sve-intrinsics-gather-loads-32bit-scaled-offsets.ll
13.6 KB
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
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
; RUN: llc -mtriple=aarch64-linux-gnu -mattr=+sve < %s | FileCheck %s
;
; LD1H, LD1W, LD1D: base + 32-bit scaled offset, sign (sxtw) or zero (uxtw)
; extended to 64 bits
; e.g. ld1h z0.d, p0/z, [x0, z0.d, uxtw #1]
;
; LD1H
define <vscale x 4 x i32> @gld1h_s_uxtw_index(<vscale x 4 x i1> %pg, i16* %base, <vscale x 4 x i32> %b) {
; CHECK-LABEL: gld1h_s_uxtw_index:
; CHECK: ld1h { z0.s }, p0/z, [x0, z0.s, uxtw #1]
; CHECK-NEXT: ret
%load = call <vscale x 4 x i16> @llvm.aarch64.sve.ld1.gather.uxtw.index.nxv4i16(<vscale x 4 x i1> %pg,
i16* %base,
<vscale x 4 x i32> %b)
%res = zext <vscale x 4 x i16> %load to <vscale x 4 x i32>
ret <vscale x 4 x i32> %res
}
define <vscale x 4 x i32> @gld1h_s_sxtw_index(<vscale x 4 x i1> %pg, i16* %base, <vscale x 4 x i32> %b) {
; CHECK-LABEL: gld1h_s_sxtw_index:
; CHECK: ld1h { z0.s }, p0/z, [x0, z0.s, sxtw #1]
; CHECK-NEXT: ret
%load = call <vscale x 4 x i16> @llvm.aarch64.sve.ld1.gather.sxtw.index.nxv4i16(<vscale x 4 x i1> %pg,
i16* %base,
<vscale x 4 x i32> %b)
%res = zext <vscale x 4 x i16> %load to <vscale x 4 x i32>
ret <vscale x 4 x i32> %res
}
define <vscale x 2 x i64> @gld1h_d_uxtw_index(<vscale x 2 x i1> %pg, i16* %base, <vscale x 2 x i32> %b) {
; CHECK-LABEL: gld1h_d_uxtw_index:
; CHECK: ld1h { z0.d }, p0/z, [x0, z0.d, uxtw #1]
; CHECK-NEXT: ret
%load = call <vscale x 2 x i16> @llvm.aarch64.sve.ld1.gather.uxtw.index.nxv2i16(<vscale x 2 x i1> %pg,
i16* %base,
<vscale x 2 x i32> %b)
%res = zext <vscale x 2 x i16> %load to <vscale x 2 x i64>
ret <vscale x 2 x i64> %res
}
define <vscale x 2 x i64> @gld1h_d_sxtw_index(<vscale x 2 x i1> %pg, i16* %base, <vscale x 2 x i32> %b) {
; CHECK-LABEL: gld1h_d_sxtw_index:
; CHECK: ld1h { z0.d }, p0/z, [x0, z0.d, sxtw #1]
; CHECK-NEXT: ret
%load = call <vscale x 2 x i16> @llvm.aarch64.sve.ld1.gather.sxtw.index.nxv2i16(<vscale x 2 x i1> %pg,
i16* %base,
<vscale x 2 x i32> %b)
%res = zext <vscale x 2 x i16> %load to <vscale x 2 x i64>
ret <vscale x 2 x i64> %res
}
; LD1W
define <vscale x 4 x i32> @gld1w_s_uxtw_index(<vscale x 4 x i1> %pg, i32* %base, <vscale x 4 x i32> %b) {
; CHECK-LABEL: gld1w_s_uxtw_index:
; CHECK: ld1w { z0.s }, p0/z, [x0, z0.s, uxtw #2]
; CHECK-NEXT: ret
%load = call <vscale x 4 x i32> @llvm.aarch64.sve.ld1.gather.uxtw.index.nxv4i32(<vscale x 4 x i1> %pg,
i32* %base,
<vscale x 4 x i32> %b)
ret <vscale x 4 x i32> %load
}
define <vscale x 4 x i32> @gld1w_s_sxtw_index(<vscale x 4 x i1> %pg, i32* %base, <vscale x 4 x i32> %b) {
; CHECK-LABEL: gld1w_s_sxtw_index:
; CHECK: ld1w { z0.s }, p0/z, [x0, z0.s, sxtw #2]
; CHECK-NEXT: ret
%load = call <vscale x 4 x i32> @llvm.aarch64.sve.ld1.gather.sxtw.index.nxv4i32(<vscale x 4 x i1> %pg,
i32* %base,
<vscale x 4 x i32> %b)
ret <vscale x 4 x i32> %load
}
define <vscale x 2 x i64> @gld1w_d_uxtw_index(<vscale x 2 x i1> %pg, i32* %base, <vscale x 2 x i32> %b) {
; CHECK-LABEL: gld1w_d_uxtw_index:
; CHECK: ld1w { z0.d }, p0/z, [x0, z0.d, uxtw #2]
; CHECK-NEXT: ret
%load = call <vscale x 2 x i32> @llvm.aarch64.sve.ld1.gather.uxtw.index.nxv2i32(<vscale x 2 x i1> %pg,
i32* %base,
<vscale x 2 x i32> %b)
%res = zext <vscale x 2 x i32> %load to <vscale x 2 x i64>
ret <vscale x 2 x i64> %res
}
define <vscale x 2 x i64> @gld1w_d_sxtw_index(<vscale x 2 x i1> %pg, i32* %base, <vscale x 2 x i32> %b) {
; CHECK-LABEL: gld1w_d_sxtw_index:
; CHECK: ld1w { z0.d }, p0/z, [x0, z0.d, sxtw #2]
; CHECK-NEXT: ret
%load = call <vscale x 2 x i32> @llvm.aarch64.sve.ld1.gather.sxtw.index.nxv2i32(<vscale x 2 x i1> %pg,
i32* %base,
<vscale x 2 x i32> %b)
%res = zext <vscale x 2 x i32> %load to <vscale x 2 x i64>
ret <vscale x 2 x i64> %res
}
define <vscale x 4 x float> @gld1w_s_uxtw_index_float(<vscale x 4 x i1> %pg, float* %base, <vscale x 4 x i32> %b) {
; CHECK-LABEL: gld1w_s_uxtw_index_float:
; CHECK: ld1w { z0.s }, p0/z, [x0, z0.s, uxtw #2]
; CHECK-NEXT: ret
%load = call <vscale x 4 x float> @llvm.aarch64.sve.ld1.gather.uxtw.index.nxv4f32(<vscale x 4 x i1> %pg,
float* %base,
<vscale x 4 x i32> %b)
ret <vscale x 4 x float> %load
}
define <vscale x 4 x float> @gld1w_s_sxtw_index_float(<vscale x 4 x i1> %pg, float* %base, <vscale x 4 x i32> %b) {
; CHECK-LABEL: gld1w_s_sxtw_index_float:
; CHECK: ld1w { z0.s }, p0/z, [x0, z0.s, sxtw #2]
; CHECK-NEXT: ret
%load = call <vscale x 4 x float> @llvm.aarch64.sve.ld1.gather.sxtw.index.nxv4f32(<vscale x 4 x i1> %pg,
float* %base,
<vscale x 4 x i32> %b)
ret <vscale x 4 x float> %load
}
; LD1D
define <vscale x 2 x i64> @gld1d_s_uxtw_index(<vscale x 2 x i1> %pg, i64* %base, <vscale x 2 x i32> %b) {
; CHECK-LABEL: gld1d_s_uxtw_index:
; CHECK: ld1d { z0.d }, p0/z, [x0, z0.d, uxtw #3]
; CHECK-NEXT: ret
%load = call <vscale x 2 x i64> @llvm.aarch64.sve.ld1.gather.uxtw.index.nxv2i64(<vscale x 2 x i1> %pg,
i64* %base,
<vscale x 2 x i32> %b)
ret <vscale x 2 x i64> %load
}
define <vscale x 2 x i64> @gld1d_sxtw_index(<vscale x 2 x i1> %pg, i64* %base, <vscale x 2 x i32> %b) {
; CHECK-LABEL: gld1d_sxtw_index:
; CHECK: ld1d { z0.d }, p0/z, [x0, z0.d, sxtw #3]
; CHECK-NEXT: ret
%load = call <vscale x 2 x i64> @llvm.aarch64.sve.ld1.gather.sxtw.index.nxv2i64(<vscale x 2 x i1> %pg,
i64* %base,
<vscale x 2 x i32> %b)
ret <vscale x 2 x i64> %load
}
define <vscale x 2 x double> @gld1d_uxtw_index_double(<vscale x 2 x i1> %pg, double* %base, <vscale x 2 x i32> %b) {
; CHECK-LABEL: gld1d_uxtw_index_double:
; CHECK: ld1d { z0.d }, p0/z, [x0, z0.d, uxtw #3]
; CHECK-NEXT: ret
%load = call <vscale x 2 x double> @llvm.aarch64.sve.ld1.gather.uxtw.index.nxv2f64(<vscale x 2 x i1> %pg,
double* %base,
<vscale x 2 x i32> %b)
ret <vscale x 2 x double> %load
}
define <vscale x 2 x double> @gld1d_sxtw_index_double(<vscale x 2 x i1> %pg, double* %base, <vscale x 2 x i32> %b) {
; CHECK-LABEL: gld1d_sxtw_index_double:
; CHECK: ld1d { z0.d }, p0/z, [x0, z0.d, sxtw #3]
; CHECK-NEXT: ret
%load = call <vscale x 2 x double> @llvm.aarch64.sve.ld1.gather.sxtw.index.nxv2f64(<vscale x 2 x i1> %pg,
double* %base,
<vscale x 2 x i32> %b)
ret <vscale x 2 x double> %load
}
;
; LD1SH, LD1SW, LD1SD: base + 32-bit scaled offset, sign (sxtw) or zero (uxtw)
; extended to 64 bits
; e.g. ld1sh z0.d, p0/z, [x0, z0.d, uxtw #1]
;
; LD1SH
define <vscale x 4 x i32> @gld1sh_s_uxtw_index(<vscale x 4 x i1> %pg, i16* %base, <vscale x 4 x i32> %b) {
; CHECK-LABEL: gld1sh_s_uxtw_index:
; CHECK: ld1sh { z0.s }, p0/z, [x0, z0.s, uxtw #1]
; CHECK-NEXT: ret
%load = call <vscale x 4 x i16> @llvm.aarch64.sve.ld1.gather.uxtw.index.nxv4i16(<vscale x 4 x i1> %pg,
i16* %base,
<vscale x 4 x i32> %b)
%res = sext <vscale x 4 x i16> %load to <vscale x 4 x i32>
ret <vscale x 4 x i32> %res
}
define <vscale x 4 x i32> @gld1sh_s_sxtw_index(<vscale x 4 x i1> %pg, i16* %base, <vscale x 4 x i32> %b) {
; CHECK-LABEL: gld1sh_s_sxtw_index:
; CHECK: ld1sh { z0.s }, p0/z, [x0, z0.s, sxtw #1]
; CHECK-NEXT: ret
%load = call <vscale x 4 x i16> @llvm.aarch64.sve.ld1.gather.sxtw.index.nxv4i16(<vscale x 4 x i1> %pg,
i16* %base,
<vscale x 4 x i32> %b)
%res = sext <vscale x 4 x i16> %load to <vscale x 4 x i32>
ret <vscale x 4 x i32> %res
}
define <vscale x 2 x i64> @gld1sh_d_uxtw_index(<vscale x 2 x i1> %pg, i16* %base, <vscale x 2 x i32> %b) {
; CHECK-LABEL: gld1sh_d_uxtw_index:
; CHECK: ld1sh { z0.d }, p0/z, [x0, z0.d, uxtw #1]
; CHECK-NEXT: ret
%load = call <vscale x 2 x i16> @llvm.aarch64.sve.ld1.gather.uxtw.index.nxv2i16(<vscale x 2 x i1> %pg,
i16* %base,
<vscale x 2 x i32> %b)
%res = sext <vscale x 2 x i16> %load to <vscale x 2 x i64>
ret <vscale x 2 x i64> %res
}
define <vscale x 2 x i64> @gld1sh_d_sxtw_index(<vscale x 2 x i1> %pg, i16* %base, <vscale x 2 x i32> %b) {
; CHECK-LABEL: gld1sh_d_sxtw_index:
; CHECK: ld1sh { z0.d }, p0/z, [x0, z0.d, sxtw #1]
; CHECK-NEXT: ret
%load = call <vscale x 2 x i16> @llvm.aarch64.sve.ld1.gather.sxtw.index.nxv2i16(<vscale x 2 x i1> %pg,
i16* %base,
<vscale x 2 x i32> %b)
%res = sext <vscale x 2 x i16> %load to <vscale x 2 x i64>
ret <vscale x 2 x i64> %res
}
; LD1SW
define <vscale x 2 x i64> @gld1sw_d_uxtw_index(<vscale x 2 x i1> %pg, i32* %base, <vscale x 2 x i32> %b) {
; CHECK-LABEL: gld1sw_d_uxtw_index:
; CHECK: ld1sw { z0.d }, p0/z, [x0, z0.d, uxtw #2]
; CHECK-NEXT: ret
%load = call <vscale x 2 x i32> @llvm.aarch64.sve.ld1.gather.uxtw.index.nxv2i32(<vscale x 2 x i1> %pg,
i32* %base,
<vscale x 2 x i32> %b)
%res = sext <vscale x 2 x i32> %load to <vscale x 2 x i64>
ret <vscale x 2 x i64> %res
}
define <vscale x 2 x i64> @gld1sw_d_sxtw_index(<vscale x 2 x i1> %pg, i32* %base, <vscale x 2 x i32> %b) {
; CHECK-LABEL: gld1sw_d_sxtw_index:
; CHECK: ld1sw { z0.d }, p0/z, [x0, z0.d, sxtw #2]
; CHECK-NEXT: ret
%load = call <vscale x 2 x i32> @llvm.aarch64.sve.ld1.gather.sxtw.index.nxv2i32(<vscale x 2 x i1> %pg,
i32* %base,
<vscale x 2 x i32> %b)
%res = sext <vscale x 2 x i32> %load to <vscale x 2 x i64>
ret <vscale x 2 x i64> %res
}
; LD1H/LD1SH
declare <vscale x 4 x i16> @llvm.aarch64.sve.ld1.gather.uxtw.index.nxv4i16(<vscale x 4 x i1>, i16*, <vscale x 4 x i32>)
declare <vscale x 4 x i16> @llvm.aarch64.sve.ld1.gather.sxtw.index.nxv4i16(<vscale x 4 x i1>, i16*, <vscale x 4 x i32>)
declare <vscale x 2 x i16> @llvm.aarch64.sve.ld1.gather.uxtw.index.nxv2i16(<vscale x 2 x i1>, i16*, <vscale x 2 x i32>)
declare <vscale x 2 x i16> @llvm.aarch64.sve.ld1.gather.sxtw.index.nxv2i16(<vscale x 2 x i1>, i16*, <vscale x 2 x i32>)
; LD1W/LD1SW
declare <vscale x 4 x i32> @llvm.aarch64.sve.ld1.gather.uxtw.index.nxv4i32(<vscale x 4 x i1>, i32*, <vscale x 4 x i32>)
declare <vscale x 4 x i32> @llvm.aarch64.sve.ld1.gather.sxtw.index.nxv4i32(<vscale x 4 x i1>, i32*, <vscale x 4 x i32>)
declare <vscale x 2 x i32> @llvm.aarch64.sve.ld1.gather.uxtw.index.nxv2i32(<vscale x 2 x i1>, i32*, <vscale x 2 x i32>)
declare <vscale x 2 x i32> @llvm.aarch64.sve.ld1.gather.sxtw.index.nxv2i32(<vscale x 2 x i1>, i32*, <vscale x 2 x i32>)
declare <vscale x 4 x float> @llvm.aarch64.sve.ld1.gather.uxtw.index.nxv4f32(<vscale x 4 x i1>, float*, <vscale x 4 x i32>)
declare <vscale x 4 x float> @llvm.aarch64.sve.ld1.gather.sxtw.index.nxv4f32(<vscale x 4 x i1>, float*, <vscale x 4 x i32>)
; LD1D
declare <vscale x 2 x i64> @llvm.aarch64.sve.ld1.gather.uxtw.index.nxv2i64(<vscale x 2 x i1>, i64*, <vscale x 2 x i32>)
declare <vscale x 2 x i64> @llvm.aarch64.sve.ld1.gather.sxtw.index.nxv2i64(<vscale x 2 x i1>, i64*, <vscale x 2 x i32>)
declare <vscale x 2 x double> @llvm.aarch64.sve.ld1.gather.uxtw.index.nxv2f64(<vscale x 2 x i1>, double*, <vscale x 2 x i32>)
declare <vscale x 2 x double> @llvm.aarch64.sve.ld1.gather.sxtw.index.nxv2f64(<vscale x 2 x i1>, double*, <vscale x 2 x i32>)