Index: lib/Target/X86/X86ISelLowering.cpp
===================================================================
--- lib/Target/X86/X86ISelLowering.cpp
+++ lib/Target/X86/X86ISelLowering.cpp
@@ -24002,6 +24002,7 @@
     SDValue Op1 = N.getOperand(1);
     SDValue Op2 = N.getOperand(2);
     unsigned InsertPSMask = cast<ConstantSDNode>(Op2)->getZExtValue();
+    unsigned SrcIdx = (InsertPSMask >> 6) & 0x3;
     unsigned DstIdx = (InsertPSMask >> 4) & 0x3;
     unsigned ZeroMask = InsertPSMask & 0xF;
 
@@ -24015,15 +24016,33 @@
       return DAG.getNode(X86ISD::INSERTPS, DL, VT, Op0, DAG.getUNDEF(VT),
                          DAG.getConstant(InsertPSMask, DL, MVT::i8));
 
-    SmallVector<int, 8> TargetMask;
-    if (!setTargetShuffleZeroElements(Op0, TargetMask))
+    // Attempt to merge insertps Op1 with an inner target shuffle node. 
+    SmallVector<int, 8> TargetMask1; 
+    if (setTargetShuffleZeroElements(Op1, TargetMask1)) { 
+      int M = TargetMask1[SrcIdx];
+      if (M < 0) { 
+        // Zero/UNDEF insertion - zero out element and remove dependency.
+        InsertPSMask |= (1u << DstIdx); 
+        return DAG.getNode(X86ISD::INSERTPS, DL, VT, Op0, DAG.getUNDEF(VT), 
+                           DAG.getConstant(InsertPSMask, DL, MVT::i8)); 
+      }
+      // Update insertps mask srcidx and reference the source input directly.
+      InsertPSMask = (InsertPSMask & 0x3f) | ((M & 0x3) << 6); 
+      Op1 = Op1.getOperand(M < 4 ? 0 : 1); 
+      return DAG.getNode(X86ISD::INSERTPS, DL, VT, Op0, Op1, 
+                         DAG.getConstant(InsertPSMask, DL, MVT::i8)); 
+    } 
+
+    // Attempt to merge insertps Op0 with an inner target shuffle node.
+    SmallVector<int, 8> TargetMask0;
+    if (!setTargetShuffleZeroElements(Op0, TargetMask0))
       return SDValue();
 
     bool Updated = false;
     bool UseInput00 = false;
     bool UseInput01 = false;
     for (int i = 0; i != 4; ++i) {
-      int M = TargetMask[i];
+      int M = TargetMask0[i];
       if ((InsertPSMask & (1u << i)) || (i == (int)DstIdx)) {
         // No change if element is already zero or the inserted element.
         continue;
Index: test/CodeGen/X86/sse41.ll
===================================================================
--- test/CodeGen/X86/sse41.ll
+++ test/CodeGen/X86/sse41.ll
@@ -846,16 +846,12 @@
 ; X32:       ## BB#0:
 ; X32-NEXT:    movl {{[0-9]+}}(%esp), %eax
 ; X32-NEXT:    movl {{[0-9]+}}(%esp), %ecx
-; X32-NEXT:    movss {{.*#+}} xmm1 = mem[0],zero,zero,zero
-; X32-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,0,0,0]
-; X32-NEXT:    insertps {{.*#+}} xmm0 = xmm0[0,1,2],xmm1[0]
+; X32-NEXT:    insertps {{.*#+}} xmm0 = xmm0[0,1,2],mem[0]
 ; X32-NEXT:    retl
 ;
 ; X64-LABEL: insertps_from_broadcast_loadf32:
 ; X64:       ## BB#0:
-; X64-NEXT:    movss {{.*#+}} xmm1 = mem[0],zero,zero,zero
-; X64-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,0,0,0]
-; X64-NEXT:    insertps {{.*#+}} xmm0 = xmm0[0,1,2],xmm1[0]
+; X64-NEXT:    insertps {{.*#+}} xmm0 = xmm0[0,1,2],mem[0]
 ; X64-NEXT:    retq
   %1 = getelementptr inbounds float, float* %fb, i64 %index
   %2 = load float, float* %1, align 4
@@ -871,16 +867,12 @@
 ; X32-LABEL: insertps_from_broadcast_loadv4f32:
 ; X32:       ## BB#0:
 ; X32-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X32-NEXT:    movups (%eax), %xmm1
-; X32-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,0,0,0]
-; X32-NEXT:    insertps {{.*#+}} xmm0 = xmm0[0,1,2],xmm1[0]
+; X32-NEXT:    insertps {{.*#+}} xmm0 = xmm0[0,1,2],mem[0]
 ; X32-NEXT:    retl
 ;
 ; X64-LABEL: insertps_from_broadcast_loadv4f32:
 ; X64:       ## BB#0:
-; X64-NEXT:    movups (%rdi), %xmm1
-; X64-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,0,0,0]
-; X64-NEXT:    insertps {{.*#+}} xmm0 = xmm0[0,1,2],xmm1[0]
+; X64-NEXT:    insertps {{.*#+}} xmm0 = xmm0[0,1,2],mem[0]
 ; X64-NEXT:    retq
   %1 = load <4 x float>, <4 x float>* %b, align 4
   %2 = extractelement <4 x float> %1, i32 0
@@ -892,14 +884,12 @@
   ret <4 x float> %7
 }
 
-;; FIXME: We're emitting an extraneous pshufd/vbroadcast.
 define <4 x float> @insertps_from_broadcast_multiple_use(<4 x float> %a, <4 x float> %b, <4 x float> %c, <4 x float> %d, float* nocapture readonly %fb, i64 %index) {
 ; X32-LABEL: insertps_from_broadcast_multiple_use:
 ; X32:       ## BB#0:
 ; X32-NEXT:    movl {{[0-9]+}}(%esp), %eax
 ; X32-NEXT:    movl {{[0-9]+}}(%esp), %ecx
 ; X32-NEXT:    movss {{.*#+}} xmm4 = mem[0],zero,zero,zero
-; X32-NEXT:    shufps {{.*#+}} xmm4 = xmm4[0,0,0,0]
 ; X32-NEXT:    insertps {{.*#+}} xmm0 = xmm0[0,1,2],xmm4[0]
 ; X32-NEXT:    insertps {{.*#+}} xmm1 = xmm1[0,1,2],xmm4[0]
 ; X32-NEXT:    insertps {{.*#+}} xmm2 = xmm2[0,1,2],xmm4[0]
@@ -912,7 +902,6 @@
 ; X64-LABEL: insertps_from_broadcast_multiple_use:
 ; X64:       ## BB#0:
 ; X64-NEXT:    movss {{.*#+}} xmm4 = mem[0],zero,zero,zero
-; X64-NEXT:    shufps {{.*#+}} xmm4 = xmm4[0,0,0,0]
 ; X64-NEXT:    insertps {{.*#+}} xmm0 = xmm0[0,1,2],xmm4[0]
 ; X64-NEXT:    insertps {{.*#+}} xmm1 = xmm1[0,1,2],xmm4[0]
 ; X64-NEXT:    insertps {{.*#+}} xmm2 = xmm2[0,1,2],xmm4[0]