76 Commits (1dac2dd6779a7a5b35bd222f74bee1fe8c6e560d)

Author	SHA1	Message	Date
Thomas Munro	101c37cd34	Disable clang 16's -Wcast-function-type-strict. ... Clang 16 is still in development, but seawasp reveals that it has started warning about many of our casts of function pointers (those introduced by commit 1c27d16e, and some older ones). Disable the new warning for now, since otherwise buildfarm animal seawasp fails, and we have no current plans to change our strategy for these callback function types. May be back-patched with other Clang/LLVM 16 changes around release time. Discussion: https://postgr.es/m/CA%2BhUKGJvX%2BL3aMN84ksT-cGy08VHErRNip3nV-WmTx7f6Pqhyw%40mail.gmail.com	3 years ago
Andres Freund	fc7852c6cb	meson: Improve PG_VERSION_STR generation ... Previously the host operating system and 32/64 bit were not included and the build machine's cpu was used, which is potentially wrong for cross builds. Author: Juan José Santamaría Flecha <juanjo.santamaria@gmail.com> Author: Andres Freund <andres@anarazel.de> Discussion: https://postgr.es/m/CAC+AXB16gwYhKCdS+t0pk3U7kKtpVj5L-ynmhK3Gbea330At3w@mail.gmail.com	3 years ago
Andres Freund	d3b111e320	Add option to specify segment size in blocks ... The tests don't have much coverage of segment related code, as we don't create large enough tables. To make it easier to test these paths, add a new option specifying the segment size in blocks. Set the new option to 6 blocks in one of the CI tasks. Smaller numbers currently fail one of the tests, for understandable reasons. While at it, fix some segment size related issues in the meson build. Author: Andres Freund <andres@anarazel.de> Discussion: https://postgr.es/m/20221107171355.c23fzwanfzq2pmgt@awork3.anarazel.de	3 years ago
Andres Freund	3f0e786ccb	meson: Add 'running' test setup, as a replacement for installcheck ... To run all tests that support running against existing server: $ meson test --setup running To run just the main pg_regress tests against existing server: $ meson test --setup running regress-running/regress To ensure the 'running' setup continues to work, test it as part of the freebsd CI task. Discussion: https://postgr.es/m/CAH2-Wz=XDQcmLoo7RR_i6FKQdDmcyb9q5gStnfuuQXrOGhB2sQ@mail.gmail.com	3 years ago
Andres Freund	5bdd0cfb91	meson: Add basic PGXS compatibility ... Generate a Makefile.global that's complete enough for PGXS to work for some extensions. It is likely that this compatibility layer will not suffice for every extension and not all platforms - we can expand it over time. This allows extensions to use a single buildsystem across all the supported postgres versions. Once all supported PG versions support meson, we can remove the compatibility layer. Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Discussion: https://postgr.es/m/20221005200710.luvw5evhwf6clig6@awork3.anarazel.de	3 years ago
Andres Freund	79f7c482f6	meson: Basic cygwin support ... There likely are further issues, but as evidenced by the CI task proposed by Justin in the referenced thread, this suffices to build and run basic tests in cygwin (some fixes for the test infrastructure are needed, but that's independent of the meson aspect). Author: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Andres Freund <andres@anarazel.de> Discussion: https://postgr.es/m/20221021034040.GT16921@telsasoft.com	3 years ago
Andres Freund	f0cd57f852	meson: Fix binary mismatch for MSVC plperl vs gcc built perl libs ... This mirrors ccc59a83cd.	3 years ago
Peter Eisentraut	373679c4a8	meson: Define HAVE_LOCALE_T for msvc ... Meson doesn't see the redefinition of locale_t done in src/include/port/win32_port.h, so it is not defining HAVE_LOCALE_T, HAVE_WCSTOMBS_L nor HAVE_MBSTOWCS_L as the current src/tools/msvc/build.pl script does. Add manual overrides to fix. Author: Author: Juan Jose Santamaria Flecha <juanjo.santamaria@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/CAC%2BAXB1wJEqfKCuVcNpoH%3Dgxd61N%3D7c2fR3Ew6YRPpSfEUA%3DyQ%40mail.gmail.com	3 years ago
Andres Freund	ab72a31f6c	meson: Mark PROVE as not required ... In the meson build the prove binary is currently not even used. It will soon be, for PGXS compatibility, but even then we should build without it around. Author: Justin Pryzby <pryzbyj@telsasoft.com> Discussion: https://postgr.es/m/20221021034040.GT16921@telsasoft.com Discussion: https://postgr.es/m/20221104235412.GE16921@telsasoft.com	3 years ago
Peter Eisentraut	c8e4030d1b	Make finding openssl program a configure or meson option ... Various test suites use the "openssl" program as part of their setup. There isn't a way to override which openssl program is to be used, other than by fiddling with the path, perhaps. This has gotten increasingly problematic because different versions of openssl have different capabilities and do different things by default. This patch checks for an openssl binary in configure and meson setup, with appropriate ways to override it. This is similar to how "lz4" and "zstd" are handled, for example. The meson build system actually already did this, but the result was only used in some places. This is now applied more uniformly. Reviewed-by: Michael Paquier <michael@paquier.xyz> Discussion: https://www.postgresql.org/message-id/flat/dc638b75-a16a-007d-9e1c-d16ed6cf0ad2%40enterprisedb.com	3 years ago
Andres Freund	9a95a510ad	meson: macos: Use -Wl,-undefined,error for modules ... meson defaults to -Wl,-undefined,dynamic_lookup for modules, which we don't want because a) it's different from what we do for autoconf, b) it causes warnings starting in macOS Ventura. Discussion: https://postgr.es/m/20221015211955.q4cwbsfkyk3c4ty3@awork3.anarazel.de	3 years ago
Andres Freund	57bab33302	meson: catch up to a few configure changes ... I (Andres) missed a few recent changes to configure when merging e6927270cd "meson: Add initial version of meson based build system". Mirror the changes in - ec3c9cc202 "Add definition pg_attribute_aligned() for MSVC" - b086a47a27 "Bump minimum version of Bison to 2.3" - 8b878bffa8 "Bump minimum version of Flex to 2.5.35" As MSVC does not implement 128 bit integers, the oversight of not using pg_attribute_aligned() should not have current practical consequences. But of course the code from c.h should still be correctly mirrored. I (Andres) also hadn't implemented the minimum perl version check. Added that now. Reported-by: Junwang Zhao <zhjwpku@gmail.com> Author: Junwang Zhao <zhjwpku@gmail.com> Author: Andres Freund <andres@anarazel.de> Discussion: https://postgr.es/m/CAEG8a3K9c87EwAwmdOgmS0Li1J6P_7r-Uc0-zN6cJtrMr7VvPg@mail.gmail.com	3 years ago
David Rowley	0fe954c285	Add -Wshadow=compatible-local to the standard compilation flags ... Since cd4e8caaa, we've been able to build the source tree with -Wshadow=compatible-local without any warnings. Lots of work was done by Justin Pryzby and I (David) to get all our code to compile warning free with that flag. In that process, 2 bugs (16d69ec29 and af7d270dd) were discovered and fixed. Additionally, "git log --grep=shadow" shows that there is no shortage of other bugs that have been fixed over the years which were caused by variable shadowing. In light of the above, it seems very much worthwhile to add at least -Wshadow=compatible-local to our standard compilation flags. We *may* want to go further and take this to -Wshadow=local in the future, but we're not ready for that today, so let's add -Wshadow=compatible-local now to help make sure we don't introduce further local variable shadowing. Author: Andres Freund Discussion: https://postgr.es/m/20221006003920.6xlqaoccxwisza5k@awork3.anarazel.de	3 years ago
Andres Freund	e5555657ba	meson: Add support for building with precompiled headers ... This substantially speeds up building for windows, due to the vast amount of headers included via windows.h. A cross build from linux targetting mingw goes from 994.11user 136.43system 0:31.58elapsed 3579%CPU to 422.41user 89.05system 0:14.35elapsed 3562%CPU The wins on windows are similar-ish (but I don't have a system at hand just now for actual numbers). Targetting other operating systems the wins are far smaller (tested linux, macOS, FreeBSD). For now precompiled headers are disabled by default, it's not clear how well they work on all platforms. E.g. on FreeBSD gcc doesn't seem to have working support, but clang does. When doing a full build precompiled headers are only beneficial for targets with multiple .c files, as meson builds a separate precompiled header for each target (so that different compilation options take effect). This commit therefore only changes target with at least two .c files to use precompiled headers. Because this commit adds b_pch=false to the default_options new build directories will have precompiled headers disabled by default, however existing build directories will continue use the default value of b_pch, which is true. Note that using precompiled headers with ccache requires setting CCACHE_SLOPPINESS=pch_defines,time_macros to get hits. Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Discussion: https://postgr.es/m/CA+hUKG+50eOUbN++ocDc0Qnp9Pvmou23DSXu=ZA6fepOcftKqA@mail.gmail.com Discussion: https://postgr.es/m/c5736f70-bb6d-8d25-e35c-e3d886e4e905@enterprisedb.com Discussion: https://postgr.es/m/20190826054000.GE7005%40paquier.xyz	3 years ago
Andres Freund	902ab2fcef	meson: Add windows resource files ... The generated resource files aren't exactly the same ones as the old buildsystems generate. Previously "InternalName" and "OriginalFileName" were mostly wrong / not set (despite being required), but that was hard to fix in at least the make build. Additionally, the meson build falls back to a "auto-generated" description when not set, and doesn't set it in a few cases - unlikely that anybody looks at these descriptions in detail. Author: Andres Freund <andres@anarazel.de> Author: Nazir Bilal Yavuz <byavuz81@gmail.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com>	3 years ago
Andres Freund	089c0bc7a7	meson: libpq: Revise static / shared library setup ... Improvements: - we don't need -DFRONTEND for libpq anymore since 1d77afefbd - the .pc file contents for a static libpq were wrong (referencing {pgport, common}_shlib) - incidentally fixes meson not supporting link_whole on AIX yet - added explanatory comments Previously I tried to avoid building libpq's sources twice, once for the static and once for the shared library. We could still do so, but it's not clear that it's worth the complication. Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com>	3 years ago
Andres Freund	0ae5db28d0	meson: respect -Dldap=disabled ... I noticed during some manual testing that -Dldap=disabled (or --auto-features=disabled) doesn't disable ldap if available - that's obviously wrong.	3 years ago
Andres Freund	1360943d1a	meson: Add prefix=/usr/local/pgsql to default_options ... autoconf set PREFIX to /usr/local/pgsql, so using the same default for meson makes sense. The effect on windows is that installation defaults to installing to C:/usr/local/pgsql rather than meson's default of C:/, which doesn't seem perfect, but OK enough. Signed-off-by: Junwang Zhao <zhjwpku@gmail.com> Author: Junwang Zhao <zhjwpku@gmail.com> Discussion: https://postgr.es/CAEG8a3LGWE-gG6vuddmH91RORhi8gWs0mMB-hcTmP3_NVgM7dg@mail.gmail.com	3 years ago
Andres Freund	d0366bfb3b	meson: windows: Determine path to tmp_install + prefix using meson ... Previously some paths (like c:\ or d:/) worked, but plenty others (like /path/to or //computer/share/path) didn't. As we'd like to change the default prefix to /usr/local/pgsql, that's a problem. Instead of trying to do this in meson.build, call out to the implementation meson install uses. This isn't pretty, but it's more reliable than what we had before. Discussion: https://postgr.es/CAEG8a3LGWE-gG6vuddmH91RORhi8gWs0mMB-hcTmP3_NVgM7dg@mail.gmail.com	3 years ago
Andres Freund	24ef127b6d	meson: mingw: Add -Wl,--disable-auto-import, enable when linking with readline ... I hadn't ported -Wl,--disable-auto-import over from the win32 template as I had focused on msvc for windows. The flag is desirable as it makes it easier to find problems one would have with msvc, particularly useful during cross compilation. This turned out to be a somewhat happy accident, as it allowed me to realize that readline actually works on windows these days, as long as auto imports to enable. Therefore enable auto-import again as part of linking to readline. We perhaps can come up with a better solution for the readline issue, but this seems good enough for now. Discussion: http://postgr.es/m/20220928022724.erzuk5v4ai4b53do@awork3.anarazel.de	3 years ago
Andres Freund	a1b3bca1c8	meson: Implement getopt logic from autoconf ... Not replacing getopt/getopt_long definitely causes issues on mingw. It's not as clear whether the solaris & openbsd aspect is still needed, but if not, we should remove it from both autoconf and meson. Discussion: http://postgr.es/m/20220928022724.erzuk5v4ai4b53do@awork3.anarazel.de	3 years ago
Andres Freund	caeeabd110	meson: mingw: Allow multiple definitions ... I didn't carry this forward from the win32 template. It's not needed anymore for the reason stated therein, but it turns out to be required to e.g. override getopt. Possibly a better solution exists, but that's for later. Discussion: http://postgr.es/m/20220928022724.erzuk5v4ai4b53do@awork3.anarazel.de	3 years ago
Andres Freund	e3a892539b	meson: windows: Normalize slashes in prefix ... This fixes a build issue on windows, when the prefix is set to a path with forward slashes. Windows python defaults to a path with a backslash, but mingw ucrt python defaults to a forward slash. This in turn lead to a wrong PATH set during tests, causing tests to fail. Reported-by: Melih Mutlu <m.melihmutlu@gmail.com> Discussion: http://postgr.es/m/20220928022724.erzuk5v4ai4b53do@awork3.anarazel.de	3 years ago
Andres Freund	a1860071b3	meson: Set up absolute rpaths to libdir ... While I (Andres) had planned to use relative rpaths, it looks like agreeing on the precise path might take a bit. So set up absolute rpaths for now. I'm pushing this fairly quickly after posting the patch as several hackers fought with this issue. Discussion: https://postgr.es/m/20220927011951.j3h4o7n6bhf7dwau@awork3.anarazel.de	3 years ago
Andres Freund	1330dcdec0	meson: Include CFLAGS/c_args in summary and pg_config output ... Previously arguments passed in via CFLAGS/-Dc_args were neither displayed in meson's summary, nor in pg_config's output. Reported-by: "wangw.fnst@fujitsu.com" <wangw.fnst@fujitsu.com> Discussion: https://postgr.es/m/OS3PR01MB62751847BC9CD2DB7B29AC129E529@OS3PR01MB6275.jpnprd01.prod.outlook.com	3 years ago
Andres Freund	e6927270cd	meson: Add initial version of meson based build system ... Autoconf is showing its age, fewer and fewer contributors know how to wrangle it. Recursive make has a lot of hard to resolve dependency issues and slow incremental rebuilds. Our home-grown MSVC build system is hard to maintain for developers not using Windows and runs tests serially. While these and other issues could individually be addressed with incremental improvements, together they seem best addressed by moving to a more modern build system. After evaluating different build system choices, we chose to use meson, to a good degree based on the adoption by other open source projects. We decided that it's more realistic to commit a relatively early version of the new build system and mature it in tree. This commit adds an initial version of a meson based build system. It supports building postgres on at least AIX, FreeBSD, Linux, macOS, NetBSD, OpenBSD, Solaris and Windows (however only gcc is supported on aix, solaris). For Windows/MSVC postgres can now be built with ninja (faster, particularly for incremental builds) and msbuild (supporting the visual studio GUI, but building slower). Several aspects (e.g. Windows rc file generation, PGXS compatibility, LLVM bitcode generation, documentation adjustments) are done in subsequent commits requiring further review. Other aspects (e.g. not installing test-only extensions) are not yet addressed. When building on Windows with msbuild, builds are slower when using a visual studio version older than 2019, because those versions do not support MultiToolTask, required by meson for intra-target parallelism. The plan is to remove the MSVC specific build system in src/tools/msvc soon after reaching feature parity. However, we're not planning to remove the autoconf/make build system in the near future. Likely we're going to keep at least the parts required for PGXS to keep working around until all supported versions build with meson. Some initial help for postgres developers is at https://wiki.postgresql.org/wiki/Meson With contributions from Thomas Munro, John Naylor, Stone Tickle and others. Author: Andres Freund <andres@anarazel.de> Author: Nazir Bilal Yavuz <byavuz81@gmail.com> Author: Peter Eisentraut <peter@eisentraut.org> Reviewed-By: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Discussion: https://postgr.es/m/20211012083721.hvixq4pnh2pixr3j@alap3.anarazel.de	3 years ago