Google Test has recently (02/09/2018) switched the default death test
style from "fast" to "threadsafe" in
ec7faa943d
Threadsafe death tests have been used internally for a while, and are
proven to be a better default.
However, adopting this better default can be challenging for large
projects with a significant investment in custom infrastructure built on
top of Google Test. The same custom infrastructure can make it difficult
for large projects to switch back to the old default by passing in
--gtest_death_test_style=fast.
For the reasons above, the default switch is considered too disruptive,
and this CL reverts it. This CL also introduces the
GTEST_DEFAULT_DEATH_TEST_STYLE preprocesor macro, which replaces the
hard-coded default. The macro can be defined in
gtest/internal/custom/gtest-port.h by projects that are ready to migrate
to thread-safe death tests.
_HAS_EXCEPTIONS is specific to the MSVC STL and defining it to 0 causes
problems with libc++, so libc++ users may leave it undefined. This can
cause GTEST_HAS_EXCEPTIONS to be defined incorrectly if the user has
disabled exceptions via the compiler, which can lead to build errors.
_CPPUNWIND is a builtin macro provided by the compiler so it should
work with both STLs.
clang-cl is clang for Windows running in MSVC mode. Chromium uses it for
Windows builds. clang-cl is weird in that it defines __clang__ and
_MSC_VER, but *NOT* __GNUC__. This is vaguely analogous to how normal
clang defines __clang__ (what it is) and __GNUC__ (what it is compatible
with).
However, clang-cl still implements most GCC extensions, being clang.
Notably, the way to control -Wformat-literal is still with
__attribute__((__format__)). For better error-checking and strict
-Wformatl-literal compatibility (see
53c478d639), define
GTEST_ATTRIBUTE_PRINTF_ in clang-cl too.
This makes it easier to use GTest in projects that build with the
-Wmissing-declarations warning. This fixes the warning in headers and
source files, though not GTest's own tests as it is rather noisy there.
On platforms with std::tuple and not std::tr1::tuple, GTEST_HAS_COMBINE
gets turned off when it works fine (due to GTEST_TUPLE_NAMESPACE_).
Elsewhere in the project, several GTEST_HAS_TR1_TUPLE checks
additionally check GTEST_HAS_STD_TUPLE_, so use that formulation.
(The ones that don't are specific to std::tr1::tuple and are followed by
an identical GTEST_HAS_STD_TUPLE_ version underneath it.)
In particular, this fixes testing::Combine on MSVC 2017, which regressed
here:
https://github.com/google/googletest/pull/1348#issuecomment-353879010
ThreadLocal class needs to be have default visibility.
Root cause is gtest uses typeinfo for the ThreadLocal class.
The problem manifests When gtest/gmock are built as a shared library
with libc++. When a class is used in typeinfo, it must have default
visibility.
There is an explanation about typeinfo and visibility here:
https://developer.apple.com/library/content/documentation/DeveloperTools/Conceptual/CppRuntimeEnv/Articles/SymbolVisibility.html
When libc++ is used with gtest in shared library mode, any tests
that are compiled with -fvisibility=hidden and exercise the
macro EXPECT_CALL, it results in an abort like:
[ FATAL ] /usr/include/gtest/internal/gtest-port.h:1394::
Condition typeid(*base) == typeid(Derived) failed.
This is because the typeinfo for ThreadLocal class is not visible.
Therefore, linker failed to match it to the shared library symbol, creating a
new symbol instead.
This fixes https://github.com/google/googletest/issues/1207.
TempDir() function is declared twice, once in `internal/gtest-port.h`
and a second time in `gtest.h`.
Fixes a warning with GCC when -Wredundant-decls is given.
As mentioned in issue #360:
"Now that all the platforms gtest supports work with value-parameterized
tests, we should remove the uses of the GTEST_HAS_PARAM_TESTS macro from
the codebase everywhere."
https://github.com/google/googletest/issues/360
MSVC has an optional warning which flags when 32-bit pointers get cast
into a 64-bit value. This is a little overaggressive I think, but to
ease compiling in projects with aggressive warnings, fix this by just
casting to const void * directly. Modern GCCs seem to compile it just
fine.
This merges a Google-internal change (117235625).
Original CL description:
This CL was created manually in about an hour with sed, a Python script
to find all the places unqualified 'string' was mentioned, and some help
from Emacs to add the "std::" qualifications, plus a few manual tweaks.
This upstreams a Google-internal change.
Original CL description:
The C++ standard says that function pointers are not implicitly
convertible to object pointers. Visual Studio disregards that and allows
implicit conversion between function pointers and object points, and
enough code relies on this that clang follows suit in
Microsoft-compatibility mode.
However, clang emits a -Wmicrosoft-cast warning when such a conversion
is done:
E:\b\c\b\win_clang\src\sandbox\win\src\sync_dispatcher.cc(42,7):
warning: implicit conversion between pointer-to-function and
pointer-to-object is a Microsoft extension [-Wmicrosoft-cast]
This change fixes this warning in gtest, while hopefully not changing
any behavior. The change does two things:
1. It replaces the if in DefaultPrintTo with SFINAE
2. In C++11 mode, it uses enable_if<is_function<>> instead of
ImplicitlyConvertible<T*, const void*> to check if the
explicit cast is needed.
With this change, functions will use the branch with the reintpret_casts
with Visual Studio and clang/win, and clang no longer needs to warn
that it implicitly converts a function pointer to a void pointer.
Fixes the following errors:
hash_map(17): error C2338: <hash_map> is deprecated and will be REMOVED. Please use <unordered_map>. You can define _SILENCE_STDEXT_HASH_DEPRECATION_WARNINGS to acknowledge that you have received this warning. [googlemock\gtest\gtest-printers_test.vcxproj]
hash_set(17): error C2338: <hash_set> is deprecated and will be REMOVED. Please use <unordered_set>. You can define _SILENCE_STDEXT_HASH_DEPRECATION_WARNINGS to acknowledge that you have received this warning. [googlemock\gtest\gtest-printers_test.vcxproj]
googletest\test\gtest_catch_exceptions_test_.cc(152): error C2220: warning treated as error - no 'object' file generated
googletest\test\gtest_catch_exceptions_test_.cc(152): warning C4297: 'CxxExceptionInDestructorTest::~CxxExceptionInDestructorTest': function assumed not to throw an exception but does
googletest\test\gtest_catch_exceptions_test_.cc(152): note: destructor or deallocator has a (possibly implicit) non-throwing exception specification
There were already some checks for _MSC_VER in the code, so this commit
continues in that vein.
If REGISTER_TYPED_TEST_CASE_P is included in a header file, but
the .cc file does not declare INSTANTIATE_TYPED_TEST_CASE_P, an
unused-variable warning may be raised by the compiler.
googletest doesn't currently build with clang's very aggressive
-Wformat-nonliteral warning. It requires that all non-literal format
strings come from the argument of a function annotated with a compatible
format attribute.
Fixing that reports that ColoredPrintf's callers weren't passing the
normal -Wformat warning. Some messages were passed directly into the
format string rather than via "%s".
If $XML_OUTPUT_FILE is set, and $GTEST_OUTPUT and --gtest_output are not
specified, produce output as if GTEST_OUTPUT=xml:$XML_OUTPUT_FILE had
been set.
