; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc < %s -disable-peephole -mtriple=x86_64-unknown-unknown | FileCheck %s --check-prefixes=SSE,SSE2
; RUN: llc < %s -disable-peephole -mtriple=x86_64-unknown-unknown -mattr=+sse3 | FileCheck %s --check-prefixes=SSE,SSE3
; RUN: llc < %s -disable-peephole -mtriple=x86_64-unknown-unknown -mattr=+ssse3 | FileCheck %s --check-prefixes=SSE,SSSE3
; RUN: llc < %s -disable-peephole -mtriple=x86_64-unknown-unknown -mattr=+sse4.1 | FileCheck %s --check-prefixes=SSE,SSE41
; RUN: llc < %s -disable-peephole -mtriple=x86_64-unknown-unknown -mattr=+avx | FileCheck %s --check-prefixes=AVX,AVX1OR2,AVX1
; RUN: llc < %s -disable-peephole -mtriple=x86_64-unknown-unknown -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1OR2,AVX2OR512VL,AVX2,AVX2-SLOW
; RUN: llc < %s -disable-peephole -mtriple=x86_64-unknown-unknown -mattr=+avx2,+fast-variable-shuffle | FileCheck %s --check-prefixes=AVX,AVX1OR2,AVX2OR512VL,AVX2,AVX2-FAST
; RUN: llc < %s -disable-peephole -mtriple=x86_64-unknown-unknown -mattr=+avx512vl,+avx512dq,+fast-variable-shuffle | FileCheck %s --check-prefixes=AVX,AVX2OR512VL,AVX512VL

define <4 x i32> @shuffle_v4i32_0001(<4 x i32> %a, <4 x i32> %b) {
; SSE-LABEL: shuffle_v4i32_0001:
; SSE:       # %bb.0:
; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,0,0,1]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_0001:
; AVX:       # %bb.0:
; AVX-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[0,0,0,1]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 0, i32 0, i32 0, i32 1>
  ret <4 x i32> %shuffle
}
define <4 x i32> @shuffle_v4i32_0020(<4 x i32> %a, <4 x i32> %b) {
; SSE-LABEL: shuffle_v4i32_0020:
; SSE:       # %bb.0:
; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,0,2,0]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_0020:
; AVX:       # %bb.0:
; AVX-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[0,0,2,0]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 0, i32 0, i32 2, i32 0>
  ret <4 x i32> %shuffle
}
define <4 x i32> @shuffle_v4i32_0112(<4 x i32> %a, <4 x i32> %b) {
; SSE-LABEL: shuffle_v4i32_0112:
; SSE:       # %bb.0:
; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,1,1,2]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_0112:
; AVX:       # %bb.0:
; AVX-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[0,1,1,2]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 0, i32 1, i32 1, i32 2>
  ret <4 x i32> %shuffle
}
define <4 x i32> @shuffle_v4i32_0300(<4 x i32> %a, <4 x i32> %b) {
; SSE-LABEL: shuffle_v4i32_0300:
; SSE:       # %bb.0:
; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,3,0,0]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_0300:
; AVX:       # %bb.0:
; AVX-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[0,3,0,0]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 0, i32 3, i32 0, i32 0>
  ret <4 x i32> %shuffle
}
define <4 x i32> @shuffle_v4i32_1000(<4 x i32> %a, <4 x i32> %b) {
; SSE-LABEL: shuffle_v4i32_1000:
; SSE:       # %bb.0:
; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[1,0,0,0]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_1000:
; AVX:       # %bb.0:
; AVX-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[1,0,0,0]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 1, i32 0, i32 0, i32 0>
  ret <4 x i32> %shuffle
}
define <4 x i32> @shuffle_v4i32_2200(<4 x i32> %a, <4 x i32> %b) {
; SSE-LABEL: shuffle_v4i32_2200:
; SSE:       # %bb.0:
; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,2,0,0]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_2200:
; AVX:       # %bb.0:
; AVX-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[2,2,0,0]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 2, i32 2, i32 0, i32 0>
  ret <4 x i32> %shuffle
}
define <4 x i32> @shuffle_v4i32_3330(<4 x i32> %a, <4 x i32> %b) {
; SSE-LABEL: shuffle_v4i32_3330:
; SSE:       # %bb.0:
; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[3,3,3,0]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_3330:
; AVX:       # %bb.0:
; AVX-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[3,3,3,0]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 3, i32 3, i32 3, i32 0>
  ret <4 x i32> %shuffle
}
define <4 x i32> @shuffle_v4i32_3210(<4 x i32> %a, <4 x i32> %b) {
; SSE-LABEL: shuffle_v4i32_3210:
; SSE:       # %bb.0:
; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[3,2,1,0]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_3210:
; AVX:       # %bb.0:
; AVX-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[3,2,1,0]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
  ret <4 x i32> %shuffle
}

define <4 x i32> @shuffle_v4i32_2121(<4 x i32> %a, <4 x i32> %b) {
; SSE-LABEL: shuffle_v4i32_2121:
; SSE:       # %bb.0:
; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,1,2,1]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_2121:
; AVX:       # %bb.0:
; AVX-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[2,1,2,1]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 2, i32 1, i32 2, i32 1>
  ret <4 x i32> %shuffle
}

define <4 x float> @shuffle_v4f32_0001(<4 x float> %a, <4 x float> %b) {
; SSE-LABEL: shuffle_v4f32_0001:
; SSE:       # %bb.0:
; SSE-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,0,0,1]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_0001:
; AVX:       # %bb.0:
; AVX-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[0,0,0,1]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 0, i32 0, i32 1>
  ret <4 x float> %shuffle
}
define <4 x float> @shuffle_v4f32_0020(<4 x float> %a, <4 x float> %b) {
; SSE-LABEL: shuffle_v4f32_0020:
; SSE:       # %bb.0:
; SSE-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,0,2,0]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_0020:
; AVX:       # %bb.0:
; AVX-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[0,0,2,0]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 0, i32 2, i32 0>
  ret <4 x float> %shuffle
}
define <4 x float> @shuffle_v4f32_0300(<4 x float> %a, <4 x float> %b) {
; SSE-LABEL: shuffle_v4f32_0300:
; SSE:       # %bb.0:
; SSE-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,3,0,0]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_0300:
; AVX:       # %bb.0:
; AVX-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[0,3,0,0]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 3, i32 0, i32 0>
  ret <4 x float> %shuffle
}
define <4 x float> @shuffle_v4f32_1000(<4 x float> %a, <4 x float> %b) {
; SSE-LABEL: shuffle_v4f32_1000:
; SSE:       # %bb.0:
; SSE-NEXT:    shufps {{.*#+}} xmm0 = xmm0[1,0,0,0]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_1000:
; AVX:       # %bb.0:
; AVX-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[1,0,0,0]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 1, i32 0, i32 0, i32 0>
  ret <4 x float> %shuffle
}
define <4 x float> @shuffle_v4f32_2200(<4 x float> %a, <4 x float> %b) {
; SSE-LABEL: shuffle_v4f32_2200:
; SSE:       # %bb.0:
; SSE-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,2,0,0]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_2200:
; AVX:       # %bb.0:
; AVX-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[2,2,0,0]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 2, i32 2, i32 0, i32 0>
  ret <4 x float> %shuffle
}
define <4 x float> @shuffle_v4f32_3330(<4 x float> %a, <4 x float> %b) {
; SSE-LABEL: shuffle_v4f32_3330:
; SSE:       # %bb.0:
; SSE-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,3,3,0]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_3330:
; AVX:       # %bb.0:
; AVX-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[3,3,3,0]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 3, i32 3, i32 3, i32 0>
  ret <4 x float> %shuffle
}
define <4 x float> @shuffle_v4f32_3210(<4 x float> %a, <4 x float> %b) {
; SSE-LABEL: shuffle_v4f32_3210:
; SSE:       # %bb.0:
; SSE-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,2,1,0]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_3210:
; AVX:       # %bb.0:
; AVX-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[3,2,1,0]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
  ret <4 x float> %shuffle
}
define <4 x float> @shuffle_v4f32_0011(<4 x float> %a, <4 x float> %b) {
; SSE-LABEL: shuffle_v4f32_0011:
; SSE:       # %bb.0:
; SSE-NEXT:    unpcklps {{.*#+}} xmm0 = xmm0[0,0,1,1]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_0011:
; AVX:       # %bb.0:
; AVX-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[0,0,1,1]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 0, i32 1, i32 1>
  ret <4 x float> %shuffle
}
define <4 x float> @shuffle_v4f32_2233(<4 x float> %a, <4 x float> %b) {
; SSE-LABEL: shuffle_v4f32_2233:
; SSE:       # %bb.0:
; SSE-NEXT:    unpckhps {{.*#+}} xmm0 = xmm0[2,2,3,3]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_2233:
; AVX:       # %bb.0:
; AVX-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[2,2,3,3]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 2, i32 2, i32 3, i32 3>
  ret <4 x float> %shuffle
}
define <4 x float> @shuffle_v4f32_0022(<4 x float> %a, <4 x float> %b) {
; SSE2-LABEL: shuffle_v4f32_0022:
; SSE2:       # %bb.0:
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,0,2,2]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4f32_0022:
; SSE3:       # %bb.0:
; SSE3-NEXT:    movsldup {{.*#+}} xmm0 = xmm0[0,0,2,2]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4f32_0022:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    movsldup {{.*#+}} xmm0 = xmm0[0,0,2,2]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4f32_0022:
; SSE41:       # %bb.0:
; SSE41-NEXT:    movsldup {{.*#+}} xmm0 = xmm0[0,0,2,2]
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_0022:
; AVX:       # %bb.0:
; AVX-NEXT:    vmovsldup {{.*#+}} xmm0 = xmm0[0,0,2,2]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 0, i32 2, i32 2>
  ret <4 x float> %shuffle
}
define <4 x float> @shuffle_v4f32_1133(<4 x float> %a, <4 x float> %b) {
; SSE2-LABEL: shuffle_v4f32_1133:
; SSE2:       # %bb.0:
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[1,1,3,3]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4f32_1133:
; SSE3:       # %bb.0:
; SSE3-NEXT:    movshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4f32_1133:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    movshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4f32_1133:
; SSE41:       # %bb.0:
; SSE41-NEXT:    movshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_1133:
; AVX:       # %bb.0:
; AVX-NEXT:    vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 1, i32 1, i32 3, i32 3>
  ret <4 x float> %shuffle
}

define <4 x float> @shuffle_v4f32_0145(<4 x float> %a, <4 x float> %b) {
; SSE-LABEL: shuffle_v4f32_0145:
; SSE:       # %bb.0:
; SSE-NEXT:    movlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_0145:
; AVX:       # %bb.0:
; AVX-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 1, i32 4, i32 5>
  ret <4 x float> %shuffle
}

define <4 x float> @shuffle_v4f32_6723(<4 x float> %a, <4 x float> %b) {
; SSE-LABEL: shuffle_v4f32_6723:
; SSE:       # %bb.0:
; SSE-NEXT:    movhlps {{.*#+}} xmm0 = xmm1[1],xmm0[1]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_6723:
; AVX:       # %bb.0:
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm0 = xmm1[1],xmm0[1]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 6, i32 7, i32 2, i32 3>
  ret <4 x float> %shuffle
}

define <4 x i32> @shuffle_v4i32_0124(<4 x i32> %a, <4 x i32> %b) {
; SSE2-LABEL: shuffle_v4i32_0124:
; SSE2:       # %bb.0:
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,1],xmm0[2,3]
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,0]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4i32_0124:
; SSE3:       # %bb.0:
; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,1],xmm0[2,3]
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,0]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4i32_0124:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,1],xmm0[2,3]
; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,0]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4i32_0124:
; SSE41:       # %bb.0:
; SSE41-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[0,0,0,0]
; SSE41-NEXT:    pblendw {{.*#+}} xmm0 = xmm0[0,1,2,3,4,5],xmm1[6,7]
; SSE41-NEXT:    retq
;
; AVX1-LABEL: shuffle_v4i32_0124:
; AVX1:       # %bb.0:
; AVX1-NEXT:    vpermilps {{.*#+}} xmm1 = xmm1[0,0,0,0]
; AVX1-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0,1,2],xmm1[3]
; AVX1-NEXT:    retq
;
; AVX2-LABEL: shuffle_v4i32_0124:
; AVX2:       # %bb.0:
; AVX2-NEXT:    vbroadcastss %xmm1, %xmm1
; AVX2-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0,1,2],xmm1[3]
; AVX2-NEXT:    retq
;
; AVX512VL-LABEL: shuffle_v4i32_0124:
; AVX512VL:       # %bb.0:
; AVX512VL-NEXT:    vmovdqa {{.*#+}} xmm2 = [0,1,2,4]
; AVX512VL-NEXT:    vpermt2d %xmm1, %xmm2, %xmm0
; AVX512VL-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 0, i32 1, i32 2, i32 4>
  ret <4 x i32> %shuffle
}
define <4 x i32> @shuffle_v4i32_0142(<4 x i32> %a, <4 x i32> %b) {
; SSE2-LABEL: shuffle_v4i32_0142:
; SSE2:       # %bb.0:
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,1],xmm0[2,3]
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[0,2]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4i32_0142:
; SSE3:       # %bb.0:
; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,1],xmm0[2,3]
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[0,2]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4i32_0142:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,1],xmm0[2,3]
; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[0,2]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4i32_0142:
; SSE41:       # %bb.0:
; SSE41-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[0,1,0,1]
; SSE41-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,1,2,2]
; SSE41-NEXT:    pblendw {{.*#+}} xmm0 = xmm0[0,1,2,3],xmm1[4,5],xmm0[6,7]
; SSE41-NEXT:    retq
;
; AVX1-LABEL: shuffle_v4i32_0142:
; AVX1:       # %bb.0:
; AVX1-NEXT:    vpermilps {{.*#+}} xmm1 = xmm1[0,1,0,1]
; AVX1-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[0,1,2,2]
; AVX1-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2],xmm0[3]
; AVX1-NEXT:    retq
;
; AVX2-LABEL: shuffle_v4i32_0142:
; AVX2:       # %bb.0:
; AVX2-NEXT:    vbroadcastss %xmm1, %xmm1
; AVX2-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[0,1,2,2]
; AVX2-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2],xmm0[3]
; AVX2-NEXT:    retq
;
; AVX512VL-LABEL: shuffle_v4i32_0142:
; AVX512VL:       # %bb.0:
; AVX512VL-NEXT:    vmovdqa {{.*#+}} xmm2 = [0,1,4,2]
; AVX512VL-NEXT:    vpermt2d %xmm1, %xmm2, %xmm0
; AVX512VL-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 0, i32 1, i32 4, i32 2>
  ret <4 x i32> %shuffle
}
define <4 x i32> @shuffle_v4i32_0412(<4 x i32> %a, <4 x i32> %b) {
; SSE2-LABEL: shuffle_v4i32_0412:
; SSE2:       # %bb.0:
; SSE2-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[2,0],xmm0[1,2]
; SSE2-NEXT:    movaps %xmm1, %xmm0
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4i32_0412:
; SSE3:       # %bb.0:
; SSE3-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[2,0],xmm0[1,2]
; SSE3-NEXT:    movaps %xmm1, %xmm0
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4i32_0412:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[2,0],xmm0[1,2]
; SSSE3-NEXT:    movaps %xmm1, %xmm0
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4i32_0412:
; SSE41:       # %bb.0:
; SSE41-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[0,0,1,1]
; SSE41-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,1,1,2]
; SSE41-NEXT:    pblendw {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,3],xmm0[4,5,6,7]
; SSE41-NEXT:    retq
;
; AVX1-LABEL: shuffle_v4i32_0412:
; AVX1:       # %bb.0:
; AVX1-NEXT:    vpermilps {{.*#+}} xmm1 = xmm1[0,0,1,1]
; AVX1-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[0,1,1,2]
; AVX1-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2,3]
; AVX1-NEXT:    retq
;
; AVX2-LABEL: shuffle_v4i32_0412:
; AVX2:       # %bb.0:
; AVX2-NEXT:    vbroadcastss %xmm1, %xmm1
; AVX2-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[0,1,1,2]
; AVX2-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2,3]
; AVX2-NEXT:    retq
;
; AVX512VL-LABEL: shuffle_v4i32_0412:
; AVX512VL:       # %bb.0:
; AVX512VL-NEXT:    vmovdqa {{.*#+}} xmm2 = [0,4,1,2]
; AVX512VL-NEXT:    vpermt2d %xmm1, %xmm2, %xmm0
; AVX512VL-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 2>
  ret <4 x i32> %shuffle
}
define <4 x i32> @shuffle_v4i32_4012(<4 x i32> %a, <4 x i32> %b) {
; SSE2-LABEL: shuffle_v4i32_4012:
; SSE2:       # %bb.0:
; SSE2-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,2],xmm0[1,2]
; SSE2-NEXT:    movaps %xmm1, %xmm0
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4i32_4012:
; SSE3:       # %bb.0:
; SSE3-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,2],xmm0[1,2]
; SSE3-NEXT:    movaps %xmm1, %xmm0
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4i32_4012:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,2],xmm0[1,2]
; SSSE3-NEXT:    movaps %xmm1, %xmm0
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4i32_4012:
; SSE41:       # %bb.0:
; SSE41-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,0,1,2]
; SSE41-NEXT:    pblendw {{.*#+}} xmm0 = xmm1[0,1],xmm0[2,3,4,5,6,7]
; SSE41-NEXT:    retq
;
; AVX1OR2-LABEL: shuffle_v4i32_4012:
; AVX1OR2:       # %bb.0:
; AVX1OR2-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[0,0,1,2]
; AVX1OR2-NEXT:    vblendps {{.*#+}} xmm0 = xmm1[0],xmm0[1,2,3]
; AVX1OR2-NEXT:    retq
;
; AVX512VL-LABEL: shuffle_v4i32_4012:
; AVX512VL:       # %bb.0:
; AVX512VL-NEXT:    vmovdqa {{.*#+}} xmm2 = [4,0,1,2]
; AVX512VL-NEXT:    vpermt2d %xmm1, %xmm2, %xmm0
; AVX512VL-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 4, i32 0, i32 1, i32 2>
  ret <4 x i32> %shuffle
}
define <4 x i32> @shuffle_v4i32_0145(<4 x i32> %a, <4 x i32> %b) {
; SSE-LABEL: shuffle_v4i32_0145:
; SSE:       # %bb.0:
; SSE-NEXT:    movlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_0145:
; AVX:       # %bb.0:
; AVX-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 0, i32 1, i32 4, i32 5>
  ret <4 x i32> %shuffle
}
define <4 x i32> @shuffle_v4i32_0451(<4 x i32> %a, <4 x i32> %b) {
; SSE2-LABEL: shuffle_v4i32_0451:
; SSE2:       # %bb.0:
; SSE2-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,1,3,2]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4i32_0451:
; SSE3:       # %bb.0:
; SSE3-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
; SSE3-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,1,3,2]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4i32_0451:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
; SSSE3-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,1,3,2]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4i32_0451:
; SSE41:       # %bb.0:
; SSE41-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[0,0,1,1]
; SSE41-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,1,0,1]
; SSE41-NEXT:    pblendw {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,3,4,5],xmm0[6,7]
; SSE41-NEXT:    retq
;
; AVX1-LABEL: shuffle_v4i32_0451:
; AVX1:       # %bb.0:
; AVX1-NEXT:    vpermilps {{.*#+}} xmm1 = xmm1[0,0,1,1]
; AVX1-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[0,1,0,1]
; AVX1-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0],xmm1[1,2],xmm0[3]
; AVX1-NEXT:    retq
;
; AVX2-LABEL: shuffle_v4i32_0451:
; AVX2:       # %bb.0:
; AVX2-NEXT:    vpermilps {{.*#+}} xmm1 = xmm1[0,0,1,1]
; AVX2-NEXT:    vmovddup {{.*#+}} xmm0 = xmm0[0,0]
; AVX2-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0],xmm1[1,2],xmm0[3]
; AVX2-NEXT:    retq
;
; AVX512VL-LABEL: shuffle_v4i32_0451:
; AVX512VL:       # %bb.0:
; AVX512VL-NEXT:    vmovdqa {{.*#+}} xmm2 = [0,4,5,1]
; AVX512VL-NEXT:    vpermt2d %xmm1, %xmm2, %xmm0
; AVX512VL-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 0, i32 4, i32 5, i32 1>
  ret <4 x i32> %shuffle
}
define <4 x i32> @shuffle_v4i32_4501(<4 x i32> %a, <4 x i32> %b) {
; SSE-LABEL: shuffle_v4i32_4501:
; SSE:       # %bb.0:
; SSE-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE-NEXT:    movaps %xmm1, %xmm0
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_4501:
; AVX:       # %bb.0:
; AVX-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm1[0],xmm0[0]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 4, i32 5, i32 0, i32 1>
  ret <4 x i32> %shuffle
}
define <4 x i32> @shuffle_v4i32_4015(<4 x i32> %a, <4 x i32> %b) {
; SSE2-LABEL: shuffle_v4i32_4015:
; SSE2:       # %bb.0:
; SSE2-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[1,0,2,3]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4i32_4015:
; SSE3:       # %bb.0:
; SSE3-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
; SSE3-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[1,0,2,3]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4i32_4015:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
; SSSE3-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[1,0,2,3]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4i32_4015:
; SSE41:       # %bb.0:
; SSE41-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[0,1,0,1]
; SSE41-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,0,1,1]
; SSE41-NEXT:    pblendw {{.*#+}} xmm0 = xmm1[0,1],xmm0[2,3,4,5],xmm1[6,7]
; SSE41-NEXT:    retq
;
; AVX1-LABEL: shuffle_v4i32_4015:
; AVX1:       # %bb.0:
; AVX1-NEXT:    vpermilps {{.*#+}} xmm1 = xmm1[0,1,0,1]
; AVX1-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[0,0,1,1]
; AVX1-NEXT:    vblendps {{.*#+}} xmm0 = xmm1[0],xmm0[1,2],xmm1[3]
; AVX1-NEXT:    retq
;
; AVX2-LABEL: shuffle_v4i32_4015:
; AVX2:       # %bb.0:
; AVX2-NEXT:    vmovddup {{.*#+}} xmm1 = xmm1[0,0]
; AVX2-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[0,0,1,1]
; AVX2-NEXT:    vblendps {{.*#+}} xmm0 = xmm1[0],xmm0[1,2],xmm1[3]
; AVX2-NEXT:    retq
;
; AVX512VL-LABEL: shuffle_v4i32_4015:
; AVX512VL:       # %bb.0:
; AVX512VL-NEXT:    vmovdqa {{.*#+}} xmm2 = [4,0,1,5]
; AVX512VL-NEXT:    vpermt2d %xmm1, %xmm2, %xmm0
; AVX512VL-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 4, i32 0, i32 1, i32 5>
  ret <4 x i32> %shuffle
}

define <4 x float> @shuffle_v4f32_4zzz(<4 x float> %a) {
; SSE2-LABEL: shuffle_v4f32_4zzz:
; SSE2:       # %bb.0:
; SSE2-NEXT:    xorps %xmm1, %xmm1
; SSE2-NEXT:    movss {{.*#+}} xmm1 = xmm0[0],xmm1[1,2,3]
; SSE2-NEXT:    movaps %xmm1, %xmm0
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4f32_4zzz:
; SSE3:       # %bb.0:
; SSE3-NEXT:    xorps %xmm1, %xmm1
; SSE3-NEXT:    movss {{.*#+}} xmm1 = xmm0[0],xmm1[1,2,3]
; SSE3-NEXT:    movaps %xmm1, %xmm0
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4f32_4zzz:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    xorps %xmm1, %xmm1
; SSSE3-NEXT:    movss {{.*#+}} xmm1 = xmm0[0],xmm1[1,2,3]
; SSSE3-NEXT:    movaps %xmm1, %xmm0
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4f32_4zzz:
; SSE41:       # %bb.0:
; SSE41-NEXT:    xorps %xmm1, %xmm1
; SSE41-NEXT:    blendps {{.*#+}} xmm0 = xmm0[0],xmm1[1,2,3]
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_4zzz:
; AVX:       # %bb.0:
; AVX-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; AVX-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0],xmm1[1,2,3]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> zeroinitializer, <4 x float> %a, <4 x i32> <i32 4, i32 1, i32 2, i32 3>
  ret <4 x float> %shuffle
}

define <4 x float> @shuffle_v4f32_z4zz(<4 x float> %a) {
; SSE2-LABEL: shuffle_v4f32_z4zz:
; SSE2:       # %bb.0:
; SSE2-NEXT:    movq {{.*#+}} xmm0 = xmm0[0],zero
; SSE2-NEXT:    xorps %xmm1, %xmm1
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,0],xmm1[2,3]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4f32_z4zz:
; SSE3:       # %bb.0:
; SSE3-NEXT:    movq {{.*#+}} xmm0 = xmm0[0],zero
; SSE3-NEXT:    xorps %xmm1, %xmm1
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,0],xmm1[2,3]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4f32_z4zz:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    movq {{.*#+}} xmm0 = xmm0[0],zero
; SSSE3-NEXT:    xorps %xmm1, %xmm1
; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,0],xmm1[2,3]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4f32_z4zz:
; SSE41:       # %bb.0:
; SSE41-NEXT:    insertps {{.*#+}} xmm0 = zero,xmm0[0],zero,zero
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_z4zz:
; AVX:       # %bb.0:
; AVX-NEXT:    vinsertps {{.*#+}} xmm0 = zero,xmm0[0],zero,zero
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> zeroinitializer, <4 x float> %a, <4 x i32> <i32 2, i32 4, i32 3, i32 0>
  ret <4 x float> %shuffle
}

define <4 x float> @shuffle_v4f32_zz4z(<4 x float> %a) {
; SSE2-LABEL: shuffle_v4f32_zz4z:
; SSE2:       # %bb.0:
; SSE2-NEXT:    movq {{.*#+}} xmm1 = xmm0[0],zero
; SSE2-NEXT:    pxor %xmm0, %xmm0
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[0,2]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4f32_zz4z:
; SSE3:       # %bb.0:
; SSE3-NEXT:    movq {{.*#+}} xmm1 = xmm0[0],zero
; SSE3-NEXT:    pxor %xmm0, %xmm0
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[0,2]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4f32_zz4z:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    movq {{.*#+}} xmm1 = xmm0[0],zero
; SSSE3-NEXT:    pxor %xmm0, %xmm0
; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[0,2]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4f32_zz4z:
; SSE41:       # %bb.0:
; SSE41-NEXT:    insertps {{.*#+}} xmm0 = zero,zero,xmm0[0],zero
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_zz4z:
; AVX:       # %bb.0:
; AVX-NEXT:    vinsertps {{.*#+}} xmm0 = zero,zero,xmm0[0],zero
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> zeroinitializer, <4 x float> %a, <4 x i32> <i32 0, i32 0, i32 4, i32 0>
  ret <4 x float> %shuffle
}

define <4 x float> @shuffle_v4f32_zuu4(<4 x float> %a) {
; SSE2-LABEL: shuffle_v4f32_zuu4:
; SSE2:       # %bb.0:
; SSE2-NEXT:    xorps %xmm1, %xmm1
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,0],xmm0[0,0]
; SSE2-NEXT:    movaps %xmm1, %xmm0
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4f32_zuu4:
; SSE3:       # %bb.0:
; SSE3-NEXT:    xorps %xmm1, %xmm1
; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,0],xmm0[0,0]
; SSE3-NEXT:    movaps %xmm1, %xmm0
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4f32_zuu4:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    xorps %xmm1, %xmm1
; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,0],xmm0[0,0]
; SSSE3-NEXT:    movaps %xmm1, %xmm0
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4f32_zuu4:
; SSE41:       # %bb.0:
; SSE41-NEXT:    insertps {{.*#+}} xmm0 = zero,zero,zero,xmm0[0]
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_zuu4:
; AVX:       # %bb.0:
; AVX-NEXT:    vinsertps {{.*#+}} xmm0 = zero,zero,zero,xmm0[0]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> zeroinitializer, <4 x float> %a, <4 x i32> <i32 0, i32 undef, i32 undef, i32 4>
  ret <4 x float> %shuffle
}

define <4 x float> @shuffle_v4f32_zzz7(<4 x float> %a) {
; SSE2-LABEL: shuffle_v4f32_zzz7:
; SSE2:       # %bb.0:
; SSE2-NEXT:    xorps %xmm1, %xmm1
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,0],xmm1[2,0]
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,1],xmm0[2,0]
; SSE2-NEXT:    movaps %xmm1, %xmm0
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4f32_zzz7:
; SSE3:       # %bb.0:
; SSE3-NEXT:    xorps %xmm1, %xmm1
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,0],xmm1[2,0]
; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,1],xmm0[2,0]
; SSE3-NEXT:    movaps %xmm1, %xmm0
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4f32_zzz7:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    xorps %xmm1, %xmm1
; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,0],xmm1[2,0]
; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,1],xmm0[2,0]
; SSSE3-NEXT:    movaps %xmm1, %xmm0
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4f32_zzz7:
; SSE41:       # %bb.0:
; SSE41-NEXT:    xorps %xmm1, %xmm1
; SSE41-NEXT:    blendps {{.*#+}} xmm0 = xmm1[0,1,2],xmm0[3]
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_zzz7:
; AVX:       # %bb.0:
; AVX-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; AVX-NEXT:    vblendps {{.*#+}} xmm0 = xmm1[0,1,2],xmm0[3]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> zeroinitializer, <4 x float> %a, <4 x i32> <i32 0, i32 1, i32 2, i32 7>
  ret <4 x float> %shuffle
}

define <4 x float> @shuffle_v4f32_z6zz(<4 x float> %a) {
; SSE2-LABEL: shuffle_v4f32_z6zz:
; SSE2:       # %bb.0:
; SSE2-NEXT:    xorps %xmm1, %xmm1
; SSE2-NEXT:    unpckhpd {{.*#+}} xmm0 = xmm0[1],xmm1[1]
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,0],xmm1[2,3]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4f32_z6zz:
; SSE3:       # %bb.0:
; SSE3-NEXT:    xorps %xmm1, %xmm1
; SSE3-NEXT:    unpckhpd {{.*#+}} xmm0 = xmm0[1],xmm1[1]
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,0],xmm1[2,3]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4f32_z6zz:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    xorps %xmm1, %xmm1
; SSSE3-NEXT:    unpckhpd {{.*#+}} xmm0 = xmm0[1],xmm1[1]
; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,0],xmm1[2,3]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4f32_z6zz:
; SSE41:       # %bb.0:
; SSE41-NEXT:    insertps {{.*#+}} xmm0 = zero,xmm0[2],zero,zero
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_z6zz:
; AVX:       # %bb.0:
; AVX-NEXT:    vinsertps {{.*#+}} xmm0 = zero,xmm0[2],zero,zero
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> zeroinitializer, <4 x float> %a, <4 x i32> <i32 0, i32 6, i32 2, i32 3>
  ret <4 x float> %shuffle
}

define <4 x float> @shuffle_v4f32_0z23(<4 x float> %a) {
; SSE2-LABEL: shuffle_v4f32_0z23:
; SSE2:       # %bb.0:
; SSE2-NEXT:    xorps %xmm1, %xmm1
; SSE2-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[2,0],xmm0[2,3]
; SSE2-NEXT:    movaps %xmm1, %xmm0
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4f32_0z23:
; SSE3:       # %bb.0:
; SSE3-NEXT:    xorps %xmm1, %xmm1
; SSE3-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[2,0],xmm0[2,3]
; SSE3-NEXT:    movaps %xmm1, %xmm0
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4f32_0z23:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    xorps %xmm1, %xmm1
; SSSE3-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[2,0],xmm0[2,3]
; SSSE3-NEXT:    movaps %xmm1, %xmm0
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4f32_0z23:
; SSE41:       # %bb.0:
; SSE41-NEXT:    xorps %xmm1, %xmm1
; SSE41-NEXT:    blendps {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2,3]
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_0z23:
; AVX:       # %bb.0:
; AVX-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; AVX-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2,3]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> %a, <4 x float> zeroinitializer, <4 x i32> <i32 0, i32 4, i32 2, i32 3>
  ret <4 x float> %shuffle
}

define <4 x float> @shuffle_v4f32_01z3(<4 x float> %a) {
; SSE2-LABEL: shuffle_v4f32_01z3:
; SSE2:       # %bb.0:
; SSE2-NEXT:    xorps %xmm1, %xmm1
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[2,0],xmm0[3,0]
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[0,2]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4f32_01z3:
; SSE3:       # %bb.0:
; SSE3-NEXT:    xorps %xmm1, %xmm1
; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[2,0],xmm0[3,0]
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[0,2]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4f32_01z3:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    xorps %xmm1, %xmm1
; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[2,0],xmm0[3,0]
; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[0,2]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4f32_01z3:
; SSE41:       # %bb.0:
; SSE41-NEXT:    xorps %xmm1, %xmm1
; SSE41-NEXT:    blendps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2],xmm0[3]
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_01z3:
; AVX:       # %bb.0:
; AVX-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; AVX-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2],xmm0[3]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> %a, <4 x float> zeroinitializer, <4 x i32> <i32 0, i32 1, i32 4, i32 3>
  ret <4 x float> %shuffle
}

define <4 x float> @shuffle_v4f32_012z(<4 x float> %a) {
; SSE2-LABEL: shuffle_v4f32_012z:
; SSE2:       # %bb.0:
; SSE2-NEXT:    xorps %xmm1, %xmm1
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[3,0],xmm0[2,0]
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,0]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4f32_012z:
; SSE3:       # %bb.0:
; SSE3-NEXT:    xorps %xmm1, %xmm1
; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[3,0],xmm0[2,0]
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,0]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4f32_012z:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    xorps %xmm1, %xmm1
; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[3,0],xmm0[2,0]
; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,0]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4f32_012z:
; SSE41:       # %bb.0:
; SSE41-NEXT:    xorps %xmm1, %xmm1
; SSE41-NEXT:    blendps {{.*#+}} xmm0 = xmm0[0,1,2],xmm1[3]
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_012z:
; AVX:       # %bb.0:
; AVX-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; AVX-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0,1,2],xmm1[3]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> %a, <4 x float> zeroinitializer, <4 x i32> <i32 0, i32 1, i32 2, i32 7>
  ret <4 x float> %shuffle
}

define <4 x float> @shuffle_v4f32_0zz3(<4 x float> %a) {
; SSE2-LABEL: shuffle_v4f32_0zz3:
; SSE2:       # %bb.0:
; SSE2-NEXT:    xorps %xmm1, %xmm1
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,3],xmm1[1,2]
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,2,3,1]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4f32_0zz3:
; SSE3:       # %bb.0:
; SSE3-NEXT:    xorps %xmm1, %xmm1
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,3],xmm1[1,2]
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,2,3,1]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4f32_0zz3:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    xorps %xmm1, %xmm1
; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,3],xmm1[1,2]
; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,2,3,1]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4f32_0zz3:
; SSE41:       # %bb.0:
; SSE41-NEXT:    xorps %xmm1, %xmm1
; SSE41-NEXT:    blendps {{.*#+}} xmm0 = xmm0[0],xmm1[1,2],xmm0[3]
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_0zz3:
; AVX:       # %bb.0:
; AVX-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; AVX-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0],xmm1[1,2],xmm0[3]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> %a, <4 x float> zeroinitializer, <4 x i32> <i32 0, i32 4, i32 4, i32 3>
  ret <4 x float> %shuffle
}

define <4 x float> @shuffle_v4f32_0z2z(<4 x float> %v) {
; SSE2-LABEL: shuffle_v4f32_0z2z:
; SSE2:       # %bb.0:
; SSE2-NEXT:    xorps %xmm1, %xmm1
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[0,0]
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,2,1,3]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4f32_0z2z:
; SSE3:       # %bb.0:
; SSE3-NEXT:    xorps %xmm1, %xmm1
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[0,0]
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,2,1,3]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4f32_0z2z:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    xorps %xmm1, %xmm1
; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[0,0]
; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,2,1,3]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4f32_0z2z:
; SSE41:       # %bb.0:
; SSE41-NEXT:    xorps %xmm1, %xmm1
; SSE41-NEXT:    blendps {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3]
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_0z2z:
; AVX:       # %bb.0:
; AVX-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; AVX-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> %v, <4 x float> <float 0.000000e+00, float undef, float undef, float undef>, <4 x i32> <i32 0, i32 4, i32 2, i32 4>
  ret <4 x float> %shuffle
}

define <4 x float> @shuffle_v4f32_u051(<4 x float> %a, <4 x float> %b) {
; SSE-LABEL: shuffle_v4f32_u051:
; SSE:       # %bb.0:
; SSE-NEXT:    unpcklps {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
; SSE-NEXT:    movaps %xmm1, %xmm0
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_u051:
; AVX:       # %bb.0:
; AVX-NEXT:    vunpcklps {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 undef, i32 0, i32 5, i32 1>
  ret <4 x float> %shuffle
}

define <4 x float> @shuffle_v4f32_0zz4(<4 x float> %a, <4 x float> %b) {
; SSE2-LABEL: shuffle_v4f32_0zz4:
; SSE2:       # %bb.0:
; SSE2-NEXT:    movq {{.*#+}} xmm2 = xmm1[0],zero
; SSE2-NEXT:    pxor %xmm1, %xmm1
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,1],xmm2[2,0]
; SSE2-NEXT:    movss {{.*#+}} xmm1 = xmm0[0],xmm1[1,2,3]
; SSE2-NEXT:    movaps %xmm1, %xmm0
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4f32_0zz4:
; SSE3:       # %bb.0:
; SSE3-NEXT:    movq {{.*#+}} xmm2 = xmm1[0],zero
; SSE3-NEXT:    pxor %xmm1, %xmm1
; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,1],xmm2[2,0]
; SSE3-NEXT:    movss {{.*#+}} xmm1 = xmm0[0],xmm1[1,2,3]
; SSE3-NEXT:    movaps %xmm1, %xmm0
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4f32_0zz4:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    movq {{.*#+}} xmm2 = xmm1[0],zero
; SSSE3-NEXT:    pxor %xmm1, %xmm1
; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,1],xmm2[2,0]
; SSSE3-NEXT:    movss {{.*#+}} xmm1 = xmm0[0],xmm1[1,2,3]
; SSSE3-NEXT:    movaps %xmm1, %xmm0
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4f32_0zz4:
; SSE41:       # %bb.0:
; SSE41-NEXT:    insertps {{.*#+}} xmm0 = xmm0[0],zero,zero,xmm1[0]
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_0zz4:
; AVX:       # %bb.0:
; AVX-NEXT:    vinsertps {{.*#+}} xmm0 = xmm0[0],zero,zero,xmm1[0]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> %b, <4 x float> zeroinitializer, <4 x i32> <i32 undef, i32 5, i32 6, i32 0>
  %shuffle1 = shufflevector <4 x float> %a, <4 x float> %shuffle, <4 x i32> <i32 0, i32 5, i32 6, i32 7>
  ret <4 x float> %shuffle1
}

define <4 x float> @shuffle_v4f32_0zz6(<4 x float> %a, <4 x float> %b) {
; SSE2-LABEL: shuffle_v4f32_0zz6:
; SSE2:       # %bb.0:
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,2]
; SSE2-NEXT:    xorps %xmm1, %xmm1
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[1,2],xmm0[0,3]
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[2,0,1,3]
; SSE2-NEXT:    movaps %xmm1, %xmm0
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4f32_0zz6:
; SSE3:       # %bb.0:
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,2]
; SSE3-NEXT:    xorps %xmm1, %xmm1
; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[1,2],xmm0[0,3]
; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[2,0,1,3]
; SSE3-NEXT:    movaps %xmm1, %xmm0
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4f32_0zz6:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,2]
; SSSE3-NEXT:    xorps %xmm1, %xmm1
; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[1,2],xmm0[0,3]
; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[2,0,1,3]
; SSSE3-NEXT:    movaps %xmm1, %xmm0
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4f32_0zz6:
; SSE41:       # %bb.0:
; SSE41-NEXT:    insertps {{.*#+}} xmm0 = xmm0[0],zero,zero,xmm1[2]
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_0zz6:
; AVX:       # %bb.0:
; AVX-NEXT:    vinsertps {{.*#+}} xmm0 = xmm0[0],zero,zero,xmm1[2]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 undef, i32 undef, i32 6>
  %shuffle1 = shufflevector <4 x float> zeroinitializer, <4 x float> %shuffle, <4 x i32> <i32 4, i32 1, i32 2, i32 7>
  ret <4 x float> %shuffle1
}

define <4 x float> @shuffle_v4f32_0z24(<4 x float> %a, <4 x float> %b) {
; SSE2-LABEL: shuffle_v4f32_0z24:
; SSE2:       # %bb.0:
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,1],xmm0[2,3]
; SSE2-NEXT:    xorps %xmm2, %xmm2
; SSE2-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm0[0]
; SSE2-NEXT:    shufps {{.*#+}} xmm2 = xmm2[2,0],xmm1[2,0]
; SSE2-NEXT:    movaps %xmm2, %xmm0
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4f32_0z24:
; SSE3:       # %bb.0:
; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,1],xmm0[2,3]
; SSE3-NEXT:    xorps %xmm2, %xmm2
; SSE3-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm0[0]
; SSE3-NEXT:    shufps {{.*#+}} xmm2 = xmm2[2,0],xmm1[2,0]
; SSE3-NEXT:    movaps %xmm2, %xmm0
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4f32_0z24:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,1],xmm0[2,3]
; SSSE3-NEXT:    xorps %xmm2, %xmm2
; SSSE3-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm0[0]
; SSSE3-NEXT:    shufps {{.*#+}} xmm2 = xmm2[2,0],xmm1[2,0]
; SSSE3-NEXT:    movaps %xmm2, %xmm0
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4f32_0z24:
; SSE41:       # %bb.0:
; SSE41-NEXT:    insertps {{.*#+}} xmm0 = xmm0[0],zero,xmm0[2],xmm1[0]
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_0z24:
; AVX:       # %bb.0:
; AVX-NEXT:    vinsertps {{.*#+}} xmm0 = xmm0[0],zero,xmm0[2],xmm1[0]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 undef, i32 2, i32 4>
  %shuffle1 = shufflevector <4 x float> zeroinitializer, <4 x float> %shuffle, <4 x i32> <i32 4, i32 1, i32 6, i32 7>
  ret <4 x float> %shuffle1
}

define <4 x i32> @shuffle_v4i32_4zzz(<4 x i32> %a) {
; SSE2-LABEL: shuffle_v4i32_4zzz:
; SSE2:       # %bb.0:
; SSE2-NEXT:    xorps %xmm1, %xmm1
; SSE2-NEXT:    movss {{.*#+}} xmm1 = xmm0[0],xmm1[1,2,3]
; SSE2-NEXT:    movaps %xmm1, %xmm0
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4i32_4zzz:
; SSE3:       # %bb.0:
; SSE3-NEXT:    xorps %xmm1, %xmm1
; SSE3-NEXT:    movss {{.*#+}} xmm1 = xmm0[0],xmm1[1,2,3]
; SSE3-NEXT:    movaps %xmm1, %xmm0
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4i32_4zzz:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    xorps %xmm1, %xmm1
; SSSE3-NEXT:    movss {{.*#+}} xmm1 = xmm0[0],xmm1[1,2,3]
; SSSE3-NEXT:    movaps %xmm1, %xmm0
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4i32_4zzz:
; SSE41:       # %bb.0:
; SSE41-NEXT:    xorps %xmm1, %xmm1
; SSE41-NEXT:    blendps {{.*#+}} xmm0 = xmm0[0],xmm1[1,2,3]
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_4zzz:
; AVX:       # %bb.0:
; AVX-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; AVX-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0],xmm1[1,2,3]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> zeroinitializer, <4 x i32> %a, <4 x i32> <i32 4, i32 1, i32 2, i32 3>
  ret <4 x i32> %shuffle
}

define <4 x i32> @shuffle_v4i32_z4zz(<4 x i32> %a) {
; SSE2-LABEL: shuffle_v4i32_z4zz:
; SSE2:       # %bb.0:
; SSE2-NEXT:    xorps %xmm1, %xmm1
; SSE2-NEXT:    movss {{.*#+}} xmm1 = xmm0[0],xmm1[1,2,3]
; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm1[1,0,1,1]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4i32_z4zz:
; SSE3:       # %bb.0:
; SSE3-NEXT:    xorps %xmm1, %xmm1
; SSE3-NEXT:    movss {{.*#+}} xmm1 = xmm0[0],xmm1[1,2,3]
; SSE3-NEXT:    pshufd {{.*#+}} xmm0 = xmm1[1,0,1,1]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4i32_z4zz:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    xorps %xmm1, %xmm1
; SSSE3-NEXT:    movss {{.*#+}} xmm1 = xmm0[0],xmm1[1,2,3]
; SSSE3-NEXT:    pshufd {{.*#+}} xmm0 = xmm1[1,0,1,1]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4i32_z4zz:
; SSE41:       # %bb.0:
; SSE41-NEXT:    pxor %xmm1, %xmm1
; SSE41-NEXT:    pblendw {{.*#+}} xmm1 = xmm0[0,1],xmm1[2,3,4,5,6,7]
; SSE41-NEXT:    pshufd {{.*#+}} xmm0 = xmm1[1,0,1,1]
; SSE41-NEXT:    retq
;
; AVX1-LABEL: shuffle_v4i32_z4zz:
; AVX1:       # %bb.0:
; AVX1-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; AVX1-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0],xmm1[1,2,3]
; AVX1-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[1,0,1,1]
; AVX1-NEXT:    retq
;
; AVX2-SLOW-LABEL: shuffle_v4i32_z4zz:
; AVX2-SLOW:       # %bb.0:
; AVX2-SLOW-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; AVX2-SLOW-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0],xmm1[1,2,3]
; AVX2-SLOW-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[1,0,1,1]
; AVX2-SLOW-NEXT:    retq
;
; AVX2-FAST-LABEL: shuffle_v4i32_z4zz:
; AVX2-FAST:       # %bb.0:
; AVX2-FAST-NEXT:    vpshufb {{.*#+}} xmm0 = zero,zero,zero,zero,xmm0[0,1,2,3],zero,zero,zero,zero,zero,zero,zero,zero
; AVX2-FAST-NEXT:    retq
;
; AVX512VL-LABEL: shuffle_v4i32_z4zz:
; AVX512VL:       # %bb.0:
; AVX512VL-NEXT:    vpshufb {{.*#+}} xmm0 = zero,zero,zero,zero,xmm0[0,1,2,3],zero,zero,zero,zero,zero,zero,zero,zero
; AVX512VL-NEXT:    retq
  %shuffle = shufflevector <4 x i32> zeroinitializer, <4 x i32> %a, <4 x i32> <i32 2, i32 4, i32 3, i32 0>
  ret <4 x i32> %shuffle
}

define <4 x i32> @shuffle_v4i32_zz4z(<4 x i32> %a) {
; SSE2-LABEL: shuffle_v4i32_zz4z:
; SSE2:       # %bb.0:
; SSE2-NEXT:    xorps %xmm1, %xmm1
; SSE2-NEXT:    movss {{.*#+}} xmm1 = xmm0[0],xmm1[1,2,3]
; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm1[1,1,0,1]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4i32_zz4z:
; SSE3:       # %bb.0:
; SSE3-NEXT:    xorps %xmm1, %xmm1
; SSE3-NEXT:    movss {{.*#+}} xmm1 = xmm0[0],xmm1[1,2,3]
; SSE3-NEXT:    pshufd {{.*#+}} xmm0 = xmm1[1,1,0,1]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4i32_zz4z:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    xorps %xmm1, %xmm1
; SSSE3-NEXT:    movss {{.*#+}} xmm1 = xmm0[0],xmm1[1,2,3]
; SSSE3-NEXT:    pshufd {{.*#+}} xmm0 = xmm1[1,1,0,1]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4i32_zz4z:
; SSE41:       # %bb.0:
; SSE41-NEXT:    pxor %xmm1, %xmm1
; SSE41-NEXT:    pblendw {{.*#+}} xmm1 = xmm0[0,1],xmm1[2,3,4,5,6,7]
; SSE41-NEXT:    pshufd {{.*#+}} xmm0 = xmm1[1,1,0,1]
; SSE41-NEXT:    retq
;
; AVX1-LABEL: shuffle_v4i32_zz4z:
; AVX1:       # %bb.0:
; AVX1-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; AVX1-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0],xmm1[1,2,3]
; AVX1-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[1,1,0,1]
; AVX1-NEXT:    retq
;
; AVX2-SLOW-LABEL: shuffle_v4i32_zz4z:
; AVX2-SLOW:       # %bb.0:
; AVX2-SLOW-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; AVX2-SLOW-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0],xmm1[1,2,3]
; AVX2-SLOW-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[1,1,0,1]
; AVX2-SLOW-NEXT:    retq
;
; AVX2-FAST-LABEL: shuffle_v4i32_zz4z:
; AVX2-FAST:       # %bb.0:
; AVX2-FAST-NEXT:    vpshufb {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2,3],zero,zero,zero,zero
; AVX2-FAST-NEXT:    retq
;
; AVX512VL-LABEL: shuffle_v4i32_zz4z:
; AVX512VL:       # %bb.0:
; AVX512VL-NEXT:    vpshufb {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2,3],zero,zero,zero,zero
; AVX512VL-NEXT:    retq
  %shuffle = shufflevector <4 x i32> zeroinitializer, <4 x i32> %a, <4 x i32> <i32 0, i32 0, i32 4, i32 0>
  ret <4 x i32> %shuffle
}

define <4 x i32> @shuffle_v4i32_zuu4(<4 x i32> %a) {
; SSE-LABEL: shuffle_v4i32_zuu4:
; SSE:       # %bb.0:
; SSE-NEXT:    pslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2,3]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_zuu4:
; AVX:       # %bb.0:
; AVX-NEXT:    vpslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2,3]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> zeroinitializer, <4 x i32> %a, <4 x i32> <i32 0, i32 undef, i32 undef, i32 4>
  ret <4 x i32> %shuffle
}

define <4 x i32> @shuffle_v4i32_z6zz(<4 x i32> %a) {
; SSE2-LABEL: shuffle_v4i32_z6zz:
; SSE2:       # %bb.0:
; SSE2-NEXT:    psrldq {{.*#+}} xmm0 = xmm0[8,9,10,11,12,13,14,15],zero,zero,zero,zero,zero,zero,zero,zero
; SSE2-NEXT:    xorps %xmm1, %xmm1
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,0],xmm1[2,3]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4i32_z6zz:
; SSE3:       # %bb.0:
; SSE3-NEXT:    psrldq {{.*#+}} xmm0 = xmm0[8,9,10,11,12,13,14,15],zero,zero,zero,zero,zero,zero,zero,zero
; SSE3-NEXT:    xorps %xmm1, %xmm1
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,0],xmm1[2,3]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4i32_z6zz:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    psrldq {{.*#+}} xmm0 = xmm0[8,9,10,11,12,13,14,15],zero,zero,zero,zero,zero,zero,zero,zero
; SSSE3-NEXT:    xorps %xmm1, %xmm1
; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,0],xmm1[2,3]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4i32_z6zz:
; SSE41:       # %bb.0:
; SSE41-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[2,2,3,3]
; SSE41-NEXT:    pxor %xmm0, %xmm0
; SSE41-NEXT:    pblendw {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,3],xmm0[4,5,6,7]
; SSE41-NEXT:    retq
;
; AVX1-LABEL: shuffle_v4i32_z6zz:
; AVX1:       # %bb.0:
; AVX1-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[2,2,3,3]
; AVX1-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; AVX1-NEXT:    vblendps {{.*#+}} xmm0 = xmm1[0],xmm0[1],xmm1[2,3]
; AVX1-NEXT:    retq
;
; AVX2-SLOW-LABEL: shuffle_v4i32_z6zz:
; AVX2-SLOW:       # %bb.0:
; AVX2-SLOW-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[2,2,3,3]
; AVX2-SLOW-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; AVX2-SLOW-NEXT:    vblendps {{.*#+}} xmm0 = xmm1[0],xmm0[1],xmm1[2,3]
; AVX2-SLOW-NEXT:    retq
;
; AVX2-FAST-LABEL: shuffle_v4i32_z6zz:
; AVX2-FAST:       # %bb.0:
; AVX2-FAST-NEXT:    vpshufb {{.*#+}} xmm0 = zero,zero,zero,zero,xmm0[8,9,10,11],zero,zero,zero,zero,zero,zero,zero,zero
; AVX2-FAST-NEXT:    retq
;
; AVX512VL-LABEL: shuffle_v4i32_z6zz:
; AVX512VL:       # %bb.0:
; AVX512VL-NEXT:    vpshufb {{.*#+}} xmm0 = zero,zero,zero,zero,xmm0[8,9,10,11],zero,zero,zero,zero,zero,zero,zero,zero
; AVX512VL-NEXT:    retq
  %shuffle = shufflevector <4 x i32> zeroinitializer, <4 x i32> %a, <4 x i32> <i32 0, i32 6, i32 2, i32 3>
  ret <4 x i32> %shuffle
}

define <4 x i32> @shuffle_v4i32_7012(<4 x i32> %a, <4 x i32> %b) {
; SSE2-LABEL: shuffle_v4i32_7012:
; SSE2:       # %bb.0:
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[3,0],xmm0[0,0]
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,2],xmm0[1,2]
; SSE2-NEXT:    movaps %xmm1, %xmm0
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4i32_7012:
; SSE3:       # %bb.0:
; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[3,0],xmm0[0,0]
; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,2],xmm0[1,2]
; SSE3-NEXT:    movaps %xmm1, %xmm0
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4i32_7012:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    palignr {{.*#+}} xmm0 = xmm1[12,13,14,15],xmm0[0,1,2,3,4,5,6,7,8,9,10,11]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4i32_7012:
; SSE41:       # %bb.0:
; SSE41-NEXT:    palignr {{.*#+}} xmm0 = xmm1[12,13,14,15],xmm0[0,1,2,3,4,5,6,7,8,9,10,11]
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_7012:
; AVX:       # %bb.0:
; AVX-NEXT:    vpalignr {{.*#+}} xmm0 = xmm1[12,13,14,15],xmm0[0,1,2,3,4,5,6,7,8,9,10,11]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 7, i32 0, i32 1, i32 2>
  ret <4 x i32> %shuffle
}

define <4 x i32> @shuffle_v4i32_6701(<4 x i32> %a, <4 x i32> %b) {
; SSE2-LABEL: shuffle_v4i32_6701:
; SSE2:       # %bb.0:
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[2,3],xmm0[0,1]
; SSE2-NEXT:    movaps %xmm1, %xmm0
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4i32_6701:
; SSE3:       # %bb.0:
; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[2,3],xmm0[0,1]
; SSE3-NEXT:    movaps %xmm1, %xmm0
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4i32_6701:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    palignr {{.*#+}} xmm0 = xmm1[8,9,10,11,12,13,14,15],xmm0[0,1,2,3,4,5,6,7]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4i32_6701:
; SSE41:       # %bb.0:
; SSE41-NEXT:    palignr {{.*#+}} xmm0 = xmm1[8,9,10,11,12,13,14,15],xmm0[0,1,2,3,4,5,6,7]
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_6701:
; AVX:       # %bb.0:
; AVX-NEXT:    vpalignr {{.*#+}} xmm0 = xmm1[8,9,10,11,12,13,14,15],xmm0[0,1,2,3,4,5,6,7]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 6, i32 7, i32 0, i32 1>
  ret <4 x i32> %shuffle
}

define <4 x i32> @shuffle_v4i32_5670(<4 x i32> %a, <4 x i32> %b) {
; SSE2-LABEL: shuffle_v4i32_5670:
; SSE2:       # %bb.0:
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,0],xmm1[3,0]
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[1,2],xmm0[2,0]
; SSE2-NEXT:    movaps %xmm1, %xmm0
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4i32_5670:
; SSE3:       # %bb.0:
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,0],xmm1[3,0]
; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[1,2],xmm0[2,0]
; SSE3-NEXT:    movaps %xmm1, %xmm0
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4i32_5670:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    palignr {{.*#+}} xmm0 = xmm1[4,5,6,7,8,9,10,11,12,13,14,15],xmm0[0,1,2,3]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4i32_5670:
; SSE41:       # %bb.0:
; SSE41-NEXT:    palignr {{.*#+}} xmm0 = xmm1[4,5,6,7,8,9,10,11,12,13,14,15],xmm0[0,1,2,3]
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_5670:
; AVX:       # %bb.0:
; AVX-NEXT:    vpalignr {{.*#+}} xmm0 = xmm1[4,5,6,7,8,9,10,11,12,13,14,15],xmm0[0,1,2,3]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 5, i32 6, i32 7, i32 0>
  ret <4 x i32> %shuffle
}

define <4 x i32> @shuffle_v4i32_1234(<4 x i32> %a, <4 x i32> %b) {
; SSE2-LABEL: shuffle_v4i32_1234:
; SSE2:       # %bb.0:
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,0],xmm0[3,0]
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[1,2],xmm1[2,0]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4i32_1234:
; SSE3:       # %bb.0:
; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,0],xmm0[3,0]
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[1,2],xmm1[2,0]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4i32_1234:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    palignr {{.*#+}} xmm1 = xmm0[4,5,6,7,8,9,10,11,12,13,14,15],xmm1[0,1,2,3]
; SSSE3-NEXT:    movdqa %xmm1, %xmm0
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4i32_1234:
; SSE41:       # %bb.0:
; SSE41-NEXT:    palignr {{.*#+}} xmm1 = xmm0[4,5,6,7,8,9,10,11,12,13,14,15],xmm1[0,1,2,3]
; SSE41-NEXT:    movdqa %xmm1, %xmm0
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_1234:
; AVX:       # %bb.0:
; AVX-NEXT:    vpalignr {{.*#+}} xmm0 = xmm0[4,5,6,7,8,9,10,11,12,13,14,15],xmm1[0,1,2,3]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 1, i32 2, i32 3, i32 4>
  ret <4 x i32> %shuffle
}

define <4 x i32> @shuffle_v4i32_2345(<4 x i32> %a, <4 x i32> %b) {
; SSE2-LABEL: shuffle_v4i32_2345:
; SSE2:       # %bb.0:
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,3],xmm1[0,1]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4i32_2345:
; SSE3:       # %bb.0:
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,3],xmm1[0,1]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4i32_2345:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    palignr {{.*#+}} xmm1 = xmm0[8,9,10,11,12,13,14,15],xmm1[0,1,2,3,4,5,6,7]
; SSSE3-NEXT:    movdqa %xmm1, %xmm0
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4i32_2345:
; SSE41:       # %bb.0:
; SSE41-NEXT:    palignr {{.*#+}} xmm1 = xmm0[8,9,10,11,12,13,14,15],xmm1[0,1,2,3,4,5,6,7]
; SSE41-NEXT:    movdqa %xmm1, %xmm0
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_2345:
; AVX:       # %bb.0:
; AVX-NEXT:    vpalignr {{.*#+}} xmm0 = xmm0[8,9,10,11,12,13,14,15],xmm1[0,1,2,3,4,5,6,7]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 2, i32 3, i32 4, i32 5>
  ret <4 x i32> %shuffle
}

; PR22391
define <4 x i32> @shuffle_v4i32_2456(<4 x i32> %a, <4 x i32> %b) {
; SSE2-LABEL: shuffle_v4i32_2456:
; SSE2:       # %bb.0:
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,3],xmm1[0,1]
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[1,2]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4i32_2456:
; SSE3:       # %bb.0:
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,3],xmm1[0,1]
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[1,2]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4i32_2456:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,2,2,2]
; SSSE3-NEXT:    palignr {{.*#+}} xmm1 = xmm0[12,13,14,15],xmm1[0,1,2,3,4,5,6,7,8,9,10,11]
; SSSE3-NEXT:    movdqa %xmm1, %xmm0
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4i32_2456:
; SSE41:       # %bb.0:
; SSE41-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,2,2,2]
; SSE41-NEXT:    palignr {{.*#+}} xmm1 = xmm0[12,13,14,15],xmm1[0,1,2,3,4,5,6,7,8,9,10,11]
; SSE41-NEXT:    movdqa %xmm1, %xmm0
; SSE41-NEXT:    retq
;
; AVX1OR2-LABEL: shuffle_v4i32_2456:
; AVX1OR2:       # %bb.0:
; AVX1OR2-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[2,2,2,2]
; AVX1OR2-NEXT:    vpalignr {{.*#+}} xmm0 = xmm0[12,13,14,15],xmm1[0,1,2,3,4,5,6,7,8,9,10,11]
; AVX1OR2-NEXT:    retq
;
; AVX512VL-LABEL: shuffle_v4i32_2456:
; AVX512VL:       # %bb.0:
; AVX512VL-NEXT:    vmovdqa {{.*#+}} xmm2 = [6,0,1,2]
; AVX512VL-NEXT:    vpermi2d %xmm0, %xmm1, %xmm2
; AVX512VL-NEXT:    vmovdqa %xmm2, %xmm0
; AVX512VL-NEXT:    retq
  %s1 = shufflevector <4 x i32> %a, <4 x i32> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 2>
  %s2 = shufflevector <4 x i32> %s1, <4 x i32> %b, <4 x i32> <i32 3, i32 4, i32 5, i32 6>
  ret <4 x i32> %s2
}

define <4 x i32> @shuffle_v4i32_40u1(<4 x i32> %a, <4 x i32> %b) {
; SSE-LABEL: shuffle_v4i32_40u1:
; SSE:       # %bb.0:
; SSE-NEXT:    unpcklps {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
; SSE-NEXT:    movaps %xmm1, %xmm0
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_40u1:
; AVX:       # %bb.0:
; AVX-NEXT:    vunpcklps {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 4, i32 0, i32 undef, i32 1>
  ret <4 x i32> %shuffle
}

define <4 x i32> @shuffle_v4i32_3456(<4 x i32> %a, <4 x i32> %b) {
; SSE2-LABEL: shuffle_v4i32_3456:
; SSE2:       # %bb.0:
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,0],xmm1[0,0]
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[1,2]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4i32_3456:
; SSE3:       # %bb.0:
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,0],xmm1[0,0]
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[1,2]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4i32_3456:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    palignr {{.*#+}} xmm1 = xmm0[12,13,14,15],xmm1[0,1,2,3,4,5,6,7,8,9,10,11]
; SSSE3-NEXT:    movdqa %xmm1, %xmm0
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4i32_3456:
; SSE41:       # %bb.0:
; SSE41-NEXT:    palignr {{.*#+}} xmm1 = xmm0[12,13,14,15],xmm1[0,1,2,3,4,5,6,7,8,9,10,11]
; SSE41-NEXT:    movdqa %xmm1, %xmm0
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_3456:
; AVX:       # %bb.0:
; AVX-NEXT:    vpalignr {{.*#+}} xmm0 = xmm0[12,13,14,15],xmm1[0,1,2,3,4,5,6,7,8,9,10,11]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 3, i32 4, i32 5, i32 6>
  ret <4 x i32> %shuffle
}

define <4 x i32> @shuffle_v4i32_0u1u(<4 x i32> %a, <4 x i32> %b) {
; SSE2-LABEL: shuffle_v4i32_0u1u:
; SSE2:       # %bb.0:
; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,1,1,3]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4i32_0u1u:
; SSE3:       # %bb.0:
; SSE3-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,1,1,3]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4i32_0u1u:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,1,1,3]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4i32_0u1u:
; SSE41:       # %bb.0:
; SSE41-NEXT:    pmovzxdq {{.*#+}} xmm0 = xmm0[0],zero,xmm0[1],zero
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_0u1u:
; AVX:       # %bb.0:
; AVX-NEXT:    vpmovzxdq {{.*#+}} xmm0 = xmm0[0],zero,xmm0[1],zero
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 0, i32 undef, i32 1, i32 undef>
  ret <4 x i32> %shuffle
}

define <4 x i32> @shuffle_v4i32_0z1z(<4 x i32> %a) {
; SSE2-LABEL: shuffle_v4i32_0z1z:
; SSE2:       # %bb.0:
; SSE2-NEXT:    xorps %xmm1, %xmm1
; SSE2-NEXT:    unpcklps {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4i32_0z1z:
; SSE3:       # %bb.0:
; SSE3-NEXT:    xorps %xmm1, %xmm1
; SSE3-NEXT:    unpcklps {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4i32_0z1z:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    xorps %xmm1, %xmm1
; SSSE3-NEXT:    unpcklps {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4i32_0z1z:
; SSE41:       # %bb.0:
; SSE41-NEXT:    pmovzxdq {{.*#+}} xmm0 = xmm0[0],zero,xmm0[1],zero
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_0z1z:
; AVX:       # %bb.0:
; AVX-NEXT:    vpmovzxdq {{.*#+}} xmm0 = xmm0[0],zero,xmm0[1],zero
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> zeroinitializer, <4 x i32> <i32 0, i32 5, i32 1, i32 7>
  ret <4 x i32> %shuffle
}

define <4 x i32> @shuffle_v4i32_01zu(<4 x i32> %a) {
; SSE-LABEL: shuffle_v4i32_01zu:
; SSE:       # %bb.0:
; SSE-NEXT:    movq {{.*#+}} xmm0 = xmm0[0],zero
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_01zu:
; AVX:       # %bb.0:
; AVX-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> zeroinitializer, <4 x i32> <i32 0, i32 1, i32 7, i32 undef>
  ret <4 x i32> %shuffle
}

define <4 x i32> @shuffle_v4i32_0z23(<4 x i32> %a) {
; SSE2-LABEL: shuffle_v4i32_0z23:
; SSE2:       # %bb.0:
; SSE2-NEXT:    andps {{.*}}(%rip), %xmm0
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4i32_0z23:
; SSE3:       # %bb.0:
; SSE3-NEXT:    andps {{.*}}(%rip), %xmm0
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4i32_0z23:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    andps {{.*}}(%rip), %xmm0
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4i32_0z23:
; SSE41:       # %bb.0:
; SSE41-NEXT:    xorps %xmm1, %xmm1
; SSE41-NEXT:    blendps {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2,3]
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_0z23:
; AVX:       # %bb.0:
; AVX-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; AVX-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2,3]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> zeroinitializer, <4 x i32> <i32 0, i32 4, i32 2, i32 3>
  ret <4 x i32> %shuffle
}

define <4 x i32> @shuffle_v4i32_01z3(<4 x i32> %a) {
; SSE2-LABEL: shuffle_v4i32_01z3:
; SSE2:       # %bb.0:
; SSE2-NEXT:    andps {{.*}}(%rip), %xmm0
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4i32_01z3:
; SSE3:       # %bb.0:
; SSE3-NEXT:    andps {{.*}}(%rip), %xmm0
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4i32_01z3:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    andps {{.*}}(%rip), %xmm0
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4i32_01z3:
; SSE41:       # %bb.0:
; SSE41-NEXT:    xorps %xmm1, %xmm1
; SSE41-NEXT:    blendps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2],xmm0[3]
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_01z3:
; AVX:       # %bb.0:
; AVX-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; AVX-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2],xmm0[3]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> zeroinitializer, <4 x i32> <i32 0, i32 1, i32 4, i32 3>
  ret <4 x i32> %shuffle
}

define <4 x i32> @shuffle_v4i32_012z(<4 x i32> %a) {
; SSE2-LABEL: shuffle_v4i32_012z:
; SSE2:       # %bb.0:
; SSE2-NEXT:    andps {{.*}}(%rip), %xmm0
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4i32_012z:
; SSE3:       # %bb.0:
; SSE3-NEXT:    andps {{.*}}(%rip), %xmm0
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4i32_012z:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    andps {{.*}}(%rip), %xmm0
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4i32_012z:
; SSE41:       # %bb.0:
; SSE41-NEXT:    xorps %xmm1, %xmm1
; SSE41-NEXT:    blendps {{.*#+}} xmm0 = xmm0[0,1,2],xmm1[3]
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_012z:
; AVX:       # %bb.0:
; AVX-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; AVX-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0,1,2],xmm1[3]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> zeroinitializer, <4 x i32> <i32 0, i32 1, i32 2, i32 7>
  ret <4 x i32> %shuffle
}

define <4 x i32> @shuffle_v4i32_0zz3(<4 x i32> %a) {
; SSE2-LABEL: shuffle_v4i32_0zz3:
; SSE2:       # %bb.0:
; SSE2-NEXT:    andps {{.*}}(%rip), %xmm0
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_v4i32_0zz3:
; SSE3:       # %bb.0:
; SSE3-NEXT:    andps {{.*}}(%rip), %xmm0
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_v4i32_0zz3:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    andps {{.*}}(%rip), %xmm0
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_v4i32_0zz3:
; SSE41:       # %bb.0:
; SSE41-NEXT:    xorps %xmm1, %xmm1
; SSE41-NEXT:    blendps {{.*#+}} xmm0 = xmm0[0],xmm1[1,2],xmm0[3]
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_0zz3:
; AVX:       # %bb.0:
; AVX-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; AVX-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0],xmm1[1,2],xmm0[3]
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> zeroinitializer, <4 x i32> <i32 0, i32 4, i32 4, i32 3>
  ret <4 x i32> %shuffle
}

define <4 x i32> @shuffle_v4i32_bitcast_0415(<4 x i32> %a, <4 x i32> %b) {
; SSE-LABEL: shuffle_v4i32_bitcast_0415:
; SSE:       # %bb.0:
; SSE-NEXT:    unpcklps {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_bitcast_0415:
; AVX:       # %bb.0:
; AVX-NEXT:    vunpcklps {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
; AVX-NEXT:    retq
  %shuffle32 = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 1, i32 5, i32 0, i32 4>
  %bitcast64 = bitcast <4 x i32> %shuffle32 to <2 x double>
  %shuffle64 = shufflevector <2 x double> %bitcast64, <2 x double> undef, <2 x i32> <i32 1, i32 0>
  %bitcast32 = bitcast <2 x double> %shuffle64 to <4 x i32>
  ret <4 x i32> %bitcast32
}

define <4 x float> @shuffle_v4f32_bitcast_4401(<4 x float> %a, <4 x i32> %b) {
; SSE-LABEL: shuffle_v4f32_bitcast_4401:
; SSE:       # %bb.0:
; SSE-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,0],xmm0[0,1]
; SSE-NEXT:    movaps %xmm1, %xmm0
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_bitcast_4401:
; AVX:       # %bb.0:
; AVX-NEXT:    vshufps {{.*#+}} xmm0 = xmm1[0,0],xmm0[0,1]
; AVX-NEXT:    retq
  %1 = shufflevector <4 x i32> %b, <4 x i32> undef, <4 x i32> <i32 0, i32 0, i32 1, i32 1>
  %2 = bitcast <4 x i32> %1 to <2 x double>
  %3 = bitcast <4 x float> %a to <2 x double>
  %4 = shufflevector <2 x double> %2, <2 x double> %3, <2 x i32> <i32 0, i32 2>
  %5 = bitcast <2 x double> %4 to <4 x float>
  ret <4 x float> %5
}

define <4 x float> @shuffle_v4f32_bitcast_0045(<4 x float> %a, <4 x i32> %b) {
; SSE-LABEL: shuffle_v4f32_bitcast_0045:
; SSE:       # %bb.0:
; SSE-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,0],xmm1[0,1]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4f32_bitcast_0045:
; AVX:       # %bb.0:
; AVX-NEXT:    vshufps {{.*#+}} xmm0 = xmm0[0,0],xmm1[0,1]
; AVX-NEXT:    retq
  %1 = shufflevector <4 x float> %a, <4 x float> undef, <4 x i32> <i32 0, i32 0, i32 1, i32 1>
  %2 = bitcast <4 x i32> %b to <4 x float>
  %3 = shufflevector <4 x float> %1, <4 x float> %2, <4 x i32> <i32 1, i32 0, i32 4, i32 5>
  ret <4 x float> %3
}

define <4 x float> @mask_v4f32_4127(<4 x float> %a, <4 x float> %b) {
; SSE2-LABEL: mask_v4f32_4127:
; SSE2:       # %bb.0:
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,3],xmm0[1,2]
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,2,3,1]
; SSE2-NEXT:    movaps %xmm1, %xmm0
; SSE2-NEXT:    retq
;
; SSE3-LABEL: mask_v4f32_4127:
; SSE3:       # %bb.0:
; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,3],xmm0[1,2]
; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,2,3,1]
; SSE3-NEXT:    movaps %xmm1, %xmm0
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: mask_v4f32_4127:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,3],xmm0[1,2]
; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,2,3,1]
; SSSE3-NEXT:    movaps %xmm1, %xmm0
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: mask_v4f32_4127:
; SSE41:       # %bb.0:
; SSE41-NEXT:    blendps {{.*#+}} xmm0 = xmm1[0],xmm0[1,2],xmm1[3]
; SSE41-NEXT:    retq
;
; AVX-LABEL: mask_v4f32_4127:
; AVX:       # %bb.0:
; AVX-NEXT:    vblendps {{.*#+}} xmm0 = xmm1[0],xmm0[1,2],xmm1[3]
; AVX-NEXT:    retq
  %1 = bitcast <4 x float> %a to <4 x i32>
  %2 = bitcast <4 x float> %b to <4 x i32>
  %3 = and <4 x i32> %1, <i32  0, i32 -1, i32 -1, i32  0>
  %4 = and <4 x i32> %2, <i32 -1, i32  0, i32  0, i32 -1>
  %5 = or <4 x i32> %4, %3
  %6 = bitcast <4 x i32> %5 to <4 x float>
  ret <4 x float> %6
}

define <4 x float> @mask_v4f32_0127(<4 x float> %a, <4 x float> %b) {
; SSE2-LABEL: mask_v4f32_0127:
; SSE2:       # %bb.0:
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,0],xmm1[2,0]
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,1],xmm0[2,0]
; SSE2-NEXT:    movaps %xmm1, %xmm0
; SSE2-NEXT:    retq
;
; SSE3-LABEL: mask_v4f32_0127:
; SSE3:       # %bb.0:
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,0],xmm1[2,0]
; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,1],xmm0[2,0]
; SSE3-NEXT:    movaps %xmm1, %xmm0
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: mask_v4f32_0127:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,0],xmm1[2,0]
; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,1],xmm0[2,0]
; SSSE3-NEXT:    movaps %xmm1, %xmm0
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: mask_v4f32_0127:
; SSE41:       # %bb.0:
; SSE41-NEXT:    blendps {{.*#+}} xmm0 = xmm1[0,1,2],xmm0[3]
; SSE41-NEXT:    retq
;
; AVX-LABEL: mask_v4f32_0127:
; AVX:       # %bb.0:
; AVX-NEXT:    vblendps {{.*#+}} xmm0 = xmm1[0,1,2],xmm0[3]
; AVX-NEXT:    retq
  %1 = bitcast <4 x float> %a to <2 x i64>
  %2 = bitcast <4 x float> %b to <2 x i64>
  %3 = and <2 x i64> %1, <i64 0, i64 -4294967296>
  %4 = and <2 x i64> %2, <i64 -1, i64 4294967295>
  %5 = or <2 x i64> %4, %3
  %6 = bitcast <2 x i64> %5 to <4 x float>
  ret <4 x float> %6
}

define <4 x i32> @mask_v4i32_0127(<4 x i32> %a, <4 x i32> %b) {
; SSE2-LABEL: mask_v4i32_0127:
; SSE2:       # %bb.0:
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,0],xmm1[2,0]
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,1],xmm0[2,0]
; SSE2-NEXT:    movaps %xmm1, %xmm0
; SSE2-NEXT:    retq
;
; SSE3-LABEL: mask_v4i32_0127:
; SSE3:       # %bb.0:
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,0],xmm1[2,0]
; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,1],xmm0[2,0]
; SSE3-NEXT:    movaps %xmm1, %xmm0
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: mask_v4i32_0127:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,0],xmm1[2,0]
; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,1],xmm0[2,0]
; SSSE3-NEXT:    movaps %xmm1, %xmm0
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: mask_v4i32_0127:
; SSE41:       # %bb.0:
; SSE41-NEXT:    blendps {{.*#+}} xmm0 = xmm1[0,1,2],xmm0[3]
; SSE41-NEXT:    retq
;
; AVX-LABEL: mask_v4i32_0127:
; AVX:       # %bb.0:
; AVX-NEXT:    vblendps {{.*#+}} xmm0 = xmm1[0,1,2],xmm0[3]
; AVX-NEXT:    retq
  %1 = bitcast <4 x i32> %a to <2 x i64>
  %2 = bitcast <4 x i32> %b to <2 x i64>
  %3 = and <2 x i64> %1, <i64 0, i64 -4294967296>
  %4 = and <2 x i64> %2, <i64 -1, i64 4294967295>
  %5 = or <2 x i64> %4, %3
  %6 = bitcast <2 x i64> %5 to <4 x i32>
  ret <4 x i32> %6
}

define <4 x float> @broadcast_v4f32_0101_from_v2f32(<2 x float>* %x) {
; SSE2-LABEL: broadcast_v4f32_0101_from_v2f32:
; SSE2:       # %bb.0:
; SSE2-NEXT:    movsd {{.*#+}} xmm0 = mem[0],zero
; SSE2-NEXT:    movlhps {{.*#+}} xmm0 = xmm0[0,0]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: broadcast_v4f32_0101_from_v2f32:
; SSE3:       # %bb.0:
; SSE3-NEXT:    movddup {{.*#+}} xmm0 = mem[0,0]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: broadcast_v4f32_0101_from_v2f32:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    movddup {{.*#+}} xmm0 = mem[0,0]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: broadcast_v4f32_0101_from_v2f32:
; SSE41:       # %bb.0:
; SSE41-NEXT:    movddup {{.*#+}} xmm0 = mem[0,0]
; SSE41-NEXT:    retq
;
; AVX-LABEL: broadcast_v4f32_0101_from_v2f32:
; AVX:       # %bb.0:
; AVX-NEXT:    vmovddup {{.*#+}} xmm0 = mem[0,0]
; AVX-NEXT:    retq
  %1 = load <2 x float>, <2 x float>* %x, align 1
  %2 = shufflevector <2 x float> %1, <2 x float> undef, <4 x i32> <i32 0, i32 1, i32 0, i32 1>
  ret <4 x float> %2
}

define <4 x i32> @extract3_insert0_v4i32_7123(<4 x i32> %a0, <4 x i32> %a1) {
; SSE2-LABEL: extract3_insert0_v4i32_7123:
; SSE2:       # %bb.0:
; SSE2-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[3,3,3,3]
; SSE2-NEXT:    movd %xmm1, %eax
; SSE2-NEXT:    movd %eax, %xmm1
; SSE2-NEXT:    movss {{.*#+}} xmm0 = xmm1[0],xmm0[1,2,3]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: extract3_insert0_v4i32_7123:
; SSE3:       # %bb.0:
; SSE3-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[3,3,3,3]
; SSE3-NEXT:    movd %xmm1, %eax
; SSE3-NEXT:    movd %eax, %xmm1
; SSE3-NEXT:    movss {{.*#+}} xmm0 = xmm1[0],xmm0[1,2,3]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: extract3_insert0_v4i32_7123:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[3,3,3,3]
; SSSE3-NEXT:    movd %xmm1, %eax
; SSSE3-NEXT:    movd %eax, %xmm1
; SSSE3-NEXT:    movss {{.*#+}} xmm0 = xmm1[0],xmm0[1,2,3]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: extract3_insert0_v4i32_7123:
; SSE41:       # %bb.0:
; SSE41-NEXT:    extractps $3, %xmm1, %eax
; SSE41-NEXT:    pinsrd $0, %eax, %xmm0
; SSE41-NEXT:    retq
;
; AVX-LABEL: extract3_insert0_v4i32_7123:
; AVX:       # %bb.0:
; AVX-NEXT:    vextractps $3, %xmm1, %eax
; AVX-NEXT:    vpinsrd $0, %eax, %xmm0, %xmm0
; AVX-NEXT:    retq
  %1 = extractelement <4 x i32> %a1, i32 3
  %2 = insertelement <4 x i32> %a0, i32 %1, i32 0
  ret <4 x i32> %2
}

define <4 x i32> @extract3_insert3_v4i32_0127(<4 x i32> %a0, <4 x i32> %a1) {
; SSE2-LABEL: extract3_insert3_v4i32_0127:
; SSE2:       # %bb.0:
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[3,1],xmm0[2,3]
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,0]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: extract3_insert3_v4i32_0127:
; SSE3:       # %bb.0:
; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[3,1],xmm0[2,3]
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,0]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: extract3_insert3_v4i32_0127:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[3,1],xmm0[2,3]
; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,0]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: extract3_insert3_v4i32_0127:
; SSE41:       # %bb.0:
; SSE41-NEXT:    blendps {{.*#+}} xmm0 = xmm0[0,1,2],xmm1[3]
; SSE41-NEXT:    retq
;
; AVX-LABEL: extract3_insert3_v4i32_0127:
; AVX:       # %bb.0:
; AVX-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0,1,2],xmm1[3]
; AVX-NEXT:    retq
  %1 = extractelement <4 x i32> %a1, i32 3
  %2 = insertelement <4 x i32> %a0, i32 %1, i32 3
  ret <4 x i32> %2
}

define <4 x i32> @insert_reg_and_zero_v4i32(i32 %a) {
; SSE-LABEL: insert_reg_and_zero_v4i32:
; SSE:       # %bb.0:
; SSE-NEXT:    movd %edi, %xmm0
; SSE-NEXT:    retq
;
; AVX-LABEL: insert_reg_and_zero_v4i32:
; AVX:       # %bb.0:
; AVX-NEXT:    vmovd %edi, %xmm0
; AVX-NEXT:    retq
  %v = insertelement <4 x i32> undef, i32 %a, i32 0
  %shuffle = shufflevector <4 x i32> %v, <4 x i32> zeroinitializer, <4 x i32> <i32 0, i32 5, i32 6, i32 7>
  ret <4 x i32> %shuffle
}

define <4 x i32> @insert_mem_and_zero_v4i32(i32* %ptr) {
; SSE-LABEL: insert_mem_and_zero_v4i32:
; SSE:       # %bb.0:
; SSE-NEXT:    movss {{.*#+}} xmm0 = mem[0],zero,zero,zero
; SSE-NEXT:    retq
;
; AVX-LABEL: insert_mem_and_zero_v4i32:
; AVX:       # %bb.0:
; AVX-NEXT:    vmovss {{.*#+}} xmm0 = mem[0],zero,zero,zero
; AVX-NEXT:    retq
  %a = load i32, i32* %ptr
  %v = insertelement <4 x i32> undef, i32 %a, i32 0
  %shuffle = shufflevector <4 x i32> %v, <4 x i32> zeroinitializer, <4 x i32> <i32 0, i32 5, i32 6, i32 7>
  ret <4 x i32> %shuffle
}

define <4 x float> @insert_reg_and_zero_v4f32(float %a) {
; SSE2-LABEL: insert_reg_and_zero_v4f32:
; SSE2:       # %bb.0:
; SSE2-NEXT:    xorps %xmm1, %xmm1
; SSE2-NEXT:    movss {{.*#+}} xmm1 = xmm0[0],xmm1[1,2,3]
; SSE2-NEXT:    movaps %xmm1, %xmm0
; SSE2-NEXT:    retq
;
; SSE3-LABEL: insert_reg_and_zero_v4f32:
; SSE3:       # %bb.0:
; SSE3-NEXT:    xorps %xmm1, %xmm1
; SSE3-NEXT:    movss {{.*#+}} xmm1 = xmm0[0],xmm1[1,2,3]
; SSE3-NEXT:    movaps %xmm1, %xmm0
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: insert_reg_and_zero_v4f32:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    xorps %xmm1, %xmm1
; SSSE3-NEXT:    movss {{.*#+}} xmm1 = xmm0[0],xmm1[1,2,3]
; SSSE3-NEXT:    movaps %xmm1, %xmm0
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: insert_reg_and_zero_v4f32:
; SSE41:       # %bb.0:
; SSE41-NEXT:    xorps %xmm1, %xmm1
; SSE41-NEXT:    blendps {{.*#+}} xmm0 = xmm0[0],xmm1[1,2,3]
; SSE41-NEXT:    retq
;
; AVX-LABEL: insert_reg_and_zero_v4f32:
; AVX:       # %bb.0:
; AVX-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; AVX-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0],xmm1[1,2,3]
; AVX-NEXT:    retq
  %v = insertelement <4 x float> undef, float %a, i32 0
  %shuffle = shufflevector <4 x float> %v, <4 x float> zeroinitializer, <4 x i32> <i32 0, i32 5, i32 6, i32 7>
  ret <4 x float> %shuffle
}

define <4 x float> @insert_mem_and_zero_v4f32(float* %ptr) {
; SSE-LABEL: insert_mem_and_zero_v4f32:
; SSE:       # %bb.0:
; SSE-NEXT:    movss {{.*#+}} xmm0 = mem[0],zero,zero,zero
; SSE-NEXT:    retq
;
; AVX-LABEL: insert_mem_and_zero_v4f32:
; AVX:       # %bb.0:
; AVX-NEXT:    vmovss {{.*#+}} xmm0 = mem[0],zero,zero,zero
; AVX-NEXT:    retq
  %a = load float, float* %ptr
  %v = insertelement <4 x float> undef, float %a, i32 0
  %shuffle = shufflevector <4 x float> %v, <4 x float> zeroinitializer, <4 x i32> <i32 0, i32 5, i32 6, i32 7>
  ret <4 x float> %shuffle
}

define <4 x i32> @insert_reg_lo_v4i32(i64 %a, <4 x i32> %b) {
; SSE2-LABEL: insert_reg_lo_v4i32:
; SSE2:       # %bb.0:
; SSE2-NEXT:    movq %rdi, %xmm1
; SSE2-NEXT:    movsd {{.*#+}} xmm0 = xmm1[0],xmm0[1]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: insert_reg_lo_v4i32:
; SSE3:       # %bb.0:
; SSE3-NEXT:    movq %rdi, %xmm1
; SSE3-NEXT:    movsd {{.*#+}} xmm0 = xmm1[0],xmm0[1]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: insert_reg_lo_v4i32:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    movq %rdi, %xmm1
; SSSE3-NEXT:    movsd {{.*#+}} xmm0 = xmm1[0],xmm0[1]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: insert_reg_lo_v4i32:
; SSE41:       # %bb.0:
; SSE41-NEXT:    movq %rdi, %xmm1
; SSE41-NEXT:    pblendw {{.*#+}} xmm0 = xmm1[0,1,2,3],xmm0[4,5,6,7]
; SSE41-NEXT:    retq
;
; AVX1-LABEL: insert_reg_lo_v4i32:
; AVX1:       # %bb.0:
; AVX1-NEXT:    vmovq %rdi, %xmm1
; AVX1-NEXT:    vpblendw {{.*#+}} xmm0 = xmm1[0,1,2,3],xmm0[4,5,6,7]
; AVX1-NEXT:    retq
;
; AVX2OR512VL-LABEL: insert_reg_lo_v4i32:
; AVX2OR512VL:       # %bb.0:
; AVX2OR512VL-NEXT:    vmovq %rdi, %xmm1
; AVX2OR512VL-NEXT:    vpblendd {{.*#+}} xmm0 = xmm1[0,1],xmm0[2,3]
; AVX2OR512VL-NEXT:    retq
  %a.cast = bitcast i64 %a to <2 x i32>
  %v = shufflevector <2 x i32> %a.cast, <2 x i32> undef, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
  %shuffle = shufflevector <4 x i32> %v, <4 x i32> %b, <4 x i32> <i32 0, i32 1, i32 6, i32 7>
  ret <4 x i32> %shuffle
}

define <4 x i32> @insert_mem_lo_v4i32(<2 x i32>* %ptr, <4 x i32> %b) {
; SSE2-LABEL: insert_mem_lo_v4i32:
; SSE2:       # %bb.0:
; SSE2-NEXT:    movlps {{.*#+}} xmm0 = mem[0,1],xmm0[2,3]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: insert_mem_lo_v4i32:
; SSE3:       # %bb.0:
; SSE3-NEXT:    movlps {{.*#+}} xmm0 = mem[0,1],xmm0[2,3]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: insert_mem_lo_v4i32:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    movlps {{.*#+}} xmm0 = mem[0,1],xmm0[2,3]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: insert_mem_lo_v4i32:
; SSE41:       # %bb.0:
; SSE41-NEXT:    movsd {{.*#+}} xmm1 = mem[0],zero
; SSE41-NEXT:    blendps {{.*#+}} xmm0 = xmm1[0,1],xmm0[2,3]
; SSE41-NEXT:    retq
;
; AVX-LABEL: insert_mem_lo_v4i32:
; AVX:       # %bb.0:
; AVX-NEXT:    vmovsd {{.*#+}} xmm1 = mem[0],zero
; AVX-NEXT:    vblendps {{.*#+}} xmm0 = xmm1[0,1],xmm0[2,3]
; AVX-NEXT:    retq
  %a = load <2 x i32>, <2 x i32>* %ptr
  %v = shufflevector <2 x i32> %a, <2 x i32> undef, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
  %shuffle = shufflevector <4 x i32> %v, <4 x i32> %b, <4 x i32> <i32 0, i32 1, i32 6, i32 7>
  ret <4 x i32> %shuffle
}

define <4 x i32> @insert_reg_hi_v4i32(i64 %a, <4 x i32> %b) {
; SSE-LABEL: insert_reg_hi_v4i32:
; SSE:       # %bb.0:
; SSE-NEXT:    movq %rdi, %xmm1
; SSE-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; SSE-NEXT:    retq
;
; AVX-LABEL: insert_reg_hi_v4i32:
; AVX:       # %bb.0:
; AVX-NEXT:    vmovq %rdi, %xmm1
; AVX-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX-NEXT:    retq
  %a.cast = bitcast i64 %a to <2 x i32>
  %v = shufflevector <2 x i32> %a.cast, <2 x i32> undef, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
  %shuffle = shufflevector <4 x i32> %v, <4 x i32> %b, <4 x i32> <i32 4, i32 5, i32 0, i32 1>
  ret <4 x i32> %shuffle
}

define <4 x i32> @insert_mem_hi_v4i32(<2 x i32>* %ptr, <4 x i32> %b) {
; SSE-LABEL: insert_mem_hi_v4i32:
; SSE:       # %bb.0:
; SSE-NEXT:    movsd {{.*#+}} xmm1 = mem[0],zero
; SSE-NEXT:    movlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; SSE-NEXT:    retq
;
; AVX-LABEL: insert_mem_hi_v4i32:
; AVX:       # %bb.0:
; AVX-NEXT:    vmovsd {{.*#+}} xmm1 = mem[0],zero
; AVX-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX-NEXT:    retq
  %a = load <2 x i32>, <2 x i32>* %ptr
  %v = shufflevector <2 x i32> %a, <2 x i32> undef, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
  %shuffle = shufflevector <4 x i32> %v, <4 x i32> %b, <4 x i32> <i32 4, i32 5, i32 0, i32 1>
  ret <4 x i32> %shuffle
}

define <4 x float> @insert_reg_lo_v4f32(double %a, <4 x float> %b) {
; SSE2-LABEL: insert_reg_lo_v4f32:
; SSE2:       # %bb.0:
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,3]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: insert_reg_lo_v4f32:
; SSE3:       # %bb.0:
; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,3]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: insert_reg_lo_v4f32:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,3]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: insert_reg_lo_v4f32:
; SSE41:       # %bb.0:
; SSE41-NEXT:    blendps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,3]
; SSE41-NEXT:    retq
;
; AVX-LABEL: insert_reg_lo_v4f32:
; AVX:       # %bb.0:
; AVX-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,3]
; AVX-NEXT:    retq
  %a.cast = bitcast double %a to <2 x float>
  %v = shufflevector <2 x float> %a.cast, <2 x float> undef, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
  %shuffle = shufflevector <4 x float> %v, <4 x float> %b, <4 x i32> <i32 0, i32 1, i32 6, i32 7>
  ret <4 x float> %shuffle
}

define <4 x float> @insert_mem_lo_v4f32(<2 x float>* %ptr, <4 x float> %b) {
; SSE-LABEL: insert_mem_lo_v4f32:
; SSE:       # %bb.0:
; SSE-NEXT:    movlps {{.*#+}} xmm0 = mem[0,1],xmm0[2,3]
; SSE-NEXT:    retq
;
; AVX-LABEL: insert_mem_lo_v4f32:
; AVX:       # %bb.0:
; AVX-NEXT:    vmovlps {{.*#+}} xmm0 = mem[0,1],xmm0[2,3]
; AVX-NEXT:    retq
  %a = load <2 x float>, <2 x float>* %ptr
  %v = shufflevector <2 x float> %a, <2 x float> undef, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
  %shuffle = shufflevector <4 x float> %v, <4 x float> %b, <4 x i32> <i32 0, i32 1, i32 6, i32 7>
  ret <4 x float> %shuffle
}

define <4 x float> @insert_reg_hi_v4f32(double %a, <4 x float> %b) {
; SSE-LABEL: insert_reg_hi_v4f32:
; SSE:       # %bb.0:
; SSE-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE-NEXT:    movaps %xmm1, %xmm0
; SSE-NEXT:    retq
;
; AVX-LABEL: insert_reg_hi_v4f32:
; AVX:       # %bb.0:
; AVX-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm1[0],xmm0[0]
; AVX-NEXT:    retq
  %a.cast = bitcast double %a to <2 x float>
  %v = shufflevector <2 x float> %a.cast, <2 x float> undef, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
  %shuffle = shufflevector <4 x float> %v, <4 x float> %b, <4 x i32> <i32 4, i32 5, i32 0, i32 1>
  ret <4 x float> %shuffle
}

define <4 x float> @insert_mem_hi_v4f32(<2 x float>* %ptr, <4 x float> %b) {
; SSE-LABEL: insert_mem_hi_v4f32:
; SSE:       # %bb.0:
; SSE-NEXT:    movhps {{.*#+}} xmm0 = xmm0[0,1],mem[0,1]
; SSE-NEXT:    retq
;
; AVX-LABEL: insert_mem_hi_v4f32:
; AVX:       # %bb.0:
; AVX-NEXT:    vmovhps {{.*#+}} xmm0 = xmm0[0,1],mem[0,1]
; AVX-NEXT:    retq
  %a = load <2 x float>, <2 x float>* %ptr
  %v = shufflevector <2 x float> %a, <2 x float> undef, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
  %shuffle = shufflevector <4 x float> %v, <4 x float> %b, <4 x i32> <i32 4, i32 5, i32 0, i32 1>
  ret <4 x float> %shuffle
}

; PR21137
define <4 x float> @shuffle_mem_v4f32_3210(<4 x float>* %ptr) {
; SSE-LABEL: shuffle_mem_v4f32_3210:
; SSE:       # %bb.0:
; SSE-NEXT:    movaps (%rdi), %xmm0
; SSE-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,2,1,0]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_mem_v4f32_3210:
; AVX:       # %bb.0:
; AVX-NEXT:    vpermilps {{.*#+}} xmm0 = mem[3,2,1,0]
; AVX-NEXT:    retq
  %a = load <4 x float>, <4 x float>* %ptr
  %shuffle = shufflevector <4 x float> %a, <4 x float> undef, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
  ret <4 x float> %shuffle
}

define <4 x i32> @insert_dup_mem_v4i32(i32* %ptr) {
; SSE-LABEL: insert_dup_mem_v4i32:
; SSE:       # %bb.0:
; SSE-NEXT:    movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,0,0,0]
; SSE-NEXT:    retq
;
; AVX-LABEL: insert_dup_mem_v4i32:
; AVX:       # %bb.0:
; AVX-NEXT:    vbroadcastss (%rdi), %xmm0
; AVX-NEXT:    retq
  %tmp = load i32, i32* %ptr, align 4
  %tmp1 = insertelement <4 x i32> zeroinitializer, i32 %tmp, i32 0
  %tmp2 = shufflevector <4 x i32> %tmp1, <4 x i32> undef, <4 x i32> zeroinitializer
  ret <4 x i32> %tmp2
}

; PR41249
define <4 x float> @shuffle_mem_pmovzx_v4f32(<2 x float>* %p0, <4 x float>* %p1) {
; SSE-LABEL: shuffle_mem_pmovzx_v4f32:
; SSE:       # %bb.0:
; SSE-NEXT:    movsd {{.*#+}} xmm0 = mem[0],zero
; SSE-NEXT:    xorps %xmm1, %xmm1
; SSE-NEXT:    movaps %xmm0, %xmm2
; SSE-NEXT:    unpcklps {{.*#+}} xmm2 = xmm2[0],xmm1[0],xmm2[1],xmm1[1]
; SSE-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,0,0,0]
; SSE-NEXT:    movaps %xmm2, (%rsi)
; SSE-NEXT:    retq
;
; AVX1-LABEL: shuffle_mem_pmovzx_v4f32:
; AVX1:       # %bb.0:
; AVX1-NEXT:    vmovsd {{.*#+}} xmm0 = mem[0],zero
; AVX1-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; AVX1-NEXT:    vunpcklps {{.*#+}} xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
; AVX1-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[0,0,0,0]
; AVX1-NEXT:    vmovaps %xmm1, (%rsi)
; AVX1-NEXT:    retq
;
; AVX2OR512VL-LABEL: shuffle_mem_pmovzx_v4f32:
; AVX2OR512VL:       # %bb.0:
; AVX2OR512VL-NEXT:    vmovsd {{.*#+}} xmm0 = mem[0],zero
; AVX2OR512VL-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; AVX2OR512VL-NEXT:    vunpcklps {{.*#+}} xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
; AVX2OR512VL-NEXT:    vbroadcastss %xmm0, %xmm0
; AVX2OR512VL-NEXT:    vmovaps %xmm1, (%rsi)
; AVX2OR512VL-NEXT:    retq
  %1 = load <2 x float>, <2 x float>* %p0
  %2 = shufflevector <2 x float> %1, <2 x float> undef, <4 x i32> <i32 undef, i32 undef, i32 0, i32 1>
  %3 = shufflevector <4 x float> %2, <4 x float> <float undef, float undef, float 0.000000e+00, float 0.000000e+00>, <4 x i32> <i32 2, i32 6, i32 3, i32 7>
  %4 = shufflevector <2 x float> %1, <2 x float> undef, <4 x i32> zeroinitializer
  store <4 x float> %3, <4 x float>* %p1
  ret <4 x float> %4
}

;
; Shuffle to logical bit shifts
;

define <4 x i32> @shuffle_v4i32_z0zX(<4 x i32> %a) {
; SSE-LABEL: shuffle_v4i32_z0zX:
; SSE:       # %bb.0:
; SSE-NEXT:    psllq $32, %xmm0
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_z0zX:
; AVX:       # %bb.0:
; AVX-NEXT:    vpsllq $32, %xmm0, %xmm0
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> zeroinitializer, <4 x i32> <i32 4, i32 0, i32 4, i32 undef>
  ret <4 x i32> %shuffle
}

define <4 x i32> @shuffle_v4i32_1z3z(<4 x i32> %a) {
; SSE-LABEL: shuffle_v4i32_1z3z:
; SSE:       # %bb.0:
; SSE-NEXT:    psrlq $32, %xmm0
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_v4i32_1z3z:
; AVX:       # %bb.0:
; AVX-NEXT:    vpsrlq $32, %xmm0, %xmm0
; AVX-NEXT:    retq
  %shuffle = shufflevector <4 x i32> %a, <4 x i32> zeroinitializer, <4 x i32> <i32 1, i32 4, i32 3, i32 4>
  ret <4 x i32> %shuffle
}

define <4 x float> @shuffle_mem_v4f32_0145(<4 x float> %a, <4 x float>* %pb) {
; SSE-LABEL: shuffle_mem_v4f32_0145:
; SSE:       # %bb.0:
; SSE-NEXT:    movhps {{.*#+}} xmm0 = xmm0[0,1],mem[0,1]
; SSE-NEXT:    retq
;
; AVX-LABEL: shuffle_mem_v4f32_0145:
; AVX:       # %bb.0:
; AVX-NEXT:    vunpcklpd {{.*#+}} xmm0 = xmm0[0],mem[0]
; AVX-NEXT:    retq
  %b = load <4 x float>, <4 x float>* %pb, align 1
  %shuffle = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 1, i32 4, i32 5>
  ret <4 x float> %shuffle
}

define <4 x float> @shuffle_mem_v4f32_4523(<4 x float> %a, <4 x float>* %pb) {
; SSE2-LABEL: shuffle_mem_v4f32_4523:
; SSE2:       # %bb.0:
; SSE2-NEXT:    movlps {{.*#+}} xmm0 = mem[0,1],xmm0[2,3]
; SSE2-NEXT:    retq
;
; SSE3-LABEL: shuffle_mem_v4f32_4523:
; SSE3:       # %bb.0:
; SSE3-NEXT:    movlps {{.*#+}} xmm0 = mem[0,1],xmm0[2,3]
; SSE3-NEXT:    retq
;
; SSSE3-LABEL: shuffle_mem_v4f32_4523:
; SSSE3:       # %bb.0:
; SSSE3-NEXT:    movlps {{.*#+}} xmm0 = mem[0,1],xmm0[2,3]
; SSSE3-NEXT:    retq
;
; SSE41-LABEL: shuffle_mem_v4f32_4523:
; SSE41:       # %bb.0:
; SSE41-NEXT:    movups (%rdi), %xmm1
; SSE41-NEXT:    blendps {{.*#+}} xmm0 = xmm1[0,1],xmm0[2,3]
; SSE41-NEXT:    retq
;
; AVX-LABEL: shuffle_mem_v4f32_4523:
; AVX:       # %bb.0:
; AVX-NEXT:    vblendps {{.*#+}} xmm0 = mem[0,1],xmm0[2,3]
; AVX-NEXT:    retq
  %b = load <4 x float>, <4 x float>* %pb, align 1
  %shuffle = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 4, i32 5, i32 2, i32 3>
  ret <4 x float> %shuffle
}

define  <4 x float> @shuffle_mem_v4f32_0624(<4 x float> %a0, <4 x float>* %a1) {
; SSE-LABEL: shuffle_mem_v4f32_0624:
; SSE:       # %bb.0:
; SSE-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,0],mem[0,2]
; SSE-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,0,3,1]
; SSE-NEXT:    retq
;
; AVX1OR2-LABEL: shuffle_mem_v4f32_0624:
; AVX1OR2:       # %bb.0:
; AVX1OR2-NEXT:    vshufps {{.*#+}} xmm0 = xmm0[2,0],mem[0,2]
; AVX1OR2-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[2,0,3,1]
; AVX1OR2-NEXT:    retq
;
; AVX512VL-LABEL: shuffle_mem_v4f32_0624:
; AVX512VL:       # %bb.0:
; AVX512VL-NEXT:    vmovaps (%rdi), %xmm2
; AVX512VL-NEXT:    vmovaps {{.*#+}} xmm1 = [0,6,2,4]
; AVX512VL-NEXT:    vpermi2ps %xmm0, %xmm2, %xmm1
; AVX512VL-NEXT:    vmovaps %xmm1, %xmm0
; AVX512VL-NEXT:    retq
  %1 = load <4 x float>, <4 x float>* %a1
  %2 = shufflevector <4 x float> %1, <4 x float> %a0, <4 x i32> <i32 0, i32 6, i32 2, i32 4>
  ret <4 x float> %2
}

define  <4 x float> @shuffle_mem_v4f32_4760(<4 x float> %a0, <4 x float>* %a1) {
; SSE-LABEL: shuffle_mem_v4f32_4760:
; SSE:       # %bb.0:
; SSE-NEXT:    movaps %xmm0, %xmm1
; SSE-NEXT:    shufps {{.*#+}} xmm1 = xmm1[2,0],mem[0,0]
; SSE-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,3],xmm1[0,2]
; SSE-NEXT:    retq
;
; AVX1OR2-LABEL: shuffle_mem_v4f32_4760:
; AVX1OR2:       # %bb.0:
; AVX1OR2-NEXT:    vshufps {{.*#+}} xmm1 = xmm0[2,0],mem[0,0]
; AVX1OR2-NEXT:    vshufps {{.*#+}} xmm0 = xmm0[0,3],xmm1[0,2]
; AVX1OR2-NEXT:    retq
;
; AVX512VL-LABEL: shuffle_mem_v4f32_4760:
; AVX512VL:       # %bb.0:
; AVX512VL-NEXT:    vmovaps {{.*#+}} xmm1 = [0,3,2,4]
; AVX512VL-NEXT:    vpermt2ps (%rdi), %xmm1, %xmm0
; AVX512VL-NEXT:    retq
  %1 = load <4 x float>, <4 x float>* %a1
  %2 = shufflevector <4 x float> %1, <4 x float> %a0, <4 x i32> <i32 4, i32 7, i32 6, i32 0>
  ret <4 x float> %2
}