[PATCH] D137422: [DAGCombine] Generalize foldSelectCCToShiftAnd

[Mikhail Gudim via Phabricator via llvm-commits]

mgudim updated this revision to Diff 475094.
mgudim added a comment.

@craig.topper

When I previously said that now optimization works in all the cases I was wrong. You're right that we don't need to consider the non-strict inequalities. But we still need to extend the logic to work on the "inverted" patterns where the zero is the true value in the select instruction.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D137422/new/

https://reviews.llvm.org/D137422

Files:
  llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
  llvm/test/CodeGen/AArch64/arm64-fmax.ll
  llvm/test/CodeGen/RISCV/select-to-shift-and.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D137422.475094.patch
Type: text/x-patch
Size: 9917 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20221114/e3c39dd2/attachment.bin>