Commit Graph

25920 Commits

Author SHA1 Message Date
Brad King
3106cf4e3d include_external_msproject: Restore support for EXCLUDE_FROM_ALL
In commit dc6888573d (Pass EXCLUDE_FROM_ALL from directory to targets,
2019-01-15, v3.14.0-rc1~83^2) all `AddNewTarget` call sites were updated
to copy the directory-level `EXCLUDE_FROM_ALL` into the target property
of the same name, except that the one for `include_external_msproject`
was incorrectly missed.  Add it now.

Furthermore, refactoring in commit b99129d2d8 (ENH: some code cleanup,
2007-03-12, v2.6.0~2020) accidentally set the `EXCLUDE_FROM_ALL` target
property of `include_external_msproject`-generated targets to `FALSE`
instead of simply leaving it unset.  This was not necessary but had no
effect until the above commit gave it a meaning.  Drop that.

Fixes: #18986
2019-02-26 10:20:01 -05:00
Brad King
6ebe40c6ae Merge branch 'remove-exclude-from-all-warning' into release-3.14
Merge-request: !3002
2019-02-22 11:19:09 -05:00
Craig Scott
d3b765d451 EXCLUDE_FROM_ALL: Don't warn if installing target excluded from all
The original warning pre-dates support for install components.
There are now legitimate scenarios where an install(TARGETS)
command may list a target that is excluded from all, e.g.
hierarchical projects that will never install the component such a
target belongs to.

Fixes: #18938
2019-02-22 07:55:56 +11:00
Brad King
c925166764 Merge branch 'link-options' into release-3.14
Merge-request: !2996
2019-02-20 14:32:37 -05:00
Marc Chevrier
98a2d42301 PIE link options: No warning when policy CMP0083 is not set.
Fixes: #18955
2019-02-20 14:30:57 -05:00
Brad King
c36bf6b13c Merge branch 'fix-implicit-includes-fortran' into release-3.14
Merge-request: !2994
2019-02-20 08:57:20 -05:00
Brad King
3dc81a48ff Fortran: Do not suppress explicit use of implicit include directories
Since commit 2e91627dea (ParseImplicitIncludeInfo: add Fortran implicit
include handling, 2019-01-25, v3.14.0-rc1~73^2) we actually populate
`CMAKE_Fortran_IMPLICIT_INCLUDE_DIRECTORIES` for the first time.  This
value may be useful to project code to pass to other tooling that wants
to preprocess the way Fortran does, so we should compute the value.
However, compilers like `gfortran` do not actually search their own
implicit include directories for `.mod` files.  The directories must be
passed via `-I` in order for `.mod` files in them to be found.

Since Fortran has no standard library header files that we need to avoid
overriding, it is safe to *not* filter out implicit include directories
from those passed explicitly via `-I` options.  Skip this filtering so
that include directories specified by project code to find `.mod` files
will be searched by the compiler even if they happen to be implicitly
searched by the preprocessor.

Fixes: #18914
2019-02-20 08:13:57 -05:00
Brad King
0794dd3018 Merge branch 'autogen_cache_binary_checks' into release-3.14
Merge-request: !2990
2019-02-19 13:29:48 -05:00
Sebastian Holtermann
b5befac154 Autogen: Use output caching GetExecutableTestOutput
Use the output caching cmQtAutoGenGlobalInitializer::GetExecutableTestOutput
method to avoid identical calls to moc, uic and rcc.

Closes #18947
2019-02-19 13:11:44 -05:00
Sebastian Holtermann
a4e01d6707 Autogen: Add output caching GetExecutableTestOutput
This adds the cmQtAutoGenGlobalInitializer::GetExecutableTestOutput method
which caches the output of the called executable and returns the cached value
on any subsequent call.
2019-02-19 13:11:44 -05:00
Brad King
4f5bb64c56 Merge branch 'vs-win-sdk' into release-3.14
Merge-request: !2989
2019-02-19 09:33:34 -05:00
Brad King
4dab8e69bd VS: Tell VS 2019 to use Windows SDK 8.1 explicitly when needed
VS 2019 does not default to the 8.1 SDK as VS 2017 and VS 2015 did.
When `CMAKE_SYSTEM_VERSION` is 8.1 or lower, select the 8.1 SDK
explicitly.

Fixes: #18927
2019-02-19 09:31:50 -05:00
Brad King
35bf9ded3b VS: Factor out a method to set the Windows SDK version internally 2019-02-19 09:31:50 -05:00
Brad King
13f020198f Merge branch 'fix-legacy-implicit-includes' into release-3.14
Merge-request: !2981
2019-02-19 07:50:07 -05:00
Brad King
890bae524c Do not explicitly report "standard" include directories as implicit
In commit 1293ed8507 (ParseImplicitIncludeInfo: keep implicit incl.
consistent when rerunning cmake, 2019-01-30, v3.14.0-rc1~26^2) we did
not account for `CMAKE_<LANG>_STANDARD_INCLUDE_DIRECTORIES`.  This
variable lets platform modules or toolchain files specify directories
that are to be explicitly passed as standard include directories.  These
include directories are used by the test project from which we extract
implicit include directories so they appear in the parsed results
whether or not the compiler really considers them implicit.  Exclude
these entries from the computed implicit include directories since they
are not actually implied by the compiler when we invoke it with
"standard" include directories passed explicitly.

Instead teach the build system generators to treat the "standard"
directories as implicit for purposes of excluding them from appearing
earlier in the compiler command line due to `include_directories` and
`target_include_directories` calls.

Issue: #18936, #18944
2019-02-18 17:12:14 -05:00
Brad King
5c171ca898 Restore unconditional use of "standard" include directories
`CMAKE_<LANG>_STANDARD_INCLUDE_DIRECTORIES` is meant to unconditionally
add explicitly specified include directories to compile lines.  In
commit 5f34bdc7f9 (cmLocalGenerator: Refactor
`GetIncludeDirectoriesImplicit` method, 2019-01-25, v3.14.0-rc1~65^2~1)
a condition was accidentally added to exclude implicit include
directories.  Drop that condition.

Fixes: #18936
2019-02-18 17:10:07 -05:00
Brad King
9502276f82 Prefix implicit include directories with sysroot on construction
Since commit 7cd65c97fa (Add CMAKE_SYSROOT variable to set --sysroot
when cross compiling., 2013-04-13, v3.0.0-rc1~342^2) we have prefixed
the value of `CMAKE_SYSROOT` to implicit include directories.  This was
done because we hard-coded `/usr/include` as an implicit include
directory without accounting for the sysroot.  Instead we should prefix
the hard-coded paths when they are constructed.  Update the
`Platform/UnixPaths` module to do this as `Platform/Darwin` already
does.

Since commit 5990ecb741 (Compute implicit include directories from
compiler output, 2018-12-07, v3.14.0-rc1~108^2) the values of the
`CMAKE_<LANG>_IMPLICIT_INCLUDE_DIRECTORIES` variables are computed from
a real compiler invocation so they already account for the sysroot
prefix.  In commit 6fc3382944 (Update logic for sysroot in detected
implicit include directories, 2019-02-13, v3.14.0-rc2~6^2) we attempted
to apply the prefix conditionally, but that is incorrect because the
compiler's real implicit include directories are not all under the
sysroot.  Instead assume that all implicit include directories in
`CMAKE_<LANG>_IMPLICIT_INCLUDE_DIRECTORIES` already have the sysroot
prefix if needed.  Code that constructs the value must be responsible
for that because it is the only place that knows.
2019-02-18 17:09:26 -05:00
Brad King
e6897c72e7 CMake 3.14.0-rc2 2019-02-15 07:46:43 -05:00
Brad King
113a395ee7 Merge branch 'try_compile-expand-compile-defs' into release-3.14
Merge-request: !2965
2019-02-15 06:47:59 -05:00
Brad King
cde2596a19 try_compile: Restore expansion of ;-list in COMPILE_DEFINITIONS
The quoting added by commit 8c5221fb1f (try_compile: Preserve special
characters in COMPILE_DEFINITIONS, 2019-01-21, v3.14.0-rc1~108^2~3)
broke the case that the `COMPILE_DEFINITIONS` value contains a `;`.
Without the quoting the `;` would be generated literally in an unquoted
argument in the test `CMakeLists.txt` file and would then be expanded.
With quoting the `;` is preserved, which is not the old behavior.

Fix this by expanding the `;`-list ahead of time.  Add test cases for
behavior with both `#` and `;`.

This was noticed with the PGI compiler where we set
`CMAKE_CXX*_STANDARD_COMPILE_OPTION` to values like `--c++17;-A`.  The
symptom had also been observed while preparing commit ef8f237686
(ParseImplicitIncludeInfo: add SunPro Fortran and PGI compiler, Cray
fix, 2019-01-29, v3.14.0-rc1~26^2~2) but was not recognized at the time
as a regression.  Revert the workaround added by that commit.

Fixes: #18919
2019-02-15 06:21:19 -05:00
Brad King
9d3710affe Merge branch 'vs2017-sdk81' into release-3.14
Merge-request: !2962
2019-02-14 11:24:16 -05:00
Brad King
2f51f281a8 Merge branch 'fortran-submodule-names' into release-3.14
Merge-request: !2958
2019-02-14 10:24:43 -05:00
Brad King
d80ecba5c2 Fortran: Fix submodule file names across compilers
The naming convention for submodule files varies across compilers.  Add
a table to the compiler information modules and thread the information
through to the Fortran module dependency parser.  Fill out the table for
compiler ids known to support Fortran submodules.

Fixes: #18746
2019-02-14 10:23:02 -05:00
Brad King
72057d9c15 Fortran: Thread compiler id through to internal Fortran parser 2019-02-14 10:23:02 -05:00
Brad King
7ae329e2ed Fortran: Factor out .mod and .smod file name construction 2019-02-14 10:23:02 -05:00
Brad King
5c26e3c5e3 VS: Fix validation of Windows 8.1 SDK
The check added by commit 0a29a31161 (VS2017: Verify Windows 8.1 SDK
before using it, 2017-04-25, v3.8.1~2^2) used the wrong path to
`windows.h` within the SDK, leading to it never being detected.

Fixes: #18923
2019-02-14 08:25:26 -05:00
Brad King
9f351b93ab Merge branch 'fix-pointer-truncation' into release-3.14
Merge-request: !2956
2019-02-13 10:28:02 -05:00
Zsolt Parragi
76a5ac2100 cmListFileLexer: Add missing include to avoid possible pointer truncation
The `cmsys/Enconding.h` include had a typo in its surrounding ifdef,
possibly causing a missing function declaration (`cmsysEncoding_DupToWide`).
As this is C code, this resulted in the code compiling, but with a truncated
return value, possibly causing crashes.
2019-02-13 10:22:59 -05:00
Brad King
30783addc1 Merge branch 'fix-legacy-implicit-includes' into release-3.14
Merge-request: !2957
2019-02-13 10:05:48 -05:00
Brad King
6fc3382944 Update logic for sysroot in detected implicit include directories
Since commit 5990ecb741 (Compute implicit include directories from
compiler output, 2018-12-07, v3.14.0-rc1~108^2) the values of the
`CMAKE_<LANG>_IMPLICIT_INCLUDE_DIRECTORIES` variables are computed from
a real compiler invocation.  In this case the paths under the sysroot
should already have the sysroot prefix so we should no longer have to
add the sysroot prefix.  However, it is also possible for project code
to add its own paths to `CMAKE_<LANG>_IMPLICIT_INCLUDE_DIRECTORIES`
without the sysroot prefix and expect the historical addition of the
sysroot prefix to be preserved.

Try to account for both cases by conditionally adding the sysroot prefix
on implicit include directories that do not already have it.
2019-02-13 10:03:56 -05:00
Brad King
2ad14ef4ea cmAlgorithms: Add cmHasPrefix to match existing cmHasSuffix 2019-02-13 09:52:20 -05:00
Brad King
557b2d6e65 Fix regression in -I/usr/include exclusion logic
The change in commit 15ad830062 (Refactor exclusion of -I/usr/include to
avoid per-language values, 2019-01-21, v3.14.0-rc1~108^2~4) caused the
exclusion to apply to Fortran, but it was only meant for C, CXX, and
CUDA.  The purpose of the change was to prepare for the value of
`CMAKE_<LANG>_IMPLICIT_INCLUDE_DIRECTORIES` to be computed from the
actual compiler instead of hard-coded.  We need to preserve exclusion of
`-I/usr/include` if the compiler has any implicit include directory that
looks intended to replace it, e.g. `<sdk>/usr/include` on macOS.

Fixes: #18914
2019-02-13 09:52:20 -05:00
Brad King
1c4570e22e Merge branch 'genex-GENEX_EVAL-fix-recursion' into release-3.14
Merge-request: !2954
2019-02-13 08:37:38 -05:00
Marc Chevrier
e429e9af42 genex: Fix erroneous handling of recursion for $<GENEX_EVAL:>
Fixes: #18894
2019-02-13 08:37:02 -05:00
Brad King
88031dd09a Merge branch 'ghs-updates' into release-3.14
Merge-request: !2943
2019-02-08 14:02:14 -05:00
Brad King
3b58b647ae Merge branch 'fix-exclude-dir-with-iface' into release-3.14
Merge-request: !2937
2019-02-08 14:02:03 -05:00
Fred Baksik
bcd99fac60 GHS: Document usage of GHS_NO_SOURCE_GROUP_FILE
-- Also change variable name to CMAKE_GHS_NO_SOURCE_GROUP_FILE
2019-02-08 13:07:00 -05:00
Brad King
f87e724e8c Fix EXCLUDE_FROM_ALL on directory with an interface library
Since commit dc6888573d (Pass EXCLUDE_FROM_ALL from directory to
targets, 2019-01-15, v3.14.0-rc1~83^2) we automatically forward the
`EXCLUDE_FROM_ALL` to targets as they are created.  This regressed
support for interface libraries on which the property is not allowed.

Skip forwarding the `EXCLUDE_FROM_ALL` property for interface libraries.
It is not needed on them because they do not participate in the
generated build system anyway.

Fixes: #18896
2019-02-08 08:17:48 -05:00
Brad King
f657bcbd48 Merge branch 'xcode-stdlib-flags' into release-3.14
Merge-request: !2919
2019-02-07 06:44:19 -05:00
Gregor Jasny
8af334f5ba Xcode: Derive stdlib from CXX flags
Closes: #18396
2019-02-07 06:43:51 -05:00
Brad King
f835526d8f Merge branch 'vs-csharp-nowarn-numbers' into release-3.14
Merge-request: !2923
2019-02-07 06:41:22 -05:00
Wil Stark
822697996e VS: Fix nowarn compiler option to accept warning numbers.
Warning disables are transferred to the VS IDE `<NoWarn>` node.

Fixes: #18878
2019-02-07 06:39:45 -05:00
Brad King
bf1c2a2593 CMake 3.14.0-rc1 version update 2019-02-06 10:08:35 -05:00
Brad King
d6729505cb Merge topic 'relax_CUDA_RESOLVE_DEVICE_SYMBOLS_constraints'
850ef90a66 CUDA: Honor CUDA_RESOLVE_DEVICE_SYMBOLS for more target types

Acked-by: Kitware Robot <kwrobot@kitware.com>
Acked-by: Robert Maynard <robert.maynard@kitware.com>
Merge-request: !2900
2019-02-06 06:51:30 -05:00
Brad King
4400ac07f7 Merge topic 'autogen_same_name_different_extension'
a28caabf45 Autogen: AUTOMOC support for files with the same name but different extensions
1f802295f2 Autogen: Extend the SameName test with same name but different extension files

Acked-by: Kitware Robot <kwrobot@kitware.com>
Merge-request: !2920
2019-02-06 06:50:36 -05:00
Kitware Robot
3a2387bce7 CMake Nightly Date Stamp 2019-02-06 00:01:07 -05:00
Robert Maynard
850ef90a66 CUDA: Honor CUDA_RESOLVE_DEVICE_SYMBOLS for more target types
`CUDA_RESOLVE_DEVICE_SYMBOLS` can be used with shared, module, and
executable target types.  This relaxation is to allow for better
interoperability with linkers that automatically do CUDA device symbol
resolution and have no way to disable it.
2019-02-05 11:09:48 -05:00
Sebastian Holtermann
a28caabf45 Autogen: AUTOMOC support for files with the same name but different extensions
This adds support for AUTOMOC to moc header files with the same but different
extensions (e.g `obj.h`, `obj.hpp`, `obj.hxx`).  If a moc file would appear
multiple times in `mocs_compilation.cpp`, a number suffix is appended
to the name to make it unique.

Closes #14489
2019-02-05 15:51:53 +01:00
Brad King
f47d980f73 Merge topic 'update-kwsys'
9ddbf4c6e6 Merge branch 'upstream-KWSys' into update-kwsys
8372c05472 KWSys 2019-02-05 (65802a20)

Acked-by: Kitware Robot <kwrobot@kitware.com>
Merge-request: !2918
2019-02-05 07:49:07 -05:00
Brad King
ab01ce91ee Merge topic 'autogen_single_all_sources_iteration'
b6cf086267 Autogen: Iterate over makefile sources only once
ed0fa784eb cmSystemTools: Let `GetFileFormat` accept a `std::stding const&`

Acked-by: Kitware Robot <kwrobot@kitware.com>
Merge-request: !2909
2019-02-05 07:47:14 -05:00