test/CodeGen/X86/stack-folding-int-avx2.ll

; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -O3 -disable-peephole -mtriple=x86_64-unknown-unknown -mattr=+avx2 < %s | FileCheck %s

target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
target triple = "x86_64-unknown-unknown"

; Stack reload folding tests.
;
; By including a nop call with sideeffects we can force a partial register spill of the
; relevant registers and check that the reload is correctly folded into the instruction.

define <4 x double> @stack_fold_broadcastsd_ymm(<2 x double> %a0) {
; CHECK-LABEL: stack_fold_broadcastsd_ymm:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vbroadcastsd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    vmovsd {{.*#+}} xmm1 = mem[0],zero
; CHECK-NEXT:    vaddpd %ymm1, %ymm0, %ymm0
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <2 x double> %a0, <2 x double> undef, <4 x i32> zeroinitializer
  ; fadd forces execution domain
  %3 = fadd <4 x double> %2, <double 0x1, double 0x0, double 0x0, double 0x0>
  ret <4 x double> %3
}

define <4 x float> @stack_fold_broadcastss(<4 x float> %a0) {
; CHECK-LABEL: stack_fold_broadcastss:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vbroadcastss {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    vmovss {{.*#+}} xmm1 = mem[0],zero,zero,zero
; CHECK-NEXT:    vaddps %xmm1, %xmm0, %xmm0
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <4 x float> %a0, <4 x float> undef, <4 x i32> zeroinitializer
  ; fadd forces execution domain
  %3 = fadd <4 x float> %2, <float 1.0, float 0x0, float 0x0, float 0x0>
  ret <4 x float> %3
}

define <8 x float> @stack_fold_broadcastss_ymm(<4 x float> %a0) {
; CHECK-LABEL: stack_fold_broadcastss_ymm:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vbroadcastss {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    vmovss {{.*#+}} xmm1 = mem[0],zero,zero,zero
; CHECK-NEXT:    vaddps %ymm1, %ymm0, %ymm0
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <4 x float> %a0, <4 x float> undef, <8 x i32> zeroinitializer
  ; fadd forces execution domain
  %3 = fadd <8 x float> %2, <float 1.0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0>
  ret <8 x float> %3
}

define <4 x i32> @stack_fold_extracti128(<8 x i16> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_extracti128:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vpmovzxwd {{.*#+}} ymm0 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero
; CHECK-NEXT:    vextracti128 $1, %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Folded Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vmovaps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; CHECK-NEXT:    vzeroupper
; CHECK-NEXT:    retq
  ; zext forces execution domain
  %t1 = zext <8 x i16> %a0 to <8 x i32>
  %t2 = shufflevector <8 x i32> %t1, <8 x i32> %a1, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
  %t3 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  ret <4 x i32> %t2
}

define <8 x i32> @stack_fold_inserti128(<4 x i32> %a0, <4 x i32> %a1) {
; CHECK-LABEL: stack_fold_inserti128:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    # kill: def $xmm0 killed $xmm0 def $ymm0
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vinserti128 $1, {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    vpaddd {{.*}}(%rip), %ymm0, %ymm0
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <4 x i32> %a0, <4 x i32> %a1, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
  ; add forces execution domain
  %3 = add <8 x i32> %2, <i32 2, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
  ret <8 x i32> %3
}

define <16 x i16> @stack_fold_mpsadbw(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_mpsadbw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vmpsadbw $7, {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <16 x i16> @llvm.x86.avx2.mpsadbw(<32 x i8> %a0, <32 x i8> %a1, i8 7)
  ret <16 x i16> %2
}
declare <16 x i16> @llvm.x86.avx2.mpsadbw(<32 x i8>, <32 x i8>, i8) nounwind readnone

define <32 x i8> @stack_fold_pabsb(<32 x i8> %a0) {
; CHECK-LABEL: stack_fold_pabsb:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpabsb {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = icmp sgt <32 x i8> %a0, zeroinitializer
  %3 = sub <32 x i8> zeroinitializer, %a0
  %4 = select <32 x i1> %2, <32 x i8> %a0, <32 x i8> %3
  ret <32 x i8> %4
}

define <8 x i32> @stack_fold_pabsd(<8 x i32> %a0) {
; CHECK-LABEL: stack_fold_pabsd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpabsd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = icmp sgt <8 x i32> %a0, zeroinitializer
  %3 = sub <8 x i32> zeroinitializer, %a0
  %4 = select <8 x i1> %2, <8 x i32> %a0, <8 x i32> %3
  ret <8 x i32> %4
}

define <16 x i16> @stack_fold_pabsw(<16 x i16> %a0) {
; CHECK-LABEL: stack_fold_pabsw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpabsw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = icmp sgt <16 x i16> %a0, zeroinitializer
  %3 = sub <16 x i16> zeroinitializer, %a0
  %4 = select <16 x i1> %2, <16 x i16> %a0, <16 x i16> %3
  ret <16 x i16> %4
}

define <16 x i16> @stack_fold_packssdw(<8 x i32> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_packssdw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpackssdw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <16 x i16> @llvm.x86.avx2.packssdw(<8 x i32> %a0, <8 x i32> %a1)
  ret <16 x i16> %2
}
declare <16 x i16> @llvm.x86.avx2.packssdw(<8 x i32>, <8 x i32>) nounwind readnone

define <32 x i8> @stack_fold_packsswb(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_packsswb:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpacksswb {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <32 x i8> @llvm.x86.avx2.packsswb(<16 x i16> %a0, <16 x i16> %a1)
  ret <32 x i8> %2
}
declare <32 x i8> @llvm.x86.avx2.packsswb(<16 x i16>, <16 x i16>) nounwind readnone

define <16 x i16> @stack_fold_packusdw(<8 x i32> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_packusdw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpackusdw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <16 x i16> @llvm.x86.avx2.packusdw(<8 x i32> %a0, <8 x i32> %a1)
  ret <16 x i16> %2
}
declare <16 x i16> @llvm.x86.avx2.packusdw(<8 x i32>, <8 x i32>) nounwind readnone

define <32 x i8> @stack_fold_packuswb(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_packuswb:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpackuswb {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <32 x i8> @llvm.x86.avx2.packuswb(<16 x i16> %a0, <16 x i16> %a1)
  ret <32 x i8> %2
}
declare <32 x i8> @llvm.x86.avx2.packuswb(<16 x i16>, <16 x i16>) nounwind readnone

define <32 x i8> @stack_fold_paddb(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_paddb:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpaddb {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = add <32 x i8> %a0, %a1
  ret <32 x i8> %2
}

define <8 x i32> @stack_fold_paddd(<8 x i32> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_paddd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpaddd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = add <8 x i32> %a0, %a1
  ret <8 x i32> %2
}

define <4 x i64> @stack_fold_paddq(<4 x i64> %a0, <4 x i64> %a1) {
; CHECK-LABEL: stack_fold_paddq:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpaddq {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = add <4 x i64> %a0, %a1
  ret <4 x i64> %2
}

define <32 x i8> @stack_fold_paddsb(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_paddsb:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpaddsb {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> %a0, <32 x i8> %a1)
  ret <32 x i8> %2
}
declare <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8>, <32 x i8>) nounwind readnone

define <16 x i16> @stack_fold_paddsw(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_paddsw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpaddsw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> %a0, <16 x i16> %a1)
  ret <16 x i16> %2
}
declare <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16>, <16 x i16>) nounwind readnone

define <32 x i8> @stack_fold_paddusb(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_paddusb:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpaddusb {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> %a0, <32 x i8> %a1)
  ret <32 x i8> %2
}
declare <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8>, <32 x i8>) nounwind readnone

define <16 x i16> @stack_fold_paddusw(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_paddusw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpaddusw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> %a0, <16 x i16> %a1)
  ret <16 x i16> %2
}
declare <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16>, <16 x i16>) nounwind readnone

define <16 x i16> @stack_fold_paddw(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_paddw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpaddw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = add <16 x i16> %a0, %a1
  ret <16 x i16> %2
}

define <32 x i8> @stack_fold_palignr(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_palignr:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpalignr $1, {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = mem[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15],ymm0[0],mem[17,18,19,20,21,22,23,24,25,26,27,28,29,30,31],ymm0[16]
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <32 x i8> %a1, <32 x i8> %a0, <32 x i32> <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 32, 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 48>
  ret <32 x i8> %2
}

define <32 x i8> @stack_fold_pand(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_pand:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpand {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    vpaddb {{.*}}(%rip), %ymm0, %ymm0
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = and <32 x i8> %a0, %a1
  ; add forces execution domain
  %3 = add <32 x i8> %2, <i8 2, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1>
  ret <32 x i8> %3
}

define <32 x i8> @stack_fold_pandn(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_pandn:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpandn {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    vpaddb {{.*}}(%rip), %ymm0, %ymm0
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = xor <32 x i8> %a0, <i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1>
  %3 = and <32 x i8> %2, %a1
  ; add forces execution domain
  %4 = add <32 x i8> %3, <i8 2, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1>
  ret <32 x i8> %4
}

define <32 x i8> @stack_fold_pavgb(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_pavgb:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpavgb {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = zext <32 x i8> %a0 to <32 x i16>
  %3 = zext <32 x i8> %a1 to <32 x i16>
  %4 = add <32 x i16> %2, %3
  %5 = add <32 x i16> %4, <i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1>
  %6 = lshr <32 x i16> %5, <i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1>
  %7 = trunc <32 x i16> %6 to <32 x i8>
  ret <32 x i8> %7
}

define <16 x i16> @stack_fold_pavgw(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_pavgw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpavgw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = zext <16 x i16> %a0 to <16 x i32>
  %3 = zext <16 x i16> %a1 to <16 x i32>
  %4 = add <16 x i32> %2, %3
  %5 = add <16 x i32> %4, <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
  %6 = lshr <16 x i32> %5, <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
  %7 = trunc <16 x i32> %6 to <16 x i16>
  ret <16 x i16> %7
}

define <4 x i32> @stack_fold_pblendd(<4 x i32> %a0, <4 x i32> %a1) {
; CHECK-LABEL: stack_fold_pblendd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpblendd $7, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    # xmm0 = mem[0,1,2],xmm0[3]
; CHECK-NEXT:    vpcmpeqd %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vpsubd %xmm1, %xmm0, %xmm0
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <4 x i32> %a0, <4 x i32> %a1, <4 x i32> <i32 4, i32 5, i32 6, i32 3>
  ; add forces execution domain
  %3 = add <4 x i32> %2, <i32 1, i32 1, i32 1, i32 1>
  ret <4 x i32> %3
}

define <8 x i32> @stack_fold_pblendd_ymm(<8 x i32> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_pblendd_ymm:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpblendd $7, {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = mem[0,1,2],ymm0[3,4,5,6,7]
; CHECK-NEXT:    vpaddd {{.*}}(%rip), %ymm0, %ymm0
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <8 x i32> %a0, <8 x i32> %a1, <8 x i32> <i32 8, i32 9, i32 10, i32 3, i32 4, i32 5, i32 6, i32 7>
  ; add forces execution domain
  %3 = add <8 x i32> %2, <i32 2, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
  ret <8 x i32> %3
}

define <32 x i8> @stack_fold_pblendvb(<32 x i8> %a0, <32 x i8> %a1, <32 x i8> %c) {
; CHECK-LABEL: stack_fold_pblendvb:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm2, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpblendvb %ymm0, {{[-0-9]+}}(%r{{[sb]}}p), %ymm1, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <32 x i8> @llvm.x86.avx2.pblendvb(<32 x i8> %a1, <32 x i8> %c, <32 x i8> %a0)
  ret <32 x i8> %2
}
declare <32 x i8> @llvm.x86.avx2.pblendvb(<32 x i8>, <32 x i8>, <32 x i8>) nounwind readnone

define <16 x i16> @stack_fold_pblendw(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_pblendw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpblendw $7, {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = mem[0,1,2],ymm0[3,4,5,6,7],mem[8,9,10],ymm0[11,12,13,14,15]
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <16 x i16> %a0, <16 x i16> %a1, <16 x i32> <i32 16, i32 17, i32 18, i32 3, i32 4, i32 5, i32 6, i32 7, i32 24, i32 25, i32 26, i32 11, i32 12, i32 13, i32 14, i32 15>
  ret <16 x i16> %2
}

define <16 x i8> @stack_fold_pbroadcastb(<16 x i8> %a0) {
; CHECK-LABEL: stack_fold_pbroadcastb:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpbroadcastb {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <16 x i8> %a0, <16 x i8> undef, <16 x i32> zeroinitializer
  ret <16 x i8> %2
}

define <32 x i8> @stack_fold_pbroadcastb_ymm(<16 x i8> %a0) {
; CHECK-LABEL: stack_fold_pbroadcastb_ymm:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpbroadcastb {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <16 x i8> %a0, <16 x i8> undef, <32 x i32> zeroinitializer
  ret <32 x i8> %2
}

define <4 x i32> @stack_fold_pbroadcastd(<4 x i32> %a0) {
; CHECK-LABEL: stack_fold_pbroadcastd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpbroadcastd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    vpaddd {{.*}}(%rip), %xmm0, %xmm0
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <4 x i32> %a0, <4 x i32> undef, <4 x i32> zeroinitializer
  ; add forces execution domain
  %3 = add <4 x i32> %2, <i32 2, i32 1, i32 1, i32 1>
  ret <4 x i32> %3
}

define <8 x i32> @stack_fold_pbroadcastd_ymm(<4 x i32> %a0) {
; CHECK-LABEL: stack_fold_pbroadcastd_ymm:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpbroadcastd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    vpaddd {{.*}}(%rip), %ymm0, %ymm0
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <4 x i32> %a0, <4 x i32> undef, <8 x i32> zeroinitializer
  ; add forces execution domain
  %3 = add <8 x i32> %2, <i32 2, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
  ret <8 x i32> %3
}

define <2 x i64> @stack_fold_pbroadcastq(<2 x i64> %a0) {
; CHECK-LABEL: stack_fold_pbroadcastq:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpbroadcastq {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    vpaddq {{.*}}(%rip), %xmm0, %xmm0
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <2 x i64> %a0, <2 x i64> undef, <2 x i32> zeroinitializer
  ; add forces execution domain
  %3 = add <2 x i64> %2, <i64 2, i64 1>
  ret <2 x i64> %3
}

define <4 x i64> @stack_fold_pbroadcastq_ymm(<2 x i64> %a0) {
; CHECK-LABEL: stack_fold_pbroadcastq_ymm:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpbroadcastq {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    vpaddq {{.*}}(%rip), %ymm0, %ymm0
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <2 x i64> %a0, <2 x i64> undef, <4 x i32> zeroinitializer
  ; add forces execution domain
  %3 = add <4 x i64> %2, <i64 2, i64 1, i64 1, i64 1>
  ret <4 x i64> %3
}

define <8 x i16> @stack_fold_pbroadcastw(<8 x i16> %a0) {
; CHECK-LABEL: stack_fold_pbroadcastw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpbroadcastw {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <8 x i16> %a0, <8 x i16> undef, <8 x i32> zeroinitializer
  ret <8 x i16> %2
}

define <16 x i16> @stack_fold_pbroadcastw_ymm(<8 x i16> %a0) {
; CHECK-LABEL: stack_fold_pbroadcastw_ymm:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpbroadcastw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <8 x i16> %a0, <8 x i16> undef, <16 x i32> zeroinitializer
  ret <16 x i16> %2
}

define <32 x i8> @stack_fold_pcmpeqb(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_pcmpeqb:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpcmpeqb {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = icmp eq <32 x i8> %a0, %a1
  %3 = sext <32 x i1> %2 to <32 x i8>
  ret <32 x i8> %3
}

define <8 x i32> @stack_fold_pcmpeqd(<8 x i32> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_pcmpeqd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpcmpeqd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = icmp eq <8 x i32> %a0, %a1
  %3 = sext <8 x i1> %2 to <8 x i32>
  ret <8 x i32> %3
}

define <4 x i64> @stack_fold_pcmpeqq(<4 x i64> %a0, <4 x i64> %a1) {
; CHECK-LABEL: stack_fold_pcmpeqq:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpcmpeqq {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = icmp eq <4 x i64> %a0, %a1
  %3 = sext <4 x i1> %2 to <4 x i64>
  ret <4 x i64> %3
}

define <16 x i16> @stack_fold_pcmpeqw(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_pcmpeqw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpcmpeqw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = icmp eq <16 x i16> %a0, %a1
  %3 = sext <16 x i1> %2 to <16 x i16>
  ret <16 x i16> %3
}

define <32 x i8> @stack_fold_pcmpgtb(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_pcmpgtb:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpcmpgtb {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = icmp sgt <32 x i8> %a0, %a1
  %3 = sext <32 x i1> %2 to <32 x i8>
  ret <32 x i8> %3
}

define <8 x i32> @stack_fold_pcmpgtd(<8 x i32> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_pcmpgtd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpcmpgtd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = icmp sgt <8 x i32> %a0, %a1
  %3 = sext <8 x i1> %2 to <8 x i32>
  ret <8 x i32> %3
}

define <4 x i64> @stack_fold_pcmpgtq(<4 x i64> %a0, <4 x i64> %a1) {
; CHECK-LABEL: stack_fold_pcmpgtq:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpcmpgtq {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = icmp sgt <4 x i64> %a0, %a1
  %3 = sext <4 x i1> %2 to <4 x i64>
  ret <4 x i64> %3
}

define <16 x i16> @stack_fold_pcmpgtw(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_pcmpgtw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpcmpgtw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = icmp sgt <16 x i16> %a0, %a1
  %3 = sext <16 x i1> %2 to <16 x i16>
  ret <16 x i16> %3
}

define <8 x i32> @stack_fold_perm2i128(<8 x i32> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_perm2i128:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vperm2i128 $33, {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = ymm0[2,3],mem[0,1]
; CHECK-NEXT:    vpaddd {{.*}}(%rip), %ymm0, %ymm0
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <8 x i32> %a0, <8 x i32> %a1, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11>
  ; add forces execution domain
  %3 = add <8 x i32> %2, <i32 2, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
  ret <8 x i32> %3
}

define <8 x i32> @stack_fold_permd(<8 x i32> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_permd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpermd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    vpaddd {{.*}}(%rip), %ymm0, %ymm0
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <8 x i32> @llvm.x86.avx2.permd(<8 x i32> %a1, <8 x i32> %a0)
  ; add forces execution domain
  %3 = add <8 x i32> %2, <i32 2, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
  ret <8 x i32> %3
}
declare <8 x i32> @llvm.x86.avx2.permd(<8 x i32>, <8 x i32>) nounwind readonly

define <4 x double> @stack_fold_permpd(<4 x double> %a0) {
; CHECK-LABEL: stack_fold_permpd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpermpd $235, {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = mem[3,2,2,3]
; CHECK-NEXT:    vxorpd %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vaddpd %ymm1, %ymm0, %ymm0
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <4 x double> %a0, <4 x double> undef, <4 x i32> <i32 3, i32 2, i32 2, i32 3>
  ; fadd forces execution domain
  %3 = fadd <4 x double> %2, <double 0x0, double 0x0, double 0x0, double 0x0>
  ret <4 x double> %3
}

define <8 x float> @stack_fold_permps(<8 x i32> %a0, <8 x float> %a1) {
; CHECK-LABEL: stack_fold_permps:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpermps {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <8 x float> @llvm.x86.avx2.permps(<8 x float> %a1, <8 x i32> %a0)
  ret <8 x float> %2
}
declare <8 x float> @llvm.x86.avx2.permps(<8 x float>, <8 x i32>) nounwind readonly

define <4 x i64> @stack_fold_permq(<4 x i64> %a0) {
; CHECK-LABEL: stack_fold_permq:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpermq $235, {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = mem[3,2,2,3]
; CHECK-NEXT:    vpaddq {{.*}}(%rip), %ymm0, %ymm0
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <4 x i64> %a0, <4 x i64> undef, <4 x i32> <i32 3, i32 2, i32 2, i32 3>
  ; add forces execution domain
  %3 = add <4 x i64> %2, <i64 2, i64 1, i64 1, i64 1>
  ret <4 x i64> %3
}

define <8 x i32> @stack_fold_phaddd(<8 x i32> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_phaddd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vphaddd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <8 x i32> @llvm.x86.avx2.phadd.d(<8 x i32> %a0, <8 x i32> %a1)
  ret <8 x i32> %2
}
declare <8 x i32> @llvm.x86.avx2.phadd.d(<8 x i32>, <8 x i32>) nounwind readnone

define <16 x i16> @stack_fold_phaddsw(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_phaddsw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vphaddsw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <16 x i16> @llvm.x86.avx2.phadd.sw(<16 x i16> %a0, <16 x i16> %a1)
  ret <16 x i16> %2
}
declare <16 x i16> @llvm.x86.avx2.phadd.sw(<16 x i16>, <16 x i16>) nounwind readnone

define <16 x i16> @stack_fold_phaddw(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_phaddw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vphaddw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <16 x i16> @llvm.x86.avx2.phadd.w(<16 x i16> %a0, <16 x i16> %a1)
  ret <16 x i16> %2
}
declare <16 x i16> @llvm.x86.avx2.phadd.w(<16 x i16>, <16 x i16>) nounwind readnone

define <8 x i32> @stack_fold_phsubd(<8 x i32> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_phsubd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vphsubd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <8 x i32> @llvm.x86.avx2.phsub.d(<8 x i32> %a0, <8 x i32> %a1)
  ret <8 x i32> %2
}
declare <8 x i32> @llvm.x86.avx2.phsub.d(<8 x i32>, <8 x i32>) nounwind readnone

define <16 x i16> @stack_fold_phsubsw(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_phsubsw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vphsubsw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <16 x i16> @llvm.x86.avx2.phsub.sw(<16 x i16> %a0, <16 x i16> %a1)
  ret <16 x i16> %2
}
declare <16 x i16> @llvm.x86.avx2.phsub.sw(<16 x i16>, <16 x i16>) nounwind readnone

define <16 x i16> @stack_fold_phsubw(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_phsubw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vphsubw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <16 x i16> @llvm.x86.avx2.phsub.w(<16 x i16> %a0, <16 x i16> %a1)
  ret <16 x i16> %2
}
declare <16 x i16> @llvm.x86.avx2.phsub.w(<16 x i16>, <16 x i16>) nounwind readnone

define <16 x i16> @stack_fold_pmaddubsw(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_pmaddubsw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmaddubsw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <16 x i16> @llvm.x86.avx2.pmadd.ub.sw(<32 x i8> %a0, <32 x i8> %a1)
  ret <16 x i16> %2
}
declare <16 x i16> @llvm.x86.avx2.pmadd.ub.sw(<32 x i8>, <32 x i8>) nounwind readnone

define <8 x i32> @stack_fold_pmaddwd(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_pmaddwd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmaddwd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <8 x i32> @llvm.x86.avx2.pmadd.wd(<16 x i16> %a0, <16 x i16> %a1)
  ret <8 x i32> %2
}
declare <8 x i32> @llvm.x86.avx2.pmadd.wd(<16 x i16>, <16 x i16>) nounwind readnone

define <32 x i8> @stack_fold_pmaxsb(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_pmaxsb:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmaxsb {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = icmp sgt <32 x i8> %a0, %a1
  %3 = select <32 x i1> %2, <32 x i8> %a0, <32 x i8> %a1
  ret <32 x i8> %3
}

define <8 x i32> @stack_fold_pmaxsd(<8 x i32> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_pmaxsd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmaxsd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = icmp sgt <8 x i32> %a0, %a1
  %3 = select <8 x i1> %2, <8 x i32> %a0, <8 x i32> %a1
  ret <8 x i32> %3
}

define <16 x i16> @stack_fold_pmaxsw(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_pmaxsw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmaxsw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = icmp sgt <16 x i16> %a0, %a1
  %3 = select <16 x i1> %2, <16 x i16> %a0, <16 x i16> %a1
  ret <16 x i16> %3
}

define <32 x i8> @stack_fold_pmaxub(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_pmaxub:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmaxub {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = icmp ugt <32 x i8> %a0, %a1
  %3 = select <32 x i1> %2, <32 x i8> %a0, <32 x i8> %a1
  ret <32 x i8> %3
}

define <8 x i32> @stack_fold_pmaxud(<8 x i32> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_pmaxud:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmaxud {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = icmp ugt <8 x i32> %a0, %a1
  %3 = select <8 x i1> %2, <8 x i32> %a0, <8 x i32> %a1
  ret <8 x i32> %3
}

define <16 x i16> @stack_fold_pmaxuw(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_pmaxuw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmaxuw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = icmp ugt <16 x i16> %a0, %a1
  %3 = select <16 x i1> %2, <16 x i16> %a0, <16 x i16> %a1
  ret <16 x i16> %3
}

define <32 x i8> @stack_fold_pminsb(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_pminsb:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpminsb {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = icmp slt <32 x i8> %a0, %a1
  %3 = select <32 x i1> %2, <32 x i8> %a0, <32 x i8> %a1
  ret <32 x i8> %3
}

define <8 x i32> @stack_fold_pminsd(<8 x i32> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_pminsd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpminsd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = icmp slt <8 x i32> %a0, %a1
  %3 = select <8 x i1> %2, <8 x i32> %a0, <8 x i32> %a1
  ret <8 x i32> %3
}

define <16 x i16> @stack_fold_pminsw(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_pminsw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpminsw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = icmp slt <16 x i16> %a0, %a1
  %3 = select <16 x i1> %2, <16 x i16> %a0, <16 x i16> %a1
  ret <16 x i16> %3
}

define <32 x i8> @stack_fold_pminub(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_pminub:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpminub {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = icmp ult <32 x i8> %a0, %a1
  %3 = select <32 x i1> %2, <32 x i8> %a0, <32 x i8> %a1
  ret <32 x i8> %3
}

define <8 x i32> @stack_fold_pminud(<8 x i32> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_pminud:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpminud {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = icmp ult <8 x i32> %a0, %a1
  %3 = select <8 x i1> %2, <8 x i32> %a0, <8 x i32> %a1
  ret <8 x i32> %3
}

define <16 x i16> @stack_fold_pminuw(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_pminuw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpminuw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = icmp ult <16 x i16> %a0, %a1
  %3 = select <16 x i1> %2, <16 x i16> %a0, <16 x i16> %a1
  ret <16 x i16> %3
}

define <8 x i32> @stack_fold_pmovsxbd(<16 x i8> %a0) {
; CHECK-LABEL: stack_fold_pmovsxbd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmovsxbd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <16 x i8> %a0, <16 x i8> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
  %3 = sext <8 x i8> %2 to <8 x i32>
  ret <8 x i32> %3
}

define <4 x i64> @stack_fold_pmovsxbq(<16 x i8> %a0) {
; CHECK-LABEL: stack_fold_pmovsxbq:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmovsxbq {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <16 x i8> %a0, <16 x i8> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
  %3 = sext <4 x i8> %2 to <4 x i64>
  ret <4 x i64> %3
}

define <16 x i16> @stack_fold_pmovsxbw(<16 x i8> %a0) {
; CHECK-LABEL: stack_fold_pmovsxbw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmovsxbw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = sext <16 x i8> %a0 to <16 x i16>
  ret <16 x i16> %2
}

define <4 x i64> @stack_fold_pmovsxdq(<4 x i32> %a0) {
; CHECK-LABEL: stack_fold_pmovsxdq:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmovsxdq {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = sext <4 x i32> %a0 to <4 x i64>
  ret <4 x i64> %2
}

define <8 x i32> @stack_fold_pmovsxwd(<8 x i16> %a0) {
; CHECK-LABEL: stack_fold_pmovsxwd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmovsxwd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = sext <8 x i16> %a0 to <8 x i32>
  ret <8 x i32> %2
}

define <4 x i64> @stack_fold_pmovsxwq(<8 x i16> %a0) {
; CHECK-LABEL: stack_fold_pmovsxwq:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmovsxwq {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <8 x i16> %a0, <8 x i16> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
  %3 = sext <4 x i16> %2 to <4 x i64>
  ret <4 x i64> %3
}

define <8 x i32> @stack_fold_pmovzxbd(<16 x i8> %a0) {
; CHECK-LABEL: stack_fold_pmovzxbd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmovzxbd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    # ymm0 = mem[0],zero,zero,zero,mem[1],zero,zero,zero,mem[2],zero,zero,zero,mem[3],zero,zero,zero,mem[4],zero,zero,zero,mem[5],zero,zero,zero,mem[6],zero,zero,zero,mem[7],zero,zero,zero
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <16 x i8> %a0, <16 x i8> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
  %3 = zext <8 x i8> %2 to <8 x i32>
  ret <8 x i32> %3
}

define <4 x i64> @stack_fold_pmovzxbq(<16 x i8> %a0) {
; CHECK-LABEL: stack_fold_pmovzxbq:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmovzxbq {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    # ymm0 = mem[0],zero,zero,zero,zero,zero,zero,zero,mem[1],zero,zero,zero,zero,zero,zero,zero,mem[2],zero,zero,zero,zero,zero,zero,zero,mem[3],zero,zero,zero,zero,zero,zero,zero
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <16 x i8> %a0, <16 x i8> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
  %3 = zext <4 x i8> %2 to <4 x i64>
  ret <4 x i64> %3
}

define <16 x i16> @stack_fold_pmovzxbw(<16 x i8> %a0) {
; CHECK-LABEL: stack_fold_pmovzxbw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmovzxbw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    # ymm0 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero,mem[8],zero,mem[9],zero,mem[10],zero,mem[11],zero,mem[12],zero,mem[13],zero,mem[14],zero,mem[15],zero
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = zext <16 x i8> %a0 to <16 x i16>
  ret <16 x i16> %2
}

define <4 x i64> @stack_fold_pmovzxdq(<4 x i32> %a0) {
; CHECK-LABEL: stack_fold_pmovzxdq:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmovzxdq {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    # ymm0 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = zext <4 x i32> %a0 to <4 x i64>
  ret <4 x i64> %2
}

define <8 x i32> @stack_fold_pmovzxwd(<8 x i16> %a0) {
; CHECK-LABEL: stack_fold_pmovzxwd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmovzxwd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    # ymm0 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = zext <8 x i16> %a0 to <8 x i32>
  ret <8 x i32> %2
}

define <4 x i64> @stack_fold_pmovzxwq(<8 x i16> %a0) {
; CHECK-LABEL: stack_fold_pmovzxwq:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmovzxwq {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    # ymm0 = mem[0],zero,zero,zero,mem[1],zero,zero,zero,mem[2],zero,zero,zero,mem[3],zero,zero,zero
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <8 x i16> %a0, <8 x i16> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
  %3 = zext <4 x i16> %2 to <4 x i64>
  ret <4 x i64> %3
}

define <4 x i64> @stack_fold_pmuldq(<8 x i32> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_pmuldq:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmuldq {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = bitcast <8 x i32> %a0 to <4 x i64>
  %3 = bitcast <8 x i32> %a1 to <4 x i64>
  %4 = shl <4 x i64> %2, <i64 32, i64 32, i64 32, i64 32>
  %5 = ashr <4 x i64> %4, <i64 32, i64 32, i64 32, i64 32>
  %6 = shl <4 x i64> %3, <i64 32, i64 32, i64 32, i64 32>
  %7 = ashr <4 x i64> %6, <i64 32, i64 32, i64 32, i64 32>
  %8 = mul <4 x i64> %5, %7
  ret <4 x i64> %8
}

define <16 x i16> @stack_fold_pmulhrsw(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_pmulhrsw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmulhrsw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <16 x i16> @llvm.x86.avx2.pmul.hr.sw(<16 x i16> %a0, <16 x i16> %a1)
  ret <16 x i16> %2
}
declare <16 x i16> @llvm.x86.avx2.pmul.hr.sw(<16 x i16>, <16 x i16>) nounwind readnone

define <16 x i16> @stack_fold_pmulhuw(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_pmulhuw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmulhuw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <16 x i16> @llvm.x86.avx2.pmulhu.w(<16 x i16> %a0, <16 x i16> %a1)
  ret <16 x i16> %2
}
declare <16 x i16> @llvm.x86.avx2.pmulhu.w(<16 x i16>, <16 x i16>) nounwind readnone

define <16 x i16> @stack_fold_pmulhw(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_pmulhw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmulhw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <16 x i16> @llvm.x86.avx2.pmulh.w(<16 x i16> %a0, <16 x i16> %a1)
  ret <16 x i16> %2
}
declare <16 x i16> @llvm.x86.avx2.pmulh.w(<16 x i16>, <16 x i16>) nounwind readnone

define <8 x i32> @stack_fold_pmulld(<8 x i32> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_pmulld:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmulld {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = mul <8 x i32> %a0, %a1
  ret <8 x i32> %2
}

define <16 x i16> @stack_fold_pmullw(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_pmullw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmullw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = mul <16 x i16> %a0, %a1
  ret <16 x i16> %2
}

define <4 x i64> @stack_fold_pmuludq(<8 x i32> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_pmuludq:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpmuludq {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = bitcast <8 x i32> %a0 to <4 x i64>
  %3 = bitcast <8 x i32> %a1 to <4 x i64>
  %4 = and <4 x i64> %2, <i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295>
  %5 = and <4 x i64> %3, <i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295>
  %6 = mul <4 x i64> %4, %5
  ret <4 x i64> %6
}

define <32 x i8> @stack_fold_por(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_por:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpor {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    vpaddb {{.*}}(%rip), %ymm0, %ymm0
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = or <32 x i8> %a0, %a1
  ; add forces execution domain
  %3 = add <32 x i8> %2, <i8 2, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1>
  ret <32 x i8> %3
}

define <4 x i64> @stack_fold_psadbw(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_psadbw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsadbw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <4 x i64> @llvm.x86.avx2.psad.bw(<32 x i8> %a0, <32 x i8> %a1)
  ret <4 x i64> %2
}
declare <4 x i64> @llvm.x86.avx2.psad.bw(<32 x i8>, <32 x i8>) nounwind readnone

define <32 x i8> @stack_fold_pshufb(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_pshufb:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpshufb {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <32 x i8> @llvm.x86.avx2.pshuf.b(<32 x i8> %a0, <32 x i8> %a1)
  ret <32 x i8> %2
}
declare <32 x i8> @llvm.x86.avx2.pshuf.b(<32 x i8>, <32 x i8>) nounwind readnone

define <8 x i32> @stack_fold_pshufd(<8 x i32> %a0) {
; CHECK-LABEL: stack_fold_pshufd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpshufd $27, {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = mem[3,2,1,0,7,6,5,4]
; CHECK-NEXT:    vpaddd {{.*}}(%rip), %ymm0, %ymm0
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <8 x i32> %a0, <8 x i32> undef, <8 x i32> <i32 3, i32 2, i32 1, i32 0, i32 7, i32 6, i32 5, i32 4>
  ; add forces execution domain
  %3 = add <8 x i32> %2, <i32 2, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
  ret <8 x i32> %3
}

define <16 x i16> @stack_fold_vpshufhw(<16 x i16> %a0) {
; CHECK-LABEL: stack_fold_vpshufhw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpshufhw $27, {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = mem[0,1,2,3,7,6,5,4,8,9,10,11,15,14,13,12]
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <16 x i16> %a0, <16 x i16> undef, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 7, i32 6, i32 5, i32 4, i32 8, i32 9, i32 10, i32 11, i32 15, i32 14, i32 13, i32 12>
  ret <16 x i16> %2
}

define <16 x i16> @stack_fold_vpshuflw(<16 x i16> %a0) {
; CHECK-LABEL: stack_fold_vpshuflw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm0, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpshuflw $27, {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = mem[3,2,1,0,4,5,6,7,11,10,9,8,12,13,14,15]
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <16 x i16> %a0, <16 x i16> undef, <16 x i32> <i32 3, i32 2, i32 1, i32 0, i32 4, i32 5, i32 6, i32 7, i32 11, i32 10, i32 9, i32 8, i32 12, i32 13, i32 14, i32 15>
  ret <16 x i16> %2
}

define <32 x i8> @stack_fold_psignb(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_psignb:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsignb {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <32 x i8> @llvm.x86.avx2.psign.b(<32 x i8> %a0, <32 x i8> %a1)
  ret <32 x i8> %2
}
declare <32 x i8> @llvm.x86.avx2.psign.b(<32 x i8>, <32 x i8>) nounwind readnone

define <8 x i32> @stack_fold_psignd(<8 x i32> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_psignd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsignd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <8 x i32> @llvm.x86.avx2.psign.d(<8 x i32> %a0, <8 x i32> %a1)
  ret <8 x i32> %2
}
declare <8 x i32> @llvm.x86.avx2.psign.d(<8 x i32>, <8 x i32>) nounwind readnone

define <16 x i16> @stack_fold_psignw(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_psignw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsignw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <16 x i16> @llvm.x86.avx2.psign.w(<16 x i16> %a0, <16 x i16> %a1)
  ret <16 x i16> %2
}
declare <16 x i16> @llvm.x86.avx2.psign.w(<16 x i16>, <16 x i16>) nounwind readnone

define <8 x i32> @stack_fold_pslld(<8 x i32> %a0, <4 x i32> %a1) {
; CHECK-LABEL: stack_fold_pslld:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpslld {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <8 x i32> @llvm.x86.avx2.psll.d(<8 x i32> %a0, <4 x i32> %a1)
  ret <8 x i32> %2
}
declare <8 x i32> @llvm.x86.avx2.psll.d(<8 x i32>, <4 x i32>) nounwind readnone

define <4 x i64> @stack_fold_psllq(<4 x i64> %a0, <2 x i64> %a1) {
; CHECK-LABEL: stack_fold_psllq:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsllq {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <4 x i64> @llvm.x86.avx2.psll.q(<4 x i64> %a0, <2 x i64> %a1)
  ret <4 x i64> %2
}
declare <4 x i64> @llvm.x86.avx2.psll.q(<4 x i64>, <2 x i64>) nounwind readnone

define <4 x i32> @stack_fold_psllvd(<4 x i32> %a0, <4 x i32> %a1) {
; CHECK-LABEL: stack_fold_psllvd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsllvd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <4 x i32> @llvm.x86.avx2.psllv.d(<4 x i32> %a0, <4 x i32> %a1)
  ret <4 x i32> %2
}
declare <4 x i32> @llvm.x86.avx2.psllv.d(<4 x i32>, <4 x i32>) nounwind readnone

define <8 x i32> @stack_fold_psllvd_ymm(<8 x i32> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_psllvd_ymm:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsllvd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <8 x i32> @llvm.x86.avx2.psllv.d.256(<8 x i32> %a0, <8 x i32> %a1)
  ret <8 x i32> %2
}
declare <8 x i32> @llvm.x86.avx2.psllv.d.256(<8 x i32>, <8 x i32>) nounwind readnone

define <2 x i64> @stack_fold_psllvq(<2 x i64> %a0, <2 x i64> %a1) {
; CHECK-LABEL: stack_fold_psllvq:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsllvq {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <2 x i64> @llvm.x86.avx2.psllv.q(<2 x i64> %a0, <2 x i64> %a1)
  ret <2 x i64> %2
}
declare <2 x i64> @llvm.x86.avx2.psllv.q(<2 x i64>, <2 x i64>) nounwind readnone

define <4 x i64> @stack_fold_psllvq_ymm(<4 x i64> %a0, <4 x i64> %a1) {
; CHECK-LABEL: stack_fold_psllvq_ymm:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsllvq {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <4 x i64> @llvm.x86.avx2.psllv.q.256(<4 x i64> %a0, <4 x i64> %a1)
  ret <4 x i64> %2
}
declare <4 x i64> @llvm.x86.avx2.psllv.q.256(<4 x i64>, <4 x i64>) nounwind readnone

define <16 x i16> @stack_fold_psllw(<16 x i16> %a0, <8 x i16> %a1) {
; CHECK-LABEL: stack_fold_psllw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsllw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <16 x i16> @llvm.x86.avx2.psll.w(<16 x i16> %a0, <8 x i16> %a1)
  ret <16 x i16> %2
}
declare <16 x i16> @llvm.x86.avx2.psll.w(<16 x i16>, <8 x i16>) nounwind readnone

define <8 x i32> @stack_fold_psrad(<8 x i32> %a0, <4 x i32> %a1) {
; CHECK-LABEL: stack_fold_psrad:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsrad {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <8 x i32> @llvm.x86.avx2.psra.d(<8 x i32> %a0, <4 x i32> %a1)
  ret <8 x i32> %2
}
declare <8 x i32> @llvm.x86.avx2.psra.d(<8 x i32>, <4 x i32>) nounwind readnone

define <4 x i32> @stack_fold_psravd(<4 x i32> %a0, <4 x i32> %a1) {
; CHECK-LABEL: stack_fold_psravd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsravd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <4 x i32> @llvm.x86.avx2.psrav.d(<4 x i32> %a0, <4 x i32> %a1)
  ret <4 x i32> %2
}
declare <4 x i32> @llvm.x86.avx2.psrav.d(<4 x i32>, <4 x i32>) nounwind readnone

define <8 x i32> @stack_fold_psravd_ymm(<8 x i32> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_psravd_ymm:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsravd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <8 x i32> @llvm.x86.avx2.psrav.d.256(<8 x i32> %a0, <8 x i32> %a1)
  ret <8 x i32> %2
}
declare <8 x i32> @llvm.x86.avx2.psrav.d.256(<8 x i32>, <8 x i32>) nounwind readnone

define <16 x i16> @stack_fold_psraw(<16 x i16> %a0, <8 x i16> %a1) {
; CHECK-LABEL: stack_fold_psraw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsraw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <16 x i16> @llvm.x86.avx2.psra.w(<16 x i16> %a0, <8 x i16> %a1)
  ret <16 x i16> %2
}
declare <16 x i16> @llvm.x86.avx2.psra.w(<16 x i16>, <8 x i16>) nounwind readnone

define <8 x i32> @stack_fold_psrld(<8 x i32> %a0, <4 x i32> %a1) {
; CHECK-LABEL: stack_fold_psrld:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsrld {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <8 x i32> @llvm.x86.avx2.psrl.d(<8 x i32> %a0, <4 x i32> %a1)
  ret <8 x i32> %2
}
declare <8 x i32> @llvm.x86.avx2.psrl.d(<8 x i32>, <4 x i32>) nounwind readnone

define <4 x i64> @stack_fold_psrlq(<4 x i64> %a0, <2 x i64> %a1) {
; CHECK-LABEL: stack_fold_psrlq:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsrlq {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <4 x i64> @llvm.x86.avx2.psrl.q(<4 x i64> %a0, <2 x i64> %a1)
  ret <4 x i64> %2
}
declare <4 x i64> @llvm.x86.avx2.psrl.q(<4 x i64>, <2 x i64>) nounwind readnone

define <4 x i32> @stack_fold_psrlvd(<4 x i32> %a0, <4 x i32> %a1) {
; CHECK-LABEL: stack_fold_psrlvd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsrlvd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <4 x i32> @llvm.x86.avx2.psrlv.d(<4 x i32> %a0, <4 x i32> %a1)
  ret <4 x i32> %2
}
declare <4 x i32> @llvm.x86.avx2.psrlv.d(<4 x i32>, <4 x i32>) nounwind readnone

define <8 x i32> @stack_fold_psrlvd_ymm(<8 x i32> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_psrlvd_ymm:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsrlvd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <8 x i32> @llvm.x86.avx2.psrlv.d.256(<8 x i32> %a0, <8 x i32> %a1)
  ret <8 x i32> %2
}
declare <8 x i32> @llvm.x86.avx2.psrlv.d.256(<8 x i32>, <8 x i32>) nounwind readnone

define <2 x i64> @stack_fold_psrlvq(<2 x i64> %a0, <2 x i64> %a1) {
; CHECK-LABEL: stack_fold_psrlvq:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsrlvq {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <2 x i64> @llvm.x86.avx2.psrlv.q(<2 x i64> %a0, <2 x i64> %a1)
  ret <2 x i64> %2
}
declare <2 x i64> @llvm.x86.avx2.psrlv.q(<2 x i64>, <2 x i64>) nounwind readnone

define <4 x i64> @stack_fold_psrlvq_ymm(<4 x i64> %a0, <4 x i64> %a1) {
; CHECK-LABEL: stack_fold_psrlvq_ymm:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsrlvq {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <4 x i64> @llvm.x86.avx2.psrlv.q.256(<4 x i64> %a0, <4 x i64> %a1)
  ret <4 x i64> %2
}
declare <4 x i64> @llvm.x86.avx2.psrlv.q.256(<4 x i64>, <4 x i64>) nounwind readnone

define <16 x i16> @stack_fold_psrlw(<16 x i16> %a0, <8 x i16> %a1) {
; CHECK-LABEL: stack_fold_psrlw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsrlw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 16-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <16 x i16> @llvm.x86.avx2.psrl.w(<16 x i16> %a0, <8 x i16> %a1)
  ret <16 x i16> %2
}
declare <16 x i16> @llvm.x86.avx2.psrl.w(<16 x i16>, <8 x i16>) nounwind readnone

define <32 x i8> @stack_fold_psubb(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_psubb:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsubb {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = sub <32 x i8> %a0, %a1
  ret <32 x i8> %2
}

define <8 x i32> @stack_fold_psubd(<8 x i32> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_psubd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsubd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = sub <8 x i32> %a0, %a1
  ret <8 x i32> %2
}

define <4 x i64> @stack_fold_psubq(<4 x i64> %a0, <4 x i64> %a1) {
; CHECK-LABEL: stack_fold_psubq:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsubq {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = sub <4 x i64> %a0, %a1
  ret <4 x i64> %2
}

define <32 x i8> @stack_fold_psubsb(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_psubsb:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsubsb {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> %a0, <32 x i8> %a1)
  ret <32 x i8> %2
}
declare <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8>, <32 x i8>) nounwind readnone

define <16 x i16> @stack_fold_psubsw(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_psubsw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsubsw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> %a0, <16 x i16> %a1)
  ret <16 x i16> %2
}
declare <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16>, <16 x i16>) nounwind readnone

define <32 x i8> @stack_fold_psubusb(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_psubusb:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsubusb {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> %a0, <32 x i8> %a1)
  ret <32 x i8> %2
}
declare <32 x i8> @llvm.usub.sat.v32i8(<32 x i8>, <32 x i8>) nounwind readnone

define <16 x i16> @stack_fold_psubusw(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_psubusw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsubusw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> %a0, <16 x i16> %a1)
  ret <16 x i16> %2
}
declare <16 x i16> @llvm.usub.sat.v16i16(<16 x i16>, <16 x i16>) nounwind readnone

define <16 x i16> @stack_fold_psubw(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_psubw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpsubw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = sub <16 x i16> %a0, %a1
  ret <16 x i16> %2
}

define <32 x i8> @stack_fold_punpckhbw(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_punpckhbw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpunpckhbw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = ymm0[8],mem[8],ymm0[9],mem[9],ymm0[10],mem[10],ymm0[11],mem[11],ymm0[12],mem[12],ymm0[13],mem[13],ymm0[14],mem[14],ymm0[15],mem[15],ymm0[24],mem[24],ymm0[25],mem[25],ymm0[26],mem[26],ymm0[27],mem[27],ymm0[28],mem[28],ymm0[29],mem[29],ymm0[30],mem[30],ymm0[31],mem[31]
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <32 x i8> %a0, <32 x i8> %a1, <32 x i32> <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 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>
  ret <32 x i8> %2
}

define <8 x i32> @stack_fold_punpckhdq(<8 x i32> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_punpckhdq:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpunpckhdq {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = ymm0[2],mem[2],ymm0[3],mem[3],ymm0[6],mem[6],ymm0[7],mem[7]
; CHECK-NEXT:    vpaddd {{.*}}(%rip), %ymm0, %ymm0
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <8 x i32> %a0, <8 x i32> %a1, <8 x i32> <i32 2, i32 10, i32 3, i32 11, i32 6, i32 14, i32 7, i32 15>
  ; add forces execution domain
  %3 = add <8 x i32> %2, <i32 2, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
  ret <8 x i32> %3
}

define <4 x i64> @stack_fold_punpckhqdq(<4 x i64> %a0, <4 x i64> %a1) {
; CHECK-LABEL: stack_fold_punpckhqdq:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpunpckhqdq {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = ymm0[1],mem[1],ymm0[3],mem[3]
; CHECK-NEXT:    vpaddq {{.*}}(%rip), %ymm0, %ymm0
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <4 x i64> %a0, <4 x i64> %a1, <4 x i32> <i32 1, i32 5, i32 3, i32 7>
  ; add forces execution domain
  %3 = add <4 x i64> %2, <i64 2, i64 1, i64 1, i64 1>
  ret <4 x i64> %3
}

define <16 x i16> @stack_fold_punpckhwd(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_punpckhwd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpunpckhwd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = ymm0[4],mem[4],ymm0[5],mem[5],ymm0[6],mem[6],ymm0[7],mem[7],ymm0[12],mem[12],ymm0[13],mem[13],ymm0[14],mem[14],ymm0[15],mem[15]
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <16 x i16> %a0, <16 x i16> %a1, <16 x i32> <i32 4, i32 20, i32 5, i32 21, i32 6, i32 22, i32 7, i32 23, i32 12, i32 28, i32 13, i32 29, i32 14, i32 30, i32 15, i32 31>
  ret <16 x i16> %2
}

define <32 x i8> @stack_fold_punpcklbw(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_punpcklbw:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpunpcklbw {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[2],mem[2],ymm0[3],mem[3],ymm0[4],mem[4],ymm0[5],mem[5],ymm0[6],mem[6],ymm0[7],mem[7],ymm0[16],mem[16],ymm0[17],mem[17],ymm0[18],mem[18],ymm0[19],mem[19],ymm0[20],mem[20],ymm0[21],mem[21],ymm0[22],mem[22],ymm0[23],mem[23]
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <32 x i8> %a0, <32 x i8> %a1, <32 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 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>
  ret <32 x i8> %2
}

define <8 x i32> @stack_fold_punpckldq(<8 x i32> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_punpckldq:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpunpckldq {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[4],mem[4],ymm0[5],mem[5]
; CHECK-NEXT:    vpaddd {{.*}}(%rip), %ymm0, %ymm0
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <8 x i32> %a0, <8 x i32> %a1, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 4, i32 12, i32 5, i32 13>
  ; add forces execution domain
  %3 = add <8 x i32> %2, <i32 2, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
  ret <8 x i32> %3
}

define <4 x i64> @stack_fold_punpcklqdq(<4 x i64> %a0, <4 x i64> %a1) {
; CHECK-LABEL: stack_fold_punpcklqdq:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpunpcklqdq {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = ymm0[0],mem[0],ymm0[2],mem[2]
; CHECK-NEXT:    vpaddq {{.*}}(%rip), %ymm0, %ymm0
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <4 x i64> %a0, <4 x i64> %a1, <4 x i32> <i32 0, i32 4, i32 2, i32 6>
  ; add forces execution domain
  %3 = add <4 x i64> %2, <i64 2, i64 1, i64 1, i64 1>
  ret <4 x i64> %3
}

define <16 x i16> @stack_fold_punpcklwd(<16 x i16> %a0, <16 x i16> %a1) {
; CHECK-LABEL: stack_fold_punpcklwd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpunpcklwd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = ymm0[0],mem[0],ymm0[1],mem[1],ymm0[2],mem[2],ymm0[3],mem[3],ymm0[8],mem[8],ymm0[9],mem[9],ymm0[10],mem[10],ymm0[11],mem[11]
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <16 x i16> %a0, <16 x i16> %a1, <16 x i32> <i32 0, i32 16, i32 1, i32 17, i32 2, i32 18, i32 3, i32 19, i32 8, i32 24, i32 9, i32 25, i32 10, i32 26, i32 11, i32 27>
  ret <16 x i16> %2
}

define <32 x i8> @stack_fold_pxor(<32 x i8> %a0, <32 x i8> %a1) {
; CHECK-LABEL: stack_fold_pxor:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovups %ymm1, {{[-0-9]+}}(%r{{[sb]}}p) # 32-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vpxor {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    vpaddb {{.*}}(%rip), %ymm0, %ymm0
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = xor <32 x i8> %a0, %a1
  ; add forces execution domain
  %3 = add <32 x i8> %2, <i8 2, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1>
  ret <32 x i8> %3
}