[AArch64] Utilize XAR for certain vector rotates

Resolves #137162

For cases when there isn't any `XOR` in the transformation,
replace with a zero register.

Author: Rajveer100

llvm/lib/Target/AArch64/AArch64ISelDAGToDAG.cpp

@@ -4558,9 +4558,27 @@ bool AArch64DAGToDAGISel::trySelectXAR(SDNode *N) {
         !TLI->isAllActivePredicate(*CurDAG, N1.getOperand(0)))
       return false;
 
-    SDValue XOR = N0.getOperand(1);
-    if (XOR.getOpcode() != ISD::XOR || XOR != N1.getOperand(1))
-      return false;
+    SDValue R1, R2;
+    if (N0.getOperand(1).getOpcode() != ISD::XOR) {
+      if (N0.getOperand(1) != N1.getOperand(1))
+        return false;
+      SDLoc DL(N1->getOperand(0));
+      EVT VT = N1->getOperand(0).getValueType();
+
+      SDValue Zero = CurDAG->getTargetConstant(0, DL, MVT::i64);
+      SDNode *MOV = CurDAG->getMachineNode(AArch64::MOVIv2d_ns, DL, VT, Zero);
+      SDValue MOVIV = SDValue(MOV, 0);
+
+      SDValue ZSub = CurDAG->getTargetConstant(AArch64::zsub, DL, MVT::i64);
+      SDNode *SubRegToReg = CurDAG->getMachineNode(AArch64::SUBREG_TO_REG, DL,
+                                                   MVT::i64, Zero, MOVIV, ZSub);
+
+      R1 = N1->getOperand(0);
+      R2 = SDValue(SubRegToReg, 0);
+    } else {
+      R1 = N0.getOperand(1);
+      R2 = N1.getOperand(1);
+    }
 
     APInt ShlAmt, ShrAmt;
     if (!ISD::isConstantSplatVector(N0.getOperand(2).getNode(), ShlAmt) ||
@@ -4574,7 +4592,7 @@ bool AArch64DAGToDAGISel::trySelectXAR(SDNode *N) {
     SDValue Imm =
         CurDAG->getTargetConstant(ShrAmt.getZExtValue(), DL, MVT::i32);
 
-    SDValue Ops[] = {XOR.getOperand(0), XOR.getOperand(1), Imm};
+    SDValue Ops[] = {R1, R2, Imm};
     if (auto Opc = SelectOpcodeFromVT<SelectTypeKind::Int>(
             VT, {AArch64::XAR_ZZZI_B, AArch64::XAR_ZZZI_H, AArch64::XAR_ZZZI_S,
                  AArch64::XAR_ZZZI_D})) {
@@ -4591,13 +4609,24 @@ bool AArch64DAGToDAGISel::trySelectXAR(SDNode *N) {
       N1->getOpcode() != AArch64ISD::VLSHR)
     return false;
 
-  if (N0->getOperand(0) != N1->getOperand(0) ||
-      N1->getOperand(0)->getOpcode() != ISD::XOR)
+  if (N0->getOperand(0) != N1->getOperand(0))
     return false;
 
-  SDValue XOR = N0.getOperand(0);
-  SDValue R1 = XOR.getOperand(0);
-  SDValue R2 = XOR.getOperand(1);
+  SDValue R1, R2;
+  if (N1->getOperand(0)->getOpcode() != ISD::XOR) {
+    SDLoc DL(N1->getOperand(0));
+    EVT VT = N1->getOperand(0).getValueType();
+
+    SDValue Zero = CurDAG->getTargetConstant(0, DL, MVT::i64);
+    SDNode *MOV = CurDAG->getMachineNode(AArch64::MOVIv2d_ns, DL, VT, Zero);
+    SDValue MOVIV = SDValue(MOV, 0);
+    R1 = N1->getOperand(0);
+    R2 = MOVIV;
+  } else {
+    SDValue XOR = N0.getOperand(0);
+    R1 = XOR.getOperand(0);
+    R2 = XOR.getOperand(1);
+  }
 
   unsigned HsAmt = N0.getConstantOperandVal(1);
   unsigned ShAmt = N1.getConstantOperandVal(1);