llvm-for-llvmta/test/CodeGen/X86/stack-folding-fp-avx1.ll

; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -O3 -verify-machineinstrs -disable-peephole -mtriple=x86_64-unknown-unknown -mattr=+avx,+f16c < %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 <2 x double> @stack_fold_addpd(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_addpd:
; 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:    vaddpd {{[-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 = fadd <2 x double> %a0, %a1
  ret <2 x double> %2
}

define <4 x double> @stack_fold_addpd_ymm(<4 x double> %a0, <4 x double> %a1) {
; CHECK-LABEL: stack_fold_addpd_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:    vaddpd {{[-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 = fadd <4 x double> %a0, %a1
  ret <4 x double> %2
}

define <4 x float> @stack_fold_addps(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_addps:
; 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:    vaddps {{[-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 = fadd <4 x float> %a0, %a1
  ret <4 x float> %2
}

define <8 x float> @stack_fold_addps_ymm(<8 x float> %a0, <8 x float> %a1) {
; CHECK-LABEL: stack_fold_addps_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:    vaddps {{[-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 = fadd <8 x float> %a0, %a1
  ret <8 x float> %2
}

define double @stack_fold_addsd(double %a0, double %a1) {
; CHECK-LABEL: stack_fold_addsd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovsd %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 8-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vaddsd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 8-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 = fadd double %a0, %a1
  ret double %2
}

define <2 x double> @stack_fold_addsd_int(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_addsd_int:
; 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:    vaddsd {{[-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 = extractelement <2 x double> %a0, i32 0
  %3 = extractelement <2 x double> %a1, i32 0
  %4 = fadd double %2, %3
  %5 = insertelement <2 x double> %a0, double %4, i32 0
  ret <2 x double> %5
}
declare <2 x double> @llvm.x86.sse2.add.sd(<2 x double>, <2 x double>) nounwind readnone

define float @stack_fold_addss(float %a0, float %a1) {
; CHECK-LABEL: stack_fold_addss:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovss %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vaddss {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 4-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 = fadd float %a0, %a1
  ret float %2
}

define <4 x float> @stack_fold_addss_int(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_addss_int:
; 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:    vaddss {{[-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 = extractelement <4 x float> %a0, i32 0
  %3 = extractelement <4 x float> %a1, i32 0
  %4 = fadd float %2, %3
  %5 = insertelement <4 x float> %a0, float %4, i32 0
  ret <4 x float> %5
}
declare <4 x float> @llvm.x86.sse.add.ss(<4 x float>, <4 x float>) nounwind readnone

define <2 x double> @stack_fold_addsubpd(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_addsubpd:
; 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:    vaddsubpd {{[-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 double> @llvm.x86.sse3.addsub.pd(<2 x double> %a0, <2 x double> %a1)
  ret <2 x double> %2
}
declare <2 x double> @llvm.x86.sse3.addsub.pd(<2 x double>, <2 x double>) nounwind readnone

define <4 x double> @stack_fold_addsubpd_ymm(<4 x double> %a0, <4 x double> %a1) {
; CHECK-LABEL: stack_fold_addsubpd_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:    vaddsubpd {{[-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 double> @llvm.x86.avx.addsub.pd.256(<4 x double> %a0, <4 x double> %a1)
  ret <4 x double> %2
}
declare <4 x double> @llvm.x86.avx.addsub.pd.256(<4 x double>, <4 x double>) nounwind readnone

define <4 x float> @stack_fold_addsubps(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_addsubps:
; 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:    vaddsubps {{[-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 float> @llvm.x86.sse3.addsub.ps(<4 x float> %a0, <4 x float> %a1)
  ret <4 x float> %2
}
declare <4 x float> @llvm.x86.sse3.addsub.ps(<4 x float>, <4 x float>) nounwind readnone

define <8 x float> @stack_fold_addsubps_ymm(<8 x float> %a0, <8 x float> %a1) {
; CHECK-LABEL: stack_fold_addsubps_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:    vaddsubps {{[-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.avx.addsub.ps.256(<8 x float> %a0, <8 x float> %a1)
  ret <8 x float> %2
}
declare <8 x float> @llvm.x86.avx.addsub.ps.256(<8 x float>, <8 x float>) nounwind readnone

define <2 x double> @stack_fold_andnpd(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_andnpd:
; 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:    vandnpd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    vxorpd %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vaddpd %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 = bitcast <2 x double> %a0 to <2 x i64>
  %3 = bitcast <2 x double> %a1 to <2 x i64>
  %4 = xor <2 x i64> %2, <i64 -1, i64 -1>
  %5 = and <2 x i64> %4, %3
  %6 = bitcast <2 x i64> %5 to <2 x double>
  ; fadd forces execution domain
  %7 = fadd <2 x double> %6, <double 0x0, double 0x0>
  ret <2 x double> %7
}

define <4 x double> @stack_fold_andnpd_ymm(<4 x double> %a0, <4 x double> %a1) {
; CHECK-LABEL: stack_fold_andnpd_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:    vandnpd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; 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,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = bitcast <4 x double> %a0 to <4 x i64>
  %3 = bitcast <4 x double> %a1 to <4 x i64>
  %4 = xor <4 x i64> %2, <i64 -1, i64 -1, i64 -1, i64 -1>
  %5 = and <4 x i64> %4, %3
  %6 = bitcast <4 x i64> %5 to <4 x double>
  ; fadd forces execution domain
  %7 = fadd <4 x double> %6, <double 0x0, double 0x0, double 0x0, double 0x0>
  ret <4 x double> %7
}

define <4 x float> @stack_fold_andnps(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_andnps:
; 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:    vandnps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vaddps %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 = bitcast <4 x float> %a0 to <2 x i64>
  %3 = bitcast <4 x float> %a1 to <2 x i64>
  %4 = xor <2 x i64> %2, <i64 -1, i64 -1>
  %5 = and <2 x i64> %4, %3
  %6 = bitcast <2 x i64> %5 to <4 x float>
  ; fadd forces execution domain
  %7 = fadd <4 x float> %6, <float 0x0, float 0x0, float 0x0, float 0x0>
  ret <4 x float> %7
}

define <8 x float> @stack_fold_andnps_ymm(<8 x float> %a0, <8 x float> %a1) {
; CHECK-LABEL: stack_fold_andnps_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:    vandnps {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vaddps %ymm1, %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 = bitcast <8 x float> %a0 to <4 x i64>
  %3 = bitcast <8 x float> %a1 to <4 x i64>
  %4 = xor <4 x i64> %2, <i64 -1, i64 -1, i64 -1, i64 -1>
  %5 = and <4 x i64> %4, %3
  %6 = bitcast <4 x i64> %5 to <8 x float>
  ; fadd forces execution domain
  %7 = fadd <8 x float> %6, <float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0>
  ret <8 x float> %7
}

define <2 x double> @stack_fold_andpd(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_andpd:
; 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:    vandpd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    vxorpd %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vaddpd %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 = bitcast <2 x double> %a0 to <2 x i64>
  %3 = bitcast <2 x double> %a1 to <2 x i64>
  %4 = and <2 x i64> %2, %3
  %5 = bitcast <2 x i64> %4 to <2 x double>
  ; fadd forces execution domain
  %6 = fadd <2 x double> %5, <double 0x0, double 0x0>
  ret <2 x double> %6
}

define <4 x double> @stack_fold_andpd_ymm(<4 x double> %a0, <4 x double> %a1) {
; CHECK-LABEL: stack_fold_andpd_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:    vandpd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; 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,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = bitcast <4 x double> %a0 to <4 x i64>
  %3 = bitcast <4 x double> %a1 to <4 x i64>
  %4 = and <4 x i64> %2, %3
  %5 = bitcast <4 x i64> %4 to <4 x double>
  ; fadd forces execution domain
  %6 = fadd <4 x double> %5, <double 0x0, double 0x0, double 0x0, double 0x0>
  ret <4 x double> %6
}

define <4 x float> @stack_fold_andps(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_andps:
; 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:    vandps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vaddps %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 = bitcast <4 x float> %a0 to <2 x i64>
  %3 = bitcast <4 x float> %a1 to <2 x i64>
  %4 = and <2 x i64> %2, %3
  %5 = bitcast <2 x i64> %4 to <4 x float>
  ; fadd forces execution domain
  %6 = fadd <4 x float> %5, <float 0x0, float 0x0, float 0x0, float 0x0>
  ret <4 x float> %6
}

define <8 x float> @stack_fold_andps_ymm(<8 x float> %a0, <8 x float> %a1) {
; CHECK-LABEL: stack_fold_andps_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:    vandps {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vaddps %ymm1, %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 = bitcast <8 x float> %a0 to <4 x i64>
  %3 = bitcast <8 x float> %a1 to <4 x i64>
  %4 = and <4 x i64> %2, %3
  %5 = bitcast <4 x i64> %4 to <8 x float>
  ; fadd forces execution domain
  %6 = fadd <8 x float> %5, <float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0>
  ret <8 x float> %6
}

define <2 x double> @stack_fold_blendpd(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_blendpd:
; 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:    vblendpd $2, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    # xmm0 = xmm0[0],mem[1]
; CHECK-NEXT:    vxorpd %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vaddpd %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 = select <2 x i1> <i1 1, i1 0>, <2 x double> %a0, <2 x double> %a1
  ; fadd forces execution domain
  %3 = fadd <2 x double> %2, <double 0x0, double 0x0>
  ret <2 x double> %3
}

define <4 x double> @stack_fold_blendpd_ymm(<4 x double> %a0, <4 x double> %a1) {
; CHECK-LABEL: stack_fold_blendpd_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:    vblendpd $6, {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = ymm0[0],mem[1,2],ymm0[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,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = select <4 x i1> <i1 1, i1 0, i1 0, i1 1>, <4 x double> %a0, <4 x double> %a1
  ; fadd forces execution domain
  %3 = fadd <4 x double> %2, <double 0x0, double 0x0, double 0x0, double 0x0>
  ret <4 x double> %3}

define <4 x float> @stack_fold_blendps(<4 x float> %a0, <4 x float> %a1) {
  %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 = select <4 x i1> <i1 1, i1 0, i1 0, i1 1>, <4 x float> %a0, <4 x float> %a1
  ; fadd forces execution domain
  %3 = fadd <4 x float> %2, <float 0x0, float 0x0, float 0x0, float 0x0>
  ret <4 x float> %3
}

define <8 x float> @stack_fold_blendps_ymm(<8 x float> %a0, <8 x float> %a1) {
; CHECK-LABEL: stack_fold_blendps_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:    vblendps $102, {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = ymm0[0],mem[1,2],ymm0[3,4],mem[5,6],ymm0[7]
; CHECK-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vaddps %ymm1, %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 = select <8 x i1> <i1 1, i1 0, i1 0, i1 1, i1 1, i1 0, i1 0, i1 1>, <8 x float> %a0, <8 x float> %a1
  ; fadd forces execution domain
  %3 = fadd <8 x float> %2, <float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0>
  ret <8 x float> %3
}

define <2 x double> @stack_fold_blendvpd(<2 x double> %a0, <2 x double> %a1, <2 x double> %c) {
; CHECK-LABEL: stack_fold_blendvpd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vblendvpd %xmm0, {{[-0-9]+}}(%r{{[sb]}}p), %xmm1, %xmm0 # 16-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 <2 x double> @llvm.x86.sse41.blendvpd(<2 x double> %a1, <2 x double> %c, <2 x double> %a0)
  ret <2 x double> %2
}
declare <2 x double> @llvm.x86.sse41.blendvpd(<2 x double>, <2 x double>, <2 x double>) nounwind readnone

define <4 x double> @stack_fold_blendvpd_ymm(<4 x double> %a0, <4 x double> %a1, <4 x double> %c) {
; CHECK-LABEL: stack_fold_blendvpd_ymm:
; 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:    vblendvpd %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 <4 x double> @llvm.x86.avx.blendv.pd.256(<4 x double> %a1, <4 x double> %c, <4 x double> %a0)
  ret <4 x double> %2
}
declare <4 x double> @llvm.x86.avx.blendv.pd.256(<4 x double>, <4 x double>, <4 x double>) nounwind readnone

define <4 x float> @stack_fold_blendvps(<4 x float> %a0, <4 x float> %a1, <4 x float> %c) {
; CHECK-LABEL: stack_fold_blendvps:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm2, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vblendvps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p), %xmm1, %xmm0 # 16-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 <4 x float> @llvm.x86.sse41.blendvps(<4 x float> %a1, <4 x float> %c, <4 x float> %a0)
  ret <4 x float> %2
}
declare <4 x float> @llvm.x86.sse41.blendvps(<4 x float>, <4 x float>, <4 x float>) nounwind readnone

define <8 x float> @stack_fold_blendvps_ymm(<8 x float> %a0, <8 x float> %a1, <8 x float> %c) {
; CHECK-LABEL: stack_fold_blendvps_ymm:
; 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:    vblendvps %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 <8 x float> @llvm.x86.avx.blendv.ps.256(<8 x float> %a1, <8 x float> %c, <8 x float> %a0)
  ret <8 x float> %2
}
declare <8 x float> @llvm.x86.avx.blendv.ps.256(<8 x float>, <8 x float>, <8 x float>) nounwind readnone

define <2 x double> @stack_fold_cmppd(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_cmppd:
; 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:    vcmpeqpd {{[-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 double> @llvm.x86.sse2.cmp.pd(<2 x double> %a0, <2 x double> %a1, i8 0)
  ret <2 x double> %2
}
declare <2 x double> @llvm.x86.sse2.cmp.pd(<2 x double>, <2 x double>, i8) nounwind readnone

define <4 x double> @stack_fold_cmppd_ymm(<4 x double> %a0, <4 x double> %a1) {
; CHECK-LABEL: stack_fold_cmppd_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:    vcmpeqpd {{[-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 double> @llvm.x86.avx.cmp.pd.256(<4 x double> %a0, <4 x double> %a1, i8 0)
  ret <4 x double> %2
}
declare <4 x double> @llvm.x86.avx.cmp.pd.256(<4 x double>, <4 x double>, i8) nounwind readnone

define <4 x float> @stack_fold_cmpps(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_cmpps:
; 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:    vcmpeqps {{[-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 float> @llvm.x86.sse.cmp.ps(<4 x float> %a0, <4 x float> %a1, i8 0)
  ret <4 x float> %2
}
declare <4 x float> @llvm.x86.sse.cmp.ps(<4 x float>, <4 x float>, i8) nounwind readnone

define <8 x float> @stack_fold_cmpps_ymm(<8 x float> %a0, <8 x float> %a1) {
; CHECK-LABEL: stack_fold_cmpps_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:    vcmpeqps {{[-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.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a1, i8 0)
  ret <8 x float> %2
}
declare <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float>, <8 x float>, i8) nounwind readnone

define i32 @stack_fold_cmpsd(double %a0, double %a1) {
; CHECK-LABEL: stack_fold_cmpsd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovsd %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 8-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vcmpeqsd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 8-byte Folded Reload
; CHECK-NEXT:    vmovq %xmm0, %rax
; CHECK-NEXT:    andl $1, %eax
; CHECK-NEXT:    # kill: def $eax killed $eax killed $rax
; 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 = fcmp oeq double %a0, %a1
  %3 = zext i1 %2 to i32
  ret i32 %3
}

define <2 x double> @stack_fold_cmpsd_int(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_cmpsd_int:
; 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:    vcmpeqsd {{[-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 double> @llvm.x86.sse2.cmp.sd(<2 x double> %a0, <2 x double> %a1, i8 0)
  ret <2 x double> %2
}
declare <2 x double> @llvm.x86.sse2.cmp.sd(<2 x double>, <2 x double>, i8) nounwind readnone

define i32 @stack_fold_cmpss(float %a0, float %a1) {
; CHECK-LABEL: stack_fold_cmpss:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovss %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vcmpeqss {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 4-byte Folded Reload
; CHECK-NEXT:    vmovd %xmm0, %eax
; CHECK-NEXT:    andl $1, %eax
; 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 = fcmp oeq float %a0, %a1
  %3 = zext i1 %2 to i32
  ret i32 %3
}

define <4 x float> @stack_fold_cmpss_int(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_cmpss_int:
; 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:    vcmpeqss {{[-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 float> @llvm.x86.sse.cmp.ss(<4 x float> %a0, <4 x float> %a1, i8 0)
  ret <4 x float> %2
}
declare <4 x float> @llvm.x86.sse.cmp.ss(<4 x float>, <4 x float>, i8) nounwind readnone

; TODO stack_fold_comisd

define i32 @stack_fold_comisd_int(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_comisd_int:
; 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:    vcomisd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    setnp %al
; CHECK-NEXT:    sete %cl
; CHECK-NEXT:    andb %al, %cl
; CHECK-NEXT:    movzbl %cl, %eax
; 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 i32 @llvm.x86.sse2.comieq.sd(<2 x double> %a0, <2 x double> %a1)
  ret i32 %2
}
declare i32 @llvm.x86.sse2.comieq.sd(<2 x double>, <2 x double>) nounwind readnone

; TODO stack_fold_comiss

define i32 @stack_fold_comiss_int(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_comiss_int:
; 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:    vcomiss {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    setnp %al
; CHECK-NEXT:    sete %cl
; CHECK-NEXT:    andb %al, %cl
; CHECK-NEXT:    movzbl %cl, %eax
; 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 i32 @llvm.x86.sse.comieq.ss(<4 x float> %a0, <4 x float> %a1)
  ret i32 %2
}
declare i32 @llvm.x86.sse.comieq.ss(<4 x float>, <4 x float>) nounwind readnone

define <2 x double> @stack_fold_cvtdq2pd(<4 x i32> %a0) {
; CHECK-LABEL: stack_fold_cvtdq2pd:
; 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:    vcvtdq2pd {{[-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 <4 x i32> %a0, <4 x i32> undef, <2 x i32> <i32 0, i32 1>
  %3 = sitofp <2 x i32> %2 to <2 x double>
  ret <2 x double> %3
}
define <2 x double> @stack_fold_cvtdq2pd_int(<4 x i32> %a0) {
; CHECK-LABEL: stack_fold_cvtdq2pd_int:
; 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:    vcvtdq2pd {{[-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 <4 x i32> %a0, <4 x i32> %a0, <2 x i32> <i32 0, i32 1>
  %cvt = sitofp <2 x i32> %2 to <2 x double>
  ret <2 x double> %cvt
}

define <4 x double> @stack_fold_cvtdq2pd_ymm(<4 x i32> %a0) {
; CHECK-LABEL: stack_fold_cvtdq2pd_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:    vcvtdq2pd {{[-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 = sitofp <4 x i32> %a0 to <4 x double>
  ret <4 x double> %2
}

define <4 x double> @stack_fold_cvtdq2pd_ymm_int(<4 x i32> %a0) {
; CHECK-LABEL: stack_fold_cvtdq2pd_ymm_int:
; 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:    vcvtdq2pd {{[-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}"()
  %cvt = sitofp <4 x i32> %a0 to <4 x double>
  ret <4 x double> %cvt
}

define <4 x float> @stack_fold_cvtdq2ps(<4 x i32> %a0) {
; CHECK-LABEL: stack_fold_cvtdq2ps:
; 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:    vcvtdq2ps {{[-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 = sitofp <4 x i32> %a0 to <4 x float>
  ret <4 x float> %2
}

define <8 x float> @stack_fold_cvtdq2ps_ymm(<8 x i32> %a0) {
; CHECK-LABEL: stack_fold_cvtdq2ps_ymm:
; 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:    vcvtdq2ps {{[-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 = sitofp <8 x i32> %a0 to <8 x float>
  ret <8 x float> %2
}

define <4 x i32> @stack_fold_cvtpd2dq(<2 x double> %a0) {
; CHECK-LABEL: stack_fold_cvtpd2dq:
; 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:    vcvtpd2dqx {{[-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 = call <4 x i32> @llvm.x86.sse2.cvtpd2dq(<2 x double> %a0)
  ret <4 x i32> %2
}
declare <4 x i32> @llvm.x86.sse2.cvtpd2dq(<2 x double>) nounwind readnone

define <4 x i32> @stack_fold_cvtpd2dq_ymm(<4 x double> %a0) {
; CHECK-LABEL: stack_fold_cvtpd2dq_ymm:
; 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:    vcvtpd2dqy {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 32-byte Folded Reload
; CHECK-NEXT:    vzeroupper
; 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 = call <4 x i32> @llvm.x86.avx.cvt.pd2dq.256(<4 x double> %a0)
  ret <4 x i32> %2
}
declare <4 x i32> @llvm.x86.avx.cvt.pd2dq.256(<4 x double>) nounwind readnone

define <2 x float> @stack_fold_cvtpd2ps(<2 x double> %a0) {
; CHECK-LABEL: stack_fold_cvtpd2ps:
; 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:    vcvtpd2psx {{[-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 = fptrunc <2 x double> %a0 to <2 x float>
  ret <2 x float> %2
}

define <4 x float> @stack_fold_cvtpd2ps_ymm(<4 x double> %a0) {
; CHECK-LABEL: stack_fold_cvtpd2ps_ymm:
; 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:    vcvtpd2psy {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 32-byte Folded Reload
; CHECK-NEXT:    vzeroupper
; 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 = fptrunc <4 x double> %a0 to <4 x float>
  ret <4 x float> %2
}

define <4 x float> @stack_fold_cvtph2ps(<8 x i16> %a0) {
; CHECK-LABEL: stack_fold_cvtph2ps:
; 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:    vcvtph2ps {{[-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 = call <4 x float> @llvm.x86.vcvtph2ps.128(<8 x i16> %a0)
  ret <4 x float> %2
}
declare <4 x float> @llvm.x86.vcvtph2ps.128(<8 x i16>) nounwind readonly

define <8 x float> @stack_fold_cvtph2ps_ymm(<8 x i16> %a0) {
; CHECK-LABEL: stack_fold_cvtph2ps_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:    vcvtph2ps {{[-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 = call <8 x float> @llvm.x86.vcvtph2ps.256(<8 x i16> %a0)
  ret <8 x float> %2
}
declare <8 x float> @llvm.x86.vcvtph2ps.256(<8 x i16>) nounwind readonly

define <4 x i32> @stack_fold_cvtps2dq(<4 x float> %a0) {
; CHECK-LABEL: stack_fold_cvtps2dq:
; 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:    vcvtps2dq {{[-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 = call <4 x i32> @llvm.x86.sse2.cvtps2dq(<4 x float> %a0)
  ret <4 x i32> %2
}
declare <4 x i32> @llvm.x86.sse2.cvtps2dq(<4 x float>) nounwind readnone

define <8 x i32> @stack_fold_cvtps2dq_ymm(<8 x float> %a0) {
; CHECK-LABEL: stack_fold_cvtps2dq_ymm:
; 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:    vcvtps2dq {{[-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 = call <8 x i32> @llvm.x86.avx.cvt.ps2dq.256(<8 x float> %a0)
  ret <8 x i32> %2
}
declare <8 x i32> @llvm.x86.avx.cvt.ps2dq.256(<8 x float>) nounwind readnone

define <2 x double> @stack_fold_cvtps2pd(<4 x float> %a0) {
; CHECK-LABEL: stack_fold_cvtps2pd:
; 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:    vcvtps2pd {{[-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 <4 x float> %a0, <4 x float> undef, <2 x i32> <i32 0, i32 1>
  %3 = fpext <2 x float> %2 to <2 x double>
  ret <2 x double> %3
}

define <2 x double> @stack_fold_cvtps2pd_int(<4 x float> %a0) {
; CHECK-LABEL: stack_fold_cvtps2pd_int:
; 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:    vcvtps2pd {{[-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 <4 x float> %a0, <4 x float> %a0, <2 x i32> <i32 0, i32 1>
  %cvtps2pd = fpext <2 x float> %2 to <2 x double>
  ret <2 x double> %cvtps2pd
}

define <4 x double> @stack_fold_cvtps2pd_ymm(<4 x float> %a0) {
; CHECK-LABEL: stack_fold_cvtps2pd_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:    vcvtps2pd {{[-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 = fpext <4 x float> %a0 to <4 x double>
  ret <4 x double> %2
}

define <4 x double> @stack_fold_cvtps2pd_ymm_int(<4 x float> %a0) {
; CHECK-LABEL: stack_fold_cvtps2pd_ymm_int:
; 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:    vcvtps2pd {{[-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}"()
  %cvtps2pd = fpext <4 x float> %a0 to <4 x double>
  ret <4 x double> %cvtps2pd
}

define <8 x i16> @stack_fold_cvtps2ph_ymm(<8 x float> %a0) {
; CHECK-LABEL: stack_fold_cvtps2ph_ymm:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vcvtps2ph $0, %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
  %1 = call <8 x i16> @llvm.x86.vcvtps2ph.256(<8 x float> %a0, i32 0)
  %2 = 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 <8 x i16> %1
}
declare <8 x i16> @llvm.x86.vcvtps2ph.256(<8 x float>, i32) nounwind readonly

; TODO stack_fold_cvtsd2si

define i32 @stack_fold_cvtsd2si_int(<2 x double> %a0) {
; CHECK-LABEL: stack_fold_cvtsd2si_int:
; 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:    vcvtsd2si {{[-0-9]+}}(%r{{[sb]}}p), %eax # 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 = call i32 @llvm.x86.sse2.cvtsd2si(<2 x double> %a0)
  ret i32 %2
}
declare i32 @llvm.x86.sse2.cvtsd2si(<2 x double>) nounwind readnone

; TODO stack_fold_cvtsd2si64

define i64 @stack_fold_cvtsd2si64_int(<2 x double> %a0) {
; CHECK-LABEL: stack_fold_cvtsd2si64_int:
; 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:    vcvtsd2si {{[-0-9]+}}(%r{{[sb]}}p), %rax # 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 = call i64 @llvm.x86.sse2.cvtsd2si64(<2 x double> %a0)
  ret i64 %2
}
declare i64 @llvm.x86.sse2.cvtsd2si64(<2 x double>) nounwind readnone

define double @stack_fold_cvtsi2sd(i32 %a0) {
; CHECK-LABEL: stack_fold_cvtsi2sd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    pushq %rbp
; CHECK-NEXT:    .cfi_def_cfa_offset 16
; CHECK-NEXT:    pushq %r15
; CHECK-NEXT:    .cfi_def_cfa_offset 24
; CHECK-NEXT:    pushq %r14
; CHECK-NEXT:    .cfi_def_cfa_offset 32
; CHECK-NEXT:    pushq %r13
; CHECK-NEXT:    .cfi_def_cfa_offset 40
; CHECK-NEXT:    pushq %r12
; CHECK-NEXT:    .cfi_def_cfa_offset 48
; CHECK-NEXT:    pushq %rbx
; CHECK-NEXT:    .cfi_def_cfa_offset 56
; CHECK-NEXT:    .cfi_offset %rbx, -56
; CHECK-NEXT:    .cfi_offset %r12, -48
; CHECK-NEXT:    .cfi_offset %r13, -40
; CHECK-NEXT:    .cfi_offset %r14, -32
; CHECK-NEXT:    .cfi_offset %r15, -24
; CHECK-NEXT:    .cfi_offset %rbp, -16
; CHECK-NEXT:    movl %edi, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vcvtsi2sdl {{[-0-9]+}}(%r{{[sb]}}p), %xmm1, %xmm0 # 4-byte Folded Reload
; CHECK-NEXT:    popq %rbx
; CHECK-NEXT:    .cfi_def_cfa_offset 48
; CHECK-NEXT:    popq %r12
; CHECK-NEXT:    .cfi_def_cfa_offset 40
; CHECK-NEXT:    popq %r13
; CHECK-NEXT:    .cfi_def_cfa_offset 32
; CHECK-NEXT:    popq %r14
; CHECK-NEXT:    .cfi_def_cfa_offset 24
; CHECK-NEXT:    popq %r15
; CHECK-NEXT:    .cfi_def_cfa_offset 16
; CHECK-NEXT:    popq %rbp
; CHECK-NEXT:    .cfi_def_cfa_offset 8
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{rax},~{rbx},~{rcx},~{rdx},~{rsi},~{rdi},~{rbp},~{r8},~{r9},~{r10},~{r11},~{r12},~{r13},~{r14},~{r15}"()
  %2 = sitofp i32 %a0 to double
  ret double %2
}

define <2 x double> @stack_fold_cvtsi2sd_int(i32 %a0) {
; CHECK-LABEL: stack_fold_cvtsi2sd_int:
; CHECK:       # %bb.0:
; CHECK-NEXT:    pushq %rbp
; CHECK-NEXT:    .cfi_def_cfa_offset 16
; CHECK-NEXT:    pushq %r15
; CHECK-NEXT:    .cfi_def_cfa_offset 24
; CHECK-NEXT:    pushq %r14
; CHECK-NEXT:    .cfi_def_cfa_offset 32
; CHECK-NEXT:    pushq %r13
; CHECK-NEXT:    .cfi_def_cfa_offset 40
; CHECK-NEXT:    pushq %r12
; CHECK-NEXT:    .cfi_def_cfa_offset 48
; CHECK-NEXT:    pushq %rbx
; CHECK-NEXT:    .cfi_def_cfa_offset 56
; CHECK-NEXT:    .cfi_offset %rbx, -56
; CHECK-NEXT:    .cfi_offset %r12, -48
; CHECK-NEXT:    .cfi_offset %r13, -40
; CHECK-NEXT:    .cfi_offset %r14, -32
; CHECK-NEXT:    .cfi_offset %r15, -24
; CHECK-NEXT:    .cfi_offset %rbp, -16
; CHECK-NEXT:    movl %edi, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vcvtsi2sdl {{[-0-9]+}}(%r{{[sb]}}p), %xmm1, %xmm0 # 4-byte Folded Reload
; CHECK-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
; CHECK-NEXT:    popq %rbx
; CHECK-NEXT:    .cfi_def_cfa_offset 48
; CHECK-NEXT:    popq %r12
; CHECK-NEXT:    .cfi_def_cfa_offset 40
; CHECK-NEXT:    popq %r13
; CHECK-NEXT:    .cfi_def_cfa_offset 32
; CHECK-NEXT:    popq %r14
; CHECK-NEXT:    .cfi_def_cfa_offset 24
; CHECK-NEXT:    popq %r15
; CHECK-NEXT:    .cfi_def_cfa_offset 16
; CHECK-NEXT:    popq %rbp
; CHECK-NEXT:    .cfi_def_cfa_offset 8
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{rax},~{rbx},~{rcx},~{rdx},~{rsi},~{rdi},~{rbp},~{r8},~{r9},~{r10},~{r11},~{r12},~{r13},~{r14},~{r15}"()
  %2 = sitofp i32 %a0 to double
  %3 = insertelement <2 x double> zeroinitializer, double %2, i64 0
  ret <2 x double> %3
}

define double @stack_fold_cvtsi642sd(i64 %a0) {
; CHECK-LABEL: stack_fold_cvtsi642sd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    pushq %rbp
; CHECK-NEXT:    .cfi_def_cfa_offset 16
; CHECK-NEXT:    pushq %r15
; CHECK-NEXT:    .cfi_def_cfa_offset 24
; CHECK-NEXT:    pushq %r14
; CHECK-NEXT:    .cfi_def_cfa_offset 32
; CHECK-NEXT:    pushq %r13
; CHECK-NEXT:    .cfi_def_cfa_offset 40
; CHECK-NEXT:    pushq %r12
; CHECK-NEXT:    .cfi_def_cfa_offset 48
; CHECK-NEXT:    pushq %rbx
; CHECK-NEXT:    .cfi_def_cfa_offset 56
; CHECK-NEXT:    .cfi_offset %rbx, -56
; CHECK-NEXT:    .cfi_offset %r12, -48
; CHECK-NEXT:    .cfi_offset %r13, -40
; CHECK-NEXT:    .cfi_offset %r14, -32
; CHECK-NEXT:    .cfi_offset %r15, -24
; CHECK-NEXT:    .cfi_offset %rbp, -16
; CHECK-NEXT:    movq %rdi, {{[-0-9]+}}(%r{{[sb]}}p) # 8-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vcvtsi2sdq {{[-0-9]+}}(%r{{[sb]}}p), %xmm1, %xmm0 # 8-byte Folded Reload
; CHECK-NEXT:    popq %rbx
; CHECK-NEXT:    .cfi_def_cfa_offset 48
; CHECK-NEXT:    popq %r12
; CHECK-NEXT:    .cfi_def_cfa_offset 40
; CHECK-NEXT:    popq %r13
; CHECK-NEXT:    .cfi_def_cfa_offset 32
; CHECK-NEXT:    popq %r14
; CHECK-NEXT:    .cfi_def_cfa_offset 24
; CHECK-NEXT:    popq %r15
; CHECK-NEXT:    .cfi_def_cfa_offset 16
; CHECK-NEXT:    popq %rbp
; CHECK-NEXT:    .cfi_def_cfa_offset 8
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{rax},~{rbx},~{rcx},~{rdx},~{rsi},~{rdi},~{rbp},~{r8},~{r9},~{r10},~{r11},~{r12},~{r13},~{r14},~{r15}"()
  %2 = sitofp i64 %a0 to double
  ret double %2
}

define <2 x double> @stack_fold_cvtsi642sd_int(i64 %a0) {
; CHECK-LABEL: stack_fold_cvtsi642sd_int:
; CHECK:       # %bb.0:
; CHECK-NEXT:    pushq %rbp
; CHECK-NEXT:    .cfi_def_cfa_offset 16
; CHECK-NEXT:    pushq %r15
; CHECK-NEXT:    .cfi_def_cfa_offset 24
; CHECK-NEXT:    pushq %r14
; CHECK-NEXT:    .cfi_def_cfa_offset 32
; CHECK-NEXT:    pushq %r13
; CHECK-NEXT:    .cfi_def_cfa_offset 40
; CHECK-NEXT:    pushq %r12
; CHECK-NEXT:    .cfi_def_cfa_offset 48
; CHECK-NEXT:    pushq %rbx
; CHECK-NEXT:    .cfi_def_cfa_offset 56
; CHECK-NEXT:    .cfi_offset %rbx, -56
; CHECK-NEXT:    .cfi_offset %r12, -48
; CHECK-NEXT:    .cfi_offset %r13, -40
; CHECK-NEXT:    .cfi_offset %r14, -32
; CHECK-NEXT:    .cfi_offset %r15, -24
; CHECK-NEXT:    .cfi_offset %rbp, -16
; CHECK-NEXT:    movq %rdi, {{[-0-9]+}}(%r{{[sb]}}p) # 8-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vcvtsi2sdq {{[-0-9]+}}(%r{{[sb]}}p), %xmm1, %xmm0 # 8-byte Folded Reload
; CHECK-NEXT:    vmovq {{.*#+}} xmm0 = xmm0[0],zero
; CHECK-NEXT:    popq %rbx
; CHECK-NEXT:    .cfi_def_cfa_offset 48
; CHECK-NEXT:    popq %r12
; CHECK-NEXT:    .cfi_def_cfa_offset 40
; CHECK-NEXT:    popq %r13
; CHECK-NEXT:    .cfi_def_cfa_offset 32
; CHECK-NEXT:    popq %r14
; CHECK-NEXT:    .cfi_def_cfa_offset 24
; CHECK-NEXT:    popq %r15
; CHECK-NEXT:    .cfi_def_cfa_offset 16
; CHECK-NEXT:    popq %rbp
; CHECK-NEXT:    .cfi_def_cfa_offset 8
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{rax},~{rbx},~{rcx},~{rdx},~{rsi},~{rdi},~{rbp},~{r8},~{r9},~{r10},~{r11},~{r12},~{r13},~{r14},~{r15}"()
  %2 = sitofp i64 %a0 to double
  %3 = insertelement <2 x double> zeroinitializer, double %2, i64 0
  ret <2 x double> %3
}

define float @stack_fold_cvtsi2ss(i32 %a0) {
; CHECK-LABEL: stack_fold_cvtsi2ss:
; CHECK:       # %bb.0:
; CHECK-NEXT:    pushq %rbp
; CHECK-NEXT:    .cfi_def_cfa_offset 16
; CHECK-NEXT:    pushq %r15
; CHECK-NEXT:    .cfi_def_cfa_offset 24
; CHECK-NEXT:    pushq %r14
; CHECK-NEXT:    .cfi_def_cfa_offset 32
; CHECK-NEXT:    pushq %r13
; CHECK-NEXT:    .cfi_def_cfa_offset 40
; CHECK-NEXT:    pushq %r12
; CHECK-NEXT:    .cfi_def_cfa_offset 48
; CHECK-NEXT:    pushq %rbx
; CHECK-NEXT:    .cfi_def_cfa_offset 56
; CHECK-NEXT:    .cfi_offset %rbx, -56
; CHECK-NEXT:    .cfi_offset %r12, -48
; CHECK-NEXT:    .cfi_offset %r13, -40
; CHECK-NEXT:    .cfi_offset %r14, -32
; CHECK-NEXT:    .cfi_offset %r15, -24
; CHECK-NEXT:    .cfi_offset %rbp, -16
; CHECK-NEXT:    movl %edi, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vcvtsi2ssl {{[-0-9]+}}(%r{{[sb]}}p), %xmm1, %xmm0 # 4-byte Folded Reload
; CHECK-NEXT:    popq %rbx
; CHECK-NEXT:    .cfi_def_cfa_offset 48
; CHECK-NEXT:    popq %r12
; CHECK-NEXT:    .cfi_def_cfa_offset 40
; CHECK-NEXT:    popq %r13
; CHECK-NEXT:    .cfi_def_cfa_offset 32
; CHECK-NEXT:    popq %r14
; CHECK-NEXT:    .cfi_def_cfa_offset 24
; CHECK-NEXT:    popq %r15
; CHECK-NEXT:    .cfi_def_cfa_offset 16
; CHECK-NEXT:    popq %rbp
; CHECK-NEXT:    .cfi_def_cfa_offset 8
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{rax},~{rbx},~{rcx},~{rdx},~{rsi},~{rdi},~{rbp},~{r8},~{r9},~{r10},~{r11},~{r12},~{r13},~{r14},~{r15}"()
  %2 = sitofp i32 %a0 to float
  ret float %2
}

define <4 x float> @stack_fold_cvtsi2ss_int(i32 %a0) {
; CHECK-LABEL: stack_fold_cvtsi2ss_int:
; CHECK:       # %bb.0:
; CHECK-NEXT:    pushq %rbp
; CHECK-NEXT:    .cfi_def_cfa_offset 16
; CHECK-NEXT:    pushq %r15
; CHECK-NEXT:    .cfi_def_cfa_offset 24
; CHECK-NEXT:    pushq %r14
; CHECK-NEXT:    .cfi_def_cfa_offset 32
; CHECK-NEXT:    pushq %r13
; CHECK-NEXT:    .cfi_def_cfa_offset 40
; CHECK-NEXT:    pushq %r12
; CHECK-NEXT:    .cfi_def_cfa_offset 48
; CHECK-NEXT:    pushq %rbx
; CHECK-NEXT:    .cfi_def_cfa_offset 56
; CHECK-NEXT:    .cfi_offset %rbx, -56
; CHECK-NEXT:    .cfi_offset %r12, -48
; CHECK-NEXT:    .cfi_offset %r13, -40
; CHECK-NEXT:    .cfi_offset %r14, -32
; CHECK-NEXT:    .cfi_offset %r15, -24
; CHECK-NEXT:    .cfi_offset %rbp, -16
; CHECK-NEXT:    movl %edi, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vcvtsi2ssl {{[-0-9]+}}(%r{{[sb]}}p), %xmm1, %xmm0 # 4-byte Folded Reload
; CHECK-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0],xmm1[1,2,3]
; CHECK-NEXT:    popq %rbx
; CHECK-NEXT:    .cfi_def_cfa_offset 48
; CHECK-NEXT:    popq %r12
; CHECK-NEXT:    .cfi_def_cfa_offset 40
; CHECK-NEXT:    popq %r13
; CHECK-NEXT:    .cfi_def_cfa_offset 32
; CHECK-NEXT:    popq %r14
; CHECK-NEXT:    .cfi_def_cfa_offset 24
; CHECK-NEXT:    popq %r15
; CHECK-NEXT:    .cfi_def_cfa_offset 16
; CHECK-NEXT:    popq %rbp
; CHECK-NEXT:    .cfi_def_cfa_offset 8
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{rax},~{rbx},~{rcx},~{rdx},~{rsi},~{rdi},~{rbp},~{r8},~{r9},~{r10},~{r11},~{r12},~{r13},~{r14},~{r15}"()
  %2 = sitofp i32 %a0 to float
  %3 = insertelement <4 x float> zeroinitializer, float %2, i64 0
  ret <4 x float> %3
}

define float @stack_fold_cvtsi642ss(i64 %a0) {
; CHECK-LABEL: stack_fold_cvtsi642ss:
; CHECK:       # %bb.0:
; CHECK-NEXT:    pushq %rbp
; CHECK-NEXT:    .cfi_def_cfa_offset 16
; CHECK-NEXT:    pushq %r15
; CHECK-NEXT:    .cfi_def_cfa_offset 24
; CHECK-NEXT:    pushq %r14
; CHECK-NEXT:    .cfi_def_cfa_offset 32
; CHECK-NEXT:    pushq %r13
; CHECK-NEXT:    .cfi_def_cfa_offset 40
; CHECK-NEXT:    pushq %r12
; CHECK-NEXT:    .cfi_def_cfa_offset 48
; CHECK-NEXT:    pushq %rbx
; CHECK-NEXT:    .cfi_def_cfa_offset 56
; CHECK-NEXT:    .cfi_offset %rbx, -56
; CHECK-NEXT:    .cfi_offset %r12, -48
; CHECK-NEXT:    .cfi_offset %r13, -40
; CHECK-NEXT:    .cfi_offset %r14, -32
; CHECK-NEXT:    .cfi_offset %r15, -24
; CHECK-NEXT:    .cfi_offset %rbp, -16
; CHECK-NEXT:    movq %rdi, {{[-0-9]+}}(%r{{[sb]}}p) # 8-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vcvtsi2ssq {{[-0-9]+}}(%r{{[sb]}}p), %xmm1, %xmm0 # 8-byte Folded Reload
; CHECK-NEXT:    popq %rbx
; CHECK-NEXT:    .cfi_def_cfa_offset 48
; CHECK-NEXT:    popq %r12
; CHECK-NEXT:    .cfi_def_cfa_offset 40
; CHECK-NEXT:    popq %r13
; CHECK-NEXT:    .cfi_def_cfa_offset 32
; CHECK-NEXT:    popq %r14
; CHECK-NEXT:    .cfi_def_cfa_offset 24
; CHECK-NEXT:    popq %r15
; CHECK-NEXT:    .cfi_def_cfa_offset 16
; CHECK-NEXT:    popq %rbp
; CHECK-NEXT:    .cfi_def_cfa_offset 8
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{rax},~{rbx},~{rcx},~{rdx},~{rsi},~{rdi},~{rbp},~{r8},~{r9},~{r10},~{r11},~{r12},~{r13},~{r14},~{r15}"()
  %2 = sitofp i64 %a0 to float
  ret float %2
}

define <4 x float> @stack_fold_cvtsi642ss_int(i64 %a0) {
; CHECK-LABEL: stack_fold_cvtsi642ss_int:
; CHECK:       # %bb.0:
; CHECK-NEXT:    pushq %rbp
; CHECK-NEXT:    .cfi_def_cfa_offset 16
; CHECK-NEXT:    pushq %r15
; CHECK-NEXT:    .cfi_def_cfa_offset 24
; CHECK-NEXT:    pushq %r14
; CHECK-NEXT:    .cfi_def_cfa_offset 32
; CHECK-NEXT:    pushq %r13
; CHECK-NEXT:    .cfi_def_cfa_offset 40
; CHECK-NEXT:    pushq %r12
; CHECK-NEXT:    .cfi_def_cfa_offset 48
; CHECK-NEXT:    pushq %rbx
; CHECK-NEXT:    .cfi_def_cfa_offset 56
; CHECK-NEXT:    .cfi_offset %rbx, -56
; CHECK-NEXT:    .cfi_offset %r12, -48
; CHECK-NEXT:    .cfi_offset %r13, -40
; CHECK-NEXT:    .cfi_offset %r14, -32
; CHECK-NEXT:    .cfi_offset %r15, -24
; CHECK-NEXT:    .cfi_offset %rbp, -16
; CHECK-NEXT:    movq %rdi, {{[-0-9]+}}(%r{{[sb]}}p) # 8-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vcvtsi2ssq {{[-0-9]+}}(%r{{[sb]}}p), %xmm1, %xmm0 # 8-byte Folded Reload
; CHECK-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0],xmm1[1,2,3]
; CHECK-NEXT:    popq %rbx
; CHECK-NEXT:    .cfi_def_cfa_offset 48
; CHECK-NEXT:    popq %r12
; CHECK-NEXT:    .cfi_def_cfa_offset 40
; CHECK-NEXT:    popq %r13
; CHECK-NEXT:    .cfi_def_cfa_offset 32
; CHECK-NEXT:    popq %r14
; CHECK-NEXT:    .cfi_def_cfa_offset 24
; CHECK-NEXT:    popq %r15
; CHECK-NEXT:    .cfi_def_cfa_offset 16
; CHECK-NEXT:    popq %rbp
; CHECK-NEXT:    .cfi_def_cfa_offset 8
; CHECK-NEXT:    retq
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{rax},~{rbx},~{rcx},~{rdx},~{rsi},~{rdi},~{rbp},~{r8},~{r9},~{r10},~{r11},~{r12},~{r13},~{r14},~{r15}"()
  %2 = sitofp i64 %a0 to float
  %3 = insertelement <4 x float> zeroinitializer, float %2, i64 0
  ret <4 x float> %3
}

; TODO stack_fold_cvtss2si

define i32 @stack_fold_cvtss2si_int(<4 x float> %a0) {
; CHECK-LABEL: stack_fold_cvtss2si_int:
; 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:    vcvtss2si {{[-0-9]+}}(%r{{[sb]}}p), %eax # 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 = call i32 @llvm.x86.sse.cvtss2si(<4 x float> %a0)
  ret i32 %2
}
declare i32 @llvm.x86.sse.cvtss2si(<4 x float>) nounwind readnone

; TODO stack_fold_cvtss2si64

define i64 @stack_fold_cvtss2si64_int(<4 x float> %a0) {
; CHECK-LABEL: stack_fold_cvtss2si64_int:
; 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:    vcvtss2si {{[-0-9]+}}(%r{{[sb]}}p), %rax # 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 = call i64 @llvm.x86.sse.cvtss2si64(<4 x float> %a0)
  ret i64 %2
}
declare i64 @llvm.x86.sse.cvtss2si64(<4 x float>) nounwind readnone

define <4 x i32> @stack_fold_cvttpd2dq(<2 x double> %a0) {
; CHECK-LABEL: stack_fold_cvttpd2dq:
; 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:    vcvttpd2dqx {{[-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 = call <4 x i32> @llvm.x86.sse2.cvttpd2dq(<2 x double> %a0)
  ret <4 x i32> %2
}
declare <4 x i32> @llvm.x86.sse2.cvttpd2dq(<2 x double>) nounwind readnone

define <4 x i32> @stack_fold_cvttpd2dq_ymm(<4 x double> %a0) {
; CHECK-LABEL: stack_fold_cvttpd2dq_ymm:
; 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:    vcvttpd2dqy {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 32-byte Folded Reload
; CHECK-NEXT:    vzeroupper
; 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 = fptosi <4 x double> %a0 to <4 x i32>
  ret <4 x i32> %2
}

define <4 x i32> @stack_fold_cvttps2dq(<4 x float> %a0) {
; CHECK-LABEL: stack_fold_cvttps2dq:
; 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:    vcvttps2dq {{[-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 = fptosi <4 x float> %a0 to <4 x i32>
  ret <4 x i32> %2
}

define <8 x i32> @stack_fold_cvttps2dq_ymm(<8 x float> %a0) {
; CHECK-LABEL: stack_fold_cvttps2dq_ymm:
; 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:    vcvttps2dq {{[-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 = fptosi <8 x float> %a0 to <8 x i32>
  ret <8 x i32> %2
}

define i32 @stack_fold_cvttsd2si(double %a0) {
; CHECK-LABEL: stack_fold_cvttsd2si:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovsd %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 8-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vcvttsd2si {{[-0-9]+}}(%r{{[sb]}}p), %eax # 8-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 = fptosi double %a0 to i32
  ret i32 %2
}

define i32 @stack_fold_cvttsd2si_int(<2 x double> %a0) {
; CHECK-LABEL: stack_fold_cvttsd2si_int:
; 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:    vcvttsd2si {{[-0-9]+}}(%r{{[sb]}}p), %eax # 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 = call i32 @llvm.x86.sse2.cvttsd2si(<2 x double> %a0)
  ret i32 %2
}
declare i32 @llvm.x86.sse2.cvttsd2si(<2 x double>) nounwind readnone

define i64 @stack_fold_cvttsd2si64(double %a0) {
; CHECK-LABEL: stack_fold_cvttsd2si64:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovsd %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 8-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vcvttsd2si {{[-0-9]+}}(%r{{[sb]}}p), %rax # 8-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 = fptosi double %a0 to i64
  ret i64 %2
}

define i64 @stack_fold_cvttsd2si64_int(<2 x double> %a0) {
; CHECK-LABEL: stack_fold_cvttsd2si64_int:
; 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:    vcvttsd2si {{[-0-9]+}}(%r{{[sb]}}p), %rax # 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 = call i64 @llvm.x86.sse2.cvttsd2si64(<2 x double> %a0)
  ret i64 %2
}
declare i64 @llvm.x86.sse2.cvttsd2si64(<2 x double>) nounwind readnone

define i32 @stack_fold_cvttss2si(float %a0) {
; CHECK-LABEL: stack_fold_cvttss2si:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vcvttss2si {{[-0-9]+}}(%r{{[sb]}}p), %eax # 4-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 = fptosi float %a0 to i32
  ret i32 %2
}

define i32 @stack_fold_cvttss2si_int(<4 x float> %a0) {
; CHECK-LABEL: stack_fold_cvttss2si_int:
; 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:    vcvttss2si {{[-0-9]+}}(%r{{[sb]}}p), %eax # 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 = call i32 @llvm.x86.sse.cvttss2si(<4 x float> %a0)
  ret i32 %2
}
declare i32 @llvm.x86.sse.cvttss2si(<4 x float>) nounwind readnone

define i64 @stack_fold_cvttss2si64(float %a0) {
; CHECK-LABEL: stack_fold_cvttss2si64:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vcvttss2si {{[-0-9]+}}(%r{{[sb]}}p), %rax # 4-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 = fptosi float %a0 to i64
  ret i64 %2
}

define i64 @stack_fold_cvttss2si64_int(<4 x float> %a0) {
; CHECK-LABEL: stack_fold_cvttss2si64_int:
; 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:    vcvttss2si {{[-0-9]+}}(%r{{[sb]}}p), %rax # 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 = call i64 @llvm.x86.sse.cvttss2si64(<4 x float> %a0)
  ret i64 %2
}
declare i64 @llvm.x86.sse.cvttss2si64(<4 x float>) nounwind readnone

define <2 x double> @stack_fold_divpd(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_divpd:
; 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:    vdivpd {{[-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 = fdiv <2 x double> %a0, %a1
  ret <2 x double> %2
}

define <4 x double> @stack_fold_divpd_ymm(<4 x double> %a0, <4 x double> %a1) {
; CHECK-LABEL: stack_fold_divpd_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:    vdivpd {{[-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 = fdiv <4 x double> %a0, %a1
  ret <4 x double> %2
}

define <4 x float> @stack_fold_divps(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_divps:
; 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:    vdivps {{[-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 = fdiv <4 x float> %a0, %a1
  ret <4 x float> %2
}

define <8 x float> @stack_fold_divps_ymm(<8 x float> %a0, <8 x float> %a1) {
; CHECK-LABEL: stack_fold_divps_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:    vdivps {{[-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 = fdiv <8 x float> %a0, %a1
  ret <8 x float> %2
}

define double @stack_fold_divsd(double %a0, double %a1) {
; CHECK-LABEL: stack_fold_divsd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovsd %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 8-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vdivsd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 8-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 = fdiv double %a0, %a1
  ret double %2
}

define <2 x double> @stack_fold_divsd_int(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_divsd_int:
; 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:    vdivsd {{[-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 = extractelement <2 x double> %a0, i32 0
  %3 = extractelement <2 x double> %a1, i32 0
  %4 = fdiv double %2, %3
  %5 = insertelement <2 x double> %a0, double %4, i32 0
  ret <2 x double> %5
}

define float @stack_fold_divss(float %a0, float %a1) {
; CHECK-LABEL: stack_fold_divss:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovss %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vdivss {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 4-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 = fdiv float %a0, %a1
  ret float %2
}

define <4 x float> @stack_fold_divss_int(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_divss_int:
; 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:    vdivss {{[-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 = extractelement <4 x float> %a0, i32 0
  %3 = extractelement <4 x float> %a1, i32 0
  %4 = fdiv float %2, %3
  %5 = insertelement <4 x float> %a0, float %4, i32 0
  ret <4 x float> %5
}

define <2 x double> @stack_fold_dppd(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_dppd:
; 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:    vdppd $7, {{[-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 double> @llvm.x86.sse41.dppd(<2 x double> %a0, <2 x double> %a1, i8 7)
  ret <2 x double> %2
}
declare <2 x double> @llvm.x86.sse41.dppd(<2 x double>, <2 x double>, i8) nounwind readnone

define <4 x float> @stack_fold_dpps(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_dpps:
; 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:    vdpps $7, {{[-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 float> @llvm.x86.sse41.dpps(<4 x float> %a0, <4 x float> %a1, i8 7)
  ret <4 x float> %2
}
declare <4 x float> @llvm.x86.sse41.dpps(<4 x float>, <4 x float>, i8) nounwind readnone

define <8 x float> @stack_fold_dpps_ymm(<8 x float> %a0, <8 x float> %a1) {
; CHECK-LABEL: stack_fold_dpps_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:    vdpps $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 <8 x float> @llvm.x86.avx.dp.ps.256(<8 x float> %a0, <8 x float> %a1, i8 7)
  ret <8 x float> %2
}
declare <8 x float> @llvm.x86.avx.dp.ps.256(<8 x float>, <8 x float>, i8) nounwind readnone

define <4 x float> @stack_fold_extractf128(<8 x float> %a0, <8 x float> %a1) {
; CHECK-LABEL: stack_fold_extractf128:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vextractf128 $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
  %1 = shufflevector <8 x float> %a0, <8 x float> %a1, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
  %2 = 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 float> %1
}

define i32 @stack_fold_extractps(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_extractps:
; CHECK:       # %bb.0:
; CHECK-NEXT:    pushq %rbp
; CHECK-NEXT:    .cfi_def_cfa_offset 16
; CHECK-NEXT:    pushq %r15
; CHECK-NEXT:    .cfi_def_cfa_offset 24
; CHECK-NEXT:    pushq %r14
; CHECK-NEXT:    .cfi_def_cfa_offset 32
; CHECK-NEXT:    pushq %r13
; CHECK-NEXT:    .cfi_def_cfa_offset 40
; CHECK-NEXT:    pushq %r12
; CHECK-NEXT:    .cfi_def_cfa_offset 48
; CHECK-NEXT:    pushq %rbx
; CHECK-NEXT:    .cfi_def_cfa_offset 56
; CHECK-NEXT:    .cfi_offset %rbx, -56
; CHECK-NEXT:    .cfi_offset %r12, -48
; CHECK-NEXT:    .cfi_offset %r13, -40
; CHECK-NEXT:    .cfi_offset %r14, -32
; CHECK-NEXT:    .cfi_offset %r15, -24
; CHECK-NEXT:    .cfi_offset %rbp, -16
; CHECK-NEXT:    vaddps %xmm1, %xmm0, %xmm0
; CHECK-NEXT:    vextractps $1, %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Folded Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 4-byte Reload
; CHECK-NEXT:    popq %rbx
; CHECK-NEXT:    .cfi_def_cfa_offset 48
; CHECK-NEXT:    popq %r12
; CHECK-NEXT:    .cfi_def_cfa_offset 40
; CHECK-NEXT:    popq %r13
; CHECK-NEXT:    .cfi_def_cfa_offset 32
; CHECK-NEXT:    popq %r14
; CHECK-NEXT:    .cfi_def_cfa_offset 24
; CHECK-NEXT:    popq %r15
; CHECK-NEXT:    .cfi_def_cfa_offset 16
; CHECK-NEXT:    popq %rbp
; CHECK-NEXT:    .cfi_def_cfa_offset 8
; CHECK-NEXT:    retq
  ; fadd forces execution domain
  %1 = fadd <4 x float> %a0, %a1
  %2 = extractelement <4 x float> %1, i32 1
  %3 = bitcast float %2 to i32
  %4 = tail call <2 x i64> asm sideeffect "nop", "=x,~{rax},~{rbx},~{rcx},~{rdx},~{rsi},~{rdi},~{rbp},~{r8},~{r9},~{r10},~{r11},~{r12},~{r13},~{r14},~{r15}"()
  ret i32 %3
}

define <2 x double> @stack_fold_haddpd(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_haddpd:
; 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:    vhaddpd {{[-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 double> @llvm.x86.sse3.hadd.pd(<2 x double> %a0, <2 x double> %a1)
  ret <2 x double> %2
}
declare <2 x double> @llvm.x86.sse3.hadd.pd(<2 x double>, <2 x double>) nounwind readnone

define <4 x double> @stack_fold_haddpd_ymm(<4 x double> %a0, <4 x double> %a1) {
; CHECK-LABEL: stack_fold_haddpd_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:    vhaddpd {{[-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 double> @llvm.x86.avx.hadd.pd.256(<4 x double> %a0, <4 x double> %a1)
  ret <4 x double> %2
}
declare <4 x double> @llvm.x86.avx.hadd.pd.256(<4 x double>, <4 x double>) nounwind readnone

define <4 x float> @stack_fold_haddps(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_haddps:
; 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:    vhaddps {{[-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 float> @llvm.x86.sse3.hadd.ps(<4 x float> %a0, <4 x float> %a1)
  ret <4 x float> %2
}
declare <4 x float> @llvm.x86.sse3.hadd.ps(<4 x float>, <4 x float>) nounwind readnone

define <8 x float> @stack_fold_haddps_ymm(<8 x float> %a0, <8 x float> %a1) {
; CHECK-LABEL: stack_fold_haddps_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:    vhaddps {{[-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.avx.hadd.ps.256(<8 x float> %a0, <8 x float> %a1)
  ret <8 x float> %2
}
declare <8 x float> @llvm.x86.avx.hadd.ps.256(<8 x float>, <8 x float>) nounwind readnone

define <2 x double> @stack_fold_hsubpd(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_hsubpd:
; 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:    vhsubpd {{[-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 double> @llvm.x86.sse3.hsub.pd(<2 x double> %a0, <2 x double> %a1)
  ret <2 x double> %2
}
declare <2 x double> @llvm.x86.sse3.hsub.pd(<2 x double>, <2 x double>) nounwind readnone

define <4 x double> @stack_fold_hsubpd_ymm(<4 x double> %a0, <4 x double> %a1) {
; CHECK-LABEL: stack_fold_hsubpd_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:    vhsubpd {{[-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 double> @llvm.x86.avx.hsub.pd.256(<4 x double> %a0, <4 x double> %a1)
  ret <4 x double> %2
}
declare <4 x double> @llvm.x86.avx.hsub.pd.256(<4 x double>, <4 x double>) nounwind readnone

define <4 x float> @stack_fold_hsubps(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_hsubps:
; 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:    vhsubps {{[-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 float> @llvm.x86.sse3.hsub.ps(<4 x float> %a0, <4 x float> %a1)
  ret <4 x float> %2
}
declare <4 x float> @llvm.x86.sse3.hsub.ps(<4 x float>, <4 x float>) nounwind readnone

define <8 x float> @stack_fold_hsubps_ymm(<8 x float> %a0, <8 x float> %a1) {
; CHECK-LABEL: stack_fold_hsubps_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:    vhsubps {{[-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.avx.hsub.ps.256(<8 x float> %a0, <8 x float> %a1)
  ret <8 x float> %2
}
declare <8 x float> @llvm.x86.avx.hsub.ps.256(<8 x float>, <8 x float>) nounwind readnone

define <8 x float> @stack_fold_insertf128(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_insertf128:
; 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:    vinsertf128 $1, {{[-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 = shufflevector <4 x float> %a0, <4 x float> %a1, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
  ret <8 x float> %2
}

define <4 x float> @stack_fold_insertps(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_insertps:
; 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:    vinsertps $17, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    # xmm0 = zero,mem[0],xmm0[2,3]
; 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 float> @llvm.x86.sse41.insertps(<4 x float> %a0, <4 x float> %a1, i8 209)
  ret <4 x float> %2
}
declare <4 x float> @llvm.x86.sse41.insertps(<4 x float>, <4 x float>, i8) nounwind readnone

define <2 x double> @stack_fold_maxpd(<2 x double> %a0, <2 x double> %a1) #0 {
; CHECK-LABEL: stack_fold_maxpd:
; 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:    vmaxpd {{[-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 double> @llvm.x86.sse2.max.pd(<2 x double> %a0, <2 x double> %a1)
  ret <2 x double> %2
}
declare <2 x double> @llvm.x86.sse2.max.pd(<2 x double>, <2 x double>) nounwind readnone

define <2 x double> @stack_fold_maxpd_commutable(<2 x double> %a0, <2 x double> %a1) #1 {
; CHECK-LABEL: stack_fold_maxpd_commutable:
; 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:    vmaxpd {{[-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 double> @llvm.x86.sse2.max.pd(<2 x double> %a0, <2 x double> %a1)
  ret <2 x double> %2
}

define <4 x double> @stack_fold_maxpd_ymm(<4 x double> %a0, <4 x double> %a1) #0 {
; CHECK-LABEL: stack_fold_maxpd_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:    vmaxpd {{[-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 double> @llvm.x86.avx.max.pd.256(<4 x double> %a0, <4 x double> %a1)
  ret <4 x double> %2
}
declare <4 x double> @llvm.x86.avx.max.pd.256(<4 x double>, <4 x double>) nounwind readnone

define <4 x double> @stack_fold_maxpd_ymm_commutable(<4 x double> %a0, <4 x double> %a1) #1 {
; CHECK-LABEL: stack_fold_maxpd_ymm_commutable:
; 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:    vmaxpd {{[-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 double> @llvm.x86.avx.max.pd.256(<4 x double> %a0, <4 x double> %a1)
  ret <4 x double> %2
}

define <4 x float> @stack_fold_maxps(<4 x float> %a0, <4 x float> %a1) #0 {
; CHECK-LABEL: stack_fold_maxps:
; 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:    vmaxps {{[-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 float> @llvm.x86.sse.max.ps(<4 x float> %a0, <4 x float> %a1)
  ret <4 x float> %2
}
declare <4 x float> @llvm.x86.sse.max.ps(<4 x float>, <4 x float>) nounwind readnone

define <4 x float> @stack_fold_maxps_commutable(<4 x float> %a0, <4 x float> %a1) #1 {
; CHECK-LABEL: stack_fold_maxps_commutable:
; 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:    vmaxps {{[-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 float> @llvm.x86.sse.max.ps(<4 x float> %a0, <4 x float> %a1)
  ret <4 x float> %2
}

define <8 x float> @stack_fold_maxps_ymm(<8 x float> %a0, <8 x float> %a1) #0 {
; CHECK-LABEL: stack_fold_maxps_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:    vmaxps {{[-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.avx.max.ps.256(<8 x float> %a0, <8 x float> %a1)
  ret <8 x float> %2
}
declare <8 x float> @llvm.x86.avx.max.ps.256(<8 x float>, <8 x float>) nounwind readnone

define <8 x float> @stack_fold_maxps_ymm_commutable(<8 x float> %a0, <8 x float> %a1) #1 {
; CHECK-LABEL: stack_fold_maxps_ymm_commutable:
; 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:    vmaxps {{[-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.avx.max.ps.256(<8 x float> %a0, <8 x float> %a1)
  ret <8 x float> %2
}

define double @stack_fold_maxsd(double %a0, double %a1) #0 {
; CHECK-LABEL: stack_fold_maxsd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovsd %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 8-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vmaxsd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 8-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 = fcmp ogt double %a0, %a1
  %3 = select i1 %2, double %a0, double %a1
  ret double %3
}

define double @stack_fold_maxsd_commutable(double %a0, double %a1) #1 {
; CHECK-LABEL: stack_fold_maxsd_commutable:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovsd %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 8-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vmaxsd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 8-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 = fcmp ogt double %a0, %a1
  %3 = select i1 %2, double %a0, double %a1
  ret double %3
}

define <2 x double> @stack_fold_maxsd_int(<2 x double> %a0, <2 x double> %a1) #0 {
; CHECK-LABEL: stack_fold_maxsd_int:
; 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:    vmaxsd {{[-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 double> @llvm.x86.sse2.max.sd(<2 x double> %a0, <2 x double> %a1)
  ret <2 x double> %2
}
declare <2 x double> @llvm.x86.sse2.max.sd(<2 x double>, <2 x double>) nounwind readnone

define float @stack_fold_maxss(float %a0, float %a1) #0 {
; CHECK-LABEL: stack_fold_maxss:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovss %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vmaxss {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 4-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 = fcmp ogt float %a0, %a1
  %3 = select i1 %2, float %a0, float %a1
  ret float %3
}

define float @stack_fold_maxss_commutable(float %a0, float %a1) #1 {
; CHECK-LABEL: stack_fold_maxss_commutable:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovss %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vmaxss {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 4-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 = fcmp ogt float %a0, %a1
  %3 = select i1 %2, float %a0, float %a1
  ret float %3
}

define <4 x float> @stack_fold_maxss_int(<4 x float> %a0, <4 x float> %a1) #0 {
; CHECK-LABEL: stack_fold_maxss_int:
; 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:    vmaxss {{[-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 float> @llvm.x86.sse.max.ss(<4 x float> %a0, <4 x float> %a1)
  ret <4 x float> %2
}
declare <4 x float> @llvm.x86.sse.max.ss(<4 x float>, <4 x float>) nounwind readnone

define <2 x double> @stack_fold_minpd(<2 x double> %a0, <2 x double> %a1) #0 {
; CHECK-LABEL: stack_fold_minpd:
; 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:    vminpd {{[-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 double> @llvm.x86.sse2.min.pd(<2 x double> %a0, <2 x double> %a1)
  ret <2 x double> %2
}
declare <2 x double> @llvm.x86.sse2.min.pd(<2 x double>, <2 x double>) nounwind readnone

define <2 x double> @stack_fold_minpd_commutable(<2 x double> %a0, <2 x double> %a1) #1 {
; CHECK-LABEL: stack_fold_minpd_commutable:
; 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:    vminpd {{[-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 double> @llvm.x86.sse2.min.pd(<2 x double> %a0, <2 x double> %a1)
  ret <2 x double> %2
}

define <4 x double> @stack_fold_minpd_ymm(<4 x double> %a0, <4 x double> %a1) #0 {
; CHECK-LABEL: stack_fold_minpd_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:    vminpd {{[-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 double> @llvm.x86.avx.min.pd.256(<4 x double> %a0, <4 x double> %a1)
  ret <4 x double> %2
}
declare <4 x double> @llvm.x86.avx.min.pd.256(<4 x double>, <4 x double>) nounwind readnone

define <4 x double> @stack_fold_minpd_ymm_commutable(<4 x double> %a0, <4 x double> %a1) #1 {
; CHECK-LABEL: stack_fold_minpd_ymm_commutable:
; 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:    vminpd {{[-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 double> @llvm.x86.avx.min.pd.256(<4 x double> %a0, <4 x double> %a1)
  ret <4 x double> %2
}

define <4 x float> @stack_fold_minps(<4 x float> %a0, <4 x float> %a1) #0 {
; CHECK-LABEL: stack_fold_minps:
; 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:    vminps {{[-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 float> @llvm.x86.sse.min.ps(<4 x float> %a0, <4 x float> %a1)
  ret <4 x float> %2
}
declare <4 x float> @llvm.x86.sse.min.ps(<4 x float>, <4 x float>) nounwind readnone

define <4 x float> @stack_fold_minps_commutable(<4 x float> %a0, <4 x float> %a1) #1 {
; CHECK-LABEL: stack_fold_minps_commutable:
; 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:    vminps {{[-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 float> @llvm.x86.sse.min.ps(<4 x float> %a0, <4 x float> %a1)
  ret <4 x float> %2
}

define <8 x float> @stack_fold_minps_ymm(<8 x float> %a0, <8 x float> %a1) #0 {
; CHECK-LABEL: stack_fold_minps_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:    vminps {{[-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.avx.min.ps.256(<8 x float> %a0, <8 x float> %a1)
  ret <8 x float> %2
}
declare <8 x float> @llvm.x86.avx.min.ps.256(<8 x float>, <8 x float>) nounwind readnone

define <8 x float> @stack_fold_minps_ymm_commutable(<8 x float> %a0, <8 x float> %a1) #1 {
; CHECK-LABEL: stack_fold_minps_ymm_commutable:
; 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:    vminps {{[-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.avx.min.ps.256(<8 x float> %a0, <8 x float> %a1)
  ret <8 x float> %2
}

define double @stack_fold_minsd(double %a0, double %a1) #0 {
; CHECK-LABEL: stack_fold_minsd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovsd %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 8-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vminsd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 8-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 = fcmp olt double %a0, %a1
  %3 = select i1 %2, double %a0, double %a1
  ret double %3
}

define double @stack_fold_minsd_commutable(double %a0, double %a1) #1 {
; CHECK-LABEL: stack_fold_minsd_commutable:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovsd %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 8-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vminsd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 8-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 = fcmp olt double %a0, %a1
  %3 = select i1 %2, double %a0, double %a1
  ret double %3
}

define <2 x double> @stack_fold_minsd_int(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_minsd_int:
; 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:    vminsd {{[-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 double> @llvm.x86.sse2.min.sd(<2 x double> %a0, <2 x double> %a1)
  ret <2 x double> %2
}
declare <2 x double> @llvm.x86.sse2.min.sd(<2 x double>, <2 x double>) nounwind readnone

define float @stack_fold_minss(float %a0, float %a1) #0 {
; CHECK-LABEL: stack_fold_minss:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovss %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vminss {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 4-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 = fcmp olt float %a0, %a1
  %3 = select i1 %2, float %a0, float %a1
  ret float %3
}

define float @stack_fold_minss_commutable(float %a0, float %a1) #1 {
; CHECK-LABEL: stack_fold_minss_commutable:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovss %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vminss {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 4-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 = fcmp olt float %a0, %a1
  %3 = select i1 %2, float %a0, float %a1
  ret float %3
}

define <4 x float> @stack_fold_minss_int(<4 x float> %a0, <4 x float> %a1) #0 {
; CHECK-LABEL: stack_fold_minss_int:
; 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:    vminss {{[-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 float> @llvm.x86.sse.min.ss(<4 x float> %a0, <4 x float> %a1)
  ret <4 x float> %2
}
declare <4 x float> @llvm.x86.sse.min.ss(<4 x float>, <4 x float>) nounwind readnone

define <2 x double> @stack_fold_movddup(<2 x double> %a0) {
; CHECK-LABEL: stack_fold_movddup:
; 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:    vmovddup {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    # xmm0 = mem[0,0]
; 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, <2 x i32> <i32 0, i32 0>
  ret <2 x double> %2
}

define <4 x double> @stack_fold_movddup_ymm(<4 x double> %a0) {
; CHECK-LABEL: stack_fold_movddup_ymm:
; 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:    vmovddup {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = mem[0,0,2,2]
; 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 0, i32 0, i32 2, i32 2>
  ret <4 x double> %2
}

; TODO stack_fold_movhpd (load / store)
; TODO stack_fold_movhps (load / store)

; TODO stack_fold_movlpd (load / store)
; TODO stack_fold_movlps (load / store)

define <4 x float> @stack_fold_movshdup(<4 x float> %a0) {
; CHECK-LABEL: stack_fold_movshdup:
; 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:    vmovshdup {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    # xmm0 = mem[1,1,3,3]
; 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> <i32 1, i32 1, i32 3, i32 3>
  ret <4 x float> %2
}

define <8 x float> @stack_fold_movshdup_ymm(<8 x float> %a0) {
; CHECK-LABEL: stack_fold_movshdup_ymm:
; 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:    vmovshdup {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = mem[1,1,3,3,5,5,7,7]
; 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 float> %a0, <8 x float> undef, <8 x i32> <i32 1, i32 1, i32 3, i32 3, i32 5, i32 5, i32 7, i32 7>
  ret <8 x float> %2
}

define <4 x float> @stack_fold_movsldup(<4 x float> %a0) {
; CHECK-LABEL: stack_fold_movsldup:
; 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:    vmovsldup {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    # xmm0 = mem[0,0,2,2]
; 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> <i32 0, i32 0, i32 2, i32 2>
  ret <4 x float> %2
}

define <8 x float> @stack_fold_movsldup_ymm(<8 x float> %a0) {
; CHECK-LABEL: stack_fold_movsldup_ymm:
; 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:    vmovsldup {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = mem[0,0,2,2,4,4,6,6]
; 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 float> %a0, <8 x float> undef, <8 x i32> <i32 0, i32 0, i32 2, i32 2, i32 4, i32 4, i32 6, i32 6>
  ret <8 x float> %2
}

define <2 x double> @stack_fold_mulpd(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_mulpd:
; 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:    vmulpd {{[-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 = fmul <2 x double> %a0, %a1
  ret <2 x double> %2
}

define <4 x double> @stack_fold_mulpd_ymm(<4 x double> %a0, <4 x double> %a1) {
; CHECK-LABEL: stack_fold_mulpd_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:    vmulpd {{[-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 = fmul <4 x double> %a0, %a1
  ret <4 x double> %2
}

define <4 x float> @stack_fold_mulps(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_mulps:
; 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:    vmulps {{[-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 = fmul <4 x float> %a0, %a1
  ret <4 x float> %2
}

define <8 x float> @stack_fold_mulps_ymm(<8 x float> %a0, <8 x float> %a1) {
; CHECK-LABEL: stack_fold_mulps_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:    vmulps {{[-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 = fmul <8 x float> %a0, %a1
  ret <8 x float> %2
}

define double @stack_fold_mulsd(double %a0, double %a1) {
; CHECK-LABEL: stack_fold_mulsd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovsd %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 8-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vmulsd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 8-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 = fmul double %a0, %a1
  ret double %2
}

define <2 x double> @stack_fold_mulsd_int(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_mulsd_int:
; 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:    vmulsd {{[-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 = extractelement <2 x double> %a0, i32 0
  %3 = extractelement <2 x double> %a1, i32 0
  %4 = fmul double %2, %3
  %5 = insertelement <2 x double> %a0, double %4, i32 0
  ret <2 x double> %5
}

define float @stack_fold_mulss(float %a0, float %a1) {
; CHECK-LABEL: stack_fold_mulss:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovss %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vmulss {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 4-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 = fmul float %a0, %a1
  ret float %2
}

define <4 x float> @stack_fold_mulss_int(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_mulss_int:
; 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:    vmulss {{[-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 = extractelement <4 x float> %a0, i32 0
  %3 = extractelement <4 x float> %a1, i32 0
  %4 = fmul float %2, %3
  %5 = insertelement <4 x float> %a0, float %4, i32 0
  ret <4 x float> %5
}

define <2 x double> @stack_fold_orpd(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_orpd:
; 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:    vorpd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    vxorpd %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vaddpd %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 = bitcast <2 x double> %a0 to <2 x i64>
  %3 = bitcast <2 x double> %a1 to <2 x i64>
  %4 = or <2 x i64> %2, %3
  %5 = bitcast <2 x i64> %4 to <2 x double>
  ; fadd forces execution domain
  %6 = fadd <2 x double> %5, <double 0x0, double 0x0>
  ret <2 x double> %6
}

define <4 x double> @stack_fold_orpd_ymm(<4 x double> %a0, <4 x double> %a1) {
; CHECK-LABEL: stack_fold_orpd_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:    vorpd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; 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,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = bitcast <4 x double> %a0 to <4 x i64>
  %3 = bitcast <4 x double> %a1 to <4 x i64>
  %4 = or <4 x i64> %2, %3
  %5 = bitcast <4 x i64> %4 to <4 x double>
  ; fadd forces execution domain
  %6 = fadd <4 x double> %5, <double 0x0, double 0x0, double 0x0, double 0x0>
  ret <4 x double> %6
}

define <4 x float> @stack_fold_orps(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_orps:
; 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:    vorps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vaddps %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 = bitcast <4 x float> %a0 to <2 x i64>
  %3 = bitcast <4 x float> %a1 to <2 x i64>
  %4 = or <2 x i64> %2, %3
  %5 = bitcast <2 x i64> %4 to <4 x float>
  ; fadd forces execution domain
  %6 = fadd <4 x float> %5, <float 0x0, float 0x0, float 0x0, float 0x0>
  ret <4 x float> %6
}

define <8 x float> @stack_fold_orps_ymm(<8 x float> %a0, <8 x float> %a1) {
; CHECK-LABEL: stack_fold_orps_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:    vorps {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vaddps %ymm1, %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 = bitcast <8 x float> %a0 to <4 x i64>
  %3 = bitcast <8 x float> %a1 to <4 x i64>
  %4 = or <4 x i64> %2, %3
  %5 = bitcast <4 x i64> %4 to <8 x float>
  ; fadd forces execution domain
  %6 = fadd <8 x float> %5, <float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0>
  ret <8 x float> %6
}

define <8 x float> @stack_fold_perm2f128(<8 x float> %a0, <8 x float> %a1) {
; CHECK-LABEL: stack_fold_perm2f128:
; 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:    vperm2f128 $33, {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = ymm0[2,3],mem[0,1]
; 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 float> %a0, <8 x float> %a1, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11>
  ret <8 x float> %2
}

define <2 x double> @stack_fold_permilpd(<2 x double> %a0) {
; CHECK-LABEL: stack_fold_permilpd:
; 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:    vpermilpd $1, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    # xmm0 = mem[1,0]
; 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, <2 x i32> <i32 1, i32 0>
  ret <2 x double> %2
}

define <4 x double> @stack_fold_permilpd_ymm(<4 x double> %a0) {
; CHECK-LABEL: stack_fold_permilpd_ymm:
; 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:    vpermilpd $5, {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = mem[1,0,3,2]
; 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 1, i32 0, i32 3, i32 2>
  ret <4 x double> %2
}

define <2 x double> @stack_fold_permilpdvar(<2 x double> %a0, <2 x i64> %a1) {
; CHECK-LABEL: stack_fold_permilpdvar:
; 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:    vpermilpd {{[-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 double> @llvm.x86.avx.vpermilvar.pd(<2 x double> %a0, <2 x i64> %a1)
  ret <2 x double> %2
}
declare <2 x double> @llvm.x86.avx.vpermilvar.pd(<2 x double>, <2 x i64>) nounwind readnone

define <4 x double> @stack_fold_permilpdvar_ymm(<4 x double> %a0, <4 x i64> %a1) {
; CHECK-LABEL: stack_fold_permilpdvar_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:    vpermilpd {{[-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 double> @llvm.x86.avx.vpermilvar.pd.256(<4 x double> %a0, <4 x i64> %a1)
  ret <4 x double> %2
}
declare <4 x double> @llvm.x86.avx.vpermilvar.pd.256(<4 x double>, <4 x i64>) nounwind readnone

define <4 x float> @stack_fold_permilps(<4 x float> %a0) {
; CHECK-LABEL: stack_fold_permilps:
; 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:    vpermilps $27, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    # xmm0 = mem[3,2,1,0]
; 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> <i32 3, i32 2, i32 1, i32 0>
  ret <4 x float> %2
}

define <8 x float> @stack_fold_permilps_ymm(<8 x float> %a0) {
; CHECK-LABEL: stack_fold_permilps_ymm:
; 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:    vpermilps $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:    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 float> %a0, <8 x float> undef, <8 x i32> <i32 3, i32 2, i32 1, i32 0, i32 7, i32 6, i32 5, i32 4>
  ret <8 x float> %2
}

define <4 x float> @stack_fold_permilpsvar(<4 x float> %a0, <4 x i32> %a1) {
; CHECK-LABEL: stack_fold_permilpsvar:
; 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:    vpermilps {{[-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 float> @llvm.x86.avx.vpermilvar.ps(<4 x float> %a0, <4 x i32> %a1)
  ret <4 x float> %2
}
declare <4 x float> @llvm.x86.avx.vpermilvar.ps(<4 x float>, <4 x i32>) nounwind readnone

define <8 x float> @stack_fold_permilpsvar_ymm(<8 x float> %a0, <8 x i32> %a1) {
; CHECK-LABEL: stack_fold_permilpsvar_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:    vpermilps {{[-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.avx.vpermilvar.ps.256(<8 x float> %a0, <8 x i32> %a1)
  ret <8 x float> %2
}
declare <8 x float> @llvm.x86.avx.vpermilvar.ps.256(<8 x float>, <8 x i32>) nounwind readnone

; TODO stack_fold_rcpps

define <4 x float> @stack_fold_rcpps_int(<4 x float> %a0) {
; CHECK-LABEL: stack_fold_rcpps_int:
; 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:    vrcpps {{[-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 = call <4 x float> @llvm.x86.sse.rcp.ps(<4 x float> %a0)
  ret <4 x float> %2
}
declare <4 x float> @llvm.x86.sse.rcp.ps(<4 x float>) nounwind readnone

; TODO stack_fold_rcpps_ymm

define <8 x float> @stack_fold_rcpps_ymm_int(<8 x float> %a0) {
; CHECK-LABEL: stack_fold_rcpps_ymm_int:
; 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:    vrcpps {{[-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 = call <8 x float> @llvm.x86.avx.rcp.ps.256(<8 x float> %a0)
  ret <8 x float> %2
}
declare <8 x float> @llvm.x86.avx.rcp.ps.256(<8 x float>) nounwind readnone

; TODO stack_fold_rcpss
; TODO stack_fold_rcpss_int

define <2 x double> @stack_fold_roundpd(<2 x double> %a0) {
; CHECK-LABEL: stack_fold_roundpd:
; 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:    vroundpd $7, {{[-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 = call <2 x double> @llvm.x86.sse41.round.pd(<2 x double> %a0, i32 7)
  ret <2 x double> %2
}
declare <2 x double> @llvm.x86.sse41.round.pd(<2 x double>, i32) nounwind readnone

define <4 x double> @stack_fold_roundpd_ymm(<4 x double> %a0) {
; CHECK-LABEL: stack_fold_roundpd_ymm:
; 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:    vroundpd $7, {{[-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 = call <4 x double> @llvm.x86.avx.round.pd.256(<4 x double> %a0, i32 7)
  ret <4 x double> %2
}
declare <4 x double> @llvm.x86.avx.round.pd.256(<4 x double>, i32) nounwind readnone

define <4 x float> @stack_fold_roundps(<4 x float> %a0) {
; CHECK-LABEL: stack_fold_roundps:
; 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:    vroundps $7, {{[-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 = call <4 x float> @llvm.x86.sse41.round.ps(<4 x float> %a0, i32 7)
  ret <4 x float> %2
}
declare <4 x float> @llvm.x86.sse41.round.ps(<4 x float>, i32) nounwind readnone

define <8 x float> @stack_fold_roundps_ymm(<8 x float> %a0) {
; CHECK-LABEL: stack_fold_roundps_ymm:
; 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:    vroundps $7, {{[-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 = call <8 x float> @llvm.x86.avx.round.ps.256(<8 x float> %a0, i32 7)
  ret <8 x float> %2
}
declare <8 x float> @llvm.x86.avx.round.ps.256(<8 x float>, i32) nounwind readnone

define double @stack_fold_roundsd(double %a0) optsize {
; CHECK-LABEL: stack_fold_roundsd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovsd %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 8-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vxorps %xmm0, %xmm0, %xmm0
; CHECK-NEXT:    vroundsd $9, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 8-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 = call double @llvm.floor.f64(double %a0)
  ret double %2
}

define double @stack_fold_roundsd_minsize(double %a0) minsize {
; CHECK-LABEL: stack_fold_roundsd_minsize:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovsd %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 8-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vroundsd $9, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 8-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 = call double @llvm.floor.f64(double %a0)
  ret double %2
}
declare double @llvm.floor.f64(double) nounwind readnone

define <2 x double> @stack_fold_roundsd_int(<2 x double> %a0, <2 x double> %a1) optsize {
; CHECK-LABEL: stack_fold_roundsd_int:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vmovapd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Reload
; CHECK-NEXT:    vroundsd $7, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %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 = call <2 x double> @llvm.x86.sse41.round.sd(<2 x double> %a0, <2 x double> %a1, i32 7)
  ret <2 x double> %2
}
declare <2 x double> @llvm.x86.sse41.round.sd(<2 x double>, <2 x double>, i32) nounwind readnone

define float @stack_fold_roundss(float %a0) optsize {
; CHECK-LABEL: stack_fold_roundss:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vxorps %xmm0, %xmm0, %xmm0
; CHECK-NEXT:    vroundss $9, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 4-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 = call float @llvm.floor.f32(float %a0)
  ret float %2
}
declare float @llvm.floor.f32(float) nounwind readnone

define <4 x float> @stack_fold_roundss_int(<4 x float> %a0, <4 x float> %a1) optsize {
; CHECK-LABEL: stack_fold_roundss_int:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovaps %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte Spill
; CHECK-NEXT:    vmovaps %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 16-byte 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:    vroundss $7, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %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 = call <4 x float> @llvm.x86.sse41.round.ss(<4 x float> %a0, <4 x float> %a1, i32 7)
  ret <4 x float> %2
}
declare <4 x float> @llvm.x86.sse41.round.ss(<4 x float>, <4 x float>, i32) nounwind readnone

; TODO stack_fold_rsqrtps

define <4 x float> @stack_fold_rsqrtps_int(<4 x float> %a0) {
; CHECK-LABEL: stack_fold_rsqrtps_int:
; 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:    vrsqrtps {{[-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 = call <4 x float> @llvm.x86.sse.rsqrt.ps(<4 x float> %a0)
  ret <4 x float> %2
}
declare <4 x float> @llvm.x86.sse.rsqrt.ps(<4 x float>) nounwind readnone

; TODO stack_fold_rsqrtps_ymm

define <8 x float> @stack_fold_rsqrtps_ymm_int(<8 x float> %a0) {
; CHECK-LABEL: stack_fold_rsqrtps_ymm_int:
; 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:    vrsqrtps {{[-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 = call <8 x float> @llvm.x86.avx.rsqrt.ps.256(<8 x float> %a0)
  ret <8 x float> %2
}
declare <8 x float> @llvm.x86.avx.rsqrt.ps.256(<8 x float>) nounwind readnone

; TODO stack_fold_rsqrtss
; TODO stack_fold_rsqrtss_int

define <2 x double> @stack_fold_shufpd(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_shufpd:
; 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:    vshufpd $1, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    # xmm0 = xmm0[1],mem[0]
; CHECK-NEXT:    vxorpd %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vaddpd %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 <2 x double> %a0, <2 x double> %a1, <2 x i32> <i32 1, i32 2>
  ; fadd forces execution domain
  %3 = fadd <2 x double> %2, <double 0x0, double 0x0>
  ret <2 x double> %3
}

define <4 x double> @stack_fold_shufpd_ymm(<4 x double> %a0, <4 x double> %a1) {
; CHECK-LABEL: stack_fold_shufpd_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:    vshufpd $5, {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = ymm0[1],mem[0],ymm0[3],mem[2]
; 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,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <4 x double> %a0, <4 x double> %a1, <4 x i32> <i32 1, i32 4, i32 3, i32 6>
  ; fadd forces execution domain
  %3 = fadd <4 x double> %2, <double 0x0, double 0x0, double 0x0, double 0x0>
  ret <4 x double> %3
}

define <4 x float> @stack_fold_shufps(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_shufps:
; 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:    vshufps $200, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    # xmm0 = xmm0[0,2],mem[0,3]
; 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 float> %a0, <4 x float> %a1, <4 x i32> <i32 0, i32 2, i32 4, i32 7>
  ret <4 x float> %2
}

define <8 x float> @stack_fold_shufps_ymm(<8 x float> %a0, <8 x float> %a1) {
; CHECK-LABEL: stack_fold_shufps_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:    vshufps $148, {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = ymm0[0,1],mem[1,2],ymm0[4,5],mem[5,6]
; 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 float> %a0, <8 x float> %a1, <8 x i32> <i32 0, i32 1, i32 9, i32 10, i32 4, i32 5, i32 13, i32 14>
  ret <8 x float> %2
}

define <2 x double> @stack_fold_sqrtpd(<2 x double> %a0) {
; CHECK-LABEL: stack_fold_sqrtpd:
; 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:    vsqrtpd {{[-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 = call <2 x double> @llvm.sqrt.v2f64(<2 x double> %a0)
  ret <2 x double> %2
}
declare <2 x double> @llvm.sqrt.v2f64(<2 x double>)

define <4 x double> @stack_fold_sqrtpd_ymm(<4 x double> %a0) {
; CHECK-LABEL: stack_fold_sqrtpd_ymm:
; 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:    vsqrtpd {{[-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 = call <4 x double> @llvm.sqrt.v4f64(<4 x double> %a0)
  ret <4 x double> %2
}
declare <4 x double> @llvm.sqrt.v4f64(<4 x double>)

define <4 x float> @stack_fold_sqrtps(<4 x float> %a0) {
; CHECK-LABEL: stack_fold_sqrtps:
; 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:    vsqrtps {{[-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 = call <4 x float> @llvm.sqrt.v4f32(<4 x float> %a0)
  ret <4 x float> %2
}
declare <4 x float> @llvm.sqrt.v4f32(<4 x float>)

define <8 x float> @stack_fold_sqrtps_ymm(<8 x float> %a0) {
; CHECK-LABEL: stack_fold_sqrtps_ymm:
; 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:    vsqrtps {{[-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 = call <8 x float> @llvm.sqrt.v8f32(<8 x float> %a0)
  ret <8 x float> %2
}
declare <8 x float> @llvm.sqrt.v8f32(<8 x float>)

define double @stack_fold_sqrtsd(double %a0) optsize {
; CHECK-LABEL: stack_fold_sqrtsd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovsd %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 8-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vxorps %xmm0, %xmm0, %xmm0
; CHECK-NEXT:    vsqrtsd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 8-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 = call double @llvm.sqrt.f64(double %a0)
  ret double %2
}
declare double @llvm.sqrt.f64(double) nounwind readnone

; TODO stack_fold_sqrtsd_int

define float @stack_fold_sqrtss(float %a0) optsize {
; CHECK-LABEL: stack_fold_sqrtss:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovss %xmm0, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vxorps %xmm0, %xmm0, %xmm0
; CHECK-NEXT:    vsqrtss {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 4-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 = call float @llvm.sqrt.f32(float %a0)
  ret float %2
}
declare float @llvm.sqrt.f32(float) nounwind readnone

; TODO stack_fold_sqrtss_int

define <2 x double> @stack_fold_subpd(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_subpd:
; 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:    vsubpd {{[-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 = fsub <2 x double> %a0, %a1
  ret <2 x double> %2
}

define <4 x double> @stack_fold_subpd_ymm(<4 x double> %a0, <4 x double> %a1) {
; CHECK-LABEL: stack_fold_subpd_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:    vsubpd {{[-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 = fsub <4 x double> %a0, %a1
  ret <4 x double> %2
}

define <4 x float> @stack_fold_subps(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_subps:
; 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:    vsubps {{[-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 = fsub <4 x float> %a0, %a1
  ret <4 x float> %2
}

define <8 x float> @stack_fold_subps_ymm(<8 x float> %a0, <8 x float> %a1) {
; CHECK-LABEL: stack_fold_subps_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:    vsubps {{[-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 = fsub <8 x float> %a0, %a1
  ret <8 x float> %2
}

define double @stack_fold_subsd(double %a0, double %a1) {
; CHECK-LABEL: stack_fold_subsd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovsd %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 8-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vsubsd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 8-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 = fsub double %a0, %a1
  ret double %2
}

define <2 x double> @stack_fold_subsd_int(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_subsd_int:
; 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:    vsubsd {{[-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 = extractelement <2 x double> %a0, i32 0
  %3 = extractelement <2 x double> %a1, i32 0
  %4 = fsub double %2, %3
  %5 = insertelement <2 x double> %a0, double %4, i32 0
  ret <2 x double> %5
}

define float @stack_fold_subss(float %a0, float %a1) {
; CHECK-LABEL: stack_fold_subss:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovss %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    vsubss {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 4-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 = fsub float %a0, %a1
  ret float %2
}

define <4 x float> @stack_fold_subss_int(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_subss_int:
; 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:    vsubss {{[-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 = extractelement <4 x float> %a0, i32 0
  %3 = extractelement <4 x float> %a1, i32 0
  %4 = fsub float %2, %3
  %5 = insertelement <4 x float> %a0, float %4, i32 0
  ret <4 x float> %5
}

define i32 @stack_fold_testpd(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_testpd:
; 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:    xorl %eax, %eax
; CHECK-NEXT:    vtestpd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    setb %al
; 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 i32 @llvm.x86.avx.vtestc.pd(<2 x double> %a0, <2 x double> %a1)
  ret i32 %2
}
declare i32 @llvm.x86.avx.vtestc.pd(<2 x double>, <2 x double>) nounwind readnone

define i32 @stack_fold_testpd_ymm(<4 x double> %a0, <4 x double> %a1) {
; CHECK-LABEL: stack_fold_testpd_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:    xorl %eax, %eax
; CHECK-NEXT:    vtestpd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    setb %al
; CHECK-NEXT:    vzeroupper
; 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 i32 @llvm.x86.avx.vtestc.pd.256(<4 x double> %a0, <4 x double> %a1)
  ret i32 %2
}
declare i32 @llvm.x86.avx.vtestc.pd.256(<4 x double>, <4 x double>) nounwind readnone

define i32 @stack_fold_testps(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_testps:
; 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:    xorl %eax, %eax
; CHECK-NEXT:    vtestps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    setb %al
; 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 i32 @llvm.x86.avx.vtestc.ps(<4 x float> %a0, <4 x float> %a1)
  ret i32 %2
}
declare i32 @llvm.x86.avx.vtestc.ps(<4 x float>, <4 x float>) nounwind readnone

define i32 @stack_fold_testps_ymm(<8 x float> %a0, <8 x float> %a1) {
; CHECK-LABEL: stack_fold_testps_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:    xorl %eax, %eax
; CHECK-NEXT:    vtestps {{[-0-9]+}}(%r{{[sb]}}p), %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    setb %al
; CHECK-NEXT:    vzeroupper
; 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 i32 @llvm.x86.avx.vtestc.ps.256(<8 x float> %a0, <8 x float> %a1)
  ret i32 %2
}
declare i32 @llvm.x86.avx.vtestc.ps.256(<8 x float>, <8 x float>) nounwind readnone

define i32 @stack_fold_ucomisd(double %a0, double %a1) {
; CHECK-LABEL: stack_fold_ucomisd:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovsd %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 8-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    xorl %eax, %eax
; CHECK-NEXT:    vucomisd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 8-byte Folded Reload
; CHECK-NEXT:    sete %al
; CHECK-NEXT:    leal -1(%rax,%rax), %eax
; 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 = fcmp ueq double %a0, %a1
  %3 = select i1 %2, i32 1, i32 -1
  ret i32 %3
}

define i32 @stack_fold_ucomisd_int(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_ucomisd_int:
; 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:    vucomisd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    setnp %al
; CHECK-NEXT:    sete %cl
; CHECK-NEXT:    andb %al, %cl
; CHECK-NEXT:    movzbl %cl, %eax
; 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 i32 @llvm.x86.sse2.ucomieq.sd(<2 x double> %a0, <2 x double> %a1)
  ret i32 %2
}
declare i32 @llvm.x86.sse2.ucomieq.sd(<2 x double>, <2 x double>) nounwind readnone

define i32 @stack_fold_ucomiss(float %a0, float %a1) {
; CHECK-LABEL: stack_fold_ucomiss:
; CHECK:       # %bb.0:
; CHECK-NEXT:    vmovss %xmm1, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
; CHECK-NEXT:    #APP
; CHECK-NEXT:    nop
; CHECK-NEXT:    #NO_APP
; CHECK-NEXT:    xorl %eax, %eax
; CHECK-NEXT:    vucomiss {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 4-byte Folded Reload
; CHECK-NEXT:    sete %al
; CHECK-NEXT:    leal -1(%rax,%rax), %eax
; 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 = fcmp ueq float %a0, %a1
  %3 = select i1 %2, i32 1, i32 -1
  ret i32 %3
}

define i32 @stack_fold_ucomiss_int(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_ucomiss_int:
; 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:    vucomiss {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    setnp %al
; CHECK-NEXT:    sete %cl
; CHECK-NEXT:    andb %al, %cl
; CHECK-NEXT:    movzbl %cl, %eax
; 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 i32 @llvm.x86.sse.ucomieq.ss(<4 x float> %a0, <4 x float> %a1)
  ret i32 %2
}
declare i32 @llvm.x86.sse.ucomieq.ss(<4 x float>, <4 x float>) nounwind readnone

define <2 x double> @stack_fold_unpckhpd(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_unpckhpd:
; 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:    vunpckhpd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    # xmm0 = xmm0[1],mem[1]
; CHECK-NEXT:    vxorpd %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vaddpd %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 <2 x double> %a0, <2 x double> %a1, <2 x i32> <i32 1, i32 3>
  ; fadd forces execution domain
  %3 = fadd <2 x double> %2, <double 0x0, double 0x0>
  ret <2 x double> %3
}

define <4 x double> @stack_fold_unpckhpd_ymm(<4 x double> %a0, <4 x double> %a1) {
; CHECK-LABEL: stack_fold_unpckhpd_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:    vunpckhpd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = ymm0[1],mem[1],ymm0[3],mem[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,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <4 x double> %a0, <4 x double> %a1, <4 x i32> <i32 1, i32 5, i32 3, i32 7>
  ; fadd forces execution domain
  %3 = fadd <4 x double> %2, <double 0x0, double 0x0, double 0x0, double 0x0>
  ret <4 x double> %3
}

define <4 x float> @stack_fold_unpckhps(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_unpckhps:
; 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:    vunpckhps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    # xmm0 = xmm0[2],mem[2],xmm0[3],mem[3]
; CHECK-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vaddps %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 float> %a0, <4 x float> %a1, <4 x i32> <i32 2, i32 6, i32 3, i32 7>
  ; fadd forces execution domain
  %3 = fadd <4 x float> %2, <float 0x0, float 0x0, float 0x0, float 0x0>
  ret <4 x float> %3
}

define <8 x float> @stack_fold_unpckhps_ymm(<8 x float> %a0, <8 x float> %a1) {
; CHECK-LABEL: stack_fold_unpckhps_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:    vunpckhps {{[-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:    vxorps %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vaddps %ymm1, %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 float> %a0, <8 x float> %a1, <8 x i32> <i32 2, i32 10, i32 3, i32 11, i32 6, i32 14, i32 7, i32 15>
  ; fadd forces execution domain
  %3 = fadd <8 x float> %2, <float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0>
  ret <8 x float> %3
}

define <2 x double> @stack_fold_unpcklpd(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_unpcklpd:
; 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:    vunpcklpd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    # xmm0 = xmm0[0],mem[0]
; CHECK-NEXT:    vxorpd %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vaddpd %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 <2 x double> %a0, <2 x double> %a1, <2 x i32> <i32 0, i32 2>
  ; fadd forces execution domain
  %3 = fadd <2 x double> %2, <double 0x0, double 0x0>
  ret <2 x double> %3
}

define <4 x double> @stack_fold_unpcklpd_ymm(<4 x double> %a0, <4 x double> %a1) {
; CHECK-LABEL: stack_fold_unpcklpd_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:    vunpcklpd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    # ymm0 = ymm0[0],mem[0],ymm0[2],mem[2]
; 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,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = shufflevector <4 x double> %a0, <4 x double> %a1, <4 x i32> <i32 0, i32 4, i32 2, i32 6>
  ; fadd forces execution domain
  %3 = fadd <4 x double> %2, <double 0x0, double 0x0, double 0x0, double 0x0>
  ret <4 x double> %3
}

define <4 x float> @stack_fold_unpcklps(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_unpcklps:
; 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:    vunpcklps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    # xmm0 = xmm0[0],mem[0],xmm0[1],mem[1]
; CHECK-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vaddps %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 float> %a0, <4 x float> %a1, <4 x i32> <i32 0, i32 4, i32 1, i32 5>
  ; fadd forces execution domain
  %3 = fadd <4 x float> %2, <float 0x0, float 0x0, float 0x0, float 0x0>
  ret <4 x float> %3
}

define <8 x float> @stack_fold_unpcklps_ymm(<8 x float> %a0, <8 x float> %a1) {
; CHECK-LABEL: stack_fold_unpcklps_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:    vunpcklps {{[-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:    vxorps %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vaddps %ymm1, %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 float> %a0, <8 x float> %a1, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 4, i32 12, i32 5, i32 13>
  ; fadd forces execution domain
  %3 = fadd <8 x float> %2, <float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0>
  ret <8 x float> %3
}

define <2 x double> @stack_fold_xorpd(<2 x double> %a0, <2 x double> %a1) {
; CHECK-LABEL: stack_fold_xorpd:
; 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:    vxorpd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    vxorpd %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vaddpd %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 = bitcast <2 x double> %a0 to <2 x i64>
  %3 = bitcast <2 x double> %a1 to <2 x i64>
  %4 = xor <2 x i64> %2, %3
  %5 = bitcast <2 x i64> %4 to <2 x double>
  ; fadd forces execution domain
  %6 = fadd <2 x double> %5, <double 0x0, double 0x0>
  ret <2 x double> %6
}

define <4 x double> @stack_fold_xorpd_ymm(<4 x double> %a0, <4 x double> %a1) {
; CHECK-LABEL: stack_fold_xorpd_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:    vxorpd {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; 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,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
  %2 = bitcast <4 x double> %a0 to <4 x i64>
  %3 = bitcast <4 x double> %a1 to <4 x i64>
  %4 = xor <4 x i64> %2, %3
  %5 = bitcast <4 x i64> %4 to <4 x double>
  ; fadd forces execution domain
  %6 = fadd <4 x double> %5, <double 0x0, double 0x0, double 0x0, double 0x0>
  ret <4 x double> %6
}

define <4 x float> @stack_fold_xorps(<4 x float> %a0, <4 x float> %a1) {
; CHECK-LABEL: stack_fold_xorps:
; 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:    vxorps {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 16-byte Folded Reload
; CHECK-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vaddps %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 = bitcast <4 x float> %a0 to <2 x i64>
  %3 = bitcast <4 x float> %a1 to <2 x i64>
  %4 = xor <2 x i64> %2, %3
  %5 = bitcast <2 x i64> %4 to <4 x float>
  ; fadd forces execution domain
  %6 = fadd <4 x float> %5, <float 0x0, float 0x0, float 0x0, float 0x0>
  ret <4 x float> %6
}

define <8 x float> @stack_fold_xorps_ymm(<8 x float> %a0, <8 x float> %a1) {
; CHECK-LABEL: stack_fold_xorps_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:    vxorps {{[-0-9]+}}(%r{{[sb]}}p), %ymm0, %ymm0 # 32-byte Folded Reload
; CHECK-NEXT:    vxorps %xmm1, %xmm1, %xmm1
; CHECK-NEXT:    vaddps %ymm1, %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 = bitcast <8 x float> %a0 to <4 x i64>
  %3 = bitcast <8 x float> %a1 to <4 x i64>
  %4 = xor <4 x i64> %2, %3
  %5 = bitcast <4 x i64> %4 to <8 x float>
  ; fadd forces execution domain
  %6 = fadd <8 x float> %5, <float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0, float 0x0>
  ret <8 x float> %6
}

attributes #0 = { "unsafe-fp-math"="false" }
attributes #1 = { "unsafe-fp-math"="true" }