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
|
/*
* Copyright (c) 2008 Siarhei Siamashka <ssvb@users.sourceforge.net>
*
* This file is part of FFmpeg.
*
* FFmpeg is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* FFmpeg is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
#include "libavcodec/dsputil.h"
/*
* VFP is a floating point coprocessor used in some ARM cores. VFP11 has 1 cycle
* throughput for almost all the instructions (except for double precision
* arithmetics), but rather high latency. Latency is 4 cycles for loads and 8 cycles
* for arithmetic operations. Scheduling code to avoid pipeline stalls is very
* important for performance. One more interesting feature is that VFP has
* independent load/store and arithmetics pipelines, so it is possible to make
* them work simultaneously and get more than 1 operation per cycle. Load/store
* pipeline can process 2 single precision floating point values per cycle and
* supports bulk loads and stores for large sets of registers. Arithmetic operations
* can be done on vectors, which allows to keep the arithmetics pipeline busy,
* while the processor may issue and execute other instructions. Detailed
* optimization manuals can be found at http://www.arm.com
*/
/**
* ARM VFP optimized implementation of 'vector_fmul_c' function.
* Assume that len is a positive number and is multiple of 8
*/
static void vector_fmul_vfp(float *dst, const float *src, int len)
{
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 %[dst_r]!, {s0-s3}\n\t"
"fldmias %[src]!, {s8-s11}\n\t"
"fldmias %[dst_r]!, {s4-s7}\n\t"
"fldmias %[src]!, {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 %[dst_r]!, {s16-s19}\n\t"
"fldmiasge %[src]!, {s24-s27}\n\t"
"fldmiasge %[dst_r]!, {s20-s23}\n\t"
"fldmiasge %[src]!, {s28-s31}\n\t"
"fmulsge s24, s16, s24\n\t"
"fstmias %[dst_w]!, {s8-s11}\n\t"
"fstmias %[dst_w]!, {s12-s15}\n\t"
"fmulsge s28, s20, s28\n\t"
"fldmiasgt %[dst_r]!, {s0-s3}\n\t"
"fldmiasgt %[src]!, {s8-s11}\n\t"
"fldmiasgt %[dst_r]!, {s4-s7}\n\t"
"fldmiasgt %[src]!, {s12-s15}\n\t"
"fmulsge s8, s0, s8\n\t"
"fstmiasge %[dst_w]!, {s24-s27}\n\t"
"fstmiasge %[dst_w]!, {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_w] "+&r" (dst), [dst_r] "+&r" (dst), [src] "+&r" (src), [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");
}
/**
* ARM VFP optimized implementation of 'vector_fmul_reverse_c' function.
* Assume that len is a positive number and is multiple of 8
*/
static void vector_fmul_reverse_vfp(float *dst, const float *src0, const float *src1, int len)
{
src1 += len;
asm volatile(
"fldmdbs %[src1]!, {s0-s3}\n\t"
"fldmias %[src0]!, {s8-s11}\n\t"
"fldmdbs %[src1]!, {s4-s7}\n\t"
"fldmias %[src0]!, {s12-s15}\n\t"
"fmuls s8, s3, s8\n\t"
"fmuls s9, s2, s9\n\t"
"fmuls s10, s1, s10\n\t"
"fmuls s11, s0, s11\n\t"
"1:\n\t"
"subs %[len], %[len], #16\n\t"
"fldmdbsge %[src1]!, {s16-s19}\n\t"
"fmuls s12, s7, s12\n\t"
"fldmiasge %[src0]!, {s24-s27}\n\t"
"fmuls s13, s6, s13\n\t"
"fldmdbsge %[src1]!, {s20-s23}\n\t"
"fmuls s14, s5, s14\n\t"
"fldmiasge %[src0]!, {s28-s31}\n\t"
"fmuls s15, s4, s15\n\t"
"fmulsge s24, s19, s24\n\t"
"fldmdbsgt %[src1]!, {s0-s3}\n\t"
"fmulsge s25, s18, s25\n\t"
"fstmias %[dst]!, {s8-s13}\n\t"
"fmulsge s26, s17, s26\n\t"
"fldmiasgt %[src0]!, {s8-s11}\n\t"
"fmulsge s27, s16, s27\n\t"
"fmulsge s28, s23, s28\n\t"
"fldmdbsgt %[src1]!, {s4-s7}\n\t"
"fmulsge s29, s22, s29\n\t"
"fstmias %[dst]!, {s14-s15}\n\t"
"fmulsge s30, s21, s30\n\t"
"fmulsge s31, s20, s31\n\t"
"fmulsge s8, s3, s8\n\t"
"fldmiasgt %[src0]!, {s12-s15}\n\t"
"fmulsge s9, s2, s9\n\t"
"fmulsge s10, s1, s10\n\t"
"fstmiasge %[dst]!, {s24-s27}\n\t"
"fmulsge s11, s0, s11\n\t"
"fstmiasge %[dst]!, {s28-s31}\n\t"
"bgt 1b\n\t"
: [dst] "+&r" (dst), [src0] "+&r" (src0), [src1] "+&r" (src1), [len] "+&r" (len)
:
: "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");
}
#ifdef HAVE_ARMV6
/**
* ARM VFP optimized float to int16 conversion.
* Assume that len is a positive number and is multiple of 8, destination
* buffer is at least 4 bytes aligned (8 bytes alignment is better for
* performance), little endian byte sex
*/
void float_to_int16_vfp(int16_t *dst, const float *src, int len)
{
asm volatile(
"fldmias %[src]!, {s16-s23}\n\t"
"ftosis s0, s16\n\t"
"ftosis s1, s17\n\t"
"ftosis s2, s18\n\t"
"ftosis s3, s19\n\t"
"ftosis s4, s20\n\t"
"ftosis s5, s21\n\t"
"ftosis s6, s22\n\t"
"ftosis s7, s23\n\t"
"1:\n\t"
"subs %[len], %[len], #8\n\t"
"fmrrs r3, r4, {s0, s1}\n\t"
"fmrrs r5, r6, {s2, s3}\n\t"
"fmrrs r7, r8, {s4, s5}\n\t"
"fmrrs ip, lr, {s6, s7}\n\t"
"fldmiasgt %[src]!, {s16-s23}\n\t"
"ssat r4, #16, r4\n\t"
"ssat r3, #16, r3\n\t"
"ssat r6, #16, r6\n\t"
"ssat r5, #16, r5\n\t"
"pkhbt r3, r3, r4, lsl #16\n\t"
"pkhbt r4, r5, r6, lsl #16\n\t"
"ftosisgt s0, s16\n\t"
"ftosisgt s1, s17\n\t"
"ftosisgt s2, s18\n\t"
"ftosisgt s3, s19\n\t"
"ftosisgt s4, s20\n\t"
"ftosisgt s5, s21\n\t"
"ftosisgt s6, s22\n\t"
"ftosisgt s7, s23\n\t"
"ssat r8, #16, r8\n\t"
"ssat r7, #16, r7\n\t"
"ssat lr, #16, lr\n\t"
"ssat ip, #16, ip\n\t"
"pkhbt r5, r7, r8, lsl #16\n\t"
"pkhbt r6, ip, lr, lsl #16\n\t"
"stmia %[dst]!, {r3-r6}\n\t"
"bgt 1b\n\t"
: [dst] "+&r" (dst), [src] "+&r" (src), [len] "+&r" (len)
:
: "s0", "s1", "s2", "s3", "s4", "s5", "s6", "s7",
"s16", "s17", "s18", "s19", "s20", "s21", "s22", "s23",
"r3", "r4", "r5", "r6", "r7", "r8", "ip", "lr",
"cc", "memory");
}
#endif
void ff_float_init_arm_vfp(DSPContext* c, AVCodecContext *avctx)
{
c->vector_fmul = vector_fmul_vfp;
c->vector_fmul_reverse = vector_fmul_reverse_vfp;
#ifdef HAVE_ARMV6
c->float_to_int16 = float_to_int16_vfp;
#endif
}
|
