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
|
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: sed 's/iXLen/i32/g' %s | llc -mtriple=riscv32 -mattr=+v,+zvfbfmin,+xsfvfwmaccqqq \
; RUN: -verify-machineinstrs | FileCheck %s --check-prefixes=CHECK
; RUN: sed 's/iXLen/i64/g' %s | llc -mtriple=riscv64 -mattr=+v,+zvfbfmin,+xsfvfwmaccqqq \
; RUN: -verify-machineinstrs | FileCheck %s --check-prefixes=CHECK
declare <vscale x 1 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv1f32.nxv4bf16.nxv1bf16.iXLen(
<vscale x 1 x float>,
<vscale x 4 x bfloat>,
<vscale x 1 x bfloat>,
iXLen, iXLen);
define <vscale x 1 x float> @intrinsic_vfwmacc_4x4x4_tu_f32mf2(<vscale x 1 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 1 x bfloat> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_tu_f32mf2:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, tu, ma
; CHECK-NEXT: sf.vfwmacc.4x4x4 v8, v9, v10
; CHECK-NEXT: ret
entry:
%a = call <vscale x 1 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv1f32.nxv4bf16.nxv1bf16.iXLen(
<vscale x 1 x float> %0,
<vscale x 4 x bfloat> %1,
<vscale x 1 x bfloat> %2,
iXLen %3, iXLen 2)
ret <vscale x 1 x float> %a
}
define <vscale x 1 x float> @intrinsic_vfwmacc_4x4x4_ta_f32mf2(<vscale x 1 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 1 x bfloat> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_ta_f32mf2:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
; CHECK-NEXT: sf.vfwmacc.4x4x4 v8, v9, v10
; CHECK-NEXT: ret
entry:
%a = call <vscale x 1 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv1f32.nxv4bf16.nxv1bf16.iXLen(
<vscale x 1 x float> %0,
<vscale x 4 x bfloat> %1,
<vscale x 1 x bfloat> %2,
iXLen %3, iXLen 3)
ret <vscale x 1 x float> %a
}
declare <vscale x 2 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv2f32.nxv4bf16.nxv2bf16.iXLen(
<vscale x 2 x float>,
<vscale x 4 x bfloat>,
<vscale x 2 x bfloat>,
iXLen, iXLen);
define <vscale x 2 x float> @intrinsic_vfwmacc_4x4x4_tu_f32m1(<vscale x 2 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 2 x bfloat> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_tu_f32m1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, tu, ma
; CHECK-NEXT: sf.vfwmacc.4x4x4 v8, v9, v10
; CHECK-NEXT: ret
entry:
%a = call <vscale x 2 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv2f32.nxv4bf16.nxv2bf16.iXLen(
<vscale x 2 x float> %0,
<vscale x 4 x bfloat> %1,
<vscale x 2 x bfloat> %2,
iXLen %3, iXLen 2)
ret <vscale x 2 x float> %a
}
define <vscale x 2 x float> @intrinsic_vfwmacc_4x4x4_ta_f32m1(<vscale x 2 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 2 x bfloat> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_ta_f32m1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
; CHECK-NEXT: sf.vfwmacc.4x4x4 v8, v9, v10
; CHECK-NEXT: ret
entry:
%a = call <vscale x 2 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv2f32.nxv4bf16.nxv2bf16.iXLen(
<vscale x 2 x float> %0,
<vscale x 4 x bfloat> %1,
<vscale x 2 x bfloat> %2,
iXLen %3, iXLen 3)
ret <vscale x 2 x float> %a
}
declare <vscale x 4 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv4f32.nxv4bf16.nxv4bf16.iXLen(
<vscale x 4 x float>,
<vscale x 4 x bfloat>,
<vscale x 4 x bfloat>,
iXLen, iXLen);
define <vscale x 4 x float> @intrinsic_vfwmacc_4x4x4_tu_f32m2(<vscale x 4 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 4 x bfloat> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_tu_f32m2:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli zero, a0, e16, m1, tu, ma
; CHECK-NEXT: sf.vfwmacc.4x4x4 v8, v10, v11
; CHECK-NEXT: ret
entry:
%a = call <vscale x 4 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv4f32.nxv4bf16.nxv4bf16.iXLen(
<vscale x 4 x float> %0,
<vscale x 4 x bfloat> %1,
<vscale x 4 x bfloat> %2,
iXLen %3, iXLen 2)
ret <vscale x 4 x float> %a
}
define <vscale x 4 x float> @intrinsic_vfwmacc_4x4x4_ta_f32m2(<vscale x 4 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 4 x bfloat> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_ta_f32m2:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, ma
; CHECK-NEXT: sf.vfwmacc.4x4x4 v8, v10, v11
; CHECK-NEXT: ret
entry:
%a = call <vscale x 4 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv4f32.nxv4bf16.nxv4bf16.iXLen(
<vscale x 4 x float> %0,
<vscale x 4 x bfloat> %1,
<vscale x 4 x bfloat> %2,
iXLen %3, iXLen 3)
ret <vscale x 4 x float> %a
}
declare <vscale x 8 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv8f32.nxv4bf16.nxv8bf16.iXLen(
<vscale x 8 x float>,
<vscale x 4 x bfloat>,
<vscale x 8 x bfloat>,
iXLen, iXLen);
define <vscale x 8 x float> @intrinsic_vfwmacc_4x4x4_tu_f32m4(<vscale x 8 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 8 x bfloat> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_tu_f32m4:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli zero, a0, e16, m2, tu, ma
; CHECK-NEXT: sf.vfwmacc.4x4x4 v8, v12, v14
; CHECK-NEXT: ret
entry:
%a = call <vscale x 8 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv8f32.nxv4bf16.nxv8bf16.iXLen(
<vscale x 8 x float> %0,
<vscale x 4 x bfloat> %1,
<vscale x 8 x bfloat> %2,
iXLen %3, iXLen 2)
ret <vscale x 8 x float> %a
}
define <vscale x 8 x float> @intrinsic_vfwmacc_4x4x4_ta_f32m4(<vscale x 8 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 8 x bfloat> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_ta_f32m4:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, ma
; CHECK-NEXT: sf.vfwmacc.4x4x4 v8, v12, v14
; CHECK-NEXT: ret
entry:
%a = call <vscale x 8 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv8f32.nxv4bf16.nxv8bf16.iXLen(
<vscale x 8 x float> %0,
<vscale x 4 x bfloat> %1,
<vscale x 8 x bfloat> %2,
iXLen %3, iXLen 3)
ret <vscale x 8 x float> %a
}
declare <vscale x 16 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv16f32.nxv4bf16.nxv16bf16.iXLen(
<vscale x 16 x float>,
<vscale x 4 x bfloat>,
<vscale x 16 x bfloat>,
iXLen, iXLen);
define <vscale x 16 x float> @intrinsic_vfwmacc_4x4x4_tu_f32m8(<vscale x 16 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 16 x bfloat> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_tu_f32m8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli zero, a0, e16, m4, tu, ma
; CHECK-NEXT: sf.vfwmacc.4x4x4 v8, v16, v20
; CHECK-NEXT: ret
entry:
%a = call <vscale x 16 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv16f32.nxv4bf16.nxv16bf16.iXLen(
<vscale x 16 x float> %0,
<vscale x 4 x bfloat> %1,
<vscale x 16 x bfloat> %2,
iXLen %3, iXLen 2)
ret <vscale x 16 x float> %a
}
define <vscale x 16 x float> @intrinsic_vfwmacc_4x4x4_ta_f32m8(<vscale x 16 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 16 x bfloat> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_ta_f32m8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
; CHECK-NEXT: sf.vfwmacc.4x4x4 v8, v16, v20
; CHECK-NEXT: ret
entry:
%a = call <vscale x 16 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv16f32.nxv4bf16.nxv16bf16.iXLen(
<vscale x 16 x float> %0,
<vscale x 4 x bfloat> %1,
<vscale x 16 x bfloat> %2,
iXLen %3, iXLen 3)
ret <vscale x 16 x float> %a
}
|
