FFmpeg/libavcodec/mpegvideo.c

/*
 * The simplest mpeg encoder (well, it was the simplest!)
 * Copyright (c) 2000,2001 Fabrice Bellard
 * Copyright (c) 2002-2004 Michael Niedermayer <michaelni@gmx.at>
 *
 * 4MV & hq & B-frame encoding stuff by Michael Niedermayer <michaelni@gmx.at>
 *
 * This file is part of FFmpeg.
 *
 * FFmpeg is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
 * version 2.1 of the License, or (at your option) any later version.
 *
 * FFmpeg is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
 * License along with FFmpeg; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
 */

/**
 * @file
 * The simplest mpeg encoder (well, it was the simplest!).
 */

#include "config_components.h"

#include "libavutil/attributes.h"
#include "libavutil/avassert.h"
#include "libavutil/imgutils.h"
#include "libavutil/internal.h"

#include "avcodec.h"
#include "blockdsp.h"
#include "h264chroma.h"
#include "idctdsp.h"
#include "mathops.h"
#include "mpeg_er.h"
#include "mpegutils.h"
#include "mpegvideo.h"
#include "mpeg4videodec.h"
#include "mpegvideodata.h"
#include "qpeldsp.h"
#include "threadframe.h"
#include "wmv2dec.h"
#include <limits.h>

static void dct_unquantize_mpeg1_intra_c(MpegEncContext *s,
                                   int16_t *block, int n, int qscale)
{
    int i, level, nCoeffs;
    const uint16_t *quant_matrix;

    nCoeffs= s->block_last_index[n];

    block[0] *= n < 4 ? s->y_dc_scale : s->c_dc_scale;
    /* XXX: only MPEG-1 */
    quant_matrix = s->intra_matrix;
    for(i=1;i<=nCoeffs;i++) {
        int j= s->intra_scantable.permutated[i];
        level = block[j];
        if (level) {
            if (level < 0) {
                level = -level;
                level = (int)(level * qscale * quant_matrix[j]) >> 3;
                level = (level - 1) | 1;
                level = -level;
            } else {
                level = (int)(level * qscale * quant_matrix[j]) >> 3;
                level = (level - 1) | 1;
            }
            block[j] = level;
        }
    }
}

static void dct_unquantize_mpeg1_inter_c(MpegEncContext *s,
                                   int16_t *block, int n, int qscale)
{
    int i, level, nCoeffs;
    const uint16_t *quant_matrix;

    nCoeffs= s->block_last_index[n];

    quant_matrix = s->inter_matrix;
    for(i=0; i<=nCoeffs; i++) {
        int j= s->intra_scantable.permutated[i];
        level = block[j];
        if (level) {
            if (level < 0) {
                level = -level;
                level = (((level << 1) + 1) * qscale *
                         ((int) (quant_matrix[j]))) >> 4;
                level = (level - 1) | 1;
                level = -level;
            } else {
                level = (((level << 1) + 1) * qscale *
                         ((int) (quant_matrix[j]))) >> 4;
                level = (level - 1) | 1;
            }
            block[j] = level;
        }
    }
}

static void dct_unquantize_mpeg2_intra_c(MpegEncContext *s,
                                   int16_t *block, int n, int qscale)
{
    int i, level, nCoeffs;
    const uint16_t *quant_matrix;

    if (s->q_scale_type) qscale = ff_mpeg2_non_linear_qscale[qscale];
    else                 qscale <<= 1;

    if(s->alternate_scan) nCoeffs= 63;
    else nCoeffs= s->block_last_index[n];

    block[0] *= n < 4 ? s->y_dc_scale : s->c_dc_scale;
    quant_matrix = s->intra_matrix;
    for(i=1;i<=nCoeffs;i++) {
        int j= s->intra_scantable.permutated[i];
        level = block[j];
        if (level) {
            if (level < 0) {
                level = -level;
                level = (int)(level * qscale * quant_matrix[j]) >> 4;
                level = -level;
            } else {
                level = (int)(level * qscale * quant_matrix[j]) >> 4;
            }
            block[j] = level;
        }
    }
}

static void dct_unquantize_mpeg2_intra_bitexact(MpegEncContext *s,
                                   int16_t *block, int n, int qscale)
{
    int i, level, nCoeffs;
    const uint16_t *quant_matrix;
    int sum=-1;

    if (s->q_scale_type) qscale = ff_mpeg2_non_linear_qscale[qscale];
    else                 qscale <<= 1;

    if(s->alternate_scan) nCoeffs= 63;
    else nCoeffs= s->block_last_index[n];

    block[0] *= n < 4 ? s->y_dc_scale : s->c_dc_scale;
    sum += block[0];
    quant_matrix = s->intra_matrix;
    for(i=1;i<=nCoeffs;i++) {
        int j= s->intra_scantable.permutated[i];
        level = block[j];
        if (level) {
            if (level < 0) {
                level = -level;
                level = (int)(level * qscale * quant_matrix[j]) >> 4;
                level = -level;
            } else {
                level = (int)(level * qscale * quant_matrix[j]) >> 4;
            }
            block[j] = level;
            sum+=level;
        }
    }
    block[63]^=sum&1;
}

static void dct_unquantize_mpeg2_inter_c(MpegEncContext *s,
                                   int16_t *block, int n, int qscale)
{
    int i, level, nCoeffs;
    const uint16_t *quant_matrix;
    int sum=-1;

    if (s->q_scale_type) qscale = ff_mpeg2_non_linear_qscale[qscale];
    else                 qscale <<= 1;

    if(s->alternate_scan) nCoeffs= 63;
    else nCoeffs= s->block_last_index[n];

    quant_matrix = s->inter_matrix;
    for(i=0; i<=nCoeffs; i++) {
        int j= s->intra_scantable.permutated[i];
        level = block[j];
        if (level) {
            if (level < 0) {
                level = -level;
                level = (((level << 1) + 1) * qscale *
                         ((int) (quant_matrix[j]))) >> 5;
                level = -level;
            } else {
                level = (((level << 1) + 1) * qscale *
                         ((int) (quant_matrix[j]))) >> 5;
            }
            block[j] = level;
            sum+=level;
        }
    }
    block[63]^=sum&1;
}

static void dct_unquantize_h263_intra_c(MpegEncContext *s,
                                  int16_t *block, int n, int qscale)
{
    int i, level, qmul, qadd;
    int nCoeffs;

    av_assert2(s->block_last_index[n]>=0 || s->h263_aic);

    qmul = qscale << 1;

    if (!s->h263_aic) {
        block[0] *= n < 4 ? s->y_dc_scale : s->c_dc_scale;
        qadd = (qscale - 1) | 1;
    }else{
        qadd = 0;
    }
    if(s->ac_pred)
        nCoeffs=63;
    else
        nCoeffs= s->intra_scantable.raster_end[ s->block_last_index[n] ];

    for(i=1; i<=nCoeffs; i++) {
        level = block[i];
        if (level) {
            if (level < 0) {
                level = level * qmul - qadd;
            } else {
                level = level * qmul + qadd;
            }
            block[i] = level;
        }
    }
}

static void dct_unquantize_h263_inter_c(MpegEncContext *s,
                                  int16_t *block, int n, int qscale)
{
    int i, level, qmul, qadd;
    int nCoeffs;

    av_assert2(s->block_last_index[n]>=0);

    qadd = (qscale - 1) | 1;
    qmul = qscale << 1;

    nCoeffs= s->inter_scantable.raster_end[ s->block_last_index[n] ];

    for(i=0; i<=nCoeffs; i++) {
        level = block[i];
        if (level) {
            if (level < 0) {
                level = level * qmul - qadd;
            } else {
                level = level * qmul + qadd;
            }
            block[i] = level;
        }
    }
}


static void gray16(uint8_t *dst, const uint8_t *src, ptrdiff_t linesize, int h)
{
    while(h--)
        memset(dst + h*linesize, 128, 16);
}

static void gray8(uint8_t *dst, const uint8_t *src, ptrdiff_t linesize, int h)
{
    while(h--)
        memset(dst + h*linesize, 128, 8);
}

/* init common dct for both encoder and decoder */
static av_cold int dct_init(MpegEncContext *s)
{
    ff_blockdsp_init(&s->bdsp, s->avctx);
    ff_h264chroma_init(&s->h264chroma, 8); //for lowres
    ff_hpeldsp_init(&s->hdsp, s->avctx->flags);
    ff_mpegvideodsp_init(&s->mdsp);
    ff_videodsp_init(&s->vdsp, s->avctx->bits_per_raw_sample);

    if (s->avctx->debug & FF_DEBUG_NOMC) {
        int i;
        for (i=0; i<4; i++) {
            s->hdsp.avg_pixels_tab[0][i] = gray16;
            s->hdsp.put_pixels_tab[0][i] = gray16;
            s->hdsp.put_no_rnd_pixels_tab[0][i] = gray16;

            s->hdsp.avg_pixels_tab[1][i] = gray8;
            s->hdsp.put_pixels_tab[1][i] = gray8;
            s->hdsp.put_no_rnd_pixels_tab[1][i] = gray8;
        }
    }

    s->dct_unquantize_h263_intra = dct_unquantize_h263_intra_c;
    s->dct_unquantize_h263_inter = dct_unquantize_h263_inter_c;
    s->dct_unquantize_mpeg1_intra = dct_unquantize_mpeg1_intra_c;
    s->dct_unquantize_mpeg1_inter = dct_unquantize_mpeg1_inter_c;
    s->dct_unquantize_mpeg2_intra = dct_unquantize_mpeg2_intra_c;
    if (s->avctx->flags & AV_CODEC_FLAG_BITEXACT)
        s->dct_unquantize_mpeg2_intra = dct_unquantize_mpeg2_intra_bitexact;
    s->dct_unquantize_mpeg2_inter = dct_unquantize_mpeg2_inter_c;

#if HAVE_INTRINSICS_NEON
    ff_mpv_common_init_neon(s);
#endif

#if ARCH_ALPHA
    ff_mpv_common_init_axp(s);
#elif ARCH_ARM
    ff_mpv_common_init_arm(s);
#elif ARCH_PPC
    ff_mpv_common_init_ppc(s);
#elif ARCH_X86
    ff_mpv_common_init_x86(s);
#elif ARCH_MIPS
    ff_mpv_common_init_mips(s);
#endif

    return 0;
}

av_cold void ff_mpv_idct_init(MpegEncContext *s)
{
    if (s->codec_id == AV_CODEC_ID_MPEG4)
        s->idsp.mpeg4_studio_profile = s->studio_profile;
    ff_idctdsp_init(&s->idsp, s->avctx);

    /* load & permutate scantables
     * note: only wmv uses different ones
     */
    if (s->alternate_scan) {
        ff_init_scantable(s->idsp.idct_permutation, &s->inter_scantable, ff_alternate_vertical_scan);
        ff_init_scantable(s->idsp.idct_permutation, &s->intra_scantable, ff_alternate_vertical_scan);
    } else {
        ff_init_scantable(s->idsp.idct_permutation, &s->inter_scantable, ff_zigzag_direct);
        ff_init_scantable(s->idsp.idct_permutation, &s->intra_scantable, ff_zigzag_direct);
    }
    ff_init_scantable(s->idsp.idct_permutation, &s->intra_h_scantable, ff_alternate_horizontal_scan);
    ff_init_scantable(s->idsp.idct_permutation, &s->intra_v_scantable, ff_alternate_vertical_scan);
}

static int init_duplicate_context(MpegEncContext *s)
{
    int y_size = s->b8_stride * (2 * s->mb_height + 1);
    int c_size = s->mb_stride * (s->mb_height + 1);
    int yc_size = y_size + 2 * c_size;
    int i;

    if (s->mb_height & 1)
        yc_size += 2*s->b8_stride + 2*s->mb_stride;

    if (s->encoding) {
        if (!FF_ALLOCZ_TYPED_ARRAY(s->me.map,       ME_MAP_SIZE) ||
            !FF_ALLOCZ_TYPED_ARRAY(s->me.score_map, ME_MAP_SIZE))
            return AVERROR(ENOMEM);

        if (s->noise_reduction) {
            if (!FF_ALLOCZ_TYPED_ARRAY(s->dct_error_sum,  2))
                return AVERROR(ENOMEM);
        }
    }
    if (!FF_ALLOCZ_TYPED_ARRAY(s->blocks,  2))
        return AVERROR(ENOMEM);
    s->block = s->blocks[0];

    for (i = 0; i < 12; i++) {
        s->pblocks[i] = &s->block[i];
    }

    if (s->avctx->codec_tag == AV_RL32("VCR2")) {
        // exchange uv
        FFSWAP(void *, s->pblocks[4], s->pblocks[5]);
    }

    if (s->out_format == FMT_H263) {
        /* ac values */
        if (!FF_ALLOCZ_TYPED_ARRAY(s->ac_val_base,  yc_size))
            return AVERROR(ENOMEM);
        s->ac_val[0] = s->ac_val_base + s->b8_stride + 1;
        s->ac_val[1] = s->ac_val_base + y_size + s->mb_stride + 1;
        s->ac_val[2] = s->ac_val[1] + c_size;
    }

    return 0;
}

int ff_mpv_init_duplicate_contexts(MpegEncContext *s)
{
    int nb_slices = s->slice_context_count, ret;

    /* We initialize the copies before the original so that
     * fields allocated in init_duplicate_context are NULL after
     * copying. This prevents double-frees upon allocation error. */
    for (int i = 1; i < nb_slices; i++) {
        s->thread_context[i] = av_memdup(s, sizeof(MpegEncContext));
        if (!s->thread_context[i])
            return AVERROR(ENOMEM);
        if ((ret = init_duplicate_context(s->thread_context[i])) < 0)
            return ret;
        s->thread_context[i]->start_mb_y =
            (s->mb_height * (i    ) + nb_slices / 2) / nb_slices;
        s->thread_context[i]->end_mb_y   =
            (s->mb_height * (i + 1) + nb_slices / 2) / nb_slices;
    }
    s->start_mb_y = 0;
    s->end_mb_y   = nb_slices > 1 ? (s->mb_height + nb_slices / 2) / nb_slices
                                  : s->mb_height;
    return init_duplicate_context(s);
}

static void free_duplicate_context(MpegEncContext *s)
{
    if (!s)
        return;

    av_freep(&s->sc.edge_emu_buffer);
    av_freep(&s->me.scratchpad);
    s->me.temp =
    s->sc.rd_scratchpad =
    s->sc.b_scratchpad =
    s->sc.obmc_scratchpad = NULL;

    av_freep(&s->dct_error_sum);
    av_freep(&s->me.map);
    av_freep(&s->me.score_map);
    av_freep(&s->blocks);
    av_freep(&s->ac_val_base);
    s->block = NULL;
}

static void free_duplicate_contexts(MpegEncContext *s)
{
    for (int i = 1; i < s->slice_context_count; i++) {
        free_duplicate_context(s->thread_context[i]);
        av_freep(&s->thread_context[i]);
    }
    free_duplicate_context(s);
}

static void backup_duplicate_context(MpegEncContext *bak, MpegEncContext *src)
{
#define COPY(a) bak->a = src->a
    COPY(sc.edge_emu_buffer);
    COPY(me.scratchpad);
    COPY(me.temp);
    COPY(sc.rd_scratchpad);
    COPY(sc.b_scratchpad);
    COPY(sc.obmc_scratchpad);
    COPY(me.map);
    COPY(me.score_map);
    COPY(blocks);
    COPY(block);
    COPY(start_mb_y);
    COPY(end_mb_y);
    COPY(me.map_generation);
    COPY(pb);
    COPY(dct_error_sum);
    COPY(dct_count[0]);
    COPY(dct_count[1]);
    COPY(ac_val_base);
    COPY(ac_val[0]);
    COPY(ac_val[1]);
    COPY(ac_val[2]);
#undef COPY
}

int ff_update_duplicate_context(MpegEncContext *dst, const MpegEncContext *src)
{
    MpegEncContext bak;
    int i, ret;
    // FIXME copy only needed parts
    backup_duplicate_context(&bak, dst);
    memcpy(dst, src, sizeof(MpegEncContext));
    backup_duplicate_context(dst, &bak);
    for (i = 0; i < 12; i++) {
        dst->pblocks[i] = &dst->block[i];
    }
    if (dst->avctx->codec_tag == AV_RL32("VCR2")) {
        // exchange uv
        FFSWAP(void *, dst->pblocks[4], dst->pblocks[5]);
    }
    if (!dst->sc.edge_emu_buffer &&
        (ret = ff_mpeg_framesize_alloc(dst->avctx, &dst->me,
                                       &dst->sc, dst->linesize)) < 0) {
        av_log(dst->avctx, AV_LOG_ERROR, "failed to allocate context "
               "scratch buffers.\n");
        return ret;
    }
    return 0;
}

/**
 * Set the given MpegEncContext to common defaults
 * (same for encoding and decoding).
 * The changed fields will not depend upon the
 * prior state of the MpegEncContext.
 */
void ff_mpv_common_defaults(MpegEncContext *s)
{
    s->y_dc_scale_table      =
    s->c_dc_scale_table      = ff_mpeg1_dc_scale_table;
    s->chroma_qscale_table   = ff_default_chroma_qscale_table;
    s->progressive_frame     = 1;
    s->progressive_sequence  = 1;
    s->picture_structure     = PICT_FRAME;

    s->coded_picture_number  = 0;
    s->picture_number        = 0;

    s->f_code                = 1;
    s->b_code                = 1;

    s->slice_context_count   = 1;
}

int ff_mpv_init_context_frame(MpegEncContext *s)
{
    int y_size, c_size, yc_size, i, mb_array_size, mv_table_size, x, y;

    s->mb_width   = (s->width + 15) / 16;
    s->mb_stride  = s->mb_width + 1;
    s->b8_stride  = s->mb_width * 2 + 1;
    mb_array_size = s->mb_height * s->mb_stride;
    mv_table_size = (s->mb_height + 2) * s->mb_stride + 1;

    /* set default edge pos, will be overridden
     * in decode_header if needed */
    s->h_edge_pos = s->mb_width * 16;
    s->v_edge_pos = s->mb_height * 16;

    s->mb_num     = s->mb_width * s->mb_height;

    s->block_wrap[0] =
    s->block_wrap[1] =
    s->block_wrap[2] =
    s->block_wrap[3] = s->b8_stride;
    s->block_wrap[4] =
    s->block_wrap[5] = s->mb_stride;

    y_size  = s->b8_stride * (2 * s->mb_height + 1);
    c_size  = s->mb_stride * (s->mb_height + 1);
    yc_size = y_size + 2   * c_size;

    if (s->mb_height & 1)
        yc_size += 2*s->b8_stride + 2*s->mb_stride;

    if (!FF_ALLOCZ_TYPED_ARRAY(s->mb_index2xy, s->mb_num + 1))
        return AVERROR(ENOMEM);
    for (y = 0; y < s->mb_height; y++)
        for (x = 0; x < s->mb_width; x++)
            s->mb_index2xy[x + y * s->mb_width] = x + y * s->mb_stride;

    s->mb_index2xy[s->mb_height * s->mb_width] = (s->mb_height - 1) * s->mb_stride + s->mb_width; // FIXME really needed?

    if (s->encoding) {
        /* Allocate MV tables */
        if (!FF_ALLOCZ_TYPED_ARRAY(s->p_mv_table_base,            mv_table_size) ||
            !FF_ALLOCZ_TYPED_ARRAY(s->b_forw_mv_table_base,       mv_table_size) ||
            !FF_ALLOCZ_TYPED_ARRAY(s->b_back_mv_table_base,       mv_table_size) ||
            !FF_ALLOCZ_TYPED_ARRAY(s->b_bidir_forw_mv_table_base, mv_table_size) ||
            !FF_ALLOCZ_TYPED_ARRAY(s->b_bidir_back_mv_table_base, mv_table_size) ||
            !FF_ALLOCZ_TYPED_ARRAY(s->b_direct_mv_table_base,     mv_table_size))
            return AVERROR(ENOMEM);
        s->p_mv_table            = s->p_mv_table_base + s->mb_stride + 1;
        s->b_forw_mv_table       = s->b_forw_mv_table_base + s->mb_stride + 1;
        s->b_back_mv_table       = s->b_back_mv_table_base + s->mb_stride + 1;
        s->b_bidir_forw_mv_table = s->b_bidir_forw_mv_table_base + s->mb_stride + 1;
        s->b_bidir_back_mv_table = s->b_bidir_back_mv_table_base + s->mb_stride + 1;
        s->b_direct_mv_table     = s->b_direct_mv_table_base + s->mb_stride + 1;

        /* Allocate MB type table */
        if (!FF_ALLOCZ_TYPED_ARRAY(s->mb_type,      mb_array_size) ||
            !FF_ALLOCZ_TYPED_ARRAY(s->lambda_table, mb_array_size) ||
            !FF_ALLOC_TYPED_ARRAY (s->cplx_tab,     mb_array_size) ||
            !FF_ALLOC_TYPED_ARRAY (s->bits_tab,     mb_array_size))
            return AVERROR(ENOMEM);

#define ALLOCZ_ARRAYS(p, mult, numb) ((p) = av_calloc(numb, mult * sizeof(*(p))))
        if (s->codec_id == AV_CODEC_ID_MPEG4 ||
            (s->avctx->flags & AV_CODEC_FLAG_INTERLACED_ME)) {
            int16_t (*tmp1)[2];
            uint8_t *tmp2;
            if (!(tmp1 = ALLOCZ_ARRAYS(s->b_field_mv_table_base, 8, mv_table_size)) ||
                !(tmp2 = ALLOCZ_ARRAYS(s->b_field_select_table[0][0], 2 * 4, mv_table_size)) ||
                !ALLOCZ_ARRAYS(s->p_field_select_table[0], 2 * 2, mv_table_size))
                return AVERROR(ENOMEM);

            s->p_field_select_table[1] = s->p_field_select_table[0] + 2 * mv_table_size;
            tmp1 += s->mb_stride + 1;

            for (int i = 0; i < 2; i++) {
                for (int j = 0; j < 2; j++) {
                    for (int k = 0; k < 2; k++) {
                        s->b_field_mv_table[i][j][k] = tmp1;
                        tmp1 += mv_table_size;
                    }
                    s->b_field_select_table[i][j] = tmp2;
                    tmp2 += 2 * mv_table_size;
                }
            }
        }
    }

    if (s->codec_id == AV_CODEC_ID_MPEG4 ||
        (s->avctx->flags & AV_CODEC_FLAG_INTERLACED_ME)) {
        int16_t (*tmp)[2];
        /* interlaced direct mode decoding tables */
        if (!(tmp = ALLOCZ_ARRAYS(s->p_field_mv_table_base, 4, mv_table_size)))
            return AVERROR(ENOMEM);
        tmp += s->mb_stride + 1;
        for (int i = 0; i < 2; i++) {
            for (int j = 0; j < 2; j++) {
                s->p_field_mv_table[i][j] = tmp;
                tmp += mv_table_size;
            }
        }
    }

    if (s->out_format == FMT_H263) {
        /* cbp values, cbp, ac_pred, pred_dir */
        if (!(s->coded_block_base = av_mallocz(y_size + (s->mb_height&1)*2*s->b8_stride)) ||
            !(s->cbp_table        = av_mallocz(mb_array_size)) ||
            !(s->pred_dir_table   = av_mallocz(mb_array_size)))
            return AVERROR(ENOMEM);
        s->coded_block = s->coded_block_base + s->b8_stride + 1;
    }

    if (s->h263_pred || s->h263_plus || !s->encoding) {
        /* dc values */
        // MN: we need these for error resilience of intra-frames
        if (!FF_ALLOCZ_TYPED_ARRAY(s->dc_val_base, yc_size))
            return AVERROR(ENOMEM);
        s->dc_val[0] = s->dc_val_base + s->b8_stride + 1;
        s->dc_val[1] = s->dc_val_base + y_size + s->mb_stride + 1;
        s->dc_val[2] = s->dc_val[1] + c_size;
        for (i = 0; i < yc_size; i++)
            s->dc_val_base[i] = 1024;
    }

    /* which mb is an intra block,  init macroblock skip table */
    if (!(s->mbintra_table = av_mallocz(mb_array_size)) ||
        // Note the + 1 is for a quicker MPEG-4 slice_end detection
        !(s->mbskip_table  = av_mallocz(mb_array_size + 2)))
        return AVERROR(ENOMEM);
    memset(s->mbintra_table, 1, mb_array_size);

    return !CONFIG_MPEGVIDEODEC || s->encoding ? 0 : ff_mpeg_er_init(s);
}

static void clear_context(MpegEncContext *s)
{
    int i, j, k;

    memset(&s->next_picture, 0, sizeof(s->next_picture));
    memset(&s->last_picture, 0, sizeof(s->last_picture));
    memset(&s->current_picture, 0, sizeof(s->current_picture));
    memset(&s->new_picture, 0, sizeof(s->new_picture));

    memset(s->thread_context, 0, sizeof(s->thread_context));

    s->me.map = NULL;
    s->me.score_map = NULL;
    s->dct_error_sum = NULL;
    s->block = NULL;
    s->blocks = NULL;
    memset(s->pblocks, 0, sizeof(s->pblocks));
    s->ac_val_base = NULL;
    s->ac_val[0] =
    s->ac_val[1] =
    s->ac_val[2] =NULL;
    s->sc.edge_emu_buffer = NULL;
    s->me.scratchpad = NULL;
    s->me.temp =
    s->sc.rd_scratchpad =
    s->sc.b_scratchpad =
    s->sc.obmc_scratchpad = NULL;


    s->bitstream_buffer = NULL;
    s->allocated_bitstream_buffer_size = 0;
    s->picture          = NULL;
    s->mb_type          = NULL;
    s->p_mv_table_base  = NULL;
    s->b_forw_mv_table_base = NULL;
    s->b_back_mv_table_base = NULL;
    s->b_bidir_forw_mv_table_base = NULL;
    s->b_bidir_back_mv_table_base = NULL;
    s->b_direct_mv_table_base = NULL;
    s->p_mv_table            = NULL;
    s->b_forw_mv_table       = NULL;
    s->b_back_mv_table       = NULL;
    s->b_bidir_forw_mv_table = NULL;
    s->b_bidir_back_mv_table = NULL;
    s->b_direct_mv_table     = NULL;
    s->b_field_mv_table_base = NULL;
    s->p_field_mv_table_base = NULL;
    for (i = 0; i < 2; i++) {
        for (j = 0; j < 2; j++) {
            for (k = 0; k < 2; k++) {
                s->b_field_mv_table[i][j][k] = NULL;
            }
            s->b_field_select_table[i][j] = NULL;
            s->p_field_mv_table[i][j] = NULL;
        }
        s->p_field_select_table[i] = NULL;
    }

    s->dc_val_base = NULL;
    s->coded_block_base = NULL;
    s->mbintra_table = NULL;
    s->cbp_table = NULL;
    s->pred_dir_table = NULL;

    s->mbskip_table = NULL;

    s->er.error_status_table = NULL;
    s->er.er_temp_buffer = NULL;
    s->mb_index2xy = NULL;
    s->lambda_table = NULL;

    s->cplx_tab = NULL;
    s->bits_tab = NULL;
}

/**
 * init common structure for both encoder and decoder.
 * this assumes that some variables like width/height are already set
 */
av_cold int ff_mpv_common_init(MpegEncContext *s)
{
    int i, ret;
    int nb_slices = (HAVE_THREADS &&
                     s->avctx->active_thread_type & FF_THREAD_SLICE) ?
                    s->avctx->thread_count : 1;

    clear_context(s);

    if (s->encoding && s->avctx->slices)
        nb_slices = s->avctx->slices;

    if (s->codec_id == AV_CODEC_ID_MPEG2VIDEO && !s->progressive_sequence)
        s->mb_height = (s->height + 31) / 32 * 2;
    else
        s->mb_height = (s->height + 15) / 16;

    if (s->avctx->pix_fmt == AV_PIX_FMT_NONE) {
        av_log(s->avctx, AV_LOG_ERROR,
               "decoding to AV_PIX_FMT_NONE is not supported.\n");
        return AVERROR(EINVAL);
    }

    if (nb_slices > MAX_THREADS || (nb_slices > s->mb_height && s->mb_height)) {
        int max_slices;
        if (s->mb_height)
            max_slices = FFMIN(MAX_THREADS, s->mb_height);
        else
            max_slices = MAX_THREADS;
        av_log(s->avctx, AV_LOG_WARNING, "too many threads/slices (%d),"
               " reducing to %d\n", nb_slices, max_slices);
        nb_slices = max_slices;
    }

    if ((s->width || s->height) &&
        av_image_check_size(s->width, s->height, 0, s->avctx))
        return AVERROR(EINVAL);

    dct_init(s);

    /* set chroma shifts */
    ret = av_pix_fmt_get_chroma_sub_sample(s->avctx->pix_fmt,
                                           &s->chroma_x_shift,
                                           &s->chroma_y_shift);
    if (ret)
        return ret;

    if (!FF_ALLOCZ_TYPED_ARRAY(s->picture, MAX_PICTURE_COUNT))
        return AVERROR(ENOMEM);
    for (i = 0; i < MAX_PICTURE_COUNT; i++) {
        s->picture[i].f = av_frame_alloc();
        if (!s->picture[i].f)
            goto fail_nomem;
    }

    if (!(s->next_picture.f    = av_frame_alloc()) ||
        !(s->last_picture.f    = av_frame_alloc()) ||
        !(s->current_picture.f = av_frame_alloc()) ||
        !(s->new_picture       = av_frame_alloc()))
        goto fail_nomem;

    if ((ret = ff_mpv_init_context_frame(s)))
        goto fail;

#if FF_API_FLAG_TRUNCATED
    s->parse_context.state = -1;
#endif

    s->context_initialized = 1;
    memset(s->thread_context, 0, sizeof(s->thread_context));
    s->thread_context[0]   = s;
    s->slice_context_count = nb_slices;

//     if (s->width && s->height) {
    ret = ff_mpv_init_duplicate_contexts(s);
    if (ret < 0)
        goto fail;
//     }

    return 0;
 fail_nomem:
    ret = AVERROR(ENOMEM);
 fail:
    ff_mpv_common_end(s);
    return ret;
}

void ff_mpv_free_context_frame(MpegEncContext *s)
{
    int i, j, k;

    free_duplicate_contexts(s);

    av_freep(&s->mb_type);
    av_freep(&s->p_mv_table_base);
    av_freep(&s->b_forw_mv_table_base);
    av_freep(&s->b_back_mv_table_base);
    av_freep(&s->b_bidir_forw_mv_table_base);
    av_freep(&s->b_bidir_back_mv_table_base);
    av_freep(&s->b_direct_mv_table_base);
    s->p_mv_table            = NULL;
    s->b_forw_mv_table       = NULL;
    s->b_back_mv_table       = NULL;
    s->b_bidir_forw_mv_table = NULL;
    s->b_bidir_back_mv_table = NULL;
    s->b_direct_mv_table     = NULL;
    av_freep(&s->b_field_mv_table_base);
    av_freep(&s->b_field_select_table[0][0]);
    av_freep(&s->p_field_mv_table_base);
    av_freep(&s->p_field_select_table[0]);
    for (i = 0; i < 2; i++) {
        for (j = 0; j < 2; j++) {
            for (k = 0; k < 2; k++) {
                s->b_field_mv_table[i][j][k] = NULL;
            }
            s->b_field_select_table[i][j] = NULL;
            s->p_field_mv_table[i][j] = NULL;
        }
        s->p_field_select_table[i] = NULL;
    }

    av_freep(&s->dc_val_base);
    av_freep(&s->coded_block_base);
    av_freep(&s->mbintra_table);
    av_freep(&s->cbp_table);
    av_freep(&s->pred_dir_table);

    av_freep(&s->mbskip_table);

    av_freep(&s->er.error_status_table);
    av_freep(&s->er.er_temp_buffer);
    av_freep(&s->mb_index2xy);
    av_freep(&s->lambda_table);

    av_freep(&s->cplx_tab);
    av_freep(&s->bits_tab);

    s->linesize = s->uvlinesize = 0;
}

/* init common structure for both encoder and decoder */
void ff_mpv_common_end(MpegEncContext *s)
{
    if (!s)
        return;

    ff_mpv_free_context_frame(s);
    if (s->slice_context_count > 1)
        s->slice_context_count = 1;

#if FF_API_FLAG_TRUNCATED
    av_freep(&s->parse_context.buffer);
    s->parse_context.buffer_size = 0;
#endif

    av_freep(&s->bitstream_buffer);
    s->allocated_bitstream_buffer_size = 0;

    if (!s->avctx)
        return;

    if (s->picture) {
        for (int i = 0; i < MAX_PICTURE_COUNT; i++)
            ff_mpv_picture_free(s->avctx, &s->picture[i]);
    }
    av_freep(&s->picture);
    ff_mpv_picture_free(s->avctx, &s->last_picture);
    ff_mpv_picture_free(s->avctx, &s->current_picture);
    ff_mpv_picture_free(s->avctx, &s->next_picture);
    av_frame_free(&s->new_picture);

    s->context_initialized      = 0;
    s->context_reinit           = 0;
    s->last_picture_ptr         =
    s->next_picture_ptr         =
    s->current_picture_ptr      = NULL;
    s->linesize = s->uvlinesize = 0;
}


static inline int hpel_motion_lowres(MpegEncContext *s,
                                     uint8_t *dest, uint8_t *src,
                                     int field_based, int field_select,
                                     int src_x, int src_y,
                                     int width, int height, ptrdiff_t stride,
                                     int h_edge_pos, int v_edge_pos,
                                     int w, int h, h264_chroma_mc_func *pix_op,
                                     int motion_x, int motion_y)
{
    const int lowres   = s->avctx->lowres;
    const int op_index = FFMIN(lowres, 3);
    const int s_mask   = (2 << lowres) - 1;
    int emu = 0;
    int sx, sy;

    if (s->quarter_sample) {
        motion_x /= 2;
        motion_y /= 2;
    }

    sx = motion_x & s_mask;
    sy = motion_y & s_mask;
    src_x += motion_x >> lowres + 1;
    src_y += motion_y >> lowres + 1;

    src   += src_y * stride + src_x;

    if ((unsigned)src_x > FFMAX( h_edge_pos - (!!sx) - w,                 0) ||
        (unsigned)src_y > FFMAX((v_edge_pos >> field_based) - (!!sy) - h, 0)) {
        s->vdsp.emulated_edge_mc(s->sc.edge_emu_buffer, src,
                                 s->linesize, s->linesize,
                                 w + 1, (h + 1) << field_based,
                                 src_x, src_y   << field_based,
                                 h_edge_pos, v_edge_pos);
        src = s->sc.edge_emu_buffer;
        emu = 1;
    }

    sx = (sx << 2) >> lowres;
    sy = (sy << 2) >> lowres;
    if (field_select)
        src += s->linesize;
    pix_op[op_index](dest, src, stride, h, sx, sy);
    return emu;
}

/* apply one mpeg motion vector to the three components */
static av_always_inline void mpeg_motion_lowres(MpegEncContext *s,
                                                uint8_t *dest_y,
                                                uint8_t *dest_cb,
                                                uint8_t *dest_cr,
                                                int field_based,
                                                int bottom_field,
                                                int field_select,
                                                uint8_t **ref_picture,
                                                h264_chroma_mc_func *pix_op,
                                                int motion_x, int motion_y,
                                                int h, int mb_y)
{
    uint8_t *ptr_y, *ptr_cb, *ptr_cr;
    int mx, my, src_x, src_y, uvsrc_x, uvsrc_y, sx, sy, uvsx, uvsy;
    ptrdiff_t uvlinesize, linesize;
    const int lowres     = s->avctx->lowres;
    const int op_index   = FFMIN(lowres-1+s->chroma_x_shift, 3);
    const int block_s    = 8>>lowres;
    const int s_mask     = (2 << lowres) - 1;
    const int h_edge_pos = s->h_edge_pos >> lowres;
    const int v_edge_pos = s->v_edge_pos >> lowres;
    linesize   = s->current_picture.f->linesize[0] << field_based;
    uvlinesize = s->current_picture.f->linesize[1] << field_based;

    // FIXME obviously not perfect but qpel will not work in lowres anyway
    if (s->quarter_sample) {
        motion_x /= 2;
        motion_y /= 2;
    }

    if(field_based){
        motion_y += (bottom_field - field_select)*((1 << lowres)-1);
    }

    sx = motion_x & s_mask;
    sy = motion_y & s_mask;
    src_x = s->mb_x * 2 * block_s + (motion_x >> lowres + 1);
    src_y = (mb_y * 2 * block_s >> field_based) + (motion_y >> lowres + 1);

    if (s->out_format == FMT_H263) {
        uvsx    = ((motion_x >> 1) & s_mask) | (sx & 1);
        uvsy    = ((motion_y >> 1) & s_mask) | (sy & 1);
        uvsrc_x = src_x >> 1;
        uvsrc_y = src_y >> 1;
    } else if (s->out_format == FMT_H261) {
        // even chroma mv's are full pel in H261
        mx      = motion_x / 4;
        my      = motion_y / 4;
        uvsx    = (2 * mx) & s_mask;
        uvsy    = (2 * my) & s_mask;
        uvsrc_x = s->mb_x * block_s + (mx >> lowres);
        uvsrc_y =    mb_y * block_s + (my >> lowres);
    } else {
        if(s->chroma_y_shift){
            mx      = motion_x / 2;
            my      = motion_y / 2;
            uvsx    = mx & s_mask;
            uvsy    = my & s_mask;
            uvsrc_x = s->mb_x * block_s                 + (mx >> lowres + 1);
            uvsrc_y =   (mb_y * block_s >> field_based) + (my >> lowres + 1);
        } else {
            if(s->chroma_x_shift){
            //Chroma422
                mx = motion_x / 2;
                uvsx = mx & s_mask;
                uvsy = motion_y & s_mask;
                uvsrc_y = src_y;
                uvsrc_x = s->mb_x*block_s               + (mx >> (lowres+1));
            } else {
            //Chroma444
                uvsx = motion_x & s_mask;
                uvsy = motion_y & s_mask;
                uvsrc_x = src_x;
                uvsrc_y = src_y;
            }
        }
    }

    ptr_y  = ref_picture[0] + src_y   * linesize   + src_x;
    ptr_cb = ref_picture[1] + uvsrc_y * uvlinesize + uvsrc_x;
    ptr_cr = ref_picture[2] + uvsrc_y * uvlinesize + uvsrc_x;

    if ((unsigned) src_x > FFMAX( h_edge_pos - (!!sx) - 2 * block_s,       0) || uvsrc_y<0 ||
        (unsigned) src_y > FFMAX((v_edge_pos >> field_based) - (!!sy) - h, 0)) {
        s->vdsp.emulated_edge_mc(s->sc.edge_emu_buffer, ptr_y,
                                 linesize >> field_based, linesize >> field_based,
                                 17, 17 + field_based,
                                src_x, src_y << field_based, h_edge_pos,
                                v_edge_pos);
        ptr_y = s->sc.edge_emu_buffer;
        if (!CONFIG_GRAY || !(s->avctx->flags & AV_CODEC_FLAG_GRAY)) {
            uint8_t *ubuf = s->sc.edge_emu_buffer + 18 * s->linesize;
            uint8_t *vbuf =ubuf + 10 * s->uvlinesize;
            if (s->workaround_bugs & FF_BUG_IEDGE)
                vbuf -= s->uvlinesize;
            s->vdsp.emulated_edge_mc(ubuf,  ptr_cb,
                                     uvlinesize >> field_based, uvlinesize >> field_based,
                                     9, 9 + field_based,
                                    uvsrc_x, uvsrc_y << field_based,
                                    h_edge_pos >> 1, v_edge_pos >> 1);
            s->vdsp.emulated_edge_mc(vbuf,  ptr_cr,
                                     uvlinesize >> field_based,uvlinesize >> field_based,
                                     9, 9 + field_based,
                                    uvsrc_x, uvsrc_y << field_based,
                                    h_edge_pos >> 1, v_edge_pos >> 1);
            ptr_cb = ubuf;
            ptr_cr = vbuf;
        }
    }

    // FIXME use this for field pix too instead of the obnoxious hack which changes picture.f->data
    if (bottom_field) {
        dest_y  += s->linesize;
        dest_cb += s->uvlinesize;
        dest_cr += s->uvlinesize;
    }

    if (field_select) {
        ptr_y   += s->linesize;
        ptr_cb  += s->uvlinesize;
        ptr_cr  += s->uvlinesize;
    }

    sx = (sx << 2) >> lowres;
    sy = (sy << 2) >> lowres;
    pix_op[lowres - 1](dest_y, ptr_y, linesize, h, sx, sy);

    if (!CONFIG_GRAY || !(s->avctx->flags & AV_CODEC_FLAG_GRAY)) {
        int hc = s->chroma_y_shift ? (h+1-bottom_field)>>1 : h;
        uvsx = (uvsx << 2) >> lowres;
        uvsy = (uvsy << 2) >> lowres;
        if (hc) {
            pix_op[op_index](dest_cb, ptr_cb, uvlinesize, hc, uvsx, uvsy);
            pix_op[op_index](dest_cr, ptr_cr, uvlinesize, hc, uvsx, uvsy);
        }
    }
    // FIXME h261 lowres loop filter
}

static inline void chroma_4mv_motion_lowres(MpegEncContext *s,
                                            uint8_t *dest_cb, uint8_t *dest_cr,
                                            uint8_t **ref_picture,
                                            h264_chroma_mc_func * pix_op,
                                            int mx, int my)
{
    const int lowres     = s->avctx->lowres;
    const int op_index   = FFMIN(lowres, 3);
    const int block_s    = 8 >> lowres;
    const int s_mask     = (2 << lowres) - 1;
    const int h_edge_pos = s->h_edge_pos >> lowres + 1;
    const int v_edge_pos = s->v_edge_pos >> lowres + 1;
    int emu = 0, src_x, src_y, sx, sy;
    ptrdiff_t offset;
    uint8_t *ptr;

    if (s->quarter_sample) {
        mx /= 2;
        my /= 2;
    }

    /* In case of 8X8, we construct a single chroma motion vector
       with a special rounding */
    mx = ff_h263_round_chroma(mx);
    my = ff_h263_round_chroma(my);

    sx = mx & s_mask;
    sy = my & s_mask;
    src_x = s->mb_x * block_s + (mx >> lowres + 1);
    src_y = s->mb_y * block_s + (my >> lowres + 1);

    offset = src_y * s->uvlinesize + src_x;
    ptr = ref_picture[1] + offset;
    if ((unsigned) src_x > FFMAX(h_edge_pos - (!!sx) - block_s, 0) ||
        (unsigned) src_y > FFMAX(v_edge_pos - (!!sy) - block_s, 0)) {
        s->vdsp.emulated_edge_mc(s->sc.edge_emu_buffer, ptr,
                                 s->uvlinesize, s->uvlinesize,
                                 9, 9,
                                 src_x, src_y, h_edge_pos, v_edge_pos);
        ptr = s->sc.edge_emu_buffer;
        emu = 1;
    }
    sx = (sx << 2) >> lowres;
    sy = (sy << 2) >> lowres;
    pix_op[op_index](dest_cb, ptr, s->uvlinesize, block_s, sx, sy);

    ptr = ref_picture[2] + offset;
    if (emu) {
        s->vdsp.emulated_edge_mc(s->sc.edge_emu_buffer, ptr,
                                 s->uvlinesize, s->uvlinesize,
                                 9, 9,
                                 src_x, src_y, h_edge_pos, v_edge_pos);
        ptr = s->sc.edge_emu_buffer;
    }
    pix_op[op_index](dest_cr, ptr, s->uvlinesize, block_s, sx, sy);
}

/**
 * motion compensation of a single macroblock
 * @param s context
 * @param dest_y luma destination pointer
 * @param dest_cb chroma cb/u destination pointer
 * @param dest_cr chroma cr/v destination pointer
 * @param dir direction (0->forward, 1->backward)
 * @param ref_picture array[3] of pointers to the 3 planes of the reference picture
 * @param pix_op halfpel motion compensation function (average or put normally)
 * the motion vectors are taken from s->mv and the MV type from s->mv_type
 */
static inline void MPV_motion_lowres(MpegEncContext *s,
                                     uint8_t *dest_y, uint8_t *dest_cb,
                                     uint8_t *dest_cr,
                                     int dir, uint8_t **ref_picture,
                                     h264_chroma_mc_func *pix_op)
{
    int mx, my;
    int mb_x, mb_y, i;
    const int lowres  = s->avctx->lowres;
    const int block_s = 8 >>lowres;

    mb_x = s->mb_x;
    mb_y = s->mb_y;

    switch (s->mv_type) {
    case MV_TYPE_16X16:
        mpeg_motion_lowres(s, dest_y, dest_cb, dest_cr,
                           0, 0, 0,
                           ref_picture, pix_op,
                           s->mv[dir][0][0], s->mv[dir][0][1],
                           2 * block_s, mb_y);
        break;
    case MV_TYPE_8X8:
        mx = 0;
        my = 0;
        for (i = 0; i < 4; i++) {
            hpel_motion_lowres(s, dest_y + ((i & 1) + (i >> 1) *
                               s->linesize) * block_s,
                               ref_picture[0], 0, 0,
                               (2 * mb_x + (i & 1)) * block_s,
                               (2 * mb_y + (i >> 1)) * block_s,
                               s->width, s->height, s->linesize,
                               s->h_edge_pos >> lowres, s->v_edge_pos >> lowres,
                               block_s, block_s, pix_op,
                               s->mv[dir][i][0], s->mv[dir][i][1]);

            mx += s->mv[dir][i][0];
            my += s->mv[dir][i][1];
        }

        if (!CONFIG_GRAY || !(s->avctx->flags & AV_CODEC_FLAG_GRAY))
            chroma_4mv_motion_lowres(s, dest_cb, dest_cr, ref_picture,
                                     pix_op, mx, my);
        break;
    case MV_TYPE_FIELD:
        if (s->picture_structure == PICT_FRAME) {
            /* top field */
            mpeg_motion_lowres(s, dest_y, dest_cb, dest_cr,
                               1, 0, s->field_select[dir][0],
                               ref_picture, pix_op,
                               s->mv[dir][0][0], s->mv[dir][0][1],
                               block_s, mb_y);
            /* bottom field */
            mpeg_motion_lowres(s, dest_y, dest_cb, dest_cr,
                               1, 1, s->field_select[dir][1],
                               ref_picture, pix_op,
                               s->mv[dir][1][0], s->mv[dir][1][1],
                               block_s, mb_y);
        } else {
            if (s->picture_structure != s->field_select[dir][0] + 1 &&
                s->pict_type != AV_PICTURE_TYPE_B && !s->first_field) {
                ref_picture = s->current_picture_ptr->f->data;

            }
            mpeg_motion_lowres(s, dest_y, dest_cb, dest_cr,
                               0, 0, s->field_select[dir][0],
                               ref_picture, pix_op,
                               s->mv[dir][0][0],
                               s->mv[dir][0][1], 2 * block_s, mb_y >> 1);
            }
        break;
    case MV_TYPE_16X8:
        for (i = 0; i < 2; i++) {
            uint8_t **ref2picture;

            if (s->picture_structure == s->field_select[dir][i] + 1 ||
                s->pict_type == AV_PICTURE_TYPE_B || s->first_field) {
                ref2picture = ref_picture;
            } else {
                ref2picture = s->current_picture_ptr->f->data;
            }

            mpeg_motion_lowres(s, dest_y, dest_cb, dest_cr,
                               0, 0, s->field_select[dir][i],
                               ref2picture, pix_op,
                               s->mv[dir][i][0], s->mv[dir][i][1] +
                               2 * block_s * i, block_s, mb_y >> 1);

            dest_y  +=  2 * block_s *  s->linesize;
            dest_cb += (2 * block_s >> s->chroma_y_shift) * s->uvlinesize;
            dest_cr += (2 * block_s >> s->chroma_y_shift) * s->uvlinesize;
        }
        break;
    case MV_TYPE_DMV:
        if (s->picture_structure == PICT_FRAME) {
            for (i = 0; i < 2; i++) {
                int j;
                for (j = 0; j < 2; j++) {
                    mpeg_motion_lowres(s, dest_y, dest_cb, dest_cr,
                                       1, j, j ^ i,
                                       ref_picture, pix_op,
                                       s->mv[dir][2 * i + j][0],
                                       s->mv[dir][2 * i + j][1],
                                       block_s, mb_y);
                }
                pix_op = s->h264chroma.avg_h264_chroma_pixels_tab;
            }
        } else {
            for (i = 0; i < 2; i++) {
                mpeg_motion_lowres(s, dest_y, dest_cb, dest_cr,
                                   0, 0, s->picture_structure != i + 1,
                                   ref_picture, pix_op,
                                   s->mv[dir][2 * i][0],s->mv[dir][2 * i][1],
                                   2 * block_s, mb_y >> 1);

                // after put we make avg of the same block
                pix_op = s->h264chroma.avg_h264_chroma_pixels_tab;

                // opposite parity is always in the same
                // frame if this is second field
                if (!s->first_field) {
                    ref_picture = s->current_picture_ptr->f->data;
                }
            }
        }
        break;
    default:
        av_assert2(0);
    }
}

/**
 * find the lowest MB row referenced in the MVs
 */
static int lowest_referenced_row(MpegEncContext *s, int dir)
{
    int my_max = INT_MIN, my_min = INT_MAX, qpel_shift = !s->quarter_sample;
    int my, off, i, mvs;

    if (s->picture_structure != PICT_FRAME || s->mcsel)
        goto unhandled;

    switch (s->mv_type) {
        case MV_TYPE_16X16:
            mvs = 1;
            break;
        case MV_TYPE_16X8:
            mvs = 2;
            break;
        case MV_TYPE_8X8:
            mvs = 4;
            break;
        default:
            goto unhandled;
    }

    for (i = 0; i < mvs; i++) {
        my = s->mv[dir][i][1];
        my_max = FFMAX(my_max, my);
        my_min = FFMIN(my_min, my);
    }

    off = ((FFMAX(-my_min, my_max)<<qpel_shift) + 63) >> 6;

    return av_clip(s->mb_y + off, 0, s->mb_height - 1);
unhandled:
    return s->mb_height-1;
}

/* put block[] to dest[] */
static inline void put_dct(MpegEncContext *s,
                           int16_t *block, int i, uint8_t *dest, int line_size, int qscale)
{
    s->dct_unquantize_intra(s, block, i, qscale);
    s->idsp.idct_put(dest, line_size, block);
}

/* add block[] to dest[] */
static inline void add_dct(MpegEncContext *s,
                           int16_t *block, int i, uint8_t *dest, int line_size)
{
    if (s->block_last_index[i] >= 0) {
        s->idsp.idct_add(dest, line_size, block);
    }
}

static inline void add_dequant_dct(MpegEncContext *s,
                           int16_t *block, int i, uint8_t *dest, int line_size, int qscale)
{
    if (s->block_last_index[i] >= 0) {
        s->dct_unquantize_inter(s, block, i, qscale);

        s->idsp.idct_add(dest, line_size, block);
    }
}

/**
 * Clean dc, ac, coded_block for the current non-intra MB.
 */
void ff_clean_intra_table_entries(MpegEncContext *s)
{
    int wrap = s->b8_stride;
    int xy = s->block_index[0];

    s->dc_val[0][xy           ] =
    s->dc_val[0][xy + 1       ] =
    s->dc_val[0][xy     + wrap] =
    s->dc_val[0][xy + 1 + wrap] = 1024;
    /* ac pred */
    memset(s->ac_val[0][xy       ], 0, 32 * sizeof(int16_t));
    memset(s->ac_val[0][xy + wrap], 0, 32 * sizeof(int16_t));
    if (s->msmpeg4_version>=3) {
        s->coded_block[xy           ] =
        s->coded_block[xy + 1       ] =
        s->coded_block[xy     + wrap] =
        s->coded_block[xy + 1 + wrap] = 0;
    }
    /* chroma */
    wrap = s->mb_stride;
    xy = s->mb_x + s->mb_y * wrap;
    s->dc_val[1][xy] =
    s->dc_val[2][xy] = 1024;
    /* ac pred */
    memset(s->ac_val[1][xy], 0, 16 * sizeof(int16_t));
    memset(s->ac_val[2][xy], 0, 16 * sizeof(int16_t));

    s->mbintra_table[xy]= 0;
}

/* generic function called after a macroblock has been parsed by the
   decoder or after it has been encoded by the encoder.

   Important variables used:
   s->mb_intra : true if intra macroblock
   s->mv_dir   : motion vector direction
   s->mv_type  : motion vector type
   s->mv       : motion vector
   s->interlaced_dct : true if interlaced dct used (mpeg2)
 */
static av_always_inline
void mpv_reconstruct_mb_internal(MpegEncContext *s, int16_t block[12][64],
                            int lowres_flag, int is_mpeg12)
{
#define IS_ENCODER(s) (CONFIG_MPEGVIDEOENC && !lowres_flag && (s)->encoding)
#define IS_MPEG12(s) (CONFIG_SMALL ? ((s)->out_format == FMT_MPEG1) : is_mpeg12)
    const int mb_xy = s->mb_y * s->mb_stride + s->mb_x;

    s->current_picture.qscale_table[mb_xy] = s->qscale;

    /* update DC predictors for P macroblocks */
    if (!s->mb_intra) {
        if (!is_mpeg12 && (s->h263_pred || s->h263_aic)) {
            if(s->mbintra_table[mb_xy])
                ff_clean_intra_table_entries(s);
        } else {
            s->last_dc[0] =
            s->last_dc[1] =
            s->last_dc[2] = 128 << s->intra_dc_precision;
        }
    }
    else if (!is_mpeg12 && (s->h263_pred || s->h263_aic))
        s->mbintra_table[mb_xy]=1;

    if (!IS_ENCODER(s) || (s->avctx->flags & AV_CODEC_FLAG_PSNR) || s->frame_skip_threshold || s->frame_skip_factor ||
        !((s->intra_only || s->pict_type == AV_PICTURE_TYPE_B) &&
          s->avctx->mb_decision != FF_MB_DECISION_RD)) { // FIXME precalc
        uint8_t *dest_y, *dest_cb, *dest_cr;
        int dct_linesize, dct_offset;
        op_pixels_func (*op_pix)[4];
        qpel_mc_func (*op_qpix)[16];
        const int linesize   = s->current_picture.f->linesize[0]; //not s->linesize as this would be wrong for field pics
        const int uvlinesize = s->current_picture.f->linesize[1];
        const int readable = s->pict_type != AV_PICTURE_TYPE_B || IS_ENCODER(s) || s->avctx->draw_horiz_band || lowres_flag;
        const int block_size= lowres_flag ? 8>>s->avctx->lowres : 8;

        /* avoid copy if macroblock skipped in last frame too */
        /* skip only during decoding as we might trash the buffers during encoding a bit */
        if (!IS_ENCODER(s)) {
            uint8_t *mbskip_ptr = &s->mbskip_table[mb_xy];

            if (s->mb_skipped) {
                s->mb_skipped= 0;
                av_assert2(s->pict_type!=AV_PICTURE_TYPE_I);
                *mbskip_ptr = 1;
            } else if(!s->current_picture.reference) {
                *mbskip_ptr = 1;
            } else{
                *mbskip_ptr = 0; /* not skipped */
            }
        }

        dct_linesize = linesize << s->interlaced_dct;
        dct_offset   = s->interlaced_dct ? linesize : linesize * block_size;

        if(readable){
            dest_y=  s->dest[0];
            dest_cb= s->dest[1];
            dest_cr= s->dest[2];
        }else{
            dest_y = s->sc.b_scratchpad;
            dest_cb= s->sc.b_scratchpad+16*linesize;
            dest_cr= s->sc.b_scratchpad+32*linesize;
        }

        if (!s->mb_intra) {
            /* motion handling */
            /* decoding or more than one mb_type (MC was already done otherwise) */
            if (!IS_ENCODER(s)) {

                if(HAVE_THREADS && s->avctx->active_thread_type&FF_THREAD_FRAME) {
                    if (s->mv_dir & MV_DIR_FORWARD) {
                        ff_thread_await_progress(&s->last_picture_ptr->tf,
                                                 lowest_referenced_row(s, 0),
                                                 0);
                    }
                    if (s->mv_dir & MV_DIR_BACKWARD) {
                        ff_thread_await_progress(&s->next_picture_ptr->tf,
                                                 lowest_referenced_row(s, 1),
                                                 0);
                    }
                }

                if(lowres_flag){
                    h264_chroma_mc_func *op_pix = s->h264chroma.put_h264_chroma_pixels_tab;

                    if (s->mv_dir & MV_DIR_FORWARD) {
                        MPV_motion_lowres(s, dest_y, dest_cb, dest_cr, 0, s->last_picture.f->data, op_pix);
                        op_pix = s->h264chroma.avg_h264_chroma_pixels_tab;
                    }
                    if (s->mv_dir & MV_DIR_BACKWARD) {
                        MPV_motion_lowres(s, dest_y, dest_cb, dest_cr, 1, s->next_picture.f->data, op_pix);
                    }
                }else{
                    op_qpix = s->me.qpel_put;
                    if ((is_mpeg12 || !s->no_rounding) || s->pict_type == AV_PICTURE_TYPE_B) {
                        op_pix = s->hdsp.put_pixels_tab;
                    }else{
                        op_pix = s->hdsp.put_no_rnd_pixels_tab;
                    }
                    if (s->mv_dir & MV_DIR_FORWARD) {
                        ff_mpv_motion(s, dest_y, dest_cb, dest_cr, 0, s->last_picture.f->data, op_pix, op_qpix);
                        op_pix = s->hdsp.avg_pixels_tab;
                        op_qpix= s->me.qpel_avg;
                    }
                    if (s->mv_dir & MV_DIR_BACKWARD) {
                        ff_mpv_motion(s, dest_y, dest_cb, dest_cr, 1, s->next_picture.f->data, op_pix, op_qpix);
                    }
                }
            }

            /* skip dequant / idct if we are really late ;) */
            if(s->avctx->skip_idct){
                if(  (s->avctx->skip_idct >= AVDISCARD_NONREF && s->pict_type == AV_PICTURE_TYPE_B)
                   ||(s->avctx->skip_idct >= AVDISCARD_NONKEY && s->pict_type != AV_PICTURE_TYPE_I)
                   || s->avctx->skip_idct >= AVDISCARD_ALL)
                    goto skip_idct;
            }

            /* add dct residue */
            if (IS_ENCODER(s) || !(IS_MPEG12(s) || s->msmpeg4_version
                                || (s->codec_id==AV_CODEC_ID_MPEG4 && !s->mpeg_quant))){
                add_dequant_dct(s, block[0], 0, dest_y                          , dct_linesize, s->qscale);
                add_dequant_dct(s, block[1], 1, dest_y              + block_size, dct_linesize, s->qscale);
                add_dequant_dct(s, block[2], 2, dest_y + dct_offset             , dct_linesize, s->qscale);
                add_dequant_dct(s, block[3], 3, dest_y + dct_offset + block_size, dct_linesize, s->qscale);

                if (!CONFIG_GRAY || !(s->avctx->flags & AV_CODEC_FLAG_GRAY)) {
                    if (s->chroma_y_shift){
                        add_dequant_dct(s, block[4], 4, dest_cb, uvlinesize, s->chroma_qscale);
                        add_dequant_dct(s, block[5], 5, dest_cr, uvlinesize, s->chroma_qscale);
                    }else{
                        dct_linesize >>= 1;
                        dct_offset >>=1;
                        add_dequant_dct(s, block[4], 4, dest_cb,              dct_linesize, s->chroma_qscale);
                        add_dequant_dct(s, block[5], 5, dest_cr,              dct_linesize, s->chroma_qscale);
                        add_dequant_dct(s, block[6], 6, dest_cb + dct_offset, dct_linesize, s->chroma_qscale);
                        add_dequant_dct(s, block[7], 7, dest_cr + dct_offset, dct_linesize, s->chroma_qscale);
                    }
                }
            } else if(is_mpeg12 || (s->codec_id != AV_CODEC_ID_WMV2)){
                add_dct(s, block[0], 0, dest_y                          , dct_linesize);
                add_dct(s, block[1], 1, dest_y              + block_size, dct_linesize);
                add_dct(s, block[2], 2, dest_y + dct_offset             , dct_linesize);
                add_dct(s, block[3], 3, dest_y + dct_offset + block_size, dct_linesize);

                if (!CONFIG_GRAY || !(s->avctx->flags & AV_CODEC_FLAG_GRAY)) {
                    if(s->chroma_y_shift){//Chroma420
                        add_dct(s, block[4], 4, dest_cb, uvlinesize);
                        add_dct(s, block[5], 5, dest_cr, uvlinesize);
                    }else{
                        //chroma422
                        dct_linesize = uvlinesize << s->interlaced_dct;
                        dct_offset   = s->interlaced_dct ? uvlinesize : uvlinesize*block_size;

                        add_dct(s, block[4], 4, dest_cb, dct_linesize);
                        add_dct(s, block[5], 5, dest_cr, dct_linesize);
                        add_dct(s, block[6], 6, dest_cb+dct_offset, dct_linesize);
                        add_dct(s, block[7], 7, dest_cr+dct_offset, dct_linesize);
                        if(!s->chroma_x_shift){//Chroma444
                            add_dct(s, block[8], 8, dest_cb+block_size, dct_linesize);
                            add_dct(s, block[9], 9, dest_cr+block_size, dct_linesize);
                            add_dct(s, block[10], 10, dest_cb+block_size+dct_offset, dct_linesize);
                            add_dct(s, block[11], 11, dest_cr+block_size+dct_offset, dct_linesize);
                        }
                    }
                }//fi gray
            } else if (CONFIG_WMV2_DECODER) {
                ff_wmv2_add_mb(s, block, dest_y, dest_cb, dest_cr);
            }
        } else {
            /* Only MPEG-4 Simple Studio Profile is supported in > 8-bit mode.
               TODO: Integrate 10-bit properly into mpegvideo.c so that ER works properly */
            if (!is_mpeg12 && CONFIG_MPEG4_DECODER && /* s->codec_id == AV_CODEC_ID_MPEG4 && */
                s->avctx->bits_per_raw_sample > 8) {
                ff_mpeg4_decode_studio(s, dest_y, dest_cb, dest_cr, block_size,
                                       uvlinesize, dct_linesize, dct_offset);
            }
            /* dct only in intra block */
            else if (IS_ENCODER(s) || !IS_MPEG12(s)) {
                put_dct(s, block[0], 0, dest_y                          , dct_linesize, s->qscale);
                put_dct(s, block[1], 1, dest_y              + block_size, dct_linesize, s->qscale);
                put_dct(s, block[2], 2, dest_y + dct_offset             , dct_linesize, s->qscale);
                put_dct(s, block[3], 3, dest_y + dct_offset + block_size, dct_linesize, s->qscale);

                if (!CONFIG_GRAY || !(s->avctx->flags & AV_CODEC_FLAG_GRAY)) {
                    if(s->chroma_y_shift){
                        put_dct(s, block[4], 4, dest_cb, uvlinesize, s->chroma_qscale);
                        put_dct(s, block[5], 5, dest_cr, uvlinesize, s->chroma_qscale);
                    }else{
                        dct_offset >>=1;
                        dct_linesize >>=1;
                        put_dct(s, block[4], 4, dest_cb,              dct_linesize, s->chroma_qscale);
                        put_dct(s, block[5], 5, dest_cr,              dct_linesize, s->chroma_qscale);
                        put_dct(s, block[6], 6, dest_cb + dct_offset, dct_linesize, s->chroma_qscale);
                        put_dct(s, block[7], 7, dest_cr + dct_offset, dct_linesize, s->chroma_qscale);
                    }
                }
            }else{
                s->idsp.idct_put(dest_y,                           dct_linesize, block[0]);
                s->idsp.idct_put(dest_y              + block_size, dct_linesize, block[1]);
                s->idsp.idct_put(dest_y + dct_offset,              dct_linesize, block[2]);
                s->idsp.idct_put(dest_y + dct_offset + block_size, dct_linesize, block[3]);

                if (!CONFIG_GRAY || !(s->avctx->flags & AV_CODEC_FLAG_GRAY)) {
                    if(s->chroma_y_shift){
                        s->idsp.idct_put(dest_cb, uvlinesize, block[4]);
                        s->idsp.idct_put(dest_cr, uvlinesize, block[5]);
                    }else{

                        dct_linesize = uvlinesize << s->interlaced_dct;
                        dct_offset   = s->interlaced_dct ? uvlinesize : uvlinesize*block_size;

                        s->idsp.idct_put(dest_cb,              dct_linesize, block[4]);
                        s->idsp.idct_put(dest_cr,              dct_linesize, block[5]);
                        s->idsp.idct_put(dest_cb + dct_offset, dct_linesize, block[6]);
                        s->idsp.idct_put(dest_cr + dct_offset, dct_linesize, block[7]);
                        if(!s->chroma_x_shift){//Chroma444
                            s->idsp.idct_put(dest_cb + block_size,              dct_linesize, block[8]);
                            s->idsp.idct_put(dest_cr + block_size,              dct_linesize, block[9]);
                            s->idsp.idct_put(dest_cb + block_size + dct_offset, dct_linesize, block[10]);
                            s->idsp.idct_put(dest_cr + block_size + dct_offset, dct_linesize, block[11]);
                        }
                    }
                }//gray
            }
        }
skip_idct:
        if(!readable){
            s->hdsp.put_pixels_tab[0][0](s->dest[0], dest_y ,   linesize,16);
            if (!CONFIG_GRAY || !(s->avctx->flags & AV_CODEC_FLAG_GRAY)) {
                s->hdsp.put_pixels_tab[s->chroma_x_shift][0](s->dest[1], dest_cb, uvlinesize,16 >> s->chroma_y_shift);
                s->hdsp.put_pixels_tab[s->chroma_x_shift][0](s->dest[2], dest_cr, uvlinesize,16 >> s->chroma_y_shift);
            }
        }
    }
}

void ff_mpv_reconstruct_mb(MpegEncContext *s, int16_t block[12][64])
{
    if (s->avctx->debug & FF_DEBUG_DCT_COEFF) {
       /* print DCT coefficients */
       av_log(s->avctx, AV_LOG_DEBUG, "DCT coeffs of MB at %dx%d:\n", s->mb_x, s->mb_y);
       for (int i = 0; i < 6; i++) {
           for (int j = 0; j < 64; j++) {
               av_log(s->avctx, AV_LOG_DEBUG, "%5d",
                      block[i][s->idsp.idct_permutation[j]]);
           }
           av_log(s->avctx, AV_LOG_DEBUG, "\n");
       }
    }

#if !CONFIG_SMALL
    if(s->out_format == FMT_MPEG1) {
        if(s->avctx->lowres) mpv_reconstruct_mb_internal(s, block, 1, 1);
        else                 mpv_reconstruct_mb_internal(s, block, 0, 1);
    } else
#endif
    if(s->avctx->lowres) mpv_reconstruct_mb_internal(s, block, 1, 0);
    else                  mpv_reconstruct_mb_internal(s, block, 0, 0);
}

void ff_init_block_index(MpegEncContext *s){ //FIXME maybe rename
    const int linesize   = s->current_picture.f->linesize[0]; //not s->linesize as this would be wrong for field pics
    const int uvlinesize = s->current_picture.f->linesize[1];
    const int width_of_mb = (4 + (s->avctx->bits_per_raw_sample > 8)) - s->avctx->lowres;
    const int height_of_mb = 4 - s->avctx->lowres;

    s->block_index[0]= s->b8_stride*(s->mb_y*2    ) - 2 + s->mb_x*2;
    s->block_index[1]= s->b8_stride*(s->mb_y*2    ) - 1 + s->mb_x*2;
    s->block_index[2]= s->b8_stride*(s->mb_y*2 + 1) - 2 + s->mb_x*2;
    s->block_index[3]= s->b8_stride*(s->mb_y*2 + 1) - 1 + s->mb_x*2;
    s->block_index[4]= s->mb_stride*(s->mb_y + 1)                + s->b8_stride*s->mb_height*2 + s->mb_x - 1;
    s->block_index[5]= s->mb_stride*(s->mb_y + s->mb_height + 2) + s->b8_stride*s->mb_height*2 + s->mb_x - 1;
    //block_index is not used by mpeg2, so it is not affected by chroma_format

    s->dest[0] = s->current_picture.f->data[0] + (int)((s->mb_x - 1U) <<  width_of_mb);
    s->dest[1] = s->current_picture.f->data[1] + (int)((s->mb_x - 1U) << (width_of_mb - s->chroma_x_shift));
    s->dest[2] = s->current_picture.f->data[2] + (int)((s->mb_x - 1U) << (width_of_mb - s->chroma_x_shift));

    if(!(s->pict_type==AV_PICTURE_TYPE_B && s->avctx->draw_horiz_band && s->picture_structure==PICT_FRAME))
    {
        if(s->picture_structure==PICT_FRAME){
        s->dest[0] += s->mb_y *   linesize << height_of_mb;
        s->dest[1] += s->mb_y * uvlinesize << (height_of_mb - s->chroma_y_shift);
        s->dest[2] += s->mb_y * uvlinesize << (height_of_mb - s->chroma_y_shift);
        }else{
            s->dest[0] += (s->mb_y>>1) *   linesize << height_of_mb;
            s->dest[1] += (s->mb_y>>1) * uvlinesize << (height_of_mb - s->chroma_y_shift);
            s->dest[2] += (s->mb_y>>1) * uvlinesize << (height_of_mb - s->chroma_y_shift);
            av_assert1((s->mb_y&1) == (s->picture_structure == PICT_BOTTOM_FIELD));
        }
    }
}

/**
 * set qscale and update qscale dependent variables.
 */
void ff_set_qscale(MpegEncContext * s, int qscale)
{
    if (qscale < 1)
        qscale = 1;
    else if (qscale > 31)
        qscale = 31;

    s->qscale = qscale;
    s->chroma_qscale= s->chroma_qscale_table[qscale];

    s->y_dc_scale= s->y_dc_scale_table[ qscale ];
    s->c_dc_scale= s->c_dc_scale_table[ s->chroma_qscale ];
}