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
256
257
258
259
260
261
262
|
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 2
; RUN: llc -mtriple=aarch64--linux-gnu -mattr=+sve < %s | FileCheck %s
;
; Masked Loads
;
define <vscale x 2 x i64> @masked_zload_nxv2i8(ptr %src, <vscale x 2 x i1> %mask) {
; CHECK-LABEL: masked_zload_nxv2i8:
; CHECK: // %bb.0:
; CHECK-NEXT: ld1b { z0.d }, p0/z, [x0]
; CHECK-NEXT: ret
%load = call <vscale x 2 x i8> @llvm.masked.load.nxv2i8(ptr %src, i32 1, <vscale x 2 x i1> %mask, <vscale x 2 x i8> undef)
%ext = zext <vscale x 2 x i8> %load to <vscale x 2 x i64>
ret <vscale x 2 x i64> %ext
}
define <vscale x 2 x i64> @masked_zload_nxv2i16(ptr %src, <vscale x 2 x i1> %mask) {
; CHECK-LABEL: masked_zload_nxv2i16:
; CHECK: // %bb.0:
; CHECK-NEXT: ld1h { z0.d }, p0/z, [x0]
; CHECK-NEXT: ret
%load = call <vscale x 2 x i16> @llvm.masked.load.nxv2i16(ptr %src, i32 1, <vscale x 2 x i1> %mask, <vscale x 2 x i16> undef)
%ext = zext <vscale x 2 x i16> %load to <vscale x 2 x i64>
ret <vscale x 2 x i64> %ext
}
define <vscale x 2 x i64> @masked_zload_nxv2i32(ptr %src, <vscale x 2 x i1> %mask) {
; CHECK-LABEL: masked_zload_nxv2i32:
; CHECK: // %bb.0:
; CHECK-NEXT: ld1w { z0.d }, p0/z, [x0]
; CHECK-NEXT: ret
%load = call <vscale x 2 x i32> @llvm.masked.load.nxv2i32(ptr %src, i32 1, <vscale x 2 x i1> %mask, <vscale x 2 x i32> undef)
%ext = zext <vscale x 2 x i32> %load to <vscale x 2 x i64>
ret <vscale x 2 x i64> %ext
}
define <vscale x 4 x i32> @masked_zload_nxv4i8(ptr %src, <vscale x 4 x i1> %mask) {
; CHECK-LABEL: masked_zload_nxv4i8:
; CHECK: // %bb.0:
; CHECK-NEXT: ld1b { z0.s }, p0/z, [x0]
; CHECK-NEXT: ret
%load = call <vscale x 4 x i8> @llvm.masked.load.nxv4i8(ptr %src, i32 1, <vscale x 4 x i1> %mask, <vscale x 4 x i8> undef)
%ext = zext <vscale x 4 x i8> %load to <vscale x 4 x i32>
ret <vscale x 4 x i32> %ext
}
define <vscale x 4 x i32> @masked_zload_nxv4i16(ptr %src, <vscale x 4 x i1> %mask) {
; CHECK-LABEL: masked_zload_nxv4i16:
; CHECK: // %bb.0:
; CHECK-NEXT: ld1h { z0.s }, p0/z, [x0]
; CHECK-NEXT: ret
%load = call <vscale x 4 x i16> @llvm.masked.load.nxv4i16(ptr %src, i32 1, <vscale x 4 x i1> %mask, <vscale x 4 x i16> undef)
%ext = zext <vscale x 4 x i16> %load to <vscale x 4 x i32>
ret <vscale x 4 x i32> %ext
}
define <vscale x 8 x i16> @masked_zload_nxv8i8(ptr %src, <vscale x 8 x i1> %mask) {
; CHECK-LABEL: masked_zload_nxv8i8:
; CHECK: // %bb.0:
; CHECK-NEXT: ld1b { z0.h }, p0/z, [x0]
; CHECK-NEXT: ret
%load = call <vscale x 8 x i8> @llvm.masked.load.nxv8i8(ptr %src, i32 1, <vscale x 8 x i1> %mask, <vscale x 8 x i8> undef)
%ext = zext <vscale x 8 x i8> %load to <vscale x 8 x i16>
ret <vscale x 8 x i16> %ext
}
define <vscale x 2 x i64> @masked_zload_passthru(ptr %src, <vscale x 2 x i1> %mask, <vscale x 2 x i32> %passthru) {
; CHECK-LABEL: masked_zload_passthru:
; CHECK: // %bb.0:
; CHECK-NEXT: ld1w { z1.d }, p0/z, [x0]
; CHECK-NEXT: and z0.d, z0.d, #0xffffffff
; CHECK-NEXT: mov z0.d, p0/m, z1.d
; CHECK-NEXT: ret
%load = call <vscale x 2 x i32> @llvm.masked.load.nxv2i32(ptr %src, i32 1, <vscale x 2 x i1> %mask, <vscale x 2 x i32> %passthru)
%ext = zext <vscale x 2 x i32> %load to <vscale x 2 x i64>
ret <vscale x 2 x i64> %ext
}
; Return type requires splitting
define <vscale x 8 x i64> @masked_zload_nxv8i16(ptr %a, <vscale x 8 x i1> %mask) {
; CHECK-LABEL: masked_zload_nxv8i16:
; CHECK: // %bb.0:
; CHECK-NEXT: ld1h { z0.h }, p0/z, [x0]
; CHECK-NEXT: uunpklo z1.s, z0.h
; CHECK-NEXT: uunpkhi z3.s, z0.h
; CHECK-NEXT: uunpklo z0.d, z1.s
; CHECK-NEXT: uunpkhi z1.d, z1.s
; CHECK-NEXT: uunpklo z2.d, z3.s
; CHECK-NEXT: uunpkhi z3.d, z3.s
; CHECK-NEXT: ret
%load = call <vscale x 8 x i16> @llvm.masked.load.nxv8i16(ptr %a, i32 2, <vscale x 8 x i1> %mask, <vscale x 8 x i16> undef)
%ext = zext <vscale x 8 x i16> %load to <vscale x 8 x i64>
ret <vscale x 8 x i64> %ext
}
; Masked load requires promotion
define <vscale x 2 x double> @masked_zload_2i16_2f64(ptr noalias %in, <vscale x 2 x i1> %mask) {
; CHECK-LABEL: masked_zload_2i16_2f64:
; CHECK: // %bb.0:
; CHECK-NEXT: ld1h { z0.d }, p0/z, [x0]
; CHECK-NEXT: ptrue p0.d
; CHECK-NEXT: ucvtf z0.d, p0/m, z0.d
; CHECK-NEXT: ret
%wide.load = call <vscale x 2 x i16> @llvm.masked.load.nxv2i16(ptr %in, i32 2, <vscale x 2 x i1> %mask, <vscale x 2 x i16> undef)
%zext = zext <vscale x 2 x i16> %wide.load to <vscale x 2 x i32>
%res = uitofp <vscale x 2 x i32> %zext to <vscale x 2 x double>
ret <vscale x 2 x double> %res
}
; Extending loads from unpacked to wide illegal types
define <vscale x 4 x i64> @masked_zload_4i8_4i64(ptr %a, <vscale x 4 x i1> %b) {
; CHECK-LABEL: masked_zload_4i8_4i64:
; CHECK: // %bb.0:
; CHECK-NEXT: ld1b { z1.s }, p0/z, [x0]
; CHECK-NEXT: uunpklo z0.d, z1.s
; CHECK-NEXT: uunpkhi z1.d, z1.s
; CHECK-NEXT: ret
%aval = call <vscale x 4 x i8> @llvm.masked.load.nxv4i8(ptr %a, i32 16, <vscale x 4 x i1> %b, <vscale x 4 x i8> zeroinitializer)
%aext = zext <vscale x 4 x i8> %aval to <vscale x 4 x i64>
ret <vscale x 4 x i64> %aext
}
define <vscale x 4 x i64> @masked_zload_4i16_4i64(ptr %a, <vscale x 4 x i1> %b) {
; CHECK-LABEL: masked_zload_4i16_4i64:
; CHECK: // %bb.0:
; CHECK-NEXT: ld1h { z1.s }, p0/z, [x0]
; CHECK-NEXT: uunpklo z0.d, z1.s
; CHECK-NEXT: uunpkhi z1.d, z1.s
; CHECK-NEXT: ret
%aval = call <vscale x 4 x i16> @llvm.masked.load.nxv4i16(ptr %a, i32 16, <vscale x 4 x i1> %b, <vscale x 4 x i16> zeroinitializer)
%aext = zext <vscale x 4 x i16> %aval to <vscale x 4 x i64>
ret <vscale x 4 x i64> %aext
}
define <vscale x 8 x i32> @masked_zload_8i8_8i32(ptr %a, <vscale x 8 x i1> %b) {
; CHECK-LABEL: masked_zload_8i8_8i32:
; CHECK: // %bb.0:
; CHECK-NEXT: ld1b { z1.h }, p0/z, [x0]
; CHECK-NEXT: uunpklo z0.s, z1.h
; CHECK-NEXT: uunpkhi z1.s, z1.h
; CHECK-NEXT: ret
%aval = call <vscale x 8 x i8> @llvm.masked.load.nxv8i8(ptr %a, i32 16, <vscale x 8 x i1> %b, <vscale x 8 x i8> zeroinitializer)
%aext = zext <vscale x 8 x i8> %aval to <vscale x 8 x i32>
ret <vscale x 8 x i32> %aext
}
define <vscale x 8 x i64> @masked_zload_8i8_8i64(ptr %a, <vscale x 8 x i1> %b) {
; CHECK-LABEL: masked_zload_8i8_8i64:
; CHECK: // %bb.0:
; CHECK-NEXT: ld1b { z0.h }, p0/z, [x0]
; CHECK-NEXT: uunpklo z1.s, z0.h
; CHECK-NEXT: uunpkhi z3.s, z0.h
; CHECK-NEXT: uunpklo z0.d, z1.s
; CHECK-NEXT: uunpkhi z1.d, z1.s
; CHECK-NEXT: uunpklo z2.d, z3.s
; CHECK-NEXT: uunpkhi z3.d, z3.s
; CHECK-NEXT: ret
%aval = call <vscale x 8 x i8> @llvm.masked.load.nxv8i8(ptr %a, i32 16, <vscale x 8 x i1> %b, <vscale x 8 x i8> zeroinitializer)
%aext = zext <vscale x 8 x i8> %aval to <vscale x 8 x i64>
ret <vscale x 8 x i64> %aext
}
define <vscale x 4 x i64> @masked_zload_x2_4i8_4i64(ptr %a, ptr %b, <vscale x 4 x i1> %c) {
; CHECK-LABEL: masked_zload_x2_4i8_4i64:
; CHECK: // %bb.0:
; CHECK-NEXT: punpkhi p1.h, p0.b
; CHECK-NEXT: punpklo p0.h, p0.b
; CHECK-NEXT: ld1b { z1.d }, p1/z, [x0, #1, mul vl]
; CHECK-NEXT: ld1b { z2.d }, p1/z, [x1, #1, mul vl]
; CHECK-NEXT: ld1b { z0.d }, p0/z, [x0]
; CHECK-NEXT: ld1b { z3.d }, p0/z, [x1]
; CHECK-NEXT: add z1.d, z1.d, z2.d
; CHECK-NEXT: add z0.d, z0.d, z3.d
; CHECK-NEXT: ret
%aval = call <vscale x 4 x i8> @llvm.masked.load.nxv4i8(ptr %a, i32 16, <vscale x 4 x i1> %c, <vscale x 4 x i8> zeroinitializer)
%bval = call <vscale x 4 x i8> @llvm.masked.load.nxv4i8(ptr %b, i32 16, <vscale x 4 x i1> %c, <vscale x 4 x i8> zeroinitializer)
%aext = zext <vscale x 4 x i8> %aval to <vscale x 4 x i64>
%bext = zext <vscale x 4 x i8> %bval to <vscale x 4 x i64>
%res = add <vscale x 4 x i64> %aext, %bext
ret <vscale x 4 x i64> %res
}
define <vscale x 4 x i64> @masked_zload_x2_4i16_4i64(ptr %a, ptr %b, <vscale x 4 x i1> %c) {
; CHECK-LABEL: masked_zload_x2_4i16_4i64:
; CHECK: // %bb.0:
; CHECK-NEXT: punpkhi p1.h, p0.b
; CHECK-NEXT: punpklo p0.h, p0.b
; CHECK-NEXT: ld1h { z1.d }, p1/z, [x0, #1, mul vl]
; CHECK-NEXT: ld1h { z2.d }, p1/z, [x1, #1, mul vl]
; CHECK-NEXT: ld1h { z0.d }, p0/z, [x0]
; CHECK-NEXT: ld1h { z3.d }, p0/z, [x1]
; CHECK-NEXT: add z1.d, z1.d, z2.d
; CHECK-NEXT: add z0.d, z0.d, z3.d
; CHECK-NEXT: ret
%aval = call <vscale x 4 x i16> @llvm.masked.load.nxv4i16(ptr %a, i32 16, <vscale x 4 x i1> %c, <vscale x 4 x i16> zeroinitializer)
%bval = call <vscale x 4 x i16> @llvm.masked.load.nxv4i16(ptr %b, i32 16, <vscale x 4 x i1> %c, <vscale x 4 x i16> zeroinitializer)
%aext = zext <vscale x 4 x i16> %aval to <vscale x 4 x i64>
%bext = zext <vscale x 4 x i16> %bval to <vscale x 4 x i64>
%res = add <vscale x 4 x i64> %aext, %bext
ret <vscale x 4 x i64> %res
}
define <vscale x 8 x i32> @masked_zload_x2_8i8_8i32(ptr %a, ptr %b, <vscale x 8 x i1> %c) {
; CHECK-LABEL: masked_zload_x2_8i8_8i32:
; CHECK: // %bb.0:
; CHECK-NEXT: punpkhi p1.h, p0.b
; CHECK-NEXT: punpklo p0.h, p0.b
; CHECK-NEXT: ld1b { z1.s }, p1/z, [x0, #1, mul vl]
; CHECK-NEXT: ld1b { z2.s }, p1/z, [x1, #1, mul vl]
; CHECK-NEXT: ld1b { z0.s }, p0/z, [x0]
; CHECK-NEXT: ld1b { z3.s }, p0/z, [x1]
; CHECK-NEXT: add z1.s, z1.s, z2.s
; CHECK-NEXT: add z0.s, z0.s, z3.s
; CHECK-NEXT: ret
%aval = call <vscale x 8 x i8> @llvm.masked.load.nxv8i8(ptr %a, i32 16, <vscale x 8 x i1> %c, <vscale x 8 x i8> zeroinitializer)
%bval = call <vscale x 8 x i8> @llvm.masked.load.nxv8i8(ptr %b, i32 16, <vscale x 8 x i1> %c, <vscale x 8 x i8> zeroinitializer)
%aext = zext <vscale x 8 x i8> %aval to <vscale x 8 x i32>
%bext = zext <vscale x 8 x i8> %bval to <vscale x 8 x i32>
%res = add <vscale x 8 x i32> %aext, %bext
ret <vscale x 8 x i32> %res
}
define <vscale x 8 x i64> @masked_zload_x2_8i8_8i64(ptr %a, ptr %b, <vscale x 8 x i1> %c) {
; CHECK-LABEL: masked_zload_x2_8i8_8i64:
; CHECK: // %bb.0:
; CHECK-NEXT: punpkhi p1.h, p0.b
; CHECK-NEXT: punpklo p0.h, p0.b
; CHECK-NEXT: punpkhi p2.h, p1.b
; CHECK-NEXT: punpklo p1.h, p1.b
; CHECK-NEXT: punpkhi p3.h, p0.b
; CHECK-NEXT: ld1b { z3.d }, p2/z, [x0, #3, mul vl]
; CHECK-NEXT: ld1b { z5.d }, p2/z, [x1, #3, mul vl]
; CHECK-NEXT: punpklo p0.h, p0.b
; CHECK-NEXT: ld1b { z2.d }, p1/z, [x0, #2, mul vl]
; CHECK-NEXT: ld1b { z6.d }, p1/z, [x1, #2, mul vl]
; CHECK-NEXT: ld1b { z1.d }, p3/z, [x0, #1, mul vl]
; CHECK-NEXT: ld1b { z7.d }, p3/z, [x1, #1, mul vl]
; CHECK-NEXT: ld1b { z0.d }, p0/z, [x0]
; CHECK-NEXT: ld1b { z4.d }, p0/z, [x1]
; CHECK-NEXT: add z3.d, z3.d, z5.d
; CHECK-NEXT: add z2.d, z2.d, z6.d
; CHECK-NEXT: add z1.d, z1.d, z7.d
; CHECK-NEXT: add z0.d, z0.d, z4.d
; CHECK-NEXT: ret
%aval = call <vscale x 8 x i8> @llvm.masked.load.nxv8i8(ptr %a, i32 16, <vscale x 8 x i1> %c, <vscale x 8 x i8> zeroinitializer)
%bval = call <vscale x 8 x i8> @llvm.masked.load.nxv8i8(ptr %b, i32 16, <vscale x 8 x i1> %c, <vscale x 8 x i8> zeroinitializer)
%aext = zext <vscale x 8 x i8> %aval to <vscale x 8 x i64>
%bext = zext <vscale x 8 x i8> %bval to <vscale x 8 x i64>
%res = add <vscale x 8 x i64> %aext, %bext
ret <vscale x 8 x i64> %res
}
declare <vscale x 2 x i8> @llvm.masked.load.nxv2i8(ptr, i32, <vscale x 2 x i1>, <vscale x 2 x i8>)
declare <vscale x 2 x i16> @llvm.masked.load.nxv2i16(ptr, i32, <vscale x 2 x i1>, <vscale x 2 x i16>)
declare <vscale x 2 x i32> @llvm.masked.load.nxv2i32(ptr, i32, <vscale x 2 x i1>, <vscale x 2 x i32>)
declare <vscale x 4 x i8> @llvm.masked.load.nxv4i8(ptr, i32, <vscale x 4 x i1>, <vscale x 4 x i8>)
declare <vscale x 4 x i16> @llvm.masked.load.nxv4i16(ptr, i32, <vscale x 4 x i1>, <vscale x 4 x i16>)
declare <vscale x 8 x i8> @llvm.masked.load.nxv8i8(ptr, i32, <vscale x 8 x i1>, <vscale x 8 x i8>)
declare <vscale x 8 x i16> @llvm.masked.load.nxv8i16(ptr, i32, <vscale x 8 x i1>, <vscale x 8 x i16>)
|
