Index: include/llvm/Target/TargetLowering.h
===================================================================
--- include/llvm/Target/TargetLowering.h
+++ include/llvm/Target/TargetLowering.h
@@ -2726,6 +2726,20 @@
     return true;
   }
 
+  // Return true if it is profitable to combine a BUILD_VECTOR to a TRUNCATE for
+  // given operand and result types.
+  // Example of such a combine:
+  // v4i32 build_vector((extract_elt V, 0),
+  //                    (extract_elt V, 2),
+  //                    (extract_elt V, 4),
+  //                    (extract_elt V, 6))
+  //  -->
+  // v4i32 truncate (bitcast V to v4i64)
+  virtual bool isDesirableToCombineBuildVectorToTruncate(EVT OpVT,
+                                                         EVT ResVT) const {
+    return false;
+  }
+
   /// Return true if the target has native support for the specified value type
   /// and it is 'desirable' to use the type for the given node type. e.g. On x86
   /// i16 is legal, but undesirable since i16 instruction encodings are longer
Index: lib/CodeGen/SelectionDAG/DAGCombiner.cpp
===================================================================
--- lib/CodeGen/SelectionDAG/DAGCombiner.cpp
+++ lib/CodeGen/SelectionDAG/DAGCombiner.cpp
@@ -400,6 +400,7 @@
     SDValue reduceBuildVecExtToExtBuildVec(SDNode *N);
     SDValue reduceBuildVecConvertToConvertBuildVec(SDNode *N);
     SDValue reduceBuildVecToShuffle(SDNode *N);
+    SDValue reduceBuildVecToTrunc(SDNode *N);
     SDValue createBuildVecShuffle(const SDLoc &DL, SDNode *N,
                                   ArrayRef<int> VectorMask, SDValue VecIn1,
                                   SDValue VecIn2, unsigned LeftIdx);
@@ -14170,6 +14171,79 @@
   return Shuffles[0];
 }
 
+// Check to see if this is a BUILD_VECTOR of a bunch of EXTRACT_VECTOR_ELT
+// operations which be matched for a truncate.
+SDValue DAGCombiner::reduceBuildVecToTrunc(SDNode *N) {
+  // TODO: Add support for big-endian.
+  if (DAG.getDataLayout().isBigEndian())
+    return SDValue();
+  if (N->getNumOperands() < 2)
+    return SDValue();
+  SDLoc DL(N);
+  EVT VT = N->getValueType(0);
+  unsigned NumElems = N->getNumOperands();
+
+  if (!isTypeLegal(VT))
+    return SDValue();
+
+  // If the input is something other than an EXTRACT_VECTOR_ELT with a constant
+  // index, bail out.
+  // TODO: Allow undef elements in some cases?
+  if (any_of(make_range(N->op_begin(), N->op_end()), [VT](SDValue Op) {
+        return Op.getOpcode() != ISD::EXTRACT_VECTOR_ELT ||
+               !isa<ConstantSDNode>(Op.getOperand(1)) ||
+               Op.getValueType() != VT.getVectorElementType();
+      }))
+    return SDValue();
+
+  // Helper for obtaining an EXTRACT_VECTOR_ELT's constant index
+  auto GetExtractIdx = [](SDValue Extract) {
+    return cast<ConstantSDNode>(Extract.getOperand(1))->getSExtValue();
+  };
+
+  // The first BUILD_VECTOR operand must be an an extract from index zero
+  // (assuming no undef and little-endian).
+  if (GetExtractIdx(N->getOperand(0)) != 0)
+    return SDValue();
+
+  // Compute the stride from the first index.
+  int Stride = GetExtractIdx(N->getOperand(1));
+  SDValue ExtractedFromVec = N->getOperand(1).getOperand(0);
+
+  // Proceed only if the stride and the types can be matched to a truncate.
+  if ((Stride == 1 || !isPowerOf2_32(Stride)) ||
+      (ExtractedFromVec.getValueType().getVectorNumElements() !=
+       Stride * NumElems) ||
+      (VT.getScalarSizeInBits() * Stride > 64))
+    return SDValue();
+
+  // Check remaining operands are consistent with the computed stride.
+  for (unsigned i = 1; i != NumElems; ++i) {
+    SDValue Op = N->getOperand(i);
+
+    if ((Op.getOperand(0) != ExtractedFromVec) ||
+        (GetExtractIdx(Op) != Stride * i))
+      return SDValue();
+  }
+
+  EVT NewVT = VT.getVectorVT(
+      *DAG.getContext(), MVT::getIntegerVT(VT.getScalarSizeInBits() * Stride),
+      NumElems);
+  EVT TruncVT =
+      VT.isFloatingPoint() ? VT.changeVectorElementTypeToInteger() : VT;
+
+  // Bail out if combine not profitable for target.
+  if (!TLI.isDesirableToCombineBuildVectorToTruncate(NewVT, TruncVT))
+    return SDValue();
+
+  // All checks were ok, construct the truncate.
+  SDValue Res = DAG.getBitcast(NewVT, ExtractedFromVec);
+  Res = DAG.getNode(ISD::TRUNCATE, SDLoc(N), TruncVT, Res);
+  if (VT.isFloatingPoint())
+    Res = DAG.getBitcast(VT, Res);
+  return Res;
+}
+
 SDValue DAGCombiner::visitBUILD_VECTOR(SDNode *N) {
   EVT VT = N->getValueType(0);
 
@@ -14212,6 +14286,9 @@
   if (SDValue V = reduceBuildVecConvertToConvertBuildVec(N))
     return V;
 
+  if (SDValue V = reduceBuildVecToTrunc(N))
+    return V;
+
   if (SDValue V = reduceBuildVecToShuffle(N))
     return V;
 
Index: lib/Target/X86/X86ISelLowering.h
===================================================================
--- lib/Target/X86/X86ISelLowering.h
+++ lib/Target/X86/X86ISelLowering.h
@@ -767,6 +767,20 @@
 
     SDValue PerformDAGCombine(SDNode *N, DAGCombinerInfo &DCI) const override;
 
+    // Return true if it is profitable to combine a BUILD_VECTOR to a TRUNCATE
+    // for given operand and result types.
+    // Example of such a combine:
+    // v4i32 build_vector((extract_elt V, 0),
+    //                    (extract_elt V, 2),
+    //                    (extract_elt V, 4),
+    //                    (extract_elt V, 6))
+    //  -->
+    // v4i32 truncate (bitcast V to v4i64)
+    bool isDesirableToCombineBuildVectorToTruncate(EVT OpVT,
+                                                   EVT ResVT) const override {
+      return true;
+    }
+
     /// Return true if the target has native support for
     /// the specified value type and it is 'desirable' to use the type for the
     /// given node type. e.g. On x86 i16 is legal, but undesirable since i16
Index: test/CodeGen/X86/shuffle-vs-trunc-256.ll
===================================================================
--- test/CodeGen/X86/shuffle-vs-trunc-256.ll
+++ test/CodeGen/X86/shuffle-vs-trunc-256.ll
@@ -37,24 +37,16 @@
 ;
 ; AVX512F-LABEL: shuffle_v32i8_to_v16i8:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512F-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512F-NEXT:    vmovdqa {{.*#+}} xmm2 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
-; AVX512F-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
-; AVX512F-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
-; AVX512F-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; AVX512F-NEXT:    vpmovsxwd (%rdi), %zmm0
+; AVX512F-NEXT:    vpmovdb %zmm0, %xmm0
 ; AVX512F-NEXT:    vmovdqa %xmm0, (%rsi)
 ; AVX512F-NEXT:    vzeroupper
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512VL-LABEL: shuffle_v32i8_to_v16i8:
 ; AVX512VL:       # BB#0:
-; AVX512VL-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512VL-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512VL-NEXT:    vmovdqa {{.*#+}} xmm2 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
-; AVX512VL-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
-; AVX512VL-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
-; AVX512VL-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; AVX512VL-NEXT:    vpmovsxwd (%rdi), %zmm0
+; AVX512VL-NEXT:    vpmovdb %zmm0, %xmm0
 ; AVX512VL-NEXT:    vmovdqa %xmm0, (%rsi)
 ; AVX512VL-NEXT:    vzeroupper
 ; AVX512VL-NEXT:    retq
@@ -62,11 +54,7 @@
 ; AVX512BW-LABEL: shuffle_v32i8_to_v16i8:
 ; AVX512BW:       # BB#0:
 ; AVX512BW-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512BW-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512BW-NEXT:    vmovdqa {{.*#+}} xmm2 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
-; AVX512BW-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
-; AVX512BW-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
-; AVX512BW-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; AVX512BW-NEXT:    vpmovwb %zmm0, %ymm0
 ; AVX512BW-NEXT:    vmovdqa %xmm0, (%rsi)
 ; AVX512BW-NEXT:    vzeroupper
 ; AVX512BW-NEXT:    retq
@@ -74,12 +62,7 @@
 ; AVX512BWVL-LABEL: shuffle_v32i8_to_v16i8:
 ; AVX512BWVL:       # BB#0:
 ; AVX512BWVL-NEXT:    vmovdqu (%rdi), %ymm0
-; AVX512BWVL-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512BWVL-NEXT:    vmovdqu {{.*#+}} xmm2 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
-; AVX512BWVL-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
-; AVX512BWVL-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; AVX512BWVL-NEXT:    vmovdqu %xmm0, (%rsi)
+; AVX512BWVL-NEXT:    vpmovwb %ymm0, (%rsi)
 ; AVX512BWVL-NEXT:    vzeroupper
 ; AVX512BWVL-NEXT:    retq
   %vec = load <32 x i8>, <32 x i8>* %L
@@ -166,11 +149,8 @@
 ; AVX2-LABEL: shuffle_v16i16_to_v8i16:
 ; AVX2:       # BB#0:
 ; AVX2-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX2-NEXT:    vmovdqa {{.*#+}} xmm2 = [0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15]
-; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
-; AVX2-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
-; AVX2-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15,16,17,20,21,24,25,28,29,24,25,28,29,28,29,30,31]
+; AVX2-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,2,2,3]
 ; AVX2-NEXT:    vmovdqa %xmm0, (%rsi)
 ; AVX2-NEXT:    vzeroupper
 ; AVX2-NEXT:    retq
@@ -178,11 +158,7 @@
 ; AVX512F-LABEL: shuffle_v16i16_to_v8i16:
 ; AVX512F:       # BB#0:
 ; AVX512F-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512F-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512F-NEXT:    vmovdqa {{.*#+}} xmm2 = [0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15]
-; AVX512F-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
-; AVX512F-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
-; AVX512F-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; AVX512F-NEXT:    vpmovdw %zmm0, %ymm0
 ; AVX512F-NEXT:    vmovdqa %xmm0, (%rsi)
 ; AVX512F-NEXT:    vzeroupper
 ; AVX512F-NEXT:    retq
@@ -190,42 +166,22 @@
 ; AVX512VL-LABEL: shuffle_v16i16_to_v8i16:
 ; AVX512VL:       # BB#0:
 ; AVX512VL-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512VL-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512VL-NEXT:    vpshuflw {{.*#+}} xmm1 = xmm1[0,2,2,3,4,5,6,7]
-; AVX512VL-NEXT:    vpshufhw {{.*#+}} xmm1 = xmm1[0,1,2,3,4,6,6,7]
-; AVX512VL-NEXT:    vpshufd {{.*#+}} xmm1 = xmm1[0,2,2,3]
-; AVX512VL-NEXT:    vpshuflw {{.*#+}} xmm0 = xmm0[0,2,2,3,4,5,6,7]
-; AVX512VL-NEXT:    vpshufhw {{.*#+}} xmm0 = xmm0[0,1,2,3,4,6,6,7]
-; AVX512VL-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
-; AVX512VL-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; AVX512VL-NEXT:    vmovdqa %xmm0, (%rsi)
+; AVX512VL-NEXT:    vpmovdw %ymm0, (%rsi)
 ; AVX512VL-NEXT:    vzeroupper
 ; AVX512VL-NEXT:    retq
 ;
 ; AVX512BW-LABEL: shuffle_v16i16_to_v8i16:
 ; AVX512BW:       # BB#0:
 ; AVX512BW-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512BW-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512BW-NEXT:    vmovdqa {{.*#+}} xmm2 = [0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15]
-; AVX512BW-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
-; AVX512BW-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
-; AVX512BW-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; AVX512BW-NEXT:    vpmovdw %zmm0, %ymm0
 ; AVX512BW-NEXT:    vmovdqa %xmm0, (%rsi)
 ; AVX512BW-NEXT:    vzeroupper
 ; AVX512BW-NEXT:    retq
 ;
 ; AVX512BWVL-LABEL: shuffle_v16i16_to_v8i16:
 ; AVX512BWVL:       # BB#0:
-; AVX512BWVL-NEXT:    vmovdqu (%rdi), %ymm0
-; AVX512BWVL-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512BWVL-NEXT:    vpshuflw {{.*#+}} xmm1 = xmm1[0,2,2,3,4,5,6,7]
-; AVX512BWVL-NEXT:    vpshufhw {{.*#+}} xmm1 = xmm1[0,1,2,3,4,6,6,7]
-; AVX512BWVL-NEXT:    vpshufd {{.*#+}} xmm1 = xmm1[0,2,2,3]
-; AVX512BWVL-NEXT:    vpshuflw {{.*#+}} xmm0 = xmm0[0,2,2,3,4,5,6,7]
-; AVX512BWVL-NEXT:    vpshufhw {{.*#+}} xmm0 = xmm0[0,1,2,3,4,6,6,7]
-; AVX512BWVL-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
-; AVX512BWVL-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; AVX512BWVL-NEXT:    vmovdqu %xmm0, (%rsi)
+; AVX512BWVL-NEXT:    vmovdqa (%rdi), %ymm0
+; AVX512BWVL-NEXT:    vpmovdw %ymm0, (%rsi)
 ; AVX512BWVL-NEXT:    vzeroupper
 ; AVX512BWVL-NEXT:    retq
   %vec = load <16 x i16>, <16 x i16>* %L
@@ -293,48 +249,50 @@
 }
 
 define void @shuffle_v8i32_to_v4i32(<8 x i32>* %L, <4 x i32>* %S) nounwind {
-; AVX-LABEL: shuffle_v8i32_to_v4i32:
-; AVX:       # BB#0:
-; AVX-NEXT:    vmovaps (%rdi), %ymm0
-; AVX-NEXT:    vextractf128 $1, %ymm0, %xmm1
-; AVX-NEXT:    vshufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[0,2]
-; AVX-NEXT:    vmovaps %xmm0, (%rsi)
-; AVX-NEXT:    vzeroupper
-; AVX-NEXT:    retq
+; AVX1-LABEL: shuffle_v8i32_to_v4i32:
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vmovaps (%rdi), %ymm0
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vshufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[0,2]
+; AVX1-NEXT:    vmovaps %xmm0, (%rsi)
+; AVX1-NEXT:    vzeroupper
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: shuffle_v8i32_to_v4i32:
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshufd {{.*#+}} ymm0 = mem[0,2,2,3,4,6,6,7]
+; AVX2-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,2,2,3]
+; AVX2-NEXT:    vmovdqa %xmm0, (%rsi)
+; AVX2-NEXT:    vzeroupper
+; AVX2-NEXT:    retq
 ;
 ; AVX512F-LABEL: shuffle_v8i32_to_v4i32:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovaps (%rdi), %ymm0
-; AVX512F-NEXT:    vextractf128 $1, %ymm0, %xmm1
-; AVX512F-NEXT:    vshufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[0,2]
-; AVX512F-NEXT:    vmovaps %xmm0, (%rsi)
+; AVX512F-NEXT:    vmovdqa (%rdi), %ymm0
+; AVX512F-NEXT:    vpmovqd %zmm0, %ymm0
+; AVX512F-NEXT:    vmovdqa %xmm0, (%rsi)
 ; AVX512F-NEXT:    vzeroupper
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512VL-LABEL: shuffle_v8i32_to_v4i32:
 ; AVX512VL:       # BB#0:
 ; AVX512VL-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512VL-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512VL-NEXT:    vshufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[0,2]
-; AVX512VL-NEXT:    vmovaps %xmm0, (%rsi)
+; AVX512VL-NEXT:    vpmovqd %ymm0, (%rsi)
 ; AVX512VL-NEXT:    vzeroupper
 ; AVX512VL-NEXT:    retq
 ;
 ; AVX512BW-LABEL: shuffle_v8i32_to_v4i32:
 ; AVX512BW:       # BB#0:
-; AVX512BW-NEXT:    vmovaps (%rdi), %ymm0
-; AVX512BW-NEXT:    vextractf128 $1, %ymm0, %xmm1
-; AVX512BW-NEXT:    vshufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[0,2]
-; AVX512BW-NEXT:    vmovaps %xmm0, (%rsi)
+; AVX512BW-NEXT:    vmovdqa (%rdi), %ymm0
+; AVX512BW-NEXT:    vpmovqd %zmm0, %ymm0
+; AVX512BW-NEXT:    vmovdqa %xmm0, (%rsi)
 ; AVX512BW-NEXT:    vzeroupper
 ; AVX512BW-NEXT:    retq
 ;
 ; AVX512BWVL-LABEL: shuffle_v8i32_to_v4i32:
 ; AVX512BWVL:       # BB#0:
 ; AVX512BWVL-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512BWVL-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512BWVL-NEXT:    vshufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[0,2]
-; AVX512BWVL-NEXT:    vmovaps %xmm0, (%rsi)
+; AVX512BWVL-NEXT:    vpmovqd %ymm0, (%rsi)
 ; AVX512BWVL-NEXT:    vzeroupper
 ; AVX512BWVL-NEXT:    retq
   %vec = load <8 x i32>, <8 x i32>* %L
@@ -413,11 +371,9 @@
 ; AVX2-LABEL: shuffle_v32i8_to_v8i8:
 ; AVX2:       # BB#0:
 ; AVX2-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX2-NEXT:    vmovdqa {{.*#+}} xmm2 = <0,4,8,12,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
-; AVX2-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
-; AVX2-NEXT:    vpunpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15,16,17,20,21,24,25,28,29,24,25,28,29,28,29,30,31]
+; AVX2-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,2,2,3]
+; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u]
 ; AVX2-NEXT:    vmovq %xmm0, (%rsi)
 ; AVX2-NEXT:    vzeroupper
 ; AVX2-NEXT:    retq
@@ -425,11 +381,8 @@
 ; AVX512F-LABEL: shuffle_v32i8_to_v8i8:
 ; AVX512F:       # BB#0:
 ; AVX512F-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512F-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512F-NEXT:    vmovdqa {{.*#+}} xmm2 = <0,4,8,12,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512F-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
-; AVX512F-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
-; AVX512F-NEXT:    vpunpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; AVX512F-NEXT:    vpmovdw %zmm0, %ymm0
+; AVX512F-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u]
 ; AVX512F-NEXT:    vmovq %xmm0, (%rsi)
 ; AVX512F-NEXT:    vzeroupper
 ; AVX512F-NEXT:    retq
@@ -437,39 +390,23 @@
 ; AVX512VL-LABEL: shuffle_v32i8_to_v8i8:
 ; AVX512VL:       # BB#0:
 ; AVX512VL-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512VL-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512VL-NEXT:    vmovdqa {{.*#+}} xmm2 = <0,4,8,12,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512VL-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
-; AVX512VL-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
-; AVX512VL-NEXT:    vpunpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
-; AVX512VL-NEXT:    vmovq %xmm0, (%rsi)
+; AVX512VL-NEXT:    vpmovdb %ymm0, (%rsi)
 ; AVX512VL-NEXT:    vzeroupper
 ; AVX512VL-NEXT:    retq
 ;
 ; AVX512BW-LABEL: shuffle_v32i8_to_v8i8:
 ; AVX512BW:       # BB#0:
 ; AVX512BW-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512BW-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512BW-NEXT:    vmovdqa {{.*#+}} xmm2 = <0,4,8,12,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512BW-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
-; AVX512BW-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
-; AVX512BW-NEXT:    vpunpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; AVX512BW-NEXT:    vpmovdw %zmm0, %ymm0
+; AVX512BW-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u]
 ; AVX512BW-NEXT:    vmovq %xmm0, (%rsi)
 ; AVX512BW-NEXT:    vzeroupper
 ; AVX512BW-NEXT:    retq
 ;
 ; AVX512BWVL-LABEL: shuffle_v32i8_to_v8i8:
 ; AVX512BWVL:       # BB#0:
-; AVX512BWVL-NEXT:    vmovdqu (%rdi), %ymm0
-; AVX512BWVL-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512BWVL-NEXT:    vpshuflw {{.*#+}} xmm1 = xmm1[0,2,2,3,4,5,6,7]
-; AVX512BWVL-NEXT:    vpshufhw {{.*#+}} xmm1 = xmm1[0,1,2,3,4,6,6,7]
-; AVX512BWVL-NEXT:    vpshufd {{.*#+}} xmm1 = xmm1[0,2,2,3]
-; AVX512BWVL-NEXT:    vpshuflw {{.*#+}} xmm0 = xmm0[0,2,2,3,4,5,6,7]
-; AVX512BWVL-NEXT:    vpshufhw {{.*#+}} xmm0 = xmm0[0,1,2,3,4,6,6,7]
-; AVX512BWVL-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
-; AVX512BWVL-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; AVX512BWVL-NEXT:    vpmovwb %xmm0, (%rsi)
+; AVX512BWVL-NEXT:    vmovdqa (%rdi), %ymm0
+; AVX512BWVL-NEXT:    vpmovdb %ymm0, (%rsi)
 ; AVX512BWVL-NEXT:    vzeroupper
 ; AVX512BWVL-NEXT:    retq
   %vec = load <32 x i8>, <32 x i8>* %L
@@ -542,26 +479,19 @@
 define void @shuffle_v16i16_to_v4i16(<16 x i16>* %L, <4 x i16>* %S) nounwind {
 ; AVX1-LABEL: shuffle_v16i16_to_v4i16:
 ; AVX1:       # BB#0:
-; AVX1-NEXT:    vmovdqa (%rdi), %ymm0
+; AVX1-NEXT:    vmovaps (%rdi), %ymm0
 ; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
-; AVX1-NEXT:    vpshufd {{.*#+}} xmm1 = xmm1[0,2,2,3]
-; AVX1-NEXT:    vpshuflw {{.*#+}} xmm1 = xmm1[0,2,2,3,4,5,6,7]
-; AVX1-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
-; AVX1-NEXT:    vpshuflw {{.*#+}} xmm0 = xmm0[0,2,2,3,4,5,6,7]
-; AVX1-NEXT:    vpunpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; AVX1-NEXT:    vshufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[0,2]
+; AVX1-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15]
 ; AVX1-NEXT:    vmovq %xmm0, (%rsi)
 ; AVX1-NEXT:    vzeroupper
 ; AVX1-NEXT:    retq
 ;
 ; AVX2-LABEL: shuffle_v16i16_to_v4i16:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX2-NEXT:    vpshufd {{.*#+}} xmm1 = xmm1[0,2,2,3]
-; AVX2-NEXT:    vpshuflw {{.*#+}} xmm1 = xmm1[0,2,2,3,4,5,6,7]
-; AVX2-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
-; AVX2-NEXT:    vpshuflw {{.*#+}} xmm0 = xmm0[0,2,2,3,4,5,6,7]
-; AVX2-NEXT:    vpunpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; AVX2-NEXT:    vpshufd {{.*#+}} ymm0 = mem[0,2,2,3,4,6,6,7]
+; AVX2-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,2,2,3]
+; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15]
 ; AVX2-NEXT:    vmovq %xmm0, (%rsi)
 ; AVX2-NEXT:    vzeroupper
 ; AVX2-NEXT:    retq
@@ -569,12 +499,8 @@
 ; AVX512F-LABEL: shuffle_v16i16_to_v4i16:
 ; AVX512F:       # BB#0:
 ; AVX512F-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512F-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512F-NEXT:    vpshufd {{.*#+}} xmm1 = xmm1[0,2,2,3]
-; AVX512F-NEXT:    vpshuflw {{.*#+}} xmm1 = xmm1[0,2,2,3,4,5,6,7]
-; AVX512F-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
-; AVX512F-NEXT:    vpshuflw {{.*#+}} xmm0 = xmm0[0,2,2,3,4,5,6,7]
-; AVX512F-NEXT:    vpunpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; AVX512F-NEXT:    vpmovqd %zmm0, %ymm0
+; AVX512F-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15]
 ; AVX512F-NEXT:    vmovq %xmm0, (%rsi)
 ; AVX512F-NEXT:    vzeroupper
 ; AVX512F-NEXT:    retq
@@ -582,31 +508,23 @@
 ; AVX512VL-LABEL: shuffle_v16i16_to_v4i16:
 ; AVX512VL:       # BB#0:
 ; AVX512VL-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512VL-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512VL-NEXT:    vshufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[0,2]
-; AVX512VL-NEXT:    vpmovdw %xmm0, (%rsi)
+; AVX512VL-NEXT:    vpmovqw %ymm0, (%rsi)
 ; AVX512VL-NEXT:    vzeroupper
 ; AVX512VL-NEXT:    retq
 ;
 ; AVX512BW-LABEL: shuffle_v16i16_to_v4i16:
 ; AVX512BW:       # BB#0:
 ; AVX512BW-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512BW-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512BW-NEXT:    vpshufd {{.*#+}} xmm1 = xmm1[0,2,2,3]
-; AVX512BW-NEXT:    vpshuflw {{.*#+}} xmm1 = xmm1[0,2,2,3,4,5,6,7]
-; AVX512BW-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
-; AVX512BW-NEXT:    vpshuflw {{.*#+}} xmm0 = xmm0[0,2,2,3,4,5,6,7]
-; AVX512BW-NEXT:    vpunpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; AVX512BW-NEXT:    vpmovqd %zmm0, %ymm0
+; AVX512BW-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15]
 ; AVX512BW-NEXT:    vmovq %xmm0, (%rsi)
 ; AVX512BW-NEXT:    vzeroupper
 ; AVX512BW-NEXT:    retq
 ;
 ; AVX512BWVL-LABEL: shuffle_v16i16_to_v4i16:
 ; AVX512BWVL:       # BB#0:
-; AVX512BWVL-NEXT:    vmovdqu (%rdi), %ymm0
-; AVX512BWVL-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512BWVL-NEXT:    vshufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[0,2]
-; AVX512BWVL-NEXT:    vpmovdw %xmm0, (%rsi)
+; AVX512BWVL-NEXT:    vmovdqa (%rdi), %ymm0
+; AVX512BWVL-NEXT:    vpmovqw %ymm0, (%rsi)
 ; AVX512BWVL-NEXT:    vzeroupper
 ; AVX512BWVL-NEXT:    retq
   %vec = load <16 x i16>, <16 x i16>* %L
@@ -676,24 +594,19 @@
 define void @shuffle_v32i8_to_v4i8(<32 x i8>* %L, <4 x i8>* %S) nounwind {
 ; AVX1-LABEL: shuffle_v32i8_to_v4i8:
 ; AVX1:       # BB#0:
-; AVX1-NEXT:    vmovdqa (%rdi), %ymm0
+; AVX1-NEXT:    vmovaps (%rdi), %ymm0
 ; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
-; AVX1-NEXT:    vmovdqa {{.*#+}} xmm2 = <0,8,u,u,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX1-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
-; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
-; AVX1-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX1-NEXT:    vshufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[0,2]
+; AVX1-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,4,8,12,u,u,u,u,u,u,u,u,u,u,u,u]
 ; AVX1-NEXT:    vmovd %xmm0, (%rsi)
 ; AVX1-NEXT:    vzeroupper
 ; AVX1-NEXT:    retq
 ;
 ; AVX2-LABEL: shuffle_v32i8_to_v4i8:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX2-NEXT:    vmovdqa {{.*#+}} xmm2 = <0,8,u,u,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
-; AVX2-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
-; AVX2-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX2-NEXT:    vpshufd {{.*#+}} ymm0 = mem[0,2,2,3,4,6,6,7]
+; AVX2-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,2,2,3]
+; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,4,8,12,u,u,u,u,u,u,u,u,u,u,u,u]
 ; AVX2-NEXT:    vmovd %xmm0, (%rsi)
 ; AVX2-NEXT:    vzeroupper
 ; AVX2-NEXT:    retq
@@ -701,11 +614,8 @@
 ; AVX512F-LABEL: shuffle_v32i8_to_v4i8:
 ; AVX512F:       # BB#0:
 ; AVX512F-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512F-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512F-NEXT:    vmovdqa {{.*#+}} xmm2 = <0,8,u,u,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512F-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
-; AVX512F-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
-; AVX512F-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX512F-NEXT:    vpmovqd %zmm0, %ymm0
+; AVX512F-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,4,8,12,u,u,u,u,u,u,u,u,u,u,u,u]
 ; AVX512F-NEXT:    vmovd %xmm0, (%rsi)
 ; AVX512F-NEXT:    vzeroupper
 ; AVX512F-NEXT:    retq
@@ -713,30 +623,23 @@
 ; AVX512VL-LABEL: shuffle_v32i8_to_v4i8:
 ; AVX512VL:       # BB#0:
 ; AVX512VL-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512VL-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512VL-NEXT:    vshufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[0,2]
-; AVX512VL-NEXT:    vpmovdb %xmm0, (%rsi)
+; AVX512VL-NEXT:    vpmovqb %ymm0, (%rsi)
 ; AVX512VL-NEXT:    vzeroupper
 ; AVX512VL-NEXT:    retq
 ;
 ; AVX512BW-LABEL: shuffle_v32i8_to_v4i8:
 ; AVX512BW:       # BB#0:
 ; AVX512BW-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512BW-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512BW-NEXT:    vmovdqa {{.*#+}} xmm2 = <0,8,u,u,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512BW-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
-; AVX512BW-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
-; AVX512BW-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX512BW-NEXT:    vpmovqd %zmm0, %ymm0
+; AVX512BW-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,4,8,12,u,u,u,u,u,u,u,u,u,u,u,u]
 ; AVX512BW-NEXT:    vmovd %xmm0, (%rsi)
 ; AVX512BW-NEXT:    vzeroupper
 ; AVX512BW-NEXT:    retq
 ;
 ; AVX512BWVL-LABEL: shuffle_v32i8_to_v4i8:
 ; AVX512BWVL:       # BB#0:
-; AVX512BWVL-NEXT:    vmovdqu (%rdi), %ymm0
-; AVX512BWVL-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512BWVL-NEXT:    vshufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[0,2]
-; AVX512BWVL-NEXT:    vpmovdb %xmm0, (%rsi)
+; AVX512BWVL-NEXT:    vmovdqa (%rdi), %ymm0
+; AVX512BWVL-NEXT:    vpmovqb %ymm0, (%rsi)
 ; AVX512BWVL-NEXT:    vzeroupper
 ; AVX512BWVL-NEXT:    retq
   %vec = load <32 x i8>, <32 x i8>* %L
Index: test/CodeGen/X86/shuffle-vs-trunc-512.ll
===================================================================
--- test/CodeGen/X86/shuffle-vs-trunc-512.ll
+++ test/CodeGen/X86/shuffle-vs-trunc-512.ll
@@ -11,49 +11,37 @@
 define void @shuffle_v64i8_to_v32i8(<64 x i8>* %L, <32 x i8>* %S) nounwind {
 ; AVX512F-LABEL: shuffle_v64i8_to_v32i8:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512F-NEXT:    vmovdqa 32(%rdi), %ymm1
-; AVX512F-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,u,u,u,u,u,u,u,0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u,16,18,20,22,24,26,28,30]
-; AVX512F-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u,16,18,20,22,24,26,28,30,u,u,u,u,u,u,u,u]
-; AVX512F-NEXT:    vpblendd {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3],ymm0[4,5],ymm1[6,7]
-; AVX512F-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,2,1,3]
+; AVX512F-NEXT:    vpmovsxwd (%rdi), %zmm0
+; AVX512F-NEXT:    vpmovdb %zmm0, %xmm0
+; AVX512F-NEXT:    vpmovsxwd 32(%rdi), %zmm1
+; AVX512F-NEXT:    vpmovdb %zmm1, %xmm1
+; AVX512F-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
 ; AVX512F-NEXT:    vmovdqa %ymm0, (%rsi)
 ; AVX512F-NEXT:    vzeroupper
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512VL-LABEL: shuffle_v64i8_to_v32i8:
 ; AVX512VL:       # BB#0:
-; AVX512VL-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512VL-NEXT:    vmovdqa 32(%rdi), %ymm1
-; AVX512VL-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,u,u,u,u,u,u,u,0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u,16,18,20,22,24,26,28,30]
-; AVX512VL-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u,16,18,20,22,24,26,28,30,u,u,u,u,u,u,u,u]
-; AVX512VL-NEXT:    vpblendd {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3],ymm0[4,5],ymm1[6,7]
-; AVX512VL-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,2,1,3]
+; AVX512VL-NEXT:    vpmovsxwd (%rdi), %zmm0
+; AVX512VL-NEXT:    vpmovdb %zmm0, %xmm0
+; AVX512VL-NEXT:    vpmovsxwd 32(%rdi), %zmm1
+; AVX512VL-NEXT:    vpmovdb %zmm1, %xmm1
+; AVX512VL-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
 ; AVX512VL-NEXT:    vmovdqa %ymm0, (%rsi)
 ; AVX512VL-NEXT:    vzeroupper
 ; AVX512VL-NEXT:    retq
 ;
 ; AVX512BW-LABEL: shuffle_v64i8_to_v32i8:
 ; AVX512BW:       # BB#0:
-; AVX512BW-NEXT:    vmovdqu8 (%rdi), %zmm0
-; AVX512BW-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
-; AVX512BW-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,u,u,u,u,u,u,u,0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u,16,18,20,22,24,26,28,30]
-; AVX512BW-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u,16,18,20,22,24,26,28,30,u,u,u,u,u,u,u,u]
-; AVX512BW-NEXT:    vpblendd {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3],ymm0[4,5],ymm1[6,7]
-; AVX512BW-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,2,1,3]
-; AVX512BW-NEXT:    vmovdqa %ymm0, (%rsi)
+; AVX512BW-NEXT:    vmovdqu16 (%rdi), %zmm0
+; AVX512BW-NEXT:    vpmovwb %zmm0, (%rsi)
 ; AVX512BW-NEXT:    vzeroupper
 ; AVX512BW-NEXT:    retq
 ;
 ; AVX512BWVL-LABEL: shuffle_v64i8_to_v32i8:
 ; AVX512BWVL:       # BB#0:
-; AVX512BWVL-NEXT:    vmovdqu8 (%rdi), %zmm0
-; AVX512BWVL-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
-; AVX512BWVL-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,u,u,u,u,u,u,u,0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u,16,18,20,22,24,26,28,30]
-; AVX512BWVL-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u,16,18,20,22,24,26,28,30,u,u,u,u,u,u,u,u]
-; AVX512BWVL-NEXT:    vpblendd {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3],ymm0[4,5],ymm1[6,7]
-; AVX512BWVL-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,2,1,3]
-; AVX512BWVL-NEXT:    vmovdqu %ymm0, (%rsi)
+; AVX512BWVL-NEXT:    vmovdqu16 (%rdi), %zmm0
+; AVX512BWVL-NEXT:    vpmovwb %zmm0, (%rsi)
 ; AVX512BWVL-NEXT:    vzeroupper
 ; AVX512BWVL-NEXT:    retq
   %vec = load <64 x i8>, <64 x i8>* %L
@@ -106,54 +94,12 @@
 }
 
 define void @shuffle_v32i16_to_v16i16(<32 x i16>* %L, <16 x i16>* %S) nounwind {
-; AVX512F-LABEL: shuffle_v32i16_to_v16i16:
-; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vpshuflw {{.*#+}} ymm0 = mem[0,2,2,3,4,5,6,7,8,10,10,11,12,13,14,15]
-; AVX512F-NEXT:    vpshufhw {{.*#+}} ymm0 = ymm0[0,1,2,3,4,6,6,7,8,9,10,11,12,14,14,15]
-; AVX512F-NEXT:    vpshuflw {{.*#+}} ymm1 = mem[0,2,2,3,4,5,6,7,8,10,10,11,12,13,14,15]
-; AVX512F-NEXT:    vpshufhw {{.*#+}} ymm1 = ymm1[0,1,2,3,4,6,6,7,8,9,10,11,12,14,14,15]
-; AVX512F-NEXT:    vshufps {{.*#+}} ymm0 = ymm1[0,2],ymm0[0,2],ymm1[4,6],ymm0[4,6]
-; AVX512F-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,2,1,3]
-; AVX512F-NEXT:    vmovdqa %ymm0, (%rsi)
-; AVX512F-NEXT:    vzeroupper
-; AVX512F-NEXT:    retq
-;
-; AVX512VL-LABEL: shuffle_v32i16_to_v16i16:
-; AVX512VL:       # BB#0:
-; AVX512VL-NEXT:    vpshuflw {{.*#+}} ymm0 = mem[0,2,2,3,4,5,6,7,8,10,10,11,12,13,14,15]
-; AVX512VL-NEXT:    vpshufhw {{.*#+}} ymm0 = ymm0[0,1,2,3,4,6,6,7,8,9,10,11,12,14,14,15]
-; AVX512VL-NEXT:    vpshuflw {{.*#+}} ymm1 = mem[0,2,2,3,4,5,6,7,8,10,10,11,12,13,14,15]
-; AVX512VL-NEXT:    vpshufhw {{.*#+}} ymm1 = ymm1[0,1,2,3,4,6,6,7,8,9,10,11,12,14,14,15]
-; AVX512VL-NEXT:    vshufps {{.*#+}} ymm0 = ymm1[0,2],ymm0[0,2],ymm1[4,6],ymm0[4,6]
-; AVX512VL-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,2,1,3]
-; AVX512VL-NEXT:    vmovdqa %ymm0, (%rsi)
-; AVX512VL-NEXT:    vzeroupper
-; AVX512VL-NEXT:    retq
-;
-; AVX512BW-LABEL: shuffle_v32i16_to_v16i16:
-; AVX512BW:       # BB#0:
-; AVX512BW-NEXT:    vmovdqu16 (%rdi), %zmm0
-; AVX512BW-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
-; AVX512BW-NEXT:    vpshuflw {{.*#+}} ymm1 = ymm1[0,2,2,3,4,5,6,7,8,10,10,11,12,13,14,15]
-; AVX512BW-NEXT:    vpshufhw {{.*#+}} ymm1 = ymm1[0,1,2,3,4,6,6,7,8,9,10,11,12,14,14,15]
-; AVX512BW-NEXT:    vpshuflw {{.*#+}} ymm0 = ymm0[0,2,2,3,4,5,6,7,8,10,10,11,12,13,14,15]
-; AVX512BW-NEXT:    vpshufhw {{.*#+}} ymm0 = ymm0[0,1,2,3,4,6,6,7,8,9,10,11,12,14,14,15]
-; AVX512BW-NEXT:    vshufps {{.*#+}} ymm0 = ymm0[0,2],ymm1[0,2],ymm0[4,6],ymm1[4,6]
-; AVX512BW-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,2,1,3]
-; AVX512BW-NEXT:    vmovdqa %ymm0, (%rsi)
-; AVX512BW-NEXT:    vzeroupper
-; AVX512BW-NEXT:    retq
-;
-; AVX512BWVL-LABEL: shuffle_v32i16_to_v16i16:
-; AVX512BWVL:       # BB#0:
-; AVX512BWVL-NEXT:    vmovdqu16 (%rdi), %zmm0
-; AVX512BWVL-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
-; AVX512BWVL-NEXT:    vmovdqu {{.*#+}} ymm2 = [0,2,4,6,16,18,20,22,8,10,12,14,24,26,28,30]
-; AVX512BWVL-NEXT:    vpermi2w %ymm1, %ymm0, %ymm2
-; AVX512BWVL-NEXT:    vpermq {{.*#+}} ymm0 = ymm2[0,2,1,3]
-; AVX512BWVL-NEXT:    vmovdqu %ymm0, (%rsi)
-; AVX512BWVL-NEXT:    vzeroupper
-; AVX512BWVL-NEXT:    retq
+; AVX512-LABEL: shuffle_v32i16_to_v16i16:
+; AVX512:       # BB#0:
+; AVX512-NEXT:    vmovdqa32 (%rdi), %zmm0
+; AVX512-NEXT:    vpmovdw %zmm0, (%rsi)
+; AVX512-NEXT:    vzeroupper
+; AVX512-NEXT:    retq
   %vec = load <32 x i16>, <32 x i16>* %L
   %strided.vec = shufflevector <32 x i16> %vec, <32 x i16> undef, <16 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14, i32 16, i32 18, i32 20, i32 22, i32 24, i32 26, i32 28, i32 30>
   store <16 x i16> %strided.vec, <16 x i16>* %S
@@ -177,11 +123,8 @@
 define void @shuffle_v16i32_to_v8i32(<16 x i32>* %L, <8 x i32>* %S) nounwind {
 ; AVX512-LABEL: shuffle_v16i32_to_v8i32:
 ; AVX512:       # BB#0:
-; AVX512-NEXT:    vmovdqa32 (%rdi), %zmm0
-; AVX512-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
-; AVX512-NEXT:    vshufps {{.*#+}} ymm0 = ymm0[0,2],ymm1[0,2],ymm0[4,6],ymm1[4,6]
-; AVX512-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,2,1,3]
-; AVX512-NEXT:    vmovdqa %ymm0, (%rsi)
+; AVX512-NEXT:    vmovdqa64 (%rdi), %zmm0
+; AVX512-NEXT:    vpmovqd %zmm0, (%rsi)
 ; AVX512-NEXT:    vzeroupper
 ; AVX512-NEXT:    retq
   %vec = load <16 x i32>, <16 x i32>* %L
@@ -205,127 +148,12 @@
 }
 
 define void @shuffle_v64i8_to_v16i8(<64 x i8>* %L, <16 x i8>* %S) nounwind {
-; AVX512F-LABEL: shuffle_v64i8_to_v16i8:
-; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512F-NEXT:    vmovdqa 32(%rdi), %ymm1
-; AVX512F-NEXT:    vextracti128 $1, %ymm1, %xmm2
-; AVX512F-NEXT:    vmovdqa {{.*#+}} xmm3 = <u,u,u,u,0,4,8,12,u,u,u,u,u,u,u,u>
-; AVX512F-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512F-NEXT:    vpshufb %xmm3, %xmm1, %xmm1
-; AVX512F-NEXT:    vpunpckldq {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1]
-; AVX512F-NEXT:    vextracti128 $1, %ymm0, %xmm2
-; AVX512F-NEXT:    vmovdqa {{.*#+}} xmm3 = <0,4,8,12,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512F-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512F-NEXT:    vpshufb %xmm3, %xmm0, %xmm0
-; AVX512F-NEXT:    vpunpckldq {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1]
-; AVX512F-NEXT:    vpblendd {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,3]
-; AVX512F-NEXT:    vmovdqa %xmm0, (%rsi)
-; AVX512F-NEXT:    vzeroupper
-; AVX512F-NEXT:    retq
-;
-; AVX512VL-LABEL: shuffle_v64i8_to_v16i8:
-; AVX512VL:       # BB#0:
-; AVX512VL-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512VL-NEXT:    vmovdqa 32(%rdi), %ymm1
-; AVX512VL-NEXT:    vextracti128 $1, %ymm1, %xmm2
-; AVX512VL-NEXT:    vmovdqa {{.*#+}} xmm3 = <u,u,u,u,0,4,8,12,u,u,u,u,u,u,u,u>
-; AVX512VL-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512VL-NEXT:    vpshufb %xmm3, %xmm1, %xmm1
-; AVX512VL-NEXT:    vpunpckldq {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1]
-; AVX512VL-NEXT:    vextracti128 $1, %ymm0, %xmm2
-; AVX512VL-NEXT:    vmovdqa {{.*#+}} xmm3 = <0,4,8,12,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512VL-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512VL-NEXT:    vpshufb %xmm3, %xmm0, %xmm0
-; AVX512VL-NEXT:    vpunpckldq {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1]
-; AVX512VL-NEXT:    vpblendd {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,3]
-; AVX512VL-NEXT:    vmovdqa %xmm0, (%rsi)
-; AVX512VL-NEXT:    vzeroupper
-; AVX512VL-NEXT:    retq
-;
-; AVX512BW-LABEL: shuffle_v64i8_to_v16i8:
-; AVX512BW:       # BB#0:
-; AVX512BW-NEXT:    vmovdqu8 (%rdi), %zmm0
-; AVX512BW-NEXT:    vpextrb $4, %xmm0, %eax
-; AVX512BW-NEXT:    vpextrb $0, %xmm0, %ecx
-; AVX512BW-NEXT:    vmovd %ecx, %xmm1
-; AVX512BW-NEXT:    vpinsrb $1, %eax, %xmm1, %xmm1
-; AVX512BW-NEXT:    vpextrb $8, %xmm0, %eax
-; AVX512BW-NEXT:    vpinsrb $2, %eax, %xmm1, %xmm1
-; AVX512BW-NEXT:    vpextrb $12, %xmm0, %eax
-; AVX512BW-NEXT:    vpinsrb $3, %eax, %xmm1, %xmm1
-; AVX512BW-NEXT:    vextracti32x4 $1, %zmm0, %xmm2
-; AVX512BW-NEXT:    vpextrb $0, %xmm2, %eax
-; AVX512BW-NEXT:    vpinsrb $4, %eax, %xmm1, %xmm1
-; AVX512BW-NEXT:    vpextrb $4, %xmm2, %eax
-; AVX512BW-NEXT:    vpinsrb $5, %eax, %xmm1, %xmm1
-; AVX512BW-NEXT:    vpextrb $8, %xmm2, %eax
-; AVX512BW-NEXT:    vpinsrb $6, %eax, %xmm1, %xmm1
-; AVX512BW-NEXT:    vpextrb $12, %xmm2, %eax
-; AVX512BW-NEXT:    vpinsrb $7, %eax, %xmm1, %xmm1
-; AVX512BW-NEXT:    vextracti32x4 $2, %zmm0, %xmm2
-; AVX512BW-NEXT:    vpextrb $0, %xmm2, %eax
-; AVX512BW-NEXT:    vpinsrb $8, %eax, %xmm1, %xmm1
-; AVX512BW-NEXT:    vpextrb $4, %xmm2, %eax
-; AVX512BW-NEXT:    vpinsrb $9, %eax, %xmm1, %xmm1
-; AVX512BW-NEXT:    vpextrb $8, %xmm2, %eax
-; AVX512BW-NEXT:    vpinsrb $10, %eax, %xmm1, %xmm1
-; AVX512BW-NEXT:    vpextrb $12, %xmm2, %eax
-; AVX512BW-NEXT:    vpinsrb $11, %eax, %xmm1, %xmm1
-; AVX512BW-NEXT:    vextracti32x4 $3, %zmm0, %xmm0
-; AVX512BW-NEXT:    vpextrb $0, %xmm0, %eax
-; AVX512BW-NEXT:    vpinsrb $12, %eax, %xmm1, %xmm1
-; AVX512BW-NEXT:    vpextrb $4, %xmm0, %eax
-; AVX512BW-NEXT:    vpinsrb $13, %eax, %xmm1, %xmm1
-; AVX512BW-NEXT:    vpextrb $8, %xmm0, %eax
-; AVX512BW-NEXT:    vpinsrb $14, %eax, %xmm1, %xmm1
-; AVX512BW-NEXT:    vpextrb $12, %xmm0, %eax
-; AVX512BW-NEXT:    vpinsrb $15, %eax, %xmm1, %xmm0
-; AVX512BW-NEXT:    vmovdqa %xmm0, (%rsi)
-; AVX512BW-NEXT:    vzeroupper
-; AVX512BW-NEXT:    retq
-;
-; AVX512BWVL-LABEL: shuffle_v64i8_to_v16i8:
-; AVX512BWVL:       # BB#0:
-; AVX512BWVL-NEXT:    vmovdqu8 (%rdi), %zmm0
-; AVX512BWVL-NEXT:    vpextrb $4, %xmm0, %eax
-; AVX512BWVL-NEXT:    vpextrb $0, %xmm0, %ecx
-; AVX512BWVL-NEXT:    vmovd %ecx, %xmm1
-; AVX512BWVL-NEXT:    vpinsrb $1, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vpextrb $8, %xmm0, %eax
-; AVX512BWVL-NEXT:    vpinsrb $2, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vpextrb $12, %xmm0, %eax
-; AVX512BWVL-NEXT:    vpinsrb $3, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vextracti32x4 $1, %zmm0, %xmm2
-; AVX512BWVL-NEXT:    vpextrb $0, %xmm2, %eax
-; AVX512BWVL-NEXT:    vpinsrb $4, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vpextrb $4, %xmm2, %eax
-; AVX512BWVL-NEXT:    vpinsrb $5, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vpextrb $8, %xmm2, %eax
-; AVX512BWVL-NEXT:    vpinsrb $6, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vpextrb $12, %xmm2, %eax
-; AVX512BWVL-NEXT:    vpinsrb $7, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vextracti32x4 $2, %zmm0, %xmm2
-; AVX512BWVL-NEXT:    vpextrb $0, %xmm2, %eax
-; AVX512BWVL-NEXT:    vpinsrb $8, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vpextrb $4, %xmm2, %eax
-; AVX512BWVL-NEXT:    vpinsrb $9, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vpextrb $8, %xmm2, %eax
-; AVX512BWVL-NEXT:    vpinsrb $10, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vpextrb $12, %xmm2, %eax
-; AVX512BWVL-NEXT:    vpinsrb $11, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vextracti32x4 $3, %zmm0, %xmm0
-; AVX512BWVL-NEXT:    vpextrb $0, %xmm0, %eax
-; AVX512BWVL-NEXT:    vpinsrb $12, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vpextrb $4, %xmm0, %eax
-; AVX512BWVL-NEXT:    vpinsrb $13, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vpextrb $8, %xmm0, %eax
-; AVX512BWVL-NEXT:    vpinsrb $14, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vpextrb $12, %xmm0, %eax
-; AVX512BWVL-NEXT:    vpinsrb $15, %eax, %xmm1, %xmm0
-; AVX512BWVL-NEXT:    vmovdqu %xmm0, (%rsi)
-; AVX512BWVL-NEXT:    vzeroupper
-; AVX512BWVL-NEXT:    retq
+; AVX512-LABEL: shuffle_v64i8_to_v16i8:
+; AVX512:       # BB#0:
+; AVX512-NEXT:    vmovdqa32 (%rdi), %zmm0
+; AVX512-NEXT:    vpmovdb %zmm0, (%rsi)
+; AVX512-NEXT:    vzeroupper
+; AVX512-NEXT:    retq
   %vec = load <64 x i8>, <64 x i8>* %L
   %strided.vec = shufflevector <64 x i8> %vec, <64 x i8> undef, <16 x i32> <i32 0, i32 4, i32 8, i32 12, i32 16, i32 20, i32 24, i32 28, i32 32, i32 36, i32 40, i32 44, i32 48, i32 52, i32 56, i32 60>
   store <16 x i8> %strided.vec, <16 x i8>* %S
@@ -347,99 +175,12 @@
 }
 
 define void @shuffle_v32i16_to_v8i16(<32 x i16>* %L, <8 x i16>* %S) nounwind {
-; AVX512F-LABEL: shuffle_v32i16_to_v8i16:
-; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512F-NEXT:    vmovdqa 32(%rdi), %ymm1
-; AVX512F-NEXT:    vextracti128 $1, %ymm1, %xmm2
-; AVX512F-NEXT:    vpshufd {{.*#+}} xmm2 = xmm2[0,2,2,3]
-; AVX512F-NEXT:    vpshuflw {{.*#+}} xmm2 = xmm2[0,1,0,2,4,5,6,7]
-; AVX512F-NEXT:    vpshufd {{.*#+}} xmm1 = xmm1[0,2,2,3]
-; AVX512F-NEXT:    vpshuflw {{.*#+}} xmm1 = xmm1[0,1,0,2,4,5,6,7]
-; AVX512F-NEXT:    vpunpckldq {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1]
-; AVX512F-NEXT:    vextracti128 $1, %ymm0, %xmm2
-; AVX512F-NEXT:    vpshufd {{.*#+}} xmm2 = xmm2[0,2,2,3]
-; AVX512F-NEXT:    vpshuflw {{.*#+}} xmm2 = xmm2[0,2,2,3,4,5,6,7]
-; AVX512F-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
-; AVX512F-NEXT:    vpshuflw {{.*#+}} xmm0 = xmm0[0,2,2,3,4,5,6,7]
-; AVX512F-NEXT:    vpunpckldq {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1]
-; AVX512F-NEXT:    vpblendd {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,3]
-; AVX512F-NEXT:    vmovdqa %xmm0, (%rsi)
-; AVX512F-NEXT:    vzeroupper
-; AVX512F-NEXT:    retq
-;
-; AVX512VL-LABEL: shuffle_v32i16_to_v8i16:
-; AVX512VL:       # BB#0:
-; AVX512VL-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512VL-NEXT:    vmovdqa 32(%rdi), %ymm1
-; AVX512VL-NEXT:    vextracti128 $1, %ymm1, %xmm2
-; AVX512VL-NEXT:    vpshufd {{.*#+}} xmm2 = xmm2[0,2,2,3]
-; AVX512VL-NEXT:    vpshuflw {{.*#+}} xmm2 = xmm2[0,1,0,2,4,5,6,7]
-; AVX512VL-NEXT:    vpshufd {{.*#+}} xmm1 = xmm1[0,2,2,3]
-; AVX512VL-NEXT:    vpshuflw {{.*#+}} xmm1 = xmm1[0,1,0,2,4,5,6,7]
-; AVX512VL-NEXT:    vpunpckldq {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1]
-; AVX512VL-NEXT:    vextracti128 $1, %ymm0, %xmm2
-; AVX512VL-NEXT:    vpshufd {{.*#+}} xmm2 = xmm2[0,2,2,3]
-; AVX512VL-NEXT:    vpshuflw {{.*#+}} xmm2 = xmm2[0,2,2,3,4,5,6,7]
-; AVX512VL-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
-; AVX512VL-NEXT:    vpshuflw {{.*#+}} xmm0 = xmm0[0,2,2,3,4,5,6,7]
-; AVX512VL-NEXT:    vpunpckldq {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1]
-; AVX512VL-NEXT:    vpblendd {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,3]
-; AVX512VL-NEXT:    vmovdqa %xmm0, (%rsi)
-; AVX512VL-NEXT:    vzeroupper
-; AVX512VL-NEXT:    retq
-;
-; AVX512BW-LABEL: shuffle_v32i16_to_v8i16:
-; AVX512BW:       # BB#0:
-; AVX512BW-NEXT:    vmovdqu16 (%rdi), %zmm0
-; AVX512BW-NEXT:    vxorps %xmm1, %xmm1, %xmm1
-; AVX512BW-NEXT:    vmovss {{.*#+}} xmm1 = xmm0[0],xmm1[1,2,3]
-; AVX512BW-NEXT:    vpextrw $4, %xmm0, %eax
-; AVX512BW-NEXT:    vpinsrw $1, %eax, %xmm1, %xmm1
-; AVX512BW-NEXT:    vextracti32x4 $1, %zmm0, %xmm2
-; AVX512BW-NEXT:    vmovd %xmm2, %eax
-; AVX512BW-NEXT:    vpinsrw $2, %eax, %xmm1, %xmm1
-; AVX512BW-NEXT:    vpextrw $4, %xmm2, %eax
-; AVX512BW-NEXT:    vpinsrw $3, %eax, %xmm1, %xmm1
-; AVX512BW-NEXT:    vextracti32x4 $2, %zmm0, %xmm2
-; AVX512BW-NEXT:    vmovd %xmm2, %eax
-; AVX512BW-NEXT:    vpinsrw $4, %eax, %xmm1, %xmm1
-; AVX512BW-NEXT:    vpextrw $4, %xmm2, %eax
-; AVX512BW-NEXT:    vpinsrw $5, %eax, %xmm1, %xmm1
-; AVX512BW-NEXT:    vextracti32x4 $3, %zmm0, %xmm0
-; AVX512BW-NEXT:    vmovd %xmm0, %eax
-; AVX512BW-NEXT:    vpinsrw $6, %eax, %xmm1, %xmm1
-; AVX512BW-NEXT:    vpextrw $4, %xmm0, %eax
-; AVX512BW-NEXT:    vpinsrw $7, %eax, %xmm1, %xmm0
-; AVX512BW-NEXT:    vmovdqa %xmm0, (%rsi)
-; AVX512BW-NEXT:    vzeroupper
-; AVX512BW-NEXT:    retq
-;
-; AVX512BWVL-LABEL: shuffle_v32i16_to_v8i16:
-; AVX512BWVL:       # BB#0:
-; AVX512BWVL-NEXT:    vmovdqu16 (%rdi), %zmm0
-; AVX512BWVL-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vmovss {{.*#+}} xmm1 = xmm0[0],xmm1[1,2,3]
-; AVX512BWVL-NEXT:    vpextrw $4, %xmm0, %eax
-; AVX512BWVL-NEXT:    vpinsrw $1, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vextracti32x4 $1, %zmm0, %xmm2
-; AVX512BWVL-NEXT:    vmovd %xmm2, %eax
-; AVX512BWVL-NEXT:    vpinsrw $2, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vpextrw $4, %xmm2, %eax
-; AVX512BWVL-NEXT:    vpinsrw $3, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vextracti32x4 $2, %zmm0, %xmm2
-; AVX512BWVL-NEXT:    vmovd %xmm2, %eax
-; AVX512BWVL-NEXT:    vpinsrw $4, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vpextrw $4, %xmm2, %eax
-; AVX512BWVL-NEXT:    vpinsrw $5, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vextracti32x4 $3, %zmm0, %xmm0
-; AVX512BWVL-NEXT:    vmovd %xmm0, %eax
-; AVX512BWVL-NEXT:    vpinsrw $6, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vpextrw $4, %xmm0, %eax
-; AVX512BWVL-NEXT:    vpinsrw $7, %eax, %xmm1, %xmm0
-; AVX512BWVL-NEXT:    vmovdqu %xmm0, (%rsi)
-; AVX512BWVL-NEXT:    vzeroupper
-; AVX512BWVL-NEXT:    retq
+; AVX512-LABEL: shuffle_v32i16_to_v8i16:
+; AVX512:       # BB#0:
+; AVX512-NEXT:    vmovdqa64 (%rdi), %zmm0
+; AVX512-NEXT:    vpmovqw %zmm0, (%rsi)
+; AVX512-NEXT:    vzeroupper
+; AVX512-NEXT:    retq
   %vec = load <32 x i16>, <32 x i16>* %L
   %strided.vec = shufflevector <32 x i16> %vec, <32 x i16> undef, <8 x i32> <i32 0, i32 4, i32 8, i32 12, i32 16, i32 20, i32 24, i32 28>
   store <8 x i16> %strided.vec, <8 x i16>* %S
@@ -461,95 +202,12 @@
 }
 
 define void @shuffle_v64i8_to_v8i8(<64 x i8>* %L, <8 x i8>* %S) nounwind {
-; AVX512F-LABEL: shuffle_v64i8_to_v8i8:
-; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512F-NEXT:    vmovdqa 32(%rdi), %ymm1
-; AVX512F-NEXT:    vextracti128 $1, %ymm1, %xmm2
-; AVX512F-NEXT:    vmovdqa {{.*#+}} xmm3 = <u,u,0,8,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512F-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512F-NEXT:    vpshufb %xmm3, %xmm1, %xmm1
-; AVX512F-NEXT:    vpunpcklwd {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1],xmm1[2],xmm2[2],xmm1[3],xmm2[3]
-; AVX512F-NEXT:    vextracti128 $1, %ymm0, %xmm2
-; AVX512F-NEXT:    vmovdqa {{.*#+}} xmm3 = <0,8,u,u,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512F-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512F-NEXT:    vpshufb %xmm3, %xmm0, %xmm0
-; AVX512F-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3]
-; AVX512F-NEXT:    vpblendd {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2,3]
-; AVX512F-NEXT:    vmovq %xmm0, (%rsi)
-; AVX512F-NEXT:    vzeroupper
-; AVX512F-NEXT:    retq
-;
-; AVX512VL-LABEL: shuffle_v64i8_to_v8i8:
-; AVX512VL:       # BB#0:
-; AVX512VL-NEXT:    vmovdqa (%rdi), %ymm0
-; AVX512VL-NEXT:    vmovdqa 32(%rdi), %ymm1
-; AVX512VL-NEXT:    vextracti128 $1, %ymm1, %xmm2
-; AVX512VL-NEXT:    vmovdqa {{.*#+}} xmm3 = <u,u,0,8,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512VL-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512VL-NEXT:    vpshufb %xmm3, %xmm1, %xmm1
-; AVX512VL-NEXT:    vpunpcklwd {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1],xmm1[2],xmm2[2],xmm1[3],xmm2[3]
-; AVX512VL-NEXT:    vextracti128 $1, %ymm0, %xmm2
-; AVX512VL-NEXT:    vmovdqa {{.*#+}} xmm3 = <0,8,u,u,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512VL-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512VL-NEXT:    vpshufb %xmm3, %xmm0, %xmm0
-; AVX512VL-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3]
-; AVX512VL-NEXT:    vpblendd {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2,3]
-; AVX512VL-NEXT:    vmovq %xmm0, (%rsi)
-; AVX512VL-NEXT:    vzeroupper
-; AVX512VL-NEXT:    retq
-;
-; AVX512BW-LABEL: shuffle_v64i8_to_v8i8:
-; AVX512BW:       # BB#0:
-; AVX512BW-NEXT:    vmovdqu8 (%rdi), %zmm0
-; AVX512BW-NEXT:    vextracti32x4 $3, %zmm0, %xmm1
-; AVX512BW-NEXT:    vpextrb $8, %xmm1, %r8d
-; AVX512BW-NEXT:    vpextrb $0, %xmm1, %r9d
-; AVX512BW-NEXT:    vextracti32x4 $2, %zmm0, %xmm1
-; AVX512BW-NEXT:    vpextrb $8, %xmm1, %r10d
-; AVX512BW-NEXT:    vpextrb $0, %xmm1, %r11d
-; AVX512BW-NEXT:    vextracti32x4 $1, %zmm0, %xmm1
-; AVX512BW-NEXT:    vpextrb $8, %xmm1, %eax
-; AVX512BW-NEXT:    vpextrb $0, %xmm1, %ecx
-; AVX512BW-NEXT:    vpextrb $8, %xmm0, %edx
-; AVX512BW-NEXT:    vpextrb $0, %xmm0, %edi
-; AVX512BW-NEXT:    vmovd %edi, %xmm0
-; AVX512BW-NEXT:    vpinsrb $1, %edx, %xmm0, %xmm0
-; AVX512BW-NEXT:    vpinsrb $2, %ecx, %xmm0, %xmm0
-; AVX512BW-NEXT:    vpinsrb $3, %eax, %xmm0, %xmm0
-; AVX512BW-NEXT:    vpinsrb $4, %r11d, %xmm0, %xmm0
-; AVX512BW-NEXT:    vpinsrb $5, %r10d, %xmm0, %xmm0
-; AVX512BW-NEXT:    vpinsrb $6, %r9d, %xmm0, %xmm0
-; AVX512BW-NEXT:    vpinsrb $7, %r8d, %xmm0, %xmm0
-; AVX512BW-NEXT:    vmovq %xmm0, (%rsi)
-; AVX512BW-NEXT:    vzeroupper
-; AVX512BW-NEXT:    retq
-;
-; AVX512BWVL-LABEL: shuffle_v64i8_to_v8i8:
-; AVX512BWVL:       # BB#0:
-; AVX512BWVL-NEXT:    vmovdqu8 (%rdi), %zmm0
-; AVX512BWVL-NEXT:    vpextrb $8, %xmm0, %eax
-; AVX512BWVL-NEXT:    vpextrb $0, %xmm0, %ecx
-; AVX512BWVL-NEXT:    vmovd %ecx, %xmm1
-; AVX512BWVL-NEXT:    vpinsrb $2, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vextracti32x4 $1, %zmm0, %xmm2
-; AVX512BWVL-NEXT:    vpextrb $0, %xmm2, %eax
-; AVX512BWVL-NEXT:    vpinsrb $4, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vpextrb $8, %xmm2, %eax
-; AVX512BWVL-NEXT:    vpinsrb $6, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vextracti32x4 $2, %zmm0, %xmm2
-; AVX512BWVL-NEXT:    vpextrb $0, %xmm2, %eax
-; AVX512BWVL-NEXT:    vpinsrb $8, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vpextrb $8, %xmm2, %eax
-; AVX512BWVL-NEXT:    vpinsrb $10, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vextracti32x4 $3, %zmm0, %xmm0
-; AVX512BWVL-NEXT:    vpextrb $0, %xmm0, %eax
-; AVX512BWVL-NEXT:    vpinsrb $12, %eax, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vpextrb $8, %xmm0, %eax
-; AVX512BWVL-NEXT:    vpinsrb $14, %eax, %xmm1, %xmm0
-; AVX512BWVL-NEXT:    vpmovwb %xmm0, (%rsi)
-; AVX512BWVL-NEXT:    vzeroupper
-; AVX512BWVL-NEXT:    retq
+; AVX512-LABEL: shuffle_v64i8_to_v8i8:
+; AVX512:       # BB#0:
+; AVX512-NEXT:    vmovdqa64 (%rdi), %zmm0
+; AVX512-NEXT:    vpmovqb %zmm0, (%rsi)
+; AVX512-NEXT:    vzeroupper
+; AVX512-NEXT:    retq
   %vec = load <64 x i8>, <64 x i8>* %L
   %strided.vec = shufflevector <64 x i8> %vec, <64 x i8> undef, <8 x i32> <i32 0, i32 8, i32 16, i32 24, i32 32, i32 40, i32 48, i32 56>
   store <8 x i8> %strided.vec, <8 x i8>* %S