summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* announce-gen: VPATH issuesAkim Demaille2012-06-123-4/+16
| | | | | | * build-aux/announce-gen (--srcdir): New option, used to trim the $srcdir part of the path from $builddir to NEWS. * top/maint.mk (announcement): Adjust.
* gnu-web-doc-update: VPATH buildsAkim Demaille2012-06-123-31/+76
| | | | | | | | | | | | | * build-aux/gnu-web-doc-update (--builddir): New option. Revamp the handling of options. Prefer $(...) to `...`. Don't pass --tmpdir=. to mktemp, it is useless given that we specify the template, and it is GNU mktemp specific. Prefer set -e to long series of &&. Restore the initial git branch, not "master". Properly initialize submodules (don't rely only on bootstrap) Do not reconfigure blindly, use config.status. * top/README-release: Update instructions for gnu-web-doc-update.
* autoupdateKarl Berry2012-06-111-2/+2
|
* maint.mk: revert most of the previous change re "all these"Jim Meyering2012-06-112-1/+8
| | | | | | * top/maint.mk (prohibit_undesirable_word_seq_RE_): Remove that pair. For rationale, see the discussion at http://thread.gmane.org/gmane.comp.lib.gnulib.bugs/30912
* autoupdate missed x bitKarl Berry2012-06-101-0/+0
|
* do not ask for gpg pw with --dry-runKarl Berry2012-06-102-5/+9
|
* implement --replace for ftp-upload protocol v1.2Karl Berry2012-06-102-2/+18
|
* parse-duration: Relicense under LGPLv2+.Bruce Korb2012-06-112-1/+6
| | | | * modules/parse-duration (License): Change to LGPLv2+.
* maint.mk: prohibit common grammar error: "all these"Jim Meyering2012-06-103-2/+9
| | | | | | * top/maint.mk (prohibit_undesirable_word_seq_RE_): Add "all these" to the list of prohibited word sequences. It should be "all of these". * lib/tempname.c (__gen_tempname): Fix one of them.
* do-release-commit-and-tag: support VPATH buildsAkim Demaille2012-06-073-37/+78
| | | | | | | | | | | | | | | | * build-aux/do-release-commit-and-tag: Prefer $(...) to `...`. (noteworthy): Defined earlier to factor its value. (noteworthy_stub): New. Use it to factor. (help_version): Split into... (help, version): these. Adjust the option processing part. Support "--option=value" in addition to "--option value". (builddir): New. (--builddir): New option. * top/README-release: Document this. Reword slightly so that the reader cannot understand that he has to do these steps before calling do-release-commit-and-tag.
* readme-release: also require announce-gen and maintainer-makefileAkim Demaille2012-06-076-4/+16
| | | | | | | * modules/readme-release (Depends-on): here. * modules/announce-gen, modules/do-release-commit-and-tag, modules/gnu-web-doc-update, modules/maintainer-makefile (Description): Point to readme-release.
* maint.mk: fix VPATH issues.Akim Demaille2012-06-072-2/+8
| | | | | * top/maint.mk (news-check): GNU Make understand $< very well. (release-prep): NEWS is in $(srcdir).
* readme-release: require the promoted modules.Akim Demaille2012-06-072-0/+10
| | | | | | * modules/readme-release (Depends-on): Add do-release-commit-and-tag, gnupload, and gnu-web-doc-update, used in this text.
* update from texinfoKarl Berry2012-06-051-2/+2
|
* error, strerror-override: Support mingw64 from Fedora 17.Paolo Bonzini2012-06-035-1/+22
| | | | | | | | | * lib/errno.in.h (GNULIB_defined_ESTREAMS): Use a different indicator for ETXTBSY, ENODATA, ENOSR, ENOSTR, ETIME, EOTHER, compared to EINPROGRESS. * lib/strerror-override.h (strerror_override): Test it. * lib/strerror-override.c (strerror_override): Likewise. * m4/errno_h.m4 (gl_HEADER_ERRNO_H): Test also ETXTBSY.
* error, strerror-override: Support mingw64 from Fedora 17.Paolo Bonzini2012-06-034-3/+21
| | | | | | | * lib/errno.in.h (GNULIB_defined_ENOTRECOVERABLE): Use a different indicator for ENOTRECOVERABLE, compared to EOWNERDEAD. * lib/strerror-override.h (strerror_override): Test it. * lib/strerror-override.c (strerror_override): Likewise.
* error, strerror-override: Support new errno values from POSIX:2008.Bruno Haible2012-06-037-8/+74
| | | | | | | | | | | | | | | * m4/errno_h.m4 (gl_HEADER_ERRNO_H): Test also EOWNERDEAD and ENOTRECOVERABLE. * lib/errno.in.h (EOWNERDEAD, ENOTRECOVERABLE): Define on all platforms. * lib/strerror-override.c (strerror_override): Conditionalize the EOWNERDEAD, ENOTRECOVERABLE handling on GNULIB_defined_EOWNERDEAD. * lib/strerror-override.h (strerror_override): Declare also if GNULIB_defined_EOWNERDEAD is defined. * tests/test-errno.c (e130, e131): New variables. * doc/posix-headers/errno.texi: Mention the status for EOWNERDEAD, ENOTRECOVERABLE. Reported by Paolo Bonzini.
* autoupdateKarl Berry2012-06-021-14/+22
|
* savewd: add missing dependency on sys_wait moduleJim Meyering2012-05-312-0/+7
| | | | | * modules/savewd (Depends-on): Add sys_wait, needed at least for MSVC. Report and suggested change by Michael Goffioul.
* system-quote-tests: port to CentOS 5Paul Eggert2012-05-292-0/+12
| | | | | | Problem reported by Tom G. Christensen in <http://lists.gnu.org/archive/html/bug-gnulib/2012-05/msg00255.html>. * tests/test-system-quote-child.c (fclose, fprintf): Undef.
* maint: fix typos in comments and ChangeLogJim Meyering2012-05-294-14/+26
| | | | | | | | | | | Culprits identified and fixed mostly automatically using these commands: git ls-files | misspellings -f - |grep -v '^ERROR:' |perl -pe \ 's/^(.*?)\[(\d+)\]: (\w+) -> "(.*?)"$/sed -i '\''${2}s!$3!$4!'\'' $1/' using http://github.com/lyda/misspell-check * ChangeLog: Fix typos. * doc/solaris-versions: Likewise. * lib/regexec.c (re_search_stub): Likewise. * m4/mmap-anon.m4 (gl_FUNC_MMAP_ANON): Likewise.
* autoupdateKarl Berry2012-05-281-2/+4
|
* manywarnings: remove duplicate -Wmultichar entryPaul Eggert2012-05-272-2/+8
| | | | | | * m4/manywarnings.m4 (gl_MANYWARN_ALL_GCC): Remove duplicate entry for -Wmultichar. -Wno-multichar is in the GCC 4.7.0 manual, so keep the entry marked as documented.
* do not sync mktime.c from libc, last one left, not worth itKarl Berry2012-05-272-1/+8
|
* autoupdateKarl Berry2012-05-271-4/+4
|
* regex: don't assume uint64_t or uint32_tPaul Eggert2012-05-262-6/+19
| | | | | | | | * lib/regcomp.c (init_word_char): Don't assume that the types uint64_t and uint32_t exist. The C standard doesn't guarantee them, and on some 32-bit compilers there is no uint64_t. Problem reported by Gianluigi Tiesi in <http://lists.gnu.org/archive/html/bug-gnulib/2012-03/msg00154.html>.
* * lib/mktime.c: Revert autoupdate, which fetched the wrong version.Paul Eggert2012-05-251-80/+155
|
* maint.mk: add strncpy-prohibiting syntax-check ruleJim Meyering2012-05-252-0/+13
| | | | * top/maint.mk (sc_prohibit_strncpy): New rule, from coreutils.
* autoupdateKarl Berry2012-05-252-155/+83
|
* maint.mk: compute $(gpg_key_ID) more portablyJim Meyering2012-05-242-1/+9
| | | | | | | * top/maint.mk (gpg_key_ID): Use awk in place of sed '{...;...;}'. That use of sed is not portable to some fringe systems. Reported by Paul Eggert in http://thread.gmane.org/gmane.comp.lib.gnulib.bugs/30793/focus=30802
* mktime: sync from glibcPaul Eggert2012-05-233-195/+210
| | | | | | | | | | | * config/srclist.txt: Uncomment mktime.c. * lib/mktime.c: Sync from glibc master. This incorporates 3 changes. First, indent with tabs, since glibc uses tabs and doesn't want to change and we'd rather be identical to glibc. Also, two small coding changes: (isdst_differ): Use &&, not &, as && is the usual style. (__mktime_internal): Rename local var from abs_diff to approx_abs_diff for clarity.
* announce-gen: du -h is more portable than du --humanAkim Demaille2012-05-232-2/+8
| | | | | | * build-aux/announce-gen (sizes): Invoke du with -h instead of --human. Accept leading white space in its output. These changes are required to accommodate OS/X's du.
* announce-gen: improve getopt-emitted diagnosticsAkim Demaille2012-05-232-14/+30
| | | | | | * build-aux/announce-gen: When parsing command line options, prefer "announce-gen: option --release-type requires an argument" to "Option release-type requires an argument".
* maint.mk: gpg_key_ID: use sed more portablyAkim Demaille2012-05-232-2/+9
| | | | | | * top/maint.mk (gpg_key_ID): End sed block with a semicolon before the closing brace. (refresh-po): Fuse two sed invocations into one.
* gitlog-to-changelog: support the log message format used in Bison.Akim Demaille2012-05-222-2/+26
| | | | | * build-aux/gitlog-to-changelog: Support --strip-tab and --strip-cherry-picked.
* Revert "Revert "select: Fix build error on IRIX 6.5.""Simon Josefsson2012-05-222-0/+6
| | | | This reverts commit 4bacd42cee65eb44bd3aabfb01ac682fa9c7a3ce.
* Revert "select: Fix build error on IRIX 6.5."Simon Josefsson2012-05-222-6/+0
| | | | This reverts commit 0b1066b5ab2c8e0e8a8bec718c7eaa8b3a685d89.
* Revert "strdup: undeprecate, IRIX 6.5 needs it"Simon Josefsson2012-05-223-8/+7
| | | | This reverts commit 7e2a535f8c362d5d3905cac9ebd7df4622604992.
* poll/select: prevent busy-waitingPaolo Bonzini2012-05-223-19/+55
| | | | | | | | | | | | | | 2012-05-21 Paolo Bonzini <bonzini@gnu.org> poll/select: prevent busy-waiting. SwitchToThread() only gives away the rest of the current time slice to another thread in the current process. So if the thread that feeds the file decscriptor we're polling is not in the current process, we get busy-waiting. * lib/poll.c: Use SleepEx(1, TRUE) instead of SwitchToThread(). Patch from Theodore Leblond. * lib/select.c: Split polling out of the loop that sets the output fd_sets. Check for zero result and loop if the wait timeout is infinite.
* strdup: undeprecate, IRIX 6.5 needs itEric Blake2012-05-213-7/+8
| | | | | | | | | Reported by Simon Josefsson. * modules/strdup (Status): No longer obsolete. * doc/posix-functions/strdup.texi (strdup): Document this. Signed-off-by: Eric Blake <eblake@redhat.com>
* select: Fix build error on IRIX 6.5.Simon Josefsson2012-05-212-0/+6
|
* autoupdateKarl Berry2012-05-211-14/+50
|
* Fix ChangeLog entry in last commit (silent change).Simon Josefsson2012-05-211-0/+1
|
* gc: fix libgcrypt detection on older machines.Simon Josefsson2012-05-212-3/+17
| | | | | | * m4/gc.m4: Reject libgcrypt earlier than 1.4.4. Collapse copyright years because the file has been distributed every year since it was created.
* autoupdateKarl Berry2012-05-201-12/+19
|
* crypto: fix bug in large buffer handlingPaul Eggert2012-05-182-5/+7
| | | | | * lib/sha512.c (sha512_process_block): Work even if size_t is wider than 64 bits.
* crypto: fix bug in large buffer handlingPaul Eggert2012-05-183-2/+18
| | | | | | | | * lib/sha512.c (sha512_process_block): Don't assume the buffer length is less than 2**32. Here, the bug is present only in the rare case where the host does not support uint64_t; use u64size to work around the problem. * lib/u64.h (u64size): New macro.
* crypto: fix bug in large buffer handlingPaul Eggert2012-05-185-12/+24
| | | | | | | | | | | Problem reported by Serge Belyshev for glibc in <http://sourceware.org/bugzilla/show_bug.cgi?id=14090> and for gnulib in <http://lists.gnu.org/archive/html/bug-gnulib/2012-05/msg00226.html>. * lib/md4.c (md4_process_block): * lib/md5.c (md5_process_block): * lib/sha1.c (sha1_process_block): * lib/sha256.c (sha256_process_block): Don't assume the buffer length is less than 2**32.
* autoupdateKarl Berry2012-05-181-5/+5
|
* update from texinfoKarl Berry2012-05-161-20/+22
|