Skip to content

Commit 81c72b6

Browse files
4vtomat4vtomat
committed
[Auto-gen] Update bfloat16 documents under ../auto-generated. (make git-commit-autogen-zvfh-doc)
1 parent ae95472 commit 81c72b6

28 files changed

+65338
-0
lines changed

‎auto-generated/zvfh/intrinsic_funcs.adoc‎

Lines changed: 6148 additions & 0 deletions
Large diffs are not rendered by default.

‎auto-generated/zvfh/intrinsic_funcs/00_vector_loads_and_stores_intrinsics.adoc‎

Lines changed: 262 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,262 @@
1+
2+
=== Vector Loads and Stores Intrinsics
3+
4+
[[vector-unit-stride-load]]
5+
==== Vector Unit-Stride Load Intrinsics
6+
7+
[,c]
8+
----
9+
vfloat16mf4_t __riscv_vle16_v_f16mf4(const _Float16 *rs1, size_t vl);
10+
vfloat16mf2_t __riscv_vle16_v_f16mf2(const _Float16 *rs1, size_t vl);
11+
vfloat16m1_t __riscv_vle16_v_f16m1(const _Float16 *rs1, size_t vl);
12+
vfloat16m2_t __riscv_vle16_v_f16m2(const _Float16 *rs1, size_t vl);
13+
vfloat16m4_t __riscv_vle16_v_f16m4(const _Float16 *rs1, size_t vl);
14+
vfloat16m8_t __riscv_vle16_v_f16m8(const _Float16 *rs1, size_t vl);
15+
// masked functions
16+
vfloat16mf4_t __riscv_vle16_v_f16mf4_m(vbool64_t vm, const _Float16 *rs1,
17+
size_t vl);
18+
vfloat16mf2_t __riscv_vle16_v_f16mf2_m(vbool32_t vm, const _Float16 *rs1,
19+
size_t vl);
20+
vfloat16m1_t __riscv_vle16_v_f16m1_m(vbool16_t vm, const _Float16 *rs1,
21+
size_t vl);
22+
vfloat16m2_t __riscv_vle16_v_f16m2_m(vbool8_t vm, const _Float16 *rs1,
23+
size_t vl);
24+
vfloat16m4_t __riscv_vle16_v_f16m4_m(vbool4_t vm, const _Float16 *rs1,
25+
size_t vl);
26+
vfloat16m8_t __riscv_vle16_v_f16m8_m(vbool2_t vm, const _Float16 *rs1,
27+
size_t vl);
28+
----
29+
30+
[[vector-unit-stride-store]]
31+
==== Vector Unit-Stride Store Intrinsics
32+
33+
[,c]
34+
----
35+
void __riscv_vse16_v_f16mf4(_Float16 *rs1, vfloat16mf4_t vs3, size_t vl);
36+
void __riscv_vse16_v_f16mf2(_Float16 *rs1, vfloat16mf2_t vs3, size_t vl);
37+
void __riscv_vse16_v_f16m1(_Float16 *rs1, vfloat16m1_t vs3, size_t vl);
38+
void __riscv_vse16_v_f16m2(_Float16 *rs1, vfloat16m2_t vs3, size_t vl);
39+
void __riscv_vse16_v_f16m4(_Float16 *rs1, vfloat16m4_t vs3, size_t vl);
40+
void __riscv_vse16_v_f16m8(_Float16 *rs1, vfloat16m8_t vs3, size_t vl);
41+
// masked functions
42+
void __riscv_vse16_v_f16mf4_m(vbool64_t vm, _Float16 *rs1, vfloat16mf4_t vs3,
43+
size_t vl);
44+
void __riscv_vse16_v_f16mf2_m(vbool32_t vm, _Float16 *rs1, vfloat16mf2_t vs3,
45+
size_t vl);
46+
void __riscv_vse16_v_f16m1_m(vbool16_t vm, _Float16 *rs1, vfloat16m1_t vs3,
47+
size_t vl);
48+
void __riscv_vse16_v_f16m2_m(vbool8_t vm, _Float16 *rs1, vfloat16m2_t vs3,
49+
size_t vl);
50+
void __riscv_vse16_v_f16m4_m(vbool4_t vm, _Float16 *rs1, vfloat16m4_t vs3,
51+
size_t vl);
52+
void __riscv_vse16_v_f16m8_m(vbool2_t vm, _Float16 *rs1, vfloat16m8_t vs3,
53+
size_t vl);
54+
----
55+
56+
[[vector-strided-load]]
57+
==== Vector Strided Load Intrinsics
58+
59+
[,c]
60+
----
61+
vfloat16mf4_t __riscv_vlse16_v_f16mf4(const _Float16 *rs1, ptrdiff_t rs2,
62+
size_t vl);
63+
vfloat16mf2_t __riscv_vlse16_v_f16mf2(const _Float16 *rs1, ptrdiff_t rs2,
64+
size_t vl);
65+
vfloat16m1_t __riscv_vlse16_v_f16m1(const _Float16 *rs1, ptrdiff_t rs2,
66+
size_t vl);
67+
vfloat16m2_t __riscv_vlse16_v_f16m2(const _Float16 *rs1, ptrdiff_t rs2,
68+
size_t vl);
69+
vfloat16m4_t __riscv_vlse16_v_f16m4(const _Float16 *rs1, ptrdiff_t rs2,
70+
size_t vl);
71+
vfloat16m8_t __riscv_vlse16_v_f16m8(const _Float16 *rs1, ptrdiff_t rs2,
72+
size_t vl);
73+
// masked functions
74+
vfloat16mf4_t __riscv_vlse16_v_f16mf4_m(vbool64_t vm, const _Float16 *rs1,
75+
ptrdiff_t rs2, size_t vl);
76+
vfloat16mf2_t __riscv_vlse16_v_f16mf2_m(vbool32_t vm, const _Float16 *rs1,
77+
ptrdiff_t rs2, size_t vl);
78+
vfloat16m1_t __riscv_vlse16_v_f16m1_m(vbool16_t vm, const _Float16 *rs1,
79+
ptrdiff_t rs2, size_t vl);
80+
vfloat16m2_t __riscv_vlse16_v_f16m2_m(vbool8_t vm, const _Float16 *rs1,
81+
ptrdiff_t rs2, size_t vl);
82+
vfloat16m4_t __riscv_vlse16_v_f16m4_m(vbool4_t vm, const _Float16 *rs1,
83+
ptrdiff_t rs2, size_t vl);
84+
vfloat16m8_t __riscv_vlse16_v_f16m8_m(vbool2_t vm, const _Float16 *rs1,
85+
ptrdiff_t rs2, size_t vl);
86+
----
87+
88+
[[vector-strided-store]]
89+
==== Vector Strided Store Intrinsics
90+
91+
[,c]
92+
----
93+
void __riscv_vsse16_v_f16mf4(_Float16 *rs1, ptrdiff_t rs2, vfloat16mf4_t vs3,
94+
size_t vl);
95+
void __riscv_vsse16_v_f16mf2(_Float16 *rs1, ptrdiff_t rs2, vfloat16mf2_t vs3,
96+
size_t vl);
97+
void __riscv_vsse16_v_f16m1(_Float16 *rs1, ptrdiff_t rs2, vfloat16m1_t vs3,
98+
size_t vl);
99+
void __riscv_vsse16_v_f16m2(_Float16 *rs1, ptrdiff_t rs2, vfloat16m2_t vs3,
100+
size_t vl);
101+
void __riscv_vsse16_v_f16m4(_Float16 *rs1, ptrdiff_t rs2, vfloat16m4_t vs3,
102+
size_t vl);
103+
void __riscv_vsse16_v_f16m8(_Float16 *rs1, ptrdiff_t rs2, vfloat16m8_t vs3,
104+
size_t vl);
105+
// masked functions
106+
void __riscv_vsse16_v_f16mf4_m(vbool64_t vm, _Float16 *rs1, ptrdiff_t rs2,
107+
vfloat16mf4_t vs3, size_t vl);
108+
void __riscv_vsse16_v_f16mf2_m(vbool32_t vm, _Float16 *rs1, ptrdiff_t rs2,
109+
vfloat16mf2_t vs3, size_t vl);
110+
void __riscv_vsse16_v_f16m1_m(vbool16_t vm, _Float16 *rs1, ptrdiff_t rs2,
111+
vfloat16m1_t vs3, size_t vl);
112+
void __riscv_vsse16_v_f16m2_m(vbool8_t vm, _Float16 *rs1, ptrdiff_t rs2,
113+
vfloat16m2_t vs3, size_t vl);
114+
void __riscv_vsse16_v_f16m4_m(vbool4_t vm, _Float16 *rs1, ptrdiff_t rs2,
115+
vfloat16m4_t vs3, size_t vl);
116+
void __riscv_vsse16_v_f16m8_m(vbool2_t vm, _Float16 *rs1, ptrdiff_t rs2,
117+
vfloat16m8_t vs3, size_t vl);
118+
----
119+
120+
[[vector-indexed-load]]
121+
==== Vector Indexed Load Intrinsics
122+
123+
[,c]
124+
----
125+
vfloat16mf4_t __riscv_vloxei16_v_f16mf4(const _Float16 *rs1, vuint16mf4_t rs2,
126+
size_t vl);
127+
vfloat16mf2_t __riscv_vloxei16_v_f16mf2(const _Float16 *rs1, vuint16mf2_t rs2,
128+
size_t vl);
129+
vfloat16m1_t __riscv_vloxei16_v_f16m1(const _Float16 *rs1, vuint16m1_t rs2,
130+
size_t vl);
131+
vfloat16m2_t __riscv_vloxei16_v_f16m2(const _Float16 *rs1, vuint16m2_t rs2,
132+
size_t vl);
133+
vfloat16m4_t __riscv_vloxei16_v_f16m4(const _Float16 *rs1, vuint16m4_t rs2,
134+
size_t vl);
135+
vfloat16m8_t __riscv_vloxei16_v_f16m8(const _Float16 *rs1, vuint16m8_t rs2,
136+
size_t vl);
137+
vfloat16mf4_t __riscv_vluxei16_v_f16mf4(const _Float16 *rs1, vuint16mf4_t rs2,
138+
size_t vl);
139+
vfloat16mf2_t __riscv_vluxei16_v_f16mf2(const _Float16 *rs1, vuint16mf2_t rs2,
140+
size_t vl);
141+
vfloat16m1_t __riscv_vluxei16_v_f16m1(const _Float16 *rs1, vuint16m1_t rs2,
142+
size_t vl);
143+
vfloat16m2_t __riscv_vluxei16_v_f16m2(const _Float16 *rs1, vuint16m2_t rs2,
144+
size_t vl);
145+
vfloat16m4_t __riscv_vluxei16_v_f16m4(const _Float16 *rs1, vuint16m4_t rs2,
146+
size_t vl);
147+
vfloat16m8_t __riscv_vluxei16_v_f16m8(const _Float16 *rs1, vuint16m8_t rs2,
148+
size_t vl);
149+
// masked functions
150+
vfloat16mf4_t __riscv_vloxei16_v_f16mf4_m(vbool64_t vm, const _Float16 *rs1,
151+
vuint16mf4_t rs2, size_t vl);
152+
vfloat16mf2_t __riscv_vloxei16_v_f16mf2_m(vbool32_t vm, const _Float16 *rs1,
153+
vuint16mf2_t rs2, size_t vl);
154+
vfloat16m1_t __riscv_vloxei16_v_f16m1_m(vbool16_t vm, const _Float16 *rs1,
155+
vuint16m1_t rs2, size_t vl);
156+
vfloat16m2_t __riscv_vloxei16_v_f16m2_m(vbool8_t vm, const _Float16 *rs1,
157+
vuint16m2_t rs2, size_t vl);
158+
vfloat16m4_t __riscv_vloxei16_v_f16m4_m(vbool4_t vm, const _Float16 *rs1,
159+
vuint16m4_t rs2, size_t vl);
160+
vfloat16m8_t __riscv_vloxei16_v_f16m8_m(vbool2_t vm, const _Float16 *rs1,
161+
vuint16m8_t rs2, size_t vl);
162+
vfloat16mf4_t __riscv_vluxei16_v_f16mf4_m(vbool64_t vm, const _Float16 *rs1,
163+
vuint16mf4_t rs2, size_t vl);
164+
vfloat16mf2_t __riscv_vluxei16_v_f16mf2_m(vbool32_t vm, const _Float16 *rs1,
165+
vuint16mf2_t rs2, size_t vl);
166+
vfloat16m1_t __riscv_vluxei16_v_f16m1_m(vbool16_t vm, const _Float16 *rs1,
167+
vuint16m1_t rs2, size_t vl);
168+
vfloat16m2_t __riscv_vluxei16_v_f16m2_m(vbool8_t vm, const _Float16 *rs1,
169+
vuint16m2_t rs2, size_t vl);
170+
vfloat16m4_t __riscv_vluxei16_v_f16m4_m(vbool4_t vm, const _Float16 *rs1,
171+
vuint16m4_t rs2, size_t vl);
172+
vfloat16m8_t __riscv_vluxei16_v_f16m8_m(vbool2_t vm, const _Float16 *rs1,
173+
vuint16m8_t rs2, size_t vl);
174+
----
175+
176+
[[vector-indexed-store]]
177+
==== Vector Indexed Store Intrinsics
178+
179+
[,c]
180+
----
181+
void __riscv_vsoxei16_v_f16mf4(_Float16 *rs1, vuint16mf4_t rs2,
182+
vfloat16mf4_t vs3, size_t vl);
183+
void __riscv_vsoxei16_v_f16mf2(_Float16 *rs1, vuint16mf2_t rs2,
184+
vfloat16mf2_t vs3, size_t vl);
185+
void __riscv_vsoxei16_v_f16m1(_Float16 *rs1, vuint16m1_t rs2, vfloat16m1_t vs3,
186+
size_t vl);
187+
void __riscv_vsoxei16_v_f16m2(_Float16 *rs1, vuint16m2_t rs2, vfloat16m2_t vs3,
188+
size_t vl);
189+
void __riscv_vsoxei16_v_f16m4(_Float16 *rs1, vuint16m4_t rs2, vfloat16m4_t vs3,
190+
size_t vl);
191+
void __riscv_vsoxei16_v_f16m8(_Float16 *rs1, vuint16m8_t rs2, vfloat16m8_t vs3,
192+
size_t vl);
193+
void __riscv_vsuxei16_v_f16mf4(_Float16 *rs1, vuint16mf4_t rs2,
194+
vfloat16mf4_t vs3, size_t vl);
195+
void __riscv_vsuxei16_v_f16mf2(_Float16 *rs1, vuint16mf2_t rs2,
196+
vfloat16mf2_t vs3, size_t vl);
197+
void __riscv_vsuxei16_v_f16m1(_Float16 *rs1, vuint16m1_t rs2, vfloat16m1_t vs3,
198+
size_t vl);
199+
void __riscv_vsuxei16_v_f16m2(_Float16 *rs1, vuint16m2_t rs2, vfloat16m2_t vs3,
200+
size_t vl);
201+
void __riscv_vsuxei16_v_f16m4(_Float16 *rs1, vuint16m4_t rs2, vfloat16m4_t vs3,
202+
size_t vl);
203+
void __riscv_vsuxei16_v_f16m8(_Float16 *rs1, vuint16m8_t rs2, vfloat16m8_t vs3,
204+
size_t vl);
205+
// masked functions
206+
void __riscv_vsoxei16_v_f16mf4_m(vbool64_t vm, _Float16 *rs1, vuint16mf4_t rs2,
207+
vfloat16mf4_t vs3, size_t vl);
208+
void __riscv_vsoxei16_v_f16mf2_m(vbool32_t vm, _Float16 *rs1, vuint16mf2_t rs2,
209+
vfloat16mf2_t vs3, size_t vl);
210+
void __riscv_vsoxei16_v_f16m1_m(vbool16_t vm, _Float16 *rs1, vuint16m1_t rs2,
211+
vfloat16m1_t vs3, size_t vl);
212+
void __riscv_vsoxei16_v_f16m2_m(vbool8_t vm, _Float16 *rs1, vuint16m2_t rs2,
213+
vfloat16m2_t vs3, size_t vl);
214+
void __riscv_vsoxei16_v_f16m4_m(vbool4_t vm, _Float16 *rs1, vuint16m4_t rs2,
215+
vfloat16m4_t vs3, size_t vl);
216+
void __riscv_vsoxei16_v_f16m8_m(vbool2_t vm, _Float16 *rs1, vuint16m8_t rs2,
217+
vfloat16m8_t vs3, size_t vl);
218+
void __riscv_vsuxei16_v_f16mf4_m(vbool64_t vm, _Float16 *rs1, vuint16mf4_t rs2,
219+
vfloat16mf4_t vs3, size_t vl);
220+
void __riscv_vsuxei16_v_f16mf2_m(vbool32_t vm, _Float16 *rs1, vuint16mf2_t rs2,
221+
vfloat16mf2_t vs3, size_t vl);
222+
void __riscv_vsuxei16_v_f16m1_m(vbool16_t vm, _Float16 *rs1, vuint16m1_t rs2,
223+
vfloat16m1_t vs3, size_t vl);
224+
void __riscv_vsuxei16_v_f16m2_m(vbool8_t vm, _Float16 *rs1, vuint16m2_t rs2,
225+
vfloat16m2_t vs3, size_t vl);
226+
void __riscv_vsuxei16_v_f16m4_m(vbool4_t vm, _Float16 *rs1, vuint16m4_t rs2,
227+
vfloat16m4_t vs3, size_t vl);
228+
void __riscv_vsuxei16_v_f16m8_m(vbool2_t vm, _Float16 *rs1, vuint16m8_t rs2,
229+
vfloat16m8_t vs3, size_t vl);
230+
----
231+
232+
[[unit-stride-fault-only-first-loads]]
233+
==== Unit-stride Fault-Only-First Loads Intrinsics
234+
235+
[,c]
236+
----
237+
vfloat16mf4_t __riscv_vle16ff_v_f16mf4(const _Float16 *rs1, size_t *new_vl,
238+
size_t vl);
239+
vfloat16mf2_t __riscv_vle16ff_v_f16mf2(const _Float16 *rs1, size_t *new_vl,
240+
size_t vl);
241+
vfloat16m1_t __riscv_vle16ff_v_f16m1(const _Float16 *rs1, size_t *new_vl,
242+
size_t vl);
243+
vfloat16m2_t __riscv_vle16ff_v_f16m2(const _Float16 *rs1, size_t *new_vl,
244+
size_t vl);
245+
vfloat16m4_t __riscv_vle16ff_v_f16m4(const _Float16 *rs1, size_t *new_vl,
246+
size_t vl);
247+
vfloat16m8_t __riscv_vle16ff_v_f16m8(const _Float16 *rs1, size_t *new_vl,
248+
size_t vl);
249+
// masked functions
250+
vfloat16mf4_t __riscv_vle16ff_v_f16mf4_m(vbool64_t vm, const _Float16 *rs1,
251+
size_t *new_vl, size_t vl);
252+
vfloat16mf2_t __riscv_vle16ff_v_f16mf2_m(vbool32_t vm, const _Float16 *rs1,
253+
size_t *new_vl, size_t vl);
254+
vfloat16m1_t __riscv_vle16ff_v_f16m1_m(vbool16_t vm, const _Float16 *rs1,
255+
size_t *new_vl, size_t vl);
256+
vfloat16m2_t __riscv_vle16ff_v_f16m2_m(vbool8_t vm, const _Float16 *rs1,
257+
size_t *new_vl, size_t vl);
258+
vfloat16m4_t __riscv_vle16ff_v_f16m4_m(vbool4_t vm, const _Float16 *rs1,
259+
size_t *new_vl, size_t vl);
260+
vfloat16m8_t __riscv_vle16ff_v_f16m8_m(vbool2_t vm, const _Float16 *rs1,
261+
size_t *new_vl, size_t vl);
262+
----

0 commit comments

Comments
 (0)