| 1 | /* |
| 2 | * Copyright (c) 2010 Mans Rullgard <mans@mansr.com> |
| 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 <stdint.h> |
| 22 | |
| 23 | #include "libavutil/attributes.h" |
| 24 | #include "libavutil/cpu.h" |
| 25 | #include "libavutil/aarch64/cpu.h" |
| 26 | #include "libavcodec/h264dsp.h" |
| 27 | |
| 28 | void ff_h264_v_loop_filter_luma_neon(uint8_t *pix, int stride, int alpha, |
| 29 | int beta, int8_t *tc0); |
| 30 | void ff_h264_h_loop_filter_luma_neon(uint8_t *pix, int stride, int alpha, |
| 31 | int beta, int8_t *tc0); |
| 32 | void ff_h264_v_loop_filter_chroma_neon(uint8_t *pix, int stride, int alpha, |
| 33 | int beta, int8_t *tc0); |
| 34 | void ff_h264_h_loop_filter_chroma_neon(uint8_t *pix, int stride, int alpha, |
| 35 | int beta, int8_t *tc0); |
| 36 | |
| 37 | void ff_weight_h264_pixels_16_neon(uint8_t *dst, int stride, int height, |
| 38 | int log2_den, int weight, int offset); |
| 39 | void ff_weight_h264_pixels_8_neon(uint8_t *dst, int stride, int height, |
| 40 | int log2_den, int weight, int offset); |
| 41 | void ff_weight_h264_pixels_4_neon(uint8_t *dst, int stride, int height, |
| 42 | int log2_den, int weight, int offset); |
| 43 | |
| 44 | void ff_biweight_h264_pixels_16_neon(uint8_t *dst, uint8_t *src, int stride, |
| 45 | int height, int log2_den, int weightd, |
| 46 | int weights, int offset); |
| 47 | void ff_biweight_h264_pixels_8_neon(uint8_t *dst, uint8_t *src, int stride, |
| 48 | int height, int log2_den, int weightd, |
| 49 | int weights, int offset); |
| 50 | void ff_biweight_h264_pixels_4_neon(uint8_t *dst, uint8_t *src, int stride, |
| 51 | int height, int log2_den, int weightd, |
| 52 | int weights, int offset); |
| 53 | |
| 54 | void ff_h264_idct_add_neon(uint8_t *dst, int16_t *block, int stride); |
| 55 | void ff_h264_idct_dc_add_neon(uint8_t *dst, int16_t *block, int stride); |
| 56 | void ff_h264_idct_add16_neon(uint8_t *dst, const int *block_offset, |
| 57 | int16_t *block, int stride, |
| 58 | const uint8_t nnzc[6*8]); |
| 59 | void ff_h264_idct_add16intra_neon(uint8_t *dst, const int *block_offset, |
| 60 | int16_t *block, int stride, |
| 61 | const uint8_t nnzc[6*8]); |
| 62 | void ff_h264_idct_add8_neon(uint8_t **dest, const int *block_offset, |
| 63 | int16_t *block, int stride, |
| 64 | const uint8_t nnzc[6*8]); |
| 65 | |
| 66 | void ff_h264_idct8_add_neon(uint8_t *dst, int16_t *block, int stride); |
| 67 | void ff_h264_idct8_dc_add_neon(uint8_t *dst, int16_t *block, int stride); |
| 68 | void ff_h264_idct8_add4_neon(uint8_t *dst, const int *block_offset, |
| 69 | int16_t *block, int stride, |
| 70 | const uint8_t nnzc[6*8]); |
| 71 | |
| 72 | av_cold void ff_h264dsp_init_aarch64(H264DSPContext *c, const int bit_depth, |
| 73 | const int chroma_format_idc) |
| 74 | { |
| 75 | int cpu_flags = av_get_cpu_flags(); |
| 76 | |
| 77 | if (have_neon(cpu_flags) && bit_depth == 8) { |
| 78 | c->h264_v_loop_filter_luma = ff_h264_v_loop_filter_luma_neon; |
| 79 | c->h264_h_loop_filter_luma = ff_h264_h_loop_filter_luma_neon; |
| 80 | c->h264_v_loop_filter_chroma = ff_h264_v_loop_filter_chroma_neon; |
| 81 | c->h264_h_loop_filter_chroma = ff_h264_h_loop_filter_chroma_neon; |
| 82 | |
| 83 | c->weight_h264_pixels_tab[0] = ff_weight_h264_pixels_16_neon; |
| 84 | c->weight_h264_pixels_tab[1] = ff_weight_h264_pixels_8_neon; |
| 85 | c->weight_h264_pixels_tab[2] = ff_weight_h264_pixels_4_neon; |
| 86 | |
| 87 | c->biweight_h264_pixels_tab[0] = ff_biweight_h264_pixels_16_neon; |
| 88 | c->biweight_h264_pixels_tab[1] = ff_biweight_h264_pixels_8_neon; |
| 89 | c->biweight_h264_pixels_tab[2] = ff_biweight_h264_pixels_4_neon; |
| 90 | |
| 91 | c->h264_idct_add = ff_h264_idct_add_neon; |
| 92 | c->h264_idct_dc_add = ff_h264_idct_dc_add_neon; |
| 93 | c->h264_idct_add16 = ff_h264_idct_add16_neon; |
| 94 | c->h264_idct_add16intra = ff_h264_idct_add16intra_neon; |
| 95 | if (chroma_format_idc <= 1) |
| 96 | c->h264_idct_add8 = ff_h264_idct_add8_neon; |
| 97 | c->h264_idct8_add = ff_h264_idct8_add_neon; |
| 98 | c->h264_idct8_dc_add = ff_h264_idct8_dc_add_neon; |
| 99 | c->h264_idct8_add4 = ff_h264_idct8_add4_neon; |
| 100 | } |
| 101 | } |