Mercurial > libavcodec.hg
annotate arm/dsputil_init_neon.c @ 11637:f7281af560fe libavcodec
vp3: DC-only IDCT
2-4% faster overall decode
author | conrad |
---|---|
date | Sat, 17 Apr 2010 02:04:30 +0000 |
parents | 2a4dc3c0b012 |
children | 659f16d04776 |
rev | line source |
---|---|
8334 | 1 /* |
2 * ARM NEON optimised DSP functions | |
3 * Copyright (c) 2008 Mans Rullgard <mans@mansr.com> | |
4 * | |
5 * This file is part of FFmpeg. | |
6 * | |
7 * FFmpeg is free software; you can redistribute it and/or | |
8 * modify it under the terms of the GNU Lesser General Public | |
9 * License as published by the Free Software Foundation; either | |
10 * version 2.1 of the License, or (at your option) any later version. | |
11 * | |
12 * FFmpeg is distributed in the hope that it will be useful, | |
13 * but WITHOUT ANY WARRANTY; without even the implied warranty of | |
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU | |
15 * Lesser General Public License for more details. | |
16 * | |
17 * You should have received a copy of the GNU Lesser General Public | |
18 * License along with FFmpeg; if not, write to the Free Software | |
19 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA | |
20 */ | |
21 | |
22 #include <stdint.h> | |
23 | |
24 #include "libavcodec/avcodec.h" | |
25 #include "libavcodec/dsputil.h" | |
10359 | 26 #include "dsputil_arm.h" |
27 | |
28 void ff_simple_idct_neon(DCTELEM *data); | |
29 void ff_simple_idct_put_neon(uint8_t *dest, int line_size, DCTELEM *data); | |
30 void ff_simple_idct_add_neon(uint8_t *dest, int line_size, DCTELEM *data); | |
31 | |
32 void ff_vp3_idct_neon(DCTELEM *data); | |
33 void ff_vp3_idct_put_neon(uint8_t *dest, int line_size, DCTELEM *data); | |
34 void ff_vp3_idct_add_neon(uint8_t *dest, int line_size, DCTELEM *data); | |
11637 | 35 void ff_vp3_idct_dc_add_neon(uint8_t *dest, int line_size, const DCTELEM *data); |
8334 | 36 |
37 void ff_put_pixels16_neon(uint8_t *, const uint8_t *, int, int); | |
38 void ff_put_pixels16_x2_neon(uint8_t *, const uint8_t *, int, int); | |
39 void ff_put_pixels16_y2_neon(uint8_t *, const uint8_t *, int, int); | |
40 void ff_put_pixels16_xy2_neon(uint8_t *, const uint8_t *, int, int); | |
41 void ff_put_pixels8_neon(uint8_t *, const uint8_t *, int, int); | |
42 void ff_put_pixels8_x2_neon(uint8_t *, const uint8_t *, int, int); | |
43 void ff_put_pixels8_y2_neon(uint8_t *, const uint8_t *, int, int); | |
44 void ff_put_pixels8_xy2_neon(uint8_t *, const uint8_t *, int, int); | |
45 void ff_put_pixels16_x2_no_rnd_neon(uint8_t *, const uint8_t *, int, int); | |
46 void ff_put_pixels16_y2_no_rnd_neon(uint8_t *, const uint8_t *, int, int); | |
47 void ff_put_pixels16_xy2_no_rnd_neon(uint8_t *, const uint8_t *, int, int); | |
48 void ff_put_pixels8_x2_no_rnd_neon(uint8_t *, const uint8_t *, int, int); | |
49 void ff_put_pixels8_y2_no_rnd_neon(uint8_t *, const uint8_t *, int, int); | |
50 void ff_put_pixels8_xy2_no_rnd_neon(uint8_t *, const uint8_t *, int, int); | |
51 | |
52 void ff_avg_pixels16_neon(uint8_t *, const uint8_t *, int, int); | |
10375 | 53 void ff_avg_pixels8_neon(uint8_t *, const uint8_t *, int, int); |
8334 | 54 |
9344 | 55 void ff_add_pixels_clamped_neon(const DCTELEM *, uint8_t *, int); |
9580 | 56 void ff_put_pixels_clamped_neon(const DCTELEM *, uint8_t *, int); |
9345 | 57 void ff_put_signed_pixels_clamped_neon(const DCTELEM *, uint8_t *, int); |
9344 | 58 |
8334 | 59 void ff_put_h264_qpel16_mc00_neon(uint8_t *, uint8_t *, int); |
8338 | 60 void ff_put_h264_qpel16_mc10_neon(uint8_t *, uint8_t *, int); |
61 void ff_put_h264_qpel16_mc20_neon(uint8_t *, uint8_t *, int); | |
62 void ff_put_h264_qpel16_mc30_neon(uint8_t *, uint8_t *, int); | |
63 void ff_put_h264_qpel16_mc01_neon(uint8_t *, uint8_t *, int); | |
64 void ff_put_h264_qpel16_mc11_neon(uint8_t *, uint8_t *, int); | |
65 void ff_put_h264_qpel16_mc21_neon(uint8_t *, uint8_t *, int); | |
66 void ff_put_h264_qpel16_mc31_neon(uint8_t *, uint8_t *, int); | |
67 void ff_put_h264_qpel16_mc02_neon(uint8_t *, uint8_t *, int); | |
68 void ff_put_h264_qpel16_mc12_neon(uint8_t *, uint8_t *, int); | |
69 void ff_put_h264_qpel16_mc22_neon(uint8_t *, uint8_t *, int); | |
70 void ff_put_h264_qpel16_mc32_neon(uint8_t *, uint8_t *, int); | |
71 void ff_put_h264_qpel16_mc03_neon(uint8_t *, uint8_t *, int); | |
72 void ff_put_h264_qpel16_mc13_neon(uint8_t *, uint8_t *, int); | |
73 void ff_put_h264_qpel16_mc23_neon(uint8_t *, uint8_t *, int); | |
74 void ff_put_h264_qpel16_mc33_neon(uint8_t *, uint8_t *, int); | |
75 | |
8334 | 76 void ff_put_h264_qpel8_mc00_neon(uint8_t *, uint8_t *, int); |
8338 | 77 void ff_put_h264_qpel8_mc10_neon(uint8_t *, uint8_t *, int); |
78 void ff_put_h264_qpel8_mc20_neon(uint8_t *, uint8_t *, int); | |
79 void ff_put_h264_qpel8_mc30_neon(uint8_t *, uint8_t *, int); | |
80 void ff_put_h264_qpel8_mc01_neon(uint8_t *, uint8_t *, int); | |
81 void ff_put_h264_qpel8_mc11_neon(uint8_t *, uint8_t *, int); | |
82 void ff_put_h264_qpel8_mc21_neon(uint8_t *, uint8_t *, int); | |
83 void ff_put_h264_qpel8_mc31_neon(uint8_t *, uint8_t *, int); | |
84 void ff_put_h264_qpel8_mc02_neon(uint8_t *, uint8_t *, int); | |
85 void ff_put_h264_qpel8_mc12_neon(uint8_t *, uint8_t *, int); | |
86 void ff_put_h264_qpel8_mc22_neon(uint8_t *, uint8_t *, int); | |
87 void ff_put_h264_qpel8_mc32_neon(uint8_t *, uint8_t *, int); | |
88 void ff_put_h264_qpel8_mc03_neon(uint8_t *, uint8_t *, int); | |
89 void ff_put_h264_qpel8_mc13_neon(uint8_t *, uint8_t *, int); | |
90 void ff_put_h264_qpel8_mc23_neon(uint8_t *, uint8_t *, int); | |
91 void ff_put_h264_qpel8_mc33_neon(uint8_t *, uint8_t *, int); | |
8334 | 92 |
93 void ff_avg_h264_qpel16_mc00_neon(uint8_t *, uint8_t *, int); | |
10616 | 94 void ff_avg_h264_qpel16_mc10_neon(uint8_t *, uint8_t *, int); |
95 void ff_avg_h264_qpel16_mc20_neon(uint8_t *, uint8_t *, int); | |
96 void ff_avg_h264_qpel16_mc30_neon(uint8_t *, uint8_t *, int); | |
97 void ff_avg_h264_qpel16_mc01_neon(uint8_t *, uint8_t *, int); | |
98 void ff_avg_h264_qpel16_mc11_neon(uint8_t *, uint8_t *, int); | |
99 void ff_avg_h264_qpel16_mc21_neon(uint8_t *, uint8_t *, int); | |
100 void ff_avg_h264_qpel16_mc31_neon(uint8_t *, uint8_t *, int); | |
101 void ff_avg_h264_qpel16_mc02_neon(uint8_t *, uint8_t *, int); | |
102 void ff_avg_h264_qpel16_mc12_neon(uint8_t *, uint8_t *, int); | |
103 void ff_avg_h264_qpel16_mc22_neon(uint8_t *, uint8_t *, int); | |
104 void ff_avg_h264_qpel16_mc32_neon(uint8_t *, uint8_t *, int); | |
105 void ff_avg_h264_qpel16_mc03_neon(uint8_t *, uint8_t *, int); | |
106 void ff_avg_h264_qpel16_mc13_neon(uint8_t *, uint8_t *, int); | |
107 void ff_avg_h264_qpel16_mc23_neon(uint8_t *, uint8_t *, int); | |
108 void ff_avg_h264_qpel16_mc33_neon(uint8_t *, uint8_t *, int); | |
8334 | 109 |
10375 | 110 void ff_avg_h264_qpel8_mc00_neon(uint8_t *, uint8_t *, int); |
10616 | 111 void ff_avg_h264_qpel8_mc10_neon(uint8_t *, uint8_t *, int); |
112 void ff_avg_h264_qpel8_mc20_neon(uint8_t *, uint8_t *, int); | |
113 void ff_avg_h264_qpel8_mc30_neon(uint8_t *, uint8_t *, int); | |
114 void ff_avg_h264_qpel8_mc01_neon(uint8_t *, uint8_t *, int); | |
115 void ff_avg_h264_qpel8_mc11_neon(uint8_t *, uint8_t *, int); | |
116 void ff_avg_h264_qpel8_mc21_neon(uint8_t *, uint8_t *, int); | |
117 void ff_avg_h264_qpel8_mc31_neon(uint8_t *, uint8_t *, int); | |
118 void ff_avg_h264_qpel8_mc02_neon(uint8_t *, uint8_t *, int); | |
119 void ff_avg_h264_qpel8_mc12_neon(uint8_t *, uint8_t *, int); | |
120 void ff_avg_h264_qpel8_mc22_neon(uint8_t *, uint8_t *, int); | |
121 void ff_avg_h264_qpel8_mc32_neon(uint8_t *, uint8_t *, int); | |
122 void ff_avg_h264_qpel8_mc03_neon(uint8_t *, uint8_t *, int); | |
123 void ff_avg_h264_qpel8_mc13_neon(uint8_t *, uint8_t *, int); | |
124 void ff_avg_h264_qpel8_mc23_neon(uint8_t *, uint8_t *, int); | |
125 void ff_avg_h264_qpel8_mc33_neon(uint8_t *, uint8_t *, int); | |
10375 | 126 |
8336 | 127 void ff_put_h264_chroma_mc8_neon(uint8_t *, uint8_t *, int, int, int, int); |
128 void ff_put_h264_chroma_mc4_neon(uint8_t *, uint8_t *, int, int, int, int); | |
10617 | 129 void ff_put_h264_chroma_mc2_neon(uint8_t *, uint8_t *, int, int, int, int); |
8336 | 130 |
131 void ff_avg_h264_chroma_mc8_neon(uint8_t *, uint8_t *, int, int, int, int); | |
132 void ff_avg_h264_chroma_mc4_neon(uint8_t *, uint8_t *, int, int, int, int); | |
10617 | 133 void ff_avg_h264_chroma_mc2_neon(uint8_t *, uint8_t *, int, int, int, int); |
8336 | 134 |
9692 | 135 void ff_vp3_v_loop_filter_neon(uint8_t *, int, int *); |
136 void ff_vp3_h_loop_filter_neon(uint8_t *, int, int *); | |
137 | |
8697 | 138 void ff_vector_fmul_neon(float *dst, const float *src, int len); |
8698 | 139 void ff_vector_fmul_window_neon(float *dst, const float *src0, |
140 const float *src1, const float *win, | |
141 float add_bias, int len); | |
10221 | 142 void ff_vector_fmul_scalar_neon(float *dst, const float *src, float mul, |
143 int len); | |
144 void ff_vector_fmul_sv_scalar_2_neon(float *dst, const float *src, | |
145 const float **vp, float mul, int len); | |
146 void ff_vector_fmul_sv_scalar_4_neon(float *dst, const float *src, | |
147 const float **vp, float mul, int len); | |
148 void ff_sv_fmul_scalar_2_neon(float *dst, const float **vp, float mul, | |
149 int len); | |
150 void ff_sv_fmul_scalar_4_neon(float *dst, const float **vp, float mul, | |
151 int len); | |
152 void ff_butterflies_float_neon(float *v1, float *v2, int len); | |
10228 | 153 float ff_scalarproduct_float_neon(const float *v1, const float *v2, int len); |
10253 | 154 void ff_int32_to_float_fmul_scalar_neon(float *dst, const int *src, |
155 float mul, int len); | |
10274 | 156 void ff_vector_fmul_reverse_neon(float *dst, const float *src0, |
157 const float *src1, int len); | |
10302 | 158 void ff_vector_fmul_add_neon(float *dst, const float *src0, const float *src1, |
159 const float *src2, int len); | |
8697 | 160 |
10276 | 161 void ff_vector_clipf_neon(float *dst, const float *src, float min, float max, |
162 int len); | |
8492 | 163 void ff_float_to_int16_neon(int16_t *, const float *, long); |
164 void ff_float_to_int16_interleave_neon(int16_t *, const float **, long, int); | |
165 | |
10046 | 166 void ff_vorbis_inverse_coupling_neon(float *mag, float *ang, int blocksize); |
167 | |
11243
e71b0be9ac79
ARM: NEON scalarproduct_int16 and scalarproduct_and_madd_int16
mru
parents:
10617
diff
changeset
|
168 int32_t ff_scalarproduct_int16_neon(int16_t *v1, int16_t *v2, int len, |
e71b0be9ac79
ARM: NEON scalarproduct_int16 and scalarproduct_and_madd_int16
mru
parents:
10617
diff
changeset
|
169 int shift); |
e71b0be9ac79
ARM: NEON scalarproduct_int16 and scalarproduct_and_madd_int16
mru
parents:
10617
diff
changeset
|
170 int32_t ff_scalarproduct_and_madd_int16_neon(int16_t *v1, int16_t *v2, |
e71b0be9ac79
ARM: NEON scalarproduct_int16 and scalarproduct_and_madd_int16
mru
parents:
10617
diff
changeset
|
171 int16_t *v3, int len, int mul); |
e71b0be9ac79
ARM: NEON scalarproduct_int16 and scalarproduct_and_madd_int16
mru
parents:
10617
diff
changeset
|
172 |
8334 | 173 void ff_dsputil_init_neon(DSPContext *c, AVCodecContext *avctx) |
174 { | |
10359 | 175 if (!avctx->lowres) { |
176 if (avctx->idct_algo == FF_IDCT_AUTO || | |
177 avctx->idct_algo == FF_IDCT_SIMPLENEON) { | |
10362 | 178 c->idct_put = ff_simple_idct_put_neon; |
179 c->idct_add = ff_simple_idct_add_neon; | |
180 c->idct = ff_simple_idct_neon; | |
10359 | 181 c->idct_permutation_type = FF_PARTTRANS_IDCT_PERM; |
182 } else if ((CONFIG_VP3_DECODER || CONFIG_VP5_DECODER || | |
183 CONFIG_VP6_DECODER) && | |
184 avctx->idct_algo == FF_IDCT_VP3) { | |
10362 | 185 c->idct_put = ff_vp3_idct_put_neon; |
186 c->idct_add = ff_vp3_idct_add_neon; | |
187 c->idct = ff_vp3_idct_neon; | |
10359 | 188 c->idct_permutation_type = FF_TRANSPOSE_IDCT_PERM; |
189 } | |
190 } | |
191 | |
8334 | 192 c->put_pixels_tab[0][0] = ff_put_pixels16_neon; |
193 c->put_pixels_tab[0][1] = ff_put_pixels16_x2_neon; | |
194 c->put_pixels_tab[0][2] = ff_put_pixels16_y2_neon; | |
195 c->put_pixels_tab[0][3] = ff_put_pixels16_xy2_neon; | |
196 c->put_pixels_tab[1][0] = ff_put_pixels8_neon; | |
197 c->put_pixels_tab[1][1] = ff_put_pixels8_x2_neon; | |
198 c->put_pixels_tab[1][2] = ff_put_pixels8_y2_neon; | |
199 c->put_pixels_tab[1][3] = ff_put_pixels8_xy2_neon; | |
200 | |
201 c->put_no_rnd_pixels_tab[0][0] = ff_put_pixels16_neon; | |
202 c->put_no_rnd_pixels_tab[0][1] = ff_put_pixels16_x2_no_rnd_neon; | |
203 c->put_no_rnd_pixels_tab[0][2] = ff_put_pixels16_y2_no_rnd_neon; | |
204 c->put_no_rnd_pixels_tab[0][3] = ff_put_pixels16_xy2_no_rnd_neon; | |
205 c->put_no_rnd_pixels_tab[1][0] = ff_put_pixels8_neon; | |
206 c->put_no_rnd_pixels_tab[1][1] = ff_put_pixels8_x2_no_rnd_neon; | |
207 c->put_no_rnd_pixels_tab[1][2] = ff_put_pixels8_y2_no_rnd_neon; | |
208 c->put_no_rnd_pixels_tab[1][3] = ff_put_pixels8_xy2_no_rnd_neon; | |
209 | |
210 c->avg_pixels_tab[0][0] = ff_avg_pixels16_neon; | |
10375 | 211 c->avg_pixels_tab[1][0] = ff_avg_pixels8_neon; |
8334 | 212 |
9344 | 213 c->add_pixels_clamped = ff_add_pixels_clamped_neon; |
9580 | 214 c->put_pixels_clamped = ff_put_pixels_clamped_neon; |
9345 | 215 c->put_signed_pixels_clamped = ff_put_signed_pixels_clamped_neon; |
9344 | 216 |
9976
e52cd349e708
Only compile in NEON optimizations for H.264 when the H.264 decoder is enabled.
diego
parents:
9975
diff
changeset
|
217 if (CONFIG_H264_DECODER) { |
9977 | 218 c->put_h264_chroma_pixels_tab[0] = ff_put_h264_chroma_mc8_neon; |
219 c->put_h264_chroma_pixels_tab[1] = ff_put_h264_chroma_mc4_neon; | |
10617 | 220 c->put_h264_chroma_pixels_tab[2] = ff_put_h264_chroma_mc2_neon; |
8336 | 221 |
9977 | 222 c->avg_h264_chroma_pixels_tab[0] = ff_avg_h264_chroma_mc8_neon; |
223 c->avg_h264_chroma_pixels_tab[1] = ff_avg_h264_chroma_mc4_neon; | |
10617 | 224 c->avg_h264_chroma_pixels_tab[2] = ff_avg_h264_chroma_mc2_neon; |
8336 | 225 |
9977 | 226 c->put_h264_qpel_pixels_tab[0][ 0] = ff_put_h264_qpel16_mc00_neon; |
227 c->put_h264_qpel_pixels_tab[0][ 1] = ff_put_h264_qpel16_mc10_neon; | |
228 c->put_h264_qpel_pixels_tab[0][ 2] = ff_put_h264_qpel16_mc20_neon; | |
229 c->put_h264_qpel_pixels_tab[0][ 3] = ff_put_h264_qpel16_mc30_neon; | |
230 c->put_h264_qpel_pixels_tab[0][ 4] = ff_put_h264_qpel16_mc01_neon; | |
231 c->put_h264_qpel_pixels_tab[0][ 5] = ff_put_h264_qpel16_mc11_neon; | |
232 c->put_h264_qpel_pixels_tab[0][ 6] = ff_put_h264_qpel16_mc21_neon; | |
233 c->put_h264_qpel_pixels_tab[0][ 7] = ff_put_h264_qpel16_mc31_neon; | |
234 c->put_h264_qpel_pixels_tab[0][ 8] = ff_put_h264_qpel16_mc02_neon; | |
235 c->put_h264_qpel_pixels_tab[0][ 9] = ff_put_h264_qpel16_mc12_neon; | |
236 c->put_h264_qpel_pixels_tab[0][10] = ff_put_h264_qpel16_mc22_neon; | |
237 c->put_h264_qpel_pixels_tab[0][11] = ff_put_h264_qpel16_mc32_neon; | |
238 c->put_h264_qpel_pixels_tab[0][12] = ff_put_h264_qpel16_mc03_neon; | |
239 c->put_h264_qpel_pixels_tab[0][13] = ff_put_h264_qpel16_mc13_neon; | |
240 c->put_h264_qpel_pixels_tab[0][14] = ff_put_h264_qpel16_mc23_neon; | |
241 c->put_h264_qpel_pixels_tab[0][15] = ff_put_h264_qpel16_mc33_neon; | |
8338 | 242 |
9977 | 243 c->put_h264_qpel_pixels_tab[1][ 0] = ff_put_h264_qpel8_mc00_neon; |
244 c->put_h264_qpel_pixels_tab[1][ 1] = ff_put_h264_qpel8_mc10_neon; | |
245 c->put_h264_qpel_pixels_tab[1][ 2] = ff_put_h264_qpel8_mc20_neon; | |
246 c->put_h264_qpel_pixels_tab[1][ 3] = ff_put_h264_qpel8_mc30_neon; | |
247 c->put_h264_qpel_pixels_tab[1][ 4] = ff_put_h264_qpel8_mc01_neon; | |
248 c->put_h264_qpel_pixels_tab[1][ 5] = ff_put_h264_qpel8_mc11_neon; | |
249 c->put_h264_qpel_pixels_tab[1][ 6] = ff_put_h264_qpel8_mc21_neon; | |
250 c->put_h264_qpel_pixels_tab[1][ 7] = ff_put_h264_qpel8_mc31_neon; | |
251 c->put_h264_qpel_pixels_tab[1][ 8] = ff_put_h264_qpel8_mc02_neon; | |
252 c->put_h264_qpel_pixels_tab[1][ 9] = ff_put_h264_qpel8_mc12_neon; | |
253 c->put_h264_qpel_pixels_tab[1][10] = ff_put_h264_qpel8_mc22_neon; | |
254 c->put_h264_qpel_pixels_tab[1][11] = ff_put_h264_qpel8_mc32_neon; | |
255 c->put_h264_qpel_pixels_tab[1][12] = ff_put_h264_qpel8_mc03_neon; | |
256 c->put_h264_qpel_pixels_tab[1][13] = ff_put_h264_qpel8_mc13_neon; | |
257 c->put_h264_qpel_pixels_tab[1][14] = ff_put_h264_qpel8_mc23_neon; | |
258 c->put_h264_qpel_pixels_tab[1][15] = ff_put_h264_qpel8_mc33_neon; | |
8334 | 259 |
9977 | 260 c->avg_h264_qpel_pixels_tab[0][ 0] = ff_avg_h264_qpel16_mc00_neon; |
10616 | 261 c->avg_h264_qpel_pixels_tab[0][ 1] = ff_avg_h264_qpel16_mc10_neon; |
262 c->avg_h264_qpel_pixels_tab[0][ 2] = ff_avg_h264_qpel16_mc20_neon; | |
263 c->avg_h264_qpel_pixels_tab[0][ 3] = ff_avg_h264_qpel16_mc30_neon; | |
264 c->avg_h264_qpel_pixels_tab[0][ 4] = ff_avg_h264_qpel16_mc01_neon; | |
265 c->avg_h264_qpel_pixels_tab[0][ 5] = ff_avg_h264_qpel16_mc11_neon; | |
266 c->avg_h264_qpel_pixels_tab[0][ 6] = ff_avg_h264_qpel16_mc21_neon; | |
267 c->avg_h264_qpel_pixels_tab[0][ 7] = ff_avg_h264_qpel16_mc31_neon; | |
268 c->avg_h264_qpel_pixels_tab[0][ 8] = ff_avg_h264_qpel16_mc02_neon; | |
269 c->avg_h264_qpel_pixels_tab[0][ 9] = ff_avg_h264_qpel16_mc12_neon; | |
270 c->avg_h264_qpel_pixels_tab[0][10] = ff_avg_h264_qpel16_mc22_neon; | |
271 c->avg_h264_qpel_pixels_tab[0][11] = ff_avg_h264_qpel16_mc32_neon; | |
272 c->avg_h264_qpel_pixels_tab[0][12] = ff_avg_h264_qpel16_mc03_neon; | |
273 c->avg_h264_qpel_pixels_tab[0][13] = ff_avg_h264_qpel16_mc13_neon; | |
274 c->avg_h264_qpel_pixels_tab[0][14] = ff_avg_h264_qpel16_mc23_neon; | |
275 c->avg_h264_qpel_pixels_tab[0][15] = ff_avg_h264_qpel16_mc33_neon; | |
8337 | 276 |
10375 | 277 c->avg_h264_qpel_pixels_tab[1][ 0] = ff_avg_h264_qpel8_mc00_neon; |
10616 | 278 c->avg_h264_qpel_pixels_tab[1][ 1] = ff_avg_h264_qpel8_mc10_neon; |
279 c->avg_h264_qpel_pixels_tab[1][ 2] = ff_avg_h264_qpel8_mc20_neon; | |
280 c->avg_h264_qpel_pixels_tab[1][ 3] = ff_avg_h264_qpel8_mc30_neon; | |
281 c->avg_h264_qpel_pixels_tab[1][ 4] = ff_avg_h264_qpel8_mc01_neon; | |
282 c->avg_h264_qpel_pixels_tab[1][ 5] = ff_avg_h264_qpel8_mc11_neon; | |
283 c->avg_h264_qpel_pixels_tab[1][ 6] = ff_avg_h264_qpel8_mc21_neon; | |
284 c->avg_h264_qpel_pixels_tab[1][ 7] = ff_avg_h264_qpel8_mc31_neon; | |
285 c->avg_h264_qpel_pixels_tab[1][ 8] = ff_avg_h264_qpel8_mc02_neon; | |
286 c->avg_h264_qpel_pixels_tab[1][ 9] = ff_avg_h264_qpel8_mc12_neon; | |
287 c->avg_h264_qpel_pixels_tab[1][10] = ff_avg_h264_qpel8_mc22_neon; | |
288 c->avg_h264_qpel_pixels_tab[1][11] = ff_avg_h264_qpel8_mc32_neon; | |
289 c->avg_h264_qpel_pixels_tab[1][12] = ff_avg_h264_qpel8_mc03_neon; | |
290 c->avg_h264_qpel_pixels_tab[1][13] = ff_avg_h264_qpel8_mc13_neon; | |
291 c->avg_h264_qpel_pixels_tab[1][14] = ff_avg_h264_qpel8_mc23_neon; | |
292 c->avg_h264_qpel_pixels_tab[1][15] = ff_avg_h264_qpel8_mc33_neon; | |
9976
e52cd349e708
Only compile in NEON optimizations for H.264 when the H.264 decoder is enabled.
diego
parents:
9975
diff
changeset
|
293 } |
8492 | 294 |
9975
d6d7e8d4a04d
Do not redundantly check for both CONFIG_THEORA_DECODER and CONFIG_VP3_DECODER.
diego
parents:
9692
diff
changeset
|
295 if (CONFIG_VP3_DECODER) { |
9692 | 296 c->vp3_v_loop_filter = ff_vp3_v_loop_filter_neon; |
297 c->vp3_h_loop_filter = ff_vp3_h_loop_filter_neon; | |
11637 | 298 c->vp3_idct_dc_add = ff_vp3_idct_dc_add_neon; |
9692 | 299 } |
300 | |
10362 | 301 c->vector_fmul = ff_vector_fmul_neon; |
302 c->vector_fmul_window = ff_vector_fmul_window_neon; | |
303 c->vector_fmul_scalar = ff_vector_fmul_scalar_neon; | |
304 c->butterflies_float = ff_butterflies_float_neon; | |
305 c->scalarproduct_float = ff_scalarproduct_float_neon; | |
10253 | 306 c->int32_to_float_fmul_scalar = ff_int32_to_float_fmul_scalar_neon; |
10362 | 307 c->vector_fmul_reverse = ff_vector_fmul_reverse_neon; |
308 c->vector_fmul_add = ff_vector_fmul_add_neon; | |
309 c->vector_clipf = ff_vector_clipf_neon; | |
10221 | 310 |
311 c->vector_fmul_sv_scalar[0] = ff_vector_fmul_sv_scalar_2_neon; | |
312 c->vector_fmul_sv_scalar[1] = ff_vector_fmul_sv_scalar_4_neon; | |
313 | |
314 c->sv_fmul_scalar[0] = ff_sv_fmul_scalar_2_neon; | |
315 c->sv_fmul_scalar[1] = ff_sv_fmul_scalar_4_neon; | |
8697 | 316 |
8492 | 317 if (!(avctx->flags & CODEC_FLAG_BITEXACT)) { |
10362 | 318 c->float_to_int16 = ff_float_to_int16_neon; |
8492 | 319 c->float_to_int16_interleave = ff_float_to_int16_interleave_neon; |
320 } | |
10046 | 321 |
322 if (CONFIG_VORBIS_DECODER) | |
323 c->vorbis_inverse_coupling = ff_vorbis_inverse_coupling_neon; | |
11243
e71b0be9ac79
ARM: NEON scalarproduct_int16 and scalarproduct_and_madd_int16
mru
parents:
10617
diff
changeset
|
324 |
e71b0be9ac79
ARM: NEON scalarproduct_int16 and scalarproduct_and_madd_int16
mru
parents:
10617
diff
changeset
|
325 c->scalarproduct_int16 = ff_scalarproduct_int16_neon; |
e71b0be9ac79
ARM: NEON scalarproduct_int16 and scalarproduct_and_madd_int16
mru
parents:
10617
diff
changeset
|
326 c->scalarproduct_and_madd_int16 = ff_scalarproduct_and_madd_int16_neon; |
8334 | 327 } |