[all-commits] [llvm/llvm-project] 6f88ac: [SLP][X86] Improve reordering to consider alternat...

[vporpo via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 6f88acf410b48f3e6c1526df2dc32ed86f249685
      https://github.com/llvm/llvm-project/commit/6f88acf410b48f3e6c1526df2dc32ed86f249685
  Author: Vasileios Porpodas <vporpodas at google.com>
  Date:   2022-06-21 (Tue, 21 Jun 2022)

  Changed paths:
    M llvm/include/llvm/Analysis/TargetTransformInfo.h
    M llvm/include/llvm/Analysis/TargetTransformInfoImpl.h
    M llvm/lib/Analysis/TargetTransformInfo.cpp
    M llvm/lib/Target/X86/X86TargetTransformInfo.cpp
    M llvm/lib/Target/X86/X86TargetTransformInfo.h
    M llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
    M llvm/test/Transforms/SLPVectorizer/X86/reorder_with_external_users.ll

  Log Message:
  -----------
  [SLP][X86] Improve reordering to consider alternate instruction bundles

During the reordering transformation we should try to avoid reordering bundles
like fadd,fsub because this may block them being matched into a single vector
instruction in x86.
We do this by checking if a TreeEntry is such a pattern and adding it to the
list of TreeEntries with orders that need to be considered.

Differential Revision: https://reviews.llvm.org/D125712