← Back to branch summary

~ubuntu-branches/ubuntu/maverick/blender/maverick

« back to all changes in this revision

Viewing changes to extern/fftw/dft/codelets/standard/n1_12.c

  • Committer: Bazaar Package Importer
  • Author(s): Khashayar Naderehvandi, Khashayar Naderehvandi, Alessio Treglia
  • Date: 2009-01-22 16:53:59 UTC
  • mfrom: (14.1.1 experimental)
  • Revision ID: james.westby@ubuntu.com-20090122165359-v0996tn7fbit64ni
Tags: 2.48a+dfsg-1ubuntu1
https://launchpad.net/bugs/320045
[ Khashayar Naderehvandi ]
* Merge from debian experimental (LP: #320045), Ubuntu remaining changes:
  - Add patch correcting header file locations.
  - Add libvorbis-dev and libgsm1-dev to Build-Depends.
  - Use avcodec_decode_audio2() in source/blender/src/hddaudio.c

[ Alessio Treglia ]
* Add missing previous changelog entries.

Show diffs side-by-side

added added

removed removed

Lines of Context:
extern/fftw/dft/codelets/standard/n1_12.c
 
1
/*
 
2
 * Copyright (c) 2003, 2006 Matteo Frigo
 
3
 * Copyright (c) 2003, 2006 Massachusetts Institute of Technology
 
4
 *
 
5
 * This program is free software; you can redistribute it and/or modify
 
6
 * it under the terms of the GNU General Public License as published by
 
7
 * the Free Software Foundation; either version 2 of the License, or
 
8
 * (at your option) any later version.
 
9
 *
 
10
 * This program is distributed in the hope that it will be useful,
 
11
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 
12
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 
13
 * GNU General Public License for more details.
 
14
 *
 
15
 * You should have received a copy of the GNU General Public License
 
16
 * along with this program; if not, write to the Free Software
 
17
 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
 
18
 *
 
19
 */
 
20
 
 
21
/* This file was automatically generated --- DO NOT EDIT */
 
22
/* Generated on Sat Jul  1 13:54:27 EDT 2006 */
 
23
 
 
24
#include "codelet-dft.h"
 
25
 
 
26
#ifdef HAVE_FMA
 
27
 
 
28
/* Generated by: ../../../genfft/gen_notw -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 12 -name n1_12 -include n.h */
 
29
 
 
30
/*
 
31
 * This function contains 96 FP additions, 24 FP multiplications,
 
32
 * (or, 72 additions, 0 multiplications, 24 fused multiply/add),
 
33
 * 63 stack variables, and 48 memory accesses
 
34
 */
 
35
/*
 
36
 * Generator Id's : 
 
37
 * $Id: algsimp.ml,v 1.9 2006-02-12 23:34:12 athena Exp $
 
38
 * $Id: fft.ml,v 1.4 2006-01-05 03:04:27 stevenj Exp $
 
39
 * $Id: gen_notw.ml,v 1.30 2006-02-12 23:34:12 athena Exp $
 
40
 */
 
41
 
 
42
#include "n.h"
 
43
 
 
44
static void n1_12(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
 
45
{
 
46
     DK(KP866025403, +0.866025403784438646763723170752936183471402627);
 
47
     DK(KP500000000, +0.500000000000000000000000000000000000000000000);
 
48
     INT i;
 
49
     for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(is), MAKE_VOLATILE_STRIDE(os)) {
 
50
          E TT, TW, TF, T1q, TY, TQ, TX, T1n;
 
51
          {
 
52
               E TA, TS, TR, T5, Ts, Tz, TD, TV, TU, Ta, Tx, TC, T1d, Th, TJ;
 
53
               E TG, Tg, T1u, T1c, T1f, TM, TN, Tk, T1i;
 
54
               {
 
55
                    E T6, Tt, Tu, Tv, T9;
 
56
                    {
 
57
                         E T1, To, Tp, Tq, T4, T2, T3, T7, T8, Tr;
 
58
                         T1 = ri[0];
 
59
                         T2 = ri[WS(is, 4)];
 
60
                         T3 = ri[WS(is, 8)];
 
61
                         To = ii[0];
 
62
                         Tp = ii[WS(is, 4)];
 
63
                         Tq = ii[WS(is, 8)];
 
64
                         T4 = T2 + T3;
 
65
                         TA = T3 - T2;
 
66
                         T6 = ri[WS(is, 6)];
 
67
                         TS = Tp - Tq;
 
68
                         Tr = Tp + Tq;
 
69
                         TR = FNMS(KP500000000, T4, T1);
 
70
                         T5 = T1 + T4;
 
71
                         T7 = ri[WS(is, 10)];
 
72
                         Ts = To + Tr;
 
73
                         Tz = FNMS(KP500000000, Tr, To);
 
74
                         T8 = ri[WS(is, 2)];
 
75
                         Tt = ii[WS(is, 6)];
 
76
                         Tu = ii[WS(is, 10)];
 
77
                         Tv = ii[WS(is, 2)];
 
78
                         T9 = T7 + T8;
 
79
                         TD = T8 - T7;
 
80
                    }
 
81
                    {
 
82
                         E Tc, T1a, TH, TI, Tf, Td, Te, Tw, Ti, Tj, T1b;
 
83
                         Tc = ri[WS(is, 3)];
 
84
                         TV = Tu - Tv;
 
85
                         Tw = Tu + Tv;
 
86
                         TU = FNMS(KP500000000, T9, T6);
 
87
                         Ta = T6 + T9;
 
88
                         Td = ri[WS(is, 7)];
 
89
                         Tx = Tt + Tw;
 
90
                         TC = FNMS(KP500000000, Tw, Tt);
 
91
                         Te = ri[WS(is, 11)];
 
92
                         T1a = ii[WS(is, 3)];
 
93
                         TH = ii[WS(is, 7)];
 
94
                         TI = ii[WS(is, 11)];
 
95
                         Tf = Td + Te;
 
96
                         T1d = Te - Td;
 
97
                         Th = ri[WS(is, 9)];
 
98
                         TJ = TH - TI;
 
99
                         T1b = TH + TI;
 
100
                         TG = FNMS(KP500000000, Tf, Tc);
 
101
                         Tg = Tc + Tf;
 
102
                         Ti = ri[WS(is, 1)];
 
103
                         T1u = T1a + T1b;
 
104
                         T1c = FNMS(KP500000000, T1b, T1a);
 
105
                         Tj = ri[WS(is, 5)];
 
106
                         T1f = ii[WS(is, 9)];
 
107
                         TM = ii[WS(is, 1)];
 
108
                         TN = ii[WS(is, 5)];
 
109
                         Tk = Ti + Tj;
 
110
                         T1i = Tj - Ti;
 
111
                    }
 
112
               }
 
113
               {
 
114
                    E T1t, TO, TL, T1h, T1w, Tb, T1g, Tl;
 
115
                    T1t = T5 - Ta;
 
116
                    Tb = T5 + Ta;
 
117
                    TO = TM - TN;
 
118
                    T1g = TM + TN;
 
119
                    TL = FNMS(KP500000000, Tk, Th);
 
120
                    Tl = Th + Tk;
 
121
                    {
 
122
                         E T1x, Ty, T1v, Tn, Tm, T1y;
 
123
                         T1x = Ts + Tx;
 
124
                         Ty = Ts - Tx;
 
125
                         T1v = T1f + T1g;
 
126
                         T1h = FNMS(KP500000000, T1g, T1f);
 
127
                         Tn = Tg - Tl;
 
128
                         Tm = Tg + Tl;
 
129
                         T1y = T1u + T1v;
 
130
                         T1w = T1u - T1v;
 
131
                         ro[0] = Tb + Tm;
 
132
                         ro[WS(os, 6)] = Tb - Tm;
 
133
                         io[WS(os, 3)] = Tn + Ty;
 
134
                         io[0] = T1x + T1y;
 
135
                         io[WS(os, 6)] = T1x - T1y;
 
136
                         io[WS(os, 9)] = Ty - Tn;
 
137
                    }
 
138
                    {
 
139
                         E TB, TE, T1o, T11, T1p, TK, TP, T15, T1k, T18, T14, T16, T1l, T1m;
 
140
                         {
 
141
                              E T1e, T1j, TZ, T10, T12, T13;
 
142
                              TB = FNMS(KP866025403, TA, Tz);
 
143
                              TZ = FMA(KP866025403, TA, Tz);
 
144
                              T10 = FMA(KP866025403, TD, TC);
 
145
                              TE = FNMS(KP866025403, TD, TC);
 
146
                              T1o = FNMS(KP866025403, T1d, T1c);
 
147
                              T1e = FMA(KP866025403, T1d, T1c);
 
148
                              ro[WS(os, 9)] = T1t + T1w;
 
149
                              ro[WS(os, 3)] = T1t - T1w;
 
150
                              T1l = TZ + T10;
 
151
                              T11 = TZ - T10;
 
152
                              T1j = FMA(KP866025403, T1i, T1h);
 
153
                              T1p = FNMS(KP866025403, T1i, T1h);
 
154
                              TK = FNMS(KP866025403, TJ, TG);
 
155
                              T12 = FMA(KP866025403, TJ, TG);
 
156
                              T13 = FMA(KP866025403, TO, TL);
 
157
                              TP = FNMS(KP866025403, TO, TL);
 
158
                              TT = FNMS(KP866025403, TS, TR);
 
159
                              T15 = FMA(KP866025403, TS, TR);
 
160
                              T1m = T1e + T1j;
 
161
                              T1k = T1e - T1j;
 
162
                              T18 = T12 + T13;
 
163
                              T14 = T12 - T13;
 
164
                              T16 = FMA(KP866025403, TV, TU);
 
165
                              TW = FNMS(KP866025403, TV, TU);
 
166
                         }
 
167
                         io[WS(os, 10)] = T1l - T1m;
 
168
                         io[WS(os, 4)] = T1l + T1m;
 
169
                         io[WS(os, 7)] = T11 + T14;
 
170
                         io[WS(os, 1)] = T11 - T14;
 
171
                         {
 
172
                              E T17, T19, T1r, T1s;
 
173
                              T17 = T15 + T16;
 
174
                              T19 = T15 - T16;
 
175
                              ro[WS(os, 7)] = T19 - T1k;
 
176
                              ro[WS(os, 1)] = T19 + T1k;
 
177
                              ro[WS(os, 4)] = T17 + T18;
 
178
                              ro[WS(os, 10)] = T17 - T18;
 
179
                              T1r = TB + TE;
 
180
                              TF = TB - TE;
 
181
                              T1s = T1o + T1p;
 
182
                              T1q = T1o - T1p;
 
183
                              TY = TK + TP;
 
184
                              TQ = TK - TP;
 
185
                              io[WS(os, 2)] = T1r - T1s;
 
186
                              io[WS(os, 8)] = T1r + T1s;
 
187
                         }
 
188
                    }
 
189
               }
 
190
          }
 
191
          io[WS(os, 11)] = TF + TQ;
 
192
          io[WS(os, 5)] = TF - TQ;
 
193
          TX = TT + TW;
 
194
          T1n = TT - TW;
 
195
          ro[WS(os, 11)] = T1n - T1q;
 
196
          ro[WS(os, 5)] = T1n + T1q;
 
197
          ro[WS(os, 8)] = TX + TY;
 
198
          ro[WS(os, 2)] = TX - TY;
 
199
     }
 
200
}
 
201
 
 
202
static const kdft_desc desc = { 12, "n1_12", {72, 0, 24, 0}, &GENUS, 0, 0, 0, 0 };
 
203
void X(codelet_n1_12) (planner *p) {
 
204
     X(kdft_register) (p, n1_12, &desc);
 
205
}
 
206
 
 
207
#else                           /* HAVE_FMA */
 
208
 
 
209
/* Generated by: ../../../genfft/gen_notw -compact -variables 4 -pipeline-latency 4 -n 12 -name n1_12 -include n.h */
 
210
 
 
211
/*
 
212
 * This function contains 96 FP additions, 16 FP multiplications,
 
213
 * (or, 88 additions, 8 multiplications, 8 fused multiply/add),
 
214
 * 43 stack variables, and 48 memory accesses
 
215
 */
 
216
/*
 
217
 * Generator Id's : 
 
218
 * $Id: algsimp.ml,v 1.9 2006-02-12 23:34:12 athena Exp $
 
219
 * $Id: fft.ml,v 1.4 2006-01-05 03:04:27 stevenj Exp $
 
220
 * $Id: gen_notw.ml,v 1.30 2006-02-12 23:34:12 athena Exp $
 
221
 */
 
222
 
 
223
#include "n.h"
 
224
 
 
225
static void n1_12(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs)
 
226
{
 
227
     DK(KP866025403, +0.866025403784438646763723170752936183471402627);
 
228
     DK(KP500000000, +0.500000000000000000000000000000000000000000000);
 
229
     INT i;
 
230
     for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(is), MAKE_VOLATILE_STRIDE(os)) {
 
231
          E T5, TR, TA, Ts, TS, Tz, Ta, TU, TD, Tx, TV, TC, Tg, T1a, TG;
 
232
          E TJ, T1u, T1d, Tl, T1f, TL, TO, T1v, T1i;
 
233
          {
 
234
               E T1, T2, T3, T4;
 
235
               T1 = ri[0];
 
236
               T2 = ri[WS(is, 4)];
 
237
               T3 = ri[WS(is, 8)];
 
238
               T4 = T2 + T3;
 
239
               T5 = T1 + T4;
 
240
               TR = FNMS(KP500000000, T4, T1);
 
241
               TA = KP866025403 * (T3 - T2);
 
242
          }
 
243
          {
 
244
               E To, Tp, Tq, Tr;
 
245
               To = ii[0];
 
246
               Tp = ii[WS(is, 4)];
 
247
               Tq = ii[WS(is, 8)];
 
248
               Tr = Tp + Tq;
 
249
               Ts = To + Tr;
 
250
               TS = KP866025403 * (Tp - Tq);
 
251
               Tz = FNMS(KP500000000, Tr, To);
 
252
          }
 
253
          {
 
254
               E T6, T7, T8, T9;
 
255
               T6 = ri[WS(is, 6)];
 
256
               T7 = ri[WS(is, 10)];
 
257
               T8 = ri[WS(is, 2)];
 
258
               T9 = T7 + T8;
 
259
               Ta = T6 + T9;
 
260
               TU = FNMS(KP500000000, T9, T6);
 
261
               TD = KP866025403 * (T8 - T7);
 
262
          }
 
263
          {
 
264
               E Tt, Tu, Tv, Tw;
 
265
               Tt = ii[WS(is, 6)];
 
266
               Tu = ii[WS(is, 10)];
 
267
               Tv = ii[WS(is, 2)];
 
268
               Tw = Tu + Tv;
 
269
               Tx = Tt + Tw;
 
270
               TV = KP866025403 * (Tu - Tv);
 
271
               TC = FNMS(KP500000000, Tw, Tt);
 
272
          }
 
273
          {
 
274
               E Tc, Td, Te, Tf;
 
275
               Tc = ri[WS(is, 3)];
 
276
               Td = ri[WS(is, 7)];
 
277
               Te = ri[WS(is, 11)];
 
278
               Tf = Td + Te;
 
279
               Tg = Tc + Tf;
 
280
               T1a = KP866025403 * (Te - Td);
 
281
               TG = FNMS(KP500000000, Tf, Tc);
 
282
          }
 
283
          {
 
284
               E T1b, TH, TI, T1c;
 
285
               T1b = ii[WS(is, 3)];
 
286
               TH = ii[WS(is, 7)];
 
287
               TI = ii[WS(is, 11)];
 
288
               T1c = TH + TI;
 
289
               TJ = KP866025403 * (TH - TI);
 
290
               T1u = T1b + T1c;
 
291
               T1d = FNMS(KP500000000, T1c, T1b);
 
292
          }
 
293
          {
 
294
               E Th, Ti, Tj, Tk;
 
295
               Th = ri[WS(is, 9)];
 
296
               Ti = ri[WS(is, 1)];
 
297
               Tj = ri[WS(is, 5)];
 
298
               Tk = Ti + Tj;
 
299
               Tl = Th + Tk;
 
300
               T1f = KP866025403 * (Tj - Ti);
 
301
               TL = FNMS(KP500000000, Tk, Th);
 
302
          }
 
303
          {
 
304
               E T1g, TM, TN, T1h;
 
305
               T1g = ii[WS(is, 9)];
 
306
               TM = ii[WS(is, 1)];
 
307
               TN = ii[WS(is, 5)];
 
308
               T1h = TM + TN;
 
309
               TO = KP866025403 * (TM - TN);
 
310
               T1v = T1g + T1h;
 
311
               T1i = FNMS(KP500000000, T1h, T1g);
 
312
          }
 
313
          {
 
314
               E Tb, Tm, T1t, T1w;
 
315
               Tb = T5 + Ta;
 
316
               Tm = Tg + Tl;
 
317
               ro[WS(os, 6)] = Tb - Tm;
 
318
               ro[0] = Tb + Tm;
 
319
               {
 
320
                    E T1x, T1y, Tn, Ty;
 
321
                    T1x = Ts + Tx;
 
322
                    T1y = T1u + T1v;
 
323
                    io[WS(os, 6)] = T1x - T1y;
 
324
                    io[0] = T1x + T1y;
 
325
                    Tn = Tg - Tl;
 
326
                    Ty = Ts - Tx;
 
327
                    io[WS(os, 3)] = Tn + Ty;
 
328
                    io[WS(os, 9)] = Ty - Tn;
 
329
               }
 
330
               T1t = T5 - Ta;
 
331
               T1w = T1u - T1v;
 
332
               ro[WS(os, 3)] = T1t - T1w;
 
333
               ro[WS(os, 9)] = T1t + T1w;
 
334
               {
 
335
                    E T11, T1l, T1k, T1m, T14, T18, T17, T19;
 
336
                    {
 
337
                         E TZ, T10, T1e, T1j;
 
338
                         TZ = TA + Tz;
 
339
                         T10 = TD + TC;
 
340
                         T11 = TZ - T10;
 
341
                         T1l = TZ + T10;
 
342
                         T1e = T1a + T1d;
 
343
                         T1j = T1f + T1i;
 
344
                         T1k = T1e - T1j;
 
345
                         T1m = T1e + T1j;
 
346
                    }
 
347
                    {
 
348
                         E T12, T13, T15, T16;
 
349
                         T12 = TG + TJ;
 
350
                         T13 = TL + TO;
 
351
                         T14 = T12 - T13;
 
352
                         T18 = T12 + T13;
 
353
                         T15 = TR + TS;
 
354
                         T16 = TU + TV;
 
355
                         T17 = T15 + T16;
 
356
                         T19 = T15 - T16;
 
357
                    }
 
358
                    io[WS(os, 1)] = T11 - T14;
 
359
                    ro[WS(os, 1)] = T19 + T1k;
 
360
                    io[WS(os, 7)] = T11 + T14;
 
361
                    ro[WS(os, 7)] = T19 - T1k;
 
362
                    ro[WS(os, 10)] = T17 - T18;
 
363
                    io[WS(os, 10)] = T1l - T1m;
 
364
                    ro[WS(os, 4)] = T17 + T18;
 
365
                    io[WS(os, 4)] = T1l + T1m;
 
366
               }
 
367
               {
 
368
                    E TF, T1r, T1q, T1s, TQ, TY, TX, T1n;
 
369
                    {
 
370
                         E TB, TE, T1o, T1p;
 
371
                         TB = Tz - TA;
 
372
                         TE = TC - TD;
 
373
                         TF = TB - TE;
 
374
                         T1r = TB + TE;
 
375
                         T1o = T1d - T1a;
 
376
                         T1p = T1i - T1f;
 
377
                         T1q = T1o - T1p;
 
378
                         T1s = T1o + T1p;
 
379
                    }
 
380
                    {
 
381
                         E TK, TP, TT, TW;
 
382
                         TK = TG - TJ;
 
383
                         TP = TL - TO;
 
384
                         TQ = TK - TP;
 
385
                         TY = TK + TP;
 
386
                         TT = TR - TS;
 
387
                         TW = TU - TV;
 
388
                         TX = TT + TW;
 
389
                         T1n = TT - TW;
 
390
                    }
 
391
                    io[WS(os, 5)] = TF - TQ;
 
392
                    ro[WS(os, 5)] = T1n + T1q;
 
393
                    io[WS(os, 11)] = TF + TQ;
 
394
                    ro[WS(os, 11)] = T1n - T1q;
 
395
                    ro[WS(os, 2)] = TX - TY;
 
396
                    io[WS(os, 2)] = T1r - T1s;
 
397
                    ro[WS(os, 8)] = TX + TY;
 
398
                    io[WS(os, 8)] = T1r + T1s;
 
399
               }
 
400
          }
 
401
     }
 
402
}
 
403
 
 
404
static const kdft_desc desc = { 12, "n1_12", {88, 8, 8, 0}, &GENUS, 0, 0, 0, 0 };
 
405
void X(codelet_n1_12) (planner *p) {
 
406
     X(kdft_register) (p, n1_12, &desc);
 
407
}
 
408
 
 
409
#endif                          /* HAVE_FMA */