Move some mpegaudio functions to new mpegaudiodsp subsystem
[libav.git] / libavcodec / ppc / mpegaudiodec_altivec.c
1 /*
2 * Altivec optimized MP3 decoding functions
3 * Copyright (c) 2010 Vitor Sessak
4 *
5 * This file is part of Libav.
6 *
7 * Libav is free software; you can redistribute it and/or
8 * modify it under the terms of the GNU Lesser General Public
9 * License as published by the Free Software Foundation; either
10 * version 2.1 of the License, or (at your option) any later version.
11 *
12 * Libav is distributed in the hope that it will be useful,
13 * but WITHOUT ANY WARRANTY; without even the implied warranty of
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
15 * Lesser General Public License for more details.
16 *
17 * You should have received a copy of the GNU Lesser General Public
18 * License along with Libav; if not, write to the Free Software
19 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
20 */
21
22 #include "dsputil_altivec.h"
23 #include "util_altivec.h"
24 #include "libavcodec/dsputil.h"
25 #include "libavcodec/mpegaudiodsp.h"
26
27 #define MACS(rt, ra, rb) rt+=(ra)*(rb)
28 #define MLSS(rt, ra, rb) rt-=(ra)*(rb)
29
30 #define SUM8(op, sum, w, p) \
31 { \
32 op(sum, (w)[0 * 64], (p)[0 * 64]); \
33 op(sum, (w)[1 * 64], (p)[1 * 64]); \
34 op(sum, (w)[2 * 64], (p)[2 * 64]); \
35 op(sum, (w)[3 * 64], (p)[3 * 64]); \
36 op(sum, (w)[4 * 64], (p)[4 * 64]); \
37 op(sum, (w)[5 * 64], (p)[5 * 64]); \
38 op(sum, (w)[6 * 64], (p)[6 * 64]); \
39 op(sum, (w)[7 * 64], (p)[7 * 64]); \
40 }
41
42 static void apply_window(const float *buf, const float *win1,
43 const float *win2, float *sum1, float *sum2, int len)
44 {
45 const vector float *win1a = (const vector float *) win1;
46 const vector float *win2a = (const vector float *) win2;
47 const vector float *bufa = (const vector float *) buf;
48 vector float *sum1a = (vector float *) sum1;
49 vector float *sum2a = (vector float *) sum2;
50 vector float av_uninit(v0), av_uninit(v4);
51 vector float v1, v2, v3;
52
53 len = len >> 2;
54
55 #define MULT(a, b) \
56 { \
57 v1 = vec_ld(a, win1a); \
58 v2 = vec_ld(b, win2a); \
59 v3 = vec_ld(a, bufa); \
60 v0 = vec_madd(v3, v1, v0); \
61 v4 = vec_madd(v2, v3, v4); \
62 }
63
64 while (len--) {
65 v0 = vec_xor(v0, v0);
66 v4 = vec_xor(v4, v4);
67
68 MULT( 0, 0);
69 MULT( 256, 64);
70 MULT( 512, 128);
71 MULT( 768, 192);
72 MULT(1024, 256);
73 MULT(1280, 320);
74 MULT(1536, 384);
75 MULT(1792, 448);
76
77 vec_st(v0, 0, sum1a);
78 vec_st(v4, 0, sum2a);
79 sum1a++;
80 sum2a++;
81 win1a++;
82 win2a++;
83 bufa++;
84 }
85 }
86
87 static void apply_window_mp3(float *in, float *win, int *unused, float *out,
88 int incr)
89 {
90 LOCAL_ALIGNED_16(float, suma, [17]);
91 LOCAL_ALIGNED_16(float, sumb, [17]);
92 LOCAL_ALIGNED_16(float, sumc, [17]);
93 LOCAL_ALIGNED_16(float, sumd, [17]);
94
95 float sum;
96 int j;
97 float *out2 = out + 32 * incr;
98
99 /* copy to avoid wrap */
100 memcpy(in + 512, in, 32 * sizeof(*in));
101
102 apply_window(in + 16, win , win + 512, suma, sumc, 16);
103 apply_window(in + 32, win + 48, win + 640, sumb, sumd, 16);
104
105 SUM8(MLSS, suma[0], win + 32, in + 48);
106
107 sumc[ 0] = 0;
108 sumb[16] = 0;
109 sumd[16] = 0;
110
111 out[0 ] = suma[ 0];
112 out += incr;
113 out2 -= incr;
114 for(j=1;j<16;j++) {
115 *out = suma[ j] - sumd[16-j];
116 *out2 = -sumb[16-j] - sumc[ j];
117 out += incr;
118 out2 -= incr;
119 }
120
121 sum = 0;
122 SUM8(MLSS, sum, win + 16 + 32, in + 32);
123 *out = sum;
124 }
125
126 void ff_mpadsp_init_altivec(MPADSPContext *s)
127 {
128 s->apply_window_float = apply_window_mp3;
129 }