[all-commits] [llvm/llvm-project] 93c9b3: [X86] Fix MOVMSK(CONCAT(X, Y)) -> MOVMSK(AND/OR(X, Y...

[Simon Pilgrim via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 93c9b39d25ce842a911e800f68432668403b8aca
      https://github.com/llvm/llvm-project/commit/93c9b39d25ce842a911e800f68432668403b8aca
  Author: Simon Pilgrim <llvm-dev at redking.me.uk>
  Date:   2022-01-28 (Fri, 28 Jan 2022)

  Changed paths:
    M llvm/lib/Target/X86/X86ISelLowering.cpp
    M llvm/test/CodeGen/X86/combine-movmsk-avx.ll

  Log Message:
  -----------
  [X86] Fix MOVMSK(CONCAT(X,Y)) -> MOVMSK(AND/OR(X,Y)) fold for float types and demanded elements

rG9103b73fe052 was assuming that we could OR/AND with the source vector, but that will fail on float/double vectors without bitcasting - it also missed the case that any_of checks might be testing less than all the source elements