[PATCH] D118185: [AMDGPU] Mark v_cmp* convergent

[Ruiling, Song via Phabricator via llvm-commits]

ruiling added a comment.

In D118185#3272249 <https://reviews.llvm.org/D118185#3272249>, @foad wrote:

> Maybe another solution is to try to deprecate and remove amdgcn.icmp in favour of amdcgn.ballot, and try to lower ballot to a single machine instruction that is marked "convergent".

A second look, I agree with @foad using amdgcn.ballot is a good idea. Sounds like we need to fix the existing implementation of amdgcn.ballot as @foad's suggested.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D118185/new/

https://reviews.llvm.org/D118185