edb0a70d1f
STYLE: remove debug output
2007-07-16 13:29:41 -04:00
fac8604810
BUG: GET_DIRECTORY_PROPERTY(INCLUDE_DIRECTORIES|LINK_DIRECTORIES) wasn't
...
working, for both the result was always empty, since
cmMakefile::GetProperty() recognized it as a special property, constructed a
correct return value and called cmMakefile::SetProperty() with this list of
directories, which then didn't actually set the property, but applied it to
the internal vector of include/link directories. The following
getPropertyValue in cmMakefile::GetProperty() then still didn't find it and
returned nothing. Now for all special property the static string output is
used and its content is returned. I'm not sure it is the right way to fix
this problem but at least it seems to work and it fixes the Paraview3 build
Alex
2007-07-16 13:26:56 -04:00
c8010cd7fb
ENH: add LIST(CONTAINS ...) patch from "Miguel A. Figueroa-Villanueva, miguelf (AT) ieee.org
...
added tests for LIST(CONTAINS, SORT, REVERSE)
Alex
2007-07-12 11:56:45 -04:00
1f7eaa1c6a
COMP: hopefully fix test, finally
...
Alex
2007-07-09 08:16:04 -04:00
0cd81e8a8c
COMP: fix test
...
Alex
2007-07-08 22:06:32 -04:00
03ea7bf2e1
BUG: fix test
...
Alex
2007-07-06 14:02:42 -04:00
ef7b647b3b
COMP: OPTIONAL was missing in ENABLE_LANGUAGE()
...
-the assembler file seems to work for Linux and FreeBSD
-try to fix main() for HP-UX compiler
Alex
2007-07-06 08:53:16 -04:00
7627b96dd1
STYLE: some more output
...
Alex
2007-07-05 16:38:16 -04:00
0589ff74ae
COMP: skip APPLE, since there with universal binaries the assembler file would be built for both architectures
...
Alex
2007-07-05 16:32:40 -04:00
f9fb18f875
COMP: let's see if this assembler file works also on other platforms than linux...
...
Alex
2007-07-05 16:11:20 -04:00
5021eb90b9
ENH: add a simple assembler test
...
Alex
2007-07-05 09:05:40 -04:00
930bb0cd37
ENH: remove support for presetting CMAKE_SYSTEM_INFO_FILE,
...
CMAKE_SYSTEM_AND_C_COMPILER_INFO_FILE, CMAKE_SYSTEM_AND_CXX_COMPILER_INFO_FILE,
CMAKE_SYSTEM_AND_C_COMPILER_AND_PROCESSOR_INFO_FILE and CMAKE_SYSTEM_AND_CXX_COMPILER_AND_PROCESSOR_INFO_FILE
Instead of presetting these variables to arbitrary filenames, users should
set up CMAKE_SYSTEM_NAME and the compilers correctly and also create a
Platform/ directory so these files will all follow the official cmake style,
which should make it easier to understand and debug project which have their
own platform/toolchain support files.
-remove support for a suffix to MS crosscompilers, since this is not (yet)
supported by cmake and might confuse users
Alex
2007-07-02 13:29:36 -04:00
c8b263c674
ENH: Enable versioned executable test everywhere but XCode.
2007-07-02 09:58:08 -04:00
04525424af
BUG: For in-source version do not use a custom command output and custom target with the same name. This accidentally worked before but with a circular dependency.
2007-06-27 13:22:45 -04:00
07613ff7f8
COMP: fix tests where the building cmake doesn't have GET_TEST_PROPERTY
...
Alex
2007-06-27 08:43:06 -04:00
ae47302529
ENH: add a SimpleCOnly test, this is needed e.g. for testing sdcc since this
...
doesn't support C++ and also doesn't have a printf() implementation by
default
-add a test for mingw cross compiler
Alex
2007-06-26 15:30:02 -04:00
25a425ecf0
ENH:
2007-06-26 15:15:52 -04:00
509f1fb9cb
ENH: some property cleanup and added GetProperty
2007-06-25 09:51:37 -04:00
7f1fbe4e4c
STYLE: add some more output, so it is easier to understand
...
Alex
2007-06-21 14:06:52 -04:00
9a696669dc
COMP: big timeout for building kdelibs
...
Alex
2007-06-15 08:42:34 -04:00
53e5cf3a38
ENH: add test for buildingn kdelibs alpha1 ( http://websvn.kde.org/tags/KDE/3.90.1 ) with cmake
...
requires Qt >= 4.3.0, DBus, kdesupport (http://websvn.kde.org/trunk/kdesupport/ )
and the EasyDashboard scripts.
Alex
2007-06-14 13:55:23 -04:00
7203990661
COMP: removing the directory at the beginning breaks the test for in-source
...
builds
Alex
2007-06-14 08:33:21 -04:00
92270d5bf8
COMP: fix test, in some cases stdout from bar was not captured correctly,
...
probably because the process was killed before the fflush() worked because
the busy loop blocked the processor (failing midworld test)
Alex
2007-06-13 16:26:50 -04:00
227b942584
ENH: remove test
2007-06-11 17:15:00 -04:00
23ef759b40
ENH: add ConvLib test back for some time
2007-06-11 15:00:23 -04:00
0f4ad19ef8
ENH: Re-arranged code to test adding a custom command to generate a source file after the file has been added to a target. This is supported by the current implementation because of the use of source lists in the target implementation. When we later convert to creating cmSourceFile instances immediately for the target we need to make sure the mentioned case still works.
2007-06-11 13:22:17 -04:00
db0f26e852
BUG: Fixed @ONLY configuration to not try to parse ${} syntax at all. This fixes the original fix to bug#4393 and adds a test.
2007-06-06 16:20:02 -04:00
c8067b3904
ENH: fix it
2007-06-06 13:43:46 -04:00
b4f4a06102
ENH: use lower case for file compare on windows
2007-06-06 13:32:01 -04:00
61ccdc4e92
ENH: add missing file
2007-06-04 17:18:31 -04:00
64a8eb990a
ENH: add one of the headers to the regular sources
2007-06-04 15:28:05 -04:00
eddf1cf39f
ENH: improve TRY_RUN() for crosscompiling: instead of just failing, it now
...
creates two cache variables, one for the RUN_RESULT, one for the RUN_OUTPUT
(if required), which can be set or preset by the user. It has now also two
new arguments: RUN_OUTPUT_VARIABLE and COMPILE_OUTPUT_VARIABLE (the old
OUTPUT_VARIABLE merges both), so if only COMPILE_OUTPUT_VARIABLE is used the
run time output of the TRY_RUN is unused and the user doesn't have to care
about the output when crosscompiling. This is now used in FindThreads.cmake,
CheckC/CXXSourceRuns.cmake and TestBigEndian.cmake, which used the output
only for the logfile (compile output is still there). Test/TryCompile/ now
also tests the behaviour of OUTPUT_VARIABLE, RUN_OUTPUT_VARIABLE and
COMPILE_OUTPUT_VARIABLE.
Alex
2007-06-01 11:16:29 -04:00
2f07e8e8f8
ENH: added new test for SourceFile objects and properties
2007-05-31 10:29:13 -04:00
eed7c8dec1
ENH: always provide CMAKE_SYSTEM_XXX() and MAKE_HOST_SYSTEM_XXX() variables,
...
so when cross compiling the build host platform can be tested
Alex
2007-05-29 11:36:07 -04:00
06472d620a
ENH: Added more install rules to increase coverage of the command.
2007-05-28 11:41:29 -04:00
f4eb541880
ENH: make the compiler id detection work, even if the output file name of
...
the compiler is completely unknown and even if it produces intel hex or
motorola s-record files, with test
Alex
2007-05-25 15:22:22 -04:00
a99c60b0ed
ENH: Added testing of REGEX option to INSTALL(DIRECTORY). Added tests to cover all forms of old-style install commands.
2007-05-25 11:41:37 -04:00
3124c60938
ENH: Add test to see if INSTALL_FILES actually worked.
2007-05-25 11:09:22 -04:00
7147c3e1cc
ENH: add COPY_FILE argument to TRY_COMPILE, so the compiled executable can
...
be used e.g. for getting strings out of it.
Alex
2007-05-24 12:06:59 -04:00
7d7aba292c
ENH: add two simple tests for TRY_RUN()
...
STYLE: create a new base class cmCoreTryCompile, from which
cmTryCompileCommand and cmTryRunCommand are derived, so there are no public
static functions with lots of arguments anymore
Alex
2007-05-24 11:27:51 -04:00
6a145fb2be
BUG: Report proper error message when project does not build the first time. Also added hack to rebuild subproject several times for Xcode. The generator should be fixed and the hack removed.
2007-05-23 17:58:14 -04:00
c3474ad029
ENH: Executable bar should rebuild when its generated header changes. It does not need to link to the foo library anymore.
2007-05-23 17:01:26 -04:00
303b4312b2
BUG: Target names in the COMMAND part of a custom command should not create a file-level dependency that forces the command to rerun when the executable target rebuilds, but the target-level dependency should still be created. Target names in a DEPENDS should do both a target-level and file-level dependency. Updated the BuildDepends test to check that this works.
2007-05-23 13:30:23 -04:00
c288da754a
BUG: Target names in the COMMAND part of a custom command should not create a file-level dependency that forces the command to rerun when the executable target rebuilds, but the target-level dependency should still be created. Target names in a DEPENDS should do both a target-level and file-level dependency. Updated the BuildDepends test to check that this works.
2007-05-23 13:27:00 -04:00
3b4aa3bf35
BUG: always search for ar, ranlib, etc. except under MSVC -> this should fix the mingw fortran test
...
-also generate the fortran test with the kdevelop generator
Alex
2007-05-21 10:15:42 -04:00
7416d0d67d
BUG: Fix name of project to build for LoadCommandOneConfig now that it has been renamed for new name of LoadCommand test.
2007-05-20 10:11:48 -04:00
5e45ca6fbf
BUG: Finish fixing test for new name.
2007-05-19 10:15:24 -04:00
e2ea9391e6
BUG: fix project name for test
2007-05-19 10:10:29 -04:00
df1a77c225
ENH: some cleanup, condensing some tests, removing arguments that were not needed but rather were cut and paste copies etc
2007-05-18 15:11:20 -04:00
9bfb3cacbb
BUG: Replace "with space" in custom command argument tests with "w s" to still have whitespace but be shorter. The test was failing because the custom command line length was simply too long for the VS IDE.
2007-05-18 09:33:59 -04:00
Alexander Neundorf
Brad King
Ken Martin
Bill Hoffman