STM32F769IDiscovery  1.00
uDANTE Audio Networking with STM32F7 DISCO board
arm_cfft_radix8_f32.c
Go to the documentation of this file.
1 /* ----------------------------------------------------------------------
2 * Copyright (C) 2010-2014 ARM Limited. All rights reserved.
3 *
4 * $Date: 19. March 2015
5 * $Revision: V.1.4.5
6 *
7 * Project: CMSIS DSP Library
8 * Title: arm_cfft_radix8_f32.c
9 *
10 * Description: Radix-8 Decimation in Frequency CFFT & CIFFT Floating point processing function
11 *
12 * Target Processor: Cortex-M4/Cortex-M3/Cortex-M0
13 *
14 * Redistribution and use in source and binary forms, with or without
15 * modification, are permitted provided that the following conditions
16 * are met:
17 * - Redistributions of source code must retain the above copyright
18 * notice, this list of conditions and the following disclaimer.
19 * - Redistributions in binary form must reproduce the above copyright
20 * notice, this list of conditions and the following disclaimer in
21 * the documentation and/or other materials provided with the
22 * distribution.
23 * - Neither the name of ARM LIMITED nor the names of its contributors
24 * may be used to endorse or promote products derived from this
25 * software without specific prior written permission.
26 *
27 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
28 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
29 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
30 * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
31 * COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
32 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
33 * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
34 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
35 * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
36 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
37 * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
38 * POSSIBILITY OF SUCH DAMAGE.
39 * -------------------------------------------------------------------- */
40 
41 #include "arm_math.h"
42 
131 /*
132 * @brief Core function for the floating-point CFFT butterfly process.
133 * @param[in, out] *pSrc points to the in-place buffer of floating-point data type.
134 * @param[in] fftLen length of the FFT.
135 * @param[in] *pCoef points to the twiddle coefficient buffer.
136 * @param[in] twidCoefModifier twiddle coefficient modifier that supports different size FFTs with the same twiddle factor table.
137 * @return none.
138 */
139 
140 void arm_radix8_butterfly_f32(
141 float32_t * pSrc,
142 uint16_t fftLen,
143 const float32_t * pCoef,
144 uint16_t twidCoefModifier)
145 {
146  uint32_t ia1, ia2, ia3, ia4, ia5, ia6, ia7;
147  uint32_t i1, i2, i3, i4, i5, i6, i7, i8;
148  uint32_t id;
149  uint32_t n1, n2, j;
150 
151  float32_t r1, r2, r3, r4, r5, r6, r7, r8;
152  float32_t t1, t2;
153  float32_t s1, s2, s3, s4, s5, s6, s7, s8;
154  float32_t p1, p2, p3, p4;
155  float32_t co2, co3, co4, co5, co6, co7, co8;
156  float32_t si2, si3, si4, si5, si6, si7, si8;
157  const float32_t C81 = 0.70710678118f;
158 
159  n2 = fftLen;
160 
161  do
162  {
163  n1 = n2;
164  n2 = n2 >> 3;
165  i1 = 0;
166 
167  do
168  {
169  i2 = i1 + n2;
170  i3 = i2 + n2;
171  i4 = i3 + n2;
172  i5 = i4 + n2;
173  i6 = i5 + n2;
174  i7 = i6 + n2;
175  i8 = i7 + n2;
176  r1 = pSrc[2 * i1] + pSrc[2 * i5];
177  r5 = pSrc[2 * i1] - pSrc[2 * i5];
178  r2 = pSrc[2 * i2] + pSrc[2 * i6];
179  r6 = pSrc[2 * i2] - pSrc[2 * i6];
180  r3 = pSrc[2 * i3] + pSrc[2 * i7];
181  r7 = pSrc[2 * i3] - pSrc[2 * i7];
182  r4 = pSrc[2 * i4] + pSrc[2 * i8];
183  r8 = pSrc[2 * i4] - pSrc[2 * i8];
184  t1 = r1 - r3;
185  r1 = r1 + r3;
186  r3 = r2 - r4;
187  r2 = r2 + r4;
188  pSrc[2 * i1] = r1 + r2;
189  pSrc[2 * i5] = r1 - r2;
190  r1 = pSrc[2 * i1 + 1] + pSrc[2 * i5 + 1];
191  s5 = pSrc[2 * i1 + 1] - pSrc[2 * i5 + 1];
192  r2 = pSrc[2 * i2 + 1] + pSrc[2 * i6 + 1];
193  s6 = pSrc[2 * i2 + 1] - pSrc[2 * i6 + 1];
194  s3 = pSrc[2 * i3 + 1] + pSrc[2 * i7 + 1];
195  s7 = pSrc[2 * i3 + 1] - pSrc[2 * i7 + 1];
196  r4 = pSrc[2 * i4 + 1] + pSrc[2 * i8 + 1];
197  s8 = pSrc[2 * i4 + 1] - pSrc[2 * i8 + 1];
198  t2 = r1 - s3;
199  r1 = r1 + s3;
200  s3 = r2 - r4;
201  r2 = r2 + r4;
202  pSrc[2 * i1 + 1] = r1 + r2;
203  pSrc[2 * i5 + 1] = r1 - r2;
204  pSrc[2 * i3] = t1 + s3;
205  pSrc[2 * i7] = t1 - s3;
206  pSrc[2 * i3 + 1] = t2 - r3;
207  pSrc[2 * i7 + 1] = t2 + r3;
208  r1 = (r6 - r8) * C81;
209  r6 = (r6 + r8) * C81;
210  r2 = (s6 - s8) * C81;
211  s6 = (s6 + s8) * C81;
212  t1 = r5 - r1;
213  r5 = r5 + r1;
214  r8 = r7 - r6;
215  r7 = r7 + r6;
216  t2 = s5 - r2;
217  s5 = s5 + r2;
218  s8 = s7 - s6;
219  s7 = s7 + s6;
220  pSrc[2 * i2] = r5 + s7;
221  pSrc[2 * i8] = r5 - s7;
222  pSrc[2 * i6] = t1 + s8;
223  pSrc[2 * i4] = t1 - s8;
224  pSrc[2 * i2 + 1] = s5 - r7;
225  pSrc[2 * i8 + 1] = s5 + r7;
226  pSrc[2 * i6 + 1] = t2 - r8;
227  pSrc[2 * i4 + 1] = t2 + r8;
228 
229  i1 += n1;
230  } while(i1 < fftLen);
231 
232  if(n2 < 8)
233  break;
234 
235  ia1 = 0;
236  j = 1;
237 
238  do
239  {
240  /* index calculation for the coefficients */
241  id = ia1 + twidCoefModifier;
242  ia1 = id;
243  ia2 = ia1 + id;
244  ia3 = ia2 + id;
245  ia4 = ia3 + id;
246  ia5 = ia4 + id;
247  ia6 = ia5 + id;
248  ia7 = ia6 + id;
249 
250  co2 = pCoef[2 * ia1];
251  co3 = pCoef[2 * ia2];
252  co4 = pCoef[2 * ia3];
253  co5 = pCoef[2 * ia4];
254  co6 = pCoef[2 * ia5];
255  co7 = pCoef[2 * ia6];
256  co8 = pCoef[2 * ia7];
257  si2 = pCoef[2 * ia1 + 1];
258  si3 = pCoef[2 * ia2 + 1];
259  si4 = pCoef[2 * ia3 + 1];
260  si5 = pCoef[2 * ia4 + 1];
261  si6 = pCoef[2 * ia5 + 1];
262  si7 = pCoef[2 * ia6 + 1];
263  si8 = pCoef[2 * ia7 + 1];
264 
265  i1 = j;
266 
267  do
268  {
269  /* index calculation for the input */
270  i2 = i1 + n2;
271  i3 = i2 + n2;
272  i4 = i3 + n2;
273  i5 = i4 + n2;
274  i6 = i5 + n2;
275  i7 = i6 + n2;
276  i8 = i7 + n2;
277  r1 = pSrc[2 * i1] + pSrc[2 * i5];
278  r5 = pSrc[2 * i1] - pSrc[2 * i5];
279  r2 = pSrc[2 * i2] + pSrc[2 * i6];
280  r6 = pSrc[2 * i2] - pSrc[2 * i6];
281  r3 = pSrc[2 * i3] + pSrc[2 * i7];
282  r7 = pSrc[2 * i3] - pSrc[2 * i7];
283  r4 = pSrc[2 * i4] + pSrc[2 * i8];
284  r8 = pSrc[2 * i4] - pSrc[2 * i8];
285  t1 = r1 - r3;
286  r1 = r1 + r3;
287  r3 = r2 - r4;
288  r2 = r2 + r4;
289  pSrc[2 * i1] = r1 + r2;
290  r2 = r1 - r2;
291  s1 = pSrc[2 * i1 + 1] + pSrc[2 * i5 + 1];
292  s5 = pSrc[2 * i1 + 1] - pSrc[2 * i5 + 1];
293  s2 = pSrc[2 * i2 + 1] + pSrc[2 * i6 + 1];
294  s6 = pSrc[2 * i2 + 1] - pSrc[2 * i6 + 1];
295  s3 = pSrc[2 * i3 + 1] + pSrc[2 * i7 + 1];
296  s7 = pSrc[2 * i3 + 1] - pSrc[2 * i7 + 1];
297  s4 = pSrc[2 * i4 + 1] + pSrc[2 * i8 + 1];
298  s8 = pSrc[2 * i4 + 1] - pSrc[2 * i8 + 1];
299  t2 = s1 - s3;
300  s1 = s1 + s3;
301  s3 = s2 - s4;
302  s2 = s2 + s4;
303  r1 = t1 + s3;
304  t1 = t1 - s3;
305  pSrc[2 * i1 + 1] = s1 + s2;
306  s2 = s1 - s2;
307  s1 = t2 - r3;
308  t2 = t2 + r3;
309  p1 = co5 * r2;
310  p2 = si5 * s2;
311  p3 = co5 * s2;
312  p4 = si5 * r2;
313  pSrc[2 * i5] = p1 + p2;
314  pSrc[2 * i5 + 1] = p3 - p4;
315  p1 = co3 * r1;
316  p2 = si3 * s1;
317  p3 = co3 * s1;
318  p4 = si3 * r1;
319  pSrc[2 * i3] = p1 + p2;
320  pSrc[2 * i3 + 1] = p3 - p4;
321  p1 = co7 * t1;
322  p2 = si7 * t2;
323  p3 = co7 * t2;
324  p4 = si7 * t1;
325  pSrc[2 * i7] = p1 + p2;
326  pSrc[2 * i7 + 1] = p3 - p4;
327  r1 = (r6 - r8) * C81;
328  r6 = (r6 + r8) * C81;
329  s1 = (s6 - s8) * C81;
330  s6 = (s6 + s8) * C81;
331  t1 = r5 - r1;
332  r5 = r5 + r1;
333  r8 = r7 - r6;
334  r7 = r7 + r6;
335  t2 = s5 - s1;
336  s5 = s5 + s1;
337  s8 = s7 - s6;
338  s7 = s7 + s6;
339  r1 = r5 + s7;
340  r5 = r5 - s7;
341  r6 = t1 + s8;
342  t1 = t1 - s8;
343  s1 = s5 - r7;
344  s5 = s5 + r7;
345  s6 = t2 - r8;
346  t2 = t2 + r8;
347  p1 = co2 * r1;
348  p2 = si2 * s1;
349  p3 = co2 * s1;
350  p4 = si2 * r1;
351  pSrc[2 * i2] = p1 + p2;
352  pSrc[2 * i2 + 1] = p3 - p4;
353  p1 = co8 * r5;
354  p2 = si8 * s5;
355  p3 = co8 * s5;
356  p4 = si8 * r5;
357  pSrc[2 * i8] = p1 + p2;
358  pSrc[2 * i8 + 1] = p3 - p4;
359  p1 = co6 * r6;
360  p2 = si6 * s6;
361  p3 = co6 * s6;
362  p4 = si6 * r6;
363  pSrc[2 * i6] = p1 + p2;
364  pSrc[2 * i6 + 1] = p3 - p4;
365  p1 = co4 * t1;
366  p2 = si4 * t2;
367  p3 = co4 * t2;
368  p4 = si4 * t1;
369  pSrc[2 * i4] = p1 + p2;
370  pSrc[2 * i4 + 1] = p3 - p4;
371 
372  i1 += n1;
373  } while(i1 < fftLen);
374 
375  j++;
376  } while(j < n2);
377 
378  twidCoefModifier <<= 3;
379  } while(n2 > 7);
380 }
381 
float32_t
float float32_t
32-bit floating-point type definition.
Definition: arm_math.h:407
arm_math.h
arm_radix8_butterfly_f32
void arm_radix8_butterfly_f32(float32_t *pSrc, uint16_t fftLen, const float32_t *pCoef, uint16_t twidCoefModifier)
Definition: arm_cfft_radix8_f32.c:140
Generated by   doxygen 1.8.11