[all-commits] [llvm/llvm-project] 7e9747: [X86][F16C] Remove cvtph2ps intrinsics and use gen...

[Simon Pilgrim via All-commits]

  Branch: refs/heads/master
  Home:   https://github.com/llvm/llvm-project
  Commit: 7e9747b50bcb1be28d4a3236571e8050835497a6
      https://github.com/llvm/llvm-project/commit/7e9747b50bcb1be28d4a3236571e8050835497a6
  Author: Simon Pilgrim <llvm-dev at redking.me.uk>
  Date:   2020-02-29 (Sat, 29 Feb 2020)

  Changed paths:
    M clang/lib/CodeGen/CGBuiltin.cpp
    M clang/test/CodeGen/avx512f-builtins-constrained.c
    M clang/test/CodeGen/avx512f-builtins.c
    M clang/test/CodeGen/avx512vl-builtins-constrained.c
    M clang/test/CodeGen/avx512vl-builtins.c
    M clang/test/CodeGen/f16c-builtins-constrained.c
    M clang/test/CodeGen/f16c-builtins.c
    M llvm/include/llvm/IR/IntrinsicsX86.td
    M llvm/lib/IR/AutoUpgrade.cpp
    M llvm/lib/Target/X86/X86IntrinsicsInfo.h
    M llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp
    M llvm/test/CodeGen/X86/avx512-intrinsics-fast-isel.ll
    M llvm/test/CodeGen/X86/avx512-intrinsics-upgrade.ll
    M llvm/test/CodeGen/X86/avx512-intrinsics.ll
    M llvm/test/CodeGen/X86/avx512vl-intrinsics-fast-isel.ll
    M llvm/test/CodeGen/X86/avx512vl-intrinsics-upgrade.ll
    M llvm/test/CodeGen/X86/avx512vl-intrinsics.ll
    M llvm/test/CodeGen/X86/f16c-intrinsics-fast-isel.ll
    A llvm/test/CodeGen/X86/f16c-intrinsics-upgrade.ll
    M llvm/test/CodeGen/X86/f16c-intrinsics.ll
    M llvm/test/Transforms/InstCombine/X86/x86-f16c.ll

  Log Message:
  -----------
  [X86][F16C] Remove cvtph2ps intrinsics and use generic half2float conversion (PR37554)

This removes everything but int_x86_avx512_mask_vcvtph2ps_512 which provides the SAE variant, but even this can use the fpext generic if the rounding control is the default.

Differential Revision: https://reviews.llvm.org/D75162