Commit graph

246 commits

Author SHA1 Message Date
Tor Didriksen
f325fc42e8 Bug#14300733 CMAKE DOES NOT CHECK FOR ZLIB VERSION
Add check for compressBound() and deflateBound()
Keep existing functionality: fallback to bundled zlib if functions not found.
2013-02-20 11:24:16 +01:00
Joerg Bruehe
7e613db478 Placement change:
Top level "SPECIFIC-ULN/" was inappropriate,
put the files to create RPMs for ULN into
"packaging/rpm-uln/".
2012-11-07 17:45:02 +01:00
Tor Didriksen
4baab59e93 Bug#14737559 BZR JOIN PLUGIN TREES INTO INTERNAL/PLUGIN
Part three:
Fix some search paths.
2012-10-24 17:06:43 +02:00
Tor Didriksen
7873a6e4d7 Use MESSAGE(STATUS ...) as we don't want any popup from the build system. 2012-10-23 13:41:34 +02:00
Joerg Bruehe
4828c75af1 Fix formatting in 'INFO_BIN' on Windows - backport. 2012-10-19 20:31:22 +02:00
Tor Didriksen
30d35590a3 Bug#13713525 CREATE_INITIAL_DB.CMAKE IS FAILING ON WINDOWS, STILL "DEVENV" RETURNS 0
This bug depends on cmake version.

For cmake 2.6 (which is still in use for some pushbuild trees)
the main build would succeed, even if create_initial_db failed.

The problem was the chaining of commands in the CUSTOM_COMMAND
to produce 'initdb.dep'. It first invokes cmake to run mysqld,
then invokes 'touch' to create the file. Moving the 'touch'
command makes the error propagate properly for both cmake 2.6 and 2.8
2012-10-03 16:05:07 +02:00
Joerg Bruehe
fc03d92ebe Merge changes for ULN RPMs into main 5.5 2012-09-21 16:01:53 +02:00
Tor Didriksen
b143676297 Bug#14593123 CONFIGURE.PL WITH CMAKE2.8 BREAKS THE BUILD
Ignore --with-client-ldflags 
it's not supported by the cmake scripts anyways.

Ignore --with-mysqld-ldflags
it's only used with --with-mysqld-ldflags=-static
and that doesn't work.
2012-09-07 10:12:32 +02:00
Harin Vadodaria
7baba64497 Bug#13904906: YASSL PRE-AUTH CRASH WITH 5.1.62, 5.5.22
Problem: Valgrind reports errors when an invalid certificate is used on the
         client.

Solution: Updated yaSSL to version 2.2.2.
2012-07-24 18:45:58 +05:30
Joerg Bruehe
6cd3c6b6b0 More changes to supporting the build of RPMs for ULN
from the MySQL 5.5 source tree.

Contrary to the comment, the spec file for these ULN RPMs
was missing in the previous changeset (blame ".bzrignore").

This change now brings said spec file, and it adds cmake
handling for the new files.

Still for internal tool tests only, not yet ready for publishing.


CMakeLists.txt:
  Add the subdirectory holding specific stuff
  for the RedHat-compatible RPMs for ULN.
cmake/install_layout.cmake:
  Explicitly mention my the ULN RPMs are handled.
2012-06-11 12:11:02 +02:00
Bjorn Munch
26c08ffa40 Merged some fixes from 7.2 release branches 2012-03-22 16:16:11 +01:00
MySQL Build Team
7a35cb9150 Updated/added copyright headers 2012-02-16 10:48:16 +01:00
Tor Didriksen
b507df6145 Bug#13463417 63487: ANNOYING TRACE MESSAGE IN CMAKE CODE
Remove it.
2011-12-15 10:44:33 +01:00
Daniel Fischer
7450044eb7 merge from 5.5.16 2011-09-21 12:40:41 +02:00
Jonathan Perkin
1f9fe1b4d5 Update email address for Release Engineering. 2011-09-20 17:47:53 +02:00
Daniel Fischer
888b8793a2 Source plugin library files list from cmake-generated file in RPM builds 2011-08-15 12:10:08 +02:00
Tor Didriksen
0fd2624780 Bug#12710832 - PB2 DAILY-TRUNK BUILD FAILS ON SOLARIS X86_64
Compiler bug:
http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6879978

Workaround is to always specify "-xOn".


cmake/build_configurations/mysql_release.cmake:
  Enable basic optimization for debug builds.
2011-07-04 14:27:08 +02:00
Kent Boortz
789aa8c485 Updated/added copyright headers 2011-07-04 01:25:49 +02:00
Kent Boortz
02e07e3b51 Updated/added copyright headers 2011-06-30 17:46:53 +02:00
Daniel Fischer
f089d2e870 RETURN doesn't exactly return from the macro... 2011-06-10 14:57:42 +02:00
Tor Didriksen
155df6c657 Bug#12641810 - MYSQL MAKE DIST DOESN'T WORK WHEN USING MYSQL TREE + PLUGIN TREE(S)
cmake/make_dist.cmake.in:
  Run 'bzr export' for plugins.
cmake/plugin.cmake:
  Lookup plugins with bzr repos.
2011-06-10 09:12:10 +02:00
Bjorn Munch
aca402b62b Followup to 12607800, testing it in PB2 didn't work, trying again
Be more explicit about path to (potential) plugin tests dirs
2011-06-01 15:19:36 +02:00
Bjorn Munch
fd717fb560 Bug #12607800 ADD HOOK TO INSTALL TESTS FROM IMPORTED FEATURE TREES
Sets INSTALL_PLUGINTESTDIR if any plugin/*/tests exist
2011-06-01 12:15:01 +02:00
Bjorn Munch
94f96bc472 Bug #12598603 HAVE COLLECTIONS FILES IN FEATURE TREES AUTO-APPENDED TO COMMON FILES
Do this in the common plugin.cmake but only if running in PB2
  (If done in manual builds it would create a bzr diff)
2011-05-27 14:43:15 +02:00
Bjorn Munch
b5009ee6be merge from 5.5-mtr 2011-05-26 10:18:21 +02:00
Tor Didriksen
c41e866357 Don't check for FIONREAD on windows.
Execution of platforms tests are slow/flaky when building on windows.
in PB:mysql-next-mr-opt-team on 2011-05-18 for win x86 debug_max, i see:
-- Looking for FIONREAD
-- Looking for FIONREAD - found
and the build fails.
2011-05-26 09:54:19 +02:00
Bjorn Munch
73d1992ef4 Bug #12586211 ENABLE GCOV BUILDS WITH CMAKE, INCLUDING PLUGINS
Added --with-gcov option to configure.pl and use that from SETUP.sh
2011-05-25 12:39:11 +02:00
Nirbhay Choubey
51721df03a Merge of fix for bug#11757855 from mysql-5.1 -> mysql-5.5. 2011-04-30 10:54:38 +05:30
Georgi Kodinov
c7b12ecc6e Backport of (part of) bug #11760838 to 5.5.
Enabled the ABI check to run on MacOSX.
2011-04-28 13:13:36 +03:00
Guilhem Bichot
fe1ec4a6e7 Fix for Bug#11892055 - "GCC COMPILER FLAG -WOVERLOADED-VIRTUAL NOT USED, WHICH LETS BUGS IN" 2011-04-26 11:18:29 +02:00
Magne Mahre
b71c5d51c6 Bug#11766320 MYSQL SYMBOLIC LINKS NOT WORKING
When MySQL converted from autotools to CMake, the
preprocessor symbol USE_SYMDIR was omitted by mistake.

Without this symbol, the code for checking .sym files
is not built.

This patch defines USE_SYMDIR when built on MS Windows.
2011-04-14 10:08:12 +02:00
Davi Arnaut
8aadfb4065 Remove some leftovers from the removal of the gethostbyname wrappers. 2011-04-13 16:05:26 -03:00
unknown
753c406994 Merge from mysql-5.5.10-release 2011-03-16 15:11:20 +01:00
MySQL Build Team
a471109971 Implementing Bug #57175 MySQL source tarballs should include generated spec file 2011-03-03 22:46:12 +01:00
Kent Boortz
fe12c639b1 The client shared library major version was changed to 18, to reflect
ABI changes (Bug#60061)

Bumping the version from 16 to 18, instead of 17, was done to avoid a
library conflict on Mac OS X between MySQL 5.1 and MySQL 5.5. In MySQL
5.1 GNU libtool was used, that made the ABI version used in the file
name to be 16, and the one stored inside the binary to be 17.

MySQL 5.5 uses CMake as a build tool, that will store the same ABI
number in the file name as inside the binary in Mac OS X, and then
bumping the ABI number two steps avoids a conflict on Mac OS X.
2011-02-17 19:36:19 +01:00
Joerg Bruehe
a7662cf0a9 Merge bug fix 42969 into main 5.5 branch. 2011-02-14 15:53:51 +01:00
Joerg Bruehe
bb956c1ea2 Fix bug#42969 Please add a MANIFEST to each build
With this change, there will be new files "INFO_SRC"
and "INFO_BIN", which describe the source and the
binaries.
They will be contained in all packages:
- in "tar.gz" and derived packages, in "docs/",
- in RPMs, in "/usr/share/doc/packages/MySQL-server".

"INFO_SRC" is also part of a source tarball.
It gives the version as exact as possible, preferably
by calling "bzr version-info" on the source tree.
If that is not possible, it just contains the three
level version number.

"INFO_BIN" contains some info when and where the
binaries were built, the options given to the compiler,
and the flags controlling the included features.

The tests (test "mysql" in the main suite) are extended
to verify the existence of both "INFO_SRC" and "INFO_BIN",
as well as some of the expected contents.


CMakeLists.txt:
  For the new files describing the source and the build
  ("INFO_SRC" and "INFO_BIN"), we need a new file
  "cmake/info_macros.cmake.in" with the build rules.
  
  1) This file must be configured with the current variables.
  
  2) "INFO_SRC" can be created during the cmake phase,
     but this should be repeated with each "make" to
  protect against a developer doing only "make" after
  a "bzr pull" (or "bzr commit").
  So have it both as a cmake rule and as a custom target.
  
  3) "INFO_BIN" must be created during the make phase
     only, because it contains information from files
  which will be written at the end of the cmake phase only.
  Therefore, it must be a custom target which is included
  in all "make" targets.
  
  4) The resulting "INFO_*" files must be included in packages.
cmake/info_bin.cmake:
  This is the file to create "INFO_BIN",
  by calling the "CREATE_INFO_BIN()" macro.
  
  It must be a separate file, so that the macro
  definitions can be included in other cmake scripts
  without that file inclusion causing a side effect,
  the macro call.
  That call would modify the source tree which should
  be trated read-only.
cmake/info_macros.cmake.in:
  This new file contains the macros to create the
  "INFO_*" files during various steps of the build,
  the calls will be at other places.
  
  1) For source: If running from a BZR tree, always create
     (update) "INFO_SRC" by running "bzr version-info".
     Outside a BZR tree, try to take it from exported
     sources, and create it only if missing, in that
     case put the three level version number into it.
  
  2) "INFO_BIN" contains
     - date/time and host name of the build host,
     - information about the platform,
     - information about the C and CXX compiler
       and the options given to them (Unix only),
     - the feature flags as reported by "cmake -L".
cmake/info_src.cmake:
  This is the file to create "INFO_SRC",
  by calling the "CREATE_INFO_SRC()" macro.
  
  It must be a separate file, so that the macro
  definitions can be included in other cmake scripts
  without that file inclusion causing a side effect,
  the macro call.
  That call would modify the source tree which should
  be trated read-only.
cmake/make_dist.cmake.in:
  Create a "VERSION_src" file during "make dist".
  
  In case it already exists from a preceding "cmake" run
  or tree export (which is quite likely), a new
  "make dist" must not modify it.
mysql-test/r/file_contents.result:
  Result of test for bug#42969.
mysql-test/t/file_contents.test:
  Perl test scriptlet for bug#42969.
support-files/mysql.spec.sh:
  Add "INFO_SRC" and "INFO_BIN" to the RPM contents.
2011-02-11 15:55:25 +01:00
Tor Didriksen
7f3decd9b2 Bug #58991 DEFAULT_CHARSET and DEFAULT_COLLATION does not work with CMake.
cmake/character_sets.cmake:
  Fix typo in variable name.
  Some cosmetics.
config.h.cmake:
  Do not use hard-coded defaults for charset and collation.
2011-01-04 11:23:45 +01:00
Tor Didriksen
029a5df870 Bug #58699 cannot build with gcc dbg on solaris
cmake/os/SunOS.cmake:
  Remove TARGET_OS_SOLARIS
config.h.cmake:
  Remove TARGET_OS_SOLARIS
  Add PTHREAD_ONCE_INITIALIZER
configure.cmake:
  Add function for testing whether we need { PTHREAD_ONCE_INIT } rather than PTHREAD_ONCE_INIT
include/my_pthread.h:
  Use PTHREAD_ONCE_INITIALIZER if set by cmake.
include/mysql/psi/mysql_file.h:
  Include my_global.h first, to get correct platform definitions.
mysys/ptr_cmp.c:
  Hide the unused static functions in #ifdef's on solaris.
  Use __sun (defined by both gcc and SunPro cc) rather than TARGET_OS_SOLARIS
sql/my_decimal.cc:
  Include my_global.h first, to get correct platform definitions.
sql/mysqld.cc:
  Fix signed/unsigned comparison warning.
sql/sql_audit.h:
  Include my_global.h first, to get correct platform definitions.
sql/sql_plugin.h:
  Include my_global.h first, to get correct platform definitions.
sql/sql_show.cc:
  Fix: warning: cast from pointer to integer of different size
sql/sys_vars.h:
  Use reinterpret_cast rather than c-style cast.
storage/perfschema/pfs_instr.cc:
  Include my_global.h first, to get correct platform definitions.
2010-12-21 13:00:26 +01:00
Alexander Nozdrin
3237e4908a Manual merge from mysql-5.5. 2010-12-16 21:43:21 +03:00
Georgi Kodinov
74e2520cd9 merge mysql-5.5->mysql-5.5-bugteam 2010-12-16 18:44:17 +02:00
Jonathan Perkin
5752c16ca9 bug#58955: Must -DBUILD_CONFIG=mysql_release require libaio on Linux
Allow users to build without aio if they really want to, by passing
-DIGNORE_AIO_CHECK to cmake.
2010-12-16 09:27:55 +00:00
Davi Arnaut
650d9cc5b0 Bug#58871: Reorganize maintainer mode compiler flags to allow
option for specific compilers

Reorganize the maintainer mode cmake code to allow options
for specific compilers. For now, enable -Wcheck for ICC,
but do not turn warnings into errors.

CMakeLists.txt:
  Move the code that sets options to cmake/maintainer.cmake
cmake/maintainer.cmake:
  Add macros for each specific compiler.
2010-12-15 08:30:09 -02:00
Jonathan Perkin
5046dc160b bug#58766: Server binary was compiled without fast-mutexes
Re-enable fast mutexes on Linux for release builds.
2010-12-07 11:40:38 +00:00
Jonathan Perkin
9836927774 Fix the .deb package prefix. 2010-11-25 00:21:28 +00:00
Vladislav Vaintroub
199cb4b7b1 Post-fix for Bug#58158 on Solaris
Problem : The build might fail with 
make[2]: *** No rule to make target `../sql/z',
needed by `sql/mysqld_dtrace_all.o'.  Stop.
if one of plugins would depends on system libz 
library

Fix: Filter out non-static dependent libraries
when dtracing static libs
2010-11-25 00:20:36 +00:00
Tor Didriksen
20f94499fa Bug #58372 cmake should not write temporary files in the source directory
cmake/do_abi_check.cmake:
  Put temporary files in ${BINARY_DIR} rather than in source directory.
  Slight reformatting for readability of commands.
2010-11-23 10:26:26 +01:00
Magnus Blåudd
6ad007c7ad Merge 5.5-bug58158 into 5.5-bugteam 2010-11-24 11:49:10 +01:00
Magnus Blåudd
41cab32111 WL#5665: Removal of the autotools-based build system
- Make the cmake files depend on VERSION, causing
   cmake rerun when VERSION changes
2010-11-24 11:37:59 +01:00
Sunanda Menon
377c783e75 merge 2010-11-22 11:35:48 +01:00