[PATCH] D37348: Tighten conditions for converting ISD::CTTZ_ZERO_UNDEF to ISD::CTTZ

[Brian Sumner via Phabricator via llvm-commits]

b-sumner added a comment.

In https://reviews.llvm.org/D37348#859119, @wdng wrote:

> Just add a custom lowering ISD:CTTZ to ISD::CTTZ_ZERO_UNDEF


I don't think that will help.  Why not follow exactly how CTLZ* is handled now and implement AMDGPUTargetLowering::LowerCTTZ making use of ffbl?


Repository:
  rL LLVM

https://reviews.llvm.org/D37348