yading@10: /*
yading@10:  * Copyright (c) 2010 Mans Rullgard <mans@mansr.com>
yading@10:  *
yading@10:  * This file is part of FFmpeg.
yading@10:  *
yading@10:  * FFmpeg is free software; you can redistribute it and/or
yading@10:  * modify it under the terms of the GNU Lesser General Public
yading@10:  * License as published by the Free Software Foundation; either
yading@10:  * version 2.1 of the License, or (at your option) any later version.
yading@10:  *
yading@10:  * FFmpeg is distributed in the hope that it will be useful,
yading@10:  * but WITHOUT ANY WARRANTY; without even the implied warranty of
yading@10:  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
yading@10:  * Lesser General Public License for more details.
yading@10:  *
yading@10:  * You should have received a copy of the GNU Lesser General Public
yading@10:  * License along with FFmpeg; if not, write to the Free Software
yading@10:  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
yading@10:  */
yading@10: 
yading@10: #include <stdint.h>
yading@10: 
yading@10: #include "libavutil/attributes.h"
yading@10: #include "libavutil/arm/cpu.h"
yading@10: #include "libavcodec/h264dsp.h"
yading@10: 
yading@10: void ff_h264_v_loop_filter_luma_neon(uint8_t *pix, int stride, int alpha,
yading@10:                                      int beta, int8_t *tc0);
yading@10: void ff_h264_h_loop_filter_luma_neon(uint8_t *pix, int stride, int alpha,
yading@10:                                      int beta, int8_t *tc0);
yading@10: void ff_h264_v_loop_filter_chroma_neon(uint8_t *pix, int stride, int alpha,
yading@10:                                        int beta, int8_t *tc0);
yading@10: void ff_h264_h_loop_filter_chroma_neon(uint8_t *pix, int stride, int alpha,
yading@10:                                        int beta, int8_t *tc0);
yading@10: 
yading@10: void ff_weight_h264_pixels_16_neon(uint8_t *dst, int stride, int height,
yading@10:                                    int log2_den, int weight, int offset);
yading@10: void ff_weight_h264_pixels_8_neon(uint8_t *dst, int stride, int height,
yading@10:                                   int log2_den, int weight, int offset);
yading@10: void ff_weight_h264_pixels_4_neon(uint8_t *dst, int stride, int height,
yading@10:                                   int log2_den, int weight, int offset);
yading@10: 
yading@10: void ff_biweight_h264_pixels_16_neon(uint8_t *dst, uint8_t *src, int stride,
yading@10:                                      int height, int log2_den, int weightd,
yading@10:                                      int weights, int offset);
yading@10: void ff_biweight_h264_pixels_8_neon(uint8_t *dst, uint8_t *src, int stride,
yading@10:                                     int height, int log2_den, int weightd,
yading@10:                                     int weights, int offset);
yading@10: void ff_biweight_h264_pixels_4_neon(uint8_t *dst, uint8_t *src, int stride,
yading@10:                                     int height, int log2_den, int weightd,
yading@10:                                     int weights, int offset);
yading@10: 
yading@10: void ff_h264_idct_add_neon(uint8_t *dst, int16_t *block, int stride);
yading@10: void ff_h264_idct_dc_add_neon(uint8_t *dst, int16_t *block, int stride);
yading@10: void ff_h264_idct_add16_neon(uint8_t *dst, const int *block_offset,
yading@10:                              int16_t *block, int stride,
yading@10:                              const uint8_t nnzc[6*8]);
yading@10: void ff_h264_idct_add16intra_neon(uint8_t *dst, const int *block_offset,
yading@10:                                   int16_t *block, int stride,
yading@10:                                   const uint8_t nnzc[6*8]);
yading@10: void ff_h264_idct_add8_neon(uint8_t **dest, const int *block_offset,
yading@10:                             int16_t *block, int stride,
yading@10:                             const uint8_t nnzc[6*8]);
yading@10: 
yading@10: void ff_h264_idct8_add_neon(uint8_t *dst, int16_t *block, int stride);
yading@10: void ff_h264_idct8_dc_add_neon(uint8_t *dst, int16_t *block, int stride);
yading@10: void ff_h264_idct8_add4_neon(uint8_t *dst, const int *block_offset,
yading@10:                              int16_t *block, int stride,
yading@10:                              const uint8_t nnzc[6*8]);
yading@10: 
yading@10: static av_cold void ff_h264dsp_init_neon(H264DSPContext *c, const int bit_depth,
yading@10:                                          const int chroma_format_idc)
yading@10: {
yading@10: #if HAVE_NEON
yading@10:     if (bit_depth == 8) {
yading@10:     c->h264_v_loop_filter_luma   = ff_h264_v_loop_filter_luma_neon;
yading@10:     c->h264_h_loop_filter_luma   = ff_h264_h_loop_filter_luma_neon;
yading@10:     if(chroma_format_idc == 1){
yading@10:     c->h264_v_loop_filter_chroma = ff_h264_v_loop_filter_chroma_neon;
yading@10:     c->h264_h_loop_filter_chroma = ff_h264_h_loop_filter_chroma_neon;
yading@10:     }
yading@10: 
yading@10:     c->weight_h264_pixels_tab[0] = ff_weight_h264_pixels_16_neon;
yading@10:     c->weight_h264_pixels_tab[1] = ff_weight_h264_pixels_8_neon;
yading@10:     c->weight_h264_pixels_tab[2] = ff_weight_h264_pixels_4_neon;
yading@10: 
yading@10:     c->biweight_h264_pixels_tab[0] = ff_biweight_h264_pixels_16_neon;
yading@10:     c->biweight_h264_pixels_tab[1] = ff_biweight_h264_pixels_8_neon;
yading@10:     c->biweight_h264_pixels_tab[2] = ff_biweight_h264_pixels_4_neon;
yading@10: 
yading@10:     c->h264_idct_add        = ff_h264_idct_add_neon;
yading@10:     c->h264_idct_dc_add     = ff_h264_idct_dc_add_neon;
yading@10:     c->h264_idct_add16      = ff_h264_idct_add16_neon;
yading@10:     c->h264_idct_add16intra = ff_h264_idct_add16intra_neon;
yading@10:     if (chroma_format_idc == 1)
yading@10:         c->h264_idct_add8   = ff_h264_idct_add8_neon;
yading@10:     c->h264_idct8_add       = ff_h264_idct8_add_neon;
yading@10:     c->h264_idct8_dc_add    = ff_h264_idct8_dc_add_neon;
yading@10:     c->h264_idct8_add4      = ff_h264_idct8_add4_neon;
yading@10:     }
yading@10: #endif // HAVE_NEON
yading@10: }
yading@10: 
yading@10: av_cold void ff_h264dsp_init_arm(H264DSPContext *c, const int bit_depth,
yading@10:                                  const int chroma_format_idc)
yading@10: {
yading@10:     int cpu_flags = av_get_cpu_flags();
yading@10: 
yading@10:     if (have_neon(cpu_flags))
yading@10:         ff_h264dsp_init_neon(c, bit_depth, chroma_format_idc);
yading@10: }