[PATCH] D153246: [AArch64] Mark vecreduce_fminimum and vecreduce_fmaximum as legal

[Dave Green via Phabricator via llvm-commits]

dmgreen created this revision.
dmgreen added reviewers: SjoerdMeijer, samtebbs, jaykang10, john.brawn, efriedma, anna.
Herald added subscribers: hiraditya, kristof.beyls.
Herald added a project: All.
dmgreen requested review of this revision.
Herald added a project: LLVM.

This adds some simple lowering of vecreduce_fminimum and vecreduce_fmaximum to fminv/fmaxv instructions, in the same way that vecreduce_fmax / vecreduce_fmin is lowered to fminnmv / fmaxnmv.


https://reviews.llvm.org/D153246

Files:
  llvm/include/llvm/Target/TargetSelectionDAG.td
  llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
  llvm/lib/Target/AArch64/AArch64InstrInfo.td
  llvm/test/CodeGen/AArch64/vecreduce-fmaximum.ll
  llvm/test/CodeGen/AArch64/vecreduce-fminimum.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D153246.532543.patch
Type: text/x-patch
Size: 10133 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20230619/4c54bc71/attachment.bin>