; RUN: llc -O3 -disable-peephole -mtriple=x86_64-unknown-unknown -mattr=+avx512vl < %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:       vaddpd {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}}, {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{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:       vaddpd {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%ymm[0-9][0-9]*}} {{.*#+}} 32-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{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:       vaddps {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}}, {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{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:       vaddps {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%ymm[0-9][0-9]*}} {{.*#+}} 32-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
  %2 = fadd <8 x float> %a0, %a1
  ret <8 x float> %2
}

define <2 x double> @stack_fold_andnpd(<2 x double> %a0, <2 x double> %a1) {
  ;CHECK-LABEL: stack_fold_andnpd
  ;CHECK:       vpandnq {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}}, {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
  %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:       vpandnq {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%ymm[0-9][0-9]*}} {{.*#+}} 32-byte Folded Reload
  %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:       vpandnq {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}}, {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
  %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:       vpandnq {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%ymm[0-9][0-9]*}} {{.*#+}} 32-byte Folded Reload
  %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:       vpandq {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}}, {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{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:       vpandq {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%ymm[0-9][0-9]*}} {{.*#+}} 32-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{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:       vpandq {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}}, {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
  %2 = bitcast <4 x float> %a0 to <4 x i32>
  %3 = bitcast <4 x float> %a1 to <4 x i32>
  %4 = and <4 x i32> %2, %3
  %5 = bitcast <4 x i32> %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:       vpandq {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%ymm[0-9][0-9]*}} {{.*#+}} 32-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
  %2 = bitcast <8 x float> %a0 to <8 x i32>
  %3 = bitcast <8 x float> %a1 to <8 x i32>
  %4 = and <8 x i32> %2, %3
  %5 = bitcast <8 x i32> %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 i8 @stack_fold_cmppd(<2 x double> %a0, <2 x double> %a1) {
  ;CHECK-LABEL: stack_fold_cmppd
  ;CHECK:       vcmpeqpd {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}}, {{%k[0-9]}} {{.*#+}} 16-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
  %res = call i8 @llvm.x86.avx512.mask.cmp.pd.128(<2 x double> %a0, <2 x double> %a1, i32 0, i8 -1)
  ret i8 %res
}
declare i8 @llvm.x86.avx512.mask.cmp.pd.128(<2 x double> , <2 x double> , i32, i8)

define i8 @stack_fold_cmppd_ymm(<4 x double> %a0, <4 x double> %a1) {
  ;CHECK-LABEL: stack_fold_cmppd_ymm
  ;CHECK:       vcmpeqpd {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%k[0-9]}} {{.*#+}} 32-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
  %res = call i8 @llvm.x86.avx512.mask.cmp.pd.256(<4 x double> %a0, <4 x double> %a1, i32 0, i8 -1)
  ret i8 %res
}
declare i8 @llvm.x86.avx512.mask.cmp.pd.256(<4 x double> , <4 x double> , i32, i8)

define i8 @stack_fold_cmpps(<4 x float> %a0, <4 x float> %a1) {
  ;CHECK-LABEL: stack_fold_cmpps
  ;CHECK:       vcmpeqps {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}}, {{%k[0-9]*}} {{.*#+}} 16-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
  %res = call i8 @llvm.x86.avx512.mask.cmp.ps.128(<4 x float> %a0, <4 x float> %a1, i32 0, i8 -1)
  ret i8 %res
}
declare i8 @llvm.x86.avx512.mask.cmp.ps.128(<4 x float> , <4 x float> , i32, i8)

define i8 @stack_fold_cmpps_ymm(<8 x float> %a0, <8 x float> %a1) {
  ;CHECK-LABEL: stack_fold_cmpps_ymm
  ;CHECK:       vcmpeqps {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%k[0-9]*}} {{.*#+}} 32-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
  %res = call i8 @llvm.x86.avx512.mask.cmp.ps.256(<8 x float> %a0, <8 x float> %a1, i32 0, i8 -1)
  ret i8 %res
}
declare i8 @llvm.x86.avx512.mask.cmp.ps.256(<8 x float> , <8 x float> , i32, i8)

define <2 x double> @stack_fold_maxpd(<2 x double> %a0, <2 x double> %a1) #0 {
  ;CHECK-LABEL: stack_fold_maxpd
  ;CHECK:       vmaxpd {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}}, {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{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:       vmaxpd {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}}, {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{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:       vmaxpd {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%ymm[0-9][0-9]*}} {{.*#+}} 32-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{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:       vmaxpd {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%ymm[0-9][0-9]*}} {{.*#+}} 32-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{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:       vmaxps {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}}, {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{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:       vmaxps {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}}, {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{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:       vmaxps {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%ymm[0-9][0-9]*}} {{.*#+}} 32-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{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:       vmaxps {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%ymm[0-9][0-9]*}} {{.*#+}} 32-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{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 <4 x float> @stack_fold_minps(<4 x float> %a0, <4 x float> %a1) #0 {
  ;CHECK-LABEL: stack_fold_minps
  ;CHECK:       vminps {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}}, {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
  %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:       vminps {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}}, {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
  %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:       vminps {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%ymm[0-9][0-9]*}} {{.*#+}} 32-byte Folded Reload
  %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:       vminps {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%ymm[0-9][0-9]*}} {{.*#+}} 32-byte Folded Reload
  %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 <2 x double> @stack_fold_orpd(<2 x double> %a0, <2 x double> %a1) {
  ;CHECK-LABEL: stack_fold_orpd
  ;CHECK:       vporq {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}}, {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{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:       vporq {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%ymm[0-9][0-9]*}} {{.*#+}} 32-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{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:       vporq {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}}, {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
  %2 = bitcast <4 x float> %a0 to <4 x i32>
  %3 = bitcast <4 x float> %a1 to <4 x i32>
  %4 = or <4 x i32> %2, %3
  %5 = bitcast <4 x i32> %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:       vporq {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%ymm[0-9][0-9]*}} {{.*#+}} 32-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
  %2 = bitcast <8 x float> %a0 to <8 x i32>
  %3 = bitcast <8 x float> %a1 to <8 x i32>
  %4 = or <8 x i32> %2, %3
  %5 = bitcast <8 x i32> %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_subpd(<2 x double> %a0, <2 x double> %a1) {
  ;CHECK-LABEL: stack_fold_subpd
  ;CHECK:       vsubpd {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}}, {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{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:       vsubpd {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%ymm[0-9][0-9]*}} {{.*#+}} 32-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{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:       vsubps {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}}, {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{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:       vsubps {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%ymm[0-9][0-9]*}} {{.*#+}} 32-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
  %2 = fsub <8 x float> %a0, %a1
  ret <8 x float> %2
}

define <2 x double> @stack_fold_xorpd(<2 x double> %a0, <2 x double> %a1) {
  ;CHECK-LABEL: stack_fold_xorpd
  ;CHECK:       vpxorq {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}}, {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{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:       vpxorq {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%ymm[0-9][0-9]*}} {{.*#+}} 32-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{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:       vpxorq {{-?[0-9]*}}(%rsp), {{%xmm[0-9][0-9]*}}, {{%xmm[0-9][0-9]*}} {{.*#+}} 16-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
  %2 = bitcast <4 x float> %a0 to <4 x i32>
  %3 = bitcast <4 x float> %a1 to <4 x i32>
  %4 = xor <4 x i32> %2, %3
  %5 = bitcast <4 x i32> %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:       vpxorq {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%ymm[0-9][0-9]*}} {{.*#+}} 32-byte Folded Reload
  %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{xmm16},~{xmm17},~{xmm18},~{xmm19},~{xmm20},~{xmm21},~{xmm22},~{xmm23},~{xmm24},~{xmm25},~{xmm26},~{xmm27},~{xmm28},~{xmm29},~{xmm30},~{xmm31},~{flags}"()
  %2 = bitcast <8 x float> %a0 to <8 x i32>
  %3 = bitcast <8 x float> %a1 to <8 x i32>
  %4 = xor <8 x i32> %2, %3
  %5 = bitcast <8 x i32> %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" }