[all-commits] [llvm/llvm-project] 78a28b: [X86] Support icmp_i64(i64 X, i64 Y) --> icmp_i32(...

[Simon Pilgrim via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 78a28b3ef7de2a07c8d01cace58019068a17a0bf
      https://github.com/llvm/llvm-project/commit/78a28b3ef7de2a07c8d01cace58019068a17a0bf
  Author: Simon Pilgrim <llvm-dev at redking.me.uk>
  Date:   2024-05-05 (Sun, 05 May 2024)

  Changed paths:
    M llvm/lib/Target/X86/X86ISelLowering.cpp
    M llvm/test/CodeGen/X86/cmp-shiftX-maskX.ll
    M llvm/test/CodeGen/X86/vector-shift-by-select-loop.ll

  Log Message:
  -----------
  [X86] Support icmp_i64(i64 X, i64 Y) --> icmp_i32(trunc(X), trunc(Y)) for non-constant Y (#91085)

Relax the constraint on the fold to allow any value for X and Y iff all upper 32-bits are known zero

Minor code-size saving by moving to i32 bit instructions



To unsubscribe from these emails, change your notification settings at https://github.com/llvm/llvm-project/settings/notifications