Mercurial > libavcodec.hg
annotate arm/dsputil_init_neon.c @ 12530:63edd10ad4bc libavcodec tip
Try to fix crashes introduced by r25218
r25218 made assumptions about the existence of past reference frames that
weren't necessarily true.
| author | darkshikari |
|---|---|
| date | Tue, 28 Sep 2010 09:06:22 +0000 |
| parents | d3551fcf4c1c |
| children |
| 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 |
| 11807 | 37 void ff_clear_block_neon(DCTELEM *block); |
| 38 void ff_clear_blocks_neon(DCTELEM *blocks); | |
| 39 | |
| 8334 | 40 void ff_put_pixels16_neon(uint8_t *, const uint8_t *, int, int); |
| 41 void ff_put_pixels16_x2_neon(uint8_t *, const uint8_t *, int, int); | |
| 42 void ff_put_pixels16_y2_neon(uint8_t *, const uint8_t *, int, int); | |
| 43 void ff_put_pixels16_xy2_neon(uint8_t *, const uint8_t *, int, int); | |
| 44 void ff_put_pixels8_neon(uint8_t *, const uint8_t *, int, int); | |
| 45 void ff_put_pixels8_x2_neon(uint8_t *, const uint8_t *, int, int); | |
| 46 void ff_put_pixels8_y2_neon(uint8_t *, const uint8_t *, int, int); | |
| 47 void ff_put_pixels8_xy2_neon(uint8_t *, const uint8_t *, int, int); | |
| 48 void ff_put_pixels16_x2_no_rnd_neon(uint8_t *, const uint8_t *, int, int); | |
| 49 void ff_put_pixels16_y2_no_rnd_neon(uint8_t *, const uint8_t *, int, int); | |
| 50 void ff_put_pixels16_xy2_no_rnd_neon(uint8_t *, const uint8_t *, int, int); | |
| 51 void ff_put_pixels8_x2_no_rnd_neon(uint8_t *, const uint8_t *, int, int); | |
| 52 void ff_put_pixels8_y2_no_rnd_neon(uint8_t *, const uint8_t *, int, int); | |
| 53 void ff_put_pixels8_xy2_no_rnd_neon(uint8_t *, const uint8_t *, int, int); | |
| 54 | |
| 55 void ff_avg_pixels16_neon(uint8_t *, const uint8_t *, int, int); | |
| 10375 | 56 void ff_avg_pixels8_neon(uint8_t *, const uint8_t *, int, int); |
| 8334 | 57 |
| 9344 | 58 void ff_add_pixels_clamped_neon(const DCTELEM *, uint8_t *, int); |
| 9580 | 59 void ff_put_pixels_clamped_neon(const DCTELEM *, uint8_t *, int); |
| 9345 | 60 void ff_put_signed_pixels_clamped_neon(const DCTELEM *, uint8_t *, int); |
| 9344 | 61 |
| 8334 | 62 void ff_put_h264_qpel16_mc00_neon(uint8_t *, uint8_t *, int); |
| 8338 | 63 void ff_put_h264_qpel16_mc10_neon(uint8_t *, uint8_t *, int); |
| 64 void ff_put_h264_qpel16_mc20_neon(uint8_t *, uint8_t *, int); | |
| 65 void ff_put_h264_qpel16_mc30_neon(uint8_t *, uint8_t *, int); | |
| 66 void ff_put_h264_qpel16_mc01_neon(uint8_t *, uint8_t *, int); | |
| 67 void ff_put_h264_qpel16_mc11_neon(uint8_t *, uint8_t *, int); | |
| 68 void ff_put_h264_qpel16_mc21_neon(uint8_t *, uint8_t *, int); | |
| 69 void ff_put_h264_qpel16_mc31_neon(uint8_t *, uint8_t *, int); | |
| 70 void ff_put_h264_qpel16_mc02_neon(uint8_t *, uint8_t *, int); | |
| 71 void ff_put_h264_qpel16_mc12_neon(uint8_t *, uint8_t *, int); | |
| 72 void ff_put_h264_qpel16_mc22_neon(uint8_t *, uint8_t *, int); | |
| 73 void ff_put_h264_qpel16_mc32_neon(uint8_t *, uint8_t *, int); | |
| 74 void ff_put_h264_qpel16_mc03_neon(uint8_t *, uint8_t *, int); | |
| 75 void ff_put_h264_qpel16_mc13_neon(uint8_t *, uint8_t *, int); | |
| 76 void ff_put_h264_qpel16_mc23_neon(uint8_t *, uint8_t *, int); | |
| 77 void ff_put_h264_qpel16_mc33_neon(uint8_t *, uint8_t *, int); | |
| 78 | |
| 8334 | 79 void ff_put_h264_qpel8_mc00_neon(uint8_t *, uint8_t *, int); |
| 8338 | 80 void ff_put_h264_qpel8_mc10_neon(uint8_t *, uint8_t *, int); |
| 81 void ff_put_h264_qpel8_mc20_neon(uint8_t *, uint8_t *, int); | |
| 82 void ff_put_h264_qpel8_mc30_neon(uint8_t *, uint8_t *, int); | |
| 83 void ff_put_h264_qpel8_mc01_neon(uint8_t *, uint8_t *, int); | |
| 84 void ff_put_h264_qpel8_mc11_neon(uint8_t *, uint8_t *, int); | |
| 85 void ff_put_h264_qpel8_mc21_neon(uint8_t *, uint8_t *, int); | |
| 86 void ff_put_h264_qpel8_mc31_neon(uint8_t *, uint8_t *, int); | |
| 87 void ff_put_h264_qpel8_mc02_neon(uint8_t *, uint8_t *, int); | |
| 88 void ff_put_h264_qpel8_mc12_neon(uint8_t *, uint8_t *, int); | |
| 89 void ff_put_h264_qpel8_mc22_neon(uint8_t *, uint8_t *, int); | |
| 90 void ff_put_h264_qpel8_mc32_neon(uint8_t *, uint8_t *, int); | |
| 91 void ff_put_h264_qpel8_mc03_neon(uint8_t *, uint8_t *, int); | |
| 92 void ff_put_h264_qpel8_mc13_neon(uint8_t *, uint8_t *, int); | |
| 93 void ff_put_h264_qpel8_mc23_neon(uint8_t *, uint8_t *, int); | |
| 94 void ff_put_h264_qpel8_mc33_neon(uint8_t *, uint8_t *, int); | |
| 8334 | 95 |
| 96 void ff_avg_h264_qpel16_mc00_neon(uint8_t *, uint8_t *, int); | |
| 10616 | 97 void ff_avg_h264_qpel16_mc10_neon(uint8_t *, uint8_t *, int); |
| 98 void ff_avg_h264_qpel16_mc20_neon(uint8_t *, uint8_t *, int); | |
| 99 void ff_avg_h264_qpel16_mc30_neon(uint8_t *, uint8_t *, int); | |
| 100 void ff_avg_h264_qpel16_mc01_neon(uint8_t *, uint8_t *, int); | |
| 101 void ff_avg_h264_qpel16_mc11_neon(uint8_t *, uint8_t *, int); | |
| 102 void ff_avg_h264_qpel16_mc21_neon(uint8_t *, uint8_t *, int); | |
| 103 void ff_avg_h264_qpel16_mc31_neon(uint8_t *, uint8_t *, int); | |
| 104 void ff_avg_h264_qpel16_mc02_neon(uint8_t *, uint8_t *, int); | |
| 105 void ff_avg_h264_qpel16_mc12_neon(uint8_t *, uint8_t *, int); | |
| 106 void ff_avg_h264_qpel16_mc22_neon(uint8_t *, uint8_t *, int); | |
| 107 void ff_avg_h264_qpel16_mc32_neon(uint8_t *, uint8_t *, int); | |
| 108 void ff_avg_h264_qpel16_mc03_neon(uint8_t *, uint8_t *, int); | |
| 109 void ff_avg_h264_qpel16_mc13_neon(uint8_t *, uint8_t *, int); | |
| 110 void ff_avg_h264_qpel16_mc23_neon(uint8_t *, uint8_t *, int); | |
| 111 void ff_avg_h264_qpel16_mc33_neon(uint8_t *, uint8_t *, int); | |
| 8334 | 112 |
| 10375 | 113 void ff_avg_h264_qpel8_mc00_neon(uint8_t *, uint8_t *, int); |
| 10616 | 114 void ff_avg_h264_qpel8_mc10_neon(uint8_t *, uint8_t *, int); |
| 115 void ff_avg_h264_qpel8_mc20_neon(uint8_t *, uint8_t *, int); | |
| 116 void ff_avg_h264_qpel8_mc30_neon(uint8_t *, uint8_t *, int); | |
| 117 void ff_avg_h264_qpel8_mc01_neon(uint8_t *, uint8_t *, int); | |
| 118 void ff_avg_h264_qpel8_mc11_neon(uint8_t *, uint8_t *, int); | |
| 119 void ff_avg_h264_qpel8_mc21_neon(uint8_t *, uint8_t *, int); | |
| 120 void ff_avg_h264_qpel8_mc31_neon(uint8_t *, uint8_t *, int); | |
| 121 void ff_avg_h264_qpel8_mc02_neon(uint8_t *, uint8_t *, int); | |
| 122 void ff_avg_h264_qpel8_mc12_neon(uint8_t *, uint8_t *, int); | |
| 123 void ff_avg_h264_qpel8_mc22_neon(uint8_t *, uint8_t *, int); | |
| 124 void ff_avg_h264_qpel8_mc32_neon(uint8_t *, uint8_t *, int); | |
| 125 void ff_avg_h264_qpel8_mc03_neon(uint8_t *, uint8_t *, int); | |
| 126 void ff_avg_h264_qpel8_mc13_neon(uint8_t *, uint8_t *, int); | |
| 127 void ff_avg_h264_qpel8_mc23_neon(uint8_t *, uint8_t *, int); | |
| 128 void ff_avg_h264_qpel8_mc33_neon(uint8_t *, uint8_t *, int); | |
| 10375 | 129 |
| 8336 | 130 void ff_put_h264_chroma_mc8_neon(uint8_t *, uint8_t *, int, int, int, int); |
| 131 void ff_put_h264_chroma_mc4_neon(uint8_t *, uint8_t *, int, int, int, int); | |
| 10617 | 132 void ff_put_h264_chroma_mc2_neon(uint8_t *, uint8_t *, int, int, int, int); |
| 8336 | 133 |
| 134 void ff_avg_h264_chroma_mc8_neon(uint8_t *, uint8_t *, int, int, int, int); | |
| 135 void ff_avg_h264_chroma_mc4_neon(uint8_t *, uint8_t *, int, int, int, int); | |
| 10617 | 136 void ff_avg_h264_chroma_mc2_neon(uint8_t *, uint8_t *, int, int, int, int); |
| 8336 | 137 |
| 9692 | 138 void ff_vp3_v_loop_filter_neon(uint8_t *, int, int *); |
| 139 void ff_vp3_h_loop_filter_neon(uint8_t *, int, int *); | |
| 140 | |
| 8697 | 141 void ff_vector_fmul_neon(float *dst, const float *src, int len); |
| 8698 | 142 void ff_vector_fmul_window_neon(float *dst, const float *src0, |
| 143 const float *src1, const float *win, | |
| 144 float add_bias, int len); | |
| 10221 | 145 void ff_vector_fmul_scalar_neon(float *dst, const float *src, float mul, |
| 146 int len); | |
| 147 void ff_vector_fmul_sv_scalar_2_neon(float *dst, const float *src, | |
| 148 const float **vp, float mul, int len); | |
| 149 void ff_vector_fmul_sv_scalar_4_neon(float *dst, const float *src, | |
| 150 const float **vp, float mul, int len); | |
| 151 void ff_sv_fmul_scalar_2_neon(float *dst, const float **vp, float mul, | |
| 152 int len); | |
| 153 void ff_sv_fmul_scalar_4_neon(float *dst, const float **vp, float mul, | |
| 154 int len); | |
| 155 void ff_butterflies_float_neon(float *v1, float *v2, int len); | |
| 10228 | 156 float ff_scalarproduct_float_neon(const float *v1, const float *v2, int len); |
| 10253 | 157 void ff_int32_to_float_fmul_scalar_neon(float *dst, const int *src, |
| 158 float mul, int len); | |
| 10274 | 159 void ff_vector_fmul_reverse_neon(float *dst, const float *src0, |
| 160 const float *src1, int len); | |
| 10302 | 161 void ff_vector_fmul_add_neon(float *dst, const float *src0, const float *src1, |
| 162 const float *src2, int len); | |
| 8697 | 163 |
| 10276 | 164 void ff_vector_clipf_neon(float *dst, const float *src, float min, float max, |
| 165 int len); | |
| 8492 | 166 void ff_float_to_int16_neon(int16_t *, const float *, long); |
| 167 void ff_float_to_int16_interleave_neon(int16_t *, const float **, long, int); | |
| 168 | |
| 10046 | 169 void ff_vorbis_inverse_coupling_neon(float *mag, float *ang, int blocksize); |
| 170 | |
| 11981 | 171 int32_t ff_scalarproduct_int16_neon(const int16_t *v1, const int16_t *v2, int len, |
|
11243
e71b0be9ac79
ARM: NEON scalarproduct_int16 and scalarproduct_and_madd_int16
mru
parents:
10617
diff
changeset
|
172 int shift); |
| 11981 | 173 int32_t ff_scalarproduct_and_madd_int16_neon(int16_t *v1, const int16_t *v2, |
| 174 const int16_t *v3, int len, int mul); | |
|
11243
e71b0be9ac79
ARM: NEON scalarproduct_int16 and scalarproduct_and_madd_int16
mru
parents:
10617
diff
changeset
|
175 |
| 8334 | 176 void ff_dsputil_init_neon(DSPContext *c, AVCodecContext *avctx) |
| 177 { | |
| 10359 | 178 if (!avctx->lowres) { |
| 179 if (avctx->idct_algo == FF_IDCT_AUTO || | |
| 180 avctx->idct_algo == FF_IDCT_SIMPLENEON) { | |
| 10362 | 181 c->idct_put = ff_simple_idct_put_neon; |
| 182 c->idct_add = ff_simple_idct_add_neon; | |
| 183 c->idct = ff_simple_idct_neon; | |
| 10359 | 184 c->idct_permutation_type = FF_PARTTRANS_IDCT_PERM; |
| 185 } else if ((CONFIG_VP3_DECODER || CONFIG_VP5_DECODER || | |
| 186 CONFIG_VP6_DECODER) && | |
| 187 avctx->idct_algo == FF_IDCT_VP3) { | |
| 10362 | 188 c->idct_put = ff_vp3_idct_put_neon; |
| 189 c->idct_add = ff_vp3_idct_add_neon; | |
| 190 c->idct = ff_vp3_idct_neon; | |
| 10359 | 191 c->idct_permutation_type = FF_TRANSPOSE_IDCT_PERM; |
| 192 } | |
| 193 } | |
| 194 | |
| 11807 | 195 c->clear_block = ff_clear_block_neon; |
| 196 c->clear_blocks = ff_clear_blocks_neon; | |
| 197 | |
| 8334 | 198 c->put_pixels_tab[0][0] = ff_put_pixels16_neon; |
| 199 c->put_pixels_tab[0][1] = ff_put_pixels16_x2_neon; | |
| 200 c->put_pixels_tab[0][2] = ff_put_pixels16_y2_neon; | |
| 201 c->put_pixels_tab[0][3] = ff_put_pixels16_xy2_neon; | |
| 202 c->put_pixels_tab[1][0] = ff_put_pixels8_neon; | |
| 203 c->put_pixels_tab[1][1] = ff_put_pixels8_x2_neon; | |
| 204 c->put_pixels_tab[1][2] = ff_put_pixels8_y2_neon; | |
| 205 c->put_pixels_tab[1][3] = ff_put_pixels8_xy2_neon; | |
| 206 | |
| 207 c->put_no_rnd_pixels_tab[0][0] = ff_put_pixels16_neon; | |
| 208 c->put_no_rnd_pixels_tab[0][1] = ff_put_pixels16_x2_no_rnd_neon; | |
| 209 c->put_no_rnd_pixels_tab[0][2] = ff_put_pixels16_y2_no_rnd_neon; | |
| 210 c->put_no_rnd_pixels_tab[0][3] = ff_put_pixels16_xy2_no_rnd_neon; | |
| 211 c->put_no_rnd_pixels_tab[1][0] = ff_put_pixels8_neon; | |
| 212 c->put_no_rnd_pixels_tab[1][1] = ff_put_pixels8_x2_no_rnd_neon; | |
| 213 c->put_no_rnd_pixels_tab[1][2] = ff_put_pixels8_y2_no_rnd_neon; | |
| 214 c->put_no_rnd_pixels_tab[1][3] = ff_put_pixels8_xy2_no_rnd_neon; | |
| 215 | |
| 216 c->avg_pixels_tab[0][0] = ff_avg_pixels16_neon; | |
| 10375 | 217 c->avg_pixels_tab[1][0] = ff_avg_pixels8_neon; |
| 8334 | 218 |
| 9344 | 219 c->add_pixels_clamped = ff_add_pixels_clamped_neon; |
| 9580 | 220 c->put_pixels_clamped = ff_put_pixels_clamped_neon; |
| 9345 | 221 c->put_signed_pixels_clamped = ff_put_signed_pixels_clamped_neon; |
| 9344 | 222 |
|
9976
e52cd349e708
Only compile in NEON optimizations for H.264 when the H.264 decoder is enabled.
diego
parents:
9975
diff
changeset
|
223 if (CONFIG_H264_DECODER) { |
| 9977 | 224 c->put_h264_chroma_pixels_tab[0] = ff_put_h264_chroma_mc8_neon; |
| 225 c->put_h264_chroma_pixels_tab[1] = ff_put_h264_chroma_mc4_neon; | |
| 10617 | 226 c->put_h264_chroma_pixels_tab[2] = ff_put_h264_chroma_mc2_neon; |
| 8336 | 227 |
| 9977 | 228 c->avg_h264_chroma_pixels_tab[0] = ff_avg_h264_chroma_mc8_neon; |
| 229 c->avg_h264_chroma_pixels_tab[1] = ff_avg_h264_chroma_mc4_neon; | |
| 10617 | 230 c->avg_h264_chroma_pixels_tab[2] = ff_avg_h264_chroma_mc2_neon; |
| 8336 | 231 |
| 9977 | 232 c->put_h264_qpel_pixels_tab[0][ 0] = ff_put_h264_qpel16_mc00_neon; |
| 233 c->put_h264_qpel_pixels_tab[0][ 1] = ff_put_h264_qpel16_mc10_neon; | |
| 234 c->put_h264_qpel_pixels_tab[0][ 2] = ff_put_h264_qpel16_mc20_neon; | |
| 235 c->put_h264_qpel_pixels_tab[0][ 3] = ff_put_h264_qpel16_mc30_neon; | |
| 236 c->put_h264_qpel_pixels_tab[0][ 4] = ff_put_h264_qpel16_mc01_neon; | |
| 237 c->put_h264_qpel_pixels_tab[0][ 5] = ff_put_h264_qpel16_mc11_neon; | |
| 238 c->put_h264_qpel_pixels_tab[0][ 6] = ff_put_h264_qpel16_mc21_neon; | |
| 239 c->put_h264_qpel_pixels_tab[0][ 7] = ff_put_h264_qpel16_mc31_neon; | |
| 240 c->put_h264_qpel_pixels_tab[0][ 8] = ff_put_h264_qpel16_mc02_neon; | |
| 241 c->put_h264_qpel_pixels_tab[0][ 9] = ff_put_h264_qpel16_mc12_neon; | |
| 242 c->put_h264_qpel_pixels_tab[0][10] = ff_put_h264_qpel16_mc22_neon; | |
| 243 c->put_h264_qpel_pixels_tab[0][11] = ff_put_h264_qpel16_mc32_neon; | |
| 244 c->put_h264_qpel_pixels_tab[0][12] = ff_put_h264_qpel16_mc03_neon; | |
| 245 c->put_h264_qpel_pixels_tab[0][13] = ff_put_h264_qpel16_mc13_neon; | |
| 246 c->put_h264_qpel_pixels_tab[0][14] = ff_put_h264_qpel16_mc23_neon; | |
| 247 c->put_h264_qpel_pixels_tab[0][15] = ff_put_h264_qpel16_mc33_neon; | |
| 8338 | 248 |
| 9977 | 249 c->put_h264_qpel_pixels_tab[1][ 0] = ff_put_h264_qpel8_mc00_neon; |
| 250 c->put_h264_qpel_pixels_tab[1][ 1] = ff_put_h264_qpel8_mc10_neon; | |
| 251 c->put_h264_qpel_pixels_tab[1][ 2] = ff_put_h264_qpel8_mc20_neon; | |
| 252 c->put_h264_qpel_pixels_tab[1][ 3] = ff_put_h264_qpel8_mc30_neon; | |
| 253 c->put_h264_qpel_pixels_tab[1][ 4] = ff_put_h264_qpel8_mc01_neon; | |
| 254 c->put_h264_qpel_pixels_tab[1][ 5] = ff_put_h264_qpel8_mc11_neon; | |
| 255 c->put_h264_qpel_pixels_tab[1][ 6] = ff_put_h264_qpel8_mc21_neon; | |
| 256 c->put_h264_qpel_pixels_tab[1][ 7] = ff_put_h264_qpel8_mc31_neon; | |
| 257 c->put_h264_qpel_pixels_tab[1][ 8] = ff_put_h264_qpel8_mc02_neon; | |
| 258 c->put_h264_qpel_pixels_tab[1][ 9] = ff_put_h264_qpel8_mc12_neon; | |
| 259 c->put_h264_qpel_pixels_tab[1][10] = ff_put_h264_qpel8_mc22_neon; | |
| 260 c->put_h264_qpel_pixels_tab[1][11] = ff_put_h264_qpel8_mc32_neon; | |
| 261 c->put_h264_qpel_pixels_tab[1][12] = ff_put_h264_qpel8_mc03_neon; | |
| 262 c->put_h264_qpel_pixels_tab[1][13] = ff_put_h264_qpel8_mc13_neon; | |
| 263 c->put_h264_qpel_pixels_tab[1][14] = ff_put_h264_qpel8_mc23_neon; | |
| 264 c->put_h264_qpel_pixels_tab[1][15] = ff_put_h264_qpel8_mc33_neon; | |
| 8334 | 265 |
| 9977 | 266 c->avg_h264_qpel_pixels_tab[0][ 0] = ff_avg_h264_qpel16_mc00_neon; |
| 10616 | 267 c->avg_h264_qpel_pixels_tab[0][ 1] = ff_avg_h264_qpel16_mc10_neon; |
| 268 c->avg_h264_qpel_pixels_tab[0][ 2] = ff_avg_h264_qpel16_mc20_neon; | |
| 269 c->avg_h264_qpel_pixels_tab[0][ 3] = ff_avg_h264_qpel16_mc30_neon; | |
| 270 c->avg_h264_qpel_pixels_tab[0][ 4] = ff_avg_h264_qpel16_mc01_neon; | |
| 271 c->avg_h264_qpel_pixels_tab[0][ 5] = ff_avg_h264_qpel16_mc11_neon; | |
| 272 c->avg_h264_qpel_pixels_tab[0][ 6] = ff_avg_h264_qpel16_mc21_neon; | |
| 273 c->avg_h264_qpel_pixels_tab[0][ 7] = ff_avg_h264_qpel16_mc31_neon; | |
| 274 c->avg_h264_qpel_pixels_tab[0][ 8] = ff_avg_h264_qpel16_mc02_neon; | |
| 275 c->avg_h264_qpel_pixels_tab[0][ 9] = ff_avg_h264_qpel16_mc12_neon; | |
| 276 c->avg_h264_qpel_pixels_tab[0][10] = ff_avg_h264_qpel16_mc22_neon; | |
| 277 c->avg_h264_qpel_pixels_tab[0][11] = ff_avg_h264_qpel16_mc32_neon; | |
| 278 c->avg_h264_qpel_pixels_tab[0][12] = ff_avg_h264_qpel16_mc03_neon; | |
| 279 c->avg_h264_qpel_pixels_tab[0][13] = ff_avg_h264_qpel16_mc13_neon; | |
| 280 c->avg_h264_qpel_pixels_tab[0][14] = ff_avg_h264_qpel16_mc23_neon; | |
| 281 c->avg_h264_qpel_pixels_tab[0][15] = ff_avg_h264_qpel16_mc33_neon; | |
| 8337 | 282 |
| 10375 | 283 c->avg_h264_qpel_pixels_tab[1][ 0] = ff_avg_h264_qpel8_mc00_neon; |
| 10616 | 284 c->avg_h264_qpel_pixels_tab[1][ 1] = ff_avg_h264_qpel8_mc10_neon; |
| 285 c->avg_h264_qpel_pixels_tab[1][ 2] = ff_avg_h264_qpel8_mc20_neon; | |
| 286 c->avg_h264_qpel_pixels_tab[1][ 3] = ff_avg_h264_qpel8_mc30_neon; | |
| 287 c->avg_h264_qpel_pixels_tab[1][ 4] = ff_avg_h264_qpel8_mc01_neon; | |
| 288 c->avg_h264_qpel_pixels_tab[1][ 5] = ff_avg_h264_qpel8_mc11_neon; | |
| 289 c->avg_h264_qpel_pixels_tab[1][ 6] = ff_avg_h264_qpel8_mc21_neon; | |
| 290 c->avg_h264_qpel_pixels_tab[1][ 7] = ff_avg_h264_qpel8_mc31_neon; | |
| 291 c->avg_h264_qpel_pixels_tab[1][ 8] = ff_avg_h264_qpel8_mc02_neon; | |
| 292 c->avg_h264_qpel_pixels_tab[1][ 9] = ff_avg_h264_qpel8_mc12_neon; | |
| 293 c->avg_h264_qpel_pixels_tab[1][10] = ff_avg_h264_qpel8_mc22_neon; | |
| 294 c->avg_h264_qpel_pixels_tab[1][11] = ff_avg_h264_qpel8_mc32_neon; | |
| 295 c->avg_h264_qpel_pixels_tab[1][12] = ff_avg_h264_qpel8_mc03_neon; | |
| 296 c->avg_h264_qpel_pixels_tab[1][13] = ff_avg_h264_qpel8_mc13_neon; | |
| 297 c->avg_h264_qpel_pixels_tab[1][14] = ff_avg_h264_qpel8_mc23_neon; | |
| 298 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
|
299 } |
| 8492 | 300 |
|
9975
d6d7e8d4a04d
Do not redundantly check for both CONFIG_THEORA_DECODER and CONFIG_VP3_DECODER.
diego
parents:
9692
diff
changeset
|
301 if (CONFIG_VP3_DECODER) { |
| 9692 | 302 c->vp3_v_loop_filter = ff_vp3_v_loop_filter_neon; |
| 303 c->vp3_h_loop_filter = ff_vp3_h_loop_filter_neon; | |
| 11637 | 304 c->vp3_idct_dc_add = ff_vp3_idct_dc_add_neon; |
| 9692 | 305 } |
| 306 | |
| 10362 | 307 c->vector_fmul = ff_vector_fmul_neon; |
| 308 c->vector_fmul_window = ff_vector_fmul_window_neon; | |
| 309 c->vector_fmul_scalar = ff_vector_fmul_scalar_neon; | |
| 310 c->butterflies_float = ff_butterflies_float_neon; | |
| 311 c->scalarproduct_float = ff_scalarproduct_float_neon; | |
| 10253 | 312 c->int32_to_float_fmul_scalar = ff_int32_to_float_fmul_scalar_neon; |
| 10362 | 313 c->vector_fmul_reverse = ff_vector_fmul_reverse_neon; |
| 314 c->vector_fmul_add = ff_vector_fmul_add_neon; | |
| 315 c->vector_clipf = ff_vector_clipf_neon; | |
| 10221 | 316 |
| 317 c->vector_fmul_sv_scalar[0] = ff_vector_fmul_sv_scalar_2_neon; | |
| 318 c->vector_fmul_sv_scalar[1] = ff_vector_fmul_sv_scalar_4_neon; | |
| 319 | |
| 320 c->sv_fmul_scalar[0] = ff_sv_fmul_scalar_2_neon; | |
| 321 c->sv_fmul_scalar[1] = ff_sv_fmul_scalar_4_neon; | |
| 8697 | 322 |
| 8492 | 323 if (!(avctx->flags & CODEC_FLAG_BITEXACT)) { |
| 10362 | 324 c->float_to_int16 = ff_float_to_int16_neon; |
| 8492 | 325 c->float_to_int16_interleave = ff_float_to_int16_interleave_neon; |
| 326 } | |
| 10046 | 327 |
| 328 if (CONFIG_VORBIS_DECODER) | |
| 329 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
|
330 |
|
e71b0be9ac79
ARM: NEON scalarproduct_int16 and scalarproduct_and_madd_int16
mru
parents:
10617
diff
changeset
|
331 c->scalarproduct_int16 = ff_scalarproduct_int16_neon; |
|
e71b0be9ac79
ARM: NEON scalarproduct_int16 and scalarproduct_and_madd_int16
mru
parents:
10617
diff
changeset
|
332 c->scalarproduct_and_madd_int16 = ff_scalarproduct_and_madd_int16_neon; |
| 8334 | 333 } |
