Fix an ancient bug in removeCopyByCommutingDef().

Before rewriting uses of one value in A to register B, check that there
are no tied uses. That would require multiple A values to be rewritten.

This bug can't bite in the current version of the code for a fairly
subtle reason: A tied use would have caused 2-addr to insert a copy
before the use. If the copy has been coalesced, it will be found by the
same loop changed by this patch, and the optimization is aborted.

This was exposed by 400.perlbench and lua after applying a patch that
deletes joined copies aggressively.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@157130 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/CodeGen/RegisterCoalescer.cpp b/lib/CodeGen/RegisterCoalescer.cpp
index dbd2cdd8d38ce00fb79d2840be9a6b91067250ef..5e876df8662ab1706d30a4d5c2ce8f45b3331dda 100644 (file)
--- a/lib/CodeGen/RegisterCoalescer.cpp
+++ b/lib/CodeGen/RegisterCoalescer.cpp
@@ -687,9 +687,12 @@ bool RegisterCoalescer::removeCopyByCommutingDef(const CoalescerPair &CP,
     MachineInstr *UseMI = &*UI;
     SlotIndex UseIdx = LIS->getInstructionIndex(UseMI);
     LiveInterval::iterator ULR = IntA.FindLiveRangeContaining(UseIdx);
-    if (ULR == IntA.end())
+    if (ULR == IntA.end() || ULR->valno != AValNo)
       continue;
-    if (ULR->valno == AValNo && JoinedCopies.count(UseMI))
+    if (JoinedCopies.count(UseMI))
+      return false;
+    // If this use is tied to a def, we can't rewrite the register.
+    if (UseMI->isRegTiedToDefOperand(UI.getOperandNo()))
       return false;
   }