[PATCH] D123147: [X86] Enable enableAggressiveFMAFusion to true for FMA capable targets (PR36826)

[Simon Pilgrim via Phabricator via llvm-commits]

RKSimon created this revision.
RKSimon added reviewers: craig.topper, pengfei, spatel, andrew.w.kaylor.
Herald added subscribers: StephenFan, hiraditya.
Herald added a project: All.
RKSimon requested review of this revision.
Herald added a project: LLVM.

If the X86 subtarget supports FMA, then allow it to aggressively generate FMA nodes, even if it means we have duplicated mul(x,y) and fma(x,y,z) cases

This demonstrates a likely flaw in the existing enableAggressiveFMAFusion folds - should we fold fadd(fmul(x,y), fmul(x,y)) -> fma(x,y,fmul(x,y)) ?


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D123147

Files:
  llvm/lib/Target/X86/X86ISelLowering.cpp
  llvm/lib/Target/X86/X86ISelLowering.h
  llvm/test/CodeGen/X86/dag-fmf-cse.ll
  llvm/test/CodeGen/X86/fmsubadd-combine.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D123147.420573.patch
Type: text/x-patch
Size: 5087 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20220405/56467344/attachment.bin>