[all-commits] [llvm/llvm-project] ba7d42: [PowerPC] Optimize vec_splats of small FP values o...

[Maryam Moghadas via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: ba7d42b096cbef00320abe2b2e12a0125a6dc2f3
      https://github.com/llvm/llvm-project/commit/ba7d42b096cbef00320abe2b2e12a0125a6dc2f3
  Author: Maryam Moghadas <maryammo at ca.ibm.com>
  Date:   2026-05-28 (Thu, 28 May 2026)

  Changed paths:
    M llvm/lib/Target/PowerPC/PPCISelLowering.cpp
    M llvm/lib/Target/PowerPC/PPCISelLowering.h
    M llvm/lib/Target/PowerPC/PPCInstrVSX.td
    M llvm/test/CodeGen/PowerPC/combine-fneg.ll
    M llvm/test/CodeGen/PowerPC/fma-combine.ll
    M llvm/test/CodeGen/PowerPC/recipest.ll
    M llvm/test/CodeGen/PowerPC/repeated-fp-divisors.ll
    A llvm/test/CodeGen/PowerPC/vec-splats-opt.ll

  Log Message:
  -----------
  [PowerPC] Optimize vec_splats of small FP values on Power8/9 (#199538)

Recognize small constant FP splats representable as integers and lower
them to vspltisw + xvcvsxwdp/xvcvsxwsp on Power8/9, avoiding constant
pool loads.



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