Index: lib/Target/X86/X86ISelLowering.cpp
===================================================================
--- lib/Target/X86/X86ISelLowering.cpp
+++ lib/Target/X86/X86ISelLowering.cpp
@@ -5595,6 +5595,24 @@
     }
   }
 
+  auto CreateLoad = [&DAG, &DL](EVT VT, LoadSDNode *LDBase) {
+    SDValue NewLd = DAG.getLoad(VT, DL, LDBase->getChain(),
+                                LDBase->getBasePtr(), LDBase->getPointerInfo(),
+                                LDBase->isVolatile(), LDBase->isNonTemporal(),
+                                LDBase->isInvariant(), LDBase->getAlignment());
+
+    if (LDBase->hasAnyUseOfValue(1)) {
+      SDValue NewChain =
+          DAG.getNode(ISD::TokenFactor, DL, MVT::Other, SDValue(LDBase, 1),
+                      SDValue(NewLd.getNode(), 1));
+      DAG.ReplaceAllUsesOfValueWith(SDValue(LDBase, 1), NewChain);
+      DAG.UpdateNodeOperands(NewChain.getNode(), SDValue(LDBase, 1),
+                             SDValue(NewLd.getNode(), 1));
+    }
+
+    return NewLd;
+  };
+
   // LOAD - all consecutive load/undefs (must start/end with a load).
   // If we have found an entire vector of loads and undefs, then return a large
   // load of the entire vector width starting at the base pointer.
@@ -5610,23 +5628,7 @@
     if (isAfterLegalize && !TLI.isOperationLegal(ISD::LOAD, VT))
       return SDValue();
 
-    SDValue NewLd = SDValue();
-
-    NewLd = DAG.getLoad(VT, DL, LDBase->getChain(), LDBase->getBasePtr(),
-                        LDBase->getPointerInfo(), LDBase->isVolatile(),
-                        LDBase->isNonTemporal(), LDBase->isInvariant(),
-                        LDBase->getAlignment());
-
-    if (LDBase->hasAnyUseOfValue(1)) {
-      SDValue NewChain =
-          DAG.getNode(ISD::TokenFactor, DL, MVT::Other, SDValue(LDBase, 1),
-                      SDValue(NewLd.getNode(), 1));
-      DAG.ReplaceAllUsesOfValueWith(SDValue(LDBase, 1), NewChain);
-      DAG.UpdateNodeOperands(NewChain.getNode(), SDValue(LDBase, 1),
-                             SDValue(NewLd.getNode(), 1));
-    }
-
-    return NewLd;
+    return CreateLoad(VT, LDBase);
   }
 
   int LoadSize =
@@ -5660,6 +5662,18 @@
 
     return DAG.getBitcast(VT, ResNode);
   }
+
+  // VZEXT_MOVL - consecutive 32-bit load/undefs followed by zeros/undefs.
+  if (IsConsecutiveLoad && FirstLoadedElt == 0 && LoadSize == 32 &&
+      ((VT.is128BitVector() && TLI.isTypeLegal(MVT::v4f32)) ||
+       (VT.is256BitVector() && TLI.isTypeLegal(MVT::v8f32)))) {
+    MVT VecVT = VT.is256BitVector() ? MVT::v8f32 : MVT::v4f32;
+    SDValue V = CreateLoad(MVT::f32, LDBase);
+    V = DAG.getNode(ISD::SCALAR_TO_VECTOR, DL, VecVT, V);
+    V = DAG.getNode(X86ISD::VZEXT_MOVL, DL, VecVT, V);
+    return DAG.getBitcast(VT, V);
+  }
+
   return SDValue();
 }
 
Index: test/CodeGen/X86/merge-consecutive-loads-128.ll
===================================================================
--- test/CodeGen/X86/merge-consecutive-loads-128.ll
+++ test/CodeGen/X86/merge-consecutive-loads-128.ll
@@ -326,14 +326,12 @@
 define <8 x i16> @merge_8i16_i16_34uuuuuu(i16* %ptr) nounwind uwtable noinline ssp {
 ; SSE-LABEL: merge_8i16_i16_34uuuuuu:
 ; SSE:       # BB#0:
-; SSE-NEXT:    pinsrw $0, 6(%rdi), %xmm0
-; SSE-NEXT:    pinsrw $1, 8(%rdi), %xmm0
+; SSE-NEXT:    movss {{.*#+}} xmm0 = mem[0],zero,zero,zero
 ; SSE-NEXT:    retq
 ;
 ; AVX-LABEL: merge_8i16_i16_34uuuuuu:
 ; AVX:       # BB#0:
-; AVX-NEXT:    vpinsrw $0, 6(%rdi), %xmm0, %xmm0
-; AVX-NEXT:    vpinsrw $1, 8(%rdi), %xmm0, %xmm0
+; AVX-NEXT:    vmovss {{.*#+}} xmm0 = mem[0],zero,zero,zero
 ; AVX-NEXT:    retq
   %ptr0 = getelementptr inbounds i16, i16* %ptr, i64 3
   %ptr1 = getelementptr inbounds i16, i16* %ptr, i64 4
@@ -426,33 +424,14 @@
 }
 
 define <16 x i8> @merge_16i8_i8_01u3uuzzuuuuuzzz(i8* %ptr) nounwind uwtable noinline ssp {
-; SSE2-LABEL: merge_16i8_i8_01u3uuzzuuuuuzzz:
-; SSE2:       # BB#0:
-; SSE2-NEXT:    movzbl (%rdi), %eax
-; SSE2-NEXT:    movzbl 1(%rdi), %ecx
-; SSE2-NEXT:    shll $8, %ecx
-; SSE2-NEXT:    orl %eax, %ecx
-; SSE2-NEXT:    pxor %xmm0, %xmm0
-; SSE2-NEXT:    pinsrw $0, %ecx, %xmm0
-; SSE2-NEXT:    movzbl 3(%rdi), %eax
-; SSE2-NEXT:    shll $8, %eax
-; SSE2-NEXT:    pinsrw $1, %eax, %xmm0
-; SSE2-NEXT:    retq
-;
-; SSE41-LABEL: merge_16i8_i8_01u3uuzzuuuuuzzz:
-; SSE41:       # BB#0:
-; SSE41-NEXT:    pxor %xmm0, %xmm0
-; SSE41-NEXT:    pinsrb $0, (%rdi), %xmm0
-; SSE41-NEXT:    pinsrb $1, 1(%rdi), %xmm0
-; SSE41-NEXT:    pinsrb $3, 3(%rdi), %xmm0
-; SSE41-NEXT:    retq
+; SSE-LABEL: merge_16i8_i8_01u3uuzzuuuuuzzz:
+; SSE:       # BB#0:
+; SSE-NEXT:    movss {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; SSE-NEXT:    retq
 ;
 ; AVX-LABEL: merge_16i8_i8_01u3uuzzuuuuuzzz:
 ; AVX:       # BB#0:
-; AVX-NEXT:    vpxor %xmm0, %xmm0, %xmm0
-; AVX-NEXT:    vpinsrb $0, (%rdi), %xmm0, %xmm0
-; AVX-NEXT:    vpinsrb $1, 1(%rdi), %xmm0, %xmm0
-; AVX-NEXT:    vpinsrb $3, 3(%rdi), %xmm0, %xmm0
+; AVX-NEXT:    vmovss {{.*#+}} xmm0 = mem[0],zero,zero,zero
 ; AVX-NEXT:    retq
   %ptr0 = getelementptr inbounds i8, i8* %ptr, i64 0
   %ptr1 = getelementptr inbounds i8, i8* %ptr, i64 1
Index: test/CodeGen/X86/merge-consecutive-loads-256.ll
===================================================================
--- test/CodeGen/X86/merge-consecutive-loads-256.ll
+++ test/CodeGen/X86/merge-consecutive-loads-256.ll
@@ -347,21 +347,10 @@
 }
 
 define <16 x i16> @merge_16i16_i16_89zzzuuuuuuuuuuuz(i16* %ptr) nounwind uwtable noinline ssp {
-; AVX1-LABEL: merge_16i16_i16_89zzzuuuuuuuuuuuz:
-; AVX1:       # BB#0:
-; AVX1-NEXT:    vpxor %xmm0, %xmm0, %xmm0
-; AVX1-NEXT:    vpinsrw $0, 16(%rdi), %xmm0, %xmm1
-; AVX1-NEXT:    vpinsrw $1, 18(%rdi), %xmm1, %xmm1
-; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
-; AVX1-NEXT:    retq
-;
-; AVX2-LABEL: merge_16i16_i16_89zzzuuuuuuuuuuuz:
-; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %xmm0, %xmm0, %xmm0
-; AVX2-NEXT:    vpinsrw $0, 16(%rdi), %xmm0, %xmm1
-; AVX2-NEXT:    vpinsrw $1, 18(%rdi), %xmm1, %xmm1
-; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
-; AVX2-NEXT:    retq
+; AVX-LABEL: merge_16i16_i16_89zzzuuuuuuuuuuuz:
+; AVX:       # BB#0:
+; AVX-NEXT:    vmovss {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; AVX-NEXT:    retq
   %ptr0 = getelementptr inbounds i16, i16* %ptr, i64 8
   %ptr1 = getelementptr inbounds i16, i16* %ptr, i64 9
   %val0 = load i16, i16* %ptr0
@@ -459,9 +448,7 @@
 define <32 x i8> @merge_32i8_i8_45u7uuuuuuuuuuuuuuuuuuuuuuuuuuuu(i8* %ptr) nounwind uwtable noinline ssp {
 ; AVX-LABEL: merge_32i8_i8_45u7uuuuuuuuuuuuuuuuuuuuuuuuuuuu:
 ; AVX:       # BB#0:
-; AVX-NEXT:    vpinsrb $0, 4(%rdi), %xmm0, %xmm0
-; AVX-NEXT:    vpinsrb $1, 5(%rdi), %xmm0, %xmm0
-; AVX-NEXT:    vpinsrb $3, 7(%rdi), %xmm0, %xmm0
+; AVX-NEXT:    vmovss {{.*#+}} xmm0 = mem[0],zero,zero,zero
 ; AVX-NEXT:    retq
   %ptr0 = getelementptr inbounds i8, i8* %ptr, i64 4
   %ptr1 = getelementptr inbounds i8, i8* %ptr, i64 5
@@ -476,23 +463,10 @@
 }
 
 define <32 x i8> @merge_32i8_i8_23u5uuuuuuuuuuzzzzuuuuuuuuuuuuuu(i8* %ptr) nounwind uwtable noinline ssp {
-; AVX1-LABEL: merge_32i8_i8_23u5uuuuuuuuuuzzzzuuuuuuuuuuuuuu:
-; AVX1:       # BB#0:
-; AVX1-NEXT:    vpxor %xmm0, %xmm0, %xmm0
-; AVX1-NEXT:    vpinsrb $0, 4(%rdi), %xmm0, %xmm1
-; AVX1-NEXT:    vpinsrb $1, 5(%rdi), %xmm1, %xmm1
-; AVX1-NEXT:    vpinsrb $3, 7(%rdi), %xmm1, %xmm1
-; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
-; AVX1-NEXT:    retq
-;
-; AVX2-LABEL: merge_32i8_i8_23u5uuuuuuuuuuzzzzuuuuuuuuuuuuuu:
-; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %xmm0, %xmm0, %xmm0
-; AVX2-NEXT:    vpinsrb $0, 4(%rdi), %xmm0, %xmm1
-; AVX2-NEXT:    vpinsrb $1, 5(%rdi), %xmm1, %xmm1
-; AVX2-NEXT:    vpinsrb $3, 7(%rdi), %xmm1, %xmm1
-; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
-; AVX2-NEXT:    retq
+; AVX-LABEL: merge_32i8_i8_23u5uuuuuuuuuuzzzzuuuuuuuuuuuuuu:
+; AVX:       # BB#0:
+; AVX-NEXT:    vmovss {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; AVX-NEXT:    retq
   %ptr0 = getelementptr inbounds i8, i8* %ptr, i64 4
   %ptr1 = getelementptr inbounds i8, i8* %ptr, i64 5
   %ptr3 = getelementptr inbounds i8, i8* %ptr, i64 7