llvm-6502/test/CodeGen/X86/avx512-intrinsics.ll
Elena Demikhovsky 3bf51cf302 AVX-512: Removed "z" suffix from AVX-512 instructions, since it is incompatible with GCC.
I moved a test from avx512-vbroadcast-crash.ll to avx512-vbroadcast.ll
I defined HasAVX512 predicate as AssemblerPredicate. It means that you should invoke llvm-mc with "-mcpu=knl" to get encoding for AVX-512 instructions. I need this to let AsmMatcher to set different encoding for AVX and AVX-512 instructions that have the same mnemonic and operands (all scalar instructions).


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@197041 91177308-0d34-0410-b5e6-96231b3b80d8
2013-12-11 14:31:04 +00:00

423 lines
17 KiB
LLVM

; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=knl --show-mc-encoding| FileCheck %s
declare i32 @llvm.x86.avx512.kortestz.w(i16, i16) nounwind readnone
; CHECK-LABEL: test_kortestz
; CHECK: kortestw
; CHECK: sete
define i32 @test_kortestz(i16 %a0, i16 %a1) {
%res = call i32 @llvm.x86.avx512.kortestz.w(i16 %a0, i16 %a1)
ret i32 %res
}
declare i32 @llvm.x86.avx512.kortestc.w(i16, i16) nounwind readnone
; CHECK-LABEL: test_kortestc
; CHECK: kortestw
; CHECK: sbbl
define i32 @test_kortestc(i16 %a0, i16 %a1) {
%res = call i32 @llvm.x86.avx512.kortestc.w(i16 %a0, i16 %a1)
ret i32 %res
}
declare i16 @llvm.x86.avx512.kand.w(i16, i16) nounwind readnone
; CHECK-LABEL: test_kand
; CHECK: kandw
; CHECK: kandw
define i16 @test_kand(i16 %a0, i16 %a1) {
%t1 = call i16 @llvm.x86.avx512.kand.w(i16 %a0, i16 8)
%t2 = call i16 @llvm.x86.avx512.kand.w(i16 %t1, i16 %a1)
ret i16 %t2
}
declare i16 @llvm.x86.avx512.knot.w(i16) nounwind readnone
; CHECK-LABEL: test_knot
; CHECK: knotw
define i16 @test_knot(i16 %a0) {
%res = call i16 @llvm.x86.avx512.knot.w(i16 %a0)
ret i16 %res
}
declare i16 @llvm.x86.avx512.kunpck.bw(i16, i16) nounwind readnone
; CHECK-LABEL: unpckbw_test
; CHECK: kunpckbw
; CHECK:ret
define i16 @unpckbw_test(i16 %a0, i16 %a1) {
%res = call i16 @llvm.x86.avx512.kunpck.bw(i16 %a0, i16 %a1)
ret i16 %res
}
define <16 x float> @test_rcp_ps_512(<16 x float> %a0) {
; CHECK: vrcp14ps
%res = call <16 x float> @llvm.x86.avx512.rcp14.ps.512(<16 x float> %a0) ; <<16 x float>> [#uses=1]
ret <16 x float> %res
}
declare <16 x float> @llvm.x86.avx512.rcp14.ps.512(<16 x float>) nounwind readnone
define <8 x double> @test_rcp_pd_512(<8 x double> %a0) {
; CHECK: vrcp14pd
%res = call <8 x double> @llvm.x86.avx512.rcp14.pd.512(<8 x double> %a0) ; <<8 x double>> [#uses=1]
ret <8 x double> %res
}
declare <8 x double> @llvm.x86.avx512.rcp14.pd.512(<8 x double>) nounwind readnone
define <16 x float> @test_rcp28_ps_512(<16 x float> %a0) {
; CHECK: vrcp28ps
%res = call <16 x float> @llvm.x86.avx512.rcp28.ps.512(<16 x float> %a0) ; <<16 x float>> [#uses=1]
ret <16 x float> %res
}
declare <16 x float> @llvm.x86.avx512.rcp28.ps.512(<16 x float>) nounwind readnone
define <8 x double> @test_rcp28_pd_512(<8 x double> %a0) {
; CHECK: vrcp28pd
%res = call <8 x double> @llvm.x86.avx512.rcp28.pd.512(<8 x double> %a0) ; <<8 x double>> [#uses=1]
ret <8 x double> %res
}
declare <8 x double> @llvm.x86.avx512.rcp28.pd.512(<8 x double>) nounwind readnone
define <8 x double> @test_rndscale_pd_512(<8 x double> %a0) {
; CHECK: vrndscale
%res = call <8 x double> @llvm.x86.avx512.rndscale.pd.512(<8 x double> %a0, i32 7) ; <<8 x double>> [#uses=1]
ret <8 x double> %res
}
declare <8 x double> @llvm.x86.avx512.rndscale.pd.512(<8 x double>, i32) nounwind readnone
define <16 x float> @test_rndscale_ps_512(<16 x float> %a0) {
; CHECK: vrndscale
%res = call <16 x float> @llvm.x86.avx512.rndscale.ps.512(<16 x float> %a0, i32 7) ; <<16 x float>> [#uses=1]
ret <16 x float> %res
}
declare <16 x float> @llvm.x86.avx512.rndscale.ps.512(<16 x float>, i32) nounwind readnone
define <16 x float> @test_rsqrt_ps_512(<16 x float> %a0) {
; CHECK: vrsqrt14ps
%res = call <16 x float> @llvm.x86.avx512.rsqrt14.ps.512(<16 x float> %a0) ; <<16 x float>> [#uses=1]
ret <16 x float> %res
}
declare <16 x float> @llvm.x86.avx512.rsqrt14.ps.512(<16 x float>) nounwind readnone
define <16 x float> @test_rsqrt28_ps_512(<16 x float> %a0) {
; CHECK: vrsqrt28ps
%res = call <16 x float> @llvm.x86.avx512.rsqrt28.ps.512(<16 x float> %a0) ; <<16 x float>> [#uses=1]
ret <16 x float> %res
}
declare <16 x float> @llvm.x86.avx512.rsqrt28.ps.512(<16 x float>) nounwind readnone
define <4 x float> @test_rsqrt14_ss(<4 x float> %a0) {
; CHECK: vrsqrt14ss
%res = call <4 x float> @llvm.x86.avx512.rsqrt14.ss(<4 x float> %a0) ; <<4 x float>> [#uses=1]
ret <4 x float> %res
}
declare <4 x float> @llvm.x86.avx512.rsqrt14.ss(<4 x float>) nounwind readnone
define <4 x float> @test_rsqrt28_ss(<4 x float> %a0) {
; CHECK: vrsqrt28ss
%res = call <4 x float> @llvm.x86.avx512.rsqrt28.ss(<4 x float> %a0) ; <<4 x float>> [#uses=1]
ret <4 x float> %res
}
declare <4 x float> @llvm.x86.avx512.rsqrt28.ss(<4 x float>) nounwind readnone
define <4 x float> @test_rcp14_ss(<4 x float> %a0) {
; CHECK: vrcp14ss
%res = call <4 x float> @llvm.x86.avx512.rcp14.ss(<4 x float> %a0) ; <<4 x float>> [#uses=1]
ret <4 x float> %res
}
declare <4 x float> @llvm.x86.avx512.rcp14.ss(<4 x float>) nounwind readnone
define <4 x float> @test_rcp28_ss(<4 x float> %a0) {
; CHECK: vrcp28ss
%res = call <4 x float> @llvm.x86.avx512.rcp28.ss(<4 x float> %a0) ; <<4 x float>> [#uses=1]
ret <4 x float> %res
}
declare <4 x float> @llvm.x86.avx512.rcp28.ss(<4 x float>) nounwind readnone
define <8 x double> @test_sqrt_pd_512(<8 x double> %a0) {
; CHECK: vsqrtpd
%res = call <8 x double> @llvm.x86.avx512.sqrt.pd.512(<8 x double> %a0) ; <<8 x double>> [#uses=1]
ret <8 x double> %res
}
declare <8 x double> @llvm.x86.avx512.sqrt.pd.512(<8 x double>) nounwind readnone
define <16 x float> @test_sqrt_ps_512(<16 x float> %a0) {
; CHECK: vsqrtps
%res = call <16 x float> @llvm.x86.avx512.sqrt.ps.512(<16 x float> %a0) ; <<16 x float>> [#uses=1]
ret <16 x float> %res
}
declare <16 x float> @llvm.x86.avx512.sqrt.ps.512(<16 x float>) nounwind readnone
define <4 x float> @test_sqrt_ss(<4 x float> %a0, <4 x float> %a1) {
; CHECK: vsqrtss {{.*}}encoding: [0x62
%res = call <4 x float> @llvm.x86.avx512.sqrt.ss(<4 x float> %a0, <4 x float> %a1) ; <<4 x float>> [#uses=1]
ret <4 x float> %res
}
declare <4 x float> @llvm.x86.avx512.sqrt.ss(<4 x float>, <4 x float>) nounwind readnone
define <2 x double> @test_sqrt_sd(<2 x double> %a0, <2 x double> %a1) {
; CHECK: vsqrtsd {{.*}}encoding: [0x62
%res = call <2 x double> @llvm.x86.avx512.sqrt.sd(<2 x double> %a0, <2 x double> %a1) ; <<2 x double>> [#uses=1]
ret <2 x double> %res
}
declare <2 x double> @llvm.x86.avx512.sqrt.sd(<2 x double>, <2 x double>) nounwind readnone
define i64 @test_x86_sse2_cvtsd2si64(<2 x double> %a0) {
; CHECK: vcvtsd2si {{.*}}encoding: [0x62
%res = call i64 @llvm.x86.sse2.cvtsd2si64(<2 x double> %a0) ; <i64> [#uses=1]
ret i64 %res
}
declare i64 @llvm.x86.sse2.cvtsd2si64(<2 x double>) nounwind readnone
define <2 x double> @test_x86_sse2_cvtsi642sd(<2 x double> %a0, i64 %a1) {
; CHECK: vcvtsi2sdq {{.*}}encoding: [0x62
%res = call <2 x double> @llvm.x86.sse2.cvtsi642sd(<2 x double> %a0, i64 %a1) ; <<2 x double>> [#uses=1]
ret <2 x double> %res
}
declare <2 x double> @llvm.x86.sse2.cvtsi642sd(<2 x double>, i64) nounwind readnone
define <2 x double> @test_x86_avx512_cvtusi642sd(<2 x double> %a0, i64 %a1) {
; CHECK: vcvtusi2sdq {{.*}}encoding: [0x62
%res = call <2 x double> @llvm.x86.avx512.cvtusi642sd(<2 x double> %a0, i64 %a1) ; <<2 x double>> [#uses=1]
ret <2 x double> %res
}
declare <2 x double> @llvm.x86.avx512.cvtusi642sd(<2 x double>, i64) nounwind readnone
define i64 @test_x86_sse2_cvttsd2si64(<2 x double> %a0) {
; CHECK: vcvttsd2si {{.*}}encoding: [0x62
%res = call i64 @llvm.x86.sse2.cvttsd2si64(<2 x double> %a0) ; <i64> [#uses=1]
ret i64 %res
}
declare i64 @llvm.x86.sse2.cvttsd2si64(<2 x double>) nounwind readnone
define i64 @test_x86_sse_cvtss2si64(<4 x float> %a0) {
; CHECK: vcvtss2si {{.*}}encoding: [0x62
%res = call i64 @llvm.x86.sse.cvtss2si64(<4 x float> %a0) ; <i64> [#uses=1]
ret i64 %res
}
declare i64 @llvm.x86.sse.cvtss2si64(<4 x float>) nounwind readnone
define <4 x float> @test_x86_sse_cvtsi642ss(<4 x float> %a0, i64 %a1) {
; CHECK: vcvtsi2ssq {{.*}}encoding: [0x62
%res = call <4 x float> @llvm.x86.sse.cvtsi642ss(<4 x float> %a0, i64 %a1) ; <<4 x float>> [#uses=1]
ret <4 x float> %res
}
declare <4 x float> @llvm.x86.sse.cvtsi642ss(<4 x float>, i64) nounwind readnone
define i64 @test_x86_sse_cvttss2si64(<4 x float> %a0) {
; CHECK: vcvttss2si {{.*}}encoding: [0x62
%res = call i64 @llvm.x86.sse.cvttss2si64(<4 x float> %a0) ; <i64> [#uses=1]
ret i64 %res
}
declare i64 @llvm.x86.sse.cvttss2si64(<4 x float>) nounwind readnone
define i64 @test_x86_avx512_cvtsd2usi64(<2 x double> %a0) {
; CHECK: vcvtsd2usi {{.*}}encoding: [0x62
%res = call i64 @llvm.x86.avx512.cvtsd2usi64(<2 x double> %a0) ; <i64> [#uses=1]
ret i64 %res
}
declare i64 @llvm.x86.avx512.cvtsd2usi64(<2 x double>) nounwind readnone
define <16 x float> @test_x86_vcvtph2ps_512(<16 x i16> %a0) {
; CHECK: vcvtph2ps
%res = call <16 x float> @llvm.x86.avx512.vcvtph2ps.512(<16 x i16> %a0)
ret <16 x float> %res
}
declare <16 x float> @llvm.x86.avx512.vcvtph2ps.512(<16 x i16>) nounwind readonly
define <16 x i16> @test_x86_vcvtps2ph_256(<16 x float> %a0) {
; CHECK: vcvtps2ph
%res = call <16 x i16> @llvm.x86.avx512.vcvtps2ph.512(<16 x float> %a0, i32 0)
ret <16 x i16> %res
}
declare <16 x i16> @llvm.x86.avx512.vcvtps2ph.512(<16 x float>, i32) nounwind readonly
define <16 x float> @test_x86_vbroadcast_ss_512(i8* %a0) {
; CHECK: vbroadcastss
%res = call <16 x float> @llvm.x86.avx512.vbroadcast.ss.512(i8* %a0) ; <<16 x float>> [#uses=1]
ret <16 x float> %res
}
declare <16 x float> @llvm.x86.avx512.vbroadcast.ss.512(i8*) nounwind readonly
define <8 x double> @test_x86_vbroadcast_sd_512(i8* %a0) {
; CHECK: vbroadcastsd
%res = call <8 x double> @llvm.x86.avx512.vbroadcast.sd.512(i8* %a0) ; <<8 x double>> [#uses=1]
ret <8 x double> %res
}
declare <8 x double> @llvm.x86.avx512.vbroadcast.sd.512(i8*) nounwind readonly
define <16 x float> @test_x86_vbroadcast_ss_ps_512(<4 x float> %a0) {
; CHECK: vbroadcastss
%res = call <16 x float> @llvm.x86.avx512.vbroadcast.ss.ps.512(<4 x float> %a0) ; <<16 x float>> [#uses=1]
ret <16 x float> %res
}
declare <16 x float> @llvm.x86.avx512.vbroadcast.ss.ps.512(<4 x float>) nounwind readonly
define <8 x double> @test_x86_vbroadcast_sd_pd_512(<2 x double> %a0) {
; CHECK: vbroadcastsd
%res = call <8 x double> @llvm.x86.avx512.vbroadcast.sd.pd.512(<2 x double> %a0) ; <<8 x double>> [#uses=1]
ret <8 x double> %res
}
declare <8 x double> @llvm.x86.avx512.vbroadcast.sd.pd.512(<2 x double>) nounwind readonly
define <16 x i32> @test_x86_pbroadcastd_512(<4 x i32> %a0) {
; CHECK: vpbroadcastd
%res = call <16 x i32> @llvm.x86.avx512.pbroadcastd.512(<4 x i32> %a0) ; <<16 x i32>> [#uses=1]
ret <16 x i32> %res
}
declare <16 x i32> @llvm.x86.avx512.pbroadcastd.512(<4 x i32>) nounwind readonly
define <16 x i32> @test_x86_pbroadcastd_i32_512(i32 %a0) {
; CHECK: vpbroadcastd
%res = call <16 x i32> @llvm.x86.avx512.pbroadcastd.i32.512(i32 %a0) ; <<16 x i32>> [#uses=1]
ret <16 x i32> %res
}
declare <16 x i32> @llvm.x86.avx512.pbroadcastd.i32.512(i32) nounwind readonly
define <8 x i64> @test_x86_pbroadcastq_512(<2 x i64> %a0) {
; CHECK: vpbroadcastq
%res = call <8 x i64> @llvm.x86.avx512.pbroadcastq.512(<2 x i64> %a0) ; <<8 x i64>> [#uses=1]
ret <8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.pbroadcastq.512(<2 x i64>) nounwind readonly
define <8 x i64> @test_x86_pbroadcastq_i64_512(i64 %a0) {
; CHECK: vpbroadcastq
%res = call <8 x i64> @llvm.x86.avx512.pbroadcastq.i64.512(i64 %a0) ; <<8 x i64>> [#uses=1]
ret <8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.pbroadcastq.i64.512(i64) nounwind readonly
define <16 x i32> @test_x86_pmaxu_d(<16 x i32> %a0, <16 x i32> %a1) {
; CHECK: vpmaxud
%res = call <16 x i32> @llvm.x86.avx512.pmaxu.d(<16 x i32> %a0, <16 x i32> %a1) ; <<16 x i32>> [#uses=1]
ret <16 x i32> %res
}
declare <16 x i32> @llvm.x86.avx512.pmaxu.d(<16 x i32>, <16 x i32>) nounwind readonly
define <8 x i64> @test_x86_pmaxu_q(<8 x i64> %a0, <8 x i64> %a1) {
; CHECK: vpmaxuq
%res = call <8 x i64> @llvm.x86.avx512.pmaxu.q(<8 x i64> %a0, <8 x i64> %a1) ; <<8 x i64>> [#uses=1]
ret <8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.pmaxu.q(<8 x i64>, <8 x i64>) nounwind readonly
define <16 x i32> @test_x86_pmaxs_d(<16 x i32> %a0, <16 x i32> %a1) {
; CHECK: vpmaxsd
%res = call <16 x i32> @llvm.x86.avx512.pmaxs.d(<16 x i32> %a0, <16 x i32> %a1) ; <<16 x i32>> [#uses=1]
ret <16 x i32> %res
}
declare <16 x i32> @llvm.x86.avx512.pmaxs.d(<16 x i32>, <16 x i32>) nounwind readonly
define <8 x i64> @test_x86_pmaxs_q(<8 x i64> %a0, <8 x i64> %a1) {
; CHECK: vpmaxsq
%res = call <8 x i64> @llvm.x86.avx512.pmaxs.q(<8 x i64> %a0, <8 x i64> %a1) ; <<8 x i64>> [#uses=1]
ret <8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.pmaxs.q(<8 x i64>, <8 x i64>) nounwind readonly
define <16 x i32> @test_x86_pminu_d(<16 x i32> %a0, <16 x i32> %a1) {
; CHECK: vpminud
%res = call <16 x i32> @llvm.x86.avx512.pminu.d(<16 x i32> %a0, <16 x i32> %a1) ; <<16 x i32>> [#uses=1]
ret <16 x i32> %res
}
declare <16 x i32> @llvm.x86.avx512.pminu.d(<16 x i32>, <16 x i32>) nounwind readonly
define <8 x i64> @test_x86_pminu_q(<8 x i64> %a0, <8 x i64> %a1) {
; CHECK: vpminuq
%res = call <8 x i64> @llvm.x86.avx512.pminu.q(<8 x i64> %a0, <8 x i64> %a1) ; <<8 x i64>> [#uses=1]
ret <8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.pminu.q(<8 x i64>, <8 x i64>) nounwind readonly
define <16 x i32> @test_x86_pmins_d(<16 x i32> %a0, <16 x i32> %a1) {
; CHECK: vpminsd
%res = call <16 x i32> @llvm.x86.avx512.pmins.d(<16 x i32> %a0, <16 x i32> %a1) ; <<16 x i32>> [#uses=1]
ret <16 x i32> %res
}
declare <16 x i32> @llvm.x86.avx512.pmins.d(<16 x i32>, <16 x i32>) nounwind readonly
define <8 x i64> @test_x86_pmins_q(<8 x i64> %a0, <8 x i64> %a1) {
; CHECK: vpminsq
%res = call <8 x i64> @llvm.x86.avx512.pmins.q(<8 x i64> %a0, <8 x i64> %a1) ; <<8 x i64>> [#uses=1]
ret <8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.pmins.q(<8 x i64>, <8 x i64>) nounwind readonly
define <16 x i32> @test_conflict_d(<16 x i32> %a) {
; CHECK: movw $-1, %ax
; CHECK: vpxor
; CHECK: vpconflictd
%res = call <16 x i32> @llvm.x86.avx512.mask.conflict.d.512(<16 x i32> %a, <16 x i32> zeroinitializer, i16 -1)
ret <16 x i32> %res
}
declare <16 x i32> @llvm.x86.avx512.mask.conflict.d.512(<16 x i32>, <16 x i32>, i16) nounwind readonly
define <8 x i64> @test_conflict_q(<8 x i64> %a) {
; CHECK: movb $-1, %al
; CHECK: vpxor
; CHECK: vpconflictq
%res = call <8 x i64> @llvm.x86.avx512.mask.conflict.q.512(<8 x i64> %a, <8 x i64> zeroinitializer, i8 -1)
ret <8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.mask.conflict.q.512(<8 x i64>, <8 x i64>, i8) nounwind readonly
define <16 x i32> @test_maskz_conflict_d(<16 x i32> %a, i16 %mask) {
; CHECK: vpconflictd
%res = call <16 x i32> @llvm.x86.avx512.mask.conflict.d.512(<16 x i32> %a, <16 x i32> zeroinitializer, i16 %mask)
ret <16 x i32> %res
}
define <8 x i64> @test_mask_conflict_q(<8 x i64> %a, <8 x i64> %b, i8 %mask) {
; CHECK: vpconflictq
%res = call <8 x i64> @llvm.x86.avx512.mask.conflict.q.512(<8 x i64> %a, <8 x i64> %b, i8 %mask)
ret <8 x i64> %res
}
define <16 x float> @test_x86_mask_blend_ps_512(i16 %a0, <16 x float> %a1, <16 x float> %a2) {
; CHECK: vblendmps
%m0 = bitcast i16 %a0 to <16 x i1>
%res = call <16 x float> @llvm.x86.avx512.mask.blend.ps.512(<16 x i1> %m0, <16 x float> %a1, <16 x float> %a2) ; <<16 x float>> [#uses=1]
ret <16 x float> %res
}
declare <16 x float> @llvm.x86.avx512.mask.blend.ps.512(<16 x i1> %a0, <16 x float> %a1, <16 x float> %a2) nounwind readonly
define <8 x double> @test_x86_mask_blend_pd_512(i8 %a0, <8 x double> %a1, <8 x double> %a2) {
; CHECK: vblendmpd
%m0 = bitcast i8 %a0 to <8 x i1>
%res = call <8 x double> @llvm.x86.avx512.mask.blend.pd.512(<8 x i1> %m0, <8 x double> %a1, <8 x double> %a2) ; <<8 x double>> [#uses=1]
ret <8 x double> %res
}
define <8 x double> @test_x86_mask_blend_pd_512_memop(<8 x double> %a, <8 x double>* %ptr, i8 %mask) {
; CHECK-LABEL: test_x86_mask_blend_pd_512_memop
; CHECK: vblendmpd {{.*}}, {{%zmm[0-9]}}, {{%zmm[0-9]}} {%k1}
%vmask = bitcast i8 %mask to <8 x i1>
%b = load <8 x double>* %ptr
%res = call <8 x double> @llvm.x86.avx512.mask.blend.pd.512(<8 x i1> %vmask, <8 x double> %a, <8 x double> %b) ; <<8 x double>> [#uses=1]
ret <8 x double> %res
}
declare <8 x double> @llvm.x86.avx512.mask.blend.pd.512(<8 x i1> %a0, <8 x double> %a1, <8 x double> %a2) nounwind readonly
define <16 x i32> @test_x86_mask_blend_d_512(i16 %a0, <16 x i32> %a1, <16 x i32> %a2) {
; CHECK: vpblendmd
%m0 = bitcast i16 %a0 to <16 x i1>
%res = call <16 x i32> @llvm.x86.avx512.mask.blend.d.512(<16 x i1> %m0, <16 x i32> %a1, <16 x i32> %a2) ; <<16 x i32>> [#uses=1]
ret <16 x i32> %res
}
declare <16 x i32> @llvm.x86.avx512.mask.blend.d.512(<16 x i1> %a0, <16 x i32> %a1, <16 x i32> %a2) nounwind readonly
define <8 x i64> @test_x86_mask_blend_q_512(i8 %a0, <8 x i64> %a1, <8 x i64> %a2) {
; CHECK: vpblendmq
%m0 = bitcast i8 %a0 to <8 x i1>
%res = call <8 x i64> @llvm.x86.avx512.mask.blend.q.512(<8 x i1> %m0, <8 x i64> %a1, <8 x i64> %a2) ; <<8 x i64>> [#uses=1]
ret <8 x i64> %res
}
declare <8 x i64> @llvm.x86.avx512.mask.blend.q.512(<8 x i1> %a0, <8 x i64> %a1, <8 x i64> %a2) nounwind readonly