[all-commits] [llvm/llvm-project] 711dff: [X86] Add matchTruncateWithPACK helper for matchin...

[Simon Pilgrim via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 711dff45772e04bbaff284dfb16501cbc96bef1c
      https://github.com/llvm/llvm-project/commit/711dff45772e04bbaff284dfb16501cbc96bef1c
  Author: Simon Pilgrim <llvm-dev at redking.me.uk>
  Date:   2023-08-07 (Mon, 07 Aug 2023)

  Changed paths:
    M llvm/lib/Target/X86/X86ISelLowering.cpp
    M llvm/test/CodeGen/X86/vector-trunc-packus.ll
    M llvm/test/CodeGen/X86/vector-trunc-ssat.ll
    M llvm/test/CodeGen/X86/vector-trunc-usat.ll

  Log Message:
  -----------
  [X86] Add matchTruncateWithPACK helper for matching signbits/knownbits for PACKSS/PACKUS

Begin to consolidate the similar matching code we have - all have semi-similar constraints that still need merging together to ensure we get consistent codegen depending on when the truncate is lowered.


  Commit: 0d1f8532bcde2a6001fb672d056673e1a6907187
      https://github.com/llvm/llvm-project/commit/0d1f8532bcde2a6001fb672d056673e1a6907187
  Author: Simon Pilgrim <llvm-dev at redking.me.uk>
  Date:   2023-08-07 (Mon, 07 Aug 2023)

  Changed paths:
    M llvm/lib/Target/X86/X86ISelLowering.cpp
    M llvm/test/CodeGen/X86/vector-pack-128.ll

  Log Message:
  -----------
  [X86] truncateVectorWithPACK - ensure we don't truncate to <1 x iXX> vector types

Fuzz testing noticed that the sub-128-bit vector splitting added in ef4330f4f3cc didn't correctly halt at <2 x iXX> truncations.


Compare: https://github.com/llvm/llvm-project/compare/bbb95893ded9...0d1f8532bcde