[all-commits] [llvm/llvm-project] 67c3f2: [X86] mayFoldIntoStore - peek through oneuse bitca...

[Simon Pilgrim via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 67c3f2b4303972a6dc8ada54efe1d5d80d119a51
      https://github.com/llvm/llvm-project/commit/67c3f2b4303972a6dc8ada54efe1d5d80d119a51
  Author: Simon Pilgrim <llvm-dev at redking.me.uk>
  Date:   2025-01-18 (Sat, 18 Jan 2025)

  Changed paths:
    M llvm/lib/Target/X86/X86ISelLowering.cpp
    M llvm/test/CodeGen/X86/canonicalize-vars-f16-type.ll
    M llvm/test/CodeGen/X86/cvt16.ll
    M llvm/test/CodeGen/X86/fp-strict-scalar-fp16.ll
    M llvm/test/CodeGen/X86/fp16-libcalls.ll
    M llvm/test/CodeGen/X86/half-constrained.ll
    M llvm/test/CodeGen/X86/half-darwin.ll
    M llvm/test/CodeGen/X86/half.ll
    M llvm/test/CodeGen/X86/pr91005.ll
    M llvm/test/CodeGen/X86/pr95278.ll
    M llvm/test/CodeGen/X86/vector-half-conversions.ll

  Log Message:
  -----------
  [X86] mayFoldIntoStore - peek through oneuse bitcase users to find a store node (#123366)

mayFoldIntoStore currently just checks the direct (oneuse) user of a
SDValue to check its stored, which prevents cases where we bitcast the
value prior to storing (usually the bitcast will be removed later).

This patch peeks up through oneuse BITCAST nodes chain to see if its
eventually stored.

The main use of mayFoldIntoStore is v8i16 EXTRACT_VECTOR_ELT lowering
which will only use PEXTRW/PEXTRB for index0 extractions (vs the faster
MOVD) if the extracted value will be folded into a store on SSE41+
targets.

Fixes #107086



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