[PATCH] D94746: [AMDGPU] Move kill lowering to WQM pass and add live mask tracking

[Piotr Sobczak via Phabricator via llvm-commits]

piotr added inline comments.


================
Comment at: llvm/lib/Target/AMDGPU/SIWholeQuadMode.cpp:618-626
+bool SIWholeQuadMode::canSplitBlockAt(MachineBasicBlock *BB, MachineInstr *MI) {
+  // Cannot split immediately before the epilog
+  // because there are values in physical registers
+  if (MI->getOpcode() == AMDGPU::SI_RETURN_TO_EPILOG) {
+    return false;
+  }
+
----------------
Unused?


================
Comment at: llvm/lib/Target/AMDGPU/SIWholeQuadMode.cpp:636
+
+  // Convert last instruction in to a terminator.
+  // Note: this only covers the expected patterns
----------------
A missing word after "in"?


================
Comment at: llvm/lib/Target/AMDGPU/SIWholeQuadMode.cpp:866
+  LIS->InsertMachineInstrInMaps(*EarlyTermMI);
+  if (WQMMaskMI)
+    LIS->InsertMachineInstrInMaps(*WQMMaskMI);
----------------
WQMMaskMI is always nullptr here.


================
Comment at: llvm/lib/Target/AMDGPU/SIWholeQuadMode.cpp:1048
                               MachineBasicBlock::iterator Before,
-                              unsigned SavedOrig) {
+                              Register SavedOrig, char NonWWMState) {
   MachineInstr *MI;
----------------
It looks the new param NonWWMState is unused?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D94746/new/

https://reviews.llvm.org/D94746