; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=x86_64-apple-darwin -mcpu=x86-64 -mattr=+avx2 < %s | FileCheck %s --check-prefix=AVX2
; RUN: llc -mtriple=x86_64-apple-darwin -mcpu=x86-64 -mattr=+avx512f,+avx512vl,+avx512bw < %s | FileCheck %s --check-prefix=AVX512

define i16 @v16i16(<16 x i16> %a, <16 x i16> %b) {
; AVX2-LABEL: v16i16:
; AVX2:       ## BB#0:
; AVX2-NEXT:    vpcmpgtw %ymm1, %ymm0, %ymm0
; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
; AVX2-NEXT:    vpacksswb %xmm1, %xmm0, %xmm0
; AVX2-NEXT:    vpmovmskb %xmm0, %eax
; AVX2-NEXT:    ## kill: %AX<def> %AX<kill> %EAX<kill>
; AVX2-NEXT:    vzeroupper
; AVX2-NEXT:    retq
;
; AVX512-LABEL: v16i16:
; AVX512:       ## BB#0:
; AVX512-NEXT:    vpcmpgtw %ymm1, %ymm0, %k0
; AVX512-NEXT:    kmovd %k0, %eax
; AVX512-NEXT:    ## kill: %AX<def> %AX<kill> %EAX<kill>
; AVX512-NEXT:    vzeroupper
; AVX512-NEXT:    retq
  %x = icmp sgt <16 x i16> %a, %b
  %res = bitcast <16 x i1> %x to i16
  ret i16 %res
}

define i8 @v8i32(<8 x i32> %a, <8 x i32> %b) {
; AVX2-LABEL: v8i32:
; AVX2:       ## BB#0:
; AVX2-NEXT:    vpcmpgtd %ymm1, %ymm0, %ymm0
; AVX2-NEXT:    vmovmskps %ymm0, %eax
; AVX2-NEXT:    ## kill: %AL<def> %AL<kill> %EAX<kill>
; AVX2-NEXT:    vzeroupper
; AVX2-NEXT:    retq
;
; AVX512-LABEL: v8i32:
; AVX512:       ## BB#0:
; AVX512-NEXT:    vpcmpgtd %ymm1, %ymm0, %k0
; AVX512-NEXT:    kmovd %k0, %eax
; AVX512-NEXT:    ## kill: %AL<def> %AL<kill> %EAX<kill>
; AVX512-NEXT:    vzeroupper
; AVX512-NEXT:    retq
  %x = icmp sgt <8 x i32> %a, %b
  %res = bitcast <8 x i1> %x to i8
  ret i8 %res
}

define i8 @v8f32(<8 x float> %a, <8 x float> %b) {
; AVX2-LABEL: v8f32:
; AVX2:       ## BB#0:
; AVX2-NEXT:    vcmpltps %ymm0, %ymm1, %ymm0
; AVX2-NEXT:    vmovmskps %ymm0, %eax
; AVX2-NEXT:    ## kill: %AL<def> %AL<kill> %EAX<kill>
; AVX2-NEXT:    vzeroupper
; AVX2-NEXT:    retq
;
; AVX512-LABEL: v8f32:
; AVX512:       ## BB#0:
; AVX512-NEXT:    vcmpltps %ymm0, %ymm1, %k0
; AVX512-NEXT:    kmovd %k0, %eax
; AVX512-NEXT:    ## kill: %AL<def> %AL<kill> %EAX<kill>
; AVX512-NEXT:    vzeroupper
; AVX512-NEXT:    retq
  %x = fcmp ogt <8 x float> %a, %b
  %res = bitcast <8 x i1> %x to i8
  ret i8 %res
}

define i32 @v32i8(<32 x i8> %a, <32 x i8> %b) {
; AVX2-LABEL: v32i8:
; AVX2:       ## BB#0:
; AVX2-NEXT:    vpcmpgtb %ymm1, %ymm0, %ymm0
; AVX2-NEXT:    vpmovmskb %ymm0, %eax
; AVX2-NEXT:    vzeroupper
; AVX2-NEXT:    retq
;
; AVX512-LABEL: v32i8:
; AVX512:       ## BB#0:
; AVX512-NEXT:    vpcmpgtb %ymm1, %ymm0, %k0
; AVX512-NEXT:    kmovd %k0, %eax
; AVX512-NEXT:    vzeroupper
; AVX512-NEXT:    retq
  %x = icmp sgt <32 x i8> %a, %b
  %res = bitcast <32 x i1> %x to i32
  ret i32 %res
}

define i4 @v4i64(<4 x i64> %a, <4 x i64> %b) {
; AVX2-LABEL: v4i64:
; AVX2:       ## BB#0:
; AVX2-NEXT:    vpcmpgtq %ymm1, %ymm0, %ymm0
; AVX2-NEXT:    vmovmskpd %ymm0, %eax
; AVX2-NEXT:    ## kill: %AL<def> %AL<kill> %EAX<kill>
; AVX2-NEXT:    vzeroupper
; AVX2-NEXT:    retq
;
; AVX512-LABEL: v4i64:
; AVX512:       ## BB#0:
; AVX512-NEXT:    vpcmpgtq %ymm1, %ymm0, %k0
; AVX512-NEXT:    kmovd %k0, %eax
; AVX512-NEXT:    movb %al, -{{[0-9]+}}(%rsp)
; AVX512-NEXT:    movb -{{[0-9]+}}(%rsp), %al
; AVX512-NEXT:    vzeroupper
; AVX512-NEXT:    retq
  %x = icmp sgt <4 x i64> %a, %b
  %res = bitcast <4 x i1> %x to i4
  ret i4 %res
}

define i4 @v4f64(<4 x double> %a, <4 x double> %b) {
; AVX2-LABEL: v4f64:
; AVX2:       ## BB#0:
; AVX2-NEXT:    vcmpltpd %ymm0, %ymm1, %ymm0
; AVX2-NEXT:    vmovmskpd %ymm0, %eax
; AVX2-NEXT:    ## kill: %AL<def> %AL<kill> %EAX<kill>
; AVX2-NEXT:    vzeroupper
; AVX2-NEXT:    retq
;
; AVX512-LABEL: v4f64:
; AVX512:       ## BB#0:
; AVX512-NEXT:    vcmpltpd %ymm0, %ymm1, %k0
; AVX512-NEXT:    kmovd %k0, %eax
; AVX512-NEXT:    movb %al, -{{[0-9]+}}(%rsp)
; AVX512-NEXT:    movb -{{[0-9]+}}(%rsp), %al
; AVX512-NEXT:    vzeroupper
; AVX512-NEXT:    retq
  %x = fcmp ogt <4 x double> %a, %b
  %res = bitcast <4 x i1> %x to i4
  ret i4 %res
}