-
Notifications
You must be signed in to change notification settings - Fork 556
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
porting/cpphdrcheck.t: test perl's headers with C++ compilers #22232
Conversation
|
Is there anyone familiar enough with C++ headers to give this proposal for a porting test a review? IMO it's the sort of thing that, if we're going to pull it into core, we should do so now, at an early point in the current development cycle. |
Adding _GNU_SOURCE, which was done much later, should give us the GNU-ish symbols (though POSIX_C_SOURCE would be more general.) I couldn't find the reason for this being added, but the C++ headers react badly to it: In file included from /usr/lib/gcc/x86_64-pc-cygwin/11/include/c++/iostream:38, from source.cpp:1: /usr/lib/gcc/x86_64-pc-cygwin/11/include/c++/x86_64-pc-cygwin/bits/c++config.h:573:2: warning: #warning "__STRICT_ANSI__ seems to have been undefined; this is not supported" [-Wcpp] 573 | #warning "__STRICT_ANSI__ seems to have been undefined; this is not supported" | ^~~~~~~ In file included from /usr/lib/gcc/x86_64-pc-cygwin/11/include/c++/bits/max_size_type.h:37, from /usr/lib/gcc/x86_64-pc-cygwin/11/include/c++/bits/ranges_base.h:38, from /usr/lib/gcc/x86_64-pc-cygwin/11/include/c++/string_view:48, from /usr/lib/gcc/x86_64-pc-cygwin/11/include/c++/bits/basic_string.h:48, from /usr/lib/gcc/x86_64-pc-cygwin/11/include/c++/string:55, from /usr/lib/gcc/x86_64-pc-cygwin/11/include/c++/bits/locale_classes.h:40, from /usr/lib/gcc/x86_64-pc-cygwin/11/include/c++/bits/ios_base.h:41, from /usr/lib/gcc/x86_64-pc-cygwin/11/include/c++/ios:42, from /usr/lib/gcc/x86_64-pc-cygwin/11/include/c++/ostream:38, from /usr/lib/gcc/x86_64-pc-cygwin/11/include/c++/iostream:39, from source.cpp:1: /usr/lib/gcc/x86_64-pc-cygwin/11/include/c++/numbers:139:9: error: unable to find numeric literal operator 'operator""Q' 139 | = 2.718281828459045235360287471352662498Q; | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ...
This searches for a C++ compiler based on the supplied C compiler, and checks that compiler for any options controlling the C++ standard requested, including simple checks that the compiler supports that standard. If a C++ compiler is found, test compilation of the same simple code as above but with the perl headers included after any C++ headers. Ideally we'd also test runtime, but would require more complex test code, which I leave to later contributors (which may still be me). Tested at various times with: - MSVC - gcc - clang - Oracle/Sun Development Workshop cc (CC is the C++ compiler), on Oracle Linux - Intel oneAPI compiler (llvm based apparently, and now free to use) - Intel classic compiler (discontinued) Currently this probes the compiler for C++ sanity with the perl ccflags, since icc (Intel classic) would successfully build the sample without perl's ccflags, but then fail with both the headers and perl's ccflags. It turned out to fail with just the ccflags, and since the primary intent is to test the headers, I probe *with* ccflags. The Sun Workshop compiler failed to build the C++11 or 14 sample at all in my testing, which may have been due to an installation problem.
ac23c4f
to
2d61a8e
Compare
Fixed everything I think:
|
This searches for a C++ compiler based on the supplied C compiler, and
checks that compiler for any options controlling the C++ standard
requested, including simple checks that the compiler supports that
standard.
If a C++ compiler is found, test compilation of the same simple code
as above but with the perl headers included after any C++ headers.
Ideally we'd also test runtime, but would require more complex test
code, which I leave to later contributors (which may still be me).
Tested at various times with:
Currently this probes the compiler for C++ sanity with the perl
ccflags, since icc (Intel classic) would successfully build the sample
without perl's ccflags, but then fail with both the headers and perl's
ccflags. It turned out to fail with just the ccflags, and since the
primary intent is to test the headers, I probe with ccflags.
The Sun Workshop compiler failed to build the C++11 or 14 sample at
all in my testing, which may have been due to an installation problem.
This branch also removes an option added to ccflags by the cygwin hints
that prevented C++ code from compiling.