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
|
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=riscv32 -mattr=+m,+v -riscv-v-vector-bits-min=128 -riscv-v-fixed-length-vector-lmul-max=8 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,LMULMAX8
; RUN: llc -mtriple=riscv64 -mattr=+m,+v -riscv-v-vector-bits-min=128 -riscv-v-fixed-length-vector-lmul-max=8 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,LMULMAX8
; RUN: llc -mtriple=riscv32 -mattr=+m,+v -riscv-v-vector-bits-min=128 -riscv-v-fixed-length-vector-lmul-max=2 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,LMULMAX2
; RUN: llc -mtriple=riscv64 -mattr=+m,+v -riscv-v-vector-bits-min=128 -riscv-v-fixed-length-vector-lmul-max=2 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,LMULMAX2
; RUN: llc -mtriple=riscv32 -mattr=+m,+v -riscv-v-vector-bits-min=128 -riscv-v-fixed-length-vector-lmul-max=1 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,LMULMAX1
; RUN: llc -mtriple=riscv64 -mattr=+m,+v -riscv-v-vector-bits-min=128 -riscv-v-fixed-length-vector-lmul-max=1 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,LMULMAX1
define void @sext_v4i8_v4i32(<4 x i8>* %x, <4 x i32>* %z) {
; CHECK-LABEL: sext_v4i8_v4i32:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, mu
; CHECK-NEXT: vle8.v v8, (a0)
; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, mu
; CHECK-NEXT: vsext.vf4 v9, v8
; CHECK-NEXT: vse32.v v9, (a1)
; CHECK-NEXT: ret
%a = load <4 x i8>, <4 x i8>* %x
%b = sext <4 x i8> %a to <4 x i32>
store <4 x i32> %b, <4 x i32>* %z
ret void
}
define void @zext_v4i8_v4i32(<4 x i8>* %x, <4 x i32>* %z) {
; CHECK-LABEL: zext_v4i8_v4i32:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, mu
; CHECK-NEXT: vle8.v v8, (a0)
; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, mu
; CHECK-NEXT: vzext.vf4 v9, v8
; CHECK-NEXT: vse32.v v9, (a1)
; CHECK-NEXT: ret
%a = load <4 x i8>, <4 x i8>* %x
%b = zext <4 x i8> %a to <4 x i32>
store <4 x i32> %b, <4 x i32>* %z
ret void
}
define void @sext_v8i8_v8i32(<8 x i8>* %x, <8 x i32>* %z) {
; LMULMAX8-LABEL: sext_v8i8_v8i32:
; LMULMAX8: # %bb.0:
; LMULMAX8-NEXT: vsetivli zero, 8, e8, mf2, ta, mu
; LMULMAX8-NEXT: vle8.v v8, (a0)
; LMULMAX8-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; LMULMAX8-NEXT: vsext.vf4 v10, v8
; LMULMAX8-NEXT: vse32.v v10, (a1)
; LMULMAX8-NEXT: ret
;
; LMULMAX2-LABEL: sext_v8i8_v8i32:
; LMULMAX2: # %bb.0:
; LMULMAX2-NEXT: vsetivli zero, 8, e8, mf2, ta, mu
; LMULMAX2-NEXT: vle8.v v8, (a0)
; LMULMAX2-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; LMULMAX2-NEXT: vsext.vf4 v10, v8
; LMULMAX2-NEXT: vse32.v v10, (a1)
; LMULMAX2-NEXT: ret
;
; LMULMAX1-LABEL: sext_v8i8_v8i32:
; LMULMAX1: # %bb.0:
; LMULMAX1-NEXT: vsetivli zero, 8, e8, mf2, ta, mu
; LMULMAX1-NEXT: vle8.v v8, (a0)
; LMULMAX1-NEXT: vsetivli zero, 4, e8, mf2, ta, mu
; LMULMAX1-NEXT: vslidedown.vi v9, v8, 4
; LMULMAX1-NEXT: vsetivli zero, 4, e32, m1, ta, mu
; LMULMAX1-NEXT: vsext.vf4 v10, v9
; LMULMAX1-NEXT: vsext.vf4 v9, v8
; LMULMAX1-NEXT: addi a0, a1, 16
; LMULMAX1-NEXT: vse32.v v10, (a0)
; LMULMAX1-NEXT: vse32.v v9, (a1)
; LMULMAX1-NEXT: ret
%a = load <8 x i8>, <8 x i8>* %x
%b = sext <8 x i8> %a to <8 x i32>
store <8 x i32> %b, <8 x i32>* %z
ret void
}
define void @sext_v32i8_v32i32(<32 x i8>* %x, <32 x i32>* %z) {
; LMULMAX8-LABEL: sext_v32i8_v32i32:
; LMULMAX8: # %bb.0:
; LMULMAX8-NEXT: li a2, 32
; LMULMAX8-NEXT: vsetvli zero, a2, e8, m2, ta, mu
; LMULMAX8-NEXT: vle8.v v8, (a0)
; LMULMAX8-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; LMULMAX8-NEXT: vsext.vf4 v16, v8
; LMULMAX8-NEXT: vse32.v v16, (a1)
; LMULMAX8-NEXT: ret
;
; LMULMAX2-LABEL: sext_v32i8_v32i32:
; LMULMAX2: # %bb.0:
; LMULMAX2-NEXT: li a2, 32
; LMULMAX2-NEXT: vsetvli zero, a2, e8, m2, ta, mu
; LMULMAX2-NEXT: vle8.v v8, (a0)
; LMULMAX2-NEXT: vsetivli zero, 8, e8, m1, ta, mu
; LMULMAX2-NEXT: vslidedown.vi v10, v8, 8
; LMULMAX2-NEXT: vsetivli zero, 8, e32, m2, ta, mu
; LMULMAX2-NEXT: vsext.vf4 v12, v10
; LMULMAX2-NEXT: vsetivli zero, 16, e8, m2, ta, mu
; LMULMAX2-NEXT: vslidedown.vi v10, v8, 16
; LMULMAX2-NEXT: vsetivli zero, 8, e8, m1, ta, mu
; LMULMAX2-NEXT: vslidedown.vi v14, v10, 8
; LMULMAX2-NEXT: vsetivli zero, 8, e32, m2, ta, mu
; LMULMAX2-NEXT: vsext.vf4 v16, v14
; LMULMAX2-NEXT: vsext.vf4 v14, v8
; LMULMAX2-NEXT: vsext.vf4 v8, v10
; LMULMAX2-NEXT: addi a0, a1, 64
; LMULMAX2-NEXT: vse32.v v8, (a0)
; LMULMAX2-NEXT: vse32.v v14, (a1)
; LMULMAX2-NEXT: addi a0, a1, 96
; LMULMAX2-NEXT: vse32.v v16, (a0)
; LMULMAX2-NEXT: addi a0, a1, 32
; LMULMAX2-NEXT: vse32.v v12, (a0)
; LMULMAX2-NEXT: ret
;
; LMULMAX1-LABEL: sext_v32i8_v32i32:
; LMULMAX1: # %bb.0:
; LMULMAX1-NEXT: vsetivli zero, 16, e8, m1, ta, mu
; LMULMAX1-NEXT: addi a2, a0, 16
; LMULMAX1-NEXT: vle8.v v8, (a2)
; LMULMAX1-NEXT: vle8.v v9, (a0)
; LMULMAX1-NEXT: vsetivli zero, 4, e8, mf2, ta, mu
; LMULMAX1-NEXT: vslidedown.vi v10, v8, 4
; LMULMAX1-NEXT: vsetivli zero, 4, e32, m1, ta, mu
; LMULMAX1-NEXT: vsext.vf4 v11, v10
; LMULMAX1-NEXT: vsetivli zero, 8, e8, m1, ta, mu
; LMULMAX1-NEXT: vslidedown.vi v10, v8, 8
; LMULMAX1-NEXT: vsetivli zero, 4, e8, mf2, ta, mu
; LMULMAX1-NEXT: vslidedown.vi v12, v10, 4
; LMULMAX1-NEXT: vsetivli zero, 4, e32, m1, ta, mu
; LMULMAX1-NEXT: vsext.vf4 v13, v12
; LMULMAX1-NEXT: vsetivli zero, 4, e8, mf2, ta, mu
; LMULMAX1-NEXT: vslidedown.vi v12, v9, 4
; LMULMAX1-NEXT: vsetivli zero, 4, e32, m1, ta, mu
; LMULMAX1-NEXT: vsext.vf4 v14, v12
; LMULMAX1-NEXT: vsetivli zero, 8, e8, m1, ta, mu
; LMULMAX1-NEXT: vslidedown.vi v12, v9, 8
; LMULMAX1-NEXT: vsetivli zero, 4, e8, mf2, ta, mu
; LMULMAX1-NEXT: vslidedown.vi v15, v12, 4
; LMULMAX1-NEXT: vsetivli zero, 4, e32, m1, ta, mu
; LMULMAX1-NEXT: vsext.vf4 v16, v15
; LMULMAX1-NEXT: vsext.vf4 v15, v10
; LMULMAX1-NEXT: vsext.vf4 v10, v12
; LMULMAX1-NEXT: vsext.vf4 v12, v8
; LMULMAX1-NEXT: vsext.vf4 v8, v9
; LMULMAX1-NEXT: addi a0, a1, 32
; LMULMAX1-NEXT: vse32.v v10, (a0)
; LMULMAX1-NEXT: vse32.v v8, (a1)
; LMULMAX1-NEXT: addi a0, a1, 96
; LMULMAX1-NEXT: vse32.v v15, (a0)
; LMULMAX1-NEXT: addi a0, a1, 64
; LMULMAX1-NEXT: vse32.v v12, (a0)
; LMULMAX1-NEXT: addi a0, a1, 48
; LMULMAX1-NEXT: vse32.v v16, (a0)
; LMULMAX1-NEXT: addi a0, a1, 16
; LMULMAX1-NEXT: vse32.v v14, (a0)
; LMULMAX1-NEXT: addi a0, a1, 112
; LMULMAX1-NEXT: vse32.v v13, (a0)
; LMULMAX1-NEXT: addi a0, a1, 80
; LMULMAX1-NEXT: vse32.v v11, (a0)
; LMULMAX1-NEXT: ret
%a = load <32 x i8>, <32 x i8>* %x
%b = sext <32 x i8> %a to <32 x i32>
store <32 x i32> %b, <32 x i32>* %z
ret void
}
define void @trunc_v4i8_v4i32(<4 x i32>* %x, <4 x i8>* %z) {
; CHECK-LABEL: trunc_v4i8_v4i32:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, mu
; CHECK-NEXT: vle32.v v8, (a0)
; CHECK-NEXT: vsetvli zero, zero, e16, mf2, ta, mu
; CHECK-NEXT: vnsrl.wx v8, v8, zero
; CHECK-NEXT: vsetvli zero, zero, e8, mf4, ta, mu
; CHECK-NEXT: vnsrl.wx v8, v8, zero
; CHECK-NEXT: vse8.v v8, (a1)
; CHECK-NEXT: ret
%a = load <4 x i32>, <4 x i32>* %x
%b = trunc <4 x i32> %a to <4 x i8>
store <4 x i8> %b, <4 x i8>* %z
ret void
}
define void @trunc_v8i8_v8i32(<8 x i32>* %x, <8 x i8>* %z) {
; LMULMAX8-LABEL: trunc_v8i8_v8i32:
; LMULMAX8: # %bb.0:
; LMULMAX8-NEXT: vsetivli zero, 8, e32, m2, ta, mu
; LMULMAX8-NEXT: vle32.v v8, (a0)
; LMULMAX8-NEXT: vsetvli zero, zero, e16, m1, ta, mu
; LMULMAX8-NEXT: vnsrl.wx v10, v8, zero
; LMULMAX8-NEXT: vsetvli zero, zero, e8, mf2, ta, mu
; LMULMAX8-NEXT: vnsrl.wx v8, v10, zero
; LMULMAX8-NEXT: vse8.v v8, (a1)
; LMULMAX8-NEXT: ret
;
; LMULMAX2-LABEL: trunc_v8i8_v8i32:
; LMULMAX2: # %bb.0:
; LMULMAX2-NEXT: vsetivli zero, 8, e32, m2, ta, mu
; LMULMAX2-NEXT: vle32.v v8, (a0)
; LMULMAX2-NEXT: vsetvli zero, zero, e16, m1, ta, mu
; LMULMAX2-NEXT: vnsrl.wx v10, v8, zero
; LMULMAX2-NEXT: vsetvli zero, zero, e8, mf2, ta, mu
; LMULMAX2-NEXT: vnsrl.wx v8, v10, zero
; LMULMAX2-NEXT: vse8.v v8, (a1)
; LMULMAX2-NEXT: ret
;
; LMULMAX1-LABEL: trunc_v8i8_v8i32:
; LMULMAX1: # %bb.0:
; LMULMAX1-NEXT: vsetivli zero, 4, e32, m1, ta, mu
; LMULMAX1-NEXT: vle32.v v8, (a0)
; LMULMAX1-NEXT: addi a0, a0, 16
; LMULMAX1-NEXT: vle32.v v9, (a0)
; LMULMAX1-NEXT: vsetvli zero, zero, e16, mf2, ta, mu
; LMULMAX1-NEXT: vnsrl.wx v8, v8, zero
; LMULMAX1-NEXT: vsetvli zero, zero, e8, mf4, ta, mu
; LMULMAX1-NEXT: vnsrl.wx v8, v8, zero
; LMULMAX1-NEXT: vsetvli zero, zero, e16, mf2, ta, mu
; LMULMAX1-NEXT: vnsrl.wx v9, v9, zero
; LMULMAX1-NEXT: vsetvli zero, zero, e8, mf4, ta, mu
; LMULMAX1-NEXT: vnsrl.wx v9, v9, zero
; LMULMAX1-NEXT: vsetivli zero, 8, e8, mf2, tu, mu
; LMULMAX1-NEXT: vslideup.vi v8, v9, 4
; LMULMAX1-NEXT: vse8.v v8, (a1)
; LMULMAX1-NEXT: ret
%a = load <8 x i32>, <8 x i32>* %x
%b = trunc <8 x i32> %a to <8 x i8>
store <8 x i8> %b, <8 x i8>* %z
ret void
}
|
