[clang] d930ed1 - Disallow use of __has_c_attribute in C++ mode.

Fri Feb 28 07:05:27 PST 2020

On Thu, Feb 27, 2020 at 6:04 PM James Y Knight <jyknight at google.com> wrote:
>
> That all makes sense -- especially the bits about the dates needing to be different.
>
> But, even with all that, I'm not sure why we shouldn't implement both __has_cpp_attribute AND __has_c_attribute in C++ mode?
>
> The subset of C attributes which retain their C-defined semantics in C++ (which, hopefully, is all of them, but doesn't need to be) should return the appropriate C-standard date with which the (C++ mode) implementation is compatible. That is, in -std=c++11, we may have __has_cpp_attribute(deprecated) == 201309, while __has_c_attribute(deprecated) == 2020XXYY.

If I'm understanding you properly, then I think that this leads to
some strange (to me, at least) behavior. We have some attributes that
are C only attributes, some attributes that are C++ only, but most of
the attributes are supported in both C and C++ with the same
semantics. For the attributes that are in both C and C++, what you
suggest makes sense. For the attributes that exist only in one
language, I'm less convinced. It would be weird for
__has_c_attribute(gnu::transparent_union) (which is a C-only
attribute) to report true in C and false in C++ because we do support
that C attribute. Similar for
__has_cpp_attribute(clang::lifetimebound) (which is a C++-only
attribute) to report true in C++ and false in C. Having the language
in the name of the feature test macro is what makes this so awkward.

> I'd hope/expect that whichever version of the C++ standard updates its baseline to C2x will adopt this feature "automatically", and if it doesn't, that it'll be included explicitly. But, ISTM we probably don't need to wait for that to occur in order to implement that behavior in clang?

WG21 doesn't automatically incorporate the language parts of C, only
the library parts, so this would require an explicit change to the
standard. I would not be surprised if WG21 was just as unhappy about
supporting __has_c_attribute as WG14 is about the idea of supporting
__has_cpp_attribute for the same reasons.

Another possible option is to add __has_std_attribute with the
behavior you describe, and possibly a pile of macros for the feature
test dates. This negates the naming concern and retains the same
semantics, but it comes at the expense of adding yet another way to
feature test for attributes. Once we have field experience (and esp if
we could convince GCC to adopt the same feature), hopefully both
committees would have some appetite for standardizing it as an
existing practice.

~Aaron

>
>
>
> On Mon, Nov 25, 2019, 7:07 PM Aaron Ballman <aaron at aaronballman.com> wrote:
>>
>> On Mon, Nov 25, 2019 at 5:59 PM James Y Knight <jyknight at google.com> wrote:
>> >
>> > Isn't this unnecessarily annoying to users? You have the same syntax to use the attributes, and the attributes are expected to be compatible when named the same way, but you can't use the same #if conditional to check for availability, when writing a header intended to work in both modes?
>>
>> Yes, I do think it's needlessly annoying to users.
>>
>> WG21 adopted __has_cpp_attribute for C++2a based on the existing Clang
>> implementation. WG14 is looking to add a similar feature testing
>> macro, but cannot use __has_cpp_attribute due to the name and so I
>> proposed __has_c_attribute for parity with __has_cpp_attribute and
>> implemented it in Clang. After a lot of discussions with users,
>> committee members, and other implementers, I realized this is silly:
>> we could have a single macro that handles both C and C++ depending on
>> the language mode being compiled for. Because of that, as a liaison I
>> filed an NB comment with WG21 to have them reconsider the name
>> __has_cpp_attribute before it shipped in a standard so that we could
>> use a language-neutral name for both languages. In Belfast, EWG
>> reaffirmed that they want the name __has_cpp_attribute. They found my
>> arguments to be unpersuasive and didn't think it would be possible for
>> the semantics of attributes to align between the languages, or to
>> commit to such a guarantee (despite WG14 having a chartered mandate to
>> be compatible with features adopted from C++).
>>
>> Rather than leaving C out in the cold because EWG came to the decision
>> they did, we're kind of stuck with __has_c_attribute.
>>
>> One problem this does solve is that __has_cpp_attribute and
>> __has_c_attribute both return a numeric value that is date-like for
>> standard attributes, but those standard attributes are adopted into
>> different standards at different times. Because returned values are
>> different, it makes use harder when dealing with attribute
>> modifications over time. This isn't impossible to solve, we could
>> introduce macros for the return values and base their values on which
>> language mode is being preprocessed, but we no longer have to.
>>
>> ~Aaron
>>
>> >
>> > On Mon, Nov 25, 2019 at 5:35 PM Aaron Ballman via cfe-commits <cfe-commits at lists.llvm.org> wrote:
>> >>
>> >>
>> >> Author: Aaron Ballman
>> >> Date: 2019-11-25T17:35:12-05:00
>> >> New Revision: d930ed1acc0ea49d4b3aae7e95b4c6d9cd310578
>> >>
>> >> URL: https://github.com/llvm/llvm-project/commit/d930ed1acc0ea49d4b3aae7e95b4c6d9cd310578
>> >> DIFF: https://github.com/llvm/llvm-project/commit/d930ed1acc0ea49d4b3aae7e95b4c6d9cd310578.diff
>> >>
>> >> LOG: Disallow use of __has_c_attribute in C++ mode.
>> >>
>> >> __has_cpp_attribute is not available in C mode, and __has_c_attribute
>> >> should not be available in C++ mode. This also adds a test to
>> >> demonstrate that we properly handle scoped attribute tokens even in C
>> >> mode.
>> >>
>> >> Added:
>> >>     clang/test/Preprocessor/has_c_attribute.cpp
>> >>
>> >> Modified:
>> >>     clang/lib/Lex/PPMacroExpansion.cpp
>> >>     clang/test/Preprocessor/has_c_attribute.c
>> >>
>> >> Removed:
>> >>
>> >>
>> >>
>> >> ################################################################################
>> >> diff  --git a/clang/lib/Lex/PPMacroExpansion.cpp b/clang/lib/Lex/PPMacroExpansion.cpp
>> >> index e6e00b1c1700..a69c4dbb3a2a 100644
>> >> --- a/clang/lib/Lex/PPMacroExpansion.cpp
>> >> +++ b/clang/lib/Lex/PPMacroExpansion.cpp
>> >> @@ -370,7 +370,11 @@ void Preprocessor::RegisterBuiltinMacros() {
>> >>    Ident__has_extension    = RegisterBuiltinMacro(*this, "__has_extension");
>> >>    Ident__has_builtin      = RegisterBuiltinMacro(*this, "__has_builtin");
>> >>    Ident__has_attribute    = RegisterBuiltinMacro(*this, "__has_attribute");
>> >> -  Ident__has_c_attribute  = RegisterBuiltinMacro(*this, "__has_c_attribute");
>> >> +  if (!LangOpts.CPlusPlus)
>> >> +    Ident__has_c_attribute = RegisterBuiltinMacro(*this, "__has_c_attribute");
>> >> +  else
>> >> +    Ident__has_c_attribute = nullptr;
>> >> +
>> >>    Ident__has_declspec = RegisterBuiltinMacro(*this, "__has_declspec_attribute");
>> >>    Ident__has_include      = RegisterBuiltinMacro(*this, "__has_include");
>> >>    Ident__has_include_next = RegisterBuiltinMacro(*this, "__has_include_next");
>> >>
>> >> diff  --git a/clang/test/Preprocessor/has_c_attribute.c b/clang/test/Preprocessor/has_c_attribute.c
>> >> index 843a67a2646c..f8b0b364faa5 100644
>> >> --- a/clang/test/Preprocessor/has_c_attribute.c
>> >> +++ b/clang/test/Preprocessor/has_c_attribute.c
>> >> @@ -1,4 +1,5 @@
>> >>  // RUN: %clang_cc1 -fdouble-square-bracket-attributes -std=c11 -E %s -o - | FileCheck %s
>> >> +// RUN: %clang_cc1 -std=c2x -E %s -o - | FileCheck %s
>> >>
>> >>  // CHECK: has_fallthrough
>> >>  #if __has_c_attribute(fallthrough)
>> >> @@ -14,3 +15,8 @@
>> >>  #if __has_c_attribute(__nodiscard__)
>> >>    int has_nodiscard_underscore();
>> >>  #endif
>> >> +
>> >> +// CHECK: has_clang_annotate
>> >> +#if __has_c_attribute(clang::annotate)
>> >> +  int has_clang_annotate();
>> >> +#endif
>> >>
>> >> diff  --git a/clang/test/Preprocessor/has_c_attribute.cpp b/clang/test/Preprocessor/has_c_attribute.cpp
>> >> new file mode 100644
>> >> index 000000000000..0bde73067178
>> >> --- /dev/null
>> >> +++ b/clang/test/Preprocessor/has_c_attribute.cpp
>> >> @@ -0,0 +1,8 @@
>> >> +// RUN: %clang_cc1 -std=c++11 %s -verify
>> >> +
>> >> +#if __has_c_attribute(fallthrough) // expected-error {{function-like macro '__has_c_attribute' is not defined}}
>> >> +#endif
>> >> +
>> >> +#if __has_c_attribute(gnu::transparent_union) // expected-error {{function-like macro '__has_c_attribute' is not defined}}
>> >> +#endif
>> >> +
>> >>
>> >>
>> >>
>> >> _______________________________________________
>> >> cfe-commits mailing list
>> >> cfe-commits at lists.llvm.org
>> >> https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits