diff --git a/lib/Target/X86/X86ISelLowering.cpp b/lib/Target/X86/X86ISelLowering.cpp
index 07451a5e749..4409389e7c4 100644
--- a/lib/Target/X86/X86ISelLowering.cpp
+++ b/lib/Target/X86/X86ISelLowering.cpp
@@ -5161,11 +5161,9 @@ X86TargetLowering::LowerBUILD_VECTOR(SDValue Op, SelectionDAG &DAG) const {
       if (ExtVT == MVT::i32 || ExtVT == MVT::f32 || ExtVT == MVT::f64 ||
           (ExtVT == MVT::i64 && Subtarget->is64Bit())) {
         if (VT.getSizeInBits() == 256) {
-          EVT VT128 = EVT::getVectorVT(*DAG.getContext(), ExtVT, NumElems / 2);
-          Item = DAG.getNode(ISD::SCALAR_TO_VECTOR, dl, VT128, Item);
           SDValue ZeroVec = getZeroVector(VT, true, DAG, dl);
-          return Insert128BitVector(ZeroVec, Item, DAG.getConstant(0, MVT::i32),
-                              DAG, dl);
+          return DAG.getNode(ISD::INSERT_VECTOR_ELT, dl, VT, ZeroVec,
+                             Item, DAG.getIntPtrConstant(0));
         }
         assert(VT.getSizeInBits() == 128 && "Expected an SSE value type!");
         Item = DAG.getNode(ISD::SCALAR_TO_VECTOR, dl, VT, Item);
diff --git a/test/CodeGen/X86/2012-1-10-buildvector.ll b/test/CodeGen/X86/2012-1-10-buildvector.ll
index 4e7cbc6c14e..ff6be369dc5 100644
--- a/test/CodeGen/X86/2012-1-10-buildvector.ll
+++ b/test/CodeGen/X86/2012-1-10-buildvector.ll
@@ -1,13 +1,26 @@
-; RUN: llc < %s -march=x86 -mcpu=corei7-avx -mattr=+avx -mtriple=i686-pc-win32
+; RUN: llc < %s -march=x86 -mcpu=corei7-avx -mattr=+avx -mtriple=i686-pc-win32 | FileCheck %s
 
 target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-f80:128:128-v64:64:64-v128:128:128-a0:0:64-f80:32:32-n8:16:32-S32"
 target triple = "i686-pc-win32"
 
+;CHECK: bad_cast
 define void @bad_cast() {
 entry:
   %vext.i = shufflevector <2 x i64> undef, <2 x i64> undef, <3 x i32> <i32 0, i32 1, i32 undef>
   %vecinit8.i = shufflevector <3 x i64> zeroinitializer, <3 x i64> %vext.i, <3 x i32> <i32 0, i32 3, i32 4>
   store <3 x i64> %vecinit8.i, <3 x i64>* undef, align 32
-  unreachable
+;CHECK: ret
+  ret void
+}
+
+
+;CHECK: bad_insert
+define void @bad_insert(i32 %t) {
+entry:
+;CHECK: vpinsrd
+  %v2 = insertelement <8 x i32> zeroinitializer, i32 %t, i32 0
+  store <8 x i32> %v2, <8 x i32> addrspace(1)* undef, align 32
+;CHECK: ret
+  ret void
 }
 
diff --git a/test/CodeGen/X86/avx-load-store.ll b/test/CodeGen/X86/avx-load-store.ll
index 07a63efd71f..c9fc66a8a79 100644
--- a/test/CodeGen/X86/avx-load-store.ll
+++ b/test/CodeGen/X86/avx-load-store.ll
@@ -25,20 +25,26 @@ declare void @dummy(<4 x double>, <8 x float>, <4 x i64>)
 
 ;;
 ;; The two tests below check that we must fold load + scalar_to_vector
-;; + ins_subvec+ zext into only a single vmovss or vmovsd
+;; + ins_subvec+ zext into only a single vmovss or vmovsd or vinsertps from memory
 
-; CHECK: vmovss (%
+; CHECK: mov00
 define <8 x float> @mov00(<8 x float> %v, float * %ptr) nounwind {
   %val = load float* %ptr
+; CHECK: vinsertps
+; CHECK: vinsertf128
   %i0 = insertelement <8 x float> zeroinitializer, float %val, i32 0
   ret <8 x float> %i0
+; CHECK: ret
 }
 
-; CHECK: vmovsd (%
+; CHECK: mov01
 define <4 x double> @mov01(<4 x double> %v, double * %ptr) nounwind {
   %val = load double* %ptr
+; CHECK: vmovlpd
+; CHECK: vinsertf128
   %i0 = insertelement <4 x double> zeroinitializer, double %val, i32 0
   ret <4 x double> %i0
+; CHECK: ret
 }
 
 ; CHECK: vmovaps  %ymm