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
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
|
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=aarch64-linux-gnu -mattr=+sve2p1,+bf16 < %s | FileCheck %s
;
; LD1Q: vector base + unscaled offset
; e.g. ld1q { z0.q }, p0/z, [z0.d, x0]
;
define <vscale x 16 x i8> @ld1q_gather_u64base_i8(<vscale x 1 x i1> %pg, <vscale x 2 x i64> %base, i64 %offset) {
; CHECK-LABEL: ld1q_gather_u64base_i8:
; CHECK: // %bb.0:
; CHECK-NEXT: ld1q { z0.q }, p0/z, [z0.d, x0]
; CHECK-NEXT: ret
%load = call <vscale x 16 x i8> @llvm.aarch64.sve.ld1q.gather.scalar.offset.nxv16i8.nxv2i64(<vscale x 1 x i1> %pg,
<vscale x 2 x i64> %base,
i64 %offset)
ret <vscale x 16 x i8> %load
}
define <vscale x 8 x i16> @ld1q_gather_u64base_i16(<vscale x 1 x i1> %pg, <vscale x 2 x i64> %base, i64 %offset) {
; CHECK-LABEL: ld1q_gather_u64base_i16:
; CHECK: // %bb.0:
; CHECK-NEXT: ld1q { z0.q }, p0/z, [z0.d, x0]
; CHECK-NEXT: ret
%load = call <vscale x 8 x i16> @llvm.aarch64.sve.ld1q.gather.scalar.offset.nxv8i16.nxv2i64(<vscale x 1 x i1> %pg,
<vscale x 2 x i64> %base,
i64 %offset)
ret <vscale x 8 x i16> %load
}
define <vscale x 4 x i32> @ld1q_gather_u64base_i32(<vscale x 1 x i1> %pg, <vscale x 2 x i64> %base, i64 %offset) {
; CHECK-LABEL: ld1q_gather_u64base_i32:
; CHECK: // %bb.0:
; CHECK-NEXT: ld1q { z0.q }, p0/z, [z0.d, x0]
; CHECK-NEXT: ret
%load = call <vscale x 4 x i32> @llvm.aarch64.sve.ld1q.gather.scalar.offset.nxv4i32.nxv2i64(<vscale x 1 x i1> %pg,
<vscale x 2 x i64> %base,
i64 %offset)
ret <vscale x 4 x i32> %load
}
define <vscale x 2 x i64> @ld1q_gather_u64base_i64(<vscale x 1 x i1> %pg, <vscale x 2 x i64> %base, i64 %offset) {
; CHECK-LABEL: ld1q_gather_u64base_i64:
; CHECK: // %bb.0:
; CHECK-NEXT: ld1q { z0.q }, p0/z, [z0.d, x0]
; CHECK-NEXT: ret
%load = call <vscale x 2 x i64> @llvm.aarch64.sve.ld1q.gather.scalar.offset.nxv2i64.nxv2i64(<vscale x 1 x i1> %pg,
<vscale x 2 x i64> %base,
i64 %offset)
ret <vscale x 2 x i64> %load
}
define <vscale x 8 x half> @ld1q_gather_u64base_f16(<vscale x 1 x i1> %pg, <vscale x 2 x i64> %base, i64 %offset) {
; CHECK-LABEL: ld1q_gather_u64base_f16:
; CHECK: // %bb.0:
; CHECK-NEXT: ld1q { z0.q }, p0/z, [z0.d, x0]
; CHECK-NEXT: ret
%load = call <vscale x 8 x half> @llvm.aarch64.sve.ld1q.gather.scalar.offset.nxv8f16.nxv2i64(<vscale x 1 x i1> %pg,
<vscale x 2 x i64> %base,
i64 %offset)
ret <vscale x 8 x half> %load
}
define <vscale x 4 x float> @ld1q_gather_u64base_f32(<vscale x 1 x i1> %pg, <vscale x 2 x i64> %base, i64 %offset) {
; CHECK-LABEL: ld1q_gather_u64base_f32:
; CHECK: // %bb.0:
; CHECK-NEXT: ld1q { z0.q }, p0/z, [z0.d, x0]
; CHECK-NEXT: ret
%load = call <vscale x 4 x float> @llvm.aarch64.sve.ld1q.gather.scalar.offset.nxv4f32.nxv2i64(<vscale x 1 x i1> %pg,
<vscale x 2 x i64> %base,
i64 %offset)
ret <vscale x 4 x float> %load
}
define <vscale x 2 x double> @ld1q_gather_u64base_f64(<vscale x 1 x i1> %pg, <vscale x 2 x i64> %base, i64 %offset) {
; CHECK-LABEL: ld1q_gather_u64base_f64:
; CHECK: // %bb.0:
; CHECK-NEXT: ld1q { z0.q }, p0/z, [z0.d, x0]
; CHECK-NEXT: ret
%load = call <vscale x 2 x double> @llvm.aarch64.sve.ld1q.gather.scalar.offset.nxv2f64.nxv2i64(<vscale x 1 x i1> %pg,
<vscale x 2 x i64> %base,
i64 %offset)
ret <vscale x 2 x double> %load
}
define <vscale x 8 x bfloat> @ld1q_gather_u64base_bf16(<vscale x 1 x i1> %pg, <vscale x 2 x i64> %base, i64 %offset) {
; CHECK-LABEL: ld1q_gather_u64base_bf16:
; CHECK: // %bb.0:
; CHECK-NEXT: ld1q { z0.q }, p0/z, [z0.d, x0]
; CHECK-NEXT: ret
%load = call <vscale x 8 x bfloat> @llvm.aarch64.sve.ld1q.gather.scalar.offset.nxv8bf16.nxv2i64(<vscale x 1 x i1> %pg,
<vscale x 2 x i64> %base,
i64 %offset)
ret <vscale x 8 x bfloat> %load
}
define <vscale x 16 x i8> @test_svdl1q_gather_u64offset_s8(<vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %off) {
; CHECK-LABEL: test_svdl1q_gather_u64offset_s8:
; CHECK: // %bb.0: // %entry
; CHECK-NEXT: ld1q { z0.q }, p0/z, [z0.d, x0]
; CHECK-NEXT: ret
entry:
%0 = tail call <vscale x 16 x i8> @llvm.aarch64.sve.ld1q.gather.vector.offset.nxv16i8(<vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %off)
ret <vscale x 16 x i8> %0
}
define <vscale x 16 x i8> @test_svdl1q_gather_u64offset_u8(<vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %off) {
; CHECK-LABEL: test_svdl1q_gather_u64offset_u8:
; CHECK: // %bb.0: // %entry
; CHECK-NEXT: ld1q { z0.q }, p0/z, [z0.d, x0]
; CHECK-NEXT: ret
entry:
%0 = tail call <vscale x 16 x i8> @llvm.aarch64.sve.ld1q.gather.vector.offset.nxv16i8(<vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %off)
ret <vscale x 16 x i8> %0
}
define <vscale x 8 x i16> @test_svdl1q_gather_u64offset_s16(<vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %off) {
; CHECK-LABEL: test_svdl1q_gather_u64offset_s16:
; CHECK: // %bb.0: // %entry
; CHECK-NEXT: ld1q { z0.q }, p0/z, [z0.d, x0]
; CHECK-NEXT: ret
entry:
%0 = tail call <vscale x 8 x i16> @llvm.aarch64.sve.ld1q.gather.vector.offset.nxv8i16(<vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %off)
ret <vscale x 8 x i16> %0
}
define <vscale x 8 x i16> @test_svdl1q_gather_u64offset_u16(<vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %off) {
; CHECK-LABEL: test_svdl1q_gather_u64offset_u16:
; CHECK: // %bb.0: // %entry
; CHECK-NEXT: ld1q { z0.q }, p0/z, [z0.d, x0]
; CHECK-NEXT: ret
entry:
%0 = tail call <vscale x 8 x i16> @llvm.aarch64.sve.ld1q.gather.vector.offset.nxv8i16(<vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %off)
ret <vscale x 8 x i16> %0
}
define <vscale x 4 x i32> @test_svdl1q_gather_u64offset_s32(<vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %off) {
; CHECK-LABEL: test_svdl1q_gather_u64offset_s32:
; CHECK: // %bb.0: // %entry
; CHECK-NEXT: ld1q { z0.q }, p0/z, [z0.d, x0]
; CHECK-NEXT: ret
entry:
%0 = tail call <vscale x 4 x i32> @llvm.aarch64.sve.ld1q.gather.vector.offset.nxv4i32(<vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %off)
ret <vscale x 4 x i32> %0
}
define <vscale x 4 x i32> @test_svdl1q_gather_u64offset_u32(<vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %off) {
; CHECK-LABEL: test_svdl1q_gather_u64offset_u32:
; CHECK: // %bb.0: // %entry
; CHECK-NEXT: ld1q { z0.q }, p0/z, [z0.d, x0]
; CHECK-NEXT: ret
entry:
%0 = tail call <vscale x 4 x i32> @llvm.aarch64.sve.ld1q.gather.vector.offset.nxv4i32(<vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %off)
ret <vscale x 4 x i32> %0
}
define <vscale x 2 x i64> @test_svdl1q_gather_u64offset_s64(<vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %off) {
; CHECK-LABEL: test_svdl1q_gather_u64offset_s64:
; CHECK: // %bb.0: // %entry
; CHECK-NEXT: ld1q { z0.q }, p0/z, [z0.d, x0]
; CHECK-NEXT: ret
entry:
%0 = tail call <vscale x 2 x i64> @llvm.aarch64.sve.ld1q.gather.vector.offset.nxv2i64(<vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %off)
ret <vscale x 2 x i64> %0
}
define <vscale x 2 x i64> @test_svdl1q_gather_u64offset_u64(<vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %off) {
; CHECK-LABEL: test_svdl1q_gather_u64offset_u64:
; CHECK: // %bb.0: // %entry
; CHECK-NEXT: ld1q { z0.q }, p0/z, [z0.d, x0]
; CHECK-NEXT: ret
entry:
%0 = tail call <vscale x 2 x i64> @llvm.aarch64.sve.ld1q.gather.vector.offset.nxv2i64(<vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %off)
ret <vscale x 2 x i64> %0
}
define <vscale x 8 x bfloat> @test_svdl1q_gather_u64offset_bf16(<vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %off) {
; CHECK-LABEL: test_svdl1q_gather_u64offset_bf16:
; CHECK: // %bb.0: // %entry
; CHECK-NEXT: ld1q { z0.q }, p0/z, [z0.d, x0]
; CHECK-NEXT: ret
entry:
%0 = tail call <vscale x 8 x bfloat> @llvm.aarch64.sve.ld1q.gather.vector.offset.nxv8bf16(<vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %off)
ret <vscale x 8 x bfloat> %0
}
define <vscale x 8 x half> @test_svdl1q_gather_u64offset_f16(<vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %off) {
; CHECK-LABEL: test_svdl1q_gather_u64offset_f16:
; CHECK: // %bb.0: // %entry
; CHECK-NEXT: ld1q { z0.q }, p0/z, [z0.d, x0]
; CHECK-NEXT: ret
entry:
%0 = tail call <vscale x 8 x half> @llvm.aarch64.sve.ld1q.gather.vector.offset.nxv8f16(<vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %off)
ret <vscale x 8 x half> %0
}
define <vscale x 4 x float> @test_svdl1q_gather_u64offset_f32(<vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %off) {
; CHECK-LABEL: test_svdl1q_gather_u64offset_f32:
; CHECK: // %bb.0: // %entry
; CHECK-NEXT: ld1q { z0.q }, p0/z, [z0.d, x0]
; CHECK-NEXT: ret
entry:
%0 = tail call <vscale x 4 x float> @llvm.aarch64.sve.ld1q.gather.vector.offset.nxv4f32(<vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %off)
ret <vscale x 4 x float> %0
}
define <vscale x 2 x double> @test_svdl1q_gather_u64offset_f64(<vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %off) {
; CHECK-LABEL: test_svdl1q_gather_u64offset_f64:
; CHECK: // %bb.0: // %entry
; CHECK-NEXT: ld1q { z0.q }, p0/z, [z0.d, x0]
; CHECK-NEXT: ret
entry:
%0 = tail call <vscale x 2 x double> @llvm.aarch64.sve.ld1q.gather.vector.offset.nxv2f64(<vscale x 1 x i1> %pg, ptr %base, <vscale x 2 x i64> %off)
ret <vscale x 2 x double> %0
}
declare <vscale x 16 x i8> @llvm.aarch64.sve.ld1q.gather.scalar.offset.nxv16i8.nxv2i64(<vscale x 1 x i1>, <vscale x 2 x i64>, i64)
declare <vscale x 8 x i16> @llvm.aarch64.sve.ld1q.gather.scalar.offset.nxv8i16.nxv2i64(<vscale x 1 x i1>, <vscale x 2 x i64>, i64)
declare <vscale x 4 x i32> @llvm.aarch64.sve.ld1q.gather.scalar.offset.nxv4i32.nxv2i64(<vscale x 1 x i1>, <vscale x 2 x i64>, i64)
declare <vscale x 2 x i64> @llvm.aarch64.sve.ld1q.gather.scalar.offset.nxv2i64.nxv2i64(<vscale x 1 x i1>, <vscale x 2 x i64>, i64)
declare <vscale x 8 x half> @llvm.aarch64.sve.ld1q.gather.scalar.offset.nxv8f16.nxv2i64(<vscale x 1 x i1>, <vscale x 2 x i64>, i64)
declare <vscale x 4 x float> @llvm.aarch64.sve.ld1q.gather.scalar.offset.nxv4f32.nxv2i64(<vscale x 1 x i1>, <vscale x 2 x i64>, i64)
declare <vscale x 2 x double> @llvm.aarch64.sve.ld1q.gather.scalar.offset.nxv2f64.nxv2i64(<vscale x 1 x i1>, <vscale x 2 x i64>, i64)
declare <vscale x 8 x bfloat> @llvm.aarch64.sve.ld1q.gather.scalar.offset.nxv8bf16.nxv2i64(<vscale x 1 x i1>, <vscale x 2 x i64>, i64)
declare <vscale x 16 x i8> @llvm.aarch64.sve.ld1q.gather.vector.offset.nxv16i8(<vscale x 1 x i1>, ptr, <vscale x 2 x i64>)
declare <vscale x 8 x i16> @llvm.aarch64.sve.ld1q.gather.vector.offset.nxv8i16(<vscale x 1 x i1>, ptr, <vscale x 2 x i64>)
declare <vscale x 4 x i32> @llvm.aarch64.sve.ld1q.gather.vector.offset.nxv4i32(<vscale x 1 x i1>, ptr, <vscale x 2 x i64>)
declare <vscale x 2 x i64> @llvm.aarch64.sve.ld1q.gather.vector.offset.nxv2i64(<vscale x 1 x i1>, ptr, <vscale x 2 x i64>)
declare <vscale x 8 x bfloat> @llvm.aarch64.sve.ld1q.gather.vector.offset.nxv8bf16(<vscale x 1 x i1>, ptr, <vscale x 2 x i64>)
declare <vscale x 8 x half> @llvm.aarch64.sve.ld1q.gather.vector.offset.nxv8f16(<vscale x 1 x i1>, ptr, <vscale x 2 x i64>)
declare <vscale x 4 x float> @llvm.aarch64.sve.ld1q.gather.vector.offset.nxv4f32(<vscale x 1 x i1>, ptr, <vscale x 2 x i64>)
declare <vscale x 2 x double> @llvm.aarch64.sve.ld1q.gather.vector.offset.nxv2f64(<vscale x 1 x i1>, ptr, <vscale x 2 x i64>)
|
