[Bug libstdc++/126849] New: std::numbers wrong check for existence of bfloat16_t

gcc at maxmitti dot at via Gcc-bugs <[email protected]> Thu, 13 Aug 2026 14:35:07 +0000

Newsgroups	gmane.comp.gcc.bugs
Message-ID	<[email protected]/bugzilla/>

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=126849

            Bug ID: 126849
           Summary: std::numbers wrong check for existence of bfloat16_t
           Product: gcc
           Version: 16.2.1
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: libstdc++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: gcc at maxmitti dot at
  Target Milestone: ---

The specialization of values in std::numbers for bfloat16_t are guarded like
this:

#ifdef __STDCPP_BFLOAT128_T__
__glibcxx_numbers (__gnu_cxx::__bfloat16_t, BF16);
#endif

Here:
https://github.com/gcc-mirror/gcc/blob/194c8f2bc612df9f6ef1472d6279328c0b4e603e/libstdc%2B%2B-v3/include/std/numbers#L225

I think the macro that should be checked for is actually __STDCPP_BFLOAT16_T__