x264源代码简单分析：编码器主干部分-2

x264_slicetype_analyse()

//分析帧的类型（I、B、P）
void x264_slicetype_analyse( x264_t *h, int intra_minigop )
{
    x264_mb_analysis_t a;
    x264_frame_t *frames[X264_LOOKAHEAD_MAX+3] = { NULL, };
    int num_frames, orig_num_frames, keyint_limit, framecnt;
    int i_mb_count = NUM_MBS;
    int cost1p0, cost2p0, cost1b1, cost2p1;
    // 确定最大的搜索长度
    // 在我的调试当中， h->lookahead->next.i_size = 4
    int i_max_search = X264_MIN( h->lookahead->next.i_size, X264_LOOKAHEAD_MAX );
    int vbv_lookahead = h->param.rc.i_vbv_buffer_size && h->param.rc.i_lookahead;
    /* For determinism we should limit the search to the number of frames lookahead has for sure
     * in h->lookahead->next.list buffer, except at the end of stream.
     * For normal calls with (intra_minigop == 0) that is h->lookahead->i_slicetype_length + 1 frames.
     * And for I-frame calls (intra_minigop != 0) we already removed intra_minigop frames from there. */
    if( h->param.b_deterministic )
        i_max_search = X264_MIN( i_max_search, h->lookahead->i_slicetype_length + 1 - intra_minigop );
    int keyframe = !!intra_minigop;

    assert( h->frames.b_have_lowres );

    if( !h->lookahead->last_nonb )
        return;
    //frames[0]指向上一次的非B帧
    frames[0] = h->lookahead->last_nonb;
    //frames[] 依次指向 lookahead->next链表中的帧
    for( framecnt = 0; framecnt < i_max_search && h->lookahead->next.list[framecnt]->i_type == X264_TYPE_AUTO; framecnt++ )
        frames[framecnt+1] = h->lookahead->next.list[framecnt];

    x264_lowres_context_init( h, &a );

    if( !framecnt )
    {
        if( h->param.rc.b_mb_tree )
            x264_macroblock_tree( h, &a, frames, 0, keyframe );
        return;
    }

    keyint_limit = h->param.i_keyint_max - frames[0]->i_frame + h->lookahead->i_last_keyframe - 1;
    orig_num_frames = num_frames = h->param.b_intra_refresh ? framecnt : X264_MIN( framecnt, keyint_limit );

    /* This is important psy-wise: if we have a non-scenecut keyframe,
     * there will be significant visual artifacts if the frames just before
     * go down in quality due to being referenced less, despite it being
     * more RD-optimal. */
    if( (h->param.analyse.b_psy && h->param.rc.b_mb_tree) || vbv_lookahead )
        num_frames = framecnt;
    else if( h->param.b_open_gop && num_frames < framecnt )
        num_frames++;
    else if( num_frames == 0 )
    { 
        frames[1]->i_type = X264_TYPE_I;
        return;
    }

    int num_bframes = 0;
    int num_analysed_frames = num_frames;
    int reset_start;
    //通过scenecut()函数判断是否有场景切换，从而确定I帧
    if( h->param.i_scenecut_threshold && scenecut( h, &a, frames, 0, 1, 1, orig_num_frames, i_max_search ) )
    {
        frames[1]->i_type = X264_TYPE_I;
        return;
    }

#if H***E_OPENCL
    x264_opencl_slicetype_prep( h, frames, num_frames, a.i_lambda );
#endif
    //允许有B帧的时候
    if( h->param.i_bframe )
    {
        if( h->param.i_bframe_adaptive == X264_B_ADAPT_TRELLIS )
        {
            if( num_frames > 1 )
            {
                char best_paths[X264_BFRAME_MAX+1][X264_LOOKAHEAD_MAX+1] = {"","P"};
                int best_path_index = num_frames % (X264_BFRAME_MAX+1);

                /* Perform the frametype analysis. */
                for( int j = 2; j <= num_frames; j++ )
                    x264_slicetype_path( h, &a, frames, j, best_paths );

                num_bframes = strspn( best_paths[best_path_index], "B" );
                /* Load the results of the analysis into the frame types. */
                for( int j = 1; j < num_frames; j++ )
                    frames[j]->i_type = best_paths[best_path_index][j-1] == 'B' ? X264_TYPE_B : X264_TYPE_P;
            }
            frames[num_frames]->i_type = X264_TYPE_P;
        }
        else if( h->param.i_bframe_adaptive == X264_B_ADAPT_FAST )
        {
            for( int i = 0; i <= num_frames-2; )
            {
            	//i+2作为Ｐ帧编码的代价
                //注：i+2始终为P帧
            	cost2p1 = x264_slicetype_frame_cost( h, &a, frames, i+0, i+2, i+2, 1 );
                if( frames[i+2]->i_intra_mbs[2] > i_mb_count / 2 )
                {
                    frames[i+1]->i_type = X264_TYPE_P;
                    frames[i+2]->i_type = X264_TYPE_P;
                    i += 2;
                    continue;
                }

#if H***E_OPENCL
                if( h->param.b_opencl )
                {
                    int b_work_done = 0;
                    b_work_done |= x264_opencl_precalculate_frame_cost(h, frames, a.i_lambda, i+0, i+2, i+1 );
                    b_work_done |= x264_opencl_precalculate_frame_cost(h, frames, a.i_lambda, i+0, i+1, i+1 );
                    b_work_done |= x264_opencl_precalculate_frame_cost(h, frames, a.i_lambda, i+1, i+2, i+2 );
                    if( b_work_done )
                        x264_opencl_flush( h );
                }
#endif
                //计算代价
                //x264_slicetype_frame_cost(,,,p0,p1,b,)
                //p0 b p1
                //p1!=b为B帧，否则为P帧

                // i + 1 作为B帧编码的代价
                cost1b1 = x264_slicetype_frame_cost( h, &a, frames, i+0, i+2, i+1, 0 );
                // i + 1 作为P帧编码的代价
                cost1p0 = x264_slicetype_frame_cost( h, &a, frames, i+0, i+1, i+1, 0 );
                // i + 2 作为P帧编码的代价
                cost2p0 = x264_slicetype_frame_cost( h, &a, frames, i+1, i+2, i+2, 0 );
                //如果i+1作为P帧编码的代价　+ i+2作为P帧编码的代价
                //小于 i+1作为B帧编码的代价   + i+2作为P帧编码的代价
                if( cost1p0 + cost2p0 < cost1b1 + cost2p1 )
                {
                    //那么i+1将作为P帧编码
                    //然后直接continue
                    frames[i+1]->i_type = X264_TYPE_P;
                    i += 1;
                    continue;
                }

                // arbitrary and untuned
                #define INTER_THRESH 300
                #define P_SENS_BIAS (50 - h->param.i_bframe_bias)

                // i+1 将作为B帧编码
                frames[i+1]->i_type = X264_TYPE_B;

                int j;
                for( j = i+2; j <= X264_MIN( i+h->param.i_bframe, num_frames-1 ); j++ )
                {
                    int pthresh = X264_MAX(INTER_THRESH - P_SENS_BIAS * (j-i-1), INTER_THRESH/10);
                    // 预测j+1作为P帧编码代价
                    int pcost = x264_slicetype_frame_cost( h, &a, frames, i+0, j+1, j+1, 1 );
                    // 如果pcost 满足下述条件， 则确定了一个P帧，跳出循环
                    if( pcost > pthresh*i_mb_count || frames[j+1]->i_intra_mbs[j-i+1] > i_mb_count/3 )
                        break;
                    // 否则就是B帧
                    frames[j]->i_type = X264_TYPE_B;
                }
                // 将j帧确定为P帧
                frames[j]->i_type = X264_TYPE_P;
                i = j;
            }
            // 最后一帧确定为P帧
            frames[num_frames]->i_type = X264_TYPE_P;
            num_bframes = 0;
            // 确定有多少个B帧
            while( num_bframes < num_frames && frames[num_bframes+1]->i_type == X264_TYPE_B )
                num_bframes++;
        }
        else
        {
        	// 确定多少B帧
            num_bframes = X264_MIN(num_frames-1, h->param.i_bframe);
            // 每num_bframes + 1一个P帧， 其余皆为B帧
            for( int j = 1; j < num_frames; j++ )
                frames[j]->i_type = (j%(num_bframes+1)) ? X264_TYPE_B : X264_TYPE_P;
            // 最后一帧为P帧
            frames[num_frames]->i_type = X264_TYPE_P;
        }

        /* Check scenecut on the first minigop. */
        // 如果B帧中， 有帧有场景切换， 则改变其为P帧
        for( int j = 1; j < num_bframes+1; j++ )
            if( h->param.i_scenecut_threshold && scenecut( h, &a, frames, j, j+1, 0, orig_num_frames, i_max_search ) )
            {
                frames[j]->i_type = X264_TYPE_P;
                num_analysed_frames = j;
                break;
            }

        reset_start = keyframe ? 1 : X264_MIN( num_bframes+2, num_analysed_frames+1 );
    }
    else
    {
    	//h->param.i_bframe为 0
    	//则所有的帧皆为P帧
        for( int j = 1; j <= num_frames; j++ )
            frames[j]->i_type = X264_TYPE_P;
        reset_start = !keyframe + 1;
        num_bframes = 0;
    }

    /* Perform the actual macroblock tree analysis.
     * Don't go farther than the maximum keyframe interval; this helps in short GOPs. */
    if( h->param.rc.b_mb_tree )
        x264_macroblock_tree( h, &a, frames, X264_MIN(num_frames, h->param.i_keyint_max), keyframe );

    /* Enforce keyframe limit. */
    if( !h->param.b_intra_refresh )
        for( int i = keyint_limit+1; i <= num_frames; i += h->param.i_keyint_max )
        {
        	//迫使为I帧
            frames[i]->i_type = X264_TYPE_I;
            reset_start = X264_MIN( reset_start, i+1 );
            if( h->param.b_open_gop && h->param.b_bluray_compat )
                while( IS_X264_TYPE_B( frames[i-1]->i_type ) )
                    i--;
        }

    if( vbv_lookahead )
        x264_vbv_lookahead( h, &a, frames, num_frames, keyframe );

    /* Restore frametypes for all frames that haven't actually been decided yet. */
    for( int j = reset_start; j <= num_frames; j++ )
        frames[j]->i_type = X264_TYPE_AUTO;

#if H***E_OPENCL
    x264_opencl_slicetype_end( h );
#endif
}

x264_slicetype_frame_cost()

//一帧图像的开销
//x264_slicetype_frame_cost(,,,p0,p1,b,)
// p0 b p1
static int x264_slicetype_frame_cost( x264_t *h, x264_mb_analysis_t *a,
                                      x264_frame_t **frames, int p0, int p1, int b,
                                      int b_intra_penalty )
{
    int i_score = 0;
    int do_search[2];
    const x264_weight_t *w = x264_weight_none;
    x264_frame_t *fenc = frames[b];

    /* Check whether we already evaluated this frame
     * If we have tried this frame as P, then we have also tried
     * the preceding frames as B. (is this still true?) */
    /* Also check that we already calculated the row SATDs for the current frame. */
    //如果已经计算过就不用算了
    if( fenc->i_cost_est[b-p0][p1-b] >= 0 && (!h->param.rc.i_vbv_buffer_size || fenc->i_row_satds[b-p0][p1-b][0] != -1) )
        i_score = fenc->i_cost_est[b-p0][p1-b];
    else
    {
        int dist_scale_factor = 128;

        /* For each list, check to see whether we have lowres motion-searched this reference frame before. */
        do_search[0] = b != p0 && fenc->lowres_mvs[0][b-p0-1][0][0] == 0x7FFF;
        do_search[1] = b != p1 && fenc->lowres_mvs[1][p1-b-1][0][0] == 0x7FFF;
        if( do_search[0] )
        {
            if( h->param.analyse.i_weighted_pred && b == p1 )
            {
                x264_emms();
                x264_weights_analyse( h, fenc, frames[p0], 1 );
                w = fenc->weight[0];
            }
            fenc->lowres_mvs[0][b-p0-1][0][0] = 0;
        }
        if( do_search[1] ) fenc->lowres_mvs[1][p1-b-1][0][0] = 0;

        if( p1 != p0 )
            dist_scale_factor = ( ((b-p0) << 8) + ((p1-p0) >> 1) ) / (p1-p0);

        int output_buf_size = h->mb.i_mb_height + (NUM_INTS + PAD_SIZE) * h->param.i_lookahead_threads;
        int *output_inter[X264_LOOKAHEAD_THREAD_MAX+1];
        int *output_intra[X264_LOOKAHEAD_THREAD_MAX+1];
        output_inter[0] = h->scratch_buffer2;
        output_intra[0] = output_inter[0] + output_buf_size;

#if H***E_OPENCL
        if( h->param.b_opencl )
        {
            x264_opencl_lowres_init(h, fenc, a->i_lambda );
            if( do_search[0] )
            {
                x264_opencl_lowres_init( h, frames[p0], a->i_lambda );
                x264_opencl_motionsearch( h, frames, b, p0, 0, a->i_lambda, w );
            }
            if( do_search[1] )
            {
                x264_opencl_lowres_init( h, frames[p1], a->i_lambda );
                x264_opencl_motionsearch( h, frames, b, p1, 1, a->i_lambda, NULL );
            }
            if( b != p0 )
                x264_opencl_finalize_cost( h, a->i_lambda, frames, p0, p1, b, dist_scale_factor );
            x264_opencl_flush( h );

            i_score = fenc->i_cost_est[b-p0][p1-b];
        }
        else
#endif
        {
            if( h->param.i_lookahead_threads > 1 )
            {
                x264_slicetype_slice_t s[X264_LOOKAHEAD_THREAD_MAX];

                for( int i = 0; i < h->param.i_lookahead_threads; i++ )
                {
                    x264_t *t = h->lookahead_thread[i];

                    /* FIXME move this somewhere else */
                    t->mb.i_me_method = h->mb.i_me_method;
                    t->mb.i_subpel_refine = h->mb.i_subpel_refine;
                    t->mb.b_chroma_me = h->mb.b_chroma_me;

                    s[i] = (x264_slicetype_slice_t){ t, a, frames, p0, p1, b, dist_scale_factor, do_search, w,
                        output_inter[i], output_intra[i] };

                    t->i_threadslice_start = ((h->mb.i_mb_height *  i    + h->param.i_lookahead_threads/2) / h->param.i_lookahead_threads);
                    t->i_threadslice_end   = ((h->mb.i_mb_height * (i+1) + h->param.i_lookahead_threads/2) / h->param.i_lookahead_threads);

                    int thread_height = t->i_threadslice_end - t->i_threadslice_start;
                    int thread_output_size = thread_height + NUM_INTS;
                    memset( output_inter[i], 0, thread_output_size * sizeof(int) );
                    memset( output_intra[i], 0, thread_output_size * sizeof(int) );
                    output_inter[i][NUM_ROWS] = output_intra[i][NUM_ROWS] = thread_height;

                    output_inter[i+1] = output_inter[i] + thread_output_size + PAD_SIZE;
                    output_intra[i+1] = output_intra[i] + thread_output_size + PAD_SIZE;

                    x264_threadpool_run( h->lookaheadpool, (void*)x264_slicetype_slice_cost, &s[i] );
                }
                for( int i = 0; i < h->param.i_lookahead_threads; i++ )
                    x264_threadpool_wait( h->lookaheadpool, &s[i] );
            }
            else
            {
                h->i_threadslice_start = 0;
                h->i_threadslice_end = h->mb.i_mb_height;
                memset( output_inter[0], 0, (output_buf_size - PAD_SIZE) * sizeof(int) );
                memset( output_intra[0], 0, (output_buf_size - PAD_SIZE) * sizeof(int) );
                output_inter[0][NUM_ROWS] = output_intra[0][NUM_ROWS] = h->mb.i_mb_height;
                //作为参数的结构体
                x264_slicetype_slice_t s = (x264_slicetype_slice_t){ h, a, frames, p0, p1, b, dist_scale_factor, do_search, w,
                    output_inter[0], output_intra[0] };
                //一个slice的开销
                //输入输出参数都在s结构体中
                x264_slicetype_slice_cost( &s );
            }

            /* Sum up accumulators */
            if( b == p1 )
                fenc->i_intra_mbs[b-p0] = 0;
            if( !fenc->b_intra_calculated )
            {
                fenc->i_cost_est[0][0] = 0;
                fenc->i_cost_est_aq[0][0] = 0;
            }
            fenc->i_cost_est[b-p0][p1-b] = 0;
            fenc->i_cost_est_aq[b-p0][p1-b] = 0;

            int *row_satd_inter = fenc->i_row_satds[b-p0][p1-b];
            int *row_satd_intra = fenc->i_row_satds[0][0];
            for( int i = 0; i < h->param.i_lookahead_threads; i++ )
            {
            	//累加output_inter[]或output_intra[]
            	//这2个变量中存储了整帧的开销
                if( b == p1 )
                    fenc->i_intra_mbs[b-p0] += output_inter[i][INTRA_MBS];
                if( !fenc->b_intra_calculated )
                {
                	//帧内编码的代价
                    fenc->i_cost_est[0][0] += output_intra[i][COST_EST];
                    fenc->i_cost_est_aq[0][0] += output_intra[i][COST_EST_AQ];
                }

                //帧间编码的代价
                fenc->i_cost_est[b-p0][p1-b] += output_inter[i][COST_EST];
                fenc->i_cost_est_aq[b-p0][p1-b] += output_inter[i][COST_EST_AQ];

                if( h->param.rc.i_vbv_buffer_size )
                {
                    int row_count = output_inter[i][NUM_ROWS];
                    memcpy( row_satd_inter, output_inter[i] + NUM_INTS, row_count * sizeof(int) );
                    if( !fenc->b_intra_calculated )
                        memcpy( row_satd_intra, output_intra[i] + NUM_INTS, row_count * sizeof(int) );
                    row_satd_inter += row_count;
                    row_satd_intra += row_count;
                }
            }
            //一帧的开销
            i_score = fenc->i_cost_est[b-p0][p1-b];
            if( b != p1 )//B帧
                i_score = (uint64_t)i_score * 100 / (120 + h->param.i_bframe_bias);
            else
                fenc->b_intra_calculated = 1;

            fenc->i_cost_est[b-p0][p1-b] = i_score;
            x264_emms();
        }
    }

    if( b_intra_penalty )
    {
        // arbitrary penalty for I-blocks after B-frames
        int nmb = NUM_MBS;
        i_score += (uint64_t)i_score * fenc->i_intra_mbs[b-p0] / (nmb * 8);
    }
    //返回一帧的开销值
    return i_score;
}

x264_slicetype_slice_cost()

//一个slice的开销
static void x264_slicetype_slice_cost( x264_slicetype_slice_t *s )
{
    x264_t *h = s->h;

    /* Lowres lookahead goes backwards because the MVs are used as predictors in the main encode.
     * This considerably improves MV prediction overall. */

    /* The edge mbs seem to reduce the predictive quality of the
     * whole frame's score, but are needed for a spatial distribution. */
    int do_edges = h->param.rc.b_mb_tree || h->param.rc.i_vbv_buffer_size || h->mb.i_mb_width <= 2 || h->mb.i_mb_height <= 2;

    int start_y = X264_MIN( h->i_threadslice_end - 1, h->mb.i_mb_height - 2 + do_edges );
    int end_y = X264_MAX( h->i_threadslice_start, 1 - do_edges );
    int start_x = h->mb.i_mb_width - 2 + do_edges;
    int end_x = 1 - do_edges;

    //逐个计算每个MB的开销
    for( h->mb.i_mb_y = start_y; h->mb.i_mb_y >= end_y; h->mb.i_mb_y-- )
        for( h->mb.i_mb_x = start_x; h->mb.i_mb_x >= end_x; h->mb.i_mb_x-- )
            x264_slicetype_mb_cost( h, s->a, s->frames, s->p0, s->p1, s->b, s->dist_scale_factor,
                                    s->do_search, s->w, s->output_inter, s->output_intra );
}

x264_slicetype_mb_cost()

//一个MB的开销
static void x264_slicetype_mb_cost( x264_t *h, x264_mb_analysis_t *a,
                                    x264_frame_t **frames, int p0, int p1, int b,
                                    int dist_scale_factor, int do_search[2], const x264_weight_t *w,
                                    int *output_inter, int *output_intra )
{
    x264_frame_t *fref0 = frames[p0];
    x264_frame_t *fref1 = frames[p1];
    x264_frame_t *fenc  = frames[b];
    const int b_bidir = (b < p1);
    const int i_mb_x = h->mb.i_mb_x;
    const int i_mb_y = h->mb.i_mb_y;
    const int i_mb_stride = h->mb.i_mb_width;
    const int i_mb_xy = i_mb_x + i_mb_y * i_mb_stride;
    const int i_stride = fenc->i_stride_lowres;
    const int i_pel_offset = 8 * (i_mb_x + i_mb_y * i_stride);
    const int i_bipred_weight = h->param.analyse.b_weighted_bipred ? 64 - (dist_scale_factor>>2) : 32;
    int16_t (*fenc_mvs[2])[2] = { &fenc->lowres_mvs[0][b-p0-1][i_mb_xy], &fenc->lowres_mvs[1][p1-b-1][i_mb_xy] };
    int (*fenc_costs[2]) = { &fenc->lowres_mv_costs[0][b-p0-1][i_mb_xy], &fenc->lowres_mv_costs[1][p1-b-1][i_mb_xy] };
    int b_frame_score_mb = (i_mb_x > 0 && i_mb_x < h->mb.i_mb_width - 1 &&
                            i_mb_y > 0 && i_mb_y < h->mb.i_mb_height - 1) ||
                            h->mb.i_mb_width <= 2 || h->mb.i_mb_height <= 2;

    ALIGNED_ARRAY_16( pixel, pix1,[9*FDEC_STRIDE] );
    pixel *pix2 = pix1+8;
    x264_me_t m[2];
    int i_bcost = COST_MAX;
    int list_used = 0;
    /* A small, arbitrary bias to avoid VBV problems caused by zero-residual lookahead blocks. */
    int lowres_penalty = 4;
    //计算只涉及一个分量
    h->mb.pic.p_fenc[0] = h->mb.pic.fenc_buf;
    //从低分辨率（1/2线性内插）图像中拷贝数据
    h->mc.copy[PIXEL_8x8]( h->mb.pic.p_fenc[0], FENC_STRIDE, &fenc->lowres[0][i_pel_offset], i_stride, 8 );

    if( p0 == p1 )
        goto lowres_intra_mb;

    // no need for h->mb.mv_min[]
    h->mb.mv_limit_fpel[0][0] = -8*h->mb.i_mb_x - 4;
    h->mb.mv_limit_fpel[1][0] = 8*( h->mb.i_mb_width - h->mb.i_mb_x - 1 ) + 4;
    h->mb.mv_min_spel[0] = 4*( h->mb.mv_limit_fpel[0][0] - 8 );
    h->mb.mv_max_spel[0] = 4*( h->mb.mv_limit_fpel[1][0] + 8 );
    if( h->mb.i_mb_x >= h->mb.i_mb_width - 2 )
    {
        h->mb.mv_limit_fpel[0][1] = -8*h->mb.i_mb_y - 4;
        h->mb.mv_limit_fpel[1][1] = 8*( h->mb.i_mb_height - h->mb.i_mb_y - 1 ) + 4;
        h->mb.mv_min_spel[1] = 4*( h->mb.mv_limit_fpel[0][1] - 8 );
        h->mb.mv_max_spel[1] = 4*( h->mb.mv_limit_fpel[1][1] + 8 );
    }

#define LOAD_HPELS_LUMA(dst, src) \
    { \
        (dst)[0] = &(src)[0][i_pel_offset]; \
        (dst)[1] = &(src)[1][i_pel_offset]; \
        (dst)[2] = &(src)[2][i_pel_offset]; \
        (dst)[3] = &(src)[3][i_pel_offset]; \
    }
#define LOAD_WPELS_LUMA(dst,src) \
    (dst) = &(src)[i_pel_offset];

#define CLIP_MV( mv ) \
    { \
        mv[0] = x264_clip3( mv[0], h->mb.mv_min_spel[0], h->mb.mv_max_spel[0] ); \
        mv[1] = x264_clip3( mv[1], h->mb.mv_min_spel[1], h->mb.mv_max_spel[1] ); \
    }
#define TRY_BIDIR( mv0, mv1, penalty ) \
    { \
        int i_cost; \
        if( h->param.analyse.i_subpel_refine <= 1 ) \
        { \
            int hpel_idx1 = (((mv0)[0]&2)>>1) + ((mv0)[1]&2); \
            int hpel_idx2 = (((mv1)[0]&2)>>1) + ((mv1)[1]&2); \
            pixel *src1 = m[0].p_fref[hpel_idx1] + ((mv0)[0]>>2) + ((mv0)[1]>>2) * m[0].i_stride[0]; \
            pixel *src2 = m[1].p_fref[hpel_idx2] + ((mv1)[0]>>2) + ((mv1)[1]>>2) * m[1].i_stride[0]; \
            h->mc.avg[PIXEL_8x8]( pix1, 16, src1, m[0].i_stride[0], src2, m[1].i_stride[0], i_bipred_weight ); \
        } \
        else \
        { \
            intptr_t stride1 = 16, stride2 = 16; \
            pixel *src1, *src2; \
            src1 = h->mc.get_ref( pix1, &stride1, m[0].p_fref, m[0].i_stride[0], \
                                  (mv0)[0], (mv0)[1], 8, 8, w ); \
            src2 = h->mc.get_ref( pix2, &stride2, m[1].p_fref, m[1].i_stride[0], \
                                  (mv1)[0], (mv1)[1], 8, 8, w ); \
            h->mc.avg[PIXEL_8x8]( pix1, 16, src1, stride1, src2, stride2, i_bipred_weight ); \
        } \
        i_cost = penalty * a->i_lambda + h->pixf.mbcmp[PIXEL_8x8]( \
                           m[0].p_fenc[0], FENC_STRIDE, pix1, 16 ); \
        COPY2_IF_LT( i_bcost, i_cost, list_used, 3 ); \
    }

    //帧间编码（后面还有帧内编码）

    //处理m[0]
    m[0].i_pixel = PIXEL_8x8;
    m[0].p_cost_mv = a->p_cost_mv;
    m[0].i_stride[0] = i_stride;
    m[0].p_fenc[0] = h->mb.pic.p_fenc[0];
    m[0].weight = w;
    m[0].i_ref = 0;
    //加载1/2插值像素点
    LOAD_HPELS_LUMA( m[0].p_fref, fref0->lowres );
    m[0].p_fref_w = m[0].p_fref[0];
    if( w[0].weightfn )
        LOAD_WPELS_LUMA( m[0].p_fref_w, fenc->weighted[0] );
    //双线预测，处理m[1]
    if( b_bidir )
    {
        int16_t *mvr = fref1->lowres_mvs[0][p1-p0-1][i_mb_xy];
        ALIGNED_ARRAY_8( int16_t, dmv,[2],[2] );

        m[1].i_pixel = PIXEL_8x8;
        m[1].p_cost_mv = a->p_cost_mv;
        m[1].i_stride[0] = i_stride;
        m[1].p_fenc[0] = h->mb.pic.p_fenc[0];
        m[1].i_ref = 0;
        m[1].weight = x264_weight_none;
        LOAD_HPELS_LUMA( m[1].p_fref, fref1->lowres );
        m[1].p_fref_w = m[1].p_fref[0];

        dmv[0][0] = ( mvr[0] * dist_scale_factor + 128 ) >> 8;
        dmv[0][1] = ( mvr[1] * dist_scale_factor + 128 ) >> 8;
        dmv[1][0] = dmv[0][0] - mvr[0];
        dmv[1][1] = dmv[0][1] - mvr[1];
        CLIP_MV( dmv[0] );
        CLIP_MV( dmv[1] );
        if( h->param.analyse.i_subpel_refine <= 1 )
            M64( dmv ) &= ~0x0001000100010001ULL; /* mv & ~1 */

        //双向预测，其中包含了mc.avg[PIXEL_8x8]()
        TRY_BIDIR( dmv[0], dmv[1], 0 );
        if( M64( dmv ) )
        {
            int i_cost;
            h->mc.avg[PIXEL_8x8]( pix1, 16, m[0].p_fref[0], m[0].i_stride[0], m[1].p_fref[0], m[1].i_stride[0], i_bipred_weight );
            i_cost = h->pixf.mbcmp[PIXEL_8x8]( m[0].p_fenc[0], FENC_STRIDE, pix1, 16 );
            COPY2_IF_LT( i_bcost, i_cost, list_used, 3 );
        }
    }

    for( int l = 0; l < 1 + b_bidir; l++ )
    {
        if( do_search[l] )
        {
            int i_mvc = 0;
            int16_t (*fenc_mv)[2] = fenc_mvs[l];
            ALIGNED_4( int16_t mvc[4][2] );

            /* Reverse-order MV prediction. */
            M32( mvc[0] ) = 0;
            M32( mvc[2] ) = 0;
#define MVC(mv) { CP32( mvc[i_mvc], mv ); i_mvc++; }
            if( i_mb_x < h->mb.i_mb_width - 1 )
                MVC( fenc_mv[1] );
            if( i_mb_y < h->i_threadslice_end - 1 )
            {
                MVC( fenc_mv[i_mb_stride] );
                if( i_mb_x > 0 )
                    MVC( fenc_mv[i_mb_stride-1] );
                if( i_mb_x < h->mb.i_mb_width - 1 )
                    MVC( fenc_mv[i_mb_stride+1] );
            }
#undef MVC
            if( i_mvc <= 1 )
                CP32( m[l].mvp, mvc[0] );
            else
                x264_median_mv( m[l].mvp, mvc[0], mvc[1], mvc[2] );

            /* Fast skip for cases of near-zero residual.  Shortcut: don't bother except in the mv0 case,
             * since anything else is likely to have enough residual to not trigger the skip. */
            if( !M32( m[l].mvp ) )
            {
                m[l].cost = h->pixf.mbcmp[PIXEL_8x8]( m[l].p_fenc[0], FENC_STRIDE, m[l].p_fref[0], m[l].i_stride[0] );
                if( m[l].cost < 64 )
                {
                    M32( m[l].mv ) = 0;
                    goto skip_motionest;
                }
            }
            //运动搜索，开销存在m[l].cost中
            x264_me_search( h, &m[l], mvc, i_mvc );
            m[l].cost -= a->p_cost_mv[0]; // remove mvcost from skip mbs
            if( M32( m[l].mv ) )
                m[l].cost += 5 * a->i_lambda;

skip_motionest:
            CP32( fenc_mvs[l], m[l].mv );
            *fenc_costs[l] = m[l].cost;
        }
        else
        {
            CP32( m[l].mv, fenc_mvs[l] );
            m[l].cost = *fenc_costs[l];
        }
        //如果更小就拷贝
        //帧间编码开销，存储于i_bcost
        COPY2_IF_LT( i_bcost, m[l].cost, list_used, l+1 );
    }

    if( b_bidir && ( M32( m[0].mv ) || M32( m[1].mv ) ) )
        TRY_BIDIR( m[0].mv, m[1].mv, 5 );

lowres_intra_mb:
	//帧内编码
    if( !fenc->b_intra_calculated )
    {
        ALIGNED_ARRAY_16( pixel, edge,[36] );
        pixel *pix = &pix1[8+FDEC_STRIDE];
        pixel *src = &fenc->lowres[0][i_pel_offset];
        const int intra_penalty = 5 * a->i_lambda;
        int satds[3];
        int pixoff = 4 / sizeof(pixel);

        /* Avoid store forwarding stalls by writing larger chunks */
        memcpy( pix-FDEC_STRIDE, src-i_stride, 16 * sizeof(pixel) );
        for( int i = -1; i < 8; i++ )
            M32( &pix[i*FDEC_STRIDE-pixoff] ) = M32( &src[i*i_stride-pixoff] );

        //8x8块的SAD/SATD计算
        //x3打表计算了V，H，DC三种模式，开销存储在satds[3]数组的3个元素中
        h->pixf.intra_mbcmp_x3_8x8c( h->mb.pic.p_fenc[0], pix, satds );
        //帧内编码开销，存储于i_icost
        int i_icost = X264_MIN3( satds[0], satds[1], satds[2] );

        if( h->param.analyse.i_subpel_refine > 1 )
        {
            h->predict_8x8c[I_PRED_CHROMA_P]( pix );
            int satd = h->pixf.mbcmp[PIXEL_8x8]( pix, FDEC_STRIDE, h->mb.pic.p_fenc[0], FENC_STRIDE );
            i_icost = X264_MIN( i_icost, satd );
            h->predict_8x8_filter( pix, edge, ALL_NEIGHBORS, ALL_NEIGHBORS );
            for( int i = 3; i < 9; i++ )
            {
                h->predict_8x8[i]( pix, edge );
                satd = h->pixf.mbcmp[PIXEL_8x8]( pix, FDEC_STRIDE, h->mb.pic.p_fenc[0], FENC_STRIDE );
                i_icost = X264_MIN( i_icost, satd );
            }
        }

        i_icost += intra_penalty + lowres_penalty;
        //存一下
        fenc->i_intra_cost[i_mb_xy] = i_icost;
        int i_icost_aq = i_icost;
        if( h->param.rc.i_aq_mode )
            i_icost_aq = (i_icost_aq * fenc->i_inv_qscale_factor[i_mb_xy] + 128) >> 8;
        output_intra[ROW_SATD] += i_icost_aq;
        if( b_frame_score_mb )
        {
        	//累加。[COST_EST]用于整帧的开销计算
            output_intra[COST_EST] += i_icost;
            output_intra[COST_EST_AQ] += i_icost_aq;
        }
    }
    i_bcost += lowres_penalty;

    /* forbid intra-mbs in B-frames, because it's rare and not worth checking */
    /* FIXME: Should we still forbid them now that we cache intra scores? */
    if( !b_bidir )
    {
        int i_icost = fenc->i_intra_cost[i_mb_xy];
        //帧内开销比帧间更小，b_intra就会取1
        int b_intra = i_icost < i_bcost;
        if( b_intra )
        {
        	//赋值给i_bcost
            i_bcost = i_icost;
            list_used = 0;
        }
        if( b_frame_score_mb )
            output_inter[INTRA_MBS] += b_intra;//[INTRA_MBS]统计有多少个帧内模式的宏块
    }

    /* In an I-frame, we've already added the results above in the intra section. */
    if( p0 != p1 )
    {
        int i_bcost_aq = i_bcost;
        if( h->param.rc.i_aq_mode )
            i_bcost_aq = (i_bcost_aq * fenc->i_inv_qscale_factor[i_mb_xy] + 128) >> 8;
        output_inter[ROW_SATD] += i_bcost_aq;
        if( b_frame_score_mb )
        {
            /* Don't use AQ-weighted costs for slicetype decision, only for ratecontrol. */
        	//累加。[COST_EST]用于整帧的开销计算
            output_inter[COST_EST] += i_bcost;
            output_inter[COST_EST_AQ] += i_bcost_aq;
        }
    }
    //存储开销i_bcost
    fenc->lowres_costs[b-p0][p1-b][i_mb_xy] = X264_MIN( i_bcost, LOWRES_COST_MASK ) + (list_used << LOWRES_COST_SHIFT);
}
#undef TRY_BIDIR

x264_frame_shift()

//从队列的头部取出一帧
x264_frame_t *x264_frame_shift( x264_frame_t **list )
{
    x264_frame_t *frame = list[0];
    int i;
    for( i = 0; list[i]; i++ )
        list[i] = list[i+1];
    assert(frame);
    return frame;
}

x264_reference_update()

//更新参考帧队列，若为非参考B帧则不更新
//重建帧移植参考帧列表，新建一个重建帧
static inline int x264_reference_update( x264_t *h )
{
	//如果不是被参考的帧
    if( !h->fdec->b_kept_as_ref )
    {
        if( h->i_thread_frames > 1 )
        {
            x264_frame_push_unused( h, h->fdec );
            h->fdec = x264_frame_pop_unused( h, 1 );
            if( !h->fdec )
                return -1;
        }
        return 0;
    }

    /* apply mmco from previous frame. */
    for( int i = 0; i < h->sh.i_mmco_command_count; i++ )
        for( int j = 0; h->frames.reference[j]; j++ )
            if( h->frames.reference[j]->i_poc == h->sh.mmco[i].i_poc )
                x264_frame_push_unused( h, x264_frame_shift( &h->frames.reference[j] ) );

    /* move frame in the buffer */
    //重建帧加入参考帧列表
    x264_frame_push( h->frames.reference, h->fdec );
    //列表满了，则要移除1帧
    if( h->frames.reference[h->sps->i_num_ref_frames] )
        x264_frame_push_unused( h, x264_frame_shift( h->frames.reference ) );
    //重新初始化重建帧fdec
    h->fdec = x264_frame_pop_unused( h, 1 );
    if( !h->fdec )
        return -1;
    return 0;
}

x264_reference_reset()

//清空所有参考帧
static inline void x264_reference_reset( x264_t *h )
{
	//把frames.reference[]中所有帧移动到frames.unused[]
    while( h->frames.reference[0] )
        x264_frame_push_unused( h, x264_frame_pop( h->frames.reference ) );
    h->fdec->i_poc =
    h->fenc->i_poc = 0;
}

x264_slice_init()

//创建Slice Header
static inline void x264_slice_init( x264_t *h, int i_nal_type, int i_global_qp )
{
    /* ------------------------ Create slice header  ----------------------- */
    if( i_nal_type == NAL_SLICE_IDR )
    {
    	//I帧

    	//对x264_slice_header_t进行赋值
        x264_slice_header_init( h, &h->sh, h->sps, h->pps, h->i_idr_pic_id, h->i_frame_num, i_global_qp );

        /* alternate id */
        if( h->param.i_avcintra_class )
        {
            switch( h->i_idr_pic_id )
            {
                case 5:
                    h->i_idr_pic_id = 3;
                    break;
                case 3:
                    h->i_idr_pic_id = 4;
                    break;
                case 4:
                default:
                    h->i_idr_pic_id = 5;
                    break;
            }
        }
        else
            h->i_idr_pic_id ^= 1;
    }
    else
    {
    	//非IDR帧
        x264_slice_header_init( h, &h->sh, h->sps, h->pps, -1, h->i_frame_num, i_global_qp );
        //参考帧列表
        h->sh.i_num_ref_idx_l0_active = h->i_ref[0] <= 0 ? 1 : h->i_ref[0];
        h->sh.i_num_ref_idx_l1_active = h->i_ref[1] <= 0 ? 1 : h->i_ref[1];
        if( h->sh.i_num_ref_idx_l0_active != h->pps->i_num_ref_idx_l0_default_active ||
            (h->sh.i_type == SLICE_TYPE_B && h->sh.i_num_ref_idx_l1_active != h->pps->i_num_ref_idx_l1_default_active) )
        {
            h->sh.b_num_ref_idx_override = 1;
        }
    }

    if( h->fenc->i_type == X264_TYPE_BREF && h->param.b_bluray_compat && h->sh.i_mmco_command_count )
    {
        h->b_sh_backup = 1;
        h->sh_backup = h->sh;
    }

    h->fdec->i_frame_num = h->sh.i_frame_num;

    if( h->sps->i_poc_type == 0 )
    {
        h->sh.i_poc = h->fdec->i_poc;
        if( PARAM_INTERLACED )
        {
            h->sh.i_delta_poc_bottom = h->param.b_tff ? 1 : -1;
            h->sh.i_poc += h->sh.i_delta_poc_bottom == -1;
        }
        else
            h->sh.i_delta_poc_bottom = 0;
        h->fdec->i_delta_poc[0] = h->sh.i_delta_poc_bottom == -1;
        h->fdec->i_delta_poc[1] = h->sh.i_delta_poc_bottom ==  1;
    }
    else
    {
        /* Nothing to do ? */
    }
    //主要对mb结构体赋初值
    x264_macroblock_slice_init( h );
}

x264_slice_header_init()

/* Fill "default" values */
//对x264_slice_header_t进行赋值
static void x264_slice_header_init( x264_t *h, x264_slice_header_t *sh,
                                    x264_sps_t *sps, x264_pps_t *pps,
                                    int i_idr_pic_id, int i_frame, int i_qp )
{
    x264_param_t *param = &h->param;

    /* First we fill all fields */
    sh->sps = sps;
    sh->pps = pps;

    sh->i_first_mb  = 0;
    sh->i_last_mb   = h->mb.i_mb_count - 1;
    sh->i_pps_id    = pps->i_id;

    sh->i_frame_num = i_frame;

    sh->b_mbaff = PARAM_INTERLACED;
    sh->b_field_pic = 0;    /* no field support for now */
    sh->b_bottom_field = 0; /* not yet used */

    sh->i_idr_pic_id = i_idr_pic_id;

    /* poc stuff, fixed later */
    sh->i_poc = 0;
    sh->i_delta_poc_bottom = 0;
    sh->i_delta_poc[0] = 0;
    sh->i_delta_poc[1] = 0;

    sh->i_redundant_pic_cnt = 0;

    h->mb.b_direct_auto_write = h->param.analyse.i_direct_mv_pred == X264_DIRECT_PRED_AUTO
                                && h->param.i_bframe
                                && ( h->param.rc.b_stat_write || !h->param.rc.b_stat_read );

    if( !h->mb.b_direct_auto_read && sh->i_type == SLICE_TYPE_B )
    {
        if( h->fref[1][0]->i_poc_l0ref0 == h->fref[0][0]->i_poc )
        {
            if( h->mb.b_direct_auto_write )
                sh->b_direct_spatial_mv_pred = ( h->stat.i_direct_score[1] > h->stat.i_direct_score[0] );
            else
                sh->b_direct_spatial_mv_pred = ( param->analyse.i_direct_mv_pred == X264_DIRECT_PRED_SPATIAL );
        }
        else
        {
            h->mb.b_direct_auto_write = 0;
            sh->b_direct_spatial_mv_pred = 1;
        }
    }
    /* else b_direct_spatial_mv_pred was read from the 2pass statsfile */

    sh->b_num_ref_idx_override = 0;
    sh->i_num_ref_idx_l0_active = 1;
    sh->i_num_ref_idx_l1_active = 1;

    sh->b_ref_pic_list_reordering[0] = h->b_ref_reorder[0];
    sh->b_ref_pic_list_reordering[1] = h->b_ref_reorder[1];

    /* If the ref list isn't in the default order, construct reordering header */
    for( int list = 0; list < 2; list++ )
    {
        if( sh->b_ref_pic_list_reordering[list] )
        {
            int pred_frame_num = i_frame;
            for( int i = 0; i < h->i_ref[list]; i++ )
            {
                int diff = h->fref[list][i]->i_frame_num - pred_frame_num;
                sh->ref_pic_list_order[list][i].idc = ( diff > 0 );
                sh->ref_pic_list_order[list][i].arg = (abs(diff) - 1) & ((1 << sps->i_log2_max_frame_num) - 1);
                pred_frame_num = h->fref[list][i]->i_frame_num;
            }
        }
    }

    sh->i_cabac_init_idc = param->i_cabac_init_idc;

    sh->i_qp = SPEC_QP(i_qp);
    sh->i_qp_delta = sh->i_qp - pps->i_pic_init_qp;
    sh->b_sp_for_swidth = 0;
    sh->i_qs_delta = 0;

    int deblock_thresh = i_qp + 2 * X264_MIN(param->i_deblocking_filter_alphac0, param->i_deblocking_filter_beta);
    /* If effective qp <= 15, deblocking would have no effect anyway */
    if( param->b_deblocking_filter && (h->mb.b_variable_qp || 15 < deblock_thresh ) )
        sh->i_disable_deblocking_filter_idc = param->b_sliced_threads ? 2 : 0;
    else
        sh->i_disable_deblocking_filter_idc = 1;
    sh->i_alpha_c0_offset = param->i_deblocking_filter_alphac0 << 1;
    sh->i_beta_offset = param->i_deblocking_filter_beta << 1;
}

x264_slices_write()

//真正的编码——编码1个图像帧
//注意“slice”后面有一个“s”
//它其中又调用了一个x264_slice_write()
//这一点要区分开
static void *x264_slices_write( x264_t *h )
{
    int i_slice_num = 0;
    int last_thread_mb = h->sh.i_last_mb;

    /* init stats */
    memset( &h->stat.frame, 0, sizeof(h->stat.frame) );
    h->mb.b_reencode_mb = 0;
    //循环每一个slice（一幅图像可以由多个Slice构成）
    while( h->sh.i_first_mb + SLICE_MBAFF*h->mb.i_mb_stride <= last_thread_mb )
    {
        h->sh.i_last_mb = last_thread_mb;
        if( !i_slice_num || !x264_frame_new_slice( h, h->fdec ) )
        {
            if( h->param.i_slice_max_mbs )
            {
                if( SLICE_MBAFF )
                {
                    // convert first to mbaff form, add slice-max-mbs, then convert back to normal form
                    int last_mbaff = 2*(h->sh.i_first_mb % h->mb.i_mb_width)
                        + h->mb.i_mb_width*(h->sh.i_first_mb / h->mb.i_mb_width)
                        + h->param.i_slice_max_mbs - 1;
                    int last_x = (last_mbaff % (2*h->mb.i_mb_width))/2;
                    int last_y = (last_mbaff / (2*h->mb.i_mb_width))*2 + 1;
                    h->sh.i_last_mb = last_x + h->mb.i_mb_stride*last_y;
                }
                else
                {
                    h->sh.i_last_mb = h->sh.i_first_mb + h->param.i_slice_max_mbs - 1;
                    if( h->sh.i_last_mb < last_thread_mb && last_thread_mb - h->sh.i_last_mb < h->param.i_slice_min_mbs )
                        h->sh.i_last_mb = last_thread_mb - h->param.i_slice_min_mbs;
                }
                i_slice_num++;
            }
            else if( h->param.i_slice_count && !h->param.b_sliced_threads )
            {
                int height = h->mb.i_mb_height >> PARAM_INTERLACED;
                int width = h->mb.i_mb_width << PARAM_INTERLACED;
                i_slice_num++;
                h->sh.i_last_mb = (height * i_slice_num + h->param.i_slice_count/2) / h->param.i_slice_count * width - 1;
            }
        }
        h->sh.i_last_mb = X264_MIN( h->sh.i_last_mb, last_thread_mb );
        //真正的编码——编码1个Slice
        //x264_stack_align()应该是平台优化过程中内存对齐的工作
        //实际上就是调用x264_slice_write()
        if( x264_stack_align( x264_slice_write, h ) )
            goto fail;
        //注意这里对i_first_mb进行了赋值
        h->sh.i_first_mb = h->sh.i_last_mb + 1;
        // if i_first_mb is not the last mb in a row then go to the next mb in MBAFF order
        if( SLICE_MBAFF && h->sh.i_first_mb % h->mb.i_mb_width )
            h->sh.i_first_mb -= h->mb.i_mb_stride;
    }

    return (void *)0;

fail:
    /* Tell other threads we're done, so they wouldn't wait for it */
    if( h->param.b_sliced_threads )
        x264_threadslice_cond_broadcast( h, 2 );
    return (void *)-1;
}

x264_slice_write()

x264_encoder_frame_end()

//结束的时候做一些处理，记录一些统计信息
//pp_nal：输出的NALU
//pic_out：输出的重建帧
static int x264_encoder_frame_end( x264_t *h, x264_t *thread_current,
                                   x264_nal_t **pp_nal, int *pi_nal,
                                   x264_picture_t *pic_out )
{
    char psz_message[80];

    if( !h->param.b_sliced_threads && h->b_thread_active )
    {
        h->b_thread_active = 0;
        if( (intptr_t)x264_threadpool_wait( h->threadpool, h ) )
            return -1;
    }
    if( !h->out.i_nal )
    {
        pic_out->i_type = X264_TYPE_AUTO;
        return 0;
    }

    x264_emms();

    /* generate buffering period sei and insert it into place */
    if( h->i_thread_frames > 1 && h->fenc->b_keyframe && h->sps->vui.b_nal_hrd_parameters_present )
    {
        x264_hrd_fullness( h );
        x264_nal_start( h, NAL_SEI, NAL_PRIORITY_DISPOSABLE );
        x264_sei_buffering_period_write( h, &h->out.bs );
        if( x264_nal_end( h ) )
           return -1;
        /* buffering period sei must follow AUD, SPS and PPS and precede all other SEIs */
        int idx = 0;
        while( h->out.nal[idx].i_type == NAL_AUD ||
               h->out.nal[idx].i_type == NAL_SPS ||
               h->out.nal[idx].i_type == NAL_PPS )
            idx++;
        x264_nal_t nal_tmp = h->out.nal[h->out.i_nal-1];
        memmove( &h->out.nal[idx+1], &h->out.nal[idx], (h->out.i_nal-idx-1)*sizeof(x264_nal_t) );
        h->out.nal[idx] = nal_tmp;
    }
    //封装一帧数据对应的NALU.
    //例如给NALU添加起始码0x00000001
    int frame_size = x264_encoder_encapsulate_nals( h, 0 );
    if( frame_size < 0 )
        return -1;

    /* Set output picture properties */
    //pic_out为x264_picture_t类型结构体。是libx264对外的结构体
    //fenc,fdec是x264_frame_t类型结构体。是libx264的内部结构体
    pic_out->i_type = h->fenc->i_type;

    pic_out->b_keyframe = h->fenc->b_keyframe;
    pic_out->i_pic_struct = h->fenc->i_pic_struct;

    pic_out->i_pts = h->fdec->i_pts;
    pic_out->i_dts = h->fdec->i_dts;

    if( pic_out->i_pts < pic_out->i_dts )
        x264_log( h, X264_LOG_WARNING, "invalid DTS: PTS is less than DTS\n" );

    pic_out->opaque = h->fenc->opaque;

    pic_out->img.i_csp = h->fdec->i_csp;
#if HIGH_BIT_DEPTH
    pic_out->img.i_csp |= X264_CSP_HIGH_DEPTH;
#endif
    pic_out->img.i_plane = h->fdec->i_plane;
    //图像数据
    for( int i = 0; i < pic_out->img.i_plane; i++ )
    {
        pic_out->img.i_stride[i] = h->fdec->i_stride[i] * sizeof(pixel);
        pic_out->img.plane[i] = (uint8_t*)h->fdec->plane[i];
    }
    //回收用过的编码帧fenc
    x264_frame_push_unused( thread_current, h->fenc );

    /* ---------------------- Update encoder state ------------------------- */

    /* update rc */
    int filler = 0;
    if( x264_ratecontrol_end( h, frame_size * 8, &filler ) < 0 )
        return -1;

    pic_out->hrd_timing = h->fenc->hrd_timing;
    pic_out->prop.f_crf_avg = h->fdec->f_crf_avg;

    /* Filler in ***C-Intra mode is written as zero bytes to the last slice
     * We don't know the size of the last slice until encapsulation so we add filler to the encapsulated NAL */
    if( h->param.i_avcintra_class )
    {
        x264_t *h0 = h->thread[0];
        int ret = x264_check_encapsulated_buffer( h, h0, h->out.i_nal, frame_size, frame_size + filler );
        if( ret < 0 )
            return -1;
        memset( h->out.nal[0].p_payload + frame_size, 0, filler );
        h->out.nal[h->out.i_nal-1].i_payload += filler;
        h->out.nal[h->out.i_nal-1].i_padding = filler;
        frame_size += filler;
    }
    else
    {
        while( filler > 0 )
        {
            int f, overhead;
            overhead = (FILLER_OVERHEAD - h->param.b_annexb);
            if( h->param.i_slice_max_size && filler > h->param.i_slice_max_size )
            {
                int next_size = filler - h->param.i_slice_max_size;
                int overflow = X264_MAX( overhead - next_size, 0 );
                f = h->param.i_slice_max_size - overhead - overflow;
            }
            else
                f = X264_MAX( 0, filler - overhead );

            if( x264_bitstream_check_buffer_filler( h, f ) )
                return -1;
            x264_nal_start( h, NAL_FILLER, NAL_PRIORITY_DISPOSABLE );
            x264_filler_write( h, &h->out.bs, f );
            if( x264_nal_end( h ) )
                return -1;
            int total_size = x264_encoder_encapsulate_nals( h, h->out.i_nal-1 );
            if( total_size < 0 )
                return -1;
            frame_size += total_size;
            filler -= total_size;
        }
    }

    /* End bitstream, set output  */
    *pi_nal = h->out.i_nal;
    *pp_nal = h->out.nal;

    h->out.i_nal = 0;

    x264_noise_reduction_update( h );

    /* ---------------------- Compute/Print statistics --------------------- */
    x264_thread_sync_stat( h, h->thread[0] );

    /* Slice stat */
    //stat中存储了统计信息
    //帧数+1 （根据类型）
    h->stat.i_frame_count[h->sh.i_type]++;
    //帧大小
    h->stat.i_frame_size[h->sh.i_type] += frame_size;
    h->stat.f_frame_qp[h->sh.i_type] += h->fdec->f_qp_avg_aq;
    //统计MB个数，把不同类型的累加起来
    for( int i = 0; i < X264_MBTYPE_MAX; i++ )
        h->stat.i_mb_count[h->sh.i_type][i] += h->stat.frame.i_mb_count[i];
    for( int i = 0; i < X264_PARTTYPE_MAX; i++ )
        h->stat.i_mb_partition[h->sh.i_type][i] += h->stat.frame.i_mb_partition[i];
    for( int i = 0; i < 2; i++ )
        h->stat.i_mb_count_8x8dct[i] += h->stat.frame.i_mb_count_8x8dct[i];
    for( int i = 0; i < 6; i++ )
        h->stat.i_mb_cbp[i] += h->stat.frame.i_mb_cbp[i];
    for( int i = 0; i < 4; i++ )
        for( int j = 0; j < 13; j++ )
            h->stat.i_mb_pred_mode[i][j] += h->stat.frame.i_mb_pred_mode[i][j];
    if( h->sh.i_type != SLICE_TYPE_I )
        for( int i_list = 0; i_list < 2; i_list++ )
            for( int i = 0; i < X264_REF_MAX*2; i++ )
                h->stat.i_mb_count_ref[h->sh.i_type][i_list][i] += h->stat.frame.i_mb_count_ref[i_list][i];
    for( int i = 0; i < 3; i++ )
        h->stat.i_mb_field[i] += h->stat.frame.i_mb_field[i];
    if( h->sh.i_type == SLICE_TYPE_P && h->param.analyse.i_weighted_pred >= X264_WEIGHTP_SIMPLE )
    {
        h->stat.i_wpred[0] += !!h->sh.weight[0][0].weightfn;
        h->stat.i_wpred[1] += !!h->sh.weight[0][1].weightfn || !!h->sh.weight[0][2].weightfn;
    }
    if( h->sh.i_type == SLICE_TYPE_B )
    {
        h->stat.i_direct_frames[ h->sh.b_direct_spatial_mv_pred ] ++;
        if( h->mb.b_direct_auto_write )
        {
            //FIXME somewhat arbitrary time constants
            if( h->stat.i_direct_score[0] + h->stat.i_direct_score[1] > h->mb.i_mb_count )
                for( int i = 0; i < 2; i++ )
                    h->stat.i_direct_score[i] = h->stat.i_direct_score[i] * 9/10;
            for( int i = 0; i < 2; i++ )
                h->stat.i_direct_score[i] += h->stat.frame.i_direct_score[i];
        }
    }
    else
        h->stat.i_consecutive_bframes[h->fenc->i_bframes]++;

    psz_message[0] = '\0';
    double dur = h->fenc->f_duration;
    h->stat.f_frame_duration[h->sh.i_type] += dur;

    //需要计算PSNR
    if( h->param.analyse.b_psnr )
    {
    	//SSD（Sum of Squared Difference）即差值的平方和
        int64_t ssd[3] =
        {
            h->stat.frame.i_ssd[0],
            h->stat.frame.i_ssd[1],
            h->stat.frame.i_ssd[2],
        };
        int luma_size = h->param.i_width * h->param.i_height;
        int chroma_size = CHROMA_SIZE( luma_size );

        //SSD是已经在“滤波”环节计算过的
        //SSD简单换算成PSNR，调用x264_psnr()
        pic_out->prop.f_psnr[0] = x264_psnr( ssd[0], luma_size );
        pic_out->prop.f_psnr[1] = x264_psnr( ssd[1], chroma_size );
        pic_out->prop.f_psnr[2] = x264_psnr( ssd[2], chroma_size );
        //平均值
        pic_out->prop.f_psnr_avg = x264_psnr( ssd[0] + ssd[1] + ssd[2], luma_size + chroma_size*2 );
        //mean系列的需要累加
        h->stat.f_ssd_global[h->sh.i_type]   += dur * (ssd[0] + ssd[1] + ssd[2]);
        h->stat.f_psnr_average[h->sh.i_type] += dur * pic_out->prop.f_psnr_avg;
        h->stat.f_psnr_mean_y[h->sh.i_type]  += dur * pic_out->prop.f_psnr[0];
        h->stat.f_psnr_mean_u[h->sh.i_type]  += dur * pic_out->prop.f_psnr[1];
        h->stat.f_psnr_mean_v[h->sh.i_type]  += dur * pic_out->prop.f_psnr[2];

        snprintf( psz_message, 80, " PSNR Y:%5.2f U:%5.2f V:%5.2f", pic_out->prop.f_psnr[0],
                                                                    pic_out->prop.f_psnr[1],
                                                                    pic_out->prop.f_psnr[2] );
    }

    //需要计算SSIM
    if( h->param.analyse.b_ssim )
    {
    	//SSIM是已经在“滤波”环节计算过的
        pic_out->prop.f_ssim = h->stat.frame.f_ssim / h->stat.frame.i_ssim_cnt;
        //mean系列的需要累加
        h->stat.f_ssim_mean_y[h->sh.i_type] += pic_out->prop.f_ssim * dur;
        snprintf( psz_message + strlen(psz_message), 80 - strlen(psz_message),
                  " SSIM Y:%.5f", pic_out->prop.f_ssim );
    }
    psz_message[79] = '\0';
    //Debug时候输出
    x264_log( h, X264_LOG_DEBUG,
                  "frame=%4d QP=%.2f NAL=%d Slice:%c Poc:%-3d I:%-4d P:%-4d SKIP:%-4d size=%d bytes%s\n",
              h->i_frame,
              h->fdec->f_qp_avg_aq,
              h->i_nal_ref_idc,
              h->sh.i_type == SLICE_TYPE_I ? 'I' : (h->sh.i_type == SLICE_TYPE_P ? 'P' : 'B' ),
              h->fdec->i_poc,
              h->stat.frame.i_mb_count_i,
              h->stat.frame.i_mb_count_p,
              h->stat.frame.i_mb_count_skip,
              frame_size,
              psz_message );

    // keep stats all in one place
    x264_thread_sync_stat( h->thread[0], h );
    // for the use of the next frame
    x264_thread_sync_stat( thread_current, h );

#ifdef DEBUG_MB_TYPE
{
    static const char mb_chars[] = { 'i', 'i', 'I', 'C', 'P', '8', 'S',
        'D', '<', 'X', 'B', 'X', '>', 'B', 'B', 'B', 'B', '8', 'S' };
    for( int mb_xy = 0; mb_xy < h->mb.i_mb_width * h->mb.i_mb_height; mb_xy++ )
    {
        if( h->mb.type[mb_xy] < X264_MBTYPE_MAX && h->mb.type[mb_xy] >= 0 )
            fprintf( stderr, "%c ", mb_chars[ h->mb.type[mb_xy] ] );
        else
            fprintf( stderr, "? " );

        if( (mb_xy+1) % h->mb.i_mb_width == 0 )
            fprintf( stderr, "\n" );
    }
}
#endif

    /* Remove duplicates, must be done near the end as breaks h->fref0 array
     * by freeing some of its pointers. */
    for( int i = 0; i < h->i_ref[0]; i++ )
        if( h->fref[0][i] && h->fref[0][i]->b_duplicate )
        {
            x264_frame_push_blank_unused( h, h->fref[0][i] );
            h->fref[0][i] = 0;
        }

    if( h->param.psz_dump_yuv )
        x264_frame_dump( h );
    x264_emms();

    return frame_size;
}

x264_encoder_encapsulate_nals()

//封装一帧数据对应的NALU.
//例如给NALU添加起始码0x00000001
static int x264_encoder_encapsulate_nals( x264_t *h, int start )
{
    x264_t *h0 = h->thread[0];
    int nal_size = 0, previous_nal_size = 0;

    if( h->param.nalu_process )
    {
        for( int i = start; i < h->out.i_nal; i++ )
            nal_size += h->out.nal[i].i_payload;
        return nal_size;
    }

    for( int i = 0; i < start; i++ )
        previous_nal_size += h->out.nal[i].i_payload;

    for( int i = start; i < h->out.i_nal; i++ )
        nal_size += h->out.nal[i].i_payload;

    /* Worst-case NAL unit escaping: reallocate the buffer if it's too small. */
    int necessary_size = previous_nal_size + nal_size * 3/2 + h->out.i_nal * 4 + 4 + 64;
    for( int i = start; i < h->out.i_nal; i++ )
        necessary_size += h->out.nal[i].i_padding;
    if( x264_check_encapsulated_buffer( h, h0, start, previous_nal_size, necessary_size ) )
        return -1;

    uint8_t *nal_buffer = h0->nal_buffer + previous_nal_size;

    //一个一个NALU处理
    for( int i = start; i < h->out.i_nal; i++ )
    {
        int old_payload_len = h->out.nal[i].i_payload;
        h->out.nal[i].b_long_startcode = !i || h->out.nal[i].i_type == NAL_SPS || h->out.nal[i].i_type == NAL_PPS ||
                                         h->param.i_avcintra_class;
        //添加起始码
        x264_nal_encode( h, nal_buffer, &h->out.nal[i] );
        nal_buffer += h->out.nal[i].i_payload;
        if( h->param.i_avcintra_class )
        {
            h->out.nal[i].i_padding -= h->out.nal[i].i_payload - (old_payload_len + NALU_OVERHEAD);
            if( h->out.nal[i].i_padding > 0 )
            {
                memset( nal_buffer, 0, h->out.nal[i].i_padding );
                nal_buffer += h->out.nal[i].i_padding;
                h->out.nal[i].i_payload += h->out.nal[i].i_padding;
            }
            h->out.nal[i].i_padding = X264_MAX( h->out.nal[i].i_padding, 0 );
        }
    }

    x264_emms();

    return nal_buffer - (h0->nal_buffer + previous_nal_size);
}