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
|
/*
* FFT transform, optimized with VSX built-in functions
* Copyright (c) 2014 Rong Yan
*
* This algorithm (though not any of the implementation details) is
* based on libdjbfft by D. J. Bernstein.
*
* 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 "config.h"
#include "libavutil/cpu.h"
#include "libavutil/ppc/types_altivec.h"
#include "libavutil/ppc/util_altivec.h"
#include "libavcodec/fft.h"
#include "libavcodec/fft-internal.h"
#include "fft_vsx.h"
#if HAVE_VSX
static void fft32_vsx_interleave(FFTComplex *z)
{
fft16_vsx_interleave(z);
fft8_vsx_interleave(z+16);
fft8_vsx_interleave(z+24);
pass_vsx_interleave(z,ff_cos_32,4);
}
static void fft64_vsx_interleave(FFTComplex *z)
{
fft32_vsx_interleave(z);
fft16_vsx_interleave(z+32);
fft16_vsx_interleave(z+48);
pass_vsx_interleave(z,ff_cos_64, 8);
}
static void fft128_vsx_interleave(FFTComplex *z)
{
fft64_vsx_interleave(z);
fft32_vsx_interleave(z+64);
fft32_vsx_interleave(z+96);
pass_vsx_interleave(z,ff_cos_128,16);
}
static void fft256_vsx_interleave(FFTComplex *z)
{
fft128_vsx_interleave(z);
fft64_vsx_interleave(z+128);
fft64_vsx_interleave(z+192);
pass_vsx_interleave(z,ff_cos_256,32);
}
static void fft512_vsx_interleave(FFTComplex *z)
{
fft256_vsx_interleave(z);
fft128_vsx_interleave(z+256);
fft128_vsx_interleave(z+384);
pass_vsx_interleave(z,ff_cos_512,64);
}
static void fft1024_vsx_interleave(FFTComplex *z)
{
fft512_vsx_interleave(z);
fft256_vsx_interleave(z+512);
fft256_vsx_interleave(z+768);
pass_vsx_interleave(z,ff_cos_1024,128);
}
static void fft2048_vsx_interleave(FFTComplex *z)
{
fft1024_vsx_interleave(z);
fft512_vsx_interleave(z+1024);
fft512_vsx_interleave(z+1536);
pass_vsx_interleave(z,ff_cos_2048,256);
}
static void fft4096_vsx_interleave(FFTComplex *z)
{
fft2048_vsx_interleave(z);
fft1024_vsx_interleave(z+2048);
fft1024_vsx_interleave(z+3072);
pass_vsx_interleave(z,ff_cos_4096, 512);
}
static void fft8192_vsx_interleave(FFTComplex *z)
{
fft4096_vsx_interleave(z);
fft2048_vsx_interleave(z+4096);
fft2048_vsx_interleave(z+6144);
pass_vsx_interleave(z,ff_cos_8192,1024);
}
static void fft16384_vsx_interleave(FFTComplex *z)
{
fft8192_vsx_interleave(z);
fft4096_vsx_interleave(z+8192);
fft4096_vsx_interleave(z+12288);
pass_vsx_interleave(z,ff_cos_16384,2048);
}
static void fft32768_vsx_interleave(FFTComplex *z)
{
fft16384_vsx_interleave(z);
fft8192_vsx_interleave(z+16384);
fft8192_vsx_interleave(z+24576);
pass_vsx_interleave(z,ff_cos_32768,4096);
}
static void fft65536_vsx_interleave(FFTComplex *z)
{
fft32768_vsx_interleave(z);
fft16384_vsx_interleave(z+32768);
fft16384_vsx_interleave(z+49152);
pass_vsx_interleave(z,ff_cos_65536,8192);
}
static void fft32_vsx(FFTComplex *z)
{
fft16_vsx(z);
fft8_vsx(z+16);
fft8_vsx(z+24);
pass_vsx(z,ff_cos_32,4);
}
static void fft64_vsx(FFTComplex *z)
{
fft32_vsx(z);
fft16_vsx(z+32);
fft16_vsx(z+48);
pass_vsx(z,ff_cos_64, 8);
}
static void fft128_vsx(FFTComplex *z)
{
fft64_vsx(z);
fft32_vsx(z+64);
fft32_vsx(z+96);
pass_vsx(z,ff_cos_128,16);
}
static void fft256_vsx(FFTComplex *z)
{
fft128_vsx(z);
fft64_vsx(z+128);
fft64_vsx(z+192);
pass_vsx(z,ff_cos_256,32);
}
static void fft512_vsx(FFTComplex *z)
{
fft256_vsx(z);
fft128_vsx(z+256);
fft128_vsx(z+384);
pass_vsx(z,ff_cos_512,64);
}
static void fft1024_vsx(FFTComplex *z)
{
fft512_vsx(z);
fft256_vsx(z+512);
fft256_vsx(z+768);
pass_vsx(z,ff_cos_1024,128);
}
static void fft2048_vsx(FFTComplex *z)
{
fft1024_vsx(z);
fft512_vsx(z+1024);
fft512_vsx(z+1536);
pass_vsx(z,ff_cos_2048,256);
}
static void fft4096_vsx(FFTComplex *z)
{
fft2048_vsx(z);
fft1024_vsx(z+2048);
fft1024_vsx(z+3072);
pass_vsx(z,ff_cos_4096, 512);
}
static void fft8192_vsx(FFTComplex *z)
{
fft4096_vsx(z);
fft2048_vsx(z+4096);
fft2048_vsx(z+6144);
pass_vsx(z,ff_cos_8192,1024);
}
static void fft16384_vsx(FFTComplex *z)
{
fft8192_vsx(z);
fft4096_vsx(z+8192);
fft4096_vsx(z+12288);
pass_vsx(z,ff_cos_16384,2048);
}
static void fft32768_vsx(FFTComplex *z)
{
fft16384_vsx(z);
fft8192_vsx(z+16384);
fft8192_vsx(z+24576);
pass_vsx(z,ff_cos_32768,4096);
}
static void fft65536_vsx(FFTComplex *z)
{
fft32768_vsx(z);
fft16384_vsx(z+32768);
fft16384_vsx(z+49152);
pass_vsx(z,ff_cos_65536,8192);
}
static void (* const fft_dispatch_vsx[])(FFTComplex*) = {
fft4_vsx, fft8_vsx, fft16_vsx, fft32_vsx, fft64_vsx, fft128_vsx, fft256_vsx, fft512_vsx, fft1024_vsx,
fft2048_vsx, fft4096_vsx, fft8192_vsx, fft16384_vsx, fft32768_vsx, fft65536_vsx,
};
static void (* const fft_dispatch_vsx_interleave[])(FFTComplex*) = {
fft4_vsx_interleave, fft8_vsx_interleave, fft16_vsx_interleave, fft32_vsx_interleave, fft64_vsx_interleave,
fft128_vsx_interleave, fft256_vsx_interleave, fft512_vsx_interleave, fft1024_vsx_interleave,
fft2048_vsx_interleave, fft4096_vsx_interleave, fft8192_vsx_interleave, fft16384_vsx_interleave, fft32768_vsx_interleave, fft65536_vsx_interleave,
};
void ff_fft_calc_interleave_vsx(FFTContext *s, FFTComplex *z)
{
fft_dispatch_vsx_interleave[s->nbits-2](z);
}
void ff_fft_calc_vsx(FFTContext *s, FFTComplex *z)
{
fft_dispatch_vsx[s->nbits-2](z);
}
#endif /* HAVE_VSX */
|
