Until now postgres built extension libraries with global visibility, i.e.
exporting all symbols. On the one platform where that behavior is not
natively available, namely windows, we emulate it by analyzing the input files
to the shared library and exporting all the symbols therein.
Not exporting all symbols is actually desirable, as it can improve loading
speed, reduces the likelihood of symbol conflicts and can improve intra
extension library function call performance. It also makes the non-windows
builds more similar to windows builds.
Additionally, with meson implementing the export-all-symbols behavior for
windows, turns out to be more verbose than desirable.
This patch adds support for hiding symbols by default and, to counteract that,
explicit symbol visibility annotation for compilers that support
__attribute__((visibility("default"))) and -fvisibility=hidden. That is
expected to be most, if not all, compilers except msvc (for which we already
support explicit symbol export annotations).
Now that extension library symbols are explicitly exported, we don't need to
export all symbols on windows anymore, hence remove that behavior from
src/tools/msvc. The supporting code can't be removed, as we still need to
export all symbols from the main postgres binary.
Author: Andres Freund <andres@anarazel.de>
Author: Tom Lane <tgl@sss.pgh.pa.us>
Discussion: https://postgr.es/m/20211101020311.av6hphdl6xbjbuif@alap3.anarazel.de
Apply in libpq_pipeline test makefile, so that the test file is not installed into tmp_install. Reviewed-by: Alvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://www.postgresql.org/message-id/flat/cb9d16a6-760f-cd44-28d6-b091d5fb6ca7%40enterprisedb.com
When $(MODULES) and $(MODULE_big) are empty, derive the database name from the first element of $(REGRESS) instead of using a constant string. When deriving the database name from $(MODULES), use its first element instead of the entire list; the earlier approach would fail if any multi-module directory had $(REGRESS) tests. Treat isolation suites and src/pl correspondingly. Under USE_MODULE_DB=1, installcheck-world and check-world no longer reuse any database name in a given postmaster. Buildfarm members axolotl, mandrill and frogfish saw spurious "is being accessed by other users" failures that would not have happened without database name reuse. (The CountOtherDBBackends() 5s deadline expired during DROP DATABASE; a backend for an earlier test suite had used the same database name and had not yet exited.) Back-patch to 9.4 (all supported versions), except bits pertaining to isolation suites. Concept reviewed by Andrew Dunstan, Andres Freund and Tom Lane. Discussion: https://postgr.es/m/20190401135213.GE891537@rfd.leadboat.com
Add PG_CFLAGS, PG_CXXFLAGS, and PG_LDFLAGS variables to pgxs.mk which will be appended or prepended to the corresponding make variables. Notably, there was previously no way to pass custom CXXFLAGS to third party extension module builds, COPT and PROFILE supporting only CFLAGS and LDFLAGS. Backpatch all the way down to ease integration with existing extensions. Author: Christoph Berg Reviewed-by: Andres Freund, Tom Lane, Michael Paquier Discussion: https://postgr.es/m/20181113104005.GA32154@msg.credativ.de Backpatch-through: 9.4
This closes a race condition in "make -j check-world"; the symptom was EEXIST errors. Back-patch to v10, before which parallel check-world had worse problems. Discussion: https://postgr.es/m/20181224221601.GA3227827@rfd.leadboat.com
The following options are added for extensions: - TAP_TESTS, to allow an extention to run TAP tests which are the ones present in t/*.pl. A subset of tests can always be run with the existing PROVE_TESTS for developers. - ISOLATION, to define a list of isolation tests. - ISOLATION_OPTS, to pass custom options to isolation_tester. A couple of custom Makefile rules have been accumulated across the tree to cover the lack of facility in PGXS for a couple of releases when using those test suites, which are all now replaced with the new flags, without reducing the test coverage. Note that tests of contrib/bloom/ are not enabled yet, as those are proving unstable in the buildfarm. Author: Michael Paquier Reviewed-by: Adam Berlin, Álvaro Herrera, Tom Lane, Nikolay Shaplov, Arthur Zakirov Discussion: https://postgr.es/m/20180906014849.GG2726@paquier.xyz
A set of failures in buildfarm machines are proving that this is not quite ready yet because of another set of issues: - MSVC scripts assume that REGRESS_OPTS can only use top_builddir. Some test suites actually finish by using top_srcdir, like pg_stat_statements which cause the regression tests to never run. - Trying to enforce top_builddir does not work either when using VPATH as this is not recognized properly. - TAP tests of bloom are unstable on various platforms, causing various failures.
The following options are added for extensions: - TAP_TESTS, to allow an extention to run TAP tests which are the ones present in t/*.pl. A subset of tests can always be run with the existing PROVE_TESTS for developers. - ISOLATION, to define a list of isolation tests. - ISOLATION_OPTS, to pass custom options to isolation_tester. A couple of custom Makefile targets have been accumulated across the tree to cover the lack of facility in PGXS for a couple of releases when using those test suites, which are all now replaced with the new flags, without reducing the test coverage. This also fixes an issue with contrib/bloom/, which had a custom target to trigger its TAP tests of its own not part of the main check runs. Author: Michael Paquier Reviewed-by: Adam Berlin, Álvaro Herrera, Tom Lane, Nikolay Shaplov, Arthur Zakirov Discussion: https://postgr.es/m/20180906014849.GG2726@paquier.xyz
Commit be54b3777 failed on gmake 3.80 due to a chained conditional, which on closer examination could be removed entirely with some refactoring elsewhere for a net simplification and more robustness against empty expansions. Along the way, add some more comments. Also make explicit in the documentation and comments that built headers are not removed by 'make clean', since we don't typically want that for headers generated by a separate ./configure step, and it's much easier to add your own 'distclean' rule or use EXTRA_CLEAN than to try and override a deletion rule in pgxs.mk. Per buildfarm member prariedog and comments by Michael Paquier, though all the actual changes are my fault.
Commit df163230b overlooked the case that an out-of-tree extension might need to build its header files (e.g. via ./configure). If it is also doing a VPATH build, the HEADERS_* rules in the original commit would then fail to find the files, since they would be looking only under $(srcdir) and not in the build directory. Fix by adding HEADERS_built and HEADERS_built_$(MODULE) which behave like DATA_built in that they look in the build dir rather than the source dir (and also make the files dependencies of the "all" target). No Windows support appears to be needed for this, since it is only relevant to out-of-tree builds (no support exists in Mkvcbuild.pm to build extension header files in any case).