[all-commits] [llvm/llvm-project] e06a9c: [LLVM][CodeGen][SVE] Improve lowering of fixed len...

[Paul Walker via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: e06a9ca2cb596c19c23c32a248e1b3b69d7e4c4f
      https://github.com/llvm/llvm-project/commit/e06a9ca2cb596c19c23c32a248e1b3b69d7e4c4f
  Author: Paul Walker <paul.walker at arm.com>
  Date:   2025-04-08 (Tue, 08 Apr 2025)

  Changed paths:
    M llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
    M llvm/test/CodeGen/AArch64/sve-fixed-length-masked-gather.ll
    M llvm/test/CodeGen/AArch64/sve-fixed-length-masked-loads.ll
    M llvm/test/CodeGen/AArch64/sve-fixed-length-masked-scatter.ll

  Log Message:
  -----------
  [LLVM][CodeGen][SVE] Improve lowering of fixed length masked mem ops. (#134402)

Converting fixed length masks, as used by MLOAD, to scalable vectors is
done by comparing the mask to zero. When the mask is the result of a
compare we can instead promote the operands and regenerate the original
compare. At worst this reduces the dependecy chain and in most cases
removes the need for multiple compares.



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