[PATCH] D154085: AMDGPU: Fold away mbcnt.hi in wave32 mode

[Jay Foad via Phabricator via llvm-commits]

foad accepted this revision.
foad added inline comments.
This revision is now accepted and ready to land.


================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUInstCombineIntrinsic.cpp:906
+  case Intrinsic::amdgcn_mbcnt_hi: {
+    // exec_hi is all 0, so this is just a copy.
+    if (ST->isWave32())
----------------
Nit: I find the comment slightly misleading, since it suggests to me that something is being ANDed with exec which is not true. Maybe say something like "ThreadPosition < 32" instead of "exec_hi is all 0"?


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D154085/new/

https://reviews.llvm.org/D154085