1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
|
; RUN: llc -mtriple=aarch64--linux-gnu -mattr=+sve --asm-verbose=false < %s | FileCheck %s
; 2-lane non-temporal load/stores
define void @test_masked_ldst_sv2i64(i64* %base, <vscale x 2 x i1> %mask, i64 %offset) nounwind {
; CHECK-LABEL: test_masked_ldst_sv2i64:
; CHECK-NEXT: ldnt1d { z[[DATA:[0-9]+]].d }, p0/z, [x0, x1, lsl #3]
; CHECK-NEXT: stnt1d { z[[DATA]].d }, p0, [x0, x1, lsl #3]
; CHECK-NEXT: ret
%gep = getelementptr i64, i64* %base, i64 %offset
%data = call <vscale x 2 x i64> @llvm.aarch64.sve.ldnt1.nxv2i64(<vscale x 2 x i1> %mask,
i64* %gep)
call void @llvm.aarch64.sve.stnt1.nxv2i64(<vscale x 2 x i64> %data,
<vscale x 2 x i1> %mask,
i64* %gep)
ret void
}
define void @test_masked_ldst_sv2f64(double* %base, <vscale x 2 x i1> %mask, i64 %offset) nounwind {
; CHECK-LABEL: test_masked_ldst_sv2f64:
; CHECK-NEXT: ldnt1d { z[[DATA:[0-9]+]].d }, p0/z, [x0, x1, lsl #3]
; CHECK-NEXT: stnt1d { z[[DATA]].d }, p0, [x0, x1, lsl #3]
; CHECK-NEXT: ret
%gep = getelementptr double, double* %base, i64 %offset
%data = call <vscale x 2 x double> @llvm.aarch64.sve.ldnt1.nxv2f64(<vscale x 2 x i1> %mask,
double* %gep)
call void @llvm.aarch64.sve.stnt1.nxv2f64(<vscale x 2 x double> %data,
<vscale x 2 x i1> %mask,
double* %gep)
ret void
}
; 4-lane non-temporal load/stores.
define void @test_masked_ldst_sv4i32(i32* %base, <vscale x 4 x i1> %mask, i64 %offset) nounwind {
; CHECK-LABEL: test_masked_ldst_sv4i32:
; CHECK-NEXT: ldnt1w { z[[DATA:[0-9]+]].s }, p0/z, [x0, x1, lsl #2]
; CHECK-NEXT: stnt1w { z[[DATA]].s }, p0, [x0, x1, lsl #2]
; CHECK-NEXT: ret
%gep = getelementptr i32, i32* %base, i64 %offset
%data = call <vscale x 4 x i32> @llvm.aarch64.sve.ldnt1.nxv4i32(<vscale x 4 x i1> %mask,
i32* %gep)
call void @llvm.aarch64.sve.stnt1.nxv4i32(<vscale x 4 x i32> %data,
<vscale x 4 x i1> %mask,
i32* %gep)
ret void
}
define void @test_masked_ldst_sv4f32(float* %base, <vscale x 4 x i1> %mask, i64 %offset) nounwind {
; CHECK-LABEL: test_masked_ldst_sv4f32:
; CHECK-NEXT: ldnt1w { z[[DATA:[0-9]+]].s }, p0/z, [x0, x1, lsl #2]
; CHECK-NEXT: stnt1w { z[[DATA]].s }, p0, [x0, x1, lsl #2]
; CHECK-NEXT: ret
%gep = getelementptr float, float* %base, i64 %offset
%data = call <vscale x 4 x float> @llvm.aarch64.sve.ldnt1.nxv4f32(<vscale x 4 x i1> %mask,
float* %gep)
call void @llvm.aarch64.sve.stnt1.nxv4f32(<vscale x 4 x float> %data,
<vscale x 4 x i1> %mask,
float* %gep)
ret void
}
; 8-lane non-temporal load/stores.
define void @test_masked_ldst_sv8i16(i16* %base, <vscale x 8 x i1> %mask, i64 %offset) nounwind {
; CHECK-LABEL: test_masked_ldst_sv8i16:
; CHECK-NEXT: ldnt1h { z[[DATA:[0-9]+]].h }, p0/z, [x0, x1, lsl #1]
; CHECK-NEXT: stnt1h { z[[DATA]].h }, p0, [x0, x1, lsl #1]
; CHECK-NEXT: ret
%gep = getelementptr i16, i16* %base, i64 %offset
%data = call <vscale x 8 x i16> @llvm.aarch64.sve.ldnt1.nxv8i16(<vscale x 8 x i1> %mask,
i16* %gep)
call void @llvm.aarch64.sve.stnt1.nxv8i16(<vscale x 8 x i16> %data,
<vscale x 8 x i1> %mask,
i16* %gep)
ret void
}
define void @test_masked_ldst_sv8f16(half* %base, <vscale x 8 x i1> %mask, i64 %offset) nounwind {
; CHECK-LABEL: test_masked_ldst_sv8f16:
; CHECK-NEXT: ldnt1h { z[[DATA:[0-9]+]].h }, p0/z, [x0, x1, lsl #1]
; CHECK-NEXT: stnt1h { z[[DATA]].h }, p0, [x0, x1, lsl #1]
; CHECK-NEXT: ret
%gep = getelementptr half, half* %base, i64 %offset
%data = call <vscale x 8 x half> @llvm.aarch64.sve.ldnt1.nxv8f16(<vscale x 8 x i1> %mask,
half* %gep)
call void @llvm.aarch64.sve.stnt1.nxv8f16(<vscale x 8 x half> %data,
<vscale x 8 x i1> %mask,
half* %gep)
ret void
}
define void @test_masked_ldst_sv8bf16(bfloat* %base, <vscale x 8 x i1> %mask, i64 %offset) nounwind #0 {
; CHECK-LABEL: test_masked_ldst_sv8bf16:
; CHECK-NEXT: ldnt1h { z[[DATA:[0-9]+]].h }, p0/z, [x0, x1, lsl #1]
; CHECK-NEXT: stnt1h { z[[DATA]].h }, p0, [x0, x1, lsl #1]
; CHECK-NEXT: ret
%gep = getelementptr bfloat, bfloat* %base, i64 %offset
%data = call <vscale x 8 x bfloat> @llvm.aarch64.sve.ldnt1.nxv8bf16(<vscale x 8 x i1> %mask,
bfloat* %gep)
call void @llvm.aarch64.sve.stnt1.nxv8bf16(<vscale x 8 x bfloat> %data,
<vscale x 8 x i1> %mask,
bfloat* %gep)
ret void
}
; 16-lane non-temporal load/stores.
define void @test_masked_ldst_sv16i8(i8* %base, <vscale x 16 x i1> %mask, i64 %offset) nounwind {
; CHECK-LABEL: test_masked_ldst_sv16i8:
; CHECK-NEXT: ldnt1b { z[[DATA:[0-9]+]].b }, p0/z, [x0, x1]
; CHECK-NEXT: stnt1b { z[[DATA]].b }, p0, [x0, x1]
; CHECK-NEXT: ret
%gep = getelementptr i8, i8* %base, i64 %offset
%data = call <vscale x 16 x i8> @llvm.aarch64.sve.ldnt1.nxv16i8(<vscale x 16 x i1> %mask,
i8* %gep)
call void @llvm.aarch64.sve.stnt1.nxv16i8(<vscale x 16 x i8> %data,
<vscale x 16 x i1> %mask,
i8* %gep)
ret void
}
; 2-element non-temporal loads.
declare <vscale x 2 x i64> @llvm.aarch64.sve.ldnt1.nxv2i64(<vscale x 2 x i1>, i64*)
declare <vscale x 2 x double> @llvm.aarch64.sve.ldnt1.nxv2f64(<vscale x 2 x i1>, double*)
; 4-element non-temporal loads.
declare <vscale x 4 x i32> @llvm.aarch64.sve.ldnt1.nxv4i32(<vscale x 4 x i1>, i32*)
declare <vscale x 4 x float> @llvm.aarch64.sve.ldnt1.nxv4f32(<vscale x 4 x i1>, float*)
; 8-element non-temporal loads.
declare <vscale x 8 x i16> @llvm.aarch64.sve.ldnt1.nxv8i16(<vscale x 8 x i1>, i16*)
declare <vscale x 8 x half> @llvm.aarch64.sve.ldnt1.nxv8f16(<vscale x 8 x i1>, half*)
declare <vscale x 8 x bfloat> @llvm.aarch64.sve.ldnt1.nxv8bf16(<vscale x 8 x i1>, bfloat*)
; 16-element non-temporal loads.
declare <vscale x 16 x i8> @llvm.aarch64.sve.ldnt1.nxv16i8(<vscale x 16 x i1>, i8*)
; 2-element non-temporal stores.
declare void @llvm.aarch64.sve.stnt1.nxv2i64(<vscale x 2 x i64>, <vscale x 2 x i1>, i64*)
declare void @llvm.aarch64.sve.stnt1.nxv2f64(<vscale x 2 x double>, <vscale x 2 x i1>, double*)
; 4-element non-temporal stores.
declare void @llvm.aarch64.sve.stnt1.nxv4i32(<vscale x 4 x i32>, <vscale x 4 x i1>, i32*)
declare void @llvm.aarch64.sve.stnt1.nxv4f32(<vscale x 4 x float>, <vscale x 4 x i1>, float*)
; 8-element non-temporal stores.
declare void @llvm.aarch64.sve.stnt1.nxv8i16(<vscale x 8 x i16>, <vscale x 8 x i1>, i16*)
declare void @llvm.aarch64.sve.stnt1.nxv8f16(<vscale x 8 x half>, <vscale x 8 x i1>, half*)
declare void @llvm.aarch64.sve.stnt1.nxv8bf16(<vscale x 8 x bfloat>, <vscale x 8 x i1>, bfloat*)
; 16-element non-temporal stores.
declare void @llvm.aarch64.sve.stnt1.nxv16i8(<vscale x 16 x i8>, <vscale x 16 x i1>, i8*)
; +bf16 is required for the bfloat version.
attributes #0 = { "target-features"="+sve,+bf16" }
|
