FFmpeg
 All Data Structures Namespaces Files Functions Variables Typedefs Enumerations Enumerator Macros Groups Pages
h264dsp.c
Go to the documentation of this file.
1 /*
2  * Copyright (c) 2004 Romain Dolbeau <romain@dolbeau.org>
3  *
4  * This file is part of FFmpeg.
5  *
6  * FFmpeg is free software; you can redistribute it and/or
7  * modify it under the terms of the GNU Lesser General Public
8  * License as published by the Free Software Foundation; either
9  * version 2.1 of the License, or (at your option) any later version.
10  *
11  * FFmpeg is distributed in the hope that it will be useful,
12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
14  * Lesser General Public License for more details.
15  *
16  * You should have received a copy of the GNU Lesser General Public
17  * License along with FFmpeg; if not, write to the Free Software
18  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
19  */
20 
21 #include "config.h"
22 
23 #include <stdint.h>
24 #include <string.h>
25 
26 #include "libavutil/attributes.h"
27 #include "libavutil/cpu.h"
28 #include "libavutil/intreadwrite.h"
29 #include "libavutil/mem.h"
30 #include "libavutil/ppc/cpu.h"
33 
34 #include "libavcodec/h264dec.h"
35 #include "libavcodec/h264dsp.h"
36 
37 #if HAVE_ALTIVEC
38 
39 /****************************************************************************
40  * IDCT transform:
41  ****************************************************************************/
42 
43 #define VEC_1D_DCT(vb0,vb1,vb2,vb3,va0,va1,va2,va3) \
44  /* 1st stage */ \
45  vz0 = vec_add(vb0,vb2); /* temp[0] = Y[0] + Y[2] */ \
46  vz1 = vec_sub(vb0,vb2); /* temp[1] = Y[0] - Y[2] */ \
47  vz2 = vec_sra(vb1,vec_splat_u16(1)); \
48  vz2 = vec_sub(vz2,vb3); /* temp[2] = Y[1].1/2 - Y[3] */ \
49  vz3 = vec_sra(vb3,vec_splat_u16(1)); \
50  vz3 = vec_add(vb1,vz3); /* temp[3] = Y[1] + Y[3].1/2 */ \
51  /* 2nd stage: output */ \
52  va0 = vec_add(vz0,vz3); /* x[0] = temp[0] + temp[3] */ \
53  va1 = vec_add(vz1,vz2); /* x[1] = temp[1] + temp[2] */ \
54  va2 = vec_sub(vz1,vz2); /* x[2] = temp[1] - temp[2] */ \
55  va3 = vec_sub(vz0,vz3) /* x[3] = temp[0] - temp[3] */
56 
57 #define VEC_TRANSPOSE_4(a0,a1,a2,a3,b0,b1,b2,b3) \
58  b0 = vec_mergeh( a0, a0 ); \
59  b1 = vec_mergeh( a1, a0 ); \
60  b2 = vec_mergeh( a2, a0 ); \
61  b3 = vec_mergeh( a3, a0 ); \
62  a0 = vec_mergeh( b0, b2 ); \
63  a1 = vec_mergel( b0, b2 ); \
64  a2 = vec_mergeh( b1, b3 ); \
65  a3 = vec_mergel( b1, b3 ); \
66  b0 = vec_mergeh( a0, a2 ); \
67  b1 = vec_mergel( a0, a2 ); \
68  b2 = vec_mergeh( a1, a3 ); \
69  b3 = vec_mergel( a1, a3 )
70 
71 #if HAVE_BIGENDIAN
72 #define vdst_load(d) \
73  vdst_orig = vec_ld(0, dst); \
74  vdst = vec_perm(vdst_orig, zero_u8v, vdst_mask);
75 #else
76 #define vdst_load(d) vdst = vec_vsx_ld(0, dst)
77 #endif
78 
79 #define VEC_LOAD_U8_ADD_S16_STORE_U8(va) \
80  vdst_load(); \
81  vdst_ss = (vec_s16) VEC_MERGEH(zero_u8v, vdst); \
82  va = vec_add(va, vdst_ss); \
83  va_u8 = vec_packsu(va, zero_s16v); \
84  va_u32 = vec_splat((vec_u32)va_u8, 0); \
85  vec_ste(va_u32, element, (uint32_t*)dst);
86 
87 static void h264_idct_add_altivec(uint8_t *dst, int16_t *block, int stride)
88 {
89  vec_s16 va0, va1, va2, va3;
90  vec_s16 vz0, vz1, vz2, vz3;
91  vec_s16 vtmp0, vtmp1, vtmp2, vtmp3;
92  vec_u8 va_u8;
93  vec_u32 va_u32;
94  vec_s16 vdst_ss;
95  const vec_u16 v6us = vec_splat_u16(6);
96  vec_u8 vdst, vdst_orig;
97  vec_u8 vdst_mask = vec_lvsl(0, dst);
98  int element = ((unsigned long)dst & 0xf) >> 2;
99  LOAD_ZERO;
100 
101  block[0] += 32; /* add 32 as a DC-level for rounding */
102 
103  vtmp0 = vec_ld(0,block);
104  vtmp1 = vec_sld(vtmp0, vtmp0, 8);
105  vtmp2 = vec_ld(16,block);
106  vtmp3 = vec_sld(vtmp2, vtmp2, 8);
107  memset(block, 0, 16 * sizeof(int16_t));
108 
109  VEC_1D_DCT(vtmp0,vtmp1,vtmp2,vtmp3,va0,va1,va2,va3);
110  VEC_TRANSPOSE_4(va0,va1,va2,va3,vtmp0,vtmp1,vtmp2,vtmp3);
111  VEC_1D_DCT(vtmp0,vtmp1,vtmp2,vtmp3,va0,va1,va2,va3);
112 
113  va0 = vec_sra(va0,v6us);
114  va1 = vec_sra(va1,v6us);
115  va2 = vec_sra(va2,v6us);
116  va3 = vec_sra(va3,v6us);
117 
118  VEC_LOAD_U8_ADD_S16_STORE_U8(va0);
119  dst += stride;
120  VEC_LOAD_U8_ADD_S16_STORE_U8(va1);
121  dst += stride;
122  VEC_LOAD_U8_ADD_S16_STORE_U8(va2);
123  dst += stride;
124  VEC_LOAD_U8_ADD_S16_STORE_U8(va3);
125 }
126 
127 #define IDCT8_1D_ALTIVEC(s0, s1, s2, s3, s4, s5, s6, s7, d0, d1, d2, d3, d4, d5, d6, d7) {\
128  /* a0 = SRC(0) + SRC(4); */ \
129  vec_s16 a0v = vec_add(s0, s4); \
130  /* a2 = SRC(0) - SRC(4); */ \
131  vec_s16 a2v = vec_sub(s0, s4); \
132  /* a4 = (SRC(2)>>1) - SRC(6); */ \
133  vec_s16 a4v = vec_sub(vec_sra(s2, onev), s6); \
134  /* a6 = (SRC(6)>>1) + SRC(2); */ \
135  vec_s16 a6v = vec_add(vec_sra(s6, onev), s2); \
136  /* b0 = a0 + a6; */ \
137  vec_s16 b0v = vec_add(a0v, a6v); \
138  /* b2 = a2 + a4; */ \
139  vec_s16 b2v = vec_add(a2v, a4v); \
140  /* b4 = a2 - a4; */ \
141  vec_s16 b4v = vec_sub(a2v, a4v); \
142  /* b6 = a0 - a6; */ \
143  vec_s16 b6v = vec_sub(a0v, a6v); \
144  /* a1 = SRC(5) - SRC(3) - SRC(7) - (SRC(7)>>1); */ \
145  /* a1 = (SRC(5)-SRC(3)) - (SRC(7) + (SRC(7)>>1)); */ \
146  vec_s16 a1v = vec_sub( vec_sub(s5, s3), vec_add(s7, vec_sra(s7, onev)) ); \
147  /* a3 = SRC(7) + SRC(1) - SRC(3) - (SRC(3)>>1); */ \
148  /* a3 = (SRC(7)+SRC(1)) - (SRC(3) + (SRC(3)>>1)); */ \
149  vec_s16 a3v = vec_sub( vec_add(s7, s1), vec_add(s3, vec_sra(s3, onev)) );\
150  /* a5 = SRC(7) - SRC(1) + SRC(5) + (SRC(5)>>1); */ \
151  /* a5 = (SRC(7)-SRC(1)) + SRC(5) + (SRC(5)>>1); */ \
152  vec_s16 a5v = vec_add( vec_sub(s7, s1), vec_add(s5, vec_sra(s5, onev)) );\
153  /* a7 = SRC(5)+SRC(3) + SRC(1) + (SRC(1)>>1); */ \
154  vec_s16 a7v = vec_add( vec_add(s5, s3), vec_add(s1, vec_sra(s1, onev)) );\
155  /* b1 = (a7>>2) + a1; */ \
156  vec_s16 b1v = vec_add( vec_sra(a7v, twov), a1v); \
157  /* b3 = a3 + (a5>>2); */ \
158  vec_s16 b3v = vec_add(a3v, vec_sra(a5v, twov)); \
159  /* b5 = (a3>>2) - a5; */ \
160  vec_s16 b5v = vec_sub( vec_sra(a3v, twov), a5v); \
161  /* b7 = a7 - (a1>>2); */ \
162  vec_s16 b7v = vec_sub( a7v, vec_sra(a1v, twov)); \
163  /* DST(0, b0 + b7); */ \
164  d0 = vec_add(b0v, b7v); \
165  /* DST(1, b2 + b5); */ \
166  d1 = vec_add(b2v, b5v); \
167  /* DST(2, b4 + b3); */ \
168  d2 = vec_add(b4v, b3v); \
169  /* DST(3, b6 + b1); */ \
170  d3 = vec_add(b6v, b1v); \
171  /* DST(4, b6 - b1); */ \
172  d4 = vec_sub(b6v, b1v); \
173  /* DST(5, b4 - b3); */ \
174  d5 = vec_sub(b4v, b3v); \
175  /* DST(6, b2 - b5); */ \
176  d6 = vec_sub(b2v, b5v); \
177  /* DST(7, b0 - b7); */ \
178  d7 = vec_sub(b0v, b7v); \
179 }
180 
181 #if HAVE_BIGENDIAN
182 #define GET_2PERM(ldv, stv, d) \
183  ldv = vec_lvsl(0, d); \
184  stv = vec_lvsr(8, d);
185 #define dstv_load(d) \
186  vec_u8 hv = vec_ld( 0, d ); \
187  vec_u8 lv = vec_ld( 7, d); \
188  vec_u8 dstv = vec_perm( hv, lv, (vec_u8)perm_ldv );
189 #define dest_unligned_store(d) \
190  vec_u8 edgehv; \
191  vec_u8 bodyv = vec_perm( idstsum8, idstsum8, perm_stv ); \
192  vec_u8 edgelv = vec_perm( sel, zero_u8v, perm_stv ); \
193  lv = vec_sel( lv, bodyv, edgelv ); \
194  vec_st( lv, 7, d ); \
195  hv = vec_ld( 0, d ); \
196  edgehv = vec_perm( zero_u8v, sel, perm_stv ); \
197  hv = vec_sel( hv, bodyv, edgehv ); \
198  vec_st( hv, 0, d );
199 #else
200 
201 #define GET_2PERM(ldv, stv, d) {}
202 #define dstv_load(d) vec_u8 dstv = vec_vsx_ld(0, d)
203 #define dest_unligned_store(d)\
204  vec_u8 dst8 = vec_perm((vec_u8)idstsum8, dstv, vcprm(2,3,s2,s3));\
205  vec_vsx_st(dst8, 0, d)
206 #endif /* HAVE_BIGENDIAN */
207 
208 #define ALTIVEC_STORE_SUM_CLIP(dest, idctv, perm_ldv, perm_stv, sel) { \
209  /* unaligned load */ \
210  dstv_load(dest); \
211  vec_s16 idct_sh6 = vec_sra(idctv, sixv); \
212  vec_u16 dst16 = (vec_u16)VEC_MERGEH(zero_u8v, dstv); \
213  vec_s16 idstsum = vec_adds(idct_sh6, (vec_s16)dst16); \
214  vec_u8 idstsum8 = vec_packsu(zero_s16v, idstsum); \
215  /* unaligned store */ \
216  dest_unligned_store(dest);\
217 }
218 
219 static void h264_idct8_add_altivec(uint8_t *dst, int16_t *dct, int stride)
220 {
221  vec_s16 s0, s1, s2, s3, s4, s5, s6, s7;
222  vec_s16 d0, d1, d2, d3, d4, d5, d6, d7;
223  vec_s16 idct0, idct1, idct2, idct3, idct4, idct5, idct6, idct7;
224 
225  vec_u8 perm_ldv, perm_stv;
226  GET_2PERM(perm_ldv, perm_stv, dst);
227 
228  const vec_u16 onev = vec_splat_u16(1);
229  const vec_u16 twov = vec_splat_u16(2);
230  const vec_u16 sixv = vec_splat_u16(6);
231 
232  const vec_u8 sel = (vec_u8) {0,0,0,0,0,0,0,0,-1,-1,-1,-1,-1,-1,-1,-1};
233  LOAD_ZERO;
234 
235  dct[0] += 32; // rounding for the >>6 at the end
236 
237  s0 = vec_ld(0x00, (int16_t*)dct);
238  s1 = vec_ld(0x10, (int16_t*)dct);
239  s2 = vec_ld(0x20, (int16_t*)dct);
240  s3 = vec_ld(0x30, (int16_t*)dct);
241  s4 = vec_ld(0x40, (int16_t*)dct);
242  s5 = vec_ld(0x50, (int16_t*)dct);
243  s6 = vec_ld(0x60, (int16_t*)dct);
244  s7 = vec_ld(0x70, (int16_t*)dct);
245  memset(dct, 0, 64 * sizeof(int16_t));
246 
247  IDCT8_1D_ALTIVEC(s0, s1, s2, s3, s4, s5, s6, s7,
248  d0, d1, d2, d3, d4, d5, d6, d7);
249 
250  TRANSPOSE8( d0, d1, d2, d3, d4, d5, d6, d7 );
251 
252  IDCT8_1D_ALTIVEC(d0, d1, d2, d3, d4, d5, d6, d7,
253  idct0, idct1, idct2, idct3, idct4, idct5, idct6, idct7);
254 
255  ALTIVEC_STORE_SUM_CLIP(&dst[0*stride], idct0, perm_ldv, perm_stv, sel);
256  ALTIVEC_STORE_SUM_CLIP(&dst[1*stride], idct1, perm_ldv, perm_stv, sel);
257  ALTIVEC_STORE_SUM_CLIP(&dst[2*stride], idct2, perm_ldv, perm_stv, sel);
258  ALTIVEC_STORE_SUM_CLIP(&dst[3*stride], idct3, perm_ldv, perm_stv, sel);
259  ALTIVEC_STORE_SUM_CLIP(&dst[4*stride], idct4, perm_ldv, perm_stv, sel);
260  ALTIVEC_STORE_SUM_CLIP(&dst[5*stride], idct5, perm_ldv, perm_stv, sel);
261  ALTIVEC_STORE_SUM_CLIP(&dst[6*stride], idct6, perm_ldv, perm_stv, sel);
262  ALTIVEC_STORE_SUM_CLIP(&dst[7*stride], idct7, perm_ldv, perm_stv, sel);
263 }
264 
265 #if HAVE_BIGENDIAN
266 #define DST_LD vec_ld
267 #else
268 #define DST_LD vec_vsx_ld
269 #endif
270 static av_always_inline void h264_idct_dc_add_internal(uint8_t *dst, int16_t *block, int stride, int size)
271 {
272  vec_s16 dc16;
273  vec_u8 dcplus, dcminus, v0, v1, v2, v3, aligner;
274  vec_s32 v_dc32;
275  LOAD_ZERO;
276  DECLARE_ALIGNED(16, int, dc);
277  int i;
278 
279  dc = (block[0] + 32) >> 6;
280  block[0] = 0;
281  v_dc32 = vec_lde(0, &dc);
282  dc16 = VEC_SPLAT16((vec_s16)v_dc32, 1);
283 
284  if (size == 4)
285  dc16 = VEC_SLD16(dc16, zero_s16v, 8);
286  dcplus = vec_packsu(dc16, zero_s16v);
287  dcminus = vec_packsu(vec_sub(zero_s16v, dc16), zero_s16v);
288 
289 #if HAVE_BIGENDIAN
290  aligner = vec_lvsr(0, dst);
291  dcplus = vec_perm(dcplus, dcplus, aligner);
292  dcminus = vec_perm(dcminus, dcminus, aligner);
293 #endif
294 
295  for (i = 0; i < size; i += 4) {
296  v0 = DST_LD(0, dst+0*stride);
297  v1 = DST_LD(0, dst+1*stride);
298  v2 = DST_LD(0, dst+2*stride);
299  v3 = DST_LD(0, dst+3*stride);
300 
301  v0 = vec_adds(v0, dcplus);
302  v1 = vec_adds(v1, dcplus);
303  v2 = vec_adds(v2, dcplus);
304  v3 = vec_adds(v3, dcplus);
305 
306  v0 = vec_subs(v0, dcminus);
307  v1 = vec_subs(v1, dcminus);
308  v2 = vec_subs(v2, dcminus);
309  v3 = vec_subs(v3, dcminus);
310 
311  VEC_ST(v0, 0, dst+0*stride);
312  VEC_ST(v1, 0, dst+1*stride);
313  VEC_ST(v2, 0, dst+2*stride);
314  VEC_ST(v3, 0, dst+3*stride);
315 
316  dst += 4*stride;
317  }
318 }
319 
320 static void h264_idct_dc_add_altivec(uint8_t *dst, int16_t *block, int stride)
321 {
322  h264_idct_dc_add_internal(dst, block, stride, 4);
323 }
324 
325 static void h264_idct8_dc_add_altivec(uint8_t *dst, int16_t *block, int stride)
326 {
327  h264_idct_dc_add_internal(dst, block, stride, 8);
328 }
329 
330 static void h264_idct_add16_altivec(uint8_t *dst, const int *block_offset,
331  int16_t *block, int stride,
332  const uint8_t nnzc[15 * 8])
333 {
334  int i;
335  for(i=0; i<16; i++){
336  int nnz = nnzc[ scan8[i] ];
337  if(nnz){
338  if(nnz==1 && block[i*16]) h264_idct_dc_add_altivec(dst + block_offset[i], block + i*16, stride);
339  else h264_idct_add_altivec(dst + block_offset[i], block + i*16, stride);
340  }
341  }
342 }
343 
344 static void h264_idct_add16intra_altivec(uint8_t *dst, const int *block_offset,
345  int16_t *block, int stride,
346  const uint8_t nnzc[15 * 8])
347 {
348  int i;
349  for(i=0; i<16; i++){
350  if(nnzc[ scan8[i] ]) h264_idct_add_altivec(dst + block_offset[i], block + i*16, stride);
351  else if(block[i*16]) h264_idct_dc_add_altivec(dst + block_offset[i], block + i*16, stride);
352  }
353 }
354 
355 static void h264_idct8_add4_altivec(uint8_t *dst, const int *block_offset,
356  int16_t *block, int stride,
357  const uint8_t nnzc[15 * 8])
358 {
359  int i;
360  for(i=0; i<16; i+=4){
361  int nnz = nnzc[ scan8[i] ];
362  if(nnz){
363  if(nnz==1 && block[i*16]) h264_idct8_dc_add_altivec(dst + block_offset[i], block + i*16, stride);
364  else h264_idct8_add_altivec(dst + block_offset[i], block + i*16, stride);
365  }
366  }
367 }
368 
369 static void h264_idct_add8_altivec(uint8_t **dest, const int *block_offset,
370  int16_t *block, int stride,
371  const uint8_t nnzc[15 * 8])
372 {
373  int i, j;
374  for (j = 1; j < 3; j++) {
375  for(i = j * 16; i < j * 16 + 4; i++){
376  if(nnzc[ scan8[i] ])
377  h264_idct_add_altivec(dest[j-1] + block_offset[i], block + i*16, stride);
378  else if(block[i*16])
379  h264_idct_dc_add_altivec(dest[j-1] + block_offset[i], block + i*16, stride);
380  }
381  }
382 }
383 
384 #define transpose4x16(r0, r1, r2, r3) { \
385  register vec_u8 r4; \
386  register vec_u8 r5; \
387  register vec_u8 r6; \
388  register vec_u8 r7; \
389  \
390  r4 = vec_mergeh(r0, r2); /*0, 2 set 0*/ \
391  r5 = vec_mergel(r0, r2); /*0, 2 set 1*/ \
392  r6 = vec_mergeh(r1, r3); /*1, 3 set 0*/ \
393  r7 = vec_mergel(r1, r3); /*1, 3 set 1*/ \
394  \
395  r0 = vec_mergeh(r4, r6); /*all set 0*/ \
396  r1 = vec_mergel(r4, r6); /*all set 1*/ \
397  r2 = vec_mergeh(r5, r7); /*all set 2*/ \
398  r3 = vec_mergel(r5, r7); /*all set 3*/ \
399 }
400 
401 static inline void write16x4(uint8_t *dst, int dst_stride,
402  register vec_u8 r0, register vec_u8 r1,
403  register vec_u8 r2, register vec_u8 r3) {
404  DECLARE_ALIGNED(16, unsigned char, result)[64];
405  uint32_t *src_int = (uint32_t *)result, *dst_int = (uint32_t *)dst;
406  int int_dst_stride = dst_stride/4;
407 
408  vec_st(r0, 0, result);
409  vec_st(r1, 16, result);
410  vec_st(r2, 32, result);
411  vec_st(r3, 48, result);
412  /* FIXME: there has to be a better way!!!! */
413  *dst_int = *src_int;
414  *(dst_int+ int_dst_stride) = *(src_int + 1);
415  *(dst_int+ 2*int_dst_stride) = *(src_int + 2);
416  *(dst_int+ 3*int_dst_stride) = *(src_int + 3);
417  *(dst_int+ 4*int_dst_stride) = *(src_int + 4);
418  *(dst_int+ 5*int_dst_stride) = *(src_int + 5);
419  *(dst_int+ 6*int_dst_stride) = *(src_int + 6);
420  *(dst_int+ 7*int_dst_stride) = *(src_int + 7);
421  *(dst_int+ 8*int_dst_stride) = *(src_int + 8);
422  *(dst_int+ 9*int_dst_stride) = *(src_int + 9);
423  *(dst_int+10*int_dst_stride) = *(src_int + 10);
424  *(dst_int+11*int_dst_stride) = *(src_int + 11);
425  *(dst_int+12*int_dst_stride) = *(src_int + 12);
426  *(dst_int+13*int_dst_stride) = *(src_int + 13);
427  *(dst_int+14*int_dst_stride) = *(src_int + 14);
428  *(dst_int+15*int_dst_stride) = *(src_int + 15);
429 }
430 
431 /** @brief performs a 6x16 transpose of data in src, and stores it to dst
432  @todo FIXME: see if we can't spare some vec_lvsl() by them factorizing
433  out of unaligned_load() */
434 #define readAndTranspose16x6(src, src_stride, r8, r9, r10, r11, r12, r13) {\
435  register vec_u8 r0 = unaligned_load(0, src); \
436  register vec_u8 r1 = unaligned_load( src_stride, src); \
437  register vec_u8 r2 = unaligned_load(2* src_stride, src); \
438  register vec_u8 r3 = unaligned_load(3* src_stride, src); \
439  register vec_u8 r4 = unaligned_load(4* src_stride, src); \
440  register vec_u8 r5 = unaligned_load(5* src_stride, src); \
441  register vec_u8 r6 = unaligned_load(6* src_stride, src); \
442  register vec_u8 r7 = unaligned_load(7* src_stride, src); \
443  register vec_u8 r14 = unaligned_load(14*src_stride, src); \
444  register vec_u8 r15 = unaligned_load(15*src_stride, src); \
445  \
446  r8 = unaligned_load( 8*src_stride, src); \
447  r9 = unaligned_load( 9*src_stride, src); \
448  r10 = unaligned_load(10*src_stride, src); \
449  r11 = unaligned_load(11*src_stride, src); \
450  r12 = unaligned_load(12*src_stride, src); \
451  r13 = unaligned_load(13*src_stride, src); \
452  \
453  /*Merge first pairs*/ \
454  r0 = vec_mergeh(r0, r8); /*0, 8*/ \
455  r1 = vec_mergeh(r1, r9); /*1, 9*/ \
456  r2 = vec_mergeh(r2, r10); /*2,10*/ \
457  r3 = vec_mergeh(r3, r11); /*3,11*/ \
458  r4 = vec_mergeh(r4, r12); /*4,12*/ \
459  r5 = vec_mergeh(r5, r13); /*5,13*/ \
460  r6 = vec_mergeh(r6, r14); /*6,14*/ \
461  r7 = vec_mergeh(r7, r15); /*7,15*/ \
462  \
463  /*Merge second pairs*/ \
464  r8 = vec_mergeh(r0, r4); /*0,4, 8,12 set 0*/ \
465  r9 = vec_mergel(r0, r4); /*0,4, 8,12 set 1*/ \
466  r10 = vec_mergeh(r1, r5); /*1,5, 9,13 set 0*/ \
467  r11 = vec_mergel(r1, r5); /*1,5, 9,13 set 1*/ \
468  r12 = vec_mergeh(r2, r6); /*2,6,10,14 set 0*/ \
469  r13 = vec_mergel(r2, r6); /*2,6,10,14 set 1*/ \
470  r14 = vec_mergeh(r3, r7); /*3,7,11,15 set 0*/ \
471  r15 = vec_mergel(r3, r7); /*3,7,11,15 set 1*/ \
472  \
473  /*Third merge*/ \
474  r0 = vec_mergeh(r8, r12); /*0,2,4,6,8,10,12,14 set 0*/ \
475  r1 = vec_mergel(r8, r12); /*0,2,4,6,8,10,12,14 set 1*/ \
476  r2 = vec_mergeh(r9, r13); /*0,2,4,6,8,10,12,14 set 2*/ \
477  r4 = vec_mergeh(r10, r14); /*1,3,5,7,9,11,13,15 set 0*/ \
478  r5 = vec_mergel(r10, r14); /*1,3,5,7,9,11,13,15 set 1*/ \
479  r6 = vec_mergeh(r11, r15); /*1,3,5,7,9,11,13,15 set 2*/ \
480  /* Don't need to compute 3 and 7*/ \
481  \
482  /*Final merge*/ \
483  r8 = vec_mergeh(r0, r4); /*all set 0*/ \
484  r9 = vec_mergel(r0, r4); /*all set 1*/ \
485  r10 = vec_mergeh(r1, r5); /*all set 2*/ \
486  r11 = vec_mergel(r1, r5); /*all set 3*/ \
487  r12 = vec_mergeh(r2, r6); /*all set 4*/ \
488  r13 = vec_mergel(r2, r6); /*all set 5*/ \
489  /* Don't need to compute 14 and 15*/ \
490  \
491 }
492 
493 // out: o = |x-y| < a
494 static inline vec_u8 diff_lt_altivec ( register vec_u8 x,
495  register vec_u8 y,
496  register vec_u8 a) {
497 
498  register vec_u8 diff = vec_subs(x, y);
499  register vec_u8 diffneg = vec_subs(y, x);
500  register vec_u8 o = vec_or(diff, diffneg); /* |x-y| */
501  o = (vec_u8)vec_cmplt(o, a);
502  return o;
503 }
504 
505 static inline vec_u8 h264_deblock_mask ( register vec_u8 p0,
506  register vec_u8 p1,
507  register vec_u8 q0,
508  register vec_u8 q1,
509  register vec_u8 alpha,
510  register vec_u8 beta) {
511 
512  register vec_u8 mask;
513  register vec_u8 tempmask;
514 
515  mask = diff_lt_altivec(p0, q0, alpha);
516  tempmask = diff_lt_altivec(p1, p0, beta);
517  mask = vec_and(mask, tempmask);
518  tempmask = diff_lt_altivec(q1, q0, beta);
519  mask = vec_and(mask, tempmask);
520 
521  return mask;
522 }
523 
524 // out: newp1 = clip((p2 + ((p0 + q0 + 1) >> 1)) >> 1, p1-tc0, p1+tc0)
525 static inline vec_u8 h264_deblock_q1(register vec_u8 p0,
526  register vec_u8 p1,
527  register vec_u8 p2,
528  register vec_u8 q0,
529  register vec_u8 tc0) {
530 
531  register vec_u8 average = vec_avg(p0, q0);
532  register vec_u8 temp;
533  register vec_u8 unclipped;
534  register vec_u8 ones;
535  register vec_u8 max;
536  register vec_u8 min;
537  register vec_u8 newp1;
538 
539  temp = vec_xor(average, p2);
540  average = vec_avg(average, p2); /*avg(p2, avg(p0, q0)) */
541  ones = vec_splat_u8(1);
542  temp = vec_and(temp, ones); /*(p2^avg(p0, q0)) & 1 */
543  unclipped = vec_subs(average, temp); /*(p2+((p0+q0+1)>>1))>>1 */
544  max = vec_adds(p1, tc0);
545  min = vec_subs(p1, tc0);
546  newp1 = vec_max(min, unclipped);
547  newp1 = vec_min(max, newp1);
548  return newp1;
549 }
550 
551 #define h264_deblock_p0_q0(p0, p1, q0, q1, tc0masked) { \
552  \
553  const vec_u8 A0v = vec_sl(vec_splat_u8(10), vec_splat_u8(4)); \
554  \
555  register vec_u8 pq0bit = vec_xor(p0,q0); \
556  register vec_u8 q1minus; \
557  register vec_u8 p0minus; \
558  register vec_u8 stage1; \
559  register vec_u8 stage2; \
560  register vec_u8 vec160; \
561  register vec_u8 delta; \
562  register vec_u8 deltaneg; \
563  \
564  q1minus = vec_nor(q1, q1); /* 255 - q1 */ \
565  stage1 = vec_avg(p1, q1minus); /* (p1 - q1 + 256)>>1 */ \
566  stage2 = vec_sr(stage1, vec_splat_u8(1)); /* (p1 - q1 + 256)>>2 = 64 + (p1 - q1) >> 2 */ \
567  p0minus = vec_nor(p0, p0); /* 255 - p0 */ \
568  stage1 = vec_avg(q0, p0minus); /* (q0 - p0 + 256)>>1 */ \
569  pq0bit = vec_and(pq0bit, vec_splat_u8(1)); \
570  stage2 = vec_avg(stage2, pq0bit); /* 32 + ((q0 - p0)&1 + (p1 - q1) >> 2 + 1) >> 1 */ \
571  stage2 = vec_adds(stage2, stage1); /* 160 + ((p0 - q0) + (p1 - q1) >> 2 + 1) >> 1 */ \
572  vec160 = vec_ld(0, &A0v); \
573  deltaneg = vec_subs(vec160, stage2); /* -d */ \
574  delta = vec_subs(stage2, vec160); /* d */ \
575  deltaneg = vec_min(tc0masked, deltaneg); \
576  delta = vec_min(tc0masked, delta); \
577  p0 = vec_subs(p0, deltaneg); \
578  q0 = vec_subs(q0, delta); \
579  p0 = vec_adds(p0, delta); \
580  q0 = vec_adds(q0, deltaneg); \
581 }
582 
583 #define h264_loop_filter_luma_altivec(p2, p1, p0, q0, q1, q2, alpha, beta, tc0) { \
584  DECLARE_ALIGNED(16, unsigned char, temp)[16]; \
585  register vec_u8 alphavec; \
586  register vec_u8 betavec; \
587  register vec_u8 mask; \
588  register vec_u8 p1mask; \
589  register vec_u8 q1mask; \
590  register vector signed char tc0vec; \
591  register vec_u8 finaltc0; \
592  register vec_u8 tc0masked; \
593  register vec_u8 newp1; \
594  register vec_u8 newq1; \
595  \
596  temp[0] = alpha; \
597  temp[1] = beta; \
598  alphavec = vec_ld(0, temp); \
599  betavec = vec_splat(alphavec, 0x1); \
600  alphavec = vec_splat(alphavec, 0x0); \
601  mask = h264_deblock_mask(p0, p1, q0, q1, alphavec, betavec); /*if in block */ \
602  \
603  AV_COPY32(temp, tc0); \
604  tc0vec = vec_ld(0, (signed char*)temp); \
605  tc0vec = vec_mergeh(tc0vec, tc0vec); \
606  tc0vec = vec_mergeh(tc0vec, tc0vec); \
607  mask = vec_and(mask, vec_cmpgt(tc0vec, vec_splat_s8(-1))); /* if tc0[i] >= 0 */ \
608  finaltc0 = vec_and((vec_u8)tc0vec, mask); /* tc = tc0 */ \
609  \
610  p1mask = diff_lt_altivec(p2, p0, betavec); \
611  p1mask = vec_and(p1mask, mask); /* if ( |p2 - p0| < beta) */ \
612  tc0masked = vec_and(p1mask, (vec_u8)tc0vec); \
613  finaltc0 = vec_sub(finaltc0, p1mask); /* tc++ */ \
614  newp1 = h264_deblock_q1(p0, p1, p2, q0, tc0masked); \
615  /*end if*/ \
616  \
617  q1mask = diff_lt_altivec(q2, q0, betavec); \
618  q1mask = vec_and(q1mask, mask); /* if ( |q2 - q0| < beta ) */\
619  tc0masked = vec_and(q1mask, (vec_u8)tc0vec); \
620  finaltc0 = vec_sub(finaltc0, q1mask); /* tc++ */ \
621  newq1 = h264_deblock_q1(p0, q1, q2, q0, tc0masked); \
622  /*end if*/ \
623  \
624  h264_deblock_p0_q0(p0, p1, q0, q1, finaltc0); \
625  p1 = newp1; \
626  q1 = newq1; \
627 }
628 
629 static void h264_v_loop_filter_luma_altivec(uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0) {
630 
631  if ((tc0[0] & tc0[1] & tc0[2] & tc0[3]) >= 0) {
632  register vec_u8 p2 = vec_ld(-3*stride, pix);
633  register vec_u8 p1 = vec_ld(-2*stride, pix);
634  register vec_u8 p0 = vec_ld(-1*stride, pix);
635  register vec_u8 q0 = vec_ld(0, pix);
636  register vec_u8 q1 = vec_ld(stride, pix);
637  register vec_u8 q2 = vec_ld(2*stride, pix);
638  h264_loop_filter_luma_altivec(p2, p1, p0, q0, q1, q2, alpha, beta, tc0);
639  vec_st(p1, -2*stride, pix);
640  vec_st(p0, -1*stride, pix);
641  vec_st(q0, 0, pix);
642  vec_st(q1, stride, pix);
643  }
644 }
645 
646 static void h264_h_loop_filter_luma_altivec(uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0) {
647 
648  register vec_u8 line0, line1, line2, line3, line4, line5;
649  if ((tc0[0] & tc0[1] & tc0[2] & tc0[3]) < 0)
650  return;
651  readAndTranspose16x6(pix-3, stride, line0, line1, line2, line3, line4, line5);
652  h264_loop_filter_luma_altivec(line0, line1, line2, line3, line4, line5, alpha, beta, tc0);
653  transpose4x16(line1, line2, line3, line4);
654  write16x4(pix-2, stride, line1, line2, line3, line4);
655 }
656 
657 static av_always_inline
658 void weight_h264_W_altivec(uint8_t *block, int stride, int height,
659  int log2_denom, int weight, int offset, int w)
660 {
661  int y, aligned;
662  vec_u8 vblock;
663  vec_s16 vtemp, vweight, voffset, v0, v1;
664  vec_u16 vlog2_denom;
665  DECLARE_ALIGNED(16, int32_t, temp)[4];
666  LOAD_ZERO;
667 
668  offset <<= log2_denom;
669  if(log2_denom) offset += 1<<(log2_denom-1);
670  temp[0] = log2_denom;
671  temp[1] = weight;
672  temp[2] = offset;
673 
674  vtemp = (vec_s16)vec_ld(0, temp);
675 #if !HAVE_BIGENDIAN
676  vtemp =(vec_s16)vec_perm(vtemp, vtemp, vcswapi2s(0,1,2,3));
677 #endif
678  vlog2_denom = (vec_u16)vec_splat(vtemp, 1);
679  vweight = vec_splat(vtemp, 3);
680  voffset = vec_splat(vtemp, 5);
681  aligned = !((unsigned long)block & 0xf);
682 
683  for (y = 0; y < height; y++) {
684  vblock = vec_ld(0, block);
685 
686  v0 = (vec_s16)VEC_MERGEH(zero_u8v, vblock);
687  v1 = (vec_s16)VEC_MERGEL(zero_u8v, vblock);
688 
689  if (w == 16 || aligned) {
690  v0 = vec_mladd(v0, vweight, zero_s16v);
691  v0 = vec_adds(v0, voffset);
692  v0 = vec_sra(v0, vlog2_denom);
693  }
694  if (w == 16 || !aligned) {
695  v1 = vec_mladd(v1, vweight, zero_s16v);
696  v1 = vec_adds(v1, voffset);
697  v1 = vec_sra(v1, vlog2_denom);
698  }
699  vblock = vec_packsu(v0, v1);
700  vec_st(vblock, 0, block);
701 
702  block += stride;
703  }
704 }
705 
706 static av_always_inline
707 void biweight_h264_W_altivec(uint8_t *dst, uint8_t *src, int stride, int height,
708  int log2_denom, int weightd, int weights, int offset, int w)
709 {
710  int y, dst_aligned, src_aligned;
711  vec_u8 vsrc, vdst;
712  vec_s16 vtemp, vweights, vweightd, voffset, v0, v1, v2, v3;
713  vec_u16 vlog2_denom;
714  DECLARE_ALIGNED(16, int32_t, temp)[4];
715  LOAD_ZERO;
716 
717  offset = ((offset + 1) | 1) << log2_denom;
718  temp[0] = log2_denom+1;
719  temp[1] = weights;
720  temp[2] = weightd;
721  temp[3] = offset;
722 
723  vtemp = (vec_s16)vec_ld(0, temp);
724 #if !HAVE_BIGENDIAN
725  vtemp =(vec_s16)vec_perm(vtemp, vtemp, vcswapi2s(0,1,2,3));
726 #endif
727  vlog2_denom = (vec_u16)vec_splat(vtemp, 1);
728  vweights = vec_splat(vtemp, 3);
729  vweightd = vec_splat(vtemp, 5);
730  voffset = vec_splat(vtemp, 7);
731  dst_aligned = !((unsigned long)dst & 0xf);
732  src_aligned = !((unsigned long)src & 0xf);
733 
734  for (y = 0; y < height; y++) {
735  vdst = vec_ld(0, dst);
736  vsrc = vec_ld(0, src);
737 
738  v0 = (vec_s16)VEC_MERGEH(zero_u8v, vdst);
739  v1 = (vec_s16)VEC_MERGEL(zero_u8v, vdst);
740  v2 = (vec_s16)VEC_MERGEH(zero_u8v, vsrc);
741  v3 = (vec_s16)VEC_MERGEL(zero_u8v, vsrc);
742 
743  if (w == 8) {
744  if (src_aligned)
745  v3 = v2;
746  else
747  v2 = v3;
748  }
749 
750  if (w == 16 || dst_aligned) {
751  v0 = vec_mladd(v0, vweightd, zero_s16v);
752  v2 = vec_mladd(v2, vweights, zero_s16v);
753 
754  v0 = vec_adds(v0, voffset);
755  v0 = vec_adds(v0, v2);
756  v0 = vec_sra(v0, vlog2_denom);
757  }
758  if (w == 16 || !dst_aligned) {
759  v1 = vec_mladd(v1, vweightd, zero_s16v);
760  v3 = vec_mladd(v3, vweights, zero_s16v);
761 
762  v1 = vec_adds(v1, voffset);
763  v1 = vec_adds(v1, v3);
764  v1 = vec_sra(v1, vlog2_denom);
765  }
766  vdst = vec_packsu(v0, v1);
767  vec_st(vdst, 0, dst);
768 
769  dst += stride;
770  src += stride;
771  }
772 }
773 
774 #define H264_WEIGHT(W) \
775 static void weight_h264_pixels ## W ## _altivec(uint8_t *block, int stride, int height, \
776  int log2_denom, int weight, int offset) \
777 { \
778  weight_h264_W_altivec(block, stride, height, log2_denom, weight, offset, W); \
779 }\
780 static void biweight_h264_pixels ## W ## _altivec(uint8_t *dst, uint8_t *src, int stride, int height, \
781  int log2_denom, int weightd, int weights, int offset) \
782 { \
783  biweight_h264_W_altivec(dst, src, stride, height, log2_denom, weightd, weights, offset, W); \
784 }
785 
786 H264_WEIGHT(16)
787 H264_WEIGHT( 8)
788 #endif /* HAVE_ALTIVEC */
789 
791  const int chroma_format_idc)
792 {
793 #if HAVE_ALTIVEC
795  return;
796 
797  if (bit_depth == 8) {
798  c->h264_idct_add = h264_idct_add_altivec;
799  if (chroma_format_idc <= 1)
800  c->h264_idct_add8 = h264_idct_add8_altivec;
801  c->h264_idct_add16 = h264_idct_add16_altivec;
802  c->h264_idct_add16intra = h264_idct_add16intra_altivec;
803  c->h264_idct_dc_add= h264_idct_dc_add_altivec;
804  c->h264_idct8_dc_add = h264_idct8_dc_add_altivec;
805  c->h264_idct8_add = h264_idct8_add_altivec;
806  c->h264_idct8_add4 = h264_idct8_add4_altivec;
807  c->h264_v_loop_filter_luma= h264_v_loop_filter_luma_altivec;
808  c->h264_h_loop_filter_luma= h264_h_loop_filter_luma_altivec;
809 
810  c->weight_h264_pixels_tab[0] = weight_h264_pixels16_altivec;
811  c->weight_h264_pixels_tab[1] = weight_h264_pixels8_altivec;
812  c->biweight_h264_pixels_tab[0] = biweight_h264_pixels16_altivec;
813  c->biweight_h264_pixels_tab[1] = biweight_h264_pixels8_altivec;
814  }
815 #endif /* HAVE_ALTIVEC */
816 }
void(* h264_idct_add)(uint8_t *dst, int16_t *block, int stride)
Definition: h264dsp.h:81
Memory handling functions.
else temp
Definition: vf_mcdeint.c:259
#define vec_s32
Definition: types_altivec.h:32
GLfloat v0
Definition: opengl_enc.c:107
static const uint8_t q1[256]
Definition: twofish.c:96
#define src
Definition: vp8dsp.c:254
H.264 DSP functions.
Macro definitions for various function/variable attributes.
static int16_t block[64]
Definition: dct.c:115
void(* h264_idct_add16)(uint8_t *dst, const int *blockoffset, int16_t *block, int stride, const uint8_t nnzc[15 *8])
Definition: h264dsp.h:90
uint8_t
#define av_cold
Definition: attributes.h:82
void(* h264_idct_add8)(uint8_t **dst, const int *blockoffset, int16_t *block, int stride, const uint8_t nnzc[15 *8])
Definition: h264dsp.h:96
h264_weight_func weight_h264_pixels_tab[4]
Definition: h264dsp.h:44
#define DECLARE_ALIGNED(n, t, v)
Declare a variable that is aligned in memory.
Definition: mem.h:104
#define height
#define vec_s16
Definition: types_altivec.h:30
#define zero_s16v
Definition: types_altivec.h:43
ptrdiff_t size
Definition: opengl_enc.c:101
static double alpha(void *priv, double x, double y)
Definition: vf_geq.c:99
#define s2
Definition: regdef.h:39
static const uint16_t mask[17]
Definition: lzw.c:38
#define PPC_ALTIVEC(flags)
Definition: cpu.h:26
void(* h264_idct8_add)(uint8_t *dst, int16_t *block, int stride)
Definition: h264dsp.h:83
#define s0
Definition: regdef.h:37
#define vec_u16
Definition: types_altivec.h:29
#define H264_WEIGHT(W)
#define LOAD_ZERO
Definition: types_altivec.h:38
static const uint8_t offset[127][2]
Definition: vf_spp.c:92
static const uint8_t q0[256]
Definition: twofish.c:77
#define s5
Definition: regdef.h:42
static void idct6(int pre_mant[6])
Calculate 6-point IDCT of the pre-mantissas.
Definition: eac3dec.c:173
h264_biweight_func biweight_h264_pixels_tab[4]
Definition: h264dsp.h:45
Context for storing H.264 DSP functions.
Definition: h264dsp.h:42
static void bit_depth(AudioStatsContext *s, uint64_t mask, uint64_t imask, AVRational *depth)
Definition: af_astats.c:150
void(* h264_idct_dc_add)(uint8_t *dst, int16_t *block, int stride)
Definition: h264dsp.h:85
int32_t
#define vec_u32
Definition: types_altivec.h:31
H.264 / AVC / MPEG-4 part10 codec.
av_cold void ff_h264dsp_init_ppc(H264DSPContext *c, const int bit_depth, const int chroma_format_idc)
Definition: h264dsp.c:790
#define s4
Definition: regdef.h:41
#define s3
Definition: regdef.h:40
#define vec_u8
Definition: types_altivec.h:27
void(* h264_v_loop_filter_luma)(uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0)
Definition: h264dsp.h:48
static const uint8_t scan8[16 *3+3]
Definition: h264dec.h:638
void(* h264_idct8_add4)(uint8_t *dst, const int *blockoffset, int16_t *block, int stride, const uint8_t nnzc[15 *8])
Definition: h264dsp.h:93
#define s1
Definition: regdef.h:38
int av_get_cpu_flags(void)
Return the flags which specify extensions supported by the CPU.
Definition: cpu.c:89
static int weight(int i, int blen, int offset)
Definition: diracdec.c:1523
void(* h264_idct8_dc_add)(uint8_t *dst, int16_t *block, int stride)
Definition: h264dsp.h:87
Contains misc utility macros and inline functions.
GLint GLenum GLboolean GLsizei stride
Definition: opengl_enc.c:105
static double c[64]
static av_always_inline int diff(const uint32_t a, const uint32_t b)
void(* h264_idct_add16intra)(uint8_t *dst, const int *blockoffset, int16_t *block, int stride, const uint8_t nnzc[15 *8])
Definition: h264dsp.h:99
void(* h264_h_loop_filter_luma)(uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0)
Definition: h264dsp.h:50
uint8_t pi<< 24) CONV_FUNC_GROUP(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_U8, uint8_t,(*(constuint8_t *) pi-0x80)*(1.0f/(1<< 7))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_U8, uint8_t,(*(constuint8_t *) pi-0x80)*(1.0/(1<< 7))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_S16, int16_t,(*(constint16_t *) pi >>8)+0x80) CONV_FUNC_GROUP(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_S16, int16_t,*(constint16_t *) pi *(1.0f/(1<< 15))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_S16, int16_t,*(constint16_t *) pi *(1.0/(1<< 15))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_S32, int32_t,(*(constint32_t *) pi >>24)+0x80) CONV_FUNC_GROUP(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_S32, int32_t,*(constint32_t *) pi *(1.0f/(1U<< 31))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_S32, int32_t,*(constint32_t *) pi *(1.0/(1U<< 31))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_FLT, float, av_clip_uint8(lrintf(*(constfloat *) pi *(1<< 7))+0x80)) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S16, int16_t, AV_SAMPLE_FMT_FLT, float, av_clip_int16(lrintf(*(constfloat *) pi *(1<< 15)))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S32, int32_t, AV_SAMPLE_FMT_FLT, float, av_clipl_int32(llrintf(*(constfloat *) pi *(1U<< 31)))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_DBL, double, av_clip_uint8(lrint(*(constdouble *) pi *(1<< 7))+0x80)) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S16, int16_t, AV_SAMPLE_FMT_DBL, double, av_clip_int16(lrint(*(constdouble *) pi *(1<< 15)))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S32, int32_t, AV_SAMPLE_FMT_DBL, double, av_clipl_int32(llrint(*(constdouble *) pi *(1U<< 31))))#defineSET_CONV_FUNC_GROUP(ofmt, ifmt) staticvoidset_generic_function(AudioConvert *ac){}voidff_audio_convert_free(AudioConvert **ac){if(!*ac) return;ff_dither_free(&(*ac) ->dc);av_freep(ac);}AudioConvert *ff_audio_convert_alloc(AVAudioResampleContext *avr, enumAVSampleFormatout_fmt, enumAVSampleFormatin_fmt, intchannels, intsample_rate, intapply_map){AudioConvert *ac;intin_planar, out_planar;ac=av_mallocz(sizeof(*ac));if(!ac) returnNULL;ac->avr=avr;ac->out_fmt=out_fmt;ac->in_fmt=in_fmt;ac->channels=channels;ac->apply_map=apply_map;if(avr->dither_method!=AV_RESAMPLE_DITHER_NONE &&av_get_packed_sample_fmt(out_fmt)==AV_SAMPLE_FMT_S16 &&av_get_bytes_per_sample(in_fmt)>2){ac->dc=ff_dither_alloc(avr, out_fmt, in_fmt, channels, sample_rate, apply_map);if(!ac->dc){av_free(ac);returnNULL;}returnac;}in_planar=ff_sample_fmt_is_planar(in_fmt, channels);out_planar=ff_sample_fmt_is_planar(out_fmt, channels);if(in_planar==out_planar){ac->func_type=CONV_FUNC_TYPE_FLAT;ac->planes=in_planar?ac->channels:1;}elseif(in_planar) ac->func_type=CONV_FUNC_TYPE_INTERLEAVE;elseac->func_type=CONV_FUNC_TYPE_DEINTERLEAVE;set_generic_function(ac);if(ARCH_AARCH64) ff_audio_convert_init_aarch64(ac);if(ARCH_ARM) ff_audio_convert_init_arm(ac);if(ARCH_X86) ff_audio_convert_init_x86(ac);returnac;}intff_audio_convert(AudioConvert *ac, AudioData *out, AudioData *in){intuse_generic=1;intlen=in->nb_samples;intp;if(ac->dc){av_log(ac->avr, AV_LOG_TRACE,"%dsamples-audio_convert:%sto%s(dithered)\n", len, av_get_sample_fmt_name(ac->in_fmt), av_get_sample_fmt_name(ac->out_fmt));returnff_convert_dither(ac-> dc
#define s6
Definition: regdef.h:43
#define av_always_inline
Definition: attributes.h:39
#define stride
float min
#define zero_u8v
Definition: types_altivec.h:40