falkTX
/
FFmpeg
mirror of https://github.com/falkTX/FFmpeg.git
1
0
Fork 0
Code Issues 0 Releases 338 Wiki Activity
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.
94858 Commits
32 Branches
293MB
Tree: d620b1f9de
Branches Tags
${ item.name }
Create branch ${ searchTerm }
from 'd620b1f9de'
${ noResults }
FFmpeg/libavcodec/fft_template.c

640 lines
18KB
Raw Blame History 

  1. /*

  2.  * FFT/IFFT transforms

  3.  * Copyright (c) 2008 Loren Merritt

  4.  * Copyright (c) 2002 Fabrice Bellard

  5.  * Partly based on libdjbfft by D. J. Bernstein

  6.  *

  7.  * This file is part of FFmpeg.

  8.  *

  9.  * FFmpeg is free software; you can redistribute it and/or

  10.  * modify it under the terms of the GNU Lesser General Public

  11.  * License as published by the Free Software Foundation; either

  12.  * version 2.1 of the License, or (at your option) any later version.

  13.  *

  14.  * FFmpeg is distributed in the hope that it will be useful,

  15.  * but WITHOUT ANY WARRANTY; without even the implied warranty of

  16.  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU

  17.  * Lesser General Public License for more details.

  18.  *

  19.  * You should have received a copy of the GNU Lesser General Public

  20.  * License along with FFmpeg; if not, write to the Free Software

  21.  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA

  22.  */

  23. 

  24. /**

  25.  * @file

  26.  * FFT/IFFT transforms.

  27.  */

  28. 

  29. #include <stdlib.h>

  30. #include <string.h>

  31. #include "libavutil/mathematics.h"

  32. #include "libavutil/thread.h"

  33. #include "fft.h"

  34. #include "fft-internal.h"

  35. 

  36. #if FFT_FIXED_32

  37. #include "fft_table.h"

  38. 

  39. static void av_cold fft_lut_init(void)

  40. {

  41.     int n = 0;

  42.     ff_fft_lut_init(ff_fft_offsets_lut, 0, 1 << 17, &n);

  43. }

  44. 

  45. #else /* FFT_FIXED_32 */

  46. 

  47. /* cos(2*pi*x/n) for 0<=x<=n/4, followed by its reverse */

  48. #if !CONFIG_HARDCODED_TABLES

  49. COSTABLE(16);

  50. COSTABLE(32);

  51. COSTABLE(64);

  52. COSTABLE(128);

  53. COSTABLE(256);

  54. COSTABLE(512);

  55. COSTABLE(1024);

  56. COSTABLE(2048);

  57. COSTABLE(4096);

  58. COSTABLE(8192);

  59. COSTABLE(16384);

  60. COSTABLE(32768);

  61. COSTABLE(65536);

  62. COSTABLE(131072);

  63. 

  64. static av_cold void init_ff_cos_tabs(int index)

  65. {

  66.     int i;

  67.     int m = 1<<index;

  68.     double freq = 2*M_PI/m;

  69.     FFTSample *tab = FFT_NAME(ff_cos_tabs)[index];

  70.     for(i=0; i<=m/4; i++)

  71.         tab[i] = FIX15(cos(i*freq));

  72.     for(i=1; i<m/4; i++)

  73.         tab[m/2-i] = tab[i];

  74. }

  75. 

  76. typedef struct CosTabsInitOnce {

  77.     void (*func)(void);

  78.     AVOnce control;

  79. } CosTabsInitOnce;

  80. 

  81. #define INIT_FF_COS_TABS_FUNC(index, size)          \

  82. static av_cold void init_ff_cos_tabs_ ## size (void)\

  83. {                                                   \

  84.     init_ff_cos_tabs(index);                        \

  85. }

  86. 

  87. INIT_FF_COS_TABS_FUNC(4, 16)

  88. INIT_FF_COS_TABS_FUNC(5, 32)

  89. INIT_FF_COS_TABS_FUNC(6, 64)

  90. INIT_FF_COS_TABS_FUNC(7, 128)

  91. INIT_FF_COS_TABS_FUNC(8, 256)

  92. INIT_FF_COS_TABS_FUNC(9, 512)

  93. INIT_FF_COS_TABS_FUNC(10, 1024)

  94. INIT_FF_COS_TABS_FUNC(11, 2048)

  95. INIT_FF_COS_TABS_FUNC(12, 4096)

  96. INIT_FF_COS_TABS_FUNC(13, 8192)

  97. INIT_FF_COS_TABS_FUNC(14, 16384)

  98. INIT_FF_COS_TABS_FUNC(15, 32768)

  99. INIT_FF_COS_TABS_FUNC(16, 65536)

  100. INIT_FF_COS_TABS_FUNC(17, 131072)

  101. 

  102. static CosTabsInitOnce cos_tabs_init_once[] = {

  103.     { NULL },

  104.     { NULL },

  105.     { NULL },

  106.     { NULL },

  107.     { init_ff_cos_tabs_16, AV_ONCE_INIT },

  108.     { init_ff_cos_tabs_32, AV_ONCE_INIT },

  109.     { init_ff_cos_tabs_64, AV_ONCE_INIT },

  110.     { init_ff_cos_tabs_128, AV_ONCE_INIT },

  111.     { init_ff_cos_tabs_256, AV_ONCE_INIT },

  112.     { init_ff_cos_tabs_512, AV_ONCE_INIT },

  113.     { init_ff_cos_tabs_1024, AV_ONCE_INIT },

  114.     { init_ff_cos_tabs_2048, AV_ONCE_INIT },

  115.     { init_ff_cos_tabs_4096, AV_ONCE_INIT },

  116.     { init_ff_cos_tabs_8192, AV_ONCE_INIT },

  117.     { init_ff_cos_tabs_16384, AV_ONCE_INIT },

  118.     { init_ff_cos_tabs_32768, AV_ONCE_INIT },

  119.     { init_ff_cos_tabs_65536, AV_ONCE_INIT },

  120.     { init_ff_cos_tabs_131072, AV_ONCE_INIT },

  121. };

  122. 

  123. #endif

  124. COSTABLE_CONST FFTSample * const FFT_NAME(ff_cos_tabs)[] = {

  125.     NULL, NULL, NULL, NULL,

  126.     FFT_NAME(ff_cos_16),

  127.     FFT_NAME(ff_cos_32),

  128.     FFT_NAME(ff_cos_64),

  129.     FFT_NAME(ff_cos_128),

  130.     FFT_NAME(ff_cos_256),

  131.     FFT_NAME(ff_cos_512),

  132.     FFT_NAME(ff_cos_1024),

  133.     FFT_NAME(ff_cos_2048),

  134.     FFT_NAME(ff_cos_4096),

  135.     FFT_NAME(ff_cos_8192),

  136.     FFT_NAME(ff_cos_16384),

  137.     FFT_NAME(ff_cos_32768),

  138.     FFT_NAME(ff_cos_65536),

  139.     FFT_NAME(ff_cos_131072),

  140. };

  141. 

  142. #endif /* FFT_FIXED_32 */

  143. 

  144. static void fft_permute_c(FFTContext *s, FFTComplex *z);

  145. static void fft_calc_c(FFTContext *s, FFTComplex *z);

  146. 

  147. static int split_radix_permutation(int i, int n, int inverse)

  148. {

  149.     int m;

  150.     if(n <= 2) return i&1;

  151.     m = n >> 1;

  152.     if(!(i&m))            return split_radix_permutation(i, m, inverse)*2;

  153.     m >>= 1;

  154.     if(inverse == !(i&m)) return split_radix_permutation(i, m, inverse)*4 + 1;

  155.     else                  return split_radix_permutation(i, m, inverse)*4 - 1;

  156. }

  157. 

  158. av_cold void ff_init_ff_cos_tabs(int index)

  159. {

  160. #if (!CONFIG_HARDCODED_TABLES) && (!FFT_FIXED_32)

  161.     ff_thread_once(&cos_tabs_init_once[index].control, cos_tabs_init_once[index].func);

  162. #endif

  163. }

  164. 

  165. static const int avx_tab[] = {

  166.     0, 4, 1, 5, 8, 12, 9, 13, 2, 6, 3, 7, 10, 14, 11, 15

  167. };

  168. 

  169. static int is_second_half_of_fft32(int i, int n)

  170. {

  171.     if (n <= 32)

  172.         return i >= 16;

  173.     else if (i < n/2)

  174.         return is_second_half_of_fft32(i, n/2);

  175.     else if (i < 3*n/4)

  176.         return is_second_half_of_fft32(i - n/2, n/4);

  177.     else

  178.         return is_second_half_of_fft32(i - 3*n/4, n/4);

  179. }

  180. 

  181. static av_cold void fft_perm_avx(FFTContext *s)

  182. {

  183.     int i;

  184.     int n = 1 << s->nbits;

  185. 

  186.     for (i = 0; i < n; i += 16) {

  187.         int k;

  188.         if (is_second_half_of_fft32(i, n)) {

  189.             for (k = 0; k < 16; k++)

  190.                 s->revtab[-split_radix_permutation(i + k, n, s->inverse) & (n - 1)] =

  191.                     i + avx_tab[k];

  192. 

  193.         } else {

  194.             for (k = 0; k < 16; k++) {

  195.                 int j = i + k;

  196.                 j = (j & ~7) | ((j >> 1) & 3) | ((j << 2) & 4);

  197.                 s->revtab[-split_radix_permutation(i + k, n, s->inverse) & (n - 1)] = j;

  198.             }

  199.         }

  200.     }

  201. }

  202. 

  203. av_cold int ff_fft_init(FFTContext *s, int nbits, int inverse)

  204. {

  205.     int i, j, n;

  206. 

  207.     s->revtab = NULL;

  208.     s->revtab32 = NULL;

  209. 

  210.     if (nbits < 2 || nbits > 17)

  211.         goto fail;

  212.     s->nbits = nbits;

  213.     n = 1 << nbits;

  214. 

  215.     if (nbits <= 16) {

  216.         s->revtab = av_malloc(n * sizeof(uint16_t));

  217.         if (!s->revtab)

  218.             goto fail;

  219.     } else {

  220.         s->revtab32 = av_malloc(n * sizeof(uint32_t));

  221.         if (!s->revtab32)

  222.             goto fail;

  223.     }

  224.     s->tmp_buf = av_malloc(n * sizeof(FFTComplex));

  225.     if (!s->tmp_buf)

  226.         goto fail;

  227.     s->inverse = inverse;

  228.     s->fft_permutation = FF_FFT_PERM_DEFAULT;

  229. 

  230.     s->fft_permute = fft_permute_c;

  231.     s->fft_calc    = fft_calc_c;

  232. #if CONFIG_MDCT

  233.     s->imdct_calc  = ff_imdct_calc_c;

  234.     s->imdct_half  = ff_imdct_half_c;

  235.     s->mdct_calc   = ff_mdct_calc_c;

  236. #endif

  237. 

  238. #if FFT_FIXED_32

  239.     {

  240.         static AVOnce control = AV_ONCE_INIT;

  241.         ff_thread_once(&control, fft_lut_init);

  242.     }

  243. #else /* FFT_FIXED_32 */

  244. #if FFT_FLOAT

  245.     if (ARCH_AARCH64) ff_fft_init_aarch64(s);

  246.     if (ARCH_ARM)     ff_fft_init_arm(s);

  247.     if (ARCH_PPC)     ff_fft_init_ppc(s);

  248.     if (ARCH_X86)     ff_fft_init_x86(s);

  249.     if (CONFIG_MDCT)  s->mdct_calcw = s->mdct_calc;

  250.     if (HAVE_MIPSFPU) ff_fft_init_mips(s);

  251. #else

  252.     if (CONFIG_MDCT)  s->mdct_calcw = ff_mdct_calcw_c;

  253.     if (ARCH_ARM)     ff_fft_fixed_init_arm(s);

  254. #endif

  255.     for(j=4; j<=nbits; j++) {

  256.         ff_init_ff_cos_tabs(j);

  257.     }

  258. #endif /* FFT_FIXED_32 */

  259. 

  260. 

  261.     if (s->fft_permutation == FF_FFT_PERM_AVX) {

  262.         fft_perm_avx(s);

  263.     } else {

  264. #define PROCESS_FFT_PERM_SWAP_LSBS(num) do {\

  265.     for(i = 0; i < n; i++) {\

  266.         int k;\

  267.         j = i;\

  268.         j = (j & ~3) | ((j >> 1) & 1) | ((j << 1) & 2);\

  269.         k = -split_radix_permutation(i, n, s->inverse) & (n - 1);\

  270.         s->revtab##num[k] = j;\

  271.     } \

  272. } while(0);

  273. 

  274. #define PROCESS_FFT_PERM_DEFAULT(num) do {\

  275.     for(i = 0; i < n; i++) {\

  276.         int k;\

  277.         j = i;\

  278.         k = -split_radix_permutation(i, n, s->inverse) & (n - 1);\

  279.         s->revtab##num[k] = j;\

  280.     } \

  281. } while(0);

  282. 

  283. #define SPLIT_RADIX_PERMUTATION(num) do { \

  284.     if (s->fft_permutation == FF_FFT_PERM_SWAP_LSBS) {\

  285.         PROCESS_FFT_PERM_SWAP_LSBS(num) \

  286.     } else {\

  287.         PROCESS_FFT_PERM_DEFAULT(num) \

  288.     }\

  289. } while(0);

  290. 

  291.     if (s->revtab)

  292.         SPLIT_RADIX_PERMUTATION()

  293.     if (s->revtab32)

  294.         SPLIT_RADIX_PERMUTATION(32)

  295. 

  296. #undef PROCESS_FFT_PERM_DEFAULT

  297. #undef PROCESS_FFT_PERM_SWAP_LSBS

  298. #undef SPLIT_RADIX_PERMUTATION

  299.     }

  300. 

  301.     return 0;

  302.  fail:

  303.     av_freep(&s->revtab);

  304.     av_freep(&s->revtab32);

  305.     av_freep(&s->tmp_buf);

  306.     return -1;

  307. }

  308. 

  309. static void fft_permute_c(FFTContext *s, FFTComplex *z)

  310. {

  311.     int j, np;

  312.     const uint16_t *revtab = s->revtab;

  313.     const uint32_t *revtab32 = s->revtab32;

  314.     np = 1 << s->nbits;

  315.     /* TODO: handle split-radix permute in a more optimal way, probably in-place */

  316.     if (revtab) {

  317.         for(j=0;j<np;j++) s->tmp_buf[revtab[j]] = z[j];

  318.     } else

  319.         for(j=0;j<np;j++) s->tmp_buf[revtab32[j]] = z[j];

  320. 

  321.     memcpy(z, s->tmp_buf, np * sizeof(FFTComplex));

  322. }

  323. 

  324. av_cold void ff_fft_end(FFTContext *s)

  325. {

  326.     av_freep(&s->revtab);

  327.     av_freep(&s->revtab32);

  328.     av_freep(&s->tmp_buf);

  329. }

  330. 

  331. #if FFT_FIXED_32

  332. 

  333. static void fft_calc_c(FFTContext *s, FFTComplex *z) {

  334. 

  335.     int nbits, i, n, num_transforms, offset, step;

  336.     int n4, n2, n34;

  337.     unsigned tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7, tmp8;

  338.     FFTComplex *tmpz;

  339.     const int fft_size = (1 << s->nbits);

  340.     int64_t accu;

  341. 

  342.     num_transforms = (0x2aab >> (16 - s->nbits)) | 1;

  343. 

  344.     for (n=0; n<num_transforms; n++){

  345.         offset = ff_fft_offsets_lut[n] << 2;

  346.         tmpz = z + offset;

  347. 

  348.         tmp1 = tmpz[0].re + (unsigned)tmpz[1].re;

  349.         tmp5 = tmpz[2].re + (unsigned)tmpz[3].re;

  350.         tmp2 = tmpz[0].im + (unsigned)tmpz[1].im;

  351.         tmp6 = tmpz[2].im + (unsigned)tmpz[3].im;

  352.         tmp3 = tmpz[0].re - (unsigned)tmpz[1].re;

  353.         tmp8 = tmpz[2].im - (unsigned)tmpz[3].im;

  354.         tmp4 = tmpz[0].im - (unsigned)tmpz[1].im;

  355.         tmp7 = tmpz[2].re - (unsigned)tmpz[3].re;

  356. 

  357.         tmpz[0].re = tmp1 + tmp5;

  358.         tmpz[2].re = tmp1 - tmp5;

  359.         tmpz[0].im = tmp2 + tmp6;

  360.         tmpz[2].im = tmp2 - tmp6;

  361.         tmpz[1].re = tmp3 + tmp8;

  362.         tmpz[3].re = tmp3 - tmp8;

  363.         tmpz[1].im = tmp4 - tmp7;

  364.         tmpz[3].im = tmp4 + tmp7;

  365.     }

  366. 

  367.     if (fft_size < 8)

  368.         return;

  369. 

  370.     num_transforms = (num_transforms >> 1) | 1;

  371. 

  372.     for (n=0; n<num_transforms; n++){

  373.         offset = ff_fft_offsets_lut[n] << 3;

  374.         tmpz = z + offset;

  375. 

  376.         tmp1 = tmpz[4].re + (unsigned)tmpz[5].re;

  377.         tmp3 = tmpz[6].re + (unsigned)tmpz[7].re;

  378.         tmp2 = tmpz[4].im + (unsigned)tmpz[5].im;

  379.         tmp4 = tmpz[6].im + (unsigned)tmpz[7].im;

  380.         tmp5 = tmp1 + tmp3;

  381.         tmp7 = tmp1 - tmp3;

  382.         tmp6 = tmp2 + tmp4;

  383.         tmp8 = tmp2 - tmp4;

  384. 

  385.         tmp1 = tmpz[4].re - (unsigned)tmpz[5].re;

  386.         tmp2 = tmpz[4].im - (unsigned)tmpz[5].im;

  387.         tmp3 = tmpz[6].re - (unsigned)tmpz[7].re;

  388.         tmp4 = tmpz[6].im - (unsigned)tmpz[7].im;

  389. 

  390.         tmpz[4].re = tmpz[0].re - tmp5;

  391.         tmpz[0].re = tmpz[0].re + tmp5;

  392.         tmpz[4].im = tmpz[0].im - tmp6;

  393.         tmpz[0].im = tmpz[0].im + tmp6;

  394.         tmpz[6].re = tmpz[2].re - tmp8;

  395.         tmpz[2].re = tmpz[2].re + tmp8;

  396.         tmpz[6].im = tmpz[2].im + tmp7;

  397.         tmpz[2].im = tmpz[2].im - tmp7;

  398. 

  399.         accu = (int64_t)Q31(M_SQRT1_2)*(int)(tmp1 + tmp2);

  400.         tmp5 = (int32_t)((accu + 0x40000000) >> 31);

  401.         accu = (int64_t)Q31(M_SQRT1_2)*(int)(tmp3 - tmp4);

  402.         tmp7 = (int32_t)((accu + 0x40000000) >> 31);

  403.         accu = (int64_t)Q31(M_SQRT1_2)*(int)(tmp2 - tmp1);

  404.         tmp6 = (int32_t)((accu + 0x40000000) >> 31);

  405.         accu = (int64_t)Q31(M_SQRT1_2)*(int)(tmp3 + tmp4);

  406.         tmp8 = (int32_t)((accu + 0x40000000) >> 31);

  407.         tmp1 = tmp5 + tmp7;

  408.         tmp3 = tmp5 - tmp7;

  409.         tmp2 = tmp6 + tmp8;

  410.         tmp4 = tmp6 - tmp8;

  411. 

  412.         tmpz[5].re = tmpz[1].re - tmp1;

  413.         tmpz[1].re = tmpz[1].re + tmp1;

  414.         tmpz[5].im = tmpz[1].im - tmp2;

  415.         tmpz[1].im = tmpz[1].im + tmp2;

  416.         tmpz[7].re = tmpz[3].re - tmp4;

  417.         tmpz[3].re = tmpz[3].re + tmp4;

  418.         tmpz[7].im = tmpz[3].im + tmp3;

  419.         tmpz[3].im = tmpz[3].im - tmp3;

  420.     }

  421. 

  422.     step = 1 << ((MAX_LOG2_NFFT-4) - 4);

  423.     n4 = 4;

  424. 

  425.     for (nbits=4; nbits<=s->nbits; nbits++){

  426.         n2  = 2*n4;

  427.         n34 = 3*n4;

  428.         num_transforms = (num_transforms >> 1) | 1;

  429. 

  430.         for (n=0; n<num_transforms; n++){

  431.             const FFTSample *w_re_ptr = ff_w_tab_sr + step;

  432.             const FFTSample *w_im_ptr = ff_w_tab_sr + MAX_FFT_SIZE/(4*16) - step;

  433.             offset = ff_fft_offsets_lut[n] << nbits;

  434.             tmpz = z + offset;

  435. 

  436.             tmp5 = tmpz[ n2].re + (unsigned)tmpz[n34].re;

  437.             tmp1 = tmpz[ n2].re - (unsigned)tmpz[n34].re;

  438.             tmp6 = tmpz[ n2].im + (unsigned)tmpz[n34].im;

  439.             tmp2 = tmpz[ n2].im - (unsigned)tmpz[n34].im;

  440. 

  441.             tmpz[ n2].re = tmpz[ 0].re - tmp5;

  442.             tmpz[  0].re = tmpz[ 0].re + tmp5;

  443.             tmpz[ n2].im = tmpz[ 0].im - tmp6;

  444.             tmpz[  0].im = tmpz[ 0].im + tmp6;

  445.             tmpz[n34].re = tmpz[n4].re - tmp2;

  446.             tmpz[ n4].re = tmpz[n4].re + tmp2;

  447.             tmpz[n34].im = tmpz[n4].im + tmp1;

  448.             tmpz[ n4].im = tmpz[n4].im - tmp1;

  449. 

  450.             for (i=1; i<n4; i++){

  451.                 FFTSample w_re = w_re_ptr[0];

  452.                 FFTSample w_im = w_im_ptr[0];

  453.                 accu  = (int64_t)w_re*tmpz[ n2+i].re;

  454.                 accu += (int64_t)w_im*tmpz[ n2+i].im;

  455.                 tmp1 = (int32_t)((accu + 0x40000000) >> 31);

  456.                 accu  = (int64_t)w_re*tmpz[ n2+i].im;

  457.                 accu -= (int64_t)w_im*tmpz[ n2+i].re;

  458.                 tmp2 = (int32_t)((accu + 0x40000000) >> 31);

  459.                 accu  = (int64_t)w_re*tmpz[n34+i].re;

  460.                 accu -= (int64_t)w_im*tmpz[n34+i].im;

  461.                 tmp3 = (int32_t)((accu + 0x40000000) >> 31);

  462.                 accu  = (int64_t)w_re*tmpz[n34+i].im;

  463.                 accu += (int64_t)w_im*tmpz[n34+i].re;

  464.                 tmp4 = (int32_t)((accu + 0x40000000) >> 31);

  465. 

  466.                 tmp5 = tmp1 + tmp3;

  467.                 tmp1 = tmp1 - tmp3;

  468.                 tmp6 = tmp2 + tmp4;

  469.                 tmp2 = tmp2 - tmp4;

  470. 

  471.                 tmpz[ n2+i].re = tmpz[   i].re - tmp5;

  472.                 tmpz[    i].re = tmpz[   i].re + tmp5;

  473.                 tmpz[ n2+i].im = tmpz[   i].im - tmp6;

  474.                 tmpz[    i].im = tmpz[   i].im + tmp6;

  475.                 tmpz[n34+i].re = tmpz[n4+i].re - tmp2;

  476.                 tmpz[ n4+i].re = tmpz[n4+i].re + tmp2;

  477.                 tmpz[n34+i].im = tmpz[n4+i].im + tmp1;

  478.                 tmpz[ n4+i].im = tmpz[n4+i].im - tmp1;

  479. 

  480.                 w_re_ptr += step;

  481.                 w_im_ptr -= step;

  482.             }

  483.         }

  484.         step >>= 1;

  485.         n4   <<= 1;

  486.     }

  487. }

  488. 

  489. #else /* FFT_FIXED_32 */

  490. 

  491. #define BUTTERFLIES(a0,a1,a2,a3) {\

  492.     BF(t3, t5, t5, t1);\

  493.     BF(a2.re, a0.re, a0.re, t5);\

  494.     BF(a3.im, a1.im, a1.im, t3);\

  495.     BF(t4, t6, t2, t6);\

  496.     BF(a3.re, a1.re, a1.re, t4);\

  497.     BF(a2.im, a0.im, a0.im, t6);\

  498. }

  499. 

  500. // force loading all the inputs before storing any.

  501. // this is slightly slower for small data, but avoids store->load aliasing

  502. // for addresses separated by large powers of 2.

  503. #define BUTTERFLIES_BIG(a0,a1,a2,a3) {\

  504.     FFTSample r0=a0.re, i0=a0.im, r1=a1.re, i1=a1.im;\

  505.     BF(t3, t5, t5, t1);\

  506.     BF(a2.re, a0.re, r0, t5);\

  507.     BF(a3.im, a1.im, i1, t3);\

  508.     BF(t4, t6, t2, t6);\

  509.     BF(a3.re, a1.re, r1, t4);\

  510.     BF(a2.im, a0.im, i0, t6);\

  511. }

  512. 

  513. #define TRANSFORM(a0,a1,a2,a3,wre,wim) {\

  514.     CMUL(t1, t2, a2.re, a2.im, wre, -wim);\

  515.     CMUL(t5, t6, a3.re, a3.im, wre,  wim);\

  516.     BUTTERFLIES(a0,a1,a2,a3)\

  517. }

  518. 

  519. #define TRANSFORM_ZERO(a0,a1,a2,a3) {\

  520.     t1 = a2.re;\

  521.     t2 = a2.im;\

  522.     t5 = a3.re;\

  523.     t6 = a3.im;\

  524.     BUTTERFLIES(a0,a1,a2,a3)\

  525. }

  526. 

  527. /* z[0...8n-1], w[1...2n-1] */

  528. #define PASS(name)\

  529. static void name(FFTComplex *z, const FFTSample *wre, unsigned int n)\

  530. {\

  531.     FFTDouble t1, t2, t3, t4, t5, t6;\

  532.     int o1 = 2*n;\

  533.     int o2 = 4*n;\

  534.     int o3 = 6*n;\

  535.     const FFTSample *wim = wre+o1;\

  536.     n--;\

  537. \

  538.     TRANSFORM_ZERO(z[0],z[o1],z[o2],z[o3]);\

  539.     TRANSFORM(z[1],z[o1+1],z[o2+1],z[o3+1],wre[1],wim[-1]);\

  540.     do {\

  541.         z += 2;\

  542.         wre += 2;\

  543.         wim -= 2;\

  544.         TRANSFORM(z[0],z[o1],z[o2],z[o3],wre[0],wim[0]);\

  545.         TRANSFORM(z[1],z[o1+1],z[o2+1],z[o3+1],wre[1],wim[-1]);\

  546.     } while(--n);\

  547. }

  548. 

  549. PASS(pass)

  550. #if !CONFIG_SMALL

  551. #undef BUTTERFLIES

  552. #define BUTTERFLIES BUTTERFLIES_BIG

  553. PASS(pass_big)

  554. #endif

  555. 

  556. #define DECL_FFT(n,n2,n4)\

  557. static void fft##n(FFTComplex *z)\

  558. {\

  559.     fft##n2(z);\

  560.     fft##n4(z+n4*2);\

  561.     fft##n4(z+n4*3);\

  562.     pass(z,FFT_NAME(ff_cos_##n),n4/2);\

  563. }

  564. 

  565. static void fft4(FFTComplex *z)

  566. {

  567.     FFTDouble t1, t2, t3, t4, t5, t6, t7, t8;

  568. 

  569.     BF(t3, t1, z[0].re, z[1].re);

  570.     BF(t8, t6, z[3].re, z[2].re);

  571.     BF(z[2].re, z[0].re, t1, t6);

  572.     BF(t4, t2, z[0].im, z[1].im);

  573.     BF(t7, t5, z[2].im, z[3].im);

  574.     BF(z[3].im, z[1].im, t4, t8);

  575.     BF(z[3].re, z[1].re, t3, t7);

  576.     BF(z[2].im, z[0].im, t2, t5);

  577. }

  578. 

  579. static void fft8(FFTComplex *z)

  580. {

  581.     FFTDouble t1, t2, t3, t4, t5, t6;

  582. 

  583.     fft4(z);

  584. 

  585.     BF(t1, z[5].re, z[4].re, -z[5].re);

  586.     BF(t2, z[5].im, z[4].im, -z[5].im);

  587.     BF(t5, z[7].re, z[6].re, -z[7].re);

  588.     BF(t6, z[7].im, z[6].im, -z[7].im);

  589. 

  590.     BUTTERFLIES(z[0],z[2],z[4],z[6]);

  591.     TRANSFORM(z[1],z[3],z[5],z[7],sqrthalf,sqrthalf);

  592. }

  593. 

  594. #if !CONFIG_SMALL

  595. static void fft16(FFTComplex *z)

  596. {

  597.     FFTDouble t1, t2, t3, t4, t5, t6;

  598.     FFTSample cos_16_1 = FFT_NAME(ff_cos_16)[1];

  599.     FFTSample cos_16_3 = FFT_NAME(ff_cos_16)[3];

  600. 

  601.     fft8(z);

  602.     fft4(z+8);

  603.     fft4(z+12);

  604. 

  605.     TRANSFORM_ZERO(z[0],z[4],z[8],z[12]);

  606.     TRANSFORM(z[2],z[6],z[10],z[14],sqrthalf,sqrthalf);

  607.     TRANSFORM(z[1],z[5],z[9],z[13],cos_16_1,cos_16_3);

  608.     TRANSFORM(z[3],z[7],z[11],z[15],cos_16_3,cos_16_1);

  609. }

  610. #else

  611. DECL_FFT(16,8,4)

  612. #endif

  613. DECL_FFT(32,16,8)

  614. DECL_FFT(64,32,16)

  615. DECL_FFT(128,64,32)

  616. DECL_FFT(256,128,64)

  617. DECL_FFT(512,256,128)

  618. #if !CONFIG_SMALL

  619. #define pass pass_big

  620. #endif

  621. DECL_FFT(1024,512,256)

  622. DECL_FFT(2048,1024,512)

  623. DECL_FFT(4096,2048,1024)

  624. DECL_FFT(8192,4096,2048)

  625. DECL_FFT(16384,8192,4096)

  626. DECL_FFT(32768,16384,8192)

  627. DECL_FFT(65536,32768,16384)

  628. DECL_FFT(131072,65536,32768)

  629. 

  630. static void (* const fft_dispatch[])(FFTComplex*) = {

  631.     fft4, fft8, fft16, fft32, fft64, fft128, fft256, fft512, fft1024,

  632.     fft2048, fft4096, fft8192, fft16384, fft32768, fft65536, fft131072

  633. };

  634. 

  635. static void fft_calc_c(FFTContext *s, FFTComplex *z)

  636. {

  637.     fft_dispatch[s->nbits-2](z);

  638. }

  639. #endif /* FFT_FIXED_32 */