diff --git a/llvm/lib/Target/X86/X86ISelLowering.cpp b/llvm/lib/Target/X86/X86ISelLowering.cpp
--- a/llvm/lib/Target/X86/X86ISelLowering.cpp
+++ b/llvm/lib/Target/X86/X86ISelLowering.cpp
@@ -36226,12 +36226,61 @@
       IsBlend = false;
       break;
     }
-    if (IsBlend &&
-        DAG.computeKnownBits(V1, DemandedZeroV1).isZero() &&
-        DAG.computeKnownBits(V2, DemandedZeroV2).isZero()) {
-      Shuffle = ISD::OR;
-      SrcVT = DstVT = MaskVT.changeTypeToInteger();
-      return true;
+    if (IsBlend) {
+      if (DAG.computeKnownBits(V1, DemandedZeroV1).isZero() &&
+          DAG.computeKnownBits(V2, DemandedZeroV2).isZero()) {
+        Shuffle = ISD::OR;
+        SrcVT = DstVT = MaskVT.changeTypeToInteger();
+        return true;
+      } else if (NumV1Elts == NumV2Elts && NumV1Elts == NumMaskElts) {
+        APInt V1KnownZeroElts = APInt::getZero(NumV1Elts);
+        APInt V1KnownOneElts = APInt::getZero(NumV1Elts);
+        for (unsigned V1EltIdx = 0; V1EltIdx != NumV1Elts; ++V1EltIdx) {
+          APInt Mask = APInt::getZero(NumV1Elts);
+          Mask.setBit(V1EltIdx);
+          KnownBits ZZ = DAG.computeKnownBits(V1, Mask);
+          if (ZZ.isZero())
+            V1KnownZeroElts.setBit(V1EltIdx);
+          if (ZZ.isAllOnes())
+            V1KnownOneElts.setBit(V1EltIdx);
+        }
+
+        APInt V2KnownZeroElts = APInt::getZero(NumV2Elts);
+        APInt V2KnownOneElts = APInt::getZero(NumV2Elts);
+        for (unsigned V2EltIdx = 0; V2EltIdx != NumV2Elts; ++V2EltIdx) {
+          APInt Mask = APInt::getZero(NumV2Elts);
+          Mask.setBit(V2EltIdx);
+          KnownBits ZZ = DAG.computeKnownBits(V2, Mask);
+          if (ZZ.isZero())
+            V2KnownZeroElts.setBit(V2EltIdx);
+          if (ZZ.isAllOnes())
+            V2KnownOneElts.setBit(V2EltIdx);
+        }
+
+        for (unsigned i = 0; i != NumMaskElts && IsBlend; ++i) {
+          int M = Mask[i];
+          if (M == SM_SentinelUndef)
+            continue;
+          if (M == SM_SentinelZero) {
+            IsBlend &= V1KnownZeroElts[i] && V2KnownZeroElts[i];
+            continue;
+          }
+          if (M == (int)i) {
+            IsBlend &= V2KnownZeroElts[i] || V1KnownOneElts[i];
+            continue;
+          }
+          if (M == (int)(i + NumMaskElts)) {
+            IsBlend &= V1KnownZeroElts[i] || V2KnownOneElts[i];
+            continue;
+          }
+          llvm_unreachable("will not get here.");
+        }
+        if (IsBlend) {
+          Shuffle = ISD::OR;
+          SrcVT = DstVT = MaskVT.changeTypeToInteger();
+          return true;
+        }
+      }
     }
   }
 
diff --git a/llvm/test/CodeGen/X86/insertelement-ones.ll b/llvm/test/CodeGen/X86/insertelement-ones.ll
--- a/llvm/test/CodeGen/X86/insertelement-ones.ll
+++ b/llvm/test/CodeGen/X86/insertelement-ones.ll
@@ -280,11 +280,8 @@
 ;
 ; AVX1-LABEL: insert_v16i16_x12345x789ABCDEx:
 ; AVX1:       # %bb.0:
-; AVX1-NEXT:    vandps {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %ymm0, %ymm0
 ; AVX1-NEXT:    vorps {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %ymm0, %ymm0
-; AVX1-NEXT:    vandps {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %ymm0, %ymm0
 ; AVX1-NEXT:    vorps {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %ymm0, %ymm0
-; AVX1-NEXT:    vandps {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %ymm0, %ymm0
 ; AVX1-NEXT:    vorps {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %ymm0, %ymm0
 ; AVX1-NEXT:    retq
 ;
@@ -315,26 +312,22 @@
 ; SSE2-LABEL: insert_v16i8_x123456789ABCDEx:
 ; SSE2:       # %bb.0:
 ; SSE2-NEXT:    movdqa {{.*#+}} xmm1 = [0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
-; SSE2-NEXT:    pand %xmm1, %xmm0
 ; SSE2-NEXT:    movl $255, %eax
 ; SSE2-NEXT:    movd %eax, %xmm2
 ; SSE2-NEXT:    pandn %xmm2, %xmm1
-; SSE2-NEXT:    por %xmm1, %xmm0
-; SSE2-NEXT:    pand {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
 ; SSE2-NEXT:    pslldq {{.*#+}} xmm2 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm2[0]
+; SSE2-NEXT:    por %xmm1, %xmm2
 ; SSE2-NEXT:    por %xmm2, %xmm0
 ; SSE2-NEXT:    retq
 ;
 ; SSE3-LABEL: insert_v16i8_x123456789ABCDEx:
 ; SSE3:       # %bb.0:
 ; SSE3-NEXT:    movdqa {{.*#+}} xmm1 = [0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
-; SSE3-NEXT:    pand %xmm1, %xmm0
 ; SSE3-NEXT:    movl $255, %eax
 ; SSE3-NEXT:    movd %eax, %xmm2
 ; SSE3-NEXT:    pandn %xmm2, %xmm1
-; SSE3-NEXT:    por %xmm1, %xmm0
-; SSE3-NEXT:    pand {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
 ; SSE3-NEXT:    pslldq {{.*#+}} xmm2 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm2[0]
+; SSE3-NEXT:    por %xmm1, %xmm2
 ; SSE3-NEXT:    por %xmm2, %xmm0
 ; SSE3-NEXT:    retq
 ;
@@ -344,7 +337,7 @@
 ; SSSE3-NEXT:    movd %eax, %xmm1
 ; SSSE3-NEXT:    movdqa %xmm1, %xmm2
 ; SSSE3-NEXT:    palignr {{.*#+}} xmm2 = xmm0[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15],xmm2[0]
-; SSSE3-NEXT:    pshufb {{.*#+}} xmm2 = xmm2[15,0,1,2,3,4,5,6,7,8,9,10,11,12,13],zero
+; SSSE3-NEXT:    palignr {{.*#+}} xmm2 = xmm2[15,0,1,2,3,4,5,6,7,8,9,10,11,12,13,14]
 ; SSSE3-NEXT:    pslldq {{.*#+}} xmm1 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm1[0]
 ; SSSE3-NEXT:    por %xmm2, %xmm1
 ; SSSE3-NEXT:    movdqa %xmm1, %xmm0
@@ -372,41 +365,31 @@
 ; SSE2-LABEL: insert_v32i8_x123456789ABCDEzGHIJKLMNOPQRSTxx:
 ; SSE2:       # %bb.0:
 ; SSE2-NEXT:    movdqa {{.*#+}} xmm2 = [0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
-; SSE2-NEXT:    pand %xmm2, %xmm0
 ; SSE2-NEXT:    movl $255, %eax
 ; SSE2-NEXT:    movd %eax, %xmm3
 ; SSE2-NEXT:    pandn %xmm3, %xmm2
-; SSE2-NEXT:    por %xmm2, %xmm0
-; SSE2-NEXT:    movdqa {{.*#+}} xmm2 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0]
-; SSE2-NEXT:    pand %xmm2, %xmm0
 ; SSE2-NEXT:    movdqa %xmm3, %xmm4
 ; SSE2-NEXT:    pslldq {{.*#+}} xmm4 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm4[0]
-; SSE2-NEXT:    por %xmm4, %xmm0
-; SSE2-NEXT:    pand {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm1
+; SSE2-NEXT:    por %xmm4, %xmm2
+; SSE2-NEXT:    por %xmm2, %xmm0
 ; SSE2-NEXT:    pslldq {{.*#+}} xmm3 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm3[0,1]
+; SSE2-NEXT:    por %xmm4, %xmm3
 ; SSE2-NEXT:    por %xmm3, %xmm1
-; SSE2-NEXT:    pand %xmm2, %xmm1
-; SSE2-NEXT:    por %xmm4, %xmm1
 ; SSE2-NEXT:    retq
 ;
 ; SSE3-LABEL: insert_v32i8_x123456789ABCDEzGHIJKLMNOPQRSTxx:
 ; SSE3:       # %bb.0:
 ; SSE3-NEXT:    movdqa {{.*#+}} xmm2 = [0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
-; SSE3-NEXT:    pand %xmm2, %xmm0
 ; SSE3-NEXT:    movl $255, %eax
 ; SSE3-NEXT:    movd %eax, %xmm3
 ; SSE3-NEXT:    pandn %xmm3, %xmm2
-; SSE3-NEXT:    por %xmm2, %xmm0
-; SSE3-NEXT:    movdqa {{.*#+}} xmm2 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0]
-; SSE3-NEXT:    pand %xmm2, %xmm0
 ; SSE3-NEXT:    movdqa %xmm3, %xmm4
 ; SSE3-NEXT:    pslldq {{.*#+}} xmm4 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm4[0]
-; SSE3-NEXT:    por %xmm4, %xmm0
-; SSE3-NEXT:    pand {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm1
+; SSE3-NEXT:    por %xmm4, %xmm2
+; SSE3-NEXT:    por %xmm2, %xmm0
 ; SSE3-NEXT:    pslldq {{.*#+}} xmm3 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm3[0,1]
+; SSE3-NEXT:    por %xmm4, %xmm3
 ; SSE3-NEXT:    por %xmm3, %xmm1
-; SSE3-NEXT:    pand %xmm2, %xmm1
-; SSE3-NEXT:    por %xmm4, %xmm1
 ; SSE3-NEXT:    retq
 ;
 ; SSSE3-LABEL: insert_v32i8_x123456789ABCDEzGHIJKLMNOPQRSTxx:
@@ -415,15 +398,13 @@
 ; SSSE3-NEXT:    movd %eax, %xmm3
 ; SSSE3-NEXT:    movdqa %xmm3, %xmm2
 ; SSSE3-NEXT:    palignr {{.*#+}} xmm2 = xmm0[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15],xmm2[0]
-; SSSE3-NEXT:    pshufb {{.*#+}} xmm2 = xmm2[15,0,1,2,3,4,5,6,7,8,9,10,11,12,13],zero
+; SSSE3-NEXT:    palignr {{.*#+}} xmm2 = xmm2[15,0,1,2,3,4,5,6,7,8,9,10,11,12,13,14]
 ; SSSE3-NEXT:    movdqa %xmm3, %xmm0
 ; SSSE3-NEXT:    pslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0]
 ; SSSE3-NEXT:    por %xmm0, %xmm2
-; SSSE3-NEXT:    pshufb {{.*#+}} xmm1 = xmm1[0,1,2,3,4,5,6,7,8,9,10,11,12,13],zero,xmm1[u]
-; SSSE3-NEXT:    pslldq {{.*#+}} xmm3 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm3[0,1]
+; SSSE3-NEXT:    pshufb {{.*#+}} xmm3 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm3[0],zero
+; SSSE3-NEXT:    por %xmm0, %xmm3
 ; SSSE3-NEXT:    por %xmm3, %xmm1
-; SSSE3-NEXT:    pshufb {{.*#+}} xmm1 = xmm1[0,1,2,3,4,5,6,7,8,9,10,11,12,13,14],zero
-; SSSE3-NEXT:    por %xmm0, %xmm1
 ; SSSE3-NEXT:    movdqa %xmm2, %xmm0
 ; SSSE3-NEXT:    retq
 ;