You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

130 lines
3.7KB

  1. /*
  2. * Altivec optimized MP3 decoding functions
  3. * Copyright (c) 2010 Vitor Sessak
  4. *
  5. * This file is part of FFmpeg.
  6. *
  7. * FFmpeg 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. * FFmpeg 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 FFmpeg; if not, write to the Free Software
  19. * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  20. */
  21. #include "dsputil_altivec.h"
  22. #include "util_altivec.h"
  23. #include "libavcodec/dsputil.h"
  24. #include "libavcodec/mpegaudiodsp.h"
  25. #define MACS(rt, ra, rb) rt+=(ra)*(rb)
  26. #define MLSS(rt, ra, rb) rt-=(ra)*(rb)
  27. #define SUM8(op, sum, w, p) \
  28. { \
  29. op(sum, (w)[0 * 64], (p)[0 * 64]); \
  30. op(sum, (w)[1 * 64], (p)[1 * 64]); \
  31. op(sum, (w)[2 * 64], (p)[2 * 64]); \
  32. op(sum, (w)[3 * 64], (p)[3 * 64]); \
  33. op(sum, (w)[4 * 64], (p)[4 * 64]); \
  34. op(sum, (w)[5 * 64], (p)[5 * 64]); \
  35. op(sum, (w)[6 * 64], (p)[6 * 64]); \
  36. op(sum, (w)[7 * 64], (p)[7 * 64]); \
  37. }
  38. static void apply_window(const float *buf, const float *win1,
  39. const float *win2, float *sum1, float *sum2, int len)
  40. {
  41. const vector float *win1a = (const vector float *) win1;
  42. const vector float *win2a = (const vector float *) win2;
  43. const vector float *bufa = (const vector float *) buf;
  44. vector float *sum1a = (vector float *) sum1;
  45. vector float *sum2a = (vector float *) sum2;
  46. vector float av_uninit(v0), av_uninit(v4);
  47. vector float v1, v2, v3;
  48. len = len >> 2;
  49. #define MULT(a, b) \
  50. { \
  51. v1 = vec_ld(a, win1a); \
  52. v2 = vec_ld(b, win2a); \
  53. v3 = vec_ld(a, bufa); \
  54. v0 = vec_madd(v3, v1, v0); \
  55. v4 = vec_madd(v2, v3, v4); \
  56. }
  57. while (len--) {
  58. v0 = vec_xor(v0, v0);
  59. v4 = vec_xor(v4, v4);
  60. MULT( 0, 0);
  61. MULT( 256, 64);
  62. MULT( 512, 128);
  63. MULT( 768, 192);
  64. MULT(1024, 256);
  65. MULT(1280, 320);
  66. MULT(1536, 384);
  67. MULT(1792, 448);
  68. vec_st(v0, 0, sum1a);
  69. vec_st(v4, 0, sum2a);
  70. sum1a++;
  71. sum2a++;
  72. win1a++;
  73. win2a++;
  74. bufa++;
  75. }
  76. }
  77. static void apply_window_mp3(float *in, float *win, int *unused, float *out,
  78. int incr)
  79. {
  80. LOCAL_ALIGNED_16(float, suma, [17]);
  81. LOCAL_ALIGNED_16(float, sumb, [17]);
  82. LOCAL_ALIGNED_16(float, sumc, [17]);
  83. LOCAL_ALIGNED_16(float, sumd, [17]);
  84. float sum;
  85. int j;
  86. float *out2 = out + 32 * incr;
  87. /* copy to avoid wrap */
  88. memcpy(in + 512, in, 32 * sizeof(*in));
  89. apply_window(in + 16, win , win + 512, suma, sumc, 16);
  90. apply_window(in + 32, win + 48, win + 640, sumb, sumd, 16);
  91. SUM8(MLSS, suma[0], win + 32, in + 48);
  92. sumc[ 0] = 0;
  93. sumb[16] = 0;
  94. sumd[16] = 0;
  95. out[0 ] = suma[ 0];
  96. out += incr;
  97. out2 -= incr;
  98. for(j=1;j<16;j++) {
  99. *out = suma[ j] - sumd[16-j];
  100. *out2 = -sumb[16-j] - sumc[ j];
  101. out += incr;
  102. out2 -= incr;
  103. }
  104. sum = 0;
  105. SUM8(MLSS, sum, win + 16 + 32, in + 32);
  106. *out = sum;
  107. }
  108. void ff_mpadsp_init_altivec(MPADSPContext *s)
  109. {
  110. s->apply_window_float = apply_window_mp3;
  111. }