[PATCH] D85530: [AMDGPU] Don't cluster stores

[Jay Foad via Phabricator via llvm-commits]

foad added a comment.

Some statistics for this change, from statically compiling 9785 graphics shaders from games for GFX9:
average vgpr_count decreases by 0.04%
average sgpr_count decreases by 0.001%
total number of instructions decreases by 21 (0.0004%)
number of s_waitcnt instructions decreases by 17 (0.0003%)
So the numbers are tiny but at least they moved in the right direction!


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D85530/new/

https://reviews.llvm.org/D85530