[all-commits] [llvm/llvm-project] 660739: AMDGPU: Allow vectorization of round intrinsic

[Matt Arsenault via All-commits]

  Branch: refs/heads/master
  Home:   https://github.com/llvm/llvm-project
  Commit: 66073953a5cb7ecac32bcb897033fdc337c56b5e
      https://github.com/llvm/llvm-project/commit/66073953a5cb7ecac32bcb897033fdc337c56b5e
  Author: Matt Arsenault <Matthew.Arsenault at amd.com>
  Date:   2020-03-23 (Mon, 23 Mar 2020)

  Changed paths:
    M llvm/lib/Target/AMDGPU/AMDGPUTargetTransformInfo.cpp
    A llvm/test/Transforms/SLPVectorizer/AMDGPU/round.ll

  Log Message:
  -----------
  AMDGPU: Allow vectorization of round intrinsic

There seems to be a small benefit to the legalized sequence for v2f16
round with packed instructions, so allow vectorizing it by reducing
the cost.

An unintended side effect is vectorization of f32 round also
happens. The current FMA logic seems off to me, and isn't checking for
packed instructions.