[PATCH] D101405: [AMDGPU] Change FLAT SADDR to VADDR form in moveToVALU

[Stanislav Mekhanoshin via Phabricator via llvm-commits]

rampitec added inline comments.


================
Comment at: llvm/lib/Target/AMDGPU/SIInstrInfo.cpp:5049-5063
+  if (OldVAddrIdx == NewVAddrIdx) {
+    assert(OldVAddrIdx == NewVAddrIdx);
+    MachineOperand &NewVAddr = Inst.getOperand(NewVAddrIdx);
+    MRI.removeRegOperandFromUseList(&NewVAddr);
+    MRI.moveOperands(&NewVAddr, &SAddr, 1);
+    Inst.RemoveOperand(OldSAddrIdx);
+    MRI.removeRegOperandFromUseList(&NewVAddr);
----------------
arsenm wrote:
> At this point wouldn't it be simpler to just create a fresh new instruction and delete the old one?
Unfortunately callers expect iterator to be intact. Otherwise probably yes, although it would not be less code.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D101405/new/

https://reviews.llvm.org/D101405