llvm/llvm/test/CodeGen/X86/vector-interleaved-store-i64-stride-2.ll

; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc < %s -mtriple=x86_64-- -mattr=+sse2 | FileCheck %s --check-prefixes=SSE
; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx  | FileCheck %s --check-prefixes=AVX
; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX2
; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX2-FP
; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX2-FCP
; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX512
; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX512-FCP
; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq | FileCheck %s --check-prefixes=AVX512DQ
; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX512DQ-FCP
; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw | FileCheck %s --check-prefixes=AVX512BW
; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX512BW-FCP
; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefixes=AVX512DQ-BW
; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl,+avx512dq,+avx512bw,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=AVX512DQ-BW-FCP

; These patterns are produced by LoopVectorizer for interleaved stores.

define void @store_i64_stride2_vf2(ptr %in.vecptr0, ptr %in.vecptr1, ptr %out.vec) nounwind {
; SSE-LABEL: store_i64_stride2_vf2:
; SSE:       # %bb.0:
; SSE-NEXT:    movaps (%rdi), %xmm0
; SSE-NEXT:    movaps (%rsi), %xmm1
; SSE-NEXT:    movaps %xmm0, %xmm2
; SSE-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm1[0]
; SSE-NEXT:    unpckhpd {{.*#+}} xmm0 = xmm0[1],xmm1[1]
; SSE-NEXT:    movaps %xmm0, 16(%rdx)
; SSE-NEXT:    movaps %xmm2, (%rdx)
; SSE-NEXT:    retq
;
; AVX-LABEL: store_i64_stride2_vf2:
; AVX:       # %bb.0:
; AVX-NEXT:    vbroadcastf128 {{.*#+}} ymm0 = mem[0,1,0,1]
; AVX-NEXT:    vbroadcastf128 {{.*#+}} ymm1 = mem[0,1,0,1]
; AVX-NEXT:    vshufpd {{.*#+}} ymm0 = ymm1[0],ymm0[0],ymm1[3],ymm0[3]
; AVX-NEXT:    vmovapd %ymm0, (%rdx)
; AVX-NEXT:    vzeroupper
; AVX-NEXT:    retq
;
; AVX2-LABEL: store_i64_stride2_vf2:
; AVX2:       # %bb.0:
; AVX2-NEXT:    vmovaps (%rdi), %xmm0
; AVX2-NEXT:    vinsertf128 $1, (%rsi), %ymm0, %ymm0
; AVX2-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,2,1,3]
; AVX2-NEXT:    vmovaps %ymm0, (%rdx)
; AVX2-NEXT:    vzeroupper
; AVX2-NEXT:    retq
;
; AVX2-FP-LABEL: store_i64_stride2_vf2:
; AVX2-FP:       # %bb.0:
; AVX2-FP-NEXT:    vmovaps (%rdi), %xmm0
; AVX2-FP-NEXT:    vinsertf128 $1, (%rsi), %ymm0, %ymm0
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,2,1,3]
; AVX2-FP-NEXT:    vmovaps %ymm0, (%rdx)
; AVX2-FP-NEXT:    vzeroupper
; AVX2-FP-NEXT:    retq
;
; AVX2-FCP-LABEL: store_i64_stride2_vf2:
; AVX2-FCP:       # %bb.0:
; AVX2-FCP-NEXT:    vmovaps (%rdi), %xmm0
; AVX2-FCP-NEXT:    vinsertf128 $1, (%rsi), %ymm0, %ymm0
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,2,1,3]
; AVX2-FCP-NEXT:    vmovaps %ymm0, (%rdx)
; AVX2-FCP-NEXT:    vzeroupper
; AVX2-FCP-NEXT:    retq
;
; AVX512-LABEL: store_i64_stride2_vf2:
; AVX512:       # %bb.0:
; AVX512-NEXT:    vmovaps (%rdi), %xmm0
; AVX512-NEXT:    vinsertf128 $1, (%rsi), %ymm0, %ymm0
; AVX512-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,2,1,3]
; AVX512-NEXT:    vmovaps %ymm0, (%rdx)
; AVX512-NEXT:    vzeroupper
; AVX512-NEXT:    retq
;
; AVX512-FCP-LABEL: store_i64_stride2_vf2:
; AVX512-FCP:       # %bb.0:
; AVX512-FCP-NEXT:    vmovaps (%rdi), %xmm0
; AVX512-FCP-NEXT:    vinsertf128 $1, (%rsi), %ymm0, %ymm0
; AVX512-FCP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,2,1,3]
; AVX512-FCP-NEXT:    vmovaps %ymm0, (%rdx)
; AVX512-FCP-NEXT:    vzeroupper
; AVX512-FCP-NEXT:    retq
;
; AVX512DQ-LABEL: store_i64_stride2_vf2:
; AVX512DQ:       # %bb.0:
; AVX512DQ-NEXT:    vmovaps (%rdi), %xmm0
; AVX512DQ-NEXT:    vinsertf128 $1, (%rsi), %ymm0, %ymm0
; AVX512DQ-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,2,1,3]
; AVX512DQ-NEXT:    vmovaps %ymm0, (%rdx)
; AVX512DQ-NEXT:    vzeroupper
; AVX512DQ-NEXT:    retq
;
; AVX512DQ-FCP-LABEL: store_i64_stride2_vf2:
; AVX512DQ-FCP:       # %bb.0:
; AVX512DQ-FCP-NEXT:    vmovaps (%rdi), %xmm0
; AVX512DQ-FCP-NEXT:    vinsertf128 $1, (%rsi), %ymm0, %ymm0
; AVX512DQ-FCP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,2,1,3]
; AVX512DQ-FCP-NEXT:    vmovaps %ymm0, (%rdx)
; AVX512DQ-FCP-NEXT:    vzeroupper
; AVX512DQ-FCP-NEXT:    retq
;
; AVX512BW-LABEL: store_i64_stride2_vf2:
; AVX512BW:       # %bb.0:
; AVX512BW-NEXT:    vmovaps (%rdi), %xmm0
; AVX512BW-NEXT:    vinsertf128 $1, (%rsi), %ymm0, %ymm0
; AVX512BW-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,2,1,3]
; AVX512BW-NEXT:    vmovaps %ymm0, (%rdx)
; AVX512BW-NEXT:    vzeroupper
; AVX512BW-NEXT:    retq
;
; AVX512BW-FCP-LABEL: store_i64_stride2_vf2:
; AVX512BW-FCP:       # %bb.0:
; AVX512BW-FCP-NEXT:    vmovaps (%rdi), %xmm0
; AVX512BW-FCP-NEXT:    vinsertf128 $1, (%rsi), %ymm0, %ymm0
; AVX512BW-FCP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,2,1,3]
; AVX512BW-FCP-NEXT:    vmovaps %ymm0, (%rdx)
; AVX512BW-FCP-NEXT:    vzeroupper
; AVX512BW-FCP-NEXT:    retq
;
; AVX512DQ-BW-LABEL: store_i64_stride2_vf2:
; AVX512DQ-BW:       # %bb.0:
; AVX512DQ-BW-NEXT:    vmovaps (%rdi), %xmm0
; AVX512DQ-BW-NEXT:    vinsertf128 $1, (%rsi), %ymm0, %ymm0
; AVX512DQ-BW-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,2,1,3]
; AVX512DQ-BW-NEXT:    vmovaps %ymm0, (%rdx)
; AVX512DQ-BW-NEXT:    vzeroupper
; AVX512DQ-BW-NEXT:    retq
;
; AVX512DQ-BW-FCP-LABEL: store_i64_stride2_vf2:
; AVX512DQ-BW-FCP:       # %bb.0:
; AVX512DQ-BW-FCP-NEXT:    vmovaps (%rdi), %xmm0
; AVX512DQ-BW-FCP-NEXT:    vinsertf128 $1, (%rsi), %ymm0, %ymm0
; AVX512DQ-BW-FCP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,2,1,3]
; AVX512DQ-BW-FCP-NEXT:    vmovaps %ymm0, (%rdx)
; AVX512DQ-BW-FCP-NEXT:    vzeroupper
; AVX512DQ-BW-FCP-NEXT:    retq
  %in.vec0 = load <2 x i64>, ptr %in.vecptr0, align 64
  %in.vec1 = load <2 x i64>, ptr %in.vecptr1, align 64
  %1 = shufflevector <2 x i64> %in.vec0, <2 x i64> %in.vec1, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
  %interleaved.vec = shufflevector <4 x i64> %1, <4 x i64> poison, <4 x i32> <i32 0, i32 2, i32 1, i32 3>
  store <4 x i64> %interleaved.vec, ptr %out.vec, align 64
  ret void
}

define void @store_i64_stride2_vf4(ptr %in.vecptr0, ptr %in.vecptr1, ptr %out.vec) nounwind {
; SSE-LABEL: store_i64_stride2_vf4:
; SSE:       # %bb.0:
; SSE-NEXT:    movaps (%rdi), %xmm0
; SSE-NEXT:    movaps 16(%rdi), %xmm1
; SSE-NEXT:    movaps (%rsi), %xmm2
; SSE-NEXT:    movaps 16(%rsi), %xmm3
; SSE-NEXT:    movaps %xmm0, %xmm4
; SSE-NEXT:    unpckhpd {{.*#+}} xmm4 = xmm4[1],xmm2[1]
; SSE-NEXT:    movlhps {{.*#+}} xmm0 = xmm0[0],xmm2[0]
; SSE-NEXT:    movaps %xmm1, %xmm2
; SSE-NEXT:    unpckhpd {{.*#+}} xmm2 = xmm2[1],xmm3[1]
; SSE-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm3[0]
; SSE-NEXT:    movaps %xmm1, 32(%rdx)
; SSE-NEXT:    movaps %xmm2, 48(%rdx)
; SSE-NEXT:    movaps %xmm0, (%rdx)
; SSE-NEXT:    movaps %xmm4, 16(%rdx)
; SSE-NEXT:    retq
;
; AVX-LABEL: store_i64_stride2_vf4:
; AVX:       # %bb.0:
; AVX-NEXT:    vmovaps (%rsi), %xmm0
; AVX-NEXT:    vmovaps (%rdi), %xmm1
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm2 = xmm1[1],xmm0[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm1[0],xmm0[0]
; AVX-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm1 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm2 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm1 = ymm2[0],ymm1[0],ymm2[3],ymm1[3]
; AVX-NEXT:    vmovapd %ymm1, 32(%rdx)
; AVX-NEXT:    vmovapd %ymm0, (%rdx)
; AVX-NEXT:    vzeroupper
; AVX-NEXT:    retq
;
; AVX2-LABEL: store_i64_stride2_vf4:
; AVX2:       # %bb.0:
; AVX2-NEXT:    vmovaps (%rdi), %ymm0
; AVX2-NEXT:    vmovaps (%rsi), %ymm1
; AVX2-NEXT:    vpermpd {{.*#+}} ymm2 = ymm1[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm3 = ymm0[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm2 = ymm3[0,1],ymm2[2,3],ymm3[4,5],ymm2[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3],ymm0[4,5],ymm1[6,7]
; AVX2-NEXT:    vmovaps %ymm0, 32(%rdx)
; AVX2-NEXT:    vmovaps %ymm2, (%rdx)
; AVX2-NEXT:    vzeroupper
; AVX2-NEXT:    retq
;
; AVX2-FP-LABEL: store_i64_stride2_vf4:
; AVX2-FP:       # %bb.0:
; AVX2-FP-NEXT:    vmovaps (%rdi), %ymm0
; AVX2-FP-NEXT:    vmovaps (%rsi), %ymm1
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm1[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm3 = ymm0[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm2 = ymm3[0,1],ymm2[2,3],ymm3[4,5],ymm2[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3],ymm0[4,5],ymm1[6,7]
; AVX2-FP-NEXT:    vmovaps %ymm0, 32(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm2, (%rdx)
; AVX2-FP-NEXT:    vzeroupper
; AVX2-FP-NEXT:    retq
;
; AVX2-FCP-LABEL: store_i64_stride2_vf4:
; AVX2-FCP:       # %bb.0:
; AVX2-FCP-NEXT:    vmovaps (%rdi), %ymm0
; AVX2-FCP-NEXT:    vmovaps (%rsi), %ymm1
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm1[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm3 = ymm0[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm2 = ymm3[0,1],ymm2[2,3],ymm3[4,5],ymm2[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3],ymm0[4,5],ymm1[6,7]
; AVX2-FCP-NEXT:    vmovaps %ymm0, 32(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm2, (%rdx)
; AVX2-FCP-NEXT:    vzeroupper
; AVX2-FCP-NEXT:    retq
;
; AVX512-LABEL: store_i64_stride2_vf4:
; AVX512:       # %bb.0:
; AVX512-NEXT:    vmovaps (%rdi), %ymm0
; AVX512-NEXT:    vinsertf64x4 $1, (%rsi), %zmm0, %zmm0
; AVX512-NEXT:    vmovaps {{.*#+}} zmm1 = [0,4,1,5,2,6,3,7]
; AVX512-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
; AVX512-NEXT:    vmovaps %zmm0, (%rdx)
; AVX512-NEXT:    vzeroupper
; AVX512-NEXT:    retq
;
; AVX512-FCP-LABEL: store_i64_stride2_vf4:
; AVX512-FCP:       # %bb.0:
; AVX512-FCP-NEXT:    vmovaps (%rdi), %ymm0
; AVX512-FCP-NEXT:    vinsertf64x4 $1, (%rsi), %zmm0, %zmm0
; AVX512-FCP-NEXT:    vmovaps {{.*#+}} zmm1 = [0,4,1,5,2,6,3,7]
; AVX512-FCP-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
; AVX512-FCP-NEXT:    vmovaps %zmm0, (%rdx)
; AVX512-FCP-NEXT:    vzeroupper
; AVX512-FCP-NEXT:    retq
;
; AVX512DQ-LABEL: store_i64_stride2_vf4:
; AVX512DQ:       # %bb.0:
; AVX512DQ-NEXT:    vmovaps (%rdi), %ymm0
; AVX512DQ-NEXT:    vinsertf64x4 $1, (%rsi), %zmm0, %zmm0
; AVX512DQ-NEXT:    vmovaps {{.*#+}} zmm1 = [0,4,1,5,2,6,3,7]
; AVX512DQ-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
; AVX512DQ-NEXT:    vmovaps %zmm0, (%rdx)
; AVX512DQ-NEXT:    vzeroupper
; AVX512DQ-NEXT:    retq
;
; AVX512DQ-FCP-LABEL: store_i64_stride2_vf4:
; AVX512DQ-FCP:       # %bb.0:
; AVX512DQ-FCP-NEXT:    vmovaps (%rdi), %ymm0
; AVX512DQ-FCP-NEXT:    vinsertf64x4 $1, (%rsi), %zmm0, %zmm0
; AVX512DQ-FCP-NEXT:    vmovaps {{.*#+}} zmm1 = [0,4,1,5,2,6,3,7]
; AVX512DQ-FCP-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
; AVX512DQ-FCP-NEXT:    vmovaps %zmm0, (%rdx)
; AVX512DQ-FCP-NEXT:    vzeroupper
; AVX512DQ-FCP-NEXT:    retq
;
; AVX512BW-LABEL: store_i64_stride2_vf4:
; AVX512BW:       # %bb.0:
; AVX512BW-NEXT:    vmovaps (%rdi), %ymm0
; AVX512BW-NEXT:    vinsertf64x4 $1, (%rsi), %zmm0, %zmm0
; AVX512BW-NEXT:    vmovaps {{.*#+}} zmm1 = [0,4,1,5,2,6,3,7]
; AVX512BW-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
; AVX512BW-NEXT:    vmovaps %zmm0, (%rdx)
; AVX512BW-NEXT:    vzeroupper
; AVX512BW-NEXT:    retq
;
; AVX512BW-FCP-LABEL: store_i64_stride2_vf4:
; AVX512BW-FCP:       # %bb.0:
; AVX512BW-FCP-NEXT:    vmovaps (%rdi), %ymm0
; AVX512BW-FCP-NEXT:    vinsertf64x4 $1, (%rsi), %zmm0, %zmm0
; AVX512BW-FCP-NEXT:    vmovaps {{.*#+}} zmm1 = [0,4,1,5,2,6,3,7]
; AVX512BW-FCP-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
; AVX512BW-FCP-NEXT:    vmovaps %zmm0, (%rdx)
; AVX512BW-FCP-NEXT:    vzeroupper
; AVX512BW-FCP-NEXT:    retq
;
; AVX512DQ-BW-LABEL: store_i64_stride2_vf4:
; AVX512DQ-BW:       # %bb.0:
; AVX512DQ-BW-NEXT:    vmovaps (%rdi), %ymm0
; AVX512DQ-BW-NEXT:    vinsertf64x4 $1, (%rsi), %zmm0, %zmm0
; AVX512DQ-BW-NEXT:    vmovaps {{.*#+}} zmm1 = [0,4,1,5,2,6,3,7]
; AVX512DQ-BW-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
; AVX512DQ-BW-NEXT:    vmovaps %zmm0, (%rdx)
; AVX512DQ-BW-NEXT:    vzeroupper
; AVX512DQ-BW-NEXT:    retq
;
; AVX512DQ-BW-FCP-LABEL: store_i64_stride2_vf4:
; AVX512DQ-BW-FCP:       # %bb.0:
; AVX512DQ-BW-FCP-NEXT:    vmovaps (%rdi), %ymm0
; AVX512DQ-BW-FCP-NEXT:    vinsertf64x4 $1, (%rsi), %zmm0, %zmm0
; AVX512DQ-BW-FCP-NEXT:    vmovaps {{.*#+}} zmm1 = [0,4,1,5,2,6,3,7]
; AVX512DQ-BW-FCP-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
; AVX512DQ-BW-FCP-NEXT:    vmovaps %zmm0, (%rdx)
; AVX512DQ-BW-FCP-NEXT:    vzeroupper
; AVX512DQ-BW-FCP-NEXT:    retq
  %in.vec0 = load <4 x i64>, ptr %in.vecptr0, align 64
  %in.vec1 = load <4 x i64>, ptr %in.vecptr1, align 64
  %1 = shufflevector <4 x i64> %in.vec0, <4 x i64> %in.vec1, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
  %interleaved.vec = shufflevector <8 x i64> %1, <8 x i64> poison, <8 x i32> <i32 0, i32 4, i32 1, i32 5, i32 2, i32 6, i32 3, i32 7>
  store <8 x i64> %interleaved.vec, ptr %out.vec, align 64
  ret void
}

define void @store_i64_stride2_vf8(ptr %in.vecptr0, ptr %in.vecptr1, ptr %out.vec) nounwind {
; SSE-LABEL: store_i64_stride2_vf8:
; SSE:       # %bb.0:
; SSE-NEXT:    movaps (%rdi), %xmm0
; SSE-NEXT:    movaps 16(%rdi), %xmm1
; SSE-NEXT:    movaps 32(%rdi), %xmm2
; SSE-NEXT:    movaps 48(%rdi), %xmm3
; SSE-NEXT:    movaps (%rsi), %xmm4
; SSE-NEXT:    movaps 16(%rsi), %xmm5
; SSE-NEXT:    movaps 32(%rsi), %xmm6
; SSE-NEXT:    movaps 48(%rsi), %xmm7
; SSE-NEXT:    movaps %xmm0, %xmm8
; SSE-NEXT:    unpckhpd {{.*#+}} xmm8 = xmm8[1],xmm4[1]
; SSE-NEXT:    movlhps {{.*#+}} xmm0 = xmm0[0],xmm4[0]
; SSE-NEXT:    movaps %xmm1, %xmm4
; SSE-NEXT:    unpckhpd {{.*#+}} xmm4 = xmm4[1],xmm5[1]
; SSE-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm5[0]
; SSE-NEXT:    movaps %xmm2, %xmm5
; SSE-NEXT:    unpckhpd {{.*#+}} xmm5 = xmm5[1],xmm6[1]
; SSE-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm6[0]
; SSE-NEXT:    movaps %xmm3, %xmm6
; SSE-NEXT:    unpckhpd {{.*#+}} xmm6 = xmm6[1],xmm7[1]
; SSE-NEXT:    movlhps {{.*#+}} xmm3 = xmm3[0],xmm7[0]
; SSE-NEXT:    movaps %xmm3, 96(%rdx)
; SSE-NEXT:    movaps %xmm6, 112(%rdx)
; SSE-NEXT:    movaps %xmm2, 64(%rdx)
; SSE-NEXT:    movaps %xmm5, 80(%rdx)
; SSE-NEXT:    movaps %xmm1, 32(%rdx)
; SSE-NEXT:    movaps %xmm4, 48(%rdx)
; SSE-NEXT:    movaps %xmm0, (%rdx)
; SSE-NEXT:    movaps %xmm8, 16(%rdx)
; SSE-NEXT:    retq
;
; AVX-LABEL: store_i64_stride2_vf8:
; AVX:       # %bb.0:
; AVX-NEXT:    vmovaps (%rsi), %xmm0
; AVX-NEXT:    vmovaps 32(%rsi), %xmm1
; AVX-NEXT:    vmovaps (%rdi), %xmm2
; AVX-NEXT:    vmovaps 32(%rdi), %xmm3
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm4 = xmm2[1],xmm0[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm2[0],xmm0[0]
; AVX-NEXT:    vinsertf128 $1, %xmm4, %ymm0, %ymm0
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm2 = xmm3[1],xmm1[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm1 = xmm3[0],xmm1[0]
; AVX-NEXT:    vinsertf128 $1, %xmm2, %ymm1, %ymm1
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm2 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm3 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm2 = ymm3[0],ymm2[0],ymm3[3],ymm2[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm3 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm4 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm3 = ymm4[0],ymm3[0],ymm4[3],ymm3[3]
; AVX-NEXT:    vmovapd %ymm3, 96(%rdx)
; AVX-NEXT:    vmovapd %ymm2, 32(%rdx)
; AVX-NEXT:    vmovaps %ymm1, 64(%rdx)
; AVX-NEXT:    vmovapd %ymm0, (%rdx)
; AVX-NEXT:    vzeroupper
; AVX-NEXT:    retq
;
; AVX2-LABEL: store_i64_stride2_vf8:
; AVX2:       # %bb.0:
; AVX2-NEXT:    vmovaps (%rdi), %ymm0
; AVX2-NEXT:    vmovaps 32(%rdi), %ymm1
; AVX2-NEXT:    vmovaps (%rsi), %ymm2
; AVX2-NEXT:    vmovaps 32(%rsi), %ymm3
; AVX2-NEXT:    vpermpd {{.*#+}} ymm4 = ymm2[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm5 = ymm0[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm4 = ymm5[0,1],ymm4[2,3],ymm5[4,5],ymm4[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm2 = ymm2[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm0 = ymm0[0,1],ymm2[2,3],ymm0[4,5],ymm2[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm2 = ymm3[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm5 = ymm1[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm2 = ymm5[0,1],ymm2[2,3],ymm5[4,5],ymm2[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm3 = ymm3[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm1 = ymm1[0,1],ymm3[2,3],ymm1[4,5],ymm3[6,7]
; AVX2-NEXT:    vmovaps %ymm1, 64(%rdx)
; AVX2-NEXT:    vmovaps %ymm2, 96(%rdx)
; AVX2-NEXT:    vmovaps %ymm0, (%rdx)
; AVX2-NEXT:    vmovaps %ymm4, 32(%rdx)
; AVX2-NEXT:    vzeroupper
; AVX2-NEXT:    retq
;
; AVX2-FP-LABEL: store_i64_stride2_vf8:
; AVX2-FP:       # %bb.0:
; AVX2-FP-NEXT:    vmovaps (%rdi), %ymm0
; AVX2-FP-NEXT:    vmovaps 32(%rdi), %ymm1
; AVX2-FP-NEXT:    vmovaps (%rsi), %ymm2
; AVX2-FP-NEXT:    vmovaps 32(%rsi), %ymm3
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm4 = ymm2[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm5 = ymm0[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm4 = ymm5[0,1],ymm4[2,3],ymm5[4,5],ymm4[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm2[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm0 = ymm0[0,1],ymm2[2,3],ymm0[4,5],ymm2[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm3[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm5 = ymm1[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm2 = ymm5[0,1],ymm2[2,3],ymm5[4,5],ymm2[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm3 = ymm3[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm1 = ymm1[0,1],ymm3[2,3],ymm1[4,5],ymm3[6,7]
; AVX2-FP-NEXT:    vmovaps %ymm1, 64(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm2, 96(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm0, (%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm4, 32(%rdx)
; AVX2-FP-NEXT:    vzeroupper
; AVX2-FP-NEXT:    retq
;
; AVX2-FCP-LABEL: store_i64_stride2_vf8:
; AVX2-FCP:       # %bb.0:
; AVX2-FCP-NEXT:    vmovaps (%rdi), %ymm0
; AVX2-FCP-NEXT:    vmovaps 32(%rdi), %ymm1
; AVX2-FCP-NEXT:    vmovaps (%rsi), %ymm2
; AVX2-FCP-NEXT:    vmovaps 32(%rsi), %ymm3
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm4 = ymm2[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm5 = ymm0[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm4 = ymm5[0,1],ymm4[2,3],ymm5[4,5],ymm4[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm2[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm0 = ymm0[0,1],ymm2[2,3],ymm0[4,5],ymm2[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm3[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm5 = ymm1[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm2 = ymm5[0,1],ymm2[2,3],ymm5[4,5],ymm2[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm3 = ymm3[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm1 = ymm1[0,1],ymm3[2,3],ymm1[4,5],ymm3[6,7]
; AVX2-FCP-NEXT:    vmovaps %ymm1, 64(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm2, 96(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm0, (%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm4, 32(%rdx)
; AVX2-FCP-NEXT:    vzeroupper
; AVX2-FCP-NEXT:    retq
;
; AVX512-LABEL: store_i64_stride2_vf8:
; AVX512:       # %bb.0:
; AVX512-NEXT:    vmovdqa64 (%rdi), %zmm0
; AVX512-NEXT:    vmovdqa64 (%rsi), %zmm1
; AVX512-NEXT:    vpmovsxbq {{.*#+}} zmm2 = [0,8,1,9,2,10,3,11]
; AVX512-NEXT:    vpermi2q %zmm1, %zmm0, %zmm2
; AVX512-NEXT:    vpmovsxbq {{.*#+}} zmm3 = [4,12,5,13,6,14,7,15]
; AVX512-NEXT:    vpermi2q %zmm1, %zmm0, %zmm3
; AVX512-NEXT:    vmovdqa64 %zmm3, 64(%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm2, (%rdx)
; AVX512-NEXT:    vzeroupper
; AVX512-NEXT:    retq
;
; AVX512-FCP-LABEL: store_i64_stride2_vf8:
; AVX512-FCP:       # %bb.0:
; AVX512-FCP-NEXT:    vmovdqa64 (%rdi), %zmm0
; AVX512-FCP-NEXT:    vmovdqa64 (%rsi), %zmm1
; AVX512-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm2 = [0,8,1,9,2,10,3,11]
; AVX512-FCP-NEXT:    vpermi2q %zmm1, %zmm0, %zmm2
; AVX512-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm3 = [4,12,5,13,6,14,7,15]
; AVX512-FCP-NEXT:    vpermi2q %zmm1, %zmm0, %zmm3
; AVX512-FCP-NEXT:    vmovdqa64 %zmm3, 64(%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm2, (%rdx)
; AVX512-FCP-NEXT:    vzeroupper
; AVX512-FCP-NEXT:    retq
;
; AVX512DQ-LABEL: store_i64_stride2_vf8:
; AVX512DQ:       # %bb.0:
; AVX512DQ-NEXT:    vmovdqa64 (%rdi), %zmm0
; AVX512DQ-NEXT:    vmovdqa64 (%rsi), %zmm1
; AVX512DQ-NEXT:    vpmovsxbq {{.*#+}} zmm2 = [0,8,1,9,2,10,3,11]
; AVX512DQ-NEXT:    vpermi2q %zmm1, %zmm0, %zmm2
; AVX512DQ-NEXT:    vpmovsxbq {{.*#+}} zmm3 = [4,12,5,13,6,14,7,15]
; AVX512DQ-NEXT:    vpermi2q %zmm1, %zmm0, %zmm3
; AVX512DQ-NEXT:    vmovdqa64 %zmm3, 64(%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm2, (%rdx)
; AVX512DQ-NEXT:    vzeroupper
; AVX512DQ-NEXT:    retq
;
; AVX512DQ-FCP-LABEL: store_i64_stride2_vf8:
; AVX512DQ-FCP:       # %bb.0:
; AVX512DQ-FCP-NEXT:    vmovdqa64 (%rdi), %zmm0
; AVX512DQ-FCP-NEXT:    vmovdqa64 (%rsi), %zmm1
; AVX512DQ-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm2 = [0,8,1,9,2,10,3,11]
; AVX512DQ-FCP-NEXT:    vpermi2q %zmm1, %zmm0, %zmm2
; AVX512DQ-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm3 = [4,12,5,13,6,14,7,15]
; AVX512DQ-FCP-NEXT:    vpermi2q %zmm1, %zmm0, %zmm3
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm3, 64(%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm2, (%rdx)
; AVX512DQ-FCP-NEXT:    vzeroupper
; AVX512DQ-FCP-NEXT:    retq
;
; AVX512BW-LABEL: store_i64_stride2_vf8:
; AVX512BW:       # %bb.0:
; AVX512BW-NEXT:    vmovdqa64 (%rdi), %zmm0
; AVX512BW-NEXT:    vmovdqa64 (%rsi), %zmm1
; AVX512BW-NEXT:    vpmovsxbq {{.*#+}} zmm2 = [0,8,1,9,2,10,3,11]
; AVX512BW-NEXT:    vpermi2q %zmm1, %zmm0, %zmm2
; AVX512BW-NEXT:    vpmovsxbq {{.*#+}} zmm3 = [4,12,5,13,6,14,7,15]
; AVX512BW-NEXT:    vpermi2q %zmm1, %zmm0, %zmm3
; AVX512BW-NEXT:    vmovdqa64 %zmm3, 64(%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm2, (%rdx)
; AVX512BW-NEXT:    vzeroupper
; AVX512BW-NEXT:    retq
;
; AVX512BW-FCP-LABEL: store_i64_stride2_vf8:
; AVX512BW-FCP:       # %bb.0:
; AVX512BW-FCP-NEXT:    vmovdqa64 (%rdi), %zmm0
; AVX512BW-FCP-NEXT:    vmovdqa64 (%rsi), %zmm1
; AVX512BW-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm2 = [0,8,1,9,2,10,3,11]
; AVX512BW-FCP-NEXT:    vpermi2q %zmm1, %zmm0, %zmm2
; AVX512BW-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm3 = [4,12,5,13,6,14,7,15]
; AVX512BW-FCP-NEXT:    vpermi2q %zmm1, %zmm0, %zmm3
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm3, 64(%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm2, (%rdx)
; AVX512BW-FCP-NEXT:    vzeroupper
; AVX512BW-FCP-NEXT:    retq
;
; AVX512DQ-BW-LABEL: store_i64_stride2_vf8:
; AVX512DQ-BW:       # %bb.0:
; AVX512DQ-BW-NEXT:    vmovdqa64 (%rdi), %zmm0
; AVX512DQ-BW-NEXT:    vmovdqa64 (%rsi), %zmm1
; AVX512DQ-BW-NEXT:    vpmovsxbq {{.*#+}} zmm2 = [0,8,1,9,2,10,3,11]
; AVX512DQ-BW-NEXT:    vpermi2q %zmm1, %zmm0, %zmm2
; AVX512DQ-BW-NEXT:    vpmovsxbq {{.*#+}} zmm3 = [4,12,5,13,6,14,7,15]
; AVX512DQ-BW-NEXT:    vpermi2q %zmm1, %zmm0, %zmm3
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm3, 64(%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm2, (%rdx)
; AVX512DQ-BW-NEXT:    vzeroupper
; AVX512DQ-BW-NEXT:    retq
;
; AVX512DQ-BW-FCP-LABEL: store_i64_stride2_vf8:
; AVX512DQ-BW-FCP:       # %bb.0:
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 (%rdi), %zmm0
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 (%rsi), %zmm1
; AVX512DQ-BW-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm2 = [0,8,1,9,2,10,3,11]
; AVX512DQ-BW-FCP-NEXT:    vpermi2q %zmm1, %zmm0, %zmm2
; AVX512DQ-BW-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm3 = [4,12,5,13,6,14,7,15]
; AVX512DQ-BW-FCP-NEXT:    vpermi2q %zmm1, %zmm0, %zmm3
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm3, 64(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm2, (%rdx)
; AVX512DQ-BW-FCP-NEXT:    vzeroupper
; AVX512DQ-BW-FCP-NEXT:    retq
  %in.vec0 = load <8 x i64>, ptr %in.vecptr0, align 64
  %in.vec1 = load <8 x i64>, ptr %in.vecptr1, align 64
  %1 = shufflevector <8 x i64> %in.vec0, <8 x i64> %in.vec1, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
  %interleaved.vec = shufflevector <16 x i64> %1, <16 x i64> poison, <16 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11, i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15>
  store <16 x i64> %interleaved.vec, ptr %out.vec, align 64
  ret void
}

define void @store_i64_stride2_vf16(ptr %in.vecptr0, ptr %in.vecptr1, ptr %out.vec) nounwind {
; SSE-LABEL: store_i64_stride2_vf16:
; SSE:       # %bb.0:
; SSE-NEXT:    movaps 112(%rdi), %xmm0
; SSE-NEXT:    movaps 96(%rdi), %xmm6
; SSE-NEXT:    movaps 80(%rdi), %xmm4
; SSE-NEXT:    movaps 64(%rdi), %xmm3
; SSE-NEXT:    movaps (%rdi), %xmm8
; SSE-NEXT:    movaps 16(%rdi), %xmm1
; SSE-NEXT:    movaps 32(%rdi), %xmm2
; SSE-NEXT:    movaps 48(%rdi), %xmm5
; SSE-NEXT:    movaps 96(%rsi), %xmm11
; SSE-NEXT:    movaps 80(%rsi), %xmm12
; SSE-NEXT:    movaps 64(%rsi), %xmm13
; SSE-NEXT:    movaps (%rsi), %xmm9
; SSE-NEXT:    movaps 16(%rsi), %xmm10
; SSE-NEXT:    movaps 32(%rsi), %xmm14
; SSE-NEXT:    movaps 48(%rsi), %xmm15
; SSE-NEXT:    movaps %xmm8, %xmm7
; SSE-NEXT:    unpckhpd {{.*#+}} xmm7 = xmm7[1],xmm9[1]
; SSE-NEXT:    movaps %xmm7, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movlhps {{.*#+}} xmm8 = xmm8[0],xmm9[0]
; SSE-NEXT:    movaps %xmm1, %xmm9
; SSE-NEXT:    unpckhpd {{.*#+}} xmm9 = xmm9[1],xmm10[1]
; SSE-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm10[0]
; SSE-NEXT:    movaps %xmm2, %xmm10
; SSE-NEXT:    unpckhpd {{.*#+}} xmm10 = xmm10[1],xmm14[1]
; SSE-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm14[0]
; SSE-NEXT:    movaps %xmm5, %xmm14
; SSE-NEXT:    unpckhpd {{.*#+}} xmm14 = xmm14[1],xmm15[1]
; SSE-NEXT:    movlhps {{.*#+}} xmm5 = xmm5[0],xmm15[0]
; SSE-NEXT:    movaps %xmm3, %xmm15
; SSE-NEXT:    unpckhpd {{.*#+}} xmm15 = xmm15[1],xmm13[1]
; SSE-NEXT:    movlhps {{.*#+}} xmm3 = xmm3[0],xmm13[0]
; SSE-NEXT:    movaps %xmm4, %xmm13
; SSE-NEXT:    unpckhpd {{.*#+}} xmm13 = xmm13[1],xmm12[1]
; SSE-NEXT:    movlhps {{.*#+}} xmm4 = xmm4[0],xmm12[0]
; SSE-NEXT:    movaps %xmm6, %xmm12
; SSE-NEXT:    unpckhpd {{.*#+}} xmm12 = xmm12[1],xmm11[1]
; SSE-NEXT:    movlhps {{.*#+}} xmm6 = xmm6[0],xmm11[0]
; SSE-NEXT:    movaps 112(%rsi), %xmm11
; SSE-NEXT:    movaps %xmm0, %xmm7
; SSE-NEXT:    unpckhpd {{.*#+}} xmm7 = xmm7[1],xmm11[1]
; SSE-NEXT:    movlhps {{.*#+}} xmm0 = xmm0[0],xmm11[0]
; SSE-NEXT:    movaps %xmm0, 224(%rdx)
; SSE-NEXT:    movaps %xmm7, 240(%rdx)
; SSE-NEXT:    movaps %xmm6, 192(%rdx)
; SSE-NEXT:    movaps %xmm12, 208(%rdx)
; SSE-NEXT:    movaps %xmm4, 160(%rdx)
; SSE-NEXT:    movaps %xmm13, 176(%rdx)
; SSE-NEXT:    movaps %xmm3, 128(%rdx)
; SSE-NEXT:    movaps %xmm15, 144(%rdx)
; SSE-NEXT:    movaps %xmm5, 96(%rdx)
; SSE-NEXT:    movaps %xmm14, 112(%rdx)
; SSE-NEXT:    movaps %xmm2, 64(%rdx)
; SSE-NEXT:    movaps %xmm10, 80(%rdx)
; SSE-NEXT:    movaps %xmm1, 32(%rdx)
; SSE-NEXT:    movaps %xmm9, 48(%rdx)
; SSE-NEXT:    movaps %xmm8, (%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 16(%rdx)
; SSE-NEXT:    retq
;
; AVX-LABEL: store_i64_stride2_vf16:
; AVX:       # %bb.0:
; AVX-NEXT:    vmovaps (%rsi), %xmm0
; AVX-NEXT:    vmovaps 32(%rsi), %xmm1
; AVX-NEXT:    vmovaps 64(%rsi), %xmm2
; AVX-NEXT:    vmovaps 96(%rsi), %xmm3
; AVX-NEXT:    vmovaps (%rdi), %xmm4
; AVX-NEXT:    vmovaps 32(%rdi), %xmm5
; AVX-NEXT:    vmovaps 64(%rdi), %xmm6
; AVX-NEXT:    vmovaps 96(%rdi), %xmm7
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm8 = xmm7[1],xmm3[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm3 = xmm7[0],xmm3[0]
; AVX-NEXT:    vinsertf128 $1, %xmm8, %ymm3, %ymm3
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm7 = xmm6[1],xmm2[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm2 = xmm6[0],xmm2[0]
; AVX-NEXT:    vinsertf128 $1, %xmm7, %ymm2, %ymm2
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm6 = xmm4[1],xmm0[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm4[0],xmm0[0]
; AVX-NEXT:    vinsertf128 $1, %xmm6, %ymm0, %ymm0
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm4 = xmm5[1],xmm1[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm1 = xmm5[0],xmm1[0]
; AVX-NEXT:    vinsertf128 $1, %xmm4, %ymm1, %ymm1
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm4 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm5 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm4 = ymm5[0],ymm4[0],ymm5[3],ymm4[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm5 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm6 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm5 = ymm6[0],ymm5[0],ymm6[3],ymm5[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm6 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm7 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm6 = ymm7[0],ymm6[0],ymm7[3],ymm6[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm7 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm8 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm7 = ymm8[0],ymm7[0],ymm8[3],ymm7[3]
; AVX-NEXT:    vmovapd %ymm7, 32(%rdx)
; AVX-NEXT:    vmovapd %ymm6, 96(%rdx)
; AVX-NEXT:    vmovapd %ymm5, 160(%rdx)
; AVX-NEXT:    vmovapd %ymm4, 224(%rdx)
; AVX-NEXT:    vmovaps %ymm1, 64(%rdx)
; AVX-NEXT:    vmovapd %ymm0, (%rdx)
; AVX-NEXT:    vmovaps %ymm2, 128(%rdx)
; AVX-NEXT:    vmovaps %ymm3, 192(%rdx)
; AVX-NEXT:    vzeroupper
; AVX-NEXT:    retq
;
; AVX2-LABEL: store_i64_stride2_vf16:
; AVX2:       # %bb.0:
; AVX2-NEXT:    vmovaps (%rdi), %ymm0
; AVX2-NEXT:    vmovaps 32(%rdi), %ymm1
; AVX2-NEXT:    vmovaps 64(%rdi), %ymm2
; AVX2-NEXT:    vmovaps 96(%rdi), %ymm3
; AVX2-NEXT:    vmovaps (%rsi), %ymm4
; AVX2-NEXT:    vmovaps 32(%rsi), %ymm5
; AVX2-NEXT:    vmovaps 64(%rsi), %ymm6
; AVX2-NEXT:    vmovaps 96(%rsi), %ymm7
; AVX2-NEXT:    vpermpd {{.*#+}} ymm8 = ymm4[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm9 = ymm0[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm8 = ymm9[0,1],ymm8[2,3],ymm9[4,5],ymm8[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm4 = ymm4[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm0 = ymm0[0,1],ymm4[2,3],ymm0[4,5],ymm4[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm4 = ymm5[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm9 = ymm1[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm4 = ymm9[0,1],ymm4[2,3],ymm9[4,5],ymm4[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm5 = ymm5[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm1 = ymm1[0,1],ymm5[2,3],ymm1[4,5],ymm5[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm5 = ymm6[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm9 = ymm2[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm5 = ymm9[0,1],ymm5[2,3],ymm9[4,5],ymm5[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm6 = ymm6[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm2 = ymm2[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm2 = ymm2[0,1],ymm6[2,3],ymm2[4,5],ymm6[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm6 = ymm7[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm9 = ymm3[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm6 = ymm9[0,1],ymm6[2,3],ymm9[4,5],ymm6[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm7 = ymm7[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm3 = ymm3[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm3 = ymm3[0,1],ymm7[2,3],ymm3[4,5],ymm7[6,7]
; AVX2-NEXT:    vmovaps %ymm3, 192(%rdx)
; AVX2-NEXT:    vmovaps %ymm6, 224(%rdx)
; AVX2-NEXT:    vmovaps %ymm2, 128(%rdx)
; AVX2-NEXT:    vmovaps %ymm5, 160(%rdx)
; AVX2-NEXT:    vmovaps %ymm1, 64(%rdx)
; AVX2-NEXT:    vmovaps %ymm4, 96(%rdx)
; AVX2-NEXT:    vmovaps %ymm0, (%rdx)
; AVX2-NEXT:    vmovaps %ymm8, 32(%rdx)
; AVX2-NEXT:    vzeroupper
; AVX2-NEXT:    retq
;
; AVX2-FP-LABEL: store_i64_stride2_vf16:
; AVX2-FP:       # %bb.0:
; AVX2-FP-NEXT:    vmovaps (%rdi), %ymm0
; AVX2-FP-NEXT:    vmovaps 32(%rdi), %ymm1
; AVX2-FP-NEXT:    vmovaps 64(%rdi), %ymm2
; AVX2-FP-NEXT:    vmovaps 96(%rdi), %ymm3
; AVX2-FP-NEXT:    vmovaps (%rsi), %ymm4
; AVX2-FP-NEXT:    vmovaps 32(%rsi), %ymm5
; AVX2-FP-NEXT:    vmovaps 64(%rsi), %ymm6
; AVX2-FP-NEXT:    vmovaps 96(%rsi), %ymm7
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm8 = ymm4[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm9 = ymm0[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm8 = ymm9[0,1],ymm8[2,3],ymm9[4,5],ymm8[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm4 = ymm4[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm0 = ymm0[0,1],ymm4[2,3],ymm0[4,5],ymm4[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm4 = ymm5[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm9 = ymm1[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm4 = ymm9[0,1],ymm4[2,3],ymm9[4,5],ymm4[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm5 = ymm5[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm1 = ymm1[0,1],ymm5[2,3],ymm1[4,5],ymm5[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm5 = ymm6[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm9 = ymm2[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm5 = ymm9[0,1],ymm5[2,3],ymm9[4,5],ymm5[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm6 = ymm6[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm2[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm2 = ymm2[0,1],ymm6[2,3],ymm2[4,5],ymm6[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm6 = ymm7[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm9 = ymm3[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm6 = ymm9[0,1],ymm6[2,3],ymm9[4,5],ymm6[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm7 = ymm7[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm3 = ymm3[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm3 = ymm3[0,1],ymm7[2,3],ymm3[4,5],ymm7[6,7]
; AVX2-FP-NEXT:    vmovaps %ymm3, 192(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm6, 224(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm2, 128(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm5, 160(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm1, 64(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm4, 96(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm0, (%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm8, 32(%rdx)
; AVX2-FP-NEXT:    vzeroupper
; AVX2-FP-NEXT:    retq
;
; AVX2-FCP-LABEL: store_i64_stride2_vf16:
; AVX2-FCP:       # %bb.0:
; AVX2-FCP-NEXT:    vmovaps (%rdi), %ymm0
; AVX2-FCP-NEXT:    vmovaps 32(%rdi), %ymm1
; AVX2-FCP-NEXT:    vmovaps 64(%rdi), %ymm2
; AVX2-FCP-NEXT:    vmovaps 96(%rdi), %ymm3
; AVX2-FCP-NEXT:    vmovaps (%rsi), %ymm4
; AVX2-FCP-NEXT:    vmovaps 32(%rsi), %ymm5
; AVX2-FCP-NEXT:    vmovaps 64(%rsi), %ymm6
; AVX2-FCP-NEXT:    vmovaps 96(%rsi), %ymm7
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm8 = ymm4[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm9 = ymm0[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm8 = ymm9[0,1],ymm8[2,3],ymm9[4,5],ymm8[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm4 = ymm4[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm0 = ymm0[0,1],ymm4[2,3],ymm0[4,5],ymm4[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm4 = ymm5[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm9 = ymm1[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm4 = ymm9[0,1],ymm4[2,3],ymm9[4,5],ymm4[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm5 = ymm5[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm1 = ymm1[0,1],ymm5[2,3],ymm1[4,5],ymm5[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm5 = ymm6[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm9 = ymm2[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm5 = ymm9[0,1],ymm5[2,3],ymm9[4,5],ymm5[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm6 = ymm6[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm2[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm2 = ymm2[0,1],ymm6[2,3],ymm2[4,5],ymm6[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm6 = ymm7[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm9 = ymm3[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm6 = ymm9[0,1],ymm6[2,3],ymm9[4,5],ymm6[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm7 = ymm7[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm3 = ymm3[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm3 = ymm3[0,1],ymm7[2,3],ymm3[4,5],ymm7[6,7]
; AVX2-FCP-NEXT:    vmovaps %ymm3, 192(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm6, 224(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm2, 128(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm5, 160(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm1, 64(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm4, 96(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm0, (%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm8, 32(%rdx)
; AVX2-FCP-NEXT:    vzeroupper
; AVX2-FCP-NEXT:    retq
;
; AVX512-LABEL: store_i64_stride2_vf16:
; AVX512:       # %bb.0:
; AVX512-NEXT:    vmovdqa64 (%rdi), %zmm0
; AVX512-NEXT:    vmovdqa64 64(%rdi), %zmm1
; AVX512-NEXT:    vmovdqa64 (%rsi), %zmm2
; AVX512-NEXT:    vmovdqa64 64(%rsi), %zmm3
; AVX512-NEXT:    vpmovsxbq {{.*#+}} zmm4 = [4,12,5,13,6,14,7,15]
; AVX512-NEXT:    vmovdqa64 %zmm0, %zmm5
; AVX512-NEXT:    vpermt2q %zmm2, %zmm4, %zmm5
; AVX512-NEXT:    vpmovsxbq {{.*#+}} zmm6 = [0,8,1,9,2,10,3,11]
; AVX512-NEXT:    vpermt2q %zmm2, %zmm6, %zmm0
; AVX512-NEXT:    vpermi2q %zmm3, %zmm1, %zmm4
; AVX512-NEXT:    vpermt2q %zmm3, %zmm6, %zmm1
; AVX512-NEXT:    vmovdqa64 %zmm1, 128(%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm4, 192(%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm0, (%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm5, 64(%rdx)
; AVX512-NEXT:    vzeroupper
; AVX512-NEXT:    retq
;
; AVX512-FCP-LABEL: store_i64_stride2_vf16:
; AVX512-FCP:       # %bb.0:
; AVX512-FCP-NEXT:    vmovdqa64 (%rdi), %zmm0
; AVX512-FCP-NEXT:    vmovdqa64 64(%rdi), %zmm1
; AVX512-FCP-NEXT:    vmovdqa64 (%rsi), %zmm2
; AVX512-FCP-NEXT:    vmovdqa64 64(%rsi), %zmm3
; AVX512-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm4 = [4,12,5,13,6,14,7,15]
; AVX512-FCP-NEXT:    vmovdqa64 %zmm0, %zmm5
; AVX512-FCP-NEXT:    vpermt2q %zmm2, %zmm4, %zmm5
; AVX512-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm6 = [0,8,1,9,2,10,3,11]
; AVX512-FCP-NEXT:    vpermt2q %zmm2, %zmm6, %zmm0
; AVX512-FCP-NEXT:    vpermi2q %zmm3, %zmm1, %zmm4
; AVX512-FCP-NEXT:    vpermt2q %zmm3, %zmm6, %zmm1
; AVX512-FCP-NEXT:    vmovdqa64 %zmm1, 128(%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm4, 192(%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm0, (%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm5, 64(%rdx)
; AVX512-FCP-NEXT:    vzeroupper
; AVX512-FCP-NEXT:    retq
;
; AVX512DQ-LABEL: store_i64_stride2_vf16:
; AVX512DQ:       # %bb.0:
; AVX512DQ-NEXT:    vmovdqa64 (%rdi), %zmm0
; AVX512DQ-NEXT:    vmovdqa64 64(%rdi), %zmm1
; AVX512DQ-NEXT:    vmovdqa64 (%rsi), %zmm2
; AVX512DQ-NEXT:    vmovdqa64 64(%rsi), %zmm3
; AVX512DQ-NEXT:    vpmovsxbq {{.*#+}} zmm4 = [4,12,5,13,6,14,7,15]
; AVX512DQ-NEXT:    vmovdqa64 %zmm0, %zmm5
; AVX512DQ-NEXT:    vpermt2q %zmm2, %zmm4, %zmm5
; AVX512DQ-NEXT:    vpmovsxbq {{.*#+}} zmm6 = [0,8,1,9,2,10,3,11]
; AVX512DQ-NEXT:    vpermt2q %zmm2, %zmm6, %zmm0
; AVX512DQ-NEXT:    vpermi2q %zmm3, %zmm1, %zmm4
; AVX512DQ-NEXT:    vpermt2q %zmm3, %zmm6, %zmm1
; AVX512DQ-NEXT:    vmovdqa64 %zmm1, 128(%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm4, 192(%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm0, (%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm5, 64(%rdx)
; AVX512DQ-NEXT:    vzeroupper
; AVX512DQ-NEXT:    retq
;
; AVX512DQ-FCP-LABEL: store_i64_stride2_vf16:
; AVX512DQ-FCP:       # %bb.0:
; AVX512DQ-FCP-NEXT:    vmovdqa64 (%rdi), %zmm0
; AVX512DQ-FCP-NEXT:    vmovdqa64 64(%rdi), %zmm1
; AVX512DQ-FCP-NEXT:    vmovdqa64 (%rsi), %zmm2
; AVX512DQ-FCP-NEXT:    vmovdqa64 64(%rsi), %zmm3
; AVX512DQ-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm4 = [4,12,5,13,6,14,7,15]
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm0, %zmm5
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm2, %zmm4, %zmm5
; AVX512DQ-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm6 = [0,8,1,9,2,10,3,11]
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm2, %zmm6, %zmm0
; AVX512DQ-FCP-NEXT:    vpermi2q %zmm3, %zmm1, %zmm4
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm3, %zmm6, %zmm1
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm1, 128(%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm4, 192(%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm0, (%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm5, 64(%rdx)
; AVX512DQ-FCP-NEXT:    vzeroupper
; AVX512DQ-FCP-NEXT:    retq
;
; AVX512BW-LABEL: store_i64_stride2_vf16:
; AVX512BW:       # %bb.0:
; AVX512BW-NEXT:    vmovdqa64 (%rdi), %zmm0
; AVX512BW-NEXT:    vmovdqa64 64(%rdi), %zmm1
; AVX512BW-NEXT:    vmovdqa64 (%rsi), %zmm2
; AVX512BW-NEXT:    vmovdqa64 64(%rsi), %zmm3
; AVX512BW-NEXT:    vpmovsxbq {{.*#+}} zmm4 = [4,12,5,13,6,14,7,15]
; AVX512BW-NEXT:    vmovdqa64 %zmm0, %zmm5
; AVX512BW-NEXT:    vpermt2q %zmm2, %zmm4, %zmm5
; AVX512BW-NEXT:    vpmovsxbq {{.*#+}} zmm6 = [0,8,1,9,2,10,3,11]
; AVX512BW-NEXT:    vpermt2q %zmm2, %zmm6, %zmm0
; AVX512BW-NEXT:    vpermi2q %zmm3, %zmm1, %zmm4
; AVX512BW-NEXT:    vpermt2q %zmm3, %zmm6, %zmm1
; AVX512BW-NEXT:    vmovdqa64 %zmm1, 128(%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm4, 192(%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm0, (%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm5, 64(%rdx)
; AVX512BW-NEXT:    vzeroupper
; AVX512BW-NEXT:    retq
;
; AVX512BW-FCP-LABEL: store_i64_stride2_vf16:
; AVX512BW-FCP:       # %bb.0:
; AVX512BW-FCP-NEXT:    vmovdqa64 (%rdi), %zmm0
; AVX512BW-FCP-NEXT:    vmovdqa64 64(%rdi), %zmm1
; AVX512BW-FCP-NEXT:    vmovdqa64 (%rsi), %zmm2
; AVX512BW-FCP-NEXT:    vmovdqa64 64(%rsi), %zmm3
; AVX512BW-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm4 = [4,12,5,13,6,14,7,15]
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm0, %zmm5
; AVX512BW-FCP-NEXT:    vpermt2q %zmm2, %zmm4, %zmm5
; AVX512BW-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm6 = [0,8,1,9,2,10,3,11]
; AVX512BW-FCP-NEXT:    vpermt2q %zmm2, %zmm6, %zmm0
; AVX512BW-FCP-NEXT:    vpermi2q %zmm3, %zmm1, %zmm4
; AVX512BW-FCP-NEXT:    vpermt2q %zmm3, %zmm6, %zmm1
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm1, 128(%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm4, 192(%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm0, (%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm5, 64(%rdx)
; AVX512BW-FCP-NEXT:    vzeroupper
; AVX512BW-FCP-NEXT:    retq
;
; AVX512DQ-BW-LABEL: store_i64_stride2_vf16:
; AVX512DQ-BW:       # %bb.0:
; AVX512DQ-BW-NEXT:    vmovdqa64 (%rdi), %zmm0
; AVX512DQ-BW-NEXT:    vmovdqa64 64(%rdi), %zmm1
; AVX512DQ-BW-NEXT:    vmovdqa64 (%rsi), %zmm2
; AVX512DQ-BW-NEXT:    vmovdqa64 64(%rsi), %zmm3
; AVX512DQ-BW-NEXT:    vpmovsxbq {{.*#+}} zmm4 = [4,12,5,13,6,14,7,15]
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm0, %zmm5
; AVX512DQ-BW-NEXT:    vpermt2q %zmm2, %zmm4, %zmm5
; AVX512DQ-BW-NEXT:    vpmovsxbq {{.*#+}} zmm6 = [0,8,1,9,2,10,3,11]
; AVX512DQ-BW-NEXT:    vpermt2q %zmm2, %zmm6, %zmm0
; AVX512DQ-BW-NEXT:    vpermi2q %zmm3, %zmm1, %zmm4
; AVX512DQ-BW-NEXT:    vpermt2q %zmm3, %zmm6, %zmm1
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm1, 128(%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm4, 192(%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm0, (%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm5, 64(%rdx)
; AVX512DQ-BW-NEXT:    vzeroupper
; AVX512DQ-BW-NEXT:    retq
;
; AVX512DQ-BW-FCP-LABEL: store_i64_stride2_vf16:
; AVX512DQ-BW-FCP:       # %bb.0:
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 (%rdi), %zmm0
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 64(%rdi), %zmm1
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 (%rsi), %zmm2
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 64(%rsi), %zmm3
; AVX512DQ-BW-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm4 = [4,12,5,13,6,14,7,15]
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm0, %zmm5
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm2, %zmm4, %zmm5
; AVX512DQ-BW-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm6 = [0,8,1,9,2,10,3,11]
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm2, %zmm6, %zmm0
; AVX512DQ-BW-FCP-NEXT:    vpermi2q %zmm3, %zmm1, %zmm4
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm3, %zmm6, %zmm1
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm1, 128(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm4, 192(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm0, (%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm5, 64(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vzeroupper
; AVX512DQ-BW-FCP-NEXT:    retq
  %in.vec0 = load <16 x i64>, ptr %in.vecptr0, align 64
  %in.vec1 = load <16 x i64>, ptr %in.vecptr1, align 64
  %1 = shufflevector <16 x i64> %in.vec0, <16 x i64> %in.vec1, <32 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31>
  %interleaved.vec = shufflevector <32 x i64> %1, <32 x i64> poison, <32 x i32> <i32 0, i32 16, i32 1, i32 17, i32 2, i32 18, i32 3, i32 19, i32 4, i32 20, i32 5, i32 21, i32 6, i32 22, i32 7, i32 23, i32 8, i32 24, i32 9, i32 25, i32 10, i32 26, i32 11, i32 27, i32 12, i32 28, i32 13, i32 29, i32 14, i32 30, i32 15, i32 31>
  store <32 x i64> %interleaved.vec, ptr %out.vec, align 64
  ret void
}

define void @store_i64_stride2_vf32(ptr %in.vecptr0, ptr %in.vecptr1, ptr %out.vec) nounwind {
; SSE-LABEL: store_i64_stride2_vf32:
; SSE:       # %bb.0:
; SSE-NEXT:    subq $152, %rsp
; SSE-NEXT:    movaps 112(%rdi), %xmm14
; SSE-NEXT:    movaps 96(%rdi), %xmm13
; SSE-NEXT:    movaps 80(%rdi), %xmm11
; SSE-NEXT:    movaps 64(%rdi), %xmm10
; SSE-NEXT:    movaps (%rdi), %xmm7
; SSE-NEXT:    movaps 16(%rdi), %xmm8
; SSE-NEXT:    movaps 32(%rdi), %xmm9
; SSE-NEXT:    movaps 48(%rdi), %xmm12
; SSE-NEXT:    movaps 96(%rsi), %xmm0
; SSE-NEXT:    movaps 80(%rsi), %xmm1
; SSE-NEXT:    movaps 64(%rsi), %xmm2
; SSE-NEXT:    movaps (%rsi), %xmm3
; SSE-NEXT:    movaps 16(%rsi), %xmm4
; SSE-NEXT:    movaps 32(%rsi), %xmm5
; SSE-NEXT:    movaps 48(%rsi), %xmm6
; SSE-NEXT:    movaps %xmm7, %xmm15
; SSE-NEXT:    movlhps {{.*#+}} xmm15 = xmm15[0],xmm3[0]
; SSE-NEXT:    movaps %xmm15, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm7 = xmm7[1],xmm3[1]
; SSE-NEXT:    movaps %xmm7, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps %xmm8, %xmm7
; SSE-NEXT:    movlhps {{.*#+}} xmm7 = xmm7[0],xmm4[0]
; SSE-NEXT:    movaps %xmm7, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm8 = xmm8[1],xmm4[1]
; SSE-NEXT:    movaps %xmm8, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps %xmm9, %xmm4
; SSE-NEXT:    movlhps {{.*#+}} xmm4 = xmm4[0],xmm5[0]
; SSE-NEXT:    movaps %xmm4, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm9 = xmm9[1],xmm5[1]
; SSE-NEXT:    movaps %xmm9, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps %xmm12, %xmm4
; SSE-NEXT:    movlhps {{.*#+}} xmm4 = xmm4[0],xmm6[0]
; SSE-NEXT:    movaps %xmm4, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm12 = xmm12[1],xmm6[1]
; SSE-NEXT:    movaps %xmm12, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps %xmm10, %xmm3
; SSE-NEXT:    movlhps {{.*#+}} xmm3 = xmm3[0],xmm2[0]
; SSE-NEXT:    movaps %xmm3, (%rsp) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm10 = xmm10[1],xmm2[1]
; SSE-NEXT:    movaps %xmm10, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps %xmm11, %xmm2
; SSE-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm1[0]
; SSE-NEXT:    movaps %xmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm11 = xmm11[1],xmm1[1]
; SSE-NEXT:    movaps %xmm11, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps %xmm13, %xmm1
; SSE-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm13 = xmm13[1],xmm0[1]
; SSE-NEXT:    movaps %xmm13, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps 112(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm14, %xmm1
; SSE-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm14 = xmm14[1],xmm0[1]
; SSE-NEXT:    movaps %xmm14, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps 128(%rdi), %xmm15
; SSE-NEXT:    movaps 128(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm15, %xmm1
; SSE-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm15 = xmm15[1],xmm0[1]
; SSE-NEXT:    movaps 144(%rdi), %xmm13
; SSE-NEXT:    movaps 144(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm13, %xmm14
; SSE-NEXT:    movlhps {{.*#+}} xmm14 = xmm14[0],xmm0[0]
; SSE-NEXT:    unpckhpd {{.*#+}} xmm13 = xmm13[1],xmm0[1]
; SSE-NEXT:    movaps 160(%rdi), %xmm10
; SSE-NEXT:    movaps 160(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm10, %xmm12
; SSE-NEXT:    movlhps {{.*#+}} xmm12 = xmm12[0],xmm0[0]
; SSE-NEXT:    unpckhpd {{.*#+}} xmm10 = xmm10[1],xmm0[1]
; SSE-NEXT:    movaps 176(%rdi), %xmm8
; SSE-NEXT:    movaps 176(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm8, %xmm11
; SSE-NEXT:    movlhps {{.*#+}} xmm11 = xmm11[0],xmm0[0]
; SSE-NEXT:    unpckhpd {{.*#+}} xmm8 = xmm8[1],xmm0[1]
; SSE-NEXT:    movaps 192(%rdi), %xmm6
; SSE-NEXT:    movaps 192(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm6, %xmm9
; SSE-NEXT:    movlhps {{.*#+}} xmm9 = xmm9[0],xmm0[0]
; SSE-NEXT:    unpckhpd {{.*#+}} xmm6 = xmm6[1],xmm0[1]
; SSE-NEXT:    movaps 208(%rdi), %xmm5
; SSE-NEXT:    movaps 208(%rsi), %xmm1
; SSE-NEXT:    movaps %xmm5, %xmm7
; SSE-NEXT:    movlhps {{.*#+}} xmm7 = xmm7[0],xmm1[0]
; SSE-NEXT:    unpckhpd {{.*#+}} xmm5 = xmm5[1],xmm1[1]
; SSE-NEXT:    movaps 224(%rdi), %xmm1
; SSE-NEXT:    movaps 224(%rsi), %xmm3
; SSE-NEXT:    movaps %xmm1, %xmm2
; SSE-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm3[0]
; SSE-NEXT:    unpckhpd {{.*#+}} xmm1 = xmm1[1],xmm3[1]
; SSE-NEXT:    movaps 240(%rdi), %xmm3
; SSE-NEXT:    movaps 240(%rsi), %xmm4
; SSE-NEXT:    movaps %xmm3, %xmm0
; SSE-NEXT:    movlhps {{.*#+}} xmm0 = xmm0[0],xmm4[0]
; SSE-NEXT:    unpckhpd {{.*#+}} xmm3 = xmm3[1],xmm4[1]
; SSE-NEXT:    movaps %xmm3, 496(%rdx)
; SSE-NEXT:    movaps %xmm0, 480(%rdx)
; SSE-NEXT:    movaps %xmm1, 464(%rdx)
; SSE-NEXT:    movaps %xmm2, 448(%rdx)
; SSE-NEXT:    movaps %xmm5, 432(%rdx)
; SSE-NEXT:    movaps %xmm7, 416(%rdx)
; SSE-NEXT:    movaps %xmm6, 400(%rdx)
; SSE-NEXT:    movaps %xmm9, 384(%rdx)
; SSE-NEXT:    movaps %xmm8, 368(%rdx)
; SSE-NEXT:    movaps %xmm11, 352(%rdx)
; SSE-NEXT:    movaps %xmm10, 336(%rdx)
; SSE-NEXT:    movaps %xmm12, 320(%rdx)
; SSE-NEXT:    movaps %xmm13, 304(%rdx)
; SSE-NEXT:    movaps %xmm14, 288(%rdx)
; SSE-NEXT:    movaps %xmm15, 272(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 256(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 240(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 224(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 208(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 192(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 176(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 160(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 144(%rdx)
; SSE-NEXT:    movaps (%rsp), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 128(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 112(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 96(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 80(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 64(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 48(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 32(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 16(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, (%rdx)
; SSE-NEXT:    addq $152, %rsp
; SSE-NEXT:    retq
;
; AVX-LABEL: store_i64_stride2_vf32:
; AVX:       # %bb.0:
; AVX-NEXT:    vmovaps 224(%rsi), %xmm0
; AVX-NEXT:    vmovaps 224(%rdi), %xmm1
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm2 = xmm1[1],xmm0[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm1[0],xmm0[0]
; AVX-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
; AVX-NEXT:    vmovupd %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX-NEXT:    vmovaps 128(%rsi), %xmm1
; AVX-NEXT:    vmovaps 128(%rdi), %xmm2
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm3 = xmm2[1],xmm1[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm1 = xmm2[0],xmm1[0]
; AVX-NEXT:    vinsertf128 $1, %xmm3, %ymm1, %ymm1
; AVX-NEXT:    vmovaps (%rsi), %xmm2
; AVX-NEXT:    vmovaps 32(%rsi), %xmm3
; AVX-NEXT:    vmovaps 64(%rsi), %xmm4
; AVX-NEXT:    vmovaps 96(%rsi), %xmm5
; AVX-NEXT:    vmovaps (%rdi), %xmm6
; AVX-NEXT:    vmovaps 32(%rdi), %xmm7
; AVX-NEXT:    vmovaps 64(%rdi), %xmm8
; AVX-NEXT:    vmovaps 96(%rdi), %xmm9
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm10 = xmm6[1],xmm2[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm2 = xmm6[0],xmm2[0]
; AVX-NEXT:    vinsertf128 $1, %xmm10, %ymm2, %ymm2
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm6 = xmm7[1],xmm3[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm3 = xmm7[0],xmm3[0]
; AVX-NEXT:    vinsertf128 $1, %xmm6, %ymm3, %ymm3
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm6 = xmm8[1],xmm4[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm4 = xmm8[0],xmm4[0]
; AVX-NEXT:    vinsertf128 $1, %xmm6, %ymm4, %ymm4
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm6 = xmm9[1],xmm5[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm5 = xmm9[0],xmm5[0]
; AVX-NEXT:    vinsertf128 $1, %xmm6, %ymm5, %ymm5
; AVX-NEXT:    vmovaps 160(%rsi), %xmm6
; AVX-NEXT:    vmovaps 160(%rdi), %xmm7
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm8 = xmm7[1],xmm6[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm6 = xmm7[0],xmm6[0]
; AVX-NEXT:    vinsertf128 $1, %xmm8, %ymm6, %ymm6
; AVX-NEXT:    vmovaps 192(%rsi), %xmm7
; AVX-NEXT:    vmovaps 192(%rdi), %xmm8
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm9 = xmm8[1],xmm7[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm7 = xmm8[0],xmm7[0]
; AVX-NEXT:    vinsertf128 $1, %xmm9, %ymm7, %ymm7
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm8 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm9 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm8 = ymm9[0],ymm8[0],ymm9[3],ymm8[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm9 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm10 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm9 = ymm10[0],ymm9[0],ymm10[3],ymm9[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm10 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm11 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm10 = ymm11[0],ymm10[0],ymm11[3],ymm10[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm11 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm12 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm11 = ymm12[0],ymm11[0],ymm12[3],ymm11[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm12 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm13 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm12 = ymm13[0],ymm12[0],ymm13[3],ymm12[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm13 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm14 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm13 = ymm14[0],ymm13[0],ymm14[3],ymm13[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm14 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm15 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm14 = ymm15[0],ymm14[0],ymm15[3],ymm14[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm15 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm0 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm0 = ymm0[0],ymm15[0],ymm0[3],ymm15[3]
; AVX-NEXT:    vmovapd %ymm0, 416(%rdx)
; AVX-NEXT:    vmovapd %ymm14, 352(%rdx)
; AVX-NEXT:    vmovapd %ymm13, 224(%rdx)
; AVX-NEXT:    vmovapd %ymm12, 32(%rdx)
; AVX-NEXT:    vmovapd %ymm11, 96(%rdx)
; AVX-NEXT:    vmovapd %ymm10, 160(%rdx)
; AVX-NEXT:    vmovapd %ymm9, 288(%rdx)
; AVX-NEXT:    vmovapd %ymm8, 480(%rdx)
; AVX-NEXT:    vmovaps %ymm7, 384(%rdx)
; AVX-NEXT:    vmovaps %ymm6, 320(%rdx)
; AVX-NEXT:    vmovaps %ymm5, 192(%rdx)
; AVX-NEXT:    vmovaps %ymm4, 128(%rdx)
; AVX-NEXT:    vmovaps %ymm3, 64(%rdx)
; AVX-NEXT:    vmovaps %ymm2, (%rdx)
; AVX-NEXT:    vmovaps %ymm1, 256(%rdx)
; AVX-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX-NEXT:    vmovaps %ymm0, 448(%rdx)
; AVX-NEXT:    vzeroupper
; AVX-NEXT:    retq
;
; AVX2-LABEL: store_i64_stride2_vf32:
; AVX2:       # %bb.0:
; AVX2-NEXT:    vmovaps 192(%rdi), %ymm2
; AVX2-NEXT:    vmovaps 160(%rdi), %ymm5
; AVX2-NEXT:    vmovaps 128(%rdi), %ymm9
; AVX2-NEXT:    vmovaps (%rdi), %ymm1
; AVX2-NEXT:    vmovaps 32(%rdi), %ymm4
; AVX2-NEXT:    vmovaps 64(%rdi), %ymm8
; AVX2-NEXT:    vmovaps 96(%rdi), %ymm12
; AVX2-NEXT:    vmovaps 192(%rsi), %ymm6
; AVX2-NEXT:    vmovaps 160(%rsi), %ymm10
; AVX2-NEXT:    vmovaps 128(%rsi), %ymm13
; AVX2-NEXT:    vmovaps (%rsi), %ymm3
; AVX2-NEXT:    vmovaps 32(%rsi), %ymm7
; AVX2-NEXT:    vmovaps 64(%rsi), %ymm11
; AVX2-NEXT:    vmovaps 96(%rsi), %ymm14
; AVX2-NEXT:    vpermpd {{.*#+}} ymm0 = ymm3[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm15 = ymm1[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm0 = ymm15[0,1],ymm0[2,3],ymm15[4,5],ymm0[6,7]
; AVX2-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-NEXT:    vpermpd {{.*#+}} ymm3 = ymm3[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm0 = ymm1[0,1],ymm3[2,3],ymm1[4,5],ymm3[6,7]
; AVX2-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-NEXT:    vpermpd {{.*#+}} ymm3 = ymm7[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm15 = ymm4[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm3 = ymm15[0,1],ymm3[2,3],ymm15[4,5],ymm3[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm7 = ymm7[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm4 = ymm4[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm4 = ymm4[0,1],ymm7[2,3],ymm4[4,5],ymm7[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm7 = ymm11[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm15 = ymm8[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm7 = ymm15[0,1],ymm7[2,3],ymm15[4,5],ymm7[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm11 = ymm11[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm8 = ymm8[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm8 = ymm8[0,1],ymm11[2,3],ymm8[4,5],ymm11[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm11 = ymm14[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm15 = ymm12[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm11 = ymm15[0,1],ymm11[2,3],ymm15[4,5],ymm11[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm14 = ymm14[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm12 = ymm12[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm12 = ymm12[0,1],ymm14[2,3],ymm12[4,5],ymm14[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm14 = ymm13[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm15 = ymm9[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm14 = ymm15[0,1],ymm14[2,3],ymm15[4,5],ymm14[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm13 = ymm13[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm9 = ymm9[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm9 = ymm9[0,1],ymm13[2,3],ymm9[4,5],ymm13[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm13 = ymm10[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm15 = ymm5[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm13 = ymm15[0,1],ymm13[2,3],ymm15[4,5],ymm13[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm10 = ymm10[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm5 = ymm5[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm5 = ymm5[0,1],ymm10[2,3],ymm5[4,5],ymm10[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm10 = ymm6[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm15 = ymm2[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm10 = ymm15[0,1],ymm10[2,3],ymm15[4,5],ymm10[6,7]
; AVX2-NEXT:    vmovaps 224(%rdi), %ymm15
; AVX2-NEXT:    vpermpd {{.*#+}} ymm6 = ymm6[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm2 = ymm2[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm2 = ymm2[0,1],ymm6[2,3],ymm2[4,5],ymm6[6,7]
; AVX2-NEXT:    vmovaps 224(%rsi), %ymm6
; AVX2-NEXT:    vpermpd {{.*#+}} ymm0 = ymm6[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm1 = ymm15[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm0 = ymm1[0,1],ymm0[2,3],ymm1[4,5],ymm0[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm1 = ymm6[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm6 = ymm15[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm1 = ymm6[0,1],ymm1[2,3],ymm6[4,5],ymm1[6,7]
; AVX2-NEXT:    vmovaps %ymm1, 448(%rdx)
; AVX2-NEXT:    vmovaps %ymm0, 480(%rdx)
; AVX2-NEXT:    vmovaps %ymm2, 384(%rdx)
; AVX2-NEXT:    vmovaps %ymm10, 416(%rdx)
; AVX2-NEXT:    vmovaps %ymm5, 320(%rdx)
; AVX2-NEXT:    vmovaps %ymm13, 352(%rdx)
; AVX2-NEXT:    vmovaps %ymm9, 256(%rdx)
; AVX2-NEXT:    vmovaps %ymm14, 288(%rdx)
; AVX2-NEXT:    vmovaps %ymm12, 192(%rdx)
; AVX2-NEXT:    vmovaps %ymm11, 224(%rdx)
; AVX2-NEXT:    vmovaps %ymm8, 128(%rdx)
; AVX2-NEXT:    vmovaps %ymm7, 160(%rdx)
; AVX2-NEXT:    vmovaps %ymm4, 64(%rdx)
; AVX2-NEXT:    vmovaps %ymm3, 96(%rdx)
; AVX2-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT:    vmovaps %ymm0, (%rdx)
; AVX2-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT:    vmovaps %ymm0, 32(%rdx)
; AVX2-NEXT:    vzeroupper
; AVX2-NEXT:    retq
;
; AVX2-FP-LABEL: store_i64_stride2_vf32:
; AVX2-FP:       # %bb.0:
; AVX2-FP-NEXT:    vmovaps 192(%rdi), %ymm2
; AVX2-FP-NEXT:    vmovaps 160(%rdi), %ymm5
; AVX2-FP-NEXT:    vmovaps 128(%rdi), %ymm9
; AVX2-FP-NEXT:    vmovaps (%rdi), %ymm1
; AVX2-FP-NEXT:    vmovaps 32(%rdi), %ymm4
; AVX2-FP-NEXT:    vmovaps 64(%rdi), %ymm8
; AVX2-FP-NEXT:    vmovaps 96(%rdi), %ymm12
; AVX2-FP-NEXT:    vmovaps 192(%rsi), %ymm6
; AVX2-FP-NEXT:    vmovaps 160(%rsi), %ymm10
; AVX2-FP-NEXT:    vmovaps 128(%rsi), %ymm13
; AVX2-FP-NEXT:    vmovaps (%rsi), %ymm3
; AVX2-FP-NEXT:    vmovaps 32(%rsi), %ymm7
; AVX2-FP-NEXT:    vmovaps 64(%rsi), %ymm11
; AVX2-FP-NEXT:    vmovaps 96(%rsi), %ymm14
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm3[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm15 = ymm1[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm0 = ymm15[0,1],ymm0[2,3],ymm15[4,5],ymm0[6,7]
; AVX2-FP-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm3 = ymm3[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm0 = ymm1[0,1],ymm3[2,3],ymm1[4,5],ymm3[6,7]
; AVX2-FP-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm3 = ymm7[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm15 = ymm4[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm3 = ymm15[0,1],ymm3[2,3],ymm15[4,5],ymm3[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm7 = ymm7[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm4 = ymm4[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm4 = ymm4[0,1],ymm7[2,3],ymm4[4,5],ymm7[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm7 = ymm11[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm15 = ymm8[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm7 = ymm15[0,1],ymm7[2,3],ymm15[4,5],ymm7[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm11 = ymm11[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm8 = ymm8[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm8 = ymm8[0,1],ymm11[2,3],ymm8[4,5],ymm11[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm11 = ymm14[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm15 = ymm12[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm11 = ymm15[0,1],ymm11[2,3],ymm15[4,5],ymm11[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm14 = ymm14[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm12 = ymm12[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm12 = ymm12[0,1],ymm14[2,3],ymm12[4,5],ymm14[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm14 = ymm13[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm15 = ymm9[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm14 = ymm15[0,1],ymm14[2,3],ymm15[4,5],ymm14[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm13 = ymm13[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm9 = ymm9[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm9 = ymm9[0,1],ymm13[2,3],ymm9[4,5],ymm13[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm13 = ymm10[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm15 = ymm5[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm13 = ymm15[0,1],ymm13[2,3],ymm15[4,5],ymm13[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm10 = ymm10[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm5 = ymm5[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm5 = ymm5[0,1],ymm10[2,3],ymm5[4,5],ymm10[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm10 = ymm6[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm15 = ymm2[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm10 = ymm15[0,1],ymm10[2,3],ymm15[4,5],ymm10[6,7]
; AVX2-FP-NEXT:    vmovaps 224(%rdi), %ymm15
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm6 = ymm6[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm2[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm2 = ymm2[0,1],ymm6[2,3],ymm2[4,5],ymm6[6,7]
; AVX2-FP-NEXT:    vmovaps 224(%rsi), %ymm6
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm6[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm15[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm0 = ymm1[0,1],ymm0[2,3],ymm1[4,5],ymm0[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm6[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm6 = ymm15[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm1 = ymm6[0,1],ymm1[2,3],ymm6[4,5],ymm1[6,7]
; AVX2-FP-NEXT:    vmovaps %ymm1, 448(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm0, 480(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm2, 384(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm10, 416(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm5, 320(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm13, 352(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm9, 256(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm14, 288(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm12, 192(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm11, 224(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm8, 128(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm7, 160(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm4, 64(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm3, 96(%rdx)
; AVX2-FP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FP-NEXT:    vmovaps %ymm0, (%rdx)
; AVX2-FP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FP-NEXT:    vmovaps %ymm0, 32(%rdx)
; AVX2-FP-NEXT:    vzeroupper
; AVX2-FP-NEXT:    retq
;
; AVX2-FCP-LABEL: store_i64_stride2_vf32:
; AVX2-FCP:       # %bb.0:
; AVX2-FCP-NEXT:    vmovaps 192(%rdi), %ymm2
; AVX2-FCP-NEXT:    vmovaps 160(%rdi), %ymm5
; AVX2-FCP-NEXT:    vmovaps 128(%rdi), %ymm9
; AVX2-FCP-NEXT:    vmovaps (%rdi), %ymm1
; AVX2-FCP-NEXT:    vmovaps 32(%rdi), %ymm4
; AVX2-FCP-NEXT:    vmovaps 64(%rdi), %ymm8
; AVX2-FCP-NEXT:    vmovaps 96(%rdi), %ymm12
; AVX2-FCP-NEXT:    vmovaps 192(%rsi), %ymm6
; AVX2-FCP-NEXT:    vmovaps 160(%rsi), %ymm10
; AVX2-FCP-NEXT:    vmovaps 128(%rsi), %ymm13
; AVX2-FCP-NEXT:    vmovaps (%rsi), %ymm3
; AVX2-FCP-NEXT:    vmovaps 32(%rsi), %ymm7
; AVX2-FCP-NEXT:    vmovaps 64(%rsi), %ymm11
; AVX2-FCP-NEXT:    vmovaps 96(%rsi), %ymm14
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm3[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm15 = ymm1[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm0 = ymm15[0,1],ymm0[2,3],ymm15[4,5],ymm0[6,7]
; AVX2-FCP-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm3 = ymm3[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm0 = ymm1[0,1],ymm3[2,3],ymm1[4,5],ymm3[6,7]
; AVX2-FCP-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm3 = ymm7[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm15 = ymm4[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm3 = ymm15[0,1],ymm3[2,3],ymm15[4,5],ymm3[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm7 = ymm7[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm4 = ymm4[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm4 = ymm4[0,1],ymm7[2,3],ymm4[4,5],ymm7[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm7 = ymm11[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm15 = ymm8[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm7 = ymm15[0,1],ymm7[2,3],ymm15[4,5],ymm7[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm11 = ymm11[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm8 = ymm8[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm8 = ymm8[0,1],ymm11[2,3],ymm8[4,5],ymm11[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm11 = ymm14[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm15 = ymm12[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm11 = ymm15[0,1],ymm11[2,3],ymm15[4,5],ymm11[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm14 = ymm14[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm12 = ymm12[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm12 = ymm12[0,1],ymm14[2,3],ymm12[4,5],ymm14[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm14 = ymm13[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm15 = ymm9[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm14 = ymm15[0,1],ymm14[2,3],ymm15[4,5],ymm14[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm13 = ymm13[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm9 = ymm9[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm9 = ymm9[0,1],ymm13[2,3],ymm9[4,5],ymm13[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm13 = ymm10[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm15 = ymm5[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm13 = ymm15[0,1],ymm13[2,3],ymm15[4,5],ymm13[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm10 = ymm10[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm5 = ymm5[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm5 = ymm5[0,1],ymm10[2,3],ymm5[4,5],ymm10[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm10 = ymm6[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm15 = ymm2[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm10 = ymm15[0,1],ymm10[2,3],ymm15[4,5],ymm10[6,7]
; AVX2-FCP-NEXT:    vmovaps 224(%rdi), %ymm15
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm6 = ymm6[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm2[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm2 = ymm2[0,1],ymm6[2,3],ymm2[4,5],ymm6[6,7]
; AVX2-FCP-NEXT:    vmovaps 224(%rsi), %ymm6
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm6[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm15[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm0 = ymm1[0,1],ymm0[2,3],ymm1[4,5],ymm0[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm6[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm6 = ymm15[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm1 = ymm6[0,1],ymm1[2,3],ymm6[4,5],ymm1[6,7]
; AVX2-FCP-NEXT:    vmovaps %ymm1, 448(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm0, 480(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm2, 384(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm10, 416(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm5, 320(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm13, 352(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm9, 256(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm14, 288(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm12, 192(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm11, 224(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm8, 128(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm7, 160(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm4, 64(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm3, 96(%rdx)
; AVX2-FCP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FCP-NEXT:    vmovaps %ymm0, (%rdx)
; AVX2-FCP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FCP-NEXT:    vmovaps %ymm0, 32(%rdx)
; AVX2-FCP-NEXT:    vzeroupper
; AVX2-FCP-NEXT:    retq
;
; AVX512-LABEL: store_i64_stride2_vf32:
; AVX512:       # %bb.0:
; AVX512-NEXT:    vmovdqa64 (%rdi), %zmm0
; AVX512-NEXT:    vmovdqa64 64(%rdi), %zmm1
; AVX512-NEXT:    vmovdqa64 128(%rdi), %zmm2
; AVX512-NEXT:    vmovdqa64 192(%rdi), %zmm3
; AVX512-NEXT:    vmovdqa64 (%rsi), %zmm4
; AVX512-NEXT:    vmovdqa64 64(%rsi), %zmm5
; AVX512-NEXT:    vmovdqa64 128(%rsi), %zmm6
; AVX512-NEXT:    vmovdqa64 192(%rsi), %zmm7
; AVX512-NEXT:    vpmovsxbq {{.*#+}} zmm8 = [4,12,5,13,6,14,7,15]
; AVX512-NEXT:    vmovdqa64 %zmm0, %zmm9
; AVX512-NEXT:    vpermt2q %zmm4, %zmm8, %zmm9
; AVX512-NEXT:    vpmovsxbq {{.*#+}} zmm10 = [0,8,1,9,2,10,3,11]
; AVX512-NEXT:    vpermt2q %zmm4, %zmm10, %zmm0
; AVX512-NEXT:    vmovdqa64 %zmm1, %zmm4
; AVX512-NEXT:    vpermt2q %zmm5, %zmm8, %zmm4
; AVX512-NEXT:    vpermt2q %zmm5, %zmm10, %zmm1
; AVX512-NEXT:    vmovdqa64 %zmm2, %zmm5
; AVX512-NEXT:    vpermt2q %zmm6, %zmm8, %zmm5
; AVX512-NEXT:    vpermt2q %zmm6, %zmm10, %zmm2
; AVX512-NEXT:    vpermi2q %zmm7, %zmm3, %zmm8
; AVX512-NEXT:    vpermt2q %zmm7, %zmm10, %zmm3
; AVX512-NEXT:    vmovdqa64 %zmm3, 384(%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm8, 448(%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm2, 256(%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm5, 320(%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm1, 128(%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm4, 192(%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm0, (%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm9, 64(%rdx)
; AVX512-NEXT:    vzeroupper
; AVX512-NEXT:    retq
;
; AVX512-FCP-LABEL: store_i64_stride2_vf32:
; AVX512-FCP:       # %bb.0:
; AVX512-FCP-NEXT:    vmovdqa64 (%rdi), %zmm0
; AVX512-FCP-NEXT:    vmovdqa64 64(%rdi), %zmm1
; AVX512-FCP-NEXT:    vmovdqa64 128(%rdi), %zmm2
; AVX512-FCP-NEXT:    vmovdqa64 192(%rdi), %zmm3
; AVX512-FCP-NEXT:    vmovdqa64 (%rsi), %zmm4
; AVX512-FCP-NEXT:    vmovdqa64 64(%rsi), %zmm5
; AVX512-FCP-NEXT:    vmovdqa64 128(%rsi), %zmm6
; AVX512-FCP-NEXT:    vmovdqa64 192(%rsi), %zmm7
; AVX512-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm8 = [4,12,5,13,6,14,7,15]
; AVX512-FCP-NEXT:    vmovdqa64 %zmm0, %zmm9
; AVX512-FCP-NEXT:    vpermt2q %zmm4, %zmm8, %zmm9
; AVX512-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm10 = [0,8,1,9,2,10,3,11]
; AVX512-FCP-NEXT:    vpermt2q %zmm4, %zmm10, %zmm0
; AVX512-FCP-NEXT:    vmovdqa64 %zmm1, %zmm4
; AVX512-FCP-NEXT:    vpermt2q %zmm5, %zmm8, %zmm4
; AVX512-FCP-NEXT:    vpermt2q %zmm5, %zmm10, %zmm1
; AVX512-FCP-NEXT:    vmovdqa64 %zmm2, %zmm5
; AVX512-FCP-NEXT:    vpermt2q %zmm6, %zmm8, %zmm5
; AVX512-FCP-NEXT:    vpermt2q %zmm6, %zmm10, %zmm2
; AVX512-FCP-NEXT:    vpermi2q %zmm7, %zmm3, %zmm8
; AVX512-FCP-NEXT:    vpermt2q %zmm7, %zmm10, %zmm3
; AVX512-FCP-NEXT:    vmovdqa64 %zmm3, 384(%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm8, 448(%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm2, 256(%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm5, 320(%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm1, 128(%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm4, 192(%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm0, (%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm9, 64(%rdx)
; AVX512-FCP-NEXT:    vzeroupper
; AVX512-FCP-NEXT:    retq
;
; AVX512DQ-LABEL: store_i64_stride2_vf32:
; AVX512DQ:       # %bb.0:
; AVX512DQ-NEXT:    vmovdqa64 (%rdi), %zmm0
; AVX512DQ-NEXT:    vmovdqa64 64(%rdi), %zmm1
; AVX512DQ-NEXT:    vmovdqa64 128(%rdi), %zmm2
; AVX512DQ-NEXT:    vmovdqa64 192(%rdi), %zmm3
; AVX512DQ-NEXT:    vmovdqa64 (%rsi), %zmm4
; AVX512DQ-NEXT:    vmovdqa64 64(%rsi), %zmm5
; AVX512DQ-NEXT:    vmovdqa64 128(%rsi), %zmm6
; AVX512DQ-NEXT:    vmovdqa64 192(%rsi), %zmm7
; AVX512DQ-NEXT:    vpmovsxbq {{.*#+}} zmm8 = [4,12,5,13,6,14,7,15]
; AVX512DQ-NEXT:    vmovdqa64 %zmm0, %zmm9
; AVX512DQ-NEXT:    vpermt2q %zmm4, %zmm8, %zmm9
; AVX512DQ-NEXT:    vpmovsxbq {{.*#+}} zmm10 = [0,8,1,9,2,10,3,11]
; AVX512DQ-NEXT:    vpermt2q %zmm4, %zmm10, %zmm0
; AVX512DQ-NEXT:    vmovdqa64 %zmm1, %zmm4
; AVX512DQ-NEXT:    vpermt2q %zmm5, %zmm8, %zmm4
; AVX512DQ-NEXT:    vpermt2q %zmm5, %zmm10, %zmm1
; AVX512DQ-NEXT:    vmovdqa64 %zmm2, %zmm5
; AVX512DQ-NEXT:    vpermt2q %zmm6, %zmm8, %zmm5
; AVX512DQ-NEXT:    vpermt2q %zmm6, %zmm10, %zmm2
; AVX512DQ-NEXT:    vpermi2q %zmm7, %zmm3, %zmm8
; AVX512DQ-NEXT:    vpermt2q %zmm7, %zmm10, %zmm3
; AVX512DQ-NEXT:    vmovdqa64 %zmm3, 384(%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm8, 448(%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm2, 256(%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm5, 320(%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm1, 128(%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm4, 192(%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm0, (%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm9, 64(%rdx)
; AVX512DQ-NEXT:    vzeroupper
; AVX512DQ-NEXT:    retq
;
; AVX512DQ-FCP-LABEL: store_i64_stride2_vf32:
; AVX512DQ-FCP:       # %bb.0:
; AVX512DQ-FCP-NEXT:    vmovdqa64 (%rdi), %zmm0
; AVX512DQ-FCP-NEXT:    vmovdqa64 64(%rdi), %zmm1
; AVX512DQ-FCP-NEXT:    vmovdqa64 128(%rdi), %zmm2
; AVX512DQ-FCP-NEXT:    vmovdqa64 192(%rdi), %zmm3
; AVX512DQ-FCP-NEXT:    vmovdqa64 (%rsi), %zmm4
; AVX512DQ-FCP-NEXT:    vmovdqa64 64(%rsi), %zmm5
; AVX512DQ-FCP-NEXT:    vmovdqa64 128(%rsi), %zmm6
; AVX512DQ-FCP-NEXT:    vmovdqa64 192(%rsi), %zmm7
; AVX512DQ-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm8 = [4,12,5,13,6,14,7,15]
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm0, %zmm9
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm4, %zmm8, %zmm9
; AVX512DQ-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm10 = [0,8,1,9,2,10,3,11]
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm4, %zmm10, %zmm0
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm1, %zmm4
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm5, %zmm8, %zmm4
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm5, %zmm10, %zmm1
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm2, %zmm5
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm6, %zmm8, %zmm5
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm6, %zmm10, %zmm2
; AVX512DQ-FCP-NEXT:    vpermi2q %zmm7, %zmm3, %zmm8
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm7, %zmm10, %zmm3
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm3, 384(%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm8, 448(%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm2, 256(%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm5, 320(%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm1, 128(%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm4, 192(%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm0, (%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm9, 64(%rdx)
; AVX512DQ-FCP-NEXT:    vzeroupper
; AVX512DQ-FCP-NEXT:    retq
;
; AVX512BW-LABEL: store_i64_stride2_vf32:
; AVX512BW:       # %bb.0:
; AVX512BW-NEXT:    vmovdqa64 (%rdi), %zmm0
; AVX512BW-NEXT:    vmovdqa64 64(%rdi), %zmm1
; AVX512BW-NEXT:    vmovdqa64 128(%rdi), %zmm2
; AVX512BW-NEXT:    vmovdqa64 192(%rdi), %zmm3
; AVX512BW-NEXT:    vmovdqa64 (%rsi), %zmm4
; AVX512BW-NEXT:    vmovdqa64 64(%rsi), %zmm5
; AVX512BW-NEXT:    vmovdqa64 128(%rsi), %zmm6
; AVX512BW-NEXT:    vmovdqa64 192(%rsi), %zmm7
; AVX512BW-NEXT:    vpmovsxbq {{.*#+}} zmm8 = [4,12,5,13,6,14,7,15]
; AVX512BW-NEXT:    vmovdqa64 %zmm0, %zmm9
; AVX512BW-NEXT:    vpermt2q %zmm4, %zmm8, %zmm9
; AVX512BW-NEXT:    vpmovsxbq {{.*#+}} zmm10 = [0,8,1,9,2,10,3,11]
; AVX512BW-NEXT:    vpermt2q %zmm4, %zmm10, %zmm0
; AVX512BW-NEXT:    vmovdqa64 %zmm1, %zmm4
; AVX512BW-NEXT:    vpermt2q %zmm5, %zmm8, %zmm4
; AVX512BW-NEXT:    vpermt2q %zmm5, %zmm10, %zmm1
; AVX512BW-NEXT:    vmovdqa64 %zmm2, %zmm5
; AVX512BW-NEXT:    vpermt2q %zmm6, %zmm8, %zmm5
; AVX512BW-NEXT:    vpermt2q %zmm6, %zmm10, %zmm2
; AVX512BW-NEXT:    vpermi2q %zmm7, %zmm3, %zmm8
; AVX512BW-NEXT:    vpermt2q %zmm7, %zmm10, %zmm3
; AVX512BW-NEXT:    vmovdqa64 %zmm3, 384(%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm8, 448(%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm2, 256(%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm5, 320(%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm1, 128(%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm4, 192(%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm0, (%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm9, 64(%rdx)
; AVX512BW-NEXT:    vzeroupper
; AVX512BW-NEXT:    retq
;
; AVX512BW-FCP-LABEL: store_i64_stride2_vf32:
; AVX512BW-FCP:       # %bb.0:
; AVX512BW-FCP-NEXT:    vmovdqa64 (%rdi), %zmm0
; AVX512BW-FCP-NEXT:    vmovdqa64 64(%rdi), %zmm1
; AVX512BW-FCP-NEXT:    vmovdqa64 128(%rdi), %zmm2
; AVX512BW-FCP-NEXT:    vmovdqa64 192(%rdi), %zmm3
; AVX512BW-FCP-NEXT:    vmovdqa64 (%rsi), %zmm4
; AVX512BW-FCP-NEXT:    vmovdqa64 64(%rsi), %zmm5
; AVX512BW-FCP-NEXT:    vmovdqa64 128(%rsi), %zmm6
; AVX512BW-FCP-NEXT:    vmovdqa64 192(%rsi), %zmm7
; AVX512BW-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm8 = [4,12,5,13,6,14,7,15]
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm0, %zmm9
; AVX512BW-FCP-NEXT:    vpermt2q %zmm4, %zmm8, %zmm9
; AVX512BW-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm10 = [0,8,1,9,2,10,3,11]
; AVX512BW-FCP-NEXT:    vpermt2q %zmm4, %zmm10, %zmm0
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm1, %zmm4
; AVX512BW-FCP-NEXT:    vpermt2q %zmm5, %zmm8, %zmm4
; AVX512BW-FCP-NEXT:    vpermt2q %zmm5, %zmm10, %zmm1
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm2, %zmm5
; AVX512BW-FCP-NEXT:    vpermt2q %zmm6, %zmm8, %zmm5
; AVX512BW-FCP-NEXT:    vpermt2q %zmm6, %zmm10, %zmm2
; AVX512BW-FCP-NEXT:    vpermi2q %zmm7, %zmm3, %zmm8
; AVX512BW-FCP-NEXT:    vpermt2q %zmm7, %zmm10, %zmm3
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm3, 384(%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm8, 448(%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm2, 256(%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm5, 320(%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm1, 128(%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm4, 192(%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm0, (%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm9, 64(%rdx)
; AVX512BW-FCP-NEXT:    vzeroupper
; AVX512BW-FCP-NEXT:    retq
;
; AVX512DQ-BW-LABEL: store_i64_stride2_vf32:
; AVX512DQ-BW:       # %bb.0:
; AVX512DQ-BW-NEXT:    vmovdqa64 (%rdi), %zmm0
; AVX512DQ-BW-NEXT:    vmovdqa64 64(%rdi), %zmm1
; AVX512DQ-BW-NEXT:    vmovdqa64 128(%rdi), %zmm2
; AVX512DQ-BW-NEXT:    vmovdqa64 192(%rdi), %zmm3
; AVX512DQ-BW-NEXT:    vmovdqa64 (%rsi), %zmm4
; AVX512DQ-BW-NEXT:    vmovdqa64 64(%rsi), %zmm5
; AVX512DQ-BW-NEXT:    vmovdqa64 128(%rsi), %zmm6
; AVX512DQ-BW-NEXT:    vmovdqa64 192(%rsi), %zmm7
; AVX512DQ-BW-NEXT:    vpmovsxbq {{.*#+}} zmm8 = [4,12,5,13,6,14,7,15]
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm0, %zmm9
; AVX512DQ-BW-NEXT:    vpermt2q %zmm4, %zmm8, %zmm9
; AVX512DQ-BW-NEXT:    vpmovsxbq {{.*#+}} zmm10 = [0,8,1,9,2,10,3,11]
; AVX512DQ-BW-NEXT:    vpermt2q %zmm4, %zmm10, %zmm0
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm1, %zmm4
; AVX512DQ-BW-NEXT:    vpermt2q %zmm5, %zmm8, %zmm4
; AVX512DQ-BW-NEXT:    vpermt2q %zmm5, %zmm10, %zmm1
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm2, %zmm5
; AVX512DQ-BW-NEXT:    vpermt2q %zmm6, %zmm8, %zmm5
; AVX512DQ-BW-NEXT:    vpermt2q %zmm6, %zmm10, %zmm2
; AVX512DQ-BW-NEXT:    vpermi2q %zmm7, %zmm3, %zmm8
; AVX512DQ-BW-NEXT:    vpermt2q %zmm7, %zmm10, %zmm3
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm3, 384(%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm8, 448(%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm2, 256(%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm5, 320(%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm1, 128(%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm4, 192(%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm0, (%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm9, 64(%rdx)
; AVX512DQ-BW-NEXT:    vzeroupper
; AVX512DQ-BW-NEXT:    retq
;
; AVX512DQ-BW-FCP-LABEL: store_i64_stride2_vf32:
; AVX512DQ-BW-FCP:       # %bb.0:
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 (%rdi), %zmm0
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 64(%rdi), %zmm1
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 128(%rdi), %zmm2
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 192(%rdi), %zmm3
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 (%rsi), %zmm4
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 64(%rsi), %zmm5
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 128(%rsi), %zmm6
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 192(%rsi), %zmm7
; AVX512DQ-BW-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm8 = [4,12,5,13,6,14,7,15]
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm0, %zmm9
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm4, %zmm8, %zmm9
; AVX512DQ-BW-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm10 = [0,8,1,9,2,10,3,11]
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm4, %zmm10, %zmm0
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm1, %zmm4
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm5, %zmm8, %zmm4
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm5, %zmm10, %zmm1
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm2, %zmm5
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm6, %zmm8, %zmm5
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm6, %zmm10, %zmm2
; AVX512DQ-BW-FCP-NEXT:    vpermi2q %zmm7, %zmm3, %zmm8
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm7, %zmm10, %zmm3
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm3, 384(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm8, 448(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm2, 256(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm5, 320(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm1, 128(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm4, 192(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm0, (%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm9, 64(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vzeroupper
; AVX512DQ-BW-FCP-NEXT:    retq
  %in.vec0 = load <32 x i64>, ptr %in.vecptr0, align 64
  %in.vec1 = load <32 x i64>, ptr %in.vecptr1, align 64
  %1 = shufflevector <32 x i64> %in.vec0, <32 x i64> %in.vec1, <64 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31, i32 32, i32 33, i32 34, i32 35, i32 36, i32 37, i32 38, i32 39, i32 40, i32 41, i32 42, i32 43, i32 44, i32 45, i32 46, i32 47, i32 48, i32 49, i32 50, i32 51, i32 52, i32 53, i32 54, i32 55, i32 56, i32 57, i32 58, i32 59, i32 60, i32 61, i32 62, i32 63>
  %interleaved.vec = shufflevector <64 x i64> %1, <64 x i64> poison, <64 x i32> <i32 0, i32 32, i32 1, i32 33, i32 2, i32 34, i32 3, i32 35, i32 4, i32 36, i32 5, i32 37, i32 6, i32 38, i32 7, i32 39, i32 8, i32 40, i32 9, i32 41, i32 10, i32 42, i32 11, i32 43, i32 12, i32 44, i32 13, i32 45, i32 14, i32 46, i32 15, i32 47, i32 16, i32 48, i32 17, i32 49, i32 18, i32 50, i32 19, i32 51, i32 20, i32 52, i32 21, i32 53, i32 22, i32 54, i32 23, i32 55, i32 24, i32 56, i32 25, i32 57, i32 26, i32 58, i32 27, i32 59, i32 28, i32 60, i32 29, i32 61, i32 30, i32 62, i32 31, i32 63>
  store <64 x i64> %interleaved.vec, ptr %out.vec, align 64
  ret void
}

define void @store_i64_stride2_vf64(ptr %in.vecptr0, ptr %in.vecptr1, ptr %out.vec) nounwind {
; SSE-LABEL: store_i64_stride2_vf64:
; SSE:       # %bb.0:
; SSE-NEXT:    subq $664, %rsp # imm = 0x298
; SSE-NEXT:    movaps 112(%rdi), %xmm14
; SSE-NEXT:    movaps 96(%rdi), %xmm13
; SSE-NEXT:    movaps 80(%rdi), %xmm11
; SSE-NEXT:    movaps 64(%rdi), %xmm10
; SSE-NEXT:    movaps (%rdi), %xmm7
; SSE-NEXT:    movaps 16(%rdi), %xmm8
; SSE-NEXT:    movaps 32(%rdi), %xmm9
; SSE-NEXT:    movaps 48(%rdi), %xmm12
; SSE-NEXT:    movaps 96(%rsi), %xmm0
; SSE-NEXT:    movaps 80(%rsi), %xmm1
; SSE-NEXT:    movaps 64(%rsi), %xmm2
; SSE-NEXT:    movaps (%rsi), %xmm3
; SSE-NEXT:    movaps 16(%rsi), %xmm4
; SSE-NEXT:    movaps 32(%rsi), %xmm5
; SSE-NEXT:    movaps 48(%rsi), %xmm6
; SSE-NEXT:    movaps %xmm7, %xmm15
; SSE-NEXT:    movlhps {{.*#+}} xmm15 = xmm15[0],xmm3[0]
; SSE-NEXT:    movaps %xmm15, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm7 = xmm7[1],xmm3[1]
; SSE-NEXT:    movaps %xmm7, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps %xmm8, %xmm3
; SSE-NEXT:    movlhps {{.*#+}} xmm3 = xmm3[0],xmm4[0]
; SSE-NEXT:    movaps %xmm3, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm8 = xmm8[1],xmm4[1]
; SSE-NEXT:    movaps %xmm8, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps %xmm9, %xmm3
; SSE-NEXT:    movlhps {{.*#+}} xmm3 = xmm3[0],xmm5[0]
; SSE-NEXT:    movaps %xmm3, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm9 = xmm9[1],xmm5[1]
; SSE-NEXT:    movaps %xmm9, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps %xmm12, %xmm3
; SSE-NEXT:    movlhps {{.*#+}} xmm3 = xmm3[0],xmm6[0]
; SSE-NEXT:    movaps %xmm3, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm12 = xmm12[1],xmm6[1]
; SSE-NEXT:    movaps %xmm12, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps %xmm10, %xmm3
; SSE-NEXT:    movlhps {{.*#+}} xmm3 = xmm3[0],xmm2[0]
; SSE-NEXT:    movaps %xmm3, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm10 = xmm10[1],xmm2[1]
; SSE-NEXT:    movaps %xmm10, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps %xmm11, %xmm2
; SSE-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm1[0]
; SSE-NEXT:    movaps %xmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm11 = xmm11[1],xmm1[1]
; SSE-NEXT:    movaps %xmm11, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps %xmm13, %xmm1
; SSE-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm13 = xmm13[1],xmm0[1]
; SSE-NEXT:    movaps %xmm13, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps 112(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm14, %xmm1
; SSE-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm14 = xmm14[1],xmm0[1]
; SSE-NEXT:    movaps %xmm14, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps 128(%rdi), %xmm1
; SSE-NEXT:    movaps 128(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm1, %xmm2
; SSE-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm0[0]
; SSE-NEXT:    movaps %xmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm1 = xmm1[1],xmm0[1]
; SSE-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps 144(%rdi), %xmm1
; SSE-NEXT:    movaps 144(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm1, %xmm2
; SSE-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm0[0]
; SSE-NEXT:    movaps %xmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm1 = xmm1[1],xmm0[1]
; SSE-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps 160(%rdi), %xmm1
; SSE-NEXT:    movaps 160(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm1, %xmm2
; SSE-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm0[0]
; SSE-NEXT:    movaps %xmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm1 = xmm1[1],xmm0[1]
; SSE-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps 176(%rdi), %xmm1
; SSE-NEXT:    movaps 176(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm1, %xmm2
; SSE-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm0[0]
; SSE-NEXT:    movaps %xmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm1 = xmm1[1],xmm0[1]
; SSE-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps 192(%rdi), %xmm1
; SSE-NEXT:    movaps 192(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm1, %xmm2
; SSE-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm0[0]
; SSE-NEXT:    movaps %xmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm1 = xmm1[1],xmm0[1]
; SSE-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps 208(%rdi), %xmm1
; SSE-NEXT:    movaps 208(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm1, %xmm2
; SSE-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm0[0]
; SSE-NEXT:    movaps %xmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm1 = xmm1[1],xmm0[1]
; SSE-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps 224(%rdi), %xmm1
; SSE-NEXT:    movaps 224(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm1, %xmm2
; SSE-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm0[0]
; SSE-NEXT:    movaps %xmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm1 = xmm1[1],xmm0[1]
; SSE-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps 240(%rdi), %xmm1
; SSE-NEXT:    movaps 240(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm1, %xmm2
; SSE-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm0[0]
; SSE-NEXT:    movaps %xmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm1 = xmm1[1],xmm0[1]
; SSE-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps 256(%rdi), %xmm1
; SSE-NEXT:    movaps 256(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm1, %xmm2
; SSE-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm0[0]
; SSE-NEXT:    movaps %xmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm1 = xmm1[1],xmm0[1]
; SSE-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps 272(%rdi), %xmm1
; SSE-NEXT:    movaps 272(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm1, %xmm2
; SSE-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm0[0]
; SSE-NEXT:    movaps %xmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm1 = xmm1[1],xmm0[1]
; SSE-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps 288(%rdi), %xmm1
; SSE-NEXT:    movaps 288(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm1, %xmm2
; SSE-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm0[0]
; SSE-NEXT:    movaps %xmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm1 = xmm1[1],xmm0[1]
; SSE-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps 304(%rdi), %xmm1
; SSE-NEXT:    movaps 304(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm1, %xmm2
; SSE-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm0[0]
; SSE-NEXT:    movaps %xmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm1 = xmm1[1],xmm0[1]
; SSE-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps 320(%rdi), %xmm1
; SSE-NEXT:    movaps 320(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm1, %xmm2
; SSE-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm0[0]
; SSE-NEXT:    movaps %xmm2, (%rsp) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm1 = xmm1[1],xmm0[1]
; SSE-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps 336(%rdi), %xmm1
; SSE-NEXT:    movaps 336(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm1, %xmm2
; SSE-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm0[0]
; SSE-NEXT:    movaps %xmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm1 = xmm1[1],xmm0[1]
; SSE-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps 352(%rdi), %xmm1
; SSE-NEXT:    movaps 352(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm1, %xmm2
; SSE-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm0[0]
; SSE-NEXT:    movaps %xmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm1 = xmm1[1],xmm0[1]
; SSE-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    movaps 368(%rdi), %xmm15
; SSE-NEXT:    movaps 368(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm15, %xmm1
; SSE-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm15 = xmm15[1],xmm0[1]
; SSE-NEXT:    movaps 384(%rdi), %xmm13
; SSE-NEXT:    movaps 384(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm13, %xmm1
; SSE-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm13 = xmm13[1],xmm0[1]
; SSE-NEXT:    movaps 400(%rdi), %xmm11
; SSE-NEXT:    movaps 400(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm11, %xmm1
; SSE-NEXT:    movlhps {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE-NEXT:    movaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; SSE-NEXT:    unpckhpd {{.*#+}} xmm11 = xmm11[1],xmm0[1]
; SSE-NEXT:    movaps 416(%rdi), %xmm12
; SSE-NEXT:    movaps 416(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm12, %xmm14
; SSE-NEXT:    movlhps {{.*#+}} xmm14 = xmm14[0],xmm0[0]
; SSE-NEXT:    unpckhpd {{.*#+}} xmm12 = xmm12[1],xmm0[1]
; SSE-NEXT:    movaps 432(%rdi), %xmm8
; SSE-NEXT:    movaps 432(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm8, %xmm10
; SSE-NEXT:    movlhps {{.*#+}} xmm10 = xmm10[0],xmm0[0]
; SSE-NEXT:    unpckhpd {{.*#+}} xmm8 = xmm8[1],xmm0[1]
; SSE-NEXT:    movaps 448(%rdi), %xmm6
; SSE-NEXT:    movaps 448(%rsi), %xmm0
; SSE-NEXT:    movaps %xmm6, %xmm9
; SSE-NEXT:    movlhps {{.*#+}} xmm9 = xmm9[0],xmm0[0]
; SSE-NEXT:    unpckhpd {{.*#+}} xmm6 = xmm6[1],xmm0[1]
; SSE-NEXT:    movaps 464(%rdi), %xmm5
; SSE-NEXT:    movaps 464(%rsi), %xmm1
; SSE-NEXT:    movaps %xmm5, %xmm7
; SSE-NEXT:    movlhps {{.*#+}} xmm7 = xmm7[0],xmm1[0]
; SSE-NEXT:    unpckhpd {{.*#+}} xmm5 = xmm5[1],xmm1[1]
; SSE-NEXT:    movaps 480(%rdi), %xmm1
; SSE-NEXT:    movaps 480(%rsi), %xmm3
; SSE-NEXT:    movaps %xmm1, %xmm2
; SSE-NEXT:    movlhps {{.*#+}} xmm2 = xmm2[0],xmm3[0]
; SSE-NEXT:    unpckhpd {{.*#+}} xmm1 = xmm1[1],xmm3[1]
; SSE-NEXT:    movaps 496(%rdi), %xmm3
; SSE-NEXT:    movaps 496(%rsi), %xmm4
; SSE-NEXT:    movaps %xmm3, %xmm0
; SSE-NEXT:    movlhps {{.*#+}} xmm0 = xmm0[0],xmm4[0]
; SSE-NEXT:    unpckhpd {{.*#+}} xmm3 = xmm3[1],xmm4[1]
; SSE-NEXT:    movaps %xmm3, 1008(%rdx)
; SSE-NEXT:    movaps %xmm0, 992(%rdx)
; SSE-NEXT:    movaps %xmm1, 976(%rdx)
; SSE-NEXT:    movaps %xmm2, 960(%rdx)
; SSE-NEXT:    movaps %xmm5, 944(%rdx)
; SSE-NEXT:    movaps %xmm7, 928(%rdx)
; SSE-NEXT:    movaps %xmm6, 912(%rdx)
; SSE-NEXT:    movaps %xmm9, 896(%rdx)
; SSE-NEXT:    movaps %xmm8, 880(%rdx)
; SSE-NEXT:    movaps %xmm10, 864(%rdx)
; SSE-NEXT:    movaps %xmm12, 848(%rdx)
; SSE-NEXT:    movaps %xmm14, 832(%rdx)
; SSE-NEXT:    movaps %xmm11, 816(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 800(%rdx)
; SSE-NEXT:    movaps %xmm13, 784(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 768(%rdx)
; SSE-NEXT:    movaps %xmm15, 752(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 736(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 720(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 704(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 688(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 672(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 656(%rdx)
; SSE-NEXT:    movaps (%rsp), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 640(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 624(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 608(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 592(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 576(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 560(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 544(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 528(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 512(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 496(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 480(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 464(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 448(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 432(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 416(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 400(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 384(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 368(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 352(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 336(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 320(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 304(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 288(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 272(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 256(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 240(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 224(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 208(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 192(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 176(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 160(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 144(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 128(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 112(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 96(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 80(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 64(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 48(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 32(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, 16(%rdx)
; SSE-NEXT:    movaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; SSE-NEXT:    movaps %xmm0, (%rdx)
; SSE-NEXT:    addq $664, %rsp # imm = 0x298
; SSE-NEXT:    retq
;
; AVX-LABEL: store_i64_stride2_vf64:
; AVX:       # %bb.0:
; AVX-NEXT:    subq $424, %rsp # imm = 0x1A8
; AVX-NEXT:    vmovaps (%rsi), %xmm0
; AVX-NEXT:    vmovaps 32(%rsi), %xmm1
; AVX-NEXT:    vmovaps 64(%rsi), %xmm2
; AVX-NEXT:    vmovaps 96(%rsi), %xmm3
; AVX-NEXT:    vmovaps (%rdi), %xmm4
; AVX-NEXT:    vmovaps 32(%rdi), %xmm5
; AVX-NEXT:    vmovaps 64(%rdi), %xmm6
; AVX-NEXT:    vmovaps 96(%rdi), %xmm7
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm8 = xmm4[1],xmm0[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm4[0],xmm0[0]
; AVX-NEXT:    vinsertf128 $1, %xmm8, %ymm0, %ymm0
; AVX-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm0 = xmm5[1],xmm1[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm1 = xmm5[0],xmm1[0]
; AVX-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
; AVX-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm0 = xmm6[1],xmm2[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm1 = xmm6[0],xmm2[0]
; AVX-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
; AVX-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm0 = xmm7[1],xmm3[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm1 = xmm7[0],xmm3[0]
; AVX-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
; AVX-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX-NEXT:    vmovaps 128(%rsi), %xmm0
; AVX-NEXT:    vmovaps 128(%rdi), %xmm1
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm2 = xmm1[1],xmm0[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm1[0],xmm0[0]
; AVX-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
; AVX-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX-NEXT:    vmovaps 160(%rsi), %xmm0
; AVX-NEXT:    vmovaps 160(%rdi), %xmm1
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm2 = xmm1[1],xmm0[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm1[0],xmm0[0]
; AVX-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
; AVX-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX-NEXT:    vmovaps 192(%rsi), %xmm0
; AVX-NEXT:    vmovaps 192(%rdi), %xmm1
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm2 = xmm1[1],xmm0[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm1[0],xmm0[0]
; AVX-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
; AVX-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX-NEXT:    vmovaps 224(%rsi), %xmm0
; AVX-NEXT:    vmovaps 224(%rdi), %xmm1
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm2 = xmm1[1],xmm0[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm1[0],xmm0[0]
; AVX-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
; AVX-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX-NEXT:    vmovaps 256(%rsi), %xmm0
; AVX-NEXT:    vmovaps 256(%rdi), %xmm1
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm2 = xmm1[1],xmm0[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm1[0],xmm0[0]
; AVX-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
; AVX-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX-NEXT:    vmovaps 288(%rsi), %xmm0
; AVX-NEXT:    vmovaps 288(%rdi), %xmm1
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm2 = xmm1[1],xmm0[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm1[0],xmm0[0]
; AVX-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
; AVX-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX-NEXT:    vmovaps 320(%rsi), %xmm0
; AVX-NEXT:    vmovaps 320(%rdi), %xmm1
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm2 = xmm1[1],xmm0[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm1[0],xmm0[0]
; AVX-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
; AVX-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX-NEXT:    vmovaps 352(%rsi), %xmm0
; AVX-NEXT:    vmovaps 352(%rdi), %xmm1
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm2 = xmm1[1],xmm0[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm1[0],xmm0[0]
; AVX-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
; AVX-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX-NEXT:    vmovaps 384(%rsi), %xmm0
; AVX-NEXT:    vmovaps 384(%rdi), %xmm1
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm2 = xmm1[1],xmm0[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm1[0],xmm0[0]
; AVX-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
; AVX-NEXT:    vmovups %ymm0, (%rsp) # 32-byte Spill
; AVX-NEXT:    vmovaps 416(%rsi), %xmm0
; AVX-NEXT:    vmovaps 416(%rdi), %xmm1
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm2 = xmm1[1],xmm0[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm1[0],xmm0[0]
; AVX-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
; AVX-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX-NEXT:    vmovaps 448(%rsi), %xmm0
; AVX-NEXT:    vmovaps 448(%rdi), %xmm1
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm2 = xmm1[1],xmm0[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm1[0],xmm0[0]
; AVX-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
; AVX-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX-NEXT:    vmovaps 480(%rsi), %xmm0
; AVX-NEXT:    vmovaps 480(%rdi), %xmm1
; AVX-NEXT:    vunpckhpd {{.*#+}} xmm2 = xmm1[1],xmm0[1]
; AVX-NEXT:    vmovlhps {{.*#+}} xmm0 = xmm1[0],xmm0[0]
; AVX-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
; AVX-NEXT:    vmovupd %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm0 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm1 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm0 = ymm1[0],ymm0[0],ymm1[3],ymm0[3]
; AVX-NEXT:    vmovupd %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm0 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm1 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm6 = ymm1[0],ymm0[0],ymm1[3],ymm0[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm0 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm1 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm5 = ymm1[0],ymm0[0],ymm1[3],ymm0[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm0 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm1 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm4 = ymm1[0],ymm0[0],ymm1[3],ymm0[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm0 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm1 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm3 = ymm1[0],ymm0[0],ymm1[3],ymm0[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm0 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm1 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm2 = ymm1[0],ymm0[0],ymm1[3],ymm0[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm0 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm1 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm1 = ymm1[0],ymm0[0],ymm1[3],ymm0[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm0 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm8 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm0 = ymm8[0],ymm0[0],ymm8[3],ymm0[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm8 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm9 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm8 = ymm9[0],ymm8[0],ymm9[3],ymm8[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm9 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm10 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm9 = ymm10[0],ymm9[0],ymm10[3],ymm9[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm10 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm11 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm10 = ymm11[0],ymm10[0],ymm11[3],ymm10[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm11 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm12 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm11 = ymm12[0],ymm11[0],ymm12[3],ymm11[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm12 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm13 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm12 = ymm13[0],ymm12[0],ymm13[3],ymm12[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm13 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm14 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm13 = ymm14[0],ymm13[0],ymm14[3],ymm13[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm14 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm15 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm14 = ymm15[0],ymm14[0],ymm15[3],ymm14[3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm15 = mem[2,3,2,3]
; AVX-NEXT:    vperm2f128 {{.*#+}} ymm7 = mem[2,3,2,3]
; AVX-NEXT:    vshufpd {{.*#+}} ymm7 = ymm7[0],ymm15[0],ymm7[3],ymm15[3]
; AVX-NEXT:    vmovapd %ymm7, 992(%rdx)
; AVX-NEXT:    vmovapd %ymm14, 928(%rdx)
; AVX-NEXT:    vmovapd %ymm13, 864(%rdx)
; AVX-NEXT:    vmovapd %ymm12, 800(%rdx)
; AVX-NEXT:    vmovapd %ymm11, 736(%rdx)
; AVX-NEXT:    vmovapd %ymm10, 672(%rdx)
; AVX-NEXT:    vmovapd %ymm9, 608(%rdx)
; AVX-NEXT:    vmovapd %ymm8, 544(%rdx)
; AVX-NEXT:    vmovapd %ymm0, 480(%rdx)
; AVX-NEXT:    vmovapd %ymm1, 416(%rdx)
; AVX-NEXT:    vmovapd %ymm2, 352(%rdx)
; AVX-NEXT:    vmovapd %ymm3, 288(%rdx)
; AVX-NEXT:    vmovapd %ymm4, 224(%rdx)
; AVX-NEXT:    vmovapd %ymm5, 160(%rdx)
; AVX-NEXT:    vmovapd %ymm6, 96(%rdx)
; AVX-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX-NEXT:    vmovaps %ymm0, 32(%rdx)
; AVX-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX-NEXT:    vmovaps %ymm0, 960(%rdx)
; AVX-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX-NEXT:    vmovaps %ymm0, 896(%rdx)
; AVX-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX-NEXT:    vmovaps %ymm0, 832(%rdx)
; AVX-NEXT:    vmovups (%rsp), %ymm0 # 32-byte Reload
; AVX-NEXT:    vmovaps %ymm0, 768(%rdx)
; AVX-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX-NEXT:    vmovaps %ymm0, 704(%rdx)
; AVX-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX-NEXT:    vmovaps %ymm0, 640(%rdx)
; AVX-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX-NEXT:    vmovaps %ymm0, 576(%rdx)
; AVX-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX-NEXT:    vmovaps %ymm0, 512(%rdx)
; AVX-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX-NEXT:    vmovaps %ymm0, 448(%rdx)
; AVX-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX-NEXT:    vmovaps %ymm0, 384(%rdx)
; AVX-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX-NEXT:    vmovaps %ymm0, 320(%rdx)
; AVX-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX-NEXT:    vmovaps %ymm0, 256(%rdx)
; AVX-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX-NEXT:    vmovaps %ymm0, 192(%rdx)
; AVX-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX-NEXT:    vmovaps %ymm0, 128(%rdx)
; AVX-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX-NEXT:    vmovaps %ymm0, 64(%rdx)
; AVX-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX-NEXT:    vmovaps %ymm0, (%rdx)
; AVX-NEXT:    addq $424, %rsp # imm = 0x1A8
; AVX-NEXT:    vzeroupper
; AVX-NEXT:    retq
;
; AVX2-LABEL: store_i64_stride2_vf64:
; AVX2:       # %bb.0:
; AVX2-NEXT:    subq $456, %rsp # imm = 0x1C8
; AVX2-NEXT:    vmovaps 192(%rdi), %ymm0
; AVX2-NEXT:    vmovaps 160(%rdi), %ymm1
; AVX2-NEXT:    vmovaps 128(%rdi), %ymm2
; AVX2-NEXT:    vmovaps (%rdi), %ymm7
; AVX2-NEXT:    vmovaps 32(%rdi), %ymm8
; AVX2-NEXT:    vmovaps 64(%rdi), %ymm6
; AVX2-NEXT:    vmovaps 96(%rdi), %ymm4
; AVX2-NEXT:    vmovaps 192(%rsi), %ymm3
; AVX2-NEXT:    vmovaps 160(%rsi), %ymm5
; AVX2-NEXT:    vmovaps 128(%rsi), %ymm9
; AVX2-NEXT:    vmovaps (%rsi), %ymm10
; AVX2-NEXT:    vmovaps 32(%rsi), %ymm11
; AVX2-NEXT:    vmovaps 64(%rsi), %ymm12
; AVX2-NEXT:    vmovaps 96(%rsi), %ymm13
; AVX2-NEXT:    vpermpd {{.*#+}} ymm14 = ymm10[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm15 = ymm7[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm14 = ymm15[0,1],ymm14[2,3],ymm15[4,5],ymm14[6,7]
; AVX2-NEXT:    vmovups %ymm14, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-NEXT:    vpermpd {{.*#+}} ymm10 = ymm10[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm7 = ymm7[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm7 = ymm7[0,1],ymm10[2,3],ymm7[4,5],ymm10[6,7]
; AVX2-NEXT:    vmovups %ymm7, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-NEXT:    vpermpd {{.*#+}} ymm7 = ymm11[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm10 = ymm8[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm7 = ymm10[0,1],ymm7[2,3],ymm10[4,5],ymm7[6,7]
; AVX2-NEXT:    vmovups %ymm7, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-NEXT:    vpermpd {{.*#+}} ymm7 = ymm11[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm8 = ymm8[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm7 = ymm8[0,1],ymm7[2,3],ymm8[4,5],ymm7[6,7]
; AVX2-NEXT:    vmovups %ymm7, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-NEXT:    vpermpd {{.*#+}} ymm7 = ymm12[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm8 = ymm6[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm7 = ymm8[0,1],ymm7[2,3],ymm8[4,5],ymm7[6,7]
; AVX2-NEXT:    vmovups %ymm7, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-NEXT:    vpermpd {{.*#+}} ymm7 = ymm12[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm6 = ymm6[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm6 = ymm6[0,1],ymm7[2,3],ymm6[4,5],ymm7[6,7]
; AVX2-NEXT:    vmovups %ymm6, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-NEXT:    vpermpd {{.*#+}} ymm6 = ymm13[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm7 = ymm4[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm6 = ymm7[0,1],ymm6[2,3],ymm7[4,5],ymm6[6,7]
; AVX2-NEXT:    vmovups %ymm6, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-NEXT:    vpermpd {{.*#+}} ymm6 = ymm13[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm4 = ymm4[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm4 = ymm4[0,1],ymm6[2,3],ymm4[4,5],ymm6[6,7]
; AVX2-NEXT:    vmovups %ymm4, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-NEXT:    vpermpd {{.*#+}} ymm4 = ymm9[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm6 = ymm2[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm4 = ymm6[0,1],ymm4[2,3],ymm6[4,5],ymm4[6,7]
; AVX2-NEXT:    vmovups %ymm4, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-NEXT:    vpermpd {{.*#+}} ymm4 = ymm9[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm2 = ymm2[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm2 = ymm2[0,1],ymm4[2,3],ymm2[4,5],ymm4[6,7]
; AVX2-NEXT:    vmovups %ymm2, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-NEXT:    vpermpd {{.*#+}} ymm2 = ymm5[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm4 = ymm1[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm2 = ymm4[0,1],ymm2[2,3],ymm4[4,5],ymm2[6,7]
; AVX2-NEXT:    vmovups %ymm2, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-NEXT:    vpermpd {{.*#+}} ymm2 = ymm5[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm1 = ymm1[0,1],ymm2[2,3],ymm1[4,5],ymm2[6,7]
; AVX2-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-NEXT:    vpermpd {{.*#+}} ymm1 = ymm3[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm2 = ymm0[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm1 = ymm2[0,1],ymm1[2,3],ymm2[4,5],ymm1[6,7]
; AVX2-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-NEXT:    vmovaps 224(%rdi), %ymm1
; AVX2-NEXT:    vpermpd {{.*#+}} ymm2 = ymm3[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm0 = ymm0[0,1],ymm2[2,3],ymm0[4,5],ymm2[6,7]
; AVX2-NEXT:    vmovups %ymm0, (%rsp) # 32-byte Spill
; AVX2-NEXT:    vmovaps 224(%rsi), %ymm0
; AVX2-NEXT:    vpermpd {{.*#+}} ymm2 = ymm0[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm3 = ymm1[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm2 = ymm3[0,1],ymm2[2,3],ymm3[4,5],ymm2[6,7]
; AVX2-NEXT:    vmovups %ymm2, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm0 = ymm1[0,1],ymm0[2,3],ymm1[4,5],ymm0[6,7]
; AVX2-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-NEXT:    vmovaps 256(%rdi), %ymm0
; AVX2-NEXT:    vmovaps 256(%rsi), %ymm1
; AVX2-NEXT:    vpermpd {{.*#+}} ymm2 = ymm1[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm3 = ymm0[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm2 = ymm3[0,1],ymm2[2,3],ymm3[4,5],ymm2[6,7]
; AVX2-NEXT:    vmovups %ymm2, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3],ymm0[4,5],ymm1[6,7]
; AVX2-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-NEXT:    vmovaps 288(%rdi), %ymm0
; AVX2-NEXT:    vmovaps 288(%rsi), %ymm1
; AVX2-NEXT:    vpermpd {{.*#+}} ymm2 = ymm1[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm3 = ymm0[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm12 = ymm3[0,1],ymm2[2,3],ymm3[4,5],ymm2[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm10 = ymm0[0,1],ymm1[2,3],ymm0[4,5],ymm1[6,7]
; AVX2-NEXT:    vmovaps 320(%rdi), %ymm0
; AVX2-NEXT:    vmovaps 320(%rsi), %ymm1
; AVX2-NEXT:    vpermpd {{.*#+}} ymm2 = ymm1[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm3 = ymm0[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm8 = ymm3[0,1],ymm2[2,3],ymm3[4,5],ymm2[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm7 = ymm0[0,1],ymm1[2,3],ymm0[4,5],ymm1[6,7]
; AVX2-NEXT:    vmovaps 352(%rdi), %ymm0
; AVX2-NEXT:    vmovaps 352(%rsi), %ymm1
; AVX2-NEXT:    vpermpd {{.*#+}} ymm2 = ymm1[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm9 = ymm0[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm6 = ymm9[0,1],ymm2[2,3],ymm9[4,5],ymm2[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm5 = ymm0[0,1],ymm1[2,3],ymm0[4,5],ymm1[6,7]
; AVX2-NEXT:    vmovaps 384(%rdi), %ymm9
; AVX2-NEXT:    vmovaps 384(%rsi), %ymm0
; AVX2-NEXT:    vpermpd {{.*#+}} ymm1 = ymm0[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm11 = ymm9[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm4 = ymm11[0,1],ymm1[2,3],ymm11[4,5],ymm1[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm9 = ymm9[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm3 = ymm9[0,1],ymm0[2,3],ymm9[4,5],ymm0[6,7]
; AVX2-NEXT:    vmovaps 416(%rdi), %ymm9
; AVX2-NEXT:    vmovaps 416(%rsi), %ymm11
; AVX2-NEXT:    vpermpd {{.*#+}} ymm0 = ymm11[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm13 = ymm9[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm2 = ymm13[0,1],ymm0[2,3],ymm13[4,5],ymm0[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm11 = ymm11[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm9 = ymm9[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm9 = ymm9[0,1],ymm11[2,3],ymm9[4,5],ymm11[6,7]
; AVX2-NEXT:    vmovaps 448(%rdi), %ymm11
; AVX2-NEXT:    vmovaps 448(%rsi), %ymm13
; AVX2-NEXT:    vpermpd {{.*#+}} ymm0 = ymm13[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm15 = ymm11[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm1 = ymm15[0,1],ymm0[2,3],ymm15[4,5],ymm0[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm13 = ymm13[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm11 = ymm11[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm11 = ymm11[0,1],ymm13[2,3],ymm11[4,5],ymm13[6,7]
; AVX2-NEXT:    vmovaps 480(%rdi), %ymm13
; AVX2-NEXT:    vmovaps 480(%rsi), %ymm15
; AVX2-NEXT:    vpermpd {{.*#+}} ymm0 = ymm15[0,0,2,1]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm14 = ymm13[0,1,1,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm0 = ymm14[0,1],ymm0[2,3],ymm14[4,5],ymm0[6,7]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm14 = ymm15[0,2,2,3]
; AVX2-NEXT:    vpermpd {{.*#+}} ymm13 = ymm13[2,1,3,3]
; AVX2-NEXT:    vblendps {{.*#+}} ymm13 = ymm13[0,1],ymm14[2,3],ymm13[4,5],ymm14[6,7]
; AVX2-NEXT:    vmovaps %ymm13, 992(%rdx)
; AVX2-NEXT:    vmovaps %ymm0, 960(%rdx)
; AVX2-NEXT:    vmovaps %ymm11, 928(%rdx)
; AVX2-NEXT:    vmovaps %ymm1, 896(%rdx)
; AVX2-NEXT:    vmovaps %ymm9, 864(%rdx)
; AVX2-NEXT:    vmovaps %ymm2, 832(%rdx)
; AVX2-NEXT:    vmovaps %ymm3, 800(%rdx)
; AVX2-NEXT:    vmovaps %ymm4, 768(%rdx)
; AVX2-NEXT:    vmovaps %ymm5, 736(%rdx)
; AVX2-NEXT:    vmovaps %ymm6, 704(%rdx)
; AVX2-NEXT:    vmovaps %ymm7, 672(%rdx)
; AVX2-NEXT:    vmovaps %ymm8, 640(%rdx)
; AVX2-NEXT:    vmovaps %ymm10, 608(%rdx)
; AVX2-NEXT:    vmovaps %ymm12, 576(%rdx)
; AVX2-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT:    vmovaps %ymm0, 544(%rdx)
; AVX2-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT:    vmovaps %ymm0, 512(%rdx)
; AVX2-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT:    vmovaps %ymm0, 480(%rdx)
; AVX2-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT:    vmovaps %ymm0, 448(%rdx)
; AVX2-NEXT:    vmovups (%rsp), %ymm0 # 32-byte Reload
; AVX2-NEXT:    vmovaps %ymm0, 416(%rdx)
; AVX2-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT:    vmovaps %ymm0, 384(%rdx)
; AVX2-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT:    vmovaps %ymm0, 352(%rdx)
; AVX2-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT:    vmovaps %ymm0, 320(%rdx)
; AVX2-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT:    vmovaps %ymm0, 288(%rdx)
; AVX2-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT:    vmovaps %ymm0, 256(%rdx)
; AVX2-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT:    vmovaps %ymm0, 224(%rdx)
; AVX2-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT:    vmovaps %ymm0, 192(%rdx)
; AVX2-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT:    vmovaps %ymm0, 160(%rdx)
; AVX2-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT:    vmovaps %ymm0, 128(%rdx)
; AVX2-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT:    vmovaps %ymm0, 96(%rdx)
; AVX2-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT:    vmovaps %ymm0, 64(%rdx)
; AVX2-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT:    vmovaps %ymm0, 32(%rdx)
; AVX2-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-NEXT:    vmovaps %ymm0, (%rdx)
; AVX2-NEXT:    addq $456, %rsp # imm = 0x1C8
; AVX2-NEXT:    vzeroupper
; AVX2-NEXT:    retq
;
; AVX2-FP-LABEL: store_i64_stride2_vf64:
; AVX2-FP:       # %bb.0:
; AVX2-FP-NEXT:    subq $456, %rsp # imm = 0x1C8
; AVX2-FP-NEXT:    vmovaps 192(%rdi), %ymm0
; AVX2-FP-NEXT:    vmovaps 160(%rdi), %ymm1
; AVX2-FP-NEXT:    vmovaps 128(%rdi), %ymm2
; AVX2-FP-NEXT:    vmovaps (%rdi), %ymm7
; AVX2-FP-NEXT:    vmovaps 32(%rdi), %ymm8
; AVX2-FP-NEXT:    vmovaps 64(%rdi), %ymm6
; AVX2-FP-NEXT:    vmovaps 96(%rdi), %ymm4
; AVX2-FP-NEXT:    vmovaps 192(%rsi), %ymm3
; AVX2-FP-NEXT:    vmovaps 160(%rsi), %ymm5
; AVX2-FP-NEXT:    vmovaps 128(%rsi), %ymm9
; AVX2-FP-NEXT:    vmovaps (%rsi), %ymm10
; AVX2-FP-NEXT:    vmovaps 32(%rsi), %ymm11
; AVX2-FP-NEXT:    vmovaps 64(%rsi), %ymm12
; AVX2-FP-NEXT:    vmovaps 96(%rsi), %ymm13
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm14 = ymm10[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm15 = ymm7[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm14 = ymm15[0,1],ymm14[2,3],ymm15[4,5],ymm14[6,7]
; AVX2-FP-NEXT:    vmovups %ymm14, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm10 = ymm10[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm7 = ymm7[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm7 = ymm7[0,1],ymm10[2,3],ymm7[4,5],ymm10[6,7]
; AVX2-FP-NEXT:    vmovups %ymm7, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm7 = ymm11[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm10 = ymm8[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm7 = ymm10[0,1],ymm7[2,3],ymm10[4,5],ymm7[6,7]
; AVX2-FP-NEXT:    vmovups %ymm7, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm7 = ymm11[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm8 = ymm8[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm7 = ymm8[0,1],ymm7[2,3],ymm8[4,5],ymm7[6,7]
; AVX2-FP-NEXT:    vmovups %ymm7, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm7 = ymm12[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm8 = ymm6[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm7 = ymm8[0,1],ymm7[2,3],ymm8[4,5],ymm7[6,7]
; AVX2-FP-NEXT:    vmovups %ymm7, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm7 = ymm12[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm6 = ymm6[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm6 = ymm6[0,1],ymm7[2,3],ymm6[4,5],ymm7[6,7]
; AVX2-FP-NEXT:    vmovups %ymm6, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm6 = ymm13[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm7 = ymm4[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm6 = ymm7[0,1],ymm6[2,3],ymm7[4,5],ymm6[6,7]
; AVX2-FP-NEXT:    vmovups %ymm6, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm6 = ymm13[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm4 = ymm4[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm4 = ymm4[0,1],ymm6[2,3],ymm4[4,5],ymm6[6,7]
; AVX2-FP-NEXT:    vmovups %ymm4, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm4 = ymm9[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm6 = ymm2[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm4 = ymm6[0,1],ymm4[2,3],ymm6[4,5],ymm4[6,7]
; AVX2-FP-NEXT:    vmovups %ymm4, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm4 = ymm9[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm2[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm2 = ymm2[0,1],ymm4[2,3],ymm2[4,5],ymm4[6,7]
; AVX2-FP-NEXT:    vmovups %ymm2, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm5[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm4 = ymm1[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm2 = ymm4[0,1],ymm2[2,3],ymm4[4,5],ymm2[6,7]
; AVX2-FP-NEXT:    vmovups %ymm2, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm5[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm1 = ymm1[0,1],ymm2[2,3],ymm1[4,5],ymm2[6,7]
; AVX2-FP-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm3[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm0[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm1 = ymm2[0,1],ymm1[2,3],ymm2[4,5],ymm1[6,7]
; AVX2-FP-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FP-NEXT:    vmovaps 224(%rdi), %ymm1
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm3[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm0 = ymm0[0,1],ymm2[2,3],ymm0[4,5],ymm2[6,7]
; AVX2-FP-NEXT:    vmovups %ymm0, (%rsp) # 32-byte Spill
; AVX2-FP-NEXT:    vmovaps 224(%rsi), %ymm0
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm0[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm3 = ymm1[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm2 = ymm3[0,1],ymm2[2,3],ymm3[4,5],ymm2[6,7]
; AVX2-FP-NEXT:    vmovups %ymm2, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm0 = ymm1[0,1],ymm0[2,3],ymm1[4,5],ymm0[6,7]
; AVX2-FP-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FP-NEXT:    vmovaps 256(%rdi), %ymm0
; AVX2-FP-NEXT:    vmovaps 256(%rsi), %ymm1
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm1[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm3 = ymm0[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm2 = ymm3[0,1],ymm2[2,3],ymm3[4,5],ymm2[6,7]
; AVX2-FP-NEXT:    vmovups %ymm2, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3],ymm0[4,5],ymm1[6,7]
; AVX2-FP-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FP-NEXT:    vmovaps 288(%rdi), %ymm0
; AVX2-FP-NEXT:    vmovaps 288(%rsi), %ymm1
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm1[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm3 = ymm0[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm12 = ymm3[0,1],ymm2[2,3],ymm3[4,5],ymm2[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm10 = ymm0[0,1],ymm1[2,3],ymm0[4,5],ymm1[6,7]
; AVX2-FP-NEXT:    vmovaps 320(%rdi), %ymm0
; AVX2-FP-NEXT:    vmovaps 320(%rsi), %ymm1
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm1[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm3 = ymm0[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm8 = ymm3[0,1],ymm2[2,3],ymm3[4,5],ymm2[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm7 = ymm0[0,1],ymm1[2,3],ymm0[4,5],ymm1[6,7]
; AVX2-FP-NEXT:    vmovaps 352(%rdi), %ymm0
; AVX2-FP-NEXT:    vmovaps 352(%rsi), %ymm1
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm1[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm9 = ymm0[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm6 = ymm9[0,1],ymm2[2,3],ymm9[4,5],ymm2[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm5 = ymm0[0,1],ymm1[2,3],ymm0[4,5],ymm1[6,7]
; AVX2-FP-NEXT:    vmovaps 384(%rdi), %ymm9
; AVX2-FP-NEXT:    vmovaps 384(%rsi), %ymm0
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm0[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm11 = ymm9[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm4 = ymm11[0,1],ymm1[2,3],ymm11[4,5],ymm1[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm9 = ymm9[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm3 = ymm9[0,1],ymm0[2,3],ymm9[4,5],ymm0[6,7]
; AVX2-FP-NEXT:    vmovaps 416(%rdi), %ymm9
; AVX2-FP-NEXT:    vmovaps 416(%rsi), %ymm11
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm11[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm13 = ymm9[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm2 = ymm13[0,1],ymm0[2,3],ymm13[4,5],ymm0[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm11 = ymm11[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm9 = ymm9[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm9 = ymm9[0,1],ymm11[2,3],ymm9[4,5],ymm11[6,7]
; AVX2-FP-NEXT:    vmovaps 448(%rdi), %ymm11
; AVX2-FP-NEXT:    vmovaps 448(%rsi), %ymm13
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm13[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm15 = ymm11[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm1 = ymm15[0,1],ymm0[2,3],ymm15[4,5],ymm0[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm13 = ymm13[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm11 = ymm11[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm11 = ymm11[0,1],ymm13[2,3],ymm11[4,5],ymm13[6,7]
; AVX2-FP-NEXT:    vmovaps 480(%rdi), %ymm13
; AVX2-FP-NEXT:    vmovaps 480(%rsi), %ymm15
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm15[0,0,2,1]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm14 = ymm13[0,1,1,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm0 = ymm14[0,1],ymm0[2,3],ymm14[4,5],ymm0[6,7]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm14 = ymm15[0,2,2,3]
; AVX2-FP-NEXT:    vpermpd {{.*#+}} ymm13 = ymm13[2,1,3,3]
; AVX2-FP-NEXT:    vblendps {{.*#+}} ymm13 = ymm13[0,1],ymm14[2,3],ymm13[4,5],ymm14[6,7]
; AVX2-FP-NEXT:    vmovaps %ymm13, 992(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm0, 960(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm11, 928(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm1, 896(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm9, 864(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm2, 832(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm3, 800(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm4, 768(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm5, 736(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm6, 704(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm7, 672(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm8, 640(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm10, 608(%rdx)
; AVX2-FP-NEXT:    vmovaps %ymm12, 576(%rdx)
; AVX2-FP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FP-NEXT:    vmovaps %ymm0, 544(%rdx)
; AVX2-FP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FP-NEXT:    vmovaps %ymm0, 512(%rdx)
; AVX2-FP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FP-NEXT:    vmovaps %ymm0, 480(%rdx)
; AVX2-FP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FP-NEXT:    vmovaps %ymm0, 448(%rdx)
; AVX2-FP-NEXT:    vmovups (%rsp), %ymm0 # 32-byte Reload
; AVX2-FP-NEXT:    vmovaps %ymm0, 416(%rdx)
; AVX2-FP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FP-NEXT:    vmovaps %ymm0, 384(%rdx)
; AVX2-FP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FP-NEXT:    vmovaps %ymm0, 352(%rdx)
; AVX2-FP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FP-NEXT:    vmovaps %ymm0, 320(%rdx)
; AVX2-FP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FP-NEXT:    vmovaps %ymm0, 288(%rdx)
; AVX2-FP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FP-NEXT:    vmovaps %ymm0, 256(%rdx)
; AVX2-FP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FP-NEXT:    vmovaps %ymm0, 224(%rdx)
; AVX2-FP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FP-NEXT:    vmovaps %ymm0, 192(%rdx)
; AVX2-FP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FP-NEXT:    vmovaps %ymm0, 160(%rdx)
; AVX2-FP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FP-NEXT:    vmovaps %ymm0, 128(%rdx)
; AVX2-FP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FP-NEXT:    vmovaps %ymm0, 96(%rdx)
; AVX2-FP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FP-NEXT:    vmovaps %ymm0, 64(%rdx)
; AVX2-FP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FP-NEXT:    vmovaps %ymm0, 32(%rdx)
; AVX2-FP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FP-NEXT:    vmovaps %ymm0, (%rdx)
; AVX2-FP-NEXT:    addq $456, %rsp # imm = 0x1C8
; AVX2-FP-NEXT:    vzeroupper
; AVX2-FP-NEXT:    retq
;
; AVX2-FCP-LABEL: store_i64_stride2_vf64:
; AVX2-FCP:       # %bb.0:
; AVX2-FCP-NEXT:    subq $456, %rsp # imm = 0x1C8
; AVX2-FCP-NEXT:    vmovaps 192(%rdi), %ymm0
; AVX2-FCP-NEXT:    vmovaps 160(%rdi), %ymm1
; AVX2-FCP-NEXT:    vmovaps 128(%rdi), %ymm2
; AVX2-FCP-NEXT:    vmovaps (%rdi), %ymm7
; AVX2-FCP-NEXT:    vmovaps 32(%rdi), %ymm8
; AVX2-FCP-NEXT:    vmovaps 64(%rdi), %ymm6
; AVX2-FCP-NEXT:    vmovaps 96(%rdi), %ymm4
; AVX2-FCP-NEXT:    vmovaps 192(%rsi), %ymm3
; AVX2-FCP-NEXT:    vmovaps 160(%rsi), %ymm5
; AVX2-FCP-NEXT:    vmovaps 128(%rsi), %ymm9
; AVX2-FCP-NEXT:    vmovaps (%rsi), %ymm10
; AVX2-FCP-NEXT:    vmovaps 32(%rsi), %ymm11
; AVX2-FCP-NEXT:    vmovaps 64(%rsi), %ymm12
; AVX2-FCP-NEXT:    vmovaps 96(%rsi), %ymm13
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm14 = ymm10[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm15 = ymm7[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm14 = ymm15[0,1],ymm14[2,3],ymm15[4,5],ymm14[6,7]
; AVX2-FCP-NEXT:    vmovups %ymm14, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm10 = ymm10[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm7 = ymm7[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm7 = ymm7[0,1],ymm10[2,3],ymm7[4,5],ymm10[6,7]
; AVX2-FCP-NEXT:    vmovups %ymm7, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm7 = ymm11[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm10 = ymm8[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm7 = ymm10[0,1],ymm7[2,3],ymm10[4,5],ymm7[6,7]
; AVX2-FCP-NEXT:    vmovups %ymm7, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm7 = ymm11[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm8 = ymm8[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm7 = ymm8[0,1],ymm7[2,3],ymm8[4,5],ymm7[6,7]
; AVX2-FCP-NEXT:    vmovups %ymm7, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm7 = ymm12[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm8 = ymm6[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm7 = ymm8[0,1],ymm7[2,3],ymm8[4,5],ymm7[6,7]
; AVX2-FCP-NEXT:    vmovups %ymm7, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm7 = ymm12[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm6 = ymm6[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm6 = ymm6[0,1],ymm7[2,3],ymm6[4,5],ymm7[6,7]
; AVX2-FCP-NEXT:    vmovups %ymm6, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm6 = ymm13[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm7 = ymm4[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm6 = ymm7[0,1],ymm6[2,3],ymm7[4,5],ymm6[6,7]
; AVX2-FCP-NEXT:    vmovups %ymm6, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm6 = ymm13[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm4 = ymm4[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm4 = ymm4[0,1],ymm6[2,3],ymm4[4,5],ymm6[6,7]
; AVX2-FCP-NEXT:    vmovups %ymm4, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm4 = ymm9[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm6 = ymm2[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm4 = ymm6[0,1],ymm4[2,3],ymm6[4,5],ymm4[6,7]
; AVX2-FCP-NEXT:    vmovups %ymm4, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm4 = ymm9[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm2[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm2 = ymm2[0,1],ymm4[2,3],ymm2[4,5],ymm4[6,7]
; AVX2-FCP-NEXT:    vmovups %ymm2, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm5[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm4 = ymm1[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm2 = ymm4[0,1],ymm2[2,3],ymm4[4,5],ymm2[6,7]
; AVX2-FCP-NEXT:    vmovups %ymm2, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm5[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm1 = ymm1[0,1],ymm2[2,3],ymm1[4,5],ymm2[6,7]
; AVX2-FCP-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm3[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm0[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm1 = ymm2[0,1],ymm1[2,3],ymm2[4,5],ymm1[6,7]
; AVX2-FCP-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FCP-NEXT:    vmovaps 224(%rdi), %ymm1
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm3[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm0 = ymm0[0,1],ymm2[2,3],ymm0[4,5],ymm2[6,7]
; AVX2-FCP-NEXT:    vmovups %ymm0, (%rsp) # 32-byte Spill
; AVX2-FCP-NEXT:    vmovaps 224(%rsi), %ymm0
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm0[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm3 = ymm1[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm2 = ymm3[0,1],ymm2[2,3],ymm3[4,5],ymm2[6,7]
; AVX2-FCP-NEXT:    vmovups %ymm2, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm0 = ymm1[0,1],ymm0[2,3],ymm1[4,5],ymm0[6,7]
; AVX2-FCP-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FCP-NEXT:    vmovaps 256(%rdi), %ymm0
; AVX2-FCP-NEXT:    vmovaps 256(%rsi), %ymm1
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm1[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm3 = ymm0[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm2 = ymm3[0,1],ymm2[2,3],ymm3[4,5],ymm2[6,7]
; AVX2-FCP-NEXT:    vmovups %ymm2, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3],ymm0[4,5],ymm1[6,7]
; AVX2-FCP-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; AVX2-FCP-NEXT:    vmovaps 288(%rdi), %ymm0
; AVX2-FCP-NEXT:    vmovaps 288(%rsi), %ymm1
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm1[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm3 = ymm0[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm12 = ymm3[0,1],ymm2[2,3],ymm3[4,5],ymm2[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm10 = ymm0[0,1],ymm1[2,3],ymm0[4,5],ymm1[6,7]
; AVX2-FCP-NEXT:    vmovaps 320(%rdi), %ymm0
; AVX2-FCP-NEXT:    vmovaps 320(%rsi), %ymm1
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm1[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm3 = ymm0[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm8 = ymm3[0,1],ymm2[2,3],ymm3[4,5],ymm2[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm7 = ymm0[0,1],ymm1[2,3],ymm0[4,5],ymm1[6,7]
; AVX2-FCP-NEXT:    vmovaps 352(%rdi), %ymm0
; AVX2-FCP-NEXT:    vmovaps 352(%rsi), %ymm1
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm2 = ymm1[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm9 = ymm0[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm6 = ymm9[0,1],ymm2[2,3],ymm9[4,5],ymm2[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm5 = ymm0[0,1],ymm1[2,3],ymm0[4,5],ymm1[6,7]
; AVX2-FCP-NEXT:    vmovaps 384(%rdi), %ymm9
; AVX2-FCP-NEXT:    vmovaps 384(%rsi), %ymm0
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm1 = ymm0[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm11 = ymm9[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm4 = ymm11[0,1],ymm1[2,3],ymm11[4,5],ymm1[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm9 = ymm9[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm3 = ymm9[0,1],ymm0[2,3],ymm9[4,5],ymm0[6,7]
; AVX2-FCP-NEXT:    vmovaps 416(%rdi), %ymm9
; AVX2-FCP-NEXT:    vmovaps 416(%rsi), %ymm11
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm11[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm13 = ymm9[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm2 = ymm13[0,1],ymm0[2,3],ymm13[4,5],ymm0[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm11 = ymm11[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm9 = ymm9[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm9 = ymm9[0,1],ymm11[2,3],ymm9[4,5],ymm11[6,7]
; AVX2-FCP-NEXT:    vmovaps 448(%rdi), %ymm11
; AVX2-FCP-NEXT:    vmovaps 448(%rsi), %ymm13
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm13[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm15 = ymm11[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm1 = ymm15[0,1],ymm0[2,3],ymm15[4,5],ymm0[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm13 = ymm13[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm11 = ymm11[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm11 = ymm11[0,1],ymm13[2,3],ymm11[4,5],ymm13[6,7]
; AVX2-FCP-NEXT:    vmovaps 480(%rdi), %ymm13
; AVX2-FCP-NEXT:    vmovaps 480(%rsi), %ymm15
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm0 = ymm15[0,0,2,1]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm14 = ymm13[0,1,1,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm0 = ymm14[0,1],ymm0[2,3],ymm14[4,5],ymm0[6,7]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm14 = ymm15[0,2,2,3]
; AVX2-FCP-NEXT:    vpermpd {{.*#+}} ymm13 = ymm13[2,1,3,3]
; AVX2-FCP-NEXT:    vblendps {{.*#+}} ymm13 = ymm13[0,1],ymm14[2,3],ymm13[4,5],ymm14[6,7]
; AVX2-FCP-NEXT:    vmovaps %ymm13, 992(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm0, 960(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm11, 928(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm1, 896(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm9, 864(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm2, 832(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm3, 800(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm4, 768(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm5, 736(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm6, 704(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm7, 672(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm8, 640(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm10, 608(%rdx)
; AVX2-FCP-NEXT:    vmovaps %ymm12, 576(%rdx)
; AVX2-FCP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FCP-NEXT:    vmovaps %ymm0, 544(%rdx)
; AVX2-FCP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FCP-NEXT:    vmovaps %ymm0, 512(%rdx)
; AVX2-FCP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FCP-NEXT:    vmovaps %ymm0, 480(%rdx)
; AVX2-FCP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FCP-NEXT:    vmovaps %ymm0, 448(%rdx)
; AVX2-FCP-NEXT:    vmovups (%rsp), %ymm0 # 32-byte Reload
; AVX2-FCP-NEXT:    vmovaps %ymm0, 416(%rdx)
; AVX2-FCP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FCP-NEXT:    vmovaps %ymm0, 384(%rdx)
; AVX2-FCP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FCP-NEXT:    vmovaps %ymm0, 352(%rdx)
; AVX2-FCP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FCP-NEXT:    vmovaps %ymm0, 320(%rdx)
; AVX2-FCP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FCP-NEXT:    vmovaps %ymm0, 288(%rdx)
; AVX2-FCP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FCP-NEXT:    vmovaps %ymm0, 256(%rdx)
; AVX2-FCP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FCP-NEXT:    vmovaps %ymm0, 224(%rdx)
; AVX2-FCP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FCP-NEXT:    vmovaps %ymm0, 192(%rdx)
; AVX2-FCP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FCP-NEXT:    vmovaps %ymm0, 160(%rdx)
; AVX2-FCP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FCP-NEXT:    vmovaps %ymm0, 128(%rdx)
; AVX2-FCP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FCP-NEXT:    vmovaps %ymm0, 96(%rdx)
; AVX2-FCP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FCP-NEXT:    vmovaps %ymm0, 64(%rdx)
; AVX2-FCP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FCP-NEXT:    vmovaps %ymm0, 32(%rdx)
; AVX2-FCP-NEXT:    vmovups {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Reload
; AVX2-FCP-NEXT:    vmovaps %ymm0, (%rdx)
; AVX2-FCP-NEXT:    addq $456, %rsp # imm = 0x1C8
; AVX2-FCP-NEXT:    vzeroupper
; AVX2-FCP-NEXT:    retq
;
; AVX512-LABEL: store_i64_stride2_vf64:
; AVX512:       # %bb.0:
; AVX512-NEXT:    vmovdqa64 448(%rdi), %zmm0
; AVX512-NEXT:    vmovdqa64 384(%rdi), %zmm1
; AVX512-NEXT:    vmovdqa64 320(%rdi), %zmm2
; AVX512-NEXT:    vmovdqa64 256(%rdi), %zmm3
; AVX512-NEXT:    vmovdqa64 (%rdi), %zmm4
; AVX512-NEXT:    vmovdqa64 64(%rdi), %zmm5
; AVX512-NEXT:    vmovdqa64 128(%rdi), %zmm6
; AVX512-NEXT:    vmovdqa64 192(%rdi), %zmm7
; AVX512-NEXT:    vmovdqa64 448(%rsi), %zmm8
; AVX512-NEXT:    vmovdqa64 384(%rsi), %zmm9
; AVX512-NEXT:    vmovdqa64 320(%rsi), %zmm10
; AVX512-NEXT:    vmovdqa64 256(%rsi), %zmm11
; AVX512-NEXT:    vmovdqa64 (%rsi), %zmm12
; AVX512-NEXT:    vmovdqa64 64(%rsi), %zmm13
; AVX512-NEXT:    vmovdqa64 128(%rsi), %zmm14
; AVX512-NEXT:    vmovdqa64 192(%rsi), %zmm15
; AVX512-NEXT:    vpmovsxbq {{.*#+}} zmm16 = [4,12,5,13,6,14,7,15]
; AVX512-NEXT:    vmovdqa64 %zmm4, %zmm17
; AVX512-NEXT:    vpermt2q %zmm12, %zmm16, %zmm17
; AVX512-NEXT:    vpmovsxbq {{.*#+}} zmm18 = [0,8,1,9,2,10,3,11]
; AVX512-NEXT:    vpermt2q %zmm12, %zmm18, %zmm4
; AVX512-NEXT:    vmovdqa64 %zmm5, %zmm12
; AVX512-NEXT:    vpermt2q %zmm13, %zmm16, %zmm12
; AVX512-NEXT:    vpermt2q %zmm13, %zmm18, %zmm5
; AVX512-NEXT:    vmovdqa64 %zmm6, %zmm13
; AVX512-NEXT:    vpermt2q %zmm14, %zmm16, %zmm13
; AVX512-NEXT:    vpermt2q %zmm14, %zmm18, %zmm6
; AVX512-NEXT:    vmovdqa64 %zmm7, %zmm14
; AVX512-NEXT:    vpermt2q %zmm15, %zmm16, %zmm14
; AVX512-NEXT:    vpermt2q %zmm15, %zmm18, %zmm7
; AVX512-NEXT:    vmovdqa64 %zmm3, %zmm15
; AVX512-NEXT:    vpermt2q %zmm11, %zmm16, %zmm15
; AVX512-NEXT:    vpermt2q %zmm11, %zmm18, %zmm3
; AVX512-NEXT:    vmovdqa64 %zmm2, %zmm11
; AVX512-NEXT:    vpermt2q %zmm10, %zmm16, %zmm11
; AVX512-NEXT:    vpermt2q %zmm10, %zmm18, %zmm2
; AVX512-NEXT:    vmovdqa64 %zmm1, %zmm10
; AVX512-NEXT:    vpermt2q %zmm9, %zmm16, %zmm10
; AVX512-NEXT:    vpermt2q %zmm9, %zmm18, %zmm1
; AVX512-NEXT:    vpermi2q %zmm8, %zmm0, %zmm16
; AVX512-NEXT:    vpermt2q %zmm8, %zmm18, %zmm0
; AVX512-NEXT:    vmovdqa64 %zmm0, 896(%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm16, 960(%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm1, 768(%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm10, 832(%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm2, 640(%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm11, 704(%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm3, 512(%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm15, 576(%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm7, 384(%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm14, 448(%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm6, 256(%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm13, 320(%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm5, 128(%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm12, 192(%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm4, (%rdx)
; AVX512-NEXT:    vmovdqa64 %zmm17, 64(%rdx)
; AVX512-NEXT:    vzeroupper
; AVX512-NEXT:    retq
;
; AVX512-FCP-LABEL: store_i64_stride2_vf64:
; AVX512-FCP:       # %bb.0:
; AVX512-FCP-NEXT:    vmovdqa64 448(%rdi), %zmm0
; AVX512-FCP-NEXT:    vmovdqa64 384(%rdi), %zmm1
; AVX512-FCP-NEXT:    vmovdqa64 320(%rdi), %zmm2
; AVX512-FCP-NEXT:    vmovdqa64 256(%rdi), %zmm3
; AVX512-FCP-NEXT:    vmovdqa64 (%rdi), %zmm4
; AVX512-FCP-NEXT:    vmovdqa64 64(%rdi), %zmm5
; AVX512-FCP-NEXT:    vmovdqa64 128(%rdi), %zmm6
; AVX512-FCP-NEXT:    vmovdqa64 192(%rdi), %zmm7
; AVX512-FCP-NEXT:    vmovdqa64 448(%rsi), %zmm8
; AVX512-FCP-NEXT:    vmovdqa64 384(%rsi), %zmm9
; AVX512-FCP-NEXT:    vmovdqa64 320(%rsi), %zmm10
; AVX512-FCP-NEXT:    vmovdqa64 256(%rsi), %zmm11
; AVX512-FCP-NEXT:    vmovdqa64 (%rsi), %zmm12
; AVX512-FCP-NEXT:    vmovdqa64 64(%rsi), %zmm13
; AVX512-FCP-NEXT:    vmovdqa64 128(%rsi), %zmm14
; AVX512-FCP-NEXT:    vmovdqa64 192(%rsi), %zmm15
; AVX512-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm16 = [4,12,5,13,6,14,7,15]
; AVX512-FCP-NEXT:    vmovdqa64 %zmm4, %zmm17
; AVX512-FCP-NEXT:    vpermt2q %zmm12, %zmm16, %zmm17
; AVX512-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm18 = [0,8,1,9,2,10,3,11]
; AVX512-FCP-NEXT:    vpermt2q %zmm12, %zmm18, %zmm4
; AVX512-FCP-NEXT:    vmovdqa64 %zmm5, %zmm12
; AVX512-FCP-NEXT:    vpermt2q %zmm13, %zmm16, %zmm12
; AVX512-FCP-NEXT:    vpermt2q %zmm13, %zmm18, %zmm5
; AVX512-FCP-NEXT:    vmovdqa64 %zmm6, %zmm13
; AVX512-FCP-NEXT:    vpermt2q %zmm14, %zmm16, %zmm13
; AVX512-FCP-NEXT:    vpermt2q %zmm14, %zmm18, %zmm6
; AVX512-FCP-NEXT:    vmovdqa64 %zmm7, %zmm14
; AVX512-FCP-NEXT:    vpermt2q %zmm15, %zmm16, %zmm14
; AVX512-FCP-NEXT:    vpermt2q %zmm15, %zmm18, %zmm7
; AVX512-FCP-NEXT:    vmovdqa64 %zmm3, %zmm15
; AVX512-FCP-NEXT:    vpermt2q %zmm11, %zmm16, %zmm15
; AVX512-FCP-NEXT:    vpermt2q %zmm11, %zmm18, %zmm3
; AVX512-FCP-NEXT:    vmovdqa64 %zmm2, %zmm11
; AVX512-FCP-NEXT:    vpermt2q %zmm10, %zmm16, %zmm11
; AVX512-FCP-NEXT:    vpermt2q %zmm10, %zmm18, %zmm2
; AVX512-FCP-NEXT:    vmovdqa64 %zmm1, %zmm10
; AVX512-FCP-NEXT:    vpermt2q %zmm9, %zmm16, %zmm10
; AVX512-FCP-NEXT:    vpermt2q %zmm9, %zmm18, %zmm1
; AVX512-FCP-NEXT:    vpermi2q %zmm8, %zmm0, %zmm16
; AVX512-FCP-NEXT:    vpermt2q %zmm8, %zmm18, %zmm0
; AVX512-FCP-NEXT:    vmovdqa64 %zmm0, 896(%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm16, 960(%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm1, 768(%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm10, 832(%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm2, 640(%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm11, 704(%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm3, 512(%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm15, 576(%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm7, 384(%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm14, 448(%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm6, 256(%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm13, 320(%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm5, 128(%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm12, 192(%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm4, (%rdx)
; AVX512-FCP-NEXT:    vmovdqa64 %zmm17, 64(%rdx)
; AVX512-FCP-NEXT:    vzeroupper
; AVX512-FCP-NEXT:    retq
;
; AVX512DQ-LABEL: store_i64_stride2_vf64:
; AVX512DQ:       # %bb.0:
; AVX512DQ-NEXT:    vmovdqa64 448(%rdi), %zmm0
; AVX512DQ-NEXT:    vmovdqa64 384(%rdi), %zmm1
; AVX512DQ-NEXT:    vmovdqa64 320(%rdi), %zmm2
; AVX512DQ-NEXT:    vmovdqa64 256(%rdi), %zmm3
; AVX512DQ-NEXT:    vmovdqa64 (%rdi), %zmm4
; AVX512DQ-NEXT:    vmovdqa64 64(%rdi), %zmm5
; AVX512DQ-NEXT:    vmovdqa64 128(%rdi), %zmm6
; AVX512DQ-NEXT:    vmovdqa64 192(%rdi), %zmm7
; AVX512DQ-NEXT:    vmovdqa64 448(%rsi), %zmm8
; AVX512DQ-NEXT:    vmovdqa64 384(%rsi), %zmm9
; AVX512DQ-NEXT:    vmovdqa64 320(%rsi), %zmm10
; AVX512DQ-NEXT:    vmovdqa64 256(%rsi), %zmm11
; AVX512DQ-NEXT:    vmovdqa64 (%rsi), %zmm12
; AVX512DQ-NEXT:    vmovdqa64 64(%rsi), %zmm13
; AVX512DQ-NEXT:    vmovdqa64 128(%rsi), %zmm14
; AVX512DQ-NEXT:    vmovdqa64 192(%rsi), %zmm15
; AVX512DQ-NEXT:    vpmovsxbq {{.*#+}} zmm16 = [4,12,5,13,6,14,7,15]
; AVX512DQ-NEXT:    vmovdqa64 %zmm4, %zmm17
; AVX512DQ-NEXT:    vpermt2q %zmm12, %zmm16, %zmm17
; AVX512DQ-NEXT:    vpmovsxbq {{.*#+}} zmm18 = [0,8,1,9,2,10,3,11]
; AVX512DQ-NEXT:    vpermt2q %zmm12, %zmm18, %zmm4
; AVX512DQ-NEXT:    vmovdqa64 %zmm5, %zmm12
; AVX512DQ-NEXT:    vpermt2q %zmm13, %zmm16, %zmm12
; AVX512DQ-NEXT:    vpermt2q %zmm13, %zmm18, %zmm5
; AVX512DQ-NEXT:    vmovdqa64 %zmm6, %zmm13
; AVX512DQ-NEXT:    vpermt2q %zmm14, %zmm16, %zmm13
; AVX512DQ-NEXT:    vpermt2q %zmm14, %zmm18, %zmm6
; AVX512DQ-NEXT:    vmovdqa64 %zmm7, %zmm14
; AVX512DQ-NEXT:    vpermt2q %zmm15, %zmm16, %zmm14
; AVX512DQ-NEXT:    vpermt2q %zmm15, %zmm18, %zmm7
; AVX512DQ-NEXT:    vmovdqa64 %zmm3, %zmm15
; AVX512DQ-NEXT:    vpermt2q %zmm11, %zmm16, %zmm15
; AVX512DQ-NEXT:    vpermt2q %zmm11, %zmm18, %zmm3
; AVX512DQ-NEXT:    vmovdqa64 %zmm2, %zmm11
; AVX512DQ-NEXT:    vpermt2q %zmm10, %zmm16, %zmm11
; AVX512DQ-NEXT:    vpermt2q %zmm10, %zmm18, %zmm2
; AVX512DQ-NEXT:    vmovdqa64 %zmm1, %zmm10
; AVX512DQ-NEXT:    vpermt2q %zmm9, %zmm16, %zmm10
; AVX512DQ-NEXT:    vpermt2q %zmm9, %zmm18, %zmm1
; AVX512DQ-NEXT:    vpermi2q %zmm8, %zmm0, %zmm16
; AVX512DQ-NEXT:    vpermt2q %zmm8, %zmm18, %zmm0
; AVX512DQ-NEXT:    vmovdqa64 %zmm0, 896(%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm16, 960(%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm1, 768(%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm10, 832(%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm2, 640(%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm11, 704(%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm3, 512(%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm15, 576(%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm7, 384(%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm14, 448(%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm6, 256(%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm13, 320(%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm5, 128(%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm12, 192(%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm4, (%rdx)
; AVX512DQ-NEXT:    vmovdqa64 %zmm17, 64(%rdx)
; AVX512DQ-NEXT:    vzeroupper
; AVX512DQ-NEXT:    retq
;
; AVX512DQ-FCP-LABEL: store_i64_stride2_vf64:
; AVX512DQ-FCP:       # %bb.0:
; AVX512DQ-FCP-NEXT:    vmovdqa64 448(%rdi), %zmm0
; AVX512DQ-FCP-NEXT:    vmovdqa64 384(%rdi), %zmm1
; AVX512DQ-FCP-NEXT:    vmovdqa64 320(%rdi), %zmm2
; AVX512DQ-FCP-NEXT:    vmovdqa64 256(%rdi), %zmm3
; AVX512DQ-FCP-NEXT:    vmovdqa64 (%rdi), %zmm4
; AVX512DQ-FCP-NEXT:    vmovdqa64 64(%rdi), %zmm5
; AVX512DQ-FCP-NEXT:    vmovdqa64 128(%rdi), %zmm6
; AVX512DQ-FCP-NEXT:    vmovdqa64 192(%rdi), %zmm7
; AVX512DQ-FCP-NEXT:    vmovdqa64 448(%rsi), %zmm8
; AVX512DQ-FCP-NEXT:    vmovdqa64 384(%rsi), %zmm9
; AVX512DQ-FCP-NEXT:    vmovdqa64 320(%rsi), %zmm10
; AVX512DQ-FCP-NEXT:    vmovdqa64 256(%rsi), %zmm11
; AVX512DQ-FCP-NEXT:    vmovdqa64 (%rsi), %zmm12
; AVX512DQ-FCP-NEXT:    vmovdqa64 64(%rsi), %zmm13
; AVX512DQ-FCP-NEXT:    vmovdqa64 128(%rsi), %zmm14
; AVX512DQ-FCP-NEXT:    vmovdqa64 192(%rsi), %zmm15
; AVX512DQ-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm16 = [4,12,5,13,6,14,7,15]
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm4, %zmm17
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm12, %zmm16, %zmm17
; AVX512DQ-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm18 = [0,8,1,9,2,10,3,11]
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm12, %zmm18, %zmm4
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm5, %zmm12
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm13, %zmm16, %zmm12
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm13, %zmm18, %zmm5
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm6, %zmm13
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm14, %zmm16, %zmm13
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm14, %zmm18, %zmm6
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm7, %zmm14
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm15, %zmm16, %zmm14
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm15, %zmm18, %zmm7
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm3, %zmm15
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm11, %zmm16, %zmm15
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm11, %zmm18, %zmm3
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm2, %zmm11
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm10, %zmm16, %zmm11
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm10, %zmm18, %zmm2
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm1, %zmm10
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm9, %zmm16, %zmm10
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm9, %zmm18, %zmm1
; AVX512DQ-FCP-NEXT:    vpermi2q %zmm8, %zmm0, %zmm16
; AVX512DQ-FCP-NEXT:    vpermt2q %zmm8, %zmm18, %zmm0
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm0, 896(%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm16, 960(%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm1, 768(%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm10, 832(%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm2, 640(%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm11, 704(%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm3, 512(%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm15, 576(%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm7, 384(%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm14, 448(%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm6, 256(%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm13, 320(%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm5, 128(%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm12, 192(%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm4, (%rdx)
; AVX512DQ-FCP-NEXT:    vmovdqa64 %zmm17, 64(%rdx)
; AVX512DQ-FCP-NEXT:    vzeroupper
; AVX512DQ-FCP-NEXT:    retq
;
; AVX512BW-LABEL: store_i64_stride2_vf64:
; AVX512BW:       # %bb.0:
; AVX512BW-NEXT:    vmovdqa64 448(%rdi), %zmm0
; AVX512BW-NEXT:    vmovdqa64 384(%rdi), %zmm1
; AVX512BW-NEXT:    vmovdqa64 320(%rdi), %zmm2
; AVX512BW-NEXT:    vmovdqa64 256(%rdi), %zmm3
; AVX512BW-NEXT:    vmovdqa64 (%rdi), %zmm4
; AVX512BW-NEXT:    vmovdqa64 64(%rdi), %zmm5
; AVX512BW-NEXT:    vmovdqa64 128(%rdi), %zmm6
; AVX512BW-NEXT:    vmovdqa64 192(%rdi), %zmm7
; AVX512BW-NEXT:    vmovdqa64 448(%rsi), %zmm8
; AVX512BW-NEXT:    vmovdqa64 384(%rsi), %zmm9
; AVX512BW-NEXT:    vmovdqa64 320(%rsi), %zmm10
; AVX512BW-NEXT:    vmovdqa64 256(%rsi), %zmm11
; AVX512BW-NEXT:    vmovdqa64 (%rsi), %zmm12
; AVX512BW-NEXT:    vmovdqa64 64(%rsi), %zmm13
; AVX512BW-NEXT:    vmovdqa64 128(%rsi), %zmm14
; AVX512BW-NEXT:    vmovdqa64 192(%rsi), %zmm15
; AVX512BW-NEXT:    vpmovsxbq {{.*#+}} zmm16 = [4,12,5,13,6,14,7,15]
; AVX512BW-NEXT:    vmovdqa64 %zmm4, %zmm17
; AVX512BW-NEXT:    vpermt2q %zmm12, %zmm16, %zmm17
; AVX512BW-NEXT:    vpmovsxbq {{.*#+}} zmm18 = [0,8,1,9,2,10,3,11]
; AVX512BW-NEXT:    vpermt2q %zmm12, %zmm18, %zmm4
; AVX512BW-NEXT:    vmovdqa64 %zmm5, %zmm12
; AVX512BW-NEXT:    vpermt2q %zmm13, %zmm16, %zmm12
; AVX512BW-NEXT:    vpermt2q %zmm13, %zmm18, %zmm5
; AVX512BW-NEXT:    vmovdqa64 %zmm6, %zmm13
; AVX512BW-NEXT:    vpermt2q %zmm14, %zmm16, %zmm13
; AVX512BW-NEXT:    vpermt2q %zmm14, %zmm18, %zmm6
; AVX512BW-NEXT:    vmovdqa64 %zmm7, %zmm14
; AVX512BW-NEXT:    vpermt2q %zmm15, %zmm16, %zmm14
; AVX512BW-NEXT:    vpermt2q %zmm15, %zmm18, %zmm7
; AVX512BW-NEXT:    vmovdqa64 %zmm3, %zmm15
; AVX512BW-NEXT:    vpermt2q %zmm11, %zmm16, %zmm15
; AVX512BW-NEXT:    vpermt2q %zmm11, %zmm18, %zmm3
; AVX512BW-NEXT:    vmovdqa64 %zmm2, %zmm11
; AVX512BW-NEXT:    vpermt2q %zmm10, %zmm16, %zmm11
; AVX512BW-NEXT:    vpermt2q %zmm10, %zmm18, %zmm2
; AVX512BW-NEXT:    vmovdqa64 %zmm1, %zmm10
; AVX512BW-NEXT:    vpermt2q %zmm9, %zmm16, %zmm10
; AVX512BW-NEXT:    vpermt2q %zmm9, %zmm18, %zmm1
; AVX512BW-NEXT:    vpermi2q %zmm8, %zmm0, %zmm16
; AVX512BW-NEXT:    vpermt2q %zmm8, %zmm18, %zmm0
; AVX512BW-NEXT:    vmovdqa64 %zmm0, 896(%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm16, 960(%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm1, 768(%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm10, 832(%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm2, 640(%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm11, 704(%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm3, 512(%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm15, 576(%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm7, 384(%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm14, 448(%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm6, 256(%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm13, 320(%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm5, 128(%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm12, 192(%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm4, (%rdx)
; AVX512BW-NEXT:    vmovdqa64 %zmm17, 64(%rdx)
; AVX512BW-NEXT:    vzeroupper
; AVX512BW-NEXT:    retq
;
; AVX512BW-FCP-LABEL: store_i64_stride2_vf64:
; AVX512BW-FCP:       # %bb.0:
; AVX512BW-FCP-NEXT:    vmovdqa64 448(%rdi), %zmm0
; AVX512BW-FCP-NEXT:    vmovdqa64 384(%rdi), %zmm1
; AVX512BW-FCP-NEXT:    vmovdqa64 320(%rdi), %zmm2
; AVX512BW-FCP-NEXT:    vmovdqa64 256(%rdi), %zmm3
; AVX512BW-FCP-NEXT:    vmovdqa64 (%rdi), %zmm4
; AVX512BW-FCP-NEXT:    vmovdqa64 64(%rdi), %zmm5
; AVX512BW-FCP-NEXT:    vmovdqa64 128(%rdi), %zmm6
; AVX512BW-FCP-NEXT:    vmovdqa64 192(%rdi), %zmm7
; AVX512BW-FCP-NEXT:    vmovdqa64 448(%rsi), %zmm8
; AVX512BW-FCP-NEXT:    vmovdqa64 384(%rsi), %zmm9
; AVX512BW-FCP-NEXT:    vmovdqa64 320(%rsi), %zmm10
; AVX512BW-FCP-NEXT:    vmovdqa64 256(%rsi), %zmm11
; AVX512BW-FCP-NEXT:    vmovdqa64 (%rsi), %zmm12
; AVX512BW-FCP-NEXT:    vmovdqa64 64(%rsi), %zmm13
; AVX512BW-FCP-NEXT:    vmovdqa64 128(%rsi), %zmm14
; AVX512BW-FCP-NEXT:    vmovdqa64 192(%rsi), %zmm15
; AVX512BW-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm16 = [4,12,5,13,6,14,7,15]
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm4, %zmm17
; AVX512BW-FCP-NEXT:    vpermt2q %zmm12, %zmm16, %zmm17
; AVX512BW-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm18 = [0,8,1,9,2,10,3,11]
; AVX512BW-FCP-NEXT:    vpermt2q %zmm12, %zmm18, %zmm4
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm5, %zmm12
; AVX512BW-FCP-NEXT:    vpermt2q %zmm13, %zmm16, %zmm12
; AVX512BW-FCP-NEXT:    vpermt2q %zmm13, %zmm18, %zmm5
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm6, %zmm13
; AVX512BW-FCP-NEXT:    vpermt2q %zmm14, %zmm16, %zmm13
; AVX512BW-FCP-NEXT:    vpermt2q %zmm14, %zmm18, %zmm6
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm7, %zmm14
; AVX512BW-FCP-NEXT:    vpermt2q %zmm15, %zmm16, %zmm14
; AVX512BW-FCP-NEXT:    vpermt2q %zmm15, %zmm18, %zmm7
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm3, %zmm15
; AVX512BW-FCP-NEXT:    vpermt2q %zmm11, %zmm16, %zmm15
; AVX512BW-FCP-NEXT:    vpermt2q %zmm11, %zmm18, %zmm3
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm2, %zmm11
; AVX512BW-FCP-NEXT:    vpermt2q %zmm10, %zmm16, %zmm11
; AVX512BW-FCP-NEXT:    vpermt2q %zmm10, %zmm18, %zmm2
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm1, %zmm10
; AVX512BW-FCP-NEXT:    vpermt2q %zmm9, %zmm16, %zmm10
; AVX512BW-FCP-NEXT:    vpermt2q %zmm9, %zmm18, %zmm1
; AVX512BW-FCP-NEXT:    vpermi2q %zmm8, %zmm0, %zmm16
; AVX512BW-FCP-NEXT:    vpermt2q %zmm8, %zmm18, %zmm0
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm0, 896(%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm16, 960(%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm1, 768(%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm10, 832(%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm2, 640(%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm11, 704(%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm3, 512(%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm15, 576(%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm7, 384(%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm14, 448(%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm6, 256(%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm13, 320(%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm5, 128(%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm12, 192(%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm4, (%rdx)
; AVX512BW-FCP-NEXT:    vmovdqa64 %zmm17, 64(%rdx)
; AVX512BW-FCP-NEXT:    vzeroupper
; AVX512BW-FCP-NEXT:    retq
;
; AVX512DQ-BW-LABEL: store_i64_stride2_vf64:
; AVX512DQ-BW:       # %bb.0:
; AVX512DQ-BW-NEXT:    vmovdqa64 448(%rdi), %zmm0
; AVX512DQ-BW-NEXT:    vmovdqa64 384(%rdi), %zmm1
; AVX512DQ-BW-NEXT:    vmovdqa64 320(%rdi), %zmm2
; AVX512DQ-BW-NEXT:    vmovdqa64 256(%rdi), %zmm3
; AVX512DQ-BW-NEXT:    vmovdqa64 (%rdi), %zmm4
; AVX512DQ-BW-NEXT:    vmovdqa64 64(%rdi), %zmm5
; AVX512DQ-BW-NEXT:    vmovdqa64 128(%rdi), %zmm6
; AVX512DQ-BW-NEXT:    vmovdqa64 192(%rdi), %zmm7
; AVX512DQ-BW-NEXT:    vmovdqa64 448(%rsi), %zmm8
; AVX512DQ-BW-NEXT:    vmovdqa64 384(%rsi), %zmm9
; AVX512DQ-BW-NEXT:    vmovdqa64 320(%rsi), %zmm10
; AVX512DQ-BW-NEXT:    vmovdqa64 256(%rsi), %zmm11
; AVX512DQ-BW-NEXT:    vmovdqa64 (%rsi), %zmm12
; AVX512DQ-BW-NEXT:    vmovdqa64 64(%rsi), %zmm13
; AVX512DQ-BW-NEXT:    vmovdqa64 128(%rsi), %zmm14
; AVX512DQ-BW-NEXT:    vmovdqa64 192(%rsi), %zmm15
; AVX512DQ-BW-NEXT:    vpmovsxbq {{.*#+}} zmm16 = [4,12,5,13,6,14,7,15]
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm4, %zmm17
; AVX512DQ-BW-NEXT:    vpermt2q %zmm12, %zmm16, %zmm17
; AVX512DQ-BW-NEXT:    vpmovsxbq {{.*#+}} zmm18 = [0,8,1,9,2,10,3,11]
; AVX512DQ-BW-NEXT:    vpermt2q %zmm12, %zmm18, %zmm4
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm5, %zmm12
; AVX512DQ-BW-NEXT:    vpermt2q %zmm13, %zmm16, %zmm12
; AVX512DQ-BW-NEXT:    vpermt2q %zmm13, %zmm18, %zmm5
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm6, %zmm13
; AVX512DQ-BW-NEXT:    vpermt2q %zmm14, %zmm16, %zmm13
; AVX512DQ-BW-NEXT:    vpermt2q %zmm14, %zmm18, %zmm6
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm7, %zmm14
; AVX512DQ-BW-NEXT:    vpermt2q %zmm15, %zmm16, %zmm14
; AVX512DQ-BW-NEXT:    vpermt2q %zmm15, %zmm18, %zmm7
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm3, %zmm15
; AVX512DQ-BW-NEXT:    vpermt2q %zmm11, %zmm16, %zmm15
; AVX512DQ-BW-NEXT:    vpermt2q %zmm11, %zmm18, %zmm3
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm2, %zmm11
; AVX512DQ-BW-NEXT:    vpermt2q %zmm10, %zmm16, %zmm11
; AVX512DQ-BW-NEXT:    vpermt2q %zmm10, %zmm18, %zmm2
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm1, %zmm10
; AVX512DQ-BW-NEXT:    vpermt2q %zmm9, %zmm16, %zmm10
; AVX512DQ-BW-NEXT:    vpermt2q %zmm9, %zmm18, %zmm1
; AVX512DQ-BW-NEXT:    vpermi2q %zmm8, %zmm0, %zmm16
; AVX512DQ-BW-NEXT:    vpermt2q %zmm8, %zmm18, %zmm0
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm0, 896(%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm16, 960(%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm1, 768(%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm10, 832(%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm2, 640(%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm11, 704(%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm3, 512(%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm15, 576(%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm7, 384(%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm14, 448(%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm6, 256(%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm13, 320(%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm5, 128(%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm12, 192(%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm4, (%rdx)
; AVX512DQ-BW-NEXT:    vmovdqa64 %zmm17, 64(%rdx)
; AVX512DQ-BW-NEXT:    vzeroupper
; AVX512DQ-BW-NEXT:    retq
;
; AVX512DQ-BW-FCP-LABEL: store_i64_stride2_vf64:
; AVX512DQ-BW-FCP:       # %bb.0:
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 448(%rdi), %zmm0
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 384(%rdi), %zmm1
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 320(%rdi), %zmm2
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 256(%rdi), %zmm3
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 (%rdi), %zmm4
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 64(%rdi), %zmm5
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 128(%rdi), %zmm6
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 192(%rdi), %zmm7
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 448(%rsi), %zmm8
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 384(%rsi), %zmm9
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 320(%rsi), %zmm10
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 256(%rsi), %zmm11
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 (%rsi), %zmm12
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 64(%rsi), %zmm13
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 128(%rsi), %zmm14
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 192(%rsi), %zmm15
; AVX512DQ-BW-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm16 = [4,12,5,13,6,14,7,15]
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm4, %zmm17
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm12, %zmm16, %zmm17
; AVX512DQ-BW-FCP-NEXT:    vpmovsxbq {{.*#+}} zmm18 = [0,8,1,9,2,10,3,11]
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm12, %zmm18, %zmm4
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm5, %zmm12
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm13, %zmm16, %zmm12
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm13, %zmm18, %zmm5
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm6, %zmm13
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm14, %zmm16, %zmm13
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm14, %zmm18, %zmm6
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm7, %zmm14
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm15, %zmm16, %zmm14
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm15, %zmm18, %zmm7
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm3, %zmm15
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm11, %zmm16, %zmm15
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm11, %zmm18, %zmm3
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm2, %zmm11
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm10, %zmm16, %zmm11
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm10, %zmm18, %zmm2
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm1, %zmm10
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm9, %zmm16, %zmm10
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm9, %zmm18, %zmm1
; AVX512DQ-BW-FCP-NEXT:    vpermi2q %zmm8, %zmm0, %zmm16
; AVX512DQ-BW-FCP-NEXT:    vpermt2q %zmm8, %zmm18, %zmm0
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm0, 896(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm16, 960(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm1, 768(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm10, 832(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm2, 640(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm11, 704(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm3, 512(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm15, 576(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm7, 384(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm14, 448(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm6, 256(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm13, 320(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm5, 128(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm12, 192(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm4, (%rdx)
; AVX512DQ-BW-FCP-NEXT:    vmovdqa64 %zmm17, 64(%rdx)
; AVX512DQ-BW-FCP-NEXT:    vzeroupper
; AVX512DQ-BW-FCP-NEXT:    retq
  %in.vec0 = load <64 x i64>, ptr %in.vecptr0, align 64
  %in.vec1 = load <64 x i64>, ptr %in.vecptr1, align 64
  %1 = shufflevector <64 x i64> %in.vec0, <64 x i64> %in.vec1, <128 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31, i32 32, i32 33, i32 34, i32 35, i32 36, i32 37, i32 38, i32 39, i32 40, i32 41, i32 42, i32 43, i32 44, i32 45, i32 46, i32 47, i32 48, i32 49, i32 50, i32 51, i32 52, i32 53, i32 54, i32 55, i32 56, i32 57, i32 58, i32 59, i32 60, i32 61, i32 62, i32 63, i32 64, i32 65, i32 66, i32 67, i32 68, i32 69, i32 70, i32 71, i32 72, i32 73, i32 74, i32 75, i32 76, i32 77, i32 78, i32 79, i32 80, i32 81, i32 82, i32 83, i32 84, i32 85, i32 86, i32 87, i32 88, i32 89, i32 90, i32 91, i32 92, i32 93, i32 94, i32 95, i32 96, i32 97, i32 98, i32 99, i32 100, i32 101, i32 102, i32 103, i32 104, i32 105, i32 106, i32 107, i32 108, i32 109, i32 110, i32 111, i32 112, i32 113, i32 114, i32 115, i32 116, i32 117, i32 118, i32 119, i32 120, i32 121, i32 122, i32 123, i32 124, i32 125, i32 126, i32 127>
  %interleaved.vec = shufflevector <128 x i64> %1, <128 x i64> poison, <128 x i32> <i32 0, i32 64, i32 1, i32 65, i32 2, i32 66, i32 3, i32 67, i32 4, i32 68, i32 5, i32 69, i32 6, i32 70, i32 7, i32 71, i32 8, i32 72, i32 9, i32 73, i32 10, i32 74, i32 11, i32 75, i32 12, i32 76, i32 13, i32 77, i32 14, i32 78, i32 15, i32 79, i32 16, i32 80, i32 17, i32 81, i32 18, i32 82, i32 19, i32 83, i32 20, i32 84, i32 21, i32 85, i32 22, i32 86, i32 23, i32 87, i32 24, i32 88, i32 25, i32 89, i32 26, i32 90, i32 27, i32 91, i32 28, i32 92, i32 29, i32 93, i32 30, i32 94, i32 31, i32 95, i32 32, i32 96, i32 33, i32 97, i32 34, i32 98, i32 35, i32 99, i32 36, i32 100, i32 37, i32 101, i32 38, i32 102, i32 39, i32 103, i32 40, i32 104, i32 41, i32 105, i32 42, i32 106, i32 43, i32 107, i32 44, i32 108, i32 45, i32 109, i32 46, i32 110, i32 47, i32 111, i32 48, i32 112, i32 49, i32 113, i32 50, i32 114, i32 51, i32 115, i32 52, i32 116, i32 53, i32 117, i32 54, i32 118, i32 55, i32 119, i32 56, i32 120, i32 57, i32 121, i32 58, i32 122, i32 59, i32 123, i32 60, i32 124, i32 61, i32 125, i32 62, i32 126, i32 63, i32 127>
  store <128 x i64> %interleaved.vec, ptr %out.vec, align 64
  ret void
}