[libc-commits] [PATCH] D144340: [libc][math] Set floating point exceptions for exp*f, sinhf, and coshf.

[Siva Chandra via Phabricator via libc-commits]

sivachandra added a comment.

In D144340#4145803 <https://reviews.llvm.org/D144340#4145803>, @lntue wrote:

> In D144340#4145798 <https://reviews.llvm.org/D144340#4145798>, @sivachandra wrote:
>
>> Should the `errno` be set and exceptions raised based on `math_errhandling`?
>
> I skipped the check for `math_errhandling` because of the following line from https://en.cppreference.com/w/cpp/numeric/math/math_errhandling :
>
>   If the implementation supports IEEE floating-point arithmetic (IEC 60559), math_errhandling & MATH_ERREXCEPT is required to be non-zero.

I think that is true when the implementation is conformant. But, under __FAST_MATH__, it is expected that `math_errhandling` evaluates to zero: https://github.com/llvm/llvm-project/blob/main/libc/config/linux/api.td#L57. There is `__NO_MATH_ERRNO__` also.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D144340/new/

https://reviews.llvm.org/D144340