Previously the command did not allow naming targets on the LHS that
were not created in the calling directory. Lift this restriction to
enable more flexible use by projects.
Targets named on the RHS will need to be looked up during generation in
the scope of the call site rather than the scope of the LHS target.
Introduce an internal syntax in `[INTERFACE_]LINK_LIBRARIES` properties
to specify target names that need to be looked up in a directory other
than that containing the target on which the property is set. Add
minimal documentation of the syntax to help users that encounter it.
Unfortunately CMake previously did allow such calls in the case that
only `INTERFACE` libraries are specified, but those libraries would be
looked up in the target's directory rather than the caller's. Add
policy `CMP0079` to enable the new behavior with new lookup scope in a
compatible way.
Fixes: #17943
The `AddLinkLibrary` method takes a `cmMakefile` pointer to represent
the scope of the caller that wants to link to the named library.
Currently in all call sites this is the same as the target's `Makefile`
member, but in principle the library named by the caller is visible in
its scope so we should use the `cmMakefile` it provided to look up the
library target.
While collecting usage requirements from the `INTERFACE_*` properties of
directly linked targets, we internally generate `TARGET_PROPERTY:` and
`TARGET_OBJECTS:` generator expressions to refer to those properties on
those targets. At the point we generate these expressions we already
have a pointer to an exact `cmGeneratorTarget` instance.
Switch from using the target name in these generator expressions to
using an internal unique name generated for each `cmGeneratorTarget`
instance to be referenced. This avoids depending on the user-facing
target name to find the same target we already have.
The CTestTestStopTime test has been failing sporadically because the
stop time causes the first internal test to have a timeout short enough
that we might hit it and start the second test just before the stop time
is reached. Instead we should track when a timeout is shortened in
order to stay within the stop time. If a test times out for this reason
then we should consider the stop time reached and not start any more
tests.
On filesystems with 1s resolution the `run_sub_cmake` cases fail
occasionally when the 1 second sleep does not cause files to have
a different time. Use 3 seconds instead.
In commit v3.9.0-rc1~55^2 (Improve Doxygen support, 2017-04-10)
use of the `IN_LIST` condition was added, but this is only supported
when policy CMP0057 is set to NEW. Add a policy scope around the
module and enable the policy within it. Otherwise it works only
in projects that happen to enable the policy.
Fixes: #18361
The Ninja and Makefile generators were adding a space to the value both
before and after expanding rule variables. Only the latter is needed.
While at it, revise some outdated comments since the rule variable
expansion is no longer responsible for inserting the launcher.
Suggested-by: Mate Pek <mate.pek@gmail.com>
Fixes: #18340
Use the normal target link interface computation logic to get the
list of libraries in the link interface instead of trying to
interpret the `INTERFACE_LINK_LIBRARIES` property directly.
The IAR 6.50.6 compiler places extra/truncated copies of the
compiler id `INFO:` strings into binaries with a prefix like
`?<Constant "`. Teach CMakeDetermineCompilerId to ignore them.
Fixes: #18333
31c82143bf GoogleTest: gtest_add_tests() fails if any source file is empty
d6b06d8d87 GoogleTest: Modify test to verify that empty files can be scanned
Acked-by: Kitware Robot <kwrobot@kitware.com>
Merge-request: !2342
Refactoring in commit v3.10.0-rc1~103^2~2 (FindMPI: Modernization from
ground up, 2017-04-25) accidentally left this variable set as a copy of
the `;`-list in `MPI_<LANG>_COMPILE_OPTIONS`, but the flags variable is
documented as a command-line string. Restore it now.
Fixes: #18349
The format for the TAG file was changed in 3.12, and the way it was
read caused a regression which changed how the track was decided. This
commit fixes the regression.
Fixes#18347.
If these are not detected then we cannot support the IAR compiler.
Fail early with an explicit message instead of silently proceeding
and getting strange errors.
Issue: #18333
Brad King
Roger Leigh
Craig Scott
Marc Chevrier
Kitware Robot
Daniel Schürmann
Kyle Edwards
Vitaly Stakhovsky
Brian Heim
Raul Tambre