[PATCH] D126234: [AArch64] Add support for FMA intrinsics to shouldSinkOperands.

Florian Hahn via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Mon May 23 12:33:47 PDT 2022


fhahn created this revision.
fhahn added reviewers: dmgreen, samparker, t.p.northover.
Herald added subscribers: hiraditya, kristof.beyls.
Herald added a project: All.
fhahn requested review of this revision.
Herald added a project: LLVM.

If the fma operates on a legal vector type, the indexed variants can be
used, if the second operand is a splat of a valid index.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D126234

Files:
  llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
  llvm/test/Transforms/CodeGenPrepare/AArch64/sink-free-instructions.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D126234.431453.patch
Type: text/x-patch
Size: 13945 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20220523/bc7cb009/attachment.bin>


More information about the llvm-commits mailing list