[all-commits] [llvm/llvm-project] ec294d: [X86] LowerVectorAllZero - lower to CMP(MOVMSK(NOT...

[Simon Pilgrim via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: ec294d2f8f1839e11c13ee32279cd28c1f46f66f
      https://github.com/llvm/llvm-project/commit/ec294d2f8f1839e11c13ee32279cd28c1f46f66f
  Author: Simon Pilgrim <llvm-dev at redking.me.uk>
  Date:   2023-03-23 (Thu, 23 Mar 2023)

  Changed paths:
    M llvm/lib/Target/X86/X86ISelLowering.cpp
    M llvm/test/CodeGen/X86/pr45378.ll
    M llvm/test/CodeGen/X86/ptest.ll
    M llvm/test/CodeGen/X86/vector-reduce-or-cmp.ll

  Log Message:
  -----------
  [X86] LowerVectorAllZero - lower to CMP(MOVMSK(NOT(X)),0) instead of CMP(MOVMSK(X),65535)

In most cases the NOT will still be scalarized, but it allows us to perform the CMP(X,0) combines inside combineCMP()