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
263
264
265
266
267
268
269
|
#pragma once
#include <TargetConditionals.h>
#define IPHONE_VEC
inline void initVFP() {
#if !TARGET_IPHONE_SIMULATOR
int tmp;
__asm__ __volatile__("fmrx %[tmp], fpscr\n\t"
"orr %[tmp], %[tmp], #(3 << 16)\n\t" /* set vector size to 4 */
"fmxr fpscr, %[tmp]\n\t"
: [ tmp ] "=&r"(tmp)
:
: "cc", "memory");
#endif
}
inline void releaseVFP() {
#if !TARGET_IPHONE_SIMULATOR
int tmp;
__asm__ __volatile__("fmrx %[tmp], fpscr\n\t"
"bic %[tmp], %[tmp], #(7 << 16)\n\t" /* set vector size back to 1 */
"fmxr fpscr, %[tmp]\n\t"
: [ tmp ] "=&r"(tmp)
:
: "cc", "memory");
#endif
}
inline void vfill(float* dest, float val, int len) {
#if !TARGET_IPHONE_SIMULATOR
float t[4] = { val, val, val, val };
float* v = (float*)t;
int tmp;
__asm__ __volatile__("fmrx %[tmp], fpscr\n\t"
"orr %[tmp], %[tmp], #(3 << 16)\n\t" /* set vector size to 4 */
"fmxr fpscr, %[tmp]\n\t"
"fldmias %[src1], {s8-s11}\n\t"
"1:\n\t"
"subs %[len], %[len], #16\n\t"
"fstmias %[dst]!, {s8-s11}\n\t"
"fstmias %[dst]!, {s8-s11}\n\t"
"fstmiasge %[dst]!, {s8-s11}\n\t"
"fstmiasge %[dst]!, {s8-s11}\n\t"
"bgt 1b\n\t"
"bic %[tmp], %[tmp], #(7 << 16)\n\t" /* set vector size back to 1 */
"fmxr fpscr, %[tmp]\n\t"
: [ dst ] "+&r"(dest), [ src1 ] "+&r"(v), [ len ] "+&r"(len), [ tmp ] "=&r"(tmp)
:
: "s0", "s1", "s2", "s3", "s4", "s5", "s6", "s7", "s8", "s9", "s10", "s11", "s12", "s13",
"s14", "s15", "s16", "s17", "s18", "s19", "s20", "s21", "s22", "s23", "s24", "s25", "s26",
"s27", "s28", "s29", "s30", "s31", "cc", "memory");
#endif
}
// a + b*c
inline void vmuladd(float* dest, float* a, float* b, float* c, int len) {
#if !TARGET_IPHONE_SIMULATOR
int tmp;
__asm__ __volatile__("fmrx %[tmp], fpscr\n\t"
"orr %[tmp], %[tmp], #(3 << 16)\n\t" /* set vector size to 4 */
"fmxr fpscr, %[tmp]\n\t"
"fldmias %[src2]!, {s0-s3}\n\t"
"fldmias %[src1]!, {s8-s11}\n\t"
"fldmias %[src3]!, {s24-s27}\n\t"
"fldmias %[src2]!, {s4-s7}\n\t"
"fldmias %[src1]!, {s12-s15}\n\t"
"fldmias %[src3]!, {s28-s31}\n\t"
"fmacs s8, s0, s24\n\t"
"1:\n\t"
"subs %[len], %[len], #16\n\t"
"fmacs s12, s4, s28\n\t"
"fldmiasge %[src2]!, {s0-s3}\n\t"
"fldmiasge %[src1]!, {s16-s19}\n\t"
"fldmiasge %[src3]!, {s24-s27}\n\t"
"fldmiasge %[src2]!, {s4-s7}\n\t"
"fldmiasge %[src1]!, {s20-s23}\n\t"
"fldmiasge %[src3]!, {s28-s31}\n\t"
"fmacsge s16, s0, s24\n\t"
"fstmias %[dst]!, {s8-s11}\n\t"
"fstmias %[dst]!, {s12-s15}\n\t"
"fmacsge s20, s4, s28\n\t"
"fldmiasgt %[src2]!, {s0-s3}\n\t"
"fldmiasgt %[src1]!, {s8-s11}\n\t"
"fldmiasgt %[src3]!, {s24-s27}\n\t"
"fldmiasgt %[src2]!, {s4-s7}\n\t"
"fldmiasgt %[src1]!, {s12-s15}\n\t"
"fldmiasgt %[src3]!, {s28-s31}\n\t"
"fmacsge s8, s0, s24\n\t"
"fstmiasge %[dst]!, {s16-s19}\n\t"
"fstmiasge %[dst]!, {s20-s23}\n\t"
"bgt 1b\n\t"
"bic %[tmp], %[tmp], #(7 << 16)\n\t" /* set vector size back to 1 */
"fmxr fpscr, %[tmp]\n\t"
: [ dst ] "+&r"(dest), [ src1 ] "+&r"(a), [ src2 ] "+&r"(b), [ src3 ] "+&r"(c),
[ len ] "+&r"(len), [ tmp ] "=&r"(tmp)
:
: "s0", "s1", "s2", "s3", "s4", "s5", "s6", "s7", "s8", "s9", "s10", "s11", "s12", "s13",
"s14", "s15", "s16", "s17", "s18", "s19", "s20", "s21", "s22", "s23", "s24", "s25", "s26",
"s27", "s28", "s29", "s30", "s31", "cc", "memory");
#endif
}
inline void vscalarmul(float* dest, float scalar, float* b, int len) {
#if !TARGET_IPHONE_SIMULATOR
float t[4] = { scalar, scalar, scalar, scalar };
float* s = t;
int tmp;
__asm__ __volatile__("fmrx %[tmp], fpscr\n\t"
"orr %[tmp], %[tmp], #(3 << 16)\n\t" /* set vector size to 4 */
"fmxr fpscr, %[tmp]\n\t"
"fldmias %[src1], {s0-s3}\n\t"
"fldmias %[src2]!, {s8-s11}\n\t"
"fldmias %[src2]!, {s12-s15}\n\t"
"fmuls s8, s8, s0\n\t"
"1:\n\t"
"subs %[len], %[len], #16\n\t"
"fmuls s12, s12, s0\n\t"
"fldmiasge %[src2]!, {s24-s27}\n\t"
"fldmiasge %[src2]!, {s28-s31}\n\t"
"fmulsge s24, s24, s0\n\t"
"fstmias %[dst]!, {s8-s11}\n\t"
"fstmias %[dst]!, {s12-s15}\n\t"
"fmulsge s28, s28, s0\n\t"
"fldmiasgt %[src2]!, {s8-s11}\n\t"
"fldmiasgt %[src2]!, {s12-s15}\n\t"
"fmulsge s8, s8, s0\n\t"
"fstmiasge %[dst]!, {s24-s27}\n\t"
"fstmiasge %[dst]!, {s28-s31}\n\t"
"bgt 1b\n\t"
"bic %[tmp], %[tmp], #(7 << 16)\n\t" /* set vector size back to 1 */
"fmxr fpscr, %[tmp]\n\t"
: [ dst ] "+&r"(dest), [ src1 ] "+&r"(s), [ src2 ] "+&r"(b), [ len ] "+&r"(len),
[ tmp ] "=&r"(tmp)
:
: "s0", "s1", "s2", "s3", "s4", "s5", "s6", "s7", "s8", "s9", "s10", "s11", "s12", "s13",
"s14", "s15", "s16", "s17", "s18", "s19", "s20", "s21", "s22", "s23", "s24", "s25", "s26",
"s27", "s28", "s29", "s30", "s31", "cc", "memory");
#endif
}
inline void vmul(float* dest, float* a, const float* b, int len) {
#if !TARGET_IPHONE_SIMULATOR
int tmp;
__asm__ __volatile__("fmrx %[tmp], fpscr\n\t"
"orr %[tmp], %[tmp], #(3 << 16)\n\t" /* set vector size to 4 */
"fmxr fpscr, %[tmp]\n\t"
"fldmias %[src1]!, {s0-s3}\n\t"
"fldmias %[src2]!, {s8-s11}\n\t"
"fldmias %[src1]!, {s4-s7}\n\t"
"fldmias %[src2]!, {s12-s15}\n\t"
"fmuls s8, s0, s8\n\t"
"1:\n\t"
"subs %[len], %[len], #16\n\t"
"fmuls s12, s4, s12\n\t"
"fldmiasge %[src1]!, {s16-s19}\n\t"
"fldmiasge %[src2]!, {s24-s27}\n\t"
"fldmiasge %[src1]!, {s20-s23}\n\t"
"fldmiasge %[src2]!, {s28-s31}\n\t"
"fmulsge s24, s16, s24\n\t"
"fstmias %[dst]!, {s8-s11}\n\t"
"fstmias %[dst]!, {s12-s15}\n\t"
"fmulsge s28, s20, s28\n\t"
"fldmiasgt %[src1]!, {s0-s3}\n\t"
"fldmiasgt %[src2]!, {s8-s11}\n\t"
"fldmiasgt %[src1]!, {s4-s7}\n\t"
"fldmiasgt %[src2]!, {s12-s15}\n\t"
"fmulsge s8, s0, s8\n\t"
"fstmiasge %[dst]!, {s24-s27}\n\t"
"fstmiasge %[dst]!, {s28-s31}\n\t"
"bgt 1b\n\t"
"bic %[tmp], %[tmp], #(7 << 16)\n\t" /* set vector size back to 1 */
"fmxr fpscr, %[tmp]\n\t"
: [ dst ] "+&r"(dest), [ src1 ] "+&r"(a), [ src2 ] "+&r"(b), [ len ] "+&r"(len),
[ tmp ] "=&r"(tmp)
:
: "s0", "s1", "s2", "s3", "s4", "s5", "s6", "s7", "s8", "s9", "s10", "s11", "s12", "s13",
"s14", "s15", "s16", "s17", "s18", "s19", "s20", "s21", "s22", "s23", "s24", "s25", "s26",
"s27", "s28", "s29", "s30", "s31", "cc", "memory");
#endif
}
inline void vadd(float* dest, float* a, const float* b, int len) {
#if !TARGET_IPHONE_SIMULATOR
int tmp;
__asm__ __volatile__("fmrx %[tmp], fpscr\n\t"
"orr %[tmp], %[tmp], #(3 << 16)\n\t" /* set vector size to 4 */
"fmxr fpscr, %[tmp]\n\t"
"fldmias %[src1]!, {s0-s3}\n\t"
"fldmias %[src2]!, {s8-s11}\n\t"
"fldmias %[src1]!, {s4-s7}\n\t"
"fldmias %[src2]!, {s12-s15}\n\t"
"fadds s8, s0, s8\n\t"
"1:\n\t"
"subs %[len], %[len], #16\n\t"
"fadds s12, s4, s12\n\t"
"fldmiasge %[src1]!, {s16-s19}\n\t"
"fldmiasge %[src2]!, {s24-s27}\n\t"
"fldmiasge %[src1]!, {s20-s23}\n\t"
"fldmiasge %[src2]!, {s28-s31}\n\t"
"faddsge s24, s16, s24\n\t"
"fstmias %[dst]!, {s8-s11}\n\t"
"fstmias %[dst]!, {s12-s15}\n\t"
"faddsge s28, s20, s28\n\t"
"fldmiasgt %[src1]!, {s0-s3}\n\t"
"fldmiasgt %[src2]!, {s8-s11}\n\t"
"fldmiasgt %[src1]!, {s4-s7}\n\t"
"fldmiasgt %[src2]!, {s12-s15}\n\t"
"faddsge s8, s0, s8\n\t"
"fstmiasge %[dst]!, {s24-s27}\n\t"
"fstmiasge %[dst]!, {s28-s31}\n\t"
"bgt 1b\n\t"
"bic %[tmp], %[tmp], #(7 << 16)\n\t" /* set vector size back to 1 */
"fmxr fpscr, %[tmp]\n\t"
: [ dst ] "+&r"(dest), [ src1 ] "+&r"(a), [ src2 ] "+&r"(b), [ len ] "+&r"(len),
[ tmp ] "=&r"(tmp)
:
: "s0", "s1", "s2", "s3", "s4", "s5", "s6", "s7", "s8", "s9", "s10", "s11", "s12", "s13",
"s14", "s15", "s16", "s17", "s18", "s19", "s20", "s21", "s22", "s23", "s24", "s25", "s26",
"s27", "s28", "s29", "s30", "s31", "cc", "memory");
#endif
}
inline void vcopy(float* dest, float* a, int len) {
#if !TARGET_IPHONE_SIMULATOR
int tmp;
__asm__ __volatile__("fmrx %[tmp], fpscr\n\t"
"orr %[tmp], %[tmp], #(3 << 16)\n\t" /* set vector size to 4 */
"fmxr fpscr, %[tmp]\n\t"
"fldmias %[src1]!, {s8-s11}\n\t"
"fldmias %[src1]!, {s12-s15}\n\t"
"1:\n\t"
"subs %[len], %[len], #16\n\t"
"fldmiasge %[src1]!, {s24-s27}\n\t"
"fldmiasge %[src1]!, {s28-s31}\n\t"
"fstmias %[dst]!, {s8-s11}\n\t"
"fstmias %[dst]!, {s12-s15}\n\t"
"fldmiasgt %[src1]!, {s8-s11}\n\t"
"fldmiasgt %[src1]!, {s12-s15}\n\t"
"fstmiasge %[dst]!, {s24-s27}\n\t"
"fstmiasge %[dst]!, {s28-s31}\n\t"
"bgt 1b\n\t"
"bic %[tmp], %[tmp], #(7 << 16)\n\t" /* set vector size back to 1 */
"fmxr fpscr, %[tmp]\n\t"
: [ dst ] "+&r"(dest), [ src1 ] "+&r"(a), [ len ] "+&r"(len), [ tmp ] "=&r"(tmp)
:
: "s0", "s1", "s2", "s3", "s4", "s5", "s6", "s7", "s8", "s9", "s10", "s11", "s12", "s13",
"s14", "s15", "s16", "s17", "s18", "s19", "s20", "s21", "s22", "s23", "s24", "s25", "s26",
"s27", "s28", "s29", "s30", "s31", "cc", "memory");
#endif
}
|
