[Bug c++/116319] std::fma doesn't compute std::bfloat16_t as if to infinite precision and rounded only once.

[pinskia at gcc dot gnu.org via Gcc-bugs]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=116319

--- Comment #7 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
(In reply to SHIH YEN-TE from comment #5)
> Sorry, could you please confirm if there is any standard function in library
> can  compute bfloat16_t without rounding twice as if to infinite precision?
> Thank you.

There is none unless you have bfloat16_t support in HW and even then 

Note you need -frounding-math if you use fesetround.