llvm/test/CodeGen/X86/combine-umax.ll

; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+sse2 | FileCheck %s --check-prefixes=SSE,SSE2
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+sse4.1 | FileCheck %s --check-prefixes=SSE,SSE41
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+sse4.2 | FileCheck %s --check-prefixes=SSE,SSE42
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx | FileCheck %s --check-prefixes=AVX,AVX1OR2
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX1OR2
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512f | FileCheck %s --check-prefixes=AVX,AVX512F
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512bw | FileCheck %s --check-prefixes=AVX,AVX512BW

define <8 x i16> @test_v8i16_nosignbit(<8 x i16> %a, <8 x i16> %b) {
; SSE2-LABEL: test_v8i16_nosignbit:
; SSE2:       # %bb.0:
; SSE2-NEXT:    pand {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; SSE2-NEXT:    psrlw $1, %xmm1
; SSE2-NEXT:    pminsw %xmm1, %xmm0
; SSE2-NEXT:    retq
;
; SSE41-LABEL: test_v8i16_nosignbit:
; SSE41:       # %bb.0:
; SSE41-NEXT:    pand {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; SSE41-NEXT:    psrlw $1, %xmm1
; SSE41-NEXT:    pminuw %xmm1, %xmm0
; SSE41-NEXT:    retq
;
; SSE42-LABEL: test_v8i16_nosignbit:
; SSE42:       # %bb.0:
; SSE42-NEXT:    pand {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; SSE42-NEXT:    psrlw $1, %xmm1
; SSE42-NEXT:    pminuw %xmm1, %xmm0
; SSE42-NEXT:    retq
;
; AVX-LABEL: test_v8i16_nosignbit:
; AVX:       # %bb.0:
; AVX-NEXT:    vpand {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
; AVX-NEXT:    vpsrlw $1, %xmm1, %xmm1
; AVX-NEXT:    vpminuw %xmm1, %xmm0, %xmm0
; AVX-NEXT:    retq
  %1 = and <8 x i16> %a, <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
  %2 = lshr <8 x i16> %b, <i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1, i16 1>
  %3 = icmp ult <8 x i16> %1, %2
  %4 = select <8 x i1> %3, <8 x i16> %1, <8 x i16> %2
  ret <8 x i16> %4
}

define <16 x i8> @test_v16i8_reassociation(<16 x i8> %a) {
; SSE-LABEL: test_v16i8_reassociation:
; SSE:       # %bb.0:
; SSE-NEXT:    pmaxub {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; SSE-NEXT:    retq
;
; AVX-LABEL: test_v16i8_reassociation:
; AVX:       # %bb.0:
; AVX-NEXT:    vpmaxub {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
; AVX-NEXT:    retq
  %1 = call <16 x i8> @llvm.umax.v16i8(<16 x i8> %a, <16 x i8> <i8 0, i8 1, i8 2, i8 3, i8 4, i8 5, i8 6, i8 7, i8 8, i8 9, i8 10, i8 11, i8 12, i8 13, i8 14, i8 15>)
  %2 = call <16 x i8> @llvm.umax.v16i8(<16 x i8> %1, <16 x i8> <i8 0, i8 1, i8 2, i8 3, i8 4, i8 5, i8 6, i8 7, i8 8, i8 9, i8 10, i8 11, i8 12, i8 13, i8 14, i8 15>)
  ret <16 x i8> %2
}

define <16 x i8> @test_v16i8_demandedbits(<16 x i8> %x, <16 x i8> %y, <16 x i8> %a, <16 x i8> %b) {
; SSE2-LABEL: test_v16i8_demandedbits:
; SSE2:       # %bb.0:
; SSE2-NEXT:    pmaxub %xmm1, %xmm0
; SSE2-NEXT:    pxor %xmm1, %xmm1
; SSE2-NEXT:    pcmpgtb %xmm0, %xmm1
; SSE2-NEXT:    pand %xmm1, %xmm3
; SSE2-NEXT:    pandn %xmm2, %xmm1
; SSE2-NEXT:    por %xmm3, %xmm1
; SSE2-NEXT:    movdqa %xmm1, %xmm0
; SSE2-NEXT:    retq
;
; SSE41-LABEL: test_v16i8_demandedbits:
; SSE41:       # %bb.0:
; SSE41-NEXT:    orps %xmm1, %xmm0
; SSE41-NEXT:    pblendvb %xmm0, %xmm3, %xmm2
; SSE41-NEXT:    movdqa %xmm2, %xmm0
; SSE41-NEXT:    retq
;
; SSE42-LABEL: test_v16i8_demandedbits:
; SSE42:       # %bb.0:
; SSE42-NEXT:    orps %xmm1, %xmm0
; SSE42-NEXT:    pblendvb %xmm0, %xmm3, %xmm2
; SSE42-NEXT:    movdqa %xmm2, %xmm0
; SSE42-NEXT:    retq
;
; AVX1OR2-LABEL: test_v16i8_demandedbits:
; AVX1OR2:       # %bb.0:
; AVX1OR2-NEXT:    vpor %xmm1, %xmm0, %xmm0
; AVX1OR2-NEXT:    vpblendvb %xmm0, %xmm3, %xmm2, %xmm0
; AVX1OR2-NEXT:    retq
;
; AVX512F-LABEL: test_v16i8_demandedbits:
; AVX512F:       # %bb.0:
; AVX512F-NEXT:    vpor %xmm1, %xmm0, %xmm0
; AVX512F-NEXT:    vpblendvb %xmm0, %xmm3, %xmm2, %xmm0
; AVX512F-NEXT:    retq
;
; AVX512BW-LABEL: test_v16i8_demandedbits:
; AVX512BW:       # %bb.0:
; AVX512BW-NEXT:    # kill: def $xmm3 killed $xmm3 def $zmm3
; AVX512BW-NEXT:    # kill: def $xmm2 killed $xmm2 def $zmm2
; AVX512BW-NEXT:    vpmaxub %xmm1, %xmm0, %xmm0
; AVX512BW-NEXT:    vpxor %xmm1, %xmm1, %xmm1
; AVX512BW-NEXT:    vpcmpnltb %zmm1, %zmm0, %k1
; AVX512BW-NEXT:    vpblendmb %zmm2, %zmm3, %zmm0 {%k1}
; AVX512BW-NEXT:    # kill: def $xmm0 killed $xmm0 killed $zmm0
; AVX512BW-NEXT:    vzeroupper
; AVX512BW-NEXT:    retq
  %umax = tail call <16 x i8> @llvm.umax.v16i8(<16 x i8> %x, <16 x i8> %y)
  %cmp = icmp sge <16 x i8> %umax, zeroinitializer
  %res = select <16 x i1> %cmp, <16 x i8> %a, <16 x i8> %b
  ret <16 x i8> %res
}

declare <16 x i8> @llvm.umax.v16i8(<16 x i8> %x, <16 x i8> %y)