diff --git a/lib/Target/X86/X86InstrAVX512.td b/lib/Target/X86/X86InstrAVX512.td
index 6322d02580f..06c87b9884c 100644
--- a/lib/Target/X86/X86InstrAVX512.td
+++ b/lib/Target/X86/X86InstrAVX512.td
@@ -1183,7 +1183,7 @@ defm VPBROADCASTMB2Q : avx512_mask_broadcast<0x2A, "vpbroadcastmb2q",
 //===----------------------------------------------------------------------===//
 // -- VPERMI2 - 3 source operands form --
 multiclass avx512_perm_i<bits<8> opc, string OpcodeStr, X86VectorVTInfo _> {
-let Constraints = "$src1 = $dst" in {
+let Constraints = "$src1 = $dst", ExeDomain = _.ExeDomain in {
   // The index operand in the pattern should really be an integer type. However,
   // if we do that and it happens to come from a bitcast, then it becomes
   // difficult to find the bitcast needed to convert the index to the
@@ -1205,7 +1205,7 @@ let Constraints = "$src1 = $dst" in {
 }
 multiclass avx512_perm_i_mb<bits<8> opc, string OpcodeStr,
                             X86VectorVTInfo _> {
-  let Constraints = "$src1 = $dst" in
+  let Constraints = "$src1 = $dst", ExeDomain = _.ExeDomain in
   defm rmb: AVX512_maskable_3src<opc, MRMSrcMem, _, (outs _.RC:$dst),
               (ins _.RC:$src2, _.ScalarMemOp:$src3),
               OpcodeStr,   !strconcat("${src3}", _.BroadcastStr,", $src2"),
@@ -1256,7 +1256,7 @@ defm VPERMI2PD : avx512_perm_i_sizes<0x77, "vpermi2pd",
 // VPERMT2
 multiclass avx512_perm_t<bits<8> opc, string OpcodeStr,
                          X86VectorVTInfo _, X86VectorVTInfo IdxVT> {
-let Constraints = "$src1 = $dst" in {
+let Constraints = "$src1 = $dst", ExeDomain = _.ExeDomain in {
   defm rr: AVX512_maskable_3src<opc, MRMSrcReg, _, (outs _.RC:$dst),
           (ins IdxVT.RC:$src2, _.RC:$src3),
           OpcodeStr, "$src3, $src2", "$src2, $src3",
@@ -1273,7 +1273,7 @@ let Constraints = "$src1 = $dst" in {
 }
 multiclass avx512_perm_t_mb<bits<8> opc, string OpcodeStr,
                             X86VectorVTInfo _, X86VectorVTInfo IdxVT> {
-  let Constraints = "$src1 = $dst" in
+  let Constraints = "$src1 = $dst", ExeDomain = _.ExeDomain in
   defm rmb: AVX512_maskable_3src<opc, MRMSrcMem, _, (outs _.RC:$dst),
               (ins IdxVT.RC:$src2, _.ScalarMemOp:$src3),
               OpcodeStr,   !strconcat("${src3}", _.BroadcastStr,", $src2"),
diff --git a/test/CodeGen/X86/avx512-build-vector.ll b/test/CodeGen/X86/avx512-build-vector.ll
index 0b90a18ff3c..618507a08bb 100644
--- a/test/CodeGen/X86/avx512-build-vector.ll
+++ b/test/CodeGen/X86/avx512-build-vector.ll
@@ -15,10 +15,10 @@ define <16 x float> @test3(<4 x float> %a) {
 ; CHECK-LABEL: test3:
 ; CHECK:       ## BB#0:
 ; CHECK-NEXT:    ## kill: %XMM0<def> %XMM0<kill> %ZMM0<def>
-; CHECK-NEXT:    vmovdqa32 {{.*#+}} zmm2 = [0,1,2,3,4,18,16,7,8,9,10,11,12,13,14,15]
+; CHECK-NEXT:    vmovaps {{.*#+}} zmm2 = [0,1,2,3,4,18,16,7,8,9,10,11,12,13,14,15]
 ; CHECK-NEXT:    vpxord %zmm1, %zmm1, %zmm1
 ; CHECK-NEXT:    vpermt2ps %zmm0, %zmm2, %zmm1
-; CHECK-NEXT:    vmovdqa64 %zmm1, %zmm0
+; CHECK-NEXT:    vmovaps %zmm1, %zmm0
 ; CHECK-NEXT:    retq
   %b = extractelement <4 x float> %a, i32 2
   %c = insertelement <16 x float> <float 0.000000e+00, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00, float undef, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00>, float %b, i32 5
diff --git a/test/CodeGen/X86/avx512-intrinsics.ll b/test/CodeGen/X86/avx512-intrinsics.ll
index 3c74aec55ba..fb0af943555 100644
--- a/test/CodeGen/X86/avx512-intrinsics.ll
+++ b/test/CodeGen/X86/avx512-intrinsics.ll
@@ -3272,7 +3272,7 @@ define <8 x double>@test_int_x86_avx512_mask_vpermi2var_pd_512(<8 x double> %x0,
 ; CHECK-LABEL: test_int_x86_avx512_mask_vpermi2var_pd_512:
 ; CHECK:       ## BB#0:
 ; CHECK-NEXT:    kmovw %edi, %k1
-; CHECK-NEXT:    vmovdqa64 %zmm1, %zmm3
+; CHECK-NEXT:    vmovapd %zmm1, %zmm3
 ; CHECK-NEXT:    vpermi2pd %zmm2, %zmm0, %zmm3 {%k1}
 ; CHECK-NEXT:    vpermi2pd %zmm2, %zmm0, %zmm1
 ; CHECK-NEXT:    vaddpd %zmm1, %zmm3, %zmm0
@@ -3289,7 +3289,7 @@ define <16 x float>@test_int_x86_avx512_mask_vpermi2var_ps_512(<16 x float> %x0,
 ; CHECK-LABEL: test_int_x86_avx512_mask_vpermi2var_ps_512:
 ; CHECK:       ## BB#0:
 ; CHECK-NEXT:    kmovw %edi, %k1
-; CHECK-NEXT:    vmovdqa64 %zmm1, %zmm3
+; CHECK-NEXT:    vmovaps %zmm1, %zmm3
 ; CHECK-NEXT:    vpermi2ps %zmm2, %zmm0, %zmm3 {%k1}
 ; CHECK-NEXT:    vpermi2ps %zmm2, %zmm0, %zmm1
 ; CHECK-NEXT:    vaddps %zmm1, %zmm3, %zmm0
@@ -3341,7 +3341,7 @@ define <8 x double>@test_int_x86_avx512_maskz_vpermt2var_pd_512(<8 x i64> %x0, <
 ; CHECK-LABEL: test_int_x86_avx512_maskz_vpermt2var_pd_512:
 ; CHECK:       ## BB#0:
 ; CHECK-NEXT:    kmovw %esi, %k1
-; CHECK-NEXT:    vmovdqa64 %zmm1, %zmm2
+; CHECK-NEXT:    vmovapd %zmm1, %zmm2
 ; CHECK-NEXT:    vpermt2pd (%rdi){1to8}, %zmm0, %zmm2 {%k1} {z}
 ; CHECK-NEXT:    vpermt2pd %zmm1, %zmm0, %zmm1
 ; CHECK-NEXT:    vaddpd %zmm1, %zmm2, %zmm0
@@ -3361,7 +3361,7 @@ define <16 x float>@test_int_x86_avx512_maskz_vpermt2var_ps_512(<16 x i32> %x0,
 ; CHECK-LABEL: test_int_x86_avx512_maskz_vpermt2var_ps_512:
 ; CHECK:       ## BB#0:
 ; CHECK-NEXT:    kmovw %edi, %k1
-; CHECK-NEXT:    vmovdqa64 %zmm1, %zmm3
+; CHECK-NEXT:    vmovaps %zmm1, %zmm3
 ; CHECK-NEXT:    vpermt2ps %zmm2, %zmm0, %zmm3 {%k1} {z}
 ; CHECK-NEXT:    vpermt2ps %zmm2, %zmm0, %zmm1
 ; CHECK-NEXT:    vaddps %zmm1, %zmm3, %zmm0
diff --git a/test/CodeGen/X86/avx512vl-intrinsics.ll b/test/CodeGen/X86/avx512vl-intrinsics.ll
index 5e4e8fd529b..e2448979528 100644
--- a/test/CodeGen/X86/avx512vl-intrinsics.ll
+++ b/test/CodeGen/X86/avx512vl-intrinsics.ll
@@ -2035,7 +2035,7 @@ define <2 x double>@test_int_x86_avx512_mask_vpermi2var_pd_128(<2 x double> %x0,
 ; CHECK-LABEL: test_int_x86_avx512_mask_vpermi2var_pd_128:
 ; CHECK:       ## BB#0:
 ; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
-; CHECK-NEXT:    vmovdqa64 %xmm1, %xmm3 ## encoding: [0x62,0xf1,0xfd,0x08,0x6f,0xd9]
+; CHECK-NEXT:    vmovapd %xmm1, %xmm3 ## encoding: [0x62,0xf1,0xfd,0x08,0x28,0xd9]
 ; CHECK-NEXT:    vpermi2pd %xmm2, %xmm0, %xmm3 {%k1} ## encoding: [0x62,0xf2,0xfd,0x09,0x77,0xda]
 ; CHECK-NEXT:    vpermi2pd %xmm2, %xmm0, %xmm1 ## encoding: [0x62,0xf2,0xfd,0x08,0x77,0xca]
 ; CHECK-NEXT:    vaddpd %xmm1, %xmm3, %xmm0 ## encoding: [0x62,0xf1,0xe5,0x08,0x58,0xc1]
@@ -2052,7 +2052,7 @@ define <4 x double>@test_int_x86_avx512_mask_vpermi2var_pd_256(<4 x double> %x0,
 ; CHECK-LABEL: test_int_x86_avx512_mask_vpermi2var_pd_256:
 ; CHECK:       ## BB#0:
 ; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
-; CHECK-NEXT:    vmovdqa64 %ymm1, %ymm3 ## encoding: [0x62,0xf1,0xfd,0x28,0x6f,0xd9]
+; CHECK-NEXT:    vmovapd %ymm1, %ymm3 ## encoding: [0x62,0xf1,0xfd,0x28,0x28,0xd9]
 ; CHECK-NEXT:    vpermi2pd %ymm2, %ymm0, %ymm3 {%k1} ## encoding: [0x62,0xf2,0xfd,0x29,0x77,0xda]
 ; CHECK-NEXT:    vpermi2pd %ymm2, %ymm0, %ymm1 ## encoding: [0x62,0xf2,0xfd,0x28,0x77,0xca]
 ; CHECK-NEXT:    vaddpd %ymm1, %ymm3, %ymm0 ## encoding: [0x62,0xf1,0xe5,0x28,0x58,0xc1]
@@ -2069,7 +2069,7 @@ define <4 x float>@test_int_x86_avx512_mask_vpermi2var_ps_128(<4 x float> %x0, <
 ; CHECK-LABEL: test_int_x86_avx512_mask_vpermi2var_ps_128:
 ; CHECK:       ## BB#0:
 ; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
-; CHECK-NEXT:    vmovdqa64 %xmm1, %xmm3 ## encoding: [0x62,0xf1,0xfd,0x08,0x6f,0xd9]
+; CHECK-NEXT:    vmovaps %xmm1, %xmm3 ## encoding: [0x62,0xf1,0x7c,0x08,0x28,0xd9]
 ; CHECK-NEXT:    vpermi2ps %xmm2, %xmm0, %xmm3 {%k1} ## encoding: [0x62,0xf2,0x7d,0x09,0x77,0xda]
 ; CHECK-NEXT:    vpermi2ps %xmm2, %xmm0, %xmm1 ## encoding: [0x62,0xf2,0x7d,0x08,0x77,0xca]
 ; CHECK-NEXT:    vaddps %xmm1, %xmm3, %xmm0 ## encoding: [0x62,0xf1,0x64,0x08,0x58,0xc1]
@@ -2085,7 +2085,7 @@ define <4 x float>@test_int_x86_avx512_mask_vpermi2var_ps_128_cast(<4 x float> %
 ; CHECK:       ## BB#0:
 ; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
 ; CHECK-NEXT:    vpermi2ps %xmm2, %xmm0, %xmm1 {%k1} ## encoding: [0x62,0xf2,0x7d,0x09,0x77,0xca]
-; CHECK-NEXT:    vmovdqa64 %xmm1, %xmm0 ## encoding: [0x62,0xf1,0xfd,0x08,0x6f,0xc1]
+; CHECK-NEXT:    vmovaps %xmm1, %xmm0 ## encoding: [0x62,0xf1,0x7c,0x08,0x28,0xc1]
 ; CHECK-NEXT:    retq ## encoding: [0xc3]
   %x1cast = bitcast <2 x i64> %x1 to <4 x i32>
   %res = call <4 x float> @llvm.x86.avx512.mask.vpermi2var.ps.128(<4 x float> %x0, <4 x i32> %x1cast, <4 x float> %x2, i8 %x3)
@@ -2098,7 +2098,7 @@ define <8 x float>@test_int_x86_avx512_mask_vpermi2var_ps_256(<8 x float> %x0, <
 ; CHECK-LABEL: test_int_x86_avx512_mask_vpermi2var_ps_256:
 ; CHECK:       ## BB#0:
 ; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
-; CHECK-NEXT:    vmovdqa64 %ymm1, %ymm3 ## encoding: [0x62,0xf1,0xfd,0x28,0x6f,0xd9]
+; CHECK-NEXT:    vmovaps %ymm1, %ymm3 ## encoding: [0x62,0xf1,0x7c,0x28,0x28,0xd9]
 ; CHECK-NEXT:    vpermi2ps %ymm2, %ymm0, %ymm3 {%k1} ## encoding: [0x62,0xf2,0x7d,0x29,0x77,0xda]
 ; CHECK-NEXT:    vpermi2ps %ymm2, %ymm0, %ymm1 ## encoding: [0x62,0xf2,0x7d,0x28,0x77,0xca]
 ; CHECK-NEXT:    vaddps %ymm1, %ymm3, %ymm0 ## encoding: [0x62,0xf1,0x64,0x28,0x58,0xc1]
diff --git a/test/CodeGen/X86/merge-consecutive-loads-512.ll b/test/CodeGen/X86/merge-consecutive-loads-512.ll
index 6334a8be3ea..dbcd99fbc6d 100644
--- a/test/CodeGen/X86/merge-consecutive-loads-512.ll
+++ b/test/CodeGen/X86/merge-consecutive-loads-512.ll
@@ -138,18 +138,18 @@ define <8 x double> @merge_8f64_f64_12zzuuzz(double* %ptr) nounwind uwtable noin
 define <8 x double> @merge_8f64_f64_1u3u5zu8(double* %ptr) nounwind uwtable noinline ssp {
 ; ALL-LABEL: merge_8f64_f64_1u3u5zu8:
 ; ALL:       # BB#0:
-; ALL-NEXT:    vmovdqu64 8(%rdi), %zmm0
+; ALL-NEXT:    vmovupd 8(%rdi), %zmm0
 ; ALL-NEXT:    vpxord %zmm1, %zmm1, %zmm1
-; ALL-NEXT:    vmovdqa64 {{.*#+}} zmm2 = <0,u,2,u,4,13,u,7>
+; ALL-NEXT:    vmovapd {{.*#+}} zmm2 = <0,u,2,u,4,13,u,7>
 ; ALL-NEXT:    vpermt2pd %zmm1, %zmm2, %zmm0
 ; ALL-NEXT:    retq
 ;
 ; X32-AVX512F-LABEL: merge_8f64_f64_1u3u5zu8:
 ; X32-AVX512F:       # BB#0:
 ; X32-AVX512F-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X32-AVX512F-NEXT:    vmovdqu64 8(%eax), %zmm0
+; X32-AVX512F-NEXT:    vmovupd 8(%eax), %zmm0
 ; X32-AVX512F-NEXT:    vpxord %zmm1, %zmm1, %zmm1
-; X32-AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm2 = <0,0,u,u,2,0,u,u,4,0,13,0,u,u,7,0>
+; X32-AVX512F-NEXT:    vmovapd {{.*#+}} zmm2 = <0,0,u,u,2,0,u,u,4,0,13,0,u,u,7,0>
 ; X32-AVX512F-NEXT:    vpermt2pd %zmm1, %zmm2, %zmm0
 ; X32-AVX512F-NEXT:    retl
   %ptr0 = getelementptr inbounds double, double* %ptr, i64 1
@@ -334,18 +334,18 @@ define <16 x float> @merge_16f32_f32_0uu3uuuuuuuuCuEF(float* %ptr) nounwind uwta
 define <16 x float> @merge_16f32_f32_0uu3zzuuuuuzCuEF(float* %ptr) nounwind uwtable noinline ssp {
 ; ALL-LABEL: merge_16f32_f32_0uu3zzuuuuuzCuEF:
 ; ALL:       # BB#0:
-; ALL-NEXT:    vmovdqu64 (%rdi), %zmm0
+; ALL-NEXT:    vmovups (%rdi), %zmm0
 ; ALL-NEXT:    vpxord %zmm1, %zmm1, %zmm1
-; ALL-NEXT:    vmovdqa32 {{.*#+}} zmm2 = <0,u,u,3,20,21,u,u,u,u,u,u,12,29,14,15>
+; ALL-NEXT:    vmovaps {{.*#+}} zmm2 = <0,u,u,3,20,21,u,u,u,u,u,u,12,29,14,15>
 ; ALL-NEXT:    vpermt2ps %zmm1, %zmm2, %zmm0
 ; ALL-NEXT:    retq
 ;
 ; X32-AVX512F-LABEL: merge_16f32_f32_0uu3zzuuuuuzCuEF:
 ; X32-AVX512F:       # BB#0:
 ; X32-AVX512F-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X32-AVX512F-NEXT:    vmovdqu64 (%eax), %zmm0
+; X32-AVX512F-NEXT:    vmovups (%eax), %zmm0
 ; X32-AVX512F-NEXT:    vpxord %zmm1, %zmm1, %zmm1
-; X32-AVX512F-NEXT:    vmovdqa32 {{.*#+}} zmm2 = <0,u,u,3,20,21,u,u,u,u,u,u,12,29,14,15>
+; X32-AVX512F-NEXT:    vmovaps {{.*#+}} zmm2 = <0,u,u,3,20,21,u,u,u,u,u,u,12,29,14,15>
 ; X32-AVX512F-NEXT:    vpermt2ps %zmm1, %zmm2, %zmm0
 ; X32-AVX512F-NEXT:    retl
   %ptr0 = getelementptr inbounds float, float* %ptr, i64 0
diff --git a/test/CodeGen/X86/vector-shuffle-512-v16.ll b/test/CodeGen/X86/vector-shuffle-512-v16.ll
index 39932cb073f..697662ac2a7 100644
--- a/test/CodeGen/X86/vector-shuffle-512-v16.ll
+++ b/test/CodeGen/X86/vector-shuffle-512-v16.ll
@@ -220,7 +220,7 @@ define <16 x i32> @shuffle_v16i32_0f_1f_0e_16_0d_1d_04_1e_0b_1b_0a_1a_09_19_08_1
 define <16 x float> @shuffle_v16f32_0f_1f_0e_16_0d_1d_04_1e_0b_1b_0a_1a_09_19_08_18(<16 x float> %a, <16 x float> %b)  {
 ; ALL-LABEL: shuffle_v16f32_0f_1f_0e_16_0d_1d_04_1e_0b_1b_0a_1a_09_19_08_18:
 ; ALL:       # BB#0:
-; ALL-NEXT:    vmovdqa32 {{.*#+}} zmm2 = [15,31,14,22,13,29,4,28,11,27,10,26,9,25,8,24]
+; ALL-NEXT:    vmovaps {{.*#+}} zmm2 = [15,31,14,22,13,29,4,28,11,27,10,26,9,25,8,24]
 ; ALL-NEXT:    vpermt2ps %zmm1, %zmm2, %zmm0
 ; ALL-NEXT:    retq
   %c = shufflevector <16 x float> %a, <16 x float> %b, <16 x i32> <i32 15, i32 31, i32 14, i32 22, i32 13, i32 29, i32 4, i32 28, i32 11, i32 27, i32 10, i32 26, i32 9, i32 25, i32 8, i32 24>
@@ -230,7 +230,7 @@ define <16 x float> @shuffle_v16f32_0f_1f_0e_16_0d_1d_04_1e_0b_1b_0a_1a_09_19_08
 define <16 x float> @shuffle_v16f32_load_0f_1f_0e_16_0d_1d_04_1e_0b_1b_0a_1a_09_19_08_18(<16 x float> %a, <16 x float>* %b)  {
 ; ALL-LABEL: shuffle_v16f32_load_0f_1f_0e_16_0d_1d_04_1e_0b_1b_0a_1a_09_19_08_18:
 ; ALL:       # BB#0:
-; ALL-NEXT:    vmovdqa32 {{.*#+}} zmm1 = [15,31,14,22,13,29,4,28,11,27,10,26,9,25,8,24]
+; ALL-NEXT:    vmovaps {{.*#+}} zmm1 = [15,31,14,22,13,29,4,28,11,27,10,26,9,25,8,24]
 ; ALL-NEXT:    vpermt2ps (%rdi), %zmm1, %zmm0
 ; ALL-NEXT:    retq
   %c = load <16 x float>, <16 x float>* %b
diff --git a/test/CodeGen/X86/vector-shuffle-512-v8.ll b/test/CodeGen/X86/vector-shuffle-512-v8.ll
index fbe414bbc3f..ae2e93f55dc 100644
--- a/test/CodeGen/X86/vector-shuffle-512-v8.ll
+++ b/test/CodeGen/X86/vector-shuffle-512-v8.ll
@@ -227,13 +227,13 @@ define <8 x double> @shuffle_v8f64_08080808(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_08080808:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [0,8,0,8,0,8,0,8]
+; AVX512F-NEXT:    vmovapd {{.*#+}} zmm2 = [0,8,0,8,0,8,0,8]
 ; AVX512F-NEXT:    vpermt2pd %zmm1, %zmm2, %zmm0
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_08080808:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [0,0,8,0,0,0,8,0,0,0,8,0,0,0,8,0]
+; AVX512F-32-NEXT:    vmovapd {{.*#+}} zmm2 = [0,0,8,0,0,0,8,0,0,0,8,0,0,0,8,0]
 ; AVX512F-32-NEXT:    vpermt2pd %zmm1, %zmm2, %zmm0
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 0, i32 8, i32 0, i32 8, i32 0, i32 8, i32 0, i32 8>
@@ -244,13 +244,13 @@ define <8 x double> @shuffle_v8f64_08084c4c(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_08084c4c:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [0,8,0,8,4,12,4,12]
+; AVX512F-NEXT:    vmovapd {{.*#+}} zmm2 = [0,8,0,8,4,12,4,12]
 ; AVX512F-NEXT:    vpermt2pd %zmm1, %zmm2, %zmm0
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_08084c4c:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [0,0,8,0,0,0,8,0,4,0,12,0,4,0,12,0]
+; AVX512F-32-NEXT:    vmovapd {{.*#+}} zmm2 = [0,0,8,0,0,0,8,0,4,0,12,0,4,0,12,0]
 ; AVX512F-32-NEXT:    vpermt2pd %zmm1, %zmm2, %zmm0
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 0, i32 8, i32 0, i32 8, i32 4, i32 12, i32 4, i32 12>
@@ -261,16 +261,16 @@ define <8 x double> @shuffle_v8f64_8823cc67(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_8823cc67:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [0,0,10,11,4,4,14,15]
+; AVX512F-NEXT:    vmovapd {{.*#+}} zmm2 = [0,0,10,11,4,4,14,15]
 ; AVX512F-NEXT:    vpermt2pd %zmm0, %zmm2, %zmm1
-; AVX512F-NEXT:    vmovdqa64 %zmm1, %zmm0
+; AVX512F-NEXT:    vmovapd %zmm1, %zmm0
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_8823cc67:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [0,0,0,0,10,0,11,0,4,0,4,0,14,0,15,0]
+; AVX512F-32-NEXT:    vmovapd {{.*#+}} zmm2 = [0,0,0,0,10,0,11,0,4,0,4,0,14,0,15,0]
 ; AVX512F-32-NEXT:    vpermt2pd %zmm0, %zmm2, %zmm1
-; AVX512F-32-NEXT:    vmovdqa64 %zmm1, %zmm0
+; AVX512F-32-NEXT:    vmovapd %zmm1, %zmm0
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 8, i32 8, i32 2, i32 3, i32 12, i32 12, i32 6, i32 7>
   ret <8 x double> %shuffle
@@ -280,16 +280,16 @@ define <8 x double> @shuffle_v8f64_9832dc76(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_9832dc76:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [1,0,11,10,5,4,15,14]
+; AVX512F-NEXT:    vmovapd {{.*#+}} zmm2 = [1,0,11,10,5,4,15,14]
 ; AVX512F-NEXT:    vpermt2pd %zmm0, %zmm2, %zmm1
-; AVX512F-NEXT:    vmovdqa64 %zmm1, %zmm0
+; AVX512F-NEXT:    vmovapd %zmm1, %zmm0
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_9832dc76:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [1,0,0,0,11,0,10,0,5,0,4,0,15,0,14,0]
+; AVX512F-32-NEXT:    vmovapd {{.*#+}} zmm2 = [1,0,0,0,11,0,10,0,5,0,4,0,15,0,14,0]
 ; AVX512F-32-NEXT:    vpermt2pd %zmm0, %zmm2, %zmm1
-; AVX512F-32-NEXT:    vmovdqa64 %zmm1, %zmm0
+; AVX512F-32-NEXT:    vmovapd %zmm1, %zmm0
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 9, i32 8, i32 3, i32 2, i32 13, i32 12, i32 7, i32 6>
   ret <8 x double> %shuffle
@@ -299,16 +299,16 @@ define <8 x double> @shuffle_v8f64_9810dc54(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_9810dc54:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [1,0,9,8,5,4,13,12]
+; AVX512F-NEXT:    vmovapd {{.*#+}} zmm2 = [1,0,9,8,5,4,13,12]
 ; AVX512F-NEXT:    vpermt2pd %zmm0, %zmm2, %zmm1
-; AVX512F-NEXT:    vmovdqa64 %zmm1, %zmm0
+; AVX512F-NEXT:    vmovapd %zmm1, %zmm0
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_9810dc54:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [1,0,0,0,9,0,8,0,5,0,4,0,13,0,12,0]
+; AVX512F-32-NEXT:    vmovapd {{.*#+}} zmm2 = [1,0,0,0,9,0,8,0,5,0,4,0,13,0,12,0]
 ; AVX512F-32-NEXT:    vpermt2pd %zmm0, %zmm2, %zmm1
-; AVX512F-32-NEXT:    vmovdqa64 %zmm1, %zmm0
+; AVX512F-32-NEXT:    vmovapd %zmm1, %zmm0
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 9, i32 8, i32 1, i32 0, i32 13, i32 12, i32 5, i32 4>
   ret <8 x double> %shuffle
@@ -318,13 +318,13 @@ define <8 x double> @shuffle_v8f64_08194c5d(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_08194c5d:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [0,8,1,9,4,12,5,13]
+; AVX512F-NEXT:    vmovapd {{.*#+}} zmm2 = [0,8,1,9,4,12,5,13]
 ; AVX512F-NEXT:    vpermt2pd %zmm1, %zmm2, %zmm0
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_08194c5d:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [0,0,8,0,1,0,9,0,4,0,12,0,5,0,13,0]
+; AVX512F-32-NEXT:    vmovapd {{.*#+}} zmm2 = [0,0,8,0,1,0,9,0,4,0,12,0,5,0,13,0]
 ; AVX512F-32-NEXT:    vpermt2pd %zmm1, %zmm2, %zmm0
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 4, i32 12, i32 5, i32 13>
@@ -335,13 +335,13 @@ define <8 x double> @shuffle_v8f64_2a3b6e7f(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_2a3b6e7f:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [2,10,3,11,6,14,7,15]
+; AVX512F-NEXT:    vmovapd {{.*#+}} zmm2 = [2,10,3,11,6,14,7,15]
 ; AVX512F-NEXT:    vpermt2pd %zmm1, %zmm2, %zmm0
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_2a3b6e7f:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [2,0,10,0,3,0,11,0,6,0,14,0,7,0,15,0]
+; AVX512F-32-NEXT:    vmovapd {{.*#+}} zmm2 = [2,0,10,0,3,0,11,0,6,0,14,0,7,0,15,0]
 ; AVX512F-32-NEXT:    vpermt2pd %zmm1, %zmm2, %zmm0
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 2, i32 10, i32 3, i32 11, i32 6, i32 14, i32 7, i32 15>
@@ -352,13 +352,13 @@ define <8 x double> @shuffle_v8f64_08192a3b(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_08192a3b:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [0,8,1,9,2,10,3,11]
+; AVX512F-NEXT:    vmovapd {{.*#+}} zmm2 = [0,8,1,9,2,10,3,11]
 ; AVX512F-NEXT:    vpermt2pd %zmm1, %zmm2, %zmm0
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_08192a3b:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [0,0,8,0,1,0,9,0,2,0,10,0,3,0,11,0]
+; AVX512F-32-NEXT:    vmovapd {{.*#+}} zmm2 = [0,0,8,0,1,0,9,0,2,0,10,0,3,0,11,0]
 ; AVX512F-32-NEXT:    vpermt2pd %zmm1, %zmm2, %zmm0
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11>
@@ -369,16 +369,16 @@ define <8 x double> @shuffle_v8f64_08991abb(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_08991abb:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [8,0,1,1,9,2,3,3]
+; AVX512F-NEXT:    vmovapd {{.*#+}} zmm2 = [8,0,1,1,9,2,3,3]
 ; AVX512F-NEXT:    vpermt2pd %zmm0, %zmm2, %zmm1
-; AVX512F-NEXT:    vmovdqa64 %zmm1, %zmm0
+; AVX512F-NEXT:    vmovapd %zmm1, %zmm0
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_08991abb:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [8,0,0,0,1,0,1,0,9,0,2,0,3,0,3,0]
+; AVX512F-32-NEXT:    vmovapd {{.*#+}} zmm2 = [8,0,0,0,1,0,1,0,9,0,2,0,3,0,3,0]
 ; AVX512F-32-NEXT:    vpermt2pd %zmm0, %zmm2, %zmm1
-; AVX512F-32-NEXT:    vmovdqa64 %zmm1, %zmm0
+; AVX512F-32-NEXT:    vmovapd %zmm1, %zmm0
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 0, i32 8, i32 9, i32 9, i32 1, i32 10, i32 11, i32 11>
   ret <8 x double> %shuffle
@@ -388,13 +388,13 @@ define <8 x double> @shuffle_v8f64_091b2d3f(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_091b2d3f:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [0,9,1,11,2,13,3,15]
+; AVX512F-NEXT:    vmovapd {{.*#+}} zmm2 = [0,9,1,11,2,13,3,15]
 ; AVX512F-NEXT:    vpermt2pd %zmm1, %zmm2, %zmm0
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_091b2d3f:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [0,0,9,0,1,0,11,0,2,0,13,0,3,0,15,0]
+; AVX512F-32-NEXT:    vmovapd {{.*#+}} zmm2 = [0,0,9,0,1,0,11,0,2,0,13,0,3,0,15,0]
 ; AVX512F-32-NEXT:    vpermt2pd %zmm1, %zmm2, %zmm0
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 0, i32 9, i32 1, i32 11, i32 2, i32 13, i32 3, i32 15>
@@ -405,16 +405,16 @@ define <8 x double> @shuffle_v8f64_09ab1def(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_09ab1def:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [8,1,2,3,9,5,6,7]
+; AVX512F-NEXT:    vmovapd {{.*#+}} zmm2 = [8,1,2,3,9,5,6,7]
 ; AVX512F-NEXT:    vpermt2pd %zmm0, %zmm2, %zmm1
-; AVX512F-NEXT:    vmovdqa64 %zmm1, %zmm0
+; AVX512F-NEXT:    vmovapd %zmm1, %zmm0
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_09ab1def:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [8,0,1,0,2,0,3,0,9,0,5,0,6,0,7,0]
+; AVX512F-32-NEXT:    vmovapd {{.*#+}} zmm2 = [8,0,1,0,2,0,3,0,9,0,5,0,6,0,7,0]
 ; AVX512F-32-NEXT:    vpermt2pd %zmm0, %zmm2, %zmm1
-; AVX512F-32-NEXT:    vmovdqa64 %zmm1, %zmm0
+; AVX512F-32-NEXT:    vmovapd %zmm1, %zmm0
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 0, i32 9, i32 10, i32 11, i32 1, i32 13, i32 14, i32 15>
   ret <8 x double> %shuffle
@@ -926,16 +926,16 @@ define <8 x double> @shuffle_v8f64_c348cda0(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_c348cda0:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [4,11,12,0,4,5,2,8]
+; AVX512F-NEXT:    vmovapd {{.*#+}} zmm2 = [4,11,12,0,4,5,2,8]
 ; AVX512F-NEXT:    vpermt2pd %zmm0, %zmm2, %zmm1
-; AVX512F-NEXT:    vmovdqa64 %zmm1, %zmm0
+; AVX512F-NEXT:    vmovapd %zmm1, %zmm0
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_c348cda0:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [4,0,11,0,12,0,0,0,4,0,5,0,2,0,8,0]
+; AVX512F-32-NEXT:    vmovapd {{.*#+}} zmm2 = [4,0,11,0,12,0,0,0,4,0,5,0,2,0,8,0]
 ; AVX512F-32-NEXT:    vpermt2pd %zmm0, %zmm2, %zmm1
-; AVX512F-32-NEXT:    vmovdqa64 %zmm1, %zmm0
+; AVX512F-32-NEXT:    vmovapd %zmm1, %zmm0
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 12, i32 3, i32 4, i32 8, i32 12, i32 13, i32 10, i32 0>
   ret <8 x double> %shuffle
@@ -945,13 +945,13 @@ define <8 x double> @shuffle_v8f64_f511235a(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_f511235a:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [15,5,1,1,2,3,5,10]
+; AVX512F-NEXT:    vmovapd {{.*#+}} zmm2 = [15,5,1,1,2,3,5,10]
 ; AVX512F-NEXT:    vpermt2pd %zmm1, %zmm2, %zmm0
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_f511235a:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [15,0,5,0,1,0,1,0,2,0,3,0,5,0,10,0]
+; AVX512F-32-NEXT:    vmovapd {{.*#+}} zmm2 = [15,0,5,0,1,0,1,0,2,0,3,0,5,0,10,0]
 ; AVX512F-32-NEXT:    vpermt2pd %zmm1, %zmm2, %zmm0
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 15, i32 5, i32 1, i32 1, i32 2, i32 3, i32 5, i32 10>
diff --git a/test/CodeGen/X86/vector-shuffle-combining-avx512bw.ll b/test/CodeGen/X86/vector-shuffle-combining-avx512bw.ll
index 14a870fcbd7..12fb8a2003e 100644
--- a/test/CodeGen/X86/vector-shuffle-combining-avx512bw.ll
+++ b/test/CodeGen/X86/vector-shuffle-combining-avx512bw.ll
@@ -105,9 +105,9 @@ define <8 x i64> @combine_permvar_8i64_identity_mask(<8 x i64> %x0, <8 x i64> %x
 define <8 x double> @combine_vpermt2var_8f64_identity(<8 x double> %x0, <8 x double> %x1) {
 ; X32-LABEL: combine_vpermt2var_8f64_identity:
 ; X32:       # BB#0:
-; X32-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [7,0,6,0,5,0,4,0,3,0,2,0,1,0,0,0]
+; X32-NEXT:    vmovapd {{.*#+}} zmm2 = [7,0,6,0,5,0,4,0,3,0,2,0,1,0,0,0]
 ; X32-NEXT:    vpermt2pd %zmm1, %zmm2, %zmm0
-; X32-NEXT:    vmovdqa64 {{.*#+}} zmm1 = [7,0,14,0,5,0,12,0,3,0,10,0,1,0,8,0]
+; X32-NEXT:    vmovapd {{.*#+}} zmm1 = [7,0,14,0,5,0,12,0,3,0,10,0,1,0,8,0]
 ; X32-NEXT:    vpermt2pd %zmm0, %zmm1, %zmm0
 ; X32-NEXT:    retl
 ;
@@ -123,18 +123,18 @@ define <8 x double> @combine_vpermt2var_8f64_identity_mask(<8 x double> %x0, <8
 ; X32:       # BB#0:
 ; X32-NEXT:    movzbl {{[0-9]+}}(%esp), %eax
 ; X32-NEXT:    kmovd %eax, %k1
-; X32-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [7,0,6,0,5,0,4,0,3,0,2,0,1,0,0,0]
+; X32-NEXT:    vmovapd {{.*#+}} zmm2 = [7,0,6,0,5,0,4,0,3,0,2,0,1,0,0,0]
 ; X32-NEXT:    vpermt2pd %zmm1, %zmm2, %zmm0 {%k1} {z}
-; X32-NEXT:    vmovdqa64 {{.*#+}} zmm1 = [7,0,14,0,5,0,12,0,3,0,10,0,1,0,8,0]
+; X32-NEXT:    vmovapd {{.*#+}} zmm1 = [7,0,14,0,5,0,12,0,3,0,10,0,1,0,8,0]
 ; X32-NEXT:    vpermt2pd %zmm0, %zmm1, %zmm0 {%k1} {z}
 ; X32-NEXT:    retl
 ;
 ; X64-LABEL: combine_vpermt2var_8f64_identity_mask:
 ; X64:       # BB#0:
 ; X64-NEXT:    kmovw %edi, %k1
-; X64-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [7,6,5,4,3,2,1,0]
+; X64-NEXT:    vmovapd {{.*#+}} zmm2 = [7,6,5,4,3,2,1,0]
 ; X64-NEXT:    vpermt2pd %zmm1, %zmm2, %zmm0 {%k1} {z}
-; X64-NEXT:    vmovdqa64 {{.*#+}} zmm1 = [7,14,5,12,3,10,1,8]
+; X64-NEXT:    vmovapd {{.*#+}} zmm1 = [7,14,5,12,3,10,1,8]
 ; X64-NEXT:    vpermt2pd %zmm0, %zmm1, %zmm0 {%k1} {z}
 ; X64-NEXT:    retq
   %res0 = call <8 x double> @llvm.x86.avx512.maskz.vpermt2var.pd.512(<8 x i64> <i64 7, i64 6, i64 5, i64 4, i64 3, i64 2, i64 1, i64 0>, <8 x double> %x0, <8 x double> %x1, i8 %m)
@@ -145,7 +145,7 @@ define <8 x double> @combine_vpermt2var_8f64_identity_mask(<8 x double> %x0, <8
 define <8 x double> @combine_vpermt2var_8f64_movddup(<8 x double> %x0, <8 x double> %x1) {
 ; X32-LABEL: combine_vpermt2var_8f64_movddup:
 ; X32:       # BB#0:
-; X32-NEXT:    vmovdqa64 {{.*#+}} zmm2 = <0,0,0,0,2,0,2,0,4,0,4,0,u,u,u,u>
+; X32-NEXT:    vmovapd {{.*#+}} zmm2 = <0,0,0,0,2,0,2,0,4,0,4,0,u,u,u,u>
 ; X32-NEXT:    vpermt2pd %zmm1, %zmm2, %zmm0
 ; X32-NEXT:    retl
 ;
@@ -160,10 +160,10 @@ define <8 x double> @combine_vpermt2var_8f64_movddup_load(<8 x double> *%p0, <8
 ; X32-LABEL: combine_vpermt2var_8f64_movddup_load:
 ; X32:       # BB#0:
 ; X32-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X32-NEXT:    vmovdqa64 (%eax), %zmm1
-; X32-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [0,0,0,0,2,0,2,0,4,0,4,0,6,0,6,0]
+; X32-NEXT:    vmovapd (%eax), %zmm1
+; X32-NEXT:    vmovapd {{.*#+}} zmm2 = [0,0,0,0,2,0,2,0,4,0,4,0,6,0,6,0]
 ; X32-NEXT:    vpermt2pd %zmm0, %zmm2, %zmm1
-; X32-NEXT:    vmovdqa64 %zmm1, %zmm0
+; X32-NEXT:    vmovapd %zmm1, %zmm0
 ; X32-NEXT:    retl
 ;
 ; X64-LABEL: combine_vpermt2var_8f64_movddup_load:
@@ -179,7 +179,7 @@ define <8 x double> @combine_vpermt2var_8f64_movddup_mask(<8 x double> %x0, <8 x
 ; X32:       # BB#0:
 ; X32-NEXT:    movzbl {{[0-9]+}}(%esp), %eax
 ; X32-NEXT:    kmovd %eax, %k1
-; X32-NEXT:    vmovdqa64 {{.*#+}} zmm2 = [0,0,0,0,2,0,2,0,4,0,4,0,6,0,6,0]
+; X32-NEXT:    vmovapd {{.*#+}} zmm2 = [0,0,0,0,2,0,2,0,4,0,4,0,6,0,6,0]
 ; X32-NEXT:    vpermt2pd %zmm1, %zmm2, %zmm0 {%k1} {z}
 ; X32-NEXT:    retl
 ;
@@ -248,18 +248,18 @@ define <16 x float> @combine_vpermt2var_16f32_identity_mask(<16 x float> %x0, <1
 ; X32-LABEL: combine_vpermt2var_16f32_identity_mask:
 ; X32:       # BB#0:
 ; X32-NEXT:    kmovw {{[0-9]+}}(%esp), %k1
-; X32-NEXT:    vmovdqa32 {{.*#+}} zmm2 = [15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0]
+; X32-NEXT:    vmovaps {{.*#+}} zmm2 = [15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0]
 ; X32-NEXT:    vpermt2ps %zmm1, %zmm2, %zmm0 {%k1} {z}
-; X32-NEXT:    vmovdqa32 {{.*#+}} zmm1 = [15,30,13,28,11,26,9,24,7,22,5,20,3,18,1,16]
+; X32-NEXT:    vmovaps {{.*#+}} zmm1 = [15,30,13,28,11,26,9,24,7,22,5,20,3,18,1,16]
 ; X32-NEXT:    vpermt2ps %zmm0, %zmm1, %zmm0 {%k1} {z}
 ; X32-NEXT:    retl
 ;
 ; X64-LABEL: combine_vpermt2var_16f32_identity_mask:
 ; X64:       # BB#0:
 ; X64-NEXT:    kmovw %edi, %k1
-; X64-NEXT:    vmovdqa32 {{.*#+}} zmm2 = [15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0]
+; X64-NEXT:    vmovaps {{.*#+}} zmm2 = [15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0]
 ; X64-NEXT:    vpermt2ps %zmm1, %zmm2, %zmm0 {%k1} {z}
-; X64-NEXT:    vmovdqa32 {{.*#+}} zmm1 = [15,30,13,28,11,26,9,24,7,22,5,20,3,18,1,16]
+; X64-NEXT:    vmovaps {{.*#+}} zmm1 = [15,30,13,28,11,26,9,24,7,22,5,20,3,18,1,16]
 ; X64-NEXT:    vpermt2ps %zmm0, %zmm1, %zmm0 {%k1} {z}
 ; X64-NEXT:    retq
   %res0 = call <16 x float> @llvm.x86.avx512.maskz.vpermt2var.ps.512(<16 x i32> <i32 15, i32 14, i32 13, i32 12, i32 11, i32 10, i32 9, i32 8, i32 7, i32 6, i32 5, i32 4, i32 3, i32 2, i32 1, i32 0>, <16 x float> %x0, <16 x float> %x1, i16 %m)
@@ -270,13 +270,13 @@ define <16 x float> @combine_vpermt2var_16f32_identity_mask(<16 x float> %x0, <1
 define <16 x float> @combine_vpermt2var_16f32_vmovddup(<16 x float> %x0, <16 x float> %x1) {
 ; X32-LABEL: combine_vpermt2var_16f32_vmovddup:
 ; X32:       # BB#0:
-; X32-NEXT:    vmovdqa32 {{.*#+}} zmm2 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
+; X32-NEXT:    vmovaps {{.*#+}} zmm2 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
 ; X32-NEXT:    vpermt2ps %zmm1, %zmm2, %zmm0
 ; X32-NEXT:    retl
 ;
 ; X64-LABEL: combine_vpermt2var_16f32_vmovddup:
 ; X64:       # BB#0:
-; X64-NEXT:    vmovdqa32 {{.*#+}} zmm2 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
+; X64-NEXT:    vmovaps {{.*#+}} zmm2 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
 ; X64-NEXT:    vpermt2ps %zmm1, %zmm2, %zmm0
 ; X64-NEXT:    retq
   %res0 = call <16 x float> @llvm.x86.avx512.maskz.vpermt2var.ps.512(<16 x i32> <i32 0, i32 1, i32 0, i32 1, i32 4, i32 5, i32 4, i32 5, i32 8, i32 9, i32 8, i32 9, i32 12, i32 13, i32 12, i32 13>, <16 x float> %x0, <16 x float> %x1, i16 -1)
@@ -286,18 +286,18 @@ define <16 x float> @combine_vpermt2var_16f32_vmovddup_load(<16 x float> *%p0, <
 ; X32-LABEL: combine_vpermt2var_16f32_vmovddup_load:
 ; X32:       # BB#0:
 ; X32-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X32-NEXT:    vmovdqa64 (%eax), %zmm1
-; X32-NEXT:    vmovdqa32 {{.*#+}} zmm2 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
+; X32-NEXT:    vmovaps (%eax), %zmm1
+; X32-NEXT:    vmovaps {{.*#+}} zmm2 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
 ; X32-NEXT:    vpermt2ps %zmm0, %zmm2, %zmm1
-; X32-NEXT:    vmovdqa64 %zmm1, %zmm0
+; X32-NEXT:    vmovaps %zmm1, %zmm0
 ; X32-NEXT:    retl
 ;
 ; X64-LABEL: combine_vpermt2var_16f32_vmovddup_load:
 ; X64:       # BB#0:
-; X64-NEXT:    vmovdqa64 (%rdi), %zmm1
-; X64-NEXT:    vmovdqa32 {{.*#+}} zmm2 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
+; X64-NEXT:    vmovaps (%rdi), %zmm1
+; X64-NEXT:    vmovaps {{.*#+}} zmm2 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
 ; X64-NEXT:    vpermt2ps %zmm0, %zmm2, %zmm1
-; X64-NEXT:    vmovdqa64 %zmm1, %zmm0
+; X64-NEXT:    vmovaps %zmm1, %zmm0
 ; X64-NEXT:    retq
   %x0 = load <16 x float>, <16 x float> *%p0
   %res0 = call <16 x float> @llvm.x86.avx512.maskz.vpermt2var.ps.512(<16 x i32> <i32 0, i32 1, i32 0, i32 1, i32 4, i32 5, i32 4, i32 5, i32 8, i32 9, i32 8, i32 9, i32 12, i32 13, i32 12, i32 13>, <16 x float> %x0, <16 x float> %x1, i16 -1)
@@ -307,14 +307,14 @@ define <16 x float> @combine_vpermt2var_16f32_vmovddup_mask(<16 x float> %x0, <1
 ; X32-LABEL: combine_vpermt2var_16f32_vmovddup_mask:
 ; X32:       # BB#0:
 ; X32-NEXT:    kmovw {{[0-9]+}}(%esp), %k1
-; X32-NEXT:    vmovdqa32 {{.*#+}} zmm2 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
+; X32-NEXT:    vmovaps {{.*#+}} zmm2 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
 ; X32-NEXT:    vpermt2ps %zmm1, %zmm2, %zmm0 {%k1} {z}
 ; X32-NEXT:    retl
 ;
 ; X64-LABEL: combine_vpermt2var_16f32_vmovddup_mask:
 ; X64:       # BB#0:
 ; X64-NEXT:    kmovw %edi, %k1
-; X64-NEXT:    vmovdqa32 {{.*#+}} zmm2 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
+; X64-NEXT:    vmovaps {{.*#+}} zmm2 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
 ; X64-NEXT:    vpermt2ps %zmm1, %zmm2, %zmm0 {%k1} {z}
 ; X64-NEXT:    retq
   %res0 = call <16 x float> @llvm.x86.avx512.maskz.vpermt2var.ps.512(<16 x i32> <i32 0, i32 1, i32 0, i32 1, i32 4, i32 5, i32 4, i32 5, i32 8, i32 9, i32 8, i32 9, i32 12, i32 13, i32 12, i32 13>, <16 x float> %x0, <16 x float> %x1, i16 %m)
@@ -325,19 +325,19 @@ define <16 x float> @combine_vpermt2var_16f32_vmovddup_mask_load(<16 x float> *%
 ; X32:       # BB#0:
 ; X32-NEXT:    kmovw {{[0-9]+}}(%esp), %k1
 ; X32-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X32-NEXT:    vmovdqa64 (%eax), %zmm1
-; X32-NEXT:    vmovdqa32 {{.*#+}} zmm2 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
+; X32-NEXT:    vmovaps (%eax), %zmm1
+; X32-NEXT:    vmovaps {{.*#+}} zmm2 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
 ; X32-NEXT:    vpermt2ps %zmm0, %zmm2, %zmm1 {%k1} {z}
-; X32-NEXT:    vmovdqa64 %zmm1, %zmm0
+; X32-NEXT:    vmovaps %zmm1, %zmm0
 ; X32-NEXT:    retl
 ;
 ; X64-LABEL: combine_vpermt2var_16f32_vmovddup_mask_load:
 ; X64:       # BB#0:
 ; X64-NEXT:    kmovw %esi, %k1
-; X64-NEXT:    vmovdqa64 (%rdi), %zmm1
-; X64-NEXT:    vmovdqa32 {{.*#+}} zmm2 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
+; X64-NEXT:    vmovaps (%rdi), %zmm1
+; X64-NEXT:    vmovaps {{.*#+}} zmm2 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
 ; X64-NEXT:    vpermt2ps %zmm0, %zmm2, %zmm1 {%k1} {z}
-; X64-NEXT:    vmovdqa64 %zmm1, %zmm0
+; X64-NEXT:    vmovaps %zmm1, %zmm0
 ; X64-NEXT:    retq
   %x0 = load <16 x float>, <16 x float> *%p0
   %res0 = call <16 x float> @llvm.x86.avx512.maskz.vpermt2var.ps.512(<16 x i32> <i32 0, i32 1, i32 0, i32 1, i32 4, i32 5, i32 4, i32 5, i32 8, i32 9, i32 8, i32 9, i32 12, i32 13, i32 12, i32 13>, <16 x float> %x0, <16 x float> %x1, i16 %m)