[PATCH] D124671: [AMDGPU] Do not raise wave priority beyond a specific number of VALU instructions.

[Jay Foad via Phabricator via llvm-commits]

foad added a comment.

I'm not sure how to review this. Can you explain why this heuristic makes sense intuitively? Or do you have any benchmarks to back it up? Or preferably both? :)



================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUSetWavePriority.cpp:106
+  // that backedges/loops, branch probabilities and other details can be
+  // ignored.
   SmallVector<const MachineBasicBlock *, 16> Worklist;
----------------
If you're ignoring loops, can you described succinctly what you *are* counting? Is it something like the minimum number of VALU instructions along any path from the start of the function to the VMEM load in question?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D124671/new/

https://reviews.llvm.org/D124671