diff --git a/llvm/lib/Target/X86/X86ISelLowering.cpp b/llvm/lib/Target/X86/X86ISelLowering.cpp
index dae29807acf0..a3f61b5ee352 100644
--- a/llvm/lib/Target/X86/X86ISelLowering.cpp
+++ b/llvm/lib/Target/X86/X86ISelLowering.cpp
@@ -937,6 +937,15 @@ X86TargetLowering::X86TargetLowering(const X86TargetMachine &TM,
     setOperationAction(ISD::SIGN_EXTEND_VECTOR_INREG, MVT::v4i32, Custom);
     setOperationAction(ISD::SIGN_EXTEND_VECTOR_INREG, MVT::v8i16, Custom);
 
+    if (ExperimentalVectorWideningLegalization) {
+      setOperationAction(ISD::TRUNCATE, MVT::v2i8,  Custom);
+      setOperationAction(ISD::TRUNCATE, MVT::v2i16, Custom);
+      setOperationAction(ISD::TRUNCATE, MVT::v2i32, Custom);
+      setOperationAction(ISD::TRUNCATE, MVT::v4i8,  Custom);
+      setOperationAction(ISD::TRUNCATE, MVT::v4i16, Custom);
+      setOperationAction(ISD::TRUNCATE, MVT::v8i8,  Custom);
+    }
+
     // In the customized shift lowering, the legal v4i32/v2i64 cases
     // in AVX2 will be recognized.
     for (auto VT : { MVT::v16i8, MVT::v8i16, MVT::v4i32, MVT::v2i64 }) {
@@ -17917,6 +17926,10 @@ SDValue X86TargetLowering::LowerTRUNCATE(SDValue Op, SelectionDAG &DAG) const {
   assert(VT.getVectorNumElements() == InVT.getVectorNumElements() &&
          "Invalid TRUNCATE operation");
 
+  // If called by the legalizer just return.
+  if (!DAG.getTargetLoweringInfo().isTypeLegal(InVT))
+    return SDValue();
+
   if (VT.getVectorElementType() == MVT::i1)
     return LowerTruncateVecI1(Op, DAG, Subtarget);
 
@@ -26200,6 +26213,57 @@ void X86TargetLowering::ReplaceNodeResults(SDNode *N,
     Results.push_back(V);
     return;
   }
+  case ISD::TRUNCATE: {
+    MVT VT = N->getSimpleValueType(0);
+    if (getTypeAction(*DAG.getContext(), VT) != TypeWidenVector)
+      return;
+
+    // The generic legalizer will try to widen the input type to the same
+    // number of elements as the widened result type. But this isn't always
+    // the best thing so do some custom legalization to avoid some cases.
+    MVT WidenVT = getTypeToTransformTo(*DAG.getContext(), VT).getSimpleVT();
+    SDValue In = N->getOperand(0);
+    EVT InVT = In.getValueType();
+
+    unsigned InBits = InVT.getSizeInBits();
+    if (128 % InBits == 0) {
+      // 128 bit and smaller inputs should avoid truncate all together and
+      // just use a build_vector that will become a shuffle.
+      // TODO: Widen and use a shuffle directly?
+      MVT InEltVT = InVT.getSimpleVT().getVectorElementType();
+      EVT EltVT = VT.getVectorElementType();
+      unsigned WidenNumElts = WidenVT.getVectorNumElements();
+      SmallVector<SDValue, 16> Ops(WidenNumElts, DAG.getUNDEF(EltVT));
+      // Use the original element count so we don't do more scalar opts than
+      // necessary.
+      unsigned MinElts = VT.getVectorNumElements();
+      for (unsigned i=0; i < MinElts; ++i) {
+        SDValue Val = DAG.getNode(ISD::EXTRACT_VECTOR_ELT, dl, InEltVT, In,
+                                  DAG.getIntPtrConstant(i, dl));
+        Ops[i] = DAG.getNode(ISD::TRUNCATE, dl, EltVT, Val);
+      }
+      Results.push_back(DAG.getBuildVector(WidenVT, dl, Ops));
+      return;
+    }
+    // With AVX512 there are some cases that can use a target specific
+    // truncate node to go from 256/512 to less than 128 with zeros in the
+    // upper elements of the 128 bit result.
+    if (Subtarget.hasAVX512() && isTypeLegal(InVT)) {
+      // We can use VTRUNC directly if for 256 bits with VLX or for any 512.
+      if ((InBits == 256 && Subtarget.hasVLX()) || InBits == 512) {
+        Results.push_back(DAG.getNode(X86ISD::VTRUNC, dl, WidenVT, In));
+        return;
+      }
+      // There's one case we can widen to 512 bits and use VTRUNC.
+      if (InVT == MVT::v4i64 && VT == MVT::v4i8 && isTypeLegal(MVT::v8i64)) {
+        In = DAG.getNode(ISD::CONCAT_VECTORS, dl, MVT::v8i64, In,
+                         DAG.getUNDEF(MVT::v4i64));
+        Results.push_back(DAG.getNode(X86ISD::VTRUNC, dl, WidenVT, In));
+        return;
+      }
+    }
+    return;
+  }
   case ISD::FP_TO_SINT:
   case ISD::FP_TO_UINT: {
     bool IsSigned = N->getOpcode() == ISD::FP_TO_SINT;
diff --git a/llvm/test/CodeGen/X86/avx512-cvt-widen.ll b/llvm/test/CodeGen/X86/avx512-cvt-widen.ll
index 0ea613184ab3..a4d9a8272a38 100644
--- a/llvm/test/CodeGen/X86/avx512-cvt-widen.ll
+++ b/llvm/test/CodeGen/X86/avx512-cvt-widen.ll
@@ -1645,9 +1645,8 @@ define <2 x double> @sbto2f64(<2 x double> %a) {
 ; NOVL:       # %bb.0:
 ; NOVL-NEXT:    vxorpd %xmm1, %xmm1, %xmm1
 ; NOVL-NEXT:    vcmpltpd %xmm0, %xmm1, %xmm0
-; NOVL-NEXT:    vpmovqd %zmm0, %ymm0
+; NOVL-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[0,2,2,3]
 ; NOVL-NEXT:    vcvtdq2pd %xmm0, %xmm0
-; NOVL-NEXT:    vzeroupper
 ; NOVL-NEXT:    retq
 ;
 ; VLDQ-LABEL: sbto2f64:
diff --git a/llvm/test/CodeGen/X86/shuffle-vs-trunc-128-widen.ll b/llvm/test/CodeGen/X86/shuffle-vs-trunc-128-widen.ll
index 67901e992f4e..fae69ea5d80e 100644
--- a/llvm/test/CodeGen/X86/shuffle-vs-trunc-128-widen.ll
+++ b/llvm/test/CodeGen/X86/shuffle-vs-trunc-128-widen.ll
@@ -75,27 +75,22 @@ define void @trunc_v8i16_to_v8i8(<16 x i8>* %L, <8 x i8>* %S) nounwind {
 ; AVX512F-LABEL: trunc_v8i16_to_v8i8:
 ; AVX512F:       # %bb.0:
 ; AVX512F-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512F-NEXT:    vpmovzxwd {{.*#+}} zmm0 = ymm0[0],zero,ymm0[1],zero,ymm0[2],zero,ymm0[3],zero,ymm0[4],zero,ymm0[5],zero,ymm0[6],zero,ymm0[7],zero,ymm0[8],zero,ymm0[9],zero,ymm0[10],zero,ymm0[11],zero,ymm0[12],zero,ymm0[13],zero,ymm0[14],zero,ymm0[15],zero
-; AVX512F-NEXT:    vpmovdb %zmm0, %xmm0
+; 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
 ;
 ; AVX512VL-LABEL: trunc_v8i16_to_v8i8:
 ; AVX512VL:       # %bb.0:
 ; AVX512VL-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512VL-NEXT:    vpmovzxwd {{.*#+}} zmm0 = ymm0[0],zero,ymm0[1],zero,ymm0[2],zero,ymm0[3],zero,ymm0[4],zero,ymm0[5],zero,ymm0[6],zero,ymm0[7],zero,ymm0[8],zero,ymm0[9],zero,ymm0[10],zero,ymm0[11],zero,ymm0[12],zero,ymm0[13],zero,ymm0[14],zero,ymm0[15],zero
-; AVX512VL-NEXT:    vpmovdb %zmm0, %xmm0
+; AVX512VL-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u]
 ; AVX512VL-NEXT:    vmovq %xmm0, (%rsi)
-; AVX512VL-NEXT:    vzeroupper
 ; AVX512VL-NEXT:    retq
 ;
 ; AVX512BW-LABEL: trunc_v8i16_to_v8i8:
 ; AVX512BW:       # %bb.0:
 ; AVX512BW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512BW-NEXT:    vpmovwb %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: trunc_v8i16_to_v8i8:
@@ -171,9 +166,8 @@ define void @trunc_v4i32_to_v4i16(<8 x i16>* %L, <4 x i16>* %S) nounwind {
 ; AVX512F-LABEL: trunc_v4i32_to_v4i16:
 ; AVX512F:       # %bb.0:
 ; AVX512F-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512F-NEXT:    vpmovdw %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
 ;
 ; AVX512VL-LABEL: trunc_v4i32_to_v4i16:
@@ -185,9 +179,8 @@ define void @trunc_v4i32_to_v4i16(<8 x i16>* %L, <4 x i16>* %S) nounwind {
 ; AVX512BW-LABEL: trunc_v4i32_to_v4i16:
 ; AVX512BW:       # %bb.0:
 ; AVX512BW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512BW-NEXT:    vpmovdw %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: trunc_v4i32_to_v4i16:
@@ -233,35 +226,16 @@ define void @trunc_v2i64_to_v2i32(<4 x i32>* %L, <2 x i32>* %S) nounwind {
 ; SSE-NEXT:    movq %xmm0, (%rsi)
 ; SSE-NEXT:    retq
 ;
-; AVX1-LABEL: trunc_v2i64_to_v2i32:
-; AVX1:       # %bb.0:
-; AVX1-NEXT:    vpermilps {{.*#+}} xmm0 = mem[0,2,2,3]
-; AVX1-NEXT:    vmovlps %xmm0, (%rsi)
-; AVX1-NEXT:    retq
-;
-; AVX2-SLOW-LABEL: trunc_v2i64_to_v2i32:
-; AVX2-SLOW:       # %bb.0:
-; AVX2-SLOW-NEXT:    vmovaps (%rdi), %xmm0
-; AVX2-SLOW-NEXT:    vpermilps {{.*#+}} ymm0 = ymm0[0,2,2,3,4,6,6,7]
-; AVX2-SLOW-NEXT:    vmovlps %xmm0, (%rsi)
-; AVX2-SLOW-NEXT:    vzeroupper
-; AVX2-SLOW-NEXT:    retq
-;
-; AVX2-FAST-LABEL: trunc_v2i64_to_v2i32:
-; AVX2-FAST:       # %bb.0:
-; AVX2-FAST-NEXT:    vmovaps (%rdi), %xmm0
-; AVX2-FAST-NEXT:    vmovaps {{.*#+}} ymm1 = [0,2,4,6,4,6,6,7]
-; AVX2-FAST-NEXT:    vpermps %ymm0, %ymm1, %ymm0
-; AVX2-FAST-NEXT:    vmovlps %xmm0, (%rsi)
-; AVX2-FAST-NEXT:    vzeroupper
-; AVX2-FAST-NEXT:    retq
+; AVX-LABEL: trunc_v2i64_to_v2i32:
+; AVX:       # %bb.0:
+; AVX-NEXT:    vpermilps {{.*#+}} xmm0 = mem[0,2,2,3]
+; AVX-NEXT:    vmovlps %xmm0, (%rsi)
+; AVX-NEXT:    retq
 ;
 ; AVX512F-LABEL: trunc_v2i64_to_v2i32:
 ; AVX512F:       # %bb.0:
-; AVX512F-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512F-NEXT:    vpmovqd %zmm0, %ymm0
-; AVX512F-NEXT:    vmovq %xmm0, (%rsi)
-; AVX512F-NEXT:    vzeroupper
+; AVX512F-NEXT:    vpermilps {{.*#+}} xmm0 = mem[0,2,2,3]
+; AVX512F-NEXT:    vmovlps %xmm0, (%rsi)
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512VL-LABEL: trunc_v2i64_to_v2i32:
@@ -272,10 +246,8 @@ define void @trunc_v2i64_to_v2i32(<4 x i32>* %L, <2 x i32>* %S) nounwind {
 ;
 ; AVX512BW-LABEL: trunc_v2i64_to_v2i32:
 ; AVX512BW:       # %bb.0:
-; AVX512BW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512BW-NEXT:    vpmovqd %zmm0, %ymm0
-; AVX512BW-NEXT:    vmovq %xmm0, (%rsi)
-; AVX512BW-NEXT:    vzeroupper
+; AVX512BW-NEXT:    vpermilps {{.*#+}} xmm0 = mem[0,2,2,3]
+; AVX512BW-NEXT:    vmovlps %xmm0, (%rsi)
 ; AVX512BW-NEXT:    retq
 ;
 ; AVX512BWVL-LABEL: trunc_v2i64_to_v2i32:
@@ -353,9 +325,8 @@ define void @trunc_v4i32_to_v4i8(<16 x i8>* %L, <4 x i8>* %S) nounwind {
 ; AVX512F-LABEL: trunc_v4i32_to_v4i8:
 ; AVX512F:       # %bb.0:
 ; AVX512F-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512F-NEXT:    vpmovdb %zmm0, %xmm0
+; 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
 ;
 ; AVX512VL-LABEL: trunc_v4i32_to_v4i8:
@@ -367,9 +338,8 @@ define void @trunc_v4i32_to_v4i8(<16 x i8>* %L, <4 x i8>* %S) nounwind {
 ; AVX512BW-LABEL: trunc_v4i32_to_v4i8:
 ; AVX512BW:       # %bb.0:
 ; AVX512BW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512BW-NEXT:    vpmovdb %zmm0, %xmm0
+; 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: trunc_v4i32_to_v4i8:
@@ -477,10 +447,9 @@ define void @trunc_v2i64_to_v2i16(<8 x i16>* %L, <2 x i16>* %S) nounwind {
 ;
 ; AVX512F-LABEL: trunc_v2i64_to_v2i16:
 ; AVX512F:       # %bb.0:
-; AVX512F-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512F-NEXT:    vpmovqw %zmm0, %xmm0
+; AVX512F-NEXT:    vpshufd {{.*#+}} xmm0 = mem[0,2,2,3]
+; AVX512F-NEXT:    vpshuflw {{.*#+}} xmm0 = xmm0[0,2,2,3,4,5,6,7]
 ; AVX512F-NEXT:    vmovd %xmm0, (%rsi)
-; AVX512F-NEXT:    vzeroupper
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512VL-LABEL: trunc_v2i64_to_v2i16:
@@ -492,9 +461,8 @@ define void @trunc_v2i64_to_v2i16(<8 x i16>* %L, <2 x i16>* %S) nounwind {
 ; AVX512BW-LABEL: trunc_v2i64_to_v2i16:
 ; AVX512BW:       # %bb.0:
 ; AVX512BW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512BW-NEXT:    vpmovqw %zmm0, %xmm0
+; AVX512BW-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,8,9,8,9,10,11,8,9,10,11,12,13,14,15]
 ; AVX512BW-NEXT:    vmovd %xmm0, (%rsi)
-; AVX512BW-NEXT:    vzeroupper
 ; AVX512BW-NEXT:    retq
 ;
 ; AVX512BWVL-LABEL: trunc_v2i64_to_v2i16:
diff --git a/llvm/test/CodeGen/X86/shuffle-vs-trunc-256-widen.ll b/llvm/test/CodeGen/X86/shuffle-vs-trunc-256-widen.ll
index 7ac59a667c1c..940f475080f1 100644
--- a/llvm/test/CodeGen/X86/shuffle-vs-trunc-256-widen.ll
+++ b/llvm/test/CodeGen/X86/shuffle-vs-trunc-256-widen.ll
@@ -381,13 +381,42 @@ define <2 x i64> @trunc_v8i32_to_v8i8_return_v2i64(<8 x i32> %vec) nounwind {
 ; AVX2-NEXT:    vzeroupper
 ; AVX2-NEXT:    retq
 ;
-; AVX512-LABEL: trunc_v8i32_to_v8i8_return_v2i64:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    # kill: def $ymm0 killed $ymm0 def $zmm0
-; AVX512-NEXT:    vpmovdb %zmm0, %xmm0
-; AVX512-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
-; AVX512-NEXT:    vzeroupper
-; AVX512-NEXT:    retq
+; AVX512F-LABEL: trunc_v8i32_to_v8i8_return_v2i64:
+; AVX512F:       # %bb.0:
+; AVX512F-NEXT:    # kill: def $ymm0 killed $ymm0 def $zmm0
+; AVX512F-NEXT:    vpmovdb %zmm0, %xmm0
+; AVX512F-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
+; AVX512F-NEXT:    vzeroupper
+; AVX512F-NEXT:    retq
+;
+; AVX512VL-LABEL: trunc_v8i32_to_v8i8_return_v2i64:
+; AVX512VL:       # %bb.0:
+; AVX512VL-NEXT:    vpmovdb %ymm0, %xmm0
+; AVX512VL-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
+; AVX512VL-NEXT:    vzeroupper
+; AVX512VL-NEXT:    retq
+;
+; AVX512BW-LABEL: trunc_v8i32_to_v8i8_return_v2i64:
+; AVX512BW:       # %bb.0:
+; AVX512BW-NEXT:    # kill: def $ymm0 killed $ymm0 def $zmm0
+; AVX512BW-NEXT:    vpmovdb %zmm0, %xmm0
+; AVX512BW-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
+; AVX512BW-NEXT:    vzeroupper
+; AVX512BW-NEXT:    retq
+;
+; AVX512BWVL-LABEL: trunc_v8i32_to_v8i8_return_v2i64:
+; AVX512BWVL:       # %bb.0:
+; AVX512BWVL-NEXT:    vpmovdb %ymm0, %xmm0
+; AVX512BWVL-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
+; AVX512BWVL-NEXT:    vzeroupper
+; AVX512BWVL-NEXT:    retq
+;
+; AVX512VBMIVL-LABEL: trunc_v8i32_to_v8i8_return_v2i64:
+; AVX512VBMIVL:       # %bb.0:
+; AVX512VBMIVL-NEXT:    vpmovdb %ymm0, %xmm0
+; AVX512VBMIVL-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
+; AVX512VBMIVL-NEXT:    vzeroupper
+; AVX512VBMIVL-NEXT:    retq
   %truncated.vec = trunc <8 x i32> %vec to <8 x i8>
   %bc = bitcast <8 x i8> %truncated.vec to i64
   %result = insertelement <2 x i64> zeroinitializer, i64 %bc, i32 0
@@ -536,13 +565,42 @@ define <16 x i8> @trunc_v8i32_to_v8i8_return_v16i8(<8 x i32> %vec) nounwind {
 ; AVX2-NEXT:    vzeroupper
 ; AVX2-NEXT:    retq
 ;
-; AVX512-LABEL: trunc_v8i32_to_v8i8_return_v16i8:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    # kill: def $ymm0 killed $ymm0 def $zmm0
-; AVX512-NEXT:    vpmovdb %zmm0, %xmm0
-; AVX512-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
-; AVX512-NEXT:    vzeroupper
-; AVX512-NEXT:    retq
+; AVX512F-LABEL: trunc_v8i32_to_v8i8_return_v16i8:
+; AVX512F:       # %bb.0:
+; AVX512F-NEXT:    # kill: def $ymm0 killed $ymm0 def $zmm0
+; AVX512F-NEXT:    vpmovdb %zmm0, %xmm0
+; AVX512F-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
+; AVX512F-NEXT:    vzeroupper
+; AVX512F-NEXT:    retq
+;
+; AVX512VL-LABEL: trunc_v8i32_to_v8i8_return_v16i8:
+; AVX512VL:       # %bb.0:
+; AVX512VL-NEXT:    vpmovdb %ymm0, %xmm0
+; AVX512VL-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
+; AVX512VL-NEXT:    vzeroupper
+; AVX512VL-NEXT:    retq
+;
+; AVX512BW-LABEL: trunc_v8i32_to_v8i8_return_v16i8:
+; AVX512BW:       # %bb.0:
+; AVX512BW-NEXT:    # kill: def $ymm0 killed $ymm0 def $zmm0
+; AVX512BW-NEXT:    vpmovdb %zmm0, %xmm0
+; AVX512BW-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
+; AVX512BW-NEXT:    vzeroupper
+; AVX512BW-NEXT:    retq
+;
+; AVX512BWVL-LABEL: trunc_v8i32_to_v8i8_return_v16i8:
+; AVX512BWVL:       # %bb.0:
+; AVX512BWVL-NEXT:    vpmovdb %ymm0, %xmm0
+; AVX512BWVL-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
+; AVX512BWVL-NEXT:    vzeroupper
+; AVX512BWVL-NEXT:    retq
+;
+; AVX512VBMIVL-LABEL: trunc_v8i32_to_v8i8_return_v16i8:
+; AVX512VBMIVL:       # %bb.0:
+; AVX512VBMIVL-NEXT:    vpmovdb %ymm0, %xmm0
+; AVX512VBMIVL-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
+; AVX512VBMIVL-NEXT:    vzeroupper
+; AVX512VBMIVL-NEXT:    retq
   %truncated = trunc <8 x i32> %vec to <8 x i8>
   %result = shufflevector <8 x i8> %truncated, <8 x i8> zeroinitializer, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
   ret <16 x i8> %result
@@ -586,13 +644,42 @@ define <2 x i64> @trunc_v4i64_to_v4i16_return_v2i64(<4 x i64> %vec) nounwind {
 ; AVX2-FAST-NEXT:    vzeroupper
 ; AVX2-FAST-NEXT:    retq
 ;
-; AVX512-LABEL: trunc_v4i64_to_v4i16_return_v2i64:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    # kill: def $ymm0 killed $ymm0 def $zmm0
-; AVX512-NEXT:    vpmovqw %zmm0, %xmm0
-; AVX512-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
-; AVX512-NEXT:    vzeroupper
-; AVX512-NEXT:    retq
+; AVX512F-LABEL: trunc_v4i64_to_v4i16_return_v2i64:
+; AVX512F:       # %bb.0:
+; AVX512F-NEXT:    # kill: def $ymm0 killed $ymm0 def $zmm0
+; AVX512F-NEXT:    vpmovqw %zmm0, %xmm0
+; AVX512F-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
+; AVX512F-NEXT:    vzeroupper
+; AVX512F-NEXT:    retq
+;
+; AVX512VL-LABEL: trunc_v4i64_to_v4i16_return_v2i64:
+; AVX512VL:       # %bb.0:
+; AVX512VL-NEXT:    vpmovqw %ymm0, %xmm0
+; AVX512VL-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
+; AVX512VL-NEXT:    vzeroupper
+; AVX512VL-NEXT:    retq
+;
+; AVX512BW-LABEL: trunc_v4i64_to_v4i16_return_v2i64:
+; AVX512BW:       # %bb.0:
+; AVX512BW-NEXT:    # kill: def $ymm0 killed $ymm0 def $zmm0
+; AVX512BW-NEXT:    vpmovqw %zmm0, %xmm0
+; AVX512BW-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
+; AVX512BW-NEXT:    vzeroupper
+; AVX512BW-NEXT:    retq
+;
+; AVX512BWVL-LABEL: trunc_v4i64_to_v4i16_return_v2i64:
+; AVX512BWVL:       # %bb.0:
+; AVX512BWVL-NEXT:    vpmovqw %ymm0, %xmm0
+; AVX512BWVL-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
+; AVX512BWVL-NEXT:    vzeroupper
+; AVX512BWVL-NEXT:    retq
+;
+; AVX512VBMIVL-LABEL: trunc_v4i64_to_v4i16_return_v2i64:
+; AVX512VBMIVL:       # %bb.0:
+; AVX512VBMIVL-NEXT:    vpmovqw %ymm0, %xmm0
+; AVX512VBMIVL-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
+; AVX512VBMIVL-NEXT:    vzeroupper
+; AVX512VBMIVL-NEXT:    retq
   %truncated = trunc <4 x i64> %vec to <4 x i16>
   %bc = bitcast <4 x i16> %truncated to i64
   %result = insertelement <2 x i64> zeroinitializer, i64 %bc, i32 0
@@ -764,13 +851,42 @@ define <8 x i16> @trunc_v4i64_to_v4i16_return_v8i16(<4 x i64> %vec) nounwind {
 ; AVX2-FAST-NEXT:    vzeroupper
 ; AVX2-FAST-NEXT:    retq
 ;
-; AVX512-LABEL: trunc_v4i64_to_v4i16_return_v8i16:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    # kill: def $ymm0 killed $ymm0 def $zmm0
-; AVX512-NEXT:    vpmovqw %zmm0, %xmm0
-; AVX512-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
-; AVX512-NEXT:    vzeroupper
-; AVX512-NEXT:    retq
+; AVX512F-LABEL: trunc_v4i64_to_v4i16_return_v8i16:
+; AVX512F:       # %bb.0:
+; AVX512F-NEXT:    # kill: def $ymm0 killed $ymm0 def $zmm0
+; AVX512F-NEXT:    vpmovqw %zmm0, %xmm0
+; AVX512F-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
+; AVX512F-NEXT:    vzeroupper
+; AVX512F-NEXT:    retq
+;
+; AVX512VL-LABEL: trunc_v4i64_to_v4i16_return_v8i16:
+; AVX512VL:       # %bb.0:
+; AVX512VL-NEXT:    vpmovqw %ymm0, %xmm0
+; AVX512VL-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
+; AVX512VL-NEXT:    vzeroupper
+; AVX512VL-NEXT:    retq
+;
+; AVX512BW-LABEL: trunc_v4i64_to_v4i16_return_v8i16:
+; AVX512BW:       # %bb.0:
+; AVX512BW-NEXT:    # kill: def $ymm0 killed $ymm0 def $zmm0
+; AVX512BW-NEXT:    vpmovqw %zmm0, %xmm0
+; AVX512BW-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
+; AVX512BW-NEXT:    vzeroupper
+; AVX512BW-NEXT:    retq
+;
+; AVX512BWVL-LABEL: trunc_v4i64_to_v4i16_return_v8i16:
+; AVX512BWVL:       # %bb.0:
+; AVX512BWVL-NEXT:    vpmovqw %ymm0, %xmm0
+; AVX512BWVL-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
+; AVX512BWVL-NEXT:    vzeroupper
+; AVX512BWVL-NEXT:    retq
+;
+; AVX512VBMIVL-LABEL: trunc_v4i64_to_v4i16_return_v8i16:
+; AVX512VBMIVL:       # %bb.0:
+; AVX512VBMIVL-NEXT:    vpmovqw %ymm0, %xmm0
+; AVX512VBMIVL-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
+; AVX512VBMIVL-NEXT:    vzeroupper
+; AVX512VBMIVL-NEXT:    retq
   %truncated = trunc <4 x i64> %vec to <4 x i16>
   %result = shufflevector <4 x i16> %truncated, <4 x i16> zeroinitializer, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
   ret <8 x i16> %result
@@ -801,17 +917,47 @@ define <16 x i8> @trunc_v4i64_to_v4i8_return_v16i8(<4 x i64> %vec) nounwind {
 ; AVX2-NEXT:    vzeroupper
 ; AVX2-NEXT:    retq
 ;
-; AVX512-LABEL: trunc_v4i64_to_v4i8_return_v16i8:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512-NEXT:    vmovdqa {{.*#+}} xmm2 = <0,8,u,u,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
-; AVX512-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
-; AVX512-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; AVX512-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX512-NEXT:    vpblendw {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,3,4,5,6,7]
-; AVX512-NEXT:    vzeroupper
-; AVX512-NEXT:    retq
+; AVX512F-LABEL: trunc_v4i64_to_v4i8_return_v16i8:
+; AVX512F:       # %bb.0:
+; AVX512F-NEXT:    # kill: def $ymm0 killed $ymm0 def $zmm0
+; AVX512F-NEXT:    vpmovqb %zmm0, %xmm0
+; AVX512F-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX512F-NEXT:    vpblendw {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,3,4,5,6,7]
+; AVX512F-NEXT:    vzeroupper
+; AVX512F-NEXT:    retq
+;
+; AVX512VL-LABEL: trunc_v4i64_to_v4i8_return_v16i8:
+; AVX512VL:       # %bb.0:
+; AVX512VL-NEXT:    vpmovqb %ymm0, %xmm0
+; AVX512VL-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX512VL-NEXT:    vpblendw {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,3,4,5,6,7]
+; AVX512VL-NEXT:    vzeroupper
+; AVX512VL-NEXT:    retq
+;
+; AVX512BW-LABEL: trunc_v4i64_to_v4i8_return_v16i8:
+; AVX512BW:       # %bb.0:
+; AVX512BW-NEXT:    # kill: def $ymm0 killed $ymm0 def $zmm0
+; AVX512BW-NEXT:    vpmovqb %zmm0, %xmm0
+; AVX512BW-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX512BW-NEXT:    vpblendw {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,3,4,5,6,7]
+; AVX512BW-NEXT:    vzeroupper
+; AVX512BW-NEXT:    retq
+;
+; AVX512BWVL-LABEL: trunc_v4i64_to_v4i8_return_v16i8:
+; AVX512BWVL:       # %bb.0:
+; AVX512BWVL-NEXT:    vpmovqb %ymm0, %xmm0
+; AVX512BWVL-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX512BWVL-NEXT:    vpblendw {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,3,4,5,6,7]
+; AVX512BWVL-NEXT:    vzeroupper
+; AVX512BWVL-NEXT:    retq
+;
+; AVX512VBMIVL-LABEL: trunc_v4i64_to_v4i8_return_v16i8:
+; AVX512VBMIVL:       # %bb.0:
+; AVX512VBMIVL-NEXT:    vpmovqb %ymm0, %xmm0
+; AVX512VBMIVL-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX512VBMIVL-NEXT:    vpblendw {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,3,4,5,6,7]
+; AVX512VBMIVL-NEXT:    vzeroupper
+; AVX512VBMIVL-NEXT:    retq
   %truncated = trunc <4 x i64> %vec to <4 x i8>
   %result = shufflevector <4 x i8> %truncated, <4 x i8> zeroinitializer, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 4, i32 5, i32 6, i32 7, i32 5, i32 5, i32 undef, i32 7>
   ret <16 x i8> %result
@@ -1025,13 +1171,10 @@ define void @trunc_v4i64_to_v4i8(<32 x i8>* %L, <4 x i8>* %S) nounwind {
 ;
 ; AVX512F-LABEL: trunc_v4i64_to_v4i8:
 ; AVX512F:       # %bb.0:
-; AVX512F-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512F-NEXT:    vmovdqa 16(%rdi), %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:    vmovdqa (%rdi), %ymm0
+; AVX512F-NEXT:    vpmovqb %zmm0, %xmm0
 ; AVX512F-NEXT:    vmovd %xmm0, (%rsi)
+; AVX512F-NEXT:    vzeroupper
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512VL-LABEL: trunc_v4i64_to_v4i8:
@@ -1043,13 +1186,10 @@ define void @trunc_v4i64_to_v4i8(<32 x i8>* %L, <4 x i8>* %S) nounwind {
 ;
 ; AVX512BW-LABEL: trunc_v4i64_to_v4i8:
 ; AVX512BW:       # %bb.0:
-; AVX512BW-NEXT:    vmovdqa (%rdi), %xmm0
-; AVX512BW-NEXT:    vmovdqa 16(%rdi), %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:    vmovdqa (%rdi), %ymm0
+; AVX512BW-NEXT:    vpmovqb %zmm0, %xmm0
 ; AVX512BW-NEXT:    vmovd %xmm0, (%rsi)
+; AVX512BW-NEXT:    vzeroupper
 ; AVX512BW-NEXT:    retq
 ;
 ; AVX512BWVL-LABEL: trunc_v4i64_to_v4i8:
diff --git a/llvm/test/CodeGen/X86/shuffle-vs-trunc-512-widen.ll b/llvm/test/CodeGen/X86/shuffle-vs-trunc-512-widen.ll
index a9e45243381f..e78d0e050a85 100644
--- a/llvm/test/CodeGen/X86/shuffle-vs-trunc-512-widen.ll
+++ b/llvm/test/CodeGen/X86/shuffle-vs-trunc-512-widen.ll
@@ -911,118 +911,12 @@ define <4 x double> @PR34175(<32 x i16>* %p) {
 }
 
 define <16 x i8> @trunc_v8i64_to_v8i8_return_v16i8(<8 x i64> %vec) nounwind {
-; AVX512F-LABEL: trunc_v8i64_to_v8i8_return_v16i8:
-; AVX512F:       # %bb.0:
-; AVX512F-NEXT:    vextracti32x4 $3, %zmm0, %xmm1
-; AVX512F-NEXT:    vpextrq $1, %xmm1, %r8
-; AVX512F-NEXT:    vmovq %xmm1, %r9
-; AVX512F-NEXT:    vextracti32x4 $2, %zmm0, %xmm1
-; AVX512F-NEXT:    vpextrq $1, %xmm1, %r10
-; AVX512F-NEXT:    vmovq %xmm1, %rsi
-; AVX512F-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512F-NEXT:    vpextrq $1, %xmm1, %rdi
-; AVX512F-NEXT:    vmovq %xmm1, %rax
-; AVX512F-NEXT:    vpextrq $1, %xmm0, %rcx
-; AVX512F-NEXT:    vmovq %xmm0, %rdx
-; AVX512F-NEXT:    vpxor %xmm0, %xmm0, %xmm0
-; AVX512F-NEXT:    vpinsrb $0, %edx, %xmm0, %xmm0
-; AVX512F-NEXT:    vpinsrb $1, %ecx, %xmm0, %xmm0
-; AVX512F-NEXT:    vpinsrb $2, %eax, %xmm0, %xmm0
-; AVX512F-NEXT:    vpinsrb $3, %edi, %xmm0, %xmm0
-; AVX512F-NEXT:    vpinsrb $4, %esi, %xmm0, %xmm0
-; AVX512F-NEXT:    vpinsrb $5, %r10d, %xmm0, %xmm0
-; AVX512F-NEXT:    vpinsrb $6, %r9d, %xmm0, %xmm0
-; AVX512F-NEXT:    vpinsrb $7, %r8d, %xmm0, %xmm0
-; AVX512F-NEXT:    vzeroupper
-; AVX512F-NEXT:    retq
-;
-; AVX512VL-LABEL: trunc_v8i64_to_v8i8_return_v16i8:
-; AVX512VL:       # %bb.0:
-; AVX512VL-NEXT:    vextracti32x4 $3, %zmm0, %xmm1
-; AVX512VL-NEXT:    vpextrq $1, %xmm1, %r8
-; AVX512VL-NEXT:    vmovq %xmm1, %r9
-; AVX512VL-NEXT:    vextracti32x4 $2, %zmm0, %xmm1
-; AVX512VL-NEXT:    vpextrq $1, %xmm1, %r10
-; AVX512VL-NEXT:    vmovq %xmm1, %rsi
-; AVX512VL-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX512VL-NEXT:    vpextrq $1, %xmm1, %rdi
-; AVX512VL-NEXT:    vmovq %xmm1, %rax
-; AVX512VL-NEXT:    vpextrq $1, %xmm0, %rcx
-; AVX512VL-NEXT:    vmovq %xmm0, %rdx
-; AVX512VL-NEXT:    vpxor %xmm0, %xmm0, %xmm0
-; AVX512VL-NEXT:    vpinsrb $0, %edx, %xmm0, %xmm0
-; AVX512VL-NEXT:    vpinsrb $1, %ecx, %xmm0, %xmm0
-; AVX512VL-NEXT:    vpinsrb $2, %eax, %xmm0, %xmm0
-; AVX512VL-NEXT:    vpinsrb $3, %edi, %xmm0, %xmm0
-; AVX512VL-NEXT:    vpinsrb $4, %esi, %xmm0, %xmm0
-; AVX512VL-NEXT:    vpinsrb $5, %r10d, %xmm0, %xmm0
-; AVX512VL-NEXT:    vpinsrb $6, %r9d, %xmm0, %xmm0
-; AVX512VL-NEXT:    vpinsrb $7, %r8d, %xmm0, %xmm0
-; AVX512VL-NEXT:    vzeroupper
-; AVX512VL-NEXT:    retq
-;
-; AVX512BW-LABEL: trunc_v8i64_to_v8i8_return_v16i8:
-; AVX512BW:       # %bb.0:
-; AVX512BW-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
-; AVX512BW-NEXT:    vextracti128 $1, %ymm1, %xmm2
-; AVX512BW-NEXT:    vmovdqa {{.*#+}} xmm3 = <u,u,0,8,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512BW-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512BW-NEXT:    vpshufb %xmm3, %xmm1, %xmm1
-; AVX512BW-NEXT:    vpunpcklwd {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1],xmm1[2],xmm2[2],xmm1[3],xmm2[3]
-; AVX512BW-NEXT:    vextracti128 $1, %ymm0, %xmm2
-; AVX512BW-NEXT:    vmovdqa {{.*#+}} xmm3 = <0,8,u,u,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512BW-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512BW-NEXT:    vpshufb %xmm3, %xmm0, %xmm0
-; AVX512BW-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3]
-; AVX512BW-NEXT:    vpblendd {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2,3]
-; AVX512BW-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
-; AVX512BW-NEXT:    vzeroupper
-; AVX512BW-NEXT:    retq
-;
-; AVX512BWVL-LABEL: trunc_v8i64_to_v8i8_return_v16i8:
-; AVX512BWVL:       # %bb.0:
-; AVX512BWVL-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
-; AVX512BWVL-NEXT:    vextracti128 $1, %ymm1, %xmm2
-; AVX512BWVL-NEXT:    vmovdqa {{.*#+}} xmm3 = <u,u,0,8,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512BWVL-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512BWVL-NEXT:    vpshufb %xmm3, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vpunpcklwd {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1],xmm1[2],xmm2[2],xmm1[3],xmm2[3]
-; AVX512BWVL-NEXT:    vextracti128 $1, %ymm0, %xmm2
-; AVX512BWVL-NEXT:    vmovdqa {{.*#+}} xmm3 = <0,8,u,u,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512BWVL-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512BWVL-NEXT:    vpshufb %xmm3, %xmm0, %xmm0
-; AVX512BWVL-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3]
-; AVX512BWVL-NEXT:    vpblendd {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2,3]
-; AVX512BWVL-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
-; AVX512BWVL-NEXT:    vzeroupper
-; AVX512BWVL-NEXT:    retq
-;
-; AVX512VBMI-LABEL: trunc_v8i64_to_v8i8_return_v16i8:
-; AVX512VBMI:       # %bb.0:
-; AVX512VBMI-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
-; AVX512VBMI-NEXT:    vextracti128 $1, %ymm1, %xmm2
-; AVX512VBMI-NEXT:    vmovdqa {{.*#+}} xmm3 = <u,u,0,8,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512VBMI-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512VBMI-NEXT:    vpshufb %xmm3, %xmm1, %xmm1
-; AVX512VBMI-NEXT:    vpunpcklwd {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1],xmm1[2],xmm2[2],xmm1[3],xmm2[3]
-; AVX512VBMI-NEXT:    vextracti128 $1, %ymm0, %xmm2
-; AVX512VBMI-NEXT:    vmovdqa {{.*#+}} xmm3 = <0,8,u,u,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512VBMI-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512VBMI-NEXT:    vpshufb %xmm3, %xmm0, %xmm0
-; AVX512VBMI-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3]
-; AVX512VBMI-NEXT:    vpblendd {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2,3]
-; AVX512VBMI-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
-; AVX512VBMI-NEXT:    vzeroupper
-; AVX512VBMI-NEXT:    retq
-;
-; AVX512VBMIVL-LABEL: trunc_v8i64_to_v8i8_return_v16i8:
-; AVX512VBMIVL:       # %bb.0:
-; AVX512VBMIVL-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
-; AVX512VBMIVL-NEXT:    vmovdqa {{.*#+}} ymm2 = <0,8,16,24,32,40,48,56,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512VBMIVL-NEXT:    vpermi2b %ymm1, %ymm0, %ymm2
-; AVX512VBMIVL-NEXT:    vmovq {{.*#+}} xmm0 = xmm2[0],zero
-; AVX512VBMIVL-NEXT:    vzeroupper
-; AVX512VBMIVL-NEXT:    retq
+; AVX512-LABEL: trunc_v8i64_to_v8i8_return_v16i8:
+; AVX512:       # %bb.0:
+; AVX512-NEXT:    vpmovqb %zmm0, %xmm0
+; AVX512-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
+; AVX512-NEXT:    vzeroupper
+; AVX512-NEXT:    retq
   %truncated = trunc <8 x i64> %vec to <8 x i8>
   %result = shufflevector <8 x i8> %truncated, <8 x i8> zeroinitializer, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
   ret <16 x i8> %result
diff --git a/llvm/test/CodeGen/X86/vector-trunc-packus-widen.ll b/llvm/test/CodeGen/X86/vector-trunc-packus-widen.ll
index 82904ef49fa2..2b456dde5240 100644
--- a/llvm/test/CodeGen/X86/vector-trunc-packus-widen.ll
+++ b/llvm/test/CodeGen/X86/vector-trunc-packus-widen.ll
@@ -1735,99 +1735,14 @@ define <8 x i8> @trunc_packus_v8i64_v8i8(<8 x i64> %a0) {
 ; AVX2-NEXT:    vzeroupper
 ; AVX2-NEXT:    retq
 ;
-; AVX512F-LABEL: trunc_packus_v8i64_v8i8:
-; AVX512F:       # %bb.0:
-; AVX512F-NEXT:    vpminsq {{.*}}(%rip){1to8}, %zmm0, %zmm0
-; AVX512F-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX512F-NEXT:    vpmaxsq %zmm1, %zmm0, %zmm0
-; AVX512F-NEXT:    vpextrq $1, %xmm0, %rax
-; AVX512F-NEXT:    vmovq %xmm0, %rcx
-; AVX512F-NEXT:    vmovd %ecx, %xmm1
-; AVX512F-NEXT:    vpinsrb $1, %eax, %xmm1, %xmm1
-; AVX512F-NEXT:    vextracti128 $1, %ymm0, %xmm2
-; AVX512F-NEXT:    vmovq %xmm2, %rax
-; AVX512F-NEXT:    vpinsrb $2, %eax, %xmm1, %xmm1
-; AVX512F-NEXT:    vpextrq $1, %xmm2, %rax
-; AVX512F-NEXT:    vpinsrb $3, %eax, %xmm1, %xmm1
-; AVX512F-NEXT:    vextracti32x4 $2, %zmm0, %xmm2
-; AVX512F-NEXT:    vmovq %xmm2, %rax
-; AVX512F-NEXT:    vpinsrb $4, %eax, %xmm1, %xmm1
-; AVX512F-NEXT:    vpextrq $1, %xmm2, %rax
-; AVX512F-NEXT:    vpinsrb $5, %eax, %xmm1, %xmm1
-; AVX512F-NEXT:    vextracti32x4 $3, %zmm0, %xmm0
-; AVX512F-NEXT:    vmovq %xmm0, %rax
-; AVX512F-NEXT:    vpinsrb $6, %eax, %xmm1, %xmm1
-; AVX512F-NEXT:    vpextrq $1, %xmm0, %rax
-; AVX512F-NEXT:    vpinsrb $7, %eax, %xmm1, %xmm0
-; AVX512F-NEXT:    vzeroupper
-; AVX512F-NEXT:    retq
-;
-; AVX512VL-LABEL: trunc_packus_v8i64_v8i8:
-; AVX512VL:       # %bb.0:
-; AVX512VL-NEXT:    vpminsq {{.*}}(%rip){1to8}, %zmm0, %zmm0
-; AVX512VL-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX512VL-NEXT:    vpmaxsq %zmm1, %zmm0, %zmm0
-; AVX512VL-NEXT:    vpextrq $1, %xmm0, %rax
-; AVX512VL-NEXT:    vmovq %xmm0, %rcx
-; AVX512VL-NEXT:    vmovd %ecx, %xmm1
-; AVX512VL-NEXT:    vpinsrb $1, %eax, %xmm1, %xmm1
-; AVX512VL-NEXT:    vextracti128 $1, %ymm0, %xmm2
-; AVX512VL-NEXT:    vmovq %xmm2, %rax
-; AVX512VL-NEXT:    vpinsrb $2, %eax, %xmm1, %xmm1
-; AVX512VL-NEXT:    vpextrq $1, %xmm2, %rax
-; AVX512VL-NEXT:    vpinsrb $3, %eax, %xmm1, %xmm1
-; AVX512VL-NEXT:    vextracti32x4 $2, %zmm0, %xmm2
-; AVX512VL-NEXT:    vmovq %xmm2, %rax
-; AVX512VL-NEXT:    vpinsrb $4, %eax, %xmm1, %xmm1
-; AVX512VL-NEXT:    vpextrq $1, %xmm2, %rax
-; AVX512VL-NEXT:    vpinsrb $5, %eax, %xmm1, %xmm1
-; AVX512VL-NEXT:    vextracti32x4 $3, %zmm0, %xmm0
-; AVX512VL-NEXT:    vmovq %xmm0, %rax
-; AVX512VL-NEXT:    vpinsrb $6, %eax, %xmm1, %xmm1
-; AVX512VL-NEXT:    vpextrq $1, %xmm0, %rax
-; AVX512VL-NEXT:    vpinsrb $7, %eax, %xmm1, %xmm0
-; AVX512VL-NEXT:    vzeroupper
-; AVX512VL-NEXT:    retq
-;
-; AVX512BW-LABEL: trunc_packus_v8i64_v8i8:
-; AVX512BW:       # %bb.0:
-; AVX512BW-NEXT:    vpminsq {{.*}}(%rip){1to8}, %zmm0, %zmm0
-; AVX512BW-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX512BW-NEXT:    vpmaxsq %zmm1, %zmm0, %zmm0
-; AVX512BW-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
-; AVX512BW-NEXT:    vextracti128 $1, %ymm1, %xmm2
-; AVX512BW-NEXT:    vmovdqa {{.*#+}} xmm3 = <u,u,0,8,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512BW-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512BW-NEXT:    vpshufb %xmm3, %xmm1, %xmm1
-; AVX512BW-NEXT:    vpunpcklwd {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1],xmm1[2],xmm2[2],xmm1[3],xmm2[3]
-; AVX512BW-NEXT:    vextracti128 $1, %ymm0, %xmm2
-; AVX512BW-NEXT:    vmovdqa {{.*#+}} xmm3 = <0,8,u,u,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512BW-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512BW-NEXT:    vpshufb %xmm3, %xmm0, %xmm0
-; AVX512BW-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3]
-; AVX512BW-NEXT:    vpblendd {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2,3]
-; AVX512BW-NEXT:    vzeroupper
-; AVX512BW-NEXT:    retq
-;
-; AVX512BWVL-LABEL: trunc_packus_v8i64_v8i8:
-; AVX512BWVL:       # %bb.0:
-; AVX512BWVL-NEXT:    vpminsq {{.*}}(%rip){1to8}, %zmm0, %zmm0
-; AVX512BWVL-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vpmaxsq %zmm1, %zmm0, %zmm0
-; AVX512BWVL-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
-; AVX512BWVL-NEXT:    vextracti128 $1, %ymm1, %xmm2
-; AVX512BWVL-NEXT:    vmovdqa {{.*#+}} xmm3 = <u,u,0,8,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512BWVL-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512BWVL-NEXT:    vpshufb %xmm3, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vpunpcklwd {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1],xmm1[2],xmm2[2],xmm1[3],xmm2[3]
-; AVX512BWVL-NEXT:    vextracti128 $1, %ymm0, %xmm2
-; AVX512BWVL-NEXT:    vmovdqa {{.*#+}} xmm3 = <0,8,u,u,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512BWVL-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512BWVL-NEXT:    vpshufb %xmm3, %xmm0, %xmm0
-; AVX512BWVL-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3]
-; AVX512BWVL-NEXT:    vpblendd {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2,3]
-; AVX512BWVL-NEXT:    vzeroupper
-; AVX512BWVL-NEXT:    retq
+; AVX512-LABEL: trunc_packus_v8i64_v8i8:
+; AVX512:       # %bb.0:
+; AVX512-NEXT:    vpminsq {{.*}}(%rip){1to8}, %zmm0, %zmm0
+; AVX512-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX512-NEXT:    vpmaxsq %zmm1, %zmm0, %zmm0
+; AVX512-NEXT:    vpmovqb %zmm0, %xmm0
+; AVX512-NEXT:    vzeroupper
+; AVX512-NEXT:    retq
   %1 = icmp slt <8 x i64> %a0, <i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255>
   %2 = select <8 x i1> %1, <8 x i64> %a0, <8 x i64> <i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255>
   %3 = icmp sgt <8 x i64> %2, zeroinitializer
@@ -3022,7 +2937,7 @@ define <8 x i8> @trunc_packus_v8i32_v8i8(<8 x i32> %a0) {
 ; AVX512VL-NEXT:    vpminsd {{.*}}(%rip){1to8}, %ymm0, %ymm0
 ; AVX512VL-NEXT:    vpxor %xmm1, %xmm1, %xmm1
 ; AVX512VL-NEXT:    vpmaxsd %ymm1, %ymm0, %ymm0
-; AVX512VL-NEXT:    vpmovdb %zmm0, %xmm0
+; AVX512VL-NEXT:    vpmovdb %ymm0, %xmm0
 ; AVX512VL-NEXT:    vzeroupper
 ; AVX512VL-NEXT:    retq
 ;
@@ -3041,7 +2956,7 @@ define <8 x i8> @trunc_packus_v8i32_v8i8(<8 x i32> %a0) {
 ; AVX512BWVL-NEXT:    vpminsd {{.*}}(%rip){1to8}, %ymm0, %ymm0
 ; AVX512BWVL-NEXT:    vpxor %xmm1, %xmm1, %xmm1
 ; AVX512BWVL-NEXT:    vpmaxsd %ymm1, %ymm0, %ymm0
-; AVX512BWVL-NEXT:    vpmovdb %zmm0, %xmm0
+; AVX512BWVL-NEXT:    vpmovdb %ymm0, %xmm0
 ; AVX512BWVL-NEXT:    vzeroupper
 ; AVX512BWVL-NEXT:    retq
   %1 = icmp slt <8 x i32> %a0, <i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255>
diff --git a/llvm/test/CodeGen/X86/vector-trunc-ssat-widen.ll b/llvm/test/CodeGen/X86/vector-trunc-ssat-widen.ll
index 11ed156316ab..d25d52549b4b 100644
--- a/llvm/test/CodeGen/X86/vector-trunc-ssat-widen.ll
+++ b/llvm/test/CodeGen/X86/vector-trunc-ssat-widen.ll
@@ -1633,95 +1633,13 @@ define <8 x i8> @trunc_ssat_v8i64_v8i8(<8 x i64> %a0) {
 ; AVX2-NEXT:    vzeroupper
 ; AVX2-NEXT:    retq
 ;
-; AVX512F-LABEL: trunc_ssat_v8i64_v8i8:
-; AVX512F:       # %bb.0:
-; AVX512F-NEXT:    vpminsq {{.*}}(%rip){1to8}, %zmm0, %zmm0
-; AVX512F-NEXT:    vpmaxsq {{.*}}(%rip){1to8}, %zmm0, %zmm0
-; AVX512F-NEXT:    vpextrq $1, %xmm0, %rax
-; AVX512F-NEXT:    vmovq %xmm0, %rcx
-; AVX512F-NEXT:    vmovd %ecx, %xmm1
-; AVX512F-NEXT:    vpinsrb $1, %eax, %xmm1, %xmm1
-; AVX512F-NEXT:    vextracti128 $1, %ymm0, %xmm2
-; AVX512F-NEXT:    vmovq %xmm2, %rax
-; AVX512F-NEXT:    vpinsrb $2, %eax, %xmm1, %xmm1
-; AVX512F-NEXT:    vpextrq $1, %xmm2, %rax
-; AVX512F-NEXT:    vpinsrb $3, %eax, %xmm1, %xmm1
-; AVX512F-NEXT:    vextracti32x4 $2, %zmm0, %xmm2
-; AVX512F-NEXT:    vmovq %xmm2, %rax
-; AVX512F-NEXT:    vpinsrb $4, %eax, %xmm1, %xmm1
-; AVX512F-NEXT:    vpextrq $1, %xmm2, %rax
-; AVX512F-NEXT:    vpinsrb $5, %eax, %xmm1, %xmm1
-; AVX512F-NEXT:    vextracti32x4 $3, %zmm0, %xmm0
-; AVX512F-NEXT:    vmovq %xmm0, %rax
-; AVX512F-NEXT:    vpinsrb $6, %eax, %xmm1, %xmm1
-; AVX512F-NEXT:    vpextrq $1, %xmm0, %rax
-; AVX512F-NEXT:    vpinsrb $7, %eax, %xmm1, %xmm0
-; AVX512F-NEXT:    vzeroupper
-; AVX512F-NEXT:    retq
-;
-; AVX512VL-LABEL: trunc_ssat_v8i64_v8i8:
-; AVX512VL:       # %bb.0:
-; AVX512VL-NEXT:    vpminsq {{.*}}(%rip){1to8}, %zmm0, %zmm0
-; AVX512VL-NEXT:    vpmaxsq {{.*}}(%rip){1to8}, %zmm0, %zmm0
-; AVX512VL-NEXT:    vpextrq $1, %xmm0, %rax
-; AVX512VL-NEXT:    vmovq %xmm0, %rcx
-; AVX512VL-NEXT:    vmovd %ecx, %xmm1
-; AVX512VL-NEXT:    vpinsrb $1, %eax, %xmm1, %xmm1
-; AVX512VL-NEXT:    vextracti128 $1, %ymm0, %xmm2
-; AVX512VL-NEXT:    vmovq %xmm2, %rax
-; AVX512VL-NEXT:    vpinsrb $2, %eax, %xmm1, %xmm1
-; AVX512VL-NEXT:    vpextrq $1, %xmm2, %rax
-; AVX512VL-NEXT:    vpinsrb $3, %eax, %xmm1, %xmm1
-; AVX512VL-NEXT:    vextracti32x4 $2, %zmm0, %xmm2
-; AVX512VL-NEXT:    vmovq %xmm2, %rax
-; AVX512VL-NEXT:    vpinsrb $4, %eax, %xmm1, %xmm1
-; AVX512VL-NEXT:    vpextrq $1, %xmm2, %rax
-; AVX512VL-NEXT:    vpinsrb $5, %eax, %xmm1, %xmm1
-; AVX512VL-NEXT:    vextracti32x4 $3, %zmm0, %xmm0
-; AVX512VL-NEXT:    vmovq %xmm0, %rax
-; AVX512VL-NEXT:    vpinsrb $6, %eax, %xmm1, %xmm1
-; AVX512VL-NEXT:    vpextrq $1, %xmm0, %rax
-; AVX512VL-NEXT:    vpinsrb $7, %eax, %xmm1, %xmm0
-; AVX512VL-NEXT:    vzeroupper
-; AVX512VL-NEXT:    retq
-;
-; AVX512BW-LABEL: trunc_ssat_v8i64_v8i8:
-; AVX512BW:       # %bb.0:
-; AVX512BW-NEXT:    vpminsq {{.*}}(%rip){1to8}, %zmm0, %zmm0
-; AVX512BW-NEXT:    vpmaxsq {{.*}}(%rip){1to8}, %zmm0, %zmm0
-; AVX512BW-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
-; AVX512BW-NEXT:    vextracti128 $1, %ymm1, %xmm2
-; AVX512BW-NEXT:    vmovdqa {{.*#+}} xmm3 = <u,u,0,8,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512BW-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512BW-NEXT:    vpshufb %xmm3, %xmm1, %xmm1
-; AVX512BW-NEXT:    vpunpcklwd {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1],xmm1[2],xmm2[2],xmm1[3],xmm2[3]
-; AVX512BW-NEXT:    vextracti128 $1, %ymm0, %xmm2
-; AVX512BW-NEXT:    vmovdqa {{.*#+}} xmm3 = <0,8,u,u,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512BW-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512BW-NEXT:    vpshufb %xmm3, %xmm0, %xmm0
-; AVX512BW-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3]
-; AVX512BW-NEXT:    vpblendd {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2,3]
-; AVX512BW-NEXT:    vzeroupper
-; AVX512BW-NEXT:    retq
-;
-; AVX512BWVL-LABEL: trunc_ssat_v8i64_v8i8:
-; AVX512BWVL:       # %bb.0:
-; AVX512BWVL-NEXT:    vpminsq {{.*}}(%rip){1to8}, %zmm0, %zmm0
-; AVX512BWVL-NEXT:    vpmaxsq {{.*}}(%rip){1to8}, %zmm0, %zmm0
-; AVX512BWVL-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
-; AVX512BWVL-NEXT:    vextracti128 $1, %ymm1, %xmm2
-; AVX512BWVL-NEXT:    vmovdqa {{.*#+}} xmm3 = <u,u,0,8,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512BWVL-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512BWVL-NEXT:    vpshufb %xmm3, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vpunpcklwd {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1],xmm1[2],xmm2[2],xmm1[3],xmm2[3]
-; AVX512BWVL-NEXT:    vextracti128 $1, %ymm0, %xmm2
-; AVX512BWVL-NEXT:    vmovdqa {{.*#+}} xmm3 = <0,8,u,u,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512BWVL-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512BWVL-NEXT:    vpshufb %xmm3, %xmm0, %xmm0
-; AVX512BWVL-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3]
-; AVX512BWVL-NEXT:    vpblendd {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2,3]
-; AVX512BWVL-NEXT:    vzeroupper
-; AVX512BWVL-NEXT:    retq
+; AVX512-LABEL: trunc_ssat_v8i64_v8i8:
+; AVX512:       # %bb.0:
+; AVX512-NEXT:    vpminsq {{.*}}(%rip){1to8}, %zmm0, %zmm0
+; AVX512-NEXT:    vpmaxsq {{.*}}(%rip){1to8}, %zmm0, %zmm0
+; AVX512-NEXT:    vpmovqb %zmm0, %xmm0
+; AVX512-NEXT:    vzeroupper
+; AVX512-NEXT:    retq
   %1 = icmp slt <8 x i64> %a0, <i64 127, i64 127, i64 127, i64 127, i64 127, i64 127, i64 127, i64 127>
   %2 = select <8 x i1> %1, <8 x i64> %a0, <8 x i64> <i64 127, i64 127, i64 127, i64 127, i64 127, i64 127, i64 127, i64 127>
   %3 = icmp sgt <8 x i64> %2, <i64 -128, i64 -128, i64 -128, i64 -128, i64 -128, i64 -128, i64 -128, i64 -128>
@@ -2975,7 +2893,7 @@ define <8 x i8> @trunc_ssat_v8i32_v8i8(<8 x i32> %a0) {
 ; AVX512VL:       # %bb.0:
 ; AVX512VL-NEXT:    vpminsd {{.*}}(%rip){1to8}, %ymm0, %ymm0
 ; AVX512VL-NEXT:    vpmaxsd {{.*}}(%rip){1to8}, %ymm0, %ymm0
-; AVX512VL-NEXT:    vpmovdb %zmm0, %xmm0
+; AVX512VL-NEXT:    vpmovdb %ymm0, %xmm0
 ; AVX512VL-NEXT:    vzeroupper
 ; AVX512VL-NEXT:    retq
 ;
@@ -2993,7 +2911,7 @@ define <8 x i8> @trunc_ssat_v8i32_v8i8(<8 x i32> %a0) {
 ; AVX512BWVL:       # %bb.0:
 ; AVX512BWVL-NEXT:    vpminsd {{.*}}(%rip){1to8}, %ymm0, %ymm0
 ; AVX512BWVL-NEXT:    vpmaxsd {{.*}}(%rip){1to8}, %ymm0, %ymm0
-; AVX512BWVL-NEXT:    vpmovdb %zmm0, %xmm0
+; AVX512BWVL-NEXT:    vpmovdb %ymm0, %xmm0
 ; AVX512BWVL-NEXT:    vzeroupper
 ; AVX512BWVL-NEXT:    retq
   %1 = icmp slt <8 x i32> %a0, <i32 127, i32 127, i32 127, i32 127, i32 127, i32 127, i32 127, i32 127>
diff --git a/llvm/test/CodeGen/X86/vector-trunc-usat-widen.ll b/llvm/test/CodeGen/X86/vector-trunc-usat-widen.ll
index 3dfa844ff829..9dd97b33069d 100644
--- a/llvm/test/CodeGen/X86/vector-trunc-usat-widen.ll
+++ b/llvm/test/CodeGen/X86/vector-trunc-usat-widen.ll
@@ -1216,91 +1216,12 @@ define <8 x i8> @trunc_usat_v8i64_v8i8(<8 x i64> %a0) {
 ; AVX2-NEXT:    vzeroupper
 ; AVX2-NEXT:    retq
 ;
-; AVX512F-LABEL: trunc_usat_v8i64_v8i8:
-; AVX512F:       # %bb.0:
-; AVX512F-NEXT:    vpminuq {{.*}}(%rip){1to8}, %zmm0, %zmm0
-; AVX512F-NEXT:    vpextrq $1, %xmm0, %rax
-; AVX512F-NEXT:    vmovq %xmm0, %rcx
-; AVX512F-NEXT:    vmovd %ecx, %xmm1
-; AVX512F-NEXT:    vpinsrb $1, %eax, %xmm1, %xmm1
-; AVX512F-NEXT:    vextracti128 $1, %ymm0, %xmm2
-; AVX512F-NEXT:    vmovq %xmm2, %rax
-; AVX512F-NEXT:    vpinsrb $2, %eax, %xmm1, %xmm1
-; AVX512F-NEXT:    vpextrq $1, %xmm2, %rax
-; AVX512F-NEXT:    vpinsrb $3, %eax, %xmm1, %xmm1
-; AVX512F-NEXT:    vextracti32x4 $2, %zmm0, %xmm2
-; AVX512F-NEXT:    vmovq %xmm2, %rax
-; AVX512F-NEXT:    vpinsrb $4, %eax, %xmm1, %xmm1
-; AVX512F-NEXT:    vpextrq $1, %xmm2, %rax
-; AVX512F-NEXT:    vpinsrb $5, %eax, %xmm1, %xmm1
-; AVX512F-NEXT:    vextracti32x4 $3, %zmm0, %xmm0
-; AVX512F-NEXT:    vmovq %xmm0, %rax
-; AVX512F-NEXT:    vpinsrb $6, %eax, %xmm1, %xmm1
-; AVX512F-NEXT:    vpextrq $1, %xmm0, %rax
-; AVX512F-NEXT:    vpinsrb $7, %eax, %xmm1, %xmm0
-; AVX512F-NEXT:    vzeroupper
-; AVX512F-NEXT:    retq
-;
-; AVX512VL-LABEL: trunc_usat_v8i64_v8i8:
-; AVX512VL:       # %bb.0:
-; AVX512VL-NEXT:    vpminuq {{.*}}(%rip){1to8}, %zmm0, %zmm0
-; AVX512VL-NEXT:    vpextrq $1, %xmm0, %rax
-; AVX512VL-NEXT:    vmovq %xmm0, %rcx
-; AVX512VL-NEXT:    vmovd %ecx, %xmm1
-; AVX512VL-NEXT:    vpinsrb $1, %eax, %xmm1, %xmm1
-; AVX512VL-NEXT:    vextracti128 $1, %ymm0, %xmm2
-; AVX512VL-NEXT:    vmovq %xmm2, %rax
-; AVX512VL-NEXT:    vpinsrb $2, %eax, %xmm1, %xmm1
-; AVX512VL-NEXT:    vpextrq $1, %xmm2, %rax
-; AVX512VL-NEXT:    vpinsrb $3, %eax, %xmm1, %xmm1
-; AVX512VL-NEXT:    vextracti32x4 $2, %zmm0, %xmm2
-; AVX512VL-NEXT:    vmovq %xmm2, %rax
-; AVX512VL-NEXT:    vpinsrb $4, %eax, %xmm1, %xmm1
-; AVX512VL-NEXT:    vpextrq $1, %xmm2, %rax
-; AVX512VL-NEXT:    vpinsrb $5, %eax, %xmm1, %xmm1
-; AVX512VL-NEXT:    vextracti32x4 $3, %zmm0, %xmm0
-; AVX512VL-NEXT:    vmovq %xmm0, %rax
-; AVX512VL-NEXT:    vpinsrb $6, %eax, %xmm1, %xmm1
-; AVX512VL-NEXT:    vpextrq $1, %xmm0, %rax
-; AVX512VL-NEXT:    vpinsrb $7, %eax, %xmm1, %xmm0
-; AVX512VL-NEXT:    vzeroupper
-; AVX512VL-NEXT:    retq
-;
-; AVX512BW-LABEL: trunc_usat_v8i64_v8i8:
-; AVX512BW:       # %bb.0:
-; AVX512BW-NEXT:    vpminuq {{.*}}(%rip){1to8}, %zmm0, %zmm0
-; AVX512BW-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
-; AVX512BW-NEXT:    vextracti128 $1, %ymm1, %xmm2
-; AVX512BW-NEXT:    vmovdqa {{.*#+}} xmm3 = <u,u,0,8,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512BW-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512BW-NEXT:    vpshufb %xmm3, %xmm1, %xmm1
-; AVX512BW-NEXT:    vpunpcklwd {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1],xmm1[2],xmm2[2],xmm1[3],xmm2[3]
-; AVX512BW-NEXT:    vextracti128 $1, %ymm0, %xmm2
-; AVX512BW-NEXT:    vmovdqa {{.*#+}} xmm3 = <0,8,u,u,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512BW-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512BW-NEXT:    vpshufb %xmm3, %xmm0, %xmm0
-; AVX512BW-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3]
-; AVX512BW-NEXT:    vpblendd {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2,3]
-; AVX512BW-NEXT:    vzeroupper
-; AVX512BW-NEXT:    retq
-;
-; AVX512BWVL-LABEL: trunc_usat_v8i64_v8i8:
-; AVX512BWVL:       # %bb.0:
-; AVX512BWVL-NEXT:    vpminuq {{.*}}(%rip){1to8}, %zmm0, %zmm0
-; AVX512BWVL-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
-; AVX512BWVL-NEXT:    vextracti128 $1, %ymm1, %xmm2
-; AVX512BWVL-NEXT:    vmovdqa {{.*#+}} xmm3 = <u,u,0,8,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512BWVL-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512BWVL-NEXT:    vpshufb %xmm3, %xmm1, %xmm1
-; AVX512BWVL-NEXT:    vpunpcklwd {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1],xmm1[2],xmm2[2],xmm1[3],xmm2[3]
-; AVX512BWVL-NEXT:    vextracti128 $1, %ymm0, %xmm2
-; AVX512BWVL-NEXT:    vmovdqa {{.*#+}} xmm3 = <0,8,u,u,u,u,u,u,u,u,u,u,u,u,u,u>
-; AVX512BWVL-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
-; AVX512BWVL-NEXT:    vpshufb %xmm3, %xmm0, %xmm0
-; AVX512BWVL-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3]
-; AVX512BWVL-NEXT:    vpblendd {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2,3]
-; AVX512BWVL-NEXT:    vzeroupper
-; AVX512BWVL-NEXT:    retq
+; AVX512-LABEL: trunc_usat_v8i64_v8i8:
+; AVX512:       # %bb.0:
+; AVX512-NEXT:    vpminuq {{.*}}(%rip){1to8}, %zmm0, %zmm0
+; AVX512-NEXT:    vpmovqb %zmm0, %xmm0
+; AVX512-NEXT:    vzeroupper
+; AVX512-NEXT:    retq
   %1 = icmp ult <8 x i64> %a0, <i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255>
   %2 = select <8 x i1> %1, <8 x i64> %a0, <8 x i64> <i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255>
   %3 = trunc <8 x i64> %2 to <8 x i8>
@@ -2110,7 +2031,7 @@ define <8 x i8> @trunc_usat_v8i32_v8i8(<8 x i32> %a0) {
 ; AVX512VL-LABEL: trunc_usat_v8i32_v8i8:
 ; AVX512VL:       # %bb.0:
 ; AVX512VL-NEXT:    vpminud {{.*}}(%rip){1to8}, %ymm0, %ymm0
-; AVX512VL-NEXT:    vpmovdb %zmm0, %xmm0
+; AVX512VL-NEXT:    vpmovdb %ymm0, %xmm0
 ; AVX512VL-NEXT:    vzeroupper
 ; AVX512VL-NEXT:    retq
 ;
@@ -2125,7 +2046,7 @@ define <8 x i8> @trunc_usat_v8i32_v8i8(<8 x i32> %a0) {
 ; AVX512BWVL-LABEL: trunc_usat_v8i32_v8i8:
 ; AVX512BWVL:       # %bb.0:
 ; AVX512BWVL-NEXT:    vpminud {{.*}}(%rip){1to8}, %ymm0, %ymm0
-; AVX512BWVL-NEXT:    vpmovdb %zmm0, %xmm0
+; AVX512BWVL-NEXT:    vpmovdb %ymm0, %xmm0
 ; AVX512BWVL-NEXT:    vzeroupper
 ; AVX512BWVL-NEXT:    retq
   %1 = icmp ult <8 x i32> %a0, <i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255>
diff --git a/llvm/test/CodeGen/X86/vector-trunc-widen.ll b/llvm/test/CodeGen/X86/vector-trunc-widen.ll
index e05bdf753737..a15b2aa6779e 100644
--- a/llvm/test/CodeGen/X86/vector-trunc-widen.ll
+++ b/llvm/test/CodeGen/X86/vector-trunc-widen.ll
@@ -1562,15 +1562,41 @@ define <8 x i16> @trunc2x4i64_8i16(<4 x i64> %a, <4 x i64> %b) {
 ; AVX2-FAST-NEXT:    vzeroupper
 ; AVX2-FAST-NEXT:    retq
 ;
-; AVX512-LABEL: trunc2x4i64_8i16:
-; AVX512:       # %bb.0: # %entry
-; AVX512-NEXT:    # kill: def $ymm1 killed $ymm1 def $zmm1
-; AVX512-NEXT:    # kill: def $ymm0 killed $ymm0 def $zmm0
-; AVX512-NEXT:    vpmovqw %zmm0, %xmm0
-; AVX512-NEXT:    vpmovqw %zmm1, %xmm1
-; AVX512-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; AVX512-NEXT:    vzeroupper
-; AVX512-NEXT:    retq
+; AVX512F-LABEL: trunc2x4i64_8i16:
+; AVX512F:       # %bb.0: # %entry
+; AVX512F-NEXT:    # kill: def $ymm1 killed $ymm1 def $zmm1
+; AVX512F-NEXT:    # kill: def $ymm0 killed $ymm0 def $zmm0
+; AVX512F-NEXT:    vpmovqw %zmm0, %xmm0
+; AVX512F-NEXT:    vpmovqw %zmm1, %xmm1
+; AVX512F-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; AVX512F-NEXT:    vzeroupper
+; AVX512F-NEXT:    retq
+;
+; AVX512VL-LABEL: trunc2x4i64_8i16:
+; AVX512VL:       # %bb.0: # %entry
+; AVX512VL-NEXT:    vpmovqw %ymm0, %xmm0
+; AVX512VL-NEXT:    vpmovqw %ymm1, %xmm1
+; AVX512VL-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; AVX512VL-NEXT:    vzeroupper
+; AVX512VL-NEXT:    retq
+;
+; AVX512BW-LABEL: trunc2x4i64_8i16:
+; AVX512BW:       # %bb.0: # %entry
+; AVX512BW-NEXT:    # kill: def $ymm1 killed $ymm1 def $zmm1
+; AVX512BW-NEXT:    # kill: def $ymm0 killed $ymm0 def $zmm0
+; AVX512BW-NEXT:    vpmovqw %zmm0, %xmm0
+; AVX512BW-NEXT:    vpmovqw %zmm1, %xmm1
+; AVX512BW-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; AVX512BW-NEXT:    vzeroupper
+; AVX512BW-NEXT:    retq
+;
+; AVX512BWVL-LABEL: trunc2x4i64_8i16:
+; AVX512BWVL:       # %bb.0: # %entry
+; AVX512BWVL-NEXT:    vpmovqw %ymm0, %xmm0
+; AVX512BWVL-NEXT:    vpmovqw %ymm1, %xmm1
+; AVX512BWVL-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; AVX512BWVL-NEXT:    vzeroupper
+; AVX512BWVL-NEXT:    retq
 entry:
   %0 = trunc <4 x i64> %a to <4 x i16>
   %1 = trunc <4 x i64> %b to <4 x i16>
@@ -1584,71 +1610,15 @@ define <4 x i32> @trunc2x2i64_4i32(<2 x i64> %a, <2 x i64> %b) {
 ; SSE-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[0,2]
 ; SSE-NEXT:    retq
 ;
-; AVX1-LABEL: trunc2x2i64_4i32:
-; AVX1:       # %bb.0: # %entry
-; AVX1-NEXT:    vshufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[0,2]
-; AVX1-NEXT:    retq
+; AVX-LABEL: trunc2x2i64_4i32:
+; AVX:       # %bb.0: # %entry
+; AVX-NEXT:    vshufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[0,2]
+; AVX-NEXT:    retq
 ;
-; AVX2-SLOW-LABEL: trunc2x2i64_4i32:
-; AVX2-SLOW:       # %bb.0: # %entry
-; AVX2-SLOW-NEXT:    # kill: def $xmm1 killed $xmm1 def $ymm1
-; AVX2-SLOW-NEXT:    # kill: def $xmm0 killed $xmm0 def $ymm0
-; AVX2-SLOW-NEXT:    vpermilps {{.*#+}} ymm0 = ymm0[0,2,2,3,4,6,6,7]
-; AVX2-SLOW-NEXT:    vpermilps {{.*#+}} ymm1 = ymm1[0,2,2,3,4,6,6,7]
-; AVX2-SLOW-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; AVX2-SLOW-NEXT:    vzeroupper
-; AVX2-SLOW-NEXT:    retq
-;
-; AVX2-FAST-LABEL: trunc2x2i64_4i32:
-; AVX2-FAST:       # %bb.0: # %entry
-; AVX2-FAST-NEXT:    # kill: def $xmm1 killed $xmm1 def $ymm1
-; AVX2-FAST-NEXT:    # kill: def $xmm0 killed $xmm0 def $ymm0
-; AVX2-FAST-NEXT:    vmovaps {{.*#+}} ymm2 = [0,2,4,6,4,6,6,7]
-; AVX2-FAST-NEXT:    vpermps %ymm0, %ymm2, %ymm0
-; AVX2-FAST-NEXT:    vpermps %ymm1, %ymm2, %ymm1
-; AVX2-FAST-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; AVX2-FAST-NEXT:    vzeroupper
-; AVX2-FAST-NEXT:    retq
-;
-; AVX512F-LABEL: trunc2x2i64_4i32:
-; AVX512F:       # %bb.0: # %entry
-; AVX512F-NEXT:    # kill: def $xmm1 killed $xmm1 def $zmm1
-; AVX512F-NEXT:    # kill: def $xmm0 killed $xmm0 def $zmm0
-; AVX512F-NEXT:    vpmovqd %zmm0, %ymm0
-; AVX512F-NEXT:    vpmovqd %zmm1, %ymm1
-; AVX512F-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; AVX512F-NEXT:    vzeroupper
-; AVX512F-NEXT:    retq
-;
-; AVX512VL-LABEL: trunc2x2i64_4i32:
-; AVX512VL:       # %bb.0: # %entry
-; AVX512VL-NEXT:    # kill: def $xmm1 killed $xmm1 def $ymm1
-; AVX512VL-NEXT:    # kill: def $xmm0 killed $xmm0 def $ymm0
-; AVX512VL-NEXT:    vpmovqd %ymm0, %xmm0
-; AVX512VL-NEXT:    vpmovqd %ymm1, %xmm1
-; AVX512VL-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; AVX512VL-NEXT:    vzeroupper
-; AVX512VL-NEXT:    retq
-;
-; AVX512BW-LABEL: trunc2x2i64_4i32:
-; AVX512BW:       # %bb.0: # %entry
-; AVX512BW-NEXT:    # kill: def $xmm1 killed $xmm1 def $zmm1
-; AVX512BW-NEXT:    # kill: def $xmm0 killed $xmm0 def $zmm0
-; AVX512BW-NEXT:    vpmovqd %zmm0, %ymm0
-; AVX512BW-NEXT:    vpmovqd %zmm1, %ymm1
-; AVX512BW-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; AVX512BW-NEXT:    vzeroupper
-; AVX512BW-NEXT:    retq
-;
-; AVX512BWVL-LABEL: trunc2x2i64_4i32:
-; AVX512BWVL:       # %bb.0: # %entry
-; AVX512BWVL-NEXT:    # kill: def $xmm1 killed $xmm1 def $ymm1
-; AVX512BWVL-NEXT:    # kill: def $xmm0 killed $xmm0 def $ymm0
-; AVX512BWVL-NEXT:    vpmovqd %ymm0, %xmm0
-; AVX512BWVL-NEXT:    vpmovqd %ymm1, %xmm1
-; AVX512BWVL-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; AVX512BWVL-NEXT:    vzeroupper
-; AVX512BWVL-NEXT:    retq
+; AVX512-LABEL: trunc2x2i64_4i32:
+; AVX512:       # %bb.0: # %entry
+; AVX512-NEXT:    vshufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[0,2]
+; AVX512-NEXT:    retq
 entry:
   %0 = trunc <2 x i64> %a to <2 x i32>
   %1 = trunc <2 x i64> %b to <2 x i32>
@@ -1663,60 +1633,17 @@ define i64 @trunc2i64_i64(<2 x i64> %inval) {
 ; SSE-NEXT:    movq %xmm0, %rax
 ; SSE-NEXT:    retq
 ;
-; AVX1-LABEL: trunc2i64_i64:
-; AVX1:       # %bb.0: # %entry
-; AVX1-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
-; AVX1-NEXT:    vmovq %xmm0, %rax
-; AVX1-NEXT:    retq
+; AVX-LABEL: trunc2i64_i64:
+; AVX:       # %bb.0: # %entry
+; AVX-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
+; AVX-NEXT:    vmovq %xmm0, %rax
+; AVX-NEXT:    retq
 ;
-; AVX2-SLOW-LABEL: trunc2i64_i64:
-; AVX2-SLOW:       # %bb.0: # %entry
-; AVX2-SLOW-NEXT:    # kill: def $xmm0 killed $xmm0 def $ymm0
-; AVX2-SLOW-NEXT:    vpshufd {{.*#+}} ymm0 = ymm0[0,2,2,3,4,6,6,7]
-; AVX2-SLOW-NEXT:    vmovq %xmm0, %rax
-; AVX2-SLOW-NEXT:    vzeroupper
-; AVX2-SLOW-NEXT:    retq
-;
-; AVX2-FAST-LABEL: trunc2i64_i64:
-; AVX2-FAST:       # %bb.0: # %entry
-; AVX2-FAST-NEXT:    # kill: def $xmm0 killed $xmm0 def $ymm0
-; AVX2-FAST-NEXT:    vmovdqa {{.*#+}} ymm1 = [0,2,4,6,4,6,6,7]
-; AVX2-FAST-NEXT:    vpermd %ymm0, %ymm1, %ymm0
-; AVX2-FAST-NEXT:    vmovq %xmm0, %rax
-; AVX2-FAST-NEXT:    vzeroupper
-; AVX2-FAST-NEXT:    retq
-;
-; AVX512F-LABEL: trunc2i64_i64:
-; AVX512F:       # %bb.0: # %entry
-; AVX512F-NEXT:    # kill: def $xmm0 killed $xmm0 def $zmm0
-; AVX512F-NEXT:    vpmovqd %zmm0, %ymm0
-; AVX512F-NEXT:    vmovq %xmm0, %rax
-; AVX512F-NEXT:    vzeroupper
-; AVX512F-NEXT:    retq
-;
-; AVX512VL-LABEL: trunc2i64_i64:
-; AVX512VL:       # %bb.0: # %entry
-; AVX512VL-NEXT:    # kill: def $xmm0 killed $xmm0 def $ymm0
-; AVX512VL-NEXT:    vpmovqd %ymm0, %xmm0
-; AVX512VL-NEXT:    vmovq %xmm0, %rax
-; AVX512VL-NEXT:    vzeroupper
-; AVX512VL-NEXT:    retq
-;
-; AVX512BW-LABEL: trunc2i64_i64:
-; AVX512BW:       # %bb.0: # %entry
-; AVX512BW-NEXT:    # kill: def $xmm0 killed $xmm0 def $zmm0
-; AVX512BW-NEXT:    vpmovqd %zmm0, %ymm0
-; AVX512BW-NEXT:    vmovq %xmm0, %rax
-; AVX512BW-NEXT:    vzeroupper
-; AVX512BW-NEXT:    retq
-;
-; AVX512BWVL-LABEL: trunc2i64_i64:
-; AVX512BWVL:       # %bb.0: # %entry
-; AVX512BWVL-NEXT:    # kill: def $xmm0 killed $xmm0 def $ymm0
-; AVX512BWVL-NEXT:    vpmovqd %ymm0, %xmm0
-; AVX512BWVL-NEXT:    vmovq %xmm0, %rax
-; AVX512BWVL-NEXT:    vzeroupper
-; AVX512BWVL-NEXT:    retq
+; AVX512-LABEL: trunc2i64_i64:
+; AVX512:       # %bb.0: # %entry
+; AVX512-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
+; AVX512-NEXT:    vmovq %xmm0, %rax
+; AVX512-NEXT:    retq
 entry:
   %0 = trunc <2 x i64> %inval to <2 x i32>
   %1 = bitcast <2 x i32> %0 to i64
@@ -1751,61 +1678,21 @@ define <8 x i16> @trunc2x4i32_8i16(<4 x i32> %a, <4 x i32> %b) {
 ; SSE41-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
 ; SSE41-NEXT:    retq
 ;
-; AVX1-LABEL: trunc2x4i32_8i16:
-; AVX1:       # %bb.0: # %entry
-; AVX1-NEXT:    vmovdqa {{.*#+}} xmm2 = [0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15]
-; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
-; AVX1-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
-; AVX1-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; AVX1-NEXT:    retq
+; AVX-LABEL: trunc2x4i32_8i16:
+; AVX:       # %bb.0: # %entry
+; AVX-NEXT:    vmovdqa {{.*#+}} xmm2 = [0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15]
+; AVX-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; AVX-NEXT:    retq
 ;
-; AVX2-LABEL: trunc2x4i32_8i16:
-; AVX2:       # %bb.0: # %entry
-; AVX2-NEXT:    vmovdqa {{.*#+}} xmm2 = [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:    vpshufb %xmm2, %xmm0, %xmm0
-; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
-; AVX2-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; AVX2-NEXT:    retq
-;
-; AVX512F-LABEL: trunc2x4i32_8i16:
-; AVX512F:       # %bb.0: # %entry
-; AVX512F-NEXT:    # kill: def $xmm1 killed $xmm1 def $zmm1
-; AVX512F-NEXT:    # kill: def $xmm0 killed $xmm0 def $zmm0
-; AVX512F-NEXT:    vpmovdw %zmm0, %ymm0
-; AVX512F-NEXT:    vpmovdw %zmm1, %ymm1
-; AVX512F-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; AVX512F-NEXT:    vzeroupper
-; AVX512F-NEXT:    retq
-;
-; AVX512VL-LABEL: trunc2x4i32_8i16:
-; AVX512VL:       # %bb.0: # %entry
-; AVX512VL-NEXT:    # kill: def $xmm1 killed $xmm1 def $ymm1
-; AVX512VL-NEXT:    # kill: def $xmm0 killed $xmm0 def $ymm0
-; AVX512VL-NEXT:    vpmovdw %ymm0, %xmm0
-; AVX512VL-NEXT:    vpmovdw %ymm1, %xmm1
-; AVX512VL-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; AVX512VL-NEXT:    vzeroupper
-; AVX512VL-NEXT:    retq
-;
-; AVX512BW-LABEL: trunc2x4i32_8i16:
-; AVX512BW:       # %bb.0: # %entry
-; AVX512BW-NEXT:    # kill: def $xmm1 killed $xmm1 def $zmm1
-; AVX512BW-NEXT:    # kill: def $xmm0 killed $xmm0 def $zmm0
-; AVX512BW-NEXT:    vpmovdw %zmm0, %ymm0
-; AVX512BW-NEXT:    vpmovdw %zmm1, %ymm1
-; AVX512BW-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; AVX512BW-NEXT:    vzeroupper
-; AVX512BW-NEXT:    retq
-;
-; AVX512BWVL-LABEL: trunc2x4i32_8i16:
-; AVX512BWVL:       # %bb.0: # %entry
-; AVX512BWVL-NEXT:    # kill: def $xmm1 killed $xmm1 def $ymm1
-; AVX512BWVL-NEXT:    # kill: def $xmm0 killed $xmm0 def $ymm0
-; AVX512BWVL-NEXT:    vpmovdw %ymm0, %xmm0
-; AVX512BWVL-NEXT:    vpmovdw %ymm1, %xmm1
-; AVX512BWVL-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; AVX512BWVL-NEXT:    vzeroupper
-; AVX512BWVL-NEXT:    retq
+; AVX512-LABEL: trunc2x4i32_8i16:
+; AVX512:       # %bb.0: # %entry
+; AVX512-NEXT:    vmovdqa {{.*#+}} xmm2 = [0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15]
+; AVX512-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX512-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX512-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; AVX512-NEXT:    retq
 entry:
   %0 = trunc <4 x i32> %a to <4 x i16>
   %1 = trunc <4 x i32> %b to <4 x i16>
@@ -1841,37 +1728,11 @@ define i64 @trunc4i32_i64(<4 x i32> %inval) {
 ; AVX-NEXT:    vmovq %xmm0, %rax
 ; AVX-NEXT:    retq
 ;
-; AVX512F-LABEL: trunc4i32_i64:
-; AVX512F:       # %bb.0: # %entry
-; AVX512F-NEXT:    # kill: def $xmm0 killed $xmm0 def $zmm0
-; AVX512F-NEXT:    vpmovdw %zmm0, %ymm0
-; AVX512F-NEXT:    vmovq %xmm0, %rax
-; AVX512F-NEXT:    vzeroupper
-; AVX512F-NEXT:    retq
-;
-; AVX512VL-LABEL: trunc4i32_i64:
-; AVX512VL:       # %bb.0: # %entry
-; AVX512VL-NEXT:    # kill: def $xmm0 killed $xmm0 def $ymm0
-; AVX512VL-NEXT:    vpmovdw %ymm0, %xmm0
-; AVX512VL-NEXT:    vmovq %xmm0, %rax
-; AVX512VL-NEXT:    vzeroupper
-; AVX512VL-NEXT:    retq
-;
-; AVX512BW-LABEL: trunc4i32_i64:
-; AVX512BW:       # %bb.0: # %entry
-; AVX512BW-NEXT:    # kill: def $xmm0 killed $xmm0 def $zmm0
-; AVX512BW-NEXT:    vpmovdw %zmm0, %ymm0
-; AVX512BW-NEXT:    vmovq %xmm0, %rax
-; AVX512BW-NEXT:    vzeroupper
-; AVX512BW-NEXT:    retq
-;
-; AVX512BWVL-LABEL: trunc4i32_i64:
-; AVX512BWVL:       # %bb.0: # %entry
-; AVX512BWVL-NEXT:    # kill: def $xmm0 killed $xmm0 def $ymm0
-; AVX512BWVL-NEXT:    vpmovdw %ymm0, %xmm0
-; AVX512BWVL-NEXT:    vmovq %xmm0, %rax
-; AVX512BWVL-NEXT:    vzeroupper
-; AVX512BWVL-NEXT:    retq
+; AVX512-LABEL: trunc4i32_i64:
+; AVX512:       # %bb.0: # %entry
+; AVX512-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15]
+; AVX512-NEXT:    vmovq %xmm0, %rax
+; AVX512-NEXT:    retq
 entry:
   %0 = trunc <4 x i32> %inval to <4 x i16>
   %1 = bitcast <4 x i16> %0 to i64
@@ -1906,54 +1767,18 @@ define <16 x i8> @trunc2x8i16_16i8(<8 x i16> %a, <8 x i16> %b) {
 ; AVX-LABEL: trunc2x8i16_16i8:
 ; AVX:       # %bb.0: # %entry
 ; AVX-NEXT:    vmovdqa {{.*#+}} xmm2 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
-; AVX-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
 ; AVX-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
 ; AVX-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
 ; AVX-NEXT:    retq
 ;
-; AVX512F-LABEL: trunc2x8i16_16i8:
-; AVX512F:       # %bb.0: # %entry
-; AVX512F-NEXT:    # kill: def $xmm1 killed $xmm1 def $ymm1
-; AVX512F-NEXT:    # kill: def $xmm0 killed $xmm0 def $ymm0
-; AVX512F-NEXT:    vpmovzxwd {{.*#+}} zmm0 = ymm0[0],zero,ymm0[1],zero,ymm0[2],zero,ymm0[3],zero,ymm0[4],zero,ymm0[5],zero,ymm0[6],zero,ymm0[7],zero,ymm0[8],zero,ymm0[9],zero,ymm0[10],zero,ymm0[11],zero,ymm0[12],zero,ymm0[13],zero,ymm0[14],zero,ymm0[15],zero
-; AVX512F-NEXT:    vpmovdb %zmm0, %xmm0
-; AVX512F-NEXT:    vpmovzxwd {{.*#+}} zmm1 = ymm1[0],zero,ymm1[1],zero,ymm1[2],zero,ymm1[3],zero,ymm1[4],zero,ymm1[5],zero,ymm1[6],zero,ymm1[7],zero,ymm1[8],zero,ymm1[9],zero,ymm1[10],zero,ymm1[11],zero,ymm1[12],zero,ymm1[13],zero,ymm1[14],zero,ymm1[15],zero
-; AVX512F-NEXT:    vpmovdb %zmm1, %xmm1
-; AVX512F-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; AVX512F-NEXT:    vzeroupper
-; AVX512F-NEXT:    retq
-;
-; AVX512VL-LABEL: trunc2x8i16_16i8:
-; AVX512VL:       # %bb.0: # %entry
-; AVX512VL-NEXT:    # kill: def $xmm1 killed $xmm1 def $ymm1
-; AVX512VL-NEXT:    # kill: def $xmm0 killed $xmm0 def $ymm0
-; AVX512VL-NEXT:    vpmovzxwd {{.*#+}} zmm0 = ymm0[0],zero,ymm0[1],zero,ymm0[2],zero,ymm0[3],zero,ymm0[4],zero,ymm0[5],zero,ymm0[6],zero,ymm0[7],zero,ymm0[8],zero,ymm0[9],zero,ymm0[10],zero,ymm0[11],zero,ymm0[12],zero,ymm0[13],zero,ymm0[14],zero,ymm0[15],zero
-; AVX512VL-NEXT:    vpmovdb %zmm0, %xmm0
-; AVX512VL-NEXT:    vpmovzxwd {{.*#+}} zmm1 = ymm1[0],zero,ymm1[1],zero,ymm1[2],zero,ymm1[3],zero,ymm1[4],zero,ymm1[5],zero,ymm1[6],zero,ymm1[7],zero,ymm1[8],zero,ymm1[9],zero,ymm1[10],zero,ymm1[11],zero,ymm1[12],zero,ymm1[13],zero,ymm1[14],zero,ymm1[15],zero
-; AVX512VL-NEXT:    vpmovdb %zmm1, %xmm1
-; AVX512VL-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; AVX512VL-NEXT:    vzeroupper
-; AVX512VL-NEXT:    retq
-;
-; AVX512BW-LABEL: trunc2x8i16_16i8:
-; AVX512BW:       # %bb.0: # %entry
-; AVX512BW-NEXT:    # kill: def $xmm1 killed $xmm1 def $zmm1
-; AVX512BW-NEXT:    # kill: def $xmm0 killed $xmm0 def $zmm0
-; AVX512BW-NEXT:    vpmovwb %zmm0, %ymm0
-; AVX512BW-NEXT:    vpmovwb %zmm1, %ymm1
-; AVX512BW-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; AVX512BW-NEXT:    vzeroupper
-; AVX512BW-NEXT:    retq
-;
-; AVX512BWVL-LABEL: trunc2x8i16_16i8:
-; AVX512BWVL:       # %bb.0: # %entry
-; AVX512BWVL-NEXT:    # kill: def $xmm1 killed $xmm1 def $ymm1
-; AVX512BWVL-NEXT:    # kill: def $xmm0 killed $xmm0 def $ymm0
-; AVX512BWVL-NEXT:    vpmovwb %ymm0, %xmm0
-; AVX512BWVL-NEXT:    vpmovwb %ymm1, %xmm1
-; AVX512BWVL-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; AVX512BWVL-NEXT:    vzeroupper
-; AVX512BWVL-NEXT:    retq
+; AVX512-LABEL: trunc2x8i16_16i8:
+; AVX512:       # %bb.0: # %entry
+; AVX512-NEXT:    vmovdqa {{.*#+}} xmm2 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
+; AVX512-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX512-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX512-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; AVX512-NEXT:    retq
 entry:
   %0 = trunc <8 x i16> %a to <8 x i8>
   %1 = trunc <8 x i16> %b to <8 x i8>
@@ -1988,39 +1813,11 @@ define i64 @trunc8i16_i64(<8 x i16> %inval) {
 ; AVX-NEXT:    vmovq %xmm0, %rax
 ; AVX-NEXT:    retq
 ;
-; AVX512F-LABEL: trunc8i16_i64:
-; AVX512F:       # %bb.0: # %entry
-; AVX512F-NEXT:    # kill: def $xmm0 killed $xmm0 def $ymm0
-; AVX512F-NEXT:    vpmovzxwd {{.*#+}} zmm0 = ymm0[0],zero,ymm0[1],zero,ymm0[2],zero,ymm0[3],zero,ymm0[4],zero,ymm0[5],zero,ymm0[6],zero,ymm0[7],zero,ymm0[8],zero,ymm0[9],zero,ymm0[10],zero,ymm0[11],zero,ymm0[12],zero,ymm0[13],zero,ymm0[14],zero,ymm0[15],zero
-; AVX512F-NEXT:    vpmovdb %zmm0, %xmm0
-; AVX512F-NEXT:    vmovq %xmm0, %rax
-; AVX512F-NEXT:    vzeroupper
-; AVX512F-NEXT:    retq
-;
-; AVX512VL-LABEL: trunc8i16_i64:
-; AVX512VL:       # %bb.0: # %entry
-; AVX512VL-NEXT:    # kill: def $xmm0 killed $xmm0 def $ymm0
-; AVX512VL-NEXT:    vpmovzxwd {{.*#+}} zmm0 = ymm0[0],zero,ymm0[1],zero,ymm0[2],zero,ymm0[3],zero,ymm0[4],zero,ymm0[5],zero,ymm0[6],zero,ymm0[7],zero,ymm0[8],zero,ymm0[9],zero,ymm0[10],zero,ymm0[11],zero,ymm0[12],zero,ymm0[13],zero,ymm0[14],zero,ymm0[15],zero
-; AVX512VL-NEXT:    vpmovdb %zmm0, %xmm0
-; AVX512VL-NEXT:    vmovq %xmm0, %rax
-; AVX512VL-NEXT:    vzeroupper
-; AVX512VL-NEXT:    retq
-;
-; AVX512BW-LABEL: trunc8i16_i64:
-; AVX512BW:       # %bb.0: # %entry
-; AVX512BW-NEXT:    # kill: def $xmm0 killed $xmm0 def $zmm0
-; AVX512BW-NEXT:    vpmovwb %zmm0, %ymm0
-; AVX512BW-NEXT:    vmovq %xmm0, %rax
-; AVX512BW-NEXT:    vzeroupper
-; AVX512BW-NEXT:    retq
-;
-; AVX512BWVL-LABEL: trunc8i16_i64:
-; AVX512BWVL:       # %bb.0: # %entry
-; AVX512BWVL-NEXT:    # kill: def $xmm0 killed $xmm0 def $ymm0
-; AVX512BWVL-NEXT:    vpmovwb %ymm0, %xmm0
-; AVX512BWVL-NEXT:    vmovq %xmm0, %rax
-; AVX512BWVL-NEXT:    vzeroupper
-; AVX512BWVL-NEXT:    retq
+; AVX512-LABEL: trunc8i16_i64:
+; AVX512:       # %bb.0: # %entry
+; AVX512-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u]
+; AVX512-NEXT:    vmovq %xmm0, %rax
+; AVX512-NEXT:    retq
 entry:
   %0 = trunc <8 x i16> %inval to <8 x i8>
   %1 = bitcast <8 x i8> %0 to i64