_OBJS) $(TESTS_UNIT_OBJS) $(TESTS_STRESS_OBJS): $(objroot)include/jemalloc/internal/private_namespace_jet.h + +$(C_SYM_OBJS) $(C_OBJS) $(C_PIC_OBJS) $(C_JET_SYM_OBJS) $(C_JET_OBJS) $(C_TESTLIB_OBJS) $(TESTS_OBJS): %.$(O): + @mkdir -p $(@D) + $(CC) $(CFLAGS) -c $(CPPFLAGS) $(CTARGET) $< +ifdef CC_MM + @$(CC) -MM $(CPPFLAGS) -MT $@ -o $(@:%.$(O)=%.d) $< +endif + +$(C_SYMS): %.sym: + @mkdir -p $(@D) + $(DUMP_SYMS) $< | $(AWK) -f $(objroot)include/jemalloc/internal/private_symbols.awk > $@ + +$(C_JET_SYMS): %.sym: + @mkdir -p $(@D) + $(DUMP_SYMS) $< | $(AWK) -f $(objroot)include/jemalloc/internal/private_symbols_jet.awk > $@ + +$(objroot)include/jemalloc/internal/private_namespace.gen.h: $(C_SYMS) + $(SHELL) $(srcroot)include/jemalloc/internal/private_namespace.sh $^ > $@ + +$(objroot)include/jemalloc/internal/private_namespace_jet.gen.h: $(C_JET_SYMS) + $(SHELL) $(srcroot)include/jemalloc/internal/private_namespace.sh $^ > $@ + +%.h: %.gen.h + @if ! `cmp -s $< $@` ; then echo "cp $< $<"; cp $< $@ ; fi + +$(CPP_OBJS) $(CPP_PIC_OBJS) $(TESTS_CPP_OBJS): %.$(O): + @mkdir -p $(@D) + $(CXX) $(CXXFLAGS) -c $(CPPFLAGS) $(CTARGET) $< +ifdef CC_MM + @$(CXX) -MM $(CPPFLAGS) -MT $@ -o $(@:%.$(O)=%.d) $< +endif + +ifneq ($(SOREV),$(SO)) +%.$(SO) : %.$(SOREV) + @mkdir -p $(@D) + ln -sf $(<F) $@ +endif + +$(objroot)lib/$(LIBJEMALLOC).$(SOREV) : $(if $(PIC_CFLAGS),$(C_PIC_OBJS),$(C_OBJS)) $(if $(PIC_CFLAGS),$(CPP_PIC_OBJS),$(CPP_OBJS)) + @mkdir -p $(@D) + $(CC) $(DSO_LDFLAGS) $(call RPATH,$(RPATH_EXTRA)) $(LDTARGET) $+ $(LDFLAGS) $(LIBS) $(EXTRA_LDFLAGS) + +$(objroot)lib/$(LIBJEMALLOC)_pic.$(A) : $(C_PIC_OBJS) $(CPP_PIC_OBJS) +$(objroot)lib/$(LIBJEMALLOC).$(A) : $(C_OBJS) $(CPP_OBJS) +$(objroot)lib/$(LIBJEMALLOC)_s.$(A) : $(C_OBJS) $(CPP_OBJS) + +$(STATIC_LIBS): + @mkdir -p $(@D) + $(AR) $(ARFLAGS)@AROUT@ $+ + +$(objroot)test/unit/%$(EXE): $(objroot)test/unit/%.$(O) $(C_JET_OBJS) $(C_TESTLIB_UNIT_OBJS) + @mkdir -p $(@D) + $(CC) $(LDTARGET) $(filter %.$(O),$^) $(call RPATH,$(objroot)lib) $(LDFLAGS) $(filter-out -lm,$(LIBS)) $(LM) $(EXTRA_LDFLAGS) + +$(objroot)test/integration/%$(EXE): $(objroot)test/integration/%.$(O) $(C_TESTLIB_INTEGRATION_OBJS) $(C_UTIL_INTEGRATION_OBJS) $(objroot)lib/$(LIBJEMALLOC).$(IMPORTLIB) + @mkdir -p $(@D) + $(CC) $(TEST_LD_MODE) $(LDTARGET) $(filter %.$(O),$^) $(call RPATH,$(objroot)lib) $(LJEMALLOC) $(LDFLAGS) $(filter-out -lm,$(filter -lrt -pthread -lstdc++,$(LIBS))) $(LM) $(EXTRA_LDFLAGS) + +$(objroot)test/integration/cpp/%$(EXE): $(objroot)test/integration/cpp/%.$(O) $(C_TESTLIB_INTEGRATION_OBJS) $(C_UTIL_INTEGRATION_OBJS) $(objroot)lib/$(LIBJEMALLOC).$(IMPORTLIB) + @mkdir -p $(@D) + $(CXX) $(LDTARGET) $(filter %.$(O),$^) $(call RPATH,$(objroot)lib) $(objroot)lib/$(LIBJEMALLOC).$(IMPORTLIB) $(LDFLAGS) $(filter-out -lm,$(LIBS)) -lm $(EXTRA_LDFLAGS) + +$(objroot)test/stress/%$(EXE): $(objroot)test/stress/%.$(O) $(C_JET_OBJS) $(C_TESTLIB_STRESS_OBJS) $(objroot)lib/$(LIBJEMALLOC).$(IMPORTLIB) + @mkdir -p $(@D) + $(CC) $(TEST_LD_MODE) $(LDTARGET) $(filter %.$(O),$^) $(call RPATH,$(objroot)lib) $(objroot)lib/$(LIBJEMALLOC).$(IMPORTLIB) $(LDFLAGS) $(filter-out -lm,$(LIBS)) $(LM) $(EXTRA_LDFLAGS) + +build_lib_shared: $(DSOS) +build_lib_static: $(STATIC_LIBS) +ifeq ($(enable_shared), 1) +build_lib: build_lib_shared +endif +ifeq ($(enable_static), 1) +build_lib: build_lib_static +endif + +install_bin: + $(INSTALL) -d $(BINDIR) + @for b in $(BINS); do \ + echo "$(INSTALL) -m 755 $$b $(BINDIR)"; \ + $(INSTALL) -m 755 $$b $(BINDIR); \ +done + +install_include: + $(INSTALL) -d $(INCLUDEDIR)/jemalloc + @for h in $(C_HDRS); do \ + echo "$(INSTALL) -m 644 $$h $(INCLUDEDIR)/jemalloc"; \ + $(INSTALL) -m 644 $$h $(INCLUDEDIR)/jemalloc; \ +done + +install_lib_shared: $(DSOS) + $(INSTALL) -d $(LIBDIR) + $(INSTALL) -m 755 $(objroot)lib/$(LIBJEMALLOC).$(SOREV) $(LIBDIR) +ifneq ($(SOREV),$(SO)) + ln -sf $(LIBJEMALLOC).$(SOREV) $(LIBDIR)/$(LIBJEMALLOC).$(SO) +endif + +install_lib_static: $(STATIC_LIBS) + $(INSTALL) -d $(LIBDIR) + @for l in $(STATIC_LIBS); do \ + echo "$(INSTALL) -m 755 $$l $(LIBDIR)"; \ + $(INSTALL) -m 755 $$l $(LIBDIR); \ +done + +install_lib_pc: $(PC) + $(INSTALL) -d $(LIBDIR)/pkgconfig + @for l in $(PC); do \ + echo "$(INSTALL) -m 644 $$l $(LIBDIR)/pkgconfig"; \ + $(INSTALL) -m 644 $$l $(LIBDIR)/pkgconfig; \ +done + +ifeq ($(enable_shared), 1) +install_lib: install_lib_shared +endif +ifeq ($(enable_static), 1) +install_lib: install_lib_static +endif +install_lib: install_lib_pc + +install_doc_html: + $(INSTALL) -d $(DATADIR)/doc/jemalloc$(install_suffix) + @for d in $(DOCS_HTML); do \ + echo "$(INSTALL) -m 644 $$d $(DATADIR)/doc/jemalloc$(install_suffix)"; \ + $(INSTALL) -m 644 $$d $(DATADIR)/doc/jemalloc$(install_suffix); \ +done + +install_doc_man: + $(INSTALL) -d $(MANDIR)/man3 + @for d in $(DOCS_MAN3); do \ + echo "$(INSTALL) -m 644 $$d $(MANDIR)/man3"; \ + $(INSTALL) -m 644 $$d $(MANDIR)/man3; \ +done + +install_doc: build_doc install_doc_html install_doc_man + +install: install_bin install_include install_lib + +ifeq ($(enable_doc), 1) +install: install_doc +endif + +tests_unit: $(TESTS_UNIT:$(srcroot)%.c=$(objroot)%$(EXE)) +tests_integration: $(TESTS_INTEGRATION:$(srcroot)%.c=$(objroot)%$(EXE)) $(TESTS_INTEGRATION_CPP:$(srcroot)%.cpp=$(objroot)%$(EXE)) +tests_stress: $(TESTS_STRESS:$(srcroot)%.c=$(objroot)%$(EXE)) +tests: tests_unit tests_integration tests_stress + +check_unit_dir: + @mkdir -p $(objroot)test/unit +check_integration_dir: + @mkdir -p $(objroot)test/integration +stress_dir: + @mkdir -p $(objroot)test/stress +check_dir: check_unit_dir check_integration_dir + +check_unit: tests_unit check_unit_dir + $(SHELL) $(objroot)test/test.sh $(TESTS_UNIT:$(srcroot)%.c=$(objroot)%) +check_integration_prof: tests_integration check_integration_dir +ifeq ($(enable_prof), 1) + $(MALLOC_CONF)="prof:true" $(SHELL) $(objroot)test/test.sh $(TESTS_INTEGRATION:$(srcroot)%.c=$(objroot)%) $(TESTS_INTEGRATION_CPP:$(srcroot)%.cpp=$(objroot)%) + $(MALLOC_CONF)="prof:true,prof_active:false" $(SHELL) $(objroot)test/test.sh $(TESTS_INTEGRATION:$(srcroot)%.c=$(objroot)%) $(TESTS_INTEGRATION_CPP:$(srcroot)%.cpp=$(objroot)%) +endif +check_integration_decay: tests_integration check_integration_dir + $(MALLOC_CONF)="dirty_decay_ms:-1,muzzy_decay_ms:-1" $(SHELL) $(objroot)test/test.sh $(TESTS_INTEGRATION:$(srcroot)%.c=$(objroot)%) $(TESTS_INTEGRATION_CPP:$(srcroot)%.cpp=$(objroot)%) + $(MALLOC_CONF)="dirty_decay_ms:0,muzzy_decay_ms:0" $(SHELL) $(objroot)test/test.sh $(TESTS_INTEGRATION:$(srcroot)%.c=$(objroot)%) $(TESTS_INTEGRATION_CPP:$(srcroot)%.cpp=$(objroot)%) +check_integration: tests_integration check_integration_dir + $(SHELL) $(objroot)test/test.sh $(TESTS_INTEGRATION:$(srcroot)%.c=$(objroot)%) $(TESTS_INTEGRATION_CPP:$(srcroot)%.cpp=$(objroot)%) +stress: tests_stress stress_dir + $(SHELL) $(objroot)test/test.sh $(TESTS_STRESS:$(srcroot)%.c=$(objroot)%) +check: check_unit check_integration check_integration_decay check_integration_prof + +clean: + rm -f $(PRIVATE_NAMESPACE_HDRS) + rm -f $(PRIVATE_NAMESPACE_GEN_HDRS) + rm -f $(C_SYM_OBJS) + rm -f $(C_SYMS) + rm -f $(C_OBJS) + rm -f $(CPP_OBJS) + rm -f $(C_PIC_OBJS) + rm -f $(CPP_PIC_OBJS) + rm -f $(C_JET_SYM_OBJS) + rm -f $(C_JET_SYMS) + rm -f $(C_JET_OBJS) + rm -f $(C_TESTLIB_OBJS) + rm -f $(C_SYM_OBJS:%.$(O)=%.d) + rm -f $(C_OBJS:%.$(O)=%.d) + rm -f $(CPP_OBJS:%.$(O)=%.d) + rm -f $(C_PIC_OBJS:%.$(O)=%.d) + rm -f $(CPP_PIC_OBJS:%.$(O)=%.d) + rm -f $(C_JET_SYM_OBJS:%.$(O)=%.d) + rm -f $(C_JET_OBJS:%.$(O)=%.d) + rm -f $(C_TESTLIB_OBJS:%.$(O)=%.d) + rm -f $(TESTS_OBJS:%.$(O)=%$(EXE)) + rm -f $(TESTS_OBJS) + rm -f $(TESTS_OBJS:%.$(O)=%.d) + rm -f $(TESTS_OBJS:%.$(O)=%.out) + rm -f $(TESTS_CPP_OBJS:%.$(O)=%$(EXE)) + rm -f $(TESTS_CPP_OBJS) + rm -f $(TESTS_CPP_OBJS:%.$(O)=%.d) + rm -f $(TESTS_CPP_OBJS:%.$(O)=%.out) + rm -f $(DSOS) $(STATIC_LIBS) + +distclean: clean + rm -f $(objroot)bin/jemalloc-config + rm -f $(objroot)bin/jemalloc.sh + rm -f $(objroot)bin/jeprof + rm -f $(objroot)config.log + rm -f $(objroot)config.status + rm -f $(objroot)config.stamp + rm -f $(cfghdrs_out) + rm -f $(cfgoutputs_out) + +relclean: distclean + rm -f $(objroot)configure + rm -f $(objroot)VERSION + rm -f $(DOCS_HTML) + rm -f $(DOCS_MAN3) + +#=============================================================================== +# Re-configuration rules. + +ifeq ($(enable_autogen), 1) +$(srcroot)configure : $(srcroot)configure.ac + cd ./$(srcroot) && $(AUTOCONF) + +$(objroot)config.status : $(srcroot)configure + ./$(objroot)config.status --recheck + +$(srcroot)config.stamp.in : $(srcroot)configure.ac + echo stamp > $(srcroot)config.stamp.in + +$(objroot)config.stamp : $(cfgoutputs_in) $(cfghdrs_in) $(srcroot)configure + ./$(objroot)config.status + @touch $@ + +# There must be some action in order for make to re-read Makefile when it is +# out of date. +$(cfgoutputs_out) $(cfghdrs_out) : $(objroot)config.stamp + @true +endif diff --git a/deps/jemalloc/README b/deps/jemalloc/README new file mode 100644 index 000000000..3a6e0d272 --- /dev/null +++ b/deps/jemalloc/README @@ -0,0 +1,20 @@ +jemalloc is a general purpose malloc(3) implementation that emphasizes +fragmentation avoidance and scalable concurrency support. jemalloc first came +into use as the FreeBSD libc allocator in 2005, and since then it has found its +way into numerous applications that rely on its predictable behavior. In 2010 +jemalloc development efforts broadened to include developer support features +such as heap profiling and extensive monitoring/tuning hooks. Modern jemalloc +releases continue to be integrated back into FreeBSD, and therefore versatility +remains critical. Ongoing development efforts trend toward making jemalloc +among the best allocators for a broad range of demanding applications, and +eliminating/mitigating weaknesses that have practical repercussions for real +world applications. + +The COPYING file contains copyright and licensing information. + +The INSTALL file contains information on how to configure, build, and install +jemalloc. + +The ChangeLog file contains a brief summary of changes for each release. + +URL: http://jemalloc.net/ diff --git a/deps/jemalloc/TUNING.md b/deps/jemalloc/TUNING.md new file mode 100644 index 000000000..34fca05b4 --- /dev/null +++ b/deps/jemalloc/TUNING.md @@ -0,0 +1,129 @@ +This document summarizes the common approaches for performance fine tuning with +jemalloc (as of 5.1.0). The default configuration of jemalloc tends to work +reasonably well in practice, and most applications should not have to tune any +options. However, in order to cover a wide range of applications and avoid +pathological cases, the default setting is sometimes kept conservative and +suboptimal, even for many common workloads. When jemalloc is properly tuned for +a specific application / workload, it is common to improve system level metrics +by a few percent, or make favorable trade-offs. + + +## Notable runtime options for performance tuning + +Runtime options can be set via +[malloc_conf](http://jemalloc.net/jemalloc.3.html#tuning). + +* [background_thread](http://jemalloc.net/jemalloc.3.html#background_thread) + + Enabling jemalloc background threads generally improves the tail latency for + application threads, since unused memory purging is shifted to the dedicated + background threads. In addition, unintended purging delay caused by + application inactivity is avoided with background threads. + + Suggested: `background_thread:true` when jemalloc managed threads can be + allowed. + +* [metadata_thp](http://jemalloc.net/jemalloc.3.html#opt.metadata_thp) + + Allowing jemalloc to utilize transparent huge pages for its internal + metadata usually reduces TLB misses significantly, especially for programs + with large memory footprint and frequent allocation / deallocation + activities. Metadata memory usage may increase due to the use of huge + pages. + + Suggested for allocation intensive programs: `metadata_thp:auto` or + `metadata_thp:always`, which is expected to improve CPU utilization at a + small memory cost. + +* [dirty_decay_ms](http://jemalloc.net/jemalloc.3.html#opt.dirty_decay_ms) and + [muzzy_decay_ms](http://jemalloc.net/jemalloc.3.html#opt.muzzy_decay_ms) + + Decay time determines how fast jemalloc returns unused pages back to the + operating system, and therefore provides a fairly straightforward trade-off + between CPU and memory usage. Shorter decay time purges unused pages faster + to reduces memory usage (usually at the cost of more CPU cycles spent on + purging), and vice versa. + + Suggested: tune the values based on the desired trade-offs. + +* [narenas](http://jemalloc.net/jemalloc.3.html#opt.narenas) + + By default jemalloc uses multiple arenas to reduce internal lock contention. + However high arena count may also increase overall memory fragmentation, + since arenas manage memory independently. When high degree of parallelism + is not expected at the allocator level, lower number of arenas often + improves memory usage. + + Suggested: if low parallelism is expected, try lower arena count while + monitoring CPU and memory usage. + +* [percpu_arena](http://jemalloc.net/jemalloc.3.html#opt.percpu_arena) + + Enable dynamic thread to arena association based on running CPU. This has + the potential to improve locality, e.g. when thread to CPU affinity is + present. + + Suggested: try `percpu_arena:percpu` or `percpu_arena:phycpu` if + thread migration between processors is expected to be infrequent. + +Examples: + +* High resource consumption application, prioritizing CPU utilization: + + `background_thread:true,metadata_thp:auto` combined with relaxed decay time + (increased `dirty_decay_ms` and / or `muzzy_decay_ms`, + e.g. `dirty_decay_ms:30000,muzzy_decay_ms:30000`). + +* High resource consumption application, prioritizing memory usage: + + `background_thread:true` combined with shorter decay time (decreased + `dirty_decay_ms` and / or `muzzy_decay_ms`, + e.g. `dirty_decay_ms:5000,muzzy_decay_ms:5000`), and lower arena count + (e.g. number of CPUs). + +* Low resource consumption application: + + `narenas:1,lg_tcache_max:13` combined with shorter decay time (decreased + `dirty_decay_ms` and / or `muzzy_decay_ms`,e.g. + `dirty_decay_ms:1000,muzzy_decay_ms:0`). + +* Extremely conservative -- minimize memory usage at all costs, only suitable when +allocation activity is very rare: + + `narenas:1,tcache:false,dirty_decay_ms:0,muzzy_decay_ms:0` + +Note that it is recommended to combine the options with `abort_conf:true` which +aborts immediately on illegal options. + +## Beyond runtime options + +In addition to the runtime options, there are a number of programmatic ways to +improve application performance with jemalloc. + +* [Explicit arenas](http://jemalloc.net/jemalloc.3.html#arenas.create) + + Manually created arenas can help performance in various ways, e.g. by + managing locality and contention for specific usages. For example, + applications can explicitly allocate frequently accessed objects from a + dedicated arena with + [mallocx()](http://jemalloc.net/jemalloc.3.html#MALLOCX_ARENA) to improve + locality. In addition, explicit arenas often benefit from individually + tuned options, e.g. relaxed [decay + time](http://jemalloc.net/jemalloc.3.html#arena.i.dirty_decay_ms) if + frequent reuse is expected. + +* [Extent hooks](http://jemalloc.net/jemalloc.3.html#arena.i.extent_hooks) + + Extent hooks allow customization for managing underlying memory. One use + case for performance purpose is to utilize huge pages -- for example, + [HHVM](https://github.com/facebook/hhvm/blob/master/hphp/util/alloc.cpp) + uses explicit arenas with customized extent hooks to manage 1GB huge pages + for frequently accessed data, which reduces TLB misses significantly. + +* [Explicit thread-to-arena + binding](http://jemalloc.net/jemalloc.3.html#thread.arena) + + It is common for some threads in an application to have different memory + access / allocation patterns. Threads with heavy workloads often benefit + from explicit binding, e.g. binding very active threads to dedicated arenas + may reduce contention at the allocator level. diff --git a/deps/jemalloc/autogen.sh b/deps/jemalloc/autogen.sh new file mode 100755 index 000000000..75f32da68 --- /dev/null +++ b/deps/jemalloc/autogen.sh @@ -0,0 +1,17 @@ +#!/bin/sh + +for i in autoconf; do + echo "$i" + $i + if [ $? -ne 0 ]; then + echo "Error $? in $i" + exit 1 + fi +done + +echo "./configure --enable-autogen $@" +./configure --enable-autogen $@ +if [ $? -ne 0 ]; then + echo "Error $? in ./configure" + exit 1 +fi diff --git a/deps/jemalloc/bin/jemalloc-config.in b/deps/jemalloc/bin/jemalloc-config.in new file mode 100644 index 000000000..80eca2e64 --- /dev/null +++ b/deps/jemalloc/bin/jemalloc-config.in @@ -0,0 +1,83 @@ +#!/bin/sh + +usage() { + cat <<EOF +Usage: + @BINDIR@/jemalloc-config <option> +Options: + --help | -h : Print usage. + --version : Print jemalloc version. + --revision : Print shared library revision number. + --config : Print configure options used to build jemalloc. + --prefix : Print installation directory prefix. + --bindir : Print binary installation directory. + --datadir : Print data installation directory. + --includedir : Print include installation directory. + --libdir : Print library installation directory. + --mandir : Print manual page installation directory. + --cc : Print compiler used to build jemalloc. + --cflags : Print compiler flags used to build jemalloc. + --cppflags : Print preprocessor flags used to build jemalloc. + --cxxflags : Print C++ compiler flags used to build jemalloc. + --ldflags : Print library flags used to build jemalloc. + --libs : Print libraries jemalloc was linked against. +EOF +} + +prefix="@prefix@" +exec_prefix="@exec_prefix@" + +case "$1" in +--help | -h) + usage + exit 0 + ;; +--version) + echo "@jemalloc_version@" + ;; +--revision) + echo "@rev@" + ;; +--config) + echo "@CONFIG@" + ;; +--prefix) + echo "@PREFIX@" + ;; +--bindir) + echo "@BINDIR@" + ;; +--datadir) + echo "@DATADIR@" + ;; +--includedir) + echo "@INCLUDEDIR@" + ;; +--libdir) + echo "@LIBDIR@" + ;; +--mandir) + echo "@MANDIR@" + ;; +--cc) + echo "@CC@" + ;; +--cflags) + echo "@CFLAGS@" + ;; +--cppflags) + echo "@CPPFLAGS@" + ;; +--cxxflags) + echo "@CXXFLAGS@" + ;; +--ldflags) + echo "@LDFLAGS@ @EXTRA_LDFLAGS@" + ;; +--libs) + echo "@LIBS@" + ;; +*) + usage + exit 1 +esac diff --git a/deps/jemalloc/bin/jemalloc.sh.in b/deps/jemalloc/bin/jemalloc.sh.in new file mode 100644 index 000000000..cdf367375 --- /dev/null +++ b/deps/jemalloc/bin/jemalloc.sh.in @@ -0,0 +1,9 @@ +#!/bin/sh + +prefix=@prefix@ +exec_prefix=@exec_prefix@ +libdir=@libdir@ + +@LD_PRELOAD_VAR@=${libdir}/libjemalloc.@SOREV@ +export @LD_PRELOAD_VAR@ +exec "$@" diff --git a/deps/jemalloc/bin/jeprof.in b/deps/jemalloc/bin/jeprof.in new file mode 100644 index 000000000..3ed408c9c --- /dev/null +++ b/deps/jemalloc/bin/jeprof.in @@ -0,0 +1,5625 @@ +#! /usr/bin/env perl + +# Copyright (c) 1998-2007, Google Inc. +# All rights reserved. +# +# Redistribution and use in source and binary forms, with or without +# modification, are permitted provided that the following conditions are +# met: +# +# * Redistributions of source code must retain the above copyright +# notice, this list of conditions and the following disclaimer. +# * Redistributions in binary form must reproduce the above +# copyright notice, this list of conditions and the following disclaimer +# in the documentation and/or other materials provided with the +# distribution. +# * Neither the name of Google Inc. nor the names of its +# contributors may be used to endorse or promote products derived from +# this software without specific prior written permission. +# +# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +# OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY +# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + +# --- +# Program for printing the profile generated by common/profiler.cc, +# or by the heap profiler (common/debugallocation.cc) +# +# The profile contains a sequence of entries of the form: +# <count> <stack trace> +# This program parses the profile, and generates user-readable +# output. +# +# Examples: +# +# % tools/jeprof "program" "profile" +# Enters "interactive" mode +# +# % tools/jeprof --text "program" "profile" +# Generates one line per procedure +# +# % tools/jeprof --gv "program" "profile" +# Generates annotated call-graph and displays via "gv" +# +# % tools/jeprof --gv --focus=Mutex "program" "profile" +# Restrict to code paths that involve an entry that matches "Mutex" +# +# % tools/jeprof --gv --focus=Mutex --ignore=string "program" "profile" +# Restrict to code paths that involve an entry that matches "Mutex" +# and does not match "string" +# +# % tools/jeprof --list=IBF_CheckDocid "program" "profile" +# Generates disassembly listing of all routines with at least one +# sample that match the --list=<regexp> pattern. The listing is +# annotated with the flat and cumulative sample counts at each line. +# +# % tools/jeprof --disasm=IBF_CheckDocid "program" "profile" +# Generates disassembly listing of all routines with at least one +# sample that match the --disasm=<regexp> pattern. The listing is +# annotated with the flat and cumulative sample counts at each PC value. +# +# TODO: Use color to indicate files? + +use strict; +use warnings; +use Getopt::Long; +use Cwd; + +my $JEPROF_VERSION = "@jemalloc_version@"; +my $PPROF_VERSION = "2.0"; + +# These are the object tools we use which can come from a +# user-specified location using --tools, from the JEPROF_TOOLS +# environment variable, or from the environment. +my %obj_tool_map = ( + "objdump" => "objdump", + "nm" => "nm", + "addr2line" => "addr2line", + "c++filt" => "c++filt", + ## ConfigureObjTools may add architecture-specific entries: + #"nm_pdb" => "nm-pdb", # for reading windows (PDB-format) executables + #"addr2line_pdb" => "addr2line-pdb", # ditto + #"otool" => "otool", # equivalent of objdump on OS X +); +# NOTE: these are lists, so you can put in commandline flags if you want. +my @DOT = ("dot"); # leave non-absolute, since it may be in /usr/local +my @GV = ("gv"); +my @EVINCE = ("evince"); # could also be xpdf or perhaps acroread +my @KCACHEGRIND = ("kcachegrind"); +my @PS2PDF = ("ps2pdf"); +# These are used for dynamic profiles +my @URL_FETCHER = ("curl", "-s", "--fail"); + +# These are the web pages that servers need to support for dynamic profiles +my $HEAP_PAGE = "/pprof/heap"; +my $PROFILE_PAGE = "/pprof/profile"; # must support cgi-param "?seconds=#" +my $PMUPROFILE_PAGE = "/pprof/pmuprofile(?:\\?.*)?"; # must support cgi-param + # ?seconds=#&event=x&period=n +my $GROWTH_PAGE = "/pprof/growth"; +my $CONTENTION_PAGE = "/pprof/contention"; +my $WALL_PAGE = "/pprof/wall(?:\\?.*)?"; # accepts options like namefilter +my $FILTEREDPROFILE_PAGE = "/pprof/filteredprofile(?:\\?.*)?"; +my $CENSUSPROFILE_PAGE = "/pprof/censusprofile(?:\\?.*)?"; # must support cgi-param + # "?seconds=#", + # "?tags_regexp=#" and + # "?type=#". +my $SYMBOL_PAGE = "/pprof/symbol"; # must support symbol lookup via POST +my $PROGRAM_NAME_PAGE = "/pprof/cmdline"; + +# These are the web pages that can be named on the command line. +# All the alternatives must begin with /. +my $PROFILES = "($HEAP_PAGE|$PROFILE_PAGE|$PMUPROFILE_PAGE|" . + "$GROWTH_PAGE|$CONTENTION_PAGE|$WALL_PAGE|" . + "$FILTEREDPROFILE_PAGE|$CENSUSPROFILE_PAGE)"; + +# default binary name +my $UNKNOWN_BINARY = "(unknown)"; + +# There is a pervasive dependency on the length (in hex characters, +# i.e., nibbles) of an address, distinguishing between 32-bit and +# 64-bit profiles. To err on the safe size, default to 64-bit here: +my $address_length = 16; + +my $dev_null = "/dev/null"; +if (! -e $dev_null && $^O =~ /MSWin/) { # $^O is the OS perl was built for + $dev_null = "nul"; +} + +# A list of paths to search for shared object files +my @prefix_list = (); + +# Special routine name that should not have any symbols. +# Used as separator to parse "addr2line -i" output. +my $sep_symbol = '_fini'; +my $sep_address = undef; + +##### Argument parsing ##### + +sub usage_string { + return <<EOF; +Usage: +jeprof [options] <program> <profiles> + <profiles> is a space separated list of profile names. +jeprof [options] <symbolized-profiles> + <symbolized-profiles> is a list of profile files where each file contains + the necessary symbol mappings as well as profile data (likely generated + with --raw). +jeprof [options] <profile> + <profile> is a remote form. Symbols are obtained from host:port$SYMBOL_PAGE + + Each name can be: + /path/to/profile - a path to a profile file + host:port[/<service>] - a location of a service to get profile from + + The /<service> can be $HEAP_PAGE, $PROFILE_PAGE, /pprof/pmuprofile, + $GROWTH_PAGE, $CONTENTION_PAGE, /pprof/wall, + $CENSUSPROFILE_PAGE, or /pprof/filteredprofile. + For instance: + jeprof http://myserver.com:80$HEAP_PAGE + If /<service> is omitted, the service defaults to $PROFILE_PAGE (cpu profiling). +jeprof --symbols <program> + Maps addresses to symbol names. In this mode, stdin should be a + list of library mappings, in the same format as is found in the heap- + and cpu-profile files (this loosely matches that of /proc/self/maps + on linux), followed by a list of hex addresses to map, one per line. + + For more help with querying remote servers, including how to add the + necessary server-side support code, see this filename (or one like it): + + /usr/doc/gperftools-$PPROF_VERSION/pprof_remote_servers.html + +Options: + --cum Sort by cumulative data + --base=<base> Subtract <base> from <profile> before display + --interactive Run in interactive mode (interactive "help" gives help) [default] + --seconds=<n> Length of time for dynamic profiles [default=30 secs] + --add_lib=<file> Read additional symbols and line info from the given library + --lib_prefix=<dir> Comma separated list of library path prefixes + +Reporting Granularity: + --addresses Report at address level + --lines Report at source line level + --functions Report at function level [default] + --files Report at source file level + +Output type: + --text Generate text report + --callgrind Generate callgrind format to stdout + --gv Generate Postscript and display + --evince Generate PDF and display + --web Generate SVG and display + --list=<regexp> Generate source listing of matching routines + --disasm=<regexp> Generate disassembly of matching routines + --symbols Print demangled symbol names found at given addresses + --dot Generate DOT file to stdout + --ps Generate Postcript to stdout + --pdf Generate PDF to stdout + --svg Generate SVG to stdout + --gif Generate GIF to stdout + --raw Generate symbolized jeprof data (useful with remote fetch) + +Heap-Profile Options: + --inuse_space Display in-use (mega)bytes [default] + --inuse_objects Display in-use objects + --alloc_space Display allocated (mega)bytes + --alloc_objects Display allocated objects + --show_bytes Display space in bytes + --drop_negative Ignore negative differences + +Contention-profile options: + --total_delay Display total delay at each region [default] + --contentions Display number of delays at each region + --mean_delay Display mean delay at each region + +Call-graph Options: + --nodecount=<n> Show at most so many nodes [default=80] + --nodefraction=<f> Hide nodes below <f>*total [default=.005] + --edgefraction=<f> Hide edges below <f>*total [default=.001] + --maxdegree=<n> Max incoming/outgoing edges per node [default=8] + --focus=<regexp> Focus on backtraces with nodes matching <regexp> + --thread=<n> Show profile for thread <n> + --ignore=<regexp> Ignore backtraces with nodes matching <regexp> + --scale=<n> Set GV scaling [default=0] + --heapcheck Make nodes with non-0 object counts + (i.e. direct leak generators) more visible + --retain=<regexp> Retain only nodes that match <regexp> + --exclude=<regexp> Exclude all nodes that match <regexp> + +Miscellaneous: + --tools=<prefix or binary:fullpath>[,...] \$PATH for object tool pathnames + --test Run unit tests + --help This message + --version Version information + +Environment Variables: + JEPROF_TMPDIR Profiles directory. Defaults to \$HOME/jeprof + JEPROF_TOOLS Prefix for object tools pathnames + +Examples: + +jeprof /bin/ls ls.prof + Enters "interactive" mode +jeprof --text /bin/ls ls.prof + Outputs one line per procedure +jeprof --web /bin/ls ls.prof + Displays annotated call-graph in web browser +jeprof --gv /bin/ls ls.prof + Displays annotated call-graph via 'gv' +jeprof --gv --focus=Mutex /bin/ls ls.prof + Restricts to code paths including a .*Mutex.* entry +jeprof --gv --focus=Mutex --ignore=string /bin/ls ls.prof + Code paths including Mutex but not string +jeprof --list=getdir /bin/ls ls.prof + (Per-line) annotated source listing for getdir() +jeprof --disasm=getdir /bin/ls ls.prof + (Per-PC) annotated disassembly for getdir() + +jeprof http://localhost:1234/ + Enters "interactive" mode +jeprof --text localhost:1234 + Outputs one line per procedure for localhost:1234 +jeprof --raw localhost:1234 > ./local.raw +jeprof --text ./local.raw + Fetches a remote profile for later analysis and then + analyzes it in text mode. +EOF +} + +sub version_string { + return <<EOF +jeprof (part of jemalloc $JEPROF_VERSION) +based on pprof (part of gperftools $PPROF_VERSION) + +Copyright 1998-2007 Google Inc. + +This is BSD licensed software; see the source for copying conditions +and license information. +There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A +PARTICULAR PURPOSE. +EOF +} + +sub usage { + my $msg = shift; + print STDERR "$msg\n\n"; + print STDERR usage_string(); + print STDERR "\nFATAL ERROR: $msg\n"; # just as a reminder + exit(1); +} + +sub Init() { + # Setup tmp-file name and handler to clean it up. + # We do this in the very beginning so that we can use + # error() and cleanup() function anytime here after. + $main::tmpfile_sym = "/tmp/jeprof$$.sym"; + $main::tmpfile_ps = "/tmp/jeprof$$"; + $main::next_tmpfile = 0; + $SIG{'INT'} = \&sighandler; + + # Cache from filename/linenumber to source code + $main::source_cache = (); + + $main::opt_help = 0; + $main::opt_version = 0; + + $main::opt_cum = 0; + $main::opt_base = ''; + $main::opt_addresses = 0; + $main::opt_lines = 0; + $main::opt_functions = 0; + $main::opt_files = 0; + $main::opt_lib_prefix = ""; + + $main::opt_text = 0; + $main::opt_callgrind = 0; + $main::opt_list = ""; + $main::opt_disasm = ""; + $main::opt_symbols = 0; + $main::opt_gv = 0; + $main::opt_evince = 0; + $main::opt_web = 0; + $main::opt_dot = 0; + $main::opt_ps = 0; + $main::opt_pdf = 0; + $main::opt_gif = 0; + $main::opt_svg = 0; + $main::opt_raw = 0; + + $main::opt_nodecount = 80; + $main::opt_nodefraction = 0.005; + $main::opt_edgefraction = 0.001; + $main::opt_maxdegree = 8; + $main::opt_focus = ''; + $main::opt_thread = undef; + $main::opt_ignore = ''; + $main::opt_scale = 0; + $main::opt_heapcheck = 0; + $main::opt_retain = ''; + $main::opt_exclude = ''; + $main::opt_seconds = 30; + $main::opt_lib = ""; + + $main::opt_inuse_space = 0; + $main::opt_inuse_objects = 0; + $main::opt_alloc_space = 0; + $main::opt_alloc_objects = 0; + $main::opt_show_bytes = 0; + $main::opt_drop_negative = 0; + $main::opt_interactive = 0; + + $main::opt_total_delay = 0; + $main::opt_contentions = 0; + $main::opt_mean_delay = 0; + + $main::opt_tools = ""; + $main::opt_debug = 0; + $main::opt_test = 0; + + # These are undocumented flags used only by unittests. + $main::opt_test_stride = 0; + + # Are we using $SYMBOL_PAGE? + $main::use_symbol_page = 0; + + # Files returned by TempName. + %main::tempnames = (); + + # Type of profile we are dealing with + # Supported types: + # cpu + # heap + # growth + # contention + $main::profile_type = ''; # Empty type means "unknown" + + GetOptions("help!" => \$main::opt_help, + "version!" => \$main::opt_version, + "cum!" => \$main::opt_cum, + "base=s" => \$main::opt_base, + "seconds=i" => \$main::opt_seconds, + "add_lib=s" => \$main::opt_lib, + "lib_prefix=s" => \$main::opt_lib_prefix, + "functions!" => \$main::opt_functions, + "lines!" => \$main::opt_lines, + "addresses!" => \$main::opt_addresses, + "files!" => \$main::opt_files, + "text!" => \$main::opt_text, + "callgrind!" => \$main::opt_callgrind, + "list=s" => \$main::opt_list, + "disasm=s" => \$main::opt_disasm, + "symbols!" => \$main::opt_symbols, + "gv!" => \$main::opt_gv, + "evince!" => \$main::opt_evince, + "web!" => \$main::opt_web, + "dot!" => \$main::opt_dot, + "ps!" => \$main::opt_ps, + "pdf!" => \$main::opt_pdf, + "svg!" => \$main::opt_svg, + "gif!" => \$main::opt_gif, + "raw!" => \$main::opt_raw, + "interactive!" => \$main::opt_interactive, + "nodecount=i" => \$main::opt_nodecount, + "nodefraction=f" => \$main::opt_nodefraction, + "edgefraction=f" => \$main::opt_edgefraction, + "maxdegree=i" => \$main::opt_maxdegree, + "focus=s" => \$main::opt_focus, + "thread=s" => \$main::opt_thread, + "ignore=s" => \$main::opt_ignore, + "scale=i" => \$main::opt_scale, + "heapcheck" => \$main::opt_heapcheck, + "retain=s" => \$main::opt_retain, + "exclude=s" => \$main::opt_exclude, + "inuse_space!" => \$main::opt_inuse_space, + "inuse_objects!" => \$main::opt_inuse_objects, + "alloc_space!" => \$main::opt_alloc_space, + "alloc_objects!" => \$main::opt_alloc_objects, + "show_bytes!" => \$main::opt_show_bytes, + "drop_negative!" => \$main::opt_drop_negative, + "total_delay!" => \$main::opt_total_delay, + "contentions!" => \$main::opt_contentions, + "mean_delay!" => \$main::opt_mean_delay, + "tools=s" => \$main::opt_tools, + "test!" => \$main::opt_test, + "debug!" => \$main::opt_debug, + # Undocumented flags used only by unittests: + "test_stride=i" => \$main::opt_test_stride, + ) || usage("Invalid option(s)"); + + # Deal with the standard --help and --version + if ($main::opt_help) { + print usage_string(); + exit(0); + } + + if ($main::opt_version) { + print version_string(); + exit(0); + } + + # Disassembly/listing/symbols mode requires address-level info + if ($main::opt_disasm || $main::opt_list || $main::opt_symbols) { + $main::opt_functions = 0; + $main::opt_lines = 0; + $main::opt_addresses = 1; + $main::opt_files = 0; + } + + # Check heap-profiling flags + if ($main::opt_inuse_space + + $main::opt_inuse_objects + + $main::opt_alloc_space + + $main::opt_alloc_objects > 1) { + usage("Specify at most on of --inuse/--alloc options"); + } + + # Check output granularities + my $grains = + $main::opt_functions + + $main::opt_lines + + $main::opt_addresses + + $main::opt_files + + 0; + if ($grains > 1) { + usage("Only specify one output granularity option"); + } + if ($grains == 0) { + $main::opt_functions = 1; + } + + # Check output modes + my $modes = + $main::opt_text + + $main::opt_callgrind + + ($main::opt_list eq '' ? 0 : 1) + + ($main::opt_disasm eq '' ? 0 : 1) + + ($main::opt_symbols == 0 ? 0 : 1) + + $main::opt_gv + + $main::opt_evince + + $main::opt_web + + $main::opt_dot + + $main::opt_ps + + $main::opt_pdf + + $main::opt_svg + + $main::opt_gif + + $main::opt_raw + + $main::opt_interactive + + 0; + if ($modes > 1) { + usage("Only specify one output mode"); + } + if ($modes == 0) { + if (-t STDOUT) { # If STDOUT is a tty, activate interactive mode + $main::opt_interactive = 1; + } else { + $main::opt_text = 1; + } + } + + if ($main::opt_test) { + RunUnitTests(); + # Should not return + exit(1); + } + + # Binary name and profile arguments list + $main::prog = ""; + @main::pfile_args = (); + + # Remote profiling without a binary (using $SYMBOL_PAGE instead) + if (@ARGV > 0) { + if (IsProfileURL($ARGV[0])) { + $main::use_symbol_page = 1; + } elsif (IsSymbolizedProfileFile($ARGV[0])) { + $main::use_symbolized_profile = 1; + $main::prog = $UNKNOWN_BINARY; # will be set later from the profile file + } + } + + if ($main::use_symbol_page || $main::use_symbolized_profile) { + # We don't need a binary! + my %disabled = ('--lines' => $main::opt_lines, + '--disasm' => $main::opt_disasm); + for my $option (keys %disabled) { + usage("$option cannot be used without a binary") if $disabled{$option}; + } + # Set $main::prog later... + scalar(@ARGV) || usage("Did not specify profile file"); + } elsif ($main::opt_symbols) { + # --symbols needs a binary-name (to run nm on, etc) but not profiles + $main::prog = shift(@ARGV) || usage("Did not specify program"); + } else { + $main::prog = shift(@ARGV) || usage("Did not specify program"); + scalar(@ARGV) || usage("Did not specify profile file"); + } + + # Parse profile file/location arguments + foreach my $farg (@ARGV) { + if ($farg =~ m/(.*)\@([0-9]+)(|\/.*)$/ ) { + my $machine = $1; + my $num_machines = $2; + my $path = $3; + for (my $i = 0; $i < $num_machines; $i++) { + unshift(@main::pfile_args, "$i.$machine$path"); + } + } else { + unshift(@main::pfile_args, $farg); + } + } + + if ($main::use_symbol_page) { + unless (IsProfileURL($main::pfile_args[0])) { + error("The first profile should be a remote form to use $SYMBOL_PAGE\n"); + } + CheckSymbolPage(); + $main::prog = FetchProgramName(); + } elsif (!$main::use_symbolized_profile) { # may not need objtools! + ConfigureObjTools($main::prog) + } + + # Break the opt_lib_prefix into the prefix_list array + @prefix_list = split (',', $main::opt_lib_prefix); + + # Remove trailing / from the prefixes, in the list to prevent + # searching things like /my/path//lib/mylib.so + foreach (@prefix_list) { + s|/+$||; + } +} + +sub FilterAndPrint { + my ($profile, $symbols, $libs, $thread) = @_; + + # Get total data in profile + my $total = TotalProfile($profile); + + # Remove uniniteresting stack items + $profile = RemoveUninterestingFrames($symbols, $profile); + + # Focus? + if ($main::opt_focus ne '') { + $profile = FocusProfile($symbols, $profile, $main::opt_focus); + } + + # Ignore? + if ($main::opt_ignore ne '') { + $profile = IgnoreProfile($symbols, $profile, $main::opt_ignore); + } + + my $calls = ExtractCalls($symbols, $profile); + + # Reduce profiles to required output granularity, and also clean + # each stack trace so a given entry exists at most once. + my $reduced = ReduceProfile($symbols, $profile); + + # Get derived profiles + my $flat = FlatProfile($reduced); + my $cumulative = CumulativeProfile($reduced); + + # Print + if (!$main::opt_interactive) { + if ($main::opt_disasm) { + PrintDisassembly($libs, $flat, $cumulative, $main::opt_disasm); + } elsif ($main::opt_list) { + PrintListing($total, $libs, $flat, $cumulative, $main::opt_list, 0); + } elsif ($main::opt_text) { + # Make sure the output is empty when have nothing to report + # (only matters when --heapcheck is given but we must be + # compatible with old branches that did not pass --heapcheck always): + if ($total != 0) { + printf("Total%s: %s %s\n", + (defined($thread) ? " (t$thread)" : ""), + Unparse($total), Units()); + } + PrintText($symbols, $flat, $cumulative, -1); + } elsif ($main::opt_raw) { + PrintSymbolizedProfile($symbols, $profile, $main::prog); + } elsif ($main::opt_callgrind) { + PrintCallgrind($calls); + } else { + if (PrintDot($main::prog, $symbols, $profile, $flat, $cumulative, $total)) { + if ($main::opt_gv) { + RunGV(TempName($main::next_tmpfile, "ps"), ""); + } elsif ($main::opt_evince) { + RunEvince(TempName($main::next_tmpfile, "pdf"), ""); + } elsif ($main::opt_web) { + my $tmp = TempName($main::next_tmpfile, "svg"); + RunWeb($tmp); + # The command we run might hand the file name off + # to an already running browser instance and then exit. + # Normally, we'd remove $tmp on exit (right now), + # but fork a child to remove $tmp a little later, so that the + # browser has time to load it first. + delete $main::tempnames{$tmp}; + if (fork() == 0) { + sleep 5; + unlink($tmp); + exit(0); + } + } + } else { + cleanup(); + exit(1); + } + } + } else { + InteractiveMode($profile, $symbols, $libs, $total); + } +} + +sub Main() { + Init(); + $main::collected_profile = undef; + @main::profile_files = (); + $main::op_time = time(); + + # Printing symbols is special and requires a lot less info that most. + if ($main::opt_symbols) { + PrintSymbols(*STDIN); # Get /proc/maps and symbols output from stdin + return; + } + + # Fetch all profile data + FetchDynamicProfiles(); + + # this will hold symbols that we read from the profile files + my $symbol_map = {}; + + # Read one profile, pick the last item on the list + my $data = ReadProfile($main::prog, pop(@main::profile_files)); + my $profile = $data->{profile}; + my $pcs = $data->{pcs}; + my $libs = $data->{libs}; # Info about main program and shared libraries + $symbol_map = MergeSymbols($symbol_map, $data->{symbols}); + + # Add additional profiles, if available. + if (scalar(@main::profile_files) > 0) { + foreach my $pname (@main::profile_files) { + my $data2 = ReadProfile($main::prog, $pname); + $profile = AddProfile($profile, $data2->{profile}); + $pcs = AddPcs($pcs, $data2->{pcs}); + $symbol_map = MergeSymbols($symbol_map, $data2->{symbols}); + } + } + + # Subtract base from profile, if specified + if ($main::opt_base ne '') { + my $base = ReadProfile($main::prog, $main::opt_base); + $profile = SubtractProfile($profile, $base->{profile}); + $pcs = AddPcs($pcs, $base->{pcs}); + $symbol_map = MergeSymbols($symbol_map, $base->{symbols}); + } + + # Collect symbols + my $symbols; + if ($main::use_symbolized_profile) { + $symbols = FetchSymbols($pcs, $symbol_map); + } elsif ($main::use_symbol_page) { + $symbols = FetchSymbols($pcs); + } else { + # TODO(csilvers): $libs uses the /proc/self/maps data from profile1, + # which may differ from the data from subsequent profiles, especially + # if they were run on different machines. Use appropriate libs for + # each pc somehow. + $symbols = ExtractSymbols($libs, $pcs); + } + + if (!defined($main::opt_thread)) { + FilterAndPrint($profile, $symbols, $libs); + } + if (defined($data->{threads})) { + foreach my $thread (sort { $a <=> $b } keys(%{$data->{threads}})) { + if (defined($main::opt_thread) && + ($main::opt_thread eq '*' || $main::opt_thread == $thread)) { + my $thread_profile = $data->{threads}{$thread}; + FilterAndPrint($thread_profile, $symbols, $libs, $thread); + } + } + } + + cleanup(); + exit(0); +} + +##### Entry Point ##### + +Main(); + +# Temporary code to detect if we're running on a Goobuntu system. +# These systems don't have the right stuff installed for the special +# Readline libraries to work, so as a temporary workaround, we default +# to using the normal stdio code, rather than the fancier readline-based +# code +sub ReadlineMightFail { + if (-e '/lib/libtermcap.so.2') { + return 0; # libtermcap exists, so readline should be okay + } else { + return 1; + } +} + +sub RunGV { + my $fname = shift; + my $bg = shift; # "" or " &" if we should run in background + if (!system(ShellEscape(@GV, "--version") . " >$dev_null 2>&1")) { + # Options using double dash are supported by this gv version. + # Also, turn on noantialias to better handle bug in gv for + # postscript files with large dimensions. + # TODO: Maybe we should not pass the --noantialias flag + # if the gv version is known to work properly without the flag. + system(ShellEscape(@GV, "--scale=$main::opt_scale", "--noantialias", $fname) + . $bg); + } else { + # Old gv version - only supports options that use single dash. + print STDERR ShellEscape(@GV, "-scale", $main::opt_scale) . "\n"; + system(ShellEscape(@GV, "-scale", "$main::opt_scale", $fname) . $bg); + } +} + +sub RunEvince { + my $fname = shift; + my $bg = shift; # "" or " &" if we should run in background + system(ShellEscape(@EVINCE, $fname) . $bg); +} + +sub RunWeb { + my $fname = shift; + print STDERR "Loading web page file:///$fname\n"; + + if (`uname` =~ /Darwin/) { + # OS X: open will use standard preference for SVG files. + system("/usr/bin/open", $fname); + return; + } + + # Some kind of Unix; try generic symlinks, then specific browsers. + # (Stop once we find one.) + # Works best if the browser is already running. + my @alt = ( + "/etc/alternatives/gnome-www-browser", + "/etc/alternatives/x-www-browser", + "google-chrome", + "firefox", + ); + foreach my $b (@alt) { + if (system($b, $fname) == 0) { + return; + } + } + + print STDERR "Could not load web browser.\n"; +} + +sub RunKcachegrind { + my $fname = shift; + my $bg = shift; # "" or " &" if we should run in background + print STDERR "Starting '@KCACHEGRIND " . $fname . $bg . "'\n"; + system(ShellEscape(@KCACHEGRIND, $fname) . $bg); +} + + +##### Interactive helper routines ##### + +sub InteractiveMode { + $| = 1; # Make output unbuffered for interactive mode + my ($orig_profile, $symbols, $libs, $total) = @_; + + print STDERR "Welcome to jeprof! For help, type 'help'.\n"; + + # Use ReadLine if it's installed and input comes from a console. + if ( -t STDIN && + !ReadlineMightFail() && + defined(eval {require Term::ReadLine}) ) { + my $term = new Term::ReadLine 'jeprof'; + while ( defined ($_ = $term->readline('(jeprof) '))) { + $term->addhistory($_) if /\S/; + if (!InteractiveCommand($orig_profile, $symbols, $libs, $total, $_)) { + last; # exit when we get an interactive command to quit + } + } + } else { # don't have readline + while (1) { + print STDERR "(jeprof) "; + $_ = <STDIN>; + last if ! defined $_ ; + s/\r//g; # turn windows-looking lines into unix-looking lines + + # Save some flags that might be reset by InteractiveCommand() + my $save_opt_lines = $main::opt_lines; + + if (!InteractiveCommand($orig_profile, $symbols, $libs, $total, $_)) { + last; # exit when we get an interactive command to quit + } + + # Restore flags + $main::opt_lines = $save_opt_lines; + } + } +} + +# Takes two args: orig profile, and command to run. +# Returns 1 if we should keep going, or 0 if we were asked to quit +sub InteractiveCommand { + my($orig_profile, $symbols, $libs, $total, $command) = @_; + $_ = $command; # just to make future m//'s easier + if (!defined($_)) { + print STDERR "\n"; + return 0; + } + if (m/^\s*quit/) { + return 0; + } + if (m/^\s*help/) { + InteractiveHelpMessage(); + return 1; + } + # Clear all the mode options -- mode is controlled by "$command" + $main::opt_text = 0; + $main::opt_callgrind = 0; + $main::opt_disasm = 0; + $main::opt_list = 0; + $main::opt_gv = 0; + $main::opt_evince = 0; + $main::opt_cum = 0; + + if (m/^\s*(text|top)(\d*)\s*(.*)/) { + $main::opt_text = 1; + + my $line_limit = ($2 ne "") ? int($2) : 10; + + my $routine; + my $ignore; + ($routine, $ignore) = ParseInteractiveArgs($3); + + my $profile = ProcessProfile($total, $orig_profile, $symbols, "", $ignore); + my $reduced = ReduceProfile($symbols, $profile); + + # Get derived profiles + my $flat = FlatProfile($reduced); + my $cumulative = CumulativeProfile($reduced); + + PrintText($symbols, $flat, $cumulative, $line_limit); + return 1; + } + if (m/^\s*callgrind\s*([^ \n]*)/) { + $main::opt_callgrind = 1; + + # Get derived profiles + my $calls = ExtractCalls($symbols, $orig_profile); + my $filename = $1; + if ( $1 eq '' ) { + $filename = TempName($main::next_tmpfile, "callgrind"); + } + PrintCallgrind($calls, $filename); + if ( $1 eq '' ) { + RunKcachegrind($filename, " & "); + $main::next_tmpfile++; + } + + return 1; + } + if (m/^\s*(web)?list\s*(.+)/) { + my $html = (defined($1) && ($1 eq "web")); + $main::opt_list = 1; + + my $routine; + my $ignore; + ($routine, $ignore) = ParseInteractiveArgs($2); + + my $profile = ProcessProfile($total, $orig_profile, $symbols, "", $ignore); + my $reduced = ReduceProfile($symbols, $profile); + + # Get derived profiles + my $flat = FlatProfile($reduced); + my $cumulative = CumulativeProfile($reduced); + + PrintListing($total, $libs, $flat, $cumulative, $routine, $html); + return 1; + } + if (m/^\s*disasm\s*(.+)/) { + $main::opt_disasm = 1; + + my $routine; + my $ignore; + ($routine, $ignore) = ParseInteractiveArgs($1); + + # Process current profile to account for various settings + my $profile = ProcessProfile($total, $orig_profile, $symbols, "", $ignore); + my $reduced = ReduceProfile($symbols, $profile); + + # Get derived profiles + my $flat = FlatProfile($reduced); + my $cumulative = CumulativeProfile($reduced); + + PrintDisassembly($libs, $flat, $cumulative, $routine); + return 1; + } + if (m/^\s*(gv|web|evince)\s*(.*)/) { + $main::opt_gv = 0; + $main::opt_evince = 0; + $main::opt_web = 0; + if ($1 eq "gv") { + $main::opt_gv = 1; + } elsif ($1 eq "evince") { + $main::opt_evince = 1; + } elsif ($1 eq "web") { + $main::opt_web = 1; + } + + my $focus; + my $ignore; + ($focus, $ignore) = ParseInteractiveArgs($2); + + # Process current profile to account for various settings + my $profile = ProcessProfile($total, $orig_profile, $symbols, + $focus, $ignore); + my $reduced = ReduceProfile($symbols, $profile); + + # Get derived profiles + my $flat = FlatProfile($reduced); + my $cumulative = CumulativeProfile($reduced); + + if (PrintDot($main::prog, $symbols, $profile, $flat, $cumulative, $total)) { + if ($main::opt_gv) { + RunGV(TempName($main::next_tmpfile, "ps"), " &"); + } elsif ($main::opt_evince) { + RunEvince(TempName($main::next_tmpfile, "pdf"), " &"); + } elsif ($main::opt_web) { + RunWeb(TempName($main::next_tmpfile, "svg")); + } + $main::next_tmpfile++; + } + return 1; + } + if (m/^\s*$/) { + return 1; + } + print STDERR "Unknown command: try 'help'.\n"; + return 1; +} + + +sub ProcessProfile { + my $total_count = shift; + my $orig_profile = shift; + my $symbols = shift; + my $focus = shift; + my $ignore = shift; + + # Process current profile to account for various settings + my $profile = $orig_profile; + printf("Total: %s %s\n", Unparse($total_count), Units()); + if ($focus ne '') { + $profile = FocusProfile($symbols, $profile, $focus); + my $focus_count = TotalProfile($profile); + printf("After focusing on '%s': %s %s of %s (%0.1f%%)\n", + $focus, + Unparse($focus_count), Units(), + Unparse($total_count), ($focus_count*100.0) / $total_count); + } + if ($ignore ne '') { + $profile = IgnoreProfile($symbols, $profile, $ignore); + my $ignore_count = TotalProfile($profile); + printf("After ignoring '%s': %s %s of %s (%0.1f%%)\n", + $ignore, + Unparse($ignore_count), Units(), + Unparse($total_count), + ($ignore_count*100.0) / $total_count); + } + + return $profile; +} + +sub InteractiveHelpMessage { + print STDERR <<ENDOFHELP; +Interactive jeprof mode + +Commands: + gv + gv [focus] [-ignore1] [-ignore2] + Show graphical hierarchical display of current profile. Without + any arguments, shows all samples in the profile. With the optional + "focus" argument, restricts the samples shown to just those where + the "focus" regular expression matches a routine name on the stack + trace. + + web + web [focus] [-ignore1] [-ignore2] + Like GV, but displays profile in your web browser instead of using + Ghostview. Works best if your web browser is already running. + To change the browser that gets used: + On Linux, set the /etc/alternatives/gnome-www-browser symlink. + On OS X, change the Finder association for SVG files. + + list [routine_regexp] [-ignore1] [-ignore2] + Show source listing of routines whose names match "routine_regexp" + + weblist [routine_regexp] [-ignore1] [-ignore2] + Displays a source listing of routines whose names match "routine_regexp" + in a web browser. You can click on source lines to view the + corresponding disassembly. + + top [--cum] [-ignore1] [-ignore2] + top20 [--cum] [-ignore1] [-ignore2] + top37 [--cum] [-ignore1] [-ignore2] + Show top lines ordered by flat profile count, or cumulative count + if --cum is specified. If a number is present after 'top', the + top K routines will be shown (defaults to showing the top 10) + + disasm [routine_regexp] [-ignore1] [-ignore2] + Show disassembly of routines whose names match "routine_regexp", + annotated with sample counts. + + callgrind + callgrind [filename] + Generates callgrind file. If no filename is given, kcachegrind is called. + + help - This listing + quit or ^D - End jeprof + +For commands that accept optional -ignore tags, samples where any routine in +the stack trace matches the regular expression in any of the -ignore +parameters will be ignored. + +Further pprof details are available at this location (or one similar): + + /usr/doc/gperftools-$PPROF_VERSION/cpu_profiler.html + /usr/doc/gperftools-$PPROF_VERSION/heap_profiler.html + +ENDOFHELP +} +sub ParseInteractiveArgs { + my $args = shift; + my $focus = ""; + my $ignore = ""; + my @x = split(/ +/, $args); + foreach $a (@x) { + if ($a =~ m/^(--|-)lines$/) { + $main::opt_lines = 1; + } elsif ($a =~ m/^(--|-)cum$/) { + $main::opt_cum = 1; + } elsif ($a =~ m/^-(.*)/) { + $ignore .= (($ignore ne "") ? "|" : "" ) . $1; + } else { + $focus .= (($focus ne "") ? "|" : "" ) . $a; + } + } + if ($ignore ne "") { + print STDERR "Ignoring samples in call stacks that match '$ignore'\n"; + } + return ($focus, $ignore); +} + +##### Output code ##### + +sub TempName { + my $fnum = shift; + my $ext = shift; + my $file = "$main::tmpfile_ps.$fnum.$ext"; + $main::tempnames{$file} = 1; + return $file; +} + +# Print profile data in packed binary format (64-bit) to standard out +sub PrintProfileData { + my $profile = shift; + + # print header (64-bit style) + # (zero) (header-size) (version) (sample-period) (zero) + print pack('L*', 0, 0, 3, 0, 0, 0, 1, 0, 0, 0); + + foreach my $k (keys(%{$profile})) { + my $count = $profile->{$k}; + my @addrs = split(/\n/, $k); + if ($#addrs >= 0) { + my $depth = $#addrs + 1; + # int(foo / 2**32) is the only reliable way to get rid of bottom + # 32 bits on both 32- and 64-bit systems. + print pack('L*', $count & 0xFFFFFFFF, int($count / 2**32)); + print pack('L*', $depth & 0xFFFFFFFF, int($depth / 2**32)); + + foreach my $full_addr (@addrs) { + my $addr = $full_addr; + $addr =~ s/0x0*//; # strip off leading 0x, zeroes + if (length($addr) > 16) { + print STDERR "Invalid address in profile: $full_addr\n"; + next; + } + my $low_addr = substr($addr, -8); # get last 8 hex chars + my $high_addr = substr($addr, -16, 8); # get up to 8 more hex chars + print pack('L*', hex('0x' . $low_addr), hex('0x' . $high_addr)); + } + } + } +} + +# Print symbols and profile data +sub PrintSymbolizedProfile { + my $symbols = shift; + my $profile = shift; + my $prog = shift; + + $SYMBOL_PAGE =~ m,[^/]+$,; # matches everything after the last slash + my $symbol_marker = $&; + + print '--- ', $symbol_marker, "\n"; + if (defined($prog)) { + print 'binary=', $prog, "\n"; + } + while (my ($pc, $name) = each(%{$symbols})) { + my $sep = ' '; + print '0x', $pc; + # We have a list of function names, which include the inlined + # calls. They are separated (and terminated) by --, which is + # illegal in function names. + for (my $j = 2; $j <= $#{$name}; $j += 3) { + print $sep, $name->[$j]; + $sep = '--'; + } + print "\n"; + } + print '---', "\n"; + + my $profile_marker; + if ($main::profile_type eq 'heap') { + $HEAP_PAGE =~ m,[^/]+$,; # matches everything after the last slash + $profile_marker = $&; + } elsif ($main::profile_type eq 'growth') { + $GROWTH_PAGE =~ m,[^/]+$,; # matches everything after the last slash + $profile_marker = $&; + } elsif ($main::profile_type eq 'contention') { + $CONTENTION_PAGE =~ m,[^/]+$,; # matches everything after the last slash + $profile_marker = $&; + } else { # elsif ($main::profile_type eq 'cpu') + $PROFILE_PAGE =~ m,[^/]+$,; # matches everything after the last slash + $profile_marker = $&; + } + + print '--- ', $profile_marker, "\n"; + if (defined($main::collected_profile)) { + # if used with remote fetch, simply dump the collected profile to output. + open(SRC, "<$main::collected_profile"); + while (<SRC>) { + print $_; + } + close(SRC); + } else { + # --raw/http: For everything to work correctly for non-remote profiles, we + # would need to extend PrintProfileData() to handle all possible profile + # types, re-enable the code that is currently disabled in ReadCPUProfile() + # and FixCallerAddresses(), and remove the remote profile dumping code in + # the block above. + die "--raw/http: jeprof can only dump remote profiles for --raw\n"; + # dump a cpu-format profile to standard out + PrintProfileData($profile); + } +} + +# Print text output +sub PrintText { + my $symbols = shift; + my $flat = shift; + my $cumulative = shift; + my $line_limit = shift; + + my $total = TotalProfile($flat); + + # Which profile to sort by? + my $s = $main::opt_cum ? $cumulative : $flat; + + my $running_sum = 0; + my $lines = 0; + foreach my $k (sort { GetEntry($s, $b) <=> GetEntry($s, $a) || $a cmp $b } + keys(%{$cumulative})) { + my $f = GetEntry($flat, $k); + my $c = GetEntry($cumulative, $k); + $running_sum += $f; + + my $sym = $k; + if (exists($symbols->{$k})) { + $sym = $symbols->{$k}->[0] . " " . $symbols->{$k}->[1]; + if ($main::opt_addresses) { + $sym = $k . " " . $sym; + } + } + + if ($f != 0 || $c != 0) { + printf("%8s %6s %6s %8s %6s %s\n", + Unparse($f), + Percent($f, $total), + Percent($running_sum, $total), + Unparse($c), + Percent($c, $total), + $sym); + } + $lines++; + last if ($line_limit >= 0 && $lines >= $line_limit); + } +} + +# Callgrind format has a compression for repeated function and file +# names. You show the name the first time, and just use its number +# subsequently. This can cut down the file to about a third or a +# quarter of its uncompressed size. $key and $val are the key/value +# pair that would normally be printed by callgrind; $map is a map from +# value to number. +sub CompressedCGName { + my($key, $val, $map) = @_; + my $idx = $map->{$val}; + # For very short keys, providing an index hurts rather than helps. + if (length($val) <= 3) { + return "$key=$val\n"; + } elsif (defined($idx)) { + return "$key=($idx)\n"; + } else { + # scalar(keys $map) gives the number of items in the map. + $idx = scalar(keys(%{$map})) + 1; + $map->{$val} = $idx; + return "$key=($idx) $val\n"; + } +} + +# Print the call graph in a way that's suiteable for callgrind. +sub PrintCallgrind { + my $calls = shift; + my $filename; + my %filename_to_index_map; + my %fnname_to_index_map; + + if ($main::opt_interactive) { + $filename = shift; + print STDERR "Writing callgrind file to '$filename'.\n" + } else { + $filename = "&STDOUT"; + } + open(CG, ">$filename"); + printf CG ("events: Hits\n\n"); + foreach my $call ( map { $_->[0] } + sort { $a->[1] cmp $b ->[1] || + $a->[2] <=> $b->[2] } + map { /([^:]+):(\d+):([^ ]+)( -> ([^:]+):(\d+):(.+))?/; + [$_, $1, $2] } + keys %$calls ) { + my $count = int($calls->{$call}); + $call =~ /([^:]+):(\d+):([^ ]+)( -> ([^:]+):(\d+):(.+))?/; + my ( $caller_file, $caller_line, $caller_function, + $callee_file, $callee_line, $callee_function ) = + ( $1, $2, $3, $5, $6, $7 ); + + # TODO(csilvers): for better compression, collect all the + # caller/callee_files and functions first, before printing + # anything, and only compress those referenced more than once. + printf CG CompressedCGName("fl", $caller_file, \%filename_to_index_map); + printf CG CompressedCGName("fn", $caller_function, \%fnname_to_index_map); + if (defined $6) { + printf CG CompressedCGName("cfl", $callee_file, \%filename_to_index_map); + printf CG CompressedCGName("cfn", $callee_function, \%fnname_to_index_map); + printf CG ("calls=$count $callee_line\n"); + } + printf CG ("$caller_line $count\n\n"); + } +} + +# Print disassembly for all all routines that match $main::opt_disasm +sub PrintDisassembly { + my $libs = shift; + my $flat = shift; + my $cumulative = shift; + my $disasm_opts = shift; + + my $total = TotalProfile($flat); + + foreach my $lib (@{$libs}) { + my $symbol_table = GetProcedureBoundaries($lib->[0], $disasm_opts); + my $offset = AddressSub($lib->[1], $lib->[3]); + foreach my $routine (sort ByName keys(%{$symbol_table})) { + my $start_addr = $symbol_table->{$routine}->[0]; + my $end_addr = $symbol_table->{$routine}->[1]; + # See if there are any samples in this routine + my $length = hex(AddressSub($end_addr, $start_addr)); + my $addr = AddressAdd($start_addr, $offset); + for (my $i = 0; $i < $length; $i++) { + if (defined($cumulative->{$addr})) { + PrintDisassembledFunction($lib->[0], $offset, + $routine, $flat, $cumulative, + $start_addr, $end_addr, $total); + last; + } + $addr = AddressInc($addr); + } + } + } +} + +# Return reference to array of tuples of the form: +# [start_address, filename, linenumber, instruction, limit_address] +# E.g., +# ["0x806c43d", "/foo/bar.cc", 131, "ret", "0x806c440"] +sub Disassemble { + my $prog = shift; + my $offset = shift; + my $start_addr = shift; + my $end_addr = shift; + + my $objdump = $obj_tool_map{"objdump"}; + my $cmd = ShellEscape($objdump, "-C", "-d", "-l", "--no-show-raw-insn", + "--start-address=0x$start_addr", + "--stop-address=0x$end_addr", $prog); + open(OBJDUMP, "$cmd |") || error("$cmd: $!\n"); + my @result = (); + my $filename = ""; + my $linenumber = -1; + my $last = ["", "", "", ""]; + while (<OBJDUMP>) { + s/\r//g; # turn windows-looking lines into unix-looking lines + chop; + if (m|\s*([^:\s]+):(\d+)\s*$|) { + # Location line of the form: + # <filename>:<linenumber> + $filename = $1; + $linenumber = $2; + } elsif (m/^ +([0-9a-f]+):\s*(.*)/) { + # Disassembly line -- zero-extend address to full length + my $addr = HexExtend($1); + my $k = AddressAdd($addr, $offset); + $last->[4] = $k; # Store ending address for previous instruction + $last = [$k, $filename, $linenumber, $2, $end_addr]; + push(@result, $last); + } + } + close(OBJDUMP); + return @result; +} + +# The input file should contain lines of the form /proc/maps-like +# output (same format as expected from the profiles) or that looks +# like hex addresses (like "0xDEADBEEF"). We will parse all +# /proc/maps output, and for all the hex addresses, we will output +# "short" symbol names, one per line, in the same order as the input. +sub PrintSymbols { + my $maps_and_symbols_file = shift; + + # ParseLibraries expects pcs to be in a set. Fine by us... + my @pclist = (); # pcs in sorted order + my $pcs = {}; + my $map = ""; + foreach my $line (<$maps_and_symbols_file>) { + $line =~ s/\r//g; # turn windows-looking lines into unix-looking lines + if ($line =~ /\b(0x[0-9a-f]+)\b/i) { + push(@pclist, HexExtend($1)); + $pcs->{$pclist[-1]} = 1; + } else { + $map .= $line; + } + } + + my $libs = ParseLibraries($main::prog, $map, $pcs); + my $symbols = ExtractSymbols($libs, $pcs); + + foreach my $pc (@pclist) { + # ->[0] is the shortname, ->[2] is the full name + print(($symbols->{$pc}->[0] || "??") . "\n"); + } +} + + +# For sorting functions by name +sub ByName { + return ShortFunctionName($a) cmp ShortFunctionName($b); +} + +# Print source-listing for all all routines that match $list_opts +sub PrintListing { + my $total = shift; + my $libs = shift; + my $flat = shift; + my $cumulative = shift; + my $list_opts = shift; + my $html = shift; + + my $output = \*STDOUT; + my $fname = ""; + + if ($html) { + # Arrange to write the output to a temporary file + $fname = TempName($main::next_tmpfile, "html"); + $main::next_tmpfile++; + if (!open(TEMP, ">$fname")) { + print STDERR "$fname: $!\n"; + return; + } + $output = \*TEMP; + print $output HtmlListingHeader(); + printf $output ("<div class=\"legend\">%s<br>Total: %s %s</div>\n", + $main::prog, Unparse($total), Units()); + } + + my $listed = 0; + foreach my $lib (@{$libs}) { + my $symbol_table = GetProcedureBoundaries($lib->[0], $list_opts); + my $offset = AddressSub($lib->[1], $lib->[3]); + foreach my $routine (sort ByName keys(%{$symbol_table})) { + # Print if there are any samples in this routine + my $start_addr = $symbol_table->{$routine}->[0]; + my $end_addr = $symbol_table->{$routine}->[1]; + my $length = hex(AddressSub($end_addr, $start_addr)); + my $addr = AddressAdd($start_addr, $offset); + for (my $i = 0; $i < $length; $i++) { + if (defined($cumulative->{$addr})) { + $listed += PrintSource( + $lib->[0], $offset, + $routine, $flat, $cumulative, + $start_addr, $end_addr, + $html, + $output); + last; + } + $addr = AddressInc($addr); + } + } + } + + if ($html) { + if ($listed > 0) { + print $output HtmlListingFooter(); + close($output); + RunWeb($fname); + } else { + close($output); + unlink($fname); + } + } +} + +sub HtmlListingHeader { + return <<'EOF'; +<DOCTYPE html> +<html> +<head> +<title>Pprof listing</title> +<style type="text/css"> +body { + font-family: sans-serif; +} +h1 { + font-size: 1.5em; + margin-bottom: 4px; +} +.legend { + font-size: 1.25em; +} +.line { + color: #aaaaaa; +} +.nop { + color: #aaaaaa; +} +.unimportant { + color: #cccccc; +} +.disasmloc { + color: #000000; +} +.deadsrc { + cursor: pointer; +} +.deadsrc:hover { + background-color: #eeeeee; +} +.livesrc { + color: #0000ff; + cursor: pointer; +} +.livesrc:hover { + background-color: #eeeeee; +} +.asm { + color: #008800; + display: none; +} +</style> +<script type="text/javascript"> +function jeprof_toggle_asm(e) { + var target; + if (!e) e = window.event; + if (e.target) target = e.target; + else if (e.srcElement) target = e.srcElement; + + if (target) { + var asm = target.nextSibling; + if (asm && asm.className == "asm") { + asm.style.display = (asm.style.display == "block" ? "" : "block"); + e.preventDefault(); + return false; + } + } +} +</script> +</head> +<body> +EOF +} + +sub HtmlListingFooter { + return <<'EOF'; +</body> +</html> +EOF +} + +sub HtmlEscape { + my $text = shift; + $text =~ s/&/&/g; + $text =~ s/</</g; + $text =~ s/>/>/g; + return $text; +} + +# Returns the indentation of the line, if it has any non-whitespace +# characters. Otherwise, returns -1. +sub Indentation { + my $line = shift; + if (m/^(\s*)\S/) { + return length($1); + } else { + return -1; + } +} + +# If the symbol table contains inlining info, Disassemble() may tag an +# instruction with a location inside an inlined function. But for +# source listings, we prefer to use the location in the function we +# are listing. So use MapToSymbols() to fetch full location +# information for each instruction and then pick out the first +# location from a location list (location list contains callers before +# callees in case of inlining). +# +# After this routine has run, each entry in $instructions contains: +# [0] start address +# [1] filename for function we are listing +# [2] line number for function we are listing +# [3] disassembly +# [4] limit address +# [5] most specific filename (may be different from [1] due to inlining) +# [6] most specific line number (may be different from [2] due to inlining) +sub GetTopLevelLineNumbers { + my ($lib, $offset, $instructions) = @_; + my $pcs = []; + for (my $i = 0; $i <= $#{$instructions}; $i++) { + push(@{$pcs}, $instructions->[$i]->[0]); + } + my $symbols = {}; + MapToSymbols($lib, $offset, $pcs, $symbols); + for (my $i = 0; $i <= $#{$instructions}; $i++) { + my $e = $instructions->[$i]; + push(@{$e}, $e->[1]); + push(@{$e}, $e->[2]); + my $addr = $e->[0]; + my $sym = $symbols->{$addr}; + if (defined($sym)) { + if ($#{$sym} >= 2 && $sym->[1] =~ m/^(.*):(\d+)$/) { + $e->[1] = $1; # File name + $e->[2] = $2; # Line number + } + } + } +} + +# Print source-listing for one routine +sub PrintSource { + my $prog = shift; + my $offset = shift; + my $routine = shift; + my $flat = shift; + my $cumulative = shift; + my $start_addr = shift; + my $end_addr = shift; + my $html = shift; + my $output = shift; + + # Disassemble all instructions (just to get line numbers) + my @instructions = Disassemble($prog, $offset, $start_addr, $end_addr); + GetTopLevelLineNumbers($prog, $offset, \@instructions); + + # Hack 1: assume that the first source file encountered in the + # disassembly contains the routine + my $filename = undef; + for (my $i = 0; $i <= $#instructions; $i++) { + if ($instructions[$i]->[2] >= 0) { + $filename = $instructions[$i]->[1]; + last; + } + } + if (!defined($filename)) { + print STDERR "no filename found in $routine\n"; + return 0; + } + + # Hack 2: assume that the largest line number from $filename is the + # end of the procedure. This is typically safe since if P1 contains + # an inlined call to P2, then P2 usually occurs earlier in the + # source file. If this does not work, we might have to compute a + # density profile or just print all regions we find. + my $lastline = 0; + for (my $i = 0; $i <= $#instructions; $i++) { + my $f = $instructions[$i]->[1]; + my $l = $instructions[$i]->[2]; + if (($f eq $filename) && ($l > $lastline)) { + $lastline = $l; + } + } + + # Hack 3: assume the first source location from "filename" is the start of + # the source code. + my $firstline = 1; + for (my $i = 0; $i <= $#instructions; $i++) { + if ($instructions[$i]->[1] eq $filename) { + $firstline = $instructions[$i]->[2]; + last; + } + } + + # Hack 4: Extend last line forward until its indentation is less than + # the indentation we saw on $firstline + my $oldlastline = $lastline; + { + if (!open(FILE, "<$filename")) { + print STDERR "$filename: $!\n"; + return 0; + } + my $l = 0; + my $first_indentation = -1; + while (<FILE>) { + s/\r//g; # turn windows-looking lines into unix-looking lines + $l++; + my $indent = Indentation($_); + if ($l >= $firstline) { + if ($first_indentation < 0 && $indent >= 0) { + $first_indentation = $indent; + last if ($first_indentation == 0); + } + } + if ($l >= $lastline && $indent >= 0) { + if ($indent >= $first_indentation) { + $lastline = $l+1; + } else { + last; + } + } + } + close(FILE); + } + + # Assign all samples to the range $firstline,$lastline, + # Hack 4: If an instruction does not occur in the range, its samples + # are moved to the next instruction that occurs in the range. + my $samples1 = {}; # Map from line number to flat count + my $samples2 = {}; # Map from line number to cumulative count + my $running1 = 0; # Unassigned flat counts + my $running2 = 0; # Unassigned cumulative counts + my $total1 = 0; # Total flat counts + my $total2 = 0; # Total cumulative counts + my %disasm = (); # Map from line number to disassembly + my $running_disasm = ""; # Unassigned disassembly + my $skip_marker = "---\n"; + if ($html) { + $skip_marker = ""; + for (my $l = $firstline; $l <= $lastline; $l++) { + $disasm{$l} = ""; + } + } + my $last_dis_filename = ''; + my $last_dis_linenum = -1; + my $last_touched_line = -1; # To detect gaps in disassembly for a line + foreach my $e (@instructions) { + # Add up counts for all address that fall inside this instruction + my $c1 = 0; + my $c2 = 0; + for (my $a = $e->[0]; $a lt $e->[4]; $a = AddressInc($a)) { + $c1 += GetEntry($flat, $a); + $c2 += GetEntry($cumulative, $a); + } + + if ($html) { + my $dis = sprintf(" %6s %6s \t\t%8s: %s ", + HtmlPrintNumber($c1), + HtmlPrintNumber($c2), + UnparseAddress($offset, $e->[0]), + CleanDisassembly($e->[3])); + + # Append the most specific source line associated with this instruction + if (length($dis) < 80) { $dis .= (' ' x (80 - length($dis))) }; + $dis = HtmlEscape($dis); + my $f = $e->[5]; + my $l = $e->[6]; + if ($f ne $last_dis_filename) { + $dis .= sprintf("<span class=disasmloc>%s:%d</span>", + HtmlEscape(CleanFileName($f)), $l); + } elsif ($l ne $last_dis_linenum) { + # De-emphasize the unchanged file name portion + $dis .= sprintf("<span class=unimportant>%s</span>" . + "<span class=disasmloc>:%d</span>", + HtmlEscape(CleanFileName($f)), $l); + } else { + # De-emphasize the entire location + $dis .= sprintf("<span class=unimportant>%s:%d</span>", + HtmlEscape(CleanFileName($f)), $l); + } + $last_dis_filename = $f; + $last_dis_linenum = $l; + $running_disasm .= $dis; + $running_disasm .= "\n"; + } + + $running1 += $c1; + $running2 += $c2; + $total1 += $c1; + $total2 += $c2; + my $file = $e->[1]; + my $line = $e->[2]; + if (($file eq $filename) && + ($line >= $firstline) && + ($line <= $lastline)) { + # Assign all accumulated samples to this line + AddEntry($samples1, $line, $running1); + AddEntry($samples2, $line, $running2); + $running1 = 0; + $running2 = 0; + if ($html) { + if ($line != $last_touched_line && $disasm{$line} ne '') { + $disasm{$line} .= "\n"; + } + $disasm{$line} .= $running_disasm; + $running_disasm = ''; + $last_touched_line = $line; + } + } + } + + # Assign any leftover samples to $lastline + AddEntry($samples1, $lastline, $running1); + AddEntry($samples2, $lastline, $running2); + if ($html) { + if ($lastline != $last_touched_line && $disasm{$lastline} ne '') { + $disasm{$lastline} .= "\n"; + } + $disasm{$lastline} .= $running_disasm; + } + + if ($html) { + printf $output ( + "<h1>%s</h1>%s\n<pre onClick=\"jeprof_toggle_asm()\">\n" . + "Total:%6s %6s (flat / cumulative %s)\n", + HtmlEscape(ShortFunctionName($routine)), + HtmlEscape(CleanFileName($filename)), + Unparse($total1), + Unparse($total2), + Units()); + } else { + printf $output ( + "ROUTINE ====================== %s in %s\n" . + "%6s %6s Total %s (flat / cumulative)\n", + ShortFunctionName($routine), + CleanFileName($filename), + Unparse($total1), + Unparse($total2), + Units()); + } + if (!open(FILE, "<$filename")) { + print STDERR "$filename: $!\n"; + return 0; + } + my $l = 0; + while (<FILE>) { + s/\r//g; # turn windows-looking lines into unix-looking lines + $l++; + if ($l >= $firstline - 5 && + (($l <= $oldlastline + 5) || ($l <= $lastline))) { + chop; + my $text = $_; + if ($l == $firstline) { print $output $skip_marker; } + my $n1 = GetEntry($samples1, $l); + my $n2 = GetEntry($samples2, $l); + if ($html) { + # Emit a span that has one of the following classes: + # livesrc -- has samples + # deadsrc -- has disassembly, but with no samples + # nop -- has no matching disasembly + # Also emit an optional span containing disassembly. + my $dis = $disasm{$l}; + my $asm = ""; + if (defined($dis) && $dis ne '') { + $asm = "<span class=\"asm\">" . $dis . "</span>"; + } + my $source_class = (($n1 + $n2 > 0) + ? "livesrc" + : (($asm ne "") ? "deadsrc" : "nop")); + printf $output ( + "<span class=\"line\">%5d</span> " . + "<span class=\"%s\">%6s %6s %s</span>%s\n", + $l, $source_class, + HtmlPrintNumber($n1), + HtmlPrintNumber($n2), + HtmlEscape($text), + $asm); + } else { + printf $output( + "%6s %6s %4d: %s\n", + UnparseAlt($n1), + UnparseAlt($n2), + $l, + $text); + } + if ($l == $lastline) { print $output $skip_marker; } + }; + } + close(FILE); + if ($html) { + print $output "</pre>\n"; + } + return 1; +} + +# Return the source line for the specified file/linenumber. +# Returns undef if not found. +sub SourceLine { + my $file = shift; + my $line = shift; + + # Look in cache + if (!defined($main::source_cache{$file})) { + if (100 < scalar keys(%main::source_cache)) { + # Clear the cache when it gets too big + $main::source_cache = (); + } + + # Read all lines from the file + if (!open(FILE, "<$file")) { + print STDERR "$file: $!\n"; + $main::source_cache{$file} = []; # Cache the negative result + return undef; + } + my $lines = []; + push(@{$lines}, ""); # So we can use 1-based line numbers as indices + while (<FILE>) { + push(@{$lines}, $_); + } + close(FILE); + + # Save the lines in the cache + $main::source_cache{$file} = $lines; + } + + my $lines = $main::source_cache{$file}; + if (($line < 0) || ($line > $#{$lines})) { + return undef; + } else { + return $lines->[$line]; + } +} + +# Print disassembly for one routine with interspersed source if available +sub PrintDisassembledFunction { + my $prog = shift; + my $offset = shift; + my $routine = shift; + my $flat = shift; + my $cumulative = shift; + my $start_addr = shift; + my $end_addr = shift; + my $total = shift; + + # Disassemble all instructions + my @instructions = Disassemble($prog, $offset, $start_addr, $end_addr); + + # Make array of counts per instruction + my @flat_count = (); + my @cum_count = (); + my $flat_total = 0; + my $cum_total = 0; + foreach my $e (@instructions) { + # Add up counts for all address that fall inside this instruction + my $c1 = 0; + my $c2 = 0; + for (my $a = $e->[0]; $a lt $e->[4]; $a = AddressInc($a)) { + $c1 += GetEntry($flat, $a); + $c2 += GetEntry($cumulative, $a); + } + push(@flat_count, $c1); + push(@cum_count, $c2); + $flat_total += $c1; + $cum_total += $c2; + } + + # Print header with total counts + printf("ROUTINE ====================== %s\n" . + "%6s %6s %s (flat, cumulative) %.1f%% of total\n", + ShortFunctionName($routine), + Unparse($flat_total), + Unparse($cum_total), + Units(), + ($cum_total * 100.0) / $total); + + # Process instructions in order + my $current_file = ""; + for (my $i = 0; $i <= $#instructions; ) { + my $e = $instructions[$i]; + + # Print the new file name whenever we switch files + if ($e->[1] ne $current_file) { + $current_file = $e->[1]; + my $fname = $current_file; + $fname =~ s|^\./||; # Trim leading "./" + + # Shorten long file names + if (length($fname) >= 58) { + $fname = "..." . substr($fname, -55); + } + printf("-------------------- %s\n", $fname); + } + + # TODO: Compute range of lines to print together to deal with + # small reorderings. + my $first_line = $e->[2]; + my $last_line = $first_line; + my %flat_sum = (); + my %cum_sum = (); + for (my $l = $first_line; $l <= $last_line; $l++) { + $flat_sum{$l} = 0; + $cum_sum{$l} = 0; + } + + # Find run of instructions for this range of source lines + my $first_inst = $i; + while (($i <= $#instructions) && + ($instructions[$i]->[2] >= $first_line) && + ($instructions[$i]->[2] <= $last_line)) { + $e = $instructions[$i]; + $flat_sum{$e->[2]} += $flat_count[$i]; + $cum_sum{$e->[2]} += $cum_count[$i]; + $i++; + } + my $last_inst = $i - 1; + + # Print source lines + for (my $l = $first_line; $l <= $last_line; $l++) { + my $line = SourceLine($current_file, $l); + if (!defined($line)) { + $line = "?\n"; + next; + } else { + $line =~ s/^\s+//; + } + printf("%6s %6s %5d: %s", + UnparseAlt($flat_sum{$l}), + UnparseAlt($cum_sum{$l}), + $l, + $line); + } + + # Print disassembly + for (my $x = $first_inst; $x <= $last_inst; $x++) { + my $e = $instructions[$x]; + printf("%6s %6s %8s: %6s\n", + UnparseAlt($flat_count[$x]), + UnparseAlt($cum_count[$x]), + UnparseAddress($offset, $e->[0]), + CleanDisassembly($e->[3])); + } + } +} + +# Print DOT graph +sub PrintDot { + my $prog = shift; + my $symbols = shift; + my $raw = shift; + my $flat = shift; + my $cumulative = shift; + my $overall_total = shift; + + # Get total + my $local_total = TotalProfile($flat); + my $nodelimit = int($main::opt_nodefraction * $local_total); + my $edgelimit = int($main::opt_edgefraction * $local_total); + my $nodecount = $main::opt_nodecount; + + # Find nodes to include + my @list = (sort { abs(GetEntry($cumulative, $b)) <=> + abs(GetEntry($cumulative, $a)) + || $a cmp $b } + keys(%{$cumulative})); + my $last = $nodecount - 1; + if ($last > $#list) { + $last = $#list; + } + while (($last >= 0) && + (abs(GetEntry($cumulative, $list[$last])) <= $nodelimit)) { + $last--; + } + if ($last < 0) { + print STDERR "No nodes to print\n"; + return 0; + } + + if ($nodelimit > 0 || $edgelimit > 0) { + printf STDERR ("Dropping nodes with <= %s %s; edges with <= %s abs(%s)\n", + Unparse($nodelimit), Units(), + Unparse($edgelimit), Units()); + } + + # Open DOT output file + my $output; + my $escaped_dot = ShellEscape(@DOT); + my $escaped_ps2pdf = ShellEscape(@PS2PDF); + if ($main::opt_gv) { + my $escaped_outfile = ShellEscape(TempName($main::next_tmpfile, "ps")); + $output = "| $escaped_dot -Tps2 >$escaped_outfile"; + } elsif ($main::opt_evince) { + my $escaped_outfile = ShellEscape(TempName($main::next_tmpfile, "pdf")); + $output = "| $escaped_dot -Tps2 | $escaped_ps2pdf - $escaped_outfile"; + } elsif ($main::opt_ps) { + $output = "| $escaped_dot -Tps2"; + } elsif ($main::opt_pdf) { + $output = "| $escaped_dot -Tps2 | $escaped_ps2pdf - -"; + } elsif ($main::opt_web || $main::opt_svg) { + # We need to post-process the SVG, so write to a temporary file always. + my $escaped_outfile = ShellEscape(TempName($main::next_tmpfile, "svg")); + $output = "| $escaped_dot -Tsvg >$escaped_outfile"; + } elsif ($main::opt_gif) { + $output = "| $escaped_dot -Tgif"; + } else { + $output = ">&STDOUT"; + } + open(DOT, $output) || error("$output: $!\n"); + + # Title + printf DOT ("digraph \"%s; %s %s\" {\n", + $prog, + Unparse($overall_total), + Units()); + if ($main::opt_pdf) { + # The output is more printable if we set the page size for dot. + printf DOT ("size=\"8,11\"\n"); + } + printf DOT ("node [width=0.375,height=0.25];\n"); + + # Print legend + printf DOT ("Legend [shape=box,fontsize=24,shape=plaintext," . + "label=\"%s\\l%s\\l%s\\l%s\\l%s\\l\"];\n", + $prog, + sprintf("Total %s: %s", Units(), Unparse($overall_total)), + sprintf("Focusing on: %s", Unparse($local_total)), + sprintf("Dropped nodes with <= %s abs(%s)", + Unparse($nodelimit), Units()), + sprintf("Dropped edges with <= %s %s", + Unparse($edgelimit), Units()) + ); + + # Print nodes + my %node = (); + my $nextnode = 1; + foreach my $a (@list[0..$last]) { + # Pick font size + my $f = GetEntry($flat, $a); + my $c = GetEntry($cumulative, $a); + + my $fs = 8; + if ($local_total > 0) { + $fs = 8 + (50.0 * sqrt(abs($f * 1.0 / $local_total))); + } + + $node{$a} = $nextnode++; + my $sym = $a; + $sym =~ s/\s+/\\n/g; + $sym =~ s/::/\\n/g; + + # Extra cumulative info to print for non-leaves + my $extra = ""; + if ($f != $c) { + $extra = sprintf("\\rof %s (%s)", + Unparse($c), + Percent($c, $local_total)); + } + my $style = ""; + if ($main::opt_heapcheck) { + if ($f > 0) { + # make leak-causing nodes more visible (add a background) + $style = ",style=filled,fillcolor=gray" + } elsif ($f < 0) { + # make anti-leak-causing nodes (which almost never occur) + # stand out as well (triple border) + $style = ",peripheries=3" + } + } + + printf DOT ("N%d [label=\"%s\\n%s (%s)%s\\r" . + "\",shape=box,fontsize=%.1f%s];\n", + $node{$a}, + $sym, + Unparse($f), + Percent($f, $local_total), + $extra, + $fs, + $style, + ); + } + + # Get edges and counts per edge + my %edge = (); + my $n; + my $fullname_to_shortname_map = {}; + FillFullnameToShortnameMap($symbols, $fullname_to_shortname_map); + foreach my $k (keys(%{$raw})) { + # TODO: omit low %age edges + $n = $raw->{$k}; + my @translated = TranslateStack($symbols, $fullname_to_shortname_map, $k); + for (my $i = 1; $i <= $#translated; $i++) { + my $src = $translated[$i]; + my $dst = $translated[$i-1]; + #next if ($src eq $dst); # Avoid self-edges? + if (exists($node{$src}) && exists($node{$dst})) { + my $edge_label = "$src\001$dst"; + if (!exists($edge{$edge_label})) { + $edge{$edge_label} = 0; + } + $edge{$edge_label} += $n; + } + } + } + + # Print edges (process in order of decreasing counts) + my %indegree = (); # Number of incoming edges added per node so far + my %outdegree = (); # Number of outgoing edges added per node so far + foreach my $e (sort { $edge{$b} <=> $edge{$a} } keys(%edge)) { + my @x = split(/\001/, $e); + $n = $edge{$e}; + + # Initialize degree of kept incoming and outgoing edges if necessary + my $src = $x[0]; + my $dst = $x[1]; + if (!exists($outdegree{$src})) { $outdegree{$src} = 0; } + if (!exists($indegree{$dst})) { $indegree{$dst} = 0; } + + my $keep; + if ($indegree{$dst} == 0) { + # Keep edge if needed for reachability + $keep = 1; + } elsif (abs($n) <= $edgelimit) { + # Drop if we are below --edgefraction + $keep = 0; + } elsif ($outdegree{$src} >= $main::opt_maxdegree || + $indegree{$dst} >= $main::opt_maxdegree) { + # Keep limited number of in/out edges per node + $keep = 0; + } else { + $keep = 1; + } + + if ($keep) { + $outdegree{$src}++; + $indegree{$dst}++; + + # Compute line width based on edge count + my $fraction = abs($local_total ? (3 * ($n / $local_total)) : 0); + if ($fraction > 1) { $fraction = 1; } + my $w = $fraction * 2; + if ($w < 1 && ($main::opt_web || $main::opt_svg)) { + # SVG output treats line widths < 1 poorly. + $w = 1; + } + + # Dot sometimes segfaults if given edge weights that are too large, so + # we cap the weights at a large value + my $edgeweight = abs($n) ** 0.7; + if ($edgeweight > 100000) { $edgeweight = 100000; } + $edgeweight = int($edgeweight); + + my $style = sprintf("setlinewidth(%f)", $w); + if ($x[1] =~ m/\(inline\)/) { + $style .= ",dashed"; + } + + # Use a slightly squashed function of the edge count as the weight + printf DOT ("N%s -> N%s [label=%s, weight=%d, style=\"%s\"];\n", + $node{$x[0]}, + $node{$x[1]}, + Unparse($n), + $edgeweight, + $style); + } + } + + print DOT ("}\n"); + close(DOT); + + if ($main::opt_web || $main::opt_svg) { + # Rewrite SVG to be more usable inside web browser. + RewriteSvg(TempName($main::next_tmpfile, "svg")); + } + + return 1; +} + +sub RewriteSvg { + my $svgfile = shift; + + open(SVG, $svgfile) || die "open temp svg: $!"; + my @svg = <SVG>; + close(SVG); + unlink $svgfile; + my $svg = join('', @svg); + + # Dot's SVG output is + # + # <svg width="___" height="___" + # viewBox="___" xmlns=...> + # <g id="graph0" transform="..."> + # ... + # </g> + # </svg> + # + # Change it to + # + # <svg width="100%" height="100%" + # xmlns=...> + # $svg_javascript + # <g id="viewport" transform="translate(0,0)"> + # <g id="graph0" transform="..."> + # ... + # </g> + # </g> + # </svg> + + # Fix width, height; drop viewBox. + $svg =~ s/(?s)<svg width="[^"]+" height="[^"]+"(.*?)viewBox="[^"]+"/<svg width="100%" height="100%"$1/; + + # Insert script, viewport <g> above first <g> + my $svg_javascript = SvgJavascript(); + my $viewport = "<g id=\"viewport\" transform=\"translate(0,0)\">\n"; + $svg =~ s/<g id="graph\d"/$svg_javascript$viewport$&/; + + # Insert final </g> above </svg>. + $svg =~ s/(.*)(<\/svg>)/$1<\/g>$2/; + $svg =~ s/<g id="graph\d"(.*?)/<g id="viewport"$1/; + + if ($main::opt_svg) { + # --svg: write to standard output. + print $svg; + } else { + # Write back to temporary file. + open(SVG, ">$svgfile") || die "open $svgfile: $!"; + print SVG $svg; + close(SVG); + } +} + +sub SvgJavascript { + return <<'EOF'; +<script type="text/ecmascript"><![CDATA[ +// SVGPan +// http://www.cyberz.org/blog/2009/12/08/svgpan-a-javascript-svg-panzoomdrag-library/ +// Local modification: if(true || ...) below to force panning, never moving. + +/** + * SVGPan library 1.2 + * ==================== + * + * Given an unique existing element with id "viewport", including the + * the library into any SVG adds the following capabilities: + * + * - Mouse panning + * - Mouse zooming (using the wheel) + * - Object dargging + * + * Known issues: + * + * - Zooming (while panning) on Safari has still some issues + * + * Releases: + * + * 1.2, Sat Mar 20 08:42:50 GMT 2010, Zeng Xiaohui + * Fixed a bug with browser mouse handler interaction + * + * 1.1, Wed Feb 3 17:39:33 GMT 2010, Zeng Xiaohui + * Updated the zoom code to support the mouse wheel on Safari/Chrome + * + * 1.0, Andrea Leofreddi + * First release + * + * This code is licensed under the following BSD license: + * + * Copyright 2009-2010 Andrea Leofreddi <a.leofreddi@itcharm.com>. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without modification, are + * permitted provided that the following conditions are met: + * + * 1. Redistributions of source code must retain the above copyright notice, this list of + * conditions and the following disclaimer. + * + * 2. Redistributions in binary form must reproduce the above copyright notice, this list + * of conditions and the following disclaimer in the documentation and/or other materials + * provided with the distribution. + * + * THIS SOFTWARE IS PROVIDED BY Andrea Leofreddi ``AS IS'' AND ANY EXPRESS OR IMPLIED + * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND + * FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL Andrea Leofreddi OR + * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR + * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON + * ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING + * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF + * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * The views and conclusions contained in the software and documentation are those of the + * authors and should not be interpreted as representing official policies, either expressed + * or implied, of Andrea Leofreddi. + */ + +var root = document.documentElement; + +var state = 'none', stateTarget, stateOrigin, stateTf; + +setupHandlers(root); + +/** + * Register handlers + */ +function setupHandlers(root){ + setAttributes(root, { + "onmouseup" : "add(evt)", + "onmousedown" : "handleMouseDown(evt)", + "onmousemove" : "handleMouseMove(evt)", + "onmouseup" : "handleMouseUp(evt)", + //"onmouseout" : "handleMouseUp(evt)", // Decomment this to stop the pan functionality when dragging out of the SVG element + }); + + if(navigator.userAgent.toLowerCase().indexOf('webkit') >= 0) + window.addEventListener('mousewheel', handleMouseWheel, false); // Chrome/Safari + else + window.addEventListener('DOMMouseScroll', handleMouseWheel, false); // Others + + var g = svgDoc.getElementById("svg"); + g.width = "100%"; + g.height = "100%"; +} + +/** + * Instance an SVGPoint object with given event coordinates. + */ +function getEventPoint(evt) { + var p = root.createSVGPoint(); + + p.x = evt.clientX; + p.y = evt.clientY; + + return p; +} + +/** + * Sets the current transform matrix of an element. + */ +function setCTM(element, matrix) { + var s = "matrix(" + matrix.a + "," + matrix.b + "," + matrix.c + "," + matrix.d + "," + matrix.e + "," + matrix.f + ")"; + + element.setAttribute("transform", s); +} + +/** + * Dumps a matrix to a string (useful for debug). + */ +function dumpMatrix(matrix) { + var s = "[ " + matrix.a + ", " + matrix.c + ", " + matrix.e + "\n " + matrix.b + ", " + matrix.d + ", " + matrix.f + "\n 0, 0, 1 ]"; + + return s; +} + +/** + * Sets attributes of an element. + */ +function setAttributes(element, attributes){ + for (i in attributes) + element.setAttributeNS(null, i, attributes[i]); +} + +/** + * Handle mouse move event. + */ +function handleMouseWheel(evt) { + if(evt.preventDefault) + evt.preventDefault(); + + evt.returnValue = false; + + var svgDoc = evt.target.ownerDocument; + + var delta; + + if(evt.wheelDelta) + delta = evt.wheelDelta / 3600; // Chrome/Safari + else + delta = evt.detail / -90; // Mozilla + + var z = 1 + delta; // Zoom factor: 0.9/1.1 + + var g = svgDoc.getElementById("viewport"); + + var p = getEventPoint(evt); + + p = p.matrixTransform(g.getCTM().inverse()); + + // Compute new scale matrix in current mouse position + var k = root.createSVGMatrix().translate(p.x, p.y).scale(z).translate(-p.x, -p.y); + + setCTM(g, g.getCTM().multiply(k)); + + stateTf = stateTf.multiply(k.inverse()); +} + +/** + * Handle mouse move event. + */ +function handleMouseMove(evt) { + if(evt.preventDefault) + evt.preventDefault(); + + evt.returnValue = false; + + var svgDoc = evt.target.ownerDocument; + + var g = svgDoc.getElementById("viewport"); + + if(state == 'pan') { + // Pan mode + var p = getEventPoint(evt).matrixTransform(stateTf); + + setCTM(g, stateTf.inverse().translate(p.x - stateOrigin.x, p.y - stateOrigin.y)); + } else if(state == 'move') { + // Move mode + var p = getEventPoint(evt).matrixTransform(g.getCTM().inverse()); + + setCTM(stateTarget, root.createSVGMatrix().translate(p.x - stateOrigin.x, p.y - stateOrigin.y).multiply(g.getCTM().inverse()).multiply(stateTarget.getCTM())); + + stateOrigin = p; + } +} + +/** + * Handle click event. + */ +function handleMouseDown(evt) { + if(evt.preventDefault) + evt.preventDefault(); + + evt.returnValue = false; + + var svgDoc = evt.target.ownerDocument; + + var g = svgDoc.getElementById("viewport"); + + if(true || evt.target.tagName == "svg") { + // Pan mode + state = 'pan'; + + stateTf = g.getCTM().inverse(); + + stateOrigin = getEventPoint(evt).matrixTransform(stateTf); + } else { + // Move mode + state = 'move'; + + stateTarget = evt.target; + + stateTf = g.getCTM().inverse(); + + stateOrigin = getEventPoint(evt).matrixTransform(stateTf); + } +} + +/** + * Handle mouse button release event. + */ +function handleMouseUp(evt) { + if(evt.preventDefault) + evt.preventDefault(); + + evt.returnValue = false; + + var svgDoc = evt.target.ownerDocument; + + if(state == 'pan' || state == 'move') { + // Quit pan mode + state = ''; + } +} + +]]></script> +EOF +} + +# Provides a map from fullname to shortname for cases where the +# shortname is ambiguous. The symlist has both the fullname and +# shortname for all symbols, which is usually fine, but sometimes -- +# such as overloaded functions -- two different fullnames can map to +# the same shortname. In that case, we use the address of the +# function to disambiguate the two. This function fills in a map that +# maps fullnames to modified shortnames in such cases. If a fullname +# is not present in the map, the 'normal' shortname provided by the +# symlist is the appropriate one to use. +sub FillFullnameToShortnameMap { + my $symbols = shift; + my $fullname_to_shortname_map = shift; + my $shortnames_seen_once = {}; + my $shortnames_seen_more_than_once = {}; + + foreach my $symlist (values(%{$symbols})) { + # TODO(csilvers): deal with inlined symbols too. + my $shortname = $symlist->[0]; + my $fullname = $symlist->[2]; + if ($fullname !~ /<[0-9a-fA-F]+>$/) { # fullname doesn't end in an address + next; # the only collisions we care about are when addresses differ + } + if (defined($shortnames_seen_once->{$shortname}) && + $shortnames_seen_once->{$shortname} ne $fullname) { + $shortnames_seen_more_than_once->{$shortname} = 1; + } else { + $shortnames_seen_once->{$shortname} = $fullname; + } + } + + foreach my $symlist (values(%{$symbols})) { + my $shortname = $symlist->[0]; + my $fullname = $symlist->[2]; + # TODO(csilvers): take in a list of addresses we care about, and only + # store in the map if $symlist->[1] is in that list. Saves space. + next if defined($fullname_to_shortname_map->{$fullname}); + if (defined($shortnames_seen_more_than_once->{$shortname})) { + if ($fullname =~ /<0*([^>]*)>$/) { # fullname has address at end of it + $fullname_to_shortname_map->{$fullname} = "$shortname\@$1"; + } + } + } +} + +# Return a small number that identifies the argument. +# Multiple calls with the same argument will return the same number. +# Calls with different arguments will return different numbers. +sub ShortIdFor { + my $key = shift; + my $id = $main::uniqueid{$key}; + if (!defined($id)) { + $id = keys(%main::uniqueid) + 1; + $main::uniqueid{$key} = $id; + } + return $id; +} + +# Translate a stack of addresses into a stack of symbols +sub TranslateStack { + my $symbols = shift; + my $fullname_to_shortname_map = shift; + my $k = shift; + + my @addrs = split(/\n/, $k); + my @result = (); + for (my $i = 0; $i <= $#addrs; $i++) { + my $a = $addrs[$i]; + + # Skip large addresses since they sometimes show up as fake entries on RH9 + if (length($a) > 8 && $a gt "7fffffffffffffff") { + next; + } + + if ($main::opt_disasm || $main::opt_list) { + # We want just the address for the key + push(@result, $a); + next; + } + + my $symlist = $symbols->{$a}; + if (!defined($symlist)) { + $symlist = [$a, "", $a]; + } + + # We can have a sequence of symbols for a particular entry + # (more than one symbol in the case of inlining). Callers + # come before callees in symlist, so walk backwards since + # the translated stack should contain callees before callers. + for (my $j = $#{$symlist}; $j >= 2; $j -= 3) { + my $func = $symlist->[$j-2]; + my $fileline = $symlist->[$j-1]; + my $fullfunc = $symlist->[$j]; + if (defined($fullname_to_shortname_map->{$fullfunc})) { + $func = $fullname_to_shortname_map->{$fullfunc}; + } + if ($j > 2) { + $func = "$func (inline)"; + } + + # Do not merge nodes corresponding to Callback::Run since that + # causes confusing cycles in dot display. Instead, we synthesize + # a unique name for this frame per caller. + if ($func =~ m/Callback.*::Run$/) { + my $caller = ($i > 0) ? $addrs[$i-1] : 0; + $func = "Run#" . ShortIdFor($caller); + } + + if ($main::opt_addresses) { + push(@result, "$a $func $fileline"); + } elsif ($main::opt_lines) { + if ($func eq '??' && $fileline eq '??:0') { + push(@result, "$a"); + } else { + push(@result, "$func $fileline"); + } + } elsif ($main::opt_functions) { + if ($func eq '??') { + push(@result, "$a"); + } else { + push(@result, $func); + } + } elsif ($main::opt_files) { + if ($fileline eq '??:0' || $fileline eq '') { + push(@result, "$a"); + } else { + my $f = $fileline; + $f =~ s/:\d+$//; + push(@result, $f); + } + } else { + push(@result, $a); + last; # Do not print inlined info + } + } + } + + # print join(",", @addrs), " => ", join(",", @result), "\n"; + return @result; +} + +# Generate percent string for a number and a total +sub Percent { + my $num = shift; + my $tot = shift; + if ($tot != 0) { + return sprintf("%.1f%%", $num * 100.0 / $tot); + } else { + return ($num == 0) ? "nan" : (($num > 0) ? "+inf" : "-inf"); + } +} + +# Generate pretty-printed form of number +sub Unparse { + my $num = shift; + if ($main::profile_type eq 'heap' || $main::profile_type eq 'growth') { + if ($main::opt_inuse_objects || $main::opt_alloc_objects) { + return sprintf("%d", $num); + } else { + if ($main::opt_show_bytes) { + return sprintf("%d", $num); + } else { + return sprintf("%.1f", $num / 1048576.0); + } + } + } elsif ($main::profile_type eq 'contention' && !$main::opt_contentions) { + return sprintf("%.3f", $num / 1e9); # Convert nanoseconds to seconds + } else { + return sprintf("%d", $num); + } +} + +# Alternate pretty-printed form: 0 maps to "." +sub UnparseAlt { + my $num = shift; + if ($num == 0) { + return "."; + } else { + return Unparse($num); + } +} + +# Alternate pretty-printed form: 0 maps to "" +sub HtmlPrintNumber { + my $num = shift; + if ($num == 0) { + return ""; + } else { + return Unparse($num); + } +} + +# Return output units +sub Units { + if ($main::profile_type eq 'heap' || $main::profile_type eq 'growth') { + if ($main::opt_inuse_objects || $main::opt_alloc_objects) { + return "objects"; + } else { + if ($main::opt_show_bytes) { + return "B"; + } else { + return "MB"; + } + } + } elsif ($main::profile_type eq 'contention' && !$main::opt_contentions) { + return "seconds"; + } else { + return "samples"; + } +} + +##### Profile manipulation code ##### + +# Generate flattened profile: +# If count is charged to stack [a,b,c,d], in generated profile, +# it will be charged to [a] +sub FlatProfile { + my $profile = shift; + my $result = {}; + foreach my $k (keys(%{$profile})) { + my $count = $profile->{$k}; + my @addrs = split(/\n/, $k); + if ($#addrs >= 0) { + AddEntry($result, $addrs[0], $count); + } + } + return $result; +} + +# Generate cumulative profile: +# If count is charged to stack [a,b,c,d], in generated profile, +# it will be charged to [a], [b], [c], [d] +sub CumulativeProfile { + my $profile = shift; + my $result = {}; + foreach my $k (keys(%{$profile})) { + my $count = $profile->{$k}; + my @addrs = split(/\n/, $k); + foreach my $a (@addrs) { + AddEntry($result, $a, $count); + } + } + return $result; +} + +# If the second-youngest PC on the stack is always the same, returns +# that pc. Otherwise, returns undef. +sub IsSecondPcAlwaysTheSame { + my $profile = shift; + + my $second_pc = undef; + foreach my $k (keys(%{$profile})) { + my @addrs = split(/\n/, $k); + if ($#addrs < 1) { + return undef; + } + if (not defined $second_pc) { + $second_pc = $addrs[1]; + } else { + if ($second_pc ne $addrs[1]) { + return undef; + } + } + } + return $second_pc; +} + +sub ExtractSymbolLocation { + my $symbols = shift; + my $address = shift; + # 'addr2line' outputs "??:0" for unknown locations; we do the + # same to be consistent. + my $location = "??:0:unknown"; + if (exists $symbols->{$address}) { + my $file = $symbols->{$address}->[1]; + if ($file eq "?") { + $file = "??:0" + } + $location = $file . ":" . $symbols->{$address}->[0]; + } + return $location; +} + +# Extracts a graph of calls. +sub ExtractCalls { + my $symbols = shift; + my $profile = shift; + + my $calls = {}; + while( my ($stack_trace, $count) = each %$profile ) { + my @address = split(/\n/, $stack_trace); + my $destination = ExtractSymbolLocation($symbols, $address[0]); + AddEntry($calls, $destination, $count); + for (my $i = 1; $i <= $#address; $i++) { + my $source = ExtractSymbolLocation($symbols, $address[$i]); + my $call = "$source -> $destination"; + AddEntry($calls, $call, $count); + $destination = $source; + } + } + + return $calls; +} + +sub FilterFrames { + my $symbols = shift; + my $profile = shift; + + if ($main::opt_retain eq '' && $main::opt_exclude eq '') { + return $profile; + } + + my $result = {}; + foreach my $k (keys(%{$profile})) { + my $count = $profile->{$k}; + my @addrs = split(/\n/, $k); + my @path = (); + foreach my $a (@addrs) { + my $sym; + if (exists($symbols->{$a})) { + $sym = $symbols->{$a}->[0]; + } else { + $sym = $a; + } + if ($main::opt_retain ne '' && $sym !~ m/$main::opt_retain/) { + next; + } + if ($main::opt_exclude ne '' && $sym =~ m/$main::opt_exclude/) { + next; + } + push(@path, $a); + } + if (scalar(@path) > 0) { + my $reduced_path = join("\n", @path); + AddEntry($result, $reduced_path, $count); + } + } + + return $result; +} + +sub RemoveUninterestingFrames { + my $symbols = shift; + my $profile = shift; + + # List of function names to skip + my %skip = (); + my $skip_regexp = 'NOMATCH'; + if ($main::profile_type eq 'heap' || $main::profile_type eq 'growth') { + foreach my $name ('@JEMALLOC_PREFIX@calloc', + 'cfree', + '@JEMALLOC_PREFIX@malloc', + 'newImpl', + 'void* newImpl', + '@JEMALLOC_PREFIX@free', + '@JEMALLOC_PREFIX@memalign', + '@JEMALLOC_PREFIX@posix_memalign', + '@JEMALLOC_PREFIX@aligned_alloc', + 'pvalloc', + '@JEMALLOC_PREFIX@valloc', + '@JEMALLOC_PREFIX@realloc', + '@JEMALLOC_PREFIX@mallocx', + '@JEMALLOC_PREFIX@rallocx', + '@JEMALLOC_PREFIX@xallocx', + '@JEMALLOC_PREFIX@dallocx', + '@JEMALLOC_PREFIX@sdallocx', + '@JEMALLOC_PREFIX@sdallocx_noflags', + 'tc_calloc', + 'tc_cfree', + 'tc_malloc', + 'tc_free', + 'tc_memalign', + 'tc_posix_memalign', + 'tc_pvalloc', + 'tc_valloc', + 'tc_realloc', + 'tc_new', + 'tc_delete', + 'tc_newarray', + 'tc_deletearray', + 'tc_new_nothrow', + 'tc_newarray_nothrow', + 'do_malloc', + '::do_malloc', # new name -- got moved to an unnamed ns + '::do_malloc_or_cpp_alloc', + 'DoSampledAllocation', + 'simple_alloc::allocate', + '__malloc_alloc_template::allocate', + '__builtin_delete', + '__builtin_new', + '__builtin_vec_delete', + '__builtin_vec_new', + 'operator new', + 'operator new[]', + # The entry to our memory-allocation routines on OS X + 'malloc_zone_malloc', + 'malloc_zone_calloc', + 'malloc_zone_valloc', + 'malloc_zone_realloc', + 'malloc_zone_memalign', + 'malloc_zone_free', + # These mark the beginning/end of our custom sections + '__start_google_malloc', + '__stop_google_malloc', + '__start_malloc_hook', + '__stop_malloc_hook') { + $skip{$name} = 1; + $skip{"_" . $name} = 1; # Mach (OS X) adds a _ prefix to everything + } + # TODO: Remove TCMalloc once everything has been + # moved into the tcmalloc:: namespace and we have flushed + # old code out of the system. + $skip_regexp = "TCMalloc|^tcmalloc::"; + } elsif ($main::profile_type eq 'contention') { + foreach my $vname ('base::RecordLockProfileData', + 'base::SubmitMutexProfileData', + 'base::SubmitSpinLockProfileData', + 'Mutex::Unlock', + 'Mutex::UnlockSlow', + 'Mutex::ReaderUnlock', + 'MutexLock::~MutexLock', + 'SpinLock::Unlock', + 'SpinLock::SlowUnlock', + 'SpinLockHolder::~SpinLockHolder') { + $skip{$vname} = 1; + } + } elsif ($main::profile_type eq 'cpu') { + # Drop signal handlers used for CPU profile collection + # TODO(dpeng): this should not be necessary; it's taken + # care of by the general 2nd-pc mechanism below. + foreach my $name ('ProfileData::Add', # historical + 'ProfileData::prof_handler', # historical + 'CpuProfiler::prof_handler', + '__FRAME_END__', + '__pthread_sighandler', + '__restore') { + $skip{$name} = 1; + } + } else { + # Nothing skipped for unknown types + } + + if ($main::profile_type eq 'cpu') { + # If all the second-youngest program counters are the same, + # this STRONGLY suggests that it is an artifact of measurement, + # i.e., stack frames pushed by the CPU profiler signal handler. + # Hence, we delete them. + # (The topmost PC is read from the signal structure, not from + # the stack, so it does not get involved.) + while (my $second_pc = IsSecondPcAlwaysTheSame($profile)) { + my $result = {}; + my $func = ''; + if (exists($symbols->{$second_pc})) { + $second_pc = $symbols->{$second_pc}->[0]; + } + print STDERR "Removing $second_pc from all stack traces.\n"; + foreach my $k (keys(%{$profile})) { + my $count = $profile->{$k}; + my @addrs = split(/\n/, $k); + splice @addrs, 1, 1; + my $reduced_path = join("\n", @addrs); + AddEntry($result, $reduced_path, $count); + } + $profile = $result; + } + } + + my $result = {}; + foreach my $k (keys(%{$profile})) { + my $count = $profile->{$k}; + my @addrs = split(/\n/, $k); + my @path = (); + foreach my $a (@addrs) { + if (exists($symbols->{$a})) { + my $func = $symbols->{$a}->[0]; + if ($skip{$func} || ($func =~ m/$skip_regexp/)) { + # Throw away the portion of the backtrace seen so far, under the + # assumption that previous frames were for functions internal to the + # allocator. + @path = (); + next; + } + } + push(@path, $a); + } + my $reduced_path = join("\n", @path); + AddEntry($result, $reduced_path, $count); + } + + $result = FilterFrames($symbols, $result); + + return $result; +} + +# Reduce profile to granularity given by user +sub ReduceProfile { + my $symbols = shift; + my $profile = shift; + my $result = {}; + my $fullname_to_shortname_map = {}; + FillFullnameToShortnameMap($symbols, $fullname_to_shortname_map); + foreach my $k (keys(%{$profile})) { + my $count = $profile->{$k}; + my @translated = TranslateStack($symbols, $fullname_to_shortname_map, $k); + my @path = (); + my %seen = (); + $seen{''} = 1; # So that empty keys are skipped + foreach my $e (@translated) { + # To avoid double-counting due to recursion, skip a stack-trace + # entry if it has already been seen + if (!$seen{$e}) { + $seen{$e} = 1; + push(@path, $e); + } + } + my $reduced_path = join("\n", @path); + AddEntry($result, $reduced_path, $count); + } + return $result; +} + +# Does the specified symbol array match the regexp? +sub SymbolMatches { + my $sym = shift; + my $re = shift; + if (defined($sym)) { + for (my $i = 0; $i < $#{$sym}; $i += 3) { + if ($sym->[$i] =~ m/$re/ || $sym->[$i+1] =~ m/$re/) { + return 1; + } + } + } + return 0; +} + +# Focus only on paths involving specified regexps +sub FocusProfile { + my $symbols = shift; + my $profile = shift; + my $focus = shift; + my $result = {}; + foreach my $k (keys(%{$profile})) { + my $count = $profile->{$k}; + my @addrs = split(/\n/, $k); + foreach my $a (@addrs) { + # Reply if it matches either the address/shortname/fileline + if (($a =~ m/$focus/) || SymbolMatches($symbols->{$a}, $focus)) { + AddEntry($result, $k, $count); + last; + } + } + } + return $result; +} + +# Focus only on paths not involving specified regexps +sub IgnoreProfile { + my $symbols = shift; + my $profile = shift; + my $ignore = shift; + my $result = {}; + foreach my $k (keys(%{$profile})) { + my $count = $profile->{$k}; + my @addrs = split(/\n/, $k); + my $matched = 0; + foreach my $a (@addrs) { + # Reply if it matches either the address/shortname/fileline + if (($a =~ m/$ignore/) || SymbolMatches($symbols->{$a}, $ignore)) { + $matched = 1; + last; + } + } + if (!$matched) { + AddEntry($result, $k, $count); + } + } + return $result; +} + +# Get total count in profile +sub TotalProfile { + my $profile = shift; + my $result = 0; + foreach my $k (keys(%{$profile})) { + $result += $profile->{$k}; + } + return $result; +} + +# Add A to B +sub AddProfile { + my $A = shift; + my $B = shift; + + my $R = {}; + # add all keys in A + foreach my $k (keys(%{$A})) { + my $v = $A->{$k}; + AddEntry($R, $k, $v); + } + # add all keys in B + foreach my $k (keys(%{$B})) { + my $v = $B->{$k}; + AddEntry($R, $k, $v); + } + return $R; +} + +# Merges symbol maps +sub MergeSymbols { + my $A = shift; + my $B = shift; + + my $R = {}; + foreach my $k (keys(%{$A})) { + $R->{$k} = $A->{$k}; + } + if (defined($B)) { + foreach my $k (keys(%{$B})) { + $R->{$k} = $B->{$k}; + } + } + return $R; +} + + +# Add A to B +sub AddPcs { + my $A = shift; + my $B = shift; + + my $R = {}; + # add all keys in A + foreach my $k (keys(%{$A})) { + $R->{$k} = 1 + } + # add all keys in B + foreach my $k (keys(%{$B})) { + $R->{$k} = 1 + } + return $R; +} + +# Subtract B from A +sub SubtractProfile { + my $A = shift; + my $B = shift; + + my $R = {}; + foreach my $k (keys(%{$A})) { + my $v = $A->{$k} - GetEntry($B, $k); + if ($v < 0 && $main::opt_drop_negative) { + $v = 0; + } + AddEntry($R, $k, $v); + } + if (!$main::opt_drop_negative) { + # Take care of when subtracted profile has more entries + foreach my $k (keys(%{$B})) { + if (!exists($A->{$k})) { + AddEntry($R, $k, 0 - $B->{$k}); + } + } + } + return $R; +} + +# Get entry from profile; zero if not present +sub GetEntry { + my $profile = shift; + my $k = shift; + if (exists($profile->{$k})) { + return $profile->{$k}; + } else { + return 0; + } +} + +# Add entry to specified profile +sub AddEntry { + my $profile = shift; + my $k = shift; + my $n = shift; + if (!exists($profile->{$k})) { + $profile->{$k} = 0; + } + $profile->{$k} += $n; +} + +# Add a stack of entries to specified profile, and add them to the $pcs +# list. +sub AddEntries { + my $profile = shift; + my $pcs = shift; + my $stack = shift; + my $count = shift; + my @k = (); + + foreach my $e (split(/\s+/, $stack)) { + my $pc = HexExtend($e); + $pcs->{$pc} = 1; + push @k, $pc; + } + AddEntry($profile, (join "\n", @k), $count); +} + +##### Code to profile a server dynamically ##### + +sub CheckSymbolPage { + my $url = SymbolPageURL(); + my $command = ShellEscape(@URL_FETCHER, $url); + open(SYMBOL, "$command |") or error($command); + my $line = <SYMBOL>; + $line =~ s/\r//g; # turn windows-looking lines into unix-looking lines + close(SYMBOL); + unless (defined($line)) { + error("$url doesn't exist\n"); + } + + if ($line =~ /^num_symbols:\s+(\d+)$/) { + if ($1 == 0) { + error("Stripped binary. No symbols available.\n"); + } + } else { + error("Failed to get the number of symbols from $url\n"); + } +} + +sub IsProfileURL { + my $profile_name = shift; + if (-f $profile_name) { + printf STDERR "Using local file $profile_name.\n"; + return 0; + } + return 1; +} + +sub ParseProfileURL { + my $profile_name = shift; + + if (!defined($profile_name) || $profile_name eq "") { + return (); + } + + # Split profile URL - matches all non-empty strings, so no test. + $profile_name =~ m,^(https?://)?([^/]+)(.*?)(/|$PROFILES)?$,; + + my $proto = $1 || "http://"; + my $hostport = $2; + my $prefix = $3; + my $profile = $4 || "/"; + + my $host = $hostport; + $host =~ s/:.*//; + + my $baseurl = "$proto$hostport$prefix"; + return ($host, $baseurl, $profile); +} + +# We fetch symbols from the first profile argument. +sub SymbolPageURL { + my ($host, $baseURL, $path) = ParseProfileURL($main::pfile_args[0]); + return "$baseURL$SYMBOL_PAGE"; +} + +sub FetchProgramName() { + my ($host, $baseURL, $path) = ParseProfileURL($main::pfile_args[0]); + my $url = "$baseURL$PROGRAM_NAME_PAGE"; + my $command_line = ShellEscape(@URL_FETCHER, $url); + open(CMDLINE, "$command_line |") or error($command_line); + my $cmdline = <CMDLINE>; + $cmdline =~ s/\r//g; # turn windows-looking lines into unix-looking lines + close(CMDLINE); + error("Failed to get program name from $url\n") unless defined($cmdline); + $cmdline =~ s/\x00.+//; # Remove argv[1] and latters. + $cmdline =~ s!\n!!g; # Remove LFs. + return $cmdline; +} + +# Gee, curl's -L (--location) option isn't reliable at least +# with its 7.12.3 version. Curl will forget to post data if +# there is a redirection. This function is a workaround for +# curl. Redirection happens on borg hosts. +sub ResolveRedirectionForCurl { + my $url = shift; + my $command_line = ShellEscape(@URL_FETCHER, "--head", $url); + open(CMDLINE, "$command_line |") or error($command_line); + while (<CMDLINE>) { + s/\r//g; # turn windows-looking lines into unix-looking lines + if (/^Location: (.*)/) { + $url = $1; + } + } + close(CMDLINE); + return $url; +} + +# Add a timeout flat to URL_FETCHER. Returns a new list. +sub AddFetchTimeout { + my $timeout = shift; + my @fetcher = @_; + if (defined($timeout)) { + if (join(" ", @fetcher) =~ m/\bcurl -s/) { + push(@fetcher, "--max-time", sprintf("%d", $timeout)); + } elsif (join(" ", @fetcher) =~ m/\brpcget\b/) { + push(@fetcher, sprintf("--deadline=%d", $timeout)); + } + } + return @fetcher; +} + +# Reads a symbol map from the file handle name given as $1, returning +# the resulting symbol map. Also processes variables relating to symbols. +# Currently, the only variable processed is 'binary=<value>' which updates +# $main::prog to have the correct program name. +sub ReadSymbols { + my $in = shift; + my $map = {}; + while (<$in>) { + s/\r//g; # turn windows-looking lines into unix-looking lines + # Removes all the leading zeroes from the symbols, see comment below. + if (m/^0x0*([0-9a-f]+)\s+(.+)/) { + $map->{$1} = $2; + } elsif (m/^---/) { + last; + } elsif (m/^([a-z][^=]*)=(.*)$/ ) { + my ($variable, $value) = ($1, $2); + for ($variable, $value) { + s/^\s+//; + s/\s+$//; + } + if ($variable eq "binary") { + if ($main::prog ne $UNKNOWN_BINARY && $main::prog ne $value) { + printf STDERR ("Warning: Mismatched binary name '%s', using '%s'.\n", + $main::prog, $value); + } + $main::prog = $value; + } else { + printf STDERR ("Ignoring unknown variable in symbols list: " . + "'%s' = '%s'\n", $variable, $value); + } + } + } + return $map; +} + +sub URLEncode { + my $str = shift; + $str =~ s/([^A-Za-z0-9\-_.!~*'()])/ sprintf "%%%02x", ord $1 /eg; + return $str; +} + +sub AppendSymbolFilterParams { + my $url = shift; + my @params = (); + if ($main::opt_retain ne '') { + push(@params, sprintf("retain=%s", URLEncode($main::opt_retain))); + } + if ($main::opt_exclude ne '') { + push(@params, sprintf("exclude=%s", URLEncode($main::opt_exclude))); + } + if (scalar @params > 0) { + $url = sprintf("%s?%s", $url, join("&", @params)); + } + return $url; +} + +# Fetches and processes symbols to prepare them for use in the profile output +# code. If the optional 'symbol_map' arg is not given, fetches symbols from +# $SYMBOL_PAGE for all PC values found in profile. Otherwise, the raw symbols +# are assumed to have already been fetched into 'symbol_map' and are simply +# extracted and processed. +sub FetchSymbols { + my $pcset = shift; + my $symbol_map = shift; + + my %seen = (); + my @pcs = grep { !$seen{$_}++ } keys(%$pcset); # uniq + + if (!defined($symbol_map)) { + my $post_data = join("+", sort((map {"0x" . "$_"} @pcs))); + + open(POSTFILE, ">$main::tmpfile_sym"); + print POSTFILE $post_data; + close(POSTFILE); + + my $url = SymbolPageURL(); + + my $command_line; + if (join(" ", @URL_FETCHER) =~ m/\bcurl -s/) { + $url = ResolveRedirectionForCurl($url); + $url = AppendSymbolFilterParams($url); + $command_line = ShellEscape(@URL_FETCHER, "-d", "\@$main::tmpfile_sym", + $url); + } else { + $url = AppendSymbolFilterParams($url); + $command_line = (ShellEscape(@URL_FETCHER, "--post", $url) + . " < " . ShellEscape($main::tmpfile_sym)); + } + # We use c++filt in case $SYMBOL_PAGE gives us mangled symbols. + my $escaped_cppfilt = ShellEscape($obj_tool_map{"c++filt"}); + open(SYMBOL, "$command_line | $escaped_cppfilt |") or error($command_line); + $symbol_map = ReadSymbols(*SYMBOL{IO}); + close(SYMBOL); + } + + my $symbols = {}; + foreach my $pc (@pcs) { + my $fullname; + # For 64 bits binaries, symbols are extracted with 8 leading zeroes. + # Then /symbol reads the long symbols in as uint64, and outputs + # the result with a "0x%08llx" format which get rid of the zeroes. + # By removing all the leading zeroes in both $pc and the symbols from + # /symbol, the symbols match and are retrievable from the map. + my $shortpc = $pc; + $shortpc =~ s/^0*//; + # Each line may have a list of names, which includes the function + # and also other functions it has inlined. They are separated (in + # PrintSymbolizedProfile), by --, which is illegal in function names. + my $fullnames; + if (defined($symbol_map->{$shortpc})) { + $fullnames = $symbol_map->{$shortpc}; + } else { + $fullnames = "0x" . $pc; # Just use addresses + } + my $sym = []; + $symbols->{$pc} = $sym; + foreach my $fullname (split("--", $fullnames)) { + my $name = ShortFunctionName($fullname); + push(@{$sym}, $name, "?", $fullname); + } + } + return $symbols; +} + +sub BaseName { + my $file_name = shift; + $file_name =~ s!^.*/!!; # Remove directory name + return $file_name; +} + +sub MakeProfileBaseName { + my ($binary_name, $profile_name) = @_; + my ($host, $baseURL, $path) = ParseProfileURL($profile_name); + my $binary_shortname = BaseName($binary_name); + return sprintf("%s.%s.%s", + $binary_shortname, $main::op_time, $host); +} + +sub FetchDynamicProfile { + my $binary_name = shift; + my $profile_name = shift; + my $fetch_name_only = shift; + my $encourage_patience = shift; + + if (!IsProfileURL($profile_name)) { + return $profile_name; + } else { + my ($host, $baseURL, $path) = ParseProfileURL($profile_name); + if ($path eq "" || $path eq "/") { + # Missing type specifier defaults to cpu-profile + $path = $PROFILE_PAGE; + } + + my $profile_file = MakeProfileBaseName($binary_name, $profile_name); + + my $url = "$baseURL$path"; + my $fetch_timeout = undef; + if ($path =~ m/$PROFILE_PAGE|$PMUPROFILE_PAGE/) { + if ($path =~ m/[?]/) { + $url .= "&"; + } else { + $url .= "?"; + } + $url .= sprintf("seconds=%d", $main::opt_seconds); + $fetch_timeout = $main::opt_seconds * 1.01 + 60; + # Set $profile_type for consumption by PrintSymbolizedProfile. + $main::profile_type = 'cpu'; + } else { + # For non-CPU profiles, we add a type-extension to + # the target profile file name. + my $suffix = $path; + $suffix =~ s,/,.,g; + $profile_file .= $suffix; + # Set $profile_type for consumption by PrintSymbolizedProfile. + if ($path =~ m/$HEAP_PAGE/) { + $main::profile_type = 'heap'; + } elsif ($path =~ m/$GROWTH_PAGE/) { + $main::profile_type = 'growth'; + } elsif ($path =~ m/$CONTENTION_PAGE/) { + $main::profile_type = 'contention'; + } + } + + my $profile_dir = $ENV{"JEPROF_TMPDIR"} || ($ENV{HOME} . "/jeprof"); + if (! -d $profile_dir) { + mkdir($profile_dir) + || die("Unable to create profile directory $profile_dir: $!\n"); + } + my $tmp_profile = "$profile_dir/.tmp.$profile_file"; + my $real_profile = "$profile_dir/$profile_file"; + + if ($fetch_name_only > 0) { + return $real_profile; + } + + my @fetcher = AddFetchTimeout($fetch_timeout, @URL_FETCHER); + my $cmd = ShellEscape(@fetcher, $url) . " > " . ShellEscape($tmp_profile); + if ($path =~ m/$PROFILE_PAGE|$PMUPROFILE_PAGE|$CENSUSPROFILE_PAGE/){ + print STDERR "Gathering CPU profile from $url for $main::opt_seconds seconds to\n ${real_profile}\n"; + if ($encourage_patience) { + print STDERR "Be patient...\n"; + } + } else { + print STDERR "Fetching $path profile from $url to\n ${real_profile}\n"; + } + + (system($cmd) == 0) || error("Failed to get profile: $cmd: $!\n"); + (system("mv", $tmp_profile, $real_profile) == 0) || error("Unable to rename profile\n"); + print STDERR "Wrote profile to $real_profile\n"; + $main::collected_profile = $real_profile; + return $main::collected_profile; + } +} + +# Collect profiles in parallel +sub FetchDynamicProfiles { + my $items = scalar(@main::pfile_args); + my $levels = log($items) / log(2); + + if ($items == 1) { + $main::profile_files[0] = FetchDynamicProfile($main::prog, $main::pfile_args[0], 0, 1); + } else { + # math rounding issues + if ((2 ** $levels) < $items) { + $levels++; + } + my $count = scalar(@main::pfile_args); + for (my $i = 0; $i < $count; $i++) { + $main::profile_files[$i] = FetchDynamicProfile($main::prog, $main::pfile_args[$i], 1, 0); + } + print STDERR "Fetching $count profiles, Be patient...\n"; + FetchDynamicProfilesRecurse($levels, 0, 0); + $main::collected_profile = join(" \\\n ", @main::profile_files); + } +} + +# Recursively fork a process to get enough processes +# collecting profiles +sub FetchDynamicProfilesRecurse { + my $maxlevel = shift; + my $level = shift; + my $position = shift; + + if (my $pid = fork()) { + $position = 0 | ($position << 1); + TryCollectProfile($maxlevel, $level, $position); + wait; + } else { + $position = 1 | ($position << 1); + TryCollectProfile($maxlevel, $level, $position); + cleanup(); + exit(0); + } +} + +# Collect a single profile +sub TryCollectProfile { + my $maxlevel = shift; + my $level = shift; + my $position = shift; + + if ($level >= ($maxlevel - 1)) { + if ($position < scalar(@main::pfile_args)) { + FetchDynamicProfile($main::prog, $main::pfile_args[$position], 0, 0); + } + } else { + FetchDynamicProfilesRecurse($maxlevel, $level+1, $position); + } +} + +##### Parsing code ##### + +# Provide a small streaming-read module to handle very large +# cpu-profile files. Stream in chunks along a sliding window. +# Provides an interface to get one 'slot', correctly handling +# endian-ness differences. A slot is one 32-bit or 64-bit word +# (depending on the input profile). We tell endianness and bit-size +# for the profile by looking at the first 8 bytes: in cpu profiles, +# the second slot is always 3 (we'll accept anything that's not 0). +BEGIN { + package CpuProfileStream; + + sub new { + my ($class, $file, $fname) = @_; + my $self = { file => $file, + base => 0, + stride => 512 * 1024, # must be a multiple of bitsize/8 + slots => [], + unpack_code => "", # N for big-endian, V for little + perl_is_64bit => 1, # matters if profile is 64-bit + }; + bless $self, $class; + # Let unittests adjust the stride + if ($main::opt_test_stride > 0) { + $self->{stride} = $main::opt_test_stride; + } + # Read the first two slots to figure out bitsize and endianness. + my $slots = $self->{slots}; + my $str; + read($self->{file}, $str, 8); + # Set the global $address_length based on what we see here. + # 8 is 32-bit (8 hexadecimal chars); 16 is 64-bit (16 hexadecimal chars). + $address_length = ($str eq (chr(0)x8)) ? 16 : 8; + if ($address_length == 8) { + if (substr($str, 6, 2) eq chr(0)x2) { + $self->{unpack_code} = 'V'; # Little-endian. + } elsif (substr($str, 4, 2) eq chr(0)x2) { + $self->{unpack_code} = 'N'; # Big-endian + } else { + ::error("$fname: header size >= 2**16\n"); + } + @$slots = unpack($self->{unpack_code} . "*", $str); + } else { + # If we're a 64-bit profile, check if we're a 64-bit-capable + # perl. Otherwise, each slot will be represented as a float + # instead of an int64, losing precision and making all the + # 64-bit addresses wrong. We won't complain yet, but will + # later if we ever see a value that doesn't fit in 32 bits. + my $has_q = 0; + eval { $has_q = pack("Q", "1") ? 1 : 1; }; + if (!$has_q) { + $self->{perl_is_64bit} = 0; + } + read($self->{file}, $str, 8); + if (substr($str, 4, 4) eq chr(0)x4) { + # We'd love to use 'Q', but it's a) not universal, b) not endian-proof. + $self->{unpack_code} = 'V'; # Little-endian. + } elsif (substr($str, 0, 4) eq chr(0)x4) { + $self->{unpack_code} = 'N'; # Big-endian + } else { + ::error("$fname: header size >= 2**32\n"); + } + my @pair = unpack($self->{unpack_code} . "*", $str); + # Since we know one of the pair is 0, it's fine to just add them. + @$slots = (0, $pair[0] + $pair[1]); + } + return $self; + } + + # Load more data when we access slots->get(X) which is not yet in memory. + sub overflow { + my ($self) = @_; + my $slots = $self->{slots}; + $self->{base} += $#$slots + 1; # skip over data we're replacing + my $str; + read($self->{file}, $str, $self->{stride}); + if ($address_length == 8) { # the 32-bit case + # This is the easy case: unpack provides 32-bit unpacking primitives. + @$slots = unpack($self->{unpack_code} . "*", $str); + } else { + # We need to unpack 32 bits at a time and combine. + my @b32_values = unpack($self->{unpack_code} . "*", $str); + my @b64_values = (); + for (my $i = 0; $i < $#b32_values; $i += 2) { + # TODO(csilvers): if this is a 32-bit perl, the math below + # could end up in a too-large int, which perl will promote + # to a double, losing necessary precision. Deal with that. + # Right now, we just die. + my ($lo, $hi) = ($b32_values[$i], $b32_values[$i+1]); + if ($self->{unpack_code} eq 'N') { # big-endian + ($lo, $hi) = ($hi, $lo); + } + my $value = $lo + $hi * (2**32); + if (!$self->{perl_is_64bit} && # check value is exactly represented + (($value % (2**32)) != $lo || int($value / (2**32)) != $hi)) { + ::error("Need a 64-bit perl to process this 64-bit profile.\n"); + } + push(@b64_values, $value); + } + @$slots = @b64_values; + } + } + + # Access the i-th long in the file (logically), or -1 at EOF. + sub get { + my ($self, $idx) = @_; + my $slots = $self->{slots}; + while ($#$slots >= 0) { + if ($idx < $self->{base}) { + # The only time we expect a reference to $slots[$i - something] + # after referencing $slots[$i] is reading the very first header. + # Since $stride > |header|, that shouldn't cause any lookback + # errors. And everything after the header is sequential. + print STDERR "Unexpected look-back reading CPU profile"; + return -1; # shrug, don't know what better to return + } elsif ($idx > $self->{base} + $#$slots) { + $self->overflow(); + } else { + return $slots->[$idx - $self->{base}]; + } + } + # If we get here, $slots is [], which means we've reached EOF + return -1; # unique since slots is supposed to hold unsigned numbers + } +} + +# Reads the top, 'header' section of a profile, and returns the last +# line of the header, commonly called a 'header line'. The header +# section of a profile consists of zero or more 'command' lines that +# are instructions to jeprof, which jeprof executes when reading the +# header. All 'command' lines start with a %. After the command +# lines is the 'header line', which is a profile-specific line that +# indicates what type of profile it is, and perhaps other global +# information about the profile. For instance, here's a header line +# for a heap profile: +# heap profile: 53: 38236 [ 5525: 1284029] @ heapprofile +# For historical reasons, the CPU profile does not contain a text- +# readable header line. If the profile looks like a CPU profile, +# this function returns "". If no header line could be found, this +# function returns undef. +# +# The following commands are recognized: +# %warn -- emit the rest of this line to stderr, prefixed by 'WARNING:' +# +# The input file should be in binmode. +sub ReadProfileHeader { + local *PROFILE = shift; + my $firstchar = ""; + my $line = ""; + read(PROFILE, $firstchar, 1); + seek(PROFILE, -1, 1); # unread the firstchar + if ($firstchar !~ /[[:print:]]/) { # is not a text character + return ""; + } + while (defined($line = <PROFILE>)) { + $line =~ s/\r//g; # turn windows-looking lines into unix-looking lines + if ($line =~ /^%warn\s+(.*)/) { # 'warn' command + # Note this matches both '%warn blah\n' and '%warn\n'. + print STDERR "WARNING: $1\n"; # print the rest of the line + } elsif ($line =~ /^%/) { + print STDERR "Ignoring unknown command from profile header: $line"; + } else { + # End of commands, must be the header line. + return $line; + } + } + return undef; # got to EOF without seeing a header line +} + +sub IsSymbolizedProfileFile { + my $file_name = shift; + if (!(-e $file_name) || !(-r $file_name)) { + return 0; + } + # Check if the file contains a symbol-section marker. + open(TFILE, "<$file_name"); + binmode TFILE; + my $firstline = ReadProfileHeader(*TFILE); + close(TFILE); + if (!$firstline) { + return 0; + } + $SYMBOL_PAGE =~ m,[^/]+$,; # matches everything after the last slash + my $symbol_marker = $&; + return $firstline =~ /^--- *$symbol_marker/; +} + +# Parse profile generated by common/profiler.cc and return a reference +# to a map: +# $result->{version} Version number of profile file +# $result->{period} Sampling period (in microseconds) +# $result->{profile} Profile object +# $result->{threads} Map of thread IDs to profile objects +# $result->{map} Memory map info from profile +# $result->{pcs} Hash of all PC values seen, key is hex address +sub ReadProfile { + my $prog = shift; + my $fname = shift; + my $result; # return value + + $CONTENTION_PAGE =~ m,[^/]+$,; # matches everything after the last slash + my $contention_marker = $&; + $GROWTH_PAGE =~ m,[^/]+$,; # matches everything after the last slash + my $growth_marker = $&; + $SYMBOL_PAGE =~ m,[^/]+$,; # matches everything after the last slash + my $symbol_marker = $&; + $PROFILE_PAGE =~ m,[^/]+$,; # matches everything after the last slash + my $profile_marker = $&; + $HEAP_PAGE =~ m,[^/]+$,; # matches everything after the last slash + my $heap_marker = $&; + + # Look at first line to see if it is a heap or a CPU profile. + # CPU profile may start with no header at all, and just binary data + # (starting with \0\0\0\0) -- in that case, don't try to read the + # whole firstline, since it may be gigabytes(!) of data. + open(PROFILE, "<$fname") || error("$fname: $!\n"); + binmode PROFILE; # New perls do UTF-8 processing + my $header = ReadProfileHeader(*PROFILE); + if (!defined($header)) { # means "at EOF" + error("Profile is empty.\n"); + } + + my $symbols; + if ($header =~ m/^--- *$symbol_marker/o) { + # Verify that the user asked for a symbolized profile + if (!$main::use_symbolized_profile) { + # we have both a binary and symbolized profiles, abort + error("FATAL ERROR: Symbolized profile\n $fname\ncannot be used with " . + "a binary arg. Try again without passing\n $prog\n"); + } + # Read the symbol section of the symbolized profile file. + $symbols = ReadSymbols(*PROFILE{IO}); + # Read the next line to get the header for the remaining profile. + $header = ReadProfileHeader(*PROFILE) || ""; + } + + if ($header =~ m/^--- *($heap_marker|$growth_marker)/o) { + # Skip "--- ..." line for profile types that have their own headers. + $header = ReadProfileHeader(*PROFILE) || ""; + } + + $main::profile_type = ''; + + if ($header =~ m/^heap profile:.*$growth_marker/o) { + $main::profile_type = 'growth'; + $result = ReadHeapProfile($prog, *PROFILE, $header); + } elsif ($header =~ m/^heap profile:/) { + $main::profile_type = 'heap'; + $result = ReadHeapProfile($prog, *PROFILE, $header); + } elsif ($header =~ m/^heap/) { + $main::profile_type = 'heap'; + $result = ReadThreadedHeapProfile($prog, $fname, $header); + } elsif ($header =~ m/^--- *$contention_marker/o) { + $main::profile_type = 'contention'; + $result = ReadSynchProfile($prog, *PROFILE); + } elsif ($header =~ m/^--- *Stacks:/) { + print STDERR + "Old format contention profile: mistakenly reports " . + "condition variable signals as lock contentions.\n"; + $main::profile_type = 'contention'; + $result = ReadSynchProfile($prog, *PROFILE); + } elsif ($header =~ m/^--- *$profile_marker/) { + # the binary cpu profile data starts immediately after this line + $main::profile_type = 'cpu'; + $result = ReadCPUProfile($prog, $fname, *PROFILE); + } else { + if (defined($symbols)) { + # a symbolized profile contains a format we don't recognize, bail out + error("$fname: Cannot recognize profile section after symbols.\n"); + } + # no ascii header present -- must be a CPU profile + $main::profile_type = 'cpu'; + $result = ReadCPUProfile($prog, $fname, *PROFILE); + } + + close(PROFILE); + + # if we got symbols along with the profile, return those as well + if (defined($symbols)) { + $result->{symbols} = $symbols; + } + + return $result; +} + +# Subtract one from caller pc so we map back to call instr. +# However, don't do this if we're reading a symbolized profile +# file, in which case the subtract-one was done when the file +# was written. +# +# We apply the same logic to all readers, though ReadCPUProfile uses an +# independent implementation. +sub FixCallerAddresses { + my $stack = shift; + # --raw/http: Always subtract one from pc's, because PrintSymbolizedProfile() + # dumps unadjusted profiles. + { + $stack =~ /(\s)/; + my $delimiter = $1; + my @addrs = split(' ', $stack); + my @fixedaddrs; + $#fixedaddrs = $#addrs; + if ($#addrs >= 0) { + $fixedaddrs[0] = $addrs[0]; + } + for (my $i = 1; $i <= $#addrs; $i++) { + $fixedaddrs[$i] = AddressSub($addrs[$i], "0x1"); + } + return join $delimiter, @fixedaddrs; + } +} + +# CPU profile reader +sub ReadCPUProfile { + my $prog = shift; + my $fname = shift; # just used for logging + local *PROFILE = shift; + my $version; + my $period; + my $i; + my $profile = {}; + my $pcs = {}; + + # Parse string into array of slots. + my $slots = CpuProfileStream->new(*PROFILE, $fname); + + # Read header. The current header version is a 5-element structure + # containing: + # 0: header count (always 0) + # 1: header "words" (after this one: 3) + # 2: format version (0) + # 3: sampling period (usec) + # 4: unused padding (always 0) + if ($slots->get(0) != 0 ) { + error("$fname: not a profile file, or old format profile file\n"); + } + $i = 2 + $slots->get(1); + $version = $slots->get(2); + $period = $slots->get(3); + # Do some sanity checking on these header values. + if ($version > (2**32) || $period > (2**32) || $i > (2**32) || $i < 5) { + error("$fname: not a profile file, or corrupted profile file\n"); + } + + # Parse profile + while ($slots->get($i) != -1) { + my $n = $slots->get($i++); + my $d = $slots->get($i++); + if ($d > (2**16)) { # TODO(csilvers): what's a reasonable max-stack-depth? + my $addr = sprintf("0%o", $i * ($address_length == 8 ? 4 : 8)); + print STDERR "At index $i (address $addr):\n"; + error("$fname: stack trace depth >= 2**32\n"); + } + if ($slots->get($i) == 0) { + # End of profile data marker + $i += $d; + last; + } + + # Make key out of the stack entries + my @k = (); + for (my $j = 0; $j < $d; $j++) { + my $pc = $slots->get($i+$j); + # Subtract one from caller pc so we map back to call instr. + $pc--; + $pc = sprintf("%0*x", $address_length, $pc); + $pcs->{$pc} = 1; + push @k, $pc; + } + + AddEntry($profile, (join "\n", @k), $n); + $i += $d; + } + + # Parse map + my $map = ''; + seek(PROFILE, $i * 4, 0); + read(PROFILE, $map, (stat PROFILE)[7]); + + my $r = {}; + $r->{version} = $version; + $r->{period} = $period; + $r->{profile} = $profile; + $r->{libs} = ParseLibraries($prog, $map, $pcs); + $r->{pcs} = $pcs; + + return $r; +} + +sub HeapProfileIndex { + my $index = 1; + if ($main::opt_inuse_space) { + $index = 1; + } elsif ($main::opt_inuse_objects) { + $index = 0; + } elsif ($main::opt_alloc_space) { + $index = 3; + } elsif ($main::opt_alloc_objects) { + $index = 2; + } + return $index; +} + +sub ReadMappedLibraries { + my $fh = shift; + my $map = ""; + # Read the /proc/self/maps data + while (<$fh>) { + s/\r//g; # turn windows-looking lines into unix-looking lines + $map .= $_; + } + return $map; +} + +sub ReadMemoryMap { + my $fh = shift; + my $map = ""; + # Read /proc/self/maps data as formatted by DumpAddressMap() + my $buildvar = ""; + while (<PROFILE>) { + s/\r//g; # turn windows-looking lines into unix-looking lines + # Parse "build=<dir>" specification if supplied + if (m/^\s*build=(.*)\n/) { + $buildvar = $1; + } + + # Expand "$build" variable if available + $_ =~ s/\$build\b/$buildvar/g; + + $map .= $_; + } + return $map; +} + +sub AdjustSamples { + my ($sample_adjustment, $sampling_algorithm, $n1, $s1, $n2, $s2) = @_; + if ($sample_adjustment) { + if ($sampling_algorithm == 2) { + # Remote-heap version 2 + # The sampling frequency is the rate of a Poisson process. + # This means that the probability of sampling an allocation of + # size X with sampling rate Y is 1 - exp(-X/Y) + if ($n1 != 0) { + my $ratio = (($s1*1.0)/$n1)/($sample_adjustment); + my $scale_factor = 1/(1 - exp(-$ratio)); + $n1 *= $scale_factor; + $s1 *= $scale_factor; + } + if ($n2 != 0) { + my $ratio = (($s2*1.0)/$n2)/($sample_adjustment); + my $scale_factor = 1/(1 - exp(-$ratio)); + $n2 *= $scale_factor; + $s2 *= $scale_factor; + } + } else { + # Remote-heap version 1 + my $ratio; + $ratio = (($s1*1.0)/$n1)/($sample_adjustment); + if ($ratio < 1) { + $n1 /= $ratio; + $s1 /= $ratio; + } + $ratio = (($s2*1.0)/$n2)/($sample_adjustment); + if ($ratio < 1) { + $n2 /= $ratio; + $s2 /= $ratio; + } + } + } + return ($n1, $s1, $n2, $s2); +} + +sub ReadHeapProfile { + my $prog = shift; + local *PROFILE = shift; + my $header = shift; + + my $index = HeapProfileIndex(); + + # Find the type of this profile. The header line looks like: + # heap profile: 1246: 8800744 [ 1246: 8800744] @ <heap-url>/266053 + # There are two pairs <count: size>, the first inuse objects/space, and the + # second allocated objects/space. This is followed optionally by a profile + # type, and if that is present, optionally by a sampling frequency. + # For remote heap profiles (v1): + # The interpretation of the sampling frequency is that the profiler, for + # each sample, calculates a uniformly distributed random integer less than + # the given value, and records the next sample after that many bytes have + # been allocated. Therefore, the expected sample interval is half of the + # given frequency. By default, if not specified, the expected sample + # interval is 128KB. Only remote-heap-page profiles are adjusted for + # sample size. + # For remote heap profiles (v2): + # The sampling frequency is the rate of a Poisson process. This means that + # the probability of sampling an allocation of size X with sampling rate Y + # is 1 - exp(-X/Y) + # For version 2, a typical header line might look like this: + # heap profile: 1922: 127792360 [ 1922: 127792360] @ <heap-url>_v2/524288 + # the trailing number (524288) is the sampling rate. (Version 1 showed + # double the 'rate' here) + my $sampling_algorithm = 0; + my $sample_adjustment = 0; + chomp($header); + my $type = "unknown"; + if ($header =~ m"^heap profile:\s*(\d+):\s+(\d+)\s+\[\s*(\d+):\s+(\d+)\](\s*@\s*([^/]*)(/(\d+))?)?") { + if (defined($6) && ($6 ne '')) { + $type = $6; + my $sample_period = $8; + # $type is "heapprofile" for profiles generated by the + # heap-profiler, and either "heap" or "heap_v2" for profiles + # generated by sampling directly within tcmalloc. It can also + # be "growth" for heap-growth profiles. The first is typically + # found for profiles generated locally, and the others for + # remote profiles. + if (($type eq "heapprofile") || ($type !~ /heap/) ) { + # No need to adjust for the sampling rate with heap-profiler-derived data + $sampling_algorithm = 0; + } elsif ($type =~ /_v2/) { + $sampling_algorithm = 2; # version 2 sampling + if (defined($sample_period) && ($sample_period ne '')) { + $sample_adjustment = int($sample_period); + } + } else { + $sampling_algorithm = 1; # version 1 sampling + if (defined($sample_period) && ($sample_period ne '')) { + $sample_adjustment = int($sample_period)/2; + } + } + } else { + # We detect whether or not this is a remote-heap profile by checking + # that the total-allocated stats ($n2,$s2) are exactly the + # same as the in-use stats ($n1,$s1). It is remotely conceivable + # that a non-remote-heap profile may pass this check, but it is hard + # to imagine how that could happen. + # In this case it's so old it's guaranteed to be remote-heap version 1. + my ($n1, $s1, $n2, $s2) = ($1, $2, $3, $4); + if (($n1 == $n2) && ($s1 == $s2)) { + # This is likely to be a remote-heap based sample profile + $sampling_algorithm = 1; + } + } + } + + if ($sampling_algorithm > 0) { + # For remote-heap generated profiles, adjust the counts and sizes to + # account for the sample rate (we sample once every 128KB by default). + if ($sample_adjustment == 0) { + # Turn on profile adjustment. + $sample_adjustment = 128*1024; + print STDERR "Adjusting heap profiles for 1-in-128KB sampling rate\n"; + } else { + printf STDERR ("Adjusting heap profiles for 1-in-%d sampling rate\n", + $sample_adjustment); + } + if ($sampling_algorithm > 1) { + # We don't bother printing anything for the original version (version 1) + printf STDERR "Heap version $sampling_algorithm\n"; + } + } + + my $profile = {}; + my $pcs = {}; + my $map = ""; + + while (<PROFILE>) { + s/\r//g; # turn windows-looking lines into unix-looking lines + if (/^MAPPED_LIBRARIES:/) { + $map .= ReadMappedLibraries(*PROFILE); + last; + } + + if (/^--- Memory map:/) { + $map .= ReadMemoryMap(*PROFILE); + last; + } + + # Read entry of the form: + # <count1>: <bytes1> [<count2>: <bytes2>] @ a1 a2 a3 ... an + s/^\s*//; + s/\s*$//; + if (m/^\s*(\d+):\s+(\d+)\s+\[\s*(\d+):\s+(\d+)\]\s+@\s+(.*)$/) { + my $stack = $5; + my ($n1, $s1, $n2, $s2) = ($1, $2, $3, $4); + my @counts = AdjustSamples($sample_adjustment, $sampling_algorithm, + $n1, $s1, $n2, $s2); + AddEntries($profile, $pcs, FixCallerAddresses($stack), $counts[$index]); + } + } + + my $r = {}; + $r->{version} = "heap"; + $r->{period} = 1; + $r->{profile} = $profile; + $r->{libs} = ParseLibraries($prog, $map, $pcs); + $r->{pcs} = $pcs; + return $r; +} + +sub ReadThreadedHeapProfile { + my ($prog, $fname, $header) = @_; + + my $index = HeapProfileIndex(); + my $sampling_algorithm = 0; + my $sample_adjustment = 0; + chomp($header); + my $type = "unknown"; + # Assuming a very specific type of header for now. + if ($header =~ m"^heap_v2/(\d+)") { + $type = "_v2"; + $sampling_algorithm = 2; + $sample_adjustment = int($1); + } + if ($type ne "_v2" || !defined($sample_adjustment)) { + die "Threaded heap profiles require v2 sampling with a sample rate\n"; + } + + my $profile = {}; + my $thread_profiles = {}; + my $pcs = {}; + my $map = ""; + my $stack = ""; + + while (<PROFILE>) { + s/\r//g; + if (/^MAPPED_LIBRARIES:/) { + $map .= ReadMappedLibraries(*PROFILE); + last; + } + + if (/^--- Memory map:/) { + $map .= ReadMemoryMap(*PROFILE); + last; + } + + # Read entry of the form: + # @ a1 a2 ... an + # t*: <count1>: <bytes1> [<count2>: <bytes2>] + # t1: <count1>: <bytes1> [<count2>: <bytes2>] + # ... + # tn: <count1>: <bytes1> [<count2>: <bytes2>] + s/^\s*//; + s/\s*$//; + if (m/^@\s+(.*)$/) { + $stack = $1; + } elsif (m/^\s*(t(\*|\d+)):\s+(\d+):\s+(\d+)\s+\[\s*(\d+):\s+(\d+)\]$/) { + if ($stack eq "") { + # Still in the header, so this is just a per-thread summary. + next; + } + my $thread = $2; + my ($n1, $s1, $n2, $s2) = ($3, $4, $5, $6); + my @counts = AdjustSamples($sample_adjustment, $sampling_algorithm, + $n1, $s1, $n2, $s2); + if ($thread eq "*") { + AddEntries($profile, $pcs, FixCallerAddresses($stack), $counts[$index]); + } else { + if (!exists($thread_profiles->{$thread})) { + $thread_profiles->{$thread} = {}; + } + AddEntries($thread_profiles->{$thread}, $pcs, + FixCallerAddresses($stack), $counts[$index]); + } + } + } + + my $r = {}; + $r->{version} = "heap"; + $r->{period} = 1; + $r->{profile} = $profile; + $r->{threads} = $thread_profiles; + $r->{libs} = ParseLibraries($prog, $map, $pcs); + $r->{pcs} = $pcs; + return $r; +} + +sub ReadSynchProfile { + my $prog = shift; + local *PROFILE = shift; + my $header = shift; + + my $map = ''; + my $profile = {}; + my $pcs = {}; + my $sampling_period = 1; + my $cyclespernanosec = 2.8; # Default assumption for old binaries + my $seen_clockrate = 0; + my $line; + + my $index = 0; + if ($main::opt_total_delay) { + $index = 0; + } elsif ($main::opt_contentions) { + $index = 1; + } elsif ($main::opt_mean_delay) { + $index = 2; + } + + while ( $line = <PROFILE> ) { + $line =~ s/\r//g; # turn windows-looking lines into unix-looking lines + if ( $line =~ /^\s*(\d+)\s+(\d+) \@\s*(.*?)\s*$/ ) { + my ($cycles, $count, $stack) = ($1, $2, $3); + + # Convert cycles to nanoseconds + $cycles /= $cyclespernanosec; + + # Adjust for sampling done by application + $cycles *= $sampling_period; + $count *= $sampling_period; + + my @values = ($cycles, $count, $cycles / $count); + AddEntries($profile, $pcs, FixCallerAddresses($stack), $values[$index]); + + } elsif ( $line =~ /^(slow release).*thread \d+ \@\s*(.*?)\s*$/ || + $line =~ /^\s*(\d+) \@\s*(.*?)\s*$/ ) { + my ($cycles, $stack) = ($1, $2); + if ($cycles !~ /^\d+$/) { + next; + } + + # Convert cycles to nanoseconds + $cycles /= $cyclespernanosec; + + # Adjust for sampling done by application + $cycles *= $sampling_period; + + AddEntries($profile, $pcs, FixCallerAddresses($stack), $cycles); + + } elsif ( $line =~ m/^([a-z][^=]*)=(.*)$/ ) { + my ($variable, $value) = ($1,$2); + for ($variable, $value) { + s/^\s+//; + s/\s+$//; + } + if ($variable eq "cycles/second") { + $cyclespernanosec = $value / 1e9; + $seen_clockrate = 1; + } elsif ($variable eq "sampling period") { + $sampling_period = $value; + } elsif ($variable eq "ms since reset") { + # Currently nothing is done with this value in jeprof + # So we just silently ignore it for now + } elsif ($variable eq "discarded samples") { + # Currently nothing is done with this value in jeprof + # So we just silently ignore it for now + } else { + printf STDERR ("Ignoring unnknown variable in /contention output: " . + "'%s' = '%s'\n",$variable,$value); + } + } else { + # Memory map entry + $map .= $line; + } + } + + if (!$seen_clockrate) { + printf STDERR ("No cycles/second entry in profile; Guessing %.1f GHz\n", + $cyclespernanosec); + } + + my $r = {}; + $r->{version} = 0; + $r->{period} = $sampling_period; + $r->{profile} = $profile; + $r->{libs} = ParseLibraries($prog, $map, $pcs); + $r->{pcs} = $pcs; + return $r; +} + +# Given a hex value in the form "0x1abcd" or "1abcd", return either +# "0001abcd" or "000000000001abcd", depending on the current (global) +# address length. +sub HexExtend { + my $addr = shift; + + $addr =~ s/^(0x)?0*//; + my $zeros_needed = $address_length - length($addr); + if ($zeros_needed < 0) { + printf STDERR "Warning: address $addr is longer than address length $address_length\n"; + return $addr; + } + return ("0" x $zeros_needed) . $addr; +} + +##### Symbol extraction ##### + +# Aggressively search the lib_prefix values for the given library +# If all else fails, just return the name of the library unmodified. +# If the lib_prefix is "/my/path,/other/path" and $file is "/lib/dir/mylib.so" +# it will search the following locations in this order, until it finds a file: +# /my/path/lib/dir/mylib.so +# /other/path/lib/dir/mylib.so +# /my/path/dir/mylib.so +# /other/path/dir/mylib.so +# /my/path/mylib.so +# /other/path/mylib.so +# /lib/dir/mylib.so (returned as last resort) +sub FindLibrary { + my $file = shift; + my $suffix = $file; + + # Search for the library as described above + do { + foreach my $prefix (@prefix_list) { + my $fullpath = $prefix . $suffix; + if (-e $fullpath) { + return $fullpath; + } + } + } while ($suffix =~ s|^/[^/]+/|/|); + return $file; +} + +# Return path to library with debugging symbols. +# For libc libraries, the copy in /usr/lib/debug contains debugging symbols +sub DebuggingLibrary { + my $file = shift; + if ($file =~ m|^/|) { + if (-f "/usr/lib/debug$file") { + return "/usr/lib/debug$file"; + } elsif (-f "/usr/lib/debug$file.debug") { + return "/usr/lib/debug$file.debug"; + } + } + return undef; +} + +# Parse text section header of a library using objdump +sub ParseTextSectionHeaderFromObjdump { + my $lib = shift; + + my $size = undef; + my $vma; + my $file_offset; + # Get objdump output from the library file to figure out how to + # map between mapped addresses and addresses in the library. + my $cmd = ShellEscape($obj_tool_map{"objdump"}, "-h", $lib); + open(OBJDUMP, "$cmd |") || error("$cmd: $!\n"); + while (<OBJDUMP>) { + s/\r//g; # turn windows-looking lines into unix-looking lines + # Idx Name Size VMA LMA File off Algn + # 10 .text 00104b2c 420156f0 420156f0 000156f0 2**4 + # For 64-bit objects, VMA and LMA will be 16 hex digits, size and file + # offset may still be 8. But AddressSub below will still handle that. + my @x = split; + if (($#x >= 6) && ($x[1] eq '.text')) { + $size = $x[2]; + $vma = $x[3]; + $file_offset = $x[5]; + last; + } + } + close(OBJDUMP); + + if (!defined($size)) { + return undef; + } + + my $r = {}; + $r->{size} = $size; + $r->{vma} = $vma; + $r->{file_offset} = $file_offset; + + return $r; +} + +# Parse text section header of a library using otool (on OS X) +sub ParseTextSectionHeaderFromOtool { + my $lib = shift; + + my $size = undef; + my $vma = undef; + my $file_offset = undef; + # Get otool output from the library file to figure out how to + # map between mapped addresses and addresses in the library. + my $command = ShellEscape($obj_tool_map{"otool"}, "-l", $lib); + open(OTOOL, "$command |") || error("$command: $!\n"); + my $cmd = ""; + my $sectname = ""; + my $segname = ""; + foreach my $line (<OTOOL>) { + $line =~ s/\r//g; # turn windows-looking lines into unix-looking lines + # Load command <#> + # cmd LC_SEGMENT + # [...] + # Section + # sectname __text + # segname __TEXT + # addr 0x000009f8 + # size 0x00018b9e + # offset 2552 + # align 2^2 (4) + # We will need to strip off the leading 0x from the hex addresses, + # and convert the offset into hex. + if ($line =~ /Load command/) { + $cmd = ""; + $sectname = ""; + $segname = ""; + } elsif ($line =~ /Section/) { + $sectname = ""; + $segname = ""; + } elsif ($line =~ /cmd (\w+)/) { + $cmd = $1; + } elsif ($line =~ /sectname (\w+)/) { + $sectname = $1; + } elsif ($line =~ /segname (\w+)/) { + $segname = $1; + } elsif (!(($cmd eq "LC_SEGMENT" || $cmd eq "LC_SEGMENT_64") && + $sectname eq "__text" && + $segname eq "__TEXT")) { + next; + } elsif ($line =~ /\baddr 0x([0-9a-fA-F]+)/) { + $vma = $1; + } elsif ($line =~ /\bsize 0x([0-9a-fA-F]+)/) { + $size = $1; + } elsif ($line =~ /\boffset ([0-9]+)/) { + $file_offset = sprintf("%016x", $1); + } + if (defined($vma) && defined($size) && defined($file_offset)) { + last; + } + } + close(OTOOL); + + if (!defined($vma) || !defined($size) || !defined($file_offset)) { + return undef; + } + + my $r = {}; + $r->{size} = $size; + $r->{vma} = $vma; + $r->{file_offset} = $file_offset; + + return $r; +} + +sub ParseTextSectionHeader { + # obj_tool_map("otool") is only defined if we're in a Mach-O environment + if (defined($obj_tool_map{"otool"})) { + my $r = ParseTextSectionHeaderFromOtool(@_); + if (defined($r)){ + return $r; + } + } + # If otool doesn't work, or we don't have it, fall back to objdump + return ParseTextSectionHeaderFromObjdump(@_); +} + +# Split /proc/pid/maps dump into a list of libraries +sub ParseLibraries { + return if $main::use_symbol_page; # We don't need libraries info. + my $prog = Cwd::abs_path(shift); + my $map = shift; + my $pcs = shift; + + my $result = []; + my $h = "[a-f0-9]+"; + my $zero_offset = HexExtend("0"); + + my $buildvar = ""; + foreach my $l (split("\n", $map)) { + if ($l =~ m/^\s*build=(.*)$/) { + $buildvar = $1; + } + + my $start; + my $finish; + my $offset; + my $lib; + if ($l =~ /^($h)-($h)\s+..x.\s+($h)\s+\S+:\S+\s+\d+\s+(\S+\.(so|dll|dylib|bundle)((\.\d+)+\w*(\.\d+){0,3})?)$/i) { + # Full line from /proc/self/maps. Example: + # 40000000-40015000 r-xp 00000000 03:01 12845071 /lib/ld-2.3.2.so + $start = HexExtend($1); + $finish = HexExtend($2); + $offset = HexExtend($3); + $lib = $4; + $lib =~ s|\\|/|g; # turn windows-style paths into unix-style paths + } elsif ($l =~ /^\s*($h)-($h):\s*(\S+\.so(\.\d+)*)/) { + # Cooked line from DumpAddressMap. Example: + # 40000000-40015000: /lib/ld-2.3.2.so + $start = HexExtend($1); + $finish = HexExtend($2); + $offset = $zero_offset; + $lib = $3; + } elsif (($l =~ /^($h)-($h)\s+..x.\s+($h)\s+\S+:\S+\s+\d+\s+(\S+)$/i) && ($4 eq $prog)) { + # PIEs and address space randomization do not play well with our + # default assumption that main executable is at lowest + # addresses. So we're detecting main executable in + # /proc/self/maps as well. + $start = HexExtend($1); + $finish = HexExtend($2); + $offset = HexExtend($3); + $lib = $4; + $lib =~ s|\\|/|g; # turn windows-style paths into unix-style paths + } + # FreeBSD 10.0 virtual memory map /proc/curproc/map as defined in + # function procfs_doprocmap (sys/fs/procfs/procfs_map.c) + # + # Example: + # 0x800600000 0x80061a000 26 0 0xfffff800035a0000 r-x 75 33 0x1004 COW NC vnode /libexec/ld-elf.s + # o.1 NCH -1 + elsif ($l =~ /^(0x$h)\s(0x$h)\s\d+\s\d+\s0x$h\sr-x\s\d+\s\d+\s0x\d+\s(COW|NCO)\s(NC|NNC)\svnode\s(\S+\.so(\.\d+)*)/) { + $start = HexExtend($1); + $finish = HexExtend($2); + $offset = $zero_offset; + $lib = FindLibrary($5); + + } else { + next; + } + + # Expand "$build" variable if available + $lib =~ s/\$build\b/$buildvar/g; + + $lib = FindLibrary($lib); + + # Check for pre-relocated libraries, which use pre-relocated symbol tables + # and thus require adjusting the offset that we'll use to translate + # VM addresses into symbol table addresses. + # Only do this if we're not going to fetch the symbol table from a + # debugging copy of the library. + if (!DebuggingLibrary($lib)) { + my $text = ParseTextSectionHeader($lib); + if (defined($text)) { + my $vma_offset = AddressSub($text->{vma}, $text->{file_offset}); + $offset = AddressAdd($offset, $vma_offset); + } + } + + if($main::opt_debug) { printf STDERR "$start:$finish ($offset) $lib\n"; } + push(@{$result}, [$lib, $start, $finish, $offset]); + } + + # Append special entry for additional library (not relocated) + if ($main::opt_lib ne "") { + my $text = ParseTextSectionHeader($main::opt_lib); + if (defined($text)) { + my $start = $text->{vma}; + my $finish = AddressAdd($start, $text->{size}); + + push(@{$result}, [$main::opt_lib, $start, $finish, $start]); + } + } + + # Append special entry for the main program. This covers + # 0..max_pc_value_seen, so that we assume pc values not found in one + # of the library ranges will be treated as coming from the main + # program binary. + my $min_pc = HexExtend("0"); + my $max_pc = $min_pc; # find the maximal PC value in any sample + foreach my $pc (keys(%{$pcs})) { + if (HexExtend($pc) gt $max_pc) { $max_pc = HexExtend($pc); } + } + push(@{$result}, [$prog, $min_pc, $max_pc, $zero_offset]); + + return $result; +} + +# Add two hex addresses of length $address_length. +# Run jeprof --test for unit test if this is changed. +sub AddressAdd { + my $addr1 = shift; + my $addr2 = shift; + my $sum; + + if ($address_length == 8) { + # Perl doesn't cope with wraparound arithmetic, so do it explicitly: + $sum = (hex($addr1)+hex($addr2)) % (0x10000000 * 16); + return sprintf("%08x", $sum); + + } else { + # Do the addition in 7-nibble chunks to trivialize carry handling. + + if ($main::opt_debug and $main::opt_test) { + print STDERR "AddressAdd $addr1 + $addr2 = "; + } + + my $a1 = substr($addr1,-7); + $addr1 = substr($addr1,0,-7); + my $a2 = substr($addr2,-7); + $addr2 = substr($addr2,0,-7); + $sum = hex($a1) + hex($a2); + my $c = 0; + if ($sum > 0xfffffff) { + $c = 1; + $sum -= 0x10000000; + } + my $r = sprintf("%07x", $sum); + + $a1 = substr($addr1,-7); + $addr1 = substr($addr1,0,-7); + $a2 = substr($addr2,-7); + $addr2 = substr($addr2,0,-7); + $sum = hex($a1) + hex($a2) + $c; + $c = 0; + if ($sum > 0xfffffff) { + $c = 1; + $sum -= 0x10000000; + } + $r = sprintf("%07x", $sum) . $r; + + $sum = hex($addr1) + hex($addr2) + $c; + if ($sum > 0xff) { $sum -= 0x100; } + $r = sprintf("%02x", $sum) . $r; + + if ($main::opt_debug and $main::opt_test) { print STDERR "$r\n"; } + + return $r; + } +} + + +# Subtract two hex addresses of length $address_length. +# Run jeprof --test for unit test if this is changed. +sub AddressSub { + my $addr1 = shift; + my $addr2 = shift; + my $diff; + + if ($address_length == 8) { + # Perl doesn't cope with wraparound arithmetic, so do it explicitly: + $diff = (hex($addr1)-hex($addr2)) % (0x10000000 * 16); + return sprintf("%08x", $diff); + + } else { + # Do the addition in 7-nibble chunks to trivialize borrow handling. + # if ($main::opt_debug) { print STDERR "AddressSub $addr1 - $addr2 = "; } + + my $a1 = hex(substr($addr1,-7)); + $addr1 = substr($addr1,0,-7); + my $a2 = hex(substr($addr2,-7)); + $addr2 = substr($addr2,0,-7); + my $b = 0; + if ($a2 > $a1) { + $b = 1; + $a1 += 0x10000000; + } + $diff = $a1 - $a2; + my $r = sprintf("%07x", $diff); + + $a1 = hex(substr($addr1,-7)); + $addr1 = substr($addr1,0,-7); + $a2 = hex(substr($addr2,-7)) + $b; + $addr2 = substr($addr2,0,-7); + $b = 0; + if ($a2 > $a1) { + $b = 1; + $a1 += 0x10000000; + } + $diff = $a1 - $a2; + $r = sprintf("%07x", $diff) . $r; + + $a1 = hex($addr1); + $a2 = hex($addr2) + $b; + if ($a2 > $a1) { $a1 += 0x100; } + $diff = $a1 - $a2; + $r = sprintf("%02x", $diff) . $r; + + # if ($main::opt_debug) { print STDERR "$r\n"; } + + return $r; + } +} + +# Increment a hex addresses of length $address_length. +# Run jeprof --test for unit test if this is changed. +sub AddressInc { + my $addr = shift; + my $sum; + + if ($address_length == 8) { + # Perl doesn't cope with wraparound arithmetic, so do it explicitly: + $sum = (hex($addr)+1) % (0x10000000 * 16); + return sprintf("%08x", $sum); + + } else { + # Do the addition in 7-nibble chunks to trivialize carry handling. + # We are always doing this to step through the addresses in a function, + # and will almost never overflow the first chunk, so we check for this + # case and exit early. + + # if ($main::opt_debug) { print STDERR "AddressInc $addr1 = "; } + + my $a1 = substr($addr,-7); + $addr = substr($addr,0,-7); + $sum = hex($a1) + 1; + my $r = sprintf("%07x", $sum); + if ($sum <= 0xfffffff) { + $r = $addr . $r; + # if ($main::opt_debug) { print STDERR "$r\n"; } + return HexExtend($r); + } else { + $r = "0000000"; + } + + $a1 = substr($addr,-7); + $addr = substr($addr,0,-7); + $sum = hex($a1) + 1; + $r = sprintf("%07x", $sum) . $r; + if ($sum <= 0xfffffff) { + $r = $addr . $r; + # if ($main::opt_debug) { print STDERR "$r\n"; } + return HexExtend($r); + } else { + $r = "00000000000000"; + } + + $sum = hex($addr) + 1; + if ($sum > 0xff) { $sum -= 0x100; } + $r = sprintf("%02x", $sum) . $r; + + # if ($main::opt_debug) { print STDERR "$r\n"; } + return $r; + } +} + +# Extract symbols for all PC values found in profile +sub ExtractSymbols { + my $libs = shift; + my $pcset = shift; + + my $symbols = {}; + + # Map each PC value to the containing library. To make this faster, + # we sort libraries by their starting pc value (highest first), and + # advance through the libraries as we advance the pc. Sometimes the + # addresses of libraries may overlap with the addresses of the main + # binary, so to make sure the libraries 'win', we iterate over the + # libraries in reverse order (which assumes the binary doesn't start + # in the middle of a library, which seems a fair assumption). + my @pcs = (sort { $a cmp $b } keys(%{$pcset})); # pcset is 0-extended strings + foreach my $lib (sort {$b->[1] cmp $a->[1]} @{$libs}) { + my $libname = $lib->[0]; + my $start = $lib->[1]; + my $finish = $lib->[2]; + my $offset = $lib->[3]; + + # Use debug library if it exists + my $debug_libname = DebuggingLibrary($libname); + if ($debug_libname) { + $libname = $debug_libname; + } + + # Get list of pcs that belong in this library. + my $contained = []; + my ($start_pc_index, $finish_pc_index); + # Find smallest finish_pc_index such that $finish < $pc[$finish_pc_index]. + for ($finish_pc_index = $#pcs + 1; $finish_pc_index > 0; + $finish_pc_index--) { + last if $pcs[$finish_pc_index - 1] le $finish; + } + # Find smallest start_pc_index such that $start <= $pc[$start_pc_index]. + for ($start_pc_index = $finish_pc_index; $start_pc_index > 0; + $start_pc_index--) { + last if $pcs[$start_pc_index - 1] lt $start; + } + # This keeps PC values higher than $pc[$finish_pc_index] in @pcs, + # in case there are overlaps in libraries and the main binary. + @{$contained} = splice(@pcs, $start_pc_index, + $finish_pc_index - $start_pc_index); + # Map to symbols + MapToSymbols($libname, AddressSub($start, $offset), $contained, $symbols); + } + + return $symbols; +} + +# Map list of PC values to symbols for a given image +sub MapToSymbols { + my $image = shift; + my $offset = shift; + my $pclist = shift; + my $symbols = shift; + + my $debug = 0; + + # Ignore empty binaries + if ($#{$pclist} < 0) { return; } + + # Figure out the addr2line command to use + my $addr2line = $obj_tool_map{"addr2line"}; + my $cmd = ShellEscape($addr2line, "-f", "-C", "-e", $image); + if (exists $obj_tool_map{"addr2line_pdb"}) { + $addr2line = $obj_tool_map{"addr2line_pdb"}; + $cmd = ShellEscape($addr2line, "--demangle", "-f", "-C", "-e", $image); + } + + # If "addr2line" isn't installed on the system at all, just use + # nm to get what info we can (function names, but not line numbers). + if (system(ShellEscape($addr2line, "--help") . " >$dev_null 2>&1") != 0) { + MapSymbolsWithNM($image, $offset, $pclist, $symbols); + return; + } + + # "addr2line -i" can produce a variable number of lines per input + # address, with no separator that allows us to tell when data for + # the next address starts. So we find the address for a special + # symbol (_fini) and interleave this address between all real + # addresses passed to addr2line. The name of this special symbol + # can then be used as a separator. + $sep_address = undef; # May be filled in by MapSymbolsWithNM() + my $nm_symbols = {}; + MapSymbolsWithNM($image, $offset, $pclist, $nm_symbols); + if (defined($sep_address)) { + # Only add " -i" to addr2line if the binary supports it. + # addr2line --help returns 0, but not if it sees an unknown flag first. + if (system("$cmd -i --help >$dev_null 2>&1") == 0) { + $cmd .= " -i"; + } else { + $sep_address = undef; # no need for sep_address if we don't support -i + } + } + + # Make file with all PC values with intervening 'sep_address' so + # that we can reliably detect the end of inlined function list + open(ADDRESSES, ">$main::tmpfile_sym") || error("$main::tmpfile_sym: $!\n"); + if ($debug) { print("---- $image ---\n"); } + for (my $i = 0; $i <= $#{$pclist}; $i++) { + # addr2line always reads hex addresses, and does not need '0x' prefix. + if ($debug) { printf STDERR ("%s\n", $pclist->[$i]); } + printf ADDRESSES ("%s\n", AddressSub($pclist->[$i], $offset)); + if (defined($sep_address)) { + printf ADDRESSES ("%s\n", $sep_address); + } + } + close(ADDRESSES); + if ($debug) { + print("----\n"); + system("cat", $main::tmpfile_sym); + print("----\n"); + system("$cmd < " . ShellEscape($main::tmpfile_sym)); + print("----\n"); + } + + open(SYMBOLS, "$cmd <" . ShellEscape($main::tmpfile_sym) . " |") + || error("$cmd: $!\n"); + my $count = 0; # Index in pclist + while (<SYMBOLS>) { + # Read fullfunction and filelineinfo from next pair of lines + s/\r?\n$//g; + my $fullfunction = $_; + $_ = <SYMBOLS>; + s/\r?\n$//g; + my $filelinenum = $_; + + if (defined($sep_address) && $fullfunction eq $sep_symbol) { + # Terminating marker for data for this address + $count++; + next; + } + + $filelinenum =~ s|\\|/|g; # turn windows-style paths into unix-style paths + + my $pcstr = $pclist->[$count]; + my $function = ShortFunctionName($fullfunction); + my $nms = $nm_symbols->{$pcstr}; + if (defined($nms)) { + if ($fullfunction eq '??') { + # nm found a symbol for us. + $function = $nms->[0]; + $fullfunction = $nms->[2]; + } else { + # MapSymbolsWithNM tags each routine with its starting address, + # useful in case the image has multiple occurrences of this + # routine. (It uses a syntax that resembles template paramters, + # that are automatically stripped out by ShortFunctionName().) + # addr2line does not provide the same information. So we check + # if nm disambiguated our symbol, and if so take the annotated + # (nm) version of the routine-name. TODO(csilvers): this won't + # catch overloaded, inlined symbols, which nm doesn't see. + # Better would be to do a check similar to nm's, in this fn. + if ($nms->[2] =~ m/^\Q$function\E/) { # sanity check it's the right fn + $function = $nms->[0]; + $fullfunction = $nms->[2]; + } + } + } + + # Prepend to accumulated symbols for pcstr + # (so that caller comes before callee) + my $sym = $symbols->{$pcstr}; + if (!defined($sym)) { + $sym = []; + $symbols->{$pcstr} = $sym; + } + unshift(@{$sym}, $function, $filelinenum, $fullfunction); + if ($debug) { printf STDERR ("%s => [%s]\n", $pcstr, join(" ", @{$sym})); } + if (!defined($sep_address)) { + # Inlining is off, so this entry ends immediately + $count++; + } + } + close(SYMBOLS); +} + +# Use nm to map the list of referenced PCs to symbols. Return true iff we +# are able to read procedure information via nm. +sub MapSymbolsWithNM { + my $image = shift; + my $offset = shift; + my $pclist = shift; + my $symbols = shift; + + # Get nm output sorted by increasing address + my $symbol_table = GetProcedureBoundaries($image, "."); + if (!%{$symbol_table}) { + return 0; + } + # Start addresses are already the right length (8 or 16 hex digits). + my @names = sort { $symbol_table->{$a}->[0] cmp $symbol_table->{$b}->[0] } + keys(%{$symbol_table}); + + if ($#names < 0) { + # No symbols: just use addresses + foreach my $pc (@{$pclist}) { + my $pcstr = "0x" . $pc; + $symbols->{$pc} = [$pcstr, "?", $pcstr]; + } + return 0; + } + + # Sort addresses so we can do a join against nm output + my $index = 0; + my $fullname = $names[0]; + my $name = ShortFunctionName($fullname); + foreach my $pc (sort { $a cmp $b } @{$pclist}) { + # Adjust for mapped offset + my $mpc = AddressSub($pc, $offset); + while (($index < $#names) && ($mpc ge $symbol_table->{$fullname}->[1])){ + $index++; + $fullname = $names[$index]; + $name = ShortFunctionName($fullname); + } + if ($mpc lt $symbol_table->{$fullname}->[1]) { + $symbols->{$pc} = [$name, "?", $fullname]; + } else { + my $pcstr = "0x" . $pc; + $symbols->{$pc} = [$pcstr, "?", $pcstr]; + } + } + return 1; +} + +sub ShortFunctionName { + my $function = shift; + while ($function =~ s/\([^()]*\)(\s*const)?//g) { } # Argument types + while ($function =~ s/<[^<>]*>//g) { } # Remove template arguments + $function =~ s/^.*\s+(\w+::)/$1/; # Remove leading type + return $function; +} + +# Trim overly long symbols found in disassembler output +sub CleanDisassembly { + my $d = shift; + while ($d =~ s/\([^()%]*\)(\s*const)?//g) { } # Argument types, not (%rax) + while ($d =~ s/(\w+)<[^<>]*>/$1/g) { } # Remove template arguments + return $d; +} + +# Clean file name for display +sub CleanFileName { + my ($f) = @_; + $f =~ s|^/proc/self/cwd/||; + $f =~ s|^\./||; + return $f; +} + +# Make address relative to section and clean up for display +sub UnparseAddress { + my ($offset, $address) = @_; + $address = AddressSub($address, $offset); + $address =~ s/^0x//; + $address =~ s/^0*//; + return $address; +} + +##### Miscellaneous ##### + +# Find the right versions of the above object tools to use. The +# argument is the program file being analyzed, and should be an ELF +# 32-bit or ELF 64-bit executable file. The location of the tools +# is determined by considering the following options in this order: +# 1) --tools option, if set +# 2) JEPROF_TOOLS environment variable, if set +# 3) the environment +sub ConfigureObjTools { + my $prog_file = shift; + + # Check for the existence of $prog_file because /usr/bin/file does not + # predictably return error status in prod. + (-e $prog_file) || error("$prog_file does not exist.\n"); + + my $file_type = undef; + if (-e "/usr/bin/file") { + # Follow symlinks (at least for systems where "file" supports that). + my $escaped_prog_file = ShellEscape($prog_file); + $file_type = `/usr/bin/file -L $escaped_prog_file 2>$dev_null || + /usr/bin/file $escaped_prog_file`; + } elsif ($^O == "MSWin32") { + $file_type = "MS Windows"; + } else { + print STDERR "WARNING: Can't determine the file type of $prog_file"; + } + + if ($file_type =~ /64-bit/) { + # Change $address_length to 16 if the program file is ELF 64-bit. + # We can't detect this from many (most?) heap or lock contention + # profiles, since the actual addresses referenced are generally in low + # memory even for 64-bit programs. + $address_length = 16; + } + + if ($file_type =~ /MS Windows/) { + # For windows, we provide a version of nm and addr2line as part of + # the opensource release, which is capable of parsing + # Windows-style PDB executables. It should live in the path, or + # in the same directory as jeprof. + $obj_tool_map{"nm_pdb"} = "nm-pdb"; + $obj_tool_map{"addr2line_pdb"} = "addr2line-pdb"; + } + + if ($file_type =~ /Mach-O/) { + # OS X uses otool to examine Mach-O files, rather than objdump. + $obj_tool_map{"otool"} = "otool"; + $obj_tool_map{"addr2line"} = "false"; # no addr2line + $obj_tool_map{"objdump"} = "false"; # no objdump + } + + # Go fill in %obj_tool_map with the pathnames to use: + foreach my $tool (keys %obj_tool_map) { + $obj_tool_map{$tool} = ConfigureTool($obj_tool_map{$tool}); + } +} + +# Returns the path of a caller-specified object tool. If --tools or +# JEPROF_TOOLS are specified, then returns the full path to the tool +# with that prefix. Otherwise, returns the path unmodified (which +# means we will look for it on PATH). +sub ConfigureTool { + my $tool = shift; + my $path; + + # --tools (or $JEPROF_TOOLS) is a comma separated list, where each + # item is either a) a pathname prefix, or b) a map of the form + # <tool>:<path>. First we look for an entry of type (b) for our + # tool. If one is found, we use it. Otherwise, we consider all the + # pathname prefixes in turn, until one yields an existing file. If + # none does, we use a default path. + my $tools = $main::opt_tools || $ENV{"JEPROF_TOOLS"} || ""; + if ($tools =~ m/(,|^)\Q$tool\E:([^,]*)/) { + $path = $2; + # TODO(csilvers): sanity-check that $path exists? Hard if it's relative. + } elsif ($tools ne '') { + foreach my $prefix (split(',', $tools)) { + next if ($prefix =~ /:/); # ignore "tool:fullpath" entries in the list + if (-x $prefix . $tool) { + $path = $prefix . $tool; + last; + } + } + if (!$path) { + error("No '$tool' found with prefix specified by " . + "--tools (or \$JEPROF_TOOLS) '$tools'\n"); + } + } else { + # ... otherwise use the version that exists in the same directory as + # jeprof. If there's nothing there, use $PATH. + $0 =~ m,[^/]*$,; # this is everything after the last slash + my $dirname = $`; # this is everything up to and including the last slash + if (-x "$dirname$tool") { + $path = "$dirname$tool"; + } else { + $path = $tool; + } + } + if ($main::opt_debug) { print STDERR "Using '$path' for '$tool'.\n"; } + return $path; +} + +sub ShellEscape { + my @escaped_words = (); + foreach my $word (@_) { + my $escaped_word = $word; + if ($word =~ m![^a-zA-Z0-9/.,_=-]!) { # check for anything not in whitelist + $escaped_word =~ s/'/'\\''/; + $escaped_word = "'$escaped_word'"; + } + push(@escaped_words, $escaped_word); + } + return join(" ", @escaped_words); +} + +sub cleanup { + unlink($main::tmpfile_sym); + unlink(keys %main::tempnames); + + # We leave any collected profiles in $HOME/jeprof in case the user wants + # to look at them later. We print a message informing them of this. + if ((scalar(@main::profile_files) > 0) && + defined($main::collected_profile)) { + if (scalar(@main::profile_files) == 1) { + print STDERR "Dynamically gathered profile is in $main::collected_profile\n"; + } + print STDERR "If you want to investigate this profile further, you can do:\n"; + print STDERR "\n"; + print STDERR " jeprof \\\n"; + print STDERR " $main::prog \\\n"; + print STDERR " $main::collected_profile\n"; + print STDERR "\n"; + } +} + +sub sighandler { + cleanup(); + exit(1); +} + +sub error { + my $msg = shift; + print STDERR $msg; + cleanup(); + exit(1); +} + + +# Run $nm_command and get all the resulting procedure boundaries whose +# names match "$regexp" and returns them in a hashtable mapping from +# procedure name to a two-element vector of [start address, end address] +sub GetProcedureBoundariesViaNm { + my $escaped_nm_command = shift; # shell-escaped + my $regexp = shift; + + my $symbol_table = {}; + open(NM, "$escaped_nm_command |") || error("$escaped_nm_command: $!\n"); + my $last_start = "0"; + my $routine = ""; + while (<NM>) { + s/\r//g; # turn windows-looking lines into unix-looking lines + if (m/^\s*([0-9a-f]+) (.) (..*)/) { + my $start_val = $1; + my $type = $2; + my $this_routine = $3; + + # It's possible for two symbols to share the same address, if + # one is a zero-length variable (like __start_google_malloc) or + # one symbol is a weak alias to another (like __libc_malloc). + # In such cases, we want to ignore all values except for the + # actual symbol, which in nm-speak has type "T". The logic + # below does this, though it's a bit tricky: what happens when + # we have a series of lines with the same address, is the first + # one gets queued up to be processed. However, it won't + # *actually* be processed until later, when we read a line with + # a different address. That means that as long as we're reading + # lines with the same address, we have a chance to replace that + # item in the queue, which we do whenever we see a 'T' entry -- + # that is, a line with type 'T'. If we never see a 'T' entry, + # we'll just go ahead and process the first entry (which never + # got touched in the queue), and ignore the others. + if ($start_val eq $last_start && $type =~ /t/i) { + # We are the 'T' symbol at this address, replace previous symbol. + $routine = $this_routine; + next; + } elsif ($start_val eq $last_start) { + # We're not the 'T' symbol at this address, so ignore us. + next; + } + + if ($this_routine eq $sep_symbol) { + $sep_address = HexExtend($start_val); + } + + # Tag this routine with the starting address in case the image + # has multiple occurrences of this routine. We use a syntax + # that resembles template parameters that are automatically + # stripped out by ShortFunctionName() + $this_routine .= "<$start_val>"; + + if (defined($routine) && $routine =~ m/$regexp/) { + $symbol_table->{$routine} = [HexExtend($last_start), + HexExtend($start_val)]; + } + $last_start = $start_val; + $routine = $this_routine; + } elsif (m/^Loaded image name: (.+)/) { + # The win32 nm workalike emits information about the binary it is using. + if ($main::opt_debug) { print STDERR "Using Image $1\n"; } + } elsif (m/^PDB file name: (.+)/) { + # The win32 nm workalike emits information about the pdb it is using. + if ($main::opt_debug) { print STDERR "Using PDB $1\n"; } + } + } + close(NM); + # Handle the last line in the nm output. Unfortunately, we don't know + # how big this last symbol is, because we don't know how big the file + # is. For now, we just give it a size of 0. + # TODO(csilvers): do better here. + if (defined($routine) && $routine =~ m/$regexp/) { + $symbol_table->{$routine} = [HexExtend($last_start), + HexExtend($last_start)]; + } + return $symbol_table; +} + +# Gets the procedure boundaries for all routines in "$image" whose names +# match "$regexp" and returns them in a hashtable mapping from procedure +# name to a two-element vector of [start address, end address]. +# Will return an empty map if nm is not installed or not working properly. +sub GetProcedureBoundaries { + my $image = shift; + my $regexp = shift; + + # If $image doesn't start with /, then put ./ in front of it. This works + # around an obnoxious bug in our probing of nm -f behavior. + # "nm -f $image" is supposed to fail on GNU nm, but if: + # + # a. $image starts with [BbSsPp] (for example, bin/foo/bar), AND + # b. you have a.out in your current directory (a not uncommon occurence) + # + # then "nm -f $image" succeeds because -f only looks at the first letter of + # the argument, which looks valid because it's [BbSsPp], and then since + # there's no image provided, it looks for a.out and finds it. + # + # This regex makes sure that $image starts with . or /, forcing the -f + # parsing to fail since . and / are not valid formats. + $image =~ s#^[^/]#./$&#; + + # For libc libraries, the copy in /usr/lib/debug contains debugging symbols + my $debugging = DebuggingLibrary($image); + if ($debugging) { + $image = $debugging; + } + + my $nm = $obj_tool_map{"nm"}; + my $cppfilt = $obj_tool_map{"c++filt"}; + + # nm can fail for two reasons: 1) $image isn't a debug library; 2) nm + # binary doesn't support --demangle. In addition, for OS X we need + # to use the -f flag to get 'flat' nm output (otherwise we don't sort + # properly and get incorrect results). Unfortunately, GNU nm uses -f + # in an incompatible way. So first we test whether our nm supports + # --demangle and -f. + my $demangle_flag = ""; + my $cppfilt_flag = ""; + my $to_devnull = ">$dev_null 2>&1"; + if (system(ShellEscape($nm, "--demangle", $image) . $to_devnull) == 0) { + # In this mode, we do "nm --demangle <foo>" + $demangle_flag = "--demangle"; + $cppfilt_flag = ""; + } elsif (system(ShellEscape($cppfilt, $image) . $to_devnull) == 0) { + # In this mode, we do "nm <foo> | c++filt" + $cppfilt_flag = " | " . ShellEscape($cppfilt); + }; + my $flatten_flag = ""; + if (system(ShellEscape($nm, "-f", $image) . $to_devnull) == 0) { + $flatten_flag = "-f"; + } + + # Finally, in the case $imagie isn't a debug library, we try again with + # -D to at least get *exported* symbols. If we can't use --demangle, + # we use c++filt instead, if it exists on this system. + my @nm_commands = (ShellEscape($nm, "-n", $flatten_flag, $demangle_flag, + $image) . " 2>$dev_null $cppfilt_flag", + ShellEscape($nm, "-D", "-n", $flatten_flag, $demangle_flag, + $image) . " 2>$dev_null $cppfilt_flag", + # 6nm is for Go binaries + ShellEscape("6nm", "$image") . " 2>$dev_null | sort", + ); + + # If the executable is an MS Windows PDB-format executable, we'll + # have set up obj_tool_map("nm_pdb"). In this case, we actually + # want to use both unix nm and windows-specific nm_pdb, since + # PDB-format executables can apparently include dwarf .o files. + if (exists $obj_tool_map{"nm_pdb"}) { + push(@nm_commands, + ShellEscape($obj_tool_map{"nm_pdb"}, "--demangle", $image) + . " 2>$dev_null"); + } + + foreach my $nm_command (@nm_commands) { + my $symbol_table = GetProcedureBoundariesViaNm($nm_command, $regexp); + return $symbol_table if (%{$symbol_table}); + } + my $symbol_table = {}; + return $symbol_table; +} + + +# The test vectors for AddressAdd/Sub/Inc are 8-16-nibble hex strings. +# To make them more readable, we add underscores at interesting places. +# This routine removes the underscores, producing the canonical representation +# used by jeprof to represent addresses, particularly in the tested routines. +sub CanonicalHex { + my $arg = shift; + return join '', (split '_',$arg); +} + + +# Unit test for AddressAdd: +sub AddressAddUnitTest { + my $test_data_8 = shift; + my $test_data_16 = shift; + my $error_count = 0; + my $fail_count = 0; + my $pass_count = 0; + # print STDERR "AddressAddUnitTest: ", 1+$#{$test_data_8}, " tests\n"; + + # First a few 8-nibble addresses. Note that this implementation uses + # plain old arithmetic, so a quick sanity check along with verifying what + # happens to overflow (we want it to wrap): + $address_length = 8; + foreach my $row (@{$test_data_8}) { + if ($main::opt_debug and $main::opt_test) { print STDERR "@{$row}\n"; } + my $sum = AddressAdd ($row->[0], $row->[1]); + if ($sum ne $row->[2]) { + printf STDERR "ERROR: %s != %s + %s = %s\n", $sum, + $row->[0], $row->[1], $row->[2]; + ++$fail_count; + } else { + ++$pass_count; + } + } + printf STDERR "AddressAdd 32-bit tests: %d passes, %d failures\n", + $pass_count, $fail_count; + $error_count = $fail_count; + $fail_count = 0; + $pass_count = 0; + + # Now 16-nibble addresses. + $address_length = 16; + foreach my $row (@{$test_data_16}) { + if ($main::opt_debug and $main::opt_test) { print STDERR "@{$row}\n"; } + my $sum = AddressAdd (CanonicalHex($row->[0]), CanonicalHex($row->[1])); + my $expected = join '', (split '_',$row->[2]); + if ($sum ne CanonicalHex($row->[2])) { + printf STDERR "ERROR: %s != %s + %s = %s\n", $sum, + $row->[0], $row->[1], $row->[2]; + ++$fail_count; + } else { + ++$pass_count; + } + } + printf STDERR "AddressAdd 64-bit tests: %d passes, %d failures\n", + $pass_count, $fail_count; + $error_count += $fail_count; + + return $error_count; +} + + +# Unit test for AddressSub: +sub AddressSubUnitTest { + my $test_data_8 = shift; + my $test_data_16 = shift; + my $error_count = 0; + my $fail_count = 0; + my $pass_count = 0; + # print STDERR "AddressSubUnitTest: ", 1+$#{$test_data_8}, " tests\n"; + + # First a few 8-nibble addresses. Note that this implementation uses + # plain old arithmetic, so a quick sanity check along with verifying what + # happens to overflow (we want it to wrap): + $address_length = 8; + foreach my $row (@{$test_data_8}) { + if ($main::opt_debug and $main::opt_test) { print STDERR "@{$row}\n"; } + my $sum = AddressSub ($row->[0], $row->[1]); + if ($sum ne $row->[3]) { + printf STDERR "ERROR: %s != %s - %s = %s\n", $sum, + $row->[0], $row->[1], $row->[3]; + ++$fail_count; + } else { + ++$pass_count; + } + } + printf STDERR "AddressSub 32-bit tests: %d passes, %d failures\n", + $pass_count, $fail_count; + $error_count = $fail_count; + $fail_count = 0; + $pass_count = 0; + + # Now 16-nibble addresses. + $address_length = 16; + foreach my $row (@{$test_data_16}) { + if ($main::opt_debug and $main::opt_test) { print STDERR "@{$row}\n"; } + my $sum = AddressSub (CanonicalHex($row->[0]), CanonicalHex($row->[1])); + if ($sum ne CanonicalHex($row->[3])) { + printf STDERR "ERROR: %s != %s - %s = %s\n", $sum, + $row->[0], $row->[1], $row->[3]; + ++$fail_count; + } else { + ++$pass_count; + } + } + printf STDERR "AddressSub 64-bit tests: %d passes, %d failures\n", + $pass_count, $fail_count; + $error_count += $fail_count; + + return $error_count; +} + + +# Unit test for AddressInc: +sub AddressIncUnitTest { + my $test_data_8 = shift; + my $test_data_16 = shift; + my $error_count = 0; + my $fail_count = 0; + my $pass_count = 0; + # print STDERR "AddressIncUnitTest: ", 1+$#{$test_data_8}, " tests\n"; + + # First a few 8-nibble addresses. Note that this implementation uses + # plain old arithmetic, so a quick sanity check along with verifying what + # happens to overflow (we want it to wrap): + $address_length = 8; + foreach my $row (@{$test_data_8}) { + if ($main::opt_debug and $main::opt_test) { print STDERR "@{$row}\n"; } + my $sum = AddressInc ($row->[0]); + if ($sum ne $row->[4]) { + printf STDERR "ERROR: %s != %s + 1 = %s\n", $sum, + $row->[0], $row->[4]; + ++$fail_count; + } else { + ++$pass_count; + } + } + printf STDERR "AddressInc 32-bit tests: %d passes, %d failures\n", + $pass_count, $fail_count; + $error_count = $fail_count; + $fail_count = 0; + $pass_count = 0; + + # Now 16-nibble addresses. + $address_length = 16; + foreach my $row (@{$test_data_16}) { + if ($main::opt_debug and $main::opt_test) { print STDERR "@{$row}\n"; } + my $sum = AddressInc (CanonicalHex($row->[0])); + if ($sum ne CanonicalHex($row->[4])) { + printf STDERR "ERROR: %s != %s + 1 = %s\n", $sum, + $row->[0], $row->[4]; + ++$fail_count; + } else { + ++$pass_count; + } + } + printf STDERR "AddressInc 64-bit tests: %d passes, %d failures\n", + $pass_count, $fail_count; + $error_count += $fail_count; + + return $error_count; +} + + +# Driver for unit tests. +# Currently just the address add/subtract/increment routines for 64-bit. +sub RunUnitTests { + my $error_count = 0; + + # This is a list of tuples [a, b, a+b, a-b, a+1] + my $unit_test_data_8 = [ + [qw(aaaaaaaa 50505050 fafafafa 5a5a5a5a aaaaaaab)], + [qw(50505050 aaaaaaaa fafafafa a5a5a5a6 50505051)], + [qw(ffffffff aaaaaaaa aaaaaaa9 55555555 00000000)], + [qw(00000001 ffffffff 00000000 00000002 00000002)], + [qw(00000001 fffffff0 fffffff1 00000011 00000002)], + ]; + my $unit_test_data_16 = [ + # The implementation handles data in 7-nibble chunks, so those are the + # interesting boundaries. + [qw(aaaaaaaa 50505050 + 00_000000f_afafafa 00_0000005_a5a5a5a 00_000000a_aaaaaab)], + [qw(50505050 aaaaaaaa + 00_000000f_afafafa ff_ffffffa_5a5a5a6 00_0000005_0505051)], + [qw(ffffffff aaaaaaaa + 00_000001a_aaaaaa9 00_0000005_5555555 00_0000010_0000000)], + [qw(00000001 ffffffff + 00_0000010_0000000 ff_ffffff0_0000002 00_0000000_0000002)], + [qw(00000001 fffffff0 + 00_000000f_ffffff1 ff_ffffff0_0000011 00_0000000_0000002)], + + [qw(00_a00000a_aaaaaaa 50505050 + 00_a00000f_afafafa 00_a000005_a5a5a5a 00_a00000a_aaaaaab)], + [qw(0f_fff0005_0505050 aaaaaaaa + 0f_fff000f_afafafa 0f_ffefffa_5a5a5a6 0f_fff0005_0505051)], + [qw(00_000000f_fffffff 01_800000a_aaaaaaa + 01_800001a_aaaaaa9 fe_8000005_5555555 00_0000010_0000000)], + [qw(00_0000000_0000001 ff_fffffff_fffffff + 00_0000000_0000000 00_0000000_0000002 00_0000000_0000002)], + [qw(00_0000000_0000001 ff_fffffff_ffffff0 + ff_fffffff_ffffff1 00_0000000_0000011 00_0000000_0000002)], + ]; + + $error_count += AddressAddUnitTest($unit_test_data_8, $unit_test_data_16); + $error_count += AddressSubUnitTest($unit_test_data_8, $unit_test_data_16); + $error_count += AddressIncUnitTest($unit_test_data_8, $unit_test_data_16); + if ($error_count > 0) { + print STDERR $error_count, " errors: FAILED\n"; + } else { + print STDERR "PASS\n"; + } + exit ($error_count); +} diff --git a/deps/jemalloc/build-aux/config.guess b/deps/jemalloc/build-aux/config.guess new file mode 100755 index 000000000..2e9ad7fe8 --- /dev/null +++ b/deps/jemalloc/build-aux/config.guess @@ -0,0 +1,1462 @@ +#! /bin/sh +# Attempt to guess a canonical system name. +# Copyright 1992-2016 Free Software Foundation, Inc. + +timestamp='2016-10-02' + +# This file is free software; you can redistribute it and/or modify it +# under the terms of the GNU General Public License as published by +# the Free Software Foundation; either version 3 of the License, or +# (at your option) any later version. +# +# This program is distributed in the hope that it will be useful, but +# WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# General Public License for more details. +# +# You should have received a copy of the GNU General Public License +# along with this program; if not, see <http://www.gnu.org/licenses/>. +# +# As a special exception to the GNU General Public License, if you +# distribute this file as part of a program that contains a +# configuration script generated by Autoconf, you may include it under +# the same distribution terms that you use for the rest of that +# program. This Exception is an additional permission under section 7 +# of the GNU General Public License, version 3 ("GPLv3"). +# +# Originally written by Per Bothner; maintained since 2000 by Ben Elliston. +# +# You can get the latest version of this script from: +# http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.guess +# +# Please send patches to <config-patches@gnu.org>. + + +me=`echo "$0" | sed -e 's,.*/,,'` + +usage="\ +Usage: $0 [OPTION] + +Output the configuration name of the system \`$me' is run on. + +Operation modes: + -h, --help print this help, then exit + -t, --time-stamp print date of last modification, then exit + -v, --version print version number, then exit + +Report bugs and patches to <config-patches@gnu.org>." + +version="\ +GNU config.guess ($timestamp) + +Originally written by Per Bothner. +Copyright 1992-2016 Free Software Foundation, Inc. + +This is free software; see the source for copying conditions. There is NO +warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE." + +help=" +Try \`$me --help' for more information." + +# Parse command line +while test $# -gt 0 ; do + case $1 in + --time-stamp | --time* | -t ) + echo "$timestamp" ; exit ;; + --version | -v ) + echo "$version" ; exit ;; + --help | --h* | -h ) + echo "$usage"; exit ;; + -- ) # Stop option processing + shift; break ;; + - ) # Use stdin as input. + break ;; + -* ) + echo "$me: invalid option $1$help" >&2 + exit 1 ;; + * ) + break ;; + esac +done + +if test $# != 0; then + echo "$me: too many arguments$help" >&2 + exit 1 +fi + +trap 'exit 1' 1 2 15 + +# CC_FOR_BUILD -- compiler used by this script. Note that the use of a +# compiler to aid in system detection is discouraged as it requires +# temporary files to be created and, as you can see below, it is a +# headache to deal with in a portable fashion. + +# Historically, `CC_FOR_BUILD' used to be named `HOST_CC'. We still +# use `HOST_CC' if defined, but it is deprecated. + +# Portable tmp directory creation inspired by the Autoconf team. + +set_cc_for_build=' +trap "exitcode=\$?; (rm -f \$tmpfiles 2>/dev/null; rmdir \$tmp 2>/dev/null) && exit \$exitcode" 0 ; +trap "rm -f \$tmpfiles 2>/dev/null; rmdir \$tmp 2>/dev/null; exit 1" 1 2 13 15 ; +: ${TMPDIR=/tmp} ; + { tmp=`(umask 077 && mktemp -d "$TMPDIR/cgXXXXXX") 2>/dev/null` && test -n "$tmp" && test -d "$tmp" ; } || + { test -n "$RANDOM" && tmp=$TMPDIR/cg$$-$RANDOM && (umask 077 && mkdir $tmp) ; } || + { tmp=$TMPDIR/cg-$$ && (umask 077 && mkdir $tmp) && echo "Warning: creating insecure temp directory" >&2 ; } || + { echo "$me: cannot create a temporary directory in $TMPDIR" >&2 ; exit 1 ; } ; +dummy=$tmp/dummy ; +tmpfiles="$dummy.c $dummy.o $dummy.rel $dummy" ; +case $CC_FOR_BUILD,$HOST_CC,$CC in + ,,) echo "int x;" > $dummy.c ; + for c in cc gcc c89 c99 ; do + if ($c -c -o $dummy.o $dummy.c) >/dev/null 2>&1 ; then + CC_FOR_BUILD="$c"; break ; + fi ; + done ; + if test x"$CC_FOR_BUILD" = x ; then + CC_FOR_BUILD=no_compiler_found ; + fi + ;; + ,,*) CC_FOR_BUILD=$CC ;; + ,*,*) CC_FOR_BUILD=$HOST_CC ;; +esac ; set_cc_for_build= ;' + +# This is needed to find uname on a Pyramid OSx when run in the BSD universe. +# (ghazi@noc.rutgers.edu 1994-08-24) +if (test -f /.attbin/uname) >/dev/null 2>&1 ; then + PATH=$PATH:/.attbin ; export PATH +fi + +UNAME_MACHINE=`(uname -m) 2>/dev/null` || UNAME_MACHINE=unknown +UNAME_RELEASE=`(uname -r) 2>/dev/null` || UNAME_RELEASE=unknown +UNAME_SYSTEM=`(uname -s) 2>/dev/null` || UNAME_SYSTEM=unknown +UNAME_VERSION=`(uname -v) 2>/dev/null` || UNAME_VERSION=unknown + +case "${UNAME_SYSTEM}" in +Linux|GNU|GNU/*) + # If the system lacks a compiler, then just pick glibc. + # We could probably try harder. + LIBC=gnu + + eval $set_cc_for_build + cat <<-EOF > $dummy.c + #include <features.h> + #if defined(__UCLIBC__) + LIBC=uclibc + #elif defined(__dietlibc__) + LIBC=dietlibc + #else + LIBC=gnu + #endif + EOF + eval `$CC_FOR_BUILD -E $dummy.c 2>/dev/null | grep '^LIBC' | sed 's, ,,g'` + ;; +esac + +# Note: order is significant - the case branches are not exclusive. + +case "${UNAME_MACHINE}:${UNAME_SYSTEM}:${UNAME_RELEASE}:${UNAME_VERSION}" in + *:NetBSD:*:*) + # NetBSD (nbsd) targets should (where applicable) match one or + # more of the tuples: *-*-netbsdelf*, *-*-netbsdaout*, + # *-*-netbsdecoff* and *-*-netbsd*. For targets that recently + # switched to ELF, *-*-netbsd* would select the old + # object file format. This provides both forward + # compatibility and a consistent mechanism for selecting the + # object file format. + # + # Note: NetBSD doesn't particularly care about the vendor + # portion of the name. We always set it to "unknown". + sysctl="sysctl -n hw.machine_arch" + UNAME_MACHINE_ARCH=`(uname -p 2>/dev/null || \ + /sbin/$sysctl 2>/dev/null || \ + /usr/sbin/$sysctl 2>/dev/null || \ + echo unknown)` + case "${UNAME_MACHINE_ARCH}" in + armeb) machine=armeb-unknown ;; + arm*) machine=arm-unknown ;; + sh3el) machine=shl-unknown ;; + sh3eb) machine=sh-unknown ;; + sh5el) machine=sh5le-unknown ;; + earmv*) + arch=`echo ${UNAME_MACHINE_ARCH} | sed -e 's,^e\(armv[0-9]\).*$,\1,'` + endian=`echo ${UNAME_MACHINE_ARCH} | sed -ne 's,^.*\(eb\)$,\1,p'` + machine=${arch}${endian}-unknown + ;; + *) machine=${UNAME_MACHINE_ARCH}-unknown ;; + esac + # The Operating System including object format, if it has switched + # to ELF recently (or will in the future) and ABI. + case "${UNAME_MACHINE_ARCH}" in + earm*) + os=netbsdelf + ;; + arm*|i386|m68k|ns32k|sh3*|sparc|vax) + eval $set_cc_for_build + if echo __ELF__ | $CC_FOR_BUILD -E - 2>/dev/null \ + | grep -q __ELF__ + then + # Once all utilities can be ECOFF (netbsdecoff) or a.out (netbsdaout). + # Return netbsd for either. FIX? + os=netbsd + else + os=netbsdelf + fi + ;; + *) + os=netbsd + ;; + esac + # Determine ABI tags. + case "${UNAME_MACHINE_ARCH}" in + earm*) + expr='s/^earmv[0-9]/-eabi/;s/eb$//' + abi=`echo ${UNAME_MACHINE_ARCH} | sed -e "$expr"` + ;; + esac + # The OS release + # Debian GNU/NetBSD machines have a different userland, and + # thus, need a distinct triplet. However, they do not need + # kernel version information, so it can be replaced with a + # suitable tag, in the style of linux-gnu. + case "${UNAME_VERSION}" in + Debian*) + release='-gnu' + ;; + *) + release=`echo ${UNAME_RELEASE} | sed -e 's/[-_].*//' | cut -d. -f1,2` + ;; + esac + # Since CPU_TYPE-MANUFACTURER-KERNEL-OPERATING_SYSTEM: + # contains redundant information, the shorter form: + # CPU_TYPE-MANUFACTURER-OPERATING_SYSTEM is used. + echo "${machine}-${os}${release}${abi}" + exit ;; + *:Bitrig:*:*) + UNAME_MACHINE_ARCH=`arch | sed 's/Bitrig.//'` + echo ${UNAME_MACHINE_ARCH}-unknown-bitrig${UNAME_RELEASE} + exit ;; + *:OpenBSD:*:*) + UNAME_MACHINE_ARCH=`arch | sed 's/OpenBSD.//'` + echo ${UNAME_MACHINE_ARCH}-unknown-openbsd${UNAME_RELEASE} + exit ;; + *:LibertyBSD:*:*) + UNAME_MACHINE_ARCH=`arch | sed 's/^.*BSD\.//'` + echo ${UNAME_MACHINE_ARCH}-unknown-libertybsd${UNAME_RELEASE} + exit ;; + *:ekkoBSD:*:*) + echo ${UNAME_MACHINE}-unknown-ekkobsd${UNAME_RELEASE} + exit ;; + *:SolidBSD:*:*) + echo ${UNAME_MACHINE}-unknown-solidbsd${UNAME_RELEASE} + exit ;; + macppc:MirBSD:*:*) + echo powerpc-unknown-mirbsd${UNAME_RELEASE} + exit ;; + *:MirBSD:*:*) + echo ${UNAME_MACHINE}-unknown-mirbsd${UNAME_RELEASE} + exit ;; + *:Sortix:*:*) + echo ${UNAME_MACHINE}-unknown-sortix + exit ;; + alpha:OSF1:*:*) + case $UNAME_RELEASE in + *4.0) + UNAME_RELEASE=`/usr/sbin/sizer -v | awk '{print $3}'` + ;; + *5.*) + UNAME_RELEASE=`/usr/sbin/sizer -v | awk '{print $4}'` + ;; + esac + # According to Compaq, /usr/sbin/psrinfo has been available on + # OSF/1 and Tru64 systems produced since 1995. I hope that + # covers most systems running today. This code pipes the CPU + # types through head -n 1, so we only detect the type of CPU 0. + ALPHA_CPU_TYPE=`/usr/sbin/psrinfo -v | sed -n -e 's/^ The alpha \(.*\) processor.*$/\1/p' | head -n 1` + case "$ALPHA_CPU_TYPE" in + "EV4 (21064)") + UNAME_MACHINE=alpha ;; + "EV4.5 (21064)") + UNAME_MACHINE=alpha ;; + "LCA4 (21066/21068)") + UNAME_MACHINE=alpha ;; + "EV5 (21164)") + UNAME_MACHINE=alphaev5 ;; + "EV5.6 (21164A)") + UNAME_MACHINE=alphaev56 ;; + "EV5.6 (21164PC)") + UNAME_MACHINE=alphapca56 ;; + "EV5.7 (21164PC)") + UNAME_MACHINE=alphapca57 ;; + "EV6 (21264)") + UNAME_MACHINE=alphaev6 ;; + "EV6.7 (21264A)") + UNAME_MACHINE=alphaev67 ;; + "EV6.8CB (21264C)") + UNAME_MACHINE=alphaev68 ;; + "EV6.8AL (21264B)") + UNAME_MACHINE=alphaev68 ;; + "EV6.8CX (21264D)") + UNAME_MACHINE=alphaev68 ;; + "EV6.9A (21264/EV69A)") + UNAME_MACHINE=alphaev69 ;; + "EV7 (21364)") + UNAME_MACHINE=alphaev7 ;; + "EV7.9 (21364A)") + UNAME_MACHINE=alphaev79 ;; + esac + # A Pn.n version is a patched version. + # A Vn.n version is a released version. + # A Tn.n version is a released field test version. + # A Xn.n version is an unreleased experimental baselevel. + # 1.2 uses "1.2" for uname -r. + echo ${UNAME_MACHINE}-dec-osf`echo ${UNAME_RELEASE} | sed -e 's/^[PVTX]//' | tr ABCDEFGHIJKLMNOPQRSTUVWXYZ abcdefghijklmnopqrstuvwxyz` + # Reset EXIT trap before exiting to avoid spurious non-zero exit code. + exitcode=$? + trap '' 0 + exit $exitcode ;; + Alpha\ *:Windows_NT*:*) + # How do we know it's Interix rather than the generic POSIX subsystem? + # Should we change UNAME_MACHINE based on the output of uname instead + # of the specific Alpha model? + echo alpha-pc-interix + exit ;; + 21064:Windows_NT:50:3) + echo alpha-dec-winnt3.5 + exit ;; + Amiga*:UNIX_System_V:4.0:*) + echo m68k-unknown-sysv4 + exit ;; + *:[Aa]miga[Oo][Ss]:*:*) + echo ${UNAME_MACHINE}-unknown-amigaos + exit ;; + *:[Mm]orph[Oo][Ss]:*:*) + echo ${UNAME_MACHINE}-unknown-morphos + exit ;; + *:OS/390:*:*) + echo i370-ibm-openedition + exit ;; + *:z/VM:*:*) + echo s390-ibm-zvmoe + exit ;; + *:OS400:*:*) + echo powerpc-ibm-os400 + exit ;; + arm:RISC*:1.[012]*:*|arm:riscix:1.[012]*:*) + echo arm-acorn-riscix${UNAME_RELEASE} + exit ;; + arm*:riscos:*:*|arm*:RISCOS:*:*) + echo arm-unknown-riscos + exit ;; + SR2?01:HI-UX/MPP:*:* | SR8000:HI-UX/MPP:*:*) + echo hppa1.1-hitachi-hiuxmpp + exit ;; + Pyramid*:OSx*:*:* | MIS*:OSx*:*:* | MIS*:SMP_DC-OSx*:*:*) + # akee@wpdis03.wpafb.af.mil (Earle F. Ake) contributed MIS and NILE. + if test "`(/bin/universe) 2>/dev/null`" = att ; then + echo pyramid-pyramid-sysv3 + else + echo pyramid-pyramid-bsd + fi + exit ;; + NILE*:*:*:dcosx) + echo pyramid-pyramid-svr4 + exit ;; + DRS?6000:unix:4.0:6*) + echo sparc-icl-nx6 + exit ;; + DRS?6000:UNIX_SV:4.2*:7* | DRS?6000:isis:4.2*:7*) + case `/usr/bin/uname -p` in + sparc) echo sparc-icl-nx7; exit ;; + esac ;; + s390x:SunOS:*:*) + echo ${UNAME_MACHINE}-ibm-solaris2`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'` + exit ;; + sun4H:SunOS:5.*:*) + echo sparc-hal-solaris2`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'` + exit ;; + sun4*:SunOS:5.*:* | tadpole*:SunOS:5.*:*) + echo sparc-sun-solaris2`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'` + exit ;; + i86pc:AuroraUX:5.*:* | i86xen:AuroraUX:5.*:*) + echo i386-pc-auroraux${UNAME_RELEASE} + exit ;; + i86pc:SunOS:5.*:* | i86xen:SunOS:5.*:*) + eval $set_cc_for_build + SUN_ARCH=i386 + # If there is a compiler, see if it is configured for 64-bit objects. + # Note that the Sun cc does not turn __LP64__ into 1 like gcc does. + # This test works for both compilers. + if [ "$CC_FOR_BUILD" != no_compiler_found ]; then + if (echo '#ifdef __amd64'; echo IS_64BIT_ARCH; echo '#endif') | \ + (CCOPTS="" $CC_FOR_BUILD -E - 2>/dev/null) | \ + grep IS_64BIT_ARCH >/dev/null + then + SUN_ARCH=x86_64 + fi + fi + echo ${SUN_ARCH}-pc-solaris2`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'` + exit ;; + sun4*:SunOS:6*:*) + # According to config.sub, this is the proper way to canonicalize + # SunOS6. Hard to guess exactly what SunOS6 will be like, but + # it's likely to be more like Solaris than SunOS4. + echo sparc-sun-solaris3`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'` + exit ;; + sun4*:SunOS:*:*) + case "`/usr/bin/arch -k`" in + Series*|S4*) + UNAME_RELEASE=`uname -v` + ;; + esac + # Japanese Language versions have a version number like `4.1.3-JL'. + echo sparc-sun-sunos`echo ${UNAME_RELEASE}|sed -e 's/-/_/'` + exit ;; + sun3*:SunOS:*:*) + echo m68k-sun-sunos${UNAME_RELEASE} + exit ;; + sun*:*:4.2BSD:*) + UNAME_RELEASE=`(sed 1q /etc/motd | awk '{print substr($5,1,3)}') 2>/dev/null` + test "x${UNAME_RELEASE}" = x && UNAME_RELEASE=3 + case "`/bin/arch`" in + sun3) + echo m68k-sun-sunos${UNAME_RELEASE} + ;; + sun4) + echo sparc-sun-sunos${UNAME_RELEASE} + ;; + esac + exit ;; + aushp:SunOS:*:*) + echo sparc-auspex-sunos${UNAME_RELEASE} + exit ;; + # The situation for MiNT is a little confusing. The machine name + # can be virtually everything (everything which is not + # "atarist" or "atariste" at least should have a processor + # > m68000). The system name ranges from "MiNT" over "FreeMiNT" + # to the lowercase version "mint" (or "freemint"). Finally + # the system name "TOS" denotes a system which is actually not + # MiNT. But MiNT is downward compatible to TOS, so this should + # be no problem. + atarist[e]:*MiNT:*:* | atarist[e]:*mint:*:* | atarist[e]:*TOS:*:*) + echo m68k-atari-mint${UNAME_RELEASE} + exit ;; + atari*:*MiNT:*:* | atari*:*mint:*:* | atarist[e]:*TOS:*:*) + echo m68k-atari-mint${UNAME_RELEASE} + exit ;; + *falcon*:*MiNT:*:* | *falcon*:*mint:*:* | *falcon*:*TOS:*:*) + echo m68k-atari-mint${UNAME_RELEASE} + exit ;; + milan*:*MiNT:*:* | milan*:*mint:*:* | *milan*:*TOS:*:*) + echo m68k-milan-mint${UNAME_RELEASE} + exit ;; + hades*:*MiNT:*:* | hades*:*mint:*:* | *hades*:*TOS:*:*) + echo m68k-hades-mint${UNAME_RELEASE} + exit ;; + *:*MiNT:*:* | *:*mint:*:* | *:*TOS:*:*) + echo m68k-unknown-mint${UNAME_RELEASE} + exit ;; + m68k:machten:*:*) + echo m68k-apple-machten${UNAME_RELEASE} + exit ;; + powerpc:machten:*:*) + echo powerpc-apple-machten${UNAME_RELEASE} + exit ;; + RISC*:Mach:*:*) + echo mips-dec-mach_bsd4.3 + exit ;; + RISC*:ULTRIX:*:*) + echo mips-dec-ultrix${UNAME_RELEASE} + exit ;; + VAX*:ULTRIX*:*:*) + echo vax-dec-ultrix${UNAME_RELEASE} + exit ;; + 2020:CLIX:*:* | 2430:CLIX:*:*) + echo clipper-intergraph-clix${UNAME_RELEASE} + exit ;; + mips:*:*:UMIPS | mips:*:*:RISCos) + eval $set_cc_for_build + sed 's/^ //' << EOF >$dummy.c +#ifdef __cplusplus +#include <stdio.h> /* for printf() prototype */ + int main (int argc, char *argv[]) { +#else + int main (argc, argv) int argc; char *argv[]; { +#endif + #if defined (host_mips) && defined (MIPSEB) + #if defined (SYSTYPE_SYSV) + printf ("mips-mips-riscos%ssysv\n", argv[1]); exit (0); + #endif + #if defined (SYSTYPE_SVR4) + printf ("mips-mips-riscos%ssvr4\n", argv[1]); exit (0); + #endif + #if defined (SYSTYPE_BSD43) || defined(SYSTYPE_BSD) + printf ("mips-mips-riscos%sbsd\n", argv[1]); exit (0); + #endif + #endif + exit (-1); + } +EOF + $CC_FOR_BUILD -o $dummy $dummy.c && + dummyarg=`echo "${UNAME_RELEASE}" | sed -n 's/\([0-9]*\).*/\1/p'` && + SYSTEM_NAME=`$dummy $dummyarg` && + { echo "$SYSTEM_NAME"; exit; } + echo mips-mips-riscos${UNAME_RELEASE} + exit ;; + Motorola:PowerMAX_OS:*:*) + echo powerpc-motorola-powermax + exit ;; + Motorola:*:4.3:PL8-*) + echo powerpc-harris-powermax + exit ;; + Night_Hawk:*:*:PowerMAX_OS | Synergy:PowerMAX_OS:*:*) + echo powerpc-harris-powermax + exit ;; + Night_Hawk:Power_UNIX:*:*) + echo powerpc-harris-powerunix + exit ;; + m88k:CX/UX:7*:*) + echo m88k-harris-cxux7 + exit ;; + m88k:*:4*:R4*) + echo m88k-motorola-sysv4 + exit ;; + m88k:*:3*:R3*) + echo m88k-motorola-sysv3 + exit ;; + AViiON:dgux:*:*) + # DG/UX returns AViiON for all architectures + UNAME_PROCESSOR=`/usr/bin/uname -p` + if [ $UNAME_PROCESSOR = mc88100 ] || [ $UNAME_PROCESSOR = mc88110 ] + then + if [ ${TARGET_BINARY_INTERFACE}x = m88kdguxelfx ] || \ + [ ${TARGET_BINARY_INTERFACE}x = x ] + then + echo m88k-dg-dgux${UNAME_RELEASE} + else + echo m88k-dg-dguxbcs${UNAME_RELEASE} + fi + else + echo i586-dg-dgux${UNAME_RELEASE} + fi + exit ;; + M88*:DolphinOS:*:*) # DolphinOS (SVR3) + echo m88k-dolphin-sysv3 + exit ;; + M88*:*:R3*:*) + # Delta 88k system running SVR3 + echo m88k-motorola-sysv3 + exit ;; + XD88*:*:*:*) # Tektronix XD88 system running UTekV (SVR3) + echo m88k-tektronix-sysv3 + exit ;; + Tek43[0-9][0-9]:UTek:*:*) # Tektronix 4300 system running UTek (BSD) + echo m68k-tektronix-bsd + exit ;; + *:IRIX*:*:*) + echo mips-sgi-irix`echo ${UNAME_RELEASE}|sed -e 's/-/_/g'` + exit ;; + ????????:AIX?:[12].1:2) # AIX 2.2.1 or AIX 2.1.1 is RT/PC AIX. + echo romp-ibm-aix # uname -m gives an 8 hex-code CPU id + exit ;; # Note that: echo "'`uname -s`'" gives 'AIX ' + i*86:AIX:*:*) + echo i386-ibm-aix + exit ;; + ia64:AIX:*:*) + if [ -x /usr/bin/oslevel ] ; then + IBM_REV=`/usr/bin/oslevel` + else + IBM_REV=${UNAME_VERSION}.${UNAME_RELEASE} + fi + echo ${UNAME_MACHINE}-ibm-aix${IBM_REV} + exit ;; + *:AIX:2:3) + if grep bos325 /usr/include/stdio.h >/dev/null 2>&1; then + eval $set_cc_for_build + sed 's/^ //' << EOF >$dummy.c + #include <sys/systemcfg.h> + + main() + { + if (!__power_pc()) + exit(1); + puts("powerpc-ibm-aix3.2.5"); + exit(0); + } +EOF + if $CC_FOR_BUILD -o $dummy $dummy.c && SYSTEM_NAME=`$dummy` + then + echo "$SYSTEM_NAME" + else + echo rs6000-ibm-aix3.2.5 + fi + elif grep bos324 /usr/include/stdio.h >/dev/null 2>&1; then + echo rs6000-ibm-aix3.2.4 + else + echo rs6000-ibm-aix3.2 + fi + exit ;; + *:AIX:*:[4567]) + IBM_CPU_ID=`/usr/sbin/lsdev -C -c processor -S available | sed 1q | awk '{ print $1 }'` + if /usr/sbin/lsattr -El ${IBM_CPU_ID} | grep ' POWER' >/dev/null 2>&1; then + IBM_ARCH=rs6000 + else + IBM_ARCH=powerpc + fi + if [ -x /usr/bin/lslpp ] ; then + IBM_REV=`/usr/bin/lslpp -Lqc bos.rte.libc | + awk -F: '{ print $3 }' | sed s/[0-9]*$/0/` + else + IBM_REV=${UNAME_VERSION}.${UNAME_RELEASE} + fi + echo ${IBM_ARCH}-ibm-aix${IBM_REV} + exit ;; + *:AIX:*:*) + echo rs6000-ibm-aix + exit ;; + ibmrt:4.4BSD:*|romp-ibm:BSD:*) + echo romp-ibm-bsd4.4 + exit ;; + ibmrt:*BSD:*|romp-ibm:BSD:*) # covers RT/PC BSD and + echo romp-ibm-bsd${UNAME_RELEASE} # 4.3 with uname added to + exit ;; # report: romp-ibm BSD 4.3 + *:BOSX:*:*) + echo rs6000-bull-bosx + exit ;; + DPX/2?00:B.O.S.:*:*) + echo m68k-bull-sysv3 + exit ;; + 9000/[34]??:4.3bsd:1.*:*) + echo m68k-hp-bsd + exit ;; + hp300:4.4BSD:*:* | 9000/[34]??:4.3bsd:2.*:*) + echo m68k-hp-bsd4.4 + exit ;; + 9000/[34678]??:HP-UX:*:*) + HPUX_REV=`echo ${UNAME_RELEASE}|sed -e 's/[^.]*.[0B]*//'` + case "${UNAME_MACHINE}" in + 9000/31? ) HP_ARCH=m68000 ;; + 9000/[34]?? ) HP_ARCH=m68k ;; + 9000/[678][0-9][0-9]) + if [ -x /usr/bin/getconf ]; then + sc_cpu_version=`/usr/bin/getconf SC_CPU_VERSION 2>/dev/null` + sc_kernel_bits=`/usr/bin/getconf SC_KERNEL_BITS 2>/dev/null` + case "${sc_cpu_version}" in + 523) HP_ARCH=hppa1.0 ;; # CPU_PA_RISC1_0 + 528) HP_ARCH=hppa1.1 ;; # CPU_PA_RISC1_1 + 532) # CPU_PA_RISC2_0 + case "${sc_kernel_bits}" in + 32) HP_ARCH=hppa2.0n ;; + 64) HP_ARCH=hppa2.0w ;; + '') HP_ARCH=hppa2.0 ;; # HP-UX 10.20 + esac ;; + esac + fi + if [ "${HP_ARCH}" = "" ]; then + eval $set_cc_for_build + sed 's/^ //' << EOF >$dummy.c + + #define _HPUX_SOURCE + #include <stdlib.h> + #include <unistd.h> + + int main () + { + #if defined(_SC_KERNEL_BITS) + long bits = sysconf(_SC_KERNEL_BITS); + #endif + long cpu = sysconf (_SC_CPU_VERSION); + + switch (cpu) + { + case CPU_PA_RISC1_0: puts ("hppa1.0"); break; + case CPU_PA_RISC1_1: puts ("hppa1.1"); break; + case CPU_PA_RISC2_0: + #if defined(_SC_KERNEL_BITS) + switch (bits) + { + case 64: puts ("hppa2.0w"); break; + case 32: puts ("hppa2.0n"); break; + default: puts ("hppa2.0"); break; + } break; + #else /* !defined(_SC_KERNEL_BITS) */ + puts ("hppa2.0"); break; + #endif + default: puts ("hppa1.0"); break; + } + exit (0); + } +EOF + (CCOPTS="" $CC_FOR_BUILD -o $dummy $dummy.c 2>/dev/null) && HP_ARCH=`$dummy` + test -z "$HP_ARCH" && HP_ARCH=hppa + fi ;; + esac + if [ ${HP_ARCH} = hppa2.0w ] + then + eval $set_cc_for_build + + # hppa2.0w-hp-hpux* has a 64-bit kernel and a compiler generating + # 32-bit code. hppa64-hp-hpux* has the same kernel and a compiler + # generating 64-bit code. GNU and HP use different nomenclature: + # + # $ CC_FOR_BUILD=cc ./config.guess + # => hppa2.0w-hp-hpux11.23 + # $ CC_FOR_BUILD="cc +DA2.0w" ./config.guess + # => hppa64-hp-hpux11.23 + + if echo __LP64__ | (CCOPTS="" $CC_FOR_BUILD -E - 2>/dev/null) | + grep -q __LP64__ + then + HP_ARCH=hppa2.0w + else + HP_ARCH=hppa64 + fi + fi + echo ${HP_ARCH}-hp-hpux${HPUX_REV} + exit ;; + ia64:HP-UX:*:*) + HPUX_REV=`echo ${UNAME_RELEASE}|sed -e 's/[^.]*.[0B]*//'` + echo ia64-hp-hpux${HPUX_REV} + exit ;; + 3050*:HI-UX:*:*) + eval $set_cc_for_build + sed 's/^ //' << EOF >$dummy.c + #include <unistd.h> + int + main () + { + long cpu = sysconf (_SC_CPU_VERSION); + /* The order matters, because CPU_IS_HP_MC68K erroneously returns + true for CPU_PA_RISC1_0. CPU_IS_PA_RISC returns correct + results, however. */ + if (CPU_IS_PA_RISC (cpu)) + { + switch (cpu) + { + case CPU_PA_RISC1_0: puts ("hppa1.0-hitachi-hiuxwe2"); break; + case CPU_PA_RISC1_1: puts ("hppa1.1-hitachi-hiuxwe2"); break; + case CPU_PA_RISC2_0: puts ("hppa2.0-hitachi-hiuxwe2"); break; + default: puts ("hppa-hitachi-hiuxwe2"); break; + } + } + else if (CPU_IS_HP_MC68K (cpu)) + puts ("m68k-hitachi-hiuxwe2"); + else puts ("unknown-hitachi-hiuxwe2"); + exit (0); + } +EOF + $CC_FOR_BUILD -o $dummy $dummy.c && SYSTEM_NAME=`$dummy` && + { echo "$SYSTEM_NAME"; exit; } + echo unknown-hitachi-hiuxwe2 + exit ;; + 9000/7??:4.3bsd:*:* | 9000/8?[79]:4.3bsd:*:* ) + echo hppa1.1-hp-bsd + exit ;; + 9000/8??:4.3bsd:*:*) + echo hppa1.0-hp-bsd + exit ;; + *9??*:MPE/iX:*:* | *3000*:MPE/iX:*:*) + echo hppa1.0-hp-mpeix + exit ;; + hp7??:OSF1:*:* | hp8?[79]:OSF1:*:* ) + echo hppa1.1-hp-osf + exit ;; + hp8??:OSF1:*:*) + echo hppa1.0-hp-osf + exit ;; + i*86:OSF1:*:*) + if [ -x /usr/sbin/sysversion ] ; then + echo ${UNAME_MACHINE}-unknown-osf1mk + else + echo ${UNAME_MACHINE}-unknown-osf1 + fi + exit ;; + parisc*:Lites*:*:*) + echo hppa1.1-hp-lites + exit ;; + C1*:ConvexOS:*:* | convex:ConvexOS:C1*:*) + echo c1-convex-bsd + exit ;; + C2*:ConvexOS:*:* | convex:ConvexOS:C2*:*) + if getsysinfo -f scalar_acc + then echo c32-convex-bsd + else echo c2-convex-bsd + fi + exit ;; + C34*:ConvexOS:*:* | convex:ConvexOS:C34*:*) + echo c34-convex-bsd + exit ;; + C38*:ConvexOS:*:* | convex:ConvexOS:C38*:*) + echo c38-convex-bsd + exit ;; + C4*:ConvexOS:*:* | convex:ConvexOS:C4*:*) + echo c4-convex-bsd + exit ;; + CRAY*Y-MP:*:*:*) + echo ymp-cray-unicos${UNAME_RELEASE} | sed -e 's/\.[^.]*$/.X/' + exit ;; + CRAY*[A-Z]90:*:*:*) + echo ${UNAME_MACHINE}-cray-unicos${UNAME_RELEASE} \ + | sed -e 's/CRAY.*\([A-Z]90\)/\1/' \ + -e y/ABCDEFGHIJKLMNOPQRSTUVWXYZ/abcdefghijklmnopqrstuvwxyz/ \ + -e 's/\.[^.]*$/.X/' + exit ;; + CRAY*TS:*:*:*) + echo t90-cray-unicos${UNAME_RELEASE} | sed -e 's/\.[^.]*$/.X/' + exit ;; + CRAY*T3E:*:*:*) + echo alphaev5-cray-unicosmk${UNAME_RELEASE} | sed -e 's/\.[^.]*$/.X/' + exit ;; + CRAY*SV1:*:*:*) + echo sv1-cray-unicos${UNAME_RELEASE} | sed -e 's/\.[^.]*$/.X/' + exit ;; + *:UNICOS/mp:*:*) + echo craynv-cray-unicosmp${UNAME_RELEASE} | sed -e 's/\.[^.]*$/.X/' + exit ;; + F30[01]:UNIX_System_V:*:* | F700:UNIX_System_V:*:*) + FUJITSU_PROC=`uname -m | tr ABCDEFGHIJKLMNOPQRSTUVWXYZ abcdefghijklmnopqrstuvwxyz` + FUJITSU_SYS=`uname -p | tr ABCDEFGHIJKLMNOPQRSTUVWXYZ abcdefghijklmnopqrstuvwxyz | sed -e 's/\///'` + FUJITSU_REL=`echo ${UNAME_RELEASE} | sed -e 's/ /_/'` + echo "${FUJITSU_PROC}-fujitsu-${FUJITSU_SYS}${FUJITSU_REL}" + exit ;; + 5000:UNIX_System_V:4.*:*) + FUJITSU_SYS=`uname -p | tr ABCDEFGHIJKLMNOPQRSTUVWXYZ abcdefghijklmnopqrstuvwxyz | sed -e 's/\///'` + FUJITSU_REL=`echo ${UNAME_RELEASE} | tr ABCDEFGHIJKLMNOPQRSTUVWXYZ abcdefghijklmnopqrstuvwxyz | sed -e 's/ /_/'` + echo "sparc-fujitsu-${FUJITSU_SYS}${FUJITSU_REL}" + exit ;; + i*86:BSD/386:*:* | i*86:BSD/OS:*:* | *:Ascend\ Embedded/OS:*:*) + echo ${UNAME_MACHINE}-pc-bsdi${UNAME_RELEASE} + exit ;; + sparc*:BSD/OS:*:*) + echo sparc-unknown-bsdi${UNAME_RELEASE} + exit ;; + *:BSD/OS:*:*) + echo ${UNAME_MACHINE}-unknown-bsdi${UNAME_RELEASE} + exit ;; + *:FreeBSD:*:*) + UNAME_PROCESSOR=`/usr/bin/uname -p` + case ${UNAME_PROCESSOR} in + amd64) + echo x86_64-unknown-freebsd`echo ${UNAME_RELEASE}|sed -e 's/[-(].*//'` ;; + *) + echo ${UNAME_PROCESSOR}-unknown-freebsd`echo ${UNAME_RELEASE}|sed -e 's/[-(].*//'` ;; + esac + exit ;; + i*:CYGWIN*:*) + echo ${UNAME_MACHINE}-pc-cygwin + exit ;; + *:MINGW64*:*) + echo ${UNAME_MACHINE}-pc-mingw64 + exit ;; + *:MINGW*:*) + echo ${UNAME_MACHINE}-pc-mingw32 + exit ;; + *:MSYS*:*) + echo ${UNAME_MACHINE}-pc-msys + exit ;; + i*:windows32*:*) + # uname -m includes "-pc" on this system. + echo ${UNAME_MACHINE}-mingw32 + exit ;; + i*:PW*:*) + echo ${UNAME_MACHINE}-pc-pw32 + exit ;; + *:Interix*:*) + case ${UNAME_MACHINE} in + x86) + echo i586-pc-interix${UNAME_RELEASE} + exit ;; + authenticamd | genuineintel | EM64T) + echo x86_64-unknown-interix${UNAME_RELEASE} + exit ;; + IA64) + echo ia64-unknown-interix${UNAME_RELEASE} + exit ;; + esac ;; + [345]86:Windows_95:* | [345]86:Windows_98:* | [345]86:Windows_NT:*) + echo i${UNAME_MACHINE}-pc-mks + exit ;; + 8664:Windows_NT:*) + echo x86_64-pc-mks + exit ;; + i*:Windows_NT*:* | Pentium*:Windows_NT*:*) + # How do we know it's Interix rather than the generic POSIX subsystem? + # It also conflicts with pre-2.0 versions of AT&T UWIN. Should we + # UNAME_MACHINE based on the output of uname instead of i386? + echo i586-pc-interix + exit ;; + i*:UWIN*:*) + echo ${UNAME_MACHINE}-pc-uwin + exit ;; + amd64:CYGWIN*:*:* | x86_64:CYGWIN*:*:*) + echo x86_64-unknown-cygwin + exit ;; + p*:CYGWIN*:*) + echo powerpcle-unknown-cygwin + exit ;; + prep*:SunOS:5.*:*) + echo powerpcle-unknown-solaris2`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'` + exit ;; + *:GNU:*:*) + # the GNU system + echo `echo ${UNAME_MACHINE}|sed -e 's,[-/].*$,,'`-unknown-${LIBC}`echo ${UNAME_RELEASE}|sed -e 's,/.*$,,'` + exit ;; + *:GNU/*:*:*) + # other systems with GNU libc and userland + echo ${UNAME_MACHINE}-unknown-`echo ${UNAME_SYSTEM} | sed 's,^[^/]*/,,' | tr "[:upper:]" "[:lower:]"``echo ${UNAME_RELEASE}|sed -e 's/[-(].*//'`-${LIBC} + exit ;; + i*86:Minix:*:*) + echo ${UNAME_MACHINE}-pc-minix + exit ;; + aarch64:Linux:*:*) + echo ${UNAME_MACHINE}-unknown-linux-${LIBC} + exit ;; + aarch64_be:Linux:*:*) + UNAME_MACHINE=aarch64_be + echo ${UNAME_MACHINE}-unknown-linux-${LIBC} + exit ;; + alpha:Linux:*:*) + case `sed -n '/^cpu model/s/^.*: \(.*\)/\1/p' < /proc/cpuinfo` in + EV5) UNAME_MACHINE=alphaev5 ;; + EV56) UNAME_MACHINE=alphaev56 ;; + PCA56) UNAME_MACHINE=alphapca56 ;; + PCA57) UNAME_MACHINE=alphapca56 ;; + EV6) UNAME_MACHINE=alphaev6 ;; + EV67) UNAME_MACHINE=alphaev67 ;; + EV68*) UNAME_MACHINE=alphaev68 ;; + esac + objdump --private-headers /bin/sh | grep -q ld.so.1 + if test "$?" = 0 ; then LIBC=gnulibc1 ; fi + echo ${UNAME_MACHINE}-unknown-linux-${LIBC} + exit ;; + arc:Linux:*:* | arceb:Linux:*:*) + echo ${UNAME_MACHINE}-unknown-linux-${LIBC} + exit ;; + arm*:Linux:*:*) + eval $set_cc_for_build + if echo __ARM_EABI__ | $CC_FOR_BUILD -E - 2>/dev/null \ + | grep -q __ARM_EABI__ + then + echo ${UNAME_MACHINE}-unknown-linux-${LIBC} + else + if echo __ARM_PCS_VFP | $CC_FOR_BUILD -E - 2>/dev/null \ + | grep -q __ARM_PCS_VFP + then + echo ${UNAME_MACHINE}-unknown-linux-${LIBC}eabi + else + echo ${UNAME_MACHINE}-unknown-linux-${LIBC}eabihf + fi + fi + exit ;; + avr32*:Linux:*:*) + echo ${UNAME_MACHINE}-unknown-linux-${LIBC} + exit ;; + cris:Linux:*:*) + echo ${UNAME_MACHINE}-axis-linux-${LIBC} + exit ;; + crisv32:Linux:*:*) + echo ${UNAME_MACHINE}-axis-linux-${LIBC} + exit ;; + e2k:Linux:*:*) + echo ${UNAME_MACHINE}-unknown-linux-${LIBC} + exit ;; + frv:Linux:*:*) + echo ${UNAME_MACHINE}-unknown-linux-${LIBC} + exit ;; + hexagon:Linux:*:*) + echo ${UNAME_MACHINE}-unknown-linux-${LIBC} + exit ;; + i*86:Linux:*:*) + echo ${UNAME_MACHINE}-pc-linux-${LIBC} + exit ;; + ia64:Linux:*:*) + echo ${UNAME_MACHINE}-unknown-linux-${LIBC} + exit ;; + k1om:Linux:*:*) + echo ${UNAME_MACHINE}-unknown-linux-${LIBC} + exit ;; + m32r*:Linux:*:*) + echo ${UNAME_MACHINE}-unknown-linux-${LIBC} + exit ;; + m68*:Linux:*:*) + echo ${UNAME_MACHINE}-unknown-linux-${LIBC} + exit ;; + mips:Linux:*:* | mips64:Linux:*:*) + eval $set_cc_for_build + sed 's/^ //' << EOF >$dummy.c + #undef CPU + #undef ${UNAME_MACHINE} + #undef ${UNAME_MACHINE}el + #if defined(__MIPSEL__) || defined(__MIPSEL) || defined(_MIPSEL) || defined(MIPSEL) + CPU=${UNAME_MACHINE}el + #else + #if defined(__MIPSEB__) || defined(__MIPSEB) || defined(_MIPSEB) || defined(MIPSEB) + CPU=${UNAME_MACHINE} + #else + CPU= + #endif + #endif +EOF + eval `$CC_FOR_BUILD -E $dummy.c 2>/dev/null | grep '^CPU'` + test x"${CPU}" != x && { echo "${CPU}-unknown-linux-${LIBC}"; exit; } + ;; + mips64el:Linux:*:*) + echo ${UNAME_MACHINE}-unknown-linux-${LIBC} + exit ;; + openrisc*:Linux:*:*) + echo or1k-unknown-linux-${LIBC} + exit ;; + or32:Linux:*:* | or1k*:Linux:*:*) + echo ${UNAME_MACHINE}-unknown-linux-${LIBC} + exit ;; + padre:Linux:*:*) + echo sparc-unknown-linux-${LIBC} + exit ;; + parisc64:Linux:*:* | hppa64:Linux:*:*) + echo hppa64-unknown-linux-${LIBC} + exit ;; + parisc:Linux:*:* | hppa:Linux:*:*) + # Look for CPU level + case `grep '^cpu[^a-z]*:' /proc/cpuinfo 2>/dev/null | cut -d' ' -f2` in + PA7*) echo hppa1.1-unknown-linux-${LIBC} ;; + PA8*) echo hppa2.0-unknown-linux-${LIBC} ;; + *) echo hppa-unknown-linux-${LIBC} ;; + esac + exit ;; + ppc64:Linux:*:*) + echo powerpc64-unknown-linux-${LIBC} + exit ;; + ppc:Linux:*:*) + echo powerpc-unknown-linux-${LIBC} + exit ;; + ppc64le:Linux:*:*) + echo powerpc64le-unknown-linux-${LIBC} + exit ;; + ppcle:Linux:*:*) + echo powerpcle-unknown-linux-${LIBC} + exit ;; + riscv32:Linux:*:* | riscv64:Linux:*:*) + echo ${UNAME_MACHINE}-unknown-linux-${LIBC} + exit ;; + s390:Linux:*:* | s390x:Linux:*:*) + echo ${UNAME_MACHINE}-ibm-linux-${LIBC} + exit ;; + sh64*:Linux:*:*) + echo ${UNAME_MACHINE}-unknown-linux-${LIBC} + exit ;; + sh*:Linux:*:*) + echo ${UNAME_MACHINE}-unknown-linux-${LIBC} + exit ;; + sparc:Linux:*:* | sparc64:Linux:*:*) + echo ${UNAME_MACHINE}-unknown-linux-${LIBC} + exit ;; + tile*:Linux:*:*) + echo ${UNAME_MACHINE}-unknown-linux-${LIBC} + exit ;; + vax:Linux:*:*) + echo ${UNAME_MACHINE}-dec-linux-${LIBC} + exit ;; + x86_64:Linux:*:*) + echo ${UNAME_MACHINE}-pc-linux-${LIBC} + exit ;; + xtensa*:Linux:*:*) + echo ${UNAME_MACHINE}-unknown-linux-${LIBC} + exit ;; + i*86:DYNIX/ptx:4*:*) + # ptx 4.0 does uname -s correctly, with DYNIX/ptx in there. + # earlier versions are messed up and put the nodename in both + # sysname and nodename. + echo i386-sequent-sysv4 + exit ;; + i*86:UNIX_SV:4.2MP:2.*) + # Unixware is an offshoot of SVR4, but it has its own version + # number series starting with 2... + # I am not positive that other SVR4 systems won't match this, + # I just have to hope. -- rms. + # Use sysv4.2uw... so that sysv4* matches it. + echo ${UNAME_MACHINE}-pc-sysv4.2uw${UNAME_VERSION} + exit ;; + i*86:OS/2:*:*) + # If we were able to find `uname', then EMX Unix compatibility + # is probably installed. + echo ${UNAME_MACHINE}-pc-os2-emx + exit ;; + i*86:XTS-300:*:STOP) + echo ${UNAME_MACHINE}-unknown-stop + exit ;; + i*86:atheos:*:*) + echo ${UNAME_MACHINE}-unknown-atheos + exit ;; + i*86:syllable:*:*) + echo ${UNAME_MACHINE}-pc-syllable + exit ;; + i*86:LynxOS:2.*:* | i*86:LynxOS:3.[01]*:* | i*86:LynxOS:4.[02]*:*) + echo i386-unknown-lynxos${UNAME_RELEASE} + exit ;; + i*86:*DOS:*:*) + echo ${UNAME_MACHINE}-pc-msdosdjgpp + exit ;; + i*86:*:4.*:* | i*86:SYSTEM_V:4.*:*) + UNAME_REL=`echo ${UNAME_RELEASE} | sed 's/\/MP$//'` + if grep Novell /usr/include/link.h >/dev/null 2>/dev/null; then + echo ${UNAME_MACHINE}-univel-sysv${UNAME_REL} + else + echo ${UNAME_MACHINE}-pc-sysv${UNAME_REL} + fi + exit ;; + i*86:*:5:[678]*) + # UnixWare 7.x, OpenUNIX and OpenServer 6. + case `/bin/uname -X | grep "^Machine"` in + *486*) UNAME_MACHINE=i486 ;; + *Pentium) UNAME_MACHINE=i586 ;; + *Pent*|*Celeron) UNAME_MACHINE=i686 ;; + esac + echo ${UNAME_MACHINE}-unknown-sysv${UNAME_RELEASE}${UNAME_SYSTEM}${UNAME_VERSION} + exit ;; + i*86:*:3.2:*) + if test -f /usr/options/cb.name; then + UNAME_REL=`sed -n 's/.*Version //p' </usr/options/cb.name` + echo ${UNAME_MACHINE}-pc-isc$UNAME_REL + elif /bin/uname -X 2>/dev/null >/dev/null ; then + UNAME_REL=`(/bin/uname -X|grep Release|sed -e 's/.*= //')` + (/bin/uname -X|grep i80486 >/dev/null) && UNAME_MACHINE=i486 + (/bin/uname -X|grep '^Machine.*Pentium' >/dev/null) \ + && UNAME_MACHINE=i586 + (/bin/uname -X|grep '^Machine.*Pent *II' >/dev/null) \ + && UNAME_MACHINE=i686 + (/bin/uname -X|grep '^Machine.*Pentium Pro' >/dev/null) \ + && UNAME_MACHINE=i686 + echo ${UNAME_MACHINE}-pc-sco$UNAME_REL + else + echo ${UNAME_MACHINE}-pc-sysv32 + fi + exit ;; + pc:*:*:*) + # Left here for compatibility: + # uname -m prints for DJGPP always 'pc', but it prints nothing about + # the processor, so we play safe by assuming i586. + # Note: whatever this is, it MUST be the same as what config.sub + # prints for the "djgpp" host, or else GDB configure will decide that + # this is a cross-build. + echo i586-pc-msdosdjgpp + exit ;; + Intel:Mach:3*:*) + echo i386-pc-mach3 + exit ;; + paragon:*:*:*) + echo i860-intel-osf1 + exit ;; + i860:*:4.*:*) # i860-SVR4 + if grep Stardent /usr/include/sys/uadmin.h >/dev/null 2>&1 ; then + echo i860-stardent-sysv${UNAME_RELEASE} # Stardent Vistra i860-SVR4 + else # Add other i860-SVR4 vendors below as they are discovered. + echo i860-unknown-sysv${UNAME_RELEASE} # Unknown i860-SVR4 + fi + exit ;; + mini*:CTIX:SYS*5:*) + # "miniframe" + echo m68010-convergent-sysv + exit ;; + mc68k:UNIX:SYSTEM5:3.51m) + echo m68k-convergent-sysv + exit ;; + M680?0:D-NIX:5.3:*) + echo m68k-diab-dnix + exit ;; + M68*:*:R3V[5678]*:*) + test -r /sysV68 && { echo 'm68k-motorola-sysv'; exit; } ;; + 3[345]??:*:4.0:3.0 | 3[34]??A:*:4.0:3.0 | 3[34]??,*:*:4.0:3.0 | 3[34]??/*:*:4.0:3.0 | 4400:*:4.0:3.0 | 4850:*:4.0:3.0 | SKA40:*:4.0:3.0 | SDS2:*:4.0:3.0 | SHG2:*:4.0:3.0 | S7501*:*:4.0:3.0) + OS_REL='' + test -r /etc/.relid \ + && OS_REL=.`sed -n 's/[^ ]* [^ ]* \([0-9][0-9]\).*/\1/p' < /etc/.relid` + /bin/uname -p 2>/dev/null | grep 86 >/dev/null \ + && { echo i486-ncr-sysv4.3${OS_REL}; exit; } + /bin/uname -p 2>/dev/null | /bin/grep entium >/dev/null \ + && { echo i586-ncr-sysv4.3${OS_REL}; exit; } ;; + 3[34]??:*:4.0:* | 3[34]??,*:*:4.0:*) + /bin/uname -p 2>/dev/null | grep 86 >/dev/null \ + && { echo i486-ncr-sysv4; exit; } ;; + NCR*:*:4.2:* | MPRAS*:*:4.2:*) + OS_REL='.3' + test -r /etc/.relid \ + && OS_REL=.`sed -n 's/[^ ]* [^ ]* \([0-9][0-9]\).*/\1/p' < /etc/.relid` + /bin/uname -p 2>/dev/null | grep 86 >/dev/null \ + && { echo i486-ncr-sysv4.3${OS_REL}; exit; } + /bin/uname -p 2>/dev/null | /bin/grep entium >/dev/null \ + && { echo i586-ncr-sysv4.3${OS_REL}; exit; } + /bin/uname -p 2>/dev/null | /bin/grep pteron >/dev/null \ + && { echo i586-ncr-sysv4.3${OS_REL}; exit; } ;; + m68*:LynxOS:2.*:* | m68*:LynxOS:3.0*:*) + echo m68k-unknown-lynxos${UNAME_RELEASE} + exit ;; + mc68030:UNIX_System_V:4.*:*) + echo m68k-atari-sysv4 + exit ;; + TSUNAMI:LynxOS:2.*:*) + echo sparc-unknown-lynxos${UNAME_RELEASE} + exit ;; + rs6000:LynxOS:2.*:*) + echo rs6000-unknown-lynxos${UNAME_RELEASE} + exit ;; + PowerPC:LynxOS:2.*:* | PowerPC:LynxOS:3.[01]*:* | PowerPC:LynxOS:4.[02]*:*) + echo powerpc-unknown-lynxos${UNAME_RELEASE} + exit ;; + SM[BE]S:UNIX_SV:*:*) + echo mips-dde-sysv${UNAME_RELEASE} + exit ;; + RM*:ReliantUNIX-*:*:*) + echo mips-sni-sysv4 + exit ;; + RM*:SINIX-*:*:*) + echo mips-sni-sysv4 + exit ;; + *:SINIX-*:*:*) + if uname -p 2>/dev/null >/dev/null ; then + UNAME_MACHINE=`(uname -p) 2>/dev/null` + echo ${UNAME_MACHINE}-sni-sysv4 + else + echo ns32k-sni-sysv + fi + exit ;; + PENTIUM:*:4.0*:*) # Unisys `ClearPath HMP IX 4000' SVR4/MP effort + # says <Richard.M.Bartel@ccMail.Census.GOV> + echo i586-unisys-sysv4 + exit ;; + *:UNIX_System_V:4*:FTX*) + # From Gerald Hewes <hewes@openmarket.com>. + # How about differentiating between stratus architectures? -djm + echo hppa1.1-stratus-sysv4 + exit ;; + *:*:*:FTX*) + # From seanf@swdc.stratus.com. + echo i860-stratus-sysv4 + exit ;; + i*86:VOS:*:*) + # From Paul.Green@stratus.com. + echo ${UNAME_MACHINE}-stratus-vos + exit ;; + *:VOS:*:*) + # From Paul.Green@stratus.com. + echo hppa1.1-stratus-vos + exit ;; + mc68*:A/UX:*:*) + echo m68k-apple-aux${UNAME_RELEASE} + exit ;; + news*:NEWS-OS:6*:*) + echo mips-sony-newsos6 + exit ;; + R[34]000:*System_V*:*:* | R4000:UNIX_SYSV:*:* | R*000:UNIX_SV:*:*) + if [ -d /usr/nec ]; then + echo mips-nec-sysv${UNAME_RELEASE} + else + echo mips-unknown-sysv${UNAME_RELEASE} + fi + exit ;; + BeBox:BeOS:*:*) # BeOS running on hardware made by Be, PPC only. + echo powerpc-be-beos + exit ;; + BeMac:BeOS:*:*) # BeOS running on Mac or Mac clone, PPC only. + echo powerpc-apple-beos + exit ;; + BePC:BeOS:*:*) # BeOS running on Intel PC compatible. + echo i586-pc-beos + exit ;; + BePC:Haiku:*:*) # Haiku running on Intel PC compatible. + echo i586-pc-haiku + exit ;; + x86_64:Haiku:*:*) + echo x86_64-unknown-haiku + exit ;; + SX-4:SUPER-UX:*:*) + echo sx4-nec-superux${UNAME_RELEASE} + exit ;; + SX-5:SUPER-UX:*:*) + echo sx5-nec-superux${UNAME_RELEASE} + exit ;; + SX-6:SUPER-UX:*:*) + echo sx6-nec-superux${UNAME_RELEASE} + exit ;; + SX-7:SUPER-UX:*:*) + echo sx7-nec-superux${UNAME_RELEASE} + exit ;; + SX-8:SUPER-UX:*:*) + echo sx8-nec-superux${UNAME_RELEASE} + exit ;; + SX-8R:SUPER-UX:*:*) + echo sx8r-nec-superux${UNAME_RELEASE} + exit ;; + SX-ACE:SUPER-UX:*:*) + echo sxace-nec-superux${UNAME_RELEASE} + exit ;; + Power*:Rhapsody:*:*) + echo powerpc-apple-rhapsody${UNAME_RELEASE} + exit ;; + *:Rhapsody:*:*) + echo ${UNAME_MACHINE}-apple-rhapsody${UNAME_RELEASE} + exit ;; + *:Darwin:*:*) + UNAME_PROCESSOR=`uname -p` || UNAME_PROCESSOR=unknown + eval $set_cc_for_build + if test "$UNAME_PROCESSOR" = unknown ; then + UNAME_PROCESSOR=powerpc + fi + if test `echo "$UNAME_RELEASE" | sed -e 's/\..*//'` -le 10 ; then + if [ "$CC_FOR_BUILD" != no_compiler_found ]; then + if (echo '#ifdef __LP64__'; echo IS_64BIT_ARCH; echo '#endif') | \ + (CCOPTS="" $CC_FOR_BUILD -E - 2>/dev/null) | \ + grep IS_64BIT_ARCH >/dev/null + then + case $UNAME_PROCESSOR in + i386) UNAME_PROCESSOR=x86_64 ;; + powerpc) UNAME_PROCESSOR=powerpc64 ;; + esac + fi + fi + elif test "$UNAME_PROCESSOR" = i386 ; then + # Avoid executing cc on OS X 10.9, as it ships with a stub + # that puts up a graphical alert prompting to install + # developer tools. Any system running Mac OS X 10.7 or + # later (Darwin 11 and later) is required to have a 64-bit + # processor. This is not true of the ARM version of Darwin + # that Apple uses in portable devices. + UNAME_PROCESSOR=x86_64 + fi + echo ${UNAME_PROCESSOR}-apple-darwin${UNAME_RELEASE} + exit ;; + *:procnto*:*:* | *:QNX:[0123456789]*:*) + UNAME_PROCESSOR=`uname -p` + if test "$UNAME_PROCESSOR" = x86; then + UNAME_PROCESSOR=i386 + UNAME_MACHINE=pc + fi + echo ${UNAME_PROCESSOR}-${UNAME_MACHINE}-nto-qnx${UNAME_RELEASE} + exit ;; + *:QNX:*:4*) + echo i386-pc-qnx + exit ;; + NEO-?:NONSTOP_KERNEL:*:*) + echo neo-tandem-nsk${UNAME_RELEASE} + exit ;; + NSE-*:NONSTOP_KERNEL:*:*) + echo nse-tandem-nsk${UNAME_RELEASE} + exit ;; + NSR-?:NONSTOP_KERNEL:*:*) + echo nsr-tandem-nsk${UNAME_RELEASE} + exit ;; + *:NonStop-UX:*:*) + echo mips-compaq-nonstopux + exit ;; + BS2000:POSIX*:*:*) + echo bs2000-siemens-sysv + exit ;; + DS/*:UNIX_System_V:*:*) + echo ${UNAME_MACHINE}-${UNAME_SYSTEM}-${UNAME_RELEASE} + exit ;; + *:Plan9:*:*) + # "uname -m" is not consistent, so use $cputype instead. 386 + # is converted to i386 for consistency with other x86 + # operating systems. + if test "$cputype" = 386; then + UNAME_MACHINE=i386 + else + UNAME_MACHINE="$cputype" + fi + echo ${UNAME_MACHINE}-unknown-plan9 + exit ;; + *:TOPS-10:*:*) + echo pdp10-unknown-tops10 + exit ;; + *:TENEX:*:*) + echo pdp10-unknown-tenex + exit ;; + KS10:TOPS-20:*:* | KL10:TOPS-20:*:* | TYPE4:TOPS-20:*:*) + echo pdp10-dec-tops20 + exit ;; + XKL-1:TOPS-20:*:* | TYPE5:TOPS-20:*:*) + echo pdp10-xkl-tops20 + exit ;; + *:TOPS-20:*:*) + echo pdp10-unknown-tops20 + exit ;; + *:ITS:*:*) + echo pdp10-unknown-its + exit ;; + SEI:*:*:SEIUX) + echo mips-sei-seiux${UNAME_RELEASE} + exit ;; + *:DragonFly:*:*) + echo ${UNAME_MACHINE}-unknown-dragonfly`echo ${UNAME_RELEASE}|sed -e 's/[-(].*//'` + exit ;; + *:*VMS:*:*) + UNAME_MACHINE=`(uname -p) 2>/dev/null` + case "${UNAME_MACHINE}" in + A*) echo alpha-dec-vms ; exit ;; + I*) echo ia64-dec-vms ; exit ;; + V*) echo vax-dec-vms ; exit ;; + esac ;; + *:XENIX:*:SysV) + echo i386-pc-xenix + exit ;; + i*86:skyos:*:*) + echo ${UNAME_MACHINE}-pc-skyos`echo ${UNAME_RELEASE} | sed -e 's/ .*$//'` + exit ;; + i*86:rdos:*:*) + echo ${UNAME_MACHINE}-pc-rdos + exit ;; + i*86:AROS:*:*) + echo ${UNAME_MACHINE}-pc-aros + exit ;; + x86_64:VMkernel:*:*) + echo ${UNAME_MACHINE}-unknown-esx + exit ;; + amd64:Isilon\ OneFS:*:*) + echo x86_64-unknown-onefs + exit ;; +esac + +cat >&2 <<EOF +$0: unable to guess system type + +This script (version $timestamp), has failed to recognize the +operating system you are using. If your script is old, overwrite +config.guess and config.sub with the latest versions from: + + http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.guess +and + http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.sub + +If $0 has already been updated, send the following data and any +information you think might be pertinent to config-patches@gnu.org to +provide the necessary information to handle your system. + +config.guess timestamp = $timestamp + +uname -m = `(uname -m) 2>/dev/null || echo unknown` +uname -r = `(uname -r) 2>/dev/null || echo unknown` +uname -s = `(uname -s) 2>/dev/null || echo unknown` +uname -v = `(uname -v) 2>/dev/null || echo unknown` + +/usr/bin/uname -p = `(/usr/bin/uname -p) 2>/dev/null` +/bin/uname -X = `(/bin/uname -X) 2>/dev/null` + +hostinfo = `(hostinfo) 2>/dev/null` +/bin/universe = `(/bin/universe) 2>/dev/null` +/usr/bin/arch -k = `(/usr/bin/arch -k) 2>/dev/null` +/bin/arch = `(/bin/arch) 2>/dev/null` +/usr/bin/oslevel = `(/usr/bin/oslevel) 2>/dev/null` +/usr/convex/getsysinfo = `(/usr/convex/getsysinfo) 2>/dev/null` + +UNAME_MACHINE = ${UNAME_MACHINE} +UNAME_RELEASE = ${UNAME_RELEASE} +UNAME_SYSTEM = ${UNAME_SYSTEM} +UNAME_VERSION = ${UNAME_VERSION} +EOF + +exit 1 + +# Local variables: +# eval: (add-hook 'write-file-hooks 'time-stamp) +# time-stamp-start: "timestamp='" +# time-stamp-format: "%:y-%02m-%02d" +# time-stamp-end: "'" +# End: diff --git a/deps/jemalloc/build-aux/config.sub b/deps/jemalloc/build-aux/config.sub new file mode 100755 index 000000000..dd2ca93c6 --- /dev/null +++ b/deps/jemalloc/build-aux/config.sub @@ -0,0 +1,1825 @@ +#! /bin/sh +# Configuration validation subroutine script. +# Copyright 1992-2016 Free Software Foundation, Inc. + +timestamp='2016-11-04' + +# This file is free software; you can redistribute it and/or modify it +# under the terms of the GNU General Public License as published by +# the Free Software Foundation; either version 3 of the License, or +# (at your option) any later version. +# +# This program is distributed in the hope that it will be useful, but +# WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# General Public License for more details. +# +# You should have received a copy of the GNU General Public License +# along with this program; if not, see <http://www.gnu.org/licenses/>. +# +# As a special exception to the GNU General Public License, if you +# distribute this file as part of a program that contains a +# configuration script generated by Autoconf, you may include it under +# the same distribution terms that you use for the rest of that +# program. This Exception is an additional permission under section 7 +# of the GNU General Public License, version 3 ("GPLv3"). + + +# Please send patches to <config-patches@gnu.org>. +# +# Configuration subroutine to validate and canonicalize a configuration type. +# Supply the specified configuration type as an argument. +# If it is invalid, we print an error message on stderr and exit with code 1. +# Otherwise, we print the canonical config type on stdout and succeed. + +# You can get the latest version of this script from: +# http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.sub + +# This file is supposed to be the same for all GNU packages +# and recognize all the CPU types, system types and aliases +# that are meaningful with *any* GNU software. +# Each package is responsible for reporting which valid configurations +# it does not support. The user should be able to distinguish +# a failure to support a valid configuration from a meaningless +# configuration. + +# The goal of this file is to map all the various variations of a given +# machine specification into a single specification in the form: +# CPU_TYPE-MANUFACTURER-OPERATING_SYSTEM +# or in some cases, the newer four-part form: +# CPU_TYPE-MANUFACTURER-KERNEL-OPERATING_SYSTEM +# It is wrong to echo any other type of specification. + +me=`echo "$0" | sed -e 's,.*/,,'` + +usage="\ +Usage: $0 [OPTION] CPU-MFR-OPSYS or ALIAS + +Canonicalize a configuration name. + +Operation modes: + -h, --help print this help, then exit + -t, --time-stamp print date of last modification, then exit + -v, --version print version number, then exit + +Report bugs and patches to <config-patches@gnu.org>." + +version="\ +GNU config.sub ($timestamp) + +Copyright 1992-2016 Free Software Foundation, Inc. + +This is free software; see the source for copying conditions. There is NO +warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE." + +help=" +Try \`$me --help' for more information." + +# Parse command line +while test $# -gt 0 ; do + case $1 in + --time-stamp | --time* | -t ) + echo "$timestamp" ; exit ;; + --version | -v ) + echo "$version" ; exit ;; + --help | --h* | -h ) + echo "$usage"; exit ;; + -- ) # Stop option processing + shift; break ;; + - ) # Use stdin as input. + break ;; + -* ) + echo "$me: invalid option $1$help" + exit 1 ;; + + *local*) + # First pass through any local machine types. + echo $1 + exit ;; + + * ) + break ;; + esac +done + +case $# in + 0) echo "$me: missing argument$help" >&2 + exit 1;; + 1) ;; + *) echo "$me: too many arguments$help" >&2 + exit 1;; +esac + +# Separate what the user gave into CPU-COMPANY and OS or KERNEL-OS (if any). +# Here we must recognize all the valid KERNEL-OS combinations. +maybe_os=`echo $1 | sed 's/^\(.*\)-\([^-]*-[^-]*\)$/\2/'` +case $maybe_os in + nto-qnx* | linux-gnu* | linux-android* | linux-dietlibc | linux-newlib* | \ + linux-musl* | linux-uclibc* | uclinux-uclibc* | uclinux-gnu* | kfreebsd*-gnu* | \ + knetbsd*-gnu* | netbsd*-gnu* | netbsd*-eabi* | \ + kopensolaris*-gnu* | cloudabi*-eabi* | \ + storm-chaos* | os2-emx* | rtmk-nova*) + os=-$maybe_os + basic_machine=`echo $1 | sed 's/^\(.*\)-\([^-]*-[^-]*\)$/\1/'` + ;; + android-linux) + os=-linux-android + basic_machine=`echo $1 | sed 's/^\(.*\)-\([^-]*-[^-]*\)$/\1/'`-unknown + ;; + *) + basic_machine=`echo $1 | sed 's/-[^-]*$//'` + if [ $basic_machine != $1 ] + then os=`echo $1 | sed 's/.*-/-/'` + else os=; fi + ;; +esac + +### Let's recognize common machines as not being operating systems so +### that things like config.sub decstation-3100 work. We also +### recognize some manufacturers as not being operating systems, so we +### can provide default operating systems below. +case $os in + -sun*os*) + # Prevent following clause from handling this invalid input. + ;; + -dec* | -mips* | -sequent* | -encore* | -pc532* | -sgi* | -sony* | \ + -att* | -7300* | -3300* | -delta* | -motorola* | -sun[234]* | \ + -unicom* | -ibm* | -next | -hp | -isi* | -apollo | -altos* | \ + -convergent* | -ncr* | -news | -32* | -3600* | -3100* | -hitachi* |\ + -c[123]* | -convex* | -sun | -crds | -omron* | -dg | -ultra | -tti* | \ + -harris | -dolphin | -highlevel | -gould | -cbm | -ns | -masscomp | \ + -apple | -axis | -knuth | -cray | -microblaze*) + os= + basic_machine=$1 + ;; + -bluegene*) + os=-cnk + ;; + -sim | -cisco | -oki | -wec | -winbond) + os= + basic_machine=$1 + ;; + -scout) + ;; + -wrs) + os=-vxworks + basic_machine=$1 + ;; + -chorusos*) + os=-chorusos + basic_machine=$1 + ;; + -chorusrdb) + os=-chorusrdb + basic_machine=$1 + ;; + -hiux*) + os=-hiuxwe2 + ;; + -sco6) + os=-sco5v6 + basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'` + ;; + -sco5) + os=-sco3.2v5 + basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'` + ;; + -sco4) + os=-sco3.2v4 + basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'` + ;; + -sco3.2.[4-9]*) + os=`echo $os | sed -e 's/sco3.2./sco3.2v/'` + basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'` + ;; + -sco3.2v[4-9]*) + # Don't forget version if it is 3.2v4 or newer. + basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'` + ;; + -sco5v6*) + # Don't forget version if it is 3.2v4 or newer. + basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'` + ;; + -sco*) + os=-sco3.2v2 + basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'` + ;; + -udk*) + basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'` + ;; + -isc) + os=-isc2.2 + basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'` + ;; + -clix*) + basic_machine=clipper-intergraph + ;; + -isc*) + basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'` + ;; + -lynx*178) + os=-lynxos178 + ;; + -lynx*5) + os=-lynxos5 + ;; + -lynx*) + os=-lynxos + ;; + -ptx*) + basic_machine=`echo $1 | sed -e 's/86-.*/86-sequent/'` + ;; + -windowsnt*) + os=`echo $os | sed -e 's/windowsnt/winnt/'` + ;; + -psos*) + os=-psos + ;; + -mint | -mint[0-9]*) + basic_machine=m68k-atari + os=-mint + ;; +esac + +# Decode aliases for certain CPU-COMPANY combinations. +case $basic_machine in + # Recognize the basic CPU types without company name. + # Some are omitted here because they have special meanings below. + 1750a | 580 \ + | a29k \ + | aarch64 | aarch64_be \ + | alpha | alphaev[4-8] | alphaev56 | alphaev6[78] | alphapca5[67] \ + | alpha64 | alpha64ev[4-8] | alpha64ev56 | alpha64ev6[78] | alpha64pca5[67] \ + | am33_2.0 \ + | arc | arceb \ + | arm | arm[bl]e | arme[lb] | armv[2-8] | armv[3-8][lb] | armv7[arm] \ + | avr | avr32 \ + | ba \ + | be32 | be64 \ + | bfin \ + | c4x | c8051 | clipper \ + | d10v | d30v | dlx | dsp16xx \ + | e2k | epiphany \ + | fido | fr30 | frv | ft32 \ + | h8300 | h8500 | hppa | hppa1.[01] | hppa2.0 | hppa2.0[nw] | hppa64 \ + | hexagon \ + | i370 | i860 | i960 | ia64 \ + | ip2k | iq2000 \ + | k1om \ + | le32 | le64 \ + | lm32 \ + | m32c | m32r | m32rle | m68000 | m68k | m88k \ + | maxq | mb | microblaze | microblazeel | mcore | mep | metag \ + | mips | mipsbe | mipseb | mipsel | mipsle \ + | mips16 \ + | mips64 | mips64el \ + | mips64octeon | mips64octeonel \ + | mips64orion | mips64orionel \ + | mips64r5900 | mips64r5900el \ + | mips64vr | mips64vrel \ + | mips64vr4100 | mips64vr4100el \ + | mips64vr4300 | mips64vr4300el \ + | mips64vr5000 | mips64vr5000el \ + | mips64vr5900 | mips64vr5900el \ + | mipsisa32 | mipsisa32el \ + | mipsisa32r2 | mipsisa32r2el \ + | mipsisa32r6 | mipsisa32r6el \ + | mipsisa64 | mipsisa64el \ + | mipsisa64r2 | mipsisa64r2el \ + | mipsisa64r6 | mipsisa64r6el \ + | mipsisa64sb1 | mipsisa64sb1el \ + | mipsisa64sr71k | mipsisa64sr71kel \ + | mipsr5900 | mipsr5900el \ + | mipstx39 | mipstx39el \ + | mn10200 | mn10300 \ + | moxie \ + | mt \ + | msp430 \ + | nds32 | nds32le | nds32be \ + | nios | nios2 | nios2eb | nios2el \ + | ns16k | ns32k \ + | open8 | or1k | or1knd | or32 \ + | pdp10 | pdp11 | pj | pjl \ + | powerpc | powerpc64 | powerpc64le | powerpcle \ + | pru \ + | pyramid \ + | riscv32 | riscv64 \ + | rl78 | rx \ + | score \ + | sh | sh[1234] | sh[24]a | sh[24]aeb | sh[23]e | sh[234]eb | sheb | shbe | shle | sh[1234]le | sh3ele \ + | sh64 | sh64le \ + | sparc | sparc64 | sparc64b | sparc64v | sparc86x | sparclet | sparclite \ + | sparcv8 | sparcv9 | sparcv9b | sparcv9v \ + | spu \ + | tahoe | tic4x | tic54x | tic55x | tic6x | tic80 | tron \ + | ubicom32 \ + | v850 | v850e | v850e1 | v850e2 | v850es | v850e2v3 \ + | visium \ + | we32k \ + | x86 | xc16x | xstormy16 | xtensa \ + | z8k | z80) + basic_machine=$basic_machine-unknown + ;; + c54x) + basic_machine=tic54x-unknown + ;; + c55x) + basic_machine=tic55x-unknown + ;; + c6x) + basic_machine=tic6x-unknown + ;; + leon|leon[3-9]) + basic_machine=sparc-$basic_machine + ;; + m6811 | m68hc11 | m6812 | m68hc12 | m68hcs12x | nvptx | picochip) + basic_machine=$basic_machine-unknown + os=-none + ;; + m88110 | m680[12346]0 | m683?2 | m68360 | m5200 | v70 | w65 | z8k) + ;; + ms1) + basic_machine=mt-unknown + ;; + + strongarm | thumb | xscale) + basic_machine=arm-unknown + ;; + xgate) + basic_machine=$basic_machine-unknown + os=-none + ;; + xscaleeb) + basic_machine=armeb-unknown + ;; + + xscaleel) + basic_machine=armel-unknown + ;; + + # We use `pc' rather than `unknown' + # because (1) that's what they normally are, and + # (2) the word "unknown" tends to confuse beginning users. + i*86 | x86_64) + basic_machine=$basic_machine-pc + ;; + # Object if more than one company name word. + *-*-*) + echo Invalid configuration \`$1\': machine \`$basic_machine\' not recognized 1>&2 + exit 1 + ;; + # Recognize the basic CPU types with company name. + 580-* \ + | a29k-* \ + | aarch64-* | aarch64_be-* \ + | alpha-* | alphaev[4-8]-* | alphaev56-* | alphaev6[78]-* \ + | alpha64-* | alpha64ev[4-8]-* | alpha64ev56-* | alpha64ev6[78]-* \ + | alphapca5[67]-* | alpha64pca5[67]-* | arc-* | arceb-* \ + | arm-* | armbe-* | armle-* | armeb-* | armv*-* \ + | avr-* | avr32-* \ + | ba-* \ + | be32-* | be64-* \ + | bfin-* | bs2000-* \ + | c[123]* | c30-* | [cjt]90-* | c4x-* \ + | c8051-* | clipper-* | craynv-* | cydra-* \ + | d10v-* | d30v-* | dlx-* \ + | e2k-* | elxsi-* \ + | f30[01]-* | f700-* | fido-* | fr30-* | frv-* | fx80-* \ + | h8300-* | h8500-* \ + | hppa-* | hppa1.[01]-* | hppa2.0-* | hppa2.0[nw]-* | hppa64-* \ + | hexagon-* \ + | i*86-* | i860-* | i960-* | ia64-* \ + | ip2k-* | iq2000-* \ + | k1om-* \ + | le32-* | le64-* \ + | lm32-* \ + | m32c-* | m32r-* | m32rle-* \ + | m68000-* | m680[012346]0-* | m68360-* | m683?2-* | m68k-* \ + | m88110-* | m88k-* | maxq-* | mcore-* | metag-* \ + | microblaze-* | microblazeel-* \ + | mips-* | mipsbe-* | mipseb-* | mipsel-* | mipsle-* \ + | mips16-* \ + | mips64-* | mips64el-* \ + | mips64octeon-* | mips64octeonel-* \ + | mips64orion-* | mips64orionel-* \ + | mips64r5900-* | mips64r5900el-* \ + | mips64vr-* | mips64vrel-* \ + | mips64vr4100-* | mips64vr4100el-* \ + | mips64vr4300-* | mips64vr4300el-* \ + | mips64vr5000-* | mips64vr5000el-* \ + | mips64vr5900-* | mips64vr5900el-* \ + | mipsisa32-* | mipsisa32el-* \ + | mipsisa32r2-* | mipsisa32r2el-* \ + | mipsisa32r6-* | mipsisa32r6el-* \ + | mipsisa64-* | mipsisa64el-* \ + | mipsisa64r2-* | mipsisa64r2el-* \ + | mipsisa64r6-* | mipsisa64r6el-* \ + | mipsisa64sb1-* | mipsisa64sb1el-* \ + | mipsisa64sr71k-* | mipsisa64sr71kel-* \ + | mipsr5900-* | mipsr5900el-* \ + | mipstx39-* | mipstx39el-* \ + | mmix-* \ + | mt-* \ + | msp430-* \ + | nds32-* | nds32le-* | nds32be-* \ + | nios-* | nios2-* | nios2eb-* | nios2el-* \ + | none-* | np1-* | ns16k-* | ns32k-* \ + | open8-* \ + | or1k*-* \ + | orion-* \ + | pdp10-* | pdp11-* | pj-* | pjl-* | pn-* | power-* \ + | powerpc-* | powerpc64-* | powerpc64le-* | powerpcle-* \ + | pru-* \ + | pyramid-* \ + | riscv32-* | riscv64-* \ + | rl78-* | romp-* | rs6000-* | rx-* \ + | sh-* | sh[1234]-* | sh[24]a-* | sh[24]aeb-* | sh[23]e-* | sh[34]eb-* | sheb-* | shbe-* \ + | shle-* | sh[1234]le-* | sh3ele-* | sh64-* | sh64le-* \ + | sparc-* | sparc64-* | sparc64b-* | sparc64v-* | sparc86x-* | sparclet-* \ + | sparclite-* \ + | sparcv8-* | sparcv9-* | sparcv9b-* | sparcv9v-* | sv1-* | sx*-* \ + | tahoe-* \ + | tic30-* | tic4x-* | tic54x-* | tic55x-* | tic6x-* | tic80-* \ + | tile*-* \ + | tron-* \ + | ubicom32-* \ + | v850-* | v850e-* | v850e1-* | v850es-* | v850e2-* | v850e2v3-* \ + | vax-* \ + | visium-* \ + | we32k-* \ + | x86-* | x86_64-* | xc16x-* | xps100-* \ + | xstormy16-* | xtensa*-* \ + | ymp-* \ + | z8k-* | z80-*) + ;; + # Recognize the basic CPU types without company name, with glob match. + xtensa*) + basic_machine=$basic_machine-unknown + ;; + # Recognize the various machine names and aliases which stand + # for a CPU type and a company and sometimes even an OS. + 386bsd) + basic_machine=i386-unknown + os=-bsd + ;; + 3b1 | 7300 | 7300-att | att-7300 | pc7300 | safari | unixpc) + basic_machine=m68000-att + ;; + 3b*) + basic_machine=we32k-att + ;; + a29khif) + basic_machine=a29k-amd + os=-udi + ;; + abacus) + basic_machine=abacus-unknown + ;; + adobe68k) + basic_machine=m68010-adobe + os=-scout + ;; + alliant | fx80) + basic_machine=fx80-alliant + ;; + altos | altos3068) + basic_machine=m68k-altos + ;; + am29k) + basic_machine=a29k-none + os=-bsd + ;; + amd64) + basic_machine=x86_64-pc + ;; + amd64-*) + basic_machine=x86_64-`echo $basic_machine | sed 's/^[^-]*-//'` + ;; + amdahl) + basic_machine=580-amdahl + os=-sysv + ;; + amiga | amiga-*) + basic_machine=m68k-unknown + ;; + amigaos | amigados) + basic_machine=m68k-unknown + os=-amigaos + ;; + amigaunix | amix) + basic_machine=m68k-unknown + os=-sysv4 + ;; + apollo68) + basic_machine=m68k-apollo + os=-sysv + ;; + apollo68bsd) + basic_machine=m68k-apollo + os=-bsd + ;; + aros) + basic_machine=i386-pc + os=-aros + ;; + asmjs) + basic_machine=asmjs-unknown + ;; + aux) + basic_machine=m68k-apple + os=-aux + ;; + balance) + basic_machine=ns32k-sequent + os=-dynix + ;; + blackfin) + basic_machine=bfin-unknown + os=-linux + ;; + blackfin-*) + basic_machine=bfin-`echo $basic_machine | sed 's/^[^-]*-//'` + os=-linux + ;; + bluegene*) + basic_machine=powerpc-ibm + os=-cnk + ;; + c54x-*) + basic_machine=tic54x-`echo $basic_machine | sed 's/^[^-]*-//'` + ;; + c55x-*) + basic_machine=tic55x-`echo $basic_machine | sed 's/^[^-]*-//'` + ;; + c6x-*) + basic_machine=tic6x-`echo $basic_machine | sed 's/^[^-]*-//'` + ;; + c90) + basic_machine=c90-cray + os=-unicos + ;; + cegcc) + basic_machine=arm-unknown + os=-cegcc + ;; + convex-c1) + basic_machine=c1-convex + os=-bsd + ;; + convex-c2) + basic_machine=c2-convex + os=-bsd + ;; + convex-c32) + basic_machine=c32-convex + os=-bsd + ;; + convex-c34) + basic_machine=c34-convex + os=-bsd + ;; + convex-c38) + basic_machine=c38-convex + os=-bsd + ;; + cray | j90) + basic_machine=j90-cray + os=-unicos + ;; + craynv) + basic_machine=craynv-cray + os=-unicosmp + ;; + cr16 | cr16-*) + basic_machine=cr16-unknown + os=-elf + ;; + crds | unos) + basic_machine=m68k-crds + ;; + crisv32 | crisv32-* | etraxfs*) + basic_machine=crisv32-axis + ;; + cris | cris-* | etrax*) + basic_machine=cris-axis + ;; + crx) + basic_machine=crx-unknown + os=-elf + ;; + da30 | da30-*) + basic_machine=m68k-da30 + ;; + decstation | decstation-3100 | pmax | pmax-* | pmin | dec3100 | decstatn) + basic_machine=mips-dec + ;; + decsystem10* | dec10*) + basic_machine=pdp10-dec + os=-tops10 + ;; + decsystem20* | dec20*) + basic_machine=pdp10-dec + os=-tops20 + ;; + delta | 3300 | motorola-3300 | motorola-delta \ + | 3300-motorola | delta-motorola) + basic_machine=m68k-motorola + ;; + delta88) + basic_machine=m88k-motorola + os=-sysv3 + ;; + dicos) + basic_machine=i686-pc + os=-dicos + ;; + djgpp) + basic_machine=i586-pc + os=-msdosdjgpp + ;; + dpx20 | dpx20-*) + basic_machine=rs6000-bull + os=-bosx + ;; + dpx2* | dpx2*-bull) + basic_machine=m68k-bull + os=-sysv3 + ;; + e500v[12]) + basic_machine=powerpc-unknown + os=$os"spe" + ;; + e500v[12]-*) + basic_machine=powerpc-`echo $basic_machine | sed 's/^[^-]*-//'` + os=$os"spe" + ;; + ebmon29k) + basic_machine=a29k-amd + os=-ebmon + ;; + elxsi) + basic_machine=elxsi-elxsi + os=-bsd + ;; + encore | umax | mmax) + basic_machine=ns32k-encore + ;; + es1800 | OSE68k | ose68k | ose | OSE) + basic_machine=m68k-ericsson + os=-ose + ;; + fx2800) + basic_machine=i860-alliant + ;; + genix) + basic_machine=ns32k-ns + ;; + gmicro) + basic_machine=tron-gmicro + os=-sysv + ;; + go32) + basic_machine=i386-pc + os=-go32 + ;; + h3050r* | hiux*) + basic_machine=hppa1.1-hitachi + os=-hiuxwe2 + ;; + h8300hms) + basic_machine=h8300-hitachi + os=-hms + ;; + h8300xray) + basic_machine=h8300-hitachi + os=-xray + ;; + h8500hms) + basic_machine=h8500-hitachi + os=-hms + ;; + harris) + basic_machine=m88k-harris + os=-sysv3 + ;; + hp300-*) + basic_machine=m68k-hp + ;; + hp300bsd) + basic_machine=m68k-hp + os=-bsd + ;; + hp300hpux) + basic_machine=m68k-hp + os=-hpux + ;; + hp3k9[0-9][0-9] | hp9[0-9][0-9]) + basic_machine=hppa1.0-hp + ;; + hp9k2[0-9][0-9] | hp9k31[0-9]) + basic_machine=m68000-hp + ;; + hp9k3[2-9][0-9]) + basic_machine=m68k-hp + ;; + hp9k6[0-9][0-9] | hp6[0-9][0-9]) + basic_machine=hppa1.0-hp + ;; + hp9k7[0-79][0-9] | hp7[0-79][0-9]) + basic_machine=hppa1.1-hp + ;; + hp9k78[0-9] | hp78[0-9]) + # FIXME: really hppa2.0-hp + basic_machine=hppa1.1-hp + ;; + hp9k8[67]1 | hp8[67]1 | hp9k80[24] | hp80[24] | hp9k8[78]9 | hp8[78]9 | hp9k893 | hp893) + # FIXME: really hppa2.0-hp + basic_machine=hppa1.1-hp + ;; + hp9k8[0-9][13679] | hp8[0-9][13679]) + basic_machine=hppa1.1-hp + ;; + hp9k8[0-9][0-9] | hp8[0-9][0-9]) + basic_machine=hppa1.0-hp + ;; + hppa-next) + os=-nextstep3 + ;; + hppaosf) + basic_machine=hppa1.1-hp + os=-osf + ;; + hppro) + basic_machine=hppa1.1-hp + os=-proelf + ;; + i370-ibm* | ibm*) + basic_machine=i370-ibm + ;; + i*86v32) + basic_machine=`echo $1 | sed -e 's/86.*/86-pc/'` + os=-sysv32 + ;; + i*86v4*) + basic_machine=`echo $1 | sed -e 's/86.*/86-pc/'` + os=-sysv4 + ;; + i*86v) + basic_machine=`echo $1 | sed -e 's/86.*/86-pc/'` + os=-sysv + ;; + i*86sol2) + basic_machine=`echo $1 | sed -e 's/86.*/86-pc/'` + os=-solaris2 + ;; + i386mach) + basic_machine=i386-mach + os=-mach + ;; + i386-vsta | vsta) + basic_machine=i386-unknown + os=-vsta + ;; + iris | iris4d) + basic_machine=mips-sgi + case $os in + -irix*) + ;; + *) + os=-irix4 + ;; + esac + ;; + isi68 | isi) + basic_machine=m68k-isi + os=-sysv + ;; + leon-*|leon[3-9]-*) + basic_machine=sparc-`echo $basic_machine | sed 's/-.*//'` + ;; + m68knommu) + basic_machine=m68k-unknown + os=-linux + ;; + m68knommu-*) + basic_machine=m68k-`echo $basic_machine | sed 's/^[^-]*-//'` + os=-linux + ;; + m88k-omron*) + basic_machine=m88k-omron + ;; + magnum | m3230) + basic_machine=mips-mips + os=-sysv + ;; + merlin) + basic_machine=ns32k-utek + os=-sysv + ;; + microblaze*) + basic_machine=microblaze-xilinx + ;; + mingw64) + basic_machine=x86_64-pc + os=-mingw64 + ;; + mingw32) + basic_machine=i686-pc + os=-mingw32 + ;; + mingw32ce) + basic_machine=arm-unknown + os=-mingw32ce + ;; + miniframe) + basic_machine=m68000-convergent + ;; + *mint | -mint[0-9]* | *MiNT | *MiNT[0-9]*) + basic_machine=m68k-atari + os=-mint + ;; + mips3*-*) + basic_machine=`echo $basic_machine | sed -e 's/mips3/mips64/'` + ;; + mips3*) + basic_machine=`echo $basic_machine | sed -e 's/mips3/mips64/'`-unknown + ;; + monitor) + basic_machine=m68k-rom68k + os=-coff + ;; + morphos) + basic_machine=powerpc-unknown + os=-morphos + ;; + moxiebox) + basic_machine=moxie-unknown + os=-moxiebox + ;; + msdos) + basic_machine=i386-pc + os=-msdos + ;; + ms1-*) + basic_machine=`echo $basic_machine | sed -e 's/ms1-/mt-/'` + ;; + msys) + basic_machine=i686-pc + os=-msys + ;; + mvs) + basic_machine=i370-ibm + os=-mvs + ;; + nacl) + basic_machine=le32-unknown + os=-nacl + ;; + ncr3000) + basic_machine=i486-ncr + os=-sysv4 + ;; + netbsd386) + basic_machine=i386-unknown + os=-netbsd + ;; + netwinder) + basic_machine=armv4l-rebel + os=-linux + ;; + news | news700 | news800 | news900) + basic_machine=m68k-sony + os=-newsos + ;; + news1000) + basic_machine=m68030-sony + os=-newsos + ;; + news-3600 | risc-news) + basic_machine=mips-sony + os=-newsos + ;; + necv70) + basic_machine=v70-nec + os=-sysv + ;; + next | m*-next ) + basic_machine=m68k-next + case $os in + -nextstep* ) + ;; + -ns2*) + os=-nextstep2 + ;; + *) + os=-nextstep3 + ;; + esac + ;; + nh3000) + basic_machine=m68k-harris + os=-cxux + ;; + nh[45]000) + basic_machine=m88k-harris + os=-cxux + ;; + nindy960) + basic_machine=i960-intel + os=-nindy + ;; + mon960) + basic_machine=i960-intel + os=-mon960 + ;; + nonstopux) + basic_machine=mips-compaq + os=-nonstopux + ;; + np1) + basic_machine=np1-gould + ;; + neo-tandem) + basic_machine=neo-tandem + ;; + nse-tandem) + basic_machine=nse-tandem + ;; + nsr-tandem) + basic_machine=nsr-tandem + ;; + op50n-* | op60c-*) + basic_machine=hppa1.1-oki + os=-proelf + ;; + openrisc | openrisc-*) + basic_machine=or32-unknown + ;; + os400) + basic_machine=powerpc-ibm + os=-os400 + ;; + OSE68000 | ose68000) + basic_machine=m68000-ericsson + os=-ose + ;; + os68k) + basic_machine=m68k-none + os=-os68k + ;; + pa-hitachi) + basic_machine=hppa1.1-hitachi + os=-hiuxwe2 + ;; + paragon) + basic_machine=i860-intel + os=-osf + ;; + parisc) + basic_machine=hppa-unknown + os=-linux + ;; + parisc-*) + basic_machine=hppa-`echo $basic_machine | sed 's/^[^-]*-//'` + os=-linux + ;; + pbd) + basic_machine=sparc-tti + ;; + pbb) + basic_machine=m68k-tti + ;; + pc532 | pc532-*) + basic_machine=ns32k-pc532 + ;; + pc98) + basic_machine=i386-pc + ;; + pc98-*) + basic_machine=i386-`echo $basic_machine | sed 's/^[^-]*-//'` + ;; + pentium | p5 | k5 | k6 | nexgen | viac3) + basic_machine=i586-pc + ;; + pentiumpro | p6 | 6x86 | athlon | athlon_*) + basic_machine=i686-pc + ;; + pentiumii | pentium2 | pentiumiii | pentium3) + basic_machine=i686-pc + ;; + pentium4) + basic_machine=i786-pc + ;; + pentium-* | p5-* | k5-* | k6-* | nexgen-* | viac3-*) + basic_machine=i586-`echo $basic_machine | sed 's/^[^-]*-//'` + ;; + pentiumpro-* | p6-* | 6x86-* | athlon-*) + basic_machine=i686-`echo $basic_machine | sed 's/^[^-]*-//'` + ;; + pentiumii-* | pentium2-* | pentiumiii-* | pentium3-*) + basic_machine=i686-`echo $basic_machine | sed 's/^[^-]*-//'` + ;; + pentium4-*) + basic_machine=i786-`echo $basic_machine | sed 's/^[^-]*-//'` + ;; + pn) + basic_machine=pn-gould + ;; + power) basic_machine=power-ibm + ;; + ppc | ppcbe) basic_machine=powerpc-unknown + ;; + ppc-* | ppcbe-*) + basic_machine=powerpc-`echo $basic_machine | sed 's/^[^-]*-//'` + ;; + ppcle | powerpclittle) + basic_machine=powerpcle-unknown + ;; + ppcle-* | powerpclittle-*) + basic_machine=powerpcle-`echo $basic_machine | sed 's/^[^-]*-//'` + ;; + ppc64) basic_machine=powerpc64-unknown + ;; + ppc64-*) basic_machine=powerpc64-`echo $basic_machine | sed 's/^[^-]*-//'` + ;; + ppc64le | powerpc64little) + basic_machine=powerpc64le-unknown + ;; + ppc64le-* | powerpc64little-*) + basic_machine=powerpc64le-`echo $basic_machine | sed 's/^[^-]*-//'` + ;; + ps2) + basic_machine=i386-ibm + ;; + pw32) + basic_machine=i586-unknown + os=-pw32 + ;; + rdos | rdos64) + basic_machine=x86_64-pc + os=-rdos + ;; + rdos32) + basic_machine=i386-pc + os=-rdos + ;; + rom68k) + basic_machine=m68k-rom68k + os=-coff + ;; + rm[46]00) + basic_machine=mips-siemens + ;; + rtpc | rtpc-*) + basic_machine=romp-ibm + ;; + s390 | s390-*) + basic_machine=s390-ibm + ;; + s390x | s390x-*) + basic_machine=s390x-ibm + ;; + sa29200) + basic_machine=a29k-amd + os=-udi + ;; + sb1) + basic_machine=mipsisa64sb1-unknown + ;; + sb1el) + basic_machine=mipsisa64sb1el-unknown + ;; + sde) + basic_machine=mipsisa32-sde + os=-elf + ;; + sei) + basic_machine=mips-sei + os=-seiux + ;; + sequent) + basic_machine=i386-sequent + ;; + sh) + basic_machine=sh-hitachi + os=-hms + ;; + sh5el) + basic_machine=sh5le-unknown + ;; + sh64) + basic_machine=sh64-unknown + ;; + sparclite-wrs | simso-wrs) + basic_machine=sparclite-wrs + os=-vxworks + ;; + sps7) + basic_machine=m68k-bull + os=-sysv2 + ;; + spur) + basic_machine=spur-unknown + ;; + st2000) + basic_machine=m68k-tandem + ;; + stratus) + basic_machine=i860-stratus + os=-sysv4 + ;; + strongarm-* | thumb-*) + basic_machine=arm-`echo $basic_machine | sed 's/^[^-]*-//'` + ;; + sun2) + basic_machine=m68000-sun + ;; + sun2os3) + basic_machine=m68000-sun + os=-sunos3 + ;; + sun2os4) + basic_machine=m68000-sun + os=-sunos4 + ;; + sun3os3) + basic_machine=m68k-sun + os=-sunos3 + ;; + sun3os4) + basic_machine=m68k-sun + os=-sunos4 + ;; + sun4os3) + basic_machine=sparc-sun + os=-sunos3 + ;; + sun4os4) + basic_machine=sparc-sun + os=-sunos4 + ;; + sun4sol2) + basic_machine=sparc-sun + os=-solaris2 + ;; + sun3 | sun3-*) + basic_machine=m68k-sun + ;; + sun4) + basic_machine=sparc-sun + ;; + sun386 | sun386i | roadrunner) + basic_machine=i386-sun + ;; + sv1) + basic_machine=sv1-cray + os=-unicos + ;; + symmetry) + basic_machine=i386-sequent + os=-dynix + ;; + t3e) + basic_machine=alphaev5-cray + os=-unicos + ;; + t90) + basic_machine=t90-cray + os=-unicos + ;; + tile*) + basic_machine=$basic_machine-unknown + os=-linux-gnu + ;; + tx39) + basic_machine=mipstx39-unknown + ;; + tx39el) + basic_machine=mipstx39el-unknown + ;; + toad1) + basic_machine=pdp10-xkl + os=-tops20 + ;; + tower | tower-32) + basic_machine=m68k-ncr + ;; + tpf) + basic_machine=s390x-ibm + os=-tpf + ;; + udi29k) + basic_machine=a29k-amd + os=-udi + ;; + ultra3) + basic_machine=a29k-nyu + os=-sym1 + ;; + v810 | necv810) + basic_machine=v810-nec + os=-none + ;; + vaxv) + basic_machine=vax-dec + os=-sysv + ;; + vms) + basic_machine=vax-dec + os=-vms + ;; + vpp*|vx|vx-*) + basic_machine=f301-fujitsu + ;; + vxworks960) + basic_machine=i960-wrs + os=-vxworks + ;; + vxworks68) + basic_machine=m68k-wrs + os=-vxworks + ;; + vxworks29k) + basic_machine=a29k-wrs + os=-vxworks + ;; + w65*) + basic_machine=w65-wdc + os=-none + ;; + w89k-*) + basic_machine=hppa1.1-winbond + os=-proelf + ;; + xbox) + basic_machine=i686-pc + os=-mingw32 + ;; + xps | xps100) + basic_machine=xps100-honeywell + ;; + xscale-* | xscalee[bl]-*) + basic_machine=`echo $basic_machine | sed 's/^xscale/arm/'` + ;; + ymp) + basic_machine=ymp-cray + os=-unicos + ;; + z8k-*-coff) + basic_machine=z8k-unknown + os=-sim + ;; + z80-*-coff) + basic_machine=z80-unknown + os=-sim + ;; + none) + basic_machine=none-none + os=-none + ;; + +# Here we handle the default manufacturer of certain CPU types. It is in +# some cases the only manufacturer, in others, it is the most popular. + w89k) + basic_machine=hppa1.1-winbond + ;; + op50n) + basic_machine=hppa1.1-oki + ;; + op60c) + basic_machine=hppa1.1-oki + ;; + romp) + basic_machine=romp-ibm + ;; + mmix) + basic_machine=mmix-knuth + ;; + rs6000) + basic_machine=rs6000-ibm + ;; + vax) + basic_machine=vax-dec + ;; + pdp10) + # there are many clones, so DEC is not a safe bet + basic_machine=pdp10-unknown + ;; + pdp11) + basic_machine=pdp11-dec + ;; + we32k) + basic_machine=we32k-att + ;; + sh[1234] | sh[24]a | sh[24]aeb | sh[34]eb | sh[1234]le | sh[23]ele) + basic_machine=sh-unknown + ;; + sparc | sparcv8 | sparcv9 | sparcv9b | sparcv9v) + basic_machine=sparc-sun + ;; + cydra) + basic_machine=cydra-cydrome + ;; + orion) + basic_machine=orion-highlevel + ;; + orion105) + basic_machine=clipper-highlevel + ;; + mac | mpw | mac-mpw) + basic_machine=m68k-apple + ;; + pmac | pmac-mpw) + basic_machine=powerpc-apple + ;; + *-unknown) + # Make sure to match an already-canonicalized machine name. + ;; + *) + echo Invalid configuration \`$1\': machine \`$basic_machine\' not recognized 1>&2 + exit 1 + ;; +esac + +# Here we canonicalize certain aliases for manufacturers. +case $basic_machine in + *-digital*) + basic_machine=`echo $basic_machine | sed 's/digital.*/dec/'` + ;; + *-commodore*) + basic_machine=`echo $basic_machine | sed 's/commodore.*/cbm/'` + ;; + *) + ;; +esac + +# Decode manufacturer-specific aliases for certain operating systems. + +if [ x"$os" != x"" ] +then +case $os in + # First match some system type aliases + # that might get confused with valid system types. + # -solaris* is a basic system type, with this one exception. + -auroraux) + os=-auroraux + ;; + -solaris1 | -solaris1.*) + os=`echo $os | sed -e 's|solaris1|sunos4|'` + ;; + -solaris) + os=-solaris2 + ;; + -svr4*) + os=-sysv4 + ;; + -unixware*) + os=-sysv4.2uw + ;; + -gnu/linux*) + os=`echo $os | sed -e 's|gnu/linux|linux-gnu|'` + ;; + # First accept the basic system types. + # The portable systems comes first. + # Each alternative MUST END IN A *, to match a version number. + # -sysv* is not here because it comes later, after sysvr4. + -gnu* | -bsd* | -mach* | -minix* | -genix* | -ultrix* | -irix* \ + | -*vms* | -sco* | -esix* | -isc* | -aix* | -cnk* | -sunos | -sunos[34]*\ + | -hpux* | -unos* | -osf* | -luna* | -dgux* | -auroraux* | -solaris* \ + | -sym* | -kopensolaris* | -plan9* \ + | -amigaos* | -amigados* | -msdos* | -newsos* | -unicos* | -aof* \ + | -aos* | -aros* | -cloudabi* | -sortix* \ + | -nindy* | -vxsim* | -vxworks* | -ebmon* | -hms* | -mvs* \ + | -clix* | -riscos* | -uniplus* | -iris* | -rtu* | -xenix* \ + | -hiux* | -386bsd* | -knetbsd* | -mirbsd* | -netbsd* \ + | -bitrig* | -openbsd* | -solidbsd* | -libertybsd* \ + | -ekkobsd* | -kfreebsd* | -freebsd* | -riscix* | -lynxos* \ + | -bosx* | -nextstep* | -cxux* | -aout* | -elf* | -oabi* \ + | -ptx* | -coff* | -ecoff* | -winnt* | -domain* | -vsta* \ + | -udi* | -eabi* | -lites* | -ieee* | -go32* | -aux* \ + | -chorusos* | -chorusrdb* | -cegcc* \ + | -cygwin* | -msys* | -pe* | -psos* | -moss* | -proelf* | -rtems* \ + | -midipix* | -mingw32* | -mingw64* | -linux-gnu* | -linux-android* \ + | -linux-newlib* | -linux-musl* | -linux-uclibc* \ + | -uxpv* | -beos* | -mpeix* | -udk* | -moxiebox* \ + | -interix* | -uwin* | -mks* | -rhapsody* | -darwin* | -opened* \ + | -openstep* | -oskit* | -conix* | -pw32* | -nonstopux* \ + | -storm-chaos* | -tops10* | -tenex* | -tops20* | -its* \ + | -os2* | -vos* | -palmos* | -uclinux* | -nucleus* \ + | -morphos* | -superux* | -rtmk* | -rtmk-nova* | -windiss* \ + | -powermax* | -dnix* | -nx6 | -nx7 | -sei* | -dragonfly* \ + | -skyos* | -haiku* | -rdos* | -toppers* | -drops* | -es* \ + | -onefs* | -tirtos* | -phoenix* | -fuchsia*) + # Remember, each alternative MUST END IN *, to match a version number. + ;; + -qnx*) + case $basic_machine in + x86-* | i*86-*) + ;; + *) + os=-nto$os + ;; + esac + ;; + -nto-qnx*) + ;; + -nto*) + os=`echo $os | sed -e 's|nto|nto-qnx|'` + ;; + -sim | -es1800* | -hms* | -xray | -os68k* | -none* | -v88r* \ + | -windows* | -osx | -abug | -netware* | -os9* | -beos* | -haiku* \ + | -macos* | -mpw* | -magic* | -mmixware* | -mon960* | -lnews*) + ;; + -mac*) + os=`echo $os | sed -e 's|mac|macos|'` + ;; + -linux-dietlibc) + os=-linux-dietlibc + ;; + -linux*) + os=`echo $os | sed -e 's|linux|linux-gnu|'` + ;; + -sunos5*) + os=`echo $os | sed -e 's|sunos5|solaris2|'` + ;; + -sunos6*) + os=`echo $os | sed -e 's|sunos6|solaris3|'` + ;; + -opened*) + os=-openedition + ;; + -os400*) + os=-os400 + ;; + -wince*) + os=-wince + ;; + -osfrose*) + os=-osfrose + ;; + -osf*) + os=-osf + ;; + -utek*) + os=-bsd + ;; + -dynix*) + os=-bsd + ;; + -acis*) + os=-aos + ;; + -atheos*) + os=-atheos + ;; + -syllable*) + os=-syllable + ;; + -386bsd) + os=-bsd + ;; + -ctix* | -uts*) + os=-sysv + ;; + -nova*) + os=-rtmk-nova + ;; + -ns2 ) + os=-nextstep2 + ;; + -nsk*) + os=-nsk + ;; + # Preserve the version number of sinix5. + -sinix5.*) + os=`echo $os | sed -e 's|sinix|sysv|'` + ;; + -sinix*) + os=-sysv4 + ;; + -tpf*) + os=-tpf + ;; + -triton*) + os=-sysv3 + ;; + -oss*) + os=-sysv3 + ;; + -svr4) + os=-sysv4 + ;; + -svr3) + os=-sysv3 + ;; + -sysvr4) + os=-sysv4 + ;; + # This must come after -sysvr4. + -sysv*) + ;; + -ose*) + os=-ose + ;; + -es1800*) + os=-ose + ;; + -xenix) + os=-xenix + ;; + -*mint | -mint[0-9]* | -*MiNT | -MiNT[0-9]*) + os=-mint + ;; + -aros*) + os=-aros + ;; + -zvmoe) + os=-zvmoe + ;; + -dicos*) + os=-dicos + ;; + -nacl*) + ;; + -ios) + ;; + -none) + ;; + *) + # Get rid of the `-' at the beginning of $os. + os=`echo $os | sed 's/[^-]*-//'` + echo Invalid configuration \`$1\': system \`$os\' not recognized 1>&2 + exit 1 + ;; +esac +else + +# Here we handle the default operating systems that come with various machines. +# The value should be what the vendor currently ships out the door with their +# machine or put another way, the most popular os provided with the machine. + +# Note that if you're going to try to match "-MANUFACTURER" here (say, +# "-sun"), then you have to tell the case statement up towards the top +# that MANUFACTURER isn't an operating system. Otherwise, code above +# will signal an error saying that MANUFACTURER isn't an operating +# system, and we'll never get to this point. + +case $basic_machine in + score-*) + os=-elf + ;; + spu-*) + os=-elf + ;; + *-acorn) + os=-riscix1.2 + ;; + arm*-rebel) + os=-linux + ;; + arm*-semi) + os=-aout + ;; + c4x-* | tic4x-*) + os=-coff + ;; + c8051-*) + os=-elf + ;; + hexagon-*) + os=-elf + ;; + tic54x-*) + os=-coff + ;; + tic55x-*) + os=-coff + ;; + tic6x-*) + os=-coff + ;; + # This must come before the *-dec entry. + pdp10-*) + os=-tops20 + ;; + pdp11-*) + os=-none + ;; + *-dec | vax-*) + os=-ultrix4.2 + ;; + m68*-apollo) + os=-domain + ;; + i386-sun) + os=-sunos4.0.2 + ;; + m68000-sun) + os=-sunos3 + ;; + m68*-cisco) + os=-aout + ;; + mep-*) + os=-elf + ;; + mips*-cisco) + os=-elf + ;; + mips*-*) + os=-elf + ;; + or32-*) + os=-coff + ;; + *-tti) # must be before sparc entry or we get the wrong os. + os=-sysv3 + ;; + sparc-* | *-sun) + os=-sunos4.1.1 + ;; + *-be) + os=-beos + ;; + *-haiku) + os=-haiku + ;; + *-ibm) + os=-aix + ;; + *-knuth) + os=-mmixware + ;; + *-wec) + os=-proelf + ;; + *-winbond) + os=-proelf + ;; + *-oki) + os=-proelf + ;; + *-hp) + os=-hpux + ;; + *-hitachi) + os=-hiux + ;; + i860-* | *-att | *-ncr | *-altos | *-motorola | *-convergent) + os=-sysv + ;; + *-cbm) + os=-amigaos + ;; + *-dg) + os=-dgux + ;; + *-dolphin) + os=-sysv3 + ;; + m68k-ccur) + os=-rtu + ;; + m88k-omron*) + os=-luna + ;; + *-next ) + os=-nextstep + ;; + *-sequent) + os=-ptx + ;; + *-crds) + os=-unos + ;; + *-ns) + os=-genix + ;; + i370-*) + os=-mvs + ;; + *-next) + os=-nextstep3 + ;; + *-gould) + os=-sysv + ;; + *-highlevel) + os=-bsd + ;; + *-encore) + os=-bsd + ;; + *-sgi) + os=-irix + ;; + *-siemens) + os=-sysv4 + ;; + *-masscomp) + os=-rtu + ;; + f30[01]-fujitsu | f700-fujitsu) + os=-uxpv + ;; + *-rom68k) + os=-coff + ;; + *-*bug) + os=-coff + ;; + *-apple) + os=-macos + ;; + *-atari*) + os=-mint + ;; + *) + os=-none + ;; +esac +fi + +# Here we handle the case where we know the os, and the CPU type, but not the +# manufacturer. We pick the logical manufacturer. +vendor=unknown +case $basic_machine in + *-unknown) + case $os in + -riscix*) + vendor=acorn + ;; + -sunos*) + vendor=sun + ;; + -cnk*|-aix*) + vendor=ibm + ;; + -beos*) + vendor=be + ;; + -hpux*) + vendor=hp + ;; + -mpeix*) + vendor=hp + ;; + -hiux*) + vendor=hitachi + ;; + -unos*) + vendor=crds + ;; + -dgux*) + vendor=dg + ;; + -luna*) + vendor=omron + ;; + -genix*) + vendor=ns + ;; + -mvs* | -opened*) + vendor=ibm + ;; + -os400*) + vendor=ibm + ;; + -ptx*) + vendor=sequent + ;; + -tpf*) + vendor=ibm + ;; + -vxsim* | -vxworks* | -windiss*) + vendor=wrs + ;; + -aux*) + vendor=apple + ;; + -hms*) + vendor=hitachi + ;; + -mpw* | -macos*) + vendor=apple + ;; + -*mint | -mint[0-9]* | -*MiNT | -MiNT[0-9]*) + vendor=atari + ;; + -vos*) + vendor=stratus + ;; + esac + basic_machine=`echo $basic_machine | sed "s/unknown/$vendor/"` + ;; +esac + +echo $basic_machine$os +exit + +# Local variables: +# eval: (add-hook 'write-file-hooks 'time-stamp) +# time-stamp-start: "timestamp='" +# time-stamp-format: "%:y-%02m-%02d" +# time-stamp-end: "'" +# End: diff --git a/deps/jemalloc/build-aux/install-sh b/deps/jemalloc/build-aux/install-sh new file mode 100755 index 000000000..ebc66913e --- /dev/null +++ b/deps/jemalloc/build-aux/install-sh @@ -0,0 +1,250 @@ +#! /bin/sh +# +# install - install a program, script, or datafile +# This comes from X11R5 (mit/util/scripts/install.sh). +# +# Copyright 1991 by the Massachusetts Institute of Technology +# +# Permission to use, copy, modify, distribute, and sell this software and its +# documentation for any purpose is hereby granted without fee, provided that +# the above copyright notice appear in all copies and that both that +# copyright notice and this permission notice appear in supporting +# documentation, and that the name of M.I.T. not be used in advertising or +# publicity pertaining to distribution of the software without specific, +# written prior permission. M.I.T. makes no representations about the +# suitability of this software for any purpose. It is provided "as is" +# without express or implied warranty. +# +# Calling this script install-sh is preferred over install.sh, to prevent +# `make' implicit rules from creating a file called install from it +# when there is no Makefile. +# +# This script is compatible with the BSD install script, but was written +# from scratch. It can only install one file at a time, a restriction +# shared with many OS's install programs. + + +# set DOITPROG to echo to test this script + +# Don't use :- since 4.3BSD and earlier shells don't like it. +doit="${DOITPROG-}" + + +# put in absolute paths if you don't have them in your path; or use env. vars. + +mvprog="${MVPROG-mv}" +cpprog="${CPPROG-cp}" +chmodprog="${CHMODPROG-chmod}" +chownprog="${CHOWNPROG-chown}" +chgrpprog="${CHGRPPROG-chgrp}" +stripprog="${STRIPPROG-strip}" +rmprog="${RMPROG-rm}" +mkdirprog="${MKDIRPROG-mkdir}" + +transformbasename="" +transform_arg="" +instcmd="$mvprog" +chmodcmd="$chmodprog 0755" +chowncmd="" +chgrpcmd="" +stripcmd="" +rmcmd="$rmprog -f" +mvcmd="$mvprog" +src="" +dst="" +dir_arg="" + +while [ x"$1" != x ]; do + case $1 in + -c) instcmd="$cpprog" + shift + continue;; + + -d) dir_arg=true + shift + continue;; + + -m) chmodcmd="$chmodprog $2" + shift + shift + continue;; + + -o) chowncmd="$chownprog $2" + shift + shift + continue;; + + -g) chgrpcmd="$chgrpprog $2" + shift + shift + continue;; + + -s) stripcmd="$stripprog" + shift + continue;; + + -t=*) transformarg=`echo $1 | sed 's/-t=//'` + shift + continue;; + + -b=*) transformbasename=`echo $1 | sed 's/-b=//'` + shift + continue;; + + *) if [ x"$src" = x ] + then + src=$1 + else + # this colon is to work around a 386BSD /bin/sh bug + : + dst=$1 + fi + shift + continue;; + esac +done + +if [ x"$src" = x ] +then + echo "install: no input file specified" + exit 1 +else + true +fi + +if [ x"$dir_arg" != x ]; then + dst=$src + src="" + + if [ -d $dst ]; then + instcmd=: + else + instcmd=mkdir + fi +else + +# Waiting for this to be detected by the "$instcmd $src $dsttmp" command +# might cause directories to be created, which would be especially bad +# if $src (and thus $dsttmp) contains '*'. + + if [ -f $src -o -d $src ] + then + true + else + echo "install: $src does not exist" + exit 1 + fi + + if [ x"$dst" = x ] + then + echo "install: no destination specified" + exit 1 + else + true + fi + +# If destination is a directory, append the input filename; if your system +# does not like double slashes in filenames, you may need to add some logic + + if [ -d $dst ] + then + dst="$dst"/`basename $src` + else + true + fi +fi + +## this sed command emulates the dirname command +dstdir=`echo $dst | sed -e 's,[^/]*$,,;s,/$,,;s,^$,.,'` + +# Make sure that the destination directory exists. +# this part is taken from Noah Friedman's mkinstalldirs script + +# Skip lots of stat calls in the usual case. +if [ ! -d "$dstdir" ]; then +defaultIFS=' +' +IFS="${IFS-${defaultIFS}}" + +oIFS="${IFS}" +# Some sh's can't handle IFS=/ for some reason. +IFS='%' +set - `echo ${dstdir} | sed -e 's@/@%@g' -e 's@^%@/@'` +IFS="${oIFS}" + +pathcomp='' + +while [ $# -ne 0 ] ; do + pathcomp="${pathcomp}${1}" + shift + + if [ ! -d "${pathcomp}" ] ; + then + $mkdirprog "${pathcomp}" + else + true + fi + + pathcomp="${pathcomp}/" +done +fi + +if [ x"$dir_arg" != x ] +then + $doit $instcmd $dst && + + if [ x"$chowncmd" != x ]; then $doit $chowncmd $dst; else true ; fi && + if [ x"$chgrpcmd" != x ]; then $doit $chgrpcmd $dst; else true ; fi && + if [ x"$stripcmd" != x ]; then $doit $stripcmd $dst; else true ; fi && + if [ x"$chmodcmd" != x ]; then $doit $chmodcmd $dst; else true ; fi +else + +# If we're going to rename the final executable, determine the name now. + + if [ x"$transformarg" = x ] + then + dstfile=`basename $dst` + else + dstfile=`basename $dst $transformbasename | + sed $transformarg`$transformbasename + fi + +# don't allow the sed command to completely eliminate the filename + + if [ x"$dstfile" = x ] + then + dstfile=`basename $dst` + else + true + fi + +# Make a temp file name in the proper directory. + + dsttmp=$dstdir/#inst.$$# + +# Move or copy the file name to the temp name + + $doit $instcmd $src $dsttmp && + + trap "rm -f ${dsttmp}" 0 && + +# and set any options; do chmod last to preserve setuid bits + +# If any of these fail, we abort the whole thing. If we want to +# ignore errors from any of these, just make sure not to ignore +# errors from the above "$doit $instcmd $src $dsttmp" command. + + if [ x"$chowncmd" != x ]; then $doit $chowncmd $dsttmp; else true;fi && + if [ x"$chgrpcmd" != x ]; then $doit $chgrpcmd $dsttmp; else true;fi && + if [ x"$stripcmd" != x ]; then $doit $stripcmd $dsttmp; else true;fi && + if [ x"$chmodcmd" != x ]; then $doit $chmodcmd $dsttmp; else true;fi && + +# Now rename the file to the real destination. + + $doit $rmcmd -f $dstdir/$dstfile && + $doit $mvcmd $dsttmp $dstdir/$dstfile + +fi && + + +exit 0 diff --git a/deps/jemalloc/config.stamp.in b/deps/jemalloc/config.stamp.in new file mode 100644 index 000000000..e69de29bb --- /dev/null +++ b/deps/jemalloc/config.stamp.in diff --git a/deps/jemalloc/configure.ac b/deps/jemalloc/configure.ac new file mode 100644 index 000000000..261d81c00 --- /dev/null +++ b/deps/jemalloc/configure.ac @@ -0,0 +1,2406 @@ +dnl Process this file with autoconf to produce a configure script. +AC_PREREQ(2.68) +AC_INIT([Makefile.in]) + +AC_CONFIG_AUX_DIR([build-aux]) + +dnl ============================================================================ +dnl Custom macro definitions. + +dnl JE_CONCAT_VVV(r, a, b) +dnl +dnl Set $r to the concatenation of $a and $b, with a space separating them iff +dnl both $a and $b are non-empty. +AC_DEFUN([JE_CONCAT_VVV], +if test "x[$]{$2}" = "x" -o "x[$]{$3}" = "x" ; then + $1="[$]{$2}[$]{$3}" +else + $1="[$]{$2} [$]{$3}" +fi +) + +dnl JE_APPEND_VS(a, b) +dnl +dnl Set $a to the concatenation of $a and b, with a space separating them iff +dnl both $a and b are non-empty. +AC_DEFUN([JE_APPEND_VS], + T_APPEND_V=$2 + JE_CONCAT_VVV($1, $1, T_APPEND_V) +) + +CONFIGURE_CFLAGS= +SPECIFIED_CFLAGS="${CFLAGS}" +dnl JE_CFLAGS_ADD(cflag) +dnl +dnl CFLAGS is the concatenation of CONFIGURE_CFLAGS and SPECIFIED_CFLAGS +dnl (ignoring EXTRA_CFLAGS, which does not impact configure tests. This macro +dnl appends to CONFIGURE_CFLAGS and regenerates CFLAGS. +AC_DEFUN([JE_CFLAGS_ADD], +[ +AC_MSG_CHECKING([whether compiler supports $1]) +T_CONFIGURE_CFLAGS="${CONFIGURE_CFLAGS}" +JE_APPEND_VS(CONFIGURE_CFLAGS, $1) +JE_CONCAT_VVV(CFLAGS, CONFIGURE_CFLAGS, SPECIFIED_CFLAGS) +AC_COMPILE_IFELSE([AC_LANG_PROGRAM( +[[ +]], [[ + return 0; +]])], + [je_cv_cflags_added=$1] + AC_MSG_RESULT([yes]), + [je_cv_cflags_added=] + AC_MSG_RESULT([no]) + [CONFIGURE_CFLAGS="${T_CONFIGURE_CFLAGS}"] +) +JE_CONCAT_VVV(CFLAGS, CONFIGURE_CFLAGS, SPECIFIED_CFLAGS) +]) + +dnl JE_CFLAGS_SAVE() +dnl JE_CFLAGS_RESTORE() +dnl +dnl Save/restore CFLAGS. Nesting is not supported. +AC_DEFUN([JE_CFLAGS_SAVE], +SAVED_CONFIGURE_CFLAGS="${CONFIGURE_CFLAGS}" +) +AC_DEFUN([JE_CFLAGS_RESTORE], +CONFIGURE_CFLAGS="${SAVED_CONFIGURE_CFLAGS}" +JE_CONCAT_VVV(CFLAGS, CONFIGURE_CFLAGS, SPECIFIED_CFLAGS) +) + +CONFIGURE_CXXFLAGS= +SPECIFIED_CXXFLAGS="${CXXFLAGS}" +dnl JE_CXXFLAGS_ADD(cxxflag) +AC_DEFUN([JE_CXXFLAGS_ADD], +[ +AC_MSG_CHECKING([whether compiler supports $1]) +T_CONFIGURE_CXXFLAGS="${CONFIGURE_CXXFLAGS}" +JE_APPEND_VS(CONFIGURE_CXXFLAGS, $1) +JE_CONCAT_VVV(CXXFLAGS, CONFIGURE_CXXFLAGS, SPECIFIED_CXXFLAGS) +AC_LANG_PUSH([C++]) +AC_COMPILE_IFELSE([AC_LANG_PROGRAM( +[[ +]], [[ + return 0; +]])], + [je_cv_cxxflags_added=$1] + AC_MSG_RESULT([yes]), + [je_cv_cxxflags_added=] + AC_MSG_RESULT([no]) + [CONFIGURE_CXXFLAGS="${T_CONFIGURE_CXXFLAGS}"] +) +AC_LANG_POP([C++]) +JE_CONCAT_VVV(CXXFLAGS, CONFIGURE_CXXFLAGS, SPECIFIED_CXXFLAGS) +]) + +dnl JE_COMPILABLE(label, hcode, mcode, rvar) +dnl +dnl Use AC_LINK_IFELSE() rather than AC_COMPILE_IFELSE() so that linker errors +dnl cause failure. +AC_DEFUN([JE_COMPILABLE], +[ +AC_CACHE_CHECK([whether $1 is compilable], + [$4], + [AC_LINK_IFELSE([AC_LANG_PROGRAM([$2], + [$3])], + [$4=yes], + [$4=no])]) +]) + +dnl ============================================================================ + +CONFIG=`echo ${ac_configure_args} | sed -e 's#'"'"'\([^ ]*\)'"'"'#\1#g'` +AC_SUBST([CONFIG]) + +dnl Library revision. +rev=2 +AC_SUBST([rev]) + +srcroot=$srcdir +if test "x${srcroot}" = "x." ; then + srcroot="" +else + srcroot="${srcroot}/" +fi +AC_SUBST([srcroot]) +abs_srcroot="`cd \"${srcdir}\"; pwd`/" +AC_SUBST([abs_srcroot]) + +objroot="" +AC_SUBST([objroot]) +abs_objroot="`pwd`/" +AC_SUBST([abs_objroot]) + +dnl Munge install path variables. +if test "x$prefix" = "xNONE" ; then + prefix="/usr/local" +fi +if test "x$exec_prefix" = "xNONE" ; then + exec_prefix=$prefix +fi +PREFIX=$prefix +AC_SUBST([PREFIX]) +BINDIR=`eval echo $bindir` +BINDIR=`eval echo $BINDIR` +AC_SUBST([BINDIR]) +INCLUDEDIR=`eval echo $includedir` +INCLUDEDIR=`eval echo $INCLUDEDIR` +AC_SUBST([INCLUDEDIR]) +LIBDIR=`eval echo $libdir` +LIBDIR=`eval echo $LIBDIR` +AC_SUBST([LIBDIR]) +DATADIR=`eval echo $datadir` +DATADIR=`eval echo $DATADIR` +AC_SUBST([DATADIR]) +MANDIR=`eval echo $mandir` +MANDIR=`eval echo $MANDIR` +AC_SUBST([MANDIR]) + +dnl Support for building documentation. +AC_PATH_PROG([XSLTPROC], [xsltproc], [false], [$PATH]) +if test -d "/usr/share/xml/docbook/stylesheet/docbook-xsl" ; then + DEFAULT_XSLROOT="/usr/share/xml/docbook/stylesheet/docbook-xsl" +elif test -d "/usr/share/sgml/docbook/xsl-stylesheets" ; then + DEFAULT_XSLROOT="/usr/share/sgml/docbook/xsl-stylesheets" +else + dnl Documentation building will fail if this default gets used. + DEFAULT_XSLROOT="" +fi +AC_ARG_WITH([xslroot], + [AS_HELP_STRING([--with-xslroot=<path>], [XSL stylesheet root path])], [ +if test "x$with_xslroot" = "xno" ; then + XSLROOT="${DEFAULT_XSLROOT}" +else + XSLROOT="${with_xslroot}" +fi +], + XSLROOT="${DEFAULT_XSLROOT}" +) +if test "x$XSLTPROC" = "xfalse" ; then + XSLROOT="" +fi +AC_SUBST([XSLROOT]) + +dnl If CFLAGS isn't defined, set CFLAGS to something reasonable. Otherwise, +dnl just prevent autoconf from molesting CFLAGS. +CFLAGS=$CFLAGS +AC_PROG_CC + +if test "x$GCC" != "xyes" ; then + AC_CACHE_CHECK([whether compiler is MSVC], + [je_cv_msvc], + [AC_COMPILE_IFELSE([AC_LANG_PROGRAM([], + [ +#ifndef _MSC_VER + int fail[-1]; +#endif +])], + [je_cv_msvc=yes], + [je_cv_msvc=no])]) +fi + +dnl check if a cray prgenv wrapper compiler is being used +je_cv_cray_prgenv_wrapper="" +if test "x${PE_ENV}" != "x" ; then + case "${CC}" in + CC|cc) + je_cv_cray_prgenv_wrapper="yes" + ;; + *) + ;; + esac +fi + +AC_CACHE_CHECK([whether compiler is cray], + [je_cv_cray], + [AC_COMPILE_IFELSE([AC_LANG_PROGRAM([], + [ +#ifndef _CRAYC + int fail[-1]; +#endif +])], + [je_cv_cray=yes], + [je_cv_cray=no])]) + +if test "x${je_cv_cray}" = "xyes" ; then + AC_CACHE_CHECK([whether cray compiler version is 8.4], + [je_cv_cray_84], + [AC_COMPILE_IFELSE([AC_LANG_PROGRAM([], + [ +#if !(_RELEASE_MAJOR == 8 && _RELEASE_MINOR == 4) + int fail[-1]; +#endif +])], + [je_cv_cray_84=yes], + [je_cv_cray_84=no])]) +fi + +if test "x$GCC" = "xyes" ; then + JE_CFLAGS_ADD([-std=gnu11]) + if test "x$je_cv_cflags_added" = "x-std=gnu11" ; then + AC_DEFINE_UNQUOTED([JEMALLOC_HAS_RESTRICT]) + else + JE_CFLAGS_ADD([-std=gnu99]) + if test "x$je_cv_cflags_added" = "x-std=gnu99" ; then + AC_DEFINE_UNQUOTED([JEMALLOC_HAS_RESTRICT]) + fi + fi + JE_CFLAGS_ADD([-Wall]) + JE_CFLAGS_ADD([-Wextra]) + JE_CFLAGS_ADD([-Wshorten-64-to-32]) + JE_CFLAGS_ADD([-Wsign-compare]) + JE_CFLAGS_ADD([-Wundef]) + JE_CFLAGS_ADD([-Wno-format-zero-length]) + JE_CFLAGS_ADD([-pipe]) + JE_CFLAGS_ADD([-g3]) +elif test "x$je_cv_msvc" = "xyes" ; then + CC="$CC -nologo" + JE_CFLAGS_ADD([-Zi]) + JE_CFLAGS_ADD([-MT]) + JE_CFLAGS_ADD([-W3]) + JE_CFLAGS_ADD([-FS]) + JE_APPEND_VS(CPPFLAGS, -I${srcdir}/include/msvc_compat) +fi +if test "x$je_cv_cray" = "xyes" ; then + dnl cray compiler 8.4 has an inlining bug + if test "x$je_cv_cray_84" = "xyes" ; then + JE_CFLAGS_ADD([-hipa2]) + JE_CFLAGS_ADD([-hnognu]) + fi + dnl ignore unreachable code warning + JE_CFLAGS_ADD([-hnomessage=128]) + dnl ignore redefinition of "malloc", "free", etc warning + JE_CFLAGS_ADD([-hnomessage=1357]) +fi +AC_SUBST([CONFIGURE_CFLAGS]) +AC_SUBST([SPECIFIED_CFLAGS]) +AC_SUBST([EXTRA_CFLAGS]) +AC_PROG_CPP + +AC_ARG_ENABLE([cxx], + [AS_HELP_STRING([--disable-cxx], [Disable C++ integration])], +if test "x$enable_cxx" = "xno" ; then + enable_cxx="0" +else + enable_cxx="1" +fi +, +enable_cxx="1" +) +if test "x$enable_cxx" = "x1" ; then + dnl Require at least c++14, which is the first version to support sized + dnl deallocation. C++ support is not compiled otherwise. + m4_include([m4/ax_cxx_compile_stdcxx.m4]) + AX_CXX_COMPILE_STDCXX([14], [noext], [optional]) + if test "x${HAVE_CXX14}" = "x1" ; then + JE_CXXFLAGS_ADD([-Wall]) + JE_CXXFLAGS_ADD([-Wextra]) + JE_CXXFLAGS_ADD([-g3]) + + SAVED_LIBS="${LIBS}" + JE_APPEND_VS(LIBS, -lstdc++) + JE_COMPILABLE([libstdc++ linkage], [ +#include <stdlib.h> +], [[ + int *arr = (int *)malloc(sizeof(int) * 42); + if (arr == NULL) + return 1; +]], [je_cv_libstdcxx]) + if test "x${je_cv_libstdcxx}" = "xno" ; then + LIBS="${SAVED_LIBS}" + fi + else + enable_cxx="0" + fi +fi +AC_SUBST([enable_cxx]) +AC_SUBST([CONFIGURE_CXXFLAGS]) +AC_SUBST([SPECIFIED_CXXFLAGS]) +AC_SUBST([EXTRA_CXXFLAGS]) + +AC_C_BIGENDIAN([ac_cv_big_endian=1], [ac_cv_big_endian=0]) +if test "x${ac_cv_big_endian}" = "x1" ; then + AC_DEFINE_UNQUOTED([JEMALLOC_BIG_ENDIAN], [ ]) +fi + +if test "x${je_cv_msvc}" = "xyes" -a "x${ac_cv_header_inttypes_h}" = "xno"; then + JE_APPEND_VS(CPPFLAGS, -I${srcdir}/include/msvc_compat/C99) +fi + +if test "x${je_cv_msvc}" = "xyes" ; then + LG_SIZEOF_PTR=LG_SIZEOF_PTR_WIN + AC_MSG_RESULT([Using a predefined value for sizeof(void *): 4 for 32-bit, 8 for 64-bit]) +else + AC_CHECK_SIZEOF([void *]) + if test "x${ac_cv_sizeof_void_p}" = "x8" ; then + LG_SIZEOF_PTR=3 + elif test "x${ac_cv_sizeof_void_p}" = "x4" ; then + LG_SIZEOF_PTR=2 + else + AC_MSG_ERROR([Unsupported pointer size: ${ac_cv_sizeof_void_p}]) + fi +fi +AC_DEFINE_UNQUOTED([LG_SIZEOF_PTR], [$LG_SIZEOF_PTR]) + +AC_CHECK_SIZEOF([int]) +if test "x${ac_cv_sizeof_int}" = "x8" ; then + LG_SIZEOF_INT=3 +elif test "x${ac_cv_sizeof_int}" = "x4" ; then + LG_SIZEOF_INT=2 +else + AC_MSG_ERROR([Unsupported int size: ${ac_cv_sizeof_int}]) +fi +AC_DEFINE_UNQUOTED([LG_SIZEOF_INT], [$LG_SIZEOF_INT]) + +AC_CHECK_SIZEOF([long]) +if test "x${ac_cv_sizeof_long}" = "x8" ; then + LG_SIZEOF_LONG=3 +elif test "x${ac_cv_sizeof_long}" = "x4" ; then + LG_SIZEOF_LONG=2 +else + AC_MSG_ERROR([Unsupported long size: ${ac_cv_sizeof_long}]) +fi +AC_DEFINE_UNQUOTED([LG_SIZEOF_LONG], [$LG_SIZEOF_LONG]) + +AC_CHECK_SIZEOF([long long]) +if test "x${ac_cv_sizeof_long_long}" = "x8" ; then + LG_SIZEOF_LONG_LONG=3 +elif test "x${ac_cv_sizeof_long_long}" = "x4" ; then + LG_SIZEOF_LONG_LONG=2 +else + AC_MSG_ERROR([Unsupported long long size: ${ac_cv_sizeof_long_long}]) +fi +AC_DEFINE_UNQUOTED([LG_SIZEOF_LONG_LONG], [$LG_SIZEOF_LONG_LONG]) + +AC_CHECK_SIZEOF([intmax_t]) +if test "x${ac_cv_sizeof_intmax_t}" = "x16" ; then + LG_SIZEOF_INTMAX_T=4 +elif test "x${ac_cv_sizeof_intmax_t}" = "x8" ; then + LG_SIZEOF_INTMAX_T=3 +elif test "x${ac_cv_sizeof_intmax_t}" = "x4" ; then + LG_SIZEOF_INTMAX_T=2 +else + AC_MSG_ERROR([Unsupported intmax_t size: ${ac_cv_sizeof_intmax_t}]) +fi +AC_DEFINE_UNQUOTED([LG_SIZEOF_INTMAX_T], [$LG_SIZEOF_INTMAX_T]) + +AC_CANONICAL_HOST +dnl CPU-specific settings. +CPU_SPINWAIT="" +case "${host_cpu}" in + i686|x86_64) + HAVE_CPU_SPINWAIT=1 + if test "x${je_cv_msvc}" = "xyes" ; then + AC_CACHE_VAL([je_cv_pause_msvc], + [JE_COMPILABLE([pause instruction MSVC], [], + [[_mm_pause(); return 0;]], + [je_cv_pause_msvc])]) + if test "x${je_cv_pause_msvc}" = "xyes" ; then + CPU_SPINWAIT='_mm_pause()' + fi + else + AC_CACHE_VAL([je_cv_pause], + [JE_COMPILABLE([pause instruction], [], + [[__asm__ volatile("pause"); return 0;]], + [je_cv_pause])]) + if test "x${je_cv_pause}" = "xyes" ; then + CPU_SPINWAIT='__asm__ volatile("pause")' + fi + fi + ;; + *) + HAVE_CPU_SPINWAIT=0 + ;; +esac +AC_DEFINE_UNQUOTED([HAVE_CPU_SPINWAIT], [$HAVE_CPU_SPINWAIT]) +AC_DEFINE_UNQUOTED([CPU_SPINWAIT], [$CPU_SPINWAIT]) + +AC_ARG_WITH([lg_vaddr], + [AS_HELP_STRING([--with-lg-vaddr=<lg-vaddr>], [Number of significant virtual address bits])], + [LG_VADDR="$with_lg_vaddr"], [LG_VADDR="detect"]) + +case "${host_cpu}" in + aarch64) + if test "x$LG_VADDR" = "xdetect"; then + AC_MSG_CHECKING([number of significant virtual address bits]) + if test "x${LG_SIZEOF_PTR}" = "x2" ; then + #aarch64 ILP32 + LG_VADDR=32 + else + #aarch64 LP64 + LG_VADDR=48 + fi + AC_MSG_RESULT([$LG_VADDR]) + fi + ;; + x86_64) + if test "x$LG_VADDR" = "xdetect"; then + AC_CACHE_CHECK([number of significant virtual address bits], + [je_cv_lg_vaddr], + AC_RUN_IFELSE([AC_LANG_PROGRAM( +[[ +#include <stdio.h> +#ifdef _WIN32 +#include <limits.h> +#include <intrin.h> +typedef unsigned __int32 uint32_t; +#else +#include <stdint.h> +#endif +]], [[ + uint32_t r[[4]]; + uint32_t eax_in = 0x80000008U; +#ifdef _WIN32 + __cpuid((int *)r, (int)eax_in); +#else + asm volatile ("cpuid" + : "=a" (r[[0]]), "=b" (r[[1]]), "=c" (r[[2]]), "=d" (r[[3]]) + : "a" (eax_in), "c" (0) + ); +#endif + uint32_t eax_out = r[[0]]; + uint32_t vaddr = ((eax_out & 0x0000ff00U) >> 8); + FILE *f = fopen("conftest.out", "w"); + if (f == NULL) { + return 1; + } + if (vaddr > (sizeof(void *) << 3)) { + vaddr = sizeof(void *) << 3; + } + fprintf(f, "%u", vaddr); + fclose(f); + return 0; +]])], + [je_cv_lg_vaddr=`cat conftest.out`], + [je_cv_lg_vaddr=error], + [je_cv_lg_vaddr=57])) + if test "x${je_cv_lg_vaddr}" != "x" ; then + LG_VADDR="${je_cv_lg_vaddr}" + fi + if test "x${LG_VADDR}" != "xerror" ; then + AC_DEFINE_UNQUOTED([LG_VADDR], [$LG_VADDR]) + else + AC_MSG_ERROR([cannot determine number of significant virtual address bits]) + fi + fi + ;; + *) + if test "x$LG_VADDR" = "xdetect"; then + AC_MSG_CHECKING([number of significant virtual address bits]) + if test "x${LG_SIZEOF_PTR}" = "x3" ; then + LG_VADDR=64 + elif test "x${LG_SIZEOF_PTR}" = "x2" ; then + LG_VADDR=32 + elif test "x${LG_SIZEOF_PTR}" = "xLG_SIZEOF_PTR_WIN" ; then + LG_VADDR="(1U << (LG_SIZEOF_PTR_WIN+3))" + else + AC_MSG_ERROR([Unsupported lg(pointer size): ${LG_SIZEOF_PTR}]) + fi + AC_MSG_RESULT([$LG_VADDR]) + fi + ;; +esac +AC_DEFINE_UNQUOTED([LG_VADDR], [$LG_VADDR]) + +LD_PRELOAD_VAR="LD_PRELOAD" +so="so" +importlib="${so}" +o="$ac_objext" +a="a" +exe="$ac_exeext" +libprefix="lib" +link_whole_archive="0" +DSO_LDFLAGS='-shared -Wl,-soname,$(@F)' +RPATH='-Wl,-rpath,$(1)' +SOREV="${so}.${rev}" +PIC_CFLAGS='-fPIC -DPIC' +CTARGET='-o $@' +LDTARGET='-o $@' +TEST_LD_MODE= +EXTRA_LDFLAGS= +ARFLAGS='crus' +AROUT=' $@' +CC_MM=1 + +if test "x$je_cv_cray_prgenv_wrapper" = "xyes" ; then + TEST_LD_MODE='-dynamic' +fi + +if test "x${je_cv_cray}" = "xyes" ; then + CC_MM= +fi + +AN_MAKEVAR([AR], [AC_PROG_AR]) +AN_PROGRAM([ar], [AC_PROG_AR]) +AC_DEFUN([AC_PROG_AR], [AC_CHECK_TOOL(AR, ar, :)]) +AC_PROG_AR + +AN_MAKEVAR([NM], [AC_PROG_NM]) +AN_PROGRAM([nm], [AC_PROG_NM]) +AC_DEFUN([AC_PROG_NM], [AC_CHECK_TOOL(NM, nm, :)]) +AC_PROG_NM + +AC_PROG_AWK + +dnl ============================================================================ +dnl jemalloc version. +dnl + +AC_ARG_WITH([version], + [AS_HELP_STRING([--with-version=<major>.<minor>.<bugfix>-<nrev>-g<gid>], + [Version string])], + [ + echo "${with_version}" | grep ['^[0-9]\+\.[0-9]\+\.[0-9]\+-[0-9]\+-g[0-9a-f]\+$'] 2>&1 1>/dev/null + if test $? -eq 0 ; then + echo "$with_version" > "${objroot}VERSION" + else + echo "${with_version}" | grep ['^VERSION$'] 2>&1 1>/dev/null + if test $? -ne 0 ; then + AC_MSG_ERROR([${with_version} does not match <major>.<minor>.<bugfix>-<nrev>-g<gid> or VERSION]) + fi + fi + ], [ + dnl Set VERSION if source directory is inside a git repository. + if test "x`test ! \"${srcroot}\" && cd \"${srcroot}\"; git rev-parse --is-inside-work-tree 2>/dev/null`" = "xtrue" ; then + dnl Pattern globs aren't powerful enough to match both single- and + dnl double-digit version numbers, so iterate over patterns to support up + dnl to version 99.99.99 without any accidental matches. + for pattern in ['[0-9].[0-9].[0-9]' '[0-9].[0-9].[0-9][0-9]' \ + '[0-9].[0-9][0-9].[0-9]' '[0-9].[0-9][0-9].[0-9][0-9]' \ + '[0-9][0-9].[0-9].[0-9]' '[0-9][0-9].[0-9].[0-9][0-9]' \ + '[0-9][0-9].[0-9][0-9].[0-9]' \ + '[0-9][0-9].[0-9][0-9].[0-9][0-9]']; do + (test ! "${srcroot}" && cd "${srcroot}"; git describe --long --abbrev=40 --match="${pattern}") > "${objroot}VERSION.tmp" 2>/dev/null + if test $? -eq 0 ; then + mv "${objroot}VERSION.tmp" "${objroot}VERSION" + break + fi + done + fi + rm -f "${objroot}VERSION.tmp" + ]) + +if test ! -e "${objroot}VERSION" ; then + if test ! -e "${srcroot}VERSION" ; then + AC_MSG_RESULT( + [Missing VERSION file, and unable to generate it; creating bogus VERSION]) + echo "0.0.0-0-g0000000000000000000000000000000000000000" > "${objroot}VERSION" + else + cp ${srcroot}VERSION ${objroot}VERSION + fi +fi +jemalloc_version=`cat "${objroot}VERSION"` +jemalloc_version_major=`echo ${jemalloc_version} | tr ".g-" " " | awk '{print [$]1}'` +jemalloc_version_minor=`echo ${jemalloc_version} | tr ".g-" " " | awk '{print [$]2}'` +jemalloc_version_bugfix=`echo ${jemalloc_version} | tr ".g-" " " | awk '{print [$]3}'` +jemalloc_version_nrev=`echo ${jemalloc_version} | tr ".g-" " " | awk '{print [$]4}'` +jemalloc_version_gid=`echo ${jemalloc_version} | tr ".g-" " " | awk '{print [$]5}'` +AC_SUBST([jemalloc_version]) +AC_SUBST([jemalloc_version_major]) +AC_SUBST([jemalloc_version_minor]) +AC_SUBST([jemalloc_version_bugfix]) +AC_SUBST([jemalloc_version_nrev]) +AC_SUBST([jemalloc_version_gid]) + +dnl Platform-specific settings. abi and RPATH can probably be determined +dnl programmatically, but doing so is error-prone, which makes it generally +dnl not worth the trouble. +dnl +dnl Define cpp macros in CPPFLAGS, rather than doing AC_DEFINE(macro), since the +dnl definitions need to be seen before any headers are included, which is a pain +dnl to make happen otherwise. +default_retain="0" +maps_coalesce="1" +DUMP_SYMS="${NM} -a" +SYM_PREFIX="" +case "${host}" in + *-*-darwin* | *-*-ios*) + abi="macho" + RPATH="" + LD_PRELOAD_VAR="DYLD_INSERT_LIBRARIES" + so="dylib" + importlib="${so}" + force_tls="0" + DSO_LDFLAGS='-shared -Wl,-install_name,$(LIBDIR)/$(@F)' + SOREV="${rev}.${so}" + sbrk_deprecated="1" + SYM_PREFIX="_" + ;; + *-*-freebsd*) + abi="elf" + AC_DEFINE([JEMALLOC_SYSCTL_VM_OVERCOMMIT], [ ]) + force_lazy_lock="1" + ;; + *-*-dragonfly*) + abi="elf" + ;; + *-*-openbsd*) + abi="elf" + force_tls="0" + ;; + *-*-bitrig*) + abi="elf" + ;; + *-*-linux-android) + dnl syscall(2) and secure_getenv(3) are exposed by _GNU_SOURCE. + JE_APPEND_VS(CPPFLAGS, -D_GNU_SOURCE) + abi="elf" + AC_DEFINE([JEMALLOC_PURGE_MADVISE_DONTNEED_ZEROS], [ ]) + AC_DEFINE([JEMALLOC_HAS_ALLOCA_H]) + AC_DEFINE([JEMALLOC_PROC_SYS_VM_OVERCOMMIT_MEMORY], [ ]) + AC_DEFINE([JEMALLOC_THREADED_INIT], [ ]) + AC_DEFINE([JEMALLOC_C11_ATOMICS]) + force_tls="0" + if test "${LG_SIZEOF_PTR}" = "3"; then + default_retain="1" + fi + ;; + *-*-linux*) + dnl syscall(2) and secure_getenv(3) are exposed by _GNU_SOURCE. + JE_APPEND_VS(CPPFLAGS, -D_GNU_SOURCE) + abi="elf" + AC_DEFINE([JEMALLOC_PURGE_MADVISE_DONTNEED_ZEROS], [ ]) + AC_DEFINE([JEMALLOC_HAS_ALLOCA_H]) + AC_DEFINE([JEMALLOC_PROC_SYS_VM_OVERCOMMIT_MEMORY], [ ]) + AC_DEFINE([JEMALLOC_THREADED_INIT], [ ]) + AC_DEFINE([JEMALLOC_USE_CXX_THROW], [ ]) + if test "${LG_SIZEOF_PTR}" = "3"; then + default_retain="1" + fi + ;; + *-*-kfreebsd*) + dnl syscall(2) and secure_getenv(3) are exposed by _GNU_SOURCE. + JE_APPEND_VS(CPPFLAGS, -D_GNU_SOURCE) + abi="elf" + AC_DEFINE([JEMALLOC_HAS_ALLOCA_H]) + AC_DEFINE([JEMALLOC_SYSCTL_VM_OVERCOMMIT], [ ]) + AC_DEFINE([JEMALLOC_THREADED_INIT], [ ]) + AC_DEFINE([JEMALLOC_USE_CXX_THROW], [ ]) + ;; + *-*-netbsd*) + AC_MSG_CHECKING([ABI]) + AC_COMPILE_IFELSE([AC_LANG_PROGRAM( +[[#ifdef __ELF__ +/* ELF */ +#else +#error aout +#endif +]])], + [abi="elf"], + [abi="aout"]) + AC_MSG_RESULT([$abi]) + ;; + *-*-solaris2*) + abi="elf" + RPATH='-Wl,-R,$(1)' + dnl Solaris needs this for sigwait(). + JE_APPEND_VS(CPPFLAGS, -D_POSIX_PTHREAD_SEMANTICS) + JE_APPEND_VS(LIBS, -lposix4 -lsocket -lnsl) + ;; + *-ibm-aix*) + if test "${LG_SIZEOF_PTR}" = "3"; then + dnl 64bit AIX + LD_PRELOAD_VAR="LDR_PRELOAD64" + else + dnl 32bit AIX + LD_PRELOAD_VAR="LDR_PRELOAD" + fi + abi="xcoff" + ;; + *-*-mingw* | *-*-cygwin*) + abi="pecoff" + force_tls="0" + maps_coalesce="0" + RPATH="" + so="dll" + if test "x$je_cv_msvc" = "xyes" ; then + importlib="lib" + DSO_LDFLAGS="-LD" + EXTRA_LDFLAGS="-link -DEBUG" + CTARGET='-Fo$@' + LDTARGET='-Fe$@' + AR='lib' + ARFLAGS='-nologo -out:' + AROUT='$@' + CC_MM= + else + importlib="${so}" + DSO_LDFLAGS="-shared" + link_whole_archive="1" + fi + case "${host}" in + *-*-cygwin*) + DUMP_SYMS="dumpbin /SYMBOLS" + ;; + *) + ;; + esac + a="lib" + libprefix="" + SOREV="${so}" + PIC_CFLAGS="" + if test "${LG_SIZEOF_PTR}" = "3"; then + default_retain="1" + fi + ;; + *) + AC_MSG_RESULT([Unsupported operating system: ${host}]) + abi="elf" + ;; +esac + +JEMALLOC_USABLE_SIZE_CONST=const +AC_CHECK_HEADERS([malloc.h], [ + AC_MSG_CHECKING([whether malloc_usable_size definition can use const argument]) + AC_COMPILE_IFELSE([AC_LANG_PROGRAM( + [#include <malloc.h> + #include <stddef.h> + size_t malloc_usable_size(const void *ptr); + ], + [])],[ + AC_MSG_RESULT([yes]) + ],[ + JEMALLOC_USABLE_SIZE_CONST= + AC_MSG_RESULT([no]) + ]) +]) +AC_DEFINE_UNQUOTED([JEMALLOC_USABLE_SIZE_CONST], [$JEMALLOC_USABLE_SIZE_CONST]) +AC_SUBST([abi]) +AC_SUBST([RPATH]) +AC_SUBST([LD_PRELOAD_VAR]) +AC_SUBST([so]) +AC_SUBST([importlib]) +AC_SUBST([o]) +AC_SUBST([a]) +AC_SUBST([exe]) +AC_SUBST([libprefix]) +AC_SUBST([link_whole_archive]) +AC_SUBST([DSO_LDFLAGS]) +AC_SUBST([EXTRA_LDFLAGS]) +AC_SUBST([SOREV]) +AC_SUBST([PIC_CFLAGS]) +AC_SUBST([CTARGET]) +AC_SUBST([LDTARGET]) +AC_SUBST([TEST_LD_MODE]) +AC_SUBST([MKLIB]) +AC_SUBST([ARFLAGS]) +AC_SUBST([AROUT]) +AC_SUBST([DUMP_SYMS]) +AC_SUBST([CC_MM]) + +dnl Determine whether libm must be linked to use e.g. log(3). +AC_SEARCH_LIBS([log], [m], , [AC_MSG_ERROR([Missing math functions])]) +if test "x$ac_cv_search_log" != "xnone required" ; then + LM="$ac_cv_search_log" +else + LM= +fi +AC_SUBST(LM) + +JE_COMPILABLE([__attribute__ syntax], + [static __attribute__((unused)) void foo(void){}], + [], + [je_cv_attribute]) +if test "x${je_cv_attribute}" = "xyes" ; then + AC_DEFINE([JEMALLOC_HAVE_ATTR], [ ]) + if test "x${GCC}" = "xyes" -a "x${abi}" = "xelf"; then + JE_CFLAGS_ADD([-fvisibility=hidden]) + JE_CXXFLAGS_ADD([-fvisibility=hidden]) + fi +fi +dnl Check for tls_model attribute support (clang 3.0 still lacks support). +JE_CFLAGS_SAVE() +JE_CFLAGS_ADD([-Werror]) +JE_CFLAGS_ADD([-herror_on_warning]) +JE_COMPILABLE([tls_model attribute], [], + [static __thread int + __attribute__((tls_model("initial-exec"), unused)) foo; + foo = 0;], + [je_cv_tls_model]) +JE_CFLAGS_RESTORE() +dnl (Setting of JEMALLOC_TLS_MODEL is done later, after we've checked for +dnl --disable-initial-exec-tls) + +dnl Check for alloc_size attribute support. +JE_CFLAGS_SAVE() +JE_CFLAGS_ADD([-Werror]) +JE_CFLAGS_ADD([-herror_on_warning]) +JE_COMPILABLE([alloc_size attribute], [#include <stdlib.h>], + [void *foo(size_t size) __attribute__((alloc_size(1)));], + [je_cv_alloc_size]) +JE_CFLAGS_RESTORE() +if test "x${je_cv_alloc_size}" = "xyes" ; then + AC_DEFINE([JEMALLOC_HAVE_ATTR_ALLOC_SIZE], [ ]) +fi +dnl Check for format(gnu_printf, ...) attribute support. +JE_CFLAGS_SAVE() +JE_CFLAGS_ADD([-Werror]) +JE_CFLAGS_ADD([-herror_on_warning]) +JE_COMPILABLE([format(gnu_printf, ...) attribute], [#include <stdlib.h>], + [void *foo(const char *format, ...) __attribute__((format(gnu_printf, 1, 2)));], + [je_cv_format_gnu_printf]) +JE_CFLAGS_RESTORE() +if test "x${je_cv_format_gnu_printf}" = "xyes" ; then + AC_DEFINE([JEMALLOC_HAVE_ATTR_FORMAT_GNU_PRINTF], [ ]) +fi +dnl Check for format(printf, ...) attribute support. +JE_CFLAGS_SAVE() +JE_CFLAGS_ADD([-Werror]) +JE_CFLAGS_ADD([-herror_on_warning]) +JE_COMPILABLE([format(printf, ...) attribute], [#include <stdlib.h>], + [void *foo(const char *format, ...) __attribute__((format(printf, 1, 2)));], + [je_cv_format_printf]) +JE_CFLAGS_RESTORE() +if test "x${je_cv_format_printf}" = "xyes" ; then + AC_DEFINE([JEMALLOC_HAVE_ATTR_FORMAT_PRINTF], [ ]) +fi + +dnl Check for format_arg(...) attribute support. +JE_CFLAGS_SAVE() +JE_CFLAGS_ADD([-Werror]) +JE_CFLAGS_ADD([-herror_on_warning]) +JE_COMPILABLE([format(printf, ...) attribute], [#include <stdlib.h>], + [const char * __attribute__((__format_arg__(1))) foo(const char *format);], + [je_cv_format_arg]) +JE_CFLAGS_RESTORE() +if test "x${je_cv_format_arg}" = "xyes" ; then + AC_DEFINE([JEMALLOC_HAVE_ATTR_FORMAT_ARG], [ ]) +fi + +dnl Support optional additions to rpath. +AC_ARG_WITH([rpath], + [AS_HELP_STRING([--with-rpath=<rpath>], [Colon-separated rpath (ELF systems only)])], +if test "x$with_rpath" = "xno" ; then + RPATH_EXTRA= +else + RPATH_EXTRA="`echo $with_rpath | tr \":\" \" \"`" +fi, + RPATH_EXTRA= +) +AC_SUBST([RPATH_EXTRA]) + +dnl Disable rules that do automatic regeneration of configure output by default. +AC_ARG_ENABLE([autogen], + [AS_HELP_STRING([--enable-autogen], [Automatically regenerate configure output])], +if test "x$enable_autogen" = "xno" ; then + enable_autogen="0" +else + enable_autogen="1" +fi +, +enable_autogen="0" +) +AC_SUBST([enable_autogen]) + +AC_PROG_INSTALL +AC_PROG_RANLIB +AC_PATH_PROG([LD], [ld], [false], [$PATH]) +AC_PATH_PROG([AUTOCONF], [autoconf], [false], [$PATH]) + +dnl Enable documentation +AC_ARG_ENABLE([doc], + [AS_HELP_STRING([--enable-documentation], [Build documentation])], +if test "x$enable_doc" = "xno" ; then + enable_doc="0" +else + enable_doc="1" +fi +, +enable_doc="1" +) +AC_SUBST([enable_doc]) + +dnl Enable shared libs +AC_ARG_ENABLE([shared], + [AS_HELP_STRING([--enable-shared], [Build shared libaries])], +if test "x$enable_shared" = "xno" ; then + enable_shared="0" +else + enable_shared="1" +fi +, +enable_shared="1" +) +AC_SUBST([enable_shared]) + +dnl Enable static libs +AC_ARG_ENABLE([static], + [AS_HELP_STRING([--enable-static], [Build static libaries])], +if test "x$enable_static" = "xno" ; then + enable_static="0" +else + enable_static="1" +fi +, +enable_static="1" +) +AC_SUBST([enable_static]) + +if test "$enable_shared$enable_static" = "00" ; then + AC_MSG_ERROR([Please enable one of shared or static builds]) +fi + +dnl Perform no name mangling by default. +AC_ARG_WITH([mangling], + [AS_HELP_STRING([--with-mangling=<map>], [Mangle symbols in <map>])], + [mangling_map="$with_mangling"], [mangling_map=""]) + +dnl Do not prefix public APIs by default. +AC_ARG_WITH([jemalloc_prefix], + [AS_HELP_STRING([--with-jemalloc-prefix=<prefix>], [Prefix to prepend to all public APIs])], + [JEMALLOC_PREFIX="$with_jemalloc_prefix"], + [if test "x$abi" != "xmacho" -a "x$abi" != "xpecoff"; then + JEMALLOC_PREFIX="" +else + JEMALLOC_PREFIX="je_" +fi] +) +if test "x$JEMALLOC_PREFIX" = "x" ; then + AC_DEFINE([JEMALLOC_IS_MALLOC]) +else + JEMALLOC_CPREFIX=`echo ${JEMALLOC_PREFIX} | tr "a-z" "A-Z"` + AC_DEFINE_UNQUOTED([JEMALLOC_PREFIX], ["$JEMALLOC_PREFIX"]) + AC_DEFINE_UNQUOTED([JEMALLOC_CPREFIX], ["$JEMALLOC_CPREFIX"]) +fi +AC_SUBST([JEMALLOC_PREFIX]) +AC_SUBST([JEMALLOC_CPREFIX]) + +AC_ARG_WITH([export], + [AS_HELP_STRING([--without-export], [disable exporting jemalloc public APIs])], + [if test "x$with_export" = "xno"; then + AC_DEFINE([JEMALLOC_EXPORT],[]) +fi] +) + +public_syms="aligned_alloc calloc dallocx free mallctl mallctlbymib mallctlnametomib malloc malloc_conf malloc_message malloc_stats_print malloc_usable_size mallocx smallocx_${jemalloc_version_gid} nallocx posix_memalign rallocx realloc sallocx sdallocx xallocx" +dnl Check for additional platform-specific public API functions. +AC_CHECK_FUNC([memalign], + [AC_DEFINE([JEMALLOC_OVERRIDE_MEMALIGN], [ ]) + public_syms="${public_syms} memalign"]) +AC_CHECK_FUNC([valloc], + [AC_DEFINE([JEMALLOC_OVERRIDE_VALLOC], [ ]) + public_syms="${public_syms} valloc"]) + +dnl Check for allocator-related functions that should be wrapped. +wrap_syms= +if test "x${JEMALLOC_PREFIX}" = "x" ; then + AC_CHECK_FUNC([__libc_calloc], + [AC_DEFINE([JEMALLOC_OVERRIDE___LIBC_CALLOC], [ ]) + wrap_syms="${wrap_syms} __libc_calloc"]) + AC_CHECK_FUNC([__libc_free], + [AC_DEFINE([JEMALLOC_OVERRIDE___LIBC_FREE], [ ]) + wrap_syms="${wrap_syms} __libc_free"]) + AC_CHECK_FUNC([__libc_malloc], + [AC_DEFINE([JEMALLOC_OVERRIDE___LIBC_MALLOC], [ ]) + wrap_syms="${wrap_syms} __libc_malloc"]) + AC_CHECK_FUNC([__libc_memalign], + [AC_DEFINE([JEMALLOC_OVERRIDE___LIBC_MEMALIGN], [ ]) + wrap_syms="${wrap_syms} __libc_memalign"]) + AC_CHECK_FUNC([__libc_realloc], + [AC_DEFINE([JEMALLOC_OVERRIDE___LIBC_REALLOC], [ ]) + wrap_syms="${wrap_syms} __libc_realloc"]) + AC_CHECK_FUNC([__libc_valloc], + [AC_DEFINE([JEMALLOC_OVERRIDE___LIBC_VALLOC], [ ]) + wrap_syms="${wrap_syms} __libc_valloc"]) + AC_CHECK_FUNC([__posix_memalign], + [AC_DEFINE([JEMALLOC_OVERRIDE___POSIX_MEMALIGN], [ ]) + wrap_syms="${wrap_syms} __posix_memalign"]) +fi + +case "${host}" in + *-*-mingw* | *-*-cygwin*) + wrap_syms="${wrap_syms} tls_callback" + ;; + *) + ;; +esac + +dnl Mangle library-private APIs. +AC_ARG_WITH([private_namespace], + [AS_HELP_STRING([--with-private-namespace=<prefix>], [Prefix to prepend to all library-private APIs])], + [JEMALLOC_PRIVATE_NAMESPACE="${with_private_namespace}je_"], + [JEMALLOC_PRIVATE_NAMESPACE="je_"] +) +AC_DEFINE_UNQUOTED([JEMALLOC_PRIVATE_NAMESPACE], [$JEMALLOC_PRIVATE_NAMESPACE]) +private_namespace="$JEMALLOC_PRIVATE_NAMESPACE" +AC_SUBST([private_namespace]) + +dnl Do not add suffix to installed files by default. +AC_ARG_WITH([install_suffix], + [AS_HELP_STRING([--with-install-suffix=<suffix>], [Suffix to append to all installed files])], + [INSTALL_SUFFIX="$with_install_suffix"], + [INSTALL_SUFFIX=] +) +install_suffix="$INSTALL_SUFFIX" +AC_SUBST([install_suffix]) + +dnl Specify default malloc_conf. +AC_ARG_WITH([malloc_conf], + [AS_HELP_STRING([--with-malloc-conf=<malloc_conf>], [config.malloc_conf options string])], + [JEMALLOC_CONFIG_MALLOC_CONF="$with_malloc_conf"], + [JEMALLOC_CONFIG_MALLOC_CONF=""] +) +config_malloc_conf="$JEMALLOC_CONFIG_MALLOC_CONF" +AC_DEFINE_UNQUOTED([JEMALLOC_CONFIG_MALLOC_CONF], ["$config_malloc_conf"]) + +dnl Substitute @je_@ in jemalloc_protos.h.in, primarily to make generation of +dnl jemalloc_protos_jet.h easy. +je_="je_" +AC_SUBST([je_]) + +cfgoutputs_in="Makefile.in" +cfgoutputs_in="${cfgoutputs_in} jemalloc.pc.in" +cfgoutputs_in="${cfgoutputs_in} doc/html.xsl.in" +cfgoutputs_in="${cfgoutputs_in} doc/manpages.xsl.in" +cfgoutputs_in="${cfgoutputs_in} doc/jemalloc.xml.in" +cfgoutputs_in="${cfgoutputs_in} include/jemalloc/jemalloc_macros.h.in" +cfgoutputs_in="${cfgoutputs_in} include/jemalloc/jemalloc_protos.h.in" +cfgoutputs_in="${cfgoutputs_in} include/jemalloc/jemalloc_typedefs.h.in" +cfgoutputs_in="${cfgoutputs_in} include/jemalloc/internal/jemalloc_preamble.h.in" +cfgoutputs_in="${cfgoutputs_in} test/test.sh.in" +cfgoutputs_in="${cfgoutputs_in} test/include/test/jemalloc_test.h.in" + +cfgoutputs_out="Makefile" +cfgoutputs_out="${cfgoutputs_out} jemalloc.pc" +cfgoutputs_out="${cfgoutputs_out} doc/html.xsl" +cfgoutputs_out="${cfgoutputs_out} doc/manpages.xsl" +cfgoutputs_out="${cfgoutputs_out} doc/jemalloc.xml" +cfgoutputs_out="${cfgoutputs_out} include/jemalloc/jemalloc_macros.h" +cfgoutputs_out="${cfgoutputs_out} include/jemalloc/jemalloc_protos.h" +cfgoutputs_out="${cfgoutputs_out} include/jemalloc/jemalloc_typedefs.h" +cfgoutputs_out="${cfgoutputs_out} include/jemalloc/internal/jemalloc_preamble.h" +cfgoutputs_out="${cfgoutputs_out} test/test.sh" +cfgoutputs_out="${cfgoutputs_out} test/include/test/jemalloc_test.h" + +cfgoutputs_tup="Makefile" +cfgoutputs_tup="${cfgoutputs_tup} jemalloc.pc:jemalloc.pc.in" +cfgoutputs_tup="${cfgoutputs_tup} doc/html.xsl:doc/html.xsl.in" +cfgoutputs_tup="${cfgoutputs_tup} doc/manpages.xsl:doc/manpages.xsl.in" +cfgoutputs_tup="${cfgoutputs_tup} doc/jemalloc.xml:doc/jemalloc.xml.in" +cfgoutputs_tup="${cfgoutputs_tup} include/jemalloc/jemalloc_macros.h:include/jemalloc/jemalloc_macros.h.in" +cfgoutputs_tup="${cfgoutputs_tup} include/jemalloc/jemalloc_protos.h:include/jemalloc/jemalloc_protos.h.in" +cfgoutputs_tup="${cfgoutputs_tup} include/jemalloc/jemalloc_typedefs.h:include/jemalloc/jemalloc_typedefs.h.in" +cfgoutputs_tup="${cfgoutputs_tup} include/jemalloc/internal/jemalloc_preamble.h" +cfgoutputs_tup="${cfgoutputs_tup} test/test.sh:test/test.sh.in" +cfgoutputs_tup="${cfgoutputs_tup} test/include/test/jemalloc_test.h:test/include/test/jemalloc_test.h.in" + +cfghdrs_in="include/jemalloc/jemalloc_defs.h.in" +cfghdrs_in="${cfghdrs_in} include/jemalloc/internal/jemalloc_internal_defs.h.in" +cfghdrs_in="${cfghdrs_in} include/jemalloc/internal/private_symbols.sh" +cfghdrs_in="${cfghdrs_in} include/jemalloc/internal/private_namespace.sh" +cfghdrs_in="${cfghdrs_in} include/jemalloc/internal/public_namespace.sh" +cfghdrs_in="${cfghdrs_in} include/jemalloc/internal/public_unnamespace.sh" +cfghdrs_in="${cfghdrs_in} include/jemalloc/jemalloc_rename.sh" +cfghdrs_in="${cfghdrs_in} include/jemalloc/jemalloc_mangle.sh" +cfghdrs_in="${cfghdrs_in} include/jemalloc/jemalloc.sh" +cfghdrs_in="${cfghdrs_in} test/include/test/jemalloc_test_defs.h.in" + +cfghdrs_out="include/jemalloc/jemalloc_defs.h" +cfghdrs_out="${cfghdrs_out} include/jemalloc/jemalloc${install_suffix}.h" +cfghdrs_out="${cfghdrs_out} include/jemalloc/internal/private_symbols.awk" +cfghdrs_out="${cfghdrs_out} include/jemalloc/internal/private_symbols_jet.awk" +cfghdrs_out="${cfghdrs_out} include/jemalloc/internal/public_symbols.txt" +cfghdrs_out="${cfghdrs_out} include/jemalloc/internal/public_namespace.h" +cfghdrs_out="${cfghdrs_out} include/jemalloc/internal/public_unnamespace.h" +cfghdrs_out="${cfghdrs_out} include/jemalloc/jemalloc_protos_jet.h" +cfghdrs_out="${cfghdrs_out} include/jemalloc/jemalloc_rename.h" +cfghdrs_out="${cfghdrs_out} include/jemalloc/jemalloc_mangle.h" +cfghdrs_out="${cfghdrs_out} include/jemalloc/jemalloc_mangle_jet.h" +cfghdrs_out="${cfghdrs_out} include/jemalloc/internal/jemalloc_internal_defs.h" +cfghdrs_out="${cfghdrs_out} test/include/test/jemalloc_test_defs.h" + +cfghdrs_tup="include/jemalloc/jemalloc_defs.h:include/jemalloc/jemalloc_defs.h.in" +cfghdrs_tup="${cfghdrs_tup} include/jemalloc/internal/jemalloc_internal_defs.h:include/jemalloc/internal/jemalloc_internal_defs.h.in" +cfghdrs_tup="${cfghdrs_tup} test/include/test/jemalloc_test_defs.h:test/include/test/jemalloc_test_defs.h.in" + +dnl ============================================================================ +dnl jemalloc build options. +dnl + +dnl Do not compile with debugging by default. +AC_ARG_ENABLE([debug], + [AS_HELP_STRING([--enable-debug], + [Build debugging code])], +[if test "x$enable_debug" = "xno" ; then + enable_debug="0" +else + enable_debug="1" +fi +], +[enable_debug="0"] +) +if test "x$enable_debug" = "x1" ; then + AC_DEFINE([JEMALLOC_DEBUG], [ ]) +fi +if test "x$enable_debug" = "x1" ; then + AC_DEFINE([JEMALLOC_DEBUG], [ ]) +fi +AC_SUBST([enable_debug]) + +dnl Only optimize if not debugging. +if test "x$enable_debug" = "x0" ; then + if test "x$GCC" = "xyes" ; then + JE_CFLAGS_ADD([-O3]) + JE_CXXFLAGS_ADD([-O3]) + JE_CFLAGS_ADD([-funroll-loops]) + elif test "x$je_cv_msvc" = "xyes" ; then + JE_CFLAGS_ADD([-O2]) + JE_CXXFLAGS_ADD([-O2]) + else + JE_CFLAGS_ADD([-O]) + JE_CXXFLAGS_ADD([-O]) + fi +fi + +dnl Enable statistics calculation by default. +AC_ARG_ENABLE([stats], + [AS_HELP_STRING([--disable-stats], + [Disable statistics calculation/reporting])], +[if test "x$enable_stats" = "xno" ; then + enable_stats="0" +else + enable_stats="1" +fi +], +[enable_stats="1"] +) +if test "x$enable_stats" = "x1" ; then + AC_DEFINE([JEMALLOC_STATS], [ ]) +fi +AC_SUBST([enable_stats]) + +dnl Do not enable smallocx by default. +AC_ARG_ENABLE([experimental_smallocx], + [AS_HELP_STRING([--enable-experimental-smallocx], [Enable experimental smallocx API])], +[if test "x$enable_experimental_smallocx" = "xno" ; then +enable_experimental_smallocx="0" +else +enable_experimental_smallocx="1" +fi +], +[enable_experimental_smallocx="0"] +) +if test "x$enable_experimental_smallocx" = "x1" ; then + AC_DEFINE([JEMALLOC_EXPERIMENTAL_SMALLOCX_API]) +fi +AC_SUBST([enable_experimental_smallocx]) + +dnl Do not enable profiling by default. +AC_ARG_ENABLE([prof], + [AS_HELP_STRING([--enable-prof], [Enable allocation profiling])], +[if test "x$enable_prof" = "xno" ; then + enable_prof="0" +else + enable_prof="1" +fi +], +[enable_prof="0"] +) +if test "x$enable_prof" = "x1" ; then + backtrace_method="" +else + backtrace_method="N/A" +fi + +AC_ARG_ENABLE([prof-libunwind], + [AS_HELP_STRING([--enable-prof-libunwind], [Use libunwind for backtracing])], +[if test "x$enable_prof_libunwind" = "xno" ; then + enable_prof_libunwind="0" +else + enable_prof_libunwind="1" +fi +], +[enable_prof_libunwind="0"] +) +AC_ARG_WITH([static_libunwind], + [AS_HELP_STRING([--with-static-libunwind=<libunwind.a>], + [Path to static libunwind library; use rather than dynamically linking])], +if test "x$with_static_libunwind" = "xno" ; then + LUNWIND="-lunwind" +else + if test ! -f "$with_static_libunwind" ; then + AC_MSG_ERROR([Static libunwind not found: $with_static_libunwind]) + fi + LUNWIND="$with_static_libunwind" +fi, + LUNWIND="-lunwind" +) +if test "x$backtrace_method" = "x" -a "x$enable_prof_libunwind" = "x1" ; then + AC_CHECK_HEADERS([libunwind.h], , [enable_prof_libunwind="0"]) + if test "x$LUNWIND" = "x-lunwind" ; then + AC_CHECK_LIB([unwind], [unw_backtrace], [JE_APPEND_VS(LIBS, $LUNWIND)], + [enable_prof_libunwind="0"]) + else + JE_APPEND_VS(LIBS, $LUNWIND) + fi + if test "x${enable_prof_libunwind}" = "x1" ; then + backtrace_method="libunwind" + AC_DEFINE([JEMALLOC_PROF_LIBUNWIND], [ ]) + fi +fi + +AC_ARG_ENABLE([prof-libgcc], + [AS_HELP_STRING([--disable-prof-libgcc], + [Do not use libgcc for backtracing])], +[if test "x$enable_prof_libgcc" = "xno" ; then + enable_prof_libgcc="0" +else + enable_prof_libgcc="1" +fi +], +[enable_prof_libgcc="1"] +) +if test "x$backtrace_method" = "x" -a "x$enable_prof_libgcc" = "x1" \ + -a "x$GCC" = "xyes" ; then + AC_CHECK_HEADERS([unwind.h], , [enable_prof_libgcc="0"]) + if test "x${enable_prof_libgcc}" = "x1" ; then + AC_CHECK_LIB([gcc], [_Unwind_Backtrace], [JE_APPEND_VS(LIBS, -lgcc)], [enable_prof_libgcc="0"]) + fi + if test "x${enable_prof_libgcc}" = "x1" ; then + backtrace_method="libgcc" + AC_DEFINE([JEMALLOC_PROF_LIBGCC], [ ]) + fi +else + enable_prof_libgcc="0" +fi + +AC_ARG_ENABLE([prof-gcc], + [AS_HELP_STRING([--disable-prof-gcc], + [Do not use gcc intrinsics for backtracing])], +[if test "x$enable_prof_gcc" = "xno" ; then + enable_prof_gcc="0" +else + enable_prof_gcc="1" +fi +], +[enable_prof_gcc="1"] +) +if test "x$backtrace_method" = "x" -a "x$enable_prof_gcc" = "x1" \ + -a "x$GCC" = "xyes" ; then + JE_CFLAGS_ADD([-fno-omit-frame-pointer]) + backtrace_method="gcc intrinsics" + AC_DEFINE([JEMALLOC_PROF_GCC], [ ]) +else + enable_prof_gcc="0" +fi + +if test "x$backtrace_method" = "x" ; then + backtrace_method="none (disabling profiling)" + enable_prof="0" +fi +AC_MSG_CHECKING([configured backtracing method]) +AC_MSG_RESULT([$backtrace_method]) +if test "x$enable_prof" = "x1" ; then + dnl Heap profiling uses the log(3) function. + JE_APPEND_VS(LIBS, $LM) + + AC_DEFINE([JEMALLOC_PROF], [ ]) +fi +AC_SUBST([enable_prof]) + +dnl Indicate whether adjacent virtual memory mappings automatically coalesce +dnl (and fragment on demand). +if test "x${maps_coalesce}" = "x1" ; then + AC_DEFINE([JEMALLOC_MAPS_COALESCE], [ ]) +fi + +dnl Indicate whether to retain memory (rather than using munmap()) by default. +if test "x$default_retain" = "x1" ; then + AC_DEFINE([JEMALLOC_RETAIN], [ ]) +fi + +dnl Enable allocation from DSS if supported by the OS. +have_dss="1" +dnl Check whether the BSD/SUSv1 sbrk() exists. If not, disable DSS support. +AC_CHECK_FUNC([sbrk], [have_sbrk="1"], [have_sbrk="0"]) +if test "x$have_sbrk" = "x1" ; then + if test "x$sbrk_deprecated" = "x1" ; then + AC_MSG_RESULT([Disabling dss allocation because sbrk is deprecated]) + have_dss="0" + fi +else + have_dss="0" +fi + +if test "x$have_dss" = "x1" ; then + AC_DEFINE([JEMALLOC_DSS], [ ]) +fi + +dnl Support the junk/zero filling option by default. +AC_ARG_ENABLE([fill], + [AS_HELP_STRING([--disable-fill], [Disable support for junk/zero filling])], +[if test "x$enable_fill" = "xno" ; then + enable_fill="0" +else + enable_fill="1" +fi +], +[enable_fill="1"] +) +if test "x$enable_fill" = "x1" ; then + AC_DEFINE([JEMALLOC_FILL], [ ]) +fi +AC_SUBST([enable_fill]) + +dnl Disable utrace(2)-based tracing by default. +AC_ARG_ENABLE([utrace], + [AS_HELP_STRING([--enable-utrace], [Enable utrace(2)-based tracing])], +[if test "x$enable_utrace" = "xno" ; then + enable_utrace="0" +else + enable_utrace="1" +fi +], +[enable_utrace="0"] +) +JE_COMPILABLE([utrace(2)], [ +#include <sys/types.h> +#include <sys/param.h> +#include <sys/time.h> +#include <sys/uio.h> +#include <sys/ktrace.h> +], [ + utrace((void *)0, 0); +], [je_cv_utrace]) +if test "x${je_cv_utrace}" = "xno" ; then + enable_utrace="0" +fi +if test "x$enable_utrace" = "x1" ; then + AC_DEFINE([JEMALLOC_UTRACE], [ ]) +fi +AC_SUBST([enable_utrace]) + +dnl Do not support the xmalloc option by default. +AC_ARG_ENABLE([xmalloc], + [AS_HELP_STRING([--enable-xmalloc], [Support xmalloc option])], +[if test "x$enable_xmalloc" = "xno" ; then + enable_xmalloc="0" +else + enable_xmalloc="1" +fi +], +[enable_xmalloc="0"] +) +if test "x$enable_xmalloc" = "x1" ; then + AC_DEFINE([JEMALLOC_XMALLOC], [ ]) +fi +AC_SUBST([enable_xmalloc]) + +dnl Support cache-oblivious allocation alignment by default. +AC_ARG_ENABLE([cache-oblivious], + [AS_HELP_STRING([--disable-cache-oblivious], + [Disable support for cache-oblivious allocation alignment])], +[if test "x$enable_cache_oblivious" = "xno" ; then + enable_cache_oblivious="0" +else + enable_cache_oblivious="1" +fi +], +[enable_cache_oblivious="1"] +) +if test "x$enable_cache_oblivious" = "x1" ; then + AC_DEFINE([JEMALLOC_CACHE_OBLIVIOUS], [ ]) +fi +AC_SUBST([enable_cache_oblivious]) + +dnl Do not log by default. +AC_ARG_ENABLE([log], + [AS_HELP_STRING([--enable-log], [Support debug logging])], +[if test "x$enable_log" = "xno" ; then + enable_log="0" +else + enable_log="1" +fi +], +[enable_log="0"] +) +if test "x$enable_log" = "x1" ; then + AC_DEFINE([JEMALLOC_LOG], [ ]) +fi +AC_SUBST([enable_log]) + +dnl Do not use readlinkat by default +AC_ARG_ENABLE([readlinkat], + [AS_HELP_STRING([--enable-readlinkat], [Use readlinkat over readlink])], +[if test "x$enable_readlinkat" = "xno" ; then + enable_readlinkat="0" +else + enable_readlinkat="1" +fi +], +[enable_readlinkat="0"] +) +if test "x$enable_readlinkat" = "x1" ; then + AC_DEFINE([JEMALLOC_READLINKAT], [ ]) +fi +AC_SUBST([enable_readlinkat]) + +dnl Avoid extra safety checks by default +AC_ARG_ENABLE([opt-safety-checks], + [AS_HELP_STRING([--enable-opt-safety-checks], + [Perform certain low-overhead checks, even in opt mode])], +[if test "x$enable_opt_safety_checks" = "xno" ; then + enable_opt_safety_checks="0" +else + enable_opt_safety_checks="1" +fi +], +[enable_opt_safety_checks="0"] +) +if test "x$enable_opt_safety_checks" = "x1" ; then + AC_DEFINE([JEMALLOC_OPT_SAFETY_CHECKS], [ ]) +fi +AC_SUBST([enable_opt_safety_checks]) + +JE_COMPILABLE([a program using __builtin_unreachable], [ +void foo (void) { + __builtin_unreachable(); +} +], [ + { + foo(); + } +], [je_cv_gcc_builtin_unreachable]) +if test "x${je_cv_gcc_builtin_unreachable}" = "xyes" ; then + AC_DEFINE([JEMALLOC_INTERNAL_UNREACHABLE], [__builtin_unreachable]) +else + AC_DEFINE([JEMALLOC_INTERNAL_UNREACHABLE], [abort]) +fi + +dnl ============================================================================ +dnl Check for __builtin_ffsl(), then ffsl(3), and fail if neither are found. +dnl One of those two functions should (theoretically) exist on all platforms +dnl that jemalloc currently has a chance of functioning on without modification. +dnl We additionally assume ffs[ll]() or __builtin_ffs[ll]() are defined if +dnl ffsl() or __builtin_ffsl() are defined, respectively. +JE_COMPILABLE([a program using __builtin_ffsl], [ +#include <stdio.h> +#include <strings.h> +#include <string.h> +], [ + { + int rv = __builtin_ffsl(0x08); + printf("%d\n", rv); + } +], [je_cv_gcc_builtin_ffsl]) +if test "x${je_cv_gcc_builtin_ffsl}" = "xyes" ; then + AC_DEFINE([JEMALLOC_INTERNAL_FFSLL], [__builtin_ffsll]) + AC_DEFINE([JEMALLOC_INTERNAL_FFSL], [__builtin_ffsl]) + AC_DEFINE([JEMALLOC_INTERNAL_FFS], [__builtin_ffs]) +else + JE_COMPILABLE([a program using ffsl], [ + #include <stdio.h> + #include <strings.h> + #include <string.h> + ], [ + { + int rv = ffsl(0x08); + printf("%d\n", rv); + } + ], [je_cv_function_ffsl]) + if test "x${je_cv_function_ffsl}" = "xyes" ; then + AC_DEFINE([JEMALLOC_INTERNAL_FFSLL], [ffsll]) + AC_DEFINE([JEMALLOC_INTERNAL_FFSL], [ffsl]) + AC_DEFINE([JEMALLOC_INTERNAL_FFS], [ffs]) + else + AC_MSG_ERROR([Cannot build without ffsl(3) or __builtin_ffsl()]) + fi +fi + +JE_COMPILABLE([a program using __builtin_popcountl], [ +#include <stdio.h> +#include <strings.h> +#include <string.h> +], [ + { + int rv = __builtin_popcountl(0x08); + printf("%d\n", rv); + } +], [je_cv_gcc_builtin_popcountl]) +if test "x${je_cv_gcc_builtin_popcountl}" = "xyes" ; then + AC_DEFINE([JEMALLOC_INTERNAL_POPCOUNT], [__builtin_popcount]) + AC_DEFINE([JEMALLOC_INTERNAL_POPCOUNTL], [__builtin_popcountl]) +fi + +AC_ARG_WITH([lg_quantum], + [AS_HELP_STRING([--with-lg-quantum=<lg-quantum>], + [Base 2 log of minimum allocation alignment])], + [LG_QUANTA="$with_lg_quantum"], + [LG_QUANTA="3 4"]) +if test "x$with_lg_quantum" != "x" ; then + AC_DEFINE_UNQUOTED([LG_QUANTUM], [$with_lg_quantum]) +fi + +AC_ARG_WITH([lg_page], + [AS_HELP_STRING([--with-lg-page=<lg-page>], [Base 2 log of system page size])], + [LG_PAGE="$with_lg_page"], [LG_PAGE="detect"]) +if test "x$LG_PAGE" = "xdetect"; then + AC_CACHE_CHECK([LG_PAGE], + [je_cv_lg_page], + AC_RUN_IFELSE([AC_LANG_PROGRAM( +[[ +#include <strings.h> +#ifdef _WIN32 +#include <windows.h> +#else +#include <unistd.h> +#endif +#include <stdio.h> +]], +[[ + int result; + FILE *f; + +#ifdef _WIN32 + SYSTEM_INFO si; + GetSystemInfo(&si); + result = si.dwPageSize; +#else + result = sysconf(_SC_PAGESIZE); +#endif + if (result == -1) { + return 1; + } + result = JEMALLOC_INTERNAL_FFSL(result) - 1; + + f = fopen("conftest.out", "w"); + if (f == NULL) { + return 1; + } + fprintf(f, "%d", result); + fclose(f); + + return 0; +]])], + [je_cv_lg_page=`cat conftest.out`], + [je_cv_lg_page=undefined], + [je_cv_lg_page=12])) +fi +if test "x${je_cv_lg_page}" != "x" ; then + LG_PAGE="${je_cv_lg_page}" +fi +if test "x${LG_PAGE}" != "xundefined" ; then + AC_DEFINE_UNQUOTED([LG_PAGE], [$LG_PAGE]) +else + AC_MSG_ERROR([cannot determine value for LG_PAGE]) +fi + +AC_ARG_WITH([lg_hugepage], + [AS_HELP_STRING([--with-lg-hugepage=<lg-hugepage>], + [Base 2 log of system huge page size])], + [je_cv_lg_hugepage="${with_lg_hugepage}"], + [je_cv_lg_hugepage=""]) +if test "x${je_cv_lg_hugepage}" = "x" ; then + dnl Look in /proc/meminfo (Linux-specific) for information on the default huge + dnl page size, if any. The relevant line looks like: + dnl + dnl Hugepagesize: 2048 kB + if test -e "/proc/meminfo" ; then + hpsk=[`cat /proc/meminfo 2>/dev/null | \ + grep -e '^Hugepagesize:[[:space:]]\+[0-9]\+[[:space:]]kB$' | \ + awk '{print $2}'`] + if test "x${hpsk}" != "x" ; then + je_cv_lg_hugepage=10 + while test "${hpsk}" -gt 1 ; do + hpsk="$((hpsk / 2))" + je_cv_lg_hugepage="$((je_cv_lg_hugepage + 1))" + done + fi + fi + + dnl Set default if unable to automatically configure. + if test "x${je_cv_lg_hugepage}" = "x" ; then + je_cv_lg_hugepage=21 + fi +fi +if test "x${LG_PAGE}" != "xundefined" -a \ + "${je_cv_lg_hugepage}" -lt "${LG_PAGE}" ; then + AC_MSG_ERROR([Huge page size (2^${je_cv_lg_hugepage}) must be at least page size (2^${LG_PAGE})]) +fi +AC_DEFINE_UNQUOTED([LG_HUGEPAGE], [${je_cv_lg_hugepage}]) + +dnl ============================================================================ +dnl Enable libdl by default. +AC_ARG_ENABLE([libdl], + [AS_HELP_STRING([--disable-libdl], + [Do not use libdl])], +[if test "x$enable_libdl" = "xno" ; then + enable_libdl="0" +else + enable_libdl="1" +fi +], +[enable_libdl="1"] +) +AC_SUBST([libdl]) + +dnl ============================================================================ +dnl Configure pthreads. + +if test "x$abi" != "xpecoff" ; then + AC_DEFINE([JEMALLOC_HAVE_PTHREAD], [ ]) + AC_CHECK_HEADERS([pthread.h], , [AC_MSG_ERROR([pthread.h is missing])]) + dnl Some systems may embed pthreads functionality in libc; check for libpthread + dnl first, but try libc too before failing. + AC_CHECK_LIB([pthread], [pthread_create], [JE_APPEND_VS(LIBS, -pthread)], + [AC_SEARCH_LIBS([pthread_create], , , + AC_MSG_ERROR([libpthread is missing]))]) + wrap_syms="${wrap_syms} pthread_create" + have_pthread="1" + +dnl Check if we have dlsym support. + if test "x$enable_libdl" = "x1" ; then + have_dlsym="1" + AC_CHECK_HEADERS([dlfcn.h], + AC_CHECK_FUNC([dlsym], [], + [AC_CHECK_LIB([dl], [dlsym], [LIBS="$LIBS -ldl"], [have_dlsym="0"])]), + [have_dlsym="0"]) + if test "x$have_dlsym" = "x1" ; then + AC_DEFINE([JEMALLOC_HAVE_DLSYM], [ ]) + fi + else + have_dlsym="0" + fi + + JE_COMPILABLE([pthread_atfork(3)], [ +#include <pthread.h> +], [ + pthread_atfork((void *)0, (void *)0, (void *)0); +], [je_cv_pthread_atfork]) + if test "x${je_cv_pthread_atfork}" = "xyes" ; then + AC_DEFINE([JEMALLOC_HAVE_PTHREAD_ATFORK], [ ]) + fi + dnl Check if pthread_setname_np is available with the expected API. + JE_COMPILABLE([pthread_setname_np(3)], [ +#include <pthread.h> +], [ + pthread_setname_np(pthread_self(), "setname_test"); +], [je_cv_pthread_setname_np]) + if test "x${je_cv_pthread_setname_np}" = "xyes" ; then + AC_DEFINE([JEMALLOC_HAVE_PTHREAD_SETNAME_NP], [ ]) + fi +fi + +JE_APPEND_VS(CPPFLAGS, -D_REENTRANT) + +dnl Check whether clock_gettime(2) is in libc or librt. +AC_SEARCH_LIBS([clock_gettime], [rt]) + +dnl Cray wrapper compiler often adds `-lrt` when using `-static`. Check with +dnl `-dynamic` as well in case a user tries to dynamically link in jemalloc +if test "x$je_cv_cray_prgenv_wrapper" = "xyes" ; then + if test "$ac_cv_search_clock_gettime" != "-lrt"; then + JE_CFLAGS_SAVE() + + unset ac_cv_search_clock_gettime + JE_CFLAGS_ADD([-dynamic]) + AC_SEARCH_LIBS([clock_gettime], [rt]) + + JE_CFLAGS_RESTORE() + fi +fi + +dnl check for CLOCK_MONOTONIC_COARSE (Linux-specific). +JE_COMPILABLE([clock_gettime(CLOCK_MONOTONIC_COARSE, ...)], [ +#include <time.h> +], [ + struct timespec ts; + + clock_gettime(CLOCK_MONOTONIC_COARSE, &ts); +], [je_cv_clock_monotonic_coarse]) +if test "x${je_cv_clock_monotonic_coarse}" = "xyes" ; then + AC_DEFINE([JEMALLOC_HAVE_CLOCK_MONOTONIC_COARSE]) +fi + +dnl check for CLOCK_MONOTONIC. +JE_COMPILABLE([clock_gettime(CLOCK_MONOTONIC, ...)], [ +#include <unistd.h> +#include <time.h> +], [ + struct timespec ts; + + clock_gettime(CLOCK_MONOTONIC, &ts); +#if !defined(_POSIX_MONOTONIC_CLOCK) || _POSIX_MONOTONIC_CLOCK < 0 +# error _POSIX_MONOTONIC_CLOCK missing/invalid +#endif +], [je_cv_clock_monotonic]) +if test "x${je_cv_clock_monotonic}" = "xyes" ; then + AC_DEFINE([JEMALLOC_HAVE_CLOCK_MONOTONIC]) +fi + +dnl Check for mach_absolute_time(). +JE_COMPILABLE([mach_absolute_time()], [ +#include <mach/mach_time.h> +], [ + mach_absolute_time(); +], [je_cv_mach_absolute_time]) +if test "x${je_cv_mach_absolute_time}" = "xyes" ; then + AC_DEFINE([JEMALLOC_HAVE_MACH_ABSOLUTE_TIME]) +fi + +dnl Use syscall(2) (if available) by default. +AC_ARG_ENABLE([syscall], + [AS_HELP_STRING([--disable-syscall], [Disable use of syscall(2)])], +[if test "x$enable_syscall" = "xno" ; then + enable_syscall="0" +else + enable_syscall="1" +fi +], +[enable_syscall="1"] +) +if test "x$enable_syscall" = "x1" ; then + dnl Check if syscall(2) is usable. Treat warnings as errors, so that e.g. OS + dnl X 10.12's deprecation warning prevents use. + JE_CFLAGS_SAVE() + JE_CFLAGS_ADD([-Werror]) + JE_COMPILABLE([syscall(2)], [ +#include <sys/syscall.h> +#include <unistd.h> +], [ + syscall(SYS_write, 2, "hello", 5); +], + [je_cv_syscall]) + JE_CFLAGS_RESTORE() + if test "x$je_cv_syscall" = "xyes" ; then + AC_DEFINE([JEMALLOC_USE_SYSCALL], [ ]) + fi +fi + +dnl Check if the GNU-specific secure_getenv function exists. +AC_CHECK_FUNC([secure_getenv], + [have_secure_getenv="1"], + [have_secure_getenv="0"] + ) +if test "x$have_secure_getenv" = "x1" ; then + AC_DEFINE([JEMALLOC_HAVE_SECURE_GETENV], [ ]) +fi + +dnl Check if the GNU-specific sched_getcpu function exists. +AC_CHECK_FUNC([sched_getcpu], + [have_sched_getcpu="1"], + [have_sched_getcpu="0"] + ) +if test "x$have_sched_getcpu" = "x1" ; then + AC_DEFINE([JEMALLOC_HAVE_SCHED_GETCPU], [ ]) +fi + +dnl Check if the GNU-specific sched_setaffinity function exists. +AC_CHECK_FUNC([sched_setaffinity], + [have_sched_setaffinity="1"], + [have_sched_setaffinity="0"] + ) +if test "x$have_sched_setaffinity" = "x1" ; then + AC_DEFINE([JEMALLOC_HAVE_SCHED_SETAFFINITY], [ ]) +fi + +dnl Check if the Solaris/BSD issetugid function exists. +AC_CHECK_FUNC([issetugid], + [have_issetugid="1"], + [have_issetugid="0"] + ) +if test "x$have_issetugid" = "x1" ; then + AC_DEFINE([JEMALLOC_HAVE_ISSETUGID], [ ]) +fi + +dnl Check whether the BSD-specific _malloc_thread_cleanup() exists. If so, use +dnl it rather than pthreads TSD cleanup functions to support cleanup during +dnl thread exit, in order to avoid pthreads library recursion during +dnl bootstrapping. +AC_CHECK_FUNC([_malloc_thread_cleanup], + [have__malloc_thread_cleanup="1"], + [have__malloc_thread_cleanup="0"] + ) +if test "x$have__malloc_thread_cleanup" = "x1" ; then + AC_DEFINE([JEMALLOC_MALLOC_THREAD_CLEANUP], [ ]) + wrap_syms="${wrap_syms} _malloc_thread_cleanup" + force_tls="1" +fi + +dnl Check whether the BSD-specific _pthread_mutex_init_calloc_cb() exists. If +dnl so, mutex initialization causes allocation, and we need to implement this +dnl callback function in order to prevent recursive allocation. +AC_CHECK_FUNC([_pthread_mutex_init_calloc_cb], + [have__pthread_mutex_init_calloc_cb="1"], + [have__pthread_mutex_init_calloc_cb="0"] + ) +if test "x$have__pthread_mutex_init_calloc_cb" = "x1" ; then + AC_DEFINE([JEMALLOC_MUTEX_INIT_CB]) + wrap_syms="${wrap_syms} _malloc_prefork _malloc_postfork" +fi + +dnl Disable lazy locking by default. +AC_ARG_ENABLE([lazy_lock], + [AS_HELP_STRING([--enable-lazy-lock], + [Enable lazy locking (only lock when multi-threaded)])], +[if test "x$enable_lazy_lock" = "xno" ; then + enable_lazy_lock="0" +else + enable_lazy_lock="1" +fi +], +[enable_lazy_lock=""] +) +if test "x${enable_lazy_lock}" = "x" ; then + if test "x${force_lazy_lock}" = "x1" ; then + AC_MSG_RESULT([Forcing lazy-lock to avoid allocator/threading bootstrap issues]) + enable_lazy_lock="1" + else + enable_lazy_lock="0" + fi +fi +if test "x${enable_lazy_lock}" = "x1" -a "x${abi}" = "xpecoff" ; then + AC_MSG_RESULT([Forcing no lazy-lock because thread creation monitoring is unimplemented]) + enable_lazy_lock="0" +fi +if test "x$enable_lazy_lock" = "x1" ; then + if test "x$have_dlsym" = "x1" ; then + AC_DEFINE([JEMALLOC_LAZY_LOCK], [ ]) + else + AC_MSG_ERROR([Missing dlsym support: lazy-lock cannot be enabled.]) + fi +fi +AC_SUBST([enable_lazy_lock]) + +dnl Automatically configure TLS. +if test "x${force_tls}" = "x1" ; then + enable_tls="1" +elif test "x${force_tls}" = "x0" ; then + enable_tls="0" +else + enable_tls="1" +fi +if test "x${enable_tls}" = "x1" ; then +AC_MSG_CHECKING([for TLS]) +AC_COMPILE_IFELSE([AC_LANG_PROGRAM( +[[ + __thread int x; +]], [[ + x = 42; + + return 0; +]])], + AC_MSG_RESULT([yes]), + AC_MSG_RESULT([no]) + enable_tls="0") +else + enable_tls="0" +fi +AC_SUBST([enable_tls]) +if test "x${enable_tls}" = "x1" ; then + AC_DEFINE_UNQUOTED([JEMALLOC_TLS], [ ]) +fi + +dnl ============================================================================ +dnl Check for C11 atomics. + +JE_COMPILABLE([C11 atomics], [ +#include <stdint.h> +#if (__STDC_VERSION__ >= 201112L) && !defined(__STDC_NO_ATOMICS__) +#include <stdatomic.h> +#else +#error Atomics not available +#endif +], [ + uint64_t *p = (uint64_t *)0; + uint64_t x = 1; + volatile atomic_uint_least64_t *a = (volatile atomic_uint_least64_t *)p; + uint64_t r = atomic_fetch_add(a, x) + x; + return r == 0; +], [je_cv_c11_atomics]) +if test "x${je_cv_c11_atomics}" = "xyes" ; then + AC_DEFINE([JEMALLOC_C11_ATOMICS]) +fi + +dnl ============================================================================ +dnl Check for GCC-style __atomic atomics. + +JE_COMPILABLE([GCC __atomic atomics], [ +], [ + int x = 0; + int val = 1; + int y = __atomic_fetch_add(&x, val, __ATOMIC_RELAXED); + int after_add = x; + return after_add == 1; +], [je_cv_gcc_atomic_atomics]) +if test "x${je_cv_gcc_atomic_atomics}" = "xyes" ; then + AC_DEFINE([JEMALLOC_GCC_ATOMIC_ATOMICS]) + + dnl check for 8-bit atomic support + JE_COMPILABLE([GCC 8-bit __atomic atomics], [ + ], [ + unsigned char x = 0; + int val = 1; + int y = __atomic_fetch_add(&x, val, __ATOMIC_RELAXED); + int after_add = (int)x; + return after_add == 1; + ], [je_cv_gcc_u8_atomic_atomics]) + if test "x${je_cv_gcc_u8_atomic_atomics}" = "xyes" ; then + AC_DEFINE([JEMALLOC_GCC_U8_ATOMIC_ATOMICS]) + fi +fi + +dnl ============================================================================ +dnl Check for GCC-style __sync atomics. + +JE_COMPILABLE([GCC __sync atomics], [ +], [ + int x = 0; + int before_add = __sync_fetch_and_add(&x, 1); + int after_add = x; + return (before_add == 0) && (after_add == 1); +], [je_cv_gcc_sync_atomics]) +if test "x${je_cv_gcc_sync_atomics}" = "xyes" ; then + AC_DEFINE([JEMALLOC_GCC_SYNC_ATOMICS]) + + dnl check for 8-bit atomic support + JE_COMPILABLE([GCC 8-bit __sync atomics], [ + ], [ + unsigned char x = 0; + int before_add = __sync_fetch_and_add(&x, 1); + int after_add = (int)x; + return (before_add == 0) && (after_add == 1); + ], [je_cv_gcc_u8_sync_atomics]) + if test "x${je_cv_gcc_u8_sync_atomics}" = "xyes" ; then + AC_DEFINE([JEMALLOC_GCC_U8_SYNC_ATOMICS]) + fi +fi + +dnl ============================================================================ +dnl Check for atomic(3) operations as provided on Darwin. +dnl We need this not for the atomic operations (which are provided above), but +dnl rather for the OS_unfair_lock type it exposes. + +JE_COMPILABLE([Darwin OSAtomic*()], [ +#include <libkern/OSAtomic.h> +#include <inttypes.h> +], [ + { + int32_t x32 = 0; + volatile int32_t *x32p = &x32; + OSAtomicAdd32(1, x32p); + } + { + int64_t x64 = 0; + volatile int64_t *x64p = &x64; + OSAtomicAdd64(1, x64p); + } +], [je_cv_osatomic]) +if test "x${je_cv_osatomic}" = "xyes" ; then + AC_DEFINE([JEMALLOC_OSATOMIC], [ ]) +fi + +dnl ============================================================================ +dnl Check for madvise(2). + +JE_COMPILABLE([madvise(2)], [ +#include <sys/mman.h> +], [ + madvise((void *)0, 0, 0); +], [je_cv_madvise]) +if test "x${je_cv_madvise}" = "xyes" ; then + AC_DEFINE([JEMALLOC_HAVE_MADVISE], [ ]) + + dnl Check for madvise(..., MADV_FREE). + JE_COMPILABLE([madvise(..., MADV_FREE)], [ +#include <sys/mman.h> +], [ + madvise((void *)0, 0, MADV_FREE); +], [je_cv_madv_free]) + if test "x${je_cv_madv_free}" = "xyes" ; then + AC_DEFINE([JEMALLOC_PURGE_MADVISE_FREE], [ ]) + elif test "x${je_cv_madvise}" = "xyes" ; then + case "${host_cpu}" in i686|x86_64) + case "${host}" in *-*-linux*) + AC_DEFINE([JEMALLOC_PURGE_MADVISE_FREE], [ ]) + AC_DEFINE([JEMALLOC_DEFINE_MADVISE_FREE], [ ]) + ;; + esac + ;; + esac + fi + + dnl Check for madvise(..., MADV_DONTNEED). + JE_COMPILABLE([madvise(..., MADV_DONTNEED)], [ +#include <sys/mman.h> +], [ + madvise((void *)0, 0, MADV_DONTNEED); +], [je_cv_madv_dontneed]) + if test "x${je_cv_madv_dontneed}" = "xyes" ; then + AC_DEFINE([JEMALLOC_PURGE_MADVISE_DONTNEED], [ ]) + fi + + dnl Check for madvise(..., MADV_DO[NT]DUMP). + JE_COMPILABLE([madvise(..., MADV_DO[[NT]]DUMP)], [ +#include <sys/mman.h> +], [ + madvise((void *)0, 0, MADV_DONTDUMP); + madvise((void *)0, 0, MADV_DODUMP); +], [je_cv_madv_dontdump]) + if test "x${je_cv_madv_dontdump}" = "xyes" ; then + AC_DEFINE([JEMALLOC_MADVISE_DONTDUMP], [ ]) + fi + + dnl Check for madvise(..., MADV_[NO]HUGEPAGE). + JE_COMPILABLE([madvise(..., MADV_[[NO]]HUGEPAGE)], [ +#include <sys/mman.h> +], [ + madvise((void *)0, 0, MADV_HUGEPAGE); + madvise((void *)0, 0, MADV_NOHUGEPAGE); +], [je_cv_thp]) +case "${host_cpu}" in + arm*) + ;; + *) + if test "x${je_cv_thp}" = "xyes" ; then + AC_DEFINE([JEMALLOC_HAVE_MADVISE_HUGE], [ ]) + fi + ;; +esac +fi + +dnl ============================================================================ +dnl Check for __builtin_clz() and __builtin_clzl(). + +AC_CACHE_CHECK([for __builtin_clz], + [je_cv_builtin_clz], + [AC_LINK_IFELSE([AC_LANG_PROGRAM([], + [ + { + unsigned x = 0; + int y = __builtin_clz(x); + } + { + unsigned long x = 0; + int y = __builtin_clzl(x); + } + ])], + [je_cv_builtin_clz=yes], + [je_cv_builtin_clz=no])]) + +if test "x${je_cv_builtin_clz}" = "xyes" ; then + AC_DEFINE([JEMALLOC_HAVE_BUILTIN_CLZ], [ ]) +fi + +dnl ============================================================================ +dnl Check for os_unfair_lock operations as provided on Darwin. + +JE_COMPILABLE([Darwin os_unfair_lock_*()], [ +#include <os/lock.h> +#include <AvailabilityMacros.h> +], [ + #if MAC_OS_X_VERSION_MIN_REQUIRED < 101200 + #error "os_unfair_lock is not supported" + #else + os_unfair_lock lock = OS_UNFAIR_LOCK_INIT; + os_unfair_lock_lock(&lock); + os_unfair_lock_unlock(&lock); + #endif +], [je_cv_os_unfair_lock]) +if test "x${je_cv_os_unfair_lock}" = "xyes" ; then + AC_DEFINE([JEMALLOC_OS_UNFAIR_LOCK], [ ]) +fi + +dnl ============================================================================ +dnl Darwin-related configuration. + +AC_ARG_ENABLE([zone-allocator], + [AS_HELP_STRING([--disable-zone-allocator], + [Disable zone allocator for Darwin])], +[if test "x$enable_zone_allocator" = "xno" ; then + enable_zone_allocator="0" +else + enable_zone_allocator="1" +fi +], +[if test "x${abi}" = "xmacho"; then + enable_zone_allocator="1" +fi +] +) +AC_SUBST([enable_zone_allocator]) + +if test "x${enable_zone_allocator}" = "x1" ; then + if test "x${abi}" != "xmacho"; then + AC_MSG_ERROR([--enable-zone-allocator is only supported on Darwin]) + fi + AC_DEFINE([JEMALLOC_ZONE], [ ]) +fi + +dnl ============================================================================ +dnl Use initial-exec TLS by default. +AC_ARG_ENABLE([initial-exec-tls], + [AS_HELP_STRING([--disable-initial-exec-tls], + [Disable the initial-exec tls model])], +[if test "x$enable_initial_exec_tls" = "xno" ; then + enable_initial_exec_tls="0" +else + enable_initial_exec_tls="1" +fi +], +[enable_initial_exec_tls="1"] +) +AC_SUBST([enable_initial_exec_tls]) + +if test "x${je_cv_tls_model}" = "xyes" -a \ + "x${enable_initial_exec_tls}" = "x1" ; then + AC_DEFINE([JEMALLOC_TLS_MODEL], + [__attribute__((tls_model("initial-exec")))]) +else + AC_DEFINE([JEMALLOC_TLS_MODEL], [ ]) +fi + +dnl ============================================================================ +dnl Enable background threads if possible. + +if test "x${have_pthread}" = "x1" -a "x${je_cv_os_unfair_lock}" != "xyes" ; then + AC_DEFINE([JEMALLOC_BACKGROUND_THREAD]) +fi + +dnl ============================================================================ +dnl Check for glibc malloc hooks + +JE_COMPILABLE([glibc malloc hook], [ +#include <stddef.h> + +extern void (* __free_hook)(void *ptr); +extern void *(* __malloc_hook)(size_t size); +extern void *(* __realloc_hook)(void *ptr, size_t size); +], [ + void *ptr = 0L; + if (__malloc_hook) ptr = __malloc_hook(1); + if (__realloc_hook) ptr = __realloc_hook(ptr, 2); + if (__free_hook && ptr) __free_hook(ptr); +], [je_cv_glibc_malloc_hook]) +if test "x${je_cv_glibc_malloc_hook}" = "xyes" ; then + if test "x${JEMALLOC_PREFIX}" = "x" ; then + AC_DEFINE([JEMALLOC_GLIBC_MALLOC_HOOK], [ ]) + wrap_syms="${wrap_syms} __free_hook __malloc_hook __realloc_hook" + fi +fi + +JE_COMPILABLE([glibc memalign hook], [ +#include <stddef.h> + +extern void *(* __memalign_hook)(size_t alignment, size_t size); +], [ + void *ptr = 0L; + if (__memalign_hook) ptr = __memalign_hook(16, 7); +], [je_cv_glibc_memalign_hook]) +if test "x${je_cv_glibc_memalign_hook}" = "xyes" ; then + if test "x${JEMALLOC_PREFIX}" = "x" ; then + AC_DEFINE([JEMALLOC_GLIBC_MEMALIGN_HOOK], [ ]) + wrap_syms="${wrap_syms} __memalign_hook" + fi +fi + +JE_COMPILABLE([pthreads adaptive mutexes], [ +#include <pthread.h> +], [ + pthread_mutexattr_t attr; + pthread_mutexattr_init(&attr); + pthread_mutexattr_settype(&attr, PTHREAD_MUTEX_ADAPTIVE_NP); + pthread_mutexattr_destroy(&attr); +], [je_cv_pthread_mutex_adaptive_np]) +if test "x${je_cv_pthread_mutex_adaptive_np}" = "xyes" ; then + AC_DEFINE([JEMALLOC_HAVE_PTHREAD_MUTEX_ADAPTIVE_NP], [ ]) +fi + +JE_CFLAGS_SAVE() +JE_CFLAGS_ADD([-D_GNU_SOURCE]) +JE_CFLAGS_ADD([-Werror]) +JE_CFLAGS_ADD([-herror_on_warning]) +JE_COMPILABLE([strerror_r returns char with gnu source], [ +#include <errno.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +], [ + char *buffer = (char *) malloc(100); + char *error = strerror_r(EINVAL, buffer, 100); + printf("%s\n", error); +], [je_cv_strerror_r_returns_char_with_gnu_source]) +JE_CFLAGS_RESTORE() +if test "x${je_cv_strerror_r_returns_char_with_gnu_source}" = "xyes" ; then + AC_DEFINE([JEMALLOC_STRERROR_R_RETURNS_CHAR_WITH_GNU_SOURCE], [ ]) +fi + +dnl ============================================================================ +dnl Check for typedefs, structures, and compiler characteristics. +AC_HEADER_STDBOOL + +dnl ============================================================================ +dnl Define commands that generate output files. + +AC_CONFIG_COMMANDS([include/jemalloc/internal/public_symbols.txt], [ + f="${objroot}include/jemalloc/internal/public_symbols.txt" + mkdir -p "${objroot}include/jemalloc/internal" + cp /dev/null "${f}" + for nm in `echo ${mangling_map} |tr ',' ' '` ; do + n=`echo ${nm} |tr ':' ' ' |awk '{print $[]1}'` + m=`echo ${nm} |tr ':' ' ' |awk '{print $[]2}'` + echo "${n}:${m}" >> "${f}" + dnl Remove name from public_syms so that it isn't redefined later. + public_syms=`for sym in ${public_syms}; do echo "${sym}"; done |grep -v "^${n}\$" |tr '\n' ' '` + done + for sym in ${public_syms} ; do + n="${sym}" + m="${JEMALLOC_PREFIX}${sym}" + echo "${n}:${m}" >> "${f}" + done +], [ + srcdir="${srcdir}" + objroot="${objroot}" + mangling_map="${mangling_map}" + public_syms="${public_syms}" + JEMALLOC_PREFIX="${JEMALLOC_PREFIX}" +]) +AC_CONFIG_COMMANDS([include/jemalloc/internal/private_symbols.awk], [ + f="${objroot}include/jemalloc/internal/private_symbols.awk" + mkdir -p "${objroot}include/jemalloc/internal" + export_syms=`for sym in ${public_syms}; do echo "${JEMALLOC_PREFIX}${sym}"; done; for sym in ${wrap_syms}; do echo "${sym}"; done;` + "${srcdir}/include/jemalloc/internal/private_symbols.sh" "${SYM_PREFIX}" ${export_syms} > "${objroot}include/jemalloc/internal/private_symbols.awk" +], [ + srcdir="${srcdir}" + objroot="${objroot}" + public_syms="${public_syms}" + wrap_syms="${wrap_syms}" + SYM_PREFIX="${SYM_PREFIX}" + JEMALLOC_PREFIX="${JEMALLOC_PREFIX}" +]) +AC_CONFIG_COMMANDS([include/jemalloc/internal/private_symbols_jet.awk], [ + f="${objroot}include/jemalloc/internal/private_symbols_jet.awk" + mkdir -p "${objroot}include/jemalloc/internal" + export_syms=`for sym in ${public_syms}; do echo "jet_${sym}"; done; for sym in ${wrap_syms}; do echo "${sym}"; done;` + "${srcdir}/include/jemalloc/internal/private_symbols.sh" "${SYM_PREFIX}" ${export_syms} > "${objroot}include/jemalloc/internal/private_symbols_jet.awk" +], [ + srcdir="${srcdir}" + objroot="${objroot}" + public_syms="${public_syms}" + wrap_syms="${wrap_syms}" + SYM_PREFIX="${SYM_PREFIX}" +]) +AC_CONFIG_COMMANDS([include/jemalloc/internal/public_namespace.h], [ + mkdir -p "${objroot}include/jemalloc/internal" + "${srcdir}/include/jemalloc/internal/public_namespace.sh" "${objroot}include/jemalloc/internal/public_symbols.txt" > "${objroot}include/jemalloc/internal/public_namespace.h" +], [ + srcdir="${srcdir}" + objroot="${objroot}" +]) +AC_CONFIG_COMMANDS([include/jemalloc/internal/public_unnamespace.h], [ + mkdir -p "${objroot}include/jemalloc/internal" + "${srcdir}/include/jemalloc/internal/public_unnamespace.sh" "${objroot}include/jemalloc/internal/public_symbols.txt" > "${objroot}include/jemalloc/internal/public_unnamespace.h" +], [ + srcdir="${srcdir}" + objroot="${objroot}" +]) +AC_CONFIG_COMMANDS([include/jemalloc/jemalloc_protos_jet.h], [ + mkdir -p "${objroot}include/jemalloc" + cat "${srcdir}/include/jemalloc/jemalloc_protos.h.in" | sed -e 's/@je_@/jet_/g' > "${objroot}include/jemalloc/jemalloc_protos_jet.h" +], [ + srcdir="${srcdir}" + objroot="${objroot}" +]) +AC_CONFIG_COMMANDS([include/jemalloc/jemalloc_rename.h], [ + mkdir -p "${objroot}include/jemalloc" + "${srcdir}/include/jemalloc/jemalloc_rename.sh" "${objroot}include/jemalloc/internal/public_symbols.txt" > "${objroot}include/jemalloc/jemalloc_rename.h" +], [ + srcdir="${srcdir}" + objroot="${objroot}" +]) +AC_CONFIG_COMMANDS([include/jemalloc/jemalloc_mangle.h], [ + mkdir -p "${objroot}include/jemalloc" + "${srcdir}/include/jemalloc/jemalloc_mangle.sh" "${objroot}include/jemalloc/internal/public_symbols.txt" je_ > "${objroot}include/jemalloc/jemalloc_mangle.h" +], [ + srcdir="${srcdir}" + objroot="${objroot}" +]) +AC_CONFIG_COMMANDS([include/jemalloc/jemalloc_mangle_jet.h], [ + mkdir -p "${objroot}include/jemalloc" + "${srcdir}/include/jemalloc/jemalloc_mangle.sh" "${objroot}include/jemalloc/internal/public_symbols.txt" jet_ > "${objroot}include/jemalloc/jemalloc_mangle_jet.h" +], [ + srcdir="${srcdir}" + objroot="${objroot}" +]) +AC_CONFIG_COMMANDS([include/jemalloc/jemalloc.h], [ + mkdir -p "${objroot}include/jemalloc" + "${srcdir}/include/jemalloc/jemalloc.sh" "${objroot}" > "${objroot}include/jemalloc/jemalloc${install_suffix}.h" +], [ + srcdir="${srcdir}" + objroot="${objroot}" + install_suffix="${install_suffix}" +]) + +dnl Process .in files. +AC_SUBST([cfghdrs_in]) +AC_SUBST([cfghdrs_out]) +AC_CONFIG_HEADERS([$cfghdrs_tup]) + +dnl ============================================================================ +dnl Generate outputs. + +AC_CONFIG_FILES([$cfgoutputs_tup config.stamp bin/jemalloc-config bin/jemalloc.sh bin/jeprof]) +AC_SUBST([cfgoutputs_in]) +AC_SUBST([cfgoutputs_out]) +AC_OUTPUT + +dnl ============================================================================ +dnl Print out the results of configuration. +AC_MSG_RESULT([===============================================================================]) +AC_MSG_RESULT([jemalloc version : ${jemalloc_version}]) +AC_MSG_RESULT([library revision : ${rev}]) +AC_MSG_RESULT([]) +AC_MSG_RESULT([CONFIG : ${CONFIG}]) +AC_MSG_RESULT([CC : ${CC}]) +AC_MSG_RESULT([CONFIGURE_CFLAGS : ${CONFIGURE_CFLAGS}]) +AC_MSG_RESULT([SPECIFIED_CFLAGS : ${SPECIFIED_CFLAGS}]) +AC_MSG_RESULT([EXTRA_CFLAGS : ${EXTRA_CFLAGS}]) +AC_MSG_RESULT([CPPFLAGS : ${CPPFLAGS}]) +AC_MSG_RESULT([CXX : ${CXX}]) +AC_MSG_RESULT([CONFIGURE_CXXFLAGS : ${CONFIGURE_CXXFLAGS}]) +AC_MSG_RESULT([SPECIFIED_CXXFLAGS : ${SPECIFIED_CXXFLAGS}]) +AC_MSG_RESULT([EXTRA_CXXFLAGS : ${EXTRA_CXXFLAGS}]) +AC_MSG_RESULT([LDFLAGS : ${LDFLAGS}]) +AC_MSG_RESULT([EXTRA_LDFLAGS : ${EXTRA_LDFLAGS}]) +AC_MSG_RESULT([DSO_LDFLAGS : ${DSO_LDFLAGS}]) +AC_MSG_RESULT([LIBS : ${LIBS}]) +AC_MSG_RESULT([RPATH_EXTRA : ${RPATH_EXTRA}]) +AC_MSG_RESULT([]) +AC_MSG_RESULT([XSLTPROC : ${XSLTPROC}]) +AC_MSG_RESULT([XSLROOT : ${XSLROOT}]) +AC_MSG_RESULT([]) +AC_MSG_RESULT([PREFIX : ${PREFIX}]) +AC_MSG_RESULT([BINDIR : ${BINDIR}]) +AC_MSG_RESULT([DATADIR : ${DATADIR}]) +AC_MSG_RESULT([INCLUDEDIR : ${INCLUDEDIR}]) +AC_MSG_RESULT([LIBDIR : ${LIBDIR}]) +AC_MSG_RESULT([MANDIR : ${MANDIR}]) +AC_MSG_RESULT([]) +AC_MSG_RESULT([srcroot : ${srcroot}]) +AC_MSG_RESULT([abs_srcroot : ${abs_srcroot}]) +AC_MSG_RESULT([objroot : ${objroot}]) +AC_MSG_RESULT([abs_objroot : ${abs_objroot}]) +AC_MSG_RESULT([]) +AC_MSG_RESULT([JEMALLOC_PREFIX : ${JEMALLOC_PREFIX}]) +AC_MSG_RESULT([JEMALLOC_PRIVATE_NAMESPACE]) +AC_MSG_RESULT([ : ${JEMALLOC_PRIVATE_NAMESPACE}]) +AC_MSG_RESULT([install_suffix : ${install_suffix}]) +AC_MSG_RESULT([malloc_conf : ${config_malloc_conf}]) +AC_MSG_RESULT([documentation : ${enable_doc}]) +AC_MSG_RESULT([shared libs : ${enable_shared}]) +AC_MSG_RESULT([static libs : ${enable_static}]) +AC_MSG_RESULT([autogen : ${enable_autogen}]) +AC_MSG_RESULT([debug : ${enable_debug}]) +AC_MSG_RESULT([stats : ${enable_stats}]) +AC_MSG_RESULT([experimetal_smallocx : ${enable_experimental_smallocx}]) +AC_MSG_RESULT([prof : ${enable_prof}]) +AC_MSG_RESULT([prof-libunwind : ${enable_prof_libunwind}]) +AC_MSG_RESULT([prof-libgcc : ${enable_prof_libgcc}]) +AC_MSG_RESULT([prof-gcc : ${enable_prof_gcc}]) +AC_MSG_RESULT([fill : ${enable_fill}]) +AC_MSG_RESULT([utrace : ${enable_utrace}]) +AC_MSG_RESULT([xmalloc : ${enable_xmalloc}]) +AC_MSG_RESULT([log : ${enable_log}]) +AC_MSG_RESULT([lazy_lock : ${enable_lazy_lock}]) +AC_MSG_RESULT([cache-oblivious : ${enable_cache_oblivious}]) +AC_MSG_RESULT([cxx : ${enable_cxx}]) +AC_MSG_RESULT([===============================================================================]) diff --git a/deps/jemalloc/doc/html.xsl.in b/deps/jemalloc/doc/html.xsl.in new file mode 100644 index 000000000..ec4fa6552 --- /dev/null +++ b/deps/jemalloc/doc/html.xsl.in @@ -0,0 +1,5 @@ +<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" version="1.0"> + <xsl:import href="@XSLROOT@/html/docbook.xsl"/> + <xsl:import href="@abs_srcroot@doc/stylesheet.xsl"/> + <xsl:output method="xml" encoding="utf-8"/> +</xsl:stylesheet> diff --git a/deps/jemalloc/doc/jemalloc.xml.in b/deps/jemalloc/doc/jemalloc.xml.in new file mode 100644 index 000000000..7fecda7cb --- /dev/null +++ b/deps/jemalloc/doc/jemalloc.xml.in @@ -0,0 +1,3513 @@ +<?xml version='1.0' encoding='UTF-8'?> +<?xml-stylesheet type="text/xsl" + href="http://docbook.sourceforge.net/release/xsl/current/manpages/docbook.xsl"?> +<!DOCTYPE refentry PUBLIC "-//OASIS//DTD DocBook XML V4.4//EN" + "http://www.oasis-open.org/docbook/xml/4.4/docbookx.dtd" [ +]> + +<refentry> + <refentryinfo> + <title>User Manual</title> + <productname>jemalloc</productname> + <releaseinfo role="version">@jemalloc_version@</releaseinfo> + <authorgroup> + <author> + <firstname>Jason</firstname> + <surname>Evans</surname> + <personblurb>Author</personblurb> + </author> + </authorgroup> + </refentryinfo> + <refmeta> + <refentrytitle>JEMALLOC</refentrytitle> + <manvolnum>3</manvolnum> + </refmeta> + <refnamediv> + <refdescriptor>jemalloc</refdescriptor> + <refname>jemalloc</refname> + <!-- Each refname causes a man page file to be created. Only if this were + the system malloc(3) implementation would these files be appropriate. + <refname>malloc</refname> + <refname>calloc</refname> + <refname>posix_memalign</refname> + <refname>aligned_alloc</refname> + <refname>realloc</refname> + <refname>free</refname> + <refname>mallocx</refname> + <refname>rallocx</refname> + <refname>xallocx</refname> + <refname>sallocx</refname> + <refname>dallocx</refname> + <refname>sdallocx</refname> + <refname>nallocx</refname> + <refname>mallctl</refname> + <refname>mallctlnametomib</refname> + <refname>mallctlbymib</refname> + <refname>malloc_stats_print</refname> + <refname>malloc_usable_size</refname> + --> + <refpurpose>general purpose memory allocation functions</refpurpose> + </refnamediv> + <refsect1 id="library"> + <title>LIBRARY</title> + <para>This manual describes jemalloc @jemalloc_version@. More information + can be found at the <ulink + url="http://jemalloc.net/">jemalloc website</ulink>.</para> + </refsect1> + <refsynopsisdiv> + <title>SYNOPSIS</title> + <funcsynopsis> + <funcsynopsisinfo>#include <<filename class="headerfile">jemalloc/jemalloc.h</filename>></funcsynopsisinfo> + <refsect2> + <title>Standard API</title> + <funcprototype> + <funcdef>void *<function>malloc</function></funcdef> + <paramdef>size_t <parameter>size</parameter></paramdef> + </funcprototype> + <funcprototype> + <funcdef>void *<function>calloc</function></funcdef> + <paramdef>size_t <parameter>number</parameter></paramdef> + <paramdef>size_t <parameter>size</parameter></paramdef> + </funcprototype> + <funcprototype> + <funcdef>int <function>posix_memalign</function></funcdef> + <paramdef>void **<parameter>ptr</parameter></paramdef> + <paramdef>size_t <parameter>alignment</parameter></paramdef> + <paramdef>size_t <parameter>size</parameter></paramdef> + </funcprototype> + <funcprototype> + <funcdef>void *<function>aligned_alloc</function></funcdef> + <paramdef>size_t <parameter>alignment</parameter></paramdef> + <paramdef>size_t <parameter>size</parameter></paramdef> + </funcprototype> + <funcprototype> + <funcdef>void *<function>realloc</function></funcdef> + <paramdef>void *<parameter>ptr</parameter></paramdef> + <paramdef>size_t <parameter>size</parameter></paramdef> + </funcprototype> + <funcprototype> + <funcdef>void <function>free</function></funcdef> + <paramdef>void *<parameter>ptr</parameter></paramdef> + </funcprototype> + </refsect2> + <refsect2> + <title>Non-standard API</title> + <funcprototype> + <funcdef>void *<function>mallocx</function></funcdef> + <paramdef>size_t <parameter>size</parameter></paramdef> + <paramdef>int <parameter>flags</parameter></paramdef> + </funcprototype> + <funcprototype> + <funcdef>void *<function>rallocx</function></funcdef> + <paramdef>void *<parameter>ptr</parameter></paramdef> + <paramdef>size_t <parameter>size</parameter></paramdef> + <paramdef>int <parameter>flags</parameter></paramdef> + </funcprototype> + <funcprototype> + <funcdef>size_t <function>xallocx</function></funcdef> + <paramdef>void *<parameter>ptr</parameter></paramdef> + <paramdef>size_t <parameter>size</parameter></paramdef> + <paramdef>size_t <parameter>extra</parameter></paramdef> + <paramdef>int <parameter>flags</parameter></paramdef> + </funcprototype> + <funcprototype> + <funcdef>size_t <function>sallocx</function></funcdef> + <paramdef>void *<parameter>ptr</parameter></paramdef> + <paramdef>int <parameter>flags</parameter></paramdef> + </funcprototype> + <funcprototype> + <funcdef>void <function>dallocx</function></funcdef> + <paramdef>void *<parameter>ptr</parameter></paramdef> + <paramdef>int <parameter>flags</parameter></paramdef> + </funcprototype> + <funcprototype> + <funcdef>void <function>sdallocx</function></funcdef> + <paramdef>void *<parameter>ptr</parameter></paramdef> + <paramdef>size_t <parameter>size</parameter></paramdef> + <paramdef>int <parameter>flags</parameter></paramdef> + </funcprototype> + <funcprototype> + <funcdef>size_t <function>nallocx</function></funcdef> + <paramdef>size_t <parameter>size</parameter></paramdef> + <paramdef>int <parameter>flags</parameter></paramdef> + </funcprototype> + <funcprototype> + <funcdef>int <function>mallctl</function></funcdef> + <paramdef>const char *<parameter>name</parameter></paramdef> + <paramdef>void *<parameter>oldp</parameter></paramdef> + <paramdef>size_t *<parameter>oldlenp</parameter></paramdef> + <paramdef>void *<parameter>newp</parameter></paramdef> + <paramdef>size_t <parameter>newlen</parameter></paramdef> + </funcprototype> + <funcprototype> + <funcdef>int <function>mallctlnametomib</function></funcdef> + <paramdef>const char *<parameter>name</parameter></paramdef> + <paramdef>size_t *<parameter>mibp</parameter></paramdef> + <paramdef>size_t *<parameter>miblenp</parameter></paramdef> + </funcprototype> + <funcprototype> + <funcdef>int <function>mallctlbymib</function></funcdef> + <paramdef>const size_t *<parameter>mib</parameter></paramdef> + <paramdef>size_t <parameter>miblen</parameter></paramdef> + <paramdef>void *<parameter>oldp</parameter></paramdef> + <paramdef>size_t *<parameter>oldlenp</parameter></paramdef> + <paramdef>void *<parameter>newp</parameter></paramdef> + <paramdef>size_t <parameter>newlen</parameter></paramdef> + </funcprototype> + <funcprototype> + <funcdef>void <function>malloc_stats_print</function></funcdef> + <paramdef>void <parameter>(*write_cb)</parameter> + <funcparams>void *, const char *</funcparams> + </paramdef> + <paramdef>void *<parameter>cbopaque</parameter></paramdef> + <paramdef>const char *<parameter>opts</parameter></paramdef> + </funcprototype> + <funcprototype> + <funcdef>size_t <function>malloc_usable_size</function></funcdef> + <paramdef>const void *<parameter>ptr</parameter></paramdef> + </funcprototype> + <funcprototype> + <funcdef>void <function>(*malloc_message)</function></funcdef> + <paramdef>void *<parameter>cbopaque</parameter></paramdef> + <paramdef>const char *<parameter>s</parameter></paramdef> + </funcprototype> + <para><type>const char *</type><varname>malloc_conf</varname>;</para> + </refsect2> + </funcsynopsis> + </refsynopsisdiv> + <refsect1 id="description"> + <title>DESCRIPTION</title> + <refsect2> + <title>Standard API</title> + + <para>The <function>malloc()</function> function allocates + <parameter>size</parameter> bytes of uninitialized memory. The allocated + space is suitably aligned (after possible pointer coercion) for storage + of any type of object.</para> + + <para>The <function>calloc()</function> function allocates + space for <parameter>number</parameter> objects, each + <parameter>size</parameter> bytes in length. The result is identical to + calling <function>malloc()</function> with an argument of + <parameter>number</parameter> * <parameter>size</parameter>, with the + exception that the allocated memory is explicitly initialized to zero + bytes.</para> + + <para>The <function>posix_memalign()</function> function + allocates <parameter>size</parameter> bytes of memory such that the + allocation's base address is a multiple of + <parameter>alignment</parameter>, and returns the allocation in the value + pointed to by <parameter>ptr</parameter>. The requested + <parameter>alignment</parameter> must be a power of 2 at least as large as + <code language="C">sizeof(<type>void *</type>)</code>.</para> + + <para>The <function>aligned_alloc()</function> function + allocates <parameter>size</parameter> bytes of memory such that the + allocation's base address is a multiple of + <parameter>alignment</parameter>. The requested + <parameter>alignment</parameter> must be a power of 2. Behavior is + undefined if <parameter>size</parameter> is not an integral multiple of + <parameter>alignment</parameter>.</para> + + <para>The <function>realloc()</function> function changes the + size of the previously allocated memory referenced by + <parameter>ptr</parameter> to <parameter>size</parameter> bytes. The + contents of the memory are unchanged up to the lesser of the new and old + sizes. If the new size is larger, the contents of the newly allocated + portion of the memory are undefined. Upon success, the memory referenced + by <parameter>ptr</parameter> is freed and a pointer to the newly + allocated memory is returned. Note that + <function>realloc()</function> may move the memory allocation, + resulting in a different return value than <parameter>ptr</parameter>. + If <parameter>ptr</parameter> is <constant>NULL</constant>, the + <function>realloc()</function> function behaves identically to + <function>malloc()</function> for the specified size.</para> + + <para>The <function>free()</function> function causes the + allocated memory referenced by <parameter>ptr</parameter> to be made + available for future allocations. If <parameter>ptr</parameter> is + <constant>NULL</constant>, no action occurs.</para> + </refsect2> + <refsect2> + <title>Non-standard API</title> + <para>The <function>mallocx()</function>, + <function>rallocx()</function>, + <function>xallocx()</function>, + <function>sallocx()</function>, + <function>dallocx()</function>, + <function>sdallocx()</function>, and + <function>nallocx()</function> functions all have a + <parameter>flags</parameter> argument that can be used to specify + options. The functions only check the options that are contextually + relevant. Use bitwise or (<code language="C">|</code>) operations to + specify one or more of the following: + <variablelist> + <varlistentry id="MALLOCX_LG_ALIGN"> + <term><constant>MALLOCX_LG_ALIGN(<parameter>la</parameter>) + </constant></term> + + <listitem><para>Align the memory allocation to start at an address + that is a multiple of <code language="C">(1 << + <parameter>la</parameter>)</code>. This macro does not validate + that <parameter>la</parameter> is within the valid + range.</para></listitem> + </varlistentry> + <varlistentry id="MALLOCX_ALIGN"> + <term><constant>MALLOCX_ALIGN(<parameter>a</parameter>) + </constant></term> + + <listitem><para>Align the memory allocation to start at an address + that is a multiple of <parameter>a</parameter>, where + <parameter>a</parameter> is a power of two. This macro does not + validate that <parameter>a</parameter> is a power of 2. + </para></listitem> + </varlistentry> + <varlistentry id="MALLOCX_ZERO"> + <term><constant>MALLOCX_ZERO</constant></term> + + <listitem><para>Initialize newly allocated memory to contain zero + bytes. In the growing reallocation case, the real size prior to + reallocation defines the boundary between untouched bytes and those + that are initialized to contain zero bytes. If this macro is + absent, newly allocated memory is uninitialized.</para></listitem> + </varlistentry> + <varlistentry id="MALLOCX_TCACHE"> + <term><constant>MALLOCX_TCACHE(<parameter>tc</parameter>) + </constant></term> + + <listitem><para>Use the thread-specific cache (tcache) specified by + the identifier <parameter>tc</parameter>, which must have been + acquired via the <link + linkend="tcache.create"><mallctl>tcache.create</mallctl></link> + mallctl. This macro does not validate that + <parameter>tc</parameter> specifies a valid + identifier.</para></listitem> + </varlistentry> + <varlistentry id="MALLOC_TCACHE_NONE"> + <term><constant>MALLOCX_TCACHE_NONE</constant></term> + + <listitem><para>Do not use a thread-specific cache (tcache). Unless + <constant>MALLOCX_TCACHE(<parameter>tc</parameter>)</constant> or + <constant>MALLOCX_TCACHE_NONE</constant> is specified, an + automatically managed tcache will be used under many circumstances. + This macro cannot be used in the same <parameter>flags</parameter> + argument as + <constant>MALLOCX_TCACHE(<parameter>tc</parameter>)</constant>.</para></listitem> + </varlistentry> + <varlistentry id="MALLOCX_ARENA"> + <term><constant>MALLOCX_ARENA(<parameter>a</parameter>) + </constant></term> + + <listitem><para>Use the arena specified by the index + <parameter>a</parameter>. This macro has no effect for regions that + were allocated via an arena other than the one specified. This + macro does not validate that <parameter>a</parameter> specifies an + arena index in the valid range.</para></listitem> + </varlistentry> + </variablelist> + </para> + + <para>The <function>mallocx()</function> function allocates at + least <parameter>size</parameter> bytes of memory, and returns a pointer + to the base address of the allocation. Behavior is undefined if + <parameter>size</parameter> is <constant>0</constant>.</para> + + <para>The <function>rallocx()</function> function resizes the + allocation at <parameter>ptr</parameter> to be at least + <parameter>size</parameter> bytes, and returns a pointer to the base + address of the resulting allocation, which may or may not have moved from + its original location. Behavior is undefined if + <parameter>size</parameter> is <constant>0</constant>.</para> + + <para>The <function>xallocx()</function> function resizes the + allocation at <parameter>ptr</parameter> in place to be at least + <parameter>size</parameter> bytes, and returns the real size of the + allocation. If <parameter>extra</parameter> is non-zero, an attempt is + made to resize the allocation to be at least <code + language="C">(<parameter>size</parameter> + + <parameter>extra</parameter>)</code> bytes, though inability to allocate + the extra byte(s) will not by itself result in failure to resize. + Behavior is undefined if <parameter>size</parameter> is + <constant>0</constant>, or if <code + language="C">(<parameter>size</parameter> + <parameter>extra</parameter> + > <constant>SIZE_T_MAX</constant>)</code>.</para> + + <para>The <function>sallocx()</function> function returns the + real size of the allocation at <parameter>ptr</parameter>.</para> + + <para>The <function>dallocx()</function> function causes the + memory referenced by <parameter>ptr</parameter> to be made available for + future allocations.</para> + + <para>The <function>sdallocx()</function> function is an + extension of <function>dallocx()</function> with a + <parameter>size</parameter> parameter to allow the caller to pass in the + allocation size as an optimization. The minimum valid input size is the + original requested size of the allocation, and the maximum valid input + size is the corresponding value returned by + <function>nallocx()</function> or + <function>sallocx()</function>.</para> + + <para>The <function>nallocx()</function> function allocates no + memory, but it performs the same size computation as the + <function>mallocx()</function> function, and returns the real + size of the allocation that would result from the equivalent + <function>mallocx()</function> function call, or + <constant>0</constant> if the inputs exceed the maximum supported size + class and/or alignment. Behavior is undefined if + <parameter>size</parameter> is <constant>0</constant>.</para> + + <para>The <function>mallctl()</function> function provides a + general interface for introspecting the memory allocator, as well as + setting modifiable parameters and triggering actions. The + period-separated <parameter>name</parameter> argument specifies a + location in a tree-structured namespace; see the <xref + linkend="mallctl_namespace" xrefstyle="template:%t"/> section for + documentation on the tree contents. To read a value, pass a pointer via + <parameter>oldp</parameter> to adequate space to contain the value, and a + pointer to its length via <parameter>oldlenp</parameter>; otherwise pass + <constant>NULL</constant> and <constant>NULL</constant>. Similarly, to + write a value, pass a pointer to the value via + <parameter>newp</parameter>, and its length via + <parameter>newlen</parameter>; otherwise pass <constant>NULL</constant> + and <constant>0</constant>.</para> + + <para>The <function>mallctlnametomib()</function> function + provides a way to avoid repeated name lookups for applications that + repeatedly query the same portion of the namespace, by translating a name + to a <quote>Management Information Base</quote> (MIB) that can be passed + repeatedly to <function>mallctlbymib()</function>. Upon + successful return from <function>mallctlnametomib()</function>, + <parameter>mibp</parameter> contains an array of + <parameter>*miblenp</parameter> integers, where + <parameter>*miblenp</parameter> is the lesser of the number of components + in <parameter>name</parameter> and the input value of + <parameter>*miblenp</parameter>. Thus it is possible to pass a + <parameter>*miblenp</parameter> that is smaller than the number of + period-separated name components, which results in a partial MIB that can + be used as the basis for constructing a complete MIB. For name + components that are integers (e.g. the 2 in + <link + linkend="arenas.bin.i.size"><mallctl>arenas.bin.2.size</mallctl></link>), + the corresponding MIB component will always be that integer. Therefore, + it is legitimate to construct code like the following: <programlisting + language="C"><![CDATA[ +unsigned nbins, i; +size_t mib[4]; +size_t len, miblen; + +len = sizeof(nbins); +mallctl("arenas.nbins", &nbins, &len, NULL, 0); + +miblen = 4; +mallctlnametomib("arenas.bin.0.size", mib, &miblen); +for (i = 0; i < nbins; i++) { + size_t bin_size; + + mib[2] = i; + len = sizeof(bin_size); + mallctlbymib(mib, miblen, (void *)&bin_size, &len, NULL, 0); + /* Do something with bin_size... */ +}]]></programlisting></para> + + <varlistentry id="malloc_stats_print_opts"> + </varlistentry> + <para>The <function>malloc_stats_print()</function> function writes + summary statistics via the <parameter>write_cb</parameter> callback + function pointer and <parameter>cbopaque</parameter> data passed to + <parameter>write_cb</parameter>, or <function>malloc_message()</function> + if <parameter>write_cb</parameter> is <constant>NULL</constant>. The + statistics are presented in human-readable form unless <quote>J</quote> is + specified as a character within the <parameter>opts</parameter> string, in + which case the statistics are presented in <ulink + url="http://www.json.org/">JSON format</ulink>. This function can be + called repeatedly. General information that never changes during + execution can be omitted by specifying <quote>g</quote> as a character + within the <parameter>opts</parameter> string. Note that + <function>malloc_stats_print()</function> uses the + <function>mallctl*()</function> functions internally, so inconsistent + statistics can be reported if multiple threads use these functions + simultaneously. If <option>--enable-stats</option> is specified during + configuration, <quote>m</quote>, <quote>d</quote>, and <quote>a</quote> + can be specified to omit merged arena, destroyed merged arena, and per + arena statistics, respectively; <quote>b</quote> and <quote>l</quote> can + be specified to omit per size class statistics for bins and large objects, + respectively; <quote>x</quote> can be specified to omit all mutex + statistics; <quote>e</quote> can be used to omit extent statistics. + Unrecognized characters are silently ignored. Note that thread caching + may prevent some statistics from being completely up to date, since extra + locking would be required to merge counters that track thread cache + operations.</para> + + <para>The <function>malloc_usable_size()</function> function + returns the usable size of the allocation pointed to by + <parameter>ptr</parameter>. The return value may be larger than the size + that was requested during allocation. The + <function>malloc_usable_size()</function> function is not a + mechanism for in-place <function>realloc()</function>; rather + it is provided solely as a tool for introspection purposes. Any + discrepancy between the requested allocation size and the size reported + by <function>malloc_usable_size()</function> should not be + depended on, since such behavior is entirely implementation-dependent. + </para> + </refsect2> + </refsect1> + <refsect1 id="tuning"> + <title>TUNING</title> + <para>Once, when the first call is made to one of the memory allocation + routines, the allocator initializes its internals based in part on various + options that can be specified at compile- or run-time.</para> + + <para>The string specified via <option>--with-malloc-conf</option>, the + string pointed to by the global variable <varname>malloc_conf</varname>, the + <quote>name</quote> of the file referenced by the symbolic link named + <filename class="symlink">/etc/malloc.conf</filename>, and the value of the + environment variable <envar>MALLOC_CONF</envar>, will be interpreted, in + that order, from left to right as options. Note that + <varname>malloc_conf</varname> may be read before + <function>main()</function> is entered, so the declaration of + <varname>malloc_conf</varname> should specify an initializer that contains + the final value to be read by jemalloc. <option>--with-malloc-conf</option> + and <varname>malloc_conf</varname> are compile-time mechanisms, whereas + <filename class="symlink">/etc/malloc.conf</filename> and + <envar>MALLOC_CONF</envar> can be safely set any time prior to program + invocation.</para> + + <para>An options string is a comma-separated list of option:value pairs. + There is one key corresponding to each <link + linkend="opt.abort"><mallctl>opt.*</mallctl></link> mallctl (see the <xref + linkend="mallctl_namespace" xrefstyle="template:%t"/> section for options + documentation). For example, <literal>abort:true,narenas:1</literal> sets + the <link linkend="opt.abort"><mallctl>opt.abort</mallctl></link> and <link + linkend="opt.narenas"><mallctl>opt.narenas</mallctl></link> options. Some + options have boolean values (true/false), others have integer values (base + 8, 10, or 16, depending on prefix), and yet others have raw string + values.</para> + </refsect1> + <refsect1 id="implementation_notes"> + <title>IMPLEMENTATION NOTES</title> + <para>Traditionally, allocators have used + <citerefentry><refentrytitle>sbrk</refentrytitle> + <manvolnum>2</manvolnum></citerefentry> to obtain memory, which is + suboptimal for several reasons, including race conditions, increased + fragmentation, and artificial limitations on maximum usable memory. If + <citerefentry><refentrytitle>sbrk</refentrytitle> + <manvolnum>2</manvolnum></citerefentry> is supported by the operating + system, this allocator uses both + <citerefentry><refentrytitle>mmap</refentrytitle> + <manvolnum>2</manvolnum></citerefentry> and + <citerefentry><refentrytitle>sbrk</refentrytitle> + <manvolnum>2</manvolnum></citerefentry>, in that order of preference; + otherwise only <citerefentry><refentrytitle>mmap</refentrytitle> + <manvolnum>2</manvolnum></citerefentry> is used.</para> + + <para>This allocator uses multiple arenas in order to reduce lock + contention for threaded programs on multi-processor systems. This works + well with regard to threading scalability, but incurs some costs. There is + a small fixed per-arena overhead, and additionally, arenas manage memory + completely independently of each other, which means a small fixed increase + in overall memory fragmentation. These overheads are not generally an + issue, given the number of arenas normally used. Note that using + substantially more arenas than the default is not likely to improve + performance, mainly due to reduced cache performance. However, it may make + sense to reduce the number of arenas if an application does not make much + use of the allocation functions.</para> + + <para>In addition to multiple arenas, this allocator supports + thread-specific caching, in order to make it possible to completely avoid + synchronization for most allocation requests. Such caching allows very fast + allocation in the common case, but it increases memory usage and + fragmentation, since a bounded number of objects can remain allocated in + each thread cache.</para> + + <para>Memory is conceptually broken into extents. Extents are always + aligned to multiples of the page size. This alignment makes it possible to + find metadata for user objects quickly. User objects are broken into two + categories according to size: small and large. Contiguous small objects + comprise a slab, which resides within a single extent, whereas large objects + each have their own extents backing them.</para> + + <para>Small objects are managed in groups by slabs. Each slab maintains + a bitmap to track which regions are in use. Allocation requests that are no + more than half the quantum (8 or 16, depending on architecture) are rounded + up to the nearest power of two that is at least <code + language="C">sizeof(<type>double</type>)</code>. All other object size + classes are multiples of the quantum, spaced such that there are four size + classes for each doubling in size, which limits internal fragmentation to + approximately 20% for all but the smallest size classes. Small size classes + are smaller than four times the page size, and large size classes extend + from four times the page size up to the largest size class that does not + exceed <constant>PTRDIFF_MAX</constant>.</para> + + <para>Allocations are packed tightly together, which can be an issue for + multi-threaded applications. If you need to assure that allocations do not + suffer from cacheline sharing, round your allocation requests up to the + nearest multiple of the cacheline size, or specify cacheline alignment when + allocating.</para> + + <para>The <function>realloc()</function>, + <function>rallocx()</function>, and + <function>xallocx()</function> functions may resize allocations + without moving them under limited circumstances. Unlike the + <function>*allocx()</function> API, the standard API does not + officially round up the usable size of an allocation to the nearest size + class, so technically it is necessary to call + <function>realloc()</function> to grow e.g. a 9-byte allocation to + 16 bytes, or shrink a 16-byte allocation to 9 bytes. Growth and shrinkage + trivially succeeds in place as long as the pre-size and post-size both round + up to the same size class. No other API guarantees are made regarding + in-place resizing, but the current implementation also tries to resize large + allocations in place, as long as the pre-size and post-size are both large. + For shrinkage to succeed, the extent allocator must support splitting (see + <link + linkend="arena.i.extent_hooks"><mallctl>arena.<i>.extent_hooks</mallctl></link>). + Growth only succeeds if the trailing memory is currently available, and the + extent allocator supports merging.</para> + + <para>Assuming 4 KiB pages and a 16-byte quantum on a 64-bit system, the + size classes in each category are as shown in <xref linkend="size_classes" + xrefstyle="template:Table %n"/>.</para> + + <table xml:id="size_classes" frame="all"> + <title>Size classes</title> + <tgroup cols="3" colsep="1" rowsep="1"> + <colspec colname="c1" align="left"/> + <colspec colname="c2" align="right"/> + <colspec colname="c3" align="left"/> + <thead> + <row> + <entry>Category</entry> + <entry>Spacing</entry> + <entry>Size</entry> + </row> + </thead> + <tbody> + <row> + <entry morerows="8">Small</entry> + <entry>lg</entry> + <entry>[8]</entry> + </row> + <row> + <entry>16</entry> + <entry>[16, 32, 48, 64, 80, 96, 112, 128]</entry> + </row> + <row> + <entry>32</entry> + <entry>[160, 192, 224, 256]</entry> + </row> + <row> + <entry>64</entry> + <entry>[320, 384, 448, 512]</entry> + </row> + <row> + <entry>128</entry> + <entry>[640, 768, 896, 1024]</entry> + </row> + <row> + <entry>256</entry> + <entry>[1280, 1536, 1792, 2048]</entry> + </row> + <row> + <entry>512</entry> + <entry>[2560, 3072, 3584, 4096]</entry> + </row> + <row> + <entry>1 KiB</entry> + <entry>[5 KiB, 6 KiB, 7 KiB, 8 KiB]</entry> + </row> + <row> + <entry>2 KiB</entry> + <entry>[10 KiB, 12 KiB, 14 KiB]</entry> + </row> + <row> + <entry morerows="15">Large</entry> + <entry>2 KiB</entry> + <entry>[16 KiB]</entry> + </row> + <row> + <entry>4 KiB</entry> + <entry>[20 KiB, 24 KiB, 28 KiB, 32 KiB]</entry> + </row> + <row> + <entry>8 KiB</entry> + <entry>[40 KiB, 48 KiB, 54 KiB, 64 KiB]</entry> + </row> + <row> + <entry>16 KiB</entry> + <entry>[80 KiB, 96 KiB, 112 KiB, 128 KiB]</entry> + </row> + <row> + <entry>32 KiB</entry> + <entry>[160 KiB, 192 KiB, 224 KiB, 256 KiB]</entry> + </row> + <row> + <entry>64 KiB</entry> + <entry>[320 KiB, 384 KiB, 448 KiB, 512 KiB]</entry> + </row> + <row> + <entry>128 KiB</entry> + <entry>[640 KiB, 768 KiB, 896 KiB, 1 MiB]</entry> + </row> + <row> + <entry>256 KiB</entry> + <entry>[1280 KiB, 1536 KiB, 1792 KiB, 2 MiB]</entry> + </row> + <row> + <entry>512 KiB</entry> + <entry>[2560 KiB, 3 MiB, 3584 KiB, 4 MiB]</entry> + </row> + <row> + <entry>1 MiB</entry> + <entry>[5 MiB, 6 MiB, 7 MiB, 8 MiB]</entry> + </row> + <row> + <entry>2 MiB</entry> + <entry>[10 MiB, 12 MiB, 14 MiB, 16 MiB]</entry> + </row> + <row> + <entry>4 MiB</entry> + <entry>[20 MiB, 24 MiB, 28 MiB, 32 MiB]</entry> + </row> + <row> + <entry>8 MiB</entry> + <entry>[40 MiB, 48 MiB, 56 MiB, 64 MiB]</entry> + </row> + <row> + <entry>...</entry> + <entry>...</entry> + </row> + <row> + <entry>512 PiB</entry> + <entry>[2560 PiB, 3 EiB, 3584 PiB, 4 EiB]</entry> + </row> + <row> + <entry>1 EiB</entry> + <entry>[5 EiB, 6 EiB, 7 EiB]</entry> + </row> + </tbody> + </tgroup> + </table> + </refsect1> + <refsect1 id="mallctl_namespace"> + <title>MALLCTL NAMESPACE</title> + <para>The following names are defined in the namespace accessible via the + <function>mallctl*()</function> functions. Value types are specified in + parentheses, their readable/writable statuses are encoded as + <literal>rw</literal>, <literal>r-</literal>, <literal>-w</literal>, or + <literal>--</literal>, and required build configuration flags follow, if + any. A name element encoded as <literal><i></literal> or + <literal><j></literal> indicates an integer component, where the + integer varies from 0 to some upper value that must be determined via + introspection. In the case of <mallctl>stats.arenas.<i>.*</mallctl> + and <mallctl>arena.<i>.{initialized,purge,decay,dss}</mallctl>, + <literal><i></literal> equal to + <constant>MALLCTL_ARENAS_ALL</constant> can be used to operate on all arenas + or access the summation of statistics from all arenas; similarly + <literal><i></literal> equal to + <constant>MALLCTL_ARENAS_DESTROYED</constant> can be used to access the + summation of statistics from all destroyed arenas. These constants can be + utilized either via <function>mallctlnametomib()</function> followed by + <function>mallctlbymib()</function>, or via code such as the following: + <programlisting language="C"><![CDATA[ +#define STRINGIFY_HELPER(x) #x +#define STRINGIFY(x) STRINGIFY_HELPER(x) + +mallctl("arena." STRINGIFY(MALLCTL_ARENAS_ALL) ".decay", + NULL, NULL, NULL, 0);]]></programlisting> + Take special note of the <link + linkend="epoch"><mallctl>epoch</mallctl></link> mallctl, which controls + refreshing of cached dynamic statistics.</para> + + <variablelist> + <varlistentry id="version"> + <term> + <mallctl>version</mallctl> + (<type>const char *</type>) + <literal>r-</literal> + </term> + <listitem><para>Return the jemalloc version string.</para></listitem> + </varlistentry> + + <varlistentry id="epoch"> + <term> + <mallctl>epoch</mallctl> + (<type>uint64_t</type>) + <literal>rw</literal> + </term> + <listitem><para>If a value is passed in, refresh the data from which + the <function>mallctl*()</function> functions report values, + and increment the epoch. Return the current epoch. This is useful for + detecting whether another thread caused a refresh.</para></listitem> + </varlistentry> + + <varlistentry id="background_thread"> + <term> + <mallctl>background_thread</mallctl> + (<type>bool</type>) + <literal>rw</literal> + </term> + <listitem><para>Enable/disable internal background worker threads. When + set to true, background threads are created on demand (the number of + background threads will be no more than the number of CPUs or active + arenas). Threads run periodically, and handle <link + linkend="arena.i.decay">purging</link> asynchronously. When switching + off, background threads are terminated synchronously. Note that after + <citerefentry><refentrytitle>fork</refentrytitle><manvolnum>2</manvolnum></citerefentry> + function, the state in the child process will be disabled regardless + the state in parent process. See <link + linkend="stats.background_thread.num_threads"><mallctl>stats.background_thread</mallctl></link> + for related stats. <link + linkend="opt.background_thread"><mallctl>opt.background_thread</mallctl></link> + can be used to set the default option. This option is only available on + selected pthread-based platforms.</para></listitem> + </varlistentry> + + <varlistentry id="max_background_threads"> + <term> + <mallctl>max_background_threads</mallctl> + (<type>size_t</type>) + <literal>rw</literal> + </term> + <listitem><para>Maximum number of background worker threads that will + be created. This value is capped at <link + linkend="opt.max_background_threads"><mallctl>opt.max_background_threads</mallctl></link> at + startup.</para></listitem> + </varlistentry> + + <varlistentry id="config.cache_oblivious"> + <term> + <mallctl>config.cache_oblivious</mallctl> + (<type>bool</type>) + <literal>r-</literal> + </term> + <listitem><para><option>--enable-cache-oblivious</option> was specified + during build configuration.</para></listitem> + </varlistentry> + + <varlistentry id="config.debug"> + <term> + <mallctl>config.debug</mallctl> + (<type>bool</type>) + <literal>r-</literal> + </term> + <listitem><para><option>--enable-debug</option> was specified during + build configuration.</para></listitem> + </varlistentry> + + <varlistentry id="config.fill"> + <term> + <mallctl>config.fill</mallctl> + (<type>bool</type>) + <literal>r-</literal> + </term> + <listitem><para><option>--enable-fill</option> was specified during + build configuration.</para></listitem> + </varlistentry> + + <varlistentry id="config.lazy_lock"> + <term> + <mallctl>config.lazy_lock</mallctl> + (<type>bool</type>) + <literal>r-</literal> + </term> + <listitem><para><option>--enable-lazy-lock</option> was specified + during build configuration.</para></listitem> + </varlistentry> + + <varlistentry id="config.malloc_conf"> + <term> + <mallctl>config.malloc_conf</mallctl> + (<type>const char *</type>) + <literal>r-</literal> + </term> + <listitem><para>Embedded configure-time-specified run-time options + string, empty unless <option>--with-malloc-conf</option> was specified + during build configuration.</para></listitem> + </varlistentry> + + <varlistentry id="config.prof"> + <term> + <mallctl>config.prof</mallctl> + (<type>bool</type>) + <literal>r-</literal> + </term> + <listitem><para><option>--enable-prof</option> was specified during + build configuration.</para></listitem> + </varlistentry> + + <varlistentry id="config.prof_libgcc"> + <term> + <mallctl>config.prof_libgcc</mallctl> + (<type>bool</type>) + <literal>r-</literal> + </term> + <listitem><para><option>--disable-prof-libgcc</option> was not + specified during build configuration.</para></listitem> + </varlistentry> + + <varlistentry id="config.prof_libunwind"> + <term> + <mallctl>config.prof_libunwind</mallctl> + (<type>bool</type>) + <literal>r-</literal> + </term> + <listitem><para><option>--enable-prof-libunwind</option> was specified + during build configuration.</para></listitem> + </varlistentry> + + <varlistentry id="config.stats"> + <term> + <mallctl>config.stats</mallctl> + (<type>bool</type>) + <literal>r-</literal> + </term> + <listitem><para><option>--enable-stats</option> was specified during + build configuration.</para></listitem> + </varlistentry> + + + <varlistentry id="config.utrace"> + <term> + <mallctl>config.utrace</mallctl> + (<type>bool</type>) + <literal>r-</literal> + </term> + <listitem><para><option>--enable-utrace</option> was specified during + build configuration.</para></listitem> + </varlistentry> + + <varlistentry id="config.xmalloc"> + <term> + <mallctl>config.xmalloc</mallctl> + (<type>bool</type>) + <literal>r-</literal> + </term> + <listitem><para><option>--enable-xmalloc</option> was specified during + build configuration.</para></listitem> + </varlistentry> + + <varlistentry id="opt.abort"> + <term> + <mallctl>opt.abort</mallctl> + (<type>bool</type>) + <literal>r-</literal> + </term> + <listitem><para>Abort-on-warning enabled/disabled. If true, most + warnings are fatal. Note that runtime option warnings are not included + (see <link + linkend="opt.abort_conf"><mallctl>opt.abort_conf</mallctl></link> for + that). The process will call + <citerefentry><refentrytitle>abort</refentrytitle> + <manvolnum>3</manvolnum></citerefentry> in these cases. This option is + disabled by default unless <option>--enable-debug</option> is + specified during configuration, in which case it is enabled by default. + </para></listitem> + </varlistentry> + + <varlistentry id="opt.confirm_conf"> + <term> + <mallctl>opt.confirm_conf</mallctl> + (<type>bool</type>) + <literal>r-</literal> + </term> + <listitem><para>Confirm-runtime-options-when-program-starts + enabled/disabled. If true, the string specified via + <option>--with-malloc-conf</option>, the string pointed to by the + global variable <varname>malloc_conf</varname>, the <quote>name</quote> + of the file referenced by the symbolic link named + <filename class="symlink">/etc/malloc.conf</filename>, and the value of + the environment variable <envar>MALLOC_CONF</envar>, will be printed in + order. Then, each option being set will be individually printed. This + option is disabled by default.</para></listitem> + </varlistentry> + + <varlistentry id="opt.abort_conf"> + <term> + <mallctl>opt.abort_conf</mallctl> + (<type>bool</type>) + <literal>r-</literal> + </term> + <listitem><para>Abort-on-invalid-configuration enabled/disabled. If + true, invalid runtime options are fatal. The process will call + <citerefentry><refentrytitle>abort</refentrytitle> + <manvolnum>3</manvolnum></citerefentry> in these cases. This option is + disabled by default unless <option>--enable-debug</option> is + specified during configuration, in which case it is enabled by default. + </para></listitem> + </varlistentry> + + <varlistentry id="opt.metadata_thp"> + <term> + <mallctl>opt.metadata_thp</mallctl> + (<type>const char *</type>) + <literal>r-</literal> + </term> + <listitem><para>Controls whether to allow jemalloc to use transparent + huge page (THP) for internal metadata (see <link + linkend="stats.metadata">stats.metadata</link>). <quote>always</quote> + allows such usage. <quote>auto</quote> uses no THP initially, but may + begin to do so when metadata usage reaches certain level. The default + is <quote>disabled</quote>.</para></listitem> + </varlistentry> + + <varlistentry id="opt.retain"> + <term> + <mallctl>opt.retain</mallctl> + (<type>bool</type>) + <literal>r-</literal> + </term> + <listitem><para>If true, retain unused virtual memory for later reuse + rather than discarding it by calling + <citerefentry><refentrytitle>munmap</refentrytitle> + <manvolnum>2</manvolnum></citerefentry> or equivalent (see <link + linkend="stats.retained">stats.retained</link> for related details). + It also makes jemalloc use <citerefentry> + <refentrytitle>mmap</refentrytitle><manvolnum>2</manvolnum> + </citerefentry> or equivalent in a more greedy way, mapping larger + chunks in one go. This option is disabled by default unless discarding + virtual memory is known to trigger platform-specific performance + problems, namely 1) for [64-bit] Linux, which has a quirk in its virtual + memory allocation algorithm that causes semi-permanent VM map holes + under normal jemalloc operation; and 2) for [64-bit] Windows, which + disallows split / merged regions with + <parameter><constant>MEM_RELEASE</constant></parameter>. Although the + same issues may present on 32-bit platforms as well, retaining virtual + memory for 32-bit Linux and Windows is disabled by default due to the + practical possibility of address space exhaustion. </para></listitem> + </varlistentry> + + <varlistentry id="opt.dss"> + <term> + <mallctl>opt.dss</mallctl> + (<type>const char *</type>) + <literal>r-</literal> + </term> + <listitem><para>dss (<citerefentry><refentrytitle>sbrk</refentrytitle> + <manvolnum>2</manvolnum></citerefentry>) allocation precedence as + related to <citerefentry><refentrytitle>mmap</refentrytitle> + <manvolnum>2</manvolnum></citerefentry> allocation. The following + settings are supported if + <citerefentry><refentrytitle>sbrk</refentrytitle> + <manvolnum>2</manvolnum></citerefentry> is supported by the operating + system: <quote>disabled</quote>, <quote>primary</quote>, and + <quote>secondary</quote>; otherwise only <quote>disabled</quote> is + supported. The default is <quote>secondary</quote> if + <citerefentry><refentrytitle>sbrk</refentrytitle> + <manvolnum>2</manvolnum></citerefentry> is supported by the operating + system; <quote>disabled</quote> otherwise. + </para></listitem> + </varlistentry> + + <varlistentry id="opt.narenas"> + <term> + <mallctl>opt.narenas</mallctl> + (<type>unsigned</type>) + <literal>r-</literal> + </term> + <listitem><para>Maximum number of arenas to use for automatic + multiplexing of threads and arenas. The default is four times the + number of CPUs, or one if there is a single CPU.</para></listitem> + </varlistentry> + + <varlistentry id="opt.oversize_threshold"> + <term> + <mallctl>opt.oversize_threshold</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + </term> + <listitem><para>The threshold in bytes of which requests are considered + oversize. Allocation requests with greater sizes are fulfilled from a + dedicated arena (automatically managed, however not within + <literal>narenas</literal>), in order to reduce fragmentation by not + mixing huge allocations with small ones. In addition, the decay API + guarantees on the extents greater than the specified threshold may be + overridden. Note that requests with arena index specified via + <constant>MALLOCX_ARENA</constant>, or threads associated with explicit + arenas will not be considered. The default threshold is 8MiB. Values + not within large size classes disables this feature.</para></listitem> + </varlistentry> + + <varlistentry id="opt.percpu_arena"> + <term> + <mallctl>opt.percpu_arena</mallctl> + (<type>const char *</type>) + <literal>r-</literal> + </term> + <listitem><para>Per CPU arena mode. Use the <quote>percpu</quote> + setting to enable this feature, which uses number of CPUs to determine + number of arenas, and bind threads to arenas dynamically based on the + CPU the thread runs on currently. <quote>phycpu</quote> setting uses + one arena per physical CPU, which means the two hyper threads on the + same CPU share one arena. Note that no runtime checking regarding the + availability of hyper threading is done at the moment. When set to + <quote>disabled</quote>, narenas and thread to arena association will + not be impacted by this option. The default is <quote>disabled</quote>. + </para></listitem> + </varlistentry> + + <varlistentry id="opt.background_thread"> + <term> + <mallctl>opt.background_thread</mallctl> + (<type>bool</type>) + <literal>r-</literal> + </term> + <listitem><para>Internal background worker threads enabled/disabled. + Because of potential circular dependencies, enabling background thread + using this option may cause crash or deadlock during initialization. For + a reliable way to use this feature, see <link + linkend="background_thread">background_thread</link> for dynamic control + options and details. This option is disabled by + default.</para></listitem> + </varlistentry> + + <varlistentry id="opt.max_background_threads"> + <term> + <mallctl>opt.max_background_threads</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + </term> + <listitem><para>Maximum number of background threads that will be created + if <link linkend="background_thread">background_thread</link> is set. + Defaults to number of cpus.</para></listitem> + </varlistentry> + + <varlistentry id="opt.dirty_decay_ms"> + <term> + <mallctl>opt.dirty_decay_ms</mallctl> + (<type>ssize_t</type>) + <literal>r-</literal> + </term> + <listitem><para>Approximate time in milliseconds from the creation of a + set of unused dirty pages until an equivalent set of unused dirty pages + is purged (i.e. converted to muzzy via e.g. + <function>madvise(<parameter>...</parameter><parameter><constant>MADV_FREE</constant></parameter>)</function> + if supported by the operating system, or converted to clean otherwise) + and/or reused. Dirty pages are defined as previously having been + potentially written to by the application, and therefore consuming + physical memory, yet having no current use. The pages are incrementally + purged according to a sigmoidal decay curve that starts and ends with + zero purge rate. A decay time of 0 causes all unused dirty pages to be + purged immediately upon creation. A decay time of -1 disables purging. + The default decay time is 10 seconds. See <link + linkend="arenas.dirty_decay_ms"><mallctl>arenas.dirty_decay_ms</mallctl></link> + and <link + linkend="arena.i.dirty_decay_ms"><mallctl>arena.<i>.dirty_decay_ms</mallctl></link> + for related dynamic control options. See <link + linkend="opt.muzzy_decay_ms"><mallctl>opt.muzzy_decay_ms</mallctl></link> + for a description of muzzy pages.for a description of muzzy pages. Note + that when the <link + linkend="opt.oversize_threshold"><mallctl>oversize_threshold</mallctl></link> + feature is enabled, the arenas reserved for oversize requests may have + its own default decay settings.</para></listitem> + </varlistentry> + + <varlistentry id="opt.muzzy_decay_ms"> + <term> + <mallctl>opt.muzzy_decay_ms</mallctl> + (<type>ssize_t</type>) + <literal>r-</literal> + </term> + <listitem><para>Approximate time in milliseconds from the creation of a + set of unused muzzy pages until an equivalent set of unused muzzy pages + is purged (i.e. converted to clean) and/or reused. Muzzy pages are + defined as previously having been unused dirty pages that were + subsequently purged in a manner that left them subject to the + reclamation whims of the operating system (e.g. + <function>madvise(<parameter>...</parameter><parameter><constant>MADV_FREE</constant></parameter>)</function>), + and therefore in an indeterminate state. The pages are incrementally + purged according to a sigmoidal decay curve that starts and ends with + zero purge rate. A decay time of 0 causes all unused muzzy pages to be + purged immediately upon creation. A decay time of -1 disables purging. + The default decay time is 10 seconds. See <link + linkend="arenas.muzzy_decay_ms"><mallctl>arenas.muzzy_decay_ms</mallctl></link> + and <link + linkend="arena.i.muzzy_decay_ms"><mallctl>arena.<i>.muzzy_decay_ms</mallctl></link> + for related dynamic control options.</para></listitem> + </varlistentry> + + <varlistentry id="opt.lg_extent_max_active_fit"> + <term> + <mallctl>opt.lg_extent_max_active_fit</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + </term> + <listitem><para>When reusing dirty extents, this determines the (log + base 2 of the) maximum ratio between the size of the active extent + selected (to split off from) and the size of the requested allocation. + This prevents the splitting of large active extents for smaller + allocations, which can reduce fragmentation over the long run + (especially for non-active extents). Lower value may reduce + fragmentation, at the cost of extra active extents. The default value + is 6, which gives a maximum ratio of 64 (2^6).</para></listitem> + </varlistentry> + + <varlistentry id="opt.stats_print"> + <term> + <mallctl>opt.stats_print</mallctl> + (<type>bool</type>) + <literal>r-</literal> + </term> + <listitem><para>Enable/disable statistics printing at exit. If + enabled, the <function>malloc_stats_print()</function> + function is called at program exit via an + <citerefentry><refentrytitle>atexit</refentrytitle> + <manvolnum>3</manvolnum></citerefentry> function. <link + linkend="opt.stats_print_opts"><mallctl>opt.stats_print_opts</mallctl></link> + can be combined to specify output options. If + <option>--enable-stats</option> is specified during configuration, this + has the potential to cause deadlock for a multi-threaded process that + exits while one or more threads are executing in the memory allocation + functions. Furthermore, <function>atexit()</function> may + allocate memory during application initialization and then deadlock + internally when jemalloc in turn calls + <function>atexit()</function>, so this option is not + universally usable (though the application can register its own + <function>atexit()</function> function with equivalent + functionality). Therefore, this option should only be used with care; + it is primarily intended as a performance tuning aid during application + development. This option is disabled by default.</para></listitem> + </varlistentry> + + <varlistentry id="opt.stats_print_opts"> + <term> + <mallctl>opt.stats_print_opts</mallctl> + (<type>const char *</type>) + <literal>r-</literal> + </term> + <listitem><para>Options (the <parameter>opts</parameter> string) to pass + to the <function>malloc_stats_print()</function> at exit (enabled + through <link + linkend="opt.stats_print"><mallctl>opt.stats_print</mallctl></link>). See + available options in <link + linkend="malloc_stats_print_opts"><function>malloc_stats_print()</function></link>. + Has no effect unless <link + linkend="opt.stats_print"><mallctl>opt.stats_print</mallctl></link> is + enabled. The default is <quote></quote>.</para></listitem> + </varlistentry> + + <varlistentry id="opt.junk"> + <term> + <mallctl>opt.junk</mallctl> + (<type>const char *</type>) + <literal>r-</literal> + [<option>--enable-fill</option>] + </term> + <listitem><para>Junk filling. If set to <quote>alloc</quote>, each byte + of uninitialized allocated memory will be initialized to + <literal>0xa5</literal>. If set to <quote>free</quote>, all deallocated + memory will be initialized to <literal>0x5a</literal>. If set to + <quote>true</quote>, both allocated and deallocated memory will be + initialized, and if set to <quote>false</quote>, junk filling be + disabled entirely. This is intended for debugging and will impact + performance negatively. This option is <quote>false</quote> by default + unless <option>--enable-debug</option> is specified during + configuration, in which case it is <quote>true</quote> by + default.</para></listitem> + </varlistentry> + + <varlistentry id="opt.zero"> + <term> + <mallctl>opt.zero</mallctl> + (<type>bool</type>) + <literal>r-</literal> + [<option>--enable-fill</option>] + </term> + <listitem><para>Zero filling enabled/disabled. If enabled, each byte + of uninitialized allocated memory will be initialized to 0. Note that + this initialization only happens once for each byte, so + <function>realloc()</function> and + <function>rallocx()</function> calls do not zero memory that + was previously allocated. This is intended for debugging and will + impact performance negatively. This option is disabled by default. + </para></listitem> + </varlistentry> + + <varlistentry id="opt.utrace"> + <term> + <mallctl>opt.utrace</mallctl> + (<type>bool</type>) + <literal>r-</literal> + [<option>--enable-utrace</option>] + </term> + <listitem><para>Allocation tracing based on + <citerefentry><refentrytitle>utrace</refentrytitle> + <manvolnum>2</manvolnum></citerefentry> enabled/disabled. This option + is disabled by default.</para></listitem> + </varlistentry> + + <varlistentry id="opt.xmalloc"> + <term> + <mallctl>opt.xmalloc</mallctl> + (<type>bool</type>) + <literal>r-</literal> + [<option>--enable-xmalloc</option>] + </term> + <listitem><para>Abort-on-out-of-memory enabled/disabled. If enabled, + rather than returning failure for any allocation function, display a + diagnostic message on <constant>STDERR_FILENO</constant> and cause the + program to drop core (using + <citerefentry><refentrytitle>abort</refentrytitle> + <manvolnum>3</manvolnum></citerefentry>). If an application is + designed to depend on this behavior, set the option at compile time by + including the following in the source code: + <programlisting language="C"><![CDATA[ +malloc_conf = "xmalloc:true";]]></programlisting> + This option is disabled by default.</para></listitem> + </varlistentry> + + <varlistentry id="opt.tcache"> + <term> + <mallctl>opt.tcache</mallctl> + (<type>bool</type>) + <literal>r-</literal> + </term> + <listitem><para>Thread-specific caching (tcache) enabled/disabled. When + there are multiple threads, each thread uses a tcache for objects up to + a certain size. Thread-specific caching allows many allocations to be + satisfied without performing any thread synchronization, at the cost of + increased memory use. See the <link + linkend="opt.lg_tcache_max"><mallctl>opt.lg_tcache_max</mallctl></link> + option for related tuning information. This option is enabled by + default.</para></listitem> + </varlistentry> + + <varlistentry id="opt.lg_tcache_max"> + <term> + <mallctl>opt.lg_tcache_max</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + </term> + <listitem><para>Maximum size class (log base 2) to cache in the + thread-specific cache (tcache). At a minimum, all small size classes + are cached, and at a maximum all large size classes are cached. The + default maximum is 32 KiB (2^15).</para></listitem> + </varlistentry> + + <varlistentry id="opt.thp"> + <term> + <mallctl>opt.thp</mallctl> + (<type>const char *</type>) + <literal>r-</literal> + </term> + <listitem><para>Transparent hugepage (THP) mode. Settings "always", + "never" and "default" are available if THP is supported by the operating + system. The "always" setting enables transparent hugepage for all user + memory mappings with + <parameter><constant>MADV_HUGEPAGE</constant></parameter>; "never" + ensures no transparent hugepage with + <parameter><constant>MADV_NOHUGEPAGE</constant></parameter>; the default + setting "default" makes no changes. Note that: this option does not + affect THP for jemalloc internal metadata (see <link + linkend="opt.metadata_thp"><mallctl>opt.metadata_thp</mallctl></link>); + in addition, for arenas with customized <link + linkend="arena.i.extent_hooks"><mallctl>extent_hooks</mallctl></link>, + this option is bypassed as it is implemented as part of the default + extent hooks.</para></listitem> + </varlistentry> + + <varlistentry id="opt.prof"> + <term> + <mallctl>opt.prof</mallctl> + (<type>bool</type>) + <literal>r-</literal> + [<option>--enable-prof</option>] + </term> + <listitem><para>Memory profiling enabled/disabled. If enabled, profile + memory allocation activity. See the <link + linkend="opt.prof_active"><mallctl>opt.prof_active</mallctl></link> + option for on-the-fly activation/deactivation. See the <link + linkend="opt.lg_prof_sample"><mallctl>opt.lg_prof_sample</mallctl></link> + option for probabilistic sampling control. See the <link + linkend="opt.prof_accum"><mallctl>opt.prof_accum</mallctl></link> + option for control of cumulative sample reporting. See the <link + linkend="opt.lg_prof_interval"><mallctl>opt.lg_prof_interval</mallctl></link> + option for information on interval-triggered profile dumping, the <link + linkend="opt.prof_gdump"><mallctl>opt.prof_gdump</mallctl></link> + option for information on high-water-triggered profile dumping, and the + <link linkend="opt.prof_final"><mallctl>opt.prof_final</mallctl></link> + option for final profile dumping. Profile output is compatible with + the <command>jeprof</command> command, which is based on the + <command>pprof</command> that is developed as part of the <ulink + url="http://code.google.com/p/gperftools/">gperftools + package</ulink>. See <link linkend="heap_profile_format">HEAP PROFILE + FORMAT</link> for heap profile format documentation.</para></listitem> + </varlistentry> + + <varlistentry id="opt.prof_prefix"> + <term> + <mallctl>opt.prof_prefix</mallctl> + (<type>const char *</type>) + <literal>r-</literal> + [<option>--enable-prof</option>] + </term> + <listitem><para>Filename prefix for profile dumps. If the prefix is + set to the empty string, no automatic dumps will occur; this is + primarily useful for disabling the automatic final heap dump (which + also disables leak reporting, if enabled). The default prefix is + <filename>jeprof</filename>.</para></listitem> + </varlistentry> + + <varlistentry id="opt.prof_active"> + <term> + <mallctl>opt.prof_active</mallctl> + (<type>bool</type>) + <literal>r-</literal> + [<option>--enable-prof</option>] + </term> + <listitem><para>Profiling activated/deactivated. This is a secondary + control mechanism that makes it possible to start the application with + profiling enabled (see the <link + linkend="opt.prof"><mallctl>opt.prof</mallctl></link> option) but + inactive, then toggle profiling at any time during program execution + with the <link + linkend="prof.active"><mallctl>prof.active</mallctl></link> mallctl. + This option is enabled by default.</para></listitem> + </varlistentry> + + <varlistentry id="opt.prof_thread_active_init"> + <term> + <mallctl>opt.prof_thread_active_init</mallctl> + (<type>bool</type>) + <literal>r-</literal> + [<option>--enable-prof</option>] + </term> + <listitem><para>Initial setting for <link + linkend="thread.prof.active"><mallctl>thread.prof.active</mallctl></link> + in newly created threads. The initial setting for newly created threads + can also be changed during execution via the <link + linkend="prof.thread_active_init"><mallctl>prof.thread_active_init</mallctl></link> + mallctl. This option is enabled by default.</para></listitem> + </varlistentry> + + <varlistentry id="opt.lg_prof_sample"> + <term> + <mallctl>opt.lg_prof_sample</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-prof</option>] + </term> + <listitem><para>Average interval (log base 2) between allocation + samples, as measured in bytes of allocation activity. Increasing the + sampling interval decreases profile fidelity, but also decreases the + computational overhead. The default sample interval is 512 KiB (2^19 + B).</para></listitem> + </varlistentry> + + <varlistentry id="opt.prof_accum"> + <term> + <mallctl>opt.prof_accum</mallctl> + (<type>bool</type>) + <literal>r-</literal> + [<option>--enable-prof</option>] + </term> + <listitem><para>Reporting of cumulative object/byte counts in profile + dumps enabled/disabled. If this option is enabled, every unique + backtrace must be stored for the duration of execution. Depending on + the application, this can impose a large memory overhead, and the + cumulative counts are not always of interest. This option is disabled + by default.</para></listitem> + </varlistentry> + + <varlistentry id="opt.lg_prof_interval"> + <term> + <mallctl>opt.lg_prof_interval</mallctl> + (<type>ssize_t</type>) + <literal>r-</literal> + [<option>--enable-prof</option>] + </term> + <listitem><para>Average interval (log base 2) between memory profile + dumps, as measured in bytes of allocation activity. The actual + interval between dumps may be sporadic because decentralized allocation + counters are used to avoid synchronization bottlenecks. Profiles are + dumped to files named according to the pattern + <filename><prefix>.<pid>.<seq>.i<iseq>.heap</filename>, + where <literal><prefix></literal> is controlled by the + <link + linkend="opt.prof_prefix"><mallctl>opt.prof_prefix</mallctl></link> + option. By default, interval-triggered profile dumping is disabled + (encoded as -1). + </para></listitem> + </varlistentry> + + <varlistentry id="opt.prof_gdump"> + <term> + <mallctl>opt.prof_gdump</mallctl> + (<type>bool</type>) + <literal>r-</literal> + [<option>--enable-prof</option>] + </term> + <listitem><para>Set the initial state of <link + linkend="prof.gdump"><mallctl>prof.gdump</mallctl></link>, which when + enabled triggers a memory profile dump every time the total virtual + memory exceeds the previous maximum. This option is disabled by + default.</para></listitem> + </varlistentry> + + <varlistentry id="opt.prof_final"> + <term> + <mallctl>opt.prof_final</mallctl> + (<type>bool</type>) + <literal>r-</literal> + [<option>--enable-prof</option>] + </term> + <listitem><para>Use an + <citerefentry><refentrytitle>atexit</refentrytitle> + <manvolnum>3</manvolnum></citerefentry> function to dump final memory + usage to a file named according to the pattern + <filename><prefix>.<pid>.<seq>.f.heap</filename>, + where <literal><prefix></literal> is controlled by the <link + linkend="opt.prof_prefix"><mallctl>opt.prof_prefix</mallctl></link> + option. Note that <function>atexit()</function> may allocate + memory during application initialization and then deadlock internally + when jemalloc in turn calls <function>atexit()</function>, so + this option is not universally usable (though the application can + register its own <function>atexit()</function> function with + equivalent functionality). This option is disabled by + default.</para></listitem> + </varlistentry> + + <varlistentry id="opt.prof_leak"> + <term> + <mallctl>opt.prof_leak</mallctl> + (<type>bool</type>) + <literal>r-</literal> + [<option>--enable-prof</option>] + </term> + <listitem><para>Leak reporting enabled/disabled. If enabled, use an + <citerefentry><refentrytitle>atexit</refentrytitle> + <manvolnum>3</manvolnum></citerefentry> function to report memory leaks + detected by allocation sampling. See the + <link linkend="opt.prof"><mallctl>opt.prof</mallctl></link> option for + information on analyzing heap profile output. This option is disabled + by default.</para></listitem> + </varlistentry> + + <varlistentry id="thread.arena"> + <term> + <mallctl>thread.arena</mallctl> + (<type>unsigned</type>) + <literal>rw</literal> + </term> + <listitem><para>Get or set the arena associated with the calling + thread. If the specified arena was not initialized beforehand (see the + <link + linkend="arena.i.initialized"><mallctl>arena.i.initialized</mallctl></link> + mallctl), it will be automatically initialized as a side effect of + calling this interface.</para></listitem> + </varlistentry> + + <varlistentry id="thread.allocated"> + <term> + <mallctl>thread.allocated</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Get the total number of bytes ever allocated by the + calling thread. This counter has the potential to wrap around; it is + up to the application to appropriately interpret the counter in such + cases.</para></listitem> + </varlistentry> + + <varlistentry id="thread.allocatedp"> + <term> + <mallctl>thread.allocatedp</mallctl> + (<type>uint64_t *</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Get a pointer to the the value that is returned by the + <link + linkend="thread.allocated"><mallctl>thread.allocated</mallctl></link> + mallctl. This is useful for avoiding the overhead of repeated + <function>mallctl*()</function> calls.</para></listitem> + </varlistentry> + + <varlistentry id="thread.deallocated"> + <term> + <mallctl>thread.deallocated</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Get the total number of bytes ever deallocated by the + calling thread. This counter has the potential to wrap around; it is + up to the application to appropriately interpret the counter in such + cases.</para></listitem> + </varlistentry> + + <varlistentry id="thread.deallocatedp"> + <term> + <mallctl>thread.deallocatedp</mallctl> + (<type>uint64_t *</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Get a pointer to the the value that is returned by the + <link + linkend="thread.deallocated"><mallctl>thread.deallocated</mallctl></link> + mallctl. This is useful for avoiding the overhead of repeated + <function>mallctl*()</function> calls.</para></listitem> + </varlistentry> + + <varlistentry id="thread.tcache.enabled"> + <term> + <mallctl>thread.tcache.enabled</mallctl> + (<type>bool</type>) + <literal>rw</literal> + </term> + <listitem><para>Enable/disable calling thread's tcache. The tcache is + implicitly flushed as a side effect of becoming + disabled (see <link + linkend="thread.tcache.flush"><mallctl>thread.tcache.flush</mallctl></link>). + </para></listitem> + </varlistentry> + + <varlistentry id="thread.tcache.flush"> + <term> + <mallctl>thread.tcache.flush</mallctl> + (<type>void</type>) + <literal>--</literal> + </term> + <listitem><para>Flush calling thread's thread-specific cache (tcache). + This interface releases all cached objects and internal data structures + associated with the calling thread's tcache. Ordinarily, this interface + need not be called, since automatic periodic incremental garbage + collection occurs, and the thread cache is automatically discarded when + a thread exits. However, garbage collection is triggered by allocation + activity, so it is possible for a thread that stops + allocating/deallocating to retain its cache indefinitely, in which case + the developer may find manual flushing useful.</para></listitem> + </varlistentry> + + <varlistentry id="thread.prof.name"> + <term> + <mallctl>thread.prof.name</mallctl> + (<type>const char *</type>) + <literal>r-</literal> or + <literal>-w</literal> + [<option>--enable-prof</option>] + </term> + <listitem><para>Get/set the descriptive name associated with the calling + thread in memory profile dumps. An internal copy of the name string is + created, so the input string need not be maintained after this interface + completes execution. The output string of this interface should be + copied for non-ephemeral uses, because multiple implementation details + can cause asynchronous string deallocation. Furthermore, each + invocation of this interface can only read or write; simultaneous + read/write is not supported due to string lifetime limitations. The + name string must be nil-terminated and comprised only of characters in + the sets recognized + by <citerefentry><refentrytitle>isgraph</refentrytitle> + <manvolnum>3</manvolnum></citerefentry> and + <citerefentry><refentrytitle>isblank</refentrytitle> + <manvolnum>3</manvolnum></citerefentry>.</para></listitem> + </varlistentry> + + <varlistentry id="thread.prof.active"> + <term> + <mallctl>thread.prof.active</mallctl> + (<type>bool</type>) + <literal>rw</literal> + [<option>--enable-prof</option>] + </term> + <listitem><para>Control whether sampling is currently active for the + calling thread. This is an activation mechanism in addition to <link + linkend="prof.active"><mallctl>prof.active</mallctl></link>; both must + be active for the calling thread to sample. This flag is enabled by + default.</para></listitem> + </varlistentry> + + <varlistentry id="tcache.create"> + <term> + <mallctl>tcache.create</mallctl> + (<type>unsigned</type>) + <literal>r-</literal> + </term> + <listitem><para>Create an explicit thread-specific cache (tcache) and + return an identifier that can be passed to the <link + linkend="MALLOCX_TCACHE"><constant>MALLOCX_TCACHE(<parameter>tc</parameter>)</constant></link> + macro to explicitly use the specified cache rather than the + automatically managed one that is used by default. Each explicit cache + can be used by only one thread at a time; the application must assure + that this constraint holds. + </para></listitem> + </varlistentry> + + <varlistentry id="tcache.flush"> + <term> + <mallctl>tcache.flush</mallctl> + (<type>unsigned</type>) + <literal>-w</literal> + </term> + <listitem><para>Flush the specified thread-specific cache (tcache). The + same considerations apply to this interface as to <link + linkend="thread.tcache.flush"><mallctl>thread.tcache.flush</mallctl></link>, + except that the tcache will never be automatically discarded. + </para></listitem> + </varlistentry> + + <varlistentry id="tcache.destroy"> + <term> + <mallctl>tcache.destroy</mallctl> + (<type>unsigned</type>) + <literal>-w</literal> + </term> + <listitem><para>Flush the specified thread-specific cache (tcache) and + make the identifier available for use during a future tcache creation. + </para></listitem> + </varlistentry> + + <varlistentry id="arena.i.initialized"> + <term> + <mallctl>arena.<i>.initialized</mallctl> + (<type>bool</type>) + <literal>r-</literal> + </term> + <listitem><para>Get whether the specified arena's statistics are + initialized (i.e. the arena was initialized prior to the current epoch). + This interface can also be nominally used to query whether the merged + statistics corresponding to <constant>MALLCTL_ARENAS_ALL</constant> are + initialized (always true).</para></listitem> + </varlistentry> + + <varlistentry id="arena.i.decay"> + <term> + <mallctl>arena.<i>.decay</mallctl> + (<type>void</type>) + <literal>--</literal> + </term> + <listitem><para>Trigger decay-based purging of unused dirty/muzzy pages + for arena <i>, or for all arenas if <i> equals + <constant>MALLCTL_ARENAS_ALL</constant>. The proportion of unused + dirty/muzzy pages to be purged depends on the current time; see <link + linkend="opt.dirty_decay_ms"><mallctl>opt.dirty_decay_ms</mallctl></link> + and <link + linkend="opt.muzzy_decay_ms"><mallctl>opt.muzy_decay_ms</mallctl></link> + for details.</para></listitem> + </varlistentry> + + <varlistentry id="arena.i.purge"> + <term> + <mallctl>arena.<i>.purge</mallctl> + (<type>void</type>) + <literal>--</literal> + </term> + <listitem><para>Purge all unused dirty pages for arena <i>, or for + all arenas if <i> equals <constant>MALLCTL_ARENAS_ALL</constant>. + </para></listitem> + </varlistentry> + + <varlistentry id="arena.i.reset"> + <term> + <mallctl>arena.<i>.reset</mallctl> + (<type>void</type>) + <literal>--</literal> + </term> + <listitem><para>Discard all of the arena's extant allocations. This + interface can only be used with arenas explicitly created via <link + linkend="arenas.create"><mallctl>arenas.create</mallctl></link>. None + of the arena's discarded/cached allocations may accessed afterward. As + part of this requirement, all thread caches which were used to + allocate/deallocate in conjunction with the arena must be flushed + beforehand.</para></listitem> + </varlistentry> + + <varlistentry id="arena.i.destroy"> + <term> + <mallctl>arena.<i>.destroy</mallctl> + (<type>void</type>) + <literal>--</literal> + </term> + <listitem><para>Destroy the arena. Discard all of the arena's extant + allocations using the same mechanism as for <link + linkend="arena.i.reset"><mallctl>arena.<i>.reset</mallctl></link> + (with all the same constraints and side effects), merge the arena stats + into those accessible at arena index + <constant>MALLCTL_ARENAS_DESTROYED</constant>, and then completely + discard all metadata associated with the arena. Future calls to <link + linkend="arenas.create"><mallctl>arenas.create</mallctl></link> may + recycle the arena index. Destruction will fail if any threads are + currently associated with the arena as a result of calls to <link + linkend="thread.arena"><mallctl>thread.arena</mallctl></link>.</para></listitem> + </varlistentry> + + <varlistentry id="arena.i.dss"> + <term> + <mallctl>arena.<i>.dss</mallctl> + (<type>const char *</type>) + <literal>rw</literal> + </term> + <listitem><para>Set the precedence of dss allocation as related to mmap + allocation for arena <i>, or for all arenas if <i> equals + <constant>MALLCTL_ARENAS_ALL</constant>. See <link + linkend="opt.dss"><mallctl>opt.dss</mallctl></link> for supported + settings.</para></listitem> + </varlistentry> + + <varlistentry id="arena.i.dirty_decay_ms"> + <term> + <mallctl>arena.<i>.dirty_decay_ms</mallctl> + (<type>ssize_t</type>) + <literal>rw</literal> + </term> + <listitem><para>Current per-arena approximate time in milliseconds from + the creation of a set of unused dirty pages until an equivalent set of + unused dirty pages is purged and/or reused. Each time this interface is + set, all currently unused dirty pages are considered to have fully + decayed, which causes immediate purging of all unused dirty pages unless + the decay time is set to -1 (i.e. purging disabled). See <link + linkend="opt.dirty_decay_ms"><mallctl>opt.dirty_decay_ms</mallctl></link> + for additional information.</para></listitem> + </varlistentry> + + <varlistentry id="arena.i.muzzy_decay_ms"> + <term> + <mallctl>arena.<i>.muzzy_decay_ms</mallctl> + (<type>ssize_t</type>) + <literal>rw</literal> + </term> + <listitem><para>Current per-arena approximate time in milliseconds from + the creation of a set of unused muzzy pages until an equivalent set of + unused muzzy pages is purged and/or reused. Each time this interface is + set, all currently unused muzzy pages are considered to have fully + decayed, which causes immediate purging of all unused muzzy pages unless + the decay time is set to -1 (i.e. purging disabled). See <link + linkend="opt.muzzy_decay_ms"><mallctl>opt.muzzy_decay_ms</mallctl></link> + for additional information.</para></listitem> + </varlistentry> + + <varlistentry id="arena.i.retain_grow_limit"> + <term> + <mallctl>arena.<i>.retain_grow_limit</mallctl> + (<type>size_t</type>) + <literal>rw</literal> + </term> + <listitem><para>Maximum size to grow retained region (only relevant when + <link linkend="opt.retain"><mallctl>opt.retain</mallctl></link> is + enabled). This controls the maximum increment to expand virtual memory, + or allocation through <link + linkend="arena.i.extent_hooks"><mallctl>arena.<i>extent_hooks</mallctl></link>. + In particular, if customized extent hooks reserve physical memory + (e.g. 1G huge pages), this is useful to control the allocation hook's + input size. The default is no limit.</para></listitem> + </varlistentry> + + <varlistentry id="arena.i.extent_hooks"> + <term> + <mallctl>arena.<i>.extent_hooks</mallctl> + (<type>extent_hooks_t *</type>) + <literal>rw</literal> + </term> + <listitem><para>Get or set the extent management hook functions for + arena <i>. The functions must be capable of operating on all + extant extents associated with arena <i>, usually by passing + unknown extents to the replaced functions. In practice, it is feasible + to control allocation for arenas explicitly created via <link + linkend="arenas.create"><mallctl>arenas.create</mallctl></link> such + that all extents originate from an application-supplied extent allocator + (by specifying the custom extent hook functions during arena creation). + However, the API guarantees for the automatically created arenas may be + relaxed -- hooks set there may be called in a "best effort" fashion; in + addition there may be extents created prior to the application having an + opportunity to take over extent allocation.</para> + + <programlisting language="C"><![CDATA[ +typedef extent_hooks_s extent_hooks_t; +struct extent_hooks_s { + extent_alloc_t *alloc; + extent_dalloc_t *dalloc; + extent_destroy_t *destroy; + extent_commit_t *commit; + extent_decommit_t *decommit; + extent_purge_t *purge_lazy; + extent_purge_t *purge_forced; + extent_split_t *split; + extent_merge_t *merge; +};]]></programlisting> + <para>The <type>extent_hooks_t</type> structure comprises function + pointers which are described individually below. jemalloc uses these + functions to manage extent lifetime, which starts off with allocation of + mapped committed memory, in the simplest case followed by deallocation. + However, there are performance and platform reasons to retain extents + for later reuse. Cleanup attempts cascade from deallocation to decommit + to forced purging to lazy purging, which gives the extent management + functions opportunities to reject the most permanent cleanup operations + in favor of less permanent (and often less costly) operations. All + operations except allocation can be universally opted out of by setting + the hook pointers to <constant>NULL</constant>, or selectively opted out + of by returning failure. Note that once the extent hook is set, the + structure is accessed directly by the associated arenas, so it must + remain valid for the entire lifetime of the arenas.</para> + + <funcsynopsis><funcprototype> + <funcdef>typedef void *<function>(extent_alloc_t)</function></funcdef> + <paramdef>extent_hooks_t *<parameter>extent_hooks</parameter></paramdef> + <paramdef>void *<parameter>new_addr</parameter></paramdef> + <paramdef>size_t <parameter>size</parameter></paramdef> + <paramdef>size_t <parameter>alignment</parameter></paramdef> + <paramdef>bool *<parameter>zero</parameter></paramdef> + <paramdef>bool *<parameter>commit</parameter></paramdef> + <paramdef>unsigned <parameter>arena_ind</parameter></paramdef> + </funcprototype></funcsynopsis> + <literallayout></literallayout> + <para>An extent allocation function conforms to the + <type>extent_alloc_t</type> type and upon success returns a pointer to + <parameter>size</parameter> bytes of mapped memory on behalf of arena + <parameter>arena_ind</parameter> such that the extent's base address is + a multiple of <parameter>alignment</parameter>, as well as setting + <parameter>*zero</parameter> to indicate whether the extent is zeroed + and <parameter>*commit</parameter> to indicate whether the extent is + committed. Upon error the function returns <constant>NULL</constant> + and leaves <parameter>*zero</parameter> and + <parameter>*commit</parameter> unmodified. The + <parameter>size</parameter> parameter is always a multiple of the page + size. The <parameter>alignment</parameter> parameter is always a power + of two at least as large as the page size. Zeroing is mandatory if + <parameter>*zero</parameter> is true upon function entry. Committing is + mandatory if <parameter>*commit</parameter> is true upon function entry. + If <parameter>new_addr</parameter> is not <constant>NULL</constant>, the + returned pointer must be <parameter>new_addr</parameter> on success or + <constant>NULL</constant> on error. Committed memory may be committed + in absolute terms as on a system that does not overcommit, or in + implicit terms as on a system that overcommits and satisfies physical + memory needs on demand via soft page faults. Note that replacing the + default extent allocation function makes the arena's <link + linkend="arena.i.dss"><mallctl>arena.<i>.dss</mallctl></link> + setting irrelevant.</para> + + <funcsynopsis><funcprototype> + <funcdef>typedef bool <function>(extent_dalloc_t)</function></funcdef> + <paramdef>extent_hooks_t *<parameter>extent_hooks</parameter></paramdef> + <paramdef>void *<parameter>addr</parameter></paramdef> + <paramdef>size_t <parameter>size</parameter></paramdef> + <paramdef>bool <parameter>committed</parameter></paramdef> + <paramdef>unsigned <parameter>arena_ind</parameter></paramdef> + </funcprototype></funcsynopsis> + <literallayout></literallayout> + <para> + An extent deallocation function conforms to the + <type>extent_dalloc_t</type> type and deallocates an extent at given + <parameter>addr</parameter> and <parameter>size</parameter> with + <parameter>committed</parameter>/decommited memory as indicated, on + behalf of arena <parameter>arena_ind</parameter>, returning false upon + success. If the function returns true, this indicates opt-out from + deallocation; the virtual memory mapping associated with the extent + remains mapped, in the same commit state, and available for future use, + in which case it will be automatically retained for later reuse.</para> + + <funcsynopsis><funcprototype> + <funcdef>typedef void <function>(extent_destroy_t)</function></funcdef> + <paramdef>extent_hooks_t *<parameter>extent_hooks</parameter></paramdef> + <paramdef>void *<parameter>addr</parameter></paramdef> + <paramdef>size_t <parameter>size</parameter></paramdef> + <paramdef>bool <parameter>committed</parameter></paramdef> + <paramdef>unsigned <parameter>arena_ind</parameter></paramdef> + </funcprototype></funcsynopsis> + <literallayout></literallayout> + <para> + An extent destruction function conforms to the + <type>extent_destroy_t</type> type and unconditionally destroys an + extent at given <parameter>addr</parameter> and + <parameter>size</parameter> with + <parameter>committed</parameter>/decommited memory as indicated, on + behalf of arena <parameter>arena_ind</parameter>. This function may be + called to destroy retained extents during arena destruction (see <link + linkend="arena.i.destroy"><mallctl>arena.<i>.destroy</mallctl></link>).</para> + + <funcsynopsis><funcprototype> + <funcdef>typedef bool <function>(extent_commit_t)</function></funcdef> + <paramdef>extent_hooks_t *<parameter>extent_hooks</parameter></paramdef> + <paramdef>void *<parameter>addr</parameter></paramdef> + <paramdef>size_t <parameter>size</parameter></paramdef> + <paramdef>size_t <parameter>offset</parameter></paramdef> + <paramdef>size_t <parameter>length</parameter></paramdef> + <paramdef>unsigned <parameter>arena_ind</parameter></paramdef> + </funcprototype></funcsynopsis> + <literallayout></literallayout> + <para>An extent commit function conforms to the + <type>extent_commit_t</type> type and commits zeroed physical memory to + back pages within an extent at given <parameter>addr</parameter> and + <parameter>size</parameter> at <parameter>offset</parameter> bytes, + extending for <parameter>length</parameter> on behalf of arena + <parameter>arena_ind</parameter>, returning false upon success. + Committed memory may be committed in absolute terms as on a system that + does not overcommit, or in implicit terms as on a system that + overcommits and satisfies physical memory needs on demand via soft page + faults. If the function returns true, this indicates insufficient + physical memory to satisfy the request.</para> + + <funcsynopsis><funcprototype> + <funcdef>typedef bool <function>(extent_decommit_t)</function></funcdef> + <paramdef>extent_hooks_t *<parameter>extent_hooks</parameter></paramdef> + <paramdef>void *<parameter>addr</parameter></paramdef> + <paramdef>size_t <parameter>size</parameter></paramdef> + <paramdef>size_t <parameter>offset</parameter></paramdef> + <paramdef>size_t <parameter>length</parameter></paramdef> + <paramdef>unsigned <parameter>arena_ind</parameter></paramdef> + </funcprototype></funcsynopsis> + <literallayout></literallayout> + <para>An extent decommit function conforms to the + <type>extent_decommit_t</type> type and decommits any physical memory + that is backing pages within an extent at given + <parameter>addr</parameter> and <parameter>size</parameter> at + <parameter>offset</parameter> bytes, extending for + <parameter>length</parameter> on behalf of arena + <parameter>arena_ind</parameter>, returning false upon success, in which + case the pages will be committed via the extent commit function before + being reused. If the function returns true, this indicates opt-out from + decommit; the memory remains committed and available for future use, in + which case it will be automatically retained for later reuse.</para> + + <funcsynopsis><funcprototype> + <funcdef>typedef bool <function>(extent_purge_t)</function></funcdef> + <paramdef>extent_hooks_t *<parameter>extent_hooks</parameter></paramdef> + <paramdef>void *<parameter>addr</parameter></paramdef> + <paramdef>size_t <parameter>size</parameter></paramdef> + <paramdef>size_t <parameter>offset</parameter></paramdef> + <paramdef>size_t <parameter>length</parameter></paramdef> + <paramdef>unsigned <parameter>arena_ind</parameter></paramdef> + </funcprototype></funcsynopsis> + <literallayout></literallayout> + <para>An extent purge function conforms to the + <type>extent_purge_t</type> type and discards physical pages + within the virtual memory mapping associated with an extent at given + <parameter>addr</parameter> and <parameter>size</parameter> at + <parameter>offset</parameter> bytes, extending for + <parameter>length</parameter> on behalf of arena + <parameter>arena_ind</parameter>. A lazy extent purge function (e.g. + implemented via + <function>madvise(<parameter>...</parameter><parameter><constant>MADV_FREE</constant></parameter>)</function>) + can delay purging indefinitely and leave the pages within the purged + virtual memory range in an indeterminite state, whereas a forced extent + purge function immediately purges, and the pages within the virtual + memory range will be zero-filled the next time they are accessed. If + the function returns true, this indicates failure to purge.</para> + + <funcsynopsis><funcprototype> + <funcdef>typedef bool <function>(extent_split_t)</function></funcdef> + <paramdef>extent_hooks_t *<parameter>extent_hooks</parameter></paramdef> + <paramdef>void *<parameter>addr</parameter></paramdef> + <paramdef>size_t <parameter>size</parameter></paramdef> + <paramdef>size_t <parameter>size_a</parameter></paramdef> + <paramdef>size_t <parameter>size_b</parameter></paramdef> + <paramdef>bool <parameter>committed</parameter></paramdef> + <paramdef>unsigned <parameter>arena_ind</parameter></paramdef> + </funcprototype></funcsynopsis> + <literallayout></literallayout> + <para>An extent split function conforms to the + <type>extent_split_t</type> type and optionally splits an extent at + given <parameter>addr</parameter> and <parameter>size</parameter> into + two adjacent extents, the first of <parameter>size_a</parameter> bytes, + and the second of <parameter>size_b</parameter> bytes, operating on + <parameter>committed</parameter>/decommitted memory as indicated, on + behalf of arena <parameter>arena_ind</parameter>, returning false upon + success. If the function returns true, this indicates that the extent + remains unsplit and therefore should continue to be operated on as a + whole.</para> + + <funcsynopsis><funcprototype> + <funcdef>typedef bool <function>(extent_merge_t)</function></funcdef> + <paramdef>extent_hooks_t *<parameter>extent_hooks</parameter></paramdef> + <paramdef>void *<parameter>addr_a</parameter></paramdef> + <paramdef>size_t <parameter>size_a</parameter></paramdef> + <paramdef>void *<parameter>addr_b</parameter></paramdef> + <paramdef>size_t <parameter>size_b</parameter></paramdef> + <paramdef>bool <parameter>committed</parameter></paramdef> + <paramdef>unsigned <parameter>arena_ind</parameter></paramdef> + </funcprototype></funcsynopsis> + <literallayout></literallayout> + <para>An extent merge function conforms to the + <type>extent_merge_t</type> type and optionally merges adjacent extents, + at given <parameter>addr_a</parameter> and <parameter>size_a</parameter> + with given <parameter>addr_b</parameter> and + <parameter>size_b</parameter> into one contiguous extent, operating on + <parameter>committed</parameter>/decommitted memory as indicated, on + behalf of arena <parameter>arena_ind</parameter>, returning false upon + success. If the function returns true, this indicates that the extents + remain distinct mappings and therefore should continue to be operated on + independently.</para> + </listitem> + </varlistentry> + + <varlistentry id="arenas.narenas"> + <term> + <mallctl>arenas.narenas</mallctl> + (<type>unsigned</type>) + <literal>r-</literal> + </term> + <listitem><para>Current limit on number of arenas.</para></listitem> + </varlistentry> + + <varlistentry id="arenas.dirty_decay_ms"> + <term> + <mallctl>arenas.dirty_decay_ms</mallctl> + (<type>ssize_t</type>) + <literal>rw</literal> + </term> + <listitem><para>Current default per-arena approximate time in + milliseconds from the creation of a set of unused dirty pages until an + equivalent set of unused dirty pages is purged and/or reused, used to + initialize <link + linkend="arena.i.dirty_decay_ms"><mallctl>arena.<i>.dirty_decay_ms</mallctl></link> + during arena creation. See <link + linkend="opt.dirty_decay_ms"><mallctl>opt.dirty_decay_ms</mallctl></link> + for additional information.</para></listitem> + </varlistentry> + + <varlistentry id="arenas.muzzy_decay_ms"> + <term> + <mallctl>arenas.muzzy_decay_ms</mallctl> + (<type>ssize_t</type>) + <literal>rw</literal> + </term> + <listitem><para>Current default per-arena approximate time in + milliseconds from the creation of a set of unused muzzy pages until an + equivalent set of unused muzzy pages is purged and/or reused, used to + initialize <link + linkend="arena.i.muzzy_decay_ms"><mallctl>arena.<i>.muzzy_decay_ms</mallctl></link> + during arena creation. See <link + linkend="opt.muzzy_decay_ms"><mallctl>opt.muzzy_decay_ms</mallctl></link> + for additional information.</para></listitem> + </varlistentry> + + <varlistentry id="arenas.quantum"> + <term> + <mallctl>arenas.quantum</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + </term> + <listitem><para>Quantum size.</para></listitem> + </varlistentry> + + <varlistentry id="arenas.page"> + <term> + <mallctl>arenas.page</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + </term> + <listitem><para>Page size.</para></listitem> + </varlistentry> + + <varlistentry id="arenas.tcache_max"> + <term> + <mallctl>arenas.tcache_max</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + </term> + <listitem><para>Maximum thread-cached size class.</para></listitem> + </varlistentry> + + <varlistentry id="arenas.nbins"> + <term> + <mallctl>arenas.nbins</mallctl> + (<type>unsigned</type>) + <literal>r-</literal> + </term> + <listitem><para>Number of bin size classes.</para></listitem> + </varlistentry> + + <varlistentry id="arenas.nhbins"> + <term> + <mallctl>arenas.nhbins</mallctl> + (<type>unsigned</type>) + <literal>r-</literal> + </term> + <listitem><para>Total number of thread cache bin size + classes.</para></listitem> + </varlistentry> + + <varlistentry id="arenas.bin.i.size"> + <term> + <mallctl>arenas.bin.<i>.size</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + </term> + <listitem><para>Maximum size supported by size class.</para></listitem> + </varlistentry> + + <varlistentry id="arenas.bin.i.nregs"> + <term> + <mallctl>arenas.bin.<i>.nregs</mallctl> + (<type>uint32_t</type>) + <literal>r-</literal> + </term> + <listitem><para>Number of regions per slab.</para></listitem> + </varlistentry> + + <varlistentry id="arenas.bin.i.slab_size"> + <term> + <mallctl>arenas.bin.<i>.slab_size</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + </term> + <listitem><para>Number of bytes per slab.</para></listitem> + </varlistentry> + + <varlistentry id="arenas.nlextents"> + <term> + <mallctl>arenas.nlextents</mallctl> + (<type>unsigned</type>) + <literal>r-</literal> + </term> + <listitem><para>Total number of large size classes.</para></listitem> + </varlistentry> + + <varlistentry id="arenas.lextent.i.size"> + <term> + <mallctl>arenas.lextent.<i>.size</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + </term> + <listitem><para>Maximum size supported by this large size + class.</para></listitem> + </varlistentry> + + <varlistentry id="arenas.create"> + <term> + <mallctl>arenas.create</mallctl> + (<type>unsigned</type>, <type>extent_hooks_t *</type>) + <literal>rw</literal> + </term> + <listitem><para>Explicitly create a new arena outside the range of + automatically managed arenas, with optionally specified extent hooks, + and return the new arena index.</para></listitem> + </varlistentry> + + <varlistentry id="arenas.lookup"> + <term> + <mallctl>arenas.lookup</mallctl> + (<type>unsigned</type>, <type>void*</type>) + <literal>rw</literal> + </term> + <listitem><para>Index of the arena to which an allocation belongs to.</para></listitem> + </varlistentry> + + <varlistentry id="prof.thread_active_init"> + <term> + <mallctl>prof.thread_active_init</mallctl> + (<type>bool</type>) + <literal>rw</literal> + [<option>--enable-prof</option>] + </term> + <listitem><para>Control the initial setting for <link + linkend="thread.prof.active"><mallctl>thread.prof.active</mallctl></link> + in newly created threads. See the <link + linkend="opt.prof_thread_active_init"><mallctl>opt.prof_thread_active_init</mallctl></link> + option for additional information.</para></listitem> + </varlistentry> + + <varlistentry id="prof.active"> + <term> + <mallctl>prof.active</mallctl> + (<type>bool</type>) + <literal>rw</literal> + [<option>--enable-prof</option>] + </term> + <listitem><para>Control whether sampling is currently active. See the + <link + linkend="opt.prof_active"><mallctl>opt.prof_active</mallctl></link> + option for additional information, as well as the interrelated <link + linkend="thread.prof.active"><mallctl>thread.prof.active</mallctl></link> + mallctl.</para></listitem> + </varlistentry> + + <varlistentry id="prof.dump"> + <term> + <mallctl>prof.dump</mallctl> + (<type>const char *</type>) + <literal>-w</literal> + [<option>--enable-prof</option>] + </term> + <listitem><para>Dump a memory profile to the specified file, or if NULL + is specified, to a file according to the pattern + <filename><prefix>.<pid>.<seq>.m<mseq>.heap</filename>, + where <literal><prefix></literal> is controlled by the + <link + linkend="opt.prof_prefix"><mallctl>opt.prof_prefix</mallctl></link> + option.</para></listitem> + </varlistentry> + + <varlistentry id="prof.gdump"> + <term> + <mallctl>prof.gdump</mallctl> + (<type>bool</type>) + <literal>rw</literal> + [<option>--enable-prof</option>] + </term> + <listitem><para>When enabled, trigger a memory profile dump every time + the total virtual memory exceeds the previous maximum. Profiles are + dumped to files named according to the pattern + <filename><prefix>.<pid>.<seq>.u<useq>.heap</filename>, + where <literal><prefix></literal> is controlled by the <link + linkend="opt.prof_prefix"><mallctl>opt.prof_prefix</mallctl></link> + option.</para></listitem> + </varlistentry> + + <varlistentry id="prof.reset"> + <term> + <mallctl>prof.reset</mallctl> + (<type>size_t</type>) + <literal>-w</literal> + [<option>--enable-prof</option>] + </term> + <listitem><para>Reset all memory profile statistics, and optionally + update the sample rate (see <link + linkend="opt.lg_prof_sample"><mallctl>opt.lg_prof_sample</mallctl></link> + and <link + linkend="prof.lg_sample"><mallctl>prof.lg_sample</mallctl></link>). + </para></listitem> + </varlistentry> + + <varlistentry id="prof.lg_sample"> + <term> + <mallctl>prof.lg_sample</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-prof</option>] + </term> + <listitem><para>Get the current sample rate (see <link + linkend="opt.lg_prof_sample"><mallctl>opt.lg_prof_sample</mallctl></link>). + </para></listitem> + </varlistentry> + + <varlistentry id="prof.interval"> + <term> + <mallctl>prof.interval</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-prof</option>] + </term> + <listitem><para>Average number of bytes allocated between + interval-based profile dumps. See the + <link + linkend="opt.lg_prof_interval"><mallctl>opt.lg_prof_interval</mallctl></link> + option for additional information.</para></listitem> + </varlistentry> + + <varlistentry id="stats.allocated"> + <term> + <mallctl>stats.allocated</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Total number of bytes allocated by the + application.</para></listitem> + </varlistentry> + + <varlistentry id="stats.active"> + <term> + <mallctl>stats.active</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Total number of bytes in active pages allocated by the + application. This is a multiple of the page size, and greater than or + equal to <link + linkend="stats.allocated"><mallctl>stats.allocated</mallctl></link>. + This does not include <link linkend="stats.arenas.i.pdirty"> + <mallctl>stats.arenas.<i>.pdirty</mallctl></link>, + <link linkend="stats.arenas.i.pmuzzy"> + <mallctl>stats.arenas.<i>.pmuzzy</mallctl></link>, nor pages + entirely devoted to allocator metadata.</para></listitem> + </varlistentry> + + <varlistentry id="stats.metadata"> + <term> + <mallctl>stats.metadata</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Total number of bytes dedicated to metadata, which + comprise base allocations used for bootstrap-sensitive allocator + metadata structures (see <link + linkend="stats.arenas.i.base"><mallctl>stats.arenas.<i>.base</mallctl></link>) + and internal allocations (see <link + linkend="stats.arenas.i.internal"><mallctl>stats.arenas.<i>.internal</mallctl></link>). + Transparent huge page (enabled with <link + linkend="opt.metadata_thp">opt.metadata_thp</link>) usage is not + considered.</para></listitem> + </varlistentry> + + <varlistentry id="stats.metadata_thp"> + <term> + <mallctl>stats.metadata_thp</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Number of transparent huge pages (THP) used for + metadata. See <link + linkend="stats.metadata"><mallctl>stats.metadata</mallctl></link> and + <link linkend="opt.metadata_thp">opt.metadata_thp</link>) for + details.</para></listitem> + </varlistentry> + + <varlistentry id="stats.resident"> + <term> + <mallctl>stats.resident</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Maximum number of bytes in physically resident data + pages mapped by the allocator, comprising all pages dedicated to + allocator metadata, pages backing active allocations, and unused dirty + pages. This is a maximum rather than precise because pages may not + actually be physically resident if they correspond to demand-zeroed + virtual memory that has not yet been touched. This is a multiple of the + page size, and is larger than <link + linkend="stats.active"><mallctl>stats.active</mallctl></link>.</para></listitem> + </varlistentry> + + <varlistentry id="stats.mapped"> + <term> + <mallctl>stats.mapped</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Total number of bytes in active extents mapped by the + allocator. This is larger than <link + linkend="stats.active"><mallctl>stats.active</mallctl></link>. This + does not include inactive extents, even those that contain unused dirty + pages, which means that there is no strict ordering between this and + <link + linkend="stats.resident"><mallctl>stats.resident</mallctl></link>.</para></listitem> + </varlistentry> + + <varlistentry id="stats.retained"> + <term> + <mallctl>stats.retained</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Total number of bytes in virtual memory mappings that + were retained rather than being returned to the operating system via + e.g. <citerefentry><refentrytitle>munmap</refentrytitle> + <manvolnum>2</manvolnum></citerefentry> or similar. Retained virtual + memory is typically untouched, decommitted, or purged, so it has no + strongly associated physical memory (see <link + linkend="arena.i.extent_hooks">extent hooks</link> for details). + Retained memory is excluded from mapped memory statistics, e.g. <link + linkend="stats.mapped"><mallctl>stats.mapped</mallctl></link>. + </para></listitem> + </varlistentry> + + <varlistentry id="stats.background_thread.num_threads"> + <term> + <mallctl>stats.background_thread.num_threads</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para> Number of <link linkend="background_thread">background + threads</link> running currently.</para></listitem> + </varlistentry> + + <varlistentry id="stats.background_thread.num_runs"> + <term> + <mallctl>stats.background_thread.num_runs</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para> Total number of runs from all <link + linkend="background_thread">background threads</link>.</para></listitem> + </varlistentry> + + <varlistentry id="stats.background_thread.run_interval"> + <term> + <mallctl>stats.background_thread.run_interval</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para> Average run interval in nanoseconds of <link + linkend="background_thread">background threads</link>.</para></listitem> + </varlistentry> + + <varlistentry id="stats.mutexes.ctl"> + <term> + <mallctl>stats.mutexes.ctl.{counter};</mallctl> + (<type>counter specific type</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Statistics on <varname>ctl</varname> mutex (global + scope; mallctl related). <mallctl>{counter}</mallctl> is one of the + counters below:</para> + <varlistentry id="mutex_counters"> + <listitem><para><varname>num_ops</varname> (<type>uint64_t</type>): + Total number of lock acquisition operations on this mutex.</para> + + <para><varname>num_spin_acq</varname> (<type>uint64_t</type>): Number + of times the mutex was spin-acquired. When the mutex is currently + locked and cannot be acquired immediately, a short period of + spin-retry within jemalloc will be performed. Acquired through spin + generally means the contention was lightweight and not causing context + switches.</para> + + <para><varname>num_wait</varname> (<type>uint64_t</type>): Number of + times the mutex was wait-acquired, which means the mutex contention + was not solved by spin-retry, and blocking operation was likely + involved in order to acquire the mutex. This event generally implies + higher cost / longer delay, and should be investigated if it happens + often.</para> + + <para><varname>max_wait_time</varname> (<type>uint64_t</type>): + Maximum length of time in nanoseconds spent on a single wait-acquired + lock operation. Note that to avoid profiling overhead on the common + path, this does not consider spin-acquired cases.</para> + + <para><varname>total_wait_time</varname> (<type>uint64_t</type>): + Cumulative time in nanoseconds spent on wait-acquired lock operations. + Similarly, spin-acquired cases are not considered.</para> + + <para><varname>max_num_thds</varname> (<type>uint32_t</type>): Maximum + number of threads waiting on this mutex simultaneously. Similarly, + spin-acquired cases are not considered.</para> + + <para><varname>num_owner_switch</varname> (<type>uint64_t</type>): + Number of times the current mutex owner is different from the previous + one. This event does not generally imply an issue; rather it is an + indicator of how often the protected data are accessed by different + threads. + </para> + </listitem> + </varlistentry> + </listitem> + </varlistentry> + + <varlistentry id="stats.mutexes.background_thread"> + <term> + <mallctl>stats.mutexes.background_thread.{counter}</mallctl> + (<type>counter specific type</type>) <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Statistics on <varname>background_thread</varname> mutex + (global scope; <link + linkend="background_thread"><mallctl>background_thread</mallctl></link> + related). <mallctl>{counter}</mallctl> is one of the counters in <link + linkend="mutex_counters">mutex profiling + counters</link>.</para></listitem> + </varlistentry> + + <varlistentry id="stats.mutexes.prof"> + <term> + <mallctl>stats.mutexes.prof.{counter}</mallctl> + (<type>counter specific type</type>) <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Statistics on <varname>prof</varname> mutex (global + scope; profiling related). <mallctl>{counter}</mallctl> is one of the + counters in <link linkend="mutex_counters">mutex profiling + counters</link>.</para></listitem> + </varlistentry> + + <varlistentry id="stats.mutexes.reset"> + <term> + <mallctl>stats.mutexes.reset</mallctl> + (<type>void</type>) <literal>--</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Reset all mutex profile statistics, including global + mutexes, arena mutexes and bin mutexes.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.dss"> + <term> + <mallctl>stats.arenas.<i>.dss</mallctl> + (<type>const char *</type>) + <literal>r-</literal> + </term> + <listitem><para>dss (<citerefentry><refentrytitle>sbrk</refentrytitle> + <manvolnum>2</manvolnum></citerefentry>) allocation precedence as + related to <citerefentry><refentrytitle>mmap</refentrytitle> + <manvolnum>2</manvolnum></citerefentry> allocation. See <link + linkend="opt.dss"><mallctl>opt.dss</mallctl></link> for details. + </para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.dirty_decay_ms"> + <term> + <mallctl>stats.arenas.<i>.dirty_decay_ms</mallctl> + (<type>ssize_t</type>) + <literal>r-</literal> + </term> + <listitem><para>Approximate time in milliseconds from the creation of a + set of unused dirty pages until an equivalent set of unused dirty pages + is purged and/or reused. See <link + linkend="opt.dirty_decay_ms"><mallctl>opt.dirty_decay_ms</mallctl></link> + for details.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.muzzy_decay_ms"> + <term> + <mallctl>stats.arenas.<i>.muzzy_decay_ms</mallctl> + (<type>ssize_t</type>) + <literal>r-</literal> + </term> + <listitem><para>Approximate time in milliseconds from the creation of a + set of unused muzzy pages until an equivalent set of unused muzzy pages + is purged and/or reused. See <link + linkend="opt.muzzy_decay_ms"><mallctl>opt.muzzy_decay_ms</mallctl></link> + for details.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.nthreads"> + <term> + <mallctl>stats.arenas.<i>.nthreads</mallctl> + (<type>unsigned</type>) + <literal>r-</literal> + </term> + <listitem><para>Number of threads currently assigned to + arena.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.uptime"> + <term> + <mallctl>stats.arenas.<i>.uptime</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + </term> + <listitem><para>Time elapsed (in nanoseconds) since the arena was + created. If <i> equals <constant>0</constant> or + <constant>MALLCTL_ARENAS_ALL</constant>, this is the uptime since malloc + initialization.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.pactive"> + <term> + <mallctl>stats.arenas.<i>.pactive</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + </term> + <listitem><para>Number of pages in active extents.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.pdirty"> + <term> + <mallctl>stats.arenas.<i>.pdirty</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + </term> + <listitem><para>Number of pages within unused extents that are + potentially dirty, and for which <function>madvise()</function> or + similar has not been called. See <link + linkend="opt.dirty_decay_ms"><mallctl>opt.dirty_decay_ms</mallctl></link> + for a description of dirty pages.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.pmuzzy"> + <term> + <mallctl>stats.arenas.<i>.pmuzzy</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + </term> + <listitem><para>Number of pages within unused extents that are muzzy. + See <link + linkend="opt.muzzy_decay_ms"><mallctl>opt.muzzy_decay_ms</mallctl></link> + for a description of muzzy pages.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.mapped"> + <term> + <mallctl>stats.arenas.<i>.mapped</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Number of mapped bytes.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.retained"> + <term> + <mallctl>stats.arenas.<i>.retained</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Number of retained bytes. See <link + linkend="stats.retained"><mallctl>stats.retained</mallctl></link> for + details.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.extent_avail"> + <term> + <mallctl>stats.arenas.<i>.extent_avail</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Number of allocated (but unused) extent structs in this + arena.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.base"> + <term> + <mallctl>stats.arenas.<i>.base</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para> + Number of bytes dedicated to bootstrap-sensitive allocator metadata + structures.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.internal"> + <term> + <mallctl>stats.arenas.<i>.internal</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Number of bytes dedicated to internal allocations. + Internal allocations differ from application-originated allocations in + that they are for internal use, and that they are omitted from heap + profiles.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.metadata_thp"> + <term> + <mallctl>stats.arenas.<i>.metadata_thp</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Number of transparent huge pages (THP) used for + metadata. See <link linkend="opt.metadata_thp">opt.metadata_thp</link> + for details.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.resident"> + <term> + <mallctl>stats.arenas.<i>.resident</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Maximum number of bytes in physically resident data + pages mapped by the arena, comprising all pages dedicated to allocator + metadata, pages backing active allocations, and unused dirty pages. + This is a maximum rather than precise because pages may not actually be + physically resident if they correspond to demand-zeroed virtual memory + that has not yet been touched. This is a multiple of the page + size.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.dirty_npurge"> + <term> + <mallctl>stats.arenas.<i>.dirty_npurge</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Number of dirty page purge sweeps performed. + </para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.dirty_nmadvise"> + <term> + <mallctl>stats.arenas.<i>.dirty_nmadvise</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Number of <function>madvise()</function> or similar + calls made to purge dirty pages.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.dirty_purged"> + <term> + <mallctl>stats.arenas.<i>.dirty_purged</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Number of dirty pages purged.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.muzzy_npurge"> + <term> + <mallctl>stats.arenas.<i>.muzzy_npurge</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Number of muzzy page purge sweeps performed. + </para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.muzzy_nmadvise"> + <term> + <mallctl>stats.arenas.<i>.muzzy_nmadvise</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Number of <function>madvise()</function> or similar + calls made to purge muzzy pages.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.muzzy_purged"> + <term> + <mallctl>stats.arenas.<i>.muzzy_purged</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Number of muzzy pages purged.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.small.allocated"> + <term> + <mallctl>stats.arenas.<i>.small.allocated</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Number of bytes currently allocated by small objects. + </para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.small.nmalloc"> + <term> + <mallctl>stats.arenas.<i>.small.nmalloc</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Cumulative number of times a small allocation was + requested from the arena's bins, whether to fill the relevant tcache if + <link linkend="opt.tcache"><mallctl>opt.tcache</mallctl></link> is + enabled, or to directly satisfy an allocation request + otherwise.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.small.ndalloc"> + <term> + <mallctl>stats.arenas.<i>.small.ndalloc</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Cumulative number of times a small allocation was + returned to the arena's bins, whether to flush the relevant tcache if + <link linkend="opt.tcache"><mallctl>opt.tcache</mallctl></link> is + enabled, or to directly deallocate an allocation + otherwise.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.small.nrequests"> + <term> + <mallctl>stats.arenas.<i>.small.nrequests</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Cumulative number of allocation requests satisfied by + all bin size classes.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.small.nfills"> + <term> + <mallctl>stats.arenas.<i>.small.nfills</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Cumulative number of tcache fills by all small size + classes.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.small.nflushes"> + <term> + <mallctl>stats.arenas.<i>.small.nflushes</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Cumulative number of tcache flushes by all small size + classes.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.large.allocated"> + <term> + <mallctl>stats.arenas.<i>.large.allocated</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Number of bytes currently allocated by large objects. + </para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.large.nmalloc"> + <term> + <mallctl>stats.arenas.<i>.large.nmalloc</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Cumulative number of times a large extent was allocated + from the arena, whether to fill the relevant tcache if <link + linkend="opt.tcache"><mallctl>opt.tcache</mallctl></link> is enabled and + the size class is within the range being cached, or to directly satisfy + an allocation request otherwise.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.large.ndalloc"> + <term> + <mallctl>stats.arenas.<i>.large.ndalloc</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Cumulative number of times a large extent was returned + to the arena, whether to flush the relevant tcache if <link + linkend="opt.tcache"><mallctl>opt.tcache</mallctl></link> is enabled and + the size class is within the range being cached, or to directly + deallocate an allocation otherwise.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.large.nrequests"> + <term> + <mallctl>stats.arenas.<i>.large.nrequests</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Cumulative number of allocation requests satisfied by + all large size classes.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.large.nfills"> + <term> + <mallctl>stats.arenas.<i>.large.nfills</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Cumulative number of tcache fills by all large size + classes.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.large.nflushes"> + <term> + <mallctl>stats.arenas.<i>.large.nflushes</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Cumulative number of tcache flushes by all large size + classes.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.bins.j.nmalloc"> + <term> + <mallctl>stats.arenas.<i>.bins.<j>.nmalloc</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Cumulative number of times a bin region of the + corresponding size class was allocated from the arena, whether to fill + the relevant tcache if <link + linkend="opt.tcache"><mallctl>opt.tcache</mallctl></link> is enabled, or + to directly satisfy an allocation request otherwise.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.bins.j.ndalloc"> + <term> + <mallctl>stats.arenas.<i>.bins.<j>.ndalloc</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Cumulative number of times a bin region of the + corresponding size class was returned to the arena, whether to flush the + relevant tcache if <link + linkend="opt.tcache"><mallctl>opt.tcache</mallctl></link> is enabled, or + to directly deallocate an allocation otherwise.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.bins.j.nrequests"> + <term> + <mallctl>stats.arenas.<i>.bins.<j>.nrequests</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Cumulative number of allocation requests satisfied by + bin regions of the corresponding size class.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.bins.j.curregs"> + <term> + <mallctl>stats.arenas.<i>.bins.<j>.curregs</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Current number of regions for this size + class.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.bins.j.nfills"> + <term> + <mallctl>stats.arenas.<i>.bins.<j>.nfills</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + </term> + <listitem><para>Cumulative number of tcache fills.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.bins.j.nflushes"> + <term> + <mallctl>stats.arenas.<i>.bins.<j>.nflushes</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + </term> + <listitem><para>Cumulative number of tcache flushes.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.bins.j.nslabs"> + <term> + <mallctl>stats.arenas.<i>.bins.<j>.nslabs</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Cumulative number of slabs created.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.bins.j.nreslabs"> + <term> + <mallctl>stats.arenas.<i>.bins.<j>.nreslabs</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Cumulative number of times the current slab from which + to allocate changed.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.bins.j.curslabs"> + <term> + <mallctl>stats.arenas.<i>.bins.<j>.curslabs</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Current number of slabs.</para></listitem> + </varlistentry> + + + <varlistentry id="stats.arenas.i.bins.j.nonfull_slabs"> + <term> + <mallctl>stats.arenas.<i>.bins.<j>.nonfull_slabs</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Current number of nonfull slabs.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.bins.mutex"> + <term> + <mallctl>stats.arenas.<i>.bins.<j>.mutex.{counter}</mallctl> + (<type>counter specific type</type>) <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Statistics on + <varname>arena.<i>.bins.<j></varname> mutex (arena bin + scope; bin operation related). <mallctl>{counter}</mallctl> is one of + the counters in <link linkend="mutex_counters">mutex profiling + counters</link>.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.extents.n"> + <term> + <mallctl>stats.arenas.<i>.extents.<j>.n{extent_type}</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para> Number of extents of the given type in this arena in + the bucket corresponding to page size index <j>. The extent type + is one of dirty, muzzy, or retained.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.extents.bytes"> + <term> + <mallctl>stats.arenas.<i>.extents.<j>.{extent_type}_bytes</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para> Sum of the bytes managed by extents of the given type + in this arena in the bucket corresponding to page size index <j>. + The extent type is one of dirty, muzzy, or retained.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.lextents.j.nmalloc"> + <term> + <mallctl>stats.arenas.<i>.lextents.<j>.nmalloc</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Cumulative number of times a large extent of the + corresponding size class was allocated from the arena, whether to fill + the relevant tcache if <link + linkend="opt.tcache"><mallctl>opt.tcache</mallctl></link> is enabled and + the size class is within the range being cached, or to directly satisfy + an allocation request otherwise.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.lextents.j.ndalloc"> + <term> + <mallctl>stats.arenas.<i>.lextents.<j>.ndalloc</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Cumulative number of times a large extent of the + corresponding size class was returned to the arena, whether to flush the + relevant tcache if <link + linkend="opt.tcache"><mallctl>opt.tcache</mallctl></link> is enabled and + the size class is within the range being cached, or to directly + deallocate an allocation otherwise.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.lextents.j.nrequests"> + <term> + <mallctl>stats.arenas.<i>.lextents.<j>.nrequests</mallctl> + (<type>uint64_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Cumulative number of allocation requests satisfied by + large extents of the corresponding size class.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.lextents.j.curlextents"> + <term> + <mallctl>stats.arenas.<i>.lextents.<j>.curlextents</mallctl> + (<type>size_t</type>) + <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Current number of large allocations for this size class. + </para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.mutexes.large"> + <term> + <mallctl>stats.arenas.<i>.mutexes.large.{counter}</mallctl> + (<type>counter specific type</type>) <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Statistics on <varname>arena.<i>.large</varname> + mutex (arena scope; large allocation related). + <mallctl>{counter}</mallctl> is one of the counters in <link + linkend="mutex_counters">mutex profiling + counters</link>.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.mutexes.extent_avail"> + <term> + <mallctl>stats.arenas.<i>.mutexes.extent_avail.{counter}</mallctl> + (<type>counter specific type</type>) <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Statistics on <varname>arena.<i>.extent_avail + </varname> mutex (arena scope; extent avail related). + <mallctl>{counter}</mallctl> is one of the counters in <link + linkend="mutex_counters">mutex profiling + counters</link>.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.mutexes.extents_dirty"> + <term> + <mallctl>stats.arenas.<i>.mutexes.extents_dirty.{counter}</mallctl> + (<type>counter specific type</type>) <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Statistics on <varname>arena.<i>.extents_dirty + </varname> mutex (arena scope; dirty extents related). + <mallctl>{counter}</mallctl> is one of the counters in <link + linkend="mutex_counters">mutex profiling + counters</link>.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.mutexes.extents_muzzy"> + <term> + <mallctl>stats.arenas.<i>.mutexes.extents_muzzy.{counter}</mallctl> + (<type>counter specific type</type>) <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Statistics on <varname>arena.<i>.extents_muzzy + </varname> mutex (arena scope; muzzy extents related). + <mallctl>{counter}</mallctl> is one of the counters in <link + linkend="mutex_counters">mutex profiling + counters</link>.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.mutexes.extents_retained"> + <term> + <mallctl>stats.arenas.<i>.mutexes.extents_retained.{counter}</mallctl> + (<type>counter specific type</type>) <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Statistics on <varname>arena.<i>.extents_retained + </varname> mutex (arena scope; retained extents related). + <mallctl>{counter}</mallctl> is one of the counters in <link + linkend="mutex_counters">mutex profiling + counters</link>.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.mutexes.decay_dirty"> + <term> + <mallctl>stats.arenas.<i>.mutexes.decay_dirty.{counter}</mallctl> + (<type>counter specific type</type>) <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Statistics on <varname>arena.<i>.decay_dirty + </varname> mutex (arena scope; decay for dirty pages related). + <mallctl>{counter}</mallctl> is one of the counters in <link + linkend="mutex_counters">mutex profiling + counters</link>.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.mutexes.decay_muzzy"> + <term> + <mallctl>stats.arenas.<i>.mutexes.decay_muzzy.{counter}</mallctl> + (<type>counter specific type</type>) <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Statistics on <varname>arena.<i>.decay_muzzy + </varname> mutex (arena scope; decay for muzzy pages related). + <mallctl>{counter}</mallctl> is one of the counters in <link + linkend="mutex_counters">mutex profiling + counters</link>.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.mutexes.base"> + <term> + <mallctl>stats.arenas.<i>.mutexes.base.{counter}</mallctl> + (<type>counter specific type</type>) <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Statistics on <varname>arena.<i>.base</varname> + mutex (arena scope; base allocator related). + <mallctl>{counter}</mallctl> is one of the counters in <link + linkend="mutex_counters">mutex profiling + counters</link>.</para></listitem> + </varlistentry> + + <varlistentry id="stats.arenas.i.mutexes.tcache_list"> + <term> + <mallctl>stats.arenas.<i>.mutexes.tcache_list.{counter}</mallctl> + (<type>counter specific type</type>) <literal>r-</literal> + [<option>--enable-stats</option>] + </term> + <listitem><para>Statistics on + <varname>arena.<i>.tcache_list</varname> mutex (arena scope; + tcache to arena association related). This mutex is expected to be + accessed less often. <mallctl>{counter}</mallctl> is one of the + counters in <link linkend="mutex_counters">mutex profiling + counters</link>.</para></listitem> + </varlistentry> + + </variablelist> + </refsect1> + <refsect1 id="heap_profile_format"> + <title>HEAP PROFILE FORMAT</title> + <para>Although the heap profiling functionality was originally designed to + be compatible with the + <command>pprof</command> command that is developed as part of the <ulink + url="http://code.google.com/p/gperftools/">gperftools + package</ulink>, the addition of per thread heap profiling functionality + required a different heap profile format. The <command>jeprof</command> + command is derived from <command>pprof</command>, with enhancements to + support the heap profile format described here.</para> + + <para>In the following hypothetical heap profile, <constant>[...]</constant> + indicates elision for the sake of compactness. <programlisting><![CDATA[ +heap_v2/524288 + t*: 28106: 56637512 [0: 0] + [...] + t3: 352: 16777344 [0: 0] + [...] + t99: 17754: 29341640 [0: 0] + [...] +@ 0x5f86da8 0x5f5a1dc [...] 0x29e4d4e 0xa200316 0xabb2988 [...] + t*: 13: 6688 [0: 0] + t3: 12: 6496 [0: ] + t99: 1: 192 [0: 0] +[...] + +MAPPED_LIBRARIES: +[...]]]></programlisting> The following matches the above heap profile, but most +tokens are replaced with <constant><description></constant> to indicate +descriptions of the corresponding fields. <programlisting><![CDATA[ +<heap_profile_format_version>/<mean_sample_interval> + <aggregate>: <curobjs>: <curbytes> [<cumobjs>: <cumbytes>] + [...] + <thread_3_aggregate>: <curobjs>: <curbytes>[<cumobjs>: <cumbytes>] + [...] + <thread_99_aggregate>: <curobjs>: <curbytes>[<cumobjs>: <cumbytes>] + [...] +@ <top_frame> <frame> [...] <frame> <frame> <frame> [...] + <backtrace_aggregate>: <curobjs>: <curbytes> [<cumobjs>: <cumbytes>] + <backtrace_thread_3>: <curobjs>: <curbytes> [<cumobjs>: <cumbytes>] + <backtrace_thread_99>: <curobjs>: <curbytes> [<cumobjs>: <cumbytes>] +[...] + +MAPPED_LIBRARIES: +</proc/<pid>/maps>]]></programlisting></para> + </refsect1> + + <refsect1 id="debugging_malloc_problems"> + <title>DEBUGGING MALLOC PROBLEMS</title> + <para>When debugging, it is a good idea to configure/build jemalloc with + the <option>--enable-debug</option> and <option>--enable-fill</option> + options, and recompile the program with suitable options and symbols for + debugger support. When so configured, jemalloc incorporates a wide variety + of run-time assertions that catch application errors such as double-free, + write-after-free, etc.</para> + + <para>Programs often accidentally depend on <quote>uninitialized</quote> + memory actually being filled with zero bytes. Junk filling + (see the <link linkend="opt.junk"><mallctl>opt.junk</mallctl></link> + option) tends to expose such bugs in the form of obviously incorrect + results and/or coredumps. Conversely, zero + filling (see the <link + linkend="opt.zero"><mallctl>opt.zero</mallctl></link> option) eliminates + the symptoms of such bugs. Between these two options, it is usually + possible to quickly detect, diagnose, and eliminate such bugs.</para> + + <para>This implementation does not provide much detail about the problems + it detects, because the performance impact for storing such information + would be prohibitive.</para> + </refsect1> + <refsect1 id="diagnostic_messages"> + <title>DIAGNOSTIC MESSAGES</title> + <para>If any of the memory allocation/deallocation functions detect an + error or warning condition, a message will be printed to file descriptor + <constant>STDERR_FILENO</constant>. Errors will result in the process + dumping core. If the <link + linkend="opt.abort"><mallctl>opt.abort</mallctl></link> option is set, most + warnings are treated as errors.</para> + + <para>The <varname>malloc_message</varname> variable allows the programmer + to override the function which emits the text strings forming the errors + and warnings if for some reason the <constant>STDERR_FILENO</constant> file + descriptor is not suitable for this. + <function>malloc_message()</function> takes the + <parameter>cbopaque</parameter> pointer argument that is + <constant>NULL</constant> unless overridden by the arguments in a call to + <function>malloc_stats_print()</function>, followed by a string + pointer. Please note that doing anything which tries to allocate memory in + this function is likely to result in a crash or deadlock.</para> + + <para>All messages are prefixed by + <quote><computeroutput><jemalloc>: </computeroutput></quote>.</para> + </refsect1> + <refsect1 id="return_values"> + <title>RETURN VALUES</title> + <refsect2> + <title>Standard API</title> + <para>The <function>malloc()</function> and + <function>calloc()</function> functions return a pointer to the + allocated memory if successful; otherwise a <constant>NULL</constant> + pointer is returned and <varname>errno</varname> is set to + <errorname>ENOMEM</errorname>.</para> + + <para>The <function>posix_memalign()</function> function + returns the value 0 if successful; otherwise it returns an error value. + The <function>posix_memalign()</function> function will fail + if: + <variablelist> + <varlistentry> + <term><errorname>EINVAL</errorname></term> + + <listitem><para>The <parameter>alignment</parameter> parameter is + not a power of 2 at least as large as + <code language="C">sizeof(<type>void *</type>)</code>. + </para></listitem> + </varlistentry> + <varlistentry> + <term><errorname>ENOMEM</errorname></term> + + <listitem><para>Memory allocation error.</para></listitem> + </varlistentry> + </variablelist> + </para> + + <para>The <function>aligned_alloc()</function> function returns + a pointer to the allocated memory if successful; otherwise a + <constant>NULL</constant> pointer is returned and + <varname>errno</varname> is set. The + <function>aligned_alloc()</function> function will fail if: + <variablelist> + <varlistentry> + <term><errorname>EINVAL</errorname></term> + + <listitem><para>The <parameter>alignment</parameter> parameter is + not a power of 2. + </para></listitem> + </varlistentry> + <varlistentry> + <term><errorname>ENOMEM</errorname></term> + + <listitem><para>Memory allocation error.</para></listitem> + </varlistentry> + </variablelist> + </para> + + <para>The <function>realloc()</function> function returns a + pointer, possibly identical to <parameter>ptr</parameter>, to the + allocated memory if successful; otherwise a <constant>NULL</constant> + pointer is returned, and <varname>errno</varname> is set to + <errorname>ENOMEM</errorname> if the error was the result of an + allocation failure. The <function>realloc()</function> + function always leaves the original buffer intact when an error occurs. + </para> + + <para>The <function>free()</function> function returns no + value.</para> + </refsect2> + <refsect2> + <title>Non-standard API</title> + <para>The <function>mallocx()</function> and + <function>rallocx()</function> functions return a pointer to + the allocated memory if successful; otherwise a <constant>NULL</constant> + pointer is returned to indicate insufficient contiguous memory was + available to service the allocation request. </para> + + <para>The <function>xallocx()</function> function returns the + real size of the resulting resized allocation pointed to by + <parameter>ptr</parameter>, which is a value less than + <parameter>size</parameter> if the allocation could not be adequately + grown in place. </para> + + <para>The <function>sallocx()</function> function returns the + real size of the allocation pointed to by <parameter>ptr</parameter>. + </para> + + <para>The <function>nallocx()</function> returns the real size + that would result from a successful equivalent + <function>mallocx()</function> function call, or zero if + insufficient memory is available to perform the size computation. </para> + + <para>The <function>mallctl()</function>, + <function>mallctlnametomib()</function>, and + <function>mallctlbymib()</function> functions return 0 on + success; otherwise they return an error value. The functions will fail + if: + <variablelist> + <varlistentry> + <term><errorname>EINVAL</errorname></term> + + <listitem><para><parameter>newp</parameter> is not + <constant>NULL</constant>, and <parameter>newlen</parameter> is too + large or too small. Alternatively, <parameter>*oldlenp</parameter> + is too large or too small; in this case as much data as possible + are read despite the error.</para></listitem> + </varlistentry> + <varlistentry> + <term><errorname>ENOENT</errorname></term> + + <listitem><para><parameter>name</parameter> or + <parameter>mib</parameter> specifies an unknown/invalid + value.</para></listitem> + </varlistentry> + <varlistentry> + <term><errorname>EPERM</errorname></term> + + <listitem><para>Attempt to read or write void value, or attempt to + write read-only value.</para></listitem> + </varlistentry> + <varlistentry> + <term><errorname>EAGAIN</errorname></term> + + <listitem><para>A memory allocation failure + occurred.</para></listitem> + </varlistentry> + <varlistentry> + <term><errorname>EFAULT</errorname></term> + + <listitem><para>An interface with side effects failed in some way + not directly related to <function>mallctl*()</function> + read/write processing.</para></listitem> + </varlistentry> + </variablelist> + </para> + + <para>The <function>malloc_usable_size()</function> function + returns the usable size of the allocation pointed to by + <parameter>ptr</parameter>. </para> + </refsect2> + </refsect1> + <refsect1 id="environment"> + <title>ENVIRONMENT</title> + <para>The following environment variable affects the execution of the + allocation functions: + <variablelist> + <varlistentry> + <term><envar>MALLOC_CONF</envar></term> + + <listitem><para>If the environment variable + <envar>MALLOC_CONF</envar> is set, the characters it contains + will be interpreted as options.</para></listitem> + </varlistentry> + </variablelist> + </para> + </refsect1> + <refsect1 id="examples"> + <title>EXAMPLES</title> + <para>To dump core whenever a problem occurs: + <screen>ln -s 'abort:true' /etc/malloc.conf</screen> + </para> + <para>To specify in the source that only one arena should be automatically + created: + <programlisting language="C"><![CDATA[ +malloc_conf = "narenas:1";]]></programlisting></para> + </refsect1> + <refsect1 id="see_also"> + <title>SEE ALSO</title> + <para><citerefentry><refentrytitle>madvise</refentrytitle> + <manvolnum>2</manvolnum></citerefentry>, + <citerefentry><refentrytitle>mmap</refentrytitle> + <manvolnum>2</manvolnum></citerefentry>, + <citerefentry><refentrytitle>sbrk</refentrytitle> + <manvolnum>2</manvolnum></citerefentry>, + <citerefentry><refentrytitle>utrace</refentrytitle> + <manvolnum>2</manvolnum></citerefentry>, + <citerefentry><refentrytitle>alloca</refentrytitle> + <manvolnum>3</manvolnum></citerefentry>, + <citerefentry><refentrytitle>atexit</refentrytitle> + <manvolnum>3</manvolnum></citerefentry>, + <citerefentry><refentrytitle>getpagesize</refentrytitle> + <manvolnum>3</manvolnum></citerefentry></para> + </refsect1> + <refsect1 id="standards"> + <title>STANDARDS</title> + <para>The <function>malloc()</function>, + <function>calloc()</function>, + <function>realloc()</function>, and + <function>free()</function> functions conform to ISO/IEC + 9899:1990 (<quote>ISO C90</quote>).</para> + + <para>The <function>posix_memalign()</function> function conforms + to IEEE Std 1003.1-2001 (<quote>POSIX.1</quote>).</para> + </refsect1> +</refentry> diff --git a/deps/jemalloc/doc/manpages.xsl.in b/deps/jemalloc/doc/manpages.xsl.in new file mode 100644 index 000000000..88b2626b9 --- /dev/null +++ b/deps/jemalloc/doc/manpages.xsl.in @@ -0,0 +1,4 @@ +<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" version="1.0"> + <xsl:import href="@XSLROOT@/manpages/docbook.xsl"/> + <xsl:import href="@abs_srcroot@doc/stylesheet.xsl"/> +</xsl:stylesheet> diff --git a/deps/jemalloc/doc/stylesheet.xsl b/deps/jemalloc/doc/stylesheet.xsl new file mode 100644 index 000000000..619365d82 --- /dev/null +++ b/deps/jemalloc/doc/stylesheet.xsl @@ -0,0 +1,10 @@ +<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" version="1.0"> + <xsl:param name="funcsynopsis.style">ansi</xsl:param> + <xsl:param name="function.parens" select="0"/> + <xsl:template match="function"> + <xsl:call-template name="inline.monoseq"/> + </xsl:template> + <xsl:template match="mallctl"> + <quote><xsl:call-template name="inline.monoseq"/></quote> + </xsl:template> +</xsl:stylesheet> diff --git a/deps/jemalloc/include/jemalloc/internal/arena_externs.h b/deps/jemalloc/include/jemalloc/internal/arena_externs.h new file mode 100644 index 000000000..a4523ae0c --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/arena_externs.h @@ -0,0 +1,104 @@ +#ifndef JEMALLOC_INTERNAL_ARENA_EXTERNS_H +#define JEMALLOC_INTERNAL_ARENA_EXTERNS_H + +#include "jemalloc/internal/bin.h" +#include "jemalloc/internal/extent_dss.h" +#include "jemalloc/internal/hook.h" +#include "jemalloc/internal/pages.h" +#include "jemalloc/internal/stats.h" + +extern ssize_t opt_dirty_decay_ms; +extern ssize_t opt_muzzy_decay_ms; + +extern percpu_arena_mode_t opt_percpu_arena; +extern const char *percpu_arena_mode_names[]; + +extern const uint64_t h_steps[SMOOTHSTEP_NSTEPS]; +extern malloc_mutex_t arenas_lock; + +extern size_t opt_oversize_threshold; +extern size_t oversize_threshold; + +void arena_basic_stats_merge(tsdn_t *tsdn, arena_t *arena, + unsigned *nthreads, const char **dss, ssize_t *dirty_decay_ms, + ssize_t *muzzy_decay_ms, size_t *nactive, size_t *ndirty, size_t *nmuzzy); +void arena_stats_merge(tsdn_t *tsdn, arena_t *arena, unsigned *nthreads, + const char **dss, ssize_t *dirty_decay_ms, ssize_t *muzzy_decay_ms, + size_t *nactive, size_t *ndirty, size_t *nmuzzy, arena_stats_t *astats, + bin_stats_t *bstats, arena_stats_large_t *lstats, + arena_stats_extents_t *estats); +void arena_extents_dirty_dalloc(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent); +#ifdef JEMALLOC_JET +size_t arena_slab_regind(extent_t *slab, szind_t binind, const void *ptr); +#endif +extent_t *arena_extent_alloc_large(tsdn_t *tsdn, arena_t *arena, + size_t usize, size_t alignment, bool *zero); +void arena_extent_dalloc_large_prep(tsdn_t *tsdn, arena_t *arena, + extent_t *extent); +void arena_extent_ralloc_large_shrink(tsdn_t *tsdn, arena_t *arena, + extent_t *extent, size_t oldsize); +void arena_extent_ralloc_large_expand(tsdn_t *tsdn, arena_t *arena, + extent_t *extent, size_t oldsize); +ssize_t arena_dirty_decay_ms_get(arena_t *arena); +bool arena_dirty_decay_ms_set(tsdn_t *tsdn, arena_t *arena, ssize_t decay_ms); +ssize_t arena_muzzy_decay_ms_get(arena_t *arena); +bool arena_muzzy_decay_ms_set(tsdn_t *tsdn, arena_t *arena, ssize_t decay_ms); +void arena_decay(tsdn_t *tsdn, arena_t *arena, bool is_background_thread, + bool all); +void arena_reset(tsd_t *tsd, arena_t *arena); +void arena_destroy(tsd_t *tsd, arena_t *arena); +void arena_tcache_fill_small(tsdn_t *tsdn, arena_t *arena, tcache_t *tcache, + cache_bin_t *tbin, szind_t binind, uint64_t prof_accumbytes); +void arena_alloc_junk_small(void *ptr, const bin_info_t *bin_info, + bool zero); + +typedef void (arena_dalloc_junk_small_t)(void *, const bin_info_t *); +extern arena_dalloc_junk_small_t *JET_MUTABLE arena_dalloc_junk_small; + +void *arena_malloc_hard(tsdn_t *tsdn, arena_t *arena, size_t size, + szind_t ind, bool zero); +void *arena_palloc(tsdn_t *tsdn, arena_t *arena, size_t usize, + size_t alignment, bool zero, tcache_t *tcache); +void arena_prof_promote(tsdn_t *tsdn, void *ptr, size_t usize); +void arena_dalloc_promoted(tsdn_t *tsdn, void *ptr, tcache_t *tcache, + bool slow_path); +void arena_dalloc_bin_junked_locked(tsdn_t *tsdn, arena_t *arena, bin_t *bin, + szind_t binind, extent_t *extent, void *ptr); +void arena_dalloc_small(tsdn_t *tsdn, void *ptr); +bool arena_ralloc_no_move(tsdn_t *tsdn, void *ptr, size_t oldsize, size_t size, + size_t extra, bool zero, size_t *newsize); +void *arena_ralloc(tsdn_t *tsdn, arena_t *arena, void *ptr, size_t oldsize, + size_t size, size_t alignment, bool zero, tcache_t *tcache, + hook_ralloc_args_t *hook_args); +dss_prec_t arena_dss_prec_get(arena_t *arena); +bool arena_dss_prec_set(arena_t *arena, dss_prec_t dss_prec); +ssize_t arena_dirty_decay_ms_default_get(void); +bool arena_dirty_decay_ms_default_set(ssize_t decay_ms); +ssize_t arena_muzzy_decay_ms_default_get(void); +bool arena_muzzy_decay_ms_default_set(ssize_t decay_ms); +bool arena_retain_grow_limit_get_set(tsd_t *tsd, arena_t *arena, + size_t *old_limit, size_t *new_limit); +unsigned arena_nthreads_get(arena_t *arena, bool internal); +void arena_nthreads_inc(arena_t *arena, bool internal); +void arena_nthreads_dec(arena_t *arena, bool internal); +size_t arena_extent_sn_next(arena_t *arena); +arena_t *arena_new(tsdn_t *tsdn, unsigned ind, extent_hooks_t *extent_hooks); +bool arena_init_huge(void); +bool arena_is_huge(unsigned arena_ind); +arena_t *arena_choose_huge(tsd_t *tsd); +bin_t *arena_bin_choose_lock(tsdn_t *tsdn, arena_t *arena, szind_t binind, + unsigned *binshard); +void arena_boot(sc_data_t *sc_data); +void arena_prefork0(tsdn_t *tsdn, arena_t *arena); +void arena_prefork1(tsdn_t *tsdn, arena_t *arena); +void arena_prefork2(tsdn_t *tsdn, arena_t *arena); +void arena_prefork3(tsdn_t *tsdn, arena_t *arena); +void arena_prefork4(tsdn_t *tsdn, arena_t *arena); +void arena_prefork5(tsdn_t *tsdn, arena_t *arena); +void arena_prefork6(tsdn_t *tsdn, arena_t *arena); +void arena_prefork7(tsdn_t *tsdn, arena_t *arena); +void arena_postfork_parent(tsdn_t *tsdn, arena_t *arena); +void arena_postfork_child(tsdn_t *tsdn, arena_t *arena); + +#endif /* JEMALLOC_INTERNAL_ARENA_EXTERNS_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/arena_inlines_a.h b/deps/jemalloc/include/jemalloc/internal/arena_inlines_a.h new file mode 100644 index 000000000..9abf7f6ac --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/arena_inlines_a.h @@ -0,0 +1,57 @@ +#ifndef JEMALLOC_INTERNAL_ARENA_INLINES_A_H +#define JEMALLOC_INTERNAL_ARENA_INLINES_A_H + +static inline unsigned +arena_ind_get(const arena_t *arena) { + return base_ind_get(arena->base); +} + +static inline void +arena_internal_add(arena_t *arena, size_t size) { + atomic_fetch_add_zu(&arena->stats.internal, size, ATOMIC_RELAXED); +} + +static inline void +arena_internal_sub(arena_t *arena, size_t size) { + atomic_fetch_sub_zu(&arena->stats.internal, size, ATOMIC_RELAXED); +} + +static inline size_t +arena_internal_get(arena_t *arena) { + return atomic_load_zu(&arena->stats.internal, ATOMIC_RELAXED); +} + +static inline bool +arena_prof_accum(tsdn_t *tsdn, arena_t *arena, uint64_t accumbytes) { + cassert(config_prof); + + if (likely(prof_interval == 0 || !prof_active_get_unlocked())) { + return false; + } + + return prof_accum_add(tsdn, &arena->prof_accum, accumbytes); +} + +static inline void +percpu_arena_update(tsd_t *tsd, unsigned cpu) { + assert(have_percpu_arena); + arena_t *oldarena = tsd_arena_get(tsd); + assert(oldarena != NULL); + unsigned oldind = arena_ind_get(oldarena); + + if (oldind != cpu) { + unsigned newind = cpu; + arena_t *newarena = arena_get(tsd_tsdn(tsd), newind, true); + assert(newarena != NULL); + + /* Set new arena/tcache associations. */ + arena_migrate(tsd, oldind, newind); + tcache_t *tcache = tcache_get(tsd); + if (tcache != NULL) { + tcache_arena_reassociate(tsd_tsdn(tsd), tcache, + newarena); + } + } +} + +#endif /* JEMALLOC_INTERNAL_ARENA_INLINES_A_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/arena_inlines_b.h b/deps/jemalloc/include/jemalloc/internal/arena_inlines_b.h new file mode 100644 index 000000000..dd926575f --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/arena_inlines_b.h @@ -0,0 +1,427 @@ +#ifndef JEMALLOC_INTERNAL_ARENA_INLINES_B_H +#define JEMALLOC_INTERNAL_ARENA_INLINES_B_H + +#include "jemalloc/internal/jemalloc_internal_types.h" +#include "jemalloc/internal/mutex.h" +#include "jemalloc/internal/rtree.h" +#include "jemalloc/internal/sc.h" +#include "jemalloc/internal/sz.h" +#include "jemalloc/internal/ticker.h" + +JEMALLOC_ALWAYS_INLINE bool +arena_has_default_hooks(arena_t *arena) { + return (extent_hooks_get(arena) == &extent_hooks_default); +} + +JEMALLOC_ALWAYS_INLINE arena_t * +arena_choose_maybe_huge(tsd_t *tsd, arena_t *arena, size_t size) { + if (arena != NULL) { + return arena; + } + + /* + * For huge allocations, use the dedicated huge arena if both are true: + * 1) is using auto arena selection (i.e. arena == NULL), and 2) the + * thread is not assigned to a manual arena. + */ + if (unlikely(size >= oversize_threshold)) { + arena_t *tsd_arena = tsd_arena_get(tsd); + if (tsd_arena == NULL || arena_is_auto(tsd_arena)) { + return arena_choose_huge(tsd); + } + } + + return arena_choose(tsd, NULL); +} + +JEMALLOC_ALWAYS_INLINE prof_tctx_t * +arena_prof_tctx_get(tsdn_t *tsdn, const void *ptr, alloc_ctx_t *alloc_ctx) { + cassert(config_prof); + assert(ptr != NULL); + + /* Static check. */ + if (alloc_ctx == NULL) { + const extent_t *extent = iealloc(tsdn, ptr); + if (unlikely(!extent_slab_get(extent))) { + return large_prof_tctx_get(tsdn, extent); + } + } else { + if (unlikely(!alloc_ctx->slab)) { + return large_prof_tctx_get(tsdn, iealloc(tsdn, ptr)); + } + } + return (prof_tctx_t *)(uintptr_t)1U; +} + +JEMALLOC_ALWAYS_INLINE void +arena_prof_tctx_set(tsdn_t *tsdn, const void *ptr, size_t usize, + alloc_ctx_t *alloc_ctx, prof_tctx_t *tctx) { + cassert(config_prof); + assert(ptr != NULL); + + /* Static check. */ + if (alloc_ctx == NULL) { + extent_t *extent = iealloc(tsdn, ptr); + if (unlikely(!extent_slab_get(extent))) { + large_prof_tctx_set(tsdn, extent, tctx); + } + } else { + if (unlikely(!alloc_ctx->slab)) { + large_prof_tctx_set(tsdn, iealloc(tsdn, ptr), tctx); + } + } +} + +static inline void +arena_prof_tctx_reset(tsdn_t *tsdn, const void *ptr, prof_tctx_t *tctx) { + cassert(config_prof); + assert(ptr != NULL); + + extent_t *extent = iealloc(tsdn, ptr); + assert(!extent_slab_get(extent)); + + large_prof_tctx_reset(tsdn, extent); +} + +JEMALLOC_ALWAYS_INLINE nstime_t +arena_prof_alloc_time_get(tsdn_t *tsdn, const void *ptr, + alloc_ctx_t *alloc_ctx) { + cassert(config_prof); + assert(ptr != NULL); + + extent_t *extent = iealloc(tsdn, ptr); + /* + * Unlike arena_prof_prof_tctx_{get, set}, we only call this once we're + * sure we have a sampled allocation. + */ + assert(!extent_slab_get(extent)); + return large_prof_alloc_time_get(extent); +} + +JEMALLOC_ALWAYS_INLINE void +arena_prof_alloc_time_set(tsdn_t *tsdn, const void *ptr, alloc_ctx_t *alloc_ctx, + nstime_t t) { + cassert(config_prof); + assert(ptr != NULL); + + extent_t *extent = iealloc(tsdn, ptr); + assert(!extent_slab_get(extent)); + large_prof_alloc_time_set(extent, t); +} + +JEMALLOC_ALWAYS_INLINE void +arena_decay_ticks(tsdn_t *tsdn, arena_t *arena, unsigned nticks) { + tsd_t *tsd; + ticker_t *decay_ticker; + + if (unlikely(tsdn_null(tsdn))) { + return; + } + tsd = tsdn_tsd(tsdn); + decay_ticker = decay_ticker_get(tsd, arena_ind_get(arena)); + if (unlikely(decay_ticker == NULL)) { + return; + } + if (unlikely(ticker_ticks(decay_ticker, nticks))) { + arena_decay(tsdn, arena, false, false); + } +} + +JEMALLOC_ALWAYS_INLINE void +arena_decay_tick(tsdn_t *tsdn, arena_t *arena) { + malloc_mutex_assert_not_owner(tsdn, &arena->decay_dirty.mtx); + malloc_mutex_assert_not_owner(tsdn, &arena->decay_muzzy.mtx); + + arena_decay_ticks(tsdn, arena, 1); +} + +/* Purge a single extent to retained / unmapped directly. */ +JEMALLOC_ALWAYS_INLINE void +arena_decay_extent(tsdn_t *tsdn,arena_t *arena, extent_hooks_t **r_extent_hooks, + extent_t *extent) { + size_t extent_size = extent_size_get(extent); + extent_dalloc_wrapper(tsdn, arena, + r_extent_hooks, extent); + if (config_stats) { + /* Update stats accordingly. */ + arena_stats_lock(tsdn, &arena->stats); + arena_stats_add_u64(tsdn, &arena->stats, + &arena->decay_dirty.stats->nmadvise, 1); + arena_stats_add_u64(tsdn, &arena->stats, + &arena->decay_dirty.stats->purged, extent_size >> LG_PAGE); + arena_stats_sub_zu(tsdn, &arena->stats, &arena->stats.mapped, + extent_size); + arena_stats_unlock(tsdn, &arena->stats); + } +} + +JEMALLOC_ALWAYS_INLINE void * +arena_malloc(tsdn_t *tsdn, arena_t *arena, size_t size, szind_t ind, bool zero, + tcache_t *tcache, bool slow_path) { + assert(!tsdn_null(tsdn) || tcache == NULL); + + if (likely(tcache != NULL)) { + if (likely(size <= SC_SMALL_MAXCLASS)) { + return tcache_alloc_small(tsdn_tsd(tsdn), arena, + tcache, size, ind, zero, slow_path); + } + if (likely(size <= tcache_maxclass)) { + return tcache_alloc_large(tsdn_tsd(tsdn), arena, + tcache, size, ind, zero, slow_path); + } + /* (size > tcache_maxclass) case falls through. */ + assert(size > tcache_maxclass); + } + + return arena_malloc_hard(tsdn, arena, size, ind, zero); +} + +JEMALLOC_ALWAYS_INLINE arena_t * +arena_aalloc(tsdn_t *tsdn, const void *ptr) { + return extent_arena_get(iealloc(tsdn, ptr)); +} + +JEMALLOC_ALWAYS_INLINE size_t +arena_salloc(tsdn_t *tsdn, const void *ptr) { + assert(ptr != NULL); + + rtree_ctx_t rtree_ctx_fallback; + rtree_ctx_t *rtree_ctx = tsdn_rtree_ctx(tsdn, &rtree_ctx_fallback); + + szind_t szind = rtree_szind_read(tsdn, &extents_rtree, rtree_ctx, + (uintptr_t)ptr, true); + assert(szind != SC_NSIZES); + + return sz_index2size(szind); +} + +JEMALLOC_ALWAYS_INLINE size_t +arena_vsalloc(tsdn_t *tsdn, const void *ptr) { + /* + * Return 0 if ptr is not within an extent managed by jemalloc. This + * function has two extra costs relative to isalloc(): + * - The rtree calls cannot claim to be dependent lookups, which induces + * rtree lookup load dependencies. + * - The lookup may fail, so there is an extra branch to check for + * failure. + */ + + rtree_ctx_t rtree_ctx_fallback; + rtree_ctx_t *rtree_ctx = tsdn_rtree_ctx(tsdn, &rtree_ctx_fallback); + + extent_t *extent; + szind_t szind; + if (rtree_extent_szind_read(tsdn, &extents_rtree, rtree_ctx, + (uintptr_t)ptr, false, &extent, &szind)) { + return 0; + } + + if (extent == NULL) { + return 0; + } + assert(extent_state_get(extent) == extent_state_active); + /* Only slab members should be looked up via interior pointers. */ + assert(extent_addr_get(extent) == ptr || extent_slab_get(extent)); + + assert(szind != SC_NSIZES); + + return sz_index2size(szind); +} + +static inline void +arena_dalloc_large_no_tcache(tsdn_t *tsdn, void *ptr, szind_t szind) { + if (config_prof && unlikely(szind < SC_NBINS)) { + arena_dalloc_promoted(tsdn, ptr, NULL, true); + } else { + extent_t *extent = iealloc(tsdn, ptr); + large_dalloc(tsdn, extent); + } +} + +static inline void +arena_dalloc_no_tcache(tsdn_t *tsdn, void *ptr) { + assert(ptr != NULL); + + rtree_ctx_t rtree_ctx_fallback; + rtree_ctx_t *rtree_ctx = tsdn_rtree_ctx(tsdn, &rtree_ctx_fallback); + + szind_t szind; + bool slab; + rtree_szind_slab_read(tsdn, &extents_rtree, rtree_ctx, (uintptr_t)ptr, + true, &szind, &slab); + + if (config_debug) { + extent_t *extent = rtree_extent_read(tsdn, &extents_rtree, + rtree_ctx, (uintptr_t)ptr, true); + assert(szind == extent_szind_get(extent)); + assert(szind < SC_NSIZES); + assert(slab == extent_slab_get(extent)); + } + + if (likely(slab)) { + /* Small allocation. */ + arena_dalloc_small(tsdn, ptr); + } else { + arena_dalloc_large_no_tcache(tsdn, ptr, szind); + } +} + +JEMALLOC_ALWAYS_INLINE void +arena_dalloc_large(tsdn_t *tsdn, void *ptr, tcache_t *tcache, szind_t szind, + bool slow_path) { + if (szind < nhbins) { + if (config_prof && unlikely(szind < SC_NBINS)) { + arena_dalloc_promoted(tsdn, ptr, tcache, slow_path); + } else { + tcache_dalloc_large(tsdn_tsd(tsdn), tcache, ptr, szind, + slow_path); + } + } else { + extent_t *extent = iealloc(tsdn, ptr); + large_dalloc(tsdn, extent); + } +} + +JEMALLOC_ALWAYS_INLINE void +arena_dalloc(tsdn_t *tsdn, void *ptr, tcache_t *tcache, + alloc_ctx_t *alloc_ctx, bool slow_path) { + assert(!tsdn_null(tsdn) || tcache == NULL); + assert(ptr != NULL); + + if (unlikely(tcache == NULL)) { + arena_dalloc_no_tcache(tsdn, ptr); + return; + } + + szind_t szind; + bool slab; + rtree_ctx_t *rtree_ctx; + if (alloc_ctx != NULL) { + szind = alloc_ctx->szind; + slab = alloc_ctx->slab; + assert(szind != SC_NSIZES); + } else { + rtree_ctx = tsd_rtree_ctx(tsdn_tsd(tsdn)); + rtree_szind_slab_read(tsdn, &extents_rtree, rtree_ctx, + (uintptr_t)ptr, true, &szind, &slab); + } + + if (config_debug) { + rtree_ctx = tsd_rtree_ctx(tsdn_tsd(tsdn)); + extent_t *extent = rtree_extent_read(tsdn, &extents_rtree, + rtree_ctx, (uintptr_t)ptr, true); + assert(szind == extent_szind_get(extent)); + assert(szind < SC_NSIZES); + assert(slab == extent_slab_get(extent)); + } + + if (likely(slab)) { + /* Small allocation. */ + tcache_dalloc_small(tsdn_tsd(tsdn), tcache, ptr, szind, + slow_path); + } else { + arena_dalloc_large(tsdn, ptr, tcache, szind, slow_path); + } +} + +static inline void +arena_sdalloc_no_tcache(tsdn_t *tsdn, void *ptr, size_t size) { + assert(ptr != NULL); + assert(size <= SC_LARGE_MAXCLASS); + + szind_t szind; + bool slab; + if (!config_prof || !opt_prof) { + /* + * There is no risk of being confused by a promoted sampled + * object, so base szind and slab on the given size. + */ + szind = sz_size2index(size); + slab = (szind < SC_NBINS); + } + + if ((config_prof && opt_prof) || config_debug) { + rtree_ctx_t rtree_ctx_fallback; + rtree_ctx_t *rtree_ctx = tsdn_rtree_ctx(tsdn, + &rtree_ctx_fallback); + + rtree_szind_slab_read(tsdn, &extents_rtree, rtree_ctx, + (uintptr_t)ptr, true, &szind, &slab); + + assert(szind == sz_size2index(size)); + assert((config_prof && opt_prof) || slab == (szind < SC_NBINS)); + + if (config_debug) { + extent_t *extent = rtree_extent_read(tsdn, + &extents_rtree, rtree_ctx, (uintptr_t)ptr, true); + assert(szind == extent_szind_get(extent)); + assert(slab == extent_slab_get(extent)); + } + } + + if (likely(slab)) { + /* Small allocation. */ + arena_dalloc_small(tsdn, ptr); + } else { + arena_dalloc_large_no_tcache(tsdn, ptr, szind); + } +} + +JEMALLOC_ALWAYS_INLINE void +arena_sdalloc(tsdn_t *tsdn, void *ptr, size_t size, tcache_t *tcache, + alloc_ctx_t *alloc_ctx, bool slow_path) { + assert(!tsdn_null(tsdn) || tcache == NULL); + assert(ptr != NULL); + assert(size <= SC_LARGE_MAXCLASS); + + if (unlikely(tcache == NULL)) { + arena_sdalloc_no_tcache(tsdn, ptr, size); + return; + } + + szind_t szind; + bool slab; + alloc_ctx_t local_ctx; + if (config_prof && opt_prof) { + if (alloc_ctx == NULL) { + /* Uncommon case and should be a static check. */ + rtree_ctx_t rtree_ctx_fallback; + rtree_ctx_t *rtree_ctx = tsdn_rtree_ctx(tsdn, + &rtree_ctx_fallback); + rtree_szind_slab_read(tsdn, &extents_rtree, rtree_ctx, + (uintptr_t)ptr, true, &local_ctx.szind, + &local_ctx.slab); + assert(local_ctx.szind == sz_size2index(size)); + alloc_ctx = &local_ctx; + } + slab = alloc_ctx->slab; + szind = alloc_ctx->szind; + } else { + /* + * There is no risk of being confused by a promoted sampled + * object, so base szind and slab on the given size. + */ + szind = sz_size2index(size); + slab = (szind < SC_NBINS); + } + + if (config_debug) { + rtree_ctx_t *rtree_ctx = tsd_rtree_ctx(tsdn_tsd(tsdn)); + rtree_szind_slab_read(tsdn, &extents_rtree, rtree_ctx, + (uintptr_t)ptr, true, &szind, &slab); + extent_t *extent = rtree_extent_read(tsdn, + &extents_rtree, rtree_ctx, (uintptr_t)ptr, true); + assert(szind == extent_szind_get(extent)); + assert(slab == extent_slab_get(extent)); + } + + if (likely(slab)) { + /* Small allocation. */ + tcache_dalloc_small(tsdn_tsd(tsdn), tcache, ptr, szind, + slow_path); + } else { + arena_dalloc_large(tsdn, ptr, tcache, szind, slow_path); + } +} + +#endif /* JEMALLOC_INTERNAL_ARENA_INLINES_B_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/arena_stats.h b/deps/jemalloc/include/jemalloc/internal/arena_stats.h new file mode 100644 index 000000000..23949ed92 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/arena_stats.h @@ -0,0 +1,271 @@ +#ifndef JEMALLOC_INTERNAL_ARENA_STATS_H +#define JEMALLOC_INTERNAL_ARENA_STATS_H + +#include "jemalloc/internal/atomic.h" +#include "jemalloc/internal/mutex.h" +#include "jemalloc/internal/mutex_prof.h" +#include "jemalloc/internal/sc.h" + +JEMALLOC_DIAGNOSTIC_DISABLE_SPURIOUS + +/* + * In those architectures that support 64-bit atomics, we use atomic updates for + * our 64-bit values. Otherwise, we use a plain uint64_t and synchronize + * externally. + */ +#ifdef JEMALLOC_ATOMIC_U64 +typedef atomic_u64_t arena_stats_u64_t; +#else +/* Must hold the arena stats mutex while reading atomically. */ +typedef uint64_t arena_stats_u64_t; +#endif + +typedef struct arena_stats_large_s arena_stats_large_t; +struct arena_stats_large_s { + /* + * Total number of allocation/deallocation requests served directly by + * the arena. + */ + arena_stats_u64_t nmalloc; + arena_stats_u64_t ndalloc; + + /* + * Number of allocation requests that correspond to this size class. + * This includes requests served by tcache, though tcache only + * periodically merges into this counter. + */ + arena_stats_u64_t nrequests; /* Partially derived. */ + /* + * Number of tcache fills / flushes for large (similarly, periodically + * merged). Note that there is no large tcache batch-fill currently + * (i.e. only fill 1 at a time); however flush may be batched. + */ + arena_stats_u64_t nfills; /* Partially derived. */ + arena_stats_u64_t nflushes; /* Partially derived. */ + + /* Current number of allocations of this size class. */ + size_t curlextents; /* Derived. */ +}; + +typedef struct arena_stats_decay_s arena_stats_decay_t; +struct arena_stats_decay_s { + /* Total number of purge sweeps. */ + arena_stats_u64_t npurge; + /* Total number of madvise calls made. */ + arena_stats_u64_t nmadvise; + /* Total number of pages purged. */ + arena_stats_u64_t purged; +}; + +typedef struct arena_stats_extents_s arena_stats_extents_t; +struct arena_stats_extents_s { + /* + * Stats for a given index in the range [0, SC_NPSIZES] in an extents_t. + * We track both bytes and # of extents: two extents in the same bucket + * may have different sizes if adjacent size classes differ by more than + * a page, so bytes cannot always be derived from # of extents. + */ + atomic_zu_t ndirty; + atomic_zu_t dirty_bytes; + atomic_zu_t nmuzzy; + atomic_zu_t muzzy_bytes; + atomic_zu_t nretained; + atomic_zu_t retained_bytes; +}; + +/* + * Arena stats. Note that fields marked "derived" are not directly maintained + * within the arena code; rather their values are derived during stats merge + * requests. + */ +typedef struct arena_stats_s arena_stats_t; +struct arena_stats_s { +#ifndef JEMALLOC_ATOMIC_U64 + malloc_mutex_t mtx; +#endif + + /* Number of bytes currently mapped, excluding retained memory. */ + atomic_zu_t mapped; /* Partially derived. */ + + /* + * Number of unused virtual memory bytes currently retained. Retained + * bytes are technically mapped (though always decommitted or purged), + * but they are excluded from the mapped statistic (above). + */ + atomic_zu_t retained; /* Derived. */ + + /* Number of extent_t structs allocated by base, but not being used. */ + atomic_zu_t extent_avail; + + arena_stats_decay_t decay_dirty; + arena_stats_decay_t decay_muzzy; + + atomic_zu_t base; /* Derived. */ + atomic_zu_t internal; + atomic_zu_t resident; /* Derived. */ + atomic_zu_t metadata_thp; + + atomic_zu_t allocated_large; /* Derived. */ + arena_stats_u64_t nmalloc_large; /* Derived. */ + arena_stats_u64_t ndalloc_large; /* Derived. */ + arena_stats_u64_t nfills_large; /* Derived. */ + arena_stats_u64_t nflushes_large; /* Derived. */ + arena_stats_u64_t nrequests_large; /* Derived. */ + + /* VM space had to be leaked (undocumented). Normally 0. */ + atomic_zu_t abandoned_vm; + + /* Number of bytes cached in tcache associated with this arena. */ + atomic_zu_t tcache_bytes; /* Derived. */ + + mutex_prof_data_t mutex_prof_data[mutex_prof_num_arena_mutexes]; + + /* One element for each large size class. */ + arena_stats_large_t lstats[SC_NSIZES - SC_NBINS]; + + /* Arena uptime. */ + nstime_t uptime; +}; + +static inline bool +arena_stats_init(tsdn_t *tsdn, arena_stats_t *arena_stats) { + if (config_debug) { + for (size_t i = 0; i < sizeof(arena_stats_t); i++) { + assert(((char *)arena_stats)[i] == 0); + } + } +#ifndef JEMALLOC_ATOMIC_U64 + if (malloc_mutex_init(&arena_stats->mtx, "arena_stats", + WITNESS_RANK_ARENA_STATS, malloc_mutex_rank_exclusive)) { + return true; + } +#endif + /* Memory is zeroed, so there is no need to clear stats. */ + return false; +} + +static inline void +arena_stats_lock(tsdn_t *tsdn, arena_stats_t *arena_stats) { +#ifndef JEMALLOC_ATOMIC_U64 + malloc_mutex_lock(tsdn, &arena_stats->mtx); +#endif +} + +static inline void +arena_stats_unlock(tsdn_t *tsdn, arena_stats_t *arena_stats) { +#ifndef JEMALLOC_ATOMIC_U64 + malloc_mutex_unlock(tsdn, &arena_stats->mtx); +#endif +} + +static inline uint64_t +arena_stats_read_u64(tsdn_t *tsdn, arena_stats_t *arena_stats, + arena_stats_u64_t *p) { +#ifdef JEMALLOC_ATOMIC_U64 + return atomic_load_u64(p, ATOMIC_RELAXED); +#else + malloc_mutex_assert_owner(tsdn, &arena_stats->mtx); + return *p; +#endif +} + +static inline void +arena_stats_add_u64(tsdn_t *tsdn, arena_stats_t *arena_stats, + arena_stats_u64_t *p, uint64_t x) { +#ifdef JEMALLOC_ATOMIC_U64 + atomic_fetch_add_u64(p, x, ATOMIC_RELAXED); +#else + malloc_mutex_assert_owner(tsdn, &arena_stats->mtx); + *p += x; +#endif +} + +static inline void +arena_stats_sub_u64(tsdn_t *tsdn, arena_stats_t *arena_stats, + arena_stats_u64_t *p, uint64_t x) { +#ifdef JEMALLOC_ATOMIC_U64 + uint64_t r = atomic_fetch_sub_u64(p, x, ATOMIC_RELAXED); + assert(r - x <= r); +#else + malloc_mutex_assert_owner(tsdn, &arena_stats->mtx); + *p -= x; + assert(*p + x >= *p); +#endif +} + +/* + * Non-atomically sets *dst += src. *dst needs external synchronization. + * This lets us avoid the cost of a fetch_add when its unnecessary (note that + * the types here are atomic). + */ +static inline void +arena_stats_accum_u64(arena_stats_u64_t *dst, uint64_t src) { +#ifdef JEMALLOC_ATOMIC_U64 + uint64_t cur_dst = atomic_load_u64(dst, ATOMIC_RELAXED); + atomic_store_u64(dst, src + cur_dst, ATOMIC_RELAXED); +#else + *dst += src; +#endif +} + +static inline size_t +arena_stats_read_zu(tsdn_t *tsdn, arena_stats_t *arena_stats, + atomic_zu_t *p) { +#ifdef JEMALLOC_ATOMIC_U64 + return atomic_load_zu(p, ATOMIC_RELAXED); +#else + malloc_mutex_assert_owner(tsdn, &arena_stats->mtx); + return atomic_load_zu(p, ATOMIC_RELAXED); +#endif +} + +static inline void +arena_stats_add_zu(tsdn_t *tsdn, arena_stats_t *arena_stats, + atomic_zu_t *p, size_t x) { +#ifdef JEMALLOC_ATOMIC_U64 + atomic_fetch_add_zu(p, x, ATOMIC_RELAXED); +#else + malloc_mutex_assert_owner(tsdn, &arena_stats->mtx); + size_t cur = atomic_load_zu(p, ATOMIC_RELAXED); + atomic_store_zu(p, cur + x, ATOMIC_RELAXED); +#endif +} + +static inline void +arena_stats_sub_zu(tsdn_t *tsdn, arena_stats_t *arena_stats, + atomic_zu_t *p, size_t x) { +#ifdef JEMALLOC_ATOMIC_U64 + size_t r = atomic_fetch_sub_zu(p, x, ATOMIC_RELAXED); + assert(r - x <= r); +#else + malloc_mutex_assert_owner(tsdn, &arena_stats->mtx); + size_t cur = atomic_load_zu(p, ATOMIC_RELAXED); + atomic_store_zu(p, cur - x, ATOMIC_RELAXED); +#endif +} + +/* Like the _u64 variant, needs an externally synchronized *dst. */ +static inline void +arena_stats_accum_zu(atomic_zu_t *dst, size_t src) { + size_t cur_dst = atomic_load_zu(dst, ATOMIC_RELAXED); + atomic_store_zu(dst, src + cur_dst, ATOMIC_RELAXED); +} + +static inline void +arena_stats_large_flush_nrequests_add(tsdn_t *tsdn, arena_stats_t *arena_stats, + szind_t szind, uint64_t nrequests) { + arena_stats_lock(tsdn, arena_stats); + arena_stats_large_t *lstats = &arena_stats->lstats[szind - SC_NBINS]; + arena_stats_add_u64(tsdn, arena_stats, &lstats->nrequests, nrequests); + arena_stats_add_u64(tsdn, arena_stats, &lstats->nflushes, 1); + arena_stats_unlock(tsdn, arena_stats); +} + +static inline void +arena_stats_mapped_add(tsdn_t *tsdn, arena_stats_t *arena_stats, size_t size) { + arena_stats_lock(tsdn, arena_stats); + arena_stats_add_zu(tsdn, arena_stats, &arena_stats->mapped, size); + arena_stats_unlock(tsdn, arena_stats); +} + +#endif /* JEMALLOC_INTERNAL_ARENA_STATS_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/arena_structs_a.h b/deps/jemalloc/include/jemalloc/internal/arena_structs_a.h new file mode 100644 index 000000000..46aa77c88 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/arena_structs_a.h @@ -0,0 +1,11 @@ +#ifndef JEMALLOC_INTERNAL_ARENA_STRUCTS_A_H +#define JEMALLOC_INTERNAL_ARENA_STRUCTS_A_H + +#include "jemalloc/internal/bitmap.h" + +struct arena_slab_data_s { + /* Per region allocated/deallocated bitmap. */ + bitmap_t bitmap[BITMAP_GROUPS_MAX]; +}; + +#endif /* JEMALLOC_INTERNAL_ARENA_STRUCTS_A_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/arena_structs_b.h b/deps/jemalloc/include/jemalloc/internal/arena_structs_b.h new file mode 100644 index 000000000..eeab57fd6 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/arena_structs_b.h @@ -0,0 +1,232 @@ +#ifndef JEMALLOC_INTERNAL_ARENA_STRUCTS_B_H +#define JEMALLOC_INTERNAL_ARENA_STRUCTS_B_H + +#include "jemalloc/internal/arena_stats.h" +#include "jemalloc/internal/atomic.h" +#include "jemalloc/internal/bin.h" +#include "jemalloc/internal/bitmap.h" +#include "jemalloc/internal/extent_dss.h" +#include "jemalloc/internal/jemalloc_internal_types.h" +#include "jemalloc/internal/mutex.h" +#include "jemalloc/internal/nstime.h" +#include "jemalloc/internal/ql.h" +#include "jemalloc/internal/sc.h" +#include "jemalloc/internal/smoothstep.h" +#include "jemalloc/internal/ticker.h" + +struct arena_decay_s { + /* Synchronizes all non-atomic fields. */ + malloc_mutex_t mtx; + /* + * True if a thread is currently purging the extents associated with + * this decay structure. + */ + bool purging; + /* + * Approximate time in milliseconds from the creation of a set of unused + * dirty pages until an equivalent set of unused dirty pages is purged + * and/or reused. + */ + atomic_zd_t time_ms; + /* time / SMOOTHSTEP_NSTEPS. */ + nstime_t interval; + /* + * Time at which the current decay interval logically started. We do + * not actually advance to a new epoch until sometime after it starts + * because of scheduling and computation delays, and it is even possible + * to completely skip epochs. In all cases, during epoch advancement we + * merge all relevant activity into the most recently recorded epoch. + */ + nstime_t epoch; + /* Deadline randomness generator. */ + uint64_t jitter_state; + /* + * Deadline for current epoch. This is the sum of interval and per + * epoch jitter which is a uniform random variable in [0..interval). + * Epochs always advance by precise multiples of interval, but we + * randomize the deadline to reduce the likelihood of arenas purging in + * lockstep. + */ + nstime_t deadline; + /* + * Number of unpurged pages at beginning of current epoch. During epoch + * advancement we use the delta between arena->decay_*.nunpurged and + * extents_npages_get(&arena->extents_*) to determine how many dirty + * pages, if any, were generated. + */ + size_t nunpurged; + /* + * Trailing log of how many unused dirty pages were generated during + * each of the past SMOOTHSTEP_NSTEPS decay epochs, where the last + * element is the most recent epoch. Corresponding epoch times are + * relative to epoch. + */ + size_t backlog[SMOOTHSTEP_NSTEPS]; + + /* + * Pointer to associated stats. These stats are embedded directly in + * the arena's stats due to how stats structures are shared between the + * arena and ctl code. + * + * Synchronization: Same as associated arena's stats field. */ + arena_stats_decay_t *stats; + /* Peak number of pages in associated extents. Used for debug only. */ + uint64_t ceil_npages; +}; + +struct arena_s { + /* + * Number of threads currently assigned to this arena. Each thread has + * two distinct assignments, one for application-serving allocation, and + * the other for internal metadata allocation. Internal metadata must + * not be allocated from arenas explicitly created via the arenas.create + * mallctl, because the arena.<i>.reset mallctl indiscriminately + * discards all allocations for the affected arena. + * + * 0: Application allocation. + * 1: Internal metadata allocation. + * + * Synchronization: atomic. + */ + atomic_u_t nthreads[2]; + + /* Next bin shard for binding new threads. Synchronization: atomic. */ + atomic_u_t binshard_next; + + /* + * When percpu_arena is enabled, to amortize the cost of reading / + * updating the current CPU id, track the most recent thread accessing + * this arena, and only read CPU if there is a mismatch. + */ + tsdn_t *last_thd; + + /* Synchronization: internal. */ + arena_stats_t stats; + + /* + * Lists of tcaches and cache_bin_array_descriptors for extant threads + * associated with this arena. Stats from these are merged + * incrementally, and at exit if opt_stats_print is enabled. + * + * Synchronization: tcache_ql_mtx. + */ + ql_head(tcache_t) tcache_ql; + ql_head(cache_bin_array_descriptor_t) cache_bin_array_descriptor_ql; + malloc_mutex_t tcache_ql_mtx; + + /* Synchronization: internal. */ + prof_accum_t prof_accum; + + /* + * PRNG state for cache index randomization of large allocation base + * pointers. + * + * Synchronization: atomic. + */ + atomic_zu_t offset_state; + + /* + * Extent serial number generator state. + * + * Synchronization: atomic. + */ + atomic_zu_t extent_sn_next; + + /* + * Represents a dss_prec_t, but atomically. + * + * Synchronization: atomic. + */ + atomic_u_t dss_prec; + + /* + * Number of pages in active extents. + * + * Synchronization: atomic. + */ + atomic_zu_t nactive; + + /* + * Extant large allocations. + * + * Synchronization: large_mtx. + */ + extent_list_t large; + /* Synchronizes all large allocation/update/deallocation. */ + malloc_mutex_t large_mtx; + + /* + * Collections of extents that were previously allocated. These are + * used when allocating extents, in an attempt to re-use address space. + * + * Synchronization: internal. + */ + extents_t extents_dirty; + extents_t extents_muzzy; + extents_t extents_retained; + + /* + * Decay-based purging state, responsible for scheduling extent state + * transitions. + * + * Synchronization: internal. + */ + arena_decay_t decay_dirty; /* dirty --> muzzy */ + arena_decay_t decay_muzzy; /* muzzy --> retained */ + + /* + * Next extent size class in a growing series to use when satisfying a + * request via the extent hooks (only if opt_retain). This limits the + * number of disjoint virtual memory ranges so that extent merging can + * be effective even if multiple arenas' extent allocation requests are + * highly interleaved. + * + * retain_grow_limit is the max allowed size ind to expand (unless the + * required size is greater). Default is no limit, and controlled + * through mallctl only. + * + * Synchronization: extent_grow_mtx + */ + pszind_t extent_grow_next; + pszind_t retain_grow_limit; + malloc_mutex_t extent_grow_mtx; + + /* + * Available extent structures that were allocated via + * base_alloc_extent(). + * + * Synchronization: extent_avail_mtx. + */ + extent_tree_t extent_avail; + atomic_zu_t extent_avail_cnt; + malloc_mutex_t extent_avail_mtx; + + /* + * bins is used to store heaps of free regions. + * + * Synchronization: internal. + */ + bins_t bins[SC_NBINS]; + + /* + * Base allocator, from which arena metadata are allocated. + * + * Synchronization: internal. + */ + base_t *base; + /* Used to determine uptime. Read-only after initialization. */ + nstime_t create_time; +}; + +/* Used in conjunction with tsd for fast arena-related context lookup. */ +struct arena_tdata_s { + ticker_t decay_ticker; +}; + +/* Used to pass rtree lookup context down the path. */ +struct alloc_ctx_s { + szind_t szind; + bool slab; +}; + +#endif /* JEMALLOC_INTERNAL_ARENA_STRUCTS_B_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/arena_types.h b/deps/jemalloc/include/jemalloc/internal/arena_types.h new file mode 100644 index 000000000..624937e4f --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/arena_types.h @@ -0,0 +1,51 @@ +#ifndef JEMALLOC_INTERNAL_ARENA_TYPES_H +#define JEMALLOC_INTERNAL_ARENA_TYPES_H + +#include "jemalloc/internal/sc.h" + +/* Maximum number of regions in one slab. */ +#define LG_SLAB_MAXREGS (LG_PAGE - SC_LG_TINY_MIN) +#define SLAB_MAXREGS (1U << LG_SLAB_MAXREGS) + +/* Default decay times in milliseconds. */ +#define DIRTY_DECAY_MS_DEFAULT ZD(10 * 1000) +#define MUZZY_DECAY_MS_DEFAULT (0) +/* Number of event ticks between time checks. */ +#define DECAY_NTICKS_PER_UPDATE 1000 + +typedef struct arena_slab_data_s arena_slab_data_t; +typedef struct arena_decay_s arena_decay_t; +typedef struct arena_s arena_t; +typedef struct arena_tdata_s arena_tdata_t; +typedef struct alloc_ctx_s alloc_ctx_t; + +typedef enum { + percpu_arena_mode_names_base = 0, /* Used for options processing. */ + + /* + * *_uninit are used only during bootstrapping, and must correspond + * to initialized variant plus percpu_arena_mode_enabled_base. + */ + percpu_arena_uninit = 0, + per_phycpu_arena_uninit = 1, + + /* All non-disabled modes must come after percpu_arena_disabled. */ + percpu_arena_disabled = 2, + + percpu_arena_mode_names_limit = 3, /* Used for options processing. */ + percpu_arena_mode_enabled_base = 3, + + percpu_arena = 3, + per_phycpu_arena = 4 /* Hyper threads share arena. */ +} percpu_arena_mode_t; + +#define PERCPU_ARENA_ENABLED(m) ((m) >= percpu_arena_mode_enabled_base) +#define PERCPU_ARENA_DEFAULT percpu_arena_disabled + +/* + * When allocation_size >= oversize_threshold, use the dedicated huge arena + * (unless have explicitly spicified arena index). 0 disables the feature. + */ +#define OVERSIZE_THRESHOLD_DEFAULT (8 << 20) + +#endif /* JEMALLOC_INTERNAL_ARENA_TYPES_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/assert.h b/deps/jemalloc/include/jemalloc/internal/assert.h new file mode 100644 index 000000000..be4d45b32 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/assert.h @@ -0,0 +1,56 @@ +#include "jemalloc/internal/malloc_io.h" +#include "jemalloc/internal/util.h" + +/* + * Define a custom assert() in order to reduce the chances of deadlock during + * assertion failure. + */ +#ifndef assert +#define assert(e) do { \ + if (unlikely(config_debug && !(e))) { \ + malloc_printf( \ + "<jemalloc>: %s:%d: Failed assertion: \"%s\"\n", \ + __FILE__, __LINE__, #e); \ + abort(); \ + } \ +} while (0) +#endif + +#ifndef not_reached +#define not_reached() do { \ + if (config_debug) { \ + malloc_printf( \ + "<jemalloc>: %s:%d: Unreachable code reached\n", \ + __FILE__, __LINE__); \ + abort(); \ + } \ + unreachable(); \ +} while (0) +#endif + +#ifndef not_implemented +#define not_implemented() do { \ + if (config_debug) { \ + malloc_printf("<jemalloc>: %s:%d: Not implemented\n", \ + __FILE__, __LINE__); \ + abort(); \ + } \ +} while (0) +#endif + +#ifndef assert_not_implemented +#define assert_not_implemented(e) do { \ + if (unlikely(config_debug && !(e))) { \ + not_implemented(); \ + } \ +} while (0) +#endif + +/* Use to assert a particular configuration, e.g., cassert(config_debug). */ +#ifndef cassert +#define cassert(c) do { \ + if (unlikely(!(c))) { \ + not_reached(); \ + } \ +} while (0) +#endif diff --git a/deps/jemalloc/include/jemalloc/internal/atomic.h b/deps/jemalloc/include/jemalloc/internal/atomic.h new file mode 100644 index 000000000..a76f54cee --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/atomic.h @@ -0,0 +1,86 @@ +#ifndef JEMALLOC_INTERNAL_ATOMIC_H +#define JEMALLOC_INTERNAL_ATOMIC_H + +#define ATOMIC_INLINE JEMALLOC_ALWAYS_INLINE + +#define JEMALLOC_U8_ATOMICS +#if defined(JEMALLOC_GCC_ATOMIC_ATOMICS) +# include "jemalloc/internal/atomic_gcc_atomic.h" +# if !defined(JEMALLOC_GCC_U8_ATOMIC_ATOMICS) +# undef JEMALLOC_U8_ATOMICS +# endif +#elif defined(JEMALLOC_GCC_SYNC_ATOMICS) +# include "jemalloc/internal/atomic_gcc_sync.h" +# if !defined(JEMALLOC_GCC_U8_SYNC_ATOMICS) +# undef JEMALLOC_U8_ATOMICS +# endif +#elif defined(_MSC_VER) +# include "jemalloc/internal/atomic_msvc.h" +#elif defined(JEMALLOC_C11_ATOMICS) +# include "jemalloc/internal/atomic_c11.h" +#else +# error "Don't have atomics implemented on this platform." +#endif + +/* + * This header gives more or less a backport of C11 atomics. The user can write + * JEMALLOC_GENERATE_ATOMICS(type, short_type, lg_sizeof_type); to generate + * counterparts of the C11 atomic functions for type, as so: + * JEMALLOC_GENERATE_ATOMICS(int *, pi, 3); + * and then write things like: + * int *some_ptr; + * atomic_pi_t atomic_ptr_to_int; + * atomic_store_pi(&atomic_ptr_to_int, some_ptr, ATOMIC_RELAXED); + * int *prev_value = atomic_exchange_pi(&ptr_to_int, NULL, ATOMIC_ACQ_REL); + * assert(some_ptr == prev_value); + * and expect things to work in the obvious way. + * + * Also included (with naming differences to avoid conflicts with the standard + * library): + * atomic_fence(atomic_memory_order_t) (mimics C11's atomic_thread_fence). + * ATOMIC_INIT (mimics C11's ATOMIC_VAR_INIT). + */ + +/* + * Pure convenience, so that we don't have to type "atomic_memory_order_" + * quite so often. + */ +#define ATOMIC_RELAXED atomic_memory_order_relaxed +#define ATOMIC_ACQUIRE atomic_memory_order_acquire +#define ATOMIC_RELEASE atomic_memory_order_release +#define ATOMIC_ACQ_REL atomic_memory_order_acq_rel +#define ATOMIC_SEQ_CST atomic_memory_order_seq_cst + +/* + * Not all platforms have 64-bit atomics. If we do, this #define exposes that + * fact. + */ +#if (LG_SIZEOF_PTR == 3 || LG_SIZEOF_INT == 3) +# define JEMALLOC_ATOMIC_U64 +#endif + +JEMALLOC_GENERATE_ATOMICS(void *, p, LG_SIZEOF_PTR) + +/* + * There's no actual guarantee that sizeof(bool) == 1, but it's true on the only + * platform that actually needs to know the size, MSVC. + */ +JEMALLOC_GENERATE_ATOMICS(bool, b, 0) + +JEMALLOC_GENERATE_INT_ATOMICS(unsigned, u, LG_SIZEOF_INT) + +JEMALLOC_GENERATE_INT_ATOMICS(size_t, zu, LG_SIZEOF_PTR) + +JEMALLOC_GENERATE_INT_ATOMICS(ssize_t, zd, LG_SIZEOF_PTR) + +JEMALLOC_GENERATE_INT_ATOMICS(uint8_t, u8, 0) + +JEMALLOC_GENERATE_INT_ATOMICS(uint32_t, u32, 2) + +#ifdef JEMALLOC_ATOMIC_U64 +JEMALLOC_GENERATE_INT_ATOMICS(uint64_t, u64, 3) +#endif + +#undef ATOMIC_INLINE + +#endif /* JEMALLOC_INTERNAL_ATOMIC_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/atomic_c11.h b/deps/jemalloc/include/jemalloc/internal/atomic_c11.h new file mode 100644 index 000000000..a5f9313a6 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/atomic_c11.h @@ -0,0 +1,97 @@ +#ifndef JEMALLOC_INTERNAL_ATOMIC_C11_H +#define JEMALLOC_INTERNAL_ATOMIC_C11_H + +#include <stdatomic.h> + +#define ATOMIC_INIT(...) ATOMIC_VAR_INIT(__VA_ARGS__) + +#define atomic_memory_order_t memory_order +#define atomic_memory_order_relaxed memory_order_relaxed +#define atomic_memory_order_acquire memory_order_acquire +#define atomic_memory_order_release memory_order_release +#define atomic_memory_order_acq_rel memory_order_acq_rel +#define atomic_memory_order_seq_cst memory_order_seq_cst + +#define atomic_fence atomic_thread_fence + +#define JEMALLOC_GENERATE_ATOMICS(type, short_type, \ + /* unused */ lg_size) \ +typedef _Atomic(type) atomic_##short_type##_t; \ + \ +ATOMIC_INLINE type \ +atomic_load_##short_type(const atomic_##short_type##_t *a, \ + atomic_memory_order_t mo) { \ + /* \ + * A strict interpretation of the C standard prevents \ + * atomic_load from taking a const argument, but it's \ + * convenient for our purposes. This cast is a workaround. \ + */ \ + atomic_##short_type##_t* a_nonconst = \ + (atomic_##short_type##_t*)a; \ + return atomic_load_explicit(a_nonconst, mo); \ +} \ + \ +ATOMIC_INLINE void \ +atomic_store_##short_type(atomic_##short_type##_t *a, \ + type val, atomic_memory_order_t mo) { \ + atomic_store_explicit(a, val, mo); \ +} \ + \ +ATOMIC_INLINE type \ +atomic_exchange_##short_type(atomic_##short_type##_t *a, type val, \ + atomic_memory_order_t mo) { \ + return atomic_exchange_explicit(a, val, mo); \ +} \ + \ +ATOMIC_INLINE bool \ +atomic_compare_exchange_weak_##short_type(atomic_##short_type##_t *a, \ + type *expected, type desired, atomic_memory_order_t success_mo, \ + atomic_memory_order_t failure_mo) { \ + return atomic_compare_exchange_weak_explicit(a, expected, \ + desired, success_mo, failure_mo); \ +} \ + \ +ATOMIC_INLINE bool \ +atomic_compare_exchange_strong_##short_type(atomic_##short_type##_t *a, \ + type *expected, type desired, atomic_memory_order_t success_mo, \ + atomic_memory_order_t failure_mo) { \ + return atomic_compare_exchange_strong_explicit(a, expected, \ + desired, success_mo, failure_mo); \ +} + +/* + * Integral types have some special operations available that non-integral ones + * lack. + */ +#define JEMALLOC_GENERATE_INT_ATOMICS(type, short_type, \ + /* unused */ lg_size) \ +JEMALLOC_GENERATE_ATOMICS(type, short_type, /* unused */ lg_size) \ + \ +ATOMIC_INLINE type \ +atomic_fetch_add_##short_type(atomic_##short_type##_t *a, \ + type val, atomic_memory_order_t mo) { \ + return atomic_fetch_add_explicit(a, val, mo); \ +} \ + \ +ATOMIC_INLINE type \ +atomic_fetch_sub_##short_type(atomic_##short_type##_t *a, \ + type val, atomic_memory_order_t mo) { \ + return atomic_fetch_sub_explicit(a, val, mo); \ +} \ +ATOMIC_INLINE type \ +atomic_fetch_and_##short_type(atomic_##short_type##_t *a, \ + type val, atomic_memory_order_t mo) { \ + return atomic_fetch_and_explicit(a, val, mo); \ +} \ +ATOMIC_INLINE type \ +atomic_fetch_or_##short_type(atomic_##short_type##_t *a, \ + type val, atomic_memory_order_t mo) { \ + return atomic_fetch_or_explicit(a, val, mo); \ +} \ +ATOMIC_INLINE type \ +atomic_fetch_xor_##short_type(atomic_##short_type##_t *a, \ + type val, atomic_memory_order_t mo) { \ + return atomic_fetch_xor_explicit(a, val, mo); \ +} + +#endif /* JEMALLOC_INTERNAL_ATOMIC_C11_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/atomic_gcc_atomic.h b/deps/jemalloc/include/jemalloc/internal/atomic_gcc_atomic.h new file mode 100644 index 000000000..471515e82 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/atomic_gcc_atomic.h @@ -0,0 +1,129 @@ +#ifndef JEMALLOC_INTERNAL_ATOMIC_GCC_ATOMIC_H +#define JEMALLOC_INTERNAL_ATOMIC_GCC_ATOMIC_H + +#include "jemalloc/internal/assert.h" + +#define ATOMIC_INIT(...) {__VA_ARGS__} + +typedef enum { + atomic_memory_order_relaxed, + atomic_memory_order_acquire, + atomic_memory_order_release, + atomic_memory_order_acq_rel, + atomic_memory_order_seq_cst +} atomic_memory_order_t; + +ATOMIC_INLINE int +atomic_enum_to_builtin(atomic_memory_order_t mo) { + switch (mo) { + case atomic_memory_order_relaxed: + return __ATOMIC_RELAXED; + case atomic_memory_order_acquire: + return __ATOMIC_ACQUIRE; + case atomic_memory_order_release: + return __ATOMIC_RELEASE; + case atomic_memory_order_acq_rel: + return __ATOMIC_ACQ_REL; + case atomic_memory_order_seq_cst: + return __ATOMIC_SEQ_CST; + } + /* Can't happen; the switch is exhaustive. */ + not_reached(); +} + +ATOMIC_INLINE void +atomic_fence(atomic_memory_order_t mo) { + __atomic_thread_fence(atomic_enum_to_builtin(mo)); +} + +#define JEMALLOC_GENERATE_ATOMICS(type, short_type, \ + /* unused */ lg_size) \ +typedef struct { \ + type repr; \ +} atomic_##short_type##_t; \ + \ +ATOMIC_INLINE type \ +atomic_load_##short_type(const atomic_##short_type##_t *a, \ + atomic_memory_order_t mo) { \ + type result; \ + __atomic_load(&a->repr, &result, atomic_enum_to_builtin(mo)); \ + return result; \ +} \ + \ +ATOMIC_INLINE void \ +atomic_store_##short_type(atomic_##short_type##_t *a, type val, \ + atomic_memory_order_t mo) { \ + __atomic_store(&a->repr, &val, atomic_enum_to_builtin(mo)); \ +} \ + \ +ATOMIC_INLINE type \ +atomic_exchange_##short_type(atomic_##short_type##_t *a, type val, \ + atomic_memory_order_t mo) { \ + type result; \ + __atomic_exchange(&a->repr, &val, &result, \ + atomic_enum_to_builtin(mo)); \ + return result; \ +} \ + \ +ATOMIC_INLINE bool \ +atomic_compare_exchange_weak_##short_type(atomic_##short_type##_t *a, \ + UNUSED type *expected, type desired, \ + atomic_memory_order_t success_mo, \ + atomic_memory_order_t failure_mo) { \ + return __atomic_compare_exchange(&a->repr, expected, &desired, \ + true, atomic_enum_to_builtin(success_mo), \ + atomic_enum_to_builtin(failure_mo)); \ +} \ + \ +ATOMIC_INLINE bool \ +atomic_compare_exchange_strong_##short_type(atomic_##short_type##_t *a, \ + UNUSED type *expected, type desired, \ + atomic_memory_order_t success_mo, \ + atomic_memory_order_t failure_mo) { \ + return __atomic_compare_exchange(&a->repr, expected, &desired, \ + false, \ + atomic_enum_to_builtin(success_mo), \ + atomic_enum_to_builtin(failure_mo)); \ +} + + +#define JEMALLOC_GENERATE_INT_ATOMICS(type, short_type, \ + /* unused */ lg_size) \ +JEMALLOC_GENERATE_ATOMICS(type, short_type, /* unused */ lg_size) \ + \ +ATOMIC_INLINE type \ +atomic_fetch_add_##short_type(atomic_##short_type##_t *a, type val, \ + atomic_memory_order_t mo) { \ + return __atomic_fetch_add(&a->repr, val, \ + atomic_enum_to_builtin(mo)); \ +} \ + \ +ATOMIC_INLINE type \ +atomic_fetch_sub_##short_type(atomic_##short_type##_t *a, type val, \ + atomic_memory_order_t mo) { \ + return __atomic_fetch_sub(&a->repr, val, \ + atomic_enum_to_builtin(mo)); \ +} \ + \ +ATOMIC_INLINE type \ +atomic_fetch_and_##short_type(atomic_##short_type##_t *a, type val, \ + atomic_memory_order_t mo) { \ + return __atomic_fetch_and(&a->repr, val, \ + atomic_enum_to_builtin(mo)); \ +} \ + \ +ATOMIC_INLINE type \ +atomic_fetch_or_##short_type(atomic_##short_type##_t *a, type val, \ + atomic_memory_order_t mo) { \ + return __atomic_fetch_or(&a->repr, val, \ + atomic_enum_to_builtin(mo)); \ +} \ + \ +ATOMIC_INLINE type \ +atomic_fetch_xor_##short_type(atomic_##short_type##_t *a, type val, \ + atomic_memory_order_t mo) { \ + return __atomic_fetch_xor(&a->repr, val, \ + atomic_enum_to_builtin(mo)); \ +} + +#endif /* JEMALLOC_INTERNAL_ATOMIC_GCC_ATOMIC_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/atomic_gcc_sync.h b/deps/jemalloc/include/jemalloc/internal/atomic_gcc_sync.h new file mode 100644 index 000000000..e02b7cbe3 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/atomic_gcc_sync.h @@ -0,0 +1,195 @@ +#ifndef JEMALLOC_INTERNAL_ATOMIC_GCC_SYNC_H +#define JEMALLOC_INTERNAL_ATOMIC_GCC_SYNC_H + +#define ATOMIC_INIT(...) {__VA_ARGS__} + +typedef enum { + atomic_memory_order_relaxed, + atomic_memory_order_acquire, + atomic_memory_order_release, + atomic_memory_order_acq_rel, + atomic_memory_order_seq_cst +} atomic_memory_order_t; + +ATOMIC_INLINE void +atomic_fence(atomic_memory_order_t mo) { + /* Easy cases first: no barrier, and full barrier. */ + if (mo == atomic_memory_order_relaxed) { + asm volatile("" ::: "memory"); + return; + } + if (mo == atomic_memory_order_seq_cst) { + asm volatile("" ::: "memory"); + __sync_synchronize(); + asm volatile("" ::: "memory"); + return; + } + asm volatile("" ::: "memory"); +# if defined(__i386__) || defined(__x86_64__) + /* This is implicit on x86. */ +# elif defined(__ppc64__) + asm volatile("lwsync"); +# elif defined(__ppc__) + asm volatile("sync"); +# elif defined(__sparc__) && defined(__arch64__) + if (mo == atomic_memory_order_acquire) { + asm volatile("membar #LoadLoad | #LoadStore"); + } else if (mo == atomic_memory_order_release) { + asm volatile("membar #LoadStore | #StoreStore"); + } else { + asm volatile("membar #LoadLoad | #LoadStore | #StoreStore"); + } +# else + __sync_synchronize(); +# endif + asm volatile("" ::: "memory"); +} + +/* + * A correct implementation of seq_cst loads and stores on weakly ordered + * architectures could do either of the following: + * 1. store() is weak-fence -> store -> strong fence, load() is load -> + * strong-fence. + * 2. store() is strong-fence -> store, load() is strong-fence -> load -> + * weak-fence. + * The tricky thing is, load() and store() above can be the load or store + * portions of a gcc __sync builtin, so we have to follow GCC's lead, which + * means going with strategy 2. + * On strongly ordered architectures, the natural strategy is to stick a strong + * fence after seq_cst stores, and have naked loads. So we want the strong + * fences in different places on different architectures. + * atomic_pre_sc_load_fence and atomic_post_sc_store_fence allow us to + * accomplish this. + */ + +ATOMIC_INLINE void +atomic_pre_sc_load_fence() { +# if defined(__i386__) || defined(__x86_64__) || \ + (defined(__sparc__) && defined(__arch64__)) + atomic_fence(atomic_memory_order_relaxed); +# else + atomic_fence(atomic_memory_order_seq_cst); +# endif +} + +ATOMIC_INLINE void +atomic_post_sc_store_fence() { +# if defined(__i386__) || defined(__x86_64__) || \ + (defined(__sparc__) && defined(__arch64__)) + atomic_fence(atomic_memory_order_seq_cst); +# else + atomic_fence(atomic_memory_order_relaxed); +# endif + +} + +#define JEMALLOC_GENERATE_ATOMICS(type, short_type, \ + /* unused */ lg_size) \ +typedef struct { \ + type volatile repr; \ +} atomic_##short_type##_t; \ + \ +ATOMIC_INLINE type \ +atomic_load_##short_type(const atomic_##short_type##_t *a, \ + atomic_memory_order_t mo) { \ + if (mo == atomic_memory_order_seq_cst) { \ + atomic_pre_sc_load_fence(); \ + } \ + type result = a->repr; \ + if (mo != atomic_memory_order_relaxed) { \ + atomic_fence(atomic_memory_order_acquire); \ + } \ + return result; \ +} \ + \ +ATOMIC_INLINE void \ +atomic_store_##short_type(atomic_##short_type##_t *a, \ + type val, atomic_memory_order_t mo) { \ + if (mo != atomic_memory_order_relaxed) { \ + atomic_fence(atomic_memory_order_release); \ + } \ + a->repr = val; \ + if (mo == atomic_memory_order_seq_cst) { \ + atomic_post_sc_store_fence(); \ + } \ +} \ + \ +ATOMIC_INLINE type \ +atomic_exchange_##short_type(atomic_##short_type##_t *a, type val, \ + atomic_memory_order_t mo) { \ + /* \ + * Because of FreeBSD, we care about gcc 4.2, which doesn't have\ + * an atomic exchange builtin. We fake it with a CAS loop. \ + */ \ + while (true) { \ + type old = a->repr; \ + if (__sync_bool_compare_and_swap(&a->repr, old, val)) { \ + return old; \ + } \ + } \ +} \ + \ +ATOMIC_INLINE bool \ +atomic_compare_exchange_weak_##short_type(atomic_##short_type##_t *a, \ + type *expected, type desired, \ + atomic_memory_order_t success_mo, \ + atomic_memory_order_t failure_mo) { \ + type prev = __sync_val_compare_and_swap(&a->repr, *expected, \ + desired); \ + if (prev == *expected) { \ + return true; \ + } else { \ + *expected = prev; \ + return false; \ + } \ +} \ +ATOMIC_INLINE bool \ +atomic_compare_exchange_strong_##short_type(atomic_##short_type##_t *a, \ + type *expected, type desired, \ + atomic_memory_order_t success_mo, \ + atomic_memory_order_t failure_mo) { \ + type prev = __sync_val_compare_and_swap(&a->repr, *expected, \ + desired); \ + if (prev == *expected) { \ + return true; \ + } else { \ + *expected = prev; \ + return false; \ + } \ +} + +#define JEMALLOC_GENERATE_INT_ATOMICS(type, short_type, \ + /* unused */ lg_size) \ +JEMALLOC_GENERATE_ATOMICS(type, short_type, /* unused */ lg_size) \ + \ +ATOMIC_INLINE type \ +atomic_fetch_add_##short_type(atomic_##short_type##_t *a, type val, \ + atomic_memory_order_t mo) { \ + return __sync_fetch_and_add(&a->repr, val); \ +} \ + \ +ATOMIC_INLINE type \ +atomic_fetch_sub_##short_type(atomic_##short_type##_t *a, type val, \ + atomic_memory_order_t mo) { \ + return __sync_fetch_and_sub(&a->repr, val); \ +} \ + \ +ATOMIC_INLINE type \ +atomic_fetch_and_##short_type(atomic_##short_type##_t *a, type val, \ + atomic_memory_order_t mo) { \ + return __sync_fetch_and_and(&a->repr, val); \ +} \ + \ +ATOMIC_INLINE type \ +atomic_fetch_or_##short_type(atomic_##short_type##_t *a, type val, \ + atomic_memory_order_t mo) { \ + return __sync_fetch_and_or(&a->repr, val); \ +} \ + \ +ATOMIC_INLINE type \ +atomic_fetch_xor_##short_type(atomic_##short_type##_t *a, type val, \ + atomic_memory_order_t mo) { \ + return __sync_fetch_and_xor(&a->repr, val); \ +} + +#endif /* JEMALLOC_INTERNAL_ATOMIC_GCC_SYNC_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/atomic_msvc.h b/deps/jemalloc/include/jemalloc/internal/atomic_msvc.h new file mode 100644 index 000000000..67057ce50 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/atomic_msvc.h @@ -0,0 +1,158 @@ +#ifndef JEMALLOC_INTERNAL_ATOMIC_MSVC_H +#define JEMALLOC_INTERNAL_ATOMIC_MSVC_H + +#define ATOMIC_INIT(...) {__VA_ARGS__} + +typedef enum { + atomic_memory_order_relaxed, + atomic_memory_order_acquire, + atomic_memory_order_release, + atomic_memory_order_acq_rel, + atomic_memory_order_seq_cst +} atomic_memory_order_t; + +typedef char atomic_repr_0_t; +typedef short atomic_repr_1_t; +typedef long atomic_repr_2_t; +typedef __int64 atomic_repr_3_t; + +ATOMIC_INLINE void +atomic_fence(atomic_memory_order_t mo) { + _ReadWriteBarrier(); +# if defined(_M_ARM) || defined(_M_ARM64) + /* ARM needs a barrier for everything but relaxed. */ + if (mo != atomic_memory_order_relaxed) { + MemoryBarrier(); + } +# elif defined(_M_IX86) || defined (_M_X64) + /* x86 needs a barrier only for seq_cst. */ + if (mo == atomic_memory_order_seq_cst) { + MemoryBarrier(); + } +# else +# error "Don't know how to create atomics for this platform for MSVC." +# endif + _ReadWriteBarrier(); +} + +#define ATOMIC_INTERLOCKED_REPR(lg_size) atomic_repr_ ## lg_size ## _t + +#define ATOMIC_CONCAT(a, b) ATOMIC_RAW_CONCAT(a, b) +#define ATOMIC_RAW_CONCAT(a, b) a ## b + +#define ATOMIC_INTERLOCKED_NAME(base_name, lg_size) ATOMIC_CONCAT( \ + base_name, ATOMIC_INTERLOCKED_SUFFIX(lg_size)) + +#define ATOMIC_INTERLOCKED_SUFFIX(lg_size) \ + ATOMIC_CONCAT(ATOMIC_INTERLOCKED_SUFFIX_, lg_size) + +#define ATOMIC_INTERLOCKED_SUFFIX_0 8 +#define ATOMIC_INTERLOCKED_SUFFIX_1 16 +#define ATOMIC_INTERLOCKED_SUFFIX_2 +#define ATOMIC_INTERLOCKED_SUFFIX_3 64 + +#define JEMALLOC_GENERATE_ATOMICS(type, short_type, lg_size) \ +typedef struct { \ + ATOMIC_INTERLOCKED_REPR(lg_size) repr; \ +} atomic_##short_type##_t; \ + \ +ATOMIC_INLINE type \ +atomic_load_##short_type(const atomic_##short_type##_t *a, \ + atomic_memory_order_t mo) { \ + ATOMIC_INTERLOCKED_REPR(lg_size) ret = a->repr; \ + if (mo != atomic_memory_order_relaxed) { \ + atomic_fence(atomic_memory_order_acquire); \ + } \ + return (type) ret; \ +} \ + \ +ATOMIC_INLINE void \ +atomic_store_##short_type(atomic_##short_type##_t *a, \ + type val, atomic_memory_order_t mo) { \ + if (mo != atomic_memory_order_relaxed) { \ + atomic_fence(atomic_memory_order_release); \ + } \ + a->repr = (ATOMIC_INTERLOCKED_REPR(lg_size)) val; \ + if (mo == atomic_memory_order_seq_cst) { \ + atomic_fence(atomic_memory_order_seq_cst); \ + } \ +} \ + \ +ATOMIC_INLINE type \ +atomic_exchange_##short_type(atomic_##short_type##_t *a, type val, \ + atomic_memory_order_t mo) { \ + return (type)ATOMIC_INTERLOCKED_NAME(_InterlockedExchange, \ + lg_size)(&a->repr, (ATOMIC_INTERLOCKED_REPR(lg_size))val); \ +} \ + \ +ATOMIC_INLINE bool \ +atomic_compare_exchange_weak_##short_type(atomic_##short_type##_t *a, \ + type *expected, type desired, atomic_memory_order_t success_mo, \ + atomic_memory_order_t failure_mo) { \ + ATOMIC_INTERLOCKED_REPR(lg_size) e = \ + (ATOMIC_INTERLOCKED_REPR(lg_size))*expected; \ + ATOMIC_INTERLOCKED_REPR(lg_size) d = \ + (ATOMIC_INTERLOCKED_REPR(lg_size))desired; \ + ATOMIC_INTERLOCKED_REPR(lg_size) old = \ + ATOMIC_INTERLOCKED_NAME(_InterlockedCompareExchange, \ + lg_size)(&a->repr, d, e); \ + if (old == e) { \ + return true; \ + } else { \ + *expected = (type)old; \ + return false; \ + } \ +} \ + \ +ATOMIC_INLINE bool \ +atomic_compare_exchange_strong_##short_type(atomic_##short_type##_t *a, \ + type *expected, type desired, atomic_memory_order_t success_mo, \ + atomic_memory_order_t failure_mo) { \ + /* We implement the weak version with strong semantics. */ \ + return atomic_compare_exchange_weak_##short_type(a, expected, \ + desired, success_mo, failure_mo); \ +} + + +#define JEMALLOC_GENERATE_INT_ATOMICS(type, short_type, lg_size) \ +JEMALLOC_GENERATE_ATOMICS(type, short_type, lg_size) \ + \ +ATOMIC_INLINE type \ +atomic_fetch_add_##short_type(atomic_##short_type##_t *a, \ + type val, atomic_memory_order_t mo) { \ + return (type)ATOMIC_INTERLOCKED_NAME(_InterlockedExchangeAdd, \ + lg_size)(&a->repr, (ATOMIC_INTERLOCKED_REPR(lg_size))val); \ +} \ + \ +ATOMIC_INLINE type \ +atomic_fetch_sub_##short_type(atomic_##short_type##_t *a, \ + type val, atomic_memory_order_t mo) { \ + /* \ + * MSVC warns on negation of unsigned operands, but for us it \ + * gives exactly the right semantics (MAX_TYPE + 1 - operand). \ + */ \ + __pragma(warning(push)) \ + __pragma(warning(disable: 4146)) \ + return atomic_fetch_add_##short_type(a, -val, mo); \ + __pragma(warning(pop)) \ +} \ +ATOMIC_INLINE type \ +atomic_fetch_and_##short_type(atomic_##short_type##_t *a, \ + type val, atomic_memory_order_t mo) { \ + return (type)ATOMIC_INTERLOCKED_NAME(_InterlockedAnd, lg_size)( \ + &a->repr, (ATOMIC_INTERLOCKED_REPR(lg_size))val); \ +} \ +ATOMIC_INLINE type \ +atomic_fetch_or_##short_type(atomic_##short_type##_t *a, \ + type val, atomic_memory_order_t mo) { \ + return (type)ATOMIC_INTERLOCKED_NAME(_InterlockedOr, lg_size)( \ + &a->repr, (ATOMIC_INTERLOCKED_REPR(lg_size))val); \ +} \ +ATOMIC_INLINE type \ +atomic_fetch_xor_##short_type(atomic_##short_type##_t *a, \ + type val, atomic_memory_order_t mo) { \ + return (type)ATOMIC_INTERLOCKED_NAME(_InterlockedXor, lg_size)( \ + &a->repr, (ATOMIC_INTERLOCKED_REPR(lg_size))val); \ +} + +#endif /* JEMALLOC_INTERNAL_ATOMIC_MSVC_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/background_thread_externs.h b/deps/jemalloc/include/jemalloc/internal/background_thread_externs.h new file mode 100644 index 000000000..0f997e18b --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/background_thread_externs.h @@ -0,0 +1,32 @@ +#ifndef JEMALLOC_INTERNAL_BACKGROUND_THREAD_EXTERNS_H +#define JEMALLOC_INTERNAL_BACKGROUND_THREAD_EXTERNS_H + +extern bool opt_background_thread; +extern size_t opt_max_background_threads; +extern malloc_mutex_t background_thread_lock; +extern atomic_b_t background_thread_enabled_state; +extern size_t n_background_threads; +extern size_t max_background_threads; +extern background_thread_info_t *background_thread_info; + +bool background_thread_create(tsd_t *tsd, unsigned arena_ind); +bool background_threads_enable(tsd_t *tsd); +bool background_threads_disable(tsd_t *tsd); +void background_thread_interval_check(tsdn_t *tsdn, arena_t *arena, + arena_decay_t *decay, size_t npages_new); +void background_thread_prefork0(tsdn_t *tsdn); +void background_thread_prefork1(tsdn_t *tsdn); +void background_thread_postfork_parent(tsdn_t *tsdn); +void background_thread_postfork_child(tsdn_t *tsdn); +bool background_thread_stats_read(tsdn_t *tsdn, + background_thread_stats_t *stats); +void background_thread_ctl_init(tsdn_t *tsdn); + +#ifdef JEMALLOC_PTHREAD_CREATE_WRAPPER +extern int pthread_create_wrapper(pthread_t *__restrict, const pthread_attr_t *, + void *(*)(void *), void *__restrict); +#endif +bool background_thread_boot0(void); +bool background_thread_boot1(tsdn_t *tsdn); + +#endif /* JEMALLOC_INTERNAL_BACKGROUND_THREAD_EXTERNS_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/background_thread_inlines.h b/deps/jemalloc/include/jemalloc/internal/background_thread_inlines.h new file mode 100644 index 000000000..f85e86fa3 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/background_thread_inlines.h @@ -0,0 +1,62 @@ +#ifndef JEMALLOC_INTERNAL_BACKGROUND_THREAD_INLINES_H +#define JEMALLOC_INTERNAL_BACKGROUND_THREAD_INLINES_H + +JEMALLOC_ALWAYS_INLINE bool +background_thread_enabled(void) { + return atomic_load_b(&background_thread_enabled_state, ATOMIC_RELAXED); +} + +JEMALLOC_ALWAYS_INLINE void +background_thread_enabled_set(tsdn_t *tsdn, bool state) { + malloc_mutex_assert_owner(tsdn, &background_thread_lock); + atomic_store_b(&background_thread_enabled_state, state, ATOMIC_RELAXED); +} + +JEMALLOC_ALWAYS_INLINE background_thread_info_t * +arena_background_thread_info_get(arena_t *arena) { + unsigned arena_ind = arena_ind_get(arena); + return &background_thread_info[arena_ind % max_background_threads]; +} + +JEMALLOC_ALWAYS_INLINE background_thread_info_t * +background_thread_info_get(size_t ind) { + return &background_thread_info[ind % max_background_threads]; +} + +JEMALLOC_ALWAYS_INLINE uint64_t +background_thread_wakeup_time_get(background_thread_info_t *info) { + uint64_t next_wakeup = nstime_ns(&info->next_wakeup); + assert(atomic_load_b(&info->indefinite_sleep, ATOMIC_ACQUIRE) == + (next_wakeup == BACKGROUND_THREAD_INDEFINITE_SLEEP)); + return next_wakeup; +} + +JEMALLOC_ALWAYS_INLINE void +background_thread_wakeup_time_set(tsdn_t *tsdn, background_thread_info_t *info, + uint64_t wakeup_time) { + malloc_mutex_assert_owner(tsdn, &info->mtx); + atomic_store_b(&info->indefinite_sleep, + wakeup_time == BACKGROUND_THREAD_INDEFINITE_SLEEP, ATOMIC_RELEASE); + nstime_init(&info->next_wakeup, wakeup_time); +} + +JEMALLOC_ALWAYS_INLINE bool +background_thread_indefinite_sleep(background_thread_info_t *info) { + return atomic_load_b(&info->indefinite_sleep, ATOMIC_ACQUIRE); +} + +JEMALLOC_ALWAYS_INLINE void +arena_background_thread_inactivity_check(tsdn_t *tsdn, arena_t *arena, + bool is_background_thread) { + if (!background_thread_enabled() || is_background_thread) { + return; + } + background_thread_info_t *info = + arena_background_thread_info_get(arena); + if (background_thread_indefinite_sleep(info)) { + background_thread_interval_check(tsdn, arena, + &arena->decay_dirty, 0); + } +} + +#endif /* JEMALLOC_INTERNAL_BACKGROUND_THREAD_INLINES_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/background_thread_structs.h b/deps/jemalloc/include/jemalloc/internal/background_thread_structs.h new file mode 100644 index 000000000..c02aa434c --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/background_thread_structs.h @@ -0,0 +1,54 @@ +#ifndef JEMALLOC_INTERNAL_BACKGROUND_THREAD_STRUCTS_H +#define JEMALLOC_INTERNAL_BACKGROUND_THREAD_STRUCTS_H + +/* This file really combines "structs" and "types", but only transitionally. */ + +#if defined(JEMALLOC_BACKGROUND_THREAD) || defined(JEMALLOC_LAZY_LOCK) +# define JEMALLOC_PTHREAD_CREATE_WRAPPER +#endif + +#define BACKGROUND_THREAD_INDEFINITE_SLEEP UINT64_MAX +#define MAX_BACKGROUND_THREAD_LIMIT MALLOCX_ARENA_LIMIT +#define DEFAULT_NUM_BACKGROUND_THREAD 4 + +typedef enum { + background_thread_stopped, + background_thread_started, + /* Thread waits on the global lock when paused (for arena_reset). */ + background_thread_paused, +} background_thread_state_t; + +struct background_thread_info_s { +#ifdef JEMALLOC_BACKGROUND_THREAD + /* Background thread is pthread specific. */ + pthread_t thread; + pthread_cond_t cond; +#endif + malloc_mutex_t mtx; + background_thread_state_t state; + /* When true, it means no wakeup scheduled. */ + atomic_b_t indefinite_sleep; + /* Next scheduled wakeup time (absolute time in ns). */ + nstime_t next_wakeup; + /* + * Since the last background thread run, newly added number of pages + * that need to be purged by the next wakeup. This is adjusted on + * epoch advance, and is used to determine whether we should signal the + * background thread to wake up earlier. + */ + size_t npages_to_purge_new; + /* Stats: total number of runs since started. */ + uint64_t tot_n_runs; + /* Stats: total sleep time since started. */ + nstime_t tot_sleep_time; +}; +typedef struct background_thread_info_s background_thread_info_t; + +struct background_thread_stats_s { + size_t num_threads; + uint64_t num_runs; + nstime_t run_interval; +}; +typedef struct background_thread_stats_s background_thread_stats_t; + +#endif /* JEMALLOC_INTERNAL_BACKGROUND_THREAD_STRUCTS_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/base_externs.h b/deps/jemalloc/include/jemalloc/internal/base_externs.h new file mode 100644 index 000000000..7b705c9b4 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/base_externs.h @@ -0,0 +1,22 @@ +#ifndef JEMALLOC_INTERNAL_BASE_EXTERNS_H +#define JEMALLOC_INTERNAL_BASE_EXTERNS_H + +extern metadata_thp_mode_t opt_metadata_thp; +extern const char *metadata_thp_mode_names[]; + +base_t *b0get(void); +base_t *base_new(tsdn_t *tsdn, unsigned ind, extent_hooks_t *extent_hooks); +void base_delete(tsdn_t *tsdn, base_t *base); +extent_hooks_t *base_extent_hooks_get(base_t *base); +extent_hooks_t *base_extent_hooks_set(base_t *base, + extent_hooks_t *extent_hooks); +void *base_alloc(tsdn_t *tsdn, base_t *base, size_t size, size_t alignment); +extent_t *base_alloc_extent(tsdn_t *tsdn, base_t *base); +void base_stats_get(tsdn_t *tsdn, base_t *base, size_t *allocated, + size_t *resident, size_t *mapped, size_t *n_thp); +void base_prefork(tsdn_t *tsdn, base_t *base); +void base_postfork_parent(tsdn_t *tsdn, base_t *base); +void base_postfork_child(tsdn_t *tsdn, base_t *base); +bool base_boot(tsdn_t *tsdn); + +#endif /* JEMALLOC_INTERNAL_BASE_EXTERNS_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/base_inlines.h b/deps/jemalloc/include/jemalloc/internal/base_inlines.h new file mode 100644 index 000000000..aec0e2e1e --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/base_inlines.h @@ -0,0 +1,13 @@ +#ifndef JEMALLOC_INTERNAL_BASE_INLINES_H +#define JEMALLOC_INTERNAL_BASE_INLINES_H + +static inline unsigned +base_ind_get(const base_t *base) { + return base->ind; +} + +static inline bool +metadata_thp_enabled(void) { + return (opt_metadata_thp != metadata_thp_disabled); +} +#endif /* JEMALLOC_INTERNAL_BASE_INLINES_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/base_structs.h b/deps/jemalloc/include/jemalloc/internal/base_structs.h new file mode 100644 index 000000000..07f214eb2 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/base_structs.h @@ -0,0 +1,59 @@ +#ifndef JEMALLOC_INTERNAL_BASE_STRUCTS_H +#define JEMALLOC_INTERNAL_BASE_STRUCTS_H + +#include "jemalloc/internal/jemalloc_internal_types.h" +#include "jemalloc/internal/mutex.h" +#include "jemalloc/internal/sc.h" + +/* Embedded at the beginning of every block of base-managed virtual memory. */ +struct base_block_s { + /* Total size of block's virtual memory mapping. */ + size_t size; + + /* Next block in list of base's blocks. */ + base_block_t *next; + + /* Tracks unused trailing space. */ + extent_t extent; +}; + +struct base_s { + /* Associated arena's index within the arenas array. */ + unsigned ind; + + /* + * User-configurable extent hook functions. Points to an + * extent_hooks_t. + */ + atomic_p_t extent_hooks; + + /* Protects base_alloc() and base_stats_get() operations. */ + malloc_mutex_t mtx; + + /* Using THP when true (metadata_thp auto mode). */ + bool auto_thp_switched; + /* + * Most recent size class in the series of increasingly large base + * extents. Logarithmic spacing between subsequent allocations ensures + * that the total number of distinct mappings remains small. + */ + pszind_t pind_last; + + /* Serial number generation state. */ + size_t extent_sn_next; + + /* Chain of all blocks associated with base. */ + base_block_t *blocks; + + /* Heap of extents that track unused trailing space within blocks. */ + extent_heap_t avail[SC_NSIZES]; + + /* Stats, only maintained if config_stats. */ + size_t allocated; + size_t resident; + size_t mapped; + /* Number of THP regions touched. */ + size_t n_thp; +}; + +#endif /* JEMALLOC_INTERNAL_BASE_STRUCTS_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/base_types.h b/deps/jemalloc/include/jemalloc/internal/base_types.h new file mode 100644 index 000000000..b6db77df7 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/base_types.h @@ -0,0 +1,33 @@ +#ifndef JEMALLOC_INTERNAL_BASE_TYPES_H +#define JEMALLOC_INTERNAL_BASE_TYPES_H + +typedef struct base_block_s base_block_t; +typedef struct base_s base_t; + +#define METADATA_THP_DEFAULT metadata_thp_disabled + +/* + * In auto mode, arenas switch to huge pages for the base allocator on the + * second base block. a0 switches to thp on the 5th block (after 20 megabytes + * of metadata), since more metadata (e.g. rtree nodes) come from a0's base. + */ + +#define BASE_AUTO_THP_THRESHOLD 2 +#define BASE_AUTO_THP_THRESHOLD_A0 5 + +typedef enum { + metadata_thp_disabled = 0, + /* + * Lazily enable hugepage for metadata. To avoid high RSS caused by THP + * + low usage arena (i.e. THP becomes a significant percentage), the + * "auto" option only starts using THP after a base allocator used up + * the first THP region. Starting from the second hugepage (in a single + * arena), "auto" behaves the same as "always", i.e. madvise hugepage + * right away. + */ + metadata_thp_auto = 1, + metadata_thp_always = 2, + metadata_thp_mode_limit = 3 +} metadata_thp_mode_t; + +#endif /* JEMALLOC_INTERNAL_BASE_TYPES_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/bin.h b/deps/jemalloc/include/jemalloc/internal/bin.h new file mode 100644 index 000000000..8547e8930 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/bin.h @@ -0,0 +1,123 @@ +#ifndef JEMALLOC_INTERNAL_BIN_H +#define JEMALLOC_INTERNAL_BIN_H + +#include "jemalloc/internal/bin_stats.h" +#include "jemalloc/internal/bin_types.h" +#include "jemalloc/internal/extent_types.h" +#include "jemalloc/internal/extent_structs.h" +#include "jemalloc/internal/mutex.h" +#include "jemalloc/internal/sc.h" + +/* + * A bin contains a set of extents that are currently being used for slab + * allocations. + */ + +/* + * Read-only information associated with each element of arena_t's bins array + * is stored separately, partly to reduce memory usage (only one copy, rather + * than one per arena), but mainly to avoid false cacheline sharing. + * + * Each slab has the following layout: + * + * /--------------------\ + * | region 0 | + * |--------------------| + * | region 1 | + * |--------------------| + * | ... | + * | ... | + * | ... | + * |--------------------| + * | region nregs-1 | + * \--------------------/ + */ +typedef struct bin_info_s bin_info_t; +struct bin_info_s { + /* Size of regions in a slab for this bin's size class. */ + size_t reg_size; + + /* Total size of a slab for this bin's size class. */ + size_t slab_size; + + /* Total number of regions in a slab for this bin's size class. */ + uint32_t nregs; + + /* Number of sharded bins in each arena for this size class. */ + uint32_t n_shards; + + /* + * Metadata used to manipulate bitmaps for slabs associated with this + * bin. + */ + bitmap_info_t bitmap_info; +}; + +extern bin_info_t bin_infos[SC_NBINS]; + +typedef struct bin_s bin_t; +struct bin_s { + /* All operations on bin_t fields require lock ownership. */ + malloc_mutex_t lock; + + /* + * Current slab being used to service allocations of this bin's size + * class. slabcur is independent of slabs_{nonfull,full}; whenever + * slabcur is reassigned, the previous slab must be deallocated or + * inserted into slabs_{nonfull,full}. + */ + extent_t *slabcur; + + /* + * Heap of non-full slabs. This heap is used to assure that new + * allocations come from the non-full slab that is oldest/lowest in + * memory. + */ + extent_heap_t slabs_nonfull; + + /* List used to track full slabs. */ + extent_list_t slabs_full; + + /* Bin statistics. */ + bin_stats_t stats; +}; + +/* A set of sharded bins of the same size class. */ +typedef struct bins_s bins_t; +struct bins_s { + /* Sharded bins. Dynamically sized. */ + bin_t *bin_shards; +}; + +void bin_shard_sizes_boot(unsigned bin_shards[SC_NBINS]); +bool bin_update_shard_size(unsigned bin_shards[SC_NBINS], size_t start_size, + size_t end_size, size_t nshards); +void bin_boot(sc_data_t *sc_data, unsigned bin_shard_sizes[SC_NBINS]); + +/* Initializes a bin to empty. Returns true on error. */ +bool bin_init(bin_t *bin); + +/* Forking. */ +void bin_prefork(tsdn_t *tsdn, bin_t *bin); +void bin_postfork_parent(tsdn_t *tsdn, bin_t *bin); +void bin_postfork_child(tsdn_t *tsdn, bin_t *bin); + +/* Stats. */ +static inline void +bin_stats_merge(tsdn_t *tsdn, bin_stats_t *dst_bin_stats, bin_t *bin) { + malloc_mutex_lock(tsdn, &bin->lock); + malloc_mutex_prof_accum(tsdn, &dst_bin_stats->mutex_data, &bin->lock); + dst_bin_stats->nmalloc += bin->stats.nmalloc; + dst_bin_stats->ndalloc += bin->stats.ndalloc; + dst_bin_stats->nrequests += bin->stats.nrequests; + dst_bin_stats->curregs += bin->stats.curregs; + dst_bin_stats->nfills += bin->stats.nfills; + dst_bin_stats->nflushes += bin->stats.nflushes; + dst_bin_stats->nslabs += bin->stats.nslabs; + dst_bin_stats->reslabs += bin->stats.reslabs; + dst_bin_stats->curslabs += bin->stats.curslabs; + dst_bin_stats->nonfull_slabs += bin->stats.nonfull_slabs; + malloc_mutex_unlock(tsdn, &bin->lock); +} + +#endif /* JEMALLOC_INTERNAL_BIN_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/bin_stats.h b/deps/jemalloc/include/jemalloc/internal/bin_stats.h new file mode 100644 index 000000000..d04519c82 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/bin_stats.h @@ -0,0 +1,54 @@ +#ifndef JEMALLOC_INTERNAL_BIN_STATS_H +#define JEMALLOC_INTERNAL_BIN_STATS_H + +#include "jemalloc/internal/mutex_prof.h" + +typedef struct bin_stats_s bin_stats_t; +struct bin_stats_s { + /* + * Total number of allocation/deallocation requests served directly by + * the bin. Note that tcache may allocate an object, then recycle it + * many times, resulting many increments to nrequests, but only one + * each to nmalloc and ndalloc. + */ + uint64_t nmalloc; + uint64_t ndalloc; + + /* + * Number of allocation requests that correspond to the size of this + * bin. This includes requests served by tcache, though tcache only + * periodically merges into this counter. + */ + uint64_t nrequests; + + /* + * Current number of regions of this size class, including regions + * currently cached by tcache. + */ + size_t curregs; + + /* Number of tcache fills from this bin. */ + uint64_t nfills; + + /* Number of tcache flushes to this bin. */ + uint64_t nflushes; + + /* Total number of slabs created for this bin's size class. */ + uint64_t nslabs; + + /* + * Total number of slabs reused by extracting them from the slabs heap + * for this bin's size class. + */ + uint64_t reslabs; + + /* Current number of slabs in this bin. */ + size_t curslabs; + + /* Current size of nonfull slabs heap in this bin. */ + size_t nonfull_slabs; + + mutex_prof_data_t mutex_data; +}; + +#endif /* JEMALLOC_INTERNAL_BIN_STATS_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/bin_types.h b/deps/jemalloc/include/jemalloc/internal/bin_types.h new file mode 100644 index 000000000..3533606b9 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/bin_types.h @@ -0,0 +1,17 @@ +#ifndef JEMALLOC_INTERNAL_BIN_TYPES_H +#define JEMALLOC_INTERNAL_BIN_TYPES_H + +#include "jemalloc/internal/sc.h" + +#define BIN_SHARDS_MAX (1 << EXTENT_BITS_BINSHARD_WIDTH) +#define N_BIN_SHARDS_DEFAULT 1 + +/* Used in TSD static initializer only. Real init in arena_bind(). */ +#define TSD_BINSHARDS_ZERO_INITIALIZER {{UINT8_MAX}} + +typedef struct tsd_binshards_s tsd_binshards_t; +struct tsd_binshards_s { + uint8_t binshard[SC_NBINS]; +}; + +#endif /* JEMALLOC_INTERNAL_BIN_TYPES_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/bit_util.h b/deps/jemalloc/include/jemalloc/internal/bit_util.h new file mode 100644 index 000000000..c045eb868 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/bit_util.h @@ -0,0 +1,239 @@ +#ifndef JEMALLOC_INTERNAL_BIT_UTIL_H +#define JEMALLOC_INTERNAL_BIT_UTIL_H + +#include "jemalloc/internal/assert.h" + +#define BIT_UTIL_INLINE static inline + +/* Sanity check. */ +#if !defined(JEMALLOC_INTERNAL_FFSLL) || !defined(JEMALLOC_INTERNAL_FFSL) \ + || !defined(JEMALLOC_INTERNAL_FFS) +# error JEMALLOC_INTERNAL_FFS{,L,LL} should have been defined by configure +#endif + + +BIT_UTIL_INLINE unsigned +ffs_llu(unsigned long long bitmap) { + return JEMALLOC_INTERNAL_FFSLL(bitmap); +} + +BIT_UTIL_INLINE unsigned +ffs_lu(unsigned long bitmap) { + return JEMALLOC_INTERNAL_FFSL(bitmap); +} + +BIT_UTIL_INLINE unsigned +ffs_u(unsigned bitmap) { + return JEMALLOC_INTERNAL_FFS(bitmap); +} + +#ifdef JEMALLOC_INTERNAL_POPCOUNTL +BIT_UTIL_INLINE unsigned +popcount_lu(unsigned long bitmap) { + return JEMALLOC_INTERNAL_POPCOUNTL(bitmap); +} +#endif + +/* + * Clears first unset bit in bitmap, and returns + * place of bit. bitmap *must not* be 0. + */ + +BIT_UTIL_INLINE size_t +cfs_lu(unsigned long* bitmap) { + size_t bit = ffs_lu(*bitmap) - 1; + *bitmap ^= ZU(1) << bit; + return bit; +} + +BIT_UTIL_INLINE unsigned +ffs_zu(size_t bitmap) { +#if LG_SIZEOF_PTR == LG_SIZEOF_INT + return ffs_u(bitmap); +#elif LG_SIZEOF_PTR == LG_SIZEOF_LONG + return ffs_lu(bitmap); +#elif LG_SIZEOF_PTR == LG_SIZEOF_LONG_LONG + return ffs_llu(bitmap); +#else +#error No implementation for size_t ffs() +#endif +} + +BIT_UTIL_INLINE unsigned +ffs_u64(uint64_t bitmap) { +#if LG_SIZEOF_LONG == 3 + return ffs_lu(bitmap); +#elif LG_SIZEOF_LONG_LONG == 3 + return ffs_llu(bitmap); +#else +#error No implementation for 64-bit ffs() +#endif +} + +BIT_UTIL_INLINE unsigned +ffs_u32(uint32_t bitmap) { +#if LG_SIZEOF_INT == 2 + return ffs_u(bitmap); +#else +#error No implementation for 32-bit ffs() +#endif + return ffs_u(bitmap); +} + +BIT_UTIL_INLINE uint64_t +pow2_ceil_u64(uint64_t x) { +#if (defined(__amd64__) || defined(__x86_64__) || defined(JEMALLOC_HAVE_BUILTIN_CLZ)) + if(unlikely(x <= 1)) { + return x; + } + size_t msb_on_index; +#if (defined(__amd64__) || defined(__x86_64__)) + asm ("bsrq %1, %0" + : "=r"(msb_on_index) // Outputs. + : "r"(x-1) // Inputs. + ); +#elif (defined(JEMALLOC_HAVE_BUILTIN_CLZ)) + msb_on_index = (63 ^ __builtin_clzll(x - 1)); +#endif + assert(msb_on_index < 63); + return 1ULL << (msb_on_index + 1); +#else + x--; + x |= x >> 1; + x |= x >> 2; + x |= x >> 4; + x |= x >> 8; + x |= x >> 16; + x |= x >> 32; + x++; + return x; +#endif +} + +BIT_UTIL_INLINE uint32_t +pow2_ceil_u32(uint32_t x) { +#if ((defined(__i386__) || defined(JEMALLOC_HAVE_BUILTIN_CLZ)) && (!defined(__s390__))) + if(unlikely(x <= 1)) { + return x; + } + size_t msb_on_index; +#if (defined(__i386__)) + asm ("bsr %1, %0" + : "=r"(msb_on_index) // Outputs. + : "r"(x-1) // Inputs. + ); +#elif (defined(JEMALLOC_HAVE_BUILTIN_CLZ)) + msb_on_index = (31 ^ __builtin_clz(x - 1)); +#endif + assert(msb_on_index < 31); + return 1U << (msb_on_index + 1); +#else + x--; + x |= x >> 1; + x |= x >> 2; + x |= x >> 4; + x |= x >> 8; + x |= x >> 16; + x++; + return x; +#endif +} + +/* Compute the smallest power of 2 that is >= x. */ +BIT_UTIL_INLINE size_t +pow2_ceil_zu(size_t x) { +#if (LG_SIZEOF_PTR == 3) + return pow2_ceil_u64(x); +#else + return pow2_ceil_u32(x); +#endif +} + +#if (defined(__i386__) || defined(__amd64__) || defined(__x86_64__)) +BIT_UTIL_INLINE unsigned +lg_floor(size_t x) { + size_t ret; + assert(x != 0); + + asm ("bsr %1, %0" + : "=r"(ret) // Outputs. + : "r"(x) // Inputs. + ); + assert(ret < UINT_MAX); + return (unsigned)ret; +} +#elif (defined(_MSC_VER)) +BIT_UTIL_INLINE unsigned +lg_floor(size_t x) { + unsigned long ret; + + assert(x != 0); + +#if (LG_SIZEOF_PTR == 3) + _BitScanReverse64(&ret, x); +#elif (LG_SIZEOF_PTR == 2) + _BitScanReverse(&ret, x); +#else +# error "Unsupported type size for lg_floor()" +#endif + assert(ret < UINT_MAX); + return (unsigned)ret; +} +#elif (defined(JEMALLOC_HAVE_BUILTIN_CLZ)) +BIT_UTIL_INLINE unsigned +lg_floor(size_t x) { + assert(x != 0); + +#if (LG_SIZEOF_PTR == LG_SIZEOF_INT) + return ((8 << LG_SIZEOF_PTR) - 1) - __builtin_clz(x); +#elif (LG_SIZEOF_PTR == LG_SIZEOF_LONG) + return ((8 << LG_SIZEOF_PTR) - 1) - __builtin_clzl(x); +#else +# error "Unsupported type size for lg_floor()" +#endif +} +#else +BIT_UTIL_INLINE unsigned +lg_floor(size_t x) { + assert(x != 0); + + x |= (x >> 1); + x |= (x >> 2); + x |= (x >> 4); + x |= (x >> 8); + x |= (x >> 16); +#if (LG_SIZEOF_PTR == 3) + x |= (x >> 32); +#endif + if (x == SIZE_T_MAX) { + return (8 << LG_SIZEOF_PTR) - 1; + } + x++; + return ffs_zu(x) - 2; +} +#endif + +BIT_UTIL_INLINE unsigned +lg_ceil(size_t x) { + return lg_floor(x) + ((x & (x - 1)) == 0 ? 0 : 1); +} + +#undef BIT_UTIL_INLINE + +/* A compile-time version of lg_floor and lg_ceil. */ +#define LG_FLOOR_1(x) 0 +#define LG_FLOOR_2(x) (x < (1ULL << 1) ? LG_FLOOR_1(x) : 1 + LG_FLOOR_1(x >> 1)) +#define LG_FLOOR_4(x) (x < (1ULL << 2) ? LG_FLOOR_2(x) : 2 + LG_FLOOR_2(x >> 2)) +#define LG_FLOOR_8(x) (x < (1ULL << 4) ? LG_FLOOR_4(x) : 4 + LG_FLOOR_4(x >> 4)) +#define LG_FLOOR_16(x) (x < (1ULL << 8) ? LG_FLOOR_8(x) : 8 + LG_FLOOR_8(x >> 8)) +#define LG_FLOOR_32(x) (x < (1ULL << 16) ? LG_FLOOR_16(x) : 16 + LG_FLOOR_16(x >> 16)) +#define LG_FLOOR_64(x) (x < (1ULL << 32) ? LG_FLOOR_32(x) : 32 + LG_FLOOR_32(x >> 32)) +#if LG_SIZEOF_PTR == 2 +# define LG_FLOOR(x) LG_FLOOR_32((x)) +#else +# define LG_FLOOR(x) LG_FLOOR_64((x)) +#endif + +#define LG_CEIL(x) (LG_FLOOR(x) + (((x) & ((x) - 1)) == 0 ? 0 : 1)) + +#endif /* JEMALLOC_INTERNAL_BIT_UTIL_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/bitmap.h b/deps/jemalloc/include/jemalloc/internal/bitmap.h new file mode 100644 index 000000000..c3f9cb490 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/bitmap.h @@ -0,0 +1,369 @@ +#ifndef JEMALLOC_INTERNAL_BITMAP_H +#define JEMALLOC_INTERNAL_BITMAP_H + +#include "jemalloc/internal/arena_types.h" +#include "jemalloc/internal/bit_util.h" +#include "jemalloc/internal/sc.h" + +typedef unsigned long bitmap_t; +#define LG_SIZEOF_BITMAP LG_SIZEOF_LONG + +/* Maximum bitmap bit count is 2^LG_BITMAP_MAXBITS. */ +#if LG_SLAB_MAXREGS > LG_CEIL(SC_NSIZES) +/* Maximum bitmap bit count is determined by maximum regions per slab. */ +# define LG_BITMAP_MAXBITS LG_SLAB_MAXREGS +#else +/* Maximum bitmap bit count is determined by number of extent size classes. */ +# define LG_BITMAP_MAXBITS LG_CEIL(SC_NSIZES) +#endif +#define BITMAP_MAXBITS (ZU(1) << LG_BITMAP_MAXBITS) + +/* Number of bits per group. */ +#define LG_BITMAP_GROUP_NBITS (LG_SIZEOF_BITMAP + 3) +#define BITMAP_GROUP_NBITS (1U << LG_BITMAP_GROUP_NBITS) +#define BITMAP_GROUP_NBITS_MASK (BITMAP_GROUP_NBITS-1) + +/* + * Do some analysis on how big the bitmap is before we use a tree. For a brute + * force linear search, if we would have to call ffs_lu() more than 2^3 times, + * use a tree instead. + */ +#if LG_BITMAP_MAXBITS - LG_BITMAP_GROUP_NBITS > 3 +# define BITMAP_USE_TREE +#endif + +/* Number of groups required to store a given number of bits. */ +#define BITMAP_BITS2GROUPS(nbits) \ + (((nbits) + BITMAP_GROUP_NBITS_MASK) >> LG_BITMAP_GROUP_NBITS) + +/* + * Number of groups required at a particular level for a given number of bits. + */ +#define BITMAP_GROUPS_L0(nbits) \ + BITMAP_BITS2GROUPS(nbits) +#define BITMAP_GROUPS_L1(nbits) \ + BITMAP_BITS2GROUPS(BITMAP_BITS2GROUPS(nbits)) +#define BITMAP_GROUPS_L2(nbits) \ + BITMAP_BITS2GROUPS(BITMAP_BITS2GROUPS(BITMAP_BITS2GROUPS((nbits)))) +#define BITMAP_GROUPS_L3(nbits) \ + BITMAP_BITS2GROUPS(BITMAP_BITS2GROUPS(BITMAP_BITS2GROUPS( \ + BITMAP_BITS2GROUPS((nbits))))) +#define BITMAP_GROUPS_L4(nbits) \ + BITMAP_BITS2GROUPS(BITMAP_BITS2GROUPS(BITMAP_BITS2GROUPS( \ + BITMAP_BITS2GROUPS(BITMAP_BITS2GROUPS((nbits)))))) + +/* + * Assuming the number of levels, number of groups required for a given number + * of bits. + */ +#define BITMAP_GROUPS_1_LEVEL(nbits) \ + BITMAP_GROUPS_L0(nbits) +#define BITMAP_GROUPS_2_LEVEL(nbits) \ + (BITMAP_GROUPS_1_LEVEL(nbits) + BITMAP_GROUPS_L1(nbits)) +#define BITMAP_GROUPS_3_LEVEL(nbits) \ + (BITMAP_GROUPS_2_LEVEL(nbits) + BITMAP_GROUPS_L2(nbits)) +#define BITMAP_GROUPS_4_LEVEL(nbits) \ + (BITMAP_GROUPS_3_LEVEL(nbits) + BITMAP_GROUPS_L3(nbits)) +#define BITMAP_GROUPS_5_LEVEL(nbits) \ + (BITMAP_GROUPS_4_LEVEL(nbits) + BITMAP_GROUPS_L4(nbits)) + +/* + * Maximum number of groups required to support LG_BITMAP_MAXBITS. + */ +#ifdef BITMAP_USE_TREE + +#if LG_BITMAP_MAXBITS <= LG_BITMAP_GROUP_NBITS +# define BITMAP_GROUPS(nbits) BITMAP_GROUPS_1_LEVEL(nbits) +# define BITMAP_GROUPS_MAX BITMAP_GROUPS_1_LEVEL(BITMAP_MAXBITS) +#elif LG_BITMAP_MAXBITS <= LG_BITMAP_GROUP_NBITS * 2 +# define BITMAP_GROUPS(nbits) BITMAP_GROUPS_2_LEVEL(nbits) +# define BITMAP_GROUPS_MAX BITMAP_GROUPS_2_LEVEL(BITMAP_MAXBITS) +#elif LG_BITMAP_MAXBITS <= LG_BITMAP_GROUP_NBITS * 3 +# define BITMAP_GROUPS(nbits) BITMAP_GROUPS_3_LEVEL(nbits) +# define BITMAP_GROUPS_MAX BITMAP_GROUPS_3_LEVEL(BITMAP_MAXBITS) +#elif LG_BITMAP_MAXBITS <= LG_BITMAP_GROUP_NBITS * 4 +# define BITMAP_GROUPS(nbits) BITMAP_GROUPS_4_LEVEL(nbits) +# define BITMAP_GROUPS_MAX BITMAP_GROUPS_4_LEVEL(BITMAP_MAXBITS) +#elif LG_BITMAP_MAXBITS <= LG_BITMAP_GROUP_NBITS * 5 +# define BITMAP_GROUPS(nbits) BITMAP_GROUPS_5_LEVEL(nbits) +# define BITMAP_GROUPS_MAX BITMAP_GROUPS_5_LEVEL(BITMAP_MAXBITS) +#else +# error "Unsupported bitmap size" +#endif + +/* + * Maximum number of levels possible. This could be statically computed based + * on LG_BITMAP_MAXBITS: + * + * #define BITMAP_MAX_LEVELS \ + * (LG_BITMAP_MAXBITS / LG_SIZEOF_BITMAP) \ + * + !!(LG_BITMAP_MAXBITS % LG_SIZEOF_BITMAP) + * + * However, that would not allow the generic BITMAP_INFO_INITIALIZER() macro, so + * instead hardcode BITMAP_MAX_LEVELS to the largest number supported by the + * various cascading macros. The only additional cost this incurs is some + * unused trailing entries in bitmap_info_t structures; the bitmaps themselves + * are not impacted. + */ +#define BITMAP_MAX_LEVELS 5 + +#define BITMAP_INFO_INITIALIZER(nbits) { \ + /* nbits. */ \ + nbits, \ + /* nlevels. */ \ + (BITMAP_GROUPS_L0(nbits) > BITMAP_GROUPS_L1(nbits)) + \ + (BITMAP_GROUPS_L1(nbits) > BITMAP_GROUPS_L2(nbits)) + \ + (BITMAP_GROUPS_L2(nbits) > BITMAP_GROUPS_L3(nbits)) + \ + (BITMAP_GROUPS_L3(nbits) > BITMAP_GROUPS_L4(nbits)) + 1, \ + /* levels. */ \ + { \ + {0}, \ + {BITMAP_GROUPS_L0(nbits)}, \ + {BITMAP_GROUPS_L1(nbits) + BITMAP_GROUPS_L0(nbits)}, \ + {BITMAP_GROUPS_L2(nbits) + BITMAP_GROUPS_L1(nbits) + \ + BITMAP_GROUPS_L0(nbits)}, \ + {BITMAP_GROUPS_L3(nbits) + BITMAP_GROUPS_L2(nbits) + \ + BITMAP_GROUPS_L1(nbits) + BITMAP_GROUPS_L0(nbits)}, \ + {BITMAP_GROUPS_L4(nbits) + BITMAP_GROUPS_L3(nbits) + \ + BITMAP_GROUPS_L2(nbits) + BITMAP_GROUPS_L1(nbits) \ + + BITMAP_GROUPS_L0(nbits)} \ + } \ +} + +#else /* BITMAP_USE_TREE */ + +#define BITMAP_GROUPS(nbits) BITMAP_BITS2GROUPS(nbits) +#define BITMAP_GROUPS_MAX BITMAP_BITS2GROUPS(BITMAP_MAXBITS) + +#define BITMAP_INFO_INITIALIZER(nbits) { \ + /* nbits. */ \ + nbits, \ + /* ngroups. */ \ + BITMAP_BITS2GROUPS(nbits) \ +} + +#endif /* BITMAP_USE_TREE */ + +typedef struct bitmap_level_s { + /* Offset of this level's groups within the array of groups. */ + size_t group_offset; +} bitmap_level_t; + +typedef struct bitmap_info_s { + /* Logical number of bits in bitmap (stored at bottom level). */ + size_t nbits; + +#ifdef BITMAP_USE_TREE + /* Number of levels necessary for nbits. */ + unsigned nlevels; + + /* + * Only the first (nlevels+1) elements are used, and levels are ordered + * bottom to top (e.g. the bottom level is stored in levels[0]). + */ + bitmap_level_t levels[BITMAP_MAX_LEVELS+1]; +#else /* BITMAP_USE_TREE */ + /* Number of groups necessary for nbits. */ + size_t ngroups; +#endif /* BITMAP_USE_TREE */ +} bitmap_info_t; + +void bitmap_info_init(bitmap_info_t *binfo, size_t nbits); +void bitmap_init(bitmap_t *bitmap, const bitmap_info_t *binfo, bool fill); +size_t bitmap_size(const bitmap_info_t *binfo); + +static inline bool +bitmap_full(bitmap_t *bitmap, const bitmap_info_t *binfo) { +#ifdef BITMAP_USE_TREE + size_t rgoff = binfo->levels[binfo->nlevels].group_offset - 1; + bitmap_t rg = bitmap[rgoff]; + /* The bitmap is full iff the root group is 0. */ + return (rg == 0); +#else + size_t i; + + for (i = 0; i < binfo->ngroups; i++) { + if (bitmap[i] != 0) { + return false; + } + } + return true; +#endif +} + +static inline bool +bitmap_get(bitmap_t *bitmap, const bitmap_info_t *binfo, size_t bit) { + size_t goff; + bitmap_t g; + + assert(bit < binfo->nbits); + goff = bit >> LG_BITMAP_GROUP_NBITS; + g = bitmap[goff]; + return !(g & (ZU(1) << (bit & BITMAP_GROUP_NBITS_MASK))); +} + +static inline void +bitmap_set(bitmap_t *bitmap, const bitmap_info_t *binfo, size_t bit) { + size_t goff; + bitmap_t *gp; + bitmap_t g; + + assert(bit < binfo->nbits); + assert(!bitmap_get(bitmap, binfo, bit)); + goff = bit >> LG_BITMAP_GROUP_NBITS; + gp = &bitmap[goff]; + g = *gp; + assert(g & (ZU(1) << (bit & BITMAP_GROUP_NBITS_MASK))); + g ^= ZU(1) << (bit & BITMAP_GROUP_NBITS_MASK); + *gp = g; + assert(bitmap_get(bitmap, binfo, bit)); +#ifdef BITMAP_USE_TREE + /* Propagate group state transitions up the tree. */ + if (g == 0) { + unsigned i; + for (i = 1; i < binfo->nlevels; i++) { + bit = goff; + goff = bit >> LG_BITMAP_GROUP_NBITS; + gp = &bitmap[binfo->levels[i].group_offset + goff]; + g = *gp; + assert(g & (ZU(1) << (bit & BITMAP_GROUP_NBITS_MASK))); + g ^= ZU(1) << (bit & BITMAP_GROUP_NBITS_MASK); + *gp = g; + if (g != 0) { + break; + } + } + } +#endif +} + +/* ffu: find first unset >= bit. */ +static inline size_t +bitmap_ffu(const bitmap_t *bitmap, const bitmap_info_t *binfo, size_t min_bit) { + assert(min_bit < binfo->nbits); + +#ifdef BITMAP_USE_TREE + size_t bit = 0; + for (unsigned level = binfo->nlevels; level--;) { + size_t lg_bits_per_group = (LG_BITMAP_GROUP_NBITS * (level + + 1)); + bitmap_t group = bitmap[binfo->levels[level].group_offset + (bit + >> lg_bits_per_group)]; + unsigned group_nmask = (unsigned)(((min_bit > bit) ? (min_bit - + bit) : 0) >> (lg_bits_per_group - LG_BITMAP_GROUP_NBITS)); + assert(group_nmask <= BITMAP_GROUP_NBITS); + bitmap_t group_mask = ~((1LU << group_nmask) - 1); + bitmap_t group_masked = group & group_mask; + if (group_masked == 0LU) { + if (group == 0LU) { + return binfo->nbits; + } + /* + * min_bit was preceded by one or more unset bits in + * this group, but there are no other unset bits in this + * group. Try again starting at the first bit of the + * next sibling. This will recurse at most once per + * non-root level. + */ + size_t sib_base = bit + (ZU(1) << lg_bits_per_group); + assert(sib_base > min_bit); + assert(sib_base > bit); + if (sib_base >= binfo->nbits) { + return binfo->nbits; + } + return bitmap_ffu(bitmap, binfo, sib_base); + } + bit += ((size_t)(ffs_lu(group_masked) - 1)) << + (lg_bits_per_group - LG_BITMAP_GROUP_NBITS); + } + assert(bit >= min_bit); + assert(bit < binfo->nbits); + return bit; +#else + size_t i = min_bit >> LG_BITMAP_GROUP_NBITS; + bitmap_t g = bitmap[i] & ~((1LU << (min_bit & BITMAP_GROUP_NBITS_MASK)) + - 1); + size_t bit; + do { + bit = ffs_lu(g); + if (bit != 0) { + return (i << LG_BITMAP_GROUP_NBITS) + (bit - 1); + } + i++; + g = bitmap[i]; + } while (i < binfo->ngroups); + return binfo->nbits; +#endif +} + +/* sfu: set first unset. */ +static inline size_t +bitmap_sfu(bitmap_t *bitmap, const bitmap_info_t *binfo) { + size_t bit; + bitmap_t g; + unsigned i; + + assert(!bitmap_full(bitmap, binfo)); + +#ifdef BITMAP_USE_TREE + i = binfo->nlevels - 1; + g = bitmap[binfo->levels[i].group_offset]; + bit = ffs_lu(g) - 1; + while (i > 0) { + i--; + g = bitmap[binfo->levels[i].group_offset + bit]; + bit = (bit << LG_BITMAP_GROUP_NBITS) + (ffs_lu(g) - 1); + } +#else + i = 0; + g = bitmap[0]; + while ((bit = ffs_lu(g)) == 0) { + i++; + g = bitmap[i]; + } + bit = (i << LG_BITMAP_GROUP_NBITS) + (bit - 1); +#endif + bitmap_set(bitmap, binfo, bit); + return bit; +} + +static inline void +bitmap_unset(bitmap_t *bitmap, const bitmap_info_t *binfo, size_t bit) { + size_t goff; + bitmap_t *gp; + bitmap_t g; + UNUSED bool propagate; + + assert(bit < binfo->nbits); + assert(bitmap_get(bitmap, binfo, bit)); + goff = bit >> LG_BITMAP_GROUP_NBITS; + gp = &bitmap[goff]; + g = *gp; + propagate = (g == 0); + assert((g & (ZU(1) << (bit & BITMAP_GROUP_NBITS_MASK))) == 0); + g ^= ZU(1) << (bit & BITMAP_GROUP_NBITS_MASK); + *gp = g; + assert(!bitmap_get(bitmap, binfo, bit)); +#ifdef BITMAP_USE_TREE + /* Propagate group state transitions up the tree. */ + if (propagate) { + unsigned i; + for (i = 1; i < binfo->nlevels; i++) { + bit = goff; + goff = bit >> LG_BITMAP_GROUP_NBITS; + gp = &bitmap[binfo->levels[i].group_offset + goff]; + g = *gp; + propagate = (g == 0); + assert((g & (ZU(1) << (bit & BITMAP_GROUP_NBITS_MASK))) + == 0); + g ^= ZU(1) << (bit & BITMAP_GROUP_NBITS_MASK); + *gp = g; + if (!propagate) { + break; + } + } + } +#endif /* BITMAP_USE_TREE */ +} + +#endif /* JEMALLOC_INTERNAL_BITMAP_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/cache_bin.h b/deps/jemalloc/include/jemalloc/internal/cache_bin.h new file mode 100644 index 000000000..d14556a3d --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/cache_bin.h @@ -0,0 +1,131 @@ +#ifndef JEMALLOC_INTERNAL_CACHE_BIN_H +#define JEMALLOC_INTERNAL_CACHE_BIN_H + +#include "jemalloc/internal/ql.h" + +/* + * The cache_bins are the mechanism that the tcache and the arena use to + * communicate. The tcache fills from and flushes to the arena by passing a + * cache_bin_t to fill/flush. When the arena needs to pull stats from the + * tcaches associated with it, it does so by iterating over its + * cache_bin_array_descriptor_t objects and reading out per-bin stats it + * contains. This makes it so that the arena need not know about the existence + * of the tcache at all. + */ + + +/* + * The count of the number of cached allocations in a bin. We make this signed + * so that negative numbers can encode "invalid" states (e.g. a low water mark + * of -1 for a cache that has been depleted). + */ +typedef int32_t cache_bin_sz_t; + +typedef struct cache_bin_stats_s cache_bin_stats_t; +struct cache_bin_stats_s { + /* + * Number of allocation requests that corresponded to the size of this + * bin. + */ + uint64_t nrequests; +}; + +/* + * Read-only information associated with each element of tcache_t's tbins array + * is stored separately, mainly to reduce memory usage. + */ +typedef struct cache_bin_info_s cache_bin_info_t; +struct cache_bin_info_s { + /* Upper limit on ncached. */ + cache_bin_sz_t ncached_max; +}; + +typedef struct cache_bin_s cache_bin_t; +struct cache_bin_s { + /* Min # cached since last GC. */ + cache_bin_sz_t low_water; + /* # of cached objects. */ + cache_bin_sz_t ncached; + /* + * ncached and stats are both modified frequently. Let's keep them + * close so that they have a higher chance of being on the same + * cacheline, thus less write-backs. + */ + cache_bin_stats_t tstats; + /* + * Stack of available objects. + * + * To make use of adjacent cacheline prefetch, the items in the avail + * stack goes to higher address for newer allocations. avail points + * just above the available space, which means that + * avail[-ncached, ... -1] are available items and the lowest item will + * be allocated first. + */ + void **avail; +}; + +typedef struct cache_bin_array_descriptor_s cache_bin_array_descriptor_t; +struct cache_bin_array_descriptor_s { + /* + * The arena keeps a list of the cache bins associated with it, for + * stats collection. + */ + ql_elm(cache_bin_array_descriptor_t) link; + /* Pointers to the tcache bins. */ + cache_bin_t *bins_small; + cache_bin_t *bins_large; +}; + +static inline void +cache_bin_array_descriptor_init(cache_bin_array_descriptor_t *descriptor, + cache_bin_t *bins_small, cache_bin_t *bins_large) { + ql_elm_new(descriptor, link); + descriptor->bins_small = bins_small; + descriptor->bins_large = bins_large; +} + +JEMALLOC_ALWAYS_INLINE void * +cache_bin_alloc_easy(cache_bin_t *bin, bool *success) { + void *ret; + + bin->ncached--; + + /* + * Check for both bin->ncached == 0 and ncached < low_water + * in a single branch. + */ + if (unlikely(bin->ncached <= bin->low_water)) { + bin->low_water = bin->ncached; + if (bin->ncached == -1) { + bin->ncached = 0; + *success = false; + return NULL; + } + } + + /* + * success (instead of ret) should be checked upon the return of this + * function. We avoid checking (ret == NULL) because there is never a + * null stored on the avail stack (which is unknown to the compiler), + * and eagerly checking ret would cause pipeline stall (waiting for the + * cacheline). + */ + *success = true; + ret = *(bin->avail - (bin->ncached + 1)); + + return ret; +} + +JEMALLOC_ALWAYS_INLINE bool +cache_bin_dalloc_easy(cache_bin_t *bin, cache_bin_info_t *bin_info, void *ptr) { + if (unlikely(bin->ncached == bin_info->ncached_max)) { + return false; + } + assert(bin->ncached < bin_info->ncached_max); + bin->ncached++; + *(bin->avail - bin->ncached) = ptr; + + return true; +} + +#endif /* JEMALLOC_INTERNAL_CACHE_BIN_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/ckh.h b/deps/jemalloc/include/jemalloc/internal/ckh.h new file mode 100644 index 000000000..7b3850bc1 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/ckh.h @@ -0,0 +1,101 @@ +#ifndef JEMALLOC_INTERNAL_CKH_H +#define JEMALLOC_INTERNAL_CKH_H + +#include "jemalloc/internal/tsd.h" + +/* Cuckoo hashing implementation. Skip to the end for the interface. */ + +/******************************************************************************/ +/* INTERNAL DEFINITIONS -- IGNORE */ +/******************************************************************************/ + +/* Maintain counters used to get an idea of performance. */ +/* #define CKH_COUNT */ +/* Print counter values in ckh_delete() (requires CKH_COUNT). */ +/* #define CKH_VERBOSE */ + +/* + * There are 2^LG_CKH_BUCKET_CELLS cells in each hash table bucket. Try to fit + * one bucket per L1 cache line. + */ +#define LG_CKH_BUCKET_CELLS (LG_CACHELINE - LG_SIZEOF_PTR - 1) + +/* Typedefs to allow easy function pointer passing. */ +typedef void ckh_hash_t (const void *, size_t[2]); +typedef bool ckh_keycomp_t (const void *, const void *); + +/* Hash table cell. */ +typedef struct { + const void *key; + const void *data; +} ckhc_t; + +/* The hash table itself. */ +typedef struct { +#ifdef CKH_COUNT + /* Counters used to get an idea of performance. */ + uint64_t ngrows; + uint64_t nshrinks; + uint64_t nshrinkfails; + uint64_t ninserts; + uint64_t nrelocs; +#endif + + /* Used for pseudo-random number generation. */ + uint64_t prng_state; + + /* Total number of items. */ + size_t count; + + /* + * Minimum and current number of hash table buckets. There are + * 2^LG_CKH_BUCKET_CELLS cells per bucket. + */ + unsigned lg_minbuckets; + unsigned lg_curbuckets; + + /* Hash and comparison functions. */ + ckh_hash_t *hash; + ckh_keycomp_t *keycomp; + + /* Hash table with 2^lg_curbuckets buckets. */ + ckhc_t *tab; +} ckh_t; + +/******************************************************************************/ +/* BEGIN PUBLIC API */ +/******************************************************************************/ + +/* Lifetime management. Minitems is the initial capacity. */ +bool ckh_new(tsd_t *tsd, ckh_t *ckh, size_t minitems, ckh_hash_t *hash, + ckh_keycomp_t *keycomp); +void ckh_delete(tsd_t *tsd, ckh_t *ckh); + +/* Get the number of elements in the set. */ +size_t ckh_count(ckh_t *ckh); + +/* + * To iterate over the elements in the table, initialize *tabind to 0 and call + * this function until it returns true. Each call that returns false will + * update *key and *data to the next element in the table, assuming the pointers + * are non-NULL. + */ +bool ckh_iter(ckh_t *ckh, size_t *tabind, void **key, void **data); + +/* + * Basic hash table operations -- insert, removal, lookup. For ckh_remove and + * ckh_search, key or data can be NULL. The hash-table only stores pointers to + * the key and value, and doesn't do any lifetime management. + */ +bool ckh_insert(tsd_t *tsd, ckh_t *ckh, const void *key, const void *data); +bool ckh_remove(tsd_t *tsd, ckh_t *ckh, const void *searchkey, void **key, + void **data); +bool ckh_search(ckh_t *ckh, const void *searchkey, void **key, void **data); + +/* Some useful hash and comparison functions for strings and pointers. */ +void ckh_string_hash(const void *key, size_t r_hash[2]); +bool ckh_string_keycomp(const void *k1, const void *k2); +void ckh_pointer_hash(const void *key, size_t r_hash[2]); +bool ckh_pointer_keycomp(const void *k1, const void *k2); + +#endif /* JEMALLOC_INTERNAL_CKH_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/ctl.h b/deps/jemalloc/include/jemalloc/internal/ctl.h new file mode 100644 index 000000000..1d1aacc6f --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/ctl.h @@ -0,0 +1,134 @@ +#ifndef JEMALLOC_INTERNAL_CTL_H +#define JEMALLOC_INTERNAL_CTL_H + +#include "jemalloc/internal/jemalloc_internal_types.h" +#include "jemalloc/internal/malloc_io.h" +#include "jemalloc/internal/mutex_prof.h" +#include "jemalloc/internal/ql.h" +#include "jemalloc/internal/sc.h" +#include "jemalloc/internal/stats.h" + +/* Maximum ctl tree depth. */ +#define CTL_MAX_DEPTH 7 + +typedef struct ctl_node_s { + bool named; +} ctl_node_t; + +typedef struct ctl_named_node_s { + ctl_node_t node; + const char *name; + /* If (nchildren == 0), this is a terminal node. */ + size_t nchildren; + const ctl_node_t *children; + int (*ctl)(tsd_t *, const size_t *, size_t, void *, size_t *, void *, + size_t); +} ctl_named_node_t; + +typedef struct ctl_indexed_node_s { + struct ctl_node_s node; + const ctl_named_node_t *(*index)(tsdn_t *, const size_t *, size_t, + size_t); +} ctl_indexed_node_t; + +typedef struct ctl_arena_stats_s { + arena_stats_t astats; + + /* Aggregate stats for small size classes, based on bin stats. */ + size_t allocated_small; + uint64_t nmalloc_small; + uint64_t ndalloc_small; + uint64_t nrequests_small; + uint64_t nfills_small; + uint64_t nflushes_small; + + bin_stats_t bstats[SC_NBINS]; + arena_stats_large_t lstats[SC_NSIZES - SC_NBINS]; + arena_stats_extents_t estats[SC_NPSIZES]; +} ctl_arena_stats_t; + +typedef struct ctl_stats_s { + size_t allocated; + size_t active; + size_t metadata; + size_t metadata_thp; + size_t resident; + size_t mapped; + size_t retained; + + background_thread_stats_t background_thread; + mutex_prof_data_t mutex_prof_data[mutex_prof_num_global_mutexes]; +} ctl_stats_t; + +typedef struct ctl_arena_s ctl_arena_t; +struct ctl_arena_s { + unsigned arena_ind; + bool initialized; + ql_elm(ctl_arena_t) destroyed_link; + + /* Basic stats, supported even if !config_stats. */ + unsigned nthreads; + const char *dss; + ssize_t dirty_decay_ms; + ssize_t muzzy_decay_ms; + size_t pactive; + size_t pdirty; + size_t pmuzzy; + + /* NULL if !config_stats. */ + ctl_arena_stats_t *astats; +}; + +typedef struct ctl_arenas_s { + uint64_t epoch; + unsigned narenas; + ql_head(ctl_arena_t) destroyed; + + /* + * Element 0 corresponds to merged stats for extant arenas (accessed via + * MALLCTL_ARENAS_ALL), element 1 corresponds to merged stats for + * destroyed arenas (accessed via MALLCTL_ARENAS_DESTROYED), and the + * remaining MALLOCX_ARENA_LIMIT elements correspond to arenas. + */ + ctl_arena_t *arenas[2 + MALLOCX_ARENA_LIMIT]; +} ctl_arenas_t; + +int ctl_byname(tsd_t *tsd, const char *name, void *oldp, size_t *oldlenp, + void *newp, size_t newlen); +int ctl_nametomib(tsd_t *tsd, const char *name, size_t *mibp, size_t *miblenp); + +int ctl_bymib(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp, + size_t *oldlenp, void *newp, size_t newlen); +bool ctl_boot(void); +void ctl_prefork(tsdn_t *tsdn); +void ctl_postfork_parent(tsdn_t *tsdn); +void ctl_postfork_child(tsdn_t *tsdn); + +#define xmallctl(name, oldp, oldlenp, newp, newlen) do { \ + if (je_mallctl(name, oldp, oldlenp, newp, newlen) \ + != 0) { \ + malloc_printf( \ + "<jemalloc>: Failure in xmallctl(\"%s\", ...)\n", \ + name); \ + abort(); \ + } \ +} while (0) + +#define xmallctlnametomib(name, mibp, miblenp) do { \ + if (je_mallctlnametomib(name, mibp, miblenp) != 0) { \ + malloc_printf("<jemalloc>: Failure in " \ + "xmallctlnametomib(\"%s\", ...)\n", name); \ + abort(); \ + } \ +} while (0) + +#define xmallctlbymib(mib, miblen, oldp, oldlenp, newp, newlen) do { \ + if (je_mallctlbymib(mib, miblen, oldp, oldlenp, newp, \ + newlen) != 0) { \ + malloc_write( \ + "<jemalloc>: Failure in xmallctlbymib()\n"); \ + abort(); \ + } \ +} while (0) + +#endif /* JEMALLOC_INTERNAL_CTL_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/div.h b/deps/jemalloc/include/jemalloc/internal/div.h new file mode 100644 index 000000000..aebae9398 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/div.h @@ -0,0 +1,41 @@ +#ifndef JEMALLOC_INTERNAL_DIV_H +#define JEMALLOC_INTERNAL_DIV_H + +#include "jemalloc/internal/assert.h" + +/* + * This module does the division that computes the index of a region in a slab, + * given its offset relative to the base. + * That is, given a divisor d, an n = i * d (all integers), we'll return i. + * We do some pre-computation to do this more quickly than a CPU division + * instruction. + * We bound n < 2^32, and don't support dividing by one. + */ + +typedef struct div_info_s div_info_t; +struct div_info_s { + uint32_t magic; +#ifdef JEMALLOC_DEBUG + size_t d; +#endif +}; + +void div_init(div_info_t *div_info, size_t divisor); + +static inline size_t +div_compute(div_info_t *div_info, size_t n) { + assert(n <= (uint32_t)-1); + /* + * This generates, e.g. mov; imul; shr on x86-64. On a 32-bit machine, + * the compilers I tried were all smart enough to turn this into the + * appropriate "get the high 32 bits of the result of a multiply" (e.g. + * mul; mov edx eax; on x86, umull on arm, etc.). + */ + size_t i = ((uint64_t)n * (uint64_t)div_info->magic) >> 32; +#ifdef JEMALLOC_DEBUG + assert(i * div_info->d == n); +#endif + return i; +} + +#endif /* JEMALLOC_INTERNAL_DIV_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/emitter.h b/deps/jemalloc/include/jemalloc/internal/emitter.h new file mode 100644 index 000000000..542bc79c3 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/emitter.h @@ -0,0 +1,486 @@ +#ifndef JEMALLOC_INTERNAL_EMITTER_H +#define JEMALLOC_INTERNAL_EMITTER_H + +#include "jemalloc/internal/ql.h" + +typedef enum emitter_output_e emitter_output_t; +enum emitter_output_e { + emitter_output_json, + emitter_output_table +}; + +typedef enum emitter_justify_e emitter_justify_t; +enum emitter_justify_e { + emitter_justify_left, + emitter_justify_right, + /* Not for users; just to pass to internal functions. */ + emitter_justify_none +}; + +typedef enum emitter_type_e emitter_type_t; +enum emitter_type_e { + emitter_type_bool, + emitter_type_int, + emitter_type_unsigned, + emitter_type_uint32, + emitter_type_uint64, + emitter_type_size, + emitter_type_ssize, + emitter_type_string, + /* + * A title is a column title in a table; it's just a string, but it's + * not quoted. + */ + emitter_type_title, +}; + +typedef struct emitter_col_s emitter_col_t; +struct emitter_col_s { + /* Filled in by the user. */ + emitter_justify_t justify; + int width; + emitter_type_t type; + union { + bool bool_val; + int int_val; + unsigned unsigned_val; + uint32_t uint32_val; + uint32_t uint32_t_val; + uint64_t uint64_val; + uint64_t uint64_t_val; + size_t size_val; + ssize_t ssize_val; + const char *str_val; + }; + + /* Filled in by initialization. */ + ql_elm(emitter_col_t) link; +}; + +typedef struct emitter_row_s emitter_row_t; +struct emitter_row_s { + ql_head(emitter_col_t) cols; +}; + +typedef struct emitter_s emitter_t; +struct emitter_s { + emitter_output_t output; + /* The output information. */ + void (*write_cb)(void *, const char *); + void *cbopaque; + int nesting_depth; + /* True if we've already emitted a value at the given depth. */ + bool item_at_depth; + /* True if we emitted a key and will emit corresponding value next. */ + bool emitted_key; +}; + +/* Internal convenience function. Write to the emitter the given string. */ +JEMALLOC_FORMAT_PRINTF(2, 3) +static inline void +emitter_printf(emitter_t *emitter, const char *format, ...) { + va_list ap; + + va_start(ap, format); + malloc_vcprintf(emitter->write_cb, emitter->cbopaque, format, ap); + va_end(ap); +} + +static inline const char * JEMALLOC_FORMAT_ARG(3) +emitter_gen_fmt(char *out_fmt, size_t out_size, const char *fmt_specifier, + emitter_justify_t justify, int width) { + size_t written; + fmt_specifier++; + if (justify == emitter_justify_none) { + written = malloc_snprintf(out_fmt, out_size, + "%%%s", fmt_specifier); + } else if (justify == emitter_justify_left) { + written = malloc_snprintf(out_fmt, out_size, + "%%-%d%s", width, fmt_specifier); + } else { + written = malloc_snprintf(out_fmt, out_size, + "%%%d%s", width, fmt_specifier); + } + /* Only happens in case of bad format string, which *we* choose. */ + assert(written < out_size); + return out_fmt; +} + +/* + * Internal. Emit the given value type in the relevant encoding (so that the + * bool true gets mapped to json "true", but the string "true" gets mapped to + * json "\"true\"", for instance. + * + * Width is ignored if justify is emitter_justify_none. + */ +static inline void +emitter_print_value(emitter_t *emitter, emitter_justify_t justify, int width, + emitter_type_t value_type, const void *value) { + size_t str_written; +#define BUF_SIZE 256 +#define FMT_SIZE 10 + /* + * We dynamically generate a format string to emit, to let us use the + * snprintf machinery. This is kinda hacky, but gets the job done + * quickly without having to think about the various snprintf edge + * cases. + */ + char fmt[FMT_SIZE]; + char buf[BUF_SIZE]; + +#define EMIT_SIMPLE(type, format) \ + emitter_printf(emitter, \ + emitter_gen_fmt(fmt, FMT_SIZE, format, justify, width), \ + *(const type *)value); + + switch (value_type) { + case emitter_type_bool: + emitter_printf(emitter, + emitter_gen_fmt(fmt, FMT_SIZE, "%s", justify, width), + *(const bool *)value ? "true" : "false"); + break; + case emitter_type_int: + EMIT_SIMPLE(int, "%d") + break; + case emitter_type_unsigned: + EMIT_SIMPLE(unsigned, "%u") + break; + case emitter_type_ssize: + EMIT_SIMPLE(ssize_t, "%zd") + break; + case emitter_type_size: + EMIT_SIMPLE(size_t, "%zu") + break; + case emitter_type_string: + str_written = malloc_snprintf(buf, BUF_SIZE, "\"%s\"", + *(const char *const *)value); + /* + * We control the strings we output; we shouldn't get anything + * anywhere near the fmt size. + */ + assert(str_written < BUF_SIZE); + emitter_printf(emitter, + emitter_gen_fmt(fmt, FMT_SIZE, "%s", justify, width), buf); + break; + case emitter_type_uint32: + EMIT_SIMPLE(uint32_t, "%" FMTu32) + break; + case emitter_type_uint64: + EMIT_SIMPLE(uint64_t, "%" FMTu64) + break; + case emitter_type_title: + EMIT_SIMPLE(char *const, "%s"); + break; + default: + unreachable(); + } +#undef BUF_SIZE +#undef FMT_SIZE +} + + +/* Internal functions. In json mode, tracks nesting state. */ +static inline void +emitter_nest_inc(emitter_t *emitter) { + emitter->nesting_depth++; + emitter->item_at_depth = false; +} + +static inline void +emitter_nest_dec(emitter_t *emitter) { + emitter->nesting_depth--; + emitter->item_at_depth = true; +} + +static inline void +emitter_indent(emitter_t *emitter) { + int amount = emitter->nesting_depth; + const char *indent_str; + if (emitter->output == emitter_output_json) { + indent_str = "\t"; + } else { + amount *= 2; + indent_str = " "; + } + for (int i = 0; i < amount; i++) { + emitter_printf(emitter, "%s", indent_str); + } +} + +static inline void +emitter_json_key_prefix(emitter_t *emitter) { + if (emitter->emitted_key) { + emitter->emitted_key = false; + return; + } + emitter_printf(emitter, "%s\n", emitter->item_at_depth ? "," : ""); + emitter_indent(emitter); +} + +/******************************************************************************/ +/* Public functions for emitter_t. */ + +static inline void +emitter_init(emitter_t *emitter, emitter_output_t emitter_output, + void (*write_cb)(void *, const char *), void *cbopaque) { + emitter->output = emitter_output; + emitter->write_cb = write_cb; + emitter->cbopaque = cbopaque; + emitter->item_at_depth = false; + emitter->emitted_key = false; + emitter->nesting_depth = 0; +} + +/******************************************************************************/ +/* JSON public API. */ + +/* + * Emits a key (e.g. as appears in an object). The next json entity emitted will + * be the corresponding value. + */ +static inline void +emitter_json_key(emitter_t *emitter, const char *json_key) { + if (emitter->output == emitter_output_json) { + emitter_json_key_prefix(emitter); + emitter_printf(emitter, "\"%s\": ", json_key); + emitter->emitted_key = true; + } +} + +static inline void +emitter_json_value(emitter_t *emitter, emitter_type_t value_type, + const void *value) { + if (emitter->output == emitter_output_json) { + emitter_json_key_prefix(emitter); + emitter_print_value(emitter, emitter_justify_none, -1, + value_type, value); + emitter->item_at_depth = true; + } +} + +/* Shorthand for calling emitter_json_key and then emitter_json_value. */ +static inline void +emitter_json_kv(emitter_t *emitter, const char *json_key, + emitter_type_t value_type, const void *value) { + emitter_json_key(emitter, json_key); + emitter_json_value(emitter, value_type, value); +} + +static inline void +emitter_json_array_begin(emitter_t *emitter) { + if (emitter->output == emitter_output_json) { + emitter_json_key_prefix(emitter); + emitter_printf(emitter, "["); + emitter_nest_inc(emitter); + } +} + +/* Shorthand for calling emitter_json_key and then emitter_json_array_begin. */ +static inline void +emitter_json_array_kv_begin(emitter_t *emitter, const char *json_key) { + emitter_json_key(emitter, json_key); + emitter_json_array_begin(emitter); +} + +static inline void +emitter_json_array_end(emitter_t *emitter) { + if (emitter->output == emitter_output_json) { + assert(emitter->nesting_depth > 0); + emitter_nest_dec(emitter); + emitter_printf(emitter, "\n"); + emitter_indent(emitter); + emitter_printf(emitter, "]"); + } +} + +static inline void +emitter_json_object_begin(emitter_t *emitter) { + if (emitter->output == emitter_output_json) { + emitter_json_key_prefix(emitter); + emitter_printf(emitter, "{"); + emitter_nest_inc(emitter); + } +} + +/* Shorthand for calling emitter_json_key and then emitter_json_object_begin. */ +static inline void +emitter_json_object_kv_begin(emitter_t *emitter, const char *json_key) { + emitter_json_key(emitter, json_key); + emitter_json_object_begin(emitter); +} + +static inline void +emitter_json_object_end(emitter_t *emitter) { + if (emitter->output == emitter_output_json) { + assert(emitter->nesting_depth > 0); + emitter_nest_dec(emitter); + emitter_printf(emitter, "\n"); + emitter_indent(emitter); + emitter_printf(emitter, "}"); + } +} + + +/******************************************************************************/ +/* Table public API. */ + +static inline void +emitter_table_dict_begin(emitter_t *emitter, const char *table_key) { + if (emitter->output == emitter_output_table) { + emitter_indent(emitter); + emitter_printf(emitter, "%s\n", table_key); + emitter_nest_inc(emitter); + } +} + +static inline void +emitter_table_dict_end(emitter_t *emitter) { + if (emitter->output == emitter_output_table) { + emitter_nest_dec(emitter); + } +} + +static inline void +emitter_table_kv_note(emitter_t *emitter, const char *table_key, + emitter_type_t value_type, const void *value, + const char *table_note_key, emitter_type_t table_note_value_type, + const void *table_note_value) { + if (emitter->output == emitter_output_table) { + emitter_indent(emitter); + emitter_printf(emitter, "%s: ", table_key); + emitter_print_value(emitter, emitter_justify_none, -1, + value_type, value); + if (table_note_key != NULL) { + emitter_printf(emitter, " (%s: ", table_note_key); + emitter_print_value(emitter, emitter_justify_none, -1, + table_note_value_type, table_note_value); + emitter_printf(emitter, ")"); + } + emitter_printf(emitter, "\n"); + } + emitter->item_at_depth = true; +} + +static inline void +emitter_table_kv(emitter_t *emitter, const char *table_key, + emitter_type_t value_type, const void *value) { + emitter_table_kv_note(emitter, table_key, value_type, value, NULL, + emitter_type_bool, NULL); +} + + +/* Write to the emitter the given string, but only in table mode. */ +JEMALLOC_FORMAT_PRINTF(2, 3) +static inline void +emitter_table_printf(emitter_t *emitter, const char *format, ...) { + if (emitter->output == emitter_output_table) { + va_list ap; + va_start(ap, format); + malloc_vcprintf(emitter->write_cb, emitter->cbopaque, format, ap); + va_end(ap); + } +} + +static inline void +emitter_table_row(emitter_t *emitter, emitter_row_t *row) { + if (emitter->output != emitter_output_table) { + return; + } + emitter_col_t *col; + ql_foreach(col, &row->cols, link) { + emitter_print_value(emitter, col->justify, col->width, + col->type, (const void *)&col->bool_val); + } + emitter_table_printf(emitter, "\n"); +} + +static inline void +emitter_row_init(emitter_row_t *row) { + ql_new(&row->cols); +} + +static inline void +emitter_col_init(emitter_col_t *col, emitter_row_t *row) { + ql_elm_new(col, link); + ql_tail_insert(&row->cols, col, link); +} + + +/******************************************************************************/ +/* + * Generalized public API. Emits using either JSON or table, according to + * settings in the emitter_t. */ + +/* + * Note emits a different kv pair as well, but only in table mode. Omits the + * note if table_note_key is NULL. + */ +static inline void +emitter_kv_note(emitter_t *emitter, const char *json_key, const char *table_key, + emitter_type_t value_type, const void *value, + const char *table_note_key, emitter_type_t table_note_value_type, + const void *table_note_value) { + if (emitter->output == emitter_output_json) { + emitter_json_key(emitter, json_key); + emitter_json_value(emitter, value_type, value); + } else { + emitter_table_kv_note(emitter, table_key, value_type, value, + table_note_key, table_note_value_type, table_note_value); + } + emitter->item_at_depth = true; +} + +static inline void +emitter_kv(emitter_t *emitter, const char *json_key, const char *table_key, + emitter_type_t value_type, const void *value) { + emitter_kv_note(emitter, json_key, table_key, value_type, value, NULL, + emitter_type_bool, NULL); +} + +static inline void +emitter_dict_begin(emitter_t *emitter, const char *json_key, + const char *table_header) { + if (emitter->output == emitter_output_json) { + emitter_json_key(emitter, json_key); + emitter_json_object_begin(emitter); + } else { + emitter_table_dict_begin(emitter, table_header); + } +} + +static inline void +emitter_dict_end(emitter_t *emitter) { + if (emitter->output == emitter_output_json) { + emitter_json_object_end(emitter); + } else { + emitter_table_dict_end(emitter); + } +} + +static inline void +emitter_begin(emitter_t *emitter) { + if (emitter->output == emitter_output_json) { + assert(emitter->nesting_depth == 0); + emitter_printf(emitter, "{"); + emitter_nest_inc(emitter); + } else { + /* + * This guarantees that we always call write_cb at least once. + * This is useful if some invariant is established by each call + * to write_cb, but doesn't hold initially: e.g., some buffer + * holds a null-terminated string. + */ + emitter_printf(emitter, "%s", ""); + } +} + +static inline void +emitter_end(emitter_t *emitter) { + if (emitter->output == emitter_output_json) { + assert(emitter->nesting_depth == 1); + emitter_nest_dec(emitter); + emitter_printf(emitter, "\n}\n"); + } +} + +#endif /* JEMALLOC_INTERNAL_EMITTER_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/extent_dss.h b/deps/jemalloc/include/jemalloc/internal/extent_dss.h new file mode 100644 index 000000000..e8f02ce2a --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/extent_dss.h @@ -0,0 +1,26 @@ +#ifndef JEMALLOC_INTERNAL_EXTENT_DSS_H +#define JEMALLOC_INTERNAL_EXTENT_DSS_H + +typedef enum { + dss_prec_disabled = 0, + dss_prec_primary = 1, + dss_prec_secondary = 2, + + dss_prec_limit = 3 +} dss_prec_t; +#define DSS_PREC_DEFAULT dss_prec_secondary +#define DSS_DEFAULT "secondary" + +extern const char *dss_prec_names[]; + +extern const char *opt_dss; + +dss_prec_t extent_dss_prec_get(void); +bool extent_dss_prec_set(dss_prec_t dss_prec); +void *extent_alloc_dss(tsdn_t *tsdn, arena_t *arena, void *new_addr, + size_t size, size_t alignment, bool *zero, bool *commit); +bool extent_in_dss(void *addr); +bool extent_dss_mergeable(void *addr_a, void *addr_b); +void extent_dss_boot(void); + +#endif /* JEMALLOC_INTERNAL_EXTENT_DSS_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/extent_externs.h b/deps/jemalloc/include/jemalloc/internal/extent_externs.h new file mode 100644 index 000000000..8aba57633 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/extent_externs.h @@ -0,0 +1,83 @@ +#ifndef JEMALLOC_INTERNAL_EXTENT_EXTERNS_H +#define JEMALLOC_INTERNAL_EXTENT_EXTERNS_H + +#include "jemalloc/internal/mutex.h" +#include "jemalloc/internal/mutex_pool.h" +#include "jemalloc/internal/ph.h" +#include "jemalloc/internal/rtree.h" + +extern size_t opt_lg_extent_max_active_fit; + +extern rtree_t extents_rtree; +extern const extent_hooks_t extent_hooks_default; +extern mutex_pool_t extent_mutex_pool; + +extent_t *extent_alloc(tsdn_t *tsdn, arena_t *arena); +void extent_dalloc(tsdn_t *tsdn, arena_t *arena, extent_t *extent); + +extent_hooks_t *extent_hooks_get(arena_t *arena); +extent_hooks_t *extent_hooks_set(tsd_t *tsd, arena_t *arena, + extent_hooks_t *extent_hooks); + +#ifdef JEMALLOC_JET +size_t extent_size_quantize_floor(size_t size); +size_t extent_size_quantize_ceil(size_t size); +#endif + +ph_proto(, extent_avail_, extent_tree_t, extent_t) +ph_proto(, extent_heap_, extent_heap_t, extent_t) + +bool extents_init(tsdn_t *tsdn, extents_t *extents, extent_state_t state, + bool delay_coalesce); +extent_state_t extents_state_get(const extents_t *extents); +size_t extents_npages_get(extents_t *extents); +/* Get the number of extents in the given page size index. */ +size_t extents_nextents_get(extents_t *extents, pszind_t ind); +/* Get the sum total bytes of the extents in the given page size index. */ +size_t extents_nbytes_get(extents_t *extents, pszind_t ind); +extent_t *extents_alloc(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extents_t *extents, void *new_addr, + size_t size, size_t pad, size_t alignment, bool slab, szind_t szind, + bool *zero, bool *commit); +void extents_dalloc(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extents_t *extents, extent_t *extent); +extent_t *extents_evict(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extents_t *extents, size_t npages_min); +void extents_prefork(tsdn_t *tsdn, extents_t *extents); +void extents_postfork_parent(tsdn_t *tsdn, extents_t *extents); +void extents_postfork_child(tsdn_t *tsdn, extents_t *extents); +extent_t *extent_alloc_wrapper(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, void *new_addr, size_t size, size_t pad, + size_t alignment, bool slab, szind_t szind, bool *zero, bool *commit); +void extent_dalloc_gap(tsdn_t *tsdn, arena_t *arena, extent_t *extent); +void extent_dalloc_wrapper(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent); +void extent_destroy_wrapper(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent); +bool extent_commit_wrapper(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent, size_t offset, + size_t length); +bool extent_decommit_wrapper(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent, size_t offset, + size_t length); +bool extent_purge_lazy_wrapper(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent, size_t offset, + size_t length); +bool extent_purge_forced_wrapper(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent, size_t offset, + size_t length); +extent_t *extent_split_wrapper(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent, size_t size_a, + szind_t szind_a, bool slab_a, size_t size_b, szind_t szind_b, bool slab_b); +bool extent_merge_wrapper(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *a, extent_t *b); + +bool extent_boot(void); + +void extent_util_stats_get(tsdn_t *tsdn, const void *ptr, + size_t *nfree, size_t *nregs, size_t *size); +void extent_util_stats_verbose_get(tsdn_t *tsdn, const void *ptr, + size_t *nfree, size_t *nregs, size_t *size, + size_t *bin_nfree, size_t *bin_nregs, void **slabcur_addr); + +#endif /* JEMALLOC_INTERNAL_EXTENT_EXTERNS_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/extent_inlines.h b/deps/jemalloc/include/jemalloc/internal/extent_inlines.h new file mode 100644 index 000000000..77fa4c4a2 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/extent_inlines.h @@ -0,0 +1,501 @@ +#ifndef JEMALLOC_INTERNAL_EXTENT_INLINES_H +#define JEMALLOC_INTERNAL_EXTENT_INLINES_H + +#include "jemalloc/internal/mutex.h" +#include "jemalloc/internal/mutex_pool.h" +#include "jemalloc/internal/pages.h" +#include "jemalloc/internal/prng.h" +#include "jemalloc/internal/ql.h" +#include "jemalloc/internal/sc.h" +#include "jemalloc/internal/sz.h" + +static inline void +extent_lock(tsdn_t *tsdn, extent_t *extent) { + assert(extent != NULL); + mutex_pool_lock(tsdn, &extent_mutex_pool, (uintptr_t)extent); +} + +static inline void +extent_unlock(tsdn_t *tsdn, extent_t *extent) { + assert(extent != NULL); + mutex_pool_unlock(tsdn, &extent_mutex_pool, (uintptr_t)extent); +} + +static inline void +extent_lock2(tsdn_t *tsdn, extent_t *extent1, extent_t *extent2) { + assert(extent1 != NULL && extent2 != NULL); + mutex_pool_lock2(tsdn, &extent_mutex_pool, (uintptr_t)extent1, + (uintptr_t)extent2); +} + +static inline void +extent_unlock2(tsdn_t *tsdn, extent_t *extent1, extent_t *extent2) { + assert(extent1 != NULL && extent2 != NULL); + mutex_pool_unlock2(tsdn, &extent_mutex_pool, (uintptr_t)extent1, + (uintptr_t)extent2); +} + +static inline unsigned +extent_arena_ind_get(const extent_t *extent) { + unsigned arena_ind = (unsigned)((extent->e_bits & + EXTENT_BITS_ARENA_MASK) >> EXTENT_BITS_ARENA_SHIFT); + assert(arena_ind < MALLOCX_ARENA_LIMIT); + + return arena_ind; +} + +static inline arena_t * +extent_arena_get(const extent_t *extent) { + unsigned arena_ind = extent_arena_ind_get(extent); + + return (arena_t *)atomic_load_p(&arenas[arena_ind], ATOMIC_ACQUIRE); +} + +static inline szind_t +extent_szind_get_maybe_invalid(const extent_t *extent) { + szind_t szind = (szind_t)((extent->e_bits & EXTENT_BITS_SZIND_MASK) >> + EXTENT_BITS_SZIND_SHIFT); + assert(szind <= SC_NSIZES); + return szind; +} + +static inline szind_t +extent_szind_get(const extent_t *extent) { + szind_t szind = extent_szind_get_maybe_invalid(extent); + assert(szind < SC_NSIZES); /* Never call when "invalid". */ + return szind; +} + +static inline size_t +extent_usize_get(const extent_t *extent) { + return sz_index2size(extent_szind_get(extent)); +} + +static inline unsigned +extent_binshard_get(const extent_t *extent) { + unsigned binshard = (unsigned)((extent->e_bits & + EXTENT_BITS_BINSHARD_MASK) >> EXTENT_BITS_BINSHARD_SHIFT); + assert(binshard < bin_infos[extent_szind_get(extent)].n_shards); + return binshard; +} + +static inline size_t +extent_sn_get(const extent_t *extent) { + return (size_t)((extent->e_bits & EXTENT_BITS_SN_MASK) >> + EXTENT_BITS_SN_SHIFT); +} + +static inline extent_state_t +extent_state_get(const extent_t *extent) { + return (extent_state_t)((extent->e_bits & EXTENT_BITS_STATE_MASK) >> + EXTENT_BITS_STATE_SHIFT); +} + +static inline bool +extent_zeroed_get(const extent_t *extent) { + return (bool)((extent->e_bits & EXTENT_BITS_ZEROED_MASK) >> + EXTENT_BITS_ZEROED_SHIFT); +} + +static inline bool +extent_committed_get(const extent_t *extent) { + return (bool)((extent->e_bits & EXTENT_BITS_COMMITTED_MASK) >> + EXTENT_BITS_COMMITTED_SHIFT); +} + +static inline bool +extent_dumpable_get(const extent_t *extent) { + return (bool)((extent->e_bits & EXTENT_BITS_DUMPABLE_MASK) >> + EXTENT_BITS_DUMPABLE_SHIFT); +} + +static inline bool +extent_slab_get(const extent_t *extent) { + return (bool)((extent->e_bits & EXTENT_BITS_SLAB_MASK) >> + EXTENT_BITS_SLAB_SHIFT); +} + +static inline unsigned +extent_nfree_get(const extent_t *extent) { + assert(extent_slab_get(extent)); + return (unsigned)((extent->e_bits & EXTENT_BITS_NFREE_MASK) >> + EXTENT_BITS_NFREE_SHIFT); +} + +static inline void * +extent_base_get(const extent_t *extent) { + assert(extent->e_addr == PAGE_ADDR2BASE(extent->e_addr) || + !extent_slab_get(extent)); + return PAGE_ADDR2BASE(extent->e_addr); +} + +static inline void * +extent_addr_get(const extent_t *extent) { + assert(extent->e_addr == PAGE_ADDR2BASE(extent->e_addr) || + !extent_slab_get(extent)); + return extent->e_addr; +} + +static inline size_t +extent_size_get(const extent_t *extent) { + return (extent->e_size_esn & EXTENT_SIZE_MASK); +} + +static inline size_t +extent_esn_get(const extent_t *extent) { + return (extent->e_size_esn & EXTENT_ESN_MASK); +} + +static inline size_t +extent_bsize_get(const extent_t *extent) { + return extent->e_bsize; +} + +static inline void * +extent_before_get(const extent_t *extent) { + return (void *)((uintptr_t)extent_base_get(extent) - PAGE); +} + +static inline void * +extent_last_get(const extent_t *extent) { + return (void *)((uintptr_t)extent_base_get(extent) + + extent_size_get(extent) - PAGE); +} + +static inline void * +extent_past_get(const extent_t *extent) { + return (void *)((uintptr_t)extent_base_get(extent) + + extent_size_get(extent)); +} + +static inline arena_slab_data_t * +extent_slab_data_get(extent_t *extent) { + assert(extent_slab_get(extent)); + return &extent->e_slab_data; +} + +static inline const arena_slab_data_t * +extent_slab_data_get_const(const extent_t *extent) { + assert(extent_slab_get(extent)); + return &extent->e_slab_data; +} + +static inline prof_tctx_t * +extent_prof_tctx_get(const extent_t *extent) { + return (prof_tctx_t *)atomic_load_p(&extent->e_prof_tctx, + ATOMIC_ACQUIRE); +} + +static inline nstime_t +extent_prof_alloc_time_get(const extent_t *extent) { + return extent->e_alloc_time; +} + +static inline void +extent_arena_set(extent_t *extent, arena_t *arena) { + unsigned arena_ind = (arena != NULL) ? arena_ind_get(arena) : ((1U << + MALLOCX_ARENA_BITS) - 1); + extent->e_bits = (extent->e_bits & ~EXTENT_BITS_ARENA_MASK) | + ((uint64_t)arena_ind << EXTENT_BITS_ARENA_SHIFT); +} + +static inline void +extent_binshard_set(extent_t *extent, unsigned binshard) { + /* The assertion assumes szind is set already. */ + assert(binshard < bin_infos[extent_szind_get(extent)].n_shards); + extent->e_bits = (extent->e_bits & ~EXTENT_BITS_BINSHARD_MASK) | + ((uint64_t)binshard << EXTENT_BITS_BINSHARD_SHIFT); +} + +static inline void +extent_addr_set(extent_t *extent, void *addr) { + extent->e_addr = addr; +} + +static inline void +extent_addr_randomize(tsdn_t *tsdn, extent_t *extent, size_t alignment) { + assert(extent_base_get(extent) == extent_addr_get(extent)); + + if (alignment < PAGE) { + unsigned lg_range = LG_PAGE - + lg_floor(CACHELINE_CEILING(alignment)); + size_t r; + if (!tsdn_null(tsdn)) { + tsd_t *tsd = tsdn_tsd(tsdn); + r = (size_t)prng_lg_range_u64( + tsd_offset_statep_get(tsd), lg_range); + } else { + r = prng_lg_range_zu( + &extent_arena_get(extent)->offset_state, + lg_range, true); + } + uintptr_t random_offset = ((uintptr_t)r) << (LG_PAGE - + lg_range); + extent->e_addr = (void *)((uintptr_t)extent->e_addr + + random_offset); + assert(ALIGNMENT_ADDR2BASE(extent->e_addr, alignment) == + extent->e_addr); + } +} + +static inline void +extent_size_set(extent_t *extent, size_t size) { + assert((size & ~EXTENT_SIZE_MASK) == 0); + extent->e_size_esn = size | (extent->e_size_esn & ~EXTENT_SIZE_MASK); +} + +static inline void +extent_esn_set(extent_t *extent, size_t esn) { + extent->e_size_esn = (extent->e_size_esn & ~EXTENT_ESN_MASK) | (esn & + EXTENT_ESN_MASK); +} + +static inline void +extent_bsize_set(extent_t *extent, size_t bsize) { + extent->e_bsize = bsize; +} + +static inline void +extent_szind_set(extent_t *extent, szind_t szind) { + assert(szind <= SC_NSIZES); /* SC_NSIZES means "invalid". */ + extent->e_bits = (extent->e_bits & ~EXTENT_BITS_SZIND_MASK) | + ((uint64_t)szind << EXTENT_BITS_SZIND_SHIFT); +} + +static inline void +extent_nfree_set(extent_t *extent, unsigned nfree) { + assert(extent_slab_get(extent)); + extent->e_bits = (extent->e_bits & ~EXTENT_BITS_NFREE_MASK) | + ((uint64_t)nfree << EXTENT_BITS_NFREE_SHIFT); +} + +static inline void +extent_nfree_binshard_set(extent_t *extent, unsigned nfree, unsigned binshard) { + /* The assertion assumes szind is set already. */ + assert(binshard < bin_infos[extent_szind_get(extent)].n_shards); + extent->e_bits = (extent->e_bits & + (~EXTENT_BITS_NFREE_MASK & ~EXTENT_BITS_BINSHARD_MASK)) | + ((uint64_t)binshard << EXTENT_BITS_BINSHARD_SHIFT) | + ((uint64_t)nfree << EXTENT_BITS_NFREE_SHIFT); +} + +static inline void +extent_nfree_inc(extent_t *extent) { + assert(extent_slab_get(extent)); + extent->e_bits += ((uint64_t)1U << EXTENT_BITS_NFREE_SHIFT); +} + +static inline void +extent_nfree_dec(extent_t *extent) { + assert(extent_slab_get(extent)); + extent->e_bits -= ((uint64_t)1U << EXTENT_BITS_NFREE_SHIFT); +} + +static inline void +extent_nfree_sub(extent_t *extent, uint64_t n) { + assert(extent_slab_get(extent)); + extent->e_bits -= (n << EXTENT_BITS_NFREE_SHIFT); +} + +static inline void +extent_sn_set(extent_t *extent, size_t sn) { + extent->e_bits = (extent->e_bits & ~EXTENT_BITS_SN_MASK) | + ((uint64_t)sn << EXTENT_BITS_SN_SHIFT); +} + +static inline void +extent_state_set(extent_t *extent, extent_state_t state) { + extent->e_bits = (extent->e_bits & ~EXTENT_BITS_STATE_MASK) | + ((uint64_t)state << EXTENT_BITS_STATE_SHIFT); +} + +static inline void +extent_zeroed_set(extent_t *extent, bool zeroed) { + extent->e_bits = (extent->e_bits & ~EXTENT_BITS_ZEROED_MASK) | + ((uint64_t)zeroed << EXTENT_BITS_ZEROED_SHIFT); +} + +static inline void +extent_committed_set(extent_t *extent, bool committed) { + extent->e_bits = (extent->e_bits & ~EXTENT_BITS_COMMITTED_MASK) | + ((uint64_t)committed << EXTENT_BITS_COMMITTED_SHIFT); +} + +static inline void +extent_dumpable_set(extent_t *extent, bool dumpable) { + extent->e_bits = (extent->e_bits & ~EXTENT_BITS_DUMPABLE_MASK) | + ((uint64_t)dumpable << EXTENT_BITS_DUMPABLE_SHIFT); +} + +static inline void +extent_slab_set(extent_t *extent, bool slab) { + extent->e_bits = (extent->e_bits & ~EXTENT_BITS_SLAB_MASK) | + ((uint64_t)slab << EXTENT_BITS_SLAB_SHIFT); +} + +static inline void +extent_prof_tctx_set(extent_t *extent, prof_tctx_t *tctx) { + atomic_store_p(&extent->e_prof_tctx, tctx, ATOMIC_RELEASE); +} + +static inline void +extent_prof_alloc_time_set(extent_t *extent, nstime_t t) { + nstime_copy(&extent->e_alloc_time, &t); +} + +static inline bool +extent_is_head_get(extent_t *extent) { + if (maps_coalesce) { + not_reached(); + } + + return (bool)((extent->e_bits & EXTENT_BITS_IS_HEAD_MASK) >> + EXTENT_BITS_IS_HEAD_SHIFT); +} + +static inline void +extent_is_head_set(extent_t *extent, bool is_head) { + if (maps_coalesce) { + not_reached(); + } + + extent->e_bits = (extent->e_bits & ~EXTENT_BITS_IS_HEAD_MASK) | + ((uint64_t)is_head << EXTENT_BITS_IS_HEAD_SHIFT); +} + +static inline void +extent_init(extent_t *extent, arena_t *arena, void *addr, size_t size, + bool slab, szind_t szind, size_t sn, extent_state_t state, bool zeroed, + bool committed, bool dumpable, extent_head_state_t is_head) { + assert(addr == PAGE_ADDR2BASE(addr) || !slab); + + extent_arena_set(extent, arena); + extent_addr_set(extent, addr); + extent_size_set(extent, size); + extent_slab_set(extent, slab); + extent_szind_set(extent, szind); + extent_sn_set(extent, sn); + extent_state_set(extent, state); + extent_zeroed_set(extent, zeroed); + extent_committed_set(extent, committed); + extent_dumpable_set(extent, dumpable); + ql_elm_new(extent, ql_link); + if (!maps_coalesce) { + extent_is_head_set(extent, (is_head == EXTENT_IS_HEAD) ? true : + false); + } + if (config_prof) { + extent_prof_tctx_set(extent, NULL); + } +} + +static inline void +extent_binit(extent_t *extent, void *addr, size_t bsize, size_t sn) { + extent_arena_set(extent, NULL); + extent_addr_set(extent, addr); + extent_bsize_set(extent, bsize); + extent_slab_set(extent, false); + extent_szind_set(extent, SC_NSIZES); + extent_sn_set(extent, sn); + extent_state_set(extent, extent_state_active); + extent_zeroed_set(extent, true); + extent_committed_set(extent, true); + extent_dumpable_set(extent, true); +} + +static inline void +extent_list_init(extent_list_t *list) { + ql_new(list); +} + +static inline extent_t * +extent_list_first(const extent_list_t *list) { + return ql_first(list); +} + +static inline extent_t * +extent_list_last(const extent_list_t *list) { + return ql_last(list, ql_link); +} + +static inline void +extent_list_append(extent_list_t *list, extent_t *extent) { + ql_tail_insert(list, extent, ql_link); +} + +static inline void +extent_list_prepend(extent_list_t *list, extent_t *extent) { + ql_head_insert(list, extent, ql_link); +} + +static inline void +extent_list_replace(extent_list_t *list, extent_t *to_remove, + extent_t *to_insert) { + ql_after_insert(to_remove, to_insert, ql_link); + ql_remove(list, to_remove, ql_link); +} + +static inline void +extent_list_remove(extent_list_t *list, extent_t *extent) { + ql_remove(list, extent, ql_link); +} + +static inline int +extent_sn_comp(const extent_t *a, const extent_t *b) { + size_t a_sn = extent_sn_get(a); + size_t b_sn = extent_sn_get(b); + + return (a_sn > b_sn) - (a_sn < b_sn); +} + +static inline int +extent_esn_comp(const extent_t *a, const extent_t *b) { + size_t a_esn = extent_esn_get(a); + size_t b_esn = extent_esn_get(b); + + return (a_esn > b_esn) - (a_esn < b_esn); +} + +static inline int +extent_ad_comp(const extent_t *a, const extent_t *b) { + uintptr_t a_addr = (uintptr_t)extent_addr_get(a); + uintptr_t b_addr = (uintptr_t)extent_addr_get(b); + + return (a_addr > b_addr) - (a_addr < b_addr); +} + +static inline int +extent_ead_comp(const extent_t *a, const extent_t *b) { + uintptr_t a_eaddr = (uintptr_t)a; + uintptr_t b_eaddr = (uintptr_t)b; + + return (a_eaddr > b_eaddr) - (a_eaddr < b_eaddr); +} + +static inline int +extent_snad_comp(const extent_t *a, const extent_t *b) { + int ret; + + ret = extent_sn_comp(a, b); + if (ret != 0) { + return ret; + } + + ret = extent_ad_comp(a, b); + return ret; +} + +static inline int +extent_esnead_comp(const extent_t *a, const extent_t *b) { + int ret; + + ret = extent_esn_comp(a, b); + if (ret != 0) { + return ret; + } + + ret = extent_ead_comp(a, b); + return ret; +} + +#endif /* JEMALLOC_INTERNAL_EXTENT_INLINES_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/extent_mmap.h b/deps/jemalloc/include/jemalloc/internal/extent_mmap.h new file mode 100644 index 000000000..55f17ee48 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/extent_mmap.h @@ -0,0 +1,10 @@ +#ifndef JEMALLOC_INTERNAL_EXTENT_MMAP_EXTERNS_H +#define JEMALLOC_INTERNAL_EXTENT_MMAP_EXTERNS_H + +extern bool opt_retain; + +void *extent_alloc_mmap(void *new_addr, size_t size, size_t alignment, + bool *zero, bool *commit); +bool extent_dalloc_mmap(void *addr, size_t size); + +#endif /* JEMALLOC_INTERNAL_EXTENT_MMAP_EXTERNS_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/extent_structs.h b/deps/jemalloc/include/jemalloc/internal/extent_structs.h new file mode 100644 index 000000000..767cd8930 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/extent_structs.h @@ -0,0 +1,256 @@ +#ifndef JEMALLOC_INTERNAL_EXTENT_STRUCTS_H +#define JEMALLOC_INTERNAL_EXTENT_STRUCTS_H + +#include "jemalloc/internal/atomic.h" +#include "jemalloc/internal/bit_util.h" +#include "jemalloc/internal/bitmap.h" +#include "jemalloc/internal/mutex.h" +#include "jemalloc/internal/ql.h" +#include "jemalloc/internal/ph.h" +#include "jemalloc/internal/sc.h" + +typedef enum { + extent_state_active = 0, + extent_state_dirty = 1, + extent_state_muzzy = 2, + extent_state_retained = 3 +} extent_state_t; + +/* Extent (span of pages). Use accessor functions for e_* fields. */ +struct extent_s { + /* + * Bitfield containing several fields: + * + * a: arena_ind + * b: slab + * c: committed + * d: dumpable + * z: zeroed + * t: state + * i: szind + * f: nfree + * s: bin_shard + * n: sn + * + * nnnnnnnn ... nnnnnnss ssssffff ffffffii iiiiiitt zdcbaaaa aaaaaaaa + * + * arena_ind: Arena from which this extent came, or all 1 bits if + * unassociated. + * + * slab: The slab flag indicates whether the extent is used for a slab + * of small regions. This helps differentiate small size classes, + * and it indicates whether interior pointers can be looked up via + * iealloc(). + * + * committed: The committed flag indicates whether physical memory is + * committed to the extent, whether explicitly or implicitly + * as on a system that overcommits and satisfies physical + * memory needs on demand via soft page faults. + * + * dumpable: The dumpable flag indicates whether or not we've set the + * memory in question to be dumpable. Note that this + * interacts somewhat subtly with user-specified extent hooks, + * since we don't know if *they* are fiddling with + * dumpability (in which case, we don't want to undo whatever + * they're doing). To deal with this scenario, we: + * - Make dumpable false only for memory allocated with the + * default hooks. + * - Only allow memory to go from non-dumpable to dumpable, + * and only once. + * - Never make the OS call to allow dumping when the + * dumpable bit is already set. + * These three constraints mean that we will never + * accidentally dump user memory that the user meant to set + * nondumpable with their extent hooks. + * + * + * zeroed: The zeroed flag is used by extent recycling code to track + * whether memory is zero-filled. + * + * state: The state flag is an extent_state_t. + * + * szind: The szind flag indicates usable size class index for + * allocations residing in this extent, regardless of whether the + * extent is a slab. Extent size and usable size often differ + * even for non-slabs, either due to sz_large_pad or promotion of + * sampled small regions. + * + * nfree: Number of free regions in slab. + * + * bin_shard: the shard of the bin from which this extent came. + * + * sn: Serial number (potentially non-unique). + * + * Serial numbers may wrap around if !opt_retain, but as long as + * comparison functions fall back on address comparison for equal + * serial numbers, stable (if imperfect) ordering is maintained. + * + * Serial numbers may not be unique even in the absence of + * wrap-around, e.g. when splitting an extent and assigning the same + * serial number to both resulting adjacent extents. + */ + uint64_t e_bits; +#define MASK(CURRENT_FIELD_WIDTH, CURRENT_FIELD_SHIFT) ((((((uint64_t)0x1U) << (CURRENT_FIELD_WIDTH)) - 1)) << (CURRENT_FIELD_SHIFT)) + +#define EXTENT_BITS_ARENA_WIDTH MALLOCX_ARENA_BITS +#define EXTENT_BITS_ARENA_SHIFT 0 +#define EXTENT_BITS_ARENA_MASK MASK(EXTENT_BITS_ARENA_WIDTH, EXTENT_BITS_ARENA_SHIFT) + +#define EXTENT_BITS_SLAB_WIDTH 1 +#define EXTENT_BITS_SLAB_SHIFT (EXTENT_BITS_ARENA_WIDTH + EXTENT_BITS_ARENA_SHIFT) +#define EXTENT_BITS_SLAB_MASK MASK(EXTENT_BITS_SLAB_WIDTH, EXTENT_BITS_SLAB_SHIFT) + +#define EXTENT_BITS_COMMITTED_WIDTH 1 +#define EXTENT_BITS_COMMITTED_SHIFT (EXTENT_BITS_SLAB_WIDTH + EXTENT_BITS_SLAB_SHIFT) +#define EXTENT_BITS_COMMITTED_MASK MASK(EXTENT_BITS_COMMITTED_WIDTH, EXTENT_BITS_COMMITTED_SHIFT) + +#define EXTENT_BITS_DUMPABLE_WIDTH 1 +#define EXTENT_BITS_DUMPABLE_SHIFT (EXTENT_BITS_COMMITTED_WIDTH + EXTENT_BITS_COMMITTED_SHIFT) +#define EXTENT_BITS_DUMPABLE_MASK MASK(EXTENT_BITS_DUMPABLE_WIDTH, EXTENT_BITS_DUMPABLE_SHIFT) + +#define EXTENT_BITS_ZEROED_WIDTH 1 +#define EXTENT_BITS_ZEROED_SHIFT (EXTENT_BITS_DUMPABLE_WIDTH + EXTENT_BITS_DUMPABLE_SHIFT) +#define EXTENT_BITS_ZEROED_MASK MASK(EXTENT_BITS_ZEROED_WIDTH, EXTENT_BITS_ZEROED_SHIFT) + +#define EXTENT_BITS_STATE_WIDTH 2 +#define EXTENT_BITS_STATE_SHIFT (EXTENT_BITS_ZEROED_WIDTH + EXTENT_BITS_ZEROED_SHIFT) +#define EXTENT_BITS_STATE_MASK MASK(EXTENT_BITS_STATE_WIDTH, EXTENT_BITS_STATE_SHIFT) + +#define EXTENT_BITS_SZIND_WIDTH LG_CEIL(SC_NSIZES) +#define EXTENT_BITS_SZIND_SHIFT (EXTENT_BITS_STATE_WIDTH + EXTENT_BITS_STATE_SHIFT) +#define EXTENT_BITS_SZIND_MASK MASK(EXTENT_BITS_SZIND_WIDTH, EXTENT_BITS_SZIND_SHIFT) + +#define EXTENT_BITS_NFREE_WIDTH (LG_SLAB_MAXREGS + 1) +#define EXTENT_BITS_NFREE_SHIFT (EXTENT_BITS_SZIND_WIDTH + EXTENT_BITS_SZIND_SHIFT) +#define EXTENT_BITS_NFREE_MASK MASK(EXTENT_BITS_NFREE_WIDTH, EXTENT_BITS_NFREE_SHIFT) + +#define EXTENT_BITS_BINSHARD_WIDTH 6 +#define EXTENT_BITS_BINSHARD_SHIFT (EXTENT_BITS_NFREE_WIDTH + EXTENT_BITS_NFREE_SHIFT) +#define EXTENT_BITS_BINSHARD_MASK MASK(EXTENT_BITS_BINSHARD_WIDTH, EXTENT_BITS_BINSHARD_SHIFT) + +#define EXTENT_BITS_IS_HEAD_WIDTH 1 +#define EXTENT_BITS_IS_HEAD_SHIFT (EXTENT_BITS_BINSHARD_WIDTH + EXTENT_BITS_BINSHARD_SHIFT) +#define EXTENT_BITS_IS_HEAD_MASK MASK(EXTENT_BITS_IS_HEAD_WIDTH, EXTENT_BITS_IS_HEAD_SHIFT) + +#define EXTENT_BITS_SN_SHIFT (EXTENT_BITS_IS_HEAD_WIDTH + EXTENT_BITS_IS_HEAD_SHIFT) +#define EXTENT_BITS_SN_MASK (UINT64_MAX << EXTENT_BITS_SN_SHIFT) + + /* Pointer to the extent that this structure is responsible for. */ + void *e_addr; + + union { + /* + * Extent size and serial number associated with the extent + * structure (different than the serial number for the extent at + * e_addr). + * + * ssssssss [...] ssssssss ssssnnnn nnnnnnnn + */ + size_t e_size_esn; + #define EXTENT_SIZE_MASK ((size_t)~(PAGE-1)) + #define EXTENT_ESN_MASK ((size_t)PAGE-1) + /* Base extent size, which may not be a multiple of PAGE. */ + size_t e_bsize; + }; + + /* + * List linkage, used by a variety of lists: + * - bin_t's slabs_full + * - extents_t's LRU + * - stashed dirty extents + * - arena's large allocations + */ + ql_elm(extent_t) ql_link; + + /* + * Linkage for per size class sn/address-ordered heaps, and + * for extent_avail + */ + phn(extent_t) ph_link; + + union { + /* Small region slab metadata. */ + arena_slab_data_t e_slab_data; + + /* Profiling data, used for large objects. */ + struct { + /* Time when this was allocated. */ + nstime_t e_alloc_time; + /* Points to a prof_tctx_t. */ + atomic_p_t e_prof_tctx; + }; + }; +}; +typedef ql_head(extent_t) extent_list_t; +typedef ph(extent_t) extent_tree_t; +typedef ph(extent_t) extent_heap_t; + +/* Quantized collection of extents, with built-in LRU queue. */ +struct extents_s { + malloc_mutex_t mtx; + + /* + * Quantized per size class heaps of extents. + * + * Synchronization: mtx. + */ + extent_heap_t heaps[SC_NPSIZES + 1]; + atomic_zu_t nextents[SC_NPSIZES + 1]; + atomic_zu_t nbytes[SC_NPSIZES + 1]; + + /* + * Bitmap for which set bits correspond to non-empty heaps. + * + * Synchronization: mtx. + */ + bitmap_t bitmap[BITMAP_GROUPS(SC_NPSIZES + 1)]; + + /* + * LRU of all extents in heaps. + * + * Synchronization: mtx. + */ + extent_list_t lru; + + /* + * Page sum for all extents in heaps. + * + * The synchronization here is a little tricky. Modifications to npages + * must hold mtx, but reads need not (though, a reader who sees npages + * without holding the mutex can't assume anything about the rest of the + * state of the extents_t). + */ + atomic_zu_t npages; + + /* All stored extents must be in the same state. */ + extent_state_t state; + + /* + * If true, delay coalescing until eviction; otherwise coalesce during + * deallocation. + */ + bool delay_coalesce; +}; + +/* + * The following two structs are for experimental purposes. See + * experimental_utilization_query_ctl and + * experimental_utilization_batch_query_ctl in src/ctl.c. + */ + +struct extent_util_stats_s { + size_t nfree; + size_t nregs; + size_t size; +}; + +struct extent_util_stats_verbose_s { + void *slabcur_addr; + size_t nfree; + size_t nregs; + size_t size; + size_t bin_nfree; + size_t bin_nregs; +}; + +#endif /* JEMALLOC_INTERNAL_EXTENT_STRUCTS_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/extent_types.h b/deps/jemalloc/include/jemalloc/internal/extent_types.h new file mode 100644 index 000000000..96925cf95 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/extent_types.h @@ -0,0 +1,23 @@ +#ifndef JEMALLOC_INTERNAL_EXTENT_TYPES_H +#define JEMALLOC_INTERNAL_EXTENT_TYPES_H + +typedef struct extent_s extent_t; +typedef struct extents_s extents_t; + +typedef struct extent_util_stats_s extent_util_stats_t; +typedef struct extent_util_stats_verbose_s extent_util_stats_verbose_t; + +#define EXTENT_HOOKS_INITIALIZER NULL + +/* + * When reuse (and split) an active extent, (1U << opt_lg_extent_max_active_fit) + * is the max ratio between the size of the active extent and the new extent. + */ +#define LG_EXTENT_MAX_ACTIVE_FIT_DEFAULT 6 + +typedef enum { + EXTENT_NOT_HEAD, + EXTENT_IS_HEAD /* Only relevant for Windows && opt.retain. */ +} extent_head_state_t; + +#endif /* JEMALLOC_INTERNAL_EXTENT_TYPES_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/hash.h b/deps/jemalloc/include/jemalloc/internal/hash.h new file mode 100644 index 000000000..0270034e8 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/hash.h @@ -0,0 +1,319 @@ +#ifndef JEMALLOC_INTERNAL_HASH_H +#define JEMALLOC_INTERNAL_HASH_H + +#include "jemalloc/internal/assert.h" + +/* + * The following hash function is based on MurmurHash3, placed into the public + * domain by Austin Appleby. See https://github.com/aappleby/smhasher for + * details. + */ + +/******************************************************************************/ +/* Internal implementation. */ +static inline uint32_t +hash_rotl_32(uint32_t x, int8_t r) { + return ((x << r) | (x >> (32 - r))); +} + +static inline uint64_t +hash_rotl_64(uint64_t x, int8_t r) { + return ((x << r) | (x >> (64 - r))); +} + +static inline uint32_t +hash_get_block_32(const uint32_t *p, int i) { + /* Handle unaligned read. */ + if (unlikely((uintptr_t)p & (sizeof(uint32_t)-1)) != 0) { + uint32_t ret; + + memcpy(&ret, (uint8_t *)(p + i), sizeof(uint32_t)); + return ret; + } + + return p[i]; +} + +static inline uint64_t +hash_get_block_64(const uint64_t *p, int i) { + /* Handle unaligned read. */ + if (unlikely((uintptr_t)p & (sizeof(uint64_t)-1)) != 0) { + uint64_t ret; + + memcpy(&ret, (uint8_t *)(p + i), sizeof(uint64_t)); + return ret; + } + + return p[i]; +} + +static inline uint32_t +hash_fmix_32(uint32_t h) { + h ^= h >> 16; + h *= 0x85ebca6b; + h ^= h >> 13; + h *= 0xc2b2ae35; + h ^= h >> 16; + + return h; +} + +static inline uint64_t +hash_fmix_64(uint64_t k) { + k ^= k >> 33; + k *= KQU(0xff51afd7ed558ccd); + k ^= k >> 33; + k *= KQU(0xc4ceb9fe1a85ec53); + k ^= k >> 33; + + return k; +} + +static inline uint32_t +hash_x86_32(const void *key, int len, uint32_t seed) { + const uint8_t *data = (const uint8_t *) key; + const int nblocks = len / 4; + + uint32_t h1 = seed; + + const uint32_t c1 = 0xcc9e2d51; + const uint32_t c2 = 0x1b873593; + + /* body */ + { + const uint32_t *blocks = (const uint32_t *) (data + nblocks*4); + int i; + + for (i = -nblocks; i; i++) { + uint32_t k1 = hash_get_block_32(blocks, i); + + k1 *= c1; + k1 = hash_rotl_32(k1, 15); + k1 *= c2; + + h1 ^= k1; + h1 = hash_rotl_32(h1, 13); + h1 = h1*5 + 0xe6546b64; + } + } + + /* tail */ + { + const uint8_t *tail = (const uint8_t *) (data + nblocks*4); + + uint32_t k1 = 0; + + switch (len & 3) { + case 3: k1 ^= tail[2] << 16; JEMALLOC_FALLTHROUGH + case 2: k1 ^= tail[1] << 8; JEMALLOC_FALLTHROUGH + case 1: k1 ^= tail[0]; k1 *= c1; k1 = hash_rotl_32(k1, 15); + k1 *= c2; h1 ^= k1; + } + } + + /* finalization */ + h1 ^= len; + + h1 = hash_fmix_32(h1); + + return h1; +} + +static inline void +hash_x86_128(const void *key, const int len, uint32_t seed, + uint64_t r_out[2]) { + const uint8_t * data = (const uint8_t *) key; + const int nblocks = len / 16; + + uint32_t h1 = seed; + uint32_t h2 = seed; + uint32_t h3 = seed; + uint32_t h4 = seed; + + const uint32_t c1 = 0x239b961b; + const uint32_t c2 = 0xab0e9789; + const uint32_t c3 = 0x38b34ae5; + const uint32_t c4 = 0xa1e38b93; + + /* body */ + { + const uint32_t *blocks = (const uint32_t *) (data + nblocks*16); + int i; + + for (i = -nblocks; i; i++) { + uint32_t k1 = hash_get_block_32(blocks, i*4 + 0); + uint32_t k2 = hash_get_block_32(blocks, i*4 + 1); + uint32_t k3 = hash_get_block_32(blocks, i*4 + 2); + uint32_t k4 = hash_get_block_32(blocks, i*4 + 3); + + k1 *= c1; k1 = hash_rotl_32(k1, 15); k1 *= c2; h1 ^= k1; + + h1 = hash_rotl_32(h1, 19); h1 += h2; + h1 = h1*5 + 0x561ccd1b; + + k2 *= c2; k2 = hash_rotl_32(k2, 16); k2 *= c3; h2 ^= k2; + + h2 = hash_rotl_32(h2, 17); h2 += h3; + h2 = h2*5 + 0x0bcaa747; + + k3 *= c3; k3 = hash_rotl_32(k3, 17); k3 *= c4; h3 ^= k3; + + h3 = hash_rotl_32(h3, 15); h3 += h4; + h3 = h3*5 + 0x96cd1c35; + + k4 *= c4; k4 = hash_rotl_32(k4, 18); k4 *= c1; h4 ^= k4; + + h4 = hash_rotl_32(h4, 13); h4 += h1; + h4 = h4*5 + 0x32ac3b17; + } + } + + /* tail */ + { + const uint8_t *tail = (const uint8_t *) (data + nblocks*16); + uint32_t k1 = 0; + uint32_t k2 = 0; + uint32_t k3 = 0; + uint32_t k4 = 0; + + switch (len & 15) { + case 15: k4 ^= tail[14] << 16; JEMALLOC_FALLTHROUGH + case 14: k4 ^= tail[13] << 8; JEMALLOC_FALLTHROUGH + case 13: k4 ^= tail[12] << 0; + k4 *= c4; k4 = hash_rotl_32(k4, 18); k4 *= c1; h4 ^= k4; + JEMALLOC_FALLTHROUGH + case 12: k3 ^= tail[11] << 24; JEMALLOC_FALLTHROUGH + case 11: k3 ^= tail[10] << 16; JEMALLOC_FALLTHROUGH + case 10: k3 ^= tail[ 9] << 8; JEMALLOC_FALLTHROUGH + case 9: k3 ^= tail[ 8] << 0; + k3 *= c3; k3 = hash_rotl_32(k3, 17); k3 *= c4; h3 ^= k3; + JEMALLOC_FALLTHROUGH + case 8: k2 ^= tail[ 7] << 24; JEMALLOC_FALLTHROUGH + case 7: k2 ^= tail[ 6] << 16; JEMALLOC_FALLTHROUGH + case 6: k2 ^= tail[ 5] << 8; JEMALLOC_FALLTHROUGH + case 5: k2 ^= tail[ 4] << 0; + k2 *= c2; k2 = hash_rotl_32(k2, 16); k2 *= c3; h2 ^= k2; + JEMALLOC_FALLTHROUGH + case 4: k1 ^= tail[ 3] << 24; JEMALLOC_FALLTHROUGH + case 3: k1 ^= tail[ 2] << 16; JEMALLOC_FALLTHROUGH + case 2: k1 ^= tail[ 1] << 8; JEMALLOC_FALLTHROUGH + case 1: k1 ^= tail[ 0] << 0; + k1 *= c1; k1 = hash_rotl_32(k1, 15); k1 *= c2; h1 ^= k1; + JEMALLOC_FALLTHROUGH + } + } + + /* finalization */ + h1 ^= len; h2 ^= len; h3 ^= len; h4 ^= len; + + h1 += h2; h1 += h3; h1 += h4; + h2 += h1; h3 += h1; h4 += h1; + + h1 = hash_fmix_32(h1); + h2 = hash_fmix_32(h2); + h3 = hash_fmix_32(h3); + h4 = hash_fmix_32(h4); + + h1 += h2; h1 += h3; h1 += h4; + h2 += h1; h3 += h1; h4 += h1; + + r_out[0] = (((uint64_t) h2) << 32) | h1; + r_out[1] = (((uint64_t) h4) << 32) | h3; +} + +static inline void +hash_x64_128(const void *key, const int len, const uint32_t seed, + uint64_t r_out[2]) { + const uint8_t *data = (const uint8_t *) key; + const int nblocks = len / 16; + + uint64_t h1 = seed; + uint64_t h2 = seed; + + const uint64_t c1 = KQU(0x87c37b91114253d5); + const uint64_t c2 = KQU(0x4cf5ad432745937f); + + /* body */ + { + const uint64_t *blocks = (const uint64_t *) (data); + int i; + + for (i = 0; i < nblocks; i++) { + uint64_t k1 = hash_get_block_64(blocks, i*2 + 0); + uint64_t k2 = hash_get_block_64(blocks, i*2 + 1); + + k1 *= c1; k1 = hash_rotl_64(k1, 31); k1 *= c2; h1 ^= k1; + + h1 = hash_rotl_64(h1, 27); h1 += h2; + h1 = h1*5 + 0x52dce729; + + k2 *= c2; k2 = hash_rotl_64(k2, 33); k2 *= c1; h2 ^= k2; + + h2 = hash_rotl_64(h2, 31); h2 += h1; + h2 = h2*5 + 0x38495ab5; + } + } + + /* tail */ + { + const uint8_t *tail = (const uint8_t*)(data + nblocks*16); + uint64_t k1 = 0; + uint64_t k2 = 0; + + switch (len & 15) { + case 15: k2 ^= ((uint64_t)(tail[14])) << 48; JEMALLOC_FALLTHROUGH + case 14: k2 ^= ((uint64_t)(tail[13])) << 40; JEMALLOC_FALLTHROUGH + case 13: k2 ^= ((uint64_t)(tail[12])) << 32; JEMALLOC_FALLTHROUGH + case 12: k2 ^= ((uint64_t)(tail[11])) << 24; JEMALLOC_FALLTHROUGH + case 11: k2 ^= ((uint64_t)(tail[10])) << 16; JEMALLOC_FALLTHROUGH + case 10: k2 ^= ((uint64_t)(tail[ 9])) << 8; JEMALLOC_FALLTHROUGH + case 9: k2 ^= ((uint64_t)(tail[ 8])) << 0; + k2 *= c2; k2 = hash_rotl_64(k2, 33); k2 *= c1; h2 ^= k2; + JEMALLOC_FALLTHROUGH + case 8: k1 ^= ((uint64_t)(tail[ 7])) << 56; JEMALLOC_FALLTHROUGH + case 7: k1 ^= ((uint64_t)(tail[ 6])) << 48; JEMALLOC_FALLTHROUGH + case 6: k1 ^= ((uint64_t)(tail[ 5])) << 40; JEMALLOC_FALLTHROUGH + case 5: k1 ^= ((uint64_t)(tail[ 4])) << 32; JEMALLOC_FALLTHROUGH + case 4: k1 ^= ((uint64_t)(tail[ 3])) << 24; JEMALLOC_FALLTHROUGH + case 3: k1 ^= ((uint64_t)(tail[ 2])) << 16; JEMALLOC_FALLTHROUGH + case 2: k1 ^= ((uint64_t)(tail[ 1])) << 8; JEMALLOC_FALLTHROUGH + case 1: k1 ^= ((uint64_t)(tail[ 0])) << 0; + k1 *= c1; k1 = hash_rotl_64(k1, 31); k1 *= c2; h1 ^= k1; + } + } + + /* finalization */ + h1 ^= len; h2 ^= len; + + h1 += h2; + h2 += h1; + + h1 = hash_fmix_64(h1); + h2 = hash_fmix_64(h2); + + h1 += h2; + h2 += h1; + + r_out[0] = h1; + r_out[1] = h2; +} + +/******************************************************************************/ +/* API. */ +static inline void +hash(const void *key, size_t len, const uint32_t seed, size_t r_hash[2]) { + assert(len <= INT_MAX); /* Unfortunate implementation limitation. */ + +#if (LG_SIZEOF_PTR == 3 && !defined(JEMALLOC_BIG_ENDIAN)) + hash_x64_128(key, (int)len, seed, (uint64_t *)r_hash); +#else + { + uint64_t hashes[2]; + hash_x86_128(key, (int)len, seed, hashes); + r_hash[0] = (size_t)hashes[0]; + r_hash[1] = (size_t)hashes[1]; + } +#endif +} + +#endif /* JEMALLOC_INTERNAL_HASH_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/hook.h b/deps/jemalloc/include/jemalloc/internal/hook.h new file mode 100644 index 000000000..ee246b1e0 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/hook.h @@ -0,0 +1,163 @@ +#ifndef JEMALLOC_INTERNAL_HOOK_H +#define JEMALLOC_INTERNAL_HOOK_H + +#include "jemalloc/internal/tsd.h" + +/* + * This API is *extremely* experimental, and may get ripped out, changed in API- + * and ABI-incompatible ways, be insufficiently or incorrectly documented, etc. + * + * It allows hooking the stateful parts of the API to see changes as they + * happen. + * + * Allocation hooks are called after the allocation is done, free hooks are + * called before the free is done, and expand hooks are called after the + * allocation is expanded. + * + * For realloc and rallocx, if the expansion happens in place, the expansion + * hook is called. If it is moved, then the alloc hook is called on the new + * location, and then the free hook is called on the old location (i.e. both + * hooks are invoked in between the alloc and the dalloc). + * + * If we return NULL from OOM, then usize might not be trustworthy. Calling + * realloc(NULL, size) only calls the alloc hook, and calling realloc(ptr, 0) + * only calls the free hook. (Calling realloc(NULL, 0) is treated as malloc(0), + * and only calls the alloc hook). + * + * Reentrancy: + * Reentrancy is guarded against from within the hook implementation. If you + * call allocator functions from within a hook, the hooks will not be invoked + * again. + * Threading: + * The installation of a hook synchronizes with all its uses. If you can + * prove the installation of a hook happens-before a jemalloc entry point, + * then the hook will get invoked (unless there's a racing removal). + * + * Hook insertion appears to be atomic at a per-thread level (i.e. if a thread + * allocates and has the alloc hook invoked, then a subsequent free on the + * same thread will also have the free hook invoked). + * + * The *removal* of a hook does *not* block until all threads are done with + * the hook. Hook authors have to be resilient to this, and need some + * out-of-band mechanism for cleaning up any dynamically allocated memory + * associated with their hook. + * Ordering: + * Order of hook execution is unspecified, and may be different than insertion + * order. + */ + +#define HOOK_MAX 4 + +enum hook_alloc_e { + hook_alloc_malloc, + hook_alloc_posix_memalign, + hook_alloc_aligned_alloc, + hook_alloc_calloc, + hook_alloc_memalign, + hook_alloc_valloc, + hook_alloc_mallocx, + + /* The reallocating functions have both alloc and dalloc variants */ + hook_alloc_realloc, + hook_alloc_rallocx, +}; +/* + * We put the enum typedef after the enum, since this file may get included by + * jemalloc_cpp.cpp, and C++ disallows enum forward declarations. + */ +typedef enum hook_alloc_e hook_alloc_t; + +enum hook_dalloc_e { + hook_dalloc_free, + hook_dalloc_dallocx, + hook_dalloc_sdallocx, + + /* + * The dalloc halves of reallocation (not called if in-place expansion + * happens). + */ + hook_dalloc_realloc, + hook_dalloc_rallocx, +}; +typedef enum hook_dalloc_e hook_dalloc_t; + + +enum hook_expand_e { + hook_expand_realloc, + hook_expand_rallocx, + hook_expand_xallocx, +}; +typedef enum hook_expand_e hook_expand_t; + +typedef void (*hook_alloc)( + void *extra, hook_alloc_t type, void *result, uintptr_t result_raw, + uintptr_t args_raw[3]); + +typedef void (*hook_dalloc)( + void *extra, hook_dalloc_t type, void *address, uintptr_t args_raw[3]); + +typedef void (*hook_expand)( + void *extra, hook_expand_t type, void *address, size_t old_usize, + size_t new_usize, uintptr_t result_raw, uintptr_t args_raw[4]); + +typedef struct hooks_s hooks_t; +struct hooks_s { + hook_alloc alloc_hook; + hook_dalloc dalloc_hook; + hook_expand expand_hook; + void *extra; +}; + +/* + * Begin implementation details; everything above this point might one day live + * in a public API. Everything below this point never will. + */ + +/* + * The realloc pathways haven't gotten any refactoring love in a while, and it's + * fairly difficult to pass information from the entry point to the hooks. We + * put the informaiton the hooks will need into a struct to encapsulate + * everything. + * + * Much of these pathways are force-inlined, so that the compiler can avoid + * materializing this struct until we hit an extern arena function. For fairly + * goofy reasons, *many* of the realloc paths hit an extern arena function. + * These paths are cold enough that it doesn't matter; eventually, we should + * rewrite the realloc code to make the expand-in-place and the + * free-then-realloc paths more orthogonal, at which point we don't need to + * spread the hook logic all over the place. + */ +typedef struct hook_ralloc_args_s hook_ralloc_args_t; +struct hook_ralloc_args_s { + /* I.e. as opposed to rallocx. */ + bool is_realloc; + /* + * The expand hook takes 4 arguments, even if only 3 are actually used; + * we add an extra one in case the user decides to memcpy without + * looking too closely at the hooked function. + */ + uintptr_t args[4]; +}; + +/* + * Returns an opaque handle to be used when removing the hook. NULL means that + * we couldn't install the hook. + */ +bool hook_boot(); + +void *hook_install(tsdn_t *tsdn, hooks_t *hooks); +/* Uninstalls the hook with the handle previously returned from hook_install. */ +void hook_remove(tsdn_t *tsdn, void *opaque); + +/* Hooks */ + +void hook_invoke_alloc(hook_alloc_t type, void *result, uintptr_t result_raw, + uintptr_t args_raw[3]); + +void hook_invoke_dalloc(hook_dalloc_t type, void *address, + uintptr_t args_raw[3]); + +void hook_invoke_expand(hook_expand_t type, void *address, size_t old_usize, + size_t new_usize, uintptr_t result_raw, uintptr_t args_raw[4]); + +#endif /* JEMALLOC_INTERNAL_HOOK_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_decls.h b/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_decls.h new file mode 100644 index 000000000..7d6053e21 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_decls.h @@ -0,0 +1,94 @@ +#ifndef JEMALLOC_INTERNAL_DECLS_H +#define JEMALLOC_INTERNAL_DECLS_H + +#include <math.h> +#ifdef _WIN32 +# include <windows.h> +# include "msvc_compat/windows_extra.h" +# ifdef _WIN64 +# if LG_VADDR <= 32 +# error Generate the headers using x64 vcargs +# endif +# else +# if LG_VADDR > 32 +# undef LG_VADDR +# define LG_VADDR 32 +# endif +# endif +#else +# include <sys/param.h> +# include <sys/mman.h> +# if !defined(__pnacl__) && !defined(__native_client__) +# include <sys/syscall.h> +# if !defined(SYS_write) && defined(__NR_write) +# define SYS_write __NR_write +# endif +# if defined(SYS_open) && defined(__aarch64__) + /* Android headers may define SYS_open to __NR_open even though + * __NR_open may not exist on AArch64 (superseded by __NR_openat). */ +# undef SYS_open +# endif +# include <sys/uio.h> +# endif +# include <pthread.h> +# ifdef __FreeBSD__ +# include <pthread_np.h> +# endif +# include <signal.h> +# ifdef JEMALLOC_OS_UNFAIR_LOCK +# include <os/lock.h> +# endif +# ifdef JEMALLOC_GLIBC_MALLOC_HOOK +# include <sched.h> +# endif +# include <errno.h> +# include <sys/time.h> +# include <time.h> +# ifdef JEMALLOC_HAVE_MACH_ABSOLUTE_TIME +# include <mach/mach_time.h> +# endif +#endif +#include <sys/types.h> + +#include <limits.h> +#ifndef SIZE_T_MAX +# define SIZE_T_MAX SIZE_MAX +#endif +#ifndef SSIZE_MAX +# define SSIZE_MAX ((ssize_t)(SIZE_T_MAX >> 1)) +#endif +#include <stdarg.h> +#include <stdbool.h> +#include <stdio.h> +#include <stdlib.h> +#include <stdint.h> +#include <stddef.h> +#ifndef offsetof +# define offsetof(type, member) ((size_t)&(((type *)NULL)->member)) +#endif +#include <string.h> +#include <strings.h> +#include <ctype.h> +#ifdef _MSC_VER +# include <io.h> +typedef intptr_t ssize_t; +# define PATH_MAX 1024 +# define STDERR_FILENO 2 +# define __func__ __FUNCTION__ +# ifdef JEMALLOC_HAS_RESTRICT +# define restrict __restrict +# endif +/* Disable warnings about deprecated system functions. */ +# pragma warning(disable: 4996) +#if _MSC_VER < 1800 +static int +isblank(int c) { + return (c == '\t' || c == ' '); +} +#endif +#else +# include <unistd.h> +#endif +#include <fcntl.h> + +#endif /* JEMALLOC_INTERNAL_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_defs.h.in b/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_defs.h.in new file mode 100644 index 000000000..c442a2191 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_defs.h.in @@ -0,0 +1,366 @@ +#ifndef JEMALLOC_INTERNAL_DEFS_H_ +#define JEMALLOC_INTERNAL_DEFS_H_ +/* + * If JEMALLOC_PREFIX is defined via --with-jemalloc-prefix, it will cause all + * public APIs to be prefixed. This makes it possible, with some care, to use + * multiple allocators simultaneously. + */ +#undef JEMALLOC_PREFIX +#undef JEMALLOC_CPREFIX + +/* + * Define overrides for non-standard allocator-related functions if they are + * present on the system. + */ +#undef JEMALLOC_OVERRIDE___LIBC_CALLOC +#undef JEMALLOC_OVERRIDE___LIBC_FREE +#undef JEMALLOC_OVERRIDE___LIBC_MALLOC +#undef JEMALLOC_OVERRIDE___LIBC_MEMALIGN +#undef JEMALLOC_OVERRIDE___LIBC_REALLOC +#undef JEMALLOC_OVERRIDE___LIBC_VALLOC +#undef JEMALLOC_OVERRIDE___POSIX_MEMALIGN + +/* + * JEMALLOC_PRIVATE_NAMESPACE is used as a prefix for all library-private APIs. + * For shared libraries, symbol visibility mechanisms prevent these symbols + * from being exported, but for static libraries, naming collisions are a real + * possibility. + */ +#undef JEMALLOC_PRIVATE_NAMESPACE + +/* + * Hyper-threaded CPUs may need a special instruction inside spin loops in + * order to yield to another virtual CPU. + */ +#undef CPU_SPINWAIT +/* 1 if CPU_SPINWAIT is defined, 0 otherwise. */ +#undef HAVE_CPU_SPINWAIT + +/* + * Number of significant bits in virtual addresses. This may be less than the + * total number of bits in a pointer, e.g. on x64, for which the uppermost 16 + * bits are the same as bit 47. + */ +#undef LG_VADDR + +/* Defined if C11 atomics are available. */ +#undef JEMALLOC_C11_ATOMICS + +/* Defined if GCC __atomic atomics are available. */ +#undef JEMALLOC_GCC_ATOMIC_ATOMICS +/* and the 8-bit variant support. */ +#undef JEMALLOC_GCC_U8_ATOMIC_ATOMICS + +/* Defined if GCC __sync atomics are available. */ +#undef JEMALLOC_GCC_SYNC_ATOMICS +/* and the 8-bit variant support. */ +#undef JEMALLOC_GCC_U8_SYNC_ATOMICS + +/* + * Defined if __builtin_clz() and __builtin_clzl() are available. + */ +#undef JEMALLOC_HAVE_BUILTIN_CLZ + +/* + * Defined if os_unfair_lock_*() functions are available, as provided by Darwin. + */ +#undef JEMALLOC_OS_UNFAIR_LOCK + +/* Defined if syscall(2) is usable. */ +#undef JEMALLOC_USE_SYSCALL + +/* + * Defined if secure_getenv(3) is available. + */ +#undef JEMALLOC_HAVE_SECURE_GETENV + +/* + * Defined if issetugid(2) is available. + */ +#undef JEMALLOC_HAVE_ISSETUGID + +/* Defined if pthread_atfork(3) is available. */ +#undef JEMALLOC_HAVE_PTHREAD_ATFORK + +/* Defined if pthread_setname_np(3) is available. */ +#undef JEMALLOC_HAVE_PTHREAD_SETNAME_NP + +/* + * Defined if clock_gettime(CLOCK_MONOTONIC_COARSE, ...) is available. + */ +#undef JEMALLOC_HAVE_CLOCK_MONOTONIC_COARSE + +/* + * Defined if clock_gettime(CLOCK_MONOTONIC, ...) is available. + */ +#undef JEMALLOC_HAVE_CLOCK_MONOTONIC + +/* + * Defined if mach_absolute_time() is available. + */ +#undef JEMALLOC_HAVE_MACH_ABSOLUTE_TIME + +/* + * Defined if _malloc_thread_cleanup() exists. At least in the case of + * FreeBSD, pthread_key_create() allocates, which if used during malloc + * bootstrapping will cause recursion into the pthreads library. Therefore, if + * _malloc_thread_cleanup() exists, use it as the basis for thread cleanup in + * malloc_tsd. + */ +#undef JEMALLOC_MALLOC_THREAD_CLEANUP + +/* + * Defined if threaded initialization is known to be safe on this platform. + * Among other things, it must be possible to initialize a mutex without + * triggering allocation in order for threaded allocation to be safe. + */ +#undef JEMALLOC_THREADED_INIT + +/* + * Defined if the pthreads implementation defines + * _pthread_mutex_init_calloc_cb(), in which case the function is used in order + * to avoid recursive allocation during mutex initialization. + */ +#undef JEMALLOC_MUTEX_INIT_CB + +/* Non-empty if the tls_model attribute is supported. */ +#undef JEMALLOC_TLS_MODEL + +/* + * JEMALLOC_DEBUG enables assertions and other sanity checks, and disables + * inline functions. + */ +#undef JEMALLOC_DEBUG + +/* JEMALLOC_STATS enables statistics calculation. */ +#undef JEMALLOC_STATS + +/* JEMALLOC_EXPERIMENTAL_SMALLOCX_API enables experimental smallocx API. */ +#undef JEMALLOC_EXPERIMENTAL_SMALLOCX_API + +/* JEMALLOC_PROF enables allocation profiling. */ +#undef JEMALLOC_PROF + +/* Use libunwind for profile backtracing if defined. */ +#undef JEMALLOC_PROF_LIBUNWIND + +/* Use libgcc for profile backtracing if defined. */ +#undef JEMALLOC_PROF_LIBGCC + +/* Use gcc intrinsics for profile backtracing if defined. */ +#undef JEMALLOC_PROF_GCC + +/* + * JEMALLOC_DSS enables use of sbrk(2) to allocate extents from the data storage + * segment (DSS). + */ +#undef JEMALLOC_DSS + +/* Support memory filling (junk/zero). */ +#undef JEMALLOC_FILL + +/* Support utrace(2)-based tracing. */ +#undef JEMALLOC_UTRACE + +/* Support optional abort() on OOM. */ +#undef JEMALLOC_XMALLOC + +/* Support lazy locking (avoid locking unless a second thread is launched). */ +#undef JEMALLOC_LAZY_LOCK + +/* + * Minimum allocation alignment is 2^LG_QUANTUM bytes (ignoring tiny size + * classes). + */ +#undef LG_QUANTUM + +/* One page is 2^LG_PAGE bytes. */ +#undef LG_PAGE + +/* + * One huge page is 2^LG_HUGEPAGE bytes. Note that this is defined even if the + * system does not explicitly support huge pages; system calls that require + * explicit huge page support are separately configured. + */ +#undef LG_HUGEPAGE + +/* + * If defined, adjacent virtual memory mappings with identical attributes + * automatically coalesce, and they fragment when changes are made to subranges. + * This is the normal order of things for mmap()/munmap(), but on Windows + * VirtualAlloc()/VirtualFree() operations must be precisely matched, i.e. + * mappings do *not* coalesce/fragment. + */ +#undef JEMALLOC_MAPS_COALESCE + +/* + * If defined, retain memory for later reuse by default rather than using e.g. + * munmap() to unmap freed extents. This is enabled on 64-bit Linux because + * common sequences of mmap()/munmap() calls will cause virtual memory map + * holes. + */ +#undef JEMALLOC_RETAIN + +/* TLS is used to map arenas and magazine caches to threads. */ +#undef JEMALLOC_TLS + +/* + * Used to mark unreachable code to quiet "end of non-void" compiler warnings. + * Don't use this directly; instead use unreachable() from util.h + */ +#undef JEMALLOC_INTERNAL_UNREACHABLE + +/* + * ffs*() functions to use for bitmapping. Don't use these directly; instead, + * use ffs_*() from util.h. + */ +#undef JEMALLOC_INTERNAL_FFSLL +#undef JEMALLOC_INTERNAL_FFSL +#undef JEMALLOC_INTERNAL_FFS + +/* + * popcount*() functions to use for bitmapping. + */ +#undef JEMALLOC_INTERNAL_POPCOUNTL +#undef JEMALLOC_INTERNAL_POPCOUNT + +/* + * If defined, explicitly attempt to more uniformly distribute large allocation + * pointer alignments across all cache indices. + */ +#undef JEMALLOC_CACHE_OBLIVIOUS + +/* + * If defined, enable logging facilities. We make this a configure option to + * avoid taking extra branches everywhere. + */ +#undef JEMALLOC_LOG + +/* + * If defined, use readlinkat() (instead of readlink()) to follow + * /etc/malloc_conf. + */ +#undef JEMALLOC_READLINKAT + +/* + * Darwin (OS X) uses zones to work around Mach-O symbol override shortcomings. + */ +#undef JEMALLOC_ZONE + +/* + * Methods for determining whether the OS overcommits. + * JEMALLOC_PROC_SYS_VM_OVERCOMMIT_MEMORY: Linux's + * /proc/sys/vm.overcommit_memory file. + * JEMALLOC_SYSCTL_VM_OVERCOMMIT: FreeBSD's vm.overcommit sysctl. + */ +#undef JEMALLOC_SYSCTL_VM_OVERCOMMIT +#undef JEMALLOC_PROC_SYS_VM_OVERCOMMIT_MEMORY + +/* Defined if madvise(2) is available. */ +#undef JEMALLOC_HAVE_MADVISE + +/* + * Defined if transparent huge pages are supported via the MADV_[NO]HUGEPAGE + * arguments to madvise(2). + */ +#undef JEMALLOC_HAVE_MADVISE_HUGE + +/* + * Methods for purging unused pages differ between operating systems. + * + * madvise(..., MADV_FREE) : This marks pages as being unused, such that they + * will be discarded rather than swapped out. + * madvise(..., MADV_DONTNEED) : If JEMALLOC_PURGE_MADVISE_DONTNEED_ZEROS is + * defined, this immediately discards pages, + * such that new pages will be demand-zeroed if + * the address region is later touched; + * otherwise this behaves similarly to + * MADV_FREE, though typically with higher + * system overhead. + */ +#undef JEMALLOC_PURGE_MADVISE_FREE +#undef JEMALLOC_PURGE_MADVISE_DONTNEED +#undef JEMALLOC_PURGE_MADVISE_DONTNEED_ZEROS + +/* Defined if madvise(2) is available but MADV_FREE is not (x86 Linux only). */ +#undef JEMALLOC_DEFINE_MADVISE_FREE + +/* + * Defined if MADV_DO[NT]DUMP is supported as an argument to madvise. + */ +#undef JEMALLOC_MADVISE_DONTDUMP + +/* + * Defined if transparent huge pages (THPs) are supported via the + * MADV_[NO]HUGEPAGE arguments to madvise(2), and THP support is enabled. + */ +#undef JEMALLOC_THP + +/* Define if operating system has alloca.h header. */ +#undef JEMALLOC_HAS_ALLOCA_H + +/* C99 restrict keyword supported. */ +#undef JEMALLOC_HAS_RESTRICT + +/* For use by hash code. */ +#undef JEMALLOC_BIG_ENDIAN + +/* sizeof(int) == 2^LG_SIZEOF_INT. */ +#undef LG_SIZEOF_INT + +/* sizeof(long) == 2^LG_SIZEOF_LONG. */ +#undef LG_SIZEOF_LONG + +/* sizeof(long long) == 2^LG_SIZEOF_LONG_LONG. */ +#undef LG_SIZEOF_LONG_LONG + +/* sizeof(intmax_t) == 2^LG_SIZEOF_INTMAX_T. */ +#undef LG_SIZEOF_INTMAX_T + +/* glibc malloc hooks (__malloc_hook, __realloc_hook, __free_hook). */ +#undef JEMALLOC_GLIBC_MALLOC_HOOK + +/* glibc memalign hook. */ +#undef JEMALLOC_GLIBC_MEMALIGN_HOOK + +/* pthread support */ +#undef JEMALLOC_HAVE_PTHREAD + +/* dlsym() support */ +#undef JEMALLOC_HAVE_DLSYM + +/* Adaptive mutex support in pthreads. */ +#undef JEMALLOC_HAVE_PTHREAD_MUTEX_ADAPTIVE_NP + +/* GNU specific sched_getcpu support */ +#undef JEMALLOC_HAVE_SCHED_GETCPU + +/* GNU specific sched_setaffinity support */ +#undef JEMALLOC_HAVE_SCHED_SETAFFINITY + +/* + * If defined, all the features necessary for background threads are present. + */ +#undef JEMALLOC_BACKGROUND_THREAD + +/* + * If defined, jemalloc symbols are not exported (doesn't work when + * JEMALLOC_PREFIX is not defined). + */ +#undef JEMALLOC_EXPORT + +/* config.malloc_conf options string. */ +#undef JEMALLOC_CONFIG_MALLOC_CONF + +/* If defined, jemalloc takes the malloc/free/etc. symbol names. */ +#undef JEMALLOC_IS_MALLOC + +/* + * Defined if strerror_r returns char * if _GNU_SOURCE is defined. + */ +#undef JEMALLOC_STRERROR_R_RETURNS_CHAR_WITH_GNU_SOURCE + +/* Performs additional safety checks when defined. */ +#undef JEMALLOC_OPT_SAFETY_CHECKS + +#endif /* JEMALLOC_INTERNAL_DEFS_H_ */ diff --git a/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_externs.h b/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_externs.h new file mode 100644 index 000000000..d291170be --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_externs.h @@ -0,0 +1,57 @@ +#ifndef JEMALLOC_INTERNAL_EXTERNS_H +#define JEMALLOC_INTERNAL_EXTERNS_H + +#include "jemalloc/internal/atomic.h" +#include "jemalloc/internal/tsd_types.h" + +/* TSD checks this to set thread local slow state accordingly. */ +extern bool malloc_slow; + +/* Run-time options. */ +extern bool opt_abort; +extern bool opt_abort_conf; +extern bool opt_confirm_conf; +extern const char *opt_junk; +extern bool opt_junk_alloc; +extern bool opt_junk_free; +extern bool opt_utrace; +extern bool opt_xmalloc; +extern bool opt_zero; +extern unsigned opt_narenas; + +/* Number of CPUs. */ +extern unsigned ncpus; + +/* Number of arenas used for automatic multiplexing of threads and arenas. */ +extern unsigned narenas_auto; + +/* Base index for manual arenas. */ +extern unsigned manual_arena_base; + +/* + * Arenas that are used to service external requests. Not all elements of the + * arenas array are necessarily used; arenas are created lazily as needed. + */ +extern atomic_p_t arenas[]; + +void *a0malloc(size_t size); +void a0dalloc(void *ptr); +void *bootstrap_malloc(size_t size); +void *bootstrap_calloc(size_t num, size_t size); +void bootstrap_free(void *ptr); +void arena_set(unsigned ind, arena_t *arena); +unsigned narenas_total_get(void); +arena_t *arena_init(tsdn_t *tsdn, unsigned ind, extent_hooks_t *extent_hooks); +arena_tdata_t *arena_tdata_get_hard(tsd_t *tsd, unsigned ind); +arena_t *arena_choose_hard(tsd_t *tsd, bool internal); +void arena_migrate(tsd_t *tsd, unsigned oldind, unsigned newind); +void iarena_cleanup(tsd_t *tsd); +void arena_cleanup(tsd_t *tsd); +void arenas_tdata_cleanup(tsd_t *tsd); +void jemalloc_prefork(void); +void jemalloc_postfork_parent(void); +void jemalloc_postfork_child(void); +bool malloc_initialized(void); +void je_sdallocx_noflags(void *ptr, size_t size); + +#endif /* JEMALLOC_INTERNAL_EXTERNS_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_includes.h b/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_includes.h new file mode 100644 index 000000000..437eaa407 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_includes.h @@ -0,0 +1,94 @@ +#ifndef JEMALLOC_INTERNAL_INCLUDES_H +#define JEMALLOC_INTERNAL_INCLUDES_H + +/* + * jemalloc can conceptually be broken into components (arena, tcache, etc.), + * but there are circular dependencies that cannot be broken without + * substantial performance degradation. + * + * Historically, we dealt with this by each header into four sections (types, + * structs, externs, and inlines), and included each header file multiple times + * in this file, picking out the portion we want on each pass using the + * following #defines: + * JEMALLOC_H_TYPES : Preprocessor-defined constants and psuedo-opaque data + * types. + * JEMALLOC_H_STRUCTS : Data structures. + * JEMALLOC_H_EXTERNS : Extern data declarations and function prototypes. + * JEMALLOC_H_INLINES : Inline functions. + * + * We're moving toward a world in which the dependencies are explicit; each file + * will #include the headers it depends on (rather than relying on them being + * implicitly available via this file including every header file in the + * project). + * + * We're now in an intermediate state: we've broken up the header files to avoid + * having to include each one multiple times, but have not yet moved the + * dependency information into the header files (i.e. we still rely on the + * ordering in this file to ensure all a header's dependencies are available in + * its translation unit). Each component is now broken up into multiple header + * files, corresponding to the sections above (e.g. instead of "foo.h", we now + * have "foo_types.h", "foo_structs.h", "foo_externs.h", "foo_inlines.h"). + * + * Those files which have been converted to explicitly include their + * inter-component dependencies are now in the initial HERMETIC HEADERS + * section. All headers may still rely on jemalloc_preamble.h (which, by fiat, + * must be included first in every translation unit) for system headers and + * global jemalloc definitions, however. + */ + +/******************************************************************************/ +/* TYPES */ +/******************************************************************************/ + +#include "jemalloc/internal/extent_types.h" +#include "jemalloc/internal/base_types.h" +#include "jemalloc/internal/arena_types.h" +#include "jemalloc/internal/tcache_types.h" +#include "jemalloc/internal/prof_types.h" + +/******************************************************************************/ +/* STRUCTS */ +/******************************************************************************/ + +#include "jemalloc/internal/arena_structs_a.h" +#include "jemalloc/internal/extent_structs.h" +#include "jemalloc/internal/base_structs.h" +#include "jemalloc/internal/prof_structs.h" +#include "jemalloc/internal/arena_structs_b.h" +#include "jemalloc/internal/tcache_structs.h" +#include "jemalloc/internal/background_thread_structs.h" + +/******************************************************************************/ +/* EXTERNS */ +/******************************************************************************/ + +#include "jemalloc/internal/jemalloc_internal_externs.h" +#include "jemalloc/internal/extent_externs.h" +#include "jemalloc/internal/base_externs.h" +#include "jemalloc/internal/arena_externs.h" +#include "jemalloc/internal/large_externs.h" +#include "jemalloc/internal/tcache_externs.h" +#include "jemalloc/internal/prof_externs.h" +#include "jemalloc/internal/background_thread_externs.h" + +/******************************************************************************/ +/* INLINES */ +/******************************************************************************/ + +#include "jemalloc/internal/jemalloc_internal_inlines_a.h" +#include "jemalloc/internal/base_inlines.h" +/* + * Include portions of arena code interleaved with tcache code in order to + * resolve circular dependencies. + */ +#include "jemalloc/internal/prof_inlines_a.h" +#include "jemalloc/internal/arena_inlines_a.h" +#include "jemalloc/internal/extent_inlines.h" +#include "jemalloc/internal/jemalloc_internal_inlines_b.h" +#include "jemalloc/internal/tcache_inlines.h" +#include "jemalloc/internal/arena_inlines_b.h" +#include "jemalloc/internal/jemalloc_internal_inlines_c.h" +#include "jemalloc/internal/prof_inlines_b.h" +#include "jemalloc/internal/background_thread_inlines.h" + +#endif /* JEMALLOC_INTERNAL_INCLUDES_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_inlines_a.h b/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_inlines_a.h new file mode 100644 index 000000000..ddde9b4e6 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_inlines_a.h @@ -0,0 +1,174 @@ +#ifndef JEMALLOC_INTERNAL_INLINES_A_H +#define JEMALLOC_INTERNAL_INLINES_A_H + +#include "jemalloc/internal/atomic.h" +#include "jemalloc/internal/bit_util.h" +#include "jemalloc/internal/jemalloc_internal_types.h" +#include "jemalloc/internal/sc.h" +#include "jemalloc/internal/ticker.h" + +JEMALLOC_ALWAYS_INLINE malloc_cpuid_t +malloc_getcpu(void) { + assert(have_percpu_arena); +#if defined(_WIN32) + return GetCurrentProcessorNumber(); +#elif defined(JEMALLOC_HAVE_SCHED_GETCPU) + return (malloc_cpuid_t)sched_getcpu(); +#else + not_reached(); + return -1; +#endif +} + +/* Return the chosen arena index based on current cpu. */ +JEMALLOC_ALWAYS_INLINE unsigned +percpu_arena_choose(void) { + assert(have_percpu_arena && PERCPU_ARENA_ENABLED(opt_percpu_arena)); + + malloc_cpuid_t cpuid = malloc_getcpu(); + assert(cpuid >= 0); + + unsigned arena_ind; + if ((opt_percpu_arena == percpu_arena) || ((unsigned)cpuid < ncpus / + 2)) { + arena_ind = cpuid; + } else { + assert(opt_percpu_arena == per_phycpu_arena); + /* Hyper threads on the same physical CPU share arena. */ + arena_ind = cpuid - ncpus / 2; + } + + return arena_ind; +} + +/* Return the limit of percpu auto arena range, i.e. arenas[0...ind_limit). */ +JEMALLOC_ALWAYS_INLINE unsigned +percpu_arena_ind_limit(percpu_arena_mode_t mode) { + assert(have_percpu_arena && PERCPU_ARENA_ENABLED(mode)); + if (mode == per_phycpu_arena && ncpus > 1) { + if (ncpus % 2) { + /* This likely means a misconfig. */ + return ncpus / 2 + 1; + } + return ncpus / 2; + } else { + return ncpus; + } +} + +static inline arena_tdata_t * +arena_tdata_get(tsd_t *tsd, unsigned ind, bool refresh_if_missing) { + arena_tdata_t *tdata; + arena_tdata_t *arenas_tdata = tsd_arenas_tdata_get(tsd); + + if (unlikely(arenas_tdata == NULL)) { + /* arenas_tdata hasn't been initialized yet. */ + return arena_tdata_get_hard(tsd, ind); + } + if (unlikely(ind >= tsd_narenas_tdata_get(tsd))) { + /* + * ind is invalid, cache is old (too small), or tdata to be + * initialized. + */ + return (refresh_if_missing ? arena_tdata_get_hard(tsd, ind) : + NULL); + } + + tdata = &arenas_tdata[ind]; + if (likely(tdata != NULL) || !refresh_if_missing) { + return tdata; + } + return arena_tdata_get_hard(tsd, ind); +} + +static inline arena_t * +arena_get(tsdn_t *tsdn, unsigned ind, bool init_if_missing) { + arena_t *ret; + + assert(ind < MALLOCX_ARENA_LIMIT); + + ret = (arena_t *)atomic_load_p(&arenas[ind], ATOMIC_ACQUIRE); + if (unlikely(ret == NULL)) { + if (init_if_missing) { + ret = arena_init(tsdn, ind, + (extent_hooks_t *)&extent_hooks_default); + } + } + return ret; +} + +static inline ticker_t * +decay_ticker_get(tsd_t *tsd, unsigned ind) { + arena_tdata_t *tdata; + + tdata = arena_tdata_get(tsd, ind, true); + if (unlikely(tdata == NULL)) { + return NULL; + } + return &tdata->decay_ticker; +} + +JEMALLOC_ALWAYS_INLINE cache_bin_t * +tcache_small_bin_get(tcache_t *tcache, szind_t binind) { + assert(binind < SC_NBINS); + return &tcache->bins_small[binind]; +} + +JEMALLOC_ALWAYS_INLINE cache_bin_t * +tcache_large_bin_get(tcache_t *tcache, szind_t binind) { + assert(binind >= SC_NBINS &&binind < nhbins); + return &tcache->bins_large[binind - SC_NBINS]; +} + +JEMALLOC_ALWAYS_INLINE bool +tcache_available(tsd_t *tsd) { + /* + * Thread specific auto tcache might be unavailable if: 1) during tcache + * initialization, or 2) disabled through thread.tcache.enabled mallctl + * or config options. This check covers all cases. + */ + if (likely(tsd_tcache_enabled_get(tsd))) { + /* Associated arena == NULL implies tcache init in progress. */ + assert(tsd_tcachep_get(tsd)->arena == NULL || + tcache_small_bin_get(tsd_tcachep_get(tsd), 0)->avail != + NULL); + return true; + } + + return false; +} + +JEMALLOC_ALWAYS_INLINE tcache_t * +tcache_get(tsd_t *tsd) { + if (!tcache_available(tsd)) { + return NULL; + } + + return tsd_tcachep_get(tsd); +} + +static inline void +pre_reentrancy(tsd_t *tsd, arena_t *arena) { + /* arena is the current context. Reentry from a0 is not allowed. */ + assert(arena != arena_get(tsd_tsdn(tsd), 0, false)); + + bool fast = tsd_fast(tsd); + assert(tsd_reentrancy_level_get(tsd) < INT8_MAX); + ++*tsd_reentrancy_levelp_get(tsd); + if (fast) { + /* Prepare slow path for reentrancy. */ + tsd_slow_update(tsd); + assert(tsd_state_get(tsd) == tsd_state_nominal_slow); + } +} + +static inline void +post_reentrancy(tsd_t *tsd) { + int8_t *reentrancy_level = tsd_reentrancy_levelp_get(tsd); + assert(*reentrancy_level > 0); + if (--*reentrancy_level == 0) { + tsd_slow_update(tsd); + } +} + +#endif /* JEMALLOC_INTERNAL_INLINES_A_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_inlines_b.h b/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_inlines_b.h new file mode 100644 index 000000000..70d6e5788 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_inlines_b.h @@ -0,0 +1,87 @@ +#ifndef JEMALLOC_INTERNAL_INLINES_B_H +#define JEMALLOC_INTERNAL_INLINES_B_H + +#include "jemalloc/internal/rtree.h" + +/* Choose an arena based on a per-thread value. */ +static inline arena_t * +arena_choose_impl(tsd_t *tsd, arena_t *arena, bool internal) { + arena_t *ret; + + if (arena != NULL) { + return arena; + } + + /* During reentrancy, arena 0 is the safest bet. */ + if (unlikely(tsd_reentrancy_level_get(tsd) > 0)) { + return arena_get(tsd_tsdn(tsd), 0, true); + } + + ret = internal ? tsd_iarena_get(tsd) : tsd_arena_get(tsd); + if (unlikely(ret == NULL)) { + ret = arena_choose_hard(tsd, internal); + assert(ret); + if (tcache_available(tsd)) { + tcache_t *tcache = tcache_get(tsd); + if (tcache->arena != NULL) { + /* See comments in tcache_data_init().*/ + assert(tcache->arena == + arena_get(tsd_tsdn(tsd), 0, false)); + if (tcache->arena != ret) { + tcache_arena_reassociate(tsd_tsdn(tsd), + tcache, ret); + } + } else { + tcache_arena_associate(tsd_tsdn(tsd), tcache, + ret); + } + } + } + + /* + * Note that for percpu arena, if the current arena is outside of the + * auto percpu arena range, (i.e. thread is assigned to a manually + * managed arena), then percpu arena is skipped. + */ + if (have_percpu_arena && PERCPU_ARENA_ENABLED(opt_percpu_arena) && + !internal && (arena_ind_get(ret) < + percpu_arena_ind_limit(opt_percpu_arena)) && (ret->last_thd != + tsd_tsdn(tsd))) { + unsigned ind = percpu_arena_choose(); + if (arena_ind_get(ret) != ind) { + percpu_arena_update(tsd, ind); + ret = tsd_arena_get(tsd); + } + ret->last_thd = tsd_tsdn(tsd); + } + + return ret; +} + +static inline arena_t * +arena_choose(tsd_t *tsd, arena_t *arena) { + return arena_choose_impl(tsd, arena, false); +} + +static inline arena_t * +arena_ichoose(tsd_t *tsd, arena_t *arena) { + return arena_choose_impl(tsd, arena, true); +} + +static inline bool +arena_is_auto(arena_t *arena) { + assert(narenas_auto > 0); + + return (arena_ind_get(arena) < manual_arena_base); +} + +JEMALLOC_ALWAYS_INLINE extent_t * +iealloc(tsdn_t *tsdn, const void *ptr) { + rtree_ctx_t rtree_ctx_fallback; + rtree_ctx_t *rtree_ctx = tsdn_rtree_ctx(tsdn, &rtree_ctx_fallback); + + return rtree_extent_read(tsdn, &extents_rtree, rtree_ctx, + (uintptr_t)ptr, true); +} + +#endif /* JEMALLOC_INTERNAL_INLINES_B_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_inlines_c.h b/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_inlines_c.h new file mode 100644 index 000000000..cdb10eb21 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_inlines_c.h @@ -0,0 +1,222 @@ +#ifndef JEMALLOC_INTERNAL_INLINES_C_H +#define JEMALLOC_INTERNAL_INLINES_C_H + +#include "jemalloc/internal/hook.h" +#include "jemalloc/internal/jemalloc_internal_types.h" +#include "jemalloc/internal/sz.h" +#include "jemalloc/internal/witness.h" + +/* + * Translating the names of the 'i' functions: + * Abbreviations used in the first part of the function name (before + * alloc/dalloc) describe what that function accomplishes: + * a: arena (query) + * s: size (query, or sized deallocation) + * e: extent (query) + * p: aligned (allocates) + * vs: size (query, without knowing that the pointer is into the heap) + * r: rallocx implementation + * x: xallocx implementation + * Abbreviations used in the second part of the function name (after + * alloc/dalloc) describe the arguments it takes + * z: whether to return zeroed memory + * t: accepts a tcache_t * parameter + * m: accepts an arena_t * parameter + */ + +JEMALLOC_ALWAYS_INLINE arena_t * +iaalloc(tsdn_t *tsdn, const void *ptr) { + assert(ptr != NULL); + + return arena_aalloc(tsdn, ptr); +} + +JEMALLOC_ALWAYS_INLINE size_t +isalloc(tsdn_t *tsdn, const void *ptr) { + assert(ptr != NULL); + + return arena_salloc(tsdn, ptr); +} + +JEMALLOC_ALWAYS_INLINE void * +iallocztm(tsdn_t *tsdn, size_t size, szind_t ind, bool zero, tcache_t *tcache, + bool is_internal, arena_t *arena, bool slow_path) { + void *ret; + + assert(!is_internal || tcache == NULL); + assert(!is_internal || arena == NULL || arena_is_auto(arena)); + if (!tsdn_null(tsdn) && tsd_reentrancy_level_get(tsdn_tsd(tsdn)) == 0) { + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, 0); + } + + ret = arena_malloc(tsdn, arena, size, ind, zero, tcache, slow_path); + if (config_stats && is_internal && likely(ret != NULL)) { + arena_internal_add(iaalloc(tsdn, ret), isalloc(tsdn, ret)); + } + return ret; +} + +JEMALLOC_ALWAYS_INLINE void * +ialloc(tsd_t *tsd, size_t size, szind_t ind, bool zero, bool slow_path) { + return iallocztm(tsd_tsdn(tsd), size, ind, zero, tcache_get(tsd), false, + NULL, slow_path); +} + +JEMALLOC_ALWAYS_INLINE void * +ipallocztm(tsdn_t *tsdn, size_t usize, size_t alignment, bool zero, + tcache_t *tcache, bool is_internal, arena_t *arena) { + void *ret; + + assert(usize != 0); + assert(usize == sz_sa2u(usize, alignment)); + assert(!is_internal || tcache == NULL); + assert(!is_internal || arena == NULL || arena_is_auto(arena)); + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, 0); + + ret = arena_palloc(tsdn, arena, usize, alignment, zero, tcache); + assert(ALIGNMENT_ADDR2BASE(ret, alignment) == ret); + if (config_stats && is_internal && likely(ret != NULL)) { + arena_internal_add(iaalloc(tsdn, ret), isalloc(tsdn, ret)); + } + return ret; +} + +JEMALLOC_ALWAYS_INLINE void * +ipalloct(tsdn_t *tsdn, size_t usize, size_t alignment, bool zero, + tcache_t *tcache, arena_t *arena) { + return ipallocztm(tsdn, usize, alignment, zero, tcache, false, arena); +} + +JEMALLOC_ALWAYS_INLINE void * +ipalloc(tsd_t *tsd, size_t usize, size_t alignment, bool zero) { + return ipallocztm(tsd_tsdn(tsd), usize, alignment, zero, + tcache_get(tsd), false, NULL); +} + +JEMALLOC_ALWAYS_INLINE size_t +ivsalloc(tsdn_t *tsdn, const void *ptr) { + return arena_vsalloc(tsdn, ptr); +} + +JEMALLOC_ALWAYS_INLINE void +idalloctm(tsdn_t *tsdn, void *ptr, tcache_t *tcache, alloc_ctx_t *alloc_ctx, + bool is_internal, bool slow_path) { + assert(ptr != NULL); + assert(!is_internal || tcache == NULL); + assert(!is_internal || arena_is_auto(iaalloc(tsdn, ptr))); + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, 0); + if (config_stats && is_internal) { + arena_internal_sub(iaalloc(tsdn, ptr), isalloc(tsdn, ptr)); + } + if (!is_internal && !tsdn_null(tsdn) && + tsd_reentrancy_level_get(tsdn_tsd(tsdn)) != 0) { + assert(tcache == NULL); + } + arena_dalloc(tsdn, ptr, tcache, alloc_ctx, slow_path); +} + +JEMALLOC_ALWAYS_INLINE void +idalloc(tsd_t *tsd, void *ptr) { + idalloctm(tsd_tsdn(tsd), ptr, tcache_get(tsd), NULL, false, true); +} + +JEMALLOC_ALWAYS_INLINE void +isdalloct(tsdn_t *tsdn, void *ptr, size_t size, tcache_t *tcache, + alloc_ctx_t *alloc_ctx, bool slow_path) { + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, 0); + arena_sdalloc(tsdn, ptr, size, tcache, alloc_ctx, slow_path); +} + +JEMALLOC_ALWAYS_INLINE void * +iralloct_realign(tsdn_t *tsdn, void *ptr, size_t oldsize, size_t size, + size_t alignment, bool zero, tcache_t *tcache, arena_t *arena, + hook_ralloc_args_t *hook_args) { + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, 0); + void *p; + size_t usize, copysize; + + usize = sz_sa2u(size, alignment); + if (unlikely(usize == 0 || usize > SC_LARGE_MAXCLASS)) { + return NULL; + } + p = ipalloct(tsdn, usize, alignment, zero, tcache, arena); + if (p == NULL) { + return NULL; + } + /* + * Copy at most size bytes (not size+extra), since the caller has no + * expectation that the extra bytes will be reliably preserved. + */ + copysize = (size < oldsize) ? size : oldsize; + memcpy(p, ptr, copysize); + hook_invoke_alloc(hook_args->is_realloc + ? hook_alloc_realloc : hook_alloc_rallocx, p, (uintptr_t)p, + hook_args->args); + hook_invoke_dalloc(hook_args->is_realloc + ? hook_dalloc_realloc : hook_dalloc_rallocx, ptr, hook_args->args); + isdalloct(tsdn, ptr, oldsize, tcache, NULL, true); + return p; +} + +/* + * is_realloc threads through the knowledge of whether or not this call comes + * from je_realloc (as opposed to je_rallocx); this ensures that we pass the + * correct entry point into any hooks. + * Note that these functions are all force-inlined, so no actual bool gets + * passed-around anywhere. + */ +JEMALLOC_ALWAYS_INLINE void * +iralloct(tsdn_t *tsdn, void *ptr, size_t oldsize, size_t size, size_t alignment, + bool zero, tcache_t *tcache, arena_t *arena, hook_ralloc_args_t *hook_args) +{ + assert(ptr != NULL); + assert(size != 0); + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, 0); + + if (alignment != 0 && ((uintptr_t)ptr & ((uintptr_t)alignment-1)) + != 0) { + /* + * Existing object alignment is inadequate; allocate new space + * and copy. + */ + return iralloct_realign(tsdn, ptr, oldsize, size, alignment, + zero, tcache, arena, hook_args); + } + + return arena_ralloc(tsdn, arena, ptr, oldsize, size, alignment, zero, + tcache, hook_args); +} + +JEMALLOC_ALWAYS_INLINE void * +iralloc(tsd_t *tsd, void *ptr, size_t oldsize, size_t size, size_t alignment, + bool zero, hook_ralloc_args_t *hook_args) { + return iralloct(tsd_tsdn(tsd), ptr, oldsize, size, alignment, zero, + tcache_get(tsd), NULL, hook_args); +} + +JEMALLOC_ALWAYS_INLINE bool +ixalloc(tsdn_t *tsdn, void *ptr, size_t oldsize, size_t size, size_t extra, + size_t alignment, bool zero, size_t *newsize) { + assert(ptr != NULL); + assert(size != 0); + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, 0); + + if (alignment != 0 && ((uintptr_t)ptr & ((uintptr_t)alignment-1)) + != 0) { + /* Existing object alignment is inadequate. */ + *newsize = oldsize; + return true; + } + + return arena_ralloc_no_move(tsdn, ptr, oldsize, size, extra, zero, + newsize); +} + +#endif /* JEMALLOC_INTERNAL_INLINES_C_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_macros.h b/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_macros.h new file mode 100644 index 000000000..d8ea06f6d --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_macros.h @@ -0,0 +1,114 @@ +#ifndef JEMALLOC_INTERNAL_MACROS_H +#define JEMALLOC_INTERNAL_MACROS_H + +#ifdef JEMALLOC_DEBUG +# define JEMALLOC_ALWAYS_INLINE static inline +#else +# define JEMALLOC_ALWAYS_INLINE JEMALLOC_ATTR(always_inline) static inline +#endif +#ifdef _MSC_VER +# define inline _inline +#endif + +#define UNUSED JEMALLOC_ATTR(unused) + +#define ZU(z) ((size_t)z) +#define ZD(z) ((ssize_t)z) +#define QU(q) ((uint64_t)q) +#define QD(q) ((int64_t)q) + +#define KZU(z) ZU(z##ULL) +#define KZD(z) ZD(z##LL) +#define KQU(q) QU(q##ULL) +#define KQD(q) QI(q##LL) + +#ifndef __DECONST +# define __DECONST(type, var) ((type)(uintptr_t)(const void *)(var)) +#endif + +#if !defined(JEMALLOC_HAS_RESTRICT) || defined(__cplusplus) +# define restrict +#endif + +/* Various function pointers are static and immutable except during testing. */ +#ifdef JEMALLOC_JET +# define JET_MUTABLE +#else +# define JET_MUTABLE const +#endif + +#define JEMALLOC_VA_ARGS_HEAD(head, ...) head +#define JEMALLOC_VA_ARGS_TAIL(head, ...) __VA_ARGS__ + +#if (defined(__GNUC__) || defined(__GNUG__)) && !defined(__clang__) \ + && defined(JEMALLOC_HAVE_ATTR) && (__GNUC__ >= 7) +#define JEMALLOC_FALLTHROUGH JEMALLOC_ATTR(fallthrough); +#else +#define JEMALLOC_FALLTHROUGH /* falls through */ +#endif + +/* Diagnostic suppression macros */ +#if defined(_MSC_VER) && !defined(__clang__) +# define JEMALLOC_DIAGNOSTIC_PUSH __pragma(warning(push)) +# define JEMALLOC_DIAGNOSTIC_POP __pragma(warning(pop)) +# define JEMALLOC_DIAGNOSTIC_IGNORE(W) __pragma(warning(disable:W)) +# define JEMALLOC_DIAGNOSTIC_IGNORE_MISSING_STRUCT_FIELD_INITIALIZERS +# define JEMALLOC_DIAGNOSTIC_IGNORE_TYPE_LIMITS +# define JEMALLOC_DIAGNOSTIC_IGNORE_ALLOC_SIZE_LARGER_THAN +# define JEMALLOC_DIAGNOSTIC_DISABLE_SPURIOUS +/* #pragma GCC diagnostic first appeared in gcc 4.6. */ +#elif (defined(__GNUC__) && ((__GNUC__ > 4) || ((__GNUC__ == 4) && \ + (__GNUC_MINOR__ > 5)))) || defined(__clang__) +/* + * The JEMALLOC_PRAGMA__ macro is an implementation detail of the GCC and Clang + * diagnostic suppression macros and should not be used anywhere else. + */ +# define JEMALLOC_PRAGMA__(X) _Pragma(#X) +# define JEMALLOC_DIAGNOSTIC_PUSH JEMALLOC_PRAGMA__(GCC diagnostic push) +# define JEMALLOC_DIAGNOSTIC_POP JEMALLOC_PRAGMA__(GCC diagnostic pop) +# define JEMALLOC_DIAGNOSTIC_IGNORE(W) \ + JEMALLOC_PRAGMA__(GCC diagnostic ignored W) + +/* + * The -Wmissing-field-initializers warning is buggy in GCC versions < 5.1 and + * all clang versions up to version 7 (currently trunk, unreleased). This macro + * suppresses the warning for the affected compiler versions only. + */ +# if ((defined(__GNUC__) && !defined(__clang__)) && (__GNUC__ < 5)) || \ + defined(__clang__) +# define JEMALLOC_DIAGNOSTIC_IGNORE_MISSING_STRUCT_FIELD_INITIALIZERS \ + JEMALLOC_DIAGNOSTIC_IGNORE("-Wmissing-field-initializers") +# else +# define JEMALLOC_DIAGNOSTIC_IGNORE_MISSING_STRUCT_FIELD_INITIALIZERS +# endif + +# define JEMALLOC_DIAGNOSTIC_IGNORE_TYPE_LIMITS \ + JEMALLOC_DIAGNOSTIC_IGNORE("-Wtype-limits") +# define JEMALLOC_DIAGNOSTIC_IGNORE_UNUSED_PARAMETER \ + JEMALLOC_DIAGNOSTIC_IGNORE("-Wunused-parameter") +# if defined(__GNUC__) && !defined(__clang__) && (__GNUC__ >= 7) +# define JEMALLOC_DIAGNOSTIC_IGNORE_ALLOC_SIZE_LARGER_THAN \ + JEMALLOC_DIAGNOSTIC_IGNORE("-Walloc-size-larger-than=") +# else +# define JEMALLOC_DIAGNOSTIC_IGNORE_ALLOC_SIZE_LARGER_THAN +# endif +# define JEMALLOC_DIAGNOSTIC_DISABLE_SPURIOUS \ + JEMALLOC_DIAGNOSTIC_PUSH \ + JEMALLOC_DIAGNOSTIC_IGNORE_UNUSED_PARAMETER +#else +# define JEMALLOC_DIAGNOSTIC_PUSH +# define JEMALLOC_DIAGNOSTIC_POP +# define JEMALLOC_DIAGNOSTIC_IGNORE(W) +# define JEMALLOC_DIAGNOSTIC_IGNORE_MISSING_STRUCT_FIELD_INITIALIZERS +# define JEMALLOC_DIAGNOSTIC_IGNORE_TYPE_LIMITS +# define JEMALLOC_DIAGNOSTIC_IGNORE_ALLOC_SIZE_LARGER_THAN +# define JEMALLOC_DIAGNOSTIC_DISABLE_SPURIOUS +#endif + +/* + * Disables spurious diagnostics for all headers. Since these headers are not + * included by users directly, it does not affect their diagnostic settings. + */ +JEMALLOC_DIAGNOSTIC_DISABLE_SPURIOUS + +#endif /* JEMALLOC_INTERNAL_MACROS_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_types.h b/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_types.h new file mode 100644 index 000000000..e296c5a7e --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_types.h @@ -0,0 +1,114 @@ +#ifndef JEMALLOC_INTERNAL_TYPES_H +#define JEMALLOC_INTERNAL_TYPES_H + +#include "jemalloc/internal/quantum.h" + +/* Page size index type. */ +typedef unsigned pszind_t; + +/* Size class index type. */ +typedef unsigned szind_t; + +/* Processor / core id type. */ +typedef int malloc_cpuid_t; + +/* + * Flags bits: + * + * a: arena + * t: tcache + * 0: unused + * z: zero + * n: alignment + * + * aaaaaaaa aaaatttt tttttttt 0znnnnnn + */ +#define MALLOCX_ARENA_BITS 12 +#define MALLOCX_TCACHE_BITS 12 +#define MALLOCX_LG_ALIGN_BITS 6 +#define MALLOCX_ARENA_SHIFT 20 +#define MALLOCX_TCACHE_SHIFT 8 +#define MALLOCX_ARENA_MASK \ + (((1 << MALLOCX_ARENA_BITS) - 1) << MALLOCX_ARENA_SHIFT) +/* NB: Arena index bias decreases the maximum number of arenas by 1. */ +#define MALLOCX_ARENA_LIMIT ((1 << MALLOCX_ARENA_BITS) - 1) +#define MALLOCX_TCACHE_MASK \ + (((1 << MALLOCX_TCACHE_BITS) - 1) << MALLOCX_TCACHE_SHIFT) +#define MALLOCX_TCACHE_MAX ((1 << MALLOCX_TCACHE_BITS) - 3) +#define MALLOCX_LG_ALIGN_MASK ((1 << MALLOCX_LG_ALIGN_BITS) - 1) +/* Use MALLOCX_ALIGN_GET() if alignment may not be specified in flags. */ +#define MALLOCX_ALIGN_GET_SPECIFIED(flags) \ + (ZU(1) << (flags & MALLOCX_LG_ALIGN_MASK)) +#define MALLOCX_ALIGN_GET(flags) \ + (MALLOCX_ALIGN_GET_SPECIFIED(flags) & (SIZE_T_MAX-1)) +#define MALLOCX_ZERO_GET(flags) \ + ((bool)(flags & MALLOCX_ZERO)) + +#define MALLOCX_TCACHE_GET(flags) \ + (((unsigned)((flags & MALLOCX_TCACHE_MASK) >> MALLOCX_TCACHE_SHIFT)) - 2) +#define MALLOCX_ARENA_GET(flags) \ + (((unsigned)(((unsigned)flags) >> MALLOCX_ARENA_SHIFT)) - 1) + +/* Smallest size class to support. */ +#define TINY_MIN (1U << LG_TINY_MIN) + +#define LONG ((size_t)(1U << LG_SIZEOF_LONG)) +#define LONG_MASK (LONG - 1) + +/* Return the smallest long multiple that is >= a. */ +#define LONG_CEILING(a) \ + (((a) + LONG_MASK) & ~LONG_MASK) + +#define SIZEOF_PTR (1U << LG_SIZEOF_PTR) +#define PTR_MASK (SIZEOF_PTR - 1) + +/* Return the smallest (void *) multiple that is >= a. */ +#define PTR_CEILING(a) \ + (((a) + PTR_MASK) & ~PTR_MASK) + +/* + * Maximum size of L1 cache line. This is used to avoid cache line aliasing. + * In addition, this controls the spacing of cacheline-spaced size classes. + * + * CACHELINE cannot be based on LG_CACHELINE because __declspec(align()) can + * only handle raw constants. + */ +#define LG_CACHELINE 6 +#define CACHELINE 64 +#define CACHELINE_MASK (CACHELINE - 1) + +/* Return the smallest cacheline multiple that is >= s. */ +#define CACHELINE_CEILING(s) \ + (((s) + CACHELINE_MASK) & ~CACHELINE_MASK) + +/* Return the nearest aligned address at or below a. */ +#define ALIGNMENT_ADDR2BASE(a, alignment) \ + ((void *)((uintptr_t)(a) & ((~(alignment)) + 1))) + +/* Return the offset between a and the nearest aligned address at or below a. */ +#define ALIGNMENT_ADDR2OFFSET(a, alignment) \ + ((size_t)((uintptr_t)(a) & (alignment - 1))) + +/* Return the smallest alignment multiple that is >= s. */ +#define ALIGNMENT_CEILING(s, alignment) \ + (((s) + (alignment - 1)) & ((~(alignment)) + 1)) + +/* Declare a variable-length array. */ +#if __STDC_VERSION__ < 199901L +# ifdef _MSC_VER +# include <malloc.h> +# define alloca _alloca +# else +# ifdef JEMALLOC_HAS_ALLOCA_H +# include <alloca.h> +# else +# include <stdlib.h> +# endif +# endif +# define VARIABLE_ARRAY(type, name, count) \ + type *name = alloca(sizeof(type) * (count)) +#else +# define VARIABLE_ARRAY(type, name, count) type name[(count)] +#endif + +#endif /* JEMALLOC_INTERNAL_TYPES_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/jemalloc_preamble.h.in b/deps/jemalloc/include/jemalloc/internal/jemalloc_preamble.h.in new file mode 100644 index 000000000..3418cbfa2 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/jemalloc_preamble.h.in @@ -0,0 +1,213 @@ +#ifndef JEMALLOC_PREAMBLE_H +#define JEMALLOC_PREAMBLE_H + +#include "jemalloc_internal_defs.h" +#include "jemalloc/internal/jemalloc_internal_decls.h" + +#ifdef JEMALLOC_UTRACE +#include <sys/ktrace.h> +#endif + +#define JEMALLOC_NO_DEMANGLE +#ifdef JEMALLOC_JET +# undef JEMALLOC_IS_MALLOC +# define JEMALLOC_N(n) jet_##n +# include "jemalloc/internal/public_namespace.h" +# define JEMALLOC_NO_RENAME +# include "../jemalloc@install_suffix@.h" +# undef JEMALLOC_NO_RENAME +#else +# define JEMALLOC_N(n) @private_namespace@##n +# include "../jemalloc@install_suffix@.h" +#endif + +#if defined(JEMALLOC_OSATOMIC) +#include <libkern/OSAtomic.h> +#endif + +#ifdef JEMALLOC_ZONE +#include <mach/mach_error.h> +#include <mach/mach_init.h> +#include <mach/vm_map.h> +#endif + +#include "jemalloc/internal/jemalloc_internal_macros.h" + +/* + * Note that the ordering matters here; the hook itself is name-mangled. We + * want the inclusion of hooks to happen early, so that we hook as much as + * possible. + */ +#ifndef JEMALLOC_NO_PRIVATE_NAMESPACE +# ifndef JEMALLOC_JET +# include "jemalloc/internal/private_namespace.h" +# else +# include "jemalloc/internal/private_namespace_jet.h" +# endif +#endif +#include "jemalloc/internal/test_hooks.h" + +#ifdef JEMALLOC_DEFINE_MADVISE_FREE +# define JEMALLOC_MADV_FREE 8 +#endif + +static const bool config_debug = +#ifdef JEMALLOC_DEBUG + true +#else + false +#endif + ; +static const bool have_dss = +#ifdef JEMALLOC_DSS + true +#else + false +#endif + ; +static const bool have_madvise_huge = +#ifdef JEMALLOC_HAVE_MADVISE_HUGE + true +#else + false +#endif + ; +static const bool config_fill = +#ifdef JEMALLOC_FILL + true +#else + false +#endif + ; +static const bool config_lazy_lock = +#ifdef JEMALLOC_LAZY_LOCK + true +#else + false +#endif + ; +static const char * const config_malloc_conf = JEMALLOC_CONFIG_MALLOC_CONF; +static const bool config_prof = +#ifdef JEMALLOC_PROF + true +#else + false +#endif + ; +static const bool config_prof_libgcc = +#ifdef JEMALLOC_PROF_LIBGCC + true +#else + false +#endif + ; +static const bool config_prof_libunwind = +#ifdef JEMALLOC_PROF_LIBUNWIND + true +#else + false +#endif + ; +static const bool maps_coalesce = +#ifdef JEMALLOC_MAPS_COALESCE + true +#else + false +#endif + ; +static const bool config_stats = +#ifdef JEMALLOC_STATS + true +#else + false +#endif + ; +static const bool config_tls = +#ifdef JEMALLOC_TLS + true +#else + false +#endif + ; +static const bool config_utrace = +#ifdef JEMALLOC_UTRACE + true +#else + false +#endif + ; +static const bool config_xmalloc = +#ifdef JEMALLOC_XMALLOC + true +#else + false +#endif + ; +static const bool config_cache_oblivious = +#ifdef JEMALLOC_CACHE_OBLIVIOUS + true +#else + false +#endif + ; +/* + * Undocumented, for jemalloc development use only at the moment. See the note + * in jemalloc/internal/log.h. + */ +static const bool config_log = +#ifdef JEMALLOC_LOG + true +#else + false +#endif + ; +/* + * Are extra safety checks enabled; things like checking the size of sized + * deallocations, double-frees, etc. + */ +static const bool config_opt_safety_checks = +#ifdef JEMALLOC_OPT_SAFETY_CHECKS + true +#elif defined(JEMALLOC_DEBUG) + /* + * This lets us only guard safety checks by one flag instead of two; fast + * checks can guard solely by config_opt_safety_checks and run in debug mode + * too. + */ + true +#else + false +#endif + ; + +#if defined(_WIN32) || defined(JEMALLOC_HAVE_SCHED_GETCPU) +/* Currently percpu_arena depends on sched_getcpu. */ +#define JEMALLOC_PERCPU_ARENA +#endif +static const bool have_percpu_arena = +#ifdef JEMALLOC_PERCPU_ARENA + true +#else + false +#endif + ; +/* + * Undocumented, and not recommended; the application should take full + * responsibility for tracking provenance. + */ +static const bool force_ivsalloc = +#ifdef JEMALLOC_FORCE_IVSALLOC + true +#else + false +#endif + ; +static const bool have_background_thread = +#ifdef JEMALLOC_BACKGROUND_THREAD + true +#else + false +#endif + ; + +#endif /* JEMALLOC_PREAMBLE_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/large_externs.h b/deps/jemalloc/include/jemalloc/internal/large_externs.h new file mode 100644 index 000000000..a05019e8a --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/large_externs.h @@ -0,0 +1,32 @@ +#ifndef JEMALLOC_INTERNAL_LARGE_EXTERNS_H +#define JEMALLOC_INTERNAL_LARGE_EXTERNS_H + +#include "jemalloc/internal/hook.h" + +void *large_malloc(tsdn_t *tsdn, arena_t *arena, size_t usize, bool zero); +void *large_palloc(tsdn_t *tsdn, arena_t *arena, size_t usize, size_t alignment, + bool zero); +bool large_ralloc_no_move(tsdn_t *tsdn, extent_t *extent, size_t usize_min, + size_t usize_max, bool zero); +void *large_ralloc(tsdn_t *tsdn, arena_t *arena, void *ptr, size_t usize, + size_t alignment, bool zero, tcache_t *tcache, + hook_ralloc_args_t *hook_args); + +typedef void (large_dalloc_junk_t)(void *, size_t); +extern large_dalloc_junk_t *JET_MUTABLE large_dalloc_junk; + +typedef void (large_dalloc_maybe_junk_t)(void *, size_t); +extern large_dalloc_maybe_junk_t *JET_MUTABLE large_dalloc_maybe_junk; + +void large_dalloc_prep_junked_locked(tsdn_t *tsdn, extent_t *extent); +void large_dalloc_finish(tsdn_t *tsdn, extent_t *extent); +void large_dalloc(tsdn_t *tsdn, extent_t *extent); +size_t large_salloc(tsdn_t *tsdn, const extent_t *extent); +prof_tctx_t *large_prof_tctx_get(tsdn_t *tsdn, const extent_t *extent); +void large_prof_tctx_set(tsdn_t *tsdn, extent_t *extent, prof_tctx_t *tctx); +void large_prof_tctx_reset(tsdn_t *tsdn, extent_t *extent); + +nstime_t large_prof_alloc_time_get(const extent_t *extent); +void large_prof_alloc_time_set(extent_t *extent, nstime_t time); + +#endif /* JEMALLOC_INTERNAL_LARGE_EXTERNS_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/log.h b/deps/jemalloc/include/jemalloc/internal/log.h new file mode 100644 index 000000000..642085863 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/log.h @@ -0,0 +1,115 @@ +#ifndef JEMALLOC_INTERNAL_LOG_H +#define JEMALLOC_INTERNAL_LOG_H + +#include "jemalloc/internal/atomic.h" +#include "jemalloc/internal/malloc_io.h" +#include "jemalloc/internal/mutex.h" + +#ifdef JEMALLOC_LOG +# define JEMALLOC_LOG_VAR_BUFSIZE 1000 +#else +# define JEMALLOC_LOG_VAR_BUFSIZE 1 +#endif + +#define JEMALLOC_LOG_BUFSIZE 4096 + +/* + * The log malloc_conf option is a '|'-delimited list of log_var name segments + * which should be logged. The names are themselves hierarchical, with '.' as + * the delimiter (a "segment" is just a prefix in the log namespace). So, if + * you have: + * + * log("arena", "log msg for arena"); // 1 + * log("arena.a", "log msg for arena.a"); // 2 + * log("arena.b", "log msg for arena.b"); // 3 + * log("arena.a.a", "log msg for arena.a.a"); // 4 + * log("extent.a", "log msg for extent.a"); // 5 + * log("extent.b", "log msg for extent.b"); // 6 + * + * And your malloc_conf option is "log=arena.a|extent", then lines 2, 4, 5, and + * 6 will print at runtime. You can enable logging from all log vars by + * writing "log=.". + * + * None of this should be regarded as a stable API for right now. It's intended + * as a debugging interface, to let us keep around some of our printf-debugging + * statements. + */ + +extern char log_var_names[JEMALLOC_LOG_VAR_BUFSIZE]; +extern atomic_b_t log_init_done; + +typedef struct log_var_s log_var_t; +struct log_var_s { + /* + * Lowest bit is "inited", second lowest is "enabled". Putting them in + * a single word lets us avoid any fences on weak architectures. + */ + atomic_u_t state; + const char *name; +}; + +#define LOG_NOT_INITIALIZED 0U +#define LOG_INITIALIZED_NOT_ENABLED 1U +#define LOG_ENABLED 2U + +#define LOG_VAR_INIT(name_str) {ATOMIC_INIT(LOG_NOT_INITIALIZED), name_str} + +/* + * Returns the value we should assume for state (which is not necessarily + * accurate; if logging is done before logging has finished initializing, then + * we default to doing the safe thing by logging everything). + */ +unsigned log_var_update_state(log_var_t *log_var); + +/* We factor out the metadata management to allow us to test more easily. */ +#define log_do_begin(log_var) \ +if (config_log) { \ + unsigned log_state = atomic_load_u(&(log_var).state, \ + ATOMIC_RELAXED); \ + if (unlikely(log_state == LOG_NOT_INITIALIZED)) { \ + log_state = log_var_update_state(&(log_var)); \ + assert(log_state != LOG_NOT_INITIALIZED); \ + } \ + if (log_state == LOG_ENABLED) { \ + { + /* User code executes here. */ +#define log_do_end(log_var) \ + } \ + } \ +} + +/* + * MSVC has some preprocessor bugs in its expansion of __VA_ARGS__ during + * preprocessing. To work around this, we take all potential extra arguments in + * a var-args functions. Since a varargs macro needs at least one argument in + * the "...", we accept the format string there, and require that the first + * argument in this "..." is a const char *. + */ +static inline void +log_impl_varargs(const char *name, ...) { + char buf[JEMALLOC_LOG_BUFSIZE]; + va_list ap; + + va_start(ap, name); + const char *format = va_arg(ap, const char *); + size_t dst_offset = 0; + dst_offset += malloc_snprintf(buf, JEMALLOC_LOG_BUFSIZE, "%s: ", name); + dst_offset += malloc_vsnprintf(buf + dst_offset, + JEMALLOC_LOG_BUFSIZE - dst_offset, format, ap); + dst_offset += malloc_snprintf(buf + dst_offset, + JEMALLOC_LOG_BUFSIZE - dst_offset, "\n"); + va_end(ap); + + malloc_write(buf); +} + +/* Call as log("log.var.str", "format_string %d", arg_for_format_string); */ +#define LOG(log_var_str, ...) \ +do { \ + static log_var_t log_var = LOG_VAR_INIT(log_var_str); \ + log_do_begin(log_var) \ + log_impl_varargs((log_var).name, __VA_ARGS__); \ + log_do_end(log_var) \ +} while (0) + +#endif /* JEMALLOC_INTERNAL_LOG_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/malloc_io.h b/deps/jemalloc/include/jemalloc/internal/malloc_io.h new file mode 100644 index 000000000..1d1a414e0 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/malloc_io.h @@ -0,0 +1,102 @@ +#ifndef JEMALLOC_INTERNAL_MALLOC_IO_H +#define JEMALLOC_INTERNAL_MALLOC_IO_H + +#ifdef _WIN32 +# ifdef _WIN64 +# define FMT64_PREFIX "ll" +# define FMTPTR_PREFIX "ll" +# else +# define FMT64_PREFIX "ll" +# define FMTPTR_PREFIX "" +# endif +# define FMTd32 "d" +# define FMTu32 "u" +# define FMTx32 "x" +# define FMTd64 FMT64_PREFIX "d" +# define FMTu64 FMT64_PREFIX "u" +# define FMTx64 FMT64_PREFIX "x" +# define FMTdPTR FMTPTR_PREFIX "d" +# define FMTuPTR FMTPTR_PREFIX "u" +# define FMTxPTR FMTPTR_PREFIX "x" +#else +# include <inttypes.h> +# define FMTd32 PRId32 +# define FMTu32 PRIu32 +# define FMTx32 PRIx32 +# define FMTd64 PRId64 +# define FMTu64 PRIu64 +# define FMTx64 PRIx64 +# define FMTdPTR PRIdPTR +# define FMTuPTR PRIuPTR +# define FMTxPTR PRIxPTR +#endif + +/* Size of stack-allocated buffer passed to buferror(). */ +#define BUFERROR_BUF 64 + +/* + * Size of stack-allocated buffer used by malloc_{,v,vc}printf(). This must be + * large enough for all possible uses within jemalloc. + */ +#define MALLOC_PRINTF_BUFSIZE 4096 + +int buferror(int err, char *buf, size_t buflen); +uintmax_t malloc_strtoumax(const char *restrict nptr, char **restrict endptr, + int base); +void malloc_write(const char *s); + +/* + * malloc_vsnprintf() supports a subset of snprintf(3) that avoids floating + * point math. + */ +size_t malloc_vsnprintf(char *str, size_t size, const char *format, + va_list ap); +size_t malloc_snprintf(char *str, size_t size, const char *format, ...) + JEMALLOC_FORMAT_PRINTF(3, 4); +/* + * The caller can set write_cb to null to choose to print with the + * je_malloc_message hook. + */ +void malloc_vcprintf(void (*write_cb)(void *, const char *), void *cbopaque, + const char *format, va_list ap); +void malloc_cprintf(void (*write_cb)(void *, const char *), void *cbopaque, + const char *format, ...) JEMALLOC_FORMAT_PRINTF(3, 4); +void malloc_printf(const char *format, ...) JEMALLOC_FORMAT_PRINTF(1, 2); + +static inline ssize_t +malloc_write_fd(int fd, const void *buf, size_t count) { +#if defined(JEMALLOC_USE_SYSCALL) && defined(SYS_write) + /* + * Use syscall(2) rather than write(2) when possible in order to avoid + * the possibility of memory allocation within libc. This is necessary + * on FreeBSD; most operating systems do not have this problem though. + * + * syscall() returns long or int, depending on platform, so capture the + * result in the widest plausible type to avoid compiler warnings. + */ + long result = syscall(SYS_write, fd, buf, count); +#else + ssize_t result = (ssize_t)write(fd, buf, +#ifdef _WIN32 + (unsigned int) +#endif + count); +#endif + return (ssize_t)result; +} + +static inline ssize_t +malloc_read_fd(int fd, void *buf, size_t count) { +#if defined(JEMALLOC_USE_SYSCALL) && defined(SYS_read) + long result = syscall(SYS_read, fd, buf, count); +#else + ssize_t result = read(fd, buf, +#ifdef _WIN32 + (unsigned int) +#endif + count); +#endif + return (ssize_t)result; +} + +#endif /* JEMALLOC_INTERNAL_MALLOC_IO_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/mutex.h b/deps/jemalloc/include/jemalloc/internal/mutex.h new file mode 100644 index 000000000..7c24f0725 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/mutex.h @@ -0,0 +1,288 @@ +#ifndef JEMALLOC_INTERNAL_MUTEX_H +#define JEMALLOC_INTERNAL_MUTEX_H + +#include "jemalloc/internal/atomic.h" +#include "jemalloc/internal/mutex_prof.h" +#include "jemalloc/internal/tsd.h" +#include "jemalloc/internal/witness.h" + +typedef enum { + /* Can only acquire one mutex of a given witness rank at a time. */ + malloc_mutex_rank_exclusive, + /* + * Can acquire multiple mutexes of the same witness rank, but in + * address-ascending order only. + */ + malloc_mutex_address_ordered +} malloc_mutex_lock_order_t; + +typedef struct malloc_mutex_s malloc_mutex_t; +struct malloc_mutex_s { + union { + struct { + /* + * prof_data is defined first to reduce cacheline + * bouncing: the data is not touched by the mutex holder + * during unlocking, while might be modified by + * contenders. Having it before the mutex itself could + * avoid prefetching a modified cacheline (for the + * unlocking thread). + */ + mutex_prof_data_t prof_data; +#ifdef _WIN32 +# if _WIN32_WINNT >= 0x0600 + SRWLOCK lock; +# else + CRITICAL_SECTION lock; +# endif +#elif (defined(JEMALLOC_OS_UNFAIR_LOCK)) + os_unfair_lock lock; +#elif (defined(JEMALLOC_MUTEX_INIT_CB)) + pthread_mutex_t lock; + malloc_mutex_t *postponed_next; +#else + pthread_mutex_t lock; +#endif + /* + * Hint flag to avoid exclusive cache line contention + * during spin waiting + */ + atomic_b_t locked; + }; + /* + * We only touch witness when configured w/ debug. However we + * keep the field in a union when !debug so that we don't have + * to pollute the code base with #ifdefs, while avoid paying the + * memory cost. + */ +#if !defined(JEMALLOC_DEBUG) + witness_t witness; + malloc_mutex_lock_order_t lock_order; +#endif + }; + +#if defined(JEMALLOC_DEBUG) + witness_t witness; + malloc_mutex_lock_order_t lock_order; +#endif +}; + +/* + * Based on benchmark results, a fixed spin with this amount of retries works + * well for our critical sections. + */ +#define MALLOC_MUTEX_MAX_SPIN 250 + +#ifdef _WIN32 +# if _WIN32_WINNT >= 0x0600 +# define MALLOC_MUTEX_LOCK(m) AcquireSRWLockExclusive(&(m)->lock) +# define MALLOC_MUTEX_UNLOCK(m) ReleaseSRWLockExclusive(&(m)->lock) +# define MALLOC_MUTEX_TRYLOCK(m) (!TryAcquireSRWLockExclusive(&(m)->lock)) +# else +# define MALLOC_MUTEX_LOCK(m) EnterCriticalSection(&(m)->lock) +# define MALLOC_MUTEX_UNLOCK(m) LeaveCriticalSection(&(m)->lock) +# define MALLOC_MUTEX_TRYLOCK(m) (!TryEnterCriticalSection(&(m)->lock)) +# endif +#elif (defined(JEMALLOC_OS_UNFAIR_LOCK)) +# define MALLOC_MUTEX_LOCK(m) os_unfair_lock_lock(&(m)->lock) +# define MALLOC_MUTEX_UNLOCK(m) os_unfair_lock_unlock(&(m)->lock) +# define MALLOC_MUTEX_TRYLOCK(m) (!os_unfair_lock_trylock(&(m)->lock)) +#else +# define MALLOC_MUTEX_LOCK(m) pthread_mutex_lock(&(m)->lock) +# define MALLOC_MUTEX_UNLOCK(m) pthread_mutex_unlock(&(m)->lock) +# define MALLOC_MUTEX_TRYLOCK(m) (pthread_mutex_trylock(&(m)->lock) != 0) +#endif + +#define LOCK_PROF_DATA_INITIALIZER \ + {NSTIME_ZERO_INITIALIZER, NSTIME_ZERO_INITIALIZER, 0, 0, 0, \ + ATOMIC_INIT(0), 0, NULL, 0} + +#ifdef _WIN32 +# define MALLOC_MUTEX_INITIALIZER +#elif (defined(JEMALLOC_OS_UNFAIR_LOCK)) +# if defined(JEMALLOC_DEBUG) +# define MALLOC_MUTEX_INITIALIZER \ + {{{LOCK_PROF_DATA_INITIALIZER, OS_UNFAIR_LOCK_INIT, ATOMIC_INIT(false)}}, \ + WITNESS_INITIALIZER("mutex", WITNESS_RANK_OMIT), 0} +# else +# define MALLOC_MUTEX_INITIALIZER \ + {{{LOCK_PROF_DATA_INITIALIZER, OS_UNFAIR_LOCK_INIT, ATOMIC_INIT(false)}}, \ + WITNESS_INITIALIZER("mutex", WITNESS_RANK_OMIT)} +# endif +#elif (defined(JEMALLOC_MUTEX_INIT_CB)) +# if (defined(JEMALLOC_DEBUG)) +# define MALLOC_MUTEX_INITIALIZER \ + {{{LOCK_PROF_DATA_INITIALIZER, PTHREAD_MUTEX_INITIALIZER, NULL, ATOMIC_INIT(false)}}, \ + WITNESS_INITIALIZER("mutex", WITNESS_RANK_OMIT), 0} +# else +# define MALLOC_MUTEX_INITIALIZER \ + {{{LOCK_PROF_DATA_INITIALIZER, PTHREAD_MUTEX_INITIALIZER, NULL, ATOMIC_INIT(false)}}, \ + WITNESS_INITIALIZER("mutex", WITNESS_RANK_OMIT)} +# endif + +#else +# define MALLOC_MUTEX_TYPE PTHREAD_MUTEX_DEFAULT +# if defined(JEMALLOC_DEBUG) +# define MALLOC_MUTEX_INITIALIZER \ + {{{LOCK_PROF_DATA_INITIALIZER, PTHREAD_MUTEX_INITIALIZER, ATOMIC_INIT(false)}}, \ + WITNESS_INITIALIZER("mutex", WITNESS_RANK_OMIT), 0} +# else +# define MALLOC_MUTEX_INITIALIZER \ + {{{LOCK_PROF_DATA_INITIALIZER, PTHREAD_MUTEX_INITIALIZER, ATOMIC_INIT(false)}}, \ + WITNESS_INITIALIZER("mutex", WITNESS_RANK_OMIT)} +# endif +#endif + +#ifdef JEMALLOC_LAZY_LOCK +extern bool isthreaded; +#else +# undef isthreaded /* Undo private_namespace.h definition. */ +# define isthreaded true +#endif + +bool malloc_mutex_init(malloc_mutex_t *mutex, const char *name, + witness_rank_t rank, malloc_mutex_lock_order_t lock_order); +void malloc_mutex_prefork(tsdn_t *tsdn, malloc_mutex_t *mutex); +void malloc_mutex_postfork_parent(tsdn_t *tsdn, malloc_mutex_t *mutex); +void malloc_mutex_postfork_child(tsdn_t *tsdn, malloc_mutex_t *mutex); +bool malloc_mutex_boot(void); +void malloc_mutex_prof_data_reset(tsdn_t *tsdn, malloc_mutex_t *mutex); + +void malloc_mutex_lock_slow(malloc_mutex_t *mutex); + +static inline void +malloc_mutex_lock_final(malloc_mutex_t *mutex) { + MALLOC_MUTEX_LOCK(mutex); + atomic_store_b(&mutex->locked, true, ATOMIC_RELAXED); +} + +static inline bool +malloc_mutex_trylock_final(malloc_mutex_t *mutex) { + return MALLOC_MUTEX_TRYLOCK(mutex); +} + +static inline void +mutex_owner_stats_update(tsdn_t *tsdn, malloc_mutex_t *mutex) { + if (config_stats) { + mutex_prof_data_t *data = &mutex->prof_data; + data->n_lock_ops++; + if (data->prev_owner != tsdn) { + data->prev_owner = tsdn; + data->n_owner_switches++; + } + } +} + +/* Trylock: return false if the lock is successfully acquired. */ +static inline bool +malloc_mutex_trylock(tsdn_t *tsdn, malloc_mutex_t *mutex) { + witness_assert_not_owner(tsdn_witness_tsdp_get(tsdn), &mutex->witness); + if (isthreaded) { + if (malloc_mutex_trylock_final(mutex)) { + atomic_store_b(&mutex->locked, true, ATOMIC_RELAXED); + return true; + } + mutex_owner_stats_update(tsdn, mutex); + } + witness_lock(tsdn_witness_tsdp_get(tsdn), &mutex->witness); + + return false; +} + +/* Aggregate lock prof data. */ +static inline void +malloc_mutex_prof_merge(mutex_prof_data_t *sum, mutex_prof_data_t *data) { + nstime_add(&sum->tot_wait_time, &data->tot_wait_time); + if (nstime_compare(&sum->max_wait_time, &data->max_wait_time) < 0) { + nstime_copy(&sum->max_wait_time, &data->max_wait_time); + } + + sum->n_wait_times += data->n_wait_times; + sum->n_spin_acquired += data->n_spin_acquired; + + if (sum->max_n_thds < data->max_n_thds) { + sum->max_n_thds = data->max_n_thds; + } + uint32_t cur_n_waiting_thds = atomic_load_u32(&sum->n_waiting_thds, + ATOMIC_RELAXED); + uint32_t new_n_waiting_thds = cur_n_waiting_thds + atomic_load_u32( + &data->n_waiting_thds, ATOMIC_RELAXED); + atomic_store_u32(&sum->n_waiting_thds, new_n_waiting_thds, + ATOMIC_RELAXED); + sum->n_owner_switches += data->n_owner_switches; + sum->n_lock_ops += data->n_lock_ops; +} + +static inline void +malloc_mutex_lock(tsdn_t *tsdn, malloc_mutex_t *mutex) { + witness_assert_not_owner(tsdn_witness_tsdp_get(tsdn), &mutex->witness); + if (isthreaded) { + if (malloc_mutex_trylock_final(mutex)) { + malloc_mutex_lock_slow(mutex); + atomic_store_b(&mutex->locked, true, ATOMIC_RELAXED); + } + mutex_owner_stats_update(tsdn, mutex); + } + witness_lock(tsdn_witness_tsdp_get(tsdn), &mutex->witness); +} + +static inline void +malloc_mutex_unlock(tsdn_t *tsdn, malloc_mutex_t *mutex) { + atomic_store_b(&mutex->locked, false, ATOMIC_RELAXED); + witness_unlock(tsdn_witness_tsdp_get(tsdn), &mutex->witness); + if (isthreaded) { + MALLOC_MUTEX_UNLOCK(mutex); + } +} + +static inline void +malloc_mutex_assert_owner(tsdn_t *tsdn, malloc_mutex_t *mutex) { + witness_assert_owner(tsdn_witness_tsdp_get(tsdn), &mutex->witness); +} + +static inline void +malloc_mutex_assert_not_owner(tsdn_t *tsdn, malloc_mutex_t *mutex) { + witness_assert_not_owner(tsdn_witness_tsdp_get(tsdn), &mutex->witness); +} + +/* Copy the prof data from mutex for processing. */ +static inline void +malloc_mutex_prof_read(tsdn_t *tsdn, mutex_prof_data_t *data, + malloc_mutex_t *mutex) { + mutex_prof_data_t *source = &mutex->prof_data; + /* Can only read holding the mutex. */ + malloc_mutex_assert_owner(tsdn, mutex); + + /* + * Not *really* allowed (we shouldn't be doing non-atomic loads of + * atomic data), but the mutex protection makes this safe, and writing + * a member-for-member copy is tedious for this situation. + */ + *data = *source; + /* n_wait_thds is not reported (modified w/o locking). */ + atomic_store_u32(&data->n_waiting_thds, 0, ATOMIC_RELAXED); +} + +static inline void +malloc_mutex_prof_accum(tsdn_t *tsdn, mutex_prof_data_t *data, + malloc_mutex_t *mutex) { + mutex_prof_data_t *source = &mutex->prof_data; + /* Can only read holding the mutex. */ + malloc_mutex_assert_owner(tsdn, mutex); + + nstime_add(&data->tot_wait_time, &source->tot_wait_time); + if (nstime_compare(&source->max_wait_time, &data->max_wait_time) > 0) { + nstime_copy(&data->max_wait_time, &source->max_wait_time); + } + data->n_wait_times += source->n_wait_times; + data->n_spin_acquired += source->n_spin_acquired; + if (data->max_n_thds < source->max_n_thds) { + data->max_n_thds = source->max_n_thds; + } + /* n_wait_thds is not reported. */ + atomic_store_u32(&data->n_waiting_thds, 0, ATOMIC_RELAXED); + data->n_owner_switches += source->n_owner_switches; + data->n_lock_ops += source->n_lock_ops; +} + +#endif /* JEMALLOC_INTERNAL_MUTEX_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/mutex_pool.h b/deps/jemalloc/include/jemalloc/internal/mutex_pool.h new file mode 100644 index 000000000..726cece90 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/mutex_pool.h @@ -0,0 +1,94 @@ +#ifndef JEMALLOC_INTERNAL_MUTEX_POOL_H +#define JEMALLOC_INTERNAL_MUTEX_POOL_H + +#include "jemalloc/internal/hash.h" +#include "jemalloc/internal/mutex.h" +#include "jemalloc/internal/witness.h" + +/* We do mod reductions by this value, so it should be kept a power of 2. */ +#define MUTEX_POOL_SIZE 256 + +typedef struct mutex_pool_s mutex_pool_t; +struct mutex_pool_s { + malloc_mutex_t mutexes[MUTEX_POOL_SIZE]; +}; + +bool mutex_pool_init(mutex_pool_t *pool, const char *name, witness_rank_t rank); + +/* Internal helper - not meant to be called outside this module. */ +static inline malloc_mutex_t * +mutex_pool_mutex(mutex_pool_t *pool, uintptr_t key) { + size_t hash_result[2]; + hash(&key, sizeof(key), 0xd50dcc1b, hash_result); + return &pool->mutexes[hash_result[0] % MUTEX_POOL_SIZE]; +} + +static inline void +mutex_pool_assert_not_held(tsdn_t *tsdn, mutex_pool_t *pool) { + for (int i = 0; i < MUTEX_POOL_SIZE; i++) { + malloc_mutex_assert_not_owner(tsdn, &pool->mutexes[i]); + } +} + +/* + * Note that a mutex pool doesn't work exactly the way an embdedded mutex would. + * You're not allowed to acquire mutexes in the pool one at a time. You have to + * acquire all the mutexes you'll need in a single function call, and then + * release them all in a single function call. + */ + +static inline void +mutex_pool_lock(tsdn_t *tsdn, mutex_pool_t *pool, uintptr_t key) { + mutex_pool_assert_not_held(tsdn, pool); + + malloc_mutex_t *mutex = mutex_pool_mutex(pool, key); + malloc_mutex_lock(tsdn, mutex); +} + +static inline void +mutex_pool_unlock(tsdn_t *tsdn, mutex_pool_t *pool, uintptr_t key) { + malloc_mutex_t *mutex = mutex_pool_mutex(pool, key); + malloc_mutex_unlock(tsdn, mutex); + + mutex_pool_assert_not_held(tsdn, pool); +} + +static inline void +mutex_pool_lock2(tsdn_t *tsdn, mutex_pool_t *pool, uintptr_t key1, + uintptr_t key2) { + mutex_pool_assert_not_held(tsdn, pool); + + malloc_mutex_t *mutex1 = mutex_pool_mutex(pool, key1); + malloc_mutex_t *mutex2 = mutex_pool_mutex(pool, key2); + if ((uintptr_t)mutex1 < (uintptr_t)mutex2) { + malloc_mutex_lock(tsdn, mutex1); + malloc_mutex_lock(tsdn, mutex2); + } else if ((uintptr_t)mutex1 == (uintptr_t)mutex2) { + malloc_mutex_lock(tsdn, mutex1); + } else { + malloc_mutex_lock(tsdn, mutex2); + malloc_mutex_lock(tsdn, mutex1); + } +} + +static inline void +mutex_pool_unlock2(tsdn_t *tsdn, mutex_pool_t *pool, uintptr_t key1, + uintptr_t key2) { + malloc_mutex_t *mutex1 = mutex_pool_mutex(pool, key1); + malloc_mutex_t *mutex2 = mutex_pool_mutex(pool, key2); + if (mutex1 == mutex2) { + malloc_mutex_unlock(tsdn, mutex1); + } else { + malloc_mutex_unlock(tsdn, mutex1); + malloc_mutex_unlock(tsdn, mutex2); + } + + mutex_pool_assert_not_held(tsdn, pool); +} + +static inline void +mutex_pool_assert_owner(tsdn_t *tsdn, mutex_pool_t *pool, uintptr_t key) { + malloc_mutex_assert_owner(tsdn, mutex_pool_mutex(pool, key)); +} + +#endif /* JEMALLOC_INTERNAL_MUTEX_POOL_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/mutex_prof.h b/deps/jemalloc/include/jemalloc/internal/mutex_prof.h new file mode 100644 index 000000000..2cb8fb0cb --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/mutex_prof.h @@ -0,0 +1,108 @@ +#ifndef JEMALLOC_INTERNAL_MUTEX_PROF_H +#define JEMALLOC_INTERNAL_MUTEX_PROF_H + +#include "jemalloc/internal/atomic.h" +#include "jemalloc/internal/nstime.h" +#include "jemalloc/internal/tsd_types.h" + +#define MUTEX_PROF_GLOBAL_MUTEXES \ + OP(background_thread) \ + OP(ctl) \ + OP(prof) + +typedef enum { +#define OP(mtx) global_prof_mutex_##mtx, + MUTEX_PROF_GLOBAL_MUTEXES +#undef OP + mutex_prof_num_global_mutexes +} mutex_prof_global_ind_t; + +#define MUTEX_PROF_ARENA_MUTEXES \ + OP(large) \ + OP(extent_avail) \ + OP(extents_dirty) \ + OP(extents_muzzy) \ + OP(extents_retained) \ + OP(decay_dirty) \ + OP(decay_muzzy) \ + OP(base) \ + OP(tcache_list) + +typedef enum { +#define OP(mtx) arena_prof_mutex_##mtx, + MUTEX_PROF_ARENA_MUTEXES +#undef OP + mutex_prof_num_arena_mutexes +} mutex_prof_arena_ind_t; + +/* + * The forth parameter is a boolean value that is true for derived rate counters + * and false for real ones. + */ +#define MUTEX_PROF_UINT64_COUNTERS \ + OP(num_ops, uint64_t, "n_lock_ops", false, num_ops) \ + OP(num_ops_ps, uint64_t, "(#/sec)", true, num_ops) \ + OP(num_wait, uint64_t, "n_waiting", false, num_wait) \ + OP(num_wait_ps, uint64_t, "(#/sec)", true, num_wait) \ + OP(num_spin_acq, uint64_t, "n_spin_acq", false, num_spin_acq) \ + OP(num_spin_acq_ps, uint64_t, "(#/sec)", true, num_spin_acq) \ + OP(num_owner_switch, uint64_t, "n_owner_switch", false, num_owner_switch) \ + OP(num_owner_switch_ps, uint64_t, "(#/sec)", true, num_owner_switch) \ + OP(total_wait_time, uint64_t, "total_wait_ns", false, total_wait_time) \ + OP(total_wait_time_ps, uint64_t, "(#/sec)", true, total_wait_time) \ + OP(max_wait_time, uint64_t, "max_wait_ns", false, max_wait_time) + +#define MUTEX_PROF_UINT32_COUNTERS \ + OP(max_num_thds, uint32_t, "max_n_thds", false, max_num_thds) + +#define MUTEX_PROF_COUNTERS \ + MUTEX_PROF_UINT64_COUNTERS \ + MUTEX_PROF_UINT32_COUNTERS + +#define OP(counter, type, human, derived, base_counter) mutex_counter_##counter, + +#define COUNTER_ENUM(counter_list, t) \ + typedef enum { \ + counter_list \ + mutex_prof_num_##t##_counters \ + } mutex_prof_##t##_counter_ind_t; + +COUNTER_ENUM(MUTEX_PROF_UINT64_COUNTERS, uint64_t) +COUNTER_ENUM(MUTEX_PROF_UINT32_COUNTERS, uint32_t) + +#undef COUNTER_ENUM +#undef OP + +typedef struct { + /* + * Counters touched on the slow path, i.e. when there is lock + * contention. We update them once we have the lock. + */ + /* Total time (in nano seconds) spent waiting on this mutex. */ + nstime_t tot_wait_time; + /* Max time (in nano seconds) spent on a single lock operation. */ + nstime_t max_wait_time; + /* # of times have to wait for this mutex (after spinning). */ + uint64_t n_wait_times; + /* # of times acquired the mutex through local spinning. */ + uint64_t n_spin_acquired; + /* Max # of threads waiting for the mutex at the same time. */ + uint32_t max_n_thds; + /* Current # of threads waiting on the lock. Atomic synced. */ + atomic_u32_t n_waiting_thds; + + /* + * Data touched on the fast path. These are modified right after we + * grab the lock, so it's placed closest to the end (i.e. right before + * the lock) so that we have a higher chance of them being on the same + * cacheline. + */ + /* # of times the mutex holder is different than the previous one. */ + uint64_t n_owner_switches; + /* Previous mutex holder, to facilitate n_owner_switches. */ + tsdn_t *prev_owner; + /* # of lock() operations in total. */ + uint64_t n_lock_ops; +} mutex_prof_data_t; + +#endif /* JEMALLOC_INTERNAL_MUTEX_PROF_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/nstime.h b/deps/jemalloc/include/jemalloc/internal/nstime.h new file mode 100644 index 000000000..17c177c7f --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/nstime.h @@ -0,0 +1,34 @@ +#ifndef JEMALLOC_INTERNAL_NSTIME_H +#define JEMALLOC_INTERNAL_NSTIME_H + +/* Maximum supported number of seconds (~584 years). */ +#define NSTIME_SEC_MAX KQU(18446744072) +#define NSTIME_ZERO_INITIALIZER {0} + +typedef struct { + uint64_t ns; +} nstime_t; + +void nstime_init(nstime_t *time, uint64_t ns); +void nstime_init2(nstime_t *time, uint64_t sec, uint64_t nsec); +uint64_t nstime_ns(const nstime_t *time); +uint64_t nstime_sec(const nstime_t *time); +uint64_t nstime_msec(const nstime_t *time); +uint64_t nstime_nsec(const nstime_t *time); +void nstime_copy(nstime_t *time, const nstime_t *source); +int nstime_compare(const nstime_t *a, const nstime_t *b); +void nstime_add(nstime_t *time, const nstime_t *addend); +void nstime_iadd(nstime_t *time, uint64_t addend); +void nstime_subtract(nstime_t *time, const nstime_t *subtrahend); +void nstime_isubtract(nstime_t *time, uint64_t subtrahend); +void nstime_imultiply(nstime_t *time, uint64_t multiplier); +void nstime_idivide(nstime_t *time, uint64_t divisor); +uint64_t nstime_divide(const nstime_t *time, const nstime_t *divisor); + +typedef bool (nstime_monotonic_t)(void); +extern nstime_monotonic_t *JET_MUTABLE nstime_monotonic; + +typedef bool (nstime_update_t)(nstime_t *); +extern nstime_update_t *JET_MUTABLE nstime_update; + +#endif /* JEMALLOC_INTERNAL_NSTIME_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/pages.h b/deps/jemalloc/include/jemalloc/internal/pages.h new file mode 100644 index 000000000..7dae633af --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/pages.h @@ -0,0 +1,88 @@ +#ifndef JEMALLOC_INTERNAL_PAGES_EXTERNS_H +#define JEMALLOC_INTERNAL_PAGES_EXTERNS_H + +/* Page size. LG_PAGE is determined by the configure script. */ +#ifdef PAGE_MASK +# undef PAGE_MASK +#endif +#define PAGE ((size_t)(1U << LG_PAGE)) +#define PAGE_MASK ((size_t)(PAGE - 1)) +/* Return the page base address for the page containing address a. */ +#define PAGE_ADDR2BASE(a) \ + ((void *)((uintptr_t)(a) & ~PAGE_MASK)) +/* Return the smallest pagesize multiple that is >= s. */ +#define PAGE_CEILING(s) \ + (((s) + PAGE_MASK) & ~PAGE_MASK) + +/* Huge page size. LG_HUGEPAGE is determined by the configure script. */ +#define HUGEPAGE ((size_t)(1U << LG_HUGEPAGE)) +#define HUGEPAGE_MASK ((size_t)(HUGEPAGE - 1)) +/* Return the huge page base address for the huge page containing address a. */ +#define HUGEPAGE_ADDR2BASE(a) \ + ((void *)((uintptr_t)(a) & ~HUGEPAGE_MASK)) +/* Return the smallest pagesize multiple that is >= s. */ +#define HUGEPAGE_CEILING(s) \ + (((s) + HUGEPAGE_MASK) & ~HUGEPAGE_MASK) + +/* PAGES_CAN_PURGE_LAZY is defined if lazy purging is supported. */ +#if defined(_WIN32) || defined(JEMALLOC_PURGE_MADVISE_FREE) +# define PAGES_CAN_PURGE_LAZY +#endif +/* + * PAGES_CAN_PURGE_FORCED is defined if forced purging is supported. + * + * The only supported way to hard-purge on Windows is to decommit and then + * re-commit, but doing so is racy, and if re-commit fails it's a pain to + * propagate the "poisoned" memory state. Since we typically decommit as the + * next step after purging on Windows anyway, there's no point in adding such + * complexity. + */ +#if !defined(_WIN32) && ((defined(JEMALLOC_PURGE_MADVISE_DONTNEED) && \ + defined(JEMALLOC_PURGE_MADVISE_DONTNEED_ZEROS)) || \ + defined(JEMALLOC_MAPS_COALESCE)) +# define PAGES_CAN_PURGE_FORCED +#endif + +static const bool pages_can_purge_lazy = +#ifdef PAGES_CAN_PURGE_LAZY + true +#else + false +#endif + ; +static const bool pages_can_purge_forced = +#ifdef PAGES_CAN_PURGE_FORCED + true +#else + false +#endif + ; + +typedef enum { + thp_mode_default = 0, /* Do not change hugepage settings. */ + thp_mode_always = 1, /* Always set MADV_HUGEPAGE. */ + thp_mode_never = 2, /* Always set MADV_NOHUGEPAGE. */ + + thp_mode_names_limit = 3, /* Used for option processing. */ + thp_mode_not_supported = 3 /* No THP support detected. */ +} thp_mode_t; + +#define THP_MODE_DEFAULT thp_mode_default +extern thp_mode_t opt_thp; +extern thp_mode_t init_system_thp_mode; /* Initial system wide state. */ +extern const char *thp_mode_names[]; + +void *pages_map(void *addr, size_t size, size_t alignment, bool *commit); +void pages_unmap(void *addr, size_t size); +bool pages_commit(void *addr, size_t size); +bool pages_decommit(void *addr, size_t size); +bool pages_purge_lazy(void *addr, size_t size); +bool pages_purge_forced(void *addr, size_t size); +bool pages_huge(void *addr, size_t size); +bool pages_nohuge(void *addr, size_t size); +bool pages_dontdump(void *addr, size_t size); +bool pages_dodump(void *addr, size_t size); +bool pages_boot(void); +void pages_set_thp_state (void *ptr, size_t size); + +#endif /* JEMALLOC_INTERNAL_PAGES_EXTERNS_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/ph.h b/deps/jemalloc/include/jemalloc/internal/ph.h new file mode 100644 index 000000000..84d6778a9 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/ph.h @@ -0,0 +1,391 @@ +/* + * A Pairing Heap implementation. + * + * "The Pairing Heap: A New Form of Self-Adjusting Heap" + * https://www.cs.cmu.edu/~sleator/papers/pairing-heaps.pdf + * + * With auxiliary twopass list, described in a follow on paper. + * + * "Pairing Heaps: Experiments and Analysis" + * http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.106.2988&rep=rep1&type=pdf + * + ******************************************************************************* + */ + +#ifndef PH_H_ +#define PH_H_ + +/* Node structure. */ +#define phn(a_type) \ +struct { \ + a_type *phn_prev; \ + a_type *phn_next; \ + a_type *phn_lchild; \ +} + +/* Root structure. */ +#define ph(a_type) \ +struct { \ + a_type *ph_root; \ +} + +/* Internal utility macros. */ +#define phn_lchild_get(a_type, a_field, a_phn) \ + (a_phn->a_field.phn_lchild) +#define phn_lchild_set(a_type, a_field, a_phn, a_lchild) do { \ + a_phn->a_field.phn_lchild = a_lchild; \ +} while (0) + +#define phn_next_get(a_type, a_field, a_phn) \ + (a_phn->a_field.phn_next) +#define phn_prev_set(a_type, a_field, a_phn, a_prev) do { \ + a_phn->a_field.phn_prev = a_prev; \ +} while (0) + +#define phn_prev_get(a_type, a_field, a_phn) \ + (a_phn->a_field.phn_prev) +#define phn_next_set(a_type, a_field, a_phn, a_next) do { \ + a_phn->a_field.phn_next = a_next; \ +} while (0) + +#define phn_merge_ordered(a_type, a_field, a_phn0, a_phn1, a_cmp) do { \ + a_type *phn0child; \ + \ + assert(a_phn0 != NULL); \ + assert(a_phn1 != NULL); \ + assert(a_cmp(a_phn0, a_phn1) <= 0); \ + \ + phn_prev_set(a_type, a_field, a_phn1, a_phn0); \ + phn0child = phn_lchild_get(a_type, a_field, a_phn0); \ + phn_next_set(a_type, a_field, a_phn1, phn0child); \ + if (phn0child != NULL) { \ + phn_prev_set(a_type, a_field, phn0child, a_phn1); \ + } \ + phn_lchild_set(a_type, a_field, a_phn0, a_phn1); \ +} while (0) + +#define phn_merge(a_type, a_field, a_phn0, a_phn1, a_cmp, r_phn) do { \ + if (a_phn0 == NULL) { \ + r_phn = a_phn1; \ + } else if (a_phn1 == NULL) { \ + r_phn = a_phn0; \ + } else if (a_cmp(a_phn0, a_phn1) < 0) { \ + phn_merge_ordered(a_type, a_field, a_phn0, a_phn1, \ + a_cmp); \ + r_phn = a_phn0; \ + } else { \ + phn_merge_ordered(a_type, a_field, a_phn1, a_phn0, \ + a_cmp); \ + r_phn = a_phn1; \ + } \ +} while (0) + +#define ph_merge_siblings(a_type, a_field, a_phn, a_cmp, r_phn) do { \ + a_type *head = NULL; \ + a_type *tail = NULL; \ + a_type *phn0 = a_phn; \ + a_type *phn1 = phn_next_get(a_type, a_field, phn0); \ + \ + /* \ + * Multipass merge, wherein the first two elements of a FIFO \ + * are repeatedly merged, and each result is appended to the \ + * singly linked FIFO, until the FIFO contains only a single \ + * element. We start with a sibling list but no reference to \ + * its tail, so we do a single pass over the sibling list to \ + * populate the FIFO. \ + */ \ + if (phn1 != NULL) { \ + a_type *phnrest = phn_next_get(a_type, a_field, phn1); \ + if (phnrest != NULL) { \ + phn_prev_set(a_type, a_field, phnrest, NULL); \ + } \ + phn_prev_set(a_type, a_field, phn0, NULL); \ + phn_next_set(a_type, a_field, phn0, NULL); \ + phn_prev_set(a_type, a_field, phn1, NULL); \ + phn_next_set(a_type, a_field, phn1, NULL); \ + phn_merge(a_type, a_field, phn0, phn1, a_cmp, phn0); \ + head = tail = phn0; \ + phn0 = phnrest; \ + while (phn0 != NULL) { \ + phn1 = phn_next_get(a_type, a_field, phn0); \ + if (phn1 != NULL) { \ + phnrest = phn_next_get(a_type, a_field, \ + phn1); \ + if (phnrest != NULL) { \ + phn_prev_set(a_type, a_field, \ + phnrest, NULL); \ + } \ + phn_prev_set(a_type, a_field, phn0, \ + NULL); \ + phn_next_set(a_type, a_field, phn0, \ + NULL); \ + phn_prev_set(a_type, a_field, phn1, \ + NULL); \ + phn_next_set(a_type, a_field, phn1, \ + NULL); \ + phn_merge(a_type, a_field, phn0, phn1, \ + a_cmp, phn0); \ + phn_next_set(a_type, a_field, tail, \ + phn0); \ + tail = phn0; \ + phn0 = phnrest; \ + } else { \ + phn_next_set(a_type, a_field, tail, \ + phn0); \ + tail = phn0; \ + phn0 = NULL; \ + } \ + } \ + phn0 = head; \ + phn1 = phn_next_get(a_type, a_field, phn0); \ + if (phn1 != NULL) { \ + while (true) { \ + head = phn_next_get(a_type, a_field, \ + phn1); \ + assert(phn_prev_get(a_type, a_field, \ + phn0) == NULL); \ + phn_next_set(a_type, a_field, phn0, \ + NULL); \ + assert(phn_prev_get(a_type, a_field, \ + phn1) == NULL); \ + phn_next_set(a_type, a_field, phn1, \ + NULL); \ + phn_merge(a_type, a_field, phn0, phn1, \ + a_cmp, phn0); \ + if (head == NULL) { \ + break; \ + } \ + phn_next_set(a_type, a_field, tail, \ + phn0); \ + tail = phn0; \ + phn0 = head; \ + phn1 = phn_next_get(a_type, a_field, \ + phn0); \ + } \ + } \ + } \ + r_phn = phn0; \ +} while (0) + +#define ph_merge_aux(a_type, a_field, a_ph, a_cmp) do { \ + a_type *phn = phn_next_get(a_type, a_field, a_ph->ph_root); \ + if (phn != NULL) { \ + phn_prev_set(a_type, a_field, a_ph->ph_root, NULL); \ + phn_next_set(a_type, a_field, a_ph->ph_root, NULL); \ + phn_prev_set(a_type, a_field, phn, NULL); \ + ph_merge_siblings(a_type, a_field, phn, a_cmp, phn); \ + assert(phn_next_get(a_type, a_field, phn) == NULL); \ + phn_merge(a_type, a_field, a_ph->ph_root, phn, a_cmp, \ + a_ph->ph_root); \ + } \ +} while (0) + +#define ph_merge_children(a_type, a_field, a_phn, a_cmp, r_phn) do { \ + a_type *lchild = phn_lchild_get(a_type, a_field, a_phn); \ + if (lchild == NULL) { \ + r_phn = NULL; \ + } else { \ + ph_merge_siblings(a_type, a_field, lchild, a_cmp, \ + r_phn); \ + } \ +} while (0) + +/* + * The ph_proto() macro generates function prototypes that correspond to the + * functions generated by an equivalently parameterized call to ph_gen(). + */ +#define ph_proto(a_attr, a_prefix, a_ph_type, a_type) \ +a_attr void a_prefix##new(a_ph_type *ph); \ +a_attr bool a_prefix##empty(a_ph_type *ph); \ +a_attr a_type *a_prefix##first(a_ph_type *ph); \ +a_attr a_type *a_prefix##any(a_ph_type *ph); \ +a_attr void a_prefix##insert(a_ph_type *ph, a_type *phn); \ +a_attr a_type *a_prefix##remove_first(a_ph_type *ph); \ +a_attr a_type *a_prefix##remove_any(a_ph_type *ph); \ +a_attr void a_prefix##remove(a_ph_type *ph, a_type *phn); + +/* + * The ph_gen() macro generates a type-specific pairing heap implementation, + * based on the above cpp macros. + */ +#define ph_gen(a_attr, a_prefix, a_ph_type, a_type, a_field, a_cmp) \ +a_attr void \ +a_prefix##new(a_ph_type *ph) { \ + memset(ph, 0, sizeof(ph(a_type))); \ +} \ +a_attr bool \ +a_prefix##empty(a_ph_type *ph) { \ + return (ph->ph_root == NULL); \ +} \ +a_attr a_type * \ +a_prefix##first(a_ph_type *ph) { \ + if (ph->ph_root == NULL) { \ + return NULL; \ + } \ + ph_merge_aux(a_type, a_field, ph, a_cmp); \ + return ph->ph_root; \ +} \ +a_attr a_type * \ +a_prefix##any(a_ph_type *ph) { \ + if (ph->ph_root == NULL) { \ + return NULL; \ + } \ + a_type *aux = phn_next_get(a_type, a_field, ph->ph_root); \ + if (aux != NULL) { \ + return aux; \ + } \ + return ph->ph_root; \ +} \ +a_attr void \ +a_prefix##insert(a_ph_type *ph, a_type *phn) { \ + memset(&phn->a_field, 0, sizeof(phn(a_type))); \ + \ + /* \ + * Treat the root as an aux list during insertion, and lazily \ + * merge during a_prefix##remove_first(). For elements that \ + * are inserted, then removed via a_prefix##remove() before the \ + * aux list is ever processed, this makes insert/remove \ + * constant-time, whereas eager merging would make insert \ + * O(log n). \ + */ \ + if (ph->ph_root == NULL) { \ + ph->ph_root = phn; \ + } else { \ + phn_next_set(a_type, a_field, phn, phn_next_get(a_type, \ + a_field, ph->ph_root)); \ + if (phn_next_get(a_type, a_field, ph->ph_root) != \ + NULL) { \ + phn_prev_set(a_type, a_field, \ + phn_next_get(a_type, a_field, ph->ph_root), \ + phn); \ + } \ + phn_prev_set(a_type, a_field, phn, ph->ph_root); \ + phn_next_set(a_type, a_field, ph->ph_root, phn); \ + } \ +} \ +a_attr a_type * \ +a_prefix##remove_first(a_ph_type *ph) { \ + a_type *ret; \ + \ + if (ph->ph_root == NULL) { \ + return NULL; \ + } \ + ph_merge_aux(a_type, a_field, ph, a_cmp); \ + \ + ret = ph->ph_root; \ + \ + ph_merge_children(a_type, a_field, ph->ph_root, a_cmp, \ + ph->ph_root); \ + \ + return ret; \ +} \ +a_attr a_type * \ +a_prefix##remove_any(a_ph_type *ph) { \ + /* \ + * Remove the most recently inserted aux list element, or the \ + * root if the aux list is empty. This has the effect of \ + * behaving as a LIFO (and insertion/removal is therefore \ + * constant-time) if a_prefix##[remove_]first() are never \ + * called. \ + */ \ + if (ph->ph_root == NULL) { \ + return NULL; \ + } \ + a_type *ret = phn_next_get(a_type, a_field, ph->ph_root); \ + if (ret != NULL) { \ + a_type *aux = phn_next_get(a_type, a_field, ret); \ + phn_next_set(a_type, a_field, ph->ph_root, aux); \ + if (aux != NULL) { \ + phn_prev_set(a_type, a_field, aux, \ + ph->ph_root); \ + } \ + return ret; \ + } \ + ret = ph->ph_root; \ + ph_merge_children(a_type, a_field, ph->ph_root, a_cmp, \ + ph->ph_root); \ + return ret; \ +} \ +a_attr void \ +a_prefix##remove(a_ph_type *ph, a_type *phn) { \ + a_type *replace, *parent; \ + \ + if (ph->ph_root == phn) { \ + /* \ + * We can delete from aux list without merging it, but \ + * we need to merge if we are dealing with the root \ + * node and it has children. \ + */ \ + if (phn_lchild_get(a_type, a_field, phn) == NULL) { \ + ph->ph_root = phn_next_get(a_type, a_field, \ + phn); \ + if (ph->ph_root != NULL) { \ + phn_prev_set(a_type, a_field, \ + ph->ph_root, NULL); \ + } \ + return; \ + } \ + ph_merge_aux(a_type, a_field, ph, a_cmp); \ + if (ph->ph_root == phn) { \ + ph_merge_children(a_type, a_field, ph->ph_root, \ + a_cmp, ph->ph_root); \ + return; \ + } \ + } \ + \ + /* Get parent (if phn is leftmost child) before mutating. */ \ + if ((parent = phn_prev_get(a_type, a_field, phn)) != NULL) { \ + if (phn_lchild_get(a_type, a_field, parent) != phn) { \ + parent = NULL; \ + } \ + } \ + /* Find a possible replacement node, and link to parent. */ \ + ph_merge_children(a_type, a_field, phn, a_cmp, replace); \ + /* Set next/prev for sibling linked list. */ \ + if (replace != NULL) { \ + if (parent != NULL) { \ + phn_prev_set(a_type, a_field, replace, parent); \ + phn_lchild_set(a_type, a_field, parent, \ + replace); \ + } else { \ + phn_prev_set(a_type, a_field, replace, \ + phn_prev_get(a_type, a_field, phn)); \ + if (phn_prev_get(a_type, a_field, phn) != \ + NULL) { \ + phn_next_set(a_type, a_field, \ + phn_prev_get(a_type, a_field, phn), \ + replace); \ + } \ + } \ + phn_next_set(a_type, a_field, replace, \ + phn_next_get(a_type, a_field, phn)); \ + if (phn_next_get(a_type, a_field, phn) != NULL) { \ + phn_prev_set(a_type, a_field, \ + phn_next_get(a_type, a_field, phn), \ + replace); \ + } \ + } else { \ + if (parent != NULL) { \ + a_type *next = phn_next_get(a_type, a_field, \ + phn); \ + phn_lchild_set(a_type, a_field, parent, next); \ + if (next != NULL) { \ + phn_prev_set(a_type, a_field, next, \ + parent); \ + } \ + } else { \ + assert(phn_prev_get(a_type, a_field, phn) != \ + NULL); \ + phn_next_set(a_type, a_field, \ + phn_prev_get(a_type, a_field, phn), \ + phn_next_get(a_type, a_field, phn)); \ + } \ + if (phn_next_get(a_type, a_field, phn) != NULL) { \ + phn_prev_set(a_type, a_field, \ + phn_next_get(a_type, a_field, phn), \ + phn_prev_get(a_type, a_field, phn)); \ + } \ + } \ +} + +#endif /* PH_H_ */ diff --git a/deps/jemalloc/include/jemalloc/internal/private_namespace.sh b/deps/jemalloc/include/jemalloc/internal/private_namespace.sh new file mode 100755 index 000000000..6ef1346a3 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/private_namespace.sh @@ -0,0 +1,5 @@ +#!/bin/sh + +for symbol in `cat "$@"` ; do + echo "#define ${symbol} JEMALLOC_N(${symbol})" +done diff --git a/deps/jemalloc/include/jemalloc/internal/private_symbols.sh b/deps/jemalloc/include/jemalloc/internal/private_symbols.sh new file mode 100755 index 000000000..442a259fd --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/private_symbols.sh @@ -0,0 +1,51 @@ +#!/bin/sh +# +# Generate private_symbols[_jet].awk. +# +# Usage: private_symbols.sh <sym_prefix> <sym>* +# +# <sym_prefix> is typically "" or "_". + +sym_prefix=$1 +shift + +cat <<EOF +#!/usr/bin/env awk -f + +BEGIN { + sym_prefix = "${sym_prefix}" + split("\\ +EOF + +for public_sym in "$@" ; do + cat <<EOF + ${sym_prefix}${public_sym} \\ +EOF +done + +cat <<"EOF" + ", exported_symbol_names) + # Store exported symbol names as keys in exported_symbols. + for (i in exported_symbol_names) { + exported_symbols[exported_symbol_names[i]] = 1 + } +} + +# Process 'nm -a <c_source.o>' output. +# +# Handle lines like: +# 0000000000000008 D opt_junk +# 0000000000007574 T malloc_initialized +(NF == 3 && $2 ~ /^[ABCDGRSTVW]$/ && !($3 in exported_symbols) && $3 ~ /^[A-Za-z0-9_]+$/) { + print substr($3, 1+length(sym_prefix), length($3)-length(sym_prefix)) +} + +# Process 'dumpbin /SYMBOLS <c_source.obj>' output. +# +# Handle lines like: +# 353 00008098 SECT4 notype External | opt_junk +# 3F1 00000000 SECT7 notype () External | malloc_initialized +($3 ~ /^SECT[0-9]+/ && $(NF-2) == "External" && !($NF in exported_symbols)) { + print $NF +} +EOF diff --git a/deps/jemalloc/include/jemalloc/internal/prng.h b/deps/jemalloc/include/jemalloc/internal/prng.h new file mode 100644 index 000000000..15cc2d18f --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/prng.h @@ -0,0 +1,185 @@ +#ifndef JEMALLOC_INTERNAL_PRNG_H +#define JEMALLOC_INTERNAL_PRNG_H + +#include "jemalloc/internal/atomic.h" +#include "jemalloc/internal/bit_util.h" + +/* + * Simple linear congruential pseudo-random number generator: + * + * prng(y) = (a*x + c) % m + * + * where the following constants ensure maximal period: + * + * a == Odd number (relatively prime to 2^n), and (a-1) is a multiple of 4. + * c == Odd number (relatively prime to 2^n). + * m == 2^32 + * + * See Knuth's TAOCP 3rd Ed., Vol. 2, pg. 17 for details on these constraints. + * + * This choice of m has the disadvantage that the quality of the bits is + * proportional to bit position. For example, the lowest bit has a cycle of 2, + * the next has a cycle of 4, etc. For this reason, we prefer to use the upper + * bits. + */ + +/******************************************************************************/ +/* INTERNAL DEFINITIONS -- IGNORE */ +/******************************************************************************/ +#define PRNG_A_32 UINT32_C(1103515241) +#define PRNG_C_32 UINT32_C(12347) + +#define PRNG_A_64 UINT64_C(6364136223846793005) +#define PRNG_C_64 UINT64_C(1442695040888963407) + +JEMALLOC_ALWAYS_INLINE uint32_t +prng_state_next_u32(uint32_t state) { + return (state * PRNG_A_32) + PRNG_C_32; +} + +JEMALLOC_ALWAYS_INLINE uint64_t +prng_state_next_u64(uint64_t state) { + return (state * PRNG_A_64) + PRNG_C_64; +} + +JEMALLOC_ALWAYS_INLINE size_t +prng_state_next_zu(size_t state) { +#if LG_SIZEOF_PTR == 2 + return (state * PRNG_A_32) + PRNG_C_32; +#elif LG_SIZEOF_PTR == 3 + return (state * PRNG_A_64) + PRNG_C_64; +#else +#error Unsupported pointer size +#endif +} + +/******************************************************************************/ +/* BEGIN PUBLIC API */ +/******************************************************************************/ + +/* + * The prng_lg_range functions give a uniform int in the half-open range [0, + * 2**lg_range). If atomic is true, they do so safely from multiple threads. + * Multithreaded 64-bit prngs aren't supported. + */ + +JEMALLOC_ALWAYS_INLINE uint32_t +prng_lg_range_u32(atomic_u32_t *state, unsigned lg_range, bool atomic) { + uint32_t ret, state0, state1; + + assert(lg_range > 0); + assert(lg_range <= 32); + + state0 = atomic_load_u32(state, ATOMIC_RELAXED); + + if (atomic) { + do { + state1 = prng_state_next_u32(state0); + } while (!atomic_compare_exchange_weak_u32(state, &state0, + state1, ATOMIC_RELAXED, ATOMIC_RELAXED)); + } else { + state1 = prng_state_next_u32(state0); + atomic_store_u32(state, state1, ATOMIC_RELAXED); + } + ret = state1 >> (32 - lg_range); + + return ret; +} + +JEMALLOC_ALWAYS_INLINE uint64_t +prng_lg_range_u64(uint64_t *state, unsigned lg_range) { + uint64_t ret, state1; + + assert(lg_range > 0); + assert(lg_range <= 64); + + state1 = prng_state_next_u64(*state); + *state = state1; + ret = state1 >> (64 - lg_range); + + return ret; +} + +JEMALLOC_ALWAYS_INLINE size_t +prng_lg_range_zu(atomic_zu_t *state, unsigned lg_range, bool atomic) { + size_t ret, state0, state1; + + assert(lg_range > 0); + assert(lg_range <= ZU(1) << (3 + LG_SIZEOF_PTR)); + + state0 = atomic_load_zu(state, ATOMIC_RELAXED); + + if (atomic) { + do { + state1 = prng_state_next_zu(state0); + } while (atomic_compare_exchange_weak_zu(state, &state0, + state1, ATOMIC_RELAXED, ATOMIC_RELAXED)); + } else { + state1 = prng_state_next_zu(state0); + atomic_store_zu(state, state1, ATOMIC_RELAXED); + } + ret = state1 >> ((ZU(1) << (3 + LG_SIZEOF_PTR)) - lg_range); + + return ret; +} + +/* + * The prng_range functions behave like the prng_lg_range, but return a result + * in [0, range) instead of [0, 2**lg_range). + */ + +JEMALLOC_ALWAYS_INLINE uint32_t +prng_range_u32(atomic_u32_t *state, uint32_t range, bool atomic) { + uint32_t ret; + unsigned lg_range; + + assert(range > 1); + + /* Compute the ceiling of lg(range). */ + lg_range = ffs_u32(pow2_ceil_u32(range)) - 1; + + /* Generate a result in [0..range) via repeated trial. */ + do { + ret = prng_lg_range_u32(state, lg_range, atomic); + } while (ret >= range); + + return ret; +} + +JEMALLOC_ALWAYS_INLINE uint64_t +prng_range_u64(uint64_t *state, uint64_t range) { + uint64_t ret; + unsigned lg_range; + + assert(range > 1); + + /* Compute the ceiling of lg(range). */ + lg_range = ffs_u64(pow2_ceil_u64(range)) - 1; + + /* Generate a result in [0..range) via repeated trial. */ + do { + ret = prng_lg_range_u64(state, lg_range); + } while (ret >= range); + + return ret; +} + +JEMALLOC_ALWAYS_INLINE size_t +prng_range_zu(atomic_zu_t *state, size_t range, bool atomic) { + size_t ret; + unsigned lg_range; + + assert(range > 1); + + /* Compute the ceiling of lg(range). */ + lg_range = ffs_u64(pow2_ceil_u64(range)) - 1; + + /* Generate a result in [0..range) via repeated trial. */ + do { + ret = prng_lg_range_zu(state, lg_range, atomic); + } while (ret >= range); + + return ret; +} + +#endif /* JEMALLOC_INTERNAL_PRNG_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/prof_externs.h b/deps/jemalloc/include/jemalloc/internal/prof_externs.h new file mode 100644 index 000000000..094f3e170 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/prof_externs.h @@ -0,0 +1,105 @@ +#ifndef JEMALLOC_INTERNAL_PROF_EXTERNS_H +#define JEMALLOC_INTERNAL_PROF_EXTERNS_H + +#include "jemalloc/internal/mutex.h" + +extern malloc_mutex_t bt2gctx_mtx; + +extern bool opt_prof; +extern bool opt_prof_active; +extern bool opt_prof_thread_active_init; +extern size_t opt_lg_prof_sample; /* Mean bytes between samples. */ +extern ssize_t opt_lg_prof_interval; /* lg(prof_interval). */ +extern bool opt_prof_gdump; /* High-water memory dumping. */ +extern bool opt_prof_final; /* Final profile dumping. */ +extern bool opt_prof_leak; /* Dump leak summary at exit. */ +extern bool opt_prof_accum; /* Report cumulative bytes. */ +extern bool opt_prof_log; /* Turn logging on at boot. */ +extern char opt_prof_prefix[ + /* Minimize memory bloat for non-prof builds. */ +#ifdef JEMALLOC_PROF + PATH_MAX + +#endif + 1]; + +/* Accessed via prof_active_[gs]et{_unlocked,}(). */ +extern bool prof_active; + +/* Accessed via prof_gdump_[gs]et{_unlocked,}(). */ +extern bool prof_gdump_val; + +/* + * Profile dump interval, measured in bytes allocated. Each arena triggers a + * profile dump when it reaches this threshold. The effect is that the + * interval between profile dumps averages prof_interval, though the actual + * interval between dumps will tend to be sporadic, and the interval will be a + * maximum of approximately (prof_interval * narenas). + */ +extern uint64_t prof_interval; + +/* + * Initialized as opt_lg_prof_sample, and potentially modified during profiling + * resets. + */ +extern size_t lg_prof_sample; + +void prof_alloc_rollback(tsd_t *tsd, prof_tctx_t *tctx, bool updated); +void prof_malloc_sample_object(tsdn_t *tsdn, const void *ptr, size_t usize, + prof_tctx_t *tctx); +void prof_free_sampled_object(tsd_t *tsd, const void *ptr, size_t usize, + prof_tctx_t *tctx); +void bt_init(prof_bt_t *bt, void **vec); +void prof_backtrace(prof_bt_t *bt); +prof_tctx_t *prof_lookup(tsd_t *tsd, prof_bt_t *bt); +#ifdef JEMALLOC_JET +size_t prof_tdata_count(void); +size_t prof_bt_count(void); +#endif +typedef int (prof_dump_open_t)(bool, const char *); +extern prof_dump_open_t *JET_MUTABLE prof_dump_open; + +typedef bool (prof_dump_header_t)(tsdn_t *, bool, const prof_cnt_t *); +extern prof_dump_header_t *JET_MUTABLE prof_dump_header; +#ifdef JEMALLOC_JET +void prof_cnt_all(uint64_t *curobjs, uint64_t *curbytes, uint64_t *accumobjs, + uint64_t *accumbytes); +#endif +bool prof_accum_init(tsdn_t *tsdn, prof_accum_t *prof_accum); +void prof_idump(tsdn_t *tsdn); +bool prof_mdump(tsd_t *tsd, const char *filename); +void prof_gdump(tsdn_t *tsdn); +prof_tdata_t *prof_tdata_init(tsd_t *tsd); +prof_tdata_t *prof_tdata_reinit(tsd_t *tsd, prof_tdata_t *tdata); +void prof_reset(tsd_t *tsd, size_t lg_sample); +void prof_tdata_cleanup(tsd_t *tsd); +bool prof_active_get(tsdn_t *tsdn); +bool prof_active_set(tsdn_t *tsdn, bool active); +const char *prof_thread_name_get(tsd_t *tsd); +int prof_thread_name_set(tsd_t *tsd, const char *thread_name); +bool prof_thread_active_get(tsd_t *tsd); +bool prof_thread_active_set(tsd_t *tsd, bool active); +bool prof_thread_active_init_get(tsdn_t *tsdn); +bool prof_thread_active_init_set(tsdn_t *tsdn, bool active_init); +bool prof_gdump_get(tsdn_t *tsdn); +bool prof_gdump_set(tsdn_t *tsdn, bool active); +void prof_boot0(void); +void prof_boot1(void); +bool prof_boot2(tsd_t *tsd); +void prof_prefork0(tsdn_t *tsdn); +void prof_prefork1(tsdn_t *tsdn); +void prof_postfork_parent(tsdn_t *tsdn); +void prof_postfork_child(tsdn_t *tsdn); +void prof_sample_threshold_update(prof_tdata_t *tdata); + +bool prof_log_start(tsdn_t *tsdn, const char *filename); +bool prof_log_stop(tsdn_t *tsdn); +#ifdef JEMALLOC_JET +size_t prof_log_bt_count(void); +size_t prof_log_alloc_count(void); +size_t prof_log_thr_count(void); +bool prof_log_is_logging(void); +bool prof_log_rep_check(void); +void prof_log_dummy_set(bool new_value); +#endif + +#endif /* JEMALLOC_INTERNAL_PROF_EXTERNS_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/prof_inlines_a.h b/deps/jemalloc/include/jemalloc/internal/prof_inlines_a.h new file mode 100644 index 000000000..471d9853c --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/prof_inlines_a.h @@ -0,0 +1,85 @@ +#ifndef JEMALLOC_INTERNAL_PROF_INLINES_A_H +#define JEMALLOC_INTERNAL_PROF_INLINES_A_H + +#include "jemalloc/internal/mutex.h" + +static inline bool +prof_accum_add(tsdn_t *tsdn, prof_accum_t *prof_accum, + uint64_t accumbytes) { + cassert(config_prof); + + bool overflow; + uint64_t a0, a1; + + /* + * If the application allocates fast enough (and/or if idump is slow + * enough), extreme overflow here (a1 >= prof_interval * 2) can cause + * idump trigger coalescing. This is an intentional mechanism that + * avoids rate-limiting allocation. + */ +#ifdef JEMALLOC_ATOMIC_U64 + a0 = atomic_load_u64(&prof_accum->accumbytes, ATOMIC_RELAXED); + do { + a1 = a0 + accumbytes; + assert(a1 >= a0); + overflow = (a1 >= prof_interval); + if (overflow) { + a1 %= prof_interval; + } + } while (!atomic_compare_exchange_weak_u64(&prof_accum->accumbytes, &a0, + a1, ATOMIC_RELAXED, ATOMIC_RELAXED)); +#else + malloc_mutex_lock(tsdn, &prof_accum->mtx); + a0 = prof_accum->accumbytes; + a1 = a0 + accumbytes; + overflow = (a1 >= prof_interval); + if (overflow) { + a1 %= prof_interval; + } + prof_accum->accumbytes = a1; + malloc_mutex_unlock(tsdn, &prof_accum->mtx); +#endif + return overflow; +} + +static inline void +prof_accum_cancel(tsdn_t *tsdn, prof_accum_t *prof_accum, + size_t usize) { + cassert(config_prof); + + /* + * Cancel out as much of the excessive prof_accumbytes increase as + * possible without underflowing. Interval-triggered dumps occur + * slightly more often than intended as a result of incomplete + * canceling. + */ + uint64_t a0, a1; +#ifdef JEMALLOC_ATOMIC_U64 + a0 = atomic_load_u64(&prof_accum->accumbytes, ATOMIC_RELAXED); + do { + a1 = (a0 >= SC_LARGE_MINCLASS - usize) + ? a0 - (SC_LARGE_MINCLASS - usize) : 0; + } while (!atomic_compare_exchange_weak_u64(&prof_accum->accumbytes, &a0, + a1, ATOMIC_RELAXED, ATOMIC_RELAXED)); +#else + malloc_mutex_lock(tsdn, &prof_accum->mtx); + a0 = prof_accum->accumbytes; + a1 = (a0 >= SC_LARGE_MINCLASS - usize) + ? a0 - (SC_LARGE_MINCLASS - usize) : 0; + prof_accum->accumbytes = a1; + malloc_mutex_unlock(tsdn, &prof_accum->mtx); +#endif +} + +JEMALLOC_ALWAYS_INLINE bool +prof_active_get_unlocked(void) { + /* + * Even if opt_prof is true, sampling can be temporarily disabled by + * setting prof_active to false. No locking is used when reading + * prof_active in the fast path, so there are no guarantees regarding + * how long it will take for all threads to notice state changes. + */ + return prof_active; +} + +#endif /* JEMALLOC_INTERNAL_PROF_INLINES_A_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/prof_inlines_b.h b/deps/jemalloc/include/jemalloc/internal/prof_inlines_b.h new file mode 100644 index 000000000..8ba8a1e1f --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/prof_inlines_b.h @@ -0,0 +1,250 @@ +#ifndef JEMALLOC_INTERNAL_PROF_INLINES_B_H +#define JEMALLOC_INTERNAL_PROF_INLINES_B_H + +#include "jemalloc/internal/safety_check.h" +#include "jemalloc/internal/sz.h" + +JEMALLOC_ALWAYS_INLINE bool +prof_gdump_get_unlocked(void) { + /* + * No locking is used when reading prof_gdump_val in the fast path, so + * there are no guarantees regarding how long it will take for all + * threads to notice state changes. + */ + return prof_gdump_val; +} + +JEMALLOC_ALWAYS_INLINE prof_tdata_t * +prof_tdata_get(tsd_t *tsd, bool create) { + prof_tdata_t *tdata; + + cassert(config_prof); + + tdata = tsd_prof_tdata_get(tsd); + if (create) { + if (unlikely(tdata == NULL)) { + if (tsd_nominal(tsd)) { + tdata = prof_tdata_init(tsd); + tsd_prof_tdata_set(tsd, tdata); + } + } else if (unlikely(tdata->expired)) { + tdata = prof_tdata_reinit(tsd, tdata); + tsd_prof_tdata_set(tsd, tdata); + } + assert(tdata == NULL || tdata->attached); + } + + return tdata; +} + +JEMALLOC_ALWAYS_INLINE prof_tctx_t * +prof_tctx_get(tsdn_t *tsdn, const void *ptr, alloc_ctx_t *alloc_ctx) { + cassert(config_prof); + assert(ptr != NULL); + + return arena_prof_tctx_get(tsdn, ptr, alloc_ctx); +} + +JEMALLOC_ALWAYS_INLINE void +prof_tctx_set(tsdn_t *tsdn, const void *ptr, size_t usize, + alloc_ctx_t *alloc_ctx, prof_tctx_t *tctx) { + cassert(config_prof); + assert(ptr != NULL); + + arena_prof_tctx_set(tsdn, ptr, usize, alloc_ctx, tctx); +} + +JEMALLOC_ALWAYS_INLINE void +prof_tctx_reset(tsdn_t *tsdn, const void *ptr, prof_tctx_t *tctx) { + cassert(config_prof); + assert(ptr != NULL); + + arena_prof_tctx_reset(tsdn, ptr, tctx); +} + +JEMALLOC_ALWAYS_INLINE nstime_t +prof_alloc_time_get(tsdn_t *tsdn, const void *ptr, alloc_ctx_t *alloc_ctx) { + cassert(config_prof); + assert(ptr != NULL); + + return arena_prof_alloc_time_get(tsdn, ptr, alloc_ctx); +} + +JEMALLOC_ALWAYS_INLINE void +prof_alloc_time_set(tsdn_t *tsdn, const void *ptr, alloc_ctx_t *alloc_ctx, + nstime_t t) { + cassert(config_prof); + assert(ptr != NULL); + + arena_prof_alloc_time_set(tsdn, ptr, alloc_ctx, t); +} + +JEMALLOC_ALWAYS_INLINE bool +prof_sample_check(tsd_t *tsd, size_t usize, bool update) { + ssize_t check = update ? 0 : usize; + + int64_t bytes_until_sample = tsd_bytes_until_sample_get(tsd); + if (update) { + bytes_until_sample -= usize; + if (tsd_nominal(tsd)) { + tsd_bytes_until_sample_set(tsd, bytes_until_sample); + } + } + if (likely(bytes_until_sample >= check)) { + return true; + } + + return false; +} + +JEMALLOC_ALWAYS_INLINE bool +prof_sample_accum_update(tsd_t *tsd, size_t usize, bool update, + prof_tdata_t **tdata_out) { + prof_tdata_t *tdata; + + cassert(config_prof); + + /* Fastpath: no need to load tdata */ + if (likely(prof_sample_check(tsd, usize, update))) { + return true; + } + + bool booted = tsd_prof_tdata_get(tsd); + tdata = prof_tdata_get(tsd, true); + if (unlikely((uintptr_t)tdata <= (uintptr_t)PROF_TDATA_STATE_MAX)) { + tdata = NULL; + } + + if (tdata_out != NULL) { + *tdata_out = tdata; + } + + if (unlikely(tdata == NULL)) { + return true; + } + + /* + * If this was the first creation of tdata, then + * prof_tdata_get() reset bytes_until_sample, so decrement and + * check it again + */ + if (!booted && prof_sample_check(tsd, usize, update)) { + return true; + } + + if (tsd_reentrancy_level_get(tsd) > 0) { + return true; + } + /* Compute new sample threshold. */ + if (update) { + prof_sample_threshold_update(tdata); + } + return !tdata->active; +} + +JEMALLOC_ALWAYS_INLINE prof_tctx_t * +prof_alloc_prep(tsd_t *tsd, size_t usize, bool prof_active, bool update) { + prof_tctx_t *ret; + prof_tdata_t *tdata; + prof_bt_t bt; + + assert(usize == sz_s2u(usize)); + + if (!prof_active || likely(prof_sample_accum_update(tsd, usize, update, + &tdata))) { + ret = (prof_tctx_t *)(uintptr_t)1U; + } else { + bt_init(&bt, tdata->vec); + prof_backtrace(&bt); + ret = prof_lookup(tsd, &bt); + } + + return ret; +} + +JEMALLOC_ALWAYS_INLINE void +prof_malloc(tsdn_t *tsdn, const void *ptr, size_t usize, alloc_ctx_t *alloc_ctx, + prof_tctx_t *tctx) { + cassert(config_prof); + assert(ptr != NULL); + assert(usize == isalloc(tsdn, ptr)); + + if (unlikely((uintptr_t)tctx > (uintptr_t)1U)) { + prof_malloc_sample_object(tsdn, ptr, usize, tctx); + } else { + prof_tctx_set(tsdn, ptr, usize, alloc_ctx, + (prof_tctx_t *)(uintptr_t)1U); + } +} + +JEMALLOC_ALWAYS_INLINE void +prof_realloc(tsd_t *tsd, const void *ptr, size_t usize, prof_tctx_t *tctx, + bool prof_active, bool updated, const void *old_ptr, size_t old_usize, + prof_tctx_t *old_tctx) { + bool sampled, old_sampled, moved; + + cassert(config_prof); + assert(ptr != NULL || (uintptr_t)tctx <= (uintptr_t)1U); + + if (prof_active && !updated && ptr != NULL) { + assert(usize == isalloc(tsd_tsdn(tsd), ptr)); + if (prof_sample_accum_update(tsd, usize, true, NULL)) { + /* + * Don't sample. The usize passed to prof_alloc_prep() + * was larger than what actually got allocated, so a + * backtrace was captured for this allocation, even + * though its actual usize was insufficient to cross the + * sample threshold. + */ + prof_alloc_rollback(tsd, tctx, true); + tctx = (prof_tctx_t *)(uintptr_t)1U; + } + } + + sampled = ((uintptr_t)tctx > (uintptr_t)1U); + old_sampled = ((uintptr_t)old_tctx > (uintptr_t)1U); + moved = (ptr != old_ptr); + + if (unlikely(sampled)) { + prof_malloc_sample_object(tsd_tsdn(tsd), ptr, usize, tctx); + } else if (moved) { + prof_tctx_set(tsd_tsdn(tsd), ptr, usize, NULL, + (prof_tctx_t *)(uintptr_t)1U); + } else if (unlikely(old_sampled)) { + /* + * prof_tctx_set() would work for the !moved case as well, but + * prof_tctx_reset() is slightly cheaper, and the proper thing + * to do here in the presence of explicit knowledge re: moved + * state. + */ + prof_tctx_reset(tsd_tsdn(tsd), ptr, tctx); + } else { + assert((uintptr_t)prof_tctx_get(tsd_tsdn(tsd), ptr, NULL) == + (uintptr_t)1U); + } + + /* + * The prof_free_sampled_object() call must come after the + * prof_malloc_sample_object() call, because tctx and old_tctx may be + * the same, in which case reversing the call order could cause the tctx + * to be prematurely destroyed as a side effect of momentarily zeroed + * counters. + */ + if (unlikely(old_sampled)) { + prof_free_sampled_object(tsd, ptr, old_usize, old_tctx); + } +} + +JEMALLOC_ALWAYS_INLINE void +prof_free(tsd_t *tsd, const void *ptr, size_t usize, alloc_ctx_t *alloc_ctx) { + prof_tctx_t *tctx = prof_tctx_get(tsd_tsdn(tsd), ptr, alloc_ctx); + + cassert(config_prof); + assert(usize == isalloc(tsd_tsdn(tsd), ptr)); + + if (unlikely((uintptr_t)tctx > (uintptr_t)1U)) { + prof_free_sampled_object(tsd, ptr, usize, tctx); + } +} + +#endif /* JEMALLOC_INTERNAL_PROF_INLINES_B_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/prof_structs.h b/deps/jemalloc/include/jemalloc/internal/prof_structs.h new file mode 100644 index 000000000..34ed4822b --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/prof_structs.h @@ -0,0 +1,200 @@ +#ifndef JEMALLOC_INTERNAL_PROF_STRUCTS_H +#define JEMALLOC_INTERNAL_PROF_STRUCTS_H + +#include "jemalloc/internal/ckh.h" +#include "jemalloc/internal/mutex.h" +#include "jemalloc/internal/prng.h" +#include "jemalloc/internal/rb.h" + +struct prof_bt_s { + /* Backtrace, stored as len program counters. */ + void **vec; + unsigned len; +}; + +#ifdef JEMALLOC_PROF_LIBGCC +/* Data structure passed to libgcc _Unwind_Backtrace() callback functions. */ +typedef struct { + prof_bt_t *bt; + unsigned max; +} prof_unwind_data_t; +#endif + +struct prof_accum_s { +#ifndef JEMALLOC_ATOMIC_U64 + malloc_mutex_t mtx; + uint64_t accumbytes; +#else + atomic_u64_t accumbytes; +#endif +}; + +struct prof_cnt_s { + /* Profiling counters. */ + uint64_t curobjs; + uint64_t curbytes; + uint64_t accumobjs; + uint64_t accumbytes; +}; + +typedef enum { + prof_tctx_state_initializing, + prof_tctx_state_nominal, + prof_tctx_state_dumping, + prof_tctx_state_purgatory /* Dumper must finish destroying. */ +} prof_tctx_state_t; + +struct prof_tctx_s { + /* Thread data for thread that performed the allocation. */ + prof_tdata_t *tdata; + + /* + * Copy of tdata->thr_{uid,discrim}, necessary because tdata may be + * defunct during teardown. + */ + uint64_t thr_uid; + uint64_t thr_discrim; + + /* Profiling counters, protected by tdata->lock. */ + prof_cnt_t cnts; + + /* Associated global context. */ + prof_gctx_t *gctx; + + /* + * UID that distinguishes multiple tctx's created by the same thread, + * but coexisting in gctx->tctxs. There are two ways that such + * coexistence can occur: + * - A dumper thread can cause a tctx to be retained in the purgatory + * state. + * - Although a single "producer" thread must create all tctx's which + * share the same thr_uid, multiple "consumers" can each concurrently + * execute portions of prof_tctx_destroy(). prof_tctx_destroy() only + * gets called once each time cnts.cur{objs,bytes} drop to 0, but this + * threshold can be hit again before the first consumer finishes + * executing prof_tctx_destroy(). + */ + uint64_t tctx_uid; + + /* Linkage into gctx's tctxs. */ + rb_node(prof_tctx_t) tctx_link; + + /* + * True during prof_alloc_prep()..prof_malloc_sample_object(), prevents + * sample vs destroy race. + */ + bool prepared; + + /* Current dump-related state, protected by gctx->lock. */ + prof_tctx_state_t state; + + /* + * Copy of cnts snapshotted during early dump phase, protected by + * dump_mtx. + */ + prof_cnt_t dump_cnts; +}; +typedef rb_tree(prof_tctx_t) prof_tctx_tree_t; + +struct prof_gctx_s { + /* Protects nlimbo, cnt_summed, and tctxs. */ + malloc_mutex_t *lock; + + /* + * Number of threads that currently cause this gctx to be in a state of + * limbo due to one of: + * - Initializing this gctx. + * - Initializing per thread counters associated with this gctx. + * - Preparing to destroy this gctx. + * - Dumping a heap profile that includes this gctx. + * nlimbo must be 1 (single destroyer) in order to safely destroy the + * gctx. + */ + unsigned nlimbo; + + /* + * Tree of profile counters, one for each thread that has allocated in + * this context. + */ + prof_tctx_tree_t tctxs; + + /* Linkage for tree of contexts to be dumped. */ + rb_node(prof_gctx_t) dump_link; + + /* Temporary storage for summation during dump. */ + prof_cnt_t cnt_summed; + + /* Associated backtrace. */ + prof_bt_t bt; + + /* Backtrace vector, variable size, referred to by bt. */ + void *vec[1]; +}; +typedef rb_tree(prof_gctx_t) prof_gctx_tree_t; + +struct prof_tdata_s { + malloc_mutex_t *lock; + + /* Monotonically increasing unique thread identifier. */ + uint64_t thr_uid; + + /* + * Monotonically increasing discriminator among tdata structures + * associated with the same thr_uid. + */ + uint64_t thr_discrim; + + /* Included in heap profile dumps if non-NULL. */ + char *thread_name; + + bool attached; + bool expired; + + rb_node(prof_tdata_t) tdata_link; + + /* + * Counter used to initialize prof_tctx_t's tctx_uid. No locking is + * necessary when incrementing this field, because only one thread ever + * does so. + */ + uint64_t tctx_uid_next; + + /* + * Hash of (prof_bt_t *)-->(prof_tctx_t *). Each thread tracks + * backtraces for which it has non-zero allocation/deallocation counters + * associated with thread-specific prof_tctx_t objects. Other threads + * may write to prof_tctx_t contents when freeing associated objects. + */ + ckh_t bt2tctx; + + /* Sampling state. */ + uint64_t prng_state; + + /* State used to avoid dumping while operating on prof internals. */ + bool enq; + bool enq_idump; + bool enq_gdump; + + /* + * Set to true during an early dump phase for tdata's which are + * currently being dumped. New threads' tdata's have this initialized + * to false so that they aren't accidentally included in later dump + * phases. + */ + bool dumping; + + /* + * True if profiling is active for this tdata's thread + * (thread.prof.active mallctl). + */ + bool active; + + /* Temporary storage for summation during dump. */ + prof_cnt_t cnt_summed; + + /* Backtrace vector, used for calls to prof_backtrace(). */ + void *vec[PROF_BT_MAX]; +}; +typedef rb_tree(prof_tdata_t) prof_tdata_tree_t; + +#endif /* JEMALLOC_INTERNAL_PROF_STRUCTS_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/prof_types.h b/deps/jemalloc/include/jemalloc/internal/prof_types.h new file mode 100644 index 000000000..1eff995ec --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/prof_types.h @@ -0,0 +1,56 @@ +#ifndef JEMALLOC_INTERNAL_PROF_TYPES_H +#define JEMALLOC_INTERNAL_PROF_TYPES_H + +typedef struct prof_bt_s prof_bt_t; +typedef struct prof_accum_s prof_accum_t; +typedef struct prof_cnt_s prof_cnt_t; +typedef struct prof_tctx_s prof_tctx_t; +typedef struct prof_gctx_s prof_gctx_t; +typedef struct prof_tdata_s prof_tdata_t; + +/* Option defaults. */ +#ifdef JEMALLOC_PROF +# define PROF_PREFIX_DEFAULT "jeprof" +#else +# define PROF_PREFIX_DEFAULT "" +#endif +#define LG_PROF_SAMPLE_DEFAULT 19 +#define LG_PROF_INTERVAL_DEFAULT -1 + +/* + * Hard limit on stack backtrace depth. The version of prof_backtrace() that + * is based on __builtin_return_address() necessarily has a hard-coded number + * of backtrace frame handlers, and should be kept in sync with this setting. + */ +#define PROF_BT_MAX 128 + +/* Initial hash table size. */ +#define PROF_CKH_MINITEMS 64 + +/* Size of memory buffer to use when writing dump files. */ +#define PROF_DUMP_BUFSIZE 65536 + +/* Size of stack-allocated buffer used by prof_printf(). */ +#define PROF_PRINTF_BUFSIZE 128 + +/* + * Number of mutexes shared among all gctx's. No space is allocated for these + * unless profiling is enabled, so it's okay to over-provision. + */ +#define PROF_NCTX_LOCKS 1024 + +/* + * Number of mutexes shared among all tdata's. No space is allocated for these + * unless profiling is enabled, so it's okay to over-provision. + */ +#define PROF_NTDATA_LOCKS 256 + +/* + * prof_tdata pointers close to NULL are used to encode state information that + * is used for cleaning up during thread shutdown. + */ +#define PROF_TDATA_STATE_REINCARNATED ((prof_tdata_t *)(uintptr_t)1) +#define PROF_TDATA_STATE_PURGATORY ((prof_tdata_t *)(uintptr_t)2) +#define PROF_TDATA_STATE_MAX PROF_TDATA_STATE_PURGATORY + +#endif /* JEMALLOC_INTERNAL_PROF_TYPES_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/public_namespace.sh b/deps/jemalloc/include/jemalloc/internal/public_namespace.sh new file mode 100755 index 000000000..4d415ba01 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/public_namespace.sh @@ -0,0 +1,6 @@ +#!/bin/sh + +for nm in `cat $1` ; do + n=`echo ${nm} |tr ':' ' ' |awk '{print $1}'` + echo "#define je_${n} JEMALLOC_N(${n})" +done diff --git a/deps/jemalloc/include/jemalloc/internal/public_unnamespace.sh b/deps/jemalloc/include/jemalloc/internal/public_unnamespace.sh new file mode 100755 index 000000000..4239d1775 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/public_unnamespace.sh @@ -0,0 +1,6 @@ +#!/bin/sh + +for nm in `cat $1` ; do + n=`echo ${nm} |tr ':' ' ' |awk '{print $1}'` + echo "#undef je_${n}" +done diff --git a/deps/jemalloc/include/jemalloc/internal/ql.h b/deps/jemalloc/include/jemalloc/internal/ql.h new file mode 100644 index 000000000..802904077 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/ql.h @@ -0,0 +1,88 @@ +#ifndef JEMALLOC_INTERNAL_QL_H +#define JEMALLOC_INTERNAL_QL_H + +#include "jemalloc/internal/qr.h" + +/* List definitions. */ +#define ql_head(a_type) \ +struct { \ + a_type *qlh_first; \ +} + +#define ql_head_initializer(a_head) {NULL} + +#define ql_elm(a_type) qr(a_type) + +/* List functions. */ +#define ql_new(a_head) do { \ + (a_head)->qlh_first = NULL; \ +} while (0) + +#define ql_elm_new(a_elm, a_field) qr_new((a_elm), a_field) + +#define ql_first(a_head) ((a_head)->qlh_first) + +#define ql_last(a_head, a_field) \ + ((ql_first(a_head) != NULL) \ + ? qr_prev(ql_first(a_head), a_field) : NULL) + +#define ql_next(a_head, a_elm, a_field) \ + ((ql_last(a_head, a_field) != (a_elm)) \ + ? qr_next((a_elm), a_field) : NULL) + +#define ql_prev(a_head, a_elm, a_field) \ + ((ql_first(a_head) != (a_elm)) ? qr_prev((a_elm), a_field) \ + : NULL) + +#define ql_before_insert(a_head, a_qlelm, a_elm, a_field) do { \ + qr_before_insert((a_qlelm), (a_elm), a_field); \ + if (ql_first(a_head) == (a_qlelm)) { \ + ql_first(a_head) = (a_elm); \ + } \ +} while (0) + +#define ql_after_insert(a_qlelm, a_elm, a_field) \ + qr_after_insert((a_qlelm), (a_elm), a_field) + +#define ql_head_insert(a_head, a_elm, a_field) do { \ + if (ql_first(a_head) != NULL) { \ + qr_before_insert(ql_first(a_head), (a_elm), a_field); \ + } \ + ql_first(a_head) = (a_elm); \ +} while (0) + +#define ql_tail_insert(a_head, a_elm, a_field) do { \ + if (ql_first(a_head) != NULL) { \ + qr_before_insert(ql_first(a_head), (a_elm), a_field); \ + } \ + ql_first(a_head) = qr_next((a_elm), a_field); \ +} while (0) + +#define ql_remove(a_head, a_elm, a_field) do { \ + if (ql_first(a_head) == (a_elm)) { \ + ql_first(a_head) = qr_next(ql_first(a_head), a_field); \ + } \ + if (ql_first(a_head) != (a_elm)) { \ + qr_remove((a_elm), a_field); \ + } else { \ + ql_first(a_head) = NULL; \ + } \ +} while (0) + +#define ql_head_remove(a_head, a_type, a_field) do { \ + a_type *t = ql_first(a_head); \ + ql_remove((a_head), t, a_field); \ +} while (0) + +#define ql_tail_remove(a_head, a_type, a_field) do { \ + a_type *t = ql_last(a_head, a_field); \ + ql_remove((a_head), t, a_field); \ +} while (0) + +#define ql_foreach(a_var, a_head, a_field) \ + qr_foreach((a_var), ql_first(a_head), a_field) + +#define ql_reverse_foreach(a_var, a_head, a_field) \ + qr_reverse_foreach((a_var), ql_first(a_head), a_field) + +#endif /* JEMALLOC_INTERNAL_QL_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/qr.h b/deps/jemalloc/include/jemalloc/internal/qr.h new file mode 100644 index 000000000..1e1056b38 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/qr.h @@ -0,0 +1,72 @@ +#ifndef JEMALLOC_INTERNAL_QR_H +#define JEMALLOC_INTERNAL_QR_H + +/* Ring definitions. */ +#define qr(a_type) \ +struct { \ + a_type *qre_next; \ + a_type *qre_prev; \ +} + +/* Ring functions. */ +#define qr_new(a_qr, a_field) do { \ + (a_qr)->a_field.qre_next = (a_qr); \ + (a_qr)->a_field.qre_prev = (a_qr); \ +} while (0) + +#define qr_next(a_qr, a_field) ((a_qr)->a_field.qre_next) + +#define qr_prev(a_qr, a_field) ((a_qr)->a_field.qre_prev) + +#define qr_before_insert(a_qrelm, a_qr, a_field) do { \ + (a_qr)->a_field.qre_prev = (a_qrelm)->a_field.qre_prev; \ + (a_qr)->a_field.qre_next = (a_qrelm); \ + (a_qr)->a_field.qre_prev->a_field.qre_next = (a_qr); \ + (a_qrelm)->a_field.qre_prev = (a_qr); \ +} while (0) + +#define qr_after_insert(a_qrelm, a_qr, a_field) do { \ + (a_qr)->a_field.qre_next = (a_qrelm)->a_field.qre_next; \ + (a_qr)->a_field.qre_prev = (a_qrelm); \ + (a_qr)->a_field.qre_next->a_field.qre_prev = (a_qr); \ + (a_qrelm)->a_field.qre_next = (a_qr); \ +} while (0) + +#define qr_meld(a_qr_a, a_qr_b, a_type, a_field) do { \ + a_type *t; \ + (a_qr_a)->a_field.qre_prev->a_field.qre_next = (a_qr_b); \ + (a_qr_b)->a_field.qre_prev->a_field.qre_next = (a_qr_a); \ + t = (a_qr_a)->a_field.qre_prev; \ + (a_qr_a)->a_field.qre_prev = (a_qr_b)->a_field.qre_prev; \ + (a_qr_b)->a_field.qre_prev = t; \ +} while (0) + +/* + * qr_meld() and qr_split() are functionally equivalent, so there's no need to + * have two copies of the code. + */ +#define qr_split(a_qr_a, a_qr_b, a_type, a_field) \ + qr_meld((a_qr_a), (a_qr_b), a_type, a_field) + +#define qr_remove(a_qr, a_field) do { \ + (a_qr)->a_field.qre_prev->a_field.qre_next \ + = (a_qr)->a_field.qre_next; \ + (a_qr)->a_field.qre_next->a_field.qre_prev \ + = (a_qr)->a_field.qre_prev; \ + (a_qr)->a_field.qre_next = (a_qr); \ + (a_qr)->a_field.qre_prev = (a_qr); \ +} while (0) + +#define qr_foreach(var, a_qr, a_field) \ + for ((var) = (a_qr); \ + (var) != NULL; \ + (var) = (((var)->a_field.qre_next != (a_qr)) \ + ? (var)->a_field.qre_next : NULL)) + +#define qr_reverse_foreach(var, a_qr, a_field) \ + for ((var) = ((a_qr) != NULL) ? qr_prev(a_qr, a_field) : NULL; \ + (var) != NULL; \ + (var) = (((var) != (a_qr)) \ + ? (var)->a_field.qre_prev : NULL)) + +#endif /* JEMALLOC_INTERNAL_QR_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/quantum.h b/deps/jemalloc/include/jemalloc/internal/quantum.h new file mode 100644 index 000000000..821086e99 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/quantum.h @@ -0,0 +1,77 @@ +#ifndef JEMALLOC_INTERNAL_QUANTUM_H +#define JEMALLOC_INTERNAL_QUANTUM_H + +/* + * Minimum allocation alignment is 2^LG_QUANTUM bytes (ignoring tiny size + * classes). + */ +#ifndef LG_QUANTUM +# if (defined(__i386__) || defined(_M_IX86)) +# define LG_QUANTUM 4 +# endif +# ifdef __ia64__ +# define LG_QUANTUM 4 +# endif +# ifdef __alpha__ +# define LG_QUANTUM 4 +# endif +# if (defined(__sparc64__) || defined(__sparcv9) || defined(__sparc_v9__)) +# define LG_QUANTUM 4 +# endif +# if (defined(__amd64__) || defined(__x86_64__) || defined(_M_X64)) +# define LG_QUANTUM 4 +# endif +# ifdef __arm__ +# define LG_QUANTUM 3 +# endif +# ifdef __aarch64__ +# define LG_QUANTUM 4 +# endif +# ifdef __hppa__ +# define LG_QUANTUM 4 +# endif +# ifdef __m68k__ +# define LG_QUANTUM 3 +# endif +# ifdef __mips__ +# define LG_QUANTUM 3 +# endif +# ifdef __nios2__ +# define LG_QUANTUM 3 +# endif +# ifdef __or1k__ +# define LG_QUANTUM 3 +# endif +# ifdef __powerpc__ +# define LG_QUANTUM 4 +# endif +# if defined(__riscv) || defined(__riscv__) +# define LG_QUANTUM 4 +# endif +# ifdef __s390__ +# define LG_QUANTUM 4 +# endif +# if (defined (__SH3E__) || defined(__SH4_SINGLE__) || defined(__SH4__) || \ + defined(__SH4_SINGLE_ONLY__)) +# define LG_QUANTUM 4 +# endif +# ifdef __tile__ +# define LG_QUANTUM 4 +# endif +# ifdef __le32__ +# define LG_QUANTUM 4 +# endif +# ifndef LG_QUANTUM +# error "Unknown minimum alignment for architecture; specify via " + "--with-lg-quantum" +# endif +#endif + +#define QUANTUM ((size_t)(1U << LG_QUANTUM)) +#define QUANTUM_MASK (QUANTUM - 1) + +/* Return the smallest quantum multiple that is >= a. */ +#define QUANTUM_CEILING(a) \ + (((a) + QUANTUM_MASK) & ~QUANTUM_MASK) + +#endif /* JEMALLOC_INTERNAL_QUANTUM_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/rb.h b/deps/jemalloc/include/jemalloc/internal/rb.h new file mode 100644 index 000000000..47fa5ca99 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/rb.h @@ -0,0 +1,1006 @@ +/*- + ******************************************************************************* + * + * cpp macro implementation of left-leaning 2-3 red-black trees. Parent + * pointers are not used, and color bits are stored in the least significant + * bit of right-child pointers (if RB_COMPACT is defined), thus making node + * linkage as compact as is possible for red-black trees. + * + * Usage: + * + * #include <stdint.h> + * #include <stdbool.h> + * #define NDEBUG // (Optional, see assert(3).) + * #include <assert.h> + * #define RB_COMPACT // (Optional, embed color bits in right-child pointers.) + * #include <rb.h> + * ... + * + ******************************************************************************* + */ + +#ifndef RB_H_ +#define RB_H_ + +#ifndef __PGI +#define RB_COMPACT +#endif + +#ifdef RB_COMPACT +/* Node structure. */ +#define rb_node(a_type) \ +struct { \ + a_type *rbn_left; \ + a_type *rbn_right_red; \ +} +#else +#define rb_node(a_type) \ +struct { \ + a_type *rbn_left; \ + a_type *rbn_right; \ + bool rbn_red; \ +} +#endif + +/* Root structure. */ +#define rb_tree(a_type) \ +struct { \ + a_type *rbt_root; \ +} + +/* Left accessors. */ +#define rbtn_left_get(a_type, a_field, a_node) \ + ((a_node)->a_field.rbn_left) +#define rbtn_left_set(a_type, a_field, a_node, a_left) do { \ + (a_node)->a_field.rbn_left = a_left; \ +} while (0) + +#ifdef RB_COMPACT +/* Right accessors. */ +#define rbtn_right_get(a_type, a_field, a_node) \ + ((a_type *) (((intptr_t) (a_node)->a_field.rbn_right_red) \ + & ((ssize_t)-2))) +#define rbtn_right_set(a_type, a_field, a_node, a_right) do { \ + (a_node)->a_field.rbn_right_red = (a_type *) (((uintptr_t) a_right) \ + | (((uintptr_t) (a_node)->a_field.rbn_right_red) & ((size_t)1))); \ +} while (0) + +/* Color accessors. */ +#define rbtn_red_get(a_type, a_field, a_node) \ + ((bool) (((uintptr_t) (a_node)->a_field.rbn_right_red) \ + & ((size_t)1))) +#define rbtn_color_set(a_type, a_field, a_node, a_red) do { \ + (a_node)->a_field.rbn_right_red = (a_type *) ((((intptr_t) \ + (a_node)->a_field.rbn_right_red) & ((ssize_t)-2)) \ + | ((ssize_t)a_red)); \ +} while (0) +#define rbtn_red_set(a_type, a_field, a_node) do { \ + (a_node)->a_field.rbn_right_red = (a_type *) (((uintptr_t) \ + (a_node)->a_field.rbn_right_red) | ((size_t)1)); \ +} while (0) +#define rbtn_black_set(a_type, a_field, a_node) do { \ + (a_node)->a_field.rbn_right_red = (a_type *) (((intptr_t) \ + (a_node)->a_field.rbn_right_red) & ((ssize_t)-2)); \ +} while (0) + +/* Node initializer. */ +#define rbt_node_new(a_type, a_field, a_rbt, a_node) do { \ + /* Bookkeeping bit cannot be used by node pointer. */ \ + assert(((uintptr_t)(a_node) & 0x1) == 0); \ + rbtn_left_set(a_type, a_field, (a_node), NULL); \ + rbtn_right_set(a_type, a_field, (a_node), NULL); \ + rbtn_red_set(a_type, a_field, (a_node)); \ +} while (0) +#else +/* Right accessors. */ +#define rbtn_right_get(a_type, a_field, a_node) \ + ((a_node)->a_field.rbn_right) +#define rbtn_right_set(a_type, a_field, a_node, a_right) do { \ + (a_node)->a_field.rbn_right = a_right; \ +} while (0) + +/* Color accessors. */ +#define rbtn_red_get(a_type, a_field, a_node) \ + ((a_node)->a_field.rbn_red) +#define rbtn_color_set(a_type, a_field, a_node, a_red) do { \ + (a_node)->a_field.rbn_red = (a_red); \ +} while (0) +#define rbtn_red_set(a_type, a_field, a_node) do { \ + (a_node)->a_field.rbn_red = true; \ +} while (0) +#define rbtn_black_set(a_type, a_field, a_node) do { \ + (a_node)->a_field.rbn_red = false; \ +} while (0) + +/* Node initializer. */ +#define rbt_node_new(a_type, a_field, a_rbt, a_node) do { \ + rbtn_left_set(a_type, a_field, (a_node), NULL); \ + rbtn_right_set(a_type, a_field, (a_node), NULL); \ + rbtn_red_set(a_type, a_field, (a_node)); \ +} while (0) +#endif + +/* Tree initializer. */ +#define rb_new(a_type, a_field, a_rbt) do { \ + (a_rbt)->rbt_root = NULL; \ +} while (0) + +/* Internal utility macros. */ +#define rbtn_first(a_type, a_field, a_rbt, a_root, r_node) do { \ + (r_node) = (a_root); \ + if ((r_node) != NULL) { \ + for (; \ + rbtn_left_get(a_type, a_field, (r_node)) != NULL; \ + (r_node) = rbtn_left_get(a_type, a_field, (r_node))) { \ + } \ + } \ +} while (0) + +#define rbtn_last(a_type, a_field, a_rbt, a_root, r_node) do { \ + (r_node) = (a_root); \ + if ((r_node) != NULL) { \ + for (; rbtn_right_get(a_type, a_field, (r_node)) != NULL; \ + (r_node) = rbtn_right_get(a_type, a_field, (r_node))) { \ + } \ + } \ +} while (0) + +#define rbtn_rotate_left(a_type, a_field, a_node, r_node) do { \ + (r_node) = rbtn_right_get(a_type, a_field, (a_node)); \ + rbtn_right_set(a_type, a_field, (a_node), \ + rbtn_left_get(a_type, a_field, (r_node))); \ + rbtn_left_set(a_type, a_field, (r_node), (a_node)); \ +} while (0) + +#define rbtn_rotate_right(a_type, a_field, a_node, r_node) do { \ + (r_node) = rbtn_left_get(a_type, a_field, (a_node)); \ + rbtn_left_set(a_type, a_field, (a_node), \ + rbtn_right_get(a_type, a_field, (r_node))); \ + rbtn_right_set(a_type, a_field, (r_node), (a_node)); \ +} while (0) + +/* + * The rb_proto() macro generates function prototypes that correspond to the + * functions generated by an equivalently parameterized call to rb_gen(). + */ + +#define rb_proto(a_attr, a_prefix, a_rbt_type, a_type) \ +a_attr void \ +a_prefix##new(a_rbt_type *rbtree); \ +a_attr bool \ +a_prefix##empty(a_rbt_type *rbtree); \ +a_attr a_type * \ +a_prefix##first(a_rbt_type *rbtree); \ +a_attr a_type * \ +a_prefix##last(a_rbt_type *rbtree); \ +a_attr a_type * \ +a_prefix##next(a_rbt_type *rbtree, a_type *node); \ +a_attr a_type * \ +a_prefix##prev(a_rbt_type *rbtree, a_type *node); \ +a_attr a_type * \ +a_prefix##search(a_rbt_type *rbtree, const a_type *key); \ +a_attr a_type * \ +a_prefix##nsearch(a_rbt_type *rbtree, const a_type *key); \ +a_attr a_type * \ +a_prefix##psearch(a_rbt_type *rbtree, const a_type *key); \ +a_attr void \ +a_prefix##insert(a_rbt_type *rbtree, a_type *node); \ +a_attr void \ +a_prefix##remove(a_rbt_type *rbtree, a_type *node); \ +a_attr a_type * \ +a_prefix##iter(a_rbt_type *rbtree, a_type *start, a_type *(*cb)( \ + a_rbt_type *, a_type *, void *), void *arg); \ +a_attr a_type * \ +a_prefix##reverse_iter(a_rbt_type *rbtree, a_type *start, \ + a_type *(*cb)(a_rbt_type *, a_type *, void *), void *arg); \ +a_attr void \ +a_prefix##destroy(a_rbt_type *rbtree, void (*cb)(a_type *, void *), \ + void *arg); + +/* + * The rb_gen() macro generates a type-specific red-black tree implementation, + * based on the above cpp macros. + * + * Arguments: + * + * a_attr : Function attribute for generated functions (ex: static). + * a_prefix : Prefix for generated functions (ex: ex_). + * a_rb_type : Type for red-black tree data structure (ex: ex_t). + * a_type : Type for red-black tree node data structure (ex: ex_node_t). + * a_field : Name of red-black tree node linkage (ex: ex_link). + * a_cmp : Node comparison function name, with the following prototype: + * int (a_cmp *)(a_type *a_node, a_type *a_other); + * ^^^^^^ + * or a_key + * Interpretation of comparison function return values: + * -1 : a_node < a_other + * 0 : a_node == a_other + * 1 : a_node > a_other + * In all cases, the a_node or a_key macro argument is the first + * argument to the comparison function, which makes it possible + * to write comparison functions that treat the first argument + * specially. + * + * Assuming the following setup: + * + * typedef struct ex_node_s ex_node_t; + * struct ex_node_s { + * rb_node(ex_node_t) ex_link; + * }; + * typedef rb_tree(ex_node_t) ex_t; + * rb_gen(static, ex_, ex_t, ex_node_t, ex_link, ex_cmp) + * + * The following API is generated: + * + * static void + * ex_new(ex_t *tree); + * Description: Initialize a red-black tree structure. + * Args: + * tree: Pointer to an uninitialized red-black tree object. + * + * static bool + * ex_empty(ex_t *tree); + * Description: Determine whether tree is empty. + * Args: + * tree: Pointer to an initialized red-black tree object. + * Ret: True if tree is empty, false otherwise. + * + * static ex_node_t * + * ex_first(ex_t *tree); + * static ex_node_t * + * ex_last(ex_t *tree); + * Description: Get the first/last node in tree. + * Args: + * tree: Pointer to an initialized red-black tree object. + * Ret: First/last node in tree, or NULL if tree is empty. + * + * static ex_node_t * + * ex_next(ex_t *tree, ex_node_t *node); + * static ex_node_t * + * ex_prev(ex_t *tree, ex_node_t *node); + * Description: Get node's successor/predecessor. + * Args: + * tree: Pointer to an initialized red-black tree object. + * node: A node in tree. + * Ret: node's successor/predecessor in tree, or NULL if node is + * last/first. + * + * static ex_node_t * + * ex_search(ex_t *tree, const ex_node_t *key); + * Description: Search for node that matches key. + * Args: + * tree: Pointer to an initialized red-black tree object. + * key : Search key. + * Ret: Node in tree that matches key, or NULL if no match. + * + * static ex_node_t * + * ex_nsearch(ex_t *tree, const ex_node_t *key); + * static ex_node_t * + * ex_psearch(ex_t *tree, const ex_node_t *key); + * Description: Search for node that matches key. If no match is found, + * return what would be key's successor/predecessor, were + * key in tree. + * Args: + * tree: Pointer to an initialized red-black tree object. + * key : Search key. + * Ret: Node in tree that matches key, or if no match, hypothetical node's + * successor/predecessor (NULL if no successor/predecessor). + * + * static void + * ex_insert(ex_t *tree, ex_node_t *node); + * Description: Insert node into tree. + * Args: + * tree: Pointer to an initialized red-black tree object. + * node: Node to be inserted into tree. + * + * static void + * ex_remove(ex_t *tree, ex_node_t *node); + * Description: Remove node from tree. + * Args: + * tree: Pointer to an initialized red-black tree object. + * node: Node in tree to be removed. + * + * static ex_node_t * + * ex_iter(ex_t *tree, ex_node_t *start, ex_node_t *(*cb)(ex_t *, + * ex_node_t *, void *), void *arg); + * static ex_node_t * + * ex_reverse_iter(ex_t *tree, ex_node_t *start, ex_node *(*cb)(ex_t *, + * ex_node_t *, void *), void *arg); + * Description: Iterate forward/backward over tree, starting at node. If + * tree is modified, iteration must be immediately + * terminated by the callback function that causes the + * modification. + * Args: + * tree : Pointer to an initialized red-black tree object. + * start: Node at which to start iteration, or NULL to start at + * first/last node. + * cb : Callback function, which is called for each node during + * iteration. Under normal circumstances the callback function + * should return NULL, which causes iteration to continue. If a + * callback function returns non-NULL, iteration is immediately + * terminated and the non-NULL return value is returned by the + * iterator. This is useful for re-starting iteration after + * modifying tree. + * arg : Opaque pointer passed to cb(). + * Ret: NULL if iteration completed, or the non-NULL callback return value + * that caused termination of the iteration. + * + * static void + * ex_destroy(ex_t *tree, void (*cb)(ex_node_t *, void *), void *arg); + * Description: Iterate over the tree with post-order traversal, remove + * each node, and run the callback if non-null. This is + * used for destroying a tree without paying the cost to + * rebalance it. The tree must not be otherwise altered + * during traversal. + * Args: + * tree: Pointer to an initialized red-black tree object. + * cb : Callback function, which, if non-null, is called for each node + * during iteration. There is no way to stop iteration once it + * has begun. + * arg : Opaque pointer passed to cb(). + */ +#define rb_gen(a_attr, a_prefix, a_rbt_type, a_type, a_field, a_cmp) \ +a_attr void \ +a_prefix##new(a_rbt_type *rbtree) { \ + rb_new(a_type, a_field, rbtree); \ +} \ +a_attr bool \ +a_prefix##empty(a_rbt_type *rbtree) { \ + return (rbtree->rbt_root == NULL); \ +} \ +a_attr a_type * \ +a_prefix##first(a_rbt_type *rbtree) { \ + a_type *ret; \ + rbtn_first(a_type, a_field, rbtree, rbtree->rbt_root, ret); \ + return ret; \ +} \ +a_attr a_type * \ +a_prefix##last(a_rbt_type *rbtree) { \ + a_type *ret; \ + rbtn_last(a_type, a_field, rbtree, rbtree->rbt_root, ret); \ + return ret; \ +} \ +a_attr a_type * \ +a_prefix##next(a_rbt_type *rbtree, a_type *node) { \ + a_type *ret; \ + if (rbtn_right_get(a_type, a_field, node) != NULL) { \ + rbtn_first(a_type, a_field, rbtree, rbtn_right_get(a_type, \ + a_field, node), ret); \ + } else { \ + a_type *tnode = rbtree->rbt_root; \ + assert(tnode != NULL); \ + ret = NULL; \ + while (true) { \ + int cmp = (a_cmp)(node, tnode); \ + if (cmp < 0) { \ + ret = tnode; \ + tnode = rbtn_left_get(a_type, a_field, tnode); \ + } else if (cmp > 0) { \ + tnode = rbtn_right_get(a_type, a_field, tnode); \ + } else { \ + break; \ + } \ + assert(tnode != NULL); \ + } \ + } \ + return ret; \ +} \ +a_attr a_type * \ +a_prefix##prev(a_rbt_type *rbtree, a_type *node) { \ + a_type *ret; \ + if (rbtn_left_get(a_type, a_field, node) != NULL) { \ + rbtn_last(a_type, a_field, rbtree, rbtn_left_get(a_type, \ + a_field, node), ret); \ + } else { \ + a_type *tnode = rbtree->rbt_root; \ + assert(tnode != NULL); \ + ret = NULL; \ + while (true) { \ + int cmp = (a_cmp)(node, tnode); \ + if (cmp < 0) { \ + tnode = rbtn_left_get(a_type, a_field, tnode); \ + } else if (cmp > 0) { \ + ret = tnode; \ + tnode = rbtn_right_get(a_type, a_field, tnode); \ + } else { \ + break; \ + } \ + assert(tnode != NULL); \ + } \ + } \ + return ret; \ +} \ +a_attr a_type * \ +a_prefix##search(a_rbt_type *rbtree, const a_type *key) { \ + a_type *ret; \ + int cmp; \ + ret = rbtree->rbt_root; \ + while (ret != NULL \ + && (cmp = (a_cmp)(key, ret)) != 0) { \ + if (cmp < 0) { \ + ret = rbtn_left_get(a_type, a_field, ret); \ + } else { \ + ret = rbtn_right_get(a_type, a_field, ret); \ + } \ + } \ + return ret; \ +} \ +a_attr a_type * \ +a_prefix##nsearch(a_rbt_type *rbtree, const a_type *key) { \ + a_type *ret; \ + a_type *tnode = rbtree->rbt_root; \ + ret = NULL; \ + while (tnode != NULL) { \ + int cmp = (a_cmp)(key, tnode); \ + if (cmp < 0) { \ + ret = tnode; \ + tnode = rbtn_left_get(a_type, a_field, tnode); \ + } else if (cmp > 0) { \ + tnode = rbtn_right_get(a_type, a_field, tnode); \ + } else { \ + ret = tnode; \ + break; \ + } \ + } \ + return ret; \ +} \ +a_attr a_type * \ +a_prefix##psearch(a_rbt_type *rbtree, const a_type *key) { \ + a_type *ret; \ + a_type *tnode = rbtree->rbt_root; \ + ret = NULL; \ + while (tnode != NULL) { \ + int cmp = (a_cmp)(key, tnode); \ + if (cmp < 0) { \ + tnode = rbtn_left_get(a_type, a_field, tnode); \ + } else if (cmp > 0) { \ + ret = tnode; \ + tnode = rbtn_right_get(a_type, a_field, tnode); \ + } else { \ + ret = tnode; \ + break; \ + } \ + } \ + return ret; \ +} \ +a_attr void \ +a_prefix##insert(a_rbt_type *rbtree, a_type *node) { \ + struct { \ + a_type *node; \ + int cmp; \ + } path[sizeof(void *) << 4], *pathp; \ + rbt_node_new(a_type, a_field, rbtree, node); \ + /* Wind. */ \ + path->node = rbtree->rbt_root; \ + for (pathp = path; pathp->node != NULL; pathp++) { \ + int cmp = pathp->cmp = a_cmp(node, pathp->node); \ + assert(cmp != 0); \ + if (cmp < 0) { \ + pathp[1].node = rbtn_left_get(a_type, a_field, \ + pathp->node); \ + } else { \ + pathp[1].node = rbtn_right_get(a_type, a_field, \ + pathp->node); \ + } \ + } \ + pathp->node = node; \ + /* Unwind. */ \ + for (pathp--; (uintptr_t)pathp >= (uintptr_t)path; pathp--) { \ + a_type *cnode = pathp->node; \ + if (pathp->cmp < 0) { \ + a_type *left = pathp[1].node; \ + rbtn_left_set(a_type, a_field, cnode, left); \ + if (rbtn_red_get(a_type, a_field, left)) { \ + a_type *leftleft = rbtn_left_get(a_type, a_field, left);\ + if (leftleft != NULL && rbtn_red_get(a_type, a_field, \ + leftleft)) { \ + /* Fix up 4-node. */ \ + a_type *tnode; \ + rbtn_black_set(a_type, a_field, leftleft); \ + rbtn_rotate_right(a_type, a_field, cnode, tnode); \ + cnode = tnode; \ + } \ + } else { \ + return; \ + } \ + } else { \ + a_type *right = pathp[1].node; \ + rbtn_right_set(a_type, a_field, cnode, right); \ + if (rbtn_red_get(a_type, a_field, right)) { \ + a_type *left = rbtn_left_get(a_type, a_field, cnode); \ + if (left != NULL && rbtn_red_get(a_type, a_field, \ + left)) { \ + /* Split 4-node. */ \ + rbtn_black_set(a_type, a_field, left); \ + rbtn_black_set(a_type, a_field, right); \ + rbtn_red_set(a_type, a_field, cnode); \ + } else { \ + /* Lean left. */ \ + a_type *tnode; \ + bool tred = rbtn_red_get(a_type, a_field, cnode); \ + rbtn_rotate_left(a_type, a_field, cnode, tnode); \ + rbtn_color_set(a_type, a_field, tnode, tred); \ + rbtn_red_set(a_type, a_field, cnode); \ + cnode = tnode; \ + } \ + } else { \ + return; \ + } \ + } \ + pathp->node = cnode; \ + } \ + /* Set root, and make it black. */ \ + rbtree->rbt_root = path->node; \ + rbtn_black_set(a_type, a_field, rbtree->rbt_root); \ +} \ +a_attr void \ +a_prefix##remove(a_rbt_type *rbtree, a_type *node) { \ + struct { \ + a_type *node; \ + int cmp; \ + } *pathp, *nodep, path[sizeof(void *) << 4]; \ + /* Wind. */ \ + nodep = NULL; /* Silence compiler warning. */ \ + path->node = rbtree->rbt_root; \ + for (pathp = path; pathp->node != NULL; pathp++) { \ + int cmp = pathp->cmp = a_cmp(node, pathp->node); \ + if (cmp < 0) { \ + pathp[1].node = rbtn_left_get(a_type, a_field, \ + pathp->node); \ + } else { \ + pathp[1].node = rbtn_right_get(a_type, a_field, \ + pathp->node); \ + if (cmp == 0) { \ + /* Find node's successor, in preparation for swap. */ \ + pathp->cmp = 1; \ + nodep = pathp; \ + for (pathp++; pathp->node != NULL; pathp++) { \ + pathp->cmp = -1; \ + pathp[1].node = rbtn_left_get(a_type, a_field, \ + pathp->node); \ + } \ + break; \ + } \ + } \ + } \ + assert(nodep->node == node); \ + pathp--; \ + if (pathp->node != node) { \ + /* Swap node with its successor. */ \ + bool tred = rbtn_red_get(a_type, a_field, pathp->node); \ + rbtn_color_set(a_type, a_field, pathp->node, \ + rbtn_red_get(a_type, a_field, node)); \ + rbtn_left_set(a_type, a_field, pathp->node, \ + rbtn_left_get(a_type, a_field, node)); \ + /* If node's successor is its right child, the following code */\ + /* will do the wrong thing for the right child pointer. */\ + /* However, it doesn't matter, because the pointer will be */\ + /* properly set when the successor is pruned. */\ + rbtn_right_set(a_type, a_field, pathp->node, \ + rbtn_right_get(a_type, a_field, node)); \ + rbtn_color_set(a_type, a_field, node, tred); \ + /* The pruned leaf node's child pointers are never accessed */\ + /* again, so don't bother setting them to nil. */\ + nodep->node = pathp->node; \ + pathp->node = node; \ + if (nodep == path) { \ + rbtree->rbt_root = nodep->node; \ + } else { \ + if (nodep[-1].cmp < 0) { \ + rbtn_left_set(a_type, a_field, nodep[-1].node, \ + nodep->node); \ + } else { \ + rbtn_right_set(a_type, a_field, nodep[-1].node, \ + nodep->node); \ + } \ + } \ + } else { \ + a_type *left = rbtn_left_get(a_type, a_field, node); \ + if (left != NULL) { \ + /* node has no successor, but it has a left child. */\ + /* Splice node out, without losing the left child. */\ + assert(!rbtn_red_get(a_type, a_field, node)); \ + assert(rbtn_red_get(a_type, a_field, left)); \ + rbtn_black_set(a_type, a_field, left); \ + if (pathp == path) { \ + rbtree->rbt_root = left; \ + } else { \ + if (pathp[-1].cmp < 0) { \ + rbtn_left_set(a_type, a_field, pathp[-1].node, \ + left); \ + } else { \ + rbtn_right_set(a_type, a_field, pathp[-1].node, \ + left); \ + } \ + } \ + return; \ + } else if (pathp == path) { \ + /* The tree only contained one node. */ \ + rbtree->rbt_root = NULL; \ + return; \ + } \ + } \ + if (rbtn_red_get(a_type, a_field, pathp->node)) { \ + /* Prune red node, which requires no fixup. */ \ + assert(pathp[-1].cmp < 0); \ + rbtn_left_set(a_type, a_field, pathp[-1].node, NULL); \ + return; \ + } \ + /* The node to be pruned is black, so unwind until balance is */\ + /* restored. */\ + pathp->node = NULL; \ + for (pathp--; (uintptr_t)pathp >= (uintptr_t)path; pathp--) { \ + assert(pathp->cmp != 0); \ + if (pathp->cmp < 0) { \ + rbtn_left_set(a_type, a_field, pathp->node, \ + pathp[1].node); \ + if (rbtn_red_get(a_type, a_field, pathp->node)) { \ + a_type *right = rbtn_right_get(a_type, a_field, \ + pathp->node); \ + a_type *rightleft = rbtn_left_get(a_type, a_field, \ + right); \ + a_type *tnode; \ + if (rightleft != NULL && rbtn_red_get(a_type, a_field, \ + rightleft)) { \ + /* In the following diagrams, ||, //, and \\ */\ + /* indicate the path to the removed node. */\ + /* */\ + /* || */\ + /* pathp(r) */\ + /* // \ */\ + /* (b) (b) */\ + /* / */\ + /* (r) */\ + /* */\ + rbtn_black_set(a_type, a_field, pathp->node); \ + rbtn_rotate_right(a_type, a_field, right, tnode); \ + rbtn_right_set(a_type, a_field, pathp->node, tnode);\ + rbtn_rotate_left(a_type, a_field, pathp->node, \ + tnode); \ + } else { \ + /* || */\ + /* pathp(r) */\ + /* // \ */\ + /* (b) (b) */\ + /* / */\ + /* (b) */\ + /* */\ + rbtn_rotate_left(a_type, a_field, pathp->node, \ + tnode); \ + } \ + /* Balance restored, but rotation modified subtree */\ + /* root. */\ + assert((uintptr_t)pathp > (uintptr_t)path); \ + if (pathp[-1].cmp < 0) { \ + rbtn_left_set(a_type, a_field, pathp[-1].node, \ + tnode); \ + } else { \ + rbtn_right_set(a_type, a_field, pathp[-1].node, \ + tnode); \ + } \ + return; \ + } else { \ + a_type *right = rbtn_right_get(a_type, a_field, \ + pathp->node); \ + a_type *rightleft = rbtn_left_get(a_type, a_field, \ + right); \ + if (rightleft != NULL && rbtn_red_get(a_type, a_field, \ + rightleft)) { \ + /* || */\ + /* pathp(b) */\ + /* // \ */\ + /* (b) (b) */\ + /* / */\ + /* (r) */\ + a_type *tnode; \ + rbtn_black_set(a_type, a_field, rightleft); \ + rbtn_rotate_right(a_type, a_field, right, tnode); \ + rbtn_right_set(a_type, a_field, pathp->node, tnode);\ + rbtn_rotate_left(a_type, a_field, pathp->node, \ + tnode); \ + /* Balance restored, but rotation modified */\ + /* subtree root, which may actually be the tree */\ + /* root. */\ + if (pathp == path) { \ + /* Set root. */ \ + rbtree->rbt_root = tnode; \ + } else { \ + if (pathp[-1].cmp < 0) { \ + rbtn_left_set(a_type, a_field, \ + pathp[-1].node, tnode); \ + } else { \ + rbtn_right_set(a_type, a_field, \ + pathp[-1].node, tnode); \ + } \ + } \ + return; \ + } else { \ + /* || */\ + /* pathp(b) */\ + /* // \ */\ + /* (b) (b) */\ + /* / */\ + /* (b) */\ + a_type *tnode; \ + rbtn_red_set(a_type, a_field, pathp->node); \ + rbtn_rotate_left(a_type, a_field, pathp->node, \ + tnode); \ + pathp->node = tnode; \ + } \ + } \ + } else { \ + a_type *left; \ + rbtn_right_set(a_type, a_field, pathp->node, \ + pathp[1].node); \ + left = rbtn_left_get(a_type, a_field, pathp->node); \ + if (rbtn_red_get(a_type, a_field, left)) { \ + a_type *tnode; \ + a_type *leftright = rbtn_right_get(a_type, a_field, \ + left); \ + a_type *leftrightleft = rbtn_left_get(a_type, a_field, \ + leftright); \ + if (leftrightleft != NULL && rbtn_red_get(a_type, \ + a_field, leftrightleft)) { \ + /* || */\ + /* pathp(b) */\ + /* / \\ */\ + /* (r) (b) */\ + /* \ */\ + /* (b) */\ + /* / */\ + /* (r) */\ + a_type *unode; \ + rbtn_black_set(a_type, a_field, leftrightleft); \ + rbtn_rotate_right(a_type, a_field, pathp->node, \ + unode); \ + rbtn_rotate_right(a_type, a_field, pathp->node, \ + tnode); \ + rbtn_right_set(a_type, a_field, unode, tnode); \ + rbtn_rotate_left(a_type, a_field, unode, tnode); \ + } else { \ + /* || */\ + /* pathp(b) */\ + /* / \\ */\ + /* (r) (b) */\ + /* \ */\ + /* (b) */\ + /* / */\ + /* (b) */\ + assert(leftright != NULL); \ + rbtn_red_set(a_type, a_field, leftright); \ + rbtn_rotate_right(a_type, a_field, pathp->node, \ + tnode); \ + rbtn_black_set(a_type, a_field, tnode); \ + } \ + /* Balance restored, but rotation modified subtree */\ + /* root, which may actually be the tree root. */\ + if (pathp == path) { \ + /* Set root. */ \ + rbtree->rbt_root = tnode; \ + } else { \ + if (pathp[-1].cmp < 0) { \ + rbtn_left_set(a_type, a_field, pathp[-1].node, \ + tnode); \ + } else { \ + rbtn_right_set(a_type, a_field, pathp[-1].node, \ + tnode); \ + } \ + } \ + return; \ + } else if (rbtn_red_get(a_type, a_field, pathp->node)) { \ + a_type *leftleft = rbtn_left_get(a_type, a_field, left);\ + if (leftleft != NULL && rbtn_red_get(a_type, a_field, \ + leftleft)) { \ + /* || */\ + /* pathp(r) */\ + /* / \\ */\ + /* (b) (b) */\ + /* / */\ + /* (r) */\ + a_type *tnode; \ + rbtn_black_set(a_type, a_field, pathp->node); \ + rbtn_red_set(a_type, a_field, left); \ + rbtn_black_set(a_type, a_field, leftleft); \ + rbtn_rotate_right(a_type, a_field, pathp->node, \ + tnode); \ + /* Balance restored, but rotation modified */\ + /* subtree root. */\ + assert((uintptr_t)pathp > (uintptr_t)path); \ + if (pathp[-1].cmp < 0) { \ + rbtn_left_set(a_type, a_field, pathp[-1].node, \ + tnode); \ + } else { \ + rbtn_right_set(a_type, a_field, pathp[-1].node, \ + tnode); \ + } \ + return; \ + } else { \ + /* || */\ + /* pathp(r) */\ + /* / \\ */\ + /* (b) (b) */\ + /* / */\ + /* (b) */\ + rbtn_red_set(a_type, a_field, left); \ + rbtn_black_set(a_type, a_field, pathp->node); \ + /* Balance restored. */ \ + return; \ + } \ + } else { \ + a_type *leftleft = rbtn_left_get(a_type, a_field, left);\ + if (leftleft != NULL && rbtn_red_get(a_type, a_field, \ + leftleft)) { \ + /* || */\ + /* pathp(b) */\ + /* / \\ */\ + /* (b) (b) */\ + /* / */\ + /* (r) */\ + a_type *tnode; \ + rbtn_black_set(a_type, a_field, leftleft); \ + rbtn_rotate_right(a_type, a_field, pathp->node, \ + tnode); \ + /* Balance restored, but rotation modified */\ + /* subtree root, which may actually be the tree */\ + /* root. */\ + if (pathp == path) { \ + /* Set root. */ \ + rbtree->rbt_root = tnode; \ + } else { \ + if (pathp[-1].cmp < 0) { \ + rbtn_left_set(a_type, a_field, \ + pathp[-1].node, tnode); \ + } else { \ + rbtn_right_set(a_type, a_field, \ + pathp[-1].node, tnode); \ + } \ + } \ + return; \ + } else { \ + /* || */\ + /* pathp(b) */\ + /* / \\ */\ + /* (b) (b) */\ + /* / */\ + /* (b) */\ + rbtn_red_set(a_type, a_field, left); \ + } \ + } \ + } \ + } \ + /* Set root. */ \ + rbtree->rbt_root = path->node; \ + assert(!rbtn_red_get(a_type, a_field, rbtree->rbt_root)); \ +} \ +a_attr a_type * \ +a_prefix##iter_recurse(a_rbt_type *rbtree, a_type *node, \ + a_type *(*cb)(a_rbt_type *, a_type *, void *), void *arg) { \ + if (node == NULL) { \ + return NULL; \ + } else { \ + a_type *ret; \ + if ((ret = a_prefix##iter_recurse(rbtree, rbtn_left_get(a_type, \ + a_field, node), cb, arg)) != NULL || (ret = cb(rbtree, node, \ + arg)) != NULL) { \ + return ret; \ + } \ + return a_prefix##iter_recurse(rbtree, rbtn_right_get(a_type, \ + a_field, node), cb, arg); \ + } \ +} \ +a_attr a_type * \ +a_prefix##iter_start(a_rbt_type *rbtree, a_type *start, a_type *node, \ + a_type *(*cb)(a_rbt_type *, a_type *, void *), void *arg) { \ + int cmp = a_cmp(start, node); \ + if (cmp < 0) { \ + a_type *ret; \ + if ((ret = a_prefix##iter_start(rbtree, start, \ + rbtn_left_get(a_type, a_field, node), cb, arg)) != NULL || \ + (ret = cb(rbtree, node, arg)) != NULL) { \ + return ret; \ + } \ + return a_prefix##iter_recurse(rbtree, rbtn_right_get(a_type, \ + a_field, node), cb, arg); \ + } else if (cmp > 0) { \ + return a_prefix##iter_start(rbtree, start, \ + rbtn_right_get(a_type, a_field, node), cb, arg); \ + } else { \ + a_type *ret; \ + if ((ret = cb(rbtree, node, arg)) != NULL) { \ + return ret; \ + } \ + return a_prefix##iter_recurse(rbtree, rbtn_right_get(a_type, \ + a_field, node), cb, arg); \ + } \ +} \ +a_attr a_type * \ +a_prefix##iter(a_rbt_type *rbtree, a_type *start, a_type *(*cb)( \ + a_rbt_type *, a_type *, void *), void *arg) { \ + a_type *ret; \ + if (start != NULL) { \ + ret = a_prefix##iter_start(rbtree, start, rbtree->rbt_root, \ + cb, arg); \ + } else { \ + ret = a_prefix##iter_recurse(rbtree, rbtree->rbt_root, cb, arg);\ + } \ + return ret; \ +} \ +a_attr a_type * \ +a_prefix##reverse_iter_recurse(a_rbt_type *rbtree, a_type *node, \ + a_type *(*cb)(a_rbt_type *, a_type *, void *), void *arg) { \ + if (node == NULL) { \ + return NULL; \ + } else { \ + a_type *ret; \ + if ((ret = a_prefix##reverse_iter_recurse(rbtree, \ + rbtn_right_get(a_type, a_field, node), cb, arg)) != NULL || \ + (ret = cb(rbtree, node, arg)) != NULL) { \ + return ret; \ + } \ + return a_prefix##reverse_iter_recurse(rbtree, \ + rbtn_left_get(a_type, a_field, node), cb, arg); \ + } \ +} \ +a_attr a_type * \ +a_prefix##reverse_iter_start(a_rbt_type *rbtree, a_type *start, \ + a_type *node, a_type *(*cb)(a_rbt_type *, a_type *, void *), \ + void *arg) { \ + int cmp = a_cmp(start, node); \ + if (cmp > 0) { \ + a_type *ret; \ + if ((ret = a_prefix##reverse_iter_start(rbtree, start, \ + rbtn_right_get(a_type, a_field, node), cb, arg)) != NULL || \ + (ret = cb(rbtree, node, arg)) != NULL) { \ + return ret; \ + } \ + return a_prefix##reverse_iter_recurse(rbtree, \ + rbtn_left_get(a_type, a_field, node), cb, arg); \ + } else if (cmp < 0) { \ + return a_prefix##reverse_iter_start(rbtree, start, \ + rbtn_left_get(a_type, a_field, node), cb, arg); \ + } else { \ + a_type *ret; \ + if ((ret = cb(rbtree, node, arg)) != NULL) { \ + return ret; \ + } \ + return a_prefix##reverse_iter_recurse(rbtree, \ + rbtn_left_get(a_type, a_field, node), cb, arg); \ + } \ +} \ +a_attr a_type * \ +a_prefix##reverse_iter(a_rbt_type *rbtree, a_type *start, \ + a_type *(*cb)(a_rbt_type *, a_type *, void *), void *arg) { \ + a_type *ret; \ + if (start != NULL) { \ + ret = a_prefix##reverse_iter_start(rbtree, start, \ + rbtree->rbt_root, cb, arg); \ + } else { \ + ret = a_prefix##reverse_iter_recurse(rbtree, rbtree->rbt_root, \ + cb, arg); \ + } \ + return ret; \ +} \ +a_attr void \ +a_prefix##destroy_recurse(a_rbt_type *rbtree, a_type *node, void (*cb)( \ + a_type *, void *), void *arg) { \ + if (node == NULL) { \ + return; \ + } \ + a_prefix##destroy_recurse(rbtree, rbtn_left_get(a_type, a_field, \ + node), cb, arg); \ + rbtn_left_set(a_type, a_field, (node), NULL); \ + a_prefix##destroy_recurse(rbtree, rbtn_right_get(a_type, a_field, \ + node), cb, arg); \ + rbtn_right_set(a_type, a_field, (node), NULL); \ + if (cb) { \ + cb(node, arg); \ + } \ +} \ +a_attr void \ +a_prefix##destroy(a_rbt_type *rbtree, void (*cb)(a_type *, void *), \ + void *arg) { \ + a_prefix##destroy_recurse(rbtree, rbtree->rbt_root, cb, arg); \ + rbtree->rbt_root = NULL; \ +} + +#endif /* RB_H_ */ diff --git a/deps/jemalloc/include/jemalloc/internal/rtree.h b/deps/jemalloc/include/jemalloc/internal/rtree.h new file mode 100644 index 000000000..16ccbebee --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/rtree.h @@ -0,0 +1,528 @@ +#ifndef JEMALLOC_INTERNAL_RTREE_H +#define JEMALLOC_INTERNAL_RTREE_H + +#include "jemalloc/internal/atomic.h" +#include "jemalloc/internal/mutex.h" +#include "jemalloc/internal/rtree_tsd.h" +#include "jemalloc/internal/sc.h" +#include "jemalloc/internal/tsd.h" + +/* + * This radix tree implementation is tailored to the singular purpose of + * associating metadata with extents that are currently owned by jemalloc. + * + ******************************************************************************* + */ + +/* Number of high insignificant bits. */ +#define RTREE_NHIB ((1U << (LG_SIZEOF_PTR+3)) - LG_VADDR) +/* Number of low insigificant bits. */ +#define RTREE_NLIB LG_PAGE +/* Number of significant bits. */ +#define RTREE_NSB (LG_VADDR - RTREE_NLIB) +/* Number of levels in radix tree. */ +#if RTREE_NSB <= 10 +# define RTREE_HEIGHT 1 +#elif RTREE_NSB <= 36 +# define RTREE_HEIGHT 2 +#elif RTREE_NSB <= 52 +# define RTREE_HEIGHT 3 +#else +# error Unsupported number of significant virtual address bits +#endif +/* Use compact leaf representation if virtual address encoding allows. */ +#if RTREE_NHIB >= LG_CEIL(SC_NSIZES) +# define RTREE_LEAF_COMPACT +#endif + +/* Needed for initialization only. */ +#define RTREE_LEAFKEY_INVALID ((uintptr_t)1) + +typedef struct rtree_node_elm_s rtree_node_elm_t; +struct rtree_node_elm_s { + atomic_p_t child; /* (rtree_{node,leaf}_elm_t *) */ +}; + +struct rtree_leaf_elm_s { +#ifdef RTREE_LEAF_COMPACT + /* + * Single pointer-width field containing all three leaf element fields. + * For example, on a 64-bit x64 system with 48 significant virtual + * memory address bits, the index, extent, and slab fields are packed as + * such: + * + * x: index + * e: extent + * b: slab + * + * 00000000 xxxxxxxx eeeeeeee [...] eeeeeeee eeee000b + */ + atomic_p_t le_bits; +#else + atomic_p_t le_extent; /* (extent_t *) */ + atomic_u_t le_szind; /* (szind_t) */ + atomic_b_t le_slab; /* (bool) */ +#endif +}; + +typedef struct rtree_level_s rtree_level_t; +struct rtree_level_s { + /* Number of key bits distinguished by this level. */ + unsigned bits; + /* + * Cumulative number of key bits distinguished by traversing to + * corresponding tree level. + */ + unsigned cumbits; +}; + +typedef struct rtree_s rtree_t; +struct rtree_s { + malloc_mutex_t init_lock; + /* Number of elements based on rtree_levels[0].bits. */ +#if RTREE_HEIGHT > 1 + rtree_node_elm_t root[1U << (RTREE_NSB/RTREE_HEIGHT)]; +#else + rtree_leaf_elm_t root[1U << (RTREE_NSB/RTREE_HEIGHT)]; +#endif +}; + +/* + * Split the bits into one to three partitions depending on number of + * significant bits. It the number of bits does not divide evenly into the + * number of levels, place one remainder bit per level starting at the leaf + * level. + */ +static const rtree_level_t rtree_levels[] = { +#if RTREE_HEIGHT == 1 + {RTREE_NSB, RTREE_NHIB + RTREE_NSB} +#elif RTREE_HEIGHT == 2 + {RTREE_NSB/2, RTREE_NHIB + RTREE_NSB/2}, + {RTREE_NSB/2 + RTREE_NSB%2, RTREE_NHIB + RTREE_NSB} +#elif RTREE_HEIGHT == 3 + {RTREE_NSB/3, RTREE_NHIB + RTREE_NSB/3}, + {RTREE_NSB/3 + RTREE_NSB%3/2, + RTREE_NHIB + RTREE_NSB/3*2 + RTREE_NSB%3/2}, + {RTREE_NSB/3 + RTREE_NSB%3 - RTREE_NSB%3/2, RTREE_NHIB + RTREE_NSB} +#else +# error Unsupported rtree height +#endif +}; + +bool rtree_new(rtree_t *rtree, bool zeroed); + +typedef rtree_node_elm_t *(rtree_node_alloc_t)(tsdn_t *, rtree_t *, size_t); +extern rtree_node_alloc_t *JET_MUTABLE rtree_node_alloc; + +typedef rtree_leaf_elm_t *(rtree_leaf_alloc_t)(tsdn_t *, rtree_t *, size_t); +extern rtree_leaf_alloc_t *JET_MUTABLE rtree_leaf_alloc; + +typedef void (rtree_node_dalloc_t)(tsdn_t *, rtree_t *, rtree_node_elm_t *); +extern rtree_node_dalloc_t *JET_MUTABLE rtree_node_dalloc; + +typedef void (rtree_leaf_dalloc_t)(tsdn_t *, rtree_t *, rtree_leaf_elm_t *); +extern rtree_leaf_dalloc_t *JET_MUTABLE rtree_leaf_dalloc; +#ifdef JEMALLOC_JET +void rtree_delete(tsdn_t *tsdn, rtree_t *rtree); +#endif +rtree_leaf_elm_t *rtree_leaf_elm_lookup_hard(tsdn_t *tsdn, rtree_t *rtree, + rtree_ctx_t *rtree_ctx, uintptr_t key, bool dependent, bool init_missing); + +JEMALLOC_ALWAYS_INLINE uintptr_t +rtree_leafkey(uintptr_t key) { + unsigned ptrbits = ZU(1) << (LG_SIZEOF_PTR+3); + unsigned cumbits = (rtree_levels[RTREE_HEIGHT-1].cumbits - + rtree_levels[RTREE_HEIGHT-1].bits); + unsigned maskbits = ptrbits - cumbits; + uintptr_t mask = ~((ZU(1) << maskbits) - 1); + return (key & mask); +} + +JEMALLOC_ALWAYS_INLINE size_t +rtree_cache_direct_map(uintptr_t key) { + unsigned ptrbits = ZU(1) << (LG_SIZEOF_PTR+3); + unsigned cumbits = (rtree_levels[RTREE_HEIGHT-1].cumbits - + rtree_levels[RTREE_HEIGHT-1].bits); + unsigned maskbits = ptrbits - cumbits; + return (size_t)((key >> maskbits) & (RTREE_CTX_NCACHE - 1)); +} + +JEMALLOC_ALWAYS_INLINE uintptr_t +rtree_subkey(uintptr_t key, unsigned level) { + unsigned ptrbits = ZU(1) << (LG_SIZEOF_PTR+3); + unsigned cumbits = rtree_levels[level].cumbits; + unsigned shiftbits = ptrbits - cumbits; + unsigned maskbits = rtree_levels[level].bits; + uintptr_t mask = (ZU(1) << maskbits) - 1; + return ((key >> shiftbits) & mask); +} + +/* + * Atomic getters. + * + * dependent: Reading a value on behalf of a pointer to a valid allocation + * is guaranteed to be a clean read even without synchronization, + * because the rtree update became visible in memory before the + * pointer came into existence. + * !dependent: An arbitrary read, e.g. on behalf of ivsalloc(), may not be + * dependent on a previous rtree write, which means a stale read + * could result if synchronization were omitted here. + */ +# ifdef RTREE_LEAF_COMPACT +JEMALLOC_ALWAYS_INLINE uintptr_t +rtree_leaf_elm_bits_read(tsdn_t *tsdn, rtree_t *rtree, + rtree_leaf_elm_t *elm, bool dependent) { + return (uintptr_t)atomic_load_p(&elm->le_bits, dependent + ? ATOMIC_RELAXED : ATOMIC_ACQUIRE); +} + +JEMALLOC_ALWAYS_INLINE extent_t * +rtree_leaf_elm_bits_extent_get(uintptr_t bits) { +# ifdef __aarch64__ + /* + * aarch64 doesn't sign extend the highest virtual address bit to set + * the higher ones. Instead, the high bits gets zeroed. + */ + uintptr_t high_bit_mask = ((uintptr_t)1 << LG_VADDR) - 1; + /* Mask off the slab bit. */ + uintptr_t low_bit_mask = ~(uintptr_t)1; + uintptr_t mask = high_bit_mask & low_bit_mask; + return (extent_t *)(bits & mask); +# else + /* Restore sign-extended high bits, mask slab bit. */ + return (extent_t *)((uintptr_t)((intptr_t)(bits << RTREE_NHIB) >> + RTREE_NHIB) & ~((uintptr_t)0x1)); +# endif +} + +JEMALLOC_ALWAYS_INLINE szind_t +rtree_leaf_elm_bits_szind_get(uintptr_t bits) { + return (szind_t)(bits >> LG_VADDR); +} + +JEMALLOC_ALWAYS_INLINE bool +rtree_leaf_elm_bits_slab_get(uintptr_t bits) { + return (bool)(bits & (uintptr_t)0x1); +} + +# endif + +JEMALLOC_ALWAYS_INLINE extent_t * +rtree_leaf_elm_extent_read(tsdn_t *tsdn, rtree_t *rtree, + rtree_leaf_elm_t *elm, bool dependent) { +#ifdef RTREE_LEAF_COMPACT + uintptr_t bits = rtree_leaf_elm_bits_read(tsdn, rtree, elm, dependent); + return rtree_leaf_elm_bits_extent_get(bits); +#else + extent_t *extent = (extent_t *)atomic_load_p(&elm->le_extent, dependent + ? ATOMIC_RELAXED : ATOMIC_ACQUIRE); + return extent; +#endif +} + +JEMALLOC_ALWAYS_INLINE szind_t +rtree_leaf_elm_szind_read(tsdn_t *tsdn, rtree_t *rtree, + rtree_leaf_elm_t *elm, bool dependent) { +#ifdef RTREE_LEAF_COMPACT + uintptr_t bits = rtree_leaf_elm_bits_read(tsdn, rtree, elm, dependent); + return rtree_leaf_elm_bits_szind_get(bits); +#else + return (szind_t)atomic_load_u(&elm->le_szind, dependent ? ATOMIC_RELAXED + : ATOMIC_ACQUIRE); +#endif +} + +JEMALLOC_ALWAYS_INLINE bool +rtree_leaf_elm_slab_read(tsdn_t *tsdn, rtree_t *rtree, + rtree_leaf_elm_t *elm, bool dependent) { +#ifdef RTREE_LEAF_COMPACT + uintptr_t bits = rtree_leaf_elm_bits_read(tsdn, rtree, elm, dependent); + return rtree_leaf_elm_bits_slab_get(bits); +#else + return atomic_load_b(&elm->le_slab, dependent ? ATOMIC_RELAXED : + ATOMIC_ACQUIRE); +#endif +} + +static inline void +rtree_leaf_elm_extent_write(tsdn_t *tsdn, rtree_t *rtree, + rtree_leaf_elm_t *elm, extent_t *extent) { +#ifdef RTREE_LEAF_COMPACT + uintptr_t old_bits = rtree_leaf_elm_bits_read(tsdn, rtree, elm, true); + uintptr_t bits = ((uintptr_t)rtree_leaf_elm_bits_szind_get(old_bits) << + LG_VADDR) | ((uintptr_t)extent & (((uintptr_t)0x1 << LG_VADDR) - 1)) + | ((uintptr_t)rtree_leaf_elm_bits_slab_get(old_bits)); + atomic_store_p(&elm->le_bits, (void *)bits, ATOMIC_RELEASE); +#else + atomic_store_p(&elm->le_extent, extent, ATOMIC_RELEASE); +#endif +} + +static inline void +rtree_leaf_elm_szind_write(tsdn_t *tsdn, rtree_t *rtree, + rtree_leaf_elm_t *elm, szind_t szind) { + assert(szind <= SC_NSIZES); + +#ifdef RTREE_LEAF_COMPACT + uintptr_t old_bits = rtree_leaf_elm_bits_read(tsdn, rtree, elm, + true); + uintptr_t bits = ((uintptr_t)szind << LG_VADDR) | + ((uintptr_t)rtree_leaf_elm_bits_extent_get(old_bits) & + (((uintptr_t)0x1 << LG_VADDR) - 1)) | + ((uintptr_t)rtree_leaf_elm_bits_slab_get(old_bits)); + atomic_store_p(&elm->le_bits, (void *)bits, ATOMIC_RELEASE); +#else + atomic_store_u(&elm->le_szind, szind, ATOMIC_RELEASE); +#endif +} + +static inline void +rtree_leaf_elm_slab_write(tsdn_t *tsdn, rtree_t *rtree, + rtree_leaf_elm_t *elm, bool slab) { +#ifdef RTREE_LEAF_COMPACT + uintptr_t old_bits = rtree_leaf_elm_bits_read(tsdn, rtree, elm, + true); + uintptr_t bits = ((uintptr_t)rtree_leaf_elm_bits_szind_get(old_bits) << + LG_VADDR) | ((uintptr_t)rtree_leaf_elm_bits_extent_get(old_bits) & + (((uintptr_t)0x1 << LG_VADDR) - 1)) | ((uintptr_t)slab); + atomic_store_p(&elm->le_bits, (void *)bits, ATOMIC_RELEASE); +#else + atomic_store_b(&elm->le_slab, slab, ATOMIC_RELEASE); +#endif +} + +static inline void +rtree_leaf_elm_write(tsdn_t *tsdn, rtree_t *rtree, + rtree_leaf_elm_t *elm, extent_t *extent, szind_t szind, bool slab) { +#ifdef RTREE_LEAF_COMPACT + uintptr_t bits = ((uintptr_t)szind << LG_VADDR) | + ((uintptr_t)extent & (((uintptr_t)0x1 << LG_VADDR) - 1)) | + ((uintptr_t)slab); + atomic_store_p(&elm->le_bits, (void *)bits, ATOMIC_RELEASE); +#else + rtree_leaf_elm_slab_write(tsdn, rtree, elm, slab); + rtree_leaf_elm_szind_write(tsdn, rtree, elm, szind); + /* + * Write extent last, since the element is atomically considered valid + * as soon as the extent field is non-NULL. + */ + rtree_leaf_elm_extent_write(tsdn, rtree, elm, extent); +#endif +} + +static inline void +rtree_leaf_elm_szind_slab_update(tsdn_t *tsdn, rtree_t *rtree, + rtree_leaf_elm_t *elm, szind_t szind, bool slab) { + assert(!slab || szind < SC_NBINS); + + /* + * The caller implicitly assures that it is the only writer to the szind + * and slab fields, and that the extent field cannot currently change. + */ + rtree_leaf_elm_slab_write(tsdn, rtree, elm, slab); + rtree_leaf_elm_szind_write(tsdn, rtree, elm, szind); +} + +JEMALLOC_ALWAYS_INLINE rtree_leaf_elm_t * +rtree_leaf_elm_lookup(tsdn_t *tsdn, rtree_t *rtree, rtree_ctx_t *rtree_ctx, + uintptr_t key, bool dependent, bool init_missing) { + assert(key != 0); + assert(!dependent || !init_missing); + + size_t slot = rtree_cache_direct_map(key); + uintptr_t leafkey = rtree_leafkey(key); + assert(leafkey != RTREE_LEAFKEY_INVALID); + + /* Fast path: L1 direct mapped cache. */ + if (likely(rtree_ctx->cache[slot].leafkey == leafkey)) { + rtree_leaf_elm_t *leaf = rtree_ctx->cache[slot].leaf; + assert(leaf != NULL); + uintptr_t subkey = rtree_subkey(key, RTREE_HEIGHT-1); + return &leaf[subkey]; + } + /* + * Search the L2 LRU cache. On hit, swap the matching element into the + * slot in L1 cache, and move the position in L2 up by 1. + */ +#define RTREE_CACHE_CHECK_L2(i) do { \ + if (likely(rtree_ctx->l2_cache[i].leafkey == leafkey)) { \ + rtree_leaf_elm_t *leaf = rtree_ctx->l2_cache[i].leaf; \ + assert(leaf != NULL); \ + if (i > 0) { \ + /* Bubble up by one. */ \ + rtree_ctx->l2_cache[i].leafkey = \ + rtree_ctx->l2_cache[i - 1].leafkey; \ + rtree_ctx->l2_cache[i].leaf = \ + rtree_ctx->l2_cache[i - 1].leaf; \ + rtree_ctx->l2_cache[i - 1].leafkey = \ + rtree_ctx->cache[slot].leafkey; \ + rtree_ctx->l2_cache[i - 1].leaf = \ + rtree_ctx->cache[slot].leaf; \ + } else { \ + rtree_ctx->l2_cache[0].leafkey = \ + rtree_ctx->cache[slot].leafkey; \ + rtree_ctx->l2_cache[0].leaf = \ + rtree_ctx->cache[slot].leaf; \ + } \ + rtree_ctx->cache[slot].leafkey = leafkey; \ + rtree_ctx->cache[slot].leaf = leaf; \ + uintptr_t subkey = rtree_subkey(key, RTREE_HEIGHT-1); \ + return &leaf[subkey]; \ + } \ +} while (0) + /* Check the first cache entry. */ + RTREE_CACHE_CHECK_L2(0); + /* Search the remaining cache elements. */ + for (unsigned i = 1; i < RTREE_CTX_NCACHE_L2; i++) { + RTREE_CACHE_CHECK_L2(i); + } +#undef RTREE_CACHE_CHECK_L2 + + return rtree_leaf_elm_lookup_hard(tsdn, rtree, rtree_ctx, key, + dependent, init_missing); +} + +static inline bool +rtree_write(tsdn_t *tsdn, rtree_t *rtree, rtree_ctx_t *rtree_ctx, uintptr_t key, + extent_t *extent, szind_t szind, bool slab) { + /* Use rtree_clear() to set the extent to NULL. */ + assert(extent != NULL); + + rtree_leaf_elm_t *elm = rtree_leaf_elm_lookup(tsdn, rtree, rtree_ctx, + key, false, true); + if (elm == NULL) { + return true; + } + + assert(rtree_leaf_elm_extent_read(tsdn, rtree, elm, false) == NULL); + rtree_leaf_elm_write(tsdn, rtree, elm, extent, szind, slab); + + return false; +} + +JEMALLOC_ALWAYS_INLINE rtree_leaf_elm_t * +rtree_read(tsdn_t *tsdn, rtree_t *rtree, rtree_ctx_t *rtree_ctx, uintptr_t key, + bool dependent) { + rtree_leaf_elm_t *elm = rtree_leaf_elm_lookup(tsdn, rtree, rtree_ctx, + key, dependent, false); + if (!dependent && elm == NULL) { + return NULL; + } + assert(elm != NULL); + return elm; +} + +JEMALLOC_ALWAYS_INLINE extent_t * +rtree_extent_read(tsdn_t *tsdn, rtree_t *rtree, rtree_ctx_t *rtree_ctx, + uintptr_t key, bool dependent) { + rtree_leaf_elm_t *elm = rtree_read(tsdn, rtree, rtree_ctx, key, + dependent); + if (!dependent && elm == NULL) { + return NULL; + } + return rtree_leaf_elm_extent_read(tsdn, rtree, elm, dependent); +} + +JEMALLOC_ALWAYS_INLINE szind_t +rtree_szind_read(tsdn_t *tsdn, rtree_t *rtree, rtree_ctx_t *rtree_ctx, + uintptr_t key, bool dependent) { + rtree_leaf_elm_t *elm = rtree_read(tsdn, rtree, rtree_ctx, key, + dependent); + if (!dependent && elm == NULL) { + return SC_NSIZES; + } + return rtree_leaf_elm_szind_read(tsdn, rtree, elm, dependent); +} + +/* + * rtree_slab_read() is intentionally omitted because slab is always read in + * conjunction with szind, which makes rtree_szind_slab_read() a better choice. + */ + +JEMALLOC_ALWAYS_INLINE bool +rtree_extent_szind_read(tsdn_t *tsdn, rtree_t *rtree, rtree_ctx_t *rtree_ctx, + uintptr_t key, bool dependent, extent_t **r_extent, szind_t *r_szind) { + rtree_leaf_elm_t *elm = rtree_read(tsdn, rtree, rtree_ctx, key, + dependent); + if (!dependent && elm == NULL) { + return true; + } + *r_extent = rtree_leaf_elm_extent_read(tsdn, rtree, elm, dependent); + *r_szind = rtree_leaf_elm_szind_read(tsdn, rtree, elm, dependent); + return false; +} + +/* + * Try to read szind_slab from the L1 cache. Returns true on a hit, + * and fills in r_szind and r_slab. Otherwise returns false. + * + * Key is allowed to be NULL in order to save an extra branch on the + * fastpath. returns false in this case. + */ +JEMALLOC_ALWAYS_INLINE bool +rtree_szind_slab_read_fast(tsdn_t *tsdn, rtree_t *rtree, rtree_ctx_t *rtree_ctx, + uintptr_t key, szind_t *r_szind, bool *r_slab) { + rtree_leaf_elm_t *elm; + + size_t slot = rtree_cache_direct_map(key); + uintptr_t leafkey = rtree_leafkey(key); + assert(leafkey != RTREE_LEAFKEY_INVALID); + + if (likely(rtree_ctx->cache[slot].leafkey == leafkey)) { + rtree_leaf_elm_t *leaf = rtree_ctx->cache[slot].leaf; + assert(leaf != NULL); + uintptr_t subkey = rtree_subkey(key, RTREE_HEIGHT-1); + elm = &leaf[subkey]; + +#ifdef RTREE_LEAF_COMPACT + uintptr_t bits = rtree_leaf_elm_bits_read(tsdn, rtree, + elm, true); + *r_szind = rtree_leaf_elm_bits_szind_get(bits); + *r_slab = rtree_leaf_elm_bits_slab_get(bits); +#else + *r_szind = rtree_leaf_elm_szind_read(tsdn, rtree, elm, true); + *r_slab = rtree_leaf_elm_slab_read(tsdn, rtree, elm, true); +#endif + return true; + } else { + return false; + } +} +JEMALLOC_ALWAYS_INLINE bool +rtree_szind_slab_read(tsdn_t *tsdn, rtree_t *rtree, rtree_ctx_t *rtree_ctx, + uintptr_t key, bool dependent, szind_t *r_szind, bool *r_slab) { + rtree_leaf_elm_t *elm = rtree_read(tsdn, rtree, rtree_ctx, key, + dependent); + if (!dependent && elm == NULL) { + return true; + } +#ifdef RTREE_LEAF_COMPACT + uintptr_t bits = rtree_leaf_elm_bits_read(tsdn, rtree, elm, dependent); + *r_szind = rtree_leaf_elm_bits_szind_get(bits); + *r_slab = rtree_leaf_elm_bits_slab_get(bits); +#else + *r_szind = rtree_leaf_elm_szind_read(tsdn, rtree, elm, dependent); + *r_slab = rtree_leaf_elm_slab_read(tsdn, rtree, elm, dependent); +#endif + return false; +} + +static inline void +rtree_szind_slab_update(tsdn_t *tsdn, rtree_t *rtree, rtree_ctx_t *rtree_ctx, + uintptr_t key, szind_t szind, bool slab) { + assert(!slab || szind < SC_NBINS); + + rtree_leaf_elm_t *elm = rtree_read(tsdn, rtree, rtree_ctx, key, true); + rtree_leaf_elm_szind_slab_update(tsdn, rtree, elm, szind, slab); +} + +static inline void +rtree_clear(tsdn_t *tsdn, rtree_t *rtree, rtree_ctx_t *rtree_ctx, + uintptr_t key) { + rtree_leaf_elm_t *elm = rtree_read(tsdn, rtree, rtree_ctx, key, true); + assert(rtree_leaf_elm_extent_read(tsdn, rtree, elm, false) != + NULL); + rtree_leaf_elm_write(tsdn, rtree, elm, NULL, SC_NSIZES, false); +} + +#endif /* JEMALLOC_INTERNAL_RTREE_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/rtree_tsd.h b/deps/jemalloc/include/jemalloc/internal/rtree_tsd.h new file mode 100644 index 000000000..562e29297 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/rtree_tsd.h @@ -0,0 +1,50 @@ +#ifndef JEMALLOC_INTERNAL_RTREE_CTX_H +#define JEMALLOC_INTERNAL_RTREE_CTX_H + +/* + * Number of leafkey/leaf pairs to cache in L1 and L2 level respectively. Each + * entry supports an entire leaf, so the cache hit rate is typically high even + * with a small number of entries. In rare cases extent activity will straddle + * the boundary between two leaf nodes. Furthermore, an arena may use a + * combination of dss and mmap. Note that as memory usage grows past the amount + * that this cache can directly cover, the cache will become less effective if + * locality of reference is low, but the consequence is merely cache misses + * while traversing the tree nodes. + * + * The L1 direct mapped cache offers consistent and low cost on cache hit. + * However collision could affect hit rate negatively. This is resolved by + * combining with a L2 LRU cache, which requires linear search and re-ordering + * on access but suffers no collision. Note that, the cache will itself suffer + * cache misses if made overly large, plus the cost of linear search in the LRU + * cache. + */ +#define RTREE_CTX_LG_NCACHE 4 +#define RTREE_CTX_NCACHE (1 << RTREE_CTX_LG_NCACHE) +#define RTREE_CTX_NCACHE_L2 8 + +/* + * Zero initializer required for tsd initialization only. Proper initialization + * done via rtree_ctx_data_init(). + */ +#define RTREE_CTX_ZERO_INITIALIZER {{{0, 0}}, {{0, 0}}} + + +typedef struct rtree_leaf_elm_s rtree_leaf_elm_t; + +typedef struct rtree_ctx_cache_elm_s rtree_ctx_cache_elm_t; +struct rtree_ctx_cache_elm_s { + uintptr_t leafkey; + rtree_leaf_elm_t *leaf; +}; + +typedef struct rtree_ctx_s rtree_ctx_t; +struct rtree_ctx_s { + /* Direct mapped cache. */ + rtree_ctx_cache_elm_t cache[RTREE_CTX_NCACHE]; + /* L2 LRU cache. */ + rtree_ctx_cache_elm_t l2_cache[RTREE_CTX_NCACHE_L2]; +}; + +void rtree_ctx_data_init(rtree_ctx_t *ctx); + +#endif /* JEMALLOC_INTERNAL_RTREE_CTX_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/safety_check.h b/deps/jemalloc/include/jemalloc/internal/safety_check.h new file mode 100644 index 000000000..53339ac12 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/safety_check.h @@ -0,0 +1,26 @@ +#ifndef JEMALLOC_INTERNAL_SAFETY_CHECK_H +#define JEMALLOC_INTERNAL_SAFETY_CHECK_H + +void safety_check_fail(const char *format, ...); +/* Can set to NULL for a default. */ +void safety_check_set_abort(void (*abort_fn)()); + +JEMALLOC_ALWAYS_INLINE void +safety_check_set_redzone(void *ptr, size_t usize, size_t bumped_usize) { + assert(usize < bumped_usize); + for (size_t i = usize; i < bumped_usize && i < usize + 32; ++i) { + *((unsigned char *)ptr + i) = 0xBC; + } +} + +JEMALLOC_ALWAYS_INLINE void +safety_check_verify_redzone(const void *ptr, size_t usize, size_t bumped_usize) +{ + for (size_t i = usize; i < bumped_usize && i < usize + 32; ++i) { + if (unlikely(*((unsigned char *)ptr + i) != 0xBC)) { + safety_check_fail("Use after free error\n"); + } + } +} + +#endif /*JEMALLOC_INTERNAL_SAFETY_CHECK_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/sc.h b/deps/jemalloc/include/jemalloc/internal/sc.h new file mode 100644 index 000000000..9a099d8b6 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/sc.h @@ -0,0 +1,333 @@ +#ifndef JEMALLOC_INTERNAL_SC_H +#define JEMALLOC_INTERNAL_SC_H + +#include "jemalloc/internal/jemalloc_internal_types.h" + +/* + * Size class computations: + * + * These are a little tricky; we'll first start by describing how things + * generally work, and then describe some of the details. + * + * Ignore the first few size classes for a moment. We can then split all the + * remaining size classes into groups. The size classes in a group are spaced + * such that they cover allocation request sizes in a power-of-2 range. The + * power of two is called the base of the group, and the size classes in it + * satisfy allocations in the half-open range (base, base * 2]. There are + * SC_NGROUP size classes in each group, equally spaced in the range, so that + * each one covers allocations for base / SC_NGROUP possible allocation sizes. + * We call that value (base / SC_NGROUP) the delta of the group. Each size class + * is delta larger than the one before it (including the initial size class in a + * group, which is delta larger than base, the largest size class in the + * previous group). + * To make the math all work out nicely, we require that SC_NGROUP is a power of + * two, and define it in terms of SC_LG_NGROUP. We'll often talk in terms of + * lg_base and lg_delta. For each of these groups then, we have that + * lg_delta == lg_base - SC_LG_NGROUP. + * The size classes in a group with a given lg_base and lg_delta (which, recall, + * can be computed from lg_base for these groups) are therefore: + * base + 1 * delta + * which covers allocations in (base, base + 1 * delta] + * base + 2 * delta + * which covers allocations in (base + 1 * delta, base + 2 * delta]. + * base + 3 * delta + * which covers allocations in (base + 2 * delta, base + 3 * delta]. + * ... + * base + SC_NGROUP * delta ( == 2 * base) + * which covers allocations in (base + (SC_NGROUP - 1) * delta, 2 * base]. + * (Note that currently SC_NGROUP is always 4, so the "..." is empty in + * practice.) + * Note that the last size class in the group is the next power of two (after + * base), so that we've set up the induction correctly for the next group's + * selection of delta. + * + * Now, let's start considering the first few size classes. Two extra constants + * come into play here: LG_QUANTUM and SC_LG_TINY_MIN. LG_QUANTUM ensures + * correct platform alignment; all objects of size (1 << LG_QUANTUM) or larger + * are at least (1 << LG_QUANTUM) aligned; this can be used to ensure that we + * never return improperly aligned memory, by making (1 << LG_QUANTUM) equal the + * highest required alignment of a platform. For allocation sizes smaller than + * (1 << LG_QUANTUM) though, we can be more relaxed (since we don't support + * platforms with types with alignment larger than their size). To allow such + * allocations (without wasting space unnecessarily), we introduce tiny size + * classes; one per power of two, up until we hit the quantum size. There are + * therefore LG_QUANTUM - SC_LG_TINY_MIN such size classes. + * + * Next, we have a size class of size (1 << LG_QUANTUM). This can't be the + * start of a group in the sense we described above (covering a power of two + * range) since, if we divided into it to pick a value of delta, we'd get a + * delta smaller than (1 << LG_QUANTUM) for sizes >= (1 << LG_QUANTUM), which + * is against the rules. + * + * The first base we can divide by SC_NGROUP while still being at least + * (1 << LG_QUANTUM) is SC_NGROUP * (1 << LG_QUANTUM). We can get there by + * having SC_NGROUP size classes, spaced (1 << LG_QUANTUM) apart. These size + * classes are: + * 1 * (1 << LG_QUANTUM) + * 2 * (1 << LG_QUANTUM) + * 3 * (1 << LG_QUANTUM) + * ... (although, as above, this "..." is empty in practice) + * SC_NGROUP * (1 << LG_QUANTUM). + * + * There are SC_NGROUP of these size classes, so we can regard it as a sort of + * pseudo-group, even though it spans multiple powers of 2, is divided + * differently, and both starts and ends on a power of 2 (as opposed to just + * ending). SC_NGROUP is itself a power of two, so the first group after the + * pseudo-group has the power-of-two base SC_NGROUP * (1 << LG_QUANTUM), for a + * lg_base of LG_QUANTUM + SC_LG_NGROUP. We can divide this base into SC_NGROUP + * sizes without violating our LG_QUANTUM requirements, so we can safely set + * lg_delta = lg_base - SC_LG_GROUP (== LG_QUANTUM). + * + * So, in order, the size classes are: + * + * Tiny size classes: + * - Count: LG_QUANTUM - SC_LG_TINY_MIN. + * - Sizes: + * 1 << SC_LG_TINY_MIN + * 1 << (SC_LG_TINY_MIN + 1) + * 1 << (SC_LG_TINY_MIN + 2) + * ... + * 1 << (LG_QUANTUM - 1) + * + * Initial pseudo-group: + * - Count: SC_NGROUP + * - Sizes: + * 1 * (1 << LG_QUANTUM) + * 2 * (1 << LG_QUANTUM) + * 3 * (1 << LG_QUANTUM) + * ... + * SC_NGROUP * (1 << LG_QUANTUM) + * + * Regular group 0: + * - Count: SC_NGROUP + * - Sizes: + * (relative to lg_base of LG_QUANTUM + SC_LG_NGROUP and lg_delta of + * lg_base - SC_LG_NGROUP) + * (1 << lg_base) + 1 * (1 << lg_delta) + * (1 << lg_base) + 2 * (1 << lg_delta) + * (1 << lg_base) + 3 * (1 << lg_delta) + * ... + * (1 << lg_base) + SC_NGROUP * (1 << lg_delta) [ == (1 << (lg_base + 1)) ] + * + * Regular group 1: + * - Count: SC_NGROUP + * - Sizes: + * (relative to lg_base of LG_QUANTUM + SC_LG_NGROUP + 1 and lg_delta of + * lg_base - SC_LG_NGROUP) + * (1 << lg_base) + 1 * (1 << lg_delta) + * (1 << lg_base) + 2 * (1 << lg_delta) + * (1 << lg_base) + 3 * (1 << lg_delta) + * ... + * (1 << lg_base) + SC_NGROUP * (1 << lg_delta) [ == (1 << (lg_base + 1)) ] + * + * ... + * + * Regular group N: + * - Count: SC_NGROUP + * - Sizes: + * (relative to lg_base of LG_QUANTUM + SC_LG_NGROUP + N and lg_delta of + * lg_base - SC_LG_NGROUP) + * (1 << lg_base) + 1 * (1 << lg_delta) + * (1 << lg_base) + 2 * (1 << lg_delta) + * (1 << lg_base) + 3 * (1 << lg_delta) + * ... + * (1 << lg_base) + SC_NGROUP * (1 << lg_delta) [ == (1 << (lg_base + 1)) ] + * + * + * Representation of metadata: + * To make the math easy, we'll mostly work in lg quantities. We record lg_base, + * lg_delta, and ndelta (i.e. number of deltas above the base) on a + * per-size-class basis, and maintain the invariant that, across all size + * classes, size == (1 << lg_base) + ndelta * (1 << lg_delta). + * + * For regular groups (i.e. those with lg_base >= LG_QUANTUM + SC_LG_NGROUP), + * lg_delta is lg_base - SC_LG_NGROUP, and ndelta goes from 1 to SC_NGROUP. + * + * For the initial tiny size classes (if any), lg_base is lg(size class size). + * lg_delta is lg_base for the first size class, and lg_base - 1 for all + * subsequent ones. ndelta is always 0. + * + * For the pseudo-group, if there are no tiny size classes, then we set + * lg_base == LG_QUANTUM, lg_delta == LG_QUANTUM, and have ndelta range from 0 + * to SC_NGROUP - 1. (Note that delta == base, so base + (SC_NGROUP - 1) * delta + * is just SC_NGROUP * base, or (1 << (SC_LG_NGROUP + LG_QUANTUM)), so we do + * indeed get a power of two that way). If there *are* tiny size classes, then + * the first size class needs to have lg_delta relative to the largest tiny size + * class. We therefore set lg_base == LG_QUANTUM - 1, + * lg_delta == LG_QUANTUM - 1, and ndelta == 1, keeping the rest of the + * pseudo-group the same. + * + * + * Other terminology: + * "Small" size classes mean those that are allocated out of bins, which is the + * same as those that are slab allocated. + * "Large" size classes are those that are not small. The cutoff for counting as + * large is page size * group size. + */ + +/* + * Size class N + (1 << SC_LG_NGROUP) twice the size of size class N. + */ +#define SC_LG_NGROUP 2 +#define SC_LG_TINY_MIN 3 + +#if SC_LG_TINY_MIN == 0 +/* The div module doesn't support division by 1, which this would require. */ +#error "Unsupported LG_TINY_MIN" +#endif + +/* + * The definitions below are all determined by the above settings and system + * characteristics. + */ +#define SC_NGROUP (1ULL << SC_LG_NGROUP) +#define SC_PTR_BITS ((1ULL << LG_SIZEOF_PTR) * 8) +#define SC_NTINY (LG_QUANTUM - SC_LG_TINY_MIN) +#define SC_LG_TINY_MAXCLASS (LG_QUANTUM > SC_LG_TINY_MIN ? LG_QUANTUM - 1 : -1) +#define SC_NPSEUDO SC_NGROUP +#define SC_LG_FIRST_REGULAR_BASE (LG_QUANTUM + SC_LG_NGROUP) +/* + * We cap allocations to be less than 2 ** (ptr_bits - 1), so the highest base + * we need is 2 ** (ptr_bits - 2). (This also means that the last group is 1 + * size class shorter than the others). + * We could probably save some space in arenas by capping this at LG_VADDR size. + */ +#define SC_LG_BASE_MAX (SC_PTR_BITS - 2) +#define SC_NREGULAR (SC_NGROUP * \ + (SC_LG_BASE_MAX - SC_LG_FIRST_REGULAR_BASE + 1) - 1) +#define SC_NSIZES (SC_NTINY + SC_NPSEUDO + SC_NREGULAR) + +/* The number of size classes that are a multiple of the page size. */ +#define SC_NPSIZES ( \ + /* Start with all the size classes. */ \ + SC_NSIZES \ + /* Subtract out those groups with too small a base. */ \ + - (LG_PAGE - 1 - SC_LG_FIRST_REGULAR_BASE) * SC_NGROUP \ + /* And the pseudo-group. */ \ + - SC_NPSEUDO \ + /* And the tiny group. */ \ + - SC_NTINY \ + /* Sizes where ndelta*delta is not a multiple of the page size. */ \ + - (SC_LG_NGROUP * SC_NGROUP)) +/* + * Note that the last line is computed as the sum of the second column in the + * following table: + * lg(base) | count of sizes to exclude + * ------------------------------|----------------------------- + * LG_PAGE - 1 | SC_NGROUP - 1 + * LG_PAGE | SC_NGROUP - 1 + * LG_PAGE + 1 | SC_NGROUP - 2 + * LG_PAGE + 2 | SC_NGROUP - 4 + * ... | ... + * LG_PAGE + (SC_LG_NGROUP - 1) | SC_NGROUP - (SC_NGROUP / 2) + */ + +/* + * We declare a size class is binnable if size < page size * group. Or, in other + * words, lg(size) < lg(page size) + lg(group size). + */ +#define SC_NBINS ( \ + /* Sub-regular size classes. */ \ + SC_NTINY + SC_NPSEUDO \ + /* Groups with lg_regular_min_base <= lg_base <= lg_base_max */ \ + + SC_NGROUP * (LG_PAGE + SC_LG_NGROUP - SC_LG_FIRST_REGULAR_BASE) \ + /* Last SC of the last group hits the bound exactly; exclude it. */ \ + - 1) + +/* + * The size2index_tab lookup table uses uint8_t to encode each bin index, so we + * cannot support more than 256 small size classes. + */ +#if (SC_NBINS > 256) +# error "Too many small size classes" +#endif + +/* The largest size class in the lookup table. */ +#define SC_LOOKUP_MAXCLASS ((size_t)1 << 12) + +/* Internal, only used for the definition of SC_SMALL_MAXCLASS. */ +#define SC_SMALL_MAX_BASE ((size_t)1 << (LG_PAGE + SC_LG_NGROUP - 1)) +#define SC_SMALL_MAX_DELTA ((size_t)1 << (LG_PAGE - 1)) + +/* The largest size class allocated out of a slab. */ +#define SC_SMALL_MAXCLASS (SC_SMALL_MAX_BASE \ + + (SC_NGROUP - 1) * SC_SMALL_MAX_DELTA) + +/* The smallest size class not allocated out of a slab. */ +#define SC_LARGE_MINCLASS ((size_t)1ULL << (LG_PAGE + SC_LG_NGROUP)) +#define SC_LG_LARGE_MINCLASS (LG_PAGE + SC_LG_NGROUP) + +/* Internal; only used for the definition of SC_LARGE_MAXCLASS. */ +#define SC_MAX_BASE ((size_t)1 << (SC_PTR_BITS - 2)) +#define SC_MAX_DELTA ((size_t)1 << (SC_PTR_BITS - 2 - SC_LG_NGROUP)) + +/* The largest size class supported. */ +#define SC_LARGE_MAXCLASS (SC_MAX_BASE + (SC_NGROUP - 1) * SC_MAX_DELTA) + +typedef struct sc_s sc_t; +struct sc_s { + /* Size class index, or -1 if not a valid size class. */ + int index; + /* Lg group base size (no deltas added). */ + int lg_base; + /* Lg delta to previous size class. */ + int lg_delta; + /* Delta multiplier. size == 1<<lg_base + ndelta<<lg_delta */ + int ndelta; + /* + * True if the size class is a multiple of the page size, false + * otherwise. + */ + bool psz; + /* + * True if the size class is a small, bin, size class. False otherwise. + */ + bool bin; + /* The slab page count if a small bin size class, 0 otherwise. */ + int pgs; + /* Same as lg_delta if a lookup table size class, 0 otherwise. */ + int lg_delta_lookup; +}; + +typedef struct sc_data_s sc_data_t; +struct sc_data_s { + /* Number of tiny size classes. */ + unsigned ntiny; + /* Number of bins supported by the lookup table. */ + int nlbins; + /* Number of small size class bins. */ + int nbins; + /* Number of size classes. */ + int nsizes; + /* Number of bits required to store NSIZES. */ + int lg_ceil_nsizes; + /* Number of size classes that are a multiple of (1U << LG_PAGE). */ + unsigned npsizes; + /* Lg of maximum tiny size class (or -1, if none). */ + int lg_tiny_maxclass; + /* Maximum size class included in lookup table. */ + size_t lookup_maxclass; + /* Maximum small size class. */ + size_t small_maxclass; + /* Lg of minimum large size class. */ + int lg_large_minclass; + /* The minimum large size class. */ + size_t large_minclass; + /* Maximum (large) size class. */ + size_t large_maxclass; + /* True if the sc_data_t has been initialized (for debugging only). */ + bool initialized; + + sc_t sc[SC_NSIZES]; +}; + +void sc_data_init(sc_data_t *data); +/* + * Updates slab sizes in [begin, end] to be pgs pages in length, if possible. + * Otherwise, does its best to accomodate the request. + */ +void sc_data_update_slab_size(sc_data_t *data, size_t begin, size_t end, + int pgs); +void sc_boot(sc_data_t *data); + +#endif /* JEMALLOC_INTERNAL_SC_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/seq.h b/deps/jemalloc/include/jemalloc/internal/seq.h new file mode 100644 index 000000000..ef2df4c6e --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/seq.h @@ -0,0 +1,55 @@ +#ifndef JEMALLOC_INTERNAL_SEQ_H +#define JEMALLOC_INTERNAL_SEQ_H + +#include "jemalloc/internal/atomic.h" + +/* + * A simple seqlock implementation. + */ + +#define seq_define(type, short_type) \ +typedef struct { \ + atomic_zu_t seq; \ + atomic_zu_t data[ \ + (sizeof(type) + sizeof(size_t) - 1) / sizeof(size_t)]; \ +} seq_##short_type##_t; \ + \ +/* \ + * No internal synchronization -- the caller must ensure that there's \ + * only a single writer at a time. \ + */ \ +static inline void \ +seq_store_##short_type(seq_##short_type##_t *dst, type *src) { \ + size_t buf[sizeof(dst->data) / sizeof(size_t)]; \ + buf[sizeof(buf) / sizeof(size_t) - 1] = 0; \ + memcpy(buf, src, sizeof(type)); \ + size_t old_seq = atomic_load_zu(&dst->seq, ATOMIC_RELAXED); \ + atomic_store_zu(&dst->seq, old_seq + 1, ATOMIC_RELAXED); \ + atomic_fence(ATOMIC_RELEASE); \ + for (size_t i = 0; i < sizeof(buf) / sizeof(size_t); i++) { \ + atomic_store_zu(&dst->data[i], buf[i], ATOMIC_RELAXED); \ + } \ + atomic_store_zu(&dst->seq, old_seq + 2, ATOMIC_RELEASE); \ +} \ + \ +/* Returns whether or not the read was consistent. */ \ +static inline bool \ +seq_try_load_##short_type(type *dst, seq_##short_type##_t *src) { \ + size_t buf[sizeof(src->data) / sizeof(size_t)]; \ + size_t seq1 = atomic_load_zu(&src->seq, ATOMIC_ACQUIRE); \ + if (seq1 % 2 != 0) { \ + return false; \ + } \ + for (size_t i = 0; i < sizeof(buf) / sizeof(size_t); i++) { \ + buf[i] = atomic_load_zu(&src->data[i], ATOMIC_RELAXED); \ + } \ + atomic_fence(ATOMIC_ACQUIRE); \ + size_t seq2 = atomic_load_zu(&src->seq, ATOMIC_RELAXED); \ + if (seq1 != seq2) { \ + return false; \ + } \ + memcpy(dst, buf, sizeof(type)); \ + return true; \ +} + +#endif /* JEMALLOC_INTERNAL_SEQ_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/smoothstep.h b/deps/jemalloc/include/jemalloc/internal/smoothstep.h new file mode 100644 index 000000000..2e14430f5 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/smoothstep.h @@ -0,0 +1,232 @@ +#ifndef JEMALLOC_INTERNAL_SMOOTHSTEP_H +#define JEMALLOC_INTERNAL_SMOOTHSTEP_H + +/* + * This file was generated by the following command: + * sh smoothstep.sh smoother 200 24 3 15 + */ +/******************************************************************************/ + +/* + * This header defines a precomputed table based on the smoothstep family of + * sigmoidal curves (https://en.wikipedia.org/wiki/Smoothstep) that grow from 0 + * to 1 in 0 <= x <= 1. The table is stored as integer fixed point values so + * that floating point math can be avoided. + * + * 3 2 + * smoothstep(x) = -2x + 3x + * + * 5 4 3 + * smootherstep(x) = 6x - 15x + 10x + * + * 7 6 5 4 + * smootheststep(x) = -20x + 70x - 84x + 35x + */ + +#define SMOOTHSTEP_VARIANT "smoother" +#define SMOOTHSTEP_NSTEPS 200 +#define SMOOTHSTEP_BFP 24 +#define SMOOTHSTEP \ + /* STEP(step, h, x, y) */ \ + STEP( 1, UINT64_C(0x0000000000000014), 0.005, 0.000001240643750) \ + STEP( 2, UINT64_C(0x00000000000000a5), 0.010, 0.000009850600000) \ + STEP( 3, UINT64_C(0x0000000000000229), 0.015, 0.000032995181250) \ + STEP( 4, UINT64_C(0x0000000000000516), 0.020, 0.000077619200000) \ + STEP( 5, UINT64_C(0x00000000000009dc), 0.025, 0.000150449218750) \ + STEP( 6, UINT64_C(0x00000000000010e8), 0.030, 0.000257995800000) \ + STEP( 7, UINT64_C(0x0000000000001aa4), 0.035, 0.000406555756250) \ + STEP( 8, UINT64_C(0x0000000000002777), 0.040, 0.000602214400000) \ + STEP( 9, UINT64_C(0x00000000000037c2), 0.045, 0.000850847793750) \ + STEP( 10, UINT64_C(0x0000000000004be6), 0.050, 0.001158125000000) \ + STEP( 11, UINT64_C(0x000000000000643c), 0.055, 0.001529510331250) \ + STEP( 12, UINT64_C(0x000000000000811f), 0.060, 0.001970265600000) \ + STEP( 13, UINT64_C(0x000000000000a2e2), 0.065, 0.002485452368750) \ + STEP( 14, UINT64_C(0x000000000000c9d8), 0.070, 0.003079934200000) \ + STEP( 15, UINT64_C(0x000000000000f64f), 0.075, 0.003758378906250) \ + STEP( 16, UINT64_C(0x0000000000012891), 0.080, 0.004525260800000) \ + STEP( 17, UINT64_C(0x00000000000160e7), 0.085, 0.005384862943750) \ + STEP( 18, UINT64_C(0x0000000000019f95), 0.090, 0.006341279400000) \ + STEP( 19, UINT64_C(0x000000000001e4dc), 0.095, 0.007398417481250) \ + STEP( 20, UINT64_C(0x00000000000230fc), 0.100, 0.008560000000000) \ + STEP( 21, UINT64_C(0x0000000000028430), 0.105, 0.009829567518750) \ + STEP( 22, UINT64_C(0x000000000002deb0), 0.110, 0.011210480600000) \ + STEP( 23, UINT64_C(0x00000000000340b1), 0.115, 0.012705922056250) \ + STEP( 24, UINT64_C(0x000000000003aa67), 0.120, 0.014318899200000) \ + STEP( 25, UINT64_C(0x0000000000041c00), 0.125, 0.016052246093750) \ + STEP( 26, UINT64_C(0x00000000000495a8), 0.130, 0.017908625800000) \ + STEP( 27, UINT64_C(0x000000000005178b), 0.135, 0.019890532631250) \ + STEP( 28, UINT64_C(0x000000000005a1cf), 0.140, 0.022000294400000) \ + STEP( 29, UINT64_C(0x0000000000063498), 0.145, 0.024240074668750) \ + STEP( 30, UINT64_C(0x000000000006d009), 0.150, 0.026611875000000) \ + STEP( 31, UINT64_C(0x000000000007743f), 0.155, 0.029117537206250) \ + STEP( 32, UINT64_C(0x0000000000082157), 0.160, 0.031758745600000) \ + STEP( 33, UINT64_C(0x000000000008d76b), 0.165, 0.034537029243750) \ + STEP( 34, UINT64_C(0x0000000000099691), 0.170, 0.037453764200000) \ + STEP( 35, UINT64_C(0x00000000000a5edf), 0.175, 0.040510175781250) \ + STEP( 36, UINT64_C(0x00000000000b3067), 0.180, 0.043707340800000) \ + STEP( 37, UINT64_C(0x00000000000c0b38), 0.185, 0.047046189818750) \ + STEP( 38, UINT64_C(0x00000000000cef5e), 0.190, 0.050527509400000) \ + STEP( 39, UINT64_C(0x00000000000ddce6), 0.195, 0.054151944356250) \ + STEP( 40, UINT64_C(0x00000000000ed3d8), 0.200, 0.057920000000000) \ + STEP( 41, UINT64_C(0x00000000000fd439), 0.205, 0.061832044393750) \ + STEP( 42, UINT64_C(0x000000000010de0e), 0.210, 0.065888310600000) \ + STEP( 43, UINT64_C(0x000000000011f158), 0.215, 0.070088898931250) \ + STEP( 44, UINT64_C(0x0000000000130e17), 0.220, 0.074433779200000) \ + STEP( 45, UINT64_C(0x0000000000143448), 0.225, 0.078922792968750) \ + STEP( 46, UINT64_C(0x00000000001563e7), 0.230, 0.083555655800000) \ + STEP( 47, UINT64_C(0x0000000000169cec), 0.235, 0.088331959506250) \ + STEP( 48, UINT64_C(0x000000000017df4f), 0.240, 0.093251174400000) \ + STEP( 49, UINT64_C(0x0000000000192b04), 0.245, 0.098312651543750) \ + STEP( 50, UINT64_C(0x00000000001a8000), 0.250, 0.103515625000000) \ + STEP( 51, UINT64_C(0x00000000001bde32), 0.255, 0.108859214081250) \ + STEP( 52, UINT64_C(0x00000000001d458b), 0.260, 0.114342425600000) \ + STEP( 53, UINT64_C(0x00000000001eb5f8), 0.265, 0.119964156118750) \ + STEP( 54, UINT64_C(0x0000000000202f65), 0.270, 0.125723194200000) \ + STEP( 55, UINT64_C(0x000000000021b1bb), 0.275, 0.131618222656250) \ + STEP( 56, UINT64_C(0x0000000000233ce3), 0.280, 0.137647820800000) \ + STEP( 57, UINT64_C(0x000000000024d0c3), 0.285, 0.143810466693750) \ + STEP( 58, UINT64_C(0x0000000000266d40), 0.290, 0.150104539400000) \ + STEP( 59, UINT64_C(0x000000000028123d), 0.295, 0.156528321231250) \ + STEP( 60, UINT64_C(0x000000000029bf9c), 0.300, 0.163080000000000) \ + STEP( 61, UINT64_C(0x00000000002b753d), 0.305, 0.169757671268750) \ + STEP( 62, UINT64_C(0x00000000002d32fe), 0.310, 0.176559340600000) \ + STEP( 63, UINT64_C(0x00000000002ef8bc), 0.315, 0.183482925806250) \ + STEP( 64, UINT64_C(0x000000000030c654), 0.320, 0.190526259200000) \ + STEP( 65, UINT64_C(0x0000000000329b9f), 0.325, 0.197687089843750) \ + STEP( 66, UINT64_C(0x0000000000347875), 0.330, 0.204963085800000) \ + STEP( 67, UINT64_C(0x0000000000365cb0), 0.335, 0.212351836381250) \ + STEP( 68, UINT64_C(0x0000000000384825), 0.340, 0.219850854400000) \ + STEP( 69, UINT64_C(0x00000000003a3aa8), 0.345, 0.227457578418750) \ + STEP( 70, UINT64_C(0x00000000003c340f), 0.350, 0.235169375000000) \ + STEP( 71, UINT64_C(0x00000000003e342b), 0.355, 0.242983540956250) \ + STEP( 72, UINT64_C(0x0000000000403ace), 0.360, 0.250897305600000) \ + STEP( 73, UINT64_C(0x00000000004247c8), 0.365, 0.258907832993750) \ + STEP( 74, UINT64_C(0x0000000000445ae9), 0.370, 0.267012224200000) \ + STEP( 75, UINT64_C(0x0000000000467400), 0.375, 0.275207519531250) \ + STEP( 76, UINT64_C(0x00000000004892d8), 0.380, 0.283490700800000) \ + STEP( 77, UINT64_C(0x00000000004ab740), 0.385, 0.291858693568750) \ + STEP( 78, UINT64_C(0x00000000004ce102), 0.390, 0.300308369400000) \ + STEP( 79, UINT64_C(0x00000000004f0fe9), 0.395, 0.308836548106250) \ + STEP( 80, UINT64_C(0x00000000005143bf), 0.400, 0.317440000000000) \ + STEP( 81, UINT64_C(0x0000000000537c4d), 0.405, 0.326115448143750) \ + STEP( 82, UINT64_C(0x000000000055b95b), 0.410, 0.334859570600000) \ + STEP( 83, UINT64_C(0x000000000057fab1), 0.415, 0.343669002681250) \ + STEP( 84, UINT64_C(0x00000000005a4015), 0.420, 0.352540339200000) \ + STEP( 85, UINT64_C(0x00000000005c894e), 0.425, 0.361470136718750) \ + STEP( 86, UINT64_C(0x00000000005ed622), 0.430, 0.370454915800000) \ + STEP( 87, UINT64_C(0x0000000000612655), 0.435, 0.379491163256250) \ + STEP( 88, UINT64_C(0x00000000006379ac), 0.440, 0.388575334400000) \ + STEP( 89, UINT64_C(0x000000000065cfeb), 0.445, 0.397703855293750) \ + STEP( 90, UINT64_C(0x00000000006828d6), 0.450, 0.406873125000000) \ + STEP( 91, UINT64_C(0x00000000006a842f), 0.455, 0.416079517831250) \ + STEP( 92, UINT64_C(0x00000000006ce1bb), 0.460, 0.425319385600000) \ + STEP( 93, UINT64_C(0x00000000006f413a), 0.465, 0.434589059868750) \ + STEP( 94, UINT64_C(0x000000000071a270), 0.470, 0.443884854200000) \ + STEP( 95, UINT64_C(0x000000000074051d), 0.475, 0.453203066406250) \ + STEP( 96, UINT64_C(0x0000000000766905), 0.480, 0.462539980800000) \ + STEP( 97, UINT64_C(0x000000000078cde7), 0.485, 0.471891870443750) \ + STEP( 98, UINT64_C(0x00000000007b3387), 0.490, 0.481254999400000) \ + STEP( 99, UINT64_C(0x00000000007d99a4), 0.495, 0.490625624981250) \ + STEP( 100, UINT64_C(0x0000000000800000), 0.500, 0.500000000000000) \ + STEP( 101, UINT64_C(0x000000000082665b), 0.505, 0.509374375018750) \ + STEP( 102, UINT64_C(0x000000000084cc78), 0.510, 0.518745000600000) \ + STEP( 103, UINT64_C(0x0000000000873218), 0.515, 0.528108129556250) \ + STEP( 104, UINT64_C(0x00000000008996fa), 0.520, 0.537460019200000) \ + STEP( 105, UINT64_C(0x00000000008bfae2), 0.525, 0.546796933593750) \ + STEP( 106, UINT64_C(0x00000000008e5d8f), 0.530, 0.556115145800000) \ + STEP( 107, UINT64_C(0x000000000090bec5), 0.535, 0.565410940131250) \ + STEP( 108, UINT64_C(0x0000000000931e44), 0.540, 0.574680614400000) \ + STEP( 109, UINT64_C(0x0000000000957bd0), 0.545, 0.583920482168750) \ + STEP( 110, UINT64_C(0x000000000097d729), 0.550, 0.593126875000000) \ + STEP( 111, UINT64_C(0x00000000009a3014), 0.555, 0.602296144706250) \ + STEP( 112, UINT64_C(0x00000000009c8653), 0.560, 0.611424665600000) \ + STEP( 113, UINT64_C(0x00000000009ed9aa), 0.565, 0.620508836743750) \ + STEP( 114, UINT64_C(0x0000000000a129dd), 0.570, 0.629545084200000) \ + STEP( 115, UINT64_C(0x0000000000a376b1), 0.575, 0.638529863281250) \ + STEP( 116, UINT64_C(0x0000000000a5bfea), 0.580, 0.647459660800000) \ + STEP( 117, UINT64_C(0x0000000000a8054e), 0.585, 0.656330997318750) \ + STEP( 118, UINT64_C(0x0000000000aa46a4), 0.590, 0.665140429400000) \ + STEP( 119, UINT64_C(0x0000000000ac83b2), 0.595, 0.673884551856250) \ + STEP( 120, UINT64_C(0x0000000000aebc40), 0.600, 0.682560000000000) \ + STEP( 121, UINT64_C(0x0000000000b0f016), 0.605, 0.691163451893750) \ + STEP( 122, UINT64_C(0x0000000000b31efd), 0.610, 0.699691630600000) \ + STEP( 123, UINT64_C(0x0000000000b548bf), 0.615, 0.708141306431250) \ + STEP( 124, UINT64_C(0x0000000000b76d27), 0.620, 0.716509299200000) \ + STEP( 125, UINT64_C(0x0000000000b98c00), 0.625, 0.724792480468750) \ + STEP( 126, UINT64_C(0x0000000000bba516), 0.630, 0.732987775800000) \ + STEP( 127, UINT64_C(0x0000000000bdb837), 0.635, 0.741092167006250) \ + STEP( 128, UINT64_C(0x0000000000bfc531), 0.640, 0.749102694400000) \ + STEP( 129, UINT64_C(0x0000000000c1cbd4), 0.645, 0.757016459043750) \ + STEP( 130, UINT64_C(0x0000000000c3cbf0), 0.650, 0.764830625000000) \ + STEP( 131, UINT64_C(0x0000000000c5c557), 0.655, 0.772542421581250) \ + STEP( 132, UINT64_C(0x0000000000c7b7da), 0.660, 0.780149145600000) \ + STEP( 133, UINT64_C(0x0000000000c9a34f), 0.665, 0.787648163618750) \ + STEP( 134, UINT64_C(0x0000000000cb878a), 0.670, 0.795036914200000) \ + STEP( 135, UINT64_C(0x0000000000cd6460), 0.675, 0.802312910156250) \ + STEP( 136, UINT64_C(0x0000000000cf39ab), 0.680, 0.809473740800000) \ + STEP( 137, UINT64_C(0x0000000000d10743), 0.685, 0.816517074193750) \ + STEP( 138, UINT64_C(0x0000000000d2cd01), 0.690, 0.823440659400000) \ + STEP( 139, UINT64_C(0x0000000000d48ac2), 0.695, 0.830242328731250) \ + STEP( 140, UINT64_C(0x0000000000d64063), 0.700, 0.836920000000000) \ + STEP( 141, UINT64_C(0x0000000000d7edc2), 0.705, 0.843471678768750) \ + STEP( 142, UINT64_C(0x0000000000d992bf), 0.710, 0.849895460600000) \ + STEP( 143, UINT64_C(0x0000000000db2f3c), 0.715, 0.856189533306250) \ + STEP( 144, UINT64_C(0x0000000000dcc31c), 0.720, 0.862352179200000) \ + STEP( 145, UINT64_C(0x0000000000de4e44), 0.725, 0.868381777343750) \ + STEP( 146, UINT64_C(0x0000000000dfd09a), 0.730, 0.874276805800000) \ + STEP( 147, UINT64_C(0x0000000000e14a07), 0.735, 0.880035843881250) \ + STEP( 148, UINT64_C(0x0000000000e2ba74), 0.740, 0.885657574400000) \ + STEP( 149, UINT64_C(0x0000000000e421cd), 0.745, 0.891140785918750) \ + STEP( 150, UINT64_C(0x0000000000e58000), 0.750, 0.896484375000000) \ + STEP( 151, UINT64_C(0x0000000000e6d4fb), 0.755, 0.901687348456250) \ + STEP( 152, UINT64_C(0x0000000000e820b0), 0.760, 0.906748825600000) \ + STEP( 153, UINT64_C(0x0000000000e96313), 0.765, 0.911668040493750) \ + STEP( 154, UINT64_C(0x0000000000ea9c18), 0.770, 0.916444344200000) \ + STEP( 155, UINT64_C(0x0000000000ebcbb7), 0.775, 0.921077207031250) \ + STEP( 156, UINT64_C(0x0000000000ecf1e8), 0.780, 0.925566220800000) \ + STEP( 157, UINT64_C(0x0000000000ee0ea7), 0.785, 0.929911101068750) \ + STEP( 158, UINT64_C(0x0000000000ef21f1), 0.790, 0.934111689400000) \ + STEP( 159, UINT64_C(0x0000000000f02bc6), 0.795, 0.938167955606250) \ + STEP( 160, UINT64_C(0x0000000000f12c27), 0.800, 0.942080000000000) \ + STEP( 161, UINT64_C(0x0000000000f22319), 0.805, 0.945848055643750) \ + STEP( 162, UINT64_C(0x0000000000f310a1), 0.810, 0.949472490600000) \ + STEP( 163, UINT64_C(0x0000000000f3f4c7), 0.815, 0.952953810181250) \ + STEP( 164, UINT64_C(0x0000000000f4cf98), 0.820, 0.956292659200000) \ + STEP( 165, UINT64_C(0x0000000000f5a120), 0.825, 0.959489824218750) \ + STEP( 166, UINT64_C(0x0000000000f6696e), 0.830, 0.962546235800000) \ + STEP( 167, UINT64_C(0x0000000000f72894), 0.835, 0.965462970756250) \ + STEP( 168, UINT64_C(0x0000000000f7dea8), 0.840, 0.968241254400000) \ + STEP( 169, UINT64_C(0x0000000000f88bc0), 0.845, 0.970882462793750) \ + STEP( 170, UINT64_C(0x0000000000f92ff6), 0.850, 0.973388125000000) \ + STEP( 171, UINT64_C(0x0000000000f9cb67), 0.855, 0.975759925331250) \ + STEP( 172, UINT64_C(0x0000000000fa5e30), 0.860, 0.977999705600000) \ + STEP( 173, UINT64_C(0x0000000000fae874), 0.865, 0.980109467368750) \ + STEP( 174, UINT64_C(0x0000000000fb6a57), 0.870, 0.982091374200000) \ + STEP( 175, UINT64_C(0x0000000000fbe400), 0.875, 0.983947753906250) \ + STEP( 176, UINT64_C(0x0000000000fc5598), 0.880, 0.985681100800000) \ + STEP( 177, UINT64_C(0x0000000000fcbf4e), 0.885, 0.987294077943750) \ + STEP( 178, UINT64_C(0x0000000000fd214f), 0.890, 0.988789519400000) \ + STEP( 179, UINT64_C(0x0000000000fd7bcf), 0.895, 0.990170432481250) \ + STEP( 180, UINT64_C(0x0000000000fdcf03), 0.900, 0.991440000000000) \ + STEP( 181, UINT64_C(0x0000000000fe1b23), 0.905, 0.992601582518750) \ + STEP( 182, UINT64_C(0x0000000000fe606a), 0.910, 0.993658720600000) \ + STEP( 183, UINT64_C(0x0000000000fe9f18), 0.915, 0.994615137056250) \ + STEP( 184, UINT64_C(0x0000000000fed76e), 0.920, 0.995474739200000) \ + STEP( 185, UINT64_C(0x0000000000ff09b0), 0.925, 0.996241621093750) \ + STEP( 186, UINT64_C(0x0000000000ff3627), 0.930, 0.996920065800000) \ + STEP( 187, UINT64_C(0x0000000000ff5d1d), 0.935, 0.997514547631250) \ + STEP( 188, UINT64_C(0x0000000000ff7ee0), 0.940, 0.998029734400000) \ + STEP( 189, UINT64_C(0x0000000000ff9bc3), 0.945, 0.998470489668750) \ + STEP( 190, UINT64_C(0x0000000000ffb419), 0.950, 0.998841875000000) \ + STEP( 191, UINT64_C(0x0000000000ffc83d), 0.955, 0.999149152206250) \ + STEP( 192, UINT64_C(0x0000000000ffd888), 0.960, 0.999397785600000) \ + STEP( 193, UINT64_C(0x0000000000ffe55b), 0.965, 0.999593444243750) \ + STEP( 194, UINT64_C(0x0000000000ffef17), 0.970, 0.999742004200000) \ + STEP( 195, UINT64_C(0x0000000000fff623), 0.975, 0.999849550781250) \ + STEP( 196, UINT64_C(0x0000000000fffae9), 0.980, 0.999922380800000) \ + STEP( 197, UINT64_C(0x0000000000fffdd6), 0.985, 0.999967004818750) \ + STEP( 198, UINT64_C(0x0000000000ffff5a), 0.990, 0.999990149400000) \ + STEP( 199, UINT64_C(0x0000000000ffffeb), 0.995, 0.999998759356250) \ + STEP( 200, UINT64_C(0x0000000001000000), 1.000, 1.000000000000000) \ + +#endif /* JEMALLOC_INTERNAL_SMOOTHSTEP_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/smoothstep.sh b/deps/jemalloc/include/jemalloc/internal/smoothstep.sh new file mode 100755 index 000000000..65de97bf4 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/smoothstep.sh @@ -0,0 +1,101 @@ +#!/bin/sh +# +# Generate a discrete lookup table for a sigmoid function in the smoothstep +# family (https://en.wikipedia.org/wiki/Smoothstep), where the lookup table +# entries correspond to x in [1/nsteps, 2/nsteps, ..., nsteps/nsteps]. Encode +# the entries using a binary fixed point representation. +# +# Usage: smoothstep.sh <variant> <nsteps> <bfp> <xprec> <yprec> +# +# <variant> is in {smooth, smoother, smoothest}. +# <nsteps> must be greater than zero. +# <bfp> must be in [0..62]; reasonable values are roughly [10..30]. +# <xprec> is x decimal precision. +# <yprec> is y decimal precision. + +#set -x + +cmd="sh smoothstep.sh $*" +variant=$1 +nsteps=$2 +bfp=$3 +xprec=$4 +yprec=$5 + +case "${variant}" in + smooth) + ;; + smoother) + ;; + smoothest) + ;; + *) + echo "Unsupported variant" + exit 1 + ;; +esac + +smooth() { + step=$1 + y=`echo ${yprec} k ${step} ${nsteps} / sx _2 lx 3 ^ '*' 3 lx 2 ^ '*' + p | dc | tr -d '\\\\\n' | sed -e 's#^\.#0.#g'` + h=`echo ${yprec} k 2 ${bfp} ^ ${y} '*' p | dc | tr -d '\\\\\n' | sed -e 's#^\.#0.#g' | tr '.' ' ' | awk '{print $1}' ` +} + +smoother() { + step=$1 + y=`echo ${yprec} k ${step} ${nsteps} / sx 6 lx 5 ^ '*' _15 lx 4 ^ '*' + 10 lx 3 ^ '*' + p | dc | tr -d '\\\\\n' | sed -e 's#^\.#0.#g'` + h=`echo ${yprec} k 2 ${bfp} ^ ${y} '*' p | dc | tr -d '\\\\\n' | sed -e 's#^\.#0.#g' | tr '.' ' ' | awk '{print $1}' ` +} + +smoothest() { + step=$1 + y=`echo ${yprec} k ${step} ${nsteps} / sx _20 lx 7 ^ '*' 70 lx 6 ^ '*' + _84 lx 5 ^ '*' + 35 lx 4 ^ '*' + p | dc | tr -d '\\\\\n' | sed -e 's#^\.#0.#g'` + h=`echo ${yprec} k 2 ${bfp} ^ ${y} '*' p | dc | tr -d '\\\\\n' | sed -e 's#^\.#0.#g' | tr '.' ' ' | awk '{print $1}' ` +} + +cat <<EOF +#ifndef JEMALLOC_INTERNAL_SMOOTHSTEP_H +#define JEMALLOC_INTERNAL_SMOOTHSTEP_H + +/* + * This file was generated by the following command: + * $cmd + */ +/******************************************************************************/ + +/* + * This header defines a precomputed table based on the smoothstep family of + * sigmoidal curves (https://en.wikipedia.org/wiki/Smoothstep) that grow from 0 + * to 1 in 0 <= x <= 1. The table is stored as integer fixed point values so + * that floating point math can be avoided. + * + * 3 2 + * smoothstep(x) = -2x + 3x + * + * 5 4 3 + * smootherstep(x) = 6x - 15x + 10x + * + * 7 6 5 4 + * smootheststep(x) = -20x + 70x - 84x + 35x + */ + +#define SMOOTHSTEP_VARIANT "${variant}" +#define SMOOTHSTEP_NSTEPS ${nsteps} +#define SMOOTHSTEP_BFP ${bfp} +#define SMOOTHSTEP \\ + /* STEP(step, h, x, y) */ \\ +EOF + +s=1 +while [ $s -le $nsteps ] ; do + $variant ${s} + x=`echo ${xprec} k ${s} ${nsteps} / p | dc | tr -d '\\\\\n' | sed -e 's#^\.#0.#g'` + printf ' STEP(%4d, UINT64_C(0x%016x), %s, %s) \\\n' ${s} ${h} ${x} ${y} + + s=$((s+1)) +done +echo + +cat <<EOF +#endif /* JEMALLOC_INTERNAL_SMOOTHSTEP_H */ +EOF diff --git a/deps/jemalloc/include/jemalloc/internal/spin.h b/deps/jemalloc/include/jemalloc/internal/spin.h new file mode 100644 index 000000000..22804c687 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/spin.h @@ -0,0 +1,40 @@ +#ifndef JEMALLOC_INTERNAL_SPIN_H +#define JEMALLOC_INTERNAL_SPIN_H + +#define SPIN_INITIALIZER {0U} + +typedef struct { + unsigned iteration; +} spin_t; + +static inline void +spin_cpu_spinwait() { +# if HAVE_CPU_SPINWAIT + CPU_SPINWAIT; +# else + volatile int x = 0; + x = x; +# endif +} + +static inline void +spin_adaptive(spin_t *spin) { + volatile uint32_t i; + + if (spin->iteration < 5) { + for (i = 0; i < (1U << spin->iteration); i++) { + spin_cpu_spinwait(); + } + spin->iteration++; + } else { +#ifdef _WIN32 + SwitchToThread(); +#else + sched_yield(); +#endif + } +} + +#undef SPIN_INLINE + +#endif /* JEMALLOC_INTERNAL_SPIN_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/stats.h b/deps/jemalloc/include/jemalloc/internal/stats.h new file mode 100644 index 000000000..3b9e0eac1 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/stats.h @@ -0,0 +1,31 @@ +#ifndef JEMALLOC_INTERNAL_STATS_H +#define JEMALLOC_INTERNAL_STATS_H + +/* OPTION(opt, var_name, default, set_value_to) */ +#define STATS_PRINT_OPTIONS \ + OPTION('J', json, false, true) \ + OPTION('g', general, true, false) \ + OPTION('m', merged, config_stats, false) \ + OPTION('d', destroyed, config_stats, false) \ + OPTION('a', unmerged, config_stats, false) \ + OPTION('b', bins, true, false) \ + OPTION('l', large, true, false) \ + OPTION('x', mutex, true, false) \ + OPTION('e', extents, true, false) + +enum { +#define OPTION(o, v, d, s) stats_print_option_num_##v, + STATS_PRINT_OPTIONS +#undef OPTION + stats_print_tot_num_options +}; + +/* Options for stats_print. */ +extern bool opt_stats_print; +extern char opt_stats_print_opts[stats_print_tot_num_options+1]; + +/* Implements je_malloc_stats_print. */ +void stats_print(void (*write_cb)(void *, const char *), void *cbopaque, + const char *opts); + +#endif /* JEMALLOC_INTERNAL_STATS_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/sz.h b/deps/jemalloc/include/jemalloc/internal/sz.h new file mode 100644 index 000000000..68e558abf --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/sz.h @@ -0,0 +1,318 @@ +#ifndef JEMALLOC_INTERNAL_SIZE_H +#define JEMALLOC_INTERNAL_SIZE_H + +#include "jemalloc/internal/bit_util.h" +#include "jemalloc/internal/pages.h" +#include "jemalloc/internal/sc.h" +#include "jemalloc/internal/util.h" + +/* + * sz module: Size computations. + * + * Some abbreviations used here: + * p: Page + * ind: Index + * s, sz: Size + * u: Usable size + * a: Aligned + * + * These are not always used completely consistently, but should be enough to + * interpret function names. E.g. sz_psz2ind converts page size to page size + * index; sz_sa2u converts a (size, alignment) allocation request to the usable + * size that would result from such an allocation. + */ + +/* + * sz_pind2sz_tab encodes the same information as could be computed by + * sz_pind2sz_compute(). + */ +extern size_t sz_pind2sz_tab[SC_NPSIZES + 1]; +/* + * sz_index2size_tab encodes the same information as could be computed (at + * unacceptable cost in some code paths) by sz_index2size_compute(). + */ +extern size_t sz_index2size_tab[SC_NSIZES]; +/* + * sz_size2index_tab is a compact lookup table that rounds request sizes up to + * size classes. In order to reduce cache footprint, the table is compressed, + * and all accesses are via sz_size2index(). + */ +extern uint8_t sz_size2index_tab[]; + +static const size_t sz_large_pad = +#ifdef JEMALLOC_CACHE_OBLIVIOUS + PAGE +#else + 0 +#endif + ; + +extern void sz_boot(const sc_data_t *sc_data); + +JEMALLOC_ALWAYS_INLINE pszind_t +sz_psz2ind(size_t psz) { + if (unlikely(psz > SC_LARGE_MAXCLASS)) { + return SC_NPSIZES; + } + pszind_t x = lg_floor((psz<<1)-1); + pszind_t shift = (x < SC_LG_NGROUP + LG_PAGE) ? + 0 : x - (SC_LG_NGROUP + LG_PAGE); + pszind_t grp = shift << SC_LG_NGROUP; + + pszind_t lg_delta = (x < SC_LG_NGROUP + LG_PAGE + 1) ? + LG_PAGE : x - SC_LG_NGROUP - 1; + + size_t delta_inverse_mask = ZU(-1) << lg_delta; + pszind_t mod = ((((psz-1) & delta_inverse_mask) >> lg_delta)) & + ((ZU(1) << SC_LG_NGROUP) - 1); + + pszind_t ind = grp + mod; + return ind; +} + +static inline size_t +sz_pind2sz_compute(pszind_t pind) { + if (unlikely(pind == SC_NPSIZES)) { + return SC_LARGE_MAXCLASS + PAGE; + } + size_t grp = pind >> SC_LG_NGROUP; + size_t mod = pind & ((ZU(1) << SC_LG_NGROUP) - 1); + + size_t grp_size_mask = ~((!!grp)-1); + size_t grp_size = ((ZU(1) << (LG_PAGE + (SC_LG_NGROUP-1))) << grp) + & grp_size_mask; + + size_t shift = (grp == 0) ? 1 : grp; + size_t lg_delta = shift + (LG_PAGE-1); + size_t mod_size = (mod+1) << lg_delta; + + size_t sz = grp_size + mod_size; + return sz; +} + +static inline size_t +sz_pind2sz_lookup(pszind_t pind) { + size_t ret = (size_t)sz_pind2sz_tab[pind]; + assert(ret == sz_pind2sz_compute(pind)); + return ret; +} + +static inline size_t +sz_pind2sz(pszind_t pind) { + assert(pind < SC_NPSIZES + 1); + return sz_pind2sz_lookup(pind); +} + +static inline size_t +sz_psz2u(size_t psz) { + if (unlikely(psz > SC_LARGE_MAXCLASS)) { + return SC_LARGE_MAXCLASS + PAGE; + } + size_t x = lg_floor((psz<<1)-1); + size_t lg_delta = (x < SC_LG_NGROUP + LG_PAGE + 1) ? + LG_PAGE : x - SC_LG_NGROUP - 1; + size_t delta = ZU(1) << lg_delta; + size_t delta_mask = delta - 1; + size_t usize = (psz + delta_mask) & ~delta_mask; + return usize; +} + +static inline szind_t +sz_size2index_compute(size_t size) { + if (unlikely(size > SC_LARGE_MAXCLASS)) { + return SC_NSIZES; + } + + if (size == 0) { + return 0; + } +#if (SC_NTINY != 0) + if (size <= (ZU(1) << SC_LG_TINY_MAXCLASS)) { + szind_t lg_tmin = SC_LG_TINY_MAXCLASS - SC_NTINY + 1; + szind_t lg_ceil = lg_floor(pow2_ceil_zu(size)); + return (lg_ceil < lg_tmin ? 0 : lg_ceil - lg_tmin); + } +#endif + { + szind_t x = lg_floor((size<<1)-1); + szind_t shift = (x < SC_LG_NGROUP + LG_QUANTUM) ? 0 : + x - (SC_LG_NGROUP + LG_QUANTUM); + szind_t grp = shift << SC_LG_NGROUP; + + szind_t lg_delta = (x < SC_LG_NGROUP + LG_QUANTUM + 1) + ? LG_QUANTUM : x - SC_LG_NGROUP - 1; + + size_t delta_inverse_mask = ZU(-1) << lg_delta; + szind_t mod = ((((size-1) & delta_inverse_mask) >> lg_delta)) & + ((ZU(1) << SC_LG_NGROUP) - 1); + + szind_t index = SC_NTINY + grp + mod; + return index; + } +} + +JEMALLOC_ALWAYS_INLINE szind_t +sz_size2index_lookup(size_t size) { + assert(size <= SC_LOOKUP_MAXCLASS); + szind_t ret = (sz_size2index_tab[(size + (ZU(1) << SC_LG_TINY_MIN) - 1) + >> SC_LG_TINY_MIN]); + assert(ret == sz_size2index_compute(size)); + return ret; +} + +JEMALLOC_ALWAYS_INLINE szind_t +sz_size2index(size_t size) { + if (likely(size <= SC_LOOKUP_MAXCLASS)) { + return sz_size2index_lookup(size); + } + return sz_size2index_compute(size); +} + +static inline size_t +sz_index2size_compute(szind_t index) { +#if (SC_NTINY > 0) + if (index < SC_NTINY) { + return (ZU(1) << (SC_LG_TINY_MAXCLASS - SC_NTINY + 1 + index)); + } +#endif + { + size_t reduced_index = index - SC_NTINY; + size_t grp = reduced_index >> SC_LG_NGROUP; + size_t mod = reduced_index & ((ZU(1) << SC_LG_NGROUP) - + 1); + + size_t grp_size_mask = ~((!!grp)-1); + size_t grp_size = ((ZU(1) << (LG_QUANTUM + + (SC_LG_NGROUP-1))) << grp) & grp_size_mask; + + size_t shift = (grp == 0) ? 1 : grp; + size_t lg_delta = shift + (LG_QUANTUM-1); + size_t mod_size = (mod+1) << lg_delta; + + size_t usize = grp_size + mod_size; + return usize; + } +} + +JEMALLOC_ALWAYS_INLINE size_t +sz_index2size_lookup(szind_t index) { + size_t ret = (size_t)sz_index2size_tab[index]; + assert(ret == sz_index2size_compute(index)); + return ret; +} + +JEMALLOC_ALWAYS_INLINE size_t +sz_index2size(szind_t index) { + assert(index < SC_NSIZES); + return sz_index2size_lookup(index); +} + +JEMALLOC_ALWAYS_INLINE size_t +sz_s2u_compute(size_t size) { + if (unlikely(size > SC_LARGE_MAXCLASS)) { + return 0; + } + + if (size == 0) { + size++; + } +#if (SC_NTINY > 0) + if (size <= (ZU(1) << SC_LG_TINY_MAXCLASS)) { + size_t lg_tmin = SC_LG_TINY_MAXCLASS - SC_NTINY + 1; + size_t lg_ceil = lg_floor(pow2_ceil_zu(size)); + return (lg_ceil < lg_tmin ? (ZU(1) << lg_tmin) : + (ZU(1) << lg_ceil)); + } +#endif + { + size_t x = lg_floor((size<<1)-1); + size_t lg_delta = (x < SC_LG_NGROUP + LG_QUANTUM + 1) + ? LG_QUANTUM : x - SC_LG_NGROUP - 1; + size_t delta = ZU(1) << lg_delta; + size_t delta_mask = delta - 1; + size_t usize = (size + delta_mask) & ~delta_mask; + return usize; + } +} + +JEMALLOC_ALWAYS_INLINE size_t +sz_s2u_lookup(size_t size) { + size_t ret = sz_index2size_lookup(sz_size2index_lookup(size)); + + assert(ret == sz_s2u_compute(size)); + return ret; +} + +/* + * Compute usable size that would result from allocating an object with the + * specified size. + */ +JEMALLOC_ALWAYS_INLINE size_t +sz_s2u(size_t size) { + if (likely(size <= SC_LOOKUP_MAXCLASS)) { + return sz_s2u_lookup(size); + } + return sz_s2u_compute(size); +} + +/* + * Compute usable size that would result from allocating an object with the + * specified size and alignment. + */ +JEMALLOC_ALWAYS_INLINE size_t +sz_sa2u(size_t size, size_t alignment) { + size_t usize; + + assert(alignment != 0 && ((alignment - 1) & alignment) == 0); + + /* Try for a small size class. */ + if (size <= SC_SMALL_MAXCLASS && alignment < PAGE) { + /* + * Round size up to the nearest multiple of alignment. + * + * This done, we can take advantage of the fact that for each + * small size class, every object is aligned at the smallest + * power of two that is non-zero in the base two representation + * of the size. For example: + * + * Size | Base 2 | Minimum alignment + * -----+----------+------------------ + * 96 | 1100000 | 32 + * 144 | 10100000 | 32 + * 192 | 11000000 | 64 + */ + usize = sz_s2u(ALIGNMENT_CEILING(size, alignment)); + if (usize < SC_LARGE_MINCLASS) { + return usize; + } + } + + /* Large size class. Beware of overflow. */ + + if (unlikely(alignment > SC_LARGE_MAXCLASS)) { + return 0; + } + + /* Make sure result is a large size class. */ + if (size <= SC_LARGE_MINCLASS) { + usize = SC_LARGE_MINCLASS; + } else { + usize = sz_s2u(size); + if (usize < size) { + /* size_t overflow. */ + return 0; + } + } + + /* + * Calculate the multi-page mapping that large_palloc() would need in + * order to guarantee the alignment. + */ + if (usize + sz_large_pad + PAGE_CEILING(alignment) - PAGE < usize) { + /* size_t overflow. */ + return 0; + } + return usize; +} + +#endif /* JEMALLOC_INTERNAL_SIZE_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/tcache_externs.h b/deps/jemalloc/include/jemalloc/internal/tcache_externs.h new file mode 100644 index 000000000..d63eafde8 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/tcache_externs.h @@ -0,0 +1,53 @@ +#ifndef JEMALLOC_INTERNAL_TCACHE_EXTERNS_H +#define JEMALLOC_INTERNAL_TCACHE_EXTERNS_H + +extern bool opt_tcache; +extern ssize_t opt_lg_tcache_max; + +extern cache_bin_info_t *tcache_bin_info; + +/* + * Number of tcache bins. There are SC_NBINS small-object bins, plus 0 or more + * large-object bins. + */ +extern unsigned nhbins; + +/* Maximum cached size class. */ +extern size_t tcache_maxclass; + +/* + * Explicit tcaches, managed via the tcache.{create,flush,destroy} mallctls and + * usable via the MALLOCX_TCACHE() flag. The automatic per thread tcaches are + * completely disjoint from this data structure. tcaches starts off as a sparse + * array, so it has no physical memory footprint until individual pages are + * touched. This allows the entire array to be allocated the first time an + * explicit tcache is created without a disproportionate impact on memory usage. + */ +extern tcaches_t *tcaches; + +size_t tcache_salloc(tsdn_t *tsdn, const void *ptr); +void tcache_event_hard(tsd_t *tsd, tcache_t *tcache); +void *tcache_alloc_small_hard(tsdn_t *tsdn, arena_t *arena, tcache_t *tcache, + cache_bin_t *tbin, szind_t binind, bool *tcache_success); +void tcache_bin_flush_small(tsd_t *tsd, tcache_t *tcache, cache_bin_t *tbin, + szind_t binind, unsigned rem); +void tcache_bin_flush_large(tsd_t *tsd, cache_bin_t *tbin, szind_t binind, + unsigned rem, tcache_t *tcache); +void tcache_arena_reassociate(tsdn_t *tsdn, tcache_t *tcache, + arena_t *arena); +tcache_t *tcache_create_explicit(tsd_t *tsd); +void tcache_cleanup(tsd_t *tsd); +void tcache_stats_merge(tsdn_t *tsdn, tcache_t *tcache, arena_t *arena); +bool tcaches_create(tsd_t *tsd, unsigned *r_ind); +void tcaches_flush(tsd_t *tsd, unsigned ind); +void tcaches_destroy(tsd_t *tsd, unsigned ind); +bool tcache_boot(tsdn_t *tsdn); +void tcache_arena_associate(tsdn_t *tsdn, tcache_t *tcache, arena_t *arena); +void tcache_prefork(tsdn_t *tsdn); +void tcache_postfork_parent(tsdn_t *tsdn); +void tcache_postfork_child(tsdn_t *tsdn); +void tcache_flush(tsd_t *tsd); +bool tsd_tcache_data_init(tsd_t *tsd); +bool tsd_tcache_enabled_data_init(tsd_t *tsd); + +#endif /* JEMALLOC_INTERNAL_TCACHE_EXTERNS_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/tcache_inlines.h b/deps/jemalloc/include/jemalloc/internal/tcache_inlines.h new file mode 100644 index 000000000..5eca20e89 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/tcache_inlines.h @@ -0,0 +1,227 @@ +#ifndef JEMALLOC_INTERNAL_TCACHE_INLINES_H +#define JEMALLOC_INTERNAL_TCACHE_INLINES_H + +#include "jemalloc/internal/bin.h" +#include "jemalloc/internal/jemalloc_internal_types.h" +#include "jemalloc/internal/sc.h" +#include "jemalloc/internal/sz.h" +#include "jemalloc/internal/ticker.h" +#include "jemalloc/internal/util.h" + +static inline bool +tcache_enabled_get(tsd_t *tsd) { + return tsd_tcache_enabled_get(tsd); +} + +static inline void +tcache_enabled_set(tsd_t *tsd, bool enabled) { + bool was_enabled = tsd_tcache_enabled_get(tsd); + + if (!was_enabled && enabled) { + tsd_tcache_data_init(tsd); + } else if (was_enabled && !enabled) { + tcache_cleanup(tsd); + } + /* Commit the state last. Above calls check current state. */ + tsd_tcache_enabled_set(tsd, enabled); + tsd_slow_update(tsd); +} + +JEMALLOC_ALWAYS_INLINE void +tcache_event(tsd_t *tsd, tcache_t *tcache) { + if (TCACHE_GC_INCR == 0) { + return; + } + + if (unlikely(ticker_tick(&tcache->gc_ticker))) { + tcache_event_hard(tsd, tcache); + } +} + +JEMALLOC_ALWAYS_INLINE void * +tcache_alloc_small(tsd_t *tsd, arena_t *arena, tcache_t *tcache, + size_t size, szind_t binind, bool zero, bool slow_path) { + void *ret; + cache_bin_t *bin; + bool tcache_success; + size_t usize JEMALLOC_CC_SILENCE_INIT(0); + + assert(binind < SC_NBINS); + bin = tcache_small_bin_get(tcache, binind); + ret = cache_bin_alloc_easy(bin, &tcache_success); + assert(tcache_success == (ret != NULL)); + if (unlikely(!tcache_success)) { + bool tcache_hard_success; + arena = arena_choose(tsd, arena); + if (unlikely(arena == NULL)) { + return NULL; + } + + ret = tcache_alloc_small_hard(tsd_tsdn(tsd), arena, tcache, + bin, binind, &tcache_hard_success); + if (tcache_hard_success == false) { + return NULL; + } + } + + assert(ret); + /* + * Only compute usize if required. The checks in the following if + * statement are all static. + */ + if (config_prof || (slow_path && config_fill) || unlikely(zero)) { + usize = sz_index2size(binind); + assert(tcache_salloc(tsd_tsdn(tsd), ret) == usize); + } + + if (likely(!zero)) { + if (slow_path && config_fill) { + if (unlikely(opt_junk_alloc)) { + arena_alloc_junk_small(ret, &bin_infos[binind], + false); + } else if (unlikely(opt_zero)) { + memset(ret, 0, usize); + } + } + } else { + if (slow_path && config_fill && unlikely(opt_junk_alloc)) { + arena_alloc_junk_small(ret, &bin_infos[binind], true); + } + memset(ret, 0, usize); + } + + if (config_stats) { + bin->tstats.nrequests++; + } + if (config_prof) { + tcache->prof_accumbytes += usize; + } + tcache_event(tsd, tcache); + return ret; +} + +JEMALLOC_ALWAYS_INLINE void * +tcache_alloc_large(tsd_t *tsd, arena_t *arena, tcache_t *tcache, size_t size, + szind_t binind, bool zero, bool slow_path) { + void *ret; + cache_bin_t *bin; + bool tcache_success; + + assert(binind >= SC_NBINS &&binind < nhbins); + bin = tcache_large_bin_get(tcache, binind); + ret = cache_bin_alloc_easy(bin, &tcache_success); + assert(tcache_success == (ret != NULL)); + if (unlikely(!tcache_success)) { + /* + * Only allocate one large object at a time, because it's quite + * expensive to create one and not use it. + */ + arena = arena_choose(tsd, arena); + if (unlikely(arena == NULL)) { + return NULL; + } + + ret = large_malloc(tsd_tsdn(tsd), arena, sz_s2u(size), zero); + if (ret == NULL) { + return NULL; + } + } else { + size_t usize JEMALLOC_CC_SILENCE_INIT(0); + + /* Only compute usize on demand */ + if (config_prof || (slow_path && config_fill) || + unlikely(zero)) { + usize = sz_index2size(binind); + assert(usize <= tcache_maxclass); + } + + if (likely(!zero)) { + if (slow_path && config_fill) { + if (unlikely(opt_junk_alloc)) { + memset(ret, JEMALLOC_ALLOC_JUNK, + usize); + } else if (unlikely(opt_zero)) { + memset(ret, 0, usize); + } + } + } else { + memset(ret, 0, usize); + } + + if (config_stats) { + bin->tstats.nrequests++; + } + if (config_prof) { + tcache->prof_accumbytes += usize; + } + } + + tcache_event(tsd, tcache); + return ret; +} + +JEMALLOC_ALWAYS_INLINE void +tcache_dalloc_small(tsd_t *tsd, tcache_t *tcache, void *ptr, szind_t binind, + bool slow_path) { + cache_bin_t *bin; + cache_bin_info_t *bin_info; + + assert(tcache_salloc(tsd_tsdn(tsd), ptr) + <= SC_SMALL_MAXCLASS); + + if (slow_path && config_fill && unlikely(opt_junk_free)) { + arena_dalloc_junk_small(ptr, &bin_infos[binind]); + } + + bin = tcache_small_bin_get(tcache, binind); + bin_info = &tcache_bin_info[binind]; + if (unlikely(!cache_bin_dalloc_easy(bin, bin_info, ptr))) { + tcache_bin_flush_small(tsd, tcache, bin, binind, + (bin_info->ncached_max >> 1)); + bool ret = cache_bin_dalloc_easy(bin, bin_info, ptr); + assert(ret); + } + + tcache_event(tsd, tcache); +} + +JEMALLOC_ALWAYS_INLINE void +tcache_dalloc_large(tsd_t *tsd, tcache_t *tcache, void *ptr, szind_t binind, + bool slow_path) { + cache_bin_t *bin; + cache_bin_info_t *bin_info; + + assert(tcache_salloc(tsd_tsdn(tsd), ptr) + > SC_SMALL_MAXCLASS); + assert(tcache_salloc(tsd_tsdn(tsd), ptr) <= tcache_maxclass); + + if (slow_path && config_fill && unlikely(opt_junk_free)) { + large_dalloc_junk(ptr, sz_index2size(binind)); + } + + bin = tcache_large_bin_get(tcache, binind); + bin_info = &tcache_bin_info[binind]; + if (unlikely(bin->ncached == bin_info->ncached_max)) { + tcache_bin_flush_large(tsd, bin, binind, + (bin_info->ncached_max >> 1), tcache); + } + assert(bin->ncached < bin_info->ncached_max); + bin->ncached++; + *(bin->avail - bin->ncached) = ptr; + + tcache_event(tsd, tcache); +} + +JEMALLOC_ALWAYS_INLINE tcache_t * +tcaches_get(tsd_t *tsd, unsigned ind) { + tcaches_t *elm = &tcaches[ind]; + if (unlikely(elm->tcache == NULL)) { + malloc_printf("<jemalloc>: invalid tcache id (%u).\n", ind); + abort(); + } else if (unlikely(elm->tcache == TCACHES_ELM_NEED_REINIT)) { + elm->tcache = tcache_create_explicit(tsd); + } + return elm->tcache; +} + +#endif /* JEMALLOC_INTERNAL_TCACHE_INLINES_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/tcache_structs.h b/deps/jemalloc/include/jemalloc/internal/tcache_structs.h new file mode 100644 index 000000000..172ef9040 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/tcache_structs.h @@ -0,0 +1,70 @@ +#ifndef JEMALLOC_INTERNAL_TCACHE_STRUCTS_H +#define JEMALLOC_INTERNAL_TCACHE_STRUCTS_H + +#include "jemalloc/internal/cache_bin.h" +#include "jemalloc/internal/ql.h" +#include "jemalloc/internal/sc.h" +#include "jemalloc/internal/ticker.h" +#include "jemalloc/internal/tsd_types.h" + +/* Various uses of this struct need it to be a named type. */ +typedef ql_elm(tsd_t) tsd_link_t; + +struct tcache_s { + /* + * To minimize our cache-footprint, we put the frequently accessed data + * together at the start of this struct. + */ + + /* Cleared after arena_prof_accum(). */ + uint64_t prof_accumbytes; + /* Drives incremental GC. */ + ticker_t gc_ticker; + /* + * The pointer stacks associated with bins follow as a contiguous array. + * During tcache initialization, the avail pointer in each element of + * tbins is initialized to point to the proper offset within this array. + */ + cache_bin_t bins_small[SC_NBINS]; + + /* + * This data is less hot; we can be a little less careful with our + * footprint here. + */ + /* Lets us track all the tcaches in an arena. */ + ql_elm(tcache_t) link; + + /* Logically scoped to tsd, but put here for cache layout reasons. */ + ql_elm(tsd_t) tsd_link; + bool in_hook; + + /* + * The descriptor lets the arena find our cache bins without seeing the + * tcache definition. This enables arenas to aggregate stats across + * tcaches without having a tcache dependency. + */ + cache_bin_array_descriptor_t cache_bin_array_descriptor; + + /* The arena this tcache is associated with. */ + arena_t *arena; + /* Next bin to GC. */ + szind_t next_gc_bin; + /* For small bins, fill (ncached_max >> lg_fill_div). */ + uint8_t lg_fill_div[SC_NBINS]; + /* + * We put the cache bins for large size classes at the end of the + * struct, since some of them might not get used. This might end up + * letting us avoid touching an extra page if we don't have to. + */ + cache_bin_t bins_large[SC_NSIZES-SC_NBINS]; +}; + +/* Linkage for list of available (previously used) explicit tcache IDs. */ +struct tcaches_s { + union { + tcache_t *tcache; + tcaches_t *next; + }; +}; + +#endif /* JEMALLOC_INTERNAL_TCACHE_STRUCTS_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/tcache_types.h b/deps/jemalloc/include/jemalloc/internal/tcache_types.h new file mode 100644 index 000000000..dce69382e --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/tcache_types.h @@ -0,0 +1,59 @@ +#ifndef JEMALLOC_INTERNAL_TCACHE_TYPES_H +#define JEMALLOC_INTERNAL_TCACHE_TYPES_H + +#include "jemalloc/internal/sc.h" + +typedef struct tcache_s tcache_t; +typedef struct tcaches_s tcaches_t; + +/* + * tcache pointers close to NULL are used to encode state information that is + * used for two purposes: preventing thread caching on a per thread basis and + * cleaning up during thread shutdown. + */ +#define TCACHE_STATE_DISABLED ((tcache_t *)(uintptr_t)1) +#define TCACHE_STATE_REINCARNATED ((tcache_t *)(uintptr_t)2) +#define TCACHE_STATE_PURGATORY ((tcache_t *)(uintptr_t)3) +#define TCACHE_STATE_MAX TCACHE_STATE_PURGATORY + +/* + * Absolute minimum number of cache slots for each small bin. + */ +#define TCACHE_NSLOTS_SMALL_MIN 20 + +/* + * Absolute maximum number of cache slots for each small bin in the thread + * cache. This is an additional constraint beyond that imposed as: twice the + * number of regions per slab for this size class. + * + * This constant must be an even number. + */ +#define TCACHE_NSLOTS_SMALL_MAX 200 + +/* Number of cache slots for large size classes. */ +#define TCACHE_NSLOTS_LARGE 20 + +/* (1U << opt_lg_tcache_max) is used to compute tcache_maxclass. */ +#define LG_TCACHE_MAXCLASS_DEFAULT 15 + +/* + * TCACHE_GC_SWEEP is the approximate number of allocation events between + * full GC sweeps. Integer rounding may cause the actual number to be + * slightly higher, since GC is performed incrementally. + */ +#define TCACHE_GC_SWEEP 8192 + +/* Number of tcache allocation/deallocation events between incremental GCs. */ +#define TCACHE_GC_INCR \ + ((TCACHE_GC_SWEEP / SC_NBINS) + ((TCACHE_GC_SWEEP / SC_NBINS == 0) ? 0 : 1)) + +/* Used in TSD static initializer only. Real init in tcache_data_init(). */ +#define TCACHE_ZERO_INITIALIZER {0} + +/* Used in TSD static initializer only. Will be initialized to opt_tcache. */ +#define TCACHE_ENABLED_ZERO_INITIALIZER false + +/* Used for explicit tcache only. Means flushed but not destroyed. */ +#define TCACHES_ELM_NEED_REINIT ((tcache_t *)(uintptr_t)1) + +#endif /* JEMALLOC_INTERNAL_TCACHE_TYPES_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/test_hooks.h b/deps/jemalloc/include/jemalloc/internal/test_hooks.h new file mode 100644 index 000000000..a6351e59a --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/test_hooks.h @@ -0,0 +1,19 @@ +#ifndef JEMALLOC_INTERNAL_TEST_HOOKS_H +#define JEMALLOC_INTERNAL_TEST_HOOKS_H + +extern JEMALLOC_EXPORT void (*test_hooks_arena_new_hook)(); +extern JEMALLOC_EXPORT void (*test_hooks_libc_hook)(); + +#define JEMALLOC_HOOK(fn, hook) ((void)(hook != NULL && (hook(), 0)), fn) + +#define open JEMALLOC_HOOK(open, test_hooks_libc_hook) +#define read JEMALLOC_HOOK(read, test_hooks_libc_hook) +#define write JEMALLOC_HOOK(write, test_hooks_libc_hook) +#define readlink JEMALLOC_HOOK(readlink, test_hooks_libc_hook) +#define close JEMALLOC_HOOK(close, test_hooks_libc_hook) +#define creat JEMALLOC_HOOK(creat, test_hooks_libc_hook) +#define secure_getenv JEMALLOC_HOOK(secure_getenv, test_hooks_libc_hook) +/* Note that this is undef'd and re-define'd in src/prof.c. */ +#define _Unwind_Backtrace JEMALLOC_HOOK(_Unwind_Backtrace, test_hooks_libc_hook) + +#endif /* JEMALLOC_INTERNAL_TEST_HOOKS_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/ticker.h b/deps/jemalloc/include/jemalloc/internal/ticker.h new file mode 100644 index 000000000..52d0db4c8 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/ticker.h @@ -0,0 +1,91 @@ +#ifndef JEMALLOC_INTERNAL_TICKER_H +#define JEMALLOC_INTERNAL_TICKER_H + +#include "jemalloc/internal/util.h" + +/** + * A ticker makes it easy to count-down events until some limit. You + * ticker_init the ticker to trigger every nticks events. You then notify it + * that an event has occurred with calls to ticker_tick (or that nticks events + * have occurred with a call to ticker_ticks), which will return true (and reset + * the counter) if the countdown hit zero. + */ + +typedef struct { + int32_t tick; + int32_t nticks; +} ticker_t; + +static inline void +ticker_init(ticker_t *ticker, int32_t nticks) { + ticker->tick = nticks; + ticker->nticks = nticks; +} + +static inline void +ticker_copy(ticker_t *ticker, const ticker_t *other) { + *ticker = *other; +} + +static inline int32_t +ticker_read(const ticker_t *ticker) { + return ticker->tick; +} + +/* + * Not intended to be a public API. Unfortunately, on x86, neither gcc nor + * clang seems smart enough to turn + * ticker->tick -= nticks; + * if (unlikely(ticker->tick < 0)) { + * fixup ticker + * return true; + * } + * return false; + * into + * subq %nticks_reg, (%ticker_reg) + * js fixup ticker + * + * unless we force "fixup ticker" out of line. In that case, gcc gets it right, + * but clang now does worse than before. So, on x86 with gcc, we force it out + * of line, but otherwise let the inlining occur. Ordinarily this wouldn't be + * worth the hassle, but this is on the fast path of both malloc and free (via + * tcache_event). + */ +#if defined(__GNUC__) && !defined(__clang__) \ + && (defined(__x86_64__) || defined(__i386__)) +JEMALLOC_NOINLINE +#endif +static bool +ticker_fixup(ticker_t *ticker) { + ticker->tick = ticker->nticks; + return true; +} + +static inline bool +ticker_ticks(ticker_t *ticker, int32_t nticks) { + ticker->tick -= nticks; + if (unlikely(ticker->tick < 0)) { + return ticker_fixup(ticker); + } + return false; +} + +static inline bool +ticker_tick(ticker_t *ticker) { + return ticker_ticks(ticker, 1); +} + +/* + * Try to tick. If ticker would fire, return true, but rely on + * slowpath to reset ticker. + */ +static inline bool +ticker_trytick(ticker_t *ticker) { + --ticker->tick; + if (unlikely(ticker->tick < 0)) { + return true; + } + return false; +} + +#endif /* JEMALLOC_INTERNAL_TICKER_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/tsd.h b/deps/jemalloc/include/jemalloc/internal/tsd.h new file mode 100644 index 000000000..9ba260045 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/tsd.h @@ -0,0 +1,415 @@ +#ifndef JEMALLOC_INTERNAL_TSD_H +#define JEMALLOC_INTERNAL_TSD_H + +#include "jemalloc/internal/arena_types.h" +#include "jemalloc/internal/assert.h" +#include "jemalloc/internal/bin_types.h" +#include "jemalloc/internal/jemalloc_internal_externs.h" +#include "jemalloc/internal/prof_types.h" +#include "jemalloc/internal/ql.h" +#include "jemalloc/internal/rtree_tsd.h" +#include "jemalloc/internal/tcache_types.h" +#include "jemalloc/internal/tcache_structs.h" +#include "jemalloc/internal/util.h" +#include "jemalloc/internal/witness.h" + +/* + * Thread-Specific-Data layout + * --- data accessed on tcache fast path: state, rtree_ctx, stats, prof --- + * s: state + * e: tcache_enabled + * m: thread_allocated (config_stats) + * f: thread_deallocated (config_stats) + * p: prof_tdata (config_prof) + * c: rtree_ctx (rtree cache accessed on deallocation) + * t: tcache + * --- data not accessed on tcache fast path: arena-related fields --- + * d: arenas_tdata_bypass + * r: reentrancy_level + * x: narenas_tdata + * i: iarena + * a: arena + * o: arenas_tdata + * Loading TSD data is on the critical path of basically all malloc operations. + * In particular, tcache and rtree_ctx rely on hot CPU cache to be effective. + * Use a compact layout to reduce cache footprint. + * +--- 64-bit and 64B cacheline; 1B each letter; First byte on the left. ---+ + * |---------------------------- 1st cacheline ----------------------------| + * | sedrxxxx mmmmmmmm ffffffff pppppppp [c * 32 ........ ........ .......] | + * |---------------------------- 2nd cacheline ----------------------------| + * | [c * 64 ........ ........ ........ ........ ........ ........ .......] | + * |---------------------------- 3nd cacheline ----------------------------| + * | [c * 32 ........ ........ .......] iiiiiiii aaaaaaaa oooooooo [t...... | + * +-------------------------------------------------------------------------+ + * Note: the entire tcache is embedded into TSD and spans multiple cachelines. + * + * The last 3 members (i, a and o) before tcache isn't really needed on tcache + * fast path. However we have a number of unused tcache bins and witnesses + * (never touched unless config_debug) at the end of tcache, so we place them + * there to avoid breaking the cachelines and possibly paging in an extra page. + */ +#ifdef JEMALLOC_JET +typedef void (*test_callback_t)(int *); +# define MALLOC_TSD_TEST_DATA_INIT 0x72b65c10 +# define MALLOC_TEST_TSD \ + O(test_data, int, int) \ + O(test_callback, test_callback_t, int) +# define MALLOC_TEST_TSD_INITIALIZER , MALLOC_TSD_TEST_DATA_INIT, NULL +#else +# define MALLOC_TEST_TSD +# define MALLOC_TEST_TSD_INITIALIZER +#endif + +/* O(name, type, nullable type */ +#define MALLOC_TSD \ + O(tcache_enabled, bool, bool) \ + O(arenas_tdata_bypass, bool, bool) \ + O(reentrancy_level, int8_t, int8_t) \ + O(narenas_tdata, uint32_t, uint32_t) \ + O(offset_state, uint64_t, uint64_t) \ + O(thread_allocated, uint64_t, uint64_t) \ + O(thread_deallocated, uint64_t, uint64_t) \ + O(bytes_until_sample, int64_t, int64_t) \ + O(prof_tdata, prof_tdata_t *, prof_tdata_t *) \ + O(rtree_ctx, rtree_ctx_t, rtree_ctx_t) \ + O(iarena, arena_t *, arena_t *) \ + O(arena, arena_t *, arena_t *) \ + O(arenas_tdata, arena_tdata_t *, arena_tdata_t *)\ + O(binshards, tsd_binshards_t, tsd_binshards_t)\ + O(tcache, tcache_t, tcache_t) \ + O(witness_tsd, witness_tsd_t, witness_tsdn_t) \ + MALLOC_TEST_TSD + +#define TSD_INITIALIZER { \ + ATOMIC_INIT(tsd_state_uninitialized), \ + TCACHE_ENABLED_ZERO_INITIALIZER, \ + false, \ + 0, \ + 0, \ + 0, \ + 0, \ + 0, \ + 0, \ + NULL, \ + RTREE_CTX_ZERO_INITIALIZER, \ + NULL, \ + NULL, \ + NULL, \ + TSD_BINSHARDS_ZERO_INITIALIZER, \ + TCACHE_ZERO_INITIALIZER, \ + WITNESS_TSD_INITIALIZER \ + MALLOC_TEST_TSD_INITIALIZER \ +} + +void *malloc_tsd_malloc(size_t size); +void malloc_tsd_dalloc(void *wrapper); +void malloc_tsd_cleanup_register(bool (*f)(void)); +tsd_t *malloc_tsd_boot0(void); +void malloc_tsd_boot1(void); +void tsd_cleanup(void *arg); +tsd_t *tsd_fetch_slow(tsd_t *tsd, bool internal); +void tsd_state_set(tsd_t *tsd, uint8_t new_state); +void tsd_slow_update(tsd_t *tsd); +void tsd_prefork(tsd_t *tsd); +void tsd_postfork_parent(tsd_t *tsd); +void tsd_postfork_child(tsd_t *tsd); + +/* + * Call ..._inc when your module wants to take all threads down the slow paths, + * and ..._dec when it no longer needs to. + */ +void tsd_global_slow_inc(tsdn_t *tsdn); +void tsd_global_slow_dec(tsdn_t *tsdn); +bool tsd_global_slow(); + +enum { + /* Common case --> jnz. */ + tsd_state_nominal = 0, + /* Initialized but on slow path. */ + tsd_state_nominal_slow = 1, + /* + * Some thread has changed global state in such a way that all nominal + * threads need to recompute their fast / slow status the next time they + * get a chance. + * + * Any thread can change another thread's status *to* recompute, but + * threads are the only ones who can change their status *from* + * recompute. + */ + tsd_state_nominal_recompute = 2, + /* + * The above nominal states should be lower values. We use + * tsd_nominal_max to separate nominal states from threads in the + * process of being born / dying. + */ + tsd_state_nominal_max = 2, + + /* + * A thread might free() during its death as its only allocator action; + * in such scenarios, we need tsd, but set up in such a way that no + * cleanup is necessary. + */ + tsd_state_minimal_initialized = 3, + /* States during which we know we're in thread death. */ + tsd_state_purgatory = 4, + tsd_state_reincarnated = 5, + /* + * What it says on the tin; tsd that hasn't been initialized. Note + * that even when the tsd struct lives in TLS, when need to keep track + * of stuff like whether or not our pthread destructors have been + * scheduled, so this really truly is different than the nominal state. + */ + tsd_state_uninitialized = 6 +}; + +/* + * Some TSD accesses can only be done in a nominal state. To enforce this, we + * wrap TSD member access in a function that asserts on TSD state, and mangle + * field names to prevent touching them accidentally. + */ +#define TSD_MANGLE(n) cant_access_tsd_items_directly_use_a_getter_or_setter_##n + +#ifdef JEMALLOC_U8_ATOMICS +# define tsd_state_t atomic_u8_t +# define tsd_atomic_load atomic_load_u8 +# define tsd_atomic_store atomic_store_u8 +# define tsd_atomic_exchange atomic_exchange_u8 +#else +# define tsd_state_t atomic_u32_t +# define tsd_atomic_load atomic_load_u32 +# define tsd_atomic_store atomic_store_u32 +# define tsd_atomic_exchange atomic_exchange_u32 +#endif + +/* The actual tsd. */ +struct tsd_s { + /* + * The contents should be treated as totally opaque outside the tsd + * module. Access any thread-local state through the getters and + * setters below. + */ + + /* + * We manually limit the state to just a single byte. Unless the 8-bit + * atomics are unavailable (which is rare). + */ + tsd_state_t state; +#define O(n, t, nt) \ + t TSD_MANGLE(n); +MALLOC_TSD +#undef O +}; + +JEMALLOC_ALWAYS_INLINE uint8_t +tsd_state_get(tsd_t *tsd) { + /* + * This should be atomic. Unfortunately, compilers right now can't tell + * that this can be done as a memory comparison, and forces a load into + * a register that hurts fast-path performance. + */ + /* return atomic_load_u8(&tsd->state, ATOMIC_RELAXED); */ + return *(uint8_t *)&tsd->state; +} + +/* + * Wrapper around tsd_t that makes it possible to avoid implicit conversion + * between tsd_t and tsdn_t, where tsdn_t is "nullable" and has to be + * explicitly converted to tsd_t, which is non-nullable. + */ +struct tsdn_s { + tsd_t tsd; +}; +#define TSDN_NULL ((tsdn_t *)0) +JEMALLOC_ALWAYS_INLINE tsdn_t * +tsd_tsdn(tsd_t *tsd) { + return (tsdn_t *)tsd; +} + +JEMALLOC_ALWAYS_INLINE bool +tsdn_null(const tsdn_t *tsdn) { + return tsdn == NULL; +} + +JEMALLOC_ALWAYS_INLINE tsd_t * +tsdn_tsd(tsdn_t *tsdn) { + assert(!tsdn_null(tsdn)); + + return &tsdn->tsd; +} + +/* + * We put the platform-specific data declarations and inlines into their own + * header files to avoid cluttering this file. They define tsd_boot0, + * tsd_boot1, tsd_boot, tsd_booted_get, tsd_get_allocates, tsd_get, and tsd_set. + */ +#ifdef JEMALLOC_MALLOC_THREAD_CLEANUP +#include "jemalloc/internal/tsd_malloc_thread_cleanup.h" +#elif (defined(JEMALLOC_TLS)) +#include "jemalloc/internal/tsd_tls.h" +#elif (defined(_WIN32)) +#include "jemalloc/internal/tsd_win.h" +#else +#include "jemalloc/internal/tsd_generic.h" +#endif + +/* + * tsd_foop_get_unsafe(tsd) returns a pointer to the thread-local instance of + * foo. This omits some safety checks, and so can be used during tsd + * initialization and cleanup. + */ +#define O(n, t, nt) \ +JEMALLOC_ALWAYS_INLINE t * \ +tsd_##n##p_get_unsafe(tsd_t *tsd) { \ + return &tsd->TSD_MANGLE(n); \ +} +MALLOC_TSD +#undef O + +/* tsd_foop_get(tsd) returns a pointer to the thread-local instance of foo. */ +#define O(n, t, nt) \ +JEMALLOC_ALWAYS_INLINE t * \ +tsd_##n##p_get(tsd_t *tsd) { \ + /* \ + * Because the state might change asynchronously if it's \ + * nominal, we need to make sure that we only read it once. \ + */ \ + uint8_t state = tsd_state_get(tsd); \ + assert(state == tsd_state_nominal || \ + state == tsd_state_nominal_slow || \ + state == tsd_state_nominal_recompute || \ + state == tsd_state_reincarnated || \ + state == tsd_state_minimal_initialized); \ + return tsd_##n##p_get_unsafe(tsd); \ +} +MALLOC_TSD +#undef O + +/* + * tsdn_foop_get(tsdn) returns either the thread-local instance of foo (if tsdn + * isn't NULL), or NULL (if tsdn is NULL), cast to the nullable pointer type. + */ +#define O(n, t, nt) \ +JEMALLOC_ALWAYS_INLINE nt * \ +tsdn_##n##p_get(tsdn_t *tsdn) { \ + if (tsdn_null(tsdn)) { \ + return NULL; \ + } \ + tsd_t *tsd = tsdn_tsd(tsdn); \ + return (nt *)tsd_##n##p_get(tsd); \ +} +MALLOC_TSD +#undef O + +/* tsd_foo_get(tsd) returns the value of the thread-local instance of foo. */ +#define O(n, t, nt) \ +JEMALLOC_ALWAYS_INLINE t \ +tsd_##n##_get(tsd_t *tsd) { \ + return *tsd_##n##p_get(tsd); \ +} +MALLOC_TSD +#undef O + +/* tsd_foo_set(tsd, val) updates the thread-local instance of foo to be val. */ +#define O(n, t, nt) \ +JEMALLOC_ALWAYS_INLINE void \ +tsd_##n##_set(tsd_t *tsd, t val) { \ + assert(tsd_state_get(tsd) != tsd_state_reincarnated && \ + tsd_state_get(tsd) != tsd_state_minimal_initialized); \ + *tsd_##n##p_get(tsd) = val; \ +} +MALLOC_TSD +#undef O + +JEMALLOC_ALWAYS_INLINE void +tsd_assert_fast(tsd_t *tsd) { + /* + * Note that our fastness assertion does *not* include global slowness + * counters; it's not in general possible to ensure that they won't + * change asynchronously from underneath us. + */ + assert(!malloc_slow && tsd_tcache_enabled_get(tsd) && + tsd_reentrancy_level_get(tsd) == 0); +} + +JEMALLOC_ALWAYS_INLINE bool +tsd_fast(tsd_t *tsd) { + bool fast = (tsd_state_get(tsd) == tsd_state_nominal); + if (fast) { + tsd_assert_fast(tsd); + } + + return fast; +} + +JEMALLOC_ALWAYS_INLINE tsd_t * +tsd_fetch_impl(bool init, bool minimal) { + tsd_t *tsd = tsd_get(init); + + if (!init && tsd_get_allocates() && tsd == NULL) { + return NULL; + } + assert(tsd != NULL); + + if (unlikely(tsd_state_get(tsd) != tsd_state_nominal)) { + return tsd_fetch_slow(tsd, minimal); + } + assert(tsd_fast(tsd)); + tsd_assert_fast(tsd); + + return tsd; +} + +/* Get a minimal TSD that requires no cleanup. See comments in free(). */ +JEMALLOC_ALWAYS_INLINE tsd_t * +tsd_fetch_min(void) { + return tsd_fetch_impl(true, true); +} + +/* For internal background threads use only. */ +JEMALLOC_ALWAYS_INLINE tsd_t * +tsd_internal_fetch(void) { + tsd_t *tsd = tsd_fetch_min(); + /* Use reincarnated state to prevent full initialization. */ + tsd_state_set(tsd, tsd_state_reincarnated); + + return tsd; +} + +JEMALLOC_ALWAYS_INLINE tsd_t * +tsd_fetch(void) { + return tsd_fetch_impl(true, false); +} + +static inline bool +tsd_nominal(tsd_t *tsd) { + return (tsd_state_get(tsd) <= tsd_state_nominal_max); +} + +JEMALLOC_ALWAYS_INLINE tsdn_t * +tsdn_fetch(void) { + if (!tsd_booted_get()) { + return NULL; + } + + return tsd_tsdn(tsd_fetch_impl(false, false)); +} + +JEMALLOC_ALWAYS_INLINE rtree_ctx_t * +tsd_rtree_ctx(tsd_t *tsd) { + return tsd_rtree_ctxp_get(tsd); +} + +JEMALLOC_ALWAYS_INLINE rtree_ctx_t * +tsdn_rtree_ctx(tsdn_t *tsdn, rtree_ctx_t *fallback) { + /* + * If tsd cannot be accessed, initialize the fallback rtree_ctx and + * return a pointer to it. + */ + if (unlikely(tsdn_null(tsdn))) { + rtree_ctx_data_init(fallback); + return fallback; + } + return tsd_rtree_ctx(tsdn_tsd(tsdn)); +} + +#endif /* JEMALLOC_INTERNAL_TSD_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/tsd_generic.h b/deps/jemalloc/include/jemalloc/internal/tsd_generic.h new file mode 100644 index 000000000..cf73c0c71 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/tsd_generic.h @@ -0,0 +1,163 @@ +#ifdef JEMALLOC_INTERNAL_TSD_GENERIC_H +#error This file should be included only once, by tsd.h. +#endif +#define JEMALLOC_INTERNAL_TSD_GENERIC_H + +typedef struct tsd_init_block_s tsd_init_block_t; +struct tsd_init_block_s { + ql_elm(tsd_init_block_t) link; + pthread_t thread; + void *data; +}; + +/* Defined in tsd.c, to allow the mutex headers to have tsd dependencies. */ +typedef struct tsd_init_head_s tsd_init_head_t; + +typedef struct { + bool initialized; + tsd_t val; +} tsd_wrapper_t; + +void *tsd_init_check_recursion(tsd_init_head_t *head, + tsd_init_block_t *block); +void tsd_init_finish(tsd_init_head_t *head, tsd_init_block_t *block); + +extern pthread_key_t tsd_tsd; +extern tsd_init_head_t tsd_init_head; +extern tsd_wrapper_t tsd_boot_wrapper; +extern bool tsd_booted; + +/* Initialization/cleanup. */ +JEMALLOC_ALWAYS_INLINE void +tsd_cleanup_wrapper(void *arg) { + tsd_wrapper_t *wrapper = (tsd_wrapper_t *)arg; + + if (wrapper->initialized) { + wrapper->initialized = false; + tsd_cleanup(&wrapper->val); + if (wrapper->initialized) { + /* Trigger another cleanup round. */ + if (pthread_setspecific(tsd_tsd, (void *)wrapper) != 0) + { + malloc_write("<jemalloc>: Error setting TSD\n"); + if (opt_abort) { + abort(); + } + } + return; + } + } + malloc_tsd_dalloc(wrapper); +} + +JEMALLOC_ALWAYS_INLINE void +tsd_wrapper_set(tsd_wrapper_t *wrapper) { + if (pthread_setspecific(tsd_tsd, (void *)wrapper) != 0) { + malloc_write("<jemalloc>: Error setting TSD\n"); + abort(); + } +} + +JEMALLOC_ALWAYS_INLINE tsd_wrapper_t * +tsd_wrapper_get(bool init) { + tsd_wrapper_t *wrapper = (tsd_wrapper_t *)pthread_getspecific(tsd_tsd); + + if (init && unlikely(wrapper == NULL)) { + tsd_init_block_t block; + wrapper = (tsd_wrapper_t *) + tsd_init_check_recursion(&tsd_init_head, &block); + if (wrapper) { + return wrapper; + } + wrapper = (tsd_wrapper_t *) + malloc_tsd_malloc(sizeof(tsd_wrapper_t)); + block.data = (void *)wrapper; + if (wrapper == NULL) { + malloc_write("<jemalloc>: Error allocating TSD\n"); + abort(); + } else { + wrapper->initialized = false; + JEMALLOC_DIAGNOSTIC_PUSH + JEMALLOC_DIAGNOSTIC_IGNORE_MISSING_STRUCT_FIELD_INITIALIZERS + tsd_t initializer = TSD_INITIALIZER; + JEMALLOC_DIAGNOSTIC_POP + wrapper->val = initializer; + } + tsd_wrapper_set(wrapper); + tsd_init_finish(&tsd_init_head, &block); + } + return wrapper; +} + +JEMALLOC_ALWAYS_INLINE bool +tsd_boot0(void) { + if (pthread_key_create(&tsd_tsd, tsd_cleanup_wrapper) != 0) { + return true; + } + tsd_wrapper_set(&tsd_boot_wrapper); + tsd_booted = true; + return false; +} + +JEMALLOC_ALWAYS_INLINE void +tsd_boot1(void) { + tsd_wrapper_t *wrapper; + wrapper = (tsd_wrapper_t *)malloc_tsd_malloc(sizeof(tsd_wrapper_t)); + if (wrapper == NULL) { + malloc_write("<jemalloc>: Error allocating TSD\n"); + abort(); + } + tsd_boot_wrapper.initialized = false; + tsd_cleanup(&tsd_boot_wrapper.val); + wrapper->initialized = false; + JEMALLOC_DIAGNOSTIC_PUSH + JEMALLOC_DIAGNOSTIC_IGNORE_MISSING_STRUCT_FIELD_INITIALIZERS + tsd_t initializer = TSD_INITIALIZER; + JEMALLOC_DIAGNOSTIC_POP + wrapper->val = initializer; + tsd_wrapper_set(wrapper); +} + +JEMALLOC_ALWAYS_INLINE bool +tsd_boot(void) { + if (tsd_boot0()) { + return true; + } + tsd_boot1(); + return false; +} + +JEMALLOC_ALWAYS_INLINE bool +tsd_booted_get(void) { + return tsd_booted; +} + +JEMALLOC_ALWAYS_INLINE bool +tsd_get_allocates(void) { + return true; +} + +/* Get/set. */ +JEMALLOC_ALWAYS_INLINE tsd_t * +tsd_get(bool init) { + tsd_wrapper_t *wrapper; + + assert(tsd_booted); + wrapper = tsd_wrapper_get(init); + if (tsd_get_allocates() && !init && wrapper == NULL) { + return NULL; + } + return &wrapper->val; +} + +JEMALLOC_ALWAYS_INLINE void +tsd_set(tsd_t *val) { + tsd_wrapper_t *wrapper; + + assert(tsd_booted); + wrapper = tsd_wrapper_get(true); + if (likely(&wrapper->val != val)) { + wrapper->val = *(val); + } + wrapper->initialized = true; +} diff --git a/deps/jemalloc/include/jemalloc/internal/tsd_malloc_thread_cleanup.h b/deps/jemalloc/include/jemalloc/internal/tsd_malloc_thread_cleanup.h new file mode 100644 index 000000000..65852d5c1 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/tsd_malloc_thread_cleanup.h @@ -0,0 +1,61 @@ +#ifdef JEMALLOC_INTERNAL_TSD_MALLOC_THREAD_CLEANUP_H +#error This file should be included only once, by tsd.h. +#endif +#define JEMALLOC_INTERNAL_TSD_MALLOC_THREAD_CLEANUP_H + +#define JEMALLOC_TSD_TYPE_ATTR(type) __thread type JEMALLOC_TLS_MODEL + +extern JEMALLOC_TSD_TYPE_ATTR(tsd_t) tsd_tls; +extern JEMALLOC_TSD_TYPE_ATTR(bool) tsd_initialized; +extern bool tsd_booted; + +/* Initialization/cleanup. */ +JEMALLOC_ALWAYS_INLINE bool +tsd_cleanup_wrapper(void) { + if (tsd_initialized) { + tsd_initialized = false; + tsd_cleanup(&tsd_tls); + } + return tsd_initialized; +} + +JEMALLOC_ALWAYS_INLINE bool +tsd_boot0(void) { + malloc_tsd_cleanup_register(&tsd_cleanup_wrapper); + tsd_booted = true; + return false; +} + +JEMALLOC_ALWAYS_INLINE void +tsd_boot1(void) { + /* Do nothing. */ +} + +JEMALLOC_ALWAYS_INLINE bool +tsd_boot(void) { + return tsd_boot0(); +} + +JEMALLOC_ALWAYS_INLINE bool +tsd_booted_get(void) { + return tsd_booted; +} + +JEMALLOC_ALWAYS_INLINE bool +tsd_get_allocates(void) { + return false; +} + +/* Get/set. */ +JEMALLOC_ALWAYS_INLINE tsd_t * +tsd_get(bool init) { + return &tsd_tls; +} +JEMALLOC_ALWAYS_INLINE void +tsd_set(tsd_t *val) { + assert(tsd_booted); + if (likely(&tsd_tls != val)) { + tsd_tls = (*val); + } + tsd_initialized = true; +} diff --git a/deps/jemalloc/include/jemalloc/internal/tsd_tls.h b/deps/jemalloc/include/jemalloc/internal/tsd_tls.h new file mode 100644 index 000000000..7d6c805be --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/tsd_tls.h @@ -0,0 +1,60 @@ +#ifdef JEMALLOC_INTERNAL_TSD_TLS_H +#error This file should be included only once, by tsd.h. +#endif +#define JEMALLOC_INTERNAL_TSD_TLS_H + +#define JEMALLOC_TSD_TYPE_ATTR(type) __thread type JEMALLOC_TLS_MODEL + +extern JEMALLOC_TSD_TYPE_ATTR(tsd_t) tsd_tls; +extern pthread_key_t tsd_tsd; +extern bool tsd_booted; + +/* Initialization/cleanup. */ +JEMALLOC_ALWAYS_INLINE bool +tsd_boot0(void) { + if (pthread_key_create(&tsd_tsd, &tsd_cleanup) != 0) { + return true; + } + tsd_booted = true; + return false; +} + +JEMALLOC_ALWAYS_INLINE void +tsd_boot1(void) { + /* Do nothing. */ +} + +JEMALLOC_ALWAYS_INLINE bool +tsd_boot(void) { + return tsd_boot0(); +} + +JEMALLOC_ALWAYS_INLINE bool +tsd_booted_get(void) { + return tsd_booted; +} + +JEMALLOC_ALWAYS_INLINE bool +tsd_get_allocates(void) { + return false; +} + +/* Get/set. */ +JEMALLOC_ALWAYS_INLINE tsd_t * +tsd_get(bool init) { + return &tsd_tls; +} + +JEMALLOC_ALWAYS_INLINE void +tsd_set(tsd_t *val) { + assert(tsd_booted); + if (likely(&tsd_tls != val)) { + tsd_tls = (*val); + } + if (pthread_setspecific(tsd_tsd, (void *)(&tsd_tls)) != 0) { + malloc_write("<jemalloc>: Error setting tsd.\n"); + if (opt_abort) { + abort(); + } + } +} diff --git a/deps/jemalloc/include/jemalloc/internal/tsd_types.h b/deps/jemalloc/include/jemalloc/internal/tsd_types.h new file mode 100644 index 000000000..6200af61f --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/tsd_types.h @@ -0,0 +1,10 @@ +#ifndef JEMALLOC_INTERNAL_TSD_TYPES_H +#define JEMALLOC_INTERNAL_TSD_TYPES_H + +#define MALLOC_TSD_CLEANUPS_MAX 2 + +typedef struct tsd_s tsd_t; +typedef struct tsdn_s tsdn_t; +typedef bool (*malloc_tsd_cleanup_t)(void); + +#endif /* JEMALLOC_INTERNAL_TSD_TYPES_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/tsd_win.h b/deps/jemalloc/include/jemalloc/internal/tsd_win.h new file mode 100644 index 000000000..cf30d18e3 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/tsd_win.h @@ -0,0 +1,139 @@ +#ifdef JEMALLOC_INTERNAL_TSD_WIN_H +#error This file should be included only once, by tsd.h. +#endif +#define JEMALLOC_INTERNAL_TSD_WIN_H + +typedef struct { + bool initialized; + tsd_t val; +} tsd_wrapper_t; + +extern DWORD tsd_tsd; +extern tsd_wrapper_t tsd_boot_wrapper; +extern bool tsd_booted; + +/* Initialization/cleanup. */ +JEMALLOC_ALWAYS_INLINE bool +tsd_cleanup_wrapper(void) { + DWORD error = GetLastError(); + tsd_wrapper_t *wrapper = (tsd_wrapper_t *)TlsGetValue(tsd_tsd); + SetLastError(error); + + if (wrapper == NULL) { + return false; + } + + if (wrapper->initialized) { + wrapper->initialized = false; + tsd_cleanup(&wrapper->val); + if (wrapper->initialized) { + /* Trigger another cleanup round. */ + return true; + } + } + malloc_tsd_dalloc(wrapper); + return false; +} + +JEMALLOC_ALWAYS_INLINE void +tsd_wrapper_set(tsd_wrapper_t *wrapper) { + if (!TlsSetValue(tsd_tsd, (void *)wrapper)) { + malloc_write("<jemalloc>: Error setting TSD\n"); + abort(); + } +} + +JEMALLOC_ALWAYS_INLINE tsd_wrapper_t * +tsd_wrapper_get(bool init) { + DWORD error = GetLastError(); + tsd_wrapper_t *wrapper = (tsd_wrapper_t *) TlsGetValue(tsd_tsd); + SetLastError(error); + + if (init && unlikely(wrapper == NULL)) { + wrapper = (tsd_wrapper_t *) + malloc_tsd_malloc(sizeof(tsd_wrapper_t)); + if (wrapper == NULL) { + malloc_write("<jemalloc>: Error allocating TSD\n"); + abort(); + } else { + wrapper->initialized = false; + /* MSVC is finicky about aggregate initialization. */ + tsd_t tsd_initializer = TSD_INITIALIZER; + wrapper->val = tsd_initializer; + } + tsd_wrapper_set(wrapper); + } + return wrapper; +} + +JEMALLOC_ALWAYS_INLINE bool +tsd_boot0(void) { + tsd_tsd = TlsAlloc(); + if (tsd_tsd == TLS_OUT_OF_INDEXES) { + return true; + } + malloc_tsd_cleanup_register(&tsd_cleanup_wrapper); + tsd_wrapper_set(&tsd_boot_wrapper); + tsd_booted = true; + return false; +} + +JEMALLOC_ALWAYS_INLINE void +tsd_boot1(void) { + tsd_wrapper_t *wrapper; + wrapper = (tsd_wrapper_t *) + malloc_tsd_malloc(sizeof(tsd_wrapper_t)); + if (wrapper == NULL) { + malloc_write("<jemalloc>: Error allocating TSD\n"); + abort(); + } + tsd_boot_wrapper.initialized = false; + tsd_cleanup(&tsd_boot_wrapper.val); + wrapper->initialized = false; + tsd_t initializer = TSD_INITIALIZER; + wrapper->val = initializer; + tsd_wrapper_set(wrapper); +} +JEMALLOC_ALWAYS_INLINE bool +tsd_boot(void) { + if (tsd_boot0()) { + return true; + } + tsd_boot1(); + return false; +} + +JEMALLOC_ALWAYS_INLINE bool +tsd_booted_get(void) { + return tsd_booted; +} + +JEMALLOC_ALWAYS_INLINE bool +tsd_get_allocates(void) { + return true; +} + +/* Get/set. */ +JEMALLOC_ALWAYS_INLINE tsd_t * +tsd_get(bool init) { + tsd_wrapper_t *wrapper; + + assert(tsd_booted); + wrapper = tsd_wrapper_get(init); + if (tsd_get_allocates() && !init && wrapper == NULL) { + return NULL; + } + return &wrapper->val; +} + +JEMALLOC_ALWAYS_INLINE void +tsd_set(tsd_t *val) { + tsd_wrapper_t *wrapper; + + assert(tsd_booted); + wrapper = tsd_wrapper_get(true); + if (likely(&wrapper->val != val)) { + wrapper->val = *(val); + } + wrapper->initialized = true; +} diff --git a/deps/jemalloc/include/jemalloc/internal/util.h b/deps/jemalloc/include/jemalloc/internal/util.h new file mode 100644 index 000000000..304cb545a --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/util.h @@ -0,0 +1,67 @@ +#ifndef JEMALLOC_INTERNAL_UTIL_H +#define JEMALLOC_INTERNAL_UTIL_H + +#define UTIL_INLINE static inline + +/* Junk fill patterns. */ +#ifndef JEMALLOC_ALLOC_JUNK +# define JEMALLOC_ALLOC_JUNK ((uint8_t)0xa5) +#endif +#ifndef JEMALLOC_FREE_JUNK +# define JEMALLOC_FREE_JUNK ((uint8_t)0x5a) +#endif + +/* + * Wrap a cpp argument that contains commas such that it isn't broken up into + * multiple arguments. + */ +#define JEMALLOC_ARG_CONCAT(...) __VA_ARGS__ + +/* cpp macro definition stringification. */ +#define STRINGIFY_HELPER(x) #x +#define STRINGIFY(x) STRINGIFY_HELPER(x) + +/* + * Silence compiler warnings due to uninitialized values. This is used + * wherever the compiler fails to recognize that the variable is never used + * uninitialized. + */ +#define JEMALLOC_CC_SILENCE_INIT(v) = v + +#ifdef __GNUC__ +# define likely(x) __builtin_expect(!!(x), 1) +# define unlikely(x) __builtin_expect(!!(x), 0) +#else +# define likely(x) !!(x) +# define unlikely(x) !!(x) +#endif + +#if !defined(JEMALLOC_INTERNAL_UNREACHABLE) +# error JEMALLOC_INTERNAL_UNREACHABLE should have been defined by configure +#endif + +#define unreachable() JEMALLOC_INTERNAL_UNREACHABLE() + +/* Set error code. */ +UTIL_INLINE void +set_errno(int errnum) { +#ifdef _WIN32 + SetLastError(errnum); +#else + errno = errnum; +#endif +} + +/* Get last error code. */ +UTIL_INLINE int +get_errno(void) { +#ifdef _WIN32 + return GetLastError(); +#else + return errno; +#endif +} + +#undef UTIL_INLINE + +#endif /* JEMALLOC_INTERNAL_UTIL_H */ diff --git a/deps/jemalloc/include/jemalloc/internal/witness.h b/deps/jemalloc/include/jemalloc/internal/witness.h new file mode 100644 index 000000000..fff9e98cb --- /dev/null +++ b/deps/jemalloc/include/jemalloc/internal/witness.h @@ -0,0 +1,347 @@ +#ifndef JEMALLOC_INTERNAL_WITNESS_H +#define JEMALLOC_INTERNAL_WITNESS_H + +#include "jemalloc/internal/ql.h" + +/******************************************************************************/ +/* LOCK RANKS */ +/******************************************************************************/ + +/* + * Witnesses with rank WITNESS_RANK_OMIT are completely ignored by the witness + * machinery. + */ + +#define WITNESS_RANK_OMIT 0U + +#define WITNESS_RANK_MIN 1U + +#define WITNESS_RANK_INIT 1U +#define WITNESS_RANK_CTL 1U +#define WITNESS_RANK_TCACHES 2U +#define WITNESS_RANK_ARENAS 3U + +#define WITNESS_RANK_BACKGROUND_THREAD_GLOBAL 4U + +#define WITNESS_RANK_PROF_DUMP 5U +#define WITNESS_RANK_PROF_BT2GCTX 6U +#define WITNESS_RANK_PROF_TDATAS 7U +#define WITNESS_RANK_PROF_TDATA 8U +#define WITNESS_RANK_PROF_LOG 9U +#define WITNESS_RANK_PROF_GCTX 10U +#define WITNESS_RANK_BACKGROUND_THREAD 11U + +/* + * Used as an argument to witness_assert_depth_to_rank() in order to validate + * depth excluding non-core locks with lower ranks. Since the rank argument to + * witness_assert_depth_to_rank() is inclusive rather than exclusive, this + * definition can have the same value as the minimally ranked core lock. + */ +#define WITNESS_RANK_CORE 12U + +#define WITNESS_RANK_DECAY 12U +#define WITNESS_RANK_TCACHE_QL 13U +#define WITNESS_RANK_EXTENT_GROW 14U +#define WITNESS_RANK_EXTENTS 15U +#define WITNESS_RANK_EXTENT_AVAIL 16U + +#define WITNESS_RANK_EXTENT_POOL 17U +#define WITNESS_RANK_RTREE 18U +#define WITNESS_RANK_BASE 19U +#define WITNESS_RANK_ARENA_LARGE 20U +#define WITNESS_RANK_HOOK 21U + +#define WITNESS_RANK_LEAF 0xffffffffU +#define WITNESS_RANK_BIN WITNESS_RANK_LEAF +#define WITNESS_RANK_ARENA_STATS WITNESS_RANK_LEAF +#define WITNESS_RANK_DSS WITNESS_RANK_LEAF +#define WITNESS_RANK_PROF_ACTIVE WITNESS_RANK_LEAF +#define WITNESS_RANK_PROF_ACCUM WITNESS_RANK_LEAF +#define WITNESS_RANK_PROF_DUMP_SEQ WITNESS_RANK_LEAF +#define WITNESS_RANK_PROF_GDUMP WITNESS_RANK_LEAF +#define WITNESS_RANK_PROF_NEXT_THR_UID WITNESS_RANK_LEAF +#define WITNESS_RANK_PROF_THREAD_ACTIVE_INIT WITNESS_RANK_LEAF + +/******************************************************************************/ +/* PER-WITNESS DATA */ +/******************************************************************************/ +#if defined(JEMALLOC_DEBUG) +# define WITNESS_INITIALIZER(name, rank) {name, rank, NULL, NULL, {NULL, NULL}} +#else +# define WITNESS_INITIALIZER(name, rank) +#endif + +typedef struct witness_s witness_t; +typedef unsigned witness_rank_t; +typedef ql_head(witness_t) witness_list_t; +typedef int witness_comp_t (const witness_t *, void *, const witness_t *, + void *); + +struct witness_s { + /* Name, used for printing lock order reversal messages. */ + const char *name; + + /* + * Witness rank, where 0 is lowest and UINT_MAX is highest. Witnesses + * must be acquired in order of increasing rank. + */ + witness_rank_t rank; + + /* + * If two witnesses are of equal rank and they have the samp comp + * function pointer, it is called as a last attempt to differentiate + * between witnesses of equal rank. + */ + witness_comp_t *comp; + + /* Opaque data, passed to comp(). */ + void *opaque; + + /* Linkage for thread's currently owned locks. */ + ql_elm(witness_t) link; +}; + +/******************************************************************************/ +/* PER-THREAD DATA */ +/******************************************************************************/ +typedef struct witness_tsd_s witness_tsd_t; +struct witness_tsd_s { + witness_list_t witnesses; + bool forking; +}; + +#define WITNESS_TSD_INITIALIZER { ql_head_initializer(witnesses), false } +#define WITNESS_TSDN_NULL ((witness_tsdn_t *)0) + +/******************************************************************************/ +/* (PER-THREAD) NULLABILITY HELPERS */ +/******************************************************************************/ +typedef struct witness_tsdn_s witness_tsdn_t; +struct witness_tsdn_s { + witness_tsd_t witness_tsd; +}; + +JEMALLOC_ALWAYS_INLINE witness_tsdn_t * +witness_tsd_tsdn(witness_tsd_t *witness_tsd) { + return (witness_tsdn_t *)witness_tsd; +} + +JEMALLOC_ALWAYS_INLINE bool +witness_tsdn_null(witness_tsdn_t *witness_tsdn) { + return witness_tsdn == NULL; +} + +JEMALLOC_ALWAYS_INLINE witness_tsd_t * +witness_tsdn_tsd(witness_tsdn_t *witness_tsdn) { + assert(!witness_tsdn_null(witness_tsdn)); + return &witness_tsdn->witness_tsd; +} + +/******************************************************************************/ +/* API */ +/******************************************************************************/ +void witness_init(witness_t *witness, const char *name, witness_rank_t rank, + witness_comp_t *comp, void *opaque); + +typedef void (witness_lock_error_t)(const witness_list_t *, const witness_t *); +extern witness_lock_error_t *JET_MUTABLE witness_lock_error; + +typedef void (witness_owner_error_t)(const witness_t *); +extern witness_owner_error_t *JET_MUTABLE witness_owner_error; + +typedef void (witness_not_owner_error_t)(const witness_t *); +extern witness_not_owner_error_t *JET_MUTABLE witness_not_owner_error; + +typedef void (witness_depth_error_t)(const witness_list_t *, + witness_rank_t rank_inclusive, unsigned depth); +extern witness_depth_error_t *JET_MUTABLE witness_depth_error; + +void witnesses_cleanup(witness_tsd_t *witness_tsd); +void witness_prefork(witness_tsd_t *witness_tsd); +void witness_postfork_parent(witness_tsd_t *witness_tsd); +void witness_postfork_child(witness_tsd_t *witness_tsd); + +/* Helper, not intended for direct use. */ +static inline bool +witness_owner(witness_tsd_t *witness_tsd, const witness_t *witness) { + witness_list_t *witnesses; + witness_t *w; + + cassert(config_debug); + + witnesses = &witness_tsd->witnesses; + ql_foreach(w, witnesses, link) { + if (w == witness) { + return true; + } + } + + return false; +} + +static inline void +witness_assert_owner(witness_tsdn_t *witness_tsdn, const witness_t *witness) { + witness_tsd_t *witness_tsd; + + if (!config_debug) { + return; + } + + if (witness_tsdn_null(witness_tsdn)) { + return; + } + witness_tsd = witness_tsdn_tsd(witness_tsdn); + if (witness->rank == WITNESS_RANK_OMIT) { + return; + } + + if (witness_owner(witness_tsd, witness)) { + return; + } + witness_owner_error(witness); +} + +static inline void +witness_assert_not_owner(witness_tsdn_t *witness_tsdn, + const witness_t *witness) { + witness_tsd_t *witness_tsd; + witness_list_t *witnesses; + witness_t *w; + + if (!config_debug) { + return; + } + + if (witness_tsdn_null(witness_tsdn)) { + return; + } + witness_tsd = witness_tsdn_tsd(witness_tsdn); + if (witness->rank == WITNESS_RANK_OMIT) { + return; + } + + witnesses = &witness_tsd->witnesses; + ql_foreach(w, witnesses, link) { + if (w == witness) { + witness_not_owner_error(witness); + } + } +} + +static inline void +witness_assert_depth_to_rank(witness_tsdn_t *witness_tsdn, + witness_rank_t rank_inclusive, unsigned depth) { + witness_tsd_t *witness_tsd; + unsigned d; + witness_list_t *witnesses; + witness_t *w; + + if (!config_debug) { + return; + } + + if (witness_tsdn_null(witness_tsdn)) { + return; + } + witness_tsd = witness_tsdn_tsd(witness_tsdn); + + d = 0; + witnesses = &witness_tsd->witnesses; + w = ql_last(witnesses, link); + if (w != NULL) { + ql_reverse_foreach(w, witnesses, link) { + if (w->rank < rank_inclusive) { + break; + } + d++; + } + } + if (d != depth) { + witness_depth_error(witnesses, rank_inclusive, depth); + } +} + +static inline void +witness_assert_depth(witness_tsdn_t *witness_tsdn, unsigned depth) { + witness_assert_depth_to_rank(witness_tsdn, WITNESS_RANK_MIN, depth); +} + +static inline void +witness_assert_lockless(witness_tsdn_t *witness_tsdn) { + witness_assert_depth(witness_tsdn, 0); +} + +static inline void +witness_lock(witness_tsdn_t *witness_tsdn, witness_t *witness) { + witness_tsd_t *witness_tsd; + witness_list_t *witnesses; + witness_t *w; + + if (!config_debug) { + return; + } + + if (witness_tsdn_null(witness_tsdn)) { + return; + } + witness_tsd = witness_tsdn_tsd(witness_tsdn); + if (witness->rank == WITNESS_RANK_OMIT) { + return; + } + + witness_assert_not_owner(witness_tsdn, witness); + + witnesses = &witness_tsd->witnesses; + w = ql_last(witnesses, link); + if (w == NULL) { + /* No other locks; do nothing. */ + } else if (witness_tsd->forking && w->rank <= witness->rank) { + /* Forking, and relaxed ranking satisfied. */ + } else if (w->rank > witness->rank) { + /* Not forking, rank order reversal. */ + witness_lock_error(witnesses, witness); + } else if (w->rank == witness->rank && (w->comp == NULL || w->comp != + witness->comp || w->comp(w, w->opaque, witness, witness->opaque) > + 0)) { + /* + * Missing/incompatible comparison function, or comparison + * function indicates rank order reversal. + */ + witness_lock_error(witnesses, witness); + } + + ql_elm_new(witness, link); + ql_tail_insert(witnesses, witness, link); +} + +static inline void +witness_unlock(witness_tsdn_t *witness_tsdn, witness_t *witness) { + witness_tsd_t *witness_tsd; + witness_list_t *witnesses; + + if (!config_debug) { + return; + } + + if (witness_tsdn_null(witness_tsdn)) { + return; + } + witness_tsd = witness_tsdn_tsd(witness_tsdn); + if (witness->rank == WITNESS_RANK_OMIT) { + return; + } + + /* + * Check whether owner before removal, rather than relying on + * witness_assert_owner() to abort, so that unit tests can test this + * function's failure mode without causing undefined behavior. + */ + if (witness_owner(witness_tsd, witness)) { + witnesses = &witness_tsd->witnesses; + ql_remove(witnesses, witness, link); + } else { + witness_assert_owner(witness_tsdn, witness); + } +} + +#endif /* JEMALLOC_INTERNAL_WITNESS_H */ diff --git a/deps/jemalloc/include/jemalloc/jemalloc.sh b/deps/jemalloc/include/jemalloc/jemalloc.sh new file mode 100755 index 000000000..b19b1548b --- /dev/null +++ b/deps/jemalloc/include/jemalloc/jemalloc.sh @@ -0,0 +1,27 @@ +#!/bin/sh + +objroot=$1 + +cat <<EOF +#ifndef JEMALLOC_H_ +#define JEMALLOC_H_ +#ifdef __cplusplus +extern "C" { +#endif + +EOF + +for hdr in jemalloc_defs.h jemalloc_rename.h jemalloc_macros.h \ + jemalloc_protos.h jemalloc_typedefs.h jemalloc_mangle.h ; do + cat "${objroot}include/jemalloc/${hdr}" \ + | grep -v 'Generated from .* by configure\.' \ + | sed -e 's/ $//g' + echo +done + +cat <<EOF +#ifdef __cplusplus +} +#endif +#endif /* JEMALLOC_H_ */ +EOF diff --git a/deps/jemalloc/include/jemalloc/jemalloc_defs.h.in b/deps/jemalloc/include/jemalloc/jemalloc_defs.h.in new file mode 100644 index 000000000..11c39181b --- /dev/null +++ b/deps/jemalloc/include/jemalloc/jemalloc_defs.h.in @@ -0,0 +1,48 @@ +/* Defined if __attribute__((...)) syntax is supported. */ +#undef JEMALLOC_HAVE_ATTR + +/* Defined if alloc_size attribute is supported. */ +#undef JEMALLOC_HAVE_ATTR_ALLOC_SIZE + +/* Defined if format_arg(...) attribute is supported. */ +#undef JEMALLOC_HAVE_ATTR_FORMAT_ARG + +/* Defined if format(gnu_printf, ...) attribute is supported. */ +#undef JEMALLOC_HAVE_ATTR_FORMAT_GNU_PRINTF + +/* Defined if format(printf, ...) attribute is supported. */ +#undef JEMALLOC_HAVE_ATTR_FORMAT_PRINTF + +/* + * Define overrides for non-standard allocator-related functions if they are + * present on the system. + */ +#undef JEMALLOC_OVERRIDE_MEMALIGN +#undef JEMALLOC_OVERRIDE_VALLOC + +/* + * At least Linux omits the "const" in: + * + * size_t malloc_usable_size(const void *ptr); + * + * Match the operating system's prototype. + */ +#undef JEMALLOC_USABLE_SIZE_CONST + +/* + * If defined, specify throw() for the public function prototypes when compiling + * with C++. The only justification for this is to match the prototypes that + * glibc defines. + */ +#undef JEMALLOC_USE_CXX_THROW + +#ifdef _MSC_VER +# ifdef _WIN64 +# define LG_SIZEOF_PTR_WIN 3 +# else +# define LG_SIZEOF_PTR_WIN 2 +# endif +#endif + +/* sizeof(void *) == 2^LG_SIZEOF_PTR. */ +#undef LG_SIZEOF_PTR diff --git a/deps/jemalloc/include/jemalloc/jemalloc_macros.h.in b/deps/jemalloc/include/jemalloc/jemalloc_macros.h.in new file mode 100644 index 000000000..59e29558c --- /dev/null +++ b/deps/jemalloc/include/jemalloc/jemalloc_macros.h.in @@ -0,0 +1,129 @@ +#include <stdlib.h> +#include <stdbool.h> +#include <stdint.h> +#include <limits.h> +#include <strings.h> + +#define JEMALLOC_VERSION "@jemalloc_version@" +#define JEMALLOC_VERSION_MAJOR @jemalloc_version_major@ +#define JEMALLOC_VERSION_MINOR @jemalloc_version_minor@ +#define JEMALLOC_VERSION_BUGFIX @jemalloc_version_bugfix@ +#define JEMALLOC_VERSION_NREV @jemalloc_version_nrev@ +#define JEMALLOC_VERSION_GID "@jemalloc_version_gid@" +#define JEMALLOC_VERSION_GID_IDENT @jemalloc_version_gid@ + +#define MALLOCX_LG_ALIGN(la) ((int)(la)) +#if LG_SIZEOF_PTR == 2 +# define MALLOCX_ALIGN(a) ((int)(ffs((int)(a))-1)) +#else +# define MALLOCX_ALIGN(a) \ + ((int)(((size_t)(a) < (size_t)INT_MAX) ? ffs((int)(a))-1 : \ + ffs((int)(((size_t)(a))>>32))+31)) +#endif +#define MALLOCX_ZERO ((int)0x40) +/* + * Bias tcache index bits so that 0 encodes "automatic tcache management", and 1 + * encodes MALLOCX_TCACHE_NONE. + */ +#define MALLOCX_TCACHE(tc) ((int)(((tc)+2) << 8)) +#define MALLOCX_TCACHE_NONE MALLOCX_TCACHE(-1) +/* + * Bias arena index bits so that 0 encodes "use an automatically chosen arena". + */ +#define MALLOCX_ARENA(a) ((((int)(a))+1) << 20) + +/* + * Use as arena index in "arena.<i>.{purge,decay,dss}" and + * "stats.arenas.<i>.*" mallctl interfaces to select all arenas. This + * definition is intentionally specified in raw decimal format to support + * cpp-based string concatenation, e.g. + * + * #define STRINGIFY_HELPER(x) #x + * #define STRINGIFY(x) STRINGIFY_HELPER(x) + * + * mallctl("arena." STRINGIFY(MALLCTL_ARENAS_ALL) ".purge", NULL, NULL, NULL, + * 0); + */ +#define MALLCTL_ARENAS_ALL 4096 +/* + * Use as arena index in "stats.arenas.<i>.*" mallctl interfaces to select + * destroyed arenas. + */ +#define MALLCTL_ARENAS_DESTROYED 4097 + +#if defined(__cplusplus) && defined(JEMALLOC_USE_CXX_THROW) +# define JEMALLOC_CXX_THROW throw() +#else +# define JEMALLOC_CXX_THROW +#endif + +#if defined(_MSC_VER) +# define JEMALLOC_ATTR(s) +# define JEMALLOC_ALIGNED(s) __declspec(align(s)) +# define JEMALLOC_ALLOC_SIZE(s) +# define JEMALLOC_ALLOC_SIZE2(s1, s2) +# ifndef JEMALLOC_EXPORT +# ifdef DLLEXPORT +# define JEMALLOC_EXPORT __declspec(dllexport) +# else +# define JEMALLOC_EXPORT __declspec(dllimport) +# endif +# endif +# define JEMALLOC_FORMAT_ARG(i) +# define JEMALLOC_FORMAT_PRINTF(s, i) +# define JEMALLOC_NOINLINE __declspec(noinline) +# ifdef __cplusplus +# define JEMALLOC_NOTHROW __declspec(nothrow) +# else +# define JEMALLOC_NOTHROW +# endif +# define JEMALLOC_SECTION(s) __declspec(allocate(s)) +# define JEMALLOC_RESTRICT_RETURN __declspec(restrict) +# if _MSC_VER >= 1900 && !defined(__EDG__) +# define JEMALLOC_ALLOCATOR __declspec(allocator) +# else +# define JEMALLOC_ALLOCATOR +# endif +#elif defined(JEMALLOC_HAVE_ATTR) +# define JEMALLOC_ATTR(s) __attribute__((s)) +# define JEMALLOC_ALIGNED(s) JEMALLOC_ATTR(aligned(s)) +# ifdef JEMALLOC_HAVE_ATTR_ALLOC_SIZE +# define JEMALLOC_ALLOC_SIZE(s) JEMALLOC_ATTR(alloc_size(s)) +# define JEMALLOC_ALLOC_SIZE2(s1, s2) JEMALLOC_ATTR(alloc_size(s1, s2)) +# else +# define JEMALLOC_ALLOC_SIZE(s) +# define JEMALLOC_ALLOC_SIZE2(s1, s2) +# endif +# ifndef JEMALLOC_EXPORT +# define JEMALLOC_EXPORT JEMALLOC_ATTR(visibility("default")) +# endif +# ifdef JEMALLOC_HAVE_ATTR_FORMAT_ARG +# define JEMALLOC_FORMAT_ARG(i) JEMALLOC_ATTR(__format_arg__(3)) +# else +# define JEMALLOC_FORMAT_ARG(i) +# endif +# ifdef JEMALLOC_HAVE_ATTR_FORMAT_GNU_PRINTF +# define JEMALLOC_FORMAT_PRINTF(s, i) JEMALLOC_ATTR(format(gnu_printf, s, i)) +# elif defined(JEMALLOC_HAVE_ATTR_FORMAT_PRINTF) +# define JEMALLOC_FORMAT_PRINTF(s, i) JEMALLOC_ATTR(format(printf, s, i)) +# else +# define JEMALLOC_FORMAT_PRINTF(s, i) +# endif +# define JEMALLOC_NOINLINE JEMALLOC_ATTR(noinline) +# define JEMALLOC_NOTHROW JEMALLOC_ATTR(nothrow) +# define JEMALLOC_SECTION(s) JEMALLOC_ATTR(section(s)) +# define JEMALLOC_RESTRICT_RETURN +# define JEMALLOC_ALLOCATOR +#else +# define JEMALLOC_ATTR(s) +# define JEMALLOC_ALIGNED(s) +# define JEMALLOC_ALLOC_SIZE(s) +# define JEMALLOC_ALLOC_SIZE2(s1, s2) +# define JEMALLOC_EXPORT +# define JEMALLOC_FORMAT_PRINTF(s, i) +# define JEMALLOC_NOINLINE +# define JEMALLOC_NOTHROW +# define JEMALLOC_SECTION(s) +# define JEMALLOC_RESTRICT_RETURN +# define JEMALLOC_ALLOCATOR +#endif diff --git a/deps/jemalloc/include/jemalloc/jemalloc_mangle.sh b/deps/jemalloc/include/jemalloc/jemalloc_mangle.sh new file mode 100755 index 000000000..c675bb469 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/jemalloc_mangle.sh @@ -0,0 +1,45 @@ +#!/bin/sh -eu + +public_symbols_txt=$1 +symbol_prefix=$2 + +cat <<EOF +/* + * By default application code must explicitly refer to mangled symbol names, + * so that it is possible to use jemalloc in conjunction with another allocator + * in the same application. Define JEMALLOC_MANGLE in order to cause automatic + * name mangling that matches the API prefixing that happened as a result of + * --with-mangling and/or --with-jemalloc-prefix configuration settings. + */ +#ifdef JEMALLOC_MANGLE +# ifndef JEMALLOC_NO_DEMANGLE +# define JEMALLOC_NO_DEMANGLE +# endif +EOF + +for nm in `cat ${public_symbols_txt}` ; do + n=`echo ${nm} |tr ':' ' ' |awk '{print $1}'` + echo "# define ${n} ${symbol_prefix}${n}" +done + +cat <<EOF +#endif + +/* + * The ${symbol_prefix}* macros can be used as stable alternative names for the + * public jemalloc API if JEMALLOC_NO_DEMANGLE is defined. This is primarily + * meant for use in jemalloc itself, but it can be used by application code to + * provide isolation from the name mangling specified via --with-mangling + * and/or --with-jemalloc-prefix. + */ +#ifndef JEMALLOC_NO_DEMANGLE +EOF + +for nm in `cat ${public_symbols_txt}` ; do + n=`echo ${nm} |tr ':' ' ' |awk '{print $1}'` + echo "# undef ${symbol_prefix}${n}" +done + +cat <<EOF +#endif +EOF diff --git a/deps/jemalloc/include/jemalloc/jemalloc_protos.h.in b/deps/jemalloc/include/jemalloc/jemalloc_protos.h.in new file mode 100644 index 000000000..a78414b19 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/jemalloc_protos.h.in @@ -0,0 +1,66 @@ +/* + * The @je_@ prefix on the following public symbol declarations is an artifact + * of namespace management, and should be omitted in application code unless + * JEMALLOC_NO_DEMANGLE is defined (see jemalloc_mangle@install_suffix@.h). + */ +extern JEMALLOC_EXPORT const char *@je_@malloc_conf; +extern JEMALLOC_EXPORT void (*@je_@malloc_message)(void *cbopaque, + const char *s); + +JEMALLOC_EXPORT JEMALLOC_ALLOCATOR JEMALLOC_RESTRICT_RETURN + void JEMALLOC_NOTHROW *@je_@malloc(size_t size) + JEMALLOC_CXX_THROW JEMALLOC_ATTR(malloc) JEMALLOC_ALLOC_SIZE(1); +JEMALLOC_EXPORT JEMALLOC_ALLOCATOR JEMALLOC_RESTRICT_RETURN + void JEMALLOC_NOTHROW *@je_@calloc(size_t num, size_t size) + JEMALLOC_CXX_THROW JEMALLOC_ATTR(malloc) JEMALLOC_ALLOC_SIZE2(1, 2); +JEMALLOC_EXPORT int JEMALLOC_NOTHROW @je_@posix_memalign(void **memptr, + size_t alignment, size_t size) JEMALLOC_CXX_THROW JEMALLOC_ATTR(nonnull(1)); +JEMALLOC_EXPORT JEMALLOC_ALLOCATOR JEMALLOC_RESTRICT_RETURN + void JEMALLOC_NOTHROW *@je_@aligned_alloc(size_t alignment, + size_t size) JEMALLOC_CXX_THROW JEMALLOC_ATTR(malloc) + JEMALLOC_ALLOC_SIZE(2); +JEMALLOC_EXPORT JEMALLOC_ALLOCATOR JEMALLOC_RESTRICT_RETURN + void JEMALLOC_NOTHROW *@je_@realloc(void *ptr, size_t size) + JEMALLOC_CXX_THROW JEMALLOC_ALLOC_SIZE(2); +JEMALLOC_EXPORT void JEMALLOC_NOTHROW @je_@free(void *ptr) + JEMALLOC_CXX_THROW; + +JEMALLOC_EXPORT JEMALLOC_ALLOCATOR JEMALLOC_RESTRICT_RETURN + void JEMALLOC_NOTHROW *@je_@mallocx(size_t size, int flags) + JEMALLOC_ATTR(malloc) JEMALLOC_ALLOC_SIZE(1); +JEMALLOC_EXPORT JEMALLOC_ALLOCATOR JEMALLOC_RESTRICT_RETURN + void JEMALLOC_NOTHROW *@je_@rallocx(void *ptr, size_t size, + int flags) JEMALLOC_ALLOC_SIZE(2); +JEMALLOC_EXPORT size_t JEMALLOC_NOTHROW @je_@xallocx(void *ptr, size_t size, + size_t extra, int flags); +JEMALLOC_EXPORT size_t JEMALLOC_NOTHROW @je_@sallocx(const void *ptr, + int flags) JEMALLOC_ATTR(pure); +JEMALLOC_EXPORT void JEMALLOC_NOTHROW @je_@dallocx(void *ptr, int flags); +JEMALLOC_EXPORT void JEMALLOC_NOTHROW @je_@sdallocx(void *ptr, size_t size, + int flags); +JEMALLOC_EXPORT size_t JEMALLOC_NOTHROW @je_@nallocx(size_t size, int flags) + JEMALLOC_ATTR(pure); + +JEMALLOC_EXPORT int JEMALLOC_NOTHROW @je_@mallctl(const char *name, + void *oldp, size_t *oldlenp, void *newp, size_t newlen); +JEMALLOC_EXPORT int JEMALLOC_NOTHROW @je_@mallctlnametomib(const char *name, + size_t *mibp, size_t *miblenp); +JEMALLOC_EXPORT int JEMALLOC_NOTHROW @je_@mallctlbymib(const size_t *mib, + size_t miblen, void *oldp, size_t *oldlenp, void *newp, size_t newlen); +JEMALLOC_EXPORT void JEMALLOC_NOTHROW @je_@malloc_stats_print( + void (*write_cb)(void *, const char *), void *@je_@cbopaque, + const char *opts); +JEMALLOC_EXPORT size_t JEMALLOC_NOTHROW @je_@malloc_usable_size( + JEMALLOC_USABLE_SIZE_CONST void *ptr) JEMALLOC_CXX_THROW; + +#ifdef JEMALLOC_OVERRIDE_MEMALIGN +JEMALLOC_EXPORT JEMALLOC_ALLOCATOR JEMALLOC_RESTRICT_RETURN + void JEMALLOC_NOTHROW *@je_@memalign(size_t alignment, size_t size) + JEMALLOC_CXX_THROW JEMALLOC_ATTR(malloc); +#endif + +#ifdef JEMALLOC_OVERRIDE_VALLOC +JEMALLOC_EXPORT JEMALLOC_ALLOCATOR JEMALLOC_RESTRICT_RETURN + void JEMALLOC_NOTHROW *@je_@valloc(size_t size) JEMALLOC_CXX_THROW + JEMALLOC_ATTR(malloc); +#endif diff --git a/deps/jemalloc/include/jemalloc/jemalloc_rename.sh b/deps/jemalloc/include/jemalloc/jemalloc_rename.sh new file mode 100755 index 000000000..f94389120 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/jemalloc_rename.sh @@ -0,0 +1,22 @@ +#!/bin/sh + +public_symbols_txt=$1 + +cat <<EOF +/* + * Name mangling for public symbols is controlled by --with-mangling and + * --with-jemalloc-prefix. With default settings the je_ prefix is stripped by + * these macro definitions. + */ +#ifndef JEMALLOC_NO_RENAME +EOF + +for nm in `cat ${public_symbols_txt}` ; do + n=`echo ${nm} |tr ':' ' ' |awk '{print $1}'` + m=`echo ${nm} |tr ':' ' ' |awk '{print $2}'` + echo "# define je_${n} ${m}" +done + +cat <<EOF +#endif +EOF diff --git a/deps/jemalloc/include/jemalloc/jemalloc_typedefs.h.in b/deps/jemalloc/include/jemalloc/jemalloc_typedefs.h.in new file mode 100644 index 000000000..1a5887430 --- /dev/null +++ b/deps/jemalloc/include/jemalloc/jemalloc_typedefs.h.in @@ -0,0 +1,77 @@ +typedef struct extent_hooks_s extent_hooks_t; + +/* + * void * + * extent_alloc(extent_hooks_t *extent_hooks, void *new_addr, size_t size, + * size_t alignment, bool *zero, bool *commit, unsigned arena_ind); + */ +typedef void *(extent_alloc_t)(extent_hooks_t *, void *, size_t, size_t, bool *, + bool *, unsigned); + +/* + * bool + * extent_dalloc(extent_hooks_t *extent_hooks, void *addr, size_t size, + * bool committed, unsigned arena_ind); + */ +typedef bool (extent_dalloc_t)(extent_hooks_t *, void *, size_t, bool, + unsigned); + +/* + * void + * extent_destroy(extent_hooks_t *extent_hooks, void *addr, size_t size, + * bool committed, unsigned arena_ind); + */ +typedef void (extent_destroy_t)(extent_hooks_t *, void *, size_t, bool, + unsigned); + +/* + * bool + * extent_commit(extent_hooks_t *extent_hooks, void *addr, size_t size, + * size_t offset, size_t length, unsigned arena_ind); + */ +typedef bool (extent_commit_t)(extent_hooks_t *, void *, size_t, size_t, size_t, + unsigned); + +/* + * bool + * extent_decommit(extent_hooks_t *extent_hooks, void *addr, size_t size, + * size_t offset, size_t length, unsigned arena_ind); + */ +typedef bool (extent_decommit_t)(extent_hooks_t *, void *, size_t, size_t, + size_t, unsigned); + +/* + * bool + * extent_purge(extent_hooks_t *extent_hooks, void *addr, size_t size, + * size_t offset, size_t length, unsigned arena_ind); + */ +typedef bool (extent_purge_t)(extent_hooks_t *, void *, size_t, size_t, size_t, + unsigned); + +/* + * bool + * extent_split(extent_hooks_t *extent_hooks, void *addr, size_t size, + * size_t size_a, size_t size_b, bool committed, unsigned arena_ind); + */ +typedef bool (extent_split_t)(extent_hooks_t *, void *, size_t, size_t, size_t, + bool, unsigned); + +/* + * bool + * extent_merge(extent_hooks_t *extent_hooks, void *addr_a, size_t size_a, + * void *addr_b, size_t size_b, bool committed, unsigned arena_ind); + */ +typedef bool (extent_merge_t)(extent_hooks_t *, void *, size_t, void *, size_t, + bool, unsigned); + +struct extent_hooks_s { + extent_alloc_t *alloc; + extent_dalloc_t *dalloc; + extent_destroy_t *destroy; + extent_commit_t *commit; + extent_decommit_t *decommit; + extent_purge_t *purge_lazy; + extent_purge_t *purge_forced; + extent_split_t *split; + extent_merge_t *merge; +}; diff --git a/deps/jemalloc/include/msvc_compat/C99/stdbool.h b/deps/jemalloc/include/msvc_compat/C99/stdbool.h new file mode 100644 index 000000000..d92160ebc --- /dev/null +++ b/deps/jemalloc/include/msvc_compat/C99/stdbool.h @@ -0,0 +1,20 @@ +#ifndef stdbool_h +#define stdbool_h + +#include <wtypes.h> + +/* MSVC doesn't define _Bool or bool in C, but does have BOOL */ +/* Note this doesn't pass autoconf's test because (bool) 0.5 != true */ +/* Clang-cl uses MSVC headers, so needs msvc_compat, but has _Bool as + * a built-in type. */ +#ifndef __clang__ +typedef BOOL _Bool; +#endif + +#define bool _Bool +#define true 1 +#define false 0 + +#define __bool_true_false_are_defined 1 + +#endif /* stdbool_h */ diff --git a/deps/jemalloc/include/msvc_compat/C99/stdint.h b/deps/jemalloc/include/msvc_compat/C99/stdint.h new file mode 100644 index 000000000..d02608a59 --- /dev/null +++ b/deps/jemalloc/include/msvc_compat/C99/stdint.h @@ -0,0 +1,247 @@ +// ISO C9x compliant stdint.h for Microsoft Visual Studio +// Based on ISO/IEC 9899:TC2 Committee draft (May 6, 2005) WG14/N1124 +// +// Copyright (c) 2006-2008 Alexander Chemeris +// +// Redistribution and use in source and binary forms, with or without +// modification, are permitted provided that the following conditions are met: +// +// 1. Redistributions of source code must retain the above copyright notice, +// this list of conditions and the following disclaimer. +// +// 2. Redistributions in binary form must reproduce the above copyright +// notice, this list of conditions and the following disclaimer in the +// documentation and/or other materials provided with the distribution. +// +// 3. The name of the author may be used to endorse or promote products +// derived from this software without specific prior written permission. +// +// THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR IMPLIED +// WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF +// MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO +// EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, +// PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; +// OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, +// WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR +// OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF +// ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. +// +/////////////////////////////////////////////////////////////////////////////// + +#ifndef _MSC_VER // [ +#error "Use this header only with Microsoft Visual C++ compilers!" +#endif // _MSC_VER ] + +#ifndef _MSC_STDINT_H_ // [ +#define _MSC_STDINT_H_ + +#if _MSC_VER > 1000 +#pragma once +#endif + +#include <limits.h> + +// For Visual Studio 6 in C++ mode and for many Visual Studio versions when +// compiling for ARM we should wrap <wchar.h> include with 'extern "C++" {}' +// or compiler give many errors like this: +// error C2733: second C linkage of overloaded function 'wmemchr' not allowed +#ifdef __cplusplus +extern "C" { +#endif +# include <wchar.h> +#ifdef __cplusplus +} +#endif + +// Define _W64 macros to mark types changing their size, like intptr_t. +#ifndef _W64 +# if !defined(__midl) && (defined(_X86_) || defined(_M_IX86)) && _MSC_VER >= 1300 +# define _W64 __w64 +# else +# define _W64 +# endif +#endif + + +// 7.18.1 Integer types + +// 7.18.1.1 Exact-width integer types + +// Visual Studio 6 and Embedded Visual C++ 4 doesn't +// realize that, e.g. char has the same size as __int8 +// so we give up on __intX for them. +#if (_MSC_VER < 1300) + typedef signed char int8_t; + typedef signed short int16_t; + typedef signed int int32_t; + typedef unsigned char uint8_t; + typedef unsigned short uint16_t; + typedef unsigned int uint32_t; +#else + typedef signed __int8 int8_t; + typedef signed __int16 int16_t; + typedef signed __int32 int32_t; + typedef unsigned __int8 uint8_t; + typedef unsigned __int16 uint16_t; + typedef unsigned __int32 uint32_t; +#endif +typedef signed __int64 int64_t; +typedef unsigned __int64 uint64_t; + + +// 7.18.1.2 Minimum-width integer types +typedef int8_t int_least8_t; +typedef int16_t int_least16_t; +typedef int32_t int_least32_t; +typedef int64_t int_least64_t; +typedef uint8_t uint_least8_t; +typedef uint16_t uint_least16_t; +typedef uint32_t uint_least32_t; +typedef uint64_t uint_least64_t; + +// 7.18.1.3 Fastest minimum-width integer types +typedef int8_t int_fast8_t; +typedef int16_t int_fast16_t; +typedef int32_t int_fast32_t; +typedef int64_t int_fast64_t; +typedef uint8_t uint_fast8_t; +typedef uint16_t uint_fast16_t; +typedef uint32_t uint_fast32_t; +typedef uint64_t uint_fast64_t; + +// 7.18.1.4 Integer types capable of holding object pointers +#ifdef _WIN64 // [ + typedef signed __int64 intptr_t; + typedef unsigned __int64 uintptr_t; +#else // _WIN64 ][ + typedef _W64 signed int intptr_t; + typedef _W64 unsigned int uintptr_t; +#endif // _WIN64 ] + +// 7.18.1.5 Greatest-width integer types +typedef int64_t intmax_t; +typedef uint64_t uintmax_t; + + +// 7.18.2 Limits of specified-width integer types + +#if !defined(__cplusplus) || defined(__STDC_LIMIT_MACROS) // [ See footnote 220 at page 257 and footnote 221 at page 259 + +// 7.18.2.1 Limits of exact-width integer types +#define INT8_MIN ((int8_t)_I8_MIN) +#define INT8_MAX _I8_MAX +#define INT16_MIN ((int16_t)_I16_MIN) +#define INT16_MAX _I16_MAX +#define INT32_MIN ((int32_t)_I32_MIN) +#define INT32_MAX _I32_MAX +#define INT64_MIN ((int64_t)_I64_MIN) +#define INT64_MAX _I64_MAX +#define UINT8_MAX _UI8_MAX +#define UINT16_MAX _UI16_MAX +#define UINT32_MAX _UI32_MAX +#define UINT64_MAX _UI64_MAX + +// 7.18.2.2 Limits of minimum-width integer types +#define INT_LEAST8_MIN INT8_MIN +#define INT_LEAST8_MAX INT8_MAX +#define INT_LEAST16_MIN INT16_MIN +#define INT_LEAST16_MAX INT16_MAX +#define INT_LEAST32_MIN INT32_MIN +#define INT_LEAST32_MAX INT32_MAX +#define INT_LEAST64_MIN INT64_MIN +#define INT_LEAST64_MAX INT64_MAX +#define UINT_LEAST8_MAX UINT8_MAX +#define UINT_LEAST16_MAX UINT16_MAX +#define UINT_LEAST32_MAX UINT32_MAX +#define UINT_LEAST64_MAX UINT64_MAX + +// 7.18.2.3 Limits of fastest minimum-width integer types +#define INT_FAST8_MIN INT8_MIN +#define INT_FAST8_MAX INT8_MAX +#define INT_FAST16_MIN INT16_MIN +#define INT_FAST16_MAX INT16_MAX +#define INT_FAST32_MIN INT32_MIN +#define INT_FAST32_MAX INT32_MAX +#define INT_FAST64_MIN INT64_MIN +#define INT_FAST64_MAX INT64_MAX +#define UINT_FAST8_MAX UINT8_MAX +#define UINT_FAST16_MAX UINT16_MAX +#define UINT_FAST32_MAX UINT32_MAX +#define UINT_FAST64_MAX UINT64_MAX + +// 7.18.2.4 Limits of integer types capable of holding object pointers +#ifdef _WIN64 // [ +# define INTPTR_MIN INT64_MIN +# define INTPTR_MAX INT64_MAX +# define UINTPTR_MAX UINT64_MAX +#else // _WIN64 ][ +# define INTPTR_MIN INT32_MIN +# define INTPTR_MAX INT32_MAX +# define UINTPTR_MAX UINT32_MAX +#endif // _WIN64 ] + +// 7.18.2.5 Limits of greatest-width integer types +#define INTMAX_MIN INT64_MIN +#define INTMAX_MAX INT64_MAX +#define UINTMAX_MAX UINT64_MAX + +// 7.18.3 Limits of other integer types + +#ifdef _WIN64 // [ +# define PTRDIFF_MIN _I64_MIN +# define PTRDIFF_MAX _I64_MAX +#else // _WIN64 ][ +# define PTRDIFF_MIN _I32_MIN +# define PTRDIFF_MAX _I32_MAX +#endif // _WIN64 ] + +#define SIG_ATOMIC_MIN INT_MIN +#define SIG_ATOMIC_MAX INT_MAX + +#ifndef SIZE_MAX // [ +# ifdef _WIN64 // [ +# define SIZE_MAX _UI64_MAX +# else // _WIN64 ][ +# define SIZE_MAX _UI32_MAX +# endif // _WIN64 ] +#endif // SIZE_MAX ] + +// WCHAR_MIN and WCHAR_MAX are also defined in <wchar.h> +#ifndef WCHAR_MIN // [ +# define WCHAR_MIN 0 +#endif // WCHAR_MIN ] +#ifndef WCHAR_MAX // [ +# define WCHAR_MAX _UI16_MAX +#endif // WCHAR_MAX ] + +#define WINT_MIN 0 +#define WINT_MAX _UI16_MAX + +#endif // __STDC_LIMIT_MACROS ] + + +// 7.18.4 Limits of other integer types + +#if !defined(__cplusplus) || defined(__STDC_CONSTANT_MACROS) // [ See footnote 224 at page 260 + +// 7.18.4.1 Macros for minimum-width integer constants + +#define INT8_C(val) val##i8 +#define INT16_C(val) val##i16 +#define INT32_C(val) val##i32 +#define INT64_C(val) val##i64 + +#define UINT8_C(val) val##ui8 +#define UINT16_C(val) val##ui16 +#define UINT32_C(val) val##ui32 +#define UINT64_C(val) val##ui64 + +// 7.18.4.2 Macros for greatest-width integer constants +#define INTMAX_C INT64_C +#define UINTMAX_C UINT64_C + +#endif // __STDC_CONSTANT_MACROS ] + + +#endif // _MSC_STDINT_H_ ] diff --git a/deps/jemalloc/include/msvc_compat/strings.h b/deps/jemalloc/include/msvc_compat/strings.h new file mode 100644 index 000000000..996f256ce --- /dev/null +++ b/deps/jemalloc/include/msvc_compat/strings.h @@ -0,0 +1,58 @@ +#ifndef strings_h +#define strings_h + +/* MSVC doesn't define ffs/ffsl. This dummy strings.h header is provided + * for both */ +#ifdef _MSC_VER +# include <intrin.h> +# pragma intrinsic(_BitScanForward) +static __forceinline int ffsl(long x) { + unsigned long i; + + if (_BitScanForward(&i, x)) { + return i + 1; + } + return 0; +} + +static __forceinline int ffs(int x) { + return ffsl(x); +} + +# ifdef _M_X64 +# pragma intrinsic(_BitScanForward64) +# endif + +static __forceinline int ffsll(unsigned __int64 x) { + unsigned long i; +#ifdef _M_X64 + if (_BitScanForward64(&i, x)) { + return i + 1; + } + return 0; +#else +// Fallback for 32-bit build where 64-bit version not available +// assuming little endian + union { + unsigned __int64 ll; + unsigned long l[2]; + } s; + + s.ll = x; + + if (_BitScanForward(&i, s.l[0])) { + return i + 1; + } else if(_BitScanForward(&i, s.l[1])) { + return i + 33; + } + return 0; +#endif +} + +#else +# define ffsll(x) __builtin_ffsll(x) +# define ffsl(x) __builtin_ffsl(x) +# define ffs(x) __builtin_ffs(x) +#endif + +#endif /* strings_h */ diff --git a/deps/jemalloc/include/msvc_compat/windows_extra.h b/deps/jemalloc/include/msvc_compat/windows_extra.h new file mode 100644 index 000000000..a6ebb9306 --- /dev/null +++ b/deps/jemalloc/include/msvc_compat/windows_extra.h @@ -0,0 +1,6 @@ +#ifndef MSVC_COMPAT_WINDOWS_EXTRA_H +#define MSVC_COMPAT_WINDOWS_EXTRA_H + +#include <errno.h> + +#endif /* MSVC_COMPAT_WINDOWS_EXTRA_H */ diff --git a/deps/jemalloc/jemalloc.pc.in b/deps/jemalloc/jemalloc.pc.in new file mode 100644 index 000000000..c428a86dc --- /dev/null +++ b/deps/jemalloc/jemalloc.pc.in @@ -0,0 +1,12 @@ +prefix=@prefix@ +exec_prefix=@exec_prefix@ +libdir=@libdir@ +includedir=@includedir@ +install_suffix=@install_suffix@ + +Name: jemalloc +Description: A general purpose malloc(3) implementation that emphasizes fragmentation avoidance and scalable concurrency support. +URL: http://jemalloc.net/ +Version: @jemalloc_version_major@.@jemalloc_version_minor@.@jemalloc_version_bugfix@_@jemalloc_version_nrev@ +Cflags: -I${includedir} +Libs: -L${libdir} -ljemalloc${install_suffix} diff --git a/deps/jemalloc/m4/ax_cxx_compile_stdcxx.m4 b/deps/jemalloc/m4/ax_cxx_compile_stdcxx.m4 new file mode 100644 index 000000000..2c18e49c5 --- /dev/null +++ b/deps/jemalloc/m4/ax_cxx_compile_stdcxx.m4 @@ -0,0 +1,562 @@ +# =========================================================================== +# http://www.gnu.org/software/autoconf-archive/ax_cxx_compile_stdcxx.html +# =========================================================================== +# +# SYNOPSIS +# +# AX_CXX_COMPILE_STDCXX(VERSION, [ext|noext], [mandatory|optional]) +# +# DESCRIPTION +# +# Check for baseline language coverage in the compiler for the specified +# version of the C++ standard. If necessary, add switches to CXX and +# CXXCPP to enable support. VERSION may be '11' (for the C++11 standard) +# or '14' (for the C++14 standard). +# +# The second argument, if specified, indicates whether you insist on an +# extended mode (e.g. -std=gnu++11) or a strict conformance mode (e.g. +# -std=c++11). If neither is specified, you get whatever works, with +# preference for an extended mode. +# +# The third argument, if specified 'mandatory' or if left unspecified, +# indicates that baseline support for the specified C++ standard is +# required and that the macro should error out if no mode with that +# support is found. If specified 'optional', then configuration proceeds +# regardless, after defining HAVE_CXX${VERSION} if and only if a +# supporting mode is found. +# +# LICENSE +# +# Copyright (c) 2008 Benjamin Kosnik <bkoz@redhat.com> +# Copyright (c) 2012 Zack Weinberg <zackw@panix.com> +# Copyright (c) 2013 Roy Stogner <roystgnr@ices.utexas.edu> +# Copyright (c) 2014, 2015 Google Inc.; contributed by Alexey Sokolov <sokolov@google.com> +# Copyright (c) 2015 Paul Norman <penorman@mac.com> +# Copyright (c) 2015 Moritz Klammler <moritz@klammler.eu> +# +# Copying and distribution of this file, with or without modification, are +# permitted in any medium without royalty provided the copyright notice +# and this notice are preserved. This file is offered as-is, without any +# warranty. + +#serial 4 + +dnl This macro is based on the code from the AX_CXX_COMPILE_STDCXX_11 macro +dnl (serial version number 13). + +AC_DEFUN([AX_CXX_COMPILE_STDCXX], [dnl + m4_if([$1], [11], [], + [$1], [14], [], + [$1], [17], [m4_fatal([support for C++17 not yet implemented in AX_CXX_COMPILE_STDCXX])], + [m4_fatal([invalid first argument `$1' to AX_CXX_COMPILE_STDCXX])])dnl + m4_if([$2], [], [], + [$2], [ext], [], + [$2], [noext], [], + [m4_fatal([invalid second argument `$2' to AX_CXX_COMPILE_STDCXX])])dnl + m4_if([$3], [], [ax_cxx_compile_cxx$1_required=true], + [$3], [mandatory], [ax_cxx_compile_cxx$1_required=true], + [$3], [optional], [ax_cxx_compile_cxx$1_required=false], + [m4_fatal([invalid third argument `$3' to AX_CXX_COMPILE_STDCXX])]) + AC_LANG_PUSH([C++])dnl + ac_success=no + AC_CACHE_CHECK(whether $CXX supports C++$1 features by default, + ax_cv_cxx_compile_cxx$1, + [AC_COMPILE_IFELSE([AC_LANG_SOURCE([_AX_CXX_COMPILE_STDCXX_testbody_$1])], + [ax_cv_cxx_compile_cxx$1=yes], + [ax_cv_cxx_compile_cxx$1=no])]) + if test x$ax_cv_cxx_compile_cxx$1 = xyes; then + ac_success=yes + fi + + m4_if([$2], [noext], [], [dnl + if test x$ac_success = xno; then + for switch in -std=gnu++$1 -std=gnu++0x; do + cachevar=AS_TR_SH([ax_cv_cxx_compile_cxx$1_$switch]) + AC_CACHE_CHECK(whether $CXX supports C++$1 features with $switch, + $cachevar, + [ac_save_CXX="$CXX" + CXX="$CXX $switch" + AC_COMPILE_IFELSE([AC_LANG_SOURCE([_AX_CXX_COMPILE_STDCXX_testbody_$1])], + [eval $cachevar=yes], + [eval $cachevar=no]) + CXX="$ac_save_CXX"]) + if eval test x\$$cachevar = xyes; then + CXX="$CXX $switch" + if test -n "$CXXCPP" ; then + CXXCPP="$CXXCPP $switch" + fi + ac_success=yes + break + fi + done + fi]) + + m4_if([$2], [ext], [], [dnl + if test x$ac_success = xno; then + dnl HP's aCC needs +std=c++11 according to: + dnl http://h21007.www2.hp.com/portal/download/files/unprot/aCxx/PDF_Release_Notes/769149-001.pdf + dnl Cray's crayCC needs "-h std=c++11" + for switch in -std=c++$1 -std=c++0x +std=c++$1 "-h std=c++$1"; do + cachevar=AS_TR_SH([ax_cv_cxx_compile_cxx$1_$switch]) + AC_CACHE_CHECK(whether $CXX supports C++$1 features with $switch, + $cachevar, + [ac_save_CXX="$CXX" + CXX="$CXX $switch" + AC_COMPILE_IFELSE([AC_LANG_SOURCE([_AX_CXX_COMPILE_STDCXX_testbody_$1])], + [eval $cachevar=yes], + [eval $cachevar=no]) + CXX="$ac_save_CXX"]) + if eval test x\$$cachevar = xyes; then + CXX="$CXX $switch" + if test -n "$CXXCPP" ; then + CXXCPP="$CXXCPP $switch" + fi + ac_success=yes + break + fi + done + fi]) + AC_LANG_POP([C++]) + if test x$ax_cxx_compile_cxx$1_required = xtrue; then + if test x$ac_success = xno; then + AC_MSG_ERROR([*** A compiler with support for C++$1 language features is required.]) + fi + fi + if test x$ac_success = xno; then + HAVE_CXX$1=0 + AC_MSG_NOTICE([No compiler with C++$1 support was found]) + else + HAVE_CXX$1=1 + AC_DEFINE(HAVE_CXX$1,1, + [define if the compiler supports basic C++$1 syntax]) + fi + AC_SUBST(HAVE_CXX$1) +]) + + +dnl Test body for checking C++11 support + +m4_define([_AX_CXX_COMPILE_STDCXX_testbody_11], + _AX_CXX_COMPILE_STDCXX_testbody_new_in_11 +) + + +dnl Test body for checking C++14 support + +m4_define([_AX_CXX_COMPILE_STDCXX_testbody_14], + _AX_CXX_COMPILE_STDCXX_testbody_new_in_11 + _AX_CXX_COMPILE_STDCXX_testbody_new_in_14 +) + + +dnl Tests for new features in C++11 + +m4_define([_AX_CXX_COMPILE_STDCXX_testbody_new_in_11], [[ + +// If the compiler admits that it is not ready for C++11, why torture it? +// Hopefully, this will speed up the test. + +#ifndef __cplusplus + +#error "This is not a C++ compiler" + +#elif __cplusplus < 201103L + +#error "This is not a C++11 compiler" + +#else + +namespace cxx11 +{ + + namespace test_static_assert + { + + template <typename T> + struct check + { + static_assert(sizeof(int) <= sizeof(T), "not big enough"); + }; + + } + + namespace test_final_override + { + + struct Base + { + virtual void f() {} + }; + + struct Derived : public Base + { + virtual void f() override {} + }; + + } + + namespace test_double_right_angle_brackets + { + + template < typename T > + struct check {}; + + typedef check<void> single_type; + typedef check<check<void>> double_type; + typedef check<check<check<void>>> triple_type; + typedef check<check<check<check<void>>>> quadruple_type; + + } + + namespace test_decltype + { + + int + f() + { + int a = 1; + decltype(a) b = 2; + return a + b; + } + + } + + namespace test_type_deduction + { + + template < typename T1, typename T2 > + struct is_same + { + static const bool value = false; + }; + + template < typename T > + struct is_same<T, T> + { + static const bool value = true; + }; + + template < typename T1, typename T2 > + auto + add(T1 a1, T2 a2) -> decltype(a1 + a2) + { + return a1 + a2; + } + + int + test(const int c, volatile int v) + { + static_assert(is_same<int, decltype(0)>::value == true, ""); + static_assert(is_same<int, decltype(c)>::value == false, ""); + static_assert(is_same<int, decltype(v)>::value == false, ""); + auto ac = c; + auto av = v; + auto sumi = ac + av + 'x'; + auto sumf = ac + av + 1.0; + static_assert(is_same<int, decltype(ac)>::value == true, ""); + static_assert(is_same<int, decltype(av)>::value == true, ""); + static_assert(is_same<int, decltype(sumi)>::value == true, ""); + static_assert(is_same<int, decltype(sumf)>::value == false, ""); + static_assert(is_same<int, decltype(add(c, v))>::value == true, ""); + return (sumf > 0.0) ? sumi : add(c, v); + } + + } + + namespace test_noexcept + { + + int f() { return 0; } + int g() noexcept { return 0; } + + static_assert(noexcept(f()) == false, ""); + static_assert(noexcept(g()) == true, ""); + + } + + namespace test_constexpr + { + + template < typename CharT > + unsigned long constexpr + strlen_c_r(const CharT *const s, const unsigned long acc) noexcept + { + return *s ? strlen_c_r(s + 1, acc + 1) : acc; + } + + template < typename CharT > + unsigned long constexpr + strlen_c(const CharT *const s) noexcept + { + return strlen_c_r(s, 0UL); + } + + static_assert(strlen_c("") == 0UL, ""); + static_assert(strlen_c("1") == 1UL, ""); + static_assert(strlen_c("example") == 7UL, ""); + static_assert(strlen_c("another\0example") == 7UL, ""); + + } + + namespace test_rvalue_references + { + + template < int N > + struct answer + { + static constexpr int value = N; + }; + + answer<1> f(int&) { return answer<1>(); } + answer<2> f(const int&) { return answer<2>(); } + answer<3> f(int&&) { return answer<3>(); } + + void + test() + { + int i = 0; + const int c = 0; + static_assert(decltype(f(i))::value == 1, ""); + static_assert(decltype(f(c))::value == 2, ""); + static_assert(decltype(f(0))::value == 3, ""); + } + + } + + namespace test_uniform_initialization + { + + struct test + { + static const int zero {}; + static const int one {1}; + }; + + static_assert(test::zero == 0, ""); + static_assert(test::one == 1, ""); + + } + + namespace test_lambdas + { + + void + test1() + { + auto lambda1 = [](){}; + auto lambda2 = lambda1; + lambda1(); + lambda2(); + } + + int + test2() + { + auto a = [](int i, int j){ return i + j; }(1, 2); + auto b = []() -> int { return '0'; }(); + auto c = [=](){ return a + b; }(); + auto d = [&](){ return c; }(); + auto e = [a, &b](int x) mutable { + const auto identity = [](int y){ return y; }; + for (auto i = 0; i < a; ++i) + a += b--; + return x + identity(a + b); + }(0); + return a + b + c + d + e; + } + + int + test3() + { + const auto nullary = [](){ return 0; }; + const auto unary = [](int x){ return x; }; + using nullary_t = decltype(nullary); + using unary_t = decltype(unary); + const auto higher1st = [](nullary_t f){ return f(); }; + const auto higher2nd = [unary](nullary_t f1){ + return [unary, f1](unary_t f2){ return f2(unary(f1())); }; + }; + return higher1st(nullary) + higher2nd(nullary)(unary); + } + + } + + namespace test_variadic_templates + { + + template <int...> + struct sum; + + template <int N0, int... N1toN> + struct sum<N0, N1toN...> + { + static constexpr auto value = N0 + sum<N1toN...>::value; + }; + + template <> + struct sum<> + { + static constexpr auto value = 0; + }; + + static_assert(sum<>::value == 0, ""); + static_assert(sum<1>::value == 1, ""); + static_assert(sum<23>::value == 23, ""); + static_assert(sum<1, 2>::value == 3, ""); + static_assert(sum<5, 5, 11>::value == 21, ""); + static_assert(sum<2, 3, 5, 7, 11, 13>::value == 41, ""); + + } + + // http://stackoverflow.com/questions/13728184/template-aliases-and-sfinae + // Clang 3.1 fails with headers of libstd++ 4.8.3 when using std::function + // because of this. + namespace test_template_alias_sfinae + { + + struct foo {}; + + template<typename T> + using member = typename T::member_type; + + template<typename T> + void func(...) {} + + template<typename T> + void func(member<T>*) {} + + void test(); + + void test() { func<foo>(0); } + + } + +} // namespace cxx11 + +#endif // __cplusplus >= 201103L + +]]) + + +dnl Tests for new features in C++14 + +m4_define([_AX_CXX_COMPILE_STDCXX_testbody_new_in_14], [[ + +// If the compiler admits that it is not ready for C++14, why torture it? +// Hopefully, this will speed up the test. + +#ifndef __cplusplus + +#error "This is not a C++ compiler" + +#elif __cplusplus < 201402L + +#error "This is not a C++14 compiler" + +#else + +namespace cxx14 +{ + + namespace test_polymorphic_lambdas + { + + int + test() + { + const auto lambda = [](auto&&... args){ + const auto istiny = [](auto x){ + return (sizeof(x) == 1UL) ? 1 : 0; + }; + const int aretiny[] = { istiny(args)... }; + return aretiny[0]; + }; + return lambda(1, 1L, 1.0f, '1'); + } + + } + + namespace test_binary_literals + { + + constexpr auto ivii = 0b0000000000101010; + static_assert(ivii == 42, "wrong value"); + + } + + namespace test_generalized_constexpr + { + + template < typename CharT > + constexpr unsigned long + strlen_c(const CharT *const s) noexcept + { + auto length = 0UL; + for (auto p = s; *p; ++p) + ++length; + return length; + } + + static_assert(strlen_c("") == 0UL, ""); + static_assert(strlen_c("x") == 1UL, ""); + static_assert(strlen_c("test") == 4UL, ""); + static_assert(strlen_c("another\0test") == 7UL, ""); + + } + + namespace test_lambda_init_capture + { + + int + test() + { + auto x = 0; + const auto lambda1 = [a = x](int b){ return a + b; }; + const auto lambda2 = [a = lambda1(x)](){ return a; }; + return lambda2(); + } + + } + + namespace test_digit_seperators + { + + constexpr auto ten_million = 100'000'000; + static_assert(ten_million == 100000000, ""); + + } + + namespace test_return_type_deduction + { + + auto f(int& x) { return x; } + decltype(auto) g(int& x) { return x; } + + template < typename T1, typename T2 > + struct is_same + { + static constexpr auto value = false; + }; + + template < typename T > + struct is_same<T, T> + { + static constexpr auto value = true; + }; + + int + test() + { + auto x = 0; + static_assert(is_same<int, decltype(f(x))>::value, ""); + static_assert(is_same<int&, decltype(g(x))>::value, ""); + return x; + } + + } + +} // namespace cxx14 + +#endif // __cplusplus >= 201402L + +]]) diff --git a/deps/jemalloc/msvc/ReadMe.txt b/deps/jemalloc/msvc/ReadMe.txt new file mode 100644 index 000000000..633a7d49f --- /dev/null +++ b/deps/jemalloc/msvc/ReadMe.txt @@ -0,0 +1,23 @@ + +How to build jemalloc for Windows +================================= + +1. Install Cygwin with at least the following packages: + * autoconf + * autogen + * gawk + * grep + * sed + +2. Install Visual Studio 2015 or 2017 with Visual C++ + +3. Add Cygwin\bin to the PATH environment variable + +4. Open "x64 Native Tools Command Prompt for VS 2017" + (note: x86/x64 doesn't matter at this point) + +5. Generate header files: + sh -c "CC=cl ./autogen.sh" + +6. Now the project can be opened and built in Visual Studio: + msvc\jemalloc_vc2017.sln diff --git a/deps/jemalloc/msvc/jemalloc_vc2015.sln b/deps/jemalloc/msvc/jemalloc_vc2015.sln new file mode 100644 index 000000000..aedd5e5ea --- /dev/null +++ b/deps/jemalloc/msvc/jemalloc_vc2015.sln @@ -0,0 +1,63 @@ + +Microsoft Visual Studio Solution File, Format Version 12.00 +# Visual Studio 14 +VisualStudioVersion = 14.0.24720.0 +MinimumVisualStudioVersion = 10.0.40219.1 +Project("{2150E333-8FDC-42A3-9474-1A3956D46DE8}") = "Solution Items", "Solution Items", "{70A99006-6DE9-472B-8F83-4CEE6C616DF3}" + ProjectSection(SolutionItems) = preProject + ReadMe.txt = ReadMe.txt + EndProjectSection +EndProject +Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "jemalloc", "projects\vc2015\jemalloc\jemalloc.vcxproj", "{8D6BB292-9E1C-413D-9F98-4864BDC1514A}" +EndProject +Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "test_threads", "projects\vc2015\test_threads\test_threads.vcxproj", "{09028CFD-4EB7-491D-869C-0708DB97ED44}" +EndProject +Global + GlobalSection(SolutionConfigurationPlatforms) = preSolution + Debug|x64 = Debug|x64 + Debug|x86 = Debug|x86 + Debug-static|x64 = Debug-static|x64 + Debug-static|x86 = Debug-static|x86 + Release|x64 = Release|x64 + Release|x86 = Release|x86 + Release-static|x64 = Release-static|x64 + Release-static|x86 = Release-static|x86 + EndGlobalSection + GlobalSection(ProjectConfigurationPlatforms) = postSolution + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Debug|x64.ActiveCfg = Debug|x64 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Debug|x64.Build.0 = Debug|x64 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Debug|x86.ActiveCfg = Debug|Win32 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Debug|x86.Build.0 = Debug|Win32 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Debug-static|x64.ActiveCfg = Debug-static|x64 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Debug-static|x64.Build.0 = Debug-static|x64 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Debug-static|x86.ActiveCfg = Debug-static|Win32 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Debug-static|x86.Build.0 = Debug-static|Win32 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Release|x64.ActiveCfg = Release|x64 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Release|x64.Build.0 = Release|x64 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Release|x86.ActiveCfg = Release|Win32 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Release|x86.Build.0 = Release|Win32 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Release-static|x64.ActiveCfg = Release-static|x64 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Release-static|x64.Build.0 = Release-static|x64 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Release-static|x86.ActiveCfg = Release-static|Win32 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Release-static|x86.Build.0 = Release-static|Win32 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Debug|x64.ActiveCfg = Debug|x64 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Debug|x64.Build.0 = Debug|x64 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Debug|x86.ActiveCfg = Debug|Win32 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Debug|x86.Build.0 = Debug|Win32 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Debug-static|x64.ActiveCfg = Debug-static|x64 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Debug-static|x64.Build.0 = Debug-static|x64 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Debug-static|x86.ActiveCfg = Debug-static|Win32 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Debug-static|x86.Build.0 = Debug-static|Win32 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Release|x64.ActiveCfg = Release|x64 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Release|x64.Build.0 = Release|x64 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Release|x86.ActiveCfg = Release|Win32 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Release|x86.Build.0 = Release|Win32 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Release-static|x64.ActiveCfg = Release-static|x64 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Release-static|x64.Build.0 = Release-static|x64 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Release-static|x86.ActiveCfg = Release-static|Win32 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Release-static|x86.Build.0 = Release-static|Win32 + EndGlobalSection + GlobalSection(SolutionProperties) = preSolution + HideSolutionNode = FALSE + EndGlobalSection +EndGlobal diff --git a/deps/jemalloc/msvc/jemalloc_vc2017.sln b/deps/jemalloc/msvc/jemalloc_vc2017.sln new file mode 100644 index 000000000..c22fcb437 --- /dev/null +++ b/deps/jemalloc/msvc/jemalloc_vc2017.sln @@ -0,0 +1,63 @@ + +Microsoft Visual Studio Solution File, Format Version 12.00 +# Visual Studio 14 +VisualStudioVersion = 14.0.24720.0 +MinimumVisualStudioVersion = 10.0.40219.1 +Project("{2150E333-8FDC-42A3-9474-1A3956D46DE8}") = "Solution Items", "Solution Items", "{70A99006-6DE9-472B-8F83-4CEE6C616DF3}" + ProjectSection(SolutionItems) = preProject + ReadMe.txt = ReadMe.txt + EndProjectSection +EndProject +Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "jemalloc", "projects\vc2017\jemalloc\jemalloc.vcxproj", "{8D6BB292-9E1C-413D-9F98-4864BDC1514A}" +EndProject +Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "test_threads", "projects\vc2017\test_threads\test_threads.vcxproj", "{09028CFD-4EB7-491D-869C-0708DB97ED44}" +EndProject +Global + GlobalSection(SolutionConfigurationPlatforms) = preSolution + Debug|x64 = Debug|x64 + Debug|x86 = Debug|x86 + Debug-static|x64 = Debug-static|x64 + Debug-static|x86 = Debug-static|x86 + Release|x64 = Release|x64 + Release|x86 = Release|x86 + Release-static|x64 = Release-static|x64 + Release-static|x86 = Release-static|x86 + EndGlobalSection + GlobalSection(ProjectConfigurationPlatforms) = postSolution + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Debug|x64.ActiveCfg = Debug|x64 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Debug|x64.Build.0 = Debug|x64 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Debug|x86.ActiveCfg = Debug|Win32 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Debug|x86.Build.0 = Debug|Win32 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Debug-static|x64.ActiveCfg = Debug-static|x64 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Debug-static|x64.Build.0 = Debug-static|x64 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Debug-static|x86.ActiveCfg = Debug-static|Win32 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Debug-static|x86.Build.0 = Debug-static|Win32 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Release|x64.ActiveCfg = Release|x64 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Release|x64.Build.0 = Release|x64 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Release|x86.ActiveCfg = Release|Win32 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Release|x86.Build.0 = Release|Win32 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Release-static|x64.ActiveCfg = Release-static|x64 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Release-static|x64.Build.0 = Release-static|x64 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Release-static|x86.ActiveCfg = Release-static|Win32 + {8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Release-static|x86.Build.0 = Release-static|Win32 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Debug|x64.ActiveCfg = Debug|x64 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Debug|x64.Build.0 = Debug|x64 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Debug|x86.ActiveCfg = Debug|Win32 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Debug|x86.Build.0 = Debug|Win32 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Debug-static|x64.ActiveCfg = Debug-static|x64 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Debug-static|x64.Build.0 = Debug-static|x64 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Debug-static|x86.ActiveCfg = Debug-static|Win32 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Debug-static|x86.Build.0 = Debug-static|Win32 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Release|x64.ActiveCfg = Release|x64 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Release|x64.Build.0 = Release|x64 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Release|x86.ActiveCfg = Release|Win32 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Release|x86.Build.0 = Release|Win32 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Release-static|x64.ActiveCfg = Release-static|x64 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Release-static|x64.Build.0 = Release-static|x64 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Release-static|x86.ActiveCfg = Release-static|Win32 + {09028CFD-4EB7-491D-869C-0708DB97ED44}.Release-static|x86.Build.0 = Release-static|Win32 + EndGlobalSection + GlobalSection(SolutionProperties) = preSolution + HideSolutionNode = FALSE + EndGlobalSection +EndGlobal diff --git a/deps/jemalloc/msvc/projects/vc2015/jemalloc/jemalloc.vcxproj b/deps/jemalloc/msvc/projects/vc2015/jemalloc/jemalloc.vcxproj new file mode 100644 index 000000000..228e8be0c --- /dev/null +++ b/deps/jemalloc/msvc/projects/vc2015/jemalloc/jemalloc.vcxproj @@ -0,0 +1,350 @@ +<?xml version="1.0" encoding="utf-8"?> +<Project DefaultTargets="Build" ToolsVersion="14.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003"> + <ItemGroup Label="ProjectConfigurations"> + <ProjectConfiguration Include="Debug-static|Win32"> + <Configuration>Debug-static</Configuration> + <Platform>Win32</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Debug-static|x64"> + <Configuration>Debug-static</Configuration> + <Platform>x64</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Debug|Win32"> + <Configuration>Debug</Configuration> + <Platform>Win32</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Release-static|Win32"> + <Configuration>Release-static</Configuration> + <Platform>Win32</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Release-static|x64"> + <Configuration>Release-static</Configuration> + <Platform>x64</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Release|Win32"> + <Configuration>Release</Configuration> + <Platform>Win32</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Debug|x64"> + <Configuration>Debug</Configuration> + <Platform>x64</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Release|x64"> + <Configuration>Release</Configuration> + <Platform>x64</Platform> + </ProjectConfiguration> + </ItemGroup> + <ItemGroup> + <ClCompile Include="..\..\..\..\src\arena.c" /> + <ClCompile Include="..\..\..\..\src\background_thread.c" /> + <ClCompile Include="..\..\..\..\src\base.c" /> + <ClCompile Include="..\..\..\..\src\bin.c" /> + <ClCompile Include="..\..\..\..\src\bitmap.c" /> + <ClCompile Include="..\..\..\..\src\ckh.c" /> + <ClCompile Include="..\..\..\..\src\ctl.c" /> + <ClCompile Include="..\..\..\..\src\div.c" /> + <ClCompile Include="..\..\..\..\src\extent.c" /> + <ClCompile Include="..\..\..\..\src\extent_dss.c" /> + <ClCompile Include="..\..\..\..\src\extent_mmap.c" /> + <ClCompile Include="..\..\..\..\src\hash.c" /> + <ClCompile Include="..\..\..\..\src\hook.c" /> + <ClCompile Include="..\..\..\..\src\jemalloc.c" /> + <ClCompile Include="..\..\..\..\src\large.c" /> + <ClCompile Include="..\..\..\..\src\log.c" /> + <ClCompile Include="..\..\..\..\src\malloc_io.c" /> + <ClCompile Include="..\..\..\..\src\mutex.c" /> + <ClCompile Include="..\..\..\..\src\mutex_pool.c" /> + <ClCompile Include="..\..\..\..\src\nstime.c" /> + <ClCompile Include="..\..\..\..\src\pages.c" /> + <ClCompile Include="..\..\..\..\src\prng.c" /> + <ClCompile Include="..\..\..\..\src\prof.c" /> + <ClCompile Include="..\..\..\..\src\rtree.c" /> + <ClCompile Include="..\..\..\..\src\sc.c" /> + <ClCompile Include="..\..\..\..\src\stats.c" /> + <ClCompile Include="..\..\..\..\src\sz.c" /> + <ClCompile Include="..\..\..\..\src\tcache.c" /> + <ClCompile Include="..\..\..\..\src\ticker.c" /> + <ClCompile Include="..\..\..\..\src\tsd.c" /> + <ClCompile Include="..\..\..\..\src\witness.c" /> + <ClCompile Include="..\..\..\..\src\safety_check.c" /> + </ItemGroup> + <PropertyGroup Label="Globals"> + <ProjectGuid>{8D6BB292-9E1C-413D-9F98-4864BDC1514A}</ProjectGuid> + <Keyword>Win32Proj</Keyword> + <RootNamespace>jemalloc</RootNamespace> + <WindowsTargetPlatformVersion>8.1</WindowsTargetPlatformVersion> + </PropertyGroup> + <Import Project="$(VCTargetsPath)\Microsoft.Cpp.Default.props" /> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'" Label="Configuration"> + <ConfigurationType>DynamicLibrary</ConfigurationType> + <UseDebugLibraries>true</UseDebugLibraries> + <PlatformToolset>v140</PlatformToolset> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|Win32'" Label="Configuration"> + <ConfigurationType>StaticLibrary</ConfigurationType> + <UseDebugLibraries>true</UseDebugLibraries> + <PlatformToolset>v140</PlatformToolset> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'" Label="Configuration"> + <ConfigurationType>DynamicLibrary</ConfigurationType> + <UseDebugLibraries>false</UseDebugLibraries> + <PlatformToolset>v140</PlatformToolset> + <WholeProgramOptimization>true</WholeProgramOptimization> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|Win32'" Label="Configuration"> + <ConfigurationType>StaticLibrary</ConfigurationType> + <UseDebugLibraries>false</UseDebugLibraries> + <PlatformToolset>v140</PlatformToolset> + <WholeProgramOptimization>true</WholeProgramOptimization> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'" Label="Configuration"> + <ConfigurationType>DynamicLibrary</ConfigurationType> + <UseDebugLibraries>true</UseDebugLibraries> + <PlatformToolset>v140</PlatformToolset> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|x64'" Label="Configuration"> + <ConfigurationType>StaticLibrary</ConfigurationType> + <UseDebugLibraries>true</UseDebugLibraries> + <PlatformToolset>v140</PlatformToolset> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'" Label="Configuration"> + <ConfigurationType>DynamicLibrary</ConfigurationType> + <UseDebugLibraries>false</UseDebugLibraries> + <PlatformToolset>v140</PlatformToolset> + <WholeProgramOptimization>true</WholeProgramOptimization> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|x64'" Label="Configuration"> + <ConfigurationType>StaticLibrary</ConfigurationType> + <UseDebugLibraries>false</UseDebugLibraries> + <PlatformToolset>v140</PlatformToolset> + <WholeProgramOptimization>true</WholeProgramOptimization> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <Import Project="$(VCTargetsPath)\Microsoft.Cpp.props" /> + <ImportGroup Label="ExtensionSettings"> + </ImportGroup> + <ImportGroup Label="Shared"> + </ImportGroup> + <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|Win32'" Label="PropertySheets"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Release|Win32'"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|Win32'" Label="PropertySheets"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Debug|x64'"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|x64'" Label="PropertySheets"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Release|x64'"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|x64'" Label="PropertySheets"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <PropertyGroup Label="UserMacros" /> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + <TargetName>$(ProjectName)d</TargetName> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|Win32'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + <TargetName>$(ProjectName)-$(PlatformToolset)-$(Configuration)</TargetName> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|Win32'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + <TargetName>$(ProjectName)-$(PlatformToolset)-$(Configuration)</TargetName> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + <TargetName>$(ProjectName)d</TargetName> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|x64'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + <TargetName>$(ProjectName)-vc$(PlatformToolsetVersion)-$(Configuration)</TargetName> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|x64'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + <TargetName>$(ProjectName)-vc$(PlatformToolsetVersion)-$(Configuration)</TargetName> + </PropertyGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'"> + <ClCompile> + <PrecompiledHeader> + </PrecompiledHeader> + <WarningLevel>Level3</WarningLevel> + <Optimization>Disabled</Optimization> + <PreprocessorDefinitions>JEMALLOC_NO_PRIVATE_NAMESPACE;_REENTRANT;_WINDLL;DLLEXPORT;JEMALLOC_DEBUG;_DEBUG;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + <DisableSpecificWarnings>4090;4146;4267;4334</DisableSpecificWarnings> + <ProgramDataBaseFileName>$(OutputPath)$(TargetName).pdb</ProgramDataBaseFileName> + </ClCompile> + <Link> + <SubSystem>Windows</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|Win32'"> + <ClCompile> + <PrecompiledHeader> + </PrecompiledHeader> + <WarningLevel>Level3</WarningLevel> + <Optimization>Disabled</Optimization> + <PreprocessorDefinitions>JEMALLOC_NO_PRIVATE_NAMESPACE;JEMALLOC_DEBUG;_REENTRANT;JEMALLOC_EXPORT=;_DEBUG;_LIB;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + <RuntimeLibrary>MultiThreadedDebug</RuntimeLibrary> + <DisableSpecificWarnings>4090;4146;4267;4334</DisableSpecificWarnings> + <ProgramDataBaseFileName>$(OutputPath)$(TargetName).pdb</ProgramDataBaseFileName> + </ClCompile> + <Link> + <SubSystem>Windows</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'"> + <ClCompile> + <PrecompiledHeader> + </PrecompiledHeader> + <WarningLevel>Level3</WarningLevel> + <Optimization>Disabled</Optimization> + <PreprocessorDefinitions>JEMALLOC_NO_PRIVATE_NAMESPACE;_REENTRANT;_WINDLL;DLLEXPORT;JEMALLOC_DEBUG;_DEBUG;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + <DisableSpecificWarnings>4090;4146;4267;4334</DisableSpecificWarnings> + <ProgramDataBaseFileName>$(OutputPath)$(TargetName).pdb</ProgramDataBaseFileName> + </ClCompile> + <Link> + <SubSystem>Windows</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|x64'"> + <ClCompile> + <PrecompiledHeader> + </PrecompiledHeader> + <WarningLevel>Level3</WarningLevel> + <Optimization>Disabled</Optimization> + <PreprocessorDefinitions>JEMALLOC_NO_PRIVATE_NAMESPACE;JEMALLOC_DEBUG;_REENTRANT;JEMALLOC_EXPORT=;_DEBUG;_LIB;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + <RuntimeLibrary>MultiThreadedDebug</RuntimeLibrary> + <DisableSpecificWarnings>4090;4146;4267;4334</DisableSpecificWarnings> + <DebugInformationFormat>OldStyle</DebugInformationFormat> + <MinimalRebuild>false</MinimalRebuild> + </ClCompile> + <Link> + <SubSystem>Windows</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'"> + <ClCompile> + <WarningLevel>Level3</WarningLevel> + <PrecompiledHeader> + </PrecompiledHeader> + <Optimization>MaxSpeed</Optimization> + <FunctionLevelLinking>true</FunctionLevelLinking> + <IntrinsicFunctions>true</IntrinsicFunctions> + <PreprocessorDefinitions>JEMALLOC_NO_PRIVATE_NAMESPACE;_REENTRANT;_WINDLL;DLLEXPORT;NDEBUG;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + <DisableSpecificWarnings>4090;4146;4267;4334</DisableSpecificWarnings> + <ProgramDataBaseFileName>$(OutputPath)$(TargetName).pdb</ProgramDataBaseFileName> + </ClCompile> + <Link> + <SubSystem>Windows</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + <EnableCOMDATFolding>true</EnableCOMDATFolding> + <OptimizeReferences>true</OptimizeReferences> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|Win32'"> + <ClCompile> + <WarningLevel>Level3</WarningLevel> + <PrecompiledHeader> + </PrecompiledHeader> + <Optimization>MaxSpeed</Optimization> + <FunctionLevelLinking>true</FunctionLevelLinking> + <IntrinsicFunctions>true</IntrinsicFunctions> + <PreprocessorDefinitions>JEMALLOC_NO_PRIVATE_NAMESPACE;_REENTRANT;JEMALLOC_EXPORT=;NDEBUG;_LIB;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + <RuntimeLibrary>MultiThreaded</RuntimeLibrary> + <DisableSpecificWarnings>4090;4146;4267;4334</DisableSpecificWarnings> + <ProgramDataBaseFileName>$(OutputPath)$(TargetName).pdb</ProgramDataBaseFileName> + </ClCompile> + <Link> + <SubSystem>Windows</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + <EnableCOMDATFolding>true</EnableCOMDATFolding> + <OptimizeReferences>true</OptimizeReferences> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'"> + <ClCompile> + <WarningLevel>Level3</WarningLevel> + <PrecompiledHeader> + </PrecompiledHeader> + <Optimization>MaxSpeed</Optimization> + <FunctionLevelLinking>true</FunctionLevelLinking> + <IntrinsicFunctions>true</IntrinsicFunctions> + <AdditionalIncludeDirectories>..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + <PreprocessorDefinitions>JEMALLOC_NO_PRIVATE_NAMESPACE;_REENTRANT;_WINDLL;DLLEXPORT;NDEBUG;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <DisableSpecificWarnings>4090;4146;4267;4334</DisableSpecificWarnings> + <ProgramDataBaseFileName>$(OutputPath)$(TargetName).pdb</ProgramDataBaseFileName> + </ClCompile> + <Link> + <SubSystem>Windows</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + <EnableCOMDATFolding>true</EnableCOMDATFolding> + <OptimizeReferences>true</OptimizeReferences> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|x64'"> + <ClCompile> + <WarningLevel>Level3</WarningLevel> + <PrecompiledHeader> + </PrecompiledHeader> + <Optimization>MaxSpeed</Optimization> + <FunctionLevelLinking>true</FunctionLevelLinking> + <IntrinsicFunctions>true</IntrinsicFunctions> + <PreprocessorDefinitions>JEMALLOC_NO_PRIVATE_NAMESPACE;_REENTRANT;JEMALLOC_EXPORT=;NDEBUG;_LIB;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + <RuntimeLibrary>MultiThreaded</RuntimeLibrary> + <DisableSpecificWarnings>4090;4146;4267;4334</DisableSpecificWarnings> + <DebugInformationFormat>OldStyle</DebugInformationFormat> + </ClCompile> + <Link> + <SubSystem>Windows</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + <EnableCOMDATFolding>true</EnableCOMDATFolding> + <OptimizeReferences>true</OptimizeReferences> + </Link> + </ItemDefinitionGroup> + <Import Project="$(VCTargetsPath)\Microsoft.Cpp.targets" /> + <ImportGroup Label="ExtensionTargets"> + </ImportGroup> +</Project> diff --git a/deps/jemalloc/msvc/projects/vc2015/jemalloc/jemalloc.vcxproj.filters b/deps/jemalloc/msvc/projects/vc2015/jemalloc/jemalloc.vcxproj.filters new file mode 100644 index 000000000..d839515b0 --- /dev/null +++ b/deps/jemalloc/msvc/projects/vc2015/jemalloc/jemalloc.vcxproj.filters @@ -0,0 +1,107 @@ +<?xml version="1.0" encoding="utf-8"?> +<Project ToolsVersion="4.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003"> + <ItemGroup> + <Filter Include="Source Files"> + <UniqueIdentifier>{4FC737F1-C7A5-4376-A066-2A32D752A2FF}</UniqueIdentifier> + <Extensions>cpp;c;cc;cxx;def;odl;idl;hpj;bat;asm;asmx</Extensions> + </Filter> + </ItemGroup> + <ItemGroup> + <ClCompile Include="..\..\..\..\src\arena.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\background_thread.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\base.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\bitmap.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\ckh.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\ctl.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\extent.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\extent_dss.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\extent_mmap.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\hash.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\hook.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\jemalloc.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\large.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\malloc_io.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\mutex.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\mutex_pool.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\nstime.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\pages.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\prng.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\prof.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\rtree.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\sc.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\stats.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\sz.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\tcache.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\ticker.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\tsd.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\witness.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\log.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\bin.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\div.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\safety_check.c"> + <Filter>Source Files</Filter> + </ClCompile> + </ItemGroup> +</Project> diff --git a/deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads.vcxproj b/deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads.vcxproj new file mode 100644 index 000000000..325876d6e --- /dev/null +++ b/deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads.vcxproj @@ -0,0 +1,327 @@ +<?xml version="1.0" encoding="utf-8"?> +<Project DefaultTargets="Build" ToolsVersion="14.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003"> + <ItemGroup Label="ProjectConfigurations"> + <ProjectConfiguration Include="Debug-static|Win32"> + <Configuration>Debug-static</Configuration> + <Platform>Win32</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Debug-static|x64"> + <Configuration>Debug-static</Configuration> + <Platform>x64</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Debug|Win32"> + <Configuration>Debug</Configuration> + <Platform>Win32</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Release-static|Win32"> + <Configuration>Release-static</Configuration> + <Platform>Win32</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Release-static|x64"> + <Configuration>Release-static</Configuration> + <Platform>x64</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Release|Win32"> + <Configuration>Release</Configuration> + <Platform>Win32</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Debug|x64"> + <Configuration>Debug</Configuration> + <Platform>x64</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Release|x64"> + <Configuration>Release</Configuration> + <Platform>x64</Platform> + </ProjectConfiguration> + </ItemGroup> + <PropertyGroup Label="Globals"> + <ProjectGuid>{09028CFD-4EB7-491D-869C-0708DB97ED44}</ProjectGuid> + <Keyword>Win32Proj</Keyword> + <RootNamespace>test_threads</RootNamespace> + <WindowsTargetPlatformVersion>8.1</WindowsTargetPlatformVersion> + </PropertyGroup> + <Import Project="$(VCTargetsPath)\Microsoft.Cpp.Default.props" /> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'" Label="Configuration"> + <ConfigurationType>Application</ConfigurationType> + <UseDebugLibraries>true</UseDebugLibraries> + <PlatformToolset>v140</PlatformToolset> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|Win32'" Label="Configuration"> + <ConfigurationType>Application</ConfigurationType> + <UseDebugLibraries>true</UseDebugLibraries> + <PlatformToolset>v140</PlatformToolset> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'" Label="Configuration"> + <ConfigurationType>Application</ConfigurationType> + <UseDebugLibraries>false</UseDebugLibraries> + <PlatformToolset>v140</PlatformToolset> + <WholeProgramOptimization>true</WholeProgramOptimization> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|Win32'" Label="Configuration"> + <ConfigurationType>Application</ConfigurationType> + <UseDebugLibraries>false</UseDebugLibraries> + <PlatformToolset>v140</PlatformToolset> + <WholeProgramOptimization>true</WholeProgramOptimization> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'" Label="Configuration"> + <ConfigurationType>Application</ConfigurationType> + <UseDebugLibraries>true</UseDebugLibraries> + <PlatformToolset>v140</PlatformToolset> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|x64'" Label="Configuration"> + <ConfigurationType>Application</ConfigurationType> + <UseDebugLibraries>true</UseDebugLibraries> + <PlatformToolset>v140</PlatformToolset> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'" Label="Configuration"> + <ConfigurationType>Application</ConfigurationType> + <UseDebugLibraries>false</UseDebugLibraries> + <PlatformToolset>v140</PlatformToolset> + <WholeProgramOptimization>true</WholeProgramOptimization> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|x64'" Label="Configuration"> + <ConfigurationType>Application</ConfigurationType> + <UseDebugLibraries>false</UseDebugLibraries> + <PlatformToolset>v140</PlatformToolset> + <WholeProgramOptimization>true</WholeProgramOptimization> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <Import Project="$(VCTargetsPath)\Microsoft.Cpp.props" /> + <ImportGroup Label="ExtensionSettings"> + </ImportGroup> + <ImportGroup Label="Shared"> + </ImportGroup> + <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|Win32'" Label="PropertySheets"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Release|Win32'"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|Win32'" Label="PropertySheets"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Debug|x64'"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|x64'" Label="PropertySheets"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Release|x64'"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|x64'" Label="PropertySheets"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <PropertyGroup Label="UserMacros" /> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + <LinkIncremental>true</LinkIncremental> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|Win32'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + <LinkIncremental>true</LinkIncremental> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'"> + <LinkIncremental>true</LinkIncremental> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|x64'"> + <LinkIncremental>true</LinkIncremental> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + <LinkIncremental>false</LinkIncremental> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|Win32'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + <LinkIncremental>false</LinkIncremental> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + <LinkIncremental>false</LinkIncremental> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|x64'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + <LinkIncremental>false</LinkIncremental> + </PropertyGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'"> + <ClCompile> + <PrecompiledHeader> + </PrecompiledHeader> + <WarningLevel>Level3</WarningLevel> + <Optimization>Disabled</Optimization> + <PreprocessorDefinitions>WIN32;_DEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\test\include;..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + </ClCompile> + <Link> + <SubSystem>Console</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + <AdditionalLibraryDirectories>$(SolutionDir)$(Platform)\$(Configuration)</AdditionalLibraryDirectories> + <AdditionalDependencies>jemallocd.lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|Win32'"> + <ClCompile> + <PrecompiledHeader> + </PrecompiledHeader> + <WarningLevel>Level3</WarningLevel> + <Optimization>Disabled</Optimization> + <PreprocessorDefinitions>JEMALLOC_EXPORT=;JEMALLOC_STATIC;_DEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\test\include;..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + <RuntimeLibrary>MultiThreadedDebug</RuntimeLibrary> + </ClCompile> + <Link> + <SubSystem>Console</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + <AdditionalLibraryDirectories>$(SolutionDir)$(Platform)\$(Configuration)</AdditionalLibraryDirectories> + <AdditionalDependencies>jemalloc-$(PlatformToolset)-$(Configuration).lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'"> + <ClCompile> + <PrecompiledHeader> + </PrecompiledHeader> + <WarningLevel>Level3</WarningLevel> + <Optimization>Disabled</Optimization> + <PreprocessorDefinitions>_DEBUG;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\test\include;..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + </ClCompile> + <Link> + <SubSystem>Console</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + <AdditionalDependencies>jemallocd.lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies> + <AdditionalLibraryDirectories>$(SolutionDir)$(Platform)\$(Configuration)</AdditionalLibraryDirectories> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|x64'"> + <ClCompile> + <PrecompiledHeader> + </PrecompiledHeader> + <WarningLevel>Level3</WarningLevel> + <Optimization>Disabled</Optimization> + <PreprocessorDefinitions>JEMALLOC_EXPORT=;JEMALLOC_STATIC;_DEBUG;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\test\include;..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + <RuntimeLibrary>MultiThreadedDebug</RuntimeLibrary> + </ClCompile> + <Link> + <SubSystem>Console</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + <AdditionalDependencies>jemalloc-vc$(PlatformToolsetVersion)-$(Configuration).lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies> + <AdditionalLibraryDirectories>$(SolutionDir)$(Platform)\$(Configuration)</AdditionalLibraryDirectories> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'"> + <ClCompile> + <WarningLevel>Level3</WarningLevel> + <PrecompiledHeader> + </PrecompiledHeader> + <Optimization>MaxSpeed</Optimization> + <FunctionLevelLinking>true</FunctionLevelLinking> + <IntrinsicFunctions>true</IntrinsicFunctions> + <PreprocessorDefinitions>WIN32;NDEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\test\include;..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + </ClCompile> + <Link> + <SubSystem>Console</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + <EnableCOMDATFolding>true</EnableCOMDATFolding> + <OptimizeReferences>true</OptimizeReferences> + <AdditionalLibraryDirectories>$(SolutionDir)$(Platform)\$(Configuration)</AdditionalLibraryDirectories> + <AdditionalDependencies>jemalloc.lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|Win32'"> + <ClCompile> + <WarningLevel>Level3</WarningLevel> + <PrecompiledHeader> + </PrecompiledHeader> + <Optimization>MaxSpeed</Optimization> + <FunctionLevelLinking>true</FunctionLevelLinking> + <IntrinsicFunctions>true</IntrinsicFunctions> + <PreprocessorDefinitions>JEMALLOC_EXPORT=;JEMALLOC_STATIC;NDEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\test\include;..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + <RuntimeLibrary>MultiThreaded</RuntimeLibrary> + </ClCompile> + <Link> + <SubSystem>Console</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + <EnableCOMDATFolding>true</EnableCOMDATFolding> + <OptimizeReferences>true</OptimizeReferences> + <AdditionalLibraryDirectories>$(SolutionDir)$(Platform)\$(Configuration)</AdditionalLibraryDirectories> + <AdditionalDependencies>jemalloc-$(PlatformToolset)-$(Configuration).lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'"> + <ClCompile> + <WarningLevel>Level3</WarningLevel> + <PrecompiledHeader> + </PrecompiledHeader> + <Optimization>MaxSpeed</Optimization> + <FunctionLevelLinking>true</FunctionLevelLinking> + <IntrinsicFunctions>true</IntrinsicFunctions> + <PreprocessorDefinitions>NDEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\test\include;..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + </ClCompile> + <Link> + <SubSystem>Console</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + <EnableCOMDATFolding>true</EnableCOMDATFolding> + <OptimizeReferences>true</OptimizeReferences> + <AdditionalLibraryDirectories>$(SolutionDir)$(Platform)\$(Configuration)</AdditionalLibraryDirectories> + <AdditionalDependencies>jemalloc.lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|x64'"> + <ClCompile> + <WarningLevel>Level3</WarningLevel> + <PrecompiledHeader> + </PrecompiledHeader> + <Optimization>MaxSpeed</Optimization> + <FunctionLevelLinking>true</FunctionLevelLinking> + <IntrinsicFunctions>true</IntrinsicFunctions> + <PreprocessorDefinitions>JEMALLOC_EXPORT=;JEMALLOC_STATIC;NDEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\test\include;..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + <RuntimeLibrary>MultiThreaded</RuntimeLibrary> + </ClCompile> + <Link> + <SubSystem>Console</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + <EnableCOMDATFolding>true</EnableCOMDATFolding> + <OptimizeReferences>true</OptimizeReferences> + <AdditionalLibraryDirectories>$(SolutionDir)$(Platform)\$(Configuration)</AdditionalLibraryDirectories> + <AdditionalDependencies>jemalloc-vc$(PlatformToolsetVersion)-$(Configuration).lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies> + </Link> + </ItemDefinitionGroup> + <ItemGroup> + <ClCompile Include="..\..\..\test_threads\test_threads.cpp" /> + <ClCompile Include="..\..\..\test_threads\test_threads_main.cpp" /> + </ItemGroup> + <ItemGroup> + <ProjectReference Include="..\jemalloc\jemalloc.vcxproj"> + <Project>{8d6bb292-9e1c-413d-9f98-4864bdc1514a}</Project> + </ProjectReference> + </ItemGroup> + <ItemGroup> + <ClInclude Include="..\..\..\test_threads\test_threads.h" /> + </ItemGroup> + <Import Project="$(VCTargetsPath)\Microsoft.Cpp.targets" /> + <ImportGroup Label="ExtensionTargets"> + </ImportGroup> +</Project>
\ No newline at end of file diff --git a/deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads.vcxproj.filters b/deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads.vcxproj.filters new file mode 100644 index 000000000..fa4588fd8 --- /dev/null +++ b/deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads.vcxproj.filters @@ -0,0 +1,26 @@ +<?xml version="1.0" encoding="utf-8"?> +<Project ToolsVersion="4.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003"> + <ItemGroup> + <Filter Include="Source Files"> + <UniqueIdentifier>{4FC737F1-C7A5-4376-A066-2A32D752A2FF}</UniqueIdentifier> + <Extensions>cpp;c;cc;cxx;def;odl;idl;hpj;bat;asm;asmx</Extensions> + </Filter> + <Filter Include="Header Files"> + <UniqueIdentifier>{93995380-89BD-4b04-88EB-625FBE52EBFB}</UniqueIdentifier> + <Extensions>h;hh;hpp;hxx;hm;inl;inc;xsd</Extensions> + </Filter> + </ItemGroup> + <ItemGroup> + <ClCompile Include="..\..\..\test_threads\test_threads.cpp"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\test_threads\test_threads_main.cpp"> + <Filter>Source Files</Filter> + </ClCompile> + </ItemGroup> + <ItemGroup> + <ClInclude Include="..\..\..\test_threads\test_threads.h"> + <Filter>Header Files</Filter> + </ClInclude> + </ItemGroup> +</Project>
\ No newline at end of file diff --git a/deps/jemalloc/msvc/projects/vc2017/jemalloc/jemalloc.vcxproj b/deps/jemalloc/msvc/projects/vc2017/jemalloc/jemalloc.vcxproj new file mode 100644 index 000000000..edcceedec --- /dev/null +++ b/deps/jemalloc/msvc/projects/vc2017/jemalloc/jemalloc.vcxproj @@ -0,0 +1,350 @@ +<?xml version="1.0" encoding="utf-8"?> +<Project DefaultTargets="Build" ToolsVersion="15.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003"> + <ItemGroup Label="ProjectConfigurations"> + <ProjectConfiguration Include="Debug-static|Win32"> + <Configuration>Debug-static</Configuration> + <Platform>Win32</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Debug-static|x64"> + <Configuration>Debug-static</Configuration> + <Platform>x64</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Debug|Win32"> + <Configuration>Debug</Configuration> + <Platform>Win32</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Release-static|Win32"> + <Configuration>Release-static</Configuration> + <Platform>Win32</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Release-static|x64"> + <Configuration>Release-static</Configuration> + <Platform>x64</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Release|Win32"> + <Configuration>Release</Configuration> + <Platform>Win32</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Debug|x64"> + <Configuration>Debug</Configuration> + <Platform>x64</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Release|x64"> + <Configuration>Release</Configuration> + <Platform>x64</Platform> + </ProjectConfiguration> + </ItemGroup> + <ItemGroup> + <ClCompile Include="..\..\..\..\src\arena.c" /> + <ClCompile Include="..\..\..\..\src\background_thread.c" /> + <ClCompile Include="..\..\..\..\src\base.c" /> + <ClCompile Include="..\..\..\..\src\bin.c" /> + <ClCompile Include="..\..\..\..\src\bitmap.c" /> + <ClCompile Include="..\..\..\..\src\ckh.c" /> + <ClCompile Include="..\..\..\..\src\ctl.c" /> + <ClCompile Include="..\..\..\..\src\div.c" /> + <ClCompile Include="..\..\..\..\src\extent.c" /> + <ClCompile Include="..\..\..\..\src\extent_dss.c" /> + <ClCompile Include="..\..\..\..\src\extent_mmap.c" /> + <ClCompile Include="..\..\..\..\src\hash.c" /> + <ClCompile Include="..\..\..\..\src\hook.c" /> + <ClCompile Include="..\..\..\..\src\jemalloc.c" /> + <ClCompile Include="..\..\..\..\src\large.c" /> + <ClCompile Include="..\..\..\..\src\log.c" /> + <ClCompile Include="..\..\..\..\src\malloc_io.c" /> + <ClCompile Include="..\..\..\..\src\mutex.c" /> + <ClCompile Include="..\..\..\..\src\mutex_pool.c" /> + <ClCompile Include="..\..\..\..\src\nstime.c" /> + <ClCompile Include="..\..\..\..\src\pages.c" /> + <ClCompile Include="..\..\..\..\src\prng.c" /> + <ClCompile Include="..\..\..\..\src\prof.c" /> + <ClCompile Include="..\..\..\..\src\rtree.c" /> + <ClCompile Include="..\..\..\..\src\sc.c" /> + <ClCompile Include="..\..\..\..\src\stats.c" /> + <ClCompile Include="..\..\..\..\src\sz.c" /> + <ClCompile Include="..\..\..\..\src\tcache.c" /> + <ClCompile Include="..\..\..\..\src\test_hooks.c" /> + <ClCompile Include="..\..\..\..\src\ticker.c" /> + <ClCompile Include="..\..\..\..\src\tsd.c" /> + <ClCompile Include="..\..\..\..\src\witness.c" /> + <ClCompile Include="..\..\..\..\src\safety_check.c" /> + </ItemGroup> + <PropertyGroup Label="Globals"> + <ProjectGuid>{8D6BB292-9E1C-413D-9F98-4864BDC1514A}</ProjectGuid> + <Keyword>Win32Proj</Keyword> + <RootNamespace>jemalloc</RootNamespace> + </PropertyGroup> + <Import Project="$(VCTargetsPath)\Microsoft.Cpp.Default.props" /> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'" Label="Configuration"> + <ConfigurationType>DynamicLibrary</ConfigurationType> + <UseDebugLibraries>true</UseDebugLibraries> + <PlatformToolset>v141</PlatformToolset> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|Win32'" Label="Configuration"> + <ConfigurationType>StaticLibrary</ConfigurationType> + <UseDebugLibraries>true</UseDebugLibraries> + <PlatformToolset>v141</PlatformToolset> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'" Label="Configuration"> + <ConfigurationType>DynamicLibrary</ConfigurationType> + <UseDebugLibraries>false</UseDebugLibraries> + <PlatformToolset>v141</PlatformToolset> + <WholeProgramOptimization>true</WholeProgramOptimization> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|Win32'" Label="Configuration"> + <ConfigurationType>StaticLibrary</ConfigurationType> + <UseDebugLibraries>false</UseDebugLibraries> + <PlatformToolset>v141</PlatformToolset> + <WholeProgramOptimization>true</WholeProgramOptimization> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'" Label="Configuration"> + <ConfigurationType>DynamicLibrary</ConfigurationType> + <UseDebugLibraries>true</UseDebugLibraries> + <PlatformToolset>v141</PlatformToolset> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|x64'" Label="Configuration"> + <ConfigurationType>StaticLibrary</ConfigurationType> + <UseDebugLibraries>true</UseDebugLibraries> + <PlatformToolset>v141</PlatformToolset> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'" Label="Configuration"> + <ConfigurationType>DynamicLibrary</ConfigurationType> + <UseDebugLibraries>false</UseDebugLibraries> + <PlatformToolset>v141</PlatformToolset> + <WholeProgramOptimization>true</WholeProgramOptimization> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|x64'" Label="Configuration"> + <ConfigurationType>StaticLibrary</ConfigurationType> + <UseDebugLibraries>false</UseDebugLibraries> + <PlatformToolset>v141</PlatformToolset> + <WholeProgramOptimization>true</WholeProgramOptimization> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <Import Project="$(VCTargetsPath)\Microsoft.Cpp.props" /> + <ImportGroup Label="ExtensionSettings"> + </ImportGroup> + <ImportGroup Label="Shared"> + </ImportGroup> + <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|Win32'" Label="PropertySheets"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Release|Win32'"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|Win32'" Label="PropertySheets"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Debug|x64'"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|x64'" Label="PropertySheets"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Release|x64'"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|x64'" Label="PropertySheets"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <PropertyGroup Label="UserMacros" /> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + <TargetName>$(ProjectName)d</TargetName> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|Win32'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + <TargetName>$(ProjectName)-$(PlatformToolset)-$(Configuration)</TargetName> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|Win32'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + <TargetName>$(ProjectName)-$(PlatformToolset)-$(Configuration)</TargetName> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + <TargetName>$(ProjectName)d</TargetName> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|x64'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + <TargetName>$(ProjectName)-vc$(PlatformToolsetVersion)-$(Configuration)</TargetName> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|x64'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + <TargetName>$(ProjectName)-vc$(PlatformToolsetVersion)-$(Configuration)</TargetName> + </PropertyGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'"> + <ClCompile> + <PrecompiledHeader> + </PrecompiledHeader> + <WarningLevel>Level3</WarningLevel> + <Optimization>Disabled</Optimization> + <PreprocessorDefinitions>_REENTRANT;_WINDLL;DLLEXPORT;JEMALLOC_DEBUG;_DEBUG;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + <DisableSpecificWarnings>4090;4146;4267;4334</DisableSpecificWarnings> + <ProgramDataBaseFileName>$(OutputPath)$(TargetName).pdb</ProgramDataBaseFileName> + </ClCompile> + <Link> + <SubSystem>Windows</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|Win32'"> + <ClCompile> + <PrecompiledHeader> + </PrecompiledHeader> + <WarningLevel>Level3</WarningLevel> + <Optimization>Disabled</Optimization> + <PreprocessorDefinitions>JEMALLOC_DEBUG;_REENTRANT;JEMALLOC_EXPORT=;_DEBUG;_LIB;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + <RuntimeLibrary>MultiThreadedDebug</RuntimeLibrary> + <DisableSpecificWarnings>4090;4146;4267;4334</DisableSpecificWarnings> + <ProgramDataBaseFileName>$(OutputPath)$(TargetName).pdb</ProgramDataBaseFileName> + </ClCompile> + <Link> + <SubSystem>Windows</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'"> + <ClCompile> + <PrecompiledHeader> + </PrecompiledHeader> + <WarningLevel>Level3</WarningLevel> + <Optimization>Disabled</Optimization> + <PreprocessorDefinitions>JEMALLOC_NO_PRIVATE_NAMESPACE;_REENTRANT;_WINDLL;DLLEXPORT;JEMALLOC_DEBUG;_DEBUG;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + <DisableSpecificWarnings>4090;4146;4267;4334</DisableSpecificWarnings> + <ProgramDataBaseFileName>$(OutputPath)$(TargetName).pdb</ProgramDataBaseFileName> + </ClCompile> + <Link> + <SubSystem>Windows</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|x64'"> + <ClCompile> + <PrecompiledHeader> + </PrecompiledHeader> + <WarningLevel>Level3</WarningLevel> + <Optimization>Disabled</Optimization> + <PreprocessorDefinitions>JEMALLOC_NO_PRIVATE_NAMESPACE;JEMALLOC_DEBUG;_REENTRANT;JEMALLOC_EXPORT=;_DEBUG;_LIB;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + <RuntimeLibrary>MultiThreadedDebug</RuntimeLibrary> + <DisableSpecificWarnings>4090;4146;4267;4334</DisableSpecificWarnings> + <DebugInformationFormat>OldStyle</DebugInformationFormat> + <MinimalRebuild>false</MinimalRebuild> + </ClCompile> + <Link> + <SubSystem>Windows</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'"> + <ClCompile> + <WarningLevel>Level3</WarningLevel> + <PrecompiledHeader> + </PrecompiledHeader> + <Optimization>MaxSpeed</Optimization> + <FunctionLevelLinking>true</FunctionLevelLinking> + <IntrinsicFunctions>true</IntrinsicFunctions> + <PreprocessorDefinitions>_REENTRANT;_WINDLL;DLLEXPORT;NDEBUG;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + <DisableSpecificWarnings>4090;4146;4267;4334</DisableSpecificWarnings> + <ProgramDataBaseFileName>$(OutputPath)$(TargetName).pdb</ProgramDataBaseFileName> + </ClCompile> + <Link> + <SubSystem>Windows</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + <EnableCOMDATFolding>true</EnableCOMDATFolding> + <OptimizeReferences>true</OptimizeReferences> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|Win32'"> + <ClCompile> + <WarningLevel>Level3</WarningLevel> + <PrecompiledHeader> + </PrecompiledHeader> + <Optimization>MaxSpeed</Optimization> + <FunctionLevelLinking>true</FunctionLevelLinking> + <IntrinsicFunctions>true</IntrinsicFunctions> + <PreprocessorDefinitions>_REENTRANT;JEMALLOC_EXPORT=;NDEBUG;_LIB;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + <RuntimeLibrary>MultiThreaded</RuntimeLibrary> + <DisableSpecificWarnings>4090;4146;4267;4334</DisableSpecificWarnings> + <ProgramDataBaseFileName>$(OutputPath)$(TargetName).pdb</ProgramDataBaseFileName> + </ClCompile> + <Link> + <SubSystem>Windows</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + <EnableCOMDATFolding>true</EnableCOMDATFolding> + <OptimizeReferences>true</OptimizeReferences> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'"> + <ClCompile> + <WarningLevel>Level3</WarningLevel> + <PrecompiledHeader> + </PrecompiledHeader> + <Optimization>MaxSpeed</Optimization> + <FunctionLevelLinking>true</FunctionLevelLinking> + <IntrinsicFunctions>true</IntrinsicFunctions> + <AdditionalIncludeDirectories>..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + <PreprocessorDefinitions>JEMALLOC_NO_PRIVATE_NAMESPACE;_REENTRANT;_WINDLL;DLLEXPORT;NDEBUG;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <DisableSpecificWarnings>4090;4146;4267;4334</DisableSpecificWarnings> + <ProgramDataBaseFileName>$(OutputPath)$(TargetName).pdb</ProgramDataBaseFileName> + </ClCompile> + <Link> + <SubSystem>Windows</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + <EnableCOMDATFolding>true</EnableCOMDATFolding> + <OptimizeReferences>true</OptimizeReferences> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|x64'"> + <ClCompile> + <WarningLevel>Level3</WarningLevel> + <PrecompiledHeader> + </PrecompiledHeader> + <Optimization>MaxSpeed</Optimization> + <FunctionLevelLinking>true</FunctionLevelLinking> + <IntrinsicFunctions>true</IntrinsicFunctions> + <PreprocessorDefinitions>JEMALLOC_NO_PRIVATE_NAMESPACE;_REENTRANT;JEMALLOC_EXPORT=;NDEBUG;_LIB;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + <RuntimeLibrary>MultiThreaded</RuntimeLibrary> + <DisableSpecificWarnings>4090;4146;4267;4334</DisableSpecificWarnings> + <DebugInformationFormat>OldStyle</DebugInformationFormat> + </ClCompile> + <Link> + <SubSystem>Windows</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + <EnableCOMDATFolding>true</EnableCOMDATFolding> + <OptimizeReferences>true</OptimizeReferences> + </Link> + </ItemDefinitionGroup> + <Import Project="$(VCTargetsPath)\Microsoft.Cpp.targets" /> + <ImportGroup Label="ExtensionTargets"> + </ImportGroup> +</Project> diff --git a/deps/jemalloc/msvc/projects/vc2017/jemalloc/jemalloc.vcxproj.filters b/deps/jemalloc/msvc/projects/vc2017/jemalloc/jemalloc.vcxproj.filters new file mode 100644 index 000000000..6df726012 --- /dev/null +++ b/deps/jemalloc/msvc/projects/vc2017/jemalloc/jemalloc.vcxproj.filters @@ -0,0 +1,110 @@ +<?xml version="1.0" encoding="utf-8"?> +<Project ToolsVersion="4.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003"> + <ItemGroup> + <Filter Include="Source Files"> + <UniqueIdentifier>{4FC737F1-C7A5-4376-A066-2A32D752A2FF}</UniqueIdentifier> + <Extensions>cpp;c;cc;cxx;def;odl;idl;hpj;bat;asm;asmx</Extensions> + </Filter> + </ItemGroup> + <ItemGroup> + <ClCompile Include="..\..\..\..\src\arena.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\background_thread.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\base.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\bitmap.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\ckh.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\ctl.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\extent.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\extent_dss.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\extent_mmap.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\hash.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\hook.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\jemalloc.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\large.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\malloc_io.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\mutex.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\mutex_pool.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\nstime.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\pages.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\prng.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\prof.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\rtree.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\sc.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\stats.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\sz.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\tcache.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\ticker.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\tsd.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\witness.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\log.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\bin.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\div.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\test_hooks.c"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\..\src\safety_check.c"> + <Filter>Source Files</Filter> + </ClCompile> + </ItemGroup> +</Project> diff --git a/deps/jemalloc/msvc/projects/vc2017/test_threads/test_threads.vcxproj b/deps/jemalloc/msvc/projects/vc2017/test_threads/test_threads.vcxproj new file mode 100644 index 000000000..c35b0f5aa --- /dev/null +++ b/deps/jemalloc/msvc/projects/vc2017/test_threads/test_threads.vcxproj @@ -0,0 +1,326 @@ +<?xml version="1.0" encoding="utf-8"?> +<Project DefaultTargets="Build" ToolsVersion="15.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003"> + <ItemGroup Label="ProjectConfigurations"> + <ProjectConfiguration Include="Debug-static|Win32"> + <Configuration>Debug-static</Configuration> + <Platform>Win32</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Debug-static|x64"> + <Configuration>Debug-static</Configuration> + <Platform>x64</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Debug|Win32"> + <Configuration>Debug</Configuration> + <Platform>Win32</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Release-static|Win32"> + <Configuration>Release-static</Configuration> + <Platform>Win32</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Release-static|x64"> + <Configuration>Release-static</Configuration> + <Platform>x64</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Release|Win32"> + <Configuration>Release</Configuration> + <Platform>Win32</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Debug|x64"> + <Configuration>Debug</Configuration> + <Platform>x64</Platform> + </ProjectConfiguration> + <ProjectConfiguration Include="Release|x64"> + <Configuration>Release</Configuration> + <Platform>x64</Platform> + </ProjectConfiguration> + </ItemGroup> + <PropertyGroup Label="Globals"> + <ProjectGuid>{09028CFD-4EB7-491D-869C-0708DB97ED44}</ProjectGuid> + <Keyword>Win32Proj</Keyword> + <RootNamespace>test_threads</RootNamespace> + </PropertyGroup> + <Import Project="$(VCTargetsPath)\Microsoft.Cpp.Default.props" /> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'" Label="Configuration"> + <ConfigurationType>Application</ConfigurationType> + <UseDebugLibraries>true</UseDebugLibraries> + <PlatformToolset>v141</PlatformToolset> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|Win32'" Label="Configuration"> + <ConfigurationType>Application</ConfigurationType> + <UseDebugLibraries>true</UseDebugLibraries> + <PlatformToolset>v141</PlatformToolset> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'" Label="Configuration"> + <ConfigurationType>Application</ConfigurationType> + <UseDebugLibraries>false</UseDebugLibraries> + <PlatformToolset>v141</PlatformToolset> + <WholeProgramOptimization>true</WholeProgramOptimization> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|Win32'" Label="Configuration"> + <ConfigurationType>Application</ConfigurationType> + <UseDebugLibraries>false</UseDebugLibraries> + <PlatformToolset>v141</PlatformToolset> + <WholeProgramOptimization>true</WholeProgramOptimization> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'" Label="Configuration"> + <ConfigurationType>Application</ConfigurationType> + <UseDebugLibraries>true</UseDebugLibraries> + <PlatformToolset>v141</PlatformToolset> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|x64'" Label="Configuration"> + <ConfigurationType>Application</ConfigurationType> + <UseDebugLibraries>true</UseDebugLibraries> + <PlatformToolset>v141</PlatformToolset> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'" Label="Configuration"> + <ConfigurationType>Application</ConfigurationType> + <UseDebugLibraries>false</UseDebugLibraries> + <PlatformToolset>v141</PlatformToolset> + <WholeProgramOptimization>true</WholeProgramOptimization> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|x64'" Label="Configuration"> + <ConfigurationType>Application</ConfigurationType> + <UseDebugLibraries>false</UseDebugLibraries> + <PlatformToolset>v141</PlatformToolset> + <WholeProgramOptimization>true</WholeProgramOptimization> + <CharacterSet>MultiByte</CharacterSet> + </PropertyGroup> + <Import Project="$(VCTargetsPath)\Microsoft.Cpp.props" /> + <ImportGroup Label="ExtensionSettings"> + </ImportGroup> + <ImportGroup Label="Shared"> + </ImportGroup> + <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|Win32'" Label="PropertySheets"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Release|Win32'"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|Win32'" Label="PropertySheets"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Debug|x64'"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|x64'" Label="PropertySheets"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Release|x64'"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|x64'" Label="PropertySheets"> + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" /> + </ImportGroup> + <PropertyGroup Label="UserMacros" /> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + <LinkIncremental>true</LinkIncremental> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|Win32'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + <LinkIncremental>true</LinkIncremental> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'"> + <LinkIncremental>true</LinkIncremental> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|x64'"> + <LinkIncremental>true</LinkIncremental> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + <LinkIncremental>false</LinkIncremental> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|Win32'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + <LinkIncremental>false</LinkIncremental> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + <LinkIncremental>false</LinkIncremental> + </PropertyGroup> + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|x64'"> + <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir> + <IntDir>$(Platform)\$(Configuration)\</IntDir> + <LinkIncremental>false</LinkIncremental> + </PropertyGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'"> + <ClCompile> + <PrecompiledHeader> + </PrecompiledHeader> + <WarningLevel>Level3</WarningLevel> + <Optimization>Disabled</Optimization> + <PreprocessorDefinitions>WIN32;_DEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\test\include;..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + </ClCompile> + <Link> + <SubSystem>Console</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + <AdditionalLibraryDirectories>$(SolutionDir)$(Platform)\$(Configuration)</AdditionalLibraryDirectories> + <AdditionalDependencies>jemallocd.lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|Win32'"> + <ClCompile> + <PrecompiledHeader> + </PrecompiledHeader> + <WarningLevel>Level3</WarningLevel> + <Optimization>Disabled</Optimization> + <PreprocessorDefinitions>JEMALLOC_EXPORT=;JEMALLOC_STATIC;_DEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\test\include;..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + <RuntimeLibrary>MultiThreadedDebug</RuntimeLibrary> + </ClCompile> + <Link> + <SubSystem>Console</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + <AdditionalLibraryDirectories>$(SolutionDir)$(Platform)\$(Configuration)</AdditionalLibraryDirectories> + <AdditionalDependencies>jemalloc-$(PlatformToolset)-$(Configuration).lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'"> + <ClCompile> + <PrecompiledHeader> + </PrecompiledHeader> + <WarningLevel>Level3</WarningLevel> + <Optimization>Disabled</Optimization> + <PreprocessorDefinitions>_DEBUG;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\test\include;..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + </ClCompile> + <Link> + <SubSystem>Console</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + <AdditionalDependencies>jemallocd.lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies> + <AdditionalLibraryDirectories>$(SolutionDir)$(Platform)\$(Configuration)</AdditionalLibraryDirectories> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|x64'"> + <ClCompile> + <PrecompiledHeader> + </PrecompiledHeader> + <WarningLevel>Level3</WarningLevel> + <Optimization>Disabled</Optimization> + <PreprocessorDefinitions>JEMALLOC_EXPORT=;JEMALLOC_STATIC;_DEBUG;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\test\include;..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + <RuntimeLibrary>MultiThreadedDebug</RuntimeLibrary> + </ClCompile> + <Link> + <SubSystem>Console</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + <AdditionalDependencies>jemalloc-vc$(PlatformToolsetVersion)-$(Configuration).lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies> + <AdditionalLibraryDirectories>$(SolutionDir)$(Platform)\$(Configuration)</AdditionalLibraryDirectories> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'"> + <ClCompile> + <WarningLevel>Level3</WarningLevel> + <PrecompiledHeader> + </PrecompiledHeader> + <Optimization>MaxSpeed</Optimization> + <FunctionLevelLinking>true</FunctionLevelLinking> + <IntrinsicFunctions>true</IntrinsicFunctions> + <PreprocessorDefinitions>WIN32;NDEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\test\include;..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + </ClCompile> + <Link> + <SubSystem>Console</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + <EnableCOMDATFolding>true</EnableCOMDATFolding> + <OptimizeReferences>true</OptimizeReferences> + <AdditionalLibraryDirectories>$(SolutionDir)$(Platform)\$(Configuration)</AdditionalLibraryDirectories> + <AdditionalDependencies>jemalloc.lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|Win32'"> + <ClCompile> + <WarningLevel>Level3</WarningLevel> + <PrecompiledHeader> + </PrecompiledHeader> + <Optimization>MaxSpeed</Optimization> + <FunctionLevelLinking>true</FunctionLevelLinking> + <IntrinsicFunctions>true</IntrinsicFunctions> + <PreprocessorDefinitions>JEMALLOC_EXPORT=;JEMALLOC_STATIC;NDEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\test\include;..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + <RuntimeLibrary>MultiThreaded</RuntimeLibrary> + </ClCompile> + <Link> + <SubSystem>Console</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + <EnableCOMDATFolding>true</EnableCOMDATFolding> + <OptimizeReferences>true</OptimizeReferences> + <AdditionalLibraryDirectories>$(SolutionDir)$(Platform)\$(Configuration)</AdditionalLibraryDirectories> + <AdditionalDependencies>jemalloc-$(PlatformToolset)-$(Configuration).lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'"> + <ClCompile> + <WarningLevel>Level3</WarningLevel> + <PrecompiledHeader> + </PrecompiledHeader> + <Optimization>MaxSpeed</Optimization> + <FunctionLevelLinking>true</FunctionLevelLinking> + <IntrinsicFunctions>true</IntrinsicFunctions> + <PreprocessorDefinitions>NDEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\test\include;..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + </ClCompile> + <Link> + <SubSystem>Console</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + <EnableCOMDATFolding>true</EnableCOMDATFolding> + <OptimizeReferences>true</OptimizeReferences> + <AdditionalLibraryDirectories>$(SolutionDir)$(Platform)\$(Configuration)</AdditionalLibraryDirectories> + <AdditionalDependencies>jemalloc.lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies> + </Link> + </ItemDefinitionGroup> + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|x64'"> + <ClCompile> + <WarningLevel>Level3</WarningLevel> + <PrecompiledHeader> + </PrecompiledHeader> + <Optimization>MaxSpeed</Optimization> + <FunctionLevelLinking>true</FunctionLevelLinking> + <IntrinsicFunctions>true</IntrinsicFunctions> + <PreprocessorDefinitions>JEMALLOC_EXPORT=;JEMALLOC_STATIC;NDEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions> + <AdditionalIncludeDirectories>..\..\..\..\test\include;..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories> + <RuntimeLibrary>MultiThreaded</RuntimeLibrary> + </ClCompile> + <Link> + <SubSystem>Console</SubSystem> + <GenerateDebugInformation>true</GenerateDebugInformation> + <EnableCOMDATFolding>true</EnableCOMDATFolding> + <OptimizeReferences>true</OptimizeReferences> + <AdditionalLibraryDirectories>$(SolutionDir)$(Platform)\$(Configuration)</AdditionalLibraryDirectories> + <AdditionalDependencies>jemalloc-vc$(PlatformToolsetVersion)-$(Configuration).lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies> + </Link> + </ItemDefinitionGroup> + <ItemGroup> + <ClCompile Include="..\..\..\test_threads\test_threads.cpp" /> + <ClCompile Include="..\..\..\test_threads\test_threads_main.cpp" /> + </ItemGroup> + <ItemGroup> + <ProjectReference Include="..\jemalloc\jemalloc.vcxproj"> + <Project>{8d6bb292-9e1c-413d-9f98-4864bdc1514a}</Project> + </ProjectReference> + </ItemGroup> + <ItemGroup> + <ClInclude Include="..\..\..\test_threads\test_threads.h" /> + </ItemGroup> + <Import Project="$(VCTargetsPath)\Microsoft.Cpp.targets" /> + <ImportGroup Label="ExtensionTargets"> + </ImportGroup> +</Project>
\ No newline at end of file diff --git a/deps/jemalloc/msvc/projects/vc2017/test_threads/test_threads.vcxproj.filters b/deps/jemalloc/msvc/projects/vc2017/test_threads/test_threads.vcxproj.filters new file mode 100644 index 000000000..fa4588fd8 --- /dev/null +++ b/deps/jemalloc/msvc/projects/vc2017/test_threads/test_threads.vcxproj.filters @@ -0,0 +1,26 @@ +<?xml version="1.0" encoding="utf-8"?> +<Project ToolsVersion="4.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003"> + <ItemGroup> + <Filter Include="Source Files"> + <UniqueIdentifier>{4FC737F1-C7A5-4376-A066-2A32D752A2FF}</UniqueIdentifier> + <Extensions>cpp;c;cc;cxx;def;odl;idl;hpj;bat;asm;asmx</Extensions> + </Filter> + <Filter Include="Header Files"> + <UniqueIdentifier>{93995380-89BD-4b04-88EB-625FBE52EBFB}</UniqueIdentifier> + <Extensions>h;hh;hpp;hxx;hm;inl;inc;xsd</Extensions> + </Filter> + </ItemGroup> + <ItemGroup> + <ClCompile Include="..\..\..\test_threads\test_threads.cpp"> + <Filter>Source Files</Filter> + </ClCompile> + <ClCompile Include="..\..\..\test_threads\test_threads_main.cpp"> + <Filter>Source Files</Filter> + </ClCompile> + </ItemGroup> + <ItemGroup> + <ClInclude Include="..\..\..\test_threads\test_threads.h"> + <Filter>Header Files</Filter> + </ClInclude> + </ItemGroup> +</Project>
\ No newline at end of file diff --git a/deps/jemalloc/msvc/test_threads/test_threads.cpp b/deps/jemalloc/msvc/test_threads/test_threads.cpp new file mode 100644 index 000000000..92e316243 --- /dev/null +++ b/deps/jemalloc/msvc/test_threads/test_threads.cpp @@ -0,0 +1,88 @@ +// jemalloc C++ threaded test +// Author: Rustam Abdullaev +// Public Domain + +#include <atomic> +#include <functional> +#include <future> +#include <random> +#include <thread> +#include <vector> +#include <stdio.h> +#include <jemalloc/jemalloc.h> + +using std::vector; +using std::thread; +using std::uniform_int_distribution; +using std::minstd_rand; + +int test_threads() { + je_malloc_conf = "narenas:3"; + int narenas = 0; + size_t sz = sizeof(narenas); + je_mallctl("opt.narenas", (void *)&narenas, &sz, NULL, 0); + if (narenas != 3) { + printf("Error: unexpected number of arenas: %d\n", narenas); + return 1; + } + static const int sizes[] = { 7, 16, 32, 60, 91, 100, 120, 144, 169, 199, 255, 400, 670, 900, 917, 1025, 3333, 5190, 13131, 49192, 99999, 123123, 255265, 2333111 }; + static const int numSizes = (int)(sizeof(sizes) / sizeof(sizes[0])); + vector<thread> workers; + static const int numThreads = narenas + 1, numAllocsMax = 25, numIter1 = 50, numIter2 = 50; + je_malloc_stats_print(NULL, NULL, NULL); + size_t allocated1; + size_t sz1 = sizeof(allocated1); + je_mallctl("stats.active", (void *)&allocated1, &sz1, NULL, 0); + printf("\nPress Enter to start threads...\n"); + getchar(); + printf("Starting %d threads x %d x %d iterations...\n", numThreads, numIter1, numIter2); + for (int i = 0; i < numThreads; i++) { + workers.emplace_back([tid=i]() { + uniform_int_distribution<int> sizeDist(0, numSizes - 1); + minstd_rand rnd(tid * 17); + uint8_t* ptrs[numAllocsMax]; + int ptrsz[numAllocsMax]; + for (int i = 0; i < numIter1; ++i) { + thread t([&]() { + for (int i = 0; i < numIter2; ++i) { + const int numAllocs = numAllocsMax - sizeDist(rnd); + for (int j = 0; j < numAllocs; j += 64) { + const int x = sizeDist(rnd); + const int sz = sizes[x]; + ptrsz[j] = sz; + ptrs[j] = (uint8_t*)je_malloc(sz); + if (!ptrs[j]) { + printf("Unable to allocate %d bytes in thread %d, iter %d, alloc %d. %d\n", sz, tid, i, j, x); + exit(1); + } + for (int k = 0; k < sz; k++) + ptrs[j][k] = tid + k; + } + for (int j = 0; j < numAllocs; j += 64) { + for (int k = 0, sz = ptrsz[j]; k < sz; k++) + if (ptrs[j][k] != (uint8_t)(tid + k)) { + printf("Memory error in thread %d, iter %d, alloc %d @ %d : %02X!=%02X\n", tid, i, j, k, ptrs[j][k], (uint8_t)(tid + k)); + exit(1); + } + je_free(ptrs[j]); + } + } + }); + t.join(); + } + }); + } + for (thread& t : workers) { + t.join(); + } + je_malloc_stats_print(NULL, NULL, NULL); + size_t allocated2; + je_mallctl("stats.active", (void *)&allocated2, &sz1, NULL, 0); + size_t leaked = allocated2 - allocated1; + printf("\nDone. Leaked: %zd bytes\n", leaked); + bool failed = leaked > 65536; // in case C++ runtime allocated something (e.g. iostream locale or facet) + printf("\nTest %s!\n", (failed ? "FAILED" : "successful")); + printf("\nPress Enter to continue...\n"); + getchar(); + return failed ? 1 : 0; +} diff --git a/deps/jemalloc/msvc/test_threads/test_threads.h b/deps/jemalloc/msvc/test_threads/test_threads.h new file mode 100644 index 000000000..64d0cdb33 --- /dev/null +++ b/deps/jemalloc/msvc/test_threads/test_threads.h @@ -0,0 +1,3 @@ +#pragma once + +int test_threads(); diff --git a/deps/jemalloc/msvc/test_threads/test_threads_main.cpp b/deps/jemalloc/msvc/test_threads/test_threads_main.cpp new file mode 100644 index 000000000..0a022fba4 --- /dev/null +++ b/deps/jemalloc/msvc/test_threads/test_threads_main.cpp @@ -0,0 +1,11 @@ +#include "test_threads.h" +#include <future> +#include <functional> +#include <chrono> + +using namespace std::chrono_literals; + +int main(int argc, char** argv) { + int rc = test_threads(); + return rc; +} diff --git a/deps/jemalloc/run_tests.sh b/deps/jemalloc/run_tests.sh new file mode 100755 index 000000000..b434f15b3 --- /dev/null +++ b/deps/jemalloc/run_tests.sh @@ -0,0 +1 @@ +$(dirname "$)")/scripts/gen_run_tests.py | bash diff --git a/deps/jemalloc/scripts/gen_run_tests.py b/deps/jemalloc/scripts/gen_run_tests.py new file mode 100755 index 000000000..a414f812a --- /dev/null +++ b/deps/jemalloc/scripts/gen_run_tests.py @@ -0,0 +1,126 @@ +#!/usr/bin/env python + +import sys +from itertools import combinations +from os import uname +from multiprocessing import cpu_count +from subprocess import call + +# Later, we want to test extended vaddr support. Apparently, the "real" way of +# checking this is flaky on OS X. +bits_64 = sys.maxsize > 2**32 + +nparallel = cpu_count() * 2 + +uname = uname()[0] + +if "BSD" in uname: + make_cmd = 'gmake' +else: + make_cmd = 'make' + +def powerset(items): + result = [] + for i in xrange(len(items) + 1): + result += combinations(items, i) + return result + +possible_compilers = [] +for cc, cxx in (['gcc', 'g++'], ['clang', 'clang++']): + try: + cmd_ret = call([cc, "-v"]) + if cmd_ret == 0: + possible_compilers.append((cc, cxx)) + except: + pass +possible_compiler_opts = [ + '-m32', +] +possible_config_opts = [ + '--enable-debug', + '--enable-prof', + '--disable-stats', + '--enable-opt-safety-checks', +] +if bits_64: + possible_config_opts.append('--with-lg-vaddr=56') + +possible_malloc_conf_opts = [ + 'tcache:false', + 'dss:primary', + 'percpu_arena:percpu', + 'background_thread:true', +] + +print 'set -e' +print 'if [ -f Makefile ] ; then %(make_cmd)s relclean ; fi' % {'make_cmd': make_cmd} +print 'autoconf' +print 'rm -rf run_tests.out' +print 'mkdir run_tests.out' +print 'cd run_tests.out' + +ind = 0 +for cc, cxx in possible_compilers: + for compiler_opts in powerset(possible_compiler_opts): + for config_opts in powerset(possible_config_opts): + for malloc_conf_opts in powerset(possible_malloc_conf_opts): + if cc is 'clang' \ + and '-m32' in possible_compiler_opts \ + and '--enable-prof' in config_opts: + continue + config_line = ( + 'EXTRA_CFLAGS=-Werror EXTRA_CXXFLAGS=-Werror ' + + 'CC="{} {}" '.format(cc, " ".join(compiler_opts)) + + 'CXX="{} {}" '.format(cxx, " ".join(compiler_opts)) + + '../../configure ' + + " ".join(config_opts) + (' --with-malloc-conf=' + + ",".join(malloc_conf_opts) if len(malloc_conf_opts) > 0 + else '') + ) + + # We don't want to test large vaddr spaces in 32-bit mode. + if ('-m32' in compiler_opts and '--with-lg-vaddr=56' in + config_opts): + continue + + # Per CPU arenas are only supported on Linux. + linux_supported = ('percpu_arena:percpu' in malloc_conf_opts \ + or 'background_thread:true' in malloc_conf_opts) + # Heap profiling and dss are not supported on OS X. + darwin_unsupported = ('--enable-prof' in config_opts or \ + 'dss:primary' in malloc_conf_opts) + if (uname == 'Linux' and linux_supported) \ + or (not linux_supported and (uname != 'Darwin' or \ + not darwin_unsupported)): + print """cat <<EOF > run_test_%(ind)d.sh +#!/bin/sh + +set -e + +abort() { + echo "==> Error" >> run_test.log + echo "Error; see run_tests.out/run_test_%(ind)d.out/run_test.log" + exit 255 # Special exit code tells xargs to terminate. +} + +# Environment variables are not supported. +run_cmd() { + echo "==> \$@" >> run_test.log + \$@ >> run_test.log 2>&1 || abort +} + +echo "=> run_test_%(ind)d: %(config_line)s" +mkdir run_test_%(ind)d.out +cd run_test_%(ind)d.out + +echo "==> %(config_line)s" >> run_test.log +%(config_line)s >> run_test.log 2>&1 || abort + +run_cmd %(make_cmd)s all tests +run_cmd %(make_cmd)s check +run_cmd %(make_cmd)s distclean +EOF +chmod 755 run_test_%(ind)d.sh""" % {'ind': ind, 'config_line': config_line, 'make_cmd': make_cmd} + ind += 1 + +print 'for i in `seq 0 %(last_ind)d` ; do echo run_test_${i}.sh ; done | xargs -P %(nparallel)d -n 1 sh' % {'last_ind': ind-1, 'nparallel': nparallel} diff --git a/deps/jemalloc/scripts/gen_travis.py b/deps/jemalloc/scripts/gen_travis.py new file mode 100755 index 000000000..f1478c62c --- /dev/null +++ b/deps/jemalloc/scripts/gen_travis.py @@ -0,0 +1,149 @@ +#!/usr/bin/env python + +from itertools import combinations + +travis_template = """\ +language: generic +dist: precise + +matrix: + include: +%s + +before_script: + - autoconf + - scripts/gen_travis.py > travis_script && diff .travis.yml travis_script + - ./configure ${COMPILER_FLAGS:+ \ + CC="$CC $COMPILER_FLAGS" \ + CXX="$CXX $COMPILER_FLAGS" } \ + $CONFIGURE_FLAGS + - make -j3 + - make -j3 tests + +script: + - make check +""" + +# The 'default' configuration is gcc, on linux, with no compiler or configure +# flags. We also test with clang, -m32, --enable-debug, --enable-prof, +# --disable-stats, and --with-malloc-conf=tcache:false. To avoid abusing +# travis though, we don't test all 2**7 = 128 possible combinations of these; +# instead, we only test combinations of up to 2 'unusual' settings, under the +# hope that bugs involving interactions of such settings are rare. +# Things at once, for C(7, 0) + C(7, 1) + C(7, 2) = 29 +MAX_UNUSUAL_OPTIONS = 2 + +os_default = 'linux' +os_unusual = 'osx' + +compilers_default = 'CC=gcc CXX=g++' +compilers_unusual = 'CC=clang CXX=clang++' + +compiler_flag_unusuals = ['-m32'] + +configure_flag_unusuals = [ + '--enable-debug', + '--enable-prof', + '--disable-stats', + '--disable-libdl', + '--enable-opt-safety-checks', +] + +malloc_conf_unusuals = [ + 'tcache:false', + 'dss:primary', + 'percpu_arena:percpu', + 'background_thread:true', +] + +all_unusuals = ( + [os_unusual] + [compilers_unusual] + compiler_flag_unusuals + + configure_flag_unusuals + malloc_conf_unusuals +) + +unusual_combinations_to_test = [] +for i in xrange(MAX_UNUSUAL_OPTIONS + 1): + unusual_combinations_to_test += combinations(all_unusuals, i) + +gcc_multilib_set = False +# Formats a job from a combination of flags +def format_job(combination): + global gcc_multilib_set + + os = os_unusual if os_unusual in combination else os_default + compilers = compilers_unusual if compilers_unusual in combination else compilers_default + + compiler_flags = [x for x in combination if x in compiler_flag_unusuals] + configure_flags = [x for x in combination if x in configure_flag_unusuals] + malloc_conf = [x for x in combination if x in malloc_conf_unusuals] + + # Filter out unsupported configurations on OS X. + if os == 'osx' and ('dss:primary' in malloc_conf or \ + 'percpu_arena:percpu' in malloc_conf or 'background_thread:true' \ + in malloc_conf): + return "" + if len(malloc_conf) > 0: + configure_flags.append('--with-malloc-conf=' + ",".join(malloc_conf)) + + # Filter out an unsupported configuration - heap profiling on OS X. + if os == 'osx' and '--enable-prof' in configure_flags: + return "" + + # We get some spurious errors when -Warray-bounds is enabled. + env_string = ('{} COMPILER_FLAGS="{}" CONFIGURE_FLAGS="{}" ' + 'EXTRA_CFLAGS="-Werror -Wno-array-bounds"').format( + compilers, " ".join(compiler_flags), " ".join(configure_flags)) + + job = "" + job += ' - os: %s\n' % os + job += ' env: %s\n' % env_string + if '-m32' in combination and os == 'linux': + job += ' addons:' + if gcc_multilib_set: + job += ' *gcc_multilib\n' + else: + job += ' &gcc_multilib\n' + job += ' apt:\n' + job += ' packages:\n' + job += ' - gcc-multilib\n' + gcc_multilib_set = True + return job + +include_rows = "" +for combination in unusual_combinations_to_test: + include_rows += format_job(combination) + +# Development build +include_rows += '''\ + # Development build + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-debug --disable-cache-oblivious --enable-stats --enable-log --enable-prof" EXTRA_CFLAGS="-Werror -Wno-array-bounds" +''' + +# Enable-expermental-smallocx +include_rows += '''\ + # --enable-expermental-smallocx: + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="--enable-debug --enable-experimental-smallocx --enable-stats --enable-prof" EXTRA_CFLAGS="-Werror -Wno-array-bounds" +''' + +# Valgrind build bots +include_rows += ''' + # Valgrind + - os: linux + env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="" EXTRA_CFLAGS="-Werror -Wno-array-bounds" JEMALLOC_TEST_PREFIX="valgrind" + addons: + apt: + packages: + - valgrind +''' + +# To enable valgrind on macosx add: +# +# - os: osx +# env: CC=gcc CXX=g++ COMPILER_FLAGS="" CONFIGURE_FLAGS="" EXTRA_CFLAGS="-Werror -Wno-array-bounds" JEMALLOC_TEST_PREFIX="valgrind" +# install: brew install valgrind +# +# It currently fails due to: https://github.com/jemalloc/jemalloc/issues/1274 + +print travis_template % include_rows diff --git a/deps/jemalloc/src/arena.c b/deps/jemalloc/src/arena.c new file mode 100644 index 000000000..ba50e4103 --- /dev/null +++ b/deps/jemalloc/src/arena.c @@ -0,0 +1,2298 @@ +#define JEMALLOC_ARENA_C_ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/assert.h" +#include "jemalloc/internal/div.h" +#include "jemalloc/internal/extent_dss.h" +#include "jemalloc/internal/extent_mmap.h" +#include "jemalloc/internal/mutex.h" +#include "jemalloc/internal/rtree.h" +#include "jemalloc/internal/safety_check.h" +#include "jemalloc/internal/util.h" + +JEMALLOC_DIAGNOSTIC_DISABLE_SPURIOUS + +/******************************************************************************/ +/* Data. */ + +/* + * Define names for both unininitialized and initialized phases, so that + * options and mallctl processing are straightforward. + */ +const char *percpu_arena_mode_names[] = { + "percpu", + "phycpu", + "disabled", + "percpu", + "phycpu" +}; +percpu_arena_mode_t opt_percpu_arena = PERCPU_ARENA_DEFAULT; + +ssize_t opt_dirty_decay_ms = DIRTY_DECAY_MS_DEFAULT; +ssize_t opt_muzzy_decay_ms = MUZZY_DECAY_MS_DEFAULT; + +static atomic_zd_t dirty_decay_ms_default; +static atomic_zd_t muzzy_decay_ms_default; + +const uint64_t h_steps[SMOOTHSTEP_NSTEPS] = { +#define STEP(step, h, x, y) \ + h, + SMOOTHSTEP +#undef STEP +}; + +static div_info_t arena_binind_div_info[SC_NBINS]; + +size_t opt_oversize_threshold = OVERSIZE_THRESHOLD_DEFAULT; +size_t oversize_threshold = OVERSIZE_THRESHOLD_DEFAULT; +static unsigned huge_arena_ind; + +/******************************************************************************/ +/* + * Function prototypes for static functions that are referenced prior to + * definition. + */ + +static void arena_decay_to_limit(tsdn_t *tsdn, arena_t *arena, + arena_decay_t *decay, extents_t *extents, bool all, size_t npages_limit, + size_t npages_decay_max, bool is_background_thread); +static bool arena_decay_dirty(tsdn_t *tsdn, arena_t *arena, + bool is_background_thread, bool all); +static void arena_dalloc_bin_slab(tsdn_t *tsdn, arena_t *arena, extent_t *slab, + bin_t *bin); +static void arena_bin_lower_slab(tsdn_t *tsdn, arena_t *arena, extent_t *slab, + bin_t *bin); + +/******************************************************************************/ + +void +arena_basic_stats_merge(tsdn_t *tsdn, arena_t *arena, unsigned *nthreads, + const char **dss, ssize_t *dirty_decay_ms, ssize_t *muzzy_decay_ms, + size_t *nactive, size_t *ndirty, size_t *nmuzzy) { + *nthreads += arena_nthreads_get(arena, false); + *dss = dss_prec_names[arena_dss_prec_get(arena)]; + *dirty_decay_ms = arena_dirty_decay_ms_get(arena); + *muzzy_decay_ms = arena_muzzy_decay_ms_get(arena); + *nactive += atomic_load_zu(&arena->nactive, ATOMIC_RELAXED); + *ndirty += extents_npages_get(&arena->extents_dirty); + *nmuzzy += extents_npages_get(&arena->extents_muzzy); +} + +void +arena_stats_merge(tsdn_t *tsdn, arena_t *arena, unsigned *nthreads, + const char **dss, ssize_t *dirty_decay_ms, ssize_t *muzzy_decay_ms, + size_t *nactive, size_t *ndirty, size_t *nmuzzy, arena_stats_t *astats, + bin_stats_t *bstats, arena_stats_large_t *lstats, + arena_stats_extents_t *estats) { + cassert(config_stats); + + arena_basic_stats_merge(tsdn, arena, nthreads, dss, dirty_decay_ms, + muzzy_decay_ms, nactive, ndirty, nmuzzy); + + size_t base_allocated, base_resident, base_mapped, metadata_thp; + base_stats_get(tsdn, arena->base, &base_allocated, &base_resident, + &base_mapped, &metadata_thp); + + arena_stats_lock(tsdn, &arena->stats); + + arena_stats_accum_zu(&astats->mapped, base_mapped + + arena_stats_read_zu(tsdn, &arena->stats, &arena->stats.mapped)); + arena_stats_accum_zu(&astats->retained, + extents_npages_get(&arena->extents_retained) << LG_PAGE); + + atomic_store_zu(&astats->extent_avail, + atomic_load_zu(&arena->extent_avail_cnt, ATOMIC_RELAXED), + ATOMIC_RELAXED); + + arena_stats_accum_u64(&astats->decay_dirty.npurge, + arena_stats_read_u64(tsdn, &arena->stats, + &arena->stats.decay_dirty.npurge)); + arena_stats_accum_u64(&astats->decay_dirty.nmadvise, + arena_stats_read_u64(tsdn, &arena->stats, + &arena->stats.decay_dirty.nmadvise)); + arena_stats_accum_u64(&astats->decay_dirty.purged, + arena_stats_read_u64(tsdn, &arena->stats, + &arena->stats.decay_dirty.purged)); + + arena_stats_accum_u64(&astats->decay_muzzy.npurge, + arena_stats_read_u64(tsdn, &arena->stats, + &arena->stats.decay_muzzy.npurge)); + arena_stats_accum_u64(&astats->decay_muzzy.nmadvise, + arena_stats_read_u64(tsdn, &arena->stats, + &arena->stats.decay_muzzy.nmadvise)); + arena_stats_accum_u64(&astats->decay_muzzy.purged, + arena_stats_read_u64(tsdn, &arena->stats, + &arena->stats.decay_muzzy.purged)); + + arena_stats_accum_zu(&astats->base, base_allocated); + arena_stats_accum_zu(&astats->internal, arena_internal_get(arena)); + arena_stats_accum_zu(&astats->metadata_thp, metadata_thp); + arena_stats_accum_zu(&astats->resident, base_resident + + (((atomic_load_zu(&arena->nactive, ATOMIC_RELAXED) + + extents_npages_get(&arena->extents_dirty) + + extents_npages_get(&arena->extents_muzzy)) << LG_PAGE))); + arena_stats_accum_zu(&astats->abandoned_vm, atomic_load_zu( + &arena->stats.abandoned_vm, ATOMIC_RELAXED)); + + for (szind_t i = 0; i < SC_NSIZES - SC_NBINS; i++) { + uint64_t nmalloc = arena_stats_read_u64(tsdn, &arena->stats, + &arena->stats.lstats[i].nmalloc); + arena_stats_accum_u64(&lstats[i].nmalloc, nmalloc); + arena_stats_accum_u64(&astats->nmalloc_large, nmalloc); + + uint64_t ndalloc = arena_stats_read_u64(tsdn, &arena->stats, + &arena->stats.lstats[i].ndalloc); + arena_stats_accum_u64(&lstats[i].ndalloc, ndalloc); + arena_stats_accum_u64(&astats->ndalloc_large, ndalloc); + + uint64_t nrequests = arena_stats_read_u64(tsdn, &arena->stats, + &arena->stats.lstats[i].nrequests); + arena_stats_accum_u64(&lstats[i].nrequests, + nmalloc + nrequests); + arena_stats_accum_u64(&astats->nrequests_large, + nmalloc + nrequests); + + /* nfill == nmalloc for large currently. */ + arena_stats_accum_u64(&lstats[i].nfills, nmalloc); + arena_stats_accum_u64(&astats->nfills_large, nmalloc); + + uint64_t nflush = arena_stats_read_u64(tsdn, &arena->stats, + &arena->stats.lstats[i].nflushes); + arena_stats_accum_u64(&lstats[i].nflushes, nflush); + arena_stats_accum_u64(&astats->nflushes_large, nflush); + + assert(nmalloc >= ndalloc); + assert(nmalloc - ndalloc <= SIZE_T_MAX); + size_t curlextents = (size_t)(nmalloc - ndalloc); + lstats[i].curlextents += curlextents; + arena_stats_accum_zu(&astats->allocated_large, + curlextents * sz_index2size(SC_NBINS + i)); + } + + for (pszind_t i = 0; i < SC_NPSIZES; i++) { + size_t dirty, muzzy, retained, dirty_bytes, muzzy_bytes, + retained_bytes; + dirty = extents_nextents_get(&arena->extents_dirty, i); + muzzy = extents_nextents_get(&arena->extents_muzzy, i); + retained = extents_nextents_get(&arena->extents_retained, i); + dirty_bytes = extents_nbytes_get(&arena->extents_dirty, i); + muzzy_bytes = extents_nbytes_get(&arena->extents_muzzy, i); + retained_bytes = + extents_nbytes_get(&arena->extents_retained, i); + + atomic_store_zu(&estats[i].ndirty, dirty, ATOMIC_RELAXED); + atomic_store_zu(&estats[i].nmuzzy, muzzy, ATOMIC_RELAXED); + atomic_store_zu(&estats[i].nretained, retained, ATOMIC_RELAXED); + atomic_store_zu(&estats[i].dirty_bytes, dirty_bytes, + ATOMIC_RELAXED); + atomic_store_zu(&estats[i].muzzy_bytes, muzzy_bytes, + ATOMIC_RELAXED); + atomic_store_zu(&estats[i].retained_bytes, retained_bytes, + ATOMIC_RELAXED); + } + + arena_stats_unlock(tsdn, &arena->stats); + + /* tcache_bytes counts currently cached bytes. */ + atomic_store_zu(&astats->tcache_bytes, 0, ATOMIC_RELAXED); + malloc_mutex_lock(tsdn, &arena->tcache_ql_mtx); + cache_bin_array_descriptor_t *descriptor; + ql_foreach(descriptor, &arena->cache_bin_array_descriptor_ql, link) { + szind_t i = 0; + for (; i < SC_NBINS; i++) { + cache_bin_t *tbin = &descriptor->bins_small[i]; + arena_stats_accum_zu(&astats->tcache_bytes, + tbin->ncached * sz_index2size(i)); + } + for (; i < nhbins; i++) { + cache_bin_t *tbin = &descriptor->bins_large[i]; + arena_stats_accum_zu(&astats->tcache_bytes, + tbin->ncached * sz_index2size(i)); + } + } + malloc_mutex_prof_read(tsdn, + &astats->mutex_prof_data[arena_prof_mutex_tcache_list], + &arena->tcache_ql_mtx); + malloc_mutex_unlock(tsdn, &arena->tcache_ql_mtx); + +#define READ_ARENA_MUTEX_PROF_DATA(mtx, ind) \ + malloc_mutex_lock(tsdn, &arena->mtx); \ + malloc_mutex_prof_read(tsdn, &astats->mutex_prof_data[ind], \ + &arena->mtx); \ + malloc_mutex_unlock(tsdn, &arena->mtx); + + /* Gather per arena mutex profiling data. */ + READ_ARENA_MUTEX_PROF_DATA(large_mtx, arena_prof_mutex_large); + READ_ARENA_MUTEX_PROF_DATA(extent_avail_mtx, + arena_prof_mutex_extent_avail) + READ_ARENA_MUTEX_PROF_DATA(extents_dirty.mtx, + arena_prof_mutex_extents_dirty) + READ_ARENA_MUTEX_PROF_DATA(extents_muzzy.mtx, + arena_prof_mutex_extents_muzzy) + READ_ARENA_MUTEX_PROF_DATA(extents_retained.mtx, + arena_prof_mutex_extents_retained) + READ_ARENA_MUTEX_PROF_DATA(decay_dirty.mtx, + arena_prof_mutex_decay_dirty) + READ_ARENA_MUTEX_PROF_DATA(decay_muzzy.mtx, + arena_prof_mutex_decay_muzzy) + READ_ARENA_MUTEX_PROF_DATA(base->mtx, + arena_prof_mutex_base) +#undef READ_ARENA_MUTEX_PROF_DATA + + nstime_copy(&astats->uptime, &arena->create_time); + nstime_update(&astats->uptime); + nstime_subtract(&astats->uptime, &arena->create_time); + + for (szind_t i = 0; i < SC_NBINS; i++) { + for (unsigned j = 0; j < bin_infos[i].n_shards; j++) { + bin_stats_merge(tsdn, &bstats[i], + &arena->bins[i].bin_shards[j]); + } + } +} + +void +arena_extents_dirty_dalloc(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent) { + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, 0); + + extents_dalloc(tsdn, arena, r_extent_hooks, &arena->extents_dirty, + extent); + if (arena_dirty_decay_ms_get(arena) == 0) { + arena_decay_dirty(tsdn, arena, false, true); + } else { + arena_background_thread_inactivity_check(tsdn, arena, false); + } +} + +static void * +arena_slab_reg_alloc(extent_t *slab, const bin_info_t *bin_info) { + void *ret; + arena_slab_data_t *slab_data = extent_slab_data_get(slab); + size_t regind; + + assert(extent_nfree_get(slab) > 0); + assert(!bitmap_full(slab_data->bitmap, &bin_info->bitmap_info)); + + regind = bitmap_sfu(slab_data->bitmap, &bin_info->bitmap_info); + ret = (void *)((uintptr_t)extent_addr_get(slab) + + (uintptr_t)(bin_info->reg_size * regind)); + extent_nfree_dec(slab); + return ret; +} + +static void +arena_slab_reg_alloc_batch(extent_t *slab, const bin_info_t *bin_info, + unsigned cnt, void** ptrs) { + arena_slab_data_t *slab_data = extent_slab_data_get(slab); + + assert(extent_nfree_get(slab) >= cnt); + assert(!bitmap_full(slab_data->bitmap, &bin_info->bitmap_info)); + +#if (! defined JEMALLOC_INTERNAL_POPCOUNTL) || (defined BITMAP_USE_TREE) + for (unsigned i = 0; i < cnt; i++) { + size_t regind = bitmap_sfu(slab_data->bitmap, + &bin_info->bitmap_info); + *(ptrs + i) = (void *)((uintptr_t)extent_addr_get(slab) + + (uintptr_t)(bin_info->reg_size * regind)); + } +#else + unsigned group = 0; + bitmap_t g = slab_data->bitmap[group]; + unsigned i = 0; + while (i < cnt) { + while (g == 0) { + g = slab_data->bitmap[++group]; + } + size_t shift = group << LG_BITMAP_GROUP_NBITS; + size_t pop = popcount_lu(g); + if (pop > (cnt - i)) { + pop = cnt - i; + } + + /* + * Load from memory locations only once, outside the + * hot loop below. + */ + uintptr_t base = (uintptr_t)extent_addr_get(slab); + uintptr_t regsize = (uintptr_t)bin_info->reg_size; + while (pop--) { + size_t bit = cfs_lu(&g); + size_t regind = shift + bit; + *(ptrs + i) = (void *)(base + regsize * regind); + + i++; + } + slab_data->bitmap[group] = g; + } +#endif + extent_nfree_sub(slab, cnt); +} + +#ifndef JEMALLOC_JET +static +#endif +size_t +arena_slab_regind(extent_t *slab, szind_t binind, const void *ptr) { + size_t diff, regind; + + /* Freeing a pointer outside the slab can cause assertion failure. */ + assert((uintptr_t)ptr >= (uintptr_t)extent_addr_get(slab)); + assert((uintptr_t)ptr < (uintptr_t)extent_past_get(slab)); + /* Freeing an interior pointer can cause assertion failure. */ + assert(((uintptr_t)ptr - (uintptr_t)extent_addr_get(slab)) % + (uintptr_t)bin_infos[binind].reg_size == 0); + + diff = (size_t)((uintptr_t)ptr - (uintptr_t)extent_addr_get(slab)); + + /* Avoid doing division with a variable divisor. */ + regind = div_compute(&arena_binind_div_info[binind], diff); + + assert(regind < bin_infos[binind].nregs); + + return regind; +} + +static void +arena_slab_reg_dalloc(extent_t *slab, arena_slab_data_t *slab_data, void *ptr) { + szind_t binind = extent_szind_get(slab); + const bin_info_t *bin_info = &bin_infos[binind]; + size_t regind = arena_slab_regind(slab, binind, ptr); + + assert(extent_nfree_get(slab) < bin_info->nregs); + /* Freeing an unallocated pointer can cause assertion failure. */ + assert(bitmap_get(slab_data->bitmap, &bin_info->bitmap_info, regind)); + + bitmap_unset(slab_data->bitmap, &bin_info->bitmap_info, regind); + extent_nfree_inc(slab); +} + +static void +arena_nactive_add(arena_t *arena, size_t add_pages) { + atomic_fetch_add_zu(&arena->nactive, add_pages, ATOMIC_RELAXED); +} + +static void +arena_nactive_sub(arena_t *arena, size_t sub_pages) { + assert(atomic_load_zu(&arena->nactive, ATOMIC_RELAXED) >= sub_pages); + atomic_fetch_sub_zu(&arena->nactive, sub_pages, ATOMIC_RELAXED); +} + +static void +arena_large_malloc_stats_update(tsdn_t *tsdn, arena_t *arena, size_t usize) { + szind_t index, hindex; + + cassert(config_stats); + + if (usize < SC_LARGE_MINCLASS) { + usize = SC_LARGE_MINCLASS; + } + index = sz_size2index(usize); + hindex = (index >= SC_NBINS) ? index - SC_NBINS : 0; + + arena_stats_add_u64(tsdn, &arena->stats, + &arena->stats.lstats[hindex].nmalloc, 1); +} + +static void +arena_large_dalloc_stats_update(tsdn_t *tsdn, arena_t *arena, size_t usize) { + szind_t index, hindex; + + cassert(config_stats); + + if (usize < SC_LARGE_MINCLASS) { + usize = SC_LARGE_MINCLASS; + } + index = sz_size2index(usize); + hindex = (index >= SC_NBINS) ? index - SC_NBINS : 0; + + arena_stats_add_u64(tsdn, &arena->stats, + &arena->stats.lstats[hindex].ndalloc, 1); +} + +static void +arena_large_ralloc_stats_update(tsdn_t *tsdn, arena_t *arena, size_t oldusize, + size_t usize) { + arena_large_dalloc_stats_update(tsdn, arena, oldusize); + arena_large_malloc_stats_update(tsdn, arena, usize); +} + +static bool +arena_may_have_muzzy(arena_t *arena) { + return (pages_can_purge_lazy && (arena_muzzy_decay_ms_get(arena) != 0)); +} + +extent_t * +arena_extent_alloc_large(tsdn_t *tsdn, arena_t *arena, size_t usize, + size_t alignment, bool *zero) { + extent_hooks_t *extent_hooks = EXTENT_HOOKS_INITIALIZER; + + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, 0); + + szind_t szind = sz_size2index(usize); + size_t mapped_add; + bool commit = true; + extent_t *extent = extents_alloc(tsdn, arena, &extent_hooks, + &arena->extents_dirty, NULL, usize, sz_large_pad, alignment, false, + szind, zero, &commit); + if (extent == NULL && arena_may_have_muzzy(arena)) { + extent = extents_alloc(tsdn, arena, &extent_hooks, + &arena->extents_muzzy, NULL, usize, sz_large_pad, alignment, + false, szind, zero, &commit); + } + size_t size = usize + sz_large_pad; + if (extent == NULL) { + extent = extent_alloc_wrapper(tsdn, arena, &extent_hooks, NULL, + usize, sz_large_pad, alignment, false, szind, zero, + &commit); + if (config_stats) { + /* + * extent may be NULL on OOM, but in that case + * mapped_add isn't used below, so there's no need to + * conditionlly set it to 0 here. + */ + mapped_add = size; + } + } else if (config_stats) { + mapped_add = 0; + } + + if (extent != NULL) { + if (config_stats) { + arena_stats_lock(tsdn, &arena->stats); + arena_large_malloc_stats_update(tsdn, arena, usize); + if (mapped_add != 0) { + arena_stats_add_zu(tsdn, &arena->stats, + &arena->stats.mapped, mapped_add); + } + arena_stats_unlock(tsdn, &arena->stats); + } + arena_nactive_add(arena, size >> LG_PAGE); + } + + return extent; +} + +void +arena_extent_dalloc_large_prep(tsdn_t *tsdn, arena_t *arena, extent_t *extent) { + if (config_stats) { + arena_stats_lock(tsdn, &arena->stats); + arena_large_dalloc_stats_update(tsdn, arena, + extent_usize_get(extent)); + arena_stats_unlock(tsdn, &arena->stats); + } + arena_nactive_sub(arena, extent_size_get(extent) >> LG_PAGE); +} + +void +arena_extent_ralloc_large_shrink(tsdn_t *tsdn, arena_t *arena, extent_t *extent, + size_t oldusize) { + size_t usize = extent_usize_get(extent); + size_t udiff = oldusize - usize; + + if (config_stats) { + arena_stats_lock(tsdn, &arena->stats); + arena_large_ralloc_stats_update(tsdn, arena, oldusize, usize); + arena_stats_unlock(tsdn, &arena->stats); + } + arena_nactive_sub(arena, udiff >> LG_PAGE); +} + +void +arena_extent_ralloc_large_expand(tsdn_t *tsdn, arena_t *arena, extent_t *extent, + size_t oldusize) { + size_t usize = extent_usize_get(extent); + size_t udiff = usize - oldusize; + + if (config_stats) { + arena_stats_lock(tsdn, &arena->stats); + arena_large_ralloc_stats_update(tsdn, arena, oldusize, usize); + arena_stats_unlock(tsdn, &arena->stats); + } + arena_nactive_add(arena, udiff >> LG_PAGE); +} + +static ssize_t +arena_decay_ms_read(arena_decay_t *decay) { + return atomic_load_zd(&decay->time_ms, ATOMIC_RELAXED); +} + +static void +arena_decay_ms_write(arena_decay_t *decay, ssize_t decay_ms) { + atomic_store_zd(&decay->time_ms, decay_ms, ATOMIC_RELAXED); +} + +static void +arena_decay_deadline_init(arena_decay_t *decay) { + /* + * Generate a new deadline that is uniformly random within the next + * epoch after the current one. + */ + nstime_copy(&decay->deadline, &decay->epoch); + nstime_add(&decay->deadline, &decay->interval); + if (arena_decay_ms_read(decay) > 0) { + nstime_t jitter; + + nstime_init(&jitter, prng_range_u64(&decay->jitter_state, + nstime_ns(&decay->interval))); + nstime_add(&decay->deadline, &jitter); + } +} + +static bool +arena_decay_deadline_reached(const arena_decay_t *decay, const nstime_t *time) { + return (nstime_compare(&decay->deadline, time) <= 0); +} + +static size_t +arena_decay_backlog_npages_limit(const arena_decay_t *decay) { + uint64_t sum; + size_t npages_limit_backlog; + unsigned i; + + /* + * For each element of decay_backlog, multiply by the corresponding + * fixed-point smoothstep decay factor. Sum the products, then divide + * to round down to the nearest whole number of pages. + */ + sum = 0; + for (i = 0; i < SMOOTHSTEP_NSTEPS; i++) { + sum += decay->backlog[i] * h_steps[i]; + } + npages_limit_backlog = (size_t)(sum >> SMOOTHSTEP_BFP); + + return npages_limit_backlog; +} + +static void +arena_decay_backlog_update_last(arena_decay_t *decay, size_t current_npages) { + size_t npages_delta = (current_npages > decay->nunpurged) ? + current_npages - decay->nunpurged : 0; + decay->backlog[SMOOTHSTEP_NSTEPS-1] = npages_delta; + + if (config_debug) { + if (current_npages > decay->ceil_npages) { + decay->ceil_npages = current_npages; + } + size_t npages_limit = arena_decay_backlog_npages_limit(decay); + assert(decay->ceil_npages >= npages_limit); + if (decay->ceil_npages > npages_limit) { + decay->ceil_npages = npages_limit; + } + } +} + +static void +arena_decay_backlog_update(arena_decay_t *decay, uint64_t nadvance_u64, + size_t current_npages) { + if (nadvance_u64 >= SMOOTHSTEP_NSTEPS) { + memset(decay->backlog, 0, (SMOOTHSTEP_NSTEPS-1) * + sizeof(size_t)); + } else { + size_t nadvance_z = (size_t)nadvance_u64; + + assert((uint64_t)nadvance_z == nadvance_u64); + + memmove(decay->backlog, &decay->backlog[nadvance_z], + (SMOOTHSTEP_NSTEPS - nadvance_z) * sizeof(size_t)); + if (nadvance_z > 1) { + memset(&decay->backlog[SMOOTHSTEP_NSTEPS - + nadvance_z], 0, (nadvance_z-1) * sizeof(size_t)); + } + } + + arena_decay_backlog_update_last(decay, current_npages); +} + +static void +arena_decay_try_purge(tsdn_t *tsdn, arena_t *arena, arena_decay_t *decay, + extents_t *extents, size_t current_npages, size_t npages_limit, + bool is_background_thread) { + if (current_npages > npages_limit) { + arena_decay_to_limit(tsdn, arena, decay, extents, false, + npages_limit, current_npages - npages_limit, + is_background_thread); + } +} + +static void +arena_decay_epoch_advance_helper(arena_decay_t *decay, const nstime_t *time, + size_t current_npages) { + assert(arena_decay_deadline_reached(decay, time)); + + nstime_t delta; + nstime_copy(&delta, time); + nstime_subtract(&delta, &decay->epoch); + + uint64_t nadvance_u64 = nstime_divide(&delta, &decay->interval); + assert(nadvance_u64 > 0); + + /* Add nadvance_u64 decay intervals to epoch. */ + nstime_copy(&delta, &decay->interval); + nstime_imultiply(&delta, nadvance_u64); + nstime_add(&decay->epoch, &delta); + + /* Set a new deadline. */ + arena_decay_deadline_init(decay); + + /* Update the backlog. */ + arena_decay_backlog_update(decay, nadvance_u64, current_npages); +} + +static void +arena_decay_epoch_advance(tsdn_t *tsdn, arena_t *arena, arena_decay_t *decay, + extents_t *extents, const nstime_t *time, bool is_background_thread) { + size_t current_npages = extents_npages_get(extents); + arena_decay_epoch_advance_helper(decay, time, current_npages); + + size_t npages_limit = arena_decay_backlog_npages_limit(decay); + /* We may unlock decay->mtx when try_purge(). Finish logging first. */ + decay->nunpurged = (npages_limit > current_npages) ? npages_limit : + current_npages; + + if (!background_thread_enabled() || is_background_thread) { + arena_decay_try_purge(tsdn, arena, decay, extents, + current_npages, npages_limit, is_background_thread); + } +} + +static void +arena_decay_reinit(arena_decay_t *decay, ssize_t decay_ms) { + arena_decay_ms_write(decay, decay_ms); + if (decay_ms > 0) { + nstime_init(&decay->interval, (uint64_t)decay_ms * + KQU(1000000)); + nstime_idivide(&decay->interval, SMOOTHSTEP_NSTEPS); + } + + nstime_init(&decay->epoch, 0); + nstime_update(&decay->epoch); + decay->jitter_state = (uint64_t)(uintptr_t)decay; + arena_decay_deadline_init(decay); + decay->nunpurged = 0; + memset(decay->backlog, 0, SMOOTHSTEP_NSTEPS * sizeof(size_t)); +} + +static bool +arena_decay_init(arena_decay_t *decay, ssize_t decay_ms, + arena_stats_decay_t *stats) { + if (config_debug) { + for (size_t i = 0; i < sizeof(arena_decay_t); i++) { + assert(((char *)decay)[i] == 0); + } + decay->ceil_npages = 0; + } + if (malloc_mutex_init(&decay->mtx, "decay", WITNESS_RANK_DECAY, + malloc_mutex_rank_exclusive)) { + return true; + } + decay->purging = false; + arena_decay_reinit(decay, decay_ms); + /* Memory is zeroed, so there is no need to clear stats. */ + if (config_stats) { + decay->stats = stats; + } + return false; +} + +static bool +arena_decay_ms_valid(ssize_t decay_ms) { + if (decay_ms < -1) { + return false; + } + if (decay_ms == -1 || (uint64_t)decay_ms <= NSTIME_SEC_MAX * + KQU(1000)) { + return true; + } + return false; +} + +static bool +arena_maybe_decay(tsdn_t *tsdn, arena_t *arena, arena_decay_t *decay, + extents_t *extents, bool is_background_thread) { + malloc_mutex_assert_owner(tsdn, &decay->mtx); + + /* Purge all or nothing if the option is disabled. */ + ssize_t decay_ms = arena_decay_ms_read(decay); + if (decay_ms <= 0) { + if (decay_ms == 0) { + arena_decay_to_limit(tsdn, arena, decay, extents, false, + 0, extents_npages_get(extents), + is_background_thread); + } + return false; + } + + nstime_t time; + nstime_init(&time, 0); + nstime_update(&time); + if (unlikely(!nstime_monotonic() && nstime_compare(&decay->epoch, &time) + > 0)) { + /* + * Time went backwards. Move the epoch back in time and + * generate a new deadline, with the expectation that time + * typically flows forward for long enough periods of time that + * epochs complete. Unfortunately, this strategy is susceptible + * to clock jitter triggering premature epoch advances, but + * clock jitter estimation and compensation isn't feasible here + * because calls into this code are event-driven. + */ + nstime_copy(&decay->epoch, &time); + arena_decay_deadline_init(decay); + } else { + /* Verify that time does not go backwards. */ + assert(nstime_compare(&decay->epoch, &time) <= 0); + } + + /* + * If the deadline has been reached, advance to the current epoch and + * purge to the new limit if necessary. Note that dirty pages created + * during the current epoch are not subject to purge until a future + * epoch, so as a result purging only happens during epoch advances, or + * being triggered by background threads (scheduled event). + */ + bool advance_epoch = arena_decay_deadline_reached(decay, &time); + if (advance_epoch) { + arena_decay_epoch_advance(tsdn, arena, decay, extents, &time, + is_background_thread); + } else if (is_background_thread) { + arena_decay_try_purge(tsdn, arena, decay, extents, + extents_npages_get(extents), + arena_decay_backlog_npages_limit(decay), + is_background_thread); + } + + return advance_epoch; +} + +static ssize_t +arena_decay_ms_get(arena_decay_t *decay) { + return arena_decay_ms_read(decay); +} + +ssize_t +arena_dirty_decay_ms_get(arena_t *arena) { + return arena_decay_ms_get(&arena->decay_dirty); +} + +ssize_t +arena_muzzy_decay_ms_get(arena_t *arena) { + return arena_decay_ms_get(&arena->decay_muzzy); +} + +static bool +arena_decay_ms_set(tsdn_t *tsdn, arena_t *arena, arena_decay_t *decay, + extents_t *extents, ssize_t decay_ms) { + if (!arena_decay_ms_valid(decay_ms)) { + return true; + } + + malloc_mutex_lock(tsdn, &decay->mtx); + /* + * Restart decay backlog from scratch, which may cause many dirty pages + * to be immediately purged. It would conceptually be possible to map + * the old backlog onto the new backlog, but there is no justification + * for such complexity since decay_ms changes are intended to be + * infrequent, either between the {-1, 0, >0} states, or a one-time + * arbitrary change during initial arena configuration. + */ + arena_decay_reinit(decay, decay_ms); + arena_maybe_decay(tsdn, arena, decay, extents, false); + malloc_mutex_unlock(tsdn, &decay->mtx); + + return false; +} + +bool +arena_dirty_decay_ms_set(tsdn_t *tsdn, arena_t *arena, + ssize_t decay_ms) { + return arena_decay_ms_set(tsdn, arena, &arena->decay_dirty, + &arena->extents_dirty, decay_ms); +} + +bool +arena_muzzy_decay_ms_set(tsdn_t *tsdn, arena_t *arena, + ssize_t decay_ms) { + return arena_decay_ms_set(tsdn, arena, &arena->decay_muzzy, + &arena->extents_muzzy, decay_ms); +} + +static size_t +arena_stash_decayed(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extents_t *extents, size_t npages_limit, + size_t npages_decay_max, extent_list_t *decay_extents) { + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, 0); + + /* Stash extents according to npages_limit. */ + size_t nstashed = 0; + extent_t *extent; + while (nstashed < npages_decay_max && + (extent = extents_evict(tsdn, arena, r_extent_hooks, extents, + npages_limit)) != NULL) { + extent_list_append(decay_extents, extent); + nstashed += extent_size_get(extent) >> LG_PAGE; + } + return nstashed; +} + +static size_t +arena_decay_stashed(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, arena_decay_t *decay, extents_t *extents, + bool all, extent_list_t *decay_extents, bool is_background_thread) { + size_t nmadvise, nunmapped; + size_t npurged; + + if (config_stats) { + nmadvise = 0; + nunmapped = 0; + } + npurged = 0; + + ssize_t muzzy_decay_ms = arena_muzzy_decay_ms_get(arena); + for (extent_t *extent = extent_list_first(decay_extents); extent != + NULL; extent = extent_list_first(decay_extents)) { + if (config_stats) { + nmadvise++; + } + size_t npages = extent_size_get(extent) >> LG_PAGE; + npurged += npages; + extent_list_remove(decay_extents, extent); + switch (extents_state_get(extents)) { + case extent_state_active: + not_reached(); + case extent_state_dirty: + if (!all && muzzy_decay_ms != 0 && + !extent_purge_lazy_wrapper(tsdn, arena, + r_extent_hooks, extent, 0, + extent_size_get(extent))) { + extents_dalloc(tsdn, arena, r_extent_hooks, + &arena->extents_muzzy, extent); + arena_background_thread_inactivity_check(tsdn, + arena, is_background_thread); + break; + } + /* Fall through. */ + case extent_state_muzzy: + extent_dalloc_wrapper(tsdn, arena, r_extent_hooks, + extent); + if (config_stats) { + nunmapped += npages; + } + break; + case extent_state_retained: + default: + not_reached(); + } + } + + if (config_stats) { + arena_stats_lock(tsdn, &arena->stats); + arena_stats_add_u64(tsdn, &arena->stats, &decay->stats->npurge, + 1); + arena_stats_add_u64(tsdn, &arena->stats, + &decay->stats->nmadvise, nmadvise); + arena_stats_add_u64(tsdn, &arena->stats, &decay->stats->purged, + npurged); + arena_stats_sub_zu(tsdn, &arena->stats, &arena->stats.mapped, + nunmapped << LG_PAGE); + arena_stats_unlock(tsdn, &arena->stats); + } + + return npurged; +} + +/* + * npages_limit: Decay at most npages_decay_max pages without violating the + * invariant: (extents_npages_get(extents) >= npages_limit). We need an upper + * bound on number of pages in order to prevent unbounded growth (namely in + * stashed), otherwise unbounded new pages could be added to extents during the + * current decay run, so that the purging thread never finishes. + */ +static void +arena_decay_to_limit(tsdn_t *tsdn, arena_t *arena, arena_decay_t *decay, + extents_t *extents, bool all, size_t npages_limit, size_t npages_decay_max, + bool is_background_thread) { + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, 1); + malloc_mutex_assert_owner(tsdn, &decay->mtx); + + if (decay->purging) { + return; + } + decay->purging = true; + malloc_mutex_unlock(tsdn, &decay->mtx); + + extent_hooks_t *extent_hooks = extent_hooks_get(arena); + + extent_list_t decay_extents; + extent_list_init(&decay_extents); + + size_t npurge = arena_stash_decayed(tsdn, arena, &extent_hooks, extents, + npages_limit, npages_decay_max, &decay_extents); + if (npurge != 0) { + size_t npurged = arena_decay_stashed(tsdn, arena, + &extent_hooks, decay, extents, all, &decay_extents, + is_background_thread); + assert(npurged == npurge); + } + + malloc_mutex_lock(tsdn, &decay->mtx); + decay->purging = false; +} + +static bool +arena_decay_impl(tsdn_t *tsdn, arena_t *arena, arena_decay_t *decay, + extents_t *extents, bool is_background_thread, bool all) { + if (all) { + malloc_mutex_lock(tsdn, &decay->mtx); + arena_decay_to_limit(tsdn, arena, decay, extents, all, 0, + extents_npages_get(extents), is_background_thread); + malloc_mutex_unlock(tsdn, &decay->mtx); + + return false; + } + + if (malloc_mutex_trylock(tsdn, &decay->mtx)) { + /* No need to wait if another thread is in progress. */ + return true; + } + + bool epoch_advanced = arena_maybe_decay(tsdn, arena, decay, extents, + is_background_thread); + size_t npages_new; + if (epoch_advanced) { + /* Backlog is updated on epoch advance. */ + npages_new = decay->backlog[SMOOTHSTEP_NSTEPS-1]; + } + malloc_mutex_unlock(tsdn, &decay->mtx); + + if (have_background_thread && background_thread_enabled() && + epoch_advanced && !is_background_thread) { + background_thread_interval_check(tsdn, arena, decay, + npages_new); + } + + return false; +} + +static bool +arena_decay_dirty(tsdn_t *tsdn, arena_t *arena, bool is_background_thread, + bool all) { + return arena_decay_impl(tsdn, arena, &arena->decay_dirty, + &arena->extents_dirty, is_background_thread, all); +} + +static bool +arena_decay_muzzy(tsdn_t *tsdn, arena_t *arena, bool is_background_thread, + bool all) { + return arena_decay_impl(tsdn, arena, &arena->decay_muzzy, + &arena->extents_muzzy, is_background_thread, all); +} + +void +arena_decay(tsdn_t *tsdn, arena_t *arena, bool is_background_thread, bool all) { + if (arena_decay_dirty(tsdn, arena, is_background_thread, all)) { + return; + } + arena_decay_muzzy(tsdn, arena, is_background_thread, all); +} + +static void +arena_slab_dalloc(tsdn_t *tsdn, arena_t *arena, extent_t *slab) { + arena_nactive_sub(arena, extent_size_get(slab) >> LG_PAGE); + + extent_hooks_t *extent_hooks = EXTENT_HOOKS_INITIALIZER; + arena_extents_dirty_dalloc(tsdn, arena, &extent_hooks, slab); +} + +static void +arena_bin_slabs_nonfull_insert(bin_t *bin, extent_t *slab) { + assert(extent_nfree_get(slab) > 0); + extent_heap_insert(&bin->slabs_nonfull, slab); + if (config_stats) { + bin->stats.nonfull_slabs++; + } +} + +static void +arena_bin_slabs_nonfull_remove(bin_t *bin, extent_t *slab) { + extent_heap_remove(&bin->slabs_nonfull, slab); + if (config_stats) { + bin->stats.nonfull_slabs--; + } +} + +static extent_t * +arena_bin_slabs_nonfull_tryget(bin_t *bin) { + extent_t *slab = extent_heap_remove_first(&bin->slabs_nonfull); + if (slab == NULL) { + return NULL; + } + if (config_stats) { + bin->stats.reslabs++; + bin->stats.nonfull_slabs--; + } + return slab; +} + +static void +arena_bin_slabs_full_insert(arena_t *arena, bin_t *bin, extent_t *slab) { + assert(extent_nfree_get(slab) == 0); + /* + * Tracking extents is required by arena_reset, which is not allowed + * for auto arenas. Bypass this step to avoid touching the extent + * linkage (often results in cache misses) for auto arenas. + */ + if (arena_is_auto(arena)) { + return; + } + extent_list_append(&bin->slabs_full, slab); +} + +static void +arena_bin_slabs_full_remove(arena_t *arena, bin_t *bin, extent_t *slab) { + if (arena_is_auto(arena)) { + return; + } + extent_list_remove(&bin->slabs_full, slab); +} + +static void +arena_bin_reset(tsd_t *tsd, arena_t *arena, bin_t *bin) { + extent_t *slab; + + malloc_mutex_lock(tsd_tsdn(tsd), &bin->lock); + if (bin->slabcur != NULL) { + slab = bin->slabcur; + bin->slabcur = NULL; + malloc_mutex_unlock(tsd_tsdn(tsd), &bin->lock); + arena_slab_dalloc(tsd_tsdn(tsd), arena, slab); + malloc_mutex_lock(tsd_tsdn(tsd), &bin->lock); + } + while ((slab = extent_heap_remove_first(&bin->slabs_nonfull)) != NULL) { + malloc_mutex_unlock(tsd_tsdn(tsd), &bin->lock); + arena_slab_dalloc(tsd_tsdn(tsd), arena, slab); + malloc_mutex_lock(tsd_tsdn(tsd), &bin->lock); + } + for (slab = extent_list_first(&bin->slabs_full); slab != NULL; + slab = extent_list_first(&bin->slabs_full)) { + arena_bin_slabs_full_remove(arena, bin, slab); + malloc_mutex_unlock(tsd_tsdn(tsd), &bin->lock); + arena_slab_dalloc(tsd_tsdn(tsd), arena, slab); + malloc_mutex_lock(tsd_tsdn(tsd), &bin->lock); + } + if (config_stats) { + bin->stats.curregs = 0; + bin->stats.curslabs = 0; + } + malloc_mutex_unlock(tsd_tsdn(tsd), &bin->lock); +} + +void +arena_reset(tsd_t *tsd, arena_t *arena) { + /* + * Locking in this function is unintuitive. The caller guarantees that + * no concurrent operations are happening in this arena, but there are + * still reasons that some locking is necessary: + * + * - Some of the functions in the transitive closure of calls assume + * appropriate locks are held, and in some cases these locks are + * temporarily dropped to avoid lock order reversal or deadlock due to + * reentry. + * - mallctl("epoch", ...) may concurrently refresh stats. While + * strictly speaking this is a "concurrent operation", disallowing + * stats refreshes would impose an inconvenient burden. + */ + + /* Large allocations. */ + malloc_mutex_lock(tsd_tsdn(tsd), &arena->large_mtx); + + for (extent_t *extent = extent_list_first(&arena->large); extent != + NULL; extent = extent_list_first(&arena->large)) { + void *ptr = extent_base_get(extent); + size_t usize; + + malloc_mutex_unlock(tsd_tsdn(tsd), &arena->large_mtx); + alloc_ctx_t alloc_ctx; + rtree_ctx_t *rtree_ctx = tsd_rtree_ctx(tsd); + rtree_szind_slab_read(tsd_tsdn(tsd), &extents_rtree, rtree_ctx, + (uintptr_t)ptr, true, &alloc_ctx.szind, &alloc_ctx.slab); + assert(alloc_ctx.szind != SC_NSIZES); + + if (config_stats || (config_prof && opt_prof)) { + usize = sz_index2size(alloc_ctx.szind); + assert(usize == isalloc(tsd_tsdn(tsd), ptr)); + } + /* Remove large allocation from prof sample set. */ + if (config_prof && opt_prof) { + prof_free(tsd, ptr, usize, &alloc_ctx); + } + large_dalloc(tsd_tsdn(tsd), extent); + malloc_mutex_lock(tsd_tsdn(tsd), &arena->large_mtx); + } + malloc_mutex_unlock(tsd_tsdn(tsd), &arena->large_mtx); + + /* Bins. */ + for (unsigned i = 0; i < SC_NBINS; i++) { + for (unsigned j = 0; j < bin_infos[i].n_shards; j++) { + arena_bin_reset(tsd, arena, + &arena->bins[i].bin_shards[j]); + } + } + + atomic_store_zu(&arena->nactive, 0, ATOMIC_RELAXED); +} + +static void +arena_destroy_retained(tsdn_t *tsdn, arena_t *arena) { + /* + * Iterate over the retained extents and destroy them. This gives the + * extent allocator underlying the extent hooks an opportunity to unmap + * all retained memory without having to keep its own metadata + * structures. In practice, virtual memory for dss-allocated extents is + * leaked here, so best practice is to avoid dss for arenas to be + * destroyed, or provide custom extent hooks that track retained + * dss-based extents for later reuse. + */ + extent_hooks_t *extent_hooks = extent_hooks_get(arena); + extent_t *extent; + while ((extent = extents_evict(tsdn, arena, &extent_hooks, + &arena->extents_retained, 0)) != NULL) { + extent_destroy_wrapper(tsdn, arena, &extent_hooks, extent); + } +} + +void +arena_destroy(tsd_t *tsd, arena_t *arena) { + assert(base_ind_get(arena->base) >= narenas_auto); + assert(arena_nthreads_get(arena, false) == 0); + assert(arena_nthreads_get(arena, true) == 0); + + /* + * No allocations have occurred since arena_reset() was called. + * Furthermore, the caller (arena_i_destroy_ctl()) purged all cached + * extents, so only retained extents may remain. + */ + assert(extents_npages_get(&arena->extents_dirty) == 0); + assert(extents_npages_get(&arena->extents_muzzy) == 0); + + /* Deallocate retained memory. */ + arena_destroy_retained(tsd_tsdn(tsd), arena); + + /* + * Remove the arena pointer from the arenas array. We rely on the fact + * that there is no way for the application to get a dirty read from the + * arenas array unless there is an inherent race in the application + * involving access of an arena being concurrently destroyed. The + * application must synchronize knowledge of the arena's validity, so as + * long as we use an atomic write to update the arenas array, the + * application will get a clean read any time after it synchronizes + * knowledge that the arena is no longer valid. + */ + arena_set(base_ind_get(arena->base), NULL); + + /* + * Destroy the base allocator, which manages all metadata ever mapped by + * this arena. + */ + base_delete(tsd_tsdn(tsd), arena->base); +} + +static extent_t * +arena_slab_alloc_hard(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, const bin_info_t *bin_info, + szind_t szind) { + extent_t *slab; + bool zero, commit; + + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, 0); + + zero = false; + commit = true; + slab = extent_alloc_wrapper(tsdn, arena, r_extent_hooks, NULL, + bin_info->slab_size, 0, PAGE, true, szind, &zero, &commit); + + if (config_stats && slab != NULL) { + arena_stats_mapped_add(tsdn, &arena->stats, + bin_info->slab_size); + } + + return slab; +} + +static extent_t * +arena_slab_alloc(tsdn_t *tsdn, arena_t *arena, szind_t binind, unsigned binshard, + const bin_info_t *bin_info) { + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, 0); + + extent_hooks_t *extent_hooks = EXTENT_HOOKS_INITIALIZER; + szind_t szind = sz_size2index(bin_info->reg_size); + bool zero = false; + bool commit = true; + extent_t *slab = extents_alloc(tsdn, arena, &extent_hooks, + &arena->extents_dirty, NULL, bin_info->slab_size, 0, PAGE, true, + binind, &zero, &commit); + if (slab == NULL && arena_may_have_muzzy(arena)) { + slab = extents_alloc(tsdn, arena, &extent_hooks, + &arena->extents_muzzy, NULL, bin_info->slab_size, 0, PAGE, + true, binind, &zero, &commit); + } + if (slab == NULL) { + slab = arena_slab_alloc_hard(tsdn, arena, &extent_hooks, + bin_info, szind); + if (slab == NULL) { + return NULL; + } + } + assert(extent_slab_get(slab)); + + /* Initialize slab internals. */ + arena_slab_data_t *slab_data = extent_slab_data_get(slab); + extent_nfree_binshard_set(slab, bin_info->nregs, binshard); + bitmap_init(slab_data->bitmap, &bin_info->bitmap_info, false); + + arena_nactive_add(arena, extent_size_get(slab) >> LG_PAGE); + + return slab; +} + +static extent_t * +arena_bin_nonfull_slab_get(tsdn_t *tsdn, arena_t *arena, bin_t *bin, + szind_t binind, unsigned binshard) { + extent_t *slab; + const bin_info_t *bin_info; + + /* Look for a usable slab. */ + slab = arena_bin_slabs_nonfull_tryget(bin); + if (slab != NULL) { + return slab; + } + /* No existing slabs have any space available. */ + + bin_info = &bin_infos[binind]; + + /* Allocate a new slab. */ + malloc_mutex_unlock(tsdn, &bin->lock); + /******************************/ + slab = arena_slab_alloc(tsdn, arena, binind, binshard, bin_info); + /********************************/ + malloc_mutex_lock(tsdn, &bin->lock); + if (slab != NULL) { + if (config_stats) { + bin->stats.nslabs++; + bin->stats.curslabs++; + } + return slab; + } + + /* + * arena_slab_alloc() failed, but another thread may have made + * sufficient memory available while this one dropped bin->lock above, + * so search one more time. + */ + slab = arena_bin_slabs_nonfull_tryget(bin); + if (slab != NULL) { + return slab; + } + + return NULL; +} + +/* Re-fill bin->slabcur, then call arena_slab_reg_alloc(). */ +static void * +arena_bin_malloc_hard(tsdn_t *tsdn, arena_t *arena, bin_t *bin, + szind_t binind, unsigned binshard) { + const bin_info_t *bin_info; + extent_t *slab; + + bin_info = &bin_infos[binind]; + if (!arena_is_auto(arena) && bin->slabcur != NULL) { + arena_bin_slabs_full_insert(arena, bin, bin->slabcur); + bin->slabcur = NULL; + } + slab = arena_bin_nonfull_slab_get(tsdn, arena, bin, binind, binshard); + if (bin->slabcur != NULL) { + /* + * Another thread updated slabcur while this one ran without the + * bin lock in arena_bin_nonfull_slab_get(). + */ + if (extent_nfree_get(bin->slabcur) > 0) { + void *ret = arena_slab_reg_alloc(bin->slabcur, + bin_info); + if (slab != NULL) { + /* + * arena_slab_alloc() may have allocated slab, + * or it may have been pulled from + * slabs_nonfull. Therefore it is unsafe to + * make any assumptions about how slab has + * previously been used, and + * arena_bin_lower_slab() must be called, as if + * a region were just deallocated from the slab. + */ + if (extent_nfree_get(slab) == bin_info->nregs) { + arena_dalloc_bin_slab(tsdn, arena, slab, + bin); + } else { + arena_bin_lower_slab(tsdn, arena, slab, + bin); + } + } + return ret; + } + + arena_bin_slabs_full_insert(arena, bin, bin->slabcur); + bin->slabcur = NULL; + } + + if (slab == NULL) { + return NULL; + } + bin->slabcur = slab; + + assert(extent_nfree_get(bin->slabcur) > 0); + + return arena_slab_reg_alloc(slab, bin_info); +} + +/* Choose a bin shard and return the locked bin. */ +bin_t * +arena_bin_choose_lock(tsdn_t *tsdn, arena_t *arena, szind_t binind, + unsigned *binshard) { + bin_t *bin; + if (tsdn_null(tsdn) || tsd_arena_get(tsdn_tsd(tsdn)) == NULL) { + *binshard = 0; + } else { + *binshard = tsd_binshardsp_get(tsdn_tsd(tsdn))->binshard[binind]; + } + assert(*binshard < bin_infos[binind].n_shards); + bin = &arena->bins[binind].bin_shards[*binshard]; + malloc_mutex_lock(tsdn, &bin->lock); + + return bin; +} + +void +arena_tcache_fill_small(tsdn_t *tsdn, arena_t *arena, tcache_t *tcache, + cache_bin_t *tbin, szind_t binind, uint64_t prof_accumbytes) { + unsigned i, nfill, cnt; + + assert(tbin->ncached == 0); + + if (config_prof && arena_prof_accum(tsdn, arena, prof_accumbytes)) { + prof_idump(tsdn); + } + + unsigned binshard; + bin_t *bin = arena_bin_choose_lock(tsdn, arena, binind, &binshard); + + for (i = 0, nfill = (tcache_bin_info[binind].ncached_max >> + tcache->lg_fill_div[binind]); i < nfill; i += cnt) { + extent_t *slab; + if ((slab = bin->slabcur) != NULL && extent_nfree_get(slab) > + 0) { + unsigned tofill = nfill - i; + cnt = tofill < extent_nfree_get(slab) ? + tofill : extent_nfree_get(slab); + arena_slab_reg_alloc_batch( + slab, &bin_infos[binind], cnt, + tbin->avail - nfill + i); + } else { + cnt = 1; + void *ptr = arena_bin_malloc_hard(tsdn, arena, bin, + binind, binshard); + /* + * OOM. tbin->avail isn't yet filled down to its first + * element, so the successful allocations (if any) must + * be moved just before tbin->avail before bailing out. + */ + if (ptr == NULL) { + if (i > 0) { + memmove(tbin->avail - i, + tbin->avail - nfill, + i * sizeof(void *)); + } + break; + } + /* Insert such that low regions get used first. */ + *(tbin->avail - nfill + i) = ptr; + } + if (config_fill && unlikely(opt_junk_alloc)) { + for (unsigned j = 0; j < cnt; j++) { + void* ptr = *(tbin->avail - nfill + i + j); + arena_alloc_junk_small(ptr, &bin_infos[binind], + true); + } + } + } + if (config_stats) { + bin->stats.nmalloc += i; + bin->stats.nrequests += tbin->tstats.nrequests; + bin->stats.curregs += i; + bin->stats.nfills++; + tbin->tstats.nrequests = 0; + } + malloc_mutex_unlock(tsdn, &bin->lock); + tbin->ncached = i; + arena_decay_tick(tsdn, arena); +} + +void +arena_alloc_junk_small(void *ptr, const bin_info_t *bin_info, bool zero) { + if (!zero) { + memset(ptr, JEMALLOC_ALLOC_JUNK, bin_info->reg_size); + } +} + +static void +arena_dalloc_junk_small_impl(void *ptr, const bin_info_t *bin_info) { + memset(ptr, JEMALLOC_FREE_JUNK, bin_info->reg_size); +} +arena_dalloc_junk_small_t *JET_MUTABLE arena_dalloc_junk_small = + arena_dalloc_junk_small_impl; + +static void * +arena_malloc_small(tsdn_t *tsdn, arena_t *arena, szind_t binind, bool zero) { + void *ret; + bin_t *bin; + size_t usize; + extent_t *slab; + + assert(binind < SC_NBINS); + usize = sz_index2size(binind); + unsigned binshard; + bin = arena_bin_choose_lock(tsdn, arena, binind, &binshard); + + if ((slab = bin->slabcur) != NULL && extent_nfree_get(slab) > 0) { + ret = arena_slab_reg_alloc(slab, &bin_infos[binind]); + } else { + ret = arena_bin_malloc_hard(tsdn, arena, bin, binind, binshard); + } + + if (ret == NULL) { + malloc_mutex_unlock(tsdn, &bin->lock); + return NULL; + } + + if (config_stats) { + bin->stats.nmalloc++; + bin->stats.nrequests++; + bin->stats.curregs++; + } + malloc_mutex_unlock(tsdn, &bin->lock); + if (config_prof && arena_prof_accum(tsdn, arena, usize)) { + prof_idump(tsdn); + } + + if (!zero) { + if (config_fill) { + if (unlikely(opt_junk_alloc)) { + arena_alloc_junk_small(ret, + &bin_infos[binind], false); + } else if (unlikely(opt_zero)) { + memset(ret, 0, usize); + } + } + } else { + if (config_fill && unlikely(opt_junk_alloc)) { + arena_alloc_junk_small(ret, &bin_infos[binind], + true); + } + memset(ret, 0, usize); + } + + arena_decay_tick(tsdn, arena); + return ret; +} + +void * +arena_malloc_hard(tsdn_t *tsdn, arena_t *arena, size_t size, szind_t ind, + bool zero) { + assert(!tsdn_null(tsdn) || arena != NULL); + + if (likely(!tsdn_null(tsdn))) { + arena = arena_choose_maybe_huge(tsdn_tsd(tsdn), arena, size); + } + if (unlikely(arena == NULL)) { + return NULL; + } + + if (likely(size <= SC_SMALL_MAXCLASS)) { + return arena_malloc_small(tsdn, arena, ind, zero); + } + return large_malloc(tsdn, arena, sz_index2size(ind), zero); +} + +void * +arena_palloc(tsdn_t *tsdn, arena_t *arena, size_t usize, size_t alignment, + bool zero, tcache_t *tcache) { + void *ret; + + if (usize <= SC_SMALL_MAXCLASS + && (alignment < PAGE + || (alignment == PAGE && (usize & PAGE_MASK) == 0))) { + /* Small; alignment doesn't require special slab placement. */ + ret = arena_malloc(tsdn, arena, usize, sz_size2index(usize), + zero, tcache, true); + } else { + if (likely(alignment <= CACHELINE)) { + ret = large_malloc(tsdn, arena, usize, zero); + } else { + ret = large_palloc(tsdn, arena, usize, alignment, zero); + } + } + return ret; +} + +void +arena_prof_promote(tsdn_t *tsdn, void *ptr, size_t usize) { + cassert(config_prof); + assert(ptr != NULL); + assert(isalloc(tsdn, ptr) == SC_LARGE_MINCLASS); + assert(usize <= SC_SMALL_MAXCLASS); + + if (config_opt_safety_checks) { + safety_check_set_redzone(ptr, usize, SC_LARGE_MINCLASS); + } + + rtree_ctx_t rtree_ctx_fallback; + rtree_ctx_t *rtree_ctx = tsdn_rtree_ctx(tsdn, &rtree_ctx_fallback); + + extent_t *extent = rtree_extent_read(tsdn, &extents_rtree, rtree_ctx, + (uintptr_t)ptr, true); + arena_t *arena = extent_arena_get(extent); + + szind_t szind = sz_size2index(usize); + extent_szind_set(extent, szind); + rtree_szind_slab_update(tsdn, &extents_rtree, rtree_ctx, (uintptr_t)ptr, + szind, false); + + prof_accum_cancel(tsdn, &arena->prof_accum, usize); + + assert(isalloc(tsdn, ptr) == usize); +} + +static size_t +arena_prof_demote(tsdn_t *tsdn, extent_t *extent, const void *ptr) { + cassert(config_prof); + assert(ptr != NULL); + + extent_szind_set(extent, SC_NBINS); + rtree_ctx_t rtree_ctx_fallback; + rtree_ctx_t *rtree_ctx = tsdn_rtree_ctx(tsdn, &rtree_ctx_fallback); + rtree_szind_slab_update(tsdn, &extents_rtree, rtree_ctx, (uintptr_t)ptr, + SC_NBINS, false); + + assert(isalloc(tsdn, ptr) == SC_LARGE_MINCLASS); + + return SC_LARGE_MINCLASS; +} + +void +arena_dalloc_promoted(tsdn_t *tsdn, void *ptr, tcache_t *tcache, + bool slow_path) { + cassert(config_prof); + assert(opt_prof); + + extent_t *extent = iealloc(tsdn, ptr); + size_t usize = extent_usize_get(extent); + size_t bumped_usize = arena_prof_demote(tsdn, extent, ptr); + if (config_opt_safety_checks && usize < SC_LARGE_MINCLASS) { + /* + * Currently, we only do redzoning for small sampled + * allocations. + */ + assert(bumped_usize == SC_LARGE_MINCLASS); + safety_check_verify_redzone(ptr, usize, bumped_usize); + } + if (bumped_usize <= tcache_maxclass && tcache != NULL) { + tcache_dalloc_large(tsdn_tsd(tsdn), tcache, ptr, + sz_size2index(bumped_usize), slow_path); + } else { + large_dalloc(tsdn, extent); + } +} + +static void +arena_dissociate_bin_slab(arena_t *arena, extent_t *slab, bin_t *bin) { + /* Dissociate slab from bin. */ + if (slab == bin->slabcur) { + bin->slabcur = NULL; + } else { + szind_t binind = extent_szind_get(slab); + const bin_info_t *bin_info = &bin_infos[binind]; + + /* + * The following block's conditional is necessary because if the + * slab only contains one region, then it never gets inserted + * into the non-full slabs heap. + */ + if (bin_info->nregs == 1) { + arena_bin_slabs_full_remove(arena, bin, slab); + } else { + arena_bin_slabs_nonfull_remove(bin, slab); + } + } +} + +static void +arena_dalloc_bin_slab(tsdn_t *tsdn, arena_t *arena, extent_t *slab, + bin_t *bin) { + assert(slab != bin->slabcur); + + malloc_mutex_unlock(tsdn, &bin->lock); + /******************************/ + arena_slab_dalloc(tsdn, arena, slab); + /****************************/ + malloc_mutex_lock(tsdn, &bin->lock); + if (config_stats) { + bin->stats.curslabs--; + } +} + +static void +arena_bin_lower_slab(tsdn_t *tsdn, arena_t *arena, extent_t *slab, + bin_t *bin) { + assert(extent_nfree_get(slab) > 0); + + /* + * Make sure that if bin->slabcur is non-NULL, it refers to the + * oldest/lowest non-full slab. It is okay to NULL slabcur out rather + * than proactively keeping it pointing at the oldest/lowest non-full + * slab. + */ + if (bin->slabcur != NULL && extent_snad_comp(bin->slabcur, slab) > 0) { + /* Switch slabcur. */ + if (extent_nfree_get(bin->slabcur) > 0) { + arena_bin_slabs_nonfull_insert(bin, bin->slabcur); + } else { + arena_bin_slabs_full_insert(arena, bin, bin->slabcur); + } + bin->slabcur = slab; + if (config_stats) { + bin->stats.reslabs++; + } + } else { + arena_bin_slabs_nonfull_insert(bin, slab); + } +} + +static void +arena_dalloc_bin_locked_impl(tsdn_t *tsdn, arena_t *arena, bin_t *bin, + szind_t binind, extent_t *slab, void *ptr, bool junked) { + arena_slab_data_t *slab_data = extent_slab_data_get(slab); + const bin_info_t *bin_info = &bin_infos[binind]; + + if (!junked && config_fill && unlikely(opt_junk_free)) { + arena_dalloc_junk_small(ptr, bin_info); + } + + arena_slab_reg_dalloc(slab, slab_data, ptr); + unsigned nfree = extent_nfree_get(slab); + if (nfree == bin_info->nregs) { + arena_dissociate_bin_slab(arena, slab, bin); + arena_dalloc_bin_slab(tsdn, arena, slab, bin); + } else if (nfree == 1 && slab != bin->slabcur) { + arena_bin_slabs_full_remove(arena, bin, slab); + arena_bin_lower_slab(tsdn, arena, slab, bin); + } + + if (config_stats) { + bin->stats.ndalloc++; + bin->stats.curregs--; + } +} + +void +arena_dalloc_bin_junked_locked(tsdn_t *tsdn, arena_t *arena, bin_t *bin, + szind_t binind, extent_t *extent, void *ptr) { + arena_dalloc_bin_locked_impl(tsdn, arena, bin, binind, extent, ptr, + true); +} + +static void +arena_dalloc_bin(tsdn_t *tsdn, arena_t *arena, extent_t *extent, void *ptr) { + szind_t binind = extent_szind_get(extent); + unsigned binshard = extent_binshard_get(extent); + bin_t *bin = &arena->bins[binind].bin_shards[binshard]; + + malloc_mutex_lock(tsdn, &bin->lock); + arena_dalloc_bin_locked_impl(tsdn, arena, bin, binind, extent, ptr, + false); + malloc_mutex_unlock(tsdn, &bin->lock); +} + +void +arena_dalloc_small(tsdn_t *tsdn, void *ptr) { + extent_t *extent = iealloc(tsdn, ptr); + arena_t *arena = extent_arena_get(extent); + + arena_dalloc_bin(tsdn, arena, extent, ptr); + arena_decay_tick(tsdn, arena); +} + +bool +arena_ralloc_no_move(tsdn_t *tsdn, void *ptr, size_t oldsize, size_t size, + size_t extra, bool zero, size_t *newsize) { + bool ret; + /* Calls with non-zero extra had to clamp extra. */ + assert(extra == 0 || size + extra <= SC_LARGE_MAXCLASS); + + extent_t *extent = iealloc(tsdn, ptr); + if (unlikely(size > SC_LARGE_MAXCLASS)) { + ret = true; + goto done; + } + + size_t usize_min = sz_s2u(size); + size_t usize_max = sz_s2u(size + extra); + if (likely(oldsize <= SC_SMALL_MAXCLASS && usize_min + <= SC_SMALL_MAXCLASS)) { + /* + * Avoid moving the allocation if the size class can be left the + * same. + */ + assert(bin_infos[sz_size2index(oldsize)].reg_size == + oldsize); + if ((usize_max > SC_SMALL_MAXCLASS + || sz_size2index(usize_max) != sz_size2index(oldsize)) + && (size > oldsize || usize_max < oldsize)) { + ret = true; + goto done; + } + + arena_decay_tick(tsdn, extent_arena_get(extent)); + ret = false; + } else if (oldsize >= SC_LARGE_MINCLASS + && usize_max >= SC_LARGE_MINCLASS) { + ret = large_ralloc_no_move(tsdn, extent, usize_min, usize_max, + zero); + } else { + ret = true; + } +done: + assert(extent == iealloc(tsdn, ptr)); + *newsize = extent_usize_get(extent); + + return ret; +} + +static void * +arena_ralloc_move_helper(tsdn_t *tsdn, arena_t *arena, size_t usize, + size_t alignment, bool zero, tcache_t *tcache) { + if (alignment == 0) { + return arena_malloc(tsdn, arena, usize, sz_size2index(usize), + zero, tcache, true); + } + usize = sz_sa2u(usize, alignment); + if (unlikely(usize == 0 || usize > SC_LARGE_MAXCLASS)) { + return NULL; + } + return ipalloct(tsdn, usize, alignment, zero, tcache, arena); +} + +void * +arena_ralloc(tsdn_t *tsdn, arena_t *arena, void *ptr, size_t oldsize, + size_t size, size_t alignment, bool zero, tcache_t *tcache, + hook_ralloc_args_t *hook_args) { + size_t usize = sz_s2u(size); + if (unlikely(usize == 0 || size > SC_LARGE_MAXCLASS)) { + return NULL; + } + + if (likely(usize <= SC_SMALL_MAXCLASS)) { + /* Try to avoid moving the allocation. */ + UNUSED size_t newsize; + if (!arena_ralloc_no_move(tsdn, ptr, oldsize, usize, 0, zero, + &newsize)) { + hook_invoke_expand(hook_args->is_realloc + ? hook_expand_realloc : hook_expand_rallocx, + ptr, oldsize, usize, (uintptr_t)ptr, + hook_args->args); + return ptr; + } + } + + if (oldsize >= SC_LARGE_MINCLASS + && usize >= SC_LARGE_MINCLASS) { + return large_ralloc(tsdn, arena, ptr, usize, + alignment, zero, tcache, hook_args); + } + + /* + * size and oldsize are different enough that we need to move the + * object. In that case, fall back to allocating new space and copying. + */ + void *ret = arena_ralloc_move_helper(tsdn, arena, usize, alignment, + zero, tcache); + if (ret == NULL) { + return NULL; + } + + hook_invoke_alloc(hook_args->is_realloc + ? hook_alloc_realloc : hook_alloc_rallocx, ret, (uintptr_t)ret, + hook_args->args); + hook_invoke_dalloc(hook_args->is_realloc + ? hook_dalloc_realloc : hook_dalloc_rallocx, ptr, hook_args->args); + + /* + * Junk/zero-filling were already done by + * ipalloc()/arena_malloc(). + */ + size_t copysize = (usize < oldsize) ? usize : oldsize; + memcpy(ret, ptr, copysize); + isdalloct(tsdn, ptr, oldsize, tcache, NULL, true); + return ret; +} + +dss_prec_t +arena_dss_prec_get(arena_t *arena) { + return (dss_prec_t)atomic_load_u(&arena->dss_prec, ATOMIC_ACQUIRE); +} + +bool +arena_dss_prec_set(arena_t *arena, dss_prec_t dss_prec) { + if (!have_dss) { + return (dss_prec != dss_prec_disabled); + } + atomic_store_u(&arena->dss_prec, (unsigned)dss_prec, ATOMIC_RELEASE); + return false; +} + +ssize_t +arena_dirty_decay_ms_default_get(void) { + return atomic_load_zd(&dirty_decay_ms_default, ATOMIC_RELAXED); +} + +bool +arena_dirty_decay_ms_default_set(ssize_t decay_ms) { + if (!arena_decay_ms_valid(decay_ms)) { + return true; + } + atomic_store_zd(&dirty_decay_ms_default, decay_ms, ATOMIC_RELAXED); + return false; +} + +ssize_t +arena_muzzy_decay_ms_default_get(void) { + return atomic_load_zd(&muzzy_decay_ms_default, ATOMIC_RELAXED); +} + +bool +arena_muzzy_decay_ms_default_set(ssize_t decay_ms) { + if (!arena_decay_ms_valid(decay_ms)) { + return true; + } + atomic_store_zd(&muzzy_decay_ms_default, decay_ms, ATOMIC_RELAXED); + return false; +} + +bool +arena_retain_grow_limit_get_set(tsd_t *tsd, arena_t *arena, size_t *old_limit, + size_t *new_limit) { + assert(opt_retain); + + pszind_t new_ind JEMALLOC_CC_SILENCE_INIT(0); + if (new_limit != NULL) { + size_t limit = *new_limit; + /* Grow no more than the new limit. */ + if ((new_ind = sz_psz2ind(limit + 1) - 1) >= SC_NPSIZES) { + return true; + } + } + + malloc_mutex_lock(tsd_tsdn(tsd), &arena->extent_grow_mtx); + if (old_limit != NULL) { + *old_limit = sz_pind2sz(arena->retain_grow_limit); + } + if (new_limit != NULL) { + arena->retain_grow_limit = new_ind; + } + malloc_mutex_unlock(tsd_tsdn(tsd), &arena->extent_grow_mtx); + + return false; +} + +unsigned +arena_nthreads_get(arena_t *arena, bool internal) { + return atomic_load_u(&arena->nthreads[internal], ATOMIC_RELAXED); +} + +void +arena_nthreads_inc(arena_t *arena, bool internal) { + atomic_fetch_add_u(&arena->nthreads[internal], 1, ATOMIC_RELAXED); +} + +void +arena_nthreads_dec(arena_t *arena, bool internal) { + atomic_fetch_sub_u(&arena->nthreads[internal], 1, ATOMIC_RELAXED); +} + +size_t +arena_extent_sn_next(arena_t *arena) { + return atomic_fetch_add_zu(&arena->extent_sn_next, 1, ATOMIC_RELAXED); +} + +arena_t * +arena_new(tsdn_t *tsdn, unsigned ind, extent_hooks_t *extent_hooks) { + arena_t *arena; + base_t *base; + unsigned i; + + if (ind == 0) { + base = b0get(); + } else { + base = base_new(tsdn, ind, extent_hooks); + if (base == NULL) { + return NULL; + } + } + + unsigned nbins_total = 0; + for (i = 0; i < SC_NBINS; i++) { + nbins_total += bin_infos[i].n_shards; + } + size_t arena_size = sizeof(arena_t) + sizeof(bin_t) * nbins_total; + arena = (arena_t *)base_alloc(tsdn, base, arena_size, CACHELINE); + if (arena == NULL) { + goto label_error; + } + + atomic_store_u(&arena->nthreads[0], 0, ATOMIC_RELAXED); + atomic_store_u(&arena->nthreads[1], 0, ATOMIC_RELAXED); + arena->last_thd = NULL; + + if (config_stats) { + if (arena_stats_init(tsdn, &arena->stats)) { + goto label_error; + } + + ql_new(&arena->tcache_ql); + ql_new(&arena->cache_bin_array_descriptor_ql); + if (malloc_mutex_init(&arena->tcache_ql_mtx, "tcache_ql", + WITNESS_RANK_TCACHE_QL, malloc_mutex_rank_exclusive)) { + goto label_error; + } + } + + if (config_prof) { + if (prof_accum_init(tsdn, &arena->prof_accum)) { + goto label_error; + } + } + + if (config_cache_oblivious) { + /* + * A nondeterministic seed based on the address of arena reduces + * the likelihood of lockstep non-uniform cache index + * utilization among identical concurrent processes, but at the + * cost of test repeatability. For debug builds, instead use a + * deterministic seed. + */ + atomic_store_zu(&arena->offset_state, config_debug ? ind : + (size_t)(uintptr_t)arena, ATOMIC_RELAXED); + } + + atomic_store_zu(&arena->extent_sn_next, 0, ATOMIC_RELAXED); + + atomic_store_u(&arena->dss_prec, (unsigned)extent_dss_prec_get(), + ATOMIC_RELAXED); + + atomic_store_zu(&arena->nactive, 0, ATOMIC_RELAXED); + + extent_list_init(&arena->large); + if (malloc_mutex_init(&arena->large_mtx, "arena_large", + WITNESS_RANK_ARENA_LARGE, malloc_mutex_rank_exclusive)) { + goto label_error; + } + + /* + * Delay coalescing for dirty extents despite the disruptive effect on + * memory layout for best-fit extent allocation, since cached extents + * are likely to be reused soon after deallocation, and the cost of + * merging/splitting extents is non-trivial. + */ + if (extents_init(tsdn, &arena->extents_dirty, extent_state_dirty, + true)) { + goto label_error; + } + /* + * Coalesce muzzy extents immediately, because operations on them are in + * the critical path much less often than for dirty extents. + */ + if (extents_init(tsdn, &arena->extents_muzzy, extent_state_muzzy, + false)) { + goto label_error; + } + /* + * Coalesce retained extents immediately, in part because they will + * never be evicted (and therefore there's no opportunity for delayed + * coalescing), but also because operations on retained extents are not + * in the critical path. + */ + if (extents_init(tsdn, &arena->extents_retained, extent_state_retained, + false)) { + goto label_error; + } + + if (arena_decay_init(&arena->decay_dirty, + arena_dirty_decay_ms_default_get(), &arena->stats.decay_dirty)) { + goto label_error; + } + if (arena_decay_init(&arena->decay_muzzy, + arena_muzzy_decay_ms_default_get(), &arena->stats.decay_muzzy)) { + goto label_error; + } + + arena->extent_grow_next = sz_psz2ind(HUGEPAGE); + arena->retain_grow_limit = sz_psz2ind(SC_LARGE_MAXCLASS); + if (malloc_mutex_init(&arena->extent_grow_mtx, "extent_grow", + WITNESS_RANK_EXTENT_GROW, malloc_mutex_rank_exclusive)) { + goto label_error; + } + + extent_avail_new(&arena->extent_avail); + if (malloc_mutex_init(&arena->extent_avail_mtx, "extent_avail", + WITNESS_RANK_EXTENT_AVAIL, malloc_mutex_rank_exclusive)) { + goto label_error; + } + + /* Initialize bins. */ + uintptr_t bin_addr = (uintptr_t)arena + sizeof(arena_t); + atomic_store_u(&arena->binshard_next, 0, ATOMIC_RELEASE); + for (i = 0; i < SC_NBINS; i++) { + unsigned nshards = bin_infos[i].n_shards; + arena->bins[i].bin_shards = (bin_t *)bin_addr; + bin_addr += nshards * sizeof(bin_t); + for (unsigned j = 0; j < nshards; j++) { + bool err = bin_init(&arena->bins[i].bin_shards[j]); + if (err) { + goto label_error; + } + } + } + assert(bin_addr == (uintptr_t)arena + arena_size); + + arena->base = base; + /* Set arena before creating background threads. */ + arena_set(ind, arena); + + nstime_init(&arena->create_time, 0); + nstime_update(&arena->create_time); + + /* We don't support reentrancy for arena 0 bootstrapping. */ + if (ind != 0) { + /* + * If we're here, then arena 0 already exists, so bootstrapping + * is done enough that we should have tsd. + */ + assert(!tsdn_null(tsdn)); + pre_reentrancy(tsdn_tsd(tsdn), arena); + if (test_hooks_arena_new_hook) { + test_hooks_arena_new_hook(); + } + post_reentrancy(tsdn_tsd(tsdn)); + } + + return arena; +label_error: + if (ind != 0) { + base_delete(tsdn, base); + } + return NULL; +} + +arena_t * +arena_choose_huge(tsd_t *tsd) { + /* huge_arena_ind can be 0 during init (will use a0). */ + if (huge_arena_ind == 0) { + assert(!malloc_initialized()); + } + + arena_t *huge_arena = arena_get(tsd_tsdn(tsd), huge_arena_ind, false); + if (huge_arena == NULL) { + /* Create the huge arena on demand. */ + assert(huge_arena_ind != 0); + huge_arena = arena_get(tsd_tsdn(tsd), huge_arena_ind, true); + if (huge_arena == NULL) { + return NULL; + } + /* + * Purge eagerly for huge allocations, because: 1) number of + * huge allocations is usually small, which means ticker based + * decay is not reliable; and 2) less immediate reuse is + * expected for huge allocations. + */ + if (arena_dirty_decay_ms_default_get() > 0) { + arena_dirty_decay_ms_set(tsd_tsdn(tsd), huge_arena, 0); + } + if (arena_muzzy_decay_ms_default_get() > 0) { + arena_muzzy_decay_ms_set(tsd_tsdn(tsd), huge_arena, 0); + } + } + + return huge_arena; +} + +bool +arena_init_huge(void) { + bool huge_enabled; + + /* The threshold should be large size class. */ + if (opt_oversize_threshold > SC_LARGE_MAXCLASS || + opt_oversize_threshold < SC_LARGE_MINCLASS) { + opt_oversize_threshold = 0; + oversize_threshold = SC_LARGE_MAXCLASS + PAGE; + huge_enabled = false; + } else { + /* Reserve the index for the huge arena. */ + huge_arena_ind = narenas_total_get(); + oversize_threshold = opt_oversize_threshold; + huge_enabled = true; + } + + return huge_enabled; +} + +bool +arena_is_huge(unsigned arena_ind) { + if (huge_arena_ind == 0) { + return false; + } + return (arena_ind == huge_arena_ind); +} + +void +arena_boot(sc_data_t *sc_data) { + arena_dirty_decay_ms_default_set(opt_dirty_decay_ms); + arena_muzzy_decay_ms_default_set(opt_muzzy_decay_ms); + for (unsigned i = 0; i < SC_NBINS; i++) { + sc_t *sc = &sc_data->sc[i]; + div_init(&arena_binind_div_info[i], + (1U << sc->lg_base) + (sc->ndelta << sc->lg_delta)); + } +} + +void +arena_prefork0(tsdn_t *tsdn, arena_t *arena) { + malloc_mutex_prefork(tsdn, &arena->decay_dirty.mtx); + malloc_mutex_prefork(tsdn, &arena->decay_muzzy.mtx); +} + +void +arena_prefork1(tsdn_t *tsdn, arena_t *arena) { + if (config_stats) { + malloc_mutex_prefork(tsdn, &arena->tcache_ql_mtx); + } +} + +void +arena_prefork2(tsdn_t *tsdn, arena_t *arena) { + malloc_mutex_prefork(tsdn, &arena->extent_grow_mtx); +} + +void +arena_prefork3(tsdn_t *tsdn, arena_t *arena) { + extents_prefork(tsdn, &arena->extents_dirty); + extents_prefork(tsdn, &arena->extents_muzzy); + extents_prefork(tsdn, &arena->extents_retained); +} + +void +arena_prefork4(tsdn_t *tsdn, arena_t *arena) { + malloc_mutex_prefork(tsdn, &arena->extent_avail_mtx); +} + +void +arena_prefork5(tsdn_t *tsdn, arena_t *arena) { + base_prefork(tsdn, arena->base); +} + +void +arena_prefork6(tsdn_t *tsdn, arena_t *arena) { + malloc_mutex_prefork(tsdn, &arena->large_mtx); +} + +void +arena_prefork7(tsdn_t *tsdn, arena_t *arena) { + for (unsigned i = 0; i < SC_NBINS; i++) { + for (unsigned j = 0; j < bin_infos[i].n_shards; j++) { + bin_prefork(tsdn, &arena->bins[i].bin_shards[j]); + } + } +} + +void +arena_postfork_parent(tsdn_t *tsdn, arena_t *arena) { + unsigned i; + + for (i = 0; i < SC_NBINS; i++) { + for (unsigned j = 0; j < bin_infos[i].n_shards; j++) { + bin_postfork_parent(tsdn, + &arena->bins[i].bin_shards[j]); + } + } + malloc_mutex_postfork_parent(tsdn, &arena->large_mtx); + base_postfork_parent(tsdn, arena->base); + malloc_mutex_postfork_parent(tsdn, &arena->extent_avail_mtx); + extents_postfork_parent(tsdn, &arena->extents_dirty); + extents_postfork_parent(tsdn, &arena->extents_muzzy); + extents_postfork_parent(tsdn, &arena->extents_retained); + malloc_mutex_postfork_parent(tsdn, &arena->extent_grow_mtx); + malloc_mutex_postfork_parent(tsdn, &arena->decay_dirty.mtx); + malloc_mutex_postfork_parent(tsdn, &arena->decay_muzzy.mtx); + if (config_stats) { + malloc_mutex_postfork_parent(tsdn, &arena->tcache_ql_mtx); + } +} + +void +arena_postfork_child(tsdn_t *tsdn, arena_t *arena) { + unsigned i; + + atomic_store_u(&arena->nthreads[0], 0, ATOMIC_RELAXED); + atomic_store_u(&arena->nthreads[1], 0, ATOMIC_RELAXED); + if (tsd_arena_get(tsdn_tsd(tsdn)) == arena) { + arena_nthreads_inc(arena, false); + } + if (tsd_iarena_get(tsdn_tsd(tsdn)) == arena) { + arena_nthreads_inc(arena, true); + } + if (config_stats) { + ql_new(&arena->tcache_ql); + ql_new(&arena->cache_bin_array_descriptor_ql); + tcache_t *tcache = tcache_get(tsdn_tsd(tsdn)); + if (tcache != NULL && tcache->arena == arena) { + ql_elm_new(tcache, link); + ql_tail_insert(&arena->tcache_ql, tcache, link); + cache_bin_array_descriptor_init( + &tcache->cache_bin_array_descriptor, + tcache->bins_small, tcache->bins_large); + ql_tail_insert(&arena->cache_bin_array_descriptor_ql, + &tcache->cache_bin_array_descriptor, link); + } + } + + for (i = 0; i < SC_NBINS; i++) { + for (unsigned j = 0; j < bin_infos[i].n_shards; j++) { + bin_postfork_child(tsdn, &arena->bins[i].bin_shards[j]); + } + } + malloc_mutex_postfork_child(tsdn, &arena->large_mtx); + base_postfork_child(tsdn, arena->base); + malloc_mutex_postfork_child(tsdn, &arena->extent_avail_mtx); + extents_postfork_child(tsdn, &arena->extents_dirty); + extents_postfork_child(tsdn, &arena->extents_muzzy); + extents_postfork_child(tsdn, &arena->extents_retained); + malloc_mutex_postfork_child(tsdn, &arena->extent_grow_mtx); + malloc_mutex_postfork_child(tsdn, &arena->decay_dirty.mtx); + malloc_mutex_postfork_child(tsdn, &arena->decay_muzzy.mtx); + if (config_stats) { + malloc_mutex_postfork_child(tsdn, &arena->tcache_ql_mtx); + } +} diff --git a/deps/jemalloc/src/background_thread.c b/deps/jemalloc/src/background_thread.c new file mode 100644 index 000000000..57b9b256b --- /dev/null +++ b/deps/jemalloc/src/background_thread.c @@ -0,0 +1,939 @@ +#define JEMALLOC_BACKGROUND_THREAD_C_ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/assert.h" + +JEMALLOC_DIAGNOSTIC_DISABLE_SPURIOUS + +/******************************************************************************/ +/* Data. */ + +/* This option should be opt-in only. */ +#define BACKGROUND_THREAD_DEFAULT false +/* Read-only after initialization. */ +bool opt_background_thread = BACKGROUND_THREAD_DEFAULT; +size_t opt_max_background_threads = MAX_BACKGROUND_THREAD_LIMIT + 1; + +/* Used for thread creation, termination and stats. */ +malloc_mutex_t background_thread_lock; +/* Indicates global state. Atomic because decay reads this w/o locking. */ +atomic_b_t background_thread_enabled_state; +size_t n_background_threads; +size_t max_background_threads; +/* Thread info per-index. */ +background_thread_info_t *background_thread_info; + +/******************************************************************************/ + +#ifdef JEMALLOC_PTHREAD_CREATE_WRAPPER + +static int (*pthread_create_fptr)(pthread_t *__restrict, const pthread_attr_t *, + void *(*)(void *), void *__restrict); + +static void +pthread_create_wrapper_init(void) { +#ifdef JEMALLOC_LAZY_LOCK + if (!isthreaded) { + isthreaded = true; + } +#endif +} + +int +pthread_create_wrapper(pthread_t *__restrict thread, const pthread_attr_t *attr, + void *(*start_routine)(void *), void *__restrict arg) { + pthread_create_wrapper_init(); + + return pthread_create_fptr(thread, attr, start_routine, arg); +} +#endif /* JEMALLOC_PTHREAD_CREATE_WRAPPER */ + +#ifndef JEMALLOC_BACKGROUND_THREAD +#define NOT_REACHED { not_reached(); } +bool background_thread_create(tsd_t *tsd, unsigned arena_ind) NOT_REACHED +bool background_threads_enable(tsd_t *tsd) NOT_REACHED +bool background_threads_disable(tsd_t *tsd) NOT_REACHED +void background_thread_interval_check(tsdn_t *tsdn, arena_t *arena, + arena_decay_t *decay, size_t npages_new) NOT_REACHED +void background_thread_prefork0(tsdn_t *tsdn) NOT_REACHED +void background_thread_prefork1(tsdn_t *tsdn) NOT_REACHED +void background_thread_postfork_parent(tsdn_t *tsdn) NOT_REACHED +void background_thread_postfork_child(tsdn_t *tsdn) NOT_REACHED +bool background_thread_stats_read(tsdn_t *tsdn, + background_thread_stats_t *stats) NOT_REACHED +void background_thread_ctl_init(tsdn_t *tsdn) NOT_REACHED +#undef NOT_REACHED +#else + +static bool background_thread_enabled_at_fork; + +static void +background_thread_info_init(tsdn_t *tsdn, background_thread_info_t *info) { + background_thread_wakeup_time_set(tsdn, info, 0); + info->npages_to_purge_new = 0; + if (config_stats) { + info->tot_n_runs = 0; + nstime_init(&info->tot_sleep_time, 0); + } +} + +static inline bool +set_current_thread_affinity(int cpu) { +#if defined(JEMALLOC_HAVE_SCHED_SETAFFINITY) + cpu_set_t cpuset; + CPU_ZERO(&cpuset); + CPU_SET(cpu, &cpuset); + int ret = sched_setaffinity(0, sizeof(cpu_set_t), &cpuset); + + return (ret != 0); +#else + return false; +#endif +} + +/* Threshold for determining when to wake up the background thread. */ +#define BACKGROUND_THREAD_NPAGES_THRESHOLD UINT64_C(1024) +#define BILLION UINT64_C(1000000000) +/* Minimal sleep interval 100 ms. */ +#define BACKGROUND_THREAD_MIN_INTERVAL_NS (BILLION / 10) + +static inline size_t +decay_npurge_after_interval(arena_decay_t *decay, size_t interval) { + size_t i; + uint64_t sum = 0; + for (i = 0; i < interval; i++) { + sum += decay->backlog[i] * h_steps[i]; + } + for (; i < SMOOTHSTEP_NSTEPS; i++) { + sum += decay->backlog[i] * (h_steps[i] - h_steps[i - interval]); + } + + return (size_t)(sum >> SMOOTHSTEP_BFP); +} + +static uint64_t +arena_decay_compute_purge_interval_impl(tsdn_t *tsdn, arena_decay_t *decay, + extents_t *extents) { + if (malloc_mutex_trylock(tsdn, &decay->mtx)) { + /* Use minimal interval if decay is contended. */ + return BACKGROUND_THREAD_MIN_INTERVAL_NS; + } + + uint64_t interval; + ssize_t decay_time = atomic_load_zd(&decay->time_ms, ATOMIC_RELAXED); + if (decay_time <= 0) { + /* Purging is eagerly done or disabled currently. */ + interval = BACKGROUND_THREAD_INDEFINITE_SLEEP; + goto label_done; + } + + uint64_t decay_interval_ns = nstime_ns(&decay->interval); + assert(decay_interval_ns > 0); + size_t npages = extents_npages_get(extents); + if (npages == 0) { + unsigned i; + for (i = 0; i < SMOOTHSTEP_NSTEPS; i++) { + if (decay->backlog[i] > 0) { + break; + } + } + if (i == SMOOTHSTEP_NSTEPS) { + /* No dirty pages recorded. Sleep indefinitely. */ + interval = BACKGROUND_THREAD_INDEFINITE_SLEEP; + goto label_done; + } + } + if (npages <= BACKGROUND_THREAD_NPAGES_THRESHOLD) { + /* Use max interval. */ + interval = decay_interval_ns * SMOOTHSTEP_NSTEPS; + goto label_done; + } + + size_t lb = BACKGROUND_THREAD_MIN_INTERVAL_NS / decay_interval_ns; + size_t ub = SMOOTHSTEP_NSTEPS; + /* Minimal 2 intervals to ensure reaching next epoch deadline. */ + lb = (lb < 2) ? 2 : lb; + if ((decay_interval_ns * ub <= BACKGROUND_THREAD_MIN_INTERVAL_NS) || + (lb + 2 > ub)) { + interval = BACKGROUND_THREAD_MIN_INTERVAL_NS; + goto label_done; + } + + assert(lb + 2 <= ub); + size_t npurge_lb, npurge_ub; + npurge_lb = decay_npurge_after_interval(decay, lb); + if (npurge_lb > BACKGROUND_THREAD_NPAGES_THRESHOLD) { + interval = decay_interval_ns * lb; + goto label_done; + } + npurge_ub = decay_npurge_after_interval(decay, ub); + if (npurge_ub < BACKGROUND_THREAD_NPAGES_THRESHOLD) { + interval = decay_interval_ns * ub; + goto label_done; + } + + unsigned n_search = 0; + size_t target, npurge; + while ((npurge_lb + BACKGROUND_THREAD_NPAGES_THRESHOLD < npurge_ub) + && (lb + 2 < ub)) { + target = (lb + ub) / 2; + npurge = decay_npurge_after_interval(decay, target); + if (npurge > BACKGROUND_THREAD_NPAGES_THRESHOLD) { + ub = target; + npurge_ub = npurge; + } else { + lb = target; + npurge_lb = npurge; + } + assert(n_search++ < lg_floor(SMOOTHSTEP_NSTEPS) + 1); + } + interval = decay_interval_ns * (ub + lb) / 2; +label_done: + interval = (interval < BACKGROUND_THREAD_MIN_INTERVAL_NS) ? + BACKGROUND_THREAD_MIN_INTERVAL_NS : interval; + malloc_mutex_unlock(tsdn, &decay->mtx); + + return interval; +} + +/* Compute purge interval for background threads. */ +static uint64_t +arena_decay_compute_purge_interval(tsdn_t *tsdn, arena_t *arena) { + uint64_t i1, i2; + i1 = arena_decay_compute_purge_interval_impl(tsdn, &arena->decay_dirty, + &arena->extents_dirty); + if (i1 == BACKGROUND_THREAD_MIN_INTERVAL_NS) { + return i1; + } + i2 = arena_decay_compute_purge_interval_impl(tsdn, &arena->decay_muzzy, + &arena->extents_muzzy); + + return i1 < i2 ? i1 : i2; +} + +static void +background_thread_sleep(tsdn_t *tsdn, background_thread_info_t *info, + uint64_t interval) { + if (config_stats) { + info->tot_n_runs++; + } + info->npages_to_purge_new = 0; + + struct timeval tv; + /* Specific clock required by timedwait. */ + gettimeofday(&tv, NULL); + nstime_t before_sleep; + nstime_init2(&before_sleep, tv.tv_sec, tv.tv_usec * 1000); + + int ret; + if (interval == BACKGROUND_THREAD_INDEFINITE_SLEEP) { + assert(background_thread_indefinite_sleep(info)); + ret = pthread_cond_wait(&info->cond, &info->mtx.lock); + assert(ret == 0); + } else { + assert(interval >= BACKGROUND_THREAD_MIN_INTERVAL_NS && + interval <= BACKGROUND_THREAD_INDEFINITE_SLEEP); + /* We need malloc clock (can be different from tv). */ + nstime_t next_wakeup; + nstime_init(&next_wakeup, 0); + nstime_update(&next_wakeup); + nstime_iadd(&next_wakeup, interval); + assert(nstime_ns(&next_wakeup) < + BACKGROUND_THREAD_INDEFINITE_SLEEP); + background_thread_wakeup_time_set(tsdn, info, + nstime_ns(&next_wakeup)); + + nstime_t ts_wakeup; + nstime_copy(&ts_wakeup, &before_sleep); + nstime_iadd(&ts_wakeup, interval); + struct timespec ts; + ts.tv_sec = (size_t)nstime_sec(&ts_wakeup); + ts.tv_nsec = (size_t)nstime_nsec(&ts_wakeup); + + assert(!background_thread_indefinite_sleep(info)); + ret = pthread_cond_timedwait(&info->cond, &info->mtx.lock, &ts); + assert(ret == ETIMEDOUT || ret == 0); + background_thread_wakeup_time_set(tsdn, info, + BACKGROUND_THREAD_INDEFINITE_SLEEP); + } + if (config_stats) { + gettimeofday(&tv, NULL); + nstime_t after_sleep; + nstime_init2(&after_sleep, tv.tv_sec, tv.tv_usec * 1000); + if (nstime_compare(&after_sleep, &before_sleep) > 0) { + nstime_subtract(&after_sleep, &before_sleep); + nstime_add(&info->tot_sleep_time, &after_sleep); + } + } +} + +static bool +background_thread_pause_check(tsdn_t *tsdn, background_thread_info_t *info) { + if (unlikely(info->state == background_thread_paused)) { + malloc_mutex_unlock(tsdn, &info->mtx); + /* Wait on global lock to update status. */ + malloc_mutex_lock(tsdn, &background_thread_lock); + malloc_mutex_unlock(tsdn, &background_thread_lock); + malloc_mutex_lock(tsdn, &info->mtx); + return true; + } + + return false; +} + +static inline void +background_work_sleep_once(tsdn_t *tsdn, background_thread_info_t *info, unsigned ind) { + uint64_t min_interval = BACKGROUND_THREAD_INDEFINITE_SLEEP; + unsigned narenas = narenas_total_get(); + + for (unsigned i = ind; i < narenas; i += max_background_threads) { + arena_t *arena = arena_get(tsdn, i, false); + if (!arena) { + continue; + } + arena_decay(tsdn, arena, true, false); + if (min_interval == BACKGROUND_THREAD_MIN_INTERVAL_NS) { + /* Min interval will be used. */ + continue; + } + uint64_t interval = arena_decay_compute_purge_interval(tsdn, + arena); + assert(interval >= BACKGROUND_THREAD_MIN_INTERVAL_NS); + if (min_interval > interval) { + min_interval = interval; + } + } + background_thread_sleep(tsdn, info, min_interval); +} + +static bool +background_threads_disable_single(tsd_t *tsd, background_thread_info_t *info) { + if (info == &background_thread_info[0]) { + malloc_mutex_assert_owner(tsd_tsdn(tsd), + &background_thread_lock); + } else { + malloc_mutex_assert_not_owner(tsd_tsdn(tsd), + &background_thread_lock); + } + + pre_reentrancy(tsd, NULL); + malloc_mutex_lock(tsd_tsdn(tsd), &info->mtx); + bool has_thread; + assert(info->state != background_thread_paused); + if (info->state == background_thread_started) { + has_thread = true; + info->state = background_thread_stopped; + pthread_cond_signal(&info->cond); + } else { + has_thread = false; + } + malloc_mutex_unlock(tsd_tsdn(tsd), &info->mtx); + + if (!has_thread) { + post_reentrancy(tsd); + return false; + } + void *ret; + if (pthread_join(info->thread, &ret)) { + post_reentrancy(tsd); + return true; + } + assert(ret == NULL); + n_background_threads--; + post_reentrancy(tsd); + + return false; +} + +static void *background_thread_entry(void *ind_arg); + +static int +background_thread_create_signals_masked(pthread_t *thread, + const pthread_attr_t *attr, void *(*start_routine)(void *), void *arg) { + /* + * Mask signals during thread creation so that the thread inherits + * an empty signal set. + */ + sigset_t set; + sigfillset(&set); + sigset_t oldset; + int mask_err = pthread_sigmask(SIG_SETMASK, &set, &oldset); + if (mask_err != 0) { + return mask_err; + } + int create_err = pthread_create_wrapper(thread, attr, start_routine, + arg); + /* + * Restore the signal mask. Failure to restore the signal mask here + * changes program behavior. + */ + int restore_err = pthread_sigmask(SIG_SETMASK, &oldset, NULL); + if (restore_err != 0) { + malloc_printf("<jemalloc>: background thread creation " + "failed (%d), and signal mask restoration failed " + "(%d)\n", create_err, restore_err); + if (opt_abort) { + abort(); + } + } + return create_err; +} + +static bool +check_background_thread_creation(tsd_t *tsd, unsigned *n_created, + bool *created_threads) { + bool ret = false; + if (likely(*n_created == n_background_threads)) { + return ret; + } + + tsdn_t *tsdn = tsd_tsdn(tsd); + malloc_mutex_unlock(tsdn, &background_thread_info[0].mtx); + for (unsigned i = 1; i < max_background_threads; i++) { + if (created_threads[i]) { + continue; + } + background_thread_info_t *info = &background_thread_info[i]; + malloc_mutex_lock(tsdn, &info->mtx); + /* + * In case of the background_thread_paused state because of + * arena reset, delay the creation. + */ + bool create = (info->state == background_thread_started); + malloc_mutex_unlock(tsdn, &info->mtx); + if (!create) { + continue; + } + + pre_reentrancy(tsd, NULL); + int err = background_thread_create_signals_masked(&info->thread, + NULL, background_thread_entry, (void *)(uintptr_t)i); + post_reentrancy(tsd); + + if (err == 0) { + (*n_created)++; + created_threads[i] = true; + } else { + malloc_printf("<jemalloc>: background thread " + "creation failed (%d)\n", err); + if (opt_abort) { + abort(); + } + } + /* Return to restart the loop since we unlocked. */ + ret = true; + break; + } + malloc_mutex_lock(tsdn, &background_thread_info[0].mtx); + + return ret; +} + +static void +background_thread0_work(tsd_t *tsd) { + /* Thread0 is also responsible for launching / terminating threads. */ + VARIABLE_ARRAY(bool, created_threads, max_background_threads); + unsigned i; + for (i = 1; i < max_background_threads; i++) { + created_threads[i] = false; + } + /* Start working, and create more threads when asked. */ + unsigned n_created = 1; + while (background_thread_info[0].state != background_thread_stopped) { + if (background_thread_pause_check(tsd_tsdn(tsd), + &background_thread_info[0])) { + continue; + } + if (check_background_thread_creation(tsd, &n_created, + (bool *)&created_threads)) { + continue; + } + background_work_sleep_once(tsd_tsdn(tsd), + &background_thread_info[0], 0); + } + + /* + * Shut down other threads at exit. Note that the ctl thread is holding + * the global background_thread mutex (and is waiting) for us. + */ + assert(!background_thread_enabled()); + for (i = 1; i < max_background_threads; i++) { + background_thread_info_t *info = &background_thread_info[i]; + assert(info->state != background_thread_paused); + if (created_threads[i]) { + background_threads_disable_single(tsd, info); + } else { + malloc_mutex_lock(tsd_tsdn(tsd), &info->mtx); + if (info->state != background_thread_stopped) { + /* The thread was not created. */ + assert(info->state == + background_thread_started); + n_background_threads--; + info->state = background_thread_stopped; + } + malloc_mutex_unlock(tsd_tsdn(tsd), &info->mtx); + } + } + background_thread_info[0].state = background_thread_stopped; + assert(n_background_threads == 1); +} + +static void +background_work(tsd_t *tsd, unsigned ind) { + background_thread_info_t *info = &background_thread_info[ind]; + + malloc_mutex_lock(tsd_tsdn(tsd), &info->mtx); + background_thread_wakeup_time_set(tsd_tsdn(tsd), info, + BACKGROUND_THREAD_INDEFINITE_SLEEP); + if (ind == 0) { + background_thread0_work(tsd); + } else { + while (info->state != background_thread_stopped) { + if (background_thread_pause_check(tsd_tsdn(tsd), + info)) { + continue; + } + background_work_sleep_once(tsd_tsdn(tsd), info, ind); + } + } + assert(info->state == background_thread_stopped); + background_thread_wakeup_time_set(tsd_tsdn(tsd), info, 0); + malloc_mutex_unlock(tsd_tsdn(tsd), &info->mtx); +} + +static void * +background_thread_entry(void *ind_arg) { + unsigned thread_ind = (unsigned)(uintptr_t)ind_arg; + assert(thread_ind < max_background_threads); +#ifdef JEMALLOC_HAVE_PTHREAD_SETNAME_NP + pthread_setname_np(pthread_self(), "jemalloc_bg_thd"); +#elif defined(__FreeBSD__) + pthread_set_name_np(pthread_self(), "jemalloc_bg_thd"); +#endif + if (opt_percpu_arena != percpu_arena_disabled) { + set_current_thread_affinity((int)thread_ind); + } + /* + * Start periodic background work. We use internal tsd which avoids + * side effects, for example triggering new arena creation (which in + * turn triggers another background thread creation). + */ + background_work(tsd_internal_fetch(), thread_ind); + assert(pthread_equal(pthread_self(), + background_thread_info[thread_ind].thread)); + + return NULL; +} + +static void +background_thread_init(tsd_t *tsd, background_thread_info_t *info) { + malloc_mutex_assert_owner(tsd_tsdn(tsd), &background_thread_lock); + info->state = background_thread_started; + background_thread_info_init(tsd_tsdn(tsd), info); + n_background_threads++; +} + +static bool +background_thread_create_locked(tsd_t *tsd, unsigned arena_ind) { + assert(have_background_thread); + malloc_mutex_assert_owner(tsd_tsdn(tsd), &background_thread_lock); + + /* We create at most NCPUs threads. */ + size_t thread_ind = arena_ind % max_background_threads; + background_thread_info_t *info = &background_thread_info[thread_ind]; + + bool need_new_thread; + malloc_mutex_lock(tsd_tsdn(tsd), &info->mtx); + need_new_thread = background_thread_enabled() && + (info->state == background_thread_stopped); + if (need_new_thread) { + background_thread_init(tsd, info); + } + malloc_mutex_unlock(tsd_tsdn(tsd), &info->mtx); + if (!need_new_thread) { + return false; + } + if (arena_ind != 0) { + /* Threads are created asynchronously by Thread 0. */ + background_thread_info_t *t0 = &background_thread_info[0]; + malloc_mutex_lock(tsd_tsdn(tsd), &t0->mtx); + assert(t0->state == background_thread_started); + pthread_cond_signal(&t0->cond); + malloc_mutex_unlock(tsd_tsdn(tsd), &t0->mtx); + + return false; + } + + pre_reentrancy(tsd, NULL); + /* + * To avoid complications (besides reentrancy), create internal + * background threads with the underlying pthread_create. + */ + int err = background_thread_create_signals_masked(&info->thread, NULL, + background_thread_entry, (void *)thread_ind); + post_reentrancy(tsd); + + if (err != 0) { + malloc_printf("<jemalloc>: arena 0 background thread creation " + "failed (%d)\n", err); + malloc_mutex_lock(tsd_tsdn(tsd), &info->mtx); + info->state = background_thread_stopped; + n_background_threads--; + malloc_mutex_unlock(tsd_tsdn(tsd), &info->mtx); + + return true; + } + + return false; +} + +/* Create a new background thread if needed. */ +bool +background_thread_create(tsd_t *tsd, unsigned arena_ind) { + assert(have_background_thread); + + bool ret; + malloc_mutex_lock(tsd_tsdn(tsd), &background_thread_lock); + ret = background_thread_create_locked(tsd, arena_ind); + malloc_mutex_unlock(tsd_tsdn(tsd), &background_thread_lock); + + return ret; +} + +bool +background_threads_enable(tsd_t *tsd) { + assert(n_background_threads == 0); + assert(background_thread_enabled()); + malloc_mutex_assert_owner(tsd_tsdn(tsd), &background_thread_lock); + + VARIABLE_ARRAY(bool, marked, max_background_threads); + unsigned i, nmarked; + for (i = 0; i < max_background_threads; i++) { + marked[i] = false; + } + nmarked = 0; + /* Thread 0 is required and created at the end. */ + marked[0] = true; + /* Mark the threads we need to create for thread 0. */ + unsigned n = narenas_total_get(); + for (i = 1; i < n; i++) { + if (marked[i % max_background_threads] || + arena_get(tsd_tsdn(tsd), i, false) == NULL) { + continue; + } + background_thread_info_t *info = &background_thread_info[ + i % max_background_threads]; + malloc_mutex_lock(tsd_tsdn(tsd), &info->mtx); + assert(info->state == background_thread_stopped); + background_thread_init(tsd, info); + malloc_mutex_unlock(tsd_tsdn(tsd), &info->mtx); + marked[i % max_background_threads] = true; + if (++nmarked == max_background_threads) { + break; + } + } + + return background_thread_create_locked(tsd, 0); +} + +bool +background_threads_disable(tsd_t *tsd) { + assert(!background_thread_enabled()); + malloc_mutex_assert_owner(tsd_tsdn(tsd), &background_thread_lock); + + /* Thread 0 will be responsible for terminating other threads. */ + if (background_threads_disable_single(tsd, + &background_thread_info[0])) { + return true; + } + assert(n_background_threads == 0); + + return false; +} + +/* Check if we need to signal the background thread early. */ +void +background_thread_interval_check(tsdn_t *tsdn, arena_t *arena, + arena_decay_t *decay, size_t npages_new) { + background_thread_info_t *info = arena_background_thread_info_get( + arena); + if (malloc_mutex_trylock(tsdn, &info->mtx)) { + /* + * Background thread may hold the mutex for a long period of + * time. We'd like to avoid the variance on application + * threads. So keep this non-blocking, and leave the work to a + * future epoch. + */ + return; + } + + if (info->state != background_thread_started) { + goto label_done; + } + if (malloc_mutex_trylock(tsdn, &decay->mtx)) { + goto label_done; + } + + ssize_t decay_time = atomic_load_zd(&decay->time_ms, ATOMIC_RELAXED); + if (decay_time <= 0) { + /* Purging is eagerly done or disabled currently. */ + goto label_done_unlock2; + } + uint64_t decay_interval_ns = nstime_ns(&decay->interval); + assert(decay_interval_ns > 0); + + nstime_t diff; + nstime_init(&diff, background_thread_wakeup_time_get(info)); + if (nstime_compare(&diff, &decay->epoch) <= 0) { + goto label_done_unlock2; + } + nstime_subtract(&diff, &decay->epoch); + if (nstime_ns(&diff) < BACKGROUND_THREAD_MIN_INTERVAL_NS) { + goto label_done_unlock2; + } + + if (npages_new > 0) { + size_t n_epoch = (size_t)(nstime_ns(&diff) / decay_interval_ns); + /* + * Compute how many new pages we would need to purge by the next + * wakeup, which is used to determine if we should signal the + * background thread. + */ + uint64_t npurge_new; + if (n_epoch >= SMOOTHSTEP_NSTEPS) { + npurge_new = npages_new; + } else { + uint64_t h_steps_max = h_steps[SMOOTHSTEP_NSTEPS - 1]; + assert(h_steps_max >= + h_steps[SMOOTHSTEP_NSTEPS - 1 - n_epoch]); + npurge_new = npages_new * (h_steps_max - + h_steps[SMOOTHSTEP_NSTEPS - 1 - n_epoch]); + npurge_new >>= SMOOTHSTEP_BFP; + } + info->npages_to_purge_new += npurge_new; + } + + bool should_signal; + if (info->npages_to_purge_new > BACKGROUND_THREAD_NPAGES_THRESHOLD) { + should_signal = true; + } else if (unlikely(background_thread_indefinite_sleep(info)) && + (extents_npages_get(&arena->extents_dirty) > 0 || + extents_npages_get(&arena->extents_muzzy) > 0 || + info->npages_to_purge_new > 0)) { + should_signal = true; + } else { + should_signal = false; + } + + if (should_signal) { + info->npages_to_purge_new = 0; + pthread_cond_signal(&info->cond); + } +label_done_unlock2: + malloc_mutex_unlock(tsdn, &decay->mtx); +label_done: + malloc_mutex_unlock(tsdn, &info->mtx); +} + +void +background_thread_prefork0(tsdn_t *tsdn) { + malloc_mutex_prefork(tsdn, &background_thread_lock); + background_thread_enabled_at_fork = background_thread_enabled(); +} + +void +background_thread_prefork1(tsdn_t *tsdn) { + for (unsigned i = 0; i < max_background_threads; i++) { + malloc_mutex_prefork(tsdn, &background_thread_info[i].mtx); + } +} + +void +background_thread_postfork_parent(tsdn_t *tsdn) { + for (unsigned i = 0; i < max_background_threads; i++) { + malloc_mutex_postfork_parent(tsdn, + &background_thread_info[i].mtx); + } + malloc_mutex_postfork_parent(tsdn, &background_thread_lock); +} + +void +background_thread_postfork_child(tsdn_t *tsdn) { + for (unsigned i = 0; i < max_background_threads; i++) { + malloc_mutex_postfork_child(tsdn, + &background_thread_info[i].mtx); + } + malloc_mutex_postfork_child(tsdn, &background_thread_lock); + if (!background_thread_enabled_at_fork) { + return; + } + + /* Clear background_thread state (reset to disabled for child). */ + malloc_mutex_lock(tsdn, &background_thread_lock); + n_background_threads = 0; + background_thread_enabled_set(tsdn, false); + for (unsigned i = 0; i < max_background_threads; i++) { + background_thread_info_t *info = &background_thread_info[i]; + malloc_mutex_lock(tsdn, &info->mtx); + info->state = background_thread_stopped; + int ret = pthread_cond_init(&info->cond, NULL); + assert(ret == 0); + background_thread_info_init(tsdn, info); + malloc_mutex_unlock(tsdn, &info->mtx); + } + malloc_mutex_unlock(tsdn, &background_thread_lock); +} + +bool +background_thread_stats_read(tsdn_t *tsdn, background_thread_stats_t *stats) { + assert(config_stats); + malloc_mutex_lock(tsdn, &background_thread_lock); + if (!background_thread_enabled()) { + malloc_mutex_unlock(tsdn, &background_thread_lock); + return true; + } + + stats->num_threads = n_background_threads; + uint64_t num_runs = 0; + nstime_init(&stats->run_interval, 0); + for (unsigned i = 0; i < max_background_threads; i++) { + background_thread_info_t *info = &background_thread_info[i]; + if (malloc_mutex_trylock(tsdn, &info->mtx)) { + /* + * Each background thread run may take a long time; + * avoid waiting on the stats if the thread is active. + */ + continue; + } + if (info->state != background_thread_stopped) { + num_runs += info->tot_n_runs; + nstime_add(&stats->run_interval, &info->tot_sleep_time); + } + malloc_mutex_unlock(tsdn, &info->mtx); + } + stats->num_runs = num_runs; + if (num_runs > 0) { + nstime_idivide(&stats->run_interval, num_runs); + } + malloc_mutex_unlock(tsdn, &background_thread_lock); + + return false; +} + +#undef BACKGROUND_THREAD_NPAGES_THRESHOLD +#undef BILLION +#undef BACKGROUND_THREAD_MIN_INTERVAL_NS + +#ifdef JEMALLOC_HAVE_DLSYM +#include <dlfcn.h> +#endif + +static bool +pthread_create_fptr_init(void) { + if (pthread_create_fptr != NULL) { + return false; + } + /* + * Try the next symbol first, because 1) when use lazy_lock we have a + * wrapper for pthread_create; and 2) application may define its own + * wrapper as well (and can call malloc within the wrapper). + */ +#ifdef JEMALLOC_HAVE_DLSYM + pthread_create_fptr = dlsym(RTLD_NEXT, "pthread_create"); +#else + pthread_create_fptr = NULL; +#endif + if (pthread_create_fptr == NULL) { + if (config_lazy_lock) { + malloc_write("<jemalloc>: Error in dlsym(RTLD_NEXT, " + "\"pthread_create\")\n"); + abort(); + } else { + /* Fall back to the default symbol. */ + pthread_create_fptr = pthread_create; + } + } + + return false; +} + +/* + * When lazy lock is enabled, we need to make sure setting isthreaded before + * taking any background_thread locks. This is called early in ctl (instead of + * wait for the pthread_create calls to trigger) because the mutex is required + * before creating background threads. + */ +void +background_thread_ctl_init(tsdn_t *tsdn) { + malloc_mutex_assert_not_owner(tsdn, &background_thread_lock); +#ifdef JEMALLOC_PTHREAD_CREATE_WRAPPER + pthread_create_fptr_init(); + pthread_create_wrapper_init(); +#endif +} + +#endif /* defined(JEMALLOC_BACKGROUND_THREAD) */ + +bool +background_thread_boot0(void) { + if (!have_background_thread && opt_background_thread) { + malloc_printf("<jemalloc>: option background_thread currently " + "supports pthread only\n"); + return true; + } +#ifdef JEMALLOC_PTHREAD_CREATE_WRAPPER + if ((config_lazy_lock || opt_background_thread) && + pthread_create_fptr_init()) { + return true; + } +#endif + return false; +} + +bool +background_thread_boot1(tsdn_t *tsdn) { +#ifdef JEMALLOC_BACKGROUND_THREAD + assert(have_background_thread); + assert(narenas_total_get() > 0); + + if (opt_max_background_threads > MAX_BACKGROUND_THREAD_LIMIT) { + opt_max_background_threads = DEFAULT_NUM_BACKGROUND_THREAD; + } + max_background_threads = opt_max_background_threads; + + background_thread_enabled_set(tsdn, opt_background_thread); + if (malloc_mutex_init(&background_thread_lock, + "background_thread_global", + WITNESS_RANK_BACKGROUND_THREAD_GLOBAL, + malloc_mutex_rank_exclusive)) { + return true; + } + + background_thread_info = (background_thread_info_t *)base_alloc(tsdn, + b0get(), opt_max_background_threads * + sizeof(background_thread_info_t), CACHELINE); + if (background_thread_info == NULL) { + return true; + } + + for (unsigned i = 0; i < max_background_threads; i++) { + background_thread_info_t *info = &background_thread_info[i]; + /* Thread mutex is rank_inclusive because of thread0. */ + if (malloc_mutex_init(&info->mtx, "background_thread", + WITNESS_RANK_BACKGROUND_THREAD, + malloc_mutex_address_ordered)) { + return true; + } + if (pthread_cond_init(&info->cond, NULL)) { + return true; + } + malloc_mutex_lock(tsdn, &info->mtx); + info->state = background_thread_stopped; + background_thread_info_init(tsdn, info); + malloc_mutex_unlock(tsdn, &info->mtx); + } +#endif + + return false; +} diff --git a/deps/jemalloc/src/base.c b/deps/jemalloc/src/base.c new file mode 100644 index 000000000..f3c61661a --- /dev/null +++ b/deps/jemalloc/src/base.c @@ -0,0 +1,514 @@ +#define JEMALLOC_BASE_C_ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/assert.h" +#include "jemalloc/internal/extent_mmap.h" +#include "jemalloc/internal/mutex.h" +#include "jemalloc/internal/sz.h" + +/******************************************************************************/ +/* Data. */ + +static base_t *b0; + +metadata_thp_mode_t opt_metadata_thp = METADATA_THP_DEFAULT; + +const char *metadata_thp_mode_names[] = { + "disabled", + "auto", + "always" +}; + +/******************************************************************************/ + +static inline bool +metadata_thp_madvise(void) { + return (metadata_thp_enabled() && + (init_system_thp_mode == thp_mode_default)); +} + +static void * +base_map(tsdn_t *tsdn, extent_hooks_t *extent_hooks, unsigned ind, size_t size) { + void *addr; + bool zero = true; + bool commit = true; + + /* Use huge page sizes and alignment regardless of opt_metadata_thp. */ + assert(size == HUGEPAGE_CEILING(size)); + size_t alignment = HUGEPAGE; + if (extent_hooks == &extent_hooks_default) { + addr = extent_alloc_mmap(NULL, size, alignment, &zero, &commit); + } else { + /* No arena context as we are creating new arenas. */ + tsd_t *tsd = tsdn_null(tsdn) ? tsd_fetch() : tsdn_tsd(tsdn); + pre_reentrancy(tsd, NULL); + addr = extent_hooks->alloc(extent_hooks, NULL, size, alignment, + &zero, &commit, ind); + post_reentrancy(tsd); + } + + return addr; +} + +static void +base_unmap(tsdn_t *tsdn, extent_hooks_t *extent_hooks, unsigned ind, void *addr, + size_t size) { + /* + * Cascade through dalloc, decommit, purge_forced, and purge_lazy, + * stopping at first success. This cascade is performed for consistency + * with the cascade in extent_dalloc_wrapper() because an application's + * custom hooks may not support e.g. dalloc. This function is only ever + * called as a side effect of arena destruction, so although it might + * seem pointless to do anything besides dalloc here, the application + * may in fact want the end state of all associated virtual memory to be + * in some consistent-but-allocated state. + */ + if (extent_hooks == &extent_hooks_default) { + if (!extent_dalloc_mmap(addr, size)) { + goto label_done; + } + if (!pages_decommit(addr, size)) { + goto label_done; + } + if (!pages_purge_forced(addr, size)) { + goto label_done; + } + if (!pages_purge_lazy(addr, size)) { + goto label_done; + } + /* Nothing worked. This should never happen. */ + not_reached(); + } else { + tsd_t *tsd = tsdn_null(tsdn) ? tsd_fetch() : tsdn_tsd(tsdn); + pre_reentrancy(tsd, NULL); + if (extent_hooks->dalloc != NULL && + !extent_hooks->dalloc(extent_hooks, addr, size, true, + ind)) { + goto label_post_reentrancy; + } + if (extent_hooks->decommit != NULL && + !extent_hooks->decommit(extent_hooks, addr, size, 0, size, + ind)) { + goto label_post_reentrancy; + } + if (extent_hooks->purge_forced != NULL && + !extent_hooks->purge_forced(extent_hooks, addr, size, 0, + size, ind)) { + goto label_post_reentrancy; + } + if (extent_hooks->purge_lazy != NULL && + !extent_hooks->purge_lazy(extent_hooks, addr, size, 0, size, + ind)) { + goto label_post_reentrancy; + } + /* Nothing worked. That's the application's problem. */ + label_post_reentrancy: + post_reentrancy(tsd); + } +label_done: + if (metadata_thp_madvise()) { + /* Set NOHUGEPAGE after unmap to avoid kernel defrag. */ + assert(((uintptr_t)addr & HUGEPAGE_MASK) == 0 && + (size & HUGEPAGE_MASK) == 0); + pages_nohuge(addr, size); + } +} + +static void +base_extent_init(size_t *extent_sn_next, extent_t *extent, void *addr, + size_t size) { + size_t sn; + + sn = *extent_sn_next; + (*extent_sn_next)++; + + extent_binit(extent, addr, size, sn); +} + +static size_t +base_get_num_blocks(base_t *base, bool with_new_block) { + base_block_t *b = base->blocks; + assert(b != NULL); + + size_t n_blocks = with_new_block ? 2 : 1; + while (b->next != NULL) { + n_blocks++; + b = b->next; + } + + return n_blocks; +} + +static void +base_auto_thp_switch(tsdn_t *tsdn, base_t *base) { + assert(opt_metadata_thp == metadata_thp_auto); + malloc_mutex_assert_owner(tsdn, &base->mtx); + if (base->auto_thp_switched) { + return; + } + /* Called when adding a new block. */ + bool should_switch; + if (base_ind_get(base) != 0) { + should_switch = (base_get_num_blocks(base, true) == + BASE_AUTO_THP_THRESHOLD); + } else { + should_switch = (base_get_num_blocks(base, true) == + BASE_AUTO_THP_THRESHOLD_A0); + } + if (!should_switch) { + return; + } + + base->auto_thp_switched = true; + assert(!config_stats || base->n_thp == 0); + /* Make the initial blocks THP lazily. */ + base_block_t *block = base->blocks; + while (block != NULL) { + assert((block->size & HUGEPAGE_MASK) == 0); + pages_huge(block, block->size); + if (config_stats) { + base->n_thp += HUGEPAGE_CEILING(block->size - + extent_bsize_get(&block->extent)) >> LG_HUGEPAGE; + } + block = block->next; + assert(block == NULL || (base_ind_get(base) == 0)); + } +} + +static void * +base_extent_bump_alloc_helper(extent_t *extent, size_t *gap_size, size_t size, + size_t alignment) { + void *ret; + + assert(alignment == ALIGNMENT_CEILING(alignment, QUANTUM)); + assert(size == ALIGNMENT_CEILING(size, alignment)); + + *gap_size = ALIGNMENT_CEILING((uintptr_t)extent_addr_get(extent), + alignment) - (uintptr_t)extent_addr_get(extent); + ret = (void *)((uintptr_t)extent_addr_get(extent) + *gap_size); + assert(extent_bsize_get(extent) >= *gap_size + size); + extent_binit(extent, (void *)((uintptr_t)extent_addr_get(extent) + + *gap_size + size), extent_bsize_get(extent) - *gap_size - size, + extent_sn_get(extent)); + return ret; +} + +static void +base_extent_bump_alloc_post(base_t *base, extent_t *extent, size_t gap_size, + void *addr, size_t size) { + if (extent_bsize_get(extent) > 0) { + /* + * Compute the index for the largest size class that does not + * exceed extent's size. + */ + szind_t index_floor = + sz_size2index(extent_bsize_get(extent) + 1) - 1; + extent_heap_insert(&base->avail[index_floor], extent); + } + + if (config_stats) { + base->allocated += size; + /* + * Add one PAGE to base_resident for every page boundary that is + * crossed by the new allocation. Adjust n_thp similarly when + * metadata_thp is enabled. + */ + base->resident += PAGE_CEILING((uintptr_t)addr + size) - + PAGE_CEILING((uintptr_t)addr - gap_size); + assert(base->allocated <= base->resident); + assert(base->resident <= base->mapped); + if (metadata_thp_madvise() && (opt_metadata_thp == + metadata_thp_always || base->auto_thp_switched)) { + base->n_thp += (HUGEPAGE_CEILING((uintptr_t)addr + size) + - HUGEPAGE_CEILING((uintptr_t)addr - gap_size)) >> + LG_HUGEPAGE; + assert(base->mapped >= base->n_thp << LG_HUGEPAGE); + } + } +} + +static void * +base_extent_bump_alloc(base_t *base, extent_t *extent, size_t size, + size_t alignment) { + void *ret; + size_t gap_size; + + ret = base_extent_bump_alloc_helper(extent, &gap_size, size, alignment); + base_extent_bump_alloc_post(base, extent, gap_size, ret, size); + return ret; +} + +/* + * Allocate a block of virtual memory that is large enough to start with a + * base_block_t header, followed by an object of specified size and alignment. + * On success a pointer to the initialized base_block_t header is returned. + */ +static base_block_t * +base_block_alloc(tsdn_t *tsdn, base_t *base, extent_hooks_t *extent_hooks, + unsigned ind, pszind_t *pind_last, size_t *extent_sn_next, size_t size, + size_t alignment) { + alignment = ALIGNMENT_CEILING(alignment, QUANTUM); + size_t usize = ALIGNMENT_CEILING(size, alignment); + size_t header_size = sizeof(base_block_t); + size_t gap_size = ALIGNMENT_CEILING(header_size, alignment) - + header_size; + /* + * Create increasingly larger blocks in order to limit the total number + * of disjoint virtual memory ranges. Choose the next size in the page + * size class series (skipping size classes that are not a multiple of + * HUGEPAGE), or a size large enough to satisfy the requested size and + * alignment, whichever is larger. + */ + size_t min_block_size = HUGEPAGE_CEILING(sz_psz2u(header_size + gap_size + + usize)); + pszind_t pind_next = (*pind_last + 1 < sz_psz2ind(SC_LARGE_MAXCLASS)) ? + *pind_last + 1 : *pind_last; + size_t next_block_size = HUGEPAGE_CEILING(sz_pind2sz(pind_next)); + size_t block_size = (min_block_size > next_block_size) ? min_block_size + : next_block_size; + base_block_t *block = (base_block_t *)base_map(tsdn, extent_hooks, ind, + block_size); + if (block == NULL) { + return NULL; + } + + if (metadata_thp_madvise()) { + void *addr = (void *)block; + assert(((uintptr_t)addr & HUGEPAGE_MASK) == 0 && + (block_size & HUGEPAGE_MASK) == 0); + if (opt_metadata_thp == metadata_thp_always) { + pages_huge(addr, block_size); + } else if (opt_metadata_thp == metadata_thp_auto && + base != NULL) { + /* base != NULL indicates this is not a new base. */ + malloc_mutex_lock(tsdn, &base->mtx); + base_auto_thp_switch(tsdn, base); + if (base->auto_thp_switched) { + pages_huge(addr, block_size); + } + malloc_mutex_unlock(tsdn, &base->mtx); + } + } + + *pind_last = sz_psz2ind(block_size); + block->size = block_size; + block->next = NULL; + assert(block_size >= header_size); + base_extent_init(extent_sn_next, &block->extent, + (void *)((uintptr_t)block + header_size), block_size - header_size); + return block; +} + +/* + * Allocate an extent that is at least as large as specified size, with + * specified alignment. + */ +static extent_t * +base_extent_alloc(tsdn_t *tsdn, base_t *base, size_t size, size_t alignment) { + malloc_mutex_assert_owner(tsdn, &base->mtx); + + extent_hooks_t *extent_hooks = base_extent_hooks_get(base); + /* + * Drop mutex during base_block_alloc(), because an extent hook will be + * called. + */ + malloc_mutex_unlock(tsdn, &base->mtx); + base_block_t *block = base_block_alloc(tsdn, base, extent_hooks, + base_ind_get(base), &base->pind_last, &base->extent_sn_next, size, + alignment); + malloc_mutex_lock(tsdn, &base->mtx); + if (block == NULL) { + return NULL; + } + block->next = base->blocks; + base->blocks = block; + if (config_stats) { + base->allocated += sizeof(base_block_t); + base->resident += PAGE_CEILING(sizeof(base_block_t)); + base->mapped += block->size; + if (metadata_thp_madvise() && + !(opt_metadata_thp == metadata_thp_auto + && !base->auto_thp_switched)) { + assert(base->n_thp > 0); + base->n_thp += HUGEPAGE_CEILING(sizeof(base_block_t)) >> + LG_HUGEPAGE; + } + assert(base->allocated <= base->resident); + assert(base->resident <= base->mapped); + assert(base->n_thp << LG_HUGEPAGE <= base->mapped); + } + return &block->extent; +} + +base_t * +b0get(void) { + return b0; +} + +base_t * +base_new(tsdn_t *tsdn, unsigned ind, extent_hooks_t *extent_hooks) { + pszind_t pind_last = 0; + size_t extent_sn_next = 0; + base_block_t *block = base_block_alloc(tsdn, NULL, extent_hooks, ind, + &pind_last, &extent_sn_next, sizeof(base_t), QUANTUM); + if (block == NULL) { + return NULL; + } + + size_t gap_size; + size_t base_alignment = CACHELINE; + size_t base_size = ALIGNMENT_CEILING(sizeof(base_t), base_alignment); + base_t *base = (base_t *)base_extent_bump_alloc_helper(&block->extent, + &gap_size, base_size, base_alignment); + base->ind = ind; + atomic_store_p(&base->extent_hooks, extent_hooks, ATOMIC_RELAXED); + if (malloc_mutex_init(&base->mtx, "base", WITNESS_RANK_BASE, + malloc_mutex_rank_exclusive)) { + base_unmap(tsdn, extent_hooks, ind, block, block->size); + return NULL; + } + base->pind_last = pind_last; + base->extent_sn_next = extent_sn_next; + base->blocks = block; + base->auto_thp_switched = false; + for (szind_t i = 0; i < SC_NSIZES; i++) { + extent_heap_new(&base->avail[i]); + } + if (config_stats) { + base->allocated = sizeof(base_block_t); + base->resident = PAGE_CEILING(sizeof(base_block_t)); + base->mapped = block->size; + base->n_thp = (opt_metadata_thp == metadata_thp_always) && + metadata_thp_madvise() ? HUGEPAGE_CEILING(sizeof(base_block_t)) + >> LG_HUGEPAGE : 0; + assert(base->allocated <= base->resident); + assert(base->resident <= base->mapped); + assert(base->n_thp << LG_HUGEPAGE <= base->mapped); + } + base_extent_bump_alloc_post(base, &block->extent, gap_size, base, + base_size); + + return base; +} + +void +base_delete(tsdn_t *tsdn, base_t *base) { + extent_hooks_t *extent_hooks = base_extent_hooks_get(base); + base_block_t *next = base->blocks; + do { + base_block_t *block = next; + next = block->next; + base_unmap(tsdn, extent_hooks, base_ind_get(base), block, + block->size); + } while (next != NULL); +} + +extent_hooks_t * +base_extent_hooks_get(base_t *base) { + return (extent_hooks_t *)atomic_load_p(&base->extent_hooks, + ATOMIC_ACQUIRE); +} + +extent_hooks_t * +base_extent_hooks_set(base_t *base, extent_hooks_t *extent_hooks) { + extent_hooks_t *old_extent_hooks = base_extent_hooks_get(base); + atomic_store_p(&base->extent_hooks, extent_hooks, ATOMIC_RELEASE); + return old_extent_hooks; +} + +static void * +base_alloc_impl(tsdn_t *tsdn, base_t *base, size_t size, size_t alignment, + size_t *esn) { + alignment = QUANTUM_CEILING(alignment); + size_t usize = ALIGNMENT_CEILING(size, alignment); + size_t asize = usize + alignment - QUANTUM; + + extent_t *extent = NULL; + malloc_mutex_lock(tsdn, &base->mtx); + for (szind_t i = sz_size2index(asize); i < SC_NSIZES; i++) { + extent = extent_heap_remove_first(&base->avail[i]); + if (extent != NULL) { + /* Use existing space. */ + break; + } + } + if (extent == NULL) { + /* Try to allocate more space. */ + extent = base_extent_alloc(tsdn, base, usize, alignment); + } + void *ret; + if (extent == NULL) { + ret = NULL; + goto label_return; + } + + ret = base_extent_bump_alloc(base, extent, usize, alignment); + if (esn != NULL) { + *esn = extent_sn_get(extent); + } +label_return: + malloc_mutex_unlock(tsdn, &base->mtx); + return ret; +} + +/* + * base_alloc() returns zeroed memory, which is always demand-zeroed for the + * auto arenas, in order to make multi-page sparse data structures such as radix + * tree nodes efficient with respect to physical memory usage. Upon success a + * pointer to at least size bytes with specified alignment is returned. Note + * that size is rounded up to the nearest multiple of alignment to avoid false + * sharing. + */ +void * +base_alloc(tsdn_t *tsdn, base_t *base, size_t size, size_t alignment) { + return base_alloc_impl(tsdn, base, size, alignment, NULL); +} + +extent_t * +base_alloc_extent(tsdn_t *tsdn, base_t *base) { + size_t esn; + extent_t *extent = base_alloc_impl(tsdn, base, sizeof(extent_t), + CACHELINE, &esn); + if (extent == NULL) { + return NULL; + } + extent_esn_set(extent, esn); + return extent; +} + +void +base_stats_get(tsdn_t *tsdn, base_t *base, size_t *allocated, size_t *resident, + size_t *mapped, size_t *n_thp) { + cassert(config_stats); + + malloc_mutex_lock(tsdn, &base->mtx); + assert(base->allocated <= base->resident); + assert(base->resident <= base->mapped); + *allocated = base->allocated; + *resident = base->resident; + *mapped = base->mapped; + *n_thp = base->n_thp; + malloc_mutex_unlock(tsdn, &base->mtx); +} + +void +base_prefork(tsdn_t *tsdn, base_t *base) { + malloc_mutex_prefork(tsdn, &base->mtx); +} + +void +base_postfork_parent(tsdn_t *tsdn, base_t *base) { + malloc_mutex_postfork_parent(tsdn, &base->mtx); +} + +void +base_postfork_child(tsdn_t *tsdn, base_t *base) { + malloc_mutex_postfork_child(tsdn, &base->mtx); +} + +bool +base_boot(tsdn_t *tsdn) { + b0 = base_new(tsdn, 0, (extent_hooks_t *)&extent_hooks_default); + return (b0 == NULL); +} diff --git a/deps/jemalloc/src/bin.c b/deps/jemalloc/src/bin.c new file mode 100644 index 000000000..bca6b12c3 --- /dev/null +++ b/deps/jemalloc/src/bin.c @@ -0,0 +1,95 @@ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/assert.h" +#include "jemalloc/internal/bin.h" +#include "jemalloc/internal/sc.h" +#include "jemalloc/internal/witness.h" + +bin_info_t bin_infos[SC_NBINS]; + +static void +bin_infos_init(sc_data_t *sc_data, unsigned bin_shard_sizes[SC_NBINS], + bin_info_t bin_infos[SC_NBINS]) { + for (unsigned i = 0; i < SC_NBINS; i++) { + bin_info_t *bin_info = &bin_infos[i]; + sc_t *sc = &sc_data->sc[i]; + bin_info->reg_size = ((size_t)1U << sc->lg_base) + + ((size_t)sc->ndelta << sc->lg_delta); + bin_info->slab_size = (sc->pgs << LG_PAGE); + bin_info->nregs = + (uint32_t)(bin_info->slab_size / bin_info->reg_size); + bin_info->n_shards = bin_shard_sizes[i]; + bitmap_info_t bitmap_info = BITMAP_INFO_INITIALIZER( + bin_info->nregs); + bin_info->bitmap_info = bitmap_info; + } +} + +bool +bin_update_shard_size(unsigned bin_shard_sizes[SC_NBINS], size_t start_size, + size_t end_size, size_t nshards) { + if (nshards > BIN_SHARDS_MAX || nshards == 0) { + return true; + } + + if (start_size > SC_SMALL_MAXCLASS) { + return false; + } + if (end_size > SC_SMALL_MAXCLASS) { + end_size = SC_SMALL_MAXCLASS; + } + + /* Compute the index since this may happen before sz init. */ + szind_t ind1 = sz_size2index_compute(start_size); + szind_t ind2 = sz_size2index_compute(end_size); + for (unsigned i = ind1; i <= ind2; i++) { + bin_shard_sizes[i] = (unsigned)nshards; + } + + return false; +} + +void +bin_shard_sizes_boot(unsigned bin_shard_sizes[SC_NBINS]) { + /* Load the default number of shards. */ + for (unsigned i = 0; i < SC_NBINS; i++) { + bin_shard_sizes[i] = N_BIN_SHARDS_DEFAULT; + } +} + +void +bin_boot(sc_data_t *sc_data, unsigned bin_shard_sizes[SC_NBINS]) { + assert(sc_data->initialized); + bin_infos_init(sc_data, bin_shard_sizes, bin_infos); +} + +bool +bin_init(bin_t *bin) { + if (malloc_mutex_init(&bin->lock, "bin", WITNESS_RANK_BIN, + malloc_mutex_rank_exclusive)) { + return true; + } + bin->slabcur = NULL; + extent_heap_new(&bin->slabs_nonfull); + extent_list_init(&bin->slabs_full); + if (config_stats) { + memset(&bin->stats, 0, sizeof(bin_stats_t)); + } + return false; +} + +void +bin_prefork(tsdn_t *tsdn, bin_t *bin) { + malloc_mutex_prefork(tsdn, &bin->lock); +} + +void +bin_postfork_parent(tsdn_t *tsdn, bin_t *bin) { + malloc_mutex_postfork_parent(tsdn, &bin->lock); +} + +void +bin_postfork_child(tsdn_t *tsdn, bin_t *bin) { + malloc_mutex_postfork_child(tsdn, &bin->lock); +} diff --git a/deps/jemalloc/src/bitmap.c b/deps/jemalloc/src/bitmap.c new file mode 100644 index 000000000..468b3178e --- /dev/null +++ b/deps/jemalloc/src/bitmap.c @@ -0,0 +1,121 @@ +#define JEMALLOC_BITMAP_C_ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/assert.h" + +/******************************************************************************/ + +#ifdef BITMAP_USE_TREE + +void +bitmap_info_init(bitmap_info_t *binfo, size_t nbits) { + unsigned i; + size_t group_count; + + assert(nbits > 0); + assert(nbits <= (ZU(1) << LG_BITMAP_MAXBITS)); + + /* + * Compute the number of groups necessary to store nbits bits, and + * progressively work upward through the levels until reaching a level + * that requires only one group. + */ + binfo->levels[0].group_offset = 0; + group_count = BITMAP_BITS2GROUPS(nbits); + for (i = 1; group_count > 1; i++) { + assert(i < BITMAP_MAX_LEVELS); + binfo->levels[i].group_offset = binfo->levels[i-1].group_offset + + group_count; + group_count = BITMAP_BITS2GROUPS(group_count); + } + binfo->levels[i].group_offset = binfo->levels[i-1].group_offset + + group_count; + assert(binfo->levels[i].group_offset <= BITMAP_GROUPS_MAX); + binfo->nlevels = i; + binfo->nbits = nbits; +} + +static size_t +bitmap_info_ngroups(const bitmap_info_t *binfo) { + return binfo->levels[binfo->nlevels].group_offset; +} + +void +bitmap_init(bitmap_t *bitmap, const bitmap_info_t *binfo, bool fill) { + size_t extra; + unsigned i; + + /* + * Bits are actually inverted with regard to the external bitmap + * interface. + */ + + if (fill) { + /* The "filled" bitmap starts out with all 0 bits. */ + memset(bitmap, 0, bitmap_size(binfo)); + return; + } + + /* + * The "empty" bitmap starts out with all 1 bits, except for trailing + * unused bits (if any). Note that each group uses bit 0 to correspond + * to the first logical bit in the group, so extra bits are the most + * significant bits of the last group. + */ + memset(bitmap, 0xffU, bitmap_size(binfo)); + extra = (BITMAP_GROUP_NBITS - (binfo->nbits & BITMAP_GROUP_NBITS_MASK)) + & BITMAP_GROUP_NBITS_MASK; + if (extra != 0) { + bitmap[binfo->levels[1].group_offset - 1] >>= extra; + } + for (i = 1; i < binfo->nlevels; i++) { + size_t group_count = binfo->levels[i].group_offset - + binfo->levels[i-1].group_offset; + extra = (BITMAP_GROUP_NBITS - (group_count & + BITMAP_GROUP_NBITS_MASK)) & BITMAP_GROUP_NBITS_MASK; + if (extra != 0) { + bitmap[binfo->levels[i+1].group_offset - 1] >>= extra; + } + } +} + +#else /* BITMAP_USE_TREE */ + +void +bitmap_info_init(bitmap_info_t *binfo, size_t nbits) { + assert(nbits > 0); + assert(nbits <= (ZU(1) << LG_BITMAP_MAXBITS)); + + binfo->ngroups = BITMAP_BITS2GROUPS(nbits); + binfo->nbits = nbits; +} + +static size_t +bitmap_info_ngroups(const bitmap_info_t *binfo) { + return binfo->ngroups; +} + +void +bitmap_init(bitmap_t *bitmap, const bitmap_info_t *binfo, bool fill) { + size_t extra; + + if (fill) { + memset(bitmap, 0, bitmap_size(binfo)); + return; + } + + memset(bitmap, 0xffU, bitmap_size(binfo)); + extra = (BITMAP_GROUP_NBITS - (binfo->nbits & BITMAP_GROUP_NBITS_MASK)) + & BITMAP_GROUP_NBITS_MASK; + if (extra != 0) { + bitmap[binfo->ngroups - 1] >>= extra; + } +} + +#endif /* BITMAP_USE_TREE */ + +size_t +bitmap_size(const bitmap_info_t *binfo) { + return (bitmap_info_ngroups(binfo) << LG_SIZEOF_BITMAP); +} diff --git a/deps/jemalloc/src/ckh.c b/deps/jemalloc/src/ckh.c new file mode 100644 index 000000000..1bf6df5a1 --- /dev/null +++ b/deps/jemalloc/src/ckh.c @@ -0,0 +1,570 @@ +/* + ******************************************************************************* + * Implementation of (2^1+,2) cuckoo hashing, where 2^1+ indicates that each + * hash bucket contains 2^n cells, for n >= 1, and 2 indicates that two hash + * functions are employed. The original cuckoo hashing algorithm was described + * in: + * + * Pagh, R., F.F. Rodler (2004) Cuckoo Hashing. Journal of Algorithms + * 51(2):122-144. + * + * Generalization of cuckoo hashing was discussed in: + * + * Erlingsson, U., M. Manasse, F. McSherry (2006) A cool and practical + * alternative to traditional hash tables. In Proceedings of the 7th + * Workshop on Distributed Data and Structures (WDAS'06), Santa Clara, CA, + * January 2006. + * + * This implementation uses precisely two hash functions because that is the + * fewest that can work, and supporting multiple hashes is an implementation + * burden. Here is a reproduction of Figure 1 from Erlingsson et al. (2006) + * that shows approximate expected maximum load factors for various + * configurations: + * + * | #cells/bucket | + * #hashes | 1 | 2 | 4 | 8 | + * --------+-------+-------+-------+-------+ + * 1 | 0.006 | 0.006 | 0.03 | 0.12 | + * 2 | 0.49 | 0.86 |>0.93< |>0.96< | + * 3 | 0.91 | 0.97 | 0.98 | 0.999 | + * 4 | 0.97 | 0.99 | 0.999 | | + * + * The number of cells per bucket is chosen such that a bucket fits in one cache + * line. So, on 32- and 64-bit systems, we use (8,2) and (4,2) cuckoo hashing, + * respectively. + * + ******************************************************************************/ +#define JEMALLOC_CKH_C_ +#include "jemalloc/internal/jemalloc_preamble.h" + +#include "jemalloc/internal/ckh.h" + +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/assert.h" +#include "jemalloc/internal/hash.h" +#include "jemalloc/internal/malloc_io.h" +#include "jemalloc/internal/prng.h" +#include "jemalloc/internal/util.h" + +/******************************************************************************/ +/* Function prototypes for non-inline static functions. */ + +static bool ckh_grow(tsd_t *tsd, ckh_t *ckh); +static void ckh_shrink(tsd_t *tsd, ckh_t *ckh); + +/******************************************************************************/ + +/* + * Search bucket for key and return the cell number if found; SIZE_T_MAX + * otherwise. + */ +static size_t +ckh_bucket_search(ckh_t *ckh, size_t bucket, const void *key) { + ckhc_t *cell; + unsigned i; + + for (i = 0; i < (ZU(1) << LG_CKH_BUCKET_CELLS); i++) { + cell = &ckh->tab[(bucket << LG_CKH_BUCKET_CELLS) + i]; + if (cell->key != NULL && ckh->keycomp(key, cell->key)) { + return (bucket << LG_CKH_BUCKET_CELLS) + i; + } + } + + return SIZE_T_MAX; +} + +/* + * Search table for key and return cell number if found; SIZE_T_MAX otherwise. + */ +static size_t +ckh_isearch(ckh_t *ckh, const void *key) { + size_t hashes[2], bucket, cell; + + assert(ckh != NULL); + + ckh->hash(key, hashes); + + /* Search primary bucket. */ + bucket = hashes[0] & ((ZU(1) << ckh->lg_curbuckets) - 1); + cell = ckh_bucket_search(ckh, bucket, key); + if (cell != SIZE_T_MAX) { + return cell; + } + + /* Search secondary bucket. */ + bucket = hashes[1] & ((ZU(1) << ckh->lg_curbuckets) - 1); + cell = ckh_bucket_search(ckh, bucket, key); + return cell; +} + +static bool +ckh_try_bucket_insert(ckh_t *ckh, size_t bucket, const void *key, + const void *data) { + ckhc_t *cell; + unsigned offset, i; + + /* + * Cycle through the cells in the bucket, starting at a random position. + * The randomness avoids worst-case search overhead as buckets fill up. + */ + offset = (unsigned)prng_lg_range_u64(&ckh->prng_state, + LG_CKH_BUCKET_CELLS); + for (i = 0; i < (ZU(1) << LG_CKH_BUCKET_CELLS); i++) { + cell = &ckh->tab[(bucket << LG_CKH_BUCKET_CELLS) + + ((i + offset) & ((ZU(1) << LG_CKH_BUCKET_CELLS) - 1))]; + if (cell->key == NULL) { + cell->key = key; + cell->data = data; + ckh->count++; + return false; + } + } + + return true; +} + +/* + * No space is available in bucket. Randomly evict an item, then try to find an + * alternate location for that item. Iteratively repeat this + * eviction/relocation procedure until either success or detection of an + * eviction/relocation bucket cycle. + */ +static bool +ckh_evict_reloc_insert(ckh_t *ckh, size_t argbucket, void const **argkey, + void const **argdata) { + const void *key, *data, *tkey, *tdata; + ckhc_t *cell; + size_t hashes[2], bucket, tbucket; + unsigned i; + + bucket = argbucket; + key = *argkey; + data = *argdata; + while (true) { + /* + * Choose a random item within the bucket to evict. This is + * critical to correct function, because without (eventually) + * evicting all items within a bucket during iteration, it + * would be possible to get stuck in an infinite loop if there + * were an item for which both hashes indicated the same + * bucket. + */ + i = (unsigned)prng_lg_range_u64(&ckh->prng_state, + LG_CKH_BUCKET_CELLS); + cell = &ckh->tab[(bucket << LG_CKH_BUCKET_CELLS) + i]; + assert(cell->key != NULL); + + /* Swap cell->{key,data} and {key,data} (evict). */ + tkey = cell->key; tdata = cell->data; + cell->key = key; cell->data = data; + key = tkey; data = tdata; + +#ifdef CKH_COUNT + ckh->nrelocs++; +#endif + + /* Find the alternate bucket for the evicted item. */ + ckh->hash(key, hashes); + tbucket = hashes[1] & ((ZU(1) << ckh->lg_curbuckets) - 1); + if (tbucket == bucket) { + tbucket = hashes[0] & ((ZU(1) << ckh->lg_curbuckets) + - 1); + /* + * It may be that (tbucket == bucket) still, if the + * item's hashes both indicate this bucket. However, + * we are guaranteed to eventually escape this bucket + * during iteration, assuming pseudo-random item + * selection (true randomness would make infinite + * looping a remote possibility). The reason we can + * never get trapped forever is that there are two + * cases: + * + * 1) This bucket == argbucket, so we will quickly + * detect an eviction cycle and terminate. + * 2) An item was evicted to this bucket from another, + * which means that at least one item in this bucket + * has hashes that indicate distinct buckets. + */ + } + /* Check for a cycle. */ + if (tbucket == argbucket) { + *argkey = key; + *argdata = data; + return true; + } + + bucket = tbucket; + if (!ckh_try_bucket_insert(ckh, bucket, key, data)) { + return false; + } + } +} + +static bool +ckh_try_insert(ckh_t *ckh, void const**argkey, void const**argdata) { + size_t hashes[2], bucket; + const void *key = *argkey; + const void *data = *argdata; + + ckh->hash(key, hashes); + + /* Try to insert in primary bucket. */ + bucket = hashes[0] & ((ZU(1) << ckh->lg_curbuckets) - 1); + if (!ckh_try_bucket_insert(ckh, bucket, key, data)) { + return false; + } + + /* Try to insert in secondary bucket. */ + bucket = hashes[1] & ((ZU(1) << ckh->lg_curbuckets) - 1); + if (!ckh_try_bucket_insert(ckh, bucket, key, data)) { + return false; + } + + /* + * Try to find a place for this item via iterative eviction/relocation. + */ + return ckh_evict_reloc_insert(ckh, bucket, argkey, argdata); +} + +/* + * Try to rebuild the hash table from scratch by inserting all items from the + * old table into the new. + */ +static bool +ckh_rebuild(ckh_t *ckh, ckhc_t *aTab) { + size_t count, i, nins; + const void *key, *data; + + count = ckh->count; + ckh->count = 0; + for (i = nins = 0; nins < count; i++) { + if (aTab[i].key != NULL) { + key = aTab[i].key; + data = aTab[i].data; + if (ckh_try_insert(ckh, &key, &data)) { + ckh->count = count; + return true; + } + nins++; + } + } + + return false; +} + +static bool +ckh_grow(tsd_t *tsd, ckh_t *ckh) { + bool ret; + ckhc_t *tab, *ttab; + unsigned lg_prevbuckets, lg_curcells; + +#ifdef CKH_COUNT + ckh->ngrows++; +#endif + + /* + * It is possible (though unlikely, given well behaved hashes) that the + * table will have to be doubled more than once in order to create a + * usable table. + */ + lg_prevbuckets = ckh->lg_curbuckets; + lg_curcells = ckh->lg_curbuckets + LG_CKH_BUCKET_CELLS; + while (true) { + size_t usize; + + lg_curcells++; + usize = sz_sa2u(sizeof(ckhc_t) << lg_curcells, CACHELINE); + if (unlikely(usize == 0 + || usize > SC_LARGE_MAXCLASS)) { + ret = true; + goto label_return; + } + tab = (ckhc_t *)ipallocztm(tsd_tsdn(tsd), usize, CACHELINE, + true, NULL, true, arena_ichoose(tsd, NULL)); + if (tab == NULL) { + ret = true; + goto label_return; + } + /* Swap in new table. */ + ttab = ckh->tab; + ckh->tab = tab; + tab = ttab; + ckh->lg_curbuckets = lg_curcells - LG_CKH_BUCKET_CELLS; + + if (!ckh_rebuild(ckh, tab)) { + idalloctm(tsd_tsdn(tsd), tab, NULL, NULL, true, true); + break; + } + + /* Rebuilding failed, so back out partially rebuilt table. */ + idalloctm(tsd_tsdn(tsd), ckh->tab, NULL, NULL, true, true); + ckh->tab = tab; + ckh->lg_curbuckets = lg_prevbuckets; + } + + ret = false; +label_return: + return ret; +} + +static void +ckh_shrink(tsd_t *tsd, ckh_t *ckh) { + ckhc_t *tab, *ttab; + size_t usize; + unsigned lg_prevbuckets, lg_curcells; + + /* + * It is possible (though unlikely, given well behaved hashes) that the + * table rebuild will fail. + */ + lg_prevbuckets = ckh->lg_curbuckets; + lg_curcells = ckh->lg_curbuckets + LG_CKH_BUCKET_CELLS - 1; + usize = sz_sa2u(sizeof(ckhc_t) << lg_curcells, CACHELINE); + if (unlikely(usize == 0 || usize > SC_LARGE_MAXCLASS)) { + return; + } + tab = (ckhc_t *)ipallocztm(tsd_tsdn(tsd), usize, CACHELINE, true, NULL, + true, arena_ichoose(tsd, NULL)); + if (tab == NULL) { + /* + * An OOM error isn't worth propagating, since it doesn't + * prevent this or future operations from proceeding. + */ + return; + } + /* Swap in new table. */ + ttab = ckh->tab; + ckh->tab = tab; + tab = ttab; + ckh->lg_curbuckets = lg_curcells - LG_CKH_BUCKET_CELLS; + + if (!ckh_rebuild(ckh, tab)) { + idalloctm(tsd_tsdn(tsd), tab, NULL, NULL, true, true); +#ifdef CKH_COUNT + ckh->nshrinks++; +#endif + return; + } + + /* Rebuilding failed, so back out partially rebuilt table. */ + idalloctm(tsd_tsdn(tsd), ckh->tab, NULL, NULL, true, true); + ckh->tab = tab; + ckh->lg_curbuckets = lg_prevbuckets; +#ifdef CKH_COUNT + ckh->nshrinkfails++; +#endif +} + +bool +ckh_new(tsd_t *tsd, ckh_t *ckh, size_t minitems, ckh_hash_t *hash, + ckh_keycomp_t *keycomp) { + bool ret; + size_t mincells, usize; + unsigned lg_mincells; + + assert(minitems > 0); + assert(hash != NULL); + assert(keycomp != NULL); + +#ifdef CKH_COUNT + ckh->ngrows = 0; + ckh->nshrinks = 0; + ckh->nshrinkfails = 0; + ckh->ninserts = 0; + ckh->nrelocs = 0; +#endif + ckh->prng_state = 42; /* Value doesn't really matter. */ + ckh->count = 0; + + /* + * Find the minimum power of 2 that is large enough to fit minitems + * entries. We are using (2+,2) cuckoo hashing, which has an expected + * maximum load factor of at least ~0.86, so 0.75 is a conservative load + * factor that will typically allow mincells items to fit without ever + * growing the table. + */ + assert(LG_CKH_BUCKET_CELLS > 0); + mincells = ((minitems + (3 - (minitems % 3))) / 3) << 2; + for (lg_mincells = LG_CKH_BUCKET_CELLS; + (ZU(1) << lg_mincells) < mincells; + lg_mincells++) { + /* Do nothing. */ + } + ckh->lg_minbuckets = lg_mincells - LG_CKH_BUCKET_CELLS; + ckh->lg_curbuckets = lg_mincells - LG_CKH_BUCKET_CELLS; + ckh->hash = hash; + ckh->keycomp = keycomp; + + usize = sz_sa2u(sizeof(ckhc_t) << lg_mincells, CACHELINE); + if (unlikely(usize == 0 || usize > SC_LARGE_MAXCLASS)) { + ret = true; + goto label_return; + } + ckh->tab = (ckhc_t *)ipallocztm(tsd_tsdn(tsd), usize, CACHELINE, true, + NULL, true, arena_ichoose(tsd, NULL)); + if (ckh->tab == NULL) { + ret = true; + goto label_return; + } + + ret = false; +label_return: + return ret; +} + +void +ckh_delete(tsd_t *tsd, ckh_t *ckh) { + assert(ckh != NULL); + +#ifdef CKH_VERBOSE + malloc_printf( + "%s(%p): ngrows: %"FMTu64", nshrinks: %"FMTu64"," + " nshrinkfails: %"FMTu64", ninserts: %"FMTu64"," + " nrelocs: %"FMTu64"\n", __func__, ckh, + (unsigned long long)ckh->ngrows, + (unsigned long long)ckh->nshrinks, + (unsigned long long)ckh->nshrinkfails, + (unsigned long long)ckh->ninserts, + (unsigned long long)ckh->nrelocs); +#endif + + idalloctm(tsd_tsdn(tsd), ckh->tab, NULL, NULL, true, true); + if (config_debug) { + memset(ckh, JEMALLOC_FREE_JUNK, sizeof(ckh_t)); + } +} + +size_t +ckh_count(ckh_t *ckh) { + assert(ckh != NULL); + + return ckh->count; +} + +bool +ckh_iter(ckh_t *ckh, size_t *tabind, void **key, void **data) { + size_t i, ncells; + + for (i = *tabind, ncells = (ZU(1) << (ckh->lg_curbuckets + + LG_CKH_BUCKET_CELLS)); i < ncells; i++) { + if (ckh->tab[i].key != NULL) { + if (key != NULL) { + *key = (void *)ckh->tab[i].key; + } + if (data != NULL) { + *data = (void *)ckh->tab[i].data; + } + *tabind = i + 1; + return false; + } + } + + return true; +} + +bool +ckh_insert(tsd_t *tsd, ckh_t *ckh, const void *key, const void *data) { + bool ret; + + assert(ckh != NULL); + assert(ckh_search(ckh, key, NULL, NULL)); + +#ifdef CKH_COUNT + ckh->ninserts++; +#endif + + while (ckh_try_insert(ckh, &key, &data)) { + if (ckh_grow(tsd, ckh)) { + ret = true; + goto label_return; + } + } + + ret = false; +label_return: + return ret; +} + +bool +ckh_remove(tsd_t *tsd, ckh_t *ckh, const void *searchkey, void **key, + void **data) { + size_t cell; + + assert(ckh != NULL); + + cell = ckh_isearch(ckh, searchkey); + if (cell != SIZE_T_MAX) { + if (key != NULL) { + *key = (void *)ckh->tab[cell].key; + } + if (data != NULL) { + *data = (void *)ckh->tab[cell].data; + } + ckh->tab[cell].key = NULL; + ckh->tab[cell].data = NULL; /* Not necessary. */ + + ckh->count--; + /* Try to halve the table if it is less than 1/4 full. */ + if (ckh->count < (ZU(1) << (ckh->lg_curbuckets + + LG_CKH_BUCKET_CELLS - 2)) && ckh->lg_curbuckets + > ckh->lg_minbuckets) { + /* Ignore error due to OOM. */ + ckh_shrink(tsd, ckh); + } + + return false; + } + + return true; +} + +bool +ckh_search(ckh_t *ckh, const void *searchkey, void **key, void **data) { + size_t cell; + + assert(ckh != NULL); + + cell = ckh_isearch(ckh, searchkey); + if (cell != SIZE_T_MAX) { + if (key != NULL) { + *key = (void *)ckh->tab[cell].key; + } + if (data != NULL) { + *data = (void *)ckh->tab[cell].data; + } + return false; + } + + return true; +} + +void +ckh_string_hash(const void *key, size_t r_hash[2]) { + hash(key, strlen((const char *)key), 0x94122f33U, r_hash); +} + +bool +ckh_string_keycomp(const void *k1, const void *k2) { + assert(k1 != NULL); + assert(k2 != NULL); + + return !strcmp((char *)k1, (char *)k2); +} + +void +ckh_pointer_hash(const void *key, size_t r_hash[2]) { + union { + const void *v; + size_t i; + } u; + + assert(sizeof(u.v) == sizeof(u.i)); + u.v = key; + hash(&u.i, sizeof(u.i), 0xd983396eU, r_hash); +} + +bool +ckh_pointer_keycomp(const void *k1, const void *k2) { + return (k1 == k2); +} diff --git a/deps/jemalloc/src/ctl.c b/deps/jemalloc/src/ctl.c new file mode 100644 index 000000000..48afaa61f --- /dev/null +++ b/deps/jemalloc/src/ctl.c @@ -0,0 +1,3435 @@ +#define JEMALLOC_CTL_C_ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/assert.h" +#include "jemalloc/internal/ctl.h" +#include "jemalloc/internal/extent_dss.h" +#include "jemalloc/internal/extent_mmap.h" +#include "jemalloc/internal/mutex.h" +#include "jemalloc/internal/nstime.h" +#include "jemalloc/internal/sc.h" +#include "jemalloc/internal/util.h" + +/******************************************************************************/ +/* Data. */ + +/* + * ctl_mtx protects the following: + * - ctl_stats->* + */ +static malloc_mutex_t ctl_mtx; +static bool ctl_initialized; +static ctl_stats_t *ctl_stats; +static ctl_arenas_t *ctl_arenas; + +/******************************************************************************/ +/* Helpers for named and indexed nodes. */ + +static const ctl_named_node_t * +ctl_named_node(const ctl_node_t *node) { + return ((node->named) ? (const ctl_named_node_t *)node : NULL); +} + +static const ctl_named_node_t * +ctl_named_children(const ctl_named_node_t *node, size_t index) { + const ctl_named_node_t *children = ctl_named_node(node->children); + + return (children ? &children[index] : NULL); +} + +static const ctl_indexed_node_t * +ctl_indexed_node(const ctl_node_t *node) { + return (!node->named ? (const ctl_indexed_node_t *)node : NULL); +} + +/******************************************************************************/ +/* Function prototypes for non-inline static functions. */ + +#define CTL_PROTO(n) \ +static int n##_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, \ + void *oldp, size_t *oldlenp, void *newp, size_t newlen); + +#define INDEX_PROTO(n) \ +static const ctl_named_node_t *n##_index(tsdn_t *tsdn, \ + const size_t *mib, size_t miblen, size_t i); + +CTL_PROTO(version) +CTL_PROTO(epoch) +CTL_PROTO(background_thread) +CTL_PROTO(max_background_threads) +CTL_PROTO(thread_tcache_enabled) +CTL_PROTO(thread_tcache_flush) +CTL_PROTO(thread_prof_name) +CTL_PROTO(thread_prof_active) +CTL_PROTO(thread_arena) +CTL_PROTO(thread_allocated) +CTL_PROTO(thread_allocatedp) +CTL_PROTO(thread_deallocated) +CTL_PROTO(thread_deallocatedp) +CTL_PROTO(config_cache_oblivious) +CTL_PROTO(config_debug) +CTL_PROTO(config_fill) +CTL_PROTO(config_lazy_lock) +CTL_PROTO(config_malloc_conf) +CTL_PROTO(config_opt_safety_checks) +CTL_PROTO(config_prof) +CTL_PROTO(config_prof_libgcc) +CTL_PROTO(config_prof_libunwind) +CTL_PROTO(config_stats) +CTL_PROTO(config_utrace) +CTL_PROTO(config_xmalloc) +CTL_PROTO(opt_abort) +CTL_PROTO(opt_abort_conf) +CTL_PROTO(opt_confirm_conf) +CTL_PROTO(opt_metadata_thp) +CTL_PROTO(opt_retain) +CTL_PROTO(opt_dss) +CTL_PROTO(opt_narenas) +CTL_PROTO(opt_percpu_arena) +CTL_PROTO(opt_oversize_threshold) +CTL_PROTO(opt_background_thread) +CTL_PROTO(opt_max_background_threads) +CTL_PROTO(opt_dirty_decay_ms) +CTL_PROTO(opt_muzzy_decay_ms) +CTL_PROTO(opt_stats_print) +CTL_PROTO(opt_stats_print_opts) +CTL_PROTO(opt_junk) +CTL_PROTO(opt_zero) +CTL_PROTO(opt_utrace) +CTL_PROTO(opt_xmalloc) +CTL_PROTO(opt_tcache) +CTL_PROTO(opt_thp) +CTL_PROTO(opt_lg_extent_max_active_fit) +CTL_PROTO(opt_lg_tcache_max) +CTL_PROTO(opt_prof) +CTL_PROTO(opt_prof_prefix) +CTL_PROTO(opt_prof_active) +CTL_PROTO(opt_prof_thread_active_init) +CTL_PROTO(opt_lg_prof_sample) +CTL_PROTO(opt_lg_prof_interval) +CTL_PROTO(opt_prof_gdump) +CTL_PROTO(opt_prof_final) +CTL_PROTO(opt_prof_leak) +CTL_PROTO(opt_prof_accum) +CTL_PROTO(tcache_create) +CTL_PROTO(tcache_flush) +CTL_PROTO(tcache_destroy) +CTL_PROTO(arena_i_initialized) +CTL_PROTO(arena_i_decay) +CTL_PROTO(arena_i_purge) +CTL_PROTO(arena_i_reset) +CTL_PROTO(arena_i_destroy) +CTL_PROTO(arena_i_dss) +CTL_PROTO(arena_i_dirty_decay_ms) +CTL_PROTO(arena_i_muzzy_decay_ms) +CTL_PROTO(arena_i_extent_hooks) +CTL_PROTO(arena_i_retain_grow_limit) +INDEX_PROTO(arena_i) +CTL_PROTO(arenas_bin_i_size) +CTL_PROTO(arenas_bin_i_nregs) +CTL_PROTO(arenas_bin_i_slab_size) +CTL_PROTO(arenas_bin_i_nshards) +INDEX_PROTO(arenas_bin_i) +CTL_PROTO(arenas_lextent_i_size) +INDEX_PROTO(arenas_lextent_i) +CTL_PROTO(arenas_narenas) +CTL_PROTO(arenas_dirty_decay_ms) +CTL_PROTO(arenas_muzzy_decay_ms) +CTL_PROTO(arenas_quantum) +CTL_PROTO(arenas_page) +CTL_PROTO(arenas_tcache_max) +CTL_PROTO(arenas_nbins) +CTL_PROTO(arenas_nhbins) +CTL_PROTO(arenas_nlextents) +CTL_PROTO(arenas_create) +CTL_PROTO(arenas_lookup) +CTL_PROTO(prof_thread_active_init) +CTL_PROTO(prof_active) +CTL_PROTO(prof_dump) +CTL_PROTO(prof_gdump) +CTL_PROTO(prof_reset) +CTL_PROTO(prof_interval) +CTL_PROTO(lg_prof_sample) +CTL_PROTO(prof_log_start) +CTL_PROTO(prof_log_stop) +CTL_PROTO(stats_arenas_i_small_allocated) +CTL_PROTO(stats_arenas_i_small_nmalloc) +CTL_PROTO(stats_arenas_i_small_ndalloc) +CTL_PROTO(stats_arenas_i_small_nrequests) +CTL_PROTO(stats_arenas_i_small_nfills) +CTL_PROTO(stats_arenas_i_small_nflushes) +CTL_PROTO(stats_arenas_i_large_allocated) +CTL_PROTO(stats_arenas_i_large_nmalloc) +CTL_PROTO(stats_arenas_i_large_ndalloc) +CTL_PROTO(stats_arenas_i_large_nrequests) +CTL_PROTO(stats_arenas_i_large_nfills) +CTL_PROTO(stats_arenas_i_large_nflushes) +CTL_PROTO(stats_arenas_i_bins_j_nmalloc) +CTL_PROTO(stats_arenas_i_bins_j_ndalloc) +CTL_PROTO(stats_arenas_i_bins_j_nrequests) +CTL_PROTO(stats_arenas_i_bins_j_curregs) +CTL_PROTO(stats_arenas_i_bins_j_nfills) +CTL_PROTO(stats_arenas_i_bins_j_nflushes) +CTL_PROTO(stats_arenas_i_bins_j_nslabs) +CTL_PROTO(stats_arenas_i_bins_j_nreslabs) +CTL_PROTO(stats_arenas_i_bins_j_curslabs) +CTL_PROTO(stats_arenas_i_bins_j_nonfull_slabs) +INDEX_PROTO(stats_arenas_i_bins_j) +CTL_PROTO(stats_arenas_i_lextents_j_nmalloc) +CTL_PROTO(stats_arenas_i_lextents_j_ndalloc) +CTL_PROTO(stats_arenas_i_lextents_j_nrequests) +CTL_PROTO(stats_arenas_i_lextents_j_curlextents) +INDEX_PROTO(stats_arenas_i_lextents_j) +CTL_PROTO(stats_arenas_i_extents_j_ndirty) +CTL_PROTO(stats_arenas_i_extents_j_nmuzzy) +CTL_PROTO(stats_arenas_i_extents_j_nretained) +CTL_PROTO(stats_arenas_i_extents_j_dirty_bytes) +CTL_PROTO(stats_arenas_i_extents_j_muzzy_bytes) +CTL_PROTO(stats_arenas_i_extents_j_retained_bytes) +INDEX_PROTO(stats_arenas_i_extents_j) +CTL_PROTO(stats_arenas_i_nthreads) +CTL_PROTO(stats_arenas_i_uptime) +CTL_PROTO(stats_arenas_i_dss) +CTL_PROTO(stats_arenas_i_dirty_decay_ms) +CTL_PROTO(stats_arenas_i_muzzy_decay_ms) +CTL_PROTO(stats_arenas_i_pactive) +CTL_PROTO(stats_arenas_i_pdirty) +CTL_PROTO(stats_arenas_i_pmuzzy) +CTL_PROTO(stats_arenas_i_mapped) +CTL_PROTO(stats_arenas_i_retained) +CTL_PROTO(stats_arenas_i_extent_avail) +CTL_PROTO(stats_arenas_i_dirty_npurge) +CTL_PROTO(stats_arenas_i_dirty_nmadvise) +CTL_PROTO(stats_arenas_i_dirty_purged) +CTL_PROTO(stats_arenas_i_muzzy_npurge) +CTL_PROTO(stats_arenas_i_muzzy_nmadvise) +CTL_PROTO(stats_arenas_i_muzzy_purged) +CTL_PROTO(stats_arenas_i_base) +CTL_PROTO(stats_arenas_i_internal) +CTL_PROTO(stats_arenas_i_metadata_thp) +CTL_PROTO(stats_arenas_i_tcache_bytes) +CTL_PROTO(stats_arenas_i_resident) +CTL_PROTO(stats_arenas_i_abandoned_vm) +INDEX_PROTO(stats_arenas_i) +CTL_PROTO(stats_allocated) +CTL_PROTO(stats_active) +CTL_PROTO(stats_background_thread_num_threads) +CTL_PROTO(stats_background_thread_num_runs) +CTL_PROTO(stats_background_thread_run_interval) +CTL_PROTO(stats_metadata) +CTL_PROTO(stats_metadata_thp) +CTL_PROTO(stats_resident) +CTL_PROTO(stats_mapped) +CTL_PROTO(stats_retained) +CTL_PROTO(experimental_hooks_install) +CTL_PROTO(experimental_hooks_remove) +CTL_PROTO(experimental_utilization_query) +CTL_PROTO(experimental_utilization_batch_query) +CTL_PROTO(experimental_arenas_i_pactivep) +INDEX_PROTO(experimental_arenas_i) + +#define MUTEX_STATS_CTL_PROTO_GEN(n) \ +CTL_PROTO(stats_##n##_num_ops) \ +CTL_PROTO(stats_##n##_num_wait) \ +CTL_PROTO(stats_##n##_num_spin_acq) \ +CTL_PROTO(stats_##n##_num_owner_switch) \ +CTL_PROTO(stats_##n##_total_wait_time) \ +CTL_PROTO(stats_##n##_max_wait_time) \ +CTL_PROTO(stats_##n##_max_num_thds) + +/* Global mutexes. */ +#define OP(mtx) MUTEX_STATS_CTL_PROTO_GEN(mutexes_##mtx) +MUTEX_PROF_GLOBAL_MUTEXES +#undef OP + +/* Per arena mutexes. */ +#define OP(mtx) MUTEX_STATS_CTL_PROTO_GEN(arenas_i_mutexes_##mtx) +MUTEX_PROF_ARENA_MUTEXES +#undef OP + +/* Arena bin mutexes. */ +MUTEX_STATS_CTL_PROTO_GEN(arenas_i_bins_j_mutex) +#undef MUTEX_STATS_CTL_PROTO_GEN + +CTL_PROTO(stats_mutexes_reset) + +/******************************************************************************/ +/* mallctl tree. */ + +#define NAME(n) {true}, n +#define CHILD(t, c) \ + sizeof(c##_node) / sizeof(ctl_##t##_node_t), \ + (ctl_node_t *)c##_node, \ + NULL +#define CTL(c) 0, NULL, c##_ctl + +/* + * Only handles internal indexed nodes, since there are currently no external + * ones. + */ +#define INDEX(i) {false}, i##_index + +static const ctl_named_node_t thread_tcache_node[] = { + {NAME("enabled"), CTL(thread_tcache_enabled)}, + {NAME("flush"), CTL(thread_tcache_flush)} +}; + +static const ctl_named_node_t thread_prof_node[] = { + {NAME("name"), CTL(thread_prof_name)}, + {NAME("active"), CTL(thread_prof_active)} +}; + +static const ctl_named_node_t thread_node[] = { + {NAME("arena"), CTL(thread_arena)}, + {NAME("allocated"), CTL(thread_allocated)}, + {NAME("allocatedp"), CTL(thread_allocatedp)}, + {NAME("deallocated"), CTL(thread_deallocated)}, + {NAME("deallocatedp"), CTL(thread_deallocatedp)}, + {NAME("tcache"), CHILD(named, thread_tcache)}, + {NAME("prof"), CHILD(named, thread_prof)} +}; + +static const ctl_named_node_t config_node[] = { + {NAME("cache_oblivious"), CTL(config_cache_oblivious)}, + {NAME("debug"), CTL(config_debug)}, + {NAME("fill"), CTL(config_fill)}, + {NAME("lazy_lock"), CTL(config_lazy_lock)}, + {NAME("malloc_conf"), CTL(config_malloc_conf)}, + {NAME("opt_safety_checks"), CTL(config_opt_safety_checks)}, + {NAME("prof"), CTL(config_prof)}, + {NAME("prof_libgcc"), CTL(config_prof_libgcc)}, + {NAME("prof_libunwind"), CTL(config_prof_libunwind)}, + {NAME("stats"), CTL(config_stats)}, + {NAME("utrace"), CTL(config_utrace)}, + {NAME("xmalloc"), CTL(config_xmalloc)} +}; + +static const ctl_named_node_t opt_node[] = { + {NAME("abort"), CTL(opt_abort)}, + {NAME("abort_conf"), CTL(opt_abort_conf)}, + {NAME("confirm_conf"), CTL(opt_confirm_conf)}, + {NAME("metadata_thp"), CTL(opt_metadata_thp)}, + {NAME("retain"), CTL(opt_retain)}, + {NAME("dss"), CTL(opt_dss)}, + {NAME("narenas"), CTL(opt_narenas)}, + {NAME("percpu_arena"), CTL(opt_percpu_arena)}, + {NAME("oversize_threshold"), CTL(opt_oversize_threshold)}, + {NAME("background_thread"), CTL(opt_background_thread)}, + {NAME("max_background_threads"), CTL(opt_max_background_threads)}, + {NAME("dirty_decay_ms"), CTL(opt_dirty_decay_ms)}, + {NAME("muzzy_decay_ms"), CTL(opt_muzzy_decay_ms)}, + {NAME("stats_print"), CTL(opt_stats_print)}, + {NAME("stats_print_opts"), CTL(opt_stats_print_opts)}, + {NAME("junk"), CTL(opt_junk)}, + {NAME("zero"), CTL(opt_zero)}, + {NAME("utrace"), CTL(opt_utrace)}, + {NAME("xmalloc"), CTL(opt_xmalloc)}, + {NAME("tcache"), CTL(opt_tcache)}, + {NAME("thp"), CTL(opt_thp)}, + {NAME("lg_extent_max_active_fit"), CTL(opt_lg_extent_max_active_fit)}, + {NAME("lg_tcache_max"), CTL(opt_lg_tcache_max)}, + {NAME("prof"), CTL(opt_prof)}, + {NAME("prof_prefix"), CTL(opt_prof_prefix)}, + {NAME("prof_active"), CTL(opt_prof_active)}, + {NAME("prof_thread_active_init"), CTL(opt_prof_thread_active_init)}, + {NAME("lg_prof_sample"), CTL(opt_lg_prof_sample)}, + {NAME("lg_prof_interval"), CTL(opt_lg_prof_interval)}, + {NAME("prof_gdump"), CTL(opt_prof_gdump)}, + {NAME("prof_final"), CTL(opt_prof_final)}, + {NAME("prof_leak"), CTL(opt_prof_leak)}, + {NAME("prof_accum"), CTL(opt_prof_accum)} +}; + +static const ctl_named_node_t tcache_node[] = { + {NAME("create"), CTL(tcache_create)}, + {NAME("flush"), CTL(tcache_flush)}, + {NAME("destroy"), CTL(tcache_destroy)} +}; + +static const ctl_named_node_t arena_i_node[] = { + {NAME("initialized"), CTL(arena_i_initialized)}, + {NAME("decay"), CTL(arena_i_decay)}, + {NAME("purge"), CTL(arena_i_purge)}, + {NAME("reset"), CTL(arena_i_reset)}, + {NAME("destroy"), CTL(arena_i_destroy)}, + {NAME("dss"), CTL(arena_i_dss)}, + {NAME("dirty_decay_ms"), CTL(arena_i_dirty_decay_ms)}, + {NAME("muzzy_decay_ms"), CTL(arena_i_muzzy_decay_ms)}, + {NAME("extent_hooks"), CTL(arena_i_extent_hooks)}, + {NAME("retain_grow_limit"), CTL(arena_i_retain_grow_limit)} +}; +static const ctl_named_node_t super_arena_i_node[] = { + {NAME(""), CHILD(named, arena_i)} +}; + +static const ctl_indexed_node_t arena_node[] = { + {INDEX(arena_i)} +}; + +static const ctl_named_node_t arenas_bin_i_node[] = { + {NAME("size"), CTL(arenas_bin_i_size)}, + {NAME("nregs"), CTL(arenas_bin_i_nregs)}, + {NAME("slab_size"), CTL(arenas_bin_i_slab_size)}, + {NAME("nshards"), CTL(arenas_bin_i_nshards)} +}; +static const ctl_named_node_t super_arenas_bin_i_node[] = { + {NAME(""), CHILD(named, arenas_bin_i)} +}; + +static const ctl_indexed_node_t arenas_bin_node[] = { + {INDEX(arenas_bin_i)} +}; + +static const ctl_named_node_t arenas_lextent_i_node[] = { + {NAME("size"), CTL(arenas_lextent_i_size)} +}; +static const ctl_named_node_t super_arenas_lextent_i_node[] = { + {NAME(""), CHILD(named, arenas_lextent_i)} +}; + +static const ctl_indexed_node_t arenas_lextent_node[] = { + {INDEX(arenas_lextent_i)} +}; + +static const ctl_named_node_t arenas_node[] = { + {NAME("narenas"), CTL(arenas_narenas)}, + {NAME("dirty_decay_ms"), CTL(arenas_dirty_decay_ms)}, + {NAME("muzzy_decay_ms"), CTL(arenas_muzzy_decay_ms)}, + {NAME("quantum"), CTL(arenas_quantum)}, + {NAME("page"), CTL(arenas_page)}, + {NAME("tcache_max"), CTL(arenas_tcache_max)}, + {NAME("nbins"), CTL(arenas_nbins)}, + {NAME("nhbins"), CTL(arenas_nhbins)}, + {NAME("bin"), CHILD(indexed, arenas_bin)}, + {NAME("nlextents"), CTL(arenas_nlextents)}, + {NAME("lextent"), CHILD(indexed, arenas_lextent)}, + {NAME("create"), CTL(arenas_create)}, + {NAME("lookup"), CTL(arenas_lookup)} +}; + +static const ctl_named_node_t prof_node[] = { + {NAME("thread_active_init"), CTL(prof_thread_active_init)}, + {NAME("active"), CTL(prof_active)}, + {NAME("dump"), CTL(prof_dump)}, + {NAME("gdump"), CTL(prof_gdump)}, + {NAME("reset"), CTL(prof_reset)}, + {NAME("interval"), CTL(prof_interval)}, + {NAME("lg_sample"), CTL(lg_prof_sample)}, + {NAME("log_start"), CTL(prof_log_start)}, + {NAME("log_stop"), CTL(prof_log_stop)} +}; +static const ctl_named_node_t stats_arenas_i_small_node[] = { + {NAME("allocated"), CTL(stats_arenas_i_small_allocated)}, + {NAME("nmalloc"), CTL(stats_arenas_i_small_nmalloc)}, + {NAME("ndalloc"), CTL(stats_arenas_i_small_ndalloc)}, + {NAME("nrequests"), CTL(stats_arenas_i_small_nrequests)}, + {NAME("nfills"), CTL(stats_arenas_i_small_nfills)}, + {NAME("nflushes"), CTL(stats_arenas_i_small_nflushes)} +}; + +static const ctl_named_node_t stats_arenas_i_large_node[] = { + {NAME("allocated"), CTL(stats_arenas_i_large_allocated)}, + {NAME("nmalloc"), CTL(stats_arenas_i_large_nmalloc)}, + {NAME("ndalloc"), CTL(stats_arenas_i_large_ndalloc)}, + {NAME("nrequests"), CTL(stats_arenas_i_large_nrequests)}, + {NAME("nfills"), CTL(stats_arenas_i_large_nfills)}, + {NAME("nflushes"), CTL(stats_arenas_i_large_nflushes)} +}; + +#define MUTEX_PROF_DATA_NODE(prefix) \ +static const ctl_named_node_t stats_##prefix##_node[] = { \ + {NAME("num_ops"), \ + CTL(stats_##prefix##_num_ops)}, \ + {NAME("num_wait"), \ + CTL(stats_##prefix##_num_wait)}, \ + {NAME("num_spin_acq"), \ + CTL(stats_##prefix##_num_spin_acq)}, \ + {NAME("num_owner_switch"), \ + CTL(stats_##prefix##_num_owner_switch)}, \ + {NAME("total_wait_time"), \ + CTL(stats_##prefix##_total_wait_time)}, \ + {NAME("max_wait_time"), \ + CTL(stats_##prefix##_max_wait_time)}, \ + {NAME("max_num_thds"), \ + CTL(stats_##prefix##_max_num_thds)} \ + /* Note that # of current waiting thread not provided. */ \ +}; + +MUTEX_PROF_DATA_NODE(arenas_i_bins_j_mutex) + +static const ctl_named_node_t stats_arenas_i_bins_j_node[] = { + {NAME("nmalloc"), CTL(stats_arenas_i_bins_j_nmalloc)}, + {NAME("ndalloc"), CTL(stats_arenas_i_bins_j_ndalloc)}, + {NAME("nrequests"), CTL(stats_arenas_i_bins_j_nrequests)}, + {NAME("curregs"), CTL(stats_arenas_i_bins_j_curregs)}, + {NAME("nfills"), CTL(stats_arenas_i_bins_j_nfills)}, + {NAME("nflushes"), CTL(stats_arenas_i_bins_j_nflushes)}, + {NAME("nslabs"), CTL(stats_arenas_i_bins_j_nslabs)}, + {NAME("nreslabs"), CTL(stats_arenas_i_bins_j_nreslabs)}, + {NAME("curslabs"), CTL(stats_arenas_i_bins_j_curslabs)}, + {NAME("nonfull_slabs"), CTL(stats_arenas_i_bins_j_nonfull_slabs)}, + {NAME("mutex"), CHILD(named, stats_arenas_i_bins_j_mutex)} +}; + +static const ctl_named_node_t super_stats_arenas_i_bins_j_node[] = { + {NAME(""), CHILD(named, stats_arenas_i_bins_j)} +}; + +static const ctl_indexed_node_t stats_arenas_i_bins_node[] = { + {INDEX(stats_arenas_i_bins_j)} +}; + +static const ctl_named_node_t stats_arenas_i_lextents_j_node[] = { + {NAME("nmalloc"), CTL(stats_arenas_i_lextents_j_nmalloc)}, + {NAME("ndalloc"), CTL(stats_arenas_i_lextents_j_ndalloc)}, + {NAME("nrequests"), CTL(stats_arenas_i_lextents_j_nrequests)}, + {NAME("curlextents"), CTL(stats_arenas_i_lextents_j_curlextents)} +}; +static const ctl_named_node_t super_stats_arenas_i_lextents_j_node[] = { + {NAME(""), CHILD(named, stats_arenas_i_lextents_j)} +}; + +static const ctl_indexed_node_t stats_arenas_i_lextents_node[] = { + {INDEX(stats_arenas_i_lextents_j)} +}; + +static const ctl_named_node_t stats_arenas_i_extents_j_node[] = { + {NAME("ndirty"), CTL(stats_arenas_i_extents_j_ndirty)}, + {NAME("nmuzzy"), CTL(stats_arenas_i_extents_j_nmuzzy)}, + {NAME("nretained"), CTL(stats_arenas_i_extents_j_nretained)}, + {NAME("dirty_bytes"), CTL(stats_arenas_i_extents_j_dirty_bytes)}, + {NAME("muzzy_bytes"), CTL(stats_arenas_i_extents_j_muzzy_bytes)}, + {NAME("retained_bytes"), CTL(stats_arenas_i_extents_j_retained_bytes)} +}; + +static const ctl_named_node_t super_stats_arenas_i_extents_j_node[] = { + {NAME(""), CHILD(named, stats_arenas_i_extents_j)} +}; + +static const ctl_indexed_node_t stats_arenas_i_extents_node[] = { + {INDEX(stats_arenas_i_extents_j)} +}; + +#define OP(mtx) MUTEX_PROF_DATA_NODE(arenas_i_mutexes_##mtx) +MUTEX_PROF_ARENA_MUTEXES +#undef OP + +static const ctl_named_node_t stats_arenas_i_mutexes_node[] = { +#define OP(mtx) {NAME(#mtx), CHILD(named, stats_arenas_i_mutexes_##mtx)}, +MUTEX_PROF_ARENA_MUTEXES +#undef OP +}; + +static const ctl_named_node_t stats_arenas_i_node[] = { + {NAME("nthreads"), CTL(stats_arenas_i_nthreads)}, + {NAME("uptime"), CTL(stats_arenas_i_uptime)}, + {NAME("dss"), CTL(stats_arenas_i_dss)}, + {NAME("dirty_decay_ms"), CTL(stats_arenas_i_dirty_decay_ms)}, + {NAME("muzzy_decay_ms"), CTL(stats_arenas_i_muzzy_decay_ms)}, + {NAME("pactive"), CTL(stats_arenas_i_pactive)}, + {NAME("pdirty"), CTL(stats_arenas_i_pdirty)}, + {NAME("pmuzzy"), CTL(stats_arenas_i_pmuzzy)}, + {NAME("mapped"), CTL(stats_arenas_i_mapped)}, + {NAME("retained"), CTL(stats_arenas_i_retained)}, + {NAME("extent_avail"), CTL(stats_arenas_i_extent_avail)}, + {NAME("dirty_npurge"), CTL(stats_arenas_i_dirty_npurge)}, + {NAME("dirty_nmadvise"), CTL(stats_arenas_i_dirty_nmadvise)}, + {NAME("dirty_purged"), CTL(stats_arenas_i_dirty_purged)}, + {NAME("muzzy_npurge"), CTL(stats_arenas_i_muzzy_npurge)}, + {NAME("muzzy_nmadvise"), CTL(stats_arenas_i_muzzy_nmadvise)}, + {NAME("muzzy_purged"), CTL(stats_arenas_i_muzzy_purged)}, + {NAME("base"), CTL(stats_arenas_i_base)}, + {NAME("internal"), CTL(stats_arenas_i_internal)}, + {NAME("metadata_thp"), CTL(stats_arenas_i_metadata_thp)}, + {NAME("tcache_bytes"), CTL(stats_arenas_i_tcache_bytes)}, + {NAME("resident"), CTL(stats_arenas_i_resident)}, + {NAME("abandoned_vm"), CTL(stats_arenas_i_abandoned_vm)}, + {NAME("small"), CHILD(named, stats_arenas_i_small)}, + {NAME("large"), CHILD(named, stats_arenas_i_large)}, + {NAME("bins"), CHILD(indexed, stats_arenas_i_bins)}, + {NAME("lextents"), CHILD(indexed, stats_arenas_i_lextents)}, + {NAME("extents"), CHILD(indexed, stats_arenas_i_extents)}, + {NAME("mutexes"), CHILD(named, stats_arenas_i_mutexes)} +}; +static const ctl_named_node_t super_stats_arenas_i_node[] = { + {NAME(""), CHILD(named, stats_arenas_i)} +}; + +static const ctl_indexed_node_t stats_arenas_node[] = { + {INDEX(stats_arenas_i)} +}; + +static const ctl_named_node_t stats_background_thread_node[] = { + {NAME("num_threads"), CTL(stats_background_thread_num_threads)}, + {NAME("num_runs"), CTL(stats_background_thread_num_runs)}, + {NAME("run_interval"), CTL(stats_background_thread_run_interval)} +}; + +#define OP(mtx) MUTEX_PROF_DATA_NODE(mutexes_##mtx) +MUTEX_PROF_GLOBAL_MUTEXES +#undef OP + +static const ctl_named_node_t stats_mutexes_node[] = { +#define OP(mtx) {NAME(#mtx), CHILD(named, stats_mutexes_##mtx)}, +MUTEX_PROF_GLOBAL_MUTEXES +#undef OP + {NAME("reset"), CTL(stats_mutexes_reset)} +}; +#undef MUTEX_PROF_DATA_NODE + +static const ctl_named_node_t stats_node[] = { + {NAME("allocated"), CTL(stats_allocated)}, + {NAME("active"), CTL(stats_active)}, + {NAME("metadata"), CTL(stats_metadata)}, + {NAME("metadata_thp"), CTL(stats_metadata_thp)}, + {NAME("resident"), CTL(stats_resident)}, + {NAME("mapped"), CTL(stats_mapped)}, + {NAME("retained"), CTL(stats_retained)}, + {NAME("background_thread"), + CHILD(named, stats_background_thread)}, + {NAME("mutexes"), CHILD(named, stats_mutexes)}, + {NAME("arenas"), CHILD(indexed, stats_arenas)} +}; + +static const ctl_named_node_t experimental_hooks_node[] = { + {NAME("install"), CTL(experimental_hooks_install)}, + {NAME("remove"), CTL(experimental_hooks_remove)} +}; + +static const ctl_named_node_t experimental_utilization_node[] = { + {NAME("query"), CTL(experimental_utilization_query)}, + {NAME("batch_query"), CTL(experimental_utilization_batch_query)} +}; + +static const ctl_named_node_t experimental_arenas_i_node[] = { + {NAME("pactivep"), CTL(experimental_arenas_i_pactivep)} +}; +static const ctl_named_node_t super_experimental_arenas_i_node[] = { + {NAME(""), CHILD(named, experimental_arenas_i)} +}; + +static const ctl_indexed_node_t experimental_arenas_node[] = { + {INDEX(experimental_arenas_i)} +}; + +static const ctl_named_node_t experimental_node[] = { + {NAME("hooks"), CHILD(named, experimental_hooks)}, + {NAME("utilization"), CHILD(named, experimental_utilization)}, + {NAME("arenas"), CHILD(indexed, experimental_arenas)} +}; + +static const ctl_named_node_t root_node[] = { + {NAME("version"), CTL(version)}, + {NAME("epoch"), CTL(epoch)}, + {NAME("background_thread"), CTL(background_thread)}, + {NAME("max_background_threads"), CTL(max_background_threads)}, + {NAME("thread"), CHILD(named, thread)}, + {NAME("config"), CHILD(named, config)}, + {NAME("opt"), CHILD(named, opt)}, + {NAME("tcache"), CHILD(named, tcache)}, + {NAME("arena"), CHILD(indexed, arena)}, + {NAME("arenas"), CHILD(named, arenas)}, + {NAME("prof"), CHILD(named, prof)}, + {NAME("stats"), CHILD(named, stats)}, + {NAME("experimental"), CHILD(named, experimental)} +}; +static const ctl_named_node_t super_root_node[] = { + {NAME(""), CHILD(named, root)} +}; + +#undef NAME +#undef CHILD +#undef CTL +#undef INDEX + +/******************************************************************************/ + +/* + * Sets *dst + *src non-atomically. This is safe, since everything is + * synchronized by the ctl mutex. + */ +static void +ctl_accum_arena_stats_u64(arena_stats_u64_t *dst, arena_stats_u64_t *src) { +#ifdef JEMALLOC_ATOMIC_U64 + uint64_t cur_dst = atomic_load_u64(dst, ATOMIC_RELAXED); + uint64_t cur_src = atomic_load_u64(src, ATOMIC_RELAXED); + atomic_store_u64(dst, cur_dst + cur_src, ATOMIC_RELAXED); +#else + *dst += *src; +#endif +} + +/* Likewise: with ctl mutex synchronization, reading is simple. */ +static uint64_t +ctl_arena_stats_read_u64(arena_stats_u64_t *p) { +#ifdef JEMALLOC_ATOMIC_U64 + return atomic_load_u64(p, ATOMIC_RELAXED); +#else + return *p; +#endif +} + +static void +accum_atomic_zu(atomic_zu_t *dst, atomic_zu_t *src) { + size_t cur_dst = atomic_load_zu(dst, ATOMIC_RELAXED); + size_t cur_src = atomic_load_zu(src, ATOMIC_RELAXED); + atomic_store_zu(dst, cur_dst + cur_src, ATOMIC_RELAXED); +} + +/******************************************************************************/ + +static unsigned +arenas_i2a_impl(size_t i, bool compat, bool validate) { + unsigned a; + + switch (i) { + case MALLCTL_ARENAS_ALL: + a = 0; + break; + case MALLCTL_ARENAS_DESTROYED: + a = 1; + break; + default: + if (compat && i == ctl_arenas->narenas) { + /* + * Provide deprecated backward compatibility for + * accessing the merged stats at index narenas rather + * than via MALLCTL_ARENAS_ALL. This is scheduled for + * removal in 6.0.0. + */ + a = 0; + } else if (validate && i >= ctl_arenas->narenas) { + a = UINT_MAX; + } else { + /* + * This function should never be called for an index + * more than one past the range of indices that have + * initialized ctl data. + */ + assert(i < ctl_arenas->narenas || (!validate && i == + ctl_arenas->narenas)); + a = (unsigned)i + 2; + } + break; + } + + return a; +} + +static unsigned +arenas_i2a(size_t i) { + return arenas_i2a_impl(i, true, false); +} + +static ctl_arena_t * +arenas_i_impl(tsd_t *tsd, size_t i, bool compat, bool init) { + ctl_arena_t *ret; + + assert(!compat || !init); + + ret = ctl_arenas->arenas[arenas_i2a_impl(i, compat, false)]; + if (init && ret == NULL) { + if (config_stats) { + struct container_s { + ctl_arena_t ctl_arena; + ctl_arena_stats_t astats; + }; + struct container_s *cont = + (struct container_s *)base_alloc(tsd_tsdn(tsd), + b0get(), sizeof(struct container_s), QUANTUM); + if (cont == NULL) { + return NULL; + } + ret = &cont->ctl_arena; + ret->astats = &cont->astats; + } else { + ret = (ctl_arena_t *)base_alloc(tsd_tsdn(tsd), b0get(), + sizeof(ctl_arena_t), QUANTUM); + if (ret == NULL) { + return NULL; + } + } + ret->arena_ind = (unsigned)i; + ctl_arenas->arenas[arenas_i2a_impl(i, compat, false)] = ret; + } + + assert(ret == NULL || arenas_i2a(ret->arena_ind) == arenas_i2a(i)); + return ret; +} + +static ctl_arena_t * +arenas_i(size_t i) { + ctl_arena_t *ret = arenas_i_impl(tsd_fetch(), i, true, false); + assert(ret != NULL); + return ret; +} + +static void +ctl_arena_clear(ctl_arena_t *ctl_arena) { + ctl_arena->nthreads = 0; + ctl_arena->dss = dss_prec_names[dss_prec_limit]; + ctl_arena->dirty_decay_ms = -1; + ctl_arena->muzzy_decay_ms = -1; + ctl_arena->pactive = 0; + ctl_arena->pdirty = 0; + ctl_arena->pmuzzy = 0; + if (config_stats) { + memset(&ctl_arena->astats->astats, 0, sizeof(arena_stats_t)); + ctl_arena->astats->allocated_small = 0; + ctl_arena->astats->nmalloc_small = 0; + ctl_arena->astats->ndalloc_small = 0; + ctl_arena->astats->nrequests_small = 0; + ctl_arena->astats->nfills_small = 0; + ctl_arena->astats->nflushes_small = 0; + memset(ctl_arena->astats->bstats, 0, SC_NBINS * + sizeof(bin_stats_t)); + memset(ctl_arena->astats->lstats, 0, (SC_NSIZES - SC_NBINS) * + sizeof(arena_stats_large_t)); + memset(ctl_arena->astats->estats, 0, SC_NPSIZES * + sizeof(arena_stats_extents_t)); + } +} + +static void +ctl_arena_stats_amerge(tsdn_t *tsdn, ctl_arena_t *ctl_arena, arena_t *arena) { + unsigned i; + + if (config_stats) { + arena_stats_merge(tsdn, arena, &ctl_arena->nthreads, + &ctl_arena->dss, &ctl_arena->dirty_decay_ms, + &ctl_arena->muzzy_decay_ms, &ctl_arena->pactive, + &ctl_arena->pdirty, &ctl_arena->pmuzzy, + &ctl_arena->astats->astats, ctl_arena->astats->bstats, + ctl_arena->astats->lstats, ctl_arena->astats->estats); + + for (i = 0; i < SC_NBINS; i++) { + ctl_arena->astats->allocated_small += + ctl_arena->astats->bstats[i].curregs * + sz_index2size(i); + ctl_arena->astats->nmalloc_small += + ctl_arena->astats->bstats[i].nmalloc; + ctl_arena->astats->ndalloc_small += + ctl_arena->astats->bstats[i].ndalloc; + ctl_arena->astats->nrequests_small += + ctl_arena->astats->bstats[i].nrequests; + ctl_arena->astats->nfills_small += + ctl_arena->astats->bstats[i].nfills; + ctl_arena->astats->nflushes_small += + ctl_arena->astats->bstats[i].nflushes; + } + } else { + arena_basic_stats_merge(tsdn, arena, &ctl_arena->nthreads, + &ctl_arena->dss, &ctl_arena->dirty_decay_ms, + &ctl_arena->muzzy_decay_ms, &ctl_arena->pactive, + &ctl_arena->pdirty, &ctl_arena->pmuzzy); + } +} + +static void +ctl_arena_stats_sdmerge(ctl_arena_t *ctl_sdarena, ctl_arena_t *ctl_arena, + bool destroyed) { + unsigned i; + + if (!destroyed) { + ctl_sdarena->nthreads += ctl_arena->nthreads; + ctl_sdarena->pactive += ctl_arena->pactive; + ctl_sdarena->pdirty += ctl_arena->pdirty; + ctl_sdarena->pmuzzy += ctl_arena->pmuzzy; + } else { + assert(ctl_arena->nthreads == 0); + assert(ctl_arena->pactive == 0); + assert(ctl_arena->pdirty == 0); + assert(ctl_arena->pmuzzy == 0); + } + + if (config_stats) { + ctl_arena_stats_t *sdstats = ctl_sdarena->astats; + ctl_arena_stats_t *astats = ctl_arena->astats; + + if (!destroyed) { + accum_atomic_zu(&sdstats->astats.mapped, + &astats->astats.mapped); + accum_atomic_zu(&sdstats->astats.retained, + &astats->astats.retained); + accum_atomic_zu(&sdstats->astats.extent_avail, + &astats->astats.extent_avail); + } + + ctl_accum_arena_stats_u64(&sdstats->astats.decay_dirty.npurge, + &astats->astats.decay_dirty.npurge); + ctl_accum_arena_stats_u64(&sdstats->astats.decay_dirty.nmadvise, + &astats->astats.decay_dirty.nmadvise); + ctl_accum_arena_stats_u64(&sdstats->astats.decay_dirty.purged, + &astats->astats.decay_dirty.purged); + + ctl_accum_arena_stats_u64(&sdstats->astats.decay_muzzy.npurge, + &astats->astats.decay_muzzy.npurge); + ctl_accum_arena_stats_u64(&sdstats->astats.decay_muzzy.nmadvise, + &astats->astats.decay_muzzy.nmadvise); + ctl_accum_arena_stats_u64(&sdstats->astats.decay_muzzy.purged, + &astats->astats.decay_muzzy.purged); + +#define OP(mtx) malloc_mutex_prof_merge( \ + &(sdstats->astats.mutex_prof_data[ \ + arena_prof_mutex_##mtx]), \ + &(astats->astats.mutex_prof_data[ \ + arena_prof_mutex_##mtx])); +MUTEX_PROF_ARENA_MUTEXES +#undef OP + if (!destroyed) { + accum_atomic_zu(&sdstats->astats.base, + &astats->astats.base); + accum_atomic_zu(&sdstats->astats.internal, + &astats->astats.internal); + accum_atomic_zu(&sdstats->astats.resident, + &astats->astats.resident); + accum_atomic_zu(&sdstats->astats.metadata_thp, + &astats->astats.metadata_thp); + } else { + assert(atomic_load_zu( + &astats->astats.internal, ATOMIC_RELAXED) == 0); + } + + if (!destroyed) { + sdstats->allocated_small += astats->allocated_small; + } else { + assert(astats->allocated_small == 0); + } + sdstats->nmalloc_small += astats->nmalloc_small; + sdstats->ndalloc_small += astats->ndalloc_small; + sdstats->nrequests_small += astats->nrequests_small; + sdstats->nfills_small += astats->nfills_small; + sdstats->nflushes_small += astats->nflushes_small; + + if (!destroyed) { + accum_atomic_zu(&sdstats->astats.allocated_large, + &astats->astats.allocated_large); + } else { + assert(atomic_load_zu(&astats->astats.allocated_large, + ATOMIC_RELAXED) == 0); + } + ctl_accum_arena_stats_u64(&sdstats->astats.nmalloc_large, + &astats->astats.nmalloc_large); + ctl_accum_arena_stats_u64(&sdstats->astats.ndalloc_large, + &astats->astats.ndalloc_large); + ctl_accum_arena_stats_u64(&sdstats->astats.nrequests_large, + &astats->astats.nrequests_large); + accum_atomic_zu(&sdstats->astats.abandoned_vm, + &astats->astats.abandoned_vm); + + accum_atomic_zu(&sdstats->astats.tcache_bytes, + &astats->astats.tcache_bytes); + + if (ctl_arena->arena_ind == 0) { + sdstats->astats.uptime = astats->astats.uptime; + } + + /* Merge bin stats. */ + for (i = 0; i < SC_NBINS; i++) { + sdstats->bstats[i].nmalloc += astats->bstats[i].nmalloc; + sdstats->bstats[i].ndalloc += astats->bstats[i].ndalloc; + sdstats->bstats[i].nrequests += + astats->bstats[i].nrequests; + if (!destroyed) { + sdstats->bstats[i].curregs += + astats->bstats[i].curregs; + } else { + assert(astats->bstats[i].curregs == 0); + } + sdstats->bstats[i].nfills += astats->bstats[i].nfills; + sdstats->bstats[i].nflushes += + astats->bstats[i].nflushes; + sdstats->bstats[i].nslabs += astats->bstats[i].nslabs; + sdstats->bstats[i].reslabs += astats->bstats[i].reslabs; + if (!destroyed) { + sdstats->bstats[i].curslabs += + astats->bstats[i].curslabs; + sdstats->bstats[i].nonfull_slabs += + astats->bstats[i].nonfull_slabs; + } else { + assert(astats->bstats[i].curslabs == 0); + assert(astats->bstats[i].nonfull_slabs == 0); + } + malloc_mutex_prof_merge(&sdstats->bstats[i].mutex_data, + &astats->bstats[i].mutex_data); + } + + /* Merge stats for large allocations. */ + for (i = 0; i < SC_NSIZES - SC_NBINS; i++) { + ctl_accum_arena_stats_u64(&sdstats->lstats[i].nmalloc, + &astats->lstats[i].nmalloc); + ctl_accum_arena_stats_u64(&sdstats->lstats[i].ndalloc, + &astats->lstats[i].ndalloc); + ctl_accum_arena_stats_u64(&sdstats->lstats[i].nrequests, + &astats->lstats[i].nrequests); + if (!destroyed) { + sdstats->lstats[i].curlextents += + astats->lstats[i].curlextents; + } else { + assert(astats->lstats[i].curlextents == 0); + } + } + + /* Merge extents stats. */ + for (i = 0; i < SC_NPSIZES; i++) { + accum_atomic_zu(&sdstats->estats[i].ndirty, + &astats->estats[i].ndirty); + accum_atomic_zu(&sdstats->estats[i].nmuzzy, + &astats->estats[i].nmuzzy); + accum_atomic_zu(&sdstats->estats[i].nretained, + &astats->estats[i].nretained); + accum_atomic_zu(&sdstats->estats[i].dirty_bytes, + &astats->estats[i].dirty_bytes); + accum_atomic_zu(&sdstats->estats[i].muzzy_bytes, + &astats->estats[i].muzzy_bytes); + accum_atomic_zu(&sdstats->estats[i].retained_bytes, + &astats->estats[i].retained_bytes); + } + } +} + +static void +ctl_arena_refresh(tsdn_t *tsdn, arena_t *arena, ctl_arena_t *ctl_sdarena, + unsigned i, bool destroyed) { + ctl_arena_t *ctl_arena = arenas_i(i); + + ctl_arena_clear(ctl_arena); + ctl_arena_stats_amerge(tsdn, ctl_arena, arena); + /* Merge into sum stats as well. */ + ctl_arena_stats_sdmerge(ctl_sdarena, ctl_arena, destroyed); +} + +static unsigned +ctl_arena_init(tsd_t *tsd, extent_hooks_t *extent_hooks) { + unsigned arena_ind; + ctl_arena_t *ctl_arena; + + if ((ctl_arena = ql_last(&ctl_arenas->destroyed, destroyed_link)) != + NULL) { + ql_remove(&ctl_arenas->destroyed, ctl_arena, destroyed_link); + arena_ind = ctl_arena->arena_ind; + } else { + arena_ind = ctl_arenas->narenas; + } + + /* Trigger stats allocation. */ + if (arenas_i_impl(tsd, arena_ind, false, true) == NULL) { + return UINT_MAX; + } + + /* Initialize new arena. */ + if (arena_init(tsd_tsdn(tsd), arena_ind, extent_hooks) == NULL) { + return UINT_MAX; + } + + if (arena_ind == ctl_arenas->narenas) { + ctl_arenas->narenas++; + } + + return arena_ind; +} + +static void +ctl_background_thread_stats_read(tsdn_t *tsdn) { + background_thread_stats_t *stats = &ctl_stats->background_thread; + if (!have_background_thread || + background_thread_stats_read(tsdn, stats)) { + memset(stats, 0, sizeof(background_thread_stats_t)); + nstime_init(&stats->run_interval, 0); + } +} + +static void +ctl_refresh(tsdn_t *tsdn) { + unsigned i; + ctl_arena_t *ctl_sarena = arenas_i(MALLCTL_ARENAS_ALL); + VARIABLE_ARRAY(arena_t *, tarenas, ctl_arenas->narenas); + + /* + * Clear sum stats, since they will be merged into by + * ctl_arena_refresh(). + */ + ctl_arena_clear(ctl_sarena); + + for (i = 0; i < ctl_arenas->narenas; i++) { + tarenas[i] = arena_get(tsdn, i, false); + } + + for (i = 0; i < ctl_arenas->narenas; i++) { + ctl_arena_t *ctl_arena = arenas_i(i); + bool initialized = (tarenas[i] != NULL); + + ctl_arena->initialized = initialized; + if (initialized) { + ctl_arena_refresh(tsdn, tarenas[i], ctl_sarena, i, + false); + } + } + + if (config_stats) { + ctl_stats->allocated = ctl_sarena->astats->allocated_small + + atomic_load_zu(&ctl_sarena->astats->astats.allocated_large, + ATOMIC_RELAXED); + ctl_stats->active = (ctl_sarena->pactive << LG_PAGE); + ctl_stats->metadata = atomic_load_zu( + &ctl_sarena->astats->astats.base, ATOMIC_RELAXED) + + atomic_load_zu(&ctl_sarena->astats->astats.internal, + ATOMIC_RELAXED); + ctl_stats->metadata_thp = atomic_load_zu( + &ctl_sarena->astats->astats.metadata_thp, ATOMIC_RELAXED); + ctl_stats->resident = atomic_load_zu( + &ctl_sarena->astats->astats.resident, ATOMIC_RELAXED); + ctl_stats->mapped = atomic_load_zu( + &ctl_sarena->astats->astats.mapped, ATOMIC_RELAXED); + ctl_stats->retained = atomic_load_zu( + &ctl_sarena->astats->astats.retained, ATOMIC_RELAXED); + + ctl_background_thread_stats_read(tsdn); + +#define READ_GLOBAL_MUTEX_PROF_DATA(i, mtx) \ + malloc_mutex_lock(tsdn, &mtx); \ + malloc_mutex_prof_read(tsdn, &ctl_stats->mutex_prof_data[i], &mtx); \ + malloc_mutex_unlock(tsdn, &mtx); + + if (config_prof && opt_prof) { + READ_GLOBAL_MUTEX_PROF_DATA(global_prof_mutex_prof, + bt2gctx_mtx); + } + if (have_background_thread) { + READ_GLOBAL_MUTEX_PROF_DATA( + global_prof_mutex_background_thread, + background_thread_lock); + } else { + memset(&ctl_stats->mutex_prof_data[ + global_prof_mutex_background_thread], 0, + sizeof(mutex_prof_data_t)); + } + /* We own ctl mutex already. */ + malloc_mutex_prof_read(tsdn, + &ctl_stats->mutex_prof_data[global_prof_mutex_ctl], + &ctl_mtx); +#undef READ_GLOBAL_MUTEX_PROF_DATA + } + ctl_arenas->epoch++; +} + +static bool +ctl_init(tsd_t *tsd) { + bool ret; + tsdn_t *tsdn = tsd_tsdn(tsd); + + malloc_mutex_lock(tsdn, &ctl_mtx); + if (!ctl_initialized) { + ctl_arena_t *ctl_sarena, *ctl_darena; + unsigned i; + + /* + * Allocate demand-zeroed space for pointers to the full + * range of supported arena indices. + */ + if (ctl_arenas == NULL) { + ctl_arenas = (ctl_arenas_t *)base_alloc(tsdn, + b0get(), sizeof(ctl_arenas_t), QUANTUM); + if (ctl_arenas == NULL) { + ret = true; + goto label_return; + } + } + + if (config_stats && ctl_stats == NULL) { + ctl_stats = (ctl_stats_t *)base_alloc(tsdn, b0get(), + sizeof(ctl_stats_t), QUANTUM); + if (ctl_stats == NULL) { + ret = true; + goto label_return; + } + } + + /* + * Allocate space for the current full range of arenas + * here rather than doing it lazily elsewhere, in order + * to limit when OOM-caused errors can occur. + */ + if ((ctl_sarena = arenas_i_impl(tsd, MALLCTL_ARENAS_ALL, false, + true)) == NULL) { + ret = true; + goto label_return; + } + ctl_sarena->initialized = true; + + if ((ctl_darena = arenas_i_impl(tsd, MALLCTL_ARENAS_DESTROYED, + false, true)) == NULL) { + ret = true; + goto label_return; + } + ctl_arena_clear(ctl_darena); + /* + * Don't toggle ctl_darena to initialized until an arena is + * actually destroyed, so that arena.<i>.initialized can be used + * to query whether the stats are relevant. + */ + + ctl_arenas->narenas = narenas_total_get(); + for (i = 0; i < ctl_arenas->narenas; i++) { + if (arenas_i_impl(tsd, i, false, true) == NULL) { + ret = true; + goto label_return; + } + } + + ql_new(&ctl_arenas->destroyed); + ctl_refresh(tsdn); + + ctl_initialized = true; + } + + ret = false; +label_return: + malloc_mutex_unlock(tsdn, &ctl_mtx); + return ret; +} + +static int +ctl_lookup(tsdn_t *tsdn, const char *name, ctl_node_t const **nodesp, + size_t *mibp, size_t *depthp) { + int ret; + const char *elm, *tdot, *dot; + size_t elen, i, j; + const ctl_named_node_t *node; + + elm = name; + /* Equivalent to strchrnul(). */ + dot = ((tdot = strchr(elm, '.')) != NULL) ? tdot : strchr(elm, '\0'); + elen = (size_t)((uintptr_t)dot - (uintptr_t)elm); + if (elen == 0) { + ret = ENOENT; + goto label_return; + } + node = super_root_node; + for (i = 0; i < *depthp; i++) { + assert(node); + assert(node->nchildren > 0); + if (ctl_named_node(node->children) != NULL) { + const ctl_named_node_t *pnode = node; + + /* Children are named. */ + for (j = 0; j < node->nchildren; j++) { + const ctl_named_node_t *child = + ctl_named_children(node, j); + if (strlen(child->name) == elen && + strncmp(elm, child->name, elen) == 0) { + node = child; + if (nodesp != NULL) { + nodesp[i] = + (const ctl_node_t *)node; + } + mibp[i] = j; + break; + } + } + if (node == pnode) { + ret = ENOENT; + goto label_return; + } + } else { + uintmax_t index; + const ctl_indexed_node_t *inode; + + /* Children are indexed. */ + index = malloc_strtoumax(elm, NULL, 10); + if (index == UINTMAX_MAX || index > SIZE_T_MAX) { + ret = ENOENT; + goto label_return; + } + + inode = ctl_indexed_node(node->children); + node = inode->index(tsdn, mibp, *depthp, (size_t)index); + if (node == NULL) { + ret = ENOENT; + goto label_return; + } + + if (nodesp != NULL) { + nodesp[i] = (const ctl_node_t *)node; + } + mibp[i] = (size_t)index; + } + + if (node->ctl != NULL) { + /* Terminal node. */ + if (*dot != '\0') { + /* + * The name contains more elements than are + * in this path through the tree. + */ + ret = ENOENT; + goto label_return; + } + /* Complete lookup successful. */ + *depthp = i + 1; + break; + } + + /* Update elm. */ + if (*dot == '\0') { + /* No more elements. */ + ret = ENOENT; + goto label_return; + } + elm = &dot[1]; + dot = ((tdot = strchr(elm, '.')) != NULL) ? tdot : + strchr(elm, '\0'); + elen = (size_t)((uintptr_t)dot - (uintptr_t)elm); + } + + ret = 0; +label_return: + return ret; +} + +int +ctl_byname(tsd_t *tsd, const char *name, void *oldp, size_t *oldlenp, + void *newp, size_t newlen) { + int ret; + size_t depth; + ctl_node_t const *nodes[CTL_MAX_DEPTH]; + size_t mib[CTL_MAX_DEPTH]; + const ctl_named_node_t *node; + + if (!ctl_initialized && ctl_init(tsd)) { + ret = EAGAIN; + goto label_return; + } + + depth = CTL_MAX_DEPTH; + ret = ctl_lookup(tsd_tsdn(tsd), name, nodes, mib, &depth); + if (ret != 0) { + goto label_return; + } + + node = ctl_named_node(nodes[depth-1]); + if (node != NULL && node->ctl) { + ret = node->ctl(tsd, mib, depth, oldp, oldlenp, newp, newlen); + } else { + /* The name refers to a partial path through the ctl tree. */ + ret = ENOENT; + } + +label_return: + return(ret); +} + +int +ctl_nametomib(tsd_t *tsd, const char *name, size_t *mibp, size_t *miblenp) { + int ret; + + if (!ctl_initialized && ctl_init(tsd)) { + ret = EAGAIN; + goto label_return; + } + + ret = ctl_lookup(tsd_tsdn(tsd), name, NULL, mibp, miblenp); +label_return: + return(ret); +} + +int +ctl_bymib(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp, + size_t *oldlenp, void *newp, size_t newlen) { + int ret; + const ctl_named_node_t *node; + size_t i; + + if (!ctl_initialized && ctl_init(tsd)) { + ret = EAGAIN; + goto label_return; + } + + /* Iterate down the tree. */ + node = super_root_node; + for (i = 0; i < miblen; i++) { + assert(node); + assert(node->nchildren > 0); + if (ctl_named_node(node->children) != NULL) { + /* Children are named. */ + if (node->nchildren <= mib[i]) { + ret = ENOENT; + goto label_return; + } + node = ctl_named_children(node, mib[i]); + } else { + const ctl_indexed_node_t *inode; + + /* Indexed element. */ + inode = ctl_indexed_node(node->children); + node = inode->index(tsd_tsdn(tsd), mib, miblen, mib[i]); + if (node == NULL) { + ret = ENOENT; + goto label_return; + } + } + } + + /* Call the ctl function. */ + if (node && node->ctl) { + ret = node->ctl(tsd, mib, miblen, oldp, oldlenp, newp, newlen); + } else { + /* Partial MIB. */ + ret = ENOENT; + } + +label_return: + return(ret); +} + +bool +ctl_boot(void) { + if (malloc_mutex_init(&ctl_mtx, "ctl", WITNESS_RANK_CTL, + malloc_mutex_rank_exclusive)) { + return true; + } + + ctl_initialized = false; + + return false; +} + +void +ctl_prefork(tsdn_t *tsdn) { + malloc_mutex_prefork(tsdn, &ctl_mtx); +} + +void +ctl_postfork_parent(tsdn_t *tsdn) { + malloc_mutex_postfork_parent(tsdn, &ctl_mtx); +} + +void +ctl_postfork_child(tsdn_t *tsdn) { + malloc_mutex_postfork_child(tsdn, &ctl_mtx); +} + +/******************************************************************************/ +/* *_ctl() functions. */ + +#define READONLY() do { \ + if (newp != NULL || newlen != 0) { \ + ret = EPERM; \ + goto label_return; \ + } \ +} while (0) + +#define WRITEONLY() do { \ + if (oldp != NULL || oldlenp != NULL) { \ + ret = EPERM; \ + goto label_return; \ + } \ +} while (0) + +#define READ_XOR_WRITE() do { \ + if ((oldp != NULL && oldlenp != NULL) && (newp != NULL || \ + newlen != 0)) { \ + ret = EPERM; \ + goto label_return; \ + } \ +} while (0) + +#define READ(v, t) do { \ + if (oldp != NULL && oldlenp != NULL) { \ + if (*oldlenp != sizeof(t)) { \ + size_t copylen = (sizeof(t) <= *oldlenp) \ + ? sizeof(t) : *oldlenp; \ + memcpy(oldp, (void *)&(v), copylen); \ + ret = EINVAL; \ + goto label_return; \ + } \ + *(t *)oldp = (v); \ + } \ +} while (0) + +#define WRITE(v, t) do { \ + if (newp != NULL) { \ + if (newlen != sizeof(t)) { \ + ret = EINVAL; \ + goto label_return; \ + } \ + (v) = *(t *)newp; \ + } \ +} while (0) + +#define MIB_UNSIGNED(v, i) do { \ + if (mib[i] > UINT_MAX) { \ + ret = EFAULT; \ + goto label_return; \ + } \ + v = (unsigned)mib[i]; \ +} while (0) + +/* + * There's a lot of code duplication in the following macros due to limitations + * in how nested cpp macros are expanded. + */ +#define CTL_RO_CLGEN(c, l, n, v, t) \ +static int \ +n##_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp, \ + size_t *oldlenp, void *newp, size_t newlen) { \ + int ret; \ + t oldval; \ + \ + if (!(c)) { \ + return ENOENT; \ + } \ + if (l) { \ + malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx); \ + } \ + READONLY(); \ + oldval = (v); \ + READ(oldval, t); \ + \ + ret = 0; \ +label_return: \ + if (l) { \ + malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx); \ + } \ + return ret; \ +} + +#define CTL_RO_CGEN(c, n, v, t) \ +static int \ +n##_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, \ + void *oldp, size_t *oldlenp, void *newp, size_t newlen) { \ + int ret; \ + t oldval; \ + \ + if (!(c)) { \ + return ENOENT; \ + } \ + malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx); \ + READONLY(); \ + oldval = (v); \ + READ(oldval, t); \ + \ + ret = 0; \ +label_return: \ + malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx); \ + return ret; \ +} + +#define CTL_RO_GEN(n, v, t) \ +static int \ +n##_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp, \ + size_t *oldlenp, void *newp, size_t newlen) { \ + int ret; \ + t oldval; \ + \ + malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx); \ + READONLY(); \ + oldval = (v); \ + READ(oldval, t); \ + \ + ret = 0; \ +label_return: \ + malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx); \ + return ret; \ +} + +/* + * ctl_mtx is not acquired, under the assumption that no pertinent data will + * mutate during the call. + */ +#define CTL_RO_NL_CGEN(c, n, v, t) \ +static int \ +n##_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, \ + void *oldp, size_t *oldlenp, void *newp, size_t newlen) { \ + int ret; \ + t oldval; \ + \ + if (!(c)) { \ + return ENOENT; \ + } \ + READONLY(); \ + oldval = (v); \ + READ(oldval, t); \ + \ + ret = 0; \ +label_return: \ + return ret; \ +} + +#define CTL_RO_NL_GEN(n, v, t) \ +static int \ +n##_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, \ + void *oldp, size_t *oldlenp, void *newp, size_t newlen) { \ + int ret; \ + t oldval; \ + \ + READONLY(); \ + oldval = (v); \ + READ(oldval, t); \ + \ + ret = 0; \ +label_return: \ + return ret; \ +} + +#define CTL_TSD_RO_NL_CGEN(c, n, m, t) \ +static int \ +n##_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp, \ + size_t *oldlenp, void *newp, size_t newlen) { \ + int ret; \ + t oldval; \ + \ + if (!(c)) { \ + return ENOENT; \ + } \ + READONLY(); \ + oldval = (m(tsd)); \ + READ(oldval, t); \ + \ + ret = 0; \ +label_return: \ + return ret; \ +} + +#define CTL_RO_CONFIG_GEN(n, t) \ +static int \ +n##_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, \ + void *oldp, size_t *oldlenp, void *newp, size_t newlen) { \ + int ret; \ + t oldval; \ + \ + READONLY(); \ + oldval = n; \ + READ(oldval, t); \ + \ + ret = 0; \ +label_return: \ + return ret; \ +} + +/******************************************************************************/ + +CTL_RO_NL_GEN(version, JEMALLOC_VERSION, const char *) + +static int +epoch_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, + void *oldp, size_t *oldlenp, void *newp, size_t newlen) { + int ret; + UNUSED uint64_t newval; + + malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx); + WRITE(newval, uint64_t); + if (newp != NULL) { + ctl_refresh(tsd_tsdn(tsd)); + } + READ(ctl_arenas->epoch, uint64_t); + + ret = 0; +label_return: + malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx); + return ret; +} + +static int +background_thread_ctl(tsd_t *tsd, const size_t *mib, + size_t miblen, void *oldp, size_t *oldlenp, + void *newp, size_t newlen) { + int ret; + bool oldval; + + if (!have_background_thread) { + return ENOENT; + } + background_thread_ctl_init(tsd_tsdn(tsd)); + + malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx); + malloc_mutex_lock(tsd_tsdn(tsd), &background_thread_lock); + if (newp == NULL) { + oldval = background_thread_enabled(); + READ(oldval, bool); + } else { + if (newlen != sizeof(bool)) { + ret = EINVAL; + goto label_return; + } + oldval = background_thread_enabled(); + READ(oldval, bool); + + bool newval = *(bool *)newp; + if (newval == oldval) { + ret = 0; + goto label_return; + } + + background_thread_enabled_set(tsd_tsdn(tsd), newval); + if (newval) { + if (background_threads_enable(tsd)) { + ret = EFAULT; + goto label_return; + } + } else { + if (background_threads_disable(tsd)) { + ret = EFAULT; + goto label_return; + } + } + } + ret = 0; +label_return: + malloc_mutex_unlock(tsd_tsdn(tsd), &background_thread_lock); + malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx); + + return ret; +} + +static int +max_background_threads_ctl(tsd_t *tsd, const size_t *mib, + size_t miblen, void *oldp, size_t *oldlenp, void *newp, + size_t newlen) { + int ret; + size_t oldval; + + if (!have_background_thread) { + return ENOENT; + } + background_thread_ctl_init(tsd_tsdn(tsd)); + + malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx); + malloc_mutex_lock(tsd_tsdn(tsd), &background_thread_lock); + if (newp == NULL) { + oldval = max_background_threads; + READ(oldval, size_t); + } else { + if (newlen != sizeof(size_t)) { + ret = EINVAL; + goto label_return; + } + oldval = max_background_threads; + READ(oldval, size_t); + + size_t newval = *(size_t *)newp; + if (newval == oldval) { + ret = 0; + goto label_return; + } + if (newval > opt_max_background_threads) { + ret = EINVAL; + goto label_return; + } + + if (background_thread_enabled()) { + background_thread_enabled_set(tsd_tsdn(tsd), false); + if (background_threads_disable(tsd)) { + ret = EFAULT; + goto label_return; + } + max_background_threads = newval; + background_thread_enabled_set(tsd_tsdn(tsd), true); + if (background_threads_enable(tsd)) { + ret = EFAULT; + goto label_return; + } + } else { + max_background_threads = newval; + } + } + ret = 0; +label_return: + malloc_mutex_unlock(tsd_tsdn(tsd), &background_thread_lock); + malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx); + + return ret; +} + +/******************************************************************************/ + +CTL_RO_CONFIG_GEN(config_cache_oblivious, bool) +CTL_RO_CONFIG_GEN(config_debug, bool) +CTL_RO_CONFIG_GEN(config_fill, bool) +CTL_RO_CONFIG_GEN(config_lazy_lock, bool) +CTL_RO_CONFIG_GEN(config_malloc_conf, const char *) +CTL_RO_CONFIG_GEN(config_opt_safety_checks, bool) +CTL_RO_CONFIG_GEN(config_prof, bool) +CTL_RO_CONFIG_GEN(config_prof_libgcc, bool) +CTL_RO_CONFIG_GEN(config_prof_libunwind, bool) +CTL_RO_CONFIG_GEN(config_stats, bool) +CTL_RO_CONFIG_GEN(config_utrace, bool) +CTL_RO_CONFIG_GEN(config_xmalloc, bool) + +/******************************************************************************/ + +CTL_RO_NL_GEN(opt_abort, opt_abort, bool) +CTL_RO_NL_GEN(opt_abort_conf, opt_abort_conf, bool) +CTL_RO_NL_GEN(opt_confirm_conf, opt_confirm_conf, bool) +CTL_RO_NL_GEN(opt_metadata_thp, metadata_thp_mode_names[opt_metadata_thp], + const char *) +CTL_RO_NL_GEN(opt_retain, opt_retain, bool) +CTL_RO_NL_GEN(opt_dss, opt_dss, const char *) +CTL_RO_NL_GEN(opt_narenas, opt_narenas, unsigned) +CTL_RO_NL_GEN(opt_percpu_arena, percpu_arena_mode_names[opt_percpu_arena], + const char *) +CTL_RO_NL_GEN(opt_oversize_threshold, opt_oversize_threshold, size_t) +CTL_RO_NL_GEN(opt_background_thread, opt_background_thread, bool) +CTL_RO_NL_GEN(opt_max_background_threads, opt_max_background_threads, size_t) +CTL_RO_NL_GEN(opt_dirty_decay_ms, opt_dirty_decay_ms, ssize_t) +CTL_RO_NL_GEN(opt_muzzy_decay_ms, opt_muzzy_decay_ms, ssize_t) +CTL_RO_NL_GEN(opt_stats_print, opt_stats_print, bool) +CTL_RO_NL_GEN(opt_stats_print_opts, opt_stats_print_opts, const char *) +CTL_RO_NL_CGEN(config_fill, opt_junk, opt_junk, const char *) +CTL_RO_NL_CGEN(config_fill, opt_zero, opt_zero, bool) +CTL_RO_NL_CGEN(config_utrace, opt_utrace, opt_utrace, bool) +CTL_RO_NL_CGEN(config_xmalloc, opt_xmalloc, opt_xmalloc, bool) +CTL_RO_NL_GEN(opt_tcache, opt_tcache, bool) +CTL_RO_NL_GEN(opt_thp, thp_mode_names[opt_thp], const char *) +CTL_RO_NL_GEN(opt_lg_extent_max_active_fit, opt_lg_extent_max_active_fit, + size_t) +CTL_RO_NL_GEN(opt_lg_tcache_max, opt_lg_tcache_max, ssize_t) +CTL_RO_NL_CGEN(config_prof, opt_prof, opt_prof, bool) +CTL_RO_NL_CGEN(config_prof, opt_prof_prefix, opt_prof_prefix, const char *) +CTL_RO_NL_CGEN(config_prof, opt_prof_active, opt_prof_active, bool) +CTL_RO_NL_CGEN(config_prof, opt_prof_thread_active_init, + opt_prof_thread_active_init, bool) +CTL_RO_NL_CGEN(config_prof, opt_lg_prof_sample, opt_lg_prof_sample, size_t) +CTL_RO_NL_CGEN(config_prof, opt_prof_accum, opt_prof_accum, bool) +CTL_RO_NL_CGEN(config_prof, opt_lg_prof_interval, opt_lg_prof_interval, ssize_t) +CTL_RO_NL_CGEN(config_prof, opt_prof_gdump, opt_prof_gdump, bool) +CTL_RO_NL_CGEN(config_prof, opt_prof_final, opt_prof_final, bool) +CTL_RO_NL_CGEN(config_prof, opt_prof_leak, opt_prof_leak, bool) + +/******************************************************************************/ + +static int +thread_arena_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, + void *oldp, size_t *oldlenp, void *newp, size_t newlen) { + int ret; + arena_t *oldarena; + unsigned newind, oldind; + + oldarena = arena_choose(tsd, NULL); + if (oldarena == NULL) { + return EAGAIN; + } + newind = oldind = arena_ind_get(oldarena); + WRITE(newind, unsigned); + READ(oldind, unsigned); + + if (newind != oldind) { + arena_t *newarena; + + if (newind >= narenas_total_get()) { + /* New arena index is out of range. */ + ret = EFAULT; + goto label_return; + } + + if (have_percpu_arena && + PERCPU_ARENA_ENABLED(opt_percpu_arena)) { + if (newind < percpu_arena_ind_limit(opt_percpu_arena)) { + /* + * If perCPU arena is enabled, thread_arena + * control is not allowed for the auto arena + * range. + */ + ret = EPERM; + goto label_return; + } + } + + /* Initialize arena if necessary. */ + newarena = arena_get(tsd_tsdn(tsd), newind, true); + if (newarena == NULL) { + ret = EAGAIN; + goto label_return; + } + /* Set new arena/tcache associations. */ + arena_migrate(tsd, oldind, newind); + if (tcache_available(tsd)) { + tcache_arena_reassociate(tsd_tsdn(tsd), + tsd_tcachep_get(tsd), newarena); + } + } + + ret = 0; +label_return: + return ret; +} + +CTL_TSD_RO_NL_CGEN(config_stats, thread_allocated, tsd_thread_allocated_get, + uint64_t) +CTL_TSD_RO_NL_CGEN(config_stats, thread_allocatedp, tsd_thread_allocatedp_get, + uint64_t *) +CTL_TSD_RO_NL_CGEN(config_stats, thread_deallocated, tsd_thread_deallocated_get, + uint64_t) +CTL_TSD_RO_NL_CGEN(config_stats, thread_deallocatedp, + tsd_thread_deallocatedp_get, uint64_t *) + +static int +thread_tcache_enabled_ctl(tsd_t *tsd, const size_t *mib, + size_t miblen, void *oldp, size_t *oldlenp, void *newp, + size_t newlen) { + int ret; + bool oldval; + + oldval = tcache_enabled_get(tsd); + if (newp != NULL) { + if (newlen != sizeof(bool)) { + ret = EINVAL; + goto label_return; + } + tcache_enabled_set(tsd, *(bool *)newp); + } + READ(oldval, bool); + + ret = 0; +label_return: + return ret; +} + +static int +thread_tcache_flush_ctl(tsd_t *tsd, const size_t *mib, + size_t miblen, void *oldp, size_t *oldlenp, void *newp, + size_t newlen) { + int ret; + + if (!tcache_available(tsd)) { + ret = EFAULT; + goto label_return; + } + + READONLY(); + WRITEONLY(); + + tcache_flush(tsd); + + ret = 0; +label_return: + return ret; +} + +static int +thread_prof_name_ctl(tsd_t *tsd, const size_t *mib, + size_t miblen, void *oldp, size_t *oldlenp, void *newp, + size_t newlen) { + int ret; + + if (!config_prof) { + return ENOENT; + } + + READ_XOR_WRITE(); + + if (newp != NULL) { + if (newlen != sizeof(const char *)) { + ret = EINVAL; + goto label_return; + } + + if ((ret = prof_thread_name_set(tsd, *(const char **)newp)) != + 0) { + goto label_return; + } + } else { + const char *oldname = prof_thread_name_get(tsd); + READ(oldname, const char *); + } + + ret = 0; +label_return: + return ret; +} + +static int +thread_prof_active_ctl(tsd_t *tsd, const size_t *mib, + size_t miblen, void *oldp, size_t *oldlenp, void *newp, + size_t newlen) { + int ret; + bool oldval; + + if (!config_prof) { + return ENOENT; + } + + oldval = prof_thread_active_get(tsd); + if (newp != NULL) { + if (newlen != sizeof(bool)) { + ret = EINVAL; + goto label_return; + } + if (prof_thread_active_set(tsd, *(bool *)newp)) { + ret = EAGAIN; + goto label_return; + } + } + READ(oldval, bool); + + ret = 0; +label_return: + return ret; +} + +/******************************************************************************/ + +static int +tcache_create_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, + void *oldp, size_t *oldlenp, void *newp, size_t newlen) { + int ret; + unsigned tcache_ind; + + READONLY(); + if (tcaches_create(tsd, &tcache_ind)) { + ret = EFAULT; + goto label_return; + } + READ(tcache_ind, unsigned); + + ret = 0; +label_return: + return ret; +} + +static int +tcache_flush_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, + void *oldp, size_t *oldlenp, void *newp, size_t newlen) { + int ret; + unsigned tcache_ind; + + WRITEONLY(); + tcache_ind = UINT_MAX; + WRITE(tcache_ind, unsigned); + if (tcache_ind == UINT_MAX) { + ret = EFAULT; + goto label_return; + } + tcaches_flush(tsd, tcache_ind); + + ret = 0; +label_return: + return ret; +} + +static int +tcache_destroy_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, + void *oldp, size_t *oldlenp, void *newp, size_t newlen) { + int ret; + unsigned tcache_ind; + + WRITEONLY(); + tcache_ind = UINT_MAX; + WRITE(tcache_ind, unsigned); + if (tcache_ind == UINT_MAX) { + ret = EFAULT; + goto label_return; + } + tcaches_destroy(tsd, tcache_ind); + + ret = 0; +label_return: + return ret; +} + +/******************************************************************************/ + +static int +arena_i_initialized_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, + void *oldp, size_t *oldlenp, void *newp, size_t newlen) { + int ret; + tsdn_t *tsdn = tsd_tsdn(tsd); + unsigned arena_ind; + bool initialized; + + READONLY(); + MIB_UNSIGNED(arena_ind, 1); + + malloc_mutex_lock(tsdn, &ctl_mtx); + initialized = arenas_i(arena_ind)->initialized; + malloc_mutex_unlock(tsdn, &ctl_mtx); + + READ(initialized, bool); + + ret = 0; +label_return: + return ret; +} + +static void +arena_i_decay(tsdn_t *tsdn, unsigned arena_ind, bool all) { + malloc_mutex_lock(tsdn, &ctl_mtx); + { + unsigned narenas = ctl_arenas->narenas; + + /* + * Access via index narenas is deprecated, and scheduled for + * removal in 6.0.0. + */ + if (arena_ind == MALLCTL_ARENAS_ALL || arena_ind == narenas) { + unsigned i; + VARIABLE_ARRAY(arena_t *, tarenas, narenas); + + for (i = 0; i < narenas; i++) { + tarenas[i] = arena_get(tsdn, i, false); + } + + /* + * No further need to hold ctl_mtx, since narenas and + * tarenas contain everything needed below. + */ + malloc_mutex_unlock(tsdn, &ctl_mtx); + + for (i = 0; i < narenas; i++) { + if (tarenas[i] != NULL) { + arena_decay(tsdn, tarenas[i], false, + all); + } + } + } else { + arena_t *tarena; + + assert(arena_ind < narenas); + + tarena = arena_get(tsdn, arena_ind, false); + + /* No further need to hold ctl_mtx. */ + malloc_mutex_unlock(tsdn, &ctl_mtx); + + if (tarena != NULL) { + arena_decay(tsdn, tarena, false, all); + } + } + } +} + +static int +arena_i_decay_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp, + size_t *oldlenp, void *newp, size_t newlen) { + int ret; + unsigned arena_ind; + + READONLY(); + WRITEONLY(); + MIB_UNSIGNED(arena_ind, 1); + arena_i_decay(tsd_tsdn(tsd), arena_ind, false); + + ret = 0; +label_return: + return ret; +} + +static int +arena_i_purge_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp, + size_t *oldlenp, void *newp, size_t newlen) { + int ret; + unsigned arena_ind; + + READONLY(); + WRITEONLY(); + MIB_UNSIGNED(arena_ind, 1); + arena_i_decay(tsd_tsdn(tsd), arena_ind, true); + + ret = 0; +label_return: + return ret; +} + +static int +arena_i_reset_destroy_helper(tsd_t *tsd, const size_t *mib, size_t miblen, + void *oldp, size_t *oldlenp, void *newp, size_t newlen, unsigned *arena_ind, + arena_t **arena) { + int ret; + + READONLY(); + WRITEONLY(); + MIB_UNSIGNED(*arena_ind, 1); + + *arena = arena_get(tsd_tsdn(tsd), *arena_ind, false); + if (*arena == NULL || arena_is_auto(*arena)) { + ret = EFAULT; + goto label_return; + } + + ret = 0; +label_return: + return ret; +} + +static void +arena_reset_prepare_background_thread(tsd_t *tsd, unsigned arena_ind) { + /* Temporarily disable the background thread during arena reset. */ + if (have_background_thread) { + malloc_mutex_lock(tsd_tsdn(tsd), &background_thread_lock); + if (background_thread_enabled()) { + background_thread_info_t *info = + background_thread_info_get(arena_ind); + assert(info->state == background_thread_started); + malloc_mutex_lock(tsd_tsdn(tsd), &info->mtx); + info->state = background_thread_paused; + malloc_mutex_unlock(tsd_tsdn(tsd), &info->mtx); + } + } +} + +static void +arena_reset_finish_background_thread(tsd_t *tsd, unsigned arena_ind) { + if (have_background_thread) { + if (background_thread_enabled()) { + background_thread_info_t *info = + background_thread_info_get(arena_ind); + assert(info->state == background_thread_paused); + malloc_mutex_lock(tsd_tsdn(tsd), &info->mtx); + info->state = background_thread_started; + malloc_mutex_unlock(tsd_tsdn(tsd), &info->mtx); + } + malloc_mutex_unlock(tsd_tsdn(tsd), &background_thread_lock); + } +} + +static int +arena_i_reset_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp, + size_t *oldlenp, void *newp, size_t newlen) { + int ret; + unsigned arena_ind; + arena_t *arena; + + ret = arena_i_reset_destroy_helper(tsd, mib, miblen, oldp, oldlenp, + newp, newlen, &arena_ind, &arena); + if (ret != 0) { + return ret; + } + + arena_reset_prepare_background_thread(tsd, arena_ind); + arena_reset(tsd, arena); + arena_reset_finish_background_thread(tsd, arena_ind); + + return ret; +} + +static int +arena_i_destroy_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp, + size_t *oldlenp, void *newp, size_t newlen) { + int ret; + unsigned arena_ind; + arena_t *arena; + ctl_arena_t *ctl_darena, *ctl_arena; + + ret = arena_i_reset_destroy_helper(tsd, mib, miblen, oldp, oldlenp, + newp, newlen, &arena_ind, &arena); + if (ret != 0) { + goto label_return; + } + + if (arena_nthreads_get(arena, false) != 0 || arena_nthreads_get(arena, + true) != 0) { + ret = EFAULT; + goto label_return; + } + + arena_reset_prepare_background_thread(tsd, arena_ind); + /* Merge stats after resetting and purging arena. */ + arena_reset(tsd, arena); + arena_decay(tsd_tsdn(tsd), arena, false, true); + ctl_darena = arenas_i(MALLCTL_ARENAS_DESTROYED); + ctl_darena->initialized = true; + ctl_arena_refresh(tsd_tsdn(tsd), arena, ctl_darena, arena_ind, true); + /* Destroy arena. */ + arena_destroy(tsd, arena); + ctl_arena = arenas_i(arena_ind); + ctl_arena->initialized = false; + /* Record arena index for later recycling via arenas.create. */ + ql_elm_new(ctl_arena, destroyed_link); + ql_tail_insert(&ctl_arenas->destroyed, ctl_arena, destroyed_link); + arena_reset_finish_background_thread(tsd, arena_ind); + + assert(ret == 0); +label_return: + return ret; +} + +static int +arena_i_dss_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp, + size_t *oldlenp, void *newp, size_t newlen) { + int ret; + const char *dss = NULL; + unsigned arena_ind; + dss_prec_t dss_prec_old = dss_prec_limit; + dss_prec_t dss_prec = dss_prec_limit; + + malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx); + WRITE(dss, const char *); + MIB_UNSIGNED(arena_ind, 1); + if (dss != NULL) { + int i; + bool match = false; + + for (i = 0; i < dss_prec_limit; i++) { + if (strcmp(dss_prec_names[i], dss) == 0) { + dss_prec = i; + match = true; + break; + } + } + + if (!match) { + ret = EINVAL; + goto label_return; + } + } + + /* + * Access via index narenas is deprecated, and scheduled for removal in + * 6.0.0. + */ + if (arena_ind == MALLCTL_ARENAS_ALL || arena_ind == + ctl_arenas->narenas) { + if (dss_prec != dss_prec_limit && + extent_dss_prec_set(dss_prec)) { + ret = EFAULT; + goto label_return; + } + dss_prec_old = extent_dss_prec_get(); + } else { + arena_t *arena = arena_get(tsd_tsdn(tsd), arena_ind, false); + if (arena == NULL || (dss_prec != dss_prec_limit && + arena_dss_prec_set(arena, dss_prec))) { + ret = EFAULT; + goto label_return; + } + dss_prec_old = arena_dss_prec_get(arena); + } + + dss = dss_prec_names[dss_prec_old]; + READ(dss, const char *); + + ret = 0; +label_return: + malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx); + return ret; +} + +static int +arena_i_decay_ms_ctl_impl(tsd_t *tsd, const size_t *mib, size_t miblen, + void *oldp, size_t *oldlenp, void *newp, size_t newlen, bool dirty) { + int ret; + unsigned arena_ind; + arena_t *arena; + + MIB_UNSIGNED(arena_ind, 1); + arena = arena_get(tsd_tsdn(tsd), arena_ind, false); + if (arena == NULL) { + ret = EFAULT; + goto label_return; + } + + if (oldp != NULL && oldlenp != NULL) { + size_t oldval = dirty ? arena_dirty_decay_ms_get(arena) : + arena_muzzy_decay_ms_get(arena); + READ(oldval, ssize_t); + } + if (newp != NULL) { + if (newlen != sizeof(ssize_t)) { + ret = EINVAL; + goto label_return; + } + if (arena_is_huge(arena_ind) && *(ssize_t *)newp > 0) { + /* + * By default the huge arena purges eagerly. If it is + * set to non-zero decay time afterwards, background + * thread might be needed. + */ + if (background_thread_create(tsd, arena_ind)) { + ret = EFAULT; + goto label_return; + } + } + if (dirty ? arena_dirty_decay_ms_set(tsd_tsdn(tsd), arena, + *(ssize_t *)newp) : arena_muzzy_decay_ms_set(tsd_tsdn(tsd), + arena, *(ssize_t *)newp)) { + ret = EFAULT; + goto label_return; + } + } + + ret = 0; +label_return: + return ret; +} + +static int +arena_i_dirty_decay_ms_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, + void *oldp, size_t *oldlenp, void *newp, size_t newlen) { + return arena_i_decay_ms_ctl_impl(tsd, mib, miblen, oldp, oldlenp, newp, + newlen, true); +} + +static int +arena_i_muzzy_decay_ms_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, + void *oldp, size_t *oldlenp, void *newp, size_t newlen) { + return arena_i_decay_ms_ctl_impl(tsd, mib, miblen, oldp, oldlenp, newp, + newlen, false); +} + +static int +arena_i_extent_hooks_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, + void *oldp, size_t *oldlenp, void *newp, size_t newlen) { + int ret; + unsigned arena_ind; + arena_t *arena; + + malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx); + MIB_UNSIGNED(arena_ind, 1); + if (arena_ind < narenas_total_get()) { + extent_hooks_t *old_extent_hooks; + arena = arena_get(tsd_tsdn(tsd), arena_ind, false); + if (arena == NULL) { + if (arena_ind >= narenas_auto) { + ret = EFAULT; + goto label_return; + } + old_extent_hooks = + (extent_hooks_t *)&extent_hooks_default; + READ(old_extent_hooks, extent_hooks_t *); + if (newp != NULL) { + /* Initialize a new arena as a side effect. */ + extent_hooks_t *new_extent_hooks + JEMALLOC_CC_SILENCE_INIT(NULL); + WRITE(new_extent_hooks, extent_hooks_t *); + arena = arena_init(tsd_tsdn(tsd), arena_ind, + new_extent_hooks); + if (arena == NULL) { + ret = EFAULT; + goto label_return; + } + } + } else { + if (newp != NULL) { + extent_hooks_t *new_extent_hooks + JEMALLOC_CC_SILENCE_INIT(NULL); + WRITE(new_extent_hooks, extent_hooks_t *); + old_extent_hooks = extent_hooks_set(tsd, arena, + new_extent_hooks); + READ(old_extent_hooks, extent_hooks_t *); + } else { + old_extent_hooks = extent_hooks_get(arena); + READ(old_extent_hooks, extent_hooks_t *); + } + } + } else { + ret = EFAULT; + goto label_return; + } + ret = 0; +label_return: + malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx); + return ret; +} + +static int +arena_i_retain_grow_limit_ctl(tsd_t *tsd, const size_t *mib, + size_t miblen, void *oldp, size_t *oldlenp, void *newp, + size_t newlen) { + int ret; + unsigned arena_ind; + arena_t *arena; + + if (!opt_retain) { + /* Only relevant when retain is enabled. */ + return ENOENT; + } + + malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx); + MIB_UNSIGNED(arena_ind, 1); + if (arena_ind < narenas_total_get() && (arena = + arena_get(tsd_tsdn(tsd), arena_ind, false)) != NULL) { + size_t old_limit, new_limit; + if (newp != NULL) { + WRITE(new_limit, size_t); + } + bool err = arena_retain_grow_limit_get_set(tsd, arena, + &old_limit, newp != NULL ? &new_limit : NULL); + if (!err) { + READ(old_limit, size_t); + ret = 0; + } else { + ret = EFAULT; + } + } else { + ret = EFAULT; + } +label_return: + malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx); + return ret; +} + +static const ctl_named_node_t * +arena_i_index(tsdn_t *tsdn, const size_t *mib, size_t miblen, + size_t i) { + const ctl_named_node_t *ret; + + malloc_mutex_lock(tsdn, &ctl_mtx); + switch (i) { + case MALLCTL_ARENAS_ALL: + case MALLCTL_ARENAS_DESTROYED: + break; + default: + if (i > ctl_arenas->narenas) { + ret = NULL; + goto label_return; + } + break; + } + + ret = super_arena_i_node; +label_return: + malloc_mutex_unlock(tsdn, &ctl_mtx); + return ret; +} + +/******************************************************************************/ + +static int +arenas_narenas_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, + void *oldp, size_t *oldlenp, void *newp, size_t newlen) { + int ret; + unsigned narenas; + + malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx); + READONLY(); + if (*oldlenp != sizeof(unsigned)) { + ret = EINVAL; + goto label_return; + } + narenas = ctl_arenas->narenas; + READ(narenas, unsigned); + + ret = 0; +label_return: + malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx); + return ret; +} + +static int +arenas_decay_ms_ctl_impl(tsd_t *tsd, const size_t *mib, + size_t miblen, void *oldp, size_t *oldlenp, void *newp, + size_t newlen, bool dirty) { + int ret; + + if (oldp != NULL && oldlenp != NULL) { + size_t oldval = (dirty ? arena_dirty_decay_ms_default_get() : + arena_muzzy_decay_ms_default_get()); + READ(oldval, ssize_t); + } + if (newp != NULL) { + if (newlen != sizeof(ssize_t)) { + ret = EINVAL; + goto label_return; + } + if (dirty ? arena_dirty_decay_ms_default_set(*(ssize_t *)newp) + : arena_muzzy_decay_ms_default_set(*(ssize_t *)newp)) { + ret = EFAULT; + goto label_return; + } + } + + ret = 0; +label_return: + return ret; +} + +static int +arenas_dirty_decay_ms_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, + void *oldp, size_t *oldlenp, void *newp, size_t newlen) { + return arenas_decay_ms_ctl_impl(tsd, mib, miblen, oldp, oldlenp, newp, + newlen, true); +} + +static int +arenas_muzzy_decay_ms_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, + void *oldp, size_t *oldlenp, void *newp, size_t newlen) { + return arenas_decay_ms_ctl_impl(tsd, mib, miblen, oldp, oldlenp, newp, + newlen, false); +} + +CTL_RO_NL_GEN(arenas_quantum, QUANTUM, size_t) +CTL_RO_NL_GEN(arenas_page, PAGE, size_t) +CTL_RO_NL_GEN(arenas_tcache_max, tcache_maxclass, size_t) +CTL_RO_NL_GEN(arenas_nbins, SC_NBINS, unsigned) +CTL_RO_NL_GEN(arenas_nhbins, nhbins, unsigned) +CTL_RO_NL_GEN(arenas_bin_i_size, bin_infos[mib[2]].reg_size, size_t) +CTL_RO_NL_GEN(arenas_bin_i_nregs, bin_infos[mib[2]].nregs, uint32_t) +CTL_RO_NL_GEN(arenas_bin_i_slab_size, bin_infos[mib[2]].slab_size, size_t) +CTL_RO_NL_GEN(arenas_bin_i_nshards, bin_infos[mib[2]].n_shards, uint32_t) +static const ctl_named_node_t * +arenas_bin_i_index(tsdn_t *tsdn, const size_t *mib, + size_t miblen, size_t i) { + if (i > SC_NBINS) { + return NULL; + } + return super_arenas_bin_i_node; +} + +CTL_RO_NL_GEN(arenas_nlextents, SC_NSIZES - SC_NBINS, unsigned) +CTL_RO_NL_GEN(arenas_lextent_i_size, sz_index2size(SC_NBINS+(szind_t)mib[2]), + size_t) +static const ctl_named_node_t * +arenas_lextent_i_index(tsdn_t *tsdn, const size_t *mib, + size_t miblen, size_t i) { + if (i > SC_NSIZES - SC_NBINS) { + return NULL; + } + return super_arenas_lextent_i_node; +} + +static int +arenas_create_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, + void *oldp, size_t *oldlenp, void *newp, size_t newlen) { + int ret; + extent_hooks_t *extent_hooks; + unsigned arena_ind; + + malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx); + + extent_hooks = (extent_hooks_t *)&extent_hooks_default; + WRITE(extent_hooks, extent_hooks_t *); + if ((arena_ind = ctl_arena_init(tsd, extent_hooks)) == UINT_MAX) { + ret = EAGAIN; + goto label_return; + } + READ(arena_ind, unsigned); + + ret = 0; +label_return: + malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx); + return ret; +} + +static int +arenas_lookup_ctl(tsd_t *tsd, const size_t *mib, + size_t miblen, void *oldp, size_t *oldlenp, void *newp, + size_t newlen) { + int ret; + unsigned arena_ind; + void *ptr; + extent_t *extent; + arena_t *arena; + + ptr = NULL; + ret = EINVAL; + malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx); + WRITE(ptr, void *); + extent = iealloc(tsd_tsdn(tsd), ptr); + if (extent == NULL) + goto label_return; + + arena = extent_arena_get(extent); + if (arena == NULL) + goto label_return; + + arena_ind = arena_ind_get(arena); + READ(arena_ind, unsigned); + + ret = 0; +label_return: + malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx); + return ret; +} + +/******************************************************************************/ + +static int +prof_thread_active_init_ctl(tsd_t *tsd, const size_t *mib, + size_t miblen, void *oldp, size_t *oldlenp, void *newp, + size_t newlen) { + int ret; + bool oldval; + + if (!config_prof) { + return ENOENT; + } + + if (newp != NULL) { + if (newlen != sizeof(bool)) { + ret = EINVAL; + goto label_return; + } + oldval = prof_thread_active_init_set(tsd_tsdn(tsd), + *(bool *)newp); + } else { + oldval = prof_thread_active_init_get(tsd_tsdn(tsd)); + } + READ(oldval, bool); + + ret = 0; +label_return: + return ret; +} + +static int +prof_active_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, + void *oldp, size_t *oldlenp, void *newp, size_t newlen) { + int ret; + bool oldval; + + if (!config_prof) { + return ENOENT; + } + + if (newp != NULL) { + if (newlen != sizeof(bool)) { + ret = EINVAL; + goto label_return; + } + oldval = prof_active_set(tsd_tsdn(tsd), *(bool *)newp); + } else { + oldval = prof_active_get(tsd_tsdn(tsd)); + } + READ(oldval, bool); + + ret = 0; +label_return: + return ret; +} + +static int +prof_dump_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, + void *oldp, size_t *oldlenp, void *newp, size_t newlen) { + int ret; + const char *filename = NULL; + + if (!config_prof) { + return ENOENT; + } + + WRITEONLY(); + WRITE(filename, const char *); + + if (prof_mdump(tsd, filename)) { + ret = EFAULT; + goto label_return; + } + + ret = 0; +label_return: + return ret; +} + +static int +prof_gdump_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, + void *oldp, size_t *oldlenp, void *newp, size_t newlen) { + int ret; + bool oldval; + + if (!config_prof) { + return ENOENT; + } + + if (newp != NULL) { + if (newlen != sizeof(bool)) { + ret = EINVAL; + goto label_return; + } + oldval = prof_gdump_set(tsd_tsdn(tsd), *(bool *)newp); + } else { + oldval = prof_gdump_get(tsd_tsdn(tsd)); + } + READ(oldval, bool); + + ret = 0; +label_return: + return ret; +} + +static int +prof_reset_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, + void *oldp, size_t *oldlenp, void *newp, size_t newlen) { + int ret; + size_t lg_sample = lg_prof_sample; + + if (!config_prof) { + return ENOENT; + } + + WRITEONLY(); + WRITE(lg_sample, size_t); + if (lg_sample >= (sizeof(uint64_t) << 3)) { + lg_sample = (sizeof(uint64_t) << 3) - 1; + } + + prof_reset(tsd, lg_sample); + + ret = 0; +label_return: + return ret; +} + +CTL_RO_NL_CGEN(config_prof, prof_interval, prof_interval, uint64_t) +CTL_RO_NL_CGEN(config_prof, lg_prof_sample, lg_prof_sample, size_t) + +static int +prof_log_start_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp, + size_t *oldlenp, void *newp, size_t newlen) { + int ret; + + const char *filename = NULL; + + if (!config_prof) { + return ENOENT; + } + + WRITEONLY(); + WRITE(filename, const char *); + + if (prof_log_start(tsd_tsdn(tsd), filename)) { + ret = EFAULT; + goto label_return; + } + + ret = 0; +label_return: + return ret; +} + +static int +prof_log_stop_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp, + size_t *oldlenp, void *newp, size_t newlen) { + if (!config_prof) { + return ENOENT; + } + + if (prof_log_stop(tsd_tsdn(tsd))) { + return EFAULT; + } + + return 0; +} + +/******************************************************************************/ + +CTL_RO_CGEN(config_stats, stats_allocated, ctl_stats->allocated, size_t) +CTL_RO_CGEN(config_stats, stats_active, ctl_stats->active, size_t) +CTL_RO_CGEN(config_stats, stats_metadata, ctl_stats->metadata, size_t) +CTL_RO_CGEN(config_stats, stats_metadata_thp, ctl_stats->metadata_thp, size_t) +CTL_RO_CGEN(config_stats, stats_resident, ctl_stats->resident, size_t) +CTL_RO_CGEN(config_stats, stats_mapped, ctl_stats->mapped, size_t) +CTL_RO_CGEN(config_stats, stats_retained, ctl_stats->retained, size_t) + +CTL_RO_CGEN(config_stats, stats_background_thread_num_threads, + ctl_stats->background_thread.num_threads, size_t) +CTL_RO_CGEN(config_stats, stats_background_thread_num_runs, + ctl_stats->background_thread.num_runs, uint64_t) +CTL_RO_CGEN(config_stats, stats_background_thread_run_interval, + nstime_ns(&ctl_stats->background_thread.run_interval), uint64_t) + +CTL_RO_GEN(stats_arenas_i_dss, arenas_i(mib[2])->dss, const char *) +CTL_RO_GEN(stats_arenas_i_dirty_decay_ms, arenas_i(mib[2])->dirty_decay_ms, + ssize_t) +CTL_RO_GEN(stats_arenas_i_muzzy_decay_ms, arenas_i(mib[2])->muzzy_decay_ms, + ssize_t) +CTL_RO_GEN(stats_arenas_i_nthreads, arenas_i(mib[2])->nthreads, unsigned) +CTL_RO_GEN(stats_arenas_i_uptime, + nstime_ns(&arenas_i(mib[2])->astats->astats.uptime), uint64_t) +CTL_RO_GEN(stats_arenas_i_pactive, arenas_i(mib[2])->pactive, size_t) +CTL_RO_GEN(stats_arenas_i_pdirty, arenas_i(mib[2])->pdirty, size_t) +CTL_RO_GEN(stats_arenas_i_pmuzzy, arenas_i(mib[2])->pmuzzy, size_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_mapped, + atomic_load_zu(&arenas_i(mib[2])->astats->astats.mapped, ATOMIC_RELAXED), + size_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_retained, + atomic_load_zu(&arenas_i(mib[2])->astats->astats.retained, ATOMIC_RELAXED), + size_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_extent_avail, + atomic_load_zu(&arenas_i(mib[2])->astats->astats.extent_avail, + ATOMIC_RELAXED), + size_t) + +CTL_RO_CGEN(config_stats, stats_arenas_i_dirty_npurge, + ctl_arena_stats_read_u64( + &arenas_i(mib[2])->astats->astats.decay_dirty.npurge), uint64_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_dirty_nmadvise, + ctl_arena_stats_read_u64( + &arenas_i(mib[2])->astats->astats.decay_dirty.nmadvise), uint64_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_dirty_purged, + ctl_arena_stats_read_u64( + &arenas_i(mib[2])->astats->astats.decay_dirty.purged), uint64_t) + +CTL_RO_CGEN(config_stats, stats_arenas_i_muzzy_npurge, + ctl_arena_stats_read_u64( + &arenas_i(mib[2])->astats->astats.decay_muzzy.npurge), uint64_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_muzzy_nmadvise, + ctl_arena_stats_read_u64( + &arenas_i(mib[2])->astats->astats.decay_muzzy.nmadvise), uint64_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_muzzy_purged, + ctl_arena_stats_read_u64( + &arenas_i(mib[2])->astats->astats.decay_muzzy.purged), uint64_t) + +CTL_RO_CGEN(config_stats, stats_arenas_i_base, + atomic_load_zu(&arenas_i(mib[2])->astats->astats.base, ATOMIC_RELAXED), + size_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_internal, + atomic_load_zu(&arenas_i(mib[2])->astats->astats.internal, ATOMIC_RELAXED), + size_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_metadata_thp, + atomic_load_zu(&arenas_i(mib[2])->astats->astats.metadata_thp, + ATOMIC_RELAXED), size_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_tcache_bytes, + atomic_load_zu(&arenas_i(mib[2])->astats->astats.tcache_bytes, + ATOMIC_RELAXED), size_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_resident, + atomic_load_zu(&arenas_i(mib[2])->astats->astats.resident, ATOMIC_RELAXED), + size_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_abandoned_vm, + atomic_load_zu(&arenas_i(mib[2])->astats->astats.abandoned_vm, + ATOMIC_RELAXED), size_t) + +CTL_RO_CGEN(config_stats, stats_arenas_i_small_allocated, + arenas_i(mib[2])->astats->allocated_small, size_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_small_nmalloc, + arenas_i(mib[2])->astats->nmalloc_small, uint64_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_small_ndalloc, + arenas_i(mib[2])->astats->ndalloc_small, uint64_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_small_nrequests, + arenas_i(mib[2])->astats->nrequests_small, uint64_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_small_nfills, + arenas_i(mib[2])->astats->nfills_small, uint64_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_small_nflushes, + arenas_i(mib[2])->astats->nflushes_small, uint64_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_large_allocated, + atomic_load_zu(&arenas_i(mib[2])->astats->astats.allocated_large, + ATOMIC_RELAXED), size_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_large_nmalloc, + ctl_arena_stats_read_u64( + &arenas_i(mib[2])->astats->astats.nmalloc_large), uint64_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_large_ndalloc, + ctl_arena_stats_read_u64( + &arenas_i(mib[2])->astats->astats.ndalloc_large), uint64_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_large_nrequests, + ctl_arena_stats_read_u64( + &arenas_i(mib[2])->astats->astats.nrequests_large), uint64_t) +/* + * Note: "nmalloc_large" here instead of "nfills" in the read. This is + * intentional (large has no batch fill). + */ +CTL_RO_CGEN(config_stats, stats_arenas_i_large_nfills, + ctl_arena_stats_read_u64( + &arenas_i(mib[2])->astats->astats.nmalloc_large), uint64_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_large_nflushes, + ctl_arena_stats_read_u64( + &arenas_i(mib[2])->astats->astats.nflushes_large), uint64_t) + +/* Lock profiling related APIs below. */ +#define RO_MUTEX_CTL_GEN(n, l) \ +CTL_RO_CGEN(config_stats, stats_##n##_num_ops, \ + l.n_lock_ops, uint64_t) \ +CTL_RO_CGEN(config_stats, stats_##n##_num_wait, \ + l.n_wait_times, uint64_t) \ +CTL_RO_CGEN(config_stats, stats_##n##_num_spin_acq, \ + l.n_spin_acquired, uint64_t) \ +CTL_RO_CGEN(config_stats, stats_##n##_num_owner_switch, \ + l.n_owner_switches, uint64_t) \ +CTL_RO_CGEN(config_stats, stats_##n##_total_wait_time, \ + nstime_ns(&l.tot_wait_time), uint64_t) \ +CTL_RO_CGEN(config_stats, stats_##n##_max_wait_time, \ + nstime_ns(&l.max_wait_time), uint64_t) \ +CTL_RO_CGEN(config_stats, stats_##n##_max_num_thds, \ + l.max_n_thds, uint32_t) + +/* Global mutexes. */ +#define OP(mtx) \ + RO_MUTEX_CTL_GEN(mutexes_##mtx, \ + ctl_stats->mutex_prof_data[global_prof_mutex_##mtx]) +MUTEX_PROF_GLOBAL_MUTEXES +#undef OP + +/* Per arena mutexes */ +#define OP(mtx) RO_MUTEX_CTL_GEN(arenas_i_mutexes_##mtx, \ + arenas_i(mib[2])->astats->astats.mutex_prof_data[arena_prof_mutex_##mtx]) +MUTEX_PROF_ARENA_MUTEXES +#undef OP + +/* tcache bin mutex */ +RO_MUTEX_CTL_GEN(arenas_i_bins_j_mutex, + arenas_i(mib[2])->astats->bstats[mib[4]].mutex_data) +#undef RO_MUTEX_CTL_GEN + +/* Resets all mutex stats, including global, arena and bin mutexes. */ +static int +stats_mutexes_reset_ctl(tsd_t *tsd, const size_t *mib, + size_t miblen, void *oldp, size_t *oldlenp, + void *newp, size_t newlen) { + if (!config_stats) { + return ENOENT; + } + + tsdn_t *tsdn = tsd_tsdn(tsd); + +#define MUTEX_PROF_RESET(mtx) \ + malloc_mutex_lock(tsdn, &mtx); \ + malloc_mutex_prof_data_reset(tsdn, &mtx); \ + malloc_mutex_unlock(tsdn, &mtx); + + /* Global mutexes: ctl and prof. */ + MUTEX_PROF_RESET(ctl_mtx); + if (have_background_thread) { + MUTEX_PROF_RESET(background_thread_lock); + } + if (config_prof && opt_prof) { + MUTEX_PROF_RESET(bt2gctx_mtx); + } + + + /* Per arena mutexes. */ + unsigned n = narenas_total_get(); + + for (unsigned i = 0; i < n; i++) { + arena_t *arena = arena_get(tsdn, i, false); + if (!arena) { + continue; + } + MUTEX_PROF_RESET(arena->large_mtx); + MUTEX_PROF_RESET(arena->extent_avail_mtx); + MUTEX_PROF_RESET(arena->extents_dirty.mtx); + MUTEX_PROF_RESET(arena->extents_muzzy.mtx); + MUTEX_PROF_RESET(arena->extents_retained.mtx); + MUTEX_PROF_RESET(arena->decay_dirty.mtx); + MUTEX_PROF_RESET(arena->decay_muzzy.mtx); + MUTEX_PROF_RESET(arena->tcache_ql_mtx); + MUTEX_PROF_RESET(arena->base->mtx); + + for (szind_t i = 0; i < SC_NBINS; i++) { + for (unsigned j = 0; j < bin_infos[i].n_shards; j++) { + bin_t *bin = &arena->bins[i].bin_shards[j]; + MUTEX_PROF_RESET(bin->lock); + } + } + } +#undef MUTEX_PROF_RESET + return 0; +} + +CTL_RO_CGEN(config_stats, stats_arenas_i_bins_j_nmalloc, + arenas_i(mib[2])->astats->bstats[mib[4]].nmalloc, uint64_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_bins_j_ndalloc, + arenas_i(mib[2])->astats->bstats[mib[4]].ndalloc, uint64_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_bins_j_nrequests, + arenas_i(mib[2])->astats->bstats[mib[4]].nrequests, uint64_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_bins_j_curregs, + arenas_i(mib[2])->astats->bstats[mib[4]].curregs, size_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_bins_j_nfills, + arenas_i(mib[2])->astats->bstats[mib[4]].nfills, uint64_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_bins_j_nflushes, + arenas_i(mib[2])->astats->bstats[mib[4]].nflushes, uint64_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_bins_j_nslabs, + arenas_i(mib[2])->astats->bstats[mib[4]].nslabs, uint64_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_bins_j_nreslabs, + arenas_i(mib[2])->astats->bstats[mib[4]].reslabs, uint64_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_bins_j_curslabs, + arenas_i(mib[2])->astats->bstats[mib[4]].curslabs, size_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_bins_j_nonfull_slabs, + arenas_i(mib[2])->astats->bstats[mib[4]].nonfull_slabs, size_t) + +static const ctl_named_node_t * +stats_arenas_i_bins_j_index(tsdn_t *tsdn, const size_t *mib, + size_t miblen, size_t j) { + if (j > SC_NBINS) { + return NULL; + } + return super_stats_arenas_i_bins_j_node; +} + +CTL_RO_CGEN(config_stats, stats_arenas_i_lextents_j_nmalloc, + ctl_arena_stats_read_u64( + &arenas_i(mib[2])->astats->lstats[mib[4]].nmalloc), uint64_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_lextents_j_ndalloc, + ctl_arena_stats_read_u64( + &arenas_i(mib[2])->astats->lstats[mib[4]].ndalloc), uint64_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_lextents_j_nrequests, + ctl_arena_stats_read_u64( + &arenas_i(mib[2])->astats->lstats[mib[4]].nrequests), uint64_t) +CTL_RO_CGEN(config_stats, stats_arenas_i_lextents_j_curlextents, + arenas_i(mib[2])->astats->lstats[mib[4]].curlextents, size_t) + +static const ctl_named_node_t * +stats_arenas_i_lextents_j_index(tsdn_t *tsdn, const size_t *mib, + size_t miblen, size_t j) { + if (j > SC_NSIZES - SC_NBINS) { + return NULL; + } + return super_stats_arenas_i_lextents_j_node; +} + +CTL_RO_CGEN(config_stats, stats_arenas_i_extents_j_ndirty, + atomic_load_zu( + &arenas_i(mib[2])->astats->estats[mib[4]].ndirty, + ATOMIC_RELAXED), size_t); +CTL_RO_CGEN(config_stats, stats_arenas_i_extents_j_nmuzzy, + atomic_load_zu( + &arenas_i(mib[2])->astats->estats[mib[4]].nmuzzy, + ATOMIC_RELAXED), size_t); +CTL_RO_CGEN(config_stats, stats_arenas_i_extents_j_nretained, + atomic_load_zu( + &arenas_i(mib[2])->astats->estats[mib[4]].nretained, + ATOMIC_RELAXED), size_t); +CTL_RO_CGEN(config_stats, stats_arenas_i_extents_j_dirty_bytes, + atomic_load_zu( + &arenas_i(mib[2])->astats->estats[mib[4]].dirty_bytes, + ATOMIC_RELAXED), size_t); +CTL_RO_CGEN(config_stats, stats_arenas_i_extents_j_muzzy_bytes, + atomic_load_zu( + &arenas_i(mib[2])->astats->estats[mib[4]].muzzy_bytes, + ATOMIC_RELAXED), size_t); +CTL_RO_CGEN(config_stats, stats_arenas_i_extents_j_retained_bytes, + atomic_load_zu( + &arenas_i(mib[2])->astats->estats[mib[4]].retained_bytes, + ATOMIC_RELAXED), size_t); + +static const ctl_named_node_t * +stats_arenas_i_extents_j_index(tsdn_t *tsdn, const size_t *mib, + size_t miblen, size_t j) { + if (j >= SC_NPSIZES) { + return NULL; + } + return super_stats_arenas_i_extents_j_node; +} + +static bool +ctl_arenas_i_verify(size_t i) { + size_t a = arenas_i2a_impl(i, true, true); + if (a == UINT_MAX || !ctl_arenas->arenas[a]->initialized) { + return true; + } + + return false; +} + +static const ctl_named_node_t * +stats_arenas_i_index(tsdn_t *tsdn, const size_t *mib, + size_t miblen, size_t i) { + const ctl_named_node_t *ret; + + malloc_mutex_lock(tsdn, &ctl_mtx); + if (ctl_arenas_i_verify(i)) { + ret = NULL; + goto label_return; + } + + ret = super_stats_arenas_i_node; +label_return: + malloc_mutex_unlock(tsdn, &ctl_mtx); + return ret; +} + +static int +experimental_hooks_install_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, + void *oldp, size_t *oldlenp, void *newp, size_t newlen) { + int ret; + if (oldp == NULL || oldlenp == NULL|| newp == NULL) { + ret = EINVAL; + goto label_return; + } + /* + * Note: this is a *private* struct. This is an experimental interface; + * forcing the user to know the jemalloc internals well enough to + * extract the ABI hopefully ensures nobody gets too comfortable with + * this API, which can change at a moment's notice. + */ + hooks_t hooks; + WRITE(hooks, hooks_t); + void *handle = hook_install(tsd_tsdn(tsd), &hooks); + if (handle == NULL) { + ret = EAGAIN; + goto label_return; + } + READ(handle, void *); + + ret = 0; +label_return: + return ret; +} + +static int +experimental_hooks_remove_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, + void *oldp, size_t *oldlenp, void *newp, size_t newlen) { + int ret; + WRITEONLY(); + void *handle = NULL; + WRITE(handle, void *); + if (handle == NULL) { + ret = EINVAL; + goto label_return; + } + hook_remove(tsd_tsdn(tsd), handle); + ret = 0; +label_return: + return ret; +} + +/* + * Output six memory utilization entries for an input pointer, the first one of + * type (void *) and the remaining five of type size_t, describing the following + * (in the same order): + * + * (a) memory address of the extent a potential reallocation would go into, + * == the five fields below describe about the extent the pointer resides in == + * (b) number of free regions in the extent, + * (c) number of regions in the extent, + * (d) size of the extent in terms of bytes, + * (e) total number of free regions in the bin the extent belongs to, and + * (f) total number of regions in the bin the extent belongs to. + * + * Note that "(e)" and "(f)" are only available when stats are enabled; + * otherwise their values are undefined. + * + * This API is mainly intended for small class allocations, where extents are + * used as slab. + * + * In case of large class allocations, "(a)" will be NULL, and "(e)" and "(f)" + * will be zero (if stats are enabled; otherwise undefined). The other three + * fields will be properly set though the values are trivial: "(b)" will be 0, + * "(c)" will be 1, and "(d)" will be the usable size. + * + * The input pointer and size are respectively passed in by newp and newlen, + * and the output fields and size are respectively oldp and *oldlenp. + * + * It can be beneficial to define the following macros to make it easier to + * access the output: + * + * #define SLABCUR_READ(out) (*(void **)out) + * #define COUNTS(out) ((size_t *)((void **)out + 1)) + * #define NFREE_READ(out) COUNTS(out)[0] + * #define NREGS_READ(out) COUNTS(out)[1] + * #define SIZE_READ(out) COUNTS(out)[2] + * #define BIN_NFREE_READ(out) COUNTS(out)[3] + * #define BIN_NREGS_READ(out) COUNTS(out)[4] + * + * and then write e.g. NFREE_READ(oldp) to fetch the output. See the unit test + * test_query in test/unit/extent_util.c for an example. + * + * For a typical defragmentation workflow making use of this API for + * understanding the fragmentation level, please refer to the comment for + * experimental_utilization_batch_query_ctl. + * + * It's up to the application how to determine the significance of + * fragmentation relying on the outputs returned. Possible choices are: + * + * (a) if extent utilization ratio is below certain threshold, + * (b) if extent memory consumption is above certain threshold, + * (c) if extent utilization ratio is significantly below bin utilization ratio, + * (d) if input pointer deviates a lot from potential reallocation address, or + * (e) some selection/combination of the above. + * + * The caller needs to make sure that the input/output arguments are valid, + * in particular, that the size of the output is correct, i.e.: + * + * *oldlenp = sizeof(void *) + sizeof(size_t) * 5 + * + * Otherwise, the function immediately returns EINVAL without touching anything. + * + * In the rare case where there's no associated extent found for the input + * pointer, the function zeros out all output fields and return. Please refer + * to the comment for experimental_utilization_batch_query_ctl to understand the + * motivation from C++. + */ +static int +experimental_utilization_query_ctl(tsd_t *tsd, const size_t *mib, + size_t miblen, void *oldp, size_t *oldlenp, void *newp, size_t newlen) { + int ret; + + assert(sizeof(extent_util_stats_verbose_t) + == sizeof(void *) + sizeof(size_t) * 5); + + if (oldp == NULL || oldlenp == NULL + || *oldlenp != sizeof(extent_util_stats_verbose_t) + || newp == NULL) { + ret = EINVAL; + goto label_return; + } + + void *ptr = NULL; + WRITE(ptr, void *); + extent_util_stats_verbose_t *util_stats + = (extent_util_stats_verbose_t *)oldp; + extent_util_stats_verbose_get(tsd_tsdn(tsd), ptr, + &util_stats->nfree, &util_stats->nregs, &util_stats->size, + &util_stats->bin_nfree, &util_stats->bin_nregs, + &util_stats->slabcur_addr); + ret = 0; + +label_return: + return ret; +} + +/* + * Given an input array of pointers, output three memory utilization entries of + * type size_t for each input pointer about the extent it resides in: + * + * (a) number of free regions in the extent, + * (b) number of regions in the extent, and + * (c) size of the extent in terms of bytes. + * + * This API is mainly intended for small class allocations, where extents are + * used as slab. In case of large class allocations, the outputs are trivial: + * "(a)" will be 0, "(b)" will be 1, and "(c)" will be the usable size. + * + * Note that multiple input pointers may reside on a same extent so the output + * fields may contain duplicates. + * + * The format of the input/output looks like: + * + * input[0]: 1st_pointer_to_query | output[0]: 1st_extent_n_free_regions + * | output[1]: 1st_extent_n_regions + * | output[2]: 1st_extent_size + * input[1]: 2nd_pointer_to_query | output[3]: 2nd_extent_n_free_regions + * | output[4]: 2nd_extent_n_regions + * | output[5]: 2nd_extent_size + * ... | ... + * + * The input array and size are respectively passed in by newp and newlen, and + * the output array and size are respectively oldp and *oldlenp. + * + * It can be beneficial to define the following macros to make it easier to + * access the output: + * + * #define NFREE_READ(out, i) out[(i) * 3] + * #define NREGS_READ(out, i) out[(i) * 3 + 1] + * #define SIZE_READ(out, i) out[(i) * 3 + 2] + * + * and then write e.g. NFREE_READ(oldp, i) to fetch the output. See the unit + * test test_batch in test/unit/extent_util.c for a concrete example. + * + * A typical workflow would be composed of the following steps: + * + * (1) flush tcache: mallctl("thread.tcache.flush", ...) + * (2) initialize input array of pointers to query fragmentation + * (3) allocate output array to hold utilization statistics + * (4) query utilization: mallctl("experimental.utilization.batch_query", ...) + * (5) (optional) decide if it's worthwhile to defragment; otherwise stop here + * (6) disable tcache: mallctl("thread.tcache.enabled", ...) + * (7) defragment allocations with significant fragmentation, e.g.: + * for each allocation { + * if it's fragmented { + * malloc(...); + * memcpy(...); + * free(...); + * } + * } + * (8) enable tcache: mallctl("thread.tcache.enabled", ...) + * + * The application can determine the significance of fragmentation themselves + * relying on the statistics returned, both at the overall level i.e. step "(5)" + * and at individual allocation level i.e. within step "(7)". Possible choices + * are: + * + * (a) whether memory utilization ratio is below certain threshold, + * (b) whether memory consumption is above certain threshold, or + * (c) some combination of the two. + * + * The caller needs to make sure that the input/output arrays are valid and + * their sizes are proper as well as matched, meaning: + * + * (a) newlen = n_pointers * sizeof(const void *) + * (b) *oldlenp = n_pointers * sizeof(size_t) * 3 + * (c) n_pointers > 0 + * + * Otherwise, the function immediately returns EINVAL without touching anything. + * + * In the rare case where there's no associated extent found for some pointers, + * rather than immediately terminating the computation and raising an error, + * the function simply zeros out the corresponding output fields and continues + * the computation until all input pointers are handled. The motivations of + * such a design are as follows: + * + * (a) The function always either processes nothing or processes everything, and + * never leaves the output half touched and half untouched. + * + * (b) It facilitates usage needs especially common in C++. A vast variety of + * C++ objects are instantiated with multiple dynamic memory allocations. For + * example, std::string and std::vector typically use at least two allocations, + * one for the metadata and one for the actual content. Other types may use + * even more allocations. When inquiring about utilization statistics, the + * caller often wants to examine into all such allocations, especially internal + * one(s), rather than just the topmost one. The issue comes when some + * implementations do certain optimizations to reduce/aggregate some internal + * allocations, e.g. putting short strings directly into the metadata, and such + * decisions are not known to the caller. Therefore, we permit pointers to + * memory usages that may not be returned by previous malloc calls, and we + * provide the caller a convenient way to identify such cases. + */ +static int +experimental_utilization_batch_query_ctl(tsd_t *tsd, const size_t *mib, + size_t miblen, void *oldp, size_t *oldlenp, void *newp, size_t newlen) { + int ret; + + assert(sizeof(extent_util_stats_t) == sizeof(size_t) * 3); + + const size_t len = newlen / sizeof(const void *); + if (oldp == NULL || oldlenp == NULL || newp == NULL || newlen == 0 + || newlen != len * sizeof(const void *) + || *oldlenp != len * sizeof(extent_util_stats_t)) { + ret = EINVAL; + goto label_return; + } + + void **ptrs = (void **)newp; + extent_util_stats_t *util_stats = (extent_util_stats_t *)oldp; + size_t i; + for (i = 0; i < len; ++i) { + extent_util_stats_get(tsd_tsdn(tsd), ptrs[i], + &util_stats[i].nfree, &util_stats[i].nregs, + &util_stats[i].size); + } + ret = 0; + +label_return: + return ret; +} + +static const ctl_named_node_t * +experimental_arenas_i_index(tsdn_t *tsdn, const size_t *mib, + size_t miblen, size_t i) { + const ctl_named_node_t *ret; + + malloc_mutex_lock(tsdn, &ctl_mtx); + if (ctl_arenas_i_verify(i)) { + ret = NULL; + goto label_return; + } + ret = super_experimental_arenas_i_node; +label_return: + malloc_mutex_unlock(tsdn, &ctl_mtx); + return ret; +} + +static int +experimental_arenas_i_pactivep_ctl(tsd_t *tsd, const size_t *mib, + size_t miblen, void *oldp, size_t *oldlenp, void *newp, size_t newlen) { + if (!config_stats) { + return ENOENT; + } + if (oldp == NULL || oldlenp == NULL || *oldlenp != sizeof(size_t *)) { + return EINVAL; + } + + unsigned arena_ind; + arena_t *arena; + int ret; + size_t *pactivep; + + malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx); + READONLY(); + MIB_UNSIGNED(arena_ind, 2); + if (arena_ind < narenas_total_get() && (arena = + arena_get(tsd_tsdn(tsd), arena_ind, false)) != NULL) { +#if defined(JEMALLOC_GCC_ATOMIC_ATOMICS) || \ + defined(JEMALLOC_GCC_SYNC_ATOMICS) || defined(_MSC_VER) + /* Expose the underlying counter for fast read. */ + pactivep = (size_t *)&(arena->nactive.repr); + READ(pactivep, size_t *); + ret = 0; +#else + ret = EFAULT; +#endif + } else { + ret = EFAULT; + } +label_return: + malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx); + return ret; +} diff --git a/deps/jemalloc/src/div.c b/deps/jemalloc/src/div.c new file mode 100644 index 000000000..808892a13 --- /dev/null +++ b/deps/jemalloc/src/div.c @@ -0,0 +1,55 @@ +#include "jemalloc/internal/jemalloc_preamble.h" + +#include "jemalloc/internal/div.h" + +#include "jemalloc/internal/assert.h" + +/* + * Suppose we have n = q * d, all integers. We know n and d, and want q = n / d. + * + * For any k, we have (here, all division is exact; not C-style rounding): + * floor(ceil(2^k / d) * n / 2^k) = floor((2^k + r) / d * n / 2^k), where + * r = (-2^k) mod d. + * + * Expanding this out: + * ... = floor(2^k / d * n / 2^k + r / d * n / 2^k) + * = floor(n / d + (r / d) * (n / 2^k)). + * + * The fractional part of n / d is 0 (because of the assumption that d divides n + * exactly), so we have: + * ... = n / d + floor((r / d) * (n / 2^k)) + * + * So that our initial expression is equal to the quantity we seek, so long as + * (r / d) * (n / 2^k) < 1. + * + * r is a remainder mod d, so r < d and r / d < 1 always. We can make + * n / 2 ^ k < 1 by setting k = 32. This gets us a value of magic that works. + */ + +void +div_init(div_info_t *div_info, size_t d) { + /* Nonsensical. */ + assert(d != 0); + /* + * This would make the value of magic too high to fit into a uint32_t + * (we would want magic = 2^32 exactly). This would mess with code gen + * on 32-bit machines. + */ + assert(d != 1); + + uint64_t two_to_k = ((uint64_t)1 << 32); + uint32_t magic = (uint32_t)(two_to_k / d); + + /* + * We want magic = ceil(2^k / d), but C gives us floor. We have to + * increment it unless the result was exact (i.e. unless d is a power of + * two). + */ + if (two_to_k % d != 0) { + magic++; + } + div_info->magic = magic; +#ifdef JEMALLOC_DEBUG + div_info->d = d; +#endif +} diff --git a/deps/jemalloc/src/extent.c b/deps/jemalloc/src/extent.c new file mode 100644 index 000000000..9237f903d --- /dev/null +++ b/deps/jemalloc/src/extent.c @@ -0,0 +1,2403 @@ +#define JEMALLOC_EXTENT_C_ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/assert.h" +#include "jemalloc/internal/extent_dss.h" +#include "jemalloc/internal/extent_mmap.h" +#include "jemalloc/internal/ph.h" +#include "jemalloc/internal/rtree.h" +#include "jemalloc/internal/mutex.h" +#include "jemalloc/internal/mutex_pool.h" + +/******************************************************************************/ +/* Data. */ + +rtree_t extents_rtree; +/* Keyed by the address of the extent_t being protected. */ +mutex_pool_t extent_mutex_pool; + +size_t opt_lg_extent_max_active_fit = LG_EXTENT_MAX_ACTIVE_FIT_DEFAULT; + +static const bitmap_info_t extents_bitmap_info = + BITMAP_INFO_INITIALIZER(SC_NPSIZES+1); + +static void *extent_alloc_default(extent_hooks_t *extent_hooks, void *new_addr, + size_t size, size_t alignment, bool *zero, bool *commit, + unsigned arena_ind); +static bool extent_dalloc_default(extent_hooks_t *extent_hooks, void *addr, + size_t size, bool committed, unsigned arena_ind); +static void extent_destroy_default(extent_hooks_t *extent_hooks, void *addr, + size_t size, bool committed, unsigned arena_ind); +static bool extent_commit_default(extent_hooks_t *extent_hooks, void *addr, + size_t size, size_t offset, size_t length, unsigned arena_ind); +static bool extent_commit_impl(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent, size_t offset, + size_t length, bool growing_retained); +static bool extent_decommit_default(extent_hooks_t *extent_hooks, + void *addr, size_t size, size_t offset, size_t length, unsigned arena_ind); +#ifdef PAGES_CAN_PURGE_LAZY +static bool extent_purge_lazy_default(extent_hooks_t *extent_hooks, void *addr, + size_t size, size_t offset, size_t length, unsigned arena_ind); +#endif +static bool extent_purge_lazy_impl(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent, size_t offset, + size_t length, bool growing_retained); +#ifdef PAGES_CAN_PURGE_FORCED +static bool extent_purge_forced_default(extent_hooks_t *extent_hooks, + void *addr, size_t size, size_t offset, size_t length, unsigned arena_ind); +#endif +static bool extent_purge_forced_impl(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent, size_t offset, + size_t length, bool growing_retained); +static bool extent_split_default(extent_hooks_t *extent_hooks, void *addr, + size_t size, size_t size_a, size_t size_b, bool committed, + unsigned arena_ind); +static extent_t *extent_split_impl(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent, size_t size_a, + szind_t szind_a, bool slab_a, size_t size_b, szind_t szind_b, bool slab_b, + bool growing_retained); +static bool extent_merge_default(extent_hooks_t *extent_hooks, void *addr_a, + size_t size_a, void *addr_b, size_t size_b, bool committed, + unsigned arena_ind); +static bool extent_merge_impl(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *a, extent_t *b, + bool growing_retained); + +const extent_hooks_t extent_hooks_default = { + extent_alloc_default, + extent_dalloc_default, + extent_destroy_default, + extent_commit_default, + extent_decommit_default +#ifdef PAGES_CAN_PURGE_LAZY + , + extent_purge_lazy_default +#else + , + NULL +#endif +#ifdef PAGES_CAN_PURGE_FORCED + , + extent_purge_forced_default +#else + , + NULL +#endif + , + extent_split_default, + extent_merge_default +}; + +/* Used exclusively for gdump triggering. */ +static atomic_zu_t curpages; +static atomic_zu_t highpages; + +/******************************************************************************/ +/* + * Function prototypes for static functions that are referenced prior to + * definition. + */ + +static void extent_deregister(tsdn_t *tsdn, extent_t *extent); +static extent_t *extent_recycle(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extents_t *extents, void *new_addr, + size_t usize, size_t pad, size_t alignment, bool slab, szind_t szind, + bool *zero, bool *commit, bool growing_retained); +static extent_t *extent_try_coalesce(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, rtree_ctx_t *rtree_ctx, extents_t *extents, + extent_t *extent, bool *coalesced, bool growing_retained); +static void extent_record(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extents_t *extents, extent_t *extent, + bool growing_retained); + +/******************************************************************************/ + +#define ATTR_NONE /* does nothing */ + +ph_gen(ATTR_NONE, extent_avail_, extent_tree_t, extent_t, ph_link, + extent_esnead_comp) + +#undef ATTR_NONE + +typedef enum { + lock_result_success, + lock_result_failure, + lock_result_no_extent +} lock_result_t; + +static lock_result_t +extent_rtree_leaf_elm_try_lock(tsdn_t *tsdn, rtree_leaf_elm_t *elm, + extent_t **result, bool inactive_only) { + extent_t *extent1 = rtree_leaf_elm_extent_read(tsdn, &extents_rtree, + elm, true); + + /* Slab implies active extents and should be skipped. */ + if (extent1 == NULL || (inactive_only && rtree_leaf_elm_slab_read(tsdn, + &extents_rtree, elm, true))) { + return lock_result_no_extent; + } + + /* + * It's possible that the extent changed out from under us, and with it + * the leaf->extent mapping. We have to recheck while holding the lock. + */ + extent_lock(tsdn, extent1); + extent_t *extent2 = rtree_leaf_elm_extent_read(tsdn, + &extents_rtree, elm, true); + + if (extent1 == extent2) { + *result = extent1; + return lock_result_success; + } else { + extent_unlock(tsdn, extent1); + return lock_result_failure; + } +} + +/* + * Returns a pool-locked extent_t * if there's one associated with the given + * address, and NULL otherwise. + */ +static extent_t * +extent_lock_from_addr(tsdn_t *tsdn, rtree_ctx_t *rtree_ctx, void *addr, + bool inactive_only) { + extent_t *ret = NULL; + rtree_leaf_elm_t *elm = rtree_leaf_elm_lookup(tsdn, &extents_rtree, + rtree_ctx, (uintptr_t)addr, false, false); + if (elm == NULL) { + return NULL; + } + lock_result_t lock_result; + do { + lock_result = extent_rtree_leaf_elm_try_lock(tsdn, elm, &ret, + inactive_only); + } while (lock_result == lock_result_failure); + return ret; +} + +extent_t * +extent_alloc(tsdn_t *tsdn, arena_t *arena) { + malloc_mutex_lock(tsdn, &arena->extent_avail_mtx); + extent_t *extent = extent_avail_first(&arena->extent_avail); + if (extent == NULL) { + malloc_mutex_unlock(tsdn, &arena->extent_avail_mtx); + return base_alloc_extent(tsdn, arena->base); + } + extent_avail_remove(&arena->extent_avail, extent); + atomic_fetch_sub_zu(&arena->extent_avail_cnt, 1, ATOMIC_RELAXED); + malloc_mutex_unlock(tsdn, &arena->extent_avail_mtx); + return extent; +} + +void +extent_dalloc(tsdn_t *tsdn, arena_t *arena, extent_t *extent) { + malloc_mutex_lock(tsdn, &arena->extent_avail_mtx); + extent_avail_insert(&arena->extent_avail, extent); + atomic_fetch_add_zu(&arena->extent_avail_cnt, 1, ATOMIC_RELAXED); + malloc_mutex_unlock(tsdn, &arena->extent_avail_mtx); +} + +extent_hooks_t * +extent_hooks_get(arena_t *arena) { + return base_extent_hooks_get(arena->base); +} + +extent_hooks_t * +extent_hooks_set(tsd_t *tsd, arena_t *arena, extent_hooks_t *extent_hooks) { + background_thread_info_t *info; + if (have_background_thread) { + info = arena_background_thread_info_get(arena); + malloc_mutex_lock(tsd_tsdn(tsd), &info->mtx); + } + extent_hooks_t *ret = base_extent_hooks_set(arena->base, extent_hooks); + if (have_background_thread) { + malloc_mutex_unlock(tsd_tsdn(tsd), &info->mtx); + } + + return ret; +} + +static void +extent_hooks_assure_initialized(arena_t *arena, + extent_hooks_t **r_extent_hooks) { + if (*r_extent_hooks == EXTENT_HOOKS_INITIALIZER) { + *r_extent_hooks = extent_hooks_get(arena); + } +} + +#ifndef JEMALLOC_JET +static +#endif +size_t +extent_size_quantize_floor(size_t size) { + size_t ret; + pszind_t pind; + + assert(size > 0); + assert((size & PAGE_MASK) == 0); + + pind = sz_psz2ind(size - sz_large_pad + 1); + if (pind == 0) { + /* + * Avoid underflow. This short-circuit would also do the right + * thing for all sizes in the range for which there are + * PAGE-spaced size classes, but it's simplest to just handle + * the one case that would cause erroneous results. + */ + return size; + } + ret = sz_pind2sz(pind - 1) + sz_large_pad; + assert(ret <= size); + return ret; +} + +#ifndef JEMALLOC_JET +static +#endif +size_t +extent_size_quantize_ceil(size_t size) { + size_t ret; + + assert(size > 0); + assert(size - sz_large_pad <= SC_LARGE_MAXCLASS); + assert((size & PAGE_MASK) == 0); + + ret = extent_size_quantize_floor(size); + if (ret < size) { + /* + * Skip a quantization that may have an adequately large extent, + * because under-sized extents may be mixed in. This only + * happens when an unusual size is requested, i.e. for aligned + * allocation, and is just one of several places where linear + * search would potentially find sufficiently aligned available + * memory somewhere lower. + */ + ret = sz_pind2sz(sz_psz2ind(ret - sz_large_pad + 1)) + + sz_large_pad; + } + return ret; +} + +/* Generate pairing heap functions. */ +ph_gen(, extent_heap_, extent_heap_t, extent_t, ph_link, extent_snad_comp) + +bool +extents_init(tsdn_t *tsdn, extents_t *extents, extent_state_t state, + bool delay_coalesce) { + if (malloc_mutex_init(&extents->mtx, "extents", WITNESS_RANK_EXTENTS, + malloc_mutex_rank_exclusive)) { + return true; + } + for (unsigned i = 0; i < SC_NPSIZES + 1; i++) { + extent_heap_new(&extents->heaps[i]); + } + bitmap_init(extents->bitmap, &extents_bitmap_info, true); + extent_list_init(&extents->lru); + atomic_store_zu(&extents->npages, 0, ATOMIC_RELAXED); + extents->state = state; + extents->delay_coalesce = delay_coalesce; + return false; +} + +extent_state_t +extents_state_get(const extents_t *extents) { + return extents->state; +} + +size_t +extents_npages_get(extents_t *extents) { + return atomic_load_zu(&extents->npages, ATOMIC_RELAXED); +} + +size_t +extents_nextents_get(extents_t *extents, pszind_t pind) { + return atomic_load_zu(&extents->nextents[pind], ATOMIC_RELAXED); +} + +size_t +extents_nbytes_get(extents_t *extents, pszind_t pind) { + return atomic_load_zu(&extents->nbytes[pind], ATOMIC_RELAXED); +} + +static void +extents_stats_add(extents_t *extent, pszind_t pind, size_t sz) { + size_t cur = atomic_load_zu(&extent->nextents[pind], ATOMIC_RELAXED); + atomic_store_zu(&extent->nextents[pind], cur + 1, ATOMIC_RELAXED); + cur = atomic_load_zu(&extent->nbytes[pind], ATOMIC_RELAXED); + atomic_store_zu(&extent->nbytes[pind], cur + sz, ATOMIC_RELAXED); +} + +static void +extents_stats_sub(extents_t *extent, pszind_t pind, size_t sz) { + size_t cur = atomic_load_zu(&extent->nextents[pind], ATOMIC_RELAXED); + atomic_store_zu(&extent->nextents[pind], cur - 1, ATOMIC_RELAXED); + cur = atomic_load_zu(&extent->nbytes[pind], ATOMIC_RELAXED); + atomic_store_zu(&extent->nbytes[pind], cur - sz, ATOMIC_RELAXED); +} + +static void +extents_insert_locked(tsdn_t *tsdn, extents_t *extents, extent_t *extent) { + malloc_mutex_assert_owner(tsdn, &extents->mtx); + assert(extent_state_get(extent) == extents->state); + + size_t size = extent_size_get(extent); + size_t psz = extent_size_quantize_floor(size); + pszind_t pind = sz_psz2ind(psz); + if (extent_heap_empty(&extents->heaps[pind])) { + bitmap_unset(extents->bitmap, &extents_bitmap_info, + (size_t)pind); + } + extent_heap_insert(&extents->heaps[pind], extent); + + if (config_stats) { + extents_stats_add(extents, pind, size); + } + + extent_list_append(&extents->lru, extent); + size_t npages = size >> LG_PAGE; + /* + * All modifications to npages hold the mutex (as asserted above), so we + * don't need an atomic fetch-add; we can get by with a load followed by + * a store. + */ + size_t cur_extents_npages = + atomic_load_zu(&extents->npages, ATOMIC_RELAXED); + atomic_store_zu(&extents->npages, cur_extents_npages + npages, + ATOMIC_RELAXED); +} + +static void +extents_remove_locked(tsdn_t *tsdn, extents_t *extents, extent_t *extent) { + malloc_mutex_assert_owner(tsdn, &extents->mtx); + assert(extent_state_get(extent) == extents->state); + + size_t size = extent_size_get(extent); + size_t psz = extent_size_quantize_floor(size); + pszind_t pind = sz_psz2ind(psz); + extent_heap_remove(&extents->heaps[pind], extent); + + if (config_stats) { + extents_stats_sub(extents, pind, size); + } + + if (extent_heap_empty(&extents->heaps[pind])) { + bitmap_set(extents->bitmap, &extents_bitmap_info, + (size_t)pind); + } + extent_list_remove(&extents->lru, extent); + size_t npages = size >> LG_PAGE; + /* + * As in extents_insert_locked, we hold extents->mtx and so don't need + * atomic operations for updating extents->npages. + */ + size_t cur_extents_npages = + atomic_load_zu(&extents->npages, ATOMIC_RELAXED); + assert(cur_extents_npages >= npages); + atomic_store_zu(&extents->npages, + cur_extents_npages - (size >> LG_PAGE), ATOMIC_RELAXED); +} + +/* + * Find an extent with size [min_size, max_size) to satisfy the alignment + * requirement. For each size, try only the first extent in the heap. + */ +static extent_t * +extents_fit_alignment(extents_t *extents, size_t min_size, size_t max_size, + size_t alignment) { + pszind_t pind = sz_psz2ind(extent_size_quantize_ceil(min_size)); + pszind_t pind_max = sz_psz2ind(extent_size_quantize_ceil(max_size)); + + for (pszind_t i = (pszind_t)bitmap_ffu(extents->bitmap, + &extents_bitmap_info, (size_t)pind); i < pind_max; i = + (pszind_t)bitmap_ffu(extents->bitmap, &extents_bitmap_info, + (size_t)i+1)) { + assert(i < SC_NPSIZES); + assert(!extent_heap_empty(&extents->heaps[i])); + extent_t *extent = extent_heap_first(&extents->heaps[i]); + uintptr_t base = (uintptr_t)extent_base_get(extent); + size_t candidate_size = extent_size_get(extent); + assert(candidate_size >= min_size); + + uintptr_t next_align = ALIGNMENT_CEILING((uintptr_t)base, + PAGE_CEILING(alignment)); + if (base > next_align || base + candidate_size <= next_align) { + /* Overflow or not crossing the next alignment. */ + continue; + } + + size_t leadsize = next_align - base; + if (candidate_size - leadsize >= min_size) { + return extent; + } + } + + return NULL; +} + +/* + * Do first-fit extent selection, i.e. select the oldest/lowest extent that is + * large enough. + */ +static extent_t * +extents_first_fit_locked(tsdn_t *tsdn, arena_t *arena, extents_t *extents, + size_t size) { + extent_t *ret = NULL; + + pszind_t pind = sz_psz2ind(extent_size_quantize_ceil(size)); + + if (!maps_coalesce && !opt_retain) { + /* + * No split / merge allowed (Windows w/o retain). Try exact fit + * only. + */ + return extent_heap_empty(&extents->heaps[pind]) ? NULL : + extent_heap_first(&extents->heaps[pind]); + } + + for (pszind_t i = (pszind_t)bitmap_ffu(extents->bitmap, + &extents_bitmap_info, (size_t)pind); + i < SC_NPSIZES + 1; + i = (pszind_t)bitmap_ffu(extents->bitmap, &extents_bitmap_info, + (size_t)i+1)) { + assert(!extent_heap_empty(&extents->heaps[i])); + extent_t *extent = extent_heap_first(&extents->heaps[i]); + assert(extent_size_get(extent) >= size); + /* + * In order to reduce fragmentation, avoid reusing and splitting + * large extents for much smaller sizes. + * + * Only do check for dirty extents (delay_coalesce). + */ + if (extents->delay_coalesce && + (sz_pind2sz(i) >> opt_lg_extent_max_active_fit) > size) { + break; + } + if (ret == NULL || extent_snad_comp(extent, ret) < 0) { + ret = extent; + } + if (i == SC_NPSIZES) { + break; + } + assert(i < SC_NPSIZES); + } + + return ret; +} + +/* + * Do first-fit extent selection, where the selection policy choice is + * based on extents->delay_coalesce. + */ +static extent_t * +extents_fit_locked(tsdn_t *tsdn, arena_t *arena, extents_t *extents, + size_t esize, size_t alignment) { + malloc_mutex_assert_owner(tsdn, &extents->mtx); + + size_t max_size = esize + PAGE_CEILING(alignment) - PAGE; + /* Beware size_t wrap-around. */ + if (max_size < esize) { + return NULL; + } + + extent_t *extent = + extents_first_fit_locked(tsdn, arena, extents, max_size); + + if (alignment > PAGE && extent == NULL) { + /* + * max_size guarantees the alignment requirement but is rather + * pessimistic. Next we try to satisfy the aligned allocation + * with sizes in [esize, max_size). + */ + extent = extents_fit_alignment(extents, esize, max_size, + alignment); + } + + return extent; +} + +static bool +extent_try_delayed_coalesce(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, rtree_ctx_t *rtree_ctx, extents_t *extents, + extent_t *extent) { + extent_state_set(extent, extent_state_active); + bool coalesced; + extent = extent_try_coalesce(tsdn, arena, r_extent_hooks, rtree_ctx, + extents, extent, &coalesced, false); + extent_state_set(extent, extents_state_get(extents)); + + if (!coalesced) { + return true; + } + extents_insert_locked(tsdn, extents, extent); + return false; +} + +extent_t * +extents_alloc(tsdn_t *tsdn, arena_t *arena, extent_hooks_t **r_extent_hooks, + extents_t *extents, void *new_addr, size_t size, size_t pad, + size_t alignment, bool slab, szind_t szind, bool *zero, bool *commit) { + assert(size + pad != 0); + assert(alignment != 0); + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, 0); + + extent_t *extent = extent_recycle(tsdn, arena, r_extent_hooks, extents, + new_addr, size, pad, alignment, slab, szind, zero, commit, false); + assert(extent == NULL || extent_dumpable_get(extent)); + return extent; +} + +void +extents_dalloc(tsdn_t *tsdn, arena_t *arena, extent_hooks_t **r_extent_hooks, + extents_t *extents, extent_t *extent) { + assert(extent_base_get(extent) != NULL); + assert(extent_size_get(extent) != 0); + assert(extent_dumpable_get(extent)); + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, 0); + + extent_addr_set(extent, extent_base_get(extent)); + extent_zeroed_set(extent, false); + + extent_record(tsdn, arena, r_extent_hooks, extents, extent, false); +} + +extent_t * +extents_evict(tsdn_t *tsdn, arena_t *arena, extent_hooks_t **r_extent_hooks, + extents_t *extents, size_t npages_min) { + rtree_ctx_t rtree_ctx_fallback; + rtree_ctx_t *rtree_ctx = tsdn_rtree_ctx(tsdn, &rtree_ctx_fallback); + + malloc_mutex_lock(tsdn, &extents->mtx); + + /* + * Get the LRU coalesced extent, if any. If coalescing was delayed, + * the loop will iterate until the LRU extent is fully coalesced. + */ + extent_t *extent; + while (true) { + /* Get the LRU extent, if any. */ + extent = extent_list_first(&extents->lru); + if (extent == NULL) { + goto label_return; + } + /* Check the eviction limit. */ + size_t extents_npages = atomic_load_zu(&extents->npages, + ATOMIC_RELAXED); + if (extents_npages <= npages_min) { + extent = NULL; + goto label_return; + } + extents_remove_locked(tsdn, extents, extent); + if (!extents->delay_coalesce) { + break; + } + /* Try to coalesce. */ + if (extent_try_delayed_coalesce(tsdn, arena, r_extent_hooks, + rtree_ctx, extents, extent)) { + break; + } + /* + * The LRU extent was just coalesced and the result placed in + * the LRU at its neighbor's position. Start over. + */ + } + + /* + * Either mark the extent active or deregister it to protect against + * concurrent operations. + */ + switch (extents_state_get(extents)) { + case extent_state_active: + not_reached(); + case extent_state_dirty: + case extent_state_muzzy: + extent_state_set(extent, extent_state_active); + break; + case extent_state_retained: + extent_deregister(tsdn, extent); + break; + default: + not_reached(); + } + +label_return: + malloc_mutex_unlock(tsdn, &extents->mtx); + return extent; +} + +/* + * This can only happen when we fail to allocate a new extent struct (which + * indicates OOM), e.g. when trying to split an existing extent. + */ +static void +extents_abandon_vm(tsdn_t *tsdn, arena_t *arena, extent_hooks_t **r_extent_hooks, + extents_t *extents, extent_t *extent, bool growing_retained) { + size_t sz = extent_size_get(extent); + if (config_stats) { + arena_stats_accum_zu(&arena->stats.abandoned_vm, sz); + } + /* + * Leak extent after making sure its pages have already been purged, so + * that this is only a virtual memory leak. + */ + if (extents_state_get(extents) == extent_state_dirty) { + if (extent_purge_lazy_impl(tsdn, arena, r_extent_hooks, + extent, 0, sz, growing_retained)) { + extent_purge_forced_impl(tsdn, arena, r_extent_hooks, + extent, 0, extent_size_get(extent), + growing_retained); + } + } + extent_dalloc(tsdn, arena, extent); +} + +void +extents_prefork(tsdn_t *tsdn, extents_t *extents) { + malloc_mutex_prefork(tsdn, &extents->mtx); +} + +void +extents_postfork_parent(tsdn_t *tsdn, extents_t *extents) { + malloc_mutex_postfork_parent(tsdn, &extents->mtx); +} + +void +extents_postfork_child(tsdn_t *tsdn, extents_t *extents) { + malloc_mutex_postfork_child(tsdn, &extents->mtx); +} + +static void +extent_deactivate_locked(tsdn_t *tsdn, arena_t *arena, extents_t *extents, + extent_t *extent) { + assert(extent_arena_get(extent) == arena); + assert(extent_state_get(extent) == extent_state_active); + + extent_state_set(extent, extents_state_get(extents)); + extents_insert_locked(tsdn, extents, extent); +} + +static void +extent_deactivate(tsdn_t *tsdn, arena_t *arena, extents_t *extents, + extent_t *extent) { + malloc_mutex_lock(tsdn, &extents->mtx); + extent_deactivate_locked(tsdn, arena, extents, extent); + malloc_mutex_unlock(tsdn, &extents->mtx); +} + +static void +extent_activate_locked(tsdn_t *tsdn, arena_t *arena, extents_t *extents, + extent_t *extent) { + assert(extent_arena_get(extent) == arena); + assert(extent_state_get(extent) == extents_state_get(extents)); + + extents_remove_locked(tsdn, extents, extent); + extent_state_set(extent, extent_state_active); +} + +static bool +extent_rtree_leaf_elms_lookup(tsdn_t *tsdn, rtree_ctx_t *rtree_ctx, + const extent_t *extent, bool dependent, bool init_missing, + rtree_leaf_elm_t **r_elm_a, rtree_leaf_elm_t **r_elm_b) { + *r_elm_a = rtree_leaf_elm_lookup(tsdn, &extents_rtree, rtree_ctx, + (uintptr_t)extent_base_get(extent), dependent, init_missing); + if (!dependent && *r_elm_a == NULL) { + return true; + } + assert(*r_elm_a != NULL); + + *r_elm_b = rtree_leaf_elm_lookup(tsdn, &extents_rtree, rtree_ctx, + (uintptr_t)extent_last_get(extent), dependent, init_missing); + if (!dependent && *r_elm_b == NULL) { + return true; + } + assert(*r_elm_b != NULL); + + return false; +} + +static void +extent_rtree_write_acquired(tsdn_t *tsdn, rtree_leaf_elm_t *elm_a, + rtree_leaf_elm_t *elm_b, extent_t *extent, szind_t szind, bool slab) { + rtree_leaf_elm_write(tsdn, &extents_rtree, elm_a, extent, szind, slab); + if (elm_b != NULL) { + rtree_leaf_elm_write(tsdn, &extents_rtree, elm_b, extent, szind, + slab); + } +} + +static void +extent_interior_register(tsdn_t *tsdn, rtree_ctx_t *rtree_ctx, extent_t *extent, + szind_t szind) { + assert(extent_slab_get(extent)); + + /* Register interior. */ + for (size_t i = 1; i < (extent_size_get(extent) >> LG_PAGE) - 1; i++) { + rtree_write(tsdn, &extents_rtree, rtree_ctx, + (uintptr_t)extent_base_get(extent) + (uintptr_t)(i << + LG_PAGE), extent, szind, true); + } +} + +static void +extent_gdump_add(tsdn_t *tsdn, const extent_t *extent) { + cassert(config_prof); + /* prof_gdump() requirement. */ + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, 0); + + if (opt_prof && extent_state_get(extent) == extent_state_active) { + size_t nadd = extent_size_get(extent) >> LG_PAGE; + size_t cur = atomic_fetch_add_zu(&curpages, nadd, + ATOMIC_RELAXED) + nadd; + size_t high = atomic_load_zu(&highpages, ATOMIC_RELAXED); + while (cur > high && !atomic_compare_exchange_weak_zu( + &highpages, &high, cur, ATOMIC_RELAXED, ATOMIC_RELAXED)) { + /* + * Don't refresh cur, because it may have decreased + * since this thread lost the highpages update race. + * Note that high is updated in case of CAS failure. + */ + } + if (cur > high && prof_gdump_get_unlocked()) { + prof_gdump(tsdn); + } + } +} + +static void +extent_gdump_sub(tsdn_t *tsdn, const extent_t *extent) { + cassert(config_prof); + + if (opt_prof && extent_state_get(extent) == extent_state_active) { + size_t nsub = extent_size_get(extent) >> LG_PAGE; + assert(atomic_load_zu(&curpages, ATOMIC_RELAXED) >= nsub); + atomic_fetch_sub_zu(&curpages, nsub, ATOMIC_RELAXED); + } +} + +static bool +extent_register_impl(tsdn_t *tsdn, extent_t *extent, bool gdump_add) { + rtree_ctx_t rtree_ctx_fallback; + rtree_ctx_t *rtree_ctx = tsdn_rtree_ctx(tsdn, &rtree_ctx_fallback); + rtree_leaf_elm_t *elm_a, *elm_b; + + /* + * We need to hold the lock to protect against a concurrent coalesce + * operation that sees us in a partial state. + */ + extent_lock(tsdn, extent); + + if (extent_rtree_leaf_elms_lookup(tsdn, rtree_ctx, extent, false, true, + &elm_a, &elm_b)) { + extent_unlock(tsdn, extent); + return true; + } + + szind_t szind = extent_szind_get_maybe_invalid(extent); + bool slab = extent_slab_get(extent); + extent_rtree_write_acquired(tsdn, elm_a, elm_b, extent, szind, slab); + if (slab) { + extent_interior_register(tsdn, rtree_ctx, extent, szind); + } + + extent_unlock(tsdn, extent); + + if (config_prof && gdump_add) { + extent_gdump_add(tsdn, extent); + } + + return false; +} + +static bool +extent_register(tsdn_t *tsdn, extent_t *extent) { + return extent_register_impl(tsdn, extent, true); +} + +static bool +extent_register_no_gdump_add(tsdn_t *tsdn, extent_t *extent) { + return extent_register_impl(tsdn, extent, false); +} + +static void +extent_reregister(tsdn_t *tsdn, extent_t *extent) { + bool err = extent_register(tsdn, extent); + assert(!err); +} + +/* + * Removes all pointers to the given extent from the global rtree indices for + * its interior. This is relevant for slab extents, for which we need to do + * metadata lookups at places other than the head of the extent. We deregister + * on the interior, then, when an extent moves from being an active slab to an + * inactive state. + */ +static void +extent_interior_deregister(tsdn_t *tsdn, rtree_ctx_t *rtree_ctx, + extent_t *extent) { + size_t i; + + assert(extent_slab_get(extent)); + + for (i = 1; i < (extent_size_get(extent) >> LG_PAGE) - 1; i++) { + rtree_clear(tsdn, &extents_rtree, rtree_ctx, + (uintptr_t)extent_base_get(extent) + (uintptr_t)(i << + LG_PAGE)); + } +} + +/* + * Removes all pointers to the given extent from the global rtree. + */ +static void +extent_deregister_impl(tsdn_t *tsdn, extent_t *extent, bool gdump) { + rtree_ctx_t rtree_ctx_fallback; + rtree_ctx_t *rtree_ctx = tsdn_rtree_ctx(tsdn, &rtree_ctx_fallback); + rtree_leaf_elm_t *elm_a, *elm_b; + extent_rtree_leaf_elms_lookup(tsdn, rtree_ctx, extent, true, false, + &elm_a, &elm_b); + + extent_lock(tsdn, extent); + + extent_rtree_write_acquired(tsdn, elm_a, elm_b, NULL, SC_NSIZES, false); + if (extent_slab_get(extent)) { + extent_interior_deregister(tsdn, rtree_ctx, extent); + extent_slab_set(extent, false); + } + + extent_unlock(tsdn, extent); + + if (config_prof && gdump) { + extent_gdump_sub(tsdn, extent); + } +} + +static void +extent_deregister(tsdn_t *tsdn, extent_t *extent) { + extent_deregister_impl(tsdn, extent, true); +} + +static void +extent_deregister_no_gdump_sub(tsdn_t *tsdn, extent_t *extent) { + extent_deregister_impl(tsdn, extent, false); +} + +/* + * Tries to find and remove an extent from extents that can be used for the + * given allocation request. + */ +static extent_t * +extent_recycle_extract(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, rtree_ctx_t *rtree_ctx, extents_t *extents, + void *new_addr, size_t size, size_t pad, size_t alignment, bool slab, + bool growing_retained) { + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, growing_retained ? 1 : 0); + assert(alignment > 0); + if (config_debug && new_addr != NULL) { + /* + * Non-NULL new_addr has two use cases: + * + * 1) Recycle a known-extant extent, e.g. during purging. + * 2) Perform in-place expanding reallocation. + * + * Regardless of use case, new_addr must either refer to a + * non-existing extent, or to the base of an extant extent, + * since only active slabs support interior lookups (which of + * course cannot be recycled). + */ + assert(PAGE_ADDR2BASE(new_addr) == new_addr); + assert(pad == 0); + assert(alignment <= PAGE); + } + + size_t esize = size + pad; + malloc_mutex_lock(tsdn, &extents->mtx); + extent_hooks_assure_initialized(arena, r_extent_hooks); + extent_t *extent; + if (new_addr != NULL) { + extent = extent_lock_from_addr(tsdn, rtree_ctx, new_addr, + false); + if (extent != NULL) { + /* + * We might null-out extent to report an error, but we + * still need to unlock the associated mutex after. + */ + extent_t *unlock_extent = extent; + assert(extent_base_get(extent) == new_addr); + if (extent_arena_get(extent) != arena || + extent_size_get(extent) < esize || + extent_state_get(extent) != + extents_state_get(extents)) { + extent = NULL; + } + extent_unlock(tsdn, unlock_extent); + } + } else { + extent = extents_fit_locked(tsdn, arena, extents, esize, + alignment); + } + if (extent == NULL) { + malloc_mutex_unlock(tsdn, &extents->mtx); + return NULL; + } + + extent_activate_locked(tsdn, arena, extents, extent); + malloc_mutex_unlock(tsdn, &extents->mtx); + + return extent; +} + +/* + * Given an allocation request and an extent guaranteed to be able to satisfy + * it, this splits off lead and trail extents, leaving extent pointing to an + * extent satisfying the allocation. + * This function doesn't put lead or trail into any extents_t; it's the caller's + * job to ensure that they can be reused. + */ +typedef enum { + /* + * Split successfully. lead, extent, and trail, are modified to extents + * describing the ranges before, in, and after the given allocation. + */ + extent_split_interior_ok, + /* + * The extent can't satisfy the given allocation request. None of the + * input extent_t *s are touched. + */ + extent_split_interior_cant_alloc, + /* + * In a potentially invalid state. Must leak (if *to_leak is non-NULL), + * and salvage what's still salvageable (if *to_salvage is non-NULL). + * None of lead, extent, or trail are valid. + */ + extent_split_interior_error +} extent_split_interior_result_t; + +static extent_split_interior_result_t +extent_split_interior(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, rtree_ctx_t *rtree_ctx, + /* The result of splitting, in case of success. */ + extent_t **extent, extent_t **lead, extent_t **trail, + /* The mess to clean up, in case of error. */ + extent_t **to_leak, extent_t **to_salvage, + void *new_addr, size_t size, size_t pad, size_t alignment, bool slab, + szind_t szind, bool growing_retained) { + size_t esize = size + pad; + size_t leadsize = ALIGNMENT_CEILING((uintptr_t)extent_base_get(*extent), + PAGE_CEILING(alignment)) - (uintptr_t)extent_base_get(*extent); + assert(new_addr == NULL || leadsize == 0); + if (extent_size_get(*extent) < leadsize + esize) { + return extent_split_interior_cant_alloc; + } + size_t trailsize = extent_size_get(*extent) - leadsize - esize; + + *lead = NULL; + *trail = NULL; + *to_leak = NULL; + *to_salvage = NULL; + + /* Split the lead. */ + if (leadsize != 0) { + *lead = *extent; + *extent = extent_split_impl(tsdn, arena, r_extent_hooks, + *lead, leadsize, SC_NSIZES, false, esize + trailsize, szind, + slab, growing_retained); + if (*extent == NULL) { + *to_leak = *lead; + *lead = NULL; + return extent_split_interior_error; + } + } + + /* Split the trail. */ + if (trailsize != 0) { + *trail = extent_split_impl(tsdn, arena, r_extent_hooks, *extent, + esize, szind, slab, trailsize, SC_NSIZES, false, + growing_retained); + if (*trail == NULL) { + *to_leak = *extent; + *to_salvage = *lead; + *lead = NULL; + *extent = NULL; + return extent_split_interior_error; + } + } + + if (leadsize == 0 && trailsize == 0) { + /* + * Splitting causes szind to be set as a side effect, but no + * splitting occurred. + */ + extent_szind_set(*extent, szind); + if (szind != SC_NSIZES) { + rtree_szind_slab_update(tsdn, &extents_rtree, rtree_ctx, + (uintptr_t)extent_addr_get(*extent), szind, slab); + if (slab && extent_size_get(*extent) > PAGE) { + rtree_szind_slab_update(tsdn, &extents_rtree, + rtree_ctx, + (uintptr_t)extent_past_get(*extent) - + (uintptr_t)PAGE, szind, slab); + } + } + } + + return extent_split_interior_ok; +} + +/* + * This fulfills the indicated allocation request out of the given extent (which + * the caller should have ensured was big enough). If there's any unused space + * before or after the resulting allocation, that space is given its own extent + * and put back into extents. + */ +static extent_t * +extent_recycle_split(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, rtree_ctx_t *rtree_ctx, extents_t *extents, + void *new_addr, size_t size, size_t pad, size_t alignment, bool slab, + szind_t szind, extent_t *extent, bool growing_retained) { + extent_t *lead; + extent_t *trail; + extent_t *to_leak; + extent_t *to_salvage; + + extent_split_interior_result_t result = extent_split_interior( + tsdn, arena, r_extent_hooks, rtree_ctx, &extent, &lead, &trail, + &to_leak, &to_salvage, new_addr, size, pad, alignment, slab, szind, + growing_retained); + + if (!maps_coalesce && result != extent_split_interior_ok + && !opt_retain) { + /* + * Split isn't supported (implies Windows w/o retain). Avoid + * leaking the extents. + */ + assert(to_leak != NULL && lead == NULL && trail == NULL); + extent_deactivate(tsdn, arena, extents, to_leak); + return NULL; + } + + if (result == extent_split_interior_ok) { + if (lead != NULL) { + extent_deactivate(tsdn, arena, extents, lead); + } + if (trail != NULL) { + extent_deactivate(tsdn, arena, extents, trail); + } + return extent; + } else { + /* + * We should have picked an extent that was large enough to + * fulfill our allocation request. + */ + assert(result == extent_split_interior_error); + if (to_salvage != NULL) { + extent_deregister(tsdn, to_salvage); + } + if (to_leak != NULL) { + void *leak = extent_base_get(to_leak); + extent_deregister_no_gdump_sub(tsdn, to_leak); + extents_abandon_vm(tsdn, arena, r_extent_hooks, extents, + to_leak, growing_retained); + assert(extent_lock_from_addr(tsdn, rtree_ctx, leak, + false) == NULL); + } + return NULL; + } + unreachable(); +} + +static bool +extent_need_manual_zero(arena_t *arena) { + /* + * Need to manually zero the extent on repopulating if either; 1) non + * default extent hooks installed (in which case the purge semantics may + * change); or 2) transparent huge pages enabled. + */ + return (!arena_has_default_hooks(arena) || + (opt_thp == thp_mode_always)); +} + +/* + * Tries to satisfy the given allocation request by reusing one of the extents + * in the given extents_t. + */ +static extent_t * +extent_recycle(tsdn_t *tsdn, arena_t *arena, extent_hooks_t **r_extent_hooks, + extents_t *extents, void *new_addr, size_t size, size_t pad, + size_t alignment, bool slab, szind_t szind, bool *zero, bool *commit, + bool growing_retained) { + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, growing_retained ? 1 : 0); + assert(new_addr == NULL || !slab); + assert(pad == 0 || !slab); + assert(!*zero || !slab); + + rtree_ctx_t rtree_ctx_fallback; + rtree_ctx_t *rtree_ctx = tsdn_rtree_ctx(tsdn, &rtree_ctx_fallback); + + extent_t *extent = extent_recycle_extract(tsdn, arena, r_extent_hooks, + rtree_ctx, extents, new_addr, size, pad, alignment, slab, + growing_retained); + if (extent == NULL) { + return NULL; + } + + extent = extent_recycle_split(tsdn, arena, r_extent_hooks, rtree_ctx, + extents, new_addr, size, pad, alignment, slab, szind, extent, + growing_retained); + if (extent == NULL) { + return NULL; + } + + if (*commit && !extent_committed_get(extent)) { + if (extent_commit_impl(tsdn, arena, r_extent_hooks, extent, + 0, extent_size_get(extent), growing_retained)) { + extent_record(tsdn, arena, r_extent_hooks, extents, + extent, growing_retained); + return NULL; + } + if (!extent_need_manual_zero(arena)) { + extent_zeroed_set(extent, true); + } + } + + if (extent_committed_get(extent)) { + *commit = true; + } + if (extent_zeroed_get(extent)) { + *zero = true; + } + + if (pad != 0) { + extent_addr_randomize(tsdn, extent, alignment); + } + assert(extent_state_get(extent) == extent_state_active); + if (slab) { + extent_slab_set(extent, slab); + extent_interior_register(tsdn, rtree_ctx, extent, szind); + } + + if (*zero) { + void *addr = extent_base_get(extent); + if (!extent_zeroed_get(extent)) { + size_t size = extent_size_get(extent); + if (extent_need_manual_zero(arena) || + pages_purge_forced(addr, size)) { + memset(addr, 0, size); + } + } else if (config_debug) { + size_t *p = (size_t *)(uintptr_t)addr; + /* Check the first page only. */ + for (size_t i = 0; i < PAGE / sizeof(size_t); i++) { + assert(p[i] == 0); + } + } + } + return extent; +} + +/* + * If the caller specifies (!*zero), it is still possible to receive zeroed + * memory, in which case *zero is toggled to true. arena_extent_alloc() takes + * advantage of this to avoid demanding zeroed extents, but taking advantage of + * them if they are returned. + */ +static void * +extent_alloc_core(tsdn_t *tsdn, arena_t *arena, void *new_addr, size_t size, + size_t alignment, bool *zero, bool *commit, dss_prec_t dss_prec) { + void *ret; + + assert(size != 0); + assert(alignment != 0); + + /* "primary" dss. */ + if (have_dss && dss_prec == dss_prec_primary && (ret = + extent_alloc_dss(tsdn, arena, new_addr, size, alignment, zero, + commit)) != NULL) { + return ret; + } + /* mmap. */ + if ((ret = extent_alloc_mmap(new_addr, size, alignment, zero, commit)) + != NULL) { + return ret; + } + /* "secondary" dss. */ + if (have_dss && dss_prec == dss_prec_secondary && (ret = + extent_alloc_dss(tsdn, arena, new_addr, size, alignment, zero, + commit)) != NULL) { + return ret; + } + + /* All strategies for allocation failed. */ + return NULL; +} + +static void * +extent_alloc_default_impl(tsdn_t *tsdn, arena_t *arena, void *new_addr, + size_t size, size_t alignment, bool *zero, bool *commit) { + void *ret = extent_alloc_core(tsdn, arena, new_addr, size, alignment, zero, + commit, (dss_prec_t)atomic_load_u(&arena->dss_prec, + ATOMIC_RELAXED)); + if (have_madvise_huge && ret) { + pages_set_thp_state(ret, size); + } + return ret; +} + +static void * +extent_alloc_default(extent_hooks_t *extent_hooks, void *new_addr, size_t size, + size_t alignment, bool *zero, bool *commit, unsigned arena_ind) { + tsdn_t *tsdn; + arena_t *arena; + + tsdn = tsdn_fetch(); + arena = arena_get(tsdn, arena_ind, false); + /* + * The arena we're allocating on behalf of must have been initialized + * already. + */ + assert(arena != NULL); + + return extent_alloc_default_impl(tsdn, arena, new_addr, size, + ALIGNMENT_CEILING(alignment, PAGE), zero, commit); +} + +static void +extent_hook_pre_reentrancy(tsdn_t *tsdn, arena_t *arena) { + tsd_t *tsd = tsdn_null(tsdn) ? tsd_fetch() : tsdn_tsd(tsdn); + if (arena == arena_get(tsd_tsdn(tsd), 0, false)) { + /* + * The only legitimate case of customized extent hooks for a0 is + * hooks with no allocation activities. One such example is to + * place metadata on pre-allocated resources such as huge pages. + * In that case, rely on reentrancy_level checks to catch + * infinite recursions. + */ + pre_reentrancy(tsd, NULL); + } else { + pre_reentrancy(tsd, arena); + } +} + +static void +extent_hook_post_reentrancy(tsdn_t *tsdn) { + tsd_t *tsd = tsdn_null(tsdn) ? tsd_fetch() : tsdn_tsd(tsdn); + post_reentrancy(tsd); +} + +/* + * If virtual memory is retained, create increasingly larger extents from which + * to split requested extents in order to limit the total number of disjoint + * virtual memory ranges retained by each arena. + */ +static extent_t * +extent_grow_retained(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, size_t size, size_t pad, size_t alignment, + bool slab, szind_t szind, bool *zero, bool *commit) { + malloc_mutex_assert_owner(tsdn, &arena->extent_grow_mtx); + assert(pad == 0 || !slab); + assert(!*zero || !slab); + + size_t esize = size + pad; + size_t alloc_size_min = esize + PAGE_CEILING(alignment) - PAGE; + /* Beware size_t wrap-around. */ + if (alloc_size_min < esize) { + goto label_err; + } + /* + * Find the next extent size in the series that would be large enough to + * satisfy this request. + */ + pszind_t egn_skip = 0; + size_t alloc_size = sz_pind2sz(arena->extent_grow_next + egn_skip); + while (alloc_size < alloc_size_min) { + egn_skip++; + if (arena->extent_grow_next + egn_skip >= + sz_psz2ind(SC_LARGE_MAXCLASS)) { + /* Outside legal range. */ + goto label_err; + } + alloc_size = sz_pind2sz(arena->extent_grow_next + egn_skip); + } + + extent_t *extent = extent_alloc(tsdn, arena); + if (extent == NULL) { + goto label_err; + } + bool zeroed = false; + bool committed = false; + + void *ptr; + if (*r_extent_hooks == &extent_hooks_default) { + ptr = extent_alloc_default_impl(tsdn, arena, NULL, + alloc_size, PAGE, &zeroed, &committed); + } else { + extent_hook_pre_reentrancy(tsdn, arena); + ptr = (*r_extent_hooks)->alloc(*r_extent_hooks, NULL, + alloc_size, PAGE, &zeroed, &committed, + arena_ind_get(arena)); + extent_hook_post_reentrancy(tsdn); + } + + extent_init(extent, arena, ptr, alloc_size, false, SC_NSIZES, + arena_extent_sn_next(arena), extent_state_active, zeroed, + committed, true, EXTENT_IS_HEAD); + if (ptr == NULL) { + extent_dalloc(tsdn, arena, extent); + goto label_err; + } + + if (extent_register_no_gdump_add(tsdn, extent)) { + extent_dalloc(tsdn, arena, extent); + goto label_err; + } + + if (extent_zeroed_get(extent) && extent_committed_get(extent)) { + *zero = true; + } + if (extent_committed_get(extent)) { + *commit = true; + } + + rtree_ctx_t rtree_ctx_fallback; + rtree_ctx_t *rtree_ctx = tsdn_rtree_ctx(tsdn, &rtree_ctx_fallback); + + extent_t *lead; + extent_t *trail; + extent_t *to_leak; + extent_t *to_salvage; + extent_split_interior_result_t result = extent_split_interior( + tsdn, arena, r_extent_hooks, rtree_ctx, &extent, &lead, &trail, + &to_leak, &to_salvage, NULL, size, pad, alignment, slab, szind, + true); + + if (result == extent_split_interior_ok) { + if (lead != NULL) { + extent_record(tsdn, arena, r_extent_hooks, + &arena->extents_retained, lead, true); + } + if (trail != NULL) { + extent_record(tsdn, arena, r_extent_hooks, + &arena->extents_retained, trail, true); + } + } else { + /* + * We should have allocated a sufficiently large extent; the + * cant_alloc case should not occur. + */ + assert(result == extent_split_interior_error); + if (to_salvage != NULL) { + if (config_prof) { + extent_gdump_add(tsdn, to_salvage); + } + extent_record(tsdn, arena, r_extent_hooks, + &arena->extents_retained, to_salvage, true); + } + if (to_leak != NULL) { + extent_deregister_no_gdump_sub(tsdn, to_leak); + extents_abandon_vm(tsdn, arena, r_extent_hooks, + &arena->extents_retained, to_leak, true); + } + goto label_err; + } + + if (*commit && !extent_committed_get(extent)) { + if (extent_commit_impl(tsdn, arena, r_extent_hooks, extent, 0, + extent_size_get(extent), true)) { + extent_record(tsdn, arena, r_extent_hooks, + &arena->extents_retained, extent, true); + goto label_err; + } + if (!extent_need_manual_zero(arena)) { + extent_zeroed_set(extent, true); + } + } + + /* + * Increment extent_grow_next if doing so wouldn't exceed the allowed + * range. + */ + if (arena->extent_grow_next + egn_skip + 1 <= + arena->retain_grow_limit) { + arena->extent_grow_next += egn_skip + 1; + } else { + arena->extent_grow_next = arena->retain_grow_limit; + } + /* All opportunities for failure are past. */ + malloc_mutex_unlock(tsdn, &arena->extent_grow_mtx); + + if (config_prof) { + /* Adjust gdump stats now that extent is final size. */ + extent_gdump_add(tsdn, extent); + } + if (pad != 0) { + extent_addr_randomize(tsdn, extent, alignment); + } + if (slab) { + rtree_ctx_t rtree_ctx_fallback; + rtree_ctx_t *rtree_ctx = tsdn_rtree_ctx(tsdn, + &rtree_ctx_fallback); + + extent_slab_set(extent, true); + extent_interior_register(tsdn, rtree_ctx, extent, szind); + } + if (*zero && !extent_zeroed_get(extent)) { + void *addr = extent_base_get(extent); + size_t size = extent_size_get(extent); + if (extent_need_manual_zero(arena) || + pages_purge_forced(addr, size)) { + memset(addr, 0, size); + } + } + + return extent; +label_err: + malloc_mutex_unlock(tsdn, &arena->extent_grow_mtx); + return NULL; +} + +static extent_t * +extent_alloc_retained(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, void *new_addr, size_t size, size_t pad, + size_t alignment, bool slab, szind_t szind, bool *zero, bool *commit) { + assert(size != 0); + assert(alignment != 0); + + malloc_mutex_lock(tsdn, &arena->extent_grow_mtx); + + extent_t *extent = extent_recycle(tsdn, arena, r_extent_hooks, + &arena->extents_retained, new_addr, size, pad, alignment, slab, + szind, zero, commit, true); + if (extent != NULL) { + malloc_mutex_unlock(tsdn, &arena->extent_grow_mtx); + if (config_prof) { + extent_gdump_add(tsdn, extent); + } + } else if (opt_retain && new_addr == NULL) { + extent = extent_grow_retained(tsdn, arena, r_extent_hooks, size, + pad, alignment, slab, szind, zero, commit); + /* extent_grow_retained() always releases extent_grow_mtx. */ + } else { + malloc_mutex_unlock(tsdn, &arena->extent_grow_mtx); + } + malloc_mutex_assert_not_owner(tsdn, &arena->extent_grow_mtx); + + return extent; +} + +static extent_t * +extent_alloc_wrapper_hard(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, void *new_addr, size_t size, size_t pad, + size_t alignment, bool slab, szind_t szind, bool *zero, bool *commit) { + size_t esize = size + pad; + extent_t *extent = extent_alloc(tsdn, arena); + if (extent == NULL) { + return NULL; + } + void *addr; + size_t palignment = ALIGNMENT_CEILING(alignment, PAGE); + if (*r_extent_hooks == &extent_hooks_default) { + /* Call directly to propagate tsdn. */ + addr = extent_alloc_default_impl(tsdn, arena, new_addr, esize, + palignment, zero, commit); + } else { + extent_hook_pre_reentrancy(tsdn, arena); + addr = (*r_extent_hooks)->alloc(*r_extent_hooks, new_addr, + esize, palignment, zero, commit, arena_ind_get(arena)); + extent_hook_post_reentrancy(tsdn); + } + if (addr == NULL) { + extent_dalloc(tsdn, arena, extent); + return NULL; + } + extent_init(extent, arena, addr, esize, slab, szind, + arena_extent_sn_next(arena), extent_state_active, *zero, *commit, + true, EXTENT_NOT_HEAD); + if (pad != 0) { + extent_addr_randomize(tsdn, extent, alignment); + } + if (extent_register(tsdn, extent)) { + extent_dalloc(tsdn, arena, extent); + return NULL; + } + + return extent; +} + +extent_t * +extent_alloc_wrapper(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, void *new_addr, size_t size, size_t pad, + size_t alignment, bool slab, szind_t szind, bool *zero, bool *commit) { + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, 0); + + extent_hooks_assure_initialized(arena, r_extent_hooks); + + extent_t *extent = extent_alloc_retained(tsdn, arena, r_extent_hooks, + new_addr, size, pad, alignment, slab, szind, zero, commit); + if (extent == NULL) { + if (opt_retain && new_addr != NULL) { + /* + * When retain is enabled and new_addr is set, we do not + * attempt extent_alloc_wrapper_hard which does mmap + * that is very unlikely to succeed (unless it happens + * to be at the end). + */ + return NULL; + } + extent = extent_alloc_wrapper_hard(tsdn, arena, r_extent_hooks, + new_addr, size, pad, alignment, slab, szind, zero, commit); + } + + assert(extent == NULL || extent_dumpable_get(extent)); + return extent; +} + +static bool +extent_can_coalesce(arena_t *arena, extents_t *extents, const extent_t *inner, + const extent_t *outer) { + assert(extent_arena_get(inner) == arena); + if (extent_arena_get(outer) != arena) { + return false; + } + + assert(extent_state_get(inner) == extent_state_active); + if (extent_state_get(outer) != extents->state) { + return false; + } + + if (extent_committed_get(inner) != extent_committed_get(outer)) { + return false; + } + + return true; +} + +static bool +extent_coalesce(tsdn_t *tsdn, arena_t *arena, extent_hooks_t **r_extent_hooks, + extents_t *extents, extent_t *inner, extent_t *outer, bool forward, + bool growing_retained) { + assert(extent_can_coalesce(arena, extents, inner, outer)); + + extent_activate_locked(tsdn, arena, extents, outer); + + malloc_mutex_unlock(tsdn, &extents->mtx); + bool err = extent_merge_impl(tsdn, arena, r_extent_hooks, + forward ? inner : outer, forward ? outer : inner, growing_retained); + malloc_mutex_lock(tsdn, &extents->mtx); + + if (err) { + extent_deactivate_locked(tsdn, arena, extents, outer); + } + + return err; +} + +static extent_t * +extent_try_coalesce_impl(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, rtree_ctx_t *rtree_ctx, extents_t *extents, + extent_t *extent, bool *coalesced, bool growing_retained, + bool inactive_only) { + /* + * We avoid checking / locking inactive neighbors for large size + * classes, since they are eagerly coalesced on deallocation which can + * cause lock contention. + */ + /* + * Continue attempting to coalesce until failure, to protect against + * races with other threads that are thwarted by this one. + */ + bool again; + do { + again = false; + + /* Try to coalesce forward. */ + extent_t *next = extent_lock_from_addr(tsdn, rtree_ctx, + extent_past_get(extent), inactive_only); + if (next != NULL) { + /* + * extents->mtx only protects against races for + * like-state extents, so call extent_can_coalesce() + * before releasing next's pool lock. + */ + bool can_coalesce = extent_can_coalesce(arena, extents, + extent, next); + + extent_unlock(tsdn, next); + + if (can_coalesce && !extent_coalesce(tsdn, arena, + r_extent_hooks, extents, extent, next, true, + growing_retained)) { + if (extents->delay_coalesce) { + /* Do minimal coalescing. */ + *coalesced = true; + return extent; + } + again = true; + } + } + + /* Try to coalesce backward. */ + extent_t *prev = extent_lock_from_addr(tsdn, rtree_ctx, + extent_before_get(extent), inactive_only); + if (prev != NULL) { + bool can_coalesce = extent_can_coalesce(arena, extents, + extent, prev); + extent_unlock(tsdn, prev); + + if (can_coalesce && !extent_coalesce(tsdn, arena, + r_extent_hooks, extents, extent, prev, false, + growing_retained)) { + extent = prev; + if (extents->delay_coalesce) { + /* Do minimal coalescing. */ + *coalesced = true; + return extent; + } + again = true; + } + } + } while (again); + + if (extents->delay_coalesce) { + *coalesced = false; + } + return extent; +} + +static extent_t * +extent_try_coalesce(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, rtree_ctx_t *rtree_ctx, extents_t *extents, + extent_t *extent, bool *coalesced, bool growing_retained) { + return extent_try_coalesce_impl(tsdn, arena, r_extent_hooks, rtree_ctx, + extents, extent, coalesced, growing_retained, false); +} + +static extent_t * +extent_try_coalesce_large(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, rtree_ctx_t *rtree_ctx, extents_t *extents, + extent_t *extent, bool *coalesced, bool growing_retained) { + return extent_try_coalesce_impl(tsdn, arena, r_extent_hooks, rtree_ctx, + extents, extent, coalesced, growing_retained, true); +} + +/* + * Does the metadata management portions of putting an unused extent into the + * given extents_t (coalesces, deregisters slab interiors, the heap operations). + */ +static void +extent_record(tsdn_t *tsdn, arena_t *arena, extent_hooks_t **r_extent_hooks, + extents_t *extents, extent_t *extent, bool growing_retained) { + rtree_ctx_t rtree_ctx_fallback; + rtree_ctx_t *rtree_ctx = tsdn_rtree_ctx(tsdn, &rtree_ctx_fallback); + + assert((extents_state_get(extents) != extent_state_dirty && + extents_state_get(extents) != extent_state_muzzy) || + !extent_zeroed_get(extent)); + + malloc_mutex_lock(tsdn, &extents->mtx); + extent_hooks_assure_initialized(arena, r_extent_hooks); + + extent_szind_set(extent, SC_NSIZES); + if (extent_slab_get(extent)) { + extent_interior_deregister(tsdn, rtree_ctx, extent); + extent_slab_set(extent, false); + } + + assert(rtree_extent_read(tsdn, &extents_rtree, rtree_ctx, + (uintptr_t)extent_base_get(extent), true) == extent); + + if (!extents->delay_coalesce) { + extent = extent_try_coalesce(tsdn, arena, r_extent_hooks, + rtree_ctx, extents, extent, NULL, growing_retained); + } else if (extent_size_get(extent) >= SC_LARGE_MINCLASS) { + assert(extents == &arena->extents_dirty); + /* Always coalesce large extents eagerly. */ + bool coalesced; + do { + assert(extent_state_get(extent) == extent_state_active); + extent = extent_try_coalesce_large(tsdn, arena, + r_extent_hooks, rtree_ctx, extents, extent, + &coalesced, growing_retained); + } while (coalesced); + if (extent_size_get(extent) >= oversize_threshold) { + /* Shortcut to purge the oversize extent eagerly. */ + malloc_mutex_unlock(tsdn, &extents->mtx); + arena_decay_extent(tsdn, arena, r_extent_hooks, extent); + return; + } + } + extent_deactivate_locked(tsdn, arena, extents, extent); + + malloc_mutex_unlock(tsdn, &extents->mtx); +} + +void +extent_dalloc_gap(tsdn_t *tsdn, arena_t *arena, extent_t *extent) { + extent_hooks_t *extent_hooks = EXTENT_HOOKS_INITIALIZER; + + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, 0); + + if (extent_register(tsdn, extent)) { + extent_dalloc(tsdn, arena, extent); + return; + } + extent_dalloc_wrapper(tsdn, arena, &extent_hooks, extent); +} + +static bool +extent_may_dalloc(void) { + /* With retain enabled, the default dalloc always fails. */ + return !opt_retain; +} + +static bool +extent_dalloc_default_impl(void *addr, size_t size) { + if (!have_dss || !extent_in_dss(addr)) { + return extent_dalloc_mmap(addr, size); + } + return true; +} + +static bool +extent_dalloc_default(extent_hooks_t *extent_hooks, void *addr, size_t size, + bool committed, unsigned arena_ind) { + return extent_dalloc_default_impl(addr, size); +} + +static bool +extent_dalloc_wrapper_try(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent) { + bool err; + + assert(extent_base_get(extent) != NULL); + assert(extent_size_get(extent) != 0); + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, 0); + + extent_addr_set(extent, extent_base_get(extent)); + + extent_hooks_assure_initialized(arena, r_extent_hooks); + /* Try to deallocate. */ + if (*r_extent_hooks == &extent_hooks_default) { + /* Call directly to propagate tsdn. */ + err = extent_dalloc_default_impl(extent_base_get(extent), + extent_size_get(extent)); + } else { + extent_hook_pre_reentrancy(tsdn, arena); + err = ((*r_extent_hooks)->dalloc == NULL || + (*r_extent_hooks)->dalloc(*r_extent_hooks, + extent_base_get(extent), extent_size_get(extent), + extent_committed_get(extent), arena_ind_get(arena))); + extent_hook_post_reentrancy(tsdn); + } + + if (!err) { + extent_dalloc(tsdn, arena, extent); + } + + return err; +} + +void +extent_dalloc_wrapper(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent) { + assert(extent_dumpable_get(extent)); + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, 0); + + /* Avoid calling the default extent_dalloc unless have to. */ + if (*r_extent_hooks != &extent_hooks_default || extent_may_dalloc()) { + /* + * Deregister first to avoid a race with other allocating + * threads, and reregister if deallocation fails. + */ + extent_deregister(tsdn, extent); + if (!extent_dalloc_wrapper_try(tsdn, arena, r_extent_hooks, + extent)) { + return; + } + extent_reregister(tsdn, extent); + } + + if (*r_extent_hooks != &extent_hooks_default) { + extent_hook_pre_reentrancy(tsdn, arena); + } + /* Try to decommit; purge if that fails. */ + bool zeroed; + if (!extent_committed_get(extent)) { + zeroed = true; + } else if (!extent_decommit_wrapper(tsdn, arena, r_extent_hooks, extent, + 0, extent_size_get(extent))) { + zeroed = true; + } else if ((*r_extent_hooks)->purge_forced != NULL && + !(*r_extent_hooks)->purge_forced(*r_extent_hooks, + extent_base_get(extent), extent_size_get(extent), 0, + extent_size_get(extent), arena_ind_get(arena))) { + zeroed = true; + } else if (extent_state_get(extent) == extent_state_muzzy || + ((*r_extent_hooks)->purge_lazy != NULL && + !(*r_extent_hooks)->purge_lazy(*r_extent_hooks, + extent_base_get(extent), extent_size_get(extent), 0, + extent_size_get(extent), arena_ind_get(arena)))) { + zeroed = false; + } else { + zeroed = false; + } + if (*r_extent_hooks != &extent_hooks_default) { + extent_hook_post_reentrancy(tsdn); + } + extent_zeroed_set(extent, zeroed); + + if (config_prof) { + extent_gdump_sub(tsdn, extent); + } + + extent_record(tsdn, arena, r_extent_hooks, &arena->extents_retained, + extent, false); +} + +static void +extent_destroy_default_impl(void *addr, size_t size) { + if (!have_dss || !extent_in_dss(addr)) { + pages_unmap(addr, size); + } +} + +static void +extent_destroy_default(extent_hooks_t *extent_hooks, void *addr, size_t size, + bool committed, unsigned arena_ind) { + extent_destroy_default_impl(addr, size); +} + +void +extent_destroy_wrapper(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent) { + assert(extent_base_get(extent) != NULL); + assert(extent_size_get(extent) != 0); + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, 0); + + /* Deregister first to avoid a race with other allocating threads. */ + extent_deregister(tsdn, extent); + + extent_addr_set(extent, extent_base_get(extent)); + + extent_hooks_assure_initialized(arena, r_extent_hooks); + /* Try to destroy; silently fail otherwise. */ + if (*r_extent_hooks == &extent_hooks_default) { + /* Call directly to propagate tsdn. */ + extent_destroy_default_impl(extent_base_get(extent), + extent_size_get(extent)); + } else if ((*r_extent_hooks)->destroy != NULL) { + extent_hook_pre_reentrancy(tsdn, arena); + (*r_extent_hooks)->destroy(*r_extent_hooks, + extent_base_get(extent), extent_size_get(extent), + extent_committed_get(extent), arena_ind_get(arena)); + extent_hook_post_reentrancy(tsdn); + } + + extent_dalloc(tsdn, arena, extent); +} + +static bool +extent_commit_default(extent_hooks_t *extent_hooks, void *addr, size_t size, + size_t offset, size_t length, unsigned arena_ind) { + return pages_commit((void *)((uintptr_t)addr + (uintptr_t)offset), + length); +} + +static bool +extent_commit_impl(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent, size_t offset, + size_t length, bool growing_retained) { + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, growing_retained ? 1 : 0); + + extent_hooks_assure_initialized(arena, r_extent_hooks); + if (*r_extent_hooks != &extent_hooks_default) { + extent_hook_pre_reentrancy(tsdn, arena); + } + bool err = ((*r_extent_hooks)->commit == NULL || + (*r_extent_hooks)->commit(*r_extent_hooks, extent_base_get(extent), + extent_size_get(extent), offset, length, arena_ind_get(arena))); + if (*r_extent_hooks != &extent_hooks_default) { + extent_hook_post_reentrancy(tsdn); + } + extent_committed_set(extent, extent_committed_get(extent) || !err); + return err; +} + +bool +extent_commit_wrapper(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent, size_t offset, + size_t length) { + return extent_commit_impl(tsdn, arena, r_extent_hooks, extent, offset, + length, false); +} + +static bool +extent_decommit_default(extent_hooks_t *extent_hooks, void *addr, size_t size, + size_t offset, size_t length, unsigned arena_ind) { + return pages_decommit((void *)((uintptr_t)addr + (uintptr_t)offset), + length); +} + +bool +extent_decommit_wrapper(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent, size_t offset, + size_t length) { + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, 0); + + extent_hooks_assure_initialized(arena, r_extent_hooks); + + if (*r_extent_hooks != &extent_hooks_default) { + extent_hook_pre_reentrancy(tsdn, arena); + } + bool err = ((*r_extent_hooks)->decommit == NULL || + (*r_extent_hooks)->decommit(*r_extent_hooks, + extent_base_get(extent), extent_size_get(extent), offset, length, + arena_ind_get(arena))); + if (*r_extent_hooks != &extent_hooks_default) { + extent_hook_post_reentrancy(tsdn); + } + extent_committed_set(extent, extent_committed_get(extent) && err); + return err; +} + +#ifdef PAGES_CAN_PURGE_LAZY +static bool +extent_purge_lazy_default(extent_hooks_t *extent_hooks, void *addr, size_t size, + size_t offset, size_t length, unsigned arena_ind) { + assert(addr != NULL); + assert((offset & PAGE_MASK) == 0); + assert(length != 0); + assert((length & PAGE_MASK) == 0); + + return pages_purge_lazy((void *)((uintptr_t)addr + (uintptr_t)offset), + length); +} +#endif + +static bool +extent_purge_lazy_impl(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent, size_t offset, + size_t length, bool growing_retained) { + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, growing_retained ? 1 : 0); + + extent_hooks_assure_initialized(arena, r_extent_hooks); + + if ((*r_extent_hooks)->purge_lazy == NULL) { + return true; + } + if (*r_extent_hooks != &extent_hooks_default) { + extent_hook_pre_reentrancy(tsdn, arena); + } + bool err = (*r_extent_hooks)->purge_lazy(*r_extent_hooks, + extent_base_get(extent), extent_size_get(extent), offset, length, + arena_ind_get(arena)); + if (*r_extent_hooks != &extent_hooks_default) { + extent_hook_post_reentrancy(tsdn); + } + + return err; +} + +bool +extent_purge_lazy_wrapper(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent, size_t offset, + size_t length) { + return extent_purge_lazy_impl(tsdn, arena, r_extent_hooks, extent, + offset, length, false); +} + +#ifdef PAGES_CAN_PURGE_FORCED +static bool +extent_purge_forced_default(extent_hooks_t *extent_hooks, void *addr, + size_t size, size_t offset, size_t length, unsigned arena_ind) { + assert(addr != NULL); + assert((offset & PAGE_MASK) == 0); + assert(length != 0); + assert((length & PAGE_MASK) == 0); + + return pages_purge_forced((void *)((uintptr_t)addr + + (uintptr_t)offset), length); +} +#endif + +static bool +extent_purge_forced_impl(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent, size_t offset, + size_t length, bool growing_retained) { + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, growing_retained ? 1 : 0); + + extent_hooks_assure_initialized(arena, r_extent_hooks); + + if ((*r_extent_hooks)->purge_forced == NULL) { + return true; + } + if (*r_extent_hooks != &extent_hooks_default) { + extent_hook_pre_reentrancy(tsdn, arena); + } + bool err = (*r_extent_hooks)->purge_forced(*r_extent_hooks, + extent_base_get(extent), extent_size_get(extent), offset, length, + arena_ind_get(arena)); + if (*r_extent_hooks != &extent_hooks_default) { + extent_hook_post_reentrancy(tsdn); + } + return err; +} + +bool +extent_purge_forced_wrapper(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent, size_t offset, + size_t length) { + return extent_purge_forced_impl(tsdn, arena, r_extent_hooks, extent, + offset, length, false); +} + +static bool +extent_split_default(extent_hooks_t *extent_hooks, void *addr, size_t size, + size_t size_a, size_t size_b, bool committed, unsigned arena_ind) { + if (!maps_coalesce) { + /* + * Without retain, only whole regions can be purged (required by + * MEM_RELEASE on Windows) -- therefore disallow splitting. See + * comments in extent_head_no_merge(). + */ + return !opt_retain; + } + + return false; +} + +/* + * Accepts the extent to split, and the characteristics of each side of the + * split. The 'a' parameters go with the 'lead' of the resulting pair of + * extents (the lower addressed portion of the split), and the 'b' parameters go + * with the trail (the higher addressed portion). This makes 'extent' the lead, + * and returns the trail (except in case of error). + */ +static extent_t * +extent_split_impl(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent, size_t size_a, + szind_t szind_a, bool slab_a, size_t size_b, szind_t szind_b, bool slab_b, + bool growing_retained) { + assert(extent_size_get(extent) == size_a + size_b); + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, growing_retained ? 1 : 0); + + extent_hooks_assure_initialized(arena, r_extent_hooks); + + if ((*r_extent_hooks)->split == NULL) { + return NULL; + } + + extent_t *trail = extent_alloc(tsdn, arena); + if (trail == NULL) { + goto label_error_a; + } + + extent_init(trail, arena, (void *)((uintptr_t)extent_base_get(extent) + + size_a), size_b, slab_b, szind_b, extent_sn_get(extent), + extent_state_get(extent), extent_zeroed_get(extent), + extent_committed_get(extent), extent_dumpable_get(extent), + EXTENT_NOT_HEAD); + + rtree_ctx_t rtree_ctx_fallback; + rtree_ctx_t *rtree_ctx = tsdn_rtree_ctx(tsdn, &rtree_ctx_fallback); + rtree_leaf_elm_t *lead_elm_a, *lead_elm_b; + { + extent_t lead; + + extent_init(&lead, arena, extent_addr_get(extent), size_a, + slab_a, szind_a, extent_sn_get(extent), + extent_state_get(extent), extent_zeroed_get(extent), + extent_committed_get(extent), extent_dumpable_get(extent), + EXTENT_NOT_HEAD); + + extent_rtree_leaf_elms_lookup(tsdn, rtree_ctx, &lead, false, + true, &lead_elm_a, &lead_elm_b); + } + rtree_leaf_elm_t *trail_elm_a, *trail_elm_b; + extent_rtree_leaf_elms_lookup(tsdn, rtree_ctx, trail, false, true, + &trail_elm_a, &trail_elm_b); + + if (lead_elm_a == NULL || lead_elm_b == NULL || trail_elm_a == NULL + || trail_elm_b == NULL) { + goto label_error_b; + } + + extent_lock2(tsdn, extent, trail); + + if (*r_extent_hooks != &extent_hooks_default) { + extent_hook_pre_reentrancy(tsdn, arena); + } + bool err = (*r_extent_hooks)->split(*r_extent_hooks, extent_base_get(extent), + size_a + size_b, size_a, size_b, extent_committed_get(extent), + arena_ind_get(arena)); + if (*r_extent_hooks != &extent_hooks_default) { + extent_hook_post_reentrancy(tsdn); + } + if (err) { + goto label_error_c; + } + + extent_size_set(extent, size_a); + extent_szind_set(extent, szind_a); + + extent_rtree_write_acquired(tsdn, lead_elm_a, lead_elm_b, extent, + szind_a, slab_a); + extent_rtree_write_acquired(tsdn, trail_elm_a, trail_elm_b, trail, + szind_b, slab_b); + + extent_unlock2(tsdn, extent, trail); + + return trail; +label_error_c: + extent_unlock2(tsdn, extent, trail); +label_error_b: + extent_dalloc(tsdn, arena, trail); +label_error_a: + return NULL; +} + +extent_t * +extent_split_wrapper(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *extent, size_t size_a, + szind_t szind_a, bool slab_a, size_t size_b, szind_t szind_b, bool slab_b) { + return extent_split_impl(tsdn, arena, r_extent_hooks, extent, size_a, + szind_a, slab_a, size_b, szind_b, slab_b, false); +} + +static bool +extent_merge_default_impl(void *addr_a, void *addr_b) { + if (!maps_coalesce && !opt_retain) { + return true; + } + if (have_dss && !extent_dss_mergeable(addr_a, addr_b)) { + return true; + } + + return false; +} + +/* + * Returns true if the given extents can't be merged because of their head bit + * settings. Assumes the second extent has the higher address. + */ +static bool +extent_head_no_merge(extent_t *a, extent_t *b) { + assert(extent_base_get(a) < extent_base_get(b)); + /* + * When coalesce is not always allowed (Windows), only merge extents + * from the same VirtualAlloc region under opt.retain (in which case + * MEM_DECOMMIT is utilized for purging). + */ + if (maps_coalesce) { + return false; + } + if (!opt_retain) { + return true; + } + /* If b is a head extent, disallow the cross-region merge. */ + if (extent_is_head_get(b)) { + /* + * Additionally, sn should not overflow with retain; sanity + * check that different regions have unique sn. + */ + assert(extent_sn_comp(a, b) != 0); + return true; + } + assert(extent_sn_comp(a, b) == 0); + + return false; +} + +static bool +extent_merge_default(extent_hooks_t *extent_hooks, void *addr_a, size_t size_a, + void *addr_b, size_t size_b, bool committed, unsigned arena_ind) { + if (!maps_coalesce) { + tsdn_t *tsdn = tsdn_fetch(); + extent_t *a = iealloc(tsdn, addr_a); + extent_t *b = iealloc(tsdn, addr_b); + if (extent_head_no_merge(a, b)) { + return true; + } + } + return extent_merge_default_impl(addr_a, addr_b); +} + +static bool +extent_merge_impl(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *a, extent_t *b, + bool growing_retained) { + witness_assert_depth_to_rank(tsdn_witness_tsdp_get(tsdn), + WITNESS_RANK_CORE, growing_retained ? 1 : 0); + assert(extent_base_get(a) < extent_base_get(b)); + + extent_hooks_assure_initialized(arena, r_extent_hooks); + + if ((*r_extent_hooks)->merge == NULL || extent_head_no_merge(a, b)) { + return true; + } + + bool err; + if (*r_extent_hooks == &extent_hooks_default) { + /* Call directly to propagate tsdn. */ + err = extent_merge_default_impl(extent_base_get(a), + extent_base_get(b)); + } else { + extent_hook_pre_reentrancy(tsdn, arena); + err = (*r_extent_hooks)->merge(*r_extent_hooks, + extent_base_get(a), extent_size_get(a), extent_base_get(b), + extent_size_get(b), extent_committed_get(a), + arena_ind_get(arena)); + extent_hook_post_reentrancy(tsdn); + } + + if (err) { + return true; + } + + /* + * The rtree writes must happen while all the relevant elements are + * owned, so the following code uses decomposed helper functions rather + * than extent_{,de}register() to do things in the right order. + */ + rtree_ctx_t rtree_ctx_fallback; + rtree_ctx_t *rtree_ctx = tsdn_rtree_ctx(tsdn, &rtree_ctx_fallback); + rtree_leaf_elm_t *a_elm_a, *a_elm_b, *b_elm_a, *b_elm_b; + extent_rtree_leaf_elms_lookup(tsdn, rtree_ctx, a, true, false, &a_elm_a, + &a_elm_b); + extent_rtree_leaf_elms_lookup(tsdn, rtree_ctx, b, true, false, &b_elm_a, + &b_elm_b); + + extent_lock2(tsdn, a, b); + + if (a_elm_b != NULL) { + rtree_leaf_elm_write(tsdn, &extents_rtree, a_elm_b, NULL, + SC_NSIZES, false); + } + if (b_elm_b != NULL) { + rtree_leaf_elm_write(tsdn, &extents_rtree, b_elm_a, NULL, + SC_NSIZES, false); + } else { + b_elm_b = b_elm_a; + } + + extent_size_set(a, extent_size_get(a) + extent_size_get(b)); + extent_szind_set(a, SC_NSIZES); + extent_sn_set(a, (extent_sn_get(a) < extent_sn_get(b)) ? + extent_sn_get(a) : extent_sn_get(b)); + extent_zeroed_set(a, extent_zeroed_get(a) && extent_zeroed_get(b)); + + extent_rtree_write_acquired(tsdn, a_elm_a, b_elm_b, a, SC_NSIZES, + false); + + extent_unlock2(tsdn, a, b); + + extent_dalloc(tsdn, extent_arena_get(b), b); + + return false; +} + +bool +extent_merge_wrapper(tsdn_t *tsdn, arena_t *arena, + extent_hooks_t **r_extent_hooks, extent_t *a, extent_t *b) { + return extent_merge_impl(tsdn, arena, r_extent_hooks, a, b, false); +} + +bool +extent_boot(void) { + if (rtree_new(&extents_rtree, true)) { + return true; + } + + if (mutex_pool_init(&extent_mutex_pool, "extent_mutex_pool", + WITNESS_RANK_EXTENT_POOL)) { + return true; + } + + if (have_dss) { + extent_dss_boot(); + } + + return false; +} + +void +extent_util_stats_get(tsdn_t *tsdn, const void *ptr, + size_t *nfree, size_t *nregs, size_t *size) { + assert(ptr != NULL && nfree != NULL && nregs != NULL && size != NULL); + + const extent_t *extent = iealloc(tsdn, ptr); + if (unlikely(extent == NULL)) { + *nfree = *nregs = *size = 0; + return; + } + + *size = extent_size_get(extent); + if (!extent_slab_get(extent)) { + *nfree = 0; + *nregs = 1; + } else { + *nfree = extent_nfree_get(extent); + *nregs = bin_infos[extent_szind_get(extent)].nregs; + assert(*nfree <= *nregs); + assert(*nfree * extent_usize_get(extent) <= *size); + } +} + +void +extent_util_stats_verbose_get(tsdn_t *tsdn, const void *ptr, + size_t *nfree, size_t *nregs, size_t *size, + size_t *bin_nfree, size_t *bin_nregs, void **slabcur_addr) { + assert(ptr != NULL && nfree != NULL && nregs != NULL && size != NULL + && bin_nfree != NULL && bin_nregs != NULL && slabcur_addr != NULL); + + const extent_t *extent = iealloc(tsdn, ptr); + if (unlikely(extent == NULL)) { + *nfree = *nregs = *size = *bin_nfree = *bin_nregs = 0; + *slabcur_addr = NULL; + return; + } + + *size = extent_size_get(extent); + if (!extent_slab_get(extent)) { + *nfree = *bin_nfree = *bin_nregs = 0; + *nregs = 1; + *slabcur_addr = NULL; + return; + } + + *nfree = extent_nfree_get(extent); + const szind_t szind = extent_szind_get(extent); + *nregs = bin_infos[szind].nregs; + assert(*nfree <= *nregs); + assert(*nfree * extent_usize_get(extent) <= *size); + + const arena_t *arena = extent_arena_get(extent); + assert(arena != NULL); + const unsigned binshard = extent_binshard_get(extent); + bin_t *bin = &arena->bins[szind].bin_shards[binshard]; + + malloc_mutex_lock(tsdn, &bin->lock); + if (config_stats) { + *bin_nregs = *nregs * bin->stats.curslabs; + assert(*bin_nregs >= bin->stats.curregs); + *bin_nfree = *bin_nregs - bin->stats.curregs; + } else { + *bin_nfree = *bin_nregs = 0; + } + *slabcur_addr = extent_addr_get(bin->slabcur); + assert(*slabcur_addr != NULL); + malloc_mutex_unlock(tsdn, &bin->lock); +} diff --git a/deps/jemalloc/src/extent_dss.c b/deps/jemalloc/src/extent_dss.c new file mode 100644 index 000000000..858178911 --- /dev/null +++ b/deps/jemalloc/src/extent_dss.c @@ -0,0 +1,271 @@ +#define JEMALLOC_EXTENT_DSS_C_ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/assert.h" +#include "jemalloc/internal/extent_dss.h" +#include "jemalloc/internal/spin.h" + +/******************************************************************************/ +/* Data. */ + +const char *opt_dss = DSS_DEFAULT; + +const char *dss_prec_names[] = { + "disabled", + "primary", + "secondary", + "N/A" +}; + +/* + * Current dss precedence default, used when creating new arenas. NB: This is + * stored as unsigned rather than dss_prec_t because in principle there's no + * guarantee that sizeof(dss_prec_t) is the same as sizeof(unsigned), and we use + * atomic operations to synchronize the setting. + */ +static atomic_u_t dss_prec_default = ATOMIC_INIT( + (unsigned)DSS_PREC_DEFAULT); + +/* Base address of the DSS. */ +static void *dss_base; +/* Atomic boolean indicating whether a thread is currently extending DSS. */ +static atomic_b_t dss_extending; +/* Atomic boolean indicating whether the DSS is exhausted. */ +static atomic_b_t dss_exhausted; +/* Atomic current upper limit on DSS addresses. */ +static atomic_p_t dss_max; + +/******************************************************************************/ + +static void * +extent_dss_sbrk(intptr_t increment) { +#ifdef JEMALLOC_DSS + return sbrk(increment); +#else + not_implemented(); + return NULL; +#endif +} + +dss_prec_t +extent_dss_prec_get(void) { + dss_prec_t ret; + + if (!have_dss) { + return dss_prec_disabled; + } + ret = (dss_prec_t)atomic_load_u(&dss_prec_default, ATOMIC_ACQUIRE); + return ret; +} + +bool +extent_dss_prec_set(dss_prec_t dss_prec) { + if (!have_dss) { + return (dss_prec != dss_prec_disabled); + } + atomic_store_u(&dss_prec_default, (unsigned)dss_prec, ATOMIC_RELEASE); + return false; +} + +static void +extent_dss_extending_start(void) { + spin_t spinner = SPIN_INITIALIZER; + while (true) { + bool expected = false; + if (atomic_compare_exchange_weak_b(&dss_extending, &expected, + true, ATOMIC_ACQ_REL, ATOMIC_RELAXED)) { + break; + } + spin_adaptive(&spinner); + } +} + +static void +extent_dss_extending_finish(void) { + assert(atomic_load_b(&dss_extending, ATOMIC_RELAXED)); + + atomic_store_b(&dss_extending, false, ATOMIC_RELEASE); +} + +static void * +extent_dss_max_update(void *new_addr) { + /* + * Get the current end of the DSS as max_cur and assure that dss_max is + * up to date. + */ + void *max_cur = extent_dss_sbrk(0); + if (max_cur == (void *)-1) { + return NULL; + } + atomic_store_p(&dss_max, max_cur, ATOMIC_RELEASE); + /* Fixed new_addr can only be supported if it is at the edge of DSS. */ + if (new_addr != NULL && max_cur != new_addr) { + return NULL; + } + return max_cur; +} + +void * +extent_alloc_dss(tsdn_t *tsdn, arena_t *arena, void *new_addr, size_t size, + size_t alignment, bool *zero, bool *commit) { + extent_t *gap; + + cassert(have_dss); + assert(size > 0); + assert(alignment == ALIGNMENT_CEILING(alignment, PAGE)); + + /* + * sbrk() uses a signed increment argument, so take care not to + * interpret a large allocation request as a negative increment. + */ + if ((intptr_t)size < 0) { + return NULL; + } + + gap = extent_alloc(tsdn, arena); + if (gap == NULL) { + return NULL; + } + + extent_dss_extending_start(); + if (!atomic_load_b(&dss_exhausted, ATOMIC_ACQUIRE)) { + /* + * The loop is necessary to recover from races with other + * threads that are using the DSS for something other than + * malloc. + */ + while (true) { + void *max_cur = extent_dss_max_update(new_addr); + if (max_cur == NULL) { + goto label_oom; + } + + /* + * Compute how much page-aligned gap space (if any) is + * necessary to satisfy alignment. This space can be + * recycled for later use. + */ + void *gap_addr_page = (void *)(PAGE_CEILING( + (uintptr_t)max_cur)); + void *ret = (void *)ALIGNMENT_CEILING( + (uintptr_t)gap_addr_page, alignment); + size_t gap_size_page = (uintptr_t)ret - + (uintptr_t)gap_addr_page; + if (gap_size_page != 0) { + extent_init(gap, arena, gap_addr_page, + gap_size_page, false, SC_NSIZES, + arena_extent_sn_next(arena), + extent_state_active, false, true, true, + EXTENT_NOT_HEAD); + } + /* + * Compute the address just past the end of the desired + * allocation space. + */ + void *dss_next = (void *)((uintptr_t)ret + size); + if ((uintptr_t)ret < (uintptr_t)max_cur || + (uintptr_t)dss_next < (uintptr_t)max_cur) { + goto label_oom; /* Wrap-around. */ + } + /* Compute the increment, including subpage bytes. */ + void *gap_addr_subpage = max_cur; + size_t gap_size_subpage = (uintptr_t)ret - + (uintptr_t)gap_addr_subpage; + intptr_t incr = gap_size_subpage + size; + + assert((uintptr_t)max_cur + incr == (uintptr_t)ret + + size); + + /* Try to allocate. */ + void *dss_prev = extent_dss_sbrk(incr); + if (dss_prev == max_cur) { + /* Success. */ + atomic_store_p(&dss_max, dss_next, + ATOMIC_RELEASE); + extent_dss_extending_finish(); + + if (gap_size_page != 0) { + extent_dalloc_gap(tsdn, arena, gap); + } else { + extent_dalloc(tsdn, arena, gap); + } + if (!*commit) { + *commit = pages_decommit(ret, size); + } + if (*zero && *commit) { + extent_hooks_t *extent_hooks = + EXTENT_HOOKS_INITIALIZER; + extent_t extent; + + extent_init(&extent, arena, ret, size, + size, false, SC_NSIZES, + extent_state_active, false, true, + true, EXTENT_NOT_HEAD); + if (extent_purge_forced_wrapper(tsdn, + arena, &extent_hooks, &extent, 0, + size)) { + memset(ret, 0, size); + } + } + return ret; + } + /* + * Failure, whether due to OOM or a race with a raw + * sbrk() call from outside the allocator. + */ + if (dss_prev == (void *)-1) { + /* OOM. */ + atomic_store_b(&dss_exhausted, true, + ATOMIC_RELEASE); + goto label_oom; + } + } + } +label_oom: + extent_dss_extending_finish(); + extent_dalloc(tsdn, arena, gap); + return NULL; +} + +static bool +extent_in_dss_helper(void *addr, void *max) { + return ((uintptr_t)addr >= (uintptr_t)dss_base && (uintptr_t)addr < + (uintptr_t)max); +} + +bool +extent_in_dss(void *addr) { + cassert(have_dss); + + return extent_in_dss_helper(addr, atomic_load_p(&dss_max, + ATOMIC_ACQUIRE)); +} + +bool +extent_dss_mergeable(void *addr_a, void *addr_b) { + void *max; + + cassert(have_dss); + + if ((uintptr_t)addr_a < (uintptr_t)dss_base && (uintptr_t)addr_b < + (uintptr_t)dss_base) { + return true; + } + + max = atomic_load_p(&dss_max, ATOMIC_ACQUIRE); + return (extent_in_dss_helper(addr_a, max) == + extent_in_dss_helper(addr_b, max)); +} + +void +extent_dss_boot(void) { + cassert(have_dss); + + dss_base = extent_dss_sbrk(0); + atomic_store_b(&dss_extending, false, ATOMIC_RELAXED); + atomic_store_b(&dss_exhausted, dss_base == (void *)-1, ATOMIC_RELAXED); + atomic_store_p(&dss_max, dss_base, ATOMIC_RELAXED); +} + +/******************************************************************************/ diff --git a/deps/jemalloc/src/extent_mmap.c b/deps/jemalloc/src/extent_mmap.c new file mode 100644 index 000000000..17fd1c8f9 --- /dev/null +++ b/deps/jemalloc/src/extent_mmap.c @@ -0,0 +1,42 @@ +#define JEMALLOC_EXTENT_MMAP_C_ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/assert.h" +#include "jemalloc/internal/extent_mmap.h" + +/******************************************************************************/ +/* Data. */ + +bool opt_retain = +#ifdef JEMALLOC_RETAIN + true +#else + false +#endif + ; + +/******************************************************************************/ + +void * +extent_alloc_mmap(void *new_addr, size_t size, size_t alignment, bool *zero, + bool *commit) { + assert(alignment == ALIGNMENT_CEILING(alignment, PAGE)); + void *ret = pages_map(new_addr, size, alignment, commit); + if (ret == NULL) { + return NULL; + } + assert(ret != NULL); + if (*commit) { + *zero = true; + } + return ret; +} + +bool +extent_dalloc_mmap(void *addr, size_t size) { + if (!opt_retain) { + pages_unmap(addr, size); + } + return opt_retain; +} diff --git a/deps/jemalloc/src/hash.c b/deps/jemalloc/src/hash.c new file mode 100644 index 000000000..7b2bdc2bd --- /dev/null +++ b/deps/jemalloc/src/hash.c @@ -0,0 +1,3 @@ +#define JEMALLOC_HASH_C_ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" diff --git a/deps/jemalloc/src/hook.c b/deps/jemalloc/src/hook.c new file mode 100644 index 000000000..9ac703cf9 --- /dev/null +++ b/deps/jemalloc/src/hook.c @@ -0,0 +1,195 @@ +#include "jemalloc/internal/jemalloc_preamble.h" + +#include "jemalloc/internal/hook.h" + +#include "jemalloc/internal/atomic.h" +#include "jemalloc/internal/mutex.h" +#include "jemalloc/internal/seq.h" + +typedef struct hooks_internal_s hooks_internal_t; +struct hooks_internal_s { + hooks_t hooks; + bool in_use; +}; + +seq_define(hooks_internal_t, hooks) + +static atomic_u_t nhooks = ATOMIC_INIT(0); +static seq_hooks_t hooks[HOOK_MAX]; +static malloc_mutex_t hooks_mu; + +bool +hook_boot() { + return malloc_mutex_init(&hooks_mu, "hooks", WITNESS_RANK_HOOK, + malloc_mutex_rank_exclusive); +} + +static void * +hook_install_locked(hooks_t *to_install) { + hooks_internal_t hooks_internal; + for (int i = 0; i < HOOK_MAX; i++) { + bool success = seq_try_load_hooks(&hooks_internal, &hooks[i]); + /* We hold mu; no concurrent access. */ + assert(success); + if (!hooks_internal.in_use) { + hooks_internal.hooks = *to_install; + hooks_internal.in_use = true; + seq_store_hooks(&hooks[i], &hooks_internal); + atomic_store_u(&nhooks, + atomic_load_u(&nhooks, ATOMIC_RELAXED) + 1, + ATOMIC_RELAXED); + return &hooks[i]; + } + } + return NULL; +} + +void * +hook_install(tsdn_t *tsdn, hooks_t *to_install) { + malloc_mutex_lock(tsdn, &hooks_mu); + void *ret = hook_install_locked(to_install); + if (ret != NULL) { + tsd_global_slow_inc(tsdn); + } + malloc_mutex_unlock(tsdn, &hooks_mu); + return ret; +} + +static void +hook_remove_locked(seq_hooks_t *to_remove) { + hooks_internal_t hooks_internal; + bool success = seq_try_load_hooks(&hooks_internal, to_remove); + /* We hold mu; no concurrent access. */ + assert(success); + /* Should only remove hooks that were added. */ + assert(hooks_internal.in_use); + hooks_internal.in_use = false; + seq_store_hooks(to_remove, &hooks_internal); + atomic_store_u(&nhooks, atomic_load_u(&nhooks, ATOMIC_RELAXED) - 1, + ATOMIC_RELAXED); +} + +void +hook_remove(tsdn_t *tsdn, void *opaque) { + if (config_debug) { + char *hooks_begin = (char *)&hooks[0]; + char *hooks_end = (char *)&hooks[HOOK_MAX]; + char *hook = (char *)opaque; + assert(hooks_begin <= hook && hook < hooks_end + && (hook - hooks_begin) % sizeof(seq_hooks_t) == 0); + } + malloc_mutex_lock(tsdn, &hooks_mu); + hook_remove_locked((seq_hooks_t *)opaque); + tsd_global_slow_dec(tsdn); + malloc_mutex_unlock(tsdn, &hooks_mu); +} + +#define FOR_EACH_HOOK_BEGIN(hooks_internal_ptr) \ +for (int for_each_hook_counter = 0; \ + for_each_hook_counter < HOOK_MAX; \ + for_each_hook_counter++) { \ + bool for_each_hook_success = seq_try_load_hooks( \ + (hooks_internal_ptr), &hooks[for_each_hook_counter]); \ + if (!for_each_hook_success) { \ + continue; \ + } \ + if (!(hooks_internal_ptr)->in_use) { \ + continue; \ + } +#define FOR_EACH_HOOK_END \ +} + +static bool * +hook_reentrantp() { + /* + * We prevent user reentrancy within hooks. This is basically just a + * thread-local bool that triggers an early-exit. + * + * We don't fold in_hook into reentrancy. There are two reasons for + * this: + * - Right now, we turn on reentrancy during things like extent hook + * execution. Allocating during extent hooks is not officially + * supported, but we don't want to break it for the time being. These + * sorts of allocations should probably still be hooked, though. + * - If a hook allocates, we may want it to be relatively fast (after + * all, it executes on every allocator operation). Turning on + * reentrancy is a fairly heavyweight mode (disabling tcache, + * redirecting to arena 0, etc.). It's possible we may one day want + * to turn on reentrant mode here, if it proves too difficult to keep + * this working. But that's fairly easy for us to see; OTOH, people + * not using hooks because they're too slow is easy for us to miss. + * + * The tricky part is + * that this code might get invoked even if we don't have access to tsd. + * This function mimics getting a pointer to thread-local data, except + * that it might secretly return a pointer to some global data if we + * know that the caller will take the early-exit path. + * If we return a bool that indicates that we are reentrant, then the + * caller will go down the early exit path, leaving the global + * untouched. + */ + static bool in_hook_global = true; + tsdn_t *tsdn = tsdn_fetch(); + tcache_t *tcache = tsdn_tcachep_get(tsdn); + if (tcache != NULL) { + return &tcache->in_hook; + } + return &in_hook_global; +} + +#define HOOK_PROLOGUE \ + if (likely(atomic_load_u(&nhooks, ATOMIC_RELAXED) == 0)) { \ + return; \ + } \ + bool *in_hook = hook_reentrantp(); \ + if (*in_hook) { \ + return; \ + } \ + *in_hook = true; + +#define HOOK_EPILOGUE \ + *in_hook = false; + +void +hook_invoke_alloc(hook_alloc_t type, void *result, uintptr_t result_raw, + uintptr_t args_raw[3]) { + HOOK_PROLOGUE + + hooks_internal_t hook; + FOR_EACH_HOOK_BEGIN(&hook) + hook_alloc h = hook.hooks.alloc_hook; + if (h != NULL) { + h(hook.hooks.extra, type, result, result_raw, args_raw); + } + FOR_EACH_HOOK_END + + HOOK_EPILOGUE +} + +void +hook_invoke_dalloc(hook_dalloc_t type, void *address, uintptr_t args_raw[3]) { + HOOK_PROLOGUE + hooks_internal_t hook; + FOR_EACH_HOOK_BEGIN(&hook) + hook_dalloc h = hook.hooks.dalloc_hook; + if (h != NULL) { + h(hook.hooks.extra, type, address, args_raw); + } + FOR_EACH_HOOK_END + HOOK_EPILOGUE +} + +void +hook_invoke_expand(hook_expand_t type, void *address, size_t old_usize, + size_t new_usize, uintptr_t result_raw, uintptr_t args_raw[4]) { + HOOK_PROLOGUE + hooks_internal_t hook; + FOR_EACH_HOOK_BEGIN(&hook) + hook_expand h = hook.hooks.expand_hook; + if (h != NULL) { + h(hook.hooks.extra, type, address, old_usize, new_usize, + result_raw, args_raw); + } + FOR_EACH_HOOK_END + HOOK_EPILOGUE +} diff --git a/deps/jemalloc/src/jemalloc.c b/deps/jemalloc/src/jemalloc.c new file mode 100644 index 000000000..ed13718d4 --- /dev/null +++ b/deps/jemalloc/src/jemalloc.c @@ -0,0 +1,3922 @@ +#define JEMALLOC_C_ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/assert.h" +#include "jemalloc/internal/atomic.h" +#include "jemalloc/internal/ctl.h" +#include "jemalloc/internal/extent_dss.h" +#include "jemalloc/internal/extent_mmap.h" +#include "jemalloc/internal/hook.h" +#include "jemalloc/internal/jemalloc_internal_types.h" +#include "jemalloc/internal/log.h" +#include "jemalloc/internal/malloc_io.h" +#include "jemalloc/internal/mutex.h" +#include "jemalloc/internal/rtree.h" +#include "jemalloc/internal/safety_check.h" +#include "jemalloc/internal/sc.h" +#include "jemalloc/internal/spin.h" +#include "jemalloc/internal/sz.h" +#include "jemalloc/internal/ticker.h" +#include "jemalloc/internal/util.h" + +/******************************************************************************/ +/* Data. */ + +/* Runtime configuration options. */ +const char *je_malloc_conf +#ifndef _WIN32 + JEMALLOC_ATTR(weak) +#endif + ; +bool opt_abort = +#ifdef JEMALLOC_DEBUG + true +#else + false +#endif + ; +bool opt_abort_conf = +#ifdef JEMALLOC_DEBUG + true +#else + false +#endif + ; +/* Intentionally default off, even with debug builds. */ +bool opt_confirm_conf = false; +const char *opt_junk = +#if (defined(JEMALLOC_DEBUG) && defined(JEMALLOC_FILL)) + "true" +#else + "false" +#endif + ; +bool opt_junk_alloc = +#if (defined(JEMALLOC_DEBUG) && defined(JEMALLOC_FILL)) + true +#else + false +#endif + ; +bool opt_junk_free = +#if (defined(JEMALLOC_DEBUG) && defined(JEMALLOC_FILL)) + true +#else + false +#endif + ; + +bool opt_utrace = false; +bool opt_xmalloc = false; +bool opt_zero = false; +unsigned opt_narenas = 0; + +unsigned ncpus; + +/* Protects arenas initialization. */ +malloc_mutex_t arenas_lock; +/* + * Arenas that are used to service external requests. Not all elements of the + * arenas array are necessarily used; arenas are created lazily as needed. + * + * arenas[0..narenas_auto) are used for automatic multiplexing of threads and + * arenas. arenas[narenas_auto..narenas_total) are only used if the application + * takes some action to create them and allocate from them. + * + * Points to an arena_t. + */ +JEMALLOC_ALIGNED(CACHELINE) +atomic_p_t arenas[MALLOCX_ARENA_LIMIT]; +static atomic_u_t narenas_total; /* Use narenas_total_*(). */ +/* Below three are read-only after initialization. */ +static arena_t *a0; /* arenas[0]. */ +unsigned narenas_auto; +unsigned manual_arena_base; + +typedef enum { + malloc_init_uninitialized = 3, + malloc_init_a0_initialized = 2, + malloc_init_recursible = 1, + malloc_init_initialized = 0 /* Common case --> jnz. */ +} malloc_init_t; +static malloc_init_t malloc_init_state = malloc_init_uninitialized; + +/* False should be the common case. Set to true to trigger initialization. */ +bool malloc_slow = true; + +/* When malloc_slow is true, set the corresponding bits for sanity check. */ +enum { + flag_opt_junk_alloc = (1U), + flag_opt_junk_free = (1U << 1), + flag_opt_zero = (1U << 2), + flag_opt_utrace = (1U << 3), + flag_opt_xmalloc = (1U << 4) +}; +static uint8_t malloc_slow_flags; + +#ifdef JEMALLOC_THREADED_INIT +/* Used to let the initializing thread recursively allocate. */ +# define NO_INITIALIZER ((unsigned long)0) +# define INITIALIZER pthread_self() +# define IS_INITIALIZER (malloc_initializer == pthread_self()) +static pthread_t malloc_initializer = NO_INITIALIZER; +#else +# define NO_INITIALIZER false +# define INITIALIZER true +# define IS_INITIALIZER malloc_initializer +static bool malloc_initializer = NO_INITIALIZER; +#endif + +/* Used to avoid initialization races. */ +#ifdef _WIN32 +#if _WIN32_WINNT >= 0x0600 +static malloc_mutex_t init_lock = SRWLOCK_INIT; +#else +static malloc_mutex_t init_lock; +static bool init_lock_initialized = false; + +JEMALLOC_ATTR(constructor) +static void WINAPI +_init_init_lock(void) { + /* + * If another constructor in the same binary is using mallctl to e.g. + * set up extent hooks, it may end up running before this one, and + * malloc_init_hard will crash trying to lock the uninitialized lock. So + * we force an initialization of the lock in malloc_init_hard as well. + * We don't try to care about atomicity of the accessed to the + * init_lock_initialized boolean, since it really only matters early in + * the process creation, before any separate thread normally starts + * doing anything. + */ + if (!init_lock_initialized) { + malloc_mutex_init(&init_lock, "init", WITNESS_RANK_INIT, + malloc_mutex_rank_exclusive); + } + init_lock_initialized = true; +} + +#ifdef _MSC_VER +# pragma section(".CRT$XCU", read) +JEMALLOC_SECTION(".CRT$XCU") JEMALLOC_ATTR(used) +static const void (WINAPI *init_init_lock)(void) = _init_init_lock; +#endif +#endif +#else +static malloc_mutex_t init_lock = MALLOC_MUTEX_INITIALIZER; +#endif + +typedef struct { + void *p; /* Input pointer (as in realloc(p, s)). */ + size_t s; /* Request size. */ + void *r; /* Result pointer. */ +} malloc_utrace_t; + +#ifdef JEMALLOC_UTRACE +# define UTRACE(a, b, c) do { \ + if (unlikely(opt_utrace)) { \ + int utrace_serrno = errno; \ + malloc_utrace_t ut; \ + ut.p = (a); \ + ut.s = (b); \ + ut.r = (c); \ + utrace(&ut, sizeof(ut)); \ + errno = utrace_serrno; \ + } \ +} while (0) +#else +# define UTRACE(a, b, c) +#endif + +/* Whether encountered any invalid config options. */ +static bool had_conf_error = false; + +/******************************************************************************/ +/* + * Function prototypes for static functions that are referenced prior to + * definition. + */ + +static bool malloc_init_hard_a0(void); +static bool malloc_init_hard(void); + +/******************************************************************************/ +/* + * Begin miscellaneous support functions. + */ + +bool +malloc_initialized(void) { + return (malloc_init_state == malloc_init_initialized); +} + +JEMALLOC_ALWAYS_INLINE bool +malloc_init_a0(void) { + if (unlikely(malloc_init_state == malloc_init_uninitialized)) { + return malloc_init_hard_a0(); + } + return false; +} + +JEMALLOC_ALWAYS_INLINE bool +malloc_init(void) { + if (unlikely(!malloc_initialized()) && malloc_init_hard()) { + return true; + } + return false; +} + +/* + * The a0*() functions are used instead of i{d,}alloc() in situations that + * cannot tolerate TLS variable access. + */ + +static void * +a0ialloc(size_t size, bool zero, bool is_internal) { + if (unlikely(malloc_init_a0())) { + return NULL; + } + + return iallocztm(TSDN_NULL, size, sz_size2index(size), zero, NULL, + is_internal, arena_get(TSDN_NULL, 0, true), true); +} + +static void +a0idalloc(void *ptr, bool is_internal) { + idalloctm(TSDN_NULL, ptr, NULL, NULL, is_internal, true); +} + +void * +a0malloc(size_t size) { + return a0ialloc(size, false, true); +} + +void +a0dalloc(void *ptr) { + a0idalloc(ptr, true); +} + +/* + * FreeBSD's libc uses the bootstrap_*() functions in bootstrap-senstive + * situations that cannot tolerate TLS variable access (TLS allocation and very + * early internal data structure initialization). + */ + +void * +bootstrap_malloc(size_t size) { + if (unlikely(size == 0)) { + size = 1; + } + + return a0ialloc(size, false, false); +} + +void * +bootstrap_calloc(size_t num, size_t size) { + size_t num_size; + + num_size = num * size; + if (unlikely(num_size == 0)) { + assert(num == 0 || size == 0); + num_size = 1; + } + + return a0ialloc(num_size, true, false); +} + +void +bootstrap_free(void *ptr) { + if (unlikely(ptr == NULL)) { + return; + } + + a0idalloc(ptr, false); +} + +void +arena_set(unsigned ind, arena_t *arena) { + atomic_store_p(&arenas[ind], arena, ATOMIC_RELEASE); +} + +static void +narenas_total_set(unsigned narenas) { + atomic_store_u(&narenas_total, narenas, ATOMIC_RELEASE); +} + +static void +narenas_total_inc(void) { + atomic_fetch_add_u(&narenas_total, 1, ATOMIC_RELEASE); +} + +unsigned +narenas_total_get(void) { + return atomic_load_u(&narenas_total, ATOMIC_ACQUIRE); +} + +/* Create a new arena and insert it into the arenas array at index ind. */ +static arena_t * +arena_init_locked(tsdn_t *tsdn, unsigned ind, extent_hooks_t *extent_hooks) { + arena_t *arena; + + assert(ind <= narenas_total_get()); + if (ind >= MALLOCX_ARENA_LIMIT) { + return NULL; + } + if (ind == narenas_total_get()) { + narenas_total_inc(); + } + + /* + * Another thread may have already initialized arenas[ind] if it's an + * auto arena. + */ + arena = arena_get(tsdn, ind, false); + if (arena != NULL) { + assert(arena_is_auto(arena)); + return arena; + } + + /* Actually initialize the arena. */ + arena = arena_new(tsdn, ind, extent_hooks); + + return arena; +} + +static void +arena_new_create_background_thread(tsdn_t *tsdn, unsigned ind) { + if (ind == 0) { + return; + } + /* + * Avoid creating a new background thread just for the huge arena, which + * purges eagerly by default. + */ + if (have_background_thread && !arena_is_huge(ind)) { + if (background_thread_create(tsdn_tsd(tsdn), ind)) { + malloc_printf("<jemalloc>: error in background thread " + "creation for arena %u. Abort.\n", ind); + abort(); + } + } +} + +arena_t * +arena_init(tsdn_t *tsdn, unsigned ind, extent_hooks_t *extent_hooks) { + arena_t *arena; + + malloc_mutex_lock(tsdn, &arenas_lock); + arena = arena_init_locked(tsdn, ind, extent_hooks); + malloc_mutex_unlock(tsdn, &arenas_lock); + + arena_new_create_background_thread(tsdn, ind); + + return arena; +} + +static void +arena_bind(tsd_t *tsd, unsigned ind, bool internal) { + arena_t *arena = arena_get(tsd_tsdn(tsd), ind, false); + arena_nthreads_inc(arena, internal); + + if (internal) { + tsd_iarena_set(tsd, arena); + } else { + tsd_arena_set(tsd, arena); + unsigned shard = atomic_fetch_add_u(&arena->binshard_next, 1, + ATOMIC_RELAXED); + tsd_binshards_t *bins = tsd_binshardsp_get(tsd); + for (unsigned i = 0; i < SC_NBINS; i++) { + assert(bin_infos[i].n_shards > 0 && + bin_infos[i].n_shards <= BIN_SHARDS_MAX); + bins->binshard[i] = shard % bin_infos[i].n_shards; + } + } +} + +void +arena_migrate(tsd_t *tsd, unsigned oldind, unsigned newind) { + arena_t *oldarena, *newarena; + + oldarena = arena_get(tsd_tsdn(tsd), oldind, false); + newarena = arena_get(tsd_tsdn(tsd), newind, false); + arena_nthreads_dec(oldarena, false); + arena_nthreads_inc(newarena, false); + tsd_arena_set(tsd, newarena); +} + +static void +arena_unbind(tsd_t *tsd, unsigned ind, bool internal) { + arena_t *arena; + + arena = arena_get(tsd_tsdn(tsd), ind, false); + arena_nthreads_dec(arena, internal); + + if (internal) { + tsd_iarena_set(tsd, NULL); + } else { + tsd_arena_set(tsd, NULL); + } +} + +arena_tdata_t * +arena_tdata_get_hard(tsd_t *tsd, unsigned ind) { + arena_tdata_t *tdata, *arenas_tdata_old; + arena_tdata_t *arenas_tdata = tsd_arenas_tdata_get(tsd); + unsigned narenas_tdata_old, i; + unsigned narenas_tdata = tsd_narenas_tdata_get(tsd); + unsigned narenas_actual = narenas_total_get(); + + /* + * Dissociate old tdata array (and set up for deallocation upon return) + * if it's too small. + */ + if (arenas_tdata != NULL && narenas_tdata < narenas_actual) { + arenas_tdata_old = arenas_tdata; + narenas_tdata_old = narenas_tdata; + arenas_tdata = NULL; + narenas_tdata = 0; + tsd_arenas_tdata_set(tsd, arenas_tdata); + tsd_narenas_tdata_set(tsd, narenas_tdata); + } else { + arenas_tdata_old = NULL; + narenas_tdata_old = 0; + } + + /* Allocate tdata array if it's missing. */ + if (arenas_tdata == NULL) { + bool *arenas_tdata_bypassp = tsd_arenas_tdata_bypassp_get(tsd); + narenas_tdata = (ind < narenas_actual) ? narenas_actual : ind+1; + + if (tsd_nominal(tsd) && !*arenas_tdata_bypassp) { + *arenas_tdata_bypassp = true; + arenas_tdata = (arena_tdata_t *)a0malloc( + sizeof(arena_tdata_t) * narenas_tdata); + *arenas_tdata_bypassp = false; + } + if (arenas_tdata == NULL) { + tdata = NULL; + goto label_return; + } + assert(tsd_nominal(tsd) && !*arenas_tdata_bypassp); + tsd_arenas_tdata_set(tsd, arenas_tdata); + tsd_narenas_tdata_set(tsd, narenas_tdata); + } + + /* + * Copy to tdata array. It's possible that the actual number of arenas + * has increased since narenas_total_get() was called above, but that + * causes no correctness issues unless two threads concurrently execute + * the arenas.create mallctl, which we trust mallctl synchronization to + * prevent. + */ + + /* Copy/initialize tickers. */ + for (i = 0; i < narenas_actual; i++) { + if (i < narenas_tdata_old) { + ticker_copy(&arenas_tdata[i].decay_ticker, + &arenas_tdata_old[i].decay_ticker); + } else { + ticker_init(&arenas_tdata[i].decay_ticker, + DECAY_NTICKS_PER_UPDATE); + } + } + if (narenas_tdata > narenas_actual) { + memset(&arenas_tdata[narenas_actual], 0, sizeof(arena_tdata_t) + * (narenas_tdata - narenas_actual)); + } + + /* Read the refreshed tdata array. */ + tdata = &arenas_tdata[ind]; +label_return: + if (arenas_tdata_old != NULL) { + a0dalloc(arenas_tdata_old); + } + return tdata; +} + +/* Slow path, called only by arena_choose(). */ +arena_t * +arena_choose_hard(tsd_t *tsd, bool internal) { + arena_t *ret JEMALLOC_CC_SILENCE_INIT(NULL); + + if (have_percpu_arena && PERCPU_ARENA_ENABLED(opt_percpu_arena)) { + unsigned choose = percpu_arena_choose(); + ret = arena_get(tsd_tsdn(tsd), choose, true); + assert(ret != NULL); + arena_bind(tsd, arena_ind_get(ret), false); + arena_bind(tsd, arena_ind_get(ret), true); + + return ret; + } + + if (narenas_auto > 1) { + unsigned i, j, choose[2], first_null; + bool is_new_arena[2]; + + /* + * Determine binding for both non-internal and internal + * allocation. + * + * choose[0]: For application allocation. + * choose[1]: For internal metadata allocation. + */ + + for (j = 0; j < 2; j++) { + choose[j] = 0; + is_new_arena[j] = false; + } + + first_null = narenas_auto; + malloc_mutex_lock(tsd_tsdn(tsd), &arenas_lock); + assert(arena_get(tsd_tsdn(tsd), 0, false) != NULL); + for (i = 1; i < narenas_auto; i++) { + if (arena_get(tsd_tsdn(tsd), i, false) != NULL) { + /* + * Choose the first arena that has the lowest + * number of threads assigned to it. + */ + for (j = 0; j < 2; j++) { + if (arena_nthreads_get(arena_get( + tsd_tsdn(tsd), i, false), !!j) < + arena_nthreads_get(arena_get( + tsd_tsdn(tsd), choose[j], false), + !!j)) { + choose[j] = i; + } + } + } else if (first_null == narenas_auto) { + /* + * Record the index of the first uninitialized + * arena, in case all extant arenas are in use. + * + * NB: It is possible for there to be + * discontinuities in terms of initialized + * versus uninitialized arenas, due to the + * "thread.arena" mallctl. + */ + first_null = i; + } + } + + for (j = 0; j < 2; j++) { + if (arena_nthreads_get(arena_get(tsd_tsdn(tsd), + choose[j], false), !!j) == 0 || first_null == + narenas_auto) { + /* + * Use an unloaded arena, or the least loaded + * arena if all arenas are already initialized. + */ + if (!!j == internal) { + ret = arena_get(tsd_tsdn(tsd), + choose[j], false); + } + } else { + arena_t *arena; + + /* Initialize a new arena. */ + choose[j] = first_null; + arena = arena_init_locked(tsd_tsdn(tsd), + choose[j], + (extent_hooks_t *)&extent_hooks_default); + if (arena == NULL) { + malloc_mutex_unlock(tsd_tsdn(tsd), + &arenas_lock); + return NULL; + } + is_new_arena[j] = true; + if (!!j == internal) { + ret = arena; + } + } + arena_bind(tsd, choose[j], !!j); + } + malloc_mutex_unlock(tsd_tsdn(tsd), &arenas_lock); + + for (j = 0; j < 2; j++) { + if (is_new_arena[j]) { + assert(choose[j] > 0); + arena_new_create_background_thread( + tsd_tsdn(tsd), choose[j]); + } + } + + } else { + ret = arena_get(tsd_tsdn(tsd), 0, false); + arena_bind(tsd, 0, false); + arena_bind(tsd, 0, true); + } + + return ret; +} + +void +iarena_cleanup(tsd_t *tsd) { + arena_t *iarena; + + iarena = tsd_iarena_get(tsd); + if (iarena != NULL) { + arena_unbind(tsd, arena_ind_get(iarena), true); + } +} + +void +arena_cleanup(tsd_t *tsd) { + arena_t *arena; + + arena = tsd_arena_get(tsd); + if (arena != NULL) { + arena_unbind(tsd, arena_ind_get(arena), false); + } +} + +void +arenas_tdata_cleanup(tsd_t *tsd) { + arena_tdata_t *arenas_tdata; + + /* Prevent tsd->arenas_tdata from being (re)created. */ + *tsd_arenas_tdata_bypassp_get(tsd) = true; + + arenas_tdata = tsd_arenas_tdata_get(tsd); + if (arenas_tdata != NULL) { + tsd_arenas_tdata_set(tsd, NULL); + a0dalloc(arenas_tdata); + } +} + +static void +stats_print_atexit(void) { + if (config_stats) { + tsdn_t *tsdn; + unsigned narenas, i; + + tsdn = tsdn_fetch(); + + /* + * Merge stats from extant threads. This is racy, since + * individual threads do not lock when recording tcache stats + * events. As a consequence, the final stats may be slightly + * out of date by the time they are reported, if other threads + * continue to allocate. + */ + for (i = 0, narenas = narenas_total_get(); i < narenas; i++) { + arena_t *arena = arena_get(tsdn, i, false); + if (arena != NULL) { + tcache_t *tcache; + + malloc_mutex_lock(tsdn, &arena->tcache_ql_mtx); + ql_foreach(tcache, &arena->tcache_ql, link) { + tcache_stats_merge(tsdn, tcache, arena); + } + malloc_mutex_unlock(tsdn, + &arena->tcache_ql_mtx); + } + } + } + je_malloc_stats_print(NULL, NULL, opt_stats_print_opts); +} + +/* + * Ensure that we don't hold any locks upon entry to or exit from allocator + * code (in a "broad" sense that doesn't count a reentrant allocation as an + * entrance or exit). + */ +JEMALLOC_ALWAYS_INLINE void +check_entry_exit_locking(tsdn_t *tsdn) { + if (!config_debug) { + return; + } + if (tsdn_null(tsdn)) { + return; + } + tsd_t *tsd = tsdn_tsd(tsdn); + /* + * It's possible we hold locks at entry/exit if we're in a nested + * allocation. + */ + int8_t reentrancy_level = tsd_reentrancy_level_get(tsd); + if (reentrancy_level != 0) { + return; + } + witness_assert_lockless(tsdn_witness_tsdp_get(tsdn)); +} + +/* + * End miscellaneous support functions. + */ +/******************************************************************************/ +/* + * Begin initialization functions. + */ + +static char * +jemalloc_secure_getenv(const char *name) { +#ifdef JEMALLOC_HAVE_SECURE_GETENV + return secure_getenv(name); +#else +# ifdef JEMALLOC_HAVE_ISSETUGID + if (issetugid() != 0) { + return NULL; + } +# endif + return getenv(name); +#endif +} + +static unsigned +malloc_ncpus(void) { + long result; + +#ifdef _WIN32 + SYSTEM_INFO si; + GetSystemInfo(&si); + result = si.dwNumberOfProcessors; +#elif defined(JEMALLOC_GLIBC_MALLOC_HOOK) && defined(CPU_COUNT) + /* + * glibc >= 2.6 has the CPU_COUNT macro. + * + * glibc's sysconf() uses isspace(). glibc allocates for the first time + * *before* setting up the isspace tables. Therefore we need a + * different method to get the number of CPUs. + */ + { + cpu_set_t set; + + pthread_getaffinity_np(pthread_self(), sizeof(set), &set); + result = CPU_COUNT(&set); + } +#else + result = sysconf(_SC_NPROCESSORS_ONLN); +#endif + return ((result == -1) ? 1 : (unsigned)result); +} + +static void +init_opt_stats_print_opts(const char *v, size_t vlen) { + size_t opts_len = strlen(opt_stats_print_opts); + assert(opts_len <= stats_print_tot_num_options); + + for (size_t i = 0; i < vlen; i++) { + switch (v[i]) { +#define OPTION(o, v, d, s) case o: break; + STATS_PRINT_OPTIONS +#undef OPTION + default: continue; + } + + if (strchr(opt_stats_print_opts, v[i]) != NULL) { + /* Ignore repeated. */ + continue; + } + + opt_stats_print_opts[opts_len++] = v[i]; + opt_stats_print_opts[opts_len] = '\0'; + assert(opts_len <= stats_print_tot_num_options); + } + assert(opts_len == strlen(opt_stats_print_opts)); +} + +/* Reads the next size pair in a multi-sized option. */ +static bool +malloc_conf_multi_sizes_next(const char **slab_size_segment_cur, + size_t *vlen_left, size_t *slab_start, size_t *slab_end, size_t *new_size) { + const char *cur = *slab_size_segment_cur; + char *end; + uintmax_t um; + + set_errno(0); + + /* First number, then '-' */ + um = malloc_strtoumax(cur, &end, 0); + if (get_errno() != 0 || *end != '-') { + return true; + } + *slab_start = (size_t)um; + cur = end + 1; + + /* Second number, then ':' */ + um = malloc_strtoumax(cur, &end, 0); + if (get_errno() != 0 || *end != ':') { + return true; + } + *slab_end = (size_t)um; + cur = end + 1; + + /* Last number */ + um = malloc_strtoumax(cur, &end, 0); + if (get_errno() != 0) { + return true; + } + *new_size = (size_t)um; + + /* Consume the separator if there is one. */ + if (*end == '|') { + end++; + } + + *vlen_left -= end - *slab_size_segment_cur; + *slab_size_segment_cur = end; + + return false; +} + +static bool +malloc_conf_next(char const **opts_p, char const **k_p, size_t *klen_p, + char const **v_p, size_t *vlen_p) { + bool accept; + const char *opts = *opts_p; + + *k_p = opts; + + for (accept = false; !accept;) { + switch (*opts) { + case 'A': case 'B': case 'C': case 'D': case 'E': case 'F': + case 'G': case 'H': case 'I': case 'J': case 'K': case 'L': + case 'M': case 'N': case 'O': case 'P': case 'Q': case 'R': + case 'S': case 'T': case 'U': case 'V': case 'W': case 'X': + case 'Y': case 'Z': + case 'a': case 'b': case 'c': case 'd': case 'e': case 'f': + case 'g': case 'h': case 'i': case 'j': case 'k': case 'l': + case 'm': case 'n': case 'o': case 'p': case 'q': case 'r': + case 's': case 't': case 'u': case 'v': case 'w': case 'x': + case 'y': case 'z': + case '0': case '1': case '2': case '3': case '4': case '5': + case '6': case '7': case '8': case '9': + case '_': + opts++; + break; + case ':': + opts++; + *klen_p = (uintptr_t)opts - 1 - (uintptr_t)*k_p; + *v_p = opts; + accept = true; + break; + case '\0': + if (opts != *opts_p) { + malloc_write("<jemalloc>: Conf string ends " + "with key\n"); + } + return true; + default: + malloc_write("<jemalloc>: Malformed conf string\n"); + return true; + } + } + + for (accept = false; !accept;) { + switch (*opts) { + case ',': + opts++; + /* + * Look ahead one character here, because the next time + * this function is called, it will assume that end of + * input has been cleanly reached if no input remains, + * but we have optimistically already consumed the + * comma if one exists. + */ + if (*opts == '\0') { + malloc_write("<jemalloc>: Conf string ends " + "with comma\n"); + } + *vlen_p = (uintptr_t)opts - 1 - (uintptr_t)*v_p; + accept = true; + break; + case '\0': + *vlen_p = (uintptr_t)opts - (uintptr_t)*v_p; + accept = true; + break; + default: + opts++; + break; + } + } + + *opts_p = opts; + return false; +} + +static void +malloc_abort_invalid_conf(void) { + assert(opt_abort_conf); + malloc_printf("<jemalloc>: Abort (abort_conf:true) on invalid conf " + "value (see above).\n"); + abort(); +} + +static void +malloc_conf_error(const char *msg, const char *k, size_t klen, const char *v, + size_t vlen) { + malloc_printf("<jemalloc>: %s: %.*s:%.*s\n", msg, (int)klen, k, + (int)vlen, v); + /* If abort_conf is set, error out after processing all options. */ + const char *experimental = "experimental_"; + if (strncmp(k, experimental, strlen(experimental)) == 0) { + /* However, tolerate experimental features. */ + return; + } + had_conf_error = true; +} + +static void +malloc_slow_flag_init(void) { + /* + * Combine the runtime options into malloc_slow for fast path. Called + * after processing all the options. + */ + malloc_slow_flags |= (opt_junk_alloc ? flag_opt_junk_alloc : 0) + | (opt_junk_free ? flag_opt_junk_free : 0) + | (opt_zero ? flag_opt_zero : 0) + | (opt_utrace ? flag_opt_utrace : 0) + | (opt_xmalloc ? flag_opt_xmalloc : 0); + + malloc_slow = (malloc_slow_flags != 0); +} + +/* Number of sources for initializing malloc_conf */ +#define MALLOC_CONF_NSOURCES 4 + +static const char * +obtain_malloc_conf(unsigned which_source, char buf[PATH_MAX + 1]) { + if (config_debug) { + static unsigned read_source = 0; + /* + * Each source should only be read once, to minimize # of + * syscalls on init. + */ + assert(read_source++ == which_source); + } + assert(which_source < MALLOC_CONF_NSOURCES); + + const char *ret; + switch (which_source) { + case 0: + ret = config_malloc_conf; + break; + case 1: + if (je_malloc_conf != NULL) { + /* Use options that were compiled into the program. */ + ret = je_malloc_conf; + } else { + /* No configuration specified. */ + ret = NULL; + } + break; + case 2: { + ssize_t linklen = 0; +#ifndef _WIN32 + int saved_errno = errno; + const char *linkname = +# ifdef JEMALLOC_PREFIX + "/etc/"JEMALLOC_PREFIX"malloc.conf" +# else + "/etc/malloc.conf" +# endif + ; + + /* + * Try to use the contents of the "/etc/malloc.conf" symbolic + * link's name. + */ +#ifndef JEMALLOC_READLINKAT + linklen = readlink(linkname, buf, PATH_MAX); +#else + linklen = readlinkat(AT_FDCWD, linkname, buf, PATH_MAX); +#endif + if (linklen == -1) { + /* No configuration specified. */ + linklen = 0; + /* Restore errno. */ + set_errno(saved_errno); + } +#endif + buf[linklen] = '\0'; + ret = buf; + break; + } case 3: { + const char *envname = +#ifdef JEMALLOC_PREFIX + JEMALLOC_CPREFIX"MALLOC_CONF" +#else + "MALLOC_CONF" +#endif + ; + + if ((ret = jemalloc_secure_getenv(envname)) != NULL) { + /* + * Do nothing; opts is already initialized to the value + * of the MALLOC_CONF environment variable. + */ + } else { + /* No configuration specified. */ + ret = NULL; + } + break; + } default: + not_reached(); + ret = NULL; + } + return ret; +} + +static void +malloc_conf_init_helper(sc_data_t *sc_data, unsigned bin_shard_sizes[SC_NBINS], + bool initial_call, const char *opts_cache[MALLOC_CONF_NSOURCES], + char buf[PATH_MAX + 1]) { + static const char *opts_explain[MALLOC_CONF_NSOURCES] = { + "string specified via --with-malloc-conf", + "string pointed to by the global variable malloc_conf", + "\"name\" of the file referenced by the symbolic link named " + "/etc/malloc.conf", + "value of the environment variable MALLOC_CONF" + }; + unsigned i; + const char *opts, *k, *v; + size_t klen, vlen; + + for (i = 0; i < MALLOC_CONF_NSOURCES; i++) { + /* Get runtime configuration. */ + if (initial_call) { + opts_cache[i] = obtain_malloc_conf(i, buf); + } + opts = opts_cache[i]; + if (!initial_call && opt_confirm_conf) { + malloc_printf( + "<jemalloc>: malloc_conf #%u (%s): \"%s\"\n", + i + 1, opts_explain[i], opts != NULL ? opts : ""); + } + if (opts == NULL) { + continue; + } + + while (*opts != '\0' && !malloc_conf_next(&opts, &k, &klen, &v, + &vlen)) { + +#define CONF_ERROR(msg, k, klen, v, vlen) \ + if (!initial_call) { \ + malloc_conf_error( \ + msg, k, klen, v, vlen); \ + cur_opt_valid = false; \ + } +#define CONF_CONTINUE { \ + if (!initial_call && opt_confirm_conf \ + && cur_opt_valid) { \ + malloc_printf("<jemalloc>: -- " \ + "Set conf value: %.*s:%.*s" \ + "\n", (int)klen, k, \ + (int)vlen, v); \ + } \ + continue; \ + } +#define CONF_MATCH(n) \ + (sizeof(n)-1 == klen && strncmp(n, k, klen) == 0) +#define CONF_MATCH_VALUE(n) \ + (sizeof(n)-1 == vlen && strncmp(n, v, vlen) == 0) +#define CONF_HANDLE_BOOL(o, n) \ + if (CONF_MATCH(n)) { \ + if (CONF_MATCH_VALUE("true")) { \ + o = true; \ + } else if (CONF_MATCH_VALUE("false")) { \ + o = false; \ + } else { \ + CONF_ERROR("Invalid conf value",\ + k, klen, v, vlen); \ + } \ + CONF_CONTINUE; \ + } + /* + * One of the CONF_MIN macros below expands, in one of the use points, + * to "unsigned integer < 0", which is always false, triggering the + * GCC -Wtype-limits warning, which we disable here and re-enable below. + */ + JEMALLOC_DIAGNOSTIC_PUSH + JEMALLOC_DIAGNOSTIC_IGNORE_TYPE_LIMITS + +#define CONF_DONT_CHECK_MIN(um, min) false +#define CONF_CHECK_MIN(um, min) ((um) < (min)) +#define CONF_DONT_CHECK_MAX(um, max) false +#define CONF_CHECK_MAX(um, max) ((um) > (max)) +#define CONF_HANDLE_T_U(t, o, n, min, max, check_min, check_max, clip) \ + if (CONF_MATCH(n)) { \ + uintmax_t um; \ + char *end; \ + \ + set_errno(0); \ + um = malloc_strtoumax(v, &end, 0); \ + if (get_errno() != 0 || (uintptr_t)end -\ + (uintptr_t)v != vlen) { \ + CONF_ERROR("Invalid conf value",\ + k, klen, v, vlen); \ + } else if (clip) { \ + if (check_min(um, (t)(min))) { \ + o = (t)(min); \ + } else if ( \ + check_max(um, (t)(max))) { \ + o = (t)(max); \ + } else { \ + o = (t)um; \ + } \ + } else { \ + if (check_min(um, (t)(min)) || \ + check_max(um, (t)(max))) { \ + CONF_ERROR( \ + "Out-of-range " \ + "conf value", \ + k, klen, v, vlen); \ + } else { \ + o = (t)um; \ + } \ + } \ + CONF_CONTINUE; \ + } +#define CONF_HANDLE_UNSIGNED(o, n, min, max, check_min, check_max, \ + clip) \ + CONF_HANDLE_T_U(unsigned, o, n, min, max, \ + check_min, check_max, clip) +#define CONF_HANDLE_SIZE_T(o, n, min, max, check_min, check_max, clip) \ + CONF_HANDLE_T_U(size_t, o, n, min, max, \ + check_min, check_max, clip) +#define CONF_HANDLE_SSIZE_T(o, n, min, max) \ + if (CONF_MATCH(n)) { \ + long l; \ + char *end; \ + \ + set_errno(0); \ + l = strtol(v, &end, 0); \ + if (get_errno() != 0 || (uintptr_t)end -\ + (uintptr_t)v != vlen) { \ + CONF_ERROR("Invalid conf value",\ + k, klen, v, vlen); \ + } else if (l < (ssize_t)(min) || l > \ + (ssize_t)(max)) { \ + CONF_ERROR( \ + "Out-of-range conf value", \ + k, klen, v, vlen); \ + } else { \ + o = l; \ + } \ + CONF_CONTINUE; \ + } +#define CONF_HANDLE_CHAR_P(o, n, d) \ + if (CONF_MATCH(n)) { \ + size_t cpylen = (vlen <= \ + sizeof(o)-1) ? vlen : \ + sizeof(o)-1; \ + strncpy(o, v, cpylen); \ + o[cpylen] = '\0'; \ + CONF_CONTINUE; \ + } + + bool cur_opt_valid = true; + + CONF_HANDLE_BOOL(opt_confirm_conf, "confirm_conf") + if (initial_call) { + continue; + } + + CONF_HANDLE_BOOL(opt_abort, "abort") + CONF_HANDLE_BOOL(opt_abort_conf, "abort_conf") + if (strncmp("metadata_thp", k, klen) == 0) { + int i; + bool match = false; + for (i = 0; i < metadata_thp_mode_limit; i++) { + if (strncmp(metadata_thp_mode_names[i], + v, vlen) == 0) { + opt_metadata_thp = i; + match = true; + break; + } + } + if (!match) { + CONF_ERROR("Invalid conf value", + k, klen, v, vlen); + } + CONF_CONTINUE; + } + CONF_HANDLE_BOOL(opt_retain, "retain") + if (strncmp("dss", k, klen) == 0) { + int i; + bool match = false; + for (i = 0; i < dss_prec_limit; i++) { + if (strncmp(dss_prec_names[i], v, vlen) + == 0) { + if (extent_dss_prec_set(i)) { + CONF_ERROR( + "Error setting dss", + k, klen, v, vlen); + } else { + opt_dss = + dss_prec_names[i]; + match = true; + break; + } + } + } + if (!match) { + CONF_ERROR("Invalid conf value", + k, klen, v, vlen); + } + CONF_CONTINUE; + } + CONF_HANDLE_UNSIGNED(opt_narenas, "narenas", 1, + UINT_MAX, CONF_CHECK_MIN, CONF_DONT_CHECK_MAX, + false) + if (CONF_MATCH("bin_shards")) { + const char *bin_shards_segment_cur = v; + size_t vlen_left = vlen; + do { + size_t size_start; + size_t size_end; + size_t nshards; + bool err = malloc_conf_multi_sizes_next( + &bin_shards_segment_cur, &vlen_left, + &size_start, &size_end, &nshards); + if (err || bin_update_shard_size( + bin_shard_sizes, size_start, + size_end, nshards)) { + CONF_ERROR( + "Invalid settings for " + "bin_shards", k, klen, v, + vlen); + break; + } + } while (vlen_left > 0); + CONF_CONTINUE; + } + CONF_HANDLE_SSIZE_T(opt_dirty_decay_ms, + "dirty_decay_ms", -1, NSTIME_SEC_MAX * KQU(1000) < + QU(SSIZE_MAX) ? NSTIME_SEC_MAX * KQU(1000) : + SSIZE_MAX); + CONF_HANDLE_SSIZE_T(opt_muzzy_decay_ms, + "muzzy_decay_ms", -1, NSTIME_SEC_MAX * KQU(1000) < + QU(SSIZE_MAX) ? NSTIME_SEC_MAX * KQU(1000) : + SSIZE_MAX); + CONF_HANDLE_BOOL(opt_stats_print, "stats_print") + if (CONF_MATCH("stats_print_opts")) { + init_opt_stats_print_opts(v, vlen); + CONF_CONTINUE; + } + if (config_fill) { + if (CONF_MATCH("junk")) { + if (CONF_MATCH_VALUE("true")) { + opt_junk = "true"; + opt_junk_alloc = opt_junk_free = + true; + } else if (CONF_MATCH_VALUE("false")) { + opt_junk = "false"; + opt_junk_alloc = opt_junk_free = + false; + } else if (CONF_MATCH_VALUE("alloc")) { + opt_junk = "alloc"; + opt_junk_alloc = true; + opt_junk_free = false; + } else if (CONF_MATCH_VALUE("free")) { + opt_junk = "free"; + opt_junk_alloc = false; + opt_junk_free = true; + } else { + CONF_ERROR( + "Invalid conf value", + k, klen, v, vlen); + } + CONF_CONTINUE; + } + CONF_HANDLE_BOOL(opt_zero, "zero") + } + if (config_utrace) { + CONF_HANDLE_BOOL(opt_utrace, "utrace") + } + if (config_xmalloc) { + CONF_HANDLE_BOOL(opt_xmalloc, "xmalloc") + } + CONF_HANDLE_BOOL(opt_tcache, "tcache") + CONF_HANDLE_SSIZE_T(opt_lg_tcache_max, "lg_tcache_max", + -1, (sizeof(size_t) << 3) - 1) + + /* + * The runtime option of oversize_threshold remains + * undocumented. It may be tweaked in the next major + * release (6.0). The default value 8M is rather + * conservative / safe. Tuning it further down may + * improve fragmentation a bit more, but may also cause + * contention on the huge arena. + */ + CONF_HANDLE_SIZE_T(opt_oversize_threshold, + "oversize_threshold", 0, SC_LARGE_MAXCLASS, + CONF_DONT_CHECK_MIN, CONF_CHECK_MAX, false) + CONF_HANDLE_SIZE_T(opt_lg_extent_max_active_fit, + "lg_extent_max_active_fit", 0, + (sizeof(size_t) << 3), CONF_DONT_CHECK_MIN, + CONF_CHECK_MAX, false) + + if (strncmp("percpu_arena", k, klen) == 0) { + bool match = false; + for (int i = percpu_arena_mode_names_base; i < + percpu_arena_mode_names_limit; i++) { + if (strncmp(percpu_arena_mode_names[i], + v, vlen) == 0) { + if (!have_percpu_arena) { + CONF_ERROR( + "No getcpu support", + k, klen, v, vlen); + } + opt_percpu_arena = i; + match = true; + break; + } + } + if (!match) { + CONF_ERROR("Invalid conf value", + k, klen, v, vlen); + } + CONF_CONTINUE; + } + CONF_HANDLE_BOOL(opt_background_thread, + "background_thread"); + CONF_HANDLE_SIZE_T(opt_max_background_threads, + "max_background_threads", 1, + opt_max_background_threads, + CONF_CHECK_MIN, CONF_CHECK_MAX, + true); + if (CONF_MATCH("slab_sizes")) { + bool err; + const char *slab_size_segment_cur = v; + size_t vlen_left = vlen; + do { + size_t slab_start; + size_t slab_end; + size_t pgs; + err = malloc_conf_multi_sizes_next( + &slab_size_segment_cur, + &vlen_left, &slab_start, &slab_end, + &pgs); + if (!err) { + sc_data_update_slab_size( + sc_data, slab_start, + slab_end, (int)pgs); + } else { + CONF_ERROR("Invalid settings " + "for slab_sizes", + k, klen, v, vlen); + } + } while (!err && vlen_left > 0); + CONF_CONTINUE; + } + if (config_prof) { + CONF_HANDLE_BOOL(opt_prof, "prof") + CONF_HANDLE_CHAR_P(opt_prof_prefix, + "prof_prefix", "jeprof") + CONF_HANDLE_BOOL(opt_prof_active, "prof_active") + CONF_HANDLE_BOOL(opt_prof_thread_active_init, + "prof_thread_active_init") + CONF_HANDLE_SIZE_T(opt_lg_prof_sample, + "lg_prof_sample", 0, (sizeof(uint64_t) << 3) + - 1, CONF_DONT_CHECK_MIN, CONF_CHECK_MAX, + true) + CONF_HANDLE_BOOL(opt_prof_accum, "prof_accum") + CONF_HANDLE_SSIZE_T(opt_lg_prof_interval, + "lg_prof_interval", -1, + (sizeof(uint64_t) << 3) - 1) + CONF_HANDLE_BOOL(opt_prof_gdump, "prof_gdump") + CONF_HANDLE_BOOL(opt_prof_final, "prof_final") + CONF_HANDLE_BOOL(opt_prof_leak, "prof_leak") + CONF_HANDLE_BOOL(opt_prof_log, "prof_log") + } + if (config_log) { + if (CONF_MATCH("log")) { + size_t cpylen = ( + vlen <= sizeof(log_var_names) ? + vlen : sizeof(log_var_names) - 1); + strncpy(log_var_names, v, cpylen); + log_var_names[cpylen] = '\0'; + CONF_CONTINUE; + } + } + if (CONF_MATCH("thp")) { + bool match = false; + for (int i = 0; i < thp_mode_names_limit; i++) { + if (strncmp(thp_mode_names[i],v, vlen) + == 0) { + if (!have_madvise_huge) { + CONF_ERROR( + "No THP support", + k, klen, v, vlen); + } + opt_thp = i; + match = true; + break; + } + } + if (!match) { + CONF_ERROR("Invalid conf value", + k, klen, v, vlen); + } + CONF_CONTINUE; + } + CONF_ERROR("Invalid conf pair", k, klen, v, vlen); +#undef CONF_ERROR +#undef CONF_CONTINUE +#undef CONF_MATCH +#undef CONF_MATCH_VALUE +#undef CONF_HANDLE_BOOL +#undef CONF_DONT_CHECK_MIN +#undef CONF_CHECK_MIN +#undef CONF_DONT_CHECK_MAX +#undef CONF_CHECK_MAX +#undef CONF_HANDLE_T_U +#undef CONF_HANDLE_UNSIGNED +#undef CONF_HANDLE_SIZE_T +#undef CONF_HANDLE_SSIZE_T +#undef CONF_HANDLE_CHAR_P + /* Re-enable diagnostic "-Wtype-limits" */ + JEMALLOC_DIAGNOSTIC_POP + } + if (opt_abort_conf && had_conf_error) { + malloc_abort_invalid_conf(); + } + } + atomic_store_b(&log_init_done, true, ATOMIC_RELEASE); +} + +static void +malloc_conf_init(sc_data_t *sc_data, unsigned bin_shard_sizes[SC_NBINS]) { + const char *opts_cache[MALLOC_CONF_NSOURCES] = {NULL, NULL, NULL, NULL}; + char buf[PATH_MAX + 1]; + + /* The first call only set the confirm_conf option and opts_cache */ + malloc_conf_init_helper(NULL, NULL, true, opts_cache, buf); + malloc_conf_init_helper(sc_data, bin_shard_sizes, false, opts_cache, + NULL); +} + +#undef MALLOC_CONF_NSOURCES + +static bool +malloc_init_hard_needed(void) { + if (malloc_initialized() || (IS_INITIALIZER && malloc_init_state == + malloc_init_recursible)) { + /* + * Another thread initialized the allocator before this one + * acquired init_lock, or this thread is the initializing + * thread, and it is recursively allocating. + */ + return false; + } +#ifdef JEMALLOC_THREADED_INIT + if (malloc_initializer != NO_INITIALIZER && !IS_INITIALIZER) { + /* Busy-wait until the initializing thread completes. */ + spin_t spinner = SPIN_INITIALIZER; + do { + malloc_mutex_unlock(TSDN_NULL, &init_lock); + spin_adaptive(&spinner); + malloc_mutex_lock(TSDN_NULL, &init_lock); + } while (!malloc_initialized()); + return false; + } +#endif + return true; +} + +static bool +malloc_init_hard_a0_locked() { + malloc_initializer = INITIALIZER; + + JEMALLOC_DIAGNOSTIC_PUSH + JEMALLOC_DIAGNOSTIC_IGNORE_MISSING_STRUCT_FIELD_INITIALIZERS + sc_data_t sc_data = {0}; + JEMALLOC_DIAGNOSTIC_POP + + /* + * Ordering here is somewhat tricky; we need sc_boot() first, since that + * determines what the size classes will be, and then + * malloc_conf_init(), since any slab size tweaking will need to be done + * before sz_boot and bin_boot, which assume that the values they read + * out of sc_data_global are final. + */ + sc_boot(&sc_data); + unsigned bin_shard_sizes[SC_NBINS]; + bin_shard_sizes_boot(bin_shard_sizes); + /* + * prof_boot0 only initializes opt_prof_prefix. We need to do it before + * we parse malloc_conf options, in case malloc_conf parsing overwrites + * it. + */ + if (config_prof) { + prof_boot0(); + } + malloc_conf_init(&sc_data, bin_shard_sizes); + sz_boot(&sc_data); + bin_boot(&sc_data, bin_shard_sizes); + + if (opt_stats_print) { + /* Print statistics at exit. */ + if (atexit(stats_print_atexit) != 0) { + malloc_write("<jemalloc>: Error in atexit()\n"); + if (opt_abort) { + abort(); + } + } + } + if (pages_boot()) { + return true; + } + if (base_boot(TSDN_NULL)) { + return true; + } + if (extent_boot()) { + return true; + } + if (ctl_boot()) { + return true; + } + if (config_prof) { + prof_boot1(); + } + arena_boot(&sc_data); + if (tcache_boot(TSDN_NULL)) { + return true; + } + if (malloc_mutex_init(&arenas_lock, "arenas", WITNESS_RANK_ARENAS, + malloc_mutex_rank_exclusive)) { + return true; + } + hook_boot(); + /* + * Create enough scaffolding to allow recursive allocation in + * malloc_ncpus(). + */ + narenas_auto = 1; + manual_arena_base = narenas_auto + 1; + memset(arenas, 0, sizeof(arena_t *) * narenas_auto); + /* + * Initialize one arena here. The rest are lazily created in + * arena_choose_hard(). + */ + if (arena_init(TSDN_NULL, 0, (extent_hooks_t *)&extent_hooks_default) + == NULL) { + return true; + } + a0 = arena_get(TSDN_NULL, 0, false); + malloc_init_state = malloc_init_a0_initialized; + + return false; +} + +static bool +malloc_init_hard_a0(void) { + bool ret; + + malloc_mutex_lock(TSDN_NULL, &init_lock); + ret = malloc_init_hard_a0_locked(); + malloc_mutex_unlock(TSDN_NULL, &init_lock); + return ret; +} + +/* Initialize data structures which may trigger recursive allocation. */ +static bool +malloc_init_hard_recursible(void) { + malloc_init_state = malloc_init_recursible; + + ncpus = malloc_ncpus(); + +#if (defined(JEMALLOC_HAVE_PTHREAD_ATFORK) && !defined(JEMALLOC_MUTEX_INIT_CB) \ + && !defined(JEMALLOC_ZONE) && !defined(_WIN32) && \ + !defined(__native_client__)) + /* LinuxThreads' pthread_atfork() allocates. */ + if (pthread_atfork(jemalloc_prefork, jemalloc_postfork_parent, + jemalloc_postfork_child) != 0) { + malloc_write("<jemalloc>: Error in pthread_atfork()\n"); + if (opt_abort) { + abort(); + } + return true; + } +#endif + + if (background_thread_boot0()) { + return true; + } + + return false; +} + +static unsigned +malloc_narenas_default(void) { + assert(ncpus > 0); + /* + * For SMP systems, create more than one arena per CPU by + * default. + */ + if (ncpus > 1) { + return ncpus << 2; + } else { + return 1; + } +} + +static percpu_arena_mode_t +percpu_arena_as_initialized(percpu_arena_mode_t mode) { + assert(!malloc_initialized()); + assert(mode <= percpu_arena_disabled); + + if (mode != percpu_arena_disabled) { + mode += percpu_arena_mode_enabled_base; + } + + return mode; +} + +static bool +malloc_init_narenas(void) { + assert(ncpus > 0); + + if (opt_percpu_arena != percpu_arena_disabled) { + if (!have_percpu_arena || malloc_getcpu() < 0) { + opt_percpu_arena = percpu_arena_disabled; + malloc_printf("<jemalloc>: perCPU arena getcpu() not " + "available. Setting narenas to %u.\n", opt_narenas ? + opt_narenas : malloc_narenas_default()); + if (opt_abort) { + abort(); + } + } else { + if (ncpus >= MALLOCX_ARENA_LIMIT) { + malloc_printf("<jemalloc>: narenas w/ percpu" + "arena beyond limit (%d)\n", ncpus); + if (opt_abort) { + abort(); + } + return true; + } + /* NB: opt_percpu_arena isn't fully initialized yet. */ + if (percpu_arena_as_initialized(opt_percpu_arena) == + per_phycpu_arena && ncpus % 2 != 0) { + malloc_printf("<jemalloc>: invalid " + "configuration -- per physical CPU arena " + "with odd number (%u) of CPUs (no hyper " + "threading?).\n", ncpus); + if (opt_abort) + abort(); + } + unsigned n = percpu_arena_ind_limit( + percpu_arena_as_initialized(opt_percpu_arena)); + if (opt_narenas < n) { + /* + * If narenas is specified with percpu_arena + * enabled, actual narenas is set as the greater + * of the two. percpu_arena_choose will be free + * to use any of the arenas based on CPU + * id. This is conservative (at a small cost) + * but ensures correctness. + * + * If for some reason the ncpus determined at + * boot is not the actual number (e.g. because + * of affinity setting from numactl), reserving + * narenas this way provides a workaround for + * percpu_arena. + */ + opt_narenas = n; + } + } + } + if (opt_narenas == 0) { + opt_narenas = malloc_narenas_default(); + } + assert(opt_narenas > 0); + + narenas_auto = opt_narenas; + /* + * Limit the number of arenas to the indexing range of MALLOCX_ARENA(). + */ + if (narenas_auto >= MALLOCX_ARENA_LIMIT) { + narenas_auto = MALLOCX_ARENA_LIMIT - 1; + malloc_printf("<jemalloc>: Reducing narenas to limit (%d)\n", + narenas_auto); + } + narenas_total_set(narenas_auto); + if (arena_init_huge()) { + narenas_total_inc(); + } + manual_arena_base = narenas_total_get(); + + return false; +} + +static void +malloc_init_percpu(void) { + opt_percpu_arena = percpu_arena_as_initialized(opt_percpu_arena); +} + +static bool +malloc_init_hard_finish(void) { + if (malloc_mutex_boot()) { + return true; + } + + malloc_init_state = malloc_init_initialized; + malloc_slow_flag_init(); + + return false; +} + +static void +malloc_init_hard_cleanup(tsdn_t *tsdn, bool reentrancy_set) { + malloc_mutex_assert_owner(tsdn, &init_lock); + malloc_mutex_unlock(tsdn, &init_lock); + if (reentrancy_set) { + assert(!tsdn_null(tsdn)); + tsd_t *tsd = tsdn_tsd(tsdn); + assert(tsd_reentrancy_level_get(tsd) > 0); + post_reentrancy(tsd); + } +} + +static bool +malloc_init_hard(void) { + tsd_t *tsd; + +#if defined(_WIN32) && _WIN32_WINNT < 0x0600 + _init_init_lock(); +#endif + malloc_mutex_lock(TSDN_NULL, &init_lock); + +#define UNLOCK_RETURN(tsdn, ret, reentrancy) \ + malloc_init_hard_cleanup(tsdn, reentrancy); \ + return ret; + + if (!malloc_init_hard_needed()) { + UNLOCK_RETURN(TSDN_NULL, false, false) + } + + if (malloc_init_state != malloc_init_a0_initialized && + malloc_init_hard_a0_locked()) { + UNLOCK_RETURN(TSDN_NULL, true, false) + } + + malloc_mutex_unlock(TSDN_NULL, &init_lock); + /* Recursive allocation relies on functional tsd. */ + tsd = malloc_tsd_boot0(); + if (tsd == NULL) { + return true; + } + if (malloc_init_hard_recursible()) { + return true; + } + + malloc_mutex_lock(tsd_tsdn(tsd), &init_lock); + /* Set reentrancy level to 1 during init. */ + pre_reentrancy(tsd, NULL); + /* Initialize narenas before prof_boot2 (for allocation). */ + if (malloc_init_narenas() || background_thread_boot1(tsd_tsdn(tsd))) { + UNLOCK_RETURN(tsd_tsdn(tsd), true, true) + } + if (config_prof && prof_boot2(tsd)) { + UNLOCK_RETURN(tsd_tsdn(tsd), true, true) + } + + malloc_init_percpu(); + + if (malloc_init_hard_finish()) { + UNLOCK_RETURN(tsd_tsdn(tsd), true, true) + } + post_reentrancy(tsd); + malloc_mutex_unlock(tsd_tsdn(tsd), &init_lock); + + witness_assert_lockless(witness_tsd_tsdn( + tsd_witness_tsdp_get_unsafe(tsd))); + malloc_tsd_boot1(); + /* Update TSD after tsd_boot1. */ + tsd = tsd_fetch(); + if (opt_background_thread) { + assert(have_background_thread); + /* + * Need to finish init & unlock first before creating background + * threads (pthread_create depends on malloc). ctl_init (which + * sets isthreaded) needs to be called without holding any lock. + */ + background_thread_ctl_init(tsd_tsdn(tsd)); + if (background_thread_create(tsd, 0)) { + return true; + } + } +#undef UNLOCK_RETURN + return false; +} + +/* + * End initialization functions. + */ +/******************************************************************************/ +/* + * Begin allocation-path internal functions and data structures. + */ + +/* + * Settings determined by the documented behavior of the allocation functions. + */ +typedef struct static_opts_s static_opts_t; +struct static_opts_s { + /* Whether or not allocation size may overflow. */ + bool may_overflow; + + /* + * Whether or not allocations (with alignment) of size 0 should be + * treated as size 1. + */ + bool bump_empty_aligned_alloc; + /* + * Whether to assert that allocations are not of size 0 (after any + * bumping). + */ + bool assert_nonempty_alloc; + + /* + * Whether or not to modify the 'result' argument to malloc in case of + * error. + */ + bool null_out_result_on_error; + /* Whether to set errno when we encounter an error condition. */ + bool set_errno_on_error; + + /* + * The minimum valid alignment for functions requesting aligned storage. + */ + size_t min_alignment; + + /* The error string to use if we oom. */ + const char *oom_string; + /* The error string to use if the passed-in alignment is invalid. */ + const char *invalid_alignment_string; + + /* + * False if we're configured to skip some time-consuming operations. + * + * This isn't really a malloc "behavior", but it acts as a useful + * summary of several other static (or at least, static after program + * initialization) options. + */ + bool slow; + /* + * Return size. + */ + bool usize; +}; + +JEMALLOC_ALWAYS_INLINE void +static_opts_init(static_opts_t *static_opts) { + static_opts->may_overflow = false; + static_opts->bump_empty_aligned_alloc = false; + static_opts->assert_nonempty_alloc = false; + static_opts->null_out_result_on_error = false; + static_opts->set_errno_on_error = false; + static_opts->min_alignment = 0; + static_opts->oom_string = ""; + static_opts->invalid_alignment_string = ""; + static_opts->slow = false; + static_opts->usize = false; +} + +/* + * These correspond to the macros in jemalloc/jemalloc_macros.h. Broadly, we + * should have one constant here per magic value there. Note however that the + * representations need not be related. + */ +#define TCACHE_IND_NONE ((unsigned)-1) +#define TCACHE_IND_AUTOMATIC ((unsigned)-2) +#define ARENA_IND_AUTOMATIC ((unsigned)-1) + +typedef struct dynamic_opts_s dynamic_opts_t; +struct dynamic_opts_s { + void **result; + size_t usize; + size_t num_items; + size_t item_size; + size_t alignment; + bool zero; + unsigned tcache_ind; + unsigned arena_ind; +}; + +JEMALLOC_ALWAYS_INLINE void +dynamic_opts_init(dynamic_opts_t *dynamic_opts) { + dynamic_opts->result = NULL; + dynamic_opts->usize = 0; + dynamic_opts->num_items = 0; + dynamic_opts->item_size = 0; + dynamic_opts->alignment = 0; + dynamic_opts->zero = false; + dynamic_opts->tcache_ind = TCACHE_IND_AUTOMATIC; + dynamic_opts->arena_ind = ARENA_IND_AUTOMATIC; +} + +/* ind is ignored if dopts->alignment > 0. */ +JEMALLOC_ALWAYS_INLINE void * +imalloc_no_sample(static_opts_t *sopts, dynamic_opts_t *dopts, tsd_t *tsd, + size_t size, size_t usize, szind_t ind) { + tcache_t *tcache; + arena_t *arena; + + /* Fill in the tcache. */ + if (dopts->tcache_ind == TCACHE_IND_AUTOMATIC) { + if (likely(!sopts->slow)) { + /* Getting tcache ptr unconditionally. */ + tcache = tsd_tcachep_get(tsd); + assert(tcache == tcache_get(tsd)); + } else { + tcache = tcache_get(tsd); + } + } else if (dopts->tcache_ind == TCACHE_IND_NONE) { + tcache = NULL; + } else { + tcache = tcaches_get(tsd, dopts->tcache_ind); + } + + /* Fill in the arena. */ + if (dopts->arena_ind == ARENA_IND_AUTOMATIC) { + /* + * In case of automatic arena management, we defer arena + * computation until as late as we can, hoping to fill the + * allocation out of the tcache. + */ + arena = NULL; + } else { + arena = arena_get(tsd_tsdn(tsd), dopts->arena_ind, true); + } + + if (unlikely(dopts->alignment != 0)) { + return ipalloct(tsd_tsdn(tsd), usize, dopts->alignment, + dopts->zero, tcache, arena); + } + + return iallocztm(tsd_tsdn(tsd), size, ind, dopts->zero, tcache, false, + arena, sopts->slow); +} + +JEMALLOC_ALWAYS_INLINE void * +imalloc_sample(static_opts_t *sopts, dynamic_opts_t *dopts, tsd_t *tsd, + size_t usize, szind_t ind) { + void *ret; + + /* + * For small allocations, sampling bumps the usize. If so, we allocate + * from the ind_large bucket. + */ + szind_t ind_large; + size_t bumped_usize = usize; + + if (usize <= SC_SMALL_MAXCLASS) { + assert(((dopts->alignment == 0) ? + sz_s2u(SC_LARGE_MINCLASS) : + sz_sa2u(SC_LARGE_MINCLASS, dopts->alignment)) + == SC_LARGE_MINCLASS); + ind_large = sz_size2index(SC_LARGE_MINCLASS); + bumped_usize = sz_s2u(SC_LARGE_MINCLASS); + ret = imalloc_no_sample(sopts, dopts, tsd, bumped_usize, + bumped_usize, ind_large); + if (unlikely(ret == NULL)) { + return NULL; + } + arena_prof_promote(tsd_tsdn(tsd), ret, usize); + } else { + ret = imalloc_no_sample(sopts, dopts, tsd, usize, usize, ind); + } + + return ret; +} + +/* + * Returns true if the allocation will overflow, and false otherwise. Sets + * *size to the product either way. + */ +JEMALLOC_ALWAYS_INLINE bool +compute_size_with_overflow(bool may_overflow, dynamic_opts_t *dopts, + size_t *size) { + /* + * This function is just num_items * item_size, except that we may have + * to check for overflow. + */ + + if (!may_overflow) { + assert(dopts->num_items == 1); + *size = dopts->item_size; + return false; + } + + /* A size_t with its high-half bits all set to 1. */ + static const size_t high_bits = SIZE_T_MAX << (sizeof(size_t) * 8 / 2); + + *size = dopts->item_size * dopts->num_items; + + if (unlikely(*size == 0)) { + return (dopts->num_items != 0 && dopts->item_size != 0); + } + + /* + * We got a non-zero size, but we don't know if we overflowed to get + * there. To avoid having to do a divide, we'll be clever and note that + * if both A and B can be represented in N/2 bits, then their product + * can be represented in N bits (without the possibility of overflow). + */ + if (likely((high_bits & (dopts->num_items | dopts->item_size)) == 0)) { + return false; + } + if (likely(*size / dopts->item_size == dopts->num_items)) { + return false; + } + return true; +} + +JEMALLOC_ALWAYS_INLINE int +imalloc_body(static_opts_t *sopts, dynamic_opts_t *dopts, tsd_t *tsd) { + /* Where the actual allocated memory will live. */ + void *allocation = NULL; + /* Filled in by compute_size_with_overflow below. */ + size_t size = 0; + /* + * For unaligned allocations, we need only ind. For aligned + * allocations, or in case of stats or profiling we need usize. + * + * These are actually dead stores, in that their values are reset before + * any branch on their value is taken. Sometimes though, it's + * convenient to pass them as arguments before this point. To avoid + * undefined behavior then, we initialize them with dummy stores. + */ + szind_t ind = 0; + size_t usize = 0; + + /* Reentrancy is only checked on slow path. */ + int8_t reentrancy_level; + + /* Compute the amount of memory the user wants. */ + if (unlikely(compute_size_with_overflow(sopts->may_overflow, dopts, + &size))) { + goto label_oom; + } + + if (unlikely(dopts->alignment < sopts->min_alignment + || (dopts->alignment & (dopts->alignment - 1)) != 0)) { + goto label_invalid_alignment; + } + + /* This is the beginning of the "core" algorithm. */ + + if (dopts->alignment == 0) { + ind = sz_size2index(size); + if (unlikely(ind >= SC_NSIZES)) { + goto label_oom; + } + if (config_stats || (config_prof && opt_prof) || sopts->usize) { + usize = sz_index2size(ind); + dopts->usize = usize; + assert(usize > 0 && usize + <= SC_LARGE_MAXCLASS); + } + } else { + if (sopts->bump_empty_aligned_alloc) { + if (unlikely(size == 0)) { + size = 1; + } + } + usize = sz_sa2u(size, dopts->alignment); + dopts->usize = usize; + if (unlikely(usize == 0 + || usize > SC_LARGE_MAXCLASS)) { + goto label_oom; + } + } + /* Validate the user input. */ + if (sopts->assert_nonempty_alloc) { + assert (size != 0); + } + + check_entry_exit_locking(tsd_tsdn(tsd)); + + /* + * If we need to handle reentrancy, we can do it out of a + * known-initialized arena (i.e. arena 0). + */ + reentrancy_level = tsd_reentrancy_level_get(tsd); + if (sopts->slow && unlikely(reentrancy_level > 0)) { + /* + * We should never specify particular arenas or tcaches from + * within our internal allocations. + */ + assert(dopts->tcache_ind == TCACHE_IND_AUTOMATIC || + dopts->tcache_ind == TCACHE_IND_NONE); + assert(dopts->arena_ind == ARENA_IND_AUTOMATIC); + dopts->tcache_ind = TCACHE_IND_NONE; + /* We know that arena 0 has already been initialized. */ + dopts->arena_ind = 0; + } + + /* If profiling is on, get our profiling context. */ + if (config_prof && opt_prof) { + /* + * Note that if we're going down this path, usize must have been + * initialized in the previous if statement. + */ + prof_tctx_t *tctx = prof_alloc_prep( + tsd, usize, prof_active_get_unlocked(), true); + + alloc_ctx_t alloc_ctx; + if (likely((uintptr_t)tctx == (uintptr_t)1U)) { + alloc_ctx.slab = (usize + <= SC_SMALL_MAXCLASS); + allocation = imalloc_no_sample( + sopts, dopts, tsd, usize, usize, ind); + } else if ((uintptr_t)tctx > (uintptr_t)1U) { + /* + * Note that ind might still be 0 here. This is fine; + * imalloc_sample ignores ind if dopts->alignment > 0. + */ + allocation = imalloc_sample( + sopts, dopts, tsd, usize, ind); + alloc_ctx.slab = false; + } else { + allocation = NULL; + } + + if (unlikely(allocation == NULL)) { + prof_alloc_rollback(tsd, tctx, true); + goto label_oom; + } + prof_malloc(tsd_tsdn(tsd), allocation, usize, &alloc_ctx, tctx); + } else { + /* + * If dopts->alignment > 0, then ind is still 0, but usize was + * computed in the previous if statement. Down the positive + * alignment path, imalloc_no_sample ignores ind and size + * (relying only on usize). + */ + allocation = imalloc_no_sample(sopts, dopts, tsd, size, usize, + ind); + if (unlikely(allocation == NULL)) { + goto label_oom; + } + } + + /* + * Allocation has been done at this point. We still have some + * post-allocation work to do though. + */ + assert(dopts->alignment == 0 + || ((uintptr_t)allocation & (dopts->alignment - 1)) == ZU(0)); + + if (config_stats) { + assert(usize == isalloc(tsd_tsdn(tsd), allocation)); + *tsd_thread_allocatedp_get(tsd) += usize; + } + + if (sopts->slow) { + UTRACE(0, size, allocation); + } + + /* Success! */ + check_entry_exit_locking(tsd_tsdn(tsd)); + *dopts->result = allocation; + return 0; + +label_oom: + if (unlikely(sopts->slow) && config_xmalloc && unlikely(opt_xmalloc)) { + malloc_write(sopts->oom_string); + abort(); + } + + if (sopts->slow) { + UTRACE(NULL, size, NULL); + } + + check_entry_exit_locking(tsd_tsdn(tsd)); + + if (sopts->set_errno_on_error) { + set_errno(ENOMEM); + } + + if (sopts->null_out_result_on_error) { + *dopts->result = NULL; + } + + return ENOMEM; + + /* + * This label is only jumped to by one goto; we move it out of line + * anyways to avoid obscuring the non-error paths, and for symmetry with + * the oom case. + */ +label_invalid_alignment: + if (config_xmalloc && unlikely(opt_xmalloc)) { + malloc_write(sopts->invalid_alignment_string); + abort(); + } + + if (sopts->set_errno_on_error) { + set_errno(EINVAL); + } + + if (sopts->slow) { + UTRACE(NULL, size, NULL); + } + + check_entry_exit_locking(tsd_tsdn(tsd)); + + if (sopts->null_out_result_on_error) { + *dopts->result = NULL; + } + + return EINVAL; +} + +JEMALLOC_ALWAYS_INLINE bool +imalloc_init_check(static_opts_t *sopts, dynamic_opts_t *dopts) { + if (unlikely(!malloc_initialized()) && unlikely(malloc_init())) { + if (config_xmalloc && unlikely(opt_xmalloc)) { + malloc_write(sopts->oom_string); + abort(); + } + UTRACE(NULL, dopts->num_items * dopts->item_size, NULL); + set_errno(ENOMEM); + *dopts->result = NULL; + + return false; + } + + return true; +} + +/* Returns the errno-style error code of the allocation. */ +JEMALLOC_ALWAYS_INLINE int +imalloc(static_opts_t *sopts, dynamic_opts_t *dopts) { + if (tsd_get_allocates() && !imalloc_init_check(sopts, dopts)) { + return ENOMEM; + } + + /* We always need the tsd. Let's grab it right away. */ + tsd_t *tsd = tsd_fetch(); + assert(tsd); + if (likely(tsd_fast(tsd))) { + /* Fast and common path. */ + tsd_assert_fast(tsd); + sopts->slow = false; + return imalloc_body(sopts, dopts, tsd); + } else { + if (!tsd_get_allocates() && !imalloc_init_check(sopts, dopts)) { + return ENOMEM; + } + + sopts->slow = true; + return imalloc_body(sopts, dopts, tsd); + } +} + +JEMALLOC_NOINLINE +void * +malloc_default(size_t size) { + void *ret; + static_opts_t sopts; + dynamic_opts_t dopts; + + LOG("core.malloc.entry", "size: %zu", size); + + static_opts_init(&sopts); + dynamic_opts_init(&dopts); + + sopts.null_out_result_on_error = true; + sopts.set_errno_on_error = true; + sopts.oom_string = "<jemalloc>: Error in malloc(): out of memory\n"; + + dopts.result = &ret; + dopts.num_items = 1; + dopts.item_size = size; + + imalloc(&sopts, &dopts); + /* + * Note that this branch gets optimized away -- it immediately follows + * the check on tsd_fast that sets sopts.slow. + */ + if (sopts.slow) { + uintptr_t args[3] = {size}; + hook_invoke_alloc(hook_alloc_malloc, ret, (uintptr_t)ret, args); + } + + LOG("core.malloc.exit", "result: %p", ret); + + return ret; +} + +/******************************************************************************/ +/* + * Begin malloc(3)-compatible functions. + */ + +/* + * malloc() fastpath. + * + * Fastpath assumes size <= SC_LOOKUP_MAXCLASS, and that we hit + * tcache. If either of these is false, we tail-call to the slowpath, + * malloc_default(). Tail-calling is used to avoid any caller-saved + * registers. + * + * fastpath supports ticker and profiling, both of which will also + * tail-call to the slowpath if they fire. + */ +JEMALLOC_EXPORT JEMALLOC_ALLOCATOR JEMALLOC_RESTRICT_RETURN +void JEMALLOC_NOTHROW * +JEMALLOC_ATTR(malloc) JEMALLOC_ALLOC_SIZE(1) +je_malloc(size_t size) { + LOG("core.malloc.entry", "size: %zu", size); + + if (tsd_get_allocates() && unlikely(!malloc_initialized())) { + return malloc_default(size); + } + + tsd_t *tsd = tsd_get(false); + if (unlikely(!tsd || !tsd_fast(tsd) || (size > SC_LOOKUP_MAXCLASS))) { + return malloc_default(size); + } + + tcache_t *tcache = tsd_tcachep_get(tsd); + + if (unlikely(ticker_trytick(&tcache->gc_ticker))) { + return malloc_default(size); + } + + szind_t ind = sz_size2index_lookup(size); + size_t usize; + if (config_stats || config_prof) { + usize = sz_index2size(ind); + } + /* Fast path relies on size being a bin. I.e. SC_LOOKUP_MAXCLASS < SC_SMALL_MAXCLASS */ + assert(ind < SC_NBINS); + assert(size <= SC_SMALL_MAXCLASS); + + if (config_prof) { + int64_t bytes_until_sample = tsd_bytes_until_sample_get(tsd); + bytes_until_sample -= usize; + tsd_bytes_until_sample_set(tsd, bytes_until_sample); + + if (unlikely(bytes_until_sample < 0)) { + /* + * Avoid a prof_active check on the fastpath. + * If prof_active is false, set bytes_until_sample to + * a large value. If prof_active is set to true, + * bytes_until_sample will be reset. + */ + if (!prof_active) { + tsd_bytes_until_sample_set(tsd, SSIZE_MAX); + } + return malloc_default(size); + } + } + + cache_bin_t *bin = tcache_small_bin_get(tcache, ind); + bool tcache_success; + void* ret = cache_bin_alloc_easy(bin, &tcache_success); + + if (tcache_success) { + if (config_stats) { + *tsd_thread_allocatedp_get(tsd) += usize; + bin->tstats.nrequests++; + } + if (config_prof) { + tcache->prof_accumbytes += usize; + } + + LOG("core.malloc.exit", "result: %p", ret); + + /* Fastpath success */ + return ret; + } + + return malloc_default(size); +} + +JEMALLOC_EXPORT int JEMALLOC_NOTHROW +JEMALLOC_ATTR(nonnull(1)) +je_posix_memalign(void **memptr, size_t alignment, size_t size) { + int ret; + static_opts_t sopts; + dynamic_opts_t dopts; + + LOG("core.posix_memalign.entry", "mem ptr: %p, alignment: %zu, " + "size: %zu", memptr, alignment, size); + + static_opts_init(&sopts); + dynamic_opts_init(&dopts); + + sopts.bump_empty_aligned_alloc = true; + sopts.min_alignment = sizeof(void *); + sopts.oom_string = + "<jemalloc>: Error allocating aligned memory: out of memory\n"; + sopts.invalid_alignment_string = + "<jemalloc>: Error allocating aligned memory: invalid alignment\n"; + + dopts.result = memptr; + dopts.num_items = 1; + dopts.item_size = size; + dopts.alignment = alignment; + + ret = imalloc(&sopts, &dopts); + if (sopts.slow) { + uintptr_t args[3] = {(uintptr_t)memptr, (uintptr_t)alignment, + (uintptr_t)size}; + hook_invoke_alloc(hook_alloc_posix_memalign, *memptr, + (uintptr_t)ret, args); + } + + LOG("core.posix_memalign.exit", "result: %d, alloc ptr: %p", ret, + *memptr); + + return ret; +} + +JEMALLOC_EXPORT JEMALLOC_ALLOCATOR JEMALLOC_RESTRICT_RETURN +void JEMALLOC_NOTHROW * +JEMALLOC_ATTR(malloc) JEMALLOC_ALLOC_SIZE(2) +je_aligned_alloc(size_t alignment, size_t size) { + void *ret; + + static_opts_t sopts; + dynamic_opts_t dopts; + + LOG("core.aligned_alloc.entry", "alignment: %zu, size: %zu\n", + alignment, size); + + static_opts_init(&sopts); + dynamic_opts_init(&dopts); + + sopts.bump_empty_aligned_alloc = true; + sopts.null_out_result_on_error = true; + sopts.set_errno_on_error = true; + sopts.min_alignment = 1; + sopts.oom_string = + "<jemalloc>: Error allocating aligned memory: out of memory\n"; + sopts.invalid_alignment_string = + "<jemalloc>: Error allocating aligned memory: invalid alignment\n"; + + dopts.result = &ret; + dopts.num_items = 1; + dopts.item_size = size; + dopts.alignment = alignment; + + imalloc(&sopts, &dopts); + if (sopts.slow) { + uintptr_t args[3] = {(uintptr_t)alignment, (uintptr_t)size}; + hook_invoke_alloc(hook_alloc_aligned_alloc, ret, + (uintptr_t)ret, args); + } + + LOG("core.aligned_alloc.exit", "result: %p", ret); + + return ret; +} + +JEMALLOC_EXPORT JEMALLOC_ALLOCATOR JEMALLOC_RESTRICT_RETURN +void JEMALLOC_NOTHROW * +JEMALLOC_ATTR(malloc) JEMALLOC_ALLOC_SIZE2(1, 2) +je_calloc(size_t num, size_t size) { + void *ret; + static_opts_t sopts; + dynamic_opts_t dopts; + + LOG("core.calloc.entry", "num: %zu, size: %zu\n", num, size); + + static_opts_init(&sopts); + dynamic_opts_init(&dopts); + + sopts.may_overflow = true; + sopts.null_out_result_on_error = true; + sopts.set_errno_on_error = true; + sopts.oom_string = "<jemalloc>: Error in calloc(): out of memory\n"; + + dopts.result = &ret; + dopts.num_items = num; + dopts.item_size = size; + dopts.zero = true; + + imalloc(&sopts, &dopts); + if (sopts.slow) { + uintptr_t args[3] = {(uintptr_t)num, (uintptr_t)size}; + hook_invoke_alloc(hook_alloc_calloc, ret, (uintptr_t)ret, args); + } + + LOG("core.calloc.exit", "result: %p", ret); + + return ret; +} + +static void * +irealloc_prof_sample(tsd_t *tsd, void *old_ptr, size_t old_usize, size_t usize, + prof_tctx_t *tctx, hook_ralloc_args_t *hook_args) { + void *p; + + if (tctx == NULL) { + return NULL; + } + if (usize <= SC_SMALL_MAXCLASS) { + p = iralloc(tsd, old_ptr, old_usize, + SC_LARGE_MINCLASS, 0, false, hook_args); + if (p == NULL) { + return NULL; + } + arena_prof_promote(tsd_tsdn(tsd), p, usize); + } else { + p = iralloc(tsd, old_ptr, old_usize, usize, 0, false, + hook_args); + } + + return p; +} + +JEMALLOC_ALWAYS_INLINE void * +irealloc_prof(tsd_t *tsd, void *old_ptr, size_t old_usize, size_t usize, + alloc_ctx_t *alloc_ctx, hook_ralloc_args_t *hook_args) { + void *p; + bool prof_active; + prof_tctx_t *old_tctx, *tctx; + + prof_active = prof_active_get_unlocked(); + old_tctx = prof_tctx_get(tsd_tsdn(tsd), old_ptr, alloc_ctx); + tctx = prof_alloc_prep(tsd, usize, prof_active, true); + if (unlikely((uintptr_t)tctx != (uintptr_t)1U)) { + p = irealloc_prof_sample(tsd, old_ptr, old_usize, usize, tctx, + hook_args); + } else { + p = iralloc(tsd, old_ptr, old_usize, usize, 0, false, + hook_args); + } + if (unlikely(p == NULL)) { + prof_alloc_rollback(tsd, tctx, true); + return NULL; + } + prof_realloc(tsd, p, usize, tctx, prof_active, true, old_ptr, old_usize, + old_tctx); + + return p; +} + +JEMALLOC_ALWAYS_INLINE void +ifree(tsd_t *tsd, void *ptr, tcache_t *tcache, bool slow_path) { + if (!slow_path) { + tsd_assert_fast(tsd); + } + check_entry_exit_locking(tsd_tsdn(tsd)); + if (tsd_reentrancy_level_get(tsd) != 0) { + assert(slow_path); + } + + assert(ptr != NULL); + assert(malloc_initialized() || IS_INITIALIZER); + + alloc_ctx_t alloc_ctx; + rtree_ctx_t *rtree_ctx = tsd_rtree_ctx(tsd); + rtree_szind_slab_read(tsd_tsdn(tsd), &extents_rtree, rtree_ctx, + (uintptr_t)ptr, true, &alloc_ctx.szind, &alloc_ctx.slab); + assert(alloc_ctx.szind != SC_NSIZES); + + size_t usize; + if (config_prof && opt_prof) { + usize = sz_index2size(alloc_ctx.szind); + prof_free(tsd, ptr, usize, &alloc_ctx); + } else if (config_stats) { + usize = sz_index2size(alloc_ctx.szind); + } + if (config_stats) { + *tsd_thread_deallocatedp_get(tsd) += usize; + } + + if (likely(!slow_path)) { + idalloctm(tsd_tsdn(tsd), ptr, tcache, &alloc_ctx, false, + false); + } else { + idalloctm(tsd_tsdn(tsd), ptr, tcache, &alloc_ctx, false, + true); + } +} + +JEMALLOC_ALWAYS_INLINE void +isfree(tsd_t *tsd, void *ptr, size_t usize, tcache_t *tcache, bool slow_path) { + if (!slow_path) { + tsd_assert_fast(tsd); + } + check_entry_exit_locking(tsd_tsdn(tsd)); + if (tsd_reentrancy_level_get(tsd) != 0) { + assert(slow_path); + } + + assert(ptr != NULL); + assert(malloc_initialized() || IS_INITIALIZER); + + alloc_ctx_t alloc_ctx, *ctx; + if (!config_cache_oblivious && ((uintptr_t)ptr & PAGE_MASK) != 0) { + /* + * When cache_oblivious is disabled and ptr is not page aligned, + * the allocation was not sampled -- usize can be used to + * determine szind directly. + */ + alloc_ctx.szind = sz_size2index(usize); + alloc_ctx.slab = true; + ctx = &alloc_ctx; + if (config_debug) { + alloc_ctx_t dbg_ctx; + rtree_ctx_t *rtree_ctx = tsd_rtree_ctx(tsd); + rtree_szind_slab_read(tsd_tsdn(tsd), &extents_rtree, + rtree_ctx, (uintptr_t)ptr, true, &dbg_ctx.szind, + &dbg_ctx.slab); + assert(dbg_ctx.szind == alloc_ctx.szind); + assert(dbg_ctx.slab == alloc_ctx.slab); + } + } else if (config_prof && opt_prof) { + rtree_ctx_t *rtree_ctx = tsd_rtree_ctx(tsd); + rtree_szind_slab_read(tsd_tsdn(tsd), &extents_rtree, rtree_ctx, + (uintptr_t)ptr, true, &alloc_ctx.szind, &alloc_ctx.slab); + assert(alloc_ctx.szind == sz_size2index(usize)); + ctx = &alloc_ctx; + } else { + ctx = NULL; + } + + if (config_prof && opt_prof) { + prof_free(tsd, ptr, usize, ctx); + } + if (config_stats) { + *tsd_thread_deallocatedp_get(tsd) += usize; + } + + if (likely(!slow_path)) { + isdalloct(tsd_tsdn(tsd), ptr, usize, tcache, ctx, false); + } else { + isdalloct(tsd_tsdn(tsd), ptr, usize, tcache, ctx, true); + } +} + +JEMALLOC_EXPORT JEMALLOC_ALLOCATOR JEMALLOC_RESTRICT_RETURN +void JEMALLOC_NOTHROW * +JEMALLOC_ALLOC_SIZE(2) +je_realloc(void *ptr, size_t arg_size) { + void *ret; + tsdn_t *tsdn JEMALLOC_CC_SILENCE_INIT(NULL); + size_t usize JEMALLOC_CC_SILENCE_INIT(0); + size_t old_usize = 0; + size_t size = arg_size; + + LOG("core.realloc.entry", "ptr: %p, size: %zu\n", ptr, size); + + if (unlikely(size == 0)) { + if (ptr != NULL) { + /* realloc(ptr, 0) is equivalent to free(ptr). */ + UTRACE(ptr, 0, 0); + tcache_t *tcache; + tsd_t *tsd = tsd_fetch(); + if (tsd_reentrancy_level_get(tsd) == 0) { + tcache = tcache_get(tsd); + } else { + tcache = NULL; + } + + uintptr_t args[3] = {(uintptr_t)ptr, size}; + hook_invoke_dalloc(hook_dalloc_realloc, ptr, args); + + ifree(tsd, ptr, tcache, true); + + LOG("core.realloc.exit", "result: %p", NULL); + return NULL; + } + size = 1; + } + + if (likely(ptr != NULL)) { + assert(malloc_initialized() || IS_INITIALIZER); + tsd_t *tsd = tsd_fetch(); + + check_entry_exit_locking(tsd_tsdn(tsd)); + + + hook_ralloc_args_t hook_args = {true, {(uintptr_t)ptr, + (uintptr_t)arg_size, 0, 0}}; + + alloc_ctx_t alloc_ctx; + rtree_ctx_t *rtree_ctx = tsd_rtree_ctx(tsd); + rtree_szind_slab_read(tsd_tsdn(tsd), &extents_rtree, rtree_ctx, + (uintptr_t)ptr, true, &alloc_ctx.szind, &alloc_ctx.slab); + assert(alloc_ctx.szind != SC_NSIZES); + old_usize = sz_index2size(alloc_ctx.szind); + assert(old_usize == isalloc(tsd_tsdn(tsd), ptr)); + if (config_prof && opt_prof) { + usize = sz_s2u(size); + if (unlikely(usize == 0 + || usize > SC_LARGE_MAXCLASS)) { + ret = NULL; + } else { + ret = irealloc_prof(tsd, ptr, old_usize, usize, + &alloc_ctx, &hook_args); + } + } else { + if (config_stats) { + usize = sz_s2u(size); + } + ret = iralloc(tsd, ptr, old_usize, size, 0, false, + &hook_args); + } + tsdn = tsd_tsdn(tsd); + } else { + /* realloc(NULL, size) is equivalent to malloc(size). */ + static_opts_t sopts; + dynamic_opts_t dopts; + + static_opts_init(&sopts); + dynamic_opts_init(&dopts); + + sopts.null_out_result_on_error = true; + sopts.set_errno_on_error = true; + sopts.oom_string = + "<jemalloc>: Error in realloc(): out of memory\n"; + + dopts.result = &ret; + dopts.num_items = 1; + dopts.item_size = size; + + imalloc(&sopts, &dopts); + if (sopts.slow) { + uintptr_t args[3] = {(uintptr_t)ptr, arg_size}; + hook_invoke_alloc(hook_alloc_realloc, ret, + (uintptr_t)ret, args); + } + + return ret; + } + + if (unlikely(ret == NULL)) { + if (config_xmalloc && unlikely(opt_xmalloc)) { + malloc_write("<jemalloc>: Error in realloc(): " + "out of memory\n"); + abort(); + } + set_errno(ENOMEM); + } + if (config_stats && likely(ret != NULL)) { + tsd_t *tsd; + + assert(usize == isalloc(tsdn, ret)); + tsd = tsdn_tsd(tsdn); + *tsd_thread_allocatedp_get(tsd) += usize; + *tsd_thread_deallocatedp_get(tsd) += old_usize; + } + UTRACE(ptr, size, ret); + check_entry_exit_locking(tsdn); + + LOG("core.realloc.exit", "result: %p", ret); + return ret; +} + +JEMALLOC_NOINLINE +void +free_default(void *ptr) { + UTRACE(ptr, 0, 0); + if (likely(ptr != NULL)) { + /* + * We avoid setting up tsd fully (e.g. tcache, arena binding) + * based on only free() calls -- other activities trigger the + * minimal to full transition. This is because free() may + * happen during thread shutdown after tls deallocation: if a + * thread never had any malloc activities until then, a + * fully-setup tsd won't be destructed properly. + */ + tsd_t *tsd = tsd_fetch_min(); + check_entry_exit_locking(tsd_tsdn(tsd)); + + tcache_t *tcache; + if (likely(tsd_fast(tsd))) { + tsd_assert_fast(tsd); + /* Unconditionally get tcache ptr on fast path. */ + tcache = tsd_tcachep_get(tsd); + ifree(tsd, ptr, tcache, false); + } else { + if (likely(tsd_reentrancy_level_get(tsd) == 0)) { + tcache = tcache_get(tsd); + } else { + tcache = NULL; + } + uintptr_t args_raw[3] = {(uintptr_t)ptr}; + hook_invoke_dalloc(hook_dalloc_free, ptr, args_raw); + ifree(tsd, ptr, tcache, true); + } + check_entry_exit_locking(tsd_tsdn(tsd)); + } +} + +JEMALLOC_ALWAYS_INLINE +bool free_fastpath(void *ptr, size_t size, bool size_hint) { + tsd_t *tsd = tsd_get(false); + if (unlikely(!tsd || !tsd_fast(tsd))) { + return false; + } + + tcache_t *tcache = tsd_tcachep_get(tsd); + + alloc_ctx_t alloc_ctx; + /* + * If !config_cache_oblivious, we can check PAGE alignment to + * detect sampled objects. Otherwise addresses are + * randomized, and we have to look it up in the rtree anyway. + * See also isfree(). + */ + if (!size_hint || config_cache_oblivious) { + rtree_ctx_t *rtree_ctx = tsd_rtree_ctx(tsd); + bool res = rtree_szind_slab_read_fast(tsd_tsdn(tsd), &extents_rtree, + rtree_ctx, (uintptr_t)ptr, + &alloc_ctx.szind, &alloc_ctx.slab); + + /* Note: profiled objects will have alloc_ctx.slab set */ + if (!res || !alloc_ctx.slab) { + return false; + } + assert(alloc_ctx.szind != SC_NSIZES); + } else { + /* + * Check for both sizes that are too large, and for sampled objects. + * Sampled objects are always page-aligned. The sampled object check + * will also check for null ptr. + */ + if (size > SC_LOOKUP_MAXCLASS || (((uintptr_t)ptr & PAGE_MASK) == 0)) { + return false; + } + alloc_ctx.szind = sz_size2index_lookup(size); + } + + if (unlikely(ticker_trytick(&tcache->gc_ticker))) { + return false; + } + + cache_bin_t *bin = tcache_small_bin_get(tcache, alloc_ctx.szind); + cache_bin_info_t *bin_info = &tcache_bin_info[alloc_ctx.szind]; + if (!cache_bin_dalloc_easy(bin, bin_info, ptr)) { + return false; + } + + if (config_stats) { + size_t usize = sz_index2size(alloc_ctx.szind); + *tsd_thread_deallocatedp_get(tsd) += usize; + } + + return true; +} + +JEMALLOC_EXPORT void JEMALLOC_NOTHROW +je_free(void *ptr) { + LOG("core.free.entry", "ptr: %p", ptr); + + if (!free_fastpath(ptr, 0, false)) { + free_default(ptr); + } + + LOG("core.free.exit", ""); +} + +/* + * End malloc(3)-compatible functions. + */ +/******************************************************************************/ +/* + * Begin non-standard override functions. + */ + +#ifdef JEMALLOC_OVERRIDE_MEMALIGN +JEMALLOC_EXPORT JEMALLOC_ALLOCATOR JEMALLOC_RESTRICT_RETURN +void JEMALLOC_NOTHROW * +JEMALLOC_ATTR(malloc) +je_memalign(size_t alignment, size_t size) { + void *ret; + static_opts_t sopts; + dynamic_opts_t dopts; + + LOG("core.memalign.entry", "alignment: %zu, size: %zu\n", alignment, + size); + + static_opts_init(&sopts); + dynamic_opts_init(&dopts); + + sopts.min_alignment = 1; + sopts.oom_string = + "<jemalloc>: Error allocating aligned memory: out of memory\n"; + sopts.invalid_alignment_string = + "<jemalloc>: Error allocating aligned memory: invalid alignment\n"; + sopts.null_out_result_on_error = true; + + dopts.result = &ret; + dopts.num_items = 1; + dopts.item_size = size; + dopts.alignment = alignment; + + imalloc(&sopts, &dopts); + if (sopts.slow) { + uintptr_t args[3] = {alignment, size}; + hook_invoke_alloc(hook_alloc_memalign, ret, (uintptr_t)ret, + args); + } + + LOG("core.memalign.exit", "result: %p", ret); + return ret; +} +#endif + +#ifdef JEMALLOC_OVERRIDE_VALLOC +JEMALLOC_EXPORT JEMALLOC_ALLOCATOR JEMALLOC_RESTRICT_RETURN +void JEMALLOC_NOTHROW * +JEMALLOC_ATTR(malloc) +je_valloc(size_t size) { + void *ret; + + static_opts_t sopts; + dynamic_opts_t dopts; + + LOG("core.valloc.entry", "size: %zu\n", size); + + static_opts_init(&sopts); + dynamic_opts_init(&dopts); + + sopts.null_out_result_on_error = true; + sopts.min_alignment = PAGE; + sopts.oom_string = + "<jemalloc>: Error allocating aligned memory: out of memory\n"; + sopts.invalid_alignment_string = + "<jemalloc>: Error allocating aligned memory: invalid alignment\n"; + + dopts.result = &ret; + dopts.num_items = 1; + dopts.item_size = size; + dopts.alignment = PAGE; + + imalloc(&sopts, &dopts); + if (sopts.slow) { + uintptr_t args[3] = {size}; + hook_invoke_alloc(hook_alloc_valloc, ret, (uintptr_t)ret, args); + } + + LOG("core.valloc.exit", "result: %p\n", ret); + return ret; +} +#endif + +#if defined(JEMALLOC_IS_MALLOC) && defined(JEMALLOC_GLIBC_MALLOC_HOOK) +/* + * glibc provides the RTLD_DEEPBIND flag for dlopen which can make it possible + * to inconsistently reference libc's malloc(3)-compatible functions + * (https://bugzilla.mozilla.org/show_bug.cgi?id=493541). + * + * These definitions interpose hooks in glibc. The functions are actually + * passed an extra argument for the caller return address, which will be + * ignored. + */ +JEMALLOC_EXPORT void (*__free_hook)(void *ptr) = je_free; +JEMALLOC_EXPORT void *(*__malloc_hook)(size_t size) = je_malloc; +JEMALLOC_EXPORT void *(*__realloc_hook)(void *ptr, size_t size) = je_realloc; +# ifdef JEMALLOC_GLIBC_MEMALIGN_HOOK +JEMALLOC_EXPORT void *(*__memalign_hook)(size_t alignment, size_t size) = + je_memalign; +# endif + +# ifdef CPU_COUNT +/* + * To enable static linking with glibc, the libc specific malloc interface must + * be implemented also, so none of glibc's malloc.o functions are added to the + * link. + */ +# define ALIAS(je_fn) __attribute__((alias (#je_fn), used)) +/* To force macro expansion of je_ prefix before stringification. */ +# define PREALIAS(je_fn) ALIAS(je_fn) +# ifdef JEMALLOC_OVERRIDE___LIBC_CALLOC +void *__libc_calloc(size_t n, size_t size) PREALIAS(je_calloc); +# endif +# ifdef JEMALLOC_OVERRIDE___LIBC_FREE +void __libc_free(void* ptr) PREALIAS(je_free); +# endif +# ifdef JEMALLOC_OVERRIDE___LIBC_MALLOC +void *__libc_malloc(size_t size) PREALIAS(je_malloc); +# endif +# ifdef JEMALLOC_OVERRIDE___LIBC_MEMALIGN +void *__libc_memalign(size_t align, size_t s) PREALIAS(je_memalign); +# endif +# ifdef JEMALLOC_OVERRIDE___LIBC_REALLOC +void *__libc_realloc(void* ptr, size_t size) PREALIAS(je_realloc); +# endif +# ifdef JEMALLOC_OVERRIDE___LIBC_VALLOC +void *__libc_valloc(size_t size) PREALIAS(je_valloc); +# endif +# ifdef JEMALLOC_OVERRIDE___POSIX_MEMALIGN +int __posix_memalign(void** r, size_t a, size_t s) PREALIAS(je_posix_memalign); +# endif +# undef PREALIAS +# undef ALIAS +# endif +#endif + +/* + * End non-standard override functions. + */ +/******************************************************************************/ +/* + * Begin non-standard functions. + */ + +#ifdef JEMALLOC_EXPERIMENTAL_SMALLOCX_API + +#define JEMALLOC_SMALLOCX_CONCAT_HELPER(x, y) x ## y +#define JEMALLOC_SMALLOCX_CONCAT_HELPER2(x, y) \ + JEMALLOC_SMALLOCX_CONCAT_HELPER(x, y) + +typedef struct { + void *ptr; + size_t size; +} smallocx_return_t; + +JEMALLOC_EXPORT JEMALLOC_ALLOCATOR JEMALLOC_RESTRICT_RETURN +smallocx_return_t JEMALLOC_NOTHROW +/* + * The attribute JEMALLOC_ATTR(malloc) cannot be used due to: + * - https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86488 + */ +JEMALLOC_SMALLOCX_CONCAT_HELPER2(je_smallocx_, JEMALLOC_VERSION_GID_IDENT) + (size_t size, int flags) { + /* + * Note: the attribute JEMALLOC_ALLOC_SIZE(1) cannot be + * used here because it makes writing beyond the `size` + * of the `ptr` undefined behavior, but the objective + * of this function is to allow writing beyond `size` + * up to `smallocx_return_t::size`. + */ + smallocx_return_t ret; + static_opts_t sopts; + dynamic_opts_t dopts; + + LOG("core.smallocx.entry", "size: %zu, flags: %d", size, flags); + + static_opts_init(&sopts); + dynamic_opts_init(&dopts); + + sopts.assert_nonempty_alloc = true; + sopts.null_out_result_on_error = true; + sopts.oom_string = "<jemalloc>: Error in mallocx(): out of memory\n"; + sopts.usize = true; + + dopts.result = &ret.ptr; + dopts.num_items = 1; + dopts.item_size = size; + if (unlikely(flags != 0)) { + if ((flags & MALLOCX_LG_ALIGN_MASK) != 0) { + dopts.alignment = MALLOCX_ALIGN_GET_SPECIFIED(flags); + } + + dopts.zero = MALLOCX_ZERO_GET(flags); + + if ((flags & MALLOCX_TCACHE_MASK) != 0) { + if ((flags & MALLOCX_TCACHE_MASK) + == MALLOCX_TCACHE_NONE) { + dopts.tcache_ind = TCACHE_IND_NONE; + } else { + dopts.tcache_ind = MALLOCX_TCACHE_GET(flags); + } + } else { + dopts.tcache_ind = TCACHE_IND_AUTOMATIC; + } + + if ((flags & MALLOCX_ARENA_MASK) != 0) + dopts.arena_ind = MALLOCX_ARENA_GET(flags); + } + + imalloc(&sopts, &dopts); + assert(dopts.usize == je_nallocx(size, flags)); + ret.size = dopts.usize; + + LOG("core.smallocx.exit", "result: %p, size: %zu", ret.ptr, ret.size); + return ret; +} +#undef JEMALLOC_SMALLOCX_CONCAT_HELPER +#undef JEMALLOC_SMALLOCX_CONCAT_HELPER2 +#endif + +JEMALLOC_EXPORT JEMALLOC_ALLOCATOR JEMALLOC_RESTRICT_RETURN +void JEMALLOC_NOTHROW * +JEMALLOC_ATTR(malloc) JEMALLOC_ALLOC_SIZE(1) +je_mallocx(size_t size, int flags) { + void *ret; + static_opts_t sopts; + dynamic_opts_t dopts; + + LOG("core.mallocx.entry", "size: %zu, flags: %d", size, flags); + + static_opts_init(&sopts); + dynamic_opts_init(&dopts); + + sopts.assert_nonempty_alloc = true; + sopts.null_out_result_on_error = true; + sopts.oom_string = "<jemalloc>: Error in mallocx(): out of memory\n"; + + dopts.result = &ret; + dopts.num_items = 1; + dopts.item_size = size; + if (unlikely(flags != 0)) { + if ((flags & MALLOCX_LG_ALIGN_MASK) != 0) { + dopts.alignment = MALLOCX_ALIGN_GET_SPECIFIED(flags); + } + + dopts.zero = MALLOCX_ZERO_GET(flags); + + if ((flags & MALLOCX_TCACHE_MASK) != 0) { + if ((flags & MALLOCX_TCACHE_MASK) + == MALLOCX_TCACHE_NONE) { + dopts.tcache_ind = TCACHE_IND_NONE; + } else { + dopts.tcache_ind = MALLOCX_TCACHE_GET(flags); + } + } else { + dopts.tcache_ind = TCACHE_IND_AUTOMATIC; + } + + if ((flags & MALLOCX_ARENA_MASK) != 0) + dopts.arena_ind = MALLOCX_ARENA_GET(flags); + } + + imalloc(&sopts, &dopts); + if (sopts.slow) { + uintptr_t args[3] = {size, flags}; + hook_invoke_alloc(hook_alloc_mallocx, ret, (uintptr_t)ret, + args); + } + + LOG("core.mallocx.exit", "result: %p", ret); + return ret; +} + +static void * +irallocx_prof_sample(tsdn_t *tsdn, void *old_ptr, size_t old_usize, + size_t usize, size_t alignment, bool zero, tcache_t *tcache, arena_t *arena, + prof_tctx_t *tctx, hook_ralloc_args_t *hook_args) { + void *p; + + if (tctx == NULL) { + return NULL; + } + if (usize <= SC_SMALL_MAXCLASS) { + p = iralloct(tsdn, old_ptr, old_usize, + SC_LARGE_MINCLASS, alignment, zero, tcache, + arena, hook_args); + if (p == NULL) { + return NULL; + } + arena_prof_promote(tsdn, p, usize); + } else { + p = iralloct(tsdn, old_ptr, old_usize, usize, alignment, zero, + tcache, arena, hook_args); + } + + return p; +} + +JEMALLOC_ALWAYS_INLINE void * +irallocx_prof(tsd_t *tsd, void *old_ptr, size_t old_usize, size_t size, + size_t alignment, size_t *usize, bool zero, tcache_t *tcache, + arena_t *arena, alloc_ctx_t *alloc_ctx, hook_ralloc_args_t *hook_args) { + void *p; + bool prof_active; + prof_tctx_t *old_tctx, *tctx; + + prof_active = prof_active_get_unlocked(); + old_tctx = prof_tctx_get(tsd_tsdn(tsd), old_ptr, alloc_ctx); + tctx = prof_alloc_prep(tsd, *usize, prof_active, false); + if (unlikely((uintptr_t)tctx != (uintptr_t)1U)) { + p = irallocx_prof_sample(tsd_tsdn(tsd), old_ptr, old_usize, + *usize, alignment, zero, tcache, arena, tctx, hook_args); + } else { + p = iralloct(tsd_tsdn(tsd), old_ptr, old_usize, size, alignment, + zero, tcache, arena, hook_args); + } + if (unlikely(p == NULL)) { + prof_alloc_rollback(tsd, tctx, false); + return NULL; + } + + if (p == old_ptr && alignment != 0) { + /* + * The allocation did not move, so it is possible that the size + * class is smaller than would guarantee the requested + * alignment, and that the alignment constraint was + * serendipitously satisfied. Additionally, old_usize may not + * be the same as the current usize because of in-place large + * reallocation. Therefore, query the actual value of usize. + */ + *usize = isalloc(tsd_tsdn(tsd), p); + } + prof_realloc(tsd, p, *usize, tctx, prof_active, false, old_ptr, + old_usize, old_tctx); + + return p; +} + +JEMALLOC_EXPORT JEMALLOC_ALLOCATOR JEMALLOC_RESTRICT_RETURN +void JEMALLOC_NOTHROW * +JEMALLOC_ALLOC_SIZE(2) +je_rallocx(void *ptr, size_t size, int flags) { + void *p; + tsd_t *tsd; + size_t usize; + size_t old_usize; + size_t alignment = MALLOCX_ALIGN_GET(flags); + bool zero = flags & MALLOCX_ZERO; + arena_t *arena; + tcache_t *tcache; + + LOG("core.rallocx.entry", "ptr: %p, size: %zu, flags: %d", ptr, + size, flags); + + + assert(ptr != NULL); + assert(size != 0); + assert(malloc_initialized() || IS_INITIALIZER); + tsd = tsd_fetch(); + check_entry_exit_locking(tsd_tsdn(tsd)); + + if (unlikely((flags & MALLOCX_ARENA_MASK) != 0)) { + unsigned arena_ind = MALLOCX_ARENA_GET(flags); + arena = arena_get(tsd_tsdn(tsd), arena_ind, true); + if (unlikely(arena == NULL)) { + goto label_oom; + } + } else { + arena = NULL; + } + + if (unlikely((flags & MALLOCX_TCACHE_MASK) != 0)) { + if ((flags & MALLOCX_TCACHE_MASK) == MALLOCX_TCACHE_NONE) { + tcache = NULL; + } else { + tcache = tcaches_get(tsd, MALLOCX_TCACHE_GET(flags)); + } + } else { + tcache = tcache_get(tsd); + } + + alloc_ctx_t alloc_ctx; + rtree_ctx_t *rtree_ctx = tsd_rtree_ctx(tsd); + rtree_szind_slab_read(tsd_tsdn(tsd), &extents_rtree, rtree_ctx, + (uintptr_t)ptr, true, &alloc_ctx.szind, &alloc_ctx.slab); + assert(alloc_ctx.szind != SC_NSIZES); + old_usize = sz_index2size(alloc_ctx.szind); + assert(old_usize == isalloc(tsd_tsdn(tsd), ptr)); + + hook_ralloc_args_t hook_args = {false, {(uintptr_t)ptr, size, flags, + 0}}; + if (config_prof && opt_prof) { + usize = (alignment == 0) ? + sz_s2u(size) : sz_sa2u(size, alignment); + if (unlikely(usize == 0 + || usize > SC_LARGE_MAXCLASS)) { + goto label_oom; + } + p = irallocx_prof(tsd, ptr, old_usize, size, alignment, &usize, + zero, tcache, arena, &alloc_ctx, &hook_args); + if (unlikely(p == NULL)) { + goto label_oom; + } + } else { + p = iralloct(tsd_tsdn(tsd), ptr, old_usize, size, alignment, + zero, tcache, arena, &hook_args); + if (unlikely(p == NULL)) { + goto label_oom; + } + if (config_stats) { + usize = isalloc(tsd_tsdn(tsd), p); + } + } + assert(alignment == 0 || ((uintptr_t)p & (alignment - 1)) == ZU(0)); + + if (config_stats) { + *tsd_thread_allocatedp_get(tsd) += usize; + *tsd_thread_deallocatedp_get(tsd) += old_usize; + } + UTRACE(ptr, size, p); + check_entry_exit_locking(tsd_tsdn(tsd)); + + LOG("core.rallocx.exit", "result: %p", p); + return p; +label_oom: + if (config_xmalloc && unlikely(opt_xmalloc)) { + malloc_write("<jemalloc>: Error in rallocx(): out of memory\n"); + abort(); + } + UTRACE(ptr, size, 0); + check_entry_exit_locking(tsd_tsdn(tsd)); + + LOG("core.rallocx.exit", "result: %p", NULL); + return NULL; +} + +JEMALLOC_ALWAYS_INLINE size_t +ixallocx_helper(tsdn_t *tsdn, void *ptr, size_t old_usize, size_t size, + size_t extra, size_t alignment, bool zero) { + size_t newsize; + + if (ixalloc(tsdn, ptr, old_usize, size, extra, alignment, zero, + &newsize)) { + return old_usize; + } + + return newsize; +} + +static size_t +ixallocx_prof_sample(tsdn_t *tsdn, void *ptr, size_t old_usize, size_t size, + size_t extra, size_t alignment, bool zero, prof_tctx_t *tctx) { + size_t usize; + + if (tctx == NULL) { + return old_usize; + } + usize = ixallocx_helper(tsdn, ptr, old_usize, size, extra, alignment, + zero); + + return usize; +} + +JEMALLOC_ALWAYS_INLINE size_t +ixallocx_prof(tsd_t *tsd, void *ptr, size_t old_usize, size_t size, + size_t extra, size_t alignment, bool zero, alloc_ctx_t *alloc_ctx) { + size_t usize_max, usize; + bool prof_active; + prof_tctx_t *old_tctx, *tctx; + + prof_active = prof_active_get_unlocked(); + old_tctx = prof_tctx_get(tsd_tsdn(tsd), ptr, alloc_ctx); + /* + * usize isn't knowable before ixalloc() returns when extra is non-zero. + * Therefore, compute its maximum possible value and use that in + * prof_alloc_prep() to decide whether to capture a backtrace. + * prof_realloc() will use the actual usize to decide whether to sample. + */ + if (alignment == 0) { + usize_max = sz_s2u(size+extra); + assert(usize_max > 0 + && usize_max <= SC_LARGE_MAXCLASS); + } else { + usize_max = sz_sa2u(size+extra, alignment); + if (unlikely(usize_max == 0 + || usize_max > SC_LARGE_MAXCLASS)) { + /* + * usize_max is out of range, and chances are that + * allocation will fail, but use the maximum possible + * value and carry on with prof_alloc_prep(), just in + * case allocation succeeds. + */ + usize_max = SC_LARGE_MAXCLASS; + } + } + tctx = prof_alloc_prep(tsd, usize_max, prof_active, false); + + if (unlikely((uintptr_t)tctx != (uintptr_t)1U)) { + usize = ixallocx_prof_sample(tsd_tsdn(tsd), ptr, old_usize, + size, extra, alignment, zero, tctx); + } else { + usize = ixallocx_helper(tsd_tsdn(tsd), ptr, old_usize, size, + extra, alignment, zero); + } + if (usize == old_usize) { + prof_alloc_rollback(tsd, tctx, false); + return usize; + } + prof_realloc(tsd, ptr, usize, tctx, prof_active, false, ptr, old_usize, + old_tctx); + + return usize; +} + +JEMALLOC_EXPORT size_t JEMALLOC_NOTHROW +je_xallocx(void *ptr, size_t size, size_t extra, int flags) { + tsd_t *tsd; + size_t usize, old_usize; + size_t alignment = MALLOCX_ALIGN_GET(flags); + bool zero = flags & MALLOCX_ZERO; + + LOG("core.xallocx.entry", "ptr: %p, size: %zu, extra: %zu, " + "flags: %d", ptr, size, extra, flags); + + assert(ptr != NULL); + assert(size != 0); + assert(SIZE_T_MAX - size >= extra); + assert(malloc_initialized() || IS_INITIALIZER); + tsd = tsd_fetch(); + check_entry_exit_locking(tsd_tsdn(tsd)); + + alloc_ctx_t alloc_ctx; + rtree_ctx_t *rtree_ctx = tsd_rtree_ctx(tsd); + rtree_szind_slab_read(tsd_tsdn(tsd), &extents_rtree, rtree_ctx, + (uintptr_t)ptr, true, &alloc_ctx.szind, &alloc_ctx.slab); + assert(alloc_ctx.szind != SC_NSIZES); + old_usize = sz_index2size(alloc_ctx.szind); + assert(old_usize == isalloc(tsd_tsdn(tsd), ptr)); + /* + * The API explicitly absolves itself of protecting against (size + + * extra) numerical overflow, but we may need to clamp extra to avoid + * exceeding SC_LARGE_MAXCLASS. + * + * Ordinarily, size limit checking is handled deeper down, but here we + * have to check as part of (size + extra) clamping, since we need the + * clamped value in the above helper functions. + */ + if (unlikely(size > SC_LARGE_MAXCLASS)) { + usize = old_usize; + goto label_not_resized; + } + if (unlikely(SC_LARGE_MAXCLASS - size < extra)) { + extra = SC_LARGE_MAXCLASS - size; + } + + if (config_prof && opt_prof) { + usize = ixallocx_prof(tsd, ptr, old_usize, size, extra, + alignment, zero, &alloc_ctx); + } else { + usize = ixallocx_helper(tsd_tsdn(tsd), ptr, old_usize, size, + extra, alignment, zero); + } + if (unlikely(usize == old_usize)) { + goto label_not_resized; + } + + if (config_stats) { + *tsd_thread_allocatedp_get(tsd) += usize; + *tsd_thread_deallocatedp_get(tsd) += old_usize; + } +label_not_resized: + if (unlikely(!tsd_fast(tsd))) { + uintptr_t args[4] = {(uintptr_t)ptr, size, extra, flags}; + hook_invoke_expand(hook_expand_xallocx, ptr, old_usize, + usize, (uintptr_t)usize, args); + } + + UTRACE(ptr, size, ptr); + check_entry_exit_locking(tsd_tsdn(tsd)); + + LOG("core.xallocx.exit", "result: %zu", usize); + return usize; +} + +JEMALLOC_EXPORT size_t JEMALLOC_NOTHROW +JEMALLOC_ATTR(pure) +je_sallocx(const void *ptr, int flags) { + size_t usize; + tsdn_t *tsdn; + + LOG("core.sallocx.entry", "ptr: %p, flags: %d", ptr, flags); + + assert(malloc_initialized() || IS_INITIALIZER); + assert(ptr != NULL); + + tsdn = tsdn_fetch(); + check_entry_exit_locking(tsdn); + + if (config_debug || force_ivsalloc) { + usize = ivsalloc(tsdn, ptr); + assert(force_ivsalloc || usize != 0); + } else { + usize = isalloc(tsdn, ptr); + } + + check_entry_exit_locking(tsdn); + + LOG("core.sallocx.exit", "result: %zu", usize); + return usize; +} + +JEMALLOC_EXPORT void JEMALLOC_NOTHROW +je_dallocx(void *ptr, int flags) { + LOG("core.dallocx.entry", "ptr: %p, flags: %d", ptr, flags); + + assert(ptr != NULL); + assert(malloc_initialized() || IS_INITIALIZER); + + tsd_t *tsd = tsd_fetch(); + bool fast = tsd_fast(tsd); + check_entry_exit_locking(tsd_tsdn(tsd)); + + tcache_t *tcache; + if (unlikely((flags & MALLOCX_TCACHE_MASK) != 0)) { + /* Not allowed to be reentrant and specify a custom tcache. */ + assert(tsd_reentrancy_level_get(tsd) == 0); + if ((flags & MALLOCX_TCACHE_MASK) == MALLOCX_TCACHE_NONE) { + tcache = NULL; + } else { + tcache = tcaches_get(tsd, MALLOCX_TCACHE_GET(flags)); + } + } else { + if (likely(fast)) { + tcache = tsd_tcachep_get(tsd); + assert(tcache == tcache_get(tsd)); + } else { + if (likely(tsd_reentrancy_level_get(tsd) == 0)) { + tcache = tcache_get(tsd); + } else { + tcache = NULL; + } + } + } + + UTRACE(ptr, 0, 0); + if (likely(fast)) { + tsd_assert_fast(tsd); + ifree(tsd, ptr, tcache, false); + } else { + uintptr_t args_raw[3] = {(uintptr_t)ptr, flags}; + hook_invoke_dalloc(hook_dalloc_dallocx, ptr, args_raw); + ifree(tsd, ptr, tcache, true); + } + check_entry_exit_locking(tsd_tsdn(tsd)); + + LOG("core.dallocx.exit", ""); +} + +JEMALLOC_ALWAYS_INLINE size_t +inallocx(tsdn_t *tsdn, size_t size, int flags) { + check_entry_exit_locking(tsdn); + + size_t usize; + if (likely((flags & MALLOCX_LG_ALIGN_MASK) == 0)) { + usize = sz_s2u(size); + } else { + usize = sz_sa2u(size, MALLOCX_ALIGN_GET_SPECIFIED(flags)); + } + check_entry_exit_locking(tsdn); + return usize; +} + +JEMALLOC_NOINLINE void +sdallocx_default(void *ptr, size_t size, int flags) { + assert(ptr != NULL); + assert(malloc_initialized() || IS_INITIALIZER); + + tsd_t *tsd = tsd_fetch(); + bool fast = tsd_fast(tsd); + size_t usize = inallocx(tsd_tsdn(tsd), size, flags); + assert(usize == isalloc(tsd_tsdn(tsd), ptr)); + check_entry_exit_locking(tsd_tsdn(tsd)); + + tcache_t *tcache; + if (unlikely((flags & MALLOCX_TCACHE_MASK) != 0)) { + /* Not allowed to be reentrant and specify a custom tcache. */ + assert(tsd_reentrancy_level_get(tsd) == 0); + if ((flags & MALLOCX_TCACHE_MASK) == MALLOCX_TCACHE_NONE) { + tcache = NULL; + } else { + tcache = tcaches_get(tsd, MALLOCX_TCACHE_GET(flags)); + } + } else { + if (likely(fast)) { + tcache = tsd_tcachep_get(tsd); + assert(tcache == tcache_get(tsd)); + } else { + if (likely(tsd_reentrancy_level_get(tsd) == 0)) { + tcache = tcache_get(tsd); + } else { + tcache = NULL; + } + } + } + + UTRACE(ptr, 0, 0); + if (likely(fast)) { + tsd_assert_fast(tsd); + isfree(tsd, ptr, usize, tcache, false); + } else { + uintptr_t args_raw[3] = {(uintptr_t)ptr, size, flags}; + hook_invoke_dalloc(hook_dalloc_sdallocx, ptr, args_raw); + isfree(tsd, ptr, usize, tcache, true); + } + check_entry_exit_locking(tsd_tsdn(tsd)); + +} + +JEMALLOC_EXPORT void JEMALLOC_NOTHROW +je_sdallocx(void *ptr, size_t size, int flags) { + LOG("core.sdallocx.entry", "ptr: %p, size: %zu, flags: %d", ptr, + size, flags); + + if (flags !=0 || !free_fastpath(ptr, size, true)) { + sdallocx_default(ptr, size, flags); + } + + LOG("core.sdallocx.exit", ""); +} + +void JEMALLOC_NOTHROW +je_sdallocx_noflags(void *ptr, size_t size) { + LOG("core.sdallocx.entry", "ptr: %p, size: %zu, flags: 0", ptr, + size); + + if (!free_fastpath(ptr, size, true)) { + sdallocx_default(ptr, size, 0); + } + + LOG("core.sdallocx.exit", ""); +} + +JEMALLOC_EXPORT size_t JEMALLOC_NOTHROW +JEMALLOC_ATTR(pure) +je_nallocx(size_t size, int flags) { + size_t usize; + tsdn_t *tsdn; + + assert(size != 0); + + if (unlikely(malloc_init())) { + LOG("core.nallocx.exit", "result: %zu", ZU(0)); + return 0; + } + + tsdn = tsdn_fetch(); + check_entry_exit_locking(tsdn); + + usize = inallocx(tsdn, size, flags); + if (unlikely(usize > SC_LARGE_MAXCLASS)) { + LOG("core.nallocx.exit", "result: %zu", ZU(0)); + return 0; + } + + check_entry_exit_locking(tsdn); + LOG("core.nallocx.exit", "result: %zu", usize); + return usize; +} + +JEMALLOC_EXPORT int JEMALLOC_NOTHROW +je_mallctl(const char *name, void *oldp, size_t *oldlenp, void *newp, + size_t newlen) { + int ret; + tsd_t *tsd; + + LOG("core.mallctl.entry", "name: %s", name); + + if (unlikely(malloc_init())) { + LOG("core.mallctl.exit", "result: %d", EAGAIN); + return EAGAIN; + } + + tsd = tsd_fetch(); + check_entry_exit_locking(tsd_tsdn(tsd)); + ret = ctl_byname(tsd, name, oldp, oldlenp, newp, newlen); + check_entry_exit_locking(tsd_tsdn(tsd)); + + LOG("core.mallctl.exit", "result: %d", ret); + return ret; +} + +JEMALLOC_EXPORT int JEMALLOC_NOTHROW +je_mallctlnametomib(const char *name, size_t *mibp, size_t *miblenp) { + int ret; + + LOG("core.mallctlnametomib.entry", "name: %s", name); + + if (unlikely(malloc_init())) { + LOG("core.mallctlnametomib.exit", "result: %d", EAGAIN); + return EAGAIN; + } + + tsd_t *tsd = tsd_fetch(); + check_entry_exit_locking(tsd_tsdn(tsd)); + ret = ctl_nametomib(tsd, name, mibp, miblenp); + check_entry_exit_locking(tsd_tsdn(tsd)); + + LOG("core.mallctlnametomib.exit", "result: %d", ret); + return ret; +} + +JEMALLOC_EXPORT int JEMALLOC_NOTHROW +je_mallctlbymib(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp, + void *newp, size_t newlen) { + int ret; + tsd_t *tsd; + + LOG("core.mallctlbymib.entry", ""); + + if (unlikely(malloc_init())) { + LOG("core.mallctlbymib.exit", "result: %d", EAGAIN); + return EAGAIN; + } + + tsd = tsd_fetch(); + check_entry_exit_locking(tsd_tsdn(tsd)); + ret = ctl_bymib(tsd, mib, miblen, oldp, oldlenp, newp, newlen); + check_entry_exit_locking(tsd_tsdn(tsd)); + LOG("core.mallctlbymib.exit", "result: %d", ret); + return ret; +} + +JEMALLOC_EXPORT void JEMALLOC_NOTHROW +je_malloc_stats_print(void (*write_cb)(void *, const char *), void *cbopaque, + const char *opts) { + tsdn_t *tsdn; + + LOG("core.malloc_stats_print.entry", ""); + + tsdn = tsdn_fetch(); + check_entry_exit_locking(tsdn); + stats_print(write_cb, cbopaque, opts); + check_entry_exit_locking(tsdn); + LOG("core.malloc_stats_print.exit", ""); +} + +JEMALLOC_EXPORT size_t JEMALLOC_NOTHROW +je_malloc_usable_size(JEMALLOC_USABLE_SIZE_CONST void *ptr) { + size_t ret; + tsdn_t *tsdn; + + LOG("core.malloc_usable_size.entry", "ptr: %p", ptr); + + assert(malloc_initialized() || IS_INITIALIZER); + + tsdn = tsdn_fetch(); + check_entry_exit_locking(tsdn); + + if (unlikely(ptr == NULL)) { + ret = 0; + } else { + if (config_debug || force_ivsalloc) { + ret = ivsalloc(tsdn, ptr); + assert(force_ivsalloc || ret != 0); + } else { + ret = isalloc(tsdn, ptr); + } + } + + check_entry_exit_locking(tsdn); + LOG("core.malloc_usable_size.exit", "result: %zu", ret); + return ret; +} + +/* + * End non-standard functions. + */ +/******************************************************************************/ +/* + * The following functions are used by threading libraries for protection of + * malloc during fork(). + */ + +/* + * If an application creates a thread before doing any allocation in the main + * thread, then calls fork(2) in the main thread followed by memory allocation + * in the child process, a race can occur that results in deadlock within the + * child: the main thread may have forked while the created thread had + * partially initialized the allocator. Ordinarily jemalloc prevents + * fork/malloc races via the following functions it registers during + * initialization using pthread_atfork(), but of course that does no good if + * the allocator isn't fully initialized at fork time. The following library + * constructor is a partial solution to this problem. It may still be possible + * to trigger the deadlock described above, but doing so would involve forking + * via a library constructor that runs before jemalloc's runs. + */ +#ifndef JEMALLOC_JET +JEMALLOC_ATTR(constructor) +static void +jemalloc_constructor(void) { + malloc_init(); +} +#endif + +#ifndef JEMALLOC_MUTEX_INIT_CB +void +jemalloc_prefork(void) +#else +JEMALLOC_EXPORT void +_malloc_prefork(void) +#endif +{ + tsd_t *tsd; + unsigned i, j, narenas; + arena_t *arena; + +#ifdef JEMALLOC_MUTEX_INIT_CB + if (!malloc_initialized()) { + return; + } +#endif + assert(malloc_initialized()); + + tsd = tsd_fetch(); + + narenas = narenas_total_get(); + + witness_prefork(tsd_witness_tsdp_get(tsd)); + /* Acquire all mutexes in a safe order. */ + ctl_prefork(tsd_tsdn(tsd)); + tcache_prefork(tsd_tsdn(tsd)); + malloc_mutex_prefork(tsd_tsdn(tsd), &arenas_lock); + if (have_background_thread) { + background_thread_prefork0(tsd_tsdn(tsd)); + } + prof_prefork0(tsd_tsdn(tsd)); + if (have_background_thread) { + background_thread_prefork1(tsd_tsdn(tsd)); + } + /* Break arena prefork into stages to preserve lock order. */ + for (i = 0; i < 8; i++) { + for (j = 0; j < narenas; j++) { + if ((arena = arena_get(tsd_tsdn(tsd), j, false)) != + NULL) { + switch (i) { + case 0: + arena_prefork0(tsd_tsdn(tsd), arena); + break; + case 1: + arena_prefork1(tsd_tsdn(tsd), arena); + break; + case 2: + arena_prefork2(tsd_tsdn(tsd), arena); + break; + case 3: + arena_prefork3(tsd_tsdn(tsd), arena); + break; + case 4: + arena_prefork4(tsd_tsdn(tsd), arena); + break; + case 5: + arena_prefork5(tsd_tsdn(tsd), arena); + break; + case 6: + arena_prefork6(tsd_tsdn(tsd), arena); + break; + case 7: + arena_prefork7(tsd_tsdn(tsd), arena); + break; + default: not_reached(); + } + } + } + } + prof_prefork1(tsd_tsdn(tsd)); + tsd_prefork(tsd); +} + +#ifndef JEMALLOC_MUTEX_INIT_CB +void +jemalloc_postfork_parent(void) +#else +JEMALLOC_EXPORT void +_malloc_postfork(void) +#endif +{ + tsd_t *tsd; + unsigned i, narenas; + +#ifdef JEMALLOC_MUTEX_INIT_CB + if (!malloc_initialized()) { + return; + } +#endif + assert(malloc_initialized()); + + tsd = tsd_fetch(); + + tsd_postfork_parent(tsd); + + witness_postfork_parent(tsd_witness_tsdp_get(tsd)); + /* Release all mutexes, now that fork() has completed. */ + for (i = 0, narenas = narenas_total_get(); i < narenas; i++) { + arena_t *arena; + + if ((arena = arena_get(tsd_tsdn(tsd), i, false)) != NULL) { + arena_postfork_parent(tsd_tsdn(tsd), arena); + } + } + prof_postfork_parent(tsd_tsdn(tsd)); + if (have_background_thread) { + background_thread_postfork_parent(tsd_tsdn(tsd)); + } + malloc_mutex_postfork_parent(tsd_tsdn(tsd), &arenas_lock); + tcache_postfork_parent(tsd_tsdn(tsd)); + ctl_postfork_parent(tsd_tsdn(tsd)); +} + +void +jemalloc_postfork_child(void) { + tsd_t *tsd; + unsigned i, narenas; + + assert(malloc_initialized()); + + tsd = tsd_fetch(); + + tsd_postfork_child(tsd); + + witness_postfork_child(tsd_witness_tsdp_get(tsd)); + /* Release all mutexes, now that fork() has completed. */ + for (i = 0, narenas = narenas_total_get(); i < narenas; i++) { + arena_t *arena; + + if ((arena = arena_get(tsd_tsdn(tsd), i, false)) != NULL) { + arena_postfork_child(tsd_tsdn(tsd), arena); + } + } + prof_postfork_child(tsd_tsdn(tsd)); + if (have_background_thread) { + background_thread_postfork_child(tsd_tsdn(tsd)); + } + malloc_mutex_postfork_child(tsd_tsdn(tsd), &arenas_lock); + tcache_postfork_child(tsd_tsdn(tsd)); + ctl_postfork_child(tsd_tsdn(tsd)); +} + +/******************************************************************************/ diff --git a/deps/jemalloc/src/jemalloc_cpp.cpp b/deps/jemalloc/src/jemalloc_cpp.cpp new file mode 100644 index 000000000..da0441a7c --- /dev/null +++ b/deps/jemalloc/src/jemalloc_cpp.cpp @@ -0,0 +1,141 @@ +#include <mutex> +#include <new> + +#define JEMALLOC_CPP_CPP_ +#ifdef __cplusplus +extern "C" { +#endif + +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#ifdef __cplusplus +} +#endif + +// All operators in this file are exported. + +// Possibly alias hidden versions of malloc and sdallocx to avoid an extra plt +// thunk? +// +// extern __typeof (sdallocx) sdallocx_int +// __attribute ((alias ("sdallocx"), +// visibility ("hidden"))); +// +// ... but it needs to work with jemalloc namespaces. + +void *operator new(std::size_t size); +void *operator new[](std::size_t size); +void *operator new(std::size_t size, const std::nothrow_t &) noexcept; +void *operator new[](std::size_t size, const std::nothrow_t &) noexcept; +void operator delete(void *ptr) noexcept; +void operator delete[](void *ptr) noexcept; +void operator delete(void *ptr, const std::nothrow_t &) noexcept; +void operator delete[](void *ptr, const std::nothrow_t &) noexcept; + +#if __cpp_sized_deallocation >= 201309 +/* C++14's sized-delete operators. */ +void operator delete(void *ptr, std::size_t size) noexcept; +void operator delete[](void *ptr, std::size_t size) noexcept; +#endif + +JEMALLOC_NOINLINE +static void * +handleOOM(std::size_t size, bool nothrow) { + void *ptr = nullptr; + + while (ptr == nullptr) { + std::new_handler handler; + // GCC-4.8 and clang 4.0 do not have std::get_new_handler. + { + static std::mutex mtx; + std::lock_guard<std::mutex> lock(mtx); + + handler = std::set_new_handler(nullptr); + std::set_new_handler(handler); + } + if (handler == nullptr) + break; + + try { + handler(); + } catch (const std::bad_alloc &) { + break; + } + + ptr = je_malloc(size); + } + + if (ptr == nullptr && !nothrow) + std::__throw_bad_alloc(); + return ptr; +} + +template <bool IsNoExcept> +JEMALLOC_ALWAYS_INLINE +void * +newImpl(std::size_t size) noexcept(IsNoExcept) { + void *ptr = je_malloc(size); + if (likely(ptr != nullptr)) + return ptr; + + return handleOOM(size, IsNoExcept); +} + +void * +operator new(std::size_t size) { + return newImpl<false>(size); +} + +void * +operator new[](std::size_t size) { + return newImpl<false>(size); +} + +void * +operator new(std::size_t size, const std::nothrow_t &) noexcept { + return newImpl<true>(size); +} + +void * +operator new[](std::size_t size, const std::nothrow_t &) noexcept { + return newImpl<true>(size); +} + +void +operator delete(void *ptr) noexcept { + je_free(ptr); +} + +void +operator delete[](void *ptr) noexcept { + je_free(ptr); +} + +void +operator delete(void *ptr, const std::nothrow_t &) noexcept { + je_free(ptr); +} + +void operator delete[](void *ptr, const std::nothrow_t &) noexcept { + je_free(ptr); +} + +#if __cpp_sized_deallocation >= 201309 + +void +operator delete(void *ptr, std::size_t size) noexcept { + if (unlikely(ptr == nullptr)) { + return; + } + je_sdallocx_noflags(ptr, size); +} + +void operator delete[](void *ptr, std::size_t size) noexcept { + if (unlikely(ptr == nullptr)) { + return; + } + je_sdallocx_noflags(ptr, size); +} + +#endif // __cpp_sized_deallocation diff --git a/deps/jemalloc/src/large.c b/deps/jemalloc/src/large.c new file mode 100644 index 000000000..8e7a781d3 --- /dev/null +++ b/deps/jemalloc/src/large.c @@ -0,0 +1,395 @@ +#define JEMALLOC_LARGE_C_ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/assert.h" +#include "jemalloc/internal/extent_mmap.h" +#include "jemalloc/internal/mutex.h" +#include "jemalloc/internal/rtree.h" +#include "jemalloc/internal/util.h" + +/******************************************************************************/ + +void * +large_malloc(tsdn_t *tsdn, arena_t *arena, size_t usize, bool zero) { + assert(usize == sz_s2u(usize)); + + return large_palloc(tsdn, arena, usize, CACHELINE, zero); +} + +void * +large_palloc(tsdn_t *tsdn, arena_t *arena, size_t usize, size_t alignment, + bool zero) { + size_t ausize; + extent_t *extent; + bool is_zeroed; + UNUSED bool idump JEMALLOC_CC_SILENCE_INIT(false); + + assert(!tsdn_null(tsdn) || arena != NULL); + + ausize = sz_sa2u(usize, alignment); + if (unlikely(ausize == 0 || ausize > SC_LARGE_MAXCLASS)) { + return NULL; + } + + if (config_fill && unlikely(opt_zero)) { + zero = true; + } + /* + * Copy zero into is_zeroed and pass the copy when allocating the + * extent, so that it is possible to make correct junk/zero fill + * decisions below, even if is_zeroed ends up true when zero is false. + */ + is_zeroed = zero; + if (likely(!tsdn_null(tsdn))) { + arena = arena_choose_maybe_huge(tsdn_tsd(tsdn), arena, usize); + } + if (unlikely(arena == NULL) || (extent = arena_extent_alloc_large(tsdn, + arena, usize, alignment, &is_zeroed)) == NULL) { + return NULL; + } + + /* See comments in arena_bin_slabs_full_insert(). */ + if (!arena_is_auto(arena)) { + /* Insert extent into large. */ + malloc_mutex_lock(tsdn, &arena->large_mtx); + extent_list_append(&arena->large, extent); + malloc_mutex_unlock(tsdn, &arena->large_mtx); + } + if (config_prof && arena_prof_accum(tsdn, arena, usize)) { + prof_idump(tsdn); + } + + if (zero) { + assert(is_zeroed); + } else if (config_fill && unlikely(opt_junk_alloc)) { + memset(extent_addr_get(extent), JEMALLOC_ALLOC_JUNK, + extent_usize_get(extent)); + } + + arena_decay_tick(tsdn, arena); + return extent_addr_get(extent); +} + +static void +large_dalloc_junk_impl(void *ptr, size_t size) { + memset(ptr, JEMALLOC_FREE_JUNK, size); +} +large_dalloc_junk_t *JET_MUTABLE large_dalloc_junk = large_dalloc_junk_impl; + +static void +large_dalloc_maybe_junk_impl(void *ptr, size_t size) { + if (config_fill && have_dss && unlikely(opt_junk_free)) { + /* + * Only bother junk filling if the extent isn't about to be + * unmapped. + */ + if (opt_retain || (have_dss && extent_in_dss(ptr))) { + large_dalloc_junk(ptr, size); + } + } +} +large_dalloc_maybe_junk_t *JET_MUTABLE large_dalloc_maybe_junk = + large_dalloc_maybe_junk_impl; + +static bool +large_ralloc_no_move_shrink(tsdn_t *tsdn, extent_t *extent, size_t usize) { + arena_t *arena = extent_arena_get(extent); + size_t oldusize = extent_usize_get(extent); + extent_hooks_t *extent_hooks = extent_hooks_get(arena); + size_t diff = extent_size_get(extent) - (usize + sz_large_pad); + + assert(oldusize > usize); + + if (extent_hooks->split == NULL) { + return true; + } + + /* Split excess pages. */ + if (diff != 0) { + extent_t *trail = extent_split_wrapper(tsdn, arena, + &extent_hooks, extent, usize + sz_large_pad, + sz_size2index(usize), false, diff, SC_NSIZES, false); + if (trail == NULL) { + return true; + } + + if (config_fill && unlikely(opt_junk_free)) { + large_dalloc_maybe_junk(extent_addr_get(trail), + extent_size_get(trail)); + } + + arena_extents_dirty_dalloc(tsdn, arena, &extent_hooks, trail); + } + + arena_extent_ralloc_large_shrink(tsdn, arena, extent, oldusize); + + return false; +} + +static bool +large_ralloc_no_move_expand(tsdn_t *tsdn, extent_t *extent, size_t usize, + bool zero) { + arena_t *arena = extent_arena_get(extent); + size_t oldusize = extent_usize_get(extent); + extent_hooks_t *extent_hooks = extent_hooks_get(arena); + size_t trailsize = usize - oldusize; + + if (extent_hooks->merge == NULL) { + return true; + } + + if (config_fill && unlikely(opt_zero)) { + zero = true; + } + /* + * Copy zero into is_zeroed_trail and pass the copy when allocating the + * extent, so that it is possible to make correct junk/zero fill + * decisions below, even if is_zeroed_trail ends up true when zero is + * false. + */ + bool is_zeroed_trail = zero; + bool commit = true; + extent_t *trail; + bool new_mapping; + if ((trail = extents_alloc(tsdn, arena, &extent_hooks, + &arena->extents_dirty, extent_past_get(extent), trailsize, 0, + CACHELINE, false, SC_NSIZES, &is_zeroed_trail, &commit)) != NULL + || (trail = extents_alloc(tsdn, arena, &extent_hooks, + &arena->extents_muzzy, extent_past_get(extent), trailsize, 0, + CACHELINE, false, SC_NSIZES, &is_zeroed_trail, &commit)) != NULL) { + if (config_stats) { + new_mapping = false; + } + } else { + if ((trail = extent_alloc_wrapper(tsdn, arena, &extent_hooks, + extent_past_get(extent), trailsize, 0, CACHELINE, false, + SC_NSIZES, &is_zeroed_trail, &commit)) == NULL) { + return true; + } + if (config_stats) { + new_mapping = true; + } + } + + if (extent_merge_wrapper(tsdn, arena, &extent_hooks, extent, trail)) { + extent_dalloc_wrapper(tsdn, arena, &extent_hooks, trail); + return true; + } + rtree_ctx_t rtree_ctx_fallback; + rtree_ctx_t *rtree_ctx = tsdn_rtree_ctx(tsdn, &rtree_ctx_fallback); + szind_t szind = sz_size2index(usize); + extent_szind_set(extent, szind); + rtree_szind_slab_update(tsdn, &extents_rtree, rtree_ctx, + (uintptr_t)extent_addr_get(extent), szind, false); + + if (config_stats && new_mapping) { + arena_stats_mapped_add(tsdn, &arena->stats, trailsize); + } + + if (zero) { + if (config_cache_oblivious) { + /* + * Zero the trailing bytes of the original allocation's + * last page, since they are in an indeterminate state. + * There will always be trailing bytes, because ptr's + * offset from the beginning of the extent is a multiple + * of CACHELINE in [0 .. PAGE). + */ + void *zbase = (void *) + ((uintptr_t)extent_addr_get(extent) + oldusize); + void *zpast = PAGE_ADDR2BASE((void *)((uintptr_t)zbase + + PAGE)); + size_t nzero = (uintptr_t)zpast - (uintptr_t)zbase; + assert(nzero > 0); + memset(zbase, 0, nzero); + } + assert(is_zeroed_trail); + } else if (config_fill && unlikely(opt_junk_alloc)) { + memset((void *)((uintptr_t)extent_addr_get(extent) + oldusize), + JEMALLOC_ALLOC_JUNK, usize - oldusize); + } + + arena_extent_ralloc_large_expand(tsdn, arena, extent, oldusize); + + return false; +} + +bool +large_ralloc_no_move(tsdn_t *tsdn, extent_t *extent, size_t usize_min, + size_t usize_max, bool zero) { + size_t oldusize = extent_usize_get(extent); + + /* The following should have been caught by callers. */ + assert(usize_min > 0 && usize_max <= SC_LARGE_MAXCLASS); + /* Both allocation sizes must be large to avoid a move. */ + assert(oldusize >= SC_LARGE_MINCLASS + && usize_max >= SC_LARGE_MINCLASS); + + if (usize_max > oldusize) { + /* Attempt to expand the allocation in-place. */ + if (!large_ralloc_no_move_expand(tsdn, extent, usize_max, + zero)) { + arena_decay_tick(tsdn, extent_arena_get(extent)); + return false; + } + /* Try again, this time with usize_min. */ + if (usize_min < usize_max && usize_min > oldusize && + large_ralloc_no_move_expand(tsdn, extent, usize_min, + zero)) { + arena_decay_tick(tsdn, extent_arena_get(extent)); + return false; + } + } + + /* + * Avoid moving the allocation if the existing extent size accommodates + * the new size. + */ + if (oldusize >= usize_min && oldusize <= usize_max) { + arena_decay_tick(tsdn, extent_arena_get(extent)); + return false; + } + + /* Attempt to shrink the allocation in-place. */ + if (oldusize > usize_max) { + if (!large_ralloc_no_move_shrink(tsdn, extent, usize_max)) { + arena_decay_tick(tsdn, extent_arena_get(extent)); + return false; + } + } + return true; +} + +static void * +large_ralloc_move_helper(tsdn_t *tsdn, arena_t *arena, size_t usize, + size_t alignment, bool zero) { + if (alignment <= CACHELINE) { + return large_malloc(tsdn, arena, usize, zero); + } + return large_palloc(tsdn, arena, usize, alignment, zero); +} + +void * +large_ralloc(tsdn_t *tsdn, arena_t *arena, void *ptr, size_t usize, + size_t alignment, bool zero, tcache_t *tcache, + hook_ralloc_args_t *hook_args) { + extent_t *extent = iealloc(tsdn, ptr); + + size_t oldusize = extent_usize_get(extent); + /* The following should have been caught by callers. */ + assert(usize > 0 && usize <= SC_LARGE_MAXCLASS); + /* Both allocation sizes must be large to avoid a move. */ + assert(oldusize >= SC_LARGE_MINCLASS + && usize >= SC_LARGE_MINCLASS); + + /* Try to avoid moving the allocation. */ + if (!large_ralloc_no_move(tsdn, extent, usize, usize, zero)) { + hook_invoke_expand(hook_args->is_realloc + ? hook_expand_realloc : hook_expand_rallocx, ptr, oldusize, + usize, (uintptr_t)ptr, hook_args->args); + return extent_addr_get(extent); + } + + /* + * usize and old size are different enough that we need to use a + * different size class. In that case, fall back to allocating new + * space and copying. + */ + void *ret = large_ralloc_move_helper(tsdn, arena, usize, alignment, + zero); + if (ret == NULL) { + return NULL; + } + + hook_invoke_alloc(hook_args->is_realloc + ? hook_alloc_realloc : hook_alloc_rallocx, ret, (uintptr_t)ret, + hook_args->args); + hook_invoke_dalloc(hook_args->is_realloc + ? hook_dalloc_realloc : hook_dalloc_rallocx, ptr, hook_args->args); + + size_t copysize = (usize < oldusize) ? usize : oldusize; + memcpy(ret, extent_addr_get(extent), copysize); + isdalloct(tsdn, extent_addr_get(extent), oldusize, tcache, NULL, true); + return ret; +} + +/* + * junked_locked indicates whether the extent's data have been junk-filled, and + * whether the arena's large_mtx is currently held. + */ +static void +large_dalloc_prep_impl(tsdn_t *tsdn, arena_t *arena, extent_t *extent, + bool junked_locked) { + if (!junked_locked) { + /* See comments in arena_bin_slabs_full_insert(). */ + if (!arena_is_auto(arena)) { + malloc_mutex_lock(tsdn, &arena->large_mtx); + extent_list_remove(&arena->large, extent); + malloc_mutex_unlock(tsdn, &arena->large_mtx); + } + large_dalloc_maybe_junk(extent_addr_get(extent), + extent_usize_get(extent)); + } else { + /* Only hold the large_mtx if necessary. */ + if (!arena_is_auto(arena)) { + malloc_mutex_assert_owner(tsdn, &arena->large_mtx); + extent_list_remove(&arena->large, extent); + } + } + arena_extent_dalloc_large_prep(tsdn, arena, extent); +} + +static void +large_dalloc_finish_impl(tsdn_t *tsdn, arena_t *arena, extent_t *extent) { + extent_hooks_t *extent_hooks = EXTENT_HOOKS_INITIALIZER; + arena_extents_dirty_dalloc(tsdn, arena, &extent_hooks, extent); +} + +void +large_dalloc_prep_junked_locked(tsdn_t *tsdn, extent_t *extent) { + large_dalloc_prep_impl(tsdn, extent_arena_get(extent), extent, true); +} + +void +large_dalloc_finish(tsdn_t *tsdn, extent_t *extent) { + large_dalloc_finish_impl(tsdn, extent_arena_get(extent), extent); +} + +void +large_dalloc(tsdn_t *tsdn, extent_t *extent) { + arena_t *arena = extent_arena_get(extent); + large_dalloc_prep_impl(tsdn, arena, extent, false); + large_dalloc_finish_impl(tsdn, arena, extent); + arena_decay_tick(tsdn, arena); +} + +size_t +large_salloc(tsdn_t *tsdn, const extent_t *extent) { + return extent_usize_get(extent); +} + +prof_tctx_t * +large_prof_tctx_get(tsdn_t *tsdn, const extent_t *extent) { + return extent_prof_tctx_get(extent); +} + +void +large_prof_tctx_set(tsdn_t *tsdn, extent_t *extent, prof_tctx_t *tctx) { + extent_prof_tctx_set(extent, tctx); +} + +void +large_prof_tctx_reset(tsdn_t *tsdn, extent_t *extent) { + large_prof_tctx_set(tsdn, extent, (prof_tctx_t *)(uintptr_t)1U); +} + +nstime_t +large_prof_alloc_time_get(const extent_t *extent) { + return extent_prof_alloc_time_get(extent); +} + +void +large_prof_alloc_time_set(extent_t *extent, nstime_t t) { + extent_prof_alloc_time_set(extent, t); +} diff --git a/deps/jemalloc/src/log.c b/deps/jemalloc/src/log.c new file mode 100644 index 000000000..778902fb9 --- /dev/null +++ b/deps/jemalloc/src/log.c @@ -0,0 +1,78 @@ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/log.h" + +char log_var_names[JEMALLOC_LOG_VAR_BUFSIZE]; +atomic_b_t log_init_done = ATOMIC_INIT(false); + +/* + * Returns true if we were able to pick out a segment. Fills in r_segment_end + * with a pointer to the first character after the end of the string. + */ +static const char * +log_var_extract_segment(const char* segment_begin) { + const char *end; + for (end = segment_begin; *end != '\0' && *end != '|'; end++) { + } + return end; +} + +static bool +log_var_matches_segment(const char *segment_begin, const char *segment_end, + const char *log_var_begin, const char *log_var_end) { + assert(segment_begin <= segment_end); + assert(log_var_begin < log_var_end); + + ptrdiff_t segment_len = segment_end - segment_begin; + ptrdiff_t log_var_len = log_var_end - log_var_begin; + /* The special '.' segment matches everything. */ + if (segment_len == 1 && *segment_begin == '.') { + return true; + } + if (segment_len == log_var_len) { + return strncmp(segment_begin, log_var_begin, segment_len) == 0; + } else if (segment_len < log_var_len) { + return strncmp(segment_begin, log_var_begin, segment_len) == 0 + && log_var_begin[segment_len] == '.'; + } else { + return false; + } +} + +unsigned +log_var_update_state(log_var_t *log_var) { + const char *log_var_begin = log_var->name; + const char *log_var_end = log_var->name + strlen(log_var->name); + + /* Pointer to one before the beginning of the current segment. */ + const char *segment_begin = log_var_names; + + /* + * If log_init done is false, we haven't parsed the malloc conf yet. To + * avoid log-spew, we default to not displaying anything. + */ + if (!atomic_load_b(&log_init_done, ATOMIC_ACQUIRE)) { + return LOG_INITIALIZED_NOT_ENABLED; + } + + while (true) { + const char *segment_end = log_var_extract_segment( + segment_begin); + assert(segment_end < log_var_names + JEMALLOC_LOG_VAR_BUFSIZE); + if (log_var_matches_segment(segment_begin, segment_end, + log_var_begin, log_var_end)) { + atomic_store_u(&log_var->state, LOG_ENABLED, + ATOMIC_RELAXED); + return LOG_ENABLED; + } + if (*segment_end == '\0') { + /* Hit the end of the segment string with no match. */ + atomic_store_u(&log_var->state, + LOG_INITIALIZED_NOT_ENABLED, ATOMIC_RELAXED); + return LOG_INITIALIZED_NOT_ENABLED; + } + /* Otherwise, skip the delimiter and continue. */ + segment_begin = segment_end + 1; + } +} diff --git a/deps/jemalloc/src/malloc_io.c b/deps/jemalloc/src/malloc_io.c new file mode 100644 index 000000000..d7cb0f528 --- /dev/null +++ b/deps/jemalloc/src/malloc_io.c @@ -0,0 +1,675 @@ +#define JEMALLOC_MALLOC_IO_C_ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/malloc_io.h" +#include "jemalloc/internal/util.h" + +#ifdef assert +# undef assert +#endif +#ifdef not_reached +# undef not_reached +#endif +#ifdef not_implemented +# undef not_implemented +#endif +#ifdef assert_not_implemented +# undef assert_not_implemented +#endif + +/* + * Define simple versions of assertion macros that won't recurse in case + * of assertion failures in malloc_*printf(). + */ +#define assert(e) do { \ + if (config_debug && !(e)) { \ + malloc_write("<jemalloc>: Failed assertion\n"); \ + abort(); \ + } \ +} while (0) + +#define not_reached() do { \ + if (config_debug) { \ + malloc_write("<jemalloc>: Unreachable code reached\n"); \ + abort(); \ + } \ + unreachable(); \ +} while (0) + +#define not_implemented() do { \ + if (config_debug) { \ + malloc_write("<jemalloc>: Not implemented\n"); \ + abort(); \ + } \ +} while (0) + +#define assert_not_implemented(e) do { \ + if (unlikely(config_debug && !(e))) { \ + not_implemented(); \ + } \ +} while (0) + +/******************************************************************************/ +/* Function prototypes for non-inline static functions. */ + +static void wrtmessage(void *cbopaque, const char *s); +#define U2S_BUFSIZE ((1U << (LG_SIZEOF_INTMAX_T + 3)) + 1) +static char *u2s(uintmax_t x, unsigned base, bool uppercase, char *s, + size_t *slen_p); +#define D2S_BUFSIZE (1 + U2S_BUFSIZE) +static char *d2s(intmax_t x, char sign, char *s, size_t *slen_p); +#define O2S_BUFSIZE (1 + U2S_BUFSIZE) +static char *o2s(uintmax_t x, bool alt_form, char *s, size_t *slen_p); +#define X2S_BUFSIZE (2 + U2S_BUFSIZE) +static char *x2s(uintmax_t x, bool alt_form, bool uppercase, char *s, + size_t *slen_p); + +/******************************************************************************/ + +/* malloc_message() setup. */ +static void +wrtmessage(void *cbopaque, const char *s) { + malloc_write_fd(STDERR_FILENO, s, strlen(s)); +} + +JEMALLOC_EXPORT void (*je_malloc_message)(void *, const char *s); + +/* + * Wrapper around malloc_message() that avoids the need for + * je_malloc_message(...) throughout the code. + */ +void +malloc_write(const char *s) { + if (je_malloc_message != NULL) { + je_malloc_message(NULL, s); + } else { + wrtmessage(NULL, s); + } +} + +/* + * glibc provides a non-standard strerror_r() when _GNU_SOURCE is defined, so + * provide a wrapper. + */ +int +buferror(int err, char *buf, size_t buflen) { +#ifdef _WIN32 + FormatMessageA(FORMAT_MESSAGE_FROM_SYSTEM, NULL, err, 0, + (LPSTR)buf, (DWORD)buflen, NULL); + return 0; +#elif defined(JEMALLOC_STRERROR_R_RETURNS_CHAR_WITH_GNU_SOURCE) && defined(_GNU_SOURCE) + char *b = strerror_r(err, buf, buflen); + if (b != buf) { + strncpy(buf, b, buflen); + buf[buflen-1] = '\0'; + } + return 0; +#else + return strerror_r(err, buf, buflen); +#endif +} + +uintmax_t +malloc_strtoumax(const char *restrict nptr, char **restrict endptr, int base) { + uintmax_t ret, digit; + unsigned b; + bool neg; + const char *p, *ns; + + p = nptr; + if (base < 0 || base == 1 || base > 36) { + ns = p; + set_errno(EINVAL); + ret = UINTMAX_MAX; + goto label_return; + } + b = base; + + /* Swallow leading whitespace and get sign, if any. */ + neg = false; + while (true) { + switch (*p) { + case '\t': case '\n': case '\v': case '\f': case '\r': case ' ': + p++; + break; + case '-': + neg = true; + /* Fall through. */ + case '+': + p++; + /* Fall through. */ + default: + goto label_prefix; + } + } + + /* Get prefix, if any. */ + label_prefix: + /* + * Note where the first non-whitespace/sign character is so that it is + * possible to tell whether any digits are consumed (e.g., " 0" vs. + * " -x"). + */ + ns = p; + if (*p == '0') { + switch (p[1]) { + case '0': case '1': case '2': case '3': case '4': case '5': + case '6': case '7': + if (b == 0) { + b = 8; + } + if (b == 8) { + p++; + } + break; + case 'X': case 'x': + switch (p[2]) { + case '0': case '1': case '2': case '3': case '4': + case '5': case '6': case '7': case '8': case '9': + case 'A': case 'B': case 'C': case 'D': case 'E': + case 'F': + case 'a': case 'b': case 'c': case 'd': case 'e': + case 'f': + if (b == 0) { + b = 16; + } + if (b == 16) { + p += 2; + } + break; + default: + break; + } + break; + default: + p++; + ret = 0; + goto label_return; + } + } + if (b == 0) { + b = 10; + } + + /* Convert. */ + ret = 0; + while ((*p >= '0' && *p <= '9' && (digit = *p - '0') < b) + || (*p >= 'A' && *p <= 'Z' && (digit = 10 + *p - 'A') < b) + || (*p >= 'a' && *p <= 'z' && (digit = 10 + *p - 'a') < b)) { + uintmax_t pret = ret; + ret *= b; + ret += digit; + if (ret < pret) { + /* Overflow. */ + set_errno(ERANGE); + ret = UINTMAX_MAX; + goto label_return; + } + p++; + } + if (neg) { + ret = (uintmax_t)(-((intmax_t)ret)); + } + + if (p == ns) { + /* No conversion performed. */ + set_errno(EINVAL); + ret = UINTMAX_MAX; + goto label_return; + } + +label_return: + if (endptr != NULL) { + if (p == ns) { + /* No characters were converted. */ + *endptr = (char *)nptr; + } else { + *endptr = (char *)p; + } + } + return ret; +} + +static char * +u2s(uintmax_t x, unsigned base, bool uppercase, char *s, size_t *slen_p) { + unsigned i; + + i = U2S_BUFSIZE - 1; + s[i] = '\0'; + switch (base) { + case 10: + do { + i--; + s[i] = "0123456789"[x % (uint64_t)10]; + x /= (uint64_t)10; + } while (x > 0); + break; + case 16: { + const char *digits = (uppercase) + ? "0123456789ABCDEF" + : "0123456789abcdef"; + + do { + i--; + s[i] = digits[x & 0xf]; + x >>= 4; + } while (x > 0); + break; + } default: { + const char *digits = (uppercase) + ? "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ" + : "0123456789abcdefghijklmnopqrstuvwxyz"; + + assert(base >= 2 && base <= 36); + do { + i--; + s[i] = digits[x % (uint64_t)base]; + x /= (uint64_t)base; + } while (x > 0); + }} + + *slen_p = U2S_BUFSIZE - 1 - i; + return &s[i]; +} + +static char * +d2s(intmax_t x, char sign, char *s, size_t *slen_p) { + bool neg; + + if ((neg = (x < 0))) { + x = -x; + } + s = u2s(x, 10, false, s, slen_p); + if (neg) { + sign = '-'; + } + switch (sign) { + case '-': + if (!neg) { + break; + } + /* Fall through. */ + case ' ': + case '+': + s--; + (*slen_p)++; + *s = sign; + break; + default: not_reached(); + } + return s; +} + +static char * +o2s(uintmax_t x, bool alt_form, char *s, size_t *slen_p) { + s = u2s(x, 8, false, s, slen_p); + if (alt_form && *s != '0') { + s--; + (*slen_p)++; + *s = '0'; + } + return s; +} + +static char * +x2s(uintmax_t x, bool alt_form, bool uppercase, char *s, size_t *slen_p) { + s = u2s(x, 16, uppercase, s, slen_p); + if (alt_form) { + s -= 2; + (*slen_p) += 2; + memcpy(s, uppercase ? "0X" : "0x", 2); + } + return s; +} + +size_t +malloc_vsnprintf(char *str, size_t size, const char *format, va_list ap) { + size_t i; + const char *f; + +#define APPEND_C(c) do { \ + if (i < size) { \ + str[i] = (c); \ + } \ + i++; \ +} while (0) +#define APPEND_S(s, slen) do { \ + if (i < size) { \ + size_t cpylen = (slen <= size - i) ? slen : size - i; \ + memcpy(&str[i], s, cpylen); \ + } \ + i += slen; \ +} while (0) +#define APPEND_PADDED_S(s, slen, width, left_justify) do { \ + /* Left padding. */ \ + size_t pad_len = (width == -1) ? 0 : ((slen < (size_t)width) ? \ + (size_t)width - slen : 0); \ + if (!left_justify && pad_len != 0) { \ + size_t j; \ + for (j = 0; j < pad_len; j++) { \ + APPEND_C(' '); \ + } \ + } \ + /* Value. */ \ + APPEND_S(s, slen); \ + /* Right padding. */ \ + if (left_justify && pad_len != 0) { \ + size_t j; \ + for (j = 0; j < pad_len; j++) { \ + APPEND_C(' '); \ + } \ + } \ +} while (0) +#define GET_ARG_NUMERIC(val, len) do { \ + switch ((unsigned char)len) { \ + case '?': \ + val = va_arg(ap, int); \ + break; \ + case '?' | 0x80: \ + val = va_arg(ap, unsigned int); \ + break; \ + case 'l': \ + val = va_arg(ap, long); \ + break; \ + case 'l' | 0x80: \ + val = va_arg(ap, unsigned long); \ + break; \ + case 'q': \ + val = va_arg(ap, long long); \ + break; \ + case 'q' | 0x80: \ + val = va_arg(ap, unsigned long long); \ + break; \ + case 'j': \ + val = va_arg(ap, intmax_t); \ + break; \ + case 'j' | 0x80: \ + val = va_arg(ap, uintmax_t); \ + break; \ + case 't': \ + val = va_arg(ap, ptrdiff_t); \ + break; \ + case 'z': \ + val = va_arg(ap, ssize_t); \ + break; \ + case 'z' | 0x80: \ + val = va_arg(ap, size_t); \ + break; \ + case 'p': /* Synthetic; used for %p. */ \ + val = va_arg(ap, uintptr_t); \ + break; \ + default: \ + not_reached(); \ + val = 0; \ + } \ +} while (0) + + i = 0; + f = format; + while (true) { + switch (*f) { + case '\0': goto label_out; + case '%': { + bool alt_form = false; + bool left_justify = false; + bool plus_space = false; + bool plus_plus = false; + int prec = -1; + int width = -1; + unsigned char len = '?'; + char *s; + size_t slen; + + f++; + /* Flags. */ + while (true) { + switch (*f) { + case '#': + assert(!alt_form); + alt_form = true; + break; + case '-': + assert(!left_justify); + left_justify = true; + break; + case ' ': + assert(!plus_space); + plus_space = true; + break; + case '+': + assert(!plus_plus); + plus_plus = true; + break; + default: goto label_width; + } + f++; + } + /* Width. */ + label_width: + switch (*f) { + case '*': + width = va_arg(ap, int); + f++; + if (width < 0) { + left_justify = true; + width = -width; + } + break; + case '0': case '1': case '2': case '3': case '4': + case '5': case '6': case '7': case '8': case '9': { + uintmax_t uwidth; + set_errno(0); + uwidth = malloc_strtoumax(f, (char **)&f, 10); + assert(uwidth != UINTMAX_MAX || get_errno() != + ERANGE); + width = (int)uwidth; + break; + } default: + break; + } + /* Width/precision separator. */ + if (*f == '.') { + f++; + } else { + goto label_length; + } + /* Precision. */ + switch (*f) { + case '*': + prec = va_arg(ap, int); + f++; + break; + case '0': case '1': case '2': case '3': case '4': + case '5': case '6': case '7': case '8': case '9': { + uintmax_t uprec; + set_errno(0); + uprec = malloc_strtoumax(f, (char **)&f, 10); + assert(uprec != UINTMAX_MAX || get_errno() != + ERANGE); + prec = (int)uprec; + break; + } + default: break; + } + /* Length. */ + label_length: + switch (*f) { + case 'l': + f++; + if (*f == 'l') { + len = 'q'; + f++; + } else { + len = 'l'; + } + break; + case 'q': case 'j': case 't': case 'z': + len = *f; + f++; + break; + default: break; + } + /* Conversion specifier. */ + switch (*f) { + case '%': + /* %% */ + APPEND_C(*f); + f++; + break; + case 'd': case 'i': { + intmax_t val JEMALLOC_CC_SILENCE_INIT(0); + char buf[D2S_BUFSIZE]; + + GET_ARG_NUMERIC(val, len); + s = d2s(val, (plus_plus ? '+' : (plus_space ? + ' ' : '-')), buf, &slen); + APPEND_PADDED_S(s, slen, width, left_justify); + f++; + break; + } case 'o': { + uintmax_t val JEMALLOC_CC_SILENCE_INIT(0); + char buf[O2S_BUFSIZE]; + + GET_ARG_NUMERIC(val, len | 0x80); + s = o2s(val, alt_form, buf, &slen); + APPEND_PADDED_S(s, slen, width, left_justify); + f++; + break; + } case 'u': { + uintmax_t val JEMALLOC_CC_SILENCE_INIT(0); + char buf[U2S_BUFSIZE]; + + GET_ARG_NUMERIC(val, len | 0x80); + s = u2s(val, 10, false, buf, &slen); + APPEND_PADDED_S(s, slen, width, left_justify); + f++; + break; + } case 'x': case 'X': { + uintmax_t val JEMALLOC_CC_SILENCE_INIT(0); + char buf[X2S_BUFSIZE]; + + GET_ARG_NUMERIC(val, len | 0x80); + s = x2s(val, alt_form, *f == 'X', buf, &slen); + APPEND_PADDED_S(s, slen, width, left_justify); + f++; + break; + } case 'c': { + unsigned char val; + char buf[2]; + + assert(len == '?' || len == 'l'); + assert_not_implemented(len != 'l'); + val = va_arg(ap, int); + buf[0] = val; + buf[1] = '\0'; + APPEND_PADDED_S(buf, 1, width, left_justify); + f++; + break; + } case 's': + assert(len == '?' || len == 'l'); + assert_not_implemented(len != 'l'); + s = va_arg(ap, char *); + slen = (prec < 0) ? strlen(s) : (size_t)prec; + APPEND_PADDED_S(s, slen, width, left_justify); + f++; + break; + case 'p': { + uintmax_t val; + char buf[X2S_BUFSIZE]; + + GET_ARG_NUMERIC(val, 'p'); + s = x2s(val, true, false, buf, &slen); + APPEND_PADDED_S(s, slen, width, left_justify); + f++; + break; + } default: not_reached(); + } + break; + } default: { + APPEND_C(*f); + f++; + break; + }} + } + label_out: + if (i < size) { + str[i] = '\0'; + } else { + str[size - 1] = '\0'; + } + +#undef APPEND_C +#undef APPEND_S +#undef APPEND_PADDED_S +#undef GET_ARG_NUMERIC + return i; +} + +JEMALLOC_FORMAT_PRINTF(3, 4) +size_t +malloc_snprintf(char *str, size_t size, const char *format, ...) { + size_t ret; + va_list ap; + + va_start(ap, format); + ret = malloc_vsnprintf(str, size, format, ap); + va_end(ap); + + return ret; +} + +void +malloc_vcprintf(void (*write_cb)(void *, const char *), void *cbopaque, + const char *format, va_list ap) { + char buf[MALLOC_PRINTF_BUFSIZE]; + + if (write_cb == NULL) { + /* + * The caller did not provide an alternate write_cb callback + * function, so use the default one. malloc_write() is an + * inline function, so use malloc_message() directly here. + */ + write_cb = (je_malloc_message != NULL) ? je_malloc_message : + wrtmessage; + } + + malloc_vsnprintf(buf, sizeof(buf), format, ap); + write_cb(cbopaque, buf); +} + +/* + * Print to a callback function in such a way as to (hopefully) avoid memory + * allocation. + */ +JEMALLOC_FORMAT_PRINTF(3, 4) +void +malloc_cprintf(void (*write_cb)(void *, const char *), void *cbopaque, + const char *format, ...) { + va_list ap; + + va_start(ap, format); + malloc_vcprintf(write_cb, cbopaque, format, ap); + va_end(ap); +} + +/* Print to stderr in such a way as to avoid memory allocation. */ +JEMALLOC_FORMAT_PRINTF(1, 2) +void +malloc_printf(const char *format, ...) { + va_list ap; + + va_start(ap, format); + malloc_vcprintf(NULL, NULL, format, ap); + va_end(ap); +} + +/* + * Restore normal assertion macros, in order to make it possible to compile all + * C files as a single concatenation. + */ +#undef assert +#undef not_reached +#undef not_implemented +#undef assert_not_implemented +#include "jemalloc/internal/assert.h" diff --git a/deps/jemalloc/src/mutex.c b/deps/jemalloc/src/mutex.c new file mode 100644 index 000000000..3f920f5b1 --- /dev/null +++ b/deps/jemalloc/src/mutex.c @@ -0,0 +1,223 @@ +#define JEMALLOC_MUTEX_C_ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/assert.h" +#include "jemalloc/internal/malloc_io.h" +#include "jemalloc/internal/spin.h" + +#ifndef _CRT_SPINCOUNT +#define _CRT_SPINCOUNT 4000 +#endif + +/******************************************************************************/ +/* Data. */ + +#ifdef JEMALLOC_LAZY_LOCK +bool isthreaded = false; +#endif +#ifdef JEMALLOC_MUTEX_INIT_CB +static bool postpone_init = true; +static malloc_mutex_t *postponed_mutexes = NULL; +#endif + +/******************************************************************************/ +/* + * We intercept pthread_create() calls in order to toggle isthreaded if the + * process goes multi-threaded. + */ + +#if defined(JEMALLOC_LAZY_LOCK) && !defined(_WIN32) +JEMALLOC_EXPORT int +pthread_create(pthread_t *__restrict thread, + const pthread_attr_t *__restrict attr, void *(*start_routine)(void *), + void *__restrict arg) { + return pthread_create_wrapper(thread, attr, start_routine, arg); +} +#endif + +/******************************************************************************/ + +#ifdef JEMALLOC_MUTEX_INIT_CB +JEMALLOC_EXPORT int _pthread_mutex_init_calloc_cb(pthread_mutex_t *mutex, + void *(calloc_cb)(size_t, size_t)); +#endif + +void +malloc_mutex_lock_slow(malloc_mutex_t *mutex) { + mutex_prof_data_t *data = &mutex->prof_data; + nstime_t before = NSTIME_ZERO_INITIALIZER; + + if (ncpus == 1) { + goto label_spin_done; + } + + int cnt = 0, max_cnt = MALLOC_MUTEX_MAX_SPIN; + do { + spin_cpu_spinwait(); + if (!atomic_load_b(&mutex->locked, ATOMIC_RELAXED) + && !malloc_mutex_trylock_final(mutex)) { + data->n_spin_acquired++; + return; + } + } while (cnt++ < max_cnt); + + if (!config_stats) { + /* Only spin is useful when stats is off. */ + malloc_mutex_lock_final(mutex); + return; + } +label_spin_done: + nstime_update(&before); + /* Copy before to after to avoid clock skews. */ + nstime_t after; + nstime_copy(&after, &before); + uint32_t n_thds = atomic_fetch_add_u32(&data->n_waiting_thds, 1, + ATOMIC_RELAXED) + 1; + /* One last try as above two calls may take quite some cycles. */ + if (!malloc_mutex_trylock_final(mutex)) { + atomic_fetch_sub_u32(&data->n_waiting_thds, 1, ATOMIC_RELAXED); + data->n_spin_acquired++; + return; + } + + /* True slow path. */ + malloc_mutex_lock_final(mutex); + /* Update more slow-path only counters. */ + atomic_fetch_sub_u32(&data->n_waiting_thds, 1, ATOMIC_RELAXED); + nstime_update(&after); + + nstime_t delta; + nstime_copy(&delta, &after); + nstime_subtract(&delta, &before); + + data->n_wait_times++; + nstime_add(&data->tot_wait_time, &delta); + if (nstime_compare(&data->max_wait_time, &delta) < 0) { + nstime_copy(&data->max_wait_time, &delta); + } + if (n_thds > data->max_n_thds) { + data->max_n_thds = n_thds; + } +} + +static void +mutex_prof_data_init(mutex_prof_data_t *data) { + memset(data, 0, sizeof(mutex_prof_data_t)); + nstime_init(&data->max_wait_time, 0); + nstime_init(&data->tot_wait_time, 0); + data->prev_owner = NULL; +} + +void +malloc_mutex_prof_data_reset(tsdn_t *tsdn, malloc_mutex_t *mutex) { + malloc_mutex_assert_owner(tsdn, mutex); + mutex_prof_data_init(&mutex->prof_data); +} + +static int +mutex_addr_comp(const witness_t *witness1, void *mutex1, + const witness_t *witness2, void *mutex2) { + assert(mutex1 != NULL); + assert(mutex2 != NULL); + uintptr_t mu1int = (uintptr_t)mutex1; + uintptr_t mu2int = (uintptr_t)mutex2; + if (mu1int < mu2int) { + return -1; + } else if (mu1int == mu2int) { + return 0; + } else { + return 1; + } +} + +bool +malloc_mutex_init(malloc_mutex_t *mutex, const char *name, + witness_rank_t rank, malloc_mutex_lock_order_t lock_order) { + mutex_prof_data_init(&mutex->prof_data); +#ifdef _WIN32 +# if _WIN32_WINNT >= 0x0600 + InitializeSRWLock(&mutex->lock); +# else + if (!InitializeCriticalSectionAndSpinCount(&mutex->lock, + _CRT_SPINCOUNT)) { + return true; + } +# endif +#elif (defined(JEMALLOC_OS_UNFAIR_LOCK)) + mutex->lock = OS_UNFAIR_LOCK_INIT; +#elif (defined(JEMALLOC_MUTEX_INIT_CB)) + if (postpone_init) { + mutex->postponed_next = postponed_mutexes; + postponed_mutexes = mutex; + } else { + if (_pthread_mutex_init_calloc_cb(&mutex->lock, + bootstrap_calloc) != 0) { + return true; + } + } +#else + pthread_mutexattr_t attr; + + if (pthread_mutexattr_init(&attr) != 0) { + return true; + } + pthread_mutexattr_settype(&attr, MALLOC_MUTEX_TYPE); + if (pthread_mutex_init(&mutex->lock, &attr) != 0) { + pthread_mutexattr_destroy(&attr); + return true; + } + pthread_mutexattr_destroy(&attr); +#endif + if (config_debug) { + mutex->lock_order = lock_order; + if (lock_order == malloc_mutex_address_ordered) { + witness_init(&mutex->witness, name, rank, + mutex_addr_comp, mutex); + } else { + witness_init(&mutex->witness, name, rank, NULL, NULL); + } + } + return false; +} + +void +malloc_mutex_prefork(tsdn_t *tsdn, malloc_mutex_t *mutex) { + malloc_mutex_lock(tsdn, mutex); +} + +void +malloc_mutex_postfork_parent(tsdn_t *tsdn, malloc_mutex_t *mutex) { + malloc_mutex_unlock(tsdn, mutex); +} + +void +malloc_mutex_postfork_child(tsdn_t *tsdn, malloc_mutex_t *mutex) { +#ifdef JEMALLOC_MUTEX_INIT_CB + malloc_mutex_unlock(tsdn, mutex); +#else + if (malloc_mutex_init(mutex, mutex->witness.name, + mutex->witness.rank, mutex->lock_order)) { + malloc_printf("<jemalloc>: Error re-initializing mutex in " + "child\n"); + if (opt_abort) { + abort(); + } + } +#endif +} + +bool +malloc_mutex_boot(void) { +#ifdef JEMALLOC_MUTEX_INIT_CB + postpone_init = false; + while (postponed_mutexes != NULL) { + if (_pthread_mutex_init_calloc_cb(&postponed_mutexes->lock, + bootstrap_calloc) != 0) { + return true; + } + postponed_mutexes = postponed_mutexes->postponed_next; + } +#endif + return false; +} diff --git a/deps/jemalloc/src/mutex_pool.c b/deps/jemalloc/src/mutex_pool.c new file mode 100644 index 000000000..f24d10e44 --- /dev/null +++ b/deps/jemalloc/src/mutex_pool.c @@ -0,0 +1,18 @@ +#define JEMALLOC_MUTEX_POOL_C_ + +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/mutex.h" +#include "jemalloc/internal/mutex_pool.h" + +bool +mutex_pool_init(mutex_pool_t *pool, const char *name, witness_rank_t rank) { + for (int i = 0; i < MUTEX_POOL_SIZE; ++i) { + if (malloc_mutex_init(&pool->mutexes[i], name, rank, + malloc_mutex_address_ordered)) { + return true; + } + } + return false; +} diff --git a/deps/jemalloc/src/nstime.c b/deps/jemalloc/src/nstime.c new file mode 100644 index 000000000..71db35396 --- /dev/null +++ b/deps/jemalloc/src/nstime.c @@ -0,0 +1,170 @@ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/nstime.h" + +#include "jemalloc/internal/assert.h" + +#define BILLION UINT64_C(1000000000) +#define MILLION UINT64_C(1000000) + +void +nstime_init(nstime_t *time, uint64_t ns) { + time->ns = ns; +} + +void +nstime_init2(nstime_t *time, uint64_t sec, uint64_t nsec) { + time->ns = sec * BILLION + nsec; +} + +uint64_t +nstime_ns(const nstime_t *time) { + return time->ns; +} + +uint64_t +nstime_msec(const nstime_t *time) { + return time->ns / MILLION; +} + +uint64_t +nstime_sec(const nstime_t *time) { + return time->ns / BILLION; +} + +uint64_t +nstime_nsec(const nstime_t *time) { + return time->ns % BILLION; +} + +void +nstime_copy(nstime_t *time, const nstime_t *source) { + *time = *source; +} + +int +nstime_compare(const nstime_t *a, const nstime_t *b) { + return (a->ns > b->ns) - (a->ns < b->ns); +} + +void +nstime_add(nstime_t *time, const nstime_t *addend) { + assert(UINT64_MAX - time->ns >= addend->ns); + + time->ns += addend->ns; +} + +void +nstime_iadd(nstime_t *time, uint64_t addend) { + assert(UINT64_MAX - time->ns >= addend); + + time->ns += addend; +} + +void +nstime_subtract(nstime_t *time, const nstime_t *subtrahend) { + assert(nstime_compare(time, subtrahend) >= 0); + + time->ns -= subtrahend->ns; +} + +void +nstime_isubtract(nstime_t *time, uint64_t subtrahend) { + assert(time->ns >= subtrahend); + + time->ns -= subtrahend; +} + +void +nstime_imultiply(nstime_t *time, uint64_t multiplier) { + assert((((time->ns | multiplier) & (UINT64_MAX << (sizeof(uint64_t) << + 2))) == 0) || ((time->ns * multiplier) / multiplier == time->ns)); + + time->ns *= multiplier; +} + +void +nstime_idivide(nstime_t *time, uint64_t divisor) { + assert(divisor != 0); + + time->ns /= divisor; +} + +uint64_t +nstime_divide(const nstime_t *time, const nstime_t *divisor) { + assert(divisor->ns != 0); + + return time->ns / divisor->ns; +} + +#ifdef _WIN32 +# define NSTIME_MONOTONIC true +static void +nstime_get(nstime_t *time) { + FILETIME ft; + uint64_t ticks_100ns; + + GetSystemTimeAsFileTime(&ft); + ticks_100ns = (((uint64_t)ft.dwHighDateTime) << 32) | ft.dwLowDateTime; + + nstime_init(time, ticks_100ns * 100); +} +#elif defined(JEMALLOC_HAVE_CLOCK_MONOTONIC_COARSE) +# define NSTIME_MONOTONIC true +static void +nstime_get(nstime_t *time) { + struct timespec ts; + + clock_gettime(CLOCK_MONOTONIC_COARSE, &ts); + nstime_init2(time, ts.tv_sec, ts.tv_nsec); +} +#elif defined(JEMALLOC_HAVE_CLOCK_MONOTONIC) +# define NSTIME_MONOTONIC true +static void +nstime_get(nstime_t *time) { + struct timespec ts; + + clock_gettime(CLOCK_MONOTONIC, &ts); + nstime_init2(time, ts.tv_sec, ts.tv_nsec); +} +#elif defined(JEMALLOC_HAVE_MACH_ABSOLUTE_TIME) +# define NSTIME_MONOTONIC true +static void +nstime_get(nstime_t *time) { + nstime_init(time, mach_absolute_time()); +} +#else +# define NSTIME_MONOTONIC false +static void +nstime_get(nstime_t *time) { + struct timeval tv; + + gettimeofday(&tv, NULL); + nstime_init2(time, tv.tv_sec, tv.tv_usec * 1000); +} +#endif + +static bool +nstime_monotonic_impl(void) { + return NSTIME_MONOTONIC; +#undef NSTIME_MONOTONIC +} +nstime_monotonic_t *JET_MUTABLE nstime_monotonic = nstime_monotonic_impl; + +static bool +nstime_update_impl(nstime_t *time) { + nstime_t old_time; + + nstime_copy(&old_time, time); + nstime_get(time); + + /* Handle non-monotonic clocks. */ + if (unlikely(nstime_compare(&old_time, time) > 0)) { + nstime_copy(time, &old_time); + return true; + } + + return false; +} +nstime_update_t *JET_MUTABLE nstime_update = nstime_update_impl; diff --git a/deps/jemalloc/src/pages.c b/deps/jemalloc/src/pages.c new file mode 100644 index 000000000..13de27a00 --- /dev/null +++ b/deps/jemalloc/src/pages.c @@ -0,0 +1,649 @@ +#define JEMALLOC_PAGES_C_ +#include "jemalloc/internal/jemalloc_preamble.h" + +#include "jemalloc/internal/pages.h" + +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/assert.h" +#include "jemalloc/internal/malloc_io.h" + +#ifdef JEMALLOC_SYSCTL_VM_OVERCOMMIT +#include <sys/sysctl.h> +#ifdef __FreeBSD__ +#include <vm/vm_param.h> +#endif +#endif + +/******************************************************************************/ +/* Data. */ + +/* Actual operating system page size, detected during bootstrap, <= PAGE. */ +static size_t os_page; + +#ifndef _WIN32 +# define PAGES_PROT_COMMIT (PROT_READ | PROT_WRITE) +# define PAGES_PROT_DECOMMIT (PROT_NONE) +static int mmap_flags; +#endif +static bool os_overcommits; + +const char *thp_mode_names[] = { + "default", + "always", + "never", + "not supported" +}; +thp_mode_t opt_thp = THP_MODE_DEFAULT; +thp_mode_t init_system_thp_mode; + +/* Runtime support for lazy purge. Irrelevant when !pages_can_purge_lazy. */ +static bool pages_can_purge_lazy_runtime = true; + +/******************************************************************************/ +/* + * Function prototypes for static functions that are referenced prior to + * definition. + */ + +static void os_pages_unmap(void *addr, size_t size); + +/******************************************************************************/ + +static void * +os_pages_map(void *addr, size_t size, size_t alignment, bool *commit) { + assert(ALIGNMENT_ADDR2BASE(addr, os_page) == addr); + assert(ALIGNMENT_CEILING(size, os_page) == size); + assert(size != 0); + + if (os_overcommits) { + *commit = true; + } + + void *ret; +#ifdef _WIN32 + /* + * If VirtualAlloc can't allocate at the given address when one is + * given, it fails and returns NULL. + */ + ret = VirtualAlloc(addr, size, MEM_RESERVE | (*commit ? MEM_COMMIT : 0), + PAGE_READWRITE); +#else + /* + * We don't use MAP_FIXED here, because it can cause the *replacement* + * of existing mappings, and we only want to create new mappings. + */ + { + int prot = *commit ? PAGES_PROT_COMMIT : PAGES_PROT_DECOMMIT; + + ret = mmap(addr, size, prot, mmap_flags, -1, 0); + } + assert(ret != NULL); + + if (ret == MAP_FAILED) { + ret = NULL; + } else if (addr != NULL && ret != addr) { + /* + * We succeeded in mapping memory, but not in the right place. + */ + os_pages_unmap(ret, size); + ret = NULL; + } +#endif + assert(ret == NULL || (addr == NULL && ret != addr) || (addr != NULL && + ret == addr)); + return ret; +} + +static void * +os_pages_trim(void *addr, size_t alloc_size, size_t leadsize, size_t size, + bool *commit) { + void *ret = (void *)((uintptr_t)addr + leadsize); + + assert(alloc_size >= leadsize + size); +#ifdef _WIN32 + os_pages_unmap(addr, alloc_size); + void *new_addr = os_pages_map(ret, size, PAGE, commit); + if (new_addr == ret) { + return ret; + } + if (new_addr != NULL) { + os_pages_unmap(new_addr, size); + } + return NULL; +#else + size_t trailsize = alloc_size - leadsize - size; + + if (leadsize != 0) { + os_pages_unmap(addr, leadsize); + } + if (trailsize != 0) { + os_pages_unmap((void *)((uintptr_t)ret + size), trailsize); + } + return ret; +#endif +} + +static void +os_pages_unmap(void *addr, size_t size) { + assert(ALIGNMENT_ADDR2BASE(addr, os_page) == addr); + assert(ALIGNMENT_CEILING(size, os_page) == size); + +#ifdef _WIN32 + if (VirtualFree(addr, 0, MEM_RELEASE) == 0) +#else + if (munmap(addr, size) == -1) +#endif + { + char buf[BUFERROR_BUF]; + + buferror(get_errno(), buf, sizeof(buf)); + malloc_printf("<jemalloc>: Error in " +#ifdef _WIN32 + "VirtualFree" +#else + "munmap" +#endif + "(): %s\n", buf); + if (opt_abort) { + abort(); + } + } +} + +static void * +pages_map_slow(size_t size, size_t alignment, bool *commit) { + size_t alloc_size = size + alignment - os_page; + /* Beware size_t wrap-around. */ + if (alloc_size < size) { + return NULL; + } + + void *ret; + do { + void *pages = os_pages_map(NULL, alloc_size, alignment, commit); + if (pages == NULL) { + return NULL; + } + size_t leadsize = ALIGNMENT_CEILING((uintptr_t)pages, alignment) + - (uintptr_t)pages; + ret = os_pages_trim(pages, alloc_size, leadsize, size, commit); + } while (ret == NULL); + + assert(ret != NULL); + assert(PAGE_ADDR2BASE(ret) == ret); + return ret; +} + +void * +pages_map(void *addr, size_t size, size_t alignment, bool *commit) { + assert(alignment >= PAGE); + assert(ALIGNMENT_ADDR2BASE(addr, alignment) == addr); + +#if defined(__FreeBSD__) && defined(MAP_EXCL) + /* + * FreeBSD has mechanisms both to mmap at specific address without + * touching existing mappings, and to mmap with specific alignment. + */ + { + if (os_overcommits) { + *commit = true; + } + + int prot = *commit ? PAGES_PROT_COMMIT : PAGES_PROT_DECOMMIT; + int flags = mmap_flags; + + if (addr != NULL) { + flags |= MAP_FIXED | MAP_EXCL; + } else { + unsigned alignment_bits = ffs_zu(alignment); + assert(alignment_bits > 1); + flags |= MAP_ALIGNED(alignment_bits - 1); + } + + void *ret = mmap(addr, size, prot, flags, -1, 0); + if (ret == MAP_FAILED) { + ret = NULL; + } + + return ret; + } +#endif + /* + * Ideally, there would be a way to specify alignment to mmap() (like + * NetBSD has), but in the absence of such a feature, we have to work + * hard to efficiently create aligned mappings. The reliable, but + * slow method is to create a mapping that is over-sized, then trim the + * excess. However, that always results in one or two calls to + * os_pages_unmap(), and it can leave holes in the process's virtual + * memory map if memory grows downward. + * + * Optimistically try mapping precisely the right amount before falling + * back to the slow method, with the expectation that the optimistic + * approach works most of the time. + */ + + void *ret = os_pages_map(addr, size, os_page, commit); + if (ret == NULL || ret == addr) { + return ret; + } + assert(addr == NULL); + if (ALIGNMENT_ADDR2OFFSET(ret, alignment) != 0) { + os_pages_unmap(ret, size); + return pages_map_slow(size, alignment, commit); + } + + assert(PAGE_ADDR2BASE(ret) == ret); + return ret; +} + +void +pages_unmap(void *addr, size_t size) { + assert(PAGE_ADDR2BASE(addr) == addr); + assert(PAGE_CEILING(size) == size); + + os_pages_unmap(addr, size); +} + +static bool +pages_commit_impl(void *addr, size_t size, bool commit) { + assert(PAGE_ADDR2BASE(addr) == addr); + assert(PAGE_CEILING(size) == size); + + if (os_overcommits) { + return true; + } + +#ifdef _WIN32 + return (commit ? (addr != VirtualAlloc(addr, size, MEM_COMMIT, + PAGE_READWRITE)) : (!VirtualFree(addr, size, MEM_DECOMMIT))); +#else + { + int prot = commit ? PAGES_PROT_COMMIT : PAGES_PROT_DECOMMIT; + void *result = mmap(addr, size, prot, mmap_flags | MAP_FIXED, + -1, 0); + if (result == MAP_FAILED) { + return true; + } + if (result != addr) { + /* + * We succeeded in mapping memory, but not in the right + * place. + */ + os_pages_unmap(result, size); + return true; + } + return false; + } +#endif +} + +bool +pages_commit(void *addr, size_t size) { + return pages_commit_impl(addr, size, true); +} + +bool +pages_decommit(void *addr, size_t size) { + return pages_commit_impl(addr, size, false); +} + +bool +pages_purge_lazy(void *addr, size_t size) { + assert(ALIGNMENT_ADDR2BASE(addr, os_page) == addr); + assert(PAGE_CEILING(size) == size); + + if (!pages_can_purge_lazy) { + return true; + } + if (!pages_can_purge_lazy_runtime) { + /* + * Built with lazy purge enabled, but detected it was not + * supported on the current system. + */ + return true; + } + +#ifdef _WIN32 + VirtualAlloc(addr, size, MEM_RESET, PAGE_READWRITE); + return false; +#elif defined(JEMALLOC_PURGE_MADVISE_FREE) + return (madvise(addr, size, +# ifdef MADV_FREE + MADV_FREE +# else + JEMALLOC_MADV_FREE +# endif + ) != 0); +#elif defined(JEMALLOC_PURGE_MADVISE_DONTNEED) && \ + !defined(JEMALLOC_PURGE_MADVISE_DONTNEED_ZEROS) + return (madvise(addr, size, MADV_DONTNEED) != 0); +#else + not_reached(); +#endif +} + +bool +pages_purge_forced(void *addr, size_t size) { + assert(PAGE_ADDR2BASE(addr) == addr); + assert(PAGE_CEILING(size) == size); + + if (!pages_can_purge_forced) { + return true; + } + +#if defined(JEMALLOC_PURGE_MADVISE_DONTNEED) && \ + defined(JEMALLOC_PURGE_MADVISE_DONTNEED_ZEROS) + return (madvise(addr, size, MADV_DONTNEED) != 0); +#elif defined(JEMALLOC_MAPS_COALESCE) + /* Try to overlay a new demand-zeroed mapping. */ + return pages_commit(addr, size); +#else + not_reached(); +#endif +} + +static bool +pages_huge_impl(void *addr, size_t size, bool aligned) { + if (aligned) { + assert(HUGEPAGE_ADDR2BASE(addr) == addr); + assert(HUGEPAGE_CEILING(size) == size); + } +#ifdef JEMALLOC_HAVE_MADVISE_HUGE + return (madvise(addr, size, MADV_HUGEPAGE) != 0); +#else + return true; +#endif +} + +bool +pages_huge(void *addr, size_t size) { + return pages_huge_impl(addr, size, true); +} + +static bool +pages_huge_unaligned(void *addr, size_t size) { + return pages_huge_impl(addr, size, false); +} + +static bool +pages_nohuge_impl(void *addr, size_t size, bool aligned) { + if (aligned) { + assert(HUGEPAGE_ADDR2BASE(addr) == addr); + assert(HUGEPAGE_CEILING(size) == size); + } + +#ifdef JEMALLOC_HAVE_MADVISE_HUGE + return (madvise(addr, size, MADV_NOHUGEPAGE) != 0); +#else + return false; +#endif +} + +bool +pages_nohuge(void *addr, size_t size) { + return pages_nohuge_impl(addr, size, true); +} + +static bool +pages_nohuge_unaligned(void *addr, size_t size) { + return pages_nohuge_impl(addr, size, false); +} + +bool +pages_dontdump(void *addr, size_t size) { + assert(PAGE_ADDR2BASE(addr) == addr); + assert(PAGE_CEILING(size) == size); +#ifdef JEMALLOC_MADVISE_DONTDUMP + return madvise(addr, size, MADV_DONTDUMP) != 0; +#else + return false; +#endif +} + +bool +pages_dodump(void *addr, size_t size) { + assert(PAGE_ADDR2BASE(addr) == addr); + assert(PAGE_CEILING(size) == size); +#ifdef JEMALLOC_MADVISE_DONTDUMP + return madvise(addr, size, MADV_DODUMP) != 0; +#else + return false; +#endif +} + + +static size_t +os_page_detect(void) { +#ifdef _WIN32 + SYSTEM_INFO si; + GetSystemInfo(&si); + return si.dwPageSize; +#elif defined(__FreeBSD__) + /* + * This returns the value obtained from + * the auxv vector, avoiding a syscall. + */ + return getpagesize(); +#else + long result = sysconf(_SC_PAGESIZE); + if (result == -1) { + return LG_PAGE; + } + return (size_t)result; +#endif +} + +#ifdef JEMALLOC_SYSCTL_VM_OVERCOMMIT +static bool +os_overcommits_sysctl(void) { + int vm_overcommit; + size_t sz; + + sz = sizeof(vm_overcommit); +#if defined(__FreeBSD__) && defined(VM_OVERCOMMIT) + int mib[2]; + + mib[0] = CTL_VM; + mib[1] = VM_OVERCOMMIT; + if (sysctl(mib, 2, &vm_overcommit, &sz, NULL, 0) != 0) { + return false; /* Error. */ + } +#else + if (sysctlbyname("vm.overcommit", &vm_overcommit, &sz, NULL, 0) != 0) { + return false; /* Error. */ + } +#endif + + return ((vm_overcommit & 0x3) == 0); +} +#endif + +#ifdef JEMALLOC_PROC_SYS_VM_OVERCOMMIT_MEMORY +/* + * Use syscall(2) rather than {open,read,close}(2) when possible to avoid + * reentry during bootstrapping if another library has interposed system call + * wrappers. + */ +static bool +os_overcommits_proc(void) { + int fd; + char buf[1]; + +#if defined(JEMALLOC_USE_SYSCALL) && defined(SYS_open) + #if defined(O_CLOEXEC) + fd = (int)syscall(SYS_open, "/proc/sys/vm/overcommit_memory", O_RDONLY | + O_CLOEXEC); + #else + fd = (int)syscall(SYS_open, "/proc/sys/vm/overcommit_memory", O_RDONLY); + if (fd != -1) { + fcntl(fd, F_SETFD, fcntl(fd, F_GETFD) | FD_CLOEXEC); + } + #endif +#elif defined(JEMALLOC_USE_SYSCALL) && defined(SYS_openat) + #if defined(O_CLOEXEC) + fd = (int)syscall(SYS_openat, + AT_FDCWD, "/proc/sys/vm/overcommit_memory", O_RDONLY | O_CLOEXEC); + #else + fd = (int)syscall(SYS_openat, + AT_FDCWD, "/proc/sys/vm/overcommit_memory", O_RDONLY); + if (fd != -1) { + fcntl(fd, F_SETFD, fcntl(fd, F_GETFD) | FD_CLOEXEC); + } + #endif +#else + #if defined(O_CLOEXEC) + fd = open("/proc/sys/vm/overcommit_memory", O_RDONLY | O_CLOEXEC); + #else + fd = open("/proc/sys/vm/overcommit_memory", O_RDONLY); + if (fd != -1) { + fcntl(fd, F_SETFD, fcntl(fd, F_GETFD) | FD_CLOEXEC); + } + #endif +#endif + + if (fd == -1) { + return false; /* Error. */ + } + + ssize_t nread = malloc_read_fd(fd, &buf, sizeof(buf)); +#if defined(JEMALLOC_USE_SYSCALL) && defined(SYS_close) + syscall(SYS_close, fd); +#else + close(fd); +#endif + + if (nread < 1) { + return false; /* Error. */ + } + /* + * /proc/sys/vm/overcommit_memory meanings: + * 0: Heuristic overcommit. + * 1: Always overcommit. + * 2: Never overcommit. + */ + return (buf[0] == '0' || buf[0] == '1'); +} +#endif + +void +pages_set_thp_state (void *ptr, size_t size) { + if (opt_thp == thp_mode_default || opt_thp == init_system_thp_mode) { + return; + } + assert(opt_thp != thp_mode_not_supported && + init_system_thp_mode != thp_mode_not_supported); + + if (opt_thp == thp_mode_always + && init_system_thp_mode != thp_mode_never) { + assert(init_system_thp_mode == thp_mode_default); + pages_huge_unaligned(ptr, size); + } else if (opt_thp == thp_mode_never) { + assert(init_system_thp_mode == thp_mode_default || + init_system_thp_mode == thp_mode_always); + pages_nohuge_unaligned(ptr, size); + } +} + +static void +init_thp_state(void) { + if (!have_madvise_huge) { + if (metadata_thp_enabled() && opt_abort) { + malloc_write("<jemalloc>: no MADV_HUGEPAGE support\n"); + abort(); + } + goto label_error; + } + + static const char sys_state_madvise[] = "always [madvise] never\n"; + static const char sys_state_always[] = "[always] madvise never\n"; + static const char sys_state_never[] = "always madvise [never]\n"; + char buf[sizeof(sys_state_madvise)]; + +#if defined(JEMALLOC_USE_SYSCALL) && defined(SYS_open) + int fd = (int)syscall(SYS_open, + "/sys/kernel/mm/transparent_hugepage/enabled", O_RDONLY); +#else + int fd = open("/sys/kernel/mm/transparent_hugepage/enabled", O_RDONLY); +#endif + if (fd == -1) { + goto label_error; + } + + ssize_t nread = malloc_read_fd(fd, &buf, sizeof(buf)); +#if defined(JEMALLOC_USE_SYSCALL) && defined(SYS_close) + syscall(SYS_close, fd); +#else + close(fd); +#endif + + if (nread < 0) { + goto label_error; + } + + if (strncmp(buf, sys_state_madvise, (size_t)nread) == 0) { + init_system_thp_mode = thp_mode_default; + } else if (strncmp(buf, sys_state_always, (size_t)nread) == 0) { + init_system_thp_mode = thp_mode_always; + } else if (strncmp(buf, sys_state_never, (size_t)nread) == 0) { + init_system_thp_mode = thp_mode_never; + } else { + goto label_error; + } + return; +label_error: + opt_thp = init_system_thp_mode = thp_mode_not_supported; +} + +bool +pages_boot(void) { + os_page = os_page_detect(); + if (os_page > PAGE) { + malloc_write("<jemalloc>: Unsupported system page size\n"); + if (opt_abort) { + abort(); + } + return true; + } + +#ifndef _WIN32 + mmap_flags = MAP_PRIVATE | MAP_ANON; +#endif + +#ifdef JEMALLOC_SYSCTL_VM_OVERCOMMIT + os_overcommits = os_overcommits_sysctl(); +#elif defined(JEMALLOC_PROC_SYS_VM_OVERCOMMIT_MEMORY) + os_overcommits = os_overcommits_proc(); +# ifdef MAP_NORESERVE + if (os_overcommits) { + mmap_flags |= MAP_NORESERVE; + } +# endif +#else + os_overcommits = false; +#endif + + init_thp_state(); + +#ifdef __FreeBSD__ + /* + * FreeBSD doesn't need the check; madvise(2) is known to work. + */ +#else + /* Detect lazy purge runtime support. */ + if (pages_can_purge_lazy) { + bool committed = false; + void *madv_free_page = os_pages_map(NULL, PAGE, PAGE, &committed); + if (madv_free_page == NULL) { + return true; + } + assert(pages_can_purge_lazy_runtime); + if (pages_purge_lazy(madv_free_page, PAGE)) { + pages_can_purge_lazy_runtime = false; + } + os_pages_unmap(madv_free_page, PAGE); + } +#endif + + return false; +} diff --git a/deps/jemalloc/src/prng.c b/deps/jemalloc/src/prng.c new file mode 100644 index 000000000..83c04bf9b --- /dev/null +++ b/deps/jemalloc/src/prng.c @@ -0,0 +1,3 @@ +#define JEMALLOC_PRNG_C_ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" diff --git a/deps/jemalloc/src/prof.c b/deps/jemalloc/src/prof.c new file mode 100644 index 000000000..13334cb4c --- /dev/null +++ b/deps/jemalloc/src/prof.c @@ -0,0 +1,3160 @@ +#define JEMALLOC_PROF_C_ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/assert.h" +#include "jemalloc/internal/ckh.h" +#include "jemalloc/internal/hash.h" +#include "jemalloc/internal/malloc_io.h" +#include "jemalloc/internal/mutex.h" +#include "jemalloc/internal/emitter.h" + +/******************************************************************************/ + +#ifdef JEMALLOC_PROF_LIBUNWIND +#define UNW_LOCAL_ONLY +#include <libunwind.h> +#endif + +#ifdef JEMALLOC_PROF_LIBGCC +/* + * We have a circular dependency -- jemalloc_internal.h tells us if we should + * use libgcc's unwinding functionality, but after we've included that, we've + * already hooked _Unwind_Backtrace. We'll temporarily disable hooking. + */ +#undef _Unwind_Backtrace +#include <unwind.h> +#define _Unwind_Backtrace JEMALLOC_HOOK(_Unwind_Backtrace, test_hooks_libc_hook) +#endif + +/******************************************************************************/ +/* Data. */ + +bool opt_prof = false; +bool opt_prof_active = true; +bool opt_prof_thread_active_init = true; +size_t opt_lg_prof_sample = LG_PROF_SAMPLE_DEFAULT; +ssize_t opt_lg_prof_interval = LG_PROF_INTERVAL_DEFAULT; +bool opt_prof_gdump = false; +bool opt_prof_final = false; +bool opt_prof_leak = false; +bool opt_prof_accum = false; +bool opt_prof_log = false; +char opt_prof_prefix[ + /* Minimize memory bloat for non-prof builds. */ +#ifdef JEMALLOC_PROF + PATH_MAX + +#endif + 1]; + +/* + * Initialized as opt_prof_active, and accessed via + * prof_active_[gs]et{_unlocked,}(). + */ +bool prof_active; +static malloc_mutex_t prof_active_mtx; + +/* + * Initialized as opt_prof_thread_active_init, and accessed via + * prof_thread_active_init_[gs]et(). + */ +static bool prof_thread_active_init; +static malloc_mutex_t prof_thread_active_init_mtx; + +/* + * Initialized as opt_prof_gdump, and accessed via + * prof_gdump_[gs]et{_unlocked,}(). + */ +bool prof_gdump_val; +static malloc_mutex_t prof_gdump_mtx; + +uint64_t prof_interval = 0; + +size_t lg_prof_sample; + +typedef enum prof_logging_state_e prof_logging_state_t; +enum prof_logging_state_e { + prof_logging_state_stopped, + prof_logging_state_started, + prof_logging_state_dumping +}; + +/* + * - stopped: log_start never called, or previous log_stop has completed. + * - started: log_start called, log_stop not called yet. Allocations are logged. + * - dumping: log_stop called but not finished; samples are not logged anymore. + */ +prof_logging_state_t prof_logging_state = prof_logging_state_stopped; + +#ifdef JEMALLOC_JET +static bool prof_log_dummy = false; +#endif + +/* Incremented for every log file that is output. */ +static uint64_t log_seq = 0; +static char log_filename[ + /* Minimize memory bloat for non-prof builds. */ +#ifdef JEMALLOC_PROF + PATH_MAX + +#endif + 1]; + +/* Timestamp for most recent call to log_start(). */ +static nstime_t log_start_timestamp = NSTIME_ZERO_INITIALIZER; + +/* Increment these when adding to the log_bt and log_thr linked lists. */ +static size_t log_bt_index = 0; +static size_t log_thr_index = 0; + +/* Linked list node definitions. These are only used in prof.c. */ +typedef struct prof_bt_node_s prof_bt_node_t; + +struct prof_bt_node_s { + prof_bt_node_t *next; + size_t index; + prof_bt_t bt; + /* Variable size backtrace vector pointed to by bt. */ + void *vec[1]; +}; + +typedef struct prof_thr_node_s prof_thr_node_t; + +struct prof_thr_node_s { + prof_thr_node_t *next; + size_t index; + uint64_t thr_uid; + /* Variable size based on thr_name_sz. */ + char name[1]; +}; + +typedef struct prof_alloc_node_s prof_alloc_node_t; + +/* This is output when logging sampled allocations. */ +struct prof_alloc_node_s { + prof_alloc_node_t *next; + /* Indices into an array of thread data. */ + size_t alloc_thr_ind; + size_t free_thr_ind; + + /* Indices into an array of backtraces. */ + size_t alloc_bt_ind; + size_t free_bt_ind; + + uint64_t alloc_time_ns; + uint64_t free_time_ns; + + size_t usize; +}; + +/* + * Created on the first call to prof_log_start and deleted on prof_log_stop. + * These are the backtraces and threads that have already been logged by an + * allocation. + */ +static bool log_tables_initialized = false; +static ckh_t log_bt_node_set; +static ckh_t log_thr_node_set; + +/* Store linked lists for logged data. */ +static prof_bt_node_t *log_bt_first = NULL; +static prof_bt_node_t *log_bt_last = NULL; +static prof_thr_node_t *log_thr_first = NULL; +static prof_thr_node_t *log_thr_last = NULL; +static prof_alloc_node_t *log_alloc_first = NULL; +static prof_alloc_node_t *log_alloc_last = NULL; + +/* Protects the prof_logging_state and any log_{...} variable. */ +static malloc_mutex_t log_mtx; + +/* + * Table of mutexes that are shared among gctx's. These are leaf locks, so + * there is no problem with using them for more than one gctx at the same time. + * The primary motivation for this sharing though is that gctx's are ephemeral, + * and destroying mutexes causes complications for systems that allocate when + * creating/destroying mutexes. + */ +static malloc_mutex_t *gctx_locks; +static atomic_u_t cum_gctxs; /* Atomic counter. */ + +/* + * Table of mutexes that are shared among tdata's. No operations require + * holding multiple tdata locks, so there is no problem with using them for more + * than one tdata at the same time, even though a gctx lock may be acquired + * while holding a tdata lock. + */ +static malloc_mutex_t *tdata_locks; + +/* + * Global hash of (prof_bt_t *)-->(prof_gctx_t *). This is the master data + * structure that knows about all backtraces currently captured. + */ +static ckh_t bt2gctx; +/* Non static to enable profiling. */ +malloc_mutex_t bt2gctx_mtx; + +/* + * Tree of all extant prof_tdata_t structures, regardless of state, + * {attached,detached,expired}. + */ +static prof_tdata_tree_t tdatas; +static malloc_mutex_t tdatas_mtx; + +static uint64_t next_thr_uid; +static malloc_mutex_t next_thr_uid_mtx; + +static malloc_mutex_t prof_dump_seq_mtx; +static uint64_t prof_dump_seq; +static uint64_t prof_dump_iseq; +static uint64_t prof_dump_mseq; +static uint64_t prof_dump_useq; + +/* + * This buffer is rather large for stack allocation, so use a single buffer for + * all profile dumps. + */ +static malloc_mutex_t prof_dump_mtx; +static char prof_dump_buf[ + /* Minimize memory bloat for non-prof builds. */ +#ifdef JEMALLOC_PROF + PROF_DUMP_BUFSIZE +#else + 1 +#endif +]; +static size_t prof_dump_buf_end; +static int prof_dump_fd; + +/* Do not dump any profiles until bootstrapping is complete. */ +static bool prof_booted = false; + +/******************************************************************************/ +/* + * Function prototypes for static functions that are referenced prior to + * definition. + */ + +static bool prof_tctx_should_destroy(tsdn_t *tsdn, prof_tctx_t *tctx); +static void prof_tctx_destroy(tsd_t *tsd, prof_tctx_t *tctx); +static bool prof_tdata_should_destroy(tsdn_t *tsdn, prof_tdata_t *tdata, + bool even_if_attached); +static void prof_tdata_destroy(tsd_t *tsd, prof_tdata_t *tdata, + bool even_if_attached); +static char *prof_thread_name_alloc(tsdn_t *tsdn, const char *thread_name); + +/* Hashtable functions for log_bt_node_set and log_thr_node_set. */ +static void prof_thr_node_hash(const void *key, size_t r_hash[2]); +static bool prof_thr_node_keycomp(const void *k1, const void *k2); +static void prof_bt_node_hash(const void *key, size_t r_hash[2]); +static bool prof_bt_node_keycomp(const void *k1, const void *k2); + +/******************************************************************************/ +/* Red-black trees. */ + +static int +prof_tctx_comp(const prof_tctx_t *a, const prof_tctx_t *b) { + uint64_t a_thr_uid = a->thr_uid; + uint64_t b_thr_uid = b->thr_uid; + int ret = (a_thr_uid > b_thr_uid) - (a_thr_uid < b_thr_uid); + if (ret == 0) { + uint64_t a_thr_discrim = a->thr_discrim; + uint64_t b_thr_discrim = b->thr_discrim; + ret = (a_thr_discrim > b_thr_discrim) - (a_thr_discrim < + b_thr_discrim); + if (ret == 0) { + uint64_t a_tctx_uid = a->tctx_uid; + uint64_t b_tctx_uid = b->tctx_uid; + ret = (a_tctx_uid > b_tctx_uid) - (a_tctx_uid < + b_tctx_uid); + } + } + return ret; +} + +rb_gen(static UNUSED, tctx_tree_, prof_tctx_tree_t, prof_tctx_t, + tctx_link, prof_tctx_comp) + +static int +prof_gctx_comp(const prof_gctx_t *a, const prof_gctx_t *b) { + unsigned a_len = a->bt.len; + unsigned b_len = b->bt.len; + unsigned comp_len = (a_len < b_len) ? a_len : b_len; + int ret = memcmp(a->bt.vec, b->bt.vec, comp_len * sizeof(void *)); + if (ret == 0) { + ret = (a_len > b_len) - (a_len < b_len); + } + return ret; +} + +rb_gen(static UNUSED, gctx_tree_, prof_gctx_tree_t, prof_gctx_t, dump_link, + prof_gctx_comp) + +static int +prof_tdata_comp(const prof_tdata_t *a, const prof_tdata_t *b) { + int ret; + uint64_t a_uid = a->thr_uid; + uint64_t b_uid = b->thr_uid; + + ret = ((a_uid > b_uid) - (a_uid < b_uid)); + if (ret == 0) { + uint64_t a_discrim = a->thr_discrim; + uint64_t b_discrim = b->thr_discrim; + + ret = ((a_discrim > b_discrim) - (a_discrim < b_discrim)); + } + return ret; +} + +rb_gen(static UNUSED, tdata_tree_, prof_tdata_tree_t, prof_tdata_t, tdata_link, + prof_tdata_comp) + +/******************************************************************************/ + +void +prof_alloc_rollback(tsd_t *tsd, prof_tctx_t *tctx, bool updated) { + prof_tdata_t *tdata; + + cassert(config_prof); + + if (updated) { + /* + * Compute a new sample threshold. This isn't very important in + * practice, because this function is rarely executed, so the + * potential for sample bias is minimal except in contrived + * programs. + */ + tdata = prof_tdata_get(tsd, true); + if (tdata != NULL) { + prof_sample_threshold_update(tdata); + } + } + + if ((uintptr_t)tctx > (uintptr_t)1U) { + malloc_mutex_lock(tsd_tsdn(tsd), tctx->tdata->lock); + tctx->prepared = false; + if (prof_tctx_should_destroy(tsd_tsdn(tsd), tctx)) { + prof_tctx_destroy(tsd, tctx); + } else { + malloc_mutex_unlock(tsd_tsdn(tsd), tctx->tdata->lock); + } + } +} + +void +prof_malloc_sample_object(tsdn_t *tsdn, const void *ptr, size_t usize, + prof_tctx_t *tctx) { + prof_tctx_set(tsdn, ptr, usize, NULL, tctx); + + /* Get the current time and set this in the extent_t. We'll read this + * when free() is called. */ + nstime_t t = NSTIME_ZERO_INITIALIZER; + nstime_update(&t); + prof_alloc_time_set(tsdn, ptr, NULL, t); + + malloc_mutex_lock(tsdn, tctx->tdata->lock); + tctx->cnts.curobjs++; + tctx->cnts.curbytes += usize; + if (opt_prof_accum) { + tctx->cnts.accumobjs++; + tctx->cnts.accumbytes += usize; + } + tctx->prepared = false; + malloc_mutex_unlock(tsdn, tctx->tdata->lock); +} + +static size_t +prof_log_bt_index(tsd_t *tsd, prof_bt_t *bt) { + assert(prof_logging_state == prof_logging_state_started); + malloc_mutex_assert_owner(tsd_tsdn(tsd), &log_mtx); + + prof_bt_node_t dummy_node; + dummy_node.bt = *bt; + prof_bt_node_t *node; + + /* See if this backtrace is already cached in the table. */ + if (ckh_search(&log_bt_node_set, (void *)(&dummy_node), + (void **)(&node), NULL)) { + size_t sz = offsetof(prof_bt_node_t, vec) + + (bt->len * sizeof(void *)); + prof_bt_node_t *new_node = (prof_bt_node_t *) + iallocztm(tsd_tsdn(tsd), sz, sz_size2index(sz), false, NULL, + true, arena_get(TSDN_NULL, 0, true), true); + if (log_bt_first == NULL) { + log_bt_first = new_node; + log_bt_last = new_node; + } else { + log_bt_last->next = new_node; + log_bt_last = new_node; + } + + new_node->next = NULL; + new_node->index = log_bt_index; + /* + * Copy the backtrace: bt is inside a tdata or gctx, which + * might die before prof_log_stop is called. + */ + new_node->bt.len = bt->len; + memcpy(new_node->vec, bt->vec, bt->len * sizeof(void *)); + new_node->bt.vec = new_node->vec; + + log_bt_index++; + ckh_insert(tsd, &log_bt_node_set, (void *)new_node, NULL); + return new_node->index; + } else { + return node->index; + } +} +static size_t +prof_log_thr_index(tsd_t *tsd, uint64_t thr_uid, const char *name) { + assert(prof_logging_state == prof_logging_state_started); + malloc_mutex_assert_owner(tsd_tsdn(tsd), &log_mtx); + + prof_thr_node_t dummy_node; + dummy_node.thr_uid = thr_uid; + prof_thr_node_t *node; + + /* See if this thread is already cached in the table. */ + if (ckh_search(&log_thr_node_set, (void *)(&dummy_node), + (void **)(&node), NULL)) { + size_t sz = offsetof(prof_thr_node_t, name) + strlen(name) + 1; + prof_thr_node_t *new_node = (prof_thr_node_t *) + iallocztm(tsd_tsdn(tsd), sz, sz_size2index(sz), false, NULL, + true, arena_get(TSDN_NULL, 0, true), true); + if (log_thr_first == NULL) { + log_thr_first = new_node; + log_thr_last = new_node; + } else { + log_thr_last->next = new_node; + log_thr_last = new_node; + } + + new_node->next = NULL; + new_node->index = log_thr_index; + new_node->thr_uid = thr_uid; + strcpy(new_node->name, name); + + log_thr_index++; + ckh_insert(tsd, &log_thr_node_set, (void *)new_node, NULL); + return new_node->index; + } else { + return node->index; + } +} + +static void +prof_try_log(tsd_t *tsd, const void *ptr, size_t usize, prof_tctx_t *tctx) { + malloc_mutex_assert_owner(tsd_tsdn(tsd), tctx->tdata->lock); + + prof_tdata_t *cons_tdata = prof_tdata_get(tsd, false); + if (cons_tdata == NULL) { + /* + * We decide not to log these allocations. cons_tdata will be + * NULL only when the current thread is in a weird state (e.g. + * it's being destroyed). + */ + return; + } + + malloc_mutex_lock(tsd_tsdn(tsd), &log_mtx); + + if (prof_logging_state != prof_logging_state_started) { + goto label_done; + } + + if (!log_tables_initialized) { + bool err1 = ckh_new(tsd, &log_bt_node_set, PROF_CKH_MINITEMS, + prof_bt_node_hash, prof_bt_node_keycomp); + bool err2 = ckh_new(tsd, &log_thr_node_set, PROF_CKH_MINITEMS, + prof_thr_node_hash, prof_thr_node_keycomp); + if (err1 || err2) { + goto label_done; + } + log_tables_initialized = true; + } + + nstime_t alloc_time = prof_alloc_time_get(tsd_tsdn(tsd), ptr, + (alloc_ctx_t *)NULL); + nstime_t free_time = NSTIME_ZERO_INITIALIZER; + nstime_update(&free_time); + + size_t sz = sizeof(prof_alloc_node_t); + prof_alloc_node_t *new_node = (prof_alloc_node_t *) + iallocztm(tsd_tsdn(tsd), sz, sz_size2index(sz), false, NULL, true, + arena_get(TSDN_NULL, 0, true), true); + + const char *prod_thr_name = (tctx->tdata->thread_name == NULL)? + "" : tctx->tdata->thread_name; + const char *cons_thr_name = prof_thread_name_get(tsd); + + prof_bt_t bt; + /* Initialize the backtrace, using the buffer in tdata to store it. */ + bt_init(&bt, cons_tdata->vec); + prof_backtrace(&bt); + prof_bt_t *cons_bt = &bt; + + /* We haven't destroyed tctx yet, so gctx should be good to read. */ + prof_bt_t *prod_bt = &tctx->gctx->bt; + + new_node->next = NULL; + new_node->alloc_thr_ind = prof_log_thr_index(tsd, tctx->tdata->thr_uid, + prod_thr_name); + new_node->free_thr_ind = prof_log_thr_index(tsd, cons_tdata->thr_uid, + cons_thr_name); + new_node->alloc_bt_ind = prof_log_bt_index(tsd, prod_bt); + new_node->free_bt_ind = prof_log_bt_index(tsd, cons_bt); + new_node->alloc_time_ns = nstime_ns(&alloc_time); + new_node->free_time_ns = nstime_ns(&free_time); + new_node->usize = usize; + + if (log_alloc_first == NULL) { + log_alloc_first = new_node; + log_alloc_last = new_node; + } else { + log_alloc_last->next = new_node; + log_alloc_last = new_node; + } + +label_done: + malloc_mutex_unlock(tsd_tsdn(tsd), &log_mtx); +} + +void +prof_free_sampled_object(tsd_t *tsd, const void *ptr, size_t usize, + prof_tctx_t *tctx) { + malloc_mutex_lock(tsd_tsdn(tsd), tctx->tdata->lock); + + assert(tctx->cnts.curobjs > 0); + assert(tctx->cnts.curbytes >= usize); + tctx->cnts.curobjs--; + tctx->cnts.curbytes -= usize; + + prof_try_log(tsd, ptr, usize, tctx); + + if (prof_tctx_should_destroy(tsd_tsdn(tsd), tctx)) { + prof_tctx_destroy(tsd, tctx); + } else { + malloc_mutex_unlock(tsd_tsdn(tsd), tctx->tdata->lock); + } +} + +void +bt_init(prof_bt_t *bt, void **vec) { + cassert(config_prof); + + bt->vec = vec; + bt->len = 0; +} + +static void +prof_enter(tsd_t *tsd, prof_tdata_t *tdata) { + cassert(config_prof); + assert(tdata == prof_tdata_get(tsd, false)); + + if (tdata != NULL) { + assert(!tdata->enq); + tdata->enq = true; + } + + malloc_mutex_lock(tsd_tsdn(tsd), &bt2gctx_mtx); +} + +static void +prof_leave(tsd_t *tsd, prof_tdata_t *tdata) { + cassert(config_prof); + assert(tdata == prof_tdata_get(tsd, false)); + + malloc_mutex_unlock(tsd_tsdn(tsd), &bt2gctx_mtx); + + if (tdata != NULL) { + bool idump, gdump; + + assert(tdata->enq); + tdata->enq = false; + idump = tdata->enq_idump; + tdata->enq_idump = false; + gdump = tdata->enq_gdump; + tdata->enq_gdump = false; + + if (idump) { + prof_idump(tsd_tsdn(tsd)); + } + if (gdump) { + prof_gdump(tsd_tsdn(tsd)); + } + } +} + +#ifdef JEMALLOC_PROF_LIBUNWIND +void +prof_backtrace(prof_bt_t *bt) { + int nframes; + + cassert(config_prof); + assert(bt->len == 0); + assert(bt->vec != NULL); + + nframes = unw_backtrace(bt->vec, PROF_BT_MAX); + if (nframes <= 0) { + return; + } + bt->len = nframes; +} +#elif (defined(JEMALLOC_PROF_LIBGCC)) +static _Unwind_Reason_Code +prof_unwind_init_callback(struct _Unwind_Context *context, void *arg) { + cassert(config_prof); + + return _URC_NO_REASON; +} + +static _Unwind_Reason_Code +prof_unwind_callback(struct _Unwind_Context *context, void *arg) { + prof_unwind_data_t *data = (prof_unwind_data_t *)arg; + void *ip; + + cassert(config_prof); + + ip = (void *)_Unwind_GetIP(context); + if (ip == NULL) { + return _URC_END_OF_STACK; + } + data->bt->vec[data->bt->len] = ip; + data->bt->len++; + if (data->bt->len == data->max) { + return _URC_END_OF_STACK; + } + + return _URC_NO_REASON; +} + +void +prof_backtrace(prof_bt_t *bt) { + prof_unwind_data_t data = {bt, PROF_BT_MAX}; + + cassert(config_prof); + + _Unwind_Backtrace(prof_unwind_callback, &data); +} +#elif (defined(JEMALLOC_PROF_GCC)) +void +prof_backtrace(prof_bt_t *bt) { +#define BT_FRAME(i) \ + if ((i) < PROF_BT_MAX) { \ + void *p; \ + if (__builtin_frame_address(i) == 0) { \ + return; \ + } \ + p = __builtin_return_address(i); \ + if (p == NULL) { \ + return; \ + } \ + bt->vec[(i)] = p; \ + bt->len = (i) + 1; \ + } else { \ + return; \ + } + + cassert(config_prof); + + BT_FRAME(0) + BT_FRAME(1) + BT_FRAME(2) + BT_FRAME(3) + BT_FRAME(4) + BT_FRAME(5) + BT_FRAME(6) + BT_FRAME(7) + BT_FRAME(8) + BT_FRAME(9) + + BT_FRAME(10) + BT_FRAME(11) + BT_FRAME(12) + BT_FRAME(13) + BT_FRAME(14) + BT_FRAME(15) + BT_FRAME(16) + BT_FRAME(17) + BT_FRAME(18) + BT_FRAME(19) + + BT_FRAME(20) + BT_FRAME(21) + BT_FRAME(22) + BT_FRAME(23) + BT_FRAME(24) + BT_FRAME(25) + BT_FRAME(26) + BT_FRAME(27) + BT_FRAME(28) + BT_FRAME(29) + + BT_FRAME(30) + BT_FRAME(31) + BT_FRAME(32) + BT_FRAME(33) + BT_FRAME(34) + BT_FRAME(35) + BT_FRAME(36) + BT_FRAME(37) + BT_FRAME(38) + BT_FRAME(39) + + BT_FRAME(40) + BT_FRAME(41) + BT_FRAME(42) + BT_FRAME(43) + BT_FRAME(44) + BT_FRAME(45) + BT_FRAME(46) + BT_FRAME(47) + BT_FRAME(48) + BT_FRAME(49) + + BT_FRAME(50) + BT_FRAME(51) + BT_FRAME(52) + BT_FRAME(53) + BT_FRAME(54) + BT_FRAME(55) + BT_FRAME(56) + BT_FRAME(57) + BT_FRAME(58) + BT_FRAME(59) + + BT_FRAME(60) + BT_FRAME(61) + BT_FRAME(62) + BT_FRAME(63) + BT_FRAME(64) + BT_FRAME(65) + BT_FRAME(66) + BT_FRAME(67) + BT_FRAME(68) + BT_FRAME(69) + + BT_FRAME(70) + BT_FRAME(71) + BT_FRAME(72) + BT_FRAME(73) + BT_FRAME(74) + BT_FRAME(75) + BT_FRAME(76) + BT_FRAME(77) + BT_FRAME(78) + BT_FRAME(79) + + BT_FRAME(80) + BT_FRAME(81) + BT_FRAME(82) + BT_FRAME(83) + BT_FRAME(84) + BT_FRAME(85) + BT_FRAME(86) + BT_FRAME(87) + BT_FRAME(88) + BT_FRAME(89) + + BT_FRAME(90) + BT_FRAME(91) + BT_FRAME(92) + BT_FRAME(93) + BT_FRAME(94) + BT_FRAME(95) + BT_FRAME(96) + BT_FRAME(97) + BT_FRAME(98) + BT_FRAME(99) + + BT_FRAME(100) + BT_FRAME(101) + BT_FRAME(102) + BT_FRAME(103) + BT_FRAME(104) + BT_FRAME(105) + BT_FRAME(106) + BT_FRAME(107) + BT_FRAME(108) + BT_FRAME(109) + + BT_FRAME(110) + BT_FRAME(111) + BT_FRAME(112) + BT_FRAME(113) + BT_FRAME(114) + BT_FRAME(115) + BT_FRAME(116) + BT_FRAME(117) + BT_FRAME(118) + BT_FRAME(119) + + BT_FRAME(120) + BT_FRAME(121) + BT_FRAME(122) + BT_FRAME(123) + BT_FRAME(124) + BT_FRAME(125) + BT_FRAME(126) + BT_FRAME(127) +#undef BT_FRAME +} +#else +void +prof_backtrace(prof_bt_t *bt) { + cassert(config_prof); + not_reached(); +} +#endif + +static malloc_mutex_t * +prof_gctx_mutex_choose(void) { + unsigned ngctxs = atomic_fetch_add_u(&cum_gctxs, 1, ATOMIC_RELAXED); + + return &gctx_locks[(ngctxs - 1) % PROF_NCTX_LOCKS]; +} + +static malloc_mutex_t * +prof_tdata_mutex_choose(uint64_t thr_uid) { + return &tdata_locks[thr_uid % PROF_NTDATA_LOCKS]; +} + +static prof_gctx_t * +prof_gctx_create(tsdn_t *tsdn, prof_bt_t *bt) { + /* + * Create a single allocation that has space for vec of length bt->len. + */ + size_t size = offsetof(prof_gctx_t, vec) + (bt->len * sizeof(void *)); + prof_gctx_t *gctx = (prof_gctx_t *)iallocztm(tsdn, size, + sz_size2index(size), false, NULL, true, arena_get(TSDN_NULL, 0, true), + true); + if (gctx == NULL) { + return NULL; + } + gctx->lock = prof_gctx_mutex_choose(); + /* + * Set nlimbo to 1, in order to avoid a race condition with + * prof_tctx_destroy()/prof_gctx_try_destroy(). + */ + gctx->nlimbo = 1; + tctx_tree_new(&gctx->tctxs); + /* Duplicate bt. */ + memcpy(gctx->vec, bt->vec, bt->len * sizeof(void *)); + gctx->bt.vec = gctx->vec; + gctx->bt.len = bt->len; + return gctx; +} + +static void +prof_gctx_try_destroy(tsd_t *tsd, prof_tdata_t *tdata_self, prof_gctx_t *gctx, + prof_tdata_t *tdata) { + cassert(config_prof); + + /* + * Check that gctx is still unused by any thread cache before destroying + * it. prof_lookup() increments gctx->nlimbo in order to avoid a race + * condition with this function, as does prof_tctx_destroy() in order to + * avoid a race between the main body of prof_tctx_destroy() and entry + * into this function. + */ + prof_enter(tsd, tdata_self); + malloc_mutex_lock(tsd_tsdn(tsd), gctx->lock); + assert(gctx->nlimbo != 0); + if (tctx_tree_empty(&gctx->tctxs) && gctx->nlimbo == 1) { + /* Remove gctx from bt2gctx. */ + if (ckh_remove(tsd, &bt2gctx, &gctx->bt, NULL, NULL)) { + not_reached(); + } + prof_leave(tsd, tdata_self); + /* Destroy gctx. */ + malloc_mutex_unlock(tsd_tsdn(tsd), gctx->lock); + idalloctm(tsd_tsdn(tsd), gctx, NULL, NULL, true, true); + } else { + /* + * Compensate for increment in prof_tctx_destroy() or + * prof_lookup(). + */ + gctx->nlimbo--; + malloc_mutex_unlock(tsd_tsdn(tsd), gctx->lock); + prof_leave(tsd, tdata_self); + } +} + +static bool +prof_tctx_should_destroy(tsdn_t *tsdn, prof_tctx_t *tctx) { + malloc_mutex_assert_owner(tsdn, tctx->tdata->lock); + + if (opt_prof_accum) { + return false; + } + if (tctx->cnts.curobjs != 0) { + return false; + } + if (tctx->prepared) { + return false; + } + return true; +} + +static bool +prof_gctx_should_destroy(prof_gctx_t *gctx) { + if (opt_prof_accum) { + return false; + } + if (!tctx_tree_empty(&gctx->tctxs)) { + return false; + } + if (gctx->nlimbo != 0) { + return false; + } + return true; +} + +static void +prof_tctx_destroy(tsd_t *tsd, prof_tctx_t *tctx) { + prof_tdata_t *tdata = tctx->tdata; + prof_gctx_t *gctx = tctx->gctx; + bool destroy_tdata, destroy_tctx, destroy_gctx; + + malloc_mutex_assert_owner(tsd_tsdn(tsd), tctx->tdata->lock); + + assert(tctx->cnts.curobjs == 0); + assert(tctx->cnts.curbytes == 0); + assert(!opt_prof_accum); + assert(tctx->cnts.accumobjs == 0); + assert(tctx->cnts.accumbytes == 0); + + ckh_remove(tsd, &tdata->bt2tctx, &gctx->bt, NULL, NULL); + destroy_tdata = prof_tdata_should_destroy(tsd_tsdn(tsd), tdata, false); + malloc_mutex_unlock(tsd_tsdn(tsd), tdata->lock); + + malloc_mutex_lock(tsd_tsdn(tsd), gctx->lock); + switch (tctx->state) { + case prof_tctx_state_nominal: + tctx_tree_remove(&gctx->tctxs, tctx); + destroy_tctx = true; + if (prof_gctx_should_destroy(gctx)) { + /* + * Increment gctx->nlimbo in order to keep another + * thread from winning the race to destroy gctx while + * this one has gctx->lock dropped. Without this, it + * would be possible for another thread to: + * + * 1) Sample an allocation associated with gctx. + * 2) Deallocate the sampled object. + * 3) Successfully prof_gctx_try_destroy(gctx). + * + * The result would be that gctx no longer exists by the + * time this thread accesses it in + * prof_gctx_try_destroy(). + */ + gctx->nlimbo++; + destroy_gctx = true; + } else { + destroy_gctx = false; + } + break; + case prof_tctx_state_dumping: + /* + * A dumping thread needs tctx to remain valid until dumping + * has finished. Change state such that the dumping thread will + * complete destruction during a late dump iteration phase. + */ + tctx->state = prof_tctx_state_purgatory; + destroy_tctx = false; + destroy_gctx = false; + break; + default: + not_reached(); + destroy_tctx = false; + destroy_gctx = false; + } + malloc_mutex_unlock(tsd_tsdn(tsd), gctx->lock); + if (destroy_gctx) { + prof_gctx_try_destroy(tsd, prof_tdata_get(tsd, false), gctx, + tdata); + } + + malloc_mutex_assert_not_owner(tsd_tsdn(tsd), tctx->tdata->lock); + + if (destroy_tdata) { + prof_tdata_destroy(tsd, tdata, false); + } + + if (destroy_tctx) { + idalloctm(tsd_tsdn(tsd), tctx, NULL, NULL, true, true); + } +} + +static bool +prof_lookup_global(tsd_t *tsd, prof_bt_t *bt, prof_tdata_t *tdata, + void **p_btkey, prof_gctx_t **p_gctx, bool *p_new_gctx) { + union { + prof_gctx_t *p; + void *v; + } gctx, tgctx; + union { + prof_bt_t *p; + void *v; + } btkey; + bool new_gctx; + + prof_enter(tsd, tdata); + if (ckh_search(&bt2gctx, bt, &btkey.v, &gctx.v)) { + /* bt has never been seen before. Insert it. */ + prof_leave(tsd, tdata); + tgctx.p = prof_gctx_create(tsd_tsdn(tsd), bt); + if (tgctx.v == NULL) { + return true; + } + prof_enter(tsd, tdata); + if (ckh_search(&bt2gctx, bt, &btkey.v, &gctx.v)) { + gctx.p = tgctx.p; + btkey.p = &gctx.p->bt; + if (ckh_insert(tsd, &bt2gctx, btkey.v, gctx.v)) { + /* OOM. */ + prof_leave(tsd, tdata); + idalloctm(tsd_tsdn(tsd), gctx.v, NULL, NULL, + true, true); + return true; + } + new_gctx = true; + } else { + new_gctx = false; + } + } else { + tgctx.v = NULL; + new_gctx = false; + } + + if (!new_gctx) { + /* + * Increment nlimbo, in order to avoid a race condition with + * prof_tctx_destroy()/prof_gctx_try_destroy(). + */ + malloc_mutex_lock(tsd_tsdn(tsd), gctx.p->lock); + gctx.p->nlimbo++; + malloc_mutex_unlock(tsd_tsdn(tsd), gctx.p->lock); + new_gctx = false; + + if (tgctx.v != NULL) { + /* Lost race to insert. */ + idalloctm(tsd_tsdn(tsd), tgctx.v, NULL, NULL, true, + true); + } + } + prof_leave(tsd, tdata); + + *p_btkey = btkey.v; + *p_gctx = gctx.p; + *p_new_gctx = new_gctx; + return false; +} + +prof_tctx_t * +prof_lookup(tsd_t *tsd, prof_bt_t *bt) { + union { + prof_tctx_t *p; + void *v; + } ret; + prof_tdata_t *tdata; + bool not_found; + + cassert(config_prof); + + tdata = prof_tdata_get(tsd, false); + if (tdata == NULL) { + return NULL; + } + + malloc_mutex_lock(tsd_tsdn(tsd), tdata->lock); + not_found = ckh_search(&tdata->bt2tctx, bt, NULL, &ret.v); + if (!not_found) { /* Note double negative! */ + ret.p->prepared = true; + } + malloc_mutex_unlock(tsd_tsdn(tsd), tdata->lock); + if (not_found) { + void *btkey; + prof_gctx_t *gctx; + bool new_gctx, error; + + /* + * This thread's cache lacks bt. Look for it in the global + * cache. + */ + if (prof_lookup_global(tsd, bt, tdata, &btkey, &gctx, + &new_gctx)) { + return NULL; + } + + /* Link a prof_tctx_t into gctx for this thread. */ + ret.v = iallocztm(tsd_tsdn(tsd), sizeof(prof_tctx_t), + sz_size2index(sizeof(prof_tctx_t)), false, NULL, true, + arena_ichoose(tsd, NULL), true); + if (ret.p == NULL) { + if (new_gctx) { + prof_gctx_try_destroy(tsd, tdata, gctx, tdata); + } + return NULL; + } + ret.p->tdata = tdata; + ret.p->thr_uid = tdata->thr_uid; + ret.p->thr_discrim = tdata->thr_discrim; + memset(&ret.p->cnts, 0, sizeof(prof_cnt_t)); + ret.p->gctx = gctx; + ret.p->tctx_uid = tdata->tctx_uid_next++; + ret.p->prepared = true; + ret.p->state = prof_tctx_state_initializing; + malloc_mutex_lock(tsd_tsdn(tsd), tdata->lock); + error = ckh_insert(tsd, &tdata->bt2tctx, btkey, ret.v); + malloc_mutex_unlock(tsd_tsdn(tsd), tdata->lock); + if (error) { + if (new_gctx) { + prof_gctx_try_destroy(tsd, tdata, gctx, tdata); + } + idalloctm(tsd_tsdn(tsd), ret.v, NULL, NULL, true, true); + return NULL; + } + malloc_mutex_lock(tsd_tsdn(tsd), gctx->lock); + ret.p->state = prof_tctx_state_nominal; + tctx_tree_insert(&gctx->tctxs, ret.p); + gctx->nlimbo--; + malloc_mutex_unlock(tsd_tsdn(tsd), gctx->lock); + } + + return ret.p; +} + +/* + * The bodies of this function and prof_leakcheck() are compiled out unless heap + * profiling is enabled, so that it is possible to compile jemalloc with + * floating point support completely disabled. Avoiding floating point code is + * important on memory-constrained systems, but it also enables a workaround for + * versions of glibc that don't properly save/restore floating point registers + * during dynamic lazy symbol loading (which internally calls into whatever + * malloc implementation happens to be integrated into the application). Note + * that some compilers (e.g. gcc 4.8) may use floating point registers for fast + * memory moves, so jemalloc must be compiled with such optimizations disabled + * (e.g. + * -mno-sse) in order for the workaround to be complete. + */ +void +prof_sample_threshold_update(prof_tdata_t *tdata) { +#ifdef JEMALLOC_PROF + if (!config_prof) { + return; + } + + if (lg_prof_sample == 0) { + tsd_bytes_until_sample_set(tsd_fetch(), 0); + return; + } + + /* + * Compute sample interval as a geometrically distributed random + * variable with mean (2^lg_prof_sample). + * + * __ __ + * | log(u) | 1 + * tdata->bytes_until_sample = | -------- |, where p = --------------- + * | log(1-p) | lg_prof_sample + * 2 + * + * For more information on the math, see: + * + * Non-Uniform Random Variate Generation + * Luc Devroye + * Springer-Verlag, New York, 1986 + * pp 500 + * (http://luc.devroye.org/rnbookindex.html) + */ + uint64_t r = prng_lg_range_u64(&tdata->prng_state, 53); + double u = (double)r * (1.0/9007199254740992.0L); + uint64_t bytes_until_sample = (uint64_t)(log(u) / + log(1.0 - (1.0 / (double)((uint64_t)1U << lg_prof_sample)))) + + (uint64_t)1U; + if (bytes_until_sample > SSIZE_MAX) { + bytes_until_sample = SSIZE_MAX; + } + tsd_bytes_until_sample_set(tsd_fetch(), bytes_until_sample); + +#endif +} + +#ifdef JEMALLOC_JET +static prof_tdata_t * +prof_tdata_count_iter(prof_tdata_tree_t *tdatas, prof_tdata_t *tdata, + void *arg) { + size_t *tdata_count = (size_t *)arg; + + (*tdata_count)++; + + return NULL; +} + +size_t +prof_tdata_count(void) { + size_t tdata_count = 0; + tsdn_t *tsdn; + + tsdn = tsdn_fetch(); + malloc_mutex_lock(tsdn, &tdatas_mtx); + tdata_tree_iter(&tdatas, NULL, prof_tdata_count_iter, + (void *)&tdata_count); + malloc_mutex_unlock(tsdn, &tdatas_mtx); + + return tdata_count; +} + +size_t +prof_bt_count(void) { + size_t bt_count; + tsd_t *tsd; + prof_tdata_t *tdata; + + tsd = tsd_fetch(); + tdata = prof_tdata_get(tsd, false); + if (tdata == NULL) { + return 0; + } + + malloc_mutex_lock(tsd_tsdn(tsd), &bt2gctx_mtx); + bt_count = ckh_count(&bt2gctx); + malloc_mutex_unlock(tsd_tsdn(tsd), &bt2gctx_mtx); + + return bt_count; +} +#endif + +static int +prof_dump_open_impl(bool propagate_err, const char *filename) { + int fd; + + fd = creat(filename, 0644); + if (fd == -1 && !propagate_err) { + malloc_printf("<jemalloc>: creat(\"%s\"), 0644) failed\n", + filename); + if (opt_abort) { + abort(); + } + } + + return fd; +} +prof_dump_open_t *JET_MUTABLE prof_dump_open = prof_dump_open_impl; + +static bool +prof_dump_flush(bool propagate_err) { + bool ret = false; + ssize_t err; + + cassert(config_prof); + + err = malloc_write_fd(prof_dump_fd, prof_dump_buf, prof_dump_buf_end); + if (err == -1) { + if (!propagate_err) { + malloc_write("<jemalloc>: write() failed during heap " + "profile flush\n"); + if (opt_abort) { + abort(); + } + } + ret = true; + } + prof_dump_buf_end = 0; + + return ret; +} + +static bool +prof_dump_close(bool propagate_err) { + bool ret; + + assert(prof_dump_fd != -1); + ret = prof_dump_flush(propagate_err); + close(prof_dump_fd); + prof_dump_fd = -1; + + return ret; +} + +static bool +prof_dump_write(bool propagate_err, const char *s) { + size_t i, slen, n; + + cassert(config_prof); + + i = 0; + slen = strlen(s); + while (i < slen) { + /* Flush the buffer if it is full. */ + if (prof_dump_buf_end == PROF_DUMP_BUFSIZE) { + if (prof_dump_flush(propagate_err) && propagate_err) { + return true; + } + } + + if (prof_dump_buf_end + slen - i <= PROF_DUMP_BUFSIZE) { + /* Finish writing. */ + n = slen - i; + } else { + /* Write as much of s as will fit. */ + n = PROF_DUMP_BUFSIZE - prof_dump_buf_end; + } + memcpy(&prof_dump_buf[prof_dump_buf_end], &s[i], n); + prof_dump_buf_end += n; + i += n; + } + assert(i == slen); + + return false; +} + +JEMALLOC_FORMAT_PRINTF(2, 3) +static bool +prof_dump_printf(bool propagate_err, const char *format, ...) { + bool ret; + va_list ap; + char buf[PROF_PRINTF_BUFSIZE]; + + va_start(ap, format); + malloc_vsnprintf(buf, sizeof(buf), format, ap); + va_end(ap); + ret = prof_dump_write(propagate_err, buf); + + return ret; +} + +static void +prof_tctx_merge_tdata(tsdn_t *tsdn, prof_tctx_t *tctx, prof_tdata_t *tdata) { + malloc_mutex_assert_owner(tsdn, tctx->tdata->lock); + + malloc_mutex_lock(tsdn, tctx->gctx->lock); + + switch (tctx->state) { + case prof_tctx_state_initializing: + malloc_mutex_unlock(tsdn, tctx->gctx->lock); + return; + case prof_tctx_state_nominal: + tctx->state = prof_tctx_state_dumping; + malloc_mutex_unlock(tsdn, tctx->gctx->lock); + + memcpy(&tctx->dump_cnts, &tctx->cnts, sizeof(prof_cnt_t)); + + tdata->cnt_summed.curobjs += tctx->dump_cnts.curobjs; + tdata->cnt_summed.curbytes += tctx->dump_cnts.curbytes; + if (opt_prof_accum) { + tdata->cnt_summed.accumobjs += + tctx->dump_cnts.accumobjs; + tdata->cnt_summed.accumbytes += + tctx->dump_cnts.accumbytes; + } + break; + case prof_tctx_state_dumping: + case prof_tctx_state_purgatory: + not_reached(); + } +} + +static void +prof_tctx_merge_gctx(tsdn_t *tsdn, prof_tctx_t *tctx, prof_gctx_t *gctx) { + malloc_mutex_assert_owner(tsdn, gctx->lock); + + gctx->cnt_summed.curobjs += tctx->dump_cnts.curobjs; + gctx->cnt_summed.curbytes += tctx->dump_cnts.curbytes; + if (opt_prof_accum) { + gctx->cnt_summed.accumobjs += tctx->dump_cnts.accumobjs; + gctx->cnt_summed.accumbytes += tctx->dump_cnts.accumbytes; + } +} + +static prof_tctx_t * +prof_tctx_merge_iter(prof_tctx_tree_t *tctxs, prof_tctx_t *tctx, void *arg) { + tsdn_t *tsdn = (tsdn_t *)arg; + + malloc_mutex_assert_owner(tsdn, tctx->gctx->lock); + + switch (tctx->state) { + case prof_tctx_state_nominal: + /* New since dumping started; ignore. */ + break; + case prof_tctx_state_dumping: + case prof_tctx_state_purgatory: + prof_tctx_merge_gctx(tsdn, tctx, tctx->gctx); + break; + default: + not_reached(); + } + + return NULL; +} + +struct prof_tctx_dump_iter_arg_s { + tsdn_t *tsdn; + bool propagate_err; +}; + +static prof_tctx_t * +prof_tctx_dump_iter(prof_tctx_tree_t *tctxs, prof_tctx_t *tctx, void *opaque) { + struct prof_tctx_dump_iter_arg_s *arg = + (struct prof_tctx_dump_iter_arg_s *)opaque; + + malloc_mutex_assert_owner(arg->tsdn, tctx->gctx->lock); + + switch (tctx->state) { + case prof_tctx_state_initializing: + case prof_tctx_state_nominal: + /* Not captured by this dump. */ + break; + case prof_tctx_state_dumping: + case prof_tctx_state_purgatory: + if (prof_dump_printf(arg->propagate_err, + " t%"FMTu64": %"FMTu64": %"FMTu64" [%"FMTu64": " + "%"FMTu64"]\n", tctx->thr_uid, tctx->dump_cnts.curobjs, + tctx->dump_cnts.curbytes, tctx->dump_cnts.accumobjs, + tctx->dump_cnts.accumbytes)) { + return tctx; + } + break; + default: + not_reached(); + } + return NULL; +} + +static prof_tctx_t * +prof_tctx_finish_iter(prof_tctx_tree_t *tctxs, prof_tctx_t *tctx, void *arg) { + tsdn_t *tsdn = (tsdn_t *)arg; + prof_tctx_t *ret; + + malloc_mutex_assert_owner(tsdn, tctx->gctx->lock); + + switch (tctx->state) { + case prof_tctx_state_nominal: + /* New since dumping started; ignore. */ + break; + case prof_tctx_state_dumping: + tctx->state = prof_tctx_state_nominal; + break; + case prof_tctx_state_purgatory: + ret = tctx; + goto label_return; + default: + not_reached(); + } + + ret = NULL; +label_return: + return ret; +} + +static void +prof_dump_gctx_prep(tsdn_t *tsdn, prof_gctx_t *gctx, prof_gctx_tree_t *gctxs) { + cassert(config_prof); + + malloc_mutex_lock(tsdn, gctx->lock); + + /* + * Increment nlimbo so that gctx won't go away before dump. + * Additionally, link gctx into the dump list so that it is included in + * prof_dump()'s second pass. + */ + gctx->nlimbo++; + gctx_tree_insert(gctxs, gctx); + + memset(&gctx->cnt_summed, 0, sizeof(prof_cnt_t)); + + malloc_mutex_unlock(tsdn, gctx->lock); +} + +struct prof_gctx_merge_iter_arg_s { + tsdn_t *tsdn; + size_t leak_ngctx; +}; + +static prof_gctx_t * +prof_gctx_merge_iter(prof_gctx_tree_t *gctxs, prof_gctx_t *gctx, void *opaque) { + struct prof_gctx_merge_iter_arg_s *arg = + (struct prof_gctx_merge_iter_arg_s *)opaque; + + malloc_mutex_lock(arg->tsdn, gctx->lock); + tctx_tree_iter(&gctx->tctxs, NULL, prof_tctx_merge_iter, + (void *)arg->tsdn); + if (gctx->cnt_summed.curobjs != 0) { + arg->leak_ngctx++; + } + malloc_mutex_unlock(arg->tsdn, gctx->lock); + + return NULL; +} + +static void +prof_gctx_finish(tsd_t *tsd, prof_gctx_tree_t *gctxs) { + prof_tdata_t *tdata = prof_tdata_get(tsd, false); + prof_gctx_t *gctx; + + /* + * Standard tree iteration won't work here, because as soon as we + * decrement gctx->nlimbo and unlock gctx, another thread can + * concurrently destroy it, which will corrupt the tree. Therefore, + * tear down the tree one node at a time during iteration. + */ + while ((gctx = gctx_tree_first(gctxs)) != NULL) { + gctx_tree_remove(gctxs, gctx); + malloc_mutex_lock(tsd_tsdn(tsd), gctx->lock); + { + prof_tctx_t *next; + + next = NULL; + do { + prof_tctx_t *to_destroy = + tctx_tree_iter(&gctx->tctxs, next, + prof_tctx_finish_iter, + (void *)tsd_tsdn(tsd)); + if (to_destroy != NULL) { + next = tctx_tree_next(&gctx->tctxs, + to_destroy); + tctx_tree_remove(&gctx->tctxs, + to_destroy); + idalloctm(tsd_tsdn(tsd), to_destroy, + NULL, NULL, true, true); + } else { + next = NULL; + } + } while (next != NULL); + } + gctx->nlimbo--; + if (prof_gctx_should_destroy(gctx)) { + gctx->nlimbo++; + malloc_mutex_unlock(tsd_tsdn(tsd), gctx->lock); + prof_gctx_try_destroy(tsd, tdata, gctx, tdata); + } else { + malloc_mutex_unlock(tsd_tsdn(tsd), gctx->lock); + } + } +} + +struct prof_tdata_merge_iter_arg_s { + tsdn_t *tsdn; + prof_cnt_t cnt_all; +}; + +static prof_tdata_t * +prof_tdata_merge_iter(prof_tdata_tree_t *tdatas, prof_tdata_t *tdata, + void *opaque) { + struct prof_tdata_merge_iter_arg_s *arg = + (struct prof_tdata_merge_iter_arg_s *)opaque; + + malloc_mutex_lock(arg->tsdn, tdata->lock); + if (!tdata->expired) { + size_t tabind; + union { + prof_tctx_t *p; + void *v; + } tctx; + + tdata->dumping = true; + memset(&tdata->cnt_summed, 0, sizeof(prof_cnt_t)); + for (tabind = 0; !ckh_iter(&tdata->bt2tctx, &tabind, NULL, + &tctx.v);) { + prof_tctx_merge_tdata(arg->tsdn, tctx.p, tdata); + } + + arg->cnt_all.curobjs += tdata->cnt_summed.curobjs; + arg->cnt_all.curbytes += tdata->cnt_summed.curbytes; + if (opt_prof_accum) { + arg->cnt_all.accumobjs += tdata->cnt_summed.accumobjs; + arg->cnt_all.accumbytes += tdata->cnt_summed.accumbytes; + } + } else { + tdata->dumping = false; + } + malloc_mutex_unlock(arg->tsdn, tdata->lock); + + return NULL; +} + +static prof_tdata_t * +prof_tdata_dump_iter(prof_tdata_tree_t *tdatas, prof_tdata_t *tdata, + void *arg) { + bool propagate_err = *(bool *)arg; + + if (!tdata->dumping) { + return NULL; + } + + if (prof_dump_printf(propagate_err, + " t%"FMTu64": %"FMTu64": %"FMTu64" [%"FMTu64": %"FMTu64"]%s%s\n", + tdata->thr_uid, tdata->cnt_summed.curobjs, + tdata->cnt_summed.curbytes, tdata->cnt_summed.accumobjs, + tdata->cnt_summed.accumbytes, + (tdata->thread_name != NULL) ? " " : "", + (tdata->thread_name != NULL) ? tdata->thread_name : "")) { + return tdata; + } + return NULL; +} + +static bool +prof_dump_header_impl(tsdn_t *tsdn, bool propagate_err, + const prof_cnt_t *cnt_all) { + bool ret; + + if (prof_dump_printf(propagate_err, + "heap_v2/%"FMTu64"\n" + " t*: %"FMTu64": %"FMTu64" [%"FMTu64": %"FMTu64"]\n", + ((uint64_t)1U << lg_prof_sample), cnt_all->curobjs, + cnt_all->curbytes, cnt_all->accumobjs, cnt_all->accumbytes)) { + return true; + } + + malloc_mutex_lock(tsdn, &tdatas_mtx); + ret = (tdata_tree_iter(&tdatas, NULL, prof_tdata_dump_iter, + (void *)&propagate_err) != NULL); + malloc_mutex_unlock(tsdn, &tdatas_mtx); + return ret; +} +prof_dump_header_t *JET_MUTABLE prof_dump_header = prof_dump_header_impl; + +static bool +prof_dump_gctx(tsdn_t *tsdn, bool propagate_err, prof_gctx_t *gctx, + const prof_bt_t *bt, prof_gctx_tree_t *gctxs) { + bool ret; + unsigned i; + struct prof_tctx_dump_iter_arg_s prof_tctx_dump_iter_arg; + + cassert(config_prof); + malloc_mutex_assert_owner(tsdn, gctx->lock); + + /* Avoid dumping such gctx's that have no useful data. */ + if ((!opt_prof_accum && gctx->cnt_summed.curobjs == 0) || + (opt_prof_accum && gctx->cnt_summed.accumobjs == 0)) { + assert(gctx->cnt_summed.curobjs == 0); + assert(gctx->cnt_summed.curbytes == 0); + assert(gctx->cnt_summed.accumobjs == 0); + assert(gctx->cnt_summed.accumbytes == 0); + ret = false; + goto label_return; + } + + if (prof_dump_printf(propagate_err, "@")) { + ret = true; + goto label_return; + } + for (i = 0; i < bt->len; i++) { + if (prof_dump_printf(propagate_err, " %#"FMTxPTR, + (uintptr_t)bt->vec[i])) { + ret = true; + goto label_return; + } + } + + if (prof_dump_printf(propagate_err, + "\n" + " t*: %"FMTu64": %"FMTu64" [%"FMTu64": %"FMTu64"]\n", + gctx->cnt_summed.curobjs, gctx->cnt_summed.curbytes, + gctx->cnt_summed.accumobjs, gctx->cnt_summed.accumbytes)) { + ret = true; + goto label_return; + } + + prof_tctx_dump_iter_arg.tsdn = tsdn; + prof_tctx_dump_iter_arg.propagate_err = propagate_err; + if (tctx_tree_iter(&gctx->tctxs, NULL, prof_tctx_dump_iter, + (void *)&prof_tctx_dump_iter_arg) != NULL) { + ret = true; + goto label_return; + } + + ret = false; +label_return: + return ret; +} + +#ifndef _WIN32 +JEMALLOC_FORMAT_PRINTF(1, 2) +static int +prof_open_maps(const char *format, ...) { + int mfd; + va_list ap; + char filename[PATH_MAX + 1]; + + va_start(ap, format); + malloc_vsnprintf(filename, sizeof(filename), format, ap); + va_end(ap); + +#if defined(O_CLOEXEC) + mfd = open(filename, O_RDONLY | O_CLOEXEC); +#else + mfd = open(filename, O_RDONLY); + if (mfd != -1) { + fcntl(mfd, F_SETFD, fcntl(mfd, F_GETFD) | FD_CLOEXEC); + } +#endif + + return mfd; +} +#endif + +static int +prof_getpid(void) { +#ifdef _WIN32 + return GetCurrentProcessId(); +#else + return getpid(); +#endif +} + +static bool +prof_dump_maps(bool propagate_err) { + bool ret; + int mfd; + + cassert(config_prof); +#ifdef __FreeBSD__ + mfd = prof_open_maps("/proc/curproc/map"); +#elif defined(_WIN32) + mfd = -1; // Not implemented +#else + { + int pid = prof_getpid(); + + mfd = prof_open_maps("/proc/%d/task/%d/maps", pid, pid); + if (mfd == -1) { + mfd = prof_open_maps("/proc/%d/maps", pid); + } + } +#endif + if (mfd != -1) { + ssize_t nread; + + if (prof_dump_write(propagate_err, "\nMAPPED_LIBRARIES:\n") && + propagate_err) { + ret = true; + goto label_return; + } + nread = 0; + do { + prof_dump_buf_end += nread; + if (prof_dump_buf_end == PROF_DUMP_BUFSIZE) { + /* Make space in prof_dump_buf before read(). */ + if (prof_dump_flush(propagate_err) && + propagate_err) { + ret = true; + goto label_return; + } + } + nread = malloc_read_fd(mfd, + &prof_dump_buf[prof_dump_buf_end], PROF_DUMP_BUFSIZE + - prof_dump_buf_end); + } while (nread > 0); + } else { + ret = true; + goto label_return; + } + + ret = false; +label_return: + if (mfd != -1) { + close(mfd); + } + return ret; +} + +/* + * See prof_sample_threshold_update() comment for why the body of this function + * is conditionally compiled. + */ +static void +prof_leakcheck(const prof_cnt_t *cnt_all, size_t leak_ngctx, + const char *filename) { +#ifdef JEMALLOC_PROF + /* + * Scaling is equivalent AdjustSamples() in jeprof, but the result may + * differ slightly from what jeprof reports, because here we scale the + * summary values, whereas jeprof scales each context individually and + * reports the sums of the scaled values. + */ + if (cnt_all->curbytes != 0) { + double sample_period = (double)((uint64_t)1 << lg_prof_sample); + double ratio = (((double)cnt_all->curbytes) / + (double)cnt_all->curobjs) / sample_period; + double scale_factor = 1.0 / (1.0 - exp(-ratio)); + uint64_t curbytes = (uint64_t)round(((double)cnt_all->curbytes) + * scale_factor); + uint64_t curobjs = (uint64_t)round(((double)cnt_all->curobjs) * + scale_factor); + + malloc_printf("<jemalloc>: Leak approximation summary: ~%"FMTu64 + " byte%s, ~%"FMTu64" object%s, >= %zu context%s\n", + curbytes, (curbytes != 1) ? "s" : "", curobjs, (curobjs != + 1) ? "s" : "", leak_ngctx, (leak_ngctx != 1) ? "s" : ""); + malloc_printf( + "<jemalloc>: Run jeprof on \"%s\" for leak detail\n", + filename); + } +#endif +} + +struct prof_gctx_dump_iter_arg_s { + tsdn_t *tsdn; + bool propagate_err; +}; + +static prof_gctx_t * +prof_gctx_dump_iter(prof_gctx_tree_t *gctxs, prof_gctx_t *gctx, void *opaque) { + prof_gctx_t *ret; + struct prof_gctx_dump_iter_arg_s *arg = + (struct prof_gctx_dump_iter_arg_s *)opaque; + + malloc_mutex_lock(arg->tsdn, gctx->lock); + + if (prof_dump_gctx(arg->tsdn, arg->propagate_err, gctx, &gctx->bt, + gctxs)) { + ret = gctx; + goto label_return; + } + + ret = NULL; +label_return: + malloc_mutex_unlock(arg->tsdn, gctx->lock); + return ret; +} + +static void +prof_dump_prep(tsd_t *tsd, prof_tdata_t *tdata, + struct prof_tdata_merge_iter_arg_s *prof_tdata_merge_iter_arg, + struct prof_gctx_merge_iter_arg_s *prof_gctx_merge_iter_arg, + prof_gctx_tree_t *gctxs) { + size_t tabind; + union { + prof_gctx_t *p; + void *v; + } gctx; + + prof_enter(tsd, tdata); + + /* + * Put gctx's in limbo and clear their counters in preparation for + * summing. + */ + gctx_tree_new(gctxs); + for (tabind = 0; !ckh_iter(&bt2gctx, &tabind, NULL, &gctx.v);) { + prof_dump_gctx_prep(tsd_tsdn(tsd), gctx.p, gctxs); + } + + /* + * Iterate over tdatas, and for the non-expired ones snapshot their tctx + * stats and merge them into the associated gctx's. + */ + prof_tdata_merge_iter_arg->tsdn = tsd_tsdn(tsd); + memset(&prof_tdata_merge_iter_arg->cnt_all, 0, sizeof(prof_cnt_t)); + malloc_mutex_lock(tsd_tsdn(tsd), &tdatas_mtx); + tdata_tree_iter(&tdatas, NULL, prof_tdata_merge_iter, + (void *)prof_tdata_merge_iter_arg); + malloc_mutex_unlock(tsd_tsdn(tsd), &tdatas_mtx); + + /* Merge tctx stats into gctx's. */ + prof_gctx_merge_iter_arg->tsdn = tsd_tsdn(tsd); + prof_gctx_merge_iter_arg->leak_ngctx = 0; + gctx_tree_iter(gctxs, NULL, prof_gctx_merge_iter, + (void *)prof_gctx_merge_iter_arg); + + prof_leave(tsd, tdata); +} + +static bool +prof_dump_file(tsd_t *tsd, bool propagate_err, const char *filename, + bool leakcheck, prof_tdata_t *tdata, + struct prof_tdata_merge_iter_arg_s *prof_tdata_merge_iter_arg, + struct prof_gctx_merge_iter_arg_s *prof_gctx_merge_iter_arg, + struct prof_gctx_dump_iter_arg_s *prof_gctx_dump_iter_arg, + prof_gctx_tree_t *gctxs) { + /* Create dump file. */ + if ((prof_dump_fd = prof_dump_open(propagate_err, filename)) == -1) { + return true; + } + + /* Dump profile header. */ + if (prof_dump_header(tsd_tsdn(tsd), propagate_err, + &prof_tdata_merge_iter_arg->cnt_all)) { + goto label_write_error; + } + + /* Dump per gctx profile stats. */ + prof_gctx_dump_iter_arg->tsdn = tsd_tsdn(tsd); + prof_gctx_dump_iter_arg->propagate_err = propagate_err; + if (gctx_tree_iter(gctxs, NULL, prof_gctx_dump_iter, + (void *)prof_gctx_dump_iter_arg) != NULL) { + goto label_write_error; + } + + /* Dump /proc/<pid>/maps if possible. */ + if (prof_dump_maps(propagate_err)) { + goto label_write_error; + } + + if (prof_dump_close(propagate_err)) { + return true; + } + + return false; +label_write_error: + prof_dump_close(propagate_err); + return true; +} + +static bool +prof_dump(tsd_t *tsd, bool propagate_err, const char *filename, + bool leakcheck) { + cassert(config_prof); + assert(tsd_reentrancy_level_get(tsd) == 0); + + prof_tdata_t * tdata = prof_tdata_get(tsd, true); + if (tdata == NULL) { + return true; + } + + pre_reentrancy(tsd, NULL); + malloc_mutex_lock(tsd_tsdn(tsd), &prof_dump_mtx); + + prof_gctx_tree_t gctxs; + struct prof_tdata_merge_iter_arg_s prof_tdata_merge_iter_arg; + struct prof_gctx_merge_iter_arg_s prof_gctx_merge_iter_arg; + struct prof_gctx_dump_iter_arg_s prof_gctx_dump_iter_arg; + prof_dump_prep(tsd, tdata, &prof_tdata_merge_iter_arg, + &prof_gctx_merge_iter_arg, &gctxs); + bool err = prof_dump_file(tsd, propagate_err, filename, leakcheck, tdata, + &prof_tdata_merge_iter_arg, &prof_gctx_merge_iter_arg, + &prof_gctx_dump_iter_arg, &gctxs); + prof_gctx_finish(tsd, &gctxs); + + malloc_mutex_unlock(tsd_tsdn(tsd), &prof_dump_mtx); + post_reentrancy(tsd); + + if (err) { + return true; + } + + if (leakcheck) { + prof_leakcheck(&prof_tdata_merge_iter_arg.cnt_all, + prof_gctx_merge_iter_arg.leak_ngctx, filename); + } + return false; +} + +#ifdef JEMALLOC_JET +void +prof_cnt_all(uint64_t *curobjs, uint64_t *curbytes, uint64_t *accumobjs, + uint64_t *accumbytes) { + tsd_t *tsd; + prof_tdata_t *tdata; + struct prof_tdata_merge_iter_arg_s prof_tdata_merge_iter_arg; + struct prof_gctx_merge_iter_arg_s prof_gctx_merge_iter_arg; + prof_gctx_tree_t gctxs; + + tsd = tsd_fetch(); + tdata = prof_tdata_get(tsd, false); + if (tdata == NULL) { + if (curobjs != NULL) { + *curobjs = 0; + } + if (curbytes != NULL) { + *curbytes = 0; + } + if (accumobjs != NULL) { + *accumobjs = 0; + } + if (accumbytes != NULL) { + *accumbytes = 0; + } + return; + } + + prof_dump_prep(tsd, tdata, &prof_tdata_merge_iter_arg, + &prof_gctx_merge_iter_arg, &gctxs); + prof_gctx_finish(tsd, &gctxs); + + if (curobjs != NULL) { + *curobjs = prof_tdata_merge_iter_arg.cnt_all.curobjs; + } + if (curbytes != NULL) { + *curbytes = prof_tdata_merge_iter_arg.cnt_all.curbytes; + } + if (accumobjs != NULL) { + *accumobjs = prof_tdata_merge_iter_arg.cnt_all.accumobjs; + } + if (accumbytes != NULL) { + *accumbytes = prof_tdata_merge_iter_arg.cnt_all.accumbytes; + } +} +#endif + +#define DUMP_FILENAME_BUFSIZE (PATH_MAX + 1) +#define VSEQ_INVALID UINT64_C(0xffffffffffffffff) +static void +prof_dump_filename(char *filename, char v, uint64_t vseq) { + cassert(config_prof); + + if (vseq != VSEQ_INVALID) { + /* "<prefix>.<pid>.<seq>.v<vseq>.heap" */ + malloc_snprintf(filename, DUMP_FILENAME_BUFSIZE, + "%s.%d.%"FMTu64".%c%"FMTu64".heap", + opt_prof_prefix, prof_getpid(), prof_dump_seq, v, vseq); + } else { + /* "<prefix>.<pid>.<seq>.<v>.heap" */ + malloc_snprintf(filename, DUMP_FILENAME_BUFSIZE, + "%s.%d.%"FMTu64".%c.heap", + opt_prof_prefix, prof_getpid(), prof_dump_seq, v); + } + prof_dump_seq++; +} + +static void +prof_fdump(void) { + tsd_t *tsd; + char filename[DUMP_FILENAME_BUFSIZE]; + + cassert(config_prof); + assert(opt_prof_final); + assert(opt_prof_prefix[0] != '\0'); + + if (!prof_booted) { + return; + } + tsd = tsd_fetch(); + assert(tsd_reentrancy_level_get(tsd) == 0); + + malloc_mutex_lock(tsd_tsdn(tsd), &prof_dump_seq_mtx); + prof_dump_filename(filename, 'f', VSEQ_INVALID); + malloc_mutex_unlock(tsd_tsdn(tsd), &prof_dump_seq_mtx); + prof_dump(tsd, false, filename, opt_prof_leak); +} + +bool +prof_accum_init(tsdn_t *tsdn, prof_accum_t *prof_accum) { + cassert(config_prof); + +#ifndef JEMALLOC_ATOMIC_U64 + if (malloc_mutex_init(&prof_accum->mtx, "prof_accum", + WITNESS_RANK_PROF_ACCUM, malloc_mutex_rank_exclusive)) { + return true; + } + prof_accum->accumbytes = 0; +#else + atomic_store_u64(&prof_accum->accumbytes, 0, ATOMIC_RELAXED); +#endif + return false; +} + +void +prof_idump(tsdn_t *tsdn) { + tsd_t *tsd; + prof_tdata_t *tdata; + + cassert(config_prof); + + if (!prof_booted || tsdn_null(tsdn) || !prof_active_get_unlocked()) { + return; + } + tsd = tsdn_tsd(tsdn); + if (tsd_reentrancy_level_get(tsd) > 0) { + return; + } + + tdata = prof_tdata_get(tsd, false); + if (tdata == NULL) { + return; + } + if (tdata->enq) { + tdata->enq_idump = true; + return; + } + + if (opt_prof_prefix[0] != '\0') { + char filename[PATH_MAX + 1]; + malloc_mutex_lock(tsd_tsdn(tsd), &prof_dump_seq_mtx); + prof_dump_filename(filename, 'i', prof_dump_iseq); + prof_dump_iseq++; + malloc_mutex_unlock(tsd_tsdn(tsd), &prof_dump_seq_mtx); + prof_dump(tsd, false, filename, false); + } +} + +bool +prof_mdump(tsd_t *tsd, const char *filename) { + cassert(config_prof); + assert(tsd_reentrancy_level_get(tsd) == 0); + + if (!opt_prof || !prof_booted) { + return true; + } + char filename_buf[DUMP_FILENAME_BUFSIZE]; + if (filename == NULL) { + /* No filename specified, so automatically generate one. */ + if (opt_prof_prefix[0] == '\0') { + return true; + } + malloc_mutex_lock(tsd_tsdn(tsd), &prof_dump_seq_mtx); + prof_dump_filename(filename_buf, 'm', prof_dump_mseq); + prof_dump_mseq++; + malloc_mutex_unlock(tsd_tsdn(tsd), &prof_dump_seq_mtx); + filename = filename_buf; + } + return prof_dump(tsd, true, filename, false); +} + +void +prof_gdump(tsdn_t *tsdn) { + tsd_t *tsd; + prof_tdata_t *tdata; + + cassert(config_prof); + + if (!prof_booted || tsdn_null(tsdn) || !prof_active_get_unlocked()) { + return; + } + tsd = tsdn_tsd(tsdn); + if (tsd_reentrancy_level_get(tsd) > 0) { + return; + } + + tdata = prof_tdata_get(tsd, false); + if (tdata == NULL) { + return; + } + if (tdata->enq) { + tdata->enq_gdump = true; + return; + } + + if (opt_prof_prefix[0] != '\0') { + char filename[DUMP_FILENAME_BUFSIZE]; + malloc_mutex_lock(tsdn, &prof_dump_seq_mtx); + prof_dump_filename(filename, 'u', prof_dump_useq); + prof_dump_useq++; + malloc_mutex_unlock(tsdn, &prof_dump_seq_mtx); + prof_dump(tsd, false, filename, false); + } +} + +static void +prof_bt_hash(const void *key, size_t r_hash[2]) { + prof_bt_t *bt = (prof_bt_t *)key; + + cassert(config_prof); + + hash(bt->vec, bt->len * sizeof(void *), 0x94122f33U, r_hash); +} + +static bool +prof_bt_keycomp(const void *k1, const void *k2) { + const prof_bt_t *bt1 = (prof_bt_t *)k1; + const prof_bt_t *bt2 = (prof_bt_t *)k2; + + cassert(config_prof); + + if (bt1->len != bt2->len) { + return false; + } + return (memcmp(bt1->vec, bt2->vec, bt1->len * sizeof(void *)) == 0); +} + +static void +prof_bt_node_hash(const void *key, size_t r_hash[2]) { + const prof_bt_node_t *bt_node = (prof_bt_node_t *)key; + prof_bt_hash((void *)(&bt_node->bt), r_hash); +} + +static bool +prof_bt_node_keycomp(const void *k1, const void *k2) { + const prof_bt_node_t *bt_node1 = (prof_bt_node_t *)k1; + const prof_bt_node_t *bt_node2 = (prof_bt_node_t *)k2; + return prof_bt_keycomp((void *)(&bt_node1->bt), + (void *)(&bt_node2->bt)); +} + +static void +prof_thr_node_hash(const void *key, size_t r_hash[2]) { + const prof_thr_node_t *thr_node = (prof_thr_node_t *)key; + hash(&thr_node->thr_uid, sizeof(uint64_t), 0x94122f35U, r_hash); +} + +static bool +prof_thr_node_keycomp(const void *k1, const void *k2) { + const prof_thr_node_t *thr_node1 = (prof_thr_node_t *)k1; + const prof_thr_node_t *thr_node2 = (prof_thr_node_t *)k2; + return thr_node1->thr_uid == thr_node2->thr_uid; +} + +static uint64_t +prof_thr_uid_alloc(tsdn_t *tsdn) { + uint64_t thr_uid; + + malloc_mutex_lock(tsdn, &next_thr_uid_mtx); + thr_uid = next_thr_uid; + next_thr_uid++; + malloc_mutex_unlock(tsdn, &next_thr_uid_mtx); + + return thr_uid; +} + +static prof_tdata_t * +prof_tdata_init_impl(tsd_t *tsd, uint64_t thr_uid, uint64_t thr_discrim, + char *thread_name, bool active) { + prof_tdata_t *tdata; + + cassert(config_prof); + + /* Initialize an empty cache for this thread. */ + tdata = (prof_tdata_t *)iallocztm(tsd_tsdn(tsd), sizeof(prof_tdata_t), + sz_size2index(sizeof(prof_tdata_t)), false, NULL, true, + arena_get(TSDN_NULL, 0, true), true); + if (tdata == NULL) { + return NULL; + } + + tdata->lock = prof_tdata_mutex_choose(thr_uid); + tdata->thr_uid = thr_uid; + tdata->thr_discrim = thr_discrim; + tdata->thread_name = thread_name; + tdata->attached = true; + tdata->expired = false; + tdata->tctx_uid_next = 0; + + if (ckh_new(tsd, &tdata->bt2tctx, PROF_CKH_MINITEMS, prof_bt_hash, + prof_bt_keycomp)) { + idalloctm(tsd_tsdn(tsd), tdata, NULL, NULL, true, true); + return NULL; + } + + tdata->prng_state = (uint64_t)(uintptr_t)tdata; + prof_sample_threshold_update(tdata); + + tdata->enq = false; + tdata->enq_idump = false; + tdata->enq_gdump = false; + + tdata->dumping = false; + tdata->active = active; + + malloc_mutex_lock(tsd_tsdn(tsd), &tdatas_mtx); + tdata_tree_insert(&tdatas, tdata); + malloc_mutex_unlock(tsd_tsdn(tsd), &tdatas_mtx); + + return tdata; +} + +prof_tdata_t * +prof_tdata_init(tsd_t *tsd) { + return prof_tdata_init_impl(tsd, prof_thr_uid_alloc(tsd_tsdn(tsd)), 0, + NULL, prof_thread_active_init_get(tsd_tsdn(tsd))); +} + +static bool +prof_tdata_should_destroy_unlocked(prof_tdata_t *tdata, bool even_if_attached) { + if (tdata->attached && !even_if_attached) { + return false; + } + if (ckh_count(&tdata->bt2tctx) != 0) { + return false; + } + return true; +} + +static bool +prof_tdata_should_destroy(tsdn_t *tsdn, prof_tdata_t *tdata, + bool even_if_attached) { + malloc_mutex_assert_owner(tsdn, tdata->lock); + + return prof_tdata_should_destroy_unlocked(tdata, even_if_attached); +} + +static void +prof_tdata_destroy_locked(tsd_t *tsd, prof_tdata_t *tdata, + bool even_if_attached) { + malloc_mutex_assert_owner(tsd_tsdn(tsd), &tdatas_mtx); + + tdata_tree_remove(&tdatas, tdata); + + assert(prof_tdata_should_destroy_unlocked(tdata, even_if_attached)); + + if (tdata->thread_name != NULL) { + idalloctm(tsd_tsdn(tsd), tdata->thread_name, NULL, NULL, true, + true); + } + ckh_delete(tsd, &tdata->bt2tctx); + idalloctm(tsd_tsdn(tsd), tdata, NULL, NULL, true, true); +} + +static void +prof_tdata_destroy(tsd_t *tsd, prof_tdata_t *tdata, bool even_if_attached) { + malloc_mutex_lock(tsd_tsdn(tsd), &tdatas_mtx); + prof_tdata_destroy_locked(tsd, tdata, even_if_attached); + malloc_mutex_unlock(tsd_tsdn(tsd), &tdatas_mtx); +} + +static void +prof_tdata_detach(tsd_t *tsd, prof_tdata_t *tdata) { + bool destroy_tdata; + + malloc_mutex_lock(tsd_tsdn(tsd), tdata->lock); + if (tdata->attached) { + destroy_tdata = prof_tdata_should_destroy(tsd_tsdn(tsd), tdata, + true); + /* + * Only detach if !destroy_tdata, because detaching would allow + * another thread to win the race to destroy tdata. + */ + if (!destroy_tdata) { + tdata->attached = false; + } + tsd_prof_tdata_set(tsd, NULL); + } else { + destroy_tdata = false; + } + malloc_mutex_unlock(tsd_tsdn(tsd), tdata->lock); + if (destroy_tdata) { + prof_tdata_destroy(tsd, tdata, true); + } +} + +prof_tdata_t * +prof_tdata_reinit(tsd_t *tsd, prof_tdata_t *tdata) { + uint64_t thr_uid = tdata->thr_uid; + uint64_t thr_discrim = tdata->thr_discrim + 1; + char *thread_name = (tdata->thread_name != NULL) ? + prof_thread_name_alloc(tsd_tsdn(tsd), tdata->thread_name) : NULL; + bool active = tdata->active; + + prof_tdata_detach(tsd, tdata); + return prof_tdata_init_impl(tsd, thr_uid, thr_discrim, thread_name, + active); +} + +static bool +prof_tdata_expire(tsdn_t *tsdn, prof_tdata_t *tdata) { + bool destroy_tdata; + + malloc_mutex_lock(tsdn, tdata->lock); + if (!tdata->expired) { + tdata->expired = true; + destroy_tdata = tdata->attached ? false : + prof_tdata_should_destroy(tsdn, tdata, false); + } else { + destroy_tdata = false; + } + malloc_mutex_unlock(tsdn, tdata->lock); + + return destroy_tdata; +} + +static prof_tdata_t * +prof_tdata_reset_iter(prof_tdata_tree_t *tdatas, prof_tdata_t *tdata, + void *arg) { + tsdn_t *tsdn = (tsdn_t *)arg; + + return (prof_tdata_expire(tsdn, tdata) ? tdata : NULL); +} + +void +prof_reset(tsd_t *tsd, size_t lg_sample) { + prof_tdata_t *next; + + assert(lg_sample < (sizeof(uint64_t) << 3)); + + malloc_mutex_lock(tsd_tsdn(tsd), &prof_dump_mtx); + malloc_mutex_lock(tsd_tsdn(tsd), &tdatas_mtx); + + lg_prof_sample = lg_sample; + + next = NULL; + do { + prof_tdata_t *to_destroy = tdata_tree_iter(&tdatas, next, + prof_tdata_reset_iter, (void *)tsd); + if (to_destroy != NULL) { + next = tdata_tree_next(&tdatas, to_destroy); + prof_tdata_destroy_locked(tsd, to_destroy, false); + } else { + next = NULL; + } + } while (next != NULL); + + malloc_mutex_unlock(tsd_tsdn(tsd), &tdatas_mtx); + malloc_mutex_unlock(tsd_tsdn(tsd), &prof_dump_mtx); +} + +void +prof_tdata_cleanup(tsd_t *tsd) { + prof_tdata_t *tdata; + + if (!config_prof) { + return; + } + + tdata = tsd_prof_tdata_get(tsd); + if (tdata != NULL) { + prof_tdata_detach(tsd, tdata); + } +} + +bool +prof_active_get(tsdn_t *tsdn) { + bool prof_active_current; + + malloc_mutex_lock(tsdn, &prof_active_mtx); + prof_active_current = prof_active; + malloc_mutex_unlock(tsdn, &prof_active_mtx); + return prof_active_current; +} + +bool +prof_active_set(tsdn_t *tsdn, bool active) { + bool prof_active_old; + + malloc_mutex_lock(tsdn, &prof_active_mtx); + prof_active_old = prof_active; + prof_active = active; + malloc_mutex_unlock(tsdn, &prof_active_mtx); + return prof_active_old; +} + +#ifdef JEMALLOC_JET +size_t +prof_log_bt_count(void) { + size_t cnt = 0; + prof_bt_node_t *node = log_bt_first; + while (node != NULL) { + cnt++; + node = node->next; + } + return cnt; +} + +size_t +prof_log_alloc_count(void) { + size_t cnt = 0; + prof_alloc_node_t *node = log_alloc_first; + while (node != NULL) { + cnt++; + node = node->next; + } + return cnt; +} + +size_t +prof_log_thr_count(void) { + size_t cnt = 0; + prof_thr_node_t *node = log_thr_first; + while (node != NULL) { + cnt++; + node = node->next; + } + return cnt; +} + +bool +prof_log_is_logging(void) { + return prof_logging_state == prof_logging_state_started; +} + +bool +prof_log_rep_check(void) { + if (prof_logging_state == prof_logging_state_stopped + && log_tables_initialized) { + return true; + } + + if (log_bt_last != NULL && log_bt_last->next != NULL) { + return true; + } + if (log_thr_last != NULL && log_thr_last->next != NULL) { + return true; + } + if (log_alloc_last != NULL && log_alloc_last->next != NULL) { + return true; + } + + size_t bt_count = prof_log_bt_count(); + size_t thr_count = prof_log_thr_count(); + size_t alloc_count = prof_log_alloc_count(); + + + if (prof_logging_state == prof_logging_state_stopped) { + if (bt_count != 0 || thr_count != 0 || alloc_count || 0) { + return true; + } + } + + prof_alloc_node_t *node = log_alloc_first; + while (node != NULL) { + if (node->alloc_bt_ind >= bt_count) { + return true; + } + if (node->free_bt_ind >= bt_count) { + return true; + } + if (node->alloc_thr_ind >= thr_count) { + return true; + } + if (node->free_thr_ind >= thr_count) { + return true; + } + if (node->alloc_time_ns > node->free_time_ns) { + return true; + } + node = node->next; + } + + return false; +} + +void +prof_log_dummy_set(bool new_value) { + prof_log_dummy = new_value; +} +#endif + +bool +prof_log_start(tsdn_t *tsdn, const char *filename) { + if (!opt_prof || !prof_booted) { + return true; + } + + bool ret = false; + size_t buf_size = PATH_MAX + 1; + + malloc_mutex_lock(tsdn, &log_mtx); + + if (prof_logging_state != prof_logging_state_stopped) { + ret = true; + } else if (filename == NULL) { + /* Make default name. */ + malloc_snprintf(log_filename, buf_size, "%s.%d.%"FMTu64".json", + opt_prof_prefix, prof_getpid(), log_seq); + log_seq++; + prof_logging_state = prof_logging_state_started; + } else if (strlen(filename) >= buf_size) { + ret = true; + } else { + strcpy(log_filename, filename); + prof_logging_state = prof_logging_state_started; + } + + if (!ret) { + nstime_update(&log_start_timestamp); + } + + malloc_mutex_unlock(tsdn, &log_mtx); + + return ret; +} + +/* Used as an atexit function to stop logging on exit. */ +static void +prof_log_stop_final(void) { + tsd_t *tsd = tsd_fetch(); + prof_log_stop(tsd_tsdn(tsd)); +} + +struct prof_emitter_cb_arg_s { + int fd; + ssize_t ret; +}; + +static void +prof_emitter_write_cb(void *opaque, const char *to_write) { + struct prof_emitter_cb_arg_s *arg = + (struct prof_emitter_cb_arg_s *)opaque; + size_t bytes = strlen(to_write); +#ifdef JEMALLOC_JET + if (prof_log_dummy) { + return; + } +#endif + arg->ret = write(arg->fd, (void *)to_write, bytes); +} + +/* + * prof_log_emit_{...} goes through the appropriate linked list, emitting each + * node to the json and deallocating it. + */ +static void +prof_log_emit_threads(tsd_t *tsd, emitter_t *emitter) { + emitter_json_array_kv_begin(emitter, "threads"); + prof_thr_node_t *thr_node = log_thr_first; + prof_thr_node_t *thr_old_node; + while (thr_node != NULL) { + emitter_json_object_begin(emitter); + + emitter_json_kv(emitter, "thr_uid", emitter_type_uint64, + &thr_node->thr_uid); + + char *thr_name = thr_node->name; + + emitter_json_kv(emitter, "thr_name", emitter_type_string, + &thr_name); + + emitter_json_object_end(emitter); + thr_old_node = thr_node; + thr_node = thr_node->next; + idalloc(tsd, thr_old_node); + } + emitter_json_array_end(emitter); +} + +static void +prof_log_emit_traces(tsd_t *tsd, emitter_t *emitter) { + emitter_json_array_kv_begin(emitter, "stack_traces"); + prof_bt_node_t *bt_node = log_bt_first; + prof_bt_node_t *bt_old_node; + /* + * Calculate how many hex digits we need: twice number of bytes, two for + * "0x", and then one more for terminating '\0'. + */ + char buf[2 * sizeof(intptr_t) + 3]; + size_t buf_sz = sizeof(buf); + while (bt_node != NULL) { + emitter_json_array_begin(emitter); + size_t i; + for (i = 0; i < bt_node->bt.len; i++) { + malloc_snprintf(buf, buf_sz, "%p", bt_node->bt.vec[i]); + char *trace_str = buf; + emitter_json_value(emitter, emitter_type_string, + &trace_str); + } + emitter_json_array_end(emitter); + + bt_old_node = bt_node; + bt_node = bt_node->next; + idalloc(tsd, bt_old_node); + } + emitter_json_array_end(emitter); +} + +static void +prof_log_emit_allocs(tsd_t *tsd, emitter_t *emitter) { + emitter_json_array_kv_begin(emitter, "allocations"); + prof_alloc_node_t *alloc_node = log_alloc_first; + prof_alloc_node_t *alloc_old_node; + while (alloc_node != NULL) { + emitter_json_object_begin(emitter); + + emitter_json_kv(emitter, "alloc_thread", emitter_type_size, + &alloc_node->alloc_thr_ind); + + emitter_json_kv(emitter, "free_thread", emitter_type_size, + &alloc_node->free_thr_ind); + + emitter_json_kv(emitter, "alloc_trace", emitter_type_size, + &alloc_node->alloc_bt_ind); + + emitter_json_kv(emitter, "free_trace", emitter_type_size, + &alloc_node->free_bt_ind); + + emitter_json_kv(emitter, "alloc_timestamp", + emitter_type_uint64, &alloc_node->alloc_time_ns); + + emitter_json_kv(emitter, "free_timestamp", emitter_type_uint64, + &alloc_node->free_time_ns); + + emitter_json_kv(emitter, "usize", emitter_type_uint64, + &alloc_node->usize); + + emitter_json_object_end(emitter); + + alloc_old_node = alloc_node; + alloc_node = alloc_node->next; + idalloc(tsd, alloc_old_node); + } + emitter_json_array_end(emitter); +} + +static void +prof_log_emit_metadata(emitter_t *emitter) { + emitter_json_object_kv_begin(emitter, "info"); + + nstime_t now = NSTIME_ZERO_INITIALIZER; + + nstime_update(&now); + uint64_t ns = nstime_ns(&now) - nstime_ns(&log_start_timestamp); + emitter_json_kv(emitter, "duration", emitter_type_uint64, &ns); + + char *vers = JEMALLOC_VERSION; + emitter_json_kv(emitter, "version", + emitter_type_string, &vers); + + emitter_json_kv(emitter, "lg_sample_rate", + emitter_type_int, &lg_prof_sample); + + int pid = prof_getpid(); + emitter_json_kv(emitter, "pid", emitter_type_int, &pid); + + emitter_json_object_end(emitter); +} + + +bool +prof_log_stop(tsdn_t *tsdn) { + if (!opt_prof || !prof_booted) { + return true; + } + + tsd_t *tsd = tsdn_tsd(tsdn); + malloc_mutex_lock(tsdn, &log_mtx); + + if (prof_logging_state != prof_logging_state_started) { + malloc_mutex_unlock(tsdn, &log_mtx); + return true; + } + + /* + * Set the state to dumping. We'll set it to stopped when we're done. + * Since other threads won't be able to start/stop/log when the state is + * dumping, we don't have to hold the lock during the whole method. + */ + prof_logging_state = prof_logging_state_dumping; + malloc_mutex_unlock(tsdn, &log_mtx); + + + emitter_t emitter; + + /* Create a file. */ + + int fd; +#ifdef JEMALLOC_JET + if (prof_log_dummy) { + fd = 0; + } else { + fd = creat(log_filename, 0644); + } +#else + fd = creat(log_filename, 0644); +#endif + + if (fd == -1) { + malloc_printf("<jemalloc>: creat() for log file \"%s\" " + " failed with %d\n", log_filename, errno); + if (opt_abort) { + abort(); + } + return true; + } + + /* Emit to json. */ + struct prof_emitter_cb_arg_s arg; + arg.fd = fd; + emitter_init(&emitter, emitter_output_json, &prof_emitter_write_cb, + (void *)(&arg)); + + emitter_begin(&emitter); + prof_log_emit_metadata(&emitter); + prof_log_emit_threads(tsd, &emitter); + prof_log_emit_traces(tsd, &emitter); + prof_log_emit_allocs(tsd, &emitter); + emitter_end(&emitter); + + /* Reset global state. */ + if (log_tables_initialized) { + ckh_delete(tsd, &log_bt_node_set); + ckh_delete(tsd, &log_thr_node_set); + } + log_tables_initialized = false; + log_bt_index = 0; + log_thr_index = 0; + log_bt_first = NULL; + log_bt_last = NULL; + log_thr_first = NULL; + log_thr_last = NULL; + log_alloc_first = NULL; + log_alloc_last = NULL; + + malloc_mutex_lock(tsdn, &log_mtx); + prof_logging_state = prof_logging_state_stopped; + malloc_mutex_unlock(tsdn, &log_mtx); + +#ifdef JEMALLOC_JET + if (prof_log_dummy) { + return false; + } +#endif + return close(fd); +} + +const char * +prof_thread_name_get(tsd_t *tsd) { + prof_tdata_t *tdata; + + tdata = prof_tdata_get(tsd, true); + if (tdata == NULL) { + return ""; + } + return (tdata->thread_name != NULL ? tdata->thread_name : ""); +} + +static char * +prof_thread_name_alloc(tsdn_t *tsdn, const char *thread_name) { + char *ret; + size_t size; + + if (thread_name == NULL) { + return NULL; + } + + size = strlen(thread_name) + 1; + if (size == 1) { + return ""; + } + + ret = iallocztm(tsdn, size, sz_size2index(size), false, NULL, true, + arena_get(TSDN_NULL, 0, true), true); + if (ret == NULL) { + return NULL; + } + memcpy(ret, thread_name, size); + return ret; +} + +int +prof_thread_name_set(tsd_t *tsd, const char *thread_name) { + prof_tdata_t *tdata; + unsigned i; + char *s; + + tdata = prof_tdata_get(tsd, true); + if (tdata == NULL) { + return EAGAIN; + } + + /* Validate input. */ + if (thread_name == NULL) { + return EFAULT; + } + for (i = 0; thread_name[i] != '\0'; i++) { + char c = thread_name[i]; + if (!isgraph(c) && !isblank(c)) { + return EFAULT; + } + } + + s = prof_thread_name_alloc(tsd_tsdn(tsd), thread_name); + if (s == NULL) { + return EAGAIN; + } + + if (tdata->thread_name != NULL) { + idalloctm(tsd_tsdn(tsd), tdata->thread_name, NULL, NULL, true, + true); + tdata->thread_name = NULL; + } + if (strlen(s) > 0) { + tdata->thread_name = s; + } + return 0; +} + +bool +prof_thread_active_get(tsd_t *tsd) { + prof_tdata_t *tdata; + + tdata = prof_tdata_get(tsd, true); + if (tdata == NULL) { + return false; + } + return tdata->active; +} + +bool +prof_thread_active_set(tsd_t *tsd, bool active) { + prof_tdata_t *tdata; + + tdata = prof_tdata_get(tsd, true); + if (tdata == NULL) { + return true; + } + tdata->active = active; + return false; +} + +bool +prof_thread_active_init_get(tsdn_t *tsdn) { + bool active_init; + + malloc_mutex_lock(tsdn, &prof_thread_active_init_mtx); + active_init = prof_thread_active_init; + malloc_mutex_unlock(tsdn, &prof_thread_active_init_mtx); + return active_init; +} + +bool +prof_thread_active_init_set(tsdn_t *tsdn, bool active_init) { + bool active_init_old; + + malloc_mutex_lock(tsdn, &prof_thread_active_init_mtx); + active_init_old = prof_thread_active_init; + prof_thread_active_init = active_init; + malloc_mutex_unlock(tsdn, &prof_thread_active_init_mtx); + return active_init_old; +} + +bool +prof_gdump_get(tsdn_t *tsdn) { + bool prof_gdump_current; + + malloc_mutex_lock(tsdn, &prof_gdump_mtx); + prof_gdump_current = prof_gdump_val; + malloc_mutex_unlock(tsdn, &prof_gdump_mtx); + return prof_gdump_current; +} + +bool +prof_gdump_set(tsdn_t *tsdn, bool gdump) { + bool prof_gdump_old; + + malloc_mutex_lock(tsdn, &prof_gdump_mtx); + prof_gdump_old = prof_gdump_val; + prof_gdump_val = gdump; + malloc_mutex_unlock(tsdn, &prof_gdump_mtx); + return prof_gdump_old; +} + +void +prof_boot0(void) { + cassert(config_prof); + + memcpy(opt_prof_prefix, PROF_PREFIX_DEFAULT, + sizeof(PROF_PREFIX_DEFAULT)); +} + +void +prof_boot1(void) { + cassert(config_prof); + + /* + * opt_prof must be in its final state before any arenas are + * initialized, so this function must be executed early. + */ + + if (opt_prof_leak && !opt_prof) { + /* + * Enable opt_prof, but in such a way that profiles are never + * automatically dumped. + */ + opt_prof = true; + opt_prof_gdump = false; + } else if (opt_prof) { + if (opt_lg_prof_interval >= 0) { + prof_interval = (((uint64_t)1U) << + opt_lg_prof_interval); + } + } +} + +bool +prof_boot2(tsd_t *tsd) { + cassert(config_prof); + + if (opt_prof) { + unsigned i; + + lg_prof_sample = opt_lg_prof_sample; + + prof_active = opt_prof_active; + if (malloc_mutex_init(&prof_active_mtx, "prof_active", + WITNESS_RANK_PROF_ACTIVE, malloc_mutex_rank_exclusive)) { + return true; + } + + prof_gdump_val = opt_prof_gdump; + if (malloc_mutex_init(&prof_gdump_mtx, "prof_gdump", + WITNESS_RANK_PROF_GDUMP, malloc_mutex_rank_exclusive)) { + return true; + } + + prof_thread_active_init = opt_prof_thread_active_init; + if (malloc_mutex_init(&prof_thread_active_init_mtx, + "prof_thread_active_init", + WITNESS_RANK_PROF_THREAD_ACTIVE_INIT, + malloc_mutex_rank_exclusive)) { + return true; + } + + if (ckh_new(tsd, &bt2gctx, PROF_CKH_MINITEMS, prof_bt_hash, + prof_bt_keycomp)) { + return true; + } + if (malloc_mutex_init(&bt2gctx_mtx, "prof_bt2gctx", + WITNESS_RANK_PROF_BT2GCTX, malloc_mutex_rank_exclusive)) { + return true; + } + + tdata_tree_new(&tdatas); + if (malloc_mutex_init(&tdatas_mtx, "prof_tdatas", + WITNESS_RANK_PROF_TDATAS, malloc_mutex_rank_exclusive)) { + return true; + } + + next_thr_uid = 0; + if (malloc_mutex_init(&next_thr_uid_mtx, "prof_next_thr_uid", + WITNESS_RANK_PROF_NEXT_THR_UID, malloc_mutex_rank_exclusive)) { + return true; + } + + if (malloc_mutex_init(&prof_dump_seq_mtx, "prof_dump_seq", + WITNESS_RANK_PROF_DUMP_SEQ, malloc_mutex_rank_exclusive)) { + return true; + } + if (malloc_mutex_init(&prof_dump_mtx, "prof_dump", + WITNESS_RANK_PROF_DUMP, malloc_mutex_rank_exclusive)) { + return true; + } + + if (opt_prof_final && opt_prof_prefix[0] != '\0' && + atexit(prof_fdump) != 0) { + malloc_write("<jemalloc>: Error in atexit()\n"); + if (opt_abort) { + abort(); + } + } + + if (opt_prof_log) { + prof_log_start(tsd_tsdn(tsd), NULL); + } + + if (atexit(prof_log_stop_final) != 0) { + malloc_write("<jemalloc>: Error in atexit() " + "for logging\n"); + if (opt_abort) { + abort(); + } + } + + if (malloc_mutex_init(&log_mtx, "prof_log", + WITNESS_RANK_PROF_LOG, malloc_mutex_rank_exclusive)) { + return true; + } + + if (ckh_new(tsd, &log_bt_node_set, PROF_CKH_MINITEMS, + prof_bt_node_hash, prof_bt_node_keycomp)) { + return true; + } + + if (ckh_new(tsd, &log_thr_node_set, PROF_CKH_MINITEMS, + prof_thr_node_hash, prof_thr_node_keycomp)) { + return true; + } + + log_tables_initialized = true; + + gctx_locks = (malloc_mutex_t *)base_alloc(tsd_tsdn(tsd), + b0get(), PROF_NCTX_LOCKS * sizeof(malloc_mutex_t), + CACHELINE); + if (gctx_locks == NULL) { + return true; + } + for (i = 0; i < PROF_NCTX_LOCKS; i++) { + if (malloc_mutex_init(&gctx_locks[i], "prof_gctx", + WITNESS_RANK_PROF_GCTX, + malloc_mutex_rank_exclusive)) { + return true; + } + } + + tdata_locks = (malloc_mutex_t *)base_alloc(tsd_tsdn(tsd), + b0get(), PROF_NTDATA_LOCKS * sizeof(malloc_mutex_t), + CACHELINE); + if (tdata_locks == NULL) { + return true; + } + for (i = 0; i < PROF_NTDATA_LOCKS; i++) { + if (malloc_mutex_init(&tdata_locks[i], "prof_tdata", + WITNESS_RANK_PROF_TDATA, + malloc_mutex_rank_exclusive)) { + return true; + } + } +#ifdef JEMALLOC_PROF_LIBGCC + /* + * Cause the backtracing machinery to allocate its internal + * state before enabling profiling. + */ + _Unwind_Backtrace(prof_unwind_init_callback, NULL); +#endif + } + prof_booted = true; + + return false; +} + +void +prof_prefork0(tsdn_t *tsdn) { + if (config_prof && opt_prof) { + unsigned i; + + malloc_mutex_prefork(tsdn, &prof_dump_mtx); + malloc_mutex_prefork(tsdn, &bt2gctx_mtx); + malloc_mutex_prefork(tsdn, &tdatas_mtx); + for (i = 0; i < PROF_NTDATA_LOCKS; i++) { + malloc_mutex_prefork(tsdn, &tdata_locks[i]); + } + for (i = 0; i < PROF_NCTX_LOCKS; i++) { + malloc_mutex_prefork(tsdn, &gctx_locks[i]); + } + } +} + +void +prof_prefork1(tsdn_t *tsdn) { + if (config_prof && opt_prof) { + malloc_mutex_prefork(tsdn, &prof_active_mtx); + malloc_mutex_prefork(tsdn, &prof_dump_seq_mtx); + malloc_mutex_prefork(tsdn, &prof_gdump_mtx); + malloc_mutex_prefork(tsdn, &next_thr_uid_mtx); + malloc_mutex_prefork(tsdn, &prof_thread_active_init_mtx); + } +} + +void +prof_postfork_parent(tsdn_t *tsdn) { + if (config_prof && opt_prof) { + unsigned i; + + malloc_mutex_postfork_parent(tsdn, + &prof_thread_active_init_mtx); + malloc_mutex_postfork_parent(tsdn, &next_thr_uid_mtx); + malloc_mutex_postfork_parent(tsdn, &prof_gdump_mtx); + malloc_mutex_postfork_parent(tsdn, &prof_dump_seq_mtx); + malloc_mutex_postfork_parent(tsdn, &prof_active_mtx); + for (i = 0; i < PROF_NCTX_LOCKS; i++) { + malloc_mutex_postfork_parent(tsdn, &gctx_locks[i]); + } + for (i = 0; i < PROF_NTDATA_LOCKS; i++) { + malloc_mutex_postfork_parent(tsdn, &tdata_locks[i]); + } + malloc_mutex_postfork_parent(tsdn, &tdatas_mtx); + malloc_mutex_postfork_parent(tsdn, &bt2gctx_mtx); + malloc_mutex_postfork_parent(tsdn, &prof_dump_mtx); + } +} + +void +prof_postfork_child(tsdn_t *tsdn) { + if (config_prof && opt_prof) { + unsigned i; + + malloc_mutex_postfork_child(tsdn, &prof_thread_active_init_mtx); + malloc_mutex_postfork_child(tsdn, &next_thr_uid_mtx); + malloc_mutex_postfork_child(tsdn, &prof_gdump_mtx); + malloc_mutex_postfork_child(tsdn, &prof_dump_seq_mtx); + malloc_mutex_postfork_child(tsdn, &prof_active_mtx); + for (i = 0; i < PROF_NCTX_LOCKS; i++) { + malloc_mutex_postfork_child(tsdn, &gctx_locks[i]); + } + for (i = 0; i < PROF_NTDATA_LOCKS; i++) { + malloc_mutex_postfork_child(tsdn, &tdata_locks[i]); + } + malloc_mutex_postfork_child(tsdn, &tdatas_mtx); + malloc_mutex_postfork_child(tsdn, &bt2gctx_mtx); + malloc_mutex_postfork_child(tsdn, &prof_dump_mtx); + } +} + +/******************************************************************************/ diff --git a/deps/jemalloc/src/rtree.c b/deps/jemalloc/src/rtree.c new file mode 100644 index 000000000..4ae41fe2f --- /dev/null +++ b/deps/jemalloc/src/rtree.c @@ -0,0 +1,320 @@ +#define JEMALLOC_RTREE_C_ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/assert.h" +#include "jemalloc/internal/mutex.h" + +/* + * Only the most significant bits of keys passed to rtree_{read,write}() are + * used. + */ +bool +rtree_new(rtree_t *rtree, bool zeroed) { +#ifdef JEMALLOC_JET + if (!zeroed) { + memset(rtree, 0, sizeof(rtree_t)); /* Clear root. */ + } +#else + assert(zeroed); +#endif + + if (malloc_mutex_init(&rtree->init_lock, "rtree", WITNESS_RANK_RTREE, + malloc_mutex_rank_exclusive)) { + return true; + } + + return false; +} + +static rtree_node_elm_t * +rtree_node_alloc_impl(tsdn_t *tsdn, rtree_t *rtree, size_t nelms) { + return (rtree_node_elm_t *)base_alloc(tsdn, b0get(), nelms * + sizeof(rtree_node_elm_t), CACHELINE); +} +rtree_node_alloc_t *JET_MUTABLE rtree_node_alloc = rtree_node_alloc_impl; + +static void +rtree_node_dalloc_impl(tsdn_t *tsdn, rtree_t *rtree, rtree_node_elm_t *node) { + /* Nodes are never deleted during normal operation. */ + not_reached(); +} +rtree_node_dalloc_t *JET_MUTABLE rtree_node_dalloc = + rtree_node_dalloc_impl; + +static rtree_leaf_elm_t * +rtree_leaf_alloc_impl(tsdn_t *tsdn, rtree_t *rtree, size_t nelms) { + return (rtree_leaf_elm_t *)base_alloc(tsdn, b0get(), nelms * + sizeof(rtree_leaf_elm_t), CACHELINE); +} +rtree_leaf_alloc_t *JET_MUTABLE rtree_leaf_alloc = rtree_leaf_alloc_impl; + +static void +rtree_leaf_dalloc_impl(tsdn_t *tsdn, rtree_t *rtree, rtree_leaf_elm_t *leaf) { + /* Leaves are never deleted during normal operation. */ + not_reached(); +} +rtree_leaf_dalloc_t *JET_MUTABLE rtree_leaf_dalloc = + rtree_leaf_dalloc_impl; + +#ifdef JEMALLOC_JET +# if RTREE_HEIGHT > 1 +static void +rtree_delete_subtree(tsdn_t *tsdn, rtree_t *rtree, rtree_node_elm_t *subtree, + unsigned level) { + size_t nchildren = ZU(1) << rtree_levels[level].bits; + if (level + 2 < RTREE_HEIGHT) { + for (size_t i = 0; i < nchildren; i++) { + rtree_node_elm_t *node = + (rtree_node_elm_t *)atomic_load_p(&subtree[i].child, + ATOMIC_RELAXED); + if (node != NULL) { + rtree_delete_subtree(tsdn, rtree, node, level + + 1); + } + } + } else { + for (size_t i = 0; i < nchildren; i++) { + rtree_leaf_elm_t *leaf = + (rtree_leaf_elm_t *)atomic_load_p(&subtree[i].child, + ATOMIC_RELAXED); + if (leaf != NULL) { + rtree_leaf_dalloc(tsdn, rtree, leaf); + } + } + } + + if (subtree != rtree->root) { + rtree_node_dalloc(tsdn, rtree, subtree); + } +} +# endif + +void +rtree_delete(tsdn_t *tsdn, rtree_t *rtree) { +# if RTREE_HEIGHT > 1 + rtree_delete_subtree(tsdn, rtree, rtree->root, 0); +# endif +} +#endif + +static rtree_node_elm_t * +rtree_node_init(tsdn_t *tsdn, rtree_t *rtree, unsigned level, + atomic_p_t *elmp) { + malloc_mutex_lock(tsdn, &rtree->init_lock); + /* + * If *elmp is non-null, then it was initialized with the init lock + * held, so we can get by with 'relaxed' here. + */ + rtree_node_elm_t *node = atomic_load_p(elmp, ATOMIC_RELAXED); + if (node == NULL) { + node = rtree_node_alloc(tsdn, rtree, ZU(1) << + rtree_levels[level].bits); + if (node == NULL) { + malloc_mutex_unlock(tsdn, &rtree->init_lock); + return NULL; + } + /* + * Even though we hold the lock, a later reader might not; we + * need release semantics. + */ + atomic_store_p(elmp, node, ATOMIC_RELEASE); + } + malloc_mutex_unlock(tsdn, &rtree->init_lock); + + return node; +} + +static rtree_leaf_elm_t * +rtree_leaf_init(tsdn_t *tsdn, rtree_t *rtree, atomic_p_t *elmp) { + malloc_mutex_lock(tsdn, &rtree->init_lock); + /* + * If *elmp is non-null, then it was initialized with the init lock + * held, so we can get by with 'relaxed' here. + */ + rtree_leaf_elm_t *leaf = atomic_load_p(elmp, ATOMIC_RELAXED); + if (leaf == NULL) { + leaf = rtree_leaf_alloc(tsdn, rtree, ZU(1) << + rtree_levels[RTREE_HEIGHT-1].bits); + if (leaf == NULL) { + malloc_mutex_unlock(tsdn, &rtree->init_lock); + return NULL; + } + /* + * Even though we hold the lock, a later reader might not; we + * need release semantics. + */ + atomic_store_p(elmp, leaf, ATOMIC_RELEASE); + } + malloc_mutex_unlock(tsdn, &rtree->init_lock); + + return leaf; +} + +static bool +rtree_node_valid(rtree_node_elm_t *node) { + return ((uintptr_t)node != (uintptr_t)0); +} + +static bool +rtree_leaf_valid(rtree_leaf_elm_t *leaf) { + return ((uintptr_t)leaf != (uintptr_t)0); +} + +static rtree_node_elm_t * +rtree_child_node_tryread(rtree_node_elm_t *elm, bool dependent) { + rtree_node_elm_t *node; + + if (dependent) { + node = (rtree_node_elm_t *)atomic_load_p(&elm->child, + ATOMIC_RELAXED); + } else { + node = (rtree_node_elm_t *)atomic_load_p(&elm->child, + ATOMIC_ACQUIRE); + } + + assert(!dependent || node != NULL); + return node; +} + +static rtree_node_elm_t * +rtree_child_node_read(tsdn_t *tsdn, rtree_t *rtree, rtree_node_elm_t *elm, + unsigned level, bool dependent) { + rtree_node_elm_t *node; + + node = rtree_child_node_tryread(elm, dependent); + if (!dependent && unlikely(!rtree_node_valid(node))) { + node = rtree_node_init(tsdn, rtree, level + 1, &elm->child); + } + assert(!dependent || node != NULL); + return node; +} + +static rtree_leaf_elm_t * +rtree_child_leaf_tryread(rtree_node_elm_t *elm, bool dependent) { + rtree_leaf_elm_t *leaf; + + if (dependent) { + leaf = (rtree_leaf_elm_t *)atomic_load_p(&elm->child, + ATOMIC_RELAXED); + } else { + leaf = (rtree_leaf_elm_t *)atomic_load_p(&elm->child, + ATOMIC_ACQUIRE); + } + + assert(!dependent || leaf != NULL); + return leaf; +} + +static rtree_leaf_elm_t * +rtree_child_leaf_read(tsdn_t *tsdn, rtree_t *rtree, rtree_node_elm_t *elm, + unsigned level, bool dependent) { + rtree_leaf_elm_t *leaf; + + leaf = rtree_child_leaf_tryread(elm, dependent); + if (!dependent && unlikely(!rtree_leaf_valid(leaf))) { + leaf = rtree_leaf_init(tsdn, rtree, &elm->child); + } + assert(!dependent || leaf != NULL); + return leaf; +} + +rtree_leaf_elm_t * +rtree_leaf_elm_lookup_hard(tsdn_t *tsdn, rtree_t *rtree, rtree_ctx_t *rtree_ctx, + uintptr_t key, bool dependent, bool init_missing) { + rtree_node_elm_t *node; + rtree_leaf_elm_t *leaf; +#if RTREE_HEIGHT > 1 + node = rtree->root; +#else + leaf = rtree->root; +#endif + + if (config_debug) { + uintptr_t leafkey = rtree_leafkey(key); + for (unsigned i = 0; i < RTREE_CTX_NCACHE; i++) { + assert(rtree_ctx->cache[i].leafkey != leafkey); + } + for (unsigned i = 0; i < RTREE_CTX_NCACHE_L2; i++) { + assert(rtree_ctx->l2_cache[i].leafkey != leafkey); + } + } + +#define RTREE_GET_CHILD(level) { \ + assert(level < RTREE_HEIGHT-1); \ + if (level != 0 && !dependent && \ + unlikely(!rtree_node_valid(node))) { \ + return NULL; \ + } \ + uintptr_t subkey = rtree_subkey(key, level); \ + if (level + 2 < RTREE_HEIGHT) { \ + node = init_missing ? \ + rtree_child_node_read(tsdn, rtree, \ + &node[subkey], level, dependent) : \ + rtree_child_node_tryread(&node[subkey], \ + dependent); \ + } else { \ + leaf = init_missing ? \ + rtree_child_leaf_read(tsdn, rtree, \ + &node[subkey], level, dependent) : \ + rtree_child_leaf_tryread(&node[subkey], \ + dependent); \ + } \ + } + /* + * Cache replacement upon hard lookup (i.e. L1 & L2 rtree cache miss): + * (1) evict last entry in L2 cache; (2) move the collision slot from L1 + * cache down to L2; and 3) fill L1. + */ +#define RTREE_GET_LEAF(level) { \ + assert(level == RTREE_HEIGHT-1); \ + if (!dependent && unlikely(!rtree_leaf_valid(leaf))) { \ + return NULL; \ + } \ + if (RTREE_CTX_NCACHE_L2 > 1) { \ + memmove(&rtree_ctx->l2_cache[1], \ + &rtree_ctx->l2_cache[0], \ + sizeof(rtree_ctx_cache_elm_t) * \ + (RTREE_CTX_NCACHE_L2 - 1)); \ + } \ + size_t slot = rtree_cache_direct_map(key); \ + rtree_ctx->l2_cache[0].leafkey = \ + rtree_ctx->cache[slot].leafkey; \ + rtree_ctx->l2_cache[0].leaf = \ + rtree_ctx->cache[slot].leaf; \ + uintptr_t leafkey = rtree_leafkey(key); \ + rtree_ctx->cache[slot].leafkey = leafkey; \ + rtree_ctx->cache[slot].leaf = leaf; \ + uintptr_t subkey = rtree_subkey(key, level); \ + return &leaf[subkey]; \ + } + if (RTREE_HEIGHT > 1) { + RTREE_GET_CHILD(0) + } + if (RTREE_HEIGHT > 2) { + RTREE_GET_CHILD(1) + } + if (RTREE_HEIGHT > 3) { + for (unsigned i = 2; i < RTREE_HEIGHT-1; i++) { + RTREE_GET_CHILD(i) + } + } + RTREE_GET_LEAF(RTREE_HEIGHT-1) +#undef RTREE_GET_CHILD +#undef RTREE_GET_LEAF + not_reached(); +} + +void +rtree_ctx_data_init(rtree_ctx_t *ctx) { + for (unsigned i = 0; i < RTREE_CTX_NCACHE; i++) { + rtree_ctx_cache_elm_t *cache = &ctx->cache[i]; + cache->leafkey = RTREE_LEAFKEY_INVALID; + cache->leaf = NULL; + } + for (unsigned i = 0; i < RTREE_CTX_NCACHE_L2; i++) { + rtree_ctx_cache_elm_t *cache = &ctx->l2_cache[i]; + cache->leafkey = RTREE_LEAFKEY_INVALID; + cache->leaf = NULL; + } +} diff --git a/deps/jemalloc/src/safety_check.c b/deps/jemalloc/src/safety_check.c new file mode 100644 index 000000000..804155dcf --- /dev/null +++ b/deps/jemalloc/src/safety_check.c @@ -0,0 +1,24 @@ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +static void (*safety_check_abort)(const char *message); + +void safety_check_set_abort(void (*abort_fn)(const char *)) { + safety_check_abort = abort_fn; +} + +void safety_check_fail(const char *format, ...) { + char buf[MALLOC_PRINTF_BUFSIZE]; + + va_list ap; + va_start(ap, format); + malloc_vsnprintf(buf, MALLOC_PRINTF_BUFSIZE, format, ap); + va_end(ap); + + if (safety_check_abort == NULL) { + malloc_write(buf); + abort(); + } else { + safety_check_abort(buf); + } +} diff --git a/deps/jemalloc/src/sc.c b/deps/jemalloc/src/sc.c new file mode 100644 index 000000000..89ddb6ba6 --- /dev/null +++ b/deps/jemalloc/src/sc.c @@ -0,0 +1,313 @@ +#include "jemalloc/internal/jemalloc_preamble.h" + +#include "jemalloc/internal/assert.h" +#include "jemalloc/internal/bit_util.h" +#include "jemalloc/internal/bitmap.h" +#include "jemalloc/internal/pages.h" +#include "jemalloc/internal/sc.h" + +/* + * This module computes the size classes used to satisfy allocations. The logic + * here was ported more or less line-by-line from a shell script, and because of + * that is not the most idiomatic C. Eventually we should fix this, but for now + * at least the damage is compartmentalized to this file. + */ + +sc_data_t sc_data_global; + +static size_t +reg_size_compute(int lg_base, int lg_delta, int ndelta) { + return (ZU(1) << lg_base) + (ZU(ndelta) << lg_delta); +} + +/* Returns the number of pages in the slab. */ +static int +slab_size(int lg_page, int lg_base, int lg_delta, int ndelta) { + size_t page = (ZU(1) << lg_page); + size_t reg_size = reg_size_compute(lg_base, lg_delta, ndelta); + + size_t try_slab_size = page; + size_t try_nregs = try_slab_size / reg_size; + size_t perfect_slab_size = 0; + bool perfect = false; + /* + * This loop continues until we find the least common multiple of the + * page size and size class size. Size classes are all of the form + * base + ndelta * delta == (ndelta + base/ndelta) * delta, which is + * (ndelta + ngroup) * delta. The way we choose slabbing strategies + * means that delta is at most the page size and ndelta < ngroup. So + * the loop executes for at most 2 * ngroup - 1 iterations, which is + * also the bound on the number of pages in a slab chosen by default. + * With the current default settings, this is at most 7. + */ + while (!perfect) { + perfect_slab_size = try_slab_size; + size_t perfect_nregs = try_nregs; + try_slab_size += page; + try_nregs = try_slab_size / reg_size; + if (perfect_slab_size == perfect_nregs * reg_size) { + perfect = true; + } + } + return (int)(perfect_slab_size / page); +} + +static void +size_class( + /* Output. */ + sc_t *sc, + /* Configuration decisions. */ + int lg_max_lookup, int lg_page, int lg_ngroup, + /* Inputs specific to the size class. */ + int index, int lg_base, int lg_delta, int ndelta) { + sc->index = index; + sc->lg_base = lg_base; + sc->lg_delta = lg_delta; + sc->ndelta = ndelta; + sc->psz = (reg_size_compute(lg_base, lg_delta, ndelta) + % (ZU(1) << lg_page) == 0); + size_t size = (ZU(1) << lg_base) + (ZU(ndelta) << lg_delta); + if (index == 0) { + assert(!sc->psz); + } + if (size < (ZU(1) << (lg_page + lg_ngroup))) { + sc->bin = true; + sc->pgs = slab_size(lg_page, lg_base, lg_delta, ndelta); + } else { + sc->bin = false; + sc->pgs = 0; + } + if (size <= (ZU(1) << lg_max_lookup)) { + sc->lg_delta_lookup = lg_delta; + } else { + sc->lg_delta_lookup = 0; + } +} + +static void +size_classes( + /* Output. */ + sc_data_t *sc_data, + /* Determined by the system. */ + size_t lg_ptr_size, int lg_quantum, + /* Configuration decisions. */ + int lg_tiny_min, int lg_max_lookup, int lg_page, int lg_ngroup) { + int ptr_bits = (1 << lg_ptr_size) * 8; + int ngroup = (1 << lg_ngroup); + int ntiny = 0; + int nlbins = 0; + int lg_tiny_maxclass = (unsigned)-1; + int nbins = 0; + int npsizes = 0; + + int index = 0; + + int ndelta = 0; + int lg_base = lg_tiny_min; + int lg_delta = lg_base; + + /* Outputs that we update as we go. */ + size_t lookup_maxclass = 0; + size_t small_maxclass = 0; + int lg_large_minclass = 0; + size_t large_maxclass = 0; + + /* Tiny size classes. */ + while (lg_base < lg_quantum) { + sc_t *sc = &sc_data->sc[index]; + size_class(sc, lg_max_lookup, lg_page, lg_ngroup, index, + lg_base, lg_delta, ndelta); + if (sc->lg_delta_lookup != 0) { + nlbins = index + 1; + } + if (sc->psz) { + npsizes++; + } + if (sc->bin) { + nbins++; + } + ntiny++; + /* Final written value is correct. */ + lg_tiny_maxclass = lg_base; + index++; + lg_delta = lg_base; + lg_base++; + } + + /* First non-tiny (pseudo) group. */ + if (ntiny != 0) { + sc_t *sc = &sc_data->sc[index]; + /* + * See the note in sc.h; the first non-tiny size class has an + * unusual encoding. + */ + lg_base--; + ndelta = 1; + size_class(sc, lg_max_lookup, lg_page, lg_ngroup, index, + lg_base, lg_delta, ndelta); + index++; + lg_base++; + lg_delta++; + if (sc->psz) { + npsizes++; + } + if (sc->bin) { + nbins++; + } + } + while (ndelta < ngroup) { + sc_t *sc = &sc_data->sc[index]; + size_class(sc, lg_max_lookup, lg_page, lg_ngroup, index, + lg_base, lg_delta, ndelta); + index++; + ndelta++; + if (sc->psz) { + npsizes++; + } + if (sc->bin) { + nbins++; + } + } + + /* All remaining groups. */ + lg_base = lg_base + lg_ngroup; + while (lg_base < ptr_bits - 1) { + ndelta = 1; + int ndelta_limit; + if (lg_base == ptr_bits - 2) { + ndelta_limit = ngroup - 1; + } else { + ndelta_limit = ngroup; + } + while (ndelta <= ndelta_limit) { + sc_t *sc = &sc_data->sc[index]; + size_class(sc, lg_max_lookup, lg_page, lg_ngroup, index, + lg_base, lg_delta, ndelta); + if (sc->lg_delta_lookup != 0) { + nlbins = index + 1; + /* Final written value is correct. */ + lookup_maxclass = (ZU(1) << lg_base) + + (ZU(ndelta) << lg_delta); + } + if (sc->psz) { + npsizes++; + } + if (sc->bin) { + nbins++; + /* Final written value is correct. */ + small_maxclass = (ZU(1) << lg_base) + + (ZU(ndelta) << lg_delta); + if (lg_ngroup > 0) { + lg_large_minclass = lg_base + 1; + } else { + lg_large_minclass = lg_base + 2; + } + } + large_maxclass = (ZU(1) << lg_base) + + (ZU(ndelta) << lg_delta); + index++; + ndelta++; + } + lg_base++; + lg_delta++; + } + /* Additional outputs. */ + int nsizes = index; + unsigned lg_ceil_nsizes = lg_ceil(nsizes); + + /* Fill in the output data. */ + sc_data->ntiny = ntiny; + sc_data->nlbins = nlbins; + sc_data->nbins = nbins; + sc_data->nsizes = nsizes; + sc_data->lg_ceil_nsizes = lg_ceil_nsizes; + sc_data->npsizes = npsizes; + sc_data->lg_tiny_maxclass = lg_tiny_maxclass; + sc_data->lookup_maxclass = lookup_maxclass; + sc_data->small_maxclass = small_maxclass; + sc_data->lg_large_minclass = lg_large_minclass; + sc_data->large_minclass = (ZU(1) << lg_large_minclass); + sc_data->large_maxclass = large_maxclass; + + /* + * We compute these values in two ways: + * - Incrementally, as above. + * - In macros, in sc.h. + * The computation is easier when done incrementally, but putting it in + * a constant makes it available to the fast paths without having to + * touch the extra global cacheline. We assert, however, that the two + * computations are equivalent. + */ + assert(sc_data->npsizes == SC_NPSIZES); + assert(sc_data->lg_tiny_maxclass == SC_LG_TINY_MAXCLASS); + assert(sc_data->small_maxclass == SC_SMALL_MAXCLASS); + assert(sc_data->large_minclass == SC_LARGE_MINCLASS); + assert(sc_data->lg_large_minclass == SC_LG_LARGE_MINCLASS); + assert(sc_data->large_maxclass == SC_LARGE_MAXCLASS); + + /* + * In the allocation fastpath, we want to assume that we can + * unconditionally subtract the requested allocation size from + * a ssize_t, and detect passing through 0 correctly. This + * results in optimal generated code. For this to work, the + * maximum allocation size must be less than SSIZE_MAX. + */ + assert(SC_LARGE_MAXCLASS < SSIZE_MAX); +} + +void +sc_data_init(sc_data_t *sc_data) { + assert(!sc_data->initialized); + + int lg_max_lookup = 12; + + size_classes(sc_data, LG_SIZEOF_PTR, LG_QUANTUM, SC_LG_TINY_MIN, + lg_max_lookup, LG_PAGE, 2); + + sc_data->initialized = true; +} + +static void +sc_data_update_sc_slab_size(sc_t *sc, size_t reg_size, size_t pgs_guess) { + size_t min_pgs = reg_size / PAGE; + if (reg_size % PAGE != 0) { + min_pgs++; + } + /* + * BITMAP_MAXBITS is actually determined by putting the smallest + * possible size-class on one page, so this can never be 0. + */ + size_t max_pgs = BITMAP_MAXBITS * reg_size / PAGE; + + assert(min_pgs <= max_pgs); + assert(min_pgs > 0); + assert(max_pgs >= 1); + if (pgs_guess < min_pgs) { + sc->pgs = (int)min_pgs; + } else if (pgs_guess > max_pgs) { + sc->pgs = (int)max_pgs; + } else { + sc->pgs = (int)pgs_guess; + } +} + +void +sc_data_update_slab_size(sc_data_t *data, size_t begin, size_t end, int pgs) { + assert(data->initialized); + for (int i = 0; i < data->nsizes; i++) { + sc_t *sc = &data->sc[i]; + if (!sc->bin) { + break; + } + size_t reg_size = reg_size_compute(sc->lg_base, sc->lg_delta, + sc->ndelta); + if (begin <= reg_size && reg_size <= end) { + sc_data_update_sc_slab_size(sc, reg_size, pgs); + } + } +} + +void +sc_boot(sc_data_t *data) { + sc_data_init(data); +} diff --git a/deps/jemalloc/src/stats.c b/deps/jemalloc/src/stats.c new file mode 100644 index 000000000..118e05d29 --- /dev/null +++ b/deps/jemalloc/src/stats.c @@ -0,0 +1,1457 @@ +#define JEMALLOC_STATS_C_ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/assert.h" +#include "jemalloc/internal/ctl.h" +#include "jemalloc/internal/emitter.h" +#include "jemalloc/internal/mutex.h" +#include "jemalloc/internal/mutex_prof.h" + +const char *global_mutex_names[mutex_prof_num_global_mutexes] = { +#define OP(mtx) #mtx, + MUTEX_PROF_GLOBAL_MUTEXES +#undef OP +}; + +const char *arena_mutex_names[mutex_prof_num_arena_mutexes] = { +#define OP(mtx) #mtx, + MUTEX_PROF_ARENA_MUTEXES +#undef OP +}; + +#define CTL_GET(n, v, t) do { \ + size_t sz = sizeof(t); \ + xmallctl(n, (void *)v, &sz, NULL, 0); \ +} while (0) + +#define CTL_M2_GET(n, i, v, t) do { \ + size_t mib[CTL_MAX_DEPTH]; \ + size_t miblen = sizeof(mib) / sizeof(size_t); \ + size_t sz = sizeof(t); \ + xmallctlnametomib(n, mib, &miblen); \ + mib[2] = (i); \ + xmallctlbymib(mib, miblen, (void *)v, &sz, NULL, 0); \ +} while (0) + +#define CTL_M2_M4_GET(n, i, j, v, t) do { \ + size_t mib[CTL_MAX_DEPTH]; \ + size_t miblen = sizeof(mib) / sizeof(size_t); \ + size_t sz = sizeof(t); \ + xmallctlnametomib(n, mib, &miblen); \ + mib[2] = (i); \ + mib[4] = (j); \ + xmallctlbymib(mib, miblen, (void *)v, &sz, NULL, 0); \ +} while (0) + +/******************************************************************************/ +/* Data. */ + +bool opt_stats_print = false; +char opt_stats_print_opts[stats_print_tot_num_options+1] = ""; + +/******************************************************************************/ + +static uint64_t +rate_per_second(uint64_t value, uint64_t uptime_ns) { + uint64_t billion = 1000000000; + if (uptime_ns == 0 || value == 0) { + return 0; + } + if (uptime_ns < billion) { + return value; + } else { + uint64_t uptime_s = uptime_ns / billion; + return value / uptime_s; + } +} + +/* Calculate x.yyy and output a string (takes a fixed sized char array). */ +static bool +get_rate_str(uint64_t dividend, uint64_t divisor, char str[6]) { + if (divisor == 0 || dividend > divisor) { + /* The rate is not supposed to be greater than 1. */ + return true; + } + if (dividend > 0) { + assert(UINT64_MAX / dividend >= 1000); + } + + unsigned n = (unsigned)((dividend * 1000) / divisor); + if (n < 10) { + malloc_snprintf(str, 6, "0.00%u", n); + } else if (n < 100) { + malloc_snprintf(str, 6, "0.0%u", n); + } else if (n < 1000) { + malloc_snprintf(str, 6, "0.%u", n); + } else { + malloc_snprintf(str, 6, "1"); + } + + return false; +} + +#define MUTEX_CTL_STR_MAX_LENGTH 128 +static void +gen_mutex_ctl_str(char *str, size_t buf_len, const char *prefix, + const char *mutex, const char *counter) { + malloc_snprintf(str, buf_len, "stats.%s.%s.%s", prefix, mutex, counter); +} + +static void +mutex_stats_init_cols(emitter_row_t *row, const char *table_name, + emitter_col_t *name, + emitter_col_t col_uint64_t[mutex_prof_num_uint64_t_counters], + emitter_col_t col_uint32_t[mutex_prof_num_uint32_t_counters]) { + mutex_prof_uint64_t_counter_ind_t k_uint64_t = 0; + mutex_prof_uint32_t_counter_ind_t k_uint32_t = 0; + + emitter_col_t *col; + + if (name != NULL) { + emitter_col_init(name, row); + name->justify = emitter_justify_left; + name->width = 21; + name->type = emitter_type_title; + name->str_val = table_name; + } + +#define WIDTH_uint32_t 12 +#define WIDTH_uint64_t 16 +#define OP(counter, counter_type, human, derived, base_counter) \ + col = &col_##counter_type[k_##counter_type]; \ + ++k_##counter_type; \ + emitter_col_init(col, row); \ + col->justify = emitter_justify_right; \ + col->width = derived ? 8 : WIDTH_##counter_type; \ + col->type = emitter_type_title; \ + col->str_val = human; + MUTEX_PROF_COUNTERS +#undef OP +#undef WIDTH_uint32_t +#undef WIDTH_uint64_t + col_uint64_t[mutex_counter_total_wait_time_ps].width = 10; +} + +static void +mutex_stats_read_global(const char *name, emitter_col_t *col_name, + emitter_col_t col_uint64_t[mutex_prof_num_uint64_t_counters], + emitter_col_t col_uint32_t[mutex_prof_num_uint32_t_counters], + uint64_t uptime) { + char cmd[MUTEX_CTL_STR_MAX_LENGTH]; + + col_name->str_val = name; + + emitter_col_t *dst; +#define EMITTER_TYPE_uint32_t emitter_type_uint32 +#define EMITTER_TYPE_uint64_t emitter_type_uint64 +#define OP(counter, counter_type, human, derived, base_counter) \ + dst = &col_##counter_type[mutex_counter_##counter]; \ + dst->type = EMITTER_TYPE_##counter_type; \ + if (!derived) { \ + gen_mutex_ctl_str(cmd, MUTEX_CTL_STR_MAX_LENGTH, \ + "mutexes", name, #counter); \ + CTL_GET(cmd, (counter_type *)&dst->bool_val, counter_type); \ + } else { \ + emitter_col_t *base = &col_##counter_type[mutex_counter_##base_counter]; \ + dst->counter_type##_val = rate_per_second(base->counter_type##_val, uptime); \ + } + MUTEX_PROF_COUNTERS +#undef OP +#undef EMITTER_TYPE_uint32_t +#undef EMITTER_TYPE_uint64_t +} + +static void +mutex_stats_read_arena(unsigned arena_ind, mutex_prof_arena_ind_t mutex_ind, + const char *name, emitter_col_t *col_name, + emitter_col_t col_uint64_t[mutex_prof_num_uint64_t_counters], + emitter_col_t col_uint32_t[mutex_prof_num_uint32_t_counters], + uint64_t uptime) { + char cmd[MUTEX_CTL_STR_MAX_LENGTH]; + + col_name->str_val = name; + + emitter_col_t *dst; +#define EMITTER_TYPE_uint32_t emitter_type_uint32 +#define EMITTER_TYPE_uint64_t emitter_type_uint64 +#define OP(counter, counter_type, human, derived, base_counter) \ + dst = &col_##counter_type[mutex_counter_##counter]; \ + dst->type = EMITTER_TYPE_##counter_type; \ + if (!derived) { \ + gen_mutex_ctl_str(cmd, MUTEX_CTL_STR_MAX_LENGTH, \ + "arenas.0.mutexes", arena_mutex_names[mutex_ind], #counter);\ + CTL_M2_GET(cmd, arena_ind, (counter_type *)&dst->bool_val, counter_type); \ + } else { \ + emitter_col_t *base = &col_##counter_type[mutex_counter_##base_counter]; \ + dst->counter_type##_val = rate_per_second(base->counter_type##_val, uptime); \ + } + MUTEX_PROF_COUNTERS +#undef OP +#undef EMITTER_TYPE_uint32_t +#undef EMITTER_TYPE_uint64_t +} + +static void +mutex_stats_read_arena_bin(unsigned arena_ind, unsigned bin_ind, + emitter_col_t col_uint64_t[mutex_prof_num_uint64_t_counters], + emitter_col_t col_uint32_t[mutex_prof_num_uint32_t_counters], + uint64_t uptime) { + char cmd[MUTEX_CTL_STR_MAX_LENGTH]; + emitter_col_t *dst; + +#define EMITTER_TYPE_uint32_t emitter_type_uint32 +#define EMITTER_TYPE_uint64_t emitter_type_uint64 +#define OP(counter, counter_type, human, derived, base_counter) \ + dst = &col_##counter_type[mutex_counter_##counter]; \ + dst->type = EMITTER_TYPE_##counter_type; \ + if (!derived) { \ + gen_mutex_ctl_str(cmd, MUTEX_CTL_STR_MAX_LENGTH, \ + "arenas.0.bins.0","mutex", #counter); \ + CTL_M2_M4_GET(cmd, arena_ind, bin_ind, \ + (counter_type *)&dst->bool_val, counter_type); \ + } else { \ + emitter_col_t *base = &col_##counter_type[mutex_counter_##base_counter]; \ + dst->counter_type##_val = rate_per_second(base->counter_type##_val, uptime); \ + } + MUTEX_PROF_COUNTERS +#undef OP +#undef EMITTER_TYPE_uint32_t +#undef EMITTER_TYPE_uint64_t +} + +/* "row" can be NULL to avoid emitting in table mode. */ +static void +mutex_stats_emit(emitter_t *emitter, emitter_row_t *row, + emitter_col_t col_uint64_t[mutex_prof_num_uint64_t_counters], + emitter_col_t col_uint32_t[mutex_prof_num_uint32_t_counters]) { + if (row != NULL) { + emitter_table_row(emitter, row); + } + + mutex_prof_uint64_t_counter_ind_t k_uint64_t = 0; + mutex_prof_uint32_t_counter_ind_t k_uint32_t = 0; + + emitter_col_t *col; + +#define EMITTER_TYPE_uint32_t emitter_type_uint32 +#define EMITTER_TYPE_uint64_t emitter_type_uint64 +#define OP(counter, type, human, derived, base_counter) \ + if (!derived) { \ + col = &col_##type[k_##type]; \ + ++k_##type; \ + emitter_json_kv(emitter, #counter, EMITTER_TYPE_##type, \ + (const void *)&col->bool_val); \ + } + MUTEX_PROF_COUNTERS; +#undef OP +#undef EMITTER_TYPE_uint32_t +#undef EMITTER_TYPE_uint64_t +} + +#define COL(row_name, column_name, left_or_right, col_width, etype) \ + emitter_col_t col_##column_name; \ + emitter_col_init(&col_##column_name, &row_name); \ + col_##column_name.justify = emitter_justify_##left_or_right; \ + col_##column_name.width = col_width; \ + col_##column_name.type = emitter_type_##etype; + +#define COL_HDR(row_name, column_name, human, left_or_right, col_width, etype) \ + COL(row_name, column_name, left_or_right, col_width, etype) \ + emitter_col_t header_##column_name; \ + emitter_col_init(&header_##column_name, &header_##row_name); \ + header_##column_name.justify = emitter_justify_##left_or_right; \ + header_##column_name.width = col_width; \ + header_##column_name.type = emitter_type_title; \ + header_##column_name.str_val = human ? human : #column_name; + + +static void +stats_arena_bins_print(emitter_t *emitter, bool mutex, unsigned i, uint64_t uptime) { + size_t page; + bool in_gap, in_gap_prev; + unsigned nbins, j; + + CTL_GET("arenas.page", &page, size_t); + + CTL_GET("arenas.nbins", &nbins, unsigned); + + emitter_row_t header_row; + emitter_row_init(&header_row); + + emitter_row_t row; + emitter_row_init(&row); + + COL_HDR(row, size, NULL, right, 20, size) + COL_HDR(row, ind, NULL, right, 4, unsigned) + COL_HDR(row, allocated, NULL, right, 13, uint64) + COL_HDR(row, nmalloc, NULL, right, 13, uint64) + COL_HDR(row, nmalloc_ps, "(#/sec)", right, 8, uint64) + COL_HDR(row, ndalloc, NULL, right, 13, uint64) + COL_HDR(row, ndalloc_ps, "(#/sec)", right, 8, uint64) + COL_HDR(row, nrequests, NULL, right, 13, uint64) + COL_HDR(row, nrequests_ps, "(#/sec)", right, 10, uint64) + COL_HDR(row, nshards, NULL, right, 9, unsigned) + COL_HDR(row, curregs, NULL, right, 13, size) + COL_HDR(row, curslabs, NULL, right, 13, size) + COL_HDR(row, nonfull_slabs, NULL, right, 15, size) + COL_HDR(row, regs, NULL, right, 5, unsigned) + COL_HDR(row, pgs, NULL, right, 4, size) + /* To buffer a right- and left-justified column. */ + COL_HDR(row, justify_spacer, NULL, right, 1, title) + COL_HDR(row, util, NULL, right, 6, title) + COL_HDR(row, nfills, NULL, right, 13, uint64) + COL_HDR(row, nfills_ps, "(#/sec)", right, 8, uint64) + COL_HDR(row, nflushes, NULL, right, 13, uint64) + COL_HDR(row, nflushes_ps, "(#/sec)", right, 8, uint64) + COL_HDR(row, nslabs, NULL, right, 13, uint64) + COL_HDR(row, nreslabs, NULL, right, 13, uint64) + COL_HDR(row, nreslabs_ps, "(#/sec)", right, 8, uint64) + + /* Don't want to actually print the name. */ + header_justify_spacer.str_val = " "; + col_justify_spacer.str_val = " "; + + emitter_col_t col_mutex64[mutex_prof_num_uint64_t_counters]; + emitter_col_t col_mutex32[mutex_prof_num_uint32_t_counters]; + + emitter_col_t header_mutex64[mutex_prof_num_uint64_t_counters]; + emitter_col_t header_mutex32[mutex_prof_num_uint32_t_counters]; + + if (mutex) { + mutex_stats_init_cols(&row, NULL, NULL, col_mutex64, + col_mutex32); + mutex_stats_init_cols(&header_row, NULL, NULL, header_mutex64, + header_mutex32); + } + + /* + * We print a "bins:" header as part of the table row; we need to adjust + * the header size column to compensate. + */ + header_size.width -=5; + emitter_table_printf(emitter, "bins:"); + emitter_table_row(emitter, &header_row); + emitter_json_array_kv_begin(emitter, "bins"); + + for (j = 0, in_gap = false; j < nbins; j++) { + uint64_t nslabs; + size_t reg_size, slab_size, curregs; + size_t curslabs; + size_t nonfull_slabs; + uint32_t nregs, nshards; + uint64_t nmalloc, ndalloc, nrequests, nfills, nflushes; + uint64_t nreslabs; + + CTL_M2_M4_GET("stats.arenas.0.bins.0.nslabs", i, j, &nslabs, + uint64_t); + in_gap_prev = in_gap; + in_gap = (nslabs == 0); + + if (in_gap_prev && !in_gap) { + emitter_table_printf(emitter, + " ---\n"); + } + + CTL_M2_GET("arenas.bin.0.size", j, ®_size, size_t); + CTL_M2_GET("arenas.bin.0.nregs", j, &nregs, uint32_t); + CTL_M2_GET("arenas.bin.0.slab_size", j, &slab_size, size_t); + CTL_M2_GET("arenas.bin.0.nshards", j, &nshards, uint32_t); + + CTL_M2_M4_GET("stats.arenas.0.bins.0.nmalloc", i, j, &nmalloc, + uint64_t); + CTL_M2_M4_GET("stats.arenas.0.bins.0.ndalloc", i, j, &ndalloc, + uint64_t); + CTL_M2_M4_GET("stats.arenas.0.bins.0.curregs", i, j, &curregs, + size_t); + CTL_M2_M4_GET("stats.arenas.0.bins.0.nrequests", i, j, + &nrequests, uint64_t); + CTL_M2_M4_GET("stats.arenas.0.bins.0.nfills", i, j, &nfills, + uint64_t); + CTL_M2_M4_GET("stats.arenas.0.bins.0.nflushes", i, j, &nflushes, + uint64_t); + CTL_M2_M4_GET("stats.arenas.0.bins.0.nreslabs", i, j, &nreslabs, + uint64_t); + CTL_M2_M4_GET("stats.arenas.0.bins.0.curslabs", i, j, &curslabs, + size_t); + CTL_M2_M4_GET("stats.arenas.0.bins.0.nonfull_slabs", i, j, &nonfull_slabs, + size_t); + + if (mutex) { + mutex_stats_read_arena_bin(i, j, col_mutex64, + col_mutex32, uptime); + } + + emitter_json_object_begin(emitter); + emitter_json_kv(emitter, "nmalloc", emitter_type_uint64, + &nmalloc); + emitter_json_kv(emitter, "ndalloc", emitter_type_uint64, + &ndalloc); + emitter_json_kv(emitter, "curregs", emitter_type_size, + &curregs); + emitter_json_kv(emitter, "nrequests", emitter_type_uint64, + &nrequests); + emitter_json_kv(emitter, "nfills", emitter_type_uint64, + &nfills); + emitter_json_kv(emitter, "nflushes", emitter_type_uint64, + &nflushes); + emitter_json_kv(emitter, "nreslabs", emitter_type_uint64, + &nreslabs); + emitter_json_kv(emitter, "curslabs", emitter_type_size, + &curslabs); + emitter_json_kv(emitter, "nonfull_slabs", emitter_type_size, + &nonfull_slabs); + if (mutex) { + emitter_json_object_kv_begin(emitter, "mutex"); + mutex_stats_emit(emitter, NULL, col_mutex64, + col_mutex32); + emitter_json_object_end(emitter); + } + emitter_json_object_end(emitter); + + size_t availregs = nregs * curslabs; + char util[6]; + if (get_rate_str((uint64_t)curregs, (uint64_t)availregs, util)) + { + if (availregs == 0) { + malloc_snprintf(util, sizeof(util), "1"); + } else if (curregs > availregs) { + /* + * Race detected: the counters were read in + * separate mallctl calls and concurrent + * operations happened in between. In this case + * no meaningful utilization can be computed. + */ + malloc_snprintf(util, sizeof(util), " race"); + } else { + not_reached(); + } + } + + col_size.size_val = reg_size; + col_ind.unsigned_val = j; + col_allocated.size_val = curregs * reg_size; + col_nmalloc.uint64_val = nmalloc; + col_nmalloc_ps.uint64_val = rate_per_second(nmalloc, uptime); + col_ndalloc.uint64_val = ndalloc; + col_ndalloc_ps.uint64_val = rate_per_second(ndalloc, uptime); + col_nrequests.uint64_val = nrequests; + col_nrequests_ps.uint64_val = rate_per_second(nrequests, uptime); + col_nshards.unsigned_val = nshards; + col_curregs.size_val = curregs; + col_curslabs.size_val = curslabs; + col_nonfull_slabs.size_val = nonfull_slabs; + col_regs.unsigned_val = nregs; + col_pgs.size_val = slab_size / page; + col_util.str_val = util; + col_nfills.uint64_val = nfills; + col_nfills_ps.uint64_val = rate_per_second(nfills, uptime); + col_nflushes.uint64_val = nflushes; + col_nflushes_ps.uint64_val = rate_per_second(nflushes, uptime); + col_nslabs.uint64_val = nslabs; + col_nreslabs.uint64_val = nreslabs; + col_nreslabs_ps.uint64_val = rate_per_second(nreslabs, uptime); + + /* + * Note that mutex columns were initialized above, if mutex == + * true. + */ + + emitter_table_row(emitter, &row); + } + emitter_json_array_end(emitter); /* Close "bins". */ + + if (in_gap) { + emitter_table_printf(emitter, " ---\n"); + } +} + +static void +stats_arena_lextents_print(emitter_t *emitter, unsigned i, uint64_t uptime) { + unsigned nbins, nlextents, j; + bool in_gap, in_gap_prev; + + CTL_GET("arenas.nbins", &nbins, unsigned); + CTL_GET("arenas.nlextents", &nlextents, unsigned); + + emitter_row_t header_row; + emitter_row_init(&header_row); + emitter_row_t row; + emitter_row_init(&row); + + COL_HDR(row, size, NULL, right, 20, size) + COL_HDR(row, ind, NULL, right, 4, unsigned) + COL_HDR(row, allocated, NULL, right, 13, size) + COL_HDR(row, nmalloc, NULL, right, 13, uint64) + COL_HDR(row, nmalloc_ps, "(#/sec)", right, 8, uint64) + COL_HDR(row, ndalloc, NULL, right, 13, uint64) + COL_HDR(row, ndalloc_ps, "(#/sec)", right, 8, uint64) + COL_HDR(row, nrequests, NULL, right, 13, uint64) + COL_HDR(row, nrequests_ps, "(#/sec)", right, 8, uint64) + COL_HDR(row, curlextents, NULL, right, 13, size) + + /* As with bins, we label the large extents table. */ + header_size.width -= 6; + emitter_table_printf(emitter, "large:"); + emitter_table_row(emitter, &header_row); + emitter_json_array_kv_begin(emitter, "lextents"); + + for (j = 0, in_gap = false; j < nlextents; j++) { + uint64_t nmalloc, ndalloc, nrequests; + size_t lextent_size, curlextents; + + CTL_M2_M4_GET("stats.arenas.0.lextents.0.nmalloc", i, j, + &nmalloc, uint64_t); + CTL_M2_M4_GET("stats.arenas.0.lextents.0.ndalloc", i, j, + &ndalloc, uint64_t); + CTL_M2_M4_GET("stats.arenas.0.lextents.0.nrequests", i, j, + &nrequests, uint64_t); + in_gap_prev = in_gap; + in_gap = (nrequests == 0); + + if (in_gap_prev && !in_gap) { + emitter_table_printf(emitter, + " ---\n"); + } + + CTL_M2_GET("arenas.lextent.0.size", j, &lextent_size, size_t); + CTL_M2_M4_GET("stats.arenas.0.lextents.0.curlextents", i, j, + &curlextents, size_t); + + emitter_json_object_begin(emitter); + emitter_json_kv(emitter, "curlextents", emitter_type_size, + &curlextents); + emitter_json_object_end(emitter); + + col_size.size_val = lextent_size; + col_ind.unsigned_val = nbins + j; + col_allocated.size_val = curlextents * lextent_size; + col_nmalloc.uint64_val = nmalloc; + col_nmalloc_ps.uint64_val = rate_per_second(nmalloc, uptime); + col_ndalloc.uint64_val = ndalloc; + col_ndalloc_ps.uint64_val = rate_per_second(ndalloc, uptime); + col_nrequests.uint64_val = nrequests; + col_nrequests_ps.uint64_val = rate_per_second(nrequests, uptime); + col_curlextents.size_val = curlextents; + + if (!in_gap) { + emitter_table_row(emitter, &row); + } + } + emitter_json_array_end(emitter); /* Close "lextents". */ + if (in_gap) { + emitter_table_printf(emitter, " ---\n"); + } +} + +static void +stats_arena_extents_print(emitter_t *emitter, unsigned i) { + unsigned j; + bool in_gap, in_gap_prev; + emitter_row_t header_row; + emitter_row_init(&header_row); + emitter_row_t row; + emitter_row_init(&row); + + COL_HDR(row, size, NULL, right, 20, size) + COL_HDR(row, ind, NULL, right, 4, unsigned) + COL_HDR(row, ndirty, NULL, right, 13, size) + COL_HDR(row, dirty, NULL, right, 13, size) + COL_HDR(row, nmuzzy, NULL, right, 13, size) + COL_HDR(row, muzzy, NULL, right, 13, size) + COL_HDR(row, nretained, NULL, right, 13, size) + COL_HDR(row, retained, NULL, right, 13, size) + COL_HDR(row, ntotal, NULL, right, 13, size) + COL_HDR(row, total, NULL, right, 13, size) + + /* Label this section. */ + header_size.width -= 8; + emitter_table_printf(emitter, "extents:"); + emitter_table_row(emitter, &header_row); + emitter_json_array_kv_begin(emitter, "extents"); + + in_gap = false; + for (j = 0; j < SC_NPSIZES; j++) { + size_t ndirty, nmuzzy, nretained, total, dirty_bytes, + muzzy_bytes, retained_bytes, total_bytes; + CTL_M2_M4_GET("stats.arenas.0.extents.0.ndirty", i, j, + &ndirty, size_t); + CTL_M2_M4_GET("stats.arenas.0.extents.0.nmuzzy", i, j, + &nmuzzy, size_t); + CTL_M2_M4_GET("stats.arenas.0.extents.0.nretained", i, j, + &nretained, size_t); + CTL_M2_M4_GET("stats.arenas.0.extents.0.dirty_bytes", i, j, + &dirty_bytes, size_t); + CTL_M2_M4_GET("stats.arenas.0.extents.0.muzzy_bytes", i, j, + &muzzy_bytes, size_t); + CTL_M2_M4_GET("stats.arenas.0.extents.0.retained_bytes", i, j, + &retained_bytes, size_t); + total = ndirty + nmuzzy + nretained; + total_bytes = dirty_bytes + muzzy_bytes + retained_bytes; + + in_gap_prev = in_gap; + in_gap = (total == 0); + + if (in_gap_prev && !in_gap) { + emitter_table_printf(emitter, + " ---\n"); + } + + emitter_json_object_begin(emitter); + emitter_json_kv(emitter, "ndirty", emitter_type_size, &ndirty); + emitter_json_kv(emitter, "nmuzzy", emitter_type_size, &nmuzzy); + emitter_json_kv(emitter, "nretained", emitter_type_size, + &nretained); + + emitter_json_kv(emitter, "dirty_bytes", emitter_type_size, + &dirty_bytes); + emitter_json_kv(emitter, "muzzy_bytes", emitter_type_size, + &muzzy_bytes); + emitter_json_kv(emitter, "retained_bytes", emitter_type_size, + &retained_bytes); + emitter_json_object_end(emitter); + + col_size.size_val = sz_pind2sz(j); + col_ind.size_val = j; + col_ndirty.size_val = ndirty; + col_dirty.size_val = dirty_bytes; + col_nmuzzy.size_val = nmuzzy; + col_muzzy.size_val = muzzy_bytes; + col_nretained.size_val = nretained; + col_retained.size_val = retained_bytes; + col_ntotal.size_val = total; + col_total.size_val = total_bytes; + + if (!in_gap) { + emitter_table_row(emitter, &row); + } + } + emitter_json_array_end(emitter); /* Close "extents". */ + if (in_gap) { + emitter_table_printf(emitter, " ---\n"); + } +} + +static void +stats_arena_mutexes_print(emitter_t *emitter, unsigned arena_ind, uint64_t uptime) { + emitter_row_t row; + emitter_col_t col_name; + emitter_col_t col64[mutex_prof_num_uint64_t_counters]; + emitter_col_t col32[mutex_prof_num_uint32_t_counters]; + + emitter_row_init(&row); + mutex_stats_init_cols(&row, "", &col_name, col64, col32); + + emitter_json_object_kv_begin(emitter, "mutexes"); + emitter_table_row(emitter, &row); + + for (mutex_prof_arena_ind_t i = 0; i < mutex_prof_num_arena_mutexes; + i++) { + const char *name = arena_mutex_names[i]; + emitter_json_object_kv_begin(emitter, name); + mutex_stats_read_arena(arena_ind, i, name, &col_name, col64, + col32, uptime); + mutex_stats_emit(emitter, &row, col64, col32); + emitter_json_object_end(emitter); /* Close the mutex dict. */ + } + emitter_json_object_end(emitter); /* End "mutexes". */ +} + +static void +stats_arena_print(emitter_t *emitter, unsigned i, bool bins, bool large, + bool mutex, bool extents) { + unsigned nthreads; + const char *dss; + ssize_t dirty_decay_ms, muzzy_decay_ms; + size_t page, pactive, pdirty, pmuzzy, mapped, retained; + size_t base, internal, resident, metadata_thp, extent_avail; + uint64_t dirty_npurge, dirty_nmadvise, dirty_purged; + uint64_t muzzy_npurge, muzzy_nmadvise, muzzy_purged; + size_t small_allocated; + uint64_t small_nmalloc, small_ndalloc, small_nrequests, small_nfills, + small_nflushes; + size_t large_allocated; + uint64_t large_nmalloc, large_ndalloc, large_nrequests, large_nfills, + large_nflushes; + size_t tcache_bytes, abandoned_vm; + uint64_t uptime; + + CTL_GET("arenas.page", &page, size_t); + + CTL_M2_GET("stats.arenas.0.nthreads", i, &nthreads, unsigned); + emitter_kv(emitter, "nthreads", "assigned threads", + emitter_type_unsigned, &nthreads); + + CTL_M2_GET("stats.arenas.0.uptime", i, &uptime, uint64_t); + emitter_kv(emitter, "uptime_ns", "uptime", emitter_type_uint64, + &uptime); + + CTL_M2_GET("stats.arenas.0.dss", i, &dss, const char *); + emitter_kv(emitter, "dss", "dss allocation precedence", + emitter_type_string, &dss); + + CTL_M2_GET("stats.arenas.0.dirty_decay_ms", i, &dirty_decay_ms, + ssize_t); + CTL_M2_GET("stats.arenas.0.muzzy_decay_ms", i, &muzzy_decay_ms, + ssize_t); + CTL_M2_GET("stats.arenas.0.pactive", i, &pactive, size_t); + CTL_M2_GET("stats.arenas.0.pdirty", i, &pdirty, size_t); + CTL_M2_GET("stats.arenas.0.pmuzzy", i, &pmuzzy, size_t); + CTL_M2_GET("stats.arenas.0.dirty_npurge", i, &dirty_npurge, uint64_t); + CTL_M2_GET("stats.arenas.0.dirty_nmadvise", i, &dirty_nmadvise, + uint64_t); + CTL_M2_GET("stats.arenas.0.dirty_purged", i, &dirty_purged, uint64_t); + CTL_M2_GET("stats.arenas.0.muzzy_npurge", i, &muzzy_npurge, uint64_t); + CTL_M2_GET("stats.arenas.0.muzzy_nmadvise", i, &muzzy_nmadvise, + uint64_t); + CTL_M2_GET("stats.arenas.0.muzzy_purged", i, &muzzy_purged, uint64_t); + + emitter_row_t decay_row; + emitter_row_init(&decay_row); + + /* JSON-style emission. */ + emitter_json_kv(emitter, "dirty_decay_ms", emitter_type_ssize, + &dirty_decay_ms); + emitter_json_kv(emitter, "muzzy_decay_ms", emitter_type_ssize, + &muzzy_decay_ms); + + emitter_json_kv(emitter, "pactive", emitter_type_size, &pactive); + emitter_json_kv(emitter, "pdirty", emitter_type_size, &pdirty); + emitter_json_kv(emitter, "pmuzzy", emitter_type_size, &pmuzzy); + + emitter_json_kv(emitter, "dirty_npurge", emitter_type_uint64, + &dirty_npurge); + emitter_json_kv(emitter, "dirty_nmadvise", emitter_type_uint64, + &dirty_nmadvise); + emitter_json_kv(emitter, "dirty_purged", emitter_type_uint64, + &dirty_purged); + + emitter_json_kv(emitter, "muzzy_npurge", emitter_type_uint64, + &muzzy_npurge); + emitter_json_kv(emitter, "muzzy_nmadvise", emitter_type_uint64, + &muzzy_nmadvise); + emitter_json_kv(emitter, "muzzy_purged", emitter_type_uint64, + &muzzy_purged); + + /* Table-style emission. */ + COL(decay_row, decay_type, right, 9, title); + col_decay_type.str_val = "decaying:"; + + COL(decay_row, decay_time, right, 6, title); + col_decay_time.str_val = "time"; + + COL(decay_row, decay_npages, right, 13, title); + col_decay_npages.str_val = "npages"; + + COL(decay_row, decay_sweeps, right, 13, title); + col_decay_sweeps.str_val = "sweeps"; + + COL(decay_row, decay_madvises, right, 13, title); + col_decay_madvises.str_val = "madvises"; + + COL(decay_row, decay_purged, right, 13, title); + col_decay_purged.str_val = "purged"; + + /* Title row. */ + emitter_table_row(emitter, &decay_row); + + /* Dirty row. */ + col_decay_type.str_val = "dirty:"; + + if (dirty_decay_ms >= 0) { + col_decay_time.type = emitter_type_ssize; + col_decay_time.ssize_val = dirty_decay_ms; + } else { + col_decay_time.type = emitter_type_title; + col_decay_time.str_val = "N/A"; + } + + col_decay_npages.type = emitter_type_size; + col_decay_npages.size_val = pdirty; + + col_decay_sweeps.type = emitter_type_uint64; + col_decay_sweeps.uint64_val = dirty_npurge; + + col_decay_madvises.type = emitter_type_uint64; + col_decay_madvises.uint64_val = dirty_nmadvise; + + col_decay_purged.type = emitter_type_uint64; + col_decay_purged.uint64_val = dirty_purged; + + emitter_table_row(emitter, &decay_row); + + /* Muzzy row. */ + col_decay_type.str_val = "muzzy:"; + + if (muzzy_decay_ms >= 0) { + col_decay_time.type = emitter_type_ssize; + col_decay_time.ssize_val = muzzy_decay_ms; + } else { + col_decay_time.type = emitter_type_title; + col_decay_time.str_val = "N/A"; + } + + col_decay_npages.type = emitter_type_size; + col_decay_npages.size_val = pmuzzy; + + col_decay_sweeps.type = emitter_type_uint64; + col_decay_sweeps.uint64_val = muzzy_npurge; + + col_decay_madvises.type = emitter_type_uint64; + col_decay_madvises.uint64_val = muzzy_nmadvise; + + col_decay_purged.type = emitter_type_uint64; + col_decay_purged.uint64_val = muzzy_purged; + + emitter_table_row(emitter, &decay_row); + + /* Small / large / total allocation counts. */ + emitter_row_t alloc_count_row; + emitter_row_init(&alloc_count_row); + + COL(alloc_count_row, count_title, left, 21, title); + col_count_title.str_val = ""; + + COL(alloc_count_row, count_allocated, right, 16, title); + col_count_allocated.str_val = "allocated"; + + COL(alloc_count_row, count_nmalloc, right, 16, title); + col_count_nmalloc.str_val = "nmalloc"; + COL(alloc_count_row, count_nmalloc_ps, right, 8, title); + col_count_nmalloc_ps.str_val = "(#/sec)"; + + COL(alloc_count_row, count_ndalloc, right, 16, title); + col_count_ndalloc.str_val = "ndalloc"; + COL(alloc_count_row, count_ndalloc_ps, right, 8, title); + col_count_ndalloc_ps.str_val = "(#/sec)"; + + COL(alloc_count_row, count_nrequests, right, 16, title); + col_count_nrequests.str_val = "nrequests"; + COL(alloc_count_row, count_nrequests_ps, right, 10, title); + col_count_nrequests_ps.str_val = "(#/sec)"; + + COL(alloc_count_row, count_nfills, right, 16, title); + col_count_nfills.str_val = "nfill"; + COL(alloc_count_row, count_nfills_ps, right, 10, title); + col_count_nfills_ps.str_val = "(#/sec)"; + + COL(alloc_count_row, count_nflushes, right, 16, title); + col_count_nflushes.str_val = "nflush"; + COL(alloc_count_row, count_nflushes_ps, right, 10, title); + col_count_nflushes_ps.str_val = "(#/sec)"; + + emitter_table_row(emitter, &alloc_count_row); + + col_count_nmalloc_ps.type = emitter_type_uint64; + col_count_ndalloc_ps.type = emitter_type_uint64; + col_count_nrequests_ps.type = emitter_type_uint64; + col_count_nfills_ps.type = emitter_type_uint64; + col_count_nflushes_ps.type = emitter_type_uint64; + +#define GET_AND_EMIT_ALLOC_STAT(small_or_large, name, valtype) \ + CTL_M2_GET("stats.arenas.0." #small_or_large "." #name, i, \ + &small_or_large##_##name, valtype##_t); \ + emitter_json_kv(emitter, #name, emitter_type_##valtype, \ + &small_or_large##_##name); \ + col_count_##name.type = emitter_type_##valtype; \ + col_count_##name.valtype##_val = small_or_large##_##name; + + emitter_json_object_kv_begin(emitter, "small"); + col_count_title.str_val = "small:"; + + GET_AND_EMIT_ALLOC_STAT(small, allocated, size) + GET_AND_EMIT_ALLOC_STAT(small, nmalloc, uint64) + col_count_nmalloc_ps.uint64_val = + rate_per_second(col_count_nmalloc.uint64_val, uptime); + GET_AND_EMIT_ALLOC_STAT(small, ndalloc, uint64) + col_count_ndalloc_ps.uint64_val = + rate_per_second(col_count_ndalloc.uint64_val, uptime); + GET_AND_EMIT_ALLOC_STAT(small, nrequests, uint64) + col_count_nrequests_ps.uint64_val = + rate_per_second(col_count_nrequests.uint64_val, uptime); + GET_AND_EMIT_ALLOC_STAT(small, nfills, uint64) + col_count_nfills_ps.uint64_val = + rate_per_second(col_count_nfills.uint64_val, uptime); + GET_AND_EMIT_ALLOC_STAT(small, nflushes, uint64) + col_count_nflushes_ps.uint64_val = + rate_per_second(col_count_nflushes.uint64_val, uptime); + + emitter_table_row(emitter, &alloc_count_row); + emitter_json_object_end(emitter); /* Close "small". */ + + emitter_json_object_kv_begin(emitter, "large"); + col_count_title.str_val = "large:"; + + GET_AND_EMIT_ALLOC_STAT(large, allocated, size) + GET_AND_EMIT_ALLOC_STAT(large, nmalloc, uint64) + col_count_nmalloc_ps.uint64_val = + rate_per_second(col_count_nmalloc.uint64_val, uptime); + GET_AND_EMIT_ALLOC_STAT(large, ndalloc, uint64) + col_count_ndalloc_ps.uint64_val = + rate_per_second(col_count_ndalloc.uint64_val, uptime); + GET_AND_EMIT_ALLOC_STAT(large, nrequests, uint64) + col_count_nrequests_ps.uint64_val = + rate_per_second(col_count_nrequests.uint64_val, uptime); + GET_AND_EMIT_ALLOC_STAT(large, nfills, uint64) + col_count_nfills_ps.uint64_val = + rate_per_second(col_count_nfills.uint64_val, uptime); + GET_AND_EMIT_ALLOC_STAT(large, nflushes, uint64) + col_count_nflushes_ps.uint64_val = + rate_per_second(col_count_nflushes.uint64_val, uptime); + + emitter_table_row(emitter, &alloc_count_row); + emitter_json_object_end(emitter); /* Close "large". */ + +#undef GET_AND_EMIT_ALLOC_STAT + + /* Aggregated small + large stats are emitter only in table mode. */ + col_count_title.str_val = "total:"; + col_count_allocated.size_val = small_allocated + large_allocated; + col_count_nmalloc.uint64_val = small_nmalloc + large_nmalloc; + col_count_ndalloc.uint64_val = small_ndalloc + large_ndalloc; + col_count_nrequests.uint64_val = small_nrequests + large_nrequests; + col_count_nfills.uint64_val = small_nfills + large_nfills; + col_count_nflushes.uint64_val = small_nflushes + large_nflushes; + col_count_nmalloc_ps.uint64_val = + rate_per_second(col_count_nmalloc.uint64_val, uptime); + col_count_ndalloc_ps.uint64_val = + rate_per_second(col_count_ndalloc.uint64_val, uptime); + col_count_nrequests_ps.uint64_val = + rate_per_second(col_count_nrequests.uint64_val, uptime); + col_count_nfills_ps.uint64_val = + rate_per_second(col_count_nfills.uint64_val, uptime); + col_count_nflushes_ps.uint64_val = + rate_per_second(col_count_nflushes.uint64_val, uptime); + emitter_table_row(emitter, &alloc_count_row); + + emitter_row_t mem_count_row; + emitter_row_init(&mem_count_row); + + emitter_col_t mem_count_title; + emitter_col_init(&mem_count_title, &mem_count_row); + mem_count_title.justify = emitter_justify_left; + mem_count_title.width = 21; + mem_count_title.type = emitter_type_title; + mem_count_title.str_val = ""; + + emitter_col_t mem_count_val; + emitter_col_init(&mem_count_val, &mem_count_row); + mem_count_val.justify = emitter_justify_right; + mem_count_val.width = 16; + mem_count_val.type = emitter_type_title; + mem_count_val.str_val = ""; + + emitter_table_row(emitter, &mem_count_row); + mem_count_val.type = emitter_type_size; + + /* Active count in bytes is emitted only in table mode. */ + mem_count_title.str_val = "active:"; + mem_count_val.size_val = pactive * page; + emitter_table_row(emitter, &mem_count_row); + +#define GET_AND_EMIT_MEM_STAT(stat) \ + CTL_M2_GET("stats.arenas.0."#stat, i, &stat, size_t); \ + emitter_json_kv(emitter, #stat, emitter_type_size, &stat); \ + mem_count_title.str_val = #stat":"; \ + mem_count_val.size_val = stat; \ + emitter_table_row(emitter, &mem_count_row); + + GET_AND_EMIT_MEM_STAT(mapped) + GET_AND_EMIT_MEM_STAT(retained) + GET_AND_EMIT_MEM_STAT(base) + GET_AND_EMIT_MEM_STAT(internal) + GET_AND_EMIT_MEM_STAT(metadata_thp) + GET_AND_EMIT_MEM_STAT(tcache_bytes) + GET_AND_EMIT_MEM_STAT(resident) + GET_AND_EMIT_MEM_STAT(abandoned_vm) + GET_AND_EMIT_MEM_STAT(extent_avail) +#undef GET_AND_EMIT_MEM_STAT + + if (mutex) { + stats_arena_mutexes_print(emitter, i, uptime); + } + if (bins) { + stats_arena_bins_print(emitter, mutex, i, uptime); + } + if (large) { + stats_arena_lextents_print(emitter, i, uptime); + } + if (extents) { + stats_arena_extents_print(emitter, i); + } +} + +static void +stats_general_print(emitter_t *emitter) { + const char *cpv; + bool bv, bv2; + unsigned uv; + uint32_t u32v; + uint64_t u64v; + ssize_t ssv, ssv2; + size_t sv, bsz, usz, ssz, sssz, cpsz; + + bsz = sizeof(bool); + usz = sizeof(unsigned); + ssz = sizeof(size_t); + sssz = sizeof(ssize_t); + cpsz = sizeof(const char *); + + CTL_GET("version", &cpv, const char *); + emitter_kv(emitter, "version", "Version", emitter_type_string, &cpv); + + /* config. */ + emitter_dict_begin(emitter, "config", "Build-time option settings"); +#define CONFIG_WRITE_BOOL(name) \ + do { \ + CTL_GET("config."#name, &bv, bool); \ + emitter_kv(emitter, #name, "config."#name, \ + emitter_type_bool, &bv); \ + } while (0) + + CONFIG_WRITE_BOOL(cache_oblivious); + CONFIG_WRITE_BOOL(debug); + CONFIG_WRITE_BOOL(fill); + CONFIG_WRITE_BOOL(lazy_lock); + emitter_kv(emitter, "malloc_conf", "config.malloc_conf", + emitter_type_string, &config_malloc_conf); + + CONFIG_WRITE_BOOL(opt_safety_checks); + CONFIG_WRITE_BOOL(prof); + CONFIG_WRITE_BOOL(prof_libgcc); + CONFIG_WRITE_BOOL(prof_libunwind); + CONFIG_WRITE_BOOL(stats); + CONFIG_WRITE_BOOL(utrace); + CONFIG_WRITE_BOOL(xmalloc); +#undef CONFIG_WRITE_BOOL + emitter_dict_end(emitter); /* Close "config" dict. */ + + /* opt. */ +#define OPT_WRITE(name, var, size, emitter_type) \ + if (je_mallctl("opt."name, (void *)&var, &size, NULL, 0) == \ + 0) { \ + emitter_kv(emitter, name, "opt."name, emitter_type, \ + &var); \ + } + +#define OPT_WRITE_MUTABLE(name, var1, var2, size, emitter_type, \ + altname) \ + if (je_mallctl("opt."name, (void *)&var1, &size, NULL, 0) == \ + 0 && je_mallctl(altname, (void *)&var2, &size, NULL, 0) \ + == 0) { \ + emitter_kv_note(emitter, name, "opt."name, \ + emitter_type, &var1, altname, emitter_type, \ + &var2); \ + } + +#define OPT_WRITE_BOOL(name) OPT_WRITE(name, bv, bsz, emitter_type_bool) +#define OPT_WRITE_BOOL_MUTABLE(name, altname) \ + OPT_WRITE_MUTABLE(name, bv, bv2, bsz, emitter_type_bool, altname) + +#define OPT_WRITE_UNSIGNED(name) \ + OPT_WRITE(name, uv, usz, emitter_type_unsigned) + +#define OPT_WRITE_SIZE_T(name) \ + OPT_WRITE(name, sv, ssz, emitter_type_size) +#define OPT_WRITE_SSIZE_T(name) \ + OPT_WRITE(name, ssv, sssz, emitter_type_ssize) +#define OPT_WRITE_SSIZE_T_MUTABLE(name, altname) \ + OPT_WRITE_MUTABLE(name, ssv, ssv2, sssz, emitter_type_ssize, \ + altname) + +#define OPT_WRITE_CHAR_P(name) \ + OPT_WRITE(name, cpv, cpsz, emitter_type_string) + + emitter_dict_begin(emitter, "opt", "Run-time option settings"); + + OPT_WRITE_BOOL("abort") + OPT_WRITE_BOOL("abort_conf") + OPT_WRITE_BOOL("confirm_conf") + OPT_WRITE_BOOL("retain") + OPT_WRITE_CHAR_P("dss") + OPT_WRITE_UNSIGNED("narenas") + OPT_WRITE_CHAR_P("percpu_arena") + OPT_WRITE_SIZE_T("oversize_threshold") + OPT_WRITE_CHAR_P("metadata_thp") + OPT_WRITE_BOOL_MUTABLE("background_thread", "background_thread") + OPT_WRITE_SSIZE_T_MUTABLE("dirty_decay_ms", "arenas.dirty_decay_ms") + OPT_WRITE_SSIZE_T_MUTABLE("muzzy_decay_ms", "arenas.muzzy_decay_ms") + OPT_WRITE_SIZE_T("lg_extent_max_active_fit") + OPT_WRITE_CHAR_P("junk") + OPT_WRITE_BOOL("zero") + OPT_WRITE_BOOL("utrace") + OPT_WRITE_BOOL("xmalloc") + OPT_WRITE_BOOL("tcache") + OPT_WRITE_SSIZE_T("lg_tcache_max") + OPT_WRITE_CHAR_P("thp") + OPT_WRITE_BOOL("prof") + OPT_WRITE_CHAR_P("prof_prefix") + OPT_WRITE_BOOL_MUTABLE("prof_active", "prof.active") + OPT_WRITE_BOOL_MUTABLE("prof_thread_active_init", + "prof.thread_active_init") + OPT_WRITE_SSIZE_T_MUTABLE("lg_prof_sample", "prof.lg_sample") + OPT_WRITE_BOOL("prof_accum") + OPT_WRITE_SSIZE_T("lg_prof_interval") + OPT_WRITE_BOOL("prof_gdump") + OPT_WRITE_BOOL("prof_final") + OPT_WRITE_BOOL("prof_leak") + OPT_WRITE_BOOL("stats_print") + OPT_WRITE_CHAR_P("stats_print_opts") + + emitter_dict_end(emitter); + +#undef OPT_WRITE +#undef OPT_WRITE_MUTABLE +#undef OPT_WRITE_BOOL +#undef OPT_WRITE_BOOL_MUTABLE +#undef OPT_WRITE_UNSIGNED +#undef OPT_WRITE_SSIZE_T +#undef OPT_WRITE_SSIZE_T_MUTABLE +#undef OPT_WRITE_CHAR_P + + /* prof. */ + if (config_prof) { + emitter_dict_begin(emitter, "prof", "Profiling settings"); + + CTL_GET("prof.thread_active_init", &bv, bool); + emitter_kv(emitter, "thread_active_init", + "prof.thread_active_init", emitter_type_bool, &bv); + + CTL_GET("prof.active", &bv, bool); + emitter_kv(emitter, "active", "prof.active", emitter_type_bool, + &bv); + + CTL_GET("prof.gdump", &bv, bool); + emitter_kv(emitter, "gdump", "prof.gdump", emitter_type_bool, + &bv); + + CTL_GET("prof.interval", &u64v, uint64_t); + emitter_kv(emitter, "interval", "prof.interval", + emitter_type_uint64, &u64v); + + CTL_GET("prof.lg_sample", &ssv, ssize_t); + emitter_kv(emitter, "lg_sample", "prof.lg_sample", + emitter_type_ssize, &ssv); + + emitter_dict_end(emitter); /* Close "prof". */ + } + + /* arenas. */ + /* + * The json output sticks arena info into an "arenas" dict; the table + * output puts them at the top-level. + */ + emitter_json_object_kv_begin(emitter, "arenas"); + + CTL_GET("arenas.narenas", &uv, unsigned); + emitter_kv(emitter, "narenas", "Arenas", emitter_type_unsigned, &uv); + + /* + * Decay settings are emitted only in json mode; in table mode, they're + * emitted as notes with the opt output, above. + */ + CTL_GET("arenas.dirty_decay_ms", &ssv, ssize_t); + emitter_json_kv(emitter, "dirty_decay_ms", emitter_type_ssize, &ssv); + + CTL_GET("arenas.muzzy_decay_ms", &ssv, ssize_t); + emitter_json_kv(emitter, "muzzy_decay_ms", emitter_type_ssize, &ssv); + + CTL_GET("arenas.quantum", &sv, size_t); + emitter_kv(emitter, "quantum", "Quantum size", emitter_type_size, &sv); + + CTL_GET("arenas.page", &sv, size_t); + emitter_kv(emitter, "page", "Page size", emitter_type_size, &sv); + + if (je_mallctl("arenas.tcache_max", (void *)&sv, &ssz, NULL, 0) == 0) { + emitter_kv(emitter, "tcache_max", + "Maximum thread-cached size class", emitter_type_size, &sv); + } + + unsigned nbins; + CTL_GET("arenas.nbins", &nbins, unsigned); + emitter_kv(emitter, "nbins", "Number of bin size classes", + emitter_type_unsigned, &nbins); + + unsigned nhbins; + CTL_GET("arenas.nhbins", &nhbins, unsigned); + emitter_kv(emitter, "nhbins", "Number of thread-cache bin size classes", + emitter_type_unsigned, &nhbins); + + /* + * We do enough mallctls in a loop that we actually want to omit them + * (not just omit the printing). + */ + if (emitter->output == emitter_output_json) { + emitter_json_array_kv_begin(emitter, "bin"); + for (unsigned i = 0; i < nbins; i++) { + emitter_json_object_begin(emitter); + + CTL_M2_GET("arenas.bin.0.size", i, &sv, size_t); + emitter_json_kv(emitter, "size", emitter_type_size, + &sv); + + CTL_M2_GET("arenas.bin.0.nregs", i, &u32v, uint32_t); + emitter_json_kv(emitter, "nregs", emitter_type_uint32, + &u32v); + + CTL_M2_GET("arenas.bin.0.slab_size", i, &sv, size_t); + emitter_json_kv(emitter, "slab_size", emitter_type_size, + &sv); + + CTL_M2_GET("arenas.bin.0.nshards", i, &u32v, uint32_t); + emitter_json_kv(emitter, "nshards", emitter_type_uint32, + &u32v); + + emitter_json_object_end(emitter); + } + emitter_json_array_end(emitter); /* Close "bin". */ + } + + unsigned nlextents; + CTL_GET("arenas.nlextents", &nlextents, unsigned); + emitter_kv(emitter, "nlextents", "Number of large size classes", + emitter_type_unsigned, &nlextents); + + if (emitter->output == emitter_output_json) { + emitter_json_array_kv_begin(emitter, "lextent"); + for (unsigned i = 0; i < nlextents; i++) { + emitter_json_object_begin(emitter); + + CTL_M2_GET("arenas.lextent.0.size", i, &sv, size_t); + emitter_json_kv(emitter, "size", emitter_type_size, + &sv); + + emitter_json_object_end(emitter); + } + emitter_json_array_end(emitter); /* Close "lextent". */ + } + + emitter_json_object_end(emitter); /* Close "arenas" */ +} + +static void +stats_print_helper(emitter_t *emitter, bool merged, bool destroyed, + bool unmerged, bool bins, bool large, bool mutex, bool extents) { + /* + * These should be deleted. We keep them around for a while, to aid in + * the transition to the emitter code. + */ + size_t allocated, active, metadata, metadata_thp, resident, mapped, + retained; + size_t num_background_threads; + uint64_t background_thread_num_runs, background_thread_run_interval; + + CTL_GET("stats.allocated", &allocated, size_t); + CTL_GET("stats.active", &active, size_t); + CTL_GET("stats.metadata", &metadata, size_t); + CTL_GET("stats.metadata_thp", &metadata_thp, size_t); + CTL_GET("stats.resident", &resident, size_t); + CTL_GET("stats.mapped", &mapped, size_t); + CTL_GET("stats.retained", &retained, size_t); + + if (have_background_thread) { + CTL_GET("stats.background_thread.num_threads", + &num_background_threads, size_t); + CTL_GET("stats.background_thread.num_runs", + &background_thread_num_runs, uint64_t); + CTL_GET("stats.background_thread.run_interval", + &background_thread_run_interval, uint64_t); + } else { + num_background_threads = 0; + background_thread_num_runs = 0; + background_thread_run_interval = 0; + } + + /* Generic global stats. */ + emitter_json_object_kv_begin(emitter, "stats"); + emitter_json_kv(emitter, "allocated", emitter_type_size, &allocated); + emitter_json_kv(emitter, "active", emitter_type_size, &active); + emitter_json_kv(emitter, "metadata", emitter_type_size, &metadata); + emitter_json_kv(emitter, "metadata_thp", emitter_type_size, + &metadata_thp); + emitter_json_kv(emitter, "resident", emitter_type_size, &resident); + emitter_json_kv(emitter, "mapped", emitter_type_size, &mapped); + emitter_json_kv(emitter, "retained", emitter_type_size, &retained); + + emitter_table_printf(emitter, "Allocated: %zu, active: %zu, " + "metadata: %zu (n_thp %zu), resident: %zu, mapped: %zu, " + "retained: %zu\n", allocated, active, metadata, metadata_thp, + resident, mapped, retained); + + /* Background thread stats. */ + emitter_json_object_kv_begin(emitter, "background_thread"); + emitter_json_kv(emitter, "num_threads", emitter_type_size, + &num_background_threads); + emitter_json_kv(emitter, "num_runs", emitter_type_uint64, + &background_thread_num_runs); + emitter_json_kv(emitter, "run_interval", emitter_type_uint64, + &background_thread_run_interval); + emitter_json_object_end(emitter); /* Close "background_thread". */ + + emitter_table_printf(emitter, "Background threads: %zu, " + "num_runs: %"FMTu64", run_interval: %"FMTu64" ns\n", + num_background_threads, background_thread_num_runs, + background_thread_run_interval); + + if (mutex) { + emitter_row_t row; + emitter_col_t name; + emitter_col_t col64[mutex_prof_num_uint64_t_counters]; + emitter_col_t col32[mutex_prof_num_uint32_t_counters]; + uint64_t uptime; + + emitter_row_init(&row); + mutex_stats_init_cols(&row, "", &name, col64, col32); + + emitter_table_row(emitter, &row); + emitter_json_object_kv_begin(emitter, "mutexes"); + + CTL_M2_GET("stats.arenas.0.uptime", 0, &uptime, uint64_t); + + for (int i = 0; i < mutex_prof_num_global_mutexes; i++) { + mutex_stats_read_global(global_mutex_names[i], &name, + col64, col32, uptime); + emitter_json_object_kv_begin(emitter, global_mutex_names[i]); + mutex_stats_emit(emitter, &row, col64, col32); + emitter_json_object_end(emitter); + } + + emitter_json_object_end(emitter); /* Close "mutexes". */ + } + + emitter_json_object_end(emitter); /* Close "stats". */ + + if (merged || destroyed || unmerged) { + unsigned narenas; + + emitter_json_object_kv_begin(emitter, "stats.arenas"); + + CTL_GET("arenas.narenas", &narenas, unsigned); + size_t mib[3]; + size_t miblen = sizeof(mib) / sizeof(size_t); + size_t sz; + VARIABLE_ARRAY(bool, initialized, narenas); + bool destroyed_initialized; + unsigned i, j, ninitialized; + + xmallctlnametomib("arena.0.initialized", mib, &miblen); + for (i = ninitialized = 0; i < narenas; i++) { + mib[1] = i; + sz = sizeof(bool); + xmallctlbymib(mib, miblen, &initialized[i], &sz, + NULL, 0); + if (initialized[i]) { + ninitialized++; + } + } + mib[1] = MALLCTL_ARENAS_DESTROYED; + sz = sizeof(bool); + xmallctlbymib(mib, miblen, &destroyed_initialized, &sz, + NULL, 0); + + /* Merged stats. */ + if (merged && (ninitialized > 1 || !unmerged)) { + /* Print merged arena stats. */ + emitter_table_printf(emitter, "Merged arenas stats:\n"); + emitter_json_object_kv_begin(emitter, "merged"); + stats_arena_print(emitter, MALLCTL_ARENAS_ALL, bins, + large, mutex, extents); + emitter_json_object_end(emitter); /* Close "merged". */ + } + + /* Destroyed stats. */ + if (destroyed_initialized && destroyed) { + /* Print destroyed arena stats. */ + emitter_table_printf(emitter, + "Destroyed arenas stats:\n"); + emitter_json_object_kv_begin(emitter, "destroyed"); + stats_arena_print(emitter, MALLCTL_ARENAS_DESTROYED, + bins, large, mutex, extents); + emitter_json_object_end(emitter); /* Close "destroyed". */ + } + + /* Unmerged stats. */ + if (unmerged) { + for (i = j = 0; i < narenas; i++) { + if (initialized[i]) { + char arena_ind_str[20]; + malloc_snprintf(arena_ind_str, + sizeof(arena_ind_str), "%u", i); + emitter_json_object_kv_begin(emitter, + arena_ind_str); + emitter_table_printf(emitter, + "arenas[%s]:\n", arena_ind_str); + stats_arena_print(emitter, i, bins, + large, mutex, extents); + /* Close "<arena-ind>". */ + emitter_json_object_end(emitter); + } + } + } + emitter_json_object_end(emitter); /* Close "stats.arenas". */ + } +} + +void +stats_print(void (*write_cb)(void *, const char *), void *cbopaque, + const char *opts) { + int err; + uint64_t epoch; + size_t u64sz; +#define OPTION(o, v, d, s) bool v = d; + STATS_PRINT_OPTIONS +#undef OPTION + + /* + * Refresh stats, in case mallctl() was called by the application. + * + * Check for OOM here, since refreshing the ctl cache can trigger + * allocation. In practice, none of the subsequent mallctl()-related + * calls in this function will cause OOM if this one succeeds. + * */ + epoch = 1; + u64sz = sizeof(uint64_t); + err = je_mallctl("epoch", (void *)&epoch, &u64sz, (void *)&epoch, + sizeof(uint64_t)); + if (err != 0) { + if (err == EAGAIN) { + malloc_write("<jemalloc>: Memory allocation failure in " + "mallctl(\"epoch\", ...)\n"); + return; + } + malloc_write("<jemalloc>: Failure in mallctl(\"epoch\", " + "...)\n"); + abort(); + } + + if (opts != NULL) { + for (unsigned i = 0; opts[i] != '\0'; i++) { + switch (opts[i]) { +#define OPTION(o, v, d, s) case o: v = s; break; + STATS_PRINT_OPTIONS +#undef OPTION + default:; + } + } + } + + emitter_t emitter; + emitter_init(&emitter, + json ? emitter_output_json : emitter_output_table, write_cb, + cbopaque); + emitter_begin(&emitter); + emitter_table_printf(&emitter, "___ Begin jemalloc statistics ___\n"); + emitter_json_object_kv_begin(&emitter, "jemalloc"); + + if (general) { + stats_general_print(&emitter); + } + if (config_stats) { + stats_print_helper(&emitter, merged, destroyed, unmerged, + bins, large, mutex, extents); + } + + emitter_json_object_end(&emitter); /* Closes the "jemalloc" dict. */ + emitter_table_printf(&emitter, "--- End jemalloc statistics ---\n"); + emitter_end(&emitter); +} diff --git a/deps/jemalloc/src/sz.c b/deps/jemalloc/src/sz.c new file mode 100644 index 000000000..8633fb050 --- /dev/null +++ b/deps/jemalloc/src/sz.c @@ -0,0 +1,64 @@ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/sz.h" + +JEMALLOC_ALIGNED(CACHELINE) +size_t sz_pind2sz_tab[SC_NPSIZES+1]; + +static void +sz_boot_pind2sz_tab(const sc_data_t *sc_data) { + int pind = 0; + for (unsigned i = 0; i < SC_NSIZES; i++) { + const sc_t *sc = &sc_data->sc[i]; + if (sc->psz) { + sz_pind2sz_tab[pind] = (ZU(1) << sc->lg_base) + + (ZU(sc->ndelta) << sc->lg_delta); + pind++; + } + } + for (int i = pind; i <= (int)SC_NPSIZES; i++) { + sz_pind2sz_tab[pind] = sc_data->large_maxclass + PAGE; + } +} + +JEMALLOC_ALIGNED(CACHELINE) +size_t sz_index2size_tab[SC_NSIZES]; + +static void +sz_boot_index2size_tab(const sc_data_t *sc_data) { + for (unsigned i = 0; i < SC_NSIZES; i++) { + const sc_t *sc = &sc_data->sc[i]; + sz_index2size_tab[i] = (ZU(1) << sc->lg_base) + + (ZU(sc->ndelta) << (sc->lg_delta)); + } +} + +/* + * To keep this table small, we divide sizes by the tiny min size, which gives + * the smallest interval for which the result can change. + */ +JEMALLOC_ALIGNED(CACHELINE) +uint8_t sz_size2index_tab[(SC_LOOKUP_MAXCLASS >> SC_LG_TINY_MIN) + 1]; + +static void +sz_boot_size2index_tab(const sc_data_t *sc_data) { + size_t dst_max = (SC_LOOKUP_MAXCLASS >> SC_LG_TINY_MIN) + 1; + size_t dst_ind = 0; + for (unsigned sc_ind = 0; sc_ind < SC_NSIZES && dst_ind < dst_max; + sc_ind++) { + const sc_t *sc = &sc_data->sc[sc_ind]; + size_t sz = (ZU(1) << sc->lg_base) + + (ZU(sc->ndelta) << sc->lg_delta); + size_t max_ind = ((sz + (ZU(1) << SC_LG_TINY_MIN) - 1) + >> SC_LG_TINY_MIN); + for (; dst_ind <= max_ind && dst_ind < dst_max; dst_ind++) { + sz_size2index_tab[dst_ind] = sc_ind; + } + } +} + +void +sz_boot(const sc_data_t *sc_data) { + sz_boot_pind2sz_tab(sc_data); + sz_boot_index2size_tab(sc_data); + sz_boot_size2index_tab(sc_data); +} diff --git a/deps/jemalloc/src/tcache.c b/deps/jemalloc/src/tcache.c new file mode 100644 index 000000000..50099a9f2 --- /dev/null +++ b/deps/jemalloc/src/tcache.c @@ -0,0 +1,798 @@ +#define JEMALLOC_TCACHE_C_ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/assert.h" +#include "jemalloc/internal/mutex.h" +#include "jemalloc/internal/safety_check.h" +#include "jemalloc/internal/sc.h" + +/******************************************************************************/ +/* Data. */ + +bool opt_tcache = true; +ssize_t opt_lg_tcache_max = LG_TCACHE_MAXCLASS_DEFAULT; + +cache_bin_info_t *tcache_bin_info; +static unsigned stack_nelms; /* Total stack elms per tcache. */ + +unsigned nhbins; +size_t tcache_maxclass; + +tcaches_t *tcaches; + +/* Index of first element within tcaches that has never been used. */ +static unsigned tcaches_past; + +/* Head of singly linked list tracking available tcaches elements. */ +static tcaches_t *tcaches_avail; + +/* Protects tcaches{,_past,_avail}. */ +static malloc_mutex_t tcaches_mtx; + +/******************************************************************************/ + +size_t +tcache_salloc(tsdn_t *tsdn, const void *ptr) { + return arena_salloc(tsdn, ptr); +} + +void +tcache_event_hard(tsd_t *tsd, tcache_t *tcache) { + szind_t binind = tcache->next_gc_bin; + + cache_bin_t *tbin; + if (binind < SC_NBINS) { + tbin = tcache_small_bin_get(tcache, binind); + } else { + tbin = tcache_large_bin_get(tcache, binind); + } + if (tbin->low_water > 0) { + /* + * Flush (ceiling) 3/4 of the objects below the low water mark. + */ + if (binind < SC_NBINS) { + tcache_bin_flush_small(tsd, tcache, tbin, binind, + tbin->ncached - tbin->low_water + (tbin->low_water + >> 2)); + /* + * Reduce fill count by 2X. Limit lg_fill_div such that + * the fill count is always at least 1. + */ + cache_bin_info_t *tbin_info = &tcache_bin_info[binind]; + if ((tbin_info->ncached_max >> + (tcache->lg_fill_div[binind] + 1)) >= 1) { + tcache->lg_fill_div[binind]++; + } + } else { + tcache_bin_flush_large(tsd, tbin, binind, tbin->ncached + - tbin->low_water + (tbin->low_water >> 2), tcache); + } + } else if (tbin->low_water < 0) { + /* + * Increase fill count by 2X for small bins. Make sure + * lg_fill_div stays greater than 0. + */ + if (binind < SC_NBINS && tcache->lg_fill_div[binind] > 1) { + tcache->lg_fill_div[binind]--; + } + } + tbin->low_water = tbin->ncached; + + tcache->next_gc_bin++; + if (tcache->next_gc_bin == nhbins) { + tcache->next_gc_bin = 0; + } +} + +void * +tcache_alloc_small_hard(tsdn_t *tsdn, arena_t *arena, tcache_t *tcache, + cache_bin_t *tbin, szind_t binind, bool *tcache_success) { + void *ret; + + assert(tcache->arena != NULL); + arena_tcache_fill_small(tsdn, arena, tcache, tbin, binind, + config_prof ? tcache->prof_accumbytes : 0); + if (config_prof) { + tcache->prof_accumbytes = 0; + } + ret = cache_bin_alloc_easy(tbin, tcache_success); + + return ret; +} + +/* Enabled with --enable-extra-size-check. */ +static void +tbin_extents_lookup_size_check(tsdn_t *tsdn, cache_bin_t *tbin, szind_t binind, + size_t nflush, extent_t **extents){ + rtree_ctx_t rtree_ctx_fallback; + rtree_ctx_t *rtree_ctx = tsdn_rtree_ctx(tsdn, &rtree_ctx_fallback); + + /* + * Verify that the items in the tcache all have the correct size; this + * is useful for catching sized deallocation bugs, also to fail early + * instead of corrupting metadata. Since this can be turned on for opt + * builds, avoid the branch in the loop. + */ + szind_t szind; + size_t sz_sum = binind * nflush; + for (unsigned i = 0 ; i < nflush; i++) { + rtree_extent_szind_read(tsdn, &extents_rtree, + rtree_ctx, (uintptr_t)*(tbin->avail - 1 - i), true, + &extents[i], &szind); + sz_sum -= szind; + } + if (sz_sum != 0) { + safety_check_fail("<jemalloc>: size mismatch in thread cache " + "detected, likely caused by sized deallocation bugs by " + "application. Abort.\n"); + abort(); + } +} + +void +tcache_bin_flush_small(tsd_t *tsd, tcache_t *tcache, cache_bin_t *tbin, + szind_t binind, unsigned rem) { + bool merged_stats = false; + + assert(binind < SC_NBINS); + assert((cache_bin_sz_t)rem <= tbin->ncached); + + arena_t *arena = tcache->arena; + assert(arena != NULL); + unsigned nflush = tbin->ncached - rem; + VARIABLE_ARRAY(extent_t *, item_extent, nflush); + + /* Look up extent once per item. */ + if (config_opt_safety_checks) { + tbin_extents_lookup_size_check(tsd_tsdn(tsd), tbin, binind, + nflush, item_extent); + } else { + for (unsigned i = 0 ; i < nflush; i++) { + item_extent[i] = iealloc(tsd_tsdn(tsd), + *(tbin->avail - 1 - i)); + } + } + while (nflush > 0) { + /* Lock the arena bin associated with the first object. */ + extent_t *extent = item_extent[0]; + unsigned bin_arena_ind = extent_arena_ind_get(extent); + arena_t *bin_arena = arena_get(tsd_tsdn(tsd), bin_arena_ind, + false); + unsigned binshard = extent_binshard_get(extent); + assert(binshard < bin_infos[binind].n_shards); + bin_t *bin = &bin_arena->bins[binind].bin_shards[binshard]; + + if (config_prof && bin_arena == arena) { + if (arena_prof_accum(tsd_tsdn(tsd), arena, + tcache->prof_accumbytes)) { + prof_idump(tsd_tsdn(tsd)); + } + tcache->prof_accumbytes = 0; + } + + malloc_mutex_lock(tsd_tsdn(tsd), &bin->lock); + if (config_stats && bin_arena == arena && !merged_stats) { + merged_stats = true; + bin->stats.nflushes++; + bin->stats.nrequests += tbin->tstats.nrequests; + tbin->tstats.nrequests = 0; + } + unsigned ndeferred = 0; + for (unsigned i = 0; i < nflush; i++) { + void *ptr = *(tbin->avail - 1 - i); + extent = item_extent[i]; + assert(ptr != NULL && extent != NULL); + + if (extent_arena_ind_get(extent) == bin_arena_ind + && extent_binshard_get(extent) == binshard) { + arena_dalloc_bin_junked_locked(tsd_tsdn(tsd), + bin_arena, bin, binind, extent, ptr); + } else { + /* + * This object was allocated via a different + * arena bin than the one that is currently + * locked. Stash the object, so that it can be + * handled in a future pass. + */ + *(tbin->avail - 1 - ndeferred) = ptr; + item_extent[ndeferred] = extent; + ndeferred++; + } + } + malloc_mutex_unlock(tsd_tsdn(tsd), &bin->lock); + arena_decay_ticks(tsd_tsdn(tsd), bin_arena, nflush - ndeferred); + nflush = ndeferred; + } + if (config_stats && !merged_stats) { + /* + * The flush loop didn't happen to flush to this thread's + * arena, so the stats didn't get merged. Manually do so now. + */ + unsigned binshard; + bin_t *bin = arena_bin_choose_lock(tsd_tsdn(tsd), arena, binind, + &binshard); + bin->stats.nflushes++; + bin->stats.nrequests += tbin->tstats.nrequests; + tbin->tstats.nrequests = 0; + malloc_mutex_unlock(tsd_tsdn(tsd), &bin->lock); + } + + memmove(tbin->avail - rem, tbin->avail - tbin->ncached, rem * + sizeof(void *)); + tbin->ncached = rem; + if (tbin->ncached < tbin->low_water) { + tbin->low_water = tbin->ncached; + } +} + +void +tcache_bin_flush_large(tsd_t *tsd, cache_bin_t *tbin, szind_t binind, + unsigned rem, tcache_t *tcache) { + bool merged_stats = false; + + assert(binind < nhbins); + assert((cache_bin_sz_t)rem <= tbin->ncached); + + arena_t *tcache_arena = tcache->arena; + assert(tcache_arena != NULL); + unsigned nflush = tbin->ncached - rem; + VARIABLE_ARRAY(extent_t *, item_extent, nflush); + +#ifndef JEMALLOC_EXTRA_SIZE_CHECK + /* Look up extent once per item. */ + for (unsigned i = 0 ; i < nflush; i++) { + item_extent[i] = iealloc(tsd_tsdn(tsd), *(tbin->avail - 1 - i)); + } +#else + tbin_extents_lookup_size_check(tsd_tsdn(tsd), tbin, binind, nflush, + item_extent); +#endif + while (nflush > 0) { + /* Lock the arena associated with the first object. */ + extent_t *extent = item_extent[0]; + unsigned locked_arena_ind = extent_arena_ind_get(extent); + arena_t *locked_arena = arena_get(tsd_tsdn(tsd), + locked_arena_ind, false); + bool idump; + + if (config_prof) { + idump = false; + } + + bool lock_large = !arena_is_auto(locked_arena); + if (lock_large) { + malloc_mutex_lock(tsd_tsdn(tsd), &locked_arena->large_mtx); + } + for (unsigned i = 0; i < nflush; i++) { + void *ptr = *(tbin->avail - 1 - i); + assert(ptr != NULL); + extent = item_extent[i]; + if (extent_arena_ind_get(extent) == locked_arena_ind) { + large_dalloc_prep_junked_locked(tsd_tsdn(tsd), + extent); + } + } + if ((config_prof || config_stats) && + (locked_arena == tcache_arena)) { + if (config_prof) { + idump = arena_prof_accum(tsd_tsdn(tsd), + tcache_arena, tcache->prof_accumbytes); + tcache->prof_accumbytes = 0; + } + if (config_stats) { + merged_stats = true; + arena_stats_large_flush_nrequests_add( + tsd_tsdn(tsd), &tcache_arena->stats, binind, + tbin->tstats.nrequests); + tbin->tstats.nrequests = 0; + } + } + if (lock_large) { + malloc_mutex_unlock(tsd_tsdn(tsd), &locked_arena->large_mtx); + } + + unsigned ndeferred = 0; + for (unsigned i = 0; i < nflush; i++) { + void *ptr = *(tbin->avail - 1 - i); + extent = item_extent[i]; + assert(ptr != NULL && extent != NULL); + + if (extent_arena_ind_get(extent) == locked_arena_ind) { + large_dalloc_finish(tsd_tsdn(tsd), extent); + } else { + /* + * This object was allocated via a different + * arena than the one that is currently locked. + * Stash the object, so that it can be handled + * in a future pass. + */ + *(tbin->avail - 1 - ndeferred) = ptr; + item_extent[ndeferred] = extent; + ndeferred++; + } + } + if (config_prof && idump) { + prof_idump(tsd_tsdn(tsd)); + } + arena_decay_ticks(tsd_tsdn(tsd), locked_arena, nflush - + ndeferred); + nflush = ndeferred; + } + if (config_stats && !merged_stats) { + /* + * The flush loop didn't happen to flush to this thread's + * arena, so the stats didn't get merged. Manually do so now. + */ + arena_stats_large_flush_nrequests_add(tsd_tsdn(tsd), + &tcache_arena->stats, binind, tbin->tstats.nrequests); + tbin->tstats.nrequests = 0; + } + + memmove(tbin->avail - rem, tbin->avail - tbin->ncached, rem * + sizeof(void *)); + tbin->ncached = rem; + if (tbin->ncached < tbin->low_water) { + tbin->low_water = tbin->ncached; + } +} + +void +tcache_arena_associate(tsdn_t *tsdn, tcache_t *tcache, arena_t *arena) { + assert(tcache->arena == NULL); + tcache->arena = arena; + + if (config_stats) { + /* Link into list of extant tcaches. */ + malloc_mutex_lock(tsdn, &arena->tcache_ql_mtx); + + ql_elm_new(tcache, link); + ql_tail_insert(&arena->tcache_ql, tcache, link); + cache_bin_array_descriptor_init( + &tcache->cache_bin_array_descriptor, tcache->bins_small, + tcache->bins_large); + ql_tail_insert(&arena->cache_bin_array_descriptor_ql, + &tcache->cache_bin_array_descriptor, link); + + malloc_mutex_unlock(tsdn, &arena->tcache_ql_mtx); + } +} + +static void +tcache_arena_dissociate(tsdn_t *tsdn, tcache_t *tcache) { + arena_t *arena = tcache->arena; + assert(arena != NULL); + if (config_stats) { + /* Unlink from list of extant tcaches. */ + malloc_mutex_lock(tsdn, &arena->tcache_ql_mtx); + if (config_debug) { + bool in_ql = false; + tcache_t *iter; + ql_foreach(iter, &arena->tcache_ql, link) { + if (iter == tcache) { + in_ql = true; + break; + } + } + assert(in_ql); + } + ql_remove(&arena->tcache_ql, tcache, link); + ql_remove(&arena->cache_bin_array_descriptor_ql, + &tcache->cache_bin_array_descriptor, link); + tcache_stats_merge(tsdn, tcache, arena); + malloc_mutex_unlock(tsdn, &arena->tcache_ql_mtx); + } + tcache->arena = NULL; +} + +void +tcache_arena_reassociate(tsdn_t *tsdn, tcache_t *tcache, arena_t *arena) { + tcache_arena_dissociate(tsdn, tcache); + tcache_arena_associate(tsdn, tcache, arena); +} + +bool +tsd_tcache_enabled_data_init(tsd_t *tsd) { + /* Called upon tsd initialization. */ + tsd_tcache_enabled_set(tsd, opt_tcache); + tsd_slow_update(tsd); + + if (opt_tcache) { + /* Trigger tcache init. */ + tsd_tcache_data_init(tsd); + } + + return false; +} + +/* Initialize auto tcache (embedded in TSD). */ +static void +tcache_init(tsd_t *tsd, tcache_t *tcache, void *avail_stack) { + memset(&tcache->link, 0, sizeof(ql_elm(tcache_t))); + tcache->prof_accumbytes = 0; + tcache->next_gc_bin = 0; + tcache->arena = NULL; + + ticker_init(&tcache->gc_ticker, TCACHE_GC_INCR); + + size_t stack_offset = 0; + assert((TCACHE_NSLOTS_SMALL_MAX & 1U) == 0); + memset(tcache->bins_small, 0, sizeof(cache_bin_t) * SC_NBINS); + memset(tcache->bins_large, 0, sizeof(cache_bin_t) * (nhbins - SC_NBINS)); + unsigned i = 0; + for (; i < SC_NBINS; i++) { + tcache->lg_fill_div[i] = 1; + stack_offset += tcache_bin_info[i].ncached_max * sizeof(void *); + /* + * avail points past the available space. Allocations will + * access the slots toward higher addresses (for the benefit of + * prefetch). + */ + tcache_small_bin_get(tcache, i)->avail = + (void **)((uintptr_t)avail_stack + (uintptr_t)stack_offset); + } + for (; i < nhbins; i++) { + stack_offset += tcache_bin_info[i].ncached_max * sizeof(void *); + tcache_large_bin_get(tcache, i)->avail = + (void **)((uintptr_t)avail_stack + (uintptr_t)stack_offset); + } + assert(stack_offset == stack_nelms * sizeof(void *)); +} + +/* Initialize auto tcache (embedded in TSD). */ +bool +tsd_tcache_data_init(tsd_t *tsd) { + tcache_t *tcache = tsd_tcachep_get_unsafe(tsd); + assert(tcache_small_bin_get(tcache, 0)->avail == NULL); + size_t size = stack_nelms * sizeof(void *); + /* Avoid false cacheline sharing. */ + size = sz_sa2u(size, CACHELINE); + + void *avail_array = ipallocztm(tsd_tsdn(tsd), size, CACHELINE, true, + NULL, true, arena_get(TSDN_NULL, 0, true)); + if (avail_array == NULL) { + return true; + } + + tcache_init(tsd, tcache, avail_array); + /* + * Initialization is a bit tricky here. After malloc init is done, all + * threads can rely on arena_choose and associate tcache accordingly. + * However, the thread that does actual malloc bootstrapping relies on + * functional tsd, and it can only rely on a0. In that case, we + * associate its tcache to a0 temporarily, and later on + * arena_choose_hard() will re-associate properly. + */ + tcache->arena = NULL; + arena_t *arena; + if (!malloc_initialized()) { + /* If in initialization, assign to a0. */ + arena = arena_get(tsd_tsdn(tsd), 0, false); + tcache_arena_associate(tsd_tsdn(tsd), tcache, arena); + } else { + arena = arena_choose(tsd, NULL); + /* This may happen if thread.tcache.enabled is used. */ + if (tcache->arena == NULL) { + tcache_arena_associate(tsd_tsdn(tsd), tcache, arena); + } + } + assert(arena == tcache->arena); + + return false; +} + +/* Created manual tcache for tcache.create mallctl. */ +tcache_t * +tcache_create_explicit(tsd_t *tsd) { + tcache_t *tcache; + size_t size, stack_offset; + + size = sizeof(tcache_t); + /* Naturally align the pointer stacks. */ + size = PTR_CEILING(size); + stack_offset = size; + size += stack_nelms * sizeof(void *); + /* Avoid false cacheline sharing. */ + size = sz_sa2u(size, CACHELINE); + + tcache = ipallocztm(tsd_tsdn(tsd), size, CACHELINE, true, NULL, true, + arena_get(TSDN_NULL, 0, true)); + if (tcache == NULL) { + return NULL; + } + + tcache_init(tsd, tcache, + (void *)((uintptr_t)tcache + (uintptr_t)stack_offset)); + tcache_arena_associate(tsd_tsdn(tsd), tcache, arena_ichoose(tsd, NULL)); + + return tcache; +} + +static void +tcache_flush_cache(tsd_t *tsd, tcache_t *tcache) { + assert(tcache->arena != NULL); + + for (unsigned i = 0; i < SC_NBINS; i++) { + cache_bin_t *tbin = tcache_small_bin_get(tcache, i); + tcache_bin_flush_small(tsd, tcache, tbin, i, 0); + + if (config_stats) { + assert(tbin->tstats.nrequests == 0); + } + } + for (unsigned i = SC_NBINS; i < nhbins; i++) { + cache_bin_t *tbin = tcache_large_bin_get(tcache, i); + tcache_bin_flush_large(tsd, tbin, i, 0, tcache); + + if (config_stats) { + assert(tbin->tstats.nrequests == 0); + } + } + + if (config_prof && tcache->prof_accumbytes > 0 && + arena_prof_accum(tsd_tsdn(tsd), tcache->arena, + tcache->prof_accumbytes)) { + prof_idump(tsd_tsdn(tsd)); + } +} + +void +tcache_flush(tsd_t *tsd) { + assert(tcache_available(tsd)); + tcache_flush_cache(tsd, tsd_tcachep_get(tsd)); +} + +static void +tcache_destroy(tsd_t *tsd, tcache_t *tcache, bool tsd_tcache) { + tcache_flush_cache(tsd, tcache); + arena_t *arena = tcache->arena; + tcache_arena_dissociate(tsd_tsdn(tsd), tcache); + + if (tsd_tcache) { + /* Release the avail array for the TSD embedded auto tcache. */ + void *avail_array = + (void *)((uintptr_t)tcache_small_bin_get(tcache, 0)->avail - + (uintptr_t)tcache_bin_info[0].ncached_max * sizeof(void *)); + idalloctm(tsd_tsdn(tsd), avail_array, NULL, NULL, true, true); + } else { + /* Release both the tcache struct and avail array. */ + idalloctm(tsd_tsdn(tsd), tcache, NULL, NULL, true, true); + } + + /* + * The deallocation and tcache flush above may not trigger decay since + * we are on the tcache shutdown path (potentially with non-nominal + * tsd). Manually trigger decay to avoid pathological cases. Also + * include arena 0 because the tcache array is allocated from it. + */ + arena_decay(tsd_tsdn(tsd), arena_get(tsd_tsdn(tsd), 0, false), + false, false); + + if (arena_nthreads_get(arena, false) == 0 && + !background_thread_enabled()) { + /* Force purging when no threads assigned to the arena anymore. */ + arena_decay(tsd_tsdn(tsd), arena, false, true); + } else { + arena_decay(tsd_tsdn(tsd), arena, false, false); + } +} + +/* For auto tcache (embedded in TSD) only. */ +void +tcache_cleanup(tsd_t *tsd) { + tcache_t *tcache = tsd_tcachep_get(tsd); + if (!tcache_available(tsd)) { + assert(tsd_tcache_enabled_get(tsd) == false); + if (config_debug) { + assert(tcache_small_bin_get(tcache, 0)->avail == NULL); + } + return; + } + assert(tsd_tcache_enabled_get(tsd)); + assert(tcache_small_bin_get(tcache, 0)->avail != NULL); + + tcache_destroy(tsd, tcache, true); + if (config_debug) { + tcache_small_bin_get(tcache, 0)->avail = NULL; + } +} + +void +tcache_stats_merge(tsdn_t *tsdn, tcache_t *tcache, arena_t *arena) { + unsigned i; + + cassert(config_stats); + + /* Merge and reset tcache stats. */ + for (i = 0; i < SC_NBINS; i++) { + cache_bin_t *tbin = tcache_small_bin_get(tcache, i); + unsigned binshard; + bin_t *bin = arena_bin_choose_lock(tsdn, arena, i, &binshard); + bin->stats.nrequests += tbin->tstats.nrequests; + malloc_mutex_unlock(tsdn, &bin->lock); + tbin->tstats.nrequests = 0; + } + + for (; i < nhbins; i++) { + cache_bin_t *tbin = tcache_large_bin_get(tcache, i); + arena_stats_large_flush_nrequests_add(tsdn, &arena->stats, i, + tbin->tstats.nrequests); + tbin->tstats.nrequests = 0; + } +} + +static bool +tcaches_create_prep(tsd_t *tsd) { + bool err; + + malloc_mutex_lock(tsd_tsdn(tsd), &tcaches_mtx); + + if (tcaches == NULL) { + tcaches = base_alloc(tsd_tsdn(tsd), b0get(), sizeof(tcache_t *) + * (MALLOCX_TCACHE_MAX+1), CACHELINE); + if (tcaches == NULL) { + err = true; + goto label_return; + } + } + + if (tcaches_avail == NULL && tcaches_past > MALLOCX_TCACHE_MAX) { + err = true; + goto label_return; + } + + err = false; +label_return: + malloc_mutex_unlock(tsd_tsdn(tsd), &tcaches_mtx); + return err; +} + +bool +tcaches_create(tsd_t *tsd, unsigned *r_ind) { + witness_assert_depth(tsdn_witness_tsdp_get(tsd_tsdn(tsd)), 0); + + bool err; + + if (tcaches_create_prep(tsd)) { + err = true; + goto label_return; + } + + tcache_t *tcache = tcache_create_explicit(tsd); + if (tcache == NULL) { + err = true; + goto label_return; + } + + tcaches_t *elm; + malloc_mutex_lock(tsd_tsdn(tsd), &tcaches_mtx); + if (tcaches_avail != NULL) { + elm = tcaches_avail; + tcaches_avail = tcaches_avail->next; + elm->tcache = tcache; + *r_ind = (unsigned)(elm - tcaches); + } else { + elm = &tcaches[tcaches_past]; + elm->tcache = tcache; + *r_ind = tcaches_past; + tcaches_past++; + } + malloc_mutex_unlock(tsd_tsdn(tsd), &tcaches_mtx); + + err = false; +label_return: + witness_assert_depth(tsdn_witness_tsdp_get(tsd_tsdn(tsd)), 0); + return err; +} + +static tcache_t * +tcaches_elm_remove(tsd_t *tsd, tcaches_t *elm, bool allow_reinit) { + malloc_mutex_assert_owner(tsd_tsdn(tsd), &tcaches_mtx); + + if (elm->tcache == NULL) { + return NULL; + } + tcache_t *tcache = elm->tcache; + if (allow_reinit) { + elm->tcache = TCACHES_ELM_NEED_REINIT; + } else { + elm->tcache = NULL; + } + + if (tcache == TCACHES_ELM_NEED_REINIT) { + return NULL; + } + return tcache; +} + +void +tcaches_flush(tsd_t *tsd, unsigned ind) { + malloc_mutex_lock(tsd_tsdn(tsd), &tcaches_mtx); + tcache_t *tcache = tcaches_elm_remove(tsd, &tcaches[ind], true); + malloc_mutex_unlock(tsd_tsdn(tsd), &tcaches_mtx); + if (tcache != NULL) { + /* Destroy the tcache; recreate in tcaches_get() if needed. */ + tcache_destroy(tsd, tcache, false); + } +} + +void +tcaches_destroy(tsd_t *tsd, unsigned ind) { + malloc_mutex_lock(tsd_tsdn(tsd), &tcaches_mtx); + tcaches_t *elm = &tcaches[ind]; + tcache_t *tcache = tcaches_elm_remove(tsd, elm, false); + elm->next = tcaches_avail; + tcaches_avail = elm; + malloc_mutex_unlock(tsd_tsdn(tsd), &tcaches_mtx); + if (tcache != NULL) { + tcache_destroy(tsd, tcache, false); + } +} + +bool +tcache_boot(tsdn_t *tsdn) { + /* If necessary, clamp opt_lg_tcache_max. */ + if (opt_lg_tcache_max < 0 || (ZU(1) << opt_lg_tcache_max) < + SC_SMALL_MAXCLASS) { + tcache_maxclass = SC_SMALL_MAXCLASS; + } else { + tcache_maxclass = (ZU(1) << opt_lg_tcache_max); + } + + if (malloc_mutex_init(&tcaches_mtx, "tcaches", WITNESS_RANK_TCACHES, + malloc_mutex_rank_exclusive)) { + return true; + } + + nhbins = sz_size2index(tcache_maxclass) + 1; + + /* Initialize tcache_bin_info. */ + tcache_bin_info = (cache_bin_info_t *)base_alloc(tsdn, b0get(), nhbins + * sizeof(cache_bin_info_t), CACHELINE); + if (tcache_bin_info == NULL) { + return true; + } + stack_nelms = 0; + unsigned i; + for (i = 0; i < SC_NBINS; i++) { + if ((bin_infos[i].nregs << 1) <= TCACHE_NSLOTS_SMALL_MIN) { + tcache_bin_info[i].ncached_max = + TCACHE_NSLOTS_SMALL_MIN; + } else if ((bin_infos[i].nregs << 1) <= + TCACHE_NSLOTS_SMALL_MAX) { + tcache_bin_info[i].ncached_max = + (bin_infos[i].nregs << 1); + } else { + tcache_bin_info[i].ncached_max = + TCACHE_NSLOTS_SMALL_MAX; + } + stack_nelms += tcache_bin_info[i].ncached_max; + } + for (; i < nhbins; i++) { + tcache_bin_info[i].ncached_max = TCACHE_NSLOTS_LARGE; + stack_nelms += tcache_bin_info[i].ncached_max; + } + + return false; +} + +void +tcache_prefork(tsdn_t *tsdn) { + if (!config_prof && opt_tcache) { + malloc_mutex_prefork(tsdn, &tcaches_mtx); + } +} + +void +tcache_postfork_parent(tsdn_t *tsdn) { + if (!config_prof && opt_tcache) { + malloc_mutex_postfork_parent(tsdn, &tcaches_mtx); + } +} + +void +tcache_postfork_child(tsdn_t *tsdn) { + if (!config_prof && opt_tcache) { + malloc_mutex_postfork_child(tsdn, &tcaches_mtx); + } +} diff --git a/deps/jemalloc/src/test_hooks.c b/deps/jemalloc/src/test_hooks.c new file mode 100644 index 000000000..ace00d9c4 --- /dev/null +++ b/deps/jemalloc/src/test_hooks.c @@ -0,0 +1,12 @@ +#include "jemalloc/internal/jemalloc_preamble.h" + +/* + * The hooks are a little bit screwy -- they're not genuinely exported in the + * sense that we want them available to end-users, but we do want them visible + * from outside the generated library, so that we can use them in test code. + */ +JEMALLOC_EXPORT +void (*test_hooks_arena_new_hook)() = NULL; + +JEMALLOC_EXPORT +void (*test_hooks_libc_hook)() = NULL; diff --git a/deps/jemalloc/src/ticker.c b/deps/jemalloc/src/ticker.c new file mode 100644 index 000000000..d7b8cd26c --- /dev/null +++ b/deps/jemalloc/src/ticker.c @@ -0,0 +1,3 @@ +#define JEMALLOC_TICKER_C_ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" diff --git a/deps/jemalloc/src/tsd.c b/deps/jemalloc/src/tsd.c new file mode 100644 index 000000000..a31f6b969 --- /dev/null +++ b/deps/jemalloc/src/tsd.c @@ -0,0 +1,534 @@ +#define JEMALLOC_TSD_C_ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/assert.h" +#include "jemalloc/internal/mutex.h" +#include "jemalloc/internal/rtree.h" + +/******************************************************************************/ +/* Data. */ + +static unsigned ncleanups; +static malloc_tsd_cleanup_t cleanups[MALLOC_TSD_CLEANUPS_MAX]; + +/* TSD_INITIALIZER triggers "-Wmissing-field-initializer" */ +JEMALLOC_DIAGNOSTIC_PUSH +JEMALLOC_DIAGNOSTIC_IGNORE_MISSING_STRUCT_FIELD_INITIALIZERS + +#ifdef JEMALLOC_MALLOC_THREAD_CLEANUP +JEMALLOC_TSD_TYPE_ATTR(tsd_t) tsd_tls = TSD_INITIALIZER; +JEMALLOC_TSD_TYPE_ATTR(bool) JEMALLOC_TLS_MODEL tsd_initialized = false; +bool tsd_booted = false; +#elif (defined(JEMALLOC_TLS)) +JEMALLOC_TSD_TYPE_ATTR(tsd_t) tsd_tls = TSD_INITIALIZER; +pthread_key_t tsd_tsd; +bool tsd_booted = false; +#elif (defined(_WIN32)) +DWORD tsd_tsd; +tsd_wrapper_t tsd_boot_wrapper = {false, TSD_INITIALIZER}; +bool tsd_booted = false; +#else + +/* + * This contains a mutex, but it's pretty convenient to allow the mutex code to + * have a dependency on tsd. So we define the struct here, and only refer to it + * by pointer in the header. + */ +struct tsd_init_head_s { + ql_head(tsd_init_block_t) blocks; + malloc_mutex_t lock; +}; + +pthread_key_t tsd_tsd; +tsd_init_head_t tsd_init_head = { + ql_head_initializer(blocks), + MALLOC_MUTEX_INITIALIZER +}; + +tsd_wrapper_t tsd_boot_wrapper = { + false, + TSD_INITIALIZER +}; +bool tsd_booted = false; +#endif + +JEMALLOC_DIAGNOSTIC_POP + +/******************************************************************************/ + +/* A list of all the tsds in the nominal state. */ +typedef ql_head(tsd_t) tsd_list_t; +static tsd_list_t tsd_nominal_tsds = ql_head_initializer(tsd_nominal_tsds); +static malloc_mutex_t tsd_nominal_tsds_lock; + +/* How many slow-path-enabling features are turned on. */ +static atomic_u32_t tsd_global_slow_count = ATOMIC_INIT(0); + +static bool +tsd_in_nominal_list(tsd_t *tsd) { + tsd_t *tsd_list; + bool found = false; + /* + * We don't know that tsd is nominal; it might not be safe to get data + * out of it here. + */ + malloc_mutex_lock(TSDN_NULL, &tsd_nominal_tsds_lock); + ql_foreach(tsd_list, &tsd_nominal_tsds, TSD_MANGLE(tcache).tsd_link) { + if (tsd == tsd_list) { + found = true; + break; + } + } + malloc_mutex_unlock(TSDN_NULL, &tsd_nominal_tsds_lock); + return found; +} + +static void +tsd_add_nominal(tsd_t *tsd) { + assert(!tsd_in_nominal_list(tsd)); + assert(tsd_state_get(tsd) <= tsd_state_nominal_max); + ql_elm_new(tsd, TSD_MANGLE(tcache).tsd_link); + malloc_mutex_lock(tsd_tsdn(tsd), &tsd_nominal_tsds_lock); + ql_tail_insert(&tsd_nominal_tsds, tsd, TSD_MANGLE(tcache).tsd_link); + malloc_mutex_unlock(tsd_tsdn(tsd), &tsd_nominal_tsds_lock); +} + +static void +tsd_remove_nominal(tsd_t *tsd) { + assert(tsd_in_nominal_list(tsd)); + assert(tsd_state_get(tsd) <= tsd_state_nominal_max); + malloc_mutex_lock(tsd_tsdn(tsd), &tsd_nominal_tsds_lock); + ql_remove(&tsd_nominal_tsds, tsd, TSD_MANGLE(tcache).tsd_link); + malloc_mutex_unlock(tsd_tsdn(tsd), &tsd_nominal_tsds_lock); +} + +static void +tsd_force_recompute(tsdn_t *tsdn) { + /* + * The stores to tsd->state here need to synchronize with the exchange + * in tsd_slow_update. + */ + atomic_fence(ATOMIC_RELEASE); + malloc_mutex_lock(tsdn, &tsd_nominal_tsds_lock); + tsd_t *remote_tsd; + ql_foreach(remote_tsd, &tsd_nominal_tsds, TSD_MANGLE(tcache).tsd_link) { + assert(tsd_atomic_load(&remote_tsd->state, ATOMIC_RELAXED) + <= tsd_state_nominal_max); + tsd_atomic_store(&remote_tsd->state, tsd_state_nominal_recompute, + ATOMIC_RELAXED); + } + malloc_mutex_unlock(tsdn, &tsd_nominal_tsds_lock); +} + +void +tsd_global_slow_inc(tsdn_t *tsdn) { + atomic_fetch_add_u32(&tsd_global_slow_count, 1, ATOMIC_RELAXED); + /* + * We unconditionally force a recompute, even if the global slow count + * was already positive. If we didn't, then it would be possible for us + * to return to the user, have the user synchronize externally with some + * other thread, and then have that other thread not have picked up the + * update yet (since the original incrementing thread might still be + * making its way through the tsd list). + */ + tsd_force_recompute(tsdn); +} + +void tsd_global_slow_dec(tsdn_t *tsdn) { + atomic_fetch_sub_u32(&tsd_global_slow_count, 1, ATOMIC_RELAXED); + /* See the note in ..._inc(). */ + tsd_force_recompute(tsdn); +} + +static bool +tsd_local_slow(tsd_t *tsd) { + return !tsd_tcache_enabled_get(tsd) + || tsd_reentrancy_level_get(tsd) > 0; +} + +bool +tsd_global_slow() { + return atomic_load_u32(&tsd_global_slow_count, ATOMIC_RELAXED) > 0; +} + +/******************************************************************************/ + +static uint8_t +tsd_state_compute(tsd_t *tsd) { + if (!tsd_nominal(tsd)) { + return tsd_state_get(tsd); + } + /* We're in *a* nominal state; but which one? */ + if (malloc_slow || tsd_local_slow(tsd) || tsd_global_slow()) { + return tsd_state_nominal_slow; + } else { + return tsd_state_nominal; + } +} + +void +tsd_slow_update(tsd_t *tsd) { + uint8_t old_state; + do { + uint8_t new_state = tsd_state_compute(tsd); + old_state = tsd_atomic_exchange(&tsd->state, new_state, + ATOMIC_ACQUIRE); + } while (old_state == tsd_state_nominal_recompute); +} + +void +tsd_state_set(tsd_t *tsd, uint8_t new_state) { + /* Only the tsd module can change the state *to* recompute. */ + assert(new_state != tsd_state_nominal_recompute); + uint8_t old_state = tsd_atomic_load(&tsd->state, ATOMIC_RELAXED); + if (old_state > tsd_state_nominal_max) { + /* + * Not currently in the nominal list, but it might need to be + * inserted there. + */ + assert(!tsd_in_nominal_list(tsd)); + tsd_atomic_store(&tsd->state, new_state, ATOMIC_RELAXED); + if (new_state <= tsd_state_nominal_max) { + tsd_add_nominal(tsd); + } + } else { + /* + * We're currently nominal. If the new state is non-nominal, + * great; we take ourselves off the list and just enter the new + * state. + */ + assert(tsd_in_nominal_list(tsd)); + if (new_state > tsd_state_nominal_max) { + tsd_remove_nominal(tsd); + tsd_atomic_store(&tsd->state, new_state, + ATOMIC_RELAXED); + } else { + /* + * This is the tricky case. We're transitioning from + * one nominal state to another. The caller can't know + * about any races that are occuring at the same time, + * so we always have to recompute no matter what. + */ + tsd_slow_update(tsd); + } + } +} + +static bool +tsd_data_init(tsd_t *tsd) { + /* + * We initialize the rtree context first (before the tcache), since the + * tcache initialization depends on it. + */ + rtree_ctx_data_init(tsd_rtree_ctxp_get_unsafe(tsd)); + + /* + * A nondeterministic seed based on the address of tsd reduces + * the likelihood of lockstep non-uniform cache index + * utilization among identical concurrent processes, but at the + * cost of test repeatability. For debug builds, instead use a + * deterministic seed. + */ + *tsd_offset_statep_get(tsd) = config_debug ? 0 : + (uint64_t)(uintptr_t)tsd; + + return tsd_tcache_enabled_data_init(tsd); +} + +static void +assert_tsd_data_cleanup_done(tsd_t *tsd) { + assert(!tsd_nominal(tsd)); + assert(!tsd_in_nominal_list(tsd)); + assert(*tsd_arenap_get_unsafe(tsd) == NULL); + assert(*tsd_iarenap_get_unsafe(tsd) == NULL); + assert(*tsd_arenas_tdata_bypassp_get_unsafe(tsd) == true); + assert(*tsd_arenas_tdatap_get_unsafe(tsd) == NULL); + assert(*tsd_tcache_enabledp_get_unsafe(tsd) == false); + assert(*tsd_prof_tdatap_get_unsafe(tsd) == NULL); +} + +static bool +tsd_data_init_nocleanup(tsd_t *tsd) { + assert(tsd_state_get(tsd) == tsd_state_reincarnated || + tsd_state_get(tsd) == tsd_state_minimal_initialized); + /* + * During reincarnation, there is no guarantee that the cleanup function + * will be called (deallocation may happen after all tsd destructors). + * We set up tsd in a way that no cleanup is needed. + */ + rtree_ctx_data_init(tsd_rtree_ctxp_get_unsafe(tsd)); + *tsd_arenas_tdata_bypassp_get(tsd) = true; + *tsd_tcache_enabledp_get_unsafe(tsd) = false; + *tsd_reentrancy_levelp_get(tsd) = 1; + assert_tsd_data_cleanup_done(tsd); + + return false; +} + +tsd_t * +tsd_fetch_slow(tsd_t *tsd, bool minimal) { + assert(!tsd_fast(tsd)); + + if (tsd_state_get(tsd) == tsd_state_nominal_slow) { + /* + * On slow path but no work needed. Note that we can't + * necessarily *assert* that we're slow, because we might be + * slow because of an asynchronous modification to global state, + * which might be asynchronously modified *back*. + */ + } else if (tsd_state_get(tsd) == tsd_state_nominal_recompute) { + tsd_slow_update(tsd); + } else if (tsd_state_get(tsd) == tsd_state_uninitialized) { + if (!minimal) { + if (tsd_booted) { + tsd_state_set(tsd, tsd_state_nominal); + tsd_slow_update(tsd); + /* Trigger cleanup handler registration. */ + tsd_set(tsd); + tsd_data_init(tsd); + } + } else { + tsd_state_set(tsd, tsd_state_minimal_initialized); + tsd_set(tsd); + tsd_data_init_nocleanup(tsd); + } + } else if (tsd_state_get(tsd) == tsd_state_minimal_initialized) { + if (!minimal) { + /* Switch to fully initialized. */ + tsd_state_set(tsd, tsd_state_nominal); + assert(*tsd_reentrancy_levelp_get(tsd) >= 1); + (*tsd_reentrancy_levelp_get(tsd))--; + tsd_slow_update(tsd); + tsd_data_init(tsd); + } else { + assert_tsd_data_cleanup_done(tsd); + } + } else if (tsd_state_get(tsd) == tsd_state_purgatory) { + tsd_state_set(tsd, tsd_state_reincarnated); + tsd_set(tsd); + tsd_data_init_nocleanup(tsd); + } else { + assert(tsd_state_get(tsd) == tsd_state_reincarnated); + } + + return tsd; +} + +void * +malloc_tsd_malloc(size_t size) { + return a0malloc(CACHELINE_CEILING(size)); +} + +void +malloc_tsd_dalloc(void *wrapper) { + a0dalloc(wrapper); +} + +#if defined(JEMALLOC_MALLOC_THREAD_CLEANUP) || defined(_WIN32) +#ifndef _WIN32 +JEMALLOC_EXPORT +#endif +void +_malloc_thread_cleanup(void) { + bool pending[MALLOC_TSD_CLEANUPS_MAX], again; + unsigned i; + + for (i = 0; i < ncleanups; i++) { + pending[i] = true; + } + + do { + again = false; + for (i = 0; i < ncleanups; i++) { + if (pending[i]) { + pending[i] = cleanups[i](); + if (pending[i]) { + again = true; + } + } + } + } while (again); +} +#endif + +void +malloc_tsd_cleanup_register(bool (*f)(void)) { + assert(ncleanups < MALLOC_TSD_CLEANUPS_MAX); + cleanups[ncleanups] = f; + ncleanups++; +} + +static void +tsd_do_data_cleanup(tsd_t *tsd) { + prof_tdata_cleanup(tsd); + iarena_cleanup(tsd); + arena_cleanup(tsd); + arenas_tdata_cleanup(tsd); + tcache_cleanup(tsd); + witnesses_cleanup(tsd_witness_tsdp_get_unsafe(tsd)); +} + +void +tsd_cleanup(void *arg) { + tsd_t *tsd = (tsd_t *)arg; + + switch (tsd_state_get(tsd)) { + case tsd_state_uninitialized: + /* Do nothing. */ + break; + case tsd_state_minimal_initialized: + /* This implies the thread only did free() in its life time. */ + /* Fall through. */ + case tsd_state_reincarnated: + /* + * Reincarnated means another destructor deallocated memory + * after the destructor was called. Cleanup isn't required but + * is still called for testing and completeness. + */ + assert_tsd_data_cleanup_done(tsd); + /* Fall through. */ + case tsd_state_nominal: + case tsd_state_nominal_slow: + tsd_do_data_cleanup(tsd); + tsd_state_set(tsd, tsd_state_purgatory); + tsd_set(tsd); + break; + case tsd_state_purgatory: + /* + * The previous time this destructor was called, we set the + * state to tsd_state_purgatory so that other destructors + * wouldn't cause re-creation of the tsd. This time, do + * nothing, and do not request another callback. + */ + break; + default: + not_reached(); + } +#ifdef JEMALLOC_JET + test_callback_t test_callback = *tsd_test_callbackp_get_unsafe(tsd); + int *data = tsd_test_datap_get_unsafe(tsd); + if (test_callback != NULL) { + test_callback(data); + } +#endif +} + +tsd_t * +malloc_tsd_boot0(void) { + tsd_t *tsd; + + ncleanups = 0; + if (malloc_mutex_init(&tsd_nominal_tsds_lock, "tsd_nominal_tsds_lock", + WITNESS_RANK_OMIT, malloc_mutex_rank_exclusive)) { + return NULL; + } + if (tsd_boot0()) { + return NULL; + } + tsd = tsd_fetch(); + *tsd_arenas_tdata_bypassp_get(tsd) = true; + return tsd; +} + +void +malloc_tsd_boot1(void) { + tsd_boot1(); + tsd_t *tsd = tsd_fetch(); + /* malloc_slow has been set properly. Update tsd_slow. */ + tsd_slow_update(tsd); + *tsd_arenas_tdata_bypassp_get(tsd) = false; +} + +#ifdef _WIN32 +static BOOL WINAPI +_tls_callback(HINSTANCE hinstDLL, DWORD fdwReason, LPVOID lpvReserved) { + switch (fdwReason) { +#ifdef JEMALLOC_LAZY_LOCK + case DLL_THREAD_ATTACH: + isthreaded = true; + break; +#endif + case DLL_THREAD_DETACH: + _malloc_thread_cleanup(); + break; + default: + break; + } + return true; +} + +/* + * We need to be able to say "read" here (in the "pragma section"), but have + * hooked "read". We won't read for the rest of the file, so we can get away + * with unhooking. + */ +#ifdef read +# undef read +#endif + +#ifdef _MSC_VER +# ifdef _M_IX86 +# pragma comment(linker, "/INCLUDE:__tls_used") +# pragma comment(linker, "/INCLUDE:_tls_callback") +# else +# pragma comment(linker, "/INCLUDE:_tls_used") +# pragma comment(linker, "/INCLUDE:" STRINGIFY(tls_callback) ) +# endif +# pragma section(".CRT$XLY",long,read) +#endif +JEMALLOC_SECTION(".CRT$XLY") JEMALLOC_ATTR(used) +BOOL (WINAPI *const tls_callback)(HINSTANCE hinstDLL, + DWORD fdwReason, LPVOID lpvReserved) = _tls_callback; +#endif + +#if (!defined(JEMALLOC_MALLOC_THREAD_CLEANUP) && !defined(JEMALLOC_TLS) && \ + !defined(_WIN32)) +void * +tsd_init_check_recursion(tsd_init_head_t *head, tsd_init_block_t *block) { + pthread_t self = pthread_self(); + tsd_init_block_t *iter; + + /* Check whether this thread has already inserted into the list. */ + malloc_mutex_lock(TSDN_NULL, &head->lock); + ql_foreach(iter, &head->blocks, link) { + if (iter->thread == self) { + malloc_mutex_unlock(TSDN_NULL, &head->lock); + return iter->data; + } + } + /* Insert block into list. */ + ql_elm_new(block, link); + block->thread = self; + ql_tail_insert(&head->blocks, block, link); + malloc_mutex_unlock(TSDN_NULL, &head->lock); + return NULL; +} + +void +tsd_init_finish(tsd_init_head_t *head, tsd_init_block_t *block) { + malloc_mutex_lock(TSDN_NULL, &head->lock); + ql_remove(&head->blocks, block, link); + malloc_mutex_unlock(TSDN_NULL, &head->lock); +} +#endif + +void +tsd_prefork(tsd_t *tsd) { + malloc_mutex_prefork(tsd_tsdn(tsd), &tsd_nominal_tsds_lock); +} + +void +tsd_postfork_parent(tsd_t *tsd) { + malloc_mutex_postfork_parent(tsd_tsdn(tsd), &tsd_nominal_tsds_lock); +} + +void +tsd_postfork_child(tsd_t *tsd) { + malloc_mutex_postfork_child(tsd_tsdn(tsd), &tsd_nominal_tsds_lock); + ql_new(&tsd_nominal_tsds); + + if (tsd_state_get(tsd) <= tsd_state_nominal_max) { + tsd_add_nominal(tsd); + } +} diff --git a/deps/jemalloc/src/witness.c b/deps/jemalloc/src/witness.c new file mode 100644 index 000000000..f42b72ad1 --- /dev/null +++ b/deps/jemalloc/src/witness.c @@ -0,0 +1,100 @@ +#define JEMALLOC_WITNESS_C_ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/assert.h" +#include "jemalloc/internal/malloc_io.h" + +void +witness_init(witness_t *witness, const char *name, witness_rank_t rank, + witness_comp_t *comp, void *opaque) { + witness->name = name; + witness->rank = rank; + witness->comp = comp; + witness->opaque = opaque; +} + +static void +witness_lock_error_impl(const witness_list_t *witnesses, + const witness_t *witness) { + witness_t *w; + + malloc_printf("<jemalloc>: Lock rank order reversal:"); + ql_foreach(w, witnesses, link) { + malloc_printf(" %s(%u)", w->name, w->rank); + } + malloc_printf(" %s(%u)\n", witness->name, witness->rank); + abort(); +} +witness_lock_error_t *JET_MUTABLE witness_lock_error = witness_lock_error_impl; + +static void +witness_owner_error_impl(const witness_t *witness) { + malloc_printf("<jemalloc>: Should own %s(%u)\n", witness->name, + witness->rank); + abort(); +} +witness_owner_error_t *JET_MUTABLE witness_owner_error = + witness_owner_error_impl; + +static void +witness_not_owner_error_impl(const witness_t *witness) { + malloc_printf("<jemalloc>: Should not own %s(%u)\n", witness->name, + witness->rank); + abort(); +} +witness_not_owner_error_t *JET_MUTABLE witness_not_owner_error = + witness_not_owner_error_impl; + +static void +witness_depth_error_impl(const witness_list_t *witnesses, + witness_rank_t rank_inclusive, unsigned depth) { + witness_t *w; + + malloc_printf("<jemalloc>: Should own %u lock%s of rank >= %u:", depth, + (depth != 1) ? "s" : "", rank_inclusive); + ql_foreach(w, witnesses, link) { + malloc_printf(" %s(%u)", w->name, w->rank); + } + malloc_printf("\n"); + abort(); +} +witness_depth_error_t *JET_MUTABLE witness_depth_error = + witness_depth_error_impl; + +void +witnesses_cleanup(witness_tsd_t *witness_tsd) { + witness_assert_lockless(witness_tsd_tsdn(witness_tsd)); + + /* Do nothing. */ +} + +void +witness_prefork(witness_tsd_t *witness_tsd) { + if (!config_debug) { + return; + } + witness_tsd->forking = true; +} + +void +witness_postfork_parent(witness_tsd_t *witness_tsd) { + if (!config_debug) { + return; + } + witness_tsd->forking = false; +} + +void +witness_postfork_child(witness_tsd_t *witness_tsd) { + if (!config_debug) { + return; + } +#ifndef JEMALLOC_MUTEX_INIT_CB + witness_list_t *witnesses; + + witnesses = &witness_tsd->witnesses; + ql_new(witnesses); +#endif + witness_tsd->forking = false; +} diff --git a/deps/jemalloc/src/zone.c b/deps/jemalloc/src/zone.c new file mode 100644 index 000000000..23dfdd04a --- /dev/null +++ b/deps/jemalloc/src/zone.c @@ -0,0 +1,469 @@ +#include "jemalloc/internal/jemalloc_preamble.h" +#include "jemalloc/internal/jemalloc_internal_includes.h" + +#include "jemalloc/internal/assert.h" + +#ifndef JEMALLOC_ZONE +# error "This source file is for zones on Darwin (OS X)." +#endif + +/* Definitions of the following structs in malloc/malloc.h might be too old + * for the built binary to run on newer versions of OSX. So use the newest + * possible version of those structs. + */ +typedef struct _malloc_zone_t { + void *reserved1; + void *reserved2; + size_t (*size)(struct _malloc_zone_t *, const void *); + void *(*malloc)(struct _malloc_zone_t *, size_t); + void *(*calloc)(struct _malloc_zone_t *, size_t, size_t); + void *(*valloc)(struct _malloc_zone_t *, size_t); + void (*free)(struct _malloc_zone_t *, void *); + void *(*realloc)(struct _malloc_zone_t *, void *, size_t); + void (*destroy)(struct _malloc_zone_t *); + const char *zone_name; + unsigned (*batch_malloc)(struct _malloc_zone_t *, size_t, void **, unsigned); + void (*batch_free)(struct _malloc_zone_t *, void **, unsigned); + struct malloc_introspection_t *introspect; + unsigned version; + void *(*memalign)(struct _malloc_zone_t *, size_t, size_t); + void (*free_definite_size)(struct _malloc_zone_t *, void *, size_t); + size_t (*pressure_relief)(struct _malloc_zone_t *, size_t); +} malloc_zone_t; + +typedef struct { + vm_address_t address; + vm_size_t size; +} vm_range_t; + +typedef struct malloc_statistics_t { + unsigned blocks_in_use; + size_t size_in_use; + size_t max_size_in_use; + size_t size_allocated; +} malloc_statistics_t; + +typedef kern_return_t memory_reader_t(task_t, vm_address_t, vm_size_t, void **); + +typedef void vm_range_recorder_t(task_t, void *, unsigned type, vm_range_t *, unsigned); + +typedef struct malloc_introspection_t { + kern_return_t (*enumerator)(task_t, void *, unsigned, vm_address_t, memory_reader_t, vm_range_recorder_t); + size_t (*good_size)(malloc_zone_t *, size_t); + boolean_t (*check)(malloc_zone_t *); + void (*print)(malloc_zone_t *, boolean_t); + void (*log)(malloc_zone_t *, void *); + void (*force_lock)(malloc_zone_t *); + void (*force_unlock)(malloc_zone_t *); + void (*statistics)(malloc_zone_t *, malloc_statistics_t *); + boolean_t (*zone_locked)(malloc_zone_t *); + boolean_t (*enable_discharge_checking)(malloc_zone_t *); + boolean_t (*disable_discharge_checking)(malloc_zone_t *); + void (*discharge)(malloc_zone_t *, void *); +#ifdef __BLOCKS__ + void (*enumerate_discharged_pointers)(malloc_zone_t *, void (^)(void *, void *)); +#else + void *enumerate_unavailable_without_blocks; +#endif + void (*reinit_lock)(malloc_zone_t *); +} malloc_introspection_t; + +extern kern_return_t malloc_get_all_zones(task_t, memory_reader_t, vm_address_t **, unsigned *); + +extern malloc_zone_t *malloc_default_zone(void); + +extern void malloc_zone_register(malloc_zone_t *zone); + +extern void malloc_zone_unregister(malloc_zone_t *zone); + +/* + * The malloc_default_purgeable_zone() function is only available on >= 10.6. + * We need to check whether it is present at runtime, thus the weak_import. + */ +extern malloc_zone_t *malloc_default_purgeable_zone(void) +JEMALLOC_ATTR(weak_import); + +/******************************************************************************/ +/* Data. */ + +static malloc_zone_t *default_zone, *purgeable_zone; +static malloc_zone_t jemalloc_zone; +static struct malloc_introspection_t jemalloc_zone_introspect; +static pid_t zone_force_lock_pid = -1; + +/******************************************************************************/ +/* Function prototypes for non-inline static functions. */ + +static size_t zone_size(malloc_zone_t *zone, const void *ptr); +static void *zone_malloc(malloc_zone_t *zone, size_t size); +static void *zone_calloc(malloc_zone_t *zone, size_t num, size_t size); +static void *zone_valloc(malloc_zone_t *zone, size_t size); +static void zone_free(malloc_zone_t *zone, void *ptr); +static void *zone_realloc(malloc_zone_t *zone, void *ptr, size_t size); +static void *zone_memalign(malloc_zone_t *zone, size_t alignment, + size_t size); +static void zone_free_definite_size(malloc_zone_t *zone, void *ptr, + size_t size); +static void zone_destroy(malloc_zone_t *zone); +static unsigned zone_batch_malloc(struct _malloc_zone_t *zone, size_t size, + void **results, unsigned num_requested); +static void zone_batch_free(struct _malloc_zone_t *zone, + void **to_be_freed, unsigned num_to_be_freed); +static size_t zone_pressure_relief(struct _malloc_zone_t *zone, size_t goal); +static size_t zone_good_size(malloc_zone_t *zone, size_t size); +static kern_return_t zone_enumerator(task_t task, void *data, unsigned type_mask, + vm_address_t zone_address, memory_reader_t reader, + vm_range_recorder_t recorder); +static boolean_t zone_check(malloc_zone_t *zone); +static void zone_print(malloc_zone_t *zone, boolean_t verbose); +static void zone_log(malloc_zone_t *zone, void *address); +static void zone_force_lock(malloc_zone_t *zone); +static void zone_force_unlock(malloc_zone_t *zone); +static void zone_statistics(malloc_zone_t *zone, + malloc_statistics_t *stats); +static boolean_t zone_locked(malloc_zone_t *zone); +static void zone_reinit_lock(malloc_zone_t *zone); + +/******************************************************************************/ +/* + * Functions. + */ + +static size_t +zone_size(malloc_zone_t *zone, const void *ptr) { + /* + * There appear to be places within Darwin (such as setenv(3)) that + * cause calls to this function with pointers that *no* zone owns. If + * we knew that all pointers were owned by *some* zone, we could split + * our zone into two parts, and use one as the default allocator and + * the other as the default deallocator/reallocator. Since that will + * not work in practice, we must check all pointers to assure that they + * reside within a mapped extent before determining size. + */ + return ivsalloc(tsdn_fetch(), ptr); +} + +static void * +zone_malloc(malloc_zone_t *zone, size_t size) { + return je_malloc(size); +} + +static void * +zone_calloc(malloc_zone_t *zone, size_t num, size_t size) { + return je_calloc(num, size); +} + +static void * +zone_valloc(malloc_zone_t *zone, size_t size) { + void *ret = NULL; /* Assignment avoids useless compiler warning. */ + + je_posix_memalign(&ret, PAGE, size); + + return ret; +} + +static void +zone_free(malloc_zone_t *zone, void *ptr) { + if (ivsalloc(tsdn_fetch(), ptr) != 0) { + je_free(ptr); + return; + } + + free(ptr); +} + +static void * +zone_realloc(malloc_zone_t *zone, void *ptr, size_t size) { + if (ivsalloc(tsdn_fetch(), ptr) != 0) { + return je_realloc(ptr, size); + } + + return realloc(ptr, size); +} + +static void * +zone_memalign(malloc_zone_t *zone, size_t alignment, size_t size) { + void *ret = NULL; /* Assignment avoids useless compiler warning. */ + + je_posix_memalign(&ret, alignment, size); + + return ret; +} + +static void +zone_free_definite_size(malloc_zone_t *zone, void *ptr, size_t size) { + size_t alloc_size; + + alloc_size = ivsalloc(tsdn_fetch(), ptr); + if (alloc_size != 0) { + assert(alloc_size == size); + je_free(ptr); + return; + } + + free(ptr); +} + +static void +zone_destroy(malloc_zone_t *zone) { + /* This function should never be called. */ + not_reached(); +} + +static unsigned +zone_batch_malloc(struct _malloc_zone_t *zone, size_t size, void **results, + unsigned num_requested) { + unsigned i; + + for (i = 0; i < num_requested; i++) { + results[i] = je_malloc(size); + if (!results[i]) + break; + } + + return i; +} + +static void +zone_batch_free(struct _malloc_zone_t *zone, void **to_be_freed, + unsigned num_to_be_freed) { + unsigned i; + + for (i = 0; i < num_to_be_freed; i++) { + zone_free(zone, to_be_freed[i]); + to_be_freed[i] = NULL; + } +} + +static size_t +zone_pressure_relief(struct _malloc_zone_t *zone, size_t goal) { + return 0; +} + +static size_t +zone_good_size(malloc_zone_t *zone, size_t size) { + if (size == 0) { + size = 1; + } + return sz_s2u(size); +} + +static kern_return_t +zone_enumerator(task_t task, void *data, unsigned type_mask, + vm_address_t zone_address, memory_reader_t reader, + vm_range_recorder_t recorder) { + return KERN_SUCCESS; +} + +static boolean_t +zone_check(malloc_zone_t *zone) { + return true; +} + +static void +zone_print(malloc_zone_t *zone, boolean_t verbose) { +} + +static void +zone_log(malloc_zone_t *zone, void *address) { +} + +static void +zone_force_lock(malloc_zone_t *zone) { + if (isthreaded) { + /* + * See the note in zone_force_unlock, below, to see why we need + * this. + */ + assert(zone_force_lock_pid == -1); + zone_force_lock_pid = getpid(); + jemalloc_prefork(); + } +} + +static void +zone_force_unlock(malloc_zone_t *zone) { + /* + * zone_force_lock and zone_force_unlock are the entry points to the + * forking machinery on OS X. The tricky thing is, the child is not + * allowed to unlock mutexes locked in the parent, even if owned by the + * forking thread (and the mutex type we use in OS X will fail an assert + * if we try). In the child, we can get away with reinitializing all + * the mutexes, which has the effect of unlocking them. In the parent, + * doing this would mean we wouldn't wake any waiters blocked on the + * mutexes we unlock. So, we record the pid of the current thread in + * zone_force_lock, and use that to detect if we're in the parent or + * child here, to decide which unlock logic we need. + */ + if (isthreaded) { + assert(zone_force_lock_pid != -1); + if (getpid() == zone_force_lock_pid) { + jemalloc_postfork_parent(); + } else { + jemalloc_postfork_child(); + } + zone_force_lock_pid = -1; + } +} + +static void +zone_statistics(malloc_zone_t *zone, malloc_statistics_t *stats) { + /* We make no effort to actually fill the values */ + stats->blocks_in_use = 0; + stats->size_in_use = 0; + stats->max_size_in_use = 0; + stats->size_allocated = 0; +} + +static boolean_t +zone_locked(malloc_zone_t *zone) { + /* Pretend no lock is being held */ + return false; +} + +static void +zone_reinit_lock(malloc_zone_t *zone) { + /* As of OSX 10.12, this function is only used when force_unlock would + * be used if the zone version were < 9. So just use force_unlock. */ + zone_force_unlock(zone); +} + +static void +zone_init(void) { + jemalloc_zone.size = zone_size; + jemalloc_zone.malloc = zone_malloc; + jemalloc_zone.calloc = zone_calloc; + jemalloc_zone.valloc = zone_valloc; + jemalloc_zone.free = zone_free; + jemalloc_zone.realloc = zone_realloc; + jemalloc_zone.destroy = zone_destroy; + jemalloc_zone.zone_name = "jemalloc_zone"; + jemalloc_zone.batch_malloc = zone_batch_malloc; + jemalloc_zone.batch_free = zone_batch_free; + jemalloc_zone.introspect = &jemalloc_zone_introspect; + jemalloc_zone.version = 9; + jemalloc_zone.memalign = zone_memalign; + jemalloc_zone.free_definite_size = zone_free_definite_size; + jemalloc_zone.pressure_relief = zone_pressure_relief; + + jemalloc_zone_introspect.enumerator = zone_enumerator; + jemalloc_zone_introspect.good_size = zone_good_size; + jemalloc_zone_introspect.check = zone_check; + jemalloc_zone_introspect.print = zone_print; + jemalloc_zone_introspect.log = zone_log; + jemalloc_zone_introspect.force_lock = zone_force_lock; + jemalloc_zone_introspect.force_unlock = zone_force_unlock; + jemalloc_zone_introspect.statistics = zone_statistics; + jemalloc_zone_introspect.zone_locked = zone_locked; + jemalloc_zone_introspect.enable_discharge_checking = NULL; + jemalloc_zone_introspect.disable_discharge_checking = NULL; + jemalloc_zone_introspect.discharge = NULL; +#ifdef __BLOCKS__ + jemalloc_zone_introspect.enumerate_discharged_pointers = NULL; +#else + jemalloc_zone_introspect.enumerate_unavailable_without_blocks = NULL; +#endif + jemalloc_zone_introspect.reinit_lock = zone_reinit_lock; +} + +static malloc_zone_t * +zone_default_get(void) { + malloc_zone_t **zones = NULL; + unsigned int num_zones = 0; + + /* + * On OSX 10.12, malloc_default_zone returns a special zone that is not + * present in the list of registered zones. That zone uses a "lite zone" + * if one is present (apparently enabled when malloc stack logging is + * enabled), or the first registered zone otherwise. In practice this + * means unless malloc stack logging is enabled, the first registered + * zone is the default. So get the list of zones to get the first one, + * instead of relying on malloc_default_zone. + */ + if (KERN_SUCCESS != malloc_get_all_zones(0, NULL, + (vm_address_t**)&zones, &num_zones)) { + /* + * Reset the value in case the failure happened after it was + * set. + */ + num_zones = 0; + } + + if (num_zones) { + return zones[0]; + } + + return malloc_default_zone(); +} + +/* As written, this function can only promote jemalloc_zone. */ +static void +zone_promote(void) { + malloc_zone_t *zone; + + do { + /* + * Unregister and reregister the default zone. On OSX >= 10.6, + * unregistering takes the last registered zone and places it + * at the location of the specified zone. Unregistering the + * default zone thus makes the last registered one the default. + * On OSX < 10.6, unregistering shifts all registered zones. + * The first registered zone then becomes the default. + */ + malloc_zone_unregister(default_zone); + malloc_zone_register(default_zone); + + /* + * On OSX 10.6, having the default purgeable zone appear before + * the default zone makes some things crash because it thinks it + * owns the default zone allocated pointers. We thus + * unregister/re-register it in order to ensure it's always + * after the default zone. On OSX < 10.6, there is no purgeable + * zone, so this does nothing. On OSX >= 10.6, unregistering + * replaces the purgeable zone with the last registered zone + * above, i.e. the default zone. Registering it again then puts + * it at the end, obviously after the default zone. + */ + if (purgeable_zone != NULL) { + malloc_zone_unregister(purgeable_zone); + malloc_zone_register(purgeable_zone); + } + + zone = zone_default_get(); + } while (zone != &jemalloc_zone); +} + +JEMALLOC_ATTR(constructor) +void +zone_register(void) { + /* + * If something else replaced the system default zone allocator, don't + * register jemalloc's. + */ + default_zone = zone_default_get(); + if (!default_zone->zone_name || strcmp(default_zone->zone_name, + "DefaultMallocZone") != 0) { + return; + } + + /* + * The default purgeable zone is created lazily by OSX's libc. It uses + * the default zone when it is created for "small" allocations + * (< 15 KiB), but assumes the default zone is a scalable_zone. This + * obviously fails when the default zone is the jemalloc zone, so + * malloc_default_purgeable_zone() is called beforehand so that the + * default purgeable zone is created when the default zone is still + * a scalable_zone. As purgeable zones only exist on >= 10.6, we need + * to check for the existence of malloc_default_purgeable_zone() at + * run time. + */ + purgeable_zone = (malloc_default_purgeable_zone == NULL) ? NULL : + malloc_default_purgeable_zone(); + + /* Register the custom zone. At this point it won't be the default. */ + zone_init(); + malloc_zone_register(&jemalloc_zone); + + /* Promote the custom zone to be default. */ + zone_promote(); +} diff --git a/deps/jemalloc/test/include/test/SFMT-alti.h b/deps/jemalloc/test/include/test/SFMT-alti.h new file mode 100644 index 000000000..a1885dbf2 --- /dev/null +++ b/deps/jemalloc/test/include/test/SFMT-alti.h @@ -0,0 +1,186 @@ +/* + * This file derives from SFMT 1.3.3 + * (http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/SFMT/index.html), which was + * released under the terms of the following license: + * + * Copyright (c) 2006,2007 Mutsuo Saito, Makoto Matsumoto and Hiroshima + * University. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions are + * met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials provided + * with the distribution. + * * Neither the name of the Hiroshima University nor the names of + * its contributors may be used to endorse or promote products + * derived from this software without specific prior written + * permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ +/** + * @file SFMT-alti.h + * + * @brief SIMD oriented Fast Mersenne Twister(SFMT) + * pseudorandom number generator + * + * @author Mutsuo Saito (Hiroshima University) + * @author Makoto Matsumoto (Hiroshima University) + * + * Copyright (C) 2007 Mutsuo Saito, Makoto Matsumoto and Hiroshima + * University. All rights reserved. + * + * The new BSD License is applied to this software. + * see LICENSE.txt + */ + +#ifndef SFMT_ALTI_H +#define SFMT_ALTI_H + +/** + * This function represents the recursion formula in AltiVec and BIG ENDIAN. + * @param a a 128-bit part of the interal state array + * @param b a 128-bit part of the interal state array + * @param c a 128-bit part of the interal state array + * @param d a 128-bit part of the interal state array + * @return output + */ +JEMALLOC_ALWAYS_INLINE +vector unsigned int vec_recursion(vector unsigned int a, + vector unsigned int b, + vector unsigned int c, + vector unsigned int d) { + + const vector unsigned int sl1 = ALTI_SL1; + const vector unsigned int sr1 = ALTI_SR1; +#ifdef ONLY64 + const vector unsigned int mask = ALTI_MSK64; + const vector unsigned char perm_sl = ALTI_SL2_PERM64; + const vector unsigned char perm_sr = ALTI_SR2_PERM64; +#else + const vector unsigned int mask = ALTI_MSK; + const vector unsigned char perm_sl = ALTI_SL2_PERM; + const vector unsigned char perm_sr = ALTI_SR2_PERM; +#endif + vector unsigned int v, w, x, y, z; + x = vec_perm(a, (vector unsigned int)perm_sl, perm_sl); + v = a; + y = vec_sr(b, sr1); + z = vec_perm(c, (vector unsigned int)perm_sr, perm_sr); + w = vec_sl(d, sl1); + z = vec_xor(z, w); + y = vec_and(y, mask); + v = vec_xor(v, x); + z = vec_xor(z, y); + z = vec_xor(z, v); + return z; +} + +/** + * This function fills the internal state array with pseudorandom + * integers. + */ +static inline void gen_rand_all(sfmt_t *ctx) { + int i; + vector unsigned int r, r1, r2; + + r1 = ctx->sfmt[N - 2].s; + r2 = ctx->sfmt[N - 1].s; + for (i = 0; i < N - POS1; i++) { + r = vec_recursion(ctx->sfmt[i].s, ctx->sfmt[i + POS1].s, r1, r2); + ctx->sfmt[i].s = r; + r1 = r2; + r2 = r; + } + for (; i < N; i++) { + r = vec_recursion(ctx->sfmt[i].s, ctx->sfmt[i + POS1 - N].s, r1, r2); + ctx->sfmt[i].s = r; + r1 = r2; + r2 = r; + } +} + +/** + * This function fills the user-specified array with pseudorandom + * integers. + * + * @param array an 128-bit array to be filled by pseudorandom numbers. + * @param size number of 128-bit pesudorandom numbers to be generated. + */ +static inline void gen_rand_array(sfmt_t *ctx, w128_t *array, int size) { + int i, j; + vector unsigned int r, r1, r2; + + r1 = ctx->sfmt[N - 2].s; + r2 = ctx->sfmt[N - 1].s; + for (i = 0; i < N - POS1; i++) { + r = vec_recursion(ctx->sfmt[i].s, ctx->sfmt[i + POS1].s, r1, r2); + array[i].s = r; + r1 = r2; + r2 = r; + } + for (; i < N; i++) { + r = vec_recursion(ctx->sfmt[i].s, array[i + POS1 - N].s, r1, r2); + array[i].s = r; + r1 = r2; + r2 = r; + } + /* main loop */ + for (; i < size - N; i++) { + r = vec_recursion(array[i - N].s, array[i + POS1 - N].s, r1, r2); + array[i].s = r; + r1 = r2; + r2 = r; + } + for (j = 0; j < 2 * N - size; j++) { + ctx->sfmt[j].s = array[j + size - N].s; + } + for (; i < size; i++) { + r = vec_recursion(array[i - N].s, array[i + POS1 - N].s, r1, r2); + array[i].s = r; + ctx->sfmt[j++].s = r; + r1 = r2; + r2 = r; + } +} + +#ifndef ONLY64 +#if defined(__APPLE__) +#define ALTI_SWAP (vector unsigned char) \ + (4, 5, 6, 7, 0, 1, 2, 3, 12, 13, 14, 15, 8, 9, 10, 11) +#else +#define ALTI_SWAP {4, 5, 6, 7, 0, 1, 2, 3, 12, 13, 14, 15, 8, 9, 10, 11} +#endif +/** + * This function swaps high and low 32-bit of 64-bit integers in user + * specified array. + * + * @param array an 128-bit array to be swaped. + * @param size size of 128-bit array. + */ +static inline void swap(w128_t *array, int size) { + int i; + const vector unsigned char perm = ALTI_SWAP; + + for (i = 0; i < size; i++) { + array[i].s = vec_perm(array[i].s, (vector unsigned int)perm, perm); + } +} +#endif + +#endif diff --git a/deps/jemalloc/test/include/test/SFMT-params.h b/deps/jemalloc/test/include/test/SFMT-params.h new file mode 100644 index 000000000..ade662220 --- /dev/null +++ b/deps/jemalloc/test/include/test/SFMT-params.h @@ -0,0 +1,132 @@ +/* + * This file derives from SFMT 1.3.3 + * (http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/SFMT/index.html), which was + * released under the terms of the following license: + * + * Copyright (c) 2006,2007 Mutsuo Saito, Makoto Matsumoto and Hiroshima + * University. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions are + * met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials provided + * with the distribution. + * * Neither the name of the Hiroshima University nor the names of + * its contributors may be used to endorse or promote products + * derived from this software without specific prior written + * permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ +#ifndef SFMT_PARAMS_H +#define SFMT_PARAMS_H + +#if !defined(MEXP) +#ifdef __GNUC__ + #warning "MEXP is not defined. I assume MEXP is 19937." +#endif + #define MEXP 19937 +#endif +/*----------------- + BASIC DEFINITIONS + -----------------*/ +/** Mersenne Exponent. The period of the sequence + * is a multiple of 2^MEXP-1. + * #define MEXP 19937 */ +/** SFMT generator has an internal state array of 128-bit integers, + * and N is its size. */ +#define N (MEXP / 128 + 1) +/** N32 is the size of internal state array when regarded as an array + * of 32-bit integers.*/ +#define N32 (N * 4) +/** N64 is the size of internal state array when regarded as an array + * of 64-bit integers.*/ +#define N64 (N * 2) + +/*---------------------- + the parameters of SFMT + following definitions are in paramsXXXX.h file. + ----------------------*/ +/** the pick up position of the array. +#define POS1 122 +*/ + +/** the parameter of shift left as four 32-bit registers. +#define SL1 18 + */ + +/** the parameter of shift left as one 128-bit register. + * The 128-bit integer is shifted by (SL2 * 8) bits. +#define SL2 1 +*/ + +/** the parameter of shift right as four 32-bit registers. +#define SR1 11 +*/ + +/** the parameter of shift right as one 128-bit register. + * The 128-bit integer is shifted by (SL2 * 8) bits. +#define SR2 1 +*/ + +/** A bitmask, used in the recursion. These parameters are introduced + * to break symmetry of SIMD. +#define MSK1 0xdfffffefU +#define MSK2 0xddfecb7fU +#define MSK3 0xbffaffffU +#define MSK4 0xbffffff6U +*/ + +/** These definitions are part of a 128-bit period certification vector. +#define PARITY1 0x00000001U +#define PARITY2 0x00000000U +#define PARITY3 0x00000000U +#define PARITY4 0xc98e126aU +*/ + +#if MEXP == 607 + #include "test/SFMT-params607.h" +#elif MEXP == 1279 + #include "test/SFMT-params1279.h" +#elif MEXP == 2281 + #include "test/SFMT-params2281.h" +#elif MEXP == 4253 + #include "test/SFMT-params4253.h" +#elif MEXP == 11213 + #include "test/SFMT-params11213.h" +#elif MEXP == 19937 + #include "test/SFMT-params19937.h" +#elif MEXP == 44497 + #include "test/SFMT-params44497.h" +#elif MEXP == 86243 + #include "test/SFMT-params86243.h" +#elif MEXP == 132049 + #include "test/SFMT-params132049.h" +#elif MEXP == 216091 + #include "test/SFMT-params216091.h" +#else +#ifdef __GNUC__ + #error "MEXP is not valid." + #undef MEXP +#else + #undef MEXP +#endif + +#endif + +#endif /* SFMT_PARAMS_H */ diff --git a/deps/jemalloc/test/include/test/SFMT-params11213.h b/deps/jemalloc/test/include/test/SFMT-params11213.h new file mode 100644 index 000000000..2994bd21d --- /dev/null +++ b/deps/jemalloc/test/include/test/SFMT-params11213.h @@ -0,0 +1,81 @@ +/* + * This file derives from SFMT 1.3.3 + * (http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/SFMT/index.html), which was + * released under the terms of the following license: + * + * Copyright (c) 2006,2007 Mutsuo Saito, Makoto Matsumoto and Hiroshima + * University. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions are + * met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials provided + * with the distribution. + * * Neither the name of the Hiroshima University nor the names of + * its contributors may be used to endorse or promote products + * derived from this software without specific prior written + * permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ +#ifndef SFMT_PARAMS11213_H +#define SFMT_PARAMS11213_H + +#define POS1 68 +#define SL1 14 +#define SL2 3 +#define SR1 7 +#define SR2 3 +#define MSK1 0xeffff7fbU +#define MSK2 0xffffffefU +#define MSK3 0xdfdfbfffU +#define MSK4 0x7fffdbfdU +#define PARITY1 0x00000001U +#define PARITY2 0x00000000U +#define PARITY3 0xe8148000U +#define PARITY4 0xd0c7afa3U + + +/* PARAMETERS FOR ALTIVEC */ +#if defined(__APPLE__) /* For OSX */ + #define ALTI_SL1 (vector unsigned int)(SL1, SL1, SL1, SL1) + #define ALTI_SR1 (vector unsigned int)(SR1, SR1, SR1, SR1) + #define ALTI_MSK (vector unsigned int)(MSK1, MSK2, MSK3, MSK4) + #define ALTI_MSK64 \ + (vector unsigned int)(MSK2, MSK1, MSK4, MSK3) + #define ALTI_SL2_PERM \ + (vector unsigned char)(3,21,21,21,7,0,1,2,11,4,5,6,15,8,9,10) + #define ALTI_SL2_PERM64 \ + (vector unsigned char)(3,4,5,6,7,29,29,29,11,12,13,14,15,0,1,2) + #define ALTI_SR2_PERM \ + (vector unsigned char)(5,6,7,0,9,10,11,4,13,14,15,8,19,19,19,12) + #define ALTI_SR2_PERM64 \ + (vector unsigned char)(13,14,15,0,1,2,3,4,19,19,19,8,9,10,11,12) +#else /* For OTHER OSs(Linux?) */ + #define ALTI_SL1 {SL1, SL1, SL1, SL1} + #define ALTI_SR1 {SR1, SR1, SR1, SR1} + #define ALTI_MSK {MSK1, MSK2, MSK3, MSK4} + #define ALTI_MSK64 {MSK2, MSK1, MSK4, MSK3} + #define ALTI_SL2_PERM {3,21,21,21,7,0,1,2,11,4,5,6,15,8,9,10} + #define ALTI_SL2_PERM64 {3,4,5,6,7,29,29,29,11,12,13,14,15,0,1,2} + #define ALTI_SR2_PERM {5,6,7,0,9,10,11,4,13,14,15,8,19,19,19,12} + #define ALTI_SR2_PERM64 {13,14,15,0,1,2,3,4,19,19,19,8,9,10,11,12} +#endif /* For OSX */ +#define IDSTR "SFMT-11213:68-14-3-7-3:effff7fb-ffffffef-dfdfbfff-7fffdbfd" + +#endif /* SFMT_PARAMS11213_H */ diff --git a/deps/jemalloc/test/include/test/SFMT-params1279.h b/deps/jemalloc/test/include/test/SFMT-params1279.h new file mode 100644 index 000000000..d7959f980 --- /dev/null +++ b/deps/jemalloc/test/include/test/SFMT-params1279.h @@ -0,0 +1,81 @@ +/* + * This file derives from SFMT 1.3.3 + * (http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/SFMT/index.html), which was + * released under the terms of the following license: + * + * Copyright (c) 2006,2007 Mutsuo Saito, Makoto Matsumoto and Hiroshima + * University. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions are + * met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials provided + * with the distribution. + * * Neither the name of the Hiroshima University nor the names of + * its contributors may be used to endorse or promote products + * derived from this software without specific prior written + * permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ +#ifndef SFMT_PARAMS1279_H +#define SFMT_PARAMS1279_H + +#define POS1 7 +#define SL1 14 +#define SL2 3 +#define SR1 5 +#define SR2 1 +#define MSK1 0xf7fefffdU +#define MSK2 0x7fefcfffU +#define MSK3 0xaff3ef3fU +#define MSK4 0xb5ffff7fU +#define PARITY1 0x00000001U +#define PARITY2 0x00000000U +#define PARITY3 0x00000000U +#define PARITY4 0x20000000U + + +/* PARAMETERS FOR ALTIVEC */ +#if defined(__APPLE__) /* For OSX */ + #define ALTI_SL1 (vector unsigned int)(SL1, SL1, SL1, SL1) + #define ALTI_SR1 (vector unsigned int)(SR1, SR1, SR1, SR1) + #define ALTI_MSK (vector unsigned int)(MSK1, MSK2, MSK3, MSK4) + #define ALTI_MSK64 \ + (vector unsigned int)(MSK2, MSK1, MSK4, MSK3) + #define ALTI_SL2_PERM \ + (vector unsigned char)(3,21,21,21,7,0,1,2,11,4,5,6,15,8,9,10) + #define ALTI_SL2_PERM64 \ + (vector unsigned char)(3,4,5,6,7,29,29,29,11,12,13,14,15,0,1,2) + #define ALTI_SR2_PERM \ + (vector unsigned char)(7,0,1,2,11,4,5,6,15,8,9,10,17,12,13,14) + #define ALTI_SR2_PERM64 \ + (vector unsigned char)(15,0,1,2,3,4,5,6,17,8,9,10,11,12,13,14) +#else /* For OTHER OSs(Linux?) */ + #define ALTI_SL1 {SL1, SL1, SL1, SL1} + #define ALTI_SR1 {SR1, SR1, SR1, SR1} + #define ALTI_MSK {MSK1, MSK2, MSK3, MSK4} + #define ALTI_MSK64 {MSK2, MSK1, MSK4, MSK3} + #define ALTI_SL2_PERM {3,21,21,21,7,0,1,2,11,4,5,6,15,8,9,10} + #define ALTI_SL2_PERM64 {3,4,5,6,7,29,29,29,11,12,13,14,15,0,1,2} + #define ALTI_SR2_PERM {7,0,1,2,11,4,5,6,15,8,9,10,17,12,13,14} + #define ALTI_SR2_PERM64 {15,0,1,2,3,4,5,6,17,8,9,10,11,12,13,14} +#endif /* For OSX */ +#define IDSTR "SFMT-1279:7-14-3-5-1:f7fefffd-7fefcfff-aff3ef3f-b5ffff7f" + +#endif /* SFMT_PARAMS1279_H */ diff --git a/deps/jemalloc/test/include/test/SFMT-params132049.h b/deps/jemalloc/test/include/test/SFMT-params132049.h new file mode 100644 index 000000000..a1dcec392 --- /dev/null +++ b/deps/jemalloc/test/include/test/SFMT-params132049.h @@ -0,0 +1,81 @@ +/* + * This file derives from SFMT 1.3.3 + * (http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/SFMT/index.html), which was + * released under the terms of the following license: + * + * Copyright (c) 2006,2007 Mutsuo Saito, Makoto Matsumoto and Hiroshima + * University. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions are + * met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials provided + * with the distribution. + * * Neither the name of the Hiroshima University nor the names of + * its contributors may be used to endorse or promote products + * derived from this software without specific prior written + * permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ +#ifndef SFMT_PARAMS132049_H +#define SFMT_PARAMS132049_H + +#define POS1 110 +#define SL1 19 +#define SL2 1 +#define SR1 21 +#define SR2 1 +#define MSK1 0xffffbb5fU +#define MSK2 0xfb6ebf95U +#define MSK3 0xfffefffaU +#define MSK4 0xcff77fffU +#define PARITY1 0x00000001U +#define PARITY2 0x00000000U +#define PARITY3 0xcb520000U +#define PARITY4 0xc7e91c7dU + + +/* PARAMETERS FOR ALTIVEC */ +#if defined(__APPLE__) /* For OSX */ + #define ALTI_SL1 (vector unsigned int)(SL1, SL1, SL1, SL1) + #define ALTI_SR1 (vector unsigned int)(SR1, SR1, SR1, SR1) + #define ALTI_MSK (vector unsigned int)(MSK1, MSK2, MSK3, MSK4) + #define ALTI_MSK64 \ + (vector unsigned int)(MSK2, MSK1, MSK4, MSK3) + #define ALTI_SL2_PERM \ + (vector unsigned char)(1,2,3,23,5,6,7,0,9,10,11,4,13,14,15,8) + #define ALTI_SL2_PERM64 \ + (vector unsigned char)(1,2,3,4,5,6,7,31,9,10,11,12,13,14,15,0) + #define ALTI_SR2_PERM \ + (vector unsigned char)(7,0,1,2,11,4,5,6,15,8,9,10,17,12,13,14) + #define ALTI_SR2_PERM64 \ + (vector unsigned char)(15,0,1,2,3,4,5,6,17,8,9,10,11,12,13,14) +#else /* For OTHER OSs(Linux?) */ + #define ALTI_SL1 {SL1, SL1, SL1, SL1} + #define ALTI_SR1 {SR1, SR1, SR1, SR1} + #define ALTI_MSK {MSK1, MSK2, MSK3, MSK4} + #define ALTI_MSK64 {MSK2, MSK1, MSK4, MSK3} + #define ALTI_SL2_PERM {1,2,3,23,5,6,7,0,9,10,11,4,13,14,15,8} + #define ALTI_SL2_PERM64 {1,2,3,4,5,6,7,31,9,10,11,12,13,14,15,0} + #define ALTI_SR2_PERM {7,0,1,2,11,4,5,6,15,8,9,10,17,12,13,14} + #define ALTI_SR2_PERM64 {15,0,1,2,3,4,5,6,17,8,9,10,11,12,13,14} +#endif /* For OSX */ +#define IDSTR "SFMT-132049:110-19-1-21-1:ffffbb5f-fb6ebf95-fffefffa-cff77fff" + +#endif /* SFMT_PARAMS132049_H */ diff --git a/deps/jemalloc/test/include/test/SFMT-params19937.h b/deps/jemalloc/test/include/test/SFMT-params19937.h new file mode 100644 index 000000000..fb92b4c9b --- /dev/null +++ b/deps/jemalloc/test/include/test/SFMT-params19937.h @@ -0,0 +1,81 @@ +/* + * This file derives from SFMT 1.3.3 + * (http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/SFMT/index.html), which was + * released under the terms of the following license: + * + * Copyright (c) 2006,2007 Mutsuo Saito, Makoto Matsumoto and Hiroshima + * University. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions are + * met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials provided + * with the distribution. + * * Neither the name of the Hiroshima University nor the names of + * its contributors may be used to endorse or promote products + * derived from this software without specific prior written + * permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ +#ifndef SFMT_PARAMS19937_H +#define SFMT_PARAMS19937_H + +#define POS1 122 +#define SL1 18 +#define SL2 1 +#define SR1 11 +#define SR2 1 +#define MSK1 0xdfffffefU +#define MSK2 0xddfecb7fU +#define MSK3 0xbffaffffU +#define MSK4 0xbffffff6U +#define PARITY1 0x00000001U +#define PARITY2 0x00000000U +#define PARITY3 0x00000000U +#define PARITY4 0x13c9e684U + + +/* PARAMETERS FOR ALTIVEC */ +#if defined(__APPLE__) /* For OSX */ + #define ALTI_SL1 (vector unsigned int)(SL1, SL1, SL1, SL1) + #define ALTI_SR1 (vector unsigned int)(SR1, SR1, SR1, SR1) + #define ALTI_MSK (vector unsigned int)(MSK1, MSK2, MSK3, MSK4) + #define ALTI_MSK64 \ + (vector unsigned int)(MSK2, MSK1, MSK4, MSK3) + #define ALTI_SL2_PERM \ + (vector unsigned char)(1,2,3,23,5,6,7,0,9,10,11,4,13,14,15,8) + #define ALTI_SL2_PERM64 \ + (vector unsigned char)(1,2,3,4,5,6,7,31,9,10,11,12,13,14,15,0) + #define ALTI_SR2_PERM \ + (vector unsigned char)(7,0,1,2,11,4,5,6,15,8,9,10,17,12,13,14) + #define ALTI_SR2_PERM64 \ + (vector unsigned char)(15,0,1,2,3,4,5,6,17,8,9,10,11,12,13,14) +#else /* For OTHER OSs(Linux?) */ + #define ALTI_SL1 {SL1, SL1, SL1, SL1} + #define ALTI_SR1 {SR1, SR1, SR1, SR1} + #define ALTI_MSK {MSK1, MSK2, MSK3, MSK4} + #define ALTI_MSK64 {MSK2, MSK1, MSK4, MSK3} + #define ALTI_SL2_PERM {1,2,3,23,5,6,7,0,9,10,11,4,13,14,15,8} + #define ALTI_SL2_PERM64 {1,2,3,4,5,6,7,31,9,10,11,12,13,14,15,0} + #define ALTI_SR2_PERM {7,0,1,2,11,4,5,6,15,8,9,10,17,12,13,14} + #define ALTI_SR2_PERM64 {15,0,1,2,3,4,5,6,17,8,9,10,11,12,13,14} +#endif /* For OSX */ +#define IDSTR "SFMT-19937:122-18-1-11-1:dfffffef-ddfecb7f-bffaffff-bffffff6" + +#endif /* SFMT_PARAMS19937_H */ diff --git a/deps/jemalloc/test/include/test/SFMT-params216091.h b/deps/jemalloc/test/include/test/SFMT-params216091.h new file mode 100644 index 000000000..125ce2820 --- /dev/null +++ b/deps/jemalloc/test/include/test/SFMT-params216091.h @@ -0,0 +1,81 @@ +/* + * This file derives from SFMT 1.3.3 + * (http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/SFMT/index.html), which was + * released under the terms of the following license: + * + * Copyright (c) 2006,2007 Mutsuo Saito, Makoto Matsumoto and Hiroshima + * University. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions are + * met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials provided + * with the distribution. + * * Neither the name of the Hiroshima University nor the names of + * its contributors may be used to endorse or promote products + * derived from this software without specific prior written + * permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ +#ifndef SFMT_PARAMS216091_H +#define SFMT_PARAMS216091_H + +#define POS1 627 +#define SL1 11 +#define SL2 3 +#define SR1 10 +#define SR2 1 +#define MSK1 0xbff7bff7U +#define MSK2 0xbfffffffU +#define MSK3 0xbffffa7fU +#define MSK4 0xffddfbfbU +#define PARITY1 0xf8000001U +#define PARITY2 0x89e80709U +#define PARITY3 0x3bd2b64bU +#define PARITY4 0x0c64b1e4U + + +/* PARAMETERS FOR ALTIVEC */ +#if defined(__APPLE__) /* For OSX */ + #define ALTI_SL1 (vector unsigned int)(SL1, SL1, SL1, SL1) + #define ALTI_SR1 (vector unsigned int)(SR1, SR1, SR1, SR1) + #define ALTI_MSK (vector unsigned int)(MSK1, MSK2, MSK3, MSK4) + #define ALTI_MSK64 \ + (vector unsigned int)(MSK2, MSK1, MSK4, MSK3) + #define ALTI_SL2_PERM \ + (vector unsigned char)(3,21,21,21,7,0,1,2,11,4,5,6,15,8,9,10) + #define ALTI_SL2_PERM64 \ + (vector unsigned char)(3,4,5,6,7,29,29,29,11,12,13,14,15,0,1,2) + #define ALTI_SR2_PERM \ + (vector unsigned char)(7,0,1,2,11,4,5,6,15,8,9,10,17,12,13,14) + #define ALTI_SR2_PERM64 \ + (vector unsigned char)(15,0,1,2,3,4,5,6,17,8,9,10,11,12,13,14) +#else /* For OTHER OSs(Linux?) */ + #define ALTI_SL1 {SL1, SL1, SL1, SL1} + #define ALTI_SR1 {SR1, SR1, SR1, SR1} + #define ALTI_MSK {MSK1, MSK2, MSK3, MSK4} + #define ALTI_MSK64 {MSK2, MSK1, MSK4, MSK3} + #define ALTI_SL2_PERM {3,21,21,21,7,0,1,2,11,4,5,6,15,8,9,10} + #define ALTI_SL2_PERM64 {3,4,5,6,7,29,29,29,11,12,13,14,15,0,1,2} + #define ALTI_SR2_PERM {7,0,1,2,11,4,5,6,15,8,9,10,17,12,13,14} + #define ALTI_SR2_PERM64 {15,0,1,2,3,4,5,6,17,8,9,10,11,12,13,14} +#endif /* For OSX */ +#define IDSTR "SFMT-216091:627-11-3-10-1:bff7bff7-bfffffff-bffffa7f-ffddfbfb" + +#endif /* SFMT_PARAMS216091_H */ diff --git a/deps/jemalloc/test/include/test/SFMT-params2281.h b/deps/jemalloc/test/include/test/SFMT-params2281.h new file mode 100644 index 000000000..0ef85c407 --- /dev/null +++ b/deps/jemalloc/test/include/test/SFMT-params2281.h @@ -0,0 +1,81 @@ +/* + * This file derives from SFMT 1.3.3 + * (http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/SFMT/index.html), which was + * released under the terms of the following license: + * + * Copyright (c) 2006,2007 Mutsuo Saito, Makoto Matsumoto and Hiroshima + * University. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions are + * met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials provided + * with the distribution. + * * Neither the name of the Hiroshima University nor the names of + * its contributors may be used to endorse or promote products + * derived from this software without specific prior written + * permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ +#ifndef SFMT_PARAMS2281_H +#define SFMT_PARAMS2281_H + +#define POS1 12 +#define SL1 19 +#define SL2 1 +#define SR1 5 +#define SR2 1 +#define MSK1 0xbff7ffbfU +#define MSK2 0xfdfffffeU +#define MSK3 0xf7ffef7fU +#define MSK4 0xf2f7cbbfU +#define PARITY1 0x00000001U +#define PARITY2 0x00000000U +#define PARITY3 0x00000000U +#define PARITY4 0x41dfa600U + + +/* PARAMETERS FOR ALTIVEC */ +#if defined(__APPLE__) /* For OSX */ + #define ALTI_SL1 (vector unsigned int)(SL1, SL1, SL1, SL1) + #define ALTI_SR1 (vector unsigned int)(SR1, SR1, SR1, SR1) + #define ALTI_MSK (vector unsigned int)(MSK1, MSK2, MSK3, MSK4) + #define ALTI_MSK64 \ + (vector unsigned int)(MSK2, MSK1, MSK4, MSK3) + #define ALTI_SL2_PERM \ + (vector unsigned char)(1,2,3,23,5,6,7,0,9,10,11,4,13,14,15,8) + #define ALTI_SL2_PERM64 \ + (vector unsigned char)(1,2,3,4,5,6,7,31,9,10,11,12,13,14,15,0) + #define ALTI_SR2_PERM \ + (vector unsigned char)(7,0,1,2,11,4,5,6,15,8,9,10,17,12,13,14) + #define ALTI_SR2_PERM64 \ + (vector unsigned char)(15,0,1,2,3,4,5,6,17,8,9,10,11,12,13,14) +#else /* For OTHER OSs(Linux?) */ + #define ALTI_SL1 {SL1, SL1, SL1, SL1} + #define ALTI_SR1 {SR1, SR1, SR1, SR1} + #define ALTI_MSK {MSK1, MSK2, MSK3, MSK4} + #define ALTI_MSK64 {MSK2, MSK1, MSK4, MSK3} + #define ALTI_SL2_PERM {1,2,3,23,5,6,7,0,9,10,11,4,13,14,15,8} + #define ALTI_SL2_PERM64 {1,2,3,4,5,6,7,31,9,10,11,12,13,14,15,0} + #define ALTI_SR2_PERM {7,0,1,2,11,4,5,6,15,8,9,10,17,12,13,14} + #define ALTI_SR2_PERM64 {15,0,1,2,3,4,5,6,17,8,9,10,11,12,13,14} +#endif /* For OSX */ +#define IDSTR "SFMT-2281:12-19-1-5-1:bff7ffbf-fdfffffe-f7ffef7f-f2f7cbbf" + +#endif /* SFMT_PARAMS2281_H */ diff --git a/deps/jemalloc/test/include/test/SFMT-params4253.h b/deps/jemalloc/test/include/test/SFMT-params4253.h new file mode 100644 index 000000000..9f07bc67e --- /dev/null +++ b/deps/jemalloc/test/include/test/SFMT-params4253.h @@ -0,0 +1,81 @@ +/* + * This file derives from SFMT 1.3.3 + * (http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/SFMT/index.html), which was + * released under the terms of the following license: + * + * Copyright (c) 2006,2007 Mutsuo Saito, Makoto Matsumoto and Hiroshima + * University. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions are + * met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials provided + * with the distribution. + * * Neither the name of the Hiroshima University nor the names of + * its contributors may be used to endorse or promote products + * derived from this software without specific prior written + * permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ +#ifndef SFMT_PARAMS4253_H +#define SFMT_PARAMS4253_H + +#define POS1 17 +#define SL1 20 +#define SL2 1 +#define SR1 7 +#define SR2 1 +#define MSK1 0x9f7bffffU +#define MSK2 0x9fffff5fU +#define MSK3 0x3efffffbU +#define MSK4 0xfffff7bbU +#define PARITY1 0xa8000001U +#define PARITY2 0xaf5390a3U +#define PARITY3 0xb740b3f8U +#define PARITY4 0x6c11486dU + + +/* PARAMETERS FOR ALTIVEC */ +#if defined(__APPLE__) /* For OSX */ + #define ALTI_SL1 (vector unsigned int)(SL1, SL1, SL1, SL1) + #define ALTI_SR1 (vector unsigned int)(SR1, SR1, SR1, SR1) + #define ALTI_MSK (vector unsigned int)(MSK1, MSK2, MSK3, MSK4) + #define ALTI_MSK64 \ + (vector unsigned int)(MSK2, MSK1, MSK4, MSK3) + #define ALTI_SL2_PERM \ + (vector unsigned char)(1,2,3,23,5,6,7,0,9,10,11,4,13,14,15,8) + #define ALTI_SL2_PERM64 \ + (vector unsigned char)(1,2,3,4,5,6,7,31,9,10,11,12,13,14,15,0) + #define ALTI_SR2_PERM \ + (vector unsigned char)(7,0,1,2,11,4,5,6,15,8,9,10,17,12,13,14) + #define ALTI_SR2_PERM64 \ + (vector unsigned char)(15,0,1,2,3,4,5,6,17,8,9,10,11,12,13,14) +#else /* For OTHER OSs(Linux?) */ + #define ALTI_SL1 {SL1, SL1, SL1, SL1} + #define ALTI_SR1 {SR1, SR1, SR1, SR1} + #define ALTI_MSK {MSK1, MSK2, MSK3, MSK4} + #define ALTI_MSK64 {MSK2, MSK1, MSK4, MSK3} + #define ALTI_SL2_PERM {1,2,3,23,5,6,7,0,9,10,11,4,13,14,15,8} + #define ALTI_SL2_PERM64 {1,2,3,4,5,6,7,31,9,10,11,12,13,14,15,0} + #define ALTI_SR2_PERM {7,0,1,2,11,4,5,6,15,8,9,10,17,12,13,14} + #define ALTI_SR2_PERM64 {15,0,1,2,3,4,5,6,17,8,9,10,11,12,13,14} +#endif /* For OSX */ +#define IDSTR "SFMT-4253:17-20-1-7-1:9f7bffff-9fffff5f-3efffffb-fffff7bb" + +#endif /* SFMT_PARAMS4253_H */ diff --git a/deps/jemalloc/test/include/test/SFMT-params44497.h b/deps/jemalloc/test/include/test/SFMT-params44497.h new file mode 100644 index 000000000..85598fed5 --- /dev/null +++ b/deps/jemalloc/test/include/test/SFMT-params44497.h @@ -0,0 +1,81 @@ +/* + * This file derives from SFMT 1.3.3 + * (http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/SFMT/index.html), which was + * released under the terms of the following license: + * + * Copyright (c) 2006,2007 Mutsuo Saito, Makoto Matsumoto and Hiroshima + * University. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions are + * met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials provided + * with the distribution. + * * Neither the name of the Hiroshima University nor the names of + * its contributors may be used to endorse or promote products + * derived from this software without specific prior written + * permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ +#ifndef SFMT_PARAMS44497_H +#define SFMT_PARAMS44497_H + +#define POS1 330 +#define SL1 5 +#define SL2 3 +#define SR1 9 +#define SR2 3 +#define MSK1 0xeffffffbU +#define MSK2 0xdfbebfffU +#define MSK3 0xbfbf7befU +#define MSK4 0x9ffd7bffU +#define PARITY1 0x00000001U +#define PARITY2 0x00000000U +#define PARITY3 0xa3ac4000U +#define PARITY4 0xecc1327aU + + +/* PARAMETERS FOR ALTIVEC */ +#if defined(__APPLE__) /* For OSX */ + #define ALTI_SL1 (vector unsigned int)(SL1, SL1, SL1, SL1) + #define ALTI_SR1 (vector unsigned int)(SR1, SR1, SR1, SR1) + #define ALTI_MSK (vector unsigned int)(MSK1, MSK2, MSK3, MSK4) + #define ALTI_MSK64 \ + (vector unsigned int)(MSK2, MSK1, MSK4, MSK3) + #define ALTI_SL2_PERM \ + (vector unsigned char)(3,21,21,21,7,0,1,2,11,4,5,6,15,8,9,10) + #define ALTI_SL2_PERM64 \ + (vector unsigned char)(3,4,5,6,7,29,29,29,11,12,13,14,15,0,1,2) + #define ALTI_SR2_PERM \ + (vector unsigned char)(5,6,7,0,9,10,11,4,13,14,15,8,19,19,19,12) + #define ALTI_SR2_PERM64 \ + (vector unsigned char)(13,14,15,0,1,2,3,4,19,19,19,8,9,10,11,12) +#else /* For OTHER OSs(Linux?) */ + #define ALTI_SL1 {SL1, SL1, SL1, SL1} + #define ALTI_SR1 {SR1, SR1, SR1, SR1} + #define ALTI_MSK {MSK1, MSK2, MSK3, MSK4} + #define ALTI_MSK64 {MSK2, MSK1, MSK4, MSK3} + #define ALTI_SL2_PERM {3,21,21,21,7,0,1,2,11,4,5,6,15,8,9,10} + #define ALTI_SL2_PERM64 {3,4,5,6,7,29,29,29,11,12,13,14,15,0,1,2} + #define ALTI_SR2_PERM {5,6,7,0,9,10,11,4,13,14,15,8,19,19,19,12} + #define ALTI_SR2_PERM64 {13,14,15,0,1,2,3,4,19,19,19,8,9,10,11,12} +#endif /* For OSX */ +#define IDSTR "SFMT-44497:330-5-3-9-3:effffffb-dfbebfff-bfbf7bef-9ffd7bff" + +#endif /* SFMT_PARAMS44497_H */ diff --git a/deps/jemalloc/test/include/test/SFMT-params607.h b/deps/jemalloc/test/include/test/SFMT-params607.h new file mode 100644 index 000000000..bc76485f8 --- /dev/null +++ b/deps/jemalloc/test/include/test/SFMT-params607.h @@ -0,0 +1,81 @@ +/* + * This file derives from SFMT 1.3.3 + * (http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/SFMT/index.html), which was + * released under the terms of the following license: + * + * Copyright (c) 2006,2007 Mutsuo Saito, Makoto Matsumoto and Hiroshima + * University. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions are + * met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials provided + * with the distribution. + * * Neither the name of the Hiroshima University nor the names of + * its contributors may be used to endorse or promote products + * derived from this software without specific prior written + * permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ +#ifndef SFMT_PARAMS607_H +#define SFMT_PARAMS607_H + +#define POS1 2 +#define SL1 15 +#define SL2 3 +#define SR1 13 +#define SR2 3 +#define MSK1 0xfdff37ffU +#define MSK2 0xef7f3f7dU +#define MSK3 0xff777b7dU +#define MSK4 0x7ff7fb2fU +#define PARITY1 0x00000001U +#define PARITY2 0x00000000U +#define PARITY3 0x00000000U +#define PARITY4 0x5986f054U + + +/* PARAMETERS FOR ALTIVEC */ +#if defined(__APPLE__) /* For OSX */ + #define ALTI_SL1 (vector unsigned int)(SL1, SL1, SL1, SL1) + #define ALTI_SR1 (vector unsigned int)(SR1, SR1, SR1, SR1) + #define ALTI_MSK (vector unsigned int)(MSK1, MSK2, MSK3, MSK4) + #define ALTI_MSK64 \ + (vector unsigned int)(MSK2, MSK1, MSK4, MSK3) + #define ALTI_SL2_PERM \ + (vector unsigned char)(3,21,21,21,7,0,1,2,11,4,5,6,15,8,9,10) + #define ALTI_SL2_PERM64 \ + (vector unsigned char)(3,4,5,6,7,29,29,29,11,12,13,14,15,0,1,2) + #define ALTI_SR2_PERM \ + (vector unsigned char)(5,6,7,0,9,10,11,4,13,14,15,8,19,19,19,12) + #define ALTI_SR2_PERM64 \ + (vector unsigned char)(13,14,15,0,1,2,3,4,19,19,19,8,9,10,11,12) +#else /* For OTHER OSs(Linux?) */ + #define ALTI_SL1 {SL1, SL1, SL1, SL1} + #define ALTI_SR1 {SR1, SR1, SR1, SR1} + #define ALTI_MSK {MSK1, MSK2, MSK3, MSK4} + #define ALTI_MSK64 {MSK2, MSK1, MSK4, MSK3} + #define ALTI_SL2_PERM {3,21,21,21,7,0,1,2,11,4,5,6,15,8,9,10} + #define ALTI_SL2_PERM64 {3,4,5,6,7,29,29,29,11,12,13,14,15,0,1,2} + #define ALTI_SR2_PERM {5,6,7,0,9,10,11,4,13,14,15,8,19,19,19,12} + #define ALTI_SR2_PERM64 {13,14,15,0,1,2,3,4,19,19,19,8,9,10,11,12} +#endif /* For OSX */ +#define IDSTR "SFMT-607:2-15-3-13-3:fdff37ff-ef7f3f7d-ff777b7d-7ff7fb2f" + +#endif /* SFMT_PARAMS607_H */ diff --git a/deps/jemalloc/test/include/test/SFMT-params86243.h b/deps/jemalloc/test/include/test/SFMT-params86243.h new file mode 100644 index 000000000..5e4d783c5 --- /dev/null +++ b/deps/jemalloc/test/include/test/SFMT-params86243.h @@ -0,0 +1,81 @@ +/* + * This file derives from SFMT 1.3.3 + * (http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/SFMT/index.html), which was + * released under the terms of the following license: + * + * Copyright (c) 2006,2007 Mutsuo Saito, Makoto Matsumoto and Hiroshima + * University. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions are + * met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials provided + * with the distribution. + * * Neither the name of the Hiroshima University nor the names of + * its contributors may be used to endorse or promote products + * derived from this software without specific prior written + * permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ +#ifndef SFMT_PARAMS86243_H +#define SFMT_PARAMS86243_H + +#define POS1 366 +#define SL1 6 +#define SL2 7 +#define SR1 19 +#define SR2 1 +#define MSK1 0xfdbffbffU +#define MSK2 0xbff7ff3fU +#define MSK3 0xfd77efffU +#define MSK4 0xbf9ff3ffU +#define PARITY1 0x00000001U +#define PARITY2 0x00000000U +#define PARITY3 0x00000000U +#define PARITY4 0xe9528d85U + + +/* PARAMETERS FOR ALTIVEC */ +#if defined(__APPLE__) /* For OSX */ + #define ALTI_SL1 (vector unsigned int)(SL1, SL1, SL1, SL1) + #define ALTI_SR1 (vector unsigned int)(SR1, SR1, SR1, SR1) + #define ALTI_MSK (vector unsigned int)(MSK1, MSK2, MSK3, MSK4) + #define ALTI_MSK64 \ + (vector unsigned int)(MSK2, MSK1, MSK4, MSK3) + #define ALTI_SL2_PERM \ + (vector unsigned char)(25,25,25,25,3,25,25,25,7,0,1,2,11,4,5,6) + #define ALTI_SL2_PERM64 \ + (vector unsigned char)(7,25,25,25,25,25,25,25,15,0,1,2,3,4,5,6) + #define ALTI_SR2_PERM \ + (vector unsigned char)(7,0,1,2,11,4,5,6,15,8,9,10,17,12,13,14) + #define ALTI_SR2_PERM64 \ + (vector unsigned char)(15,0,1,2,3,4,5,6,17,8,9,10,11,12,13,14) +#else /* For OTHER OSs(Linux?) */ + #define ALTI_SL1 {SL1, SL1, SL1, SL1} + #define ALTI_SR1 {SR1, SR1, SR1, SR1} + #define ALTI_MSK {MSK1, MSK2, MSK3, MSK4} + #define ALTI_MSK64 {MSK2, MSK1, MSK4, MSK3} + #define ALTI_SL2_PERM {25,25,25,25,3,25,25,25,7,0,1,2,11,4,5,6} + #define ALTI_SL2_PERM64 {7,25,25,25,25,25,25,25,15,0,1,2,3,4,5,6} + #define ALTI_SR2_PERM {7,0,1,2,11,4,5,6,15,8,9,10,17,12,13,14} + #define ALTI_SR2_PERM64 {15,0,1,2,3,4,5,6,17,8,9,10,11,12,13,14} +#endif /* For OSX */ +#define IDSTR "SFMT-86243:366-6-7-19-1:fdbffbff-bff7ff3f-fd77efff-bf9ff3ff" + +#endif /* SFMT_PARAMS86243_H */ diff --git a/deps/jemalloc/test/include/test/SFMT-sse2.h b/deps/jemalloc/test/include/test/SFMT-sse2.h new file mode 100644 index 000000000..169ad5581 --- /dev/null +++ b/deps/jemalloc/test/include/test/SFMT-sse2.h @@ -0,0 +1,157 @@ +/* + * This file derives from SFMT 1.3.3 + * (http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/SFMT/index.html), which was + * released under the terms of the following license: + * + * Copyright (c) 2006,2007 Mutsuo Saito, Makoto Matsumoto and Hiroshima + * University. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions are + * met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials provided + * with the distribution. + * * Neither the name of the Hiroshima University nor the names of + * its contributors may be used to endorse or promote products + * derived from this software without specific prior written + * permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ +/** + * @file SFMT-sse2.h + * @brief SIMD oriented Fast Mersenne Twister(SFMT) for Intel SSE2 + * + * @author Mutsuo Saito (Hiroshima University) + * @author Makoto Matsumoto (Hiroshima University) + * + * @note We assume LITTLE ENDIAN in this file + * + * Copyright (C) 2006, 2007 Mutsuo Saito, Makoto Matsumoto and Hiroshima + * University. All rights reserved. + * + * The new BSD License is applied to this software, see LICENSE.txt + */ + +#ifndef SFMT_SSE2_H +#define SFMT_SSE2_H + +/** + * This function represents the recursion formula. + * @param a a 128-bit part of the interal state array + * @param b a 128-bit part of the interal state array + * @param c a 128-bit part of the interal state array + * @param d a 128-bit part of the interal state array + * @param mask 128-bit mask + * @return output + */ +JEMALLOC_ALWAYS_INLINE __m128i mm_recursion(__m128i *a, __m128i *b, + __m128i c, __m128i d, __m128i mask) { + __m128i v, x, y, z; + + x = _mm_load_si128(a); + y = _mm_srli_epi32(*b, SR1); + z = _mm_srli_si128(c, SR2); + v = _mm_slli_epi32(d, SL1); + z = _mm_xor_si128(z, x); + z = _mm_xor_si128(z, v); + x = _mm_slli_si128(x, SL2); + y = _mm_and_si128(y, mask); + z = _mm_xor_si128(z, x); + z = _mm_xor_si128(z, y); + return z; +} + +/** + * This function fills the internal state array with pseudorandom + * integers. + */ +static inline void gen_rand_all(sfmt_t *ctx) { + int i; + __m128i r, r1, r2, mask; + mask = _mm_set_epi32(MSK4, MSK3, MSK2, MSK1); + + r1 = _mm_load_si128(&ctx->sfmt[N - 2].si); + r2 = _mm_load_si128(&ctx->sfmt[N - 1].si); + for (i = 0; i < N - POS1; i++) { + r = mm_recursion(&ctx->sfmt[i].si, &ctx->sfmt[i + POS1].si, r1, r2, + mask); + _mm_store_si128(&ctx->sfmt[i].si, r); + r1 = r2; + r2 = r; + } + for (; i < N; i++) { + r = mm_recursion(&ctx->sfmt[i].si, &ctx->sfmt[i + POS1 - N].si, r1, r2, + mask); + _mm_store_si128(&ctx->sfmt[i].si, r); + r1 = r2; + r2 = r; + } +} + +/** + * This function fills the user-specified array with pseudorandom + * integers. + * + * @param array an 128-bit array to be filled by pseudorandom numbers. + * @param size number of 128-bit pesudorandom numbers to be generated. + */ +static inline void gen_rand_array(sfmt_t *ctx, w128_t *array, int size) { + int i, j; + __m128i r, r1, r2, mask; + mask = _mm_set_epi32(MSK4, MSK3, MSK2, MSK1); + + r1 = _mm_load_si128(&ctx->sfmt[N - 2].si); + r2 = _mm_load_si128(&ctx->sfmt[N - 1].si); + for (i = 0; i < N - POS1; i++) { + r = mm_recursion(&ctx->sfmt[i].si, &ctx->sfmt[i + POS1].si, r1, r2, + mask); + _mm_store_si128(&array[i].si, r); + r1 = r2; + r2 = r; + } + for (; i < N; i++) { + r = mm_recursion(&ctx->sfmt[i].si, &array[i + POS1 - N].si, r1, r2, + mask); + _mm_store_si128(&array[i].si, r); + r1 = r2; + r2 = r; + } + /* main loop */ + for (; i < size - N; i++) { + r = mm_recursion(&array[i - N].si, &array[i + POS1 - N].si, r1, r2, + mask); + _mm_store_si128(&array[i].si, r); + r1 = r2; + r2 = r; + } + for (j = 0; j < 2 * N - size; j++) { + r = _mm_load_si128(&array[j + size - N].si); + _mm_store_si128(&ctx->sfmt[j].si, r); + } + for (; i < size; i++) { + r = mm_recursion(&array[i - N].si, &array[i + POS1 - N].si, r1, r2, + mask); + _mm_store_si128(&array[i].si, r); + _mm_store_si128(&ctx->sfmt[j++].si, r); + r1 = r2; + r2 = r; + } +} + +#endif diff --git a/deps/jemalloc/test/include/test/SFMT.h b/deps/jemalloc/test/include/test/SFMT.h new file mode 100644 index 000000000..863fc55e8 --- /dev/null +++ b/deps/jemalloc/test/include/test/SFMT.h @@ -0,0 +1,146 @@ +/* + * This file derives from SFMT 1.3.3 + * (http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/SFMT/index.html), which was + * released under the terms of the following license: + * + * Copyright (c) 2006,2007 Mutsuo Saito, Makoto Matsumoto and Hiroshima + * University. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions are + * met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials provided + * with the distribution. + * * Neither the name of the Hiroshima University nor the names of + * its contributors may be used to endorse or promote products + * derived from this software without specific prior written + * permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ +/** + * @file SFMT.h + * + * @brief SIMD oriented Fast Mersenne Twister(SFMT) pseudorandom + * number generator + * + * @author Mutsuo Saito (Hiroshima University) + * @author Makoto Matsumoto (Hiroshima University) + * + * Copyright (C) 2006, 2007 Mutsuo Saito, Makoto Matsumoto and Hiroshima + * University. All rights reserved. + * + * The new BSD License is applied to this software. + * see LICENSE.txt + * + * @note We assume that your system has inttypes.h. If your system + * doesn't have inttypes.h, you have to typedef uint32_t and uint64_t, + * and you have to define PRIu64 and PRIx64 in this file as follows: + * @verbatim + typedef unsigned int uint32_t + typedef unsigned long long uint64_t + #define PRIu64 "llu" + #define PRIx64 "llx" +@endverbatim + * uint32_t must be exactly 32-bit unsigned integer type (no more, no + * less), and uint64_t must be exactly 64-bit unsigned integer type. + * PRIu64 and PRIx64 are used for printf function to print 64-bit + * unsigned int and 64-bit unsigned int in hexadecimal format. + */ + +#ifndef SFMT_H +#define SFMT_H + +typedef struct sfmt_s sfmt_t; + +uint32_t gen_rand32(sfmt_t *ctx); +uint32_t gen_rand32_range(sfmt_t *ctx, uint32_t limit); +uint64_t gen_rand64(sfmt_t *ctx); +uint64_t gen_rand64_range(sfmt_t *ctx, uint64_t limit); +void fill_array32(sfmt_t *ctx, uint32_t *array, int size); +void fill_array64(sfmt_t *ctx, uint64_t *array, int size); +sfmt_t *init_gen_rand(uint32_t seed); +sfmt_t *init_by_array(uint32_t *init_key, int key_length); +void fini_gen_rand(sfmt_t *ctx); +const char *get_idstring(void); +int get_min_array_size32(void); +int get_min_array_size64(void); + +/* These real versions are due to Isaku Wada */ +/** generates a random number on [0,1]-real-interval */ +static inline double to_real1(uint32_t v) { + return v * (1.0/4294967295.0); + /* divided by 2^32-1 */ +} + +/** generates a random number on [0,1]-real-interval */ +static inline double genrand_real1(sfmt_t *ctx) { + return to_real1(gen_rand32(ctx)); +} + +/** generates a random number on [0,1)-real-interval */ +static inline double to_real2(uint32_t v) { + return v * (1.0/4294967296.0); + /* divided by 2^32 */ +} + +/** generates a random number on [0,1)-real-interval */ +static inline double genrand_real2(sfmt_t *ctx) { + return to_real2(gen_rand32(ctx)); +} + +/** generates a random number on (0,1)-real-interval */ +static inline double to_real3(uint32_t v) { + return (((double)v) + 0.5)*(1.0/4294967296.0); + /* divided by 2^32 */ +} + +/** generates a random number on (0,1)-real-interval */ +static inline double genrand_real3(sfmt_t *ctx) { + return to_real3(gen_rand32(ctx)); +} +/** These real versions are due to Isaku Wada */ + +/** generates a random number on [0,1) with 53-bit resolution*/ +static inline double to_res53(uint64_t v) { + return v * (1.0/18446744073709551616.0L); +} + +/** generates a random number on [0,1) with 53-bit resolution from two + * 32 bit integers */ +static inline double to_res53_mix(uint32_t x, uint32_t y) { + return to_res53(x | ((uint64_t)y << 32)); +} + +/** generates a random number on [0,1) with 53-bit resolution + */ +static inline double genrand_res53(sfmt_t *ctx) { + return to_res53(gen_rand64(ctx)); +} + +/** generates a random number on [0,1) with 53-bit resolution + using 32bit integer. + */ +static inline double genrand_res53_mix(sfmt_t *ctx) { + uint32_t x, y; + + x = gen_rand32(ctx); + y = gen_rand32(ctx); + return to_res53_mix(x, y); +} +#endif diff --git a/deps/jemalloc/test/include/test/btalloc.h b/deps/jemalloc/test/include/test/btalloc.h new file mode 100644 index 000000000..5877ea77e --- /dev/null +++ b/deps/jemalloc/test/include/test/btalloc.h @@ -0,0 +1,30 @@ +/* btalloc() provides a mechanism for allocating via permuted backtraces. */ +void *btalloc(size_t size, unsigned bits); + +#define btalloc_n_proto(n) \ +void *btalloc_##n(size_t size, unsigned bits); +btalloc_n_proto(0) +btalloc_n_proto(1) + +#define btalloc_n_gen(n) \ +void * \ +btalloc_##n(size_t size, unsigned bits) { \ + void *p; \ + \ + if (bits == 0) { \ + p = mallocx(size, 0); \ + } else { \ + switch (bits & 0x1U) { \ + case 0: \ + p = (btalloc_0(size, bits >> 1)); \ + break; \ + case 1: \ + p = (btalloc_1(size, bits >> 1)); \ + break; \ + default: not_reached(); \ + } \ + } \ + /* Intentionally sabotage tail call optimization. */ \ + assert_ptr_not_null(p, "Unexpected mallocx() failure"); \ + return p; \ +} diff --git a/deps/jemalloc/test/include/test/extent_hooks.h b/deps/jemalloc/test/include/test/extent_hooks.h new file mode 100644 index 000000000..1f0620154 --- /dev/null +++ b/deps/jemalloc/test/include/test/extent_hooks.h @@ -0,0 +1,289 @@ +/* + * Boilerplate code used for testing extent hooks via interception and + * passthrough. + */ + +static void *extent_alloc_hook(extent_hooks_t *extent_hooks, void *new_addr, + size_t size, size_t alignment, bool *zero, bool *commit, + unsigned arena_ind); +static bool extent_dalloc_hook(extent_hooks_t *extent_hooks, void *addr, + size_t size, bool committed, unsigned arena_ind); +static void extent_destroy_hook(extent_hooks_t *extent_hooks, void *addr, + size_t size, bool committed, unsigned arena_ind); +static bool extent_commit_hook(extent_hooks_t *extent_hooks, void *addr, + size_t size, size_t offset, size_t length, unsigned arena_ind); +static bool extent_decommit_hook(extent_hooks_t *extent_hooks, void *addr, + size_t size, size_t offset, size_t length, unsigned arena_ind); +static bool extent_purge_lazy_hook(extent_hooks_t *extent_hooks, void *addr, + size_t size, size_t offset, size_t length, unsigned arena_ind); +static bool extent_purge_forced_hook(extent_hooks_t *extent_hooks, + void *addr, size_t size, size_t offset, size_t length, unsigned arena_ind); +static bool extent_split_hook(extent_hooks_t *extent_hooks, void *addr, + size_t size, size_t size_a, size_t size_b, bool committed, + unsigned arena_ind); +static bool extent_merge_hook(extent_hooks_t *extent_hooks, void *addr_a, + size_t size_a, void *addr_b, size_t size_b, bool committed, + unsigned arena_ind); + +static extent_hooks_t *default_hooks; +static extent_hooks_t hooks = { + extent_alloc_hook, + extent_dalloc_hook, + extent_destroy_hook, + extent_commit_hook, + extent_decommit_hook, + extent_purge_lazy_hook, + extent_purge_forced_hook, + extent_split_hook, + extent_merge_hook +}; + +/* Control whether hook functions pass calls through to default hooks. */ +static bool try_alloc = true; +static bool try_dalloc = true; +static bool try_destroy = true; +static bool try_commit = true; +static bool try_decommit = true; +static bool try_purge_lazy = true; +static bool try_purge_forced = true; +static bool try_split = true; +static bool try_merge = true; + +/* Set to false prior to operations, then introspect after operations. */ +static bool called_alloc; +static bool called_dalloc; +static bool called_destroy; +static bool called_commit; +static bool called_decommit; +static bool called_purge_lazy; +static bool called_purge_forced; +static bool called_split; +static bool called_merge; + +/* Set to false prior to operations, then introspect after operations. */ +static bool did_alloc; +static bool did_dalloc; +static bool did_destroy; +static bool did_commit; +static bool did_decommit; +static bool did_purge_lazy; +static bool did_purge_forced; +static bool did_split; +static bool did_merge; + +#if 0 +# define TRACE_HOOK(fmt, ...) malloc_printf(fmt, __VA_ARGS__) +#else +# define TRACE_HOOK(fmt, ...) +#endif + +static void * +extent_alloc_hook(extent_hooks_t *extent_hooks, void *new_addr, size_t size, + size_t alignment, bool *zero, bool *commit, unsigned arena_ind) { + void *ret; + + TRACE_HOOK("%s(extent_hooks=%p, new_addr=%p, size=%zu, alignment=%zu, " + "*zero=%s, *commit=%s, arena_ind=%u)\n", __func__, extent_hooks, + new_addr, size, alignment, *zero ? "true" : "false", *commit ? + "true" : "false", arena_ind); + assert_ptr_eq(extent_hooks, &hooks, + "extent_hooks should be same as pointer used to set hooks"); + assert_ptr_eq(extent_hooks->alloc, extent_alloc_hook, + "Wrong hook function"); + called_alloc = true; + if (!try_alloc) { + return NULL; + } + ret = default_hooks->alloc(default_hooks, new_addr, size, alignment, + zero, commit, 0); + did_alloc = (ret != NULL); + return ret; +} + +static bool +extent_dalloc_hook(extent_hooks_t *extent_hooks, void *addr, size_t size, + bool committed, unsigned arena_ind) { + bool err; + + TRACE_HOOK("%s(extent_hooks=%p, addr=%p, size=%zu, committed=%s, " + "arena_ind=%u)\n", __func__, extent_hooks, addr, size, committed ? + "true" : "false", arena_ind); + assert_ptr_eq(extent_hooks, &hooks, + "extent_hooks should be same as pointer used to set hooks"); + assert_ptr_eq(extent_hooks->dalloc, extent_dalloc_hook, + "Wrong hook function"); + called_dalloc = true; + if (!try_dalloc) { + return true; + } + err = default_hooks->dalloc(default_hooks, addr, size, committed, 0); + did_dalloc = !err; + return err; +} + +static void +extent_destroy_hook(extent_hooks_t *extent_hooks, void *addr, size_t size, + bool committed, unsigned arena_ind) { + TRACE_HOOK("%s(extent_hooks=%p, addr=%p, size=%zu, committed=%s, " + "arena_ind=%u)\n", __func__, extent_hooks, addr, size, committed ? + "true" : "false", arena_ind); + assert_ptr_eq(extent_hooks, &hooks, + "extent_hooks should be same as pointer used to set hooks"); + assert_ptr_eq(extent_hooks->destroy, extent_destroy_hook, + "Wrong hook function"); + called_destroy = true; + if (!try_destroy) { + return; + } + default_hooks->destroy(default_hooks, addr, size, committed, 0); + did_destroy = true; +} + +static bool +extent_commit_hook(extent_hooks_t *extent_hooks, void *addr, size_t size, + size_t offset, size_t length, unsigned arena_ind) { + bool err; + + TRACE_HOOK("%s(extent_hooks=%p, addr=%p, size=%zu, offset=%zu, " + "length=%zu, arena_ind=%u)\n", __func__, extent_hooks, addr, size, + offset, length, arena_ind); + assert_ptr_eq(extent_hooks, &hooks, + "extent_hooks should be same as pointer used to set hooks"); + assert_ptr_eq(extent_hooks->commit, extent_commit_hook, + "Wrong hook function"); + called_commit = true; + if (!try_commit) { + return true; + } + err = default_hooks->commit(default_hooks, addr, size, offset, length, + 0); + did_commit = !err; + return err; +} + +static bool +extent_decommit_hook(extent_hooks_t *extent_hooks, void *addr, size_t size, + size_t offset, size_t length, unsigned arena_ind) { + bool err; + + TRACE_HOOK("%s(extent_hooks=%p, addr=%p, size=%zu, offset=%zu, " + "length=%zu, arena_ind=%u)\n", __func__, extent_hooks, addr, size, + offset, length, arena_ind); + assert_ptr_eq(extent_hooks, &hooks, + "extent_hooks should be same as pointer used to set hooks"); + assert_ptr_eq(extent_hooks->decommit, extent_decommit_hook, + "Wrong hook function"); + called_decommit = true; + if (!try_decommit) { + return true; + } + err = default_hooks->decommit(default_hooks, addr, size, offset, length, + 0); + did_decommit = !err; + return err; +} + +static bool +extent_purge_lazy_hook(extent_hooks_t *extent_hooks, void *addr, size_t size, + size_t offset, size_t length, unsigned arena_ind) { + bool err; + + TRACE_HOOK("%s(extent_hooks=%p, addr=%p, size=%zu, offset=%zu, " + "length=%zu arena_ind=%u)\n", __func__, extent_hooks, addr, size, + offset, length, arena_ind); + assert_ptr_eq(extent_hooks, &hooks, + "extent_hooks should be same as pointer used to set hooks"); + assert_ptr_eq(extent_hooks->purge_lazy, extent_purge_lazy_hook, + "Wrong hook function"); + called_purge_lazy = true; + if (!try_purge_lazy) { + return true; + } + err = default_hooks->purge_lazy == NULL || + default_hooks->purge_lazy(default_hooks, addr, size, offset, length, + 0); + did_purge_lazy = !err; + return err; +} + +static bool +extent_purge_forced_hook(extent_hooks_t *extent_hooks, void *addr, size_t size, + size_t offset, size_t length, unsigned arena_ind) { + bool err; + + TRACE_HOOK("%s(extent_hooks=%p, addr=%p, size=%zu, offset=%zu, " + "length=%zu arena_ind=%u)\n", __func__, extent_hooks, addr, size, + offset, length, arena_ind); + assert_ptr_eq(extent_hooks, &hooks, + "extent_hooks should be same as pointer used to set hooks"); + assert_ptr_eq(extent_hooks->purge_forced, extent_purge_forced_hook, + "Wrong hook function"); + called_purge_forced = true; + if (!try_purge_forced) { + return true; + } + err = default_hooks->purge_forced == NULL || + default_hooks->purge_forced(default_hooks, addr, size, offset, + length, 0); + did_purge_forced = !err; + return err; +} + +static bool +extent_split_hook(extent_hooks_t *extent_hooks, void *addr, size_t size, + size_t size_a, size_t size_b, bool committed, unsigned arena_ind) { + bool err; + + TRACE_HOOK("%s(extent_hooks=%p, addr=%p, size=%zu, size_a=%zu, " + "size_b=%zu, committed=%s, arena_ind=%u)\n", __func__, extent_hooks, + addr, size, size_a, size_b, committed ? "true" : "false", + arena_ind); + assert_ptr_eq(extent_hooks, &hooks, + "extent_hooks should be same as pointer used to set hooks"); + assert_ptr_eq(extent_hooks->split, extent_split_hook, + "Wrong hook function"); + called_split = true; + if (!try_split) { + return true; + } + err = (default_hooks->split == NULL || + default_hooks->split(default_hooks, addr, size, size_a, size_b, + committed, 0)); + did_split = !err; + return err; +} + +static bool +extent_merge_hook(extent_hooks_t *extent_hooks, void *addr_a, size_t size_a, + void *addr_b, size_t size_b, bool committed, unsigned arena_ind) { + bool err; + + TRACE_HOOK("%s(extent_hooks=%p, addr_a=%p, size_a=%zu, addr_b=%p " + "size_b=%zu, committed=%s, arena_ind=%u)\n", __func__, extent_hooks, + addr_a, size_a, addr_b, size_b, committed ? "true" : "false", + arena_ind); + assert_ptr_eq(extent_hooks, &hooks, + "extent_hooks should be same as pointer used to set hooks"); + assert_ptr_eq(extent_hooks->merge, extent_merge_hook, + "Wrong hook function"); + assert_ptr_eq((void *)((uintptr_t)addr_a + size_a), addr_b, + "Extents not mergeable"); + called_merge = true; + if (!try_merge) { + return true; + } + err = (default_hooks->merge == NULL || + default_hooks->merge(default_hooks, addr_a, size_a, addr_b, size_b, + committed, 0)); + did_merge = !err; + return err; +} + +static void +extent_hooks_prep(void) { + size_t sz; + + sz = sizeof(default_hooks); + assert_d_eq(mallctl("arena.0.extent_hooks", (void *)&default_hooks, &sz, + NULL, 0), 0, "Unexpected mallctl() error"); +} diff --git a/deps/jemalloc/test/include/test/jemalloc_test.h.in b/deps/jemalloc/test/include/test/jemalloc_test.h.in new file mode 100644 index 000000000..c46af5d9b --- /dev/null +++ b/deps/jemalloc/test/include/test/jemalloc_test.h.in @@ -0,0 +1,173 @@ +#ifdef __cplusplus +extern "C" { +#endif + +#include <limits.h> +#ifndef SIZE_T_MAX +# define SIZE_T_MAX SIZE_MAX +#endif +#include <stdlib.h> +#include <stdarg.h> +#include <stdbool.h> +#include <errno.h> +#include <math.h> +#include <string.h> +#ifdef _WIN32 +# include "msvc_compat/strings.h" +#endif + +#ifdef _WIN32 +# include <windows.h> +# include "msvc_compat/windows_extra.h" +#else +# include <pthread.h> +#endif + +#include "test/jemalloc_test_defs.h" + +#if defined(JEMALLOC_OSATOMIC) +# include <libkern/OSAtomic.h> +#endif + +#if defined(HAVE_ALTIVEC) && !defined(__APPLE__) +# include <altivec.h> +#endif +#ifdef HAVE_SSE2 +# include <emmintrin.h> +#endif + +/******************************************************************************/ +/* + * For unit tests, expose all public and private interfaces. + */ +#ifdef JEMALLOC_UNIT_TEST +# define JEMALLOC_JET +# define JEMALLOC_MANGLE +# include "jemalloc/internal/jemalloc_preamble.h" +# include "jemalloc/internal/jemalloc_internal_includes.h" + +/******************************************************************************/ +/* + * For integration tests, expose the public jemalloc interfaces, but only + * expose the minimum necessary internal utility code (to avoid re-implementing + * essentially identical code within the test infrastructure). + */ +#elif defined(JEMALLOC_INTEGRATION_TEST) || \ + defined(JEMALLOC_INTEGRATION_CPP_TEST) +# define JEMALLOC_MANGLE +# include "jemalloc/jemalloc@install_suffix@.h" +# include "jemalloc/internal/jemalloc_internal_defs.h" +# include "jemalloc/internal/jemalloc_internal_macros.h" + +static const bool config_debug = +#ifdef JEMALLOC_DEBUG + true +#else + false +#endif + ; + +# define JEMALLOC_N(n) @private_namespace@##n +# include "jemalloc/internal/private_namespace.h" +# include "jemalloc/internal/test_hooks.h" + +/* Hermetic headers. */ +# include "jemalloc/internal/assert.h" +# include "jemalloc/internal/malloc_io.h" +# include "jemalloc/internal/nstime.h" +# include "jemalloc/internal/util.h" + +/* Non-hermetic headers. */ +# include "jemalloc/internal/qr.h" +# include "jemalloc/internal/ql.h" + +/******************************************************************************/ +/* + * For stress tests, expose the public jemalloc interfaces with name mangling + * so that they can be tested as e.g. malloc() and free(). Also expose the + * public jemalloc interfaces with jet_ prefixes, so that stress tests can use + * a separate allocator for their internal data structures. + */ +#elif defined(JEMALLOC_STRESS_TEST) +# include "jemalloc/jemalloc@install_suffix@.h" + +# include "jemalloc/jemalloc_protos_jet.h" + +# define JEMALLOC_JET +# include "jemalloc/internal/jemalloc_preamble.h" +# include "jemalloc/internal/jemalloc_internal_includes.h" +# include "jemalloc/internal/public_unnamespace.h" +# undef JEMALLOC_JET + +# include "jemalloc/jemalloc_rename.h" +# define JEMALLOC_MANGLE +# ifdef JEMALLOC_STRESS_TESTLIB +# include "jemalloc/jemalloc_mangle_jet.h" +# else +# include "jemalloc/jemalloc_mangle.h" +# endif + +/******************************************************************************/ +/* + * This header does dangerous things, the effects of which only test code + * should be subject to. + */ +#else +# error "This header cannot be included outside a testing context" +#endif + +/******************************************************************************/ +/* + * Common test utilities. + */ +#include "test/btalloc.h" +#include "test/math.h" +#include "test/mtx.h" +#include "test/mq.h" +#include "test/test.h" +#include "test/timer.h" +#include "test/thd.h" +#define MEXP 19937 +#include "test/SFMT.h" + +/******************************************************************************/ +/* + * Define always-enabled assertion macros, so that test assertions execute even + * if assertions are disabled in the library code. + */ +#undef assert +#undef not_reached +#undef not_implemented +#undef assert_not_implemented + +#define assert(e) do { \ + if (!(e)) { \ + malloc_printf( \ + "<jemalloc>: %s:%d: Failed assertion: \"%s\"\n", \ + __FILE__, __LINE__, #e); \ + abort(); \ + } \ +} while (0) + +#define not_reached() do { \ + malloc_printf( \ + "<jemalloc>: %s:%d: Unreachable code reached\n", \ + __FILE__, __LINE__); \ + abort(); \ +} while (0) + +#define not_implemented() do { \ + malloc_printf("<jemalloc>: %s:%d: Not implemented\n", \ + __FILE__, __LINE__); \ + abort(); \ +} while (0) + +#define assert_not_implemented(e) do { \ + if (!(e)) { \ + not_implemented(); \ + } \ +} while (0) + +#ifdef __cplusplus +} +#endif diff --git a/deps/jemalloc/test/include/test/jemalloc_test_defs.h.in b/deps/jemalloc/test/include/test/jemalloc_test_defs.h.in new file mode 100644 index 000000000..5cc8532a3 --- /dev/null +++ b/deps/jemalloc/test/include/test/jemalloc_test_defs.h.in @@ -0,0 +1,9 @@ +#include "jemalloc/internal/jemalloc_internal_defs.h" +#include "jemalloc/internal/jemalloc_internal_decls.h" + +/* + * For use by SFMT. configure.ac doesn't actually define HAVE_SSE2 because its + * dependencies are notoriously unportable in practice. + */ +#undef HAVE_SSE2 +#undef HAVE_ALTIVEC diff --git a/deps/jemalloc/test/include/test/math.h b/deps/jemalloc/test/include/test/math.h new file mode 100644 index 000000000..efba086dd --- /dev/null +++ b/deps/jemalloc/test/include/test/math.h @@ -0,0 +1,306 @@ +/* + * Compute the natural log of Gamma(x), accurate to 10 decimal places. + * + * This implementation is based on: + * + * Pike, M.C., I.D. Hill (1966) Algorithm 291: Logarithm of Gamma function + * [S14]. Communications of the ACM 9(9):684. + */ +static inline double +ln_gamma(double x) { + double f, z; + + assert(x > 0.0); + + if (x < 7.0) { + f = 1.0; + z = x; + while (z < 7.0) { + f *= z; + z += 1.0; + } + x = z; + f = -log(f); + } else { + f = 0.0; + } + + z = 1.0 / (x * x); + + return f + (x-0.5) * log(x) - x + 0.918938533204673 + + (((-0.000595238095238 * z + 0.000793650793651) * z - + 0.002777777777778) * z + 0.083333333333333) / x; +} + +/* + * Compute the incomplete Gamma ratio for [0..x], where p is the shape + * parameter, and ln_gamma_p is ln_gamma(p). + * + * This implementation is based on: + * + * Bhattacharjee, G.P. (1970) Algorithm AS 32: The incomplete Gamma integral. + * Applied Statistics 19:285-287. + */ +static inline double +i_gamma(double x, double p, double ln_gamma_p) { + double acu, factor, oflo, gin, term, rn, a, b, an, dif; + double pn[6]; + unsigned i; + + assert(p > 0.0); + assert(x >= 0.0); + + if (x == 0.0) { + return 0.0; + } + + acu = 1.0e-10; + oflo = 1.0e30; + gin = 0.0; + factor = exp(p * log(x) - x - ln_gamma_p); + + if (x <= 1.0 || x < p) { + /* Calculation by series expansion. */ + gin = 1.0; + term = 1.0; + rn = p; + + while (true) { + rn += 1.0; + term *= x / rn; + gin += term; + if (term <= acu) { + gin *= factor / p; + return gin; + } + } + } else { + /* Calculation by continued fraction. */ + a = 1.0 - p; + b = a + x + 1.0; + term = 0.0; + pn[0] = 1.0; + pn[1] = x; + pn[2] = x + 1.0; + pn[3] = x * b; + gin = pn[2] / pn[3]; + + while (true) { + a += 1.0; + b += 2.0; + term += 1.0; + an = a * term; + for (i = 0; i < 2; i++) { + pn[i+4] = b * pn[i+2] - an * pn[i]; + } + if (pn[5] != 0.0) { + rn = pn[4] / pn[5]; + dif = fabs(gin - rn); + if (dif <= acu && dif <= acu * rn) { + gin = 1.0 - factor * gin; + return gin; + } + gin = rn; + } + for (i = 0; i < 4; i++) { + pn[i] = pn[i+2]; + } + + if (fabs(pn[4]) >= oflo) { + for (i = 0; i < 4; i++) { + pn[i] /= oflo; + } + } + } + } +} + +/* + * Given a value p in [0..1] of the lower tail area of the normal distribution, + * compute the limit on the definite integral from [-inf..z] that satisfies p, + * accurate to 16 decimal places. + * + * This implementation is based on: + * + * Wichura, M.J. (1988) Algorithm AS 241: The percentage points of the normal + * distribution. Applied Statistics 37(3):477-484. + */ +static inline double +pt_norm(double p) { + double q, r, ret; + + assert(p > 0.0 && p < 1.0); + + q = p - 0.5; + if (fabs(q) <= 0.425) { + /* p close to 1/2. */ + r = 0.180625 - q * q; + return q * (((((((2.5090809287301226727e3 * r + + 3.3430575583588128105e4) * r + 6.7265770927008700853e4) * r + + 4.5921953931549871457e4) * r + 1.3731693765509461125e4) * + r + 1.9715909503065514427e3) * r + 1.3314166789178437745e2) + * r + 3.3871328727963666080e0) / + (((((((5.2264952788528545610e3 * r + + 2.8729085735721942674e4) * r + 3.9307895800092710610e4) * r + + 2.1213794301586595867e4) * r + 5.3941960214247511077e3) * + r + 6.8718700749205790830e2) * r + 4.2313330701600911252e1) + * r + 1.0); + } else { + if (q < 0.0) { + r = p; + } else { + r = 1.0 - p; + } + assert(r > 0.0); + + r = sqrt(-log(r)); + if (r <= 5.0) { + /* p neither close to 1/2 nor 0 or 1. */ + r -= 1.6; + ret = ((((((((7.74545014278341407640e-4 * r + + 2.27238449892691845833e-2) * r + + 2.41780725177450611770e-1) * r + + 1.27045825245236838258e0) * r + + 3.64784832476320460504e0) * r + + 5.76949722146069140550e0) * r + + 4.63033784615654529590e0) * r + + 1.42343711074968357734e0) / + (((((((1.05075007164441684324e-9 * r + + 5.47593808499534494600e-4) * r + + 1.51986665636164571966e-2) + * r + 1.48103976427480074590e-1) * r + + 6.89767334985100004550e-1) * r + + 1.67638483018380384940e0) * r + + 2.05319162663775882187e0) * r + 1.0)); + } else { + /* p near 0 or 1. */ + r -= 5.0; + ret = ((((((((2.01033439929228813265e-7 * r + + 2.71155556874348757815e-5) * r + + 1.24266094738807843860e-3) * r + + 2.65321895265761230930e-2) * r + + 2.96560571828504891230e-1) * r + + 1.78482653991729133580e0) * r + + 5.46378491116411436990e0) * r + + 6.65790464350110377720e0) / + (((((((2.04426310338993978564e-15 * r + + 1.42151175831644588870e-7) * r + + 1.84631831751005468180e-5) * r + + 7.86869131145613259100e-4) * r + + 1.48753612908506148525e-2) * r + + 1.36929880922735805310e-1) * r + + 5.99832206555887937690e-1) + * r + 1.0)); + } + if (q < 0.0) { + ret = -ret; + } + return ret; + } +} + +/* + * Given a value p in [0..1] of the lower tail area of the Chi^2 distribution + * with df degrees of freedom, where ln_gamma_df_2 is ln_gamma(df/2.0), compute + * the upper limit on the definite integral from [0..z] that satisfies p, + * accurate to 12 decimal places. + * + * This implementation is based on: + * + * Best, D.J., D.E. Roberts (1975) Algorithm AS 91: The percentage points of + * the Chi^2 distribution. Applied Statistics 24(3):385-388. + * + * Shea, B.L. (1991) Algorithm AS R85: A remark on AS 91: The percentage + * points of the Chi^2 distribution. Applied Statistics 40(1):233-235. + */ +static inline double +pt_chi2(double p, double df, double ln_gamma_df_2) { + double e, aa, xx, c, ch, a, q, p1, p2, t, x, b, s1, s2, s3, s4, s5, s6; + unsigned i; + + assert(p >= 0.0 && p < 1.0); + assert(df > 0.0); + + e = 5.0e-7; + aa = 0.6931471805; + + xx = 0.5 * df; + c = xx - 1.0; + + if (df < -1.24 * log(p)) { + /* Starting approximation for small Chi^2. */ + ch = pow(p * xx * exp(ln_gamma_df_2 + xx * aa), 1.0 / xx); + if (ch - e < 0.0) { + return ch; + } + } else { + if (df > 0.32) { + x = pt_norm(p); + /* + * Starting approximation using Wilson and Hilferty + * estimate. + */ + p1 = 0.222222 / df; + ch = df * pow(x * sqrt(p1) + 1.0 - p1, 3.0); + /* Starting approximation for p tending to 1. */ + if (ch > 2.2 * df + 6.0) { + ch = -2.0 * (log(1.0 - p) - c * log(0.5 * ch) + + ln_gamma_df_2); + } + } else { + ch = 0.4; + a = log(1.0 - p); + while (true) { + q = ch; + p1 = 1.0 + ch * (4.67 + ch); + p2 = ch * (6.73 + ch * (6.66 + ch)); + t = -0.5 + (4.67 + 2.0 * ch) / p1 - (6.73 + ch + * (13.32 + 3.0 * ch)) / p2; + ch -= (1.0 - exp(a + ln_gamma_df_2 + 0.5 * ch + + c * aa) * p2 / p1) / t; + if (fabs(q / ch - 1.0) - 0.01 <= 0.0) { + break; + } + } + } + } + + for (i = 0; i < 20; i++) { + /* Calculation of seven-term Taylor series. */ + q = ch; + p1 = 0.5 * ch; + if (p1 < 0.0) { + return -1.0; + } + p2 = p - i_gamma(p1, xx, ln_gamma_df_2); + t = p2 * exp(xx * aa + ln_gamma_df_2 + p1 - c * log(ch)); + b = t / ch; + a = 0.5 * t - b * c; + s1 = (210.0 + a * (140.0 + a * (105.0 + a * (84.0 + a * (70.0 + + 60.0 * a))))) / 420.0; + s2 = (420.0 + a * (735.0 + a * (966.0 + a * (1141.0 + 1278.0 * + a)))) / 2520.0; + s3 = (210.0 + a * (462.0 + a * (707.0 + 932.0 * a))) / 2520.0; + s4 = (252.0 + a * (672.0 + 1182.0 * a) + c * (294.0 + a * + (889.0 + 1740.0 * a))) / 5040.0; + s5 = (84.0 + 264.0 * a + c * (175.0 + 606.0 * a)) / 2520.0; + s6 = (120.0 + c * (346.0 + 127.0 * c)) / 5040.0; + ch += t * (1.0 + 0.5 * t * s1 - b * c * (s1 - b * (s2 - b * (s3 + - b * (s4 - b * (s5 - b * s6)))))); + if (fabs(q / ch - 1.0) <= e) { + break; + } + } + + return ch; +} + +/* + * Given a value p in [0..1] and Gamma distribution shape and scale parameters, + * compute the upper limit on the definite integral from [0..z] that satisfies + * p. + */ +static inline double +pt_gamma(double p, double shape, double scale, double ln_gamma_shape) { + return pt_chi2(p, shape * 2.0, ln_gamma_shape) * 0.5 * scale; +} diff --git a/deps/jemalloc/test/include/test/mq.h b/deps/jemalloc/test/include/test/mq.h new file mode 100644 index 000000000..af2c078da --- /dev/null +++ b/deps/jemalloc/test/include/test/mq.h @@ -0,0 +1,107 @@ +void mq_nanosleep(unsigned ns); + +/* + * Simple templated message queue implementation that relies on only mutexes for + * synchronization (which reduces portability issues). Given the following + * setup: + * + * typedef struct mq_msg_s mq_msg_t; + * struct mq_msg_s { + * mq_msg(mq_msg_t) link; + * [message data] + * }; + * mq_gen(, mq_, mq_t, mq_msg_t, link) + * + * The API is as follows: + * + * bool mq_init(mq_t *mq); + * void mq_fini(mq_t *mq); + * unsigned mq_count(mq_t *mq); + * mq_msg_t *mq_tryget(mq_t *mq); + * mq_msg_t *mq_get(mq_t *mq); + * void mq_put(mq_t *mq, mq_msg_t *msg); + * + * The message queue linkage embedded in each message is to be treated as + * externally opaque (no need to initialize or clean up externally). mq_fini() + * does not perform any cleanup of messages, since it knows nothing of their + * payloads. + */ +#define mq_msg(a_mq_msg_type) ql_elm(a_mq_msg_type) + +#define mq_gen(a_attr, a_prefix, a_mq_type, a_mq_msg_type, a_field) \ +typedef struct { \ + mtx_t lock; \ + ql_head(a_mq_msg_type) msgs; \ + unsigned count; \ +} a_mq_type; \ +a_attr bool \ +a_prefix##init(a_mq_type *mq) { \ + \ + if (mtx_init(&mq->lock)) { \ + return true; \ + } \ + ql_new(&mq->msgs); \ + mq->count = 0; \ + return false; \ +} \ +a_attr void \ +a_prefix##fini(a_mq_type *mq) { \ + mtx_fini(&mq->lock); \ +} \ +a_attr unsigned \ +a_prefix##count(a_mq_type *mq) { \ + unsigned count; \ + \ + mtx_lock(&mq->lock); \ + count = mq->count; \ + mtx_unlock(&mq->lock); \ + return count; \ +} \ +a_attr a_mq_msg_type * \ +a_prefix##tryget(a_mq_type *mq) { \ + a_mq_msg_type *msg; \ + \ + mtx_lock(&mq->lock); \ + msg = ql_first(&mq->msgs); \ + if (msg != NULL) { \ + ql_head_remove(&mq->msgs, a_mq_msg_type, a_field); \ + mq->count--; \ + } \ + mtx_unlock(&mq->lock); \ + return msg; \ +} \ +a_attr a_mq_msg_type * \ +a_prefix##get(a_mq_type *mq) { \ + a_mq_msg_type *msg; \ + unsigned ns; \ + \ + msg = a_prefix##tryget(mq); \ + if (msg != NULL) { \ + return msg; \ + } \ + \ + ns = 1; \ + while (true) { \ + mq_nanosleep(ns); \ + msg = a_prefix##tryget(mq); \ + if (msg != NULL) { \ + return msg; \ + } \ + if (ns < 1000*1000*1000) { \ + /* Double sleep time, up to max 1 second. */ \ + ns <<= 1; \ + if (ns > 1000*1000*1000) { \ + ns = 1000*1000*1000; \ + } \ + } \ + } \ +} \ +a_attr void \ +a_prefix##put(a_mq_type *mq, a_mq_msg_type *msg) { \ + \ + mtx_lock(&mq->lock); \ + ql_elm_new(msg, a_field); \ + ql_tail_insert(&mq->msgs, msg, a_field); \ + mq->count++; \ + mtx_unlock(&mq->lock); \ +} diff --git a/deps/jemalloc/test/include/test/mtx.h b/deps/jemalloc/test/include/test/mtx.h new file mode 100644 index 000000000..066a21379 --- /dev/null +++ b/deps/jemalloc/test/include/test/mtx.h @@ -0,0 +1,21 @@ +/* + * mtx is a slightly simplified version of malloc_mutex. This code duplication + * is unfortunate, but there are allocator bootstrapping considerations that + * would leak into the test infrastructure if malloc_mutex were used directly + * in tests. + */ + +typedef struct { +#ifdef _WIN32 + CRITICAL_SECTION lock; +#elif (defined(JEMALLOC_OS_UNFAIR_LOCK)) + os_unfair_lock lock; +#else + pthread_mutex_t lock; +#endif +} mtx_t; + +bool mtx_init(mtx_t *mtx); +void mtx_fini(mtx_t *mtx); +void mtx_lock(mtx_t *mtx); +void mtx_unlock(mtx_t *mtx); diff --git a/deps/jemalloc/test/include/test/test.h b/deps/jemalloc/test/include/test/test.h new file mode 100644 index 000000000..fd0e5265d --- /dev/null +++ b/deps/jemalloc/test/include/test/test.h @@ -0,0 +1,338 @@ +#define ASSERT_BUFSIZE 256 + +#define assert_cmp(t, a, b, cmp, neg_cmp, pri, ...) do { \ + t a_ = (a); \ + t b_ = (b); \ + if (!(a_ cmp b_)) { \ + char prefix[ASSERT_BUFSIZE]; \ + char message[ASSERT_BUFSIZE]; \ + malloc_snprintf(prefix, sizeof(prefix), \ + "%s:%s:%d: Failed assertion: " \ + "(%s) " #cmp " (%s) --> " \ + "%" pri " " #neg_cmp " %" pri ": ", \ + __func__, __FILE__, __LINE__, \ + #a, #b, a_, b_); \ + malloc_snprintf(message, sizeof(message), __VA_ARGS__); \ + p_test_fail(prefix, message); \ + } \ +} while (0) + +#define assert_ptr_eq(a, b, ...) assert_cmp(void *, a, b, ==, \ + !=, "p", __VA_ARGS__) +#define assert_ptr_ne(a, b, ...) assert_cmp(void *, a, b, !=, \ + ==, "p", __VA_ARGS__) +#define assert_ptr_null(a, ...) assert_cmp(void *, a, NULL, ==, \ + !=, "p", __VA_ARGS__) +#define assert_ptr_not_null(a, ...) assert_cmp(void *, a, NULL, !=, \ + ==, "p", __VA_ARGS__) + +#define assert_c_eq(a, b, ...) assert_cmp(char, a, b, ==, !=, "c", __VA_ARGS__) +#define assert_c_ne(a, b, ...) assert_cmp(char, a, b, !=, ==, "c", __VA_ARGS__) +#define assert_c_lt(a, b, ...) assert_cmp(char, a, b, <, >=, "c", __VA_ARGS__) +#define assert_c_le(a, b, ...) assert_cmp(char, a, b, <=, >, "c", __VA_ARGS__) +#define assert_c_ge(a, b, ...) assert_cmp(char, a, b, >=, <, "c", __VA_ARGS__) +#define assert_c_gt(a, b, ...) assert_cmp(char, a, b, >, <=, "c", __VA_ARGS__) + +#define assert_x_eq(a, b, ...) assert_cmp(int, a, b, ==, !=, "#x", __VA_ARGS__) +#define assert_x_ne(a, b, ...) assert_cmp(int, a, b, !=, ==, "#x", __VA_ARGS__) +#define assert_x_lt(a, b, ...) assert_cmp(int, a, b, <, >=, "#x", __VA_ARGS__) +#define assert_x_le(a, b, ...) assert_cmp(int, a, b, <=, >, "#x", __VA_ARGS__) +#define assert_x_ge(a, b, ...) assert_cmp(int, a, b, >=, <, "#x", __VA_ARGS__) +#define assert_x_gt(a, b, ...) assert_cmp(int, a, b, >, <=, "#x", __VA_ARGS__) + +#define assert_d_eq(a, b, ...) assert_cmp(int, a, b, ==, !=, "d", __VA_ARGS__) +#define assert_d_ne(a, b, ...) assert_cmp(int, a, b, !=, ==, "d", __VA_ARGS__) +#define assert_d_lt(a, b, ...) assert_cmp(int, a, b, <, >=, "d", __VA_ARGS__) +#define assert_d_le(a, b, ...) assert_cmp(int, a, b, <=, >, "d", __VA_ARGS__) +#define assert_d_ge(a, b, ...) assert_cmp(int, a, b, >=, <, "d", __VA_ARGS__) +#define assert_d_gt(a, b, ...) assert_cmp(int, a, b, >, <=, "d", __VA_ARGS__) + +#define assert_u_eq(a, b, ...) assert_cmp(int, a, b, ==, !=, "u", __VA_ARGS__) +#define assert_u_ne(a, b, ...) assert_cmp(int, a, b, !=, ==, "u", __VA_ARGS__) +#define assert_u_lt(a, b, ...) assert_cmp(int, a, b, <, >=, "u", __VA_ARGS__) +#define assert_u_le(a, b, ...) assert_cmp(int, a, b, <=, >, "u", __VA_ARGS__) +#define assert_u_ge(a, b, ...) assert_cmp(int, a, b, >=, <, "u", __VA_ARGS__) +#define assert_u_gt(a, b, ...) assert_cmp(int, a, b, >, <=, "u", __VA_ARGS__) + +#define assert_ld_eq(a, b, ...) assert_cmp(long, a, b, ==, \ + !=, "ld", __VA_ARGS__) +#define assert_ld_ne(a, b, ...) assert_cmp(long, a, b, !=, \ + ==, "ld", __VA_ARGS__) +#define assert_ld_lt(a, b, ...) assert_cmp(long, a, b, <, \ + >=, "ld", __VA_ARGS__) +#define assert_ld_le(a, b, ...) assert_cmp(long, a, b, <=, \ + >, "ld", __VA_ARGS__) +#define assert_ld_ge(a, b, ...) assert_cmp(long, a, b, >=, \ + <, "ld", __VA_ARGS__) +#define assert_ld_gt(a, b, ...) assert_cmp(long, a, b, >, \ + <=, "ld", __VA_ARGS__) + +#define assert_lu_eq(a, b, ...) assert_cmp(unsigned long, \ + a, b, ==, !=, "lu", __VA_ARGS__) +#define assert_lu_ne(a, b, ...) assert_cmp(unsigned long, \ + a, b, !=, ==, "lu", __VA_ARGS__) +#define assert_lu_lt(a, b, ...) assert_cmp(unsigned long, \ + a, b, <, >=, "lu", __VA_ARGS__) +#define assert_lu_le(a, b, ...) assert_cmp(unsigned long, \ + a, b, <=, >, "lu", __VA_ARGS__) +#define assert_lu_ge(a, b, ...) assert_cmp(unsigned long, \ + a, b, >=, <, "lu", __VA_ARGS__) +#define assert_lu_gt(a, b, ...) assert_cmp(unsigned long, \ + a, b, >, <=, "lu", __VA_ARGS__) + +#define assert_qd_eq(a, b, ...) assert_cmp(long long, a, b, ==, \ + !=, "qd", __VA_ARGS__) +#define assert_qd_ne(a, b, ...) assert_cmp(long long, a, b, !=, \ + ==, "qd", __VA_ARGS__) +#define assert_qd_lt(a, b, ...) assert_cmp(long long, a, b, <, \ + >=, "qd", __VA_ARGS__) +#define assert_qd_le(a, b, ...) assert_cmp(long long, a, b, <=, \ + >, "qd", __VA_ARGS__) +#define assert_qd_ge(a, b, ...) assert_cmp(long long, a, b, >=, \ + <, "qd", __VA_ARGS__) +#define assert_qd_gt(a, b, ...) assert_cmp(long long, a, b, >, \ + <=, "qd", __VA_ARGS__) + +#define assert_qu_eq(a, b, ...) assert_cmp(unsigned long long, \ + a, b, ==, !=, "qu", __VA_ARGS__) +#define assert_qu_ne(a, b, ...) assert_cmp(unsigned long long, \ + a, b, !=, ==, "qu", __VA_ARGS__) +#define assert_qu_lt(a, b, ...) assert_cmp(unsigned long long, \ + a, b, <, >=, "qu", __VA_ARGS__) +#define assert_qu_le(a, b, ...) assert_cmp(unsigned long long, \ + a, b, <=, >, "qu", __VA_ARGS__) +#define assert_qu_ge(a, b, ...) assert_cmp(unsigned long long, \ + a, b, >=, <, "qu", __VA_ARGS__) +#define assert_qu_gt(a, b, ...) assert_cmp(unsigned long long, \ + a, b, >, <=, "qu", __VA_ARGS__) + +#define assert_jd_eq(a, b, ...) assert_cmp(intmax_t, a, b, ==, \ + !=, "jd", __VA_ARGS__) +#define assert_jd_ne(a, b, ...) assert_cmp(intmax_t, a, b, !=, \ + ==, "jd", __VA_ARGS__) +#define assert_jd_lt(a, b, ...) assert_cmp(intmax_t, a, b, <, \ + >=, "jd", __VA_ARGS__) +#define assert_jd_le(a, b, ...) assert_cmp(intmax_t, a, b, <=, \ + >, "jd", __VA_ARGS__) +#define assert_jd_ge(a, b, ...) assert_cmp(intmax_t, a, b, >=, \ + <, "jd", __VA_ARGS__) +#define assert_jd_gt(a, b, ...) assert_cmp(intmax_t, a, b, >, \ + <=, "jd", __VA_ARGS__) + +#define assert_ju_eq(a, b, ...) assert_cmp(uintmax_t, a, b, ==, \ + !=, "ju", __VA_ARGS__) +#define assert_ju_ne(a, b, ...) assert_cmp(uintmax_t, a, b, !=, \ + ==, "ju", __VA_ARGS__) +#define assert_ju_lt(a, b, ...) assert_cmp(uintmax_t, a, b, <, \ + >=, "ju", __VA_ARGS__) +#define assert_ju_le(a, b, ...) assert_cmp(uintmax_t, a, b, <=, \ + >, "ju", __VA_ARGS__) +#define assert_ju_ge(a, b, ...) assert_cmp(uintmax_t, a, b, >=, \ + <, "ju", __VA_ARGS__) +#define assert_ju_gt(a, b, ...) assert_cmp(uintmax_t, a, b, >, \ + <=, "ju", __VA_ARGS__) + +#define assert_zd_eq(a, b, ...) assert_cmp(ssize_t, a, b, ==, \ + !=, "zd", __VA_ARGS__) +#define assert_zd_ne(a, b, ...) assert_cmp(ssize_t, a, b, !=, \ + ==, "zd", __VA_ARGS__) +#define assert_zd_lt(a, b, ...) assert_cmp(ssize_t, a, b, <, \ + >=, "zd", __VA_ARGS__) +#define assert_zd_le(a, b, ...) assert_cmp(ssize_t, a, b, <=, \ + >, "zd", __VA_ARGS__) +#define assert_zd_ge(a, b, ...) assert_cmp(ssize_t, a, b, >=, \ + <, "zd", __VA_ARGS__) +#define assert_zd_gt(a, b, ...) assert_cmp(ssize_t, a, b, >, \ + <=, "zd", __VA_ARGS__) + +#define assert_zu_eq(a, b, ...) assert_cmp(size_t, a, b, ==, \ + !=, "zu", __VA_ARGS__) +#define assert_zu_ne(a, b, ...) assert_cmp(size_t, a, b, !=, \ + ==, "zu", __VA_ARGS__) +#define assert_zu_lt(a, b, ...) assert_cmp(size_t, a, b, <, \ + >=, "zu", __VA_ARGS__) +#define assert_zu_le(a, b, ...) assert_cmp(size_t, a, b, <=, \ + >, "zu", __VA_ARGS__) +#define assert_zu_ge(a, b, ...) assert_cmp(size_t, a, b, >=, \ + <, "zu", __VA_ARGS__) +#define assert_zu_gt(a, b, ...) assert_cmp(size_t, a, b, >, \ + <=, "zu", __VA_ARGS__) + +#define assert_d32_eq(a, b, ...) assert_cmp(int32_t, a, b, ==, \ + !=, FMTd32, __VA_ARGS__) +#define assert_d32_ne(a, b, ...) assert_cmp(int32_t, a, b, !=, \ + ==, FMTd32, __VA_ARGS__) +#define assert_d32_lt(a, b, ...) assert_cmp(int32_t, a, b, <, \ + >=, FMTd32, __VA_ARGS__) +#define assert_d32_le(a, b, ...) assert_cmp(int32_t, a, b, <=, \ + >, FMTd32, __VA_ARGS__) +#define assert_d32_ge(a, b, ...) assert_cmp(int32_t, a, b, >=, \ + <, FMTd32, __VA_ARGS__) +#define assert_d32_gt(a, b, ...) assert_cmp(int32_t, a, b, >, \ + <=, FMTd32, __VA_ARGS__) + +#define assert_u32_eq(a, b, ...) assert_cmp(uint32_t, a, b, ==, \ + !=, FMTu32, __VA_ARGS__) +#define assert_u32_ne(a, b, ...) assert_cmp(uint32_t, a, b, !=, \ + ==, FMTu32, __VA_ARGS__) +#define assert_u32_lt(a, b, ...) assert_cmp(uint32_t, a, b, <, \ + >=, FMTu32, __VA_ARGS__) +#define assert_u32_le(a, b, ...) assert_cmp(uint32_t, a, b, <=, \ + >, FMTu32, __VA_ARGS__) +#define assert_u32_ge(a, b, ...) assert_cmp(uint32_t, a, b, >=, \ + <, FMTu32, __VA_ARGS__) +#define assert_u32_gt(a, b, ...) assert_cmp(uint32_t, a, b, >, \ + <=, FMTu32, __VA_ARGS__) + +#define assert_d64_eq(a, b, ...) assert_cmp(int64_t, a, b, ==, \ + !=, FMTd64, __VA_ARGS__) +#define assert_d64_ne(a, b, ...) assert_cmp(int64_t, a, b, !=, \ + ==, FMTd64, __VA_ARGS__) +#define assert_d64_lt(a, b, ...) assert_cmp(int64_t, a, b, <, \ + >=, FMTd64, __VA_ARGS__) +#define assert_d64_le(a, b, ...) assert_cmp(int64_t, a, b, <=, \ + >, FMTd64, __VA_ARGS__) +#define assert_d64_ge(a, b, ...) assert_cmp(int64_t, a, b, >=, \ + <, FMTd64, __VA_ARGS__) +#define assert_d64_gt(a, b, ...) assert_cmp(int64_t, a, b, >, \ + <=, FMTd64, __VA_ARGS__) + +#define assert_u64_eq(a, b, ...) assert_cmp(uint64_t, a, b, ==, \ + !=, FMTu64, __VA_ARGS__) +#define assert_u64_ne(a, b, ...) assert_cmp(uint64_t, a, b, !=, \ + ==, FMTu64, __VA_ARGS__) +#define assert_u64_lt(a, b, ...) assert_cmp(uint64_t, a, b, <, \ + >=, FMTu64, __VA_ARGS__) +#define assert_u64_le(a, b, ...) assert_cmp(uint64_t, a, b, <=, \ + >, FMTu64, __VA_ARGS__) +#define assert_u64_ge(a, b, ...) assert_cmp(uint64_t, a, b, >=, \ + <, FMTu64, __VA_ARGS__) +#define assert_u64_gt(a, b, ...) assert_cmp(uint64_t, a, b, >, \ + <=, FMTu64, __VA_ARGS__) + +#define assert_b_eq(a, b, ...) do { \ + bool a_ = (a); \ + bool b_ = (b); \ + if (!(a_ == b_)) { \ + char prefix[ASSERT_BUFSIZE]; \ + char message[ASSERT_BUFSIZE]; \ + malloc_snprintf(prefix, sizeof(prefix), \ + "%s:%s:%d: Failed assertion: " \ + "(%s) == (%s) --> %s != %s: ", \ + __func__, __FILE__, __LINE__, \ + #a, #b, a_ ? "true" : "false", \ + b_ ? "true" : "false"); \ + malloc_snprintf(message, sizeof(message), __VA_ARGS__); \ + p_test_fail(prefix, message); \ + } \ +} while (0) +#define assert_b_ne(a, b, ...) do { \ + bool a_ = (a); \ + bool b_ = (b); \ + if (!(a_ != b_)) { \ + char prefix[ASSERT_BUFSIZE]; \ + char message[ASSERT_BUFSIZE]; \ + malloc_snprintf(prefix, sizeof(prefix), \ + "%s:%s:%d: Failed assertion: " \ + "(%s) != (%s) --> %s == %s: ", \ + __func__, __FILE__, __LINE__, \ + #a, #b, a_ ? "true" : "false", \ + b_ ? "true" : "false"); \ + malloc_snprintf(message, sizeof(message), __VA_ARGS__); \ + p_test_fail(prefix, message); \ + } \ +} while (0) +#define assert_true(a, ...) assert_b_eq(a, true, __VA_ARGS__) +#define assert_false(a, ...) assert_b_eq(a, false, __VA_ARGS__) + +#define assert_str_eq(a, b, ...) do { \ + if (strcmp((a), (b))) { \ + char prefix[ASSERT_BUFSIZE]; \ + char message[ASSERT_BUFSIZE]; \ + malloc_snprintf(prefix, sizeof(prefix), \ + "%s:%s:%d: Failed assertion: " \ + "(%s) same as (%s) --> " \ + "\"%s\" differs from \"%s\": ", \ + __func__, __FILE__, __LINE__, #a, #b, a, b); \ + malloc_snprintf(message, sizeof(message), __VA_ARGS__); \ + p_test_fail(prefix, message); \ + } \ +} while (0) +#define assert_str_ne(a, b, ...) do { \ + if (!strcmp((a), (b))) { \ + char prefix[ASSERT_BUFSIZE]; \ + char message[ASSERT_BUFSIZE]; \ + malloc_snprintf(prefix, sizeof(prefix), \ + "%s:%s:%d: Failed assertion: " \ + "(%s) differs from (%s) --> " \ + "\"%s\" same as \"%s\": ", \ + __func__, __FILE__, __LINE__, #a, #b, a, b); \ + malloc_snprintf(message, sizeof(message), __VA_ARGS__); \ + p_test_fail(prefix, message); \ + } \ +} while (0) + +#define assert_not_reached(...) do { \ + char prefix[ASSERT_BUFSIZE]; \ + char message[ASSERT_BUFSIZE]; \ + malloc_snprintf(prefix, sizeof(prefix), \ + "%s:%s:%d: Unreachable code reached: ", \ + __func__, __FILE__, __LINE__); \ + malloc_snprintf(message, sizeof(message), __VA_ARGS__); \ + p_test_fail(prefix, message); \ +} while (0) + +/* + * If this enum changes, corresponding changes in test/test.sh.in are also + * necessary. + */ +typedef enum { + test_status_pass = 0, + test_status_skip = 1, + test_status_fail = 2, + + test_status_count = 3 +} test_status_t; + +typedef void (test_t)(void); + +#define TEST_BEGIN(f) \ +static void \ +f(void) { \ + p_test_init(#f); + +#define TEST_END \ + goto label_test_end; \ +label_test_end: \ + p_test_fini(); \ +} + +#define test(...) \ + p_test(__VA_ARGS__, NULL) + +#define test_no_reentrancy(...) \ + p_test_no_reentrancy(__VA_ARGS__, NULL) + +#define test_no_malloc_init(...) \ + p_test_no_malloc_init(__VA_ARGS__, NULL) + +#define test_skip_if(e) do { \ + if (e) { \ + test_skip("%s:%s:%d: Test skipped: (%s)", \ + __func__, __FILE__, __LINE__, #e); \ + goto label_test_end; \ + } \ +} while (0) + +bool test_is_reentrant(); + +void test_skip(const char *format, ...) JEMALLOC_FORMAT_PRINTF(1, 2); +void test_fail(const char *format, ...) JEMALLOC_FORMAT_PRINTF(1, 2); + +/* For private use by macros. */ +test_status_t p_test(test_t *t, ...); +test_status_t p_test_no_reentrancy(test_t *t, ...); +test_status_t p_test_no_malloc_init(test_t *t, ...); +void p_test_init(const char *name); +void p_test_fini(void); +void p_test_fail(const char *prefix, const char *message); diff --git a/deps/jemalloc/test/include/test/thd.h b/deps/jemalloc/test/include/test/thd.h new file mode 100644 index 000000000..47a51262e --- /dev/null +++ b/deps/jemalloc/test/include/test/thd.h @@ -0,0 +1,9 @@ +/* Abstraction layer for threading in tests. */ +#ifdef _WIN32 +typedef HANDLE thd_t; +#else +typedef pthread_t thd_t; +#endif + +void thd_create(thd_t *thd, void *(*proc)(void *), void *arg); +void thd_join(thd_t thd, void **ret); diff --git a/deps/jemalloc/test/include/test/timer.h b/deps/jemalloc/test/include/test/timer.h new file mode 100644 index 000000000..ace6191b8 --- /dev/null +++ b/deps/jemalloc/test/include/test/timer.h @@ -0,0 +1,11 @@ +/* Simple timer, for use in benchmark reporting. */ + +typedef struct { + nstime_t t0; + nstime_t t1; +} timedelta_t; + +void timer_start(timedelta_t *timer); +void timer_stop(timedelta_t *timer); +uint64_t timer_usec(const timedelta_t *timer); +void timer_ratio(timedelta_t *a, timedelta_t *b, char *buf, size_t buflen); diff --git a/deps/jemalloc/test/integration/MALLOCX_ARENA.c b/deps/jemalloc/test/integration/MALLOCX_ARENA.c new file mode 100644 index 000000000..222164d69 --- /dev/null +++ b/deps/jemalloc/test/integration/MALLOCX_ARENA.c @@ -0,0 +1,66 @@ +#include "test/jemalloc_test.h" + +#define NTHREADS 10 + +static bool have_dss = +#ifdef JEMALLOC_DSS + true +#else + false +#endif + ; + +void * +thd_start(void *arg) { + unsigned thread_ind = (unsigned)(uintptr_t)arg; + unsigned arena_ind; + void *p; + size_t sz; + + sz = sizeof(arena_ind); + assert_d_eq(mallctl("arenas.create", (void *)&arena_ind, &sz, NULL, 0), + 0, "Error in arenas.create"); + + if (thread_ind % 4 != 3) { + size_t mib[3]; + size_t miblen = sizeof(mib) / sizeof(size_t); + const char *dss_precs[] = {"disabled", "primary", "secondary"}; + unsigned prec_ind = thread_ind % + (sizeof(dss_precs)/sizeof(char*)); + const char *dss = dss_precs[prec_ind]; + int expected_err = (have_dss || prec_ind == 0) ? 0 : EFAULT; + assert_d_eq(mallctlnametomib("arena.0.dss", mib, &miblen), 0, + "Error in mallctlnametomib()"); + mib[1] = arena_ind; + assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, (void *)&dss, + sizeof(const char *)), expected_err, + "Error in mallctlbymib()"); + } + + p = mallocx(1, MALLOCX_ARENA(arena_ind)); + assert_ptr_not_null(p, "Unexpected mallocx() error"); + dallocx(p, 0); + + return NULL; +} + +TEST_BEGIN(test_MALLOCX_ARENA) { + thd_t thds[NTHREADS]; + unsigned i; + + for (i = 0; i < NTHREADS; i++) { + thd_create(&thds[i], thd_start, + (void *)(uintptr_t)i); + } + + for (i = 0; i < NTHREADS; i++) { + thd_join(thds[i], NULL); + } +} +TEST_END + +int +main(void) { + return test( + test_MALLOCX_ARENA); +} diff --git a/deps/jemalloc/test/integration/aligned_alloc.c b/deps/jemalloc/test/integration/aligned_alloc.c new file mode 100644 index 000000000..4375b172a --- /dev/null +++ b/deps/jemalloc/test/integration/aligned_alloc.c @@ -0,0 +1,157 @@ +#include "test/jemalloc_test.h" + +#define MAXALIGN (((size_t)1) << 23) + +/* + * On systems which can't merge extents, tests that call this function generate + * a lot of dirty memory very quickly. Purging between cycles mitigates + * potential OOM on e.g. 32-bit Windows. + */ +static void +purge(void) { + assert_d_eq(mallctl("arena.0.purge", NULL, NULL, NULL, 0), 0, + "Unexpected mallctl error"); +} + +TEST_BEGIN(test_alignment_errors) { + size_t alignment; + void *p; + + alignment = 0; + set_errno(0); + p = aligned_alloc(alignment, 1); + assert_false(p != NULL || get_errno() != EINVAL, + "Expected error for invalid alignment %zu", alignment); + + for (alignment = sizeof(size_t); alignment < MAXALIGN; + alignment <<= 1) { + set_errno(0); + p = aligned_alloc(alignment + 1, 1); + assert_false(p != NULL || get_errno() != EINVAL, + "Expected error for invalid alignment %zu", + alignment + 1); + } +} +TEST_END + + +/* + * GCC "-Walloc-size-larger-than" warning detects when one of the memory + * allocation functions is called with a size larger than the maximum size that + * they support. Here we want to explicitly test that the allocation functions + * do indeed fail properly when this is the case, which triggers the warning. + * Therefore we disable the warning for these tests. + */ +JEMALLOC_DIAGNOSTIC_PUSH +JEMALLOC_DIAGNOSTIC_IGNORE_ALLOC_SIZE_LARGER_THAN + +TEST_BEGIN(test_oom_errors) { + size_t alignment, size; + void *p; + +#if LG_SIZEOF_PTR == 3 + alignment = UINT64_C(0x8000000000000000); + size = UINT64_C(0x8000000000000000); +#else + alignment = 0x80000000LU; + size = 0x80000000LU; +#endif + set_errno(0); + p = aligned_alloc(alignment, size); + assert_false(p != NULL || get_errno() != ENOMEM, + "Expected error for aligned_alloc(%zu, %zu)", + alignment, size); + +#if LG_SIZEOF_PTR == 3 + alignment = UINT64_C(0x4000000000000000); + size = UINT64_C(0xc000000000000001); +#else + alignment = 0x40000000LU; + size = 0xc0000001LU; +#endif + set_errno(0); + p = aligned_alloc(alignment, size); + assert_false(p != NULL || get_errno() != ENOMEM, + "Expected error for aligned_alloc(%zu, %zu)", + alignment, size); + + alignment = 0x10LU; +#if LG_SIZEOF_PTR == 3 + size = UINT64_C(0xfffffffffffffff0); +#else + size = 0xfffffff0LU; +#endif + set_errno(0); + p = aligned_alloc(alignment, size); + assert_false(p != NULL || get_errno() != ENOMEM, + "Expected error for aligned_alloc(&p, %zu, %zu)", + alignment, size); +} +TEST_END + +/* Re-enable the "-Walloc-size-larger-than=" warning */ +JEMALLOC_DIAGNOSTIC_POP + +TEST_BEGIN(test_alignment_and_size) { +#define NITER 4 + size_t alignment, size, total; + unsigned i; + void *ps[NITER]; + + for (i = 0; i < NITER; i++) { + ps[i] = NULL; + } + + for (alignment = 8; + alignment <= MAXALIGN; + alignment <<= 1) { + total = 0; + for (size = 1; + size < 3 * alignment && size < (1U << 31); + size += (alignment >> (LG_SIZEOF_PTR-1)) - 1) { + for (i = 0; i < NITER; i++) { + ps[i] = aligned_alloc(alignment, size); + if (ps[i] == NULL) { + char buf[BUFERROR_BUF]; + + buferror(get_errno(), buf, sizeof(buf)); + test_fail( + "Error for alignment=%zu, " + "size=%zu (%#zx): %s", + alignment, size, size, buf); + } + total += malloc_usable_size(ps[i]); + if (total >= (MAXALIGN << 1)) { + break; + } + } + for (i = 0; i < NITER; i++) { + if (ps[i] != NULL) { + free(ps[i]); + ps[i] = NULL; + } + } + } + purge(); + } +#undef NITER +} +TEST_END + +TEST_BEGIN(test_zero_alloc) { + void *res = aligned_alloc(8, 0); + assert(res); + size_t usable = malloc_usable_size(res); + assert(usable > 0); + free(res); +} +TEST_END + +int +main(void) { + return test( + test_alignment_errors, + test_oom_errors, + test_alignment_and_size, + test_zero_alloc); +} diff --git a/deps/jemalloc/test/integration/allocated.c b/deps/jemalloc/test/integration/allocated.c new file mode 100644 index 000000000..1425fd0aa --- /dev/null +++ b/deps/jemalloc/test/integration/allocated.c @@ -0,0 +1,124 @@ +#include "test/jemalloc_test.h" + +static const bool config_stats = +#ifdef JEMALLOC_STATS + true +#else + false +#endif + ; + +void * +thd_start(void *arg) { + int err; + void *p; + uint64_t a0, a1, d0, d1; + uint64_t *ap0, *ap1, *dp0, *dp1; + size_t sz, usize; + + sz = sizeof(a0); + if ((err = mallctl("thread.allocated", (void *)&a0, &sz, NULL, 0))) { + if (err == ENOENT) { + goto label_ENOENT; + } + test_fail("%s(): Error in mallctl(): %s", __func__, + strerror(err)); + } + sz = sizeof(ap0); + if ((err = mallctl("thread.allocatedp", (void *)&ap0, &sz, NULL, 0))) { + if (err == ENOENT) { + goto label_ENOENT; + } + test_fail("%s(): Error in mallctl(): %s", __func__, + strerror(err)); + } + assert_u64_eq(*ap0, a0, + "\"thread.allocatedp\" should provide a pointer to internal " + "storage"); + + sz = sizeof(d0); + if ((err = mallctl("thread.deallocated", (void *)&d0, &sz, NULL, 0))) { + if (err == ENOENT) { + goto label_ENOENT; + } + test_fail("%s(): Error in mallctl(): %s", __func__, + strerror(err)); + } + sz = sizeof(dp0); + if ((err = mallctl("thread.deallocatedp", (void *)&dp0, &sz, NULL, + 0))) { + if (err == ENOENT) { + goto label_ENOENT; + } + test_fail("%s(): Error in mallctl(): %s", __func__, + strerror(err)); + } + assert_u64_eq(*dp0, d0, + "\"thread.deallocatedp\" should provide a pointer to internal " + "storage"); + + p = malloc(1); + assert_ptr_not_null(p, "Unexpected malloc() error"); + + sz = sizeof(a1); + mallctl("thread.allocated", (void *)&a1, &sz, NULL, 0); + sz = sizeof(ap1); + mallctl("thread.allocatedp", (void *)&ap1, &sz, NULL, 0); + assert_u64_eq(*ap1, a1, + "Dereferenced \"thread.allocatedp\" value should equal " + "\"thread.allocated\" value"); + assert_ptr_eq(ap0, ap1, + "Pointer returned by \"thread.allocatedp\" should not change"); + + usize = malloc_usable_size(p); + assert_u64_le(a0 + usize, a1, + "Allocated memory counter should increase by at least the amount " + "explicitly allocated"); + + free(p); + + sz = sizeof(d1); + mallctl("thread.deallocated", (void *)&d1, &sz, NULL, 0); + sz = sizeof(dp1); + mallctl("thread.deallocatedp", (void *)&dp1, &sz, NULL, 0); + assert_u64_eq(*dp1, d1, + "Dereferenced \"thread.deallocatedp\" value should equal " + "\"thread.deallocated\" value"); + assert_ptr_eq(dp0, dp1, + "Pointer returned by \"thread.deallocatedp\" should not change"); + + assert_u64_le(d0 + usize, d1, + "Deallocated memory counter should increase by at least the amount " + "explicitly deallocated"); + + return NULL; +label_ENOENT: + assert_false(config_stats, + "ENOENT should only be returned if stats are disabled"); + test_skip("\"thread.allocated\" mallctl not available"); + return NULL; +} + +TEST_BEGIN(test_main_thread) { + thd_start(NULL); +} +TEST_END + +TEST_BEGIN(test_subthread) { + thd_t thd; + + thd_create(&thd, thd_start, NULL); + thd_join(thd, NULL); +} +TEST_END + +int +main(void) { + /* Run tests multiple times to check for bad interactions. */ + return test( + test_main_thread, + test_subthread, + test_main_thread, + test_subthread, + test_main_thread); +} diff --git a/deps/jemalloc/test/integration/cpp/basic.cpp b/deps/jemalloc/test/integration/cpp/basic.cpp new file mode 100644 index 000000000..65890ecd5 --- /dev/null +++ b/deps/jemalloc/test/integration/cpp/basic.cpp @@ -0,0 +1,25 @@ +#include <memory> +#include "test/jemalloc_test.h" + +TEST_BEGIN(test_basic) { + auto foo = new long(4); + assert_ptr_not_null(foo, "Unexpected new[] failure"); + delete foo; + // Test nullptr handling. + foo = nullptr; + delete foo; + + auto bar = new long; + assert_ptr_not_null(bar, "Unexpected new failure"); + delete bar; + // Test nullptr handling. + bar = nullptr; + delete bar; +} +TEST_END + +int +main() { + return test( + test_basic); +} diff --git a/deps/jemalloc/test/integration/extent.c b/deps/jemalloc/test/integration/extent.c new file mode 100644 index 000000000..b5db08766 --- /dev/null +++ b/deps/jemalloc/test/integration/extent.c @@ -0,0 +1,248 @@ +#include "test/jemalloc_test.h" + +#include "test/extent_hooks.h" + +static bool +check_background_thread_enabled(void) { + bool enabled; + size_t sz = sizeof(bool); + int ret = mallctl("background_thread", (void *)&enabled, &sz, NULL,0); + if (ret == ENOENT) { + return false; + } + assert_d_eq(ret, 0, "Unexpected mallctl error"); + return enabled; +} + +static void +test_extent_body(unsigned arena_ind) { + void *p; + size_t large0, large1, large2, sz; + size_t purge_mib[3]; + size_t purge_miblen; + int flags; + bool xallocx_success_a, xallocx_success_b, xallocx_success_c; + + flags = MALLOCX_ARENA(arena_ind) | MALLOCX_TCACHE_NONE; + + /* Get large size classes. */ + sz = sizeof(size_t); + assert_d_eq(mallctl("arenas.lextent.0.size", (void *)&large0, &sz, NULL, + 0), 0, "Unexpected arenas.lextent.0.size failure"); + assert_d_eq(mallctl("arenas.lextent.1.size", (void *)&large1, &sz, NULL, + 0), 0, "Unexpected arenas.lextent.1.size failure"); + assert_d_eq(mallctl("arenas.lextent.2.size", (void *)&large2, &sz, NULL, + 0), 0, "Unexpected arenas.lextent.2.size failure"); + + /* Test dalloc/decommit/purge cascade. */ + purge_miblen = sizeof(purge_mib)/sizeof(size_t); + assert_d_eq(mallctlnametomib("arena.0.purge", purge_mib, &purge_miblen), + 0, "Unexpected mallctlnametomib() failure"); + purge_mib[1] = (size_t)arena_ind; + called_alloc = false; + try_alloc = true; + try_dalloc = false; + try_decommit = false; + p = mallocx(large0 * 2, flags); + assert_ptr_not_null(p, "Unexpected mallocx() error"); + assert_true(called_alloc, "Expected alloc call"); + called_dalloc = false; + called_decommit = false; + did_purge_lazy = false; + did_purge_forced = false; + called_split = false; + xallocx_success_a = (xallocx(p, large0, 0, flags) == large0); + assert_d_eq(mallctlbymib(purge_mib, purge_miblen, NULL, NULL, NULL, 0), + 0, "Unexpected arena.%u.purge error", arena_ind); + if (xallocx_success_a) { + assert_true(called_dalloc, "Expected dalloc call"); + assert_true(called_decommit, "Expected decommit call"); + assert_true(did_purge_lazy || did_purge_forced, + "Expected purge"); + } + assert_true(called_split, "Expected split call"); + dallocx(p, flags); + try_dalloc = true; + + /* Test decommit/commit and observe split/merge. */ + try_dalloc = false; + try_decommit = true; + p = mallocx(large0 * 2, flags); + assert_ptr_not_null(p, "Unexpected mallocx() error"); + did_decommit = false; + did_commit = false; + called_split = false; + did_split = false; + did_merge = false; + xallocx_success_b = (xallocx(p, large0, 0, flags) == large0); + assert_d_eq(mallctlbymib(purge_mib, purge_miblen, NULL, NULL, NULL, 0), + 0, "Unexpected arena.%u.purge error", arena_ind); + if (xallocx_success_b) { + assert_true(did_split, "Expected split"); + } + xallocx_success_c = (xallocx(p, large0 * 2, 0, flags) == large0 * 2); + if (did_split) { + assert_b_eq(did_decommit, did_commit, + "Expected decommit/commit match"); + } + if (xallocx_success_b && xallocx_success_c) { + assert_true(did_merge, "Expected merge"); + } + dallocx(p, flags); + try_dalloc = true; + try_decommit = false; + + /* Make sure non-large allocation succeeds. */ + p = mallocx(42, flags); + assert_ptr_not_null(p, "Unexpected mallocx() error"); + dallocx(p, flags); +} + +static void +test_manual_hook_auto_arena(void) { + unsigned narenas; + size_t old_size, new_size, sz; + size_t hooks_mib[3]; + size_t hooks_miblen; + extent_hooks_t *new_hooks, *old_hooks; + + extent_hooks_prep(); + + sz = sizeof(unsigned); + /* Get number of auto arenas. */ + assert_d_eq(mallctl("opt.narenas", (void *)&narenas, &sz, NULL, 0), + 0, "Unexpected mallctl() failure"); + if (narenas == 1) { + return; + } + + /* Install custom extent hooks on arena 1 (might not be initialized). */ + hooks_miblen = sizeof(hooks_mib)/sizeof(size_t); + assert_d_eq(mallctlnametomib("arena.0.extent_hooks", hooks_mib, + &hooks_miblen), 0, "Unexpected mallctlnametomib() failure"); + hooks_mib[1] = 1; + old_size = sizeof(extent_hooks_t *); + new_hooks = &hooks; + new_size = sizeof(extent_hooks_t *); + assert_d_eq(mallctlbymib(hooks_mib, hooks_miblen, (void *)&old_hooks, + &old_size, (void *)&new_hooks, new_size), 0, + "Unexpected extent_hooks error"); + static bool auto_arena_created = false; + if (old_hooks != &hooks) { + assert_b_eq(auto_arena_created, false, + "Expected auto arena 1 created only once."); + auto_arena_created = true; + } +} + +static void +test_manual_hook_body(void) { + unsigned arena_ind; + size_t old_size, new_size, sz; + size_t hooks_mib[3]; + size_t hooks_miblen; + extent_hooks_t *new_hooks, *old_hooks; + + extent_hooks_prep(); + + sz = sizeof(unsigned); + assert_d_eq(mallctl("arenas.create", (void *)&arena_ind, &sz, NULL, 0), + 0, "Unexpected mallctl() failure"); + + /* Install custom extent hooks. */ + hooks_miblen = sizeof(hooks_mib)/sizeof(size_t); + assert_d_eq(mallctlnametomib("arena.0.extent_hooks", hooks_mib, + &hooks_miblen), 0, "Unexpected mallctlnametomib() failure"); + hooks_mib[1] = (size_t)arena_ind; + old_size = sizeof(extent_hooks_t *); + new_hooks = &hooks; + new_size = sizeof(extent_hooks_t *); + assert_d_eq(mallctlbymib(hooks_mib, hooks_miblen, (void *)&old_hooks, + &old_size, (void *)&new_hooks, new_size), 0, + "Unexpected extent_hooks error"); + assert_ptr_ne(old_hooks->alloc, extent_alloc_hook, + "Unexpected extent_hooks error"); + assert_ptr_ne(old_hooks->dalloc, extent_dalloc_hook, + "Unexpected extent_hooks error"); + assert_ptr_ne(old_hooks->commit, extent_commit_hook, + "Unexpected extent_hooks error"); + assert_ptr_ne(old_hooks->decommit, extent_decommit_hook, + "Unexpected extent_hooks error"); + assert_ptr_ne(old_hooks->purge_lazy, extent_purge_lazy_hook, + "Unexpected extent_hooks error"); + assert_ptr_ne(old_hooks->purge_forced, extent_purge_forced_hook, + "Unexpected extent_hooks error"); + assert_ptr_ne(old_hooks->split, extent_split_hook, + "Unexpected extent_hooks error"); + assert_ptr_ne(old_hooks->merge, extent_merge_hook, + "Unexpected extent_hooks error"); + + if (!check_background_thread_enabled()) { + test_extent_body(arena_ind); + } + + /* Restore extent hooks. */ + assert_d_eq(mallctlbymib(hooks_mib, hooks_miblen, NULL, NULL, + (void *)&old_hooks, new_size), 0, "Unexpected extent_hooks error"); + assert_d_eq(mallctlbymib(hooks_mib, hooks_miblen, (void *)&old_hooks, + &old_size, NULL, 0), 0, "Unexpected extent_hooks error"); + assert_ptr_eq(old_hooks, default_hooks, "Unexpected extent_hooks error"); + assert_ptr_eq(old_hooks->alloc, default_hooks->alloc, + "Unexpected extent_hooks error"); + assert_ptr_eq(old_hooks->dalloc, default_hooks->dalloc, + "Unexpected extent_hooks error"); + assert_ptr_eq(old_hooks->commit, default_hooks->commit, + "Unexpected extent_hooks error"); + assert_ptr_eq(old_hooks->decommit, default_hooks->decommit, + "Unexpected extent_hooks error"); + assert_ptr_eq(old_hooks->purge_lazy, default_hooks->purge_lazy, + "Unexpected extent_hooks error"); + assert_ptr_eq(old_hooks->purge_forced, default_hooks->purge_forced, + "Unexpected extent_hooks error"); + assert_ptr_eq(old_hooks->split, default_hooks->split, + "Unexpected extent_hooks error"); + assert_ptr_eq(old_hooks->merge, default_hooks->merge, + "Unexpected extent_hooks error"); +} + +TEST_BEGIN(test_extent_manual_hook) { + test_manual_hook_auto_arena(); + test_manual_hook_body(); + + /* Test failure paths. */ + try_split = false; + test_manual_hook_body(); + try_merge = false; + test_manual_hook_body(); + try_purge_lazy = false; + try_purge_forced = false; + test_manual_hook_body(); + + try_split = try_merge = try_purge_lazy = try_purge_forced = true; +} +TEST_END + +TEST_BEGIN(test_extent_auto_hook) { + unsigned arena_ind; + size_t new_size, sz; + extent_hooks_t *new_hooks; + + extent_hooks_prep(); + + sz = sizeof(unsigned); + new_hooks = &hooks; + new_size = sizeof(extent_hooks_t *); + assert_d_eq(mallctl("arenas.create", (void *)&arena_ind, &sz, + (void *)&new_hooks, new_size), 0, "Unexpected mallctl() failure"); + + test_skip_if(check_background_thread_enabled()); + test_extent_body(arena_ind); +} +TEST_END + +int +main(void) { + return test( + test_extent_manual_hook, + test_extent_auto_hook); +} diff --git a/deps/jemalloc/test/integration/extent.sh b/deps/jemalloc/test/integration/extent.sh new file mode 100644 index 000000000..0cc218737 --- /dev/null +++ b/deps/jemalloc/test/integration/extent.sh @@ -0,0 +1,5 @@ +#!/bin/sh + +if [ "x${enable_fill}" = "x1" ] ; then + export MALLOC_CONF="junk:false" +fi diff --git a/deps/jemalloc/test/integration/malloc.c b/deps/jemalloc/test/integration/malloc.c new file mode 100644 index 000000000..8b33bc8f3 --- /dev/null +++ b/deps/jemalloc/test/integration/malloc.c @@ -0,0 +1,16 @@ +#include "test/jemalloc_test.h" + +TEST_BEGIN(test_zero_alloc) { + void *res = malloc(0); + assert(res); + size_t usable = malloc_usable_size(res); + assert(usable > 0); + free(res); +} +TEST_END + +int +main(void) { + return test( + test_zero_alloc); +} diff --git a/deps/jemalloc/test/integration/mallocx.c b/deps/jemalloc/test/integration/mallocx.c new file mode 100644 index 000000000..645d4db48 --- /dev/null +++ b/deps/jemalloc/test/integration/mallocx.c @@ -0,0 +1,274 @@ +#include "test/jemalloc_test.h" + +static unsigned +get_nsizes_impl(const char *cmd) { + unsigned ret; + size_t z; + + z = sizeof(unsigned); + assert_d_eq(mallctl(cmd, (void *)&ret, &z, NULL, 0), 0, + "Unexpected mallctl(\"%s\", ...) failure", cmd); + + return ret; +} + +static unsigned +get_nlarge(void) { + return get_nsizes_impl("arenas.nlextents"); +} + +static size_t +get_size_impl(const char *cmd, size_t ind) { + size_t ret; + size_t z; + size_t mib[4]; + size_t miblen = 4; + + z = sizeof(size_t); + assert_d_eq(mallctlnametomib(cmd, mib, &miblen), + 0, "Unexpected mallctlnametomib(\"%s\", ...) failure", cmd); + mib[2] = ind; + z = sizeof(size_t); + assert_d_eq(mallctlbymib(mib, miblen, (void *)&ret, &z, NULL, 0), + 0, "Unexpected mallctlbymib([\"%s\", %zu], ...) failure", cmd, ind); + + return ret; +} + +static size_t +get_large_size(size_t ind) { + return get_size_impl("arenas.lextent.0.size", ind); +} + +/* + * On systems which can't merge extents, tests that call this function generate + * a lot of dirty memory very quickly. Purging between cycles mitigates + * potential OOM on e.g. 32-bit Windows. + */ +static void +purge(void) { + assert_d_eq(mallctl("arena.0.purge", NULL, NULL, NULL, 0), 0, + "Unexpected mallctl error"); +} + +/* + * GCC "-Walloc-size-larger-than" warning detects when one of the memory + * allocation functions is called with a size larger than the maximum size that + * they support. Here we want to explicitly test that the allocation functions + * do indeed fail properly when this is the case, which triggers the warning. + * Therefore we disable the warning for these tests. + */ +JEMALLOC_DIAGNOSTIC_PUSH +JEMALLOC_DIAGNOSTIC_IGNORE_ALLOC_SIZE_LARGER_THAN + +TEST_BEGIN(test_overflow) { + size_t largemax; + + largemax = get_large_size(get_nlarge()-1); + + assert_ptr_null(mallocx(largemax+1, 0), + "Expected OOM for mallocx(size=%#zx, 0)", largemax+1); + + assert_ptr_null(mallocx(ZU(PTRDIFF_MAX)+1, 0), + "Expected OOM for mallocx(size=%#zx, 0)", ZU(PTRDIFF_MAX)+1); + + assert_ptr_null(mallocx(SIZE_T_MAX, 0), + "Expected OOM for mallocx(size=%#zx, 0)", SIZE_T_MAX); + + assert_ptr_null(mallocx(1, MALLOCX_ALIGN(ZU(PTRDIFF_MAX)+1)), + "Expected OOM for mallocx(size=1, MALLOCX_ALIGN(%#zx))", + ZU(PTRDIFF_MAX)+1); +} +TEST_END + +static void * +remote_alloc(void *arg) { + unsigned arena; + size_t sz = sizeof(unsigned); + assert_d_eq(mallctl("arenas.create", (void *)&arena, &sz, NULL, 0), 0, + "Unexpected mallctl() failure"); + size_t large_sz; + sz = sizeof(size_t); + assert_d_eq(mallctl("arenas.lextent.0.size", (void *)&large_sz, &sz, + NULL, 0), 0, "Unexpected mallctl failure"); + + void *ptr = mallocx(large_sz, MALLOCX_ARENA(arena) + | MALLOCX_TCACHE_NONE); + void **ret = (void **)arg; + *ret = ptr; + + return NULL; +} + +TEST_BEGIN(test_remote_free) { + thd_t thd; + void *ret; + thd_create(&thd, remote_alloc, (void *)&ret); + thd_join(thd, NULL); + assert_ptr_not_null(ret, "Unexpected mallocx failure"); + + /* Avoid TCACHE_NONE to explicitly test tcache_flush(). */ + dallocx(ret, 0); + mallctl("thread.tcache.flush", NULL, NULL, NULL, 0); +} +TEST_END + +TEST_BEGIN(test_oom) { + size_t largemax; + bool oom; + void *ptrs[3]; + unsigned i; + + /* + * It should be impossible to allocate three objects that each consume + * nearly half the virtual address space. + */ + largemax = get_large_size(get_nlarge()-1); + oom = false; + for (i = 0; i < sizeof(ptrs) / sizeof(void *); i++) { + ptrs[i] = mallocx(largemax, MALLOCX_ARENA(0)); + if (ptrs[i] == NULL) { + oom = true; + } + } + assert_true(oom, + "Expected OOM during series of calls to mallocx(size=%zu, 0)", + largemax); + for (i = 0; i < sizeof(ptrs) / sizeof(void *); i++) { + if (ptrs[i] != NULL) { + dallocx(ptrs[i], 0); + } + } + purge(); + +#if LG_SIZEOF_PTR == 3 + assert_ptr_null(mallocx(0x8000000000000000ULL, + MALLOCX_ALIGN(0x8000000000000000ULL)), + "Expected OOM for mallocx()"); + assert_ptr_null(mallocx(0x8000000000000000ULL, + MALLOCX_ALIGN(0x80000000)), + "Expected OOM for mallocx()"); +#else + assert_ptr_null(mallocx(0x80000000UL, MALLOCX_ALIGN(0x80000000UL)), + "Expected OOM for mallocx()"); +#endif +} +TEST_END + +/* Re-enable the "-Walloc-size-larger-than=" warning */ +JEMALLOC_DIAGNOSTIC_POP + +TEST_BEGIN(test_basic) { +#define MAXSZ (((size_t)1) << 23) + size_t sz; + + for (sz = 1; sz < MAXSZ; sz = nallocx(sz, 0) + 1) { + size_t nsz, rsz; + void *p; + nsz = nallocx(sz, 0); + assert_zu_ne(nsz, 0, "Unexpected nallocx() error"); + p = mallocx(sz, 0); + assert_ptr_not_null(p, + "Unexpected mallocx(size=%zx, flags=0) error", sz); + rsz = sallocx(p, 0); + assert_zu_ge(rsz, sz, "Real size smaller than expected"); + assert_zu_eq(nsz, rsz, "nallocx()/sallocx() size mismatch"); + dallocx(p, 0); + + p = mallocx(sz, 0); + assert_ptr_not_null(p, + "Unexpected mallocx(size=%zx, flags=0) error", sz); + dallocx(p, 0); + + nsz = nallocx(sz, MALLOCX_ZERO); + assert_zu_ne(nsz, 0, "Unexpected nallocx() error"); + p = mallocx(sz, MALLOCX_ZERO); + assert_ptr_not_null(p, + "Unexpected mallocx(size=%zx, flags=MALLOCX_ZERO) error", + nsz); + rsz = sallocx(p, 0); + assert_zu_eq(nsz, rsz, "nallocx()/sallocx() rsize mismatch"); + dallocx(p, 0); + purge(); + } +#undef MAXSZ +} +TEST_END + +TEST_BEGIN(test_alignment_and_size) { + const char *percpu_arena; + size_t sz = sizeof(percpu_arena); + + if(mallctl("opt.percpu_arena", (void *)&percpu_arena, &sz, NULL, 0) || + strcmp(percpu_arena, "disabled") != 0) { + test_skip("test_alignment_and_size skipped: " + "not working with percpu arena."); + }; +#define MAXALIGN (((size_t)1) << 23) +#define NITER 4 + size_t nsz, rsz, alignment, total; + unsigned i; + void *ps[NITER]; + + for (i = 0; i < NITER; i++) { + ps[i] = NULL; + } + + for (alignment = 8; + alignment <= MAXALIGN; + alignment <<= 1) { + total = 0; + for (sz = 1; + sz < 3 * alignment && sz < (1U << 31); + sz += (alignment >> (LG_SIZEOF_PTR-1)) - 1) { + for (i = 0; i < NITER; i++) { + nsz = nallocx(sz, MALLOCX_ALIGN(alignment) | + MALLOCX_ZERO | MALLOCX_ARENA(0)); + assert_zu_ne(nsz, 0, + "nallocx() error for alignment=%zu, " + "size=%zu (%#zx)", alignment, sz, sz); + ps[i] = mallocx(sz, MALLOCX_ALIGN(alignment) | + MALLOCX_ZERO | MALLOCX_ARENA(0)); + assert_ptr_not_null(ps[i], + "mallocx() error for alignment=%zu, " + "size=%zu (%#zx)", alignment, sz, sz); + rsz = sallocx(ps[i], 0); + assert_zu_ge(rsz, sz, + "Real size smaller than expected for " + "alignment=%zu, size=%zu", alignment, sz); + assert_zu_eq(nsz, rsz, + "nallocx()/sallocx() size mismatch for " + "alignment=%zu, size=%zu", alignment, sz); + assert_ptr_null( + (void *)((uintptr_t)ps[i] & (alignment-1)), + "%p inadequately aligned for" + " alignment=%zu, size=%zu", ps[i], + alignment, sz); + total += rsz; + if (total >= (MAXALIGN << 1)) { + break; + } + } + for (i = 0; i < NITER; i++) { + if (ps[i] != NULL) { + dallocx(ps[i], 0); + ps[i] = NULL; + } + } + } + purge(); + } +#undef MAXALIGN +#undef NITER +} +TEST_END + +int +main(void) { + return test( + test_overflow, + test_oom, + test_remote_free, + test_basic, + test_alignment_and_size); +} diff --git a/deps/jemalloc/test/integration/mallocx.sh b/deps/jemalloc/test/integration/mallocx.sh new file mode 100644 index 000000000..0cc218737 --- /dev/null +++ b/deps/jemalloc/test/integration/mallocx.sh @@ -0,0 +1,5 @@ +#!/bin/sh + +if [ "x${enable_fill}" = "x1" ] ; then + export MALLOC_CONF="junk:false" +fi diff --git a/deps/jemalloc/test/integration/overflow.c b/deps/jemalloc/test/integration/overflow.c new file mode 100644 index 000000000..748ebb677 --- /dev/null +++ b/deps/jemalloc/test/integration/overflow.c @@ -0,0 +1,59 @@ +#include "test/jemalloc_test.h" + +/* + * GCC "-Walloc-size-larger-than" warning detects when one of the memory + * allocation functions is called with a size larger than the maximum size that + * they support. Here we want to explicitly test that the allocation functions + * do indeed fail properly when this is the case, which triggers the warning. + * Therefore we disable the warning for these tests. + */ +JEMALLOC_DIAGNOSTIC_PUSH +JEMALLOC_DIAGNOSTIC_IGNORE_ALLOC_SIZE_LARGER_THAN + +TEST_BEGIN(test_overflow) { + unsigned nlextents; + size_t mib[4]; + size_t sz, miblen, max_size_class; + void *p; + + sz = sizeof(unsigned); + assert_d_eq(mallctl("arenas.nlextents", (void *)&nlextents, &sz, NULL, + 0), 0, "Unexpected mallctl() error"); + + miblen = sizeof(mib) / sizeof(size_t); + assert_d_eq(mallctlnametomib("arenas.lextent.0.size", mib, &miblen), 0, + "Unexpected mallctlnametomib() error"); + mib[2] = nlextents - 1; + + sz = sizeof(size_t); + assert_d_eq(mallctlbymib(mib, miblen, (void *)&max_size_class, &sz, + NULL, 0), 0, "Unexpected mallctlbymib() error"); + + assert_ptr_null(malloc(max_size_class + 1), + "Expected OOM due to over-sized allocation request"); + assert_ptr_null(malloc(SIZE_T_MAX), + "Expected OOM due to over-sized allocation request"); + + assert_ptr_null(calloc(1, max_size_class + 1), + "Expected OOM due to over-sized allocation request"); + assert_ptr_null(calloc(1, SIZE_T_MAX), + "Expected OOM due to over-sized allocation request"); + + p = malloc(1); + assert_ptr_not_null(p, "Unexpected malloc() OOM"); + assert_ptr_null(realloc(p, max_size_class + 1), + "Expected OOM due to over-sized allocation request"); + assert_ptr_null(realloc(p, SIZE_T_MAX), + "Expected OOM due to over-sized allocation request"); + free(p); +} +TEST_END + +/* Re-enable the "-Walloc-size-larger-than=" warning */ +JEMALLOC_DIAGNOSTIC_POP + +int +main(void) { + return test( + test_overflow); +} diff --git a/deps/jemalloc/test/integration/posix_memalign.c b/deps/jemalloc/test/integration/posix_memalign.c new file mode 100644 index 000000000..d992260a2 --- /dev/null +++ b/deps/jemalloc/test/integration/posix_memalign.c @@ -0,0 +1,128 @@ +#include "test/jemalloc_test.h" + +#define MAXALIGN (((size_t)1) << 23) + +/* + * On systems which can't merge extents, tests that call this function generate + * a lot of dirty memory very quickly. Purging between cycles mitigates + * potential OOM on e.g. 32-bit Windows. + */ +static void +purge(void) { + assert_d_eq(mallctl("arena.0.purge", NULL, NULL, NULL, 0), 0, + "Unexpected mallctl error"); +} + +TEST_BEGIN(test_alignment_errors) { + size_t alignment; + void *p; + + for (alignment = 0; alignment < sizeof(void *); alignment++) { + assert_d_eq(posix_memalign(&p, alignment, 1), EINVAL, + "Expected error for invalid alignment %zu", + alignment); + } + + for (alignment = sizeof(size_t); alignment < MAXALIGN; + alignment <<= 1) { + assert_d_ne(posix_memalign(&p, alignment + 1, 1), 0, + "Expected error for invalid alignment %zu", + alignment + 1); + } +} +TEST_END + +TEST_BEGIN(test_oom_errors) { + size_t alignment, size; + void *p; + +#if LG_SIZEOF_PTR == 3 + alignment = UINT64_C(0x8000000000000000); + size = UINT64_C(0x8000000000000000); +#else + alignment = 0x80000000LU; + size = 0x80000000LU; +#endif + assert_d_ne(posix_memalign(&p, alignment, size), 0, + "Expected error for posix_memalign(&p, %zu, %zu)", + alignment, size); + +#if LG_SIZEOF_PTR == 3 + alignment = UINT64_C(0x4000000000000000); + size = UINT64_C(0xc000000000000001); +#else + alignment = 0x40000000LU; + size = 0xc0000001LU; +#endif + assert_d_ne(posix_memalign(&p, alignment, size), 0, + "Expected error for posix_memalign(&p, %zu, %zu)", + alignment, size); + + alignment = 0x10LU; +#if LG_SIZEOF_PTR == 3 + size = UINT64_C(0xfffffffffffffff0); +#else + size = 0xfffffff0LU; +#endif + assert_d_ne(posix_memalign(&p, alignment, size), 0, + "Expected error for posix_memalign(&p, %zu, %zu)", + alignment, size); +} +TEST_END + +TEST_BEGIN(test_alignment_and_size) { +#define NITER 4 + size_t alignment, size, total; + unsigned i; + int err; + void *ps[NITER]; + + for (i = 0; i < NITER; i++) { + ps[i] = NULL; + } + + for (alignment = 8; + alignment <= MAXALIGN; + alignment <<= 1) { + total = 0; + for (size = 0; + size < 3 * alignment && size < (1U << 31); + size += ((size == 0) ? 1 : + (alignment >> (LG_SIZEOF_PTR-1)) - 1)) { + for (i = 0; i < NITER; i++) { + err = posix_memalign(&ps[i], + alignment, size); + if (err) { + char buf[BUFERROR_BUF]; + + buferror(get_errno(), buf, sizeof(buf)); + test_fail( + "Error for alignment=%zu, " + "size=%zu (%#zx): %s", + alignment, size, size, buf); + } + total += malloc_usable_size(ps[i]); + if (total >= (MAXALIGN << 1)) { + break; + } + } + for (i = 0; i < NITER; i++) { + if (ps[i] != NULL) { + free(ps[i]); + ps[i] = NULL; + } + } + } + purge(); + } +#undef NITER +} +TEST_END + +int +main(void) { + return test( + test_alignment_errors, + test_oom_errors, + test_alignment_and_size); +} diff --git a/deps/jemalloc/test/integration/rallocx.c b/deps/jemalloc/test/integration/rallocx.c new file mode 100644 index 000000000..08ed08d3f --- /dev/null +++ b/deps/jemalloc/test/integration/rallocx.c @@ -0,0 +1,258 @@ +#include "test/jemalloc_test.h" + +static unsigned +get_nsizes_impl(const char *cmd) { + unsigned ret; + size_t z; + + z = sizeof(unsigned); + assert_d_eq(mallctl(cmd, (void *)&ret, &z, NULL, 0), 0, + "Unexpected mallctl(\"%s\", ...) failure", cmd); + + return ret; +} + +static unsigned +get_nlarge(void) { + return get_nsizes_impl("arenas.nlextents"); +} + +static size_t +get_size_impl(const char *cmd, size_t ind) { + size_t ret; + size_t z; + size_t mib[4]; + size_t miblen = 4; + + z = sizeof(size_t); + assert_d_eq(mallctlnametomib(cmd, mib, &miblen), + 0, "Unexpected mallctlnametomib(\"%s\", ...) failure", cmd); + mib[2] = ind; + z = sizeof(size_t); + assert_d_eq(mallctlbymib(mib, miblen, (void *)&ret, &z, NULL, 0), + 0, "Unexpected mallctlbymib([\"%s\", %zu], ...) failure", cmd, ind); + + return ret; +} + +static size_t +get_large_size(size_t ind) { + return get_size_impl("arenas.lextent.0.size", ind); +} + +TEST_BEGIN(test_grow_and_shrink) { + void *p, *q; + size_t tsz; +#define NCYCLES 3 + unsigned i, j; +#define NSZS 1024 + size_t szs[NSZS]; +#define MAXSZ ZU(12 * 1024 * 1024) + + p = mallocx(1, 0); + assert_ptr_not_null(p, "Unexpected mallocx() error"); + szs[0] = sallocx(p, 0); + + for (i = 0; i < NCYCLES; i++) { + for (j = 1; j < NSZS && szs[j-1] < MAXSZ; j++) { + q = rallocx(p, szs[j-1]+1, 0); + assert_ptr_not_null(q, + "Unexpected rallocx() error for size=%zu-->%zu", + szs[j-1], szs[j-1]+1); + szs[j] = sallocx(q, 0); + assert_zu_ne(szs[j], szs[j-1]+1, + "Expected size to be at least: %zu", szs[j-1]+1); + p = q; + } + + for (j--; j > 0; j--) { + q = rallocx(p, szs[j-1], 0); + assert_ptr_not_null(q, + "Unexpected rallocx() error for size=%zu-->%zu", + szs[j], szs[j-1]); + tsz = sallocx(q, 0); + assert_zu_eq(tsz, szs[j-1], + "Expected size=%zu, got size=%zu", szs[j-1], tsz); + p = q; + } + } + + dallocx(p, 0); +#undef MAXSZ +#undef NSZS +#undef NCYCLES +} +TEST_END + +static bool +validate_fill(const void *p, uint8_t c, size_t offset, size_t len) { + bool ret = false; + const uint8_t *buf = (const uint8_t *)p; + size_t i; + + for (i = 0; i < len; i++) { + uint8_t b = buf[offset+i]; + if (b != c) { + test_fail("Allocation at %p (len=%zu) contains %#x " + "rather than %#x at offset %zu", p, len, b, c, + offset+i); + ret = true; + } + } + + return ret; +} + +TEST_BEGIN(test_zero) { + void *p, *q; + size_t psz, qsz, i, j; + size_t start_sizes[] = {1, 3*1024, 63*1024, 4095*1024}; +#define FILL_BYTE 0xaaU +#define RANGE 2048 + + for (i = 0; i < sizeof(start_sizes)/sizeof(size_t); i++) { + size_t start_size = start_sizes[i]; + p = mallocx(start_size, MALLOCX_ZERO); + assert_ptr_not_null(p, "Unexpected mallocx() error"); + psz = sallocx(p, 0); + + assert_false(validate_fill(p, 0, 0, psz), + "Expected zeroed memory"); + memset(p, FILL_BYTE, psz); + assert_false(validate_fill(p, FILL_BYTE, 0, psz), + "Expected filled memory"); + + for (j = 1; j < RANGE; j++) { + q = rallocx(p, start_size+j, MALLOCX_ZERO); + assert_ptr_not_null(q, "Unexpected rallocx() error"); + qsz = sallocx(q, 0); + if (q != p || qsz != psz) { + assert_false(validate_fill(q, FILL_BYTE, 0, + psz), "Expected filled memory"); + assert_false(validate_fill(q, 0, psz, qsz-psz), + "Expected zeroed memory"); + } + if (psz != qsz) { + memset((void *)((uintptr_t)q+psz), FILL_BYTE, + qsz-psz); + psz = qsz; + } + p = q; + } + assert_false(validate_fill(p, FILL_BYTE, 0, psz), + "Expected filled memory"); + dallocx(p, 0); + } +#undef FILL_BYTE +} +TEST_END + +TEST_BEGIN(test_align) { + void *p, *q; + size_t align; +#define MAX_ALIGN (ZU(1) << 25) + + align = ZU(1); + p = mallocx(1, MALLOCX_ALIGN(align)); + assert_ptr_not_null(p, "Unexpected mallocx() error"); + + for (align <<= 1; align <= MAX_ALIGN; align <<= 1) { + q = rallocx(p, 1, MALLOCX_ALIGN(align)); + assert_ptr_not_null(q, + "Unexpected rallocx() error for align=%zu", align); + assert_ptr_null( + (void *)((uintptr_t)q & (align-1)), + "%p inadequately aligned for align=%zu", + q, align); + p = q; + } + dallocx(p, 0); +#undef MAX_ALIGN +} +TEST_END + +TEST_BEGIN(test_lg_align_and_zero) { + void *p, *q; + unsigned lg_align; + size_t sz; +#define MAX_LG_ALIGN 25 +#define MAX_VALIDATE (ZU(1) << 22) + + lg_align = 0; + p = mallocx(1, MALLOCX_LG_ALIGN(lg_align)|MALLOCX_ZERO); + assert_ptr_not_null(p, "Unexpected mallocx() error"); + + for (lg_align++; lg_align <= MAX_LG_ALIGN; lg_align++) { + q = rallocx(p, 1, MALLOCX_LG_ALIGN(lg_align)|MALLOCX_ZERO); + assert_ptr_not_null(q, + "Unexpected rallocx() error for lg_align=%u", lg_align); + assert_ptr_null( + (void *)((uintptr_t)q & ((ZU(1) << lg_align)-1)), + "%p inadequately aligned for lg_align=%u", q, lg_align); + sz = sallocx(q, 0); + if ((sz << 1) <= MAX_VALIDATE) { + assert_false(validate_fill(q, 0, 0, sz), + "Expected zeroed memory"); + } else { + assert_false(validate_fill(q, 0, 0, MAX_VALIDATE), + "Expected zeroed memory"); + assert_false(validate_fill( + (void *)((uintptr_t)q+sz-MAX_VALIDATE), + 0, 0, MAX_VALIDATE), "Expected zeroed memory"); + } + p = q; + } + dallocx(p, 0); +#undef MAX_VALIDATE +#undef MAX_LG_ALIGN +} +TEST_END + +/* + * GCC "-Walloc-size-larger-than" warning detects when one of the memory + * allocation functions is called with a size larger than the maximum size that + * they support. Here we want to explicitly test that the allocation functions + * do indeed fail properly when this is the case, which triggers the warning. + * Therefore we disable the warning for these tests. + */ +JEMALLOC_DIAGNOSTIC_PUSH +JEMALLOC_DIAGNOSTIC_IGNORE_ALLOC_SIZE_LARGER_THAN + +TEST_BEGIN(test_overflow) { + size_t largemax; + void *p; + + largemax = get_large_size(get_nlarge()-1); + + p = mallocx(1, 0); + assert_ptr_not_null(p, "Unexpected mallocx() failure"); + + assert_ptr_null(rallocx(p, largemax+1, 0), + "Expected OOM for rallocx(p, size=%#zx, 0)", largemax+1); + + assert_ptr_null(rallocx(p, ZU(PTRDIFF_MAX)+1, 0), + "Expected OOM for rallocx(p, size=%#zx, 0)", ZU(PTRDIFF_MAX)+1); + + assert_ptr_null(rallocx(p, SIZE_T_MAX, 0), + "Expected OOM for rallocx(p, size=%#zx, 0)", SIZE_T_MAX); + + assert_ptr_null(rallocx(p, 1, MALLOCX_ALIGN(ZU(PTRDIFF_MAX)+1)), + "Expected OOM for rallocx(p, size=1, MALLOCX_ALIGN(%#zx))", + ZU(PTRDIFF_MAX)+1); + + dallocx(p, 0); +} +TEST_END + +/* Re-enable the "-Walloc-size-larger-than=" warning */ +JEMALLOC_DIAGNOSTIC_POP + +int +main(void) { + return test( + test_grow_and_shrink, + test_zero, + test_align, + test_lg_align_and_zero, + test_overflow); +} diff --git a/deps/jemalloc/test/integration/sdallocx.c b/deps/jemalloc/test/integration/sdallocx.c new file mode 100644 index 000000000..ca0144855 --- /dev/null +++ b/deps/jemalloc/test/integration/sdallocx.c @@ -0,0 +1,55 @@ +#include "test/jemalloc_test.h" + +#define MAXALIGN (((size_t)1) << 22) +#define NITER 3 + +TEST_BEGIN(test_basic) { + void *ptr = mallocx(64, 0); + sdallocx(ptr, 64, 0); +} +TEST_END + +TEST_BEGIN(test_alignment_and_size) { + size_t nsz, sz, alignment, total; + unsigned i; + void *ps[NITER]; + + for (i = 0; i < NITER; i++) { + ps[i] = NULL; + } + + for (alignment = 8; + alignment <= MAXALIGN; + alignment <<= 1) { + total = 0; + for (sz = 1; + sz < 3 * alignment && sz < (1U << 31); + sz += (alignment >> (LG_SIZEOF_PTR-1)) - 1) { + for (i = 0; i < NITER; i++) { + nsz = nallocx(sz, MALLOCX_ALIGN(alignment) | + MALLOCX_ZERO); + ps[i] = mallocx(sz, MALLOCX_ALIGN(alignment) | + MALLOCX_ZERO); + total += nsz; + if (total >= (MAXALIGN << 1)) { + break; + } + } + for (i = 0; i < NITER; i++) { + if (ps[i] != NULL) { + sdallocx(ps[i], sz, + MALLOCX_ALIGN(alignment)); + ps[i] = NULL; + } + } + } + } +} +TEST_END + +int +main(void) { + return test_no_reentrancy( + test_basic, + test_alignment_and_size); +} diff --git a/deps/jemalloc/test/integration/slab_sizes.c b/deps/jemalloc/test/integration/slab_sizes.c new file mode 100644 index 000000000..af250c3f4 --- /dev/null +++ b/deps/jemalloc/test/integration/slab_sizes.c @@ -0,0 +1,80 @@ +#include "test/jemalloc_test.h" + +/* Note that this test relies on the unusual slab sizes set in slab_sizes.sh. */ + +TEST_BEGIN(test_slab_sizes) { + unsigned nbins; + size_t page; + size_t sizemib[4]; + size_t slabmib[4]; + size_t len; + + len = sizeof(nbins); + assert_d_eq(mallctl("arenas.nbins", &nbins, &len, NULL, 0), 0, + "nbins mallctl failure"); + + len = sizeof(page); + assert_d_eq(mallctl("arenas.page", &page, &len, NULL, 0), 0, + "page mallctl failure"); + + len = 4; + assert_d_eq(mallctlnametomib("arenas.bin.0.size", sizemib, &len), 0, + "bin size mallctlnametomib failure"); + + len = 4; + assert_d_eq(mallctlnametomib("arenas.bin.0.slab_size", slabmib, &len), + 0, "slab size mallctlnametomib failure"); + + size_t biggest_slab_seen = 0; + + for (unsigned i = 0; i < nbins; i++) { + size_t bin_size; + size_t slab_size; + len = sizeof(size_t); + sizemib[2] = i; + slabmib[2] = i; + assert_d_eq(mallctlbymib(sizemib, 4, (void *)&bin_size, &len, + NULL, 0), 0, "bin size mallctlbymib failure"); + + len = sizeof(size_t); + assert_d_eq(mallctlbymib(slabmib, 4, (void *)&slab_size, &len, + NULL, 0), 0, "slab size mallctlbymib failure"); + + if (bin_size < 100) { + /* + * Then we should be as close to 17 as possible. Since + * not all page sizes are valid (because of bitmap + * limitations on the number of items in a slab), we + * should at least make sure that the number of pages + * goes up. + */ + assert_zu_ge(slab_size, biggest_slab_seen, + "Slab sizes should go up"); + biggest_slab_seen = slab_size; + } else if ( + (100 <= bin_size && bin_size < 128) + || (128 < bin_size && bin_size <= 200)) { + assert_zu_eq(slab_size, page, + "Forced-small slabs should be small"); + } else if (bin_size == 128) { + assert_zu_eq(slab_size, 2 * page, + "Forced-2-page slab should be 2 pages"); + } else if (200 < bin_size && bin_size <= 4096) { + assert_zu_ge(slab_size, biggest_slab_seen, + "Slab sizes should go up"); + biggest_slab_seen = slab_size; + } + } + /* + * For any reasonable configuration, 17 pages should be a valid slab + * size for 4096-byte items. + */ + assert_zu_eq(biggest_slab_seen, 17 * page, "Didn't hit page target"); +} +TEST_END + +int +main(void) { + return test( + test_slab_sizes); +} diff --git a/deps/jemalloc/test/integration/slab_sizes.sh b/deps/jemalloc/test/integration/slab_sizes.sh new file mode 100644 index 000000000..07e3db81b --- /dev/null +++ b/deps/jemalloc/test/integration/slab_sizes.sh @@ -0,0 +1,4 @@ +#!/bin/sh + +# Some screwy-looking slab sizes. +export MALLOC_CONF="slab_sizes:1-4096:17|100-200:1|128-128:2" diff --git a/deps/jemalloc/test/integration/smallocx.c b/deps/jemalloc/test/integration/smallocx.c new file mode 100644 index 000000000..2486752be --- /dev/null +++ b/deps/jemalloc/test/integration/smallocx.c @@ -0,0 +1,312 @@ +#include "test/jemalloc_test.h" +#include "jemalloc/jemalloc_macros.h" + +#define STR_HELPER(x) #x +#define STR(x) STR_HELPER(x) + +#ifndef JEMALLOC_VERSION_GID_IDENT + #error "JEMALLOC_VERSION_GID_IDENT not defined" +#endif + +#define JOIN(x, y) x ## y +#define JOIN2(x, y) JOIN(x, y) +#define smallocx JOIN2(smallocx_, JEMALLOC_VERSION_GID_IDENT) + +typedef struct { + void *ptr; + size_t size; +} smallocx_return_t; + +extern smallocx_return_t +smallocx(size_t size, int flags); + +static unsigned +get_nsizes_impl(const char *cmd) { + unsigned ret; + size_t z; + + z = sizeof(unsigned); + assert_d_eq(mallctl(cmd, (void *)&ret, &z, NULL, 0), 0, + "Unexpected mallctl(\"%s\", ...) failure", cmd); + + return ret; +} + +static unsigned +get_nlarge(void) { + return get_nsizes_impl("arenas.nlextents"); +} + +static size_t +get_size_impl(const char *cmd, size_t ind) { + size_t ret; + size_t z; + size_t mib[4]; + size_t miblen = 4; + + z = sizeof(size_t); + assert_d_eq(mallctlnametomib(cmd, mib, &miblen), + 0, "Unexpected mallctlnametomib(\"%s\", ...) failure", cmd); + mib[2] = ind; + z = sizeof(size_t); + assert_d_eq(mallctlbymib(mib, miblen, (void *)&ret, &z, NULL, 0), + 0, "Unexpected mallctlbymib([\"%s\", %zu], ...) failure", cmd, ind); + + return ret; +} + +static size_t +get_large_size(size_t ind) { + return get_size_impl("arenas.lextent.0.size", ind); +} + +/* + * On systems which can't merge extents, tests that call this function generate + * a lot of dirty memory very quickly. Purging between cycles mitigates + * potential OOM on e.g. 32-bit Windows. + */ +static void +purge(void) { + assert_d_eq(mallctl("arena.0.purge", NULL, NULL, NULL, 0), 0, + "Unexpected mallctl error"); +} + +/* + * GCC "-Walloc-size-larger-than" warning detects when one of the memory + * allocation functions is called with a size larger than the maximum size that + * they support. Here we want to explicitly test that the allocation functions + * do indeed fail properly when this is the case, which triggers the warning. + * Therefore we disable the warning for these tests. + */ +JEMALLOC_DIAGNOSTIC_PUSH +JEMALLOC_DIAGNOSTIC_IGNORE_ALLOC_SIZE_LARGER_THAN + +TEST_BEGIN(test_overflow) { + size_t largemax; + + largemax = get_large_size(get_nlarge()-1); + + assert_ptr_null(smallocx(largemax+1, 0).ptr, + "Expected OOM for smallocx(size=%#zx, 0)", largemax+1); + + assert_ptr_null(smallocx(ZU(PTRDIFF_MAX)+1, 0).ptr, + "Expected OOM for smallocx(size=%#zx, 0)", ZU(PTRDIFF_MAX)+1); + + assert_ptr_null(smallocx(SIZE_T_MAX, 0).ptr, + "Expected OOM for smallocx(size=%#zx, 0)", SIZE_T_MAX); + + assert_ptr_null(smallocx(1, MALLOCX_ALIGN(ZU(PTRDIFF_MAX)+1)).ptr, + "Expected OOM for smallocx(size=1, MALLOCX_ALIGN(%#zx))", + ZU(PTRDIFF_MAX)+1); +} +TEST_END + +static void * +remote_alloc(void *arg) { + unsigned arena; + size_t sz = sizeof(unsigned); + assert_d_eq(mallctl("arenas.create", (void *)&arena, &sz, NULL, 0), 0, + "Unexpected mallctl() failure"); + size_t large_sz; + sz = sizeof(size_t); + assert_d_eq(mallctl("arenas.lextent.0.size", (void *)&large_sz, &sz, + NULL, 0), 0, "Unexpected mallctl failure"); + + smallocx_return_t r + = smallocx(large_sz, MALLOCX_ARENA(arena) | MALLOCX_TCACHE_NONE); + void *ptr = r.ptr; + assert_zu_eq(r.size, + nallocx(large_sz, MALLOCX_ARENA(arena) | MALLOCX_TCACHE_NONE), + "Expected smalloc(size,flags).size == nallocx(size,flags)"); + void **ret = (void **)arg; + *ret = ptr; + + return NULL; +} + +TEST_BEGIN(test_remote_free) { + thd_t thd; + void *ret; + thd_create(&thd, remote_alloc, (void *)&ret); + thd_join(thd, NULL); + assert_ptr_not_null(ret, "Unexpected smallocx failure"); + + /* Avoid TCACHE_NONE to explicitly test tcache_flush(). */ + dallocx(ret, 0); + mallctl("thread.tcache.flush", NULL, NULL, NULL, 0); +} +TEST_END + +TEST_BEGIN(test_oom) { + size_t largemax; + bool oom; + void *ptrs[3]; + unsigned i; + + /* + * It should be impossible to allocate three objects that each consume + * nearly half the virtual address space. + */ + largemax = get_large_size(get_nlarge()-1); + oom = false; + for (i = 0; i < sizeof(ptrs) / sizeof(void *); i++) { + ptrs[i] = smallocx(largemax, 0).ptr; + if (ptrs[i] == NULL) { + oom = true; + } + } + assert_true(oom, + "Expected OOM during series of calls to smallocx(size=%zu, 0)", + largemax); + for (i = 0; i < sizeof(ptrs) / sizeof(void *); i++) { + if (ptrs[i] != NULL) { + dallocx(ptrs[i], 0); + } + } + purge(); + +#if LG_SIZEOF_PTR == 3 + assert_ptr_null(smallocx(0x8000000000000000ULL, + MALLOCX_ALIGN(0x8000000000000000ULL)).ptr, + "Expected OOM for smallocx()"); + assert_ptr_null(smallocx(0x8000000000000000ULL, + MALLOCX_ALIGN(0x80000000)).ptr, + "Expected OOM for smallocx()"); +#else + assert_ptr_null(smallocx(0x80000000UL, MALLOCX_ALIGN(0x80000000UL)).ptr, + "Expected OOM for smallocx()"); +#endif +} +TEST_END + +/* Re-enable the "-Walloc-size-larger-than=" warning */ +JEMALLOC_DIAGNOSTIC_POP + +TEST_BEGIN(test_basic) { +#define MAXSZ (((size_t)1) << 23) + size_t sz; + + for (sz = 1; sz < MAXSZ; sz = nallocx(sz, 0) + 1) { + smallocx_return_t ret; + size_t nsz, rsz, smz; + void *p; + nsz = nallocx(sz, 0); + assert_zu_ne(nsz, 0, "Unexpected nallocx() error"); + ret = smallocx(sz, 0); + p = ret.ptr; + smz = ret.size; + assert_ptr_not_null(p, + "Unexpected smallocx(size=%zx, flags=0) error", sz); + rsz = sallocx(p, 0); + assert_zu_ge(rsz, sz, "Real size smaller than expected"); + assert_zu_eq(nsz, rsz, "nallocx()/sallocx() size mismatch"); + assert_zu_eq(nsz, smz, "nallocx()/smallocx() size mismatch"); + dallocx(p, 0); + + ret = smallocx(sz, 0); + p = ret.ptr; + smz = ret.size; + assert_ptr_not_null(p, + "Unexpected smallocx(size=%zx, flags=0) error", sz); + dallocx(p, 0); + + nsz = nallocx(sz, MALLOCX_ZERO); + assert_zu_ne(nsz, 0, "Unexpected nallocx() error"); + assert_zu_ne(smz, 0, "Unexpected smallocx() error"); + ret = smallocx(sz, MALLOCX_ZERO); + p = ret.ptr; + assert_ptr_not_null(p, + "Unexpected smallocx(size=%zx, flags=MALLOCX_ZERO) error", + nsz); + rsz = sallocx(p, 0); + assert_zu_eq(nsz, rsz, "nallocx()/sallocx() rsize mismatch"); + assert_zu_eq(nsz, smz, "nallocx()/smallocx() size mismatch"); + dallocx(p, 0); + purge(); + } +#undef MAXSZ +} +TEST_END + +TEST_BEGIN(test_alignment_and_size) { + const char *percpu_arena; + size_t sz = sizeof(percpu_arena); + + if(mallctl("opt.percpu_arena", (void *)&percpu_arena, &sz, NULL, 0) || + strcmp(percpu_arena, "disabled") != 0) { + test_skip("test_alignment_and_size skipped: " + "not working with percpu arena."); + }; +#define MAXALIGN (((size_t)1) << 23) +#define NITER 4 + size_t nsz, rsz, smz, alignment, total; + unsigned i; + void *ps[NITER]; + + for (i = 0; i < NITER; i++) { + ps[i] = NULL; + } + + for (alignment = 8; + alignment <= MAXALIGN; + alignment <<= 1) { + total = 0; + for (sz = 1; + sz < 3 * alignment && sz < (1U << 31); + sz += (alignment >> (LG_SIZEOF_PTR-1)) - 1) { + for (i = 0; i < NITER; i++) { + nsz = nallocx(sz, MALLOCX_ALIGN(alignment) | + MALLOCX_ZERO); + assert_zu_ne(nsz, 0, + "nallocx() error for alignment=%zu, " + "size=%zu (%#zx)", alignment, sz, sz); + smallocx_return_t ret + = smallocx(sz, MALLOCX_ALIGN(alignment) | MALLOCX_ZERO); + ps[i] = ret.ptr; + assert_ptr_not_null(ps[i], + "smallocx() error for alignment=%zu, " + "size=%zu (%#zx)", alignment, sz, sz); + rsz = sallocx(ps[i], 0); + smz = ret.size; + assert_zu_ge(rsz, sz, + "Real size smaller than expected for " + "alignment=%zu, size=%zu", alignment, sz); + assert_zu_eq(nsz, rsz, + "nallocx()/sallocx() size mismatch for " + "alignment=%zu, size=%zu", alignment, sz); + assert_zu_eq(nsz, smz, + "nallocx()/smallocx() size mismatch for " + "alignment=%zu, size=%zu", alignment, sz); + assert_ptr_null( + (void *)((uintptr_t)ps[i] & (alignment-1)), + "%p inadequately aligned for" + " alignment=%zu, size=%zu", ps[i], + alignment, sz); + total += rsz; + if (total >= (MAXALIGN << 1)) { + break; + } + } + for (i = 0; i < NITER; i++) { + if (ps[i] != NULL) { + dallocx(ps[i], 0); + ps[i] = NULL; + } + } + } + purge(); + } +#undef MAXALIGN +#undef NITER +} +TEST_END + +int +main(void) { + return test( + test_overflow, + test_oom, + test_remote_free, + test_basic, + test_alignment_and_size); +} diff --git a/deps/jemalloc/test/integration/smallocx.sh b/deps/jemalloc/test/integration/smallocx.sh new file mode 100644 index 000000000..d07f10f3c --- /dev/null +++ b/deps/jemalloc/test/integration/smallocx.sh @@ -0,0 +1,5 @@ +#!/bin/sh + +if [ "x${enable_fill}" = "x1" ] ; then + export MALLOC_CONF="junk:false" +fi diff --git a/deps/jemalloc/test/integration/thread_arena.c b/deps/jemalloc/test/integration/thread_arena.c new file mode 100644 index 000000000..1e5ec05d8 --- /dev/null +++ b/deps/jemalloc/test/integration/thread_arena.c @@ -0,0 +1,86 @@ +#include "test/jemalloc_test.h" + +#define NTHREADS 10 + +void * +thd_start(void *arg) { + unsigned main_arena_ind = *(unsigned *)arg; + void *p; + unsigned arena_ind; + size_t size; + int err; + + p = malloc(1); + assert_ptr_not_null(p, "Error in malloc()"); + free(p); + + size = sizeof(arena_ind); + if ((err = mallctl("thread.arena", (void *)&arena_ind, &size, + (void *)&main_arena_ind, sizeof(main_arena_ind)))) { + char buf[BUFERROR_BUF]; + + buferror(err, buf, sizeof(buf)); + test_fail("Error in mallctl(): %s", buf); + } + + size = sizeof(arena_ind); + if ((err = mallctl("thread.arena", (void *)&arena_ind, &size, NULL, + 0))) { + char buf[BUFERROR_BUF]; + + buferror(err, buf, sizeof(buf)); + test_fail("Error in mallctl(): %s", buf); + } + assert_u_eq(arena_ind, main_arena_ind, + "Arena index should be same as for main thread"); + + return NULL; +} + +static void +mallctl_failure(int err) { + char buf[BUFERROR_BUF]; + + buferror(err, buf, sizeof(buf)); + test_fail("Error in mallctl(): %s", buf); +} + +TEST_BEGIN(test_thread_arena) { + void *p; + int err; + thd_t thds[NTHREADS]; + unsigned i; + + p = malloc(1); + assert_ptr_not_null(p, "Error in malloc()"); + + unsigned arena_ind, old_arena_ind; + size_t sz = sizeof(unsigned); + assert_d_eq(mallctl("arenas.create", (void *)&arena_ind, &sz, NULL, 0), + 0, "Arena creation failure"); + + size_t size = sizeof(arena_ind); + if ((err = mallctl("thread.arena", (void *)&old_arena_ind, &size, + (void *)&arena_ind, sizeof(arena_ind))) != 0) { + mallctl_failure(err); + } + + for (i = 0; i < NTHREADS; i++) { + thd_create(&thds[i], thd_start, + (void *)&arena_ind); + } + + for (i = 0; i < NTHREADS; i++) { + intptr_t join_ret; + thd_join(thds[i], (void *)&join_ret); + assert_zd_eq(join_ret, 0, "Unexpected thread join error"); + } + free(p); +} +TEST_END + +int +main(void) { + return test( + test_thread_arena); +} diff --git a/deps/jemalloc/test/integration/thread_tcache_enabled.c b/deps/jemalloc/test/integration/thread_tcache_enabled.c new file mode 100644 index 000000000..95c9acc13 --- /dev/null +++ b/deps/jemalloc/test/integration/thread_tcache_enabled.c @@ -0,0 +1,87 @@ +#include "test/jemalloc_test.h" + +void * +thd_start(void *arg) { + bool e0, e1; + size_t sz = sizeof(bool); + assert_d_eq(mallctl("thread.tcache.enabled", (void *)&e0, &sz, NULL, + 0), 0, "Unexpected mallctl failure"); + + if (e0) { + e1 = false; + assert_d_eq(mallctl("thread.tcache.enabled", (void *)&e0, &sz, + (void *)&e1, sz), 0, "Unexpected mallctl() error"); + assert_true(e0, "tcache should be enabled"); + } + + e1 = true; + assert_d_eq(mallctl("thread.tcache.enabled", (void *)&e0, &sz, + (void *)&e1, sz), 0, "Unexpected mallctl() error"); + assert_false(e0, "tcache should be disabled"); + + e1 = true; + assert_d_eq(mallctl("thread.tcache.enabled", (void *)&e0, &sz, + (void *)&e1, sz), 0, "Unexpected mallctl() error"); + assert_true(e0, "tcache should be enabled"); + + e1 = false; + assert_d_eq(mallctl("thread.tcache.enabled", (void *)&e0, &sz, + (void *)&e1, sz), 0, "Unexpected mallctl() error"); + assert_true(e0, "tcache should be enabled"); + + e1 = false; + assert_d_eq(mallctl("thread.tcache.enabled", (void *)&e0, &sz, + (void *)&e1, sz), 0, "Unexpected mallctl() error"); + assert_false(e0, "tcache should be disabled"); + + free(malloc(1)); + e1 = true; + assert_d_eq(mallctl("thread.tcache.enabled", (void *)&e0, &sz, + (void *)&e1, sz), 0, "Unexpected mallctl() error"); + assert_false(e0, "tcache should be disabled"); + + free(malloc(1)); + e1 = true; + assert_d_eq(mallctl("thread.tcache.enabled", (void *)&e0, &sz, + (void *)&e1, sz), 0, "Unexpected mallctl() error"); + assert_true(e0, "tcache should be enabled"); + + free(malloc(1)); + e1 = false; + assert_d_eq(mallctl("thread.tcache.enabled", (void *)&e0, &sz, + (void *)&e1, sz), 0, "Unexpected mallctl() error"); + assert_true(e0, "tcache should be enabled"); + + free(malloc(1)); + e1 = false; + assert_d_eq(mallctl("thread.tcache.enabled", (void *)&e0, &sz, + (void *)&e1, sz), 0, "Unexpected mallctl() error"); + assert_false(e0, "tcache should be disabled"); + + free(malloc(1)); + return NULL; +} + +TEST_BEGIN(test_main_thread) { + thd_start(NULL); +} +TEST_END + +TEST_BEGIN(test_subthread) { + thd_t thd; + + thd_create(&thd, thd_start, NULL); + thd_join(thd, NULL); +} +TEST_END + +int +main(void) { + /* Run tests multiple times to check for bad interactions. */ + return test( + test_main_thread, + test_subthread, + test_main_thread, + test_subthread, + test_main_thread); +} diff --git a/deps/jemalloc/test/integration/xallocx.c b/deps/jemalloc/test/integration/xallocx.c new file mode 100644 index 000000000..cd0ca048d --- /dev/null +++ b/deps/jemalloc/test/integration/xallocx.c @@ -0,0 +1,384 @@ +#include "test/jemalloc_test.h" + +/* + * Use a separate arena for xallocx() extension/contraction tests so that + * internal allocation e.g. by heap profiling can't interpose allocations where + * xallocx() would ordinarily be able to extend. + */ +static unsigned +arena_ind(void) { + static unsigned ind = 0; + + if (ind == 0) { + size_t sz = sizeof(ind); + assert_d_eq(mallctl("arenas.create", (void *)&ind, &sz, NULL, + 0), 0, "Unexpected mallctl failure creating arena"); + } + + return ind; +} + +TEST_BEGIN(test_same_size) { + void *p; + size_t sz, tsz; + + p = mallocx(42, 0); + assert_ptr_not_null(p, "Unexpected mallocx() error"); + sz = sallocx(p, 0); + + tsz = xallocx(p, sz, 0, 0); + assert_zu_eq(tsz, sz, "Unexpected size change: %zu --> %zu", sz, tsz); + + dallocx(p, 0); +} +TEST_END + +TEST_BEGIN(test_extra_no_move) { + void *p; + size_t sz, tsz; + + p = mallocx(42, 0); + assert_ptr_not_null(p, "Unexpected mallocx() error"); + sz = sallocx(p, 0); + + tsz = xallocx(p, sz, sz-42, 0); + assert_zu_eq(tsz, sz, "Unexpected size change: %zu --> %zu", sz, tsz); + + dallocx(p, 0); +} +TEST_END + +TEST_BEGIN(test_no_move_fail) { + void *p; + size_t sz, tsz; + + p = mallocx(42, 0); + assert_ptr_not_null(p, "Unexpected mallocx() error"); + sz = sallocx(p, 0); + + tsz = xallocx(p, sz + 5, 0, 0); + assert_zu_eq(tsz, sz, "Unexpected size change: %zu --> %zu", sz, tsz); + + dallocx(p, 0); +} +TEST_END + +static unsigned +get_nsizes_impl(const char *cmd) { + unsigned ret; + size_t z; + + z = sizeof(unsigned); + assert_d_eq(mallctl(cmd, (void *)&ret, &z, NULL, 0), 0, + "Unexpected mallctl(\"%s\", ...) failure", cmd); + + return ret; +} + +static unsigned +get_nsmall(void) { + return get_nsizes_impl("arenas.nbins"); +} + +static unsigned +get_nlarge(void) { + return get_nsizes_impl("arenas.nlextents"); +} + +static size_t +get_size_impl(const char *cmd, size_t ind) { + size_t ret; + size_t z; + size_t mib[4]; + size_t miblen = 4; + + z = sizeof(size_t); + assert_d_eq(mallctlnametomib(cmd, mib, &miblen), + 0, "Unexpected mallctlnametomib(\"%s\", ...) failure", cmd); + mib[2] = ind; + z = sizeof(size_t); + assert_d_eq(mallctlbymib(mib, miblen, (void *)&ret, &z, NULL, 0), + 0, "Unexpected mallctlbymib([\"%s\", %zu], ...) failure", cmd, ind); + + return ret; +} + +static size_t +get_small_size(size_t ind) { + return get_size_impl("arenas.bin.0.size", ind); +} + +static size_t +get_large_size(size_t ind) { + return get_size_impl("arenas.lextent.0.size", ind); +} + +TEST_BEGIN(test_size) { + size_t small0, largemax; + void *p; + + /* Get size classes. */ + small0 = get_small_size(0); + largemax = get_large_size(get_nlarge()-1); + + p = mallocx(small0, 0); + assert_ptr_not_null(p, "Unexpected mallocx() error"); + + /* Test smallest supported size. */ + assert_zu_eq(xallocx(p, 1, 0, 0), small0, + "Unexpected xallocx() behavior"); + + /* Test largest supported size. */ + assert_zu_le(xallocx(p, largemax, 0, 0), largemax, + "Unexpected xallocx() behavior"); + + /* Test size overflow. */ + assert_zu_le(xallocx(p, largemax+1, 0, 0), largemax, + "Unexpected xallocx() behavior"); + assert_zu_le(xallocx(p, SIZE_T_MAX, 0, 0), largemax, + "Unexpected xallocx() behavior"); + + dallocx(p, 0); +} +TEST_END + +TEST_BEGIN(test_size_extra_overflow) { + size_t small0, largemax; + void *p; + + /* Get size classes. */ + small0 = get_small_size(0); + largemax = get_large_size(get_nlarge()-1); + + p = mallocx(small0, 0); + assert_ptr_not_null(p, "Unexpected mallocx() error"); + + /* Test overflows that can be resolved by clamping extra. */ + assert_zu_le(xallocx(p, largemax-1, 2, 0), largemax, + "Unexpected xallocx() behavior"); + assert_zu_le(xallocx(p, largemax, 1, 0), largemax, + "Unexpected xallocx() behavior"); + + /* Test overflow such that largemax-size underflows. */ + assert_zu_le(xallocx(p, largemax+1, 2, 0), largemax, + "Unexpected xallocx() behavior"); + assert_zu_le(xallocx(p, largemax+2, 3, 0), largemax, + "Unexpected xallocx() behavior"); + assert_zu_le(xallocx(p, SIZE_T_MAX-2, 2, 0), largemax, + "Unexpected xallocx() behavior"); + assert_zu_le(xallocx(p, SIZE_T_MAX-1, 1, 0), largemax, + "Unexpected xallocx() behavior"); + + dallocx(p, 0); +} +TEST_END + +TEST_BEGIN(test_extra_small) { + size_t small0, small1, largemax; + void *p; + + /* Get size classes. */ + small0 = get_small_size(0); + small1 = get_small_size(1); + largemax = get_large_size(get_nlarge()-1); + + p = mallocx(small0, 0); + assert_ptr_not_null(p, "Unexpected mallocx() error"); + + assert_zu_eq(xallocx(p, small1, 0, 0), small0, + "Unexpected xallocx() behavior"); + + assert_zu_eq(xallocx(p, small1, 0, 0), small0, + "Unexpected xallocx() behavior"); + + assert_zu_eq(xallocx(p, small0, small1 - small0, 0), small0, + "Unexpected xallocx() behavior"); + + /* Test size+extra overflow. */ + assert_zu_eq(xallocx(p, small0, largemax - small0 + 1, 0), small0, + "Unexpected xallocx() behavior"); + assert_zu_eq(xallocx(p, small0, SIZE_T_MAX - small0, 0), small0, + "Unexpected xallocx() behavior"); + + dallocx(p, 0); +} +TEST_END + +TEST_BEGIN(test_extra_large) { + int flags = MALLOCX_ARENA(arena_ind()); + size_t smallmax, large1, large2, large3, largemax; + void *p; + + /* Get size classes. */ + smallmax = get_small_size(get_nsmall()-1); + large1 = get_large_size(1); + large2 = get_large_size(2); + large3 = get_large_size(3); + largemax = get_large_size(get_nlarge()-1); + + p = mallocx(large3, flags); + assert_ptr_not_null(p, "Unexpected mallocx() error"); + + assert_zu_eq(xallocx(p, large3, 0, flags), large3, + "Unexpected xallocx() behavior"); + /* Test size decrease with zero extra. */ + assert_zu_ge(xallocx(p, large1, 0, flags), large1, + "Unexpected xallocx() behavior"); + assert_zu_ge(xallocx(p, smallmax, 0, flags), large1, + "Unexpected xallocx() behavior"); + + if (xallocx(p, large3, 0, flags) != large3) { + p = rallocx(p, large3, flags); + assert_ptr_not_null(p, "Unexpected rallocx() failure"); + } + /* Test size decrease with non-zero extra. */ + assert_zu_eq(xallocx(p, large1, large3 - large1, flags), large3, + "Unexpected xallocx() behavior"); + assert_zu_eq(xallocx(p, large2, large3 - large2, flags), large3, + "Unexpected xallocx() behavior"); + assert_zu_ge(xallocx(p, large1, large2 - large1, flags), large2, + "Unexpected xallocx() behavior"); + assert_zu_ge(xallocx(p, smallmax, large1 - smallmax, flags), large1, + "Unexpected xallocx() behavior"); + + assert_zu_ge(xallocx(p, large1, 0, flags), large1, + "Unexpected xallocx() behavior"); + /* Test size increase with zero extra. */ + assert_zu_le(xallocx(p, large3, 0, flags), large3, + "Unexpected xallocx() behavior"); + assert_zu_le(xallocx(p, largemax+1, 0, flags), large3, + "Unexpected xallocx() behavior"); + + assert_zu_ge(xallocx(p, large1, 0, flags), large1, + "Unexpected xallocx() behavior"); + /* Test size increase with non-zero extra. */ + assert_zu_le(xallocx(p, large1, SIZE_T_MAX - large1, flags), largemax, + "Unexpected xallocx() behavior"); + + assert_zu_ge(xallocx(p, large1, 0, flags), large1, + "Unexpected xallocx() behavior"); + /* Test size increase with non-zero extra. */ + assert_zu_le(xallocx(p, large1, large3 - large1, flags), large3, + "Unexpected xallocx() behavior"); + + if (xallocx(p, large3, 0, flags) != large3) { + p = rallocx(p, large3, flags); + assert_ptr_not_null(p, "Unexpected rallocx() failure"); + } + /* Test size+extra overflow. */ + assert_zu_le(xallocx(p, large3, largemax - large3 + 1, flags), largemax, + "Unexpected xallocx() behavior"); + + dallocx(p, flags); +} +TEST_END + +static void +print_filled_extents(const void *p, uint8_t c, size_t len) { + const uint8_t *pc = (const uint8_t *)p; + size_t i, range0; + uint8_t c0; + + malloc_printf(" p=%p, c=%#x, len=%zu:", p, c, len); + range0 = 0; + c0 = pc[0]; + for (i = 0; i < len; i++) { + if (pc[i] != c0) { + malloc_printf(" %#x[%zu..%zu)", c0, range0, i); + range0 = i; + c0 = pc[i]; + } + } + malloc_printf(" %#x[%zu..%zu)\n", c0, range0, i); +} + +static bool +validate_fill(const void *p, uint8_t c, size_t offset, size_t len) { + const uint8_t *pc = (const uint8_t *)p; + bool err; + size_t i; + + for (i = offset, err = false; i < offset+len; i++) { + if (pc[i] != c) { + err = true; + } + } + + if (err) { + print_filled_extents(p, c, offset + len); + } + + return err; +} + +static void +test_zero(size_t szmin, size_t szmax) { + int flags = MALLOCX_ARENA(arena_ind()) | MALLOCX_ZERO; + size_t sz, nsz; + void *p; +#define FILL_BYTE 0x7aU + + sz = szmax; + p = mallocx(sz, flags); + assert_ptr_not_null(p, "Unexpected mallocx() error"); + assert_false(validate_fill(p, 0x00, 0, sz), "Memory not filled: sz=%zu", + sz); + + /* + * Fill with non-zero so that non-debug builds are more likely to detect + * errors. + */ + memset(p, FILL_BYTE, sz); + assert_false(validate_fill(p, FILL_BYTE, 0, sz), + "Memory not filled: sz=%zu", sz); + + /* Shrink in place so that we can expect growing in place to succeed. */ + sz = szmin; + if (xallocx(p, sz, 0, flags) != sz) { + p = rallocx(p, sz, flags); + assert_ptr_not_null(p, "Unexpected rallocx() failure"); + } + assert_false(validate_fill(p, FILL_BYTE, 0, sz), + "Memory not filled: sz=%zu", sz); + + for (sz = szmin; sz < szmax; sz = nsz) { + nsz = nallocx(sz+1, flags); + if (xallocx(p, sz+1, 0, flags) != nsz) { + p = rallocx(p, sz+1, flags); + assert_ptr_not_null(p, "Unexpected rallocx() failure"); + } + assert_false(validate_fill(p, FILL_BYTE, 0, sz), + "Memory not filled: sz=%zu", sz); + assert_false(validate_fill(p, 0x00, sz, nsz-sz), + "Memory not filled: sz=%zu, nsz-sz=%zu", sz, nsz-sz); + memset((void *)((uintptr_t)p + sz), FILL_BYTE, nsz-sz); + assert_false(validate_fill(p, FILL_BYTE, 0, nsz), + "Memory not filled: nsz=%zu", nsz); + } + + dallocx(p, flags); +} + +TEST_BEGIN(test_zero_large) { + size_t large0, large1; + + /* Get size classes. */ + large0 = get_large_size(0); + large1 = get_large_size(1); + + test_zero(large1, large0 * 2); +} +TEST_END + +int +main(void) { + return test( + test_same_size, + test_extra_no_move, + test_no_move_fail, + test_size, + test_size_extra_overflow, + test_extra_small, + test_extra_large, + test_zero_large); +} diff --git a/deps/jemalloc/test/integration/xallocx.sh b/deps/jemalloc/test/integration/xallocx.sh new file mode 100644 index 000000000..0cc218737 --- /dev/null +++ b/deps/jemalloc/test/integration/xallocx.sh @@ -0,0 +1,5 @@ +#!/bin/sh + +if [ "x${enable_fill}" = "x1" ] ; then + export MALLOC_CONF="junk:false" +fi diff --git a/deps/jemalloc/test/src/SFMT.c b/deps/jemalloc/test/src/SFMT.c new file mode 100644 index 000000000..c05e2183b --- /dev/null +++ b/deps/jemalloc/test/src/SFMT.c @@ -0,0 +1,719 @@ +/* + * This file derives from SFMT 1.3.3 + * (http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/SFMT/index.html), which was + * released under the terms of the following license: + * + * Copyright (c) 2006,2007 Mutsuo Saito, Makoto Matsumoto and Hiroshima + * University. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions are + * met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials provided + * with the distribution. + * * Neither the name of the Hiroshima University nor the names of + * its contributors may be used to endorse or promote products + * derived from this software without specific prior written + * permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ +/** + * @file SFMT.c + * @brief SIMD oriented Fast Mersenne Twister(SFMT) + * + * @author Mutsuo Saito (Hiroshima University) + * @author Makoto Matsumoto (Hiroshima University) + * + * Copyright (C) 2006,2007 Mutsuo Saito, Makoto Matsumoto and Hiroshima + * University. All rights reserved. + * + * The new BSD License is applied to this software, see LICENSE.txt + */ +#define SFMT_C_ +#include "test/jemalloc_test.h" +#include "test/SFMT-params.h" + +#if defined(JEMALLOC_BIG_ENDIAN) && !defined(BIG_ENDIAN64) +#define BIG_ENDIAN64 1 +#endif +#if defined(__BIG_ENDIAN__) && !defined(__amd64) && !defined(BIG_ENDIAN64) +#define BIG_ENDIAN64 1 +#endif +#if defined(HAVE_ALTIVEC) && !defined(BIG_ENDIAN64) +#define BIG_ENDIAN64 1 +#endif +#if defined(ONLY64) && !defined(BIG_ENDIAN64) + #if defined(__GNUC__) + #error "-DONLY64 must be specified with -DBIG_ENDIAN64" + #endif +#undef ONLY64 +#endif +/*------------------------------------------------------ + 128-bit SIMD data type for Altivec, SSE2 or standard C + ------------------------------------------------------*/ +#if defined(HAVE_ALTIVEC) +/** 128-bit data structure */ +union W128_T { + vector unsigned int s; + uint32_t u[4]; +}; +/** 128-bit data type */ +typedef union W128_T w128_t; + +#elif defined(HAVE_SSE2) +/** 128-bit data structure */ +union W128_T { + __m128i si; + uint32_t u[4]; +}; +/** 128-bit data type */ +typedef union W128_T w128_t; + +#else + +/** 128-bit data structure */ +struct W128_T { + uint32_t u[4]; +}; +/** 128-bit data type */ +typedef struct W128_T w128_t; + +#endif + +struct sfmt_s { + /** the 128-bit internal state array */ + w128_t sfmt[N]; + /** index counter to the 32-bit internal state array */ + int idx; + /** a flag: it is 0 if and only if the internal state is not yet + * initialized. */ + int initialized; +}; + +/*-------------------------------------- + FILE GLOBAL VARIABLES + internal state, index counter and flag + --------------------------------------*/ + +/** a parity check vector which certificate the period of 2^{MEXP} */ +static uint32_t parity[4] = {PARITY1, PARITY2, PARITY3, PARITY4}; + +/*---------------- + STATIC FUNCTIONS + ----------------*/ +static inline int idxof(int i); +#if (!defined(HAVE_ALTIVEC)) && (!defined(HAVE_SSE2)) +static inline void rshift128(w128_t *out, w128_t const *in, int shift); +static inline void lshift128(w128_t *out, w128_t const *in, int shift); +#endif +static inline void gen_rand_all(sfmt_t *ctx); +static inline void gen_rand_array(sfmt_t *ctx, w128_t *array, int size); +static inline uint32_t func1(uint32_t x); +static inline uint32_t func2(uint32_t x); +static void period_certification(sfmt_t *ctx); +#if defined(BIG_ENDIAN64) && !defined(ONLY64) +static inline void swap(w128_t *array, int size); +#endif + +#if defined(HAVE_ALTIVEC) + #include "test/SFMT-alti.h" +#elif defined(HAVE_SSE2) + #include "test/SFMT-sse2.h" +#endif + +/** + * This function simulate a 64-bit index of LITTLE ENDIAN + * in BIG ENDIAN machine. + */ +#ifdef ONLY64 +static inline int idxof(int i) { + return i ^ 1; +} +#else +static inline int idxof(int i) { + return i; +} +#endif +/** + * This function simulates SIMD 128-bit right shift by the standard C. + * The 128-bit integer given in in is shifted by (shift * 8) bits. + * This function simulates the LITTLE ENDIAN SIMD. + * @param out the output of this function + * @param in the 128-bit data to be shifted + * @param shift the shift value + */ +#if (!defined(HAVE_ALTIVEC)) && (!defined(HAVE_SSE2)) +#ifdef ONLY64 +static inline void rshift128(w128_t *out, w128_t const *in, int shift) { + uint64_t th, tl, oh, ol; + + th = ((uint64_t)in->u[2] << 32) | ((uint64_t)in->u[3]); + tl = ((uint64_t)in->u[0] << 32) | ((uint64_t)in->u[1]); + + oh = th >> (shift * 8); + ol = tl >> (shift * 8); + ol |= th << (64 - shift * 8); + out->u[0] = (uint32_t)(ol >> 32); + out->u[1] = (uint32_t)ol; + out->u[2] = (uint32_t)(oh >> 32); + out->u[3] = (uint32_t)oh; +} +#else +static inline void rshift128(w128_t *out, w128_t const *in, int shift) { + uint64_t th, tl, oh, ol; + + th = ((uint64_t)in->u[3] << 32) | ((uint64_t)in->u[2]); + tl = ((uint64_t)in->u[1] << 32) | ((uint64_t)in->u[0]); + + oh = th >> (shift * 8); + ol = tl >> (shift * 8); + ol |= th << (64 - shift * 8); + out->u[1] = (uint32_t)(ol >> 32); + out->u[0] = (uint32_t)ol; + out->u[3] = (uint32_t)(oh >> 32); + out->u[2] = (uint32_t)oh; +} +#endif +/** + * This function simulates SIMD 128-bit left shift by the standard C. + * The 128-bit integer given in in is shifted by (shift * 8) bits. + * This function simulates the LITTLE ENDIAN SIMD. + * @param out the output of this function + * @param in the 128-bit data to be shifted + * @param shift the shift value + */ +#ifdef ONLY64 +static inline void lshift128(w128_t *out, w128_t const *in, int shift) { + uint64_t th, tl, oh, ol; + + th = ((uint64_t)in->u[2] << 32) | ((uint64_t)in->u[3]); + tl = ((uint64_t)in->u[0] << 32) | ((uint64_t)in->u[1]); + + oh = th << (shift * 8); + ol = tl << (shift * 8); + oh |= tl >> (64 - shift * 8); + out->u[0] = (uint32_t)(ol >> 32); + out->u[1] = (uint32_t)ol; + out->u[2] = (uint32_t)(oh >> 32); + out->u[3] = (uint32_t)oh; +} +#else +static inline void lshift128(w128_t *out, w128_t const *in, int shift) { + uint64_t th, tl, oh, ol; + + th = ((uint64_t)in->u[3] << 32) | ((uint64_t)in->u[2]); + tl = ((uint64_t)in->u[1] << 32) | ((uint64_t)in->u[0]); + + oh = th << (shift * 8); + ol = tl << (shift * 8); + oh |= tl >> (64 - shift * 8); + out->u[1] = (uint32_t)(ol >> 32); + out->u[0] = (uint32_t)ol; + out->u[3] = (uint32_t)(oh >> 32); + out->u[2] = (uint32_t)oh; +} +#endif +#endif + +/** + * This function represents the recursion formula. + * @param r output + * @param a a 128-bit part of the internal state array + * @param b a 128-bit part of the internal state array + * @param c a 128-bit part of the internal state array + * @param d a 128-bit part of the internal state array + */ +#if (!defined(HAVE_ALTIVEC)) && (!defined(HAVE_SSE2)) +#ifdef ONLY64 +static inline void do_recursion(w128_t *r, w128_t *a, w128_t *b, w128_t *c, + w128_t *d) { + w128_t x; + w128_t y; + + lshift128(&x, a, SL2); + rshift128(&y, c, SR2); + r->u[0] = a->u[0] ^ x.u[0] ^ ((b->u[0] >> SR1) & MSK2) ^ y.u[0] + ^ (d->u[0] << SL1); + r->u[1] = a->u[1] ^ x.u[1] ^ ((b->u[1] >> SR1) & MSK1) ^ y.u[1] + ^ (d->u[1] << SL1); + r->u[2] = a->u[2] ^ x.u[2] ^ ((b->u[2] >> SR1) & MSK4) ^ y.u[2] + ^ (d->u[2] << SL1); + r->u[3] = a->u[3] ^ x.u[3] ^ ((b->u[3] >> SR1) & MSK3) ^ y.u[3] + ^ (d->u[3] << SL1); +} +#else +static inline void do_recursion(w128_t *r, w128_t *a, w128_t *b, w128_t *c, + w128_t *d) { + w128_t x; + w128_t y; + + lshift128(&x, a, SL2); + rshift128(&y, c, SR2); + r->u[0] = a->u[0] ^ x.u[0] ^ ((b->u[0] >> SR1) & MSK1) ^ y.u[0] + ^ (d->u[0] << SL1); + r->u[1] = a->u[1] ^ x.u[1] ^ ((b->u[1] >> SR1) & MSK2) ^ y.u[1] + ^ (d->u[1] << SL1); + r->u[2] = a->u[2] ^ x.u[2] ^ ((b->u[2] >> SR1) & MSK3) ^ y.u[2] + ^ (d->u[2] << SL1); + r->u[3] = a->u[3] ^ x.u[3] ^ ((b->u[3] >> SR1) & MSK4) ^ y.u[3] + ^ (d->u[3] << SL1); +} +#endif +#endif + +#if (!defined(HAVE_ALTIVEC)) && (!defined(HAVE_SSE2)) +/** + * This function fills the internal state array with pseudorandom + * integers. + */ +static inline void gen_rand_all(sfmt_t *ctx) { + int i; + w128_t *r1, *r2; + + r1 = &ctx->sfmt[N - 2]; + r2 = &ctx->sfmt[N - 1]; + for (i = 0; i < N - POS1; i++) { + do_recursion(&ctx->sfmt[i], &ctx->sfmt[i], &ctx->sfmt[i + POS1], r1, + r2); + r1 = r2; + r2 = &ctx->sfmt[i]; + } + for (; i < N; i++) { + do_recursion(&ctx->sfmt[i], &ctx->sfmt[i], &ctx->sfmt[i + POS1 - N], r1, + r2); + r1 = r2; + r2 = &ctx->sfmt[i]; + } +} + +/** + * This function fills the user-specified array with pseudorandom + * integers. + * + * @param array an 128-bit array to be filled by pseudorandom numbers. + * @param size number of 128-bit pseudorandom numbers to be generated. + */ +static inline void gen_rand_array(sfmt_t *ctx, w128_t *array, int size) { + int i, j; + w128_t *r1, *r2; + + r1 = &ctx->sfmt[N - 2]; + r2 = &ctx->sfmt[N - 1]; + for (i = 0; i < N - POS1; i++) { + do_recursion(&array[i], &ctx->sfmt[i], &ctx->sfmt[i + POS1], r1, r2); + r1 = r2; + r2 = &array[i]; + } + for (; i < N; i++) { + do_recursion(&array[i], &ctx->sfmt[i], &array[i + POS1 - N], r1, r2); + r1 = r2; + r2 = &array[i]; + } + for (; i < size - N; i++) { + do_recursion(&array[i], &array[i - N], &array[i + POS1 - N], r1, r2); + r1 = r2; + r2 = &array[i]; + } + for (j = 0; j < 2 * N - size; j++) { + ctx->sfmt[j] = array[j + size - N]; + } + for (; i < size; i++, j++) { + do_recursion(&array[i], &array[i - N], &array[i + POS1 - N], r1, r2); + r1 = r2; + r2 = &array[i]; + ctx->sfmt[j] = array[i]; + } +} +#endif + +#if defined(BIG_ENDIAN64) && !defined(ONLY64) && !defined(HAVE_ALTIVEC) +static inline void swap(w128_t *array, int size) { + int i; + uint32_t x, y; + + for (i = 0; i < size; i++) { + x = array[i].u[0]; + y = array[i].u[2]; + array[i].u[0] = array[i].u[1]; + array[i].u[2] = array[i].u[3]; + array[i].u[1] = x; + array[i].u[3] = y; + } +} +#endif +/** + * This function represents a function used in the initialization + * by init_by_array + * @param x 32-bit integer + * @return 32-bit integer + */ +static uint32_t func1(uint32_t x) { + return (x ^ (x >> 27)) * (uint32_t)1664525UL; +} + +/** + * This function represents a function used in the initialization + * by init_by_array + * @param x 32-bit integer + * @return 32-bit integer + */ +static uint32_t func2(uint32_t x) { + return (x ^ (x >> 27)) * (uint32_t)1566083941UL; +} + +/** + * This function certificate the period of 2^{MEXP} + */ +static void period_certification(sfmt_t *ctx) { + int inner = 0; + int i, j; + uint32_t work; + uint32_t *psfmt32 = &ctx->sfmt[0].u[0]; + + for (i = 0; i < 4; i++) + inner ^= psfmt32[idxof(i)] & parity[i]; + for (i = 16; i > 0; i >>= 1) + inner ^= inner >> i; + inner &= 1; + /* check OK */ + if (inner == 1) { + return; + } + /* check NG, and modification */ + for (i = 0; i < 4; i++) { + work = 1; + for (j = 0; j < 32; j++) { + if ((work & parity[i]) != 0) { + psfmt32[idxof(i)] ^= work; + return; + } + work = work << 1; + } + } +} + +/*---------------- + PUBLIC FUNCTIONS + ----------------*/ +/** + * This function returns the identification string. + * The string shows the word size, the Mersenne exponent, + * and all parameters of this generator. + */ +const char *get_idstring(void) { + return IDSTR; +} + +/** + * This function returns the minimum size of array used for \b + * fill_array32() function. + * @return minimum size of array used for fill_array32() function. + */ +int get_min_array_size32(void) { + return N32; +} + +/** + * This function returns the minimum size of array used for \b + * fill_array64() function. + * @return minimum size of array used for fill_array64() function. + */ +int get_min_array_size64(void) { + return N64; +} + +#ifndef ONLY64 +/** + * This function generates and returns 32-bit pseudorandom number. + * init_gen_rand or init_by_array must be called before this function. + * @return 32-bit pseudorandom number + */ +uint32_t gen_rand32(sfmt_t *ctx) { + uint32_t r; + uint32_t *psfmt32 = &ctx->sfmt[0].u[0]; + + assert(ctx->initialized); + if (ctx->idx >= N32) { + gen_rand_all(ctx); + ctx->idx = 0; + } + r = psfmt32[ctx->idx++]; + return r; +} + +/* Generate a random integer in [0..limit). */ +uint32_t gen_rand32_range(sfmt_t *ctx, uint32_t limit) { + uint32_t ret, above; + + above = 0xffffffffU - (0xffffffffU % limit); + while (1) { + ret = gen_rand32(ctx); + if (ret < above) { + ret %= limit; + break; + } + } + return ret; +} +#endif +/** + * This function generates and returns 64-bit pseudorandom number. + * init_gen_rand or init_by_array must be called before this function. + * The function gen_rand64 should not be called after gen_rand32, + * unless an initialization is again executed. + * @return 64-bit pseudorandom number + */ +uint64_t gen_rand64(sfmt_t *ctx) { +#if defined(BIG_ENDIAN64) && !defined(ONLY64) + uint32_t r1, r2; + uint32_t *psfmt32 = &ctx->sfmt[0].u[0]; +#else + uint64_t r; + uint64_t *psfmt64 = (uint64_t *)&ctx->sfmt[0].u[0]; +#endif + + assert(ctx->initialized); + assert(ctx->idx % 2 == 0); + + if (ctx->idx >= N32) { + gen_rand_all(ctx); + ctx->idx = 0; + } +#if defined(BIG_ENDIAN64) && !defined(ONLY64) + r1 = psfmt32[ctx->idx]; + r2 = psfmt32[ctx->idx + 1]; + ctx->idx += 2; + return ((uint64_t)r2 << 32) | r1; +#else + r = psfmt64[ctx->idx / 2]; + ctx->idx += 2; + return r; +#endif +} + +/* Generate a random integer in [0..limit). */ +uint64_t gen_rand64_range(sfmt_t *ctx, uint64_t limit) { + uint64_t ret, above; + + above = KQU(0xffffffffffffffff) - (KQU(0xffffffffffffffff) % limit); + while (1) { + ret = gen_rand64(ctx); + if (ret < above) { + ret %= limit; + break; + } + } + return ret; +} + +#ifndef ONLY64 +/** + * This function generates pseudorandom 32-bit integers in the + * specified array[] by one call. The number of pseudorandom integers + * is specified by the argument size, which must be at least 624 and a + * multiple of four. The generation by this function is much faster + * than the following gen_rand function. + * + * For initialization, init_gen_rand or init_by_array must be called + * before the first call of this function. This function can not be + * used after calling gen_rand function, without initialization. + * + * @param array an array where pseudorandom 32-bit integers are filled + * by this function. The pointer to the array must be \b "aligned" + * (namely, must be a multiple of 16) in the SIMD version, since it + * refers to the address of a 128-bit integer. In the standard C + * version, the pointer is arbitrary. + * + * @param size the number of 32-bit pseudorandom integers to be + * generated. size must be a multiple of 4, and greater than or equal + * to (MEXP / 128 + 1) * 4. + * + * @note \b memalign or \b posix_memalign is available to get aligned + * memory. Mac OSX doesn't have these functions, but \b malloc of OSX + * returns the pointer to the aligned memory block. + */ +void fill_array32(sfmt_t *ctx, uint32_t *array, int size) { + assert(ctx->initialized); + assert(ctx->idx == N32); + assert(size % 4 == 0); + assert(size >= N32); + + gen_rand_array(ctx, (w128_t *)array, size / 4); + ctx->idx = N32; +} +#endif + +/** + * This function generates pseudorandom 64-bit integers in the + * specified array[] by one call. The number of pseudorandom integers + * is specified by the argument size, which must be at least 312 and a + * multiple of two. The generation by this function is much faster + * than the following gen_rand function. + * + * For initialization, init_gen_rand or init_by_array must be called + * before the first call of this function. This function can not be + * used after calling gen_rand function, without initialization. + * + * @param array an array where pseudorandom 64-bit integers are filled + * by this function. The pointer to the array must be "aligned" + * (namely, must be a multiple of 16) in the SIMD version, since it + * refers to the address of a 128-bit integer. In the standard C + * version, the pointer is arbitrary. + * + * @param size the number of 64-bit pseudorandom integers to be + * generated. size must be a multiple of 2, and greater than or equal + * to (MEXP / 128 + 1) * 2 + * + * @note \b memalign or \b posix_memalign is available to get aligned + * memory. Mac OSX doesn't have these functions, but \b malloc of OSX + * returns the pointer to the aligned memory block. + */ +void fill_array64(sfmt_t *ctx, uint64_t *array, int size) { + assert(ctx->initialized); + assert(ctx->idx == N32); + assert(size % 2 == 0); + assert(size >= N64); + + gen_rand_array(ctx, (w128_t *)array, size / 2); + ctx->idx = N32; + +#if defined(BIG_ENDIAN64) && !defined(ONLY64) + swap((w128_t *)array, size /2); +#endif +} + +/** + * This function initializes the internal state array with a 32-bit + * integer seed. + * + * @param seed a 32-bit integer used as the seed. + */ +sfmt_t *init_gen_rand(uint32_t seed) { + void *p; + sfmt_t *ctx; + int i; + uint32_t *psfmt32; + + if (posix_memalign(&p, sizeof(w128_t), sizeof(sfmt_t)) != 0) { + return NULL; + } + ctx = (sfmt_t *)p; + psfmt32 = &ctx->sfmt[0].u[0]; + + psfmt32[idxof(0)] = seed; + for (i = 1; i < N32; i++) { + psfmt32[idxof(i)] = 1812433253UL * (psfmt32[idxof(i - 1)] + ^ (psfmt32[idxof(i - 1)] >> 30)) + + i; + } + ctx->idx = N32; + period_certification(ctx); + ctx->initialized = 1; + + return ctx; +} + +/** + * This function initializes the internal state array, + * with an array of 32-bit integers used as the seeds + * @param init_key the array of 32-bit integers, used as a seed. + * @param key_length the length of init_key. + */ +sfmt_t *init_by_array(uint32_t *init_key, int key_length) { + void *p; + sfmt_t *ctx; + int i, j, count; + uint32_t r; + int lag; + int mid; + int size = N * 4; + uint32_t *psfmt32; + + if (posix_memalign(&p, sizeof(w128_t), sizeof(sfmt_t)) != 0) { + return NULL; + } + ctx = (sfmt_t *)p; + psfmt32 = &ctx->sfmt[0].u[0]; + + if (size >= 623) { + lag = 11; + } else if (size >= 68) { + lag = 7; + } else if (size >= 39) { + lag = 5; + } else { + lag = 3; + } + mid = (size - lag) / 2; + + memset(ctx->sfmt, 0x8b, sizeof(ctx->sfmt)); + if (key_length + 1 > N32) { + count = key_length + 1; + } else { + count = N32; + } + r = func1(psfmt32[idxof(0)] ^ psfmt32[idxof(mid)] + ^ psfmt32[idxof(N32 - 1)]); + psfmt32[idxof(mid)] += r; + r += key_length; + psfmt32[idxof(mid + lag)] += r; + psfmt32[idxof(0)] = r; + + count--; + for (i = 1, j = 0; (j < count) && (j < key_length); j++) { + r = func1(psfmt32[idxof(i)] ^ psfmt32[idxof((i + mid) % N32)] + ^ psfmt32[idxof((i + N32 - 1) % N32)]); + psfmt32[idxof((i + mid) % N32)] += r; + r += init_key[j] + i; + psfmt32[idxof((i + mid + lag) % N32)] += r; + psfmt32[idxof(i)] = r; + i = (i + 1) % N32; + } + for (; j < count; j++) { + r = func1(psfmt32[idxof(i)] ^ psfmt32[idxof((i + mid) % N32)] + ^ psfmt32[idxof((i + N32 - 1) % N32)]); + psfmt32[idxof((i + mid) % N32)] += r; + r += i; + psfmt32[idxof((i + mid + lag) % N32)] += r; + psfmt32[idxof(i)] = r; + i = (i + 1) % N32; + } + for (j = 0; j < N32; j++) { + r = func2(psfmt32[idxof(i)] + psfmt32[idxof((i + mid) % N32)] + + psfmt32[idxof((i + N32 - 1) % N32)]); + psfmt32[idxof((i + mid) % N32)] ^= r; + r -= i; + psfmt32[idxof((i + mid + lag) % N32)] ^= r; + psfmt32[idxof(i)] = r; + i = (i + 1) % N32; + } + + ctx->idx = N32; + period_certification(ctx); + ctx->initialized = 1; + + return ctx; +} + +void fini_gen_rand(sfmt_t *ctx) { + assert(ctx != NULL); + + ctx->initialized = 0; + free(ctx); +} diff --git a/deps/jemalloc/test/src/btalloc.c b/deps/jemalloc/test/src/btalloc.c new file mode 100644 index 000000000..d570952ce --- /dev/null +++ b/deps/jemalloc/test/src/btalloc.c @@ -0,0 +1,6 @@ +#include "test/jemalloc_test.h" + +void * +btalloc(size_t size, unsigned bits) { + return btalloc_0(size, bits); +} diff --git a/deps/jemalloc/test/src/btalloc_0.c b/deps/jemalloc/test/src/btalloc_0.c new file mode 100644 index 000000000..77d8904ea --- /dev/null +++ b/deps/jemalloc/test/src/btalloc_0.c @@ -0,0 +1,3 @@ +#include "test/jemalloc_test.h" + +btalloc_n_gen(0) diff --git a/deps/jemalloc/test/src/btalloc_1.c b/deps/jemalloc/test/src/btalloc_1.c new file mode 100644 index 000000000..4c126c309 --- /dev/null +++ b/deps/jemalloc/test/src/btalloc_1.c @@ -0,0 +1,3 @@ +#include "test/jemalloc_test.h" + +btalloc_n_gen(1) diff --git a/deps/jemalloc/test/src/math.c b/deps/jemalloc/test/src/math.c new file mode 100644 index 000000000..1758c6778 --- /dev/null +++ b/deps/jemalloc/test/src/math.c @@ -0,0 +1,2 @@ +#define MATH_C_ +#include "test/jemalloc_test.h" diff --git a/deps/jemalloc/test/src/mq.c b/deps/jemalloc/test/src/mq.c new file mode 100644 index 000000000..9b5f672d6 --- /dev/null +++ b/deps/jemalloc/test/src/mq.c @@ -0,0 +1,27 @@ +#include "test/jemalloc_test.h" + +/* + * Sleep for approximately ns nanoseconds. No lower *nor* upper bound on sleep + * time is guaranteed. + */ +void +mq_nanosleep(unsigned ns) { + assert(ns <= 1000*1000*1000); + +#ifdef _WIN32 + Sleep(ns / 1000); +#else + { + struct timespec timeout; + + if (ns < 1000*1000*1000) { + timeout.tv_sec = 0; + timeout.tv_nsec = ns; + } else { + timeout.tv_sec = 1; + timeout.tv_nsec = 0; + } + nanosleep(&timeout, NULL); + } +#endif +} diff --git a/deps/jemalloc/test/src/mtx.c b/deps/jemalloc/test/src/mtx.c new file mode 100644 index 000000000..d9ce375c3 --- /dev/null +++ b/deps/jemalloc/test/src/mtx.c @@ -0,0 +1,61 @@ +#include "test/jemalloc_test.h" + +#ifndef _CRT_SPINCOUNT +#define _CRT_SPINCOUNT 4000 +#endif + +bool +mtx_init(mtx_t *mtx) { +#ifdef _WIN32 + if (!InitializeCriticalSectionAndSpinCount(&mtx->lock, + _CRT_SPINCOUNT)) { + return true; + } +#elif (defined(JEMALLOC_OS_UNFAIR_LOCK)) + mtx->lock = OS_UNFAIR_LOCK_INIT; +#else + pthread_mutexattr_t attr; + + if (pthread_mutexattr_init(&attr) != 0) { + return true; + } + pthread_mutexattr_settype(&attr, PTHREAD_MUTEX_DEFAULT); + if (pthread_mutex_init(&mtx->lock, &attr) != 0) { + pthread_mutexattr_destroy(&attr); + return true; + } + pthread_mutexattr_destroy(&attr); +#endif + return false; +} + +void +mtx_fini(mtx_t *mtx) { +#ifdef _WIN32 +#elif (defined(JEMALLOC_OS_UNFAIR_LOCK)) +#else + pthread_mutex_destroy(&mtx->lock); +#endif +} + +void +mtx_lock(mtx_t *mtx) { +#ifdef _WIN32 + EnterCriticalSection(&mtx->lock); +#elif (defined(JEMALLOC_OS_UNFAIR_LOCK)) + os_unfair_lock_lock(&mtx->lock); +#else + pthread_mutex_lock(&mtx->lock); +#endif +} + +void +mtx_unlock(mtx_t *mtx) { +#ifdef _WIN32 + LeaveCriticalSection(&mtx->lock); +#elif (defined(JEMALLOC_OS_UNFAIR_LOCK)) + os_unfair_lock_unlock(&mtx->lock); +#else + pthread_mutex_unlock(&mtx->lock); +#endif +} diff --git a/deps/jemalloc/test/src/test.c b/deps/jemalloc/test/src/test.c new file mode 100644 index 000000000..f97ce4d18 --- /dev/null +++ b/deps/jemalloc/test/src/test.c @@ -0,0 +1,234 @@ +#include "test/jemalloc_test.h" + +/* Test status state. */ + +static unsigned test_count = 0; +static test_status_t test_counts[test_status_count] = {0, 0, 0}; +static test_status_t test_status = test_status_pass; +static const char * test_name = ""; + +/* Reentrancy testing helpers. */ + +#define NUM_REENTRANT_ALLOCS 20 +typedef enum { + non_reentrant = 0, + libc_reentrant = 1, + arena_new_reentrant = 2 +} reentrancy_t; +static reentrancy_t reentrancy; + +static bool libc_hook_ran = false; +static bool arena_new_hook_ran = false; + +static const char * +reentrancy_t_str(reentrancy_t r) { + switch (r) { + case non_reentrant: + return "non-reentrant"; + case libc_reentrant: + return "libc-reentrant"; + case arena_new_reentrant: + return "arena_new-reentrant"; + default: + unreachable(); + } +} + +static void +do_hook(bool *hook_ran, void (**hook)()) { + *hook_ran = true; + *hook = NULL; + + size_t alloc_size = 1; + for (int i = 0; i < NUM_REENTRANT_ALLOCS; i++) { + free(malloc(alloc_size)); + alloc_size *= 2; + } +} + +static void +libc_reentrancy_hook() { + do_hook(&libc_hook_ran, &test_hooks_libc_hook); +} + +static void +arena_new_reentrancy_hook() { + do_hook(&arena_new_hook_ran, &test_hooks_arena_new_hook); +} + +/* Actual test infrastructure. */ +bool +test_is_reentrant() { + return reentrancy != non_reentrant; +} + +JEMALLOC_FORMAT_PRINTF(1, 2) +void +test_skip(const char *format, ...) { + va_list ap; + + va_start(ap, format); + malloc_vcprintf(NULL, NULL, format, ap); + va_end(ap); + malloc_printf("\n"); + test_status = test_status_skip; +} + +JEMALLOC_FORMAT_PRINTF(1, 2) +void +test_fail(const char *format, ...) { + va_list ap; + + va_start(ap, format); + malloc_vcprintf(NULL, NULL, format, ap); + va_end(ap); + malloc_printf("\n"); + test_status = test_status_fail; +} + +static const char * +test_status_string(test_status_t test_status) { + switch (test_status) { + case test_status_pass: return "pass"; + case test_status_skip: return "skip"; + case test_status_fail: return "fail"; + default: not_reached(); + } +} + +void +p_test_init(const char *name) { + test_count++; + test_status = test_status_pass; + test_name = name; +} + +void +p_test_fini(void) { + test_counts[test_status]++; + malloc_printf("%s (%s): %s\n", test_name, reentrancy_t_str(reentrancy), + test_status_string(test_status)); +} + +static void +check_global_slow(test_status_t *status) { +#ifdef JEMALLOC_UNIT_TEST + /* + * This check needs to peek into tsd internals, which is why it's only + * exposed in unit tests. + */ + if (tsd_global_slow()) { + malloc_printf("Testing increased global slow count\n"); + *status = test_status_fail; + } +#endif +} + +static test_status_t +p_test_impl(bool do_malloc_init, bool do_reentrant, test_t *t, va_list ap) { + test_status_t ret; + + if (do_malloc_init) { + /* + * Make sure initialization occurs prior to running tests. + * Tests are special because they may use internal facilities + * prior to triggering initialization as a side effect of + * calling into the public API. + */ + if (nallocx(1, 0) == 0) { + malloc_printf("Initialization error"); + return test_status_fail; + } + } + + ret = test_status_pass; + for (; t != NULL; t = va_arg(ap, test_t *)) { + /* Non-reentrant run. */ + reentrancy = non_reentrant; + test_hooks_arena_new_hook = test_hooks_libc_hook = NULL; + t(); + if (test_status > ret) { + ret = test_status; + } + check_global_slow(&ret); + /* Reentrant run. */ + if (do_reentrant) { + reentrancy = libc_reentrant; + test_hooks_arena_new_hook = NULL; + test_hooks_libc_hook = &libc_reentrancy_hook; + t(); + if (test_status > ret) { + ret = test_status; + } + check_global_slow(&ret); + + reentrancy = arena_new_reentrant; + test_hooks_libc_hook = NULL; + test_hooks_arena_new_hook = &arena_new_reentrancy_hook; + t(); + if (test_status > ret) { + ret = test_status; + } + check_global_slow(&ret); + } + } + + malloc_printf("--- %s: %u/%u, %s: %u/%u, %s: %u/%u ---\n", + test_status_string(test_status_pass), + test_counts[test_status_pass], test_count, + test_status_string(test_status_skip), + test_counts[test_status_skip], test_count, + test_status_string(test_status_fail), + test_counts[test_status_fail], test_count); + + return ret; +} + +test_status_t +p_test(test_t *t, ...) { + test_status_t ret; + va_list ap; + + ret = test_status_pass; + va_start(ap, t); + ret = p_test_impl(true, true, t, ap); + va_end(ap); + + return ret; +} + +test_status_t +p_test_no_reentrancy(test_t *t, ...) { + test_status_t ret; + va_list ap; + + ret = test_status_pass; + va_start(ap, t); + ret = p_test_impl(true, false, t, ap); + va_end(ap); + + return ret; +} + +test_status_t +p_test_no_malloc_init(test_t *t, ...) { + test_status_t ret; + va_list ap; + + ret = test_status_pass; + va_start(ap, t); + /* + * We also omit reentrancy from bootstrapping tests, since we don't + * (yet) care about general reentrancy during bootstrapping. + */ + ret = p_test_impl(false, false, t, ap); + va_end(ap); + + return ret; +} + +void +p_test_fail(const char *prefix, const char *message) { + malloc_cprintf(NULL, NULL, "%s%s\n", prefix, message); + test_status = test_status_fail; +} diff --git a/deps/jemalloc/test/src/thd.c b/deps/jemalloc/test/src/thd.c new file mode 100644 index 000000000..9a15eabbf --- /dev/null +++ b/deps/jemalloc/test/src/thd.c @@ -0,0 +1,34 @@ +#include "test/jemalloc_test.h" + +#ifdef _WIN32 +void +thd_create(thd_t *thd, void *(*proc)(void *), void *arg) { + LPTHREAD_START_ROUTINE routine = (LPTHREAD_START_ROUTINE)proc; + *thd = CreateThread(NULL, 0, routine, arg, 0, NULL); + if (*thd == NULL) { + test_fail("Error in CreateThread()\n"); + } +} + +void +thd_join(thd_t thd, void **ret) { + if (WaitForSingleObject(thd, INFINITE) == WAIT_OBJECT_0 && ret) { + DWORD exit_code; + GetExitCodeThread(thd, (LPDWORD) &exit_code); + *ret = (void *)(uintptr_t)exit_code; + } +} + +#else +void +thd_create(thd_t *thd, void *(*proc)(void *), void *arg) { + if (pthread_create(thd, NULL, proc, arg) != 0) { + test_fail("Error in pthread_create()\n"); + } +} + +void +thd_join(thd_t thd, void **ret) { + pthread_join(thd, ret); +} +#endif diff --git a/deps/jemalloc/test/src/timer.c b/deps/jemalloc/test/src/timer.c new file mode 100644 index 000000000..c451c6391 --- /dev/null +++ b/deps/jemalloc/test/src/timer.c @@ -0,0 +1,56 @@ +#include "test/jemalloc_test.h" + +void +timer_start(timedelta_t *timer) { + nstime_init(&timer->t0, 0); + nstime_update(&timer->t0); +} + +void +timer_stop(timedelta_t *timer) { + nstime_copy(&timer->t1, &timer->t0); + nstime_update(&timer->t1); +} + +uint64_t +timer_usec(const timedelta_t *timer) { + nstime_t delta; + + nstime_copy(&delta, &timer->t1); + nstime_subtract(&delta, &timer->t0); + return nstime_ns(&delta) / 1000; +} + +void +timer_ratio(timedelta_t *a, timedelta_t *b, char *buf, size_t buflen) { + uint64_t t0 = timer_usec(a); + uint64_t t1 = timer_usec(b); + uint64_t mult; + size_t i = 0; + size_t j, n; + + /* Whole. */ + n = malloc_snprintf(&buf[i], buflen-i, "%"FMTu64, t0 / t1); + i += n; + if (i >= buflen) { + return; + } + mult = 1; + for (j = 0; j < n; j++) { + mult *= 10; + } + + /* Decimal. */ + n = malloc_snprintf(&buf[i], buflen-i, "."); + i += n; + + /* Fraction. */ + while (i < buflen-1) { + uint64_t round = (i+1 == buflen-1 && ((t0 * mult * 10 / t1) % 10 + >= 5)) ? 1 : 0; + n = malloc_snprintf(&buf[i], buflen-i, + "%"FMTu64, (t0 * mult / t1) % 10 + round); + i += n; + mult *= 10; + } +} diff --git a/deps/jemalloc/test/stress/hookbench.c b/deps/jemalloc/test/stress/hookbench.c new file mode 100644 index 000000000..97e90b0e3 --- /dev/null +++ b/deps/jemalloc/test/stress/hookbench.c @@ -0,0 +1,73 @@ +#include "test/jemalloc_test.h" + +static void +noop_alloc_hook(void *extra, hook_alloc_t type, void *result, + uintptr_t result_raw, uintptr_t args_raw[3]) { +} + +static void +noop_dalloc_hook(void *extra, hook_dalloc_t type, void *address, + uintptr_t args_raw[3]) { +} + +static void +noop_expand_hook(void *extra, hook_expand_t type, void *address, + size_t old_usize, size_t new_usize, uintptr_t result_raw, + uintptr_t args_raw[4]) { +} + +static void +malloc_free_loop(int iters) { + for (int i = 0; i < iters; i++) { + void *p = mallocx(1, 0); + free(p); + } +} + +static void +test_hooked(int iters) { + hooks_t hooks = {&noop_alloc_hook, &noop_dalloc_hook, &noop_expand_hook, + NULL}; + + int err; + void *handles[HOOK_MAX]; + size_t sz = sizeof(handles[0]); + + for (int i = 0; i < HOOK_MAX; i++) { + err = mallctl("experimental.hooks.install", &handles[i], + &sz, &hooks, sizeof(hooks)); + assert(err == 0); + + timedelta_t timer; + timer_start(&timer); + malloc_free_loop(iters); + timer_stop(&timer); + malloc_printf("With %d hook%s: %"FMTu64"us\n", i + 1, + i + 1 == 1 ? "" : "s", timer_usec(&timer)); + } + for (int i = 0; i < HOOK_MAX; i++) { + err = mallctl("experimental.hooks.remove", NULL, NULL, + &handles[i], sizeof(handles[i])); + assert(err == 0); + } +} + +static void +test_unhooked(int iters) { + timedelta_t timer; + timer_start(&timer); + malloc_free_loop(iters); + timer_stop(&timer); + + malloc_printf("Without hooks: %"FMTu64"us\n", timer_usec(&timer)); +} + +int +main(void) { + /* Initialize */ + free(mallocx(1, 0)); + int iters = 10 * 1000 * 1000; + malloc_printf("Benchmarking hooks with %d iterations:\n", iters); + test_hooked(iters); + test_unhooked(iters); +} diff --git a/deps/jemalloc/test/stress/microbench.c b/deps/jemalloc/test/stress/microbench.c new file mode 100644 index 000000000..988b7938f --- /dev/null +++ b/deps/jemalloc/test/stress/microbench.c @@ -0,0 +1,165 @@ +#include "test/jemalloc_test.h" + +static inline void +time_func(timedelta_t *timer, uint64_t nwarmup, uint64_t niter, + void (*func)(void)) { + uint64_t i; + + for (i = 0; i < nwarmup; i++) { + func(); + } + timer_start(timer); + for (i = 0; i < niter; i++) { + func(); + } + timer_stop(timer); +} + +void +compare_funcs(uint64_t nwarmup, uint64_t niter, const char *name_a, + void (*func_a), const char *name_b, void (*func_b)) { + timedelta_t timer_a, timer_b; + char ratio_buf[6]; + void *p; + + p = mallocx(1, 0); + if (p == NULL) { + test_fail("Unexpected mallocx() failure"); + return; + } + + time_func(&timer_a, nwarmup, niter, func_a); + time_func(&timer_b, nwarmup, niter, func_b); + + timer_ratio(&timer_a, &timer_b, ratio_buf, sizeof(ratio_buf)); + malloc_printf("%"FMTu64" iterations, %s=%"FMTu64"us, " + "%s=%"FMTu64"us, ratio=1:%s\n", + niter, name_a, timer_usec(&timer_a), name_b, timer_usec(&timer_b), + ratio_buf); + + dallocx(p, 0); +} + +static void +malloc_free(void) { + /* The compiler can optimize away free(malloc(1))! */ + void *p = malloc(1); + if (p == NULL) { + test_fail("Unexpected malloc() failure"); + return; + } + free(p); +} + +static void +mallocx_free(void) { + void *p = mallocx(1, 0); + if (p == NULL) { + test_fail("Unexpected mallocx() failure"); + return; + } + free(p); +} + +TEST_BEGIN(test_malloc_vs_mallocx) { + compare_funcs(10*1000*1000, 100*1000*1000, "malloc", + malloc_free, "mallocx", mallocx_free); +} +TEST_END + +static void +malloc_dallocx(void) { + void *p = malloc(1); + if (p == NULL) { + test_fail("Unexpected malloc() failure"); + return; + } + dallocx(p, 0); +} + +static void +malloc_sdallocx(void) { + void *p = malloc(1); + if (p == NULL) { + test_fail("Unexpected malloc() failure"); + return; + } + sdallocx(p, 1, 0); +} + +TEST_BEGIN(test_free_vs_dallocx) { + compare_funcs(10*1000*1000, 100*1000*1000, "free", malloc_free, + "dallocx", malloc_dallocx); +} +TEST_END + +TEST_BEGIN(test_dallocx_vs_sdallocx) { + compare_funcs(10*1000*1000, 100*1000*1000, "dallocx", malloc_dallocx, + "sdallocx", malloc_sdallocx); +} +TEST_END + +static void +malloc_mus_free(void) { + void *p; + + p = malloc(1); + if (p == NULL) { + test_fail("Unexpected malloc() failure"); + return; + } + malloc_usable_size(p); + free(p); +} + +static void +malloc_sallocx_free(void) { + void *p; + + p = malloc(1); + if (p == NULL) { + test_fail("Unexpected malloc() failure"); + return; + } + if (sallocx(p, 0) < 1) { + test_fail("Unexpected sallocx() failure"); + } + free(p); +} + +TEST_BEGIN(test_mus_vs_sallocx) { + compare_funcs(10*1000*1000, 100*1000*1000, "malloc_usable_size", + malloc_mus_free, "sallocx", malloc_sallocx_free); +} +TEST_END + +static void +malloc_nallocx_free(void) { + void *p; + + p = malloc(1); + if (p == NULL) { + test_fail("Unexpected malloc() failure"); + return; + } + if (nallocx(1, 0) < 1) { + test_fail("Unexpected nallocx() failure"); + } + free(p); +} + +TEST_BEGIN(test_sallocx_vs_nallocx) { + compare_funcs(10*1000*1000, 100*1000*1000, "sallocx", + malloc_sallocx_free, "nallocx", malloc_nallocx_free); +} +TEST_END + +int +main(void) { + return test_no_reentrancy( + test_malloc_vs_mallocx, + test_free_vs_dallocx, + test_dallocx_vs_sdallocx, + test_mus_vs_sallocx, + test_sallocx_vs_nallocx); +} diff --git a/deps/jemalloc/test/test.sh.in b/deps/jemalloc/test/test.sh.in new file mode 100644 index 000000000..39302fff4 --- /dev/null +++ b/deps/jemalloc/test/test.sh.in @@ -0,0 +1,80 @@ +#!/bin/sh + +case @abi@ in + macho) + export DYLD_FALLBACK_LIBRARY_PATH="@objroot@lib" + ;; + pecoff) + export PATH="${PATH}:@objroot@lib" + ;; + *) + ;; +esac + +# Make a copy of the @JEMALLOC_CPREFIX@MALLOC_CONF passed in to this script, so +# it can be repeatedly concatenated with per test settings. +export MALLOC_CONF_ALL=${@JEMALLOC_CPREFIX@MALLOC_CONF} +# Concatenate the individual test's MALLOC_CONF and MALLOC_CONF_ALL. +export_malloc_conf() { + if [ "x${MALLOC_CONF}" != "x" -a "x${MALLOC_CONF_ALL}" != "x" ] ; then + export @JEMALLOC_CPREFIX@MALLOC_CONF="${MALLOC_CONF},${MALLOC_CONF_ALL}" + else + export @JEMALLOC_CPREFIX@MALLOC_CONF="${MALLOC_CONF}${MALLOC_CONF_ALL}" + fi +} + +# Corresponds to test_status_t. +pass_code=0 +skip_code=1 +fail_code=2 + +pass_count=0 +skip_count=0 +fail_count=0 +for t in $@; do + if [ $pass_count -ne 0 -o $skip_count -ne 0 -o $fail_count != 0 ] ; then + echo + fi + echo "=== ${t} ===" + if [ -e "@srcroot@${t}.sh" ] ; then + # Source the shell script corresponding to the test in a subshell and + # execute the test. This allows the shell script to set MALLOC_CONF, which + # is then used to set @JEMALLOC_CPREFIX@MALLOC_CONF (thus allowing the + # per test shell script to ignore the @JEMALLOC_CPREFIX@ detail). + enable_fill=@enable_fill@ \ + enable_prof=@enable_prof@ \ + . @srcroot@${t}.sh && \ + export_malloc_conf && \ + $JEMALLOC_TEST_PREFIX ${t}@exe@ @abs_srcroot@ @abs_objroot@ + else + export MALLOC_CONF= && \ + export_malloc_conf && \ + $JEMALLOC_TEST_PREFIX ${t}@exe@ @abs_srcroot@ @abs_objroot@ + fi + result_code=$? + case ${result_code} in + ${pass_code}) + pass_count=$((pass_count+1)) + ;; + ${skip_code}) + skip_count=$((skip_count+1)) + ;; + ${fail_code}) + fail_count=$((fail_count+1)) + ;; + *) + echo "Test harness error: ${t} w/ MALLOC_CONF=\"${MALLOC_CONF}\"" 1>&2 + echo "Use prefix to debug, e.g. JEMALLOC_TEST_PREFIX=\"gdb --args\" sh test/test.sh ${t}" 1>&2 + exit 1 + esac +done + +total_count=`expr ${pass_count} + ${skip_count} + ${fail_count}` +echo +echo "Test suite summary: pass: ${pass_count}/${total_count}, skip: ${skip_count}/${total_count}, fail: ${fail_count}/${total_count}" + +if [ ${fail_count} -eq 0 ] ; then + exit 0 +else + exit 1 +fi diff --git a/deps/jemalloc/test/unit/SFMT.c b/deps/jemalloc/test/unit/SFMT.c new file mode 100644 index 000000000..1fc8cf1bc --- /dev/null +++ b/deps/jemalloc/test/unit/SFMT.c @@ -0,0 +1,1599 @@ +/* + * This file derives from SFMT 1.3.3 + * (http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/SFMT/index.html), which was + * released under the terms of the following license: + * + * Copyright (c) 2006,2007 Mutsuo Saito, Makoto Matsumoto and Hiroshima + * University. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions are + * met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials provided + * with the distribution. + * * Neither the name of the Hiroshima University nor the names of + * its contributors may be used to endorse or promote products + * derived from this software without specific prior written + * permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ +#include "test/jemalloc_test.h" + +#define BLOCK_SIZE 10000 +#define BLOCK_SIZE64 (BLOCK_SIZE / 2) +#define COUNT_1 1000 +#define COUNT_2 700 + +static const uint32_t init_gen_rand_32_expected[] = { + 3440181298U, 1564997079U, 1510669302U, 2930277156U, 1452439940U, + 3796268453U, 423124208U, 2143818589U, 3827219408U, 2987036003U, + 2674978610U, 1536842514U, 2027035537U, 2534897563U, 1686527725U, + 545368292U, 1489013321U, 1370534252U, 4231012796U, 3994803019U, + 1764869045U, 824597505U, 862581900U, 2469764249U, 812862514U, + 359318673U, 116957936U, 3367389672U, 2327178354U, 1898245200U, + 3206507879U, 2378925033U, 1040214787U, 2524778605U, 3088428700U, + 1417665896U, 964324147U, 2282797708U, 2456269299U, 313400376U, + 2245093271U, 1015729427U, 2694465011U, 3246975184U, 1992793635U, + 463679346U, 3721104591U, 3475064196U, 856141236U, 1499559719U, + 3522818941U, 3721533109U, 1954826617U, 1282044024U, 1543279136U, + 1301863085U, 2669145051U, 4221477354U, 3896016841U, 3392740262U, + 462466863U, 1037679449U, 1228140306U, 922298197U, 1205109853U, + 1872938061U, 3102547608U, 2742766808U, 1888626088U, 4028039414U, + 157593879U, 1136901695U, 4038377686U, 3572517236U, 4231706728U, + 2997311961U, 1189931652U, 3981543765U, 2826166703U, 87159245U, + 1721379072U, 3897926942U, 1790395498U, 2569178939U, 1047368729U, + 2340259131U, 3144212906U, 2301169789U, 2442885464U, 3034046771U, + 3667880593U, 3935928400U, 2372805237U, 1666397115U, 2460584504U, + 513866770U, 3810869743U, 2147400037U, 2792078025U, 2941761810U, + 3212265810U, 984692259U, 346590253U, 1804179199U, 3298543443U, + 750108141U, 2880257022U, 243310542U, 1869036465U, 1588062513U, + 2983949551U, 1931450364U, 4034505847U, 2735030199U, 1628461061U, + 2539522841U, 127965585U, 3992448871U, 913388237U, 559130076U, + 1202933193U, 4087643167U, 2590021067U, 2256240196U, 1746697293U, + 1013913783U, 1155864921U, 2715773730U, 915061862U, 1948766573U, + 2322882854U, 3761119102U, 1343405684U, 3078711943U, 3067431651U, + 3245156316U, 3588354584U, 3484623306U, 3899621563U, 4156689741U, + 3237090058U, 3880063844U, 862416318U, 4039923869U, 2303788317U, + 3073590536U, 701653667U, 2131530884U, 3169309950U, 2028486980U, + 747196777U, 3620218225U, 432016035U, 1449580595U, 2772266392U, + 444224948U, 1662832057U, 3184055582U, 3028331792U, 1861686254U, + 1104864179U, 342430307U, 1350510923U, 3024656237U, 1028417492U, + 2870772950U, 290847558U, 3675663500U, 508431529U, 4264340390U, + 2263569913U, 1669302976U, 519511383U, 2706411211U, 3764615828U, + 3883162495U, 4051445305U, 2412729798U, 3299405164U, 3991911166U, + 2348767304U, 2664054906U, 3763609282U, 593943581U, 3757090046U, + 2075338894U, 2020550814U, 4287452920U, 4290140003U, 1422957317U, + 2512716667U, 2003485045U, 2307520103U, 2288472169U, 3940751663U, + 4204638664U, 2892583423U, 1710068300U, 3904755993U, 2363243951U, + 3038334120U, 547099465U, 771105860U, 3199983734U, 4282046461U, + 2298388363U, 934810218U, 2837827901U, 3952500708U, 2095130248U, + 3083335297U, 26885281U, 3932155283U, 1531751116U, 1425227133U, + 495654159U, 3279634176U, 3855562207U, 3957195338U, 4159985527U, + 893375062U, 1875515536U, 1327247422U, 3754140693U, 1028923197U, + 1729880440U, 805571298U, 448971099U, 2726757106U, 2749436461U, + 2485987104U, 175337042U, 3235477922U, 3882114302U, 2020970972U, + 943926109U, 2762587195U, 1904195558U, 3452650564U, 108432281U, + 3893463573U, 3977583081U, 2636504348U, 1110673525U, 3548479841U, + 4258854744U, 980047703U, 4057175418U, 3890008292U, 145653646U, + 3141868989U, 3293216228U, 1194331837U, 1254570642U, 3049934521U, + 2868313360U, 2886032750U, 1110873820U, 279553524U, 3007258565U, + 1104807822U, 3186961098U, 315764646U, 2163680838U, 3574508994U, + 3099755655U, 191957684U, 3642656737U, 3317946149U, 3522087636U, + 444526410U, 779157624U, 1088229627U, 1092460223U, 1856013765U, + 3659877367U, 368270451U, 503570716U, 3000984671U, 2742789647U, + 928097709U, 2914109539U, 308843566U, 2816161253U, 3667192079U, + 2762679057U, 3395240989U, 2928925038U, 1491465914U, 3458702834U, + 3787782576U, 2894104823U, 1296880455U, 1253636503U, 989959407U, + 2291560361U, 2776790436U, 1913178042U, 1584677829U, 689637520U, + 1898406878U, 688391508U, 3385234998U, 845493284U, 1943591856U, + 2720472050U, 222695101U, 1653320868U, 2904632120U, 4084936008U, + 1080720688U, 3938032556U, 387896427U, 2650839632U, 99042991U, + 1720913794U, 1047186003U, 1877048040U, 2090457659U, 517087501U, + 4172014665U, 2129713163U, 2413533132U, 2760285054U, 4129272496U, + 1317737175U, 2309566414U, 2228873332U, 3889671280U, 1110864630U, + 3576797776U, 2074552772U, 832002644U, 3097122623U, 2464859298U, + 2679603822U, 1667489885U, 3237652716U, 1478413938U, 1719340335U, + 2306631119U, 639727358U, 3369698270U, 226902796U, 2099920751U, + 1892289957U, 2201594097U, 3508197013U, 3495811856U, 3900381493U, + 841660320U, 3974501451U, 3360949056U, 1676829340U, 728899254U, + 2047809627U, 2390948962U, 670165943U, 3412951831U, 4189320049U, + 1911595255U, 2055363086U, 507170575U, 418219594U, 4141495280U, + 2692088692U, 4203630654U, 3540093932U, 791986533U, 2237921051U, + 2526864324U, 2956616642U, 1394958700U, 1983768223U, 1893373266U, + 591653646U, 228432437U, 1611046598U, 3007736357U, 1040040725U, + 2726180733U, 2789804360U, 4263568405U, 829098158U, 3847722805U, + 1123578029U, 1804276347U, 997971319U, 4203797076U, 4185199713U, + 2811733626U, 2343642194U, 2985262313U, 1417930827U, 3759587724U, + 1967077982U, 1585223204U, 1097475516U, 1903944948U, 740382444U, + 1114142065U, 1541796065U, 1718384172U, 1544076191U, 1134682254U, + 3519754455U, 2866243923U, 341865437U, 645498576U, 2690735853U, + 1046963033U, 2493178460U, 1187604696U, 1619577821U, 488503634U, + 3255768161U, 2306666149U, 1630514044U, 2377698367U, 2751503746U, + 3794467088U, 1796415981U, 3657173746U, 409136296U, 1387122342U, + 1297726519U, 219544855U, 4270285558U, 437578827U, 1444698679U, + 2258519491U, 963109892U, 3982244073U, 3351535275U, 385328496U, + 1804784013U, 698059346U, 3920535147U, 708331212U, 784338163U, + 785678147U, 1238376158U, 1557298846U, 2037809321U, 271576218U, + 4145155269U, 1913481602U, 2763691931U, 588981080U, 1201098051U, + 3717640232U, 1509206239U, 662536967U, 3180523616U, 1133105435U, + 2963500837U, 2253971215U, 3153642623U, 1066925709U, 2582781958U, + 3034720222U, 1090798544U, 2942170004U, 4036187520U, 686972531U, + 2610990302U, 2641437026U, 1837562420U, 722096247U, 1315333033U, + 2102231203U, 3402389208U, 3403698140U, 1312402831U, 2898426558U, + 814384596U, 385649582U, 1916643285U, 1924625106U, 2512905582U, + 2501170304U, 4275223366U, 2841225246U, 1467663688U, 3563567847U, + 2969208552U, 884750901U, 102992576U, 227844301U, 3681442994U, + 3502881894U, 4034693299U, 1166727018U, 1697460687U, 1737778332U, + 1787161139U, 1053003655U, 1215024478U, 2791616766U, 2525841204U, + 1629323443U, 3233815U, 2003823032U, 3083834263U, 2379264872U, + 3752392312U, 1287475550U, 3770904171U, 3004244617U, 1502117784U, + 918698423U, 2419857538U, 3864502062U, 1751322107U, 2188775056U, + 4018728324U, 983712955U, 440071928U, 3710838677U, 2001027698U, + 3994702151U, 22493119U, 3584400918U, 3446253670U, 4254789085U, + 1405447860U, 1240245579U, 1800644159U, 1661363424U, 3278326132U, + 3403623451U, 67092802U, 2609352193U, 3914150340U, 1814842761U, + 3610830847U, 591531412U, 3880232807U, 1673505890U, 2585326991U, + 1678544474U, 3148435887U, 3457217359U, 1193226330U, 2816576908U, + 154025329U, 121678860U, 1164915738U, 973873761U, 269116100U, + 52087970U, 744015362U, 498556057U, 94298882U, 1563271621U, + 2383059628U, 4197367290U, 3958472990U, 2592083636U, 2906408439U, + 1097742433U, 3924840517U, 264557272U, 2292287003U, 3203307984U, + 4047038857U, 3820609705U, 2333416067U, 1839206046U, 3600944252U, + 3412254904U, 583538222U, 2390557166U, 4140459427U, 2810357445U, + 226777499U, 2496151295U, 2207301712U, 3283683112U, 611630281U, + 1933218215U, 3315610954U, 3889441987U, 3719454256U, 3957190521U, + 1313998161U, 2365383016U, 3146941060U, 1801206260U, 796124080U, + 2076248581U, 1747472464U, 3254365145U, 595543130U, 3573909503U, + 3758250204U, 2020768540U, 2439254210U, 93368951U, 3155792250U, + 2600232980U, 3709198295U, 3894900440U, 2971850836U, 1578909644U, + 1443493395U, 2581621665U, 3086506297U, 2443465861U, 558107211U, + 1519367835U, 249149686U, 908102264U, 2588765675U, 1232743965U, + 1001330373U, 3561331654U, 2259301289U, 1564977624U, 3835077093U, + 727244906U, 4255738067U, 1214133513U, 2570786021U, 3899704621U, + 1633861986U, 1636979509U, 1438500431U, 58463278U, 2823485629U, + 2297430187U, 2926781924U, 3371352948U, 1864009023U, 2722267973U, + 1444292075U, 437703973U, 1060414512U, 189705863U, 910018135U, + 4077357964U, 884213423U, 2644986052U, 3973488374U, 1187906116U, + 2331207875U, 780463700U, 3713351662U, 3854611290U, 412805574U, + 2978462572U, 2176222820U, 829424696U, 2790788332U, 2750819108U, + 1594611657U, 3899878394U, 3032870364U, 1702887682U, 1948167778U, + 14130042U, 192292500U, 947227076U, 90719497U, 3854230320U, + 784028434U, 2142399787U, 1563449646U, 2844400217U, 819143172U, + 2883302356U, 2328055304U, 1328532246U, 2603885363U, 3375188924U, + 933941291U, 3627039714U, 2129697284U, 2167253953U, 2506905438U, + 1412424497U, 2981395985U, 1418359660U, 2925902456U, 52752784U, + 3713667988U, 3924669405U, 648975707U, 1145520213U, 4018650664U, + 3805915440U, 2380542088U, 2013260958U, 3262572197U, 2465078101U, + 1114540067U, 3728768081U, 2396958768U, 590672271U, 904818725U, + 4263660715U, 700754408U, 1042601829U, 4094111823U, 4274838909U, + 2512692617U, 2774300207U, 2057306915U, 3470942453U, 99333088U, + 1142661026U, 2889931380U, 14316674U, 2201179167U, 415289459U, + 448265759U, 3515142743U, 3254903683U, 246633281U, 1184307224U, + 2418347830U, 2092967314U, 2682072314U, 2558750234U, 2000352263U, + 1544150531U, 399010405U, 1513946097U, 499682937U, 461167460U, + 3045570638U, 1633669705U, 851492362U, 4052801922U, 2055266765U, + 635556996U, 368266356U, 2385737383U, 3218202352U, 2603772408U, + 349178792U, 226482567U, 3102426060U, 3575998268U, 2103001871U, + 3243137071U, 225500688U, 1634718593U, 4283311431U, 4292122923U, + 3842802787U, 811735523U, 105712518U, 663434053U, 1855889273U, + 2847972595U, 1196355421U, 2552150115U, 4254510614U, 3752181265U, + 3430721819U, 3828705396U, 3436287905U, 3441964937U, 4123670631U, + 353001539U, 459496439U, 3799690868U, 1293777660U, 2761079737U, + 498096339U, 3398433374U, 4080378380U, 2304691596U, 2995729055U, + 4134660419U, 3903444024U, 3576494993U, 203682175U, 3321164857U, + 2747963611U, 79749085U, 2992890370U, 1240278549U, 1772175713U, + 2111331972U, 2655023449U, 1683896345U, 2836027212U, 3482868021U, + 2489884874U, 756853961U, 2298874501U, 4013448667U, 4143996022U, + 2948306858U, 4132920035U, 1283299272U, 995592228U, 3450508595U, + 1027845759U, 1766942720U, 3861411826U, 1446861231U, 95974993U, + 3502263554U, 1487532194U, 601502472U, 4129619129U, 250131773U, + 2050079547U, 3198903947U, 3105589778U, 4066481316U, 3026383978U, + 2276901713U, 365637751U, 2260718426U, 1394775634U, 1791172338U, + 2690503163U, 2952737846U, 1568710462U, 732623190U, 2980358000U, + 1053631832U, 1432426951U, 3229149635U, 1854113985U, 3719733532U, + 3204031934U, 735775531U, 107468620U, 3734611984U, 631009402U, + 3083622457U, 4109580626U, 159373458U, 1301970201U, 4132389302U, + 1293255004U, 847182752U, 4170022737U, 96712900U, 2641406755U, + 1381727755U, 405608287U, 4287919625U, 1703554290U, 3589580244U, + 2911403488U, 2166565U, 2647306451U, 2330535117U, 1200815358U, + 1165916754U, 245060911U, 4040679071U, 3684908771U, 2452834126U, + 2486872773U, 2318678365U, 2940627908U, 1837837240U, 3447897409U, + 4270484676U, 1495388728U, 3754288477U, 4204167884U, 1386977705U, + 2692224733U, 3076249689U, 4109568048U, 4170955115U, 4167531356U, + 4020189950U, 4261855038U, 3036907575U, 3410399885U, 3076395737U, + 1046178638U, 144496770U, 230725846U, 3349637149U, 17065717U, + 2809932048U, 2054581785U, 3608424964U, 3259628808U, 134897388U, + 3743067463U, 257685904U, 3795656590U, 1562468719U, 3589103904U, + 3120404710U, 254684547U, 2653661580U, 3663904795U, 2631942758U, + 1063234347U, 2609732900U, 2332080715U, 3521125233U, 1180599599U, + 1935868586U, 4110970440U, 296706371U, 2128666368U, 1319875791U, + 1570900197U, 3096025483U, 1799882517U, 1928302007U, 1163707758U, + 1244491489U, 3533770203U, 567496053U, 2757924305U, 2781639343U, + 2818420107U, 560404889U, 2619609724U, 4176035430U, 2511289753U, + 2521842019U, 3910553502U, 2926149387U, 3302078172U, 4237118867U, + 330725126U, 367400677U, 888239854U, 545570454U, 4259590525U, + 134343617U, 1102169784U, 1647463719U, 3260979784U, 1518840883U, + 3631537963U, 3342671457U, 1301549147U, 2083739356U, 146593792U, + 3217959080U, 652755743U, 2032187193U, 3898758414U, 1021358093U, + 4037409230U, 2176407931U, 3427391950U, 2883553603U, 985613827U, + 3105265092U, 3423168427U, 3387507672U, 467170288U, 2141266163U, + 3723870208U, 916410914U, 1293987799U, 2652584950U, 769160137U, + 3205292896U, 1561287359U, 1684510084U, 3136055621U, 3765171391U, + 639683232U, 2639569327U, 1218546948U, 4263586685U, 3058215773U, + 2352279820U, 401870217U, 2625822463U, 1529125296U, 2981801895U, + 1191285226U, 4027725437U, 3432700217U, 4098835661U, 971182783U, + 2443861173U, 3881457123U, 3874386651U, 457276199U, 2638294160U, + 4002809368U, 421169044U, 1112642589U, 3076213779U, 3387033971U, + 2499610950U, 3057240914U, 1662679783U, 461224431U, 1168395933U +}; +static const uint32_t init_by_array_32_expected[] = { + 2920711183U, 3885745737U, 3501893680U, 856470934U, 1421864068U, + 277361036U, 1518638004U, 2328404353U, 3355513634U, 64329189U, + 1624587673U, 3508467182U, 2481792141U, 3706480799U, 1925859037U, + 2913275699U, 882658412U, 384641219U, 422202002U, 1873384891U, + 2006084383U, 3924929912U, 1636718106U, 3108838742U, 1245465724U, + 4195470535U, 779207191U, 1577721373U, 1390469554U, 2928648150U, + 121399709U, 3170839019U, 4044347501U, 953953814U, 3821710850U, + 3085591323U, 3666535579U, 3577837737U, 2012008410U, 3565417471U, + 4044408017U, 433600965U, 1637785608U, 1798509764U, 860770589U, + 3081466273U, 3982393409U, 2451928325U, 3437124742U, 4093828739U, + 3357389386U, 2154596123U, 496568176U, 2650035164U, 2472361850U, + 3438299U, 2150366101U, 1577256676U, 3802546413U, 1787774626U, + 4078331588U, 3706103141U, 170391138U, 3806085154U, 1680970100U, + 1961637521U, 3316029766U, 890610272U, 1453751581U, 1430283664U, + 3051057411U, 3597003186U, 542563954U, 3796490244U, 1690016688U, + 3448752238U, 440702173U, 347290497U, 1121336647U, 2540588620U, + 280881896U, 2495136428U, 213707396U, 15104824U, 2946180358U, + 659000016U, 566379385U, 2614030979U, 2855760170U, 334526548U, + 2315569495U, 2729518615U, 564745877U, 1263517638U, 3157185798U, + 1604852056U, 1011639885U, 2950579535U, 2524219188U, 312951012U, + 1528896652U, 1327861054U, 2846910138U, 3966855905U, 2536721582U, + 855353911U, 1685434729U, 3303978929U, 1624872055U, 4020329649U, + 3164802143U, 1642802700U, 1957727869U, 1792352426U, 3334618929U, + 2631577923U, 3027156164U, 842334259U, 3353446843U, 1226432104U, + 1742801369U, 3552852535U, 3471698828U, 1653910186U, 3380330939U, + 2313782701U, 3351007196U, 2129839995U, 1800682418U, 4085884420U, + 1625156629U, 3669701987U, 615211810U, 3294791649U, 4131143784U, + 2590843588U, 3207422808U, 3275066464U, 561592872U, 3957205738U, + 3396578098U, 48410678U, 3505556445U, 1005764855U, 3920606528U, + 2936980473U, 2378918600U, 2404449845U, 1649515163U, 701203563U, + 3705256349U, 83714199U, 3586854132U, 922978446U, 2863406304U, + 3523398907U, 2606864832U, 2385399361U, 3171757816U, 4262841009U, + 3645837721U, 1169579486U, 3666433897U, 3174689479U, 1457866976U, + 3803895110U, 3346639145U, 1907224409U, 1978473712U, 1036712794U, + 980754888U, 1302782359U, 1765252468U, 459245755U, 3728923860U, + 1512894209U, 2046491914U, 207860527U, 514188684U, 2288713615U, + 1597354672U, 3349636117U, 2357291114U, 3995796221U, 945364213U, + 1893326518U, 3770814016U, 1691552714U, 2397527410U, 967486361U, + 776416472U, 4197661421U, 951150819U, 1852770983U, 4044624181U, + 1399439738U, 4194455275U, 2284037669U, 1550734958U, 3321078108U, + 1865235926U, 2912129961U, 2664980877U, 1357572033U, 2600196436U, + 2486728200U, 2372668724U, 1567316966U, 2374111491U, 1839843570U, + 20815612U, 3727008608U, 3871996229U, 824061249U, 1932503978U, + 3404541726U, 758428924U, 2609331364U, 1223966026U, 1299179808U, + 648499352U, 2180134401U, 880821170U, 3781130950U, 113491270U, + 1032413764U, 4185884695U, 2490396037U, 1201932817U, 4060951446U, + 4165586898U, 1629813212U, 2887821158U, 415045333U, 628926856U, + 2193466079U, 3391843445U, 2227540681U, 1907099846U, 2848448395U, + 1717828221U, 1372704537U, 1707549841U, 2294058813U, 2101214437U, + 2052479531U, 1695809164U, 3176587306U, 2632770465U, 81634404U, + 1603220563U, 644238487U, 302857763U, 897352968U, 2613146653U, + 1391730149U, 4245717312U, 4191828749U, 1948492526U, 2618174230U, + 3992984522U, 2178852787U, 3596044509U, 3445573503U, 2026614616U, + 915763564U, 3415689334U, 2532153403U, 3879661562U, 2215027417U, + 3111154986U, 2929478371U, 668346391U, 1152241381U, 2632029711U, + 3004150659U, 2135025926U, 948690501U, 2799119116U, 4228829406U, + 1981197489U, 4209064138U, 684318751U, 3459397845U, 201790843U, + 4022541136U, 3043635877U, 492509624U, 3263466772U, 1509148086U, + 921459029U, 3198857146U, 705479721U, 3835966910U, 3603356465U, + 576159741U, 1742849431U, 594214882U, 2055294343U, 3634861861U, + 449571793U, 3246390646U, 3868232151U, 1479156585U, 2900125656U, + 2464815318U, 3960178104U, 1784261920U, 18311476U, 3627135050U, + 644609697U, 424968996U, 919890700U, 2986824110U, 816423214U, + 4003562844U, 1392714305U, 1757384428U, 2569030598U, 995949559U, + 3875659880U, 2933807823U, 2752536860U, 2993858466U, 4030558899U, + 2770783427U, 2775406005U, 2777781742U, 1931292655U, 472147933U, + 3865853827U, 2726470545U, 2668412860U, 2887008249U, 408979190U, + 3578063323U, 3242082049U, 1778193530U, 27981909U, 2362826515U, + 389875677U, 1043878156U, 581653903U, 3830568952U, 389535942U, + 3713523185U, 2768373359U, 2526101582U, 1998618197U, 1160859704U, + 3951172488U, 1098005003U, 906275699U, 3446228002U, 2220677963U, + 2059306445U, 132199571U, 476838790U, 1868039399U, 3097344807U, + 857300945U, 396345050U, 2835919916U, 1782168828U, 1419519470U, + 4288137521U, 819087232U, 596301494U, 872823172U, 1526888217U, + 805161465U, 1116186205U, 2829002754U, 2352620120U, 620121516U, + 354159268U, 3601949785U, 209568138U, 1352371732U, 2145977349U, + 4236871834U, 1539414078U, 3558126206U, 3224857093U, 4164166682U, + 3817553440U, 3301780278U, 2682696837U, 3734994768U, 1370950260U, + 1477421202U, 2521315749U, 1330148125U, 1261554731U, 2769143688U, + 3554756293U, 4235882678U, 3254686059U, 3530579953U, 1215452615U, + 3574970923U, 4057131421U, 589224178U, 1000098193U, 171190718U, + 2521852045U, 2351447494U, 2284441580U, 2646685513U, 3486933563U, + 3789864960U, 1190528160U, 1702536782U, 1534105589U, 4262946827U, + 2726686826U, 3584544841U, 2348270128U, 2145092281U, 2502718509U, + 1027832411U, 3571171153U, 1287361161U, 4011474411U, 3241215351U, + 2419700818U, 971242709U, 1361975763U, 1096842482U, 3271045537U, + 81165449U, 612438025U, 3912966678U, 1356929810U, 733545735U, + 537003843U, 1282953084U, 884458241U, 588930090U, 3930269801U, + 2961472450U, 1219535534U, 3632251943U, 268183903U, 1441240533U, + 3653903360U, 3854473319U, 2259087390U, 2548293048U, 2022641195U, + 2105543911U, 1764085217U, 3246183186U, 482438805U, 888317895U, + 2628314765U, 2466219854U, 717546004U, 2322237039U, 416725234U, + 1544049923U, 1797944973U, 3398652364U, 3111909456U, 485742908U, + 2277491072U, 1056355088U, 3181001278U, 129695079U, 2693624550U, + 1764438564U, 3797785470U, 195503713U, 3266519725U, 2053389444U, + 1961527818U, 3400226523U, 3777903038U, 2597274307U, 4235851091U, + 4094406648U, 2171410785U, 1781151386U, 1378577117U, 654643266U, + 3424024173U, 3385813322U, 679385799U, 479380913U, 681715441U, + 3096225905U, 276813409U, 3854398070U, 2721105350U, 831263315U, + 3276280337U, 2628301522U, 3984868494U, 1466099834U, 2104922114U, + 1412672743U, 820330404U, 3491501010U, 942735832U, 710652807U, + 3972652090U, 679881088U, 40577009U, 3705286397U, 2815423480U, + 3566262429U, 663396513U, 3777887429U, 4016670678U, 404539370U, + 1142712925U, 1140173408U, 2913248352U, 2872321286U, 263751841U, + 3175196073U, 3162557581U, 2878996619U, 75498548U, 3836833140U, + 3284664959U, 1157523805U, 112847376U, 207855609U, 1337979698U, + 1222578451U, 157107174U, 901174378U, 3883717063U, 1618632639U, + 1767889440U, 4264698824U, 1582999313U, 884471997U, 2508825098U, + 3756370771U, 2457213553U, 3565776881U, 3709583214U, 915609601U, + 460833524U, 1091049576U, 85522880U, 2553251U, 132102809U, + 2429882442U, 2562084610U, 1386507633U, 4112471229U, 21965213U, + 1981516006U, 2418435617U, 3054872091U, 4251511224U, 2025783543U, + 1916911512U, 2454491136U, 3938440891U, 3825869115U, 1121698605U, + 3463052265U, 802340101U, 1912886800U, 4031997367U, 3550640406U, + 1596096923U, 610150600U, 431464457U, 2541325046U, 486478003U, + 739704936U, 2862696430U, 3037903166U, 1129749694U, 2611481261U, + 1228993498U, 510075548U, 3424962587U, 2458689681U, 818934833U, + 4233309125U, 1608196251U, 3419476016U, 1858543939U, 2682166524U, + 3317854285U, 631986188U, 3008214764U, 613826412U, 3567358221U, + 3512343882U, 1552467474U, 3316162670U, 1275841024U, 4142173454U, + 565267881U, 768644821U, 198310105U, 2396688616U, 1837659011U, + 203429334U, 854539004U, 4235811518U, 3338304926U, 3730418692U, + 3852254981U, 3032046452U, 2329811860U, 2303590566U, 2696092212U, + 3894665932U, 145835667U, 249563655U, 1932210840U, 2431696407U, + 3312636759U, 214962629U, 2092026914U, 3020145527U, 4073039873U, + 2739105705U, 1308336752U, 855104522U, 2391715321U, 67448785U, + 547989482U, 854411802U, 3608633740U, 431731530U, 537375589U, + 3888005760U, 696099141U, 397343236U, 1864511780U, 44029739U, + 1729526891U, 1993398655U, 2010173426U, 2591546756U, 275223291U, + 1503900299U, 4217765081U, 2185635252U, 1122436015U, 3550155364U, + 681707194U, 3260479338U, 933579397U, 2983029282U, 2505504587U, + 2667410393U, 2962684490U, 4139721708U, 2658172284U, 2452602383U, + 2607631612U, 1344296217U, 3075398709U, 2949785295U, 1049956168U, + 3917185129U, 2155660174U, 3280524475U, 1503827867U, 674380765U, + 1918468193U, 3843983676U, 634358221U, 2538335643U, 1873351298U, + 3368723763U, 2129144130U, 3203528633U, 3087174986U, 2691698871U, + 2516284287U, 24437745U, 1118381474U, 2816314867U, 2448576035U, + 4281989654U, 217287825U, 165872888U, 2628995722U, 3533525116U, + 2721669106U, 872340568U, 3429930655U, 3309047304U, 3916704967U, + 3270160355U, 1348884255U, 1634797670U, 881214967U, 4259633554U, + 174613027U, 1103974314U, 1625224232U, 2678368291U, 1133866707U, + 3853082619U, 4073196549U, 1189620777U, 637238656U, 930241537U, + 4042750792U, 3842136042U, 2417007212U, 2524907510U, 1243036827U, + 1282059441U, 3764588774U, 1394459615U, 2323620015U, 1166152231U, + 3307479609U, 3849322257U, 3507445699U, 4247696636U, 758393720U, + 967665141U, 1095244571U, 1319812152U, 407678762U, 2640605208U, + 2170766134U, 3663594275U, 4039329364U, 2512175520U, 725523154U, + 2249807004U, 3312617979U, 2414634172U, 1278482215U, 349206484U, + 1573063308U, 1196429124U, 3873264116U, 2400067801U, 268795167U, + 226175489U, 2961367263U, 1968719665U, 42656370U, 1010790699U, + 561600615U, 2422453992U, 3082197735U, 1636700484U, 3977715296U, + 3125350482U, 3478021514U, 2227819446U, 1540868045U, 3061908980U, + 1087362407U, 3625200291U, 361937537U, 580441897U, 1520043666U, + 2270875402U, 1009161260U, 2502355842U, 4278769785U, 473902412U, + 1057239083U, 1905829039U, 1483781177U, 2080011417U, 1207494246U, + 1806991954U, 2194674403U, 3455972205U, 807207678U, 3655655687U, + 674112918U, 195425752U, 3917890095U, 1874364234U, 1837892715U, + 3663478166U, 1548892014U, 2570748714U, 2049929836U, 2167029704U, + 697543767U, 3499545023U, 3342496315U, 1725251190U, 3561387469U, + 2905606616U, 1580182447U, 3934525927U, 4103172792U, 1365672522U, + 1534795737U, 3308667416U, 2841911405U, 3943182730U, 4072020313U, + 3494770452U, 3332626671U, 55327267U, 478030603U, 411080625U, + 3419529010U, 1604767823U, 3513468014U, 570668510U, 913790824U, + 2283967995U, 695159462U, 3825542932U, 4150698144U, 1829758699U, + 202895590U, 1609122645U, 1267651008U, 2910315509U, 2511475445U, + 2477423819U, 3932081579U, 900879979U, 2145588390U, 2670007504U, + 580819444U, 1864996828U, 2526325979U, 1019124258U, 815508628U, + 2765933989U, 1277301341U, 3006021786U, 855540956U, 288025710U, + 1919594237U, 2331223864U, 177452412U, 2475870369U, 2689291749U, + 865194284U, 253432152U, 2628531804U, 2861208555U, 2361597573U, + 1653952120U, 1039661024U, 2159959078U, 3709040440U, 3564718533U, + 2596878672U, 2041442161U, 31164696U, 2662962485U, 3665637339U, + 1678115244U, 2699839832U, 3651968520U, 3521595541U, 458433303U, + 2423096824U, 21831741U, 380011703U, 2498168716U, 861806087U, + 1673574843U, 4188794405U, 2520563651U, 2632279153U, 2170465525U, + 4171949898U, 3886039621U, 1661344005U, 3424285243U, 992588372U, + 2500984144U, 2993248497U, 3590193895U, 1535327365U, 515645636U, + 131633450U, 3729760261U, 1613045101U, 3254194278U, 15889678U, + 1493590689U, 244148718U, 2991472662U, 1401629333U, 777349878U, + 2501401703U, 4285518317U, 3794656178U, 955526526U, 3442142820U, + 3970298374U, 736025417U, 2737370764U, 1271509744U, 440570731U, + 136141826U, 1596189518U, 923399175U, 257541519U, 3505774281U, + 2194358432U, 2518162991U, 1379893637U, 2667767062U, 3748146247U, + 1821712620U, 3923161384U, 1947811444U, 2392527197U, 4127419685U, + 1423694998U, 4156576871U, 1382885582U, 3420127279U, 3617499534U, + 2994377493U, 4038063986U, 1918458672U, 2983166794U, 4200449033U, + 353294540U, 1609232588U, 243926648U, 2332803291U, 507996832U, + 2392838793U, 4075145196U, 2060984340U, 4287475136U, 88232602U, + 2491531140U, 4159725633U, 2272075455U, 759298618U, 201384554U, + 838356250U, 1416268324U, 674476934U, 90795364U, 141672229U, + 3660399588U, 4196417251U, 3249270244U, 3774530247U, 59587265U, + 3683164208U, 19392575U, 1463123697U, 1882205379U, 293780489U, + 2553160622U, 2933904694U, 675638239U, 2851336944U, 1435238743U, + 2448730183U, 804436302U, 2119845972U, 322560608U, 4097732704U, + 2987802540U, 641492617U, 2575442710U, 4217822703U, 3271835300U, + 2836418300U, 3739921620U, 2138378768U, 2879771855U, 4294903423U, + 3121097946U, 2603440486U, 2560820391U, 1012930944U, 2313499967U, + 584489368U, 3431165766U, 897384869U, 2062537737U, 2847889234U, + 3742362450U, 2951174585U, 4204621084U, 1109373893U, 3668075775U, + 2750138839U, 3518055702U, 733072558U, 4169325400U, 788493625U +}; +static const uint64_t init_gen_rand_64_expected[] = { + KQU(16924766246869039260), KQU( 8201438687333352714), + KQU( 2265290287015001750), KQU(18397264611805473832), + KQU( 3375255223302384358), KQU( 6345559975416828796), + KQU(18229739242790328073), KQU( 7596792742098800905), + KQU( 255338647169685981), KQU( 2052747240048610300), + KQU(18328151576097299343), KQU(12472905421133796567), + KQU(11315245349717600863), KQU(16594110197775871209), + KQU(15708751964632456450), KQU(10452031272054632535), + KQU(11097646720811454386), KQU( 4556090668445745441), + KQU(17116187693090663106), KQU(14931526836144510645), + KQU( 9190752218020552591), KQU( 9625800285771901401), + KQU(13995141077659972832), KQU( 5194209094927829625), + KQU( 4156788379151063303), KQU( 8523452593770139494), + KQU(14082382103049296727), KQU( 2462601863986088483), + KQU( 3030583461592840678), KQU( 5221622077872827681), + KQU( 3084210671228981236), KQU(13956758381389953823), + KQU(13503889856213423831), KQU(15696904024189836170), + KQU( 4612584152877036206), KQU( 6231135538447867881), + KQU(10172457294158869468), KQU( 6452258628466708150), + KQU(14044432824917330221), KQU( 370168364480044279), + KQU(10102144686427193359), KQU( 667870489994776076), + KQU( 2732271956925885858), KQU(18027788905977284151), + KQU(15009842788582923859), KQU( 7136357960180199542), + KQU(15901736243475578127), KQU(16951293785352615701), + KQU(10551492125243691632), KQU(17668869969146434804), + KQU(13646002971174390445), KQU( 9804471050759613248), + KQU( 5511670439655935493), KQU(18103342091070400926), + KQU(17224512747665137533), KQU(15534627482992618168), + KQU( 1423813266186582647), KQU(15821176807932930024), + KQU( 30323369733607156), KQU(11599382494723479403), + KQU( 653856076586810062), KQU( 3176437395144899659), + KQU(14028076268147963917), KQU(16156398271809666195), + KQU( 3166955484848201676), KQU( 5746805620136919390), + KQU(17297845208891256593), KQU(11691653183226428483), + KQU(17900026146506981577), KQU(15387382115755971042), + KQU(16923567681040845943), KQU( 8039057517199388606), + KQU(11748409241468629263), KQU( 794358245539076095), + KQU(13438501964693401242), KQU(14036803236515618962), + KQU( 5252311215205424721), KQU(17806589612915509081), + KQU( 6802767092397596006), KQU(14212120431184557140), + KQU( 1072951366761385712), KQU(13098491780722836296), + KQU( 9466676828710797353), KQU(12673056849042830081), + KQU(12763726623645357580), KQU(16468961652999309493), + KQU(15305979875636438926), KQU(17444713151223449734), + KQU( 5692214267627883674), KQU(13049589139196151505), + KQU( 880115207831670745), KQU( 1776529075789695498), + KQU(16695225897801466485), KQU(10666901778795346845), + KQU( 6164389346722833869), KQU( 2863817793264300475), + KQU( 9464049921886304754), KQU( 3993566636740015468), + KQU( 9983749692528514136), KQU(16375286075057755211), + KQU(16042643417005440820), KQU(11445419662923489877), + KQU( 7999038846885158836), KQU( 6721913661721511535), + KQU( 5363052654139357320), KQU( 1817788761173584205), + KQU(13290974386445856444), KQU( 4650350818937984680), + KQU( 8219183528102484836), KQU( 1569862923500819899), + KQU( 4189359732136641860), KQU(14202822961683148583), + KQU( 4457498315309429058), KQU(13089067387019074834), + KQU(11075517153328927293), KQU(10277016248336668389), + KQU( 7070509725324401122), KQU(17808892017780289380), + KQU(13143367339909287349), KQU( 1377743745360085151), + KQU( 5749341807421286485), KQU(14832814616770931325), + KQU( 7688820635324359492), KQU(10960474011539770045), + KQU( 81970066653179790), KQU(12619476072607878022), + KQU( 4419566616271201744), KQU(15147917311750568503), + KQU( 5549739182852706345), KQU( 7308198397975204770), + KQU(13580425496671289278), KQU(17070764785210130301), + KQU( 8202832846285604405), KQU( 6873046287640887249), + KQU( 6927424434308206114), KQU( 6139014645937224874), + KQU(10290373645978487639), KQU(15904261291701523804), + KQU( 9628743442057826883), KQU(18383429096255546714), + KQU( 4977413265753686967), KQU( 7714317492425012869), + KQU( 9025232586309926193), KQU(14627338359776709107), + KQU(14759849896467790763), KQU(10931129435864423252), + KQU( 4588456988775014359), KQU(10699388531797056724), + KQU( 468652268869238792), KQU( 5755943035328078086), + KQU( 2102437379988580216), KQU( 9986312786506674028), + KQU( 2654207180040945604), KQU( 8726634790559960062), + KQU( 100497234871808137), KQU( 2800137176951425819), + KQU( 6076627612918553487), KQU( 5780186919186152796), + KQU( 8179183595769929098), KQU( 6009426283716221169), + KQU( 2796662551397449358), KQU( 1756961367041986764), + KQU( 6972897917355606205), KQU(14524774345368968243), + KQU( 2773529684745706940), KQU( 4853632376213075959), + KQU( 4198177923731358102), KQU( 8271224913084139776), + KQU( 2741753121611092226), KQU(16782366145996731181), + KQU(15426125238972640790), KQU(13595497100671260342), + KQU( 3173531022836259898), KQU( 6573264560319511662), + KQU(18041111951511157441), KQU( 2351433581833135952), + KQU( 3113255578908173487), KQU( 1739371330877858784), + KQU(16046126562789165480), KQU( 8072101652214192925), + KQU(15267091584090664910), KQU( 9309579200403648940), + KQU( 5218892439752408722), KQU(14492477246004337115), + KQU(17431037586679770619), KQU( 7385248135963250480), + KQU( 9580144956565560660), KQU( 4919546228040008720), + KQU(15261542469145035584), KQU(18233297270822253102), + KQU( 5453248417992302857), KQU( 9309519155931460285), + KQU(10342813012345291756), KQU(15676085186784762381), + KQU(15912092950691300645), KQU( 9371053121499003195), + KQU( 9897186478226866746), KQU(14061858287188196327), + KQU( 122575971620788119), KQU(12146750969116317754), + KQU( 4438317272813245201), KQU( 8332576791009527119), + KQU(13907785691786542057), KQU(10374194887283287467), + KQU( 2098798755649059566), KQU( 3416235197748288894), + KQU( 8688269957320773484), KQU( 7503964602397371571), + KQU(16724977015147478236), KQU( 9461512855439858184), + KQU(13259049744534534727), KQU( 3583094952542899294), + KQU( 8764245731305528292), KQU(13240823595462088985), + KQU(13716141617617910448), KQU(18114969519935960955), + KQU( 2297553615798302206), KQU( 4585521442944663362), + KQU(17776858680630198686), KQU( 4685873229192163363), + KQU( 152558080671135627), KQU(15424900540842670088), + KQU(13229630297130024108), KQU(17530268788245718717), + KQU(16675633913065714144), KQU( 3158912717897568068), + KQU(15399132185380087288), KQU( 7401418744515677872), + KQU(13135412922344398535), KQU( 6385314346100509511), + KQU(13962867001134161139), KQU(10272780155442671999), + KQU(12894856086597769142), KQU(13340877795287554994), + KQU(12913630602094607396), KQU(12543167911119793857), + KQU(17343570372251873096), KQU(10959487764494150545), + KQU( 6966737953093821128), KQU(13780699135496988601), + KQU( 4405070719380142046), KQU(14923788365607284982), + KQU( 2869487678905148380), KQU( 6416272754197188403), + KQU(15017380475943612591), KQU( 1995636220918429487), + KQU( 3402016804620122716), KQU(15800188663407057080), + KQU(11362369990390932882), KQU(15262183501637986147), + KQU(10239175385387371494), KQU( 9352042420365748334), + KQU( 1682457034285119875), KQU( 1724710651376289644), + KQU( 2038157098893817966), KQU( 9897825558324608773), + KQU( 1477666236519164736), KQU(16835397314511233640), + KQU(10370866327005346508), KQU(10157504370660621982), + KQU(12113904045335882069), KQU(13326444439742783008), + KQU(11302769043000765804), KQU(13594979923955228484), + KQU(11779351762613475968), KQU( 3786101619539298383), + KQU( 8021122969180846063), KQU(15745904401162500495), + KQU(10762168465993897267), KQU(13552058957896319026), + KQU(11200228655252462013), KQU( 5035370357337441226), + KQU( 7593918984545500013), KQU( 5418554918361528700), + KQU( 4858270799405446371), KQU( 9974659566876282544), + KQU(18227595922273957859), KQU( 2772778443635656220), + KQU(14285143053182085385), KQU( 9939700992429600469), + KQU(12756185904545598068), KQU( 2020783375367345262), + KQU( 57026775058331227), KQU( 950827867930065454), + KQU( 6602279670145371217), KQU( 2291171535443566929), + KQU( 5832380724425010313), KQU( 1220343904715982285), + KQU(17045542598598037633), KQU(15460481779702820971), + KQU(13948388779949365130), KQU(13975040175430829518), + KQU(17477538238425541763), KQU(11104663041851745725), + KQU(15860992957141157587), KQU(14529434633012950138), + KQU( 2504838019075394203), KQU( 7512113882611121886), + KQU( 4859973559980886617), KQU( 1258601555703250219), + KQU(15594548157514316394), KQU( 4516730171963773048), + KQU(11380103193905031983), KQU( 6809282239982353344), + KQU(18045256930420065002), KQU( 2453702683108791859), + KQU( 977214582986981460), KQU( 2006410402232713466), + KQU( 6192236267216378358), KQU( 3429468402195675253), + KQU(18146933153017348921), KQU(17369978576367231139), + KQU( 1246940717230386603), KQU(11335758870083327110), + KQU(14166488801730353682), KQU( 9008573127269635732), + KQU(10776025389820643815), KQU(15087605441903942962), + KQU( 1359542462712147922), KQU(13898874411226454206), + KQU(17911176066536804411), KQU( 9435590428600085274), + KQU( 294488509967864007), KQU( 8890111397567922046), + KQU( 7987823476034328778), KQU(13263827582440967651), + KQU( 7503774813106751573), KQU(14974747296185646837), + KQU( 8504765037032103375), KQU(17340303357444536213), + KQU( 7704610912964485743), KQU( 8107533670327205061), + KQU( 9062969835083315985), KQU(16968963142126734184), + KQU(12958041214190810180), KQU( 2720170147759570200), + KQU( 2986358963942189566), KQU(14884226322219356580), + KQU( 286224325144368520), KQU(11313800433154279797), + KQU(18366849528439673248), KQU(17899725929482368789), + KQU( 3730004284609106799), KQU( 1654474302052767205), + KQU( 5006698007047077032), KQU( 8196893913601182838), + KQU(15214541774425211640), KQU(17391346045606626073), + KQU( 8369003584076969089), KQU( 3939046733368550293), + KQU(10178639720308707785), KQU( 2180248669304388697), + KQU( 62894391300126322), KQU( 9205708961736223191), + KQU( 6837431058165360438), KQU( 3150743890848308214), + KQU(17849330658111464583), KQU(12214815643135450865), + KQU(13410713840519603402), KQU( 3200778126692046802), + KQU(13354780043041779313), KQU( 800850022756886036), + KQU(15660052933953067433), KQU( 6572823544154375676), + KQU(11030281857015819266), KQU(12682241941471433835), + KQU(11654136407300274693), KQU( 4517795492388641109), + KQU( 9757017371504524244), KQU(17833043400781889277), + KQU(12685085201747792227), KQU(10408057728835019573), + KQU( 98370418513455221), KQU( 6732663555696848598), + KQU(13248530959948529780), KQU( 3530441401230622826), + KQU(18188251992895660615), KQU( 1847918354186383756), + KQU( 1127392190402660921), KQU(11293734643143819463), + KQU( 3015506344578682982), KQU(13852645444071153329), + KQU( 2121359659091349142), KQU( 1294604376116677694), + KQU( 5616576231286352318), KQU( 7112502442954235625), + KQU(11676228199551561689), KQU(12925182803007305359), + KQU( 7852375518160493082), KQU( 1136513130539296154), + KQU( 5636923900916593195), KQU( 3221077517612607747), + KQU(17784790465798152513), KQU( 3554210049056995938), + KQU(17476839685878225874), KQU( 3206836372585575732), + KQU( 2765333945644823430), KQU(10080070903718799528), + KQU( 5412370818878286353), KQU( 9689685887726257728), + KQU( 8236117509123533998), KQU( 1951139137165040214), + KQU( 4492205209227980349), KQU(16541291230861602967), + KQU( 1424371548301437940), KQU( 9117562079669206794), + KQU(14374681563251691625), KQU(13873164030199921303), + KQU( 6680317946770936731), KQU(15586334026918276214), + KQU(10896213950976109802), KQU( 9506261949596413689), + KQU( 9903949574308040616), KQU( 6038397344557204470), + KQU( 174601465422373648), KQU(15946141191338238030), + KQU(17142225620992044937), KQU( 7552030283784477064), + KQU( 2947372384532947997), KQU( 510797021688197711), + KQU( 4962499439249363461), KQU( 23770320158385357), + KQU( 959774499105138124), KQU( 1468396011518788276), + KQU( 2015698006852312308), KQU( 4149400718489980136), + KQU( 5992916099522371188), KQU(10819182935265531076), + KQU(16189787999192351131), KQU( 342833961790261950), + KQU(12470830319550495336), KQU(18128495041912812501), + KQU( 1193600899723524337), KQU( 9056793666590079770), + KQU( 2154021227041669041), KQU( 4963570213951235735), + KQU( 4865075960209211409), KQU( 2097724599039942963), + KQU( 2024080278583179845), KQU(11527054549196576736), + KQU(10650256084182390252), KQU( 4808408648695766755), + KQU( 1642839215013788844), KQU(10607187948250398390), + KQU( 7076868166085913508), KQU( 730522571106887032), + KQU(12500579240208524895), KQU( 4484390097311355324), + KQU(15145801330700623870), KQU( 8055827661392944028), + KQU( 5865092976832712268), KQU(15159212508053625143), + KQU( 3560964582876483341), KQU( 4070052741344438280), + KQU( 6032585709886855634), KQU(15643262320904604873), + KQU( 2565119772293371111), KQU( 318314293065348260), + KQU(15047458749141511872), KQU( 7772788389811528730), + KQU( 7081187494343801976), KQU( 6465136009467253947), + KQU(10425940692543362069), KQU( 554608190318339115), + KQU(14796699860302125214), KQU( 1638153134431111443), + KQU(10336967447052276248), KQU( 8412308070396592958), + KQU( 4004557277152051226), KQU( 8143598997278774834), + KQU(16413323996508783221), KQU(13139418758033994949), + KQU( 9772709138335006667), KQU( 2818167159287157659), + KQU(17091740573832523669), KQU(14629199013130751608), + KQU(18268322711500338185), KQU( 8290963415675493063), + KQU( 8830864907452542588), KQU( 1614839084637494849), + KQU(14855358500870422231), KQU( 3472996748392519937), + KQU(15317151166268877716), KQU( 5825895018698400362), + KQU(16730208429367544129), KQU(10481156578141202800), + KQU( 4746166512382823750), KQU(12720876014472464998), + KQU( 8825177124486735972), KQU(13733447296837467838), + KQU( 6412293741681359625), KQU( 8313213138756135033), + KQU(11421481194803712517), KQU( 7997007691544174032), + KQU( 6812963847917605930), KQU( 9683091901227558641), + KQU(14703594165860324713), KQU( 1775476144519618309), + KQU( 2724283288516469519), KQU( 717642555185856868), + KQU( 8736402192215092346), KQU(11878800336431381021), + KQU( 4348816066017061293), KQU( 6115112756583631307), + KQU( 9176597239667142976), KQU(12615622714894259204), + KQU(10283406711301385987), KQU( 5111762509485379420), + KQU( 3118290051198688449), KQU( 7345123071632232145), + KQU( 9176423451688682359), KQU( 4843865456157868971), + KQU(12008036363752566088), KQU(12058837181919397720), + KQU( 2145073958457347366), KQU( 1526504881672818067), + KQU( 3488830105567134848), KQU(13208362960674805143), + KQU( 4077549672899572192), KQU( 7770995684693818365), + KQU( 1398532341546313593), KQU(12711859908703927840), + KQU( 1417561172594446813), KQU(17045191024194170604), + KQU( 4101933177604931713), KQU(14708428834203480320), + KQU(17447509264469407724), KQU(14314821973983434255), + KQU(17990472271061617265), KQU( 5087756685841673942), + KQU(12797820586893859939), KQU( 1778128952671092879), + KQU( 3535918530508665898), KQU( 9035729701042481301), + KQU(14808661568277079962), KQU(14587345077537747914), + KQU(11920080002323122708), KQU( 6426515805197278753), + KQU( 3295612216725984831), KQU(11040722532100876120), + KQU(12305952936387598754), KQU(16097391899742004253), + KQU( 4908537335606182208), KQU(12446674552196795504), + KQU(16010497855816895177), KQU( 9194378874788615551), + KQU( 3382957529567613384), KQU( 5154647600754974077), + KQU( 9801822865328396141), KQU( 9023662173919288143), + KQU(17623115353825147868), KQU( 8238115767443015816), + KQU(15811444159859002560), KQU( 9085612528904059661), + KQU( 6888601089398614254), KQU( 258252992894160189), + KQU( 6704363880792428622), KQU( 6114966032147235763), + KQU(11075393882690261875), KQU( 8797664238933620407), + KQU( 5901892006476726920), KQU( 5309780159285518958), + KQU(14940808387240817367), KQU(14642032021449656698), + KQU( 9808256672068504139), KQU( 3670135111380607658), + KQU(11211211097845960152), KQU( 1474304506716695808), + KQU(15843166204506876239), KQU( 7661051252471780561), + KQU(10170905502249418476), KQU( 7801416045582028589), + KQU( 2763981484737053050), KQU( 9491377905499253054), + KQU(16201395896336915095), KQU( 9256513756442782198), + KQU( 5411283157972456034), KQU( 5059433122288321676), + KQU( 4327408006721123357), KQU( 9278544078834433377), + KQU( 7601527110882281612), KQU(11848295896975505251), + KQU(12096998801094735560), KQU(14773480339823506413), + KQU(15586227433895802149), KQU(12786541257830242872), + KQU( 6904692985140503067), KQU( 5309011515263103959), + KQU(12105257191179371066), KQU(14654380212442225037), + KQU( 2556774974190695009), KQU( 4461297399927600261), + KQU(14888225660915118646), KQU(14915459341148291824), + KQU( 2738802166252327631), KQU( 6047155789239131512), + KQU(12920545353217010338), KQU(10697617257007840205), + KQU( 2751585253158203504), KQU(13252729159780047496), + KQU(14700326134672815469), KQU(14082527904374600529), + KQU(16852962273496542070), KQU(17446675504235853907), + KQU(15019600398527572311), KQU(12312781346344081551), + KQU(14524667935039810450), KQU( 5634005663377195738), + KQU(11375574739525000569), KQU( 2423665396433260040), + KQU( 5222836914796015410), KQU( 4397666386492647387), + KQU( 4619294441691707638), KQU( 665088602354770716), + KQU(13246495665281593610), KQU( 6564144270549729409), + KQU(10223216188145661688), KQU( 3961556907299230585), + KQU(11543262515492439914), KQU(16118031437285993790), + KQU( 7143417964520166465), KQU(13295053515909486772), + KQU( 40434666004899675), KQU(17127804194038347164), + KQU( 8599165966560586269), KQU( 8214016749011284903), + KQU(13725130352140465239), KQU( 5467254474431726291), + KQU( 7748584297438219877), KQU(16933551114829772472), + KQU( 2169618439506799400), KQU( 2169787627665113463), + KQU(17314493571267943764), KQU(18053575102911354912), + KQU(11928303275378476973), KQU(11593850925061715550), + KQU(17782269923473589362), KQU( 3280235307704747039), + KQU( 6145343578598685149), KQU(17080117031114086090), + KQU(18066839902983594755), KQU( 6517508430331020706), + KQU( 8092908893950411541), KQU(12558378233386153732), + KQU( 4476532167973132976), KQU(16081642430367025016), + KQU( 4233154094369139361), KQU( 8693630486693161027), + KQU(11244959343027742285), KQU(12273503967768513508), + KQU(14108978636385284876), KQU( 7242414665378826984), + KQU( 6561316938846562432), KQU( 8601038474994665795), + KQU(17532942353612365904), KQU(17940076637020912186), + KQU( 7340260368823171304), KQU( 7061807613916067905), + KQU(10561734935039519326), KQU(17990796503724650862), + KQU( 6208732943911827159), KQU( 359077562804090617), + KQU(14177751537784403113), KQU(10659599444915362902), + KQU(15081727220615085833), KQU(13417573895659757486), + KQU(15513842342017811524), KQU(11814141516204288231), + KQU( 1827312513875101814), KQU( 2804611699894603103), + KQU(17116500469975602763), KQU(12270191815211952087), + KQU(12256358467786024988), KQU(18435021722453971267), + KQU( 671330264390865618), KQU( 476504300460286050), + KQU(16465470901027093441), KQU( 4047724406247136402), + KQU( 1322305451411883346), KQU( 1388308688834322280), + KQU( 7303989085269758176), KQU( 9323792664765233642), + KQU( 4542762575316368936), KQU(17342696132794337618), + KQU( 4588025054768498379), KQU(13415475057390330804), + KQU(17880279491733405570), KQU(10610553400618620353), + KQU( 3180842072658960139), KQU(13002966655454270120), + KQU( 1665301181064982826), KQU( 7083673946791258979), + KQU( 190522247122496820), KQU(17388280237250677740), + KQU( 8430770379923642945), KQU(12987180971921668584), + KQU( 2311086108365390642), KQU( 2870984383579822345), + KQU(14014682609164653318), KQU(14467187293062251484), + KQU( 192186361147413298), KQU(15171951713531796524), + KQU( 9900305495015948728), KQU(17958004775615466344), + KQU(14346380954498606514), KQU(18040047357617407096), + KQU( 5035237584833424532), KQU(15089555460613972287), + KQU( 4131411873749729831), KQU( 1329013581168250330), + KQU(10095353333051193949), KQU(10749518561022462716), + KQU( 9050611429810755847), KQU(15022028840236655649), + KQU( 8775554279239748298), KQU(13105754025489230502), + KQU(15471300118574167585), KQU( 89864764002355628), + KQU( 8776416323420466637), KQU( 5280258630612040891), + KQU( 2719174488591862912), KQU( 7599309137399661994), + KQU(15012887256778039979), KQU(14062981725630928925), + KQU(12038536286991689603), KQU( 7089756544681775245), + KQU(10376661532744718039), KQU( 1265198725901533130), + KQU(13807996727081142408), KQU( 2935019626765036403), + KQU( 7651672460680700141), KQU( 3644093016200370795), + KQU( 2840982578090080674), KQU(17956262740157449201), + KQU(18267979450492880548), KQU(11799503659796848070), + KQU( 9942537025669672388), KQU(11886606816406990297), + KQU( 5488594946437447576), KQU( 7226714353282744302), + KQU( 3784851653123877043), KQU( 878018453244803041), + KQU(12110022586268616085), KQU( 734072179404675123), + KQU(11869573627998248542), KQU( 469150421297783998), + KQU( 260151124912803804), KQU(11639179410120968649), + KQU( 9318165193840846253), KQU(12795671722734758075), + KQU(15318410297267253933), KQU( 691524703570062620), + KQU( 5837129010576994601), KQU(15045963859726941052), + KQU( 5850056944932238169), KQU(12017434144750943807), + KQU( 7447139064928956574), KQU( 3101711812658245019), + KQU(16052940704474982954), KQU(18195745945986994042), + KQU( 8932252132785575659), KQU(13390817488106794834), + KQU(11582771836502517453), KQU( 4964411326683611686), + KQU( 2195093981702694011), KQU(14145229538389675669), + KQU(16459605532062271798), KQU( 866316924816482864), + KQU( 4593041209937286377), KQU( 8415491391910972138), + KQU( 4171236715600528969), KQU(16637569303336782889), + KQU( 2002011073439212680), KQU(17695124661097601411), + KQU( 4627687053598611702), KQU( 7895831936020190403), + KQU( 8455951300917267802), KQU( 2923861649108534854), + KQU( 8344557563927786255), KQU( 6408671940373352556), + KQU(12210227354536675772), KQU(14294804157294222295), + KQU(10103022425071085127), KQU(10092959489504123771), + KQU( 6554774405376736268), KQU(12629917718410641774), + KQU( 6260933257596067126), KQU( 2460827021439369673), + KQU( 2541962996717103668), KQU( 597377203127351475), + KQU( 5316984203117315309), KQU( 4811211393563241961), + KQU(13119698597255811641), KQU( 8048691512862388981), + KQU(10216818971194073842), KQU( 4612229970165291764), + KQU(10000980798419974770), KQU( 6877640812402540687), + KQU( 1488727563290436992), KQU( 2227774069895697318), + KQU(11237754507523316593), KQU(13478948605382290972), + KQU( 1963583846976858124), KQU( 5512309205269276457), + KQU( 3972770164717652347), KQU( 3841751276198975037), + KQU(10283343042181903117), KQU( 8564001259792872199), + KQU(16472187244722489221), KQU( 8953493499268945921), + KQU( 3518747340357279580), KQU( 4003157546223963073), + KQU( 3270305958289814590), KQU( 3966704458129482496), + KQU( 8122141865926661939), KQU(14627734748099506653), + KQU(13064426990862560568), KQU( 2414079187889870829), + KQU( 5378461209354225306), KQU(10841985740128255566), + KQU( 538582442885401738), KQU( 7535089183482905946), + KQU(16117559957598879095), KQU( 8477890721414539741), + KQU( 1459127491209533386), KQU(17035126360733620462), + KQU( 8517668552872379126), KQU(10292151468337355014), + KQU(17081267732745344157), KQU(13751455337946087178), + KQU(14026945459523832966), KQU( 6653278775061723516), + KQU(10619085543856390441), KQU( 2196343631481122885), + KQU(10045966074702826136), KQU(10082317330452718282), + KQU( 5920859259504831242), KQU( 9951879073426540617), + KQU( 7074696649151414158), KQU(15808193543879464318), + KQU( 7385247772746953374), KQU( 3192003544283864292), + KQU(18153684490917593847), KQU(12423498260668568905), + KQU(10957758099756378169), KQU(11488762179911016040), + KQU( 2099931186465333782), KQU(11180979581250294432), + KQU( 8098916250668367933), KQU( 3529200436790763465), + KQU(12988418908674681745), KQU( 6147567275954808580), + KQU( 3207503344604030989), KQU(10761592604898615360), + KQU( 229854861031893504), KQU( 8809853962667144291), + KQU(13957364469005693860), KQU( 7634287665224495886), + KQU(12353487366976556874), KQU( 1134423796317152034), + KQU( 2088992471334107068), KQU( 7393372127190799698), + KQU( 1845367839871058391), KQU( 207922563987322884), + KQU(11960870813159944976), KQU(12182120053317317363), + KQU(17307358132571709283), KQU(13871081155552824936), + KQU(18304446751741566262), KQU( 7178705220184302849), + KQU(10929605677758824425), KQU(16446976977835806844), + KQU(13723874412159769044), KQU( 6942854352100915216), + KQU( 1726308474365729390), KQU( 2150078766445323155), + KQU(15345558947919656626), KQU(12145453828874527201), + KQU( 2054448620739726849), KQU( 2740102003352628137), + KQU(11294462163577610655), KQU( 756164283387413743), + KQU(17841144758438810880), KQU(10802406021185415861), + KQU( 8716455530476737846), KQU( 6321788834517649606), + KQU(14681322910577468426), KQU(17330043563884336387), + KQU(12701802180050071614), KQU(14695105111079727151), + KQU( 5112098511654172830), KQU( 4957505496794139973), + KQU( 8270979451952045982), KQU(12307685939199120969), + KQU(12425799408953443032), KQU( 8376410143634796588), + KQU(16621778679680060464), KQU( 3580497854566660073), + KQU( 1122515747803382416), KQU( 857664980960597599), + KQU( 6343640119895925918), KQU(12878473260854462891), + KQU(10036813920765722626), KQU(14451335468363173812), + KQU( 5476809692401102807), KQU(16442255173514366342), + KQU(13060203194757167104), KQU(14354124071243177715), + KQU(15961249405696125227), KQU(13703893649690872584), + KQU( 363907326340340064), KQU( 6247455540491754842), + KQU(12242249332757832361), KQU( 156065475679796717), + KQU( 9351116235749732355), KQU( 4590350628677701405), + KQU( 1671195940982350389), KQU(13501398458898451905), + KQU( 6526341991225002255), KQU( 1689782913778157592), + KQU( 7439222350869010334), KQU(13975150263226478308), + KQU(11411961169932682710), KQU(17204271834833847277), + KQU( 541534742544435367), KQU( 6591191931218949684), + KQU( 2645454775478232486), KQU( 4322857481256485321), + KQU( 8477416487553065110), KQU(12902505428548435048), + KQU( 971445777981341415), KQU(14995104682744976712), + KQU( 4243341648807158063), KQU( 8695061252721927661), + KQU( 5028202003270177222), KQU( 2289257340915567840), + KQU(13870416345121866007), KQU(13994481698072092233), + KQU( 6912785400753196481), KQU( 2278309315841980139), + KQU( 4329765449648304839), KQU( 5963108095785485298), + KQU( 4880024847478722478), KQU(16015608779890240947), + KQU( 1866679034261393544), KQU( 914821179919731519), + KQU( 9643404035648760131), KQU( 2418114953615593915), + KQU( 944756836073702374), KQU(15186388048737296834), + KQU( 7723355336128442206), KQU( 7500747479679599691), + KQU(18013961306453293634), KQU( 2315274808095756456), + KQU(13655308255424029566), KQU(17203800273561677098), + KQU( 1382158694422087756), KQU( 5090390250309588976), + KQU( 517170818384213989), KQU( 1612709252627729621), + KQU( 1330118955572449606), KQU( 300922478056709885), + KQU(18115693291289091987), KQU(13491407109725238321), + KQU(15293714633593827320), KQU( 5151539373053314504), + KQU( 5951523243743139207), KQU(14459112015249527975), + KQU( 5456113959000700739), KQU( 3877918438464873016), + KQU(12534071654260163555), KQU(15871678376893555041), + KQU(11005484805712025549), KQU(16353066973143374252), + KQU( 4358331472063256685), KQU( 8268349332210859288), + KQU(12485161590939658075), KQU(13955993592854471343), + KQU( 5911446886848367039), KQU(14925834086813706974), + KQU( 6590362597857994805), KQU( 1280544923533661875), + KQU( 1637756018947988164), KQU( 4734090064512686329), + KQU(16693705263131485912), KQU( 6834882340494360958), + KQU( 8120732176159658505), KQU( 2244371958905329346), + KQU(10447499707729734021), KQU( 7318742361446942194), + KQU( 8032857516355555296), KQU(14023605983059313116), + KQU( 1032336061815461376), KQU( 9840995337876562612), + KQU( 9869256223029203587), KQU(12227975697177267636), + KQU(12728115115844186033), KQU( 7752058479783205470), + KQU( 729733219713393087), KQU(12954017801239007622) +}; +static const uint64_t init_by_array_64_expected[] = { + KQU( 2100341266307895239), KQU( 8344256300489757943), + KQU(15687933285484243894), KQU( 8268620370277076319), + KQU(12371852309826545459), KQU( 8800491541730110238), + KQU(18113268950100835773), KQU( 2886823658884438119), + KQU( 3293667307248180724), KQU( 9307928143300172731), + KQU( 7688082017574293629), KQU( 900986224735166665), + KQU( 9977972710722265039), KQU( 6008205004994830552), + KQU( 546909104521689292), KQU( 7428471521869107594), + KQU(14777563419314721179), KQU(16116143076567350053), + KQU( 5322685342003142329), KQU( 4200427048445863473), + KQU( 4693092150132559146), KQU(13671425863759338582), + KQU( 6747117460737639916), KQU( 4732666080236551150), + KQU( 5912839950611941263), KQU( 3903717554504704909), + KQU( 2615667650256786818), KQU(10844129913887006352), + KQU(13786467861810997820), KQU(14267853002994021570), + KQU(13767807302847237439), KQU(16407963253707224617), + KQU( 4802498363698583497), KQU( 2523802839317209764), + KQU( 3822579397797475589), KQU( 8950320572212130610), + KQU( 3745623504978342534), KQU(16092609066068482806), + KQU( 9817016950274642398), KQU(10591660660323829098), + KQU(11751606650792815920), KQU( 5122873818577122211), + KQU(17209553764913936624), KQU( 6249057709284380343), + KQU(15088791264695071830), KQU(15344673071709851930), + KQU( 4345751415293646084), KQU( 2542865750703067928), + KQU(13520525127852368784), KQU(18294188662880997241), + KQU( 3871781938044881523), KQU( 2873487268122812184), + KQU(15099676759482679005), KQU(15442599127239350490), + KQU( 6311893274367710888), KQU( 3286118760484672933), + KQU( 4146067961333542189), KQU(13303942567897208770), + KQU( 8196013722255630418), KQU( 4437815439340979989), + KQU(15433791533450605135), KQU( 4254828956815687049), + KQU( 1310903207708286015), KQU(10529182764462398549), + KQU(14900231311660638810), KQU( 9727017277104609793), + KQU( 1821308310948199033), KQU(11628861435066772084), + KQU( 9469019138491546924), KQU( 3145812670532604988), + KQU( 9938468915045491919), KQU( 1562447430672662142), + KQU(13963995266697989134), KQU( 3356884357625028695), + KQU( 4499850304584309747), KQU( 8456825817023658122), + KQU(10859039922814285279), KQU( 8099512337972526555), + KQU( 348006375109672149), KQU(11919893998241688603), + KQU( 1104199577402948826), KQU(16689191854356060289), + KQU(10992552041730168078), KQU( 7243733172705465836), + KQU( 5668075606180319560), KQU(18182847037333286970), + KQU( 4290215357664631322), KQU( 4061414220791828613), + KQU(13006291061652989604), KQU( 7140491178917128798), + KQU(12703446217663283481), KQU( 5500220597564558267), + KQU(10330551509971296358), KQU(15958554768648714492), + KQU( 5174555954515360045), KQU( 1731318837687577735), + KQU( 3557700801048354857), KQU(13764012341928616198), + KQU(13115166194379119043), KQU( 7989321021560255519), + KQU( 2103584280905877040), KQU( 9230788662155228488), + KQU(16396629323325547654), KQU( 657926409811318051), + KQU(15046700264391400727), KQU( 5120132858771880830), + KQU( 7934160097989028561), KQU( 6963121488531976245), + KQU(17412329602621742089), KQU(15144843053931774092), + KQU(17204176651763054532), KQU(13166595387554065870), + KQU( 8590377810513960213), KQU( 5834365135373991938), + KQU( 7640913007182226243), KQU( 3479394703859418425), + KQU(16402784452644521040), KQU( 4993979809687083980), + KQU(13254522168097688865), KQU(15643659095244365219), + KQU( 5881437660538424982), KQU(11174892200618987379), + KQU( 254409966159711077), KQU(17158413043140549909), + KQU( 3638048789290376272), KQU( 1376816930299489190), + KQU( 4622462095217761923), KQU(15086407973010263515), + KQU(13253971772784692238), KQU( 5270549043541649236), + KQU(11182714186805411604), KQU(12283846437495577140), + KQU( 5297647149908953219), KQU(10047451738316836654), + KQU( 4938228100367874746), KQU(12328523025304077923), + KQU( 3601049438595312361), KQU( 9313624118352733770), + KQU(13322966086117661798), KQU(16660005705644029394), + KQU(11337677526988872373), KQU(13869299102574417795), + KQU(15642043183045645437), KQU( 3021755569085880019), + KQU( 4979741767761188161), KQU(13679979092079279587), + KQU( 3344685842861071743), KQU(13947960059899588104), + KQU( 305806934293368007), KQU( 5749173929201650029), + KQU(11123724852118844098), KQU(15128987688788879802), + KQU(15251651211024665009), KQU( 7689925933816577776), + KQU(16732804392695859449), KQU(17087345401014078468), + KQU(14315108589159048871), KQU( 4820700266619778917), + KQU(16709637539357958441), KQU( 4936227875177351374), + KQU( 2137907697912987247), KQU(11628565601408395420), + KQU( 2333250549241556786), KQU( 5711200379577778637), + KQU( 5170680131529031729), KQU(12620392043061335164), + KQU( 95363390101096078), KQU( 5487981914081709462), + KQU( 1763109823981838620), KQU( 3395861271473224396), + KQU( 1300496844282213595), KQU( 6894316212820232902), + KQU(10673859651135576674), KQU( 5911839658857903252), + KQU(17407110743387299102), KQU( 8257427154623140385), + KQU(11389003026741800267), KQU( 4070043211095013717), + KQU(11663806997145259025), KQU(15265598950648798210), + KQU( 630585789434030934), KQU( 3524446529213587334), + KQU( 7186424168495184211), KQU(10806585451386379021), + KQU(11120017753500499273), KQU( 1586837651387701301), + KQU(17530454400954415544), KQU( 9991670045077880430), + KQU( 7550997268990730180), KQU( 8640249196597379304), + KQU( 3522203892786893823), KQU(10401116549878854788), + KQU(13690285544733124852), KQU( 8295785675455774586), + KQU(15535716172155117603), KQU( 3112108583723722511), + KQU(17633179955339271113), KQU(18154208056063759375), + KQU( 1866409236285815666), KQU(13326075895396412882), + KQU( 8756261842948020025), KQU( 6281852999868439131), + KQU(15087653361275292858), KQU(10333923911152949397), + KQU( 5265567645757408500), KQU(12728041843210352184), + KQU( 6347959327507828759), KQU( 154112802625564758), + KQU(18235228308679780218), KQU( 3253805274673352418), + KQU( 4849171610689031197), KQU(17948529398340432518), + KQU(13803510475637409167), KQU(13506570190409883095), + KQU(15870801273282960805), KQU( 8451286481299170773), + KQU( 9562190620034457541), KQU( 8518905387449138364), + KQU(12681306401363385655), KQU( 3788073690559762558), + KQU( 5256820289573487769), KQU( 2752021372314875467), + KQU( 6354035166862520716), KQU( 4328956378309739069), + KQU( 449087441228269600), KQU( 5533508742653090868), + KQU( 1260389420404746988), KQU(18175394473289055097), + KQU( 1535467109660399420), KQU( 8818894282874061442), + KQU(12140873243824811213), KQU(15031386653823014946), + KQU( 1286028221456149232), KQU( 6329608889367858784), + KQU( 9419654354945132725), KQU( 6094576547061672379), + KQU(17706217251847450255), KQU( 1733495073065878126), + KQU(16918923754607552663), KQU( 8881949849954945044), + KQU(12938977706896313891), KQU(14043628638299793407), + KQU(18393874581723718233), KQU( 6886318534846892044), + KQU(14577870878038334081), KQU(13541558383439414119), + KQU(13570472158807588273), KQU(18300760537910283361), + KQU( 818368572800609205), KQU( 1417000585112573219), + KQU(12337533143867683655), KQU(12433180994702314480), + KQU( 778190005829189083), KQU(13667356216206524711), + KQU( 9866149895295225230), KQU(11043240490417111999), + KQU( 1123933826541378598), KQU( 6469631933605123610), + KQU(14508554074431980040), KQU(13918931242962026714), + KQU( 2870785929342348285), KQU(14786362626740736974), + KQU(13176680060902695786), KQU( 9591778613541679456), + KQU( 9097662885117436706), KQU( 749262234240924947), + KQU( 1944844067793307093), KQU( 4339214904577487742), + KQU( 8009584152961946551), KQU(16073159501225501777), + KQU( 3335870590499306217), KQU(17088312653151202847), + KQU( 3108893142681931848), KQU(16636841767202792021), + KQU(10423316431118400637), KQU( 8008357368674443506), + KQU(11340015231914677875), KQU(17687896501594936090), + KQU(15173627921763199958), KQU( 542569482243721959), + KQU(15071714982769812975), KQU( 4466624872151386956), + KQU( 1901780715602332461), KQU( 9822227742154351098), + KQU( 1479332892928648780), KQU( 6981611948382474400), + KQU( 7620824924456077376), KQU(14095973329429406782), + KQU( 7902744005696185404), KQU(15830577219375036920), + KQU(10287076667317764416), KQU(12334872764071724025), + KQU( 4419302088133544331), KQU(14455842851266090520), + KQU(12488077416504654222), KQU( 7953892017701886766), + KQU( 6331484925529519007), KQU( 4902145853785030022), + KQU(17010159216096443073), KQU(11945354668653886087), + KQU(15112022728645230829), KQU(17363484484522986742), + KQU( 4423497825896692887), KQU( 8155489510809067471), + KQU( 258966605622576285), KQU( 5462958075742020534), + KQU( 6763710214913276228), KQU( 2368935183451109054), + KQU(14209506165246453811), KQU( 2646257040978514881), + KQU( 3776001911922207672), KQU( 1419304601390147631), + KQU(14987366598022458284), KQU( 3977770701065815721), + KQU( 730820417451838898), KQU( 3982991703612885327), + KQU( 2803544519671388477), KQU(17067667221114424649), + KQU( 2922555119737867166), KQU( 1989477584121460932), + KQU(15020387605892337354), KQU( 9293277796427533547), + KQU(10722181424063557247), KQU(16704542332047511651), + KQU( 5008286236142089514), KQU(16174732308747382540), + KQU(17597019485798338402), KQU(13081745199110622093), + KQU( 8850305883842258115), KQU(12723629125624589005), + KQU( 8140566453402805978), KQU(15356684607680935061), + KQU(14222190387342648650), KQU(11134610460665975178), + KQU( 1259799058620984266), KQU(13281656268025610041), + KQU( 298262561068153992), KQU(12277871700239212922), + KQU(13911297774719779438), KQU(16556727962761474934), + KQU(17903010316654728010), KQU( 9682617699648434744), + KQU(14757681836838592850), KQU( 1327242446558524473), + KQU(11126645098780572792), KQU( 1883602329313221774), + KQU( 2543897783922776873), KQU(15029168513767772842), + KQU(12710270651039129878), KQU(16118202956069604504), + KQU(15010759372168680524), KQU( 2296827082251923948), + KQU(10793729742623518101), KQU(13829764151845413046), + KQU(17769301223184451213), KQU( 3118268169210783372), + KQU(17626204544105123127), KQU( 7416718488974352644), + KQU(10450751996212925994), KQU( 9352529519128770586), + KQU( 259347569641110140), KQU( 8048588892269692697), + KQU( 1774414152306494058), KQU(10669548347214355622), + KQU(13061992253816795081), KQU(18432677803063861659), + KQU( 8879191055593984333), KQU(12433753195199268041), + KQU(14919392415439730602), KQU( 6612848378595332963), + KQU( 6320986812036143628), KQU(10465592420226092859), + KQU( 4196009278962570808), KQU( 3747816564473572224), + KQU(17941203486133732898), KQU( 2350310037040505198), + KQU( 5811779859134370113), KQU(10492109599506195126), + KQU( 7699650690179541274), KQU( 1954338494306022961), + KQU(14095816969027231152), KQU( 5841346919964852061), + KQU(14945969510148214735), KQU( 3680200305887550992), + KQU( 6218047466131695792), KQU( 8242165745175775096), + KQU(11021371934053307357), KQU( 1265099502753169797), + KQU( 4644347436111321718), KQU( 3609296916782832859), + KQU( 8109807992218521571), KQU(18387884215648662020), + KQU(14656324896296392902), KQU(17386819091238216751), + KQU(17788300878582317152), KQU( 7919446259742399591), + KQU( 4466613134576358004), KQU(12928181023667938509), + KQU(13147446154454932030), KQU(16552129038252734620), + KQU( 8395299403738822450), KQU(11313817655275361164), + KQU( 434258809499511718), KQU( 2074882104954788676), + KQU( 7929892178759395518), KQU( 9006461629105745388), + KQU( 5176475650000323086), KQU(11128357033468341069), + KQU(12026158851559118955), KQU(14699716249471156500), + KQU( 448982497120206757), KQU( 4156475356685519900), + KQU( 6063816103417215727), KQU(10073289387954971479), + KQU( 8174466846138590962), KQU( 2675777452363449006), + KQU( 9090685420572474281), KQU( 6659652652765562060), + KQU(12923120304018106621), KQU(11117480560334526775), + KQU( 937910473424587511), KQU( 1838692113502346645), + KQU(11133914074648726180), KQU( 7922600945143884053), + KQU(13435287702700959550), KQU( 5287964921251123332), + KQU(11354875374575318947), KQU(17955724760748238133), + KQU(13728617396297106512), KQU( 4107449660118101255), + KQU( 1210269794886589623), KQU(11408687205733456282), + KQU( 4538354710392677887), KQU(13566803319341319267), + KQU(17870798107734050771), KQU( 3354318982568089135), + KQU( 9034450839405133651), KQU(13087431795753424314), + KQU( 950333102820688239), KQU( 1968360654535604116), + KQU(16840551645563314995), KQU( 8867501803892924995), + KQU(11395388644490626845), KQU( 1529815836300732204), + KQU(13330848522996608842), KQU( 1813432878817504265), + KQU( 2336867432693429560), KQU(15192805445973385902), + KQU( 2528593071076407877), KQU( 128459777936689248), + KQU( 9976345382867214866), KQU( 6208885766767996043), + KQU(14982349522273141706), KQU( 3099654362410737822), + KQU(13776700761947297661), KQU( 8806185470684925550), + KQU( 8151717890410585321), KQU( 640860591588072925), + KQU(14592096303937307465), KQU( 9056472419613564846), + KQU(14861544647742266352), KQU(12703771500398470216), + KQU( 3142372800384138465), KQU( 6201105606917248196), + KQU(18337516409359270184), KQU(15042268695665115339), + KQU(15188246541383283846), KQU(12800028693090114519), + KQU( 5992859621101493472), KQU(18278043971816803521), + KQU( 9002773075219424560), KQU( 7325707116943598353), + KQU( 7930571931248040822), KQU( 5645275869617023448), + KQU( 7266107455295958487), KQU( 4363664528273524411), + KQU(14313875763787479809), KQU(17059695613553486802), + KQU( 9247761425889940932), KQU(13704726459237593128), + KQU( 2701312427328909832), KQU(17235532008287243115), + KQU(14093147761491729538), KQU( 6247352273768386516), + KQU( 8268710048153268415), KQU( 7985295214477182083), + KQU(15624495190888896807), KQU( 3772753430045262788), + KQU( 9133991620474991698), KQU( 5665791943316256028), + KQU( 7551996832462193473), KQU(13163729206798953877), + KQU( 9263532074153846374), KQU( 1015460703698618353), + KQU(17929874696989519390), KQU(18257884721466153847), + KQU(16271867543011222991), KQU( 3905971519021791941), + KQU(16814488397137052085), KQU( 1321197685504621613), + KQU( 2870359191894002181), KQU(14317282970323395450), + KQU(13663920845511074366), KQU( 2052463995796539594), + KQU(14126345686431444337), KQU( 1727572121947022534), + KQU(17793552254485594241), KQU( 6738857418849205750), + KQU( 1282987123157442952), KQU(16655480021581159251), + KQU( 6784587032080183866), KQU(14726758805359965162), + KQU( 7577995933961987349), KQU(12539609320311114036), + KQU(10789773033385439494), KQU( 8517001497411158227), + KQU(10075543932136339710), KQU(14838152340938811081), + KQU( 9560840631794044194), KQU(17445736541454117475), + KQU(10633026464336393186), KQU(15705729708242246293), + KQU( 1117517596891411098), KQU( 4305657943415886942), + KQU( 4948856840533979263), KQU(16071681989041789593), + KQU(13723031429272486527), KQU( 7639567622306509462), + KQU(12670424537483090390), KQU( 9715223453097197134), + KQU( 5457173389992686394), KQU( 289857129276135145), + KQU(17048610270521972512), KQU( 692768013309835485), + KQU(14823232360546632057), KQU(18218002361317895936), + KQU( 3281724260212650204), KQU(16453957266549513795), + KQU( 8592711109774511881), KQU( 929825123473369579), + KQU(15966784769764367791), KQU( 9627344291450607588), + KQU(10849555504977813287), KQU( 9234566913936339275), + KQU( 6413807690366911210), KQU(10862389016184219267), + KQU(13842504799335374048), KQU( 1531994113376881174), + KQU( 2081314867544364459), KQU(16430628791616959932), + KQU( 8314714038654394368), KQU( 9155473892098431813), + KQU(12577843786670475704), KQU( 4399161106452401017), + KQU( 1668083091682623186), KQU( 1741383777203714216), + KQU( 2162597285417794374), KQU(15841980159165218736), + KQU( 1971354603551467079), KQU( 1206714764913205968), + KQU( 4790860439591272330), KQU(14699375615594055799), + KQU( 8374423871657449988), KQU(10950685736472937738), + KQU( 697344331343267176), KQU(10084998763118059810), + KQU(12897369539795983124), KQU(12351260292144383605), + KQU( 1268810970176811234), KQU( 7406287800414582768), + KQU( 516169557043807831), KQU( 5077568278710520380), + KQU( 3828791738309039304), KQU( 7721974069946943610), + KQU( 3534670260981096460), KQU( 4865792189600584891), + KQU(16892578493734337298), KQU( 9161499464278042590), + KQU(11976149624067055931), KQU(13219479887277343990), + KQU(14161556738111500680), KQU(14670715255011223056), + KQU( 4671205678403576558), KQU(12633022931454259781), + KQU(14821376219869187646), KQU( 751181776484317028), + KQU( 2192211308839047070), KQU(11787306362361245189), + KQU(10672375120744095707), KQU( 4601972328345244467), + KQU(15457217788831125879), KQU( 8464345256775460809), + KQU(10191938789487159478), KQU( 6184348739615197613), + KQU(11425436778806882100), KQU( 2739227089124319793), + KQU( 461464518456000551), KQU( 4689850170029177442), + KQU( 6120307814374078625), KQU(11153579230681708671), + KQU( 7891721473905347926), KQU(10281646937824872400), + KQU( 3026099648191332248), KQU( 8666750296953273818), + KQU(14978499698844363232), KQU(13303395102890132065), + KQU( 8182358205292864080), KQU(10560547713972971291), + KQU(11981635489418959093), KQU( 3134621354935288409), + KQU(11580681977404383968), KQU(14205530317404088650), + KQU( 5997789011854923157), KQU(13659151593432238041), + KQU(11664332114338865086), KQU( 7490351383220929386), + KQU( 7189290499881530378), KQU(15039262734271020220), + KQU( 2057217285976980055), KQU( 555570804905355739), + KQU(11235311968348555110), KQU(13824557146269603217), + KQU(16906788840653099693), KQU( 7222878245455661677), + KQU( 5245139444332423756), KQU( 4723748462805674292), + KQU(12216509815698568612), KQU(17402362976648951187), + KQU(17389614836810366768), KQU( 4880936484146667711), + KQU( 9085007839292639880), KQU(13837353458498535449), + KQU(11914419854360366677), KQU(16595890135313864103), + KQU( 6313969847197627222), KQU(18296909792163910431), + KQU(10041780113382084042), KQU( 2499478551172884794), + KQU(11057894246241189489), KQU( 9742243032389068555), + KQU(12838934582673196228), KQU(13437023235248490367), + KQU(13372420669446163240), KQU( 6752564244716909224), + KQU( 7157333073400313737), KQU(12230281516370654308), + KQU( 1182884552219419117), KQU( 2955125381312499218), + KQU(10308827097079443249), KQU( 1337648572986534958), + KQU(16378788590020343939), KQU( 108619126514420935), + KQU( 3990981009621629188), KQU( 5460953070230946410), + KQU( 9703328329366531883), KQU(13166631489188077236), + KQU( 1104768831213675170), KQU( 3447930458553877908), + KQU( 8067172487769945676), KQU( 5445802098190775347), + KQU( 3244840981648973873), KQU(17314668322981950060), + KQU( 5006812527827763807), KQU(18158695070225526260), + KQU( 2824536478852417853), KQU(13974775809127519886), + KQU( 9814362769074067392), KQU(17276205156374862128), + KQU(11361680725379306967), KQU( 3422581970382012542), + KQU(11003189603753241266), KQU(11194292945277862261), + KQU( 6839623313908521348), KQU(11935326462707324634), + KQU( 1611456788685878444), KQU(13112620989475558907), + KQU( 517659108904450427), KQU(13558114318574407624), + KQU(15699089742731633077), KQU( 4988979278862685458), + KQU( 8111373583056521297), KQU( 3891258746615399627), + KQU( 8137298251469718086), KQU(12748663295624701649), + KQU( 4389835683495292062), KQU( 5775217872128831729), + KQU( 9462091896405534927), KQU( 8498124108820263989), + KQU( 8059131278842839525), KQU(10503167994254090892), + KQU(11613153541070396656), KQU(18069248738504647790), + KQU( 570657419109768508), KQU( 3950574167771159665), + KQU( 5514655599604313077), KQU( 2908460854428484165), + KQU(10777722615935663114), KQU(12007363304839279486), + KQU( 9800646187569484767), KQU( 8795423564889864287), + KQU(14257396680131028419), KQU( 6405465117315096498), + KQU( 7939411072208774878), KQU(17577572378528990006), + KQU(14785873806715994850), KQU(16770572680854747390), + KQU(18127549474419396481), KQU(11637013449455757750), + KQU(14371851933996761086), KQU( 3601181063650110280), + KQU( 4126442845019316144), KQU(10198287239244320669), + KQU(18000169628555379659), KQU(18392482400739978269), + KQU( 6219919037686919957), KQU( 3610085377719446052), + KQU( 2513925039981776336), KQU(16679413537926716955), + KQU(12903302131714909434), KQU( 5581145789762985009), + KQU(12325955044293303233), KQU(17216111180742141204), + KQU( 6321919595276545740), KQU( 3507521147216174501), + KQU( 9659194593319481840), KQU(11473976005975358326), + KQU(14742730101435987026), KQU( 492845897709954780), + KQU(16976371186162599676), KQU(17712703422837648655), + KQU( 9881254778587061697), KQU( 8413223156302299551), + KQU( 1563841828254089168), KQU( 9996032758786671975), + KQU( 138877700583772667), KQU(13003043368574995989), + KQU( 4390573668650456587), KQU( 8610287390568126755), + KQU(15126904974266642199), KQU( 6703637238986057662), + KQU( 2873075592956810157), KQU( 6035080933946049418), + KQU(13382846581202353014), KQU( 7303971031814642463), + KQU(18418024405307444267), KQU( 5847096731675404647), + KQU( 4035880699639842500), KQU(11525348625112218478), + KQU( 3041162365459574102), KQU( 2604734487727986558), + KQU(15526341771636983145), KQU(14556052310697370254), + KQU(12997787077930808155), KQU( 9601806501755554499), + KQU(11349677952521423389), KQU(14956777807644899350), + KQU(16559736957742852721), KQU(12360828274778140726), + KQU( 6685373272009662513), KQU(16932258748055324130), + KQU(15918051131954158508), KQU( 1692312913140790144), + KQU( 546653826801637367), KQU( 5341587076045986652), + KQU(14975057236342585662), KQU(12374976357340622412), + KQU(10328833995181940552), KQU(12831807101710443149), + KQU(10548514914382545716), KQU( 2217806727199715993), + KQU(12627067369242845138), KQU( 4598965364035438158), + KQU( 150923352751318171), KQU(14274109544442257283), + KQU( 4696661475093863031), KQU( 1505764114384654516), + KQU(10699185831891495147), KQU( 2392353847713620519), + KQU( 3652870166711788383), KQU( 8640653276221911108), + KQU( 3894077592275889704), KQU( 4918592872135964845), + KQU(16379121273281400789), KQU(12058465483591683656), + KQU(11250106829302924945), KQU( 1147537556296983005), + KQU( 6376342756004613268), KQU(14967128191709280506), + KQU(18007449949790627628), KQU( 9497178279316537841), + KQU( 7920174844809394893), KQU(10037752595255719907), + KQU(15875342784985217697), KQU(15311615921712850696), + KQU( 9552902652110992950), KQU(14054979450099721140), + KQU( 5998709773566417349), KQU(18027910339276320187), + KQU( 8223099053868585554), KQU( 7842270354824999767), + KQU( 4896315688770080292), KQU(12969320296569787895), + KQU( 2674321489185759961), KQU( 4053615936864718439), + KQU(11349775270588617578), KQU( 4743019256284553975), + KQU( 5602100217469723769), KQU(14398995691411527813), + KQU( 7412170493796825470), KQU( 836262406131744846), + KQU( 8231086633845153022), KQU( 5161377920438552287), + KQU( 8828731196169924949), KQU(16211142246465502680), + KQU( 3307990879253687818), KQU( 5193405406899782022), + KQU( 8510842117467566693), KQU( 6070955181022405365), + KQU(14482950231361409799), KQU(12585159371331138077), + KQU( 3511537678933588148), KQU( 2041849474531116417), + KQU(10944936685095345792), KQU(18303116923079107729), + KQU( 2720566371239725320), KQU( 4958672473562397622), + KQU( 3032326668253243412), KQU(13689418691726908338), + KQU( 1895205511728843996), KQU( 8146303515271990527), + KQU(16507343500056113480), KQU( 473996939105902919), + KQU( 9897686885246881481), KQU(14606433762712790575), + KQU( 6732796251605566368), KQU( 1399778120855368916), + KQU( 935023885182833777), KQU(16066282816186753477), + KQU( 7291270991820612055), KQU(17530230393129853844), + KQU(10223493623477451366), KQU(15841725630495676683), + KQU(17379567246435515824), KQU( 8588251429375561971), + KQU(18339511210887206423), KQU(17349587430725976100), + KQU(12244876521394838088), KQU( 6382187714147161259), + KQU(12335807181848950831), KQU(16948885622305460665), + KQU(13755097796371520506), KQU(14806740373324947801), + KQU( 4828699633859287703), KQU( 8209879281452301604), + KQU(12435716669553736437), KQU(13970976859588452131), + KQU( 6233960842566773148), KQU(12507096267900505759), + KQU( 1198713114381279421), KQU(14989862731124149015), + KQU(15932189508707978949), KQU( 2526406641432708722), + KQU( 29187427817271982), KQU( 1499802773054556353), + KQU(10816638187021897173), KQU( 5436139270839738132), + KQU( 6659882287036010082), KQU( 2154048955317173697), + KQU(10887317019333757642), KQU(16281091802634424955), + KQU(10754549879915384901), KQU(10760611745769249815), + KQU( 2161505946972504002), KQU( 5243132808986265107), + KQU(10129852179873415416), KQU( 710339480008649081), + KQU( 7802129453068808528), KQU(17967213567178907213), + KQU(15730859124668605599), KQU(13058356168962376502), + KQU( 3701224985413645909), KQU(14464065869149109264), + KQU( 9959272418844311646), KQU(10157426099515958752), + KQU(14013736814538268528), KQU(17797456992065653951), + KQU(17418878140257344806), KQU(15457429073540561521), + KQU( 2184426881360949378), KQU( 2062193041154712416), + KQU( 8553463347406931661), KQU( 4913057625202871854), + KQU( 2668943682126618425), KQU(17064444737891172288), + KQU( 4997115903913298637), KQU(12019402608892327416), + KQU(17603584559765897352), KQU(11367529582073647975), + KQU( 8211476043518436050), KQU( 8676849804070323674), + KQU(18431829230394475730), KQU(10490177861361247904), + KQU( 9508720602025651349), KQU( 7409627448555722700), + KQU( 5804047018862729008), KQU(11943858176893142594), + KQU(11908095418933847092), KQU( 5415449345715887652), + KQU( 1554022699166156407), KQU( 9073322106406017161), + KQU( 7080630967969047082), KQU(18049736940860732943), + KQU(12748714242594196794), KQU( 1226992415735156741), + KQU(17900981019609531193), KQU(11720739744008710999), + KQU( 3006400683394775434), KQU(11347974011751996028), + KQU( 3316999628257954608), KQU( 8384484563557639101), + KQU(18117794685961729767), KQU( 1900145025596618194), + KQU(17459527840632892676), KQU( 5634784101865710994), + KQU( 7918619300292897158), KQU( 3146577625026301350), + KQU( 9955212856499068767), KQU( 1873995843681746975), + KQU( 1561487759967972194), KQU( 8322718804375878474), + KQU(11300284215327028366), KQU( 4667391032508998982), + KQU( 9820104494306625580), KQU(17922397968599970610), + KQU( 1784690461886786712), KQU(14940365084341346821), + KQU( 5348719575594186181), KQU(10720419084507855261), + KQU(14210394354145143274), KQU( 2426468692164000131), + KQU(16271062114607059202), KQU(14851904092357070247), + KQU( 6524493015693121897), KQU( 9825473835127138531), + KQU(14222500616268569578), KQU(15521484052007487468), + KQU(14462579404124614699), KQU(11012375590820665520), + KQU(11625327350536084927), KQU(14452017765243785417), + KQU( 9989342263518766305), KQU( 3640105471101803790), + KQU( 4749866455897513242), KQU(13963064946736312044), + KQU(10007416591973223791), KQU(18314132234717431115), + KQU( 3286596588617483450), KQU( 7726163455370818765), + KQU( 7575454721115379328), KQU( 5308331576437663422), + KQU(18288821894903530934), KQU( 8028405805410554106), + KQU(15744019832103296628), KQU( 149765559630932100), + KQU( 6137705557200071977), KQU(14513416315434803615), + KQU(11665702820128984473), KQU( 218926670505601386), + KQU( 6868675028717769519), KQU(15282016569441512302), + KQU( 5707000497782960236), KQU( 6671120586555079567), + KQU( 2194098052618985448), KQU(16849577895477330978), + KQU(12957148471017466283), KQU( 1997805535404859393), + KQU( 1180721060263860490), KQU(13206391310193756958), + KQU(12980208674461861797), KQU( 3825967775058875366), + KQU(17543433670782042631), KQU( 1518339070120322730), + KQU(16344584340890991669), KQU( 2611327165318529819), + KQU(11265022723283422529), KQU( 4001552800373196817), + KQU(14509595890079346161), KQU( 3528717165416234562), + KQU(18153222571501914072), KQU( 9387182977209744425), + KQU(10064342315985580021), KQU(11373678413215253977), + KQU( 2308457853228798099), KQU( 9729042942839545302), + KQU( 7833785471140127746), KQU( 6351049900319844436), + KQU(14454610627133496067), KQU(12533175683634819111), + KQU(15570163926716513029), KQU(13356980519185762498) +}; + +TEST_BEGIN(test_gen_rand_32) { + uint32_t array32[BLOCK_SIZE] JEMALLOC_ATTR(aligned(16)); + uint32_t array32_2[BLOCK_SIZE] JEMALLOC_ATTR(aligned(16)); + int i; + uint32_t r32; + sfmt_t *ctx; + + assert_d_le(get_min_array_size32(), BLOCK_SIZE, + "Array size too small"); + ctx = init_gen_rand(1234); + fill_array32(ctx, array32, BLOCK_SIZE); + fill_array32(ctx, array32_2, BLOCK_SIZE); + fini_gen_rand(ctx); + + ctx = init_gen_rand(1234); + for (i = 0; i < BLOCK_SIZE; i++) { + if (i < COUNT_1) { + assert_u32_eq(array32[i], init_gen_rand_32_expected[i], + "Output mismatch for i=%d", i); + } + r32 = gen_rand32(ctx); + assert_u32_eq(r32, array32[i], + "Mismatch at array32[%d]=%x, gen=%x", i, array32[i], r32); + } + for (i = 0; i < COUNT_2; i++) { + r32 = gen_rand32(ctx); + assert_u32_eq(r32, array32_2[i], + "Mismatch at array32_2[%d]=%x, gen=%x", i, array32_2[i], + r32); + } + fini_gen_rand(ctx); +} +TEST_END + +TEST_BEGIN(test_by_array_32) { + uint32_t array32[BLOCK_SIZE] JEMALLOC_ATTR(aligned(16)); + uint32_t array32_2[BLOCK_SIZE] JEMALLOC_ATTR(aligned(16)); + int i; + uint32_t ini[4] = {0x1234, 0x5678, 0x9abc, 0xdef0}; + uint32_t r32; + sfmt_t *ctx; + + assert_d_le(get_min_array_size32(), BLOCK_SIZE, + "Array size too small"); + ctx = init_by_array(ini, 4); + fill_array32(ctx, array32, BLOCK_SIZE); + fill_array32(ctx, array32_2, BLOCK_SIZE); + fini_gen_rand(ctx); + + ctx = init_by_array(ini, 4); + for (i = 0; i < BLOCK_SIZE; i++) { + if (i < COUNT_1) { + assert_u32_eq(array32[i], init_by_array_32_expected[i], + "Output mismatch for i=%d", i); + } + r32 = gen_rand32(ctx); + assert_u32_eq(r32, array32[i], + "Mismatch at array32[%d]=%x, gen=%x", i, array32[i], r32); + } + for (i = 0; i < COUNT_2; i++) { + r32 = gen_rand32(ctx); + assert_u32_eq(r32, array32_2[i], + "Mismatch at array32_2[%d]=%x, gen=%x", i, array32_2[i], + r32); + } + fini_gen_rand(ctx); +} +TEST_END + +TEST_BEGIN(test_gen_rand_64) { + uint64_t array64[BLOCK_SIZE64] JEMALLOC_ATTR(aligned(16)); + uint64_t array64_2[BLOCK_SIZE64] JEMALLOC_ATTR(aligned(16)); + int i; + uint64_t r; + sfmt_t *ctx; + + assert_d_le(get_min_array_size64(), BLOCK_SIZE64, + "Array size too small"); + ctx = init_gen_rand(4321); + fill_array64(ctx, array64, BLOCK_SIZE64); + fill_array64(ctx, array64_2, BLOCK_SIZE64); + fini_gen_rand(ctx); + + ctx = init_gen_rand(4321); + for (i = 0; i < BLOCK_SIZE64; i++) { + if (i < COUNT_1) { + assert_u64_eq(array64[i], init_gen_rand_64_expected[i], + "Output mismatch for i=%d", i); + } + r = gen_rand64(ctx); + assert_u64_eq(r, array64[i], + "Mismatch at array64[%d]=%"FMTx64", gen=%"FMTx64, i, + array64[i], r); + } + for (i = 0; i < COUNT_2; i++) { + r = gen_rand64(ctx); + assert_u64_eq(r, array64_2[i], + "Mismatch at array64_2[%d]=%"FMTx64" gen=%"FMTx64"", i, + array64_2[i], r); + } + fini_gen_rand(ctx); +} +TEST_END + +TEST_BEGIN(test_by_array_64) { + uint64_t array64[BLOCK_SIZE64] JEMALLOC_ATTR(aligned(16)); + uint64_t array64_2[BLOCK_SIZE64] JEMALLOC_ATTR(aligned(16)); + int i; + uint64_t r; + uint32_t ini[] = {5, 4, 3, 2, 1}; + sfmt_t *ctx; + + assert_d_le(get_min_array_size64(), BLOCK_SIZE64, + "Array size too small"); + ctx = init_by_array(ini, 5); + fill_array64(ctx, array64, BLOCK_SIZE64); + fill_array64(ctx, array64_2, BLOCK_SIZE64); + fini_gen_rand(ctx); + + ctx = init_by_array(ini, 5); + for (i = 0; i < BLOCK_SIZE64; i++) { + if (i < COUNT_1) { + assert_u64_eq(array64[i], init_by_array_64_expected[i], + "Output mismatch for i=%d", i); + } + r = gen_rand64(ctx); + assert_u64_eq(r, array64[i], + "Mismatch at array64[%d]=%"FMTx64" gen=%"FMTx64, i, + array64[i], r); + } + for (i = 0; i < COUNT_2; i++) { + r = gen_rand64(ctx); + assert_u64_eq(r, array64_2[i], + "Mismatch at array64_2[%d]=%"FMTx64" gen=%"FMTx64, i, + array64_2[i], r); + } + fini_gen_rand(ctx); +} +TEST_END + +int +main(void) { + return test( + test_gen_rand_32, + test_by_array_32, + test_gen_rand_64, + test_by_array_64); +} diff --git a/deps/jemalloc/test/unit/a0.c b/deps/jemalloc/test/unit/a0.c new file mode 100644 index 000000000..a27ab3f42 --- /dev/null +++ b/deps/jemalloc/test/unit/a0.c @@ -0,0 +1,16 @@ +#include "test/jemalloc_test.h" + +TEST_BEGIN(test_a0) { + void *p; + + p = a0malloc(1); + assert_ptr_not_null(p, "Unexpected a0malloc() error"); + a0dalloc(p); +} +TEST_END + +int +main(void) { + return test_no_malloc_init( + test_a0); +} diff --git a/deps/jemalloc/test/unit/arena_reset.c b/deps/jemalloc/test/unit/arena_reset.c new file mode 100644 index 000000000..b182f31a6 --- /dev/null +++ b/deps/jemalloc/test/unit/arena_reset.c @@ -0,0 +1,349 @@ +#ifndef ARENA_RESET_PROF_C_ +#include "test/jemalloc_test.h" +#endif + +#include "jemalloc/internal/extent_mmap.h" +#include "jemalloc/internal/rtree.h" + +#include "test/extent_hooks.h" + +static unsigned +get_nsizes_impl(const char *cmd) { + unsigned ret; + size_t z; + + z = sizeof(unsigned); + assert_d_eq(mallctl(cmd, (void *)&ret, &z, NULL, 0), 0, + "Unexpected mallctl(\"%s\", ...) failure", cmd); + + return ret; +} + +static unsigned +get_nsmall(void) { + return get_nsizes_impl("arenas.nbins"); +} + +static unsigned +get_nlarge(void) { + return get_nsizes_impl("arenas.nlextents"); +} + +static size_t +get_size_impl(const char *cmd, size_t ind) { + size_t ret; + size_t z; + size_t mib[4]; + size_t miblen = 4; + + z = sizeof(size_t); + assert_d_eq(mallctlnametomib(cmd, mib, &miblen), + 0, "Unexpected mallctlnametomib(\"%s\", ...) failure", cmd); + mib[2] = ind; + z = sizeof(size_t); + assert_d_eq(mallctlbymib(mib, miblen, (void *)&ret, &z, NULL, 0), + 0, "Unexpected mallctlbymib([\"%s\", %zu], ...) failure", cmd, ind); + + return ret; +} + +static size_t +get_small_size(size_t ind) { + return get_size_impl("arenas.bin.0.size", ind); +} + +static size_t +get_large_size(size_t ind) { + return get_size_impl("arenas.lextent.0.size", ind); +} + +/* Like ivsalloc(), but safe to call on discarded allocations. */ +static size_t +vsalloc(tsdn_t *tsdn, const void *ptr) { + rtree_ctx_t rtree_ctx_fallback; + rtree_ctx_t *rtree_ctx = tsdn_rtree_ctx(tsdn, &rtree_ctx_fallback); + + extent_t *extent; + szind_t szind; + if (rtree_extent_szind_read(tsdn, &extents_rtree, rtree_ctx, + (uintptr_t)ptr, false, &extent, &szind)) { + return 0; + } + + if (extent == NULL) { + return 0; + } + if (extent_state_get(extent) != extent_state_active) { + return 0; + } + + if (szind == SC_NSIZES) { + return 0; + } + + return sz_index2size(szind); +} + +static unsigned +do_arena_create(extent_hooks_t *h) { + unsigned arena_ind; + size_t sz = sizeof(unsigned); + assert_d_eq(mallctl("arenas.create", (void *)&arena_ind, &sz, + (void *)(h != NULL ? &h : NULL), (h != NULL ? sizeof(h) : 0)), 0, + "Unexpected mallctl() failure"); + return arena_ind; +} + +static void +do_arena_reset_pre(unsigned arena_ind, void ***ptrs, unsigned *nptrs) { +#define NLARGE 32 + unsigned nsmall, nlarge, i; + size_t sz; + int flags; + tsdn_t *tsdn; + + flags = MALLOCX_ARENA(arena_ind) | MALLOCX_TCACHE_NONE; + + nsmall = get_nsmall(); + nlarge = get_nlarge() > NLARGE ? NLARGE : get_nlarge(); + *nptrs = nsmall + nlarge; + *ptrs = (void **)malloc(*nptrs * sizeof(void *)); + assert_ptr_not_null(*ptrs, "Unexpected malloc() failure"); + + /* Allocate objects with a wide range of sizes. */ + for (i = 0; i < nsmall; i++) { + sz = get_small_size(i); + (*ptrs)[i] = mallocx(sz, flags); + assert_ptr_not_null((*ptrs)[i], + "Unexpected mallocx(%zu, %#x) failure", sz, flags); + } + for (i = 0; i < nlarge; i++) { + sz = get_large_size(i); + (*ptrs)[nsmall + i] = mallocx(sz, flags); + assert_ptr_not_null((*ptrs)[i], + "Unexpected mallocx(%zu, %#x) failure", sz, flags); + } + + tsdn = tsdn_fetch(); + + /* Verify allocations. */ + for (i = 0; i < *nptrs; i++) { + assert_zu_gt(ivsalloc(tsdn, (*ptrs)[i]), 0, + "Allocation should have queryable size"); + } +} + +static void +do_arena_reset_post(void **ptrs, unsigned nptrs, unsigned arena_ind) { + tsdn_t *tsdn; + unsigned i; + + tsdn = tsdn_fetch(); + + if (have_background_thread) { + malloc_mutex_lock(tsdn, + &background_thread_info_get(arena_ind)->mtx); + } + /* Verify allocations no longer exist. */ + for (i = 0; i < nptrs; i++) { + assert_zu_eq(vsalloc(tsdn, ptrs[i]), 0, + "Allocation should no longer exist"); + } + if (have_background_thread) { + malloc_mutex_unlock(tsdn, + &background_thread_info_get(arena_ind)->mtx); + } + + free(ptrs); +} + +static void +do_arena_reset_destroy(const char *name, unsigned arena_ind) { + size_t mib[3]; + size_t miblen; + + miblen = sizeof(mib)/sizeof(size_t); + assert_d_eq(mallctlnametomib(name, mib, &miblen), 0, + "Unexpected mallctlnametomib() failure"); + mib[1] = (size_t)arena_ind; + assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, NULL, 0), 0, + "Unexpected mallctlbymib() failure"); +} + +static void +do_arena_reset(unsigned arena_ind) { + do_arena_reset_destroy("arena.0.reset", arena_ind); +} + +static void +do_arena_destroy(unsigned arena_ind) { + do_arena_reset_destroy("arena.0.destroy", arena_ind); +} + +TEST_BEGIN(test_arena_reset) { + unsigned arena_ind; + void **ptrs; + unsigned nptrs; + + arena_ind = do_arena_create(NULL); + do_arena_reset_pre(arena_ind, &ptrs, &nptrs); + do_arena_reset(arena_ind); + do_arena_reset_post(ptrs, nptrs, arena_ind); +} +TEST_END + +static bool +arena_i_initialized(unsigned arena_ind, bool refresh) { + bool initialized; + size_t mib[3]; + size_t miblen, sz; + + if (refresh) { + uint64_t epoch = 1; + assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch, + sizeof(epoch)), 0, "Unexpected mallctl() failure"); + } + + miblen = sizeof(mib)/sizeof(size_t); + assert_d_eq(mallctlnametomib("arena.0.initialized", mib, &miblen), 0, + "Unexpected mallctlnametomib() failure"); + mib[1] = (size_t)arena_ind; + sz = sizeof(initialized); + assert_d_eq(mallctlbymib(mib, miblen, (void *)&initialized, &sz, NULL, + 0), 0, "Unexpected mallctlbymib() failure"); + + return initialized; +} + +TEST_BEGIN(test_arena_destroy_initial) { + assert_false(arena_i_initialized(MALLCTL_ARENAS_DESTROYED, false), + "Destroyed arena stats should not be initialized"); +} +TEST_END + +TEST_BEGIN(test_arena_destroy_hooks_default) { + unsigned arena_ind, arena_ind_another, arena_ind_prev; + void **ptrs; + unsigned nptrs; + + arena_ind = do_arena_create(NULL); + do_arena_reset_pre(arena_ind, &ptrs, &nptrs); + + assert_false(arena_i_initialized(arena_ind, false), + "Arena stats should not be initialized"); + assert_true(arena_i_initialized(arena_ind, true), + "Arena stats should be initialized"); + + /* + * Create another arena before destroying one, to better verify arena + * index reuse. + */ + arena_ind_another = do_arena_create(NULL); + + do_arena_destroy(arena_ind); + + assert_false(arena_i_initialized(arena_ind, true), + "Arena stats should not be initialized"); + assert_true(arena_i_initialized(MALLCTL_ARENAS_DESTROYED, false), + "Destroyed arena stats should be initialized"); + + do_arena_reset_post(ptrs, nptrs, arena_ind); + + arena_ind_prev = arena_ind; + arena_ind = do_arena_create(NULL); + do_arena_reset_pre(arena_ind, &ptrs, &nptrs); + assert_u_eq(arena_ind, arena_ind_prev, + "Arena index should have been recycled"); + do_arena_destroy(arena_ind); + do_arena_reset_post(ptrs, nptrs, arena_ind); + + do_arena_destroy(arena_ind_another); +} +TEST_END + +/* + * Actually unmap extents, regardless of opt_retain, so that attempts to access + * a destroyed arena's memory will segfault. + */ +static bool +extent_dalloc_unmap(extent_hooks_t *extent_hooks, void *addr, size_t size, + bool committed, unsigned arena_ind) { + TRACE_HOOK("%s(extent_hooks=%p, addr=%p, size=%zu, committed=%s, " + "arena_ind=%u)\n", __func__, extent_hooks, addr, size, committed ? + "true" : "false", arena_ind); + assert_ptr_eq(extent_hooks, &hooks, + "extent_hooks should be same as pointer used to set hooks"); + assert_ptr_eq(extent_hooks->dalloc, extent_dalloc_unmap, + "Wrong hook function"); + called_dalloc = true; + if (!try_dalloc) { + return true; + } + did_dalloc = true; + if (!maps_coalesce && opt_retain) { + return true; + } + pages_unmap(addr, size); + return false; +} + +static extent_hooks_t hooks_orig; + +static extent_hooks_t hooks_unmap = { + extent_alloc_hook, + extent_dalloc_unmap, /* dalloc */ + extent_destroy_hook, + extent_commit_hook, + extent_decommit_hook, + extent_purge_lazy_hook, + extent_purge_forced_hook, + extent_split_hook, + extent_merge_hook +}; + +TEST_BEGIN(test_arena_destroy_hooks_unmap) { + unsigned arena_ind; + void **ptrs; + unsigned nptrs; + + extent_hooks_prep(); + if (maps_coalesce) { + try_decommit = false; + } + memcpy(&hooks_orig, &hooks, sizeof(extent_hooks_t)); + memcpy(&hooks, &hooks_unmap, sizeof(extent_hooks_t)); + + did_alloc = false; + arena_ind = do_arena_create(&hooks); + do_arena_reset_pre(arena_ind, &ptrs, &nptrs); + + assert_true(did_alloc, "Expected alloc"); + + assert_false(arena_i_initialized(arena_ind, false), + "Arena stats should not be initialized"); + assert_true(arena_i_initialized(arena_ind, true), + "Arena stats should be initialized"); + + did_dalloc = false; + do_arena_destroy(arena_ind); + assert_true(did_dalloc, "Expected dalloc"); + + assert_false(arena_i_initialized(arena_ind, true), + "Arena stats should not be initialized"); + assert_true(arena_i_initialized(MALLCTL_ARENAS_DESTROYED, false), + "Destroyed arena stats should be initialized"); + + do_arena_reset_post(ptrs, nptrs, arena_ind); + + memcpy(&hooks, &hooks_orig, sizeof(extent_hooks_t)); +} +TEST_END + +int +main(void) { + return test( + test_arena_reset, + test_arena_destroy_initial, + test_arena_destroy_hooks_default, + test_arena_destroy_hooks_unmap); +} diff --git a/deps/jemalloc/test/unit/arena_reset_prof.c b/deps/jemalloc/test/unit/arena_reset_prof.c new file mode 100644 index 000000000..38d801240 --- /dev/null +++ b/deps/jemalloc/test/unit/arena_reset_prof.c @@ -0,0 +1,4 @@ +#include "test/jemalloc_test.h" +#define ARENA_RESET_PROF_C_ + +#include "arena_reset.c" diff --git a/deps/jemalloc/test/unit/arena_reset_prof.sh b/deps/jemalloc/test/unit/arena_reset_prof.sh new file mode 100644 index 000000000..041dc1c35 --- /dev/null +++ b/deps/jemalloc/test/unit/arena_reset_prof.sh @@ -0,0 +1,3 @@ +#!/bin/sh + +export MALLOC_CONF="prof:true,lg_prof_sample:0" diff --git a/deps/jemalloc/test/unit/atomic.c b/deps/jemalloc/test/unit/atomic.c new file mode 100644 index 000000000..572d8d23f --- /dev/null +++ b/deps/jemalloc/test/unit/atomic.c @@ -0,0 +1,229 @@ +#include "test/jemalloc_test.h" + +/* + * We *almost* have consistent short names (e.g. "u32" for uint32_t, "b" for + * bool, etc. The one exception is that the short name for void * is "p" in + * some places and "ptr" in others. In the long run it would be nice to unify + * these, but in the short run we'll use this shim. + */ +#define assert_p_eq assert_ptr_eq + +/* + * t: the non-atomic type, like "uint32_t". + * ta: the short name for the type, like "u32". + * val[1,2,3]: Values of the given type. The CAS tests use val2 for expected, + * and val3 for desired. + */ + +#define DO_TESTS(t, ta, val1, val2, val3) do { \ + t val; \ + t expected; \ + bool success; \ + /* This (along with the load below) also tests ATOMIC_LOAD. */ \ + atomic_##ta##_t atom = ATOMIC_INIT(val1); \ + \ + /* ATOMIC_INIT and load. */ \ + val = atomic_load_##ta(&atom, ATOMIC_RELAXED); \ + assert_##ta##_eq(val1, val, "Load or init failed"); \ + \ + /* Store. */ \ + atomic_store_##ta(&atom, val1, ATOMIC_RELAXED); \ + atomic_store_##ta(&atom, val2, ATOMIC_RELAXED); \ + val = atomic_load_##ta(&atom, ATOMIC_RELAXED); \ + assert_##ta##_eq(val2, val, "Store failed"); \ + \ + /* Exchange. */ \ + atomic_store_##ta(&atom, val1, ATOMIC_RELAXED); \ + val = atomic_exchange_##ta(&atom, val2, ATOMIC_RELAXED); \ + assert_##ta##_eq(val1, val, "Exchange returned invalid value"); \ + val = atomic_load_##ta(&atom, ATOMIC_RELAXED); \ + assert_##ta##_eq(val2, val, "Exchange store invalid value"); \ + \ + /* \ + * Weak CAS. Spurious failures are allowed, so we loop a few \ + * times. \ + */ \ + atomic_store_##ta(&atom, val1, ATOMIC_RELAXED); \ + success = false; \ + for (int i = 0; i < 10 && !success; i++) { \ + expected = val2; \ + success = atomic_compare_exchange_weak_##ta(&atom, \ + &expected, val3, ATOMIC_RELAXED, ATOMIC_RELAXED); \ + assert_##ta##_eq(val1, expected, \ + "CAS should update expected"); \ + } \ + assert_b_eq(val1 == val2, success, \ + "Weak CAS did the wrong state update"); \ + val = atomic_load_##ta(&atom, ATOMIC_RELAXED); \ + if (success) { \ + assert_##ta##_eq(val3, val, \ + "Successful CAS should update atomic"); \ + } else { \ + assert_##ta##_eq(val1, val, \ + "Unsuccessful CAS should not update atomic"); \ + } \ + \ + /* Strong CAS. */ \ + atomic_store_##ta(&atom, val1, ATOMIC_RELAXED); \ + expected = val2; \ + success = atomic_compare_exchange_strong_##ta(&atom, &expected, \ + val3, ATOMIC_RELAXED, ATOMIC_RELAXED); \ + assert_b_eq(val1 == val2, success, \ + "Strong CAS did the wrong state update"); \ + val = atomic_load_##ta(&atom, ATOMIC_RELAXED); \ + if (success) { \ + assert_##ta##_eq(val3, val, \ + "Successful CAS should update atomic"); \ + } else { \ + assert_##ta##_eq(val1, val, \ + "Unsuccessful CAS should not update atomic"); \ + } \ + \ + \ +} while (0) + +#define DO_INTEGER_TESTS(t, ta, val1, val2) do { \ + atomic_##ta##_t atom; \ + t val; \ + \ + /* Fetch-add. */ \ + atomic_store_##ta(&atom, val1, ATOMIC_RELAXED); \ + val = atomic_fetch_add_##ta(&atom, val2, ATOMIC_RELAXED); \ + assert_##ta##_eq(val1, val, \ + "Fetch-add should return previous value"); \ + val = atomic_load_##ta(&atom, ATOMIC_RELAXED); \ + assert_##ta##_eq(val1 + val2, val, \ + "Fetch-add should update atomic"); \ + \ + /* Fetch-sub. */ \ + atomic_store_##ta(&atom, val1, ATOMIC_RELAXED); \ + val = atomic_fetch_sub_##ta(&atom, val2, ATOMIC_RELAXED); \ + assert_##ta##_eq(val1, val, \ + "Fetch-sub should return previous value"); \ + val = atomic_load_##ta(&atom, ATOMIC_RELAXED); \ + assert_##ta##_eq(val1 - val2, val, \ + "Fetch-sub should update atomic"); \ + \ + /* Fetch-and. */ \ + atomic_store_##ta(&atom, val1, ATOMIC_RELAXED); \ + val = atomic_fetch_and_##ta(&atom, val2, ATOMIC_RELAXED); \ + assert_##ta##_eq(val1, val, \ + "Fetch-and should return previous value"); \ + val = atomic_load_##ta(&atom, ATOMIC_RELAXED); \ + assert_##ta##_eq(val1 & val2, val, \ + "Fetch-and should update atomic"); \ + \ + /* Fetch-or. */ \ + atomic_store_##ta(&atom, val1, ATOMIC_RELAXED); \ + val = atomic_fetch_or_##ta(&atom, val2, ATOMIC_RELAXED); \ + assert_##ta##_eq(val1, val, \ + "Fetch-or should return previous value"); \ + val = atomic_load_##ta(&atom, ATOMIC_RELAXED); \ + assert_##ta##_eq(val1 | val2, val, \ + "Fetch-or should update atomic"); \ + \ + /* Fetch-xor. */ \ + atomic_store_##ta(&atom, val1, ATOMIC_RELAXED); \ + val = atomic_fetch_xor_##ta(&atom, val2, ATOMIC_RELAXED); \ + assert_##ta##_eq(val1, val, \ + "Fetch-xor should return previous value"); \ + val = atomic_load_##ta(&atom, ATOMIC_RELAXED); \ + assert_##ta##_eq(val1 ^ val2, val, \ + "Fetch-xor should update atomic"); \ +} while (0) + +#define TEST_STRUCT(t, ta) \ +typedef struct { \ + t val1; \ + t val2; \ + t val3; \ +} ta##_test_t; + +#define TEST_CASES(t) { \ + {(t)-1, (t)-1, (t)-2}, \ + {(t)-1, (t) 0, (t)-2}, \ + {(t)-1, (t) 1, (t)-2}, \ + \ + {(t) 0, (t)-1, (t)-2}, \ + {(t) 0, (t) 0, (t)-2}, \ + {(t) 0, (t) 1, (t)-2}, \ + \ + {(t) 1, (t)-1, (t)-2}, \ + {(t) 1, (t) 0, (t)-2}, \ + {(t) 1, (t) 1, (t)-2}, \ + \ + {(t)0, (t)-(1 << 22), (t)-2}, \ + {(t)0, (t)(1 << 22), (t)-2}, \ + {(t)(1 << 22), (t)-(1 << 22), (t)-2}, \ + {(t)(1 << 22), (t)(1 << 22), (t)-2} \ +} + +#define TEST_BODY(t, ta) do { \ + const ta##_test_t tests[] = TEST_CASES(t); \ + for (unsigned i = 0; i < sizeof(tests)/sizeof(tests[0]); i++) { \ + ta##_test_t test = tests[i]; \ + DO_TESTS(t, ta, test.val1, test.val2, test.val3); \ + } \ +} while (0) + +#define INTEGER_TEST_BODY(t, ta) do { \ + const ta##_test_t tests[] = TEST_CASES(t); \ + for (unsigned i = 0; i < sizeof(tests)/sizeof(tests[0]); i++) { \ + ta##_test_t test = tests[i]; \ + DO_TESTS(t, ta, test.val1, test.val2, test.val3); \ + DO_INTEGER_TESTS(t, ta, test.val1, test.val2); \ + } \ +} while (0) + +TEST_STRUCT(uint64_t, u64); +TEST_BEGIN(test_atomic_u64) { +#if !(LG_SIZEOF_PTR == 3 || LG_SIZEOF_INT == 3) + test_skip("64-bit atomic operations not supported"); +#else + INTEGER_TEST_BODY(uint64_t, u64); +#endif +} +TEST_END + + +TEST_STRUCT(uint32_t, u32); +TEST_BEGIN(test_atomic_u32) { + INTEGER_TEST_BODY(uint32_t, u32); +} +TEST_END + +TEST_STRUCT(void *, p); +TEST_BEGIN(test_atomic_p) { + TEST_BODY(void *, p); +} +TEST_END + +TEST_STRUCT(size_t, zu); +TEST_BEGIN(test_atomic_zu) { + INTEGER_TEST_BODY(size_t, zu); +} +TEST_END + +TEST_STRUCT(ssize_t, zd); +TEST_BEGIN(test_atomic_zd) { + INTEGER_TEST_BODY(ssize_t, zd); +} +TEST_END + + +TEST_STRUCT(unsigned, u); +TEST_BEGIN(test_atomic_u) { + INTEGER_TEST_BODY(unsigned, u); +} +TEST_END + +int +main(void) { + return test( + test_atomic_u64, + test_atomic_u32, + test_atomic_p, + test_atomic_zu, + test_atomic_zd, + test_atomic_u); +} diff --git a/deps/jemalloc/test/unit/background_thread.c b/deps/jemalloc/test/unit/background_thread.c new file mode 100644 index 000000000..f7bd37c42 --- /dev/null +++ b/deps/jemalloc/test/unit/background_thread.c @@ -0,0 +1,119 @@ +#include "test/jemalloc_test.h" + +#include "jemalloc/internal/util.h" + +static void +test_switch_background_thread_ctl(bool new_val) { + bool e0, e1; + size_t sz = sizeof(bool); + + e1 = new_val; + assert_d_eq(mallctl("background_thread", (void *)&e0, &sz, + &e1, sz), 0, "Unexpected mallctl() failure"); + assert_b_eq(e0, !e1, + "background_thread should be %d before.\n", !e1); + if (e1) { + assert_zu_gt(n_background_threads, 0, + "Number of background threads should be non zero.\n"); + } else { + assert_zu_eq(n_background_threads, 0, + "Number of background threads should be zero.\n"); + } +} + +static void +test_repeat_background_thread_ctl(bool before) { + bool e0, e1; + size_t sz = sizeof(bool); + + e1 = before; + assert_d_eq(mallctl("background_thread", (void *)&e0, &sz, + &e1, sz), 0, "Unexpected mallctl() failure"); + assert_b_eq(e0, before, + "background_thread should be %d.\n", before); + if (e1) { + assert_zu_gt(n_background_threads, 0, + "Number of background threads should be non zero.\n"); + } else { + assert_zu_eq(n_background_threads, 0, + "Number of background threads should be zero.\n"); + } +} + +TEST_BEGIN(test_background_thread_ctl) { + test_skip_if(!have_background_thread); + + bool e0, e1; + size_t sz = sizeof(bool); + + assert_d_eq(mallctl("opt.background_thread", (void *)&e0, &sz, + NULL, 0), 0, "Unexpected mallctl() failure"); + assert_d_eq(mallctl("background_thread", (void *)&e1, &sz, + NULL, 0), 0, "Unexpected mallctl() failure"); + assert_b_eq(e0, e1, + "Default and opt.background_thread does not match.\n"); + if (e0) { + test_switch_background_thread_ctl(false); + } + assert_zu_eq(n_background_threads, 0, + "Number of background threads should be 0.\n"); + + for (unsigned i = 0; i < 4; i++) { + test_switch_background_thread_ctl(true); + test_repeat_background_thread_ctl(true); + test_repeat_background_thread_ctl(true); + + test_switch_background_thread_ctl(false); + test_repeat_background_thread_ctl(false); + test_repeat_background_thread_ctl(false); + } +} +TEST_END + +TEST_BEGIN(test_background_thread_running) { + test_skip_if(!have_background_thread); + test_skip_if(!config_stats); + +#if defined(JEMALLOC_BACKGROUND_THREAD) + tsd_t *tsd = tsd_fetch(); + background_thread_info_t *info = &background_thread_info[0]; + + test_repeat_background_thread_ctl(false); + test_switch_background_thread_ctl(true); + assert_b_eq(info->state, background_thread_started, + "Background_thread did not start.\n"); + + nstime_t start, now; + nstime_init(&start, 0); + nstime_update(&start); + + bool ran = false; + while (true) { + malloc_mutex_lock(tsd_tsdn(tsd), &info->mtx); + if (info->tot_n_runs > 0) { + ran = true; + } + malloc_mutex_unlock(tsd_tsdn(tsd), &info->mtx); + if (ran) { + break; + } + + nstime_init(&now, 0); + nstime_update(&now); + nstime_subtract(&now, &start); + assert_u64_lt(nstime_sec(&now), 1000, + "Background threads did not run for 1000 seconds."); + sleep(1); + } + test_switch_background_thread_ctl(false); +#endif +} +TEST_END + +int +main(void) { + /* Background_thread creation tests reentrancy naturally. */ + return test_no_reentrancy( + test_background_thread_ctl, + test_background_thread_running); +} diff --git a/deps/jemalloc/test/unit/background_thread_enable.c b/deps/jemalloc/test/unit/background_thread_enable.c new file mode 100644 index 000000000..d894e9371 --- /dev/null +++ b/deps/jemalloc/test/unit/background_thread_enable.c @@ -0,0 +1,85 @@ +#include "test/jemalloc_test.h" + +const char *malloc_conf = "background_thread:false,narenas:1,max_background_threads:20"; + +TEST_BEGIN(test_deferred) { + test_skip_if(!have_background_thread); + + unsigned id; + size_t sz_u = sizeof(unsigned); + + /* + * 10 here is somewhat arbitrary, except insofar as we want to ensure + * that the number of background threads is smaller than the number of + * arenas. I'll ragequit long before we have to spin up 10 threads per + * cpu to handle background purging, so this is a conservative + * approximation. + */ + for (unsigned i = 0; i < 10 * ncpus; i++) { + assert_d_eq(mallctl("arenas.create", &id, &sz_u, NULL, 0), 0, + "Failed to create arena"); + } + + bool enable = true; + size_t sz_b = sizeof(bool); + assert_d_eq(mallctl("background_thread", NULL, NULL, &enable, sz_b), 0, + "Failed to enable background threads"); + enable = false; + assert_d_eq(mallctl("background_thread", NULL, NULL, &enable, sz_b), 0, + "Failed to disable background threads"); +} +TEST_END + +TEST_BEGIN(test_max_background_threads) { + test_skip_if(!have_background_thread); + + size_t max_n_thds; + size_t opt_max_n_thds; + size_t sz_m = sizeof(max_n_thds); + assert_d_eq(mallctl("opt.max_background_threads", + &opt_max_n_thds, &sz_m, NULL, 0), 0, + "Failed to get opt.max_background_threads"); + assert_d_eq(mallctl("max_background_threads", &max_n_thds, &sz_m, NULL, + 0), 0, "Failed to get max background threads"); + assert_zu_eq(opt_max_n_thds, max_n_thds, + "max_background_threads and " + "opt.max_background_threads should match"); + assert_d_eq(mallctl("max_background_threads", NULL, NULL, &max_n_thds, + sz_m), 0, "Failed to set max background threads"); + + unsigned id; + size_t sz_u = sizeof(unsigned); + + for (unsigned i = 0; i < 10 * ncpus; i++) { + assert_d_eq(mallctl("arenas.create", &id, &sz_u, NULL, 0), 0, + "Failed to create arena"); + } + + bool enable = true; + size_t sz_b = sizeof(bool); + assert_d_eq(mallctl("background_thread", NULL, NULL, &enable, sz_b), 0, + "Failed to enable background threads"); + assert_zu_eq(n_background_threads, max_n_thds, + "Number of background threads should not change.\n"); + size_t new_max_thds = max_n_thds - 1; + if (new_max_thds > 0) { + assert_d_eq(mallctl("max_background_threads", NULL, NULL, + &new_max_thds, sz_m), 0, + "Failed to set max background threads"); + assert_zu_eq(n_background_threads, new_max_thds, + "Number of background threads should decrease by 1.\n"); + } + new_max_thds = 1; + assert_d_eq(mallctl("max_background_threads", NULL, NULL, &new_max_thds, + sz_m), 0, "Failed to set max background threads"); + assert_zu_eq(n_background_threads, new_max_thds, + "Number of background threads should be 1.\n"); +} +TEST_END + +int +main(void) { + return test_no_reentrancy( + test_deferred, + test_max_background_threads); +} diff --git a/deps/jemalloc/test/unit/base.c b/deps/jemalloc/test/unit/base.c new file mode 100644 index 000000000..6b792cf21 --- /dev/null +++ b/deps/jemalloc/test/unit/base.c @@ -0,0 +1,234 @@ +#include "test/jemalloc_test.h" + +#include "test/extent_hooks.h" + +static extent_hooks_t hooks_null = { + extent_alloc_hook, + NULL, /* dalloc */ + NULL, /* destroy */ + NULL, /* commit */ + NULL, /* decommit */ + NULL, /* purge_lazy */ + NULL, /* purge_forced */ + NULL, /* split */ + NULL /* merge */ +}; + +static extent_hooks_t hooks_not_null = { + extent_alloc_hook, + extent_dalloc_hook, + extent_destroy_hook, + NULL, /* commit */ + extent_decommit_hook, + extent_purge_lazy_hook, + extent_purge_forced_hook, + NULL, /* split */ + NULL /* merge */ +}; + +TEST_BEGIN(test_base_hooks_default) { + base_t *base; + size_t allocated0, allocated1, resident, mapped, n_thp; + + tsdn_t *tsdn = tsd_tsdn(tsd_fetch()); + base = base_new(tsdn, 0, (extent_hooks_t *)&extent_hooks_default); + + if (config_stats) { + base_stats_get(tsdn, base, &allocated0, &resident, &mapped, + &n_thp); + assert_zu_ge(allocated0, sizeof(base_t), + "Base header should count as allocated"); + if (opt_metadata_thp == metadata_thp_always) { + assert_zu_gt(n_thp, 0, + "Base should have 1 THP at least."); + } + } + + assert_ptr_not_null(base_alloc(tsdn, base, 42, 1), + "Unexpected base_alloc() failure"); + + if (config_stats) { + base_stats_get(tsdn, base, &allocated1, &resident, &mapped, + &n_thp); + assert_zu_ge(allocated1 - allocated0, 42, + "At least 42 bytes were allocated by base_alloc()"); + } + + base_delete(tsdn, base); +} +TEST_END + +TEST_BEGIN(test_base_hooks_null) { + extent_hooks_t hooks_orig; + base_t *base; + size_t allocated0, allocated1, resident, mapped, n_thp; + + extent_hooks_prep(); + try_dalloc = false; + try_destroy = true; + try_decommit = false; + try_purge_lazy = false; + try_purge_forced = false; + memcpy(&hooks_orig, &hooks, sizeof(extent_hooks_t)); + memcpy(&hooks, &hooks_null, sizeof(extent_hooks_t)); + + tsdn_t *tsdn = tsd_tsdn(tsd_fetch()); + base = base_new(tsdn, 0, &hooks); + assert_ptr_not_null(base, "Unexpected base_new() failure"); + + if (config_stats) { + base_stats_get(tsdn, base, &allocated0, &resident, &mapped, + &n_thp); + assert_zu_ge(allocated0, sizeof(base_t), + "Base header should count as allocated"); + if (opt_metadata_thp == metadata_thp_always) { + assert_zu_gt(n_thp, 0, + "Base should have 1 THP at least."); + } + } + + assert_ptr_not_null(base_alloc(tsdn, base, 42, 1), + "Unexpected base_alloc() failure"); + + if (config_stats) { + base_stats_get(tsdn, base, &allocated1, &resident, &mapped, + &n_thp); + assert_zu_ge(allocated1 - allocated0, 42, + "At least 42 bytes were allocated by base_alloc()"); + } + + base_delete(tsdn, base); + + memcpy(&hooks, &hooks_orig, sizeof(extent_hooks_t)); +} +TEST_END + +TEST_BEGIN(test_base_hooks_not_null) { + extent_hooks_t hooks_orig; + base_t *base; + void *p, *q, *r, *r_exp; + + extent_hooks_prep(); + try_dalloc = false; + try_destroy = true; + try_decommit = false; + try_purge_lazy = false; + try_purge_forced = false; + memcpy(&hooks_orig, &hooks, sizeof(extent_hooks_t)); + memcpy(&hooks, &hooks_not_null, sizeof(extent_hooks_t)); + + tsdn_t *tsdn = tsd_tsdn(tsd_fetch()); + did_alloc = false; + base = base_new(tsdn, 0, &hooks); + assert_ptr_not_null(base, "Unexpected base_new() failure"); + assert_true(did_alloc, "Expected alloc"); + + /* + * Check for tight packing at specified alignment under simple + * conditions. + */ + { + const size_t alignments[] = { + 1, + QUANTUM, + QUANTUM << 1, + CACHELINE, + CACHELINE << 1, + }; + unsigned i; + + for (i = 0; i < sizeof(alignments) / sizeof(size_t); i++) { + size_t alignment = alignments[i]; + size_t align_ceil = ALIGNMENT_CEILING(alignment, + QUANTUM); + p = base_alloc(tsdn, base, 1, alignment); + assert_ptr_not_null(p, + "Unexpected base_alloc() failure"); + assert_ptr_eq(p, + (void *)(ALIGNMENT_CEILING((uintptr_t)p, + alignment)), "Expected quantum alignment"); + q = base_alloc(tsdn, base, alignment, alignment); + assert_ptr_not_null(q, + "Unexpected base_alloc() failure"); + assert_ptr_eq((void *)((uintptr_t)p + align_ceil), q, + "Minimal allocation should take up %zu bytes", + align_ceil); + r = base_alloc(tsdn, base, 1, alignment); + assert_ptr_not_null(r, + "Unexpected base_alloc() failure"); + assert_ptr_eq((void *)((uintptr_t)q + align_ceil), r, + "Minimal allocation should take up %zu bytes", + align_ceil); + } + } + + /* + * Allocate an object that cannot fit in the first block, then verify + * that the first block's remaining space is considered for subsequent + * allocation. + */ + assert_zu_ge(extent_bsize_get(&base->blocks->extent), QUANTUM, + "Remainder insufficient for test"); + /* Use up all but one quantum of block. */ + while (extent_bsize_get(&base->blocks->extent) > QUANTUM) { + p = base_alloc(tsdn, base, QUANTUM, QUANTUM); + assert_ptr_not_null(p, "Unexpected base_alloc() failure"); + } + r_exp = extent_addr_get(&base->blocks->extent); + assert_zu_eq(base->extent_sn_next, 1, "One extant block expected"); + q = base_alloc(tsdn, base, QUANTUM + 1, QUANTUM); + assert_ptr_not_null(q, "Unexpected base_alloc() failure"); + assert_ptr_ne(q, r_exp, "Expected allocation from new block"); + assert_zu_eq(base->extent_sn_next, 2, "Two extant blocks expected"); + r = base_alloc(tsdn, base, QUANTUM, QUANTUM); + assert_ptr_not_null(r, "Unexpected base_alloc() failure"); + assert_ptr_eq(r, r_exp, "Expected allocation from first block"); + assert_zu_eq(base->extent_sn_next, 2, "Two extant blocks expected"); + + /* + * Check for proper alignment support when normal blocks are too small. + */ + { + const size_t alignments[] = { + HUGEPAGE, + HUGEPAGE << 1 + }; + unsigned i; + + for (i = 0; i < sizeof(alignments) / sizeof(size_t); i++) { + size_t alignment = alignments[i]; + p = base_alloc(tsdn, base, QUANTUM, alignment); + assert_ptr_not_null(p, + "Unexpected base_alloc() failure"); + assert_ptr_eq(p, + (void *)(ALIGNMENT_CEILING((uintptr_t)p, + alignment)), "Expected %zu-byte alignment", + alignment); + } + } + + called_dalloc = called_destroy = called_decommit = called_purge_lazy = + called_purge_forced = false; + base_delete(tsdn, base); + assert_true(called_dalloc, "Expected dalloc call"); + assert_true(!called_destroy, "Unexpected destroy call"); + assert_true(called_decommit, "Expected decommit call"); + assert_true(called_purge_lazy, "Expected purge_lazy call"); + assert_true(called_purge_forced, "Expected purge_forced call"); + + try_dalloc = true; + try_destroy = true; + try_decommit = true; + try_purge_lazy = true; + try_purge_forced = true; + memcpy(&hooks, &hooks_orig, sizeof(extent_hooks_t)); +} +TEST_END + +int +main(void) { + return test( + test_base_hooks_default, + test_base_hooks_null, + test_base_hooks_not_null); +} diff --git a/deps/jemalloc/test/unit/binshard.c b/deps/jemalloc/test/unit/binshard.c new file mode 100644 index 000000000..d7a8df8fc --- /dev/null +++ b/deps/jemalloc/test/unit/binshard.c @@ -0,0 +1,154 @@ +#include "test/jemalloc_test.h" + +/* Config -- "narenas:1,bin_shards:1-160:16|129-512:4|256-256:8" */ + +#define NTHREADS 16 +#define REMOTE_NALLOC 256 + +static void * +thd_producer(void *varg) { + void **mem = varg; + unsigned arena, i; + size_t sz; + + sz = sizeof(arena); + /* Remote arena. */ + assert_d_eq(mallctl("arenas.create", (void *)&arena, &sz, NULL, 0), 0, + "Unexpected mallctl() failure"); + for (i = 0; i < REMOTE_NALLOC / 2; i++) { + mem[i] = mallocx(1, MALLOCX_TCACHE_NONE | MALLOCX_ARENA(arena)); + } + + /* Remote bin. */ + for (; i < REMOTE_NALLOC; i++) { + mem[i] = mallocx(1, MALLOCX_TCACHE_NONE | MALLOCX_ARENA(0)); + } + + return NULL; +} + +TEST_BEGIN(test_producer_consumer) { + thd_t thds[NTHREADS]; + void *mem[NTHREADS][REMOTE_NALLOC]; + unsigned i; + + /* Create producer threads to allocate. */ + for (i = 0; i < NTHREADS; i++) { + thd_create(&thds[i], thd_producer, mem[i]); + } + for (i = 0; i < NTHREADS; i++) { + thd_join(thds[i], NULL); + } + /* Remote deallocation by the current thread. */ + for (i = 0; i < NTHREADS; i++) { + for (unsigned j = 0; j < REMOTE_NALLOC; j++) { + assert_ptr_not_null(mem[i][j], + "Unexpected remote allocation failure"); + dallocx(mem[i][j], 0); + } + } +} +TEST_END + +static void * +thd_start(void *varg) { + void *ptr, *ptr2; + extent_t *extent; + unsigned shard1, shard2; + + tsdn_t *tsdn = tsdn_fetch(); + /* Try triggering allocations from sharded bins. */ + for (unsigned i = 0; i < 1024; i++) { + ptr = mallocx(1, MALLOCX_TCACHE_NONE); + ptr2 = mallocx(129, MALLOCX_TCACHE_NONE); + + extent = iealloc(tsdn, ptr); + shard1 = extent_binshard_get(extent); + dallocx(ptr, 0); + assert_u_lt(shard1, 16, "Unexpected bin shard used"); + + extent = iealloc(tsdn, ptr2); + shard2 = extent_binshard_get(extent); + dallocx(ptr2, 0); + assert_u_lt(shard2, 4, "Unexpected bin shard used"); + + if (shard1 > 0 || shard2 > 0) { + /* Triggered sharded bin usage. */ + return (void *)(uintptr_t)shard1; + } + } + + return NULL; +} + +TEST_BEGIN(test_bin_shard_mt) { + test_skip_if(have_percpu_arena && + PERCPU_ARENA_ENABLED(opt_percpu_arena)); + + thd_t thds[NTHREADS]; + unsigned i; + for (i = 0; i < NTHREADS; i++) { + thd_create(&thds[i], thd_start, NULL); + } + bool sharded = false; + for (i = 0; i < NTHREADS; i++) { + void *ret; + thd_join(thds[i], &ret); + if (ret != NULL) { + sharded = true; + } + } + assert_b_eq(sharded, true, "Did not find sharded bins"); +} +TEST_END + +TEST_BEGIN(test_bin_shard) { + unsigned nbins, i; + size_t mib[4], mib2[4]; + size_t miblen, miblen2, len; + + len = sizeof(nbins); + assert_d_eq(mallctl("arenas.nbins", (void *)&nbins, &len, NULL, 0), 0, + "Unexpected mallctl() failure"); + + miblen = 4; + assert_d_eq(mallctlnametomib("arenas.bin.0.nshards", mib, &miblen), 0, + "Unexpected mallctlnametomib() failure"); + miblen2 = 4; + assert_d_eq(mallctlnametomib("arenas.bin.0.size", mib2, &miblen2), 0, + "Unexpected mallctlnametomib() failure"); + + for (i = 0; i < nbins; i++) { + uint32_t nshards; + size_t size, sz1, sz2; + + mib[2] = i; + sz1 = sizeof(nshards); + assert_d_eq(mallctlbymib(mib, miblen, (void *)&nshards, &sz1, + NULL, 0), 0, "Unexpected mallctlbymib() failure"); + + mib2[2] = i; + sz2 = sizeof(size); + assert_d_eq(mallctlbymib(mib2, miblen2, (void *)&size, &sz2, + NULL, 0), 0, "Unexpected mallctlbymib() failure"); + + if (size >= 1 && size <= 128) { + assert_u_eq(nshards, 16, "Unexpected nshards"); + } else if (size == 256) { + assert_u_eq(nshards, 8, "Unexpected nshards"); + } else if (size > 128 && size <= 512) { + assert_u_eq(nshards, 4, "Unexpected nshards"); + } else { + assert_u_eq(nshards, 1, "Unexpected nshards"); + } + } +} +TEST_END + +int +main(void) { + return test_no_reentrancy( + test_bin_shard, + test_bin_shard_mt, + test_producer_consumer); +} diff --git a/deps/jemalloc/test/unit/binshard.sh b/deps/jemalloc/test/unit/binshard.sh new file mode 100644 index 000000000..c1d58c883 --- /dev/null +++ b/deps/jemalloc/test/unit/binshard.sh @@ -0,0 +1,3 @@ +#!/bin/sh + +export MALLOC_CONF="narenas:1,bin_shards:1-160:16|129-512:4|256-256:8" diff --git a/deps/jemalloc/test/unit/bit_util.c b/deps/jemalloc/test/unit/bit_util.c new file mode 100644 index 000000000..b747deb43 --- /dev/null +++ b/deps/jemalloc/test/unit/bit_util.c @@ -0,0 +1,111 @@ +#include "test/jemalloc_test.h" + +#include "jemalloc/internal/bit_util.h" + +#define TEST_POW2_CEIL(t, suf, pri) do { \ + unsigned i, pow2; \ + t x; \ + \ + assert_##suf##_eq(pow2_ceil_##suf(0), 0, "Unexpected result"); \ + \ + for (i = 0; i < sizeof(t) * 8; i++) { \ + assert_##suf##_eq(pow2_ceil_##suf(((t)1) << i), ((t)1) \ + << i, "Unexpected result"); \ + } \ + \ + for (i = 2; i < sizeof(t) * 8; i++) { \ + assert_##suf##_eq(pow2_ceil_##suf((((t)1) << i) - 1), \ + ((t)1) << i, "Unexpected result"); \ + } \ + \ + for (i = 0; i < sizeof(t) * 8 - 1; i++) { \ + assert_##suf##_eq(pow2_ceil_##suf((((t)1) << i) + 1), \ + ((t)1) << (i+1), "Unexpected result"); \ + } \ + \ + for (pow2 = 1; pow2 < 25; pow2++) { \ + for (x = (((t)1) << (pow2-1)) + 1; x <= ((t)1) << pow2; \ + x++) { \ + assert_##suf##_eq(pow2_ceil_##suf(x), \ + ((t)1) << pow2, \ + "Unexpected result, x=%"pri, x); \ + } \ + } \ +} while (0) + +TEST_BEGIN(test_pow2_ceil_u64) { + TEST_POW2_CEIL(uint64_t, u64, FMTu64); +} +TEST_END + +TEST_BEGIN(test_pow2_ceil_u32) { + TEST_POW2_CEIL(uint32_t, u32, FMTu32); +} +TEST_END + +TEST_BEGIN(test_pow2_ceil_zu) { + TEST_POW2_CEIL(size_t, zu, "zu"); +} +TEST_END + +void +assert_lg_ceil_range(size_t input, unsigned answer) { + if (input == 1) { + assert_u_eq(0, answer, "Got %u as lg_ceil of 1", answer); + return; + } + assert_zu_le(input, (ZU(1) << answer), + "Got %u as lg_ceil of %zu", answer, input); + assert_zu_gt(input, (ZU(1) << (answer - 1)), + "Got %u as lg_ceil of %zu", answer, input); +} + +void +assert_lg_floor_range(size_t input, unsigned answer) { + if (input == 1) { + assert_u_eq(0, answer, "Got %u as lg_floor of 1", answer); + return; + } + assert_zu_ge(input, (ZU(1) << answer), + "Got %u as lg_floor of %zu", answer, input); + assert_zu_lt(input, (ZU(1) << (answer + 1)), + "Got %u as lg_floor of %zu", answer, input); +} + +TEST_BEGIN(test_lg_ceil_floor) { + for (size_t i = 1; i < 10 * 1000 * 1000; i++) { + assert_lg_ceil_range(i, lg_ceil(i)); + assert_lg_ceil_range(i, LG_CEIL(i)); + assert_lg_floor_range(i, lg_floor(i)); + assert_lg_floor_range(i, LG_FLOOR(i)); + } + for (int i = 10; i < 8 * (1 << LG_SIZEOF_PTR) - 5; i++) { + for (size_t j = 0; j < (1 << 4); j++) { + size_t num1 = ((size_t)1 << i) + - j * ((size_t)1 << (i - 4)); + size_t num2 = ((size_t)1 << i) + + j * ((size_t)1 << (i - 4)); + assert_zu_ne(num1, 0, "Invalid lg argument"); + assert_zu_ne(num2, 0, "Invalid lg argument"); + assert_lg_ceil_range(num1, lg_ceil(num1)); + assert_lg_ceil_range(num1, LG_CEIL(num1)); + assert_lg_ceil_range(num2, lg_ceil(num2)); + assert_lg_ceil_range(num2, LG_CEIL(num2)); + + assert_lg_floor_range(num1, lg_floor(num1)); + assert_lg_floor_range(num1, LG_FLOOR(num1)); + assert_lg_floor_range(num2, lg_floor(num2)); + assert_lg_floor_range(num2, LG_FLOOR(num2)); + } + } +} +TEST_END + +int +main(void) { + return test( + test_pow2_ceil_u64, + test_pow2_ceil_u32, + test_pow2_ceil_zu, + test_lg_ceil_floor); +} diff --git a/deps/jemalloc/test/unit/bitmap.c b/deps/jemalloc/test/unit/bitmap.c new file mode 100644 index 000000000..cafb2039e --- /dev/null +++ b/deps/jemalloc/test/unit/bitmap.c @@ -0,0 +1,431 @@ +#include "test/jemalloc_test.h" + +#define NBITS_TAB \ + NB( 1) \ + NB( 2) \ + NB( 3) \ + NB( 4) \ + NB( 5) \ + NB( 6) \ + NB( 7) \ + NB( 8) \ + NB( 9) \ + NB(10) \ + NB(11) \ + NB(12) \ + NB(13) \ + NB(14) \ + NB(15) \ + NB(16) \ + NB(17) \ + NB(18) \ + NB(19) \ + NB(20) \ + NB(21) \ + NB(22) \ + NB(23) \ + NB(24) \ + NB(25) \ + NB(26) \ + NB(27) \ + NB(28) \ + NB(29) \ + NB(30) \ + NB(31) \ + NB(32) \ + \ + NB(33) \ + NB(34) \ + NB(35) \ + NB(36) \ + NB(37) \ + NB(38) \ + NB(39) \ + NB(40) \ + NB(41) \ + NB(42) \ + NB(43) \ + NB(44) \ + NB(45) \ + NB(46) \ + NB(47) \ + NB(48) \ + NB(49) \ + NB(50) \ + NB(51) \ + NB(52) \ + NB(53) \ + NB(54) \ + NB(55) \ + NB(56) \ + NB(57) \ + NB(58) \ + NB(59) \ + NB(60) \ + NB(61) \ + NB(62) \ + NB(63) \ + NB(64) \ + NB(65) \ + \ + NB(126) \ + NB(127) \ + NB(128) \ + NB(129) \ + NB(130) \ + \ + NB(254) \ + NB(255) \ + NB(256) \ + NB(257) \ + NB(258) \ + \ + NB(510) \ + NB(511) \ + NB(512) \ + NB(513) \ + NB(514) \ + \ + NB(1024) \ + NB(2048) \ + NB(4096) \ + NB(8192) \ + NB(16384) \ + +static void +test_bitmap_initializer_body(const bitmap_info_t *binfo, size_t nbits) { + bitmap_info_t binfo_dyn; + bitmap_info_init(&binfo_dyn, nbits); + + assert_zu_eq(bitmap_size(binfo), bitmap_size(&binfo_dyn), + "Unexpected difference between static and dynamic initialization, " + "nbits=%zu", nbits); + assert_zu_eq(binfo->nbits, binfo_dyn.nbits, + "Unexpected difference between static and dynamic initialization, " + "nbits=%zu", nbits); +#ifdef BITMAP_USE_TREE + assert_u_eq(binfo->nlevels, binfo_dyn.nlevels, + "Unexpected difference between static and dynamic initialization, " + "nbits=%zu", nbits); + { + unsigned i; + + for (i = 0; i < binfo->nlevels; i++) { + assert_zu_eq(binfo->levels[i].group_offset, + binfo_dyn.levels[i].group_offset, + "Unexpected difference between static and dynamic " + "initialization, nbits=%zu, level=%u", nbits, i); + } + } +#else + assert_zu_eq(binfo->ngroups, binfo_dyn.ngroups, + "Unexpected difference between static and dynamic initialization"); +#endif +} + +TEST_BEGIN(test_bitmap_initializer) { +#define NB(nbits) { \ + if (nbits <= BITMAP_MAXBITS) { \ + bitmap_info_t binfo = \ + BITMAP_INFO_INITIALIZER(nbits); \ + test_bitmap_initializer_body(&binfo, nbits); \ + } \ + } + NBITS_TAB +#undef NB +} +TEST_END + +static size_t +test_bitmap_size_body(const bitmap_info_t *binfo, size_t nbits, + size_t prev_size) { + size_t size = bitmap_size(binfo); + assert_zu_ge(size, (nbits >> 3), + "Bitmap size is smaller than expected"); + assert_zu_ge(size, prev_size, "Bitmap size is smaller than expected"); + return size; +} + +TEST_BEGIN(test_bitmap_size) { + size_t nbits, prev_size; + + prev_size = 0; + for (nbits = 1; nbits <= BITMAP_MAXBITS; nbits++) { + bitmap_info_t binfo; + bitmap_info_init(&binfo, nbits); + prev_size = test_bitmap_size_body(&binfo, nbits, prev_size); + } +#define NB(nbits) { \ + bitmap_info_t binfo = BITMAP_INFO_INITIALIZER(nbits); \ + prev_size = test_bitmap_size_body(&binfo, nbits, \ + prev_size); \ + } + prev_size = 0; + NBITS_TAB +#undef NB +} +TEST_END + +static void +test_bitmap_init_body(const bitmap_info_t *binfo, size_t nbits) { + size_t i; + bitmap_t *bitmap = (bitmap_t *)malloc(bitmap_size(binfo)); + assert_ptr_not_null(bitmap, "Unexpected malloc() failure"); + + bitmap_init(bitmap, binfo, false); + for (i = 0; i < nbits; i++) { + assert_false(bitmap_get(bitmap, binfo, i), + "Bit should be unset"); + } + + bitmap_init(bitmap, binfo, true); + for (i = 0; i < nbits; i++) { + assert_true(bitmap_get(bitmap, binfo, i), "Bit should be set"); + } + + free(bitmap); +} + +TEST_BEGIN(test_bitmap_init) { + size_t nbits; + + for (nbits = 1; nbits <= BITMAP_MAXBITS; nbits++) { + bitmap_info_t binfo; + bitmap_info_init(&binfo, nbits); + test_bitmap_init_body(&binfo, nbits); + } +#define NB(nbits) { \ + bitmap_info_t binfo = BITMAP_INFO_INITIALIZER(nbits); \ + test_bitmap_init_body(&binfo, nbits); \ + } + NBITS_TAB +#undef NB +} +TEST_END + +static void +test_bitmap_set_body(const bitmap_info_t *binfo, size_t nbits) { + size_t i; + bitmap_t *bitmap = (bitmap_t *)malloc(bitmap_size(binfo)); + assert_ptr_not_null(bitmap, "Unexpected malloc() failure"); + bitmap_init(bitmap, binfo, false); + + for (i = 0; i < nbits; i++) { + bitmap_set(bitmap, binfo, i); + } + assert_true(bitmap_full(bitmap, binfo), "All bits should be set"); + free(bitmap); +} + +TEST_BEGIN(test_bitmap_set) { + size_t nbits; + + for (nbits = 1; nbits <= BITMAP_MAXBITS; nbits++) { + bitmap_info_t binfo; + bitmap_info_init(&binfo, nbits); + test_bitmap_set_body(&binfo, nbits); + } +#define NB(nbits) { \ + bitmap_info_t binfo = BITMAP_INFO_INITIALIZER(nbits); \ + test_bitmap_set_body(&binfo, nbits); \ + } + NBITS_TAB +#undef NB +} +TEST_END + +static void +test_bitmap_unset_body(const bitmap_info_t *binfo, size_t nbits) { + size_t i; + bitmap_t *bitmap = (bitmap_t *)malloc(bitmap_size(binfo)); + assert_ptr_not_null(bitmap, "Unexpected malloc() failure"); + bitmap_init(bitmap, binfo, false); + + for (i = 0; i < nbits; i++) { + bitmap_set(bitmap, binfo, i); + } + assert_true(bitmap_full(bitmap, binfo), "All bits should be set"); + for (i = 0; i < nbits; i++) { + bitmap_unset(bitmap, binfo, i); + } + for (i = 0; i < nbits; i++) { + bitmap_set(bitmap, binfo, i); + } + assert_true(bitmap_full(bitmap, binfo), "All bits should be set"); + free(bitmap); +} + +TEST_BEGIN(test_bitmap_unset) { + size_t nbits; + + for (nbits = 1; nbits <= BITMAP_MAXBITS; nbits++) { + bitmap_info_t binfo; + bitmap_info_init(&binfo, nbits); + test_bitmap_unset_body(&binfo, nbits); + } +#define NB(nbits) { \ + bitmap_info_t binfo = BITMAP_INFO_INITIALIZER(nbits); \ + test_bitmap_unset_body(&binfo, nbits); \ + } + NBITS_TAB +#undef NB +} +TEST_END + +static void +test_bitmap_xfu_body(const bitmap_info_t *binfo, size_t nbits) { + bitmap_t *bitmap = (bitmap_t *)malloc(bitmap_size(binfo)); + assert_ptr_not_null(bitmap, "Unexpected malloc() failure"); + bitmap_init(bitmap, binfo, false); + + /* Iteratively set bits starting at the beginning. */ + for (size_t i = 0; i < nbits; i++) { + assert_zu_eq(bitmap_ffu(bitmap, binfo, 0), i, + "First unset bit should be just after previous first unset " + "bit"); + assert_zu_eq(bitmap_ffu(bitmap, binfo, (i > 0) ? i-1 : i), i, + "First unset bit should be just after previous first unset " + "bit"); + assert_zu_eq(bitmap_ffu(bitmap, binfo, i), i, + "First unset bit should be just after previous first unset " + "bit"); + assert_zu_eq(bitmap_sfu(bitmap, binfo), i, + "First unset bit should be just after previous first unset " + "bit"); + } + assert_true(bitmap_full(bitmap, binfo), "All bits should be set"); + + /* + * Iteratively unset bits starting at the end, and verify that + * bitmap_sfu() reaches the unset bits. + */ + for (size_t i = nbits - 1; i < nbits; i--) { /* (nbits..0] */ + bitmap_unset(bitmap, binfo, i); + assert_zu_eq(bitmap_ffu(bitmap, binfo, 0), i, + "First unset bit should the bit previously unset"); + assert_zu_eq(bitmap_ffu(bitmap, binfo, (i > 0) ? i-1 : i), i, + "First unset bit should the bit previously unset"); + assert_zu_eq(bitmap_ffu(bitmap, binfo, i), i, + "First unset bit should the bit previously unset"); + assert_zu_eq(bitmap_sfu(bitmap, binfo), i, + "First unset bit should the bit previously unset"); + bitmap_unset(bitmap, binfo, i); + } + assert_false(bitmap_get(bitmap, binfo, 0), "Bit should be unset"); + + /* + * Iteratively set bits starting at the beginning, and verify that + * bitmap_sfu() looks past them. + */ + for (size_t i = 1; i < nbits; i++) { + bitmap_set(bitmap, binfo, i - 1); + assert_zu_eq(bitmap_ffu(bitmap, binfo, 0), i, + "First unset bit should be just after the bit previously " + "set"); + assert_zu_eq(bitmap_ffu(bitmap, binfo, (i > 0) ? i-1 : i), i, + "First unset bit should be just after the bit previously " + "set"); + assert_zu_eq(bitmap_ffu(bitmap, binfo, i), i, + "First unset bit should be just after the bit previously " + "set"); + assert_zu_eq(bitmap_sfu(bitmap, binfo), i, + "First unset bit should be just after the bit previously " + "set"); + bitmap_unset(bitmap, binfo, i); + } + assert_zu_eq(bitmap_ffu(bitmap, binfo, 0), nbits - 1, + "First unset bit should be the last bit"); + assert_zu_eq(bitmap_ffu(bitmap, binfo, (nbits > 1) ? nbits-2 : nbits-1), + nbits - 1, "First unset bit should be the last bit"); + assert_zu_eq(bitmap_ffu(bitmap, binfo, nbits - 1), nbits - 1, + "First unset bit should be the last bit"); + assert_zu_eq(bitmap_sfu(bitmap, binfo), nbits - 1, + "First unset bit should be the last bit"); + assert_true(bitmap_full(bitmap, binfo), "All bits should be set"); + + /* + * Bubble a "usu" pattern through the bitmap and verify that + * bitmap_ffu() finds the correct bit for all five min_bit cases. + */ + if (nbits >= 3) { + for (size_t i = 0; i < nbits-2; i++) { + bitmap_unset(bitmap, binfo, i); + bitmap_unset(bitmap, binfo, i+2); + if (i > 0) { + assert_zu_eq(bitmap_ffu(bitmap, binfo, i-1), i, + "Unexpected first unset bit"); + } + assert_zu_eq(bitmap_ffu(bitmap, binfo, i), i, + "Unexpected first unset bit"); + assert_zu_eq(bitmap_ffu(bitmap, binfo, i+1), i+2, + "Unexpected first unset bit"); + assert_zu_eq(bitmap_ffu(bitmap, binfo, i+2), i+2, + "Unexpected first unset bit"); + if (i + 3 < nbits) { + assert_zu_eq(bitmap_ffu(bitmap, binfo, i+3), + nbits, "Unexpected first unset bit"); + } + assert_zu_eq(bitmap_sfu(bitmap, binfo), i, + "Unexpected first unset bit"); + assert_zu_eq(bitmap_sfu(bitmap, binfo), i+2, + "Unexpected first unset bit"); + } + } + + /* + * Unset the last bit, bubble another unset bit through the bitmap, and + * verify that bitmap_ffu() finds the correct bit for all four min_bit + * cases. + */ + if (nbits >= 3) { + bitmap_unset(bitmap, binfo, nbits-1); + for (size_t i = 0; i < nbits-1; i++) { + bitmap_unset(bitmap, binfo, i); + if (i > 0) { + assert_zu_eq(bitmap_ffu(bitmap, binfo, i-1), i, + "Unexpected first unset bit"); + } + assert_zu_eq(bitmap_ffu(bitmap, binfo, i), i, + "Unexpected first unset bit"); + assert_zu_eq(bitmap_ffu(bitmap, binfo, i+1), nbits-1, + "Unexpected first unset bit"); + assert_zu_eq(bitmap_ffu(bitmap, binfo, nbits-1), + nbits-1, "Unexpected first unset bit"); + + assert_zu_eq(bitmap_sfu(bitmap, binfo), i, + "Unexpected first unset bit"); + } + assert_zu_eq(bitmap_sfu(bitmap, binfo), nbits-1, + "Unexpected first unset bit"); + } + + free(bitmap); +} + +TEST_BEGIN(test_bitmap_xfu) { + size_t nbits; + + for (nbits = 1; nbits <= BITMAP_MAXBITS; nbits++) { + bitmap_info_t binfo; + bitmap_info_init(&binfo, nbits); + test_bitmap_xfu_body(&binfo, nbits); + } +#define NB(nbits) { \ + bitmap_info_t binfo = BITMAP_INFO_INITIALIZER(nbits); \ + test_bitmap_xfu_body(&binfo, nbits); \ + } + NBITS_TAB +#undef NB +} +TEST_END + +int +main(void) { + return test( + test_bitmap_initializer, + test_bitmap_size, + test_bitmap_init, + test_bitmap_set, + test_bitmap_unset, + test_bitmap_xfu); +} diff --git a/deps/jemalloc/test/unit/ckh.c b/deps/jemalloc/test/unit/ckh.c new file mode 100644 index 000000000..707ea5f8c --- /dev/null +++ b/deps/jemalloc/test/unit/ckh.c @@ -0,0 +1,211 @@ +#include "test/jemalloc_test.h" + +TEST_BEGIN(test_new_delete) { + tsd_t *tsd; + ckh_t ckh; + + tsd = tsd_fetch(); + + assert_false(ckh_new(tsd, &ckh, 2, ckh_string_hash, + ckh_string_keycomp), "Unexpected ckh_new() error"); + ckh_delete(tsd, &ckh); + + assert_false(ckh_new(tsd, &ckh, 3, ckh_pointer_hash, + ckh_pointer_keycomp), "Unexpected ckh_new() error"); + ckh_delete(tsd, &ckh); +} +TEST_END + +TEST_BEGIN(test_count_insert_search_remove) { + tsd_t *tsd; + ckh_t ckh; + const char *strs[] = { + "a string", + "A string", + "a string.", + "A string." + }; + const char *missing = "A string not in the hash table."; + size_t i; + + tsd = tsd_fetch(); + + assert_false(ckh_new(tsd, &ckh, 2, ckh_string_hash, + ckh_string_keycomp), "Unexpected ckh_new() error"); + assert_zu_eq(ckh_count(&ckh), 0, + "ckh_count() should return %zu, but it returned %zu", ZU(0), + ckh_count(&ckh)); + + /* Insert. */ + for (i = 0; i < sizeof(strs)/sizeof(const char *); i++) { + ckh_insert(tsd, &ckh, strs[i], strs[i]); + assert_zu_eq(ckh_count(&ckh), i+1, + "ckh_count() should return %zu, but it returned %zu", i+1, + ckh_count(&ckh)); + } + + /* Search. */ + for (i = 0; i < sizeof(strs)/sizeof(const char *); i++) { + union { + void *p; + const char *s; + } k, v; + void **kp, **vp; + const char *ks, *vs; + + kp = (i & 1) ? &k.p : NULL; + vp = (i & 2) ? &v.p : NULL; + k.p = NULL; + v.p = NULL; + assert_false(ckh_search(&ckh, strs[i], kp, vp), + "Unexpected ckh_search() error"); + + ks = (i & 1) ? strs[i] : (const char *)NULL; + vs = (i & 2) ? strs[i] : (const char *)NULL; + assert_ptr_eq((void *)ks, (void *)k.s, "Key mismatch, i=%zu", + i); + assert_ptr_eq((void *)vs, (void *)v.s, "Value mismatch, i=%zu", + i); + } + assert_true(ckh_search(&ckh, missing, NULL, NULL), + "Unexpected ckh_search() success"); + + /* Remove. */ + for (i = 0; i < sizeof(strs)/sizeof(const char *); i++) { + union { + void *p; + const char *s; + } k, v; + void **kp, **vp; + const char *ks, *vs; + + kp = (i & 1) ? &k.p : NULL; + vp = (i & 2) ? &v.p : NULL; + k.p = NULL; + v.p = NULL; + assert_false(ckh_remove(tsd, &ckh, strs[i], kp, vp), + "Unexpected ckh_remove() error"); + + ks = (i & 1) ? strs[i] : (const char *)NULL; + vs = (i & 2) ? strs[i] : (const char *)NULL; + assert_ptr_eq((void *)ks, (void *)k.s, "Key mismatch, i=%zu", + i); + assert_ptr_eq((void *)vs, (void *)v.s, "Value mismatch, i=%zu", + i); + assert_zu_eq(ckh_count(&ckh), + sizeof(strs)/sizeof(const char *) - i - 1, + "ckh_count() should return %zu, but it returned %zu", + sizeof(strs)/sizeof(const char *) - i - 1, + ckh_count(&ckh)); + } + + ckh_delete(tsd, &ckh); +} +TEST_END + +TEST_BEGIN(test_insert_iter_remove) { +#define NITEMS ZU(1000) + tsd_t *tsd; + ckh_t ckh; + void **p[NITEMS]; + void *q, *r; + size_t i; + + tsd = tsd_fetch(); + + assert_false(ckh_new(tsd, &ckh, 2, ckh_pointer_hash, + ckh_pointer_keycomp), "Unexpected ckh_new() error"); + + for (i = 0; i < NITEMS; i++) { + p[i] = mallocx(i+1, 0); + assert_ptr_not_null(p[i], "Unexpected mallocx() failure"); + } + + for (i = 0; i < NITEMS; i++) { + size_t j; + + for (j = i; j < NITEMS; j++) { + assert_false(ckh_insert(tsd, &ckh, p[j], p[j]), + "Unexpected ckh_insert() failure"); + assert_false(ckh_search(&ckh, p[j], &q, &r), + "Unexpected ckh_search() failure"); + assert_ptr_eq(p[j], q, "Key pointer mismatch"); + assert_ptr_eq(p[j], r, "Value pointer mismatch"); + } + + assert_zu_eq(ckh_count(&ckh), NITEMS, + "ckh_count() should return %zu, but it returned %zu", + NITEMS, ckh_count(&ckh)); + + for (j = i + 1; j < NITEMS; j++) { + assert_false(ckh_search(&ckh, p[j], NULL, NULL), + "Unexpected ckh_search() failure"); + assert_false(ckh_remove(tsd, &ckh, p[j], &q, &r), + "Unexpected ckh_remove() failure"); + assert_ptr_eq(p[j], q, "Key pointer mismatch"); + assert_ptr_eq(p[j], r, "Value pointer mismatch"); + assert_true(ckh_search(&ckh, p[j], NULL, NULL), + "Unexpected ckh_search() success"); + assert_true(ckh_remove(tsd, &ckh, p[j], &q, &r), + "Unexpected ckh_remove() success"); + } + + { + bool seen[NITEMS]; + size_t tabind; + + memset(seen, 0, sizeof(seen)); + + for (tabind = 0; !ckh_iter(&ckh, &tabind, &q, &r);) { + size_t k; + + assert_ptr_eq(q, r, "Key and val not equal"); + + for (k = 0; k < NITEMS; k++) { + if (p[k] == q) { + assert_false(seen[k], + "Item %zu already seen", k); + seen[k] = true; + break; + } + } + } + + for (j = 0; j < i + 1; j++) { + assert_true(seen[j], "Item %zu not seen", j); + } + for (; j < NITEMS; j++) { + assert_false(seen[j], "Item %zu seen", j); + } + } + } + + for (i = 0; i < NITEMS; i++) { + assert_false(ckh_search(&ckh, p[i], NULL, NULL), + "Unexpected ckh_search() failure"); + assert_false(ckh_remove(tsd, &ckh, p[i], &q, &r), + "Unexpected ckh_remove() failure"); + assert_ptr_eq(p[i], q, "Key pointer mismatch"); + assert_ptr_eq(p[i], r, "Value pointer mismatch"); + assert_true(ckh_search(&ckh, p[i], NULL, NULL), + "Unexpected ckh_search() success"); + assert_true(ckh_remove(tsd, &ckh, p[i], &q, &r), + "Unexpected ckh_remove() success"); + dallocx(p[i], 0); + } + + assert_zu_eq(ckh_count(&ckh), 0, + "ckh_count() should return %zu, but it returned %zu", + ZU(0), ckh_count(&ckh)); + ckh_delete(tsd, &ckh); +#undef NITEMS +} +TEST_END + +int +main(void) { + return test( + test_new_delete, + test_count_insert_search_remove, + test_insert_iter_remove); +} diff --git a/deps/jemalloc/test/unit/decay.c b/deps/jemalloc/test/unit/decay.c new file mode 100644 index 000000000..cf3c07960 --- /dev/null +++ b/deps/jemalloc/test/unit/decay.c @@ -0,0 +1,605 @@ +#include "test/jemalloc_test.h" + +#include "jemalloc/internal/ticker.h" + +static nstime_monotonic_t *nstime_monotonic_orig; +static nstime_update_t *nstime_update_orig; + +static unsigned nupdates_mock; +static nstime_t time_mock; +static bool monotonic_mock; + +static bool +check_background_thread_enabled(void) { + bool enabled; + size_t sz = sizeof(bool); + int ret = mallctl("background_thread", (void *)&enabled, &sz, NULL,0); + if (ret == ENOENT) { + return false; + } + assert_d_eq(ret, 0, "Unexpected mallctl error"); + return enabled; +} + +static bool +nstime_monotonic_mock(void) { + return monotonic_mock; +} + +static bool +nstime_update_mock(nstime_t *time) { + nupdates_mock++; + if (monotonic_mock) { + nstime_copy(time, &time_mock); + } + return !monotonic_mock; +} + +static unsigned +do_arena_create(ssize_t dirty_decay_ms, ssize_t muzzy_decay_ms) { + unsigned arena_ind; + size_t sz = sizeof(unsigned); + assert_d_eq(mallctl("arenas.create", (void *)&arena_ind, &sz, NULL, 0), + 0, "Unexpected mallctl() failure"); + size_t mib[3]; + size_t miblen = sizeof(mib)/sizeof(size_t); + + assert_d_eq(mallctlnametomib("arena.0.dirty_decay_ms", mib, &miblen), + 0, "Unexpected mallctlnametomib() failure"); + mib[1] = (size_t)arena_ind; + assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, + (void *)&dirty_decay_ms, sizeof(dirty_decay_ms)), 0, + "Unexpected mallctlbymib() failure"); + + assert_d_eq(mallctlnametomib("arena.0.muzzy_decay_ms", mib, &miblen), + 0, "Unexpected mallctlnametomib() failure"); + mib[1] = (size_t)arena_ind; + assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, + (void *)&muzzy_decay_ms, sizeof(muzzy_decay_ms)), 0, + "Unexpected mallctlbymib() failure"); + + return arena_ind; +} + +static void +do_arena_destroy(unsigned arena_ind) { + size_t mib[3]; + size_t miblen = sizeof(mib)/sizeof(size_t); + assert_d_eq(mallctlnametomib("arena.0.destroy", mib, &miblen), 0, + "Unexpected mallctlnametomib() failure"); + mib[1] = (size_t)arena_ind; + assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, NULL, 0), 0, + "Unexpected mallctlbymib() failure"); +} + +void +do_epoch(void) { + uint64_t epoch = 1; + assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch, sizeof(epoch)), + 0, "Unexpected mallctl() failure"); +} + +void +do_purge(unsigned arena_ind) { + size_t mib[3]; + size_t miblen = sizeof(mib)/sizeof(size_t); + assert_d_eq(mallctlnametomib("arena.0.purge", mib, &miblen), 0, + "Unexpected mallctlnametomib() failure"); + mib[1] = (size_t)arena_ind; + assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, NULL, 0), 0, + "Unexpected mallctlbymib() failure"); +} + +void +do_decay(unsigned arena_ind) { + size_t mib[3]; + size_t miblen = sizeof(mib)/sizeof(size_t); + assert_d_eq(mallctlnametomib("arena.0.decay", mib, &miblen), 0, + "Unexpected mallctlnametomib() failure"); + mib[1] = (size_t)arena_ind; + assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, NULL, 0), 0, + "Unexpected mallctlbymib() failure"); +} + +static uint64_t +get_arena_npurge_impl(const char *mibname, unsigned arena_ind) { + size_t mib[4]; + size_t miblen = sizeof(mib)/sizeof(size_t); + assert_d_eq(mallctlnametomib(mibname, mib, &miblen), 0, + "Unexpected mallctlnametomib() failure"); + mib[2] = (size_t)arena_ind; + uint64_t npurge = 0; + size_t sz = sizeof(npurge); + assert_d_eq(mallctlbymib(mib, miblen, (void *)&npurge, &sz, NULL, 0), + config_stats ? 0 : ENOENT, "Unexpected mallctlbymib() failure"); + return npurge; +} + +static uint64_t +get_arena_dirty_npurge(unsigned arena_ind) { + do_epoch(); + return get_arena_npurge_impl("stats.arenas.0.dirty_npurge", arena_ind); +} + +static uint64_t +get_arena_dirty_purged(unsigned arena_ind) { + do_epoch(); + return get_arena_npurge_impl("stats.arenas.0.dirty_purged", arena_ind); +} + +static uint64_t +get_arena_muzzy_npurge(unsigned arena_ind) { + do_epoch(); + return get_arena_npurge_impl("stats.arenas.0.muzzy_npurge", arena_ind); +} + +static uint64_t +get_arena_npurge(unsigned arena_ind) { + do_epoch(); + return get_arena_npurge_impl("stats.arenas.0.dirty_npurge", arena_ind) + + get_arena_npurge_impl("stats.arenas.0.muzzy_npurge", arena_ind); +} + +static size_t +get_arena_pdirty(unsigned arena_ind) { + do_epoch(); + size_t mib[4]; + size_t miblen = sizeof(mib)/sizeof(size_t); + assert_d_eq(mallctlnametomib("stats.arenas.0.pdirty", mib, &miblen), 0, + "Unexpected mallctlnametomib() failure"); + mib[2] = (size_t)arena_ind; + size_t pdirty; + size_t sz = sizeof(pdirty); + assert_d_eq(mallctlbymib(mib, miblen, (void *)&pdirty, &sz, NULL, 0), 0, + "Unexpected mallctlbymib() failure"); + return pdirty; +} + +static size_t +get_arena_pmuzzy(unsigned arena_ind) { + do_epoch(); + size_t mib[4]; + size_t miblen = sizeof(mib)/sizeof(size_t); + assert_d_eq(mallctlnametomib("stats.arenas.0.pmuzzy", mib, &miblen), 0, + "Unexpected mallctlnametomib() failure"); + mib[2] = (size_t)arena_ind; + size_t pmuzzy; + size_t sz = sizeof(pmuzzy); + assert_d_eq(mallctlbymib(mib, miblen, (void *)&pmuzzy, &sz, NULL, 0), 0, + "Unexpected mallctlbymib() failure"); + return pmuzzy; +} + +static void * +do_mallocx(size_t size, int flags) { + void *p = mallocx(size, flags); + assert_ptr_not_null(p, "Unexpected mallocx() failure"); + return p; +} + +static void +generate_dirty(unsigned arena_ind, size_t size) { + int flags = MALLOCX_ARENA(arena_ind) | MALLOCX_TCACHE_NONE; + void *p = do_mallocx(size, flags); + dallocx(p, flags); +} + +TEST_BEGIN(test_decay_ticks) { + test_skip_if(check_background_thread_enabled()); + + ticker_t *decay_ticker; + unsigned tick0, tick1, arena_ind; + size_t sz, large0; + void *p; + + sz = sizeof(size_t); + assert_d_eq(mallctl("arenas.lextent.0.size", (void *)&large0, &sz, NULL, + 0), 0, "Unexpected mallctl failure"); + + /* Set up a manually managed arena for test. */ + arena_ind = do_arena_create(0, 0); + + /* Migrate to the new arena, and get the ticker. */ + unsigned old_arena_ind; + size_t sz_arena_ind = sizeof(old_arena_ind); + assert_d_eq(mallctl("thread.arena", (void *)&old_arena_ind, + &sz_arena_ind, (void *)&arena_ind, sizeof(arena_ind)), 0, + "Unexpected mallctl() failure"); + decay_ticker = decay_ticker_get(tsd_fetch(), arena_ind); + assert_ptr_not_null(decay_ticker, + "Unexpected failure getting decay ticker"); + + /* + * Test the standard APIs using a large size class, since we can't + * control tcache interactions for small size classes (except by + * completely disabling tcache for the entire test program). + */ + + /* malloc(). */ + tick0 = ticker_read(decay_ticker); + p = malloc(large0); + assert_ptr_not_null(p, "Unexpected malloc() failure"); + tick1 = ticker_read(decay_ticker); + assert_u32_ne(tick1, tick0, "Expected ticker to tick during malloc()"); + /* free(). */ + tick0 = ticker_read(decay_ticker); + free(p); + tick1 = ticker_read(decay_ticker); + assert_u32_ne(tick1, tick0, "Expected ticker to tick during free()"); + + /* calloc(). */ + tick0 = ticker_read(decay_ticker); + p = calloc(1, large0); + assert_ptr_not_null(p, "Unexpected calloc() failure"); + tick1 = ticker_read(decay_ticker); + assert_u32_ne(tick1, tick0, "Expected ticker to tick during calloc()"); + free(p); + + /* posix_memalign(). */ + tick0 = ticker_read(decay_ticker); + assert_d_eq(posix_memalign(&p, sizeof(size_t), large0), 0, + "Unexpected posix_memalign() failure"); + tick1 = ticker_read(decay_ticker); + assert_u32_ne(tick1, tick0, + "Expected ticker to tick during posix_memalign()"); + free(p); + + /* aligned_alloc(). */ + tick0 = ticker_read(decay_ticker); + p = aligned_alloc(sizeof(size_t), large0); + assert_ptr_not_null(p, "Unexpected aligned_alloc() failure"); + tick1 = ticker_read(decay_ticker); + assert_u32_ne(tick1, tick0, + "Expected ticker to tick during aligned_alloc()"); + free(p); + + /* realloc(). */ + /* Allocate. */ + tick0 = ticker_read(decay_ticker); + p = realloc(NULL, large0); + assert_ptr_not_null(p, "Unexpected realloc() failure"); + tick1 = ticker_read(decay_ticker); + assert_u32_ne(tick1, tick0, "Expected ticker to tick during realloc()"); + /* Reallocate. */ + tick0 = ticker_read(decay_ticker); + p = realloc(p, large0); + assert_ptr_not_null(p, "Unexpected realloc() failure"); + tick1 = ticker_read(decay_ticker); + assert_u32_ne(tick1, tick0, "Expected ticker to tick during realloc()"); + /* Deallocate. */ + tick0 = ticker_read(decay_ticker); + realloc(p, 0); + tick1 = ticker_read(decay_ticker); + assert_u32_ne(tick1, tick0, "Expected ticker to tick during realloc()"); + + /* + * Test the *allocx() APIs using large and small size classes, with + * tcache explicitly disabled. + */ + { + unsigned i; + size_t allocx_sizes[2]; + allocx_sizes[0] = large0; + allocx_sizes[1] = 1; + + for (i = 0; i < sizeof(allocx_sizes) / sizeof(size_t); i++) { + sz = allocx_sizes[i]; + + /* mallocx(). */ + tick0 = ticker_read(decay_ticker); + p = mallocx(sz, MALLOCX_TCACHE_NONE); + assert_ptr_not_null(p, "Unexpected mallocx() failure"); + tick1 = ticker_read(decay_ticker); + assert_u32_ne(tick1, tick0, + "Expected ticker to tick during mallocx() (sz=%zu)", + sz); + /* rallocx(). */ + tick0 = ticker_read(decay_ticker); + p = rallocx(p, sz, MALLOCX_TCACHE_NONE); + assert_ptr_not_null(p, "Unexpected rallocx() failure"); + tick1 = ticker_read(decay_ticker); + assert_u32_ne(tick1, tick0, + "Expected ticker to tick during rallocx() (sz=%zu)", + sz); + /* xallocx(). */ + tick0 = ticker_read(decay_ticker); + xallocx(p, sz, 0, MALLOCX_TCACHE_NONE); + tick1 = ticker_read(decay_ticker); + assert_u32_ne(tick1, tick0, + "Expected ticker to tick during xallocx() (sz=%zu)", + sz); + /* dallocx(). */ + tick0 = ticker_read(decay_ticker); + dallocx(p, MALLOCX_TCACHE_NONE); + tick1 = ticker_read(decay_ticker); + assert_u32_ne(tick1, tick0, + "Expected ticker to tick during dallocx() (sz=%zu)", + sz); + /* sdallocx(). */ + p = mallocx(sz, MALLOCX_TCACHE_NONE); + assert_ptr_not_null(p, "Unexpected mallocx() failure"); + tick0 = ticker_read(decay_ticker); + sdallocx(p, sz, MALLOCX_TCACHE_NONE); + tick1 = ticker_read(decay_ticker); + assert_u32_ne(tick1, tick0, + "Expected ticker to tick during sdallocx() " + "(sz=%zu)", sz); + } + } + + /* + * Test tcache fill/flush interactions for large and small size classes, + * using an explicit tcache. + */ + unsigned tcache_ind, i; + size_t tcache_sizes[2]; + tcache_sizes[0] = large0; + tcache_sizes[1] = 1; + + size_t tcache_max, sz_tcache_max; + sz_tcache_max = sizeof(tcache_max); + assert_d_eq(mallctl("arenas.tcache_max", (void *)&tcache_max, + &sz_tcache_max, NULL, 0), 0, "Unexpected mallctl() failure"); + + sz = sizeof(unsigned); + assert_d_eq(mallctl("tcache.create", (void *)&tcache_ind, &sz, + NULL, 0), 0, "Unexpected mallctl failure"); + + for (i = 0; i < sizeof(tcache_sizes) / sizeof(size_t); i++) { + sz = tcache_sizes[i]; + + /* tcache fill. */ + tick0 = ticker_read(decay_ticker); + p = mallocx(sz, MALLOCX_TCACHE(tcache_ind)); + assert_ptr_not_null(p, "Unexpected mallocx() failure"); + tick1 = ticker_read(decay_ticker); + assert_u32_ne(tick1, tick0, + "Expected ticker to tick during tcache fill " + "(sz=%zu)", sz); + /* tcache flush. */ + dallocx(p, MALLOCX_TCACHE(tcache_ind)); + tick0 = ticker_read(decay_ticker); + assert_d_eq(mallctl("tcache.flush", NULL, NULL, + (void *)&tcache_ind, sizeof(unsigned)), 0, + "Unexpected mallctl failure"); + tick1 = ticker_read(decay_ticker); + + /* Will only tick if it's in tcache. */ + if (sz <= tcache_max) { + assert_u32_ne(tick1, tick0, + "Expected ticker to tick during tcache " + "flush (sz=%zu)", sz); + } else { + assert_u32_eq(tick1, tick0, + "Unexpected ticker tick during tcache " + "flush (sz=%zu)", sz); + } + } +} +TEST_END + +static void +decay_ticker_helper(unsigned arena_ind, int flags, bool dirty, ssize_t dt, + uint64_t dirty_npurge0, uint64_t muzzy_npurge0, bool terminate_asap) { +#define NINTERVALS 101 + nstime_t time, update_interval, decay_ms, deadline; + + nstime_init(&time, 0); + nstime_update(&time); + + nstime_init2(&decay_ms, dt, 0); + nstime_copy(&deadline, &time); + nstime_add(&deadline, &decay_ms); + + nstime_init2(&update_interval, dt, 0); + nstime_idivide(&update_interval, NINTERVALS); + + /* + * Keep q's slab from being deallocated during the looping below. If a + * cached slab were to repeatedly come and go during looping, it could + * prevent the decay backlog ever becoming empty. + */ + void *p = do_mallocx(1, flags); + uint64_t dirty_npurge1, muzzy_npurge1; + do { + for (unsigned i = 0; i < DECAY_NTICKS_PER_UPDATE / 2; + i++) { + void *q = do_mallocx(1, flags); + dallocx(q, flags); + } + dirty_npurge1 = get_arena_dirty_npurge(arena_ind); + muzzy_npurge1 = get_arena_muzzy_npurge(arena_ind); + + nstime_add(&time_mock, &update_interval); + nstime_update(&time); + } while (nstime_compare(&time, &deadline) <= 0 && ((dirty_npurge1 == + dirty_npurge0 && muzzy_npurge1 == muzzy_npurge0) || + !terminate_asap)); + dallocx(p, flags); + + if (config_stats) { + assert_u64_gt(dirty_npurge1 + muzzy_npurge1, dirty_npurge0 + + muzzy_npurge0, "Expected purging to occur"); + } +#undef NINTERVALS +} + +TEST_BEGIN(test_decay_ticker) { + test_skip_if(check_background_thread_enabled()); +#define NPS 2048 + ssize_t ddt = opt_dirty_decay_ms; + ssize_t mdt = opt_muzzy_decay_ms; + unsigned arena_ind = do_arena_create(ddt, mdt); + int flags = (MALLOCX_ARENA(arena_ind) | MALLOCX_TCACHE_NONE); + void *ps[NPS]; + size_t large; + + /* + * Allocate a bunch of large objects, pause the clock, deallocate every + * other object (to fragment virtual memory), restore the clock, then + * [md]allocx() in a tight loop while advancing time rapidly to verify + * the ticker triggers purging. + */ + + size_t tcache_max; + size_t sz = sizeof(size_t); + assert_d_eq(mallctl("arenas.tcache_max", (void *)&tcache_max, &sz, NULL, + 0), 0, "Unexpected mallctl failure"); + large = nallocx(tcache_max + 1, flags); + + do_purge(arena_ind); + uint64_t dirty_npurge0 = get_arena_dirty_npurge(arena_ind); + uint64_t muzzy_npurge0 = get_arena_muzzy_npurge(arena_ind); + + for (unsigned i = 0; i < NPS; i++) { + ps[i] = do_mallocx(large, flags); + } + + nupdates_mock = 0; + nstime_init(&time_mock, 0); + nstime_update(&time_mock); + monotonic_mock = true; + + nstime_monotonic_orig = nstime_monotonic; + nstime_update_orig = nstime_update; + nstime_monotonic = nstime_monotonic_mock; + nstime_update = nstime_update_mock; + + for (unsigned i = 0; i < NPS; i += 2) { + dallocx(ps[i], flags); + unsigned nupdates0 = nupdates_mock; + do_decay(arena_ind); + assert_u_gt(nupdates_mock, nupdates0, + "Expected nstime_update() to be called"); + } + + decay_ticker_helper(arena_ind, flags, true, ddt, dirty_npurge0, + muzzy_npurge0, true); + decay_ticker_helper(arena_ind, flags, false, ddt+mdt, dirty_npurge0, + muzzy_npurge0, false); + + do_arena_destroy(arena_ind); + + nstime_monotonic = nstime_monotonic_orig; + nstime_update = nstime_update_orig; +#undef NPS +} +TEST_END + +TEST_BEGIN(test_decay_nonmonotonic) { + test_skip_if(check_background_thread_enabled()); +#define NPS (SMOOTHSTEP_NSTEPS + 1) + int flags = (MALLOCX_ARENA(0) | MALLOCX_TCACHE_NONE); + void *ps[NPS]; + uint64_t npurge0 = 0; + uint64_t npurge1 = 0; + size_t sz, large0; + unsigned i, nupdates0; + + sz = sizeof(size_t); + assert_d_eq(mallctl("arenas.lextent.0.size", (void *)&large0, &sz, NULL, + 0), 0, "Unexpected mallctl failure"); + + assert_d_eq(mallctl("arena.0.purge", NULL, NULL, NULL, 0), 0, + "Unexpected mallctl failure"); + do_epoch(); + sz = sizeof(uint64_t); + npurge0 = get_arena_npurge(0); + + nupdates_mock = 0; + nstime_init(&time_mock, 0); + nstime_update(&time_mock); + monotonic_mock = false; + + nstime_monotonic_orig = nstime_monotonic; + nstime_update_orig = nstime_update; + nstime_monotonic = nstime_monotonic_mock; + nstime_update = nstime_update_mock; + + for (i = 0; i < NPS; i++) { + ps[i] = mallocx(large0, flags); + assert_ptr_not_null(ps[i], "Unexpected mallocx() failure"); + } + + for (i = 0; i < NPS; i++) { + dallocx(ps[i], flags); + nupdates0 = nupdates_mock; + assert_d_eq(mallctl("arena.0.decay", NULL, NULL, NULL, 0), 0, + "Unexpected arena.0.decay failure"); + assert_u_gt(nupdates_mock, nupdates0, + "Expected nstime_update() to be called"); + } + + do_epoch(); + sz = sizeof(uint64_t); + npurge1 = get_arena_npurge(0); + + if (config_stats) { + assert_u64_eq(npurge0, npurge1, "Unexpected purging occurred"); + } + + nstime_monotonic = nstime_monotonic_orig; + nstime_update = nstime_update_orig; +#undef NPS +} +TEST_END + +TEST_BEGIN(test_decay_now) { + test_skip_if(check_background_thread_enabled()); + + unsigned arena_ind = do_arena_create(0, 0); + assert_zu_eq(get_arena_pdirty(arena_ind), 0, "Unexpected dirty pages"); + assert_zu_eq(get_arena_pmuzzy(arena_ind), 0, "Unexpected muzzy pages"); + size_t sizes[] = {16, PAGE<<2, HUGEPAGE<<2}; + /* Verify that dirty/muzzy pages never linger after deallocation. */ + for (unsigned i = 0; i < sizeof(sizes)/sizeof(size_t); i++) { + size_t size = sizes[i]; + generate_dirty(arena_ind, size); + assert_zu_eq(get_arena_pdirty(arena_ind), 0, + "Unexpected dirty pages"); + assert_zu_eq(get_arena_pmuzzy(arena_ind), 0, + "Unexpected muzzy pages"); + } + do_arena_destroy(arena_ind); +} +TEST_END + +TEST_BEGIN(test_decay_never) { + test_skip_if(check_background_thread_enabled() || !config_stats); + + unsigned arena_ind = do_arena_create(-1, -1); + int flags = MALLOCX_ARENA(arena_ind) | MALLOCX_TCACHE_NONE; + assert_zu_eq(get_arena_pdirty(arena_ind), 0, "Unexpected dirty pages"); + assert_zu_eq(get_arena_pmuzzy(arena_ind), 0, "Unexpected muzzy pages"); + size_t sizes[] = {16, PAGE<<2, HUGEPAGE<<2}; + void *ptrs[sizeof(sizes)/sizeof(size_t)]; + for (unsigned i = 0; i < sizeof(sizes)/sizeof(size_t); i++) { + ptrs[i] = do_mallocx(sizes[i], flags); + } + /* Verify that each deallocation generates additional dirty pages. */ + size_t pdirty_prev = get_arena_pdirty(arena_ind); + size_t pmuzzy_prev = get_arena_pmuzzy(arena_ind); + assert_zu_eq(pdirty_prev, 0, "Unexpected dirty pages"); + assert_zu_eq(pmuzzy_prev, 0, "Unexpected muzzy pages"); + for (unsigned i = 0; i < sizeof(sizes)/sizeof(size_t); i++) { + dallocx(ptrs[i], flags); + size_t pdirty = get_arena_pdirty(arena_ind); + size_t pmuzzy = get_arena_pmuzzy(arena_ind); + assert_zu_gt(pdirty + (size_t)get_arena_dirty_purged(arena_ind), + pdirty_prev, "Expected dirty pages to increase."); + assert_zu_eq(pmuzzy, 0, "Unexpected muzzy pages"); + pdirty_prev = pdirty; + } + do_arena_destroy(arena_ind); +} +TEST_END + +int +main(void) { + return test( + test_decay_ticks, + test_decay_ticker, + test_decay_nonmonotonic, + test_decay_now, + test_decay_never); +} diff --git a/deps/jemalloc/test/unit/decay.sh b/deps/jemalloc/test/unit/decay.sh new file mode 100644 index 000000000..45aeccf42 --- /dev/null +++ b/deps/jemalloc/test/unit/decay.sh @@ -0,0 +1,3 @@ +#!/bin/sh + +export MALLOC_CONF="dirty_decay_ms:1000,muzzy_decay_ms:1000,lg_tcache_max:0" diff --git a/deps/jemalloc/test/unit/div.c b/deps/jemalloc/test/unit/div.c new file mode 100644 index 000000000..b47f10b2b --- /dev/null +++ b/deps/jemalloc/test/unit/div.c @@ -0,0 +1,29 @@ +#include "test/jemalloc_test.h" + +#include "jemalloc/internal/div.h" + +TEST_BEGIN(test_div_exhaustive) { + for (size_t divisor = 2; divisor < 1000 * 1000; ++divisor) { + div_info_t div_info; + div_init(&div_info, divisor); + size_t max = 1000 * divisor; + if (max < 1000 * 1000) { + max = 1000 * 1000; + } + for (size_t dividend = 0; dividend < 1000 * divisor; + dividend += divisor) { + size_t quotient = div_compute( + &div_info, dividend); + assert_zu_eq(dividend, quotient * divisor, + "With divisor = %zu, dividend = %zu, " + "got quotient %zu", divisor, dividend, quotient); + } + } +} +TEST_END + +int +main(void) { + return test_no_reentrancy( + test_div_exhaustive); +} diff --git a/deps/jemalloc/test/unit/emitter.c b/deps/jemalloc/test/unit/emitter.c new file mode 100644 index 000000000..b4a693f4b --- /dev/null +++ b/deps/jemalloc/test/unit/emitter.c @@ -0,0 +1,469 @@ +#include "test/jemalloc_test.h" +#include "jemalloc/internal/emitter.h" + +/* + * This is so useful for debugging and feature work, we'll leave printing + * functionality committed but disabled by default. + */ +/* Print the text as it will appear. */ +static bool print_raw = false; +/* Print the text escaped, so it can be copied back into the test case. */ +static bool print_escaped = false; + +typedef struct buf_descriptor_s buf_descriptor_t; +struct buf_descriptor_s { + char *buf; + size_t len; + bool mid_quote; +}; + +/* + * Forwards all writes to the passed-in buf_v (which should be cast from a + * buf_descriptor_t *). + */ +static void +forwarding_cb(void *buf_descriptor_v, const char *str) { + buf_descriptor_t *buf_descriptor = (buf_descriptor_t *)buf_descriptor_v; + + if (print_raw) { + malloc_printf("%s", str); + } + if (print_escaped) { + const char *it = str; + while (*it != '\0') { + if (!buf_descriptor->mid_quote) { + malloc_printf("\""); + buf_descriptor->mid_quote = true; + } + switch (*it) { + case '\\': + malloc_printf("\\"); + break; + case '\"': + malloc_printf("\\\""); + break; + case '\t': + malloc_printf("\\t"); + break; + case '\n': + malloc_printf("\\n\"\n"); + buf_descriptor->mid_quote = false; + break; + default: + malloc_printf("%c", *it); + } + it++; + } + } + + size_t written = malloc_snprintf(buf_descriptor->buf, + buf_descriptor->len, "%s", str); + assert_zu_eq(written, strlen(str), "Buffer overflow!"); + buf_descriptor->buf += written; + buf_descriptor->len -= written; + assert_zu_gt(buf_descriptor->len, 0, "Buffer out of space!"); +} + +static void +assert_emit_output(void (*emit_fn)(emitter_t *), + const char *expected_json_output, const char *expected_table_output) { + emitter_t emitter; + char buf[MALLOC_PRINTF_BUFSIZE]; + buf_descriptor_t buf_descriptor; + + buf_descriptor.buf = buf; + buf_descriptor.len = MALLOC_PRINTF_BUFSIZE; + buf_descriptor.mid_quote = false; + + emitter_init(&emitter, emitter_output_json, &forwarding_cb, + &buf_descriptor); + (*emit_fn)(&emitter); + assert_str_eq(expected_json_output, buf, "json output failure"); + + buf_descriptor.buf = buf; + buf_descriptor.len = MALLOC_PRINTF_BUFSIZE; + buf_descriptor.mid_quote = false; + + emitter_init(&emitter, emitter_output_table, &forwarding_cb, + &buf_descriptor); + (*emit_fn)(&emitter); + assert_str_eq(expected_table_output, buf, "table output failure"); +} + +static void +emit_dict(emitter_t *emitter) { + bool b_false = false; + bool b_true = true; + int i_123 = 123; + const char *str = "a string"; + + emitter_begin(emitter); + emitter_dict_begin(emitter, "foo", "This is the foo table:"); + emitter_kv(emitter, "abc", "ABC", emitter_type_bool, &b_false); + emitter_kv(emitter, "def", "DEF", emitter_type_bool, &b_true); + emitter_kv_note(emitter, "ghi", "GHI", emitter_type_int, &i_123, + "note_key1", emitter_type_string, &str); + emitter_kv_note(emitter, "jkl", "JKL", emitter_type_string, &str, + "note_key2", emitter_type_bool, &b_false); + emitter_dict_end(emitter); + emitter_end(emitter); +} +static const char *dict_json = +"{\n" +"\t\"foo\": {\n" +"\t\t\"abc\": false,\n" +"\t\t\"def\": true,\n" +"\t\t\"ghi\": 123,\n" +"\t\t\"jkl\": \"a string\"\n" +"\t}\n" +"}\n"; +static const char *dict_table = +"This is the foo table:\n" +" ABC: false\n" +" DEF: true\n" +" GHI: 123 (note_key1: \"a string\")\n" +" JKL: \"a string\" (note_key2: false)\n"; + +TEST_BEGIN(test_dict) { + assert_emit_output(&emit_dict, dict_json, dict_table); +} +TEST_END + +static void +emit_table_printf(emitter_t *emitter) { + emitter_begin(emitter); + emitter_table_printf(emitter, "Table note 1\n"); + emitter_table_printf(emitter, "Table note 2 %s\n", + "with format string"); + emitter_end(emitter); +} + +static const char *table_printf_json = +"{\n" +"}\n"; + +static const char *table_printf_table = +"Table note 1\n" +"Table note 2 with format string\n"; + +TEST_BEGIN(test_table_printf) { + assert_emit_output(&emit_table_printf, table_printf_json, + table_printf_table); +} +TEST_END + +static void emit_nested_dict(emitter_t *emitter) { + int val = 123; + emitter_begin(emitter); + emitter_dict_begin(emitter, "json1", "Dict 1"); + emitter_dict_begin(emitter, "json2", "Dict 2"); + emitter_kv(emitter, "primitive", "A primitive", emitter_type_int, &val); + emitter_dict_end(emitter); /* Close 2 */ + emitter_dict_begin(emitter, "json3", "Dict 3"); + emitter_dict_end(emitter); /* Close 3 */ + emitter_dict_end(emitter); /* Close 1 */ + emitter_dict_begin(emitter, "json4", "Dict 4"); + emitter_kv(emitter, "primitive", "Another primitive", + emitter_type_int, &val); + emitter_dict_end(emitter); /* Close 4 */ + emitter_end(emitter); +} + +static const char *nested_object_json = +"{\n" +"\t\"json1\": {\n" +"\t\t\"json2\": {\n" +"\t\t\t\"primitive\": 123\n" +"\t\t},\n" +"\t\t\"json3\": {\n" +"\t\t}\n" +"\t},\n" +"\t\"json4\": {\n" +"\t\t\"primitive\": 123\n" +"\t}\n" +"}\n"; + +static const char *nested_object_table = +"Dict 1\n" +" Dict 2\n" +" A primitive: 123\n" +" Dict 3\n" +"Dict 4\n" +" Another primitive: 123\n"; + +TEST_BEGIN(test_nested_dict) { + assert_emit_output(&emit_nested_dict, nested_object_json, + nested_object_table); +} +TEST_END + +static void +emit_types(emitter_t *emitter) { + bool b = false; + int i = -123; + unsigned u = 123; + ssize_t zd = -456; + size_t zu = 456; + const char *str = "string"; + uint32_t u32 = 789; + uint64_t u64 = 10000000000ULL; + + emitter_begin(emitter); + emitter_kv(emitter, "k1", "K1", emitter_type_bool, &b); + emitter_kv(emitter, "k2", "K2", emitter_type_int, &i); + emitter_kv(emitter, "k3", "K3", emitter_type_unsigned, &u); + emitter_kv(emitter, "k4", "K4", emitter_type_ssize, &zd); + emitter_kv(emitter, "k5", "K5", emitter_type_size, &zu); + emitter_kv(emitter, "k6", "K6", emitter_type_string, &str); + emitter_kv(emitter, "k7", "K7", emitter_type_uint32, &u32); + emitter_kv(emitter, "k8", "K8", emitter_type_uint64, &u64); + /* + * We don't test the title type, since it's only used for tables. It's + * tested in the emitter_table_row tests. + */ + emitter_end(emitter); +} + +static const char *types_json = +"{\n" +"\t\"k1\": false,\n" +"\t\"k2\": -123,\n" +"\t\"k3\": 123,\n" +"\t\"k4\": -456,\n" +"\t\"k5\": 456,\n" +"\t\"k6\": \"string\",\n" +"\t\"k7\": 789,\n" +"\t\"k8\": 10000000000\n" +"}\n"; + +static const char *types_table = +"K1: false\n" +"K2: -123\n" +"K3: 123\n" +"K4: -456\n" +"K5: 456\n" +"K6: \"string\"\n" +"K7: 789\n" +"K8: 10000000000\n"; + +TEST_BEGIN(test_types) { + assert_emit_output(&emit_types, types_json, types_table); +} +TEST_END + +static void +emit_modal(emitter_t *emitter) { + int val = 123; + emitter_begin(emitter); + emitter_dict_begin(emitter, "j0", "T0"); + emitter_json_key(emitter, "j1"); + emitter_json_object_begin(emitter); + emitter_kv(emitter, "i1", "I1", emitter_type_int, &val); + emitter_json_kv(emitter, "i2", emitter_type_int, &val); + emitter_table_kv(emitter, "I3", emitter_type_int, &val); + emitter_table_dict_begin(emitter, "T1"); + emitter_kv(emitter, "i4", "I4", emitter_type_int, &val); + emitter_json_object_end(emitter); /* Close j1 */ + emitter_kv(emitter, "i5", "I5", emitter_type_int, &val); + emitter_table_dict_end(emitter); /* Close T1 */ + emitter_kv(emitter, "i6", "I6", emitter_type_int, &val); + emitter_dict_end(emitter); /* Close j0 / T0 */ + emitter_end(emitter); +} + +const char *modal_json = +"{\n" +"\t\"j0\": {\n" +"\t\t\"j1\": {\n" +"\t\t\t\"i1\": 123,\n" +"\t\t\t\"i2\": 123,\n" +"\t\t\t\"i4\": 123\n" +"\t\t},\n" +"\t\t\"i5\": 123,\n" +"\t\t\"i6\": 123\n" +"\t}\n" +"}\n"; + +const char *modal_table = +"T0\n" +" I1: 123\n" +" I3: 123\n" +" T1\n" +" I4: 123\n" +" I5: 123\n" +" I6: 123\n"; + +TEST_BEGIN(test_modal) { + assert_emit_output(&emit_modal, modal_json, modal_table); +} +TEST_END + +static void +emit_json_arr(emitter_t *emitter) { + int ival = 123; + + emitter_begin(emitter); + emitter_json_key(emitter, "dict"); + emitter_json_object_begin(emitter); + emitter_json_key(emitter, "arr"); + emitter_json_array_begin(emitter); + emitter_json_object_begin(emitter); + emitter_json_kv(emitter, "foo", emitter_type_int, &ival); + emitter_json_object_end(emitter); /* Close arr[0] */ + /* arr[1] and arr[2] are primitives. */ + emitter_json_value(emitter, emitter_type_int, &ival); + emitter_json_value(emitter, emitter_type_int, &ival); + emitter_json_object_begin(emitter); + emitter_json_kv(emitter, "bar", emitter_type_int, &ival); + emitter_json_kv(emitter, "baz", emitter_type_int, &ival); + emitter_json_object_end(emitter); /* Close arr[3]. */ + emitter_json_array_end(emitter); /* Close arr. */ + emitter_json_object_end(emitter); /* Close dict. */ + emitter_end(emitter); +} + +static const char *json_array_json = +"{\n" +"\t\"dict\": {\n" +"\t\t\"arr\": [\n" +"\t\t\t{\n" +"\t\t\t\t\"foo\": 123\n" +"\t\t\t},\n" +"\t\t\t123,\n" +"\t\t\t123,\n" +"\t\t\t{\n" +"\t\t\t\t\"bar\": 123,\n" +"\t\t\t\t\"baz\": 123\n" +"\t\t\t}\n" +"\t\t]\n" +"\t}\n" +"}\n"; + +static const char *json_array_table = ""; + +TEST_BEGIN(test_json_arr) { + assert_emit_output(&emit_json_arr, json_array_json, json_array_table); +} +TEST_END + +static void +emit_json_nested_array(emitter_t *emitter) { + int ival = 123; + char *sval = "foo"; + emitter_begin(emitter); + emitter_json_array_begin(emitter); + emitter_json_array_begin(emitter); + emitter_json_value(emitter, emitter_type_int, &ival); + emitter_json_value(emitter, emitter_type_string, &sval); + emitter_json_value(emitter, emitter_type_int, &ival); + emitter_json_value(emitter, emitter_type_string, &sval); + emitter_json_array_end(emitter); + emitter_json_array_begin(emitter); + emitter_json_value(emitter, emitter_type_int, &ival); + emitter_json_array_end(emitter); + emitter_json_array_begin(emitter); + emitter_json_value(emitter, emitter_type_string, &sval); + emitter_json_value(emitter, emitter_type_int, &ival); + emitter_json_array_end(emitter); + emitter_json_array_begin(emitter); + emitter_json_array_end(emitter); + emitter_json_array_end(emitter); + emitter_end(emitter); +} + +static const char *json_nested_array_json = +"{\n" +"\t[\n" +"\t\t[\n" +"\t\t\t123,\n" +"\t\t\t\"foo\",\n" +"\t\t\t123,\n" +"\t\t\t\"foo\"\n" +"\t\t],\n" +"\t\t[\n" +"\t\t\t123\n" +"\t\t],\n" +"\t\t[\n" +"\t\t\t\"foo\",\n" +"\t\t\t123\n" +"\t\t],\n" +"\t\t[\n" +"\t\t]\n" +"\t]\n" +"}\n"; + +TEST_BEGIN(test_json_nested_arr) { + assert_emit_output(&emit_json_nested_array, json_nested_array_json, + json_array_table); +} +TEST_END + +static void +emit_table_row(emitter_t *emitter) { + emitter_begin(emitter); + emitter_row_t row; + emitter_col_t abc = {emitter_justify_left, 10, emitter_type_title, {0}, {0, 0}}; + abc.str_val = "ABC title"; + emitter_col_t def = {emitter_justify_right, 15, emitter_type_title, {0}, {0, 0}}; + def.str_val = "DEF title"; + emitter_col_t ghi = {emitter_justify_right, 5, emitter_type_title, {0}, {0, 0}}; + ghi.str_val = "GHI"; + + emitter_row_init(&row); + emitter_col_init(&abc, &row); + emitter_col_init(&def, &row); + emitter_col_init(&ghi, &row); + + emitter_table_row(emitter, &row); + + abc.type = emitter_type_int; + def.type = emitter_type_bool; + ghi.type = emitter_type_int; + + abc.int_val = 123; + def.bool_val = true; + ghi.int_val = 456; + emitter_table_row(emitter, &row); + + abc.int_val = 789; + def.bool_val = false; + ghi.int_val = 1011; + emitter_table_row(emitter, &row); + + abc.type = emitter_type_string; + abc.str_val = "a string"; + def.bool_val = false; + ghi.type = emitter_type_title; + ghi.str_val = "ghi"; + emitter_table_row(emitter, &row); + + emitter_end(emitter); +} + +static const char *table_row_json = +"{\n" +"}\n"; + +static const char *table_row_table = +"ABC title DEF title GHI\n" +"123 true 456\n" +"789 false 1011\n" +"\"a string\" false ghi\n"; + +TEST_BEGIN(test_table_row) { + assert_emit_output(&emit_table_row, table_row_json, table_row_table); +} +TEST_END + +int +main(void) { + return test_no_reentrancy( + test_dict, + test_table_printf, + test_nested_dict, + test_types, + test_modal, + test_json_arr, + test_json_nested_arr, + test_table_row); +} diff --git a/deps/jemalloc/test/unit/extent_quantize.c b/deps/jemalloc/test/unit/extent_quantize.c new file mode 100644 index 000000000..0ca7a75d9 --- /dev/null +++ b/deps/jemalloc/test/unit/extent_quantize.c @@ -0,0 +1,141 @@ +#include "test/jemalloc_test.h" + +TEST_BEGIN(test_small_extent_size) { + unsigned nbins, i; + size_t sz, extent_size; + size_t mib[4]; + size_t miblen = sizeof(mib) / sizeof(size_t); + + /* + * Iterate over all small size classes, get their extent sizes, and + * verify that the quantized size is the same as the extent size. + */ + + sz = sizeof(unsigned); + assert_d_eq(mallctl("arenas.nbins", (void *)&nbins, &sz, NULL, 0), 0, + "Unexpected mallctl failure"); + + assert_d_eq(mallctlnametomib("arenas.bin.0.slab_size", mib, &miblen), 0, + "Unexpected mallctlnametomib failure"); + for (i = 0; i < nbins; i++) { + mib[2] = i; + sz = sizeof(size_t); + assert_d_eq(mallctlbymib(mib, miblen, (void *)&extent_size, &sz, + NULL, 0), 0, "Unexpected mallctlbymib failure"); + assert_zu_eq(extent_size, + extent_size_quantize_floor(extent_size), + "Small extent quantization should be a no-op " + "(extent_size=%zu)", extent_size); + assert_zu_eq(extent_size, + extent_size_quantize_ceil(extent_size), + "Small extent quantization should be a no-op " + "(extent_size=%zu)", extent_size); + } +} +TEST_END + +TEST_BEGIN(test_large_extent_size) { + bool cache_oblivious; + unsigned nlextents, i; + size_t sz, extent_size_prev, ceil_prev; + size_t mib[4]; + size_t miblen = sizeof(mib) / sizeof(size_t); + + /* + * Iterate over all large size classes, get their extent sizes, and + * verify that the quantized size is the same as the extent size. + */ + + sz = sizeof(bool); + assert_d_eq(mallctl("config.cache_oblivious", (void *)&cache_oblivious, + &sz, NULL, 0), 0, "Unexpected mallctl failure"); + + sz = sizeof(unsigned); + assert_d_eq(mallctl("arenas.nlextents", (void *)&nlextents, &sz, NULL, + 0), 0, "Unexpected mallctl failure"); + + assert_d_eq(mallctlnametomib("arenas.lextent.0.size", mib, &miblen), 0, + "Unexpected mallctlnametomib failure"); + for (i = 0; i < nlextents; i++) { + size_t lextent_size, extent_size, floor, ceil; + + mib[2] = i; + sz = sizeof(size_t); + assert_d_eq(mallctlbymib(mib, miblen, (void *)&lextent_size, + &sz, NULL, 0), 0, "Unexpected mallctlbymib failure"); + extent_size = cache_oblivious ? lextent_size + PAGE : + lextent_size; + floor = extent_size_quantize_floor(extent_size); + ceil = extent_size_quantize_ceil(extent_size); + + assert_zu_eq(extent_size, floor, + "Extent quantization should be a no-op for precise size " + "(lextent_size=%zu, extent_size=%zu)", lextent_size, + extent_size); + assert_zu_eq(extent_size, ceil, + "Extent quantization should be a no-op for precise size " + "(lextent_size=%zu, extent_size=%zu)", lextent_size, + extent_size); + + if (i > 0) { + assert_zu_eq(extent_size_prev, + extent_size_quantize_floor(extent_size - PAGE), + "Floor should be a precise size"); + if (extent_size_prev < ceil_prev) { + assert_zu_eq(ceil_prev, extent_size, + "Ceiling should be a precise size " + "(extent_size_prev=%zu, ceil_prev=%zu, " + "extent_size=%zu)", extent_size_prev, + ceil_prev, extent_size); + } + } + if (i + 1 < nlextents) { + extent_size_prev = floor; + ceil_prev = extent_size_quantize_ceil(extent_size + + PAGE); + } + } +} +TEST_END + +TEST_BEGIN(test_monotonic) { +#define SZ_MAX ZU(4 * 1024 * 1024) + unsigned i; + size_t floor_prev, ceil_prev; + + floor_prev = 0; + ceil_prev = 0; + for (i = 1; i <= SZ_MAX >> LG_PAGE; i++) { + size_t extent_size, floor, ceil; + + extent_size = i << LG_PAGE; + floor = extent_size_quantize_floor(extent_size); + ceil = extent_size_quantize_ceil(extent_size); + + assert_zu_le(floor, extent_size, + "Floor should be <= (floor=%zu, extent_size=%zu, ceil=%zu)", + floor, extent_size, ceil); + assert_zu_ge(ceil, extent_size, + "Ceiling should be >= (floor=%zu, extent_size=%zu, " + "ceil=%zu)", floor, extent_size, ceil); + + assert_zu_le(floor_prev, floor, "Floor should be monotonic " + "(floor_prev=%zu, floor=%zu, extent_size=%zu, ceil=%zu)", + floor_prev, floor, extent_size, ceil); + assert_zu_le(ceil_prev, ceil, "Ceiling should be monotonic " + "(floor=%zu, extent_size=%zu, ceil_prev=%zu, ceil=%zu)", + floor, extent_size, ceil_prev, ceil); + + floor_prev = floor; + ceil_prev = ceil; + } +} +TEST_END + +int +main(void) { + return test( + test_small_extent_size, + test_large_extent_size, + test_monotonic); +} diff --git a/deps/jemalloc/test/unit/extent_util.c b/deps/jemalloc/test/unit/extent_util.c new file mode 100644 index 000000000..97e55f0f6 --- /dev/null +++ b/deps/jemalloc/test/unit/extent_util.c @@ -0,0 +1,269 @@ +#include "test/jemalloc_test.h" + +#define TEST_UTIL_EINVAL(node, a, b, c, d, why_inval) do { \ + assert_d_eq(mallctl("experimental.utilization." node, \ + a, b, c, d), EINVAL, "Should fail when " why_inval); \ + assert_zu_eq(out_sz, out_sz_ref, \ + "Output size touched when given invalid arguments"); \ + assert_d_eq(memcmp(out, out_ref, out_sz_ref), 0, \ + "Output content touched when given invalid arguments"); \ +} while (0) + +#define TEST_UTIL_QUERY_EINVAL(a, b, c, d, why_inval) \ + TEST_UTIL_EINVAL("query", a, b, c, d, why_inval) +#define TEST_UTIL_BATCH_EINVAL(a, b, c, d, why_inval) \ + TEST_UTIL_EINVAL("batch_query", a, b, c, d, why_inval) + +#define TEST_UTIL_VALID(node) do { \ + assert_d_eq(mallctl("experimental.utilization." node, \ + out, &out_sz, in, in_sz), 0, \ + "Should return 0 on correct arguments"); \ + assert_zu_eq(out_sz, out_sz_ref, "incorrect output size"); \ + assert_d_ne(memcmp(out, out_ref, out_sz_ref), 0, \ + "Output content should be changed"); \ +} while (0) + +#define TEST_UTIL_BATCH_VALID TEST_UTIL_VALID("batch_query") + +#define TEST_MAX_SIZE (1 << 20) + +TEST_BEGIN(test_query) { + size_t sz; + /* + * Select some sizes that can span both small and large sizes, and are + * numerically unrelated to any size boundaries. + */ + for (sz = 7; sz <= TEST_MAX_SIZE && sz <= SC_LARGE_MAXCLASS; + sz += (sz <= SC_SMALL_MAXCLASS ? 1009 : 99989)) { + void *p = mallocx(sz, 0); + void **in = &p; + size_t in_sz = sizeof(const void *); + size_t out_sz = sizeof(void *) + sizeof(size_t) * 5; + void *out = mallocx(out_sz, 0); + void *out_ref = mallocx(out_sz, 0); + size_t out_sz_ref = out_sz; + + assert_ptr_not_null(p, + "test pointer allocation failed"); + assert_ptr_not_null(out, + "test output allocation failed"); + assert_ptr_not_null(out_ref, + "test reference output allocation failed"); + +#define SLABCUR_READ(out) (*(void **)out) +#define COUNTS(out) ((size_t *)((void **)out + 1)) +#define NFREE_READ(out) COUNTS(out)[0] +#define NREGS_READ(out) COUNTS(out)[1] +#define SIZE_READ(out) COUNTS(out)[2] +#define BIN_NFREE_READ(out) COUNTS(out)[3] +#define BIN_NREGS_READ(out) COUNTS(out)[4] + + SLABCUR_READ(out) = NULL; + NFREE_READ(out) = NREGS_READ(out) = SIZE_READ(out) = -1; + BIN_NFREE_READ(out) = BIN_NREGS_READ(out) = -1; + memcpy(out_ref, out, out_sz); + + /* Test invalid argument(s) errors */ + TEST_UTIL_QUERY_EINVAL(NULL, &out_sz, in, in_sz, + "old is NULL"); + TEST_UTIL_QUERY_EINVAL(out, NULL, in, in_sz, + "oldlenp is NULL"); + TEST_UTIL_QUERY_EINVAL(out, &out_sz, NULL, in_sz, + "newp is NULL"); + TEST_UTIL_QUERY_EINVAL(out, &out_sz, in, 0, + "newlen is zero"); + in_sz -= 1; + TEST_UTIL_QUERY_EINVAL(out, &out_sz, in, in_sz, + "invalid newlen"); + in_sz += 1; + out_sz_ref = out_sz -= 2 * sizeof(size_t); + TEST_UTIL_QUERY_EINVAL(out, &out_sz, in, in_sz, + "invalid *oldlenp"); + out_sz_ref = out_sz += 2 * sizeof(size_t); + + /* Examine output for valid call */ + TEST_UTIL_VALID("query"); + assert_zu_le(sz, SIZE_READ(out), + "Extent size should be at least allocation size"); + assert_zu_eq(SIZE_READ(out) & (PAGE - 1), 0, + "Extent size should be a multiple of page size"); + if (sz <= SC_SMALL_MAXCLASS) { + assert_zu_le(NFREE_READ(out), NREGS_READ(out), + "Extent free count exceeded region count"); + assert_zu_le(NREGS_READ(out), SIZE_READ(out), + "Extent region count exceeded size"); + assert_zu_ne(NREGS_READ(out), 0, + "Extent region count must be positive"); + assert_ptr_not_null(SLABCUR_READ(out), + "Current slab is null"); + assert_true(NFREE_READ(out) == 0 + || SLABCUR_READ(out) <= p, + "Allocation should follow first fit principle"); + if (config_stats) { + assert_zu_le(BIN_NFREE_READ(out), + BIN_NREGS_READ(out), + "Bin free count exceeded region count"); + assert_zu_ne(BIN_NREGS_READ(out), 0, + "Bin region count must be positive"); + assert_zu_le(NFREE_READ(out), + BIN_NFREE_READ(out), + "Extent free count exceeded bin free count"); + assert_zu_le(NREGS_READ(out), + BIN_NREGS_READ(out), + "Extent region count exceeded " + "bin region count"); + assert_zu_eq(BIN_NREGS_READ(out) + % NREGS_READ(out), 0, + "Bin region count isn't a multiple of " + "extent region count"); + assert_zu_le( + BIN_NFREE_READ(out) - NFREE_READ(out), + BIN_NREGS_READ(out) - NREGS_READ(out), + "Free count in other extents in the bin " + "exceeded region count in other extents " + "in the bin"); + assert_zu_le(NREGS_READ(out) - NFREE_READ(out), + BIN_NREGS_READ(out) - BIN_NFREE_READ(out), + "Extent utilized count exceeded " + "bin utilized count"); + } + } else { + assert_zu_eq(NFREE_READ(out), 0, + "Extent free count should be zero"); + assert_zu_eq(NREGS_READ(out), 1, + "Extent region count should be one"); + assert_ptr_null(SLABCUR_READ(out), + "Current slab must be null for large size classes"); + if (config_stats) { + assert_zu_eq(BIN_NFREE_READ(out), 0, + "Bin free count must be zero for " + "large sizes"); + assert_zu_eq(BIN_NREGS_READ(out), 0, + "Bin region count must be zero for " + "large sizes"); + } + } + +#undef BIN_NREGS_READ +#undef BIN_NFREE_READ +#undef SIZE_READ +#undef NREGS_READ +#undef NFREE_READ +#undef COUNTS +#undef SLABCUR_READ + + free(out_ref); + free(out); + free(p); + } +} +TEST_END + +TEST_BEGIN(test_batch) { + size_t sz; + /* + * Select some sizes that can span both small and large sizes, and are + * numerically unrelated to any size boundaries. + */ + for (sz = 17; sz <= TEST_MAX_SIZE && sz <= SC_LARGE_MAXCLASS; + sz += (sz <= SC_SMALL_MAXCLASS ? 1019 : 99991)) { + void *p = mallocx(sz, 0); + void *q = mallocx(sz, 0); + void *in[] = {p, q}; + size_t in_sz = sizeof(const void *) * 2; + size_t out[] = {-1, -1, -1, -1, -1, -1}; + size_t out_sz = sizeof(size_t) * 6; + size_t out_ref[] = {-1, -1, -1, -1, -1, -1}; + size_t out_sz_ref = out_sz; + + assert_ptr_not_null(p, "test pointer allocation failed"); + assert_ptr_not_null(q, "test pointer allocation failed"); + + /* Test invalid argument(s) errors */ + TEST_UTIL_BATCH_EINVAL(NULL, &out_sz, in, in_sz, + "old is NULL"); + TEST_UTIL_BATCH_EINVAL(out, NULL, in, in_sz, + "oldlenp is NULL"); + TEST_UTIL_BATCH_EINVAL(out, &out_sz, NULL, in_sz, + "newp is NULL"); + TEST_UTIL_BATCH_EINVAL(out, &out_sz, in, 0, + "newlen is zero"); + in_sz -= 1; + TEST_UTIL_BATCH_EINVAL(out, &out_sz, in, in_sz, + "newlen is not an exact multiple"); + in_sz += 1; + out_sz_ref = out_sz -= 2 * sizeof(size_t); + TEST_UTIL_BATCH_EINVAL(out, &out_sz, in, in_sz, + "*oldlenp is not an exact multiple"); + out_sz_ref = out_sz += 2 * sizeof(size_t); + in_sz -= sizeof(const void *); + TEST_UTIL_BATCH_EINVAL(out, &out_sz, in, in_sz, + "*oldlenp and newlen do not match"); + in_sz += sizeof(const void *); + + /* Examine output for valid calls */ +#define TEST_EQUAL_REF(i, message) \ + assert_d_eq(memcmp(out + (i) * 3, out_ref + (i) * 3, 3), 0, message) + +#define NFREE_READ(out, i) out[(i) * 3] +#define NREGS_READ(out, i) out[(i) * 3 + 1] +#define SIZE_READ(out, i) out[(i) * 3 + 2] + + out_sz_ref = out_sz /= 2; + in_sz /= 2; + TEST_UTIL_BATCH_VALID; + assert_zu_le(sz, SIZE_READ(out, 0), + "Extent size should be at least allocation size"); + assert_zu_eq(SIZE_READ(out, 0) & (PAGE - 1), 0, + "Extent size should be a multiple of page size"); + if (sz <= SC_SMALL_MAXCLASS) { + assert_zu_le(NFREE_READ(out, 0), NREGS_READ(out, 0), + "Extent free count exceeded region count"); + assert_zu_le(NREGS_READ(out, 0), SIZE_READ(out, 0), + "Extent region count exceeded size"); + assert_zu_ne(NREGS_READ(out, 0), 0, + "Extent region count must be positive"); + } else { + assert_zu_eq(NFREE_READ(out, 0), 0, + "Extent free count should be zero"); + assert_zu_eq(NREGS_READ(out, 0), 1, + "Extent region count should be one"); + } + TEST_EQUAL_REF(1, + "Should not overwrite content beyond what's needed"); + in_sz *= 2; + out_sz_ref = out_sz *= 2; + + memcpy(out_ref, out, 3 * sizeof(size_t)); + TEST_UTIL_BATCH_VALID; + TEST_EQUAL_REF(0, "Statistics should be stable across calls"); + if (sz <= SC_SMALL_MAXCLASS) { + assert_zu_le(NFREE_READ(out, 1), NREGS_READ(out, 1), + "Extent free count exceeded region count"); + } else { + assert_zu_eq(NFREE_READ(out, 0), 0, + "Extent free count should be zero"); + } + assert_zu_eq(NREGS_READ(out, 0), NREGS_READ(out, 1), + "Extent region count should be same for same region size"); + assert_zu_eq(SIZE_READ(out, 0), SIZE_READ(out, 1), + "Extent size should be same for same region size"); + +#undef SIZE_READ +#undef NREGS_READ +#undef NFREE_READ + +#undef TEST_EQUAL_REF + + free(q); + free(p); + } +} +TEST_END + +int +main(void) { + assert_zu_lt(SC_SMALL_MAXCLASS, TEST_MAX_SIZE, + "Test case cannot cover large classes"); + return test(test_query, test_batch); +} diff --git a/deps/jemalloc/test/unit/fork.c b/deps/jemalloc/test/unit/fork.c new file mode 100644 index 000000000..b1690750a --- /dev/null +++ b/deps/jemalloc/test/unit/fork.c @@ -0,0 +1,141 @@ +#include "test/jemalloc_test.h" + +#ifndef _WIN32 +#include <sys/wait.h> +#endif + +#ifndef _WIN32 +static void +wait_for_child_exit(int pid) { + int status; + while (true) { + if (waitpid(pid, &status, 0) == -1) { + test_fail("Unexpected waitpid() failure."); + } + if (WIFSIGNALED(status)) { + test_fail("Unexpected child termination due to " + "signal %d", WTERMSIG(status)); + break; + } + if (WIFEXITED(status)) { + if (WEXITSTATUS(status) != 0) { + test_fail("Unexpected child exit value %d", + WEXITSTATUS(status)); + } + break; + } + } +} +#endif + +TEST_BEGIN(test_fork) { +#ifndef _WIN32 + void *p; + pid_t pid; + + /* Set up a manually managed arena for test. */ + unsigned arena_ind; + size_t sz = sizeof(unsigned); + assert_d_eq(mallctl("arenas.create", (void *)&arena_ind, &sz, NULL, 0), + 0, "Unexpected mallctl() failure"); + + /* Migrate to the new arena. */ + unsigned old_arena_ind; + sz = sizeof(old_arena_ind); + assert_d_eq(mallctl("thread.arena", (void *)&old_arena_ind, &sz, + (void *)&arena_ind, sizeof(arena_ind)), 0, + "Unexpected mallctl() failure"); + + p = malloc(1); + assert_ptr_not_null(p, "Unexpected malloc() failure"); + + pid = fork(); + + free(p); + + p = malloc(64); + assert_ptr_not_null(p, "Unexpected malloc() failure"); + free(p); + + if (pid == -1) { + /* Error. */ + test_fail("Unexpected fork() failure"); + } else if (pid == 0) { + /* Child. */ + _exit(0); + } else { + wait_for_child_exit(pid); + } +#else + test_skip("fork(2) is irrelevant to Windows"); +#endif +} +TEST_END + +#ifndef _WIN32 +static void * +do_fork_thd(void *arg) { + malloc(1); + int pid = fork(); + if (pid == -1) { + /* Error. */ + test_fail("Unexpected fork() failure"); + } else if (pid == 0) { + /* Child. */ + char *args[] = {"true", NULL}; + execvp(args[0], args); + test_fail("Exec failed"); + } else { + /* Parent */ + wait_for_child_exit(pid); + } + return NULL; +} +#endif + +#ifndef _WIN32 +static void +do_test_fork_multithreaded() { + thd_t child; + thd_create(&child, do_fork_thd, NULL); + do_fork_thd(NULL); + thd_join(child, NULL); +} +#endif + +TEST_BEGIN(test_fork_multithreaded) { +#ifndef _WIN32 + /* + * We've seen bugs involving hanging on arenas_lock (though the same + * class of bugs can happen on any mutex). The bugs are intermittent + * though, so we want to run the test multiple times. Since we hold the + * arenas lock only early in the process lifetime, we can't just run + * this test in a loop (since, after all the arenas are initialized, we + * won't acquire arenas_lock any further). We therefore repeat the test + * with multiple processes. + */ + for (int i = 0; i < 100; i++) { + int pid = fork(); + if (pid == -1) { + /* Error. */ + test_fail("Unexpected fork() failure,"); + } else if (pid == 0) { + /* Child. */ + do_test_fork_multithreaded(); + _exit(0); + } else { + wait_for_child_exit(pid); + } + } +#else + test_skip("fork(2) is irrelevant to Windows"); +#endif +} +TEST_END + +int +main(void) { + return test_no_reentrancy( + test_fork, + test_fork_multithreaded); +} diff --git a/deps/jemalloc/test/unit/hash.c b/deps/jemalloc/test/unit/hash.c new file mode 100644 index 000000000..7cc034f8d --- /dev/null +++ b/deps/jemalloc/test/unit/hash.c @@ -0,0 +1,173 @@ +/* + * This file is based on code that is part of SMHasher + * (https://code.google.com/p/smhasher/), and is subject to the MIT license + * (http://www.opensource.org/licenses/mit-license.php). Both email addresses + * associated with the source code's revision history belong to Austin Appleby, + * and the revision history ranges from 2010 to 2012. Therefore the copyright + * and license are here taken to be: + * + * Copyright (c) 2010-2012 Austin Appleby + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#include "test/jemalloc_test.h" +#include "jemalloc/internal/hash.h" + +typedef enum { + hash_variant_x86_32, + hash_variant_x86_128, + hash_variant_x64_128 +} hash_variant_t; + +static int +hash_variant_bits(hash_variant_t variant) { + switch (variant) { + case hash_variant_x86_32: return 32; + case hash_variant_x86_128: return 128; + case hash_variant_x64_128: return 128; + default: not_reached(); + } +} + +static const char * +hash_variant_string(hash_variant_t variant) { + switch (variant) { + case hash_variant_x86_32: return "hash_x86_32"; + case hash_variant_x86_128: return "hash_x86_128"; + case hash_variant_x64_128: return "hash_x64_128"; + default: not_reached(); + } +} + +#define KEY_SIZE 256 +static void +hash_variant_verify_key(hash_variant_t variant, uint8_t *key) { + const int hashbytes = hash_variant_bits(variant) / 8; + const int hashes_size = hashbytes * 256; + VARIABLE_ARRAY(uint8_t, hashes, hashes_size); + VARIABLE_ARRAY(uint8_t, final, hashbytes); + unsigned i; + uint32_t computed, expected; + + memset(key, 0, KEY_SIZE); + memset(hashes, 0, hashes_size); + memset(final, 0, hashbytes); + + /* + * Hash keys of the form {0}, {0,1}, {0,1,2}, ..., {0,1,...,255} as the + * seed. + */ + for (i = 0; i < 256; i++) { + key[i] = (uint8_t)i; + switch (variant) { + case hash_variant_x86_32: { + uint32_t out; + out = hash_x86_32(key, i, 256-i); + memcpy(&hashes[i*hashbytes], &out, hashbytes); + break; + } case hash_variant_x86_128: { + uint64_t out[2]; + hash_x86_128(key, i, 256-i, out); + memcpy(&hashes[i*hashbytes], out, hashbytes); + break; + } case hash_variant_x64_128: { + uint64_t out[2]; + hash_x64_128(key, i, 256-i, out); + memcpy(&hashes[i*hashbytes], out, hashbytes); + break; + } default: not_reached(); + } + } + + /* Hash the result array. */ + switch (variant) { + case hash_variant_x86_32: { + uint32_t out = hash_x86_32(hashes, hashes_size, 0); + memcpy(final, &out, sizeof(out)); + break; + } case hash_variant_x86_128: { + uint64_t out[2]; + hash_x86_128(hashes, hashes_size, 0, out); + memcpy(final, out, sizeof(out)); + break; + } case hash_variant_x64_128: { + uint64_t out[2]; + hash_x64_128(hashes, hashes_size, 0, out); + memcpy(final, out, sizeof(out)); + break; + } default: not_reached(); + } + + computed = (final[0] << 0) | (final[1] << 8) | (final[2] << 16) | + (final[3] << 24); + + switch (variant) { +#ifdef JEMALLOC_BIG_ENDIAN + case hash_variant_x86_32: expected = 0x6213303eU; break; + case hash_variant_x86_128: expected = 0x266820caU; break; + case hash_variant_x64_128: expected = 0xcc622b6fU; break; +#else + case hash_variant_x86_32: expected = 0xb0f57ee3U; break; + case hash_variant_x86_128: expected = 0xb3ece62aU; break; + case hash_variant_x64_128: expected = 0x6384ba69U; break; +#endif + default: not_reached(); + } + + assert_u32_eq(computed, expected, + "Hash mismatch for %s(): expected %#x but got %#x", + hash_variant_string(variant), expected, computed); +} + +static void +hash_variant_verify(hash_variant_t variant) { +#define MAX_ALIGN 16 + uint8_t key[KEY_SIZE + (MAX_ALIGN - 1)]; + unsigned i; + + for (i = 0; i < MAX_ALIGN; i++) { + hash_variant_verify_key(variant, &key[i]); + } +#undef MAX_ALIGN +} +#undef KEY_SIZE + +TEST_BEGIN(test_hash_x86_32) { + hash_variant_verify(hash_variant_x86_32); +} +TEST_END + +TEST_BEGIN(test_hash_x86_128) { + hash_variant_verify(hash_variant_x86_128); +} +TEST_END + +TEST_BEGIN(test_hash_x64_128) { + hash_variant_verify(hash_variant_x64_128); +} +TEST_END + +int +main(void) { + return test( + test_hash_x86_32, + test_hash_x86_128, + test_hash_x64_128); +} diff --git a/deps/jemalloc/test/unit/hook.c b/deps/jemalloc/test/unit/hook.c new file mode 100644 index 000000000..72fcc433c --- /dev/null +++ b/deps/jemalloc/test/unit/hook.c @@ -0,0 +1,580 @@ +#include "test/jemalloc_test.h" + +#include "jemalloc/internal/hook.h" + +static void *arg_extra; +static int arg_type; +static void *arg_result; +static void *arg_address; +static size_t arg_old_usize; +static size_t arg_new_usize; +static uintptr_t arg_result_raw; +static uintptr_t arg_args_raw[4]; + +static int call_count = 0; + +static void +reset_args() { + arg_extra = NULL; + arg_type = 12345; + arg_result = NULL; + arg_address = NULL; + arg_old_usize = 0; + arg_new_usize = 0; + arg_result_raw = 0; + memset(arg_args_raw, 77, sizeof(arg_args_raw)); +} + +static void +alloc_free_size(size_t sz) { + void *ptr = mallocx(1, 0); + free(ptr); + ptr = mallocx(1, 0); + free(ptr); + ptr = mallocx(1, MALLOCX_TCACHE_NONE); + dallocx(ptr, MALLOCX_TCACHE_NONE); +} + +/* + * We want to support a degree of user reentrancy. This tests a variety of + * allocation scenarios. + */ +static void +be_reentrant() { + /* Let's make sure the tcache is non-empty if enabled. */ + alloc_free_size(1); + alloc_free_size(1024); + alloc_free_size(64 * 1024); + alloc_free_size(256 * 1024); + alloc_free_size(1024 * 1024); + + /* Some reallocation. */ + void *ptr = mallocx(129, 0); + ptr = rallocx(ptr, 130, 0); + free(ptr); + + ptr = mallocx(2 * 1024 * 1024, 0); + free(ptr); + ptr = mallocx(1 * 1024 * 1024, 0); + ptr = rallocx(ptr, 2 * 1024 * 1024, 0); + free(ptr); + + ptr = mallocx(1, 0); + ptr = rallocx(ptr, 1000, 0); + free(ptr); +} + +static void +set_args_raw(uintptr_t *args_raw, int nargs) { + memcpy(arg_args_raw, args_raw, sizeof(uintptr_t) * nargs); +} + +static void +assert_args_raw(uintptr_t *args_raw_expected, int nargs) { + int cmp = memcmp(args_raw_expected, arg_args_raw, + sizeof(uintptr_t) * nargs); + assert_d_eq(cmp, 0, "Raw args mismatch"); +} + +static void +reset() { + call_count = 0; + reset_args(); +} + +static void +test_alloc_hook(void *extra, hook_alloc_t type, void *result, + uintptr_t result_raw, uintptr_t args_raw[3]) { + call_count++; + arg_extra = extra; + arg_type = (int)type; + arg_result = result; + arg_result_raw = result_raw; + set_args_raw(args_raw, 3); + be_reentrant(); +} + +static void +test_dalloc_hook(void *extra, hook_dalloc_t type, void *address, + uintptr_t args_raw[3]) { + call_count++; + arg_extra = extra; + arg_type = (int)type; + arg_address = address; + set_args_raw(args_raw, 3); + be_reentrant(); +} + +static void +test_expand_hook(void *extra, hook_expand_t type, void *address, + size_t old_usize, size_t new_usize, uintptr_t result_raw, + uintptr_t args_raw[4]) { + call_count++; + arg_extra = extra; + arg_type = (int)type; + arg_address = address; + arg_old_usize = old_usize; + arg_new_usize = new_usize; + arg_result_raw = result_raw; + set_args_raw(args_raw, 4); + be_reentrant(); +} + +TEST_BEGIN(test_hooks_basic) { + /* Just verify that the record their arguments correctly. */ + hooks_t hooks = { + &test_alloc_hook, &test_dalloc_hook, &test_expand_hook, + (void *)111}; + void *handle = hook_install(TSDN_NULL, &hooks); + uintptr_t args_raw[4] = {10, 20, 30, 40}; + + /* Alloc */ + reset_args(); + hook_invoke_alloc(hook_alloc_posix_memalign, (void *)222, 333, + args_raw); + assert_ptr_eq(arg_extra, (void *)111, "Passed wrong user pointer"); + assert_d_eq((int)hook_alloc_posix_memalign, arg_type, + "Passed wrong alloc type"); + assert_ptr_eq((void *)222, arg_result, "Passed wrong result address"); + assert_u64_eq(333, arg_result_raw, "Passed wrong result"); + assert_args_raw(args_raw, 3); + + /* Dalloc */ + reset_args(); + hook_invoke_dalloc(hook_dalloc_sdallocx, (void *)222, args_raw); + assert_d_eq((int)hook_dalloc_sdallocx, arg_type, + "Passed wrong dalloc type"); + assert_ptr_eq((void *)111, arg_extra, "Passed wrong user pointer"); + assert_ptr_eq((void *)222, arg_address, "Passed wrong address"); + assert_args_raw(args_raw, 3); + + /* Expand */ + reset_args(); + hook_invoke_expand(hook_expand_xallocx, (void *)222, 333, 444, 555, + args_raw); + assert_d_eq((int)hook_expand_xallocx, arg_type, + "Passed wrong expand type"); + assert_ptr_eq((void *)111, arg_extra, "Passed wrong user pointer"); + assert_ptr_eq((void *)222, arg_address, "Passed wrong address"); + assert_zu_eq(333, arg_old_usize, "Passed wrong old usize"); + assert_zu_eq(444, arg_new_usize, "Passed wrong new usize"); + assert_zu_eq(555, arg_result_raw, "Passed wrong result"); + assert_args_raw(args_raw, 4); + + hook_remove(TSDN_NULL, handle); +} +TEST_END + +TEST_BEGIN(test_hooks_null) { + /* Null hooks should be ignored, not crash. */ + hooks_t hooks1 = {NULL, NULL, NULL, NULL}; + hooks_t hooks2 = {&test_alloc_hook, NULL, NULL, NULL}; + hooks_t hooks3 = {NULL, &test_dalloc_hook, NULL, NULL}; + hooks_t hooks4 = {NULL, NULL, &test_expand_hook, NULL}; + + void *handle1 = hook_install(TSDN_NULL, &hooks1); + void *handle2 = hook_install(TSDN_NULL, &hooks2); + void *handle3 = hook_install(TSDN_NULL, &hooks3); + void *handle4 = hook_install(TSDN_NULL, &hooks4); + + assert_ptr_ne(handle1, NULL, "Hook installation failed"); + assert_ptr_ne(handle2, NULL, "Hook installation failed"); + assert_ptr_ne(handle3, NULL, "Hook installation failed"); + assert_ptr_ne(handle4, NULL, "Hook installation failed"); + + uintptr_t args_raw[4] = {10, 20, 30, 40}; + + call_count = 0; + hook_invoke_alloc(hook_alloc_malloc, NULL, 0, args_raw); + assert_d_eq(call_count, 1, "Called wrong number of times"); + + call_count = 0; + hook_invoke_dalloc(hook_dalloc_free, NULL, args_raw); + assert_d_eq(call_count, 1, "Called wrong number of times"); + + call_count = 0; + hook_invoke_expand(hook_expand_realloc, NULL, 0, 0, 0, args_raw); + assert_d_eq(call_count, 1, "Called wrong number of times"); + + hook_remove(TSDN_NULL, handle1); + hook_remove(TSDN_NULL, handle2); + hook_remove(TSDN_NULL, handle3); + hook_remove(TSDN_NULL, handle4); +} +TEST_END + +TEST_BEGIN(test_hooks_remove) { + hooks_t hooks = {&test_alloc_hook, NULL, NULL, NULL}; + void *handle = hook_install(TSDN_NULL, &hooks); + assert_ptr_ne(handle, NULL, "Hook installation failed"); + call_count = 0; + uintptr_t args_raw[4] = {10, 20, 30, 40}; + hook_invoke_alloc(hook_alloc_malloc, NULL, 0, args_raw); + assert_d_eq(call_count, 1, "Hook not invoked"); + + call_count = 0; + hook_remove(TSDN_NULL, handle); + hook_invoke_alloc(hook_alloc_malloc, NULL, 0, NULL); + assert_d_eq(call_count, 0, "Hook invoked after removal"); + +} +TEST_END + +TEST_BEGIN(test_hooks_alloc_simple) { + /* "Simple" in the sense that we're not in a realloc variant. */ + hooks_t hooks = {&test_alloc_hook, NULL, NULL, (void *)123}; + void *handle = hook_install(TSDN_NULL, &hooks); + assert_ptr_ne(handle, NULL, "Hook installation failed"); + + /* Stop malloc from being optimized away. */ + volatile int err; + void *volatile ptr; + + /* malloc */ + reset(); + ptr = malloc(1); + assert_d_eq(call_count, 1, "Hook not called"); + assert_ptr_eq(arg_extra, (void *)123, "Wrong extra"); + assert_d_eq(arg_type, (int)hook_alloc_malloc, "Wrong hook type"); + assert_ptr_eq(ptr, arg_result, "Wrong result"); + assert_u64_eq((uintptr_t)ptr, (uintptr_t)arg_result_raw, + "Wrong raw result"); + assert_u64_eq((uintptr_t)1, arg_args_raw[0], "Wrong argument"); + free(ptr); + + /* posix_memalign */ + reset(); + err = posix_memalign((void **)&ptr, 1024, 1); + assert_d_eq(call_count, 1, "Hook not called"); + assert_ptr_eq(arg_extra, (void *)123, "Wrong extra"); + assert_d_eq(arg_type, (int)hook_alloc_posix_memalign, + "Wrong hook type"); + assert_ptr_eq(ptr, arg_result, "Wrong result"); + assert_u64_eq((uintptr_t)err, (uintptr_t)arg_result_raw, + "Wrong raw result"); + assert_u64_eq((uintptr_t)&ptr, arg_args_raw[0], "Wrong argument"); + assert_u64_eq((uintptr_t)1024, arg_args_raw[1], "Wrong argument"); + assert_u64_eq((uintptr_t)1, arg_args_raw[2], "Wrong argument"); + free(ptr); + + /* aligned_alloc */ + reset(); + ptr = aligned_alloc(1024, 1); + assert_d_eq(call_count, 1, "Hook not called"); + assert_ptr_eq(arg_extra, (void *)123, "Wrong extra"); + assert_d_eq(arg_type, (int)hook_alloc_aligned_alloc, + "Wrong hook type"); + assert_ptr_eq(ptr, arg_result, "Wrong result"); + assert_u64_eq((uintptr_t)ptr, (uintptr_t)arg_result_raw, + "Wrong raw result"); + assert_u64_eq((uintptr_t)1024, arg_args_raw[0], "Wrong argument"); + assert_u64_eq((uintptr_t)1, arg_args_raw[1], "Wrong argument"); + free(ptr); + + /* calloc */ + reset(); + ptr = calloc(11, 13); + assert_d_eq(call_count, 1, "Hook not called"); + assert_ptr_eq(arg_extra, (void *)123, "Wrong extra"); + assert_d_eq(arg_type, (int)hook_alloc_calloc, "Wrong hook type"); + assert_ptr_eq(ptr, arg_result, "Wrong result"); + assert_u64_eq((uintptr_t)ptr, (uintptr_t)arg_result_raw, + "Wrong raw result"); + assert_u64_eq((uintptr_t)11, arg_args_raw[0], "Wrong argument"); + assert_u64_eq((uintptr_t)13, arg_args_raw[1], "Wrong argument"); + free(ptr); + + /* memalign */ +#ifdef JEMALLOC_OVERRIDE_MEMALIGN + reset(); + ptr = memalign(1024, 1); + assert_d_eq(call_count, 1, "Hook not called"); + assert_ptr_eq(arg_extra, (void *)123, "Wrong extra"); + assert_d_eq(arg_type, (int)hook_alloc_memalign, "Wrong hook type"); + assert_ptr_eq(ptr, arg_result, "Wrong result"); + assert_u64_eq((uintptr_t)ptr, (uintptr_t)arg_result_raw, + "Wrong raw result"); + assert_u64_eq((uintptr_t)1024, arg_args_raw[0], "Wrong argument"); + assert_u64_eq((uintptr_t)1, arg_args_raw[1], "Wrong argument"); + free(ptr); +#endif /* JEMALLOC_OVERRIDE_MEMALIGN */ + + /* valloc */ +#ifdef JEMALLOC_OVERRIDE_VALLOC + reset(); + ptr = valloc(1); + assert_d_eq(call_count, 1, "Hook not called"); + assert_ptr_eq(arg_extra, (void *)123, "Wrong extra"); + assert_d_eq(arg_type, (int)hook_alloc_valloc, "Wrong hook type"); + assert_ptr_eq(ptr, arg_result, "Wrong result"); + assert_u64_eq((uintptr_t)ptr, (uintptr_t)arg_result_raw, + "Wrong raw result"); + assert_u64_eq((uintptr_t)1, arg_args_raw[0], "Wrong argument"); + free(ptr); +#endif /* JEMALLOC_OVERRIDE_VALLOC */ + + /* mallocx */ + reset(); + ptr = mallocx(1, MALLOCX_LG_ALIGN(10)); + assert_d_eq(call_count, 1, "Hook not called"); + assert_ptr_eq(arg_extra, (void *)123, "Wrong extra"); + assert_d_eq(arg_type, (int)hook_alloc_mallocx, "Wrong hook type"); + assert_ptr_eq(ptr, arg_result, "Wrong result"); + assert_u64_eq((uintptr_t)ptr, (uintptr_t)arg_result_raw, + "Wrong raw result"); + assert_u64_eq((uintptr_t)1, arg_args_raw[0], "Wrong argument"); + assert_u64_eq((uintptr_t)MALLOCX_LG_ALIGN(10), arg_args_raw[1], + "Wrong flags"); + free(ptr); + + hook_remove(TSDN_NULL, handle); +} +TEST_END + +TEST_BEGIN(test_hooks_dalloc_simple) { + /* "Simple" in the sense that we're not in a realloc variant. */ + hooks_t hooks = {NULL, &test_dalloc_hook, NULL, (void *)123}; + void *handle = hook_install(TSDN_NULL, &hooks); + assert_ptr_ne(handle, NULL, "Hook installation failed"); + + void *volatile ptr; + + /* free() */ + reset(); + ptr = malloc(1); + free(ptr); + assert_d_eq(call_count, 1, "Hook not called"); + assert_ptr_eq(arg_extra, (void *)123, "Wrong extra"); + assert_d_eq(arg_type, (int)hook_dalloc_free, "Wrong hook type"); + assert_ptr_eq(ptr, arg_address, "Wrong pointer freed"); + assert_u64_eq((uintptr_t)ptr, arg_args_raw[0], "Wrong raw arg"); + + /* dallocx() */ + reset(); + ptr = malloc(1); + dallocx(ptr, MALLOCX_TCACHE_NONE); + assert_d_eq(call_count, 1, "Hook not called"); + assert_ptr_eq(arg_extra, (void *)123, "Wrong extra"); + assert_d_eq(arg_type, (int)hook_dalloc_dallocx, "Wrong hook type"); + assert_ptr_eq(ptr, arg_address, "Wrong pointer freed"); + assert_u64_eq((uintptr_t)ptr, arg_args_raw[0], "Wrong raw arg"); + assert_u64_eq((uintptr_t)MALLOCX_TCACHE_NONE, arg_args_raw[1], + "Wrong raw arg"); + + /* sdallocx() */ + reset(); + ptr = malloc(1); + sdallocx(ptr, 1, MALLOCX_TCACHE_NONE); + assert_d_eq(call_count, 1, "Hook not called"); + assert_ptr_eq(arg_extra, (void *)123, "Wrong extra"); + assert_d_eq(arg_type, (int)hook_dalloc_sdallocx, "Wrong hook type"); + assert_ptr_eq(ptr, arg_address, "Wrong pointer freed"); + assert_u64_eq((uintptr_t)ptr, arg_args_raw[0], "Wrong raw arg"); + assert_u64_eq((uintptr_t)1, arg_args_raw[1], "Wrong raw arg"); + assert_u64_eq((uintptr_t)MALLOCX_TCACHE_NONE, arg_args_raw[2], + "Wrong raw arg"); + + hook_remove(TSDN_NULL, handle); +} +TEST_END + +TEST_BEGIN(test_hooks_expand_simple) { + /* "Simple" in the sense that we're not in a realloc variant. */ + hooks_t hooks = {NULL, NULL, &test_expand_hook, (void *)123}; + void *handle = hook_install(TSDN_NULL, &hooks); + assert_ptr_ne(handle, NULL, "Hook installation failed"); + + void *volatile ptr; + + /* xallocx() */ + reset(); + ptr = malloc(1); + size_t new_usize = xallocx(ptr, 100, 200, MALLOCX_TCACHE_NONE); + assert_d_eq(call_count, 1, "Hook not called"); + assert_ptr_eq(arg_extra, (void *)123, "Wrong extra"); + assert_d_eq(arg_type, (int)hook_expand_xallocx, "Wrong hook type"); + assert_ptr_eq(ptr, arg_address, "Wrong pointer expanded"); + assert_u64_eq(arg_old_usize, nallocx(1, 0), "Wrong old usize"); + assert_u64_eq(arg_new_usize, sallocx(ptr, 0), "Wrong new usize"); + assert_u64_eq(new_usize, arg_result_raw, "Wrong result"); + assert_u64_eq((uintptr_t)ptr, arg_args_raw[0], "Wrong arg"); + assert_u64_eq(100, arg_args_raw[1], "Wrong arg"); + assert_u64_eq(200, arg_args_raw[2], "Wrong arg"); + assert_u64_eq(MALLOCX_TCACHE_NONE, arg_args_raw[3], "Wrong arg"); + + hook_remove(TSDN_NULL, handle); +} +TEST_END + +TEST_BEGIN(test_hooks_realloc_as_malloc_or_free) { + hooks_t hooks = {&test_alloc_hook, &test_dalloc_hook, + &test_expand_hook, (void *)123}; + void *handle = hook_install(TSDN_NULL, &hooks); + assert_ptr_ne(handle, NULL, "Hook installation failed"); + + void *volatile ptr; + + /* realloc(NULL, size) as malloc */ + reset(); + ptr = realloc(NULL, 1); + assert_d_eq(call_count, 1, "Hook not called"); + assert_ptr_eq(arg_extra, (void *)123, "Wrong extra"); + assert_d_eq(arg_type, (int)hook_alloc_realloc, "Wrong hook type"); + assert_ptr_eq(ptr, arg_result, "Wrong result"); + assert_u64_eq((uintptr_t)ptr, (uintptr_t)arg_result_raw, + "Wrong raw result"); + assert_u64_eq((uintptr_t)NULL, arg_args_raw[0], "Wrong argument"); + assert_u64_eq((uintptr_t)1, arg_args_raw[1], "Wrong argument"); + free(ptr); + + /* realloc(ptr, 0) as free */ + ptr = malloc(1); + reset(); + realloc(ptr, 0); + assert_d_eq(call_count, 1, "Hook not called"); + assert_ptr_eq(arg_extra, (void *)123, "Wrong extra"); + assert_d_eq(arg_type, (int)hook_dalloc_realloc, "Wrong hook type"); + assert_ptr_eq(ptr, arg_address, "Wrong pointer freed"); + assert_u64_eq((uintptr_t)ptr, arg_args_raw[0], "Wrong raw arg"); + assert_u64_eq((uintptr_t)0, arg_args_raw[1], "Wrong raw arg"); + + /* realloc(NULL, 0) as malloc(0) */ + reset(); + ptr = realloc(NULL, 0); + assert_d_eq(call_count, 1, "Hook not called"); + assert_ptr_eq(arg_extra, (void *)123, "Wrong extra"); + assert_d_eq(arg_type, (int)hook_alloc_realloc, "Wrong hook type"); + assert_ptr_eq(ptr, arg_result, "Wrong result"); + assert_u64_eq((uintptr_t)ptr, (uintptr_t)arg_result_raw, + "Wrong raw result"); + assert_u64_eq((uintptr_t)NULL, arg_args_raw[0], "Wrong argument"); + assert_u64_eq((uintptr_t)0, arg_args_raw[1], "Wrong argument"); + free(ptr); + + hook_remove(TSDN_NULL, handle); +} +TEST_END + +static void +do_realloc_test(void *(*ralloc)(void *, size_t, int), int flags, + int expand_type, int dalloc_type) { + hooks_t hooks = {&test_alloc_hook, &test_dalloc_hook, + &test_expand_hook, (void *)123}; + void *handle = hook_install(TSDN_NULL, &hooks); + assert_ptr_ne(handle, NULL, "Hook installation failed"); + + void *volatile ptr; + void *volatile ptr2; + + /* Realloc in-place, small. */ + ptr = malloc(129); + reset(); + ptr2 = ralloc(ptr, 130, flags); + assert_ptr_eq(ptr, ptr2, "Small realloc moved"); + + assert_d_eq(call_count, 1, "Hook not called"); + assert_ptr_eq(arg_extra, (void *)123, "Wrong extra"); + assert_d_eq(arg_type, expand_type, "Wrong hook type"); + assert_ptr_eq(ptr, arg_address, "Wrong address"); + assert_u64_eq((uintptr_t)ptr, (uintptr_t)arg_result_raw, + "Wrong raw result"); + assert_u64_eq((uintptr_t)ptr, arg_args_raw[0], "Wrong argument"); + assert_u64_eq((uintptr_t)130, arg_args_raw[1], "Wrong argument"); + free(ptr); + + /* + * Realloc in-place, large. Since we can't guarantee the large case + * across all platforms, we stay resilient to moving results. + */ + ptr = malloc(2 * 1024 * 1024); + free(ptr); + ptr2 = malloc(1 * 1024 * 1024); + reset(); + ptr = ralloc(ptr2, 2 * 1024 * 1024, flags); + /* ptr is the new address, ptr2 is the old address. */ + if (ptr == ptr2) { + assert_d_eq(call_count, 1, "Hook not called"); + assert_d_eq(arg_type, expand_type, "Wrong hook type"); + } else { + assert_d_eq(call_count, 2, "Wrong hooks called"); + assert_ptr_eq(ptr, arg_result, "Wrong address"); + assert_d_eq(arg_type, dalloc_type, "Wrong hook type"); + } + assert_ptr_eq(arg_extra, (void *)123, "Wrong extra"); + assert_ptr_eq(ptr2, arg_address, "Wrong address"); + assert_u64_eq((uintptr_t)ptr, (uintptr_t)arg_result_raw, + "Wrong raw result"); + assert_u64_eq((uintptr_t)ptr2, arg_args_raw[0], "Wrong argument"); + assert_u64_eq((uintptr_t)2 * 1024 * 1024, arg_args_raw[1], + "Wrong argument"); + free(ptr); + + /* Realloc with move, small. */ + ptr = malloc(8); + reset(); + ptr2 = ralloc(ptr, 128, flags); + assert_ptr_ne(ptr, ptr2, "Small realloc didn't move"); + + assert_d_eq(call_count, 2, "Hook not called"); + assert_ptr_eq(arg_extra, (void *)123, "Wrong extra"); + assert_d_eq(arg_type, dalloc_type, "Wrong hook type"); + assert_ptr_eq(ptr, arg_address, "Wrong address"); + assert_ptr_eq(ptr2, arg_result, "Wrong address"); + assert_u64_eq((uintptr_t)ptr2, (uintptr_t)arg_result_raw, + "Wrong raw result"); + assert_u64_eq((uintptr_t)ptr, arg_args_raw[0], "Wrong argument"); + assert_u64_eq((uintptr_t)128, arg_args_raw[1], "Wrong argument"); + free(ptr2); + + /* Realloc with move, large. */ + ptr = malloc(1); + reset(); + ptr2 = ralloc(ptr, 2 * 1024 * 1024, flags); + assert_ptr_ne(ptr, ptr2, "Large realloc didn't move"); + + assert_d_eq(call_count, 2, "Hook not called"); + assert_ptr_eq(arg_extra, (void *)123, "Wrong extra"); + assert_d_eq(arg_type, dalloc_type, "Wrong hook type"); + assert_ptr_eq(ptr, arg_address, "Wrong address"); + assert_ptr_eq(ptr2, arg_result, "Wrong address"); + assert_u64_eq((uintptr_t)ptr2, (uintptr_t)arg_result_raw, + "Wrong raw result"); + assert_u64_eq((uintptr_t)ptr, arg_args_raw[0], "Wrong argument"); + assert_u64_eq((uintptr_t)2 * 1024 * 1024, arg_args_raw[1], + "Wrong argument"); + free(ptr2); + + hook_remove(TSDN_NULL, handle); +} + +static void * +realloc_wrapper(void *ptr, size_t size, UNUSED int flags) { + return realloc(ptr, size); +} + +TEST_BEGIN(test_hooks_realloc) { + do_realloc_test(&realloc_wrapper, 0, hook_expand_realloc, + hook_dalloc_realloc); +} +TEST_END + +TEST_BEGIN(test_hooks_rallocx) { + do_realloc_test(&rallocx, MALLOCX_TCACHE_NONE, hook_expand_rallocx, + hook_dalloc_rallocx); +} +TEST_END + +int +main(void) { + /* We assert on call counts. */ + return test_no_reentrancy( + test_hooks_basic, + test_hooks_null, + test_hooks_remove, + test_hooks_alloc_simple, + test_hooks_dalloc_simple, + test_hooks_expand_simple, + test_hooks_realloc_as_malloc_or_free, + test_hooks_realloc, + test_hooks_rallocx); +} diff --git a/deps/jemalloc/test/unit/huge.c b/deps/jemalloc/test/unit/huge.c new file mode 100644 index 000000000..ab72cf007 --- /dev/null +++ b/deps/jemalloc/test/unit/huge.c @@ -0,0 +1,108 @@ +#include "test/jemalloc_test.h" + +/* Threshold: 2 << 20 = 2097152. */ +const char *malloc_conf = "oversize_threshold:2097152"; + +#define HUGE_SZ (2 << 20) +#define SMALL_SZ (8) + +TEST_BEGIN(huge_bind_thread) { + unsigned arena1, arena2; + size_t sz = sizeof(unsigned); + + /* Bind to a manual arena. */ + assert_d_eq(mallctl("arenas.create", &arena1, &sz, NULL, 0), 0, + "Failed to create arena"); + assert_d_eq(mallctl("thread.arena", NULL, NULL, &arena1, + sizeof(arena1)), 0, "Fail to bind thread"); + + void *ptr = mallocx(HUGE_SZ, 0); + assert_ptr_not_null(ptr, "Fail to allocate huge size"); + assert_d_eq(mallctl("arenas.lookup", &arena2, &sz, &ptr, + sizeof(ptr)), 0, "Unexpected mallctl() failure"); + assert_u_eq(arena1, arena2, "Wrong arena used after binding"); + dallocx(ptr, 0); + + /* Switch back to arena 0. */ + test_skip_if(have_percpu_arena && + PERCPU_ARENA_ENABLED(opt_percpu_arena)); + arena2 = 0; + assert_d_eq(mallctl("thread.arena", NULL, NULL, &arena2, + sizeof(arena2)), 0, "Fail to bind thread"); + ptr = mallocx(SMALL_SZ, MALLOCX_TCACHE_NONE); + assert_d_eq(mallctl("arenas.lookup", &arena2, &sz, &ptr, + sizeof(ptr)), 0, "Unexpected mallctl() failure"); + assert_u_eq(arena2, 0, "Wrong arena used after binding"); + dallocx(ptr, MALLOCX_TCACHE_NONE); + + /* Then huge allocation should use the huge arena. */ + ptr = mallocx(HUGE_SZ, 0); + assert_ptr_not_null(ptr, "Fail to allocate huge size"); + assert_d_eq(mallctl("arenas.lookup", &arena2, &sz, &ptr, + sizeof(ptr)), 0, "Unexpected mallctl() failure"); + assert_u_ne(arena2, 0, "Wrong arena used after binding"); + assert_u_ne(arena1, arena2, "Wrong arena used after binding"); + dallocx(ptr, 0); +} +TEST_END + +TEST_BEGIN(huge_mallocx) { + unsigned arena1, arena2; + size_t sz = sizeof(unsigned); + + assert_d_eq(mallctl("arenas.create", &arena1, &sz, NULL, 0), 0, + "Failed to create arena"); + void *huge = mallocx(HUGE_SZ, MALLOCX_ARENA(arena1)); + assert_ptr_not_null(huge, "Fail to allocate huge size"); + assert_d_eq(mallctl("arenas.lookup", &arena2, &sz, &huge, + sizeof(huge)), 0, "Unexpected mallctl() failure"); + assert_u_eq(arena1, arena2, "Wrong arena used for mallocx"); + dallocx(huge, MALLOCX_ARENA(arena1)); + + void *huge2 = mallocx(HUGE_SZ, 0); + assert_ptr_not_null(huge, "Fail to allocate huge size"); + assert_d_eq(mallctl("arenas.lookup", &arena2, &sz, &huge2, + sizeof(huge2)), 0, "Unexpected mallctl() failure"); + assert_u_ne(arena1, arena2, + "Huge allocation should not come from the manual arena."); + assert_u_ne(arena2, 0, + "Huge allocation should not come from the arena 0."); + dallocx(huge2, 0); +} +TEST_END + +TEST_BEGIN(huge_allocation) { + unsigned arena1, arena2; + + void *ptr = mallocx(HUGE_SZ, 0); + assert_ptr_not_null(ptr, "Fail to allocate huge size"); + size_t sz = sizeof(unsigned); + assert_d_eq(mallctl("arenas.lookup", &arena1, &sz, &ptr, sizeof(ptr)), + 0, "Unexpected mallctl() failure"); + assert_u_gt(arena1, 0, "Huge allocation should not come from arena 0"); + dallocx(ptr, 0); + + ptr = mallocx(HUGE_SZ >> 1, 0); + assert_ptr_not_null(ptr, "Fail to allocate half huge size"); + assert_d_eq(mallctl("arenas.lookup", &arena2, &sz, &ptr, + sizeof(ptr)), 0, "Unexpected mallctl() failure"); + assert_u_ne(arena1, arena2, "Wrong arena used for half huge"); + dallocx(ptr, 0); + + ptr = mallocx(SMALL_SZ, MALLOCX_TCACHE_NONE); + assert_ptr_not_null(ptr, "Fail to allocate small size"); + assert_d_eq(mallctl("arenas.lookup", &arena2, &sz, &ptr, + sizeof(ptr)), 0, "Unexpected mallctl() failure"); + assert_u_ne(arena1, arena2, + "Huge and small should be from different arenas"); + dallocx(ptr, 0); +} +TEST_END + +int +main(void) { + return test( + huge_allocation, + huge_mallocx, + huge_bind_thread); +} diff --git a/deps/jemalloc/test/unit/junk.c b/deps/jemalloc/test/unit/junk.c new file mode 100644 index 000000000..57e3ad431 --- /dev/null +++ b/deps/jemalloc/test/unit/junk.c @@ -0,0 +1,141 @@ +#include "test/jemalloc_test.h" + +#include "jemalloc/internal/util.h" + +static arena_dalloc_junk_small_t *arena_dalloc_junk_small_orig; +static large_dalloc_junk_t *large_dalloc_junk_orig; +static large_dalloc_maybe_junk_t *large_dalloc_maybe_junk_orig; +static void *watch_for_junking; +static bool saw_junking; + +static void +watch_junking(void *p) { + watch_for_junking = p; + saw_junking = false; +} + +static void +arena_dalloc_junk_small_intercept(void *ptr, const bin_info_t *bin_info) { + size_t i; + + arena_dalloc_junk_small_orig(ptr, bin_info); + for (i = 0; i < bin_info->reg_size; i++) { + assert_u_eq(((uint8_t *)ptr)[i], JEMALLOC_FREE_JUNK, + "Missing junk fill for byte %zu/%zu of deallocated region", + i, bin_info->reg_size); + } + if (ptr == watch_for_junking) { + saw_junking = true; + } +} + +static void +large_dalloc_junk_intercept(void *ptr, size_t usize) { + size_t i; + + large_dalloc_junk_orig(ptr, usize); + for (i = 0; i < usize; i++) { + assert_u_eq(((uint8_t *)ptr)[i], JEMALLOC_FREE_JUNK, + "Missing junk fill for byte %zu/%zu of deallocated region", + i, usize); + } + if (ptr == watch_for_junking) { + saw_junking = true; + } +} + +static void +large_dalloc_maybe_junk_intercept(void *ptr, size_t usize) { + large_dalloc_maybe_junk_orig(ptr, usize); + if (ptr == watch_for_junking) { + saw_junking = true; + } +} + +static void +test_junk(size_t sz_min, size_t sz_max) { + uint8_t *s; + size_t sz_prev, sz, i; + + if (opt_junk_free) { + arena_dalloc_junk_small_orig = arena_dalloc_junk_small; + arena_dalloc_junk_small = arena_dalloc_junk_small_intercept; + large_dalloc_junk_orig = large_dalloc_junk; + large_dalloc_junk = large_dalloc_junk_intercept; + large_dalloc_maybe_junk_orig = large_dalloc_maybe_junk; + large_dalloc_maybe_junk = large_dalloc_maybe_junk_intercept; + } + + sz_prev = 0; + s = (uint8_t *)mallocx(sz_min, 0); + assert_ptr_not_null((void *)s, "Unexpected mallocx() failure"); + + for (sz = sallocx(s, 0); sz <= sz_max; + sz_prev = sz, sz = sallocx(s, 0)) { + if (sz_prev > 0) { + assert_u_eq(s[0], 'a', + "Previously allocated byte %zu/%zu is corrupted", + ZU(0), sz_prev); + assert_u_eq(s[sz_prev-1], 'a', + "Previously allocated byte %zu/%zu is corrupted", + sz_prev-1, sz_prev); + } + + for (i = sz_prev; i < sz; i++) { + if (opt_junk_alloc) { + assert_u_eq(s[i], JEMALLOC_ALLOC_JUNK, + "Newly allocated byte %zu/%zu isn't " + "junk-filled", i, sz); + } + s[i] = 'a'; + } + + if (xallocx(s, sz+1, 0, 0) == sz) { + uint8_t *t; + watch_junking(s); + t = (uint8_t *)rallocx(s, sz+1, 0); + assert_ptr_not_null((void *)t, + "Unexpected rallocx() failure"); + assert_zu_ge(sallocx(t, 0), sz+1, + "Unexpectedly small rallocx() result"); + if (!background_thread_enabled()) { + assert_ptr_ne(s, t, + "Unexpected in-place rallocx()"); + assert_true(!opt_junk_free || saw_junking, + "Expected region of size %zu to be " + "junk-filled", sz); + } + s = t; + } + } + + watch_junking(s); + dallocx(s, 0); + assert_true(!opt_junk_free || saw_junking, + "Expected region of size %zu to be junk-filled", sz); + + if (opt_junk_free) { + arena_dalloc_junk_small = arena_dalloc_junk_small_orig; + large_dalloc_junk = large_dalloc_junk_orig; + large_dalloc_maybe_junk = large_dalloc_maybe_junk_orig; + } +} + +TEST_BEGIN(test_junk_small) { + test_skip_if(!config_fill); + test_junk(1, SC_SMALL_MAXCLASS - 1); +} +TEST_END + +TEST_BEGIN(test_junk_large) { + test_skip_if(!config_fill); + test_junk(SC_SMALL_MAXCLASS + 1, (1U << (SC_LG_LARGE_MINCLASS + 1))); +} +TEST_END + +int +main(void) { + return test( + test_junk_small, + test_junk_large); +} diff --git a/deps/jemalloc/test/unit/junk.sh b/deps/jemalloc/test/unit/junk.sh new file mode 100644 index 000000000..97cd8ca5e --- /dev/null +++ b/deps/jemalloc/test/unit/junk.sh @@ -0,0 +1,5 @@ +#!/bin/sh + +if [ "x${enable_fill}" = "x1" ] ; then + export MALLOC_CONF="abort:false,zero:false,junk:true" +fi diff --git a/deps/jemalloc/test/unit/junk_alloc.c b/deps/jemalloc/test/unit/junk_alloc.c new file mode 100644 index 000000000..a442a0ca5 --- /dev/null +++ b/deps/jemalloc/test/unit/junk_alloc.c @@ -0,0 +1 @@ +#include "junk.c" diff --git a/deps/jemalloc/test/unit/junk_alloc.sh b/deps/jemalloc/test/unit/junk_alloc.sh new file mode 100644 index 000000000..e1008c2e1 --- /dev/null +++ b/deps/jemalloc/test/unit/junk_alloc.sh @@ -0,0 +1,5 @@ +#!/bin/sh + +if [ "x${enable_fill}" = "x1" ] ; then + export MALLOC_CONF="abort:false,zero:false,junk:alloc" +fi diff --git a/deps/jemalloc/test/unit/junk_free.c b/deps/jemalloc/test/unit/junk_free.c new file mode 100644 index 000000000..a442a0ca5 --- /dev/null +++ b/deps/jemalloc/test/unit/junk_free.c @@ -0,0 +1 @@ +#include "junk.c" diff --git a/deps/jemalloc/test/unit/junk_free.sh b/deps/jemalloc/test/unit/junk_free.sh new file mode 100644 index 000000000..402196ca6 --- /dev/null +++ b/deps/jemalloc/test/unit/junk_free.sh @@ -0,0 +1,5 @@ +#!/bin/sh + +if [ "x${enable_fill}" = "x1" ] ; then + export MALLOC_CONF="abort:false,zero:false,junk:free" +fi diff --git a/deps/jemalloc/test/unit/log.c b/deps/jemalloc/test/unit/log.c new file mode 100644 index 000000000..a52bd737d --- /dev/null +++ b/deps/jemalloc/test/unit/log.c @@ -0,0 +1,193 @@ +#include "test/jemalloc_test.h" + +#include "jemalloc/internal/log.h" + +static void +expect_no_logging(const char *names) { + log_var_t log_l1 = LOG_VAR_INIT("l1"); + log_var_t log_l2 = LOG_VAR_INIT("l2"); + log_var_t log_l2_a = LOG_VAR_INIT("l2.a"); + + strcpy(log_var_names, names); + + int count = 0; + + for (int i = 0; i < 10; i++) { + log_do_begin(log_l1) + count++; + log_do_end(log_l1) + + log_do_begin(log_l2) + count++; + log_do_end(log_l2) + + log_do_begin(log_l2_a) + count++; + log_do_end(log_l2_a) + } + assert_d_eq(count, 0, "Disabled logging not ignored!"); +} + +TEST_BEGIN(test_log_disabled) { + test_skip_if(!config_log); + atomic_store_b(&log_init_done, true, ATOMIC_RELAXED); + expect_no_logging(""); + expect_no_logging("abc"); + expect_no_logging("a.b.c"); + expect_no_logging("l12"); + expect_no_logging("l123|a456|b789"); + expect_no_logging("|||"); +} +TEST_END + +TEST_BEGIN(test_log_enabled_direct) { + test_skip_if(!config_log); + atomic_store_b(&log_init_done, true, ATOMIC_RELAXED); + log_var_t log_l1 = LOG_VAR_INIT("l1"); + log_var_t log_l1_a = LOG_VAR_INIT("l1.a"); + log_var_t log_l2 = LOG_VAR_INIT("l2"); + + int count; + + count = 0; + strcpy(log_var_names, "l1"); + for (int i = 0; i < 10; i++) { + log_do_begin(log_l1) + count++; + log_do_end(log_l1) + } + assert_d_eq(count, 10, "Mis-logged!"); + + count = 0; + strcpy(log_var_names, "l1.a"); + for (int i = 0; i < 10; i++) { + log_do_begin(log_l1_a) + count++; + log_do_end(log_l1_a) + } + assert_d_eq(count, 10, "Mis-logged!"); + + count = 0; + strcpy(log_var_names, "l1.a|abc|l2|def"); + for (int i = 0; i < 10; i++) { + log_do_begin(log_l1_a) + count++; + log_do_end(log_l1_a) + + log_do_begin(log_l2) + count++; + log_do_end(log_l2) + } + assert_d_eq(count, 20, "Mis-logged!"); +} +TEST_END + +TEST_BEGIN(test_log_enabled_indirect) { + test_skip_if(!config_log); + atomic_store_b(&log_init_done, true, ATOMIC_RELAXED); + strcpy(log_var_names, "l0|l1|abc|l2.b|def"); + + /* On. */ + log_var_t log_l1 = LOG_VAR_INIT("l1"); + /* Off. */ + log_var_t log_l1a = LOG_VAR_INIT("l1a"); + /* On. */ + log_var_t log_l1_a = LOG_VAR_INIT("l1.a"); + /* Off. */ + log_var_t log_l2_a = LOG_VAR_INIT("l2.a"); + /* On. */ + log_var_t log_l2_b_a = LOG_VAR_INIT("l2.b.a"); + /* On. */ + log_var_t log_l2_b_b = LOG_VAR_INIT("l2.b.b"); + + /* 4 are on total, so should sum to 40. */ + int count = 0; + for (int i = 0; i < 10; i++) { + log_do_begin(log_l1) + count++; + log_do_end(log_l1) + + log_do_begin(log_l1a) + count++; + log_do_end(log_l1a) + + log_do_begin(log_l1_a) + count++; + log_do_end(log_l1_a) + + log_do_begin(log_l2_a) + count++; + log_do_end(log_l2_a) + + log_do_begin(log_l2_b_a) + count++; + log_do_end(log_l2_b_a) + + log_do_begin(log_l2_b_b) + count++; + log_do_end(log_l2_b_b) + } + + assert_d_eq(count, 40, "Mis-logged!"); +} +TEST_END + +TEST_BEGIN(test_log_enabled_global) { + test_skip_if(!config_log); + atomic_store_b(&log_init_done, true, ATOMIC_RELAXED); + strcpy(log_var_names, "abc|.|def"); + + log_var_t log_l1 = LOG_VAR_INIT("l1"); + log_var_t log_l2_a_a = LOG_VAR_INIT("l2.a.a"); + + int count = 0; + for (int i = 0; i < 10; i++) { + log_do_begin(log_l1) + count++; + log_do_end(log_l1) + + log_do_begin(log_l2_a_a) + count++; + log_do_end(log_l2_a_a) + } + assert_d_eq(count, 20, "Mis-logged!"); +} +TEST_END + +TEST_BEGIN(test_logs_if_no_init) { + test_skip_if(!config_log); + atomic_store_b(&log_init_done, false, ATOMIC_RELAXED); + + log_var_t l = LOG_VAR_INIT("definitely.not.enabled"); + + int count = 0; + for (int i = 0; i < 10; i++) { + log_do_begin(l) + count++; + log_do_end(l) + } + assert_d_eq(count, 0, "Logging shouldn't happen if not initialized."); +} +TEST_END + +/* + * This really just checks to make sure that this usage compiles; we don't have + * any test code to run. + */ +TEST_BEGIN(test_log_only_format_string) { + if (false) { + LOG("log_str", "No arguments follow this format string."); + } +} +TEST_END + +int +main(void) { + return test( + test_log_disabled, + test_log_enabled_direct, + test_log_enabled_indirect, + test_log_enabled_global, + test_logs_if_no_init, + test_log_only_format_string); +} diff --git a/deps/jemalloc/test/unit/mallctl.c b/deps/jemalloc/test/unit/mallctl.c new file mode 100644 index 000000000..3a75ac040 --- /dev/null +++ b/deps/jemalloc/test/unit/mallctl.c @@ -0,0 +1,888 @@ +#include "test/jemalloc_test.h" + +#include "jemalloc/internal/hook.h" +#include "jemalloc/internal/util.h" + +TEST_BEGIN(test_mallctl_errors) { + uint64_t epoch; + size_t sz; + + assert_d_eq(mallctl("no_such_name", NULL, NULL, NULL, 0), ENOENT, + "mallctl() should return ENOENT for non-existent names"); + + assert_d_eq(mallctl("version", NULL, NULL, "0.0.0", strlen("0.0.0")), + EPERM, "mallctl() should return EPERM on attempt to write " + "read-only value"); + + assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch, + sizeof(epoch)-1), EINVAL, + "mallctl() should return EINVAL for input size mismatch"); + assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch, + sizeof(epoch)+1), EINVAL, + "mallctl() should return EINVAL for input size mismatch"); + + sz = sizeof(epoch)-1; + assert_d_eq(mallctl("epoch", (void *)&epoch, &sz, NULL, 0), EINVAL, + "mallctl() should return EINVAL for output size mismatch"); + sz = sizeof(epoch)+1; + assert_d_eq(mallctl("epoch", (void *)&epoch, &sz, NULL, 0), EINVAL, + "mallctl() should return EINVAL for output size mismatch"); +} +TEST_END + +TEST_BEGIN(test_mallctlnametomib_errors) { + size_t mib[1]; + size_t miblen; + + miblen = sizeof(mib)/sizeof(size_t); + assert_d_eq(mallctlnametomib("no_such_name", mib, &miblen), ENOENT, + "mallctlnametomib() should return ENOENT for non-existent names"); +} +TEST_END + +TEST_BEGIN(test_mallctlbymib_errors) { + uint64_t epoch; + size_t sz; + size_t mib[1]; + size_t miblen; + + miblen = sizeof(mib)/sizeof(size_t); + assert_d_eq(mallctlnametomib("version", mib, &miblen), 0, + "Unexpected mallctlnametomib() failure"); + + assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, "0.0.0", + strlen("0.0.0")), EPERM, "mallctl() should return EPERM on " + "attempt to write read-only value"); + + miblen = sizeof(mib)/sizeof(size_t); + assert_d_eq(mallctlnametomib("epoch", mib, &miblen), 0, + "Unexpected mallctlnametomib() failure"); + + assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, (void *)&epoch, + sizeof(epoch)-1), EINVAL, + "mallctlbymib() should return EINVAL for input size mismatch"); + assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, (void *)&epoch, + sizeof(epoch)+1), EINVAL, + "mallctlbymib() should return EINVAL for input size mismatch"); + + sz = sizeof(epoch)-1; + assert_d_eq(mallctlbymib(mib, miblen, (void *)&epoch, &sz, NULL, 0), + EINVAL, + "mallctlbymib() should return EINVAL for output size mismatch"); + sz = sizeof(epoch)+1; + assert_d_eq(mallctlbymib(mib, miblen, (void *)&epoch, &sz, NULL, 0), + EINVAL, + "mallctlbymib() should return EINVAL for output size mismatch"); +} +TEST_END + +TEST_BEGIN(test_mallctl_read_write) { + uint64_t old_epoch, new_epoch; + size_t sz = sizeof(old_epoch); + + /* Blind. */ + assert_d_eq(mallctl("epoch", NULL, NULL, NULL, 0), 0, + "Unexpected mallctl() failure"); + assert_zu_eq(sz, sizeof(old_epoch), "Unexpected output size"); + + /* Read. */ + assert_d_eq(mallctl("epoch", (void *)&old_epoch, &sz, NULL, 0), 0, + "Unexpected mallctl() failure"); + assert_zu_eq(sz, sizeof(old_epoch), "Unexpected output size"); + + /* Write. */ + assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&new_epoch, + sizeof(new_epoch)), 0, "Unexpected mallctl() failure"); + assert_zu_eq(sz, sizeof(old_epoch), "Unexpected output size"); + + /* Read+write. */ + assert_d_eq(mallctl("epoch", (void *)&old_epoch, &sz, + (void *)&new_epoch, sizeof(new_epoch)), 0, + "Unexpected mallctl() failure"); + assert_zu_eq(sz, sizeof(old_epoch), "Unexpected output size"); +} +TEST_END + +TEST_BEGIN(test_mallctlnametomib_short_mib) { + size_t mib[4]; + size_t miblen; + + miblen = 3; + mib[3] = 42; + assert_d_eq(mallctlnametomib("arenas.bin.0.nregs", mib, &miblen), 0, + "Unexpected mallctlnametomib() failure"); + assert_zu_eq(miblen, 3, "Unexpected mib output length"); + assert_zu_eq(mib[3], 42, + "mallctlnametomib() wrote past the end of the input mib"); +} +TEST_END + +TEST_BEGIN(test_mallctl_config) { +#define TEST_MALLCTL_CONFIG(config, t) do { \ + t oldval; \ + size_t sz = sizeof(oldval); \ + assert_d_eq(mallctl("config."#config, (void *)&oldval, &sz, \ + NULL, 0), 0, "Unexpected mallctl() failure"); \ + assert_b_eq(oldval, config_##config, "Incorrect config value"); \ + assert_zu_eq(sz, sizeof(oldval), "Unexpected output size"); \ +} while (0) + + TEST_MALLCTL_CONFIG(cache_oblivious, bool); + TEST_MALLCTL_CONFIG(debug, bool); + TEST_MALLCTL_CONFIG(fill, bool); + TEST_MALLCTL_CONFIG(lazy_lock, bool); + TEST_MALLCTL_CONFIG(malloc_conf, const char *); + TEST_MALLCTL_CONFIG(prof, bool); + TEST_MALLCTL_CONFIG(prof_libgcc, bool); + TEST_MALLCTL_CONFIG(prof_libunwind, bool); + TEST_MALLCTL_CONFIG(stats, bool); + TEST_MALLCTL_CONFIG(utrace, bool); + TEST_MALLCTL_CONFIG(xmalloc, bool); + +#undef TEST_MALLCTL_CONFIG +} +TEST_END + +TEST_BEGIN(test_mallctl_opt) { + bool config_always = true; + +#define TEST_MALLCTL_OPT(t, opt, config) do { \ + t oldval; \ + size_t sz = sizeof(oldval); \ + int expected = config_##config ? 0 : ENOENT; \ + int result = mallctl("opt."#opt, (void *)&oldval, &sz, NULL, \ + 0); \ + assert_d_eq(result, expected, \ + "Unexpected mallctl() result for opt."#opt); \ + assert_zu_eq(sz, sizeof(oldval), "Unexpected output size"); \ +} while (0) + + TEST_MALLCTL_OPT(bool, abort, always); + TEST_MALLCTL_OPT(bool, abort_conf, always); + TEST_MALLCTL_OPT(bool, confirm_conf, always); + TEST_MALLCTL_OPT(const char *, metadata_thp, always); + TEST_MALLCTL_OPT(bool, retain, always); + TEST_MALLCTL_OPT(const char *, dss, always); + TEST_MALLCTL_OPT(unsigned, narenas, always); + TEST_MALLCTL_OPT(const char *, percpu_arena, always); + TEST_MALLCTL_OPT(size_t, oversize_threshold, always); + TEST_MALLCTL_OPT(bool, background_thread, always); + TEST_MALLCTL_OPT(ssize_t, dirty_decay_ms, always); + TEST_MALLCTL_OPT(ssize_t, muzzy_decay_ms, always); + TEST_MALLCTL_OPT(bool, stats_print, always); + TEST_MALLCTL_OPT(const char *, junk, fill); + TEST_MALLCTL_OPT(bool, zero, fill); + TEST_MALLCTL_OPT(bool, utrace, utrace); + TEST_MALLCTL_OPT(bool, xmalloc, xmalloc); + TEST_MALLCTL_OPT(bool, tcache, always); + TEST_MALLCTL_OPT(size_t, lg_extent_max_active_fit, always); + TEST_MALLCTL_OPT(size_t, lg_tcache_max, always); + TEST_MALLCTL_OPT(const char *, thp, always); + TEST_MALLCTL_OPT(bool, prof, prof); + TEST_MALLCTL_OPT(const char *, prof_prefix, prof); + TEST_MALLCTL_OPT(bool, prof_active, prof); + TEST_MALLCTL_OPT(ssize_t, lg_prof_sample, prof); + TEST_MALLCTL_OPT(bool, prof_accum, prof); + TEST_MALLCTL_OPT(ssize_t, lg_prof_interval, prof); + TEST_MALLCTL_OPT(bool, prof_gdump, prof); + TEST_MALLCTL_OPT(bool, prof_final, prof); + TEST_MALLCTL_OPT(bool, prof_leak, prof); + +#undef TEST_MALLCTL_OPT +} +TEST_END + +TEST_BEGIN(test_manpage_example) { + unsigned nbins, i; + size_t mib[4]; + size_t len, miblen; + + len = sizeof(nbins); + assert_d_eq(mallctl("arenas.nbins", (void *)&nbins, &len, NULL, 0), 0, + "Unexpected mallctl() failure"); + + miblen = 4; + assert_d_eq(mallctlnametomib("arenas.bin.0.size", mib, &miblen), 0, + "Unexpected mallctlnametomib() failure"); + for (i = 0; i < nbins; i++) { + size_t bin_size; + + mib[2] = i; + len = sizeof(bin_size); + assert_d_eq(mallctlbymib(mib, miblen, (void *)&bin_size, &len, + NULL, 0), 0, "Unexpected mallctlbymib() failure"); + /* Do something with bin_size... */ + } +} +TEST_END + +TEST_BEGIN(test_tcache_none) { + test_skip_if(!opt_tcache); + + /* Allocate p and q. */ + void *p0 = mallocx(42, 0); + assert_ptr_not_null(p0, "Unexpected mallocx() failure"); + void *q = mallocx(42, 0); + assert_ptr_not_null(q, "Unexpected mallocx() failure"); + + /* Deallocate p and q, but bypass the tcache for q. */ + dallocx(p0, 0); + dallocx(q, MALLOCX_TCACHE_NONE); + + /* Make sure that tcache-based allocation returns p, not q. */ + void *p1 = mallocx(42, 0); + assert_ptr_not_null(p1, "Unexpected mallocx() failure"); + assert_ptr_eq(p0, p1, "Expected tcache to allocate cached region"); + + /* Clean up. */ + dallocx(p1, MALLOCX_TCACHE_NONE); +} +TEST_END + +TEST_BEGIN(test_tcache) { +#define NTCACHES 10 + unsigned tis[NTCACHES]; + void *ps[NTCACHES]; + void *qs[NTCACHES]; + unsigned i; + size_t sz, psz, qsz; + + psz = 42; + qsz = nallocx(psz, 0) + 1; + + /* Create tcaches. */ + for (i = 0; i < NTCACHES; i++) { + sz = sizeof(unsigned); + assert_d_eq(mallctl("tcache.create", (void *)&tis[i], &sz, NULL, + 0), 0, "Unexpected mallctl() failure, i=%u", i); + } + + /* Exercise tcache ID recycling. */ + for (i = 0; i < NTCACHES; i++) { + assert_d_eq(mallctl("tcache.destroy", NULL, NULL, + (void *)&tis[i], sizeof(unsigned)), 0, + "Unexpected mallctl() failure, i=%u", i); + } + for (i = 0; i < NTCACHES; i++) { + sz = sizeof(unsigned); + assert_d_eq(mallctl("tcache.create", (void *)&tis[i], &sz, NULL, + 0), 0, "Unexpected mallctl() failure, i=%u", i); + } + + /* Flush empty tcaches. */ + for (i = 0; i < NTCACHES; i++) { + assert_d_eq(mallctl("tcache.flush", NULL, NULL, (void *)&tis[i], + sizeof(unsigned)), 0, "Unexpected mallctl() failure, i=%u", + i); + } + + /* Cache some allocations. */ + for (i = 0; i < NTCACHES; i++) { + ps[i] = mallocx(psz, MALLOCX_TCACHE(tis[i])); + assert_ptr_not_null(ps[i], "Unexpected mallocx() failure, i=%u", + i); + dallocx(ps[i], MALLOCX_TCACHE(tis[i])); + + qs[i] = mallocx(qsz, MALLOCX_TCACHE(tis[i])); + assert_ptr_not_null(qs[i], "Unexpected mallocx() failure, i=%u", + i); + dallocx(qs[i], MALLOCX_TCACHE(tis[i])); + } + + /* Verify that tcaches allocate cached regions. */ + for (i = 0; i < NTCACHES; i++) { + void *p0 = ps[i]; + ps[i] = mallocx(psz, MALLOCX_TCACHE(tis[i])); + assert_ptr_not_null(ps[i], "Unexpected mallocx() failure, i=%u", + i); + assert_ptr_eq(ps[i], p0, + "Expected mallocx() to allocate cached region, i=%u", i); + } + + /* Verify that reallocation uses cached regions. */ + for (i = 0; i < NTCACHES; i++) { + void *q0 = qs[i]; + qs[i] = rallocx(ps[i], qsz, MALLOCX_TCACHE(tis[i])); + assert_ptr_not_null(qs[i], "Unexpected rallocx() failure, i=%u", + i); + assert_ptr_eq(qs[i], q0, + "Expected rallocx() to allocate cached region, i=%u", i); + /* Avoid undefined behavior in case of test failure. */ + if (qs[i] == NULL) { + qs[i] = ps[i]; + } + } + for (i = 0; i < NTCACHES; i++) { + dallocx(qs[i], MALLOCX_TCACHE(tis[i])); + } + + /* Flush some non-empty tcaches. */ + for (i = 0; i < NTCACHES/2; i++) { + assert_d_eq(mallctl("tcache.flush", NULL, NULL, (void *)&tis[i], + sizeof(unsigned)), 0, "Unexpected mallctl() failure, i=%u", + i); + } + + /* Destroy tcaches. */ + for (i = 0; i < NTCACHES; i++) { + assert_d_eq(mallctl("tcache.destroy", NULL, NULL, + (void *)&tis[i], sizeof(unsigned)), 0, + "Unexpected mallctl() failure, i=%u", i); + } +} +TEST_END + +TEST_BEGIN(test_thread_arena) { + unsigned old_arena_ind, new_arena_ind, narenas; + + const char *opa; + size_t sz = sizeof(opa); + assert_d_eq(mallctl("opt.percpu_arena", (void *)&opa, &sz, NULL, 0), 0, + "Unexpected mallctl() failure"); + + sz = sizeof(unsigned); + assert_d_eq(mallctl("arenas.narenas", (void *)&narenas, &sz, NULL, 0), + 0, "Unexpected mallctl() failure"); + if (opt_oversize_threshold != 0) { + narenas--; + } + assert_u_eq(narenas, opt_narenas, "Number of arenas incorrect"); + + if (strcmp(opa, "disabled") == 0) { + new_arena_ind = narenas - 1; + assert_d_eq(mallctl("thread.arena", (void *)&old_arena_ind, &sz, + (void *)&new_arena_ind, sizeof(unsigned)), 0, + "Unexpected mallctl() failure"); + new_arena_ind = 0; + assert_d_eq(mallctl("thread.arena", (void *)&old_arena_ind, &sz, + (void *)&new_arena_ind, sizeof(unsigned)), 0, + "Unexpected mallctl() failure"); + } else { + assert_d_eq(mallctl("thread.arena", (void *)&old_arena_ind, &sz, + NULL, 0), 0, "Unexpected mallctl() failure"); + new_arena_ind = percpu_arena_ind_limit(opt_percpu_arena) - 1; + if (old_arena_ind != new_arena_ind) { + assert_d_eq(mallctl("thread.arena", + (void *)&old_arena_ind, &sz, (void *)&new_arena_ind, + sizeof(unsigned)), EPERM, "thread.arena ctl " + "should not be allowed with percpu arena"); + } + } +} +TEST_END + +TEST_BEGIN(test_arena_i_initialized) { + unsigned narenas, i; + size_t sz; + size_t mib[3]; + size_t miblen = sizeof(mib) / sizeof(size_t); + bool initialized; + + sz = sizeof(narenas); + assert_d_eq(mallctl("arenas.narenas", (void *)&narenas, &sz, NULL, 0), + 0, "Unexpected mallctl() failure"); + + assert_d_eq(mallctlnametomib("arena.0.initialized", mib, &miblen), 0, + "Unexpected mallctlnametomib() failure"); + for (i = 0; i < narenas; i++) { + mib[1] = i; + sz = sizeof(initialized); + assert_d_eq(mallctlbymib(mib, miblen, &initialized, &sz, NULL, + 0), 0, "Unexpected mallctl() failure"); + } + + mib[1] = MALLCTL_ARENAS_ALL; + sz = sizeof(initialized); + assert_d_eq(mallctlbymib(mib, miblen, &initialized, &sz, NULL, 0), 0, + "Unexpected mallctl() failure"); + assert_true(initialized, + "Merged arena statistics should always be initialized"); + + /* Equivalent to the above but using mallctl() directly. */ + sz = sizeof(initialized); + assert_d_eq(mallctl( + "arena." STRINGIFY(MALLCTL_ARENAS_ALL) ".initialized", + (void *)&initialized, &sz, NULL, 0), 0, + "Unexpected mallctl() failure"); + assert_true(initialized, + "Merged arena statistics should always be initialized"); +} +TEST_END + +TEST_BEGIN(test_arena_i_dirty_decay_ms) { + ssize_t dirty_decay_ms, orig_dirty_decay_ms, prev_dirty_decay_ms; + size_t sz = sizeof(ssize_t); + + assert_d_eq(mallctl("arena.0.dirty_decay_ms", + (void *)&orig_dirty_decay_ms, &sz, NULL, 0), 0, + "Unexpected mallctl() failure"); + + dirty_decay_ms = -2; + assert_d_eq(mallctl("arena.0.dirty_decay_ms", NULL, NULL, + (void *)&dirty_decay_ms, sizeof(ssize_t)), EFAULT, + "Unexpected mallctl() success"); + + dirty_decay_ms = 0x7fffffff; + assert_d_eq(mallctl("arena.0.dirty_decay_ms", NULL, NULL, + (void *)&dirty_decay_ms, sizeof(ssize_t)), 0, + "Unexpected mallctl() failure"); + + for (prev_dirty_decay_ms = dirty_decay_ms, dirty_decay_ms = -1; + dirty_decay_ms < 20; prev_dirty_decay_ms = dirty_decay_ms, + dirty_decay_ms++) { + ssize_t old_dirty_decay_ms; + + assert_d_eq(mallctl("arena.0.dirty_decay_ms", + (void *)&old_dirty_decay_ms, &sz, (void *)&dirty_decay_ms, + sizeof(ssize_t)), 0, "Unexpected mallctl() failure"); + assert_zd_eq(old_dirty_decay_ms, prev_dirty_decay_ms, + "Unexpected old arena.0.dirty_decay_ms"); + } +} +TEST_END + +TEST_BEGIN(test_arena_i_muzzy_decay_ms) { + ssize_t muzzy_decay_ms, orig_muzzy_decay_ms, prev_muzzy_decay_ms; + size_t sz = sizeof(ssize_t); + + assert_d_eq(mallctl("arena.0.muzzy_decay_ms", + (void *)&orig_muzzy_decay_ms, &sz, NULL, 0), 0, + "Unexpected mallctl() failure"); + + muzzy_decay_ms = -2; + assert_d_eq(mallctl("arena.0.muzzy_decay_ms", NULL, NULL, + (void *)&muzzy_decay_ms, sizeof(ssize_t)), EFAULT, + "Unexpected mallctl() success"); + + muzzy_decay_ms = 0x7fffffff; + assert_d_eq(mallctl("arena.0.muzzy_decay_ms", NULL, NULL, + (void *)&muzzy_decay_ms, sizeof(ssize_t)), 0, + "Unexpected mallctl() failure"); + + for (prev_muzzy_decay_ms = muzzy_decay_ms, muzzy_decay_ms = -1; + muzzy_decay_ms < 20; prev_muzzy_decay_ms = muzzy_decay_ms, + muzzy_decay_ms++) { + ssize_t old_muzzy_decay_ms; + + assert_d_eq(mallctl("arena.0.muzzy_decay_ms", + (void *)&old_muzzy_decay_ms, &sz, (void *)&muzzy_decay_ms, + sizeof(ssize_t)), 0, "Unexpected mallctl() failure"); + assert_zd_eq(old_muzzy_decay_ms, prev_muzzy_decay_ms, + "Unexpected old arena.0.muzzy_decay_ms"); + } +} +TEST_END + +TEST_BEGIN(test_arena_i_purge) { + unsigned narenas; + size_t sz = sizeof(unsigned); + size_t mib[3]; + size_t miblen = 3; + + assert_d_eq(mallctl("arena.0.purge", NULL, NULL, NULL, 0), 0, + "Unexpected mallctl() failure"); + + assert_d_eq(mallctl("arenas.narenas", (void *)&narenas, &sz, NULL, 0), + 0, "Unexpected mallctl() failure"); + assert_d_eq(mallctlnametomib("arena.0.purge", mib, &miblen), 0, + "Unexpected mallctlnametomib() failure"); + mib[1] = narenas; + assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, NULL, 0), 0, + "Unexpected mallctlbymib() failure"); + + mib[1] = MALLCTL_ARENAS_ALL; + assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, NULL, 0), 0, + "Unexpected mallctlbymib() failure"); +} +TEST_END + +TEST_BEGIN(test_arena_i_decay) { + unsigned narenas; + size_t sz = sizeof(unsigned); + size_t mib[3]; + size_t miblen = 3; + + assert_d_eq(mallctl("arena.0.decay", NULL, NULL, NULL, 0), 0, + "Unexpected mallctl() failure"); + + assert_d_eq(mallctl("arenas.narenas", (void *)&narenas, &sz, NULL, 0), + 0, "Unexpected mallctl() failure"); + assert_d_eq(mallctlnametomib("arena.0.decay", mib, &miblen), 0, + "Unexpected mallctlnametomib() failure"); + mib[1] = narenas; + assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, NULL, 0), 0, + "Unexpected mallctlbymib() failure"); + + mib[1] = MALLCTL_ARENAS_ALL; + assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, NULL, 0), 0, + "Unexpected mallctlbymib() failure"); +} +TEST_END + +TEST_BEGIN(test_arena_i_dss) { + const char *dss_prec_old, *dss_prec_new; + size_t sz = sizeof(dss_prec_old); + size_t mib[3]; + size_t miblen; + + miblen = sizeof(mib)/sizeof(size_t); + assert_d_eq(mallctlnametomib("arena.0.dss", mib, &miblen), 0, + "Unexpected mallctlnametomib() error"); + + dss_prec_new = "disabled"; + assert_d_eq(mallctlbymib(mib, miblen, (void *)&dss_prec_old, &sz, + (void *)&dss_prec_new, sizeof(dss_prec_new)), 0, + "Unexpected mallctl() failure"); + assert_str_ne(dss_prec_old, "primary", + "Unexpected default for dss precedence"); + + assert_d_eq(mallctlbymib(mib, miblen, (void *)&dss_prec_new, &sz, + (void *)&dss_prec_old, sizeof(dss_prec_old)), 0, + "Unexpected mallctl() failure"); + + assert_d_eq(mallctlbymib(mib, miblen, (void *)&dss_prec_old, &sz, NULL, + 0), 0, "Unexpected mallctl() failure"); + assert_str_ne(dss_prec_old, "primary", + "Unexpected value for dss precedence"); + + mib[1] = narenas_total_get(); + dss_prec_new = "disabled"; + assert_d_eq(mallctlbymib(mib, miblen, (void *)&dss_prec_old, &sz, + (void *)&dss_prec_new, sizeof(dss_prec_new)), 0, + "Unexpected mallctl() failure"); + assert_str_ne(dss_prec_old, "primary", + "Unexpected default for dss precedence"); + + assert_d_eq(mallctlbymib(mib, miblen, (void *)&dss_prec_new, &sz, + (void *)&dss_prec_old, sizeof(dss_prec_new)), 0, + "Unexpected mallctl() failure"); + + assert_d_eq(mallctlbymib(mib, miblen, (void *)&dss_prec_old, &sz, NULL, + 0), 0, "Unexpected mallctl() failure"); + assert_str_ne(dss_prec_old, "primary", + "Unexpected value for dss precedence"); +} +TEST_END + +TEST_BEGIN(test_arena_i_retain_grow_limit) { + size_t old_limit, new_limit, default_limit; + size_t mib[3]; + size_t miblen; + + bool retain_enabled; + size_t sz = sizeof(retain_enabled); + assert_d_eq(mallctl("opt.retain", &retain_enabled, &sz, NULL, 0), + 0, "Unexpected mallctl() failure"); + test_skip_if(!retain_enabled); + + sz = sizeof(default_limit); + miblen = sizeof(mib)/sizeof(size_t); + assert_d_eq(mallctlnametomib("arena.0.retain_grow_limit", mib, &miblen), + 0, "Unexpected mallctlnametomib() error"); + + assert_d_eq(mallctlbymib(mib, miblen, &default_limit, &sz, NULL, 0), 0, + "Unexpected mallctl() failure"); + assert_zu_eq(default_limit, SC_LARGE_MAXCLASS, + "Unexpected default for retain_grow_limit"); + + new_limit = PAGE - 1; + assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, &new_limit, + sizeof(new_limit)), EFAULT, "Unexpected mallctl() success"); + + new_limit = PAGE + 1; + assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, &new_limit, + sizeof(new_limit)), 0, "Unexpected mallctl() failure"); + assert_d_eq(mallctlbymib(mib, miblen, &old_limit, &sz, NULL, 0), 0, + "Unexpected mallctl() failure"); + assert_zu_eq(old_limit, PAGE, + "Unexpected value for retain_grow_limit"); + + /* Expect grow less than psize class 10. */ + new_limit = sz_pind2sz(10) - 1; + assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, &new_limit, + sizeof(new_limit)), 0, "Unexpected mallctl() failure"); + assert_d_eq(mallctlbymib(mib, miblen, &old_limit, &sz, NULL, 0), 0, + "Unexpected mallctl() failure"); + assert_zu_eq(old_limit, sz_pind2sz(9), + "Unexpected value for retain_grow_limit"); + + /* Restore to default. */ + assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, &default_limit, + sizeof(default_limit)), 0, "Unexpected mallctl() failure"); +} +TEST_END + +TEST_BEGIN(test_arenas_dirty_decay_ms) { + ssize_t dirty_decay_ms, orig_dirty_decay_ms, prev_dirty_decay_ms; + size_t sz = sizeof(ssize_t); + + assert_d_eq(mallctl("arenas.dirty_decay_ms", + (void *)&orig_dirty_decay_ms, &sz, NULL, 0), 0, + "Unexpected mallctl() failure"); + + dirty_decay_ms = -2; + assert_d_eq(mallctl("arenas.dirty_decay_ms", NULL, NULL, + (void *)&dirty_decay_ms, sizeof(ssize_t)), EFAULT, + "Unexpected mallctl() success"); + + dirty_decay_ms = 0x7fffffff; + assert_d_eq(mallctl("arenas.dirty_decay_ms", NULL, NULL, + (void *)&dirty_decay_ms, sizeof(ssize_t)), 0, + "Expected mallctl() failure"); + + for (prev_dirty_decay_ms = dirty_decay_ms, dirty_decay_ms = -1; + dirty_decay_ms < 20; prev_dirty_decay_ms = dirty_decay_ms, + dirty_decay_ms++) { + ssize_t old_dirty_decay_ms; + + assert_d_eq(mallctl("arenas.dirty_decay_ms", + (void *)&old_dirty_decay_ms, &sz, (void *)&dirty_decay_ms, + sizeof(ssize_t)), 0, "Unexpected mallctl() failure"); + assert_zd_eq(old_dirty_decay_ms, prev_dirty_decay_ms, + "Unexpected old arenas.dirty_decay_ms"); + } +} +TEST_END + +TEST_BEGIN(test_arenas_muzzy_decay_ms) { + ssize_t muzzy_decay_ms, orig_muzzy_decay_ms, prev_muzzy_decay_ms; + size_t sz = sizeof(ssize_t); + + assert_d_eq(mallctl("arenas.muzzy_decay_ms", + (void *)&orig_muzzy_decay_ms, &sz, NULL, 0), 0, + "Unexpected mallctl() failure"); + + muzzy_decay_ms = -2; + assert_d_eq(mallctl("arenas.muzzy_decay_ms", NULL, NULL, + (void *)&muzzy_decay_ms, sizeof(ssize_t)), EFAULT, + "Unexpected mallctl() success"); + + muzzy_decay_ms = 0x7fffffff; + assert_d_eq(mallctl("arenas.muzzy_decay_ms", NULL, NULL, + (void *)&muzzy_decay_ms, sizeof(ssize_t)), 0, + "Expected mallctl() failure"); + + for (prev_muzzy_decay_ms = muzzy_decay_ms, muzzy_decay_ms = -1; + muzzy_decay_ms < 20; prev_muzzy_decay_ms = muzzy_decay_ms, + muzzy_decay_ms++) { + ssize_t old_muzzy_decay_ms; + + assert_d_eq(mallctl("arenas.muzzy_decay_ms", + (void *)&old_muzzy_decay_ms, &sz, (void *)&muzzy_decay_ms, + sizeof(ssize_t)), 0, "Unexpected mallctl() failure"); + assert_zd_eq(old_muzzy_decay_ms, prev_muzzy_decay_ms, + "Unexpected old arenas.muzzy_decay_ms"); + } +} +TEST_END + +TEST_BEGIN(test_arenas_constants) { +#define TEST_ARENAS_CONSTANT(t, name, expected) do { \ + t name; \ + size_t sz = sizeof(t); \ + assert_d_eq(mallctl("arenas."#name, (void *)&name, &sz, NULL, \ + 0), 0, "Unexpected mallctl() failure"); \ + assert_zu_eq(name, expected, "Incorrect "#name" size"); \ +} while (0) + + TEST_ARENAS_CONSTANT(size_t, quantum, QUANTUM); + TEST_ARENAS_CONSTANT(size_t, page, PAGE); + TEST_ARENAS_CONSTANT(unsigned, nbins, SC_NBINS); + TEST_ARENAS_CONSTANT(unsigned, nlextents, SC_NSIZES - SC_NBINS); + +#undef TEST_ARENAS_CONSTANT +} +TEST_END + +TEST_BEGIN(test_arenas_bin_constants) { +#define TEST_ARENAS_BIN_CONSTANT(t, name, expected) do { \ + t name; \ + size_t sz = sizeof(t); \ + assert_d_eq(mallctl("arenas.bin.0."#name, (void *)&name, &sz, \ + NULL, 0), 0, "Unexpected mallctl() failure"); \ + assert_zu_eq(name, expected, "Incorrect "#name" size"); \ +} while (0) + + TEST_ARENAS_BIN_CONSTANT(size_t, size, bin_infos[0].reg_size); + TEST_ARENAS_BIN_CONSTANT(uint32_t, nregs, bin_infos[0].nregs); + TEST_ARENAS_BIN_CONSTANT(size_t, slab_size, + bin_infos[0].slab_size); + TEST_ARENAS_BIN_CONSTANT(uint32_t, nshards, bin_infos[0].n_shards); + +#undef TEST_ARENAS_BIN_CONSTANT +} +TEST_END + +TEST_BEGIN(test_arenas_lextent_constants) { +#define TEST_ARENAS_LEXTENT_CONSTANT(t, name, expected) do { \ + t name; \ + size_t sz = sizeof(t); \ + assert_d_eq(mallctl("arenas.lextent.0."#name, (void *)&name, \ + &sz, NULL, 0), 0, "Unexpected mallctl() failure"); \ + assert_zu_eq(name, expected, "Incorrect "#name" size"); \ +} while (0) + + TEST_ARENAS_LEXTENT_CONSTANT(size_t, size, + SC_LARGE_MINCLASS); + +#undef TEST_ARENAS_LEXTENT_CONSTANT +} +TEST_END + +TEST_BEGIN(test_arenas_create) { + unsigned narenas_before, arena, narenas_after; + size_t sz = sizeof(unsigned); + + assert_d_eq(mallctl("arenas.narenas", (void *)&narenas_before, &sz, + NULL, 0), 0, "Unexpected mallctl() failure"); + assert_d_eq(mallctl("arenas.create", (void *)&arena, &sz, NULL, 0), 0, + "Unexpected mallctl() failure"); + assert_d_eq(mallctl("arenas.narenas", (void *)&narenas_after, &sz, NULL, + 0), 0, "Unexpected mallctl() failure"); + + assert_u_eq(narenas_before+1, narenas_after, + "Unexpected number of arenas before versus after extension"); + assert_u_eq(arena, narenas_after-1, "Unexpected arena index"); +} +TEST_END + +TEST_BEGIN(test_arenas_lookup) { + unsigned arena, arena1; + void *ptr; + size_t sz = sizeof(unsigned); + + assert_d_eq(mallctl("arenas.create", (void *)&arena, &sz, NULL, 0), 0, + "Unexpected mallctl() failure"); + ptr = mallocx(42, MALLOCX_ARENA(arena) | MALLOCX_TCACHE_NONE); + assert_ptr_not_null(ptr, "Unexpected mallocx() failure"); + assert_d_eq(mallctl("arenas.lookup", &arena1, &sz, &ptr, sizeof(ptr)), + 0, "Unexpected mallctl() failure"); + assert_u_eq(arena, arena1, "Unexpected arena index"); + dallocx(ptr, 0); +} +TEST_END + +TEST_BEGIN(test_stats_arenas) { +#define TEST_STATS_ARENAS(t, name) do { \ + t name; \ + size_t sz = sizeof(t); \ + assert_d_eq(mallctl("stats.arenas.0."#name, (void *)&name, &sz, \ + NULL, 0), 0, "Unexpected mallctl() failure"); \ +} while (0) + + TEST_STATS_ARENAS(unsigned, nthreads); + TEST_STATS_ARENAS(const char *, dss); + TEST_STATS_ARENAS(ssize_t, dirty_decay_ms); + TEST_STATS_ARENAS(ssize_t, muzzy_decay_ms); + TEST_STATS_ARENAS(size_t, pactive); + TEST_STATS_ARENAS(size_t, pdirty); + +#undef TEST_STATS_ARENAS +} +TEST_END + +static void +alloc_hook(void *extra, UNUSED hook_alloc_t type, UNUSED void *result, + UNUSED uintptr_t result_raw, UNUSED uintptr_t args_raw[3]) { + *(bool *)extra = true; +} + +static void +dalloc_hook(void *extra, UNUSED hook_dalloc_t type, + UNUSED void *address, UNUSED uintptr_t args_raw[3]) { + *(bool *)extra = true; +} + +TEST_BEGIN(test_hooks) { + bool hook_called = false; + hooks_t hooks = {&alloc_hook, &dalloc_hook, NULL, &hook_called}; + void *handle = NULL; + size_t sz = sizeof(handle); + int err = mallctl("experimental.hooks.install", &handle, &sz, &hooks, + sizeof(hooks)); + assert_d_eq(err, 0, "Hook installation failed"); + assert_ptr_ne(handle, NULL, "Hook installation gave null handle"); + void *ptr = mallocx(1, 0); + assert_true(hook_called, "Alloc hook not called"); + hook_called = false; + free(ptr); + assert_true(hook_called, "Free hook not called"); + + err = mallctl("experimental.hooks.remove", NULL, NULL, &handle, + sizeof(handle)); + assert_d_eq(err, 0, "Hook removal failed"); + hook_called = false; + ptr = mallocx(1, 0); + free(ptr); + assert_false(hook_called, "Hook called after removal"); +} +TEST_END + +TEST_BEGIN(test_hooks_exhaustion) { + bool hook_called = false; + hooks_t hooks = {&alloc_hook, &dalloc_hook, NULL, &hook_called}; + + void *handle; + void *handles[HOOK_MAX]; + size_t sz = sizeof(handle); + int err; + for (int i = 0; i < HOOK_MAX; i++) { + handle = NULL; + err = mallctl("experimental.hooks.install", &handle, &sz, + &hooks, sizeof(hooks)); + assert_d_eq(err, 0, "Error installation hooks"); + assert_ptr_ne(handle, NULL, "Got NULL handle"); + handles[i] = handle; + } + err = mallctl("experimental.hooks.install", &handle, &sz, &hooks, + sizeof(hooks)); + assert_d_eq(err, EAGAIN, "Should have failed hook installation"); + for (int i = 0; i < HOOK_MAX; i++) { + err = mallctl("experimental.hooks.remove", NULL, NULL, + &handles[i], sizeof(handles[i])); + assert_d_eq(err, 0, "Hook removal failed"); + } + /* Insertion failed, but then we removed some; it should work now. */ + handle = NULL; + err = mallctl("experimental.hooks.install", &handle, &sz, &hooks, + sizeof(hooks)); + assert_d_eq(err, 0, "Hook insertion failed"); + assert_ptr_ne(handle, NULL, "Got NULL handle"); + err = mallctl("experimental.hooks.remove", NULL, NULL, &handle, + sizeof(handle)); + assert_d_eq(err, 0, "Hook removal failed"); +} +TEST_END + +int +main(void) { + return test( + test_mallctl_errors, + test_mallctlnametomib_errors, + test_mallctlbymib_errors, + test_mallctl_read_write, + test_mallctlnametomib_short_mib, + test_mallctl_config, + test_mallctl_opt, + test_manpage_example, + test_tcache_none, + test_tcache, + test_thread_arena, + test_arena_i_initialized, + test_arena_i_dirty_decay_ms, + test_arena_i_muzzy_decay_ms, + test_arena_i_purge, + test_arena_i_decay, + test_arena_i_dss, + test_arena_i_retain_grow_limit, + test_arenas_dirty_decay_ms, + test_arenas_muzzy_decay_ms, + test_arenas_constants, + test_arenas_bin_constants, + test_arenas_lextent_constants, + test_arenas_create, + test_arenas_lookup, + test_stats_arenas, + test_hooks, + test_hooks_exhaustion); +} diff --git a/deps/jemalloc/test/unit/malloc_io.c b/deps/jemalloc/test/unit/malloc_io.c new file mode 100644 index 000000000..79ba7fc53 --- /dev/null +++ b/deps/jemalloc/test/unit/malloc_io.c @@ -0,0 +1,258 @@ +#include "test/jemalloc_test.h" + +TEST_BEGIN(test_malloc_strtoumax_no_endptr) { + int err; + + set_errno(0); + assert_ju_eq(malloc_strtoumax("0", NULL, 0), 0, "Unexpected result"); + err = get_errno(); + assert_d_eq(err, 0, "Unexpected failure"); +} +TEST_END + +TEST_BEGIN(test_malloc_strtoumax) { + struct test_s { + const char *input; + const char *expected_remainder; + int base; + int expected_errno; + const char *expected_errno_name; + uintmax_t expected_x; + }; +#define ERR(e) e, #e +#define KUMAX(x) ((uintmax_t)x##ULL) +#define KSMAX(x) ((uintmax_t)(intmax_t)x##LL) + struct test_s tests[] = { + {"0", "0", -1, ERR(EINVAL), UINTMAX_MAX}, + {"0", "0", 1, ERR(EINVAL), UINTMAX_MAX}, + {"0", "0", 37, ERR(EINVAL), UINTMAX_MAX}, + + {"", "", 0, ERR(EINVAL), UINTMAX_MAX}, + {"+", "+", 0, ERR(EINVAL), UINTMAX_MAX}, + {"++3", "++3", 0, ERR(EINVAL), UINTMAX_MAX}, + {"-", "-", 0, ERR(EINVAL), UINTMAX_MAX}, + + {"42", "", 0, ERR(0), KUMAX(42)}, + {"+42", "", 0, ERR(0), KUMAX(42)}, + {"-42", "", 0, ERR(0), KSMAX(-42)}, + {"042", "", 0, ERR(0), KUMAX(042)}, + {"+042", "", 0, ERR(0), KUMAX(042)}, + {"-042", "", 0, ERR(0), KSMAX(-042)}, + {"0x42", "", 0, ERR(0), KUMAX(0x42)}, + {"+0x42", "", 0, ERR(0), KUMAX(0x42)}, + {"-0x42", "", 0, ERR(0), KSMAX(-0x42)}, + + {"0", "", 0, ERR(0), KUMAX(0)}, + {"1", "", 0, ERR(0), KUMAX(1)}, + + {"42", "", 0, ERR(0), KUMAX(42)}, + {" 42", "", 0, ERR(0), KUMAX(42)}, + {"42 ", " ", 0, ERR(0), KUMAX(42)}, + {"0x", "x", 0, ERR(0), KUMAX(0)}, + {"42x", "x", 0, ERR(0), KUMAX(42)}, + + {"07", "", 0, ERR(0), KUMAX(7)}, + {"010", "", 0, ERR(0), KUMAX(8)}, + {"08", "8", 0, ERR(0), KUMAX(0)}, + {"0_", "_", 0, ERR(0), KUMAX(0)}, + + {"0x", "x", 0, ERR(0), KUMAX(0)}, + {"0X", "X", 0, ERR(0), KUMAX(0)}, + {"0xg", "xg", 0, ERR(0), KUMAX(0)}, + {"0XA", "", 0, ERR(0), KUMAX(10)}, + + {"010", "", 10, ERR(0), KUMAX(10)}, + {"0x3", "x3", 10, ERR(0), KUMAX(0)}, + + {"12", "2", 2, ERR(0), KUMAX(1)}, + {"78", "8", 8, ERR(0), KUMAX(7)}, + {"9a", "a", 10, ERR(0), KUMAX(9)}, + {"9A", "A", 10, ERR(0), KUMAX(9)}, + {"fg", "g", 16, ERR(0), KUMAX(15)}, + {"FG", "G", 16, ERR(0), KUMAX(15)}, + {"0xfg", "g", 16, ERR(0), KUMAX(15)}, + {"0XFG", "G", 16, ERR(0), KUMAX(15)}, + {"z_", "_", 36, ERR(0), KUMAX(35)}, + {"Z_", "_", 36, ERR(0), KUMAX(35)} + }; +#undef ERR +#undef KUMAX +#undef KSMAX + unsigned i; + + for (i = 0; i < sizeof(tests)/sizeof(struct test_s); i++) { + struct test_s *test = &tests[i]; + int err; + uintmax_t result; + char *remainder; + + set_errno(0); + result = malloc_strtoumax(test->input, &remainder, test->base); + err = get_errno(); + assert_d_eq(err, test->expected_errno, + "Expected errno %s for \"%s\", base %d", + test->expected_errno_name, test->input, test->base); + assert_str_eq(remainder, test->expected_remainder, + "Unexpected remainder for \"%s\", base %d", + test->input, test->base); + if (err == 0) { + assert_ju_eq(result, test->expected_x, + "Unexpected result for \"%s\", base %d", + test->input, test->base); + } + } +} +TEST_END + +TEST_BEGIN(test_malloc_snprintf_truncated) { +#define BUFLEN 15 + char buf[BUFLEN]; + size_t result; + size_t len; +#define TEST(expected_str_untruncated, ...) do { \ + result = malloc_snprintf(buf, len, __VA_ARGS__); \ + assert_d_eq(strncmp(buf, expected_str_untruncated, len-1), 0, \ + "Unexpected string inequality (\"%s\" vs \"%s\")", \ + buf, expected_str_untruncated); \ + assert_zu_eq(result, strlen(expected_str_untruncated), \ + "Unexpected result"); \ +} while (0) + + for (len = 1; len < BUFLEN; len++) { + TEST("012346789", "012346789"); + TEST("a0123b", "a%sb", "0123"); + TEST("a01234567", "a%s%s", "0123", "4567"); + TEST("a0123 ", "a%-6s", "0123"); + TEST("a 0123", "a%6s", "0123"); + TEST("a 012", "a%6.3s", "0123"); + TEST("a 012", "a%*.*s", 6, 3, "0123"); + TEST("a 123b", "a% db", 123); + TEST("a123b", "a%-db", 123); + TEST("a-123b", "a%-db", -123); + TEST("a+123b", "a%+db", 123); + } +#undef BUFLEN +#undef TEST +} +TEST_END + +TEST_BEGIN(test_malloc_snprintf) { +#define BUFLEN 128 + char buf[BUFLEN]; + size_t result; +#define TEST(expected_str, ...) do { \ + result = malloc_snprintf(buf, sizeof(buf), __VA_ARGS__); \ + assert_str_eq(buf, expected_str, "Unexpected output"); \ + assert_zu_eq(result, strlen(expected_str), "Unexpected result");\ +} while (0) + + TEST("hello", "hello"); + + TEST("50%, 100%", "50%%, %d%%", 100); + + TEST("a0123b", "a%sb", "0123"); + + TEST("a 0123b", "a%5sb", "0123"); + TEST("a 0123b", "a%*sb", 5, "0123"); + + TEST("a0123 b", "a%-5sb", "0123"); + TEST("a0123b", "a%*sb", -1, "0123"); + TEST("a0123 b", "a%*sb", -5, "0123"); + TEST("a0123 b", "a%-*sb", -5, "0123"); + + TEST("a012b", "a%.3sb", "0123"); + TEST("a012b", "a%.*sb", 3, "0123"); + TEST("a0123b", "a%.*sb", -3, "0123"); + + TEST("a 012b", "a%5.3sb", "0123"); + TEST("a 012b", "a%5.*sb", 3, "0123"); + TEST("a 012b", "a%*.3sb", 5, "0123"); + TEST("a 012b", "a%*.*sb", 5, 3, "0123"); + TEST("a 0123b", "a%*.*sb", 5, -3, "0123"); + + TEST("_abcd_", "_%x_", 0xabcd); + TEST("_0xabcd_", "_%#x_", 0xabcd); + TEST("_1234_", "_%o_", 01234); + TEST("_01234_", "_%#o_", 01234); + TEST("_1234_", "_%u_", 1234); + + TEST("_1234_", "_%d_", 1234); + TEST("_ 1234_", "_% d_", 1234); + TEST("_+1234_", "_%+d_", 1234); + TEST("_-1234_", "_%d_", -1234); + TEST("_-1234_", "_% d_", -1234); + TEST("_-1234_", "_%+d_", -1234); + + TEST("_-1234_", "_%d_", -1234); + TEST("_1234_", "_%d_", 1234); + TEST("_-1234_", "_%i_", -1234); + TEST("_1234_", "_%i_", 1234); + TEST("_01234_", "_%#o_", 01234); + TEST("_1234_", "_%u_", 1234); + TEST("_0x1234abc_", "_%#x_", 0x1234abc); + TEST("_0X1234ABC_", "_%#X_", 0x1234abc); + TEST("_c_", "_%c_", 'c'); + TEST("_string_", "_%s_", "string"); + TEST("_0x42_", "_%p_", ((void *)0x42)); + + TEST("_-1234_", "_%ld_", ((long)-1234)); + TEST("_1234_", "_%ld_", ((long)1234)); + TEST("_-1234_", "_%li_", ((long)-1234)); + TEST("_1234_", "_%li_", ((long)1234)); + TEST("_01234_", "_%#lo_", ((long)01234)); + TEST("_1234_", "_%lu_", ((long)1234)); + TEST("_0x1234abc_", "_%#lx_", ((long)0x1234abc)); + TEST("_0X1234ABC_", "_%#lX_", ((long)0x1234ABC)); + + TEST("_-1234_", "_%lld_", ((long long)-1234)); + TEST("_1234_", "_%lld_", ((long long)1234)); + TEST("_-1234_", "_%lli_", ((long long)-1234)); + TEST("_1234_", "_%lli_", ((long long)1234)); + TEST("_01234_", "_%#llo_", ((long long)01234)); + TEST("_1234_", "_%llu_", ((long long)1234)); + TEST("_0x1234abc_", "_%#llx_", ((long long)0x1234abc)); + TEST("_0X1234ABC_", "_%#llX_", ((long long)0x1234ABC)); + + TEST("_-1234_", "_%qd_", ((long long)-1234)); + TEST("_1234_", "_%qd_", ((long long)1234)); + TEST("_-1234_", "_%qi_", ((long long)-1234)); + TEST("_1234_", "_%qi_", ((long long)1234)); + TEST("_01234_", "_%#qo_", ((long long)01234)); + TEST("_1234_", "_%qu_", ((long long)1234)); + TEST("_0x1234abc_", "_%#qx_", ((long long)0x1234abc)); + TEST("_0X1234ABC_", "_%#qX_", ((long long)0x1234ABC)); + + TEST("_-1234_", "_%jd_", ((intmax_t)-1234)); + TEST("_1234_", "_%jd_", ((intmax_t)1234)); + TEST("_-1234_", "_%ji_", ((intmax_t)-1234)); + TEST("_1234_", "_%ji_", ((intmax_t)1234)); + TEST("_01234_", "_%#jo_", ((intmax_t)01234)); + TEST("_1234_", "_%ju_", ((intmax_t)1234)); + TEST("_0x1234abc_", "_%#jx_", ((intmax_t)0x1234abc)); + TEST("_0X1234ABC_", "_%#jX_", ((intmax_t)0x1234ABC)); + + TEST("_1234_", "_%td_", ((ptrdiff_t)1234)); + TEST("_-1234_", "_%td_", ((ptrdiff_t)-1234)); + TEST("_1234_", "_%ti_", ((ptrdiff_t)1234)); + TEST("_-1234_", "_%ti_", ((ptrdiff_t)-1234)); + + TEST("_-1234_", "_%zd_", ((ssize_t)-1234)); + TEST("_1234_", "_%zd_", ((ssize_t)1234)); + TEST("_-1234_", "_%zi_", ((ssize_t)-1234)); + TEST("_1234_", "_%zi_", ((ssize_t)1234)); + TEST("_01234_", "_%#zo_", ((ssize_t)01234)); + TEST("_1234_", "_%zu_", ((ssize_t)1234)); + TEST("_0x1234abc_", "_%#zx_", ((ssize_t)0x1234abc)); + TEST("_0X1234ABC_", "_%#zX_", ((ssize_t)0x1234ABC)); +#undef BUFLEN +} +TEST_END + +int +main(void) { + return test( + test_malloc_strtoumax_no_endptr, + test_malloc_strtoumax, + test_malloc_snprintf_truncated, + test_malloc_snprintf); +} diff --git a/deps/jemalloc/test/unit/math.c b/deps/jemalloc/test/unit/math.c new file mode 100644 index 000000000..09ef20c7b --- /dev/null +++ b/deps/jemalloc/test/unit/math.c @@ -0,0 +1,390 @@ +#include "test/jemalloc_test.h" + +#define MAX_REL_ERR 1.0e-9 +#define MAX_ABS_ERR 1.0e-9 + +#include <float.h> + +#ifdef __PGI +#undef INFINITY +#endif + +#ifndef INFINITY +#define INFINITY (DBL_MAX + DBL_MAX) +#endif + +static bool +double_eq_rel(double a, double b, double max_rel_err, double max_abs_err) { + double rel_err; + + if (fabs(a - b) < max_abs_err) { + return true; + } + rel_err = (fabs(b) > fabs(a)) ? fabs((a-b)/b) : fabs((a-b)/a); + return (rel_err < max_rel_err); +} + +static uint64_t +factorial(unsigned x) { + uint64_t ret = 1; + unsigned i; + + for (i = 2; i <= x; i++) { + ret *= (uint64_t)i; + } + + return ret; +} + +TEST_BEGIN(test_ln_gamma_factorial) { + unsigned x; + + /* exp(ln_gamma(x)) == (x-1)! for integer x. */ + for (x = 1; x <= 21; x++) { + assert_true(double_eq_rel(exp(ln_gamma(x)), + (double)factorial(x-1), MAX_REL_ERR, MAX_ABS_ERR), + "Incorrect factorial result for x=%u", x); + } +} +TEST_END + +/* Expected ln_gamma([0.0..100.0] increment=0.25). */ +static const double ln_gamma_misc_expected[] = { + INFINITY, + 1.28802252469807743, 0.57236494292470008, 0.20328095143129538, + 0.00000000000000000, -0.09827183642181320, -0.12078223763524518, + -0.08440112102048555, 0.00000000000000000, 0.12487171489239651, + 0.28468287047291918, 0.47521466691493719, 0.69314718055994529, + 0.93580193110872523, 1.20097360234707429, 1.48681557859341718, + 1.79175946922805496, 2.11445692745037128, 2.45373657084244234, + 2.80857141857573644, 3.17805383034794575, 3.56137591038669710, + 3.95781396761871651, 4.36671603662228680, 4.78749174278204581, + 5.21960398699022932, 5.66256205985714178, 6.11591589143154568, + 6.57925121201010121, 7.05218545073853953, 7.53436423675873268, + 8.02545839631598312, 8.52516136106541467, 9.03318691960512332, + 9.54926725730099690, 10.07315123968123949, 10.60460290274525086, + 11.14340011995171231, 11.68933342079726856, 12.24220494005076176, + 12.80182748008146909, 13.36802367147604720, 13.94062521940376342, + 14.51947222506051816, 15.10441257307551943, 15.69530137706046524, + 16.29200047656724237, 16.89437797963419285, 17.50230784587389010, + 18.11566950571089407, 18.73434751193644843, 19.35823122022435427, + 19.98721449566188468, 20.62119544270163018, 21.26007615624470048, + 21.90376249182879320, 22.55216385312342098, 23.20519299513386002, + 23.86276584168908954, 24.52480131594137802, 25.19122118273868338, + 25.86194990184851861, 26.53691449111561340, 27.21604439872720604, + 27.89927138384089389, 28.58652940490193828, 29.27775451504081516, + 29.97288476399884871, 30.67186010608067548, 31.37462231367769050, + 32.08111489594735843, 32.79128302226991565, 33.50507345013689076, + 34.22243445715505317, 34.94331577687681545, 35.66766853819134298, + 36.39544520803305261, 37.12659953718355865, 37.86108650896109395, + 38.59886229060776230, 39.33988418719949465, 40.08411059791735198, + 40.83150097453079752, 41.58201578195490100, 42.33561646075348506, + 43.09226539146988699, 43.85192586067515208, 44.61456202863158893, + 45.38013889847690052, 46.14862228684032885, 46.91997879580877395, + 47.69417578616628361, 48.47118135183522014, 49.25096429545256882, + 50.03349410501914463, 50.81874093156324790, 51.60667556776436982, + 52.39726942748592364, 53.19049452616926743, 53.98632346204390586, + 54.78472939811231157, 55.58568604486942633, 56.38916764371992940, + 57.19514895105859864, 58.00360522298051080, 58.81451220059079787, + 59.62784609588432261, 60.44358357816834371, 61.26170176100199427, + 62.08217818962842927, 62.90499082887649962, 63.73011805151035958, + 64.55753862700632340, 65.38723171073768015, 66.21917683354901385, + 67.05335389170279825, 67.88974313718154008, 68.72832516833013017, + 69.56908092082363737, 70.41199165894616385, 71.25703896716800045, + 72.10420474200799390, 72.95347118416940191, 73.80482079093779646, + 74.65823634883015814, 75.51370092648485866, 76.37119786778275454, + 77.23071078519033961, 78.09222355331530707, 78.95572030266725960, + 79.82118541361435859, 80.68860351052903468, 81.55795945611502873, + 82.42923834590904164, 83.30242550295004378, 84.17750647261028973, + 85.05446701758152983, 85.93329311301090456, 86.81397094178107920, + 87.69648688992882057, 88.58082754219766741, 89.46697967771913795, + 90.35493026581838194, 91.24466646193963015, 92.13617560368709292, + 93.02944520697742803, 93.92446296229978486, 94.82121673107967297, + 95.71969454214321615, 96.61988458827809723, 97.52177522288820910, + 98.42535495673848800, 99.33061245478741341, 100.23753653310367895, + 101.14611615586458981, 102.05634043243354370, 102.96819861451382394, + 103.88168009337621811, 104.79677439715833032, 105.71347118823287303, + 106.63176026064346047, 107.55163153760463501, 108.47307506906540198, + 109.39608102933323153, 110.32063971475740516, 111.24674154146920557, + 112.17437704317786995, 113.10353686902013237, 114.03421178146170689, + 114.96639265424990128, 115.90007047041454769, 116.83523632031698014, + 117.77188139974506953, 118.70999700805310795, 119.64957454634490830, + 120.59060551569974962, 121.53308151543865279, 122.47699424143097247, + 123.42233548443955726, 124.36909712850338394, 125.31727114935689826, + 126.26684961288492559, 127.21782467361175861, 128.17018857322420899, + 129.12393363912724453, 130.07905228303084755, 131.03553699956862033, + 131.99338036494577864, 132.95257503561629164, 133.91311374698926784, + 134.87498931216194364, 135.83819462068046846, 136.80272263732638294, + 137.76856640092901785, 138.73571902320256299, 139.70417368760718091, + 140.67392364823425055, 141.64496222871400732, 142.61728282114600574, + 143.59087888505104047, 144.56574394634486680, 145.54187159633210058, + 146.51925549072063859, 147.49788934865566148, 148.47776695177302031, + 149.45888214327129617, 150.44122882700193600, 151.42480096657754984, + 152.40959258449737490, 153.39559776128982094, 154.38281063467164245, + 155.37122539872302696, 156.36083630307879844, 157.35163765213474107, + 158.34362380426921391, 159.33678917107920370, 160.33112821663092973, + 161.32663545672428995, 162.32330545817117695, 163.32113283808695314, + 164.32011226319519892, 165.32023844914485267, 166.32150615984036790, + 167.32391020678358018, 168.32744544842768164, 169.33210678954270634, + 170.33788918059275375, 171.34478761712384198, 172.35279713916281707, + 173.36191283062726143, 174.37212981874515094, 175.38344327348534080, + 176.39584840699734514, 177.40934047306160437, 178.42391476654847793, + 179.43956662288721304, 180.45629141754378111, 181.47408456550741107, + 182.49294152078630304, 183.51285777591152737, 184.53382886144947861, + 185.55585034552262869, 186.57891783333786861, 187.60302696672312095, + 188.62817342367162610, 189.65435291789341932, 190.68156119837468054, + 191.70979404894376330, 192.73904728784492590, 193.76931676731820176, + 194.80059837318714244, 195.83288802445184729, 196.86618167288995096, + 197.90047530266301123, 198.93576492992946214, 199.97204660246373464, + 201.00931639928148797, 202.04757043027063901, 203.08680483582807597, + 204.12701578650228385, 205.16819948264117102, 206.21035215404597807, + 207.25347005962987623, 208.29754948708190909, 209.34258675253678916, + 210.38857820024875878, 211.43552020227099320, 212.48340915813977858, + 213.53224149456323744, 214.58201366511514152, 215.63272214993284592, + 216.68436345542014010, 217.73693411395422004, 218.79043068359703739, + 219.84484974781133815, 220.90018791517996988, 221.95644181913033322, + 223.01360811766215875, 224.07168349307951871, 225.13066465172661879, + 226.19054832372759734, 227.25133126272962159, 228.31301024565024704, + 229.37558207242807384, 230.43904356577689896, 231.50339157094342113, + 232.56862295546847008, 233.63473460895144740, 234.70172344281823484, + 235.76958639009222907, 236.83832040516844586, 237.90792246359117712, + 238.97838956183431947, 240.04971871708477238, 241.12190696702904802, + 242.19495136964280846, 243.26884900298270509, 244.34359696498191283, + 245.41919237324782443, 246.49563236486270057, 247.57291409618682110, + 248.65103474266476269, 249.72999149863338175, 250.80978157713354904, + 251.89040220972316320, 252.97185064629374551, 254.05412415488834199, + 255.13722002152300661, 256.22113555000953511, 257.30586806178126835, + 258.39141489572085675, 259.47777340799029844, 260.56494097186322279, + 261.65291497755913497, 262.74169283208021852, 263.83127195904967266, + 264.92164979855277807, 266.01282380697938379, 267.10479145686849733, + 268.19755023675537586, 269.29109765101975427, 270.38543121973674488, + 271.48054847852881721, 272.57644697842033565, 273.67312428569374561, + 274.77057798174683967, 275.86880566295326389, 276.96780494052313770, + 278.06757344036617496, 279.16810880295668085, 280.26940868320008349, + 281.37147075030043197, 282.47429268763045229, 283.57787219260217171, + 284.68220697654078322, 285.78729476455760050, 286.89313329542699194, + 287.99972032146268930, 289.10705360839756395, 290.21513093526289140, + 291.32395009427028754, 292.43350889069523646, 293.54380514276073200, + 294.65483668152336350, 295.76660135076059532, 296.87909700685889902, + 297.99232151870342022, 299.10627276756946458, 300.22094864701409733, + 301.33634706277030091, 302.45246593264130297, 303.56930318639643929, + 304.68685676566872189, 305.80512462385280514, 306.92410472600477078, + 308.04379504874236773, 309.16419358014690033, 310.28529831966631036, + 311.40710727801865687, 312.52961847709792664, 313.65282994987899201, + 314.77673974032603610, 315.90134590329950015, 317.02664650446632777, + 318.15263962020929966, 319.27932333753892635, 320.40669575400545455, + 321.53475497761127144, 322.66349912672620803, 323.79292633000159185, + 324.92303472628691452, 326.05382246454587403, 327.18528770377525916, + 328.31742861292224234, 329.45024337080525356, 330.58373016603343331, + 331.71788719692847280, 332.85271267144611329, 333.98820480709991898, + 335.12436183088397001, 336.26118197919845443, 337.39866349777429377, + 338.53680464159958774, 339.67560367484657036, 340.81505887079896411, + 341.95516851178109619, 343.09593088908627578, 344.23734430290727460, + 345.37940706226686416, 346.52211748494903532, 347.66547389743118401, + 348.80947463481720661, 349.95411804077025408, 351.09940246744753267, + 352.24532627543504759, 353.39188783368263103, 354.53908551944078908, + 355.68691771819692349, 356.83538282361303118, 357.98447923746385868, + 359.13420536957539753 +}; + +TEST_BEGIN(test_ln_gamma_misc) { + unsigned i; + + for (i = 1; i < sizeof(ln_gamma_misc_expected)/sizeof(double); i++) { + double x = (double)i * 0.25; + assert_true(double_eq_rel(ln_gamma(x), + ln_gamma_misc_expected[i], MAX_REL_ERR, MAX_ABS_ERR), + "Incorrect ln_gamma result for i=%u", i); + } +} +TEST_END + +/* Expected pt_norm([0.01..0.99] increment=0.01). */ +static const double pt_norm_expected[] = { + -INFINITY, + -2.32634787404084076, -2.05374891063182252, -1.88079360815125085, + -1.75068607125216946, -1.64485362695147264, -1.55477359459685305, + -1.47579102817917063, -1.40507156030963221, -1.34075503369021654, + -1.28155156554460081, -1.22652812003661049, -1.17498679206608991, + -1.12639112903880045, -1.08031934081495606, -1.03643338949378938, + -0.99445788320975281, -0.95416525314619416, -0.91536508784281390, + -0.87789629505122846, -0.84162123357291418, -0.80642124701824025, + -0.77219321418868492, -0.73884684918521371, -0.70630256284008752, + -0.67448975019608171, -0.64334540539291685, -0.61281299101662701, + -0.58284150727121620, -0.55338471955567281, -0.52440051270804067, + -0.49585034734745320, -0.46769879911450812, -0.43991316567323380, + -0.41246312944140462, -0.38532046640756751, -0.35845879325119373, + -0.33185334643681652, -0.30548078809939738, -0.27931903444745404, + -0.25334710313579978, -0.22754497664114931, -0.20189347914185077, + -0.17637416478086135, -0.15096921549677725, -0.12566134685507399, + -0.10043372051146975, -0.07526986209982976, -0.05015358346473352, + -0.02506890825871106, 0.00000000000000000, 0.02506890825871106, + 0.05015358346473366, 0.07526986209982990, 0.10043372051146990, + 0.12566134685507413, 0.15096921549677739, 0.17637416478086146, + 0.20189347914185105, 0.22754497664114931, 0.25334710313579978, + 0.27931903444745404, 0.30548078809939738, 0.33185334643681652, + 0.35845879325119373, 0.38532046640756762, 0.41246312944140484, + 0.43991316567323391, 0.46769879911450835, 0.49585034734745348, + 0.52440051270804111, 0.55338471955567303, 0.58284150727121620, + 0.61281299101662701, 0.64334540539291685, 0.67448975019608171, + 0.70630256284008752, 0.73884684918521371, 0.77219321418868492, + 0.80642124701824036, 0.84162123357291441, 0.87789629505122879, + 0.91536508784281423, 0.95416525314619460, 0.99445788320975348, + 1.03643338949378938, 1.08031934081495606, 1.12639112903880045, + 1.17498679206608991, 1.22652812003661049, 1.28155156554460081, + 1.34075503369021654, 1.40507156030963265, 1.47579102817917085, + 1.55477359459685394, 1.64485362695147308, 1.75068607125217102, + 1.88079360815125041, 2.05374891063182208, 2.32634787404084076 +}; + +TEST_BEGIN(test_pt_norm) { + unsigned i; + + for (i = 1; i < sizeof(pt_norm_expected)/sizeof(double); i++) { + double p = (double)i * 0.01; + assert_true(double_eq_rel(pt_norm(p), pt_norm_expected[i], + MAX_REL_ERR, MAX_ABS_ERR), + "Incorrect pt_norm result for i=%u", i); + } +} +TEST_END + +/* + * Expected pt_chi2(p=[0.01..0.99] increment=0.07, + * df={0.1, 1.1, 10.1, 100.1, 1000.1}). + */ +static const double pt_chi2_df[] = {0.1, 1.1, 10.1, 100.1, 1000.1}; +static const double pt_chi2_expected[] = { + 1.168926411457320e-40, 1.347680397072034e-22, 3.886980416666260e-17, + 8.245951724356564e-14, 2.068936347497604e-11, 1.562561743309233e-09, + 5.459543043426564e-08, 1.114775688149252e-06, 1.532101202364371e-05, + 1.553884683726585e-04, 1.239396954915939e-03, 8.153872320255721e-03, + 4.631183739647523e-02, 2.473187311701327e-01, 2.175254800183617e+00, + + 0.0003729887888876379, 0.0164409238228929513, 0.0521523015190650113, + 0.1064701372271216612, 0.1800913735793082115, 0.2748704281195626931, + 0.3939246282787986497, 0.5420727552260817816, 0.7267265822221973259, + 0.9596554296000253670, 1.2607440376386165326, 1.6671185084541604304, + 2.2604828984738705167, 3.2868613342148607082, 6.9298574921692139839, + + 2.606673548632508, 4.602913725294877, 5.646152813924212, + 6.488971315540869, 7.249823275816285, 7.977314231410841, + 8.700354939944047, 9.441728024225892, 10.224338321374127, + 11.076435368801061, 12.039320937038386, 13.183878752697167, + 14.657791935084575, 16.885728216339373, 23.361991680031817, + + 70.14844087392152, 80.92379498849355, 85.53325420085891, + 88.94433120715347, 91.83732712857017, 94.46719943606301, + 96.96896479994635, 99.43412843510363, 101.94074719829733, + 104.57228644307247, 107.43900093448734, 110.71844673417287, + 114.76616819871325, 120.57422505959563, 135.92318818757556, + + 899.0072447849649, 937.9271278858220, 953.8117189560207, + 965.3079371501154, 974.8974061207954, 983.4936235182347, + 991.5691170518946, 999.4334123954690, 1007.3391826856553, + 1015.5445154999951, 1024.3777075619569, 1034.3538789836223, + 1046.4872561869577, 1063.5717461999654, 1107.0741966053859 +}; + +TEST_BEGIN(test_pt_chi2) { + unsigned i, j; + unsigned e = 0; + + for (i = 0; i < sizeof(pt_chi2_df)/sizeof(double); i++) { + double df = pt_chi2_df[i]; + double ln_gamma_df = ln_gamma(df * 0.5); + for (j = 1; j < 100; j += 7) { + double p = (double)j * 0.01; + assert_true(double_eq_rel(pt_chi2(p, df, ln_gamma_df), + pt_chi2_expected[e], MAX_REL_ERR, MAX_ABS_ERR), + "Incorrect pt_chi2 result for i=%u, j=%u", i, j); + e++; + } + } +} +TEST_END + +/* + * Expected pt_gamma(p=[0.1..0.99] increment=0.07, + * shape=[0.5..3.0] increment=0.5). + */ +static const double pt_gamma_shape[] = {0.5, 1.0, 1.5, 2.0, 2.5, 3.0}; +static const double pt_gamma_expected[] = { + 7.854392895485103e-05, 5.043466107888016e-03, 1.788288957794883e-02, + 3.900956150232906e-02, 6.913847560638034e-02, 1.093710833465766e-01, + 1.613412523825817e-01, 2.274682115597864e-01, 3.114117323127083e-01, + 4.189466220207417e-01, 5.598106789059246e-01, 7.521856146202706e-01, + 1.036125427911119e+00, 1.532450860038180e+00, 3.317448300510606e+00, + + 0.01005033585350144, 0.08338160893905107, 0.16251892949777497, + 0.24846135929849966, 0.34249030894677596, 0.44628710262841947, + 0.56211891815354142, 0.69314718055994529, 0.84397007029452920, + 1.02165124753198167, 1.23787435600161766, 1.51412773262977574, + 1.89711998488588196, 2.52572864430825783, 4.60517018598809091, + + 0.05741590094955853, 0.24747378084860744, 0.39888572212236084, + 0.54394139997444901, 0.69048812513915159, 0.84311389861296104, + 1.00580622221479898, 1.18298694218766931, 1.38038096305861213, + 1.60627736383027453, 1.87396970522337947, 2.20749220408081070, + 2.65852391865854942, 3.37934630984842244, 5.67243336507218476, + + 0.1485547402532659, 0.4657458011640391, 0.6832386130709406, + 0.8794297834672100, 1.0700752852474524, 1.2629614217350744, + 1.4638400448580779, 1.6783469900166610, 1.9132338090606940, + 2.1778589228618777, 2.4868823970010991, 2.8664695666264195, + 3.3724415436062114, 4.1682658512758071, 6.6383520679938108, + + 0.2771490383641385, 0.7195001279643727, 0.9969081732265243, + 1.2383497880608061, 1.4675206597269927, 1.6953064251816552, + 1.9291243435606809, 2.1757300955477641, 2.4428032131216391, + 2.7406534569230616, 3.0851445039665513, 3.5043101122033367, + 4.0575997065264637, 4.9182956424675286, 7.5431362346944937, + + 0.4360451650782932, 0.9983600902486267, 1.3306365880734528, + 1.6129750834753802, 1.8767241606994294, 2.1357032436097660, + 2.3988853336865565, 2.6740603137235603, 2.9697561737517959, + 3.2971457713883265, 3.6731795898504660, 4.1275751617770631, + 4.7230515633946677, 5.6417477865306020, 8.4059469148854635 +}; + +TEST_BEGIN(test_pt_gamma_shape) { + unsigned i, j; + unsigned e = 0; + + for (i = 0; i < sizeof(pt_gamma_shape)/sizeof(double); i++) { + double shape = pt_gamma_shape[i]; + double ln_gamma_shape = ln_gamma(shape); + for (j = 1; j < 100; j += 7) { + double p = (double)j * 0.01; + assert_true(double_eq_rel(pt_gamma(p, shape, 1.0, + ln_gamma_shape), pt_gamma_expected[e], MAX_REL_ERR, + MAX_ABS_ERR), + "Incorrect pt_gamma result for i=%u, j=%u", i, j); + e++; + } + } +} +TEST_END + +TEST_BEGIN(test_pt_gamma_scale) { + double shape = 1.0; + double ln_gamma_shape = ln_gamma(shape); + + assert_true(double_eq_rel( + pt_gamma(0.5, shape, 1.0, ln_gamma_shape) * 10.0, + pt_gamma(0.5, shape, 10.0, ln_gamma_shape), MAX_REL_ERR, + MAX_ABS_ERR), + "Scale should be trivially equivalent to external multiplication"); +} +TEST_END + +int +main(void) { + return test( + test_ln_gamma_factorial, + test_ln_gamma_misc, + test_pt_norm, + test_pt_chi2, + test_pt_gamma_shape, + test_pt_gamma_scale); +} diff --git a/deps/jemalloc/test/unit/mq.c b/deps/jemalloc/test/unit/mq.c new file mode 100644 index 000000000..57a4d54e4 --- /dev/null +++ b/deps/jemalloc/test/unit/mq.c @@ -0,0 +1,89 @@ +#include "test/jemalloc_test.h" + +#define NSENDERS 3 +#define NMSGS 100000 + +typedef struct mq_msg_s mq_msg_t; +struct mq_msg_s { + mq_msg(mq_msg_t) link; +}; +mq_gen(static, mq_, mq_t, mq_msg_t, link) + +TEST_BEGIN(test_mq_basic) { + mq_t mq; + mq_msg_t msg; + + assert_false(mq_init(&mq), "Unexpected mq_init() failure"); + assert_u_eq(mq_count(&mq), 0, "mq should be empty"); + assert_ptr_null(mq_tryget(&mq), + "mq_tryget() should fail when the queue is empty"); + + mq_put(&mq, &msg); + assert_u_eq(mq_count(&mq), 1, "mq should contain one message"); + assert_ptr_eq(mq_tryget(&mq), &msg, "mq_tryget() should return msg"); + + mq_put(&mq, &msg); + assert_ptr_eq(mq_get(&mq), &msg, "mq_get() should return msg"); + + mq_fini(&mq); +} +TEST_END + +static void * +thd_receiver_start(void *arg) { + mq_t *mq = (mq_t *)arg; + unsigned i; + + for (i = 0; i < (NSENDERS * NMSGS); i++) { + mq_msg_t *msg = mq_get(mq); + assert_ptr_not_null(msg, "mq_get() should never return NULL"); + dallocx(msg, 0); + } + return NULL; +} + +static void * +thd_sender_start(void *arg) { + mq_t *mq = (mq_t *)arg; + unsigned i; + + for (i = 0; i < NMSGS; i++) { + mq_msg_t *msg; + void *p; + p = mallocx(sizeof(mq_msg_t), 0); + assert_ptr_not_null(p, "Unexpected mallocx() failure"); + msg = (mq_msg_t *)p; + mq_put(mq, msg); + } + return NULL; +} + +TEST_BEGIN(test_mq_threaded) { + mq_t mq; + thd_t receiver; + thd_t senders[NSENDERS]; + unsigned i; + + assert_false(mq_init(&mq), "Unexpected mq_init() failure"); + + thd_create(&receiver, thd_receiver_start, (void *)&mq); + for (i = 0; i < NSENDERS; i++) { + thd_create(&senders[i], thd_sender_start, (void *)&mq); + } + + thd_join(receiver, NULL); + for (i = 0; i < NSENDERS; i++) { + thd_join(senders[i], NULL); + } + + mq_fini(&mq); +} +TEST_END + +int +main(void) { + return test( + test_mq_basic, + test_mq_threaded); +} + diff --git a/deps/jemalloc/test/unit/mtx.c b/deps/jemalloc/test/unit/mtx.c new file mode 100644 index 000000000..424587b03 --- /dev/null +++ b/deps/jemalloc/test/unit/mtx.c @@ -0,0 +1,57 @@ +#include "test/jemalloc_test.h" + +#define NTHREADS 2 +#define NINCRS 2000000 + +TEST_BEGIN(test_mtx_basic) { + mtx_t mtx; + + assert_false(mtx_init(&mtx), "Unexpected mtx_init() failure"); + mtx_lock(&mtx); + mtx_unlock(&mtx); + mtx_fini(&mtx); +} +TEST_END + +typedef struct { + mtx_t mtx; + unsigned x; +} thd_start_arg_t; + +static void * +thd_start(void *varg) { + thd_start_arg_t *arg = (thd_start_arg_t *)varg; + unsigned i; + + for (i = 0; i < NINCRS; i++) { + mtx_lock(&arg->mtx); + arg->x++; + mtx_unlock(&arg->mtx); + } + return NULL; +} + +TEST_BEGIN(test_mtx_race) { + thd_start_arg_t arg; + thd_t thds[NTHREADS]; + unsigned i; + + assert_false(mtx_init(&arg.mtx), "Unexpected mtx_init() failure"); + arg.x = 0; + for (i = 0; i < NTHREADS; i++) { + thd_create(&thds[i], thd_start, (void *)&arg); + } + for (i = 0; i < NTHREADS; i++) { + thd_join(thds[i], NULL); + } + assert_u_eq(arg.x, NTHREADS * NINCRS, + "Race-related counter corruption"); +} +TEST_END + +int +main(void) { + return test( + test_mtx_basic, + test_mtx_race); +} diff --git a/deps/jemalloc/test/unit/nstime.c b/deps/jemalloc/test/unit/nstime.c new file mode 100644 index 000000000..f31378058 --- /dev/null +++ b/deps/jemalloc/test/unit/nstime.c @@ -0,0 +1,249 @@ +#include "test/jemalloc_test.h" + +#define BILLION UINT64_C(1000000000) + +TEST_BEGIN(test_nstime_init) { + nstime_t nst; + + nstime_init(&nst, 42000000043); + assert_u64_eq(nstime_ns(&nst), 42000000043, "ns incorrectly read"); + assert_u64_eq(nstime_sec(&nst), 42, "sec incorrectly read"); + assert_u64_eq(nstime_nsec(&nst), 43, "nsec incorrectly read"); +} +TEST_END + +TEST_BEGIN(test_nstime_init2) { + nstime_t nst; + + nstime_init2(&nst, 42, 43); + assert_u64_eq(nstime_sec(&nst), 42, "sec incorrectly read"); + assert_u64_eq(nstime_nsec(&nst), 43, "nsec incorrectly read"); +} +TEST_END + +TEST_BEGIN(test_nstime_copy) { + nstime_t nsta, nstb; + + nstime_init2(&nsta, 42, 43); + nstime_init(&nstb, 0); + nstime_copy(&nstb, &nsta); + assert_u64_eq(nstime_sec(&nstb), 42, "sec incorrectly copied"); + assert_u64_eq(nstime_nsec(&nstb), 43, "nsec incorrectly copied"); +} +TEST_END + +TEST_BEGIN(test_nstime_compare) { + nstime_t nsta, nstb; + + nstime_init2(&nsta, 42, 43); + nstime_copy(&nstb, &nsta); + assert_d_eq(nstime_compare(&nsta, &nstb), 0, "Times should be equal"); + assert_d_eq(nstime_compare(&nstb, &nsta), 0, "Times should be equal"); + + nstime_init2(&nstb, 42, 42); + assert_d_eq(nstime_compare(&nsta, &nstb), 1, + "nsta should be greater than nstb"); + assert_d_eq(nstime_compare(&nstb, &nsta), -1, + "nstb should be less than nsta"); + + nstime_init2(&nstb, 42, 44); + assert_d_eq(nstime_compare(&nsta, &nstb), -1, + "nsta should be less than nstb"); + assert_d_eq(nstime_compare(&nstb, &nsta), 1, + "nstb should be greater than nsta"); + + nstime_init2(&nstb, 41, BILLION - 1); + assert_d_eq(nstime_compare(&nsta, &nstb), 1, + "nsta should be greater than nstb"); + assert_d_eq(nstime_compare(&nstb, &nsta), -1, + "nstb should be less than nsta"); + + nstime_init2(&nstb, 43, 0); + assert_d_eq(nstime_compare(&nsta, &nstb), -1, + "nsta should be less than nstb"); + assert_d_eq(nstime_compare(&nstb, &nsta), 1, + "nstb should be greater than nsta"); +} +TEST_END + +TEST_BEGIN(test_nstime_add) { + nstime_t nsta, nstb; + + nstime_init2(&nsta, 42, 43); + nstime_copy(&nstb, &nsta); + nstime_add(&nsta, &nstb); + nstime_init2(&nstb, 84, 86); + assert_d_eq(nstime_compare(&nsta, &nstb), 0, + "Incorrect addition result"); + + nstime_init2(&nsta, 42, BILLION - 1); + nstime_copy(&nstb, &nsta); + nstime_add(&nsta, &nstb); + nstime_init2(&nstb, 85, BILLION - 2); + assert_d_eq(nstime_compare(&nsta, &nstb), 0, + "Incorrect addition result"); +} +TEST_END + +TEST_BEGIN(test_nstime_iadd) { + nstime_t nsta, nstb; + + nstime_init2(&nsta, 42, BILLION - 1); + nstime_iadd(&nsta, 1); + nstime_init2(&nstb, 43, 0); + assert_d_eq(nstime_compare(&nsta, &nstb), 0, + "Incorrect addition result"); + + nstime_init2(&nsta, 42, 1); + nstime_iadd(&nsta, BILLION + 1); + nstime_init2(&nstb, 43, 2); + assert_d_eq(nstime_compare(&nsta, &nstb), 0, + "Incorrect addition result"); +} +TEST_END + +TEST_BEGIN(test_nstime_subtract) { + nstime_t nsta, nstb; + + nstime_init2(&nsta, 42, 43); + nstime_copy(&nstb, &nsta); + nstime_subtract(&nsta, &nstb); + nstime_init(&nstb, 0); + assert_d_eq(nstime_compare(&nsta, &nstb), 0, + "Incorrect subtraction result"); + + nstime_init2(&nsta, 42, 43); + nstime_init2(&nstb, 41, 44); + nstime_subtract(&nsta, &nstb); + nstime_init2(&nstb, 0, BILLION - 1); + assert_d_eq(nstime_compare(&nsta, &nstb), 0, + "Incorrect subtraction result"); +} +TEST_END + +TEST_BEGIN(test_nstime_isubtract) { + nstime_t nsta, nstb; + + nstime_init2(&nsta, 42, 43); + nstime_isubtract(&nsta, 42*BILLION + 43); + nstime_init(&nstb, 0); + assert_d_eq(nstime_compare(&nsta, &nstb), 0, + "Incorrect subtraction result"); + + nstime_init2(&nsta, 42, 43); + nstime_isubtract(&nsta, 41*BILLION + 44); + nstime_init2(&nstb, 0, BILLION - 1); + assert_d_eq(nstime_compare(&nsta, &nstb), 0, + "Incorrect subtraction result"); +} +TEST_END + +TEST_BEGIN(test_nstime_imultiply) { + nstime_t nsta, nstb; + + nstime_init2(&nsta, 42, 43); + nstime_imultiply(&nsta, 10); + nstime_init2(&nstb, 420, 430); + assert_d_eq(nstime_compare(&nsta, &nstb), 0, + "Incorrect multiplication result"); + + nstime_init2(&nsta, 42, 666666666); + nstime_imultiply(&nsta, 3); + nstime_init2(&nstb, 127, 999999998); + assert_d_eq(nstime_compare(&nsta, &nstb), 0, + "Incorrect multiplication result"); +} +TEST_END + +TEST_BEGIN(test_nstime_idivide) { + nstime_t nsta, nstb; + + nstime_init2(&nsta, 42, 43); + nstime_copy(&nstb, &nsta); + nstime_imultiply(&nsta, 10); + nstime_idivide(&nsta, 10); + assert_d_eq(nstime_compare(&nsta, &nstb), 0, + "Incorrect division result"); + + nstime_init2(&nsta, 42, 666666666); + nstime_copy(&nstb, &nsta); + nstime_imultiply(&nsta, 3); + nstime_idivide(&nsta, 3); + assert_d_eq(nstime_compare(&nsta, &nstb), 0, + "Incorrect division result"); +} +TEST_END + +TEST_BEGIN(test_nstime_divide) { + nstime_t nsta, nstb, nstc; + + nstime_init2(&nsta, 42, 43); + nstime_copy(&nstb, &nsta); + nstime_imultiply(&nsta, 10); + assert_u64_eq(nstime_divide(&nsta, &nstb), 10, + "Incorrect division result"); + + nstime_init2(&nsta, 42, 43); + nstime_copy(&nstb, &nsta); + nstime_imultiply(&nsta, 10); + nstime_init(&nstc, 1); + nstime_add(&nsta, &nstc); + assert_u64_eq(nstime_divide(&nsta, &nstb), 10, + "Incorrect division result"); + + nstime_init2(&nsta, 42, 43); + nstime_copy(&nstb, &nsta); + nstime_imultiply(&nsta, 10); + nstime_init(&nstc, 1); + nstime_subtract(&nsta, &nstc); + assert_u64_eq(nstime_divide(&nsta, &nstb), 9, + "Incorrect division result"); +} +TEST_END + +TEST_BEGIN(test_nstime_monotonic) { + nstime_monotonic(); +} +TEST_END + +TEST_BEGIN(test_nstime_update) { + nstime_t nst; + + nstime_init(&nst, 0); + + assert_false(nstime_update(&nst), "Basic time update failed."); + + /* Only Rip Van Winkle sleeps this long. */ + { + nstime_t addend; + nstime_init2(&addend, 631152000, 0); + nstime_add(&nst, &addend); + } + { + nstime_t nst0; + nstime_copy(&nst0, &nst); + assert_true(nstime_update(&nst), + "Update should detect time roll-back."); + assert_d_eq(nstime_compare(&nst, &nst0), 0, + "Time should not have been modified"); + } +} +TEST_END + +int +main(void) { + return test( + test_nstime_init, + test_nstime_init2, + test_nstime_copy, + test_nstime_compare, + test_nstime_add, + test_nstime_iadd, + test_nstime_subtract, + test_nstime_isubtract, + test_nstime_imultiply, + test_nstime_idivide, + test_nstime_divide, + test_nstime_monotonic, + test_nstime_update); +} diff --git a/deps/jemalloc/test/unit/pack.c b/deps/jemalloc/test/unit/pack.c new file mode 100644 index 000000000..fc188b003 --- /dev/null +++ b/deps/jemalloc/test/unit/pack.c @@ -0,0 +1,166 @@ +#include "test/jemalloc_test.h" + +/* + * Size class that is a divisor of the page size, ideally 4+ regions per run. + */ +#if LG_PAGE <= 14 +#define SZ (ZU(1) << (LG_PAGE - 2)) +#else +#define SZ ZU(4096) +#endif + +/* + * Number of slabs to consume at high water mark. Should be at least 2 so that + * if mmap()ed memory grows downward, downward growth of mmap()ed memory is + * tested. + */ +#define NSLABS 8 + +static unsigned +binind_compute(void) { + size_t sz; + unsigned nbins, i; + + sz = sizeof(nbins); + assert_d_eq(mallctl("arenas.nbins", (void *)&nbins, &sz, NULL, 0), 0, + "Unexpected mallctl failure"); + + for (i = 0; i < nbins; i++) { + size_t mib[4]; + size_t miblen = sizeof(mib)/sizeof(size_t); + size_t size; + + assert_d_eq(mallctlnametomib("arenas.bin.0.size", mib, + &miblen), 0, "Unexpected mallctlnametomb failure"); + mib[2] = (size_t)i; + + sz = sizeof(size); + assert_d_eq(mallctlbymib(mib, miblen, (void *)&size, &sz, NULL, + 0), 0, "Unexpected mallctlbymib failure"); + if (size == SZ) { + return i; + } + } + + test_fail("Unable to compute nregs_per_run"); + return 0; +} + +static size_t +nregs_per_run_compute(void) { + uint32_t nregs; + size_t sz; + unsigned binind = binind_compute(); + size_t mib[4]; + size_t miblen = sizeof(mib)/sizeof(size_t); + + assert_d_eq(mallctlnametomib("arenas.bin.0.nregs", mib, &miblen), 0, + "Unexpected mallctlnametomb failure"); + mib[2] = (size_t)binind; + sz = sizeof(nregs); + assert_d_eq(mallctlbymib(mib, miblen, (void *)&nregs, &sz, NULL, + 0), 0, "Unexpected mallctlbymib failure"); + return nregs; +} + +static unsigned +arenas_create_mallctl(void) { + unsigned arena_ind; + size_t sz; + + sz = sizeof(arena_ind); + assert_d_eq(mallctl("arenas.create", (void *)&arena_ind, &sz, NULL, 0), + 0, "Error in arenas.create"); + + return arena_ind; +} + +static void +arena_reset_mallctl(unsigned arena_ind) { + size_t mib[3]; + size_t miblen = sizeof(mib)/sizeof(size_t); + + assert_d_eq(mallctlnametomib("arena.0.reset", mib, &miblen), 0, + "Unexpected mallctlnametomib() failure"); + mib[1] = (size_t)arena_ind; + assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, NULL, 0), 0, + "Unexpected mallctlbymib() failure"); +} + +TEST_BEGIN(test_pack) { + bool prof_enabled; + size_t sz = sizeof(prof_enabled); + if (mallctl("opt.prof", (void *)&prof_enabled, &sz, NULL, 0) == 0) { + test_skip_if(prof_enabled); + } + + unsigned arena_ind = arenas_create_mallctl(); + size_t nregs_per_run = nregs_per_run_compute(); + size_t nregs = nregs_per_run * NSLABS; + VARIABLE_ARRAY(void *, ptrs, nregs); + size_t i, j, offset; + + /* Fill matrix. */ + for (i = offset = 0; i < NSLABS; i++) { + for (j = 0; j < nregs_per_run; j++) { + void *p = mallocx(SZ, MALLOCX_ARENA(arena_ind) | + MALLOCX_TCACHE_NONE); + assert_ptr_not_null(p, + "Unexpected mallocx(%zu, MALLOCX_ARENA(%u) |" + " MALLOCX_TCACHE_NONE) failure, run=%zu, reg=%zu", + SZ, arena_ind, i, j); + ptrs[(i * nregs_per_run) + j] = p; + } + } + + /* + * Free all but one region of each run, but rotate which region is + * preserved, so that subsequent allocations exercise the within-run + * layout policy. + */ + offset = 0; + for (i = offset = 0; + i < NSLABS; + i++, offset = (offset + 1) % nregs_per_run) { + for (j = 0; j < nregs_per_run; j++) { + void *p = ptrs[(i * nregs_per_run) + j]; + if (offset == j) { + continue; + } + dallocx(p, MALLOCX_ARENA(arena_ind) | + MALLOCX_TCACHE_NONE); + } + } + + /* + * Logically refill matrix, skipping preserved regions and verifying + * that the matrix is unmodified. + */ + offset = 0; + for (i = offset = 0; + i < NSLABS; + i++, offset = (offset + 1) % nregs_per_run) { + for (j = 0; j < nregs_per_run; j++) { + void *p; + + if (offset == j) { + continue; + } + p = mallocx(SZ, MALLOCX_ARENA(arena_ind) | + MALLOCX_TCACHE_NONE); + assert_ptr_eq(p, ptrs[(i * nregs_per_run) + j], + "Unexpected refill discrepancy, run=%zu, reg=%zu\n", + i, j); + } + } + + /* Clean up. */ + arena_reset_mallctl(arena_ind); +} +TEST_END + +int +main(void) { + return test( + test_pack); +} diff --git a/deps/jemalloc/test/unit/pack.sh b/deps/jemalloc/test/unit/pack.sh new file mode 100644 index 000000000..6f451480b --- /dev/null +++ b/deps/jemalloc/test/unit/pack.sh @@ -0,0 +1,4 @@ +#!/bin/sh + +# Immediately purge to minimize fragmentation. +export MALLOC_CONF="dirty_decay_ms:0,muzzy_decay_ms:0" diff --git a/deps/jemalloc/test/unit/pages.c b/deps/jemalloc/test/unit/pages.c new file mode 100644 index 000000000..ee729eece --- /dev/null +++ b/deps/jemalloc/test/unit/pages.c @@ -0,0 +1,29 @@ +#include "test/jemalloc_test.h" + +TEST_BEGIN(test_pages_huge) { + size_t alloc_size; + bool commit; + void *pages, *hugepage; + + alloc_size = HUGEPAGE * 2 - PAGE; + commit = true; + pages = pages_map(NULL, alloc_size, PAGE, &commit); + assert_ptr_not_null(pages, "Unexpected pages_map() error"); + + if (init_system_thp_mode == thp_mode_default) { + hugepage = (void *)(ALIGNMENT_CEILING((uintptr_t)pages, HUGEPAGE)); + assert_b_ne(pages_huge(hugepage, HUGEPAGE), have_madvise_huge, + "Unexpected pages_huge() result"); + assert_false(pages_nohuge(hugepage, HUGEPAGE), + "Unexpected pages_nohuge() result"); + } + + pages_unmap(pages, alloc_size); +} +TEST_END + +int +main(void) { + return test( + test_pages_huge); +} diff --git a/deps/jemalloc/test/unit/ph.c b/deps/jemalloc/test/unit/ph.c new file mode 100644 index 000000000..88bf56f88 --- /dev/null +++ b/deps/jemalloc/test/unit/ph.c @@ -0,0 +1,318 @@ +#include "test/jemalloc_test.h" + +#include "jemalloc/internal/ph.h" + +typedef struct node_s node_t; + +struct node_s { +#define NODE_MAGIC 0x9823af7e + uint32_t magic; + phn(node_t) link; + uint64_t key; +}; + +static int +node_cmp(const node_t *a, const node_t *b) { + int ret; + + ret = (a->key > b->key) - (a->key < b->key); + if (ret == 0) { + /* + * Duplicates are not allowed in the heap, so force an + * arbitrary ordering for non-identical items with equal keys. + */ + ret = (((uintptr_t)a) > ((uintptr_t)b)) + - (((uintptr_t)a) < ((uintptr_t)b)); + } + return ret; +} + +static int +node_cmp_magic(const node_t *a, const node_t *b) { + + assert_u32_eq(a->magic, NODE_MAGIC, "Bad magic"); + assert_u32_eq(b->magic, NODE_MAGIC, "Bad magic"); + + return node_cmp(a, b); +} + +typedef ph(node_t) heap_t; +ph_gen(static, heap_, heap_t, node_t, link, node_cmp_magic); + +static void +node_print(const node_t *node, unsigned depth) { + unsigned i; + node_t *leftmost_child, *sibling; + + for (i = 0; i < depth; i++) { + malloc_printf("\t"); + } + malloc_printf("%2"FMTu64"\n", node->key); + + leftmost_child = phn_lchild_get(node_t, link, node); + if (leftmost_child == NULL) { + return; + } + node_print(leftmost_child, depth + 1); + + for (sibling = phn_next_get(node_t, link, leftmost_child); sibling != + NULL; sibling = phn_next_get(node_t, link, sibling)) { + node_print(sibling, depth + 1); + } +} + +static void +heap_print(const heap_t *heap) { + node_t *auxelm; + + malloc_printf("vvv heap %p vvv\n", heap); + if (heap->ph_root == NULL) { + goto label_return; + } + + node_print(heap->ph_root, 0); + + for (auxelm = phn_next_get(node_t, link, heap->ph_root); auxelm != NULL; + auxelm = phn_next_get(node_t, link, auxelm)) { + assert_ptr_eq(phn_next_get(node_t, link, phn_prev_get(node_t, + link, auxelm)), auxelm, + "auxelm's prev doesn't link to auxelm"); + node_print(auxelm, 0); + } + +label_return: + malloc_printf("^^^ heap %p ^^^\n", heap); +} + +static unsigned +node_validate(const node_t *node, const node_t *parent) { + unsigned nnodes = 1; + node_t *leftmost_child, *sibling; + + if (parent != NULL) { + assert_d_ge(node_cmp_magic(node, parent), 0, + "Child is less than parent"); + } + + leftmost_child = phn_lchild_get(node_t, link, node); + if (leftmost_child == NULL) { + return nnodes; + } + assert_ptr_eq((void *)phn_prev_get(node_t, link, leftmost_child), + (void *)node, "Leftmost child does not link to node"); + nnodes += node_validate(leftmost_child, node); + + for (sibling = phn_next_get(node_t, link, leftmost_child); sibling != + NULL; sibling = phn_next_get(node_t, link, sibling)) { + assert_ptr_eq(phn_next_get(node_t, link, phn_prev_get(node_t, + link, sibling)), sibling, + "sibling's prev doesn't link to sibling"); + nnodes += node_validate(sibling, node); + } + return nnodes; +} + +static unsigned +heap_validate(const heap_t *heap) { + unsigned nnodes = 0; + node_t *auxelm; + + if (heap->ph_root == NULL) { + goto label_return; + } + + nnodes += node_validate(heap->ph_root, NULL); + + for (auxelm = phn_next_get(node_t, link, heap->ph_root); auxelm != NULL; + auxelm = phn_next_get(node_t, link, auxelm)) { + assert_ptr_eq(phn_next_get(node_t, link, phn_prev_get(node_t, + link, auxelm)), auxelm, + "auxelm's prev doesn't link to auxelm"); + nnodes += node_validate(auxelm, NULL); + } + +label_return: + if (false) { + heap_print(heap); + } + return nnodes; +} + +TEST_BEGIN(test_ph_empty) { + heap_t heap; + + heap_new(&heap); + assert_true(heap_empty(&heap), "Heap should be empty"); + assert_ptr_null(heap_first(&heap), "Unexpected node"); + assert_ptr_null(heap_any(&heap), "Unexpected node"); +} +TEST_END + +static void +node_remove(heap_t *heap, node_t *node) { + heap_remove(heap, node); + + node->magic = 0; +} + +static node_t * +node_remove_first(heap_t *heap) { + node_t *node = heap_remove_first(heap); + node->magic = 0; + return node; +} + +static node_t * +node_remove_any(heap_t *heap) { + node_t *node = heap_remove_any(heap); + node->magic = 0; + return node; +} + +TEST_BEGIN(test_ph_random) { +#define NNODES 25 +#define NBAGS 250 +#define SEED 42 + sfmt_t *sfmt; + uint64_t bag[NNODES]; + heap_t heap; + node_t nodes[NNODES]; + unsigned i, j, k; + + sfmt = init_gen_rand(SEED); + for (i = 0; i < NBAGS; i++) { + switch (i) { + case 0: + /* Insert in order. */ + for (j = 0; j < NNODES; j++) { + bag[j] = j; + } + break; + case 1: + /* Insert in reverse order. */ + for (j = 0; j < NNODES; j++) { + bag[j] = NNODES - j - 1; + } + break; + default: + for (j = 0; j < NNODES; j++) { + bag[j] = gen_rand64_range(sfmt, NNODES); + } + } + + for (j = 1; j <= NNODES; j++) { + /* Initialize heap and nodes. */ + heap_new(&heap); + assert_u_eq(heap_validate(&heap), 0, + "Incorrect node count"); + for (k = 0; k < j; k++) { + nodes[k].magic = NODE_MAGIC; + nodes[k].key = bag[k]; + } + + /* Insert nodes. */ + for (k = 0; k < j; k++) { + heap_insert(&heap, &nodes[k]); + if (i % 13 == 12) { + assert_ptr_not_null(heap_any(&heap), + "Heap should not be empty"); + /* Trigger merging. */ + assert_ptr_not_null(heap_first(&heap), + "Heap should not be empty"); + } + assert_u_eq(heap_validate(&heap), k + 1, + "Incorrect node count"); + } + + assert_false(heap_empty(&heap), + "Heap should not be empty"); + + /* Remove nodes. */ + switch (i % 6) { + case 0: + for (k = 0; k < j; k++) { + assert_u_eq(heap_validate(&heap), j - k, + "Incorrect node count"); + node_remove(&heap, &nodes[k]); + assert_u_eq(heap_validate(&heap), j - k + - 1, "Incorrect node count"); + } + break; + case 1: + for (k = j; k > 0; k--) { + node_remove(&heap, &nodes[k-1]); + assert_u_eq(heap_validate(&heap), k - 1, + "Incorrect node count"); + } + break; + case 2: { + node_t *prev = NULL; + for (k = 0; k < j; k++) { + node_t *node = node_remove_first(&heap); + assert_u_eq(heap_validate(&heap), j - k + - 1, "Incorrect node count"); + if (prev != NULL) { + assert_d_ge(node_cmp(node, + prev), 0, + "Bad removal order"); + } + prev = node; + } + break; + } case 3: { + node_t *prev = NULL; + for (k = 0; k < j; k++) { + node_t *node = heap_first(&heap); + assert_u_eq(heap_validate(&heap), j - k, + "Incorrect node count"); + if (prev != NULL) { + assert_d_ge(node_cmp(node, + prev), 0, + "Bad removal order"); + } + node_remove(&heap, node); + assert_u_eq(heap_validate(&heap), j - k + - 1, "Incorrect node count"); + prev = node; + } + break; + } case 4: { + for (k = 0; k < j; k++) { + node_remove_any(&heap); + assert_u_eq(heap_validate(&heap), j - k + - 1, "Incorrect node count"); + } + break; + } case 5: { + for (k = 0; k < j; k++) { + node_t *node = heap_any(&heap); + assert_u_eq(heap_validate(&heap), j - k, + "Incorrect node count"); + node_remove(&heap, node); + assert_u_eq(heap_validate(&heap), j - k + - 1, "Incorrect node count"); + } + break; + } default: + not_reached(); + } + + assert_ptr_null(heap_first(&heap), + "Heap should be empty"); + assert_ptr_null(heap_any(&heap), + "Heap should be empty"); + assert_true(heap_empty(&heap), "Heap should be empty"); + } + } + fini_gen_rand(sfmt); +#undef NNODES +#undef SEED +} +TEST_END + +int +main(void) { + return test( + test_ph_empty, + test_ph_random); +} diff --git a/deps/jemalloc/test/unit/prng.c b/deps/jemalloc/test/unit/prng.c new file mode 100644 index 000000000..b5795c2f4 --- /dev/null +++ b/deps/jemalloc/test/unit/prng.c @@ -0,0 +1,237 @@ +#include "test/jemalloc_test.h" + +static void +test_prng_lg_range_u32(bool atomic) { + atomic_u32_t sa, sb; + uint32_t ra, rb; + unsigned lg_range; + + atomic_store_u32(&sa, 42, ATOMIC_RELAXED); + ra = prng_lg_range_u32(&sa, 32, atomic); + atomic_store_u32(&sa, 42, ATOMIC_RELAXED); + rb = prng_lg_range_u32(&sa, 32, atomic); + assert_u32_eq(ra, rb, + "Repeated generation should produce repeated results"); + + atomic_store_u32(&sb, 42, ATOMIC_RELAXED); + rb = prng_lg_range_u32(&sb, 32, atomic); + assert_u32_eq(ra, rb, + "Equivalent generation should produce equivalent results"); + + atomic_store_u32(&sa, 42, ATOMIC_RELAXED); + ra = prng_lg_range_u32(&sa, 32, atomic); + rb = prng_lg_range_u32(&sa, 32, atomic); + assert_u32_ne(ra, rb, + "Full-width results must not immediately repeat"); + + atomic_store_u32(&sa, 42, ATOMIC_RELAXED); + ra = prng_lg_range_u32(&sa, 32, atomic); + for (lg_range = 31; lg_range > 0; lg_range--) { + atomic_store_u32(&sb, 42, ATOMIC_RELAXED); + rb = prng_lg_range_u32(&sb, lg_range, atomic); + assert_u32_eq((rb & (UINT32_C(0xffffffff) << lg_range)), + 0, "High order bits should be 0, lg_range=%u", lg_range); + assert_u32_eq(rb, (ra >> (32 - lg_range)), + "Expected high order bits of full-width result, " + "lg_range=%u", lg_range); + } +} + +static void +test_prng_lg_range_u64(void) { + uint64_t sa, sb, ra, rb; + unsigned lg_range; + + sa = 42; + ra = prng_lg_range_u64(&sa, 64); + sa = 42; + rb = prng_lg_range_u64(&sa, 64); + assert_u64_eq(ra, rb, + "Repeated generation should produce repeated results"); + + sb = 42; + rb = prng_lg_range_u64(&sb, 64); + assert_u64_eq(ra, rb, + "Equivalent generation should produce equivalent results"); + + sa = 42; + ra = prng_lg_range_u64(&sa, 64); + rb = prng_lg_range_u64(&sa, 64); + assert_u64_ne(ra, rb, + "Full-width results must not immediately repeat"); + + sa = 42; + ra = prng_lg_range_u64(&sa, 64); + for (lg_range = 63; lg_range > 0; lg_range--) { + sb = 42; + rb = prng_lg_range_u64(&sb, lg_range); + assert_u64_eq((rb & (UINT64_C(0xffffffffffffffff) << lg_range)), + 0, "High order bits should be 0, lg_range=%u", lg_range); + assert_u64_eq(rb, (ra >> (64 - lg_range)), + "Expected high order bits of full-width result, " + "lg_range=%u", lg_range); + } +} + +static void +test_prng_lg_range_zu(bool atomic) { + atomic_zu_t sa, sb; + size_t ra, rb; + unsigned lg_range; + + atomic_store_zu(&sa, 42, ATOMIC_RELAXED); + ra = prng_lg_range_zu(&sa, ZU(1) << (3 + LG_SIZEOF_PTR), atomic); + atomic_store_zu(&sa, 42, ATOMIC_RELAXED); + rb = prng_lg_range_zu(&sa, ZU(1) << (3 + LG_SIZEOF_PTR), atomic); + assert_zu_eq(ra, rb, + "Repeated generation should produce repeated results"); + + atomic_store_zu(&sb, 42, ATOMIC_RELAXED); + rb = prng_lg_range_zu(&sb, ZU(1) << (3 + LG_SIZEOF_PTR), atomic); + assert_zu_eq(ra, rb, + "Equivalent generation should produce equivalent results"); + + atomic_store_zu(&sa, 42, ATOMIC_RELAXED); + ra = prng_lg_range_zu(&sa, ZU(1) << (3 + LG_SIZEOF_PTR), atomic); + rb = prng_lg_range_zu(&sa, ZU(1) << (3 + LG_SIZEOF_PTR), atomic); + assert_zu_ne(ra, rb, + "Full-width results must not immediately repeat"); + + atomic_store_zu(&sa, 42, ATOMIC_RELAXED); + ra = prng_lg_range_zu(&sa, ZU(1) << (3 + LG_SIZEOF_PTR), atomic); + for (lg_range = (ZU(1) << (3 + LG_SIZEOF_PTR)) - 1; lg_range > 0; + lg_range--) { + atomic_store_zu(&sb, 42, ATOMIC_RELAXED); + rb = prng_lg_range_zu(&sb, lg_range, atomic); + assert_zu_eq((rb & (SIZE_T_MAX << lg_range)), + 0, "High order bits should be 0, lg_range=%u", lg_range); + assert_zu_eq(rb, (ra >> ((ZU(1) << (3 + LG_SIZEOF_PTR)) - + lg_range)), "Expected high order bits of full-width " + "result, lg_range=%u", lg_range); + } +} + +TEST_BEGIN(test_prng_lg_range_u32_nonatomic) { + test_prng_lg_range_u32(false); +} +TEST_END + +TEST_BEGIN(test_prng_lg_range_u32_atomic) { + test_prng_lg_range_u32(true); +} +TEST_END + +TEST_BEGIN(test_prng_lg_range_u64_nonatomic) { + test_prng_lg_range_u64(); +} +TEST_END + +TEST_BEGIN(test_prng_lg_range_zu_nonatomic) { + test_prng_lg_range_zu(false); +} +TEST_END + +TEST_BEGIN(test_prng_lg_range_zu_atomic) { + test_prng_lg_range_zu(true); +} +TEST_END + +static void +test_prng_range_u32(bool atomic) { + uint32_t range; +#define MAX_RANGE 10000000 +#define RANGE_STEP 97 +#define NREPS 10 + + for (range = 2; range < MAX_RANGE; range += RANGE_STEP) { + atomic_u32_t s; + unsigned rep; + + atomic_store_u32(&s, range, ATOMIC_RELAXED); + for (rep = 0; rep < NREPS; rep++) { + uint32_t r = prng_range_u32(&s, range, atomic); + + assert_u32_lt(r, range, "Out of range"); + } + } +} + +static void +test_prng_range_u64(void) { + uint64_t range; +#define MAX_RANGE 10000000 +#define RANGE_STEP 97 +#define NREPS 10 + + for (range = 2; range < MAX_RANGE; range += RANGE_STEP) { + uint64_t s; + unsigned rep; + + s = range; + for (rep = 0; rep < NREPS; rep++) { + uint64_t r = prng_range_u64(&s, range); + + assert_u64_lt(r, range, "Out of range"); + } + } +} + +static void +test_prng_range_zu(bool atomic) { + size_t range; +#define MAX_RANGE 10000000 +#define RANGE_STEP 97 +#define NREPS 10 + + for (range = 2; range < MAX_RANGE; range += RANGE_STEP) { + atomic_zu_t s; + unsigned rep; + + atomic_store_zu(&s, range, ATOMIC_RELAXED); + for (rep = 0; rep < NREPS; rep++) { + size_t r = prng_range_zu(&s, range, atomic); + + assert_zu_lt(r, range, "Out of range"); + } + } +} + +TEST_BEGIN(test_prng_range_u32_nonatomic) { + test_prng_range_u32(false); +} +TEST_END + +TEST_BEGIN(test_prng_range_u32_atomic) { + test_prng_range_u32(true); +} +TEST_END + +TEST_BEGIN(test_prng_range_u64_nonatomic) { + test_prng_range_u64(); +} +TEST_END + +TEST_BEGIN(test_prng_range_zu_nonatomic) { + test_prng_range_zu(false); +} +TEST_END + +TEST_BEGIN(test_prng_range_zu_atomic) { + test_prng_range_zu(true); +} +TEST_END + +int +main(void) { + return test( + test_prng_lg_range_u32_nonatomic, + test_prng_lg_range_u32_atomic, + test_prng_lg_range_u64_nonatomic, + test_prng_lg_range_zu_nonatomic, + test_prng_lg_range_zu_atomic, + test_prng_range_u32_nonatomic, + test_prng_range_u32_atomic, + test_prng_range_u64_nonatomic, + test_prng_range_zu_nonatomic, + test_prng_range_zu_atomic); +} diff --git a/deps/jemalloc/test/unit/prof_accum.c b/deps/jemalloc/test/unit/prof_accum.c new file mode 100644 index 000000000..252200635 --- /dev/null +++ b/deps/jemalloc/test/unit/prof_accum.c @@ -0,0 +1,81 @@ +#include "test/jemalloc_test.h" + +#define NTHREADS 4 +#define NALLOCS_PER_THREAD 50 +#define DUMP_INTERVAL 1 +#define BT_COUNT_CHECK_INTERVAL 5 + +static int +prof_dump_open_intercept(bool propagate_err, const char *filename) { + int fd; + + fd = open("/dev/null", O_WRONLY); + assert_d_ne(fd, -1, "Unexpected open() failure"); + + return fd; +} + +static void * +alloc_from_permuted_backtrace(unsigned thd_ind, unsigned iteration) { + return btalloc(1, thd_ind*NALLOCS_PER_THREAD + iteration); +} + +static void * +thd_start(void *varg) { + unsigned thd_ind = *(unsigned *)varg; + size_t bt_count_prev, bt_count; + unsigned i_prev, i; + + i_prev = 0; + bt_count_prev = 0; + for (i = 0; i < NALLOCS_PER_THREAD; i++) { + void *p = alloc_from_permuted_backtrace(thd_ind, i); + dallocx(p, 0); + if (i % DUMP_INTERVAL == 0) { + assert_d_eq(mallctl("prof.dump", NULL, NULL, NULL, 0), + 0, "Unexpected error while dumping heap profile"); + } + + if (i % BT_COUNT_CHECK_INTERVAL == 0 || + i+1 == NALLOCS_PER_THREAD) { + bt_count = prof_bt_count(); + assert_zu_le(bt_count_prev+(i-i_prev), bt_count, + "Expected larger backtrace count increase"); + i_prev = i; + bt_count_prev = bt_count; + } + } + + return NULL; +} + +TEST_BEGIN(test_idump) { + bool active; + thd_t thds[NTHREADS]; + unsigned thd_args[NTHREADS]; + unsigned i; + + test_skip_if(!config_prof); + + active = true; + assert_d_eq(mallctl("prof.active", NULL, NULL, (void *)&active, + sizeof(active)), 0, + "Unexpected mallctl failure while activating profiling"); + + prof_dump_open = prof_dump_open_intercept; + + for (i = 0; i < NTHREADS; i++) { + thd_args[i] = i; + thd_create(&thds[i], thd_start, (void *)&thd_args[i]); + } + for (i = 0; i < NTHREADS; i++) { + thd_join(thds[i], NULL); + } +} +TEST_END + +int +main(void) { + return test_no_reentrancy( + test_idump); +} diff --git a/deps/jemalloc/test/unit/prof_accum.sh b/deps/jemalloc/test/unit/prof_accum.sh new file mode 100644 index 000000000..b3e13fc54 --- /dev/null +++ b/deps/jemalloc/test/unit/prof_accum.sh @@ -0,0 +1,5 @@ +#!/bin/sh + +if [ "x${enable_prof}" = "x1" ] ; then + export MALLOC_CONF="prof:true,prof_accum:true,prof_active:false,lg_prof_sample:0" +fi diff --git a/deps/jemalloc/test/unit/prof_active.c b/deps/jemalloc/test/unit/prof_active.c new file mode 100644 index 000000000..850a24a77 --- /dev/null +++ b/deps/jemalloc/test/unit/prof_active.c @@ -0,0 +1,117 @@ +#include "test/jemalloc_test.h" + +static void +mallctl_bool_get(const char *name, bool expected, const char *func, int line) { + bool old; + size_t sz; + + sz = sizeof(old); + assert_d_eq(mallctl(name, (void *)&old, &sz, NULL, 0), 0, + "%s():%d: Unexpected mallctl failure reading %s", func, line, name); + assert_b_eq(old, expected, "%s():%d: Unexpected %s value", func, line, + name); +} + +static void +mallctl_bool_set(const char *name, bool old_expected, bool val_new, + const char *func, int line) { + bool old; + size_t sz; + + sz = sizeof(old); + assert_d_eq(mallctl(name, (void *)&old, &sz, (void *)&val_new, + sizeof(val_new)), 0, + "%s():%d: Unexpected mallctl failure reading/writing %s", func, + line, name); + assert_b_eq(old, old_expected, "%s():%d: Unexpected %s value", func, + line, name); +} + +static void +mallctl_prof_active_get_impl(bool prof_active_old_expected, const char *func, + int line) { + mallctl_bool_get("prof.active", prof_active_old_expected, func, line); +} +#define mallctl_prof_active_get(a) \ + mallctl_prof_active_get_impl(a, __func__, __LINE__) + +static void +mallctl_prof_active_set_impl(bool prof_active_old_expected, + bool prof_active_new, const char *func, int line) { + mallctl_bool_set("prof.active", prof_active_old_expected, + prof_active_new, func, line); +} +#define mallctl_prof_active_set(a, b) \ + mallctl_prof_active_set_impl(a, b, __func__, __LINE__) + +static void +mallctl_thread_prof_active_get_impl(bool thread_prof_active_old_expected, + const char *func, int line) { + mallctl_bool_get("thread.prof.active", thread_prof_active_old_expected, + func, line); +} +#define mallctl_thread_prof_active_get(a) \ + mallctl_thread_prof_active_get_impl(a, __func__, __LINE__) + +static void +mallctl_thread_prof_active_set_impl(bool thread_prof_active_old_expected, + bool thread_prof_active_new, const char *func, int line) { + mallctl_bool_set("thread.prof.active", thread_prof_active_old_expected, + thread_prof_active_new, func, line); +} +#define mallctl_thread_prof_active_set(a, b) \ + mallctl_thread_prof_active_set_impl(a, b, __func__, __LINE__) + +static void +prof_sampling_probe_impl(bool expect_sample, const char *func, int line) { + void *p; + size_t expected_backtraces = expect_sample ? 1 : 0; + + assert_zu_eq(prof_bt_count(), 0, "%s():%d: Expected 0 backtraces", func, + line); + p = mallocx(1, 0); + assert_ptr_not_null(p, "Unexpected mallocx() failure"); + assert_zu_eq(prof_bt_count(), expected_backtraces, + "%s():%d: Unexpected backtrace count", func, line); + dallocx(p, 0); +} +#define prof_sampling_probe(a) \ + prof_sampling_probe_impl(a, __func__, __LINE__) + +TEST_BEGIN(test_prof_active) { + test_skip_if(!config_prof); + + mallctl_prof_active_get(true); + mallctl_thread_prof_active_get(false); + + mallctl_prof_active_set(true, true); + mallctl_thread_prof_active_set(false, false); + /* prof.active, !thread.prof.active. */ + prof_sampling_probe(false); + + mallctl_prof_active_set(true, false); + mallctl_thread_prof_active_set(false, false); + /* !prof.active, !thread.prof.active. */ + prof_sampling_probe(false); + + mallctl_prof_active_set(false, false); + mallctl_thread_prof_active_set(false, true); + /* !prof.active, thread.prof.active. */ + prof_sampling_probe(false); + + mallctl_prof_active_set(false, true); + mallctl_thread_prof_active_set(true, true); + /* prof.active, thread.prof.active. */ + prof_sampling_probe(true); + + /* Restore settings. */ + mallctl_prof_active_set(true, true); + mallctl_thread_prof_active_set(true, false); +} +TEST_END + +int +main(void) { + return test_no_reentrancy( + test_prof_active); +} diff --git a/deps/jemalloc/test/unit/prof_active.sh b/deps/jemalloc/test/unit/prof_active.sh new file mode 100644 index 000000000..0167cb10b --- /dev/null +++ b/deps/jemalloc/test/unit/prof_active.sh @@ -0,0 +1,5 @@ +#!/bin/sh + +if [ "x${enable_prof}" = "x1" ] ; then + export MALLOC_CONF="prof:true,prof_thread_active_init:false,lg_prof_sample:0" +fi diff --git a/deps/jemalloc/test/unit/prof_gdump.c b/deps/jemalloc/test/unit/prof_gdump.c new file mode 100644 index 000000000..f7e0aac76 --- /dev/null +++ b/deps/jemalloc/test/unit/prof_gdump.c @@ -0,0 +1,74 @@ +#include "test/jemalloc_test.h" + +static bool did_prof_dump_open; + +static int +prof_dump_open_intercept(bool propagate_err, const char *filename) { + int fd; + + did_prof_dump_open = true; + + fd = open("/dev/null", O_WRONLY); + assert_d_ne(fd, -1, "Unexpected open() failure"); + + return fd; +} + +TEST_BEGIN(test_gdump) { + bool active, gdump, gdump_old; + void *p, *q, *r, *s; + size_t sz; + + test_skip_if(!config_prof); + + active = true; + assert_d_eq(mallctl("prof.active", NULL, NULL, (void *)&active, + sizeof(active)), 0, + "Unexpected mallctl failure while activating profiling"); + + prof_dump_open = prof_dump_open_intercept; + + did_prof_dump_open = false; + p = mallocx((1U << SC_LG_LARGE_MINCLASS), 0); + assert_ptr_not_null(p, "Unexpected mallocx() failure"); + assert_true(did_prof_dump_open, "Expected a profile dump"); + + did_prof_dump_open = false; + q = mallocx((1U << SC_LG_LARGE_MINCLASS), 0); + assert_ptr_not_null(q, "Unexpected mallocx() failure"); + assert_true(did_prof_dump_open, "Expected a profile dump"); + + gdump = false; + sz = sizeof(gdump_old); + assert_d_eq(mallctl("prof.gdump", (void *)&gdump_old, &sz, + (void *)&gdump, sizeof(gdump)), 0, + "Unexpected mallctl failure while disabling prof.gdump"); + assert(gdump_old); + did_prof_dump_open = false; + r = mallocx((1U << SC_LG_LARGE_MINCLASS), 0); + assert_ptr_not_null(q, "Unexpected mallocx() failure"); + assert_false(did_prof_dump_open, "Unexpected profile dump"); + + gdump = true; + sz = sizeof(gdump_old); + assert_d_eq(mallctl("prof.gdump", (void *)&gdump_old, &sz, + (void *)&gdump, sizeof(gdump)), 0, + "Unexpected mallctl failure while enabling prof.gdump"); + assert(!gdump_old); + did_prof_dump_open = false; + s = mallocx((1U << SC_LG_LARGE_MINCLASS), 0); + assert_ptr_not_null(q, "Unexpected mallocx() failure"); + assert_true(did_prof_dump_open, "Expected a profile dump"); + + dallocx(p, 0); + dallocx(q, 0); + dallocx(r, 0); + dallocx(s, 0); +} +TEST_END + +int +main(void) { + return test_no_reentrancy( + test_gdump); +} diff --git a/deps/jemalloc/test/unit/prof_gdump.sh b/deps/jemalloc/test/unit/prof_gdump.sh new file mode 100644 index 000000000..3f600d200 --- /dev/null +++ b/deps/jemalloc/test/unit/prof_gdump.sh @@ -0,0 +1,6 @@ +#!/bin/sh + +if [ "x${enable_prof}" = "x1" ] ; then + export MALLOC_CONF="prof:true,prof_active:false,prof_gdump:true" +fi + diff --git a/deps/jemalloc/test/unit/prof_idump.c b/deps/jemalloc/test/unit/prof_idump.c new file mode 100644 index 000000000..1cc6c98cd --- /dev/null +++ b/deps/jemalloc/test/unit/prof_idump.c @@ -0,0 +1,42 @@ +#include "test/jemalloc_test.h" + +static bool did_prof_dump_open; + +static int +prof_dump_open_intercept(bool propagate_err, const char *filename) { + int fd; + + did_prof_dump_open = true; + + fd = open("/dev/null", O_WRONLY); + assert_d_ne(fd, -1, "Unexpected open() failure"); + + return fd; +} + +TEST_BEGIN(test_idump) { + bool active; + void *p; + + test_skip_if(!config_prof); + + active = true; + assert_d_eq(mallctl("prof.active", NULL, NULL, (void *)&active, + sizeof(active)), 0, + "Unexpected mallctl failure while activating profiling"); + + prof_dump_open = prof_dump_open_intercept; + + did_prof_dump_open = false; + p = mallocx(1, 0); + assert_ptr_not_null(p, "Unexpected mallocx() failure"); + dallocx(p, 0); + assert_true(did_prof_dump_open, "Expected a profile dump"); +} +TEST_END + +int +main(void) { + return test( + test_idump); +} diff --git a/deps/jemalloc/test/unit/prof_idump.sh b/deps/jemalloc/test/unit/prof_idump.sh new file mode 100644 index 000000000..4dc599a30 --- /dev/null +++ b/deps/jemalloc/test/unit/prof_idump.sh @@ -0,0 +1,8 @@ +#!/bin/sh + +export MALLOC_CONF="tcache:false" +if [ "x${enable_prof}" = "x1" ] ; then + export MALLOC_CONF="${MALLOC_CONF},prof:true,prof_accum:true,prof_active:false,lg_prof_sample:0,lg_prof_interval:0" +fi + + diff --git a/deps/jemalloc/test/unit/prof_log.c b/deps/jemalloc/test/unit/prof_log.c new file mode 100644 index 000000000..92fbd7cea --- /dev/null +++ b/deps/jemalloc/test/unit/prof_log.c @@ -0,0 +1,148 @@ +#include "test/jemalloc_test.h" + +#define N_PARAM 100 +#define N_THREADS 10 + +static void assert_rep() { + assert_b_eq(prof_log_rep_check(), false, "Rep check failed"); +} + +static void assert_log_empty() { + assert_zu_eq(prof_log_bt_count(), 0, + "The log has backtraces; it isn't empty"); + assert_zu_eq(prof_log_thr_count(), 0, + "The log has threads; it isn't empty"); + assert_zu_eq(prof_log_alloc_count(), 0, + "The log has allocations; it isn't empty"); +} + +void *buf[N_PARAM]; + +static void f() { + int i; + for (i = 0; i < N_PARAM; i++) { + buf[i] = malloc(100); + } + for (i = 0; i < N_PARAM; i++) { + free(buf[i]); + } +} + +TEST_BEGIN(test_prof_log_many_logs) { + int i; + + test_skip_if(!config_prof); + + for (i = 0; i < N_PARAM; i++) { + assert_b_eq(prof_log_is_logging(), false, + "Logging shouldn't have started yet"); + assert_d_eq(mallctl("prof.log_start", NULL, NULL, NULL, 0), 0, + "Unexpected mallctl failure when starting logging"); + assert_b_eq(prof_log_is_logging(), true, + "Logging should be started by now"); + assert_log_empty(); + assert_rep(); + f(); + assert_zu_eq(prof_log_thr_count(), 1, "Wrong thread count"); + assert_rep(); + assert_b_eq(prof_log_is_logging(), true, + "Logging should still be on"); + assert_d_eq(mallctl("prof.log_stop", NULL, NULL, NULL, 0), 0, + "Unexpected mallctl failure when stopping logging"); + assert_b_eq(prof_log_is_logging(), false, + "Logging should have turned off"); + } +} +TEST_END + +thd_t thr_buf[N_THREADS]; + +static void *f_thread(void *unused) { + int i; + for (i = 0; i < N_PARAM; i++) { + void *p = malloc(100); + memset(p, 100, sizeof(char)); + free(p); + } + + return NULL; +} + +TEST_BEGIN(test_prof_log_many_threads) { + + test_skip_if(!config_prof); + + int i; + assert_d_eq(mallctl("prof.log_start", NULL, NULL, NULL, 0), 0, + "Unexpected mallctl failure when starting logging"); + for (i = 0; i < N_THREADS; i++) { + thd_create(&thr_buf[i], &f_thread, NULL); + } + + for (i = 0; i < N_THREADS; i++) { + thd_join(thr_buf[i], NULL); + } + assert_zu_eq(prof_log_thr_count(), N_THREADS, + "Wrong number of thread entries"); + assert_rep(); + assert_d_eq(mallctl("prof.log_stop", NULL, NULL, NULL, 0), 0, + "Unexpected mallctl failure when stopping logging"); +} +TEST_END + +static void f3() { + void *p = malloc(100); + free(p); +} + +static void f1() { + void *p = malloc(100); + f3(); + free(p); +} + +static void f2() { + void *p = malloc(100); + free(p); +} + +TEST_BEGIN(test_prof_log_many_traces) { + + test_skip_if(!config_prof); + + assert_d_eq(mallctl("prof.log_start", NULL, NULL, NULL, 0), 0, + "Unexpected mallctl failure when starting logging"); + int i; + assert_rep(); + assert_log_empty(); + for (i = 0; i < N_PARAM; i++) { + assert_rep(); + f1(); + assert_rep(); + f2(); + assert_rep(); + f3(); + assert_rep(); + } + /* + * There should be 8 total backtraces: two for malloc/free in f1(), two + * for malloc/free in f2(), two for malloc/free in f3(), and then two + * for malloc/free in f1()'s call to f3(). However compiler + * optimizations such as loop unrolling might generate more call sites. + * So >= 8 traces are expected. + */ + assert_zu_ge(prof_log_bt_count(), 8, + "Expect at least 8 backtraces given sample workload"); + assert_d_eq(mallctl("prof.log_stop", NULL, NULL, NULL, 0), 0, + "Unexpected mallctl failure when stopping logging"); +} +TEST_END + +int +main(void) { + prof_log_dummy_set(true); + return test_no_reentrancy( + test_prof_log_many_logs, + test_prof_log_many_traces, + test_prof_log_many_threads); +} diff --git a/deps/jemalloc/test/unit/prof_log.sh b/deps/jemalloc/test/unit/prof_log.sh new file mode 100644 index 000000000..8fcc7d8a7 --- /dev/null +++ b/deps/jemalloc/test/unit/prof_log.sh @@ -0,0 +1,5 @@ +#!/bin/sh + +if [ "x${enable_prof}" = "x1" ] ; then + export MALLOC_CONF="prof:true,lg_prof_sample:0" +fi diff --git a/deps/jemalloc/test/unit/prof_reset.c b/deps/jemalloc/test/unit/prof_reset.c new file mode 100644 index 000000000..7cce42d27 --- /dev/null +++ b/deps/jemalloc/test/unit/prof_reset.c @@ -0,0 +1,286 @@ +#include "test/jemalloc_test.h" + +static int +prof_dump_open_intercept(bool propagate_err, const char *filename) { + int fd; + + fd = open("/dev/null", O_WRONLY); + assert_d_ne(fd, -1, "Unexpected open() failure"); + + return fd; +} + +static void +set_prof_active(bool active) { + assert_d_eq(mallctl("prof.active", NULL, NULL, (void *)&active, + sizeof(active)), 0, "Unexpected mallctl failure"); +} + +static size_t +get_lg_prof_sample(void) { + size_t lg_prof_sample; + size_t sz = sizeof(size_t); + + assert_d_eq(mallctl("prof.lg_sample", (void *)&lg_prof_sample, &sz, + NULL, 0), 0, + "Unexpected mallctl failure while reading profiling sample rate"); + return lg_prof_sample; +} + +static void +do_prof_reset(size_t lg_prof_sample) { + assert_d_eq(mallctl("prof.reset", NULL, NULL, + (void *)&lg_prof_sample, sizeof(size_t)), 0, + "Unexpected mallctl failure while resetting profile data"); + assert_zu_eq(lg_prof_sample, get_lg_prof_sample(), + "Expected profile sample rate change"); +} + +TEST_BEGIN(test_prof_reset_basic) { + size_t lg_prof_sample_orig, lg_prof_sample, lg_prof_sample_next; + size_t sz; + unsigned i; + + test_skip_if(!config_prof); + + sz = sizeof(size_t); + assert_d_eq(mallctl("opt.lg_prof_sample", (void *)&lg_prof_sample_orig, + &sz, NULL, 0), 0, + "Unexpected mallctl failure while reading profiling sample rate"); + assert_zu_eq(lg_prof_sample_orig, 0, + "Unexpected profiling sample rate"); + lg_prof_sample = get_lg_prof_sample(); + assert_zu_eq(lg_prof_sample_orig, lg_prof_sample, + "Unexpected disagreement between \"opt.lg_prof_sample\" and " + "\"prof.lg_sample\""); + + /* Test simple resets. */ + for (i = 0; i < 2; i++) { + assert_d_eq(mallctl("prof.reset", NULL, NULL, NULL, 0), 0, + "Unexpected mallctl failure while resetting profile data"); + lg_prof_sample = get_lg_prof_sample(); + assert_zu_eq(lg_prof_sample_orig, lg_prof_sample, + "Unexpected profile sample rate change"); + } + + /* Test resets with prof.lg_sample changes. */ + lg_prof_sample_next = 1; + for (i = 0; i < 2; i++) { + do_prof_reset(lg_prof_sample_next); + lg_prof_sample = get_lg_prof_sample(); + assert_zu_eq(lg_prof_sample, lg_prof_sample_next, + "Expected profile sample rate change"); + lg_prof_sample_next = lg_prof_sample_orig; + } + + /* Make sure the test code restored prof.lg_sample. */ + lg_prof_sample = get_lg_prof_sample(); + assert_zu_eq(lg_prof_sample_orig, lg_prof_sample, + "Unexpected disagreement between \"opt.lg_prof_sample\" and " + "\"prof.lg_sample\""); +} +TEST_END + +bool prof_dump_header_intercepted = false; +prof_cnt_t cnt_all_copy = {0, 0, 0, 0}; +static bool +prof_dump_header_intercept(tsdn_t *tsdn, bool propagate_err, + const prof_cnt_t *cnt_all) { + prof_dump_header_intercepted = true; + memcpy(&cnt_all_copy, cnt_all, sizeof(prof_cnt_t)); + + return false; +} + +TEST_BEGIN(test_prof_reset_cleanup) { + void *p; + prof_dump_header_t *prof_dump_header_orig; + + test_skip_if(!config_prof); + + set_prof_active(true); + + assert_zu_eq(prof_bt_count(), 0, "Expected 0 backtraces"); + p = mallocx(1, 0); + assert_ptr_not_null(p, "Unexpected mallocx() failure"); + assert_zu_eq(prof_bt_count(), 1, "Expected 1 backtrace"); + + prof_dump_header_orig = prof_dump_header; + prof_dump_header = prof_dump_header_intercept; + assert_false(prof_dump_header_intercepted, "Unexpected intercept"); + + assert_d_eq(mallctl("prof.dump", NULL, NULL, NULL, 0), + 0, "Unexpected error while dumping heap profile"); + assert_true(prof_dump_header_intercepted, "Expected intercept"); + assert_u64_eq(cnt_all_copy.curobjs, 1, "Expected 1 allocation"); + + assert_d_eq(mallctl("prof.reset", NULL, NULL, NULL, 0), 0, + "Unexpected error while resetting heap profile data"); + assert_d_eq(mallctl("prof.dump", NULL, NULL, NULL, 0), + 0, "Unexpected error while dumping heap profile"); + assert_u64_eq(cnt_all_copy.curobjs, 0, "Expected 0 allocations"); + assert_zu_eq(prof_bt_count(), 1, "Expected 1 backtrace"); + + prof_dump_header = prof_dump_header_orig; + + dallocx(p, 0); + assert_zu_eq(prof_bt_count(), 0, "Expected 0 backtraces"); + + set_prof_active(false); +} +TEST_END + +#define NTHREADS 4 +#define NALLOCS_PER_THREAD (1U << 13) +#define OBJ_RING_BUF_COUNT 1531 +#define RESET_INTERVAL (1U << 10) +#define DUMP_INTERVAL 3677 +static void * +thd_start(void *varg) { + unsigned thd_ind = *(unsigned *)varg; + unsigned i; + void *objs[OBJ_RING_BUF_COUNT]; + + memset(objs, 0, sizeof(objs)); + + for (i = 0; i < NALLOCS_PER_THREAD; i++) { + if (i % RESET_INTERVAL == 0) { + assert_d_eq(mallctl("prof.reset", NULL, NULL, NULL, 0), + 0, "Unexpected error while resetting heap profile " + "data"); + } + + if (i % DUMP_INTERVAL == 0) { + assert_d_eq(mallctl("prof.dump", NULL, NULL, NULL, 0), + 0, "Unexpected error while dumping heap profile"); + } + + { + void **pp = &objs[i % OBJ_RING_BUF_COUNT]; + if (*pp != NULL) { + dallocx(*pp, 0); + *pp = NULL; + } + *pp = btalloc(1, thd_ind*NALLOCS_PER_THREAD + i); + assert_ptr_not_null(*pp, + "Unexpected btalloc() failure"); + } + } + + /* Clean up any remaining objects. */ + for (i = 0; i < OBJ_RING_BUF_COUNT; i++) { + void **pp = &objs[i % OBJ_RING_BUF_COUNT]; + if (*pp != NULL) { + dallocx(*pp, 0); + *pp = NULL; + } + } + + return NULL; +} + +TEST_BEGIN(test_prof_reset) { + size_t lg_prof_sample_orig; + thd_t thds[NTHREADS]; + unsigned thd_args[NTHREADS]; + unsigned i; + size_t bt_count, tdata_count; + + test_skip_if(!config_prof); + + bt_count = prof_bt_count(); + assert_zu_eq(bt_count, 0, + "Unexpected pre-existing tdata structures"); + tdata_count = prof_tdata_count(); + + lg_prof_sample_orig = get_lg_prof_sample(); + do_prof_reset(5); + + set_prof_active(true); + + for (i = 0; i < NTHREADS; i++) { + thd_args[i] = i; + thd_create(&thds[i], thd_start, (void *)&thd_args[i]); + } + for (i = 0; i < NTHREADS; i++) { + thd_join(thds[i], NULL); + } + + assert_zu_eq(prof_bt_count(), bt_count, + "Unexpected bactrace count change"); + assert_zu_eq(prof_tdata_count(), tdata_count, + "Unexpected remaining tdata structures"); + + set_prof_active(false); + + do_prof_reset(lg_prof_sample_orig); +} +TEST_END +#undef NTHREADS +#undef NALLOCS_PER_THREAD +#undef OBJ_RING_BUF_COUNT +#undef RESET_INTERVAL +#undef DUMP_INTERVAL + +/* Test sampling at the same allocation site across resets. */ +#define NITER 10 +TEST_BEGIN(test_xallocx) { + size_t lg_prof_sample_orig; + unsigned i; + void *ptrs[NITER]; + + test_skip_if(!config_prof); + + lg_prof_sample_orig = get_lg_prof_sample(); + set_prof_active(true); + + /* Reset profiling. */ + do_prof_reset(0); + + for (i = 0; i < NITER; i++) { + void *p; + size_t sz, nsz; + + /* Reset profiling. */ + do_prof_reset(0); + + /* Allocate small object (which will be promoted). */ + p = ptrs[i] = mallocx(1, 0); + assert_ptr_not_null(p, "Unexpected mallocx() failure"); + + /* Reset profiling. */ + do_prof_reset(0); + + /* Perform successful xallocx(). */ + sz = sallocx(p, 0); + assert_zu_eq(xallocx(p, sz, 0, 0), sz, + "Unexpected xallocx() failure"); + + /* Perform unsuccessful xallocx(). */ + nsz = nallocx(sz+1, 0); + assert_zu_eq(xallocx(p, nsz, 0, 0), sz, + "Unexpected xallocx() success"); + } + + for (i = 0; i < NITER; i++) { + /* dallocx. */ + dallocx(ptrs[i], 0); + } + + set_prof_active(false); + do_prof_reset(lg_prof_sample_orig); +} +TEST_END +#undef NITER + +int +main(void) { + /* Intercept dumping prior to running any tests. */ + prof_dump_open = prof_dump_open_intercept; + + return test_no_reentrancy( + test_prof_reset_basic, + test_prof_reset_cleanup, + test_prof_reset, + test_xallocx); +} diff --git a/deps/jemalloc/test/unit/prof_reset.sh b/deps/jemalloc/test/unit/prof_reset.sh new file mode 100644 index 000000000..43c516a08 --- /dev/null +++ b/deps/jemalloc/test/unit/prof_reset.sh @@ -0,0 +1,5 @@ +#!/bin/sh + +if [ "x${enable_prof}" = "x1" ] ; then + export MALLOC_CONF="prof:true,prof_active:false,lg_prof_sample:0" +fi diff --git a/deps/jemalloc/test/unit/prof_tctx.c b/deps/jemalloc/test/unit/prof_tctx.c new file mode 100644 index 000000000..ff3b2b0ca --- /dev/null +++ b/deps/jemalloc/test/unit/prof_tctx.c @@ -0,0 +1,46 @@ +#include "test/jemalloc_test.h" + +TEST_BEGIN(test_prof_realloc) { + tsdn_t *tsdn; + int flags; + void *p, *q; + prof_tctx_t *tctx_p, *tctx_q; + uint64_t curobjs_0, curobjs_1, curobjs_2, curobjs_3; + + test_skip_if(!config_prof); + + tsdn = tsdn_fetch(); + flags = MALLOCX_TCACHE_NONE; + + prof_cnt_all(&curobjs_0, NULL, NULL, NULL); + p = mallocx(1024, flags); + assert_ptr_not_null(p, "Unexpected mallocx() failure"); + tctx_p = prof_tctx_get(tsdn, p, NULL); + assert_ptr_ne(tctx_p, (prof_tctx_t *)(uintptr_t)1U, + "Expected valid tctx"); + prof_cnt_all(&curobjs_1, NULL, NULL, NULL); + assert_u64_eq(curobjs_0 + 1, curobjs_1, + "Allocation should have increased sample size"); + + q = rallocx(p, 2048, flags); + assert_ptr_ne(p, q, "Expected move"); + assert_ptr_not_null(p, "Unexpected rmallocx() failure"); + tctx_q = prof_tctx_get(tsdn, q, NULL); + assert_ptr_ne(tctx_q, (prof_tctx_t *)(uintptr_t)1U, + "Expected valid tctx"); + prof_cnt_all(&curobjs_2, NULL, NULL, NULL); + assert_u64_eq(curobjs_1, curobjs_2, + "Reallocation should not have changed sample size"); + + dallocx(q, flags); + prof_cnt_all(&curobjs_3, NULL, NULL, NULL); + assert_u64_eq(curobjs_0, curobjs_3, + "Sample size should have returned to base level"); +} +TEST_END + +int +main(void) { + return test_no_reentrancy( + test_prof_realloc); +} diff --git a/deps/jemalloc/test/unit/prof_tctx.sh b/deps/jemalloc/test/unit/prof_tctx.sh new file mode 100644 index 000000000..8fcc7d8a7 --- /dev/null +++ b/deps/jemalloc/test/unit/prof_tctx.sh @@ -0,0 +1,5 @@ +#!/bin/sh + +if [ "x${enable_prof}" = "x1" ] ; then + export MALLOC_CONF="prof:true,lg_prof_sample:0" +fi diff --git a/deps/jemalloc/test/unit/prof_thread_name.c b/deps/jemalloc/test/unit/prof_thread_name.c new file mode 100644 index 000000000..c9c2a2b76 --- /dev/null +++ b/deps/jemalloc/test/unit/prof_thread_name.c @@ -0,0 +1,120 @@ +#include "test/jemalloc_test.h" + +static void +mallctl_thread_name_get_impl(const char *thread_name_expected, const char *func, + int line) { + const char *thread_name_old; + size_t sz; + + sz = sizeof(thread_name_old); + assert_d_eq(mallctl("thread.prof.name", (void *)&thread_name_old, &sz, + NULL, 0), 0, + "%s():%d: Unexpected mallctl failure reading thread.prof.name", + func, line); + assert_str_eq(thread_name_old, thread_name_expected, + "%s():%d: Unexpected thread.prof.name value", func, line); +} +#define mallctl_thread_name_get(a) \ + mallctl_thread_name_get_impl(a, __func__, __LINE__) + +static void +mallctl_thread_name_set_impl(const char *thread_name, const char *func, + int line) { + assert_d_eq(mallctl("thread.prof.name", NULL, NULL, + (void *)&thread_name, sizeof(thread_name)), 0, + "%s():%d: Unexpected mallctl failure reading thread.prof.name", + func, line); + mallctl_thread_name_get_impl(thread_name, func, line); +} +#define mallctl_thread_name_set(a) \ + mallctl_thread_name_set_impl(a, __func__, __LINE__) + +TEST_BEGIN(test_prof_thread_name_validation) { + const char *thread_name; + + test_skip_if(!config_prof); + + mallctl_thread_name_get(""); + mallctl_thread_name_set("hi there"); + + /* NULL input shouldn't be allowed. */ + thread_name = NULL; + assert_d_eq(mallctl("thread.prof.name", NULL, NULL, + (void *)&thread_name, sizeof(thread_name)), EFAULT, + "Unexpected mallctl result writing \"%s\" to thread.prof.name", + thread_name); + + /* '\n' shouldn't be allowed. */ + thread_name = "hi\nthere"; + assert_d_eq(mallctl("thread.prof.name", NULL, NULL, + (void *)&thread_name, sizeof(thread_name)), EFAULT, + "Unexpected mallctl result writing \"%s\" to thread.prof.name", + thread_name); + + /* Simultaneous read/write shouldn't be allowed. */ + { + const char *thread_name_old; + size_t sz; + + sz = sizeof(thread_name_old); + assert_d_eq(mallctl("thread.prof.name", + (void *)&thread_name_old, &sz, (void *)&thread_name, + sizeof(thread_name)), EPERM, + "Unexpected mallctl result writing \"%s\" to " + "thread.prof.name", thread_name); + } + + mallctl_thread_name_set(""); +} +TEST_END + +#define NTHREADS 4 +#define NRESET 25 +static void * +thd_start(void *varg) { + unsigned thd_ind = *(unsigned *)varg; + char thread_name[16] = ""; + unsigned i; + + malloc_snprintf(thread_name, sizeof(thread_name), "thread %u", thd_ind); + + mallctl_thread_name_get(""); + mallctl_thread_name_set(thread_name); + + for (i = 0; i < NRESET; i++) { + assert_d_eq(mallctl("prof.reset", NULL, NULL, NULL, 0), 0, + "Unexpected error while resetting heap profile data"); + mallctl_thread_name_get(thread_name); + } + + mallctl_thread_name_set(thread_name); + mallctl_thread_name_set(""); + + return NULL; +} + +TEST_BEGIN(test_prof_thread_name_threaded) { + thd_t thds[NTHREADS]; + unsigned thd_args[NTHREADS]; + unsigned i; + + test_skip_if(!config_prof); + + for (i = 0; i < NTHREADS; i++) { + thd_args[i] = i; + thd_create(&thds[i], thd_start, (void *)&thd_args[i]); + } + for (i = 0; i < NTHREADS; i++) { + thd_join(thds[i], NULL); + } +} +TEST_END +#undef NTHREADS +#undef NRESET + +int +main(void) { + return test( + test_prof_thread_name_validation, + test_prof_thread_name_threaded); +} diff --git a/deps/jemalloc/test/unit/prof_thread_name.sh b/deps/jemalloc/test/unit/prof_thread_name.sh new file mode 100644 index 000000000..298c1058e --- /dev/null +++ b/deps/jemalloc/test/unit/prof_thread_name.sh @@ -0,0 +1,5 @@ +#!/bin/sh + +if [ "x${enable_prof}" = "x1" ] ; then + export MALLOC_CONF="prof:true,prof_active:false" +fi diff --git a/deps/jemalloc/test/unit/ql.c b/deps/jemalloc/test/unit/ql.c new file mode 100644 index 000000000..b76c24c41 --- /dev/null +++ b/deps/jemalloc/test/unit/ql.c @@ -0,0 +1,204 @@ +#include "test/jemalloc_test.h" + +#include "jemalloc/internal/ql.h" + +/* Number of ring entries, in [2..26]. */ +#define NENTRIES 9 + +typedef struct list_s list_t; +typedef ql_head(list_t) list_head_t; + +struct list_s { + ql_elm(list_t) link; + char id; +}; + +static void +test_empty_list(list_head_t *head) { + list_t *t; + unsigned i; + + assert_ptr_null(ql_first(head), "Unexpected element for empty list"); + assert_ptr_null(ql_last(head, link), + "Unexpected element for empty list"); + + i = 0; + ql_foreach(t, head, link) { + i++; + } + assert_u_eq(i, 0, "Unexpected element for empty list"); + + i = 0; + ql_reverse_foreach(t, head, link) { + i++; + } + assert_u_eq(i, 0, "Unexpected element for empty list"); +} + +TEST_BEGIN(test_ql_empty) { + list_head_t head; + + ql_new(&head); + test_empty_list(&head); +} +TEST_END + +static void +init_entries(list_t *entries, unsigned nentries) { + unsigned i; + + for (i = 0; i < nentries; i++) { + entries[i].id = 'a' + i; + ql_elm_new(&entries[i], link); + } +} + +static void +test_entries_list(list_head_t *head, list_t *entries, unsigned nentries) { + list_t *t; + unsigned i; + + assert_c_eq(ql_first(head)->id, entries[0].id, "Element id mismatch"); + assert_c_eq(ql_last(head, link)->id, entries[nentries-1].id, + "Element id mismatch"); + + i = 0; + ql_foreach(t, head, link) { + assert_c_eq(t->id, entries[i].id, "Element id mismatch"); + i++; + } + + i = 0; + ql_reverse_foreach(t, head, link) { + assert_c_eq(t->id, entries[nentries-i-1].id, + "Element id mismatch"); + i++; + } + + for (i = 0; i < nentries-1; i++) { + t = ql_next(head, &entries[i], link); + assert_c_eq(t->id, entries[i+1].id, "Element id mismatch"); + } + assert_ptr_null(ql_next(head, &entries[nentries-1], link), + "Unexpected element"); + + assert_ptr_null(ql_prev(head, &entries[0], link), "Unexpected element"); + for (i = 1; i < nentries; i++) { + t = ql_prev(head, &entries[i], link); + assert_c_eq(t->id, entries[i-1].id, "Element id mismatch"); + } +} + +TEST_BEGIN(test_ql_tail_insert) { + list_head_t head; + list_t entries[NENTRIES]; + unsigned i; + + ql_new(&head); + init_entries(entries, sizeof(entries)/sizeof(list_t)); + for (i = 0; i < NENTRIES; i++) { + ql_tail_insert(&head, &entries[i], link); + } + + test_entries_list(&head, entries, NENTRIES); +} +TEST_END + +TEST_BEGIN(test_ql_tail_remove) { + list_head_t head; + list_t entries[NENTRIES]; + unsigned i; + + ql_new(&head); + init_entries(entries, sizeof(entries)/sizeof(list_t)); + for (i = 0; i < NENTRIES; i++) { + ql_tail_insert(&head, &entries[i], link); + } + + for (i = 0; i < NENTRIES; i++) { + test_entries_list(&head, entries, NENTRIES-i); + ql_tail_remove(&head, list_t, link); + } + test_empty_list(&head); +} +TEST_END + +TEST_BEGIN(test_ql_head_insert) { + list_head_t head; + list_t entries[NENTRIES]; + unsigned i; + + ql_new(&head); + init_entries(entries, sizeof(entries)/sizeof(list_t)); + for (i = 0; i < NENTRIES; i++) { + ql_head_insert(&head, &entries[NENTRIES-i-1], link); + } + + test_entries_list(&head, entries, NENTRIES); +} +TEST_END + +TEST_BEGIN(test_ql_head_remove) { + list_head_t head; + list_t entries[NENTRIES]; + unsigned i; + + ql_new(&head); + init_entries(entries, sizeof(entries)/sizeof(list_t)); + for (i = 0; i < NENTRIES; i++) { + ql_head_insert(&head, &entries[NENTRIES-i-1], link); + } + + for (i = 0; i < NENTRIES; i++) { + test_entries_list(&head, &entries[i], NENTRIES-i); + ql_head_remove(&head, list_t, link); + } + test_empty_list(&head); +} +TEST_END + +TEST_BEGIN(test_ql_insert) { + list_head_t head; + list_t entries[8]; + list_t *a, *b, *c, *d, *e, *f, *g, *h; + + ql_new(&head); + init_entries(entries, sizeof(entries)/sizeof(list_t)); + a = &entries[0]; + b = &entries[1]; + c = &entries[2]; + d = &entries[3]; + e = &entries[4]; + f = &entries[5]; + g = &entries[6]; + h = &entries[7]; + + /* + * ql_remove(), ql_before_insert(), and ql_after_insert() are used + * internally by other macros that are already tested, so there's no + * need to test them completely. However, insertion/deletion from the + * middle of lists is not otherwise tested; do so here. + */ + ql_tail_insert(&head, f, link); + ql_before_insert(&head, f, b, link); + ql_before_insert(&head, f, c, link); + ql_after_insert(f, h, link); + ql_after_insert(f, g, link); + ql_before_insert(&head, b, a, link); + ql_after_insert(c, d, link); + ql_before_insert(&head, f, e, link); + + test_entries_list(&head, entries, sizeof(entries)/sizeof(list_t)); +} +TEST_END + +int +main(void) { + return test( + test_ql_empty, + test_ql_tail_insert, + test_ql_tail_remove, + test_ql_head_insert, + test_ql_head_remove, + test_ql_insert); +} diff --git a/deps/jemalloc/test/unit/qr.c b/deps/jemalloc/test/unit/qr.c new file mode 100644 index 000000000..271a10953 --- /dev/null +++ b/deps/jemalloc/test/unit/qr.c @@ -0,0 +1,243 @@ +#include "test/jemalloc_test.h" + +#include "jemalloc/internal/qr.h" + +/* Number of ring entries, in [2..26]. */ +#define NENTRIES 9 +/* Split index, in [1..NENTRIES). */ +#define SPLIT_INDEX 5 + +typedef struct ring_s ring_t; + +struct ring_s { + qr(ring_t) link; + char id; +}; + +static void +init_entries(ring_t *entries) { + unsigned i; + + for (i = 0; i < NENTRIES; i++) { + qr_new(&entries[i], link); + entries[i].id = 'a' + i; + } +} + +static void +test_independent_entries(ring_t *entries) { + ring_t *t; + unsigned i, j; + + for (i = 0; i < NENTRIES; i++) { + j = 0; + qr_foreach(t, &entries[i], link) { + j++; + } + assert_u_eq(j, 1, + "Iteration over single-element ring should visit precisely " + "one element"); + } + for (i = 0; i < NENTRIES; i++) { + j = 0; + qr_reverse_foreach(t, &entries[i], link) { + j++; + } + assert_u_eq(j, 1, + "Iteration over single-element ring should visit precisely " + "one element"); + } + for (i = 0; i < NENTRIES; i++) { + t = qr_next(&entries[i], link); + assert_ptr_eq(t, &entries[i], + "Next element in single-element ring should be same as " + "current element"); + } + for (i = 0; i < NENTRIES; i++) { + t = qr_prev(&entries[i], link); + assert_ptr_eq(t, &entries[i], + "Previous element in single-element ring should be same as " + "current element"); + } +} + +TEST_BEGIN(test_qr_one) { + ring_t entries[NENTRIES]; + + init_entries(entries); + test_independent_entries(entries); +} +TEST_END + +static void +test_entries_ring(ring_t *entries) { + ring_t *t; + unsigned i, j; + + for (i = 0; i < NENTRIES; i++) { + j = 0; + qr_foreach(t, &entries[i], link) { + assert_c_eq(t->id, entries[(i+j) % NENTRIES].id, + "Element id mismatch"); + j++; + } + } + for (i = 0; i < NENTRIES; i++) { + j = 0; + qr_reverse_foreach(t, &entries[i], link) { + assert_c_eq(t->id, entries[(NENTRIES+i-j-1) % + NENTRIES].id, "Element id mismatch"); + j++; + } + } + for (i = 0; i < NENTRIES; i++) { + t = qr_next(&entries[i], link); + assert_c_eq(t->id, entries[(i+1) % NENTRIES].id, + "Element id mismatch"); + } + for (i = 0; i < NENTRIES; i++) { + t = qr_prev(&entries[i], link); + assert_c_eq(t->id, entries[(NENTRIES+i-1) % NENTRIES].id, + "Element id mismatch"); + } +} + +TEST_BEGIN(test_qr_after_insert) { + ring_t entries[NENTRIES]; + unsigned i; + + init_entries(entries); + for (i = 1; i < NENTRIES; i++) { + qr_after_insert(&entries[i - 1], &entries[i], link); + } + test_entries_ring(entries); +} +TEST_END + +TEST_BEGIN(test_qr_remove) { + ring_t entries[NENTRIES]; + ring_t *t; + unsigned i, j; + + init_entries(entries); + for (i = 1; i < NENTRIES; i++) { + qr_after_insert(&entries[i - 1], &entries[i], link); + } + + for (i = 0; i < NENTRIES; i++) { + j = 0; + qr_foreach(t, &entries[i], link) { + assert_c_eq(t->id, entries[i+j].id, + "Element id mismatch"); + j++; + } + j = 0; + qr_reverse_foreach(t, &entries[i], link) { + assert_c_eq(t->id, entries[NENTRIES - 1 - j].id, + "Element id mismatch"); + j++; + } + qr_remove(&entries[i], link); + } + test_independent_entries(entries); +} +TEST_END + +TEST_BEGIN(test_qr_before_insert) { + ring_t entries[NENTRIES]; + ring_t *t; + unsigned i, j; + + init_entries(entries); + for (i = 1; i < NENTRIES; i++) { + qr_before_insert(&entries[i - 1], &entries[i], link); + } + for (i = 0; i < NENTRIES; i++) { + j = 0; + qr_foreach(t, &entries[i], link) { + assert_c_eq(t->id, entries[(NENTRIES+i-j) % + NENTRIES].id, "Element id mismatch"); + j++; + } + } + for (i = 0; i < NENTRIES; i++) { + j = 0; + qr_reverse_foreach(t, &entries[i], link) { + assert_c_eq(t->id, entries[(i+j+1) % NENTRIES].id, + "Element id mismatch"); + j++; + } + } + for (i = 0; i < NENTRIES; i++) { + t = qr_next(&entries[i], link); + assert_c_eq(t->id, entries[(NENTRIES+i-1) % NENTRIES].id, + "Element id mismatch"); + } + for (i = 0; i < NENTRIES; i++) { + t = qr_prev(&entries[i], link); + assert_c_eq(t->id, entries[(i+1) % NENTRIES].id, + "Element id mismatch"); + } +} +TEST_END + +static void +test_split_entries(ring_t *entries) { + ring_t *t; + unsigned i, j; + + for (i = 0; i < NENTRIES; i++) { + j = 0; + qr_foreach(t, &entries[i], link) { + if (i < SPLIT_INDEX) { + assert_c_eq(t->id, + entries[(i+j) % SPLIT_INDEX].id, + "Element id mismatch"); + } else { + assert_c_eq(t->id, entries[(i+j-SPLIT_INDEX) % + (NENTRIES-SPLIT_INDEX) + SPLIT_INDEX].id, + "Element id mismatch"); + } + j++; + } + } +} + +TEST_BEGIN(test_qr_meld_split) { + ring_t entries[NENTRIES]; + unsigned i; + + init_entries(entries); + for (i = 1; i < NENTRIES; i++) { + qr_after_insert(&entries[i - 1], &entries[i], link); + } + + qr_split(&entries[0], &entries[SPLIT_INDEX], ring_t, link); + test_split_entries(entries); + + qr_meld(&entries[0], &entries[SPLIT_INDEX], ring_t, link); + test_entries_ring(entries); + + qr_meld(&entries[0], &entries[SPLIT_INDEX], ring_t, link); + test_split_entries(entries); + + qr_split(&entries[0], &entries[SPLIT_INDEX], ring_t, link); + test_entries_ring(entries); + + qr_split(&entries[0], &entries[0], ring_t, link); + test_entries_ring(entries); + + qr_meld(&entries[0], &entries[0], ring_t, link); + test_entries_ring(entries); +} +TEST_END + +int +main(void) { + return test( + test_qr_one, + test_qr_after_insert, + test_qr_remove, + test_qr_before_insert, + test_qr_meld_split); +} diff --git a/deps/jemalloc/test/unit/rb.c b/deps/jemalloc/test/unit/rb.c new file mode 100644 index 000000000..65c049207 --- /dev/null +++ b/deps/jemalloc/test/unit/rb.c @@ -0,0 +1,355 @@ +#include "test/jemalloc_test.h" + +#include "jemalloc/internal/rb.h" + +#define rbtn_black_height(a_type, a_field, a_rbt, r_height) do { \ + a_type *rbp_bh_t; \ + for (rbp_bh_t = (a_rbt)->rbt_root, (r_height) = 0; rbp_bh_t != \ + NULL; rbp_bh_t = rbtn_left_get(a_type, a_field, \ + rbp_bh_t)) { \ + if (!rbtn_red_get(a_type, a_field, rbp_bh_t)) { \ + (r_height)++; \ + } \ + } \ +} while (0) + +typedef struct node_s node_t; + +struct node_s { +#define NODE_MAGIC 0x9823af7e + uint32_t magic; + rb_node(node_t) link; + uint64_t key; +}; + +static int +node_cmp(const node_t *a, const node_t *b) { + int ret; + + assert_u32_eq(a->magic, NODE_MAGIC, "Bad magic"); + assert_u32_eq(b->magic, NODE_MAGIC, "Bad magic"); + + ret = (a->key > b->key) - (a->key < b->key); + if (ret == 0) { + /* + * Duplicates are not allowed in the tree, so force an + * arbitrary ordering for non-identical items with equal keys. + */ + ret = (((uintptr_t)a) > ((uintptr_t)b)) + - (((uintptr_t)a) < ((uintptr_t)b)); + } + return ret; +} + +typedef rb_tree(node_t) tree_t; +rb_gen(static, tree_, tree_t, node_t, link, node_cmp); + +TEST_BEGIN(test_rb_empty) { + tree_t tree; + node_t key; + + tree_new(&tree); + + assert_true(tree_empty(&tree), "Tree should be empty"); + assert_ptr_null(tree_first(&tree), "Unexpected node"); + assert_ptr_null(tree_last(&tree), "Unexpected node"); + + key.key = 0; + key.magic = NODE_MAGIC; + assert_ptr_null(tree_search(&tree, &key), "Unexpected node"); + + key.key = 0; + key.magic = NODE_MAGIC; + assert_ptr_null(tree_nsearch(&tree, &key), "Unexpected node"); + + key.key = 0; + key.magic = NODE_MAGIC; + assert_ptr_null(tree_psearch(&tree, &key), "Unexpected node"); +} +TEST_END + +static unsigned +tree_recurse(node_t *node, unsigned black_height, unsigned black_depth) { + unsigned ret = 0; + node_t *left_node; + node_t *right_node; + + if (node == NULL) { + return ret; + } + + left_node = rbtn_left_get(node_t, link, node); + right_node = rbtn_right_get(node_t, link, node); + + if (!rbtn_red_get(node_t, link, node)) { + black_depth++; + } + + /* Red nodes must be interleaved with black nodes. */ + if (rbtn_red_get(node_t, link, node)) { + if (left_node != NULL) { + assert_false(rbtn_red_get(node_t, link, left_node), + "Node should be black"); + } + if (right_node != NULL) { + assert_false(rbtn_red_get(node_t, link, right_node), + "Node should be black"); + } + } + + /* Self. */ + assert_u32_eq(node->magic, NODE_MAGIC, "Bad magic"); + + /* Left subtree. */ + if (left_node != NULL) { + ret += tree_recurse(left_node, black_height, black_depth); + } else { + ret += (black_depth != black_height); + } + + /* Right subtree. */ + if (right_node != NULL) { + ret += tree_recurse(right_node, black_height, black_depth); + } else { + ret += (black_depth != black_height); + } + + return ret; +} + +static node_t * +tree_iterate_cb(tree_t *tree, node_t *node, void *data) { + unsigned *i = (unsigned *)data; + node_t *search_node; + + assert_u32_eq(node->magic, NODE_MAGIC, "Bad magic"); + + /* Test rb_search(). */ + search_node = tree_search(tree, node); + assert_ptr_eq(search_node, node, + "tree_search() returned unexpected node"); + + /* Test rb_nsearch(). */ + search_node = tree_nsearch(tree, node); + assert_ptr_eq(search_node, node, + "tree_nsearch() returned unexpected node"); + + /* Test rb_psearch(). */ + search_node = tree_psearch(tree, node); + assert_ptr_eq(search_node, node, + "tree_psearch() returned unexpected node"); + + (*i)++; + + return NULL; +} + +static unsigned +tree_iterate(tree_t *tree) { + unsigned i; + + i = 0; + tree_iter(tree, NULL, tree_iterate_cb, (void *)&i); + + return i; +} + +static unsigned +tree_iterate_reverse(tree_t *tree) { + unsigned i; + + i = 0; + tree_reverse_iter(tree, NULL, tree_iterate_cb, (void *)&i); + + return i; +} + +static void +node_remove(tree_t *tree, node_t *node, unsigned nnodes) { + node_t *search_node; + unsigned black_height, imbalances; + + tree_remove(tree, node); + + /* Test rb_nsearch(). */ + search_node = tree_nsearch(tree, node); + if (search_node != NULL) { + assert_u64_ge(search_node->key, node->key, + "Key ordering error"); + } + + /* Test rb_psearch(). */ + search_node = tree_psearch(tree, node); + if (search_node != NULL) { + assert_u64_le(search_node->key, node->key, + "Key ordering error"); + } + + node->magic = 0; + + rbtn_black_height(node_t, link, tree, black_height); + imbalances = tree_recurse(tree->rbt_root, black_height, 0); + assert_u_eq(imbalances, 0, "Tree is unbalanced"); + assert_u_eq(tree_iterate(tree), nnodes-1, + "Unexpected node iteration count"); + assert_u_eq(tree_iterate_reverse(tree), nnodes-1, + "Unexpected node iteration count"); +} + +static node_t * +remove_iterate_cb(tree_t *tree, node_t *node, void *data) { + unsigned *nnodes = (unsigned *)data; + node_t *ret = tree_next(tree, node); + + node_remove(tree, node, *nnodes); + + return ret; +} + +static node_t * +remove_reverse_iterate_cb(tree_t *tree, node_t *node, void *data) { + unsigned *nnodes = (unsigned *)data; + node_t *ret = tree_prev(tree, node); + + node_remove(tree, node, *nnodes); + + return ret; +} + +static void +destroy_cb(node_t *node, void *data) { + unsigned *nnodes = (unsigned *)data; + + assert_u_gt(*nnodes, 0, "Destruction removed too many nodes"); + (*nnodes)--; +} + +TEST_BEGIN(test_rb_random) { +#define NNODES 25 +#define NBAGS 250 +#define SEED 42 + sfmt_t *sfmt; + uint64_t bag[NNODES]; + tree_t tree; + node_t nodes[NNODES]; + unsigned i, j, k, black_height, imbalances; + + sfmt = init_gen_rand(SEED); + for (i = 0; i < NBAGS; i++) { + switch (i) { + case 0: + /* Insert in order. */ + for (j = 0; j < NNODES; j++) { + bag[j] = j; + } + break; + case 1: + /* Insert in reverse order. */ + for (j = 0; j < NNODES; j++) { + bag[j] = NNODES - j - 1; + } + break; + default: + for (j = 0; j < NNODES; j++) { + bag[j] = gen_rand64_range(sfmt, NNODES); + } + } + + for (j = 1; j <= NNODES; j++) { + /* Initialize tree and nodes. */ + tree_new(&tree); + for (k = 0; k < j; k++) { + nodes[k].magic = NODE_MAGIC; + nodes[k].key = bag[k]; + } + + /* Insert nodes. */ + for (k = 0; k < j; k++) { + tree_insert(&tree, &nodes[k]); + + rbtn_black_height(node_t, link, &tree, + black_height); + imbalances = tree_recurse(tree.rbt_root, + black_height, 0); + assert_u_eq(imbalances, 0, + "Tree is unbalanced"); + + assert_u_eq(tree_iterate(&tree), k+1, + "Unexpected node iteration count"); + assert_u_eq(tree_iterate_reverse(&tree), k+1, + "Unexpected node iteration count"); + + assert_false(tree_empty(&tree), + "Tree should not be empty"); + assert_ptr_not_null(tree_first(&tree), + "Tree should not be empty"); + assert_ptr_not_null(tree_last(&tree), + "Tree should not be empty"); + + tree_next(&tree, &nodes[k]); + tree_prev(&tree, &nodes[k]); + } + + /* Remove nodes. */ + switch (i % 5) { + case 0: + for (k = 0; k < j; k++) { + node_remove(&tree, &nodes[k], j - k); + } + break; + case 1: + for (k = j; k > 0; k--) { + node_remove(&tree, &nodes[k-1], k); + } + break; + case 2: { + node_t *start; + unsigned nnodes = j; + + start = NULL; + do { + start = tree_iter(&tree, start, + remove_iterate_cb, (void *)&nnodes); + nnodes--; + } while (start != NULL); + assert_u_eq(nnodes, 0, + "Removal terminated early"); + break; + } case 3: { + node_t *start; + unsigned nnodes = j; + + start = NULL; + do { + start = tree_reverse_iter(&tree, start, + remove_reverse_iterate_cb, + (void *)&nnodes); + nnodes--; + } while (start != NULL); + assert_u_eq(nnodes, 0, + "Removal terminated early"); + break; + } case 4: { + unsigned nnodes = j; + tree_destroy(&tree, destroy_cb, &nnodes); + assert_u_eq(nnodes, 0, + "Destruction terminated early"); + break; + } default: + not_reached(); + } + } + } + fini_gen_rand(sfmt); +#undef NNODES +#undef NBAGS +#undef SEED +} +TEST_END + +int +main(void) { + return test( + test_rb_empty, + test_rb_random); +} diff --git a/deps/jemalloc/test/unit/retained.c b/deps/jemalloc/test/unit/retained.c new file mode 100644 index 000000000..7993fd3d9 --- /dev/null +++ b/deps/jemalloc/test/unit/retained.c @@ -0,0 +1,184 @@ +#include "test/jemalloc_test.h" + +#include "jemalloc/internal/spin.h" + +static unsigned arena_ind; +static size_t sz; +static size_t esz; +#define NEPOCHS 8 +#define PER_THD_NALLOCS 1 +static atomic_u_t epoch; +static atomic_u_t nfinished; + +static unsigned +do_arena_create(extent_hooks_t *h) { + unsigned arena_ind; + size_t sz = sizeof(unsigned); + assert_d_eq(mallctl("arenas.create", (void *)&arena_ind, &sz, + (void *)(h != NULL ? &h : NULL), (h != NULL ? sizeof(h) : 0)), 0, + "Unexpected mallctl() failure"); + return arena_ind; +} + +static void +do_arena_destroy(unsigned arena_ind) { + size_t mib[3]; + size_t miblen; + + miblen = sizeof(mib)/sizeof(size_t); + assert_d_eq(mallctlnametomib("arena.0.destroy", mib, &miblen), 0, + "Unexpected mallctlnametomib() failure"); + mib[1] = (size_t)arena_ind; + assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, NULL, 0), 0, + "Unexpected mallctlbymib() failure"); +} + +static void +do_refresh(void) { + uint64_t epoch = 1; + assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch, + sizeof(epoch)), 0, "Unexpected mallctl() failure"); +} + +static size_t +do_get_size_impl(const char *cmd, unsigned arena_ind) { + size_t mib[4]; + size_t miblen = sizeof(mib) / sizeof(size_t); + size_t z = sizeof(size_t); + + assert_d_eq(mallctlnametomib(cmd, mib, &miblen), + 0, "Unexpected mallctlnametomib(\"%s\", ...) failure", cmd); + mib[2] = arena_ind; + size_t size; + assert_d_eq(mallctlbymib(mib, miblen, (void *)&size, &z, NULL, 0), + 0, "Unexpected mallctlbymib([\"%s\"], ...) failure", cmd); + + return size; +} + +static size_t +do_get_active(unsigned arena_ind) { + return do_get_size_impl("stats.arenas.0.pactive", arena_ind) * PAGE; +} + +static size_t +do_get_mapped(unsigned arena_ind) { + return do_get_size_impl("stats.arenas.0.mapped", arena_ind); +} + +static void * +thd_start(void *arg) { + for (unsigned next_epoch = 1; next_epoch < NEPOCHS; next_epoch++) { + /* Busy-wait for next epoch. */ + unsigned cur_epoch; + spin_t spinner = SPIN_INITIALIZER; + while ((cur_epoch = atomic_load_u(&epoch, ATOMIC_ACQUIRE)) != + next_epoch) { + spin_adaptive(&spinner); + } + assert_u_eq(cur_epoch, next_epoch, "Unexpected epoch"); + + /* + * Allocate. The main thread will reset the arena, so there's + * no need to deallocate. + */ + for (unsigned i = 0; i < PER_THD_NALLOCS; i++) { + void *p = mallocx(sz, MALLOCX_ARENA(arena_ind) | + MALLOCX_TCACHE_NONE + ); + assert_ptr_not_null(p, + "Unexpected mallocx() failure\n"); + } + + /* Let the main thread know we've finished this iteration. */ + atomic_fetch_add_u(&nfinished, 1, ATOMIC_RELEASE); + } + + return NULL; +} + +TEST_BEGIN(test_retained) { + test_skip_if(!config_stats); + + arena_ind = do_arena_create(NULL); + sz = nallocx(HUGEPAGE, 0); + esz = sz + sz_large_pad; + + atomic_store_u(&epoch, 0, ATOMIC_RELAXED); + + unsigned nthreads = ncpus * 2; + if (LG_SIZEOF_PTR < 3 && nthreads > 16) { + nthreads = 16; /* 32-bit platform could run out of vaddr. */ + } + VARIABLE_ARRAY(thd_t, threads, nthreads); + for (unsigned i = 0; i < nthreads; i++) { + thd_create(&threads[i], thd_start, NULL); + } + + for (unsigned e = 1; e < NEPOCHS; e++) { + atomic_store_u(&nfinished, 0, ATOMIC_RELEASE); + atomic_store_u(&epoch, e, ATOMIC_RELEASE); + + /* Wait for threads to finish allocating. */ + spin_t spinner = SPIN_INITIALIZER; + while (atomic_load_u(&nfinished, ATOMIC_ACQUIRE) < nthreads) { + spin_adaptive(&spinner); + } + + /* + * Assert that retained is no more than the sum of size classes + * that should have been used to satisfy the worker threads' + * requests, discounting per growth fragmentation. + */ + do_refresh(); + + size_t allocated = esz * nthreads * PER_THD_NALLOCS; + size_t active = do_get_active(arena_ind); + assert_zu_le(allocated, active, "Unexpected active memory"); + size_t mapped = do_get_mapped(arena_ind); + assert_zu_le(active, mapped, "Unexpected mapped memory"); + + arena_t *arena = arena_get(tsdn_fetch(), arena_ind, false); + size_t usable = 0; + size_t fragmented = 0; + for (pszind_t pind = sz_psz2ind(HUGEPAGE); pind < + arena->extent_grow_next; pind++) { + size_t psz = sz_pind2sz(pind); + size_t psz_fragmented = psz % esz; + size_t psz_usable = psz - psz_fragmented; + /* + * Only consider size classes that wouldn't be skipped. + */ + if (psz_usable > 0) { + assert_zu_lt(usable, allocated, + "Excessive retained memory " + "(%#zx[+%#zx] > %#zx)", usable, psz_usable, + allocated); + fragmented += psz_fragmented; + usable += psz_usable; + } + } + + /* + * Clean up arena. Destroying and recreating the arena + * is simpler that specifying extent hooks that deallocate + * (rather than retaining) during reset. + */ + do_arena_destroy(arena_ind); + assert_u_eq(do_arena_create(NULL), arena_ind, + "Unexpected arena index"); + } + + for (unsigned i = 0; i < nthreads; i++) { + thd_join(threads[i], NULL); + } + + do_arena_destroy(arena_ind); +} +TEST_END + +int +main(void) { + return test( + test_retained); +} diff --git a/deps/jemalloc/test/unit/rtree.c b/deps/jemalloc/test/unit/rtree.c new file mode 100644 index 000000000..90adca134 --- /dev/null +++ b/deps/jemalloc/test/unit/rtree.c @@ -0,0 +1,228 @@ +#include "test/jemalloc_test.h" + +#include "jemalloc/internal/rtree.h" + +rtree_node_alloc_t *rtree_node_alloc_orig; +rtree_node_dalloc_t *rtree_node_dalloc_orig; +rtree_leaf_alloc_t *rtree_leaf_alloc_orig; +rtree_leaf_dalloc_t *rtree_leaf_dalloc_orig; + +/* Potentially too large to safely place on the stack. */ +rtree_t test_rtree; + +static rtree_node_elm_t * +rtree_node_alloc_intercept(tsdn_t *tsdn, rtree_t *rtree, size_t nelms) { + rtree_node_elm_t *node; + + if (rtree != &test_rtree) { + return rtree_node_alloc_orig(tsdn, rtree, nelms); + } + + malloc_mutex_unlock(tsdn, &rtree->init_lock); + node = (rtree_node_elm_t *)calloc(nelms, sizeof(rtree_node_elm_t)); + assert_ptr_not_null(node, "Unexpected calloc() failure"); + malloc_mutex_lock(tsdn, &rtree->init_lock); + + return node; +} + +static void +rtree_node_dalloc_intercept(tsdn_t *tsdn, rtree_t *rtree, + rtree_node_elm_t *node) { + if (rtree != &test_rtree) { + rtree_node_dalloc_orig(tsdn, rtree, node); + return; + } + + free(node); +} + +static rtree_leaf_elm_t * +rtree_leaf_alloc_intercept(tsdn_t *tsdn, rtree_t *rtree, size_t nelms) { + rtree_leaf_elm_t *leaf; + + if (rtree != &test_rtree) { + return rtree_leaf_alloc_orig(tsdn, rtree, nelms); + } + + malloc_mutex_unlock(tsdn, &rtree->init_lock); + leaf = (rtree_leaf_elm_t *)calloc(nelms, sizeof(rtree_leaf_elm_t)); + assert_ptr_not_null(leaf, "Unexpected calloc() failure"); + malloc_mutex_lock(tsdn, &rtree->init_lock); + + return leaf; +} + +static void +rtree_leaf_dalloc_intercept(tsdn_t *tsdn, rtree_t *rtree, + rtree_leaf_elm_t *leaf) { + if (rtree != &test_rtree) { + rtree_leaf_dalloc_orig(tsdn, rtree, leaf); + return; + } + + free(leaf); +} + +TEST_BEGIN(test_rtree_read_empty) { + tsdn_t *tsdn; + + tsdn = tsdn_fetch(); + + rtree_t *rtree = &test_rtree; + rtree_ctx_t rtree_ctx; + rtree_ctx_data_init(&rtree_ctx); + assert_false(rtree_new(rtree, false), "Unexpected rtree_new() failure"); + assert_ptr_null(rtree_extent_read(tsdn, rtree, &rtree_ctx, PAGE, + false), "rtree_extent_read() should return NULL for empty tree"); + rtree_delete(tsdn, rtree); +} +TEST_END + +#undef NTHREADS +#undef NITERS +#undef SEED + +TEST_BEGIN(test_rtree_extrema) { + extent_t extent_a, extent_b; + extent_init(&extent_a, NULL, NULL, SC_LARGE_MINCLASS, false, + sz_size2index(SC_LARGE_MINCLASS), 0, + extent_state_active, false, false, true, EXTENT_NOT_HEAD); + extent_init(&extent_b, NULL, NULL, 0, false, SC_NSIZES, 0, + extent_state_active, false, false, true, EXTENT_NOT_HEAD); + + tsdn_t *tsdn = tsdn_fetch(); + + rtree_t *rtree = &test_rtree; + rtree_ctx_t rtree_ctx; + rtree_ctx_data_init(&rtree_ctx); + assert_false(rtree_new(rtree, false), "Unexpected rtree_new() failure"); + + assert_false(rtree_write(tsdn, rtree, &rtree_ctx, PAGE, &extent_a, + extent_szind_get(&extent_a), extent_slab_get(&extent_a)), + "Unexpected rtree_write() failure"); + rtree_szind_slab_update(tsdn, rtree, &rtree_ctx, PAGE, + extent_szind_get(&extent_a), extent_slab_get(&extent_a)); + assert_ptr_eq(rtree_extent_read(tsdn, rtree, &rtree_ctx, PAGE, true), + &extent_a, + "rtree_extent_read() should return previously set value"); + + assert_false(rtree_write(tsdn, rtree, &rtree_ctx, ~((uintptr_t)0), + &extent_b, extent_szind_get_maybe_invalid(&extent_b), + extent_slab_get(&extent_b)), "Unexpected rtree_write() failure"); + assert_ptr_eq(rtree_extent_read(tsdn, rtree, &rtree_ctx, + ~((uintptr_t)0), true), &extent_b, + "rtree_extent_read() should return previously set value"); + + rtree_delete(tsdn, rtree); +} +TEST_END + +TEST_BEGIN(test_rtree_bits) { + tsdn_t *tsdn = tsdn_fetch(); + + uintptr_t keys[] = {PAGE, PAGE + 1, + PAGE + (((uintptr_t)1) << LG_PAGE) - 1}; + + extent_t extent; + extent_init(&extent, NULL, NULL, 0, false, SC_NSIZES, 0, + extent_state_active, false, false, true, EXTENT_NOT_HEAD); + + rtree_t *rtree = &test_rtree; + rtree_ctx_t rtree_ctx; + rtree_ctx_data_init(&rtree_ctx); + assert_false(rtree_new(rtree, false), "Unexpected rtree_new() failure"); + + for (unsigned i = 0; i < sizeof(keys)/sizeof(uintptr_t); i++) { + assert_false(rtree_write(tsdn, rtree, &rtree_ctx, keys[i], + &extent, SC_NSIZES, false), + "Unexpected rtree_write() failure"); + for (unsigned j = 0; j < sizeof(keys)/sizeof(uintptr_t); j++) { + assert_ptr_eq(rtree_extent_read(tsdn, rtree, &rtree_ctx, + keys[j], true), &extent, + "rtree_extent_read() should return previously set " + "value and ignore insignificant key bits; i=%u, " + "j=%u, set key=%#"FMTxPTR", get key=%#"FMTxPTR, i, + j, keys[i], keys[j]); + } + assert_ptr_null(rtree_extent_read(tsdn, rtree, &rtree_ctx, + (((uintptr_t)2) << LG_PAGE), false), + "Only leftmost rtree leaf should be set; i=%u", i); + rtree_clear(tsdn, rtree, &rtree_ctx, keys[i]); + } + + rtree_delete(tsdn, rtree); +} +TEST_END + +TEST_BEGIN(test_rtree_random) { +#define NSET 16 +#define SEED 42 + sfmt_t *sfmt = init_gen_rand(SEED); + tsdn_t *tsdn = tsdn_fetch(); + uintptr_t keys[NSET]; + rtree_t *rtree = &test_rtree; + rtree_ctx_t rtree_ctx; + rtree_ctx_data_init(&rtree_ctx); + + extent_t extent; + extent_init(&extent, NULL, NULL, 0, false, SC_NSIZES, 0, + extent_state_active, false, false, true, EXTENT_NOT_HEAD); + + assert_false(rtree_new(rtree, false), "Unexpected rtree_new() failure"); + + for (unsigned i = 0; i < NSET; i++) { + keys[i] = (uintptr_t)gen_rand64(sfmt); + rtree_leaf_elm_t *elm = rtree_leaf_elm_lookup(tsdn, rtree, + &rtree_ctx, keys[i], false, true); + assert_ptr_not_null(elm, + "Unexpected rtree_leaf_elm_lookup() failure"); + rtree_leaf_elm_write(tsdn, rtree, elm, &extent, SC_NSIZES, + false); + assert_ptr_eq(rtree_extent_read(tsdn, rtree, &rtree_ctx, + keys[i], true), &extent, + "rtree_extent_read() should return previously set value"); + } + for (unsigned i = 0; i < NSET; i++) { + assert_ptr_eq(rtree_extent_read(tsdn, rtree, &rtree_ctx, + keys[i], true), &extent, + "rtree_extent_read() should return previously set value, " + "i=%u", i); + } + + for (unsigned i = 0; i < NSET; i++) { + rtree_clear(tsdn, rtree, &rtree_ctx, keys[i]); + assert_ptr_null(rtree_extent_read(tsdn, rtree, &rtree_ctx, + keys[i], true), + "rtree_extent_read() should return previously set value"); + } + for (unsigned i = 0; i < NSET; i++) { + assert_ptr_null(rtree_extent_read(tsdn, rtree, &rtree_ctx, + keys[i], true), + "rtree_extent_read() should return previously set value"); + } + + rtree_delete(tsdn, rtree); + fini_gen_rand(sfmt); +#undef NSET +#undef SEED +} +TEST_END + +int +main(void) { + rtree_node_alloc_orig = rtree_node_alloc; + rtree_node_alloc = rtree_node_alloc_intercept; + rtree_node_dalloc_orig = rtree_node_dalloc; + rtree_node_dalloc = rtree_node_dalloc_intercept; + rtree_leaf_alloc_orig = rtree_leaf_alloc; + rtree_leaf_alloc = rtree_leaf_alloc_intercept; + rtree_leaf_dalloc_orig = rtree_leaf_dalloc; + rtree_leaf_dalloc = rtree_leaf_dalloc_intercept; + + return test( + test_rtree_read_empty, + test_rtree_extrema, + test_rtree_bits, + test_rtree_random); +} diff --git a/deps/jemalloc/test/unit/safety_check.c b/deps/jemalloc/test/unit/safety_check.c new file mode 100644 index 000000000..bf4bd86d6 --- /dev/null +++ b/deps/jemalloc/test/unit/safety_check.c @@ -0,0 +1,156 @@ +#include "test/jemalloc_test.h" + +#include "jemalloc/internal/safety_check.h" + +/* + * Note that we get called through safety_check.sh, which turns on sampling for + * everything. + */ + +bool fake_abort_called; +void fake_abort(const char *message) { + (void)message; + fake_abort_called = true; +} + +TEST_BEGIN(test_malloc_free_overflow) { + test_skip_if(!config_prof); + test_skip_if(!config_opt_safety_checks); + + safety_check_set_abort(&fake_abort); + /* Buffer overflow! */ + char* ptr = malloc(128); + ptr[128] = 0; + free(ptr); + safety_check_set_abort(NULL); + + assert_b_eq(fake_abort_called, true, "Redzone check didn't fire."); + fake_abort_called = false; +} +TEST_END + +TEST_BEGIN(test_mallocx_dallocx_overflow) { + test_skip_if(!config_prof); + test_skip_if(!config_opt_safety_checks); + + safety_check_set_abort(&fake_abort); + /* Buffer overflow! */ + char* ptr = mallocx(128, 0); + ptr[128] = 0; + dallocx(ptr, 0); + safety_check_set_abort(NULL); + + assert_b_eq(fake_abort_called, true, "Redzone check didn't fire."); + fake_abort_called = false; +} +TEST_END + +TEST_BEGIN(test_malloc_sdallocx_overflow) { + test_skip_if(!config_prof); + test_skip_if(!config_opt_safety_checks); + + safety_check_set_abort(&fake_abort); + /* Buffer overflow! */ + char* ptr = malloc(128); + ptr[128] = 0; + sdallocx(ptr, 128, 0); + safety_check_set_abort(NULL); + + assert_b_eq(fake_abort_called, true, "Redzone check didn't fire."); + fake_abort_called = false; +} +TEST_END + +TEST_BEGIN(test_realloc_overflow) { + test_skip_if(!config_prof); + test_skip_if(!config_opt_safety_checks); + + safety_check_set_abort(&fake_abort); + /* Buffer overflow! */ + char* ptr = malloc(128); + ptr[128] = 0; + ptr = realloc(ptr, 129); + safety_check_set_abort(NULL); + free(ptr); + + assert_b_eq(fake_abort_called, true, "Redzone check didn't fire."); + fake_abort_called = false; +} +TEST_END + +TEST_BEGIN(test_rallocx_overflow) { + test_skip_if(!config_prof); + test_skip_if(!config_opt_safety_checks); + + safety_check_set_abort(&fake_abort); + /* Buffer overflow! */ + char* ptr = malloc(128); + ptr[128] = 0; + ptr = rallocx(ptr, 129, 0); + safety_check_set_abort(NULL); + free(ptr); + + assert_b_eq(fake_abort_called, true, "Redzone check didn't fire."); + fake_abort_called = false; +} +TEST_END + +TEST_BEGIN(test_xallocx_overflow) { + test_skip_if(!config_prof); + test_skip_if(!config_opt_safety_checks); + + safety_check_set_abort(&fake_abort); + /* Buffer overflow! */ + char* ptr = malloc(128); + ptr[128] = 0; + size_t result = xallocx(ptr, 129, 0, 0); + assert_zu_eq(result, 128, ""); + free(ptr); + assert_b_eq(fake_abort_called, true, "Redzone check didn't fire."); + fake_abort_called = false; + safety_check_set_abort(NULL); +} +TEST_END + +TEST_BEGIN(test_realloc_no_overflow) { + char* ptr = malloc(128); + ptr = realloc(ptr, 256); + ptr[128] = 0; + ptr[255] = 0; + free(ptr); + + ptr = malloc(128); + ptr = realloc(ptr, 64); + ptr[63] = 0; + ptr[0] = 0; + free(ptr); +} +TEST_END + +TEST_BEGIN(test_rallocx_no_overflow) { + char* ptr = malloc(128); + ptr = rallocx(ptr, 256, 0); + ptr[128] = 0; + ptr[255] = 0; + free(ptr); + + ptr = malloc(128); + ptr = rallocx(ptr, 64, 0); + ptr[63] = 0; + ptr[0] = 0; + free(ptr); +} +TEST_END + +int +main(void) { + return test( + test_malloc_free_overflow, + test_mallocx_dallocx_overflow, + test_malloc_sdallocx_overflow, + test_realloc_overflow, + test_rallocx_overflow, + test_xallocx_overflow, + test_realloc_no_overflow, + test_rallocx_no_overflow); +} diff --git a/deps/jemalloc/test/unit/safety_check.sh b/deps/jemalloc/test/unit/safety_check.sh new file mode 100644 index 000000000..8fcc7d8a7 --- /dev/null +++ b/deps/jemalloc/test/unit/safety_check.sh @@ -0,0 +1,5 @@ +#!/bin/sh + +if [ "x${enable_prof}" = "x1" ] ; then + export MALLOC_CONF="prof:true,lg_prof_sample:0" +fi diff --git a/deps/jemalloc/test/unit/sc.c b/deps/jemalloc/test/unit/sc.c new file mode 100644 index 000000000..bf51d8e59 --- /dev/null +++ b/deps/jemalloc/test/unit/sc.c @@ -0,0 +1,33 @@ +#include "test/jemalloc_test.h" + +TEST_BEGIN(test_update_slab_size) { + sc_data_t data; + memset(&data, 0, sizeof(data)); + sc_data_init(&data); + sc_t *tiny = &data.sc[0]; + size_t tiny_size = (ZU(1) << tiny->lg_base) + + (ZU(tiny->ndelta) << tiny->lg_delta); + size_t pgs_too_big = (tiny_size * BITMAP_MAXBITS + PAGE - 1) / PAGE + 1; + sc_data_update_slab_size(&data, tiny_size, tiny_size, (int)pgs_too_big); + assert_zu_lt((size_t)tiny->pgs, pgs_too_big, "Allowed excessive pages"); + + sc_data_update_slab_size(&data, 1, 10 * PAGE, 1); + for (int i = 0; i < data.nbins; i++) { + sc_t *sc = &data.sc[i]; + size_t reg_size = (ZU(1) << sc->lg_base) + + (ZU(sc->ndelta) << sc->lg_delta); + if (reg_size <= PAGE) { + assert_d_eq(sc->pgs, 1, "Ignored valid page size hint"); + } else { + assert_d_gt(sc->pgs, 1, + "Allowed invalid page size hint"); + } + } +} +TEST_END + +int +main(void) { + return test( + test_update_slab_size); +} diff --git a/deps/jemalloc/test/unit/seq.c b/deps/jemalloc/test/unit/seq.c new file mode 100644 index 000000000..19613b0b2 --- /dev/null +++ b/deps/jemalloc/test/unit/seq.c @@ -0,0 +1,95 @@ +#include "test/jemalloc_test.h" + +#include "jemalloc/internal/seq.h" + +typedef struct data_s data_t; +struct data_s { + int arr[10]; +}; + +static void +set_data(data_t *data, int num) { + for (int i = 0; i < 10; i++) { + data->arr[i] = num; + } +} + +static void +assert_data(data_t *data) { + int num = data->arr[0]; + for (int i = 0; i < 10; i++) { + assert_d_eq(num, data->arr[i], "Data consistency error"); + } +} + +seq_define(data_t, data) + +typedef struct thd_data_s thd_data_t; +struct thd_data_s { + seq_data_t data; +}; + +static void * +seq_reader_thd(void *arg) { + thd_data_t *thd_data = (thd_data_t *)arg; + int iter = 0; + data_t local_data; + while (iter < 1000 * 1000 - 1) { + bool success = seq_try_load_data(&local_data, &thd_data->data); + if (success) { + assert_data(&local_data); + assert_d_le(iter, local_data.arr[0], + "Seq read went back in time."); + iter = local_data.arr[0]; + } + } + return NULL; +} + +static void * +seq_writer_thd(void *arg) { + thd_data_t *thd_data = (thd_data_t *)arg; + data_t local_data; + memset(&local_data, 0, sizeof(local_data)); + for (int i = 0; i < 1000 * 1000; i++) { + set_data(&local_data, i); + seq_store_data(&thd_data->data, &local_data); + } + return NULL; +} + +TEST_BEGIN(test_seq_threaded) { + thd_data_t thd_data; + memset(&thd_data, 0, sizeof(thd_data)); + + thd_t reader; + thd_t writer; + + thd_create(&reader, seq_reader_thd, &thd_data); + thd_create(&writer, seq_writer_thd, &thd_data); + + thd_join(reader, NULL); + thd_join(writer, NULL); +} +TEST_END + +TEST_BEGIN(test_seq_simple) { + data_t data; + seq_data_t seq; + memset(&seq, 0, sizeof(seq)); + for (int i = 0; i < 1000 * 1000; i++) { + set_data(&data, i); + seq_store_data(&seq, &data); + set_data(&data, 0); + bool success = seq_try_load_data(&data, &seq); + assert_b_eq(success, true, "Failed non-racing read"); + assert_data(&data); + } +} +TEST_END + +int main(void) { + return test_no_reentrancy( + test_seq_simple, + test_seq_threaded); +} diff --git a/deps/jemalloc/test/unit/size_classes.c b/deps/jemalloc/test/unit/size_classes.c new file mode 100644 index 000000000..694733635 --- /dev/null +++ b/deps/jemalloc/test/unit/size_classes.c @@ -0,0 +1,188 @@ +#include "test/jemalloc_test.h" + +static size_t +get_max_size_class(void) { + unsigned nlextents; + size_t mib[4]; + size_t sz, miblen, max_size_class; + + sz = sizeof(unsigned); + assert_d_eq(mallctl("arenas.nlextents", (void *)&nlextents, &sz, NULL, + 0), 0, "Unexpected mallctl() error"); + + miblen = sizeof(mib) / sizeof(size_t); + assert_d_eq(mallctlnametomib("arenas.lextent.0.size", mib, &miblen), 0, + "Unexpected mallctlnametomib() error"); + mib[2] = nlextents - 1; + + sz = sizeof(size_t); + assert_d_eq(mallctlbymib(mib, miblen, (void *)&max_size_class, &sz, + NULL, 0), 0, "Unexpected mallctlbymib() error"); + + return max_size_class; +} + +TEST_BEGIN(test_size_classes) { + size_t size_class, max_size_class; + szind_t index, max_index; + + max_size_class = get_max_size_class(); + max_index = sz_size2index(max_size_class); + + for (index = 0, size_class = sz_index2size(index); index < max_index || + size_class < max_size_class; index++, size_class = + sz_index2size(index)) { + assert_true(index < max_index, + "Loop conditionals should be equivalent; index=%u, " + "size_class=%zu (%#zx)", index, size_class, size_class); + assert_true(size_class < max_size_class, + "Loop conditionals should be equivalent; index=%u, " + "size_class=%zu (%#zx)", index, size_class, size_class); + + assert_u_eq(index, sz_size2index(size_class), + "sz_size2index() does not reverse sz_index2size(): index=%u" + " --> size_class=%zu --> index=%u --> size_class=%zu", + index, size_class, sz_size2index(size_class), + sz_index2size(sz_size2index(size_class))); + assert_zu_eq(size_class, + sz_index2size(sz_size2index(size_class)), + "sz_index2size() does not reverse sz_size2index(): index=%u" + " --> size_class=%zu --> index=%u --> size_class=%zu", + index, size_class, sz_size2index(size_class), + sz_index2size(sz_size2index(size_class))); + + assert_u_eq(index+1, sz_size2index(size_class+1), + "Next size_class does not round up properly"); + + assert_zu_eq(size_class, (index > 0) ? + sz_s2u(sz_index2size(index-1)+1) : sz_s2u(1), + "sz_s2u() does not round up to size class"); + assert_zu_eq(size_class, sz_s2u(size_class-1), + "sz_s2u() does not round up to size class"); + assert_zu_eq(size_class, sz_s2u(size_class), + "sz_s2u() does not compute same size class"); + assert_zu_eq(sz_s2u(size_class+1), sz_index2size(index+1), + "sz_s2u() does not round up to next size class"); + } + + assert_u_eq(index, sz_size2index(sz_index2size(index)), + "sz_size2index() does not reverse sz_index2size()"); + assert_zu_eq(max_size_class, sz_index2size( + sz_size2index(max_size_class)), + "sz_index2size() does not reverse sz_size2index()"); + + assert_zu_eq(size_class, sz_s2u(sz_index2size(index-1)+1), + "sz_s2u() does not round up to size class"); + assert_zu_eq(size_class, sz_s2u(size_class-1), + "sz_s2u() does not round up to size class"); + assert_zu_eq(size_class, sz_s2u(size_class), + "sz_s2u() does not compute same size class"); +} +TEST_END + +TEST_BEGIN(test_psize_classes) { + size_t size_class, max_psz; + pszind_t pind, max_pind; + + max_psz = get_max_size_class() + PAGE; + max_pind = sz_psz2ind(max_psz); + + for (pind = 0, size_class = sz_pind2sz(pind); + pind < max_pind || size_class < max_psz; + pind++, size_class = sz_pind2sz(pind)) { + assert_true(pind < max_pind, + "Loop conditionals should be equivalent; pind=%u, " + "size_class=%zu (%#zx)", pind, size_class, size_class); + assert_true(size_class < max_psz, + "Loop conditionals should be equivalent; pind=%u, " + "size_class=%zu (%#zx)", pind, size_class, size_class); + + assert_u_eq(pind, sz_psz2ind(size_class), + "sz_psz2ind() does not reverse sz_pind2sz(): pind=%u -->" + " size_class=%zu --> pind=%u --> size_class=%zu", pind, + size_class, sz_psz2ind(size_class), + sz_pind2sz(sz_psz2ind(size_class))); + assert_zu_eq(size_class, sz_pind2sz(sz_psz2ind(size_class)), + "sz_pind2sz() does not reverse sz_psz2ind(): pind=%u -->" + " size_class=%zu --> pind=%u --> size_class=%zu", pind, + size_class, sz_psz2ind(size_class), + sz_pind2sz(sz_psz2ind(size_class))); + + if (size_class == SC_LARGE_MAXCLASS) { + assert_u_eq(SC_NPSIZES, sz_psz2ind(size_class + 1), + "Next size_class does not round up properly"); + } else { + assert_u_eq(pind + 1, sz_psz2ind(size_class + 1), + "Next size_class does not round up properly"); + } + + assert_zu_eq(size_class, (pind > 0) ? + sz_psz2u(sz_pind2sz(pind-1)+1) : sz_psz2u(1), + "sz_psz2u() does not round up to size class"); + assert_zu_eq(size_class, sz_psz2u(size_class-1), + "sz_psz2u() does not round up to size class"); + assert_zu_eq(size_class, sz_psz2u(size_class), + "sz_psz2u() does not compute same size class"); + assert_zu_eq(sz_psz2u(size_class+1), sz_pind2sz(pind+1), + "sz_psz2u() does not round up to next size class"); + } + + assert_u_eq(pind, sz_psz2ind(sz_pind2sz(pind)), + "sz_psz2ind() does not reverse sz_pind2sz()"); + assert_zu_eq(max_psz, sz_pind2sz(sz_psz2ind(max_psz)), + "sz_pind2sz() does not reverse sz_psz2ind()"); + + assert_zu_eq(size_class, sz_psz2u(sz_pind2sz(pind-1)+1), + "sz_psz2u() does not round up to size class"); + assert_zu_eq(size_class, sz_psz2u(size_class-1), + "sz_psz2u() does not round up to size class"); + assert_zu_eq(size_class, sz_psz2u(size_class), + "sz_psz2u() does not compute same size class"); +} +TEST_END + +TEST_BEGIN(test_overflow) { + size_t max_size_class, max_psz; + + max_size_class = get_max_size_class(); + max_psz = max_size_class + PAGE; + + assert_u_eq(sz_size2index(max_size_class+1), SC_NSIZES, + "sz_size2index() should return NSIZES on overflow"); + assert_u_eq(sz_size2index(ZU(PTRDIFF_MAX)+1), SC_NSIZES, + "sz_size2index() should return NSIZES on overflow"); + assert_u_eq(sz_size2index(SIZE_T_MAX), SC_NSIZES, + "sz_size2index() should return NSIZES on overflow"); + + assert_zu_eq(sz_s2u(max_size_class+1), 0, + "sz_s2u() should return 0 for unsupported size"); + assert_zu_eq(sz_s2u(ZU(PTRDIFF_MAX)+1), 0, + "sz_s2u() should return 0 for unsupported size"); + assert_zu_eq(sz_s2u(SIZE_T_MAX), 0, + "sz_s2u() should return 0 on overflow"); + + assert_u_eq(sz_psz2ind(max_size_class+1), SC_NPSIZES, + "sz_psz2ind() should return NPSIZES on overflow"); + assert_u_eq(sz_psz2ind(ZU(PTRDIFF_MAX)+1), SC_NPSIZES, + "sz_psz2ind() should return NPSIZES on overflow"); + assert_u_eq(sz_psz2ind(SIZE_T_MAX), SC_NPSIZES, + "sz_psz2ind() should return NPSIZES on overflow"); + + assert_zu_eq(sz_psz2u(max_size_class+1), max_psz, + "sz_psz2u() should return (LARGE_MAXCLASS + PAGE) for unsupported" + " size"); + assert_zu_eq(sz_psz2u(ZU(PTRDIFF_MAX)+1), max_psz, + "sz_psz2u() should return (LARGE_MAXCLASS + PAGE) for unsupported " + "size"); + assert_zu_eq(sz_psz2u(SIZE_T_MAX), max_psz, + "sz_psz2u() should return (LARGE_MAXCLASS + PAGE) on overflow"); +} +TEST_END + +int +main(void) { + return test( + test_size_classes, + test_psize_classes, + test_overflow); +} diff --git a/deps/jemalloc/test/unit/slab.c b/deps/jemalloc/test/unit/slab.c new file mode 100644 index 000000000..c56af25fe --- /dev/null +++ b/deps/jemalloc/test/unit/slab.c @@ -0,0 +1,33 @@ +#include "test/jemalloc_test.h" + +TEST_BEGIN(test_arena_slab_regind) { + szind_t binind; + + for (binind = 0; binind < SC_NBINS; binind++) { + size_t regind; + extent_t slab; + const bin_info_t *bin_info = &bin_infos[binind]; + extent_init(&slab, NULL, mallocx(bin_info->slab_size, + MALLOCX_LG_ALIGN(LG_PAGE)), bin_info->slab_size, true, + binind, 0, extent_state_active, false, true, true, + EXTENT_NOT_HEAD); + assert_ptr_not_null(extent_addr_get(&slab), + "Unexpected malloc() failure"); + for (regind = 0; regind < bin_info->nregs; regind++) { + void *reg = (void *)((uintptr_t)extent_addr_get(&slab) + + (bin_info->reg_size * regind)); + assert_zu_eq(arena_slab_regind(&slab, binind, reg), + regind, + "Incorrect region index computed for size %zu", + bin_info->reg_size); + } + free(extent_addr_get(&slab)); + } +} +TEST_END + +int +main(void) { + return test( + test_arena_slab_regind); +} diff --git a/deps/jemalloc/test/unit/smoothstep.c b/deps/jemalloc/test/unit/smoothstep.c new file mode 100644 index 000000000..7c5dbb7e0 --- /dev/null +++ b/deps/jemalloc/test/unit/smoothstep.c @@ -0,0 +1,102 @@ +#include "test/jemalloc_test.h" + +static const uint64_t smoothstep_tab[] = { +#define STEP(step, h, x, y) \ + h, + SMOOTHSTEP +#undef STEP +}; + +TEST_BEGIN(test_smoothstep_integral) { + uint64_t sum, min, max; + unsigned i; + + /* + * The integral of smoothstep in the [0..1] range equals 1/2. Verify + * that the fixed point representation's integral is no more than + * rounding error distant from 1/2. Regarding rounding, each table + * element is rounded down to the nearest fixed point value, so the + * integral may be off by as much as SMOOTHSTEP_NSTEPS ulps. + */ + sum = 0; + for (i = 0; i < SMOOTHSTEP_NSTEPS; i++) { + sum += smoothstep_tab[i]; + } + + max = (KQU(1) << (SMOOTHSTEP_BFP-1)) * (SMOOTHSTEP_NSTEPS+1); + min = max - SMOOTHSTEP_NSTEPS; + + assert_u64_ge(sum, min, + "Integral too small, even accounting for truncation"); + assert_u64_le(sum, max, "Integral exceeds 1/2"); + if (false) { + malloc_printf("%"FMTu64" ulps under 1/2 (limit %d)\n", + max - sum, SMOOTHSTEP_NSTEPS); + } +} +TEST_END + +TEST_BEGIN(test_smoothstep_monotonic) { + uint64_t prev_h; + unsigned i; + + /* + * The smoothstep function is monotonic in [0..1], i.e. its slope is + * non-negative. In practice we want to parametrize table generation + * such that piecewise slope is greater than zero, but do not require + * that here. + */ + prev_h = 0; + for (i = 0; i < SMOOTHSTEP_NSTEPS; i++) { + uint64_t h = smoothstep_tab[i]; + assert_u64_ge(h, prev_h, "Piecewise non-monotonic, i=%u", i); + prev_h = h; + } + assert_u64_eq(smoothstep_tab[SMOOTHSTEP_NSTEPS-1], + (KQU(1) << SMOOTHSTEP_BFP), "Last step must equal 1"); +} +TEST_END + +TEST_BEGIN(test_smoothstep_slope) { + uint64_t prev_h, prev_delta; + unsigned i; + + /* + * The smoothstep slope strictly increases until x=0.5, and then + * strictly decreases until x=1.0. Verify the slightly weaker + * requirement of monotonicity, so that inadequate table precision does + * not cause false test failures. + */ + prev_h = 0; + prev_delta = 0; + for (i = 0; i < SMOOTHSTEP_NSTEPS / 2 + SMOOTHSTEP_NSTEPS % 2; i++) { + uint64_t h = smoothstep_tab[i]; + uint64_t delta = h - prev_h; + assert_u64_ge(delta, prev_delta, + "Slope must monotonically increase in 0.0 <= x <= 0.5, " + "i=%u", i); + prev_h = h; + prev_delta = delta; + } + + prev_h = KQU(1) << SMOOTHSTEP_BFP; + prev_delta = 0; + for (i = SMOOTHSTEP_NSTEPS-1; i >= SMOOTHSTEP_NSTEPS / 2; i--) { + uint64_t h = smoothstep_tab[i]; + uint64_t delta = prev_h - h; + assert_u64_ge(delta, prev_delta, + "Slope must monotonically decrease in 0.5 <= x <= 1.0, " + "i=%u", i); + prev_h = h; + prev_delta = delta; + } +} +TEST_END + +int +main(void) { + return test( + test_smoothstep_integral, + test_smoothstep_monotonic, + test_smoothstep_slope); +} diff --git a/deps/jemalloc/test/unit/spin.c b/deps/jemalloc/test/unit/spin.c new file mode 100644 index 000000000..b965f7427 --- /dev/null +++ b/deps/jemalloc/test/unit/spin.c @@ -0,0 +1,18 @@ +#include "test/jemalloc_test.h" + +#include "jemalloc/internal/spin.h" + +TEST_BEGIN(test_spin) { + spin_t spinner = SPIN_INITIALIZER; + + for (unsigned i = 0; i < 100; i++) { + spin_adaptive(&spinner); + } +} +TEST_END + +int +main(void) { + return test( + test_spin); +} diff --git a/deps/jemalloc/test/unit/stats.c b/deps/jemalloc/test/unit/stats.c new file mode 100644 index 000000000..646768e88 --- /dev/null +++ b/deps/jemalloc/test/unit/stats.c @@ -0,0 +1,374 @@ +#include "test/jemalloc_test.h" + +TEST_BEGIN(test_stats_summary) { + size_t sz, allocated, active, resident, mapped; + int expected = config_stats ? 0 : ENOENT; + + sz = sizeof(size_t); + assert_d_eq(mallctl("stats.allocated", (void *)&allocated, &sz, NULL, + 0), expected, "Unexpected mallctl() result"); + assert_d_eq(mallctl("stats.active", (void *)&active, &sz, NULL, 0), + expected, "Unexpected mallctl() result"); + assert_d_eq(mallctl("stats.resident", (void *)&resident, &sz, NULL, 0), + expected, "Unexpected mallctl() result"); + assert_d_eq(mallctl("stats.mapped", (void *)&mapped, &sz, NULL, 0), + expected, "Unexpected mallctl() result"); + + if (config_stats) { + assert_zu_le(allocated, active, + "allocated should be no larger than active"); + assert_zu_lt(active, resident, + "active should be less than resident"); + assert_zu_lt(active, mapped, + "active should be less than mapped"); + } +} +TEST_END + +TEST_BEGIN(test_stats_large) { + void *p; + uint64_t epoch; + size_t allocated; + uint64_t nmalloc, ndalloc, nrequests; + size_t sz; + int expected = config_stats ? 0 : ENOENT; + + p = mallocx(SC_SMALL_MAXCLASS + 1, MALLOCX_ARENA(0)); + assert_ptr_not_null(p, "Unexpected mallocx() failure"); + + assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch, sizeof(epoch)), + 0, "Unexpected mallctl() failure"); + + sz = sizeof(size_t); + assert_d_eq(mallctl("stats.arenas.0.large.allocated", + (void *)&allocated, &sz, NULL, 0), expected, + "Unexpected mallctl() result"); + sz = sizeof(uint64_t); + assert_d_eq(mallctl("stats.arenas.0.large.nmalloc", (void *)&nmalloc, + &sz, NULL, 0), expected, "Unexpected mallctl() result"); + assert_d_eq(mallctl("stats.arenas.0.large.ndalloc", (void *)&ndalloc, + &sz, NULL, 0), expected, "Unexpected mallctl() result"); + assert_d_eq(mallctl("stats.arenas.0.large.nrequests", + (void *)&nrequests, &sz, NULL, 0), expected, + "Unexpected mallctl() result"); + + if (config_stats) { + assert_zu_gt(allocated, 0, + "allocated should be greater than zero"); + assert_u64_ge(nmalloc, ndalloc, + "nmalloc should be at least as large as ndalloc"); + assert_u64_le(nmalloc, nrequests, + "nmalloc should no larger than nrequests"); + } + + dallocx(p, 0); +} +TEST_END + +TEST_BEGIN(test_stats_arenas_summary) { + void *little, *large; + uint64_t epoch; + size_t sz; + int expected = config_stats ? 0 : ENOENT; + size_t mapped; + uint64_t dirty_npurge, dirty_nmadvise, dirty_purged; + uint64_t muzzy_npurge, muzzy_nmadvise, muzzy_purged; + + little = mallocx(SC_SMALL_MAXCLASS, MALLOCX_ARENA(0)); + assert_ptr_not_null(little, "Unexpected mallocx() failure"); + large = mallocx((1U << SC_LG_LARGE_MINCLASS), + MALLOCX_ARENA(0)); + assert_ptr_not_null(large, "Unexpected mallocx() failure"); + + dallocx(little, 0); + dallocx(large, 0); + + assert_d_eq(mallctl("thread.tcache.flush", NULL, NULL, NULL, 0), + opt_tcache ? 0 : EFAULT, "Unexpected mallctl() result"); + assert_d_eq(mallctl("arena.0.purge", NULL, NULL, NULL, 0), 0, + "Unexpected mallctl() failure"); + + assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch, sizeof(epoch)), + 0, "Unexpected mallctl() failure"); + + sz = sizeof(size_t); + assert_d_eq(mallctl("stats.arenas.0.mapped", (void *)&mapped, &sz, NULL, + 0), expected, "Unexepected mallctl() result"); + + sz = sizeof(uint64_t); + assert_d_eq(mallctl("stats.arenas.0.dirty_npurge", + (void *)&dirty_npurge, &sz, NULL, 0), expected, + "Unexepected mallctl() result"); + assert_d_eq(mallctl("stats.arenas.0.dirty_nmadvise", + (void *)&dirty_nmadvise, &sz, NULL, 0), expected, + "Unexepected mallctl() result"); + assert_d_eq(mallctl("stats.arenas.0.dirty_purged", + (void *)&dirty_purged, &sz, NULL, 0), expected, + "Unexepected mallctl() result"); + assert_d_eq(mallctl("stats.arenas.0.muzzy_npurge", + (void *)&muzzy_npurge, &sz, NULL, 0), expected, + "Unexepected mallctl() result"); + assert_d_eq(mallctl("stats.arenas.0.muzzy_nmadvise", + (void *)&muzzy_nmadvise, &sz, NULL, 0), expected, + "Unexepected mallctl() result"); + assert_d_eq(mallctl("stats.arenas.0.muzzy_purged", + (void *)&muzzy_purged, &sz, NULL, 0), expected, + "Unexepected mallctl() result"); + + if (config_stats) { + if (!background_thread_enabled()) { + assert_u64_gt(dirty_npurge + muzzy_npurge, 0, + "At least one purge should have occurred"); + } + assert_u64_le(dirty_nmadvise, dirty_purged, + "dirty_nmadvise should be no greater than dirty_purged"); + assert_u64_le(muzzy_nmadvise, muzzy_purged, + "muzzy_nmadvise should be no greater than muzzy_purged"); + } +} +TEST_END + +void * +thd_start(void *arg) { + return NULL; +} + +static void +no_lazy_lock(void) { + thd_t thd; + + thd_create(&thd, thd_start, NULL); + thd_join(thd, NULL); +} + +TEST_BEGIN(test_stats_arenas_small) { + void *p; + size_t sz, allocated; + uint64_t epoch, nmalloc, ndalloc, nrequests; + int expected = config_stats ? 0 : ENOENT; + + no_lazy_lock(); /* Lazy locking would dodge tcache testing. */ + + p = mallocx(SC_SMALL_MAXCLASS, MALLOCX_ARENA(0)); + assert_ptr_not_null(p, "Unexpected mallocx() failure"); + + assert_d_eq(mallctl("thread.tcache.flush", NULL, NULL, NULL, 0), + opt_tcache ? 0 : EFAULT, "Unexpected mallctl() result"); + + assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch, sizeof(epoch)), + 0, "Unexpected mallctl() failure"); + + sz = sizeof(size_t); + assert_d_eq(mallctl("stats.arenas.0.small.allocated", + (void *)&allocated, &sz, NULL, 0), expected, + "Unexpected mallctl() result"); + sz = sizeof(uint64_t); + assert_d_eq(mallctl("stats.arenas.0.small.nmalloc", (void *)&nmalloc, + &sz, NULL, 0), expected, "Unexpected mallctl() result"); + assert_d_eq(mallctl("stats.arenas.0.small.ndalloc", (void *)&ndalloc, + &sz, NULL, 0), expected, "Unexpected mallctl() result"); + assert_d_eq(mallctl("stats.arenas.0.small.nrequests", + (void *)&nrequests, &sz, NULL, 0), expected, + "Unexpected mallctl() result"); + + if (config_stats) { + assert_zu_gt(allocated, 0, + "allocated should be greater than zero"); + assert_u64_gt(nmalloc, 0, + "nmalloc should be no greater than zero"); + assert_u64_ge(nmalloc, ndalloc, + "nmalloc should be at least as large as ndalloc"); + assert_u64_gt(nrequests, 0, + "nrequests should be greater than zero"); + } + + dallocx(p, 0); +} +TEST_END + +TEST_BEGIN(test_stats_arenas_large) { + void *p; + size_t sz, allocated; + uint64_t epoch, nmalloc, ndalloc; + int expected = config_stats ? 0 : ENOENT; + + p = mallocx((1U << SC_LG_LARGE_MINCLASS), MALLOCX_ARENA(0)); + assert_ptr_not_null(p, "Unexpected mallocx() failure"); + + assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch, sizeof(epoch)), + 0, "Unexpected mallctl() failure"); + + sz = sizeof(size_t); + assert_d_eq(mallctl("stats.arenas.0.large.allocated", + (void *)&allocated, &sz, NULL, 0), expected, + "Unexpected mallctl() result"); + sz = sizeof(uint64_t); + assert_d_eq(mallctl("stats.arenas.0.large.nmalloc", (void *)&nmalloc, + &sz, NULL, 0), expected, "Unexpected mallctl() result"); + assert_d_eq(mallctl("stats.arenas.0.large.ndalloc", (void *)&ndalloc, + &sz, NULL, 0), expected, "Unexpected mallctl() result"); + + if (config_stats) { + assert_zu_gt(allocated, 0, + "allocated should be greater than zero"); + assert_u64_gt(nmalloc, 0, + "nmalloc should be greater than zero"); + assert_u64_ge(nmalloc, ndalloc, + "nmalloc should be at least as large as ndalloc"); + } + + dallocx(p, 0); +} +TEST_END + +static void +gen_mallctl_str(char *cmd, char *name, unsigned arena_ind) { + sprintf(cmd, "stats.arenas.%u.bins.0.%s", arena_ind, name); +} + +TEST_BEGIN(test_stats_arenas_bins) { + void *p; + size_t sz, curslabs, curregs, nonfull_slabs; + uint64_t epoch, nmalloc, ndalloc, nrequests, nfills, nflushes; + uint64_t nslabs, nreslabs; + int expected = config_stats ? 0 : ENOENT; + + /* Make sure allocation below isn't satisfied by tcache. */ + assert_d_eq(mallctl("thread.tcache.flush", NULL, NULL, NULL, 0), + opt_tcache ? 0 : EFAULT, "Unexpected mallctl() result"); + + unsigned arena_ind, old_arena_ind; + sz = sizeof(unsigned); + assert_d_eq(mallctl("arenas.create", (void *)&arena_ind, &sz, NULL, 0), + 0, "Arena creation failure"); + sz = sizeof(arena_ind); + assert_d_eq(mallctl("thread.arena", (void *)&old_arena_ind, &sz, + (void *)&arena_ind, sizeof(arena_ind)), 0, + "Unexpected mallctl() failure"); + + p = malloc(bin_infos[0].reg_size); + assert_ptr_not_null(p, "Unexpected malloc() failure"); + + assert_d_eq(mallctl("thread.tcache.flush", NULL, NULL, NULL, 0), + opt_tcache ? 0 : EFAULT, "Unexpected mallctl() result"); + + assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch, sizeof(epoch)), + 0, "Unexpected mallctl() failure"); + + char cmd[128]; + sz = sizeof(uint64_t); + gen_mallctl_str(cmd, "nmalloc", arena_ind); + assert_d_eq(mallctl(cmd, (void *)&nmalloc, &sz, NULL, 0), expected, + "Unexpected mallctl() result"); + gen_mallctl_str(cmd, "ndalloc", arena_ind); + assert_d_eq(mallctl(cmd, (void *)&ndalloc, &sz, NULL, 0), expected, + "Unexpected mallctl() result"); + gen_mallctl_str(cmd, "nrequests", arena_ind); + assert_d_eq(mallctl(cmd, (void *)&nrequests, &sz, NULL, 0), expected, + "Unexpected mallctl() result"); + sz = sizeof(size_t); + gen_mallctl_str(cmd, "curregs", arena_ind); + assert_d_eq(mallctl(cmd, (void *)&curregs, &sz, NULL, 0), expected, + "Unexpected mallctl() result"); + + sz = sizeof(uint64_t); + gen_mallctl_str(cmd, "nfills", arena_ind); + assert_d_eq(mallctl(cmd, (void *)&nfills, &sz, NULL, 0), expected, + "Unexpected mallctl() result"); + gen_mallctl_str(cmd, "nflushes", arena_ind); + assert_d_eq(mallctl(cmd, (void *)&nflushes, &sz, NULL, 0), expected, + "Unexpected mallctl() result"); + + gen_mallctl_str(cmd, "nslabs", arena_ind); + assert_d_eq(mallctl(cmd, (void *)&nslabs, &sz, NULL, 0), expected, + "Unexpected mallctl() result"); + gen_mallctl_str(cmd, "nreslabs", arena_ind); + assert_d_eq(mallctl(cmd, (void *)&nreslabs, &sz, NULL, 0), expected, + "Unexpected mallctl() result"); + sz = sizeof(size_t); + gen_mallctl_str(cmd, "curslabs", arena_ind); + assert_d_eq(mallctl(cmd, (void *)&curslabs, &sz, NULL, 0), expected, + "Unexpected mallctl() result"); + gen_mallctl_str(cmd, "nonfull_slabs", arena_ind); + assert_d_eq(mallctl(cmd, (void *)&nonfull_slabs, &sz, NULL, 0), + expected, "Unexpected mallctl() result"); + + if (config_stats) { + assert_u64_gt(nmalloc, 0, + "nmalloc should be greater than zero"); + assert_u64_ge(nmalloc, ndalloc, + "nmalloc should be at least as large as ndalloc"); + assert_u64_gt(nrequests, 0, + "nrequests should be greater than zero"); + assert_zu_gt(curregs, 0, + "allocated should be greater than zero"); + if (opt_tcache) { + assert_u64_gt(nfills, 0, + "At least one fill should have occurred"); + assert_u64_gt(nflushes, 0, + "At least one flush should have occurred"); + } + assert_u64_gt(nslabs, 0, + "At least one slab should have been allocated"); + assert_zu_gt(curslabs, 0, + "At least one slab should be currently allocated"); + assert_zu_eq(nonfull_slabs, 0, + "slabs_nonfull should be empty"); + } + + dallocx(p, 0); +} +TEST_END + +TEST_BEGIN(test_stats_arenas_lextents) { + void *p; + uint64_t epoch, nmalloc, ndalloc; + size_t curlextents, sz, hsize; + int expected = config_stats ? 0 : ENOENT; + + sz = sizeof(size_t); + assert_d_eq(mallctl("arenas.lextent.0.size", (void *)&hsize, &sz, NULL, + 0), 0, "Unexpected mallctl() failure"); + + p = mallocx(hsize, MALLOCX_ARENA(0)); + assert_ptr_not_null(p, "Unexpected mallocx() failure"); + + assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch, sizeof(epoch)), + 0, "Unexpected mallctl() failure"); + + sz = sizeof(uint64_t); + assert_d_eq(mallctl("stats.arenas.0.lextents.0.nmalloc", + (void *)&nmalloc, &sz, NULL, 0), expected, + "Unexpected mallctl() result"); + assert_d_eq(mallctl("stats.arenas.0.lextents.0.ndalloc", + (void *)&ndalloc, &sz, NULL, 0), expected, + "Unexpected mallctl() result"); + sz = sizeof(size_t); + assert_d_eq(mallctl("stats.arenas.0.lextents.0.curlextents", + (void *)&curlextents, &sz, NULL, 0), expected, + "Unexpected mallctl() result"); + + if (config_stats) { + assert_u64_gt(nmalloc, 0, + "nmalloc should be greater than zero"); + assert_u64_ge(nmalloc, ndalloc, + "nmalloc should be at least as large as ndalloc"); + assert_u64_gt(curlextents, 0, + "At least one extent should be currently allocated"); + } + + dallocx(p, 0); +} +TEST_END + +int +main(void) { + return test_no_reentrancy( + test_stats_summary, + test_stats_large, + test_stats_arenas_summary, + test_stats_arenas_small, + test_stats_arenas_large, + test_stats_arenas_bins, + test_stats_arenas_lextents); +} diff --git a/deps/jemalloc/test/unit/stats_print.c b/deps/jemalloc/test/unit/stats_print.c new file mode 100644 index 000000000..014d002fd --- /dev/null +++ b/deps/jemalloc/test/unit/stats_print.c @@ -0,0 +1,999 @@ +#include "test/jemalloc_test.h" + +#include "jemalloc/internal/util.h" + +typedef enum { + TOKEN_TYPE_NONE, + TOKEN_TYPE_ERROR, + TOKEN_TYPE_EOI, + TOKEN_TYPE_NULL, + TOKEN_TYPE_FALSE, + TOKEN_TYPE_TRUE, + TOKEN_TYPE_LBRACKET, + TOKEN_TYPE_RBRACKET, + TOKEN_TYPE_LBRACE, + TOKEN_TYPE_RBRACE, + TOKEN_TYPE_COLON, + TOKEN_TYPE_COMMA, + TOKEN_TYPE_STRING, + TOKEN_TYPE_NUMBER +} token_type_t; + +typedef struct parser_s parser_t; +typedef struct { + parser_t *parser; + token_type_t token_type; + size_t pos; + size_t len; + size_t line; + size_t col; +} token_t; + +struct parser_s { + bool verbose; + char *buf; /* '\0'-terminated. */ + size_t len; /* Number of characters preceding '\0' in buf. */ + size_t pos; + size_t line; + size_t col; + token_t token; +}; + +static void +token_init(token_t *token, parser_t *parser, token_type_t token_type, + size_t pos, size_t len, size_t line, size_t col) { + token->parser = parser; + token->token_type = token_type; + token->pos = pos; + token->len = len; + token->line = line; + token->col = col; +} + +static void +token_error(token_t *token) { + if (!token->parser->verbose) { + return; + } + switch (token->token_type) { + case TOKEN_TYPE_NONE: + not_reached(); + case TOKEN_TYPE_ERROR: + malloc_printf("%zu:%zu: Unexpected character in token: ", + token->line, token->col); + break; + default: + malloc_printf("%zu:%zu: Unexpected token: ", token->line, + token->col); + break; + } + UNUSED ssize_t err = malloc_write_fd(STDERR_FILENO, + &token->parser->buf[token->pos], token->len); + malloc_printf("\n"); +} + +static void +parser_init(parser_t *parser, bool verbose) { + parser->verbose = verbose; + parser->buf = NULL; + parser->len = 0; + parser->pos = 0; + parser->line = 1; + parser->col = 0; +} + +static void +parser_fini(parser_t *parser) { + if (parser->buf != NULL) { + dallocx(parser->buf, MALLOCX_TCACHE_NONE); + } +} + +static bool +parser_append(parser_t *parser, const char *str) { + size_t len = strlen(str); + char *buf = (parser->buf == NULL) ? mallocx(len + 1, + MALLOCX_TCACHE_NONE) : rallocx(parser->buf, parser->len + len + 1, + MALLOCX_TCACHE_NONE); + if (buf == NULL) { + return true; + } + memcpy(&buf[parser->len], str, len + 1); + parser->buf = buf; + parser->len += len; + return false; +} + +static bool +parser_tokenize(parser_t *parser) { + enum { + STATE_START, + STATE_EOI, + STATE_N, STATE_NU, STATE_NUL, STATE_NULL, + STATE_F, STATE_FA, STATE_FAL, STATE_FALS, STATE_FALSE, + STATE_T, STATE_TR, STATE_TRU, STATE_TRUE, + STATE_LBRACKET, + STATE_RBRACKET, + STATE_LBRACE, + STATE_RBRACE, + STATE_COLON, + STATE_COMMA, + STATE_CHARS, + STATE_CHAR_ESCAPE, + STATE_CHAR_U, STATE_CHAR_UD, STATE_CHAR_UDD, STATE_CHAR_UDDD, + STATE_STRING, + STATE_MINUS, + STATE_LEADING_ZERO, + STATE_DIGITS, + STATE_DECIMAL, + STATE_FRAC_DIGITS, + STATE_EXP, + STATE_EXP_SIGN, + STATE_EXP_DIGITS, + STATE_ACCEPT + } state = STATE_START; + size_t token_pos JEMALLOC_CC_SILENCE_INIT(0); + size_t token_line JEMALLOC_CC_SILENCE_INIT(1); + size_t token_col JEMALLOC_CC_SILENCE_INIT(0); + + assert_zu_le(parser->pos, parser->len, + "Position is past end of buffer"); + + while (state != STATE_ACCEPT) { + char c = parser->buf[parser->pos]; + + switch (state) { + case STATE_START: + token_pos = parser->pos; + token_line = parser->line; + token_col = parser->col; + switch (c) { + case ' ': case '\b': case '\n': case '\r': case '\t': + break; + case '\0': + state = STATE_EOI; + break; + case 'n': + state = STATE_N; + break; + case 'f': + state = STATE_F; + break; + case 't': + state = STATE_T; + break; + case '[': + state = STATE_LBRACKET; + break; + case ']': + state = STATE_RBRACKET; + break; + case '{': + state = STATE_LBRACE; + break; + case '}': + state = STATE_RBRACE; + break; + case ':': + state = STATE_COLON; + break; + case ',': + state = STATE_COMMA; + break; + case '"': + state = STATE_CHARS; + break; + case '-': + state = STATE_MINUS; + break; + case '0': + state = STATE_LEADING_ZERO; + break; + case '1': case '2': case '3': case '4': + case '5': case '6': case '7': case '8': case '9': + state = STATE_DIGITS; + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_ERROR, token_pos, parser->pos + 1 + - token_pos, token_line, token_col); + return true; + } + break; + case STATE_EOI: + token_init(&parser->token, parser, + TOKEN_TYPE_EOI, token_pos, parser->pos - + token_pos, token_line, token_col); + state = STATE_ACCEPT; + break; + case STATE_N: + switch (c) { + case 'u': + state = STATE_NU; + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_ERROR, token_pos, parser->pos + 1 + - token_pos, token_line, token_col); + return true; + } + break; + case STATE_NU: + switch (c) { + case 'l': + state = STATE_NUL; + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_ERROR, token_pos, parser->pos + 1 + - token_pos, token_line, token_col); + return true; + } + break; + case STATE_NUL: + switch (c) { + case 'l': + state = STATE_NULL; + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_ERROR, token_pos, parser->pos + 1 + - token_pos, token_line, token_col); + return true; + } + break; + case STATE_NULL: + switch (c) { + case ' ': case '\b': case '\n': case '\r': case '\t': + case '\0': + case '[': case ']': case '{': case '}': case ':': + case ',': + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_ERROR, token_pos, parser->pos + 1 + - token_pos, token_line, token_col); + return true; + } + token_init(&parser->token, parser, TOKEN_TYPE_NULL, + token_pos, parser->pos - token_pos, token_line, + token_col); + state = STATE_ACCEPT; + break; + case STATE_F: + switch (c) { + case 'a': + state = STATE_FA; + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_ERROR, token_pos, parser->pos + 1 + - token_pos, token_line, token_col); + return true; + } + break; + case STATE_FA: + switch (c) { + case 'l': + state = STATE_FAL; + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_ERROR, token_pos, parser->pos + 1 + - token_pos, token_line, token_col); + return true; + } + break; + case STATE_FAL: + switch (c) { + case 's': + state = STATE_FALS; + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_ERROR, token_pos, parser->pos + 1 + - token_pos, token_line, token_col); + return true; + } + break; + case STATE_FALS: + switch (c) { + case 'e': + state = STATE_FALSE; + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_ERROR, token_pos, parser->pos + 1 + - token_pos, token_line, token_col); + return true; + } + break; + case STATE_FALSE: + switch (c) { + case ' ': case '\b': case '\n': case '\r': case '\t': + case '\0': + case '[': case ']': case '{': case '}': case ':': + case ',': + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_ERROR, token_pos, parser->pos + 1 + - token_pos, token_line, token_col); + return true; + } + token_init(&parser->token, parser, + TOKEN_TYPE_FALSE, token_pos, parser->pos - + token_pos, token_line, token_col); + state = STATE_ACCEPT; + break; + case STATE_T: + switch (c) { + case 'r': + state = STATE_TR; + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_ERROR, token_pos, parser->pos + 1 + - token_pos, token_line, token_col); + return true; + } + break; + case STATE_TR: + switch (c) { + case 'u': + state = STATE_TRU; + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_ERROR, token_pos, parser->pos + 1 + - token_pos, token_line, token_col); + return true; + } + break; + case STATE_TRU: + switch (c) { + case 'e': + state = STATE_TRUE; + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_ERROR, token_pos, parser->pos + 1 + - token_pos, token_line, token_col); + return true; + } + break; + case STATE_TRUE: + switch (c) { + case ' ': case '\b': case '\n': case '\r': case '\t': + case '\0': + case '[': case ']': case '{': case '}': case ':': + case ',': + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_ERROR, token_pos, parser->pos + 1 + - token_pos, token_line, token_col); + return true; + } + token_init(&parser->token, parser, TOKEN_TYPE_TRUE, + token_pos, parser->pos - token_pos, token_line, + token_col); + state = STATE_ACCEPT; + break; + case STATE_LBRACKET: + token_init(&parser->token, parser, TOKEN_TYPE_LBRACKET, + token_pos, parser->pos - token_pos, token_line, + token_col); + state = STATE_ACCEPT; + break; + case STATE_RBRACKET: + token_init(&parser->token, parser, TOKEN_TYPE_RBRACKET, + token_pos, parser->pos - token_pos, token_line, + token_col); + state = STATE_ACCEPT; + break; + case STATE_LBRACE: + token_init(&parser->token, parser, TOKEN_TYPE_LBRACE, + token_pos, parser->pos - token_pos, token_line, + token_col); + state = STATE_ACCEPT; + break; + case STATE_RBRACE: + token_init(&parser->token, parser, TOKEN_TYPE_RBRACE, + token_pos, parser->pos - token_pos, token_line, + token_col); + state = STATE_ACCEPT; + break; + case STATE_COLON: + token_init(&parser->token, parser, TOKEN_TYPE_COLON, + token_pos, parser->pos - token_pos, token_line, + token_col); + state = STATE_ACCEPT; + break; + case STATE_COMMA: + token_init(&parser->token, parser, TOKEN_TYPE_COMMA, + token_pos, parser->pos - token_pos, token_line, + token_col); + state = STATE_ACCEPT; + break; + case STATE_CHARS: + switch (c) { + case '\\': + state = STATE_CHAR_ESCAPE; + break; + case '"': + state = STATE_STRING; + break; + case 0x00: case 0x01: case 0x02: case 0x03: case 0x04: + case 0x05: case 0x06: case 0x07: case 0x08: case 0x09: + case 0x0a: case 0x0b: case 0x0c: case 0x0d: case 0x0e: + case 0x0f: case 0x10: case 0x11: case 0x12: case 0x13: + case 0x14: case 0x15: case 0x16: case 0x17: case 0x18: + case 0x19: case 0x1a: case 0x1b: case 0x1c: case 0x1d: + case 0x1e: case 0x1f: + token_init(&parser->token, parser, + TOKEN_TYPE_ERROR, token_pos, parser->pos + 1 + - token_pos, token_line, token_col); + return true; + default: + break; + } + break; + case STATE_CHAR_ESCAPE: + switch (c) { + case '"': case '\\': case '/': case 'b': case 'n': + case 'r': case 't': + state = STATE_CHARS; + break; + case 'u': + state = STATE_CHAR_U; + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_ERROR, token_pos, parser->pos + 1 + - token_pos, token_line, token_col); + return true; + } + break; + case STATE_CHAR_U: + switch (c) { + case '0': case '1': case '2': case '3': case '4': + case '5': case '6': case '7': case '8': case '9': + case 'a': case 'b': case 'c': case 'd': case 'e': + case 'f': + case 'A': case 'B': case 'C': case 'D': case 'E': + case 'F': + state = STATE_CHAR_UD; + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_ERROR, token_pos, parser->pos + 1 + - token_pos, token_line, token_col); + return true; + } + break; + case STATE_CHAR_UD: + switch (c) { + case '0': case '1': case '2': case '3': case '4': + case '5': case '6': case '7': case '8': case '9': + case 'a': case 'b': case 'c': case 'd': case 'e': + case 'f': + case 'A': case 'B': case 'C': case 'D': case 'E': + case 'F': + state = STATE_CHAR_UDD; + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_ERROR, token_pos, parser->pos + 1 + - token_pos, token_line, token_col); + return true; + } + break; + case STATE_CHAR_UDD: + switch (c) { + case '0': case '1': case '2': case '3': case '4': + case '5': case '6': case '7': case '8': case '9': + case 'a': case 'b': case 'c': case 'd': case 'e': + case 'f': + case 'A': case 'B': case 'C': case 'D': case 'E': + case 'F': + state = STATE_CHAR_UDDD; + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_ERROR, token_pos, parser->pos + 1 + - token_pos, token_line, token_col); + return true; + } + break; + case STATE_CHAR_UDDD: + switch (c) { + case '0': case '1': case '2': case '3': case '4': + case '5': case '6': case '7': case '8': case '9': + case 'a': case 'b': case 'c': case 'd': case 'e': + case 'f': + case 'A': case 'B': case 'C': case 'D': case 'E': + case 'F': + state = STATE_CHARS; + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_ERROR, token_pos, parser->pos + 1 + - token_pos, token_line, token_col); + return true; + } + break; + case STATE_STRING: + token_init(&parser->token, parser, TOKEN_TYPE_STRING, + token_pos, parser->pos - token_pos, token_line, + token_col); + state = STATE_ACCEPT; + break; + case STATE_MINUS: + switch (c) { + case '0': + state = STATE_LEADING_ZERO; + break; + case '1': case '2': case '3': case '4': + case '5': case '6': case '7': case '8': case '9': + state = STATE_DIGITS; + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_ERROR, token_pos, parser->pos + 1 + - token_pos, token_line, token_col); + return true; + } + break; + case STATE_LEADING_ZERO: + switch (c) { + case '.': + state = STATE_DECIMAL; + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_NUMBER, token_pos, parser->pos - + token_pos, token_line, token_col); + state = STATE_ACCEPT; + break; + } + break; + case STATE_DIGITS: + switch (c) { + case '0': case '1': case '2': case '3': case '4': + case '5': case '6': case '7': case '8': case '9': + break; + case '.': + state = STATE_DECIMAL; + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_NUMBER, token_pos, parser->pos - + token_pos, token_line, token_col); + state = STATE_ACCEPT; + break; + } + break; + case STATE_DECIMAL: + switch (c) { + case '0': case '1': case '2': case '3': case '4': + case '5': case '6': case '7': case '8': case '9': + state = STATE_FRAC_DIGITS; + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_ERROR, token_pos, parser->pos + 1 + - token_pos, token_line, token_col); + return true; + } + break; + case STATE_FRAC_DIGITS: + switch (c) { + case '0': case '1': case '2': case '3': case '4': + case '5': case '6': case '7': case '8': case '9': + break; + case 'e': case 'E': + state = STATE_EXP; + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_NUMBER, token_pos, parser->pos - + token_pos, token_line, token_col); + state = STATE_ACCEPT; + break; + } + break; + case STATE_EXP: + switch (c) { + case '-': case '+': + state = STATE_EXP_SIGN; + break; + case '0': case '1': case '2': case '3': case '4': + case '5': case '6': case '7': case '8': case '9': + state = STATE_EXP_DIGITS; + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_ERROR, token_pos, parser->pos + 1 + - token_pos, token_line, token_col); + return true; + } + break; + case STATE_EXP_SIGN: + switch (c) { + case '0': case '1': case '2': case '3': case '4': + case '5': case '6': case '7': case '8': case '9': + state = STATE_EXP_DIGITS; + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_ERROR, token_pos, parser->pos + 1 + - token_pos, token_line, token_col); + return true; + } + break; + case STATE_EXP_DIGITS: + switch (c) { + case '0': case '1': case '2': case '3': case '4': + case '5': case '6': case '7': case '8': case '9': + break; + default: + token_init(&parser->token, parser, + TOKEN_TYPE_NUMBER, token_pos, parser->pos - + token_pos, token_line, token_col); + state = STATE_ACCEPT; + break; + } + break; + default: + not_reached(); + } + + if (state != STATE_ACCEPT) { + if (c == '\n') { + parser->line++; + parser->col = 0; + } else { + parser->col++; + } + parser->pos++; + } + } + return false; +} + +static bool parser_parse_array(parser_t *parser); +static bool parser_parse_object(parser_t *parser); + +static bool +parser_parse_value(parser_t *parser) { + switch (parser->token.token_type) { + case TOKEN_TYPE_NULL: + case TOKEN_TYPE_FALSE: + case TOKEN_TYPE_TRUE: + case TOKEN_TYPE_STRING: + case TOKEN_TYPE_NUMBER: + return false; + case TOKEN_TYPE_LBRACE: + return parser_parse_object(parser); + case TOKEN_TYPE_LBRACKET: + return parser_parse_array(parser); + default: + return true; + } + not_reached(); +} + +static bool +parser_parse_pair(parser_t *parser) { + assert_d_eq(parser->token.token_type, TOKEN_TYPE_STRING, + "Pair should start with string"); + if (parser_tokenize(parser)) { + return true; + } + switch (parser->token.token_type) { + case TOKEN_TYPE_COLON: + if (parser_tokenize(parser)) { + return true; + } + return parser_parse_value(parser); + default: + return true; + } +} + +static bool +parser_parse_values(parser_t *parser) { + if (parser_parse_value(parser)) { + return true; + } + + while (true) { + if (parser_tokenize(parser)) { + return true; + } + switch (parser->token.token_type) { + case TOKEN_TYPE_COMMA: + if (parser_tokenize(parser)) { + return true; + } + if (parser_parse_value(parser)) { + return true; + } + break; + case TOKEN_TYPE_RBRACKET: + return false; + default: + return true; + } + } +} + +static bool +parser_parse_array(parser_t *parser) { + assert_d_eq(parser->token.token_type, TOKEN_TYPE_LBRACKET, + "Array should start with ["); + if (parser_tokenize(parser)) { + return true; + } + switch (parser->token.token_type) { + case TOKEN_TYPE_RBRACKET: + return false; + default: + return parser_parse_values(parser); + } + not_reached(); +} + +static bool +parser_parse_pairs(parser_t *parser) { + assert_d_eq(parser->token.token_type, TOKEN_TYPE_STRING, + "Object should start with string"); + if (parser_parse_pair(parser)) { + return true; + } + + while (true) { + if (parser_tokenize(parser)) { + return true; + } + switch (parser->token.token_type) { + case TOKEN_TYPE_COMMA: + if (parser_tokenize(parser)) { + return true; + } + switch (parser->token.token_type) { + case TOKEN_TYPE_STRING: + if (parser_parse_pair(parser)) { + return true; + } + break; + default: + return true; + } + break; + case TOKEN_TYPE_RBRACE: + return false; + default: + return true; + } + } +} + +static bool +parser_parse_object(parser_t *parser) { + assert_d_eq(parser->token.token_type, TOKEN_TYPE_LBRACE, + "Object should start with {"); + if (parser_tokenize(parser)) { + return true; + } + switch (parser->token.token_type) { + case TOKEN_TYPE_STRING: + return parser_parse_pairs(parser); + case TOKEN_TYPE_RBRACE: + return false; + default: + return true; + } + not_reached(); +} + +static bool +parser_parse(parser_t *parser) { + if (parser_tokenize(parser)) { + goto label_error; + } + if (parser_parse_value(parser)) { + goto label_error; + } + + if (parser_tokenize(parser)) { + goto label_error; + } + switch (parser->token.token_type) { + case TOKEN_TYPE_EOI: + return false; + default: + goto label_error; + } + not_reached(); + +label_error: + token_error(&parser->token); + return true; +} + +TEST_BEGIN(test_json_parser) { + size_t i; + const char *invalid_inputs[] = { + /* Tokenizer error case tests. */ + "{ \"string\": X }", + "{ \"string\": nXll }", + "{ \"string\": nuXl }", + "{ \"string\": nulX }", + "{ \"string\": nullX }", + "{ \"string\": fXlse }", + "{ \"string\": faXse }", + "{ \"string\": falXe }", + "{ \"string\": falsX }", + "{ \"string\": falseX }", + "{ \"string\": tXue }", + "{ \"string\": trXe }", + "{ \"string\": truX }", + "{ \"string\": trueX }", + "{ \"string\": \"\n\" }", + "{ \"string\": \"\\z\" }", + "{ \"string\": \"\\uX000\" }", + "{ \"string\": \"\\u0X00\" }", + "{ \"string\": \"\\u00X0\" }", + "{ \"string\": \"\\u000X\" }", + "{ \"string\": -X }", + "{ \"string\": 0.X }", + "{ \"string\": 0.0eX }", + "{ \"string\": 0.0e+X }", + + /* Parser error test cases. */ + "{\"string\": }", + "{\"string\" }", + "{\"string\": [ 0 }", + "{\"string\": {\"a\":0, 1 } }", + "{\"string\": {\"a\":0: } }", + "{", + "{}{", + }; + const char *valid_inputs[] = { + /* Token tests. */ + "null", + "false", + "true", + "{}", + "{\"a\": 0}", + "[]", + "[0, 1]", + "0", + "1", + "10", + "-10", + "10.23", + "10.23e4", + "10.23e-4", + "10.23e+4", + "10.23E4", + "10.23E-4", + "10.23E+4", + "-10.23", + "-10.23e4", + "-10.23e-4", + "-10.23e+4", + "-10.23E4", + "-10.23E-4", + "-10.23E+4", + "\"value\"", + "\" \\\" \\/ \\b \\n \\r \\t \\u0abc \\u1DEF \"", + + /* Parser test with various nesting. */ + "{\"a\":null, \"b\":[1,[{\"c\":2},3]], \"d\":{\"e\":true}}", + }; + + for (i = 0; i < sizeof(invalid_inputs)/sizeof(const char *); i++) { + const char *input = invalid_inputs[i]; + parser_t parser; + parser_init(&parser, false); + assert_false(parser_append(&parser, input), + "Unexpected input appending failure"); + assert_true(parser_parse(&parser), + "Unexpected parse success for input: %s", input); + parser_fini(&parser); + } + + for (i = 0; i < sizeof(valid_inputs)/sizeof(const char *); i++) { + const char *input = valid_inputs[i]; + parser_t parser; + parser_init(&parser, true); + assert_false(parser_append(&parser, input), + "Unexpected input appending failure"); + assert_false(parser_parse(&parser), + "Unexpected parse error for input: %s", input); + parser_fini(&parser); + } +} +TEST_END + +void +write_cb(void *opaque, const char *str) { + parser_t *parser = (parser_t *)opaque; + if (parser_append(parser, str)) { + test_fail("Unexpected input appending failure"); + } +} + +TEST_BEGIN(test_stats_print_json) { + const char *opts[] = { + "J", + "Jg", + "Jm", + "Jd", + "Jmd", + "Jgd", + "Jgm", + "Jgmd", + "Ja", + "Jb", + "Jl", + "Jx", + "Jbl", + "Jal", + "Jab", + "Jabl", + "Jax", + "Jbx", + "Jlx", + "Jablx", + "Jgmdablx", + }; + unsigned arena_ind, i; + + for (i = 0; i < 3; i++) { + unsigned j; + + switch (i) { + case 0: + break; + case 1: { + size_t sz = sizeof(arena_ind); + assert_d_eq(mallctl("arenas.create", (void *)&arena_ind, + &sz, NULL, 0), 0, "Unexpected mallctl failure"); + break; + } case 2: { + size_t mib[3]; + size_t miblen = sizeof(mib)/sizeof(size_t); + assert_d_eq(mallctlnametomib("arena.0.destroy", + mib, &miblen), 0, + "Unexpected mallctlnametomib failure"); + mib[1] = arena_ind; + assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, NULL, + 0), 0, "Unexpected mallctlbymib failure"); + break; + } default: + not_reached(); + } + + for (j = 0; j < sizeof(opts)/sizeof(const char *); j++) { + parser_t parser; + + parser_init(&parser, true); + malloc_stats_print(write_cb, (void *)&parser, opts[j]); + assert_false(parser_parse(&parser), + "Unexpected parse error, opts=\"%s\"", opts[j]); + parser_fini(&parser); + } + } +} +TEST_END + +int +main(void) { + return test( + test_json_parser, + test_stats_print_json); +} diff --git a/deps/jemalloc/test/unit/test_hooks.c b/deps/jemalloc/test/unit/test_hooks.c new file mode 100644 index 000000000..ded8698bc --- /dev/null +++ b/deps/jemalloc/test/unit/test_hooks.c @@ -0,0 +1,38 @@ +#include "test/jemalloc_test.h" + +static bool hook_called = false; + +static void +hook() { + hook_called = true; +} + +static int +func_to_hook(int arg1, int arg2) { + return arg1 + arg2; +} + +#define func_to_hook JEMALLOC_HOOK(func_to_hook, test_hooks_libc_hook) + +TEST_BEGIN(unhooked_call) { + test_hooks_libc_hook = NULL; + hook_called = false; + assert_d_eq(3, func_to_hook(1, 2), "Hooking changed return value."); + assert_false(hook_called, "Nulling out hook didn't take."); +} +TEST_END + +TEST_BEGIN(hooked_call) { + test_hooks_libc_hook = &hook; + hook_called = false; + assert_d_eq(3, func_to_hook(1, 2), "Hooking changed return value."); + assert_true(hook_called, "Hook should have executed."); +} +TEST_END + +int +main(void) { + return test( + unhooked_call, + hooked_call); +} diff --git a/deps/jemalloc/test/unit/ticker.c b/deps/jemalloc/test/unit/ticker.c new file mode 100644 index 000000000..e5790a316 --- /dev/null +++ b/deps/jemalloc/test/unit/ticker.c @@ -0,0 +1,73 @@ +#include "test/jemalloc_test.h" + +#include "jemalloc/internal/ticker.h" + +TEST_BEGIN(test_ticker_tick) { +#define NREPS 2 +#define NTICKS 3 + ticker_t ticker; + int32_t i, j; + + ticker_init(&ticker, NTICKS); + for (i = 0; i < NREPS; i++) { + for (j = 0; j < NTICKS; j++) { + assert_u_eq(ticker_read(&ticker), NTICKS - j, + "Unexpected ticker value (i=%d, j=%d)", i, j); + assert_false(ticker_tick(&ticker), + "Unexpected ticker fire (i=%d, j=%d)", i, j); + } + assert_u32_eq(ticker_read(&ticker), 0, + "Expected ticker depletion"); + assert_true(ticker_tick(&ticker), + "Expected ticker fire (i=%d)", i); + assert_u32_eq(ticker_read(&ticker), NTICKS, + "Expected ticker reset"); + } +#undef NTICKS +} +TEST_END + +TEST_BEGIN(test_ticker_ticks) { +#define NTICKS 3 + ticker_t ticker; + + ticker_init(&ticker, NTICKS); + + assert_u_eq(ticker_read(&ticker), NTICKS, "Unexpected ticker value"); + assert_false(ticker_ticks(&ticker, NTICKS), "Unexpected ticker fire"); + assert_u_eq(ticker_read(&ticker), 0, "Unexpected ticker value"); + assert_true(ticker_ticks(&ticker, NTICKS), "Expected ticker fire"); + assert_u_eq(ticker_read(&ticker), NTICKS, "Unexpected ticker value"); + + assert_true(ticker_ticks(&ticker, NTICKS + 1), "Expected ticker fire"); + assert_u_eq(ticker_read(&ticker), NTICKS, "Unexpected ticker value"); +#undef NTICKS +} +TEST_END + +TEST_BEGIN(test_ticker_copy) { +#define NTICKS 3 + ticker_t ta, tb; + + ticker_init(&ta, NTICKS); + ticker_copy(&tb, &ta); + assert_u_eq(ticker_read(&tb), NTICKS, "Unexpected ticker value"); + assert_true(ticker_ticks(&tb, NTICKS + 1), "Expected ticker fire"); + assert_u_eq(ticker_read(&tb), NTICKS, "Unexpected ticker value"); + + ticker_tick(&ta); + ticker_copy(&tb, &ta); + assert_u_eq(ticker_read(&tb), NTICKS - 1, "Unexpected ticker value"); + assert_true(ticker_ticks(&tb, NTICKS), "Expected ticker fire"); + assert_u_eq(ticker_read(&tb), NTICKS, "Unexpected ticker value"); +#undef NTICKS +} +TEST_END + +int +main(void) { + return test( + test_ticker_tick, + test_ticker_ticks, + test_ticker_copy); +} diff --git a/deps/jemalloc/test/unit/tsd.c b/deps/jemalloc/test/unit/tsd.c new file mode 100644 index 000000000..917884dcf --- /dev/null +++ b/deps/jemalloc/test/unit/tsd.c @@ -0,0 +1,267 @@ +#include "test/jemalloc_test.h" + +/* + * If we're e.g. in debug mode, we *never* enter the fast path, and so shouldn't + * be asserting that we're on one. + */ +static bool originally_fast; +static int data_cleanup_count; + +void +data_cleanup(int *data) { + if (data_cleanup_count == 0) { + assert_x_eq(*data, MALLOC_TSD_TEST_DATA_INIT, + "Argument passed into cleanup function should match tsd " + "value"); + } + ++data_cleanup_count; + + /* + * Allocate during cleanup for two rounds, in order to assure that + * jemalloc's internal tsd reinitialization happens. + */ + bool reincarnate = false; + switch (*data) { + case MALLOC_TSD_TEST_DATA_INIT: + *data = 1; + reincarnate = true; + break; + case 1: + *data = 2; + reincarnate = true; + break; + case 2: + return; + default: + not_reached(); + } + + if (reincarnate) { + void *p = mallocx(1, 0); + assert_ptr_not_null(p, "Unexpeced mallocx() failure"); + dallocx(p, 0); + } +} + +static void * +thd_start(void *arg) { + int d = (int)(uintptr_t)arg; + void *p; + + tsd_t *tsd = tsd_fetch(); + assert_x_eq(tsd_test_data_get(tsd), MALLOC_TSD_TEST_DATA_INIT, + "Initial tsd get should return initialization value"); + + p = malloc(1); + assert_ptr_not_null(p, "Unexpected malloc() failure"); + + tsd_test_data_set(tsd, d); + assert_x_eq(tsd_test_data_get(tsd), d, + "After tsd set, tsd get should return value that was set"); + + d = 0; + assert_x_eq(tsd_test_data_get(tsd), (int)(uintptr_t)arg, + "Resetting local data should have no effect on tsd"); + + tsd_test_callback_set(tsd, &data_cleanup); + + free(p); + return NULL; +} + +TEST_BEGIN(test_tsd_main_thread) { + thd_start((void *)(uintptr_t)0xa5f3e329); +} +TEST_END + +TEST_BEGIN(test_tsd_sub_thread) { + thd_t thd; + + data_cleanup_count = 0; + thd_create(&thd, thd_start, (void *)MALLOC_TSD_TEST_DATA_INIT); + thd_join(thd, NULL); + /* + * We reincarnate twice in the data cleanup, so it should execute at + * least 3 times. + */ + assert_x_ge(data_cleanup_count, 3, + "Cleanup function should have executed multiple times."); +} +TEST_END + +static void * +thd_start_reincarnated(void *arg) { + tsd_t *tsd = tsd_fetch(); + assert(tsd); + + void *p = malloc(1); + assert_ptr_not_null(p, "Unexpected malloc() failure"); + + /* Manually trigger reincarnation. */ + assert_ptr_not_null(tsd_arena_get(tsd), + "Should have tsd arena set."); + tsd_cleanup((void *)tsd); + assert_ptr_null(*tsd_arenap_get_unsafe(tsd), + "TSD arena should have been cleared."); + assert_u_eq(tsd_state_get(tsd), tsd_state_purgatory, + "TSD state should be purgatory\n"); + + free(p); + assert_u_eq(tsd_state_get(tsd), tsd_state_reincarnated, + "TSD state should be reincarnated\n"); + p = mallocx(1, MALLOCX_TCACHE_NONE); + assert_ptr_not_null(p, "Unexpected malloc() failure"); + assert_ptr_null(*tsd_arenap_get_unsafe(tsd), + "Should not have tsd arena set after reincarnation."); + + free(p); + tsd_cleanup((void *)tsd); + assert_ptr_null(*tsd_arenap_get_unsafe(tsd), + "TSD arena should have been cleared after 2nd cleanup."); + + return NULL; +} + +TEST_BEGIN(test_tsd_reincarnation) { + thd_t thd; + thd_create(&thd, thd_start_reincarnated, NULL); + thd_join(thd, NULL); +} +TEST_END + +typedef struct { + atomic_u32_t phase; + atomic_b_t error; +} global_slow_data_t; + +static void * +thd_start_global_slow(void *arg) { + /* PHASE 0 */ + global_slow_data_t *data = (global_slow_data_t *)arg; + free(mallocx(1, 0)); + + tsd_t *tsd = tsd_fetch(); + /* + * No global slowness has happened yet; there was an error if we were + * originally fast but aren't now. + */ + atomic_store_b(&data->error, originally_fast && !tsd_fast(tsd), + ATOMIC_SEQ_CST); + atomic_store_u32(&data->phase, 1, ATOMIC_SEQ_CST); + + /* PHASE 2 */ + while (atomic_load_u32(&data->phase, ATOMIC_SEQ_CST) != 2) { + } + free(mallocx(1, 0)); + atomic_store_b(&data->error, tsd_fast(tsd), ATOMIC_SEQ_CST); + atomic_store_u32(&data->phase, 3, ATOMIC_SEQ_CST); + + /* PHASE 4 */ + while (atomic_load_u32(&data->phase, ATOMIC_SEQ_CST) != 4) { + } + free(mallocx(1, 0)); + atomic_store_b(&data->error, tsd_fast(tsd), ATOMIC_SEQ_CST); + atomic_store_u32(&data->phase, 5, ATOMIC_SEQ_CST); + + /* PHASE 6 */ + while (atomic_load_u32(&data->phase, ATOMIC_SEQ_CST) != 6) { + } + free(mallocx(1, 0)); + /* Only one decrement so far. */ + atomic_store_b(&data->error, tsd_fast(tsd), ATOMIC_SEQ_CST); + atomic_store_u32(&data->phase, 7, ATOMIC_SEQ_CST); + + /* PHASE 8 */ + while (atomic_load_u32(&data->phase, ATOMIC_SEQ_CST) != 8) { + } + free(mallocx(1, 0)); + /* + * Both decrements happened; we should be fast again (if we ever + * were) + */ + atomic_store_b(&data->error, originally_fast && !tsd_fast(tsd), + ATOMIC_SEQ_CST); + atomic_store_u32(&data->phase, 9, ATOMIC_SEQ_CST); + + return NULL; +} + +TEST_BEGIN(test_tsd_global_slow) { + global_slow_data_t data = {ATOMIC_INIT(0), ATOMIC_INIT(false)}; + /* + * Note that the "mallocx" here (vs. malloc) is important, since the + * compiler is allowed to optimize away free(malloc(1)) but not + * free(mallocx(1)). + */ + free(mallocx(1, 0)); + tsd_t *tsd = tsd_fetch(); + originally_fast = tsd_fast(tsd); + + thd_t thd; + thd_create(&thd, thd_start_global_slow, (void *)&data.phase); + /* PHASE 1 */ + while (atomic_load_u32(&data.phase, ATOMIC_SEQ_CST) != 1) { + /* + * We don't have a portable condvar/semaphore mechanism. + * Spin-wait. + */ + } + assert_false(atomic_load_b(&data.error, ATOMIC_SEQ_CST), ""); + tsd_global_slow_inc(tsd_tsdn(tsd)); + free(mallocx(1, 0)); + assert_false(tsd_fast(tsd), ""); + atomic_store_u32(&data.phase, 2, ATOMIC_SEQ_CST); + + /* PHASE 3 */ + while (atomic_load_u32(&data.phase, ATOMIC_SEQ_CST) != 3) { + } + assert_false(atomic_load_b(&data.error, ATOMIC_SEQ_CST), ""); + /* Increase again, so that we can test multiple fast/slow changes. */ + tsd_global_slow_inc(tsd_tsdn(tsd)); + atomic_store_u32(&data.phase, 4, ATOMIC_SEQ_CST); + free(mallocx(1, 0)); + assert_false(tsd_fast(tsd), ""); + + /* PHASE 5 */ + while (atomic_load_u32(&data.phase, ATOMIC_SEQ_CST) != 5) { + } + assert_false(atomic_load_b(&data.error, ATOMIC_SEQ_CST), ""); + tsd_global_slow_dec(tsd_tsdn(tsd)); + atomic_store_u32(&data.phase, 6, ATOMIC_SEQ_CST); + /* We only decreased once; things should still be slow. */ + free(mallocx(1, 0)); + assert_false(tsd_fast(tsd), ""); + + /* PHASE 7 */ + while (atomic_load_u32(&data.phase, ATOMIC_SEQ_CST) != 7) { + } + assert_false(atomic_load_b(&data.error, ATOMIC_SEQ_CST), ""); + tsd_global_slow_dec(tsd_tsdn(tsd)); + atomic_store_u32(&data.phase, 8, ATOMIC_SEQ_CST); + /* We incremented and then decremented twice; we should be fast now. */ + free(mallocx(1, 0)); + assert_true(!originally_fast || tsd_fast(tsd), ""); + + /* PHASE 9 */ + while (atomic_load_u32(&data.phase, ATOMIC_SEQ_CST) != 9) { + } + assert_false(atomic_load_b(&data.error, ATOMIC_SEQ_CST), ""); + + thd_join(thd, NULL); +} +TEST_END + +int +main(void) { + /* Ensure tsd bootstrapped. */ + if (nallocx(1, 0) == 0) { + malloc_printf("Initialization error"); + return test_status_fail; + } + + return test_no_reentrancy( + test_tsd_main_thread, + test_tsd_sub_thread, + test_tsd_reincarnation, + test_tsd_global_slow); +} diff --git a/deps/jemalloc/test/unit/witness.c b/deps/jemalloc/test/unit/witness.c new file mode 100644 index 000000000..5986da400 --- /dev/null +++ b/deps/jemalloc/test/unit/witness.c @@ -0,0 +1,280 @@ +#include "test/jemalloc_test.h" + +static witness_lock_error_t *witness_lock_error_orig; +static witness_owner_error_t *witness_owner_error_orig; +static witness_not_owner_error_t *witness_not_owner_error_orig; +static witness_depth_error_t *witness_depth_error_orig; + +static bool saw_lock_error; +static bool saw_owner_error; +static bool saw_not_owner_error; +static bool saw_depth_error; + +static void +witness_lock_error_intercept(const witness_list_t *witnesses, + const witness_t *witness) { + saw_lock_error = true; +} + +static void +witness_owner_error_intercept(const witness_t *witness) { + saw_owner_error = true; +} + +static void +witness_not_owner_error_intercept(const witness_t *witness) { + saw_not_owner_error = true; +} + +static void +witness_depth_error_intercept(const witness_list_t *witnesses, + witness_rank_t rank_inclusive, unsigned depth) { + saw_depth_error = true; +} + +static int +witness_comp(const witness_t *a, void *oa, const witness_t *b, void *ob) { + assert_u_eq(a->rank, b->rank, "Witnesses should have equal rank"); + + assert(oa == (void *)a); + assert(ob == (void *)b); + + return strcmp(a->name, b->name); +} + +static int +witness_comp_reverse(const witness_t *a, void *oa, const witness_t *b, + void *ob) { + assert_u_eq(a->rank, b->rank, "Witnesses should have equal rank"); + + assert(oa == (void *)a); + assert(ob == (void *)b); + + return -strcmp(a->name, b->name); +} + +TEST_BEGIN(test_witness) { + witness_t a, b; + witness_tsdn_t witness_tsdn = { WITNESS_TSD_INITIALIZER }; + + test_skip_if(!config_debug); + + witness_assert_lockless(&witness_tsdn); + witness_assert_depth(&witness_tsdn, 0); + witness_assert_depth_to_rank(&witness_tsdn, (witness_rank_t)1U, 0); + + witness_init(&a, "a", 1, NULL, NULL); + witness_assert_not_owner(&witness_tsdn, &a); + witness_lock(&witness_tsdn, &a); + witness_assert_owner(&witness_tsdn, &a); + witness_assert_depth(&witness_tsdn, 1); + witness_assert_depth_to_rank(&witness_tsdn, (witness_rank_t)1U, 1); + witness_assert_depth_to_rank(&witness_tsdn, (witness_rank_t)2U, 0); + + witness_init(&b, "b", 2, NULL, NULL); + witness_assert_not_owner(&witness_tsdn, &b); + witness_lock(&witness_tsdn, &b); + witness_assert_owner(&witness_tsdn, &b); + witness_assert_depth(&witness_tsdn, 2); + witness_assert_depth_to_rank(&witness_tsdn, (witness_rank_t)1U, 2); + witness_assert_depth_to_rank(&witness_tsdn, (witness_rank_t)2U, 1); + witness_assert_depth_to_rank(&witness_tsdn, (witness_rank_t)3U, 0); + + witness_unlock(&witness_tsdn, &a); + witness_assert_depth(&witness_tsdn, 1); + witness_assert_depth_to_rank(&witness_tsdn, (witness_rank_t)1U, 1); + witness_assert_depth_to_rank(&witness_tsdn, (witness_rank_t)2U, 1); + witness_assert_depth_to_rank(&witness_tsdn, (witness_rank_t)3U, 0); + witness_unlock(&witness_tsdn, &b); + + witness_assert_lockless(&witness_tsdn); + witness_assert_depth(&witness_tsdn, 0); + witness_assert_depth_to_rank(&witness_tsdn, (witness_rank_t)1U, 0); +} +TEST_END + +TEST_BEGIN(test_witness_comp) { + witness_t a, b, c, d; + witness_tsdn_t witness_tsdn = { WITNESS_TSD_INITIALIZER }; + + test_skip_if(!config_debug); + + witness_assert_lockless(&witness_tsdn); + + witness_init(&a, "a", 1, witness_comp, &a); + witness_assert_not_owner(&witness_tsdn, &a); + witness_lock(&witness_tsdn, &a); + witness_assert_owner(&witness_tsdn, &a); + witness_assert_depth(&witness_tsdn, 1); + + witness_init(&b, "b", 1, witness_comp, &b); + witness_assert_not_owner(&witness_tsdn, &b); + witness_lock(&witness_tsdn, &b); + witness_assert_owner(&witness_tsdn, &b); + witness_assert_depth(&witness_tsdn, 2); + witness_unlock(&witness_tsdn, &b); + witness_assert_depth(&witness_tsdn, 1); + + witness_lock_error_orig = witness_lock_error; + witness_lock_error = witness_lock_error_intercept; + saw_lock_error = false; + + witness_init(&c, "c", 1, witness_comp_reverse, &c); + witness_assert_not_owner(&witness_tsdn, &c); + assert_false(saw_lock_error, "Unexpected witness lock error"); + witness_lock(&witness_tsdn, &c); + assert_true(saw_lock_error, "Expected witness lock error"); + witness_unlock(&witness_tsdn, &c); + witness_assert_depth(&witness_tsdn, 1); + + saw_lock_error = false; + + witness_init(&d, "d", 1, NULL, NULL); + witness_assert_not_owner(&witness_tsdn, &d); + assert_false(saw_lock_error, "Unexpected witness lock error"); + witness_lock(&witness_tsdn, &d); + assert_true(saw_lock_error, "Expected witness lock error"); + witness_unlock(&witness_tsdn, &d); + witness_assert_depth(&witness_tsdn, 1); + + witness_unlock(&witness_tsdn, &a); + + witness_assert_lockless(&witness_tsdn); + + witness_lock_error = witness_lock_error_orig; +} +TEST_END + +TEST_BEGIN(test_witness_reversal) { + witness_t a, b; + witness_tsdn_t witness_tsdn = { WITNESS_TSD_INITIALIZER }; + + test_skip_if(!config_debug); + + witness_lock_error_orig = witness_lock_error; + witness_lock_error = witness_lock_error_intercept; + saw_lock_error = false; + + witness_assert_lockless(&witness_tsdn); + + witness_init(&a, "a", 1, NULL, NULL); + witness_init(&b, "b", 2, NULL, NULL); + + witness_lock(&witness_tsdn, &b); + witness_assert_depth(&witness_tsdn, 1); + assert_false(saw_lock_error, "Unexpected witness lock error"); + witness_lock(&witness_tsdn, &a); + assert_true(saw_lock_error, "Expected witness lock error"); + + witness_unlock(&witness_tsdn, &a); + witness_assert_depth(&witness_tsdn, 1); + witness_unlock(&witness_tsdn, &b); + + witness_assert_lockless(&witness_tsdn); + + witness_lock_error = witness_lock_error_orig; +} +TEST_END + +TEST_BEGIN(test_witness_recursive) { + witness_t a; + witness_tsdn_t witness_tsdn = { WITNESS_TSD_INITIALIZER }; + + test_skip_if(!config_debug); + + witness_not_owner_error_orig = witness_not_owner_error; + witness_not_owner_error = witness_not_owner_error_intercept; + saw_not_owner_error = false; + + witness_lock_error_orig = witness_lock_error; + witness_lock_error = witness_lock_error_intercept; + saw_lock_error = false; + + witness_assert_lockless(&witness_tsdn); + + witness_init(&a, "a", 1, NULL, NULL); + + witness_lock(&witness_tsdn, &a); + assert_false(saw_lock_error, "Unexpected witness lock error"); + assert_false(saw_not_owner_error, "Unexpected witness not owner error"); + witness_lock(&witness_tsdn, &a); + assert_true(saw_lock_error, "Expected witness lock error"); + assert_true(saw_not_owner_error, "Expected witness not owner error"); + + witness_unlock(&witness_tsdn, &a); + + witness_assert_lockless(&witness_tsdn); + + witness_owner_error = witness_owner_error_orig; + witness_lock_error = witness_lock_error_orig; + +} +TEST_END + +TEST_BEGIN(test_witness_unlock_not_owned) { + witness_t a; + witness_tsdn_t witness_tsdn = { WITNESS_TSD_INITIALIZER }; + + test_skip_if(!config_debug); + + witness_owner_error_orig = witness_owner_error; + witness_owner_error = witness_owner_error_intercept; + saw_owner_error = false; + + witness_assert_lockless(&witness_tsdn); + + witness_init(&a, "a", 1, NULL, NULL); + + assert_false(saw_owner_error, "Unexpected owner error"); + witness_unlock(&witness_tsdn, &a); + assert_true(saw_owner_error, "Expected owner error"); + + witness_assert_lockless(&witness_tsdn); + + witness_owner_error = witness_owner_error_orig; +} +TEST_END + +TEST_BEGIN(test_witness_depth) { + witness_t a; + witness_tsdn_t witness_tsdn = { WITNESS_TSD_INITIALIZER }; + + test_skip_if(!config_debug); + + witness_depth_error_orig = witness_depth_error; + witness_depth_error = witness_depth_error_intercept; + saw_depth_error = false; + + witness_assert_lockless(&witness_tsdn); + witness_assert_depth(&witness_tsdn, 0); + + witness_init(&a, "a", 1, NULL, NULL); + + assert_false(saw_depth_error, "Unexpected depth error"); + witness_assert_lockless(&witness_tsdn); + witness_assert_depth(&witness_tsdn, 0); + + witness_lock(&witness_tsdn, &a); + witness_assert_lockless(&witness_tsdn); + witness_assert_depth(&witness_tsdn, 0); + assert_true(saw_depth_error, "Expected depth error"); + + witness_unlock(&witness_tsdn, &a); + + witness_assert_lockless(&witness_tsdn); + witness_assert_depth(&witness_tsdn, 0); + + witness_depth_error = witness_depth_error_orig; +} +TEST_END + +int +main(void) { + return test( + test_witness, + test_witness_comp, + test_witness_reversal, + test_witness_recursive, + test_witness_unlock_not_owned, + test_witness_depth); +} diff --git a/deps/jemalloc/test/unit/zero.c b/deps/jemalloc/test/unit/zero.c new file mode 100644 index 000000000..271fd5cba --- /dev/null +++ b/deps/jemalloc/test/unit/zero.c @@ -0,0 +1,59 @@ +#include "test/jemalloc_test.h" + +static void +test_zero(size_t sz_min, size_t sz_max) { + uint8_t *s; + size_t sz_prev, sz, i; +#define MAGIC ((uint8_t)0x61) + + sz_prev = 0; + s = (uint8_t *)mallocx(sz_min, 0); + assert_ptr_not_null((void *)s, "Unexpected mallocx() failure"); + + for (sz = sallocx(s, 0); sz <= sz_max; + sz_prev = sz, sz = sallocx(s, 0)) { + if (sz_prev > 0) { + assert_u_eq(s[0], MAGIC, + "Previously allocated byte %zu/%zu is corrupted", + ZU(0), sz_prev); + assert_u_eq(s[sz_prev-1], MAGIC, + "Previously allocated byte %zu/%zu is corrupted", + sz_prev-1, sz_prev); + } + + for (i = sz_prev; i < sz; i++) { + assert_u_eq(s[i], 0x0, + "Newly allocated byte %zu/%zu isn't zero-filled", + i, sz); + s[i] = MAGIC; + } + + if (xallocx(s, sz+1, 0, 0) == sz) { + s = (uint8_t *)rallocx(s, sz+1, 0); + assert_ptr_not_null((void *)s, + "Unexpected rallocx() failure"); + } + } + + dallocx(s, 0); +#undef MAGIC +} + +TEST_BEGIN(test_zero_small) { + test_skip_if(!config_fill); + test_zero(1, SC_SMALL_MAXCLASS - 1); +} +TEST_END + +TEST_BEGIN(test_zero_large) { + test_skip_if(!config_fill); + test_zero(SC_SMALL_MAXCLASS + 1, 1U << (SC_LG_LARGE_MINCLASS + 1)); +} +TEST_END + +int +main(void) { + return test( + test_zero_small, + test_zero_large); +} diff --git a/deps/jemalloc/test/unit/zero.sh b/deps/jemalloc/test/unit/zero.sh new file mode 100644 index 000000000..b4540b27e --- /dev/null +++ b/deps/jemalloc/test/unit/zero.sh @@ -0,0 +1,5 @@ +#!/bin/sh + +if [ "x${enable_fill}" = "x1" ] ; then + export MALLOC_CONF="abort:false,junk:false,zero:true" +fi |