1 files changed, 32 insertions, 3 deletions

diff --git a/lib/Target/ARM/ARMISelLowering.cpp b/lib/Target/ARM/ARMISelLowering.cpp
index 9dc103fa76..11ff6bf965 100644
--- a/lib/Target/ARM/ARMISelLowering.cpp
+++ b/lib/Target/ARM/ARMISelLowering.cpp
@@ -866,6 +866,7 @@ const char *ARMTargetLowering::getTargetNodeName(unsigned Opcode) const {
   case ARMISD::BFI:           return "ARMISD::BFI";
   case ARMISD::VORRIMM:       return "ARMISD::VORRIMM";
   case ARMISD::VBICIMM:       return "ARMISD::VBICIMM";
+  case ARMISD::VBSL:          return "ARMISD::VBSL";
   case ARMISD::VLD2DUP:       return "ARMISD::VLD2DUP";
   case ARMISD::VLD3DUP:       return "ARMISD::VLD3DUP";
   case ARMISD::VLD4DUP:       return "ARMISD::VLD4DUP";
@@ -5336,6 +5337,37 @@ static SDValue PerformORCombine(SDNode *N,
     }
   }
 
+  SDValue N0 = N->getOperand(0);
+  if (N0.getOpcode() != ISD::AND)
+    return SDValue();
+  SDValue N1 = N->getOperand(1);
+
+  // (or (and B, A), (and C, ~A)) => (VBSL A, B, C) when A is a constant.
+  if (Subtarget->hasNEON() && N1.getOpcode() == ISD::AND && VT.isVector() &&
+      DAG.getTargetLoweringInfo().isTypeLegal(VT)) {
+    APInt SplatUndef;
+    unsigned SplatBitSize;
+    bool HasAnyUndefs;
+
+    BuildVectorSDNode *BVN0 = dyn_cast<BuildVectorSDNode>(N0->getOperand(1));
+    APInt SplatBits0;
+    if (BVN0 && BVN0->isConstantSplat(SplatBits0, SplatUndef, SplatBitSize,
+                                  HasAnyUndefs) && !HasAnyUndefs) {
+      BuildVectorSDNode *BVN1 = dyn_cast<BuildVectorSDNode>(N1->getOperand(1));
+      APInt SplatBits1;
+      if (BVN1 && BVN1->isConstantSplat(SplatBits1, SplatUndef, SplatBitSize,
+                                    HasAnyUndefs) && !HasAnyUndefs &&
+          SplatBits0 == ~SplatBits1) {
+        // Canonicalize the vector type to make instruction selection simpler.
+        EVT CanonicalVT = VT.is128BitVector() ? MVT::v4i32 : MVT::v2i32;
+        SDValue Result = DAG.getNode(ARMISD::VBSL, dl, CanonicalVT,
+                                     N0->getOperand(1), N0->getOperand(0),
+                                     N1->getOperand(1));
+        return DAG.getNode(ISD::BITCAST, dl, VT, Result);
+      }
+    }
+  }
+
   // Try to use the ARM/Thumb2 BFI (bitfield insert) instruction when
   // reasonable.
 
@@ -5343,7 +5375,6 @@ static SDValue PerformORCombine(SDNode *N,
   if (Subtarget->isThumb1Only() || !Subtarget->hasV6T2Ops())
     return SDValue();
 
-  SDValue N0 = N->getOperand(0), N1 = N->getOperand(1);
   DebugLoc DL = N->getDebugLoc();
   // 1) or (and A, mask), val => ARMbfi A, val, mask
   //      iff (val & mask) == val
@@ -5354,8 +5385,6 @@ static SDValue PerformORCombine(SDNode *N,
   //  2b) iff isBitFieldInvertedMask(~mask) && isBitFieldInvertedMask(mask2)
   //          && ~mask == mask2
   //  (i.e., copy a bitfield value into another bitfield of the same width)
-  if (N0.getOpcode() != ISD::AND)
-    return SDValue();
 
   if (VT != MVT::i32)
     return SDValue();