[PATCH] D83626: [AMDGPU/MemOpsCluster] Guard new mem ops clustering heuristic logic by a flag

[Stanislav Mekhanoshin via Phabricator via llvm-commits]

rampitec added inline comments.


================
Comment at: llvm/lib/Target/AMDGPU/SIInstrInfo.cpp:562
+    }
+    return NumLoads <= MaxNumLoads;
   }
----------------
You have problem with extremely wide loads. I am not sure what was in the regression case, but probably something like 8 longs or so. Isn't it better to tweak it instead and just clamp based on the NumBytes as it supposed to be? You are saying you are checking NumBytes, but the return is solely based on NumLoads.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D83626/new/

https://reviews.llvm.org/D83626