[X86][AVX512] Enable v16i8/v32i8 vector shifts to use an extend+shift+truncate pattern.

Use the existing AVX2 v8i16 vector shift lowering for v16i8 (extending to v16i32) on AVX512 targets and v32i8 (extending to v32i16) on AVX512BW targets.

Cost model updates to follow.

llvm-svn: 291451

6 files changed, 279 insertions, 318 deletions

diff --git a/llvm/test/CodeGen/X86/vector-shift-ashr-128.ll b/llvm/test/CodeGen/X86/vector-shift-ashr-128.ll
index d68ccc47816..440faa689fb 100644
--- a/llvm/test/CodeGen/X86/vector-shift-ashr-128.ll
+++ b/llvm/test/CodeGen/X86/vector-shift-ashr-128.ll
@@ -509,30 +509,10 @@ define <16 x i8> @var_shift_v16i8(<16 x i8> %a, <16 x i8> %b) nounwind {
 ;
 ; AVX512-LABEL: var_shift_v16i8:
 ; AVX512:       # BB#0:
-; AVX512-NEXT:    vpsllw $5, %xmm1, %xmm1
-; AVX512-NEXT:    vpunpckhbw {{.*#+}} xmm2 = xmm0[8],xmm1[8],xmm0[9],xmm1[9],xmm0[10],xmm1[10],xmm0[11],xmm1[11],xmm0[12],xmm1[12],xmm0[13],xmm1[13],xmm0[14],xmm1[14],xmm0[15],xmm1[15]
-; AVX512-NEXT:    vpunpckhbw {{.*#+}} xmm3 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
-; AVX512-NEXT:    vpsraw $4, %xmm3, %xmm4
-; AVX512-NEXT:    vpblendvb %xmm2, %xmm4, %xmm3, %xmm3
-; AVX512-NEXT:    vpsraw $2, %xmm3, %xmm4
-; AVX512-NEXT:    vpaddw %xmm2, %xmm2, %xmm2
-; AVX512-NEXT:    vpblendvb %xmm2, %xmm4, %xmm3, %xmm3
-; AVX512-NEXT:    vpsraw $1, %xmm3, %xmm4
-; AVX512-NEXT:    vpaddw %xmm2, %xmm2, %xmm2
-; AVX512-NEXT:    vpblendvb %xmm2, %xmm4, %xmm3, %xmm2
-; AVX512-NEXT:    vpsrlw $8, %xmm2, %xmm2
-; AVX512-NEXT:    vpunpcklbw {{.*#+}} xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3],xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
-; AVX512-NEXT:    vpunpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
-; AVX512-NEXT:    vpsraw $4, %xmm0, %xmm3
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm3, %xmm0, %xmm0
-; AVX512-NEXT:    vpsraw $2, %xmm0, %xmm3
-; AVX512-NEXT:    vpaddw %xmm1, %xmm1, %xmm1
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm3, %xmm0, %xmm0
-; AVX512-NEXT:    vpsraw $1, %xmm0, %xmm3
-; AVX512-NEXT:    vpaddw %xmm1, %xmm1, %xmm1
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm3, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrlw $8, %xmm0, %xmm0
-; AVX512-NEXT:    vpackuswb %xmm2, %xmm0, %xmm0
+; AVX512-NEXT:    vpmovzxbd {{.*#+}} zmm1 = xmm1[0],zero,zero,zero,xmm1[1],zero,zero,zero,xmm1[2],zero,zero,zero,xmm1[3],zero,zero,zero,xmm1[4],zero,zero,zero,xmm1[5],zero,zero,zero,xmm1[6],zero,zero,zero,xmm1[7],zero,zero,zero,xmm1[8],zero,zero,zero,xmm1[9],zero,zero,zero,xmm1[10],zero,zero,zero,xmm1[11],zero,zero,zero,xmm1[12],zero,zero,zero,xmm1[13],zero,zero,zero,xmm1[14],zero,zero,zero,xmm1[15],zero,zero,zero
+; AVX512-NEXT:    vpmovsxbd %xmm0, %zmm0
+; AVX512-NEXT:    vpsravd %zmm1, %zmm0, %zmm0
+; AVX512-NEXT:    vpmovdb %zmm0, %xmm0
 ; AVX512-NEXT:    retq
 ;
 ; X32-SSE-LABEL: var_shift_v16i8:
@@ -921,30 +901,10 @@ define <16 x i8> @splatvar_shift_v16i8(<16 x i8> %a, <16 x i8> %b) nounwind {
 ; AVX512-LABEL: splatvar_shift_v16i8:
 ; AVX512:       # BB#0:
 ; AVX512-NEXT:    vpbroadcastb %xmm1, %xmm1
-; AVX512-NEXT:    vpsllw $5, %xmm1, %xmm1
-; AVX512-NEXT:    vpunpckhbw {{.*#+}} xmm2 = xmm0[8],xmm1[8],xmm0[9],xmm1[9],xmm0[10],xmm1[10],xmm0[11],xmm1[11],xmm0[12],xmm1[12],xmm0[13],xmm1[13],xmm0[14],xmm1[14],xmm0[15],xmm1[15]
-; AVX512-NEXT:    vpunpckhbw {{.*#+}} xmm3 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
-; AVX512-NEXT:    vpsraw $4, %xmm3, %xmm4
-; AVX512-NEXT:    vpblendvb %xmm2, %xmm4, %xmm3, %xmm3
-; AVX512-NEXT:    vpsraw $2, %xmm3, %xmm4
-; AVX512-NEXT:    vpaddw %xmm2, %xmm2, %xmm2
-; AVX512-NEXT:    vpblendvb %xmm2, %xmm4, %xmm3, %xmm3
-; AVX512-NEXT:    vpsraw $1, %xmm3, %xmm4
-; AVX512-NEXT:    vpaddw %xmm2, %xmm2, %xmm2
-; AVX512-NEXT:    vpblendvb %xmm2, %xmm4, %xmm3, %xmm2
-; AVX512-NEXT:    vpsrlw $8, %xmm2, %xmm2
-; AVX512-NEXT:    vpunpcklbw {{.*#+}} xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3],xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
-; AVX512-NEXT:    vpunpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
-; AVX512-NEXT:    vpsraw $4, %xmm0, %xmm3
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm3, %xmm0, %xmm0
-; AVX512-NEXT:    vpsraw $2, %xmm0, %xmm3
-; AVX512-NEXT:    vpaddw %xmm1, %xmm1, %xmm1
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm3, %xmm0, %xmm0
-; AVX512-NEXT:    vpsraw $1, %xmm0, %xmm3
-; AVX512-NEXT:    vpaddw %xmm1, %xmm1, %xmm1
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm3, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrlw $8, %xmm0, %xmm0
-; AVX512-NEXT:    vpackuswb %xmm2, %xmm0, %xmm0
+; AVX512-NEXT:    vpmovsxbd %xmm0, %zmm0
+; AVX512-NEXT:    vpmovzxbd {{.*#+}} zmm1 = xmm1[0],zero,zero,zero,xmm1[1],zero,zero,zero,xmm1[2],zero,zero,zero,xmm1[3],zero,zero,zero,xmm1[4],zero,zero,zero,xmm1[5],zero,zero,zero,xmm1[6],zero,zero,zero,xmm1[7],zero,zero,zero,xmm1[8],zero,zero,zero,xmm1[9],zero,zero,zero,xmm1[10],zero,zero,zero,xmm1[11],zero,zero,zero,xmm1[12],zero,zero,zero,xmm1[13],zero,zero,zero,xmm1[14],zero,zero,zero,xmm1[15],zero,zero,zero
+; AVX512-NEXT:    vpsravd %zmm1, %zmm0, %zmm0
+; AVX512-NEXT:    vpmovdb %zmm0, %xmm0
 ; AVX512-NEXT:    retq
 ;
 ; X32-SSE-LABEL: splatvar_shift_v16i8:
@@ -1402,31 +1362,9 @@ define <16 x i8> @constant_shift_v16i8(<16 x i8> %a) nounwind {
 ;
 ; AVX512-LABEL: constant_shift_v16i8:
 ; AVX512:       # BB#0:
-; AVX512-NEXT:    vmovdqa {{.*#+}} xmm1 = [0,1,2,3,4,5,6,7,7,6,5,4,3,2,1,0]
-; AVX512-NEXT:    vpsllw $5, %xmm1, %xmm1
-; AVX512-NEXT:    vpunpckhbw {{.*#+}} xmm2 = xmm0[8],xmm1[8],xmm0[9],xmm1[9],xmm0[10],xmm1[10],xmm0[11],xmm1[11],xmm0[12],xmm1[12],xmm0[13],xmm1[13],xmm0[14],xmm1[14],xmm0[15],xmm1[15]
-; AVX512-NEXT:    vpunpckhbw {{.*#+}} xmm3 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
-; AVX512-NEXT:    vpsraw $4, %xmm3, %xmm4
-; AVX512-NEXT:    vpblendvb %xmm2, %xmm4, %xmm3, %xmm3
-; AVX512-NEXT:    vpsraw $2, %xmm3, %xmm4
-; AVX512-NEXT:    vpaddw %xmm2, %xmm2, %xmm2
-; AVX512-NEXT:    vpblendvb %xmm2, %xmm4, %xmm3, %xmm3
-; AVX512-NEXT:    vpsraw $1, %xmm3, %xmm4
-; AVX512-NEXT:    vpaddw %xmm2, %xmm2, %xmm2
-; AVX512-NEXT:    vpblendvb %xmm2, %xmm4, %xmm3, %xmm2
-; AVX512-NEXT:    vpsrlw $8, %xmm2, %xmm2
-; AVX512-NEXT:    vpunpcklbw {{.*#+}} xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3],xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
-; AVX512-NEXT:    vpunpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
-; AVX512-NEXT:    vpsraw $4, %xmm0, %xmm3
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm3, %xmm0, %xmm0
-; AVX512-NEXT:    vpsraw $2, %xmm0, %xmm3
-; AVX512-NEXT:    vpaddw %xmm1, %xmm1, %xmm1
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm3, %xmm0, %xmm0
-; AVX512-NEXT:    vpsraw $1, %xmm0, %xmm3
-; AVX512-NEXT:    vpaddw %xmm1, %xmm1, %xmm1
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm3, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrlw $8, %xmm0, %xmm0
-; AVX512-NEXT:    vpackuswb %xmm2, %xmm0, %xmm0
+; AVX512-NEXT:    vpmovsxbd %xmm0, %zmm0
+; AVX512-NEXT:    vpsravd {{.*}}(%rip), %zmm0, %zmm0
+; AVX512-NEXT:    vpmovdb %zmm0, %xmm0
 ; AVX512-NEXT:    retq
 ;
 ; X32-SSE-LABEL: constant_shift_v16i8:
diff --git a/llvm/test/CodeGen/X86/vector-shift-ashr-256.ll b/llvm/test/CodeGen/X86/vector-shift-ashr-256.ll
index f95b5c3dcc6..79902acfec2 100644
--- a/llvm/test/CodeGen/X86/vector-shift-ashr-256.ll
+++ b/llvm/test/CodeGen/X86/vector-shift-ashr-256.ll
@@ -340,33 +340,41 @@ define <32 x i8> @var_shift_v32i8(<32 x i8> %a, <32 x i8> %b) nounwind {
 ; XOPAVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
 ; XOPAVX2-NEXT:    retq
 ;
-; AVX512-LABEL: var_shift_v32i8:
-; AVX512:       # BB#0:
-; AVX512-NEXT:    vpsllw $5, %ymm1, %ymm1
-; AVX512-NEXT:    vpunpckhbw {{.*#+}} ymm2 = ymm0[8],ymm1[8],ymm0[9],ymm1[9],ymm0[10],ymm1[10],ymm0[11],ymm1[11],ymm0[12],ymm1[12],ymm0[13],ymm1[13],ymm0[14],ymm1[14],ymm0[15],ymm1[15],ymm0[24],ymm1[24],ymm0[25],ymm1[25],ymm0[26],ymm1[26],ymm0[27],ymm1[27],ymm0[28],ymm1[28],ymm0[29],ymm1[29],ymm0[30],ymm1[30],ymm0[31],ymm1[31]
-; AVX512-NEXT:    vpunpckhbw {{.*#+}} ymm3 = ymm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15,24,24,25,25,26,26,27,27,28,28,29,29,30,30,31,31]
-; AVX512-NEXT:    vpsraw $4, %ymm3, %ymm4
-; AVX512-NEXT:    vpblendvb %ymm2, %ymm4, %ymm3, %ymm3
-; AVX512-NEXT:    vpsraw $2, %ymm3, %ymm4
-; AVX512-NEXT:    vpaddw %ymm2, %ymm2, %ymm2
-; AVX512-NEXT:    vpblendvb %ymm2, %ymm4, %ymm3, %ymm3
-; AVX512-NEXT:    vpsraw $1, %ymm3, %ymm4
-; AVX512-NEXT:    vpaddw %ymm2, %ymm2, %ymm2
-; AVX512-NEXT:    vpblendvb %ymm2, %ymm4, %ymm3, %ymm2
-; AVX512-NEXT:    vpsrlw $8, %ymm2, %ymm2
-; AVX512-NEXT:    vpunpcklbw {{.*#+}} ymm1 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[4],ymm1[4],ymm0[5],ymm1[5],ymm0[6],ymm1[6],ymm0[7],ymm1[7],ymm0[16],ymm1[16],ymm0[17],ymm1[17],ymm0[18],ymm1[18],ymm0[19],ymm1[19],ymm0[20],ymm1[20],ymm0[21],ymm1[21],ymm0[22],ymm1[22],ymm0[23],ymm1[23]
-; AVX512-NEXT:    vpunpcklbw {{.*#+}} ymm0 = ymm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7,16,16,17,17,18,18,19,19,20,20,21,21,22,22,23,23]
-; AVX512-NEXT:    vpsraw $4, %ymm0, %ymm3
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm3, %ymm0, %ymm0
-; AVX512-NEXT:    vpsraw $2, %ymm0, %ymm3
-; AVX512-NEXT:    vpaddw %ymm1, %ymm1, %ymm1
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm3, %ymm0, %ymm0
-; AVX512-NEXT:    vpsraw $1, %ymm0, %ymm3
-; AVX512-NEXT:    vpaddw %ymm1, %ymm1, %ymm1
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm3, %ymm0, %ymm0
-; AVX512-NEXT:    vpsrlw $8, %ymm0, %ymm0
-; AVX512-NEXT:    vpackuswb %ymm2, %ymm0, %ymm0
-; AVX512-NEXT:    retq
+; AVX512DQ-LABEL: var_shift_v32i8:
+; AVX512DQ:       # BB#0:
+; AVX512DQ-NEXT:    vpsllw $5, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpunpckhbw {{.*#+}} ymm2 = ymm0[8],ymm1[8],ymm0[9],ymm1[9],ymm0[10],ymm1[10],ymm0[11],ymm1[11],ymm0[12],ymm1[12],ymm0[13],ymm1[13],ymm0[14],ymm1[14],ymm0[15],ymm1[15],ymm0[24],ymm1[24],ymm0[25],ymm1[25],ymm0[26],ymm1[26],ymm0[27],ymm1[27],ymm0[28],ymm1[28],ymm0[29],ymm1[29],ymm0[30],ymm1[30],ymm0[31],ymm1[31]
+; AVX512DQ-NEXT:    vpunpckhbw {{.*#+}} ymm3 = ymm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15,24,24,25,25,26,26,27,27,28,28,29,29,30,30,31,31]
+; AVX512DQ-NEXT:    vpsraw $4, %ymm3, %ymm4
+; AVX512DQ-NEXT:    vpblendvb %ymm2, %ymm4, %ymm3, %ymm3
+; AVX512DQ-NEXT:    vpsraw $2, %ymm3, %ymm4
+; AVX512DQ-NEXT:    vpaddw %ymm2, %ymm2, %ymm2
+; AVX512DQ-NEXT:    vpblendvb %ymm2, %ymm4, %ymm3, %ymm3
+; AVX512DQ-NEXT:    vpsraw $1, %ymm3, %ymm4
+; AVX512DQ-NEXT:    vpaddw %ymm2, %ymm2, %ymm2
+; AVX512DQ-NEXT:    vpblendvb %ymm2, %ymm4, %ymm3, %ymm2
+; AVX512DQ-NEXT:    vpsrlw $8, %ymm2, %ymm2
+; AVX512DQ-NEXT:    vpunpcklbw {{.*#+}} ymm1 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[4],ymm1[4],ymm0[5],ymm1[5],ymm0[6],ymm1[6],ymm0[7],ymm1[7],ymm0[16],ymm1[16],ymm0[17],ymm1[17],ymm0[18],ymm1[18],ymm0[19],ymm1[19],ymm0[20],ymm1[20],ymm0[21],ymm1[21],ymm0[22],ymm1[22],ymm0[23],ymm1[23]
+; AVX512DQ-NEXT:    vpunpcklbw {{.*#+}} ymm0 = ymm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7,16,16,17,17,18,18,19,19,20,20,21,21,22,22,23,23]
+; AVX512DQ-NEXT:    vpsraw $4, %ymm0, %ymm3
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm3, %ymm0, %ymm0
+; AVX512DQ-NEXT:    vpsraw $2, %ymm0, %ymm3
+; AVX512DQ-NEXT:    vpaddw %ymm1, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm3, %ymm0, %ymm0
+; AVX512DQ-NEXT:    vpsraw $1, %ymm0, %ymm3
+; AVX512DQ-NEXT:    vpaddw %ymm1, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm3, %ymm0, %ymm0
+; AVX512DQ-NEXT:    vpsrlw $8, %ymm0, %ymm0
+; AVX512DQ-NEXT:    vpackuswb %ymm2, %ymm0, %ymm0
+; AVX512DQ-NEXT:    retq
+;
+; AVX512BW-LABEL: var_shift_v32i8:
+; AVX512BW:       # BB#0:
+; AVX512BW-NEXT:    vpmovzxbw {{.*#+}} 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,ymm1[16],zero,ymm1[17],zero,ymm1[18],zero,ymm1[19],zero,ymm1[20],zero,ymm1[21],zero,ymm1[22],zero,ymm1[23],zero,ymm1[24],zero,ymm1[25],zero,ymm1[26],zero,ymm1[27],zero,ymm1[28],zero,ymm1[29],zero,ymm1[30],zero,ymm1[31],zero
+; AVX512BW-NEXT:    vpmovsxbw %ymm0, %zmm0
+; AVX512BW-NEXT:    vpsravw %zmm1, %zmm0, %zmm0
+; AVX512BW-NEXT:    vpmovwb %zmm0, %ymm0
+; AVX512BW-NEXT:    retq
   %shift = ashr <32 x i8> %a, %b
   ret <32 x i8> %shift
 }
@@ -617,34 +625,43 @@ define <32 x i8> @splatvar_shift_v32i8(<32 x i8> %a, <32 x i8> %b) nounwind {
 ; XOPAVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
 ; XOPAVX2-NEXT:    retq
 ;
-; AVX512-LABEL: splatvar_shift_v32i8:
-; AVX512:       # BB#0:
-; AVX512-NEXT:    vpbroadcastb %xmm1, %ymm1
-; AVX512-NEXT:    vpsllw $5, %ymm1, %ymm1
-; AVX512-NEXT:    vpunpckhbw {{.*#+}} ymm2 = ymm0[8],ymm1[8],ymm0[9],ymm1[9],ymm0[10],ymm1[10],ymm0[11],ymm1[11],ymm0[12],ymm1[12],ymm0[13],ymm1[13],ymm0[14],ymm1[14],ymm0[15],ymm1[15],ymm0[24],ymm1[24],ymm0[25],ymm1[25],ymm0[26],ymm1[26],ymm0[27],ymm1[27],ymm0[28],ymm1[28],ymm0[29],ymm1[29],ymm0[30],ymm1[30],ymm0[31],ymm1[31]
-; AVX512-NEXT:    vpunpckhbw {{.*#+}} ymm3 = ymm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15,24,24,25,25,26,26,27,27,28,28,29,29,30,30,31,31]
-; AVX512-NEXT:    vpsraw $4, %ymm3, %ymm4
-; AVX512-NEXT:    vpblendvb %ymm2, %ymm4, %ymm3, %ymm3
-; AVX512-NEXT:    vpsraw $2, %ymm3, %ymm4
-; AVX512-NEXT:    vpaddw %ymm2, %ymm2, %ymm2
-; AVX512-NEXT:    vpblendvb %ymm2, %ymm4, %ymm3, %ymm3
-; AVX512-NEXT:    vpsraw $1, %ymm3, %ymm4
-; AVX512-NEXT:    vpaddw %ymm2, %ymm2, %ymm2
-; AVX512-NEXT:    vpblendvb %ymm2, %ymm4, %ymm3, %ymm2
-; AVX512-NEXT:    vpsrlw $8, %ymm2, %ymm2
-; AVX512-NEXT:    vpunpcklbw {{.*#+}} ymm1 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[4],ymm1[4],ymm0[5],ymm1[5],ymm0[6],ymm1[6],ymm0[7],ymm1[7],ymm0[16],ymm1[16],ymm0[17],ymm1[17],ymm0[18],ymm1[18],ymm0[19],ymm1[19],ymm0[20],ymm1[20],ymm0[21],ymm1[21],ymm0[22],ymm1[22],ymm0[23],ymm1[23]
-; AVX512-NEXT:    vpunpcklbw {{.*#+}} ymm0 = ymm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7,16,16,17,17,18,18,19,19,20,20,21,21,22,22,23,23]
-; AVX512-NEXT:    vpsraw $4, %ymm0, %ymm3
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm3, %ymm0, %ymm0
-; AVX512-NEXT:    vpsraw $2, %ymm0, %ymm3
-; AVX512-NEXT:    vpaddw %ymm1, %ymm1, %ymm1
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm3, %ymm0, %ymm0
-; AVX512-NEXT:    vpsraw $1, %ymm0, %ymm3
-; AVX512-NEXT:    vpaddw %ymm1, %ymm1, %ymm1
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm3, %ymm0, %ymm0
-; AVX512-NEXT:    vpsrlw $8, %ymm0, %ymm0
-; AVX512-NEXT:    vpackuswb %ymm2, %ymm0, %ymm0
-; AVX512-NEXT:    retq
+; AVX512DQ-LABEL: splatvar_shift_v32i8:
+; AVX512DQ:       # BB#0:
+; AVX512DQ-NEXT:    vpbroadcastb %xmm1, %ymm1
+; AVX512DQ-NEXT:    vpsllw $5, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpunpckhbw {{.*#+}} ymm2 = ymm0[8],ymm1[8],ymm0[9],ymm1[9],ymm0[10],ymm1[10],ymm0[11],ymm1[11],ymm0[12],ymm1[12],ymm0[13],ymm1[13],ymm0[14],ymm1[14],ymm0[15],ymm1[15],ymm0[24],ymm1[24],ymm0[25],ymm1[25],ymm0[26],ymm1[26],ymm0[27],ymm1[27],ymm0[28],ymm1[28],ymm0[29],ymm1[29],ymm0[30],ymm1[30],ymm0[31],ymm1[31]
+; AVX512DQ-NEXT:    vpunpckhbw {{.*#+}} ymm3 = ymm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15,24,24,25,25,26,26,27,27,28,28,29,29,30,30,31,31]
+; AVX512DQ-NEXT:    vpsraw $4, %ymm3, %ymm4
+; AVX512DQ-NEXT:    vpblendvb %ymm2, %ymm4, %ymm3, %ymm3
+; AVX512DQ-NEXT:    vpsraw $2, %ymm3, %ymm4
+; AVX512DQ-NEXT:    vpaddw %ymm2, %ymm2, %ymm2
+; AVX512DQ-NEXT:    vpblendvb %ymm2, %ymm4, %ymm3, %ymm3
+; AVX512DQ-NEXT:    vpsraw $1, %ymm3, %ymm4
+; AVX512DQ-NEXT:    vpaddw %ymm2, %ymm2, %ymm2
+; AVX512DQ-NEXT:    vpblendvb %ymm2, %ymm4, %ymm3, %ymm2
+; AVX512DQ-NEXT:    vpsrlw $8, %ymm2, %ymm2
+; AVX512DQ-NEXT:    vpunpcklbw {{.*#+}} ymm1 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[4],ymm1[4],ymm0[5],ymm1[5],ymm0[6],ymm1[6],ymm0[7],ymm1[7],ymm0[16],ymm1[16],ymm0[17],ymm1[17],ymm0[18],ymm1[18],ymm0[19],ymm1[19],ymm0[20],ymm1[20],ymm0[21],ymm1[21],ymm0[22],ymm1[22],ymm0[23],ymm1[23]
+; AVX512DQ-NEXT:    vpunpcklbw {{.*#+}} ymm0 = ymm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7,16,16,17,17,18,18,19,19,20,20,21,21,22,22,23,23]
+; AVX512DQ-NEXT:    vpsraw $4, %ymm0, %ymm3
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm3, %ymm0, %ymm0
+; AVX512DQ-NEXT:    vpsraw $2, %ymm0, %ymm3
+; AVX512DQ-NEXT:    vpaddw %ymm1, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm3, %ymm0, %ymm0
+; AVX512DQ-NEXT:    vpsraw $1, %ymm0, %ymm3
+; AVX512DQ-NEXT:    vpaddw %ymm1, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm3, %ymm0, %ymm0
+; AVX512DQ-NEXT:    vpsrlw $8, %ymm0, %ymm0
+; AVX512DQ-NEXT:    vpackuswb %ymm2, %ymm0, %ymm0
+; AVX512DQ-NEXT:    retq
+;
+; AVX512BW-LABEL: splatvar_shift_v32i8:
+; AVX512BW:       # BB#0:
+; AVX512BW-NEXT:    vpbroadcastb %xmm1, %ymm1
+; AVX512BW-NEXT:    vpmovsxbw %ymm0, %zmm0
+; AVX512BW-NEXT:    vpmovzxbw {{.*#+}} 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,ymm1[16],zero,ymm1[17],zero,ymm1[18],zero,ymm1[19],zero,ymm1[20],zero,ymm1[21],zero,ymm1[22],zero,ymm1[23],zero,ymm1[24],zero,ymm1[25],zero,ymm1[26],zero,ymm1[27],zero,ymm1[28],zero,ymm1[29],zero,ymm1[30],zero,ymm1[31],zero
+; AVX512BW-NEXT:    vpsravw %zmm1, %zmm0, %zmm0
+; AVX512BW-NEXT:    vpmovwb %zmm0, %ymm0
+; AVX512BW-NEXT:    retq
   %splat = shufflevector <32 x i8> %b, <32 x i8> undef, <32 x i32> zeroinitializer
   %shift = ashr <32 x i8> %a, %splat
   ret <32 x i8> %shift
@@ -929,34 +946,41 @@ define <32 x i8> @constant_shift_v32i8(<32 x i8> %a) nounwind {
 ; XOPAVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
 ; XOPAVX2-NEXT:    retq
 ;
-; AVX512-LABEL: constant_shift_v32i8:
-; AVX512:       # BB#0:
-; AVX512-NEXT:    vmovdqa {{.*#+}} ymm1 = [0,1,2,3,4,5,6,7,7,6,5,4,3,2,1,0,0,1,2,3,4,5,6,7,7,6,5,4,3,2,1,0]
-; AVX512-NEXT:    vpsllw $5, %ymm1, %ymm1
-; AVX512-NEXT:    vpunpckhbw {{.*#+}} ymm2 = ymm0[8],ymm1[8],ymm0[9],ymm1[9],ymm0[10],ymm1[10],ymm0[11],ymm1[11],ymm0[12],ymm1[12],ymm0[13],ymm1[13],ymm0[14],ymm1[14],ymm0[15],ymm1[15],ymm0[24],ymm1[24],ymm0[25],ymm1[25],ymm0[26],ymm1[26],ymm0[27],ymm1[27],ymm0[28],ymm1[28],ymm0[29],ymm1[29],ymm0[30],ymm1[30],ymm0[31],ymm1[31]
-; AVX512-NEXT:    vpunpckhbw {{.*#+}} ymm3 = ymm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15,24,24,25,25,26,26,27,27,28,28,29,29,30,30,31,31]
-; AVX512-NEXT:    vpsraw $4, %ymm3, %ymm4
-; AVX512-NEXT:    vpblendvb %ymm2, %ymm4, %ymm3, %ymm3
-; AVX512-NEXT:    vpsraw $2, %ymm3, %ymm4
-; AVX512-NEXT:    vpaddw %ymm2, %ymm2, %ymm2
-; AVX512-NEXT:    vpblendvb %ymm2, %ymm4, %ymm3, %ymm3
-; AVX512-NEXT:    vpsraw $1, %ymm3, %ymm4
-; AVX512-NEXT:    vpaddw %ymm2, %ymm2, %ymm2
-; AVX512-NEXT:    vpblendvb %ymm2, %ymm4, %ymm3, %ymm2
-; AVX512-NEXT:    vpsrlw $8, %ymm2, %ymm2
-; AVX512-NEXT:    vpunpcklbw {{.*#+}} ymm1 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[4],ymm1[4],ymm0[5],ymm1[5],ymm0[6],ymm1[6],ymm0[7],ymm1[7],ymm0[16],ymm1[16],ymm0[17],ymm1[17],ymm0[18],ymm1[18],ymm0[19],ymm1[19],ymm0[20],ymm1[20],ymm0[21],ymm1[21],ymm0[22],ymm1[22],ymm0[23],ymm1[23]
-; AVX512-NEXT:    vpunpcklbw {{.*#+}} ymm0 = ymm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7,16,16,17,17,18,18,19,19,20,20,21,21,22,22,23,23]
-; AVX512-NEXT:    vpsraw $4, %ymm0, %ymm3
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm3, %ymm0, %ymm0
-; AVX512-NEXT:    vpsraw $2, %ymm0, %ymm3
-; AVX512-NEXT:    vpaddw %ymm1, %ymm1, %ymm1
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm3, %ymm0, %ymm0
-; AVX512-NEXT:    vpsraw $1, %ymm0, %ymm3
-; AVX512-NEXT:    vpaddw %ymm1, %ymm1, %ymm1
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm3, %ymm0, %ymm0
-; AVX512-NEXT:    vpsrlw $8, %ymm0, %ymm0
-; AVX512-NEXT:    vpackuswb %ymm2, %ymm0, %ymm0
-; AVX512-NEXT:    retq
+; AVX512DQ-LABEL: constant_shift_v32i8:
+; AVX512DQ:       # BB#0:
+; AVX512DQ-NEXT:    vmovdqa {{.*#+}} ymm1 = [0,1,2,3,4,5,6,7,7,6,5,4,3,2,1,0,0,1,2,3,4,5,6,7,7,6,5,4,3,2,1,0]
+; AVX512DQ-NEXT:    vpsllw $5, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpunpckhbw {{.*#+}} ymm2 = ymm0[8],ymm1[8],ymm0[9],ymm1[9],ymm0[10],ymm1[10],ymm0[11],ymm1[11],ymm0[12],ymm1[12],ymm0[13],ymm1[13],ymm0[14],ymm1[14],ymm0[15],ymm1[15],ymm0[24],ymm1[24],ymm0[25],ymm1[25],ymm0[26],ymm1[26],ymm0[27],ymm1[27],ymm0[28],ymm1[28],ymm0[29],ymm1[29],ymm0[30],ymm1[30],ymm0[31],ymm1[31]
+; AVX512DQ-NEXT:    vpunpckhbw {{.*#+}} ymm3 = ymm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15,24,24,25,25,26,26,27,27,28,28,29,29,30,30,31,31]
+; AVX512DQ-NEXT:    vpsraw $4, %ymm3, %ymm4
+; AVX512DQ-NEXT:    vpblendvb %ymm2, %ymm4, %ymm3, %ymm3
+; AVX512DQ-NEXT:    vpsraw $2, %ymm3, %ymm4
+; AVX512DQ-NEXT:    vpaddw %ymm2, %ymm2, %ymm2
+; AVX512DQ-NEXT:    vpblendvb %ymm2, %ymm4, %ymm3, %ymm3
+; AVX512DQ-NEXT:    vpsraw $1, %ymm3, %ymm4
+; AVX512DQ-NEXT:    vpaddw %ymm2, %ymm2, %ymm2
+; AVX512DQ-NEXT:    vpblendvb %ymm2, %ymm4, %ymm3, %ymm2
+; AVX512DQ-NEXT:    vpsrlw $8, %ymm2, %ymm2
+; AVX512DQ-NEXT:    vpunpcklbw {{.*#+}} ymm1 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[4],ymm1[4],ymm0[5],ymm1[5],ymm0[6],ymm1[6],ymm0[7],ymm1[7],ymm0[16],ymm1[16],ymm0[17],ymm1[17],ymm0[18],ymm1[18],ymm0[19],ymm1[19],ymm0[20],ymm1[20],ymm0[21],ymm1[21],ymm0[22],ymm1[22],ymm0[23],ymm1[23]
+; AVX512DQ-NEXT:    vpunpcklbw {{.*#+}} ymm0 = ymm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7,16,16,17,17,18,18,19,19,20,20,21,21,22,22,23,23]
+; AVX512DQ-NEXT:    vpsraw $4, %ymm0, %ymm3
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm3, %ymm0, %ymm0
+; AVX512DQ-NEXT:    vpsraw $2, %ymm0, %ymm3
+; AVX512DQ-NEXT:    vpaddw %ymm1, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm3, %ymm0, %ymm0
+; AVX512DQ-NEXT:    vpsraw $1, %ymm0, %ymm3
+; AVX512DQ-NEXT:    vpaddw %ymm1, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm3, %ymm0, %ymm0
+; AVX512DQ-NEXT:    vpsrlw $8, %ymm0, %ymm0
+; AVX512DQ-NEXT:    vpackuswb %ymm2, %ymm0, %ymm0
+; AVX512DQ-NEXT:    retq
+;
+; AVX512BW-LABEL: constant_shift_v32i8:
+; AVX512BW:       # BB#0:
+; AVX512BW-NEXT:    vpmovsxbw %ymm0, %zmm0
+; AVX512BW-NEXT:    vpsravw {{.*}}(%rip), %zmm0, %zmm0
+; AVX512BW-NEXT:    vpmovwb %zmm0, %ymm0
+; AVX512BW-NEXT:    retq
   %shift = ashr <32 x i8> %a, <i8 0, i8 1, i8 2, i8 3, i8 4, i8 5, i8 6, i8 7, i8 7, i8 6, i8 5, i8 4, i8 3, i8 2, i8 1, i8 0, i8 0, i8 1, i8 2, i8 3, i8 4, i8 5, i8 6, i8 7, i8 7, i8 6, i8 5, i8 4, i8 3, i8 2, i8 1, i8 0>
   ret <32 x i8> %shift
 }
diff --git a/llvm/test/CodeGen/X86/vector-shift-lshr-128.ll b/llvm/test/CodeGen/X86/vector-shift-lshr-128.ll
index 4520c24c863..a7e1a531b65 100644
--- a/llvm/test/CodeGen/X86/vector-shift-lshr-128.ll
+++ b/llvm/test/CodeGen/X86/vector-shift-lshr-128.ll
@@ -427,18 +427,10 @@ define <16 x i8> @var_shift_v16i8(<16 x i8> %a, <16 x i8> %b) nounwind {
 ;
 ; AVX512-LABEL: var_shift_v16i8:
 ; AVX512:       # BB#0:
-; AVX512-NEXT:    vpsllw $5, %xmm1, %xmm1
-; AVX512-NEXT:    vpsrlw $4, %xmm0, %xmm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm2, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrlw $2, %xmm0, %xmm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
-; AVX512-NEXT:    vpaddb %xmm1, %xmm1, %xmm1
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm2, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrlw $1, %xmm0, %xmm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
-; AVX512-NEXT:    vpaddb %xmm1, %xmm1, %xmm1
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm2, %xmm0, %xmm0
+; AVX512-NEXT:    vpmovzxbd {{.*#+}} zmm1 = xmm1[0],zero,zero,zero,xmm1[1],zero,zero,zero,xmm1[2],zero,zero,zero,xmm1[3],zero,zero,zero,xmm1[4],zero,zero,zero,xmm1[5],zero,zero,zero,xmm1[6],zero,zero,zero,xmm1[7],zero,zero,zero,xmm1[8],zero,zero,zero,xmm1[9],zero,zero,zero,xmm1[10],zero,zero,zero,xmm1[11],zero,zero,zero,xmm1[12],zero,zero,zero,xmm1[13],zero,zero,zero,xmm1[14],zero,zero,zero,xmm1[15],zero,zero,zero
+; AVX512-NEXT:    vpmovzxbd {{.*#+}} zmm0 = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero,xmm0[4],zero,zero,zero,xmm0[5],zero,zero,zero,xmm0[6],zero,zero,zero,xmm0[7],zero,zero,zero,xmm0[8],zero,zero,zero,xmm0[9],zero,zero,zero,xmm0[10],zero,zero,zero,xmm0[11],zero,zero,zero,xmm0[12],zero,zero,zero,xmm0[13],zero,zero,zero,xmm0[14],zero,zero,zero,xmm0[15],zero,zero,zero
+; AVX512-NEXT:    vpsrlvd %zmm1, %zmm0, %zmm0
+; AVX512-NEXT:    vpmovdb %zmm0, %xmm0
 ; AVX512-NEXT:    retq
 ;
 ; X32-SSE-LABEL: var_shift_v16i8:
@@ -711,18 +703,10 @@ define <16 x i8> @splatvar_shift_v16i8(<16 x i8> %a, <16 x i8> %b) nounwind {
 ; AVX512-LABEL: splatvar_shift_v16i8:
 ; AVX512:       # BB#0:
 ; AVX512-NEXT:    vpbroadcastb %xmm1, %xmm1
-; AVX512-NEXT:    vpsllw $5, %xmm1, %xmm1
-; AVX512-NEXT:    vpsrlw $4, %xmm0, %xmm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm2, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrlw $2, %xmm0, %xmm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
-; AVX512-NEXT:    vpaddb %xmm1, %xmm1, %xmm1
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm2, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrlw $1, %xmm0, %xmm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
-; AVX512-NEXT:    vpaddb %xmm1, %xmm1, %xmm1
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm2, %xmm0, %xmm0
+; AVX512-NEXT:    vpmovzxbd {{.*#+}} zmm0 = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero,xmm0[4],zero,zero,zero,xmm0[5],zero,zero,zero,xmm0[6],zero,zero,zero,xmm0[7],zero,zero,zero,xmm0[8],zero,zero,zero,xmm0[9],zero,zero,zero,xmm0[10],zero,zero,zero,xmm0[11],zero,zero,zero,xmm0[12],zero,zero,zero,xmm0[13],zero,zero,zero,xmm0[14],zero,zero,zero,xmm0[15],zero,zero,zero
+; AVX512-NEXT:    vpmovzxbd {{.*#+}} zmm1 = xmm1[0],zero,zero,zero,xmm1[1],zero,zero,zero,xmm1[2],zero,zero,zero,xmm1[3],zero,zero,zero,xmm1[4],zero,zero,zero,xmm1[5],zero,zero,zero,xmm1[6],zero,zero,zero,xmm1[7],zero,zero,zero,xmm1[8],zero,zero,zero,xmm1[9],zero,zero,zero,xmm1[10],zero,zero,zero,xmm1[11],zero,zero,zero,xmm1[12],zero,zero,zero,xmm1[13],zero,zero,zero,xmm1[14],zero,zero,zero,xmm1[15],zero,zero,zero
+; AVX512-NEXT:    vpsrlvd %zmm1, %zmm0, %zmm0
+; AVX512-NEXT:    vpmovdb %zmm0, %xmm0
 ; AVX512-NEXT:    retq
 ;
 ; X32-SSE-LABEL: splatvar_shift_v16i8:
@@ -1082,19 +1066,9 @@ define <16 x i8> @constant_shift_v16i8(<16 x i8> %a) nounwind {
 ;
 ; AVX512-LABEL: constant_shift_v16i8:
 ; AVX512:       # BB#0:
-; AVX512-NEXT:    vmovdqa {{.*#+}} xmm1 = [0,1,2,3,4,5,6,7,7,6,5,4,3,2,1,0]
-; AVX512-NEXT:    vpsllw $5, %xmm1, %xmm1
-; AVX512-NEXT:    vpsrlw $4, %xmm0, %xmm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm2, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrlw $2, %xmm0, %xmm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
-; AVX512-NEXT:    vpaddb %xmm1, %xmm1, %xmm1
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm2, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrlw $1, %xmm0, %xmm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
-; AVX512-NEXT:    vpaddb %xmm1, %xmm1, %xmm1
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm2, %xmm0, %xmm0
+; AVX512-NEXT:    vpmovzxbd {{.*#+}} zmm0 = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero,xmm0[4],zero,zero,zero,xmm0[5],zero,zero,zero,xmm0[6],zero,zero,zero,xmm0[7],zero,zero,zero,xmm0[8],zero,zero,zero,xmm0[9],zero,zero,zero,xmm0[10],zero,zero,zero,xmm0[11],zero,zero,zero,xmm0[12],zero,zero,zero,xmm0[13],zero,zero,zero,xmm0[14],zero,zero,zero,xmm0[15],zero,zero,zero
+; AVX512-NEXT:    vpsrlvd {{.*}}(%rip), %zmm0, %zmm0
+; AVX512-NEXT:    vpmovdb %zmm0, %xmm0
 ; AVX512-NEXT:    retq
 ;
 ; X32-SSE-LABEL: constant_shift_v16i8:
diff --git a/llvm/test/CodeGen/X86/vector-shift-lshr-256.ll b/llvm/test/CodeGen/X86/vector-shift-lshr-256.ll
index b61ca212aad..25667e7d166 100644
--- a/llvm/test/CodeGen/X86/vector-shift-lshr-256.ll
+++ b/llvm/test/CodeGen/X86/vector-shift-lshr-256.ll
@@ -284,21 +284,29 @@ define <32 x i8> @var_shift_v32i8(<32 x i8> %a, <32 x i8> %b) nounwind {
 ; XOPAVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
 ; XOPAVX2-NEXT:    retq
 ;
-; AVX512-LABEL: var_shift_v32i8:
-; AVX512:       # BB#0:
-; AVX512-NEXT:    vpsllw $5, %ymm1, %ymm1
-; AVX512-NEXT:    vpsrlw $4, %ymm0, %ymm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
-; AVX512-NEXT:    vpsrlw $2, %ymm0, %ymm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
-; AVX512-NEXT:    vpaddb %ymm1, %ymm1, %ymm1
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
-; AVX512-NEXT:    vpsrlw $1, %ymm0, %ymm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
-; AVX512-NEXT:    vpaddb %ymm1, %ymm1, %ymm1
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
-; AVX512-NEXT:    retq
+; AVX512DQ-LABEL: var_shift_v32i8:
+; AVX512DQ:       # BB#0:
+; AVX512DQ-NEXT:    vpsllw $5, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpsrlw $4, %ymm0, %ymm2
+; AVX512DQ-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
+; AVX512DQ-NEXT:    vpsrlw $2, %ymm0, %ymm2
+; AVX512DQ-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
+; AVX512DQ-NEXT:    vpaddb %ymm1, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
+; AVX512DQ-NEXT:    vpsrlw $1, %ymm0, %ymm2
+; AVX512DQ-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
+; AVX512DQ-NEXT:    vpaddb %ymm1, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
+; AVX512DQ-NEXT:    retq
+;
+; AVX512BW-LABEL: var_shift_v32i8:
+; AVX512BW:       # BB#0:
+; AVX512BW-NEXT:    vpmovzxbw {{.*#+}} 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,ymm1[16],zero,ymm1[17],zero,ymm1[18],zero,ymm1[19],zero,ymm1[20],zero,ymm1[21],zero,ymm1[22],zero,ymm1[23],zero,ymm1[24],zero,ymm1[25],zero,ymm1[26],zero,ymm1[27],zero,ymm1[28],zero,ymm1[29],zero,ymm1[30],zero,ymm1[31],zero
+; AVX512BW-NEXT:    vpmovzxbw {{.*#+}} 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,ymm0[16],zero,ymm0[17],zero,ymm0[18],zero,ymm0[19],zero,ymm0[20],zero,ymm0[21],zero,ymm0[22],zero,ymm0[23],zero,ymm0[24],zero,ymm0[25],zero,ymm0[26],zero,ymm0[27],zero,ymm0[28],zero,ymm0[29],zero,ymm0[30],zero,ymm0[31],zero
+; AVX512BW-NEXT:    vpsrlvw %zmm1, %zmm0, %zmm0
+; AVX512BW-NEXT:    vpmovwb %zmm0, %ymm0
+; AVX512BW-NEXT:    retq
   %shift = lshr <32 x i8> %a, %b
   ret <32 x i8> %shift
 }
@@ -499,22 +507,31 @@ define <32 x i8> @splatvar_shift_v32i8(<32 x i8> %a, <32 x i8> %b) nounwind {
 ; XOPAVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
 ; XOPAVX2-NEXT:    retq
 ;
-; AVX512-LABEL: splatvar_shift_v32i8:
-; AVX512:       # BB#0:
-; AVX512-NEXT:    vpbroadcastb %xmm1, %ymm1
-; AVX512-NEXT:    vpsrlw $4, %ymm0, %ymm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
-; AVX512-NEXT:    vpsllw $5, %ymm1, %ymm1
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
-; AVX512-NEXT:    vpsrlw $2, %ymm0, %ymm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
-; AVX512-NEXT:    vpaddb %ymm1, %ymm1, %ymm1
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
-; AVX512-NEXT:    vpsrlw $1, %ymm0, %ymm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
-; AVX512-NEXT:    vpaddb %ymm1, %ymm1, %ymm1
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
-; AVX512-NEXT:    retq
+; AVX512DQ-LABEL: splatvar_shift_v32i8:
+; AVX512DQ:       # BB#0:
+; AVX512DQ-NEXT:    vpbroadcastb %xmm1, %ymm1
+; AVX512DQ-NEXT:    vpsrlw $4, %ymm0, %ymm2
+; AVX512DQ-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
+; AVX512DQ-NEXT:    vpsllw $5, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
+; AVX512DQ-NEXT:    vpsrlw $2, %ymm0, %ymm2
+; AVX512DQ-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
+; AVX512DQ-NEXT:    vpaddb %ymm1, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
+; AVX512DQ-NEXT:    vpsrlw $1, %ymm0, %ymm2
+; AVX512DQ-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
+; AVX512DQ-NEXT:    vpaddb %ymm1, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
+; AVX512DQ-NEXT:    retq
+;
+; AVX512BW-LABEL: splatvar_shift_v32i8:
+; AVX512BW:       # BB#0:
+; AVX512BW-NEXT:    vpbroadcastb %xmm1, %ymm1
+; AVX512BW-NEXT:    vpmovzxbw {{.*#+}} 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,ymm0[16],zero,ymm0[17],zero,ymm0[18],zero,ymm0[19],zero,ymm0[20],zero,ymm0[21],zero,ymm0[22],zero,ymm0[23],zero,ymm0[24],zero,ymm0[25],zero,ymm0[26],zero,ymm0[27],zero,ymm0[28],zero,ymm0[29],zero,ymm0[30],zero,ymm0[31],zero
+; AVX512BW-NEXT:    vpmovzxbw {{.*#+}} 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,ymm1[16],zero,ymm1[17],zero,ymm1[18],zero,ymm1[19],zero,ymm1[20],zero,ymm1[21],zero,ymm1[22],zero,ymm1[23],zero,ymm1[24],zero,ymm1[25],zero,ymm1[26],zero,ymm1[27],zero,ymm1[28],zero,ymm1[29],zero,ymm1[30],zero,ymm1[31],zero
+; AVX512BW-NEXT:    vpsrlvw %zmm1, %zmm0, %zmm0
+; AVX512BW-NEXT:    vpmovwb %zmm0, %ymm0
+; AVX512BW-NEXT:    retq
   %splat = shufflevector <32 x i8> %b, <32 x i8> undef, <32 x i32> zeroinitializer
   %shift = lshr <32 x i8> %a, %splat
   ret <32 x i8> %shift
@@ -755,22 +772,29 @@ define <32 x i8> @constant_shift_v32i8(<32 x i8> %a) nounwind {
 ; XOPAVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
 ; XOPAVX2-NEXT:    retq
 ;
-; AVX512-LABEL: constant_shift_v32i8:
-; AVX512:       # BB#0:
-; AVX512-NEXT:    vmovdqa {{.*#+}} ymm1 = [0,1,2,3,4,5,6,7,7,6,5,4,3,2,1,0,0,1,2,3,4,5,6,7,7,6,5,4,3,2,1,0]
-; AVX512-NEXT:    vpsllw $5, %ymm1, %ymm1
-; AVX512-NEXT:    vpsrlw $4, %ymm0, %ymm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
-; AVX512-NEXT:    vpsrlw $2, %ymm0, %ymm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
-; AVX512-NEXT:    vpaddb %ymm1, %ymm1, %ymm1
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
-; AVX512-NEXT:    vpsrlw $1, %ymm0, %ymm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
-; AVX512-NEXT:    vpaddb %ymm1, %ymm1, %ymm1
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
-; AVX512-NEXT:    retq
+; AVX512DQ-LABEL: constant_shift_v32i8:
+; AVX512DQ:       # BB#0:
+; AVX512DQ-NEXT:    vmovdqa {{.*#+}} ymm1 = [0,1,2,3,4,5,6,7,7,6,5,4,3,2,1,0,0,1,2,3,4,5,6,7,7,6,5,4,3,2,1,0]
+; AVX512DQ-NEXT:    vpsllw $5, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpsrlw $4, %ymm0, %ymm2
+; AVX512DQ-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
+; AVX512DQ-NEXT:    vpsrlw $2, %ymm0, %ymm2
+; AVX512DQ-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
+; AVX512DQ-NEXT:    vpaddb %ymm1, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
+; AVX512DQ-NEXT:    vpsrlw $1, %ymm0, %ymm2
+; AVX512DQ-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
+; AVX512DQ-NEXT:    vpaddb %ymm1, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
+; AVX512DQ-NEXT:    retq
+;
+; AVX512BW-LABEL: constant_shift_v32i8:
+; AVX512BW:       # BB#0:
+; AVX512BW-NEXT:    vpmovzxbw {{.*#+}} 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,ymm0[16],zero,ymm0[17],zero,ymm0[18],zero,ymm0[19],zero,ymm0[20],zero,ymm0[21],zero,ymm0[22],zero,ymm0[23],zero,ymm0[24],zero,ymm0[25],zero,ymm0[26],zero,ymm0[27],zero,ymm0[28],zero,ymm0[29],zero,ymm0[30],zero,ymm0[31],zero
+; AVX512BW-NEXT:    vpsrlvw {{.*}}(%rip), %zmm0, %zmm0
+; AVX512BW-NEXT:    vpmovwb %zmm0, %ymm0
+; AVX512BW-NEXT:    retq
   %shift = lshr <32 x i8> %a, <i8 0, i8 1, i8 2, i8 3, i8 4, i8 5, i8 6, i8 7, i8 7, i8 6, i8 5, i8 4, i8 3, i8 2, i8 1, i8 0, i8 0, i8 1, i8 2, i8 3, i8 4, i8 5, i8 6, i8 7, i8 7, i8 6, i8 5, i8 4, i8 3, i8 2, i8 1, i8 0>
   ret <32 x i8> %shift
 }
diff --git a/llvm/test/CodeGen/X86/vector-shift-shl-128.ll b/llvm/test/CodeGen/X86/vector-shift-shl-128.ll
index 2ed619fcc5f..8706078b40c 100644
--- a/llvm/test/CodeGen/X86/vector-shift-shl-128.ll
+++ b/llvm/test/CodeGen/X86/vector-shift-shl-128.ll
@@ -377,17 +377,10 @@ define <16 x i8> @var_shift_v16i8(<16 x i8> %a, <16 x i8> %b) nounwind {
 ;
 ; AVX512-LABEL: var_shift_v16i8:
 ; AVX512:       # BB#0:
-; AVX512-NEXT:    vpsllw $5, %xmm1, %xmm1
-; AVX512-NEXT:    vpsllw $4, %xmm0, %xmm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm2, %xmm0, %xmm0
-; AVX512-NEXT:    vpsllw $2, %xmm0, %xmm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
-; AVX512-NEXT:    vpaddb %xmm1, %xmm1, %xmm1
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm2, %xmm0, %xmm0
-; AVX512-NEXT:    vpaddb %xmm0, %xmm0, %xmm2
-; AVX512-NEXT:    vpaddb %xmm1, %xmm1, %xmm1
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm2, %xmm0, %xmm0
+; AVX512-NEXT:    vpmovzxbd {{.*#+}} zmm1 = xmm1[0],zero,zero,zero,xmm1[1],zero,zero,zero,xmm1[2],zero,zero,zero,xmm1[3],zero,zero,zero,xmm1[4],zero,zero,zero,xmm1[5],zero,zero,zero,xmm1[6],zero,zero,zero,xmm1[7],zero,zero,zero,xmm1[8],zero,zero,zero,xmm1[9],zero,zero,zero,xmm1[10],zero,zero,zero,xmm1[11],zero,zero,zero,xmm1[12],zero,zero,zero,xmm1[13],zero,zero,zero,xmm1[14],zero,zero,zero,xmm1[15],zero,zero,zero
+; AVX512-NEXT:    vpmovzxbd {{.*#+}} zmm0 = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero,xmm0[4],zero,zero,zero,xmm0[5],zero,zero,zero,xmm0[6],zero,zero,zero,xmm0[7],zero,zero,zero,xmm0[8],zero,zero,zero,xmm0[9],zero,zero,zero,xmm0[10],zero,zero,zero,xmm0[11],zero,zero,zero,xmm0[12],zero,zero,zero,xmm0[13],zero,zero,zero,xmm0[14],zero,zero,zero,xmm0[15],zero,zero,zero
+; AVX512-NEXT:    vpsllvd %zmm1, %zmm0, %zmm0
+; AVX512-NEXT:    vpmovdb %zmm0, %xmm0
 ; AVX512-NEXT:    retq
 ;
 ; X32-SSE-LABEL: var_shift_v16i8:
@@ -652,17 +645,10 @@ define <16 x i8> @splatvar_shift_v16i8(<16 x i8> %a, <16 x i8> %b) nounwind {
 ; AVX512-LABEL: splatvar_shift_v16i8:
 ; AVX512:       # BB#0:
 ; AVX512-NEXT:    vpbroadcastb %xmm1, %xmm1
-; AVX512-NEXT:    vpsllw $5, %xmm1, %xmm1
-; AVX512-NEXT:    vpsllw $4, %xmm0, %xmm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm2, %xmm0, %xmm0
-; AVX512-NEXT:    vpsllw $2, %xmm0, %xmm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
-; AVX512-NEXT:    vpaddb %xmm1, %xmm1, %xmm1
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm2, %xmm0, %xmm0
-; AVX512-NEXT:    vpaddb %xmm0, %xmm0, %xmm2
-; AVX512-NEXT:    vpaddb %xmm1, %xmm1, %xmm1
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm2, %xmm0, %xmm0
+; AVX512-NEXT:    vpmovzxbd {{.*#+}} zmm0 = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero,xmm0[4],zero,zero,zero,xmm0[5],zero,zero,zero,xmm0[6],zero,zero,zero,xmm0[7],zero,zero,zero,xmm0[8],zero,zero,zero,xmm0[9],zero,zero,zero,xmm0[10],zero,zero,zero,xmm0[11],zero,zero,zero,xmm0[12],zero,zero,zero,xmm0[13],zero,zero,zero,xmm0[14],zero,zero,zero,xmm0[15],zero,zero,zero
+; AVX512-NEXT:    vpmovzxbd {{.*#+}} zmm1 = xmm1[0],zero,zero,zero,xmm1[1],zero,zero,zero,xmm1[2],zero,zero,zero,xmm1[3],zero,zero,zero,xmm1[4],zero,zero,zero,xmm1[5],zero,zero,zero,xmm1[6],zero,zero,zero,xmm1[7],zero,zero,zero,xmm1[8],zero,zero,zero,xmm1[9],zero,zero,zero,xmm1[10],zero,zero,zero,xmm1[11],zero,zero,zero,xmm1[12],zero,zero,zero,xmm1[13],zero,zero,zero,xmm1[14],zero,zero,zero,xmm1[15],zero,zero,zero
+; AVX512-NEXT:    vpsllvd %zmm1, %zmm0, %zmm0
+; AVX512-NEXT:    vpmovdb %zmm0, %xmm0
 ; AVX512-NEXT:    retq
 ;
 ; X32-SSE-LABEL: splatvar_shift_v16i8:
@@ -934,18 +920,9 @@ define <16 x i8> @constant_shift_v16i8(<16 x i8> %a) nounwind {
 ;
 ; AVX512-LABEL: constant_shift_v16i8:
 ; AVX512:       # BB#0:
-; AVX512-NEXT:    vmovdqa {{.*#+}} xmm1 = [0,1,2,3,4,5,6,7,7,6,5,4,3,2,1,0]
-; AVX512-NEXT:    vpsllw $5, %xmm1, %xmm1
-; AVX512-NEXT:    vpsllw $4, %xmm0, %xmm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm2, %xmm0, %xmm0
-; AVX512-NEXT:    vpsllw $2, %xmm0, %xmm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
-; AVX512-NEXT:    vpaddb %xmm1, %xmm1, %xmm1
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm2, %xmm0, %xmm0
-; AVX512-NEXT:    vpaddb %xmm0, %xmm0, %xmm2
-; AVX512-NEXT:    vpaddb %xmm1, %xmm1, %xmm1
-; AVX512-NEXT:    vpblendvb %xmm1, %xmm2, %xmm0, %xmm0
+; AVX512-NEXT:    vpmovzxbd {{.*#+}} zmm0 = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero,xmm0[4],zero,zero,zero,xmm0[5],zero,zero,zero,xmm0[6],zero,zero,zero,xmm0[7],zero,zero,zero,xmm0[8],zero,zero,zero,xmm0[9],zero,zero,zero,xmm0[10],zero,zero,zero,xmm0[11],zero,zero,zero,xmm0[12],zero,zero,zero,xmm0[13],zero,zero,zero,xmm0[14],zero,zero,zero,xmm0[15],zero,zero,zero
+; AVX512-NEXT:    vpsllvd {{.*}}(%rip), %zmm0, %zmm0
+; AVX512-NEXT:    vpmovdb %zmm0, %xmm0
 ; AVX512-NEXT:    retq
 ;
 ; X32-SSE-LABEL: constant_shift_v16i8:
diff --git a/llvm/test/CodeGen/X86/vector-shift-shl-256.ll b/llvm/test/CodeGen/X86/vector-shift-shl-256.ll
index 44308e03f27..a1ef2791c1b 100644
--- a/llvm/test/CodeGen/X86/vector-shift-shl-256.ll
+++ b/llvm/test/CodeGen/X86/vector-shift-shl-256.ll
@@ -249,20 +249,28 @@ define <32 x i8> @var_shift_v32i8(<32 x i8> %a, <32 x i8> %b) nounwind {
 ; XOPAVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
 ; XOPAVX2-NEXT:    retq
 ;
-; AVX512-LABEL: var_shift_v32i8:
-; AVX512:       # BB#0:
-; AVX512-NEXT:    vpsllw $5, %ymm1, %ymm1
-; AVX512-NEXT:    vpsllw $4, %ymm0, %ymm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
-; AVX512-NEXT:    vpsllw $2, %ymm0, %ymm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
-; AVX512-NEXT:    vpaddb %ymm1, %ymm1, %ymm1
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
-; AVX512-NEXT:    vpaddb %ymm0, %ymm0, %ymm2
-; AVX512-NEXT:    vpaddb %ymm1, %ymm1, %ymm1
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
-; AVX512-NEXT:    retq
+; AVX512DQ-LABEL: var_shift_v32i8:
+; AVX512DQ:       # BB#0:
+; AVX512DQ-NEXT:    vpsllw $5, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpsllw $4, %ymm0, %ymm2
+; AVX512DQ-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
+; AVX512DQ-NEXT:    vpsllw $2, %ymm0, %ymm2
+; AVX512DQ-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
+; AVX512DQ-NEXT:    vpaddb %ymm1, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
+; AVX512DQ-NEXT:    vpaddb %ymm0, %ymm0, %ymm2
+; AVX512DQ-NEXT:    vpaddb %ymm1, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
+; AVX512DQ-NEXT:    retq
+;
+; AVX512BW-LABEL: var_shift_v32i8:
+; AVX512BW:       # BB#0:
+; AVX512BW-NEXT:    vpmovzxbw {{.*#+}} 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,ymm1[16],zero,ymm1[17],zero,ymm1[18],zero,ymm1[19],zero,ymm1[20],zero,ymm1[21],zero,ymm1[22],zero,ymm1[23],zero,ymm1[24],zero,ymm1[25],zero,ymm1[26],zero,ymm1[27],zero,ymm1[28],zero,ymm1[29],zero,ymm1[30],zero,ymm1[31],zero
+; AVX512BW-NEXT:    vpmovzxbw {{.*#+}} 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,ymm0[16],zero,ymm0[17],zero,ymm0[18],zero,ymm0[19],zero,ymm0[20],zero,ymm0[21],zero,ymm0[22],zero,ymm0[23],zero,ymm0[24],zero,ymm0[25],zero,ymm0[26],zero,ymm0[27],zero,ymm0[28],zero,ymm0[29],zero,ymm0[30],zero,ymm0[31],zero
+; AVX512BW-NEXT:    vpsllvw %zmm1, %zmm0, %zmm0
+; AVX512BW-NEXT:    vpmovwb %zmm0, %ymm0
+; AVX512BW-NEXT:    retq
   %shift = shl <32 x i8> %a, %b
   ret <32 x i8> %shift
 }
@@ -455,21 +463,30 @@ define <32 x i8> @splatvar_shift_v32i8(<32 x i8> %a, <32 x i8> %b) nounwind {
 ; XOPAVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
 ; XOPAVX2-NEXT:    retq
 ;
-; AVX512-LABEL: splatvar_shift_v32i8:
-; AVX512:       # BB#0:
-; AVX512-NEXT:    vpbroadcastb %xmm1, %ymm1
-; AVX512-NEXT:    vpsllw $4, %ymm0, %ymm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
-; AVX512-NEXT:    vpsllw $5, %ymm1, %ymm1
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
-; AVX512-NEXT:    vpsllw $2, %ymm0, %ymm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
-; AVX512-NEXT:    vpaddb %ymm1, %ymm1, %ymm1
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
-; AVX512-NEXT:    vpaddb %ymm0, %ymm0, %ymm2
-; AVX512-NEXT:    vpaddb %ymm1, %ymm1, %ymm1
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
-; AVX512-NEXT:    retq
+; AVX512DQ-LABEL: splatvar_shift_v32i8:
+; AVX512DQ:       # BB#0:
+; AVX512DQ-NEXT:    vpbroadcastb %xmm1, %ymm1
+; AVX512DQ-NEXT:    vpsllw $4, %ymm0, %ymm2
+; AVX512DQ-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
+; AVX512DQ-NEXT:    vpsllw $5, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
+; AVX512DQ-NEXT:    vpsllw $2, %ymm0, %ymm2
+; AVX512DQ-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
+; AVX512DQ-NEXT:    vpaddb %ymm1, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
+; AVX512DQ-NEXT:    vpaddb %ymm0, %ymm0, %ymm2
+; AVX512DQ-NEXT:    vpaddb %ymm1, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
+; AVX512DQ-NEXT:    retq
+;
+; AVX512BW-LABEL: splatvar_shift_v32i8:
+; AVX512BW:       # BB#0:
+; AVX512BW-NEXT:    vpbroadcastb %xmm1, %ymm1
+; AVX512BW-NEXT:    vpmovzxbw {{.*#+}} 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,ymm0[16],zero,ymm0[17],zero,ymm0[18],zero,ymm0[19],zero,ymm0[20],zero,ymm0[21],zero,ymm0[22],zero,ymm0[23],zero,ymm0[24],zero,ymm0[25],zero,ymm0[26],zero,ymm0[27],zero,ymm0[28],zero,ymm0[29],zero,ymm0[30],zero,ymm0[31],zero
+; AVX512BW-NEXT:    vpmovzxbw {{.*#+}} 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,ymm1[16],zero,ymm1[17],zero,ymm1[18],zero,ymm1[19],zero,ymm1[20],zero,ymm1[21],zero,ymm1[22],zero,ymm1[23],zero,ymm1[24],zero,ymm1[25],zero,ymm1[26],zero,ymm1[27],zero,ymm1[28],zero,ymm1[29],zero,ymm1[30],zero,ymm1[31],zero
+; AVX512BW-NEXT:    vpsllvw %zmm1, %zmm0, %zmm0
+; AVX512BW-NEXT:    vpmovwb %zmm0, %ymm0
+; AVX512BW-NEXT:    retq
   %splat = shufflevector <32 x i8> %b, <32 x i8> undef, <32 x i32> zeroinitializer
   %shift = shl <32 x i8> %a, %splat
   ret <32 x i8> %shift
@@ -659,21 +676,28 @@ define <32 x i8> @constant_shift_v32i8(<32 x i8> %a) nounwind {
 ; XOPAVX2-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
 ; XOPAVX2-NEXT:    retq
 ;
-; AVX512-LABEL: constant_shift_v32i8:
-; AVX512:       # BB#0:
-; AVX512-NEXT:    vmovdqa {{.*#+}} ymm1 = [0,1,2,3,4,5,6,7,7,6,5,4,3,2,1,0,0,1,2,3,4,5,6,7,7,6,5,4,3,2,1,0]
-; AVX512-NEXT:    vpsllw $5, %ymm1, %ymm1
-; AVX512-NEXT:    vpsllw $4, %ymm0, %ymm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
-; AVX512-NEXT:    vpsllw $2, %ymm0, %ymm2
-; AVX512-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
-; AVX512-NEXT:    vpaddb %ymm1, %ymm1, %ymm1
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
-; AVX512-NEXT:    vpaddb %ymm0, %ymm0, %ymm2
-; AVX512-NEXT:    vpaddb %ymm1, %ymm1, %ymm1
-; AVX512-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
-; AVX512-NEXT:    retq
+; AVX512DQ-LABEL: constant_shift_v32i8:
+; AVX512DQ:       # BB#0:
+; AVX512DQ-NEXT:    vmovdqa {{.*#+}} ymm1 = [0,1,2,3,4,5,6,7,7,6,5,4,3,2,1,0,0,1,2,3,4,5,6,7,7,6,5,4,3,2,1,0]
+; AVX512DQ-NEXT:    vpsllw $5, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpsllw $4, %ymm0, %ymm2
+; AVX512DQ-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
+; AVX512DQ-NEXT:    vpsllw $2, %ymm0, %ymm2
+; AVX512DQ-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
+; AVX512DQ-NEXT:    vpaddb %ymm1, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
+; AVX512DQ-NEXT:    vpaddb %ymm0, %ymm0, %ymm2
+; AVX512DQ-NEXT:    vpaddb %ymm1, %ymm1, %ymm1
+; AVX512DQ-NEXT:    vpblendvb %ymm1, %ymm2, %ymm0, %ymm0
+; AVX512DQ-NEXT:    retq
+;
+; AVX512BW-LABEL: constant_shift_v32i8:
+; AVX512BW:       # BB#0:
+; AVX512BW-NEXT:    vpmovzxbw {{.*#+}} 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,ymm0[16],zero,ymm0[17],zero,ymm0[18],zero,ymm0[19],zero,ymm0[20],zero,ymm0[21],zero,ymm0[22],zero,ymm0[23],zero,ymm0[24],zero,ymm0[25],zero,ymm0[26],zero,ymm0[27],zero,ymm0[28],zero,ymm0[29],zero,ymm0[30],zero,ymm0[31],zero
+; AVX512BW-NEXT:    vpsllvw {{.*}}(%rip), %zmm0, %zmm0
+; AVX512BW-NEXT:    vpmovwb %zmm0, %ymm0
+; AVX512BW-NEXT:    retq
   %shift = shl <32 x i8> %a, <i8 0, i8 1, i8 2, i8 3, i8 4, i8 5, i8 6, i8 7, i8 7, i8 6, i8 5, i8 4, i8 3, i8 2, i8 1, i8 0, i8 0, i8 1, i8 2, i8 3, i8 4, i8 5, i8 6, i8 7, i8 7, i8 6, i8 5, i8 4, i8 3, i8 2, i8 1, i8 0>
   ret <32 x i8> %shift
 }