[PATCH] D128252: [AMDGPU] Lowering VGPR to SGPR copies to v_readfirstlane_b32 if profitable.

[Nicolai Hähnle via Phabricator via llvm-commits]

nhaehnle added a comment.

Perhaps I'm blind, but I don't see where the heuristic takes into account the number of ALU instructions that would be moved from SALU to VALU.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D128252/new/

https://reviews.llvm.org/D128252