[llvm-dev] Compiling LLVM with GCC 5/6/7 and -std=c++1z fails due to use of std::char_traits constexpr

[David Blaikie via llvm-dev]

+ Eric (for standard library knowledge) & Richard (for general C++-ery).

I can't say I know much about __cpp_lib_constexpr_char_traits - got any
reference on how widely supported/implemented that is?

On Tue, Oct 6, 2020 at 10:17 AM Dan Bailey via llvm-dev <
llvm-dev at lists.llvm.org> wrote:

> Hi,
>
> When trying to compile LLVM using GCC 5.1 => GCC 7.2 (inclusive) with
> -std=c++1z, using StringRef errors due to the lack of  a constexpr
> std::char_traits<char>::length():
>
> .../include/llvm/ADT/StringRef.h:85:44: *error*: call to non-constexpr
> function ‘static std::size_t std::char_traits<char>::length(const
> char_type*)’
>        return std::char_traits<char>::length(Str);
>               ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~
>
> Looking into the issue, StringRef.h uses this logic to gate against using
> constexpr char_traits:
>
> #if *__cplusplus* > 201402L
>       return std::char_traits<char>::length(Str);
> #elif ...
>
> However for the GCC version range mentioned above, the -std=c++1z flag
> generates __cplusplus 201500L or higher despite char_traits not being
> constexpr.
>
> I think the correct thing to do here might be to use the feature macro
> instead:
>
> #if *__cpp_lib_constexpr_char_traits*
>       return std::char_traits<char>::length(Str);
> #elif ...
>
> Any thoughts?
>
> Thanks,
> --
>
> Dan Bailey        | R&D               | ILM Vancouver
> danbailey at ilm.com | 415 746-XXXX Desk | 778 882-XXXX Cell
>
> _______________________________________________
> LLVM Developers mailing list
> llvm-dev at lists.llvm.org
> https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20201102/865d2355/attachment.html>