From the CMake 3.27 release notes:
Compatibility with versions of CMake older than 3.5 is now
deprecated and will be removed from a future version. Calls to
cmake_minimum_required() or cmake_policy() that set the policy
version to an older value now issue a deprecation diagnostic.
This PR also removes manually setting policy CMP0048. This is
redundant since the CMake min is already 3.X
This is useful for running individual tests in a separate process, which is
useful for testing e.g. flag changes which have a process-global effect.
PiperOrigin-RevId: 540580573
Change-Id: I18a5d24d79425a9d595be3369efc44e2f655f6f8
This is populated by the clangd language service provider.
Also ignore the cmake-variants.yaml file.
This can be created locally to select various build/test configurations.
Fixes#4266
Note that RE2 has taken a dependency on Abseil, so the `main` branch
should be used from now on. The `abseil` branch will go away soon...
PiperOrigin-RevId: 536737543
Change-Id: I0e595ee8e754b924881f997f286cdec3822763bf
Use /fp:precise to override IntelLLVM's default of /fp:fast.
This makes IsInf and IsNan work as expected by googletest.
```
[build]...\googletest\googlemock\test\gmock-function-mocker_test.cc(143,21): error: 'VoidReturning' overrides a member function but is not marked 'override' [-Werror,-Winconsistent-missing-override]
[build] MOCK_METHOD(void, VoidReturning, (int n)); // NOLINT
```
```
[build] ...\googletest\googlemock\test\gmock-function-mocker_test.cc(182,20): error: exception specification of overriding function is more lax than base version [-Werror,-Wmicrosoft-exception-spec]
[build] MOCK_METHOD(int, CTNullary, (), (Calltype(STDMETHODCALLTYPE)));
```
```
[build] ...\googletest\googletest\test\googletest-death-test-test.cc(209,5): error: unused function 'DieInCRTDebugElse12' [-Werror,-Wunused-function]
[build] int DieInCRTDebugElse12(int* sideeffect) {
[build] ^
```
```
[build] ...\googletest\googletest\test\gtest_unittest.cc(4096,7): error: variable 'n' set but not used [-Werror,-Wunused-but-set-variable]
[build] int n = 0;
```
Use the same flags as Clang if the compiler id is IntelLLVM.
IntelLLVM warns if a double constant is assigned to a float.
```
[build] .../googletest/googletest/include/gtest/gtest-printers.h:516:17: warning: implicit conversion between floating point types of different sizes [-Wimplicit-float-size-conversion]
[build] mulfor6 = 1e1;
```
IntelLLVM uses fp-model=fast by default, breaking IsNan and IsInf tests.
Use -ffp-model=precise to fix this.
IntelLLVM does not support -Wchar-subscripts
Fixes#4255
Some versions of MSVC provide incorrect values for the latter and rely on _MSVC_LANG instead.
Fixes#4226.
PiperOrigin-RevId: 527919195
Change-Id: Ifcca4612074f5ebc5337094426866a187f79f90a
