[all-commits] [llvm/llvm-project] c9320b: [X86] Use correctly sized floating point literals ...

[Craig Topper via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: c9320bc87121af2a198f174c8d58f75295efbbd7
      https://github.com/llvm/llvm-project/commit/c9320bc87121af2a198f174c8d58f75295efbbd7
  Author: Craig Topper <craig.topper at sifive.com>
  Date:   2022-11-17 (Thu, 17 Nov 2022)

  Changed paths:
    M clang/lib/Headers/avx512fintrin.h
    M clang/lib/Headers/avxintrin.h
    M clang/lib/Headers/emmintrin.h
    M clang/lib/Headers/xmmintrin.h
    M clang/test/CodeGen/X86/avx512dq-builtins-constrained.c

  Log Message:
  -----------
  [X86] Use correctly sized floating point literals in *zero_ps/pd.

This avoids depending on int->float or double->float conversion.
Improving codegen with #pragma STDC FENV_ACCESS ON.

Really we should improve constant folding somewhere, but this was
a cheap and easy improvement.

Fixes PR59052.