1; RUN: llc < %s -mtriple=x86_64-pc-linux -mcpu=knl --show-mc-encoding| FileCheck %s
2
3define <16 x float> @test_rsqrt28_ps(<16 x float> %a0) {
4  ; CHECK: vrsqrt28ps {sae}, %zmm0, %zmm0 # encoding: [0x62,0xf2,0x7d,0x18,0xcc,0xc0]
5  %res = call <16 x float> @llvm.x86.avx512.rsqrt28.ps(<16 x float> %a0, <16 x float> zeroinitializer, i16 -1, i32 8)
6  ret <16 x float> %res
7}
8
9define <16 x float> @test1_rsqrt28_ps(<16 x float> %a0, <16 x float> %a1) {
10  ; CHECK: kmovw
11  ; CHECK: vrsqrt28ps {sae}, %zmm0, %zmm1 {%k1} # encoding: [0x62,0xf2,0x7d,0x19,0xcc,0xc8]
12  %res = call <16 x float> @llvm.x86.avx512.rsqrt28.ps(<16 x float> %a0, <16 x float> %a1, i16 6, i32 8)
13  ret <16 x float> %res
14}
15
16define <16 x float> @test2_rsqrt28_ps(<16 x float> %a0) {
17  ; CHECK: vrsqrt28ps %zmm0, %zmm0 {%k1} {z} # encoding: [0x62,0xf2,0x7d,0xc9,0xcc,0xc0]
18  %res = call <16 x float> @llvm.x86.avx512.rsqrt28.ps(<16 x float> %a0, <16 x float> undef, i16 6, i32 4)
19  ret <16 x float> %res
20}
21
22define <16 x float> @test3_rsqrt28_ps(<16 x float> %a0) {
23  ; CHECK: kmovw
24  ; CHECK: vrsqrt28ps %zmm0, %zmm0 {%k1} {z} # encoding: [0x62,0xf2,0x7d,0xc9,0xcc,0xc0]
25  %res = call <16 x float> @llvm.x86.avx512.rsqrt28.ps(<16 x float> %a0, <16 x float> zeroinitializer, i16 6, i32 4)
26  ret <16 x float> %res
27}
28
29define <16 x float> @test4_rsqrt28_ps(<16 x float> %a0) {
30  ; CHECK: vrsqrt28ps {sae}, %zmm0, %zmm0 {%k1} {z} # encoding: [0x62,0xf2,0x7d,0x99,0xcc,0xc0]
31  %res = call <16 x float> @llvm.x86.avx512.rsqrt28.ps(<16 x float> %a0, <16 x float> undef, i16 6, i32 8)
32  ret <16 x float> %res
33}
34
35
36declare <16 x float> @llvm.x86.avx512.rsqrt28.ps(<16 x float>, <16 x float>, i16, i32) nounwind readnone
37
38define <16 x float> @test_rcp28_ps_512(<16 x float> %a0) {
39  ; CHECK: vrcp28ps {sae}, %zmm0, %zmm0 # encoding: [0x62,0xf2,0x7d,0x18,0xca,0xc0]
40  %res = call <16 x float> @llvm.x86.avx512.rcp28.ps(<16 x float> %a0, <16 x float> zeroinitializer, i16 -1, i32 8)
41  ret <16 x float> %res
42}
43declare <16 x float> @llvm.x86.avx512.rcp28.ps(<16 x float>, <16 x float>, i16, i32) nounwind readnone
44
45define <8 x double> @test_rcp28_pd_512(<8 x double> %a0) {
46  ; CHECK: vrcp28pd {sae}, %zmm0, %zmm0 # encoding: [0x62,0xf2,0xfd,0x18,0xca,0xc0]
47  %res = call <8 x double> @llvm.x86.avx512.rcp28.pd(<8 x double> %a0, <8 x double> zeroinitializer, i8 -1, i32 8)
48  ret <8 x double> %res
49}
50declare <8 x double> @llvm.x86.avx512.rcp28.pd(<8 x double>, <8 x double>, i8, i32) nounwind readnone
51
52define <16 x float> @test_exp2_ps_512(<16 x float> %a0) {
53  ; CHECK: vexp2ps {sae}, %zmm0, %zmm0     # encoding: [0x62,0xf2,0x7d,0x18,0xc8,0xc0]
54  %res = call <16 x float> @llvm.x86.avx512.exp2.ps(<16 x float> %a0, <16 x float> zeroinitializer, i16 -1, i32 8)
55  ret <16 x float> %res
56}
57declare <16 x float> @llvm.x86.avx512.exp2.ps(<16 x float>, <16 x float>, i16, i32) nounwind readnone
58
59define <8 x double> @test_exp2_pd_512(<8 x double> %a0) {
60  ; CHECK: vexp2pd {sae}, %zmm0, %zmm0      # encoding: [0x62,0xf2,0xfd,0x18,0xc8,0xc0]
61  %res = call <8 x double> @llvm.x86.avx512.exp2.pd(<8 x double> %a0, <8 x double> zeroinitializer, i8 -1, i32 8)
62  ret <8 x double> %res
63}
64declare <8 x double> @llvm.x86.avx512.exp2.pd(<8 x double>, <8 x double>, i8, i32) nounwind readnone
65
66define <4 x float> @test_rsqrt28_ss(<4 x float> %a0) {
67  ; CHECK: vrsqrt28ss {sae}, %xmm0, %xmm0, %xmm0 # encoding: [0x62,0xf2,0x7d,0x18,0xcd,0xc0]
68  %res = call <4 x float> @llvm.x86.avx512.rsqrt28.ss(<4 x float> %a0, <4 x float> %a0, <4 x float> zeroinitializer, i8 -1, i32 8) ; <<4 x float>> [#uses=1]
69  ret <4 x float> %res
70}
71declare <4 x float> @llvm.x86.avx512.rsqrt28.ss(<4 x float>, <4 x float>, <4 x float>, i8, i32) nounwind readnone
72
73define <4 x float> @test_rcp28_ss(<4 x float> %a0) {
74  ; CHECK: vrcp28ss {sae}, %xmm0, %xmm0, %xmm0 # encoding: [0x62,0xf2,0x7d,0x18,0xcb,0xc0]
75  %res = call <4 x float> @llvm.x86.avx512.rcp28.ss(<4 x float> %a0, <4 x float> %a0, <4 x float> zeroinitializer, i8 -1, i32 8) ; <<4 x float>> [#uses=1]
76  ret <4 x float> %res
77}
78declare <4 x float> @llvm.x86.avx512.rcp28.ss(<4 x float>, <4 x float>, <4 x float>, i8, i32) nounwind readnone
79
80define <4 x float> @test_rsqrt28_ss_maskz(<4 x float> %a0) {
81  ; CHECK: vrsqrt28ss {sae}, %xmm0, %xmm0, %xmm0 {%k1} {z} # encoding: [0x62,0xf2,0x7d,0x99,0xcd,0xc0]
82  %res = call <4 x float> @llvm.x86.avx512.rsqrt28.ss(<4 x float> %a0, <4 x float> %a0, <4 x float> zeroinitializer, i8 7, i32 8) ;
83  ret <4 x float> %res
84}
85
86define <4 x float> @test_rsqrt28_ss_mask(<4 x float> %a0, <4 x float> %b0, <4 x float> %c0) {
87  ; CHECK: vrsqrt28ss {sae}, %xmm1, %xmm0, %xmm2 {%k1} # encoding: [0x62,0xf2,0x7d,0x19,0xcd,0xd1]
88  %res = call <4 x float> @llvm.x86.avx512.rsqrt28.ss(<4 x float> %a0, <4 x float> %b0, <4 x float> %c0, i8 7, i32 8) ;
89  ret <4 x float> %res
90}
91
92define <2 x double> @test_rsqrt28_sd_maskz(<2 x double> %a0) {
93  ; CHECK: vrsqrt28sd {sae}, %xmm0, %xmm0, %xmm0 {%k1} {z} # encoding: [0x62,0xf2,0xfd,0x99,0xcd,0xc0]
94  %res = call <2 x double> @llvm.x86.avx512.rsqrt28.sd(<2 x double> %a0, <2 x double> %a0, <2 x double> zeroinitializer, i8 7, i32 8) ;
95  ret <2 x double> %res
96}
97
98declare <2 x double> @llvm.x86.avx512.rsqrt28.sd(<2 x double>, <2 x double>, <2 x double>, i8, i32) nounwind readnone
99
100define <2 x double> @test_rsqrt28_sd_maskz_mem(<2 x double> %a0, double* %ptr ) {
101  ; CHECK: vrsqrt28sd (%rdi), %xmm0, %xmm0 {%k1} {z} # encoding: [0x62,0xf2,0xfd,0x89,0xcd,0x07]
102  %mem = load double , double * %ptr, align 8
103  %mem_v = insertelement <2 x double> undef, double %mem, i32 0
104  %res = call <2 x double> @llvm.x86.avx512.rsqrt28.sd(<2 x double> %a0, <2 x double> %mem_v, <2 x double> zeroinitializer, i8 7, i32 4) ;
105  ret <2 x double> %res
106}
107
108define <2 x double> @test_rsqrt28_sd_maskz_mem_offset(<2 x double> %a0, double* %ptr ) {
109  ; CHECK: vrsqrt28sd 144(%rdi), %xmm0, %xmm0 {%k1} {z} # encoding: [0x62,0xf2,0xfd,0x89,0xcd,0x47,0x12]
110  %ptr1 = getelementptr double, double* %ptr, i32 18
111  %mem = load double , double * %ptr1, align 8
112  %mem_v = insertelement <2 x double> undef, double %mem, i32 0
113  %res = call <2 x double> @llvm.x86.avx512.rsqrt28.sd(<2 x double> %a0, <2 x double> %mem_v, <2 x double> zeroinitializer, i8 7, i32 4) ;
114  ret <2 x double> %res
115}
116
117