GitLab

Use "file:///..." instead of "file:////..." when the file system path
starts in a slash.  Commit 0916cc88 (CTest.UpdateHG: Fix repo URL for
local filesystem, 2012-03-02) added a third slash after "file://"
unconditionally.  This worked for many file systems but not on Cygwin
where "file:////cygdrive/..." looks like "file://" followed by a network
file path "//cygdrive/...".  Add the slash only if the file system path
does not already start with one.

Use "file:///c:/" instead of "file://c:/" because the latter
looks like a machine:port URL instead of a local file URL.

Alexander Neundorf authored Feb 16, 2012 and Brad King committed Feb 21, 2012

Add the function cmake_expand_imported_targets() to expand imported
targets in a list of libraries into their on-disk file names for a
particular configuration.  Adapt the implementation from KDE's
HANDLE_IMPORTED_TARGETS_IN_CMAKE_REQUIRED_LIBRARIES which has been in
use for over 2 years.  Call the function from all the Check*.cmake
macros to handle imported targets named in CMAKE_REQUIRED_LIBRARIES.

Alex

For all current build machines the modules FindPkgConfig, FindFreetype, and
FindLibXslt return a version number. Enforce this to early catch when this
is not always the case.

This did not work because find_library() did only treat the given name as
complete filename if is matched "PREFIX.*SUFFIX":

find_library(MYLIB libfoo.so.2)

Now it is also taken as a whole if the name matches "PREFIX.*SUFFIX\..*".

Ancient Gnuplot versions like 3.7.1 don't know about "--version". Disable this
check so it doesn't break on machines with this version.

We do not yet support "make install" in the external project case.
Document this explicitly in the interface.  Require the caller to use an
option to "disable" the unsupported behavior.  This will allow us to add
the behavior by default in the future without clobbering existing
projects that handle the installation themselves.

Bill Hoffman authored Jan 27, 2012 and Brad King committed Feb 09, 2012

cmake_add_fortran_directory uses imported targets when using the
mingw fortran compiler. This change makes those targets global
in scope so they act just like the real targets that exist when
a fortran compiler exists and regular add_subdirectory is used.

Bill Hoffman authored Dec 16, 2011 and Brad King committed Feb 09, 2012

The ifort compiler found on some test machines does not support Mac
universal binaries or the Linux Standard Base.

Bill Hoffman authored Dec 16, 2011 and Brad King committed Feb 09, 2012

The SunPro compiler does not add the fortran runtime library when
creating a shared fortran library.  Link to the SunPro Fortran runtime
libraries explicitly.

Fix the implementation to allow full paths with spaces.  Change the
interpretation of relative paths to be with respect to the current binary
directory.  This matches the convention used in ExternalProject.  Test
both full and relative paths in the VSGNUFortran test.

Bill Hoffman authored Dec 09, 2011 and Brad King committed Feb 09, 2012

This patch adds a new module that allows for easy integration of MinGW
gfortran and the Visual Studio compiler. It is done in a function called
cmake_add_fortran_subdirectory. The patch also includes a test for this
feature.

Define a "check_language(<lang>)" macro to test whether <lang> can be
enabled.  Cache the result in CMAKE_<lang>_COMPILER.  Add a test case
covering expected results.

We have not run tests with the "como" compiler or enabled the
experimental ConvLibrary test for years.

If CXX or Fortran is enabled before C then the values of

  CMAKE_SHARED_MODULE_C_FLAGS
  CMAKE_SHARED_MODULE_CREATE_C_FLAGS

may not be available.  On platforms where MODULE library (plugin) creation
is the same as SHARED library creation initialize the MODULE creation
flags from the SHARED creation flags of the matching language instead of
assuming that C has been enabled first.

Teach the COnly and CxxOnly tests to build MODULE libraries.  The latter
covers this specific case.

FindPackageHandleStandardArgs and FindPackageMessage match the glob
expression but are nothing that will usually be fed into find_package().

Also adds test to verify this.

This removes some useless checking. The results of these things are never
properly checked so they should not count as testcases. At the end they only
needlessly clutter the output.

This allows easy spotting of modules that output crappy messages and the
like.

Now that we think that CheckSymbolExists really works for all cases it is time
to prove that. If this code fails too many other things will break.

Since commit 183b9509 (Follow all dependencies of shared library private
dependencies, 2011-12-14) we honor LINK_INTERFACE_LIBRARIES when
following dependent shared libraries.  The link interface properties may
form a cycle if set incorrectly by a project.  Furthermore, the property
LINK_DEPENDENT_LIBRARIES may form a cycle if set incorrectly by hand
(though CMake should never generate one).  In either case, do not follow
the cycle forever when following the dependent shared library closure.
We only need to add dependency edges to the constraint graph once.

Add "LinkInterfaceLoop" test to cover this case.

Some tests only need to run CMake to configure and generate a build
tree, but not actually perform the build.  Add a new "Tests/CMakeOnly"
directory dedicated for this purpose.  Add a helper script to drive each
test by creating a fresh build tree and running CMake on it.  Add macro
"add_CMakeOnly_test" to help create tests using the script.

Now that the Complex tests do not depend on cmSystemTools or other
classes from CMakeLib the COMPLEX_TEST_CMAKELIB option is useless.

Use C standard 'remove' and POSIX standard 'stat'.

This test belongs in the CMakeLibTests test driver executable which
correctly links to CMakeLib.