[PATCH] D151126: [AMDGPU] Don't flush vmcnt for loops with use/def pairs

[Jay Foad via Phabricator via llvm-commits]

foad added inline comments.


================
Comment at: llvm/lib/Target/AMDGPU/SIInsertWaitcnts.cpp:1706
 
+static bool isVMEMOrFlatVMEM(const MachineInstr &MI,
+                             const SIInsertWaitcnts &IW) {
----------------
Please make this a member function.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D151126/new/

https://reviews.llvm.org/D151126