[PATCH] D65966: AMDGPU/SILoadStoreOptimizer: Improve merging of out of order offsets

[Matt Arsenault via Phabricator via llvm-commits]

arsenm added inline comments.


================
Comment at: llvm/lib/Target/AMDGPU/SILoadStoreOptimizer.cpp:232
+  MachineBasicBlock::iterator mergeRead2Pair(CombineInfo &CI,
+                                             std::list<CombineInfo> &MergeList);
 
----------------
Why std::list?


================
Comment at: llvm/lib/Target/AMDGPU/SILoadStoreOptimizer.cpp:273-275
+  void replaceInstsInList(std::list<CombineInfo> &MergeList,
+                                             const CombineInfo &CI,
+                                             MachineBasicBlock::iterator NewMI);
----------------
Weird format


================
Comment at: llvm/lib/Target/AMDGPU/SILoadStoreOptimizer.cpp:1595-1599
+  // FIXME: Is there a better way to do this.
+  const MachineBasicBlock *MBB = First->getParent();
+  for (auto I = First, E = MBB->end(); I != E; ++I) {
+    if (I == Second)
+      return true;
----------------
Don't you know which is first from which was encountered first?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D65966/new/

https://reviews.llvm.org/D65966