[llvm-branch-commits] [llvm] [AMDGPU] Make AMDGPURewriteAGPRCopyMFMA aware of subreg reload (PR #174998)

[Christudasan Devadasan via llvm-branch-commits]

================
@@ -603,6 +603,18 @@ unsigned TargetRegisterInfo::getSubRegIdxOffset(unsigned Idx) const {
   return SubRegIdxRanges[HwMode * getNumSubRegIndices() + Idx].Offset;
 }
 
+unsigned TargetRegisterInfo::getSubRegIdxFromOffsetSize(unsigned Offset,
+                                                        unsigned Size) const {
+  unsigned NumIdx = getNumSubRegIndices();
+  unsigned Base = HwMode * NumIdx;
+  for (unsigned Idx = 1; Idx < NumIdx; Idx++) {
+    if (SubRegIdxRanges[Base + Idx].Offset == Offset &&
+        SubRegIdxRanges[Base + Idx].Size == Size)
+      return Idx;
+  }
+  return 0;
----------------
cdevadas wrote:

```suggestion
  return 0 /*NoSubRegister*/;
```

https://github.com/llvm/llvm-project/pull/174998