Judging from the questions I see on the #cmake IRC channel this is one
of the most common pitfalls for people coming from other build systems,
especially plain Makefiles. Finally put this into the documentation to
stop people getting fooled into using this.
Text by Brad King.
Provide the ability to configure CTest with settings different from the ones
available in the source tree by checking first if CTestConfig.cmake
exists in the build tree.
The motivation is to allow build system checking out external project to
test and/or package them and submit the associated results to a different
dashboard than the one specified (or not) in the source of the external
project.
For example, the build system of Slicer can checkout, build, test
and package what I will call "extensions". These extensions can be developed
by third parties who can test and submit to their own dashboard / project.
When checked out by Slicer build system, the default dashboard can now be
overwritten by adding a custom CTestConfig.cmake to the build directory.
And if not overwritten, it would avoid to create CTestConfig.cmake within
the source checkout of the extension.
Important when calling ctest commands in a loop from a script.
Each time Populate gets called, it uses the current definition
of the variable. Without the clear, it was accumulating additional
identical values each time through the loop.
66bd543 Eclipse: fix#12417, don't create wrong src pathentries
70de8bd Eclipse: detect number of CPUs, set CMAKE_ECLIPSE_MAKE_ARGUMENTS accordigly
117f2b8 Eclipse: add Build and Clean targets to targets
c3f30bd Eclipse: move code for generating links to targets into separate function
cef6bd9 Eclipse: move code for generating links to projects into separate function
b6d4de7 Eclipse: add virtual folder for each target
5e8e9ad automoc: always run moc on the cpp file if there is a foo.moc included
ea8ac9f Add copyright notices
56aa6ff automoc:run moc on the header if the source file contains include "foo.moc"
In the log entry, the newline is missing. The output without the newline
character is a bit strange, like
SetCTestConfigurationFromCMakeVariable:MemoryCheckCommand:CTEST_MEMORYCHECK_COMMANDSetCTestConfiguration:MemoryCheckCommand:/usr/bin/valgrind
Instead of
SetCTestConfigurationFromCMakeVariable:MemoryCheckCommand:CTEST_MEMORYCHECK_COMMAND
SetCTestConfiguration:MemoryCheckCommand:/usr/bin/valgrind
This patch changes this to add a newline.
Signed-off-by: Bernhard Walle <walle@corscience.de>
Define variable CMAKE_LINK_INTERFACE_LIBRARIES to initialize the
value of this property when a target is created. This allows authors
to write
set(CMAKE_LINK_INTERFACE_LIBRARIES "")
to disable transitive linking to implementation dependencies of shared
libraries on platforms where it is possible.
They were miss-aligned with the 'enter' key instruction.
Before:
----
Press [enter] to edit option CMake Version 2.8.3.20110107-g4b05a-dirty
Press [c] to configure
Press [h] for help Press [q] to quit without generating
----
After:
----
Press [enter] to edit option CMake Version 2.8.3.20110107-g4b05a-dirty
Press [c] to configure
Press [h] for help Press [q] to quit without generating
----
Actually the Eclipse generator now does not create any CDT_SOURCE
pathentries anymore, since I was not able to find out what they are
good for, not even by asking on the cdt-dev mailing list.
So, at least the warning from eclipse about bad pathentries are gone
this way, and I didn't see anything which was not working anymore.
Let's see whether we can find out what they are good for.
Alex
With this commit the virtual folder for the targets now
have "Build" and "Clean" targets associated to them, so you can
build and clean per-target now in the project explorer.
Alex
This makes automoc behaves as the documentation says.
If there is a #include "foo.moc" in the source file, moc
will be executed on foo.cpp.
Before it was also executed on foo.cpp, but only if foo.cpp
contained a Q_OBJECT macro, otherwise moc was executed on
foo.h. This was confusing, and this change also shouldn't break
anything, since the headers are moc'ed anyway if they contain
a Q_OBJECT macro.
Alex
This fixes#12533.
Before automoc did not check the header if the source file contained a
statement, now it does.
Additionally, moc is now only run on explicitely listed headers which
contain a Q_OBJECT macro.
Alex
COVERAGE_EXTRA_FLAGS is a space separated value of extra flags
that will be passed to gcov when ctest's coverage handler invokes
gcov to do coverage analysis.
Map to CoverageExtraFlags in the CTest ini file. Use default value
of "-l" to match the coverage handler's earlier behavior from ctest
2.8.4 and earlier. The fix for related issue #11717 had added a " -p"
which was the cause of both #12415 and #12490. Here, we revert that
change to the default value, so -p is no longer there by default.
The people that care to add -p may do so in their own build trees
by appending " -p" to the new cache variable COVERAGE_EXTRA_FLAGS.
For each target a virtual folder is created, which contains
one virtual folder for each sourcegroup, which contain
links to the actual source files (#12294, #12223)
Alex
dcd2459 Eclipse: better message when Eclipse version could not be determined
b4b2fc3 Eclipse: don't create VirtualFolders if not supported
5b200e3 Detect whether the current Eclipse version supports VirtualFolders
4974ec9 Eclipse generator: detect Eclipse version
32f8437 Fix line-too-long style violations
029ab31 Constify XCode generator getters to match cmGlobalGenerator
fec4b63 Fix configuration-dependent flag lookup in cmLocalGenerator::GetTargetFlags
557956f Introduce a cmGlobalGenerator::ResolveLanguageCompiler function
5b114c9 Introduce a cmLocalGenerator::ConvertToIncludeReference function
903d914 Make cmLocalGenerator::ConvertToLinkReference virtual
8a0eb78 Constify many getters of cmGlobalGenerator.
4532d36 Add const versions of some getters.
3db2973 Refactor TargetTypeNames.
Commit 8a0eb78f (Constify many getters of cmGlobalGenerator, 2011-03-26)
added const qualifiers to many cmGlobalGenerator methods but left the
resulting lines beyond our style's limit of 79 characters.
Commit 8a0eb78f (Constify many getters of cmGlobalGenerator, 2011-03-26)
added const qualifiers to many cmGlobalGenerator methods. Fix the
signature of the virtual function overrides in cmGlobalXCodeGenerator to
match.
Files for the ASM language are those assembler files which are processed
by the C/CXX compiler, and they may contain preprocessor directives, so
run the C dependency scanner also on them.
Alex
Specifically, perform configuration-dependent lookup of
STATIC_LIBRARY_FLAGS for static libraries, and use the correct prefix
for configuration-dependent lookup of LINK_FLAGS (i.e. "LINK_FLAGS_",
as opposed to the value of the LINK_FLAGS property).
Make it a static method instead of an array. It is safer for the
type checking and if we add a new target type we will be warned to add
a case to the switch.
As Dave Cole pointed out the previous commit only checked for 10.x and 12.x.
11.0 was accounted for, but 11.1, 11.2 and 11.3 were not. This patch
should make it work for those versions as well. I did a web check and there
are 11.0, 11.1, 11.2 and 11.3 versions from Intel. I assume if 12.x uses
11.0 as the version in the .vfproj file, then all of the 11.x versions would
as well.
f15945e Use makefile->IssueMessage() for better error messages
c886e31 Improve behaviour of --find-package mode with try_run/try_compile
31f43fd Remove trailing whitespace
The intel compiler for 12.0 and 12.1 are known to expect the file version
to be 11.0 in the .vfproj file. For 10.x it should be 9.10. Prior to
this fix 12.0 and 10.1 were the only values checked. If those did not match
the actual version of intel was put in the vfproj file causing an error
about future version load attempt in the IDE.
All find_* commands re-root the list of paths and then add trailing
slashes. Factor this pair of calls out into a dedicated method. The
new method would be the only caller to AddTrailingSlashes, so subsume
that method into it.
If a http server responds with a result code greater than 400 then the
data returned from the download probably do not match that expected.
Teach file(DOWNLOAD) to fail with an error in this case instead of
silently pretending that the download worked. The file(UPLOAD) command
already does this.
Instead of failing with an error message which says things have not been
setup correctly, it now says that try_compile/try_run() are not
supported in find_package mode (#12426).
Alex
Caveats apply: only valid when not overriding default value for output
dir *and* there is a CMAKE_XCODE_EFFECTIVE_PLATFORMS value.
For now, CMAKE_XCODE_EFFECTIVE_PLATFORMS must be defined in each
project, or in the cache for a build.
Code relying on value of LOCATION and similar properties may not
work if it depends on file existence on disk since CMake will not
evaluate $(EFFECTIVE_PLATFORM_NAME). This feature is only enabled
for making it easier to build Xcode iOS projects where the developer
wants to be able to switch easily between simulator and device builds.
90efed6 Xcode: Honor Fortran_FORMAT target and source file property
5c0c635 Fortran: Add support for free- and fixed-form flags
47a0c75 VS: Map Fortran free- and fixed-format flags to IDE options
d6e2a06 VS: Map per-source Fortran flags to IDE options
The '-E build build_dir' command was created and documented, but then
morphed into '--build build_dir' instead, ... and then the -E documentation
was never removed. This commit fixes that oversight.
...when building CPack archive-based packages (.tar.gz and similar)
Rather, put the symlinks-to-directories into the archive as files,
and expect/trust that the things the symlinks point to are also in
the archive.
Do not recurse through directory symlinks when adding files.
Recursing through directory symlinks will generate broken archives,
i.e., they will look something like this:
foo -> bar/bar
foo/Info <- Shouldn't be in archive.
bar/bar
bar/bar/Info
This behaviour was previously broken; regardless of the
RecurseThroughSymLinks value, symlinks to directories were
NEVER added as files in the results.
When RecurseThroughSymLinks is ON, symlinks to directories
should be recursed as if they were the actual directories
to gather the files within.
However, when RecurseThroughSymLinks is OFF, symlinks to
directories should be returned as files in the result.
Inspired-by: Johan Björk <phb@spotify.com>
Previous code was missing some matches in the output. This commit
fixes the regular expressions used for output matching to detect
numbers reported with commas in them, too.
Without this call, recent refactoring changes in the Visual
Studio generators yield no GUIDs saved in the cache. Putting
the SaveCache here *once* avoids scattering many calls to it
in specific generators.
Added hashing fuction for std::string. This adds default support for
std::strings to KWSys hashing containers.
Author: Bradley Lowekamp <blowekamp@mail.nih.gov>
Suggested-by: Arnaud Gelas <arnaud_gelas@hms.harvard.edu>
Change-Id: I7e7a0c356b73d19868a3df1db57b702ec7fffe9d
Define a "Fortran_FORMAT" target and source file property. Initialize
the target property from a "CMAKE_Fortran_FORMAT" variable. Interpret
values "FIXED" and "FREE" to indicate the source file format. Append
corresponding flags to the compiler command line.
59238dc Fix --find-package mode on Cygwin, where enable_language(RC) is called
98472e4 Require the current cmake version in --find-package mode
a6ccf3c Use $(CXXFLAGS) and $(LDFLAGS) in the --find-package test Makefile
4386918 Fix line length
7d69310 Only enable the test when using GNU make
3011149 Make the test harder by always having a space in the include dirs
ab57ff6 Make the --find-package test harder
626fc71 Much improved test, should now be executed on all UNIXes
ec6982d Disable any STATUS output in --find-package mode
e552ae7 Dont check for -isysroot and -mmacosx-version on OSX in --find-package mode
e589589 Rename helper macros print_compile_flags() to set_compile_flags_var()
aecfc1f Fix test on OpenBSD with BSD make
6bb4ca3 The makefile for the test was kindof wrong
fd15b5e Only run the test if we are using a makefile generator under UNIX
9fc87c6 Add a test for the new --find-package mode
d3ae0ff Improve documentation for --find-package mode
bf07375 Add a cmake.m4 for using cmake in autoconf projects instead of pkgconfig
b0e3578 Use the file-utility to test for 64bit if there is no /usr/lib64
53edfb2 Better support for lib64 and Debian multiarch
b8fdaa1 Fix copyright notice in new CMakeFindPackageMode.cmake
7690edf Replace cmake::GetScriptMode() with GetWorkingMode()
e4f603b Implement find-package mode of cmake
a91d662 Add find-package mode, which does nothing yet
b976e70 Make clLocalGenerator::GetTargetFlags() public
920a046 QtAutomoc: Eliminate compiler warning
b00463f QtAutomoc test: Pass QT_QMAKE_EXECUTABLE
e78ce44 Fix automoc with VS builds: apply patch from Bill
71165e9 Silence warning in automoc: use long instead of int
1879bcc Fix build: use std::ios::out|ios::trunc instead of std::ios_base::out
678e124 Only enable the automoc test after checking that Qt4 works
71c29d1 Fix bootstrap test with automoc
afb3edc Fix warnings
add30e9 Fix build: non-void function must return a value
7e6d845 Automoc.cmake is not needed anymore
2963d0b Fix logic which decides when to execute automoc test
77a5c6e Add documentation for AUTOMOC, add initialization via CMAKE_AUTOMOC
bf8ef77 Add a test for automoc
d045fd4 Nicer progress message for the automoc target
50cd6ce Move automoc processing from add_executable/library to cmGlobalGenerator
cbaac2a Remove trailing whitespace
c27607b Refactor SetupAutomocTarget() so it can be run after creating the target
24d9b7d Remove trailing whitespace
58b7fe6 Use cout instead of printf()
72caf4d Add the generated automoc.cpp file to the cleaned files
ddb517d Color output when running moc
9303295 Initialize verbose based onb the env.var.
ace1215 Move code for parsing a cpp-file from the big loop to separate function
735a5bb Fix line lengths
83b730c Add AUTOMOC to the add_library() command
126c6ea Add the cmake module required currently for automoc
de91feb Remove the need to check for .h/.cxx during buildtime
d65689a Add actual automoc code from automoc
d1c0a5f Start implementing skeleton for automoc in cmake
a65011b Start work on automoc: add empty cmQtAutomoc class
For project and target objects, save their ids in CMakeCache.txt.
Hopefully, that will be enough to allow user settings to be saved
across multiple CMake generate operations. Other object types may
also need their ids saved: if so, more code than this commit
will be necessary...
If you had a + in the name of a target with nmake, it created a variable
in the makefile that used + in its name, which is not allowed by nmake.
To make the implementation easier, + is now not allowed for any make
generators as part of a variable name.
In SystemTools::ClassInitialize, remove call to AddTranslationPath
that was originally put in place to "work around an SGI problem."
This code precluded using CMake effectively in valid directories
under "/tmp_mnt/"
c9761de Improve documentation for WriteBasicConfigVersionFile.cmake
208bb90 Set UNSUITABLE instead of not COMPATIBLE
bb03c2d Really fix copyright notice
d50a61a Fix copyright notice
4ba09bc Add some tests for write_basic_config_version_file()
02b1e4b Add example to documentation
d216a67 Provide macro write_basic_config_version_file()
Any STATUS output will be fed directly to the compiler, which
will not understand any status messages. Error messages
are fine, since they are errors and it is ok if the compiler
fails in such cases.
Alex
It already works :-)
Needs more refactoring, e.g. using the cmake facilities to turn a string
into a vector etc.
Also still missing is the part which creates the custom target.
Alex
Makes VS 7, 8 and 9 generators consistent with the VS 10
generator. Adds the "AdditionalIncludeDirectories" attribute
at the vcproj level so that all idl files inherit the /I
command line args.
For the specialization of hash<>(), the types long long and __int64
may be the same type. While the CMakeLists indicate that if __int64 is
a alias for another type the it will not be enabled, on mingw they
both appear to be the same type and enabled.
This patch only enable specialization for long long OR __int64 to
avoid the potential conflict.
Author: Bradley Lowekamp <blowekamp@mail.nih.gov>
Change-Id: I813a9ac008b296fab5a369c48e6dd5460fd0c035
Added hash function for types long long and __int64, conditional on
detection by FundamentalType.
Author: Bradley Lowekamp <blowekamp@mail.nih.gov>
Change-Id: Ie273f55cd4387ca3dbbe00b9d96ad4935e456c9e
For custom commands in VS2010 Fortran projects the INTDIR variable
is different than in the rest of the solution because Intel
fortran still uses the old VS project files even in VS2010. So,
we replace $(Configuration) directly in the project files. I have also
added a FortranOnly test that tests this feature and is run on any
generator that has Fortran abilities.
We already use GetSystemTimeAsFileTime() and gettimeofday()
unconditionally on supported Windows and non-Windows platforms,
respectively. Remove outdated portability complexity.
This deterministic ordering makes it easier for a developer
comparing the ctest output of two test runs to see what changed
from run to run.
Thanks to Fraser Hutchison for the patch.
e2e8c0a Also put builtin include dirs into CodeBlocks project file
92c0dc5 Remove useless line of code
a5683f8 Patch by Campbell Barton: puts definitions into C::B project file
b62349c FeatureSummary.cmake: update documentation
f366cf8 FeatureSummary.cmake: cosmetics
f407bb5 FeatureSummary.cmake: only higher TYPEs can override previous TYPEs
02d47ab FeatureSummary.cmake: error out when a REQUIRED package is missing
91a1527 FeatureSummary.cmake: add INCLUDE_QUIET_PACKAGES keyword
0671a02 FeatureSummary.cmake: remove "comment" field
aae13f4 Extend FeatureSummary: add PURPOSE of package and TYPE
Refactor code from CMakeFindEclipseCDT4.cmake so it can be used
alkso for CodeBlocks, and move it into new file
CMakeExtraGeneratorDetermineCompilerMacrosAndIncludeDirs.cmake.
Alex
This macro can be used to generate basic version files which can be
installed along a Config.cmake file to provide versioning support.
This (3rd try) is implemented using a macro, which maps
the COMPATIBILITY mode to a filename and configure_file()s the
resulting file.
Alex
Since commit 85c0a69a (Cygwin: Do not define 'WIN32', 2010-12-17) WIN32
is not defined on Cygwin. Update documentation of the WIN32 variable
accordingly.
Fix NSIS template to more thoroughly use CPACK_PACKAGE_INSTALL_REGISTRY_KEY.
This allows different versions of software to have a separate sections in the
registry to keep track of things (installed components, and uninstall stuff).
Change default of CPACK_PACKAGE_INSTALL_REGISTRY_KEY to follow the value of
CPACK_PACKAGE_INSTALL_DIRECTORY so if an installation overwrites another installation,
the proper registry entries are more likely to be overwritten.
Fix CPack/NSIS generator to not insert code in the NSIS template to skip installation
of already installed components. This enables a repair like behavior and also enables
installing patch releases on top of an older installation.
Thanks to Pau Garcia i Quiles for the inspiration for the patch.
I've tweaked it a bit compared to what's in the bug tracker: this
commit does not allow empty global variable names.
I also added usage of the new feature to an existing test. Although
it has no effect on the resulting Visual Studio projects, you can
verify that the VSResource test produces a non-empty globals section
in the generated .vcproj(x) files.
4a50b5f Add a basic test for CMAKE_DISABLE_FIND_PACKAGE_<package>
a66df08 Add documentation for the CMAKE_DISABLE_FIND_PACKAGE_<Name> switch
6e1d3ed Add a switch to disable a find_package() call completely
Now found packages are not stored in ENABLED/DISABLED_FEATURES
anymore, but always in PACKAGES_FOUND/NOT_FOUND.
ENABLED/DISABLED_FEATURES is now only used via
ADD_FEATURE_INFO(), e.g. for stuff set via option().
Alex
set_property() has APPEND, which creates a list. E.g. when
appending to COMPILE_FLAGS a string is needed, not a list.
With the APPEND_STRING option the value is append as string,
not as list.
Alex
In find-package mode, cmake executes Modules/CMakeFindPackage.cmake,
which calls find_package(), and this is then evaluated in cmake.cxx,
which prints an appropriate message to stdout, so it can be used
e.g. in a normal Makefile:
$ /opt/cmake-HEAD/bin/cmake --find-package -DNAME=JPEG
-DCOMPILER_ID=GNU -DLANGUAGE=C -DMODE=EXIST
JPEG found.
$ /opt/cmake-HEAD/bin/cmake --find-package -DNAME=JPEG
-DCOMPILER_ID=GNU -DLANGUAGE=C -DMODE=COMPILE
$ /opt/cmake-HEAD/bin/cmake --find-package -DNAME=JPEG
-DCOMPILER_ID=GNU -DLANGUAGE=C -DMODE=LINK
-rdynamic -ljpeg
Alex
A common mistake when using custom commands is to list the output of a
command in multiple targets that may build in parallel. Warn against
this case in the documentation and suggest a better approach.
Update cm_sha2.[hc] from sha2.[hc] in "sha2-1.1-ALPHA.tgz" downloaded
today from
http://www.aarongifford.com/computers/sha.html
with trivial whitespace cleanup. This adds SHA-224 support.
Copy cm_sha2.[hc] from sha2.[hc] in "sha2-1.0.tar.gz" downloaded today
from
http://www.aarongifford.com/computers/sha.html
with trivial whitespace cleanup. Also fix #include to account for
rename.
8555c2b Look for VCExpress as a possible build tool as well as devenv.
ed0075b Use relative paths for custom command inputs.
38368d5 Revert "With very long file names, VS 2010 was unable to compile files."
8cd66dc Use devenv instead of msbuild for vs2010.
For source files we use full paths. This allows for longer directory
names with VS2010. However, the use of full paths causes the GUI
to not display the custom commands.
As discussed, this adds a switch CMAKE_DISABLE_FIND_PACKAGE_<name>,
which, when set to TRUE, keeps the find_package() command from executing.
When the package is REQUIRED at the same time, an error is generated.
Alex
The workaround added by commit 7e92f0b4 (Hack to make echo command work
properly in mingw32-make, 2006-10-05) and updated by commit 69356d8a
(Juse use cmake -E echo instead of the native echo, 2006-10-13) no
longer seems necessary with modern mingw32-make. Furthermore it slows
performance due to the time spent loading a cmake process instead of
plain echo.
Do not require the C language to be enabled to do lib->lib64 conversion.
The check was originally added by commit a5825cd1 (check in new find
stuff, 2006-03-02) to ensure that CMAKE_SIZEOF_VOID_P is set. Since
commit 3fdf1411 (FIND_LIBRARY should not require CMAKE_SIZEOF_VOID_P,
2008-02-11) the code following the check does not fail when the variable
is not set, so just remove the original check. This allows conversion
for any language that is enabled so long as the address size is known.
Reported-by: Christoph Höger <choeger@cs.tu-berlin.de>
1ed19bc multiarch: Set CMAKE_LIBRARY_ARCHITECTURE_REGEX for Linux|Hurd|kFreeBSD
52a6ed2 Test find_package multiarch support (#12037)
b41ad3b Teach find_(library|package) about Linux multiarch (#12037)
When add_test(NAME) is called without the CONFIGURATIONS argument then
the test is intended to run in any configuration. In multi-config
generators like the VS IDE and Xcode tests created by add_test(NAME) can
only be run when testing a known configuration (otherwise there is no
way to generate the test command line). If no test command line is
known for a particular configuration, or if no configuration is given to
ctest, report the test as not run instead of silently skipping it.
Also fix CMake's own TestsWorkingDirectory test invocation to correct a
previously silent failure exposed by this change.
When generating per-config blocks in test and install scripts replace
the form
IF()
# config == A
ENDIF()
IF()
# config == B
ENDIF()
with
IF()
# config == A
ELSEIF()
# config == B
ELSE()
# no config matches
ENDIF()
for clarity and to support the else() case cleanly.
f09ba0f Fix style errors added by parent and grandparent
eeeeca1 XCode: Support target folders on XCode.
59ed84e Xcode: Support multiple level nesting of XCode folders (#10039)
d0a403f CMake: Move tokenize to cmSystemTools
Implement support for multiarch as specified here:
http://wiki.debian.org/Multiarchhttps://wiki.ubuntu.com/MultiarchSpec
Detect the <arch> part of <prefix>/lib/<arch> from the implicit library
search path from each compiler to set CMAKE_<lang>_LIBRARY_ARCHITECTURE.
Define CMAKE_LIBRARY_ARCHITECTURE using one of these values (they should
all be the same). Teach the find_library and find_package commands to
search <prefix>/lib/<arch> whenever they would search <prefix>/lib.
27aa446 One more try. Use full path by default, and relative on broken compilers.
2dfc121 Use bin tree for inclues to avoid -I with spaces in the path.
6d29b4b Append and do not clobber CMAKE_CXX_FLAGS in the test.
7815e90 Fix for bug#10798. VS10 did not append -I flags with COMPILE_FLAGS prop.
In cmMakefile.cxx GetCMakeCFGInitDirectory is replaced with
GetCMakeFilesDirectory for .rule files. In some cases with
external projects, that directory will not exist. With vs10
the .rule files must exist or the rule will run with every build.
This fix creates the path that the .rule file is in. In addition,
it is now a CMake error if the .rule file can not be created.
A post-build phase of each target invokes the XCODE_DEPEND_HELPER.make
file to erase any targets that link to it. Narrow the set of targets
tested by each post-build phase to those that depend on the newly
completed target. This avoids removing files from partially built
unrelated targets that happen to be building in parallel.
0e591ed Fix type conversion warning
9c3a0b9 We will actually compress memcheck output if the server supports it.
8024c53 Dynamic analysis test output should not be compressed.
cdc2b41 Fix CompileCommandOutput test build on Windows
7039d1f Fix CompileCommandOutput test for Make tools not supporting spaces
4268e3d run_compile_commands: Cast istream::get() result to char
c45c60b run_compile_commands: Avoid extra stl vector conversion
7c5be51 run_compile_commands: Avoid shadow in std::map<>::at workaround
169bb05 Provide std::map<>::at for use in run_compile_commands
4e2185c Make std::map usage more portable in language=>flags/defines maps
a7e7a04 Fix run_compile_commands build on Apple GCC 3.3
c9174c0 Fix signed/unsigned comparison in EscapeJSON
8346a28 Only offer the compile command output feature on unix systems
0e6b05f Adds a test for the compile command line output.
5674844 make compile command output optional
fe07b05 implement cxx command output
65c0c24 cache flags and defines
3f064ef refactor flags and defines
7ff98b7 Fix forced-seed argument type in string(RANDOM)
3d92c8c Explicitly cast time value in cmSystemTools::RandomSeed
eb6f461 VS 6: Define _WIN32_WINNT to load wincrypt.h correctly
e1b0a11 Improve string(RANDOM) default seed
Clang points out that local variable 'seed' needs to be "unsigned int":
Source/cmStringCommand.cxx:828:21: warning: operands of ? are integers
of different signs: 'int' and 'unsigned int' [-Wsign-compare]
srand(force_seed? seed : cmSystemTools::RandomSeed());
^ ~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~
Use static_cast to avoid warnings like
conversion to ‘unsigned int’ from ‘__time_t’ may alter its value
conversion to ‘unsigned int’ from ‘__suseconds_t’ may alter its value
We do not care if the value is truncated because we are looking for just
32 bits anyway.
The Sun compiler does not provide the proper vector constructor to
initialize it from an iterator pair of a non-matching type. Extend the
ParseUnixCommandLine API to provide a vector of the proper type so no
conversion is needed.
The naive time(0) seed is unique only within one second. Instead try to
read a real source of entropy and otherwise fall back to a combination
of the process id and high-resolution time.
Older versions of GCC, the HP compiler, and the SGI MIPSpro compiler do
not like the use of make_pair in this case and the conversions it
requires:
a value of type "const char *" cannot be used to initialize an entity
of type "char [1]"
/usr/include/g++-3/stl_pair.h:68: assignment of read-only location
Instead use a map lookup pattern already used throughout the rest of our
source tree.
Under Windows "locationURI" must be used for virtual folders, while
"location" must be used only for linked folders. Under Linux it doesn't
seem to matter.
Alex
08b9397 find_package: Fix system package registry test path conversion
93021ad find_package: Test system package registry when possible
b95f3ca find_package: Check both 32-bit and 64-bit registry views
a0d76c1 find_package: Search a "system package registry"
549458f find_package: Document user package registry locations
c9563db find_package: Cleanup user package registry less aggressively
4df1197 find_package: Rename implementation of user package registry
08fa5dd Also generate dependers-graphviz files.
4f96a76 GRAPHVIZ_IGNORE_TARGETS is now a list of regular expressions
5698ad2 Make it possible to exlude external libs from dot files
The system package registry is under HKEY_LOCAL_MACHINE\SOFTWARE which
has separate views for 32-bit and 64-bit applications. Look in both
views, but prefer the architecture matching the build target platform.
Generalize the "user package registry" created by commit ed0650f6 (Teach
find_package to search a "package registry", 2009-09-01). Define a
corresponding "system" registry key under HKEY_LOCAL_MACHINE. This
gives package installers a place to create a registry value that points
at the right location for find_package() to locate the package.
Delete only REG_SZ entries that are specifically detected to point to
invalid paths. This will allow future versions to add other value types
for different purposes.
This file has not been used in KWSys since commit "Moved test for large
file support into kwsysPlatformCxxTests.cxx", 2006-08-25. CMake 2.6.0
and above come with a copy of this module anyway, and KWSys has required
CMake 2.6.3 since commit "KWSys: Require at least CMake 2.6.3",
2011-03-01.
234bae7 VS10: Fix exit code of custom commands with setlocal/endlocal (#11938)
b98fdd5 VS: Use setlocal/endlocal only in VS 10 custom commands
06fcbc4 VS10: Fix working directory of consecutive custom commands (#11938)
Use the pattern
setlocal
...
endlocal & call :cmErrorLevel %errorlevel% & goto :cmDone
:cmErrorLevel
exit /b %1
:cmDone
if %errorlevel% neq 0 goto :VCEnd
in custom commands to preserve the %errorlevel% from inside the
setlocal/endlocal block.
The setlocal/endlocal and errorlevel pattern added by commit 06fcbc47
(VS10: Fix working directory of consecutive custom commands, 2011-04-08)
does not work well in VS 7.1. Restore the original behavior for VS
versions that do not need the new behavior.
Copy the files specified in CPACK_DEBIAN_PACKAGE_CONTROL_EXTRA to the right
directory when packaging components. This fixes#12061.
Signed-off-by: Eric NOULARD <eric.noulard@gmail.com>
The parent commit assumed that "cd /d" would work in all Windows shells.
While all modern versions of windows have shells that support it, the
shells used by NMake and Borland make do not. Borland make does not
seem to even support changing drive letters with "d:". Just revert the
feature for all make tools except MinGW where the shell is known to
support this feature.
Teach cmLocalUnixMakefileGenerator3::CreateCDCommand to change working
directories for make tools using a Windows shell using "cd /d" instead
of just "cd". This tells the shell to change the current drive letter
as well as the working directory on that drive.
Commit abaa0267 (When the working directory for a custom command is on
another drive..., 2007-12-17) fixed the same problem for VS IDE
generators as reported by issue #6150.
The VS 10 msbuild tool uses a single command shell to invoke all the
custom command scripts in a project. Isolate the environment and
working directory of custom commands using setlocal/endlocal. The
form of each command is
set errlev=
setlocal
cd c:\work\dir
if %errorlevel% neq 0 goto :cmEnd
c:
if %errorlevel% neq 0 goto :cmEnd
command1 ...
if %errorlevel% neq 0 goto :cmEnd
...
commandN ...
if %errorlevel% neq 0 goto :cmEnd
:cmEnd
endlocal & set errlev=%errorlevel%
if %errlev% neq 0 goto :VCEnd
so that all changes to the environment and working directory are
isolated within the script and the return code is preserved.
Some find modules call find_package recursively to locate a package
configuration file for the package instead of searching for individual
pieces. Commit 79e9b755 (Help recursive find_package calls in modules,
2008-10-03) taught find_package to forward the version number and EXACT
arguments through the recursive call automatically. Do the same for the
component list.
When reading archive entries from disk strip any xattr and acl entry
headers that may have been loaded from the filesystem (e.g. selinux).
These fields are only useful for backup tools and not for packaging and
distribution of software. Furthermore, the GNU tar 1.15.1 on at least
one Linux distribution treats unknown entry headers as an error rather
than a warning. Therefore avoiding such fields is necessary for archive
portability.
Suggested-by: Tim Kientzle <tim@kientzle.com>
3e32db7 cmCTestUploadCommand::CheckArgumentKeyword should return false if not FILES
6b6f309 Add the FILES keyword to ctest_upload command
28cdd0a Don't tar/gz ctest_upload() files
fbe4356 Change 'Files' tag to 'Upload' in Upload.xml
350546d Implement ctest_upload command
For VS2010 if a precompiled .obj file was the output of a custom commad,
it was used as part of the build. If it was not, then VS did not
use it as part of the build. This commit updates the test to check
for this issue, and fixes the problem. This fixes bugs #0011891 and
Also allow generators to override the default packaging method.
Add a ONE_PER_GROUP option so that method can be specified by the user without relying on defaults.
All commands accepting file paths should normalize the slashes so that
the string-represented names can be compared reliably. The commands
add_library and add_executable have done this for years. We taught
add_custom_command to normalize its OUTPUT names in commit a75a0a14
(Normalize add_custom_command OUTPUT names, 2010-12-15). We handled a
special case of the DEPENDS option in commit 7befc007 (Handle trailing
slashes on add_custom_command DEPENDS, 2011-01-26).
Teach both add_custom_command and add_custom_target to normalize slashes
of DEPENDS files up front. This approach subsumes the above-mentioned
special case so remove the one line added for it but keep its test.
Extend the CustomCommand test to check that slash count mismatches
between custom command OUTPUT and DEPENDS can still be linked correctly.
A search-and-replace in commit 8d32d229 (make commands lower case by
default, 2007-10-10) accidentally changed the variable reference
CMAKE_INCLUDE_DIRECTORIES_BEFORE to CMAKE_include_directories_BEFORE.
Change it back.
The curses dialog (ccmake) allows variables to be specified on the
command line. If any of these variables is used during any configure
iteration or during generate we must not warn about it.
The Qt dialog (cmake-gui) allows variables to be added and removed in
the GUI interactively. If a variable is added, removed, and then added
again we must still warn if it is unused.
Set target property RUNTIME_OUTPUT_DIRECTORY explicitly on ProcessFwd9x
and EncodeExecutable so that we know exactly where the executables will
exist on disk.
At some point in the past VS 2010 failed some tests with custom commands when
relative paths were not used. It seems that those problems have been fixed.
However, the relative paths apparently are appended to the current working
directoy before vs accesses the file. So, with a long path, relative paths
cause it to create a combined path that is too long.
86cb17b Pass include directories with response files to GNU on Windows
9a0b9bc Optionally pass include directories with response files
6e8a67f Generate target-wide flags before individual build rules
d099546 Factor old-style -D flags out from -I flag generation
Some samples of things that got unnoticed by our nightly builds:
$ JAVA_HOME= mvn
Warning: JAVA_HOME environment variable is not set.
...
$ mvn
[INFO] Scanning for projects...
[INFO] ------------------------------------------------------------------------
[ERROR] BUILD FAILURE
...
Create platform option CMAKE_<lang>_USE_RESPONSE_FILE_FOR_INCLUDES to
enable use of response files for passing the list of include directories
to compiler command lines.
This switches the internal generation order but does not affect the
results. The new order ensures that any internal state changed by
generating target-wide flags is known when the individual rules that use
those flags are generated.
Move the GetDefineFlags call from cmLocalGenerator::GetIncludeFlags to
all call sites so that the method exclusively constructs a string of
include search path flags.
Avoid GCC warning
warning: suggest parentheses around ‘&&’ within ‘||’ [-Wparentheses]
added by commit c6a8e4c7 (The link interface of MODULE libraries is
empty, 2011-03-09).
Several major Linux distributions still do not provide Git >= 1.6.5.0 in
their stable package lists. Prior to commit 1173cc4a (Update Git
submodules with --recursive, 2011-02-22) CTest was able to use older Git
versions but simply silently failed to update submodules recursively.
Instead of failing with older Git versions preserve the status quo and
add a warning in the update log. Users testing projects with recursive
submodules may simply update to a Git new enough to support them.
MODULE libraries cannot be linked into other libraries and executables
(just like executables without ENABLE_EXPORTS). Set the MODULE target
link interface to be empty. This allows such targets to be installed in
an EXPORT set without including all their private implementation
dependencies.
Commit afd7d4ca (Add target property LINK_SEARCH_END_STATIC, 2008-01-31)
defined a property to ensure that static runtime libraries get selected.
Add a property to specify that all libraries whose type is unknown, such
as "-lm", should be assumed static. Furthermore it assumes that an
option such as "-static" is also used so that no initial -Bstatic is
needed.
1a8eed1 Merge branch 'add-STRING-subcommand-FIND-issue-11795' into fix-2828-more-info-in-script-mode
006124b Avoid direct use of std::stringstream
8a8da36 Merge branch 'fix-2828-more-info-in-script-mode' into add-STRING-subcommand-FIND-issue-11795
1462561 Add a string(FIND) sub-command (#11795)
79e13af Add XML escaping for directory name in Eclipse projects (#11658)
8074bca Fix XML escaping for target names in Eclipse project files (#11658)
cb94587 Fix XML escaping for the project() name in Eclipse projects (#11658)
Using sort results in a possibly-modified sorting
when all elements are "tied" - use stable_sort instead
to preserve the original ordering of tied elements.
The Clang compiler warns about extra parenthesis in the code
if ((form->curpage == field->page))
~ ^ ~
because the idiom is commonly used when an assignment is intended
instead of a comparison. Remove the extra enclosing layer.
Since CMake 2.6.3 add_custom_target() has a SOURCES option, this was not
yet considered in the C::B generator.
Also, now GENERATED files are not added to the project.
Alex
Now a virtual folder "Subprojects" is created, and in this virtual folder
is a linked resource for each project().
This can be considered a fix for #11657
Alex
The Intel Fortran plugin forgets to create the output directory into
which it will write a DLL's import library. Utilize the fix added by
commit f4b3bdc6 (Create an exe's implib output dir for VS, 2009-06-15)
and generalized by commit 764ac980 (Generalize exe implib dir creation
for VS, 2009-06-16). Create a pre-link rule to make the directory.
To avoid name clashes.
Before this commit, name clashes can occur when invoking gcov on a
project that uses the same file name in different directories. The
--preserve-paths flag ensures all file names have a complete path,
avoiding name clashes.
Thanks to "McBen <viertelvor12@gmx.net>" for the patch.
(Did not preserve original commit author information because
we have a push check for first and last name, and do not
accept authors with only an alias...)
Make it clear that list subcommands that modify the list
do so in the current CMake variable scope, regardless of
the original scope of the list variable involved.
RPM cannot easily 'merge' differents directory into a single RPM
with shared prefix. So more flexibility has been added to generic
CPackGenerator in order to let the specific generator chose the
local installation directory for each component.
With this commit, the --graphviz option now also generates dot files
which show which other targets depend on some target.
So, now there is
* a global dot-file which shows all targets and dependencies
* a dot file which shows on what a target Foo depends
* a dot file which shows which other targets depend on Foo
Alex
This is similar e.g. to CTEST_CUSTOM_WARNING_EXCEPTION from ctest.
GRAPHVIZ_TARGET_IGNORE_REGEX is not supported anymore.
I hope this is ok, since this was 100% undocumented and can't
break a build.
Alex
Patch from Christian Ehrlicher.
By setting GRAPHVIZ_EXTERNAL_LIBS to TRUE in CMakeGraphVizOptions.cmake
you can now exclude external libraries from the produced dot file.
I.e. then you see only the dependencies within your project.
Alex
008d116 VSResource: Avoid windres /D with quoted spaces (#11695)
8f9919d Avoid space in rc /D values for VS6 and Cygwin (#11695)
78fe97f Fix line too long KWStyle issue (#11695)
6627560 VS10: Escape double quote chars in defines for rc files (#11695)
dd2f814 Merge branch 'dev/add_test-working-directory' into dev/strict-mode
949d32c Unwatch manual variables upon removal in cmake-gui
3939032 Unwatch manual variables upon removal in ccmake
8354413 Add method to unwatch a manual variable
8ed3c85 Give a better message for unused variables
729db48 Fix ArgumentExpansion test expected results
89c2544 Checking for a definition is a usage
5625dee Don't output to stderr in the GUI
ad25a96 Merge branch 'ImprovedDotSupport2' into dev/strict-mode
c128abe Merge branch 'AddCMAKE_CURRENT_LIST_DIR' into dev/strict-mode
9bcaff0 Merge branch 'cmake-guiRememberAdvancedCheckbox' into dev/strict-mode
544d0c3 Fix expected output for WarnUninitialized test
4e3bea4 Update expected messages to new format
8e8c9e4 Don't check at destruction for usage
668e005 Use cmake::IssueMessage for warnings
88cd4c1 Use 'CMake Warning' versus 'warning' for CDash
3c3b98d Initialize the class before setting warn flags
cf8b15a Ignore files under the CMakeFiles directory
fd50f06 Don't check for unused vars at configure time
447a04c Don't warn during configure when doing everything
b97ee21 Check for unused variables at the end of generate
c18c977 When checking for variables, specify a reason
3f1121f Use a long int since Line is a long as well
2507f93 Change the failure case string to 'Unexpected'
fe390a2 Add 'ArgumentExpansion' test
8dbb209 Wrong boolean value for CLI warnings
d4ee998 Hard-code the --no-warn-unused-cli flag
a267b99 Fix line lengths
82ed104 Flag that the directories have been set
5aa535b Add argument to arg parsing to not set directories
367e5c3 Revert "Revert "When calling CMake, set the args and the cache""
ab5d4e4 Revert "When calling CMake, set the args and the cache"
9b90040 When calling CMake, set the args and the cache
fe56002 Fix long lines for KWStyle
5d30cfc Set a watch on variables added through the gui
33c63b1 Add a method to put a watch for variables
535253f Initialize the warning variables earlier
cbb286c Fix the path detection to work for top-level
62be1f7 Initialize the usage stack earlier
c6e7fab Factor out the checks for unused variables
5e41ba8 When using the API, check for Add vs. Remove
dee1976 Fix typo in VariableUnusedViaUnset test
f231ce5 Remove old false positive avoidance code
a117e02 Revert "Add test for unused warnings at the end of scope"
2c82f2b Exempt CMAKE(CURRENT|PARENT)_LIST_FILE from usage
6d7d449 Ignore CLI warnings for ABI determination
7740a73 Only return local keys that are defined
bef3aee Use the API so that warnings can be tracked
05cb0f4 Check for unused variables in the dtor
91c4c99 Add test for unused warnings at the end of scope
ca90f67 Fix detection of unused variables when setting
f7438ca Add test for unused detection via setting it
995cfb0 Don't warn if the variable wasn't defined
aefc91d Add test for usage checks via unset
a8e97f8 Remove VarRemoved code since it's been superceded
59463ef Rework CheckVariableForUnused usage
f117423 Fix line lengths to be no more than 78
e49a935 Improve unused warning logic
e01e40c Mark ARGC, ARGV*, and ARGN as used
a17aff7 Ignore CMAKE_MATCH_* variables for usage
02a114d Add method to allow variables to be marked as used
a0b0d23 CMAKE_DO_TRY_COMPILE is no longer used
ae3eff3 Fix the path used for ignoring system warnings
056b441 Fix missing case for usage of a variable
980e048 Factor out checks for unused variables
83acb0a Remove now unused variables
3801463 Use built-ins for readability and maintainability
8b52015 Push the initialize and unused states when copying
439877f Be consistent with single and double quotes
4cf1706 Add documentation for check-system-vars
b74777f Fix the spelling of the flag for warn-unused-vars
b948120 Change logic of flag to turn off cli unused checks
f047a17 Add test for uninitialized variables
75bda38 Add tests for unused command line variables
300fc15 Fix detection of system files
d784e6a Run the unused variables check on the final pass
9efc057 VariableWatch is not available when bootstrapping
2e78224 Add a missing comma to the warning message
7499700 Add a flag to warn about system files
fff9f6d Rename flags again and use variablewatch for cli
786e269 Add warn-unused to the Qt interface
636e6c4 Default to marking things as used
4ff0340 Rename find-unused to warn-unused
d7999e9 Rename strict-mode to warn-uninitialized
e141bc9 Detect unused variables
d3e8eb5 Add flags to detect unused variables
f332e14 Complete strict-mode checks for uninitialized vars
52f9637 Add method to get the local scope variables
f794d58 Make --strict-mode option, and integrate with cmake-gui
48b5b85 Add a warning when variables are used uninitialized.
cd626ea For macros make sure the FilePath points to a valid pointer in the args.
d94f9c6 Only set the property if the property was given
b6c302b Default the working dir to the current binary dir
0594287 Add more tests for WorkingDirectory for tests
To get rc defines to work in the VS10 IDE requires \" when
constructing PreprocessorDefinitions strings. This is different
than defines for cl.
Also, per-file rc defines were not being generated. Fix that, too.
During a try_compile cmGlobalGenerator::EnableLanguage uses results from
the outer project. Reject attempts to enable languages in the test
project that are not "ready" in the outer project. Mark a language as
"ready" when all its information has been loaded and we are ready to
generate build rules.
This also avoids infinite recursion introduced by commit 295b5b60 (Honor
CMAKE_USER_MAKE_RULES_OVERRIDE in try_compile, 2010-06-29) for projects
that set CMAKE_USER_MAKE_RULES_OVERRIDE to a file that uses try_compile.
The file is loaded along with the information for a given langauge so
the language is not yet "ready".
If CMAKE_MAKE_PROGRAM is set to devenv, then GenerateBuildCommand
uses it just like we used to do for VS8 and VS9. Otherwise, it
still uses MSBuild.
This will let us run the CMake test suite through devenv and make
sure all the solution and project files we generate are load-able
and build-able by the VS 2010 IDE, not just MSBuild.
Inspired-By: Robert Lenhardt
WriteCLSources should skip source files with "obj" extensions
since WriteObjSources has already written them into the vcxproj
file. Likewise, WriteGroupSources should skip source files with
"obj" extensions to avoid receiving "item ... already exists under
the filter" project-load-time error messages from Visual Studio.
If the source-file form of try_compile is given a file name with
multiple '.' characters such as "a.b.c" use only the shortest extension
to check the language. This is the expected behavior and is consistent
with normal language extension determination in the method
cmSourceFileLocation::UpdateExtension.
Set CMAKE_RUNTIME_OUTPUT_DIRECTORY explicitly in try_compile projects so
that the COPY_FILE feature knows where to look. This makes the feature
robust against CMAKE_USER_MAKE_RULES_OVERRIDE files that set variables
like CMAKE_RUNTIME_OUTPUT_DIRECTORY or EXECUTABLE_OUTPUT_PATH.
This variable was introduced to help authors override CMake's default
platform information before any of it is cached. State this clearly in
the documentation. Explicitly discourage use for other purposes.
Previously this was used only in multi-configuration generators to
choose the configuration of try_compile and try_run at their build time.
Teach CMake to honor the variable in single-configuration generators as
the CMAKE_BUILD_TYPE.
Factor out generation of SccProjectName, SccLocalPath, and SccProvider
from cmLocalVisualStudio7Generator::WriteProjectStart and call it from
cmLocalVisualStudio7Generator::WriteProjectStartFortran too.
On Windows platforms source files may contain '\' in include directives:
#include "a\b.h"
Normalize these while scanning to use forward slashes. CMake will
convert from forward slashes to the direction preferred by the native
build tools when writing the path to 'depend.make' files.
Show "<variable|string>" explicitly in if() case documentation whenever
auto-dereferencing occurs. Reference its presence from the explanation
at the bottom.
The signature of get_test_property uses argument order
test property VAR
not
test VAR property
Also document the actual behavior when the property is not found.
Some values simply cannot be escaped properly in all contexts for all
native build tools. Document known limitations after the disclaimer
that states so.
Previously the error message for code like
add_executable(myexe does_not_exist/mysrc.c)
mentioned only that "mysrc.c" is not found. Report the directory too.
Remove the example explained by the misleading phrase "CMake will treat
it as if you wrote". This was originally added by commit a73071ca
(modified the if command to address bug 9123 some, 2009-06-12). Later
related information elsewhere in the documentation was corrected and
made precise by commit cb185d93 (Fix if() command and CMP0012 OLD/NEW
behavior, 2009-10-27) but the misleading example was not corrected.
Replace the example with a correct one that more directly covers the
case that typically surprises newcomers. Avoid recommending a "correct"
way to write code because this behavior is always specific to each case.
Also update the main documentation of the behavior to be more explicit.
Previously the Xcode generator would rerun CMake only if input file
dependencies in the top-level directory changed. Teach it to depend on
input files from all directories. Other generators already do this.
Reported-by: Johan Björk <phb@spotify.com>
Explicitly state up front that the source-file form of the command links
an executable and expects a 'main' to be defined. While at it, update
the command signature documentation to use a syntax more consistent with
other commands. Also tweak some wording.
Teach the ArgumentExpansion test to expect flattened lists as has always
been the case in the CMake language. Now that the test should pass
enable the failure regex even when CMAKE_STRICT is not on. Replace the
reference to the old ArgumentExpansion test behavior in the workaround
comment in cmMakefile::TryCompile with a full inline explanation.
ce28737 Remove usage of CMAKE_CURRENT_LIST_DIR now that we have CMP0017
7db8db5 Improve documentation and messages for the new CMP0017
db44848 Prefer files from CMAKE_ROOT when including from CMAKE_ROOT
1e69c6f Merge branch 'user-policy-defaults' into policy-CMP0017
65a0a2a Merge branch 'include-command-whitespace' into policy-CMP0017
e33cbda VSMidl Test: Use correct include_directories with VS6 (#11461)
262da91 Prohibit space in HOME value for VSMidl test.
13caaa3 VS10: Finish Midl support (#11461)
Although the LOCATION property is for compatibility with CMake 2.4, the
LOCATION_<CONFIG> property is modern. However, if a project reads it
and sets location-altering properties later the behavior is undefined.
See parent commit for details.
Reading the LOCATION target property currently locks down the result and
ignores any later changes to properties that affect it. This may or may
not be expected and may or may not be the behavior in earlier versions
of CMake. The property is documented as provided only for compatibility
with CMake 2.4 and alternative interfaces are now available for all
originally envisioned use cases. We want to discourage its use without
outright deprecating it. Add documentation to explicitly state that
reading the property before other properties are set is undefined.
Previously, only strings containing "http:" qualified as
URLs when found in CPACK_NSIS_MENU_LINKS. Now, we use a
regex to detect strings beginning with any of the following:
ftp://
ftps://
http://
https://
news://
mailto:
This commit also moves the caller of CreateMenuLinks outside
the "if (cpackPackageExecutables)" block, allowing clients to
use CPACK_NSIS_MENU_LINKS without also having CPACK_PACKAGE_EXECUTABLES
defined. That bit of this commit fixes the remainder of the
issue described in http://public.kitware.com/Bug/view.php?id=7828
Also, added a set(CPACK_NSIS_MENU_LINKS ...) to the CPackComponents
test to enable verifying that all of this actually works.
CMake fails to find any registry paths on Windows 2000: according to regmon
it fails with an access denied error. I double checked all the access rights
and they are fine. After checking the access modes on MSDN I found that it
says KEY_WOW64_32KEY / KEY_WOW64_64KEY are not supported on Windows 2000.
CMake does not check if the current system supports Wow64 before applying
these flags.
This commit adds a check for IsWow64Process in kernel32.dll before adding
these flags.
Author: Axel Gembe <ago@bastart.eu.org>
Signed-off-by: Axel Gembe <ago@bastart.eu.org>
MUI_FINISHPAGE_RUN is frequently used with NSIS and provides a checkbox
on the finish page of an installer which specifies whether the specified
executable should be run when the installer exits. This commit adds support
for this setting in CPack.
NSIS installers default to assuming the executables exist in a
directory named "bin" under the installation directory. As this
isn't usual for Windows programs, the addition of this variable
allows the customization of this directory and links still to be
created correctly.
Previously both CPACK_NSIS_MUI_ICON and CPACK_NSIS_MUI_UNIICON
needed to be set for either to take effect. This commit allows
either to be set rather than requiring both as users may well
want to e.g. use a default uninstall icon but a custom install
icon.
This commit addresses all of the following:
http://public.kitware.com/Bug/view.php?id=8165http://public.kitware.com/Bug/view.php?id=10687http://public.kitware.com/Bug/view.php?id=11311http://public.kitware.com/Bug/view.php?id=11461
With this commit, the midl support for VS10 is as
complete as midl support ever was for VS9 and earlier.
The VSMidl test should run on all Visual Studio
generator based dashboards.
CMake no longer sends C++ compiler /D flag values
to the midl compiler in Visual Studio generated
projects. I think if we want to add that in the
future, we should add a way to pass midl compiler
specific flags and perhaps an optional way to add
in the C++ definitions, too. For now, not sending
them along gets past the immediate problem wherein
idl files in a CMake VS generated project just didn't
work at all.
The VSMidl test added in this commit was inspired by
the patch attached to 8165.
The test had to be modified such that it will run in
a directory whose name contains no spaces. There is an
existing bug filed against VS10's midl asking Microsoft
to fix that problem. But for now, the test added in this
commit works by copying the source directory to a location
that avoids spaces in the directory names.
Inspired-By: Robert Lenhardt
971692c Build enable_language command during bootstrap
960ace1 Add testing for windows resources for mingw/msys/cygwin and remove for watcom.
060d6e8 Add support for windres to cygwin.
b2f308c Add support for windows resources with mingw/msys.
8c7b19d Only run resource test for MSVC compilers.
753e208 Disable incremental testing for this test, it crashes vs9 linker.
16e7d4b Add flags to resource builds on vs 2010 with a test.
Since commit e4beefeb (CTest: Do not munge UTF-8 output in XML files,
2009-12-08) we validate UTF-8 encoding of build and test output as it is
written to XML files. However, in cmCTestTestHandler::CleanTestOutput
we still processed test output one byte at a time and did not recognize
multi-byte UTF-8 characters. Presence of such characters caused early
truncation.
Teach CleanTestOutput to truncate test output at the limit but without
cutting it in the middle of a multi-byte encoding. Also, stop avoiding
truncation in the middle of an XML tag like "<MyElement>" because the
'<' and '>' will be properly escaped in the generated XML anyway.
This patch makes include() and find_package() prefer cmake files
located in CMAKE_ROOT over those in CMAKE_MODULE_PATH.
This makes sure that the including file gets that file included
which it expects, i.e. the one from cmake with which it was tested.
It only changes behaviour when such an included file exists both
in CMAKE_MODULE_PATH and in CMAKE_ROOT.
This comes together with a new policy CMP0017, with default
behaviour it behaves as it always did, but warns.
With NEW behaviour it includes the file from CMAKE_ROOT
instead from CMAKE_MODULE_PATH. This fixes (if CMP0017 is set)
building KDE 4.5 with cmake >= 2.8.3.
Also a basic test for this policy in included.
Check CMAKE_POLICY_DEFAULT_CMP<NNNN> for a default when policy CMP<NNNN>
would otherwise be left unset. This allows users to set policies on the
command line when the project does not set them. One may do this to
quiet warnings or test whether a project will build with new behavior
without modifying code. There may also be cases when users want to
build an existing project release using new behavior for policies
unknown to the project at the time of the release.
The header was included only under "#ifdef __linux", but not all Linux
distributions provide the header. SystemInformation uses no symbols
from this header, so do not include it.
Commit 060d6e88 (Add support for windres to cygwin, 2010-12-23) and
commit b2f308c8 (Add support for windows resources with mingw/msys,
2010-12-22) introduced enable_language(RC) for the first time in a
platform file processed by a bootstrap-built cmake.
667a90a Fix sentence break in add_test documentation
96309fc Make TestsWorkingDirectory test a C file
a4a5e37 Use iostream to make Borland happy
cfe53cd Fully specify the path to old-signature add_test
017d4e9 Group adding tests with its properties
561cc33 Only test the default cwd with Makefiles
d87bae7 Simplify the _default_cwd derivation
992c74f Use --><-- markers to denote the path
5249551 Flip slashes around on Windows
0a014da Add ctype.h include for toupper()
af12f83 Fix header includes for C++ and Visual Studio
5597aa2 Rename the project to match the test
9bf4165 Add tests for WORKING_DIRECTORY arg to add_test
42de5d0 Add WORKING_DIRECTORY argument to add_test
7679f9f Rename WorkingDirectory test
d95f817 Add the WORKING_DIRECTORY property to tests
Pass the lpClass argument of RegCreateKeyEx as a real char[] instead of
a string literal. At least one platform declares the argument as char*
instead of "const char*".
Commit 42de5d02 (Add WORKING_DIRECTORY argument to add_test, 2010-12-16)
added a new sentence to a paragraph without separating it by " " from
the previous sentence. Add the missing spaces.
4499d50 Mark CustomCommand test perconfig.out as SYMBOLIC
f0cdb60 Introduce "generator expression" syntax to custom commands (#11209)
4749e4c Record set of targets used in cmGeneratorExpression
ef9e9de Optionally suppress errors in cmGeneratorExpression
45e1953 Factor per-config sample targets out of 'Testing' test
4091bca Factor generator expression docs out of add_test
bfb7288 Record backtrace in cmCustomCommand
In vs2010 a bad project file could be generated if a .c or .cxx file
was marked with HEADER_FILE_ONLY, if it was in a library that contained
both c and c++ code. This fixes the error in the code, and adds a test
for this case.
If a .sln file refers to a project file with a leading ".\", as in
".\foo.vcxproj" instead of just "foo.vcxproj" or a full path then
msbuild behaves strangely. Whenever target foo is built as a dependency
of another target, msbuild brings multiple configurations up to date
instead of just the requested configuration!
Avoid a leading ".\" in project file references to avoid this behavior.
This alternative fix to that attempted by commit 57e71533 (Avoid msbuild
idiosyncrasy that builds multiple configs, 2010-12-10) avoids use of
full path project file references which vcbuild does not support.
This reverts commit 57e71533f4.
While "msbuild" can handle full paths to project files in solutions,
the old "vcbuild" used for VS < 10 cannot. We will need another
way to fix issue #11594.
For prior versions of Visual Studio we would intentionally pass
"/nologo-" for "verbose makefiles" (CMAKE_VERBOSE_MAKEFILE ON)
when the caller did not already explicitly specify either /nologo
or /nologo-. And we still do. For the prior versions.
This had the side effect of always passing /nologo- for try_compile
operations because try_compile generates projects that have verbose
makefiles on.
However, starting with Visual Studio 10, the compiler emits
"cl ... warning D9035: option 'nologo-' has been deprecated"
when passed "/nologo-".
Therefore, this commit removes setting "/nologo-" for verbose
makefiles in the Visual Studio 10 case to avoid emitting a
warning for every single invocation of the compiler in a given
build.
With Visual Studio 10, we do not set this flag either way
and therefore, the generated project has no value for this
setting and gets Visual Studio's default value, which is
of course "/nologo", which does not produce a warning.
With Visual Studio 10, a caller can still force "/nologo-"
if desired by adding it explicitly to CMAKE_C_FLAGS or
CMAKE_CXX_FLAGS.
One of Cygwin's goals is to build projects using the POSIX API with no
Windows awareness. Many CMake-built projects have been written to test
for UNIX and WIN32 but not CYGWIN. The preferred behavior under Cygwin
in such projects is to take the UNIX path but not the WIN32 path.
Unfortunately this change is BACKWARDS INCOMPATIBLE for Cygwin-aware
CMake projects! Some projects that previously built under Cygwin and
are Cygwin-aware when they test for WIN32 may now behave differently.
Eventually these projects will need to be updated, but to help users
build them in the meantime we print a warning about the change in
behavior. Furthermore, one may set CMAKE_LEGACY_CYGWIN_WIN32 to request
old behavior during the transition.
Normally we avoid backwards incompatible changes, but we make an
exception in this case for a few reasons:
(1) This behavior is preferred by Cygwin's design goals.
(2) A warning provides a clear path forward for everyone who may see
incompatible behavior, and CMAKE_LEGACY_CYGWIN_WIN32 provides a
compatibility option. The warning and compatibility option both
disappear when the minimum required version of CMake in a project is
sufficiently new, so this issue will simply go away over time as
projects are updated to account for the change.
(3) The fixes required to update projects are fairly insignificant.
Furthermore, the Cygwin distribution has no releases itself so project
versions that predate said fixes tend to be difficult to build anyway.
(4) This change enables many CMake-built projects that did not
previously build under Cygwin to work out-of-the-box. From bug #10122:
"I have built over 120 different source packages with (my patched)
CMake, including most of KDE4, and have found that NOT defining
WIN32 on Cygwin is much more accurate." -- Yaakov Selkowitz
A fully compatible change would require patches on top of these project
releases for Cygwin even though they otherwise need not be aware of it.
(5) Yaakov has been maintaining a fork of CMake with this change for the
Cygwin Ports distribution. It works well in practice. By accepting the
change in upstream CMake we avoid confusion between the versions.
CMake itself builds without WIN32 defined on Cygwin. Simply disable
CMAKE_LEGACY_CYGWIN_WIN32 explicitly in our own CMakeLists.txt file.
Previously, we would search in the Windows registry for the path
to makensis, and fail immediately if we could not read the registry
value, assuming that it was simply not installed.
This change looks for makensis in the PATH even if the registry value
is not there, enabling the scenario where makensis is installed without
admin privileges and never even touches HKEY_LOCAL_MACHINE during the
non-admin install.
In VS9 and previous versions, :VCReportError is the goto label
to jump to after a failed custom command. It stops the build
before it tries to go any further.
In VS10, :VCEnd is the correct label to use.
Create a method in the VS generators to provide the correct
line of script to use for each version of Visual Studio.
For more internal details, search for VCEnd in the
C:\Program Files\MSBuild directory.
The KWSys SystemTools::GetMaximumFilePathLength method is poorly
conceived and should not be used. The cmDepends code honors its own
MaxPath buffer size. Just hard-code it.
The first instinct of a lot of users is to use file(GLOB) to assemble
lists of sources. Add a warning to the help text stating that it should
not be used for this purpose and briefly explain why.
Suggested-By: Ryan Pavlik
A cycle exists when the DFS returns to the root node, not just when
multiple paths lead to the same node.
Inspired-By: Alexander Esilevich <aesilevich@pathscale.com>
Evaluate in the COMMAND arguments of custom commands the generator
expression syntax introduced in commit d2e1f2b4 (Introduce "generator
expressions" to add_test, 2009-08-11). These expressions have a syntax
like $<TARGET_FILE:mytarget> and are evaluated during build system
generation. This syntax allows per-configuration target output files to
be referenced in custom command lines.
The target_link_libraries command requires its first argument to be a
target in the current directory. Document this and update the error
message to be more specific. While at it, format the error message with
a call stack.
Previously the OUTPUT arguments of add_custom_command were not
slash-normalized but those of add_library and add_executable were.
This caused the example
add_custom_command(OUTPUT a//b.c ...)
add_library(... a//b.c ...)
to fail at build time with "no rule to make a/b.c". Fix this and modify
the CustomCommand test to try it.
This method replaces '//' with '/' to make the paths look nicer.
Originally it correctly skipped a leading '//' in a UNC path as the
comment says. However, commit "Removed extra variable initializations"
(2005-04-15) accidentally removed the "pos=1" initializer. It was then
incorrectly restored by commit "Added missing variable initialization"
(2005-04-15) as just "pos=0". Restore the proper initializer.
The test for this added by commit "better coverage" (2006-07-31)
included incorrect output for a sample UNC-like path. Fix it.
Pass the full version to cmake_minimum_required(VERSION) in generated
CMake files for try_compile projects. This tells CMake not to use any
backward compatibility behavior while processing code it generates.
27ee50a Add ECLIPSE_CDT4_GENERATE_SOURCE_PROJECT as a ADVANCED cache variable (#9631)
50d21d4 Add cache var CMAKE_ECLIPSE_MAKE_ARGUMENTS when using the Eclipse generator
10f01ae Remove unused parameter "root" in some VS generator methods
57e7153 Avoid msbuild idiosyncrasy that builds multiple configs (#11594)
7728172 Remove unused variable "rootdir" in VS generators
Remove a boolean parameter of cmGlobalGenerator::CreateGlobalTarget that
is never set to true anymore. Remove global target "consolidation" loop
because no global targets exist before it runs anymore.
Commit e01cce28 (Allow add_dependencies() on imported targets,
2010-11-19) started using cmMakefile::FindTargetToUse to follow
dependencies, including those of GLOBAL_TARGETs like INSTALL and
PACKAGE. Since global targets exist in every directory, dependencies
between them must be traced within each directory too.
Teach FindTargetToUse to check the current directory before checking
globally. For global targets this will find the local copy. For for
normal targets this will be a no-op because they are globally unique.
This variable can be set to command line arguments which will be passed
to make when eclipse invokes make, e.g. you can enter "-j8" to get
8 parallel builds (#9930)
Alex
This authorize more control because one can set
CPACK_ARCHIVE_COMPONENT_INSTALL to ON globally
and then set it selectively to OFF inside
a CPack project config file.
Sidenote: GetOption ought to be a 'const' method.
If a .sln file refers to a project file with a leading ".\", as in
".\foo.vcxproj" instead of just "foo.vcxproj" or a full path then
msbuild behaves strangely. Whenever target foo is built as a dependency
of another target, msbuild brings multiple configurations up to date
instead of just the requested configuration!
Refer to all project files by full path to avoid this behavior.
The soname generation code was compile-time selected instead of runtime
selected. The result is that a Mac-compiled cmake used to cross-compile
Mac -> Unix generates an soname of the form libfoo.x.y.so instead of
libfoo.so.x.y as expected. Instead do a runtime check based on the
target platform.
Inspired-By: George Staikos <staikos@kde.org>
Avoid tracing dependencies of GLOBAL_TARGET targets. The build system
generators are not designed to handle any dependencies that may be
discovered. Global targets are only generated by CMake and never have
commands that reference targets built in the project anyway.
The exception is when building CMake itself there is a special case to
use the just-built "cmake" binary in the "install" target so that CMake
can replace itself on Windows. Even in this special case we do not want
to let the "install" target depend on the "cmake" target. Doing so
breaks cases like "make -j4 install".
Commit "merge in changes for beos support" (2006-12-04) added a realpath
call for every directory parsed out of a PATH-style environment
variable. No reason was given in the commit message or comments.
The call incorrectly resolves symlinks in referenced paths. Remove it.
If BeOS support really needs it then it can be restored for that
platform with a full explanation.
The cmCustomCommandGenerator::GetCommand method completely replaces the
purpose of this method. Re-implement GetRealLocation inline at the only
remaining call site and remove it.
The Makefile, VS, and Xcode generators previously duplicated some custom
command line generation code. Factor this out into a separate class
cmCustomCommandGenerator shared by all generators.
A custom command may name a target created by add_custom_target in its
DEPENDS field. Treat this case as a target-level dependency only since
a custom target provides no standard file on which to add a file-level
dependency.
Imported targets do not themselves build, but we can follow dependencies
through them to find real targets. This allows imported targets to
depend on custom targets that provide the underlying files at build
time.
Commit 60cd72d0 (Cleaned up generation of symbolic rules, 2006-02-15)
incorrectly made these Makefile targets .PHONY even though the build
rule touches an actual file. Correct it so that the copy_f90_mod and
touch steps do not happen on every "make".
Commit 34e1ac24 (Create Fortran info variables for .mod behavior,
2010-11-12) incorrectly taught GetFortranModuleDirectory to return a
relative path. We really want to use "." as the module directory only
as a workaround for compilers that do not do so by default. Therefore
we need this default only when generating the compiler command line and
not when scanning dependencies.
Revert the previous change to GetFortranModuleDirectory and apply the
change only at one of its call sites.
Commit 3827991e (...fix...not being able to remove try compile code,
2008-03-26) introduced a loop of RemoveFile attempts to overcome
anti-virus locks on recently created try_compile executables. Fix the
logic in this loop to work when the file is already missing.
The Express Edition does not come with 64-bit tools, but one can install
the "Microsoft Windows SDK v7.1" to get them. Detect this case and
check for the SDK. If found, set PlatformToolset to use the SDK tools.
Otherwise, fail with a concise and informative error.
Commit 1a0c166 (Store direct dependencies in solutions for VS >= 8,
2010-08-20) disabled use of VS-specific global dependency analysis.
Avoid perfoming the analysis at all when it is not needed. This also
prevents creation of bogus and unused '_UTILITY' targets since they are
not needed for dependencies.
Previously the command
find_path(VAR DOC "")
would crash because the argument pre-processing removed the DOC ""
arguments but the rest of the parsing assumes at least 2 arguments.
Reject the call with an error instead.
8b143fa Condense parsing of cache entries
122ebf1 Support manual cache entries
90abc3a Use cmCacheManager to load entries from the cache
6fe8624 Fix parsing of cache variables without a type
e01cce2 Allow add_dependencies() on imported targets (#10395)
bc7395c Merge branch 'vs-target-dependencies' into imported-target-dependencies
fd614e6 Use modern global dependency graph for VS < 8 deps
605f4bc Record edge type in global dependency graph
82596fc Merge branch 'vs8-direct-depends' into vs-target-dependencies
78c86f4 Exclude targets from the graphviz file based on a regex
5ea1e4c Collect targets and libs on demand instead of in the ctor
2a5790a Use std::cout instead of fprintf
7ba2d36 Enable/disable generating graphs depending on the target type
84ce612 Move the code for generating dot-files into separate class cmGraphVizWriter
a60b099 Generate separate dot files for each target, and a big one with everything.
487bd57 Properly insert all targets, also those which don't link to anything.
de2b2bf Move the code for collecting targets and libraries into separate functions
f7d56df Remove trailing whitespace and minor formatting changes for the dot-code
These mainly come from the command line or manual entries in the
CMakeCache.txt file. We want to stop at the first '=' because this is
what is most likely to have been meant. The variable can be quoted if
the '=' is intended.
Caveat: What if one wants both '"' and '=' in a variable name?
Imported targets do not themselves build, but we can follow dependencies
through them to find real targets. This allows imported targets to
depend on custom targets that provide the underlying files at build
time.
VS 7.1 and below have 2 behaviors that make the cmComputeTargetDepends
result difficult to use for solution-level dependencies. Update the
method cmGlobalVisualStudioGenerator::ComputeTargetDepends to document
the behaviors and work around them. Commit 1a0c166a (Store direct
dependencies in solutions for VS >= 8, 2010-08-20) isolated VS >= 8 from
this computation so those versions should be unaffected.
This change removes the last use of cmTarget::GetLinkLibraries for
purposes other than backward compatibility with legacy interfaces
(export_library_dependencies, VS 6 custom .dsp templates). Now the
cmComputeTargetDepends results are used for all generators so global
target dependency computation is fully centralized.
In cmComputeLinkInformation we match library names with a regular
expression, possibly extracting the 'lib' prefix. The regex component
to match the prefix always allows an empty prefix to be matched, as in
"(lib|)". Avoid every adding an empty prefix option earlier in the
regex, as in "(|lib|)", because it will be preferred and 'lib' will
never match.
156ae39 Don't disable colors in the CodeBlocks generator and minor cleanup.
e547fa7 Remove the "early alpha stage" comments about Eclipse and C::B
a02987a Don't enforce VERBOSE makefiles for the CodeBlocks generator
7b1421b Remove trailing whitespace
This commit adds support for a GRAPHVIZ_TARGET_IGNORE_REGEX variable
which can be set() in CMakeGraphVizOptions.cmake.
Targets matching this regex will be skipped when generating the graphviz
graphs.
Alex
In CMakeGraphVizOptions.cmake you can now set GRAPHVIZ_EXECUTABLES,
GRAPHVIZ_STATIC_LIBS, GRAPHVIZ_SHARED_LIBS and GRAPHVIZ_MODULE_LIBS
to TRUE or FALSE depending on whether you want graphs for the
targets of the respective types.
Alex
Delay computation of the command documentation until it is needed.
It is wasteful to do it in the constructor on every call.
Inspired-By: Christian Ehrlicher <Ch.Ehrlicher@gmx.de>
Define CMAKE_Fortran_MODDIR_DEFAULT and CMAKE_Fortran_MODOUT_FLAG
variables to help some Fortran compilers generate .mod files in the
current working directory.
Instead of enforcing verbose makefile, now the generated build command
includes "VERBOSE=1" so the output will be verbose when building in
C::B.
Also removed the now unused setForceVerboseMakefiles().
Alex
Look for the "/threads", "/libs:dll", and "/dbglibs" flags and convert
them to the proper RuntimeLibrary attribute value in the IDE. This is
a 3-to-1 flag mapping and such needs special handling in the parser.
The SubSystem attribute value must be "subSystemConsole" or
"subSystemWindows", not "1" or "2". Commit 20f49730 (Reset
platform/compiler info status for each language, 2010-09-28) exposed
this bug by (correctly) passing the /libs:dll flag to the compiler,
which chokes the linker if a value for "/subsystem:" is not given.
The behavior of add_custom_command when no DEPENDS option is specified
matches that of standard Make behavior, but it does not hurt to describe
it explicitly.
Rolf Eike Beer
KWSys Robot
Jean-Christophe Fillion-Robin
David Cole
Bernhard Walle
Stephen Kelly
Brad King
Nicolas Despres
Alex Neundorf
Johan Bjork
Niels Dekker
Clinton Stimpson
Peter Collingbourne
Bill Hoffman
Raphael Kubo da Costa
Matt McCormick
Eric NOULARD
Johannes Stallkamp
Thomas Jarosch
Steven Velez
Modestas Vainius
Matej Hribernik
Sean McBride
Daniel R. Gomez
Rolf Eike Beer
Zach Mullen
Wesley Turner
Manuel Klimek
Martin Konrad
Tim Hütz
Marco Craveiro
Kovarththanan Rajaratnam
Sebastian Herbst
Philip Lowman
James Bigler
Ben Boeckel
Richard Bateman
Chuck Atkins
Mike McQuaid
Brian Bassett
Craig Scott
Campbell Barton