FFmpeg  4.4.6
h264qpel.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 "libavutil/attributes.h"
24 #include "libavutil/cpu.h"
25 #include "libavutil/intreadwrite.h"
26 #include "libavutil/mem_internal.h"
27 #include "libavutil/ppc/cpu.h"
29 
30 #include "libavcodec/h264qpel.h"
31 
32 #include "hpeldsp_altivec.h"
33 
34 #if HAVE_ALTIVEC
35 
36 #define PUT_OP_U8_ALTIVEC(d, s, dst) d = s
37 #define AVG_OP_U8_ALTIVEC(d, s, dst) d = vec_avg(dst, s)
38 
39 #define OP_U8_ALTIVEC PUT_OP_U8_ALTIVEC
40 #define PREFIX_h264_qpel16_h_lowpass_altivec put_h264_qpel16_h_lowpass_altivec
41 #define PREFIX_h264_qpel16_h_lowpass_num altivec_put_h264_qpel16_h_lowpass_num
42 #define PREFIX_h264_qpel16_v_lowpass_altivec put_h264_qpel16_v_lowpass_altivec
43 #define PREFIX_h264_qpel16_v_lowpass_num altivec_put_h264_qpel16_v_lowpass_num
44 #define PREFIX_h264_qpel16_hv_lowpass_altivec put_h264_qpel16_hv_lowpass_altivec
45 #define PREFIX_h264_qpel16_hv_lowpass_num altivec_put_h264_qpel16_hv_lowpass_num
46 #include "h264qpel_template.c"
47 #undef OP_U8_ALTIVEC
48 #undef PREFIX_h264_qpel16_h_lowpass_altivec
49 #undef PREFIX_h264_qpel16_h_lowpass_num
50 #undef PREFIX_h264_qpel16_v_lowpass_altivec
51 #undef PREFIX_h264_qpel16_v_lowpass_num
52 #undef PREFIX_h264_qpel16_hv_lowpass_altivec
53 #undef PREFIX_h264_qpel16_hv_lowpass_num
54 
55 #define OP_U8_ALTIVEC AVG_OP_U8_ALTIVEC
56 #define PREFIX_h264_qpel16_h_lowpass_altivec avg_h264_qpel16_h_lowpass_altivec
57 #define PREFIX_h264_qpel16_h_lowpass_num altivec_avg_h264_qpel16_h_lowpass_num
58 #define PREFIX_h264_qpel16_v_lowpass_altivec avg_h264_qpel16_v_lowpass_altivec
59 #define PREFIX_h264_qpel16_v_lowpass_num altivec_avg_h264_qpel16_v_lowpass_num
60 #define PREFIX_h264_qpel16_hv_lowpass_altivec avg_h264_qpel16_hv_lowpass_altivec
61 #define PREFIX_h264_qpel16_hv_lowpass_num altivec_avg_h264_qpel16_hv_lowpass_num
62 #include "h264qpel_template.c"
63 #undef OP_U8_ALTIVEC
64 #undef PREFIX_h264_qpel16_h_lowpass_altivec
65 #undef PREFIX_h264_qpel16_h_lowpass_num
66 #undef PREFIX_h264_qpel16_v_lowpass_altivec
67 #undef PREFIX_h264_qpel16_v_lowpass_num
68 #undef PREFIX_h264_qpel16_hv_lowpass_altivec
69 #undef PREFIX_h264_qpel16_hv_lowpass_num
70 
71 #define H264_MC(OPNAME, SIZE, CODETYPE) \
72 static void OPNAME ## h264_qpel ## SIZE ## _mc00_ ## CODETYPE (uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
73 {\
74  ff_ ## OPNAME ## pixels ## SIZE ## _ ## CODETYPE(dst, src, stride, SIZE);\
75 }\
76 \
77 static void OPNAME ## h264_qpel ## SIZE ## _mc10_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
78 { \
79  DECLARE_ALIGNED(16, uint8_t, half)[SIZE*SIZE];\
80  put_h264_qpel ## SIZE ## _h_lowpass_ ## CODETYPE(half, src, SIZE, stride);\
81  OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, src, half, stride, stride, SIZE);\
82 }\
83 \
84 static void OPNAME ## h264_qpel ## SIZE ## _mc20_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
85 {\
86  OPNAME ## h264_qpel ## SIZE ## _h_lowpass_ ## CODETYPE(dst, src, stride, stride);\
87 }\
88 \
89 static void OPNAME ## h264_qpel ## SIZE ## _mc30_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
90 {\
91  DECLARE_ALIGNED(16, uint8_t, half)[SIZE*SIZE];\
92  put_h264_qpel ## SIZE ## _h_lowpass_ ## CODETYPE(half, src, SIZE, stride);\
93  OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, src+1, half, stride, stride, SIZE);\
94 }\
95 \
96 static void OPNAME ## h264_qpel ## SIZE ## _mc01_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
97 {\
98  DECLARE_ALIGNED(16, uint8_t, half)[SIZE*SIZE];\
99  put_h264_qpel ## SIZE ## _v_lowpass_ ## CODETYPE(half, src, SIZE, stride);\
100  OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, src, half, stride, stride, SIZE);\
101 }\
102 \
103 static void OPNAME ## h264_qpel ## SIZE ## _mc02_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
104 {\
105  OPNAME ## h264_qpel ## SIZE ## _v_lowpass_ ## CODETYPE(dst, src, stride, stride);\
106 }\
107 \
108 static void OPNAME ## h264_qpel ## SIZE ## _mc03_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
109 {\
110  DECLARE_ALIGNED(16, uint8_t, half)[SIZE*SIZE];\
111  put_h264_qpel ## SIZE ## _v_lowpass_ ## CODETYPE(half, src, SIZE, stride);\
112  OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, src+stride, half, stride, stride, SIZE);\
113 }\
114 \
115 static void OPNAME ## h264_qpel ## SIZE ## _mc11_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
116 {\
117  DECLARE_ALIGNED(16, uint8_t, halfH)[SIZE*SIZE];\
118  DECLARE_ALIGNED(16, uint8_t, halfV)[SIZE*SIZE];\
119  put_h264_qpel ## SIZE ## _h_lowpass_ ## CODETYPE(halfH, src, SIZE, stride);\
120  put_h264_qpel ## SIZE ## _v_lowpass_ ## CODETYPE(halfV, src, SIZE, stride);\
121  OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, halfH, halfV, stride, SIZE, SIZE);\
122 }\
123 \
124 static void OPNAME ## h264_qpel ## SIZE ## _mc31_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
125 {\
126  DECLARE_ALIGNED(16, uint8_t, halfH)[SIZE*SIZE];\
127  DECLARE_ALIGNED(16, uint8_t, halfV)[SIZE*SIZE];\
128  put_h264_qpel ## SIZE ## _h_lowpass_ ## CODETYPE(halfH, src, SIZE, stride);\
129  put_h264_qpel ## SIZE ## _v_lowpass_ ## CODETYPE(halfV, src+1, SIZE, stride);\
130  OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, halfH, halfV, stride, SIZE, SIZE);\
131 }\
132 \
133 static void OPNAME ## h264_qpel ## SIZE ## _mc13_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
134 {\
135  DECLARE_ALIGNED(16, uint8_t, halfH)[SIZE*SIZE];\
136  DECLARE_ALIGNED(16, uint8_t, halfV)[SIZE*SIZE];\
137  put_h264_qpel ## SIZE ## _h_lowpass_ ## CODETYPE(halfH, src + stride, SIZE, stride);\
138  put_h264_qpel ## SIZE ## _v_lowpass_ ## CODETYPE(halfV, src, SIZE, stride);\
139  OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, halfH, halfV, stride, SIZE, SIZE);\
140 }\
141 \
142 static void OPNAME ## h264_qpel ## SIZE ## _mc33_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
143 {\
144  DECLARE_ALIGNED(16, uint8_t, halfH)[SIZE*SIZE];\
145  DECLARE_ALIGNED(16, uint8_t, halfV)[SIZE*SIZE];\
146  put_h264_qpel ## SIZE ## _h_lowpass_ ## CODETYPE(halfH, src + stride, SIZE, stride);\
147  put_h264_qpel ## SIZE ## _v_lowpass_ ## CODETYPE(halfV, src+1, SIZE, stride);\
148  OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, halfH, halfV, stride, SIZE, SIZE);\
149 }\
150 \
151 static void OPNAME ## h264_qpel ## SIZE ## _mc22_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
152 {\
153  DECLARE_ALIGNED(16, int16_t, tmp)[SIZE*(SIZE+8)];\
154  OPNAME ## h264_qpel ## SIZE ## _hv_lowpass_ ## CODETYPE(dst, tmp, src, stride, SIZE, stride);\
155 }\
156 \
157 static void OPNAME ## h264_qpel ## SIZE ## _mc21_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
158 {\
159  DECLARE_ALIGNED(16, uint8_t, halfH)[SIZE*SIZE];\
160  DECLARE_ALIGNED(16, uint8_t, halfHV)[SIZE*SIZE];\
161  DECLARE_ALIGNED(16, int16_t, tmp)[SIZE*(SIZE+8)];\
162  put_h264_qpel ## SIZE ## _h_lowpass_ ## CODETYPE(halfH, src, SIZE, stride);\
163  put_h264_qpel ## SIZE ## _hv_lowpass_ ## CODETYPE(halfHV, tmp, src, SIZE, SIZE, stride);\
164  OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, halfH, halfHV, stride, SIZE, SIZE);\
165 }\
166 \
167 static void OPNAME ## h264_qpel ## SIZE ## _mc23_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
168 {\
169  DECLARE_ALIGNED(16, uint8_t, halfH)[SIZE*SIZE];\
170  DECLARE_ALIGNED(16, uint8_t, halfHV)[SIZE*SIZE];\
171  DECLARE_ALIGNED(16, int16_t, tmp)[SIZE*(SIZE+8)];\
172  put_h264_qpel ## SIZE ## _h_lowpass_ ## CODETYPE(halfH, src + stride, SIZE, stride);\
173  put_h264_qpel ## SIZE ## _hv_lowpass_ ## CODETYPE(halfHV, tmp, src, SIZE, SIZE, stride);\
174  OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, halfH, halfHV, stride, SIZE, SIZE);\
175 }\
176 \
177 static void OPNAME ## h264_qpel ## SIZE ## _mc12_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
178 {\
179  DECLARE_ALIGNED(16, uint8_t, halfV)[SIZE*SIZE];\
180  DECLARE_ALIGNED(16, uint8_t, halfHV)[SIZE*SIZE];\
181  DECLARE_ALIGNED(16, int16_t, tmp)[SIZE*(SIZE+8)];\
182  put_h264_qpel ## SIZE ## _v_lowpass_ ## CODETYPE(halfV, src, SIZE, stride);\
183  put_h264_qpel ## SIZE ## _hv_lowpass_ ## CODETYPE(halfHV, tmp, src, SIZE, SIZE, stride);\
184  OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, halfV, halfHV, stride, SIZE, SIZE);\
185 }\
186 \
187 static void OPNAME ## h264_qpel ## SIZE ## _mc32_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
188 {\
189  DECLARE_ALIGNED(16, uint8_t, halfV)[SIZE*SIZE];\
190  DECLARE_ALIGNED(16, uint8_t, halfHV)[SIZE*SIZE];\
191  DECLARE_ALIGNED(16, int16_t, tmp)[SIZE*(SIZE+8)];\
192  put_h264_qpel ## SIZE ## _v_lowpass_ ## CODETYPE(halfV, src+1, SIZE, stride);\
193  put_h264_qpel ## SIZE ## _hv_lowpass_ ## CODETYPE(halfHV, tmp, src, SIZE, SIZE, stride);\
194  OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, halfV, halfHV, stride, SIZE, SIZE);\
195 }\
196 
197 #if HAVE_BIGENDIAN
198 #define put_unligned_store(s, dest) { \
199  tmp1 = vec_ld(0, dest); \
200  mask = vec_lvsl(0, dest); \
201  tmp2 = vec_ld(15, dest); \
202  edges = vec_perm(tmp2, tmp1, mask); \
203  align = vec_lvsr(0, dest); \
204  tmp2 = vec_perm(s, edges, align); \
205  tmp1 = vec_perm(edges, s, align); \
206  vec_st(tmp2, 15, dest); \
207  vec_st(tmp1, 0 , dest); \
208  }
209 #else
210 #define put_unligned_store(s, dest) vec_vsx_st(s, 0, dest);
211 #endif /* HAVE_BIGENDIAN */
212 
213 static inline void put_pixels16_l2_altivec( uint8_t * dst, const uint8_t * src1,
214  const uint8_t * src2, int dst_stride,
215  int src_stride1, int h)
216 {
217  int i;
218  vec_u8 a, b, d, mask_;
219 #if HAVE_BIGENDIAN
220  vec_u8 tmp1, tmp2, mask, edges, align;
221  mask_ = vec_lvsl(0, src2);
222 #endif
223 
224  for (i = 0; i < h; i++) {
225  a = unaligned_load(i * src_stride1, src1);
226  b = load_with_perm_vec(i * 16, src2, mask_);
227  d = vec_avg(a, b);
228  put_unligned_store(d, dst);
229  dst += dst_stride;
230  }
231 }
232 
233 #if HAVE_BIGENDIAN
234 #define avg_unligned_store(s, dest){ \
235  tmp1 = vec_ld(0, dest); \
236  mask = vec_lvsl(0, dest); \
237  tmp2 = vec_ld(15, dest); \
238  a = vec_avg(vec_perm(tmp1, tmp2, mask), s); \
239  edges = vec_perm(tmp2, tmp1, mask); \
240  align = vec_lvsr(0, dest); \
241  tmp2 = vec_perm(a, edges, align); \
242  tmp1 = vec_perm(edges, a, align); \
243  vec_st(tmp2, 15, dest); \
244  vec_st(tmp1, 0 , dest); \
245  }
246 #else
247 #define avg_unligned_store(s, dest){ \
248  a = vec_avg(vec_vsx_ld(0, dst), s); \
249  vec_vsx_st(a, 0, dst); \
250  }
251 #endif /* HAVE_BIGENDIAN */
252 
253 static inline void avg_pixels16_l2_altivec( uint8_t * dst, const uint8_t * src1,
254  const uint8_t * src2, int dst_stride,
255  int src_stride1, int h)
256 {
257  int i;
258  vec_u8 a, b, d, mask_;
259 
260 #if HAVE_BIGENDIAN
261  vec_u8 tmp1, tmp2, mask, edges, align;
262  mask_ = vec_lvsl(0, src2);
263 #endif
264 
265  for (i = 0; i < h; i++) {
266  a = unaligned_load(i * src_stride1, src1);
267  b = load_with_perm_vec(i * 16, src2, mask_);
268  d = vec_avg(a, b);
269  avg_unligned_store(d, dst);
270  dst += dst_stride;
271  }
272 }
273 
274 /* Implemented but could be faster
275 #define put_pixels16_l2_altivec(d,s1,s2,ds,s1s,h) put_pixels16_l2(d,s1,s2,ds,s1s,16,h)
276 #define avg_pixels16_l2_altivec(d,s1,s2,ds,s1s,h) avg_pixels16_l2(d,s1,s2,ds,s1s,16,h)
277  */
278 
279 H264_MC(put_, 16, altivec)
280 H264_MC(avg_, 16, altivec)
281 #endif /* HAVE_ALTIVEC */
282 
284 {
285 #if HAVE_ALTIVEC
286  const int high_bit_depth = bit_depth > 8;
287 
289  return;
290 
291  if (!high_bit_depth) {
292 #define dspfunc(PFX, IDX, NUM) \
293  c->PFX ## _pixels_tab[IDX][ 0] = PFX ## NUM ## _mc00_altivec; \
294  c->PFX ## _pixels_tab[IDX][ 1] = PFX ## NUM ## _mc10_altivec; \
295  c->PFX ## _pixels_tab[IDX][ 2] = PFX ## NUM ## _mc20_altivec; \
296  c->PFX ## _pixels_tab[IDX][ 3] = PFX ## NUM ## _mc30_altivec; \
297  c->PFX ## _pixels_tab[IDX][ 4] = PFX ## NUM ## _mc01_altivec; \
298  c->PFX ## _pixels_tab[IDX][ 5] = PFX ## NUM ## _mc11_altivec; \
299  c->PFX ## _pixels_tab[IDX][ 6] = PFX ## NUM ## _mc21_altivec; \
300  c->PFX ## _pixels_tab[IDX][ 7] = PFX ## NUM ## _mc31_altivec; \
301  c->PFX ## _pixels_tab[IDX][ 8] = PFX ## NUM ## _mc02_altivec; \
302  c->PFX ## _pixels_tab[IDX][ 9] = PFX ## NUM ## _mc12_altivec; \
303  c->PFX ## _pixels_tab[IDX][10] = PFX ## NUM ## _mc22_altivec; \
304  c->PFX ## _pixels_tab[IDX][11] = PFX ## NUM ## _mc32_altivec; \
305  c->PFX ## _pixels_tab[IDX][12] = PFX ## NUM ## _mc03_altivec; \
306  c->PFX ## _pixels_tab[IDX][13] = PFX ## NUM ## _mc13_altivec; \
307  c->PFX ## _pixels_tab[IDX][14] = PFX ## NUM ## _mc23_altivec; \
308  c->PFX ## _pixels_tab[IDX][15] = PFX ## NUM ## _mc33_altivec
309 
310  dspfunc(put_h264_qpel, 0, 16);
311  dspfunc(avg_h264_qpel, 0, 16);
312 #undef dspfunc
313  }
314 #endif /* HAVE_ALTIVEC */
315 }
static void bit_depth(AudioStatsContext *s, uint64_t mask, uint64_t imask, AVRational *depth)
Definition: af_astats.c:254
Macro definitions for various function/variable attributes.
#define av_cold
Definition: attributes.h:88
uint8_t
#define dspfunc(PFX, IDX, NUM)
int av_get_cpu_flags(void)
Return the flags which specify extensions supported by the CPU.
Definition: cpu.c:95
#define H264_MC(OPNAME, SIZE)
int i
Definition: input.c:407
av_cold void ff_h264qpel_init_ppc(H264QpelContext *c, int bit_depth)
Definition: h264qpel.c:283
static const uint16_t mask[17]
Definition: lzw.c:38
#define PPC_ALTIVEC(flags)
Definition: cpu.h:25
#define src1
Definition: h264qpel.c:49
Contains misc utility macros and inline functions.
#define vec_u8
Definition: util_altivec.h:34
const char * b
Definition: vf_curves.c:118
static double c[64]