[PATCH] D81991: [AMDGPU] Control num waves per EU for implicit work-group size

[Matt Arsenault via Phabricator via llvm-commits]

arsenm added a comment.

Should clang now be emitting 1 for the minimum range for OpenCL? Or should we go the other direction and have HIP bump the minimum required group size by default in the frontend?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D81991/new/

https://reviews.llvm.org/D81991