[PATCH] D113291: [WIP] Add LowerTableBasedCTZ and enable it for AARCH64 in -O3

[Djordje Todorovic via Phabricator via llvm-commits]

djtodoro added inline comments.


================
Comment at: llvm/include/llvm/InitializePasses.h:282
 void initializeLowerMatrixIntrinsicsMinimalLegacyPassPass(PassRegistry &);
+void initializeLowerTableBasedCTZLegacyPassPass(PassRegistry &);
 void initializeMIRAddFSDiscriminatorsPass(PassRegistry &);
----------------
craig.topper wrote:
> Can you use CTTZ instead of CTZ to match the llvm intrinsic name.
Sute, typo.

I am working on moving this to the AgressiveInstCombine.



================
Comment at: llvm/lib/Transforms/Scalar/LowerTableBasedCTZ.cpp:19
+// }
+// this can be lowerer __builtin_ctz(x) (where x != 0 ? __builtin_ctz(x) : 0).
+//
----------------
craig.topper wrote:
> "lowerer" -> "lowered to" I think?
Yep, thanks


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D113291/new/

https://reviews.llvm.org/D113291