[all-commits] [llvm/llvm-project] 3cf853: [X86][BF16] Improve vectorization of BF16 (#88486)

Phoebe Wang via All-commits all-commits at lists.llvm.org
Sat Apr 13 01:21:10 PDT 2024


  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 3cf8535dbf0bf5fafa99ea1f300e2384a7254fba
      https://github.com/llvm/llvm-project/commit/3cf8535dbf0bf5fafa99ea1f300e2384a7254fba
  Author: Phoebe Wang <phoebe.wang at intel.com>
  Date:   2024-04-13 (Sat, 13 Apr 2024)

  Changed paths:
    M llvm/lib/Target/X86/X86ISelLowering.cpp
    M llvm/test/CodeGen/X86/bfloat.ll
    M llvm/test/CodeGen/X86/concat-fpext-v2bf16.ll

  Log Message:
  -----------
  [X86][BF16] Improve vectorization of BF16 (#88486)

1. Move expansion to combineFP_EXTEND to help with small vectors;
2. Combine FP_ROUND to reduce fptrunc then fpextend after promotion;



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


More information about the All-commits mailing list