[PowerPC] Correct missue of getOperandConstraint in PPCInstrInfo::commuteInstructionImpl

[craig.topper]

Summary

getOperandConstraint does not return a bool, it returns an int. It
returns -1 if there is no TIED_TO.

Additionally, TIED_TO is only set on use operands not defs and it
points to the def that the use is tied to. So calling it on operand 0
is guaranteed to return -1.

As far as I can tell this code must have been copied from the
generic implementation prior to 6aa2744bed0b8.o

Unfortunately, this code is not executed in lit tests. I just happened
to notice it while looking for other uses of TIED_TO for something
I was working on.

[craig.topper]

Ping

[nemanjai]

Sorry I missed this. LGTM.

llvm/lib/Target/PowerPC/PPCInstrInfo.cpp

1176–1177

Maybe clarify here with a comment such as:

// Must be two address instruction (i.e. op1 is tied to op0).

llvm/lib/Target/PowerPC/PPCInstrInfo.cpp

MachineInstr *PPCInstrInfo::commuteInstructionImpl(MachineInstr &MI, bool NewMI,
   Register Reg2 = MI.getOperand(2).getReg();
   unsigned SubReg1 = MI.getOperand(1).getSubReg();
   unsigned SubReg2 = MI.getOperand(2).getSubReg();
   bool Reg1IsKill = MI.getOperand(1).isKill();
   bool Reg2IsKill = MI.getOperand(2).isKill();
   bool ChangeReg0 = false;
   // If machine instrs are no longer in two-address forms, update
   // destination register as well.
   if (Reg0 == Reg1) {
-    // Must be two address instruction!
-    assert(MI.getDesc().getOperandConstraint(0, MCOI::TIED_TO) &&
+    // Must be two address instruction (i.e. op1 is tied to op0).
+    assert(MI.getDesc().getOperandConstraint(1, MCOI::TIED_TO) == 0 &&
            "Expecting a two-address instruction!");
     assert(MI.getOperand(0).getSubReg() == SubReg1 && "Tied subreg mismatch");
     Reg2IsKill = false;
     ChangeReg0 = true;
   }
 
   // Masks.
   unsigned MB = MI.getOperand(4).getImm();