changeset 811:3afe5701df41 trunk

[svn] - remove h261/h263/h264 stuff
author nenolod
date Mon, 12 Mar 2007 11:38:51 -0700
parents 4174d2783be5
children 85be08903b70
files ChangeLog src/ffmpeg/libavcodec/Makefile src/ffmpeg/libavcodec/h261.c src/ffmpeg/libavcodec/h261data.h src/ffmpeg/libavcodec/h263.c src/ffmpeg/libavcodec/h263data.h src/ffmpeg/libavcodec/h263dec.c src/ffmpeg/libavcodec/h264.c src/ffmpeg/libavcodec/h264data.h src/ffmpeg/libavcodec/h264idct.c
diffstat 10 files changed, 30 insertions(+), 18832 deletions(-) [+]
line wrap: on
line diff
--- a/ChangeLog	Mon Mar 12 11:37:35 2007 -0700
+++ b/ChangeLog	Mon Mar 12 11:38:51 2007 -0700
@@ -1,3 +1,28 @@
+2007-03-12 18:37:35 +0000  William Pitcock <nenolod@sacredspiral.co.uk>
+  revision [1720]
+  - delete a lot of stuff
+  
+  trunk/src/ffmpeg/libavcodec/Makefile              |   39 
+  trunk/src/ffmpeg/libavcodec/a52dec.c              |  257 -
+  trunk/src/ffmpeg/libavcodec/ac3dec.c              |  184 -
+  trunk/src/ffmpeg/libavcodec/amr.c                 |  666 ----
+  trunk/src/ffmpeg/libavcodec/faad.c                |  332 --
+  trunk/src/ffmpeg/libavcodec/jpeg_ls.c             |  847 -----
+  trunk/src/ffmpeg/libavcodec/libgsm.c              |   97 
+  trunk/src/ffmpeg/libavcodec/mdec.c                |  270 -
+  trunk/src/ffmpeg/libavcodec/motion_est.c          | 2117 --------------
+  trunk/src/ffmpeg/libavcodec/motion_est_template.c | 1104 -------
+  trunk/src/ffmpeg/libavcodec/mp3lameaudio.c        |  221 -
+  trunk/src/ffmpeg/libavcodec/mpeg12.c              | 3316 ----------------------
+  trunk/src/ffmpeg/libavcodec/msmpeg4.c             | 1938 ------------
+  trunk/src/ffmpeg/libavcodec/wmv2.c                |  855 -----
+  trunk/src/ffmpeg/libavcodec/x264.c                |  295 -
+  trunk/src/ffmpeg/libavcodec/xvid_internal.h       |   32 
+  trunk/src/ffmpeg/libavcodec/xvid_rc.c             |  148 
+  trunk/src/ffmpeg/libavcodec/xvidff.c              |  768 -----
+  18 files changed, 19 insertions(+), 13467 deletions(-)
+
+
 2007-03-12 18:31:14 +0000  Yoshiki Yazawa <yaz@cc.rim.or.jp>
   revision [1718]
   - make use of new millisecond seek API.
--- a/src/ffmpeg/libavcodec/Makefile	Mon Mar 12 11:37:35 2007 -0700
+++ b/src/ffmpeg/libavcodec/Makefile	Mon Mar 12 11:38:51 2007 -0700
@@ -4,11 +4,11 @@
 OBJECTIVE_LIBS_NOINST=	libavcodec.a
 
 SOURCES = \
-4xm.c                     h261.c                    raw.c		\
-8bps.c                    h263.c                    resample.c	\
-h263dec.c                 resample2.c	\
-aasc.c                    h264.c                    roqvideo.c	\
-h264idct.c                rpza.c	\
+4xm.c                     raw.c		\
+8bps.c                    resample.c	\
+resample2.c	\
+aasc.c                    roqvideo.c	\
+rpza.c	\
 ac3enc.c                  huffyuv.c                 rtjpeg.c	\
 adpcm.c                   idcinvideo.c              rv10.c	\
 adx.c                     imgconvert.c              shorten.c	\
--- a/src/ffmpeg/libavcodec/h261.c	Mon Mar 12 11:37:35 2007 -0700
+++ /dev/null	Thu Jan 01 00:00:00 1970 +0000
@@ -1,1055 +0,0 @@
-/*
- * H261 decoder
- * Copyright (c) 2002-2004 Michael Niedermayer <michaelni@gmx.at>
- * Copyright (c) 2004 Maarten Daniels
- *
- * 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 h261.c
- * h261codec.
- */
-
-#include "common.h"
-#include "dsputil.h"
-#include "avcodec.h"
-#include "mpegvideo.h"
-#include "h261data.h"
-
-
-#define H261_MBA_VLC_BITS 9
-#define H261_MTYPE_VLC_BITS 6
-#define H261_MV_VLC_BITS 7
-#define H261_CBP_VLC_BITS 9
-#define TCOEFF_VLC_BITS 9
-
-#define MBA_STUFFING 33
-#define MBA_STARTCODE 34
-#define IS_FIL(a)    ((a)&MB_TYPE_H261_FIL)
-
-/**
- * H261Context
- */
-typedef struct H261Context{
-    MpegEncContext s;
-
-    int current_mba;
-    int previous_mba;
-    int mba_diff;
-    int mtype;
-    int current_mv_x;
-    int current_mv_y;
-    int gob_number;
-    int gob_start_code_skipped; // 1 if gob start code is already read before gob header is read
-}H261Context;
-
-void ff_h261_loop_filter(MpegEncContext *s){
-    H261Context * h= (H261Context*)s;
-    const int linesize  = s->linesize;
-    const int uvlinesize= s->uvlinesize;
-    uint8_t *dest_y = s->dest[0];
-    uint8_t *dest_cb= s->dest[1];
-    uint8_t *dest_cr= s->dest[2];
-
-    if(!(IS_FIL (h->mtype)))
-        return;
-
-    s->dsp.h261_loop_filter(dest_y                   , linesize);
-    s->dsp.h261_loop_filter(dest_y                + 8, linesize);
-    s->dsp.h261_loop_filter(dest_y + 8 * linesize    , linesize);
-    s->dsp.h261_loop_filter(dest_y + 8 * linesize + 8, linesize);
-    s->dsp.h261_loop_filter(dest_cb, uvlinesize);
-    s->dsp.h261_loop_filter(dest_cr, uvlinesize);
-}
-
-static int ff_h261_get_picture_format(int width, int height){
-    // QCIF
-    if (width == 176 && height == 144)
-        return 0;
-    // CIF
-    else if (width == 352 && height == 288)
-        return 1;
-    // ERROR
-    else
-        return -1;
-}
-
-static void h261_encode_block(H261Context * h, DCTELEM * block,
-                              int n);
-static int h261_decode_block(H261Context *h, DCTELEM *block,
-                             int n, int coded);
-
-void ff_h261_encode_picture_header(MpegEncContext * s, int picture_number){
-    H261Context * h = (H261Context *) s;
-    int format, temp_ref;
-
-    align_put_bits(&s->pb);
-
-    /* Update the pointer to last GOB */
-    s->ptr_lastgob = pbBufPtr(&s->pb);
-
-    put_bits(&s->pb, 20, 0x10); /* PSC */
-
-    temp_ref= s->picture_number * (int64_t)30000 * s->avctx->time_base.num /
-                         (1001 * (int64_t)s->avctx->time_base.den); //FIXME maybe this should use a timestamp
-    put_bits(&s->pb, 5, temp_ref & 0x1f); /* TemporalReference */
-
-    put_bits(&s->pb, 1, 0); /* split screen off */
-    put_bits(&s->pb, 1, 0); /* camera  off */
-    put_bits(&s->pb, 1, 0); /* freeze picture release off */
-
-    format = ff_h261_get_picture_format(s->width, s->height);
-
-    put_bits(&s->pb, 1, format); /* 0 == QCIF, 1 == CIF */
-
-    put_bits(&s->pb, 1, 0); /* still image mode */
-    put_bits(&s->pb, 1, 0); /* reserved */
-
-    put_bits(&s->pb, 1, 0); /* no PEI */
-    if(format == 0)
-        h->gob_number = -1;
-    else
-        h->gob_number = 0;
-    h->current_mba = 0;
-}
-
-/**
- * Encodes a group of blocks header.
- */
-static void h261_encode_gob_header(MpegEncContext * s, int mb_line){
-    H261Context * h = (H261Context *)s;
-    if(ff_h261_get_picture_format(s->width, s->height) == 0){
-        h->gob_number+=2; // QCIF
-    }
-    else{
-        h->gob_number++; // CIF
-    }
-    put_bits(&s->pb, 16, 1); /* GBSC */
-    put_bits(&s->pb, 4, h->gob_number); /* GN */
-    put_bits(&s->pb, 5, s->qscale); /* GQUANT */
-    put_bits(&s->pb, 1, 0); /* no GEI */
-    h->current_mba = 0;
-    h->previous_mba = 0;
-    h->current_mv_x=0;
-    h->current_mv_y=0;
-}
-
-void ff_h261_reorder_mb_index(MpegEncContext* s){
-    int index= s->mb_x + s->mb_y*s->mb_width;
-
-    if(index % 33 == 0)
-        h261_encode_gob_header(s,0);
-
-    /* for CIF the GOB's are fragmented in the middle of a scanline
-       that's why we need to adjust the x and y index of the macroblocks */
-    if(ff_h261_get_picture_format(s->width,s->height) == 1){ // CIF
-        s->mb_x =     index % 11 ; index /= 11;
-        s->mb_y =     index %  3 ; index /=  3;
-        s->mb_x+= 11*(index %  2); index /=  2;
-        s->mb_y+=  3*index;
-
-        ff_init_block_index(s);
-        ff_update_block_index(s);
-    }
-}
-
-static void h261_encode_motion(H261Context * h, int val){
-    MpegEncContext * const s = &h->s;
-    int sign, code;
-    if(val==0){
-        code = 0;
-        put_bits(&s->pb,h261_mv_tab[code][1],h261_mv_tab[code][0]);
-    }
-    else{
-        if(val > 15)
-            val -=32;
-        if(val < -16)
-            val+=32;
-        sign = val < 0;
-        code = sign ? -val : val;
-        put_bits(&s->pb,h261_mv_tab[code][1],h261_mv_tab[code][0]);
-        put_bits(&s->pb,1,sign);
-    }
-}
-
-static inline int get_cbp(MpegEncContext * s,
-                      DCTELEM block[6][64])
-{
-    int i, cbp;
-    cbp= 0;
-    for (i = 0; i < 6; i++) {
-        if (s->block_last_index[i] >= 0)
-            cbp |= 1 << (5 - i);
-    }
-    return cbp;
-}
-void ff_h261_encode_mb(MpegEncContext * s,
-         DCTELEM block[6][64],
-         int motion_x, int motion_y)
-{
-    H261Context * h = (H261Context *)s;
-    int mvd, mv_diff_x, mv_diff_y, i, cbp;
-    cbp = 63; // avoid warning
-    mvd = 0;
-
-    h->current_mba++;
-    h->mtype = 0;
-
-    if (!s->mb_intra){
-        /* compute cbp */
-        cbp= get_cbp(s, block);
-
-        /* mvd indicates if this block is motion compensated */
-        mvd = motion_x | motion_y;
-
-        if((cbp | mvd | s->dquant ) == 0) {
-            /* skip macroblock */
-            s->skip_count++;
-            h->current_mv_x=0;
-            h->current_mv_y=0;
-            return;
-        }
-    }
-
-    /* MB is not skipped, encode MBA */
-    put_bits(&s->pb, h261_mba_bits[(h->current_mba-h->previous_mba)-1], h261_mba_code[(h->current_mba-h->previous_mba)-1]);
-
-    /* calculate MTYPE */
-    if(!s->mb_intra){
-        h->mtype++;
-
-        if(mvd || s->loop_filter)
-            h->mtype+=3;
-        if(s->loop_filter)
-            h->mtype+=3;
-        if(cbp || s->dquant)
-            h->mtype++;
-        assert(h->mtype > 1);
-    }
-
-    if(s->dquant)
-        h->mtype++;
-
-    put_bits(&s->pb, h261_mtype_bits[h->mtype], h261_mtype_code[h->mtype]);
-
-    h->mtype = h261_mtype_map[h->mtype];
-
-    if(IS_QUANT(h->mtype)){
-        ff_set_qscale(s,s->qscale+s->dquant);
-        put_bits(&s->pb, 5, s->qscale);
-    }
-
-    if(IS_16X16(h->mtype)){
-        mv_diff_x = (motion_x >> 1) - h->current_mv_x;
-        mv_diff_y = (motion_y >> 1) - h->current_mv_y;
-        h->current_mv_x = (motion_x >> 1);
-        h->current_mv_y = (motion_y >> 1);
-        h261_encode_motion(h,mv_diff_x);
-        h261_encode_motion(h,mv_diff_y);
-    }
-
-    h->previous_mba = h->current_mba;
-
-    if(HAS_CBP(h->mtype)){
-        assert(cbp>0);
-        put_bits(&s->pb,h261_cbp_tab[cbp-1][1],h261_cbp_tab[cbp-1][0]);
-    }
-    for(i=0; i<6; i++) {
-        /* encode each block */
-        h261_encode_block(h, block[i], i);
-    }
-
-    if ( ( h->current_mba == 11 ) || ( h->current_mba == 22 ) || ( h->current_mba == 33 ) || ( !IS_16X16 ( h->mtype ) )){
-        h->current_mv_x=0;
-        h->current_mv_y=0;
-    }
-}
-
-void ff_h261_encode_init(MpegEncContext *s){
-    static int done = 0;
-
-    if (!done) {
-        done = 1;
-        init_rl(&h261_rl_tcoeff, 1);
-    }
-
-    s->min_qcoeff= -127;
-    s->max_qcoeff=  127;
-    s->y_dc_scale_table=
-    s->c_dc_scale_table= ff_mpeg1_dc_scale_table;
-}
-
-
-/**
- * encodes a 8x8 block.
- * @param block the 8x8 block
- * @param n block index (0-3 are luma, 4-5 are chroma)
- */
-static void h261_encode_block(H261Context * h, DCTELEM * block, int n){
-    MpegEncContext * const s = &h->s;
-    int level, run, last, i, j, last_index, last_non_zero, sign, slevel, code;
-    RLTable *rl;
-
-    rl = &h261_rl_tcoeff;
-    if (s->mb_intra) {
-        /* DC coef */
-        level = block[0];
-        /* 255 cannot be represented, so we clamp */
-        if (level > 254) {
-            level = 254;
-            block[0] = 254;
-        }
-        /* 0 cannot be represented also */
-        else if (level < 1) {
-            level = 1;
-            block[0] = 1;
-        }
-        if (level == 128)
-            put_bits(&s->pb, 8, 0xff);
-        else
-            put_bits(&s->pb, 8, level);
-        i = 1;
-    } else if((block[0]==1 || block[0] == -1) && (s->block_last_index[n] > -1)){
-        //special case
-        put_bits(&s->pb,2,block[0]>0 ? 2 : 3 );
-        i = 1;
-    } else {
-        i = 0;
-    }
-
-    /* AC coefs */
-    last_index = s->block_last_index[n];
-    last_non_zero = i - 1;
-    for (; i <= last_index; i++) {
-        j = s->intra_scantable.permutated[i];
-        level = block[j];
-        if (level) {
-            run = i - last_non_zero - 1;
-            last = (i == last_index);
-            sign = 0;
-            slevel = level;
-            if (level < 0) {
-                sign = 1;
-                level = -level;
-            }
-            code = get_rl_index(rl, 0 /*no last in H.261, EOB is used*/, run, level);
-            if(run==0 && level < 16)
-            code+=1;
-            put_bits(&s->pb, rl->table_vlc[code][1], rl->table_vlc[code][0]);
-            if (code == rl->n) {
-                put_bits(&s->pb, 6, run);
-                assert(slevel != 0);
-                assert(level <= 127);
-                put_bits(&s->pb, 8, slevel & 0xff);
-            } else {
-                put_bits(&s->pb, 1, sign);
-            }
-            last_non_zero = i;
-        }
-    }
-    if(last_index > -1){
-        put_bits(&s->pb, rl->table_vlc[0][1], rl->table_vlc[0][0]);// END OF BLOCK
-    }
-}
-
-/***********************************************/
-/* decoding */
-
-static VLC h261_mba_vlc;
-static VLC h261_mtype_vlc;
-static VLC h261_mv_vlc;
-static VLC h261_cbp_vlc;
-
-static void h261_decode_init_vlc(H261Context *h){
-    static int done = 0;
-
-    if(!done){
-        done = 1;
-        init_vlc(&h261_mba_vlc, H261_MBA_VLC_BITS, 35,
-                 h261_mba_bits, 1, 1,
-                 h261_mba_code, 1, 1, 1);
-        init_vlc(&h261_mtype_vlc, H261_MTYPE_VLC_BITS, 10,
-                 h261_mtype_bits, 1, 1,
-                 h261_mtype_code, 1, 1, 1);
-        init_vlc(&h261_mv_vlc, H261_MV_VLC_BITS, 17,
-                 &h261_mv_tab[0][1], 2, 1,
-                 &h261_mv_tab[0][0], 2, 1, 1);
-        init_vlc(&h261_cbp_vlc, H261_CBP_VLC_BITS, 63,
-                 &h261_cbp_tab[0][1], 2, 1,
-                 &h261_cbp_tab[0][0], 2, 1, 1);
-        init_rl(&h261_rl_tcoeff, 1);
-        init_vlc_rl(&h261_rl_tcoeff, 1);
-    }
-}
-
-static int h261_decode_init(AVCodecContext *avctx){
-    H261Context *h= avctx->priv_data;
-    MpegEncContext * const s = &h->s;
-
-    // set defaults
-    MPV_decode_defaults(s);
-    s->avctx = avctx;
-
-    s->width  = s->avctx->coded_width;
-    s->height = s->avctx->coded_height;
-    s->codec_id = s->avctx->codec->id;
-
-    s->out_format = FMT_H261;
-    s->low_delay= 1;
-    avctx->pix_fmt= PIX_FMT_YUV420P;
-
-    s->codec_id= avctx->codec->id;
-
-    h261_decode_init_vlc(h);
-
-    h->gob_start_code_skipped = 0;
-
-    return 0;
-}
-
-/**
- * decodes the group of blocks header or slice header.
- * @return <0 if an error occured
- */
-static int h261_decode_gob_header(H261Context *h){
-    unsigned int val;
-    MpegEncContext * const s = &h->s;
-
-    if ( !h->gob_start_code_skipped ){
-        /* Check for GOB Start Code */
-        val = show_bits(&s->gb, 15);
-        if(val)
-            return -1;
-
-        /* We have a GBSC */
-        skip_bits(&s->gb, 16);
-    }
-
-    h->gob_start_code_skipped = 0;
-
-    h->gob_number = get_bits(&s->gb, 4); /* GN */
-    s->qscale = get_bits(&s->gb, 5); /* GQUANT */
-
-    /* Check if gob_number is valid */
-    if (s->mb_height==18){ //cif
-        if ((h->gob_number<=0) || (h->gob_number>12))
-            return -1;
-    }
-    else{ //qcif
-        if ((h->gob_number!=1) && (h->gob_number!=3) && (h->gob_number!=5))
-            return -1;
-    }
-
-    /* GEI */
-    while (get_bits1(&s->gb) != 0) {
-        skip_bits(&s->gb, 8);
-    }
-
-    if(s->qscale==0)
-        return -1;
-
-    // For the first transmitted macroblock in a GOB, MBA is the absolute address. For
-    // subsequent macroblocks, MBA is the difference between the absolute addresses of
-    // the macroblock and the last transmitted macroblock.
-    h->current_mba = 0;
-    h->mba_diff = 0;
-
-    return 0;
-}
-
-/**
- * decodes the group of blocks / video packet header.
- * @return <0 if no resync found
- */
-static int ff_h261_resync(H261Context *h){
-    MpegEncContext * const s = &h->s;
-    int left, ret;
-
-    if ( h->gob_start_code_skipped ){
-        ret= h261_decode_gob_header(h);
-        if(ret>=0)
-            return 0;
-    }
-    else{
-        if(show_bits(&s->gb, 15)==0){
-            ret= h261_decode_gob_header(h);
-            if(ret>=0)
-                return 0;
-        }
-        //ok, its not where its supposed to be ...
-        s->gb= s->last_resync_gb;
-        align_get_bits(&s->gb);
-        left= s->gb.size_in_bits - get_bits_count(&s->gb);
-
-        for(;left>15+1+4+5; left-=8){
-            if(show_bits(&s->gb, 15)==0){
-                GetBitContext bak= s->gb;
-
-                ret= h261_decode_gob_header(h);
-                if(ret>=0)
-                    return 0;
-
-                s->gb= bak;
-            }
-            skip_bits(&s->gb, 8);
-        }
-    }
-
-    return -1;
-}
-
-/**
- * decodes skipped macroblocks
- * @return 0
- */
-static int h261_decode_mb_skipped(H261Context *h, int mba1, int mba2 )
-{
-    MpegEncContext * const s = &h->s;
-    int i;
-
-    s->mb_intra = 0;
-
-    for(i=mba1; i<mba2; i++){
-        int j, xy;
-
-        s->mb_x= ((h->gob_number-1) % 2) * 11 + i % 11;
-        s->mb_y= ((h->gob_number-1) / 2) * 3 + i / 11;
-        xy = s->mb_x + s->mb_y * s->mb_stride;
-        ff_init_block_index(s);
-        ff_update_block_index(s);
-
-        for(j=0;j<6;j++)
-            s->block_last_index[j] = -1;
-
-        s->mv_dir = MV_DIR_FORWARD;
-        s->mv_type = MV_TYPE_16X16;
-        s->current_picture.mb_type[xy]= MB_TYPE_SKIP | MB_TYPE_16x16 | MB_TYPE_L0;
-        s->mv[0][0][0] = 0;
-        s->mv[0][0][1] = 0;
-        s->mb_skipped = 1;
-        h->mtype &= ~MB_TYPE_H261_FIL;
-
-        MPV_decode_mb(s, s->block);
-    }
-
-    return 0;
-}
-
-static int decode_mv_component(GetBitContext *gb, int v){
-    int mv_diff = get_vlc2(gb, h261_mv_vlc.table, H261_MV_VLC_BITS, 2);
-
-    /* check if mv_diff is valid */
-    if ( mv_diff < 0 )
-        return v;
-
-    mv_diff = mvmap[mv_diff];
-
-    if(mv_diff && !get_bits1(gb))
-        mv_diff= -mv_diff;
-
-    v += mv_diff;
-    if     (v <=-16) v+= 32;
-    else if(v >= 16) v-= 32;
-
-    return v;
-}
-
-static int h261_decode_mb(H261Context *h){
-    MpegEncContext * const s = &h->s;
-    int i, cbp, xy;
-
-    cbp = 63;
-    // Read mba
-    do{
-        h->mba_diff = get_vlc2(&s->gb, h261_mba_vlc.table, H261_MBA_VLC_BITS, 2);
-
-        /* Check for slice end */
-        /* NOTE: GOB can be empty (no MB data) or exist only of MBA_stuffing */
-        if (h->mba_diff == MBA_STARTCODE){ // start code
-            h->gob_start_code_skipped = 1;
-            return SLICE_END;
-        }
-    }
-    while( h->mba_diff == MBA_STUFFING ); // stuffing
-
-    if ( h->mba_diff < 0 ){
-        if ( get_bits_count(&s->gb) + 7 >= s->gb.size_in_bits )
-            return SLICE_END;
-
-        av_log(s->avctx, AV_LOG_ERROR, "illegal mba at %d %d\n", s->mb_x, s->mb_y);
-        return SLICE_ERROR;
-    }
-
-    h->mba_diff += 1;
-    h->current_mba += h->mba_diff;
-
-    if ( h->current_mba > MBA_STUFFING )
-        return SLICE_ERROR;
-
-    s->mb_x= ((h->gob_number-1) % 2) * 11 + ((h->current_mba-1) % 11);
-    s->mb_y= ((h->gob_number-1) / 2) * 3 + ((h->current_mba-1) / 11);
-    xy = s->mb_x + s->mb_y * s->mb_stride;
-    ff_init_block_index(s);
-    ff_update_block_index(s);
-
-    // Read mtype
-    h->mtype = get_vlc2(&s->gb, h261_mtype_vlc.table, H261_MTYPE_VLC_BITS, 2);
-    h->mtype = h261_mtype_map[h->mtype];
-
-    // Read mquant
-    if ( IS_QUANT ( h->mtype ) ){
-        ff_set_qscale(s, get_bits(&s->gb, 5));
-    }
-
-    s->mb_intra = IS_INTRA4x4(h->mtype);
-
-    // Read mv
-    if ( IS_16X16 ( h->mtype ) ){
-        // Motion vector data is included for all MC macroblocks. MVD is obtained from the macroblock vector by subtracting the
-        // vector of the preceding macroblock. For this calculation the vector of the preceding macroblock is regarded as zero in the
-        // following three situations:
-        // 1) evaluating MVD for macroblocks 1, 12 and 23;
-        // 2) evaluating MVD for macroblocks in which MBA does not represent a difference of 1;
-        // 3) MTYPE of the previous macroblock was not MC.
-        if ( ( h->current_mba == 1 ) || ( h->current_mba == 12 ) || ( h->current_mba == 23 ) ||
-             ( h->mba_diff != 1))
-        {
-            h->current_mv_x = 0;
-            h->current_mv_y = 0;
-        }
-
-        h->current_mv_x= decode_mv_component(&s->gb, h->current_mv_x);
-        h->current_mv_y= decode_mv_component(&s->gb, h->current_mv_y);
-    }else{
-        h->current_mv_x = 0;
-        h->current_mv_y = 0;
-    }
-
-    // Read cbp
-    if ( HAS_CBP( h->mtype ) ){
-        cbp = get_vlc2(&s->gb, h261_cbp_vlc.table, H261_CBP_VLC_BITS, 2) + 1;
-    }
-
-    if(s->mb_intra){
-        s->current_picture.mb_type[xy]= MB_TYPE_INTRA;
-        goto intra;
-    }
-
-    //set motion vectors
-    s->mv_dir = MV_DIR_FORWARD;
-    s->mv_type = MV_TYPE_16X16;
-    s->current_picture.mb_type[xy]= MB_TYPE_16x16 | MB_TYPE_L0;
-    s->mv[0][0][0] = h->current_mv_x * 2;//gets divided by 2 in motion compensation
-    s->mv[0][0][1] = h->current_mv_y * 2;
-
-intra:
-    /* decode each block */
-    if(s->mb_intra || HAS_CBP(h->mtype)){
-        s->dsp.clear_blocks(s->block[0]);
-        for (i = 0; i < 6; i++) {
-            if (h261_decode_block(h, s->block[i], i, cbp&32) < 0){
-                return SLICE_ERROR;
-            }
-            cbp+=cbp;
-        }
-    }else{
-        for (i = 0; i < 6; i++)
-            s->block_last_index[i]= -1;
-    }
-
-    MPV_decode_mb(s, s->block);
-
-    return SLICE_OK;
-}
-
-/**
- * decodes a macroblock
- * @return <0 if an error occured
- */
-static int h261_decode_block(H261Context * h, DCTELEM * block,
-                             int n, int coded)
-{
-    MpegEncContext * const s = &h->s;
-    int code, level, i, j, run;
-    RLTable *rl = &h261_rl_tcoeff;
-    const uint8_t *scan_table;
-
-    // For the variable length encoding there are two code tables, one being used for
-    // the first transmitted LEVEL in INTER, INTER+MC and INTER+MC+FIL blocks, the second
-    // for all other LEVELs except the first one in INTRA blocks which is fixed length
-    // coded with 8 bits.
-    // NOTE: the two code tables only differ in one VLC so we handle that manually.
-    scan_table = s->intra_scantable.permutated;
-    if (s->mb_intra){
-        /* DC coef */
-        level = get_bits(&s->gb, 8);
-        // 0 (00000000b) and -128 (10000000b) are FORBIDDEN
-        if((level&0x7F) == 0){
-            av_log(s->avctx, AV_LOG_ERROR, "illegal dc %d at %d %d\n", level, s->mb_x, s->mb_y);
-            return -1;
-        }
-        // The code 1000 0000 is not used, the reconstruction level of 1024 being coded as 1111 1111.
-        if (level == 255)
-            level = 128;
-        block[0] = level;
-        i = 1;
-    }else if(coded){
-        // Run  Level   Code
-        // EOB                  Not possible for first level when cbp is available (that's why the table is different)
-        // 0    1               1s
-        // *    *               0*
-        int check = show_bits(&s->gb, 2);
-        i = 0;
-        if ( check & 0x2 ){
-            skip_bits(&s->gb, 2);
-            block[0] = ( check & 0x1 ) ? -1 : 1;
-            i = 1;
-        }
-    }else{
-        i = 0;
-    }
-    if(!coded){
-        s->block_last_index[n] = i - 1;
-        return 0;
-    }
-    for(;;){
-        code = get_vlc2(&s->gb, rl->vlc.table, TCOEFF_VLC_BITS, 2);
-        if (code < 0){
-            av_log(s->avctx, AV_LOG_ERROR, "illegal ac vlc code at %dx%d\n", s->mb_x, s->mb_y);
-            return -1;
-        }
-        if (code == rl->n) {
-            /* escape */
-            // The remaining combinations of (run, level) are encoded with a 20-bit word consisting of 6 bits escape, 6 bits run and 8 bits level.
-            run = get_bits(&s->gb, 6);
-            level = get_sbits(&s->gb, 8);
-        }else if(code == 0){
-            break;
-        }else{
-            run = rl->table_run[code];
-            level = rl->table_level[code];
-            if (get_bits1(&s->gb))
-                level = -level;
-        }
-        i += run;
-        if (i >= 64){
-            av_log(s->avctx, AV_LOG_ERROR, "run overflow at %dx%d\n", s->mb_x, s->mb_y);
-            return -1;
-        }
-        j = scan_table[i];
-        block[j] = level;
-        i++;
-    }
-    s->block_last_index[n] = i-1;
-    return 0;
-}
-
-/**
- * decodes the H261 picture header.
- * @return <0 if no startcode found
- */
-static int h261_decode_picture_header(H261Context *h){
-    MpegEncContext * const s = &h->s;
-    int format, i;
-    uint32_t startcode= 0;
-
-    for(i= s->gb.size_in_bits - get_bits_count(&s->gb); i>24; i-=1){
-        startcode = ((startcode << 1) | get_bits(&s->gb, 1)) & 0x000FFFFF;
-
-        if(startcode == 0x10)
-            break;
-    }
-
-    if (startcode != 0x10){
-        av_log(s->avctx, AV_LOG_ERROR, "Bad picture start code\n");
-        return -1;
-    }
-
-    /* temporal reference */
-    i= get_bits(&s->gb, 5); /* picture timestamp */
-    if(i < (s->picture_number&31))
-        i += 32;
-    s->picture_number = (s->picture_number&~31) + i;
-
-    s->avctx->time_base= (AVRational){1001, 30000};
-    s->current_picture.pts= s->picture_number;
-
-
-    /* PTYPE starts here */
-    skip_bits1(&s->gb); /* split screen off */
-    skip_bits1(&s->gb); /* camera  off */
-    skip_bits1(&s->gb); /* freeze picture release off */
-
-    format = get_bits1(&s->gb);
-
-    //only 2 formats possible
-    if (format == 0){//QCIF
-        s->width = 176;
-        s->height = 144;
-        s->mb_width = 11;
-        s->mb_height = 9;
-    }else{//CIF
-        s->width = 352;
-        s->height = 288;
-        s->mb_width = 22;
-        s->mb_height = 18;
-    }
-
-    s->mb_num = s->mb_width * s->mb_height;
-
-    skip_bits1(&s->gb); /* still image mode off */
-    skip_bits1(&s->gb); /* Reserved */
-
-    /* PEI */
-    while (get_bits1(&s->gb) != 0){
-        skip_bits(&s->gb, 8);
-    }
-
-    // h261 has no I-FRAMES, but if we pass I_TYPE for the first frame, the codec crashes if it does
-    // not contain all I-blocks (e.g. when a packet is lost)
-    s->pict_type = P_TYPE;
-
-    h->gob_number = 0;
-    return 0;
-}
-
-static int h261_decode_gob(H261Context *h){
-    MpegEncContext * const s = &h->s;
-
-    ff_set_qscale(s, s->qscale);
-
-    /* decode mb's */
-    while(h->current_mba <= MBA_STUFFING)
-    {
-        int ret;
-        /* DCT & quantize */
-        ret= h261_decode_mb(h);
-        if(ret<0){
-            if(ret==SLICE_END){
-                h261_decode_mb_skipped(h, h->current_mba, 33);
-                return 0;
-            }
-            av_log(s->avctx, AV_LOG_ERROR, "Error at MB: %d\n", s->mb_x + s->mb_y*s->mb_stride);
-            return -1;
-        }
-
-        h261_decode_mb_skipped(h, h->current_mba-h->mba_diff, h->current_mba-1);
-    }
-
-    return -1;
-}
-
-#ifdef CONFIG_H261_PARSER
-static int h261_find_frame_end(ParseContext *pc, AVCodecContext* avctx, const uint8_t *buf, int buf_size){
-    int vop_found, i, j;
-    uint32_t state;
-
-    vop_found= pc->frame_start_found;
-    state= pc->state;
-
-    for(i=0; i<buf_size && !vop_found; i++){
-        state= (state<<8) | buf[i];
-        for(j=0; j<8; j++){
-            if(((state>>j)&0xFFFFF) == 0x00010){
-                vop_found=1;
-                break;
-            }
-        }
-    }
-    if(vop_found){
-        for(; i<buf_size; i++){
-            state= (state<<8) | buf[i];
-            for(j=0; j<8; j++){
-                if(((state>>j)&0xFFFFF) == 0x00010){
-                    pc->frame_start_found=0;
-                    pc->state= state>>(2*8);
-                    return i-1;
-                }
-            }
-        }
-    }
-
-    pc->frame_start_found= vop_found;
-    pc->state= state;
-    return END_NOT_FOUND;
-}
-
-static int h261_parse(AVCodecParserContext *s,
-                      AVCodecContext *avctx,
-                      uint8_t **poutbuf, int *poutbuf_size,
-                      const uint8_t *buf, int buf_size)
-{
-    ParseContext *pc = s->priv_data;
-    int next;
-
-    next= h261_find_frame_end(pc,avctx, buf, buf_size);
-    if (ff_combine_frame(pc, next, (uint8_t **)&buf, &buf_size) < 0) {
-        *poutbuf = NULL;
-        *poutbuf_size = 0;
-        return buf_size;
-    }
-    *poutbuf = (uint8_t *)buf;
-    *poutbuf_size = buf_size;
-    return next;
-}
-#endif
-
-/**
- * returns the number of bytes consumed for building the current frame
- */
-static int get_consumed_bytes(MpegEncContext *s, int buf_size){
-    int pos= get_bits_count(&s->gb)>>3;
-    if(pos==0) pos=1; //avoid infinite loops (i doubt thats needed but ...)
-    if(pos+10>buf_size) pos=buf_size; // oops ;)
-
-    return pos;
-}
-
-static int h261_decode_frame(AVCodecContext *avctx,
-                             void *data, int *data_size,
-                             uint8_t *buf, int buf_size)
-{
-    H261Context *h= avctx->priv_data;
-    MpegEncContext *s = &h->s;
-    int ret;
-    AVFrame *pict = data;
-
-#ifdef DEBUG
-    av_log(avctx, AV_LOG_DEBUG, "*****frame %d size=%d\n", avctx->frame_number, buf_size);
-    av_log(avctx, AV_LOG_DEBUG, "bytes=%x %x %x %x\n", buf[0], buf[1], buf[2], buf[3]);
-#endif
-    s->flags= avctx->flags;
-    s->flags2= avctx->flags2;
-
-    h->gob_start_code_skipped=0;
-
-retry:
-
-    init_get_bits(&s->gb, buf, buf_size*8);
-
-    if(!s->context_initialized){
-        if (MPV_common_init(s) < 0) //we need the idct permutaton for reading a custom matrix
-            return -1;
-    }
-
-    //we need to set current_picture_ptr before reading the header, otherwise we cant store anyting im there
-    if(s->current_picture_ptr==NULL || s->current_picture_ptr->data[0]){
-        int i= ff_find_unused_picture(s, 0);
-        s->current_picture_ptr= &s->picture[i];
-    }
-
-    ret = h261_decode_picture_header(h);
-
-    /* skip if the header was thrashed */
-    if (ret < 0){
-        av_log(s->avctx, AV_LOG_ERROR, "header damaged\n");
-        return -1;
-    }
-
-    if (s->width != avctx->coded_width || s->height != avctx->coded_height){
-        ParseContext pc= s->parse_context; //FIXME move these demuxng hack to avformat
-        s->parse_context.buffer=0;
-        MPV_common_end(s);
-        s->parse_context= pc;
-    }
-    if (!s->context_initialized) {
-        avcodec_set_dimensions(avctx, s->width, s->height);
-
-        goto retry;
-    }
-
-    // for hurry_up==5
-    s->current_picture.pict_type= s->pict_type;
-    s->current_picture.key_frame= s->pict_type == I_TYPE;
-
-    /* skip everything if we are in a hurry>=5 */
-    if(avctx->hurry_up>=5) return get_consumed_bytes(s, buf_size);
-    if(  (avctx->skip_frame >= AVDISCARD_NONREF && s->pict_type==B_TYPE)
-       ||(avctx->skip_frame >= AVDISCARD_NONKEY && s->pict_type!=I_TYPE)
-       || avctx->skip_frame >= AVDISCARD_ALL)
-        return get_consumed_bytes(s, buf_size);
-
-    if(MPV_frame_start(s, avctx) < 0)
-        return -1;
-
-    ff_er_frame_start(s);
-
-    /* decode each macroblock */
-    s->mb_x=0;
-    s->mb_y=0;
-
-    while(h->gob_number < (s->mb_height==18 ? 12 : 5)){
-        if(ff_h261_resync(h)<0)
-            break;
-        h261_decode_gob(h);
-    }
-    MPV_frame_end(s);
-
-assert(s->current_picture.pict_type == s->current_picture_ptr->pict_type);
-assert(s->current_picture.pict_type == s->pict_type);
-    *pict= *(AVFrame*)s->current_picture_ptr;
-    ff_print_debug_info(s, pict);
-
-    *data_size = sizeof(AVFrame);
-
-    return get_consumed_bytes(s, buf_size);
-}
-
-static int h261_decode_end(AVCodecContext *avctx)
-{
-    H261Context *h= avctx->priv_data;
-    MpegEncContext *s = &h->s;
-
-    MPV_common_end(s);
-    return 0;
-}
-
-#ifdef CONFIG_ENCODERS
-AVCodec h261_encoder = {
-    "h261",
-    CODEC_TYPE_VIDEO,
-    CODEC_ID_H261,
-    sizeof(H261Context),
-    MPV_encode_init,
-    MPV_encode_picture,
-    MPV_encode_end,
-    .pix_fmts= (enum PixelFormat[]){PIX_FMT_YUV420P, -1},
-};
-#endif
-
-AVCodec h261_decoder = {
-    "h261",
-    CODEC_TYPE_VIDEO,
-    CODEC_ID_H261,
-    sizeof(H261Context),
-    h261_decode_init,
-    NULL,
-    h261_decode_end,
-    h261_decode_frame,
-    CODEC_CAP_DR1,
-};
-
-#ifdef CONFIG_H261_PARSER
-AVCodecParser h261_parser = {
-    { CODEC_ID_H261 },
-    sizeof(ParseContext),
-    NULL,
-    h261_parse,
-    ff_parse_close,
-};
-#endif
--- a/src/ffmpeg/libavcodec/h261data.h	Mon Mar 12 11:37:35 2007 -0700
+++ /dev/null	Thu Jan 01 00:00:00 1970 +0000
@@ -1,157 +0,0 @@
-/*
- * copyright (c) 2002-2004 Michael Niedermayer <michaelni@gmx.at>
- * copyright (c) 2004 Maarten Daniels
- *
- * 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 h261data.h
- * H.261 tables.
- */
-#define MB_TYPE_H261_FIL 0x800000
-
-// H.261 VLC table for macroblock addressing
-const uint8_t h261_mba_code[35] = {
-     1,  3,  2,  3,
-     2,  3,  2,  7,
-     6, 11, 10,  9,
-     8,  7,  6, 23,
-    22, 21, 20, 19,
-    18, 35, 34, 33,
-    32, 31, 30, 29,
-    28, 27, 26, 25,
-    24,
-    15,           //(MBA stuffing)
-    1             //(start code)
-};
-
-const uint8_t h261_mba_bits[35] = {
-     1,  3,  3,  4,
-     4,  5,  5,  7,
-     7,  8,  8,  8,
-     8,  8,  8, 10,
-    10, 10, 10, 10,
-    10, 11, 11, 11,
-    11, 11, 11, 11,
-    11, 11, 11, 11,
-    11,
-    11,           //(MBA stuffing)
-    16            //(start code)
-};
-
-//H.261 VLC table for macroblock type
-const uint8_t h261_mtype_code[10] = {
-    1,  1,  1,  1,
-    1,  1,  1,  1,
-    1,  1
-};
-
-const uint8_t h261_mtype_bits[10] = {
-    4,  7,  1,  5,
-    9,  8, 10,  3,
-    2,  6
-};
-
-static const int h261_mtype_map[10]= {
-        MB_TYPE_INTRA4x4,
-        MB_TYPE_INTRA4x4  |  MB_TYPE_QUANT,
-                                               MB_TYPE_CBP,
-                             MB_TYPE_QUANT  |  MB_TYPE_CBP,
-                                                               MB_TYPE_16x16,
-                                               MB_TYPE_CBP  |  MB_TYPE_16x16,
-                             MB_TYPE_QUANT  |  MB_TYPE_CBP  |  MB_TYPE_16x16,
-                                                               MB_TYPE_16x16  |  MB_TYPE_H261_FIL,
-                                               MB_TYPE_CBP  |  MB_TYPE_16x16  |  MB_TYPE_H261_FIL,
-                             MB_TYPE_QUANT  |  MB_TYPE_CBP  |  MB_TYPE_16x16  |  MB_TYPE_H261_FIL
-};
-
-//H.261 VLC table for motion vectors
-const uint8_t h261_mv_tab[17][2] = {
-    {1,1}, {1,2}, {1,3}, {1,4}, {3,6}, {5,7}, {4,7}, {3,7},
-    {11,9}, {10,9}, {9,9}, {17,10}, {16,10}, {15,10}, {14,10}, {13,10}, {12,10}
-};
-
-static const int mvmap[17] =
-{
-    0, -1, -2, -3, -4, -5, -6, -7, -8, -9, -10, -11, -12, -13, -14, -15, -16
-};
-
-//H.261 VLC table for coded block pattern
-const uint8_t h261_cbp_tab[63][2] =
-{
-    {11,5}, {9,5}, {13,6}, {13,4}, {23,7}, {19,7}, {31,8}, {12,4},
-    {22,7}, {18,7}, {30,8}, {19,5}, {27,8}, {23,8}, {19,8}, {11,4},
-    {21,7}, {17,7}, {29,8}, {17,5}, {25,8}, {21,8}, {17,8}, {15,6},
-    {15,8}, {13,8}, {3,9}, {15,5}, {11,8}, {7,8}, {7,9}, {10,4},
-    {20,7}, {16,7}, {28,8}, {14,6}, {14,8}, {12,8}, {2,9}, {16,5},
-    {24,8}, {20,8}, {16,8}, {14,5}, {10,8}, {6,8}, {6,9}, {18,5},
-    {26,8}, {22,8}, {18,8}, {13,5}, {9,8}, {5,8}, {5,9}, {12,5},
-    {8,8}, {4,8}, {4,9}, {7,3}, {10,5}, {8,5}, {12,6}
-};
-
-//H.261 VLC table for transform coefficients
-const uint16_t h261_tcoeff_vlc[65][2] = {
-{ 0x2, 2 }, { 0x3, 2 },{ 0x4, 4 },{ 0x5, 5 },
-{ 0x6, 7 },{ 0x26, 8 },{ 0x21, 8 },{ 0xa, 10 },
-{ 0x1d, 12 },{ 0x18, 12 },{ 0x13, 12 },{ 0x10 , 12 },
-{ 0x1a, 13},{ 0x19, 13 }, { 0x18, 13 }, { 0x17, 13 },
-{ 0x3, 3 }, { 0x6, 6 }, { 0x25 , 8 }, { 0xc, 10 },
-{ 0x1b, 12 }, { 0x16, 13 }, { 0x15, 13 }, { 0x5, 4},
-{ 0x4, 7}, { 0xb, 10 }, { 0x14, 12 }, { 0x14, 13 },
-{ 0x7, 5 }, { 0x24, 8 }, { 0x1c, 12 }, { 0x13, 13 },
-{ 0x6, 5 }, { 0xf, 10 }, { 0x12, 12}, { 0x7, 6},
-{ 0x9 , 10 }, { 0x12, 13 }, { 0x5, 6 }, { 0x1e, 12 },
-{ 0x4, 6 }, { 0x15, 12 }, { 0x7, 7 }, { 0x11, 12},
-{ 0x5, 7 }, { 0x11, 13 }, { 0x27, 8 }, { 0x10, 13 },
-{ 0x23, 8 }, { 0x22, 8 }, { 0x20, 8 }, { 0xe , 10 },
-{ 0xd, 10 }, { 0x8, 10 },{ 0x1f, 12 }, { 0x1a, 12 },
-{ 0x19, 12 }, { 0x17, 12 }, { 0x16, 12}, { 0x1f, 13},
-{ 0x1e, 13 }, { 0x1d, 13 }, { 0x1c, 13}, { 0x1b, 13},
-{ 0x1, 6 }                                             //escape
-};
-
-const int8_t h261_tcoeff_level[64] = {
-    0,  1,  2,  3,  4,  5,  6,  7,
-    8,  9, 10, 11, 12, 13, 14, 15,
-    1,  2,  3,  4,  5,  6,  7,  1,
-    2,  3,  4,  5,  1,  2,  3,  4,
-    1,  2,  3,  1,  2,  3,  1,  2,
-    1,  2,  1,  2,  1,  2,  1,  2,
-    1,  1,  1,  1,  1,  1,  1,  1,
-    1,  1,  1,  1,  1,  1,  1,  1
-};
-
-const int8_t h261_tcoeff_run[64] = {
-    0,
-    0,  0,  0,  0,  0,  0,  0,  0,
-    0,  0,  0,  0,  0,  0,  0,  1,
-    1,  1,  1,  1,  1,  1,  2,  2,
-    2,  2,  2,  3,  3,  3,  3,  4,
-    4,  4,  5,  5,  5,  6,  6,  7,
-    7,  8,  8,  9,  9, 10, 10, 11,
-   12, 13, 14, 15, 16, 17, 18, 19,
-   20, 21, 22, 23, 24, 25, 26
-};
-
-static RLTable h261_rl_tcoeff = {
-    64,
-    64,
-    h261_tcoeff_vlc,
-    h261_tcoeff_run,
-    h261_tcoeff_level,
-};
--- a/src/ffmpeg/libavcodec/h263.c	Mon Mar 12 11:37:35 2007 -0700
+++ /dev/null	Thu Jan 01 00:00:00 1970 +0000
@@ -1,6291 +0,0 @@
-/*
- * H263/MPEG4 backend for ffmpeg encoder and decoder
- * Copyright (c) 2000,2001 Fabrice Bellard.
- * H263+ support.
- * Copyright (c) 2001 Juan J. Sierralta P.
- * Copyright (c) 2002-2004 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
- *
- * ac prediction encoding, b-frame support, error resilience, optimizations,
- * qpel decoding, gmc decoding, interlaced decoding,
- * by Michael Niedermayer <michaelni@gmx.at>
- */
-
-/**
- * @file h263.c
- * h263/mpeg4 codec.
- */
-
-//#define DEBUG
-#include <limits.h>
-
-#include "common.h"
-#include "dsputil.h"
-#include "avcodec.h"
-#include "mpegvideo.h"
-#include "h263data.h"
-#include "mpeg4data.h"
-
-//#undef NDEBUG
-//#include <assert.h>
-
-#define INTRA_MCBPC_VLC_BITS 6
-#define INTER_MCBPC_VLC_BITS 7
-#define CBPY_VLC_BITS 6
-#define MV_VLC_BITS 9
-#define DC_VLC_BITS 9
-#define SPRITE_TRAJ_VLC_BITS 6
-#define MB_TYPE_B_VLC_BITS 4
-#define TEX_VLC_BITS 9
-#define H263_MBTYPE_B_VLC_BITS 6
-#define CBPC_B_VLC_BITS 3
-
-#ifdef CONFIG_ENCODERS
-static void h263_encode_block(MpegEncContext * s, DCTELEM * block,
-                              int n);
-static void h263p_encode_umotion(MpegEncContext * s, int val);
-static inline void mpeg4_encode_block(MpegEncContext * s, DCTELEM * block,
-                               int n, int dc, uint8_t *scan_table,
-                               PutBitContext *dc_pb, PutBitContext *ac_pb);
-static int mpeg4_get_block_length(MpegEncContext * s, DCTELEM * block, int n, int intra_dc,
-                                  uint8_t *scan_table);
-#endif
-
-static int h263_decode_motion(MpegEncContext * s, int pred, int fcode);
-static int h263p_decode_umotion(MpegEncContext * s, int pred);
-static int h263_decode_block(MpegEncContext * s, DCTELEM * block,
-                             int n, int coded);
-static inline int mpeg4_decode_dc(MpegEncContext * s, int n, int *dir_ptr);
-static inline int mpeg4_decode_block(MpegEncContext * s, DCTELEM * block,
-                              int n, int coded, int intra, int rvlc);
-#ifdef CONFIG_ENCODERS
-static int h263_pred_dc(MpegEncContext * s, int n, int16_t **dc_val_ptr);
-static void mpeg4_encode_visual_object_header(MpegEncContext * s);
-static void mpeg4_encode_vol_header(MpegEncContext * s, int vo_number, int vol_number);
-#endif //CONFIG_ENCODERS
-static void mpeg4_decode_sprite_trajectory(MpegEncContext * s, GetBitContext *gb);
-static inline int ff_mpeg4_pred_dc(MpegEncContext * s, int n, int level, int *dir_ptr, int encoding);
-
-#ifdef CONFIG_ENCODERS
-static uint8_t uni_DCtab_lum_len[512];
-static uint8_t uni_DCtab_chrom_len[512];
-static uint16_t uni_DCtab_lum_bits[512];
-static uint16_t uni_DCtab_chrom_bits[512];
-
-static uint8_t (*mv_penalty)[MAX_MV*2+1]= NULL;
-static uint8_t fcode_tab[MAX_MV*2+1];
-static uint8_t umv_fcode_tab[MAX_MV*2+1];
-
-static uint32_t uni_mpeg4_intra_rl_bits[64*64*2*2];
-static uint8_t  uni_mpeg4_intra_rl_len [64*64*2*2];
-static uint32_t uni_mpeg4_inter_rl_bits[64*64*2*2];
-static uint8_t  uni_mpeg4_inter_rl_len [64*64*2*2];
-static uint8_t  uni_h263_intra_aic_rl_len [64*64*2*2];
-static uint8_t  uni_h263_inter_rl_len [64*64*2*2];
-//#define UNI_MPEG4_ENC_INDEX(last,run,level) ((last)*128 + (run)*256 + (level))
-//#define UNI_MPEG4_ENC_INDEX(last,run,level) ((last)*128*64 + (run) + (level)*64)
-#define UNI_MPEG4_ENC_INDEX(last,run,level) ((last)*128*64 + (run)*128 + (level))
-
-/* mpeg4
-inter
-max level: 24/6
-max run: 53/63
-
-intra
-max level: 53/16
-max run: 29/41
-*/
-#endif
-
-#if 0 //3IV1 is quite rare and it slows things down a tiny bit
-#define IS_3IV1 s->avctx->codec_tag == ff_get_fourcc("3IV1")
-#else
-#define IS_3IV1 0
-#endif
-
-int h263_get_picture_format(int width, int height)
-{
-    int format;
-
-    if (width == 128 && height == 96)
-        format = 1;
-    else if (width == 176 && height == 144)
-        format = 2;
-    else if (width == 352 && height == 288)
-        format = 3;
-    else if (width == 704 && height == 576)
-        format = 4;
-    else if (width == 1408 && height == 1152)
-        format = 5;
-    else
-        format = 7;
-    return format;
-}
-
-#ifdef CONFIG_ENCODERS
-
-static void aspect_to_info(MpegEncContext * s, AVRational aspect){
-    int i;
-
-    if(aspect.num==0) aspect= (AVRational){1,1};
-
-    for(i=1; i<6; i++){
-        if(av_cmp_q(pixel_aspect[i], aspect) == 0){
-            s->aspect_ratio_info=i;
-            return;
-        }
-    }
-
-    s->aspect_ratio_info= FF_ASPECT_EXTENDED;
-}
-
-void ff_flv_encode_picture_header(MpegEncContext * s, int picture_number)
-{
-      int format;
-
-      align_put_bits(&s->pb);
-
-      put_bits(&s->pb, 17, 1);
-      put_bits(&s->pb, 5, (s->h263_flv-1)); /* 0: h263 escape codes 1: 11-bit escape codes */
-      put_bits(&s->pb, 8, (((int64_t)s->picture_number * 30 * s->avctx->time_base.num) / //FIXME use timestamp
-                           s->avctx->time_base.den) & 0xff); /* TemporalReference */
-      if (s->width == 352 && s->height == 288)
-        format = 2;
-      else if (s->width == 176 && s->height == 144)
-        format = 3;
-      else if (s->width == 128 && s->height == 96)
-        format = 4;
-      else if (s->width == 320 && s->height == 240)
-        format = 5;
-      else if (s->width == 160 && s->height == 120)
-        format = 6;
-      else if (s->width <= 255 && s->height <= 255)
-        format = 0; /* use 1 byte width & height */
-      else
-        format = 1; /* use 2 bytes width & height */
-      put_bits(&s->pb, 3, format); /* PictureSize */
-      if (format == 0) {
-        put_bits(&s->pb, 8, s->width);
-        put_bits(&s->pb, 8, s->height);
-      } else if (format == 1) {
-        put_bits(&s->pb, 16, s->width);
-        put_bits(&s->pb, 16, s->height);
-      }
-      put_bits(&s->pb, 2, s->pict_type == P_TYPE); /* PictureType */
-      put_bits(&s->pb, 1, 1); /* DeblockingFlag: on */
-      put_bits(&s->pb, 5, s->qscale); /* Quantizer */
-      put_bits(&s->pb, 1, 0); /* ExtraInformation */
-
-      if(s->h263_aic){
-        s->y_dc_scale_table=
-          s->c_dc_scale_table= ff_aic_dc_scale_table;
-      }else{
-        s->y_dc_scale_table=
-          s->c_dc_scale_table= ff_mpeg1_dc_scale_table;
-      }
-}
-
-void h263_encode_picture_header(MpegEncContext * s, int picture_number)
-{
-    int format, coded_frame_rate, coded_frame_rate_base, i, temp_ref;
-    int best_clock_code=1;
-    int best_divisor=60;
-    int best_error= INT_MAX;
-
-    if(s->h263_plus){
-        for(i=0; i<2; i++){
-            int div, error;
-            div= (s->avctx->time_base.num*1800000LL + 500LL*s->avctx->time_base.den) / ((1000LL+i)*s->avctx->time_base.den);
-            div= clip(1, div, 127);
-            error= FFABS(s->avctx->time_base.num*1800000LL - (1000LL+i)*s->avctx->time_base.den*div);
-            if(error < best_error){
-                best_error= error;
-                best_divisor= div;
-                best_clock_code= i;
-            }
-        }
-    }
-    s->custom_pcf= best_clock_code!=1 || best_divisor!=60;
-    coded_frame_rate= 1800000;
-    coded_frame_rate_base= (1000+best_clock_code)*best_divisor;
-
-    align_put_bits(&s->pb);
-
-    /* Update the pointer to last GOB */
-    s->ptr_lastgob = pbBufPtr(&s->pb);
-    put_bits(&s->pb, 22, 0x20); /* PSC */
-    temp_ref= s->picture_number * (int64_t)coded_frame_rate * s->avctx->time_base.num / //FIXME use timestamp
-                         (coded_frame_rate_base * (int64_t)s->avctx->time_base.den);
-    put_bits(&s->pb, 8, temp_ref & 0xff); /* TemporalReference */
-
-    put_bits(&s->pb, 1, 1);     /* marker */
-    put_bits(&s->pb, 1, 0);     /* h263 id */
-    put_bits(&s->pb, 1, 0);     /* split screen off */
-    put_bits(&s->pb, 1, 0);     /* camera  off */
-    put_bits(&s->pb, 1, 0);     /* freeze picture release off */
-
-    format = h263_get_picture_format(s->width, s->height);
-    if (!s->h263_plus) {
-        /* H.263v1 */
-        put_bits(&s->pb, 3, format);
-        put_bits(&s->pb, 1, (s->pict_type == P_TYPE));
-        /* By now UMV IS DISABLED ON H.263v1, since the restrictions
-        of H.263v1 UMV implies to check the predicted MV after
-        calculation of the current MB to see if we're on the limits */
-        put_bits(&s->pb, 1, 0);         /* Unrestricted Motion Vector: off */
-        put_bits(&s->pb, 1, 0);         /* SAC: off */
-        put_bits(&s->pb, 1, s->obmc);   /* Advanced Prediction */
-        put_bits(&s->pb, 1, 0);         /* only I/P frames, no PB frame */
-        put_bits(&s->pb, 5, s->qscale);
-        put_bits(&s->pb, 1, 0);         /* Continuous Presence Multipoint mode: off */
-    } else {
-        int ufep=1;
-        /* H.263v2 */
-        /* H.263 Plus PTYPE */
-
-        put_bits(&s->pb, 3, 7);
-        put_bits(&s->pb,3,ufep); /* Update Full Extended PTYPE */
-        if (format == 7)
-            put_bits(&s->pb,3,6); /* Custom Source Format */
-        else
-            put_bits(&s->pb, 3, format);
-
-        put_bits(&s->pb,1, s->custom_pcf);
-        put_bits(&s->pb,1, s->umvplus); /* Unrestricted Motion Vector */
-        put_bits(&s->pb,1,0); /* SAC: off */
-        put_bits(&s->pb,1,s->obmc); /* Advanced Prediction Mode */
-        put_bits(&s->pb,1,s->h263_aic); /* Advanced Intra Coding */
-        put_bits(&s->pb,1,s->loop_filter); /* Deblocking Filter */
-        put_bits(&s->pb,1,s->h263_slice_structured); /* Slice Structured */
-        put_bits(&s->pb,1,0); /* Reference Picture Selection: off */
-        put_bits(&s->pb,1,0); /* Independent Segment Decoding: off */
-        put_bits(&s->pb,1,s->alt_inter_vlc); /* Alternative Inter VLC */
-        put_bits(&s->pb,1,s->modified_quant); /* Modified Quantization: */
-        put_bits(&s->pb,1,1); /* "1" to prevent start code emulation */
-        put_bits(&s->pb,3,0); /* Reserved */
-
-        put_bits(&s->pb, 3, s->pict_type == P_TYPE);
-
-        put_bits(&s->pb,1,0); /* Reference Picture Resampling: off */
-        put_bits(&s->pb,1,0); /* Reduced-Resolution Update: off */
-        put_bits(&s->pb,1,s->no_rounding); /* Rounding Type */
-        put_bits(&s->pb,2,0); /* Reserved */
-        put_bits(&s->pb,1,1); /* "1" to prevent start code emulation */
-
-        /* This should be here if PLUSPTYPE */
-        put_bits(&s->pb, 1, 0); /* Continuous Presence Multipoint mode: off */
-
-                if (format == 7) {
-            /* Custom Picture Format (CPFMT) */
-            aspect_to_info(s, s->avctx->sample_aspect_ratio);
-
-            put_bits(&s->pb,4,s->aspect_ratio_info);
-            put_bits(&s->pb,9,(s->width >> 2) - 1);
-            put_bits(&s->pb,1,1); /* "1" to prevent start code emulation */
-            put_bits(&s->pb,9,(s->height >> 2));
-            if (s->aspect_ratio_info == FF_ASPECT_EXTENDED){
-                put_bits(&s->pb, 8, s->avctx->sample_aspect_ratio.num);
-                put_bits(&s->pb, 8, s->avctx->sample_aspect_ratio.den);
-            }
-        }
-        if(s->custom_pcf){
-            if(ufep){
-                put_bits(&s->pb, 1, best_clock_code);
-                put_bits(&s->pb, 7, best_divisor);
-            }
-            put_bits(&s->pb, 2, (temp_ref>>8)&3);
-        }
-
-        /* Unlimited Unrestricted Motion Vectors Indicator (UUI) */
-        if (s->umvplus)
-//            put_bits(&s->pb,1,1); /* Limited according tables of Annex D */
-//FIXME check actual requested range
-            put_bits(&s->pb,2,1); /* unlimited */
-        if(s->h263_slice_structured)
-            put_bits(&s->pb,2,0); /* no weird submodes */
-
-        put_bits(&s->pb, 5, s->qscale);
-    }
-
-    put_bits(&s->pb, 1, 0);     /* no PEI */
-
-    if(s->h263_slice_structured){
-        put_bits(&s->pb, 1, 1);
-
-        assert(s->mb_x == 0 && s->mb_y == 0);
-        ff_h263_encode_mba(s);
-
-        put_bits(&s->pb, 1, 1);
-    }
-
-    if(s->h263_aic){
-         s->y_dc_scale_table=
-         s->c_dc_scale_table= ff_aic_dc_scale_table;
-    }else{
-        s->y_dc_scale_table=
-        s->c_dc_scale_table= ff_mpeg1_dc_scale_table;
-    }
-}
-
-/**
- * Encodes a group of blocks header.
- */
-void h263_encode_gob_header(MpegEncContext * s, int mb_line)
-{
-    put_bits(&s->pb, 17, 1); /* GBSC */
-
-    if(s->h263_slice_structured){
-        put_bits(&s->pb, 1, 1);
-
-        ff_h263_encode_mba(s);
-
-        if(s->mb_num > 1583)
-            put_bits(&s->pb, 1, 1);
-        put_bits(&s->pb, 5, s->qscale); /* GQUANT */
-        put_bits(&s->pb, 1, 1);
-        put_bits(&s->pb, 2, s->pict_type == I_TYPE); /* GFID */
-    }else{
-        int gob_number= mb_line / s->gob_index;
-
-        put_bits(&s->pb, 5, gob_number); /* GN */
-        put_bits(&s->pb, 2, s->pict_type == I_TYPE); /* GFID */
-        put_bits(&s->pb, 5, s->qscale); /* GQUANT */
-    }
-}
-
-static inline int get_block_rate(MpegEncContext * s, DCTELEM block[64], int block_last_index, uint8_t scantable[64]){
-    int last=0;
-    int j;
-    int rate=0;
-
-    for(j=1; j<=block_last_index; j++){
-        const int index= scantable[j];
-        int level= block[index];
-        if(level){
-            level+= 64;
-            if((level&(~127)) == 0){
-                if(j<block_last_index) rate+= s->intra_ac_vlc_length     [UNI_AC_ENC_INDEX(j-last-1, level)];
-                else                   rate+= s->intra_ac_vlc_last_length[UNI_AC_ENC_INDEX(j-last-1, level)];
-            }else
-                rate += s->ac_esc_length;
-            level-= 64;
-
-            last= j;
-        }
-    }
-
-    return rate;
-}
-
-static inline int decide_ac_pred(MpegEncContext * s, DCTELEM block[6][64], int dir[6], uint8_t *st[6], int zigzag_last_index[6])
-{
-    int score= 0;
-    int i, n;
-    int8_t * const qscale_table= s->current_picture.qscale_table;
-
-    memcpy(zigzag_last_index, s->block_last_index, sizeof(int)*6);
-
-    for(n=0; n<6; n++){
-        int16_t *ac_val, *ac_val1;
-
-        score -= get_block_rate(s, block[n], s->block_last_index[n], s->intra_scantable.permutated);
-
-        ac_val = s->ac_val[0][0] + s->block_index[n] * 16;
-        ac_val1= ac_val;
-        if(dir[n]){
-            const int xy= s->mb_x + s->mb_y*s->mb_stride - s->mb_stride;
-            /* top prediction */
-            ac_val-= s->block_wrap[n]*16;
-            if(s->mb_y==0 || s->qscale == qscale_table[xy] || n==2 || n==3){
-                /* same qscale */
-                for(i=1; i<8; i++){
-                    const int level= block[n][s->dsp.idct_permutation[i   ]];
-                    block[n][s->dsp.idct_permutation[i   ]] = level - ac_val[i+8];
-                    ac_val1[i  ]=    block[n][s->dsp.idct_permutation[i<<3]];
-                    ac_val1[i+8]= level;
-                }
-            }else{
-                /* different qscale, we must rescale */
-                for(i=1; i<8; i++){
-                    const int level= block[n][s->dsp.idct_permutation[i   ]];
-                    block[n][s->dsp.idct_permutation[i   ]] = level - ROUNDED_DIV(ac_val[i + 8]*qscale_table[xy], s->qscale);
-                    ac_val1[i  ]=    block[n][s->dsp.idct_permutation[i<<3]];
-                    ac_val1[i+8]= level;
-                }
-            }
-            st[n]= s->intra_h_scantable.permutated;
-        }else{
-            const int xy= s->mb_x-1 + s->mb_y*s->mb_stride;
-            /* left prediction */
-            ac_val-= 16;
-            if(s->mb_x==0 || s->qscale == qscale_table[xy] || n==1 || n==3){
-                /* same qscale */
-                for(i=1; i<8; i++){
-                    const int level= block[n][s->dsp.idct_permutation[i<<3]];
-                    block[n][s->dsp.idct_permutation[i<<3]]= level - ac_val[i];
-                    ac_val1[i  ]= level;
-                    ac_val1[i+8]=    block[n][s->dsp.idct_permutation[i   ]];
-                }
-            }else{
-                /* different qscale, we must rescale */
-                for(i=1; i<8; i++){
-                    const int level= block[n][s->dsp.idct_permutation[i<<3]];
-                    block[n][s->dsp.idct_permutation[i<<3]]= level - ROUNDED_DIV(ac_val[i]*qscale_table[xy], s->qscale);
-                    ac_val1[i  ]= level;
-                    ac_val1[i+8]=    block[n][s->dsp.idct_permutation[i   ]];
-                }
-            }
-            st[n]= s->intra_v_scantable.permutated;
-        }
-
-        for(i=63; i>0; i--) //FIXME optimize
-            if(block[n][ st[n][i] ]) break;
-        s->block_last_index[n]= i;
-
-        score += get_block_rate(s, block[n], s->block_last_index[n], st[n]);
-    }
-
-    return score < 0;
-}
-
-static inline void restore_ac_coeffs(MpegEncContext * s, DCTELEM block[6][64], int dir[6], uint8_t *st[6], int zigzag_last_index[6])
-{
-    int i, n;
-    memcpy(s->block_last_index, zigzag_last_index, sizeof(int)*6);
-
-    for(n=0; n<6; n++){
-        int16_t *ac_val = s->ac_val[0][0] + s->block_index[n] * 16;
-
-        st[n]= s->intra_scantable.permutated;
-        if(dir[n]){
-            /* top prediction */
-            for(i=1; i<8; i++){
-                block[n][s->dsp.idct_permutation[i   ]] = ac_val[i+8];
-            }
-        }else{
-            /* left prediction */
-            for(i=1; i<8; i++){
-                block[n][s->dsp.idct_permutation[i<<3]]= ac_val[i  ];
-            }
-        }
-    }
-}
-
-/**
- * modify qscale so that encoding is acually possible in h263 (limit difference to -2..2)
- */
-void ff_clean_h263_qscales(MpegEncContext *s){
-    int i;
-    int8_t * const qscale_table= s->current_picture.qscale_table;
-
-    for(i=1; i<s->mb_num; i++){
-        if(qscale_table[ s->mb_index2xy[i] ] - qscale_table[ s->mb_index2xy[i-1] ] >2)
-            qscale_table[ s->mb_index2xy[i] ]= qscale_table[ s->mb_index2xy[i-1] ]+2;
-    }
-    for(i=s->mb_num-2; i>=0; i--){
-        if(qscale_table[ s->mb_index2xy[i] ] - qscale_table[ s->mb_index2xy[i+1] ] >2)
-            qscale_table[ s->mb_index2xy[i] ]= qscale_table[ s->mb_index2xy[i+1] ]+2;
-    }
-
-    if(s->codec_id != CODEC_ID_H263P){
-        for(i=1; i<s->mb_num; i++){
-            int mb_xy= s->mb_index2xy[i];
-
-            if(qscale_table[mb_xy] != qscale_table[s->mb_index2xy[i-1]] && (s->mb_type[mb_xy]&CANDIDATE_MB_TYPE_INTER4V)){
-                s->mb_type[mb_xy]&= ~CANDIDATE_MB_TYPE_INTER4V;
-                s->mb_type[mb_xy]|= CANDIDATE_MB_TYPE_INTER;
-            }
-        }
-    }
-}
-
-/**
- * modify mb_type & qscale so that encoding is acually possible in mpeg4
- */
-void ff_clean_mpeg4_qscales(MpegEncContext *s){
-    int i;
-    int8_t * const qscale_table= s->current_picture.qscale_table;
-
-    ff_clean_h263_qscales(s);
-
-    if(s->pict_type== B_TYPE){
-        int odd=0;
-        /* ok, come on, this isn't funny anymore, there's more code for handling this mpeg4 mess than for the actual adaptive quantization */
-
-        for(i=0; i<s->mb_num; i++){
-            int mb_xy= s->mb_index2xy[i];
-            odd += qscale_table[mb_xy]&1;
-        }
-
-        if(2*odd > s->mb_num) odd=1;
-        else                  odd=0;
-
-        for(i=0; i<s->mb_num; i++){
-            int mb_xy= s->mb_index2xy[i];
-            if((qscale_table[mb_xy]&1) != odd)
-                qscale_table[mb_xy]++;
-            if(qscale_table[mb_xy] > 31)
-                qscale_table[mb_xy]= 31;
-        }
-
-        for(i=1; i<s->mb_num; i++){
-            int mb_xy= s->mb_index2xy[i];
-            if(qscale_table[mb_xy] != qscale_table[s->mb_index2xy[i-1]] && (s->mb_type[mb_xy]&CANDIDATE_MB_TYPE_DIRECT)){
-                s->mb_type[mb_xy]&= ~CANDIDATE_MB_TYPE_DIRECT;
-                s->mb_type[mb_xy]|= CANDIDATE_MB_TYPE_BIDIR;
-            }
-        }
-    }
-}
-
-#endif //CONFIG_ENCODERS
-
-#define tab_size ((signed)(sizeof(s->direct_scale_mv[0])/sizeof(int16_t)))
-#define tab_bias (tab_size/2)
-
-void ff_mpeg4_init_direct_mv(MpegEncContext *s){
-    int i;
-    for(i=0; i<tab_size; i++){
-        s->direct_scale_mv[0][i] = (i-tab_bias)*s->pb_time/s->pp_time;
-        s->direct_scale_mv[1][i] = (i-tab_bias)*(s->pb_time-s->pp_time)/s->pp_time;
-    }
-}
-
-static inline void ff_mpeg4_set_one_direct_mv(MpegEncContext *s, int mx, int my, int i){
-    int xy= s->block_index[i];
-    uint16_t time_pp= s->pp_time;
-    uint16_t time_pb= s->pb_time;
-    int p_mx, p_my;
-
-    p_mx= s->next_picture.motion_val[0][xy][0];
-    if((unsigned)(p_mx + tab_bias) < tab_size){
-        s->mv[0][i][0] = s->direct_scale_mv[0][p_mx + tab_bias] + mx;
-        s->mv[1][i][0] = mx ? s->mv[0][i][0] - p_mx
-                            : s->direct_scale_mv[1][p_mx + tab_bias];
-    }else{
-        s->mv[0][i][0] = p_mx*time_pb/time_pp + mx;
-        s->mv[1][i][0] = mx ? s->mv[0][i][0] - p_mx
-                            : p_mx*(time_pb - time_pp)/time_pp;
-    }
-    p_my= s->next_picture.motion_val[0][xy][1];
-    if((unsigned)(p_my + tab_bias) < tab_size){
-        s->mv[0][i][1] = s->direct_scale_mv[0][p_my + tab_bias] + my;
-        s->mv[1][i][1] = my ? s->mv[0][i][1] - p_my
-                            : s->direct_scale_mv[1][p_my + tab_bias];
-    }else{
-        s->mv[0][i][1] = p_my*time_pb/time_pp + my;
-        s->mv[1][i][1] = my ? s->mv[0][i][1] - p_my
-                            : p_my*(time_pb - time_pp)/time_pp;
-    }
-}
-
-#undef tab_size
-#undef tab_bias
-
-/**
- *
- * @return the mb_type
- */
-int ff_mpeg4_set_direct_mv(MpegEncContext *s, int mx, int my){
-    const int mb_index= s->mb_x + s->mb_y*s->mb_stride;
-    const int colocated_mb_type= s->next_picture.mb_type[mb_index];
-    uint16_t time_pp= s->pp_time;
-    uint16_t time_pb= s->pb_time;
-    int i;
-
-    //FIXME avoid divides
-    // try special case with shifts for 1 and 3 B-frames?
-
-    if(IS_8X8(colocated_mb_type)){
-        s->mv_type = MV_TYPE_8X8;
-        for(i=0; i<4; i++){
-            ff_mpeg4_set_one_direct_mv(s, mx, my, i);
-        }
-        return MB_TYPE_DIRECT2 | MB_TYPE_8x8 | MB_TYPE_L0L1;
-    } else if(IS_INTERLACED(colocated_mb_type)){
-        s->mv_type = MV_TYPE_FIELD;
-        for(i=0; i<2; i++){
-            int field_select= s->next_picture.ref_index[0][s->block_index[2*i]];
-            s->field_select[0][i]= field_select;
-            s->field_select[1][i]= i;
-            if(s->top_field_first){
-                time_pp= s->pp_field_time - field_select + i;
-                time_pb= s->pb_field_time - field_select + i;
-            }else{
-                time_pp= s->pp_field_time + field_select - i;
-                time_pb= s->pb_field_time + field_select - i;
-            }
-            s->mv[0][i][0] = s->p_field_mv_table[i][0][mb_index][0]*time_pb/time_pp + mx;
-            s->mv[0][i][1] = s->p_field_mv_table[i][0][mb_index][1]*time_pb/time_pp + my;
-            s->mv[1][i][0] = mx ? s->mv[0][i][0] - s->p_field_mv_table[i][0][mb_index][0]
-                                : s->p_field_mv_table[i][0][mb_index][0]*(time_pb - time_pp)/time_pp;
-            s->mv[1][i][1] = my ? s->mv[0][i][1] - s->p_field_mv_table[i][0][mb_index][1]
-                                : s->p_field_mv_table[i][0][mb_index][1]*(time_pb - time_pp)/time_pp;
-        }
-        return MB_TYPE_DIRECT2 | MB_TYPE_16x8 | MB_TYPE_L0L1 | MB_TYPE_INTERLACED;
-    }else{
-        ff_mpeg4_set_one_direct_mv(s, mx, my, 0);
-        s->mv[0][1][0] = s->mv[0][2][0] = s->mv[0][3][0] = s->mv[0][0][0];
-        s->mv[0][1][1] = s->mv[0][2][1] = s->mv[0][3][1] = s->mv[0][0][1];
-        s->mv[1][1][0] = s->mv[1][2][0] = s->mv[1][3][0] = s->mv[1][0][0];
-        s->mv[1][1][1] = s->mv[1][2][1] = s->mv[1][3][1] = s->mv[1][0][1];
-        if((s->avctx->workaround_bugs & FF_BUG_DIRECT_BLOCKSIZE) || !s->quarter_sample)
-            s->mv_type= MV_TYPE_16X16;
-        else
-            s->mv_type= MV_TYPE_8X8;
-        return MB_TYPE_DIRECT2 | MB_TYPE_16x16 | MB_TYPE_L0L1; //Note see prev line
-    }
-}
-
-void ff_h263_update_motion_val(MpegEncContext * s){
-    const int mb_xy = s->mb_y * s->mb_stride + s->mb_x;
-               //FIXME a lot of that is only needed for !low_delay
-    const int wrap = s->b8_stride;
-    const int xy = s->block_index[0];
-
-    s->current_picture.mbskip_table[mb_xy]= s->mb_skipped;
-
-    if(s->mv_type != MV_TYPE_8X8){
-        int motion_x, motion_y;
-        if (s->mb_intra) {
-            motion_x = 0;
-            motion_y = 0;
-        } else if (s->mv_type == MV_TYPE_16X16) {
-            motion_x = s->mv[0][0][0];
-            motion_y = s->mv[0][0][1];
-        } else /*if (s->mv_type == MV_TYPE_FIELD)*/ {
-            int i;
-            motion_x = s->mv[0][0][0] + s->mv[0][1][0];
-            motion_y = s->mv[0][0][1] + s->mv[0][1][1];
-            motion_x = (motion_x>>1) | (motion_x&1);
-            for(i=0; i<2; i++){
-                s->p_field_mv_table[i][0][mb_xy][0]= s->mv[0][i][0];
-                s->p_field_mv_table[i][0][mb_xy][1]= s->mv[0][i][1];
-            }
-            s->current_picture.ref_index[0][xy           ]=
-            s->current_picture.ref_index[0][xy        + 1]= s->field_select[0][0];
-            s->current_picture.ref_index[0][xy + wrap    ]=
-            s->current_picture.ref_index[0][xy + wrap + 1]= s->field_select[0][1];
-        }
-
-        /* no update if 8X8 because it has been done during parsing */
-        s->current_picture.motion_val[0][xy][0] = motion_x;
-        s->current_picture.motion_val[0][xy][1] = motion_y;
-        s->current_picture.motion_val[0][xy + 1][0] = motion_x;
-        s->current_picture.motion_val[0][xy + 1][1] = motion_y;
-        s->current_picture.motion_val[0][xy + wrap][0] = motion_x;
-        s->current_picture.motion_val[0][xy + wrap][1] = motion_y;
-        s->current_picture.motion_val[0][xy + 1 + wrap][0] = motion_x;
-        s->current_picture.motion_val[0][xy + 1 + wrap][1] = motion_y;
-    }
-
-    if(s->encoding){ //FIXME encoding MUST be cleaned up
-        if (s->mv_type == MV_TYPE_8X8)
-            s->current_picture.mb_type[mb_xy]= MB_TYPE_L0 | MB_TYPE_8x8;
-        else if(s->mb_intra)
-            s->current_picture.mb_type[mb_xy]= MB_TYPE_INTRA;
-        else
-            s->current_picture.mb_type[mb_xy]= MB_TYPE_L0 | MB_TYPE_16x16;
-    }
-}
-
-#ifdef CONFIG_ENCODERS
-
-static inline int h263_get_motion_length(MpegEncContext * s, int val, int f_code){
-    int l, bit_size, code;
-
-    if (val == 0) {
-        return mvtab[0][1];
-    } else {
-        bit_size = f_code - 1;
-        /* modulo encoding */
-        l= INT_BIT - 6 - bit_size;
-        val = (val<<l)>>l;
-        val--;
-        code = (val >> bit_size) + 1;
-
-        return mvtab[code][1] + 1 + bit_size;
-    }
-}
-
-static inline void ff_h263_encode_motion_vector(MpegEncContext * s, int x, int y, int f_code){
-    if(s->flags2 & CODEC_FLAG2_NO_OUTPUT){
-        skip_put_bits(&s->pb,
-            h263_get_motion_length(s, x, f_code)
-           +h263_get_motion_length(s, y, f_code));
-    }else{
-        ff_h263_encode_motion(s, x, f_code);
-        ff_h263_encode_motion(s, y, f_code);
-    }
-}
-
-static inline int get_p_cbp(MpegEncContext * s,
-                      DCTELEM block[6][64],
-                      int motion_x, int motion_y){
-    int cbp, i;
-
-    if(s->flags & CODEC_FLAG_CBP_RD){
-        int best_cbpy_score= INT_MAX;
-        int best_cbpc_score= INT_MAX;
-        int cbpc = (-1), cbpy= (-1);
-        const int offset= (s->mv_type==MV_TYPE_16X16 ? 0 : 16) + (s->dquant ? 8 : 0);
-        const int lambda= s->lambda2 >> (FF_LAMBDA_SHIFT - 6);
-
-        for(i=0; i<4; i++){
-            int score= inter_MCBPC_bits[i + offset] * lambda;
-            if(i&1) score += s->coded_score[5];
-            if(i&2) score += s->coded_score[4];
-
-            if(score < best_cbpc_score){
-                best_cbpc_score= score;
-                cbpc= i;
-            }
-        }
-
-        for(i=0; i<16; i++){
-            int score= cbpy_tab[i ^ 0xF][1] * lambda;
-            if(i&1) score += s->coded_score[3];
-            if(i&2) score += s->coded_score[2];
-            if(i&4) score += s->coded_score[1];
-            if(i&8) score += s->coded_score[0];
-
-            if(score < best_cbpy_score){
-                best_cbpy_score= score;
-                cbpy= i;
-            }
-        }
-        cbp= cbpc + 4*cbpy;
-        if ((motion_x | motion_y | s->dquant) == 0 && s->mv_type==MV_TYPE_16X16){
-            if(best_cbpy_score + best_cbpc_score + 2*lambda >= 0)
-                cbp= 0;
-        }
-
-        for (i = 0; i < 6; i++) {
-            if (s->block_last_index[i] >= 0 && ((cbp >> (5 - i))&1)==0 ){
-                s->block_last_index[i]= -1;
-                memset(s->block[i], 0, sizeof(DCTELEM)*64);
-            }
-        }
-    }else{
-        cbp= 0;
-        for (i = 0; i < 6; i++) {
-            if (s->block_last_index[i] >= 0)
-                cbp |= 1 << (5 - i);
-        }
-    }
-    return cbp;
-}
-
-static inline int get_b_cbp(MpegEncContext * s, DCTELEM block[6][64],
-                            int motion_x, int motion_y, int mb_type){
-    int cbp=0, i;
-
-    if(s->flags & CODEC_FLAG_CBP_RD){
-        int score=0;
-        const int lambda= s->lambda2 >> (FF_LAMBDA_SHIFT - 6);
-
-        for(i=0; i<6; i++){
-            if(s->coded_score[i] < 0){
-                score += s->coded_score[i];
-                cbp |= 1 << (5 - i);
-            }
-        }
-
-        if(cbp){
-            int zero_score= -6;
-            if ((motion_x | motion_y | s->dquant | mb_type) == 0){
-                zero_score-= 4; //2*MV + mb_type + cbp bit
-            }
-
-            zero_score*= lambda;
-            if(zero_score <= score){
-                cbp=0;
-            }
-        }
-
-        for (i = 0; i < 6; i++) {
-            if (s->block_last_index[i] >= 0 && ((cbp >> (5 - i))&1)==0 ){
-                s->block_last_index[i]= -1;
-                memset(s->block[i], 0, sizeof(DCTELEM)*64);
-            }
-        }
-    }else{
-        for (i = 0; i < 6; i++) {
-            if (s->block_last_index[i] >= 0)
-                cbp |= 1 << (5 - i);
-        }
-    }
-    return cbp;
-}
-
-static inline void mpeg4_encode_blocks(MpegEncContext * s, DCTELEM block[6][64], int intra_dc[6],
-                               uint8_t **scan_table, PutBitContext *dc_pb, PutBitContext *ac_pb){
-    int i;
-
-    if(scan_table){
-        if(s->flags2 & CODEC_FLAG2_NO_OUTPUT){
-            for (i = 0; i < 6; i++) {
-                skip_put_bits(&s->pb, mpeg4_get_block_length(s, block[i], i, intra_dc[i], scan_table[i]));
-            }
-        }else{
-            /* encode each block */
-            for (i = 0; i < 6; i++) {
-                mpeg4_encode_block(s, block[i], i, intra_dc[i], scan_table[i], dc_pb, ac_pb);
-            }
-        }
-    }else{
-        if(s->flags2 & CODEC_FLAG2_NO_OUTPUT){
-            for (i = 0; i < 6; i++) {
-                skip_put_bits(&s->pb, mpeg4_get_block_length(s, block[i], i, 0, s->intra_scantable.permutated));
-            }
-        }else{
-            /* encode each block */
-            for (i = 0; i < 6; i++) {
-                mpeg4_encode_block(s, block[i], i, 0, s->intra_scantable.permutated, dc_pb, ac_pb);
-            }
-        }
-    }
-}
-
-void mpeg4_encode_mb(MpegEncContext * s,
-                    DCTELEM block[6][64],
-                    int motion_x, int motion_y)
-{
-    int cbpc, cbpy, pred_x, pred_y;
-    PutBitContext * const pb2    = s->data_partitioning                         ? &s->pb2    : &s->pb;
-    PutBitContext * const tex_pb = s->data_partitioning && s->pict_type!=B_TYPE ? &s->tex_pb : &s->pb;
-    PutBitContext * const dc_pb  = s->data_partitioning && s->pict_type!=I_TYPE ? &s->pb2    : &s->pb;
-    const int interleaved_stats= (s->flags&CODEC_FLAG_PASS1) && !s->data_partitioning ? 1 : 0;
-    const int dquant_code[5]= {1,0,9,2,3};
-
-    //    printf("**mb x=%d y=%d\n", s->mb_x, s->mb_y);
-    if (!s->mb_intra) {
-        int i, cbp;
-
-        if(s->pict_type==B_TYPE){
-            static const int mb_type_table[8]= {-1, 2, 3, 1,-1,-1,-1, 0}; /* convert from mv_dir to type */
-            int mb_type=  mb_type_table[s->mv_dir];
-
-            if(s->mb_x==0){
-                for(i=0; i<2; i++){
-                    s->last_mv[i][0][0]=
-                    s->last_mv[i][0][1]=
-                    s->last_mv[i][1][0]=
-                    s->last_mv[i][1][1]= 0;
-                }
-            }
-
-            assert(s->dquant>=-2 && s->dquant<=2);
-            assert((s->dquant&1)==0);
-            assert(mb_type>=0);
-
-            /* nothing to do if this MB was skipped in the next P Frame */
-            if(s->next_picture.mbskip_table[s->mb_y * s->mb_stride + s->mb_x]){ //FIXME avoid DCT & ...
-                s->skip_count++;
-                s->mv[0][0][0]=
-                s->mv[0][0][1]=
-                s->mv[1][0][0]=
-                s->mv[1][0][1]= 0;
-                s->mv_dir= MV_DIR_FORWARD; //doesn't matter
-                s->qscale -= s->dquant;
-//                s->mb_skipped=1;
-
-                return;
-            }
-
-            cbp= get_b_cbp(s, block, motion_x, motion_y, mb_type);
-
-            if ((cbp | motion_x | motion_y | mb_type) ==0) {
-                /* direct MB with MV={0,0} */
-                assert(s->dquant==0);
-
-                put_bits(&s->pb, 1, 1); /* mb not coded modb1=1 */
-
-                if(interleaved_stats){
-                    s->misc_bits++;
-                    s->last_bits++;
-                }
-                s->skip_count++;
-                return;
-            }
-
-            put_bits(&s->pb, 1, 0);     /* mb coded modb1=0 */
-            put_bits(&s->pb, 1, cbp ? 0 : 1); /* modb2 */ //FIXME merge
-            put_bits(&s->pb, mb_type+1, 1); // this table is so simple that we don't need it :)
-            if(cbp) put_bits(&s->pb, 6, cbp);
-
-            if(cbp && mb_type){
-                if(s->dquant)
-                    put_bits(&s->pb, 2, (s->dquant>>2)+3);
-                else
-                    put_bits(&s->pb, 1, 0);
-            }else
-                s->qscale -= s->dquant;
-
-            if(!s->progressive_sequence){
-                if(cbp)
-                    put_bits(&s->pb, 1, s->interlaced_dct);
-                if(mb_type) // not direct mode
-                    put_bits(&s->pb, 1, s->mv_type == MV_TYPE_FIELD);
-            }
-
-            if(interleaved_stats){
-                s->misc_bits+= get_bits_diff(s);
-            }
-
-            if(mb_type == 0){
-                assert(s->mv_dir & MV_DIRECT);
-                ff_h263_encode_motion_vector(s, motion_x, motion_y, 1);
-                s->b_count++;
-                s->f_count++;
-            }else{
-                assert(mb_type > 0 && mb_type < 4);
-                if(s->mv_type != MV_TYPE_FIELD){
-                    if(s->mv_dir & MV_DIR_FORWARD){
-                        ff_h263_encode_motion_vector(s, s->mv[0][0][0] - s->last_mv[0][0][0],
-                                                        s->mv[0][0][1] - s->last_mv[0][0][1], s->f_code);
-                        s->last_mv[0][0][0]= s->last_mv[0][1][0]= s->mv[0][0][0];
-                        s->last_mv[0][0][1]= s->last_mv[0][1][1]= s->mv[0][0][1];
-                        s->f_count++;
-                    }
-                    if(s->mv_dir & MV_DIR_BACKWARD){
-                        ff_h263_encode_motion_vector(s, s->mv[1][0][0] - s->last_mv[1][0][0],
-                                                        s->mv[1][0][1] - s->last_mv[1][0][1], s->b_code);
-                        s->last_mv[1][0][0]= s->last_mv[1][1][0]= s->mv[1][0][0];
-                        s->last_mv[1][0][1]= s->last_mv[1][1][1]= s->mv[1][0][1];
-                        s->b_count++;
-                    }
-                }else{
-                    if(s->mv_dir & MV_DIR_FORWARD){
-                        put_bits(&s->pb, 1, s->field_select[0][0]);
-                        put_bits(&s->pb, 1, s->field_select[0][1]);
-                    }
-                    if(s->mv_dir & MV_DIR_BACKWARD){
-                        put_bits(&s->pb, 1, s->field_select[1][0]);
-                        put_bits(&s->pb, 1, s->field_select[1][1]);
-                    }
-                    if(s->mv_dir & MV_DIR_FORWARD){
-                        for(i=0; i<2; i++){
-                            ff_h263_encode_motion_vector(s, s->mv[0][i][0] - s->last_mv[0][i][0]  ,
-                                                            s->mv[0][i][1] - s->last_mv[0][i][1]/2, s->f_code);
-                            s->last_mv[0][i][0]= s->mv[0][i][0];
-                            s->last_mv[0][i][1]= s->mv[0][i][1]*2;
-                        }
-                        s->f_count++;
-                    }
-                    if(s->mv_dir & MV_DIR_BACKWARD){
-                        for(i=0; i<2; i++){
-                            ff_h263_encode_motion_vector(s, s->mv[1][i][0] - s->last_mv[1][i][0]  ,
-                                                            s->mv[1][i][1] - s->last_mv[1][i][1]/2, s->b_code);
-                            s->last_mv[1][i][0]= s->mv[1][i][0];
-                            s->last_mv[1][i][1]= s->mv[1][i][1]*2;
-                        }
-                        s->b_count++;
-                    }
-                }
-            }
-
-            if(interleaved_stats){
-                s->mv_bits+= get_bits_diff(s);
-            }
-
-            mpeg4_encode_blocks(s, block, NULL, NULL, NULL, &s->pb);
-
-            if(interleaved_stats){
-                s->p_tex_bits+= get_bits_diff(s);
-            }
-
-        }else{ /* s->pict_type==B_TYPE */
-            cbp= get_p_cbp(s, block, motion_x, motion_y);
-
-            if ((cbp | motion_x | motion_y | s->dquant) == 0 && s->mv_type==MV_TYPE_16X16) {
-                /* check if the B frames can skip it too, as we must skip it if we skip here
-                   why didn't they just compress the skip-mb bits instead of reusing them ?! */
-                if(s->max_b_frames>0){
-                    int i;
-                    int x,y, offset;
-                    uint8_t *p_pic;
-
-                    x= s->mb_x*16;
-                    y= s->mb_y*16;
-                    if(x+16 > s->width)  x= s->width-16;
-                    if(y+16 > s->height) y= s->height-16;
-
-                    offset= x + y*s->linesize;
-                    p_pic= s->new_picture.data[0] + offset;
-
-                    s->mb_skipped=1;
-                    for(i=0; i<s->max_b_frames; i++){
-                        uint8_t *b_pic;
-                        int diff;
-                        Picture *pic= s->reordered_input_picture[i+1];
-
-                        if(pic==NULL || pic->pict_type!=B_TYPE) break;
-
-                        b_pic= pic->data[0] + offset;
-                        if(pic->type != FF_BUFFER_TYPE_SHARED)
-                            b_pic+= INPLACE_OFFSET;
-                        diff= s->dsp.sad[0](NULL, p_pic, b_pic, s->linesize, 16);
-                        if(diff>s->qscale*70){ //FIXME check that 70 is optimal
-                            s->mb_skipped=0;
-                            break;
-                        }
-                    }
-                }else
-                    s->mb_skipped=1;
-
-                if(s->mb_skipped==1){
-                    /* skip macroblock */
-                    put_bits(&s->pb, 1, 1);
-
-                    if(interleaved_stats){
-                        s->misc_bits++;
-                        s->last_bits++;
-                    }
-                    s->skip_count++;
-
-                    return;
-                }
-            }
-
-            put_bits(&s->pb, 1, 0);     /* mb coded */
-            cbpc = cbp & 3;
-            cbpy = cbp >> 2;
-            cbpy ^= 0xf;
-            if(s->mv_type==MV_TYPE_16X16){
-                if(s->dquant) cbpc+= 8;
-                put_bits(&s->pb,
-                        inter_MCBPC_bits[cbpc],
-                        inter_MCBPC_code[cbpc]);
-
-                put_bits(pb2, cbpy_tab[cbpy][1], cbpy_tab[cbpy][0]);
-                if(s->dquant)
-                    put_bits(pb2, 2, dquant_code[s->dquant+2]);
-
-                if(!s->progressive_sequence){
-                    if(cbp)
-                        put_bits(pb2, 1, s->interlaced_dct);
-                    put_bits(pb2, 1, 0);
-                }
-
-                if(interleaved_stats){
-                    s->misc_bits+= get_bits_diff(s);
-                }
-
-                /* motion vectors: 16x16 mode */
-                h263_pred_motion(s, 0, 0, &pred_x, &pred_y);
-
-                ff_h263_encode_motion_vector(s, motion_x - pred_x,
-                                                motion_y - pred_y, s->f_code);
-            }else if(s->mv_type==MV_TYPE_FIELD){
-                if(s->dquant) cbpc+= 8;
-                put_bits(&s->pb,
-                        inter_MCBPC_bits[cbpc],
-                        inter_MCBPC_code[cbpc]);
-
-                put_bits(pb2, cbpy_tab[cbpy][1], cbpy_tab[cbpy][0]);
-                if(s->dquant)
-                    put_bits(pb2, 2, dquant_code[s->dquant+2]);
-
-                assert(!s->progressive_sequence);
-                if(cbp)
-                    put_bits(pb2, 1, s->interlaced_dct);
-                put_bits(pb2, 1, 1);
-
-                if(interleaved_stats){
-                    s->misc_bits+= get_bits_diff(s);
-                }
-
-                /* motion vectors: 16x8 interlaced mode */
-                h263_pred_motion(s, 0, 0, &pred_x, &pred_y);
-                pred_y /=2;
-
-                put_bits(&s->pb, 1, s->field_select[0][0]);
-                put_bits(&s->pb, 1, s->field_select[0][1]);
-
-                ff_h263_encode_motion_vector(s, s->mv[0][0][0] - pred_x,
-                                                s->mv[0][0][1] - pred_y, s->f_code);
-                ff_h263_encode_motion_vector(s, s->mv[0][1][0] - pred_x,
-                                                s->mv[0][1][1] - pred_y, s->f_code);
-            }else{
-                assert(s->mv_type==MV_TYPE_8X8);
-                put_bits(&s->pb,
-                        inter_MCBPC_bits[cbpc+16],
-                        inter_MCBPC_code[cbpc+16]);
-                put_bits(pb2, cbpy_tab[cbpy][1], cbpy_tab[cbpy][0]);
-
-                if(!s->progressive_sequence){
-                    if(cbp)
-                        put_bits(pb2, 1, s->interlaced_dct);
-                }
-
-                if(interleaved_stats){
-                    s->misc_bits+= get_bits_diff(s);
-                }
-
-                for(i=0; i<4; i++){
-                    /* motion vectors: 8x8 mode*/
-                    h263_pred_motion(s, i, 0, &pred_x, &pred_y);
-
-                    ff_h263_encode_motion_vector(s, s->current_picture.motion_val[0][ s->block_index[i] ][0] - pred_x,
-                                                    s->current_picture.motion_val[0][ s->block_index[i] ][1] - pred_y, s->f_code);
-                }
-            }
-
-            if(interleaved_stats){
-                s->mv_bits+= get_bits_diff(s);
-            }
-
-            mpeg4_encode_blocks(s, block, NULL, NULL, NULL, tex_pb);
-
-            if(interleaved_stats){
-                s->p_tex_bits+= get_bits_diff(s);
-            }
-            s->f_count++;
-        }
-    } else {
-        int cbp;
-        int dc_diff[6];   //dc values with the dc prediction subtracted
-        int dir[6];  //prediction direction
-        int zigzag_last_index[6];
-        uint8_t *scan_table[6];
-        int i;
-
-        for(i=0; i<6; i++){
-            dc_diff[i]= ff_mpeg4_pred_dc(s, i, block[i][0], &dir[i], 1);
-        }
-
-        if(s->flags & CODEC_FLAG_AC_PRED){
-            s->ac_pred= decide_ac_pred(s, block, dir, scan_table, zigzag_last_index);
-            if(!s->ac_pred)
-                restore_ac_coeffs(s, block, dir, scan_table, zigzag_last_index);
-        }else{
-            for(i=0; i<6; i++)
-                scan_table[i]= s->intra_scantable.permutated;
-        }
-
-        /* compute cbp */
-        cbp = 0;
-        for (i = 0; i < 6; i++) {
-            if (s->block_last_index[i] >= 1)
-                cbp |= 1 << (5 - i);
-        }
-
-        cbpc = cbp & 3;
-        if (s->pict_type == I_TYPE) {
-            if(s->dquant) cbpc+=4;
-            put_bits(&s->pb,
-                intra_MCBPC_bits[cbpc],
-                intra_MCBPC_code[cbpc]);
-        } else {
-            if(s->dquant) cbpc+=8;
-            put_bits(&s->pb, 1, 0);     /* mb coded */
-            put_bits(&s->pb,
-                inter_MCBPC_bits[cbpc + 4],
-                inter_MCBPC_code[cbpc + 4]);
-        }
-        put_bits(pb2, 1, s->ac_pred);
-        cbpy = cbp >> 2;
-        put_bits(pb2, cbpy_tab[cbpy][1], cbpy_tab[cbpy][0]);
-        if(s->dquant)
-            put_bits(dc_pb, 2, dquant_code[s->dquant+2]);
-
-        if(!s->progressive_sequence){
-            put_bits(dc_pb, 1, s->interlaced_dct);
-        }
-
-        if(interleaved_stats){
-            s->misc_bits+= get_bits_diff(s);
-        }
-
-        mpeg4_encode_blocks(s, block, dc_diff, scan_table, dc_pb, tex_pb);
-
-        if(interleaved_stats){
-            s->i_tex_bits+= get_bits_diff(s);
-        }
-        s->i_count++;
-
-        /* restore ac coeffs & last_index stuff if we messed them up with the prediction */
-        if(s->ac_pred)
-            restore_ac_coeffs(s, block, dir, scan_table, zigzag_last_index);
-    }
-}
-
-void h263_encode_mb(MpegEncContext * s,
-                    DCTELEM block[6][64],
-                    int motion_x, int motion_y)
-{
-    int cbpc, cbpy, i, cbp, pred_x, pred_y;
-    int16_t pred_dc;
-    int16_t rec_intradc[6];
-    int16_t *dc_ptr[6];
-    const int interleaved_stats= (s->flags&CODEC_FLAG_PASS1);
-    const int dquant_code[5]= {1,0,9,2,3};
-
-    //printf("**mb x=%d y=%d\n", s->mb_x, s->mb_y);
-    if (!s->mb_intra) {
-        /* compute cbp */
-        cbp= get_p_cbp(s, block, motion_x, motion_y);
-
-        if ((cbp | motion_x | motion_y | s->dquant | (s->mv_type - MV_TYPE_16X16)) == 0) {
-            /* skip macroblock */
-            put_bits(&s->pb, 1, 1);
-            if(interleaved_stats){
-                s->misc_bits++;
-                s->last_bits++;
-            }
-            s->skip_count++;
-
-            return;
-        }
-        put_bits(&s->pb, 1, 0);         /* mb coded */
-
-        cbpc = cbp & 3;
-        cbpy = cbp >> 2;
-        if(s->alt_inter_vlc==0 || cbpc!=3)
-            cbpy ^= 0xF;
-        if(s->dquant) cbpc+= 8;
-        if(s->mv_type==MV_TYPE_16X16){
-            put_bits(&s->pb,
-                    inter_MCBPC_bits[cbpc],
-                    inter_MCBPC_code[cbpc]);
-
-            put_bits(&s->pb, cbpy_tab[cbpy][1], cbpy_tab[cbpy][0]);
-            if(s->dquant)
-                put_bits(&s->pb, 2, dquant_code[s->dquant+2]);
-
-            if(interleaved_stats){
-                s->misc_bits+= get_bits_diff(s);
-            }
-
-            /* motion vectors: 16x16 mode */
-            h263_pred_motion(s, 0, 0, &pred_x, &pred_y);
-
-            if (!s->umvplus) {
-                ff_h263_encode_motion_vector(s, motion_x - pred_x,
-                                                motion_y - pred_y, 1);
-            }
-            else {
-                h263p_encode_umotion(s, motion_x - pred_x);
-                h263p_encode_umotion(s, motion_y - pred_y);
-                if (((motion_x - pred_x) == 1) && ((motion_y - pred_y) == 1))
-                    /* To prevent Start Code emulation */
-                    put_bits(&s->pb,1,1);
-            }
-        }else{
-            put_bits(&s->pb,
-                    inter_MCBPC_bits[cbpc+16],
-                    inter_MCBPC_code[cbpc+16]);
-            put_bits(&s->pb, cbpy_tab[cbpy][1], cbpy_tab[cbpy][0]);
-            if(s->dquant)
-                put_bits(&s->pb, 2, dquant_code[s->dquant+2]);
-
-            if(interleaved_stats){
-                s->misc_bits+= get_bits_diff(s);
-            }
-
-            for(i=0; i<4; i++){
-                /* motion vectors: 8x8 mode*/
-                h263_pred_motion(s, i, 0, &pred_x, &pred_y);
-
-                motion_x= s->current_picture.motion_val[0][ s->block_index[i] ][0];
-                motion_y= s->current_picture.motion_val[0][ s->block_index[i] ][1];
-                if (!s->umvplus) {
-                    ff_h263_encode_motion_vector(s, motion_x - pred_x,
-                                                    motion_y - pred_y, 1);
-                }
-                else {
-                    h263p_encode_umotion(s, motion_x - pred_x);
-                    h263p_encode_umotion(s, motion_y - pred_y);
-                    if (((motion_x - pred_x) == 1) && ((motion_y - pred_y) == 1))
-                        /* To prevent Start Code emulation */
-                        put_bits(&s->pb,1,1);
-                }
-            }
-        }
-
-        if(interleaved_stats){
-            s->mv_bits+= get_bits_diff(s);
-        }
-    } else {
-        assert(s->mb_intra);
-
-        cbp = 0;
-        if (s->h263_aic) {
-            /* Predict DC */
-            for(i=0; i<6; i++) {
-                int16_t level = block[i][0];
-                int scale;
-
-                if(i<4) scale= s->y_dc_scale;
-                else    scale= s->c_dc_scale;
-
-                pred_dc = h263_pred_dc(s, i, &dc_ptr[i]);
-                level -= pred_dc;
-                /* Quant */
-                if (level >= 0)
-                    level = (level + (scale>>1))/scale;
-                else
-                    level = (level - (scale>>1))/scale;
-
-                /* AIC can change CBP */
-                if (level == 0 && s->block_last_index[i] == 0)
-                    s->block_last_index[i] = -1;
-
-                if(!s->modified_quant){
-                    if (level < -127)
-                        level = -127;
-                    else if (level > 127)
-                        level = 127;
-                }
-
-                block[i][0] = level;
-                /* Reconstruction */
-                rec_intradc[i] = scale*level + pred_dc;
-                /* Oddify */
-                rec_intradc[i] |= 1;
-                //if ((rec_intradc[i] % 2) == 0)
-                //    rec_intradc[i]++;
-                /* Clipping */
-                if (rec_intradc[i] < 0)
-                    rec_intradc[i] = 0;
-                else if (rec_intradc[i] > 2047)
-                    rec_intradc[i] = 2047;
-
-                /* Update AC/DC tables */
-                *dc_ptr[i] = rec_intradc[i];
-                if (s->block_last_index[i] >= 0)
-                    cbp |= 1 << (5 - i);
-            }
-        }else{
-            for(i=0; i<6; i++) {
-                /* compute cbp */
-                if (s->block_last_index[i] >= 1)
-                    cbp |= 1 << (5 - i);
-            }
-        }
-
-        cbpc = cbp & 3;
-        if (s->pict_type == I_TYPE) {
-            if(s->dquant) cbpc+=4;
-            put_bits(&s->pb,
-                intra_MCBPC_bits[cbpc],
-                intra_MCBPC_code[cbpc]);
-        } else {
-            if(s->dquant) cbpc+=8;
-            put_bits(&s->pb, 1, 0);     /* mb coded */
-            put_bits(&s->pb,
-                inter_MCBPC_bits[cbpc + 4],
-                inter_MCBPC_code[cbpc + 4]);
-        }
-        if (s->h263_aic) {
-            /* XXX: currently, we do not try to use ac prediction */
-            put_bits(&s->pb, 1, 0);     /* no AC prediction */
-        }
-        cbpy = cbp >> 2;
-        put_bits(&s->pb, cbpy_tab[cbpy][1], cbpy_tab[cbpy][0]);
-        if(s->dquant)
-            put_bits(&s->pb, 2, dquant_code[s->dquant+2]);
-
-        if(interleaved_stats){
-            s->misc_bits+= get_bits_diff(s);
-        }
-    }
-
-    for(i=0; i<6; i++) {
-        /* encode each block */
-        h263_encode_block(s, block[i], i);
-
-        /* Update INTRADC for decoding */
-        if (s->h263_aic && s->mb_intra) {
-            block[i][0] = rec_intradc[i];
-
-        }
-    }
-
-    if(interleaved_stats){
-        if (!s->mb_intra) {
-            s->p_tex_bits+= get_bits_diff(s);
-            s->f_count++;
-        }else{
-            s->i_tex_bits+= get_bits_diff(s);
-            s->i_count++;
-        }
-    }
-}
-#endif
-
-void ff_h263_loop_filter(MpegEncContext * s){
-    int qp_c;
-    const int linesize  = s->linesize;
-    const int uvlinesize= s->uvlinesize;
-    const int xy = s->mb_y * s->mb_stride + s->mb_x;
-    uint8_t *dest_y = s->dest[0];
-    uint8_t *dest_cb= s->dest[1];
-    uint8_t *dest_cr= s->dest[2];
-
-//    if(s->pict_type==B_TYPE && !s->readable) return;
-
-    /*
-       Diag Top
-       Left Center
-    */
-    if(!IS_SKIP(s->current_picture.mb_type[xy])){
-        qp_c= s->qscale;
-        s->dsp.h263_v_loop_filter(dest_y+8*linesize  , linesize, qp_c);
-        s->dsp.h263_v_loop_filter(dest_y+8*linesize+8, linesize, qp_c);
-    }else
-        qp_c= 0;
-
-    if(s->mb_y){
-        int qp_dt, qp_t, qp_tc;
-
-        if(IS_SKIP(s->current_picture.mb_type[xy-s->mb_stride]))
-            qp_t=0;
-        else
-            qp_t= s->current_picture.qscale_table[xy-s->mb_stride];
-
-        if(qp_c)
-            qp_tc= qp_c;
-        else
-            qp_tc= qp_t;
-
-        if(qp_tc){
-            const int chroma_qp= s->chroma_qscale_table[qp_tc];
-            s->dsp.h263_v_loop_filter(dest_y  ,   linesize, qp_tc);
-            s->dsp.h263_v_loop_filter(dest_y+8,   linesize, qp_tc);
-
-            s->dsp.h263_v_loop_filter(dest_cb , uvlinesize, chroma_qp);
-            s->dsp.h263_v_loop_filter(dest_cr , uvlinesize, chroma_qp);
-        }
-
-        if(qp_t)
-            s->dsp.h263_h_loop_filter(dest_y-8*linesize+8  ,   linesize, qp_t);
-
-        if(s->mb_x){
-            if(qp_t || IS_SKIP(s->current_picture.mb_type[xy-1-s->mb_stride]))
-                qp_dt= qp_t;
-            else
-                qp_dt= s->current_picture.qscale_table[xy-1-s->mb_stride];
-
-            if(qp_dt){
-                const int chroma_qp= s->chroma_qscale_table[qp_dt];
-                s->dsp.h263_h_loop_filter(dest_y -8*linesize  ,   linesize, qp_dt);
-                s->dsp.h263_h_loop_filter(dest_cb-8*uvlinesize, uvlinesize, chroma_qp);
-                s->dsp.h263_h_loop_filter(dest_cr-8*uvlinesize, uvlinesize, chroma_qp);
-            }
-        }
-    }
-
-    if(qp_c){
-        s->dsp.h263_h_loop_filter(dest_y +8,   linesize, qp_c);
-        if(s->mb_y + 1 == s->mb_height)
-            s->dsp.h263_h_loop_filter(dest_y+8*linesize+8,   linesize, qp_c);
-    }
-
-    if(s->mb_x){
-        int qp_lc;
-        if(qp_c || IS_SKIP(s->current_picture.mb_type[xy-1]))
-            qp_lc= qp_c;
-        else
-            qp_lc= s->current_picture.qscale_table[xy-1];
-
-        if(qp_lc){
-            s->dsp.h263_h_loop_filter(dest_y,   linesize, qp_lc);
-            if(s->mb_y + 1 == s->mb_height){
-                const int chroma_qp= s->chroma_qscale_table[qp_lc];
-                s->dsp.h263_h_loop_filter(dest_y +8*  linesize,   linesize, qp_lc);
-                s->dsp.h263_h_loop_filter(dest_cb             , uvlinesize, chroma_qp);
-                s->dsp.h263_h_loop_filter(dest_cr             , uvlinesize, chroma_qp);
-            }
-        }
-    }
-}
-
-#ifdef CONFIG_ENCODERS
-static int h263_pred_dc(MpegEncContext * s, int n, int16_t **dc_val_ptr)
-{
-    int x, y, wrap, a, c, pred_dc, scale;
-    int16_t *dc_val;
-
-    /* find prediction */
-    if (n < 4) {
-        x = 2 * s->mb_x + (n & 1);
-        y = 2 * s->mb_y + ((n & 2) >> 1);
-        wrap = s->b8_stride;
-        dc_val = s->dc_val[0];
-        scale = s->y_dc_scale;
-    } else {
-        x = s->mb_x;
-        y = s->mb_y;
-        wrap = s->mb_stride;
-        dc_val = s->dc_val[n - 4 + 1];
-        scale = s->c_dc_scale;
-    }
-    /* B C
-     * A X
-     */
-    a = dc_val[(x - 1) + (y) * wrap];
-    c = dc_val[(x) + (y - 1) * wrap];
-
-    /* No prediction outside GOB boundary */
-    if(s->first_slice_line && n!=3){
-        if(n!=2) c= 1024;
-        if(n!=1 && s->mb_x == s->resync_mb_x) a= 1024;
-    }
-    pred_dc = 1024;
-    /* just DC prediction */
-    if (a != 1024 && c != 1024)
-        pred_dc = (a + c) >> 1;
-    else if (a != 1024)
-        pred_dc = a;
-    else
-        pred_dc = c;
-
-    /* we assume pred is positive */
-    //pred_dc = (pred_dc + (scale >> 1)) / scale;
-    *dc_val_ptr = &dc_val[x + y * wrap];
-    return pred_dc;
-}
-#endif /* CONFIG_ENCODERS */
-
-static void h263_pred_acdc(MpegEncContext * s, DCTELEM *block, int n)
-{
-    int x, y, wrap, a, c, pred_dc, scale, i;
-    int16_t *dc_val, *ac_val, *ac_val1;
-
-    /* find prediction */
-    if (n < 4) {
-        x = 2 * s->mb_x + (n & 1);
-        y = 2 * s->mb_y + (n>> 1);
-        wrap = s->b8_stride;
-        dc_val = s->dc_val[0];
-        ac_val = s->ac_val[0][0];
-        scale = s->y_dc_scale;
-    } else {
-        x = s->mb_x;
-        y = s->mb_y;
-        wrap = s->mb_stride;
-        dc_val = s->dc_val[n - 4 + 1];
-        ac_val = s->ac_val[n - 4 + 1][0];
-        scale = s->c_dc_scale;
-    }
-
-    ac_val += ((y) * wrap + (x)) * 16;
-    ac_val1 = ac_val;
-
-    /* B C
-     * A X
-     */
-    a = dc_val[(x - 1) + (y) * wrap];
-    c = dc_val[(x) + (y - 1) * wrap];
-
-    /* No prediction outside GOB boundary */
-    if(s->first_slice_line && n!=3){
-        if(n!=2) c= 1024;
-        if(n!=1 && s->mb_x == s->resync_mb_x) a= 1024;
-    }
-
-    if (s->ac_pred) {
-        pred_dc = 1024;
-        if (s->h263_aic_dir) {
-            /* left prediction */
-            if (a != 1024) {
-                ac_val -= 16;
-                for(i=1;i<8;i++) {
-                    block[s->dsp.idct_permutation[i<<3]] += ac_val[i];
-                }
-                pred_dc = a;
-            }
-        } else {
-            /* top prediction */
-            if (c != 1024) {
-                ac_val -= 16 * wrap;
-                for(i=1;i<8;i++) {
-                    block[s->dsp.idct_permutation[i   ]] += ac_val[i + 8];
-                }
-                pred_dc = c;
-            }
-        }
-    } else {
-        /* just DC prediction */
-        if (a != 1024 && c != 1024)
-            pred_dc = (a + c) >> 1;
-        else if (a != 1024)
-            pred_dc = a;
-        else
-            pred_dc = c;
-    }
-
-    /* we assume pred is positive */
-    block[0]=block[0]*scale + pred_dc;
-
-    if (block[0] < 0)
-        block[0] = 0;
-    else
-        block[0] |= 1;
-
-    /* Update AC/DC tables */
-    dc_val[(x) + (y) * wrap] = block[0];
-
-    /* left copy */
-    for(i=1;i<8;i++)
-        ac_val1[i    ] = block[s->dsp.idct_permutation[i<<3]];
-    /* top copy */
-    for(i=1;i<8;i++)
-        ac_val1[8 + i] = block[s->dsp.idct_permutation[i   ]];
-}
-
-int16_t *h263_pred_motion(MpegEncContext * s, int block, int dir,
-                        int *px, int *py)
-{
-    int wrap;
-    int16_t *A, *B, *C, (*mot_val)[2];
-    static const int off[4]= {2, 1, 1, -1};
-
-    wrap = s->b8_stride;
-    mot_val = s->current_picture.motion_val[dir] + s->block_index[block];
-
-    A = mot_val[ - 1];
-    /* special case for first (slice) line */
-    if (s->first_slice_line && block<3) {
-        // we can't just change some MVs to simulate that as we need them for the B frames (and ME)
-        // and if we ever support non rectangular objects than we need to do a few ifs here anyway :(
-        if(block==0){ //most common case
-            if(s->mb_x  == s->resync_mb_x){ //rare
-                *px= *py = 0;
-            }else if(s->mb_x + 1 == s->resync_mb_x && s->h263_pred){ //rare
-                C = mot_val[off[block] - wrap];
-                if(s->mb_x==0){
-                    *px = C[0];
-                    *py = C[1];
-                }else{
-                    *px = mid_pred(A[0], 0, C[0]);
-                    *py = mid_pred(A[1], 0, C[1]);
-                }
-            }else{
-                *px = A[0];
-                *py = A[1];
-            }
-        }else if(block==1){
-            if(s->mb_x + 1 == s->resync_mb_x && s->h263_pred){ //rare
-                C = mot_val[off[block] - wrap];
-                *px = mid_pred(A[0], 0, C[0]);
-                *py = mid_pred(A[1], 0, C[1]);
-            }else{
-                *px = A[0];
-                *py = A[1];
-            }
-        }else{ /* block==2*/
-            B = mot_val[ - wrap];
-            C = mot_val[off[block] - wrap];
-            if(s->mb_x == s->resync_mb_x) //rare
-                A[0]=A[1]=0;
-
-            *px = mid_pred(A[0], B[0], C[0]);
-            *py = mid_pred(A[1], B[1], C[1]);
-        }
-    } else {
-        B = mot_val[ - wrap];
-        C = mot_val[off[block] - wrap];
-        *px = mid_pred(A[0], B[0], C[0]);
-        *py = mid_pred(A[1], B[1], C[1]);
-    }
-    return *mot_val;
-}
-
-#ifdef CONFIG_ENCODERS
-void ff_h263_encode_motion(MpegEncContext * s, int val, int f_code)
-{
-    int range, l, bit_size, sign, code, bits;
-
-    if (val == 0) {
-        /* zero vector */
-        code = 0;
-        put_bits(&s->pb, mvtab[code][1], mvtab[code][0]);
-    } else {
-        bit_size = f_code - 1;
-        range = 1 << bit_size;
-        /* modulo encoding */
-        l= INT_BIT - 6 - bit_size;
-        val = (val<<l)>>l;
-        sign = val>>31;
-        val= (val^sign)-sign;
-        sign&=1;
-
-        val--;
-        code = (val >> bit_size) + 1;
-        bits = val & (range - 1);
-
-        put_bits(&s->pb, mvtab[code][1] + 1, (mvtab[code][0] << 1) | sign);
-        if (bit_size > 0) {
-            put_bits(&s->pb, bit_size, bits);
-        }
-    }
-}
-
-/* Encode MV differences on H.263+ with Unrestricted MV mode */
-static void h263p_encode_umotion(MpegEncContext * s, int val)
-{
-    short sval = 0;
-    short i = 0;
-    short n_bits = 0;
-    short temp_val;
-    int code = 0;
-    int tcode;
-
-    if ( val == 0)
-        put_bits(&s->pb, 1, 1);
-    else if (val == 1)
-        put_bits(&s->pb, 3, 0);
-    else if (val == -1)
-        put_bits(&s->pb, 3, 2);
-    else {
-
-        sval = ((val < 0) ? (short)(-val):(short)val);
-        temp_val = sval;
-
-        while (temp_val != 0) {
-            temp_val = temp_val >> 1;
-            n_bits++;
-        }
-
-        i = n_bits - 1;
-        while (i > 0) {
-            tcode = (sval & (1 << (i-1))) >> (i-1);
-            tcode = (tcode << 1) | 1;
-            code = (code << 2) | tcode;
-            i--;
-        }
-        code = ((code << 1) | (val < 0)) << 1;
-        put_bits(&s->pb, (2*n_bits)+1, code);
-        //printf("\nVal = %d\tCode = %d", sval, code);
-    }
-}
-
-static void init_mv_penalty_and_fcode(MpegEncContext *s)
-{
-    int f_code;
-    int mv;
-
-    if(mv_penalty==NULL)
-        mv_penalty= av_mallocz( sizeof(uint8_t)*(MAX_FCODE+1)*(2*MAX_MV+1) );
-
-    for(f_code=1; f_code<=MAX_FCODE; f_code++){
-        for(mv=-MAX_MV; mv<=MAX_MV; mv++){
-            int len;
-
-            if(mv==0) len= mvtab[0][1];
-            else{
-                int val, bit_size, range, code;
-
-                bit_size = f_code - 1;
-                range = 1 << bit_size;
-
-                val=mv;
-                if (val < 0)
-                    val = -val;
-                val--;
-                code = (val >> bit_size) + 1;
-                if(code<33){
-                    len= mvtab[code][1] + 1 + bit_size;
-                }else{
-                    len= mvtab[32][1] + av_log2(code>>5) + 2 + bit_size;
-                }
-            }
-
-            mv_penalty[f_code][mv+MAX_MV]= len;
-        }
-    }
-
-    for(f_code=MAX_FCODE; f_code>0; f_code--){
-        for(mv=-(16<<f_code); mv<(16<<f_code); mv++){
-            fcode_tab[mv+MAX_MV]= f_code;
-        }
-    }
-
-    for(mv=0; mv<MAX_MV*2+1; mv++){
-        umv_fcode_tab[mv]= 1;
-    }
-}
-#endif
-
-#ifdef CONFIG_ENCODERS
-
-static void init_uni_dc_tab(void)
-{
-    int level, uni_code, uni_len;
-
-    for(level=-256; level<256; level++){
-        int size, v, l;
-        /* find number of bits */
-        size = 0;
-        v = abs(level);
-        while (v) {
-            v >>= 1;
-            size++;
-        }
-
-        if (level < 0)
-            l= (-level) ^ ((1 << size) - 1);
-        else
-            l= level;
-
-        /* luminance */
-        uni_code= DCtab_lum[size][0];
-        uni_len = DCtab_lum[size][1];
-
-        if (size > 0) {
-            uni_code<<=size; uni_code|=l;
-            uni_len+=size;
-            if (size > 8){
-                uni_code<<=1; uni_code|=1;
-                uni_len++;
-            }
-        }
-        uni_DCtab_lum_bits[level+256]= uni_code;
-        uni_DCtab_lum_len [level+256]= uni_len;
-
-        /* chrominance */
-        uni_code= DCtab_chrom[size][0];
-        uni_len = DCtab_chrom[size][1];
-
-        if (size > 0) {
-            uni_code<<=size; uni_code|=l;
-            uni_len+=size;
-            if (size > 8){
-                uni_code<<=1; uni_code|=1;
-                uni_len++;
-            }
-        }
-        uni_DCtab_chrom_bits[level+256]= uni_code;
-        uni_DCtab_chrom_len [level+256]= uni_len;
-
-    }
-}
-
-#endif //CONFIG_ENCODERS
-
-#ifdef CONFIG_ENCODERS
-static void init_uni_mpeg4_rl_tab(RLTable *rl, uint32_t *bits_tab, uint8_t *len_tab){
-    int slevel, run, last;
-
-    assert(MAX_LEVEL >= 64);
-    assert(MAX_RUN   >= 63);
-
-    for(slevel=-64; slevel<64; slevel++){
-        if(slevel==0) continue;
-        for(run=0; run<64; run++){
-            for(last=0; last<=1; last++){
-                const int index= UNI_MPEG4_ENC_INDEX(last, run, slevel+64);
-                int level= slevel < 0 ? -slevel : slevel;
-                int sign= slevel < 0 ? 1 : 0;
-                int bits, len, code;
-                int level1, run1;
-
-                len_tab[index]= 100;
-
-                /* ESC0 */
-                code= get_rl_index(rl, last, run, level);
-                bits= rl->table_vlc[code][0];
-                len=  rl->table_vlc[code][1];
-                bits=bits*2+sign; len++;
-
-                if(code!=rl->n && len < len_tab[index]){
-                    bits_tab[index]= bits;
-                    len_tab [index]= len;
-                }
-#if 1
-                /* ESC1 */
-                bits= rl->table_vlc[rl->n][0];
-                len=  rl->table_vlc[rl->n][1];
-                bits=bits*2;    len++; //esc1
-                level1= level - rl->max_level[last][run];
-                if(level1>0){
-                    code= get_rl_index(rl, last, run, level1);
-                    bits<<= rl->table_vlc[code][1];
-                    len  += rl->table_vlc[code][1];
-                    bits += rl->table_vlc[code][0];
-                    bits=bits*2+sign; len++;
-
-                    if(code!=rl->n && len < len_tab[index]){
-                        bits_tab[index]= bits;
-                        len_tab [index]= len;
-                    }
-                }
-#endif
-#if 1
-                /* ESC2 */
-                bits= rl->table_vlc[rl->n][0];
-                len=  rl->table_vlc[rl->n][1];
-                bits=bits*4+2;    len+=2; //esc2
-                run1 = run - rl->max_run[last][level] - 1;
-                if(run1>=0){
-                    code= get_rl_index(rl, last, run1, level);
-                    bits<<= rl->table_vlc[code][1];
-                    len  += rl->table_vlc[code][1];
-                    bits += rl->table_vlc[code][0];
-                    bits=bits*2+sign; len++;
-
-                    if(code!=rl->n && len < len_tab[index]){
-                        bits_tab[index]= bits;
-                        len_tab [index]= len;
-                    }
-                }
-#endif
-                /* ESC3 */
-                bits= rl->table_vlc[rl->n][0];
-                len = rl->table_vlc[rl->n][1];
-                bits=bits*4+3;    len+=2; //esc3
-                bits=bits*2+last; len++;
-                bits=bits*64+run; len+=6;
-                bits=bits*2+1;    len++;  //marker
-                bits=bits*4096+(slevel&0xfff); len+=12;
-                bits=bits*2+1;    len++;  //marker
-
-                if(len < len_tab[index]){
-                    bits_tab[index]= bits;
-                    len_tab [index]= len;
-                }
-            }
-        }
-    }
-}
-
-static void init_uni_h263_rl_tab(RLTable *rl, uint32_t *bits_tab, uint8_t *len_tab){
-    int slevel, run, last;
-
-    assert(MAX_LEVEL >= 64);
-    assert(MAX_RUN   >= 63);
-
-    for(slevel=-64; slevel<64; slevel++){
-        if(slevel==0) continue;
-        for(run=0; run<64; run++){
-            for(last=0; last<=1; last++){
-                const int index= UNI_MPEG4_ENC_INDEX(last, run, slevel+64);
-                int level= slevel < 0 ? -slevel : slevel;
-                int sign= slevel < 0 ? 1 : 0;
-                int bits, len, code;
-
-                len_tab[index]= 100;
-
-                /* ESC0 */
-                code= get_rl_index(rl, last, run, level);
-                bits= rl->table_vlc[code][0];
-                len=  rl->table_vlc[code][1];
-                bits=bits*2+sign; len++;
-
-                if(code!=rl->n && len < len_tab[index]){
-                    if(bits_tab) bits_tab[index]= bits;
-                    len_tab [index]= len;
-                }
-                /* ESC */
-                bits= rl->table_vlc[rl->n][0];
-                len = rl->table_vlc[rl->n][1];
-                bits=bits*2+last; len++;
-                bits=bits*64+run; len+=6;
-                bits=bits*256+(level&0xff); len+=8;
-
-                if(len < len_tab[index]){
-                    if(bits_tab) bits_tab[index]= bits;
-                    len_tab [index]= len;
-                }
-            }
-        }
-    }
-}
-
-void h263_encode_init(MpegEncContext *s)
-{
-    static int done = 0;
-
-    if (!done) {
-        done = 1;
-
-        init_uni_dc_tab();
-
-        init_rl(&rl_inter, 1);
-        init_rl(&rl_intra, 1);
-        init_rl(&rl_intra_aic, 1);
-
-        init_uni_mpeg4_rl_tab(&rl_intra, uni_mpeg4_intra_rl_bits, uni_mpeg4_intra_rl_len);
-        init_uni_mpeg4_rl_tab(&rl_inter, uni_mpeg4_inter_rl_bits, uni_mpeg4_inter_rl_len);
-
-        init_uni_h263_rl_tab(&rl_intra_aic, NULL, uni_h263_intra_aic_rl_len);
-        init_uni_h263_rl_tab(&rl_inter    , NULL, uni_h263_inter_rl_len);
-
-        init_mv_penalty_and_fcode(s);
-    }
-    s->me.mv_penalty= mv_penalty; //FIXME exact table for msmpeg4 & h263p
-
-    s->intra_ac_vlc_length     =s->inter_ac_vlc_length     = uni_h263_inter_rl_len;
-    s->intra_ac_vlc_last_length=s->inter_ac_vlc_last_length= uni_h263_inter_rl_len + 128*64;
-    if(s->h263_aic){
-        s->intra_ac_vlc_length     = uni_h263_intra_aic_rl_len;
-        s->intra_ac_vlc_last_length= uni_h263_intra_aic_rl_len + 128*64;
-    }
-    s->ac_esc_length= 7+1+6+8;
-
-    // use fcodes >1 only for mpeg4 & h263 & h263p FIXME
-    switch(s->codec_id){
-    case CODEC_ID_MPEG4:
-        s->fcode_tab= fcode_tab;
-        s->min_qcoeff= -2048;
-        s->max_qcoeff=  2047;
-        s->intra_ac_vlc_length     = uni_mpeg4_intra_rl_len;
-        s->intra_ac_vlc_last_length= uni_mpeg4_intra_rl_len + 128*64;
-        s->inter_ac_vlc_length     = uni_mpeg4_inter_rl_len;
-        s->inter_ac_vlc_last_length= uni_mpeg4_inter_rl_len + 128*64;
-        s->luma_dc_vlc_length= uni_DCtab_lum_len;
-        s->chroma_dc_vlc_length= uni_DCtab_chrom_len;
-        s->ac_esc_length= 7+2+1+6+1+12+1;
-        s->y_dc_scale_table= ff_mpeg4_y_dc_scale_table;
-        s->c_dc_scale_table= ff_mpeg4_c_dc_scale_table;
-
-        if(s->flags & CODEC_FLAG_GLOBAL_HEADER){
-
-            s->avctx->extradata= av_malloc(1024);
-            init_put_bits(&s->pb, s->avctx->extradata, 1024);
-
-            if(!(s->workaround_bugs & FF_BUG_MS))
-                mpeg4_encode_visual_object_header(s);
-            mpeg4_encode_vol_header(s, 0, 0);
-
-//            ff_mpeg4_stuffing(&s->pb); ?
-            flush_put_bits(&s->pb);
-            s->avctx->extradata_size= (put_bits_count(&s->pb)+7)>>3;
-        }
-
-        break;
-    case CODEC_ID_H263P:
-        if(s->umvplus)
-            s->fcode_tab= umv_fcode_tab;
-        if(s->modified_quant){
-            s->min_qcoeff= -2047;
-            s->max_qcoeff=  2047;
-        }else{
-            s->min_qcoeff= -127;
-            s->max_qcoeff=  127;
-        }
-        break;
-        //Note for mpeg4 & h263 the dc-scale table will be set per frame as needed later
-    case CODEC_ID_FLV1:
-        if (s->h263_flv > 1) {
-            s->min_qcoeff= -1023;
-            s->max_qcoeff=  1023;
-        } else {
-            s->min_qcoeff= -127;
-            s->max_qcoeff=  127;
-        }
-        s->y_dc_scale_table=
-        s->c_dc_scale_table= ff_mpeg1_dc_scale_table;
-        break;
-    default: //nothing needed - default table already set in mpegvideo.c
-        s->min_qcoeff= -127;
-        s->max_qcoeff=  127;
-        s->y_dc_scale_table=
-        s->c_dc_scale_table= ff_mpeg1_dc_scale_table;
-    }
-}
-
-/**
- * encodes a 8x8 block.
- * @param block the 8x8 block
- * @param n block index (0-3 are luma, 4-5 are chroma)
- */
-static void h263_encode_block(MpegEncContext * s, DCTELEM * block, int n)
-{
-    int level, run, last, i, j, last_index, last_non_zero, sign, slevel, code;
-    RLTable *rl;
-
-    rl = &rl_inter;
-    if (s->mb_intra && !s->h263_aic) {
-        /* DC coef */
-        level = block[0];
-        /* 255 cannot be represented, so we clamp */
-        if (level > 254) {
-            level = 254;
-            block[0] = 254;
-        }
-        /* 0 cannot be represented also */
-        else if (level < 1) {
-            level = 1;
-            block[0] = 1;
-        }
-        if (level == 128) //FIXME check rv10
-            put_bits(&s->pb, 8, 0xff);
-        else
-            put_bits(&s->pb, 8, level);
-        i = 1;
-    } else {
-        i = 0;
-        if (s->h263_aic && s->mb_intra)
-            rl = &rl_intra_aic;
-
-        if(s->alt_inter_vlc && !s->mb_intra){
-            int aic_vlc_bits=0;
-            int inter_vlc_bits=0;
-            int wrong_pos=-1;
-            int aic_code;
-
-            last_index = s->block_last_index[n];
-            last_non_zero = i - 1;
-            for (; i <= last_index; i++) {
-                j = s->intra_scantable.permutated[i];
-                level = block[j];
-                if (level) {
-                    run = i - last_non_zero - 1;
-                    last = (i == last_index);
-
-                    if(level<0) level= -level;
-
-                    code = get_rl_index(rl, last, run, level);
-                    aic_code = get_rl_index(&rl_intra_aic, last, run, level);
-                    inter_vlc_bits += rl->table_vlc[code][1]+1;
-                    aic_vlc_bits   += rl_intra_aic.table_vlc[aic_code][1]+1;
-
-                    if (code == rl->n) {
-                        inter_vlc_bits += 1+6+8-1;
-                    }
-                    if (aic_code == rl_intra_aic.n) {
-                        aic_vlc_bits += 1+6+8-1;
-                        wrong_pos += run + 1;
-                    }else
-                        wrong_pos += wrong_run[aic_code];
-                    last_non_zero = i;
-                }
-            }
-            i = 0;
-            if(aic_vlc_bits < inter_vlc_bits && wrong_pos > 63)
-                rl = &rl_intra_aic;
-        }
-    }
-
-    /* AC coefs */
-    last_index = s->block_last_index[n];
-    last_non_zero = i - 1;
-    for (; i <= last_index; i++) {
-        j = s->intra_scantable.permutated[i];
-        level = block[j];
-        if (level) {
-            run = i - last_non_zero - 1;
-            last = (i == last_index);
-            sign = 0;
-            slevel = level;
-            if (level < 0) {
-                sign = 1;
-                level = -level;
-            }
-            code = get_rl_index(rl, last, run, level);
-            put_bits(&s->pb, rl->table_vlc[code][1], rl->table_vlc[code][0]);
-            if (code == rl->n) {
-              if(s->h263_flv <= 1){
-                put_bits(&s->pb, 1, last);
-                put_bits(&s->pb, 6, run);
-
-                assert(slevel != 0);
-
-                if(level < 128)
-                    put_bits(&s->pb, 8, slevel & 0xff);
-                else{
-                    put_bits(&s->pb, 8, 128);
-                    put_bits(&s->pb, 5, slevel & 0x1f);
-                    put_bits(&s->pb, 6, (slevel>>5)&0x3f);
-                }
-              }else{
-                if(level < 64) { // 7-bit level
-                        put_bits(&s->pb, 1, 0);
-                        put_bits(&s->pb, 1, last);
-                        put_bits(&s->pb, 6, run);
-
-                        put_bits(&s->pb, 7, slevel & 0x7f);
-                    } else {
-                        /* 11-bit level */
-                        put_bits(&s->pb, 1, 1);
-                        put_bits(&s->pb, 1, last);
-                        put_bits(&s->pb, 6, run);
-
-                        put_bits(&s->pb, 11, slevel & 0x7ff);
-                    }
-              }
-            } else {
-                put_bits(&s->pb, 1, sign);
-            }
-            last_non_zero = i;
-        }
-    }
-}
-#endif
-
-#ifdef CONFIG_ENCODERS
-
-/***************************************************/
-/**
- * add mpeg4 stuffing bits (01...1)
- */
-void ff_mpeg4_stuffing(PutBitContext * pbc)
-{
-    int length;
-    put_bits(pbc, 1, 0);
-    length= (-put_bits_count(pbc))&7;
-    if(length) put_bits(pbc, length, (1<<length)-1);
-}
-
-/* must be called before writing the header */
-void ff_set_mpeg4_time(MpegEncContext * s, int picture_number){
-    int time_div, time_mod;
-
-    assert(s->current_picture_ptr->pts != AV_NOPTS_VALUE);
-    s->time= s->current_picture_ptr->pts*s->avctx->time_base.num;
-
-    time_div= s->time/s->avctx->time_base.den;
-    time_mod= s->time%s->avctx->time_base.den;
-
-    if(s->pict_type==B_TYPE){
-        s->pb_time= s->pp_time - (s->last_non_b_time - s->time);
-        assert(s->pb_time > 0 && s->pb_time < s->pp_time);
-        ff_mpeg4_init_direct_mv(s);
-    }else{
-        s->last_time_base= s->time_base;
-        s->time_base= time_div;
-        s->pp_time= s->time - s->last_non_b_time;
-        s->last_non_b_time= s->time;
-        assert(picture_number==0 || s->pp_time > 0);
-    }
-}
-
-static void mpeg4_encode_gop_header(MpegEncContext * s){
-    int hours, minutes, seconds;
-    int64_t time;
-
-    put_bits(&s->pb, 16, 0);
-    put_bits(&s->pb, 16, GOP_STARTCODE);
-
-    time= s->current_picture_ptr->pts;
-    if(s->reordered_input_picture[1])
-        time= FFMIN(time, s->reordered_input_picture[1]->pts);
-    time= time*s->avctx->time_base.num;
-
-    seconds= time/s->avctx->time_base.den;
-    minutes= seconds/60; seconds %= 60;
-    hours= minutes/60; minutes %= 60;
-    hours%=24;
-
-    put_bits(&s->pb, 5, hours);
-    put_bits(&s->pb, 6, minutes);
-    put_bits(&s->pb, 1, 1);
-    put_bits(&s->pb, 6, seconds);
-
-    put_bits(&s->pb, 1, !!(s->flags&CODEC_FLAG_CLOSED_GOP));
-    put_bits(&s->pb, 1, 0); //broken link == NO
-
-    s->last_time_base= time / s->avctx->time_base.den;
-
-    ff_mpeg4_stuffing(&s->pb);
-}
-
-static void mpeg4_encode_visual_object_header(MpegEncContext * s){
-    int profile_and_level_indication;
-    int vo_ver_id;
-
-    if(s->avctx->profile != FF_PROFILE_UNKNOWN){
-        profile_and_level_indication = s->avctx->profile << 4;
-    }else if(s->max_b_frames || s->quarter_sample){
-        profile_and_level_indication= 0xF0; // adv simple
-    }else{
-        profile_and_level_indication= 0x00; // simple
-    }
-
-    if(s->avctx->level != FF_LEVEL_UNKNOWN){
-        profile_and_level_indication |= s->avctx->level;
-    }else{
-        profile_and_level_indication |= 1; //level 1
-    }
-
-    if(profile_and_level_indication>>4 == 0xF){
-        vo_ver_id= 5;
-    }else{
-        vo_ver_id= 1;
-    }
-
-    //FIXME levels
-
-    put_bits(&s->pb, 16, 0);
-    put_bits(&s->pb, 16, VOS_STARTCODE);
-
-    put_bits(&s->pb, 8, profile_and_level_indication);
-
-    put_bits(&s->pb, 16, 0);
-    put_bits(&s->pb, 16, VISUAL_OBJ_STARTCODE);
-
-    put_bits(&s->pb, 1, 1);
-        put_bits(&s->pb, 4, vo_ver_id);
-        put_bits(&s->pb, 3, 1); //priority
-
-    put_bits(&s->pb, 4, 1); //visual obj type== video obj
-
-    put_bits(&s->pb, 1, 0); //video signal type == no clue //FIXME
-
-    ff_mpeg4_stuffing(&s->pb);
-}
-
-static void mpeg4_encode_vol_header(MpegEncContext * s, int vo_number, int vol_number)
-{
-    int vo_ver_id;
-
-    if(s->max_b_frames || s->quarter_sample){
-        vo_ver_id= 5;
-        s->vo_type= ADV_SIMPLE_VO_TYPE;
-    }else{
-        vo_ver_id= 1;
-        s->vo_type= SIMPLE_VO_TYPE;
-    }
-
-    put_bits(&s->pb, 16, 0);
-    put_bits(&s->pb, 16, 0x100 + vo_number);        /* video obj */
-    put_bits(&s->pb, 16, 0);
-    put_bits(&s->pb, 16, 0x120 + vol_number);       /* video obj layer */
-
-    put_bits(&s->pb, 1, 0);             /* random access vol */
-    put_bits(&s->pb, 8, s->vo_type);    /* video obj type indication */
-    if(s->workaround_bugs & FF_BUG_MS) {
-        put_bits(&s->pb, 1, 0);         /* is obj layer id= no */
-    } else {
-        put_bits(&s->pb, 1, 1);         /* is obj layer id= yes */
-        put_bits(&s->pb, 4, vo_ver_id); /* is obj layer ver id */
-        put_bits(&s->pb, 3, 1);         /* is obj layer priority */
-    }
-
-    aspect_to_info(s, s->avctx->sample_aspect_ratio);
-
-    put_bits(&s->pb, 4, s->aspect_ratio_info);/* aspect ratio info */
-    if (s->aspect_ratio_info == FF_ASPECT_EXTENDED){
-        put_bits(&s->pb, 8, s->avctx->sample_aspect_ratio.num);
-        put_bits(&s->pb, 8, s->avctx->sample_aspect_ratio.den);
-    }
-
-    if(s->workaround_bugs & FF_BUG_MS) { //
-        put_bits(&s->pb, 1, 0);         /* vol control parameters= no @@@ */
-    } else {
-        put_bits(&s->pb, 1, 1);         /* vol control parameters= yes */
-        put_bits(&s->pb, 2, 1);         /* chroma format YUV 420/YV12 */
-        put_bits(&s->pb, 1, s->low_delay);
-        put_bits(&s->pb, 1, 0);         /* vbv parameters= no */
-    }
-
-    put_bits(&s->pb, 2, RECT_SHAPE);    /* vol shape= rectangle */
-    put_bits(&s->pb, 1, 1);             /* marker bit */
-
-    put_bits(&s->pb, 16, s->avctx->time_base.den);
-    if (s->time_increment_bits < 1)
-        s->time_increment_bits = 1;
-    put_bits(&s->pb, 1, 1);             /* marker bit */
-    put_bits(&s->pb, 1, 0);             /* fixed vop rate=no */
-    put_bits(&s->pb, 1, 1);             /* marker bit */
-    put_bits(&s->pb, 13, s->width);     /* vol width */
-    put_bits(&s->pb, 1, 1);             /* marker bit */
-    put_bits(&s->pb, 13, s->height);    /* vol height */
-    put_bits(&s->pb, 1, 1);             /* marker bit */
-    put_bits(&s->pb, 1, s->progressive_sequence ? 0 : 1);
-    put_bits(&s->pb, 1, 1);             /* obmc disable */
-    if (vo_ver_id == 1) {
-        put_bits(&s->pb, 1, s->vol_sprite_usage);       /* sprite enable */
-    }else{
-        put_bits(&s->pb, 2, s->vol_sprite_usage);       /* sprite enable */
-    }
-
-    put_bits(&s->pb, 1, 0);             /* not 8 bit == false */
-    put_bits(&s->pb, 1, s->mpeg_quant); /* quant type= (0=h263 style)*/
-
-    if(s->mpeg_quant){
-        ff_write_quant_matrix(&s->pb, s->avctx->intra_matrix);
-        ff_write_quant_matrix(&s->pb, s->avctx->inter_matrix);
-    }
-
-    if (vo_ver_id != 1)
-        put_bits(&s->pb, 1, s->quarter_sample);
-    put_bits(&s->pb, 1, 1);             /* complexity estimation disable */
-    s->resync_marker= s->rtp_mode;
-    put_bits(&s->pb, 1, s->resync_marker ? 0 : 1);/* resync marker disable */
-    put_bits(&s->pb, 1, s->data_partitioning ? 1 : 0);
-    if(s->data_partitioning){
-        put_bits(&s->pb, 1, 0);         /* no rvlc */
-    }
-
-    if (vo_ver_id != 1){
-        put_bits(&s->pb, 1, 0);         /* newpred */
-        put_bits(&s->pb, 1, 0);         /* reduced res vop */
-    }
-    put_bits(&s->pb, 1, 0);             /* scalability */
-
-    ff_mpeg4_stuffing(&s->pb);
-
-    /* user data */
-    if(!(s->flags & CODEC_FLAG_BITEXACT)){
-        put_bits(&s->pb, 16, 0);
-        put_bits(&s->pb, 16, 0x1B2);    /* user_data */
-        ff_put_string(&s->pb, LIBAVCODEC_IDENT, 0);
-    }
-}
-
-/* write mpeg4 VOP header */
-void mpeg4_encode_picture_header(MpegEncContext * s, int picture_number)
-{
-    int time_incr;
-    int time_div, time_mod;
-
-    if(s->pict_type==I_TYPE){
-        if(!(s->flags&CODEC_FLAG_GLOBAL_HEADER)){
-            if(s->strict_std_compliance < FF_COMPLIANCE_VERY_STRICT) //HACK, the reference sw is buggy
-                mpeg4_encode_visual_object_header(s);
-            if(s->strict_std_compliance < FF_COMPLIANCE_VERY_STRICT || picture_number==0) //HACK, the reference sw is buggy
-                mpeg4_encode_vol_header(s, 0, 0);
-        }
-        if(!(s->workaround_bugs & FF_BUG_MS))
-            mpeg4_encode_gop_header(s);
-    }
-
-    s->partitioned_frame= s->data_partitioning && s->pict_type!=B_TYPE;
-
-//printf("num:%d rate:%d base:%d\n", s->picture_number, s->time_base.den, FRAME_RATE_BASE);
-
-    put_bits(&s->pb, 16, 0);                /* vop header */
-    put_bits(&s->pb, 16, VOP_STARTCODE);    /* vop header */
-    put_bits(&s->pb, 2, s->pict_type - 1);  /* pict type: I = 0 , P = 1 */
-
-    assert(s->time>=0);
-    time_div= s->time/s->avctx->time_base.den;
-    time_mod= s->time%s->avctx->time_base.den;
-    time_incr= time_div - s->last_time_base;
-    assert(time_incr >= 0);
-    while(time_incr--)
-        put_bits(&s->pb, 1, 1);
-
-    put_bits(&s->pb, 1, 0);
-
-    put_bits(&s->pb, 1, 1);                             /* marker */
-    put_bits(&s->pb, s->time_increment_bits, time_mod); /* time increment */
-    put_bits(&s->pb, 1, 1);                             /* marker */
-    put_bits(&s->pb, 1, 1);                             /* vop coded */
-    if (    s->pict_type == P_TYPE
-        || (s->pict_type == S_TYPE && s->vol_sprite_usage==GMC_SPRITE)) {
-        put_bits(&s->pb, 1, s->no_rounding);    /* rounding type */
-    }
-    put_bits(&s->pb, 3, 0);     /* intra dc VLC threshold */
-    if(!s->progressive_sequence){
-         put_bits(&s->pb, 1, s->current_picture_ptr->top_field_first);
-         put_bits(&s->pb, 1, s->alternate_scan);
-    }
-    //FIXME sprite stuff
-
-    put_bits(&s->pb, 5, s->qscale);
-
-    if (s->pict_type != I_TYPE)
-        put_bits(&s->pb, 3, s->f_code); /* fcode_for */
-    if (s->pict_type == B_TYPE)
-        put_bits(&s->pb, 3, s->b_code); /* fcode_back */
-    //    printf("****frame %d\n", picture_number);
-}
-
-#endif //CONFIG_ENCODERS
-
-/**
- * set qscale and update qscale dependant 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 ];
-}
-
-/**
- * predicts the dc.
- * encoding quantized level -> quantized diff
- * decoding quantized diff -> quantized level
- * @param n block index (0-3 are luma, 4-5 are chroma)
- * @param dir_ptr pointer to an integer where the prediction direction will be stored
- */
-static inline int ff_mpeg4_pred_dc(MpegEncContext * s, int n, int level, int *dir_ptr, int encoding)
-{
-    int a, b, c, wrap, pred, scale, ret;
-    int16_t *dc_val;
-
-    /* find prediction */
-    if (n < 4) {
-        scale = s->y_dc_scale;
-    } else {
-        scale = s->c_dc_scale;
-    }
-    if(IS_3IV1)
-        scale= 8;
-
-    wrap= s->block_wrap[n];
-    dc_val = s->dc_val[0] + s->block_index[n];
-
-    /* B C
-     * A X
-     */
-    a = dc_val[ - 1];
-    b = dc_val[ - 1 - wrap];
-    c = dc_val[ - wrap];
-
-    /* outside slice handling (we can't do that by memset as we need the dc for error resilience) */
-    if(s->first_slice_line && n!=3){
-        if(n!=2) b=c= 1024;
-        if(n!=1 && s->mb_x == s->resync_mb_x) b=a= 1024;
-    }
-    if(s->mb_x == s->resync_mb_x && s->mb_y == s->resync_mb_y+1){
-        if(n==0 || n==4 || n==5)
-            b=1024;
-    }
-
-    if (abs(a - b) < abs(b - c)) {
-        pred = c;
-        *dir_ptr = 1; /* top */
-    } else {
-        pred = a;
-        *dir_ptr = 0; /* left */
-    }
-    /* we assume pred is positive */
-    pred = FASTDIV((pred + (scale >> 1)), scale);
-
-    if(encoding){
-        ret = level - pred;
-    }else{
-        level += pred;
-        ret= level;
-        if(s->error_resilience>=3){
-            if(level<0){
-                av_log(s->avctx, AV_LOG_ERROR, "dc<0 at %dx%d\n", s->mb_x, s->mb_y);
-                return -1;
-            }
-            if(level*scale > 2048 + scale){
-                av_log(s->avctx, AV_LOG_ERROR, "dc overflow at %dx%d\n", s->mb_x, s->mb_y);
-                return -1;
-            }
-        }
-    }
-    level *=scale;
-    if(level&(~2047)){
-        if(level<0)
-            level=0;
-        else if(!(s->workaround_bugs&FF_BUG_DC_CLIP))
-            level=2047;
-    }
-    dc_val[0]= level;
-
-    return ret;
-}
-
-/**
- * predicts the ac.
- * @param n block index (0-3 are luma, 4-5 are chroma)
- * @param dir the ac prediction direction
- */
-void mpeg4_pred_ac(MpegEncContext * s, DCTELEM *block, int n,
-                   int dir)
-{
-    int i;
-    int16_t *ac_val, *ac_val1;
-    int8_t * const qscale_table= s->current_picture.qscale_table;
-
-    /* find prediction */
-    ac_val = s->ac_val[0][0] + s->block_index[n] * 16;
-    ac_val1 = ac_val;
-    if (s->ac_pred) {
-        if (dir == 0) {
-            const int xy= s->mb_x-1 + s->mb_y*s->mb_stride;
-            /* left prediction */
-            ac_val -= 16;
-
-            if(s->mb_x==0 || s->qscale == qscale_table[xy] || n==1 || n==3){
-                /* same qscale */
-                for(i=1;i<8;i++) {
-                    block[s->dsp.idct_permutation[i<<3]] += ac_val[i];
-                }
-            }else{
-                /* different qscale, we must rescale */
-                for(i=1;i<8;i++) {
-                    block[s->dsp.idct_permutation[i<<3]] += ROUNDED_DIV(ac_val[i]*qscale_table[xy], s->qscale);
-                }
-            }
-        } else {
-            const int xy= s->mb_x + s->mb_y*s->mb_stride - s->mb_stride;
-            /* top prediction */
-            ac_val -= 16 * s->block_wrap[n];
-
-            if(s->mb_y==0 || s->qscale == qscale_table[xy] || n==2 || n==3){
-                /* same qscale */
-                for(i=1;i<8;i++) {
-                    block[s->dsp.idct_permutation[i]] += ac_val[i + 8];
-                }
-            }else{
-                /* different qscale, we must rescale */
-                for(i=1;i<8;i++) {
-                    block[s->dsp.idct_permutation[i]] += ROUNDED_DIV(ac_val[i + 8]*qscale_table[xy], s->qscale);
-                }
-            }
-        }
-    }
-    /* left copy */
-    for(i=1;i<8;i++)
-        ac_val1[i    ] = block[s->dsp.idct_permutation[i<<3]];
-
-    /* top copy */
-    for(i=1;i<8;i++)
-        ac_val1[8 + i] = block[s->dsp.idct_permutation[i   ]];
-
-}
-
-#ifdef CONFIG_ENCODERS
-
-/**
- * encodes the dc value.
- * @param n block index (0-3 are luma, 4-5 are chroma)
- */
-static inline void mpeg4_encode_dc(PutBitContext * s, int level, int n)
-{
-#if 1
-//    if(level<-255 || level>255) printf("dc overflow\n");
-    level+=256;
-    if (n < 4) {
-        /* luminance */
-        put_bits(s, uni_DCtab_lum_len[level], uni_DCtab_lum_bits[level]);
-    } else {
-        /* chrominance */
-        put_bits(s, uni_DCtab_chrom_len[level], uni_DCtab_chrom_bits[level]);
-    }
-#else
-    int size, v;
-    /* find number of bits */
-    size = 0;
-    v = abs(level);
-    while (v) {
-        v >>= 1;
-        size++;
-    }
-
-    if (n < 4) {
-        /* luminance */
-        put_bits(&s->pb, DCtab_lum[size][1], DCtab_lum[size][0]);
-    } else {
-        /* chrominance */
-        put_bits(&s->pb, DCtab_chrom[size][1], DCtab_chrom[size][0]);
-    }
-
-    /* encode remaining bits */
-    if (size > 0) {
-        if (level < 0)
-            level = (-level) ^ ((1 << size) - 1);
-        put_bits(&s->pb, size, level);
-        if (size > 8)
-            put_bits(&s->pb, 1, 1);
-    }
-#endif
-}
-
-static inline int mpeg4_get_dc_length(int level, int n){
-    if (n < 4) {
-        return uni_DCtab_lum_len[level + 256];
-    } else {
-        return uni_DCtab_chrom_len[level + 256];
-    }
-}
-
-/**
- * encodes a 8x8 block
- * @param n block index (0-3 are luma, 4-5 are chroma)
- */
-static inline void mpeg4_encode_block(MpegEncContext * s, DCTELEM * block, int n, int intra_dc,
-                               uint8_t *scan_table, PutBitContext *dc_pb, PutBitContext *ac_pb)
-{
-    int i, last_non_zero;
-#if 0 //variables for the outcommented version
-    int code, sign, last;
-#endif
-    const RLTable *rl;
-    uint32_t *bits_tab;
-    uint8_t *len_tab;
-    const int last_index = s->block_last_index[n];
-
-    if (s->mb_intra) { //Note gcc (3.2.1 at least) will optimize this away
-        /* mpeg4 based DC predictor */
-        mpeg4_encode_dc(dc_pb, intra_dc, n);
-        if(last_index<1) return;
-        i = 1;
-        rl = &rl_intra;
-        bits_tab= uni_mpeg4_intra_rl_bits;
-        len_tab = uni_mpeg4_intra_rl_len;
-    } else {
-        if(last_index<0) return;
-        i = 0;
-        rl = &rl_inter;
-        bits_tab= uni_mpeg4_inter_rl_bits;
-        len_tab = uni_mpeg4_inter_rl_len;
-    }
-
-    /* AC coefs */
-    last_non_zero = i - 1;
-#if 1
-    for (; i < last_index; i++) {
-        int level = block[ scan_table[i] ];
-        if (level) {
-            int run = i - last_non_zero - 1;
-            level+=64;
-            if((level&(~127)) == 0){
-                const int index= UNI_MPEG4_ENC_INDEX(0, run, level);
-                put_bits(ac_pb, len_tab[index], bits_tab[index]);
-            }else{ //ESC3
-                put_bits(ac_pb, 7+2+1+6+1+12+1, (3<<23)+(3<<21)+(0<<20)+(run<<14)+(1<<13)+(((level-64)&0xfff)<<1)+1);
-            }
-            last_non_zero = i;
-        }
-    }
-    /*if(i<=last_index)*/{
-        int level = block[ scan_table[i] ];
-        int run = i - last_non_zero - 1;
-        level+=64;
-        if((level&(~127)) == 0){
-            const int index= UNI_MPEG4_ENC_INDEX(1, run, level);
-            put_bits(ac_pb, len_tab[index], bits_tab[index]);
-        }else{ //ESC3
-            put_bits(ac_pb, 7+2+1+6+1+12+1, (3<<23)+(3<<21)+(1<<20)+(run<<14)+(1<<13)+(((level-64)&0xfff)<<1)+1);
-        }
-    }
-#else
-    for (; i <= last_index; i++) {
-        const int slevel = block[ scan_table[i] ];
-        if (slevel) {
-            int level;
-            int run = i - last_non_zero - 1;
-            last = (i == last_index);
-            sign = 0;
-            level = slevel;
-            if (level < 0) {
-                sign = 1;
-                level = -level;
-            }
-            code = get_rl_index(rl, last, run, level);
-            put_bits(ac_pb, rl->table_vlc[code][1], rl->table_vlc[code][0]);
-            if (code == rl->n) {
-                int level1, run1;
-                level1 = level - rl->max_level[last][run];
-                if (level1 < 1)
-                    goto esc2;
-                code = get_rl_index(rl, last, run, level1);
-                if (code == rl->n) {
-                esc2:
-                    put_bits(ac_pb, 1, 1);
-                    if (level > MAX_LEVEL)
-                        goto esc3;
-                    run1 = run - rl->max_run[last][level] - 1;
-                    if (run1 < 0)
-                        goto esc3;
-                    code = get_rl_index(rl, last, run1, level);
-                    if (code == rl->n) {
-                    esc3:
-                        /* third escape */
-                        put_bits(ac_pb, 1, 1);
-                        put_bits(ac_pb, 1, last);
-                        put_bits(ac_pb, 6, run);
-                        put_bits(ac_pb, 1, 1);
-                        put_bits(ac_pb, 12, slevel & 0xfff);
-                        put_bits(ac_pb, 1, 1);
-                    } else {
-                        /* second escape */
-                        put_bits(ac_pb, 1, 0);
-                        put_bits(ac_pb, rl->table_vlc[code][1], rl->table_vlc[code][0]);
-                        put_bits(ac_pb, 1, sign);
-                    }
-                } else {
-                    /* first escape */
-                    put_bits(ac_pb, 1, 0);
-                    put_bits(ac_pb, rl->table_vlc[code][1], rl->table_vlc[code][0]);
-                    put_bits(ac_pb, 1, sign);
-                }
-            } else {
-                put_bits(ac_pb, 1, sign);
-            }
-            last_non_zero = i;
-        }
-    }
-#endif
-}
-
-static int mpeg4_get_block_length(MpegEncContext * s, DCTELEM * block, int n, int intra_dc,
-                               uint8_t *scan_table)
-{
-    int i, last_non_zero;
-    const RLTable *rl;
-    uint8_t *len_tab;
-    const int last_index = s->block_last_index[n];
-    int len=0;
-
-    if (s->mb_intra) { //Note gcc (3.2.1 at least) will optimize this away
-        /* mpeg4 based DC predictor */
-        len += mpeg4_get_dc_length(intra_dc, n);
-        if(last_index<1) return len;
-        i = 1;
-        rl = &rl_intra;
-        len_tab = uni_mpeg4_intra_rl_len;
-    } else {
-        if(last_index<0) return 0;
-        i = 0;
-        rl = &rl_inter;
-        len_tab = uni_mpeg4_inter_rl_len;
-    }
-
-    /* AC coefs */
-    last_non_zero = i - 1;
-    for (; i < last_index; i++) {
-        int level = block[ scan_table[i] ];
-        if (level) {
-            int run = i - last_non_zero - 1;
-            level+=64;
-            if((level&(~127)) == 0){
-                const int index= UNI_MPEG4_ENC_INDEX(0, run, level);
-                len += len_tab[index];
-            }else{ //ESC3
-                len += 7+2+1+6+1+12+1;
-            }
-            last_non_zero = i;
-        }
-    }
-    /*if(i<=last_index)*/{
-        int level = block[ scan_table[i] ];
-        int run = i - last_non_zero - 1;
-        level+=64;
-        if((level&(~127)) == 0){
-            const int index= UNI_MPEG4_ENC_INDEX(1, run, level);
-            len += len_tab[index];
-        }else{ //ESC3
-            len += 7+2+1+6+1+12+1;
-        }
-    }
-
-    return len;
-}
-
-#endif
-
-
-/***********************************************/
-/* decoding */
-
-static VLC intra_MCBPC_vlc;
-static VLC inter_MCBPC_vlc;
-static VLC cbpy_vlc;
-static VLC mv_vlc;
-static VLC dc_lum, dc_chrom;
-static VLC sprite_trajectory;
-static VLC mb_type_b_vlc;
-static VLC h263_mbtype_b_vlc;
-static VLC cbpc_b_vlc;
-
-void init_vlc_rl(RLTable *rl, int use_static)
-{
-    int i, q;
-
-    /* Return if static table is already initialized */
-    if(use_static && rl->rl_vlc[0])
-        return;
-
-    init_vlc(&rl->vlc, 9, rl->n + 1,
-             &rl->table_vlc[0][1], 4, 2,
-             &rl->table_vlc[0][0], 4, 2, use_static);
-
-
-    for(q=0; q<32; q++){
-        int qmul= q*2;
-        int qadd= (q-1)|1;
-
-        if(q==0){
-            qmul=1;
-            qadd=0;
-        }
-        if(use_static)
-            rl->rl_vlc[q]= av_mallocz_static(rl->vlc.table_size*sizeof(RL_VLC_ELEM));
-        else
-            rl->rl_vlc[q]= av_malloc(rl->vlc.table_size*sizeof(RL_VLC_ELEM));
-        for(i=0; i<rl->vlc.table_size; i++){
-            int code= rl->vlc.table[i][0];
-            int len = rl->vlc.table[i][1];
-            int level, run;
-
-            if(len==0){ // illegal code
-                run= 66;
-                level= MAX_LEVEL;
-            }else if(len<0){ //more bits needed
-                run= 0;
-                level= code;
-            }else{
-                if(code==rl->n){ //esc
-                    run= 66;
-                    level= 0;
-                }else{
-                    run=   rl->table_run  [code] + 1;
-                    level= rl->table_level[code] * qmul + qadd;
-                    if(code >= rl->last) run+=192;
-                }
-            }
-            rl->rl_vlc[q][i].len= len;
-            rl->rl_vlc[q][i].level= level;
-            rl->rl_vlc[q][i].run= run;
-        }
-    }
-}
-
-/* init vlcs */
-
-/* XXX: find a better solution to handle static init */
-void h263_decode_init_vlc(MpegEncContext *s)
-{
-    static int done = 0;
-
-    if (!done) {
-        done = 1;
-
-        init_vlc(&intra_MCBPC_vlc, INTRA_MCBPC_VLC_BITS, 9,
-                 intra_MCBPC_bits, 1, 1,
-                 intra_MCBPC_code, 1, 1, 1);
-        init_vlc(&inter_MCBPC_vlc, INTER_MCBPC_VLC_BITS, 28,
-                 inter_MCBPC_bits, 1, 1,
-                 inter_MCBPC_code, 1, 1, 1);
-        init_vlc(&cbpy_vlc, CBPY_VLC_BITS, 16,
-                 &cbpy_tab[0][1], 2, 1,
-                 &cbpy_tab[0][0], 2, 1, 1);
-        init_vlc(&mv_vlc, MV_VLC_BITS, 33,
-                 &mvtab[0][1], 2, 1,
-                 &mvtab[0][0], 2, 1, 1);
-        init_rl(&rl_inter, 1);
-        init_rl(&rl_intra, 1);
-        init_rl(&rvlc_rl_inter, 1);
-        init_rl(&rvlc_rl_intra, 1);
-        init_rl(&rl_intra_aic, 1);
-        init_vlc_rl(&rl_inter, 1);
-        init_vlc_rl(&rl_intra, 1);
-        init_vlc_rl(&rvlc_rl_inter, 1);
-        init_vlc_rl(&rvlc_rl_intra, 1);
-        init_vlc_rl(&rl_intra_aic, 1);
-        init_vlc(&dc_lum, DC_VLC_BITS, 10 /* 13 */,
-                 &DCtab_lum[0][1], 2, 1,
-                 &DCtab_lum[0][0], 2, 1, 1);
-        init_vlc(&dc_chrom, DC_VLC_BITS, 10 /* 13 */,
-                 &DCtab_chrom[0][1], 2, 1,
-                 &DCtab_chrom[0][0], 2, 1, 1);
-        init_vlc(&sprite_trajectory, SPRITE_TRAJ_VLC_BITS, 15,
-                 &sprite_trajectory_tab[0][1], 4, 2,
-                 &sprite_trajectory_tab[0][0], 4, 2, 1);
-        init_vlc(&mb_type_b_vlc, MB_TYPE_B_VLC_BITS, 4,
-                 &mb_type_b_tab[0][1], 2, 1,
-                 &mb_type_b_tab[0][0], 2, 1, 1);
-        init_vlc(&h263_mbtype_b_vlc, H263_MBTYPE_B_VLC_BITS, 15,
-                 &h263_mbtype_b_tab[0][1], 2, 1,
-                 &h263_mbtype_b_tab[0][0], 2, 1, 1);
-        init_vlc(&cbpc_b_vlc, CBPC_B_VLC_BITS, 4,
-                 &cbpc_b_tab[0][1], 2, 1,
-                 &cbpc_b_tab[0][0], 2, 1, 1);
-    }
-}
-
-/**
- * Get the GOB height based on picture height.
- */
-int ff_h263_get_gob_height(MpegEncContext *s){
-    if (s->height <= 400)
-        return 1;
-    else if (s->height <= 800)
-        return  2;
-    else
-        return 4;
-}
-
-int ff_h263_decode_mba(MpegEncContext *s)
-{
-    int i, mb_pos;
-
-    for(i=0; i<6; i++){
-        if(s->mb_num-1 <= ff_mba_max[i]) break;
-    }
-    mb_pos= get_bits(&s->gb, ff_mba_length[i]);
-    s->mb_x= mb_pos % s->mb_width;
-    s->mb_y= mb_pos / s->mb_width;
-
-    return mb_pos;
-}
-
-void ff_h263_encode_mba(MpegEncContext *s)
-{
-    int i, mb_pos;
-
-    for(i=0; i<6; i++){
-        if(s->mb_num-1 <= ff_mba_max[i]) break;
-    }
-    mb_pos= s->mb_x + s->mb_width*s->mb_y;
-    put_bits(&s->pb, ff_mba_length[i], mb_pos);
-}
-
-/**
- * decodes the group of blocks header or slice header.
- * @return <0 if an error occured
- */
-static int h263_decode_gob_header(MpegEncContext *s)
-{
-    unsigned int val, gfid, gob_number;
-    int left;
-
-    /* Check for GOB Start Code */
-    val = show_bits(&s->gb, 16);
-    if(val)
-        return -1;
-
-        /* We have a GBSC probably with GSTUFF */
-    skip_bits(&s->gb, 16); /* Drop the zeros */
-    left= s->gb.size_in_bits - get_bits_count(&s->gb);
-    //MN: we must check the bits left or we might end in a infinite loop (or segfault)
-    for(;left>13; left--){
-        if(get_bits1(&s->gb)) break; /* Seek the '1' bit */
-    }
-    if(left<=13)
-        return -1;
-
-    if(s->h263_slice_structured){
-        if(get_bits1(&s->gb)==0)
-            return -1;
-
-        ff_h263_decode_mba(s);
-
-        if(s->mb_num > 1583)
-            if(get_bits1(&s->gb)==0)
-                return -1;
-
-        s->qscale = get_bits(&s->gb, 5); /* SQUANT */
-        if(get_bits1(&s->gb)==0)
-            return -1;
-        gfid = get_bits(&s->gb, 2); /* GFID */
-    }else{
-        gob_number = get_bits(&s->gb, 5); /* GN */
-        s->mb_x= 0;
-        s->mb_y= s->gob_index* gob_number;
-        gfid = get_bits(&s->gb, 2); /* GFID */
-        s->qscale = get_bits(&s->gb, 5); /* GQUANT */
-    }
-
-    if(s->mb_y >= s->mb_height)
-        return -1;
-
-    if(s->qscale==0)
-        return -1;
-
-    return 0;
-}
-
-static inline void memsetw(short *tab, int val, int n)
-{
-    int i;
-    for(i=0;i<n;i++)
-        tab[i] = val;
-}
-
-#ifdef CONFIG_ENCODERS
-
-void ff_mpeg4_init_partitions(MpegEncContext *s)
-{
-    uint8_t *start= pbBufPtr(&s->pb);
-    uint8_t *end= s->pb.buf_end;
-    int size= end - start;
-    int pb_size = (((long)start + size/3)&(~3)) - (long)start;
-    int tex_size= (size - 2*pb_size)&(~3);
-
-    set_put_bits_buffer_size(&s->pb, pb_size);
-    init_put_bits(&s->tex_pb, start + pb_size           , tex_size);
-    init_put_bits(&s->pb2   , start + pb_size + tex_size, pb_size);
-}
-
-void ff_mpeg4_merge_partitions(MpegEncContext *s)
-{
-    const int pb2_len   = put_bits_count(&s->pb2   );
-    const int tex_pb_len= put_bits_count(&s->tex_pb);
-    const int bits= put_bits_count(&s->pb);
-
-    if(s->pict_type==I_TYPE){
-        put_bits(&s->pb, 19, DC_MARKER);
-        s->misc_bits+=19 + pb2_len + bits - s->last_bits;
-        s->i_tex_bits+= tex_pb_len;
-    }else{
-        put_bits(&s->pb, 17, MOTION_MARKER);
-        s->misc_bits+=17 + pb2_len;
-        s->mv_bits+= bits - s->last_bits;
-        s->p_tex_bits+= tex_pb_len;
-    }
-
-    flush_put_bits(&s->pb2);
-    flush_put_bits(&s->tex_pb);
-
-    set_put_bits_buffer_size(&s->pb, s->pb2.buf_end - s->pb.buf);
-    ff_copy_bits(&s->pb, s->pb2.buf   , pb2_len);
-    ff_copy_bits(&s->pb, s->tex_pb.buf, tex_pb_len);
-    s->last_bits= put_bits_count(&s->pb);
-}
-
-#endif //CONFIG_ENCODERS
-
-int ff_mpeg4_get_video_packet_prefix_length(MpegEncContext *s){
-    switch(s->pict_type){
-        case I_TYPE:
-            return 16;
-        case P_TYPE:
-        case S_TYPE:
-            return s->f_code+15;
-        case B_TYPE:
-            return FFMAX(FFMAX(s->f_code, s->b_code)+15, 17);
-        default:
-            return -1;
-    }
-}
-
-#ifdef CONFIG_ENCODERS
-
-void ff_mpeg4_encode_video_packet_header(MpegEncContext *s)
-{
-    int mb_num_bits= av_log2(s->mb_num - 1) + 1;
-
-    put_bits(&s->pb, ff_mpeg4_get_video_packet_prefix_length(s), 0);
-    put_bits(&s->pb, 1, 1);
-
-    put_bits(&s->pb, mb_num_bits, s->mb_x + s->mb_y*s->mb_width);
-    put_bits(&s->pb, s->quant_precision, s->qscale);
-    put_bits(&s->pb, 1, 0); /* no HEC */
-}
-
-#endif //CONFIG_ENCODERS
-
-/**
- * check if the next stuff is a resync marker or the end.
- * @return 0 if not
- */
-static inline int mpeg4_is_resync(MpegEncContext *s){
-    const int bits_count= get_bits_count(&s->gb);
-
-    if(s->workaround_bugs&FF_BUG_NO_PADDING){
-        return 0;
-    }
-
-    if(bits_count + 8 >= s->gb.size_in_bits){
-        int v= show_bits(&s->gb, 8);
-        v|= 0x7F >> (7-(bits_count&7));
-
-        if(v==0x7F)
-            return 1;
-    }else{
-        if(show_bits(&s->gb, 16) == ff_mpeg4_resync_prefix[bits_count&7]){
-            int len;
-            GetBitContext gb= s->gb;
-
-            skip_bits(&s->gb, 1);
-            align_get_bits(&s->gb);
-
-            for(len=0; len<32; len++){
-                if(get_bits1(&s->gb)) break;
-            }
-
-            s->gb= gb;
-
-            if(len>=ff_mpeg4_get_video_packet_prefix_length(s))
-                return 1;
-        }
-    }
-    return 0;
-}
-
-/**
- * decodes the next video packet.
- * @return <0 if something went wrong
- */
-static int mpeg4_decode_video_packet_header(MpegEncContext *s)
-{
-    int mb_num_bits= av_log2(s->mb_num - 1) + 1;
-    int header_extension=0, mb_num, len;
-
-    /* is there enough space left for a video packet + header */
-    if( get_bits_count(&s->gb) > s->gb.size_in_bits-20) return -1;
-
-    for(len=0; len<32; len++){
-        if(get_bits1(&s->gb)) break;
-    }
-
-    if(len!=ff_mpeg4_get_video_packet_prefix_length(s)){
-        av_log(s->avctx, AV_LOG_ERROR, "marker does not match f_code\n");
-        return -1;
-    }
-
-    if(s->shape != RECT_SHAPE){
-        header_extension= get_bits1(&s->gb);
-        //FIXME more stuff here
-    }
-
-    mb_num= get_bits(&s->gb, mb_num_bits);
-    if(mb_num>=s->mb_num){
-        av_log(s->avctx, AV_LOG_ERROR, "illegal mb_num in video packet (%d %d) \n", mb_num, s->mb_num);
-        return -1;
-    }
-    if(s->pict_type == B_TYPE){
-        while(s->next_picture.mbskip_table[ s->mb_index2xy[ mb_num ] ]) mb_num++;
-        if(mb_num >= s->mb_num) return -1; // slice contains just skipped MBs which where allready decoded
-    }
-
-    s->mb_x= mb_num % s->mb_width;
-    s->mb_y= mb_num / s->mb_width;
-
-    if(s->shape != BIN_ONLY_SHAPE){
-        int qscale= get_bits(&s->gb, s->quant_precision);
-        if(qscale)
-            s->chroma_qscale=s->qscale= qscale;
-    }
-
-    if(s->shape == RECT_SHAPE){
-        header_extension= get_bits1(&s->gb);
-    }
-    if(header_extension){
-        int time_increment;
-        int time_incr=0;
-
-        while (get_bits1(&s->gb) != 0)
-            time_incr++;
-
-        check_marker(&s->gb, "before time_increment in video packed header");
-        time_increment= get_bits(&s->gb, s->time_increment_bits);
-        check_marker(&s->gb, "before vop_coding_type in video packed header");
-
-        skip_bits(&s->gb, 2); /* vop coding type */
-        //FIXME not rect stuff here
-
-        if(s->shape != BIN_ONLY_SHAPE){
-            skip_bits(&s->gb, 3); /* intra dc vlc threshold */
-//FIXME don't just ignore everything
-            if(s->pict_type == S_TYPE && s->vol_sprite_usage==GMC_SPRITE){
-                mpeg4_decode_sprite_trajectory(s, &s->gb);
-                av_log(s->avctx, AV_LOG_ERROR, "untested\n");
-            }
-
-            //FIXME reduced res stuff here
-
-            if (s->pict_type != I_TYPE) {
-                int f_code = get_bits(&s->gb, 3);       /* fcode_for */
-                if(f_code==0){
-                    av_log(s->avctx, AV_LOG_ERROR, "Error, video packet header damaged (f_code=0)\n");
-                }
-            }
-            if (s->pict_type == B_TYPE) {
-                int b_code = get_bits(&s->gb, 3);
-                if(b_code==0){
-                    av_log(s->avctx, AV_LOG_ERROR, "Error, video packet header damaged (b_code=0)\n");
-                }
-            }
-        }
-    }
-    //FIXME new-pred stuff
-
-//printf("parse ok %d %d %d %d\n", mb_num, s->mb_x + s->mb_y*s->mb_width, get_bits_count(gb), get_bits_count(&s->gb));
-
-    return 0;
-}
-
-void ff_mpeg4_clean_buffers(MpegEncContext *s)
-{
-    int c_wrap, c_xy, l_wrap, l_xy;
-
-    l_wrap= s->b8_stride;
-    l_xy= (2*s->mb_y-1)*l_wrap + s->mb_x*2 - 1;
-    c_wrap= s->mb_stride;
-    c_xy= (s->mb_y-1)*c_wrap + s->mb_x - 1;
-
-#if 0
-    /* clean DC */
-    memsetw(s->dc_val[0] + l_xy, 1024, l_wrap*2+1);
-    memsetw(s->dc_val[1] + c_xy, 1024, c_wrap+1);
-    memsetw(s->dc_val[2] + c_xy, 1024, c_wrap+1);
-#endif
-
-    /* clean AC */
-    memset(s->ac_val[0] + l_xy, 0, (l_wrap*2+1)*16*sizeof(int16_t));
-    memset(s->ac_val[1] + c_xy, 0, (c_wrap  +1)*16*sizeof(int16_t));
-    memset(s->ac_val[2] + c_xy, 0, (c_wrap  +1)*16*sizeof(int16_t));
-
-    /* clean MV */
-    // we can't clear the MVs as they might be needed by a b frame
-//    memset(s->motion_val + l_xy, 0, (l_wrap*2+1)*2*sizeof(int16_t));
-//    memset(s->motion_val, 0, 2*sizeof(int16_t)*(2 + s->mb_width*2)*(2 + s->mb_height*2));
-    s->last_mv[0][0][0]=
-    s->last_mv[0][0][1]=
-    s->last_mv[1][0][0]=
-    s->last_mv[1][0][1]= 0;
-}
-
-/**
- * decodes the group of blocks / video packet header.
- * @return <0 if no resync found
- */
-int ff_h263_resync(MpegEncContext *s){
-    int left, ret;
-
-    if(s->codec_id==CODEC_ID_MPEG4){
-        skip_bits1(&s->gb);
-        align_get_bits(&s->gb);
-    }
-
-    if(show_bits(&s->gb, 16)==0){
-        if(s->codec_id==CODEC_ID_MPEG4)
-            ret= mpeg4_decode_video_packet_header(s);
-        else
-            ret= h263_decode_gob_header(s);
-        if(ret>=0)
-            return 0;
-    }
-    //ok, it's not where its supposed to be ...
-    s->gb= s->last_resync_gb;
-    align_get_bits(&s->gb);
-    left= s->gb.size_in_bits - get_bits_count(&s->gb);
-
-    for(;left>16+1+5+5; left-=8){
-        if(show_bits(&s->gb, 16)==0){
-            GetBitContext bak= s->gb;
-
-            if(s->codec_id==CODEC_ID_MPEG4)
-                ret= mpeg4_decode_video_packet_header(s);
-            else
-                ret= h263_decode_gob_header(s);
-            if(ret>=0)
-                return 0;
-
-            s->gb= bak;
-        }
-        skip_bits(&s->gb, 8);
-    }
-
-    return -1;
-}
-
-/**
- * gets the average motion vector for a GMC MB.
- * @param n either 0 for the x component or 1 for y
- * @returns the average MV for a GMC MB
- */
-static inline int get_amv(MpegEncContext *s, int n){
-    int x, y, mb_v, sum, dx, dy, shift;
-    int len = 1 << (s->f_code + 4);
-    const int a= s->sprite_warping_accuracy;
-
-    if(s->workaround_bugs & FF_BUG_AMV)
-        len >>= s->quarter_sample;
-
-    if(s->real_sprite_warping_points==1){
-        if(s->divx_version==500 && s->divx_build==413)
-            sum= s->sprite_offset[0][n] / (1<<(a - s->quarter_sample));
-        else
-            sum= RSHIFT(s->sprite_offset[0][n]<<s->quarter_sample, a);
-    }else{
-        dx= s->sprite_delta[n][0];
-        dy= s->sprite_delta[n][1];
-        shift= s->sprite_shift[0];
-        if(n) dy -= 1<<(shift + a + 1);
-        else  dx -= 1<<(shift + a + 1);
-        mb_v= s->sprite_offset[0][n] + dx*s->mb_x*16 + dy*s->mb_y*16;
-
-        sum=0;
-        for(y=0; y<16; y++){
-            int v;
-
-            v= mb_v + dy*y;
-            //XXX FIXME optimize
-            for(x=0; x<16; x++){
-                sum+= v>>shift;
-                v+= dx;
-            }
-        }
-        sum= RSHIFT(sum, a+8-s->quarter_sample);
-    }
-
-    if      (sum < -len) sum= -len;
-    else if (sum >= len) sum= len-1;
-
-    return sum;
-}
-
-/**
- * decodes first partition.
- * @return number of MBs decoded or <0 if an error occured
- */
-static int mpeg4_decode_partition_a(MpegEncContext *s){
-    int mb_num;
-    static const int8_t quant_tab[4] = { -1, -2, 1, 2 };
-
-    /* decode first partition */
-    mb_num=0;
-    s->first_slice_line=1;
-    for(; s->mb_y<s->mb_height; s->mb_y++){
-        ff_init_block_index(s);
-        for(; s->mb_x<s->mb_width; s->mb_x++){
-            const int xy= s->mb_x + s->mb_y*s->mb_stride;
-            int cbpc;
-            int dir=0;
-
-            mb_num++;
-            ff_update_block_index(s);
-            if(s->mb_x == s->resync_mb_x && s->mb_y == s->resync_mb_y+1)
-                s->first_slice_line=0;
-
-            if(s->pict_type==I_TYPE){
-                int i;
-
-                do{
-                    if(show_bits_long(&s->gb, 19)==DC_MARKER){
-                        return mb_num-1;
-                    }
-
-                    cbpc = get_vlc2(&s->gb, intra_MCBPC_vlc.table, INTRA_MCBPC_VLC_BITS, 2);
-                    if (cbpc < 0){
-                        av_log(s->avctx, AV_LOG_ERROR, "cbpc corrupted at %d %d\n", s->mb_x, s->mb_y);
-                        return -1;
-                    }
-                }while(cbpc == 8);
-
-                s->cbp_table[xy]= cbpc & 3;
-                s->current_picture.mb_type[xy]= MB_TYPE_INTRA;
-                s->mb_intra = 1;
-
-                if(cbpc & 4) {
-                    ff_set_qscale(s, s->qscale + quant_tab[get_bits(&s->gb, 2)]);
-                }
-                s->current_picture.qscale_table[xy]= s->qscale;
-
-                s->mbintra_table[xy]= 1;
-                for(i=0; i<6; i++){
-                    int dc_pred_dir;
-                    int dc= mpeg4_decode_dc(s, i, &dc_pred_dir);
-                    if(dc < 0){
-                        av_log(s->avctx, AV_LOG_ERROR, "DC corrupted at %d %d\n", s->mb_x, s->mb_y);
-                        return -1;
-                    }
-                    dir<<=1;
-                    if(dc_pred_dir) dir|=1;
-                }
-                s->pred_dir_table[xy]= dir;
-            }else{ /* P/S_TYPE */
-                int mx, my, pred_x, pred_y, bits;
-                int16_t * const mot_val= s->current_picture.motion_val[0][s->block_index[0]];
-                const int stride= s->b8_stride*2;
-
-try_again:
-                bits= show_bits(&s->gb, 17);
-                if(bits==MOTION_MARKER){
-                    return mb_num-1;
-                }
-                skip_bits1(&s->gb);
-                if(bits&0x10000){
-                    /* skip mb */
-                    if(s->pict_type==S_TYPE && s->vol_sprite_usage==GMC_SPRITE){
-                        s->current_picture.mb_type[xy]= MB_TYPE_SKIP | MB_TYPE_16x16 | MB_TYPE_GMC | MB_TYPE_L0;
-                        mx= get_amv(s, 0);
-                        my= get_amv(s, 1);
-                    }else{
-                        s->current_picture.mb_type[xy]= MB_TYPE_SKIP | MB_TYPE_16x16 | MB_TYPE_L0;
-                        mx=my=0;
-                    }
-                    mot_val[0       ]= mot_val[2       ]=
-                    mot_val[0+stride]= mot_val[2+stride]= mx;
-                    mot_val[1       ]= mot_val[3       ]=
-                    mot_val[1+stride]= mot_val[3+stride]= my;
-
-                    if(s->mbintra_table[xy])
-                        ff_clean_intra_table_entries(s);
-                    continue;
-                }
-
-                cbpc = get_vlc2(&s->gb, inter_MCBPC_vlc.table, INTER_MCBPC_VLC_BITS, 2);
-                if (cbpc < 0){
-                    av_log(s->avctx, AV_LOG_ERROR, "cbpc corrupted at %d %d\n", s->mb_x, s->mb_y);
-                    return -1;
-                }
-                if(cbpc == 20)
-                    goto try_again;
-
-                s->cbp_table[xy]= cbpc&(8+3); //8 is dquant
-
-                s->mb_intra = ((cbpc & 4) != 0);
-
-                if(s->mb_intra){
-                    s->current_picture.mb_type[xy]= MB_TYPE_INTRA;
-                    s->mbintra_table[xy]= 1;
-                    mot_val[0       ]= mot_val[2       ]=
-                    mot_val[0+stride]= mot_val[2+stride]= 0;
-                    mot_val[1       ]= mot_val[3       ]=
-                    mot_val[1+stride]= mot_val[3+stride]= 0;
-                }else{
-                    if(s->mbintra_table[xy])
-                        ff_clean_intra_table_entries(s);
-
-                    if(s->pict_type==S_TYPE && s->vol_sprite_usage==GMC_SPRITE && (cbpc & 16) == 0)
-                        s->mcsel= get_bits1(&s->gb);
-                    else s->mcsel= 0;
-
-                    if ((cbpc & 16) == 0) {
-                        /* 16x16 motion prediction */
-
-                        h263_pred_motion(s, 0, 0, &pred_x, &pred_y);
-                        if(!s->mcsel){
-                            mx = h263_decode_motion(s, pred_x, s->f_code);
-                            if (mx >= 0xffff)
-                                return -1;
-
-                            my = h263_decode_motion(s, pred_y, s->f_code);
-                            if (my >= 0xffff)
-                                return -1;
-                            s->current_picture.mb_type[xy]= MB_TYPE_16x16 | MB_TYPE_L0;
-                        } else {
-                            mx = get_amv(s, 0);
-                            my = get_amv(s, 1);
-                            s->current_picture.mb_type[xy]= MB_TYPE_16x16 | MB_TYPE_GMC | MB_TYPE_L0;
-                        }
-
-                        mot_val[0       ]= mot_val[2       ] =
-                        mot_val[0+stride]= mot_val[2+stride]= mx;
-                        mot_val[1       ]= mot_val[3       ]=
-                        mot_val[1+stride]= mot_val[3+stride]= my;
-                    } else {
-                        int i;
-                        s->current_picture.mb_type[xy]= MB_TYPE_8x8 | MB_TYPE_L0;
-                        for(i=0;i<4;i++) {
-                            int16_t *mot_val= h263_pred_motion(s, i, 0, &pred_x, &pred_y);
-                            mx = h263_decode_motion(s, pred_x, s->f_code);
-                            if (mx >= 0xffff)
-                                return -1;
-
-                            my = h263_decode_motion(s, pred_y, s->f_code);
-                            if (my >= 0xffff)
-                                return -1;
-                            mot_val[0] = mx;
-                            mot_val[1] = my;
-                        }
-                    }
-                }
-            }
-        }
-        s->mb_x= 0;
-    }
-
-    return mb_num;
-}
-
-/**
- * decode second partition.
- * @return <0 if an error occured
- */
-static int mpeg4_decode_partition_b(MpegEncContext *s, int mb_count){
-    int mb_num=0;
-    static const int8_t quant_tab[4] = { -1, -2, 1, 2 };
-
-    s->mb_x= s->resync_mb_x;
-    s->first_slice_line=1;
-    for(s->mb_y= s->resync_mb_y; mb_num < mb_count; s->mb_y++){
-        ff_init_block_index(s);
-        for(; mb_num < mb_count && s->mb_x<s->mb_width; s->mb_x++){
-            const int xy= s->mb_x + s->mb_y*s->mb_stride;
-
-            mb_num++;
-            ff_update_block_index(s);
-            if(s->mb_x == s->resync_mb_x && s->mb_y == s->resync_mb_y+1)
-                s->first_slice_line=0;
-
-            if(s->pict_type==I_TYPE){
-                int ac_pred= get_bits1(&s->gb);
-                int cbpy = get_vlc2(&s->gb, cbpy_vlc.table, CBPY_VLC_BITS, 1);
-                if(cbpy<0){
-                    av_log(s->avctx, AV_LOG_ERROR, "cbpy corrupted at %d %d\n", s->mb_x, s->mb_y);
-                    return -1;
-                }
-
-                s->cbp_table[xy]|= cbpy<<2;
-                s->current_picture.mb_type[xy] |= ac_pred*MB_TYPE_ACPRED;
-            }else{ /* P || S_TYPE */
-                if(IS_INTRA(s->current_picture.mb_type[xy])){
-                    int dir=0,i;
-                    int ac_pred = get_bits1(&s->gb);
-                    int cbpy = get_vlc2(&s->gb, cbpy_vlc.table, CBPY_VLC_BITS, 1);
-
-                    if(cbpy<0){
-                        av_log(s->avctx, AV_LOG_ERROR, "I cbpy corrupted at %d %d\n", s->mb_x, s->mb_y);
-                        return -1;
-                    }
-
-                    if(s->cbp_table[xy] & 8) {
-                        ff_set_qscale(s, s->qscale + quant_tab[get_bits(&s->gb, 2)]);
-                    }
-                    s->current_picture.qscale_table[xy]= s->qscale;
-
-                    for(i=0; i<6; i++){
-                        int dc_pred_dir;
-                        int dc= mpeg4_decode_dc(s, i, &dc_pred_dir);
-                        if(dc < 0){
-                            av_log(s->avctx, AV_LOG_ERROR, "DC corrupted at %d %d\n", s->mb_x, s->mb_y);
-                            return -1;
-                        }
-                        dir<<=1;
-                        if(dc_pred_dir) dir|=1;
-                    }
-                    s->cbp_table[xy]&= 3; //remove dquant
-                    s->cbp_table[xy]|= cbpy<<2;
-                    s->current_picture.mb_type[xy] |= ac_pred*MB_TYPE_ACPRED;
-                    s->pred_dir_table[xy]= dir;
-                }else if(IS_SKIP(s->current_picture.mb_type[xy])){
-                    s->current_picture.qscale_table[xy]= s->qscale;
-                    s->cbp_table[xy]= 0;
-                }else{
-                    int cbpy = get_vlc2(&s->gb, cbpy_vlc.table, CBPY_VLC_BITS, 1);
-
-                    if(cbpy<0){
-                        av_log(s->avctx, AV_LOG_ERROR, "P cbpy corrupted at %d %d\n", s->mb_x, s->mb_y);
-                        return -1;
-                    }
-
-                    if(s->cbp_table[xy] & 8) {
-                        ff_set_qscale(s, s->qscale + quant_tab[get_bits(&s->gb, 2)]);
-                    }
-                    s->current_picture.qscale_table[xy]= s->qscale;
-
-                    s->cbp_table[xy]&= 3; //remove dquant
-                    s->cbp_table[xy]|= (cbpy^0xf)<<2;
-                }
-            }
-        }
-        if(mb_num >= mb_count) return 0;
-        s->mb_x= 0;
-    }
-    return 0;
-}
-
-/**
- * decodes the first & second partition
- * @return <0 if error (and sets error type in the error_status_table)
- */
-int ff_mpeg4_decode_partitions(MpegEncContext *s)
-{
-    int mb_num;
-    const int part_a_error= s->pict_type==I_TYPE ? (DC_ERROR|MV_ERROR) : MV_ERROR;
-    const int part_a_end  = s->pict_type==I_TYPE ? (DC_END  |MV_END)   : MV_END;
-
-    mb_num= mpeg4_decode_partition_a(s);
-    if(mb_num<0){
-        ff_er_add_slice(s, s->resync_mb_x, s->resync_mb_y, s->mb_x, s->mb_y, part_a_error);
-        return -1;
-    }
-
-    if(s->resync_mb_x + s->resync_mb_y*s->mb_width + mb_num > s->mb_num){
-        av_log(s->avctx, AV_LOG_ERROR, "slice below monitor ...\n");
-        ff_er_add_slice(s, s->resync_mb_x, s->resync_mb_y, s->mb_x, s->mb_y, part_a_error);
-        return -1;
-    }
-
-    s->mb_num_left= mb_num;
-
-    if(s->pict_type==I_TYPE){
-        while(show_bits(&s->gb, 9) == 1)
-            skip_bits(&s->gb, 9);
-        if(get_bits_long(&s->gb, 19)!=DC_MARKER){
-            av_log(s->avctx, AV_LOG_ERROR, "marker missing after first I partition at %d %d\n", s->mb_x, s->mb_y);
-            return -1;
-        }
-    }else{
-        while(show_bits(&s->gb, 10) == 1)
-            skip_bits(&s->gb, 10);
-        if(get_bits(&s->gb, 17)!=MOTION_MARKER){
-            av_log(s->avctx, AV_LOG_ERROR, "marker missing after first P partition at %d %d\n", s->mb_x, s->mb_y);
-            return -1;
-        }
-    }
-    ff_er_add_slice(s, s->resync_mb_x, s->resync_mb_y, s->mb_x-1, s->mb_y, part_a_end);
-
-    if( mpeg4_decode_partition_b(s, mb_num) < 0){
-        if(s->pict_type==P_TYPE)
-            ff_er_add_slice(s, s->resync_mb_x, s->resync_mb_y, s->mb_x, s->mb_y, DC_ERROR);
-        return -1;
-    }else{
-        if(s->pict_type==P_TYPE)
-            ff_er_add_slice(s, s->resync_mb_x, s->resync_mb_y, s->mb_x-1, s->mb_y, DC_END);
-    }
-
-    return 0;
-}
-
-/**
- * decode partition C of one MB.
- * @return <0 if an error occured
- */
-static int mpeg4_decode_partitioned_mb(MpegEncContext *s, DCTELEM block[6][64])
-{
-    int cbp, mb_type;
-    const int xy= s->mb_x + s->mb_y*s->mb_stride;
-
-    mb_type= s->current_picture.mb_type[xy];
-    cbp = s->cbp_table[xy];
-
-    s->use_intra_dc_vlc= s->qscale < s->intra_dc_threshold;
-
-    if(s->current_picture.qscale_table[xy] != s->qscale){
-        ff_set_qscale(s, s->current_picture.qscale_table[xy] );
-    }
-
-    if (s->pict_type == P_TYPE || s->pict_type==S_TYPE) {
-        int i;
-        for(i=0; i<4; i++){
-            s->mv[0][i][0] = s->current_picture.motion_val[0][ s->block_index[i] ][0];
-            s->mv[0][i][1] = s->current_picture.motion_val[0][ s->block_index[i] ][1];
-        }
-        s->mb_intra = IS_INTRA(mb_type);
-
-        if (IS_SKIP(mb_type)) {
-            /* skip mb */
-            for(i=0;i<6;i++)
-                s->block_last_index[i] = -1;
-            s->mv_dir = MV_DIR_FORWARD;
-            s->mv_type = MV_TYPE_16X16;
-            if(s->pict_type==S_TYPE && s->vol_sprite_usage==GMC_SPRITE){
-                s->mcsel=1;
-                s->mb_skipped = 0;
-            }else{
-                s->mcsel=0;
-                s->mb_skipped = 1;
-            }
-        }else if(s->mb_intra){
-            s->ac_pred = IS_ACPRED(s->current_picture.mb_type[xy]);
-        }else if(!s->mb_intra){
-//            s->mcsel= 0; //FIXME do we need to init that
-
-            s->mv_dir = MV_DIR_FORWARD;
-            if (IS_8X8(mb_type)) {
-                s->mv_type = MV_TYPE_8X8;
-            } else {
-                s->mv_type = MV_TYPE_16X16;
-            }
-        }
-    } else { /* I-Frame */
-        s->mb_intra = 1;
-        s->ac_pred = IS_ACPRED(s->current_picture.mb_type[xy]);
-    }
-
-    if (!IS_SKIP(mb_type)) {
-        int i;
-        s->dsp.clear_blocks(s->block[0]);
-        /* decode each block */
-        for (i = 0; i < 6; i++) {
-            if(mpeg4_decode_block(s, block[i], i, cbp&32, s->mb_intra, s->rvlc) < 0){
-                av_log(s->avctx, AV_LOG_ERROR, "texture corrupted at %d %d %d\n", s->mb_x, s->mb_y, s->mb_intra);
-                return -1;
-            }
-            cbp+=cbp;
-        }
-    }
-
-    /* per-MB end of slice check */
-
-    if(--s->mb_num_left <= 0){
-//printf("%06X %d\n", show_bits(&s->gb, 24), s->gb.size_in_bits - get_bits_count(&s->gb));
-        if(mpeg4_is_resync(s))
-            return SLICE_END;
-        else
-            return SLICE_NOEND;
-    }else{
-        if(mpeg4_is_resync(s)){
-            const int delta= s->mb_x + 1 == s->mb_width ? 2 : 1;
-            if(s->cbp_table[xy+delta])
-                return SLICE_END;
-        }
-        return SLICE_OK;
-    }
-}
-
-/**
- * read the next MVs for OBMC. yes this is a ugly hack, feel free to send a patch :)
- */
-static void preview_obmc(MpegEncContext *s){
-    GetBitContext gb= s->gb;
-
-    int cbpc, i, pred_x, pred_y, mx, my;
-    int16_t *mot_val;
-    const int xy= s->mb_x + 1 + s->mb_y * s->mb_stride;
-    const int stride= s->b8_stride*2;
-
-    for(i=0; i<4; i++)
-        s->block_index[i]+= 2;
-    for(i=4; i<6; i++)
-        s->block_index[i]+= 1;
-    s->mb_x++;
-
-    assert(s->pict_type == P_TYPE);
-
-    do{
-        if (get_bits1(&s->gb)) {
-            /* skip mb */
-            mot_val = s->current_picture.motion_val[0][ s->block_index[0] ];
-            mot_val[0       ]= mot_val[2       ]=
-            mot_val[0+stride]= mot_val[2+stride]= 0;
-            mot_val[1       ]= mot_val[3       ]=
-            mot_val[1+stride]= mot_val[3+stride]= 0;
-
-            s->current_picture.mb_type[xy]= MB_TYPE_SKIP | MB_TYPE_16x16 | MB_TYPE_L0;
-            goto end;
-        }
-        cbpc = get_vlc2(&s->gb, inter_MCBPC_vlc.table, INTER_MCBPC_VLC_BITS, 2);
-    }while(cbpc == 20);
-
-    if(cbpc & 4){
-        s->current_picture.mb_type[xy]= MB_TYPE_INTRA;
-    }else{
-        get_vlc2(&s->gb, cbpy_vlc.table, CBPY_VLC_BITS, 1);
-        if (cbpc & 8) {
-            if(s->modified_quant){
-                if(get_bits1(&s->gb)) skip_bits(&s->gb, 1);
-                else                  skip_bits(&s->gb, 5);
-            }else
-                skip_bits(&s->gb, 2);
-        }
-
-        if ((cbpc & 16) == 0) {
-                s->current_picture.mb_type[xy]= MB_TYPE_16x16 | MB_TYPE_L0;
-                /* 16x16 motion prediction */
-                mot_val= h263_pred_motion(s, 0, 0, &pred_x, &pred_y);
-                if (s->umvplus)
-                   mx = h263p_decode_umotion(s, pred_x);
-                else
-                   mx = h263_decode_motion(s, pred_x, 1);
-
-                if (s->umvplus)
-                   my = h263p_decode_umotion(s, pred_y);
-                else
-                   my = h263_decode_motion(s, pred_y, 1);
-
-                mot_val[0       ]= mot_val[2       ]=
-                mot_val[0+stride]= mot_val[2+stride]= mx;
-                mot_val[1       ]= mot_val[3       ]=
-                mot_val[1+stride]= mot_val[3+stride]= my;
-        } else {
-            s->current_picture.mb_type[xy]= MB_TYPE_8x8 | MB_TYPE_L0;
-            for(i=0;i<4;i++) {
-                mot_val = h263_pred_motion(s, i, 0, &pred_x, &pred_y);
-                if (s->umvplus)
-                  mx = h263p_decode_umotion(s, pred_x);
-                else
-                  mx = h263_decode_motion(s, pred_x, 1);
-
-                if (s->umvplus)
-                  my = h263p_decode_umotion(s, pred_y);
-                else
-                  my = h263_decode_motion(s, pred_y, 1);
-                if (s->umvplus && (mx - pred_x) == 1 && (my - pred_y) == 1)
-                  skip_bits1(&s->gb); /* Bit stuffing to prevent PSC */
-                mot_val[0] = mx;
-                mot_val[1] = my;
-            }
-        }
-    }
-end:
-
-    for(i=0; i<4; i++)
-        s->block_index[i]-= 2;
-    for(i=4; i<6; i++)
-        s->block_index[i]-= 1;
-    s->mb_x--;
-
-    s->gb= gb;
-}
-
-static void h263_decode_dquant(MpegEncContext *s){
-    static const int8_t quant_tab[4] = { -1, -2, 1, 2 };
-
-    if(s->modified_quant){
-        if(get_bits1(&s->gb))
-            s->qscale= modified_quant_tab[get_bits1(&s->gb)][ s->qscale ];
-        else
-            s->qscale= get_bits(&s->gb, 5);
-    }else
-        s->qscale += quant_tab[get_bits(&s->gb, 2)];
-    ff_set_qscale(s, s->qscale);
-}
-
-int ff_h263_decode_mb(MpegEncContext *s,
-                      DCTELEM block[6][64])
-{
-    int cbpc, cbpy, i, cbp, pred_x, pred_y, mx, my, dquant;
-    int16_t *mot_val;
-    const int xy= s->mb_x + s->mb_y * s->mb_stride;
-
-    assert(!s->h263_pred);
-
-    if (s->pict_type == P_TYPE) {
-        do{
-            if (get_bits1(&s->gb)) {
-                /* skip mb */
-                s->mb_intra = 0;
-                for(i=0;i<6;i++)
-                    s->block_last_index[i] = -1;
-                s->mv_dir = MV_DIR_FORWARD;
-                s->mv_type = MV_TYPE_16X16;
-                s->current_picture.mb_type[xy]= MB_TYPE_SKIP | MB_TYPE_16x16 | MB_TYPE_L0;
-                s->mv[0][0][0] = 0;
-                s->mv[0][0][1] = 0;
-                s->mb_skipped = !(s->obmc | s->loop_filter);
-                goto end;
-            }
-            cbpc = get_vlc2(&s->gb, inter_MCBPC_vlc.table, INTER_MCBPC_VLC_BITS, 2);
-            //fprintf(stderr, "\tCBPC: %d", cbpc);
-            if (cbpc < 0){
-                av_log(s->avctx, AV_LOG_ERROR, "cbpc damaged at %d %d\n", s->mb_x, s->mb_y);
-                return -1;
-            }
-        }while(cbpc == 20);
-
-        s->dsp.clear_blocks(s->block[0]);
-
-        dquant = cbpc & 8;
-        s->mb_intra = ((cbpc & 4) != 0);
-        if (s->mb_intra) goto intra;
-
-        cbpy = get_vlc2(&s->gb, cbpy_vlc.table, CBPY_VLC_BITS, 1);
-
-        if(s->alt_inter_vlc==0 || (cbpc & 3)!=3)
-            cbpy ^= 0xF;
-
-        cbp = (cbpc & 3) | (cbpy << 2);
-        if (dquant) {
-            h263_decode_dquant(s);
-        }
-
-        s->mv_dir = MV_DIR_FORWARD;
-        if ((cbpc & 16) == 0) {
-            s->current_picture.mb_type[xy]= MB_TYPE_16x16 | MB_TYPE_L0;
-            /* 16x16 motion prediction */
-            s->mv_type = MV_TYPE_16X16;
-            h263_pred_motion(s, 0, 0, &pred_x, &pred_y);
-            if (s->umvplus)
-               mx = h263p_decode_umotion(s, pred_x);
-            else
-               mx = h263_decode_motion(s, pred_x, 1);
-
-            if (mx >= 0xffff)
-                return -1;
-
-            if (s->umvplus)
-               my = h263p_decode_umotion(s, pred_y);
-            else
-               my = h263_decode_motion(s, pred_y, 1);
-
-            if (my >= 0xffff)
-                return -1;
-            s->mv[0][0][0] = mx;
-            s->mv[0][0][1] = my;
-
-            if (s->umvplus && (mx - pred_x) == 1 && (my - pred_y) == 1)
-               skip_bits1(&s->gb); /* Bit stuffing to prevent PSC */
-        } else {
-            s->current_picture.mb_type[xy]= MB_TYPE_8x8 | MB_TYPE_L0;
-            s->mv_type = MV_TYPE_8X8;
-            for(i=0;i<4;i++) {
-                mot_val = h263_pred_motion(s, i, 0, &pred_x, &pred_y);
-                if (s->umvplus)
-                  mx = h263p_decode_umotion(s, pred_x);
-                else
-                  mx = h263_decode_motion(s, pred_x, 1);
-                if (mx >= 0xffff)
-                    return -1;
-
-                if (s->umvplus)
-                  my = h263p_decode_umotion(s, pred_y);
-                else
-                  my = h263_decode_motion(s, pred_y, 1);
-                if (my >= 0xffff)
-                    return -1;
-                s->mv[0][i][0] = mx;
-                s->mv[0][i][1] = my;
-                if (s->umvplus && (mx - pred_x) == 1 && (my - pred_y) == 1)
-                  skip_bits1(&s->gb); /* Bit stuffing to prevent PSC */
-                mot_val[0] = mx;
-                mot_val[1] = my;
-            }
-        }
-
-        /* decode each block */
-        for (i = 0; i < 6; i++) {
-            if (h263_decode_block(s, block[i], i, cbp&32) < 0)
-                return -1;
-            cbp+=cbp;
-        }
-
-        if(s->obmc){
-            if(s->pict_type == P_TYPE && s->mb_x+1<s->mb_width && s->mb_num_left != 1)
-                preview_obmc(s);
-        }
-    } else if(s->pict_type==B_TYPE) {
-        int mb_type;
-        const int stride= s->b8_stride;
-        int16_t *mot_val0 = s->current_picture.motion_val[0][ 2*(s->mb_x + s->mb_y*stride) ];
-        int16_t *mot_val1 = s->current_picture.motion_val[1][ 2*(s->mb_x + s->mb_y*stride) ];
-//        const int mv_xy= s->mb_x + 1 + s->mb_y * s->mb_stride;
-
-        //FIXME ugly
-        mot_val0[0       ]= mot_val0[2       ]= mot_val0[0+2*stride]= mot_val0[2+2*stride]=
-        mot_val0[1       ]= mot_val0[3       ]= mot_val0[1+2*stride]= mot_val0[3+2*stride]=
-        mot_val1[0       ]= mot_val1[2       ]= mot_val1[0+2*stride]= mot_val1[2+2*stride]=
-        mot_val1[1       ]= mot_val1[3       ]= mot_val1[1+2*stride]= mot_val1[3+2*stride]= 0;
-
-        do{
-            mb_type= get_vlc2(&s->gb, h263_mbtype_b_vlc.table, H263_MBTYPE_B_VLC_BITS, 2);
-            if (mb_type < 0){
-                av_log(s->avctx, AV_LOG_ERROR, "b mb_type damaged at %d %d\n", s->mb_x, s->mb_y);
-                return -1;
-            }
-
-            mb_type= h263_mb_type_b_map[ mb_type ];
-        }while(!mb_type);
-
-        s->mb_intra = IS_INTRA(mb_type);
-        if(HAS_CBP(mb_type)){
-            s->dsp.clear_blocks(s->block[0]);
-            cbpc = get_vlc2(&s->gb, cbpc_b_vlc.table, CBPC_B_VLC_BITS, 1);
-            if(s->mb_intra){
-                dquant = IS_QUANT(mb_type);
-                goto intra;
-            }
-
-            cbpy = get_vlc2(&s->gb, cbpy_vlc.table, CBPY_VLC_BITS, 1);
-
-            if (cbpy < 0){
-                av_log(s->avctx, AV_LOG_ERROR, "b cbpy damaged at %d %d\n", s->mb_x, s->mb_y);
-                return -1;
-            }
-
-            if(s->alt_inter_vlc==0 || (cbpc & 3)!=3)
-                cbpy ^= 0xF;
-
-            cbp = (cbpc & 3) | (cbpy << 2);
-        }else
-            cbp=0;
-
-        assert(!s->mb_intra);
-
-        if(IS_QUANT(mb_type)){
-            h263_decode_dquant(s);
-        }
-
-        if(IS_DIRECT(mb_type)){
-            s->mv_dir = MV_DIR_FORWARD | MV_DIR_BACKWARD | MV_DIRECT;
-            mb_type |= ff_mpeg4_set_direct_mv(s, 0, 0);
-        }else{
-            s->mv_dir = 0;
-            s->mv_type= MV_TYPE_16X16;
-//FIXME UMV
-
-            if(USES_LIST(mb_type, 0)){
-                int16_t *mot_val= h263_pred_motion(s, 0, 0, &mx, &my);
-                s->mv_dir = MV_DIR_FORWARD;
-
-                mx = h263_decode_motion(s, mx, 1);
-                my = h263_decode_motion(s, my, 1);
-
-                s->mv[0][0][0] = mx;
-                s->mv[0][0][1] = my;
-                mot_val[0       ]= mot_val[2       ]= mot_val[0+2*stride]= mot_val[2+2*stride]= mx;
-                mot_val[1       ]= mot_val[3       ]= mot_val[1+2*stride]= mot_val[3+2*stride]= my;
-            }
-
-            if(USES_LIST(mb_type, 1)){
-                int16_t *mot_val= h263_pred_motion(s, 0, 1, &mx, &my);
-                s->mv_dir |= MV_DIR_BACKWARD;
-
-                mx = h263_decode_motion(s, mx, 1);
-                my = h263_decode_motion(s, my, 1);
-
-                s->mv[1][0][0] = mx;
-                s->mv[1][0][1] = my;
-                mot_val[0       ]= mot_val[2       ]= mot_val[0+2*stride]= mot_val[2+2*stride]= mx;
-                mot_val[1       ]= mot_val[3       ]= mot_val[1+2*stride]= mot_val[3+2*stride]= my;
-            }
-        }
-
-        s->current_picture.mb_type[xy]= mb_type;
-
-        /* decode each block */
-        for (i = 0; i < 6; i++) {
-            if (h263_decode_block(s, block[i], i, cbp&32) < 0)
-                return -1;
-            cbp+=cbp;
-        }
-    } else { /* I-Frame */
-        do{
-            cbpc = get_vlc2(&s->gb, intra_MCBPC_vlc.table, INTRA_MCBPC_VLC_BITS, 2);
-            if (cbpc < 0){
-                av_log(s->avctx, AV_LOG_ERROR, "I cbpc damaged at %d %d\n", s->mb_x, s->mb_y);
-                return -1;
-            }
-        }while(cbpc == 8);
-
-        s->dsp.clear_blocks(s->block[0]);
-
-        dquant = cbpc & 4;
-        s->mb_intra = 1;
-intra:
-        s->current_picture.mb_type[xy]= MB_TYPE_INTRA;
-        if (s->h263_aic) {
-            s->ac_pred = get_bits1(&s->gb);
-            if(s->ac_pred){
-                s->current_picture.mb_type[xy]= MB_TYPE_INTRA | MB_TYPE_ACPRED;
-
-                s->h263_aic_dir = get_bits1(&s->gb);
-            }
-        }else
-            s->ac_pred = 0;
-
-        cbpy = get_vlc2(&s->gb, cbpy_vlc.table, CBPY_VLC_BITS, 1);
-        if(cbpy<0){
-            av_log(s->avctx, AV_LOG_ERROR, "I cbpy damaged at %d %d\n", s->mb_x, s->mb_y);
-            return -1;
-        }
-        cbp = (cbpc & 3) | (cbpy << 2);
-        if (dquant) {
-            h263_decode_dquant(s);
-        }
-
-        /* decode each block */
-        for (i = 0; i < 6; i++) {
-            if (h263_decode_block(s, block[i], i, cbp&32) < 0)
-                return -1;
-            cbp+=cbp;
-        }
-    }
-end:
-
-        /* per-MB end of slice check */
-    {
-        int v= show_bits(&s->gb, 16);
-
-        if(get_bits_count(&s->gb) + 16 > s->gb.size_in_bits){
-            v>>= get_bits_count(&s->gb) + 16 - s->gb.size_in_bits;
-        }
-
-        if(v==0)
-            return SLICE_END;
-    }
-
-    return SLICE_OK;
-}
-
-int ff_mpeg4_decode_mb(MpegEncContext *s,
-                      DCTELEM block[6][64])
-{
-    int cbpc, cbpy, i, cbp, pred_x, pred_y, mx, my, dquant;
-    int16_t *mot_val;
-    static int8_t quant_tab[4] = { -1, -2, 1, 2 };
-    const int xy= s->mb_x + s->mb_y * s->mb_stride;
-
-    assert(s->h263_pred);
-
-    if (s->pict_type == P_TYPE || s->pict_type==S_TYPE) {
-        do{
-            if (get_bits1(&s->gb)) {
-                /* skip mb */
-                s->mb_intra = 0;
-                for(i=0;i<6;i++)
-                    s->block_last_index[i] = -1;
-                s->mv_dir = MV_DIR_FORWARD;
-                s->mv_type = MV_TYPE_16X16;
-                if(s->pict_type==S_TYPE && s->vol_sprite_usage==GMC_SPRITE){
-                    s->current_picture.mb_type[xy]= MB_TYPE_SKIP | MB_TYPE_GMC | MB_TYPE_16x16 | MB_TYPE_L0;
-                    s->mcsel=1;
-                    s->mv[0][0][0]= get_amv(s, 0);
-                    s->mv[0][0][1]= get_amv(s, 1);
-
-                    s->mb_skipped = 0;
-                }else{
-                    s->current_picture.mb_type[xy]= MB_TYPE_SKIP | MB_TYPE_16x16 | MB_TYPE_L0;
-                    s->mcsel=0;
-                    s->mv[0][0][0] = 0;
-                    s->mv[0][0][1] = 0;
-                    s->mb_skipped = 1;
-                }
-                goto end;
-            }
-            cbpc = get_vlc2(&s->gb, inter_MCBPC_vlc.table, INTER_MCBPC_VLC_BITS, 2);
-            //fprintf(stderr, "\tCBPC: %d", cbpc);
-            if (cbpc < 0){
-                av_log(s->avctx, AV_LOG_ERROR, "cbpc damaged at %d %d\n", s->mb_x, s->mb_y);
-                return -1;
-            }
-        }while(cbpc == 20);
-
-        s->dsp.clear_blocks(s->block[0]);
-        dquant = cbpc & 8;
-        s->mb_intra = ((cbpc & 4) != 0);
-        if (s->mb_intra) goto intra;
-
-        if(s->pict_type==S_TYPE && s->vol_sprite_usage==GMC_SPRITE && (cbpc & 16) == 0)
-            s->mcsel= get_bits1(&s->gb);
-        else s->mcsel= 0;
-        cbpy = get_vlc2(&s->gb, cbpy_vlc.table, CBPY_VLC_BITS, 1) ^ 0x0F;
-
-        cbp = (cbpc & 3) | (cbpy << 2);
-        if (dquant) {
-            ff_set_qscale(s, s->qscale + quant_tab[get_bits(&s->gb, 2)]);
-        }
-        if((!s->progressive_sequence) && (cbp || (s->workaround_bugs&FF_BUG_XVID_ILACE)))
-            s->interlaced_dct= get_bits1(&s->gb);
-
-        s->mv_dir = MV_DIR_FORWARD;
-        if ((cbpc & 16) == 0) {
-            if(s->mcsel){
-                s->current_picture.mb_type[xy]= MB_TYPE_GMC | MB_TYPE_16x16 | MB_TYPE_L0;
-                /* 16x16 global motion prediction */
-                s->mv_type = MV_TYPE_16X16;
-                mx= get_amv(s, 0);
-                my= get_amv(s, 1);
-                s->mv[0][0][0] = mx;
-                s->mv[0][0][1] = my;
-            }else if((!s->progressive_sequence) && get_bits1(&s->gb)){
-                s->current_picture.mb_type[xy]= MB_TYPE_16x8 | MB_TYPE_L0 | MB_TYPE_INTERLACED;
-                /* 16x8 field motion prediction */
-                s->mv_type= MV_TYPE_FIELD;
-
-                s->field_select[0][0]= get_bits1(&s->gb);
-                s->field_select[0][1]= get_bits1(&s->gb);
-
-                h263_pred_motion(s, 0, 0, &pred_x, &pred_y);
-
-                for(i=0; i<2; i++){
-                    mx = h263_decode_motion(s, pred_x, s->f_code);
-                    if (mx >= 0xffff)
-                        return -1;
-
-                    my = h263_decode_motion(s, pred_y/2, s->f_code);
-                    if (my >= 0xffff)
-                        return -1;
-
-                    s->mv[0][i][0] = mx;
-                    s->mv[0][i][1] = my;
-                }
-            }else{
-                s->current_picture.mb_type[xy]= MB_TYPE_16x16 | MB_TYPE_L0;
-                /* 16x16 motion prediction */
-                s->mv_type = MV_TYPE_16X16;
-                h263_pred_motion(s, 0, 0, &pred_x, &pred_y);
-                mx = h263_decode_motion(s, pred_x, s->f_code);
-
-                if (mx >= 0xffff)
-                    return -1;
-
-                my = h263_decode_motion(s, pred_y, s->f_code);
-
-                if (my >= 0xffff)
-                    return -1;
-                s->mv[0][0][0] = mx;
-                s->mv[0][0][1] = my;
-            }
-        } else {
-            s->current_picture.mb_type[xy]= MB_TYPE_8x8 | MB_TYPE_L0;
-            s->mv_type = MV_TYPE_8X8;
-            for(i=0;i<4;i++) {
-                mot_val = h263_pred_motion(s, i, 0, &pred_x, &pred_y);
-                mx = h263_decode_motion(s, pred_x, s->f_code);
-                if (mx >= 0xffff)
-                    return -1;
-
-                my = h263_decode_motion(s, pred_y, s->f_code);
-                if (my >= 0xffff)
-                    return -1;
-                s->mv[0][i][0] = mx;
-                s->mv[0][i][1] = my;
-                mot_val[0] = mx;
-                mot_val[1] = my;
-            }
-        }
-    } else if(s->pict_type==B_TYPE) {
-        int modb1; // first bit of modb
-        int modb2; // second bit of modb
-        int mb_type;
-
-        s->mb_intra = 0; //B-frames never contain intra blocks
-        s->mcsel=0;      //     ...               true gmc blocks
-
-        if(s->mb_x==0){
-            for(i=0; i<2; i++){
-                s->last_mv[i][0][0]=
-                s->last_mv[i][0][1]=
-                s->last_mv[i][1][0]=
-                s->last_mv[i][1][1]= 0;
-            }
-        }
-
-        /* if we skipped it in the future P Frame than skip it now too */
-        s->mb_skipped= s->next_picture.mbskip_table[s->mb_y * s->mb_stride + s->mb_x]; // Note, skiptab=0 if last was GMC
-
-        if(s->mb_skipped){
-                /* skip mb */
-            for(i=0;i<6;i++)
-                s->block_last_index[i] = -1;
-
-            s->mv_dir = MV_DIR_FORWARD;
-            s->mv_type = MV_TYPE_16X16;
-            s->mv[0][0][0] = 0;
-            s->mv[0][0][1] = 0;
-            s->mv[1][0][0] = 0;
-            s->mv[1][0][1] = 0;
-            s->current_picture.mb_type[xy]= MB_TYPE_SKIP | MB_TYPE_16x16 | MB_TYPE_L0;
-            goto end;
-        }
-
-        modb1= get_bits1(&s->gb);
-        if(modb1){
-            mb_type= MB_TYPE_DIRECT2 | MB_TYPE_SKIP | MB_TYPE_L0L1; //like MB_TYPE_B_DIRECT but no vectors coded
-            cbp=0;
-        }else{
-            modb2= get_bits1(&s->gb);
-            mb_type= get_vlc2(&s->gb, mb_type_b_vlc.table, MB_TYPE_B_VLC_BITS, 1);
-            if(mb_type<0){
-                av_log(s->avctx, AV_LOG_ERROR, "illegal MB_type\n");
-                return -1;
-            }
-            mb_type= mb_type_b_map[ mb_type ];
-            if(modb2) cbp= 0;
-            else{
-                s->dsp.clear_blocks(s->block[0]);
-                cbp= get_bits(&s->gb, 6);
-            }
-
-            if ((!IS_DIRECT(mb_type)) && cbp) {
-                if(get_bits1(&s->gb)){
-                    ff_set_qscale(s, s->qscale + get_bits1(&s->gb)*4 - 2);
-                }
-            }
-
-            if(!s->progressive_sequence){
-                if(cbp)
-                    s->interlaced_dct= get_bits1(&s->gb);
-
-                if(!IS_DIRECT(mb_type) && get_bits1(&s->gb)){
-                    mb_type |= MB_TYPE_16x8 | MB_TYPE_INTERLACED;
-                    mb_type &= ~MB_TYPE_16x16;
-
-                    if(USES_LIST(mb_type, 0)){
-                        s->field_select[0][0]= get_bits1(&s->gb);
-                        s->field_select[0][1]= get_bits1(&s->gb);
-                    }
-                    if(USES_LIST(mb_type, 1)){
-                        s->field_select[1][0]= get_bits1(&s->gb);
-                        s->field_select[1][1]= get_bits1(&s->gb);
-                    }
-                }
-            }
-
-            s->mv_dir = 0;
-            if((mb_type & (MB_TYPE_DIRECT2|MB_TYPE_INTERLACED)) == 0){
-                s->mv_type= MV_TYPE_16X16;
-
-                if(USES_LIST(mb_type, 0)){
-                    s->mv_dir = MV_DIR_FORWARD;
-
-                    mx = h263_decode_motion(s, s->last_mv[0][0][0], s->f_code);
-                    my = h263_decode_motion(s, s->last_mv[0][0][1], s->f_code);
-                    s->last_mv[0][1][0]= s->last_mv[0][0][0]= s->mv[0][0][0] = mx;
-                    s->last_mv[0][1][1]= s->last_mv[0][0][1]= s->mv[0][0][1] = my;
-                }
-
-                if(USES_LIST(mb_type, 1)){
-                    s->mv_dir |= MV_DIR_BACKWARD;
-
-                    mx = h263_decode_motion(s, s->last_mv[1][0][0], s->b_code);
-                    my = h263_decode_motion(s, s->last_mv[1][0][1], s->b_code);
-                    s->last_mv[1][1][0]= s->last_mv[1][0][0]= s->mv[1][0][0] = mx;
-                    s->last_mv[1][1][1]= s->last_mv[1][0][1]= s->mv[1][0][1] = my;
-                }
-            }else if(!IS_DIRECT(mb_type)){
-                s->mv_type= MV_TYPE_FIELD;
-
-                if(USES_LIST(mb_type, 0)){
-                    s->mv_dir = MV_DIR_FORWARD;
-
-                    for(i=0; i<2; i++){
-                        mx = h263_decode_motion(s, s->last_mv[0][i][0]  , s->f_code);
-                        my = h263_decode_motion(s, s->last_mv[0][i][1]/2, s->f_code);
-                        s->last_mv[0][i][0]=  s->mv[0][i][0] = mx;
-                        s->last_mv[0][i][1]= (s->mv[0][i][1] = my)*2;
-                    }
-                }
-
-                if(USES_LIST(mb_type, 1)){
-                    s->mv_dir |= MV_DIR_BACKWARD;
-
-                    for(i=0; i<2; i++){
-                        mx = h263_decode_motion(s, s->last_mv[1][i][0]  , s->b_code);
-                        my = h263_decode_motion(s, s->last_mv[1][i][1]/2, s->b_code);
-                        s->last_mv[1][i][0]=  s->mv[1][i][0] = mx;
-                        s->last_mv[1][i][1]= (s->mv[1][i][1] = my)*2;
-                    }
-                }
-            }
-        }
-
-        if(IS_DIRECT(mb_type)){
-            if(IS_SKIP(mb_type))
-                mx=my=0;
-            else{
-                mx = h263_decode_motion(s, 0, 1);
-                my = h263_decode_motion(s, 0, 1);
-            }
-
-            s->mv_dir = MV_DIR_FORWARD | MV_DIR_BACKWARD | MV_DIRECT;
-            mb_type |= ff_mpeg4_set_direct_mv(s, mx, my);
-        }
-        s->current_picture.mb_type[xy]= mb_type;
-    } else { /* I-Frame */
-        do{
-            cbpc = get_vlc2(&s->gb, intra_MCBPC_vlc.table, INTRA_MCBPC_VLC_BITS, 2);
-            if (cbpc < 0){
-                av_log(s->avctx, AV_LOG_ERROR, "I cbpc damaged at %d %d\n", s->mb_x, s->mb_y);
-                return -1;
-            }
-        }while(cbpc == 8);
-
-        dquant = cbpc & 4;
-        s->mb_intra = 1;
-intra:
-        s->ac_pred = get_bits1(&s->gb);
-        if(s->ac_pred)
-            s->current_picture.mb_type[xy]= MB_TYPE_INTRA | MB_TYPE_ACPRED;
-        else
-            s->current_picture.mb_type[xy]= MB_TYPE_INTRA;
-
-        cbpy = get_vlc2(&s->gb, cbpy_vlc.table, CBPY_VLC_BITS, 1);
-        if(cbpy<0){
-            av_log(s->avctx, AV_LOG_ERROR, "I cbpy damaged at %d %d\n", s->mb_x, s->mb_y);
-            return -1;
-        }
-        cbp = (cbpc & 3) | (cbpy << 2);
-
-        s->use_intra_dc_vlc= s->qscale < s->intra_dc_threshold;
-
-        if (dquant) {
-            ff_set_qscale(s, s->qscale + quant_tab[get_bits(&s->gb, 2)]);
-        }
-
-        if(!s->progressive_sequence)
-            s->interlaced_dct= get_bits1(&s->gb);
-
-        s->dsp.clear_blocks(s->block[0]);
-        /* decode each block */
-        for (i = 0; i < 6; i++) {
-            if (mpeg4_decode_block(s, block[i], i, cbp&32, 1, 0) < 0)
-                return -1;
-            cbp+=cbp;
-        }
-        goto end;
-    }
-
-    /* decode each block */
-    for (i = 0; i < 6; i++) {
-        if (mpeg4_decode_block(s, block[i], i, cbp&32, 0, 0) < 0)
-            return -1;
-        cbp+=cbp;
-    }
-end:
-
-        /* per-MB end of slice check */
-    if(s->codec_id==CODEC_ID_MPEG4){
-#if 0 //http://standards.iso.org/ittf/PubliclyAvailableStandards/ISO_IEC_14496-4_2004_Conformance_Testing/video_conformance/version_1/simple/ERROR.ZIP/mit025.m4v needs this but its unclear if the mpeg4 standard allows this at all (MN)
-        if(s->pict_type != B_TYPE){
-            while(show_bits(&s->gb, 9 + (s->pict_type == P_TYPE)) == 1)
-                skip_bits(&s->gb, 9 + (s->pict_type == P_TYPE));
-        }
-#endif
-        if(mpeg4_is_resync(s)){
-            const int delta= s->mb_x + 1 == s->mb_width ? 2 : 1;
-            if(s->pict_type==B_TYPE && s->next_picture.mbskip_table[xy + delta])
-                return SLICE_OK;
-            return SLICE_END;
-        }
-    }
-
-    return SLICE_OK;
-}
-
-static int h263_decode_motion(MpegEncContext * s, int pred, int f_code)
-{
-    int code, val, sign, shift, l;
-    code = get_vlc2(&s->gb, mv_vlc.table, MV_VLC_BITS, 2);
-
-    if (code == 0)
-        return pred;
-    if (code < 0)
-        return 0xffff;
-
-    sign = get_bits1(&s->gb);
-    shift = f_code - 1;
-    val = code;
-    if (shift) {
-        val = (val - 1) << shift;
-        val |= get_bits(&s->gb, shift);
-        val++;
-    }
-    if (sign)
-        val = -val;
-    val += pred;
-
-    /* modulo decoding */
-    if (!s->h263_long_vectors) {
-        l = INT_BIT - 5 - f_code;
-        val = (val<<l)>>l;
-    } else {
-        /* horrible h263 long vector mode */
-        if (pred < -31 && val < -63)
-            val += 64;
-        if (pred > 32 && val > 63)
-            val -= 64;
-
-    }
-    return val;
-}
-
-/* Decodes RVLC of H.263+ UMV */
-static int h263p_decode_umotion(MpegEncContext * s, int pred)
-{
-   int code = 0, sign;
-
-   if (get_bits1(&s->gb)) /* Motion difference = 0 */
-      return pred;
-
-   code = 2 + get_bits1(&s->gb);
-
-   while (get_bits1(&s->gb))
-   {
-      code <<= 1;
-      code += get_bits1(&s->gb);
-   }
-   sign = code & 1;
-   code >>= 1;
-
-   code = (sign) ? (pred - code) : (pred + code);
-#ifdef DEBUG
-   av_log( s->avctx, AV_LOG_DEBUG,"H.263+ UMV Motion = %d\n", code);
-#endif
-   return code;
-
-}
-
-static int h263_decode_block(MpegEncContext * s, DCTELEM * block,
-                             int n, int coded)
-{
-    int code, level, i, j, last, run;
-    RLTable *rl = &rl_inter;
-    const uint8_t *scan_table;
-    GetBitContext gb= s->gb;
-
-    scan_table = s->intra_scantable.permutated;
-    if (s->h263_aic && s->mb_intra) {
-        rl = &rl_intra_aic;
-        i = 0;
-        if (s->ac_pred) {
-            if (s->h263_aic_dir)
-                scan_table = s->intra_v_scantable.permutated; /* left */
-            else
-                scan_table = s->intra_h_scantable.permutated; /* top */
-        }
-    } else if (s->mb_intra) {
-        /* DC coef */
-        if(s->codec_id == CODEC_ID_RV10){
-#ifdef CONFIG_RV10_DECODER
-          if (s->rv10_version == 3 && s->pict_type == I_TYPE) {
-            int component, diff;
-            component = (n <= 3 ? 0 : n - 4 + 1);
-            level = s->last_dc[component];
-            if (s->rv10_first_dc_coded[component]) {
-                diff = rv_decode_dc(s, n);
-                if (diff == 0xffff)
-                    return -1;
-                level += diff;
-                level = level & 0xff; /* handle wrap round */
-                s->last_dc[component] = level;
-            } else {
-                s->rv10_first_dc_coded[component] = 1;
-            }
-          } else {
-                level = get_bits(&s->gb, 8);
-                if (level == 255)
-                    level = 128;
-          }
-#endif
-        }else{
-            level = get_bits(&s->gb, 8);
-            if((level&0x7F) == 0){
-                av_log(s->avctx, AV_LOG_ERROR, "illegal dc %d at %d %d\n", level, s->mb_x, s->mb_y);
-                if(s->error_resilience >= FF_ER_COMPLIANT)
-                    return -1;
-            }
-            if (level == 255)
-                level = 128;
-        }
-        block[0] = level;
-        i = 1;
-    } else {
-        i = 0;
-    }
-    if (!coded) {
-        if (s->mb_intra && s->h263_aic)
-            goto not_coded;
-        s->block_last_index[n] = i - 1;
-        return 0;
-    }
-retry:
-    for(;;) {
-        code = get_vlc2(&s->gb, rl->vlc.table, TEX_VLC_BITS, 2);
-        if (code < 0){
-            av_log(s->avctx, AV_LOG_ERROR, "illegal ac vlc code at %dx%d\n", s->mb_x, s->mb_y);
-            return -1;
-        }
-        if (code == rl->n) {
-            /* escape */
-            if (s->h263_flv > 1) {
-                int is11 = get_bits1(&s->gb);
-                last = get_bits1(&s->gb);
-                run = get_bits(&s->gb, 6);
-                if(is11){
-                    level = get_sbits(&s->gb, 11);
-                } else {
-                    level = get_sbits(&s->gb, 7);
-                }
-            } else {
-                last = get_bits1(&s->gb);
-                run = get_bits(&s->gb, 6);
-                level = (int8_t)get_bits(&s->gb, 8);
-                if(level == -128){
-                    if (s->codec_id == CODEC_ID_RV10) {
-                        /* XXX: should patch encoder too */
-                        level = get_sbits(&s->gb, 12);
-                    }else{
-                        level = get_bits(&s->gb, 5);
-                        level |= get_sbits(&s->gb, 6)<<5;
-                    }
-                }
-            }
-        } else {
-            run = rl->table_run[code];
-            level = rl->table_level[code];
-            last = code >= rl->last;
-            if (get_bits1(&s->gb))
-                level = -level;
-        }
-        i += run;
-        if (i >= 64){
-            if(s->alt_inter_vlc && rl == &rl_inter && !s->mb_intra){
-                //looks like a hack but no, it's the way its supposed to work ...
-                rl = &rl_intra_aic;
-                i = 0;
-                s->gb= gb;
-                memset(block, 0, sizeof(DCTELEM)*64);
-                goto retry;
-            }
-            av_log(s->avctx, AV_LOG_ERROR, "run overflow at %dx%d i:%d\n", s->mb_x, s->mb_y, s->mb_intra);
-            return -1;
-        }
-        j = scan_table[i];
-        block[j] = level;
-        if (last)
-            break;
-        i++;
-    }
-not_coded:
-    if (s->mb_intra && s->h263_aic) {
-        h263_pred_acdc(s, block, n);
-        i = 63;
-    }
-    s->block_last_index[n] = i;
-    return 0;
-}
-
-/**
- * decodes the dc value.
- * @param n block index (0-3 are luma, 4-5 are chroma)
- * @param dir_ptr the prediction direction will be stored here
- * @return the quantized dc
- */
-static inline int mpeg4_decode_dc(MpegEncContext * s, int n, int *dir_ptr)
-{
-    int level, code;
-
-    if (n < 4)
-        code = get_vlc2(&s->gb, dc_lum.table, DC_VLC_BITS, 1);
-    else
-        code = get_vlc2(&s->gb, dc_chrom.table, DC_VLC_BITS, 1);
-    if (code < 0 || code > 9 /* && s->nbit<9 */){
-        av_log(s->avctx, AV_LOG_ERROR, "illegal dc vlc\n");
-        return -1;
-    }
-    if (code == 0) {
-        level = 0;
-    } else {
-        if(IS_3IV1){
-            if(code==1)
-                level= 2*get_bits1(&s->gb)-1;
-            else{
-                if(get_bits1(&s->gb))
-                    level = get_bits(&s->gb, code-1) + (1<<(code-1));
-                else
-                    level = -get_bits(&s->gb, code-1) - (1<<(code-1));
-            }
-        }else{
-            level = get_xbits(&s->gb, code);
-        }
-
-        if (code > 8){
-            if(get_bits1(&s->gb)==0){ /* marker */
-                if(s->error_resilience>=2){
-                    av_log(s->avctx, AV_LOG_ERROR, "dc marker bit missing\n");
-                    return -1;
-                }
-            }
-        }
-    }
-
-    return ff_mpeg4_pred_dc(s, n, level, dir_ptr, 0);
-}
-
-/**
- * decodes a block.
- * @return <0 if an error occured
- */
-static inline int mpeg4_decode_block(MpegEncContext * s, DCTELEM * block,
-                              int n, int coded, int intra, int rvlc)
-{
-    int level, i, last, run;
-    int dc_pred_dir;
-    RLTable * rl;
-    RL_VLC_ELEM * rl_vlc;
-    const uint8_t * scan_table;
-    int qmul, qadd;
-
-    //Note intra & rvlc should be optimized away if this is inlined
-
-    if(intra) {
-      if(s->use_intra_dc_vlc){
-        /* DC coef */
-        if(s->partitioned_frame){
-            level = s->dc_val[0][ s->block_index[n] ];
-            if(n<4) level= FASTDIV((level + (s->y_dc_scale>>1)), s->y_dc_scale);
-            else    level= FASTDIV((level + (s->c_dc_scale>>1)), s->c_dc_scale);
-            dc_pred_dir= (s->pred_dir_table[s->mb_x + s->mb_y*s->mb_stride]<<n)&32;
-        }else{
-            level = mpeg4_decode_dc(s, n, &dc_pred_dir);
-            if (level < 0)
-                return -1;
-        }
-        block[0] = level;
-        i = 0;
-      }else{
-            i = -1;
-            ff_mpeg4_pred_dc(s, n, 0, &dc_pred_dir, 0);
-      }
-        if (!coded)
-            goto not_coded;
-
-        if(rvlc){
-            rl = &rvlc_rl_intra;
-            rl_vlc = rvlc_rl_intra.rl_vlc[0];
-        }else{
-            rl = &rl_intra;
-            rl_vlc = rl_intra.rl_vlc[0];
-        }
-        if (s->ac_pred) {
-            if (dc_pred_dir == 0)
-                scan_table = s->intra_v_scantable.permutated; /* left */
-            else
-                scan_table = s->intra_h_scantable.permutated; /* top */
-        } else {
-            scan_table = s->intra_scantable.permutated;
-        }
-        qmul=1;
-        qadd=0;
-    } else {
-        i = -1;
-        if (!coded) {
-            s->block_last_index[n] = i;
-            return 0;
-        }
-        if(rvlc) rl = &rvlc_rl_inter;
-        else     rl = &rl_inter;
-
-        scan_table = s->intra_scantable.permutated;
-
-        if(s->mpeg_quant){
-            qmul=1;
-            qadd=0;
-            if(rvlc){
-                rl_vlc = rvlc_rl_inter.rl_vlc[0];
-            }else{
-                rl_vlc = rl_inter.rl_vlc[0];
-            }
-        }else{
-            qmul = s->qscale << 1;
-            qadd = (s->qscale - 1) | 1;
-            if(rvlc){
-                rl_vlc = rvlc_rl_inter.rl_vlc[s->qscale];
-            }else{
-                rl_vlc = rl_inter.rl_vlc[s->qscale];
-            }
-        }
-    }
-  {
-    OPEN_READER(re, &s->gb);
-    for(;;) {
-        UPDATE_CACHE(re, &s->gb);
-        GET_RL_VLC(level, run, re, &s->gb, rl_vlc, TEX_VLC_BITS, 2, 0);
-        if (level==0) {
-          /* escape */
-          if(rvlc){
-                if(SHOW_UBITS(re, &s->gb, 1)==0){
-                    av_log(s->avctx, AV_LOG_ERROR, "1. marker bit missing in rvlc esc\n");
-                    return -1;
-                }; SKIP_CACHE(re, &s->gb, 1);
-
-                last=  SHOW_UBITS(re, &s->gb, 1); SKIP_CACHE(re, &s->gb, 1);
-                run=   SHOW_UBITS(re, &s->gb, 6); LAST_SKIP_CACHE(re, &s->gb, 6);
-                SKIP_COUNTER(re, &s->gb, 1+1+6);
-                UPDATE_CACHE(re, &s->gb);
-
-                if(SHOW_UBITS(re, &s->gb, 1)==0){
-                    av_log(s->avctx, AV_LOG_ERROR, "2. marker bit missing in rvlc esc\n");
-                    return -1;
-                }; SKIP_CACHE(re, &s->gb, 1);
-
-                level= SHOW_UBITS(re, &s->gb, 11); SKIP_CACHE(re, &s->gb, 11);
-
-                if(SHOW_UBITS(re, &s->gb, 5)!=0x10){
-                    av_log(s->avctx, AV_LOG_ERROR, "reverse esc missing\n");
-                    return -1;
-                }; SKIP_CACHE(re, &s->gb, 5);
-
-                level=  level * qmul + qadd;
-                level = (level ^ SHOW_SBITS(re, &s->gb, 1)) - SHOW_SBITS(re, &s->gb, 1); LAST_SKIP_CACHE(re, &s->gb, 1);
-                SKIP_COUNTER(re, &s->gb, 1+11+5+1);
-
-                i+= run + 1;
-                if(last) i+=192;
-          }else{
-            int cache;
-            cache= GET_CACHE(re, &s->gb);
-
-            if(IS_3IV1)
-                cache ^= 0xC0000000;
-
-            if (cache&0x80000000) {
-                if (cache&0x40000000) {
-                    /* third escape */
-                    SKIP_CACHE(re, &s->gb, 2);
-                    last=  SHOW_UBITS(re, &s->gb, 1); SKIP_CACHE(re, &s->gb, 1);
-                    run=   SHOW_UBITS(re, &s->gb, 6); LAST_SKIP_CACHE(re, &s->gb, 6);
-                    SKIP_COUNTER(re, &s->gb, 2+1+6);
-                    UPDATE_CACHE(re, &s->gb);
-
-                    if(IS_3IV1){
-                        level= SHOW_SBITS(re, &s->gb, 12); LAST_SKIP_BITS(re, &s->gb, 12);
-                    }else{
-                        if(SHOW_UBITS(re, &s->gb, 1)==0){
-                            av_log(s->avctx, AV_LOG_ERROR, "1. marker bit missing in 3. esc\n");
-                            return -1;
-                        }; SKIP_CACHE(re, &s->gb, 1);
-
-                        level= SHOW_SBITS(re, &s->gb, 12); SKIP_CACHE(re, &s->gb, 12);
-
-                        if(SHOW_UBITS(re, &s->gb, 1)==0){
-                            av_log(s->avctx, AV_LOG_ERROR, "2. marker bit missing in 3. esc\n");
-                            return -1;
-                        }; LAST_SKIP_CACHE(re, &s->gb, 1);
-
-                        SKIP_COUNTER(re, &s->gb, 1+12+1);
-                    }
-
-#if 0
-                    if(s->error_resilience >= FF_ER_COMPLIANT){
-                        const int abs_level= FFABS(level);
-                        if(abs_level<=MAX_LEVEL && run<=MAX_RUN){
-                            const int run1= run - rl->max_run[last][abs_level] - 1;
-                            if(abs_level <= rl->max_level[last][run]){
-                                av_log(s->avctx, AV_LOG_ERROR, "illegal 3. esc, vlc encoding possible\n");
-                                return -1;
-                            }
-                            if(s->error_resilience > FF_ER_COMPLIANT){
-                                if(abs_level <= rl->max_level[last][run]*2){
-                                    av_log(s->avctx, AV_LOG_ERROR, "illegal 3. esc, esc 1 encoding possible\n");
-                                    return -1;
-                                }
-                                if(run1 >= 0 && abs_level <= rl->max_level[last][run1]){
-                                    av_log(s->avctx, AV_LOG_ERROR, "illegal 3. esc, esc 2 encoding possible\n");
-                                    return -1;
-                                }
-                            }
-                        }
-                    }
-#endif
-                    if (level>0) level= level * qmul + qadd;
-                    else         level= level * qmul - qadd;
-
-                    if((unsigned)(level + 2048) > 4095){
-                        if(s->error_resilience > FF_ER_COMPLIANT){
-                            if(level > 2560 || level<-2560){
-                                av_log(s->avctx, AV_LOG_ERROR, "|level| overflow in 3. esc, qp=%d\n", s->qscale);
-                                return -1;
-                            }
-                        }
-                        level= level<0 ? -2048 : 2047;
-                    }
-
-                    i+= run + 1;
-                    if(last) i+=192;
-                } else {
-                    /* second escape */
-#if MIN_CACHE_BITS < 20
-                    LAST_SKIP_BITS(re, &s->gb, 2);
-                    UPDATE_CACHE(re, &s->gb);
-#else
-                    SKIP_BITS(re, &s->gb, 2);
-#endif
-                    GET_RL_VLC(level, run, re, &s->gb, rl_vlc, TEX_VLC_BITS, 2, 1);
-                    i+= run + rl->max_run[run>>7][level/qmul] +1; //FIXME opt indexing
-                    level = (level ^ SHOW_SBITS(re, &s->gb, 1)) - SHOW_SBITS(re, &s->gb, 1);
-                    LAST_SKIP_BITS(re, &s->gb, 1);
-                }
-            } else {
-                /* first escape */
-#if MIN_CACHE_BITS < 19
-                LAST_SKIP_BITS(re, &s->gb, 1);
-                UPDATE_CACHE(re, &s->gb);
-#else
-                SKIP_BITS(re, &s->gb, 1);
-#endif
-                GET_RL_VLC(level, run, re, &s->gb, rl_vlc, TEX_VLC_BITS, 2, 1);
-                i+= run;
-                level = level + rl->max_level[run>>7][(run-1)&63] * qmul;//FIXME opt indexing
-                level = (level ^ SHOW_SBITS(re, &s->gb, 1)) - SHOW_SBITS(re, &s->gb, 1);
-                LAST_SKIP_BITS(re, &s->gb, 1);
-            }
-          }
-        } else {
-            i+= run;
-            level = (level ^ SHOW_SBITS(re, &s->gb, 1)) - SHOW_SBITS(re, &s->gb, 1);
-            LAST_SKIP_BITS(re, &s->gb, 1);
-        }
-        if (i > 62){
-            i-= 192;
-            if(i&(~63)){
-                av_log(s->avctx, AV_LOG_ERROR, "ac-tex damaged at %d %d\n", s->mb_x, s->mb_y);
-                return -1;
-            }
-
-            block[scan_table[i]] = level;
-            break;
-        }
-
-        block[scan_table[i]] = level;
-    }
-    CLOSE_READER(re, &s->gb);
-  }
- not_coded:
-    if (intra) {
-        if(!s->use_intra_dc_vlc){
-            block[0] = ff_mpeg4_pred_dc(s, n, block[0], &dc_pred_dir, 0);
-
-            i -= i>>31; //if(i == -1) i=0;
-        }
-
-        mpeg4_pred_ac(s, block, n, dc_pred_dir);
-        if (s->ac_pred) {
-            i = 63; /* XXX: not optimal */
-        }
-    }
-    s->block_last_index[n] = i;
-    return 0;
-}
-
-/* most is hardcoded. should extend to handle all h263 streams */
-int h263_decode_picture_header(MpegEncContext *s)
-{
-    int format, width, height, i;
-    uint32_t startcode;
-
-    align_get_bits(&s->gb);
-
-    startcode= get_bits(&s->gb, 22-8);
-
-    for(i= s->gb.size_in_bits - get_bits_count(&s->gb); i>24; i-=8) {
-        startcode = ((startcode << 8) | get_bits(&s->gb, 8)) & 0x003FFFFF;
-
-        if(startcode == 0x20)
-            break;
-    }
-
-    if (startcode != 0x20) {
-        av_log(s->avctx, AV_LOG_ERROR, "Bad picture start code\n");
-        return -1;
-    }
-    /* temporal reference */
-    i = get_bits(&s->gb, 8); /* picture timestamp */
-    if( (s->picture_number&~0xFF)+i < s->picture_number)
-        i+= 256;
-    s->current_picture_ptr->pts=
-    s->picture_number= (s->picture_number&~0xFF) + i;
-
-    /* PTYPE starts here */
-    if (get_bits1(&s->gb) != 1) {
-        /* marker */
-        av_log(s->avctx, AV_LOG_ERROR, "Bad marker\n");
-        return -1;
-    }
-    if (get_bits1(&s->gb) != 0) {
-        av_log(s->avctx, AV_LOG_ERROR, "Bad H263 id\n");
-        return -1;      /* h263 id */
-    }
-    skip_bits1(&s->gb);         /* split screen off */
-    skip_bits1(&s->gb);         /* camera  off */
-    skip_bits1(&s->gb);         /* freeze picture release off */
-
-    format = get_bits(&s->gb, 3);
-    /*
-        0    forbidden
-        1    sub-QCIF
-        10   QCIF
-        7       extended PTYPE (PLUSPTYPE)
-    */
-
-    if (format != 7 && format != 6) {
-        s->h263_plus = 0;
-        /* H.263v1 */
-        width = h263_format[format][0];
-        height = h263_format[format][1];
-        if (!width)
-            return -1;
-
-        s->pict_type = I_TYPE + get_bits1(&s->gb);
-
-        s->h263_long_vectors = get_bits1(&s->gb);
-
-        if (get_bits1(&s->gb) != 0) {
-            av_log(s->avctx, AV_LOG_ERROR, "H263 SAC not supported\n");
-            return -1; /* SAC: off */
-        }
-        s->obmc= get_bits1(&s->gb); /* Advanced prediction mode */
-        s->unrestricted_mv = s->h263_long_vectors || s->obmc;
-
-        if (get_bits1(&s->gb) != 0) {
-            av_log(s->avctx, AV_LOG_ERROR, "H263 PB frame not supported\n");
-            return -1; /* not PB frame */
-        }
-        s->chroma_qscale= s->qscale = get_bits(&s->gb, 5);
-        skip_bits1(&s->gb); /* Continuous Presence Multipoint mode: off */
-
-        s->width = width;
-        s->height = height;
-        s->avctx->sample_aspect_ratio= (AVRational){12,11};
-        s->avctx->time_base= (AVRational){1001, 30000};
-    } else {
-        int ufep;
-
-        /* H.263v2 */
-        s->h263_plus = 1;
-        ufep = get_bits(&s->gb, 3); /* Update Full Extended PTYPE */
-
-        /* ufep other than 0 and 1 are reserved */
-        if (ufep == 1) {
-            /* OPPTYPE */
-            format = get_bits(&s->gb, 3);
-            dprintf("ufep=1, format: %d\n", format);
-            s->custom_pcf= get_bits1(&s->gb);
-            s->umvplus = get_bits(&s->gb, 1); /* Unrestricted Motion Vector */
-            if (get_bits1(&s->gb) != 0) {
-                av_log(s->avctx, AV_LOG_ERROR, "Syntax-based Arithmetic Coding (SAC) not supported\n");
-            }
-            s->obmc= get_bits1(&s->gb); /* Advanced prediction mode */
-            s->h263_aic = get_bits1(&s->gb); /* Advanced Intra Coding (AIC) */
-            s->loop_filter= get_bits1(&s->gb);
-            s->unrestricted_mv = s->umvplus || s->obmc || s->loop_filter;
-
-            s->h263_slice_structured= get_bits1(&s->gb);
-            if (get_bits1(&s->gb) != 0) {
-                av_log(s->avctx, AV_LOG_ERROR, "Reference Picture Selection not supported\n");
-            }
-            if (get_bits1(&s->gb) != 0) {
-                av_log(s->avctx, AV_LOG_ERROR, "Independent Segment Decoding not supported\n");
-            }
-            s->alt_inter_vlc= get_bits1(&s->gb);
-            s->modified_quant= get_bits1(&s->gb);
-            if(s->modified_quant)
-                s->chroma_qscale_table= ff_h263_chroma_qscale_table;
-
-            skip_bits(&s->gb, 1); /* Prevent start code emulation */
-
-            skip_bits(&s->gb, 3); /* Reserved */
-        } else if (ufep != 0) {
-            av_log(s->avctx, AV_LOG_ERROR, "Bad UFEP type (%d)\n", ufep);
-            return -1;
-        }
-
-        /* MPPTYPE */
-        s->pict_type = get_bits(&s->gb, 3);
-        switch(s->pict_type){
-        case 0: s->pict_type= I_TYPE;break;
-        case 1: s->pict_type= P_TYPE;break;
-        case 3: s->pict_type= B_TYPE;break;
-        case 7: s->pict_type= I_TYPE;break; //ZYGO
-        default:
-            return -1;
-        }
-        skip_bits(&s->gb, 2);
-        s->no_rounding = get_bits1(&s->gb);
-        skip_bits(&s->gb, 4);
-
-        /* Get the picture dimensions */
-        if (ufep) {
-            if (format == 6) {
-                /* Custom Picture Format (CPFMT) */
-                s->aspect_ratio_info = get_bits(&s->gb, 4);
-                dprintf("aspect: %d\n", s->aspect_ratio_info);
-                /* aspect ratios:
-                0 - forbidden
-                1 - 1:1
-                2 - 12:11 (CIF 4:3)
-                3 - 10:11 (525-type 4:3)
-                4 - 16:11 (CIF 16:9)
-                5 - 40:33 (525-type 16:9)
-                6-14 - reserved
-                */
-                width = (get_bits(&s->gb, 9) + 1) * 4;
-                skip_bits1(&s->gb);
-                height = get_bits(&s->gb, 9) * 4;
-                dprintf("\nH.263+ Custom picture: %dx%d\n",width,height);
-                if (s->aspect_ratio_info == FF_ASPECT_EXTENDED) {
-                    /* aspected dimensions */
-                    s->avctx->sample_aspect_ratio.num= get_bits(&s->gb, 8);
-                    s->avctx->sample_aspect_ratio.den= get_bits(&s->gb, 8);
-                }else{
-                    s->avctx->sample_aspect_ratio= pixel_aspect[s->aspect_ratio_info];
-                }
-            } else {
-                width = h263_format[format][0];
-                height = h263_format[format][1];
-                s->avctx->sample_aspect_ratio= (AVRational){12,11};
-            }
-            if ((width == 0) || (height == 0))
-                return -1;
-            s->width = width;
-            s->height = height;
-
-            if(s->custom_pcf){
-                int gcd;
-                s->avctx->time_base.den= 1800000;
-                s->avctx->time_base.num= 1000 + get_bits1(&s->gb);
-                s->avctx->time_base.num*= get_bits(&s->gb, 7);
-                if(s->avctx->time_base.num == 0){
-                    av_log(s, AV_LOG_ERROR, "zero framerate\n");
-                    return -1;
-                }
-                gcd= ff_gcd(s->avctx->time_base.den, s->avctx->time_base.num);
-                s->avctx->time_base.den /= gcd;
-                s->avctx->time_base.num /= gcd;
-//                av_log(s->avctx, AV_LOG_DEBUG, "%d/%d\n", s->avctx->time_base.den, s->avctx->time_base.num);
-            }else{
-                s->avctx->time_base= (AVRational){1001, 30000};
-            }
-        }
-
-        if(s->custom_pcf){
-            skip_bits(&s->gb, 2); //extended Temporal reference
-        }
-
-        if (ufep) {
-            if (s->umvplus) {
-                if(get_bits1(&s->gb)==0) /* Unlimited Unrestricted Motion Vectors Indicator (UUI) */
-                    skip_bits1(&s->gb);
-            }
-            if(s->h263_slice_structured){
-                if (get_bits1(&s->gb) != 0) {
-                    av_log(s->avctx, AV_LOG_ERROR, "rectangular slices not supported\n");
-                }
-                if (get_bits1(&s->gb) != 0) {
-                    av_log(s->avctx, AV_LOG_ERROR, "unordered slices not supported\n");
-                }
-            }
-        }
-
-        s->qscale = get_bits(&s->gb, 5);
-    }
-
-    s->mb_width = (s->width  + 15) / 16;
-    s->mb_height = (s->height  + 15) / 16;
-    s->mb_num = s->mb_width * s->mb_height;
-
-    /* PEI */
-    while (get_bits1(&s->gb) != 0) {
-        skip_bits(&s->gb, 8);
-    }
-
-    if(s->h263_slice_structured){
-        if (get_bits1(&s->gb) != 1) {
-            av_log(s->avctx, AV_LOG_ERROR, "SEPB1 marker missing\n");
-            return -1;
-        }
-
-        ff_h263_decode_mba(s);
-
-        if (get_bits1(&s->gb) != 1) {
-            av_log(s->avctx, AV_LOG_ERROR, "SEPB2 marker missing\n");
-            return -1;
-        }
-    }
-    s->f_code = 1;
-
-    if(s->h263_aic){
-         s->y_dc_scale_table=
-         s->c_dc_scale_table= ff_aic_dc_scale_table;
-    }else{
-        s->y_dc_scale_table=
-        s->c_dc_scale_table= ff_mpeg1_dc_scale_table;
-    }
-
-     if(s->avctx->debug&FF_DEBUG_PICT_INFO){
-         av_log(s->avctx, AV_LOG_DEBUG, "qp:%d %c size:%d rnd:%d%s%s%s%s%s%s%s%s%s %d/%d\n",
-         s->qscale, av_get_pict_type_char(s->pict_type),
-         s->gb.size_in_bits, 1-s->no_rounding,
-         s->obmc ? " AP" : "",
-         s->umvplus ? " UMV" : "",
-         s->h263_long_vectors ? " LONG" : "",
-         s->h263_plus ? " +" : "",
-         s->h263_aic ? " AIC" : "",
-         s->alt_inter_vlc ? " AIV" : "",
-         s->modified_quant ? " MQ" : "",
-         s->loop_filter ? " LOOP" : "",
-         s->h263_slice_structured ? " SS" : "",
-         s->avctx->time_base.den, s->avctx->time_base.num
-         );
-     }
-#if 1
-    if (s->pict_type == I_TYPE && s->avctx->codec_tag == ff_get_fourcc("ZYGO")){
-        int i,j;
-        for(i=0; i<85; i++) av_log(s->avctx, AV_LOG_DEBUG, "%d", get_bits1(&s->gb));
-        av_log(s->avctx, AV_LOG_DEBUG, "\n");
-        for(i=0; i<13; i++){
-            for(j=0; j<3; j++){
-                int v= get_bits(&s->gb, 8);
-                v |= get_sbits(&s->gb, 8)<<8;
-                av_log(s->avctx, AV_LOG_DEBUG, " %5d", v);
-            }
-            av_log(s->avctx, AV_LOG_DEBUG, "\n");
-        }
-        for(i=0; i<50; i++) av_log(s->avctx, AV_LOG_DEBUG, "%d", get_bits1(&s->gb));
-    }
-#endif
-
-    return 0;
-}
-
-static void mpeg4_decode_sprite_trajectory(MpegEncContext * s, GetBitContext *gb)
-{
-    int i;
-    int a= 2<<s->sprite_warping_accuracy;
-    int rho= 3-s->sprite_warping_accuracy;
-    int r=16/a;
-    const int vop_ref[4][2]= {{0,0}, {s->width,0}, {0, s->height}, {s->width, s->height}}; // only true for rectangle shapes
-    int d[4][2]={{0,0}, {0,0}, {0,0}, {0,0}};
-    int sprite_ref[4][2];
-    int virtual_ref[2][2];
-    int w2, h2, w3, h3;
-    int alpha=0, beta=0;
-    int w= s->width;
-    int h= s->height;
-    int min_ab;
-
-    for(i=0; i<s->num_sprite_warping_points; i++){
-        int length;
-        int x=0, y=0;
-
-        length= get_vlc2(gb, sprite_trajectory.table, SPRITE_TRAJ_VLC_BITS, 3);
-        if(length){
-            x= get_xbits(gb, length);
-        }
-        if(!(s->divx_version==500 && s->divx_build==413)) skip_bits1(gb); /* marker bit */
-
-        length= get_vlc2(gb, sprite_trajectory.table, SPRITE_TRAJ_VLC_BITS, 3);
-        if(length){
-            y=get_xbits(gb, length);
-        }
-        skip_bits1(gb); /* marker bit */
-//printf("%d %d %d %d\n", x, y, i, s->sprite_warping_accuracy);
-        d[i][0]= x;
-        d[i][1]= y;
-    }
-
-    while((1<<alpha)<w) alpha++;
-    while((1<<beta )<h) beta++; // there seems to be a typo in the mpeg4 std for the definition of w' and h'
-    w2= 1<<alpha;
-    h2= 1<<beta;
-
-// Note, the 4th point isn't used for GMC
-    if(s->divx_version==500 && s->divx_build==413){
-        sprite_ref[0][0]= a*vop_ref[0][0] + d[0][0];
-        sprite_ref[0][1]= a*vop_ref[0][1] + d[0][1];
-        sprite_ref[1][0]= a*vop_ref[1][0] + d[0][0] + d[1][0];
-        sprite_ref[1][1]= a*vop_ref[1][1] + d[0][1] + d[1][1];
-        sprite_ref[2][0]= a*vop_ref[2][0] + d[0][0] + d[2][0];
-        sprite_ref[2][1]= a*vop_ref[2][1] + d[0][1] + d[2][1];
-    } else {
-        sprite_ref[0][0]= (a>>1)*(2*vop_ref[0][0] + d[0][0]);
-        sprite_ref[0][1]= (a>>1)*(2*vop_ref[0][1] + d[0][1]);
-        sprite_ref[1][0]= (a>>1)*(2*vop_ref[1][0] + d[0][0] + d[1][0]);
-        sprite_ref[1][1]= (a>>1)*(2*vop_ref[1][1] + d[0][1] + d[1][1]);
-        sprite_ref[2][0]= (a>>1)*(2*vop_ref[2][0] + d[0][0] + d[2][0]);
-        sprite_ref[2][1]= (a>>1)*(2*vop_ref[2][1] + d[0][1] + d[2][1]);
-    }
-/*    sprite_ref[3][0]= (a>>1)*(2*vop_ref[3][0] + d[0][0] + d[1][0] + d[2][0] + d[3][0]);
-    sprite_ref[3][1]= (a>>1)*(2*vop_ref[3][1] + d[0][1] + d[1][1] + d[2][1] + d[3][1]); */
-
-// this is mostly identical to the mpeg4 std (and is totally unreadable because of that ...)
-// perhaps it should be reordered to be more readable ...
-// the idea behind this virtual_ref mess is to be able to use shifts later per pixel instead of divides
-// so the distance between points is converted from w&h based to w2&h2 based which are of the 2^x form
-    virtual_ref[0][0]= 16*(vop_ref[0][0] + w2)
-        + ROUNDED_DIV(((w - w2)*(r*sprite_ref[0][0] - 16*vop_ref[0][0]) + w2*(r*sprite_ref[1][0] - 16*vop_ref[1][0])),w);
-    virtual_ref[0][1]= 16*vop_ref[0][1]
-        + ROUNDED_DIV(((w - w2)*(r*sprite_ref[0][1] - 16*vop_ref[0][1]) + w2*(r*sprite_ref[1][1] - 16*vop_ref[1][1])),w);
-    virtual_ref[1][0]= 16*vop_ref[0][0]
-        + ROUNDED_DIV(((h - h2)*(r*sprite_ref[0][0] - 16*vop_ref[0][0]) + h2*(r*sprite_ref[2][0] - 16*vop_ref[2][0])),h);
-    virtual_ref[1][1]= 16*(vop_ref[0][1] + h2)
-        + ROUNDED_DIV(((h - h2)*(r*sprite_ref[0][1] - 16*vop_ref[0][1]) + h2*(r*sprite_ref[2][1] - 16*vop_ref[2][1])),h);
-
-    switch(s->num_sprite_warping_points)
-    {
-        case 0:
-            s->sprite_offset[0][0]= 0;
-            s->sprite_offset[0][1]= 0;
-            s->sprite_offset[1][0]= 0;
-            s->sprite_offset[1][1]= 0;
-            s->sprite_delta[0][0]= a;
-            s->sprite_delta[0][1]= 0;
-            s->sprite_delta[1][0]= 0;
-            s->sprite_delta[1][1]= a;
-            s->sprite_shift[0]= 0;
-            s->sprite_shift[1]= 0;
-            break;
-        case 1: //GMC only
-            s->sprite_offset[0][0]= sprite_ref[0][0] - a*vop_ref[0][0];
-            s->sprite_offset[0][1]= sprite_ref[0][1] - a*vop_ref[0][1];
-            s->sprite_offset[1][0]= ((sprite_ref[0][0]>>1)|(sprite_ref[0][0]&1)) - a*(vop_ref[0][0]/2);
-            s->sprite_offset[1][1]= ((sprite_ref[0][1]>>1)|(sprite_ref[0][1]&1)) - a*(vop_ref[0][1]/2);
-            s->sprite_delta[0][0]= a;
-            s->sprite_delta[0][1]= 0;
-            s->sprite_delta[1][0]= 0;
-            s->sprite_delta[1][1]= a;
-            s->sprite_shift[0]= 0;
-            s->sprite_shift[1]= 0;
-            break;
-        case 2:
-            s->sprite_offset[0][0]= (sprite_ref[0][0]<<(alpha+rho))
-                                                  + (-r*sprite_ref[0][0] + virtual_ref[0][0])*(-vop_ref[0][0])
-                                                  + ( r*sprite_ref[0][1] - virtual_ref[0][1])*(-vop_ref[0][1])
-                                                  + (1<<(alpha+rho-1));
-            s->sprite_offset[0][1]= (sprite_ref[0][1]<<(alpha+rho))
-                                                  + (-r*sprite_ref[0][1] + virtual_ref[0][1])*(-vop_ref[0][0])
-                                                  + (-r*sprite_ref[0][0] + virtual_ref[0][0])*(-vop_ref[0][1])
-                                                  + (1<<(alpha+rho-1));
-            s->sprite_offset[1][0]= ( (-r*sprite_ref[0][0] + virtual_ref[0][0])*(-2*vop_ref[0][0] + 1)
-                                     +( r*sprite_ref[0][1] - virtual_ref[0][1])*(-2*vop_ref[0][1] + 1)
-                                     +2*w2*r*sprite_ref[0][0]
-                                     - 16*w2
-                                     + (1<<(alpha+rho+1)));
-            s->sprite_offset[1][1]= ( (-r*sprite_ref[0][1] + virtual_ref[0][1])*(-2*vop_ref[0][0] + 1)
-                                     +(-r*sprite_ref[0][0] + virtual_ref[0][0])*(-2*vop_ref[0][1] + 1)
-                                     +2*w2*r*sprite_ref[0][1]
-                                     - 16*w2
-                                     + (1<<(alpha+rho+1)));
-            s->sprite_delta[0][0]=   (-r*sprite_ref[0][0] + virtual_ref[0][0]);
-            s->sprite_delta[0][1]=   (+r*sprite_ref[0][1] - virtual_ref[0][1]);
-            s->sprite_delta[1][0]=   (-r*sprite_ref[0][1] + virtual_ref[0][1]);
-            s->sprite_delta[1][1]=   (-r*sprite_ref[0][0] + virtual_ref[0][0]);
-
-            s->sprite_shift[0]= alpha+rho;
-            s->sprite_shift[1]= alpha+rho+2;
-            break;
-        case 3:
-            min_ab= FFMIN(alpha, beta);
-            w3= w2>>min_ab;
-            h3= h2>>min_ab;
-            s->sprite_offset[0][0]=  (sprite_ref[0][0]<<(alpha+beta+rho-min_ab))
-                                   + (-r*sprite_ref[0][0] + virtual_ref[0][0])*h3*(-vop_ref[0][0])
-                                   + (-r*sprite_ref[0][0] + virtual_ref[1][0])*w3*(-vop_ref[0][1])
-                                   + (1<<(alpha+beta+rho-min_ab-1));
-            s->sprite_offset[0][1]=  (sprite_ref[0][1]<<(alpha+beta+rho-min_ab))
-                                   + (-r*sprite_ref[0][1] + virtual_ref[0][1])*h3*(-vop_ref[0][0])
-                                   + (-r*sprite_ref[0][1] + virtual_ref[1][1])*w3*(-vop_ref[0][1])
-                                   + (1<<(alpha+beta+rho-min_ab-1));
-            s->sprite_offset[1][0]=  (-r*sprite_ref[0][0] + virtual_ref[0][0])*h3*(-2*vop_ref[0][0] + 1)
-                                   + (-r*sprite_ref[0][0] + virtual_ref[1][0])*w3*(-2*vop_ref[0][1] + 1)
-                                   + 2*w2*h3*r*sprite_ref[0][0]
-                                   - 16*w2*h3
-                                   + (1<<(alpha+beta+rho-min_ab+1));
-            s->sprite_offset[1][1]=  (-r*sprite_ref[0][1] + virtual_ref[0][1])*h3*(-2*vop_ref[0][0] + 1)
-                                   + (-r*sprite_ref[0][1] + virtual_ref[1][1])*w3*(-2*vop_ref[0][1] + 1)
-                                   + 2*w2*h3*r*sprite_ref[0][1]
-                                   - 16*w2*h3
-                                   + (1<<(alpha+beta+rho-min_ab+1));
-            s->sprite_delta[0][0]=   (-r*sprite_ref[0][0] + virtual_ref[0][0])*h3;
-            s->sprite_delta[0][1]=   (-r*sprite_ref[0][0] + virtual_ref[1][0])*w3;
-            s->sprite_delta[1][0]=   (-r*sprite_ref[0][1] + virtual_ref[0][1])*h3;
-            s->sprite_delta[1][1]=   (-r*sprite_ref[0][1] + virtual_ref[1][1])*w3;
-
-            s->sprite_shift[0]= alpha + beta + rho - min_ab;
-            s->sprite_shift[1]= alpha + beta + rho - min_ab + 2;
-            break;
-    }
-    /* try to simplify the situation */
-    if(   s->sprite_delta[0][0] == a<<s->sprite_shift[0]
-       && s->sprite_delta[0][1] == 0
-       && s->sprite_delta[1][0] == 0
-       && s->sprite_delta[1][1] == a<<s->sprite_shift[0])
-    {
-        s->sprite_offset[0][0]>>=s->sprite_shift[0];
-        s->sprite_offset[0][1]>>=s->sprite_shift[0];
-        s->sprite_offset[1][0]>>=s->sprite_shift[1];
-        s->sprite_offset[1][1]>>=s->sprite_shift[1];
-        s->sprite_delta[0][0]= a;
-        s->sprite_delta[0][1]= 0;
-        s->sprite_delta[1][0]= 0;
-        s->sprite_delta[1][1]= a;
-        s->sprite_shift[0]= 0;
-        s->sprite_shift[1]= 0;
-        s->real_sprite_warping_points=1;
-    }
-    else{
-        int shift_y= 16 - s->sprite_shift[0];
-        int shift_c= 16 - s->sprite_shift[1];
-//printf("shifts %d %d\n", shift_y, shift_c);
-        for(i=0; i<2; i++){
-            s->sprite_offset[0][i]<<= shift_y;
-            s->sprite_offset[1][i]<<= shift_c;
-            s->sprite_delta[0][i]<<= shift_y;
-            s->sprite_delta[1][i]<<= shift_y;
-            s->sprite_shift[i]= 16;
-        }
-        s->real_sprite_warping_points= s->num_sprite_warping_points;
-    }
-#if 0
-printf("vop:%d:%d %d:%d %d:%d, sprite:%d:%d %d:%d %d:%d, virtual: %d:%d %d:%d\n",
-    vop_ref[0][0], vop_ref[0][1],
-    vop_ref[1][0], vop_ref[1][1],
-    vop_ref[2][0], vop_ref[2][1],
-    sprite_ref[0][0], sprite_ref[0][1],
-    sprite_ref[1][0], sprite_ref[1][1],
-    sprite_ref[2][0], sprite_ref[2][1],
-    virtual_ref[0][0], virtual_ref[0][1],
-    virtual_ref[1][0], virtual_ref[1][1]
-    );
-
-printf("offset: %d:%d , delta: %d %d %d %d, shift %d\n",
-    s->sprite_offset[0][0], s->sprite_offset[0][1],
-    s->sprite_delta[0][0], s->sprite_delta[0][1],
-    s->sprite_delta[1][0], s->sprite_delta[1][1],
-    s->sprite_shift[0]
-    );
-#endif
-}
-
-static int mpeg4_decode_gop_header(MpegEncContext * s, GetBitContext *gb){
-    int hours, minutes, seconds;
-
-    hours= get_bits(gb, 5);
-    minutes= get_bits(gb, 6);
-    skip_bits1(gb);
-    seconds= get_bits(gb, 6);
-
-    s->time_base= seconds + 60*(minutes + 60*hours);
-
-    skip_bits1(gb);
-    skip_bits1(gb);
-
-    return 0;
-}
-
-static int decode_vol_header(MpegEncContext *s, GetBitContext *gb){
-    int width, height, vo_ver_id;
-
-    /* vol header */
-    skip_bits(gb, 1); /* random access */
-    s->vo_type= get_bits(gb, 8);
-    if (get_bits1(gb) != 0) { /* is_ol_id */
-        vo_ver_id = get_bits(gb, 4); /* vo_ver_id */
-        skip_bits(gb, 3); /* vo_priority */
-    } else {
-        vo_ver_id = 1;
-    }
-//printf("vo type:%d\n",s->vo_type);
-    s->aspect_ratio_info= get_bits(gb, 4);
-    if(s->aspect_ratio_info == FF_ASPECT_EXTENDED){
-        s->avctx->sample_aspect_ratio.num= get_bits(gb, 8); // par_width
-        s->avctx->sample_aspect_ratio.den= get_bits(gb, 8); // par_height
-    }else{
-        s->avctx->sample_aspect_ratio= pixel_aspect[s->aspect_ratio_info];
-    }
-
-    if ((s->vol_control_parameters=get_bits1(gb))) { /* vol control parameter */
-        int chroma_format= get_bits(gb, 2);
-        if(chroma_format!=1){
-            av_log(s->avctx, AV_LOG_ERROR, "illegal chroma format\n");
-        }
-        s->low_delay= get_bits1(gb);
-        if(get_bits1(gb)){ /* vbv parameters */
-            get_bits(gb, 15);   /* first_half_bitrate */
-            skip_bits1(gb);     /* marker */
-            get_bits(gb, 15);   /* latter_half_bitrate */
-            skip_bits1(gb);     /* marker */
-            get_bits(gb, 15);   /* first_half_vbv_buffer_size */
-            skip_bits1(gb);     /* marker */
-            get_bits(gb, 3);    /* latter_half_vbv_buffer_size */
-            get_bits(gb, 11);   /* first_half_vbv_occupancy */
-            skip_bits1(gb);     /* marker */
-            get_bits(gb, 15);   /* latter_half_vbv_occupancy */
-            skip_bits1(gb);     /* marker */
-        }
-    }else{
-        // set low delay flag only once the smartest? low delay detection won't be overriden
-        if(s->picture_number==0)
-            s->low_delay=0;
-    }
-
-    s->shape = get_bits(gb, 2); /* vol shape */
-    if(s->shape != RECT_SHAPE) av_log(s->avctx, AV_LOG_ERROR, "only rectangular vol supported\n");
-    if(s->shape == GRAY_SHAPE && vo_ver_id != 1){
-        av_log(s->avctx, AV_LOG_ERROR, "Gray shape not supported\n");
-        skip_bits(gb, 4);  //video_object_layer_shape_extension
-    }
-
-    check_marker(gb, "before time_increment_resolution");
-
-    s->avctx->time_base.den = get_bits(gb, 16);
-    if(!s->avctx->time_base.den){
-        av_log(s->avctx, AV_LOG_ERROR, "time_base.den==0\n");
-        return -1;
-    }
-
-    s->time_increment_bits = av_log2(s->avctx->time_base.den - 1) + 1;
-    if (s->time_increment_bits < 1)
-        s->time_increment_bits = 1;
-
-    check_marker(gb, "before fixed_vop_rate");
-
-    if (get_bits1(gb) != 0) {   /* fixed_vop_rate  */
-        s->avctx->time_base.num = get_bits(gb, s->time_increment_bits);
-    }else
-        s->avctx->time_base.num = 1;
-
-    s->t_frame=0;
-
-    if (s->shape != BIN_ONLY_SHAPE) {
-        if (s->shape == RECT_SHAPE) {
-            skip_bits1(gb);   /* marker */
-            width = get_bits(gb, 13);
-            skip_bits1(gb);   /* marker */
-            height = get_bits(gb, 13);
-            skip_bits1(gb);   /* marker */
-            if(width && height && !(s->width && s->avctx->codec_tag == ff_get_fourcc("MP4S"))){ /* they should be non zero but who knows ... */
-                s->width = width;
-                s->height = height;
-//                printf("width/height: %d %d\n", width, height);
-            }
-        }
-
-        s->progressive_sequence=
-        s->progressive_frame= get_bits1(gb)^1;
-        s->interlaced_dct=0;
-        if(!get_bits1(gb) && (s->avctx->debug & FF_DEBUG_PICT_INFO))
-            av_log(s->avctx, AV_LOG_INFO, "MPEG4 OBMC not supported (very likely buggy encoder)\n");   /* OBMC Disable */
-        if (vo_ver_id == 1) {
-            s->vol_sprite_usage = get_bits1(gb); /* vol_sprite_usage */
-        } else {
-            s->vol_sprite_usage = get_bits(gb, 2); /* vol_sprite_usage */
-        }
-        if(s->vol_sprite_usage==STATIC_SPRITE) av_log(s->avctx, AV_LOG_ERROR, "Static Sprites not supported\n");
-        if(s->vol_sprite_usage==STATIC_SPRITE || s->vol_sprite_usage==GMC_SPRITE){
-            if(s->vol_sprite_usage==STATIC_SPRITE){
-                s->sprite_width = get_bits(gb, 13);
-                skip_bits1(gb); /* marker */
-                s->sprite_height= get_bits(gb, 13);
-                skip_bits1(gb); /* marker */
-                s->sprite_left  = get_bits(gb, 13);
-                skip_bits1(gb); /* marker */
-                s->sprite_top   = get_bits(gb, 13);
-                skip_bits1(gb); /* marker */
-            }
-            s->num_sprite_warping_points= get_bits(gb, 6);
-            s->sprite_warping_accuracy = get_bits(gb, 2);
-            s->sprite_brightness_change= get_bits1(gb);
-            if(s->vol_sprite_usage==STATIC_SPRITE)
-                s->low_latency_sprite= get_bits1(gb);
-        }
-        // FIXME sadct disable bit if verid!=1 && shape not rect
-
-        if (get_bits1(gb) == 1) {   /* not_8_bit */
-            s->quant_precision = get_bits(gb, 4); /* quant_precision */
-            if(get_bits(gb, 4)!=8) av_log(s->avctx, AV_LOG_ERROR, "N-bit not supported\n"); /* bits_per_pixel */
-            if(s->quant_precision!=5) av_log(s->avctx, AV_LOG_ERROR, "quant precision %d\n", s->quant_precision);
-        } else {
-            s->quant_precision = 5;
-        }
-
-        // FIXME a bunch of grayscale shape things
-
-        if((s->mpeg_quant=get_bits1(gb))){ /* vol_quant_type */
-            int i, v;
-
-            /* load default matrixes */
-            for(i=0; i<64; i++){
-                int j= s->dsp.idct_permutation[i];
-                v= ff_mpeg4_default_intra_matrix[i];
-                s->intra_matrix[j]= v;
-                s->chroma_intra_matrix[j]= v;
-
-                v= ff_mpeg4_default_non_intra_matrix[i];
-                s->inter_matrix[j]= v;
-                s->chroma_inter_matrix[j]= v;
-            }
-
-            /* load custom intra matrix */
-            if(get_bits1(gb)){
-                int last=0;
-                for(i=0; i<64; i++){
-                    int j;
-                    v= get_bits(gb, 8);
-                    if(v==0) break;
-
-                    last= v;
-                    j= s->dsp.idct_permutation[ ff_zigzag_direct[i] ];
-                    s->intra_matrix[j]= v;
-                    s->chroma_intra_matrix[j]= v;
-                }
-
-                /* replicate last value */
-                for(; i<64; i++){
-                    int j= s->dsp.idct_permutation[ ff_zigzag_direct[i] ];
-                    s->intra_matrix[j]= last;
-                    s->chroma_intra_matrix[j]= last;
-                }
-            }
-
-            /* load custom non intra matrix */
-            if(get_bits1(gb)){
-                int last=0;
-                for(i=0; i<64; i++){
-                    int j;
-                    v= get_bits(gb, 8);
-                    if(v==0) break;
-
-                    last= v;
-                    j= s->dsp.idct_permutation[ ff_zigzag_direct[i] ];
-                    s->inter_matrix[j]= v;
-                    s->chroma_inter_matrix[j]= v;
-                }
-
-                /* replicate last value */
-                for(; i<64; i++){
-                    int j= s->dsp.idct_permutation[ ff_zigzag_direct[i] ];
-                    s->inter_matrix[j]= last;
-                    s->chroma_inter_matrix[j]= last;
-                }
-            }
-
-            // FIXME a bunch of grayscale shape things
-        }
-
-        if(vo_ver_id != 1)
-             s->quarter_sample= get_bits1(gb);
-        else s->quarter_sample=0;
-
-        if(!get_bits1(gb)) av_log(s->avctx, AV_LOG_ERROR, "Complexity estimation not supported\n");
-
-        s->resync_marker= !get_bits1(gb); /* resync_marker_disabled */
-
-        s->data_partitioning= get_bits1(gb);
-        if(s->data_partitioning){
-            s->rvlc= get_bits1(gb);
-        }
-
-        if(vo_ver_id != 1) {
-            s->new_pred= get_bits1(gb);
-            if(s->new_pred){
-                av_log(s->avctx, AV_LOG_ERROR, "new pred not supported\n");
-                skip_bits(gb, 2); /* requested upstream message type */
-                skip_bits1(gb); /* newpred segment type */
-            }
-            s->reduced_res_vop= get_bits1(gb);
-            if(s->reduced_res_vop) av_log(s->avctx, AV_LOG_ERROR, "reduced resolution VOP not supported\n");
-        }
-        else{
-            s->new_pred=0;
-            s->reduced_res_vop= 0;
-        }
-
-        s->scalability= get_bits1(gb);
-
-        if (s->scalability) {
-            GetBitContext bak= *gb;
-            int ref_layer_id;
-            int ref_layer_sampling_dir;
-            int h_sampling_factor_n;
-            int h_sampling_factor_m;
-            int v_sampling_factor_n;
-            int v_sampling_factor_m;
-
-            s->hierachy_type= get_bits1(gb);
-            ref_layer_id= get_bits(gb, 4);
-            ref_layer_sampling_dir= get_bits1(gb);
-            h_sampling_factor_n= get_bits(gb, 5);
-            h_sampling_factor_m= get_bits(gb, 5);
-            v_sampling_factor_n= get_bits(gb, 5);
-            v_sampling_factor_m= get_bits(gb, 5);
-            s->enhancement_type= get_bits1(gb);
-
-            if(   h_sampling_factor_n==0 || h_sampling_factor_m==0
-               || v_sampling_factor_n==0 || v_sampling_factor_m==0){
-
-//                fprintf(stderr, "illegal scalability header (VERY broken encoder), trying to workaround\n");
-                s->scalability=0;
-
-                *gb= bak;
-            }else
-                av_log(s->avctx, AV_LOG_ERROR, "scalability not supported\n");
-
-            // bin shape stuff FIXME
-        }
-    }
-    return 0;
-}
-
-/**
- * decodes the user data stuff in the header.
- * allso inits divx/xvid/lavc_version/build
- */
-static int decode_user_data(MpegEncContext *s, GetBitContext *gb){
-    char buf[256];
-    int i;
-    int e;
-    int ver = 0, build = 0, ver2 = 0, ver3 = 0;
-    char last;
-
-    for(i=0; i<255 && get_bits_count(gb) < gb->size_in_bits; i++){
-        if(show_bits(gb, 23) == 0) break;
-        buf[i]= get_bits(gb, 8);
-    }
-    buf[i]=0;
-
-    /* divx detection */
-    e=sscanf(buf, "DivX%dBuild%d%c", &ver, &build, &last);
-    if(e<2)
-        e=sscanf(buf, "DivX%db%d%c", &ver, &build, &last);
-    if(e>=2){
-        s->divx_version= ver;
-        s->divx_build= build;
-        s->divx_packed= e==3 && last=='p';
-    }
-
-    /* ffmpeg detection */
-    e=sscanf(buf, "FFmpe%*[^b]b%d", &build)+3;
-    if(e!=4)
-        e=sscanf(buf, "FFmpeg v%d.%d.%d / libavcodec build: %d", &ver, &ver2, &ver3, &build);
-    if(e!=4){
-        e=sscanf(buf, "Lavc%d.%d.%d", &ver, &ver2, &ver3)+1;
-        if (e>1)
-            build= (ver<<16) + (ver2<<8) + ver3;
-    }
-    if(e!=4){
-        if(strcmp(buf, "ffmpeg")==0){
-            s->lavc_build= 4600;
-        }
-    }
-    if(e==4){
-        s->lavc_build= build;
-    }
-
-    /* xvid detection */
-    e=sscanf(buf, "XviD%d", &build);
-    if(e==1){
-        s->xvid_build= build;
-    }
-
-//printf("User Data: %s\n", buf);
-    return 0;
-}
-
-static int decode_vop_header(MpegEncContext *s, GetBitContext *gb){
-    int time_incr, time_increment;
-
-    s->pict_type = get_bits(gb, 2) + I_TYPE;        /* pict type: I = 0 , P = 1 */
-    if(s->pict_type==B_TYPE && s->low_delay && s->vol_control_parameters==0 && !(s->flags & CODEC_FLAG_LOW_DELAY)){
-        av_log(s->avctx, AV_LOG_ERROR, "low_delay flag incorrectly, clearing it\n");
-        s->low_delay=0;
-    }
-
-    s->partitioned_frame= s->data_partitioning && s->pict_type!=B_TYPE;
-    if(s->partitioned_frame)
-        s->decode_mb= mpeg4_decode_partitioned_mb;
-    else
-        s->decode_mb= ff_mpeg4_decode_mb;
-
-    time_incr=0;
-    while (get_bits1(gb) != 0)
-        time_incr++;
-
-    check_marker(gb, "before time_increment");
-
-    if(s->time_increment_bits==0 || !(show_bits(gb, s->time_increment_bits+1)&1)){
-        av_log(s->avctx, AV_LOG_ERROR, "hmm, seems the headers are not complete, trying to guess time_increment_bits\n");
-
-        for(s->time_increment_bits=1 ;s->time_increment_bits<16; s->time_increment_bits++){
-            if(show_bits(gb, s->time_increment_bits+1)&1) break;
-        }
-
-        av_log(s->avctx, AV_LOG_ERROR, "my guess is %d bits ;)\n",s->time_increment_bits);
-    }
-
-    if(IS_3IV1) time_increment= get_bits1(gb); //FIXME investigate further
-    else time_increment= get_bits(gb, s->time_increment_bits);
-
-//    printf("%d %X\n", s->time_increment_bits, time_increment);
-//av_log(s->avctx, AV_LOG_DEBUG, " type:%d modulo_time_base:%d increment:%d t_frame %d\n", s->pict_type, time_incr, time_increment, s->t_frame);
-    if(s->pict_type!=B_TYPE){
-        s->last_time_base= s->time_base;
-        s->time_base+= time_incr;
-        s->time= s->time_base*s->avctx->time_base.den + time_increment;
-        if(s->workaround_bugs&FF_BUG_UMP4){
-            if(s->time < s->last_non_b_time){
-//                fprintf(stderr, "header is not mpeg4 compatible, broken encoder, trying to workaround\n");
-                s->time_base++;
-                s->time+= s->avctx->time_base.den;
-            }
-        }
-        s->pp_time= s->time - s->last_non_b_time;
-        s->last_non_b_time= s->time;
-    }else{
-        s->time= (s->last_time_base + time_incr)*s->avctx->time_base.den + time_increment;
-        s->pb_time= s->pp_time - (s->last_non_b_time - s->time);
-        if(s->pp_time <=s->pb_time || s->pp_time <= s->pp_time - s->pb_time || s->pp_time<=0){
-//            printf("messed up order, maybe after seeking? skipping current b frame\n");
-            return FRAME_SKIPPED;
-        }
-        ff_mpeg4_init_direct_mv(s);
-
-        if(s->t_frame==0) s->t_frame= s->pb_time;
-        if(s->t_frame==0) s->t_frame=1; // 1/0 protection
-        s->pp_field_time= (  ROUNDED_DIV(s->last_non_b_time, s->t_frame)
-                           - ROUNDED_DIV(s->last_non_b_time - s->pp_time, s->t_frame))*2;
-        s->pb_field_time= (  ROUNDED_DIV(s->time, s->t_frame)
-                           - ROUNDED_DIV(s->last_non_b_time - s->pp_time, s->t_frame))*2;
-    }
-//av_log(s->avctx, AV_LOG_DEBUG, "last nonb %Ld last_base %d time %Ld pp %d pb %d t %d ppf %d pbf %d\n", s->last_non_b_time, s->last_time_base, s->time, s->pp_time, s->pb_time, s->t_frame, s->pp_field_time, s->pb_field_time);
-
-    if(s->avctx->time_base.num)
-        s->current_picture_ptr->pts= (s->time + s->avctx->time_base.num/2) / s->avctx->time_base.num;
-    else
-        s->current_picture_ptr->pts= AV_NOPTS_VALUE;
-    if(s->avctx->debug&FF_DEBUG_PTS)
-        av_log(s->avctx, AV_LOG_DEBUG, "MPEG4 PTS: %"PRId64"\n", s->current_picture_ptr->pts);
-
-    check_marker(gb, "before vop_coded");
-
-    /* vop coded */
-    if (get_bits1(gb) != 1){
-        if(s->avctx->debug&FF_DEBUG_PICT_INFO)
-            av_log(s->avctx, AV_LOG_ERROR, "vop not coded\n");
-        return FRAME_SKIPPED;
-    }
-//printf("time %d %d %d || %Ld %Ld %Ld\n", s->time_increment_bits, s->avctx->time_base.den, s->time_base,
-//s->time, s->last_non_b_time, s->last_non_b_time - s->pp_time);
-    if (s->shape != BIN_ONLY_SHAPE && ( s->pict_type == P_TYPE
-                          || (s->pict_type == S_TYPE && s->vol_sprite_usage==GMC_SPRITE))) {
-        /* rounding type for motion estimation */
-        s->no_rounding = get_bits1(gb);
-    } else {
-        s->no_rounding = 0;
-    }
-//FIXME reduced res stuff
-
-     if (s->shape != RECT_SHAPE) {
-         if (s->vol_sprite_usage != 1 || s->pict_type != I_TYPE) {
-             int width, height, hor_spat_ref, ver_spat_ref;
-
-             width = get_bits(gb, 13);
-             skip_bits1(gb);   /* marker */
-             height = get_bits(gb, 13);
-             skip_bits1(gb);   /* marker */
-             hor_spat_ref = get_bits(gb, 13); /* hor_spat_ref */
-             skip_bits1(gb);   /* marker */
-             ver_spat_ref = get_bits(gb, 13); /* ver_spat_ref */
-         }
-         skip_bits1(gb); /* change_CR_disable */
-
-         if (get_bits1(gb) != 0) {
-             skip_bits(gb, 8); /* constant_alpha_value */
-         }
-     }
-//FIXME complexity estimation stuff
-
-     if (s->shape != BIN_ONLY_SHAPE) {
-         s->intra_dc_threshold= mpeg4_dc_threshold[ get_bits(gb, 3) ];
-         if(!s->progressive_sequence){
-             s->top_field_first= get_bits1(gb);
-             s->alternate_scan= get_bits1(gb);
-         }else
-             s->alternate_scan= 0;
-     }
-
-     if(s->alternate_scan){
-         ff_init_scantable(s->dsp.idct_permutation, &s->inter_scantable  , ff_alternate_vertical_scan);
-         ff_init_scantable(s->dsp.idct_permutation, &s->intra_scantable  , ff_alternate_vertical_scan);
-         ff_init_scantable(s->dsp.idct_permutation, &s->intra_h_scantable, ff_alternate_vertical_scan);
-         ff_init_scantable(s->dsp.idct_permutation, &s->intra_v_scantable, ff_alternate_vertical_scan);
-     } else{
-         ff_init_scantable(s->dsp.idct_permutation, &s->inter_scantable  , ff_zigzag_direct);
-         ff_init_scantable(s->dsp.idct_permutation, &s->intra_scantable  , ff_zigzag_direct);
-         ff_init_scantable(s->dsp.idct_permutation, &s->intra_h_scantable, ff_alternate_horizontal_scan);
-         ff_init_scantable(s->dsp.idct_permutation, &s->intra_v_scantable, ff_alternate_vertical_scan);
-     }
-
-     if(s->pict_type == S_TYPE && (s->vol_sprite_usage==STATIC_SPRITE || s->vol_sprite_usage==GMC_SPRITE)){
-         mpeg4_decode_sprite_trajectory(s, gb);
-         if(s->sprite_brightness_change) av_log(s->avctx, AV_LOG_ERROR, "sprite_brightness_change not supported\n");
-         if(s->vol_sprite_usage==STATIC_SPRITE) av_log(s->avctx, AV_LOG_ERROR, "static sprite not supported\n");
-     }
-
-     if (s->shape != BIN_ONLY_SHAPE) {
-         s->chroma_qscale= s->qscale = get_bits(gb, s->quant_precision);
-         if(s->qscale==0){
-             av_log(s->avctx, AV_LOG_ERROR, "Error, header damaged or not MPEG4 header (qscale=0)\n");
-             return -1; // makes no sense to continue, as there is nothing left from the image then
-         }
-
-         if (s->pict_type != I_TYPE) {
-             s->f_code = get_bits(gb, 3);       /* fcode_for */
-             if(s->f_code==0){
-                 av_log(s->avctx, AV_LOG_ERROR, "Error, header damaged or not MPEG4 header (f_code=0)\n");
-                 return -1; // makes no sense to continue, as the MV decoding will break very quickly
-             }
-         }else
-             s->f_code=1;
-
-         if (s->pict_type == B_TYPE) {
-             s->b_code = get_bits(gb, 3);
-         }else
-             s->b_code=1;
-
-         if(s->avctx->debug&FF_DEBUG_PICT_INFO){
-             av_log(s->avctx, AV_LOG_DEBUG, "qp:%d fc:%d,%d %s size:%d pro:%d alt:%d top:%d %spel part:%d resync:%d w:%d a:%d rnd:%d vot:%d%s dc:%d\n",
-                 s->qscale, s->f_code, s->b_code,
-                 s->pict_type == I_TYPE ? "I" : (s->pict_type == P_TYPE ? "P" : (s->pict_type == B_TYPE ? "B" : "S")),
-                 gb->size_in_bits,s->progressive_sequence, s->alternate_scan, s->top_field_first,
-                 s->quarter_sample ? "q" : "h", s->data_partitioning, s->resync_marker, s->num_sprite_warping_points,
-                 s->sprite_warping_accuracy, 1-s->no_rounding, s->vo_type, s->vol_control_parameters ? " VOLC" : " ", s->intra_dc_threshold);
-         }
-
-         if(!s->scalability){
-             if (s->shape!=RECT_SHAPE && s->pict_type!=I_TYPE) {
-                 skip_bits1(gb); // vop shape coding type
-             }
-         }else{
-             if(s->enhancement_type){
-                 int load_backward_shape= get_bits1(gb);
-                 if(load_backward_shape){
-                     av_log(s->avctx, AV_LOG_ERROR, "load backward shape isn't supported\n");
-                 }
-             }
-             skip_bits(gb, 2); //ref_select_code
-         }
-     }
-     /* detect buggy encoders which don't set the low_delay flag (divx4/xvid/opendivx)*/
-     // note we cannot detect divx5 without b-frames easily (although it's buggy too)
-     if(s->vo_type==0 && s->vol_control_parameters==0 && s->divx_version==0 && s->picture_number==0){
-         av_log(s->avctx, AV_LOG_ERROR, "looks like this file was encoded with (divx4/(old)xvid/opendivx) -> forcing low_delay flag\n");
-         s->low_delay=1;
-     }
-
-     s->picture_number++; // better than pic number==0 always ;)
-
-     s->y_dc_scale_table= ff_mpeg4_y_dc_scale_table; //FIXME add short header support
-     s->c_dc_scale_table= ff_mpeg4_c_dc_scale_table;
-
-     if(s->workaround_bugs&FF_BUG_EDGE){
-         s->h_edge_pos= s->width;
-         s->v_edge_pos= s->height;
-     }
-     return 0;
-}
-
-/**
- * decode mpeg4 headers
- * @return <0 if no VOP found (or a damaged one)
- *         FRAME_SKIPPED if a not coded VOP is found
- *         0 if a VOP is found
- */
-int ff_mpeg4_decode_picture_header(MpegEncContext * s, GetBitContext *gb)
-{
-    int startcode, v;
-
-    /* search next start code */
-    align_get_bits(gb);
-
-    if(s->avctx->codec_tag == ff_get_fourcc("WV1F") && show_bits(gb, 24) == 0x575630){
-        skip_bits(gb, 24);
-        if(get_bits(gb, 8) == 0xF0)
-            return decode_vop_header(s, gb);
-    }
-
-    startcode = 0xff;
-    for(;;) {
-        if(get_bits_count(gb) >= gb->size_in_bits){
-            if(gb->size_in_bits==8 && (s->divx_version || s->xvid_build)){
-                av_log(s->avctx, AV_LOG_ERROR, "frame skip %d\n", gb->size_in_bits);
-                return FRAME_SKIPPED; //divx bug
-            }else
-                return -1; //end of stream
-        }
-
-        /* use the bits after the test */
-        v = get_bits(gb, 8);
-        startcode = ((startcode << 8) | v) & 0xffffffff;
-
-        if((startcode&0xFFFFFF00) != 0x100)
-            continue; //no startcode
-
-        if(s->avctx->debug&FF_DEBUG_STARTCODE){
-            av_log(s->avctx, AV_LOG_DEBUG, "startcode: %3X ", startcode);
-            if     (startcode<=0x11F) av_log(s->avctx, AV_LOG_DEBUG, "Video Object Start");
-            else if(startcode<=0x12F) av_log(s->avctx, AV_LOG_DEBUG, "Video Object Layer Start");
-            else if(startcode<=0x13F) av_log(s->avctx, AV_LOG_DEBUG, "Reserved");
-            else if(startcode<=0x15F) av_log(s->avctx, AV_LOG_DEBUG, "FGS bp start");
-            else if(startcode<=0x1AF) av_log(s->avctx, AV_LOG_DEBUG, "Reserved");
-            else if(startcode==0x1B0) av_log(s->avctx, AV_LOG_DEBUG, "Visual Object Seq Start");
-            else if(startcode==0x1B1) av_log(s->avctx, AV_LOG_DEBUG, "Visual Object Seq End");
-            else if(startcode==0x1B2) av_log(s->avctx, AV_LOG_DEBUG, "User Data");
-            else if(startcode==0x1B3) av_log(s->avctx, AV_LOG_DEBUG, "Group of VOP start");
-            else if(startcode==0x1B4) av_log(s->avctx, AV_LOG_DEBUG, "Video Session Error");
-            else if(startcode==0x1B5) av_log(s->avctx, AV_LOG_DEBUG, "Visual Object Start");
-            else if(startcode==0x1B6) av_log(s->avctx, AV_LOG_DEBUG, "Video Object Plane start");
-            else if(startcode==0x1B7) av_log(s->avctx, AV_LOG_DEBUG, "slice start");
-            else if(startcode==0x1B8) av_log(s->avctx, AV_LOG_DEBUG, "extension start");
-            else if(startcode==0x1B9) av_log(s->avctx, AV_LOG_DEBUG, "fgs start");
-            else if(startcode==0x1BA) av_log(s->avctx, AV_LOG_DEBUG, "FBA Object start");
-            else if(startcode==0x1BB) av_log(s->avctx, AV_LOG_DEBUG, "FBA Object Plane start");
-            else if(startcode==0x1BC) av_log(s->avctx, AV_LOG_DEBUG, "Mesh Object start");
-            else if(startcode==0x1BD) av_log(s->avctx, AV_LOG_DEBUG, "Mesh Object Plane start");
-            else if(startcode==0x1BE) av_log(s->avctx, AV_LOG_DEBUG, "Still Texture Object start");
-            else if(startcode==0x1BF) av_log(s->avctx, AV_LOG_DEBUG, "Texture Spatial Layer start");
-            else if(startcode==0x1C0) av_log(s->avctx, AV_LOG_DEBUG, "Texture SNR Layer start");
-            else if(startcode==0x1C1) av_log(s->avctx, AV_LOG_DEBUG, "Texture Tile start");
-            else if(startcode==0x1C2) av_log(s->avctx, AV_LOG_DEBUG, "Texture Shape Layer start");
-            else if(startcode==0x1C3) av_log(s->avctx, AV_LOG_DEBUG, "stuffing start");
-            else if(startcode<=0x1C5) av_log(s->avctx, AV_LOG_DEBUG, "reserved");
-            else if(startcode<=0x1FF) av_log(s->avctx, AV_LOG_DEBUG, "System start");
-            av_log(s->avctx, AV_LOG_DEBUG, " at %d\n", get_bits_count(gb));
-        }
-
-        if(startcode >= 0x120 && startcode <= 0x12F){
-            if(decode_vol_header(s, gb) < 0)
-                return -1;
-        }
-        else if(startcode == USER_DATA_STARTCODE){
-            decode_user_data(s, gb);
-        }
-        else if(startcode == GOP_STARTCODE){
-            mpeg4_decode_gop_header(s, gb);
-        }
-        else if(startcode == VOP_STARTCODE){
-            return decode_vop_header(s, gb);
-        }
-
-        align_get_bits(gb);
-        startcode = 0xff;
-    }
-}
-
-/* don't understand why they choose a different header ! */
-int intel_h263_decode_picture_header(MpegEncContext *s)
-{
-    int format;
-
-    /* picture header */
-    if (get_bits_long(&s->gb, 22) != 0x20) {
-        av_log(s->avctx, AV_LOG_ERROR, "Bad picture start code\n");
-        return -1;
-    }
-    s->picture_number = get_bits(&s->gb, 8); /* picture timestamp */
-
-    if (get_bits1(&s->gb) != 1) {
-        av_log(s->avctx, AV_LOG_ERROR, "Bad marker\n");
-        return -1;      /* marker */
-    }
-    if (get_bits1(&s->gb) != 0) {
-        av_log(s->avctx, AV_LOG_ERROR, "Bad H263 id\n");
-        return -1;      /* h263 id */
-    }
-    skip_bits1(&s->gb);         /* split screen off */
-    skip_bits1(&s->gb);         /* camera  off */
-    skip_bits1(&s->gb);         /* freeze picture release off */
-
-    format = get_bits(&s->gb, 3);
-    if (format != 7) {
-        av_log(s->avctx, AV_LOG_ERROR, "Intel H263 free format not supported\n");
-        return -1;
-    }
-    s->h263_plus = 0;
-
-    s->pict_type = I_TYPE + get_bits1(&s->gb);
-
-    s->unrestricted_mv = get_bits1(&s->gb);
-    s->h263_long_vectors = s->unrestricted_mv;
-
-    if (get_bits1(&s->gb) != 0) {
-        av_log(s->avctx, AV_LOG_ERROR, "SAC not supported\n");
-        return -1;      /* SAC: off */
-    }
-    if (get_bits1(&s->gb) != 0) {
-        s->obmc= 1;
-        av_log(s->avctx, AV_LOG_ERROR, "Advanced Prediction Mode not supported\n");
-//        return -1;      /* advanced prediction mode: off */
-    }
-    if (get_bits1(&s->gb) != 0) {
-        av_log(s->avctx, AV_LOG_ERROR, "PB frame mode no supported\n");
-        return -1;      /* PB frame mode */
-    }
-
-    /* skip unknown header garbage */
-    skip_bits(&s->gb, 41);
-
-    s->chroma_qscale= s->qscale = get_bits(&s->gb, 5);
-    skip_bits1(&s->gb); /* Continuous Presence Multipoint mode: off */
-
-    /* PEI */
-    while (get_bits1(&s->gb) != 0) {
-        skip_bits(&s->gb, 8);
-    }
-    s->f_code = 1;
-
-    s->y_dc_scale_table=
-    s->c_dc_scale_table= ff_mpeg1_dc_scale_table;
-
-    return 0;
-}
-
-int flv_h263_decode_picture_header(MpegEncContext *s)
-{
-    int format, width, height;
-
-    /* picture header */
-    if (get_bits_long(&s->gb, 17) != 1) {
-        av_log(s->avctx, AV_LOG_ERROR, "Bad picture start code\n");
-        return -1;
-    }
-    format = get_bits(&s->gb, 5);
-    if (format != 0 && format != 1) {
-        av_log(s->avctx, AV_LOG_ERROR, "Bad picture format\n");
-        return -1;
-    }
-    s->h263_flv = format+1;
-    s->picture_number = get_bits(&s->gb, 8); /* picture timestamp */
-    format = get_bits(&s->gb, 3);
-    switch (format) {
-    case 0:
-        width = get_bits(&s->gb, 8);
-        height = get_bits(&s->gb, 8);
-        break;
-    case 1:
-        width = get_bits(&s->gb, 16);
-        height = get_bits(&s->gb, 16);
-        break;
-    case 2:
-        width = 352;
-        height = 288;
-        break;
-    case 3:
-        width = 176;
-        height = 144;
-        break;
-    case 4:
-        width = 128;
-        height = 96;
-        break;
-    case 5:
-        width = 320;
-        height = 240;
-        break;
-    case 6:
-        width = 160;
-        height = 120;
-        break;
-    default:
-        width = height = 0;
-        break;
-    }
-    if(avcodec_check_dimensions(s->avctx, width, height))
-        return -1;
-    s->width = width;
-    s->height = height;
-
-    s->pict_type = I_TYPE + get_bits(&s->gb, 2);
-    s->dropable= s->pict_type > P_TYPE;
-    if (s->dropable)
-        s->pict_type = P_TYPE;
-
-    skip_bits1(&s->gb); /* deblocking flag */
-    s->chroma_qscale= s->qscale = get_bits(&s->gb, 5);
-
-    s->h263_plus = 0;
-
-    s->unrestricted_mv = 1;
-    s->h263_long_vectors = 0;
-
-    /* PEI */
-    while (get_bits1(&s->gb) != 0) {
-        skip_bits(&s->gb, 8);
-    }
-    s->f_code = 1;
-
-    if(s->avctx->debug & FF_DEBUG_PICT_INFO){
-        av_log(s->avctx, AV_LOG_DEBUG, "%c esc_type:%d, qp:%d num:%d\n",
-               s->dropable ? 'D' : av_get_pict_type_char(s->pict_type), s->h263_flv-1, s->qscale, s->picture_number);
-    }
-
-    s->y_dc_scale_table=
-    s->c_dc_scale_table= ff_mpeg1_dc_scale_table;
-
-    return 0;
-}
--- a/src/ffmpeg/libavcodec/h263data.h	Mon Mar 12 11:37:35 2007 -0700
+++ /dev/null	Thu Jan 01 00:00:00 1970 +0000
@@ -1,308 +0,0 @@
-/*
- * copyright (c) 2000,2001 Fabrice Bellard
- * H263+ support
- * copyright (c) 2001 Juan J. Sierralta P.
- * copyright (c) 2002-2004 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 h263data.h
- * H.263 tables.
- */
-
-
-/* intra MCBPC, mb_type = (intra), then (intraq) */
-const uint8_t intra_MCBPC_code[9] = { 1, 1, 2, 3, 1, 1, 2, 3, 1 };
-const uint8_t intra_MCBPC_bits[9] = { 1, 3, 3, 3, 4, 6, 6, 6, 9 };
-
-/* inter MCBPC, mb_type = (inter), (intra), (interq), (intraq), (inter4v) */
-/* Changed the tables for interq and inter4v+q, following the standard ** Juanjo ** */
-const uint8_t inter_MCBPC_code[28] = {
-    1, 3, 2, 5,
-    3, 4, 3, 3,
-    3, 7, 6, 5,
-    4, 4, 3, 2,
-    2, 5, 4, 5,
-    1, 0, 0, 0, /* Stuffing */
-    2, 12, 14, 15,
-};
-const uint8_t inter_MCBPC_bits[28] = {
-    1, 4, 4, 6, /* inter  */
-    5, 8, 8, 7, /* intra  */
-    3, 7, 7, 9, /* interQ */
-    6, 9, 9, 9, /* intraQ */
-    3, 7, 7, 8, /* inter4 */
-    9, 0, 0, 0, /* Stuffing */
-    11, 13, 13, 13,/* inter4Q*/
-};
-
-static const uint8_t h263_mbtype_b_tab[15][2] = {
- {1, 1},
- {3, 3},
- {1, 5},
- {4, 4},
- {5, 4},
- {6, 6},
- {2, 4},
- {3, 4},
- {7, 6},
- {4, 6},
- {5, 6},
- {1, 6},
- {1,10},
- {1, 7},
- {1, 8},
-};
-
-static const int h263_mb_type_b_map[15]= {
-    MB_TYPE_DIRECT2 | MB_TYPE_L0L1,
-    MB_TYPE_DIRECT2 | MB_TYPE_L0L1 | MB_TYPE_CBP,
-    MB_TYPE_DIRECT2 | MB_TYPE_L0L1 | MB_TYPE_CBP | MB_TYPE_QUANT,
-                      MB_TYPE_L0                                 | MB_TYPE_16x16,
-                      MB_TYPE_L0   | MB_TYPE_CBP                 | MB_TYPE_16x16,
-                      MB_TYPE_L0   | MB_TYPE_CBP | MB_TYPE_QUANT | MB_TYPE_16x16,
-                      MB_TYPE_L1                                 | MB_TYPE_16x16,
-                      MB_TYPE_L1   | MB_TYPE_CBP                 | MB_TYPE_16x16,
-                      MB_TYPE_L1   | MB_TYPE_CBP | MB_TYPE_QUANT | MB_TYPE_16x16,
-                      MB_TYPE_L0L1                               | MB_TYPE_16x16,
-                      MB_TYPE_L0L1 | MB_TYPE_CBP                 | MB_TYPE_16x16,
-                      MB_TYPE_L0L1 | MB_TYPE_CBP | MB_TYPE_QUANT | MB_TYPE_16x16,
-    0, //stuffing
-    MB_TYPE_INTRA4x4                | MB_TYPE_CBP,
-    MB_TYPE_INTRA4x4                | MB_TYPE_CBP | MB_TYPE_QUANT,
-};
-
-const uint8_t cbpc_b_tab[4][2] = {
-{0, 1},
-{2, 2},
-{7, 3},
-{6, 3},
-};
-
-const uint8_t cbpy_tab[16][2] =
-{
-  {3,4}, {5,5}, {4,5}, {9,4}, {3,5}, {7,4}, {2,6}, {11,4},
-  {2,5}, {3,6}, {5,4}, {10,4}, {4,4}, {8,4}, {6,4}, {3,2}
-};
-
-const uint8_t mvtab[33][2] =
-{
-  {1,1}, {1,2}, {1,3}, {1,4}, {3,6}, {5,7}, {4,7}, {3,7},
-  {11,9}, {10,9}, {9,9}, {17,10}, {16,10}, {15,10}, {14,10}, {13,10},
-  {12,10}, {11,10}, {10,10}, {9,10}, {8,10}, {7,10}, {6,10}, {5,10},
-  {4,10}, {7,11}, {6,11}, {5,11}, {4,11}, {3,11}, {2,11}, {3,12},
-  {2,12}
-};
-
-/* third non intra table */
-const uint16_t inter_vlc[103][2] = {
-{ 0x2, 2 },{ 0xf, 4 },{ 0x15, 6 },{ 0x17, 7 },
-{ 0x1f, 8 },{ 0x25, 9 },{ 0x24, 9 },{ 0x21, 10 },
-{ 0x20, 10 },{ 0x7, 11 },{ 0x6, 11 },{ 0x20, 11 },
-{ 0x6, 3 },{ 0x14, 6 },{ 0x1e, 8 },{ 0xf, 10 },
-{ 0x21, 11 },{ 0x50, 12 },{ 0xe, 4 },{ 0x1d, 8 },
-{ 0xe, 10 },{ 0x51, 12 },{ 0xd, 5 },{ 0x23, 9 },
-{ 0xd, 10 },{ 0xc, 5 },{ 0x22, 9 },{ 0x52, 12 },
-{ 0xb, 5 },{ 0xc, 10 },{ 0x53, 12 },{ 0x13, 6 },
-{ 0xb, 10 },{ 0x54, 12 },{ 0x12, 6 },{ 0xa, 10 },
-{ 0x11, 6 },{ 0x9, 10 },{ 0x10, 6 },{ 0x8, 10 },
-{ 0x16, 7 },{ 0x55, 12 },{ 0x15, 7 },{ 0x14, 7 },
-{ 0x1c, 8 },{ 0x1b, 8 },{ 0x21, 9 },{ 0x20, 9 },
-{ 0x1f, 9 },{ 0x1e, 9 },{ 0x1d, 9 },{ 0x1c, 9 },
-{ 0x1b, 9 },{ 0x1a, 9 },{ 0x22, 11 },{ 0x23, 11 },
-{ 0x56, 12 },{ 0x57, 12 },{ 0x7, 4 },{ 0x19, 9 },
-{ 0x5, 11 },{ 0xf, 6 },{ 0x4, 11 },{ 0xe, 6 },
-{ 0xd, 6 },{ 0xc, 6 },{ 0x13, 7 },{ 0x12, 7 },
-{ 0x11, 7 },{ 0x10, 7 },{ 0x1a, 8 },{ 0x19, 8 },
-{ 0x18, 8 },{ 0x17, 8 },{ 0x16, 8 },{ 0x15, 8 },
-{ 0x14, 8 },{ 0x13, 8 },{ 0x18, 9 },{ 0x17, 9 },
-{ 0x16, 9 },{ 0x15, 9 },{ 0x14, 9 },{ 0x13, 9 },
-{ 0x12, 9 },{ 0x11, 9 },{ 0x7, 10 },{ 0x6, 10 },
-{ 0x5, 10 },{ 0x4, 10 },{ 0x24, 11 },{ 0x25, 11 },
-{ 0x26, 11 },{ 0x27, 11 },{ 0x58, 12 },{ 0x59, 12 },
-{ 0x5a, 12 },{ 0x5b, 12 },{ 0x5c, 12 },{ 0x5d, 12 },
-{ 0x5e, 12 },{ 0x5f, 12 },{ 0x3, 7 },
-};
-
-const int8_t inter_level[102] = {
-  1,  2,  3,  4,  5,  6,  7,  8,
-  9, 10, 11, 12,  1,  2,  3,  4,
-  5,  6,  1,  2,  3,  4,  1,  2,
-  3,  1,  2,  3,  1,  2,  3,  1,
-  2,  3,  1,  2,  1,  2,  1,  2,
-  1,  2,  1,  1,  1,  1,  1,  1,
-  1,  1,  1,  1,  1,  1,  1,  1,
-  1,  1,  1,  2,  3,  1,  2,  1,
-  1,  1,  1,  1,  1,  1,  1,  1,
-  1,  1,  1,  1,  1,  1,  1,  1,
-  1,  1,  1,  1,  1,  1,  1,  1,
-  1,  1,  1,  1,  1,  1,  1,  1,
-  1,  1,  1,  1,  1,  1,
-};
-
-const int8_t inter_run[102] = {
-  0,  0,  0,  0,  0,  0,  0,  0,
-  0,  0,  0,  0,  1,  1,  1,  1,
-  1,  1,  2,  2,  2,  2,  3,  3,
-  3,  4,  4,  4,  5,  5,  5,  6,
-  6,  6,  7,  7,  8,  8,  9,  9,
- 10, 10, 11, 12, 13, 14, 15, 16,
- 17, 18, 19, 20, 21, 22, 23, 24,
- 25, 26,  0,  0,  0,  1,  1,  2,
-  3,  4,  5,  6,  7,  8,  9, 10,
- 11, 12, 13, 14, 15, 16, 17, 18,
- 19, 20, 21, 22, 23, 24, 25, 26,
- 27, 28, 29, 30, 31, 32, 33, 34,
- 35, 36, 37, 38, 39, 40,
-};
-
-static RLTable rl_inter = {
-    102,
-    58,
-    inter_vlc,
-    inter_run,
-    inter_level,
-};
-
-const uint16_t intra_vlc_aic[103][2] = {
-{  0x2,  2 }, {  0x6,  3 }, {  0xe,  4 }, {  0xc,  5 },
-{  0xd,  5 }, { 0x10,  6 }, { 0x11,  6 }, { 0x12,  6 },
-{ 0x16,  7 }, { 0x1b,  8 }, { 0x20,  9 }, { 0x21,  9 },
-{ 0x1a,  9 }, { 0x1b,  9 }, { 0x1c,  9 }, { 0x1d,  9 },
-{ 0x1e,  9 }, { 0x1f,  9 }, { 0x23, 11 }, { 0x22, 11 },
-{ 0x57, 12 }, { 0x56, 12 }, { 0x55, 12 }, { 0x54, 12 },
-{ 0x53, 12 }, {  0xf,  4 }, { 0x14,  6 }, { 0x14,  7 },
-{ 0x1e,  8 }, {  0xf, 10 }, { 0x21, 11 }, { 0x50, 12 },
-{  0xb,  5 }, { 0x15,  7 }, {  0xe, 10 }, {  0x9, 10 },
-{ 0x15,  6 }, { 0x1d,  8 }, {  0xd, 10 }, { 0x51, 12 },
-{ 0x13,  6 }, { 0x23,  9 }, {  0x7, 11 }, { 0x17,  7 },
-{ 0x22,  9 }, { 0x52, 12 }, { 0x1c,  8 }, {  0xc, 10 },
-{ 0x1f,  8 }, {  0xb, 10 }, { 0x25,  9 }, {  0xa, 10 },
-{ 0x24,  9 }, {  0x6, 11 }, { 0x21, 10 }, { 0x20, 10 },
-{  0x8, 10 }, { 0x20, 11 }, {  0x7,  4 }, {  0xc,  6 },
-{ 0x10,  7 }, { 0x13,  8 }, { 0x11,  9 }, { 0x12,  9 },
-{  0x4, 10 }, { 0x27, 11 }, { 0x26, 11 }, { 0x5f, 12 },
-{  0xf,  6 }, { 0x13,  9 }, {  0x5, 10 }, { 0x25, 11 },
-{  0xe,  6 }, { 0x14,  9 }, { 0x24, 11 }, {  0xd,  6 },
-{  0x6, 10 }, { 0x5e, 12 }, { 0x11,  7 }, {  0x7, 10 },
-{ 0x13,  7 }, { 0x5d, 12 }, { 0x12,  7 }, { 0x5c, 12 },
-{ 0x14,  8 }, { 0x5b, 12 }, { 0x15,  8 }, { 0x1a,  8 },
-{ 0x19,  8 }, { 0x18,  8 }, { 0x17,  8 }, { 0x16,  8 },
-{ 0x19,  9 }, { 0x15,  9 }, { 0x16,  9 }, { 0x18,  9 },
-{ 0x17,  9 }, {  0x4, 11 }, {  0x5, 11 }, { 0x58, 12 },
-{ 0x59, 12 }, { 0x5a, 12 }, {  0x3,  7 },
-};
-
-const int8_t intra_run_aic[102] = {
- 0,  0,  0,  0,  0,  0,  0,  0,
- 0,  0,  0,  0,  0,  0,  0,  0,
- 0,  0,  0,  0,  0,  0,  0,  0,
- 0,  1,  1,  1,  1,  1,  1,  1,
- 2,  2,  2,  2,  3,  3,  3,  3,
- 4,  4,  4,  5,  5,  5,  6,  6,
- 7,  7,  8,  8,  9,  9, 10, 11,
-12, 13,  0,  0,  0,  0,  0,  0,
- 0,  0,  0,  0,  1,  1,  1,  1,
- 2,  2,  2,  3,  3,  3,  4,  4,
- 5,  5,  6,  6,  7,  7,  8,  9,
-10, 11, 12, 13, 14, 15, 16, 17,
-18, 19, 20, 21, 22, 23,
-};
-
-const int8_t intra_level_aic[102] = {
- 1,  2,  3,  4,  5,  6,  7,  8,
- 9, 10, 11, 12, 13, 14, 15, 16,
-17, 18, 19, 20, 21, 22, 23, 24,
-25,  1,  2,  3,  4,  5,  6,  7,
- 1,  2,  3,  4,  1,  2,  3,  4,
- 1,  2,  3,  1,  2,  3,  1,  2,
- 1,  2,  1,  2,  1,  2,  1,  1,
- 1,  1,  1,  2,  3,  4,  5,  6,
- 7,  8,  9, 10,  1,  2,  3,  4,
- 1,  2,  3,  1,  2,  3,  1,  2,
- 1,  2,  1,  2,  1,  2,  1,  1,
- 1,  1,  1,  1,  1,  1,  1,  1,
- 1,  1,  1,  1,  1,  1,
-};
-
-static RLTable rl_intra_aic = {
-    102,
-    58,
-    intra_vlc_aic,
-    intra_run_aic,
-    intra_level_aic,
-};
-
-static const uint8_t wrong_run[102] = {
- 1,  2,  3,  5,  4, 10,  9,  8,
-11, 15, 17, 16, 23, 22, 21, 20,
-19, 18, 25, 24, 27, 26, 11,  7,
- 6,  1,  2, 13,  2,  2,  2,  2,
- 6, 12,  3,  9,  1,  3,  4,  3,
- 7,  4,  1,  1,  5,  5, 14,  6,
- 1,  7,  1,  8,  1,  1,  1,  1,
-10,  1,  1,  5,  9, 17, 25, 24,
-29, 33, 32, 41,  2, 23, 28, 31,
- 3, 22, 30,  4, 27, 40,  8, 26,
- 6, 39,  7, 38, 16, 37, 15, 10,
-11, 12, 13, 14,  1, 21, 20, 18,
-19,  2,  1, 34, 35, 36
-};
-
-static const uint16_t h263_format[8][2] = {
-    { 0, 0 },
-    { 128, 96 },
-    { 176, 144 },
-    { 352, 288 },
-    { 704, 576 },
-    { 1408, 1152 },
-};
-
-const uint8_t ff_aic_dc_scale_table[32]={
-//  0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
-    0, 2, 4, 6, 8,10,12,14,16,18,20,22,24,26,28,30,32,34,36,38,40,42,44,46,48,50,52,54,56,58,60,62
-};
-
-static const uint8_t modified_quant_tab[2][32]={
-//  0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
-{
-    0, 3, 1, 2, 3, 4, 5, 6, 7, 8, 9, 9,10,11,12,13,14,15,16,17,18,18,19,20,21,22,23,24,25,26,27,28
-},{
-    0, 2, 3, 4, 5, 6, 7, 8, 9,10,11,13,14,15,16,17,18,19,20,21,22,24,25,26,27,28,29,30,31,31,31,26
-}
-};
-
-const uint8_t ff_h263_chroma_qscale_table[32]={
-//  0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
-    0, 1, 2, 3, 4, 5, 6, 6, 7, 8, 9, 9,10,10,11,11,12,12,12,13,13,13,14,14,14,14,14,15,15,15,15,15
-};
-
-const uint16_t ff_mba_max[6]={
-     47,  98, 395,1583,6335,9215
-};
-
-const uint8_t ff_mba_length[7]={
-      6,   7,   9,  11,  13,  14,  14
-};
-
-const uint8_t ff_h263_loop_filter_strength[32]={
-//  0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
-    0, 1, 1, 2, 2, 3, 3, 4, 4, 4, 5, 5, 6, 6, 7, 7, 7, 8, 8, 8, 9, 9, 9,10,10,10,11,11,11,12,12,12
-};
-
--- a/src/ffmpeg/libavcodec/h263dec.c	Mon Mar 12 11:37:35 2007 -0700
+++ /dev/null	Thu Jan 01 00:00:00 1970 +0000
@@ -1,913 +0,0 @@
-/*
- * H.263 decoder
- * Copyright (c) 2001 Fabrice Bellard.
- * Copyright (c) 2002-2004 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 h263dec.c
- * H.263 decoder.
- */
-
-#include "avcodec.h"
-#include "dsputil.h"
-#include "mpegvideo.h"
-
-//#define DEBUG
-//#define PRINT_FRAME_TIME
-
-int ff_h263_decode_init(AVCodecContext *avctx)
-{
-    MpegEncContext *s = avctx->priv_data;
-
-    s->avctx = avctx;
-    s->out_format = FMT_H263;
-
-    s->width  = avctx->coded_width;
-    s->height = avctx->coded_height;
-    s->workaround_bugs= avctx->workaround_bugs;
-
-    // set defaults
-    MPV_decode_defaults(s);
-    s->quant_precision=5;
-    s->decode_mb= ff_h263_decode_mb;
-    s->low_delay= 1;
-    avctx->pix_fmt= PIX_FMT_YUV420P;
-    s->unrestricted_mv= 1;
-
-    /* select sub codec */
-    switch(avctx->codec->id) {
-    case CODEC_ID_H263:
-        s->unrestricted_mv= 0;
-        break;
-    case CODEC_ID_MPEG4:
-        s->decode_mb= ff_mpeg4_decode_mb;
-        s->time_increment_bits = 4; /* default value for broken headers */
-        s->h263_pred = 1;
-        s->low_delay = 0; //default, might be overriden in the vol header during header parsing
-        break;
-    case CODEC_ID_MSMPEG4V1:
-        s->h263_msmpeg4 = 1;
-        s->h263_pred = 1;
-        s->msmpeg4_version=1;
-        break;
-    case CODEC_ID_MSMPEG4V2:
-        s->h263_msmpeg4 = 1;
-        s->h263_pred = 1;
-        s->msmpeg4_version=2;
-        break;
-    case CODEC_ID_MSMPEG4V3:
-        s->h263_msmpeg4 = 1;
-        s->h263_pred = 1;
-        s->msmpeg4_version=3;
-        break;
-    case CODEC_ID_WMV1:
-        s->h263_msmpeg4 = 1;
-        s->h263_pred = 1;
-        s->msmpeg4_version=4;
-        break;
-    case CODEC_ID_WMV2:
-        s->h263_msmpeg4 = 1;
-        s->h263_pred = 1;
-        s->msmpeg4_version=5;
-        break;
-    case CODEC_ID_VC1:
-    case CODEC_ID_WMV3:
-        s->h263_msmpeg4 = 1;
-        s->h263_pred = 1;
-        s->msmpeg4_version=6;
-        break;
-    case CODEC_ID_H263I:
-        break;
-    case CODEC_ID_FLV1:
-        s->h263_flv = 1;
-        break;
-    default:
-        return -1;
-    }
-    s->codec_id= avctx->codec->id;
-
-    /* for h263, we allocate the images after having read the header */
-    if (avctx->codec->id != CODEC_ID_H263 && avctx->codec->id != CODEC_ID_MPEG4)
-        if (MPV_common_init(s) < 0)
-            return -1;
-
-    if (s->h263_msmpeg4)
-        ff_msmpeg4_decode_init(s);
-    else
-        h263_decode_init_vlc(s);
-
-    return 0;
-}
-
-int ff_h263_decode_end(AVCodecContext *avctx)
-{
-    MpegEncContext *s = avctx->priv_data;
-
-    MPV_common_end(s);
-    return 0;
-}
-
-/**
- * returns the number of bytes consumed for building the current frame
- */
-static int get_consumed_bytes(MpegEncContext *s, int buf_size){
-    int pos= (get_bits_count(&s->gb)+7)>>3;
-
-    if(s->divx_packed){
-        //we would have to scan through the whole buf to handle the weird reordering ...
-        return buf_size;
-    }else if(s->flags&CODEC_FLAG_TRUNCATED){
-        pos -= s->parse_context.last_index;
-        if(pos<0) pos=0; // padding is not really read so this might be -1
-        return pos;
-    }else{
-        if(pos==0) pos=1; //avoid infinite loops (i doubt thats needed but ...)
-        if(pos+10>buf_size) pos=buf_size; // oops ;)
-
-        return pos;
-    }
-}
-
-static int decode_slice(MpegEncContext *s){
-    const int part_mask= s->partitioned_frame ? (AC_END|AC_ERROR) : 0x7F;
-    const int mb_size= 16>>s->avctx->lowres;
-    s->last_resync_gb= s->gb;
-    s->first_slice_line= 1;
-
-    s->resync_mb_x= s->mb_x;
-    s->resync_mb_y= s->mb_y;
-
-    ff_set_qscale(s, s->qscale);
-
-    if(s->partitioned_frame){
-        const int qscale= s->qscale;
-
-        if(s->codec_id==CODEC_ID_MPEG4){
-            if(ff_mpeg4_decode_partitions(s) < 0)
-                return -1;
-        }
-
-        /* restore variables which were modified */
-        s->first_slice_line=1;
-        s->mb_x= s->resync_mb_x;
-        s->mb_y= s->resync_mb_y;
-        ff_set_qscale(s, qscale);
-    }
-
-    for(; s->mb_y < s->mb_height; s->mb_y++) {
-        /* per-row end of slice checks */
-        if(s->msmpeg4_version){
-            if(s->resync_mb_y + s->slice_height == s->mb_y){
-                ff_er_add_slice(s, s->resync_mb_x, s->resync_mb_y, s->mb_x-1, s->mb_y, AC_END|DC_END|MV_END);
-
-                return 0;
-            }
-        }
-
-        if(s->msmpeg4_version==1){
-            s->last_dc[0]=
-            s->last_dc[1]=
-            s->last_dc[2]= 128;
-        }
-
-        ff_init_block_index(s);
-        for(; s->mb_x < s->mb_width; s->mb_x++) {
-            int ret;
-
-            ff_update_block_index(s);
-
-            if(s->resync_mb_x == s->mb_x && s->resync_mb_y+1 == s->mb_y){
-                s->first_slice_line=0;
-            }
-
-            /* DCT & quantize */
-
-            s->mv_dir = MV_DIR_FORWARD;
-            s->mv_type = MV_TYPE_16X16;
-//            s->mb_skipped = 0;
-//printf("%d %d %06X\n", ret, get_bits_count(&s->gb), show_bits(&s->gb, 24));
-            ret= s->decode_mb(s, s->block);
-
-            if (s->pict_type!=B_TYPE)
-                ff_h263_update_motion_val(s);
-
-            if(ret<0){
-                const int xy= s->mb_x + s->mb_y*s->mb_stride;
-                if(ret==SLICE_END){
-                    MPV_decode_mb(s, s->block);
-                    if(s->loop_filter)
-                        ff_h263_loop_filter(s);
-
-//printf("%d %d %d %06X\n", s->mb_x, s->mb_y, s->gb.size*8 - get_bits_count(&s->gb), show_bits(&s->gb, 24));
-                    ff_er_add_slice(s, s->resync_mb_x, s->resync_mb_y, s->mb_x, s->mb_y, (AC_END|DC_END|MV_END)&part_mask);
-
-                    s->padding_bug_score--;
-
-                    if(++s->mb_x >= s->mb_width){
-                        s->mb_x=0;
-                        ff_draw_horiz_band(s, s->mb_y*mb_size, mb_size);
-                        s->mb_y++;
-                    }
-                    return 0;
-                }else if(ret==SLICE_NOEND){
-                    av_log(s->avctx, AV_LOG_ERROR, "Slice mismatch at MB: %d\n", xy);
-                    ff_er_add_slice(s, s->resync_mb_x, s->resync_mb_y, s->mb_x+1, s->mb_y, (AC_END|DC_END|MV_END)&part_mask);
-                    return -1;
-                }
-                av_log(s->avctx, AV_LOG_ERROR, "Error at MB: %d\n", xy);
-                ff_er_add_slice(s, s->resync_mb_x, s->resync_mb_y, s->mb_x, s->mb_y, (AC_ERROR|DC_ERROR|MV_ERROR)&part_mask);
-
-                return -1;
-            }
-
-            MPV_decode_mb(s, s->block);
-            if(s->loop_filter)
-                ff_h263_loop_filter(s);
-        }
-
-        ff_draw_horiz_band(s, s->mb_y*mb_size, mb_size);
-
-        s->mb_x= 0;
-    }
-
-    assert(s->mb_x==0 && s->mb_y==s->mb_height);
-
-    /* try to detect the padding bug */
-    if(      s->codec_id==CODEC_ID_MPEG4
-       &&   (s->workaround_bugs&FF_BUG_AUTODETECT)
-       &&    s->gb.size_in_bits - get_bits_count(&s->gb) >=0
-       &&    s->gb.size_in_bits - get_bits_count(&s->gb) < 48
-//       &&   !s->resync_marker
-       &&   !s->data_partitioning){
-
-        const int bits_count= get_bits_count(&s->gb);
-        const int bits_left = s->gb.size_in_bits - bits_count;
-
-        if(bits_left==0){
-            s->padding_bug_score+=16;
-        } else if(bits_left != 1){
-            int v= show_bits(&s->gb, 8);
-            v|= 0x7F >> (7-(bits_count&7));
-
-            if(v==0x7F && bits_left<=8)
-                s->padding_bug_score--;
-            else if(v==0x7F && ((get_bits_count(&s->gb)+8)&8) && bits_left<=16)
-                s->padding_bug_score+= 4;
-            else
-                s->padding_bug_score++;
-        }
-    }
-
-    if(s->workaround_bugs&FF_BUG_AUTODETECT){
-        if(s->padding_bug_score > -2 && !s->data_partitioning /*&& (s->divx_version || !s->resync_marker)*/)
-            s->workaround_bugs |=  FF_BUG_NO_PADDING;
-        else
-            s->workaround_bugs &= ~FF_BUG_NO_PADDING;
-    }
-
-    // handle formats which don't have unique end markers
-    if(s->msmpeg4_version || (s->workaround_bugs&FF_BUG_NO_PADDING)){ //FIXME perhaps solve this more cleanly
-        int left= s->gb.size_in_bits - get_bits_count(&s->gb);
-        int max_extra=7;
-
-        /* no markers in M$ crap */
-        if(s->msmpeg4_version && s->pict_type==I_TYPE)
-            max_extra+= 17;
-
-        /* buggy padding but the frame should still end approximately at the bitstream end */
-        if((s->workaround_bugs&FF_BUG_NO_PADDING) && s->error_resilience>=3)
-            max_extra+= 48;
-        else if((s->workaround_bugs&FF_BUG_NO_PADDING))
-            max_extra+= 256*256*256*64;
-
-        if(left>max_extra){
-            av_log(s->avctx, AV_LOG_ERROR, "discarding %d junk bits at end, next would be %X\n", left, show_bits(&s->gb, 24));
-        }
-        else if(left<0){
-            av_log(s->avctx, AV_LOG_ERROR, "overreading %d bits\n", -left);
-        }else
-            ff_er_add_slice(s, s->resync_mb_x, s->resync_mb_y, s->mb_x-1, s->mb_y, AC_END|DC_END|MV_END);
-
-        return 0;
-    }
-
-    av_log(s->avctx, AV_LOG_ERROR, "slice end not reached but screenspace end (%d left %06X, score= %d)\n",
-            s->gb.size_in_bits - get_bits_count(&s->gb),
-            show_bits(&s->gb, 24), s->padding_bug_score);
-
-    ff_er_add_slice(s, s->resync_mb_x, s->resync_mb_y, s->mb_x, s->mb_y, (AC_END|DC_END|MV_END)&part_mask);
-
-    return -1;
-}
-
-/**
- * finds the end of the current frame in the bitstream.
- * @return the position of the first byte of the next frame, or -1
- */
-int ff_mpeg4_find_frame_end(ParseContext *pc, const uint8_t *buf, int buf_size){
-    int vop_found, i;
-    uint32_t state;
-
-    vop_found= pc->frame_start_found;
-    state= pc->state;
-
-    i=0;
-    if(!vop_found){
-        for(i=0; i<buf_size; i++){
-            state= (state<<8) | buf[i];
-            if(state == 0x1B6){
-                i++;
-                vop_found=1;
-                break;
-            }
-        }
-    }
-
-    if(vop_found){
-        /* EOF considered as end of frame */
-        if (buf_size == 0)
-            return 0;
-        for(; i<buf_size; i++){
-            state= (state<<8) | buf[i];
-            if((state&0xFFFFFF00) == 0x100){
-                pc->frame_start_found=0;
-                pc->state=-1;
-                return i-3;
-            }
-        }
-    }
-    pc->frame_start_found= vop_found;
-    pc->state= state;
-    return END_NOT_FOUND;
-}
-
-static int h263_find_frame_end(ParseContext *pc, const uint8_t *buf, int buf_size){
-    int vop_found, i;
-    uint32_t state;
-
-    vop_found= pc->frame_start_found;
-    state= pc->state;
-
-    i=0;
-    if(!vop_found){
-        for(i=0; i<buf_size; i++){
-            state= (state<<8) | buf[i];
-            if(state>>(32-22) == 0x20){
-                i++;
-                vop_found=1;
-                break;
-            }
-        }
-    }
-
-    if(vop_found){
-      for(; i<buf_size; i++){
-        state= (state<<8) | buf[i];
-        if(state>>(32-22) == 0x20){
-            pc->frame_start_found=0;
-            pc->state=-1;
-            return i-3;
-        }
-      }
-    }
-    pc->frame_start_found= vop_found;
-    pc->state= state;
-
-    return END_NOT_FOUND;
-}
-
-#ifdef CONFIG_H263_PARSER
-static int h263_parse(AVCodecParserContext *s,
-                           AVCodecContext *avctx,
-                           uint8_t **poutbuf, int *poutbuf_size,
-                           const uint8_t *buf, int buf_size)
-{
-    ParseContext *pc = s->priv_data;
-    int next;
-
-    next= h263_find_frame_end(pc, buf, buf_size);
-
-    if (ff_combine_frame(pc, next, (uint8_t **)&buf, &buf_size) < 0) {
-        *poutbuf = NULL;
-        *poutbuf_size = 0;
-        return buf_size;
-    }
-
-    *poutbuf = (uint8_t *)buf;
-    *poutbuf_size = buf_size;
-    return next;
-}
-#endif
-
-int ff_h263_decode_frame(AVCodecContext *avctx,
-                             void *data, int *data_size,
-                             uint8_t *buf, int buf_size)
-{
-    MpegEncContext *s = avctx->priv_data;
-    int ret;
-    AVFrame *pict = data;
-
-#ifdef PRINT_FRAME_TIME
-uint64_t time= rdtsc();
-#endif
-#ifdef DEBUG
-    av_log(avctx, AV_LOG_DEBUG, "*****frame %d size=%d\n", avctx->frame_number, buf_size);
-    av_log(avctx, AV_LOG_DEBUG, "bytes=%x %x %x %x\n", buf[0], buf[1], buf[2], buf[3]);
-#endif
-    s->flags= avctx->flags;
-    s->flags2= avctx->flags2;
-
-    /* no supplementary picture */
-    if (buf_size == 0) {
-        /* special case for last picture */
-        if (s->low_delay==0 && s->next_picture_ptr) {
-            *pict= *(AVFrame*)s->next_picture_ptr;
-            s->next_picture_ptr= NULL;
-
-            *data_size = sizeof(AVFrame);
-        }
-
-        return 0;
-    }
-
-    if(s->flags&CODEC_FLAG_TRUNCATED){
-        int next;
-
-        if(s->codec_id==CODEC_ID_MPEG4){
-            next= ff_mpeg4_find_frame_end(&s->parse_context, buf, buf_size);
-        }else if(s->codec_id==CODEC_ID_H263){
-            next= h263_find_frame_end(&s->parse_context, buf, buf_size);
-        }else{
-            av_log(s->avctx, AV_LOG_ERROR, "this codec does not support truncated bitstreams\n");
-            return -1;
-        }
-
-        if( ff_combine_frame(&s->parse_context, next, &buf, &buf_size) < 0 )
-            return buf_size;
-    }
-
-
-retry:
-
-    if(s->bitstream_buffer_size && (s->divx_packed || buf_size<20)){ //divx 5.01+/xvid frame reorder
-        init_get_bits(&s->gb, s->bitstream_buffer, s->bitstream_buffer_size*8);
-    }else
-        init_get_bits(&s->gb, buf, buf_size*8);
-    s->bitstream_buffer_size=0;
-
-    if (!s->context_initialized) {
-        if (MPV_common_init(s) < 0) //we need the idct permutaton for reading a custom matrix
-            return -1;
-    }
-
-    //we need to set current_picture_ptr before reading the header, otherwise we cant store anyting im there
-    if(s->current_picture_ptr==NULL || s->current_picture_ptr->data[0]){
-        int i= ff_find_unused_picture(s, 0);
-        s->current_picture_ptr= &s->picture[i];
-    }
-
-    /* let's go :-) */
-    if (s->msmpeg4_version==5) {
-        ret= ff_wmv2_decode_picture_header(s);
-    } else if (s->msmpeg4_version) {
-        ret = msmpeg4_decode_picture_header(s);
-    } else if (s->h263_pred) {
-        if(s->avctx->extradata_size && s->picture_number==0){
-            GetBitContext gb;
-
-            init_get_bits(&gb, s->avctx->extradata, s->avctx->extradata_size*8);
-            ret = ff_mpeg4_decode_picture_header(s, &gb);
-        }
-        ret = ff_mpeg4_decode_picture_header(s, &s->gb);
-
-        if(s->flags& CODEC_FLAG_LOW_DELAY)
-            s->low_delay=1;
-    } else if (s->codec_id == CODEC_ID_H263I) {
-        ret = intel_h263_decode_picture_header(s);
-    } else if (s->h263_flv) {
-        ret = flv_h263_decode_picture_header(s);
-    } else {
-        ret = h263_decode_picture_header(s);
-    }
-
-    if(ret==FRAME_SKIPPED) return get_consumed_bytes(s, buf_size);
-
-    /* skip if the header was thrashed */
-    if (ret < 0){
-        av_log(s->avctx, AV_LOG_ERROR, "header damaged\n");
-        return -1;
-    }
-
-    avctx->has_b_frames= !s->low_delay;
-
-    if(s->xvid_build==0 && s->divx_version==0 && s->lavc_build==0){
-        if(s->avctx->stream_codec_tag == ff_get_fourcc("XVID") ||
-           s->avctx->codec_tag == ff_get_fourcc("XVID") || s->avctx->codec_tag == ff_get_fourcc("XVIX") ||
-           s->avctx->codec_tag == ff_get_fourcc("RMP4"))
-            s->xvid_build= -1;
-#if 0
-        if(s->avctx->codec_tag == ff_get_fourcc("DIVX") && s->vo_type==0 && s->vol_control_parameters==1
-           && s->padding_bug_score > 0 && s->low_delay) // XVID with modified fourcc
-            s->xvid_build= -1;
-#endif
-    }
-
-    if(s->xvid_build==0 && s->divx_version==0 && s->lavc_build==0){
-        if(s->avctx->codec_tag == ff_get_fourcc("DIVX") && s->vo_type==0 && s->vol_control_parameters==0)
-            s->divx_version= 400; //divx 4
-    }
-
-    if(s->xvid_build && s->divx_version){
-        s->divx_version=
-        s->divx_build= 0;
-    }
-
-    if(s->workaround_bugs&FF_BUG_AUTODETECT){
-        if(s->avctx->codec_tag == ff_get_fourcc("XVIX"))
-            s->workaround_bugs|= FF_BUG_XVID_ILACE;
-
-        if(s->avctx->codec_tag == ff_get_fourcc("UMP4")){
-            s->workaround_bugs|= FF_BUG_UMP4;
-        }
-
-        if(s->divx_version>=500){
-            s->workaround_bugs|= FF_BUG_QPEL_CHROMA;
-        }
-
-        if(s->divx_version>502){
-            s->workaround_bugs|= FF_BUG_QPEL_CHROMA2;
-        }
-
-        if(s->xvid_build && s->xvid_build<=3)
-            s->padding_bug_score= 256*256*256*64;
-
-        if(s->xvid_build && s->xvid_build<=1)
-            s->workaround_bugs|= FF_BUG_QPEL_CHROMA;
-
-        if(s->xvid_build && s->xvid_build<=12)
-            s->workaround_bugs|= FF_BUG_EDGE;
-
-        if(s->xvid_build && s->xvid_build<=32)
-            s->workaround_bugs|= FF_BUG_DC_CLIP;
-
-#define SET_QPEL_FUNC(postfix1, postfix2) \
-    s->dsp.put_ ## postfix1 = ff_put_ ## postfix2;\
-    s->dsp.put_no_rnd_ ## postfix1 = ff_put_no_rnd_ ## postfix2;\
-    s->dsp.avg_ ## postfix1 = ff_avg_ ## postfix2;
-
-        if(s->lavc_build && s->lavc_build<4653)
-            s->workaround_bugs|= FF_BUG_STD_QPEL;
-
-        if(s->lavc_build && s->lavc_build<4655)
-            s->workaround_bugs|= FF_BUG_DIRECT_BLOCKSIZE;
-
-        if(s->lavc_build && s->lavc_build<4670){
-            s->workaround_bugs|= FF_BUG_EDGE;
-        }
-
-        if(s->lavc_build && s->lavc_build<=4712)
-            s->workaround_bugs|= FF_BUG_DC_CLIP;
-
-        if(s->divx_version)
-            s->workaround_bugs|= FF_BUG_DIRECT_BLOCKSIZE;
-//printf("padding_bug_score: %d\n", s->padding_bug_score);
-        if(s->divx_version==501 && s->divx_build==20020416)
-            s->padding_bug_score= 256*256*256*64;
-
-        if(s->divx_version && s->divx_version<500){
-            s->workaround_bugs|= FF_BUG_EDGE;
-        }
-
-        if(s->divx_version)
-            s->workaround_bugs|= FF_BUG_HPEL_CHROMA;
-#if 0
-        if(s->divx_version==500)
-            s->padding_bug_score= 256*256*256*64;
-
-        /* very ugly XVID padding bug detection FIXME/XXX solve this differently
-         * lets hope this at least works
-         */
-        if(   s->resync_marker==0 && s->data_partitioning==0 && s->divx_version==0
-           && s->codec_id==CODEC_ID_MPEG4 && s->vo_type==0)
-            s->workaround_bugs|= FF_BUG_NO_PADDING;
-
-        if(s->lavc_build && s->lavc_build<4609) //FIXME not sure about the version num but a 4609 file seems ok
-            s->workaround_bugs|= FF_BUG_NO_PADDING;
-#endif
-    }
-
-    if(s->workaround_bugs& FF_BUG_STD_QPEL){
-        SET_QPEL_FUNC(qpel_pixels_tab[0][ 5], qpel16_mc11_old_c)
-        SET_QPEL_FUNC(qpel_pixels_tab[0][ 7], qpel16_mc31_old_c)
-        SET_QPEL_FUNC(qpel_pixels_tab[0][ 9], qpel16_mc12_old_c)
-        SET_QPEL_FUNC(qpel_pixels_tab[0][11], qpel16_mc32_old_c)
-        SET_QPEL_FUNC(qpel_pixels_tab[0][13], qpel16_mc13_old_c)
-        SET_QPEL_FUNC(qpel_pixels_tab[0][15], qpel16_mc33_old_c)
-
-        SET_QPEL_FUNC(qpel_pixels_tab[1][ 5], qpel8_mc11_old_c)
-        SET_QPEL_FUNC(qpel_pixels_tab[1][ 7], qpel8_mc31_old_c)
-        SET_QPEL_FUNC(qpel_pixels_tab[1][ 9], qpel8_mc12_old_c)
-        SET_QPEL_FUNC(qpel_pixels_tab[1][11], qpel8_mc32_old_c)
-        SET_QPEL_FUNC(qpel_pixels_tab[1][13], qpel8_mc13_old_c)
-        SET_QPEL_FUNC(qpel_pixels_tab[1][15], qpel8_mc33_old_c)
-    }
-
-    if(avctx->debug & FF_DEBUG_BUGS)
-        av_log(s->avctx, AV_LOG_DEBUG, "bugs: %X lavc_build:%d xvid_build:%d divx_version:%d divx_build:%d %s\n",
-               s->workaround_bugs, s->lavc_build, s->xvid_build, s->divx_version, s->divx_build,
-               s->divx_packed ? "p" : "");
-
-#if 0 // dump bits per frame / qp / complexity
-{
-    static FILE *f=NULL;
-    if(!f) f=fopen("rate_qp_cplx.txt", "w");
-    fprintf(f, "%d %d %f\n", buf_size, s->qscale, buf_size*(double)s->qscale);
-}
-#endif
-
-#if defined(HAVE_MMX) && defined(CONFIG_GPL)
-    if(s->codec_id == CODEC_ID_MPEG4 && s->xvid_build && avctx->idct_algo == FF_IDCT_AUTO && (mm_flags & MM_MMX)){
-        avctx->idct_algo= FF_IDCT_XVIDMMX;
-        avctx->coded_width= 0; // force reinit
-//        dsputil_init(&s->dsp, avctx);
-        s->picture_number=0;
-    }
-#endif
-
-        /* After H263 & mpeg4 header decode we have the height, width,*/
-        /* and other parameters. So then we could init the picture   */
-        /* FIXME: By the way H263 decoder is evolving it should have */
-        /* an H263EncContext                                         */
-
-    if (   s->width  != avctx->coded_width
-        || s->height != avctx->coded_height) {
-        /* H.263 could change picture size any time */
-        ParseContext pc= s->parse_context; //FIXME move these demuxng hack to avformat
-        s->parse_context.buffer=0;
-        MPV_common_end(s);
-        s->parse_context= pc;
-    }
-    if (!s->context_initialized) {
-        avcodec_set_dimensions(avctx, s->width, s->height);
-
-        goto retry;
-    }
-
-    if((s->codec_id==CODEC_ID_H263 || s->codec_id==CODEC_ID_H263P))
-        s->gob_index = ff_h263_get_gob_height(s);
-
-    // for hurry_up==5
-    s->current_picture.pict_type= s->pict_type;
-    s->current_picture.key_frame= s->pict_type == I_TYPE;
-
-    /* skip B-frames if we don't have reference frames */
-    if(s->last_picture_ptr==NULL && (s->pict_type==B_TYPE || s->dropable)) return get_consumed_bytes(s, buf_size);
-    /* skip b frames if we are in a hurry */
-    if(avctx->hurry_up && s->pict_type==B_TYPE) return get_consumed_bytes(s, buf_size);
-    if(   (avctx->skip_frame >= AVDISCARD_NONREF && s->pict_type==B_TYPE)
-       || (avctx->skip_frame >= AVDISCARD_NONKEY && s->pict_type!=I_TYPE)
-       ||  avctx->skip_frame >= AVDISCARD_ALL)
-        return get_consumed_bytes(s, buf_size);
-    /* skip everything if we are in a hurry>=5 */
-    if(avctx->hurry_up>=5) return get_consumed_bytes(s, buf_size);
-
-    if(s->next_p_frame_damaged){
-        if(s->pict_type==B_TYPE)
-            return get_consumed_bytes(s, buf_size);
-        else
-            s->next_p_frame_damaged=0;
-    }
-
-    if((s->avctx->flags2 & CODEC_FLAG2_FAST) && s->pict_type==B_TYPE){
-        s->me.qpel_put= s->dsp.put_2tap_qpel_pixels_tab;
-        s->me.qpel_avg= s->dsp.avg_2tap_qpel_pixels_tab;
-    }else if((!s->no_rounding) || s->pict_type==B_TYPE){
-        s->me.qpel_put= s->dsp.put_qpel_pixels_tab;
-        s->me.qpel_avg= s->dsp.avg_qpel_pixels_tab;
-    }else{
-        s->me.qpel_put= s->dsp.put_no_rnd_qpel_pixels_tab;
-        s->me.qpel_avg= s->dsp.avg_qpel_pixels_tab;
-    }
-
-    if(MPV_frame_start(s, avctx) < 0)
-        return -1;
-
-#ifdef DEBUG
-    av_log(avctx, AV_LOG_DEBUG, "qscale=%d\n", s->qscale);
-#endif
-
-    ff_er_frame_start(s);
-
-    //the second part of the wmv2 header contains the MB skip bits which are stored in current_picture->mb_type
-    //which isnt available before MPV_frame_start()
-    if (s->msmpeg4_version==5){
-        if(ff_wmv2_decode_secondary_picture_header(s) < 0)
-            return -1;
-    }
-
-    /* decode each macroblock */
-    s->mb_x=0;
-    s->mb_y=0;
-
-    decode_slice(s);
-    while(s->mb_y<s->mb_height){
-        if(s->msmpeg4_version){
-            if(s->mb_x!=0 || (s->mb_y%s->slice_height)!=0 || get_bits_count(&s->gb) > s->gb.size_in_bits)
-                break;
-        }else{
-            if(ff_h263_resync(s)<0)
-                break;
-        }
-
-        if(s->msmpeg4_version<4 && s->h263_pred)
-            ff_mpeg4_clean_buffers(s);
-
-        decode_slice(s);
-    }
-
-    if (s->h263_msmpeg4 && s->msmpeg4_version<4 && s->pict_type==I_TYPE)
-        if(msmpeg4_decode_ext_header(s, buf_size) < 0){
-            s->error_status_table[s->mb_num-1]= AC_ERROR|DC_ERROR|MV_ERROR;
-        }
-
-    /* divx 5.01+ bistream reorder stuff */
-    if(s->codec_id==CODEC_ID_MPEG4 && s->bitstream_buffer_size==0 && s->divx_packed){
-        int current_pos= get_bits_count(&s->gb)>>3;
-        int startcode_found=0;
-
-        if(buf_size - current_pos > 5){
-            int i;
-            for(i=current_pos; i<buf_size-3; i++){
-                if(buf[i]==0 && buf[i+1]==0 && buf[i+2]==1 && buf[i+3]==0xB6){
-                    startcode_found=1;
-                    break;
-                }
-            }
-        }
-        if(s->gb.buffer == s->bitstream_buffer && buf_size>20){ //xvid style
-            startcode_found=1;
-            current_pos=0;
-        }
-
-        if(startcode_found){
-            s->bitstream_buffer= av_fast_realloc(
-                s->bitstream_buffer,
-                &s->allocated_bitstream_buffer_size,
-                buf_size - current_pos + FF_INPUT_BUFFER_PADDING_SIZE);
-            memcpy(s->bitstream_buffer, buf + current_pos, buf_size - current_pos);
-            s->bitstream_buffer_size= buf_size - current_pos;
-        }
-    }
-
-    ff_er_frame_end(s);
-
-    MPV_frame_end(s);
-
-assert(s->current_picture.pict_type == s->current_picture_ptr->pict_type);
-assert(s->current_picture.pict_type == s->pict_type);
-    if (s->pict_type == B_TYPE || s->low_delay) {
-        *pict= *(AVFrame*)s->current_picture_ptr;
-    } else if (s->last_picture_ptr != NULL) {
-        *pict= *(AVFrame*)s->last_picture_ptr;
-    }
-
-    if(s->last_picture_ptr || s->low_delay){
-        *data_size = sizeof(AVFrame);
-        ff_print_debug_info(s, pict);
-    }
-
-    /* Return the Picture timestamp as the frame number */
-    /* we substract 1 because it is added on utils.c    */
-    avctx->frame_number = s->picture_number - 1;
-
-#ifdef PRINT_FRAME_TIME
-av_log(avctx, AV_LOG_DEBUG, "%Ld\n", rdtsc()-time);
-#endif
-
-    return get_consumed_bytes(s, buf_size);
-}
-
-AVCodec mpeg4_decoder = {
-    "mpeg4",
-    CODEC_TYPE_VIDEO,
-    CODEC_ID_MPEG4,
-    sizeof(MpegEncContext),
-    ff_h263_decode_init,
-    NULL,
-    ff_h263_decode_end,
-    ff_h263_decode_frame,
-    CODEC_CAP_DRAW_HORIZ_BAND | CODEC_CAP_DR1 | CODEC_CAP_TRUNCATED | CODEC_CAP_DELAY,
-    .flush= ff_mpeg_flush,
-};
-
-AVCodec h263_decoder = {
-    "h263",
-    CODEC_TYPE_VIDEO,
-    CODEC_ID_H263,
-    sizeof(MpegEncContext),
-    ff_h263_decode_init,
-    NULL,
-    ff_h263_decode_end,
-    ff_h263_decode_frame,
-    CODEC_CAP_DRAW_HORIZ_BAND | CODEC_CAP_DR1 | CODEC_CAP_TRUNCATED | CODEC_CAP_DELAY,
-    .flush= ff_mpeg_flush,
-};
-
-AVCodec msmpeg4v1_decoder = {
-    "msmpeg4v1",
-    CODEC_TYPE_VIDEO,
-    CODEC_ID_MSMPEG4V1,
-    sizeof(MpegEncContext),
-    ff_h263_decode_init,
-    NULL,
-    ff_h263_decode_end,
-    ff_h263_decode_frame,
-    CODEC_CAP_DRAW_HORIZ_BAND | CODEC_CAP_DR1,
-};
-
-AVCodec msmpeg4v2_decoder = {
-    "msmpeg4v2",
-    CODEC_TYPE_VIDEO,
-    CODEC_ID_MSMPEG4V2,
-    sizeof(MpegEncContext),
-    ff_h263_decode_init,
-    NULL,
-    ff_h263_decode_end,
-    ff_h263_decode_frame,
-    CODEC_CAP_DRAW_HORIZ_BAND | CODEC_CAP_DR1,
-};
-
-AVCodec msmpeg4v3_decoder = {
-    "msmpeg4",
-    CODEC_TYPE_VIDEO,
-    CODEC_ID_MSMPEG4V3,
-    sizeof(MpegEncContext),
-    ff_h263_decode_init,
-    NULL,
-    ff_h263_decode_end,
-    ff_h263_decode_frame,
-    CODEC_CAP_DRAW_HORIZ_BAND | CODEC_CAP_DR1,
-};
-
-AVCodec wmv1_decoder = {
-    "wmv1",
-    CODEC_TYPE_VIDEO,
-    CODEC_ID_WMV1,
-    sizeof(MpegEncContext),
-    ff_h263_decode_init,
-    NULL,
-    ff_h263_decode_end,
-    ff_h263_decode_frame,
-    CODEC_CAP_DRAW_HORIZ_BAND | CODEC_CAP_DR1,
-};
-
-AVCodec h263i_decoder = {
-    "h263i",
-    CODEC_TYPE_VIDEO,
-    CODEC_ID_H263I,
-    sizeof(MpegEncContext),
-    ff_h263_decode_init,
-    NULL,
-    ff_h263_decode_end,
-    ff_h263_decode_frame,
-    CODEC_CAP_DRAW_HORIZ_BAND | CODEC_CAP_DR1,
-};
-
-AVCodec flv_decoder = {
-    "flv",
-    CODEC_TYPE_VIDEO,
-    CODEC_ID_FLV1,
-    sizeof(MpegEncContext),
-    ff_h263_decode_init,
-    NULL,
-    ff_h263_decode_end,
-    ff_h263_decode_frame,
-    CODEC_CAP_DRAW_HORIZ_BAND | CODEC_CAP_DR1
-};
-
-#ifdef CONFIG_H263_PARSER
-AVCodecParser h263_parser = {
-    { CODEC_ID_H263 },
-    sizeof(ParseContext),
-    NULL,
-    h263_parse,
-    ff_parse_close,
-};
-#endif
--- a/src/ffmpeg/libavcodec/h264.c	Mon Mar 12 11:37:35 2007 -0700
+++ /dev/null	Thu Jan 01 00:00:00 1970 +0000
@@ -1,8655 +0,0 @@
-/*
- * H.26L/H.264/AVC/JVT/14496-10/... encoder/decoder
- * Copyright (c) 2003 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 h264.c
- * H.264 / AVC / MPEG4 part10 codec.
- * @author Michael Niedermayer <michaelni@gmx.at>
- */
-
-#include "common.h"
-#include "dsputil.h"
-#include "avcodec.h"
-#include "mpegvideo.h"
-#include "h264data.h"
-#include "golomb.h"
-
-#include "cabac.h"
-
-//#undef NDEBUG
-#include <assert.h>
-
-#define interlaced_dct interlaced_dct_is_a_bad_name
-#define mb_intra mb_intra_isnt_initalized_see_mb_type
-
-#define LUMA_DC_BLOCK_INDEX   25
-#define CHROMA_DC_BLOCK_INDEX 26
-
-#define CHROMA_DC_COEFF_TOKEN_VLC_BITS 8
-#define COEFF_TOKEN_VLC_BITS           8
-#define TOTAL_ZEROS_VLC_BITS           9
-#define CHROMA_DC_TOTAL_ZEROS_VLC_BITS 3
-#define RUN_VLC_BITS                   3
-#define RUN7_VLC_BITS                  6
-
-#define MAX_SPS_COUNT 32
-#define MAX_PPS_COUNT 256
-
-#define MAX_MMCO_COUNT 66
-
-/* Compiling in interlaced support reduces the speed
- * of progressive decoding by about 2%. */
-#define ALLOW_INTERLACE
-
-#ifdef ALLOW_INTERLACE
-#define MB_MBAFF h->mb_mbaff
-#define MB_FIELD h->mb_field_decoding_flag
-#define FRAME_MBAFF h->mb_aff_frame
-#else
-#define MB_MBAFF 0
-#define MB_FIELD 0
-#define FRAME_MBAFF 0
-#undef  IS_INTERLACED
-#define IS_INTERLACED(mb_type) 0
-#endif
-
-/**
- * Sequence parameter set
- */
-typedef struct SPS{
-
-    int profile_idc;
-    int level_idc;
-    int transform_bypass;              ///< qpprime_y_zero_transform_bypass_flag
-    int log2_max_frame_num;            ///< log2_max_frame_num_minus4 + 4
-    int poc_type;                      ///< pic_order_cnt_type
-    int log2_max_poc_lsb;              ///< log2_max_pic_order_cnt_lsb_minus4
-    int delta_pic_order_always_zero_flag;
-    int offset_for_non_ref_pic;
-    int offset_for_top_to_bottom_field;
-    int poc_cycle_length;              ///< num_ref_frames_in_pic_order_cnt_cycle
-    int ref_frame_count;               ///< num_ref_frames
-    int gaps_in_frame_num_allowed_flag;
-    int mb_width;                      ///< frame_width_in_mbs_minus1 + 1
-    int mb_height;                     ///< frame_height_in_mbs_minus1 + 1
-    int frame_mbs_only_flag;
-    int mb_aff;                        ///<mb_adaptive_frame_field_flag
-    int direct_8x8_inference_flag;
-    int crop;                   ///< frame_cropping_flag
-    int crop_left;              ///< frame_cropping_rect_left_offset
-    int crop_right;             ///< frame_cropping_rect_right_offset
-    int crop_top;               ///< frame_cropping_rect_top_offset
-    int crop_bottom;            ///< frame_cropping_rect_bottom_offset
-    int vui_parameters_present_flag;
-    AVRational sar;
-    int timing_info_present_flag;
-    uint32_t num_units_in_tick;
-    uint32_t time_scale;
-    int fixed_frame_rate_flag;
-    short offset_for_ref_frame[256]; //FIXME dyn aloc?
-    int bitstream_restriction_flag;
-    int num_reorder_frames;
-    int scaling_matrix_present;
-    uint8_t scaling_matrix4[6][16];
-    uint8_t scaling_matrix8[2][64];
-}SPS;
-
-/**
- * Picture parameter set
- */
-typedef struct PPS{
-    int sps_id;
-    int cabac;                  ///< entropy_coding_mode_flag
-    int pic_order_present;      ///< pic_order_present_flag
-    int slice_group_count;      ///< num_slice_groups_minus1 + 1
-    int mb_slice_group_map_type;
-    int ref_count[2];           ///< num_ref_idx_l0/1_active_minus1 + 1
-    int weighted_pred;          ///< weighted_pred_flag
-    int weighted_bipred_idc;
-    int init_qp;                ///< pic_init_qp_minus26 + 26
-    int init_qs;                ///< pic_init_qs_minus26 + 26
-    int chroma_qp_index_offset;
-    int deblocking_filter_parameters_present; ///< deblocking_filter_parameters_present_flag
-    int constrained_intra_pred; ///< constrained_intra_pred_flag
-    int redundant_pic_cnt_present; ///< redundant_pic_cnt_present_flag
-    int transform_8x8_mode;     ///< transform_8x8_mode_flag
-    uint8_t scaling_matrix4[6][16];
-    uint8_t scaling_matrix8[2][64];
-}PPS;
-
-/**
- * Memory management control operation opcode.
- */
-typedef enum MMCOOpcode{
-    MMCO_END=0,
-    MMCO_SHORT2UNUSED,
-    MMCO_LONG2UNUSED,
-    MMCO_SHORT2LONG,
-    MMCO_SET_MAX_LONG,
-    MMCO_RESET,
-    MMCO_LONG,
-} MMCOOpcode;
-
-/**
- * Memory management control operation.
- */
-typedef struct MMCO{
-    MMCOOpcode opcode;
-    int short_frame_num;
-    int long_index;
-} MMCO;
-
-/**
- * H264Context
- */
-typedef struct H264Context{
-    MpegEncContext s;
-    int nal_ref_idc;
-    int nal_unit_type;
-#define NAL_SLICE                1
-#define NAL_DPA                  2
-#define NAL_DPB                  3
-#define NAL_DPC                  4
-#define NAL_IDR_SLICE            5
-#define NAL_SEI                  6
-#define NAL_SPS                  7
-#define NAL_PPS                  8
-#define NAL_AUD                  9
-#define NAL_END_SEQUENCE        10
-#define NAL_END_STREAM          11
-#define NAL_FILLER_DATA         12
-#define NAL_SPS_EXT             13
-#define NAL_AUXILIARY_SLICE     19
-    uint8_t *rbsp_buffer;
-    unsigned int rbsp_buffer_size;
-
-    /**
-      * Used to parse AVC variant of h264
-      */
-    int is_avc; ///< this flag is != 0 if codec is avc1
-    int got_avcC; ///< flag used to parse avcC data only once
-    int nal_length_size; ///< Number of bytes used for nal length (1, 2 or 4)
-
-    int chroma_qp; //QPc
-
-    int prev_mb_skipped;
-    int next_mb_skipped;
-
-    //prediction stuff
-    int chroma_pred_mode;
-    int intra16x16_pred_mode;
-
-    int top_mb_xy;
-    int left_mb_xy[2];
-
-    int8_t intra4x4_pred_mode_cache[5*8];
-    int8_t (*intra4x4_pred_mode)[8];
-    void (*pred4x4  [9+3])(uint8_t *src, uint8_t *topright, int stride);//FIXME move to dsp?
-    void (*pred8x8l [9+3])(uint8_t *src, int topleft, int topright, int stride);
-    void (*pred8x8  [4+3])(uint8_t *src, int stride);
-    void (*pred16x16[4+3])(uint8_t *src, int stride);
-    unsigned int topleft_samples_available;
-    unsigned int top_samples_available;
-    unsigned int topright_samples_available;
-    unsigned int left_samples_available;
-    uint8_t (*top_borders[2])[16+2*8];
-    uint8_t left_border[2*(17+2*9)];
-
-    /**
-     * non zero coeff count cache.
-     * is 64 if not available.
-     */
-    DECLARE_ALIGNED_8(uint8_t, non_zero_count_cache[6*8]);
-    uint8_t (*non_zero_count)[16];
-
-    /**
-     * Motion vector cache.
-     */
-    DECLARE_ALIGNED_8(int16_t, mv_cache[2][5*8][2]);
-    DECLARE_ALIGNED_8(int8_t, ref_cache[2][5*8]);
-#define LIST_NOT_USED -1 //FIXME rename?
-#define PART_NOT_AVAILABLE -2
-
-    /**
-     * is 1 if the specific list MV&references are set to 0,0,-2.
-     */
-    int mv_cache_clean[2];
-
-    /**
-     * number of neighbors (top and/or left) that used 8x8 dct
-     */
-    int neighbor_transform_size;
-
-    /**
-     * block_offset[ 0..23] for frame macroblocks
-     * block_offset[24..47] for field macroblocks
-     */
-    int block_offset[2*(16+8)];
-
-    uint32_t *mb2b_xy; //FIXME are these 4 a good idea?
-    uint32_t *mb2b8_xy;
-    int b_stride; //FIXME use s->b4_stride
-    int b8_stride;
-
-    int mb_linesize;   ///< may be equal to s->linesize or s->linesize*2, for mbaff
-    int mb_uvlinesize;
-
-    int emu_edge_width;
-    int emu_edge_height;
-
-    int halfpel_flag;
-    int thirdpel_flag;
-
-    int unknown_svq3_flag;
-    int next_slice_index;
-
-    SPS sps_buffer[MAX_SPS_COUNT];
-    SPS sps; ///< current sps
-
-    PPS pps_buffer[MAX_PPS_COUNT];
-    /**
-     * current pps
-     */
-    PPS pps; //FIXME move to Picture perhaps? (->no) do we need that?
-
-    uint32_t dequant4_buffer[6][52][16];
-    uint32_t dequant8_buffer[2][52][64];
-    uint32_t (*dequant4_coeff[6])[16];
-    uint32_t (*dequant8_coeff[2])[64];
-    int dequant_coeff_pps;     ///< reinit tables when pps changes
-
-    int slice_num;
-    uint8_t *slice_table_base;
-    uint8_t *slice_table;      ///< slice_table_base + 2*mb_stride + 1
-    int slice_type;
-    int slice_type_fixed;
-
-    //interlacing specific flags
-    int mb_aff_frame;
-    int mb_field_decoding_flag;
-    int mb_mbaff;              ///< mb_aff_frame && mb_field_decoding_flag
-
-    int sub_mb_type[4];
-
-    //POC stuff
-    int poc_lsb;
-    int poc_msb;
-    int delta_poc_bottom;
-    int delta_poc[2];
-    int frame_num;
-    int prev_poc_msb;             ///< poc_msb of the last reference pic for POC type 0
-    int prev_poc_lsb;             ///< poc_lsb of the last reference pic for POC type 0
-    int frame_num_offset;         ///< for POC type 2
-    int prev_frame_num_offset;    ///< for POC type 2
-    int prev_frame_num;           ///< frame_num of the last pic for POC type 1/2
-
-    /**
-     * frame_num for frames or 2*frame_num for field pics.
-     */
-    int curr_pic_num;
-
-    /**
-     * max_frame_num or 2*max_frame_num for field pics.
-     */
-    int max_pic_num;
-
-    //Weighted pred stuff
-    int use_weight;
-    int use_weight_chroma;
-    int luma_log2_weight_denom;
-    int chroma_log2_weight_denom;
-    int luma_weight[2][48];
-    int luma_offset[2][48];
-    int chroma_weight[2][48][2];
-    int chroma_offset[2][48][2];
-    int implicit_weight[48][48];
-
-    //deblock
-    int deblocking_filter;         ///< disable_deblocking_filter_idc with 1<->0
-    int slice_alpha_c0_offset;
-    int slice_beta_offset;
-
-    int redundant_pic_count;
-
-    int direct_spatial_mv_pred;
-    int dist_scale_factor[16];
-    int dist_scale_factor_field[32];
-    int map_col_to_list0[2][16];
-    int map_col_to_list0_field[2][32];
-
-    /**
-     * num_ref_idx_l0/1_active_minus1 + 1
-     */
-    int ref_count[2];            ///< counts frames or fields, depending on current mb mode
-    Picture *short_ref[32];
-    Picture *long_ref[32];
-    Picture default_ref_list[2][32];
-    Picture ref_list[2][48];     ///< 0..15: frame refs, 16..47: mbaff field refs
-    Picture *delayed_pic[16]; //FIXME size?
-    Picture *delayed_output_pic;
-
-    /**
-     * memory management control operations buffer.
-     */
-    MMCO mmco[MAX_MMCO_COUNT];
-    int mmco_index;
-
-    int long_ref_count;  ///< number of actual long term references
-    int short_ref_count; ///< number of actual short term references
-
-    //data partitioning
-    GetBitContext intra_gb;
-    GetBitContext inter_gb;
-    GetBitContext *intra_gb_ptr;
-    GetBitContext *inter_gb_ptr;
-
-    DECLARE_ALIGNED_8(DCTELEM, mb[16*24]);
-
-    /**
-     * Cabac
-     */
-    CABACContext cabac;
-    uint8_t      cabac_state[460];
-    int          cabac_init_idc;
-
-    /* 0x100 -> non null luma_dc, 0x80/0x40 -> non null chroma_dc (cb/cr), 0x?0 -> chroma_cbp(0,1,2), 0x0? luma_cbp */
-    uint16_t     *cbp_table;
-    int cbp;
-    int top_cbp;
-    int left_cbp;
-    /* chroma_pred_mode for i4x4 or i16x16, else 0 */
-    uint8_t     *chroma_pred_mode_table;
-    int         last_qscale_diff;
-    int16_t     (*mvd_table[2])[2];
-    DECLARE_ALIGNED_8(int16_t, mvd_cache[2][5*8][2]);
-    uint8_t     *direct_table;
-    uint8_t     direct_cache[5*8];
-
-    uint8_t zigzag_scan[16];
-    uint8_t zigzag_scan8x8[64];
-    uint8_t zigzag_scan8x8_cavlc[64];
-    uint8_t field_scan[16];
-    uint8_t field_scan8x8[64];
-    uint8_t field_scan8x8_cavlc[64];
-    const uint8_t *zigzag_scan_q0;
-    const uint8_t *zigzag_scan8x8_q0;
-    const uint8_t *zigzag_scan8x8_cavlc_q0;
-    const uint8_t *field_scan_q0;
-    const uint8_t *field_scan8x8_q0;
-    const uint8_t *field_scan8x8_cavlc_q0;
-
-    int x264_build;
-}H264Context;
-
-static VLC coeff_token_vlc[4];
-static VLC chroma_dc_coeff_token_vlc;
-
-static VLC total_zeros_vlc[15];
-static VLC chroma_dc_total_zeros_vlc[3];
-
-static VLC run_vlc[6];
-static VLC run7_vlc;
-
-static void svq3_luma_dc_dequant_idct_c(DCTELEM *block, int qp);
-static void svq3_add_idct_c(uint8_t *dst, DCTELEM *block, int stride, int qp, int dc);
-static void filter_mb( H264Context *h, int mb_x, int mb_y, uint8_t *img_y, uint8_t *img_cb, uint8_t *img_cr, unsigned int linesize, unsigned int uvlinesize);
-static void filter_mb_fast( H264Context *h, int mb_x, int mb_y, uint8_t *img_y, uint8_t *img_cb, uint8_t *img_cr, unsigned int linesize, unsigned int uvlinesize);
-
-static always_inline uint32_t pack16to32(int a, int b){
-#ifdef WORDS_BIGENDIAN
-   return (b&0xFFFF) + (a<<16);
-#else
-   return (a&0xFFFF) + (b<<16);
-#endif
-}
-
-/**
- * fill a rectangle.
- * @param h height of the rectangle, should be a constant
- * @param w width of the rectangle, should be a constant
- * @param size the size of val (1 or 4), should be a constant
- */
-static always_inline void fill_rectangle(void *vp, int w, int h, int stride, uint32_t val, int size){
-    uint8_t *p= (uint8_t*)vp;
-    assert(size==1 || size==4);
-    assert(w<=4);
-
-    w      *= size;
-    stride *= size;
-
-    assert((((long)vp)&(FFMIN(w, STRIDE_ALIGN)-1)) == 0);
-    assert((stride&(w-1))==0);
-    if(w==2){
-        const uint16_t v= size==4 ? val : val*0x0101;
-        *(uint16_t*)(p + 0*stride)= v;
-        if(h==1) return;
-        *(uint16_t*)(p + 1*stride)= v;
-        if(h==2) return;
-        *(uint16_t*)(p + 2*stride)=
-        *(uint16_t*)(p + 3*stride)= v;
-    }else if(w==4){
-        const uint32_t v= size==4 ? val : val*0x01010101;
-        *(uint32_t*)(p + 0*stride)= v;
-        if(h==1) return;
-        *(uint32_t*)(p + 1*stride)= v;
-        if(h==2) return;
-        *(uint32_t*)(p + 2*stride)=
-        *(uint32_t*)(p + 3*stride)= v;
-    }else if(w==8){
-    //gcc can't optimize 64bit math on x86_32
-#if defined(ARCH_X86_64) || (defined(MP_WORDSIZE) && MP_WORDSIZE >= 64)
-        const uint64_t v= val*0x0100000001ULL;
-        *(uint64_t*)(p + 0*stride)= v;
-        if(h==1) return;
-        *(uint64_t*)(p + 1*stride)= v;
-        if(h==2) return;
-        *(uint64_t*)(p + 2*stride)=
-        *(uint64_t*)(p + 3*stride)= v;
-    }else if(w==16){
-        const uint64_t v= val*0x0100000001ULL;
-        *(uint64_t*)(p + 0+0*stride)=
-        *(uint64_t*)(p + 8+0*stride)=
-        *(uint64_t*)(p + 0+1*stride)=
-        *(uint64_t*)(p + 8+1*stride)= v;
-        if(h==2) return;
-        *(uint64_t*)(p + 0+2*stride)=
-        *(uint64_t*)(p + 8+2*stride)=
-        *(uint64_t*)(p + 0+3*stride)=
-        *(uint64_t*)(p + 8+3*stride)= v;
-#else
-        *(uint32_t*)(p + 0+0*stride)=
-        *(uint32_t*)(p + 4+0*stride)= val;
-        if(h==1) return;
-        *(uint32_t*)(p + 0+1*stride)=
-        *(uint32_t*)(p + 4+1*stride)= val;
-        if(h==2) return;
-        *(uint32_t*)(p + 0+2*stride)=
-        *(uint32_t*)(p + 4+2*stride)=
-        *(uint32_t*)(p + 0+3*stride)=
-        *(uint32_t*)(p + 4+3*stride)= val;
-    }else if(w==16){
-        *(uint32_t*)(p + 0+0*stride)=
-        *(uint32_t*)(p + 4+0*stride)=
-        *(uint32_t*)(p + 8+0*stride)=
-        *(uint32_t*)(p +12+0*stride)=
-        *(uint32_t*)(p + 0+1*stride)=
-        *(uint32_t*)(p + 4+1*stride)=
-        *(uint32_t*)(p + 8+1*stride)=
-        *(uint32_t*)(p +12+1*stride)= val;
-        if(h==2) return;
-        *(uint32_t*)(p + 0+2*stride)=
-        *(uint32_t*)(p + 4+2*stride)=
-        *(uint32_t*)(p + 8+2*stride)=
-        *(uint32_t*)(p +12+2*stride)=
-        *(uint32_t*)(p + 0+3*stride)=
-        *(uint32_t*)(p + 4+3*stride)=
-        *(uint32_t*)(p + 8+3*stride)=
-        *(uint32_t*)(p +12+3*stride)= val;
-#endif
-    }else
-        assert(0);
-    assert(h==4);
-}
-
-static void fill_caches(H264Context *h, int mb_type, int for_deblock){
-    MpegEncContext * const s = &h->s;
-    const int mb_xy= s->mb_x + s->mb_y*s->mb_stride;
-    int topleft_xy, top_xy, topright_xy, left_xy[2];
-    int topleft_type, top_type, topright_type, left_type[2];
-    int left_block[8];
-    int i;
-
-    //FIXME deblocking could skip the intra and nnz parts.
-    if(for_deblock && (h->slice_num == 1 || h->slice_table[mb_xy] == h->slice_table[mb_xy-s->mb_stride]) && !FRAME_MBAFF)
-        return;
-
-    //wow what a mess, why didn't they simplify the interlacing&intra stuff, i can't imagine that these complex rules are worth it
-
-    top_xy     = mb_xy  - s->mb_stride;
-    topleft_xy = top_xy - 1;
-    topright_xy= top_xy + 1;
-    left_xy[1] = left_xy[0] = mb_xy-1;
-    left_block[0]= 0;
-    left_block[1]= 1;
-    left_block[2]= 2;
-    left_block[3]= 3;
-    left_block[4]= 7;
-    left_block[5]= 10;
-    left_block[6]= 8;
-    left_block[7]= 11;
-    if(FRAME_MBAFF){
-        const int pair_xy          = s->mb_x     + (s->mb_y & ~1)*s->mb_stride;
-        const int top_pair_xy      = pair_xy     - s->mb_stride;
-        const int topleft_pair_xy  = top_pair_xy - 1;
-        const int topright_pair_xy = top_pair_xy + 1;
-        const int topleft_mb_frame_flag  = !IS_INTERLACED(s->current_picture.mb_type[topleft_pair_xy]);
-        const int top_mb_frame_flag      = !IS_INTERLACED(s->current_picture.mb_type[top_pair_xy]);
-        const int topright_mb_frame_flag = !IS_INTERLACED(s->current_picture.mb_type[topright_pair_xy]);
-        const int left_mb_frame_flag = !IS_INTERLACED(s->current_picture.mb_type[pair_xy-1]);
-        const int curr_mb_frame_flag = !IS_INTERLACED(mb_type);
-        const int bottom = (s->mb_y & 1);
-        tprintf("fill_caches: curr_mb_frame_flag:%d, left_mb_frame_flag:%d, topleft_mb_frame_flag:%d, top_mb_frame_flag:%d, topright_mb_frame_flag:%d\n", curr_mb_frame_flag, left_mb_frame_flag, topleft_mb_frame_flag, top_mb_frame_flag, topright_mb_frame_flag);
-        if (bottom
-                ? !curr_mb_frame_flag // bottom macroblock
-                : (!curr_mb_frame_flag && !top_mb_frame_flag) // top macroblock
-                ) {
-            top_xy -= s->mb_stride;
-        }
-        if (bottom
-                ? !curr_mb_frame_flag // bottom macroblock
-                : (!curr_mb_frame_flag && !topleft_mb_frame_flag) // top macroblock
-                ) {
-            topleft_xy -= s->mb_stride;
-        }
-        if (bottom
-                ? !curr_mb_frame_flag // bottom macroblock
-                : (!curr_mb_frame_flag && !topright_mb_frame_flag) // top macroblock
-                ) {
-            topright_xy -= s->mb_stride;
-        }
-        if (left_mb_frame_flag != curr_mb_frame_flag) {
-            left_xy[1] = left_xy[0] = pair_xy - 1;
-            if (curr_mb_frame_flag) {
-                if (bottom) {
-                    left_block[0]= 2;
-                    left_block[1]= 2;
-                    left_block[2]= 3;
-                    left_block[3]= 3;
-                    left_block[4]= 8;
-                    left_block[5]= 11;
-                    left_block[6]= 8;
-                    left_block[7]= 11;
-                } else {
-                    left_block[0]= 0;
-                    left_block[1]= 0;
-                    left_block[2]= 1;
-                    left_block[3]= 1;
-                    left_block[4]= 7;
-                    left_block[5]= 10;
-                    left_block[6]= 7;
-                    left_block[7]= 10;
-                }
-            } else {
-                left_xy[1] += s->mb_stride;
-                //left_block[0]= 0;
-                left_block[1]= 2;
-                left_block[2]= 0;
-                left_block[3]= 2;
-                //left_block[4]= 7;
-                left_block[5]= 10;
-                left_block[6]= 7;
-                left_block[7]= 10;
-            }
-        }
-    }
-
-    h->top_mb_xy = top_xy;
-    h->left_mb_xy[0] = left_xy[0];
-    h->left_mb_xy[1] = left_xy[1];
-    if(for_deblock){
-        topleft_type = 0;
-        topright_type = 0;
-        top_type     = h->slice_table[top_xy     ] < 255 ? s->current_picture.mb_type[top_xy]     : 0;
-        left_type[0] = h->slice_table[left_xy[0] ] < 255 ? s->current_picture.mb_type[left_xy[0]] : 0;
-        left_type[1] = h->slice_table[left_xy[1] ] < 255 ? s->current_picture.mb_type[left_xy[1]] : 0;
-
-        if(FRAME_MBAFF && !IS_INTRA(mb_type)){
-            int list;
-            int v = *(uint16_t*)&h->non_zero_count[mb_xy][14];
-            for(i=0; i<16; i++)
-                h->non_zero_count_cache[scan8[i]] = (v>>i)&1;
-            for(list=0; list<1+(h->slice_type==B_TYPE); list++){
-                if(USES_LIST(mb_type,list)){
-                    uint32_t *src = (uint32_t*)s->current_picture.motion_val[list][h->mb2b_xy[mb_xy]];
-                    uint32_t *dst = (uint32_t*)h->mv_cache[list][scan8[0]];
-                    int8_t *ref = &s->current_picture.ref_index[list][h->mb2b8_xy[mb_xy]];
-                    for(i=0; i<4; i++, dst+=8, src+=h->b_stride){
-                        dst[0] = src[0];
-                        dst[1] = src[1];
-                        dst[2] = src[2];
-                        dst[3] = src[3];
-                    }
-                    *(uint32_t*)&h->ref_cache[list][scan8[ 0]] =
-                    *(uint32_t*)&h->ref_cache[list][scan8[ 2]] = pack16to32(ref[0],ref[1])*0x0101;
-                    ref += h->b8_stride;
-                    *(uint32_t*)&h->ref_cache[list][scan8[ 8]] =
-                    *(uint32_t*)&h->ref_cache[list][scan8[10]] = pack16to32(ref[0],ref[1])*0x0101;
-                }else{
-                    fill_rectangle(&h-> mv_cache[list][scan8[ 0]], 4, 4, 8, 0, 4);
-                    fill_rectangle(&h->ref_cache[list][scan8[ 0]], 4, 4, 8, (uint8_t)LIST_NOT_USED, 1);
-                }
-            }
-        }
-    }else{
-        topleft_type = h->slice_table[topleft_xy ] == h->slice_num ? s->current_picture.mb_type[topleft_xy] : 0;
-        top_type     = h->slice_table[top_xy     ] == h->slice_num ? s->current_picture.mb_type[top_xy]     : 0;
-        topright_type= h->slice_table[topright_xy] == h->slice_num ? s->current_picture.mb_type[topright_xy]: 0;
-        left_type[0] = h->slice_table[left_xy[0] ] == h->slice_num ? s->current_picture.mb_type[left_xy[0]] : 0;
-        left_type[1] = h->slice_table[left_xy[1] ] == h->slice_num ? s->current_picture.mb_type[left_xy[1]] : 0;
-    }
-
-    if(IS_INTRA(mb_type)){
-        h->topleft_samples_available=
-        h->top_samples_available=
-        h->left_samples_available= 0xFFFF;
-        h->topright_samples_available= 0xEEEA;
-
-        if(!IS_INTRA(top_type) && (top_type==0 || h->pps.constrained_intra_pred)){
-            h->topleft_samples_available= 0xB3FF;
-            h->top_samples_available= 0x33FF;
-            h->topright_samples_available= 0x26EA;
-        }
-        for(i=0; i<2; i++){
-            if(!IS_INTRA(left_type[i]) && (left_type[i]==0 || h->pps.constrained_intra_pred)){
-                h->topleft_samples_available&= 0xDF5F;
-                h->left_samples_available&= 0x5F5F;
-            }
-        }
-
-        if(!IS_INTRA(topleft_type) && (topleft_type==0 || h->pps.constrained_intra_pred))
-            h->topleft_samples_available&= 0x7FFF;
-
-        if(!IS_INTRA(topright_type) && (topright_type==0 || h->pps.constrained_intra_pred))
-            h->topright_samples_available&= 0xFBFF;
-
-        if(IS_INTRA4x4(mb_type)){
-            if(IS_INTRA4x4(top_type)){
-                h->intra4x4_pred_mode_cache[4+8*0]= h->intra4x4_pred_mode[top_xy][4];
-                h->intra4x4_pred_mode_cache[5+8*0]= h->intra4x4_pred_mode[top_xy][5];
-                h->intra4x4_pred_mode_cache[6+8*0]= h->intra4x4_pred_mode[top_xy][6];
-                h->intra4x4_pred_mode_cache[7+8*0]= h->intra4x4_pred_mode[top_xy][3];
-            }else{
-                int pred;
-                if(!top_type || (IS_INTER(top_type) && h->pps.constrained_intra_pred))
-                    pred= -1;
-                else{
-                    pred= 2;
-                }
-                h->intra4x4_pred_mode_cache[4+8*0]=
-                h->intra4x4_pred_mode_cache[5+8*0]=
-                h->intra4x4_pred_mode_cache[6+8*0]=
-                h->intra4x4_pred_mode_cache[7+8*0]= pred;
-            }
-            for(i=0; i<2; i++){
-                if(IS_INTRA4x4(left_type[i])){
-                    h->intra4x4_pred_mode_cache[3+8*1 + 2*8*i]= h->intra4x4_pred_mode[left_xy[i]][left_block[0+2*i]];
-                    h->intra4x4_pred_mode_cache[3+8*2 + 2*8*i]= h->intra4x4_pred_mode[left_xy[i]][left_block[1+2*i]];
-                }else{
-                    int pred;
-                    if(!left_type[i] || (IS_INTER(left_type[i]) && h->pps.constrained_intra_pred))
-                        pred= -1;
-                    else{
-                        pred= 2;
-                    }
-                    h->intra4x4_pred_mode_cache[3+8*1 + 2*8*i]=
-                    h->intra4x4_pred_mode_cache[3+8*2 + 2*8*i]= pred;
-                }
-            }
-        }
-    }
-
-
-/*
-0 . T T. T T T T
-1 L . .L . . . .
-2 L . .L . . . .
-3 . T TL . . . .
-4 L . .L . . . .
-5 L . .. . . . .
-*/
-//FIXME constraint_intra_pred & partitioning & nnz (lets hope this is just a typo in the spec)
-    if(top_type){
-        h->non_zero_count_cache[4+8*0]= h->non_zero_count[top_xy][4];
-        h->non_zero_count_cache[5+8*0]= h->non_zero_count[top_xy][5];
-        h->non_zero_count_cache[6+8*0]= h->non_zero_count[top_xy][6];
-        h->non_zero_count_cache[7+8*0]= h->non_zero_count[top_xy][3];
-
-        h->non_zero_count_cache[1+8*0]= h->non_zero_count[top_xy][9];
-        h->non_zero_count_cache[2+8*0]= h->non_zero_count[top_xy][8];
-
-        h->non_zero_count_cache[1+8*3]= h->non_zero_count[top_xy][12];
-        h->non_zero_count_cache[2+8*3]= h->non_zero_count[top_xy][11];
-
-    }else{
-        h->non_zero_count_cache[4+8*0]=
-        h->non_zero_count_cache[5+8*0]=
-        h->non_zero_count_cache[6+8*0]=
-        h->non_zero_count_cache[7+8*0]=
-
-        h->non_zero_count_cache[1+8*0]=
-        h->non_zero_count_cache[2+8*0]=
-
-        h->non_zero_count_cache[1+8*3]=
-        h->non_zero_count_cache[2+8*3]= h->pps.cabac && !IS_INTRA(mb_type) ? 0 : 64;
-
-    }
-
-    for (i=0; i<2; i++) {
-        if(left_type[i]){
-            h->non_zero_count_cache[3+8*1 + 2*8*i]= h->non_zero_count[left_xy[i]][left_block[0+2*i]];
-            h->non_zero_count_cache[3+8*2 + 2*8*i]= h->non_zero_count[left_xy[i]][left_block[1+2*i]];
-            h->non_zero_count_cache[0+8*1 +   8*i]= h->non_zero_count[left_xy[i]][left_block[4+2*i]];
-            h->non_zero_count_cache[0+8*4 +   8*i]= h->non_zero_count[left_xy[i]][left_block[5+2*i]];
-        }else{
-            h->non_zero_count_cache[3+8*1 + 2*8*i]=
-            h->non_zero_count_cache[3+8*2 + 2*8*i]=
-            h->non_zero_count_cache[0+8*1 +   8*i]=
-            h->non_zero_count_cache[0+8*4 +   8*i]= h->pps.cabac && !IS_INTRA(mb_type) ? 0 : 64;
-        }
-    }
-
-    if( h->pps.cabac ) {
-        // top_cbp
-        if(top_type) {
-            h->top_cbp = h->cbp_table[top_xy];
-        } else if(IS_INTRA(mb_type)) {
-            h->top_cbp = 0x1C0;
-        } else {
-            h->top_cbp = 0;
-        }
-        // left_cbp
-        if (left_type[0]) {
-            h->left_cbp = h->cbp_table[left_xy[0]] & 0x1f0;
-        } else if(IS_INTRA(mb_type)) {
-            h->left_cbp = 0x1C0;
-        } else {
-            h->left_cbp = 0;
-        }
-        if (left_type[0]) {
-            h->left_cbp |= ((h->cbp_table[left_xy[0]]>>((left_block[0]&(~1))+1))&0x1) << 1;
-        }
-        if (left_type[1]) {
-            h->left_cbp |= ((h->cbp_table[left_xy[1]]>>((left_block[2]&(~1))+1))&0x1) << 3;
-        }
-    }
-
-#if 1
-    if(IS_INTER(mb_type) || IS_DIRECT(mb_type)){
-        int list;
-        for(list=0; list<1+(h->slice_type==B_TYPE); list++){
-            if(!USES_LIST(mb_type, list) && !IS_DIRECT(mb_type) && !h->deblocking_filter){
-                /*if(!h->mv_cache_clean[list]){
-                    memset(h->mv_cache [list],  0, 8*5*2*sizeof(int16_t)); //FIXME clean only input? clean at all?
-                    memset(h->ref_cache[list], PART_NOT_AVAILABLE, 8*5*sizeof(int8_t));
-                    h->mv_cache_clean[list]= 1;
-                }*/
-                continue;
-            }
-            h->mv_cache_clean[list]= 0;
-
-            if(USES_LIST(top_type, list)){
-                const int b_xy= h->mb2b_xy[top_xy] + 3*h->b_stride;
-                const int b8_xy= h->mb2b8_xy[top_xy] + h->b8_stride;
-                *(uint32_t*)h->mv_cache[list][scan8[0] + 0 - 1*8]= *(uint32_t*)s->current_picture.motion_val[list][b_xy + 0];
-                *(uint32_t*)h->mv_cache[list][scan8[0] + 1 - 1*8]= *(uint32_t*)s->current_picture.motion_val[list][b_xy + 1];
-                *(uint32_t*)h->mv_cache[list][scan8[0] + 2 - 1*8]= *(uint32_t*)s->current_picture.motion_val[list][b_xy + 2];
-                *(uint32_t*)h->mv_cache[list][scan8[0] + 3 - 1*8]= *(uint32_t*)s->current_picture.motion_val[list][b_xy + 3];
-                h->ref_cache[list][scan8[0] + 0 - 1*8]=
-                h->ref_cache[list][scan8[0] + 1 - 1*8]= s->current_picture.ref_index[list][b8_xy + 0];
-                h->ref_cache[list][scan8[0] + 2 - 1*8]=
-                h->ref_cache[list][scan8[0] + 3 - 1*8]= s->current_picture.ref_index[list][b8_xy + 1];
-            }else{
-                *(uint32_t*)h->mv_cache [list][scan8[0] + 0 - 1*8]=
-                *(uint32_t*)h->mv_cache [list][scan8[0] + 1 - 1*8]=
-                *(uint32_t*)h->mv_cache [list][scan8[0] + 2 - 1*8]=
-                *(uint32_t*)h->mv_cache [list][scan8[0] + 3 - 1*8]= 0;
-                *(uint32_t*)&h->ref_cache[list][scan8[0] + 0 - 1*8]= ((top_type ? LIST_NOT_USED : PART_NOT_AVAILABLE)&0xFF)*0x01010101;
-            }
-
-            //FIXME unify cleanup or sth
-            if(USES_LIST(left_type[0], list)){
-                const int b_xy= h->mb2b_xy[left_xy[0]] + 3;
-                const int b8_xy= h->mb2b8_xy[left_xy[0]] + 1;
-                *(uint32_t*)h->mv_cache[list][scan8[0] - 1 + 0*8]= *(uint32_t*)s->current_picture.motion_val[list][b_xy + h->b_stride*left_block[0]];
-                *(uint32_t*)h->mv_cache[list][scan8[0] - 1 + 1*8]= *(uint32_t*)s->current_picture.motion_val[list][b_xy + h->b_stride*left_block[1]];
-                h->ref_cache[list][scan8[0] - 1 + 0*8]= s->current_picture.ref_index[list][b8_xy + h->b8_stride*(left_block[0]>>1)];
-                h->ref_cache[list][scan8[0] - 1 + 1*8]= s->current_picture.ref_index[list][b8_xy + h->b8_stride*(left_block[1]>>1)];
-            }else{
-                *(uint32_t*)h->mv_cache [list][scan8[0] - 1 + 0*8]=
-                *(uint32_t*)h->mv_cache [list][scan8[0] - 1 + 1*8]= 0;
-                h->ref_cache[list][scan8[0] - 1 + 0*8]=
-                h->ref_cache[list][scan8[0] - 1 + 1*8]= left_type[0] ? LIST_NOT_USED : PART_NOT_AVAILABLE;
-            }
-
-            if(USES_LIST(left_type[1], list)){
-                const int b_xy= h->mb2b_xy[left_xy[1]] + 3;
-                const int b8_xy= h->mb2b8_xy[left_xy[1]] + 1;
-                *(uint32_t*)h->mv_cache[list][scan8[0] - 1 + 2*8]= *(uint32_t*)s->current_picture.motion_val[list][b_xy + h->b_stride*left_block[2]];
-                *(uint32_t*)h->mv_cache[list][scan8[0] - 1 + 3*8]= *(uint32_t*)s->current_picture.motion_val[list][b_xy + h->b_stride*left_block[3]];
-                h->ref_cache[list][scan8[0] - 1 + 2*8]= s->current_picture.ref_index[list][b8_xy + h->b8_stride*(left_block[2]>>1)];
-                h->ref_cache[list][scan8[0] - 1 + 3*8]= s->current_picture.ref_index[list][b8_xy + h->b8_stride*(left_block[3]>>1)];
-            }else{
-                *(uint32_t*)h->mv_cache [list][scan8[0] - 1 + 2*8]=
-                *(uint32_t*)h->mv_cache [list][scan8[0] - 1 + 3*8]= 0;
-                h->ref_cache[list][scan8[0] - 1 + 2*8]=
-                h->ref_cache[list][scan8[0] - 1 + 3*8]= left_type[0] ? LIST_NOT_USED : PART_NOT_AVAILABLE;
-                assert((!left_type[0]) == (!left_type[1]));
-            }
-
-            if((for_deblock || (IS_DIRECT(mb_type) && !h->direct_spatial_mv_pred)) && !FRAME_MBAFF)
-                continue;
-
-            if(USES_LIST(topleft_type, list)){
-                const int b_xy = h->mb2b_xy[topleft_xy] + 3 + 3*h->b_stride;
-                const int b8_xy= h->mb2b8_xy[topleft_xy] + 1 + h->b8_stride;
-                *(uint32_t*)h->mv_cache[list][scan8[0] - 1 - 1*8]= *(uint32_t*)s->current_picture.motion_val[list][b_xy];
-                h->ref_cache[list][scan8[0] - 1 - 1*8]= s->current_picture.ref_index[list][b8_xy];
-            }else{
-                *(uint32_t*)h->mv_cache[list][scan8[0] - 1 - 1*8]= 0;
-                h->ref_cache[list][scan8[0] - 1 - 1*8]= topleft_type ? LIST_NOT_USED : PART_NOT_AVAILABLE;
-            }
-
-            if(USES_LIST(topright_type, list)){
-                const int b_xy= h->mb2b_xy[topright_xy] + 3*h->b_stride;
-                const int b8_xy= h->mb2b8_xy[topright_xy] + h->b8_stride;
-                *(uint32_t*)h->mv_cache[list][scan8[0] + 4 - 1*8]= *(uint32_t*)s->current_picture.motion_val[list][b_xy];
-                h->ref_cache[list][scan8[0] + 4 - 1*8]= s->current_picture.ref_index[list][b8_xy];
-            }else{
-                *(uint32_t*)h->mv_cache [list][scan8[0] + 4 - 1*8]= 0;
-                h->ref_cache[list][scan8[0] + 4 - 1*8]= topright_type ? LIST_NOT_USED : PART_NOT_AVAILABLE;
-            }
-
-            if((IS_SKIP(mb_type) || IS_DIRECT(mb_type)) && !FRAME_MBAFF)
-                continue;
-
-            h->ref_cache[list][scan8[5 ]+1] =
-            h->ref_cache[list][scan8[7 ]+1] =
-            h->ref_cache[list][scan8[13]+1] =  //FIXME remove past 3 (init somewhere else)
-            h->ref_cache[list][scan8[4 ]] =
-            h->ref_cache[list][scan8[12]] = PART_NOT_AVAILABLE;
-            *(uint32_t*)h->mv_cache [list][scan8[5 ]+1]=
-            *(uint32_t*)h->mv_cache [list][scan8[7 ]+1]=
-            *(uint32_t*)h->mv_cache [list][scan8[13]+1]= //FIXME remove past 3 (init somewhere else)
-            *(uint32_t*)h->mv_cache [list][scan8[4 ]]=
-            *(uint32_t*)h->mv_cache [list][scan8[12]]= 0;
-
-            if( h->pps.cabac ) {
-                /* XXX beurk, Load mvd */
-                if(USES_LIST(top_type, list)){
-                    const int b_xy= h->mb2b_xy[top_xy] + 3*h->b_stride;
-                    *(uint32_t*)h->mvd_cache[list][scan8[0] + 0 - 1*8]= *(uint32_t*)h->mvd_table[list][b_xy + 0];
-                    *(uint32_t*)h->mvd_cache[list][scan8[0] + 1 - 1*8]= *(uint32_t*)h->mvd_table[list][b_xy + 1];
-                    *(uint32_t*)h->mvd_cache[list][scan8[0] + 2 - 1*8]= *(uint32_t*)h->mvd_table[list][b_xy + 2];
-                    *(uint32_t*)h->mvd_cache[list][scan8[0] + 3 - 1*8]= *(uint32_t*)h->mvd_table[list][b_xy + 3];
-                }else{
-                    *(uint32_t*)h->mvd_cache [list][scan8[0] + 0 - 1*8]=
-                    *(uint32_t*)h->mvd_cache [list][scan8[0] + 1 - 1*8]=
-                    *(uint32_t*)h->mvd_cache [list][scan8[0] + 2 - 1*8]=
-                    *(uint32_t*)h->mvd_cache [list][scan8[0] + 3 - 1*8]= 0;
-                }
-                if(USES_LIST(left_type[0], list)){
-                    const int b_xy= h->mb2b_xy[left_xy[0]] + 3;
-                    *(uint32_t*)h->mvd_cache[list][scan8[0] - 1 + 0*8]= *(uint32_t*)h->mvd_table[list][b_xy + h->b_stride*left_block[0]];
-                    *(uint32_t*)h->mvd_cache[list][scan8[0] - 1 + 1*8]= *(uint32_t*)h->mvd_table[list][b_xy + h->b_stride*left_block[1]];
-                }else{
-                    *(uint32_t*)h->mvd_cache [list][scan8[0] - 1 + 0*8]=
-                    *(uint32_t*)h->mvd_cache [list][scan8[0] - 1 + 1*8]= 0;
-                }
-                if(USES_LIST(left_type[1], list)){
-                    const int b_xy= h->mb2b_xy[left_xy[1]] + 3;
-                    *(uint32_t*)h->mvd_cache[list][scan8[0] - 1 + 2*8]= *(uint32_t*)h->mvd_table[list][b_xy + h->b_stride*left_block[2]];
-                    *(uint32_t*)h->mvd_cache[list][scan8[0] - 1 + 3*8]= *(uint32_t*)h->mvd_table[list][b_xy + h->b_stride*left_block[3]];
-                }else{
-                    *(uint32_t*)h->mvd_cache [list][scan8[0] - 1 + 2*8]=
-                    *(uint32_t*)h->mvd_cache [list][scan8[0] - 1 + 3*8]= 0;
-                }
-                *(uint32_t*)h->mvd_cache [list][scan8[5 ]+1]=
-                *(uint32_t*)h->mvd_cache [list][scan8[7 ]+1]=
-                *(uint32_t*)h->mvd_cache [list][scan8[13]+1]= //FIXME remove past 3 (init somewhere else)
-                *(uint32_t*)h->mvd_cache [list][scan8[4 ]]=
-                *(uint32_t*)h->mvd_cache [list][scan8[12]]= 0;
-
-                if(h->slice_type == B_TYPE){
-                    fill_rectangle(&h->direct_cache[scan8[0]], 4, 4, 8, 0, 1);
-
-                    if(IS_DIRECT(top_type)){
-                        *(uint32_t*)&h->direct_cache[scan8[0] - 1*8]= 0x01010101;
-                    }else if(IS_8X8(top_type)){
-                        int b8_xy = h->mb2b8_xy[top_xy] + h->b8_stride;
-                        h->direct_cache[scan8[0] + 0 - 1*8]= h->direct_table[b8_xy];
-                        h->direct_cache[scan8[0] + 2 - 1*8]= h->direct_table[b8_xy + 1];
-                    }else{
-                        *(uint32_t*)&h->direct_cache[scan8[0] - 1*8]= 0;
-                    }
-
-                    if(IS_DIRECT(left_type[0]))
-                        h->direct_cache[scan8[0] - 1 + 0*8]= 1;
-                    else if(IS_8X8(left_type[0]))
-                        h->direct_cache[scan8[0] - 1 + 0*8]= h->direct_table[h->mb2b8_xy[left_xy[0]] + 1 + h->b8_stride*(left_block[0]>>1)];
-                    else
-                        h->direct_cache[scan8[0] - 1 + 0*8]= 0;
-
-                    if(IS_DIRECT(left_type[1]))
-                        h->direct_cache[scan8[0] - 1 + 2*8]= 1;
-                    else if(IS_8X8(left_type[1]))
-                        h->direct_cache[scan8[0] - 1 + 2*8]= h->direct_table[h->mb2b8_xy[left_xy[1]] + 1 + h->b8_stride*(left_block[2]>>1)];
-                    else
-                        h->direct_cache[scan8[0] - 1 + 2*8]= 0;
-                }
-            }
-
-            if(FRAME_MBAFF){
-#define MAP_MVS\
-                    MAP_F2F(scan8[0] - 1 - 1*8, topleft_type)\
-                    MAP_F2F(scan8[0] + 0 - 1*8, top_type)\
-                    MAP_F2F(scan8[0] + 1 - 1*8, top_type)\
-                    MAP_F2F(scan8[0] + 2 - 1*8, top_type)\
-                    MAP_F2F(scan8[0] + 3 - 1*8, top_type)\
-                    MAP_F2F(scan8[0] + 4 - 1*8, topright_type)\
-                    MAP_F2F(scan8[0] - 1 + 0*8, left_type[0])\
-                    MAP_F2F(scan8[0] - 1 + 1*8, left_type[0])\
-                    MAP_F2F(scan8[0] - 1 + 2*8, left_type[1])\
-                    MAP_F2F(scan8[0] - 1 + 3*8, left_type[1])
-                if(MB_FIELD){
-#define MAP_F2F(idx, mb_type)\
-                    if(!IS_INTERLACED(mb_type) && h->ref_cache[list][idx] >= 0){\
-                        h->ref_cache[list][idx] <<= 1;\
-                        h->mv_cache[list][idx][1] /= 2;\
-                        h->mvd_cache[list][idx][1] /= 2;\
-                    }
-                    MAP_MVS
-#undef MAP_F2F
-                }else{
-#define MAP_F2F(idx, mb_type)\
-                    if(IS_INTERLACED(mb_type) && h->ref_cache[list][idx] >= 0){\
-                        h->ref_cache[list][idx] >>= 1;\
-                        h->mv_cache[list][idx][1] <<= 1;\
-                        h->mvd_cache[list][idx][1] <<= 1;\
-                    }
-                    MAP_MVS
-#undef MAP_F2F
-                }
-            }
-        }
-    }
-#endif
-
-    h->neighbor_transform_size= !!IS_8x8DCT(top_type) + !!IS_8x8DCT(left_type[0]);
-}
-
-static inline void write_back_intra_pred_mode(H264Context *h){
-    MpegEncContext * const s = &h->s;
-    const int mb_xy= s->mb_x + s->mb_y*s->mb_stride;
-
-    h->intra4x4_pred_mode[mb_xy][0]= h->intra4x4_pred_mode_cache[7+8*1];
-    h->intra4x4_pred_mode[mb_xy][1]= h->intra4x4_pred_mode_cache[7+8*2];
-    h->intra4x4_pred_mode[mb_xy][2]= h->intra4x4_pred_mode_cache[7+8*3];
-    h->intra4x4_pred_mode[mb_xy][3]= h->intra4x4_pred_mode_cache[7+8*4];
-    h->intra4x4_pred_mode[mb_xy][4]= h->intra4x4_pred_mode_cache[4+8*4];
-    h->intra4x4_pred_mode[mb_xy][5]= h->intra4x4_pred_mode_cache[5+8*4];
-    h->intra4x4_pred_mode[mb_xy][6]= h->intra4x4_pred_mode_cache[6+8*4];
-}
-
-/**
- * checks if the top & left blocks are available if needed & changes the dc mode so it only uses the available blocks.
- */
-static inline int check_intra4x4_pred_mode(H264Context *h){
-    MpegEncContext * const s = &h->s;
-    static const int8_t top [12]= {-1, 0,LEFT_DC_PRED,-1,-1,-1,-1,-1, 0};
-    static const int8_t left[12]= { 0,-1, TOP_DC_PRED, 0,-1,-1,-1, 0,-1,DC_128_PRED};
-    int i;
-
-    if(!(h->top_samples_available&0x8000)){
-        for(i=0; i<4; i++){
-            int status= top[ h->intra4x4_pred_mode_cache[scan8[0] + i] ];
-            if(status<0){
-                av_log(h->s.avctx, AV_LOG_ERROR, "top block unavailable for requested intra4x4 mode %d at %d %d\n", status, s->mb_x, s->mb_y);
-                return -1;
-            } else if(status){
-                h->intra4x4_pred_mode_cache[scan8[0] + i]= status;
-            }
-        }
-    }
-
-    if(!(h->left_samples_available&0x8000)){
-        for(i=0; i<4; i++){
-            int status= left[ h->intra4x4_pred_mode_cache[scan8[0] + 8*i] ];
-            if(status<0){
-                av_log(h->s.avctx, AV_LOG_ERROR, "left block unavailable for requested intra4x4 mode %d at %d %d\n", status, s->mb_x, s->mb_y);
-                return -1;
-            } else if(status){
-                h->intra4x4_pred_mode_cache[scan8[0] + 8*i]= status;
-            }
-        }
-    }
-
-    return 0;
-} //FIXME cleanup like next
-
-/**
- * checks if the top & left blocks are available if needed & changes the dc mode so it only uses the available blocks.
- */
-static inline int check_intra_pred_mode(H264Context *h, int mode){
-    MpegEncContext * const s = &h->s;
-    static const int8_t top [7]= {LEFT_DC_PRED8x8, 1,-1,-1};
-    static const int8_t left[7]= { TOP_DC_PRED8x8,-1, 2,-1,DC_128_PRED8x8};
-
-    if(mode < 0 || mode > 6) {
-        av_log(h->s.avctx, AV_LOG_ERROR, "out of range intra chroma pred mode at %d %d\n", s->mb_x, s->mb_y);
-        return -1;
-    }
-
-    if(!(h->top_samples_available&0x8000)){
-        mode= top[ mode ];
-        if(mode<0){
-            av_log(h->s.avctx, AV_LOG_ERROR, "top block unavailable for requested intra mode at %d %d\n", s->mb_x, s->mb_y);
-            return -1;
-        }
-    }
-
-    if(!(h->left_samples_available&0x8000)){
-        mode= left[ mode ];
-        if(mode<0){
-            av_log(h->s.avctx, AV_LOG_ERROR, "left block unavailable for requested intra mode at %d %d\n", s->mb_x, s->mb_y);
-            return -1;
-        }
-    }
-
-    return mode;
-}
-
-/**
- * gets the predicted intra4x4 prediction mode.
- */
-static inline int pred_intra_mode(H264Context *h, int n){
-    const int index8= scan8[n];
-    const int left= h->intra4x4_pred_mode_cache[index8 - 1];
-    const int top = h->intra4x4_pred_mode_cache[index8 - 8];
-    const int min= FFMIN(left, top);
-
-    tprintf("mode:%d %d min:%d\n", left ,top, min);
-
-    if(min<0) return DC_PRED;
-    else      return min;
-}
-
-static inline void write_back_non_zero_count(H264Context *h){
-    MpegEncContext * const s = &h->s;
-    const int mb_xy= s->mb_x + s->mb_y*s->mb_stride;
-
-    h->non_zero_count[mb_xy][0]= h->non_zero_count_cache[7+8*1];
-    h->non_zero_count[mb_xy][1]= h->non_zero_count_cache[7+8*2];
-    h->non_zero_count[mb_xy][2]= h->non_zero_count_cache[7+8*3];
-    h->non_zero_count[mb_xy][3]= h->non_zero_count_cache[7+8*4];
-    h->non_zero_count[mb_xy][4]= h->non_zero_count_cache[4+8*4];
-    h->non_zero_count[mb_xy][5]= h->non_zero_count_cache[5+8*4];
-    h->non_zero_count[mb_xy][6]= h->non_zero_count_cache[6+8*4];
-
-    h->non_zero_count[mb_xy][9]= h->non_zero_count_cache[1+8*2];
-    h->non_zero_count[mb_xy][8]= h->non_zero_count_cache[2+8*2];
-    h->non_zero_count[mb_xy][7]= h->non_zero_count_cache[2+8*1];
-
-    h->non_zero_count[mb_xy][12]=h->non_zero_count_cache[1+8*5];
-    h->non_zero_count[mb_xy][11]=h->non_zero_count_cache[2+8*5];
-    h->non_zero_count[mb_xy][10]=h->non_zero_count_cache[2+8*4];
-
-    if(FRAME_MBAFF){
-        // store all luma nnzs, for deblocking
-        int v = 0, i;
-        for(i=0; i<16; i++)
-            v += (!!h->non_zero_count_cache[scan8[i]]) << i;
-        *(uint16_t*)&h->non_zero_count[mb_xy][14] = v;
-    }
-}
-
-/**
- * gets the predicted number of non zero coefficients.
- * @param n block index
- */
-static inline int pred_non_zero_count(H264Context *h, int n){
-    const int index8= scan8[n];
-    const int left= h->non_zero_count_cache[index8 - 1];
-    const int top = h->non_zero_count_cache[index8 - 8];
-    int i= left + top;
-
-    if(i<64) i= (i+1)>>1;
-
-    tprintf("pred_nnz L%X T%X n%d s%d P%X\n", left, top, n, scan8[n], i&31);
-
-    return i&31;
-}
-
-static inline int fetch_diagonal_mv(H264Context *h, const int16_t **C, int i, int list, int part_width){
-    const int topright_ref= h->ref_cache[list][ i - 8 + part_width ];
-
-    /* there is no consistent mapping of mvs to neighboring locations that will
-     * make mbaff happy, so we can't move all this logic to fill_caches */
-    if(FRAME_MBAFF){
-        MpegEncContext *s = &h->s;
-        const uint32_t *mb_types = s->current_picture_ptr->mb_type;
-        const int16_t *mv;
-        *(uint32_t*)h->mv_cache[list][scan8[0]-2] = 0;
-        *C = h->mv_cache[list][scan8[0]-2];
-
-        if(!MB_FIELD
-           && (s->mb_y&1) && i < scan8[0]+8 && topright_ref != PART_NOT_AVAILABLE){
-            int topright_xy = s->mb_x + (s->mb_y-1)*s->mb_stride + (i == scan8[0]+3);
-            if(IS_INTERLACED(mb_types[topright_xy])){
-#define SET_DIAG_MV(MV_OP, REF_OP, X4, Y4)\
-                const int x4 = X4, y4 = Y4;\
-                const int mb_type = mb_types[(x4>>2)+(y4>>2)*s->mb_stride];\
-                if(!USES_LIST(mb_type,list) && !IS_8X8(mb_type))\
-                    return LIST_NOT_USED;\
-                mv = s->current_picture_ptr->motion_val[list][x4 + y4*h->b_stride];\
-                h->mv_cache[list][scan8[0]-2][0] = mv[0];\
-                h->mv_cache[list][scan8[0]-2][1] = mv[1] MV_OP;\
-                return s->current_picture_ptr->ref_index[list][(x4>>1) + (y4>>1)*h->b8_stride] REF_OP;
-
-                SET_DIAG_MV(*2, >>1, s->mb_x*4+(i&7)-4+part_width, s->mb_y*4-1);
-            }
-        }
-        if(topright_ref == PART_NOT_AVAILABLE
-           && ((s->mb_y&1) || i >= scan8[0]+8) && (i&7)==4
-           && h->ref_cache[list][scan8[0]-1] != PART_NOT_AVAILABLE){
-            if(!MB_FIELD
-               && IS_INTERLACED(mb_types[h->left_mb_xy[0]])){
-                SET_DIAG_MV(*2, >>1, s->mb_x*4-1, (s->mb_y|1)*4+(s->mb_y&1)*2+(i>>4)-1);
-            }
-            if(MB_FIELD
-               && !IS_INTERLACED(mb_types[h->left_mb_xy[0]])
-               && i >= scan8[0]+8){
-                // leftshift will turn LIST_NOT_USED into PART_NOT_AVAILABLE, but that's ok.
-                SET_DIAG_MV(>>1, <<1, s->mb_x*4-1, (s->mb_y&~1)*4 - 1 + ((i-scan8[0])>>3)*2);
-            }
-        }
-#undef SET_DIAG_MV
-    }
-
-    if(topright_ref != PART_NOT_AVAILABLE){
-        *C= h->mv_cache[list][ i - 8 + part_width ];
-        return topright_ref;
-    }else{
-        tprintf("topright MV not available\n");
-
-        *C= h->mv_cache[list][ i - 8 - 1 ];
-        return h->ref_cache[list][ i - 8 - 1 ];
-    }
-}
-
-/**
- * gets the predicted MV.
- * @param n the block index
- * @param part_width the width of the partition (4, 8,16) -> (1, 2, 4)
- * @param mx the x component of the predicted motion vector
- * @param my the y component of the predicted motion vector
- */
-static inline void pred_motion(H264Context * const h, int n, int part_width, int list, int ref, int * const mx, int * const my){
-    const int index8= scan8[n];
-    const int top_ref=      h->ref_cache[list][ index8 - 8 ];
-    const int left_ref=     h->ref_cache[list][ index8 - 1 ];
-    const int16_t * const A= h->mv_cache[list][ index8 - 1 ];
-    const int16_t * const B= h->mv_cache[list][ index8 - 8 ];
-    const int16_t * C;
-    int diagonal_ref, match_count;
-
-    assert(part_width==1 || part_width==2 || part_width==4);
-
-/* mv_cache
-  B . . A T T T T
-  U . . L . . , .
-  U . . L . . . .
-  U . . L . . , .
-  . . . L . . . .
-*/
-
-    diagonal_ref= fetch_diagonal_mv(h, &C, index8, list, part_width);
-    match_count= (diagonal_ref==ref) + (top_ref==ref) + (left_ref==ref);
-    tprintf("pred_motion match_count=%d\n", match_count);
-    if(match_count > 1){ //most common
-        *mx= mid_pred(A[0], B[0], C[0]);
-        *my= mid_pred(A[1], B[1], C[1]);
-    }else if(match_count==1){
-        if(left_ref==ref){
-            *mx= A[0];
-            *my= A[1];
-        }else if(top_ref==ref){
-            *mx= B[0];
-            *my= B[1];
-        }else{
-            *mx= C[0];
-            *my= C[1];
-        }
-    }else{
-        if(top_ref == PART_NOT_AVAILABLE && diagonal_ref == PART_NOT_AVAILABLE && left_ref != PART_NOT_AVAILABLE){
-            *mx= A[0];
-            *my= A[1];
-        }else{
-            *mx= mid_pred(A[0], B[0], C[0]);
-            *my= mid_pred(A[1], B[1], C[1]);
-        }
-    }
-
-    tprintf("pred_motion (%2d %2d %2d) (%2d %2d %2d) (%2d %2d %2d) -> (%2d %2d %2d) at %2d %2d %d list %d\n", top_ref, B[0], B[1],                    diagonal_ref, C[0], C[1], left_ref, A[0], A[1], ref, *mx, *my, h->s.mb_x, h->s.mb_y, n, list);
-}
-
-/**
- * gets the directionally predicted 16x8 MV.
- * @param n the block index
- * @param mx the x component of the predicted motion vector
- * @param my the y component of the predicted motion vector
- */
-static inline void pred_16x8_motion(H264Context * const h, int n, int list, int ref, int * const mx, int * const my){
-    if(n==0){
-        const int top_ref=      h->ref_cache[list][ scan8[0] - 8 ];
-        const int16_t * const B= h->mv_cache[list][ scan8[0] - 8 ];
-
-        tprintf("pred_16x8: (%2d %2d %2d) at %2d %2d %d list %d\n", top_ref, B[0], B[1], h->s.mb_x, h->s.mb_y, n, list);
-
-        if(top_ref == ref){
-            *mx= B[0];
-            *my= B[1];
-            return;
-        }
-    }else{
-        const int left_ref=     h->ref_cache[list][ scan8[8] - 1 ];
-        const int16_t * const A= h->mv_cache[list][ scan8[8] - 1 ];
-
-        tprintf("pred_16x8: (%2d %2d %2d) at %2d %2d %d list %d\n", left_ref, A[0], A[1], h->s.mb_x, h->s.mb_y, n, list);
-
-        if(left_ref == ref){
-            *mx= A[0];
-            *my= A[1];
-            return;
-        }
-    }
-
-    //RARE
-    pred_motion(h, n, 4, list, ref, mx, my);
-}
-
-/**
- * gets the directionally predicted 8x16 MV.
- * @param n the block index
- * @param mx the x component of the predicted motion vector
- * @param my the y component of the predicted motion vector
- */
-static inline void pred_8x16_motion(H264Context * const h, int n, int list, int ref, int * const mx, int * const my){
-    if(n==0){
-        const int left_ref=      h->ref_cache[list][ scan8[0] - 1 ];
-        const int16_t * const A=  h->mv_cache[list][ scan8[0] - 1 ];
-
-        tprintf("pred_8x16: (%2d %2d %2d) at %2d %2d %d list %d\n", left_ref, A[0], A[1], h->s.mb_x, h->s.mb_y, n, list);
-
-        if(left_ref == ref){
-            *mx= A[0];
-            *my= A[1];
-            return;
-        }
-    }else{
-        const int16_t * C;
-        int diagonal_ref;
-
-        diagonal_ref= fetch_diagonal_mv(h, &C, scan8[4], list, 2);
-
-        tprintf("pred_8x16: (%2d %2d %2d) at %2d %2d %d list %d\n", diagonal_ref, C[0], C[1], h->s.mb_x, h->s.mb_y, n, list);
-
-        if(diagonal_ref == ref){
-            *mx= C[0];
-            *my= C[1];
-            return;
-        }
-    }
-
-    //RARE
-    pred_motion(h, n, 2, list, ref, mx, my);
-}
-
-static inline void pred_pskip_motion(H264Context * const h, int * const mx, int * const my){
-    const int top_ref = h->ref_cache[0][ scan8[0] - 8 ];
-    const int left_ref= h->ref_cache[0][ scan8[0] - 1 ];
-
-    tprintf("pred_pskip: (%d) (%d) at %2d %2d\n", top_ref, left_ref, h->s.mb_x, h->s.mb_y);
-
-    if(top_ref == PART_NOT_AVAILABLE || left_ref == PART_NOT_AVAILABLE
-       || (top_ref == 0  && *(uint32_t*)h->mv_cache[0][ scan8[0] - 8 ] == 0)
-       || (left_ref == 0 && *(uint32_t*)h->mv_cache[0][ scan8[0] - 1 ] == 0)){
-
-        *mx = *my = 0;
-        return;
-    }
-
-    pred_motion(h, 0, 4, 0, 0, mx, my);
-
-    return;
-}
-
-static inline void direct_dist_scale_factor(H264Context * const h){
-    const int poc = h->s.current_picture_ptr->poc;
-    const int poc1 = h->ref_list[1][0].poc;
-    int i;
-    for(i=0; i<h->ref_count[0]; i++){
-        int poc0 = h->ref_list[0][i].poc;
-        int td = clip(poc1 - poc0, -128, 127);
-        if(td == 0 /* FIXME || pic0 is a long-term ref */){
-            h->dist_scale_factor[i] = 256;
-        }else{
-            int tb = clip(poc - poc0, -128, 127);
-            int tx = (16384 + (FFABS(td) >> 1)) / td;
-            h->dist_scale_factor[i] = clip((tb*tx + 32) >> 6, -1024, 1023);
-        }
-    }
-    if(FRAME_MBAFF){
-        for(i=0; i<h->ref_count[0]; i++){
-            h->dist_scale_factor_field[2*i] =
-            h->dist_scale_factor_field[2*i+1] = h->dist_scale_factor[i];
-        }
-    }
-}
-static inline void direct_ref_list_init(H264Context * const h){
-    MpegEncContext * const s = &h->s;
-    Picture * const ref1 = &h->ref_list[1][0];
-    Picture * const cur = s->current_picture_ptr;
-    int list, i, j;
-    if(cur->pict_type == I_TYPE)
-        cur->ref_count[0] = 0;
-    if(cur->pict_type != B_TYPE)
-        cur->ref_count[1] = 0;
-    for(list=0; list<2; list++){
-        cur->ref_count[list] = h->ref_count[list];
-        for(j=0; j<h->ref_count[list]; j++)
-            cur->ref_poc[list][j] = h->ref_list[list][j].poc;
-    }
-    if(cur->pict_type != B_TYPE || h->direct_spatial_mv_pred)
-        return;
-    for(list=0; list<2; list++){
-        for(i=0; i<ref1->ref_count[list]; i++){
-            const int poc = ref1->ref_poc[list][i];
-            h->map_col_to_list0[list][i] = 0; /* bogus; fills in for missing frames */
-            for(j=0; j<h->ref_count[list]; j++)
-                if(h->ref_list[list][j].poc == poc){
-                    h->map_col_to_list0[list][i] = j;
-                    break;
-                }
-        }
-    }
-    if(FRAME_MBAFF){
-        for(list=0; list<2; list++){
-            for(i=0; i<ref1->ref_count[list]; i++){
-                j = h->map_col_to_list0[list][i];
-                h->map_col_to_list0_field[list][2*i] = 2*j;
-                h->map_col_to_list0_field[list][2*i+1] = 2*j+1;
-            }
-        }
-    }
-}
-
-static inline void pred_direct_motion(H264Context * const h, int *mb_type){
-    MpegEncContext * const s = &h->s;
-    const int mb_xy =   s->mb_x +   s->mb_y*s->mb_stride;
-    const int b8_xy = 2*s->mb_x + 2*s->mb_y*h->b8_stride;
-    const int b4_xy = 4*s->mb_x + 4*s->mb_y*h->b_stride;
-    const int mb_type_col = h->ref_list[1][0].mb_type[mb_xy];
-    const int16_t (*l1mv0)[2] = (const int16_t (*)[2]) &h->ref_list[1][0].motion_val[0][b4_xy];
-    const int16_t (*l1mv1)[2] = (const int16_t (*)[2]) &h->ref_list[1][0].motion_val[1][b4_xy];
-    const int8_t *l1ref0 = &h->ref_list[1][0].ref_index[0][b8_xy];
-    const int8_t *l1ref1 = &h->ref_list[1][0].ref_index[1][b8_xy];
-    const int is_b8x8 = IS_8X8(*mb_type);
-    int sub_mb_type;
-    int i8, i4;
-
-#define MB_TYPE_16x16_OR_INTRA (MB_TYPE_16x16|MB_TYPE_INTRA4x4|MB_TYPE_INTRA16x16|MB_TYPE_INTRA_PCM)
-    if(IS_8X8(mb_type_col) && !h->sps.direct_8x8_inference_flag){
-        /* FIXME save sub mb types from previous frames (or derive from MVs)
-         * so we know exactly what block size to use */
-        sub_mb_type = MB_TYPE_8x8|MB_TYPE_P0L0|MB_TYPE_P0L1|MB_TYPE_DIRECT2; /* B_SUB_4x4 */
-        *mb_type =    MB_TYPE_8x8|MB_TYPE_L0L1;
-    }else if(!is_b8x8 && (mb_type_col & MB_TYPE_16x16_OR_INTRA)){
-        sub_mb_type = MB_TYPE_16x16|MB_TYPE_P0L0|MB_TYPE_P0L1|MB_TYPE_DIRECT2; /* B_SUB_8x8 */
-        *mb_type =    MB_TYPE_16x16|MB_TYPE_P0L0|MB_TYPE_P0L1|MB_TYPE_DIRECT2; /* B_16x16 */
-    }else{
-        sub_mb_type = MB_TYPE_16x16|MB_TYPE_P0L0|MB_TYPE_P0L1|MB_TYPE_DIRECT2; /* B_SUB_8x8 */
-        *mb_type =    MB_TYPE_8x8|MB_TYPE_L0L1;
-    }
-    if(!is_b8x8)
-        *mb_type |= MB_TYPE_DIRECT2;
-    if(MB_FIELD)
-        *mb_type |= MB_TYPE_INTERLACED;
-
-    tprintf("mb_type = %08x, sub_mb_type = %08x, is_b8x8 = %d, mb_type_col = %08x\n", *mb_type, sub_mb_type, is_b8x8, mb_type_col);
-
-    if(h->direct_spatial_mv_pred){
-        int ref[2];
-        int mv[2][2];
-        int list;
-
-        /* FIXME interlacing + spatial direct uses wrong colocated block positions */
-
-        /* ref = min(neighbors) */
-        for(list=0; list<2; list++){
-            int refa = h->ref_cache[list][scan8[0] - 1];
-            int refb = h->ref_cache[list][scan8[0] - 8];
-            int refc = h->ref_cache[list][scan8[0] - 8 + 4];
-            if(refc == -2)
-                refc = h->ref_cache[list][scan8[0] - 8 - 1];
-            ref[list] = refa;
-            if(ref[list] < 0 || (refb < ref[list] && refb >= 0))
-                ref[list] = refb;
-            if(ref[list] < 0 || (refc < ref[list] && refc >= 0))
-                ref[list] = refc;
-            if(ref[list] < 0)
-                ref[list] = -1;
-        }
-
-        if(ref[0] < 0 && ref[1] < 0){
-            ref[0] = ref[1] = 0;
-            mv[0][0] = mv[0][1] =
-            mv[1][0] = mv[1][1] = 0;
-        }else{
-            for(list=0; list<2; list++){
-                if(ref[list] >= 0)
-                    pred_motion(h, 0, 4, list, ref[list], &mv[list][0], &mv[list][1]);
-                else
-                    mv[list][0] = mv[list][1] = 0;
-            }
-        }
-
-        if(ref[1] < 0){
-            *mb_type &= ~MB_TYPE_P0L1;
-            sub_mb_type &= ~MB_TYPE_P0L1;
-        }else if(ref[0] < 0){
-            *mb_type &= ~MB_TYPE_P0L0;
-            sub_mb_type &= ~MB_TYPE_P0L0;
-        }
-
-        if(IS_16X16(*mb_type)){
-            fill_rectangle(&h->ref_cache[0][scan8[0]], 4, 4, 8, (uint8_t)ref[0], 1);
-            fill_rectangle(&h->ref_cache[1][scan8[0]], 4, 4, 8, (uint8_t)ref[1], 1);
-            if(!IS_INTRA(mb_type_col)
-               && (   (l1ref0[0] == 0 && FFABS(l1mv0[0][0]) <= 1 && FFABS(l1mv0[0][1]) <= 1)
-                   || (l1ref0[0]  < 0 && l1ref1[0] == 0 && FFABS(l1mv1[0][0]) <= 1 && FFABS(l1mv1[0][1]) <= 1
-                       && (h->x264_build>33 || !h->x264_build)))){
-                if(ref[0] > 0)
-                    fill_rectangle(&h->mv_cache[0][scan8[0]], 4, 4, 8, pack16to32(mv[0][0],mv[0][1]), 4);
-                else
-                    fill_rectangle(&h->mv_cache[0][scan8[0]], 4, 4, 8, 0, 4);
-                if(ref[1] > 0)
-                    fill_rectangle(&h->mv_cache[1][scan8[0]], 4, 4, 8, pack16to32(mv[1][0],mv[1][1]), 4);
-                else
-                    fill_rectangle(&h->mv_cache[1][scan8[0]], 4, 4, 8, 0, 4);
-            }else{
-                fill_rectangle(&h->mv_cache[0][scan8[0]], 4, 4, 8, pack16to32(mv[0][0],mv[0][1]), 4);
-                fill_rectangle(&h->mv_cache[1][scan8[0]], 4, 4, 8, pack16to32(mv[1][0],mv[1][1]), 4);
-            }
-        }else{
-            for(i8=0; i8<4; i8++){
-                const int x8 = i8&1;
-                const int y8 = i8>>1;
-
-                if(is_b8x8 && !IS_DIRECT(h->sub_mb_type[i8]))
-                    continue;
-                h->sub_mb_type[i8] = sub_mb_type;
-
-                fill_rectangle(&h->mv_cache[0][scan8[i8*4]], 2, 2, 8, pack16to32(mv[0][0],mv[0][1]), 4);
-                fill_rectangle(&h->mv_cache[1][scan8[i8*4]], 2, 2, 8, pack16to32(mv[1][0],mv[1][1]), 4);
-                fill_rectangle(&h->ref_cache[0][scan8[i8*4]], 2, 2, 8, (uint8_t)ref[0], 1);
-                fill_rectangle(&h->ref_cache[1][scan8[i8*4]], 2, 2, 8, (uint8_t)ref[1], 1);
-
-                /* col_zero_flag */
-                if(!IS_INTRA(mb_type_col) && (   l1ref0[x8 + y8*h->b8_stride] == 0
-                                              || (l1ref0[x8 + y8*h->b8_stride] < 0 && l1ref1[x8 + y8*h->b8_stride] == 0
-                                                  && (h->x264_build>33 || !h->x264_build)))){
-                    const int16_t (*l1mv)[2]= l1ref0[x8 + y8*h->b8_stride] == 0 ? l1mv0 : l1mv1;
-                    if(IS_SUB_8X8(sub_mb_type)){
-                        const int16_t *mv_col = l1mv[x8*3 + y8*3*h->b_stride];
-                        if(FFABS(mv_col[0]) <= 1 && FFABS(mv_col[1]) <= 1){
-                            if(ref[0] == 0)
-                                fill_rectangle(&h->mv_cache[0][scan8[i8*4]], 2, 2, 8, 0, 4);
-                            if(ref[1] == 0)
-                                fill_rectangle(&h->mv_cache[1][scan8[i8*4]], 2, 2, 8, 0, 4);
-                        }
-                    }else
-                    for(i4=0; i4<4; i4++){
-                        const int16_t *mv_col = l1mv[x8*2 + (i4&1) + (y8*2 + (i4>>1))*h->b_stride];
-                        if(FFABS(mv_col[0]) <= 1 && FFABS(mv_col[1]) <= 1){
-                            if(ref[0] == 0)
-                                *(uint32_t*)h->mv_cache[0][scan8[i8*4+i4]] = 0;
-                            if(ref[1] == 0)
-                                *(uint32_t*)h->mv_cache[1][scan8[i8*4+i4]] = 0;
-                        }
-                    }
-                }
-            }
-        }
-    }else{ /* direct temporal mv pred */
-        const int *map_col_to_list0[2] = {h->map_col_to_list0[0], h->map_col_to_list0[1]};
-        const int *dist_scale_factor = h->dist_scale_factor;
-
-        if(FRAME_MBAFF){
-            if(IS_INTERLACED(*mb_type)){
-                map_col_to_list0[0] = h->map_col_to_list0_field[0];
-                map_col_to_list0[1] = h->map_col_to_list0_field[1];
-                dist_scale_factor = h->dist_scale_factor_field;
-            }
-            if(IS_INTERLACED(*mb_type) != IS_INTERLACED(mb_type_col)){
-                /* FIXME assumes direct_8x8_inference == 1 */
-                const int pair_xy = s->mb_x + (s->mb_y&~1)*s->mb_stride;
-                int mb_types_col[2];
-                int y_shift;
-
-                *mb_type = MB_TYPE_8x8|MB_TYPE_L0L1
-                         | (is_b8x8 ? 0 : MB_TYPE_DIRECT2)
-                         | (*mb_type & MB_TYPE_INTERLACED);
-                sub_mb_type = MB_TYPE_P0L0|MB_TYPE_P0L1|MB_TYPE_DIRECT2|MB_TYPE_16x16;
-
-                if(IS_INTERLACED(*mb_type)){
-                    /* frame to field scaling */
-                    mb_types_col[0] = h->ref_list[1][0].mb_type[pair_xy];
-                    mb_types_col[1] = h->ref_list[1][0].mb_type[pair_xy+s->mb_stride];
-                    if(s->mb_y&1){
-                        l1ref0 -= 2*h->b8_stride;
-                        l1ref1 -= 2*h->b8_stride;
-                        l1mv0 -= 4*h->b_stride;
-                        l1mv1 -= 4*h->b_stride;
-                    }
-                    y_shift = 0;
-
-                    if(   (mb_types_col[0] & MB_TYPE_16x16_OR_INTRA)
-                       && (mb_types_col[1] & MB_TYPE_16x16_OR_INTRA)
-                       && !is_b8x8)
-                        *mb_type |= MB_TYPE_16x8;
-                    else
-                        *mb_type |= MB_TYPE_8x8;
-                }else{
-                    /* field to frame scaling */
-                    /* col_mb_y = (mb_y&~1) + (topAbsDiffPOC < bottomAbsDiffPOC ? 0 : 1)
-                     * but in MBAFF, top and bottom POC are equal */
-                    int dy = (s->mb_y&1) ? 1 : 2;
-                    mb_types_col[0] =
-                    mb_types_col[1] = h->ref_list[1][0].mb_type[pair_xy+s->mb_stride];
-                    l1ref0 += dy*h->b8_stride;
-                    l1ref1 += dy*h->b8_stride;
-                    l1mv0 += 2*dy*h->b_stride;
-                    l1mv1 += 2*dy*h->b_stride;
-                    y_shift = 2;
-
-                    if((mb_types_col[0] & (MB_TYPE_16x16_OR_INTRA|MB_TYPE_16x8))
-                       && !is_b8x8)
-                        *mb_type |= MB_TYPE_16x16;
-                    else
-                        *mb_type |= MB_TYPE_8x8;
-                }
-
-                for(i8=0; i8<4; i8++){
-                    const int x8 = i8&1;
-                    const int y8 = i8>>1;
-                    int ref0, scale;
-                    const int16_t (*l1mv)[2]= l1mv0;
-
-                    if(is_b8x8 && !IS_DIRECT(h->sub_mb_type[i8]))
-                        continue;
-                    h->sub_mb_type[i8] = sub_mb_type;
-
-                    fill_rectangle(&h->ref_cache[1][scan8[i8*4]], 2, 2, 8, 0, 1);
-                    if(IS_INTRA(mb_types_col[y8])){
-                        fill_rectangle(&h->ref_cache[0][scan8[i8*4]], 2, 2, 8, 0, 1);
-                        fill_rectangle(&h-> mv_cache[0][scan8[i8*4]], 2, 2, 8, 0, 4);
-                        fill_rectangle(&h-> mv_cache[1][scan8[i8*4]], 2, 2, 8, 0, 4);
-                        continue;
-                    }
-
-                    ref0 = l1ref0[x8 + (y8*2>>y_shift)*h->b8_stride];
-                    if(ref0 >= 0)
-                        ref0 = map_col_to_list0[0][ref0*2>>y_shift];
-                    else{
-                        ref0 = map_col_to_list0[1][l1ref1[x8 + (y8*2>>y_shift)*h->b8_stride]*2>>y_shift];
-                        l1mv= l1mv1;
-                    }
-                    scale = dist_scale_factor[ref0];
-                    fill_rectangle(&h->ref_cache[0][scan8[i8*4]], 2, 2, 8, ref0, 1);
-
-                    {
-                        const int16_t *mv_col = l1mv[x8*3 + (y8*6>>y_shift)*h->b_stride];
-                        int my_col = (mv_col[1]<<y_shift)/2;
-                        int mx = (scale * mv_col[0] + 128) >> 8;
-                        int my = (scale * my_col + 128) >> 8;
-                        fill_rectangle(&h->mv_cache[0][scan8[i8*4]], 2, 2, 8, pack16to32(mx,my), 4);
-                        fill_rectangle(&h->mv_cache[1][scan8[i8*4]], 2, 2, 8, pack16to32(mx-mv_col[0],my-my_col), 4);
-                    }
-                }
-                return;
-            }
-        }
-
-        /* one-to-one mv scaling */
-
-        if(IS_16X16(*mb_type)){
-            fill_rectangle(&h->ref_cache[1][scan8[0]], 4, 4, 8, 0, 1);
-            if(IS_INTRA(mb_type_col)){
-                fill_rectangle(&h->ref_cache[0][scan8[0]], 4, 4, 8, 0, 1);
-                fill_rectangle(&h-> mv_cache[0][scan8[0]], 4, 4, 8, 0, 4);
-                fill_rectangle(&h-> mv_cache[1][scan8[0]], 4, 4, 8, 0, 4);
-            }else{
-                const int ref0 = l1ref0[0] >= 0 ? map_col_to_list0[0][l1ref0[0]]
-                                                : map_col_to_list0[1][l1ref1[0]];
-                const int scale = dist_scale_factor[ref0];
-                const int16_t *mv_col = l1ref0[0] >= 0 ? l1mv0[0] : l1mv1[0];
-                int mv_l0[2];
-                mv_l0[0] = (scale * mv_col[0] + 128) >> 8;
-                mv_l0[1] = (scale * mv_col[1] + 128) >> 8;
-                fill_rectangle(&h->ref_cache[0][scan8[0]], 4, 4, 8, ref0, 1);
-                fill_rectangle(&h-> mv_cache[0][scan8[0]], 4, 4, 8, pack16to32(mv_l0[0],mv_l0[1]), 4);
-                fill_rectangle(&h-> mv_cache[1][scan8[0]], 4, 4, 8, pack16to32(mv_l0[0]-mv_col[0],mv_l0[1]-mv_col[1]), 4);
-            }
-        }else{
-            for(i8=0; i8<4; i8++){
-                const int x8 = i8&1;
-                const int y8 = i8>>1;
-                int ref0, scale;
-                const int16_t (*l1mv)[2]= l1mv0;
-
-                if(is_b8x8 && !IS_DIRECT(h->sub_mb_type[i8]))
-                    continue;
-                h->sub_mb_type[i8] = sub_mb_type;
-                fill_rectangle(&h->ref_cache[1][scan8[i8*4]], 2, 2, 8, 0, 1);
-                if(IS_INTRA(mb_type_col)){
-                    fill_rectangle(&h->ref_cache[0][scan8[i8*4]], 2, 2, 8, 0, 1);
-                    fill_rectangle(&h-> mv_cache[0][scan8[i8*4]], 2, 2, 8, 0, 4);
-                    fill_rectangle(&h-> mv_cache[1][scan8[i8*4]], 2, 2, 8, 0, 4);
-                    continue;
-                }
-
-                ref0 = l1ref0[x8 + y8*h->b8_stride];
-                if(ref0 >= 0)
-                    ref0 = map_col_to_list0[0][ref0];
-                else{
-                    ref0 = map_col_to_list0[1][l1ref1[x8 + y8*h->b8_stride]];
-                    l1mv= l1mv1;
-                }
-                scale = dist_scale_factor[ref0];
-
-                fill_rectangle(&h->ref_cache[0][scan8[i8*4]], 2, 2, 8, ref0, 1);
-                if(IS_SUB_8X8(sub_mb_type)){
-                    const int16_t *mv_col = l1mv[x8*3 + y8*3*h->b_stride];
-                    int mx = (scale * mv_col[0] + 128) >> 8;
-                    int my = (scale * mv_col[1] + 128) >> 8;
-                    fill_rectangle(&h->mv_cache[0][scan8[i8*4]], 2, 2, 8, pack16to32(mx,my), 4);
-                    fill_rectangle(&h->mv_cache[1][scan8[i8*4]], 2, 2, 8, pack16to32(mx-mv_col[0],my-mv_col[1]), 4);
-                }else
-                for(i4=0; i4<4; i4++){
-                    const int16_t *mv_col = l1mv[x8*2 + (i4&1) + (y8*2 + (i4>>1))*h->b_stride];
-                    int16_t *mv_l0 = h->mv_cache[0][scan8[i8*4+i4]];
-                    mv_l0[0] = (scale * mv_col[0] + 128) >> 8;
-                    mv_l0[1] = (scale * mv_col[1] + 128) >> 8;
-                    *(uint32_t*)h->mv_cache[1][scan8[i8*4+i4]] =
-                        pack16to32(mv_l0[0]-mv_col[0],mv_l0[1]-mv_col[1]);
-                }
-            }
-        }
-    }
-}
-
-static inline void write_back_motion(H264Context *h, int mb_type){
-    MpegEncContext * const s = &h->s;
-    const int b_xy = 4*s->mb_x + 4*s->mb_y*h->b_stride;
-    const int b8_xy= 2*s->mb_x + 2*s->mb_y*h->b8_stride;
-    int list;
-
-    if(!USES_LIST(mb_type, 0))
-        fill_rectangle(&s->current_picture.ref_index[0][b8_xy], 2, 2, h->b8_stride, (uint8_t)LIST_NOT_USED, 1);
-
-    for(list=0; list<2; list++){
-        int y;
-        if(!USES_LIST(mb_type, list))
-            continue;
-
-        for(y=0; y<4; y++){
-            *(uint64_t*)s->current_picture.motion_val[list][b_xy + 0 + y*h->b_stride]= *(uint64_t*)h->mv_cache[list][scan8[0]+0 + 8*y];
-            *(uint64_t*)s->current_picture.motion_val[list][b_xy + 2 + y*h->b_stride]= *(uint64_t*)h->mv_cache[list][scan8[0]+2 + 8*y];
-        }
-        if( h->pps.cabac ) {
-            if(IS_SKIP(mb_type))
-                fill_rectangle(h->mvd_table[list][b_xy], 4, 4, h->b_stride, 0, 4);
-            else
-            for(y=0; y<4; y++){
-                *(uint64_t*)h->mvd_table[list][b_xy + 0 + y*h->b_stride]= *(uint64_t*)h->mvd_cache[list][scan8[0]+0 + 8*y];
-                *(uint64_t*)h->mvd_table[list][b_xy + 2 + y*h->b_stride]= *(uint64_t*)h->mvd_cache[list][scan8[0]+2 + 8*y];
-            }
-        }
-
-        {
-            int8_t *ref_index = &s->current_picture.ref_index[list][b8_xy];
-            ref_index[0+0*h->b8_stride]= h->ref_cache[list][scan8[0]];
-            ref_index[1+0*h->b8_stride]= h->ref_cache[list][scan8[4]];
-            ref_index[0+1*h->b8_stride]= h->ref_cache[list][scan8[8]];
-            ref_index[1+1*h->b8_stride]= h->ref_cache[list][scan8[12]];
-        }
-    }
-
-    if(h->slice_type == B_TYPE && h->pps.cabac){
-        if(IS_8X8(mb_type)){
-            uint8_t *direct_table = &h->direct_table[b8_xy];
-            direct_table[1+0*h->b8_stride] = IS_DIRECT(h->sub_mb_type[1]) ? 1 : 0;
-            direct_table[0+1*h->b8_stride] = IS_DIRECT(h->sub_mb_type[2]) ? 1 : 0;
-            direct_table[1+1*h->b8_stride] = IS_DIRECT(h->sub_mb_type[3]) ? 1 : 0;
-        }
-    }
-}
-
-/**
- * Decodes a network abstraction layer unit.
- * @param consumed is the number of bytes used as input
- * @param length is the length of the array
- * @param dst_length is the number of decoded bytes FIXME here or a decode rbsp tailing?
- * @returns decoded bytes, might be src+1 if no escapes
- */
-static uint8_t *decode_nal(H264Context *h, uint8_t *src, int *dst_length, int *consumed, int length){
-    int i, si, di;
-    uint8_t *dst;
-
-//    src[0]&0x80;                //forbidden bit
-    h->nal_ref_idc= src[0]>>5;
-    h->nal_unit_type= src[0]&0x1F;
-
-    src++; length--;
-#if 0
-    for(i=0; i<length; i++)
-        printf("%2X ", src[i]);
-#endif
-    for(i=0; i+1<length; i+=2){
-        if(src[i]) continue;
-        if(i>0 && src[i-1]==0) i--;
-        if(i+2<length && src[i+1]==0 && src[i+2]<=3){
-            if(src[i+2]!=3){
-                /* startcode, so we must be past the end */
-                length=i;
-            }
-            break;
-        }
-    }
-
-    if(i>=length-1){ //no escaped 0
-        *dst_length= length;
-        *consumed= length+1; //+1 for the header
-        return src;
-    }
-
-    h->rbsp_buffer= av_fast_realloc(h->rbsp_buffer, &h->rbsp_buffer_size, length);
-    dst= h->rbsp_buffer;
-
-//printf("decoding esc\n");
-    si=di=0;
-    while(si<length){
-        //remove escapes (very rare 1:2^22)
-        if(si+2<length && src[si]==0 && src[si+1]==0 && src[si+2]<=3){
-            if(src[si+2]==3){ //escape
-                dst[di++]= 0;
-                dst[di++]= 0;
-                si+=3;
-                continue;
-            }else //next start code
-                break;
-        }
-
-        dst[di++]= src[si++];
-    }
-
-    *dst_length= di;
-    *consumed= si + 1;//+1 for the header
-//FIXME store exact number of bits in the getbitcontext (its needed for decoding)
-    return dst;
-}
-
-#if 0
-/**
- * @param src the data which should be escaped
- * @param dst the target buffer, dst+1 == src is allowed as a special case
- * @param length the length of the src data
- * @param dst_length the length of the dst array
- * @returns length of escaped data in bytes or -1 if an error occured
- */
-static int encode_nal(H264Context *h, uint8_t *dst, uint8_t *src, int length, int dst_length){
-    int i, escape_count, si, di;
-    uint8_t *temp;
-
-    assert(length>=0);
-    assert(dst_length>0);
-
-    dst[0]= (h->nal_ref_idc<<5) + h->nal_unit_type;
-
-    if(length==0) return 1;
-
-    escape_count= 0;
-    for(i=0; i<length; i+=2){
-        if(src[i]) continue;
-        if(i>0 && src[i-1]==0)
-            i--;
-        if(i+2<length && src[i+1]==0 && src[i+2]<=3){
-            escape_count++;
-            i+=2;
-        }
-    }
-
-    if(escape_count==0){
-        if(dst+1 != src)
-            memcpy(dst+1, src, length);
-        return length + 1;
-    }
-
-    if(length + escape_count + 1> dst_length)
-        return -1;
-
-    //this should be damn rare (hopefully)
-
-    h->rbsp_buffer= av_fast_realloc(h->rbsp_buffer, &h->rbsp_buffer_size, length + escape_count);
-    temp= h->rbsp_buffer;
-//printf("encoding esc\n");
-
-    si= 0;
-    di= 0;
-    while(si < length){
-        if(si+2<length && src[si]==0 && src[si+1]==0 && src[si+2]<=3){
-            temp[di++]= 0; si++;
-            temp[di++]= 0; si++;
-            temp[di++]= 3;
-            temp[di++]= src[si++];
-        }
-        else
-            temp[di++]= src[si++];
-    }
-    memcpy(dst+1, temp, length+escape_count);
-
-    assert(di == length+escape_count);
-
-    return di + 1;
-}
-
-/**
- * write 1,10,100,1000,... for alignment, yes its exactly inverse to mpeg4
- */
-static void encode_rbsp_trailing(PutBitContext *pb){
-    int length;
-    put_bits(pb, 1, 1);
-    length= (-put_bits_count(pb))&7;
-    if(length) put_bits(pb, length, 0);
-}
-#endif
-
-/**
- * identifies the exact end of the bitstream
- * @return the length of the trailing, or 0 if damaged
- */
-static int decode_rbsp_trailing(uint8_t *src){
-    int v= *src;
-    int r;
-
-    tprintf("rbsp trailing %X\n", v);
-
-    for(r=1; r<9; r++){
-        if(v&1) return r;
-        v>>=1;
-    }
-    return 0;
-}
-
-/**
- * idct tranforms the 16 dc values and dequantize them.
- * @param qp quantization parameter
- */
-static void h264_luma_dc_dequant_idct_c(DCTELEM *block, int qp, int qmul){
-#define stride 16
-    int i;
-    int temp[16]; //FIXME check if this is a good idea
-    static const int x_offset[4]={0, 1*stride, 4* stride,  5*stride};
-    static const int y_offset[4]={0, 2*stride, 8* stride, 10*stride};
-
-//memset(block, 64, 2*256);
-//return;
-    for(i=0; i<4; i++){
-        const int offset= y_offset[i];
-        const int z0= block[offset+stride*0] + block[offset+stride*4];
-        const int z1= block[offset+stride*0] - block[offset+stride*4];
-        const int z2= block[offset+stride*1] - block[offset+stride*5];
-        const int z3= block[offset+stride*1] + block[offset+stride*5];
-
-        temp[4*i+0]= z0+z3;
-        temp[4*i+1]= z1+z2;
-        temp[4*i+2]= z1-z2;
-        temp[4*i+3]= z0-z3;
-    }
-
-    for(i=0; i<4; i++){
-        const int offset= x_offset[i];
-        const int z0= temp[4*0+i] + temp[4*2+i];
-        const int z1= temp[4*0+i] - temp[4*2+i];
-        const int z2= temp[4*1+i] - temp[4*3+i];
-        const int z3= temp[4*1+i] + temp[4*3+i];
-
-        block[stride*0 +offset]= ((((z0 + z3)*qmul + 128 ) >> 8)); //FIXME think about merging this into decode_resdual
-        block[stride*2 +offset]= ((((z1 + z2)*qmul + 128 ) >> 8));
-        block[stride*8 +offset]= ((((z1 - z2)*qmul + 128 ) >> 8));
-        block[stride*10+offset]= ((((z0 - z3)*qmul + 128 ) >> 8));
-    }
-}
-
-#if 0
-/**
- * dct tranforms the 16 dc values.
- * @param qp quantization parameter ??? FIXME
- */
-static void h264_luma_dc_dct_c(DCTELEM *block/*, int qp*/){
-//    const int qmul= dequant_coeff[qp][0];
-    int i;
-    int temp[16]; //FIXME check if this is a good idea
-    static const int x_offset[4]={0, 1*stride, 4* stride,  5*stride};
-    static const int y_offset[4]={0, 2*stride, 8* stride, 10*stride};
-
-    for(i=0; i<4; i++){
-        const int offset= y_offset[i];
-        const int z0= block[offset+stride*0] + block[offset+stride*4];
-        const int z1= block[offset+stride*0] - block[offset+stride*4];
-        const int z2= block[offset+stride*1] - block[offset+stride*5];
-        const int z3= block[offset+stride*1] + block[offset+stride*5];
-
-        temp[4*i+0]= z0+z3;
-        temp[4*i+1]= z1+z2;
-        temp[4*i+2]= z1-z2;
-        temp[4*i+3]= z0-z3;
-    }
-
-    for(i=0; i<4; i++){
-        const int offset= x_offset[i];
-        const int z0= temp[4*0+i] + temp[4*2+i];
-        const int z1= temp[4*0+i] - temp[4*2+i];
-        const int z2= temp[4*1+i] - temp[4*3+i];
-        const int z3= temp[4*1+i] + temp[4*3+i];
-
-        block[stride*0 +offset]= (z0 + z3)>>1;
-        block[stride*2 +offset]= (z1 + z2)>>1;
-        block[stride*8 +offset]= (z1 - z2)>>1;
-        block[stride*10+offset]= (z0 - z3)>>1;
-    }
-}
-#endif
-
-#undef xStride
-#undef stride
-
-static void chroma_dc_dequant_idct_c(DCTELEM *block, int qp, int qmul){
-    const int stride= 16*2;
-    const int xStride= 16;
-    int a,b,c,d,e;
-
-    a= block[stride*0 + xStride*0];
-    b= block[stride*0 + xStride*1];
-    c= block[stride*1 + xStride*0];
-    d= block[stride*1 + xStride*1];
-
-    e= a-b;
-    a= a+b;
-    b= c-d;
-    c= c+d;
-
-    block[stride*0 + xStride*0]= ((a+c)*qmul) >> 7;
-    block[stride*0 + xStride*1]= ((e+b)*qmul) >> 7;
-    block[stride*1 + xStride*0]= ((a-c)*qmul) >> 7;
-    block[stride*1 + xStride*1]= ((e-b)*qmul) >> 7;
-}
-
-#if 0
-static void chroma_dc_dct_c(DCTELEM *block){
-    const int stride= 16*2;
-    const int xStride= 16;
-    int a,b,c,d,e;
-
-    a= block[stride*0 + xStride*0];
-    b= block[stride*0 + xStride*1];
-    c= block[stride*1 + xStride*0];
-    d= block[stride*1 + xStride*1];
-
-    e= a-b;
-    a= a+b;
-    b= c-d;
-    c= c+d;
-
-    block[stride*0 + xStride*0]= (a+c);
-    block[stride*0 + xStride*1]= (e+b);
-    block[stride*1 + xStride*0]= (a-c);
-    block[stride*1 + xStride*1]= (e-b);
-}
-#endif
-
-/**
- * gets the chroma qp.
- */
-static inline int get_chroma_qp(int chroma_qp_index_offset, int qscale){
-
-    return chroma_qp[clip(qscale + chroma_qp_index_offset, 0, 51)];
-}
-
-
-#if 0
-static void h264_diff_dct_c(DCTELEM *block, uint8_t *src1, uint8_t *src2, int stride){
-    int i;
-    //FIXME try int temp instead of block
-
-    for(i=0; i<4; i++){
-        const int d0= src1[0 + i*stride] - src2[0 + i*stride];
-        const int d1= src1[1 + i*stride] - src2[1 + i*stride];
-        const int d2= src1[2 + i*stride] - src2[2 + i*stride];
-        const int d3= src1[3 + i*stride] - src2[3 + i*stride];
-        const int z0= d0 + d3;
-        const int z3= d0 - d3;
-        const int z1= d1 + d2;
-        const int z2= d1 - d2;
-
-        block[0 + 4*i]=   z0 +   z1;
-        block[1 + 4*i]= 2*z3 +   z2;
-        block[2 + 4*i]=   z0 -   z1;
-        block[3 + 4*i]=   z3 - 2*z2;
-    }
-
-    for(i=0; i<4; i++){
-        const int z0= block[0*4 + i] + block[3*4 + i];
-        const int z3= block[0*4 + i] - block[3*4 + i];
-        const int z1= block[1*4 + i] + block[2*4 + i];
-        const int z2= block[1*4 + i] - block[2*4 + i];
-
-        block[0*4 + i]=   z0 +   z1;
-        block[1*4 + i]= 2*z3 +   z2;
-        block[2*4 + i]=   z0 -   z1;
-        block[3*4 + i]=   z3 - 2*z2;
-    }
-}
-#endif
-
-//FIXME need to check that this doesnt overflow signed 32 bit for low qp, i am not sure, it's very close
-//FIXME check that gcc inlines this (and optimizes intra & seperate_dc stuff away)
-static inline int quantize_c(DCTELEM *block, uint8_t *scantable, int qscale, int intra, int seperate_dc){
-    int i;
-    const int * const quant_table= quant_coeff[qscale];
-    const int bias= intra ? (1<<QUANT_SHIFT)/3 : (1<<QUANT_SHIFT)/6;
-    const unsigned int threshold1= (1<<QUANT_SHIFT) - bias - 1;
-    const unsigned int threshold2= (threshold1<<1);
-    int last_non_zero;
-
-    if(seperate_dc){
-        if(qscale<=18){
-            //avoid overflows
-            const int dc_bias= intra ? (1<<(QUANT_SHIFT-2))/3 : (1<<(QUANT_SHIFT-2))/6;
-            const unsigned int dc_threshold1= (1<<(QUANT_SHIFT-2)) - dc_bias - 1;
-            const unsigned int dc_threshold2= (dc_threshold1<<1);
-
-            int level= block[0]*quant_coeff[qscale+18][0];
-            if(((unsigned)(level+dc_threshold1))>dc_threshold2){
-                if(level>0){
-                    level= (dc_bias + level)>>(QUANT_SHIFT-2);
-                    block[0]= level;
-                }else{
-                    level= (dc_bias - level)>>(QUANT_SHIFT-2);
-                    block[0]= -level;
-                }
-//                last_non_zero = i;
-            }else{
-                block[0]=0;
-            }
-        }else{
-            const int dc_bias= intra ? (1<<(QUANT_SHIFT+1))/3 : (1<<(QUANT_SHIFT+1))/6;
-            const unsigned int dc_threshold1= (1<<(QUANT_SHIFT+1)) - dc_bias - 1;
-            const unsigned int dc_threshold2= (dc_threshold1<<1);
-
-            int level= block[0]*quant_table[0];
-            if(((unsigned)(level+dc_threshold1))>dc_threshold2){
-                if(level>0){
-                    level= (dc_bias + level)>>(QUANT_SHIFT+1);
-                    block[0]= level;
-                }else{
-                    level= (dc_bias - level)>>(QUANT_SHIFT+1);
-                    block[0]= -level;
-                }
-//                last_non_zero = i;
-            }else{
-                block[0]=0;
-            }
-        }
-        last_non_zero= 0;
-        i=1;
-    }else{
-        last_non_zero= -1;
-        i=0;
-    }
-
-    for(; i<16; i++){
-        const int j= scantable[i];
-        int level= block[j]*quant_table[j];
-
-//        if(   bias+level >= (1<<(QMAT_SHIFT - 3))
-//           || bias-level >= (1<<(QMAT_SHIFT - 3))){
-        if(((unsigned)(level+threshold1))>threshold2){
-            if(level>0){
-                level= (bias + level)>>QUANT_SHIFT;
-                block[j]= level;
-            }else{
-                level= (bias - level)>>QUANT_SHIFT;
-                block[j]= -level;
-            }
-            last_non_zero = i;
-        }else{
-            block[j]=0;
-        }
-    }
-
-    return last_non_zero;
-}
-
-static void pred4x4_vertical_c(uint8_t *src, uint8_t *topright, int stride){
-    const uint32_t a= ((uint32_t*)(src-stride))[0];
-    ((uint32_t*)(src+0*stride))[0]= a;
-    ((uint32_t*)(src+1*stride))[0]= a;
-    ((uint32_t*)(src+2*stride))[0]= a;
-    ((uint32_t*)(src+3*stride))[0]= a;
-}
-
-static void pred4x4_horizontal_c(uint8_t *src, uint8_t *topright, int stride){
-    ((uint32_t*)(src+0*stride))[0]= src[-1+0*stride]*0x01010101;
-    ((uint32_t*)(src+1*stride))[0]= src[-1+1*stride]*0x01010101;
-    ((uint32_t*)(src+2*stride))[0]= src[-1+2*stride]*0x01010101;
-    ((uint32_t*)(src+3*stride))[0]= src[-1+3*stride]*0x01010101;
-}
-
-static void pred4x4_dc_c(uint8_t *src, uint8_t *topright, int stride){
-    const int dc= (  src[-stride] + src[1-stride] + src[2-stride] + src[3-stride]
-                   + src[-1+0*stride] + src[-1+1*stride] + src[-1+2*stride] + src[-1+3*stride] + 4) >>3;
-
-    ((uint32_t*)(src+0*stride))[0]=
-    ((uint32_t*)(src+1*stride))[0]=
-    ((uint32_t*)(src+2*stride))[0]=
-    ((uint32_t*)(src+3*stride))[0]= dc* 0x01010101;
-}
-
-static void pred4x4_left_dc_c(uint8_t *src, uint8_t *topright, int stride){
-    const int dc= (  src[-1+0*stride] + src[-1+1*stride] + src[-1+2*stride] + src[-1+3*stride] + 2) >>2;
-
-    ((uint32_t*)(src+0*stride))[0]=
-    ((uint32_t*)(src+1*stride))[0]=
-    ((uint32_t*)(src+2*stride))[0]=
-    ((uint32_t*)(src+3*stride))[0]= dc* 0x01010101;
-}
-
-static void pred4x4_top_dc_c(uint8_t *src, uint8_t *topright, int stride){
-    const int dc= (  src[-stride] + src[1-stride] + src[2-stride] + src[3-stride] + 2) >>2;
-
-    ((uint32_t*)(src+0*stride))[0]=
-    ((uint32_t*)(src+1*stride))[0]=
-    ((uint32_t*)(src+2*stride))[0]=
-    ((uint32_t*)(src+3*stride))[0]= dc* 0x01010101;
-}
-
-static void pred4x4_128_dc_c(uint8_t *src, uint8_t *topright, int stride){
-    ((uint32_t*)(src+0*stride))[0]=
-    ((uint32_t*)(src+1*stride))[0]=
-    ((uint32_t*)(src+2*stride))[0]=
-    ((uint32_t*)(src+3*stride))[0]= 128U*0x01010101U;
-}
-
-
-#define LOAD_TOP_RIGHT_EDGE\
-    const int t4= topright[0];\
-    const int t5= topright[1];\
-    const int t6= topright[2];\
-    const int t7= topright[3];\
-
-#define LOAD_LEFT_EDGE\
-    const int l0= src[-1+0*stride];\
-    const int l1= src[-1+1*stride];\
-    const int l2= src[-1+2*stride];\
-    const int l3= src[-1+3*stride];\
-
-#define LOAD_TOP_EDGE\
-    const int t0= src[ 0-1*stride];\
-    const int t1= src[ 1-1*stride];\
-    const int t2= src[ 2-1*stride];\
-    const int t3= src[ 3-1*stride];\
-
-static void pred4x4_down_right_c(uint8_t *src, uint8_t *topright, int stride){
-    const int lt= src[-1-1*stride];
-    LOAD_TOP_EDGE
-    LOAD_LEFT_EDGE
-
-    src[0+3*stride]=(l3 + 2*l2 + l1 + 2)>>2;
-    src[0+2*stride]=
-    src[1+3*stride]=(l2 + 2*l1 + l0 + 2)>>2;
-    src[0+1*stride]=
-    src[1+2*stride]=
-    src[2+3*stride]=(l1 + 2*l0 + lt + 2)>>2;
-    src[0+0*stride]=
-    src[1+1*stride]=
-    src[2+2*stride]=
-    src[3+3*stride]=(l0 + 2*lt + t0 + 2)>>2;
-    src[1+0*stride]=
-    src[2+1*stride]=
-    src[3+2*stride]=(lt + 2*t0 + t1 + 2)>>2;
-    src[2+0*stride]=
-    src[3+1*stride]=(t0 + 2*t1 + t2 + 2)>>2;
-    src[3+0*stride]=(t1 + 2*t2 + t3 + 2)>>2;
-}
-
-static void pred4x4_down_left_c(uint8_t *src, uint8_t *topright, int stride){
-    LOAD_TOP_EDGE
-    LOAD_TOP_RIGHT_EDGE
-//    LOAD_LEFT_EDGE
-
-    src[0+0*stride]=(t0 + t2 + 2*t1 + 2)>>2;
-    src[1+0*stride]=
-    src[0+1*stride]=(t1 + t3 + 2*t2 + 2)>>2;
-    src[2+0*stride]=
-    src[1+1*stride]=
-    src[0+2*stride]=(t2 + t4 + 2*t3 + 2)>>2;
-    src[3+0*stride]=
-    src[2+1*stride]=
-    src[1+2*stride]=
-    src[0+3*stride]=(t3 + t5 + 2*t4 + 2)>>2;
-    src[3+1*stride]=
-    src[2+2*stride]=
-    src[1+3*stride]=(t4 + t6 + 2*t5 + 2)>>2;
-    src[3+2*stride]=
-    src[2+3*stride]=(t5 + t7 + 2*t6 + 2)>>2;
-    src[3+3*stride]=(t6 + 3*t7 + 2)>>2;
-}
-
-static void pred4x4_vertical_right_c(uint8_t *src, uint8_t *topright, int stride){
-    const int lt= src[-1-1*stride];
-    LOAD_TOP_EDGE
-    LOAD_LEFT_EDGE
-    const __attribute__((unused)) int unu= l3;
-
-    src[0+0*stride]=
-    src[1+2*stride]=(lt + t0 + 1)>>1;
-    src[1+0*stride]=
-    src[2+2*stride]=(t0 + t1 + 1)>>1;
-    src[2+0*stride]=
-    src[3+2*stride]=(t1 + t2 + 1)>>1;
-    src[3+0*stride]=(t2 + t3 + 1)>>1;
-    src[0+1*stride]=
-    src[1+3*stride]=(l0 + 2*lt + t0 + 2)>>2;
-    src[1+1*stride]=
-    src[2+3*stride]=(lt + 2*t0 + t1 + 2)>>2;
-    src[2+1*stride]=
-    src[3+3*stride]=(t0 + 2*t1 + t2 + 2)>>2;
-    src[3+1*stride]=(t1 + 2*t2 + t3 + 2)>>2;
-    src[0+2*stride]=(lt + 2*l0 + l1 + 2)>>2;
-    src[0+3*stride]=(l0 + 2*l1 + l2 + 2)>>2;
-}
-
-static void pred4x4_vertical_left_c(uint8_t *src, uint8_t *topright, int stride){
-    LOAD_TOP_EDGE
-    LOAD_TOP_RIGHT_EDGE
-    const __attribute__((unused)) int unu= t7;
-
-    src[0+0*stride]=(t0 + t1 + 1)>>1;
-    src[1+0*stride]=
-    src[0+2*stride]=(t1 + t2 + 1)>>1;
-    src[2+0*stride]=
-    src[1+2*stride]=(t2 + t3 + 1)>>1;
-    src[3+0*stride]=
-    src[2+2*stride]=(t3 + t4+ 1)>>1;
-    src[3+2*stride]=(t4 + t5+ 1)>>1;
-    src[0+1*stride]=(t0 + 2*t1 + t2 + 2)>>2;
-    src[1+1*stride]=
-    src[0+3*stride]=(t1 + 2*t2 + t3 + 2)>>2;
-    src[2+1*stride]=
-    src[1+3*stride]=(t2 + 2*t3 + t4 + 2)>>2;
-    src[3+1*stride]=
-    src[2+3*stride]=(t3 + 2*t4 + t5 + 2)>>2;
-    src[3+3*stride]=(t4 + 2*t5 + t6 + 2)>>2;
-}
-
-static void pred4x4_horizontal_up_c(uint8_t *src, uint8_t *topright, int stride){
-    LOAD_LEFT_EDGE
-
-    src[0+0*stride]=(l0 + l1 + 1)>>1;
-    src[1+0*stride]=(l0 + 2*l1 + l2 + 2)>>2;
-    src[2+0*stride]=
-    src[0+1*stride]=(l1 + l2 + 1)>>1;
-    src[3+0*stride]=
-    src[1+1*stride]=(l1 + 2*l2 + l3 + 2)>>2;
-    src[2+1*stride]=
-    src[0+2*stride]=(l2 + l3 + 1)>>1;
-    src[3+1*stride]=
-    src[1+2*stride]=(l2 + 2*l3 + l3 + 2)>>2;
-    src[3+2*stride]=
-    src[1+3*stride]=
-    src[0+3*stride]=
-    src[2+2*stride]=
-    src[2+3*stride]=
-    src[3+3*stride]=l3;
-}
-
-static void pred4x4_horizontal_down_c(uint8_t *src, uint8_t *topright, int stride){
-    const int lt= src[-1-1*stride];
-    LOAD_TOP_EDGE
-    LOAD_LEFT_EDGE
-    const __attribute__((unused)) int unu= t3;
-
-    src[0+0*stride]=
-    src[2+1*stride]=(lt + l0 + 1)>>1;
-    src[1+0*stride]=
-    src[3+1*stride]=(l0 + 2*lt + t0 + 2)>>2;
-    src[2+0*stride]=(lt + 2*t0 + t1 + 2)>>2;
-    src[3+0*stride]=(t0 + 2*t1 + t2 + 2)>>2;
-    src[0+1*stride]=
-    src[2+2*stride]=(l0 + l1 + 1)>>1;
-    src[1+1*stride]=
-    src[3+2*stride]=(lt + 2*l0 + l1 + 2)>>2;
-    src[0+2*stride]=
-    src[2+3*stride]=(l1 + l2+ 1)>>1;
-    src[1+2*stride]=
-    src[3+3*stride]=(l0 + 2*l1 + l2 + 2)>>2;
-    src[0+3*stride]=(l2 + l3 + 1)>>1;
-    src[1+3*stride]=(l1 + 2*l2 + l3 + 2)>>2;
-}
-
-static void pred16x16_vertical_c(uint8_t *src, int stride){
-    int i;
-    const uint32_t a= ((uint32_t*)(src-stride))[0];
-    const uint32_t b= ((uint32_t*)(src-stride))[1];
-    const uint32_t c= ((uint32_t*)(src-stride))[2];
-    const uint32_t d= ((uint32_t*)(src-stride))[3];
-
-    for(i=0; i<16; i++){
-        ((uint32_t*)(src+i*stride))[0]= a;
-        ((uint32_t*)(src+i*stride))[1]= b;
-        ((uint32_t*)(src+i*stride))[2]= c;
-        ((uint32_t*)(src+i*stride))[3]= d;
-    }
-}
-
-static void pred16x16_horizontal_c(uint8_t *src, int stride){
-    int i;
-
-    for(i=0; i<16; i++){
-        ((uint32_t*)(src+i*stride))[0]=
-        ((uint32_t*)(src+i*stride))[1]=
-        ((uint32_t*)(src+i*stride))[2]=
-        ((uint32_t*)(src+i*stride))[3]= src[-1+i*stride]*0x01010101;
-    }
-}
-
-static void pred16x16_dc_c(uint8_t *src, int stride){
-    int i, dc=0;
-
-    for(i=0;i<16; i++){
-        dc+= src[-1+i*stride];
-    }
-
-    for(i=0;i<16; i++){
-        dc+= src[i-stride];
-    }
-
-    dc= 0x01010101*((dc + 16)>>5);
-
-    for(i=0; i<16; i++){
-        ((uint32_t*)(src+i*stride))[0]=
-        ((uint32_t*)(src+i*stride))[1]=
-        ((uint32_t*)(src+i*stride))[2]=
-        ((uint32_t*)(src+i*stride))[3]= dc;
-    }
-}
-
-static void pred16x16_left_dc_c(uint8_t *src, int stride){
-    int i, dc=0;
-
-    for(i=0;i<16; i++){
-        dc+= src[-1+i*stride];
-    }
-
-    dc= 0x01010101*((dc + 8)>>4);
-
-    for(i=0; i<16; i++){
-        ((uint32_t*)(src+i*stride))[0]=
-        ((uint32_t*)(src+i*stride))[1]=
-        ((uint32_t*)(src+i*stride))[2]=
-        ((uint32_t*)(src+i*stride))[3]= dc;
-    }
-}
-
-static void pred16x16_top_dc_c(uint8_t *src, int stride){
-    int i, dc=0;
-
-    for(i=0;i<16; i++){
-        dc+= src[i-stride];
-    }
-    dc= 0x01010101*((dc + 8)>>4);
-
-    for(i=0; i<16; i++){
-        ((uint32_t*)(src+i*stride))[0]=
-        ((uint32_t*)(src+i*stride))[1]=
-        ((uint32_t*)(src+i*stride))[2]=
-        ((uint32_t*)(src+i*stride))[3]= dc;
-    }
-}
-
-static void pred16x16_128_dc_c(uint8_t *src, int stride){
-    int i;
-
-    for(i=0; i<16; i++){
-        ((uint32_t*)(src+i*stride))[0]=
-        ((uint32_t*)(src+i*stride))[1]=
-        ((uint32_t*)(src+i*stride))[2]=
-        ((uint32_t*)(src+i*stride))[3]= 0x01010101U*128U;
-    }
-}
-
-static inline void pred16x16_plane_compat_c(uint8_t *src, int stride, const int svq3){
-  int i, j, k;
-  int a;
-  uint8_t *cm = cropTbl + MAX_NEG_CROP;
-  const uint8_t * const src0 = src+7-stride;
-  const uint8_t *src1 = src+8*stride-1;
-  const uint8_t *src2 = src1-2*stride;      // == src+6*stride-1;
-  int H = src0[1] - src0[-1];
-  int V = src1[0] - src2[ 0];
-  for(k=2; k<=8; ++k) {
-    src1 += stride; src2 -= stride;
-    H += k*(src0[k] - src0[-k]);
-    V += k*(src1[0] - src2[ 0]);
-  }
-  if(svq3){
-    H = ( 5*(H/4) ) / 16;
-    V = ( 5*(V/4) ) / 16;
-
-    /* required for 100% accuracy */
-    i = H; H = V; V = i;
-  }else{
-    H = ( 5*H+32 ) >> 6;
-    V = ( 5*V+32 ) >> 6;
-  }
-
-  a = 16*(src1[0] + src2[16] + 1) - 7*(V+H);
-  for(j=16; j>0; --j) {
-    int b = a;
-    a += V;
-    for(i=-16; i<0; i+=4) {
-      src[16+i] = cm[ (b    ) >> 5 ];
-      src[17+i] = cm[ (b+  H) >> 5 ];
-      src[18+i] = cm[ (b+2*H) >> 5 ];
-      src[19+i] = cm[ (b+3*H) >> 5 ];
-      b += 4*H;
-    }
-    src += stride;
-  }
-}
-
-static void pred16x16_plane_c(uint8_t *src, int stride){
-    pred16x16_plane_compat_c(src, stride, 0);
-}
-
-static void pred8x8_vertical_c(uint8_t *src, int stride){
-    int i;
-    const uint32_t a= ((uint32_t*)(src-stride))[0];
-    const uint32_t b= ((uint32_t*)(src-stride))[1];
-
-    for(i=0; i<8; i++){
-        ((uint32_t*)(src+i*stride))[0]= a;
-        ((uint32_t*)(src+i*stride))[1]= b;
-    }
-}
-
-static void pred8x8_horizontal_c(uint8_t *src, int stride){
-    int i;
-
-    for(i=0; i<8; i++){
-        ((uint32_t*)(src+i*stride))[0]=
-        ((uint32_t*)(src+i*stride))[1]= src[-1+i*stride]*0x01010101;
-    }
-}
-
-static void pred8x8_128_dc_c(uint8_t *src, int stride){
-    int i;
-
-    for(i=0; i<8; i++){
-        ((uint32_t*)(src+i*stride))[0]=
-        ((uint32_t*)(src+i*stride))[1]= 0x01010101U*128U;
-    }
-}
-
-static void pred8x8_left_dc_c(uint8_t *src, int stride){
-    int i;
-    int dc0, dc2;
-
-    dc0=dc2=0;
-    for(i=0;i<4; i++){
-        dc0+= src[-1+i*stride];
-        dc2+= src[-1+(i+4)*stride];
-    }
-    dc0= 0x01010101*((dc0 + 2)>>2);
-    dc2= 0x01010101*((dc2 + 2)>>2);
-
-    for(i=0; i<4; i++){
-        ((uint32_t*)(src+i*stride))[0]=
-        ((uint32_t*)(src+i*stride))[1]= dc0;
-    }
-    for(i=4; i<8; i++){
-        ((uint32_t*)(src+i*stride))[0]=
-        ((uint32_t*)(src+i*stride))[1]= dc2;
-    }
-}
-
-static void pred8x8_top_dc_c(uint8_t *src, int stride){
-    int i;
-    int dc0, dc1;
-
-    dc0=dc1=0;
-    for(i=0;i<4; i++){
-        dc0+= src[i-stride];
-        dc1+= src[4+i-stride];
-    }
-    dc0= 0x01010101*((dc0 + 2)>>2);
-    dc1= 0x01010101*((dc1 + 2)>>2);
-
-    for(i=0; i<4; i++){
-        ((uint32_t*)(src+i*stride))[0]= dc0;
-        ((uint32_t*)(src+i*stride))[1]= dc1;
-    }
-    for(i=4; i<8; i++){
-        ((uint32_t*)(src+i*stride))[0]= dc0;
-        ((uint32_t*)(src+i*stride))[1]= dc1;
-    }
-}
-
-
-static void pred8x8_dc_c(uint8_t *src, int stride){
-    int i;
-    int dc0, dc1, dc2, dc3;
-
-    dc0=dc1=dc2=0;
-    for(i=0;i<4; i++){
-        dc0+= src[-1+i*stride] + src[i-stride];
-        dc1+= src[4+i-stride];
-        dc2+= src[-1+(i+4)*stride];
-    }
-    dc3= 0x01010101*((dc1 + dc2 + 4)>>3);
-    dc0= 0x01010101*((dc0 + 4)>>3);
-    dc1= 0x01010101*((dc1 + 2)>>2);
-    dc2= 0x01010101*((dc2 + 2)>>2);
-
-    for(i=0; i<4; i++){
-        ((uint32_t*)(src+i*stride))[0]= dc0;
-        ((uint32_t*)(src+i*stride))[1]= dc1;
-    }
-    for(i=4; i<8; i++){
-        ((uint32_t*)(src+i*stride))[0]= dc2;
-        ((uint32_t*)(src+i*stride))[1]= dc3;
-    }
-}
-
-static void pred8x8_plane_c(uint8_t *src, int stride){
-  int j, k;
-  int a;
-  uint8_t *cm = cropTbl + MAX_NEG_CROP;
-  const uint8_t * const src0 = src+3-stride;
-  const uint8_t *src1 = src+4*stride-1;
-  const uint8_t *src2 = src1-2*stride;      // == src+2*stride-1;
-  int H = src0[1] - src0[-1];
-  int V = src1[0] - src2[ 0];
-  for(k=2; k<=4; ++k) {
-    src1 += stride; src2 -= stride;
-    H += k*(src0[k] - src0[-k]);
-    V += k*(src1[0] - src2[ 0]);
-  }
-  H = ( 17*H+16 ) >> 5;
-  V = ( 17*V+16 ) >> 5;
-
-  a = 16*(src1[0] + src2[8]+1) - 3*(V+H);
-  for(j=8; j>0; --j) {
-    int b = a;
-    a += V;
-    src[0] = cm[ (b    ) >> 5 ];
-    src[1] = cm[ (b+  H) >> 5 ];
-    src[2] = cm[ (b+2*H) >> 5 ];
-    src[3] = cm[ (b+3*H) >> 5 ];
-    src[4] = cm[ (b+4*H) >> 5 ];
-    src[5] = cm[ (b+5*H) >> 5 ];
-    src[6] = cm[ (b+6*H) >> 5 ];
-    src[7] = cm[ (b+7*H) >> 5 ];
-    src += stride;
-  }
-}
-
-#define SRC(x,y) src[(x)+(y)*stride]
-#define PL(y) \
-    const int l##y = (SRC(-1,y-1) + 2*SRC(-1,y) + SRC(-1,y+1) + 2) >> 2;
-#define PREDICT_8x8_LOAD_LEFT \
-    const int l0 = ((has_topleft ? SRC(-1,-1) : SRC(-1,0)) \
-                     + 2*SRC(-1,0) + SRC(-1,1) + 2) >> 2; \
-    PL(1) PL(2) PL(3) PL(4) PL(5) PL(6) \
-    const int l7 attribute_unused = (SRC(-1,6) + 3*SRC(-1,7) + 2) >> 2
-
-#define PT(x) \
-    const int t##x = (SRC(x-1,-1) + 2*SRC(x,-1) + SRC(x+1,-1) + 2) >> 2;
-#define PREDICT_8x8_LOAD_TOP \
-    const int t0 = ((has_topleft ? SRC(-1,-1) : SRC(0,-1)) \
-                     + 2*SRC(0,-1) + SRC(1,-1) + 2) >> 2; \
-    PT(1) PT(2) PT(3) PT(4) PT(5) PT(6) \
-    const int t7 attribute_unused = ((has_topright ? SRC(8,-1) : SRC(7,-1)) \
-                     + 2*SRC(7,-1) + SRC(6,-1) + 2) >> 2
-
-#define PTR(x) \
-    t##x = (SRC(x-1,-1) + 2*SRC(x,-1) + SRC(x+1,-1) + 2) >> 2;
-#define PREDICT_8x8_LOAD_TOPRIGHT \
-    int t8, t9, t10, t11, t12, t13, t14, t15; \
-    if(has_topright) { \
-        PTR(8) PTR(9) PTR(10) PTR(11) PTR(12) PTR(13) PTR(14) \
-        t15 = (SRC(14,-1) + 3*SRC(15,-1) + 2) >> 2; \
-    } else t8=t9=t10=t11=t12=t13=t14=t15= SRC(7,-1);
-
-#define PREDICT_8x8_LOAD_TOPLEFT \
-    const int lt = (SRC(-1,0) + 2*SRC(-1,-1) + SRC(0,-1) + 2) >> 2
-
-#define PREDICT_8x8_DC(v) \
-    int y; \
-    for( y = 0; y < 8; y++ ) { \
-        ((uint32_t*)src)[0] = \
-        ((uint32_t*)src)[1] = v; \
-        src += stride; \
-    }
-
-static void pred8x8l_128_dc_c(uint8_t *src, int has_topleft, int has_topright, int stride)
-{
-    PREDICT_8x8_DC(0x80808080);
-}
-static void pred8x8l_left_dc_c(uint8_t *src, int has_topleft, int has_topright, int stride)
-{
-    PREDICT_8x8_LOAD_LEFT;
-    const uint32_t dc = ((l0+l1+l2+l3+l4+l5+l6+l7+4) >> 3) * 0x01010101;
-    PREDICT_8x8_DC(dc);
-}
-static void pred8x8l_top_dc_c(uint8_t *src, int has_topleft, int has_topright, int stride)
-{
-    PREDICT_8x8_LOAD_TOP;
-    const uint32_t dc = ((t0+t1+t2+t3+t4+t5+t6+t7+4) >> 3) * 0x01010101;
-    PREDICT_8x8_DC(dc);
-}
-static void pred8x8l_dc_c(uint8_t *src, int has_topleft, int has_topright, int stride)
-{
-    PREDICT_8x8_LOAD_LEFT;
-    PREDICT_8x8_LOAD_TOP;
-    const uint32_t dc = ((l0+l1+l2+l3+l4+l5+l6+l7
-                         +t0+t1+t2+t3+t4+t5+t6+t7+8) >> 4) * 0x01010101;
-    PREDICT_8x8_DC(dc);
-}
-static void pred8x8l_horizontal_c(uint8_t *src, int has_topleft, int has_topright, int stride)
-{
-    PREDICT_8x8_LOAD_LEFT;
-#define ROW(y) ((uint32_t*)(src+y*stride))[0] =\
-               ((uint32_t*)(src+y*stride))[1] = 0x01010101 * l##y
-    ROW(0); ROW(1); ROW(2); ROW(3); ROW(4); ROW(5); ROW(6); ROW(7);
-#undef ROW
-}
-static void pred8x8l_vertical_c(uint8_t *src, int has_topleft, int has_topright, int stride)
-{
-    int y;
-    PREDICT_8x8_LOAD_TOP;
-    src[0] = t0;
-    src[1] = t1;
-    src[2] = t2;
-    src[3] = t3;
-    src[4] = t4;
-    src[5] = t5;
-    src[6] = t6;
-    src[7] = t7;
-    for( y = 1; y < 8; y++ )
-        *(uint64_t*)(src+y*stride) = *(uint64_t*)src;
-}
-static void pred8x8l_down_left_c(uint8_t *src, int has_topleft, int has_topright, int stride)
-{
-    PREDICT_8x8_LOAD_TOP;
-    PREDICT_8x8_LOAD_TOPRIGHT;
-    SRC(0,0)= (t0 + 2*t1 + t2 + 2) >> 2;
-    SRC(0,1)=SRC(1,0)= (t1 + 2*t2 + t3 + 2) >> 2;
-    SRC(0,2)=SRC(1,1)=SRC(2,0)= (t2 + 2*t3 + t4 + 2) >> 2;
-    SRC(0,3)=SRC(1,2)=SRC(2,1)=SRC(3,0)= (t3 + 2*t4 + t5 + 2) >> 2;
-    SRC(0,4)=SRC(1,3)=SRC(2,2)=SRC(3,1)=SRC(4,0)= (t4 + 2*t5 + t6 + 2) >> 2;
-    SRC(0,5)=SRC(1,4)=SRC(2,3)=SRC(3,2)=SRC(4,1)=SRC(5,0)= (t5 + 2*t6 + t7 + 2) >> 2;
-    SRC(0,6)=SRC(1,5)=SRC(2,4)=SRC(3,3)=SRC(4,2)=SRC(5,1)=SRC(6,0)= (t6 + 2*t7 + t8 + 2) >> 2;
-    SRC(0,7)=SRC(1,6)=SRC(2,5)=SRC(3,4)=SRC(4,3)=SRC(5,2)=SRC(6,1)=SRC(7,0)= (t7 + 2*t8 + t9 + 2) >> 2;
-    SRC(1,7)=SRC(2,6)=SRC(3,5)=SRC(4,4)=SRC(5,3)=SRC(6,2)=SRC(7,1)= (t8 + 2*t9 + t10 + 2) >> 2;
-    SRC(2,7)=SRC(3,6)=SRC(4,5)=SRC(5,4)=SRC(6,3)=SRC(7,2)= (t9 + 2*t10 + t11 + 2) >> 2;
-    SRC(3,7)=SRC(4,6)=SRC(5,5)=SRC(6,4)=SRC(7,3)= (t10 + 2*t11 + t12 + 2) >> 2;
-    SRC(4,7)=SRC(5,6)=SRC(6,5)=SRC(7,4)= (t11 + 2*t12 + t13 + 2) >> 2;
-    SRC(5,7)=SRC(6,6)=SRC(7,5)= (t12 + 2*t13 + t14 + 2) >> 2;
-    SRC(6,7)=SRC(7,6)= (t13 + 2*t14 + t15 + 2) >> 2;
-    SRC(7,7)= (t14 + 3*t15 + 2) >> 2;
-}
-static void pred8x8l_down_right_c(uint8_t *src, int has_topleft, int has_topright, int stride)
-{
-    PREDICT_8x8_LOAD_TOP;
-    PREDICT_8x8_LOAD_LEFT;
-    PREDICT_8x8_LOAD_TOPLEFT;
-    SRC(0,7)= (l7 + 2*l6 + l5 + 2) >> 2;
-    SRC(0,6)=SRC(1,7)= (l6 + 2*l5 + l4 + 2) >> 2;
-    SRC(0,5)=SRC(1,6)=SRC(2,7)= (l5 + 2*l4 + l3 + 2) >> 2;
-    SRC(0,4)=SRC(1,5)=SRC(2,6)=SRC(3,7)= (l4 + 2*l3 + l2 + 2) >> 2;
-    SRC(0,3)=SRC(1,4)=SRC(2,5)=SRC(3,6)=SRC(4,7)= (l3 + 2*l2 + l1 + 2) >> 2;
-    SRC(0,2)=SRC(1,3)=SRC(2,4)=SRC(3,5)=SRC(4,6)=SRC(5,7)= (l2 + 2*l1 + l0 + 2) >> 2;
-    SRC(0,1)=SRC(1,2)=SRC(2,3)=SRC(3,4)=SRC(4,5)=SRC(5,6)=SRC(6,7)= (l1 + 2*l0 + lt + 2) >> 2;
-    SRC(0,0)=SRC(1,1)=SRC(2,2)=SRC(3,3)=SRC(4,4)=SRC(5,5)=SRC(6,6)=SRC(7,7)= (l0 + 2*lt + t0 + 2) >> 2;
-    SRC(1,0)=SRC(2,1)=SRC(3,2)=SRC(4,3)=SRC(5,4)=SRC(6,5)=SRC(7,6)= (lt + 2*t0 + t1 + 2) >> 2;
-    SRC(2,0)=SRC(3,1)=SRC(4,2)=SRC(5,3)=SRC(6,4)=SRC(7,5)= (t0 + 2*t1 + t2 + 2) >> 2;
-    SRC(3,0)=SRC(4,1)=SRC(5,2)=SRC(6,3)=SRC(7,4)= (t1 + 2*t2 + t3 + 2) >> 2;
-    SRC(4,0)=SRC(5,1)=SRC(6,2)=SRC(7,3)= (t2 + 2*t3 + t4 + 2) >> 2;
-    SRC(5,0)=SRC(6,1)=SRC(7,2)= (t3 + 2*t4 + t5 + 2) >> 2;
-    SRC(6,0)=SRC(7,1)= (t4 + 2*t5 + t6 + 2) >> 2;
-    SRC(7,0)= (t5 + 2*t6 + t7 + 2) >> 2;
-
-}
-static void pred8x8l_vertical_right_c(uint8_t *src, int has_topleft, int has_topright, int stride)
-{
-    PREDICT_8x8_LOAD_TOP;
-    PREDICT_8x8_LOAD_LEFT;
-    PREDICT_8x8_LOAD_TOPLEFT;
-    SRC(0,6)= (l5 + 2*l4 + l3 + 2) >> 2;
-    SRC(0,7)= (l6 + 2*l5 + l4 + 2) >> 2;
-    SRC(0,4)=SRC(1,6)= (l3 + 2*l2 + l1 + 2) >> 2;
-    SRC(0,5)=SRC(1,7)= (l4 + 2*l3 + l2 + 2) >> 2;
-    SRC(0,2)=SRC(1,4)=SRC(2,6)= (l1 + 2*l0 + lt + 2) >> 2;
-    SRC(0,3)=SRC(1,5)=SRC(2,7)= (l2 + 2*l1 + l0 + 2) >> 2;
-    SRC(0,1)=SRC(1,3)=SRC(2,5)=SRC(3,7)= (l0 + 2*lt + t0 + 2) >> 2;
-    SRC(0,0)=SRC(1,2)=SRC(2,4)=SRC(3,6)= (lt + t0 + 1) >> 1;
-    SRC(1,1)=SRC(2,3)=SRC(3,5)=SRC(4,7)= (lt + 2*t0 + t1 + 2) >> 2;
-    SRC(1,0)=SRC(2,2)=SRC(3,4)=SRC(4,6)= (t0 + t1 + 1) >> 1;
-    SRC(2,1)=SRC(3,3)=SRC(4,5)=SRC(5,7)= (t0 + 2*t1 + t2 + 2) >> 2;
-    SRC(2,0)=SRC(3,2)=SRC(4,4)=SRC(5,6)= (t1 + t2 + 1) >> 1;
-    SRC(3,1)=SRC(4,3)=SRC(5,5)=SRC(6,7)= (t1 + 2*t2 + t3 + 2) >> 2;
-    SRC(3,0)=SRC(4,2)=SRC(5,4)=SRC(6,6)= (t2 + t3 + 1) >> 1;
-    SRC(4,1)=SRC(5,3)=SRC(6,5)=SRC(7,7)= (t2 + 2*t3 + t4 + 2) >> 2;
-    SRC(4,0)=SRC(5,2)=SRC(6,4)=SRC(7,6)= (t3 + t4 + 1) >> 1;
-    SRC(5,1)=SRC(6,3)=SRC(7,5)= (t3 + 2*t4 + t5 + 2) >> 2;
-    SRC(5,0)=SRC(6,2)=SRC(7,4)= (t4 + t5 + 1) >> 1;
-    SRC(6,1)=SRC(7,3)= (t4 + 2*t5 + t6 + 2) >> 2;
-    SRC(6,0)=SRC(7,2)= (t5 + t6 + 1) >> 1;
-    SRC(7,1)= (t5 + 2*t6 + t7 + 2) >> 2;
-    SRC(7,0)= (t6 + t7 + 1) >> 1;
-}
-static void pred8x8l_horizontal_down_c(uint8_t *src, int has_topleft, int has_topright, int stride)
-{
-    PREDICT_8x8_LOAD_TOP;
-    PREDICT_8x8_LOAD_LEFT;
-    PREDICT_8x8_LOAD_TOPLEFT;
-    SRC(0,7)= (l6 + l7 + 1) >> 1;
-    SRC(1,7)= (l5 + 2*l6 + l7 + 2) >> 2;
-    SRC(0,6)=SRC(2,7)= (l5 + l6 + 1) >> 1;
-    SRC(1,6)=SRC(3,7)= (l4 + 2*l5 + l6 + 2) >> 2;
-    SRC(0,5)=SRC(2,6)=SRC(4,7)= (l4 + l5 + 1) >> 1;
-    SRC(1,5)=SRC(3,6)=SRC(5,7)= (l3 + 2*l4 + l5 + 2) >> 2;
-    SRC(0,4)=SRC(2,5)=SRC(4,6)=SRC(6,7)= (l3 + l4 + 1) >> 1;
-    SRC(1,4)=SRC(3,5)=SRC(5,6)=SRC(7,7)= (l2 + 2*l3 + l4 + 2) >> 2;
-    SRC(0,3)=SRC(2,4)=SRC(4,5)=SRC(6,6)= (l2 + l3 + 1) >> 1;
-    SRC(1,3)=SRC(3,4)=SRC(5,5)=SRC(7,6)= (l1 + 2*l2 + l3 + 2) >> 2;
-    SRC(0,2)=SRC(2,3)=SRC(4,4)=SRC(6,5)= (l1 + l2 + 1) >> 1;
-    SRC(1,2)=SRC(3,3)=SRC(5,4)=SRC(7,5)= (l0 + 2*l1 + l2 + 2) >> 2;
-    SRC(0,1)=SRC(2,2)=SRC(4,3)=SRC(6,4)= (l0 + l1 + 1) >> 1;
-    SRC(1,1)=SRC(3,2)=SRC(5,3)=SRC(7,4)= (lt + 2*l0 + l1 + 2) >> 2;
-    SRC(0,0)=SRC(2,1)=SRC(4,2)=SRC(6,3)= (lt + l0 + 1) >> 1;
-    SRC(1,0)=SRC(3,1)=SRC(5,2)=SRC(7,3)= (l0 + 2*lt + t0 + 2) >> 2;
-    SRC(2,0)=SRC(4,1)=SRC(6,2)= (t1 + 2*t0 + lt + 2) >> 2;
-    SRC(3,0)=SRC(5,1)=SRC(7,2)= (t2 + 2*t1 + t0 + 2) >> 2;
-    SRC(4,0)=SRC(6,1)= (t3 + 2*t2 + t1 + 2) >> 2;
-    SRC(5,0)=SRC(7,1)= (t4 + 2*t3 + t2 + 2) >> 2;
-    SRC(6,0)= (t5 + 2*t4 + t3 + 2) >> 2;
-    SRC(7,0)= (t6 + 2*t5 + t4 + 2) >> 2;
-}
-static void pred8x8l_vertical_left_c(uint8_t *src, int has_topleft, int has_topright, int stride)
-{
-    PREDICT_8x8_LOAD_TOP;
-    PREDICT_8x8_LOAD_TOPRIGHT;
-    SRC(0,0)= (t0 + t1 + 1) >> 1;
-    SRC(0,1)= (t0 + 2*t1 + t2 + 2) >> 2;
-    SRC(0,2)=SRC(1,0)= (t1 + t2 + 1) >> 1;
-    SRC(0,3)=SRC(1,1)= (t1 + 2*t2 + t3 + 2) >> 2;
-    SRC(0,4)=SRC(1,2)=SRC(2,0)= (t2 + t3 + 1) >> 1;
-    SRC(0,5)=SRC(1,3)=SRC(2,1)= (t2 + 2*t3 + t4 + 2) >> 2;
-    SRC(0,6)=SRC(1,4)=SRC(2,2)=SRC(3,0)= (t3 + t4 + 1) >> 1;
-    SRC(0,7)=SRC(1,5)=SRC(2,3)=SRC(3,1)= (t3 + 2*t4 + t5 + 2) >> 2;
-    SRC(1,6)=SRC(2,4)=SRC(3,2)=SRC(4,0)= (t4 + t5 + 1) >> 1;
-    SRC(1,7)=SRC(2,5)=SRC(3,3)=SRC(4,1)= (t4 + 2*t5 + t6 + 2) >> 2;
-    SRC(2,6)=SRC(3,4)=SRC(4,2)=SRC(5,0)= (t5 + t6 + 1) >> 1;
-    SRC(2,7)=SRC(3,5)=SRC(4,3)=SRC(5,1)= (t5 + 2*t6 + t7 + 2) >> 2;
-    SRC(3,6)=SRC(4,4)=SRC(5,2)=SRC(6,0)= (t6 + t7 + 1) >> 1;
-    SRC(3,7)=SRC(4,5)=SRC(5,3)=SRC(6,1)= (t6 + 2*t7 + t8 + 2) >> 2;
-    SRC(4,6)=SRC(5,4)=SRC(6,2)=SRC(7,0)= (t7 + t8 + 1) >> 1;
-    SRC(4,7)=SRC(5,5)=SRC(6,3)=SRC(7,1)= (t7 + 2*t8 + t9 + 2) >> 2;
-    SRC(5,6)=SRC(6,4)=SRC(7,2)= (t8 + t9 + 1) >> 1;
-    SRC(5,7)=SRC(6,5)=SRC(7,3)= (t8 + 2*t9 + t10 + 2) >> 2;
-    SRC(6,6)=SRC(7,4)= (t9 + t10 + 1) >> 1;
-    SRC(6,7)=SRC(7,5)= (t9 + 2*t10 + t11 + 2) >> 2;
-    SRC(7,6)= (t10 + t11 + 1) >> 1;
-    SRC(7,7)= (t10 + 2*t11 + t12 + 2) >> 2;
-}
-static void pred8x8l_horizontal_up_c(uint8_t *src, int has_topleft, int has_topright, int stride)
-{
-    PREDICT_8x8_LOAD_LEFT;
-    SRC(0,0)= (l0 + l1 + 1) >> 1;
-    SRC(1,0)= (l0 + 2*l1 + l2 + 2) >> 2;
-    SRC(0,1)=SRC(2,0)= (l1 + l2 + 1) >> 1;
-    SRC(1,1)=SRC(3,0)= (l1 + 2*l2 + l3 + 2) >> 2;
-    SRC(0,2)=SRC(2,1)=SRC(4,0)= (l2 + l3 + 1) >> 1;
-    SRC(1,2)=SRC(3,1)=SRC(5,0)= (l2 + 2*l3 + l4 + 2) >> 2;
-    SRC(0,3)=SRC(2,2)=SRC(4,1)=SRC(6,0)= (l3 + l4 + 1) >> 1;
-    SRC(1,3)=SRC(3,2)=SRC(5,1)=SRC(7,0)= (l3 + 2*l4 + l5 + 2) >> 2;
-    SRC(0,4)=SRC(2,3)=SRC(4,2)=SRC(6,1)= (l4 + l5 + 1) >> 1;
-    SRC(1,4)=SRC(3,3)=SRC(5,2)=SRC(7,1)= (l4 + 2*l5 + l6 + 2) >> 2;
-    SRC(0,5)=SRC(2,4)=SRC(4,3)=SRC(6,2)= (l5 + l6 + 1) >> 1;
-    SRC(1,5)=SRC(3,4)=SRC(5,3)=SRC(7,2)= (l5 + 2*l6 + l7 + 2) >> 2;
-    SRC(0,6)=SRC(2,5)=SRC(4,4)=SRC(6,3)= (l6 + l7 + 1) >> 1;
-    SRC(1,6)=SRC(3,5)=SRC(5,4)=SRC(7,3)= (l6 + 3*l7 + 2) >> 2;
-    SRC(0,7)=SRC(1,7)=SRC(2,6)=SRC(2,7)=SRC(3,6)=
-    SRC(3,7)=SRC(4,5)=SRC(4,6)=SRC(4,7)=SRC(5,5)=
-    SRC(5,6)=SRC(5,7)=SRC(6,4)=SRC(6,5)=SRC(6,6)=
-    SRC(6,7)=SRC(7,4)=SRC(7,5)=SRC(7,6)=SRC(7,7)= l7;
-}
-#undef PREDICT_8x8_LOAD_LEFT
-#undef PREDICT_8x8_LOAD_TOP
-#undef PREDICT_8x8_LOAD_TOPLEFT
-#undef PREDICT_8x8_LOAD_TOPRIGHT
-#undef PREDICT_8x8_DC
-#undef PTR
-#undef PT
-#undef PL
-#undef SRC
-
-static inline void mc_dir_part(H264Context *h, Picture *pic, int n, int square, int chroma_height, int delta, int list,
-                           uint8_t *dest_y, uint8_t *dest_cb, uint8_t *dest_cr,
-                           int src_x_offset, int src_y_offset,
-                           qpel_mc_func *qpix_op, h264_chroma_mc_func chroma_op){
-    MpegEncContext * const s = &h->s;
-    const int mx= h->mv_cache[list][ scan8[n] ][0] + src_x_offset*8;
-    int my=       h->mv_cache[list][ scan8[n] ][1] + src_y_offset*8;
-    const int luma_xy= (mx&3) + ((my&3)<<2);
-    uint8_t * src_y = pic->data[0] + (mx>>2) + (my>>2)*h->mb_linesize;
-    uint8_t * src_cb, * src_cr;
-    int extra_width= h->emu_edge_width;
-    int extra_height= h->emu_edge_height;
-    int emu=0;
-    const int full_mx= mx>>2;
-    const int full_my= my>>2;
-    const int pic_width  = 16*s->mb_width;
-    const int pic_height = 16*s->mb_height >> MB_MBAFF;
-
-    if(!pic->data[0])
-        return;
-
-    if(mx&7) extra_width -= 3;
-    if(my&7) extra_height -= 3;
-
-    if(   full_mx < 0-extra_width
-       || full_my < 0-extra_height
-       || full_mx + 16/*FIXME*/ > pic_width + extra_width
-       || full_my + 16/*FIXME*/ > pic_height + extra_height){
-        ff_emulated_edge_mc(s->edge_emu_buffer, src_y - 2 - 2*h->mb_linesize, h->mb_linesize, 16+5, 16+5/*FIXME*/, full_mx-2, full_my-2, pic_width, pic_height);
-            src_y= s->edge_emu_buffer + 2 + 2*h->mb_linesize;
-        emu=1;
-    }
-
-    qpix_op[luma_xy](dest_y, src_y, h->mb_linesize); //FIXME try variable height perhaps?
-    if(!square){
-        qpix_op[luma_xy](dest_y + delta, src_y + delta, h->mb_linesize);
-    }
-
-    if(s->flags&CODEC_FLAG_GRAY) return;
-
-    if(MB_MBAFF){
-        // chroma offset when predicting from a field of opposite parity
-        my += 2 * ((s->mb_y & 1) - (h->ref_cache[list][scan8[n]] & 1));
-        emu |= (my>>3) < 0 || (my>>3) + 8 >= (pic_height>>1);
-    }
-    src_cb= pic->data[1] + (mx>>3) + (my>>3)*h->mb_uvlinesize;
-    src_cr= pic->data[2] + (mx>>3) + (my>>3)*h->mb_uvlinesize;
-
-    if(emu){
-        ff_emulated_edge_mc(s->edge_emu_buffer, src_cb, h->mb_uvlinesize, 9, 9/*FIXME*/, (mx>>3), (my>>3), pic_width>>1, pic_height>>1);
-            src_cb= s->edge_emu_buffer;
-    }
-    chroma_op(dest_cb, src_cb, h->mb_uvlinesize, chroma_height, mx&7, my&7);
-
-    if(emu){
-        ff_emulated_edge_mc(s->edge_emu_buffer, src_cr, h->mb_uvlinesize, 9, 9/*FIXME*/, (mx>>3), (my>>3), pic_width>>1, pic_height>>1);
-            src_cr= s->edge_emu_buffer;
-    }
-    chroma_op(dest_cr, src_cr, h->mb_uvlinesize, chroma_height, mx&7, my&7);
-}
-
-static inline void mc_part_std(H264Context *h, int n, int square, int chroma_height, int delta,
-                           uint8_t *dest_y, uint8_t *dest_cb, uint8_t *dest_cr,
-                           int x_offset, int y_offset,
-                           qpel_mc_func *qpix_put, h264_chroma_mc_func chroma_put,
-                           qpel_mc_func *qpix_avg, h264_chroma_mc_func chroma_avg,
-                           int list0, int list1){
-    MpegEncContext * const s = &h->s;
-    qpel_mc_func *qpix_op=  qpix_put;
-    h264_chroma_mc_func chroma_op= chroma_put;
-
-    dest_y  += 2*x_offset + 2*y_offset*h->  mb_linesize;
-    dest_cb +=   x_offset +   y_offset*h->mb_uvlinesize;
-    dest_cr +=   x_offset +   y_offset*h->mb_uvlinesize;
-    x_offset += 8*s->mb_x;
-    y_offset += 8*(s->mb_y >> MB_MBAFF);
-
-    if(list0){
-        Picture *ref= &h->ref_list[0][ h->ref_cache[0][ scan8[n] ] ];
-        mc_dir_part(h, ref, n, square, chroma_height, delta, 0,
-                           dest_y, dest_cb, dest_cr, x_offset, y_offset,
-                           qpix_op, chroma_op);
-
-        qpix_op=  qpix_avg;
-        chroma_op= chroma_avg;
-    }
-
-    if(list1){
-        Picture *ref= &h->ref_list[1][ h->ref_cache[1][ scan8[n] ] ];
-        mc_dir_part(h, ref, n, square, chroma_height, delta, 1,
-                           dest_y, dest_cb, dest_cr, x_offset, y_offset,
-                           qpix_op, chroma_op);
-    }
-}
-
-static inline void mc_part_weighted(H264Context *h, int n, int square, int chroma_height, int delta,
-                           uint8_t *dest_y, uint8_t *dest_cb, uint8_t *dest_cr,
-                           int x_offset, int y_offset,
-                           qpel_mc_func *qpix_put, h264_chroma_mc_func chroma_put,
-                           h264_weight_func luma_weight_op, h264_weight_func chroma_weight_op,
-                           h264_biweight_func luma_weight_avg, h264_biweight_func chroma_weight_avg,
-                           int list0, int list1){
-    MpegEncContext * const s = &h->s;
-
-    dest_y  += 2*x_offset + 2*y_offset*h->  mb_linesize;
-    dest_cb +=   x_offset +   y_offset*h->mb_uvlinesize;
-    dest_cr +=   x_offset +   y_offset*h->mb_uvlinesize;
-    x_offset += 8*s->mb_x;
-    y_offset += 8*(s->mb_y >> MB_MBAFF);
-
-    if(list0 && list1){
-        /* don't optimize for luma-only case, since B-frames usually
-         * use implicit weights => chroma too. */
-        uint8_t *tmp_cb = s->obmc_scratchpad;
-        uint8_t *tmp_cr = s->obmc_scratchpad + 8;
-        uint8_t *tmp_y  = s->obmc_scratchpad + 8*h->mb_uvlinesize;
-        int refn0 = h->ref_cache[0][ scan8[n] ];
-        int refn1 = h->ref_cache[1][ scan8[n] ];
-
-        mc_dir_part(h, &h->ref_list[0][refn0], n, square, chroma_height, delta, 0,
-                    dest_y, dest_cb, dest_cr,
-                    x_offset, y_offset, qpix_put, chroma_put);
-        mc_dir_part(h, &h->ref_list[1][refn1], n, square, chroma_height, delta, 1,
-                    tmp_y, tmp_cb, tmp_cr,
-                    x_offset, y_offset, qpix_put, chroma_put);
-
-        if(h->use_weight == 2){
-            int weight0 = h->implicit_weight[refn0][refn1];
-            int weight1 = 64 - weight0;
-            luma_weight_avg(  dest_y,  tmp_y,  h->  mb_linesize, 5, weight0, weight1, 0);
-            chroma_weight_avg(dest_cb, tmp_cb, h->mb_uvlinesize, 5, weight0, weight1, 0);
-            chroma_weight_avg(dest_cr, tmp_cr, h->mb_uvlinesize, 5, weight0, weight1, 0);
-        }else{
-            luma_weight_avg(dest_y, tmp_y, h->mb_linesize, h->luma_log2_weight_denom,
-                            h->luma_weight[0][refn0], h->luma_weight[1][refn1],
-                            h->luma_offset[0][refn0] + h->luma_offset[1][refn1]);
-            chroma_weight_avg(dest_cb, tmp_cb, h->mb_uvlinesize, h->chroma_log2_weight_denom,
-                            h->chroma_weight[0][refn0][0], h->chroma_weight[1][refn1][0],
-                            h->chroma_offset[0][refn0][0] + h->chroma_offset[1][refn1][0]);
-            chroma_weight_avg(dest_cr, tmp_cr, h->mb_uvlinesize, h->chroma_log2_weight_denom,
-                            h->chroma_weight[0][refn0][1], h->chroma_weight[1][refn1][1],
-                            h->chroma_offset[0][refn0][1] + h->chroma_offset[1][refn1][1]);
-        }
-    }else{
-        int list = list1 ? 1 : 0;
-        int refn = h->ref_cache[list][ scan8[n] ];
-        Picture *ref= &h->ref_list[list][refn];
-        mc_dir_part(h, ref, n, square, chroma_height, delta, list,
-                    dest_y, dest_cb, dest_cr, x_offset, y_offset,
-                    qpix_put, chroma_put);
-
-        luma_weight_op(dest_y, h->mb_linesize, h->luma_log2_weight_denom,
-                       h->luma_weight[list][refn], h->luma_offset[list][refn]);
-        if(h->use_weight_chroma){
-            chroma_weight_op(dest_cb, h->mb_uvlinesize, h->chroma_log2_weight_denom,
-                             h->chroma_weight[list][refn][0], h->chroma_offset[list][refn][0]);
-            chroma_weight_op(dest_cr, h->mb_uvlinesize, h->chroma_log2_weight_denom,
-                             h->chroma_weight[list][refn][1], h->chroma_offset[list][refn][1]);
-        }
-    }
-}
-
-static inline void mc_part(H264Context *h, int n, int square, int chroma_height, int delta,
-                           uint8_t *dest_y, uint8_t *dest_cb, uint8_t *dest_cr,
-                           int x_offset, int y_offset,
-                           qpel_mc_func *qpix_put, h264_chroma_mc_func chroma_put,
-                           qpel_mc_func *qpix_avg, h264_chroma_mc_func chroma_avg,
-                           h264_weight_func *weight_op, h264_biweight_func *weight_avg,
-                           int list0, int list1){
-    if((h->use_weight==2 && list0 && list1
-        && (h->implicit_weight[ h->ref_cache[0][scan8[n]] ][ h->ref_cache[1][scan8[n]] ] != 32))
-       || h->use_weight==1)
-        mc_part_weighted(h, n, square, chroma_height, delta, dest_y, dest_cb, dest_cr,
-                         x_offset, y_offset, qpix_put, chroma_put,
-                         weight_op[0], weight_op[3], weight_avg[0], weight_avg[3], list0, list1);
-    else
-        mc_part_std(h, n, square, chroma_height, delta, dest_y, dest_cb, dest_cr,
-                    x_offset, y_offset, qpix_put, chroma_put, qpix_avg, chroma_avg, list0, list1);
-}
-
-static inline void prefetch_motion(H264Context *h, int list){
-    /* fetch pixels for estimated mv 4 macroblocks ahead
-     * optimized for 64byte cache lines */
-    MpegEncContext * const s = &h->s;
-    const int refn = h->ref_cache[list][scan8[0]];
-    if(refn >= 0){
-        const int mx= (h->mv_cache[list][scan8[0]][0]>>2) + 16*s->mb_x + 8;
-        const int my= (h->mv_cache[list][scan8[0]][1]>>2) + 16*s->mb_y;
-        uint8_t **src= h->ref_list[list][refn].data;
-        int off= mx + (my + (s->mb_x&3)*4)*h->mb_linesize + 64;
-        s->dsp.prefetch(src[0]+off, s->linesize, 4);
-        off= (mx>>1) + ((my>>1) + (s->mb_x&7))*s->uvlinesize + 64;
-        s->dsp.prefetch(src[1]+off, src[2]-src[1], 2);
-    }
-}
-
-static void hl_motion(H264Context *h, uint8_t *dest_y, uint8_t *dest_cb, uint8_t *dest_cr,
-                      qpel_mc_func (*qpix_put)[16], h264_chroma_mc_func (*chroma_put),
-                      qpel_mc_func (*qpix_avg)[16], h264_chroma_mc_func (*chroma_avg),
-                      h264_weight_func *weight_op, h264_biweight_func *weight_avg){
-    MpegEncContext * const s = &h->s;
-    const int mb_xy= s->mb_x + s->mb_y*s->mb_stride;
-    const int mb_type= s->current_picture.mb_type[mb_xy];
-
-    assert(IS_INTER(mb_type));
-
-    prefetch_motion(h, 0);
-
-    if(IS_16X16(mb_type)){
-        mc_part(h, 0, 1, 8, 0, dest_y, dest_cb, dest_cr, 0, 0,
-                qpix_put[0], chroma_put[0], qpix_avg[0], chroma_avg[0],
-                &weight_op[0], &weight_avg[0],
-                IS_DIR(mb_type, 0, 0), IS_DIR(mb_type, 0, 1));
-    }else if(IS_16X8(mb_type)){
-        mc_part(h, 0, 0, 4, 8, dest_y, dest_cb, dest_cr, 0, 0,
-                qpix_put[1], chroma_put[0], qpix_avg[1], chroma_avg[0],
-                &weight_op[1], &weight_avg[1],
-                IS_DIR(mb_type, 0, 0), IS_DIR(mb_type, 0, 1));
-        mc_part(h, 8, 0, 4, 8, dest_y, dest_cb, dest_cr, 0, 4,
-                qpix_put[1], chroma_put[0], qpix_avg[1], chroma_avg[0],
-                &weight_op[1], &weight_avg[1],
-                IS_DIR(mb_type, 1, 0), IS_DIR(mb_type, 1, 1));
-    }else if(IS_8X16(mb_type)){
-        mc_part(h, 0, 0, 8, 8*h->mb_linesize, dest_y, dest_cb, dest_cr, 0, 0,
-                qpix_put[1], chroma_put[1], qpix_avg[1], chroma_avg[1],
-                &weight_op[2], &weight_avg[2],
-                IS_DIR(mb_type, 0, 0), IS_DIR(mb_type, 0, 1));
-        mc_part(h, 4, 0, 8, 8*h->mb_linesize, dest_y, dest_cb, dest_cr, 4, 0,
-                qpix_put[1], chroma_put[1], qpix_avg[1], chroma_avg[1],
-                &weight_op[2], &weight_avg[2],
-                IS_DIR(mb_type, 1, 0), IS_DIR(mb_type, 1, 1));
-    }else{
-        int i;
-
-        assert(IS_8X8(mb_type));
-
-        for(i=0; i<4; i++){
-            const int sub_mb_type= h->sub_mb_type[i];
-            const int n= 4*i;
-            int x_offset= (i&1)<<2;
-            int y_offset= (i&2)<<1;
-
-            if(IS_SUB_8X8(sub_mb_type)){
-                mc_part(h, n, 1, 4, 0, dest_y, dest_cb, dest_cr, x_offset, y_offset,
-                    qpix_put[1], chroma_put[1], qpix_avg[1], chroma_avg[1],
-                    &weight_op[3], &weight_avg[3],
-                    IS_DIR(sub_mb_type, 0, 0), IS_DIR(sub_mb_type, 0, 1));
-            }else if(IS_SUB_8X4(sub_mb_type)){
-                mc_part(h, n  , 0, 2, 4, dest_y, dest_cb, dest_cr, x_offset, y_offset,
-                    qpix_put[2], chroma_put[1], qpix_avg[2], chroma_avg[1],
-                    &weight_op[4], &weight_avg[4],
-                    IS_DIR(sub_mb_type, 0, 0), IS_DIR(sub_mb_type, 0, 1));
-                mc_part(h, n+2, 0, 2, 4, dest_y, dest_cb, dest_cr, x_offset, y_offset+2,
-                    qpix_put[2], chroma_put[1], qpix_avg[2], chroma_avg[1],
-                    &weight_op[4], &weight_avg[4],
-                    IS_DIR(sub_mb_type, 0, 0), IS_DIR(sub_mb_type, 0, 1));
-            }else if(IS_SUB_4X8(sub_mb_type)){
-                mc_part(h, n  , 0, 4, 4*h->mb_linesize, dest_y, dest_cb, dest_cr, x_offset, y_offset,
-                    qpix_put[2], chroma_put[2], qpix_avg[2], chroma_avg[2],
-                    &weight_op[5], &weight_avg[5],
-                    IS_DIR(sub_mb_type, 0, 0), IS_DIR(sub_mb_type, 0, 1));
-                mc_part(h, n+1, 0, 4, 4*h->mb_linesize, dest_y, dest_cb, dest_cr, x_offset+2, y_offset,
-                    qpix_put[2], chroma_put[2], qpix_avg[2], chroma_avg[2],
-                    &weight_op[5], &weight_avg[5],
-                    IS_DIR(sub_mb_type, 0, 0), IS_DIR(sub_mb_type, 0, 1));
-            }else{
-                int j;
-                assert(IS_SUB_4X4(sub_mb_type));
-                for(j=0; j<4; j++){
-                    int sub_x_offset= x_offset + 2*(j&1);
-                    int sub_y_offset= y_offset +   (j&2);
-                    mc_part(h, n+j, 1, 2, 0, dest_y, dest_cb, dest_cr, sub_x_offset, sub_y_offset,
-                        qpix_put[2], chroma_put[2], qpix_avg[2], chroma_avg[2],
-                        &weight_op[6], &weight_avg[6],
-                        IS_DIR(sub_mb_type, 0, 0), IS_DIR(sub_mb_type, 0, 1));
-                }
-            }
-        }
-    }
-
-    prefetch_motion(h, 1);
-}
-
-static void decode_init_vlc(H264Context *h){
-    static int done = 0;
-
-    if (!done) {
-        int i;
-        done = 1;
-
-        init_vlc(&chroma_dc_coeff_token_vlc, CHROMA_DC_COEFF_TOKEN_VLC_BITS, 4*5,
-                 &chroma_dc_coeff_token_len [0], 1, 1,
-                 &chroma_dc_coeff_token_bits[0], 1, 1, 1);
-
-        for(i=0; i<4; i++){
-            init_vlc(&coeff_token_vlc[i], COEFF_TOKEN_VLC_BITS, 4*17,
-                     &coeff_token_len [i][0], 1, 1,
-                     &coeff_token_bits[i][0], 1, 1, 1);
-        }
-
-        for(i=0; i<3; i++){
-            init_vlc(&chroma_dc_total_zeros_vlc[i], CHROMA_DC_TOTAL_ZEROS_VLC_BITS, 4,
-                     &chroma_dc_total_zeros_len [i][0], 1, 1,
-                     &chroma_dc_total_zeros_bits[i][0], 1, 1, 1);
-        }
-        for(i=0; i<15; i++){
-            init_vlc(&total_zeros_vlc[i], TOTAL_ZEROS_VLC_BITS, 16,
-                     &total_zeros_len [i][0], 1, 1,
-                     &total_zeros_bits[i][0], 1, 1, 1);
-        }
-
-        for(i=0; i<6; i++){
-            init_vlc(&run_vlc[i], RUN_VLC_BITS, 7,
-                     &run_len [i][0], 1, 1,
-                     &run_bits[i][0], 1, 1, 1);
-        }
-        init_vlc(&run7_vlc, RUN7_VLC_BITS, 16,
-                 &run_len [6][0], 1, 1,
-                 &run_bits[6][0], 1, 1, 1);
-    }
-}
-
-/**
- * Sets the intra prediction function pointers.
- */
-static void init_pred_ptrs(H264Context *h){
-//    MpegEncContext * const s = &h->s;
-
-    h->pred4x4[VERT_PRED           ]= pred4x4_vertical_c;
-    h->pred4x4[HOR_PRED            ]= pred4x4_horizontal_c;
-    h->pred4x4[DC_PRED             ]= pred4x4_dc_c;
-    h->pred4x4[DIAG_DOWN_LEFT_PRED ]= pred4x4_down_left_c;
-    h->pred4x4[DIAG_DOWN_RIGHT_PRED]= pred4x4_down_right_c;
-    h->pred4x4[VERT_RIGHT_PRED     ]= pred4x4_vertical_right_c;
-    h->pred4x4[HOR_DOWN_PRED       ]= pred4x4_horizontal_down_c;
-    h->pred4x4[VERT_LEFT_PRED      ]= pred4x4_vertical_left_c;
-    h->pred4x4[HOR_UP_PRED         ]= pred4x4_horizontal_up_c;
-    h->pred4x4[LEFT_DC_PRED        ]= pred4x4_left_dc_c;
-    h->pred4x4[TOP_DC_PRED         ]= pred4x4_top_dc_c;
-    h->pred4x4[DC_128_PRED         ]= pred4x4_128_dc_c;
-
-    h->pred8x8l[VERT_PRED           ]= pred8x8l_vertical_c;
-    h->pred8x8l[HOR_PRED            ]= pred8x8l_horizontal_c;
-    h->pred8x8l[DC_PRED             ]= pred8x8l_dc_c;
-    h->pred8x8l[DIAG_DOWN_LEFT_PRED ]= pred8x8l_down_left_c;
-    h->pred8x8l[DIAG_DOWN_RIGHT_PRED]= pred8x8l_down_right_c;
-    h->pred8x8l[VERT_RIGHT_PRED     ]= pred8x8l_vertical_right_c;
-    h->pred8x8l[HOR_DOWN_PRED       ]= pred8x8l_horizontal_down_c;
-    h->pred8x8l[VERT_LEFT_PRED      ]= pred8x8l_vertical_left_c;
-    h->pred8x8l[HOR_UP_PRED         ]= pred8x8l_horizontal_up_c;
-    h->pred8x8l[LEFT_DC_PRED        ]= pred8x8l_left_dc_c;
-    h->pred8x8l[TOP_DC_PRED         ]= pred8x8l_top_dc_c;
-    h->pred8x8l[DC_128_PRED         ]= pred8x8l_128_dc_c;
-
-    h->pred8x8[DC_PRED8x8     ]= pred8x8_dc_c;
-    h->pred8x8[VERT_PRED8x8   ]= pred8x8_vertical_c;
-    h->pred8x8[HOR_PRED8x8    ]= pred8x8_horizontal_c;
-    h->pred8x8[PLANE_PRED8x8  ]= pred8x8_plane_c;
-    h->pred8x8[LEFT_DC_PRED8x8]= pred8x8_left_dc_c;
-    h->pred8x8[TOP_DC_PRED8x8 ]= pred8x8_top_dc_c;
-    h->pred8x8[DC_128_PRED8x8 ]= pred8x8_128_dc_c;
-
-    h->pred16x16[DC_PRED8x8     ]= pred16x16_dc_c;
-    h->pred16x16[VERT_PRED8x8   ]= pred16x16_vertical_c;
-    h->pred16x16[HOR_PRED8x8    ]= pred16x16_horizontal_c;
-    h->pred16x16[PLANE_PRED8x8  ]= pred16x16_plane_c;
-    h->pred16x16[LEFT_DC_PRED8x8]= pred16x16_left_dc_c;
-    h->pred16x16[TOP_DC_PRED8x8 ]= pred16x16_top_dc_c;
-    h->pred16x16[DC_128_PRED8x8 ]= pred16x16_128_dc_c;
-}
-
-static void free_tables(H264Context *h){
-    av_freep(&h->intra4x4_pred_mode);
-    av_freep(&h->chroma_pred_mode_table);
-    av_freep(&h->cbp_table);
-    av_freep(&h->mvd_table[0]);
-    av_freep(&h->mvd_table[1]);
-    av_freep(&h->direct_table);
-    av_freep(&h->non_zero_count);
-    av_freep(&h->slice_table_base);
-    av_freep(&h->top_borders[1]);
-    av_freep(&h->top_borders[0]);
-    h->slice_table= NULL;
-
-    av_freep(&h->mb2b_xy);
-    av_freep(&h->mb2b8_xy);
-
-    av_freep(&h->s.obmc_scratchpad);
-}
-
-static void init_dequant8_coeff_table(H264Context *h){
-    int i,q,x;
-    const int transpose = (h->s.dsp.h264_idct8_add != ff_h264_idct8_add_c); //FIXME ugly
-    h->dequant8_coeff[0] = h->dequant8_buffer[0];
-    h->dequant8_coeff[1] = h->dequant8_buffer[1];
-
-    for(i=0; i<2; i++ ){
-        if(i && !memcmp(h->pps.scaling_matrix8[0], h->pps.scaling_matrix8[1], 64*sizeof(uint8_t))){
-            h->dequant8_coeff[1] = h->dequant8_buffer[0];
-            break;
-        }
-
-        for(q=0; q<52; q++){
-            int shift = div6[q];
-            int idx = rem6[q];
-            for(x=0; x<64; x++)
-                h->dequant8_coeff[i][q][transpose ? (x>>3)|((x&7)<<3) : x] =
-                    ((uint32_t)dequant8_coeff_init[idx][ dequant8_coeff_init_scan[((x>>1)&12) | (x&3)] ] *
-                    h->pps.scaling_matrix8[i][x]) << shift;
-        }
-    }
-}
-
-static void init_dequant4_coeff_table(H264Context *h){
-    int i,j,q,x;
-    const int transpose = (h->s.dsp.h264_idct_add != ff_h264_idct_add_c); //FIXME ugly
-    for(i=0; i<6; i++ ){
-        h->dequant4_coeff[i] = h->dequant4_buffer[i];
-        for(j=0; j<i; j++){
-            if(!memcmp(h->pps.scaling_matrix4[j], h->pps.scaling_matrix4[i], 16*sizeof(uint8_t))){
-                h->dequant4_coeff[i] = h->dequant4_buffer[j];
-                break;
-            }
-        }
-        if(j<i)
-            continue;
-
-        for(q=0; q<52; q++){
-            int shift = div6[q] + 2;
-            int idx = rem6[q];
-            for(x=0; x<16; x++)
-                h->dequant4_coeff[i][q][transpose ? (x>>2)|((x<<2)&0xF) : x] =
-                    ((uint32_t)dequant4_coeff_init[idx][(x&1) + ((x>>2)&1)] *
-                    h->pps.scaling_matrix4[i][x]) << shift;
-        }
-    }
-}
-
-static void init_dequant_tables(H264Context *h){
-    int i,x;
-    init_dequant4_coeff_table(h);
-    if(h->pps.transform_8x8_mode)
-        init_dequant8_coeff_table(h);
-    if(h->sps.transform_bypass){
-        for(i=0; i<6; i++)
-            for(x=0; x<16; x++)
-                h->dequant4_coeff[i][0][x] = 1<<6;
-        if(h->pps.transform_8x8_mode)
-            for(i=0; i<2; i++)
-                for(x=0; x<64; x++)
-                    h->dequant8_coeff[i][0][x] = 1<<6;
-    }
-}
-
-
-/**
- * allocates tables.
- * needs width/height
- */
-static int alloc_tables(H264Context *h){
-    MpegEncContext * const s = &h->s;
-    const int big_mb_num= s->mb_stride * (s->mb_height+1);
-    int x,y;
-
-    CHECKED_ALLOCZ(h->intra4x4_pred_mode, big_mb_num * 8  * sizeof(uint8_t))
-
-    CHECKED_ALLOCZ(h->non_zero_count    , big_mb_num * 16 * sizeof(uint8_t))
-    CHECKED_ALLOCZ(h->slice_table_base  , (big_mb_num+s->mb_stride) * sizeof(uint8_t))
-    CHECKED_ALLOCZ(h->top_borders[0]    , s->mb_width * (16+8+8) * sizeof(uint8_t))
-    CHECKED_ALLOCZ(h->top_borders[1]    , s->mb_width * (16+8+8) * sizeof(uint8_t))
-    CHECKED_ALLOCZ(h->cbp_table, big_mb_num * sizeof(uint16_t))
-
-    if( h->pps.cabac ) {
-        CHECKED_ALLOCZ(h->chroma_pred_mode_table, big_mb_num * sizeof(uint8_t))
-        CHECKED_ALLOCZ(h->mvd_table[0], 32*big_mb_num * sizeof(uint16_t));
-        CHECKED_ALLOCZ(h->mvd_table[1], 32*big_mb_num * sizeof(uint16_t));
-        CHECKED_ALLOCZ(h->direct_table, 32*big_mb_num * sizeof(uint8_t));
-    }
-
-    memset(h->slice_table_base, -1, (big_mb_num+s->mb_stride)  * sizeof(uint8_t));
-    h->slice_table= h->slice_table_base + s->mb_stride*2 + 1;
-
-    CHECKED_ALLOCZ(h->mb2b_xy  , big_mb_num * sizeof(uint32_t));
-    CHECKED_ALLOCZ(h->mb2b8_xy , big_mb_num * sizeof(uint32_t));
-    for(y=0; y<s->mb_height; y++){
-        for(x=0; x<s->mb_width; x++){
-            const int mb_xy= x + y*s->mb_stride;
-            const int b_xy = 4*x + 4*y*h->b_stride;
-            const int b8_xy= 2*x + 2*y*h->b8_stride;
-
-            h->mb2b_xy [mb_xy]= b_xy;
-            h->mb2b8_xy[mb_xy]= b8_xy;
-        }
-    }
-
-    s->obmc_scratchpad = NULL;
-
-    if(!h->dequant4_coeff[0])
-        init_dequant_tables(h);
-
-    return 0;
-fail:
-    free_tables(h);
-    return -1;
-}
-
-static void common_init(H264Context *h){
-    MpegEncContext * const s = &h->s;
-
-    s->width = s->avctx->width;
-    s->height = s->avctx->height;
-    s->codec_id= s->avctx->codec->id;
-
-    init_pred_ptrs(h);
-
-    h->dequant_coeff_pps= -1;
-    s->unrestricted_mv=1;
-    s->decode=1; //FIXME
-
-    memset(h->pps.scaling_matrix4, 16, 6*16*sizeof(uint8_t));
-    memset(h->pps.scaling_matrix8, 16, 2*64*sizeof(uint8_t));
-}
-
-static int decode_init(AVCodecContext *avctx){
-    H264Context *h= avctx->priv_data;
-    MpegEncContext * const s = &h->s;
-
-    MPV_decode_defaults(s);
-
-    s->avctx = avctx;
-    common_init(h);
-
-    s->out_format = FMT_H264;
-    s->workaround_bugs= avctx->workaround_bugs;
-
-    // set defaults
-//    s->decode_mb= ff_h263_decode_mb;
-    s->low_delay= 1;
-    avctx->pix_fmt= PIX_FMT_YUV420P;
-
-    decode_init_vlc(h);
-
-    if(avctx->extradata_size > 0 && avctx->extradata &&
-       *(char *)avctx->extradata == 1){
-        h->is_avc = 1;
-        h->got_avcC = 0;
-    } else {
-        h->is_avc = 0;
-    }
-
-    return 0;
-}
-
-static int frame_start(H264Context *h){
-    MpegEncContext * const s = &h->s;
-    int i;
-
-    if(MPV_frame_start(s, s->avctx) < 0)
-        return -1;
-    ff_er_frame_start(s);
-
-    assert(s->linesize && s->uvlinesize);
-
-    for(i=0; i<16; i++){
-        h->block_offset[i]= 4*((scan8[i] - scan8[0])&7) + 4*s->linesize*((scan8[i] - scan8[0])>>3);
-        h->block_offset[24+i]= 4*((scan8[i] - scan8[0])&7) + 8*s->linesize*((scan8[i] - scan8[0])>>3);
-    }
-    for(i=0; i<4; i++){
-        h->block_offset[16+i]=
-        h->block_offset[20+i]= 4*((scan8[i] - scan8[0])&7) + 4*s->uvlinesize*((scan8[i] - scan8[0])>>3);
-        h->block_offset[24+16+i]=
-        h->block_offset[24+20+i]= 4*((scan8[i] - scan8[0])&7) + 8*s->uvlinesize*((scan8[i] - scan8[0])>>3);
-    }
-
-    /* can't be in alloc_tables because linesize isn't known there.
-     * FIXME: redo bipred weight to not require extra buffer? */
-    if(!s->obmc_scratchpad)
-        s->obmc_scratchpad = av_malloc(16*2*s->linesize + 8*2*s->uvlinesize);
-
-    /* some macroblocks will be accessed before they're available */
-    if(FRAME_MBAFF)
-        memset(h->slice_table, -1, (s->mb_height*s->mb_stride-1) * sizeof(uint8_t));
-
-//    s->decode= (s->flags&CODEC_FLAG_PSNR) || !s->encoding || s->current_picture.reference /*|| h->contains_intra*/ || 1;
-    return 0;
-}
-
-static inline void backup_mb_border(H264Context *h, uint8_t *src_y, uint8_t *src_cb, uint8_t *src_cr, int linesize, int uvlinesize){
-    MpegEncContext * const s = &h->s;
-    int i;
-
-    src_y  -=   linesize;
-    src_cb -= uvlinesize;
-    src_cr -= uvlinesize;
-
-    // There are two lines saved, the line above the the top macroblock of a pair,
-    // and the line above the bottom macroblock
-    h->left_border[0]= h->top_borders[0][s->mb_x][15];
-    for(i=1; i<17; i++){
-        h->left_border[i]= src_y[15+i*  linesize];
-    }
-
-    *(uint64_t*)(h->top_borders[0][s->mb_x]+0)= *(uint64_t*)(src_y +  16*linesize);
-    *(uint64_t*)(h->top_borders[0][s->mb_x]+8)= *(uint64_t*)(src_y +8+16*linesize);
-
-    if(!(s->flags&CODEC_FLAG_GRAY)){
-        h->left_border[17  ]= h->top_borders[0][s->mb_x][16+7];
-        h->left_border[17+9]= h->top_borders[0][s->mb_x][24+7];
-        for(i=1; i<9; i++){
-            h->left_border[i+17  ]= src_cb[7+i*uvlinesize];
-            h->left_border[i+17+9]= src_cr[7+i*uvlinesize];
-        }
-        *(uint64_t*)(h->top_borders[0][s->mb_x]+16)= *(uint64_t*)(src_cb+8*uvlinesize);
-        *(uint64_t*)(h->top_borders[0][s->mb_x]+24)= *(uint64_t*)(src_cr+8*uvlinesize);
-    }
-}
-
-static inline void xchg_mb_border(H264Context *h, uint8_t *src_y, uint8_t *src_cb, uint8_t *src_cr, int linesize, int uvlinesize, int xchg){
-    MpegEncContext * const s = &h->s;
-    int temp8, i;
-    uint64_t temp64;
-    int deblock_left = (s->mb_x > 0);
-    int deblock_top  = (s->mb_y > 0);
-
-    src_y  -=   linesize + 1;
-    src_cb -= uvlinesize + 1;
-    src_cr -= uvlinesize + 1;
-
-#define XCHG(a,b,t,xchg)\
-t= a;\
-if(xchg)\
-    a= b;\
-b= t;
-
-    if(deblock_left){
-        for(i = !deblock_top; i<17; i++){
-            XCHG(h->left_border[i     ], src_y [i*  linesize], temp8, xchg);
-        }
-    }
-
-    if(deblock_top){
-        XCHG(*(uint64_t*)(h->top_borders[0][s->mb_x]+0), *(uint64_t*)(src_y +1), temp64, xchg);
-        XCHG(*(uint64_t*)(h->top_borders[0][s->mb_x]+8), *(uint64_t*)(src_y +9), temp64, 1);
-        if(s->mb_x+1 < s->mb_width){
-            XCHG(*(uint64_t*)(h->top_borders[0][s->mb_x+1]), *(uint64_t*)(src_y +17), temp64, 1);
-        }
-    }
-
-    if(!(s->flags&CODEC_FLAG_GRAY)){
-        if(deblock_left){
-            for(i = !deblock_top; i<9; i++){
-                XCHG(h->left_border[i+17  ], src_cb[i*uvlinesize], temp8, xchg);
-                XCHG(h->left_border[i+17+9], src_cr[i*uvlinesize], temp8, xchg);
-            }
-        }
-        if(deblock_top){
-            XCHG(*(uint64_t*)(h->top_borders[0][s->mb_x]+16), *(uint64_t*)(src_cb+1), temp64, 1);
-            XCHG(*(uint64_t*)(h->top_borders[0][s->mb_x]+24), *(uint64_t*)(src_cr+1), temp64, 1);
-        }
-    }
-}
-
-static inline void backup_pair_border(H264Context *h, uint8_t *src_y, uint8_t *src_cb, uint8_t *src_cr, int linesize, int uvlinesize){
-    MpegEncContext * const s = &h->s;
-    int i;
-
-    src_y  -= 2 *   linesize;
-    src_cb -= 2 * uvlinesize;
-    src_cr -= 2 * uvlinesize;
-
-    // There are two lines saved, the line above the the top macroblock of a pair,
-    // and the line above the bottom macroblock
-    h->left_border[0]= h->top_borders[0][s->mb_x][15];
-    h->left_border[1]= h->top_borders[1][s->mb_x][15];
-    for(i=2; i<34; i++){
-        h->left_border[i]= src_y[15+i*  linesize];
-    }
-
-    *(uint64_t*)(h->top_borders[0][s->mb_x]+0)= *(uint64_t*)(src_y +  32*linesize);
-    *(uint64_t*)(h->top_borders[0][s->mb_x]+8)= *(uint64_t*)(src_y +8+32*linesize);
-    *(uint64_t*)(h->top_borders[1][s->mb_x]+0)= *(uint64_t*)(src_y +  33*linesize);
-    *(uint64_t*)(h->top_borders[1][s->mb_x]+8)= *(uint64_t*)(src_y +8+33*linesize);
-
-    if(!(s->flags&CODEC_FLAG_GRAY)){
-        h->left_border[34     ]= h->top_borders[0][s->mb_x][16+7];
-        h->left_border[34+   1]= h->top_borders[1][s->mb_x][16+7];
-        h->left_border[34+18  ]= h->top_borders[0][s->mb_x][24+7];
-        h->left_border[34+18+1]= h->top_borders[1][s->mb_x][24+7];
-        for(i=2; i<18; i++){
-            h->left_border[i+34   ]= src_cb[7+i*uvlinesize];
-            h->left_border[i+34+18]= src_cr[7+i*uvlinesize];
-        }
-        *(uint64_t*)(h->top_borders[0][s->mb_x]+16)= *(uint64_t*)(src_cb+16*uvlinesize);
-        *(uint64_t*)(h->top_borders[0][s->mb_x]+24)= *(uint64_t*)(src_cr+16*uvlinesize);
-        *(uint64_t*)(h->top_borders[1][s->mb_x]+16)= *(uint64_t*)(src_cb+17*uvlinesize);
-        *(uint64_t*)(h->top_borders[1][s->mb_x]+24)= *(uint64_t*)(src_cr+17*uvlinesize);
-    }
-}
-
-static inline void xchg_pair_border(H264Context *h, uint8_t *src_y, uint8_t *src_cb, uint8_t *src_cr, int linesize, int uvlinesize, int xchg){
-    MpegEncContext * const s = &h->s;
-    int temp8, i;
-    uint64_t temp64;
-    int deblock_left = (s->mb_x > 0);
-    int deblock_top  = (s->mb_y > 1);
-
-    tprintf("xchg_pair_border: src_y:%p src_cb:%p src_cr:%p ls:%d uvls:%d\n", src_y, src_cb, src_cr, linesize, uvlinesize);
-
-    src_y  -= 2 *   linesize + 1;
-    src_cb -= 2 * uvlinesize + 1;
-    src_cr -= 2 * uvlinesize + 1;
-
-#define XCHG(a,b,t,xchg)\
-t= a;\
-if(xchg)\
-    a= b;\
-b= t;
-
-    if(deblock_left){
-        for(i = (!deblock_top)<<1; i<34; i++){
-            XCHG(h->left_border[i     ], src_y [i*  linesize], temp8, xchg);
-        }
-    }
-
-    if(deblock_top){
-        XCHG(*(uint64_t*)(h->top_borders[0][s->mb_x]+0), *(uint64_t*)(src_y +1), temp64, xchg);
-        XCHG(*(uint64_t*)(h->top_borders[0][s->mb_x]+8), *(uint64_t*)(src_y +9), temp64, 1);
-        XCHG(*(uint64_t*)(h->top_borders[1][s->mb_x]+0), *(uint64_t*)(src_y +1 +linesize), temp64, xchg);
-        XCHG(*(uint64_t*)(h->top_borders[1][s->mb_x]+8), *(uint64_t*)(src_y +9 +linesize), temp64, 1);
-        if(s->mb_x+1 < s->mb_width){
-            XCHG(*(uint64_t*)(h->top_borders[0][s->mb_x+1]), *(uint64_t*)(src_y +17), temp64, 1);
-            XCHG(*(uint64_t*)(h->top_borders[1][s->mb_x+1]), *(uint64_t*)(src_y +17 +linesize), temp64, 1);
-        }
-    }
-
-    if(!(s->flags&CODEC_FLAG_GRAY)){
-        if(deblock_left){
-            for(i = (!deblock_top) << 1; i<18; i++){
-                XCHG(h->left_border[i+34   ], src_cb[i*uvlinesize], temp8, xchg);
-                XCHG(h->left_border[i+34+18], src_cr[i*uvlinesize], temp8, xchg);
-            }
-        }
-        if(deblock_top){
-            XCHG(*(uint64_t*)(h->top_borders[0][s->mb_x]+16), *(uint64_t*)(src_cb+1), temp64, 1);
-            XCHG(*(uint64_t*)(h->top_borders[0][s->mb_x]+24), *(uint64_t*)(src_cr+1), temp64, 1);
-            XCHG(*(uint64_t*)(h->top_borders[1][s->mb_x]+16), *(uint64_t*)(src_cb+1 +uvlinesize), temp64, 1);
-            XCHG(*(uint64_t*)(h->top_borders[1][s->mb_x]+24), *(uint64_t*)(src_cr+1 +uvlinesize), temp64, 1);
-        }
-    }
-}
-
-static void hl_decode_mb(H264Context *h){
-    MpegEncContext * const s = &h->s;
-    const int mb_x= s->mb_x;
-    const int mb_y= s->mb_y;
-    const int mb_xy= mb_x + mb_y*s->mb_stride;
-    const int mb_type= s->current_picture.mb_type[mb_xy];
-    uint8_t  *dest_y, *dest_cb, *dest_cr;
-    int linesize, uvlinesize /*dct_offset*/;
-    int i;
-    int *block_offset = &h->block_offset[0];
-    const unsigned int bottom = mb_y & 1;
-    const int transform_bypass = (s->qscale == 0 && h->sps.transform_bypass);
-    void (*idct_add)(uint8_t *dst, DCTELEM *block, int stride);
-    void (*idct_dc_add)(uint8_t *dst, DCTELEM *block, int stride);
-
-    if(!s->decode)
-        return;
-
-    dest_y  = s->current_picture.data[0] + (mb_y * 16* s->linesize  ) + mb_x * 16;
-    dest_cb = s->current_picture.data[1] + (mb_y * 8 * s->uvlinesize) + mb_x * 8;
-    dest_cr = s->current_picture.data[2] + (mb_y * 8 * s->uvlinesize) + mb_x * 8;
-
-    s->dsp.prefetch(dest_y + (s->mb_x&3)*4*s->linesize + 64, s->linesize, 4);
-    s->dsp.prefetch(dest_cb + (s->mb_x&7)*s->uvlinesize + 64, dest_cr - dest_cb, 2);
-
-    if (MB_FIELD) {
-        linesize   = h->mb_linesize   = s->linesize * 2;
-        uvlinesize = h->mb_uvlinesize = s->uvlinesize * 2;
-        block_offset = &h->block_offset[24];
-        if(mb_y&1){ //FIXME move out of this func?
-            dest_y -= s->linesize*15;
-            dest_cb-= s->uvlinesize*7;
-            dest_cr-= s->uvlinesize*7;
-        }
-        if(FRAME_MBAFF) {
-            int list;
-            for(list=0; list<2; list++){
-                if(!USES_LIST(mb_type, list))
-                    continue;
-                if(IS_16X16(mb_type)){
-                    int8_t *ref = &h->ref_cache[list][scan8[0]];
-                    fill_rectangle(ref, 4, 4, 8, 16+*ref^(s->mb_y&1), 1);
-                }else{
-                    for(i=0; i<16; i+=4){
-                        //FIXME can refs be smaller than 8x8 when !direct_8x8_inference ?
-                        int ref = h->ref_cache[list][scan8[i]];
-                        if(ref >= 0)
-                            fill_rectangle(&h->ref_cache[list][scan8[i]], 2, 2, 8, 16+ref^(s->mb_y&1), 1);
-                    }
-                }
-            }
-        }
-    } else {
-        linesize   = h->mb_linesize   = s->linesize;
-        uvlinesize = h->mb_uvlinesize = s->uvlinesize;
-//        dct_offset = s->linesize * 16;
-    }
-
-    if(transform_bypass){
-        idct_dc_add =
-        idct_add = IS_8x8DCT(mb_type) ? s->dsp.add_pixels8 : s->dsp.add_pixels4;
-    }else if(IS_8x8DCT(mb_type)){
-        idct_dc_add = s->dsp.h264_idct8_dc_add;
-        idct_add = s->dsp.h264_idct8_add;
-    }else{
-        idct_dc_add = s->dsp.h264_idct_dc_add;
-        idct_add = s->dsp.h264_idct_add;
-    }
-
-    if(FRAME_MBAFF && h->deblocking_filter && IS_INTRA(mb_type)
-       && (!bottom || !IS_INTRA(s->current_picture.mb_type[mb_xy-s->mb_stride]))){
-        int mbt_y = mb_y&~1;
-        uint8_t *top_y  = s->current_picture.data[0] + (mbt_y * 16* s->linesize  ) + mb_x * 16;
-        uint8_t *top_cb = s->current_picture.data[1] + (mbt_y * 8 * s->uvlinesize) + mb_x * 8;
-        uint8_t *top_cr = s->current_picture.data[2] + (mbt_y * 8 * s->uvlinesize) + mb_x * 8;
-        xchg_pair_border(h, top_y, top_cb, top_cr, s->linesize, s->uvlinesize, 1);
-    }
-
-    if (IS_INTRA_PCM(mb_type)) {
-        unsigned int x, y;
-
-        // The pixels are stored in h->mb array in the same order as levels,
-        // copy them in output in the correct order.
-        for(i=0; i<16; i++) {
-            for (y=0; y<4; y++) {
-                for (x=0; x<4; x++) {
-                    *(dest_y + block_offset[i] + y*linesize + x) = h->mb[i*16+y*4+x];
-                }
-            }
-        }
-        for(i=16; i<16+4; i++) {
-            for (y=0; y<4; y++) {
-                for (x=0; x<4; x++) {
-                    *(dest_cb + block_offset[i] + y*uvlinesize + x) = h->mb[i*16+y*4+x];
-                }
-            }
-        }
-        for(i=20; i<20+4; i++) {
-            for (y=0; y<4; y++) {
-                for (x=0; x<4; x++) {
-                    *(dest_cr + block_offset[i] + y*uvlinesize + x) = h->mb[i*16+y*4+x];
-                }
-            }
-        }
-    } else {
-        if(IS_INTRA(mb_type)){
-            if(h->deblocking_filter && !FRAME_MBAFF)
-                xchg_mb_border(h, dest_y, dest_cb, dest_cr, linesize, uvlinesize, 1);
-
-            if(!(s->flags&CODEC_FLAG_GRAY)){
-                h->pred8x8[ h->chroma_pred_mode ](dest_cb, uvlinesize);
-                h->pred8x8[ h->chroma_pred_mode ](dest_cr, uvlinesize);
-            }
-
-            if(IS_INTRA4x4(mb_type)){
-                if(!s->encoding){
-                    if(IS_8x8DCT(mb_type)){
-                        for(i=0; i<16; i+=4){
-                            uint8_t * const ptr= dest_y + block_offset[i];
-                            const int dir= h->intra4x4_pred_mode_cache[ scan8[i] ];
-                            const int nnz = h->non_zero_count_cache[ scan8[i] ];
-                            h->pred8x8l[ dir ](ptr, (h->topleft_samples_available<<i)&0x8000,
-                                                   (h->topright_samples_available<<(i+1))&0x8000, linesize);
-                            if(nnz){
-                                if(nnz == 1 && h->mb[i*16])
-                                    idct_dc_add(ptr, h->mb + i*16, linesize);
-                                else
-                                    idct_add(ptr, h->mb + i*16, linesize);
-                            }
-                        }
-                    }else
-                    for(i=0; i<16; i++){
-                        uint8_t * const ptr= dest_y + block_offset[i];
-                        uint8_t *topright;
-                        const int dir= h->intra4x4_pred_mode_cache[ scan8[i] ];
-                        int nnz, tr;
-
-                        if(dir == DIAG_DOWN_LEFT_PRED || dir == VERT_LEFT_PRED){
-                            const int topright_avail= (h->topright_samples_available<<i)&0x8000;
-                            assert(mb_y || linesize <= block_offset[i]);
-                            if(!topright_avail){
-                                tr= ptr[3 - linesize]*0x01010101;
-                                topright= (uint8_t*) &tr;
-                            }else
-                                topright= ptr + 4 - linesize;
-                        }else
-                            topright= NULL;
-
-                        h->pred4x4[ dir ](ptr, topright, linesize);
-                        nnz = h->non_zero_count_cache[ scan8[i] ];
-                        if(nnz){
-                            if(s->codec_id == CODEC_ID_H264){
-                                if(nnz == 1 && h->mb[i*16])
-                                    idct_dc_add(ptr, h->mb + i*16, linesize);
-                                else
-                                    idct_add(ptr, h->mb + i*16, linesize);
-                            }else
-                                svq3_add_idct_c(ptr, h->mb + i*16, linesize, s->qscale, 0);
-                        }
-                    }
-                }
-            }else{
-                h->pred16x16[ h->intra16x16_pred_mode ](dest_y , linesize);
-                if(s->codec_id == CODEC_ID_H264){
-                    if(!transform_bypass)
-                        h264_luma_dc_dequant_idct_c(h->mb, s->qscale, h->dequant4_coeff[IS_INTRA(mb_type) ? 0:3][s->qscale][0]);
-                }else
-                    svq3_luma_dc_dequant_idct_c(h->mb, s->qscale);
-            }
-            if(h->deblocking_filter && !FRAME_MBAFF)
-                xchg_mb_border(h, dest_y, dest_cb, dest_cr, linesize, uvlinesize, 0);
-        }else if(s->codec_id == CODEC_ID_H264){
-            hl_motion(h, dest_y, dest_cb, dest_cr,
-                      s->me.qpel_put, s->dsp.put_h264_chroma_pixels_tab,
-                      s->me.qpel_avg, s->dsp.avg_h264_chroma_pixels_tab,
-                      s->dsp.weight_h264_pixels_tab, s->dsp.biweight_h264_pixels_tab);
-        }
-
-
-        if(!IS_INTRA4x4(mb_type)){
-            if(s->codec_id == CODEC_ID_H264){
-                if(IS_INTRA16x16(mb_type)){
-                    for(i=0; i<16; i++){
-                        if(h->non_zero_count_cache[ scan8[i] ])
-                            idct_add(dest_y + block_offset[i], h->mb + i*16, linesize);
-                        else if(h->mb[i*16])
-                            idct_dc_add(dest_y + block_offset[i], h->mb + i*16, linesize);
-                    }
-                }else{
-                    const int di = IS_8x8DCT(mb_type) ? 4 : 1;
-                    for(i=0; i<16; i+=di){
-                        int nnz = h->non_zero_count_cache[ scan8[i] ];
-                        if(nnz){
-                            if(nnz==1 && h->mb[i*16])
-                                idct_dc_add(dest_y + block_offset[i], h->mb + i*16, linesize);
-                            else
-                                idct_add(dest_y + block_offset[i], h->mb + i*16, linesize);
-                        }
-                    }
-                }
-            }else{
-                for(i=0; i<16; i++){
-                    if(h->non_zero_count_cache[ scan8[i] ] || h->mb[i*16]){ //FIXME benchmark weird rule, & below
-                        uint8_t * const ptr= dest_y + block_offset[i];
-                        svq3_add_idct_c(ptr, h->mb + i*16, linesize, s->qscale, IS_INTRA(mb_type) ? 1 : 0);
-                    }
-                }
-            }
-        }
-
-        if(!(s->flags&CODEC_FLAG_GRAY)){
-            uint8_t *dest[2] = {dest_cb, dest_cr};
-            if(transform_bypass){
-                idct_add = idct_dc_add = s->dsp.add_pixels4;
-            }else{
-                idct_add = s->dsp.h264_idct_add;
-                idct_dc_add = s->dsp.h264_idct_dc_add;
-                chroma_dc_dequant_idct_c(h->mb + 16*16, h->chroma_qp, h->dequant4_coeff[IS_INTRA(mb_type) ? 1:4][h->chroma_qp][0]);
-                chroma_dc_dequant_idct_c(h->mb + 16*16+4*16, h->chroma_qp, h->dequant4_coeff[IS_INTRA(mb_type) ? 2:5][h->chroma_qp][0]);
-            }
-            if(s->codec_id == CODEC_ID_H264){
-                for(i=16; i<16+8; i++){
-                    if(h->non_zero_count_cache[ scan8[i] ])
-                        idct_add(dest[(i&4)>>2] + block_offset[i], h->mb + i*16, uvlinesize);
-                    else if(h->mb[i*16])
-                        idct_dc_add(dest[(i&4)>>2] + block_offset[i], h->mb + i*16, uvlinesize);
-                }
-            }else{
-                for(i=16; i<16+8; i++){
-                    if(h->non_zero_count_cache[ scan8[i] ] || h->mb[i*16]){
-                        uint8_t * const ptr= dest[(i&4)>>2] + block_offset[i];
-                        svq3_add_idct_c(ptr, h->mb + i*16, uvlinesize, chroma_qp[s->qscale + 12] - 12, 2);
-                    }
-                }
-            }
-        }
-    }
-    if(h->deblocking_filter) {
-        if (FRAME_MBAFF) {
-            //FIXME try deblocking one mb at a time?
-            // the reduction in load/storing mvs and such might outweigh the extra backup/xchg_border
-            const int mb_y = s->mb_y - 1;
-            uint8_t  *pair_dest_y, *pair_dest_cb, *pair_dest_cr;
-            const int mb_xy= mb_x + mb_y*s->mb_stride;
-            const int mb_type_top   = s->current_picture.mb_type[mb_xy];
-            const int mb_type_bottom= s->current_picture.mb_type[mb_xy+s->mb_stride];
-            if (!bottom) return;
-            pair_dest_y  = s->current_picture.data[0] + (mb_y * 16* s->linesize  ) + mb_x * 16;
-            pair_dest_cb = s->current_picture.data[1] + (mb_y * 8 * s->uvlinesize) + mb_x * 8;
-            pair_dest_cr = s->current_picture.data[2] + (mb_y * 8 * s->uvlinesize) + mb_x * 8;
-
-            if(IS_INTRA(mb_type_top | mb_type_bottom))
-                xchg_pair_border(h, pair_dest_y, pair_dest_cb, pair_dest_cr, s->linesize, s->uvlinesize, 0);
-
-            backup_pair_border(h, pair_dest_y, pair_dest_cb, pair_dest_cr, s->linesize, s->uvlinesize);
-            // deblock a pair
-            // top
-            s->mb_y--;
-            tprintf("call mbaff filter_mb mb_x:%d mb_y:%d pair_dest_y = %p, dest_y = %p\n", mb_x, mb_y, pair_dest_y, dest_y);
-            fill_caches(h, mb_type_top, 1); //FIXME don't fill stuff which isn't used by filter_mb
-            h->chroma_qp = get_chroma_qp(h->pps.chroma_qp_index_offset, s->current_picture.qscale_table[mb_xy]);
-            filter_mb(h, mb_x, mb_y, pair_dest_y, pair_dest_cb, pair_dest_cr, linesize, uvlinesize);
-            // bottom
-            s->mb_y++;
-            tprintf("call mbaff filter_mb\n");
-            fill_caches(h, mb_type_bottom, 1); //FIXME don't fill stuff which isn't used by filter_mb
-            h->chroma_qp = get_chroma_qp(h->pps.chroma_qp_index_offset, s->current_picture.qscale_table[mb_xy+s->mb_stride]);
-            filter_mb(h, mb_x, mb_y+1, dest_y, dest_cb, dest_cr, linesize, uvlinesize);
-        } else {
-            tprintf("call filter_mb\n");
-            backup_mb_border(h, dest_y, dest_cb, dest_cr, linesize, uvlinesize);
-            fill_caches(h, mb_type, 1); //FIXME don't fill stuff which isn't used by filter_mb
-            filter_mb_fast(h, mb_x, mb_y, dest_y, dest_cb, dest_cr, linesize, uvlinesize);
-        }
-    }
-}
-
-/**
- * fills the default_ref_list.
- */
-static int fill_default_ref_list(H264Context *h){
-    MpegEncContext * const s = &h->s;
-    int i;
-    int smallest_poc_greater_than_current = -1;
-    Picture sorted_short_ref[32];
-
-    if(h->slice_type==B_TYPE){
-        int out_i;
-        int limit= INT_MIN;
-
-        /* sort frame according to poc in B slice */
-        for(out_i=0; out_i<h->short_ref_count; out_i++){
-            int best_i=INT_MIN;
-            int best_poc=INT_MAX;
-
-            for(i=0; i<h->short_ref_count; i++){
-                const int poc= h->short_ref[i]->poc;
-                if(poc > limit && poc < best_poc){
-                    best_poc= poc;
-                    best_i= i;
-                }
-            }
-
-            assert(best_i != INT_MIN);
-
-            limit= best_poc;
-            sorted_short_ref[out_i]= *h->short_ref[best_i];
-            tprintf("sorted poc: %d->%d poc:%d fn:%d\n", best_i, out_i, sorted_short_ref[out_i].poc, sorted_short_ref[out_i].frame_num);
-            if (-1 == smallest_poc_greater_than_current) {
-                if (h->short_ref[best_i]->poc >= s->current_picture_ptr->poc) {
-                    smallest_poc_greater_than_current = out_i;
-                }
-            }
-        }
-    }
-
-    if(s->picture_structure == PICT_FRAME){
-        if(h->slice_type==B_TYPE){
-            int list;
-            tprintf("current poc: %d, smallest_poc_greater_than_current: %d\n", s->current_picture_ptr->poc, smallest_poc_greater_than_current);
-
-            // find the largest poc
-            for(list=0; list<2; list++){
-                int index = 0;
-                int j= -99;
-                int step= list ? -1 : 1;
-
-                for(i=0; i<h->short_ref_count && index < h->ref_count[list]; i++, j+=step) {
-                    while(j<0 || j>= h->short_ref_count){
-                        if(j != -99 && step == (list ? -1 : 1))
-                            return -1;
-                        step = -step;
-                        j= smallest_poc_greater_than_current + (step>>1);
-                    }
-                    if(sorted_short_ref[j].reference != 3) continue;
-                    h->default_ref_list[list][index  ]= sorted_short_ref[j];
-                    h->default_ref_list[list][index++].pic_id= sorted_short_ref[j].frame_num;
-                }
-
-                for(i = 0; i < 16 && index < h->ref_count[ list ]; i++){
-                    if(h->long_ref[i] == NULL) continue;
-                    if(h->long_ref[i]->reference != 3) continue;
-
-                    h->default_ref_list[ list ][index  ]= *h->long_ref[i];
-                    h->default_ref_list[ list ][index++].pic_id= i;;
-                }
-
-                if(list && (smallest_poc_greater_than_current<=0 || smallest_poc_greater_than_current>=h->short_ref_count) && (1 < index)){
-                    // swap the two first elements of L1 when
-                    // L0 and L1 are identical
-                    Picture temp= h->default_ref_list[1][0];
-                    h->default_ref_list[1][0] = h->default_ref_list[1][1];
-                    h->default_ref_list[1][1] = temp;
-                }
-
-                if(index < h->ref_count[ list ])
-                    memset(&h->default_ref_list[list][index], 0, sizeof(Picture)*(h->ref_count[ list ] - index));
-            }
-        }else{
-            int index=0;
-            for(i=0; i<h->short_ref_count; i++){
-                if(h->short_ref[i]->reference != 3) continue; //FIXME refernce field shit
-                h->default_ref_list[0][index  ]= *h->short_ref[i];
-                h->default_ref_list[0][index++].pic_id= h->short_ref[i]->frame_num;
-            }
-            for(i = 0; i < 16; i++){
-                if(h->long_ref[i] == NULL) continue;
-                if(h->long_ref[i]->reference != 3) continue;
-                h->default_ref_list[0][index  ]= *h->long_ref[i];
-                h->default_ref_list[0][index++].pic_id= i;;
-            }
-            if(index < h->ref_count[0])
-                memset(&h->default_ref_list[0][index], 0, sizeof(Picture)*(h->ref_count[0] - index));
-        }
-    }else{ //FIELD
-        if(h->slice_type==B_TYPE){
-        }else{
-            //FIXME second field balh
-        }
-    }
-#ifdef TRACE
-    for (i=0; i<h->ref_count[0]; i++) {
-        tprintf("List0: %s fn:%d 0x%p\n", (h->default_ref_list[0][i].long_ref ? "LT" : "ST"), h->default_ref_list[0][i].pic_id, h->default_ref_list[0][i].data[0]);
-    }
-    if(h->slice_type==B_TYPE){
-        for (i=0; i<h->ref_count[1]; i++) {
-            tprintf("List1: %s fn:%d 0x%p\n", (h->default_ref_list[1][i].long_ref ? "LT" : "ST"), h->default_ref_list[1][i].pic_id, h->default_ref_list[0][i].data[0]);
-        }
-    }
-#endif
-    return 0;
-}
-
-static void print_short_term(H264Context *h);
-static void print_long_term(H264Context *h);
-
-static int decode_ref_pic_list_reordering(H264Context *h){
-    MpegEncContext * const s = &h->s;
-    int list, index;
-
-    print_short_term(h);
-    print_long_term(h);
-    if(h->slice_type==I_TYPE || h->slice_type==SI_TYPE) return 0; //FIXME move before func
-
-    for(list=0; list<2; list++){
-        memcpy(h->ref_list[list], h->default_ref_list[list], sizeof(Picture)*h->ref_count[list]);
-
-        if(get_bits1(&s->gb)){
-            int pred= h->curr_pic_num;
-
-            for(index=0; ; index++){
-                int reordering_of_pic_nums_idc= get_ue_golomb(&s->gb);
-                int pic_id;
-                int i;
-                Picture *ref = NULL;
-
-                if(reordering_of_pic_nums_idc==3)
-                    break;
-
-                if(index >= h->ref_count[list]){
-                    av_log(h->s.avctx, AV_LOG_ERROR, "reference count overflow\n");
-                    return -1;
-                }
-
-                if(reordering_of_pic_nums_idc<3){
-                    if(reordering_of_pic_nums_idc<2){
-                        const int abs_diff_pic_num= get_ue_golomb(&s->gb) + 1;
-
-                        if(abs_diff_pic_num >= h->max_pic_num){
-                            av_log(h->s.avctx, AV_LOG_ERROR, "abs_diff_pic_num overflow\n");
-                            return -1;
-                        }
-
-                        if(reordering_of_pic_nums_idc == 0) pred-= abs_diff_pic_num;
-                        else                                pred+= abs_diff_pic_num;
-                        pred &= h->max_pic_num - 1;
-
-                        for(i= h->short_ref_count-1; i>=0; i--){
-                            ref = h->short_ref[i];
-                            assert(ref->reference == 3);
-                            assert(!ref->long_ref);
-                            if(ref->data[0] != NULL && ref->frame_num == pred && ref->long_ref == 0) // ignore non existing pictures by testing data[0] pointer
-                                break;
-                        }
-                        if(i>=0)
-                            ref->pic_id= ref->frame_num;
-                    }else{
-                        pic_id= get_ue_golomb(&s->gb); //long_term_pic_idx
-                        ref = h->long_ref[pic_id];
-                        ref->pic_id= pic_id;
-                        assert(ref->reference == 3);
-                        assert(ref->long_ref);
-                        i=0;
-                    }
-
-                    if (i < 0) {
-                        av_log(h->s.avctx, AV_LOG_ERROR, "reference picture missing during reorder\n");
-                        memset(&h->ref_list[list][index], 0, sizeof(Picture)); //FIXME
-                    } else {
-                        for(i=index; i+1<h->ref_count[list]; i++){
-                            if(ref->long_ref == h->ref_list[list][i].long_ref && ref->pic_id == h->ref_list[list][i].pic_id)
-                                break;
-                        }
-                        for(; i > index; i--){
-                            h->ref_list[list][i]= h->ref_list[list][i-1];
-                        }
-                        h->ref_list[list][index]= *ref;
-                    }
-                }else{
-                    av_log(h->s.avctx, AV_LOG_ERROR, "illegal reordering_of_pic_nums_idc\n");
-                    return -1;
-                }
-            }
-        }
-
-        if(h->slice_type!=B_TYPE) break;
-    }
-    for(list=0; list<2; list++){
-        for(index= 0; index < h->ref_count[list]; index++){
-            if(!h->ref_list[list][index].data[0])
-                h->ref_list[list][index]= s->current_picture;
-        }
-        if(h->slice_type!=B_TYPE) break;
-    }
-
-    if(h->slice_type==B_TYPE && !h->direct_spatial_mv_pred)
-        direct_dist_scale_factor(h);
-    direct_ref_list_init(h);
-    return 0;
-}
-
-static void fill_mbaff_ref_list(H264Context *h){
-    int list, i, j;
-    for(list=0; list<2; list++){
-        for(i=0; i<h->ref_count[list]; i++){
-            Picture *frame = &h->ref_list[list][i];
-            Picture *field = &h->ref_list[list][16+2*i];
-            field[0] = *frame;
-            for(j=0; j<3; j++)
-                field[0].linesize[j] <<= 1;
-            field[1] = field[0];
-            for(j=0; j<3; j++)
-                field[1].data[j] += frame->linesize[j];
-
-            h->luma_weight[list][16+2*i] = h->luma_weight[list][16+2*i+1] = h->luma_weight[list][i];
-            h->luma_offset[list][16+2*i] = h->luma_offset[list][16+2*i+1] = h->luma_offset[list][i];
-            for(j=0; j<2; j++){
-                h->chroma_weight[list][16+2*i][j] = h->chroma_weight[list][16+2*i+1][j] = h->chroma_weight[list][i][j];
-                h->chroma_offset[list][16+2*i][j] = h->chroma_offset[list][16+2*i+1][j] = h->chroma_offset[list][i][j];
-            }
-        }
-    }
-    for(j=0; j<h->ref_count[1]; j++){
-        for(i=0; i<h->ref_count[0]; i++)
-            h->implicit_weight[j][16+2*i] = h->implicit_weight[j][16+2*i+1] = h->implicit_weight[j][i];
-        memcpy(h->implicit_weight[16+2*j],   h->implicit_weight[j], sizeof(*h->implicit_weight));
-        memcpy(h->implicit_weight[16+2*j+1], h->implicit_weight[j], sizeof(*h->implicit_weight));
-    }
-}
-
-static int pred_weight_table(H264Context *h){
-    MpegEncContext * const s = &h->s;
-    int list, i;
-    int luma_def, chroma_def;
-
-    h->use_weight= 0;
-    h->use_weight_chroma= 0;
-    h->luma_log2_weight_denom= get_ue_golomb(&s->gb);
-    h->chroma_log2_weight_denom= get_ue_golomb(&s->gb);
-    luma_def = 1<<h->luma_log2_weight_denom;
-    chroma_def = 1<<h->chroma_log2_weight_denom;
-
-    for(list=0; list<2; list++){
-        for(i=0; i<h->ref_count[list]; i++){
-            int luma_weight_flag, chroma_weight_flag;
-
-            luma_weight_flag= get_bits1(&s->gb);
-            if(luma_weight_flag){
-                h->luma_weight[list][i]= get_se_golomb(&s->gb);
-                h->luma_offset[list][i]= get_se_golomb(&s->gb);
-                if(   h->luma_weight[list][i] != luma_def
-                   || h->luma_offset[list][i] != 0)
-                    h->use_weight= 1;
-            }else{
-                h->luma_weight[list][i]= luma_def;
-                h->luma_offset[list][i]= 0;
-            }
-
-            chroma_weight_flag= get_bits1(&s->gb);
-            if(chroma_weight_flag){
-                int j;
-                for(j=0; j<2; j++){
-                    h->chroma_weight[list][i][j]= get_se_golomb(&s->gb);
-                    h->chroma_offset[list][i][j]= get_se_golomb(&s->gb);
-                    if(   h->chroma_weight[list][i][j] != chroma_def
-                       || h->chroma_offset[list][i][j] != 0)
-                        h->use_weight_chroma= 1;
-                }
-            }else{
-                int j;
-                for(j=0; j<2; j++){
-                    h->chroma_weight[list][i][j]= chroma_def;
-                    h->chroma_offset[list][i][j]= 0;
-                }
-            }
-        }
-        if(h->slice_type != B_TYPE) break;
-    }
-    h->use_weight= h->use_weight || h->use_weight_chroma;
-    return 0;
-}
-
-static void implicit_weight_table(H264Context *h){
-    MpegEncContext * const s = &h->s;
-    int ref0, ref1;
-    int cur_poc = s->current_picture_ptr->poc;
-
-    if(   h->ref_count[0] == 1 && h->ref_count[1] == 1
-       && h->ref_list[0][0].poc + h->ref_list[1][0].poc == 2*cur_poc){
-        h->use_weight= 0;
-        h->use_weight_chroma= 0;
-        return;
-    }
-
-    h->use_weight= 2;
-    h->use_weight_chroma= 2;
-    h->luma_log2_weight_denom= 5;
-    h->chroma_log2_weight_denom= 5;
-
-    for(ref0=0; ref0 < h->ref_count[0]; ref0++){
-        int poc0 = h->ref_list[0][ref0].poc;
-        for(ref1=0; ref1 < h->ref_count[1]; ref1++){
-            int poc1 = h->ref_list[1][ref1].poc;
-            int td = clip(poc1 - poc0, -128, 127);
-            if(td){
-                int tb = clip(cur_poc - poc0, -128, 127);
-                int tx = (16384 + (FFABS(td) >> 1)) / td;
-                int dist_scale_factor = clip((tb*tx + 32) >> 6, -1024, 1023) >> 2;
-                if(dist_scale_factor < -64 || dist_scale_factor > 128)
-                    h->implicit_weight[ref0][ref1] = 32;
-                else
-                    h->implicit_weight[ref0][ref1] = 64 - dist_scale_factor;
-            }else
-                h->implicit_weight[ref0][ref1] = 32;
-        }
-    }
-}
-
-static inline void unreference_pic(H264Context *h, Picture *pic){
-    int i;
-    pic->reference=0;
-    if(pic == h->delayed_output_pic)
-        pic->reference=1;
-    else{
-        for(i = 0; h->delayed_pic[i]; i++)
-            if(pic == h->delayed_pic[i]){
-                pic->reference=1;
-                break;
-            }
-    }
-}
-
-/**
- * instantaneous decoder refresh.
- */
-static void idr(H264Context *h){
-    int i;
-
-    for(i=0; i<16; i++){
-        if (h->long_ref[i] != NULL) {
-            unreference_pic(h, h->long_ref[i]);
-            h->long_ref[i]= NULL;
-        }
-    }
-    h->long_ref_count=0;
-
-    for(i=0; i<h->short_ref_count; i++){
-        unreference_pic(h, h->short_ref[i]);
-        h->short_ref[i]= NULL;
-    }
-    h->short_ref_count=0;
-}
-
-/* forget old pics after a seek */
-static void flush_dpb(AVCodecContext *avctx){
-    H264Context *h= avctx->priv_data;
-    int i;
-    for(i=0; i<16; i++) {
-        if(h->delayed_pic[i])
-            h->delayed_pic[i]->reference= 0;
-        h->delayed_pic[i]= NULL;
-    }
-    if(h->delayed_output_pic)
-        h->delayed_output_pic->reference= 0;
-    h->delayed_output_pic= NULL;
-    idr(h);
-    if(h->s.current_picture_ptr)
-        h->s.current_picture_ptr->reference= 0;
-}
-
-/**
- *
- * @return the removed picture or NULL if an error occurs
- */
-static Picture * remove_short(H264Context *h, int frame_num){
-    MpegEncContext * const s = &h->s;
-    int i;
-
-    if(s->avctx->debug&FF_DEBUG_MMCO)
-        av_log(h->s.avctx, AV_LOG_DEBUG, "remove short %d count %d\n", frame_num, h->short_ref_count);
-
-    for(i=0; i<h->short_ref_count; i++){
-        Picture *pic= h->short_ref[i];
-        if(s->avctx->debug&FF_DEBUG_MMCO)
-            av_log(h->s.avctx, AV_LOG_DEBUG, "%d %d %p\n", i, pic->frame_num, pic);
-        if(pic->frame_num == frame_num){
-            h->short_ref[i]= NULL;
-            memmove(&h->short_ref[i], &h->short_ref[i+1], (h->short_ref_count - i - 1)*sizeof(Picture*));
-            h->short_ref_count--;
-            return pic;
-        }
-    }
-    return NULL;
-}
-
-/**
- *
- * @return the removed picture or NULL if an error occurs
- */
-static Picture * remove_long(H264Context *h, int i){
-    Picture *pic;
-
-    pic= h->long_ref[i];
-    h->long_ref[i]= NULL;
-    if(pic) h->long_ref_count--;
-
-    return pic;
-}
-
-/**
- * print short term list
- */
-static void print_short_term(H264Context *h) {
-    uint32_t i;
-    if(h->s.avctx->debug&FF_DEBUG_MMCO) {
-        av_log(h->s.avctx, AV_LOG_DEBUG, "short term list:\n");
-        for(i=0; i<h->short_ref_count; i++){
-            Picture *pic= h->short_ref[i];
-            av_log(h->s.avctx, AV_LOG_DEBUG, "%d fn:%d poc:%d %p\n", i, pic->frame_num, pic->poc, pic->data[0]);
-        }
-    }
-}
-
-/**
- * print long term list
- */
-static void print_long_term(H264Context *h) {
-    uint32_t i;
-    if(h->s.avctx->debug&FF_DEBUG_MMCO) {
-        av_log(h->s.avctx, AV_LOG_DEBUG, "long term list:\n");
-        for(i = 0; i < 16; i++){
-            Picture *pic= h->long_ref[i];
-            if (pic) {
-                av_log(h->s.avctx, AV_LOG_DEBUG, "%d fn:%d poc:%d %p\n", i, pic->frame_num, pic->poc, pic->data[0]);
-            }
-        }
-    }
-}
-
-/**
- * Executes the reference picture marking (memory management control operations).
- */
-static int execute_ref_pic_marking(H264Context *h, MMCO *mmco, int mmco_count){
-    MpegEncContext * const s = &h->s;
-    int i, j;
-    int current_is_long=0;
-    Picture *pic;
-
-    if((s->avctx->debug&FF_DEBUG_MMCO) && mmco_count==0)
-        av_log(h->s.avctx, AV_LOG_DEBUG, "no mmco here\n");
-
-    for(i=0; i<mmco_count; i++){
-        if(s->avctx->debug&FF_DEBUG_MMCO)
-            av_log(h->s.avctx, AV_LOG_DEBUG, "mmco:%d %d %d\n", h->mmco[i].opcode, h->mmco[i].short_frame_num, h->mmco[i].long_index);
-
-        switch(mmco[i].opcode){
-        case MMCO_SHORT2UNUSED:
-            pic= remove_short(h, mmco[i].short_frame_num);
-            if(pic)
-                unreference_pic(h, pic);
-            else if(s->avctx->debug&FF_DEBUG_MMCO)
-                av_log(h->s.avctx, AV_LOG_DEBUG, "mmco: remove_short() failure\n");
-            break;
-        case MMCO_SHORT2LONG:
-            pic= remove_long(h, mmco[i].long_index);
-            if(pic) unreference_pic(h, pic);
-
-            h->long_ref[ mmco[i].long_index ]= remove_short(h, mmco[i].short_frame_num);
-            h->long_ref[ mmco[i].long_index ]->long_ref=1;
-            h->long_ref_count++;
-            break;
-        case MMCO_LONG2UNUSED:
-            pic= remove_long(h, mmco[i].long_index);
-            if(pic)
-                unreference_pic(h, pic);
-            else if(s->avctx->debug&FF_DEBUG_MMCO)
-                av_log(h->s.avctx, AV_LOG_DEBUG, "mmco: remove_long() failure\n");
-            break;
-        case MMCO_LONG:
-            pic= remove_long(h, mmco[i].long_index);
-            if(pic) unreference_pic(h, pic);
-
-            h->long_ref[ mmco[i].long_index ]= s->current_picture_ptr;
-            h->long_ref[ mmco[i].long_index ]->long_ref=1;
-            h->long_ref_count++;
-
-            current_is_long=1;
-            break;
-        case MMCO_SET_MAX_LONG:
-            assert(mmco[i].long_index <= 16);
-            // just remove the long term which index is greater than new max
-            for(j = mmco[i].long_index; j<16; j++){
-                pic = remove_long(h, j);
-                if (pic) unreference_pic(h, pic);
-            }
-            break;
-        case MMCO_RESET:
-            while(h->short_ref_count){
-                pic= remove_short(h, h->short_ref[0]->frame_num);
-                unreference_pic(h, pic);
-            }
-            for(j = 0; j < 16; j++) {
-                pic= remove_long(h, j);
-                if(pic) unreference_pic(h, pic);
-            }
-            break;
-        default: assert(0);
-        }
-    }
-
-    if(!current_is_long){
-        pic= remove_short(h, s->current_picture_ptr->frame_num);
-        if(pic){
-            unreference_pic(h, pic);
-            av_log(h->s.avctx, AV_LOG_ERROR, "illegal short term buffer state detected\n");
-        }
-
-        if(h->short_ref_count)
-            memmove(&h->short_ref[1], &h->short_ref[0], h->short_ref_count*sizeof(Picture*));
-
-        h->short_ref[0]= s->current_picture_ptr;
-        h->short_ref[0]->long_ref=0;
-        h->short_ref_count++;
-    }
-
-    print_short_term(h);
-    print_long_term(h);
-    return 0;
-}
-
-static int decode_ref_pic_marking(H264Context *h){
-    MpegEncContext * const s = &h->s;
-    int i;
-
-    if(h->nal_unit_type == NAL_IDR_SLICE){ //FIXME fields
-        s->broken_link= get_bits1(&s->gb) -1;
-        h->mmco[0].long_index= get_bits1(&s->gb) - 1; // current_long_term_idx
-        if(h->mmco[0].long_index == -1)
-            h->mmco_index= 0;
-        else{
-            h->mmco[0].opcode= MMCO_LONG;
-            h->mmco_index= 1;
-        }
-    }else{
-        if(get_bits1(&s->gb)){ // adaptive_ref_pic_marking_mode_flag
-            for(i= 0; i<MAX_MMCO_COUNT; i++) {
-                MMCOOpcode opcode= get_ue_golomb(&s->gb);;
-
-                h->mmco[i].opcode= opcode;
-                if(opcode==MMCO_SHORT2UNUSED || opcode==MMCO_SHORT2LONG){
-                    h->mmco[i].short_frame_num= (h->frame_num - get_ue_golomb(&s->gb) - 1) & ((1<<h->sps.log2_max_frame_num)-1); //FIXME fields
-/*                    if(h->mmco[i].short_frame_num >= h->short_ref_count || h->short_ref[ h->mmco[i].short_frame_num ] == NULL){
-                        av_log(s->avctx, AV_LOG_ERROR, "illegal short ref in memory management control operation %d\n", mmco);
-                        return -1;
-                    }*/
-                }
-                if(opcode==MMCO_SHORT2LONG || opcode==MMCO_LONG2UNUSED || opcode==MMCO_LONG || opcode==MMCO_SET_MAX_LONG){
-                    h->mmco[i].long_index= get_ue_golomb(&s->gb);
-                    if(/*h->mmco[i].long_index >= h->long_ref_count || h->long_ref[ h->mmco[i].long_index ] == NULL*/ h->mmco[i].long_index >= 16){
-                        av_log(h->s.avctx, AV_LOG_ERROR, "illegal long ref in memory management control operation %d\n", opcode);
-                        return -1;
-                    }
-                }
-
-                if(opcode > MMCO_LONG){
-                    av_log(h->s.avctx, AV_LOG_ERROR, "illegal memory management control operation %d\n", opcode);
-                    return -1;
-                }
-                if(opcode == MMCO_END)
-                    break;
-            }
-            h->mmco_index= i;
-        }else{
-            assert(h->long_ref_count + h->short_ref_count <= h->sps.ref_frame_count);
-
-            if(h->long_ref_count + h->short_ref_count == h->sps.ref_frame_count){ //FIXME fields
-                h->mmco[0].opcode= MMCO_SHORT2UNUSED;
-                h->mmco[0].short_frame_num= h->short_ref[ h->short_ref_count - 1 ]->frame_num;
-                h->mmco_index= 1;
-            }else
-                h->mmco_index= 0;
-        }
-    }
-
-    return 0;
-}
-
-static int init_poc(H264Context *h){
-    MpegEncContext * const s = &h->s;
-    const int max_frame_num= 1<<h->sps.log2_max_frame_num;
-    int field_poc[2];
-
-    if(h->nal_unit_type == NAL_IDR_SLICE){
-        h->frame_num_offset= 0;
-    }else{
-        if(h->frame_num < h->prev_frame_num)
-            h->frame_num_offset= h->prev_frame_num_offset + max_frame_num;
-        else
-            h->frame_num_offset= h->prev_frame_num_offset;
-    }
-
-    if(h->sps.poc_type==0){
-        const int max_poc_lsb= 1<<h->sps.log2_max_poc_lsb;
-
-        if(h->nal_unit_type == NAL_IDR_SLICE){
-             h->prev_poc_msb=
-             h->prev_poc_lsb= 0;
-        }
-
-        if     (h->poc_lsb < h->prev_poc_lsb && h->prev_poc_lsb - h->poc_lsb >= max_poc_lsb/2)
-            h->poc_msb = h->prev_poc_msb + max_poc_lsb;
-        else if(h->poc_lsb > h->prev_poc_lsb && h->prev_poc_lsb - h->poc_lsb < -max_poc_lsb/2)
-            h->poc_msb = h->prev_poc_msb - max_poc_lsb;
-        else
-            h->poc_msb = h->prev_poc_msb;
-//printf("poc: %d %d\n", h->poc_msb, h->poc_lsb);
-        field_poc[0] =
-        field_poc[1] = h->poc_msb + h->poc_lsb;
-        if(s->picture_structure == PICT_FRAME)
-            field_poc[1] += h->delta_poc_bottom;
-    }else if(h->sps.poc_type==1){
-        int abs_frame_num, expected_delta_per_poc_cycle, expectedpoc;
-        int i;
-
-        if(h->sps.poc_cycle_length != 0)
-            abs_frame_num = h->frame_num_offset + h->frame_num;
-        else
-            abs_frame_num = 0;
-
-        if(h->nal_ref_idc==0 && abs_frame_num > 0)
-            abs_frame_num--;
-
-        expected_delta_per_poc_cycle = 0;
-        for(i=0; i < h->sps.poc_cycle_length; i++)
-            expected_delta_per_poc_cycle += h->sps.offset_for_ref_frame[ i ]; //FIXME integrate during sps parse
-
-        if(abs_frame_num > 0){
-            int poc_cycle_cnt          = (abs_frame_num - 1) / h->sps.poc_cycle_length;
-            int frame_num_in_poc_cycle = (abs_frame_num - 1) % h->sps.poc_cycle_length;
-
-            expectedpoc = poc_cycle_cnt * expected_delta_per_poc_cycle;
-            for(i = 0; i <= frame_num_in_poc_cycle; i++)
-                expectedpoc = expectedpoc + h->sps.offset_for_ref_frame[ i ];
-        } else
-            expectedpoc = 0;
-
-        if(h->nal_ref_idc == 0)
-            expectedpoc = expectedpoc + h->sps.offset_for_non_ref_pic;
-
-        field_poc[0] = expectedpoc + h->delta_poc[0];
-        field_poc[1] = field_poc[0] + h->sps.offset_for_top_to_bottom_field;
-
-        if(s->picture_structure == PICT_FRAME)
-            field_poc[1] += h->delta_poc[1];
-    }else{
-        int poc;
-        if(h->nal_unit_type == NAL_IDR_SLICE){
-            poc= 0;
-        }else{
-            if(h->nal_ref_idc) poc= 2*(h->frame_num_offset + h->frame_num);
-            else               poc= 2*(h->frame_num_offset + h->frame_num) - 1;
-        }
-        field_poc[0]= poc;
-        field_poc[1]= poc;
-    }
-
-    if(s->picture_structure != PICT_BOTTOM_FIELD)
-        s->current_picture_ptr->field_poc[0]= field_poc[0];
-    if(s->picture_structure != PICT_TOP_FIELD)
-        s->current_picture_ptr->field_poc[1]= field_poc[1];
-    if(s->picture_structure == PICT_FRAME) // FIXME field pix?
-        s->current_picture_ptr->poc= FFMIN(field_poc[0], field_poc[1]);
-
-    return 0;
-}
-
-/**
- * decodes a slice header.
- * this will allso call MPV_common_init() and frame_start() as needed
- */
-static int decode_slice_header(H264Context *h){
-    MpegEncContext * const s = &h->s;
-    int first_mb_in_slice, pps_id;
-    int num_ref_idx_active_override_flag;
-    static const uint8_t slice_type_map[5]= {P_TYPE, B_TYPE, I_TYPE, SP_TYPE, SI_TYPE};
-    int slice_type;
-    int default_ref_list_done = 0;
-
-    s->current_picture.reference= h->nal_ref_idc != 0;
-    s->dropable= h->nal_ref_idc == 0;
-
-    first_mb_in_slice= get_ue_golomb(&s->gb);
-
-    slice_type= get_ue_golomb(&s->gb);
-    if(slice_type > 9){
-        av_log(h->s.avctx, AV_LOG_ERROR, "slice type too large (%d) at %d %d\n", h->slice_type, s->mb_x, s->mb_y);
-        return -1;
-    }
-    if(slice_type > 4){
-        slice_type -= 5;
-        h->slice_type_fixed=1;
-    }else
-        h->slice_type_fixed=0;
-
-    slice_type= slice_type_map[ slice_type ];
-    if (slice_type == I_TYPE
-        || (h->slice_num != 0 && slice_type == h->slice_type) ) {
-        default_ref_list_done = 1;
-    }
-    h->slice_type= slice_type;
-
-    s->pict_type= h->slice_type; // to make a few old func happy, it's wrong though
-
-    pps_id= get_ue_golomb(&s->gb);
-    if(pps_id>255){
-        av_log(h->s.avctx, AV_LOG_ERROR, "pps_id out of range\n");
-        return -1;
-    }
-    h->pps= h->pps_buffer[pps_id];
-    if(h->pps.slice_group_count == 0){
-        av_log(h->s.avctx, AV_LOG_ERROR, "non existing PPS referenced\n");
-        return -1;
-    }
-
-    h->sps= h->sps_buffer[ h->pps.sps_id ];
-    if(h->sps.log2_max_frame_num == 0){
-        av_log(h->s.avctx, AV_LOG_ERROR, "non existing SPS referenced\n");
-        return -1;
-    }
-
-    if(h->dequant_coeff_pps != pps_id){
-        h->dequant_coeff_pps = pps_id;
-        init_dequant_tables(h);
-    }
-
-    s->mb_width= h->sps.mb_width;
-    s->mb_height= h->sps.mb_height * (2 - h->sps.frame_mbs_only_flag);
-
-    h->b_stride=  s->mb_width*4;
-    h->b8_stride= s->mb_width*2;
-
-    s->width = 16*s->mb_width - 2*(h->sps.crop_left + h->sps.crop_right );
-    if(h->sps.frame_mbs_only_flag)
-        s->height= 16*s->mb_height - 2*(h->sps.crop_top  + h->sps.crop_bottom);
-    else
-        s->height= 16*s->mb_height - 4*(h->sps.crop_top  + h->sps.crop_bottom); //FIXME recheck
-
-    if (s->context_initialized
-        && (   s->width != s->avctx->width || s->height != s->avctx->height)) {
-        free_tables(h);
-        MPV_common_end(s);
-    }
-    if (!s->context_initialized) {
-        if (MPV_common_init(s) < 0)
-            return -1;
-
-        if(s->dsp.h264_idct_add == ff_h264_idct_add_c){ //FIXME little ugly
-            memcpy(h->zigzag_scan, zigzag_scan, 16*sizeof(uint8_t));
-            memcpy(h-> field_scan,  field_scan, 16*sizeof(uint8_t));
-        }else{
-            int i;
-            for(i=0; i<16; i++){
-#define T(x) (x>>2) | ((x<<2) & 0xF)
-                h->zigzag_scan[i] = T(zigzag_scan[i]);
-                h-> field_scan[i] = T( field_scan[i]);
-#undef T
-            }
-        }
-        if(s->dsp.h264_idct8_add == ff_h264_idct8_add_c){
-            memcpy(h->zigzag_scan8x8,       zigzag_scan8x8,       64*sizeof(uint8_t));
-            memcpy(h->zigzag_scan8x8_cavlc, zigzag_scan8x8_cavlc, 64*sizeof(uint8_t));
-            memcpy(h->field_scan8x8,        field_scan8x8,        64*sizeof(uint8_t));
-            memcpy(h->field_scan8x8_cavlc,  field_scan8x8_cavlc,  64*sizeof(uint8_t));
-        }else{
-            int i;
-            for(i=0; i<64; i++){
-#define T(x) (x>>3) | ((x&7)<<3)
-                h->zigzag_scan8x8[i]       = T(zigzag_scan8x8[i]);
-                h->zigzag_scan8x8_cavlc[i] = T(zigzag_scan8x8_cavlc[i]);
-                h->field_scan8x8[i]        = T(field_scan8x8[i]);
-                h->field_scan8x8_cavlc[i]  = T(field_scan8x8_cavlc[i]);
-#undef T
-            }
-        }
-        if(h->sps.transform_bypass){ //FIXME same ugly
-            h->zigzag_scan_q0          = zigzag_scan;
-            h->zigzag_scan8x8_q0       = zigzag_scan8x8;
-            h->zigzag_scan8x8_cavlc_q0 = zigzag_scan8x8_cavlc;
-            h->field_scan_q0           = field_scan;
-            h->field_scan8x8_q0        = field_scan8x8;
-            h->field_scan8x8_cavlc_q0  = field_scan8x8_cavlc;
-        }else{
-            h->zigzag_scan_q0          = h->zigzag_scan;
-            h->zigzag_scan8x8_q0       = h->zigzag_scan8x8;
-            h->zigzag_scan8x8_cavlc_q0 = h->zigzag_scan8x8_cavlc;
-            h->field_scan_q0           = h->field_scan;
-            h->field_scan8x8_q0        = h->field_scan8x8;
-            h->field_scan8x8_cavlc_q0  = h->field_scan8x8_cavlc;
-        }
-
-        alloc_tables(h);
-
-        s->avctx->width = s->width;
-        s->avctx->height = s->height;
-        s->avctx->sample_aspect_ratio= h->sps.sar;
-        if(!s->avctx->sample_aspect_ratio.den)
-            s->avctx->sample_aspect_ratio.den = 1;
-
-        if(h->sps.timing_info_present_flag){
-            s->avctx->time_base= (AVRational){h->sps.num_units_in_tick * 2, h->sps.time_scale};
-            if(h->x264_build > 0 && h->x264_build < 44)
-                s->avctx->time_base.den *= 2;
-            av_reduce(&s->avctx->time_base.num, &s->avctx->time_base.den,
-                      s->avctx->time_base.num, s->avctx->time_base.den, 1<<30);
-        }
-    }
-
-    if(h->slice_num == 0){
-        if(frame_start(h) < 0)
-            return -1;
-    }
-
-    s->current_picture_ptr->frame_num= //FIXME frame_num cleanup
-    h->frame_num= get_bits(&s->gb, h->sps.log2_max_frame_num);
-
-    h->mb_mbaff = 0;
-    h->mb_aff_frame = 0;
-    if(h->sps.frame_mbs_only_flag){
-        s->picture_structure= PICT_FRAME;
-    }else{
-        if(get_bits1(&s->gb)) { //field_pic_flag
-            s->picture_structure= PICT_TOP_FIELD + get_bits1(&s->gb); //bottom_field_flag
-            av_log(h->s.avctx, AV_LOG_ERROR, "PAFF interlacing is not implemented\n");
-        } else {
-            s->picture_structure= PICT_FRAME;
-            h->mb_aff_frame = h->sps.mb_aff;
-        }
-    }
-
-    s->resync_mb_x = s->mb_x = first_mb_in_slice % s->mb_width;
-    s->resync_mb_y = s->mb_y = (first_mb_in_slice / s->mb_width) << h->mb_aff_frame;
-    if(s->mb_y >= s->mb_height){
-        return -1;
-    }
-
-    if(s->picture_structure==PICT_FRAME){
-        h->curr_pic_num=   h->frame_num;
-        h->max_pic_num= 1<< h->sps.log2_max_frame_num;
-    }else{
-        h->curr_pic_num= 2*h->frame_num;
-        h->max_pic_num= 1<<(h->sps.log2_max_frame_num + 1);
-    }
-
-    if(h->nal_unit_type == NAL_IDR_SLICE){
-        get_ue_golomb(&s->gb); /* idr_pic_id */
-    }
-
-    if(h->sps.poc_type==0){
-        h->poc_lsb= get_bits(&s->gb, h->sps.log2_max_poc_lsb);
-
-        if(h->pps.pic_order_present==1 && s->picture_structure==PICT_FRAME){
-            h->delta_poc_bottom= get_se_golomb(&s->gb);
-        }
-    }
-
-    if(h->sps.poc_type==1 && !h->sps.delta_pic_order_always_zero_flag){
-        h->delta_poc[0]= get_se_golomb(&s->gb);
-
-        if(h->pps.pic_order_present==1 && s->picture_structure==PICT_FRAME)
-            h->delta_poc[1]= get_se_golomb(&s->gb);
-    }
-
-    init_poc(h);
-
-    if(h->pps.redundant_pic_cnt_present){
-        h->redundant_pic_count= get_ue_golomb(&s->gb);
-    }
-
-    //set defaults, might be overriden a few line later
-    h->ref_count[0]= h->pps.ref_count[0];
-    h->ref_count[1]= h->pps.ref_count[1];
-
-    if(h->slice_type == P_TYPE || h->slice_type == SP_TYPE || h->slice_type == B_TYPE){
-        if(h->slice_type == B_TYPE){
-            h->direct_spatial_mv_pred= get_bits1(&s->gb);
-            if(h->sps.mb_aff && h->direct_spatial_mv_pred)
-                av_log(h->s.avctx, AV_LOG_ERROR, "MBAFF + spatial direct mode is not implemented\n");
-        }
-        num_ref_idx_active_override_flag= get_bits1(&s->gb);
-
-        if(num_ref_idx_active_override_flag){
-            h->ref_count[0]= get_ue_golomb(&s->gb) + 1;
-            if(h->slice_type==B_TYPE)
-                h->ref_count[1]= get_ue_golomb(&s->gb) + 1;
-
-            if(h->ref_count[0] > 32 || h->ref_count[1] > 32){
-                av_log(h->s.avctx, AV_LOG_ERROR, "reference overflow\n");
-                return -1;
-            }
-        }
-    }
-
-    if(!default_ref_list_done){
-        fill_default_ref_list(h);
-    }
-
-    if(decode_ref_pic_list_reordering(h) < 0)
-        return -1;
-
-    if(   (h->pps.weighted_pred          && (h->slice_type == P_TYPE || h->slice_type == SP_TYPE ))
-       || (h->pps.weighted_bipred_idc==1 && h->slice_type==B_TYPE ) )
-        pred_weight_table(h);
-    else if(h->pps.weighted_bipred_idc==2 && h->slice_type==B_TYPE)
-        implicit_weight_table(h);
-    else
-        h->use_weight = 0;
-
-    if(s->current_picture.reference)
-        decode_ref_pic_marking(h);
-
-    if(FRAME_MBAFF)
-        fill_mbaff_ref_list(h);
-
-    if( h->slice_type != I_TYPE && h->slice_type != SI_TYPE && h->pps.cabac )
-        h->cabac_init_idc = get_ue_golomb(&s->gb);
-
-    h->last_qscale_diff = 0;
-    s->qscale = h->pps.init_qp + get_se_golomb(&s->gb);
-    if(s->qscale<0 || s->qscale>51){
-        av_log(s->avctx, AV_LOG_ERROR, "QP %d out of range\n", s->qscale);
-        return -1;
-    }
-    h->chroma_qp = get_chroma_qp(h->pps.chroma_qp_index_offset, s->qscale);
-    //FIXME qscale / qp ... stuff
-    if(h->slice_type == SP_TYPE){
-        get_bits1(&s->gb); /* sp_for_switch_flag */
-    }
-    if(h->slice_type==SP_TYPE || h->slice_type == SI_TYPE){
-        get_se_golomb(&s->gb); /* slice_qs_delta */
-    }
-
-    h->deblocking_filter = 1;
-    h->slice_alpha_c0_offset = 0;
-    h->slice_beta_offset = 0;
-    if( h->pps.deblocking_filter_parameters_present ) {
-        h->deblocking_filter= get_ue_golomb(&s->gb);
-        if(h->deblocking_filter < 2)
-            h->deblocking_filter^= 1; // 1<->0
-
-        if( h->deblocking_filter ) {
-            h->slice_alpha_c0_offset = get_se_golomb(&s->gb) << 1;
-            h->slice_beta_offset = get_se_golomb(&s->gb) << 1;
-        }
-    }
-    if(   s->avctx->skip_loop_filter >= AVDISCARD_ALL
-       ||(s->avctx->skip_loop_filter >= AVDISCARD_NONKEY && h->slice_type != I_TYPE)
-       ||(s->avctx->skip_loop_filter >= AVDISCARD_BIDIR  && h->slice_type == B_TYPE)
-       ||(s->avctx->skip_loop_filter >= AVDISCARD_NONREF && h->nal_ref_idc == 0))
-        h->deblocking_filter= 0;
-
-#if 0 //FMO
-    if( h->pps.num_slice_groups > 1  && h->pps.mb_slice_group_map_type >= 3 && h->pps.mb_slice_group_map_type <= 5)
-        slice_group_change_cycle= get_bits(&s->gb, ?);
-#endif
-
-    h->slice_num++;
-
-    h->emu_edge_width= (s->flags&CODEC_FLAG_EMU_EDGE) ? 0 : 16;
-    h->emu_edge_height= FRAME_MBAFF ? 0 : h->emu_edge_width;
-
-    if(s->avctx->debug&FF_DEBUG_PICT_INFO){
-        av_log(h->s.avctx, AV_LOG_DEBUG, "slice:%d %s mb:%d %c pps:%d frame:%d poc:%d/%d ref:%d/%d qp:%d loop:%d:%d:%d weight:%d%s\n",
-               h->slice_num,
-               (s->picture_structure==PICT_FRAME ? "F" : s->picture_structure==PICT_TOP_FIELD ? "T" : "B"),
-               first_mb_in_slice,
-               av_get_pict_type_char(h->slice_type),
-               pps_id, h->frame_num,
-               s->current_picture_ptr->field_poc[0], s->current_picture_ptr->field_poc[1],
-               h->ref_count[0], h->ref_count[1],
-               s->qscale,
-               h->deblocking_filter, h->slice_alpha_c0_offset/2, h->slice_beta_offset/2,
-               h->use_weight,
-               h->use_weight==1 && h->use_weight_chroma ? "c" : ""
-               );
-    }
-
-    if((s->avctx->flags2 & CODEC_FLAG2_FAST) && !s->current_picture.reference){
-        s->me.qpel_put= s->dsp.put_2tap_qpel_pixels_tab;
-        s->me.qpel_avg= s->dsp.avg_2tap_qpel_pixels_tab;
-    }else{
-        s->me.qpel_put= s->dsp.put_h264_qpel_pixels_tab;
-        s->me.qpel_avg= s->dsp.avg_h264_qpel_pixels_tab;
-    }
-
-    return 0;
-}
-
-/**
- *
- */
-static inline int get_level_prefix(GetBitContext *gb){
-    unsigned int buf;
-    int log;
-
-    OPEN_READER(re, gb);
-    UPDATE_CACHE(re, gb);
-    buf=GET_CACHE(re, gb);
-
-    log= 32 - av_log2(buf);
-#ifdef TRACE
-    print_bin(buf>>(32-log), log);
-    av_log(NULL, AV_LOG_DEBUG, "%5d %2d %3d lpr @%5d in %s get_level_prefix\n", buf>>(32-log), log, log-1, get_bits_count(gb), __FILE__);
-#endif
-
-    LAST_SKIP_BITS(re, gb, log);
-    CLOSE_READER(re, gb);
-
-    return log-1;
-}
-
-static inline int get_dct8x8_allowed(H264Context *h){
-    int i;
-    for(i=0; i<4; i++){
-        if(!IS_SUB_8X8(h->sub_mb_type[i])
-           || (!h->sps.direct_8x8_inference_flag && IS_DIRECT(h->sub_mb_type[i])))
-            return 0;
-    }
-    return 1;
-}
-
-/**
- * decodes a residual block.
- * @param n block index
- * @param scantable scantable
- * @param max_coeff number of coefficients in the block
- * @return <0 if an error occured
- */
-static int decode_residual(H264Context *h, GetBitContext *gb, DCTELEM *block, int n, const uint8_t *scantable, const uint32_t *qmul, int max_coeff){
-    MpegEncContext * const s = &h->s;
-    static const int coeff_token_table_index[17]= {0, 0, 1, 1, 2, 2, 2, 2, 3, 3, 3, 3, 3, 3, 3, 3, 3};
-    int level[16];
-    int zeros_left, coeff_num, coeff_token, total_coeff, i, j, trailing_ones, run_before;
-
-    //FIXME put trailing_onex into the context
-
-    if(n == CHROMA_DC_BLOCK_INDEX){
-        coeff_token= get_vlc2(gb, chroma_dc_coeff_token_vlc.table, CHROMA_DC_COEFF_TOKEN_VLC_BITS, 1);
-        total_coeff= coeff_token>>2;
-    }else{
-        if(n == LUMA_DC_BLOCK_INDEX){
-            total_coeff= pred_non_zero_count(h, 0);
-            coeff_token= get_vlc2(gb, coeff_token_vlc[ coeff_token_table_index[total_coeff] ].table, COEFF_TOKEN_VLC_BITS, 2);
-            total_coeff= coeff_token>>2;
-        }else{
-            total_coeff= pred_non_zero_count(h, n);
-            coeff_token= get_vlc2(gb, coeff_token_vlc[ coeff_token_table_index[total_coeff] ].table, COEFF_TOKEN_VLC_BITS, 2);
-            total_coeff= coeff_token>>2;
-            h->non_zero_count_cache[ scan8[n] ]= total_coeff;
-        }
-    }
-
-    //FIXME set last_non_zero?
-
-    if(total_coeff==0)
-        return 0;
-
-    trailing_ones= coeff_token&3;
-    tprintf("trailing:%d, total:%d\n", trailing_ones, total_coeff);
-    assert(total_coeff<=16);
-
-    for(i=0; i<trailing_ones; i++){
-        level[i]= 1 - 2*get_bits1(gb);
-    }
-
-    if(i<total_coeff) {
-        int level_code, mask;
-        int suffix_length = total_coeff > 10 && trailing_ones < 3;
-        int prefix= get_level_prefix(gb);
-
-        //first coefficient has suffix_length equal to 0 or 1
-        if(prefix<14){ //FIXME try to build a large unified VLC table for all this
-            if(suffix_length)
-                level_code= (prefix<<suffix_length) + get_bits(gb, suffix_length); //part
-            else
-                level_code= (prefix<<suffix_length); //part
-        }else if(prefix==14){
-            if(suffix_length)
-                level_code= (prefix<<suffix_length) + get_bits(gb, suffix_length); //part
-            else
-                level_code= prefix + get_bits(gb, 4); //part
-        }else if(prefix==15){
-            level_code= (prefix<<suffix_length) + get_bits(gb, 12); //part
-            if(suffix_length==0) level_code+=15; //FIXME doesn't make (much)sense
-        }else{
-            av_log(h->s.avctx, AV_LOG_ERROR, "prefix too large at %d %d\n", s->mb_x, s->mb_y);
-            return -1;
-        }
-
-        if(trailing_ones < 3) level_code += 2;
-
-        suffix_length = 1;
-        if(level_code > 5)
-            suffix_length++;
-        mask= -(level_code&1);
-        level[i]= (((2+level_code)>>1) ^ mask) - mask;
-        i++;
-
-        //remaining coefficients have suffix_length > 0
-        for(;i<total_coeff;i++) {
-            static const int suffix_limit[7] = {0,5,11,23,47,95,INT_MAX };
-            prefix = get_level_prefix(gb);
-            if(prefix<15){
-                level_code = (prefix<<suffix_length) + get_bits(gb, suffix_length);
-            }else if(prefix==15){
-                level_code =  (prefix<<suffix_length) + get_bits(gb, 12);
-            }else{
-                av_log(h->s.avctx, AV_LOG_ERROR, "prefix too large at %d %d\n", s->mb_x, s->mb_y);
-                return -1;
-            }
-            mask= -(level_code&1);
-            level[i]= (((2+level_code)>>1) ^ mask) - mask;
-            if(level_code > suffix_limit[suffix_length])
-                suffix_length++;
-        }
-    }
-
-    if(total_coeff == max_coeff)
-        zeros_left=0;
-    else{
-        if(n == CHROMA_DC_BLOCK_INDEX)
-            zeros_left= get_vlc2(gb, chroma_dc_total_zeros_vlc[ total_coeff-1 ].table, CHROMA_DC_TOTAL_ZEROS_VLC_BITS, 1);
-        else
-            zeros_left= get_vlc2(gb, total_zeros_vlc[ total_coeff-1 ].table, TOTAL_ZEROS_VLC_BITS, 1);
-    }
-
-    coeff_num = zeros_left + total_coeff - 1;
-    j = scantable[coeff_num];
-    if(n > 24){
-        block[j] = level[0];
-        for(i=1;i<total_coeff;i++) {
-            if(zeros_left <= 0)
-                run_before = 0;
-            else if(zeros_left < 7){
-                run_before= get_vlc2(gb, run_vlc[zeros_left-1].table, RUN_VLC_BITS, 1);
-            }else{
-                run_before= get_vlc2(gb, run7_vlc.table, RUN7_VLC_BITS, 2);
-            }
-            zeros_left -= run_before;
-            coeff_num -= 1 + run_before;
-            j= scantable[ coeff_num ];
-
-            block[j]= level[i];
-        }
-    }else{
-        block[j] = (level[0] * qmul[j] + 32)>>6;
-        for(i=1;i<total_coeff;i++) {
-            if(zeros_left <= 0)
-                run_before = 0;
-            else if(zeros_left < 7){
-                run_before= get_vlc2(gb, run_vlc[zeros_left-1].table, RUN_VLC_BITS, 1);
-            }else{
-                run_before= get_vlc2(gb, run7_vlc.table, RUN7_VLC_BITS, 2);
-            }
-            zeros_left -= run_before;
-            coeff_num -= 1 + run_before;
-            j= scantable[ coeff_num ];
-
-            block[j]= (level[i] * qmul[j] + 32)>>6;
-        }
-    }
-
-    if(zeros_left<0){
-        av_log(h->s.avctx, AV_LOG_ERROR, "negative number of zero coeffs at %d %d\n", s->mb_x, s->mb_y);
-        return -1;
-    }
-
-    return 0;
-}
-
-static void predict_field_decoding_flag(H264Context *h){
-    MpegEncContext * const s = &h->s;
-    const int mb_xy= s->mb_x + s->mb_y*s->mb_stride;
-    int mb_type = (h->slice_table[mb_xy-1] == h->slice_num)
-                ? s->current_picture.mb_type[mb_xy-1]
-                : (h->slice_table[mb_xy-s->mb_stride] == h->slice_num)
-                ? s->current_picture.mb_type[mb_xy-s->mb_stride]
-                : 0;
-    h->mb_mbaff = h->mb_field_decoding_flag = IS_INTERLACED(mb_type) ? 1 : 0;
-}
-
-/**
- * decodes a P_SKIP or B_SKIP macroblock
- */
-static void decode_mb_skip(H264Context *h){
-    MpegEncContext * const s = &h->s;
-    const int mb_xy= s->mb_x + s->mb_y*s->mb_stride;
-    int mb_type=0;
-
-    memset(h->non_zero_count[mb_xy], 0, 16);
-    memset(h->non_zero_count_cache + 8, 0, 8*5); //FIXME ugly, remove pfui
-
-    if(MB_FIELD)
-        mb_type|= MB_TYPE_INTERLACED;
-
-    if( h->slice_type == B_TYPE )
-    {
-        // just for fill_caches. pred_direct_motion will set the real mb_type
-        mb_type|= MB_TYPE_16x16|MB_TYPE_P0L0|MB_TYPE_P0L1|MB_TYPE_DIRECT2|MB_TYPE_SKIP;
-
-        fill_caches(h, mb_type, 0); //FIXME check what is needed and what not ...
-        pred_direct_motion(h, &mb_type);
-        mb_type|= MB_TYPE_SKIP;
-    }
-    else
-    {
-        int mx, my;
-        mb_type|= MB_TYPE_16x16|MB_TYPE_P0L0|MB_TYPE_P1L0|MB_TYPE_SKIP;
-
-        fill_caches(h, mb_type, 0); //FIXME check what is needed and what not ...
-        pred_pskip_motion(h, &mx, &my);
-        fill_rectangle(&h->ref_cache[0][scan8[0]], 4, 4, 8, 0, 1);
-        fill_rectangle(  h->mv_cache[0][scan8[0]], 4, 4, 8, pack16to32(mx,my), 4);
-    }
-
-    write_back_motion(h, mb_type);
-    s->current_picture.mb_type[mb_xy]= mb_type;
-    s->current_picture.qscale_table[mb_xy]= s->qscale;
-    h->slice_table[ mb_xy ]= h->slice_num;
-    h->prev_mb_skipped= 1;
-}
-
-/**
- * decodes a macroblock
- * @returns 0 if ok, AC_ERROR / DC_ERROR / MV_ERROR if an error is noticed
- */
-static int decode_mb_cavlc(H264Context *h){
-    MpegEncContext * const s = &h->s;
-    const int mb_xy= s->mb_x + s->mb_y*s->mb_stride;
-    int mb_type, partition_count, cbp;
-    int dct8x8_allowed= h->pps.transform_8x8_mode;
-
-    s->dsp.clear_blocks(h->mb); //FIXME avoid if already clear (move after skip handlong?
-
-    tprintf("pic:%d mb:%d/%d\n", h->frame_num, s->mb_x, s->mb_y);
-    cbp = 0; /* avoid warning. FIXME: find a solution without slowing
-                down the code */
-    if(h->slice_type != I_TYPE && h->slice_type != SI_TYPE){
-        if(s->mb_skip_run==-1)
-            s->mb_skip_run= get_ue_golomb(&s->gb);
-
-        if (s->mb_skip_run--) {
-            if(FRAME_MBAFF && (s->mb_y&1) == 0){
-                if(s->mb_skip_run==0)
-                    h->mb_mbaff = h->mb_field_decoding_flag = get_bits1(&s->gb);
-                else
-                    predict_field_decoding_flag(h);
-            }
-            decode_mb_skip(h);
-            return 0;
-        }
-    }
-    if(FRAME_MBAFF){
-        if( (s->mb_y&1) == 0 )
-            h->mb_mbaff = h->mb_field_decoding_flag = get_bits1(&s->gb);
-    }else
-        h->mb_field_decoding_flag= (s->picture_structure!=PICT_FRAME);
-
-    h->prev_mb_skipped= 0;
-
-    mb_type= get_ue_golomb(&s->gb);
-    if(h->slice_type == B_TYPE){
-        if(mb_type < 23){
-            partition_count= b_mb_type_info[mb_type].partition_count;
-            mb_type=         b_mb_type_info[mb_type].type;
-        }else{
-            mb_type -= 23;
-            goto decode_intra_mb;
-        }
-    }else if(h->slice_type == P_TYPE /*|| h->slice_type == SP_TYPE */){
-        if(mb_type < 5){
-            partition_count= p_mb_type_info[mb_type].partition_count;
-            mb_type=         p_mb_type_info[mb_type].type;
-        }else{
-            mb_type -= 5;
-            goto decode_intra_mb;
-        }
-    }else{
-       assert(h->slice_type == I_TYPE);
-decode_intra_mb:
-        if(mb_type > 25){
-            av_log(h->s.avctx, AV_LOG_ERROR, "mb_type %d in %c slice too large at %d %d\n", mb_type, av_get_pict_type_char(h->slice_type), s->mb_x, s->mb_y);
-            return -1;
-        }
-        partition_count=0;
-        cbp= i_mb_type_info[mb_type].cbp;
-        h->intra16x16_pred_mode= i_mb_type_info[mb_type].pred_mode;
-        mb_type= i_mb_type_info[mb_type].type;
-    }
-
-    if(MB_FIELD)
-        mb_type |= MB_TYPE_INTERLACED;
-
-    h->slice_table[ mb_xy ]= h->slice_num;
-
-    if(IS_INTRA_PCM(mb_type)){
-        unsigned int x, y;
-
-        // we assume these blocks are very rare so we dont optimize it
-        align_get_bits(&s->gb);
-
-        // The pixels are stored in the same order as levels in h->mb array.
-        for(y=0; y<16; y++){
-            const int index= 4*(y&3) + 32*((y>>2)&1) + 128*(y>>3);
-            for(x=0; x<16; x++){
-                tprintf("LUMA ICPM LEVEL (%3d)\n", show_bits(&s->gb, 8));
-                h->mb[index + (x&3) + 16*((x>>2)&1) + 64*(x>>3)]= get_bits(&s->gb, 8);
-            }
-        }
-        for(y=0; y<8; y++){
-            const int index= 256 + 4*(y&3) + 32*(y>>2);
-            for(x=0; x<8; x++){
-                tprintf("CHROMA U ICPM LEVEL (%3d)\n", show_bits(&s->gb, 8));
-                h->mb[index + (x&3) + 16*(x>>2)]= get_bits(&s->gb, 8);
-            }
-        }
-        for(y=0; y<8; y++){
-            const int index= 256 + 64 + 4*(y&3) + 32*(y>>2);
-            for(x=0; x<8; x++){
-                tprintf("CHROMA V ICPM LEVEL (%3d)\n", show_bits(&s->gb, 8));
-                h->mb[index + (x&3) + 16*(x>>2)]= get_bits(&s->gb, 8);
-            }
-        }
-
-        // In deblocking, the quantizer is 0
-        s->current_picture.qscale_table[mb_xy]= 0;
-        h->chroma_qp = get_chroma_qp(h->pps.chroma_qp_index_offset, 0);
-        // All coeffs are present
-        memset(h->non_zero_count[mb_xy], 16, 16);
-
-        s->current_picture.mb_type[mb_xy]= mb_type;
-        return 0;
-    }
-
-    if(MB_MBAFF){
-        h->ref_count[0] <<= 1;
-        h->ref_count[1] <<= 1;
-    }
-
-    fill_caches(h, mb_type, 0);
-
-    //mb_pred
-    if(IS_INTRA(mb_type)){
-//            init_top_left_availability(h);
-            if(IS_INTRA4x4(mb_type)){
-                int i;
-                int di = 1;
-                if(dct8x8_allowed && get_bits1(&s->gb)){
-                    mb_type |= MB_TYPE_8x8DCT;
-                    di = 4;
-                }
-
-//                fill_intra4x4_pred_table(h);
-                for(i=0; i<16; i+=di){
-                    int mode= pred_intra_mode(h, i);
-
-                    if(!get_bits1(&s->gb)){
-                        const int rem_mode= get_bits(&s->gb, 3);
-                        mode = rem_mode + (rem_mode >= mode);
-                    }
-
-                    if(di==4)
-                        fill_rectangle( &h->intra4x4_pred_mode_cache[ scan8[i] ], 2, 2, 8, mode, 1 );
-                    else
-                        h->intra4x4_pred_mode_cache[ scan8[i] ] = mode;
-                }
-                write_back_intra_pred_mode(h);
-                if( check_intra4x4_pred_mode(h) < 0)
-                    return -1;
-            }else{
-                h->intra16x16_pred_mode= check_intra_pred_mode(h, h->intra16x16_pred_mode);
-                if(h->intra16x16_pred_mode < 0)
-                    return -1;
-            }
-            h->chroma_pred_mode= get_ue_golomb(&s->gb);
-
-            h->chroma_pred_mode= check_intra_pred_mode(h, h->chroma_pred_mode);
-            if(h->chroma_pred_mode < 0)
-                return -1;
-    }else if(partition_count==4){
-        int i, j, sub_partition_count[4], list, ref[2][4];
-
-        if(h->slice_type == B_TYPE){
-            for(i=0; i<4; i++){
-                h->sub_mb_type[i]= get_ue_golomb(&s->gb);
-                if(h->sub_mb_type[i] >=13){
-                    av_log(h->s.avctx, AV_LOG_ERROR, "B sub_mb_type %d out of range at %d %d\n", h->sub_mb_type[i], s->mb_x, s->mb_y);
-                    return -1;
-                }
-                sub_partition_count[i]= b_sub_mb_type_info[ h->sub_mb_type[i] ].partition_count;
-                h->sub_mb_type[i]=      b_sub_mb_type_info[ h->sub_mb_type[i] ].type;
-            }
-            if(   IS_DIRECT(h->sub_mb_type[0]) || IS_DIRECT(h->sub_mb_type[1])
-               || IS_DIRECT(h->sub_mb_type[2]) || IS_DIRECT(h->sub_mb_type[3])) {
-                pred_direct_motion(h, &mb_type);
-                h->ref_cache[0][scan8[4]] =
-                h->ref_cache[1][scan8[4]] =
-                h->ref_cache[0][scan8[12]] =
-                h->ref_cache[1][scan8[12]] = PART_NOT_AVAILABLE;
-            }
-        }else{
-            assert(h->slice_type == P_TYPE || h->slice_type == SP_TYPE); //FIXME SP correct ?
-            for(i=0; i<4; i++){
-                h->sub_mb_type[i]= get_ue_golomb(&s->gb);
-                if(h->sub_mb_type[i] >=4){
-                    av_log(h->s.avctx, AV_LOG_ERROR, "P sub_mb_type %d out of range at %d %d\n", h->sub_mb_type[i], s->mb_x, s->mb_y);
-                    return -1;
-                }
-                sub_partition_count[i]= p_sub_mb_type_info[ h->sub_mb_type[i] ].partition_count;
-                h->sub_mb_type[i]=      p_sub_mb_type_info[ h->sub_mb_type[i] ].type;
-            }
-        }
-
-        for(list=0; list<2; list++){
-            int ref_count= IS_REF0(mb_type) ? 1 : h->ref_count[list];
-            if(ref_count == 0) continue;
-            for(i=0; i<4; i++){
-                if(IS_DIRECT(h->sub_mb_type[i])) continue;
-                if(IS_DIR(h->sub_mb_type[i], 0, list)){
-                    ref[list][i] = get_te0_golomb(&s->gb, ref_count); //FIXME init to 0 before and skip?
-                }else{
-                 //FIXME
-                    ref[list][i] = -1;
-                }
-            }
-        }
-
-        if(dct8x8_allowed)
-            dct8x8_allowed = get_dct8x8_allowed(h);
-
-        for(list=0; list<2; list++){
-            const int ref_count= IS_REF0(mb_type) ? 1 : h->ref_count[list];
-            if(ref_count == 0) continue;
-
-            for(i=0; i<4; i++){
-                if(IS_DIRECT(h->sub_mb_type[i])) {
-                    h->ref_cache[list][ scan8[4*i] ] = h->ref_cache[list][ scan8[4*i]+1 ];
-                    continue;
-                }
-                h->ref_cache[list][ scan8[4*i]   ]=h->ref_cache[list][ scan8[4*i]+1 ]=
-                h->ref_cache[list][ scan8[4*i]+8 ]=h->ref_cache[list][ scan8[4*i]+9 ]= ref[list][i];
-
-                if(IS_DIR(h->sub_mb_type[i], 0, list)){
-                    const int sub_mb_type= h->sub_mb_type[i];
-                    const int block_width= (sub_mb_type & (MB_TYPE_16x16|MB_TYPE_16x8)) ? 2 : 1;
-                    for(j=0; j<sub_partition_count[i]; j++){
-                        int mx, my;
-                        const int index= 4*i + block_width*j;
-                        int16_t (* mv_cache)[2]= &h->mv_cache[list][ scan8[index] ];
-                        pred_motion(h, index, block_width, list, h->ref_cache[list][ scan8[index] ], &mx, &my);
-                        mx += get_se_golomb(&s->gb);
-                        my += get_se_golomb(&s->gb);
-                        tprintf("final mv:%d %d\n", mx, my);
-
-                        if(IS_SUB_8X8(sub_mb_type)){
-                            mv_cache[ 0 ][0]= mv_cache[ 1 ][0]=
-                            mv_cache[ 8 ][0]= mv_cache[ 9 ][0]= mx;
-                            mv_cache[ 0 ][1]= mv_cache[ 1 ][1]=
-                            mv_cache[ 8 ][1]= mv_cache[ 9 ][1]= my;
-                        }else if(IS_SUB_8X4(sub_mb_type)){
-                            mv_cache[ 0 ][0]= mv_cache[ 1 ][0]= mx;
-                            mv_cache[ 0 ][1]= mv_cache[ 1 ][1]= my;
-                        }else if(IS_SUB_4X8(sub_mb_type)){
-                            mv_cache[ 0 ][0]= mv_cache[ 8 ][0]= mx;
-                            mv_cache[ 0 ][1]= mv_cache[ 8 ][1]= my;
-                        }else{
-                            assert(IS_SUB_4X4(sub_mb_type));
-                            mv_cache[ 0 ][0]= mx;
-                            mv_cache[ 0 ][1]= my;
-                        }
-                    }
-                }else{
-                    uint32_t *p= (uint32_t *)&h->mv_cache[list][ scan8[4*i] ][0];
-                    p[0] = p[1]=
-                    p[8] = p[9]= 0;
-                }
-            }
-        }
-    }else if(IS_DIRECT(mb_type)){
-        pred_direct_motion(h, &mb_type);
-        dct8x8_allowed &= h->sps.direct_8x8_inference_flag;
-    }else{
-        int list, mx, my, i;
-         //FIXME we should set ref_idx_l? to 0 if we use that later ...
-        if(IS_16X16(mb_type)){
-            for(list=0; list<2; list++){
-                if(h->ref_count[list]>0){
-                    if(IS_DIR(mb_type, 0, list)){
-                        const int val= get_te0_golomb(&s->gb, h->ref_count[list]);
-                        fill_rectangle(&h->ref_cache[list][ scan8[0] ], 4, 4, 8, val, 1);
-                    }else
-                        fill_rectangle(&h->ref_cache[list][ scan8[0] ], 4, 4, 8, (LIST_NOT_USED&0xFF), 1);
-                }
-            }
-            for(list=0; list<2; list++){
-                if(IS_DIR(mb_type, 0, list)){
-                    pred_motion(h, 0, 4, list, h->ref_cache[list][ scan8[0] ], &mx, &my);
-                    mx += get_se_golomb(&s->gb);
-                    my += get_se_golomb(&s->gb);
-                    tprintf("final mv:%d %d\n", mx, my);
-
-                    fill_rectangle(h->mv_cache[list][ scan8[0] ], 4, 4, 8, pack16to32(mx,my), 4);
-                }else
-                    fill_rectangle(h->mv_cache[list][ scan8[0] ], 4, 4, 8, 0, 4);
-            }
-        }
-        else if(IS_16X8(mb_type)){
-            for(list=0; list<2; list++){
-                if(h->ref_count[list]>0){
-                    for(i=0; i<2; i++){
-                        if(IS_DIR(mb_type, i, list)){
-                            const int val= get_te0_golomb(&s->gb, h->ref_count[list]);
-                            fill_rectangle(&h->ref_cache[list][ scan8[0] + 16*i ], 4, 2, 8, val, 1);
-                        }else
-                            fill_rectangle(&h->ref_cache[list][ scan8[0] + 16*i ], 4, 2, 8, (LIST_NOT_USED&0xFF), 1);
-                    }
-                }
-            }
-            for(list=0; list<2; list++){
-                for(i=0; i<2; i++){
-                    if(IS_DIR(mb_type, i, list)){
-                        pred_16x8_motion(h, 8*i, list, h->ref_cache[list][scan8[0] + 16*i], &mx, &my);
-                        mx += get_se_golomb(&s->gb);
-                        my += get_se_golomb(&s->gb);
-                        tprintf("final mv:%d %d\n", mx, my);
-
-                        fill_rectangle(h->mv_cache[list][ scan8[0] + 16*i ], 4, 2, 8, pack16to32(mx,my), 4);
-                    }else
-                        fill_rectangle(h->mv_cache[list][ scan8[0] + 16*i ], 4, 2, 8, 0, 4);
-                }
-            }
-        }else{
-            assert(IS_8X16(mb_type));
-            for(list=0; list<2; list++){
-                if(h->ref_count[list]>0){
-                    for(i=0; i<2; i++){
-                        if(IS_DIR(mb_type, i, list)){ //FIXME optimize
-                            const int val= get_te0_golomb(&s->gb, h->ref_count[list]);
-                            fill_rectangle(&h->ref_cache[list][ scan8[0] + 2*i ], 2, 4, 8, val, 1);
-                        }else
-                            fill_rectangle(&h->ref_cache[list][ scan8[0] + 2*i ], 2, 4, 8, (LIST_NOT_USED&0xFF), 1);
-                    }
-                }
-            }
-            for(list=0; list<2; list++){
-                for(i=0; i<2; i++){
-                    if(IS_DIR(mb_type, i, list)){
-                        pred_8x16_motion(h, i*4, list, h->ref_cache[list][ scan8[0] + 2*i ], &mx, &my);
-                        mx += get_se_golomb(&s->gb);
-                        my += get_se_golomb(&s->gb);
-                        tprintf("final mv:%d %d\n", mx, my);
-
-                        fill_rectangle(h->mv_cache[list][ scan8[0] + 2*i ], 2, 4, 8, pack16to32(mx,my), 4);
-                    }else
-                        fill_rectangle(h->mv_cache[list][ scan8[0] + 2*i ], 2, 4, 8, 0, 4);
-                }
-            }
-        }
-    }
-
-    if(IS_INTER(mb_type))
-        write_back_motion(h, mb_type);
-
-    if(!IS_INTRA16x16(mb_type)){
-        cbp= get_ue_golomb(&s->gb);
-        if(cbp > 47){
-            av_log(h->s.avctx, AV_LOG_ERROR, "cbp too large (%d) at %d %d\n", cbp, s->mb_x, s->mb_y);
-            return -1;
-        }
-
-        if(IS_INTRA4x4(mb_type))
-            cbp= golomb_to_intra4x4_cbp[cbp];
-        else
-            cbp= golomb_to_inter_cbp[cbp];
-    }
-    h->cbp = cbp;
-
-    if(dct8x8_allowed && (cbp&15) && !IS_INTRA(mb_type)){
-        if(get_bits1(&s->gb))
-            mb_type |= MB_TYPE_8x8DCT;
-    }
-    s->current_picture.mb_type[mb_xy]= mb_type;
-
-    if(cbp || IS_INTRA16x16(mb_type)){
-        int i8x8, i4x4, chroma_idx;
-        int chroma_qp, dquant;
-        GetBitContext *gb= IS_INTRA(mb_type) ? h->intra_gb_ptr : h->inter_gb_ptr;
-        const uint8_t *scan, *scan8x8, *dc_scan;
-
-//        fill_non_zero_count_cache(h);
-
-        if(IS_INTERLACED(mb_type)){
-            scan8x8= s->qscale ? h->field_scan8x8_cavlc : h->field_scan8x8_cavlc_q0;
-            scan= s->qscale ? h->field_scan : h->field_scan_q0;
-            dc_scan= luma_dc_field_scan;
-        }else{
-            scan8x8= s->qscale ? h->zigzag_scan8x8_cavlc : h->zigzag_scan8x8_cavlc_q0;
-            scan= s->qscale ? h->zigzag_scan : h->zigzag_scan_q0;
-            dc_scan= luma_dc_zigzag_scan;
-        }
-
-        dquant= get_se_golomb(&s->gb);
-
-        if( dquant > 25 || dquant < -26 ){
-            av_log(h->s.avctx, AV_LOG_ERROR, "dquant out of range (%d) at %d %d\n", dquant, s->mb_x, s->mb_y);
-            return -1;
-        }
-
-        s->qscale += dquant;
-        if(((unsigned)s->qscale) > 51){
-            if(s->qscale<0) s->qscale+= 52;
-            else            s->qscale-= 52;
-        }
-
-        h->chroma_qp= chroma_qp= get_chroma_qp(h->pps.chroma_qp_index_offset, s->qscale);
-        if(IS_INTRA16x16(mb_type)){
-            if( decode_residual(h, h->intra_gb_ptr, h->mb, LUMA_DC_BLOCK_INDEX, dc_scan, h->dequant4_coeff[0][s->qscale], 16) < 0){
-                return -1; //FIXME continue if partitioned and other return -1 too
-            }
-
-            assert((cbp&15) == 0 || (cbp&15) == 15);
-
-            if(cbp&15){
-                for(i8x8=0; i8x8<4; i8x8++){
-                    for(i4x4=0; i4x4<4; i4x4++){
-                        const int index= i4x4 + 4*i8x8;
-                        if( decode_residual(h, h->intra_gb_ptr, h->mb + 16*index, index, scan + 1, h->dequant4_coeff[0][s->qscale], 15) < 0 ){
-                            return -1;
-                        }
-                    }
-                }
-            }else{
-                fill_rectangle(&h->non_zero_count_cache[scan8[0]], 4, 4, 8, 0, 1);
-            }
-        }else{
-            for(i8x8=0; i8x8<4; i8x8++){
-                if(cbp & (1<<i8x8)){
-                    if(IS_8x8DCT(mb_type)){
-                        DCTELEM *buf = &h->mb[64*i8x8];
-                        uint8_t *nnz;
-                        for(i4x4=0; i4x4<4; i4x4++){
-                            if( decode_residual(h, gb, buf, i4x4+4*i8x8, scan8x8+16*i4x4,
-                                                h->dequant8_coeff[IS_INTRA( mb_type ) ? 0:1][s->qscale], 16) <0 )
-                                return -1;
-                        }
-                        nnz= &h->non_zero_count_cache[ scan8[4*i8x8] ];
-                        nnz[0] += nnz[1] + nnz[8] + nnz[9];
-                    }else{
-                        for(i4x4=0; i4x4<4; i4x4++){
-                            const int index= i4x4 + 4*i8x8;
-
-                            if( decode_residual(h, gb, h->mb + 16*index, index, scan, h->dequant4_coeff[IS_INTRA( mb_type ) ? 0:3][s->qscale], 16) <0 ){
-                                return -1;
-                            }
-                        }
-                    }
-                }else{
-                    uint8_t * const nnz= &h->non_zero_count_cache[ scan8[4*i8x8] ];
-                    nnz[0] = nnz[1] = nnz[8] = nnz[9] = 0;
-                }
-            }
-        }
-
-        if(cbp&0x30){
-            for(chroma_idx=0; chroma_idx<2; chroma_idx++)
-                if( decode_residual(h, gb, h->mb + 256 + 16*4*chroma_idx, CHROMA_DC_BLOCK_INDEX, chroma_dc_scan, NULL, 4) < 0){
-                    return -1;
-                }
-        }
-
-        if(cbp&0x20){
-            for(chroma_idx=0; chroma_idx<2; chroma_idx++){
-                for(i4x4=0; i4x4<4; i4x4++){
-                    const int index= 16 + 4*chroma_idx + i4x4;
-                    if( decode_residual(h, gb, h->mb + 16*index, index, scan + 1, h->dequant4_coeff[chroma_idx+1+(IS_INTRA( mb_type ) ? 0:3)][chroma_qp], 15) < 0){
-                        return -1;
-                    }
-                }
-            }
-        }else{
-            uint8_t * const nnz= &h->non_zero_count_cache[0];
-            nnz[ scan8[16]+0 ] = nnz[ scan8[16]+1 ] =nnz[ scan8[16]+8 ] =nnz[ scan8[16]+9 ] =
-            nnz[ scan8[20]+0 ] = nnz[ scan8[20]+1 ] =nnz[ scan8[20]+8 ] =nnz[ scan8[20]+9 ] = 0;
-        }
-    }else{
-        uint8_t * const nnz= &h->non_zero_count_cache[0];
-        fill_rectangle(&nnz[scan8[0]], 4, 4, 8, 0, 1);
-        nnz[ scan8[16]+0 ] = nnz[ scan8[16]+1 ] =nnz[ scan8[16]+8 ] =nnz[ scan8[16]+9 ] =
-        nnz[ scan8[20]+0 ] = nnz[ scan8[20]+1 ] =nnz[ scan8[20]+8 ] =nnz[ scan8[20]+9 ] = 0;
-    }
-    s->current_picture.qscale_table[mb_xy]= s->qscale;
-    write_back_non_zero_count(h);
-
-    if(MB_MBAFF){
-        h->ref_count[0] >>= 1;
-        h->ref_count[1] >>= 1;
-    }
-
-    return 0;
-}
-
-static int decode_cabac_field_decoding_flag(H264Context *h) {
-    MpegEncContext * const s = &h->s;
-    const int mb_x = s->mb_x;
-    const int mb_y = s->mb_y & ~1;
-    const int mba_xy = mb_x - 1 +  mb_y   *s->mb_stride;
-    const int mbb_xy = mb_x     + (mb_y-2)*s->mb_stride;
-
-    unsigned int ctx = 0;
-
-    if( h->slice_table[mba_xy] == h->slice_num && IS_INTERLACED( s->current_picture.mb_type[mba_xy] ) ) {
-        ctx += 1;
-    }
-    if( h->slice_table[mbb_xy] == h->slice_num && IS_INTERLACED( s->current_picture.mb_type[mbb_xy] ) ) {
-        ctx += 1;
-    }
-
-    return get_cabac_noinline( &h->cabac, &h->cabac_state[70 + ctx] );
-}
-
-static int decode_cabac_intra_mb_type(H264Context *h, int ctx_base, int intra_slice) {
-    uint8_t *state= &h->cabac_state[ctx_base];
-    int mb_type;
-
-    if(intra_slice){
-        MpegEncContext * const s = &h->s;
-        const int mba_xy = h->left_mb_xy[0];
-        const int mbb_xy = h->top_mb_xy;
-        int ctx=0;
-        if( h->slice_table[mba_xy] == h->slice_num && !IS_INTRA4x4( s->current_picture.mb_type[mba_xy] ) )
-            ctx++;
-        if( h->slice_table[mbb_xy] == h->slice_num && !IS_INTRA4x4( s->current_picture.mb_type[mbb_xy] ) )
-            ctx++;
-        if( get_cabac_noinline( &h->cabac, &state[ctx] ) == 0 )
-            return 0;   /* I4x4 */
-        state += 2;
-    }else{
-        if( get_cabac_noinline( &h->cabac, &state[0] ) == 0 )
-            return 0;   /* I4x4 */
-    }
-
-    if( get_cabac_terminate( &h->cabac ) )
-        return 25;  /* PCM */
-
-    mb_type = 1; /* I16x16 */
-    mb_type += 12 * get_cabac_noinline( &h->cabac, &state[1] ); /* cbp_luma != 0 */
-    if( get_cabac_noinline( &h->cabac, &state[2] ) ) /* cbp_chroma */
-        mb_type += 4 + 4 * get_cabac_noinline( &h->cabac, &state[2+intra_slice] );
-    mb_type += 2 * get_cabac_noinline( &h->cabac, &state[3+intra_slice] );
-    mb_type += 1 * get_cabac_noinline( &h->cabac, &state[3+2*intra_slice] );
-    return mb_type;
-}
-
-static int decode_cabac_mb_type( H264Context *h ) {
-    MpegEncContext * const s = &h->s;
-
-    if( h->slice_type == I_TYPE ) {
-        return decode_cabac_intra_mb_type(h, 3, 1);
-    } else if( h->slice_type == P_TYPE ) {
-        if( get_cabac_noinline( &h->cabac, &h->cabac_state[14] ) == 0 ) {
-            /* P-type */
-            if( get_cabac_noinline( &h->cabac, &h->cabac_state[15] ) == 0 ) {
-                /* P_L0_D16x16, P_8x8 */
-                return 3 * get_cabac_noinline( &h->cabac, &h->cabac_state[16] );
-            } else {
-                /* P_L0_D8x16, P_L0_D16x8 */
-                return 2 - get_cabac_noinline( &h->cabac, &h->cabac_state[17] );
-            }
-        } else {
-            return decode_cabac_intra_mb_type(h, 17, 0) + 5;
-        }
-    } else if( h->slice_type == B_TYPE ) {
-        const int mba_xy = h->left_mb_xy[0];
-        const int mbb_xy = h->top_mb_xy;
-        int ctx = 0;
-        int bits;
-
-        if( h->slice_table[mba_xy] == h->slice_num && !IS_DIRECT( s->current_picture.mb_type[mba_xy] ) )
-            ctx++;
-        if( h->slice_table[mbb_xy] == h->slice_num && !IS_DIRECT( s->current_picture.mb_type[mbb_xy] ) )
-            ctx++;
-
-        if( !get_cabac_noinline( &h->cabac, &h->cabac_state[27+ctx] ) )
-            return 0; /* B_Direct_16x16 */
-
-        if( !get_cabac_noinline( &h->cabac, &h->cabac_state[27+3] ) ) {
-            return 1 + get_cabac_noinline( &h->cabac, &h->cabac_state[27+5] ); /* B_L[01]_16x16 */
-        }
-
-        bits = get_cabac_noinline( &h->cabac, &h->cabac_state[27+4] ) << 3;
-        bits|= get_cabac_noinline( &h->cabac, &h->cabac_state[27+5] ) << 2;
-        bits|= get_cabac_noinline( &h->cabac, &h->cabac_state[27+5] ) << 1;
-        bits|= get_cabac_noinline( &h->cabac, &h->cabac_state[27+5] );
-        if( bits < 8 )
-            return bits + 3; /* B_Bi_16x16 through B_L1_L0_16x8 */
-        else if( bits == 13 ) {
-            return decode_cabac_intra_mb_type(h, 32, 0) + 23;
-        } else if( bits == 14 )
-            return 11; /* B_L1_L0_8x16 */
-        else if( bits == 15 )
-            return 22; /* B_8x8 */
-
-        bits= ( bits<<1 ) | get_cabac_noinline( &h->cabac, &h->cabac_state[27+5] );
-        return bits - 4; /* B_L0_Bi_* through B_Bi_Bi_* */
-    } else {
-        /* TODO SI/SP frames? */
-        return -1;
-    }
-}
-
-static int decode_cabac_mb_skip( H264Context *h, int mb_x, int mb_y ) {
-    MpegEncContext * const s = &h->s;
-    int mba_xy, mbb_xy;
-    int ctx = 0;
-
-    if(FRAME_MBAFF){ //FIXME merge with the stuff in fill_caches?
-        int mb_xy = mb_x + (mb_y&~1)*s->mb_stride;
-        mba_xy = mb_xy - 1;
-        if( (mb_y&1)
-            && h->slice_table[mba_xy] == h->slice_num
-            && MB_FIELD == !!IS_INTERLACED( s->current_picture.mb_type[mba_xy] ) )
-            mba_xy += s->mb_stride;
-        if( MB_FIELD ){
-            mbb_xy = mb_xy - s->mb_stride;
-            if( !(mb_y&1)
-                && h->slice_table[mbb_xy] == h->slice_num
-                && IS_INTERLACED( s->current_picture.mb_type[mbb_xy] ) )
-                mbb_xy -= s->mb_stride;
-        }else
-            mbb_xy = mb_x + (mb_y-1)*s->mb_stride;
-    }else{
-        int mb_xy = mb_x + mb_y*s->mb_stride;
-        mba_xy = mb_xy - 1;
-        mbb_xy = mb_xy - s->mb_stride;
-    }
-
-    if( h->slice_table[mba_xy] == h->slice_num && !IS_SKIP( s->current_picture.mb_type[mba_xy] ))
-        ctx++;
-    if( h->slice_table[mbb_xy] == h->slice_num && !IS_SKIP( s->current_picture.mb_type[mbb_xy] ))
-        ctx++;
-
-    if( h->slice_type == B_TYPE )
-        ctx += 13;
-    return get_cabac_noinline( &h->cabac, &h->cabac_state[11+ctx] );
-}
-
-static int decode_cabac_mb_intra4x4_pred_mode( H264Context *h, int pred_mode ) {
-    int mode = 0;
-
-    if( get_cabac( &h->cabac, &h->cabac_state[68] ) )
-        return pred_mode;
-
-    mode += 1 * get_cabac( &h->cabac, &h->cabac_state[69] );
-    mode += 2 * get_cabac( &h->cabac, &h->cabac_state[69] );
-    mode += 4 * get_cabac( &h->cabac, &h->cabac_state[69] );
-
-    if( mode >= pred_mode )
-        return mode + 1;
-    else
-        return mode;
-}
-
-static int decode_cabac_mb_chroma_pre_mode( H264Context *h) {
-    const int mba_xy = h->left_mb_xy[0];
-    const int mbb_xy = h->top_mb_xy;
-
-    int ctx = 0;
-
-    /* No need to test for IS_INTRA4x4 and IS_INTRA16x16, as we set chroma_pred_mode_table to 0 */
-    if( h->slice_table[mba_xy] == h->slice_num && h->chroma_pred_mode_table[mba_xy] != 0 )
-        ctx++;
-
-    if( h->slice_table[mbb_xy] == h->slice_num && h->chroma_pred_mode_table[mbb_xy] != 0 )
-        ctx++;
-
-    if( get_cabac_noinline( &h->cabac, &h->cabac_state[64+ctx] ) == 0 )
-        return 0;
-
-    if( get_cabac_noinline( &h->cabac, &h->cabac_state[64+3] ) == 0 )
-        return 1;
-    if( get_cabac_noinline( &h->cabac, &h->cabac_state[64+3] ) == 0 )
-        return 2;
-    else
-        return 3;
-}
-
-static const uint8_t block_idx_x[16] = {
-    0, 1, 0, 1, 2, 3, 2, 3, 0, 1, 0, 1, 2, 3, 2, 3
-};
-static const uint8_t block_idx_y[16] = {
-    0, 0, 1, 1, 0, 0, 1, 1, 2, 2, 3, 3, 2, 2, 3, 3
-};
-static const uint8_t block_idx_xy[4][4] = {
-    { 0, 2, 8,  10},
-    { 1, 3, 9,  11},
-    { 4, 6, 12, 14},
-    { 5, 7, 13, 15}
-};
-
-static int decode_cabac_mb_cbp_luma( H264Context *h) {
-    int cbp = 0;
-    int cbp_b = -1;
-    int i8x8;
-
-    if( h->slice_table[h->top_mb_xy] == h->slice_num ) {
-        cbp_b = h->top_cbp;
-        tprintf("cbp_b = top_cbp = %x\n", cbp_b);
-    }
-
-    for( i8x8 = 0; i8x8 < 4; i8x8++ ) {
-        int cbp_a = -1;
-        int x, y;
-        int ctx = 0;
-
-        x = block_idx_x[4*i8x8];
-        y = block_idx_y[4*i8x8];
-
-        if( x > 0 )
-            cbp_a = cbp;
-        else if( h->slice_table[h->left_mb_xy[0]] == h->slice_num ) {
-            cbp_a = h->left_cbp;
-            tprintf("cbp_a = left_cbp = %x\n", cbp_a);
-        }
-
-        if( y > 0 )
-            cbp_b = cbp;
-
-        /* No need to test for skip as we put 0 for skip block */
-        /* No need to test for IPCM as we put 1 for IPCM block */
-        if( cbp_a >= 0 ) {
-            int i8x8a = block_idx_xy[(x-1)&0x03][y]/4;
-            if( ((cbp_a >> i8x8a)&0x01) == 0 )
-                ctx++;
-        }
-
-        if( cbp_b >= 0 ) {
-            int i8x8b = block_idx_xy[x][(y-1)&0x03]/4;
-            if( ((cbp_b >> i8x8b)&0x01) == 0 )
-                ctx += 2;
-        }
-
-        if( get_cabac( &h->cabac, &h->cabac_state[73 + ctx] ) ) {
-            cbp |= 1 << i8x8;
-        }
-    }
-    return cbp;
-}
-static int decode_cabac_mb_cbp_chroma( H264Context *h) {
-    int ctx;
-    int cbp_a, cbp_b;
-
-    cbp_a = (h->left_cbp>>4)&0x03;
-    cbp_b = (h-> top_cbp>>4)&0x03;
-
-    ctx = 0;
-    if( cbp_a > 0 ) ctx++;
-    if( cbp_b > 0 ) ctx += 2;
-    if( get_cabac_noinline( &h->cabac, &h->cabac_state[77 + ctx] ) == 0 )
-        return 0;
-
-    ctx = 4;
-    if( cbp_a == 2 ) ctx++;
-    if( cbp_b == 2 ) ctx += 2;
-    return 1 + get_cabac_noinline( &h->cabac, &h->cabac_state[77 + ctx] );
-}
-static int decode_cabac_mb_dqp( H264Context *h) {
-    MpegEncContext * const s = &h->s;
-    int mbn_xy;
-    int   ctx = 0;
-    int   val = 0;
-
-    if( s->mb_x > 0 )
-        mbn_xy = s->mb_x + s->mb_y*s->mb_stride - 1;
-    else
-        mbn_xy = s->mb_width - 1 + (s->mb_y-1)*s->mb_stride;
-
-    if( h->last_qscale_diff != 0 )
-        ctx++;
-
-    while( get_cabac_noinline( &h->cabac, &h->cabac_state[60 + ctx] ) ) {
-        if( ctx < 2 )
-            ctx = 2;
-        else
-            ctx = 3;
-        val++;
-        if(val > 102) //prevent infinite loop
-            return INT_MIN;
-    }
-
-    if( val&0x01 )
-        return (val + 1)/2;
-    else
-        return -(val + 1)/2;
-}
-static int decode_cabac_p_mb_sub_type( H264Context *h ) {
-    if( get_cabac( &h->cabac, &h->cabac_state[21] ) )
-        return 0;   /* 8x8 */
-    if( !get_cabac( &h->cabac, &h->cabac_state[22] ) )
-        return 1;   /* 8x4 */
-    if( get_cabac( &h->cabac, &h->cabac_state[23] ) )
-        return 2;   /* 4x8 */
-    return 3;       /* 4x4 */
-}
-static int decode_cabac_b_mb_sub_type( H264Context *h ) {
-    int type;
-    if( !get_cabac( &h->cabac, &h->cabac_state[36] ) )
-        return 0;   /* B_Direct_8x8 */
-    if( !get_cabac( &h->cabac, &h->cabac_state[37] ) )
-        return 1 + get_cabac( &h->cabac, &h->cabac_state[39] ); /* B_L0_8x8, B_L1_8x8 */
-    type = 3;
-    if( get_cabac( &h->cabac, &h->cabac_state[38] ) ) {
-        if( get_cabac( &h->cabac, &h->cabac_state[39] ) )
-            return 11 + get_cabac( &h->cabac, &h->cabac_state[39] ); /* B_L1_4x4, B_Bi_4x4 */
-        type += 4;
-    }
-    type += 2*get_cabac( &h->cabac, &h->cabac_state[39] );
-    type +=   get_cabac( &h->cabac, &h->cabac_state[39] );
-    return type;
-}
-
-static inline int decode_cabac_mb_transform_size( H264Context *h ) {
-    return get_cabac_noinline( &h->cabac, &h->cabac_state[399 + h->neighbor_transform_size] );
-}
-
-static int decode_cabac_mb_ref( H264Context *h, int list, int n ) {
-    int refa = h->ref_cache[list][scan8[n] - 1];
-    int refb = h->ref_cache[list][scan8[n] - 8];
-    int ref  = 0;
-    int ctx  = 0;
-
-    if( h->slice_type == B_TYPE) {
-        if( refa > 0 && !h->direct_cache[scan8[n] - 1] )
-            ctx++;
-        if( refb > 0 && !h->direct_cache[scan8[n] - 8] )
-            ctx += 2;
-    } else {
-        if( refa > 0 )
-            ctx++;
-        if( refb > 0 )
-            ctx += 2;
-    }
-
-    while( get_cabac( &h->cabac, &h->cabac_state[54+ctx] ) ) {
-        ref++;
-        if( ctx < 4 )
-            ctx = 4;
-        else
-            ctx = 5;
-    }
-    return ref;
-}
-
-static int decode_cabac_mb_mvd( H264Context *h, int list, int n, int l ) {
-    int amvd = abs( h->mvd_cache[list][scan8[n] - 1][l] ) +
-               abs( h->mvd_cache[list][scan8[n] - 8][l] );
-    int ctxbase = (l == 0) ? 40 : 47;
-    int ctx, mvd;
-
-    if( amvd < 3 )
-        ctx = 0;
-    else if( amvd > 32 )
-        ctx = 2;
-    else
-        ctx = 1;
-
-    if(!get_cabac(&h->cabac, &h->cabac_state[ctxbase+ctx]))
-        return 0;
-
-    mvd= 1;
-    ctx= 3;
-    while( mvd < 9 && get_cabac( &h->cabac, &h->cabac_state[ctxbase+ctx] ) ) {
-        mvd++;
-        if( ctx < 6 )
-            ctx++;
-    }
-
-    if( mvd >= 9 ) {
-        int k = 3;
-        while( get_cabac_bypass( &h->cabac ) ) {
-            mvd += 1 << k;
-            k++;
-        }
-        while( k-- ) {
-            if( get_cabac_bypass( &h->cabac ) )
-                mvd += 1 << k;
-        }
-    }
-    return get_cabac_bypass_sign( &h->cabac, -mvd );
-}
-
-static int inline get_cabac_cbf_ctx( H264Context *h, int cat, int idx ) {
-    int nza, nzb;
-    int ctx = 0;
-
-    if( cat == 0 ) {
-        nza = h->left_cbp&0x100;
-        nzb = h-> top_cbp&0x100;
-    } else if( cat == 1 || cat == 2 ) {
-        nza = h->non_zero_count_cache[scan8[idx] - 1];
-        nzb = h->non_zero_count_cache[scan8[idx] - 8];
-    } else if( cat == 3 ) {
-        nza = (h->left_cbp>>(6+idx))&0x01;
-        nzb = (h-> top_cbp>>(6+idx))&0x01;
-    } else {
-        assert(cat == 4);
-        nza = h->non_zero_count_cache[scan8[16+idx] - 1];
-        nzb = h->non_zero_count_cache[scan8[16+idx] - 8];
-    }
-
-    if( nza > 0 )
-        ctx++;
-
-    if( nzb > 0 )
-        ctx += 2;
-
-    return ctx + 4 * cat;
-}
-
-static const __attribute((used)) uint8_t last_coeff_flag_offset_8x8[63] = {
-    0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
-    2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
-    3, 3, 3, 3, 3, 3, 3, 3, 4, 4, 4, 4, 4, 4, 4, 4,
-    5, 5, 5, 5, 6, 6, 6, 6, 7, 7, 7, 7, 8, 8, 8
-};
-
-static int decode_cabac_residual( H264Context *h, DCTELEM *block, int cat, int n, const uint8_t *scantable, const uint32_t *qmul, int max_coeff) {
-    const int mb_xy  = h->s.mb_x + h->s.mb_y*h->s.mb_stride;
-    static const int significant_coeff_flag_offset[2][6] = {
-      { 105+0, 105+15, 105+29, 105+44, 105+47, 402 },
-      { 277+0, 277+15, 277+29, 277+44, 277+47, 436 }
-    };
-    static const int last_coeff_flag_offset[2][6] = {
-      { 166+0, 166+15, 166+29, 166+44, 166+47, 417 },
-      { 338+0, 338+15, 338+29, 338+44, 338+47, 451 }
-    };
-    static const int coeff_abs_level_m1_offset[6] = {
-        227+0, 227+10, 227+20, 227+30, 227+39, 426
-    };
-    static const uint8_t significant_coeff_flag_offset_8x8[2][63] = {
-      { 0, 1, 2, 3, 4, 5, 5, 4, 4, 3, 3, 4, 4, 4, 5, 5,
-        4, 4, 4, 4, 3, 3, 6, 7, 7, 7, 8, 9,10, 9, 8, 7,
-        7, 6,11,12,13,11, 6, 7, 8, 9,14,10, 9, 8, 6,11,
-       12,13,11, 6, 9,14,10, 9,11,12,13,11,14,10,12 },
-      { 0, 1, 1, 2, 2, 3, 3, 4, 5, 6, 7, 7, 7, 8, 4, 5,
-        6, 9,10,10, 8,11,12,11, 9, 9,10,10, 8,11,12,11,
-        9, 9,10,10, 8,11,12,11, 9, 9,10,10, 8,13,13, 9,
-        9,10,10, 8,13,13, 9, 9,10,10,14,14,14,14,14 }
-    };
-
-    int index[64];
-
-    int last;
-    int coeff_count = 0;
-
-    int abslevel1 = 1;
-    int abslevelgt1 = 0;
-
-    uint8_t *significant_coeff_ctx_base;
-    uint8_t *last_coeff_ctx_base;
-    uint8_t *abs_level_m1_ctx_base;
-
-#ifndef ARCH_X86
-#define CABAC_ON_STACK
-#endif
-#ifdef CABAC_ON_STACK
-#define CC &cc
-    CABACContext cc;
-    cc.range     = h->cabac.range;
-    cc.low       = h->cabac.low;
-    cc.bytestream= h->cabac.bytestream;
-#else
-#define CC &h->cabac
-#endif
-
-
-    /* cat: 0-> DC 16x16  n = 0
-     *      1-> AC 16x16  n = luma4x4idx
-     *      2-> Luma4x4   n = luma4x4idx
-     *      3-> DC Chroma n = iCbCr
-     *      4-> AC Chroma n = 4 * iCbCr + chroma4x4idx
-     *      5-> Luma8x8   n = 4 * luma8x8idx
-     */
-
-    /* read coded block flag */
-    if( cat != 5 ) {
-        if( get_cabac( CC, &h->cabac_state[85 + get_cabac_cbf_ctx( h, cat, n ) ] ) == 0 ) {
-            if( cat == 1 || cat == 2 )
-                h->non_zero_count_cache[scan8[n]] = 0;
-            else if( cat == 4 )
-                h->non_zero_count_cache[scan8[16+n]] = 0;
-#ifdef CABAC_ON_STACK
-            h->cabac.range     = cc.range     ;
-            h->cabac.low       = cc.low       ;
-            h->cabac.bytestream= cc.bytestream;
-#endif
-            return 0;
-        }
-    }
-
-    significant_coeff_ctx_base = h->cabac_state
-        + significant_coeff_flag_offset[MB_FIELD][cat];
-    last_coeff_ctx_base = h->cabac_state
-        + last_coeff_flag_offset[MB_FIELD][cat];
-    abs_level_m1_ctx_base = h->cabac_state
-        + coeff_abs_level_m1_offset[cat];
-
-    if( cat == 5 ) {
-#define DECODE_SIGNIFICANCE( coefs, sig_off, last_off ) \
-        for(last= 0; last < coefs; last++) { \
-            uint8_t *sig_ctx = significant_coeff_ctx_base + sig_off; \
-            if( get_cabac( CC, sig_ctx )) { \
-                uint8_t *last_ctx = last_coeff_ctx_base + last_off; \
-                index[coeff_count++] = last; \
-                if( get_cabac( CC, last_ctx ) ) { \
-                    last= max_coeff; \
-                    break; \
-                } \
-            } \
-        }\
-        if( last == max_coeff -1 ) {\
-            index[coeff_count++] = last;\
-        }
-        const uint8_t *sig_off = significant_coeff_flag_offset_8x8[MB_FIELD];
-#ifdef ARCH_X86
-        coeff_count= decode_significance_8x8_x86(CC, significant_coeff_ctx_base, index, sig_off);
-    } else {
-        coeff_count= decode_significance_x86(CC, max_coeff, significant_coeff_ctx_base, index);
-#else
-        DECODE_SIGNIFICANCE( 63, sig_off[last], last_coeff_flag_offset_8x8[last] );
-    } else {
-        DECODE_SIGNIFICANCE( max_coeff - 1, last, last );
-#endif
-    }
-    assert(coeff_count > 0);
-
-    if( cat == 0 )
-        h->cbp_table[mb_xy] |= 0x100;
-    else if( cat == 1 || cat == 2 )
-        h->non_zero_count_cache[scan8[n]] = coeff_count;
-    else if( cat == 3 )
-        h->cbp_table[mb_xy] |= 0x40 << n;
-    else if( cat == 4 )
-        h->non_zero_count_cache[scan8[16+n]] = coeff_count;
-    else {
-        assert( cat == 5 );
-        fill_rectangle(&h->non_zero_count_cache[scan8[n]], 2, 2, 8, coeff_count, 1);
-    }
-
-    for( coeff_count--; coeff_count >= 0; coeff_count-- ) {
-        uint8_t *ctx = (abslevelgt1 != 0 ? 0 : FFMIN( 4, abslevel1 )) + abs_level_m1_ctx_base;
-        int j= scantable[index[coeff_count]];
-
-        if( get_cabac( CC, ctx ) == 0 ) {
-            if( !qmul ) {
-                block[j] = get_cabac_bypass_sign( CC, -1);
-            }else{
-                block[j] = (get_cabac_bypass_sign( CC, -qmul[j]) + 32) >> 6;;
-            }
-
-            abslevel1++;
-        } else {
-            int coeff_abs = 2;
-            ctx = 5 + FFMIN( 4, abslevelgt1 ) + abs_level_m1_ctx_base;
-            while( coeff_abs < 15 && get_cabac( CC, ctx ) ) {
-                coeff_abs++;
-            }
-
-            if( coeff_abs >= 15 ) {
-                int j = 0;
-                while( get_cabac_bypass( CC ) ) {
-                    j++;
-                }
-
-                coeff_abs=1;
-                while( j-- ) {
-                    coeff_abs += coeff_abs + get_cabac_bypass( CC );
-                }
-                coeff_abs+= 14;
-            }
-
-            if( !qmul ) {
-                if( get_cabac_bypass( CC ) ) block[j] = -coeff_abs;
-                else                                block[j] =  coeff_abs;
-            }else{
-                if( get_cabac_bypass( CC ) ) block[j] = (-coeff_abs * qmul[j] + 32) >> 6;
-                else                                block[j] = ( coeff_abs * qmul[j] + 32) >> 6;
-            }
-
-            abslevelgt1++;
-        }
-    }
-#ifdef CABAC_ON_STACK
-            h->cabac.range     = cc.range     ;
-            h->cabac.low       = cc.low       ;
-            h->cabac.bytestream= cc.bytestream;
-#endif
-    return 0;
-}
-
-static void inline compute_mb_neighbors(H264Context *h)
-{
-    MpegEncContext * const s = &h->s;
-    const int mb_xy  = s->mb_x + s->mb_y*s->mb_stride;
-    h->top_mb_xy     = mb_xy - s->mb_stride;
-    h->left_mb_xy[0] = mb_xy - 1;
-    if(FRAME_MBAFF){
-        const int pair_xy          = s->mb_x     + (s->mb_y & ~1)*s->mb_stride;
-        const int top_pair_xy      = pair_xy     - s->mb_stride;
-        const int top_mb_frame_flag      = !IS_INTERLACED(s->current_picture.mb_type[top_pair_xy]);
-        const int left_mb_frame_flag = !IS_INTERLACED(s->current_picture.mb_type[pair_xy-1]);
-        const int curr_mb_frame_flag = !MB_FIELD;
-        const int bottom = (s->mb_y & 1);
-        if (bottom
-                ? !curr_mb_frame_flag // bottom macroblock
-                : (!curr_mb_frame_flag && !top_mb_frame_flag) // top macroblock
-                ) {
-            h->top_mb_xy -= s->mb_stride;
-        }
-        if (left_mb_frame_flag != curr_mb_frame_flag) {
-            h->left_mb_xy[0] = pair_xy - 1;
-        }
-    }
-    return;
-}
-
-/**
- * decodes a macroblock
- * @returns 0 if ok, AC_ERROR / DC_ERROR / MV_ERROR if an error is noticed
- */
-static int decode_mb_cabac(H264Context *h) {
-    MpegEncContext * const s = &h->s;
-    const int mb_xy= s->mb_x + s->mb_y*s->mb_stride;
-    int mb_type, partition_count, cbp = 0;
-    int dct8x8_allowed= h->pps.transform_8x8_mode;
-
-    s->dsp.clear_blocks(h->mb); //FIXME avoid if already clear (move after skip handlong?)
-
-    tprintf("pic:%d mb:%d/%d\n", h->frame_num, s->mb_x, s->mb_y);
-    if( h->slice_type != I_TYPE && h->slice_type != SI_TYPE ) {
-        int skip;
-        /* a skipped mb needs the aff flag from the following mb */
-        if( FRAME_MBAFF && s->mb_x==0 && (s->mb_y&1)==0 )
-            predict_field_decoding_flag(h);
-        if( FRAME_MBAFF && (s->mb_y&1)==1 && h->prev_mb_skipped )
-            skip = h->next_mb_skipped;
-        else
-            skip = decode_cabac_mb_skip( h, s->mb_x, s->mb_y );
-        /* read skip flags */
-        if( skip ) {
-            if( FRAME_MBAFF && (s->mb_y&1)==0 ){
-                s->current_picture.mb_type[mb_xy] = MB_TYPE_SKIP;
-                h->next_mb_skipped = decode_cabac_mb_skip( h, s->mb_x, s->mb_y+1 );
-                if(h->next_mb_skipped)
-                    predict_field_decoding_flag(h);
-                else
-                    h->mb_mbaff = h->mb_field_decoding_flag = decode_cabac_field_decoding_flag(h);
-            }
-
-            decode_mb_skip(h);
-
-            h->cbp_table[mb_xy] = 0;
-            h->chroma_pred_mode_table[mb_xy] = 0;
-            h->last_qscale_diff = 0;
-
-            return 0;
-
-        }
-    }
-    if(FRAME_MBAFF){
-        if( (s->mb_y&1) == 0 )
-            h->mb_mbaff =
-            h->mb_field_decoding_flag = decode_cabac_field_decoding_flag(h);
-    }else
-        h->mb_field_decoding_flag= (s->picture_structure!=PICT_FRAME);
-
-    h->prev_mb_skipped = 0;
-
-    compute_mb_neighbors(h);
-    if( ( mb_type = decode_cabac_mb_type( h ) ) < 0 ) {
-        av_log( h->s.avctx, AV_LOG_ERROR, "decode_cabac_mb_type failed\n" );
-        return -1;
-    }
-
-    if( h->slice_type == B_TYPE ) {
-        if( mb_type < 23 ){
-            partition_count= b_mb_type_info[mb_type].partition_count;
-            mb_type=         b_mb_type_info[mb_type].type;
-        }else{
-            mb_type -= 23;
-            goto decode_intra_mb;
-        }
-    } else if( h->slice_type == P_TYPE ) {
-        if( mb_type < 5) {
-            partition_count= p_mb_type_info[mb_type].partition_count;
-            mb_type=         p_mb_type_info[mb_type].type;
-        } else {
-            mb_type -= 5;
-            goto decode_intra_mb;
-        }
-    } else {
-       assert(h->slice_type == I_TYPE);
-decode_intra_mb:
-        partition_count = 0;
-        cbp= i_mb_type_info[mb_type].cbp;
-        h->intra16x16_pred_mode= i_mb_type_info[mb_type].pred_mode;
-        mb_type= i_mb_type_info[mb_type].type;
-    }
-    if(MB_FIELD)
-        mb_type |= MB_TYPE_INTERLACED;
-
-    h->slice_table[ mb_xy ]= h->slice_num;
-
-    if(IS_INTRA_PCM(mb_type)) {
-        const uint8_t *ptr;
-        unsigned int x, y;
-
-        // We assume these blocks are very rare so we dont optimize it.
-        // FIXME The two following lines get the bitstream position in the cabac
-        // decode, I think it should be done by a function in cabac.h (or cabac.c).
-        ptr= h->cabac.bytestream;
-        if (h->cabac.low&0x1) ptr-=CABAC_BITS/8;
-
-        // The pixels are stored in the same order as levels in h->mb array.
-        for(y=0; y<16; y++){
-            const int index= 4*(y&3) + 32*((y>>2)&1) + 128*(y>>3);
-            for(x=0; x<16; x++){
-                tprintf("LUMA ICPM LEVEL (%3d)\n", *ptr);
-                h->mb[index + (x&3) + 16*((x>>2)&1) + 64*(x>>3)]= *ptr++;
-            }
-        }
-        for(y=0; y<8; y++){
-            const int index= 256 + 4*(y&3) + 32*(y>>2);
-            for(x=0; x<8; x++){
-                tprintf("CHROMA U ICPM LEVEL (%3d)\n", *ptr);
-                h->mb[index + (x&3) + 16*(x>>2)]= *ptr++;
-            }
-        }
-        for(y=0; y<8; y++){
-            const int index= 256 + 64 + 4*(y&3) + 32*(y>>2);
-            for(x=0; x<8; x++){
-                tprintf("CHROMA V ICPM LEVEL (%3d)\n", *ptr);
-                h->mb[index + (x&3) + 16*(x>>2)]= *ptr++;
-            }
-        }
-
-        ff_init_cabac_decoder(&h->cabac, ptr, h->cabac.bytestream_end - ptr);
-
-        // All blocks are present
-        h->cbp_table[mb_xy] = 0x1ef;
-        h->chroma_pred_mode_table[mb_xy] = 0;
-        // In deblocking, the quantizer is 0
-        s->current_picture.qscale_table[mb_xy]= 0;
-        h->chroma_qp = get_chroma_qp(h->pps.chroma_qp_index_offset, 0);
-        // All coeffs are present
-        memset(h->non_zero_count[mb_xy], 16, 16);
-        s->current_picture.mb_type[mb_xy]= mb_type;
-        return 0;
-    }
-
-    if(MB_MBAFF){
-        h->ref_count[0] <<= 1;
-        h->ref_count[1] <<= 1;
-    }
-
-    fill_caches(h, mb_type, 0);
-
-    if( IS_INTRA( mb_type ) ) {
-        int i;
-        if( IS_INTRA4x4( mb_type ) ) {
-            if( dct8x8_allowed && decode_cabac_mb_transform_size( h ) ) {
-                mb_type |= MB_TYPE_8x8DCT;
-                for( i = 0; i < 16; i+=4 ) {
-                    int pred = pred_intra_mode( h, i );
-                    int mode = decode_cabac_mb_intra4x4_pred_mode( h, pred );
-                    fill_rectangle( &h->intra4x4_pred_mode_cache[ scan8[i] ], 2, 2, 8, mode, 1 );
-                }
-            } else {
-                for( i = 0; i < 16; i++ ) {
-                    int pred = pred_intra_mode( h, i );
-                    h->intra4x4_pred_mode_cache[ scan8[i] ] = decode_cabac_mb_intra4x4_pred_mode( h, pred );
-
-                //av_log( s->avctx, AV_LOG_ERROR, "i4x4 pred=%d mode=%d\n", pred, h->intra4x4_pred_mode_cache[ scan8[i] ] );
-                }
-            }
-            write_back_intra_pred_mode(h);
-            if( check_intra4x4_pred_mode(h) < 0 ) return -1;
-        } else {
-            h->intra16x16_pred_mode= check_intra_pred_mode( h, h->intra16x16_pred_mode );
-            if( h->intra16x16_pred_mode < 0 ) return -1;
-        }
-        h->chroma_pred_mode_table[mb_xy] =
-            h->chroma_pred_mode          = decode_cabac_mb_chroma_pre_mode( h );
-
-        h->chroma_pred_mode= check_intra_pred_mode( h, h->chroma_pred_mode );
-        if( h->chroma_pred_mode < 0 ) return -1;
-    } else if( partition_count == 4 ) {
-        int i, j, sub_partition_count[4], list, ref[2][4];
-
-        if( h->slice_type == B_TYPE ) {
-            for( i = 0; i < 4; i++ ) {
-                h->sub_mb_type[i] = decode_cabac_b_mb_sub_type( h );
-                sub_partition_count[i]= b_sub_mb_type_info[ h->sub_mb_type[i] ].partition_count;
-                h->sub_mb_type[i]=      b_sub_mb_type_info[ h->sub_mb_type[i] ].type;
-            }
-            if( IS_DIRECT(h->sub_mb_type[0] | h->sub_mb_type[1] |
-                          h->sub_mb_type[2] | h->sub_mb_type[3]) ) {
-                pred_direct_motion(h, &mb_type);
-                if( h->ref_count[0] > 1 || h->ref_count[1] > 1 ) {
-                    for( i = 0; i < 4; i++ )
-                        if( IS_DIRECT(h->sub_mb_type[i]) )
-                            fill_rectangle( &h->direct_cache[scan8[4*i]], 2, 2, 8, 1, 1 );
-                }
-            }
-        } else {
-            for( i = 0; i < 4; i++ ) {
-                h->sub_mb_type[i] = decode_cabac_p_mb_sub_type( h );
-                sub_partition_count[i]= p_sub_mb_type_info[ h->sub_mb_type[i] ].partition_count;
-                h->sub_mb_type[i]=      p_sub_mb_type_info[ h->sub_mb_type[i] ].type;
-            }
-        }
-
-        for( list = 0; list < 2; list++ ) {
-            if( h->ref_count[list] > 0 ) {
-                for( i = 0; i < 4; i++ ) {
-                    if(IS_DIRECT(h->sub_mb_type[i])) continue;
-                    if(IS_DIR(h->sub_mb_type[i], 0, list)){
-                        if( h->ref_count[list] > 1 )
-                            ref[list][i] = decode_cabac_mb_ref( h, list, 4*i );
-                        else
-                            ref[list][i] = 0;
-                    } else {
-                        ref[list][i] = -1;
-                    }
-                                                       h->ref_cache[list][ scan8[4*i]+1 ]=
-                    h->ref_cache[list][ scan8[4*i]+8 ]=h->ref_cache[list][ scan8[4*i]+9 ]= ref[list][i];
-                }
-            }
-        }
-
-        if(dct8x8_allowed)
-            dct8x8_allowed = get_dct8x8_allowed(h);
-
-        for(list=0; list<2; list++){
-            for(i=0; i<4; i++){
-                if(IS_DIRECT(h->sub_mb_type[i])){
-                    fill_rectangle(h->mvd_cache[list][scan8[4*i]], 2, 2, 8, 0, 4);
-                    continue;
-                }
-                h->ref_cache[list][ scan8[4*i]   ]=h->ref_cache[list][ scan8[4*i]+1 ];
-
-                if(IS_DIR(h->sub_mb_type[i], 0, list) && !IS_DIRECT(h->sub_mb_type[i])){
-                    const int sub_mb_type= h->sub_mb_type[i];
-                    const int block_width= (sub_mb_type & (MB_TYPE_16x16|MB_TYPE_16x8)) ? 2 : 1;
-                    for(j=0; j<sub_partition_count[i]; j++){
-                        int mpx, mpy;
-                        int mx, my;
-                        const int index= 4*i + block_width*j;
-                        int16_t (* mv_cache)[2]= &h->mv_cache[list][ scan8[index] ];
-                        int16_t (* mvd_cache)[2]= &h->mvd_cache[list][ scan8[index] ];
-                        pred_motion(h, index, block_width, list, h->ref_cache[list][ scan8[index] ], &mpx, &mpy);
-
-                        mx = mpx + decode_cabac_mb_mvd( h, list, index, 0 );
-                        my = mpy + decode_cabac_mb_mvd( h, list, index, 1 );
-                        tprintf("final mv:%d %d\n", mx, my);
-
-                        if(IS_SUB_8X8(sub_mb_type)){
-                            mv_cache[ 0 ][0]= mv_cache[ 1 ][0]=
-                            mv_cache[ 8 ][0]= mv_cache[ 9 ][0]= mx;
-                            mv_cache[ 0 ][1]= mv_cache[ 1 ][1]=
-                            mv_cache[ 8 ][1]= mv_cache[ 9 ][1]= my;
-
-                            mvd_cache[ 0 ][0]= mvd_cache[ 1 ][0]=
-                            mvd_cache[ 8 ][0]= mvd_cache[ 9 ][0]= mx - mpx;
-                            mvd_cache[ 0 ][1]= mvd_cache[ 1 ][1]=
-                            mvd_cache[ 8 ][1]= mvd_cache[ 9 ][1]= my - mpy;
-                        }else if(IS_SUB_8X4(sub_mb_type)){
-                            mv_cache[ 0 ][0]= mv_cache[ 1 ][0]= mx;
-                            mv_cache[ 0 ][1]= mv_cache[ 1 ][1]= my;
-
-                            mvd_cache[ 0 ][0]= mvd_cache[ 1 ][0]= mx- mpx;
-                            mvd_cache[ 0 ][1]= mvd_cache[ 1 ][1]= my - mpy;
-                        }else if(IS_SUB_4X8(sub_mb_type)){
-                            mv_cache[ 0 ][0]= mv_cache[ 8 ][0]= mx;
-                            mv_cache[ 0 ][1]= mv_cache[ 8 ][1]= my;
-
-                            mvd_cache[ 0 ][0]= mvd_cache[ 8 ][0]= mx - mpx;
-                            mvd_cache[ 0 ][1]= mvd_cache[ 8 ][1]= my - mpy;
-                        }else{
-                            assert(IS_SUB_4X4(sub_mb_type));
-                            mv_cache[ 0 ][0]= mx;
-                            mv_cache[ 0 ][1]= my;
-
-                            mvd_cache[ 0 ][0]= mx - mpx;
-                            mvd_cache[ 0 ][1]= my - mpy;
-                        }
-                    }
-                }else{
-                    uint32_t *p= (uint32_t *)&h->mv_cache[list][ scan8[4*i] ][0];
-                    uint32_t *pd= (uint32_t *)&h->mvd_cache[list][ scan8[4*i] ][0];
-                    p[0] = p[1] = p[8] = p[9] = 0;
-                    pd[0]= pd[1]= pd[8]= pd[9]= 0;
-                }
-            }
-        }
-    } else if( IS_DIRECT(mb_type) ) {
-        pred_direct_motion(h, &mb_type);
-        fill_rectangle(h->mvd_cache[0][scan8[0]], 4, 4, 8, 0, 4);
-        fill_rectangle(h->mvd_cache[1][scan8[0]], 4, 4, 8, 0, 4);
-        dct8x8_allowed &= h->sps.direct_8x8_inference_flag;
-    } else {
-        int list, mx, my, i, mpx, mpy;
-        if(IS_16X16(mb_type)){
-            for(list=0; list<2; list++){
-                if(IS_DIR(mb_type, 0, list)){
-                    if(h->ref_count[list] > 0 ){
-                        const int ref = h->ref_count[list] > 1 ? decode_cabac_mb_ref( h, list, 0 ) : 0;
-                        fill_rectangle(&h->ref_cache[list][ scan8[0] ], 4, 4, 8, ref, 1);
-                    }
-                }else
-                    fill_rectangle(&h->ref_cache[list][ scan8[0] ], 4, 4, 8, (uint8_t)LIST_NOT_USED, 1);
-            }
-            for(list=0; list<2; list++){
-                if(IS_DIR(mb_type, 0, list)){
-                    pred_motion(h, 0, 4, list, h->ref_cache[list][ scan8[0] ], &mpx, &mpy);
-
-                    mx = mpx + decode_cabac_mb_mvd( h, list, 0, 0 );
-                    my = mpy + decode_cabac_mb_mvd( h, list, 0, 1 );
-                    tprintf("final mv:%d %d\n", mx, my);
-
-                    fill_rectangle(h->mvd_cache[list][ scan8[0] ], 4, 4, 8, pack16to32(mx-mpx,my-mpy), 4);
-                    fill_rectangle(h->mv_cache[list][ scan8[0] ], 4, 4, 8, pack16to32(mx,my), 4);
-                }else
-                    fill_rectangle(h->mv_cache[list][ scan8[0] ], 4, 4, 8, 0, 4);
-            }
-        }
-        else if(IS_16X8(mb_type)){
-            for(list=0; list<2; list++){
-                if(h->ref_count[list]>0){
-                    for(i=0; i<2; i++){
-                        if(IS_DIR(mb_type, i, list)){
-                            const int ref= h->ref_count[list] > 1 ? decode_cabac_mb_ref( h, list, 8*i ) : 0;
-                            fill_rectangle(&h->ref_cache[list][ scan8[0] + 16*i ], 4, 2, 8, ref, 1);
-                        }else
-                            fill_rectangle(&h->ref_cache[list][ scan8[0] + 16*i ], 4, 2, 8, (LIST_NOT_USED&0xFF), 1);
-                    }
-                }
-            }
-            for(list=0; list<2; list++){
-                for(i=0; i<2; i++){
-                    if(IS_DIR(mb_type, i, list)){
-                        pred_16x8_motion(h, 8*i, list, h->ref_cache[list][scan8[0] + 16*i], &mpx, &mpy);
-                        mx = mpx + decode_cabac_mb_mvd( h, list, 8*i, 0 );
-                        my = mpy + decode_cabac_mb_mvd( h, list, 8*i, 1 );
-                        tprintf("final mv:%d %d\n", mx, my);
-
-                        fill_rectangle(h->mvd_cache[list][ scan8[0] + 16*i ], 4, 2, 8, pack16to32(mx-mpx,my-mpy), 4);
-                        fill_rectangle(h->mv_cache[list][ scan8[0] + 16*i ], 4, 2, 8, pack16to32(mx,my), 4);
-                    }else{
-                        fill_rectangle(h->mvd_cache[list][ scan8[0] + 16*i ], 4, 2, 8, 0, 4);
-                        fill_rectangle(h-> mv_cache[list][ scan8[0] + 16*i ], 4, 2, 8, 0, 4);
-                    }
-                }
-            }
-        }else{
-            assert(IS_8X16(mb_type));
-            for(list=0; list<2; list++){
-                if(h->ref_count[list]>0){
-                    for(i=0; i<2; i++){
-                        if(IS_DIR(mb_type, i, list)){ //FIXME optimize
-                            const int ref= h->ref_count[list] > 1 ? decode_cabac_mb_ref( h, list, 4*i ) : 0;
-                            fill_rectangle(&h->ref_cache[list][ scan8[0] + 2*i ], 2, 4, 8, ref, 1);
-                        }else
-                            fill_rectangle(&h->ref_cache[list][ scan8[0] + 2*i ], 2, 4, 8, (LIST_NOT_USED&0xFF), 1);
-                    }
-                }
-            }
-            for(list=0; list<2; list++){
-                for(i=0; i<2; i++){
-                    if(IS_DIR(mb_type, i, list)){
-                        pred_8x16_motion(h, i*4, list, h->ref_cache[list][ scan8[0] + 2*i ], &mpx, &mpy);
-                        mx = mpx + decode_cabac_mb_mvd( h, list, 4*i, 0 );
-                        my = mpy + decode_cabac_mb_mvd( h, list, 4*i, 1 );
-
-                        tprintf("final mv:%d %d\n", mx, my);
-                        fill_rectangle(h->mvd_cache[list][ scan8[0] + 2*i ], 2, 4, 8, pack16to32(mx-mpx,my-mpy), 4);
-                        fill_rectangle(h->mv_cache[list][ scan8[0] + 2*i ], 2, 4, 8, pack16to32(mx,my), 4);
-                    }else{
-                        fill_rectangle(h->mvd_cache[list][ scan8[0] + 2*i ], 2, 4, 8, 0, 4);
-                        fill_rectangle(h-> mv_cache[list][ scan8[0] + 2*i ], 2, 4, 8, 0, 4);
-                    }
-                }
-            }
-        }
-    }
-
-   if( IS_INTER( mb_type ) ) {
-        h->chroma_pred_mode_table[mb_xy] = 0;
-        write_back_motion( h, mb_type );
-   }
-
-    if( !IS_INTRA16x16( mb_type ) ) {
-        cbp  = decode_cabac_mb_cbp_luma( h );
-        cbp |= decode_cabac_mb_cbp_chroma( h ) << 4;
-    }
-
-    h->cbp_table[mb_xy] = h->cbp = cbp;
-
-    if( dct8x8_allowed && (cbp&15) && !IS_INTRA( mb_type ) ) {
-        if( decode_cabac_mb_transform_size( h ) )
-            mb_type |= MB_TYPE_8x8DCT;
-    }
-    s->current_picture.mb_type[mb_xy]= mb_type;
-
-    if( cbp || IS_INTRA16x16( mb_type ) ) {
-        const uint8_t *scan, *scan8x8, *dc_scan;
-        int dqp;
-
-        if(IS_INTERLACED(mb_type)){
-            scan8x8= s->qscale ? h->field_scan8x8 : h->field_scan8x8_q0;
-            scan= s->qscale ? h->field_scan : h->field_scan_q0;
-            dc_scan= luma_dc_field_scan;
-        }else{
-            scan8x8= s->qscale ? h->zigzag_scan8x8 : h->zigzag_scan8x8_q0;
-            scan= s->qscale ? h->zigzag_scan : h->zigzag_scan_q0;
-            dc_scan= luma_dc_zigzag_scan;
-        }
-
-        h->last_qscale_diff = dqp = decode_cabac_mb_dqp( h );
-        if( dqp == INT_MIN ){
-            av_log(h->s.avctx, AV_LOG_ERROR, "cabac decode of qscale diff failed at %d %d\n", s->mb_x, s->mb_y);
-            return -1;
-        }
-        s->qscale += dqp;
-        if(((unsigned)s->qscale) > 51){
-            if(s->qscale<0) s->qscale+= 52;
-            else            s->qscale-= 52;
-        }
-        h->chroma_qp = get_chroma_qp(h->pps.chroma_qp_index_offset, s->qscale);
-
-        if( IS_INTRA16x16( mb_type ) ) {
-            int i;
-            //av_log( s->avctx, AV_LOG_ERROR, "INTRA16x16 DC\n" );
-            if( decode_cabac_residual( h, h->mb, 0, 0, dc_scan, NULL, 16) < 0)
-                return -1;
-            if( cbp&15 ) {
-                for( i = 0; i < 16; i++ ) {
-                    //av_log( s->avctx, AV_LOG_ERROR, "INTRA16x16 AC:%d\n", i );
-                    if( decode_cabac_residual(h, h->mb + 16*i, 1, i, scan + 1, h->dequant4_coeff[0][s->qscale], 15) < 0 )
-                        return -1;
-                }
-            } else {
-                fill_rectangle(&h->non_zero_count_cache[scan8[0]], 4, 4, 8, 0, 1);
-            }
-        } else {
-            int i8x8, i4x4;
-            for( i8x8 = 0; i8x8 < 4; i8x8++ ) {
-                if( cbp & (1<<i8x8) ) {
-                    if( IS_8x8DCT(mb_type) ) {
-                        if( decode_cabac_residual(h, h->mb + 64*i8x8, 5, 4*i8x8,
-                            scan8x8, h->dequant8_coeff[IS_INTRA( mb_type ) ? 0:1][s->qscale], 64) < 0 )
-                            return -1;
-                    } else
-                    for( i4x4 = 0; i4x4 < 4; i4x4++ ) {
-                        const int index = 4*i8x8 + i4x4;
-                        //av_log( s->avctx, AV_LOG_ERROR, "Luma4x4: %d\n", index );
-//START_TIMER
-                        if( decode_cabac_residual(h, h->mb + 16*index, 2, index, scan, h->dequant4_coeff[IS_INTRA( mb_type ) ? 0:3][s->qscale], 16) < 0 )
-                            return -1;
-//STOP_TIMER("decode_residual")
-                    }
-                } else {
-                    uint8_t * const nnz= &h->non_zero_count_cache[ scan8[4*i8x8] ];
-                    nnz[0] = nnz[1] = nnz[8] = nnz[9] = 0;
-                }
-            }
-        }
-
-        if( cbp&0x30 ){
-            int c;
-            for( c = 0; c < 2; c++ ) {
-                //av_log( s->avctx, AV_LOG_ERROR, "INTRA C%d-DC\n",c );
-                if( decode_cabac_residual(h, h->mb + 256 + 16*4*c, 3, c, chroma_dc_scan, NULL, 4) < 0)
-                    return -1;
-            }
-        }
-
-        if( cbp&0x20 ) {
-            int c, i;
-            for( c = 0; c < 2; c++ ) {
-                for( i = 0; i < 4; i++ ) {
-                    const int index = 16 + 4 * c + i;
-                    //av_log( s->avctx, AV_LOG_ERROR, "INTRA C%d-AC %d\n",c, index - 16 );
-                    if( decode_cabac_residual(h, h->mb + 16*index, 4, index - 16, scan + 1, h->dequant4_coeff[c+1+(IS_INTRA( mb_type ) ? 0:3)][h->chroma_qp], 15) < 0)
-                        return -1;
-                }
-            }
-        } else {
-            uint8_t * const nnz= &h->non_zero_count_cache[0];
-            nnz[ scan8[16]+0 ] = nnz[ scan8[16]+1 ] =nnz[ scan8[16]+8 ] =nnz[ scan8[16]+9 ] =
-            nnz[ scan8[20]+0 ] = nnz[ scan8[20]+1 ] =nnz[ scan8[20]+8 ] =nnz[ scan8[20]+9 ] = 0;
-        }
-    } else {
-        uint8_t * const nnz= &h->non_zero_count_cache[0];
-        fill_rectangle(&nnz[scan8[0]], 4, 4, 8, 0, 1);
-        nnz[ scan8[16]+0 ] = nnz[ scan8[16]+1 ] =nnz[ scan8[16]+8 ] =nnz[ scan8[16]+9 ] =
-        nnz[ scan8[20]+0 ] = nnz[ scan8[20]+1 ] =nnz[ scan8[20]+8 ] =nnz[ scan8[20]+9 ] = 0;
-        h->last_qscale_diff = 0;
-    }
-
-    s->current_picture.qscale_table[mb_xy]= s->qscale;
-    write_back_non_zero_count(h);
-
-    if(MB_MBAFF){
-        h->ref_count[0] >>= 1;
-        h->ref_count[1] >>= 1;
-    }
-
-    return 0;
-}
-
-
-static void filter_mb_edgev( H264Context *h, uint8_t *pix, int stride, int16_t bS[4], int qp ) {
-    int i, d;
-    const int index_a = clip( qp + h->slice_alpha_c0_offset, 0, 51 );
-    const int alpha = alpha_table[index_a];
-    const int beta  = beta_table[clip( qp + h->slice_beta_offset, 0, 51 )];
-
-    if( bS[0] < 4 ) {
-        int8_t tc[4];
-        for(i=0; i<4; i++)
-            tc[i] = bS[i] ? tc0_table[index_a][bS[i] - 1] : -1;
-        h->s.dsp.h264_h_loop_filter_luma(pix, stride, alpha, beta, tc);
-    } else {
-        /* 16px edge length, because bS=4 is triggered by being at
-         * the edge of an intra MB, so all 4 bS are the same */
-            for( d = 0; d < 16; d++ ) {
-                const int p0 = pix[-1];
-                const int p1 = pix[-2];
-                const int p2 = pix[-3];
-
-                const int q0 = pix[0];
-                const int q1 = pix[1];
-                const int q2 = pix[2];
-
-                if( FFABS( p0 - q0 ) < alpha &&
-                    FFABS( p1 - p0 ) < beta &&
-                    FFABS( q1 - q0 ) < beta ) {
-
-                    if(FFABS( p0 - q0 ) < (( alpha >> 2 ) + 2 )){
-                        if( FFABS( p2 - p0 ) < beta)
-                        {
-                            const int p3 = pix[-4];
-                            /* p0', p1', p2' */
-                            pix[-1] = ( p2 + 2*p1 + 2*p0 + 2*q0 + q1 + 4 ) >> 3;
-                            pix[-2] = ( p2 + p1 + p0 + q0 + 2 ) >> 2;
-                            pix[-3] = ( 2*p3 + 3*p2 + p1 + p0 + q0 + 4 ) >> 3;
-                        } else {
-                            /* p0' */
-                            pix[-1] = ( 2*p1 + p0 + q1 + 2 ) >> 2;
-                        }
-                        if( FFABS( q2 - q0 ) < beta)
-                        {
-                            const int q3 = pix[3];
-                            /* q0', q1', q2' */
-                            pix[0] = ( p1 + 2*p0 + 2*q0 + 2*q1 + q2 + 4 ) >> 3;
-                            pix[1] = ( p0 + q0 + q1 + q2 + 2 ) >> 2;
-                            pix[2] = ( 2*q3 + 3*q2 + q1 + q0 + p0 + 4 ) >> 3;
-                        } else {
-                            /* q0' */
-                            pix[0] = ( 2*q1 + q0 + p1 + 2 ) >> 2;
-                        }
-                    }else{
-                        /* p0', q0' */
-                        pix[-1] = ( 2*p1 + p0 + q1 + 2 ) >> 2;
-                        pix[ 0] = ( 2*q1 + q0 + p1 + 2 ) >> 2;
-                    }
-                    tprintf("filter_mb_edgev i:%d d:%d\n# bS:4 -> [%02x, %02x, %02x, %02x, %02x, %02x] =>[%02x, %02x, %02x, %02x]\n", i, d, p2, p1, p0, q0, q1, q2, pix[-2], pix[-1], pix[0], pix[1]);
-                }
-                pix += stride;
-            }
-    }
-}
-static void filter_mb_edgecv( H264Context *h, uint8_t *pix, int stride, int16_t bS[4], int qp ) {
-    int i;
-    const int index_a = clip( qp + h->slice_alpha_c0_offset, 0, 51 );
-    const int alpha = alpha_table[index_a];
-    const int beta  = beta_table[clip( qp + h->slice_beta_offset, 0, 51 )];
-
-    if( bS[0] < 4 ) {
-        int8_t tc[4];
-        for(i=0; i<4; i++)
-            tc[i] = bS[i] ? tc0_table[index_a][bS[i] - 1] + 1 : 0;
-        h->s.dsp.h264_h_loop_filter_chroma(pix, stride, alpha, beta, tc);
-    } else {
-        h->s.dsp.h264_h_loop_filter_chroma_intra(pix, stride, alpha, beta);
-    }
-}
-
-static void filter_mb_mbaff_edgev( H264Context *h, uint8_t *pix, int stride, int16_t bS[8], int qp[2] ) {
-    int i;
-    for( i = 0; i < 16; i++, pix += stride) {
-        int index_a;
-        int alpha;
-        int beta;
-
-        int qp_index;
-        int bS_index = (i >> 1);
-        if (!MB_FIELD) {
-            bS_index &= ~1;
-            bS_index |= (i & 1);
-        }
-
-        if( bS[bS_index] == 0 ) {
-            continue;
-        }
-
-        qp_index = MB_FIELD ? (i >> 3) : (i & 1);
-        index_a = clip( qp[qp_index] + h->slice_alpha_c0_offset, 0, 51 );
-        alpha = alpha_table[index_a];
-        beta  = beta_table[clip( qp[qp_index] + h->slice_beta_offset, 0, 51 )];
-
-        if( bS[bS_index] < 4 ) {
-            const int tc0 = tc0_table[index_a][bS[bS_index] - 1];
-            const int p0 = pix[-1];
-            const int p1 = pix[-2];
-            const int p2 = pix[-3];
-            const int q0 = pix[0];
-            const int q1 = pix[1];
-            const int q2 = pix[2];
-
-            if( FFABS( p0 - q0 ) < alpha &&
-                FFABS( p1 - p0 ) < beta &&
-                FFABS( q1 - q0 ) < beta ) {
-                int tc = tc0;
-                int i_delta;
-
-                if( FFABS( p2 - p0 ) < beta ) {
-                    pix[-2] = p1 + clip( ( p2 + ( ( p0 + q0 + 1 ) >> 1 ) - ( p1 << 1 ) ) >> 1, -tc0, tc0 );
-                    tc++;
-                }
-                if( FFABS( q2 - q0 ) < beta ) {
-                    pix[1] = q1 + clip( ( q2 + ( ( p0 + q0 + 1 ) >> 1 ) - ( q1 << 1 ) ) >> 1, -tc0, tc0 );
-                    tc++;
-                }
-
-                i_delta = clip( (((q0 - p0 ) << 2) + (p1 - q1) + 4) >> 3, -tc, tc );
-                pix[-1] = clip_uint8( p0 + i_delta );    /* p0' */
-                pix[0]  = clip_uint8( q0 - i_delta );    /* q0' */
-                tprintf("filter_mb_mbaff_edgev i:%d, qp:%d, indexA:%d, alpha:%d, beta:%d, tc:%d\n# bS:%d -> [%02x, %02x, %02x, %02x, %02x, %02x] =>[%02x, %02x, %02x, %02x]\n", i, qp[qp_index], index_a, alpha, beta, tc, bS[bS_index], pix[-3], p1, p0, q0, q1, pix[2], p1, pix[-1], pix[0], q1);
-            }
-        }else{
-            const int p0 = pix[-1];
-            const int p1 = pix[-2];
-            const int p2 = pix[-3];
-
-            const int q0 = pix[0];
-            const int q1 = pix[1];
-            const int q2 = pix[2];
-
-            if( FFABS( p0 - q0 ) < alpha &&
-                FFABS( p1 - p0 ) < beta &&
-                FFABS( q1 - q0 ) < beta ) {
-
-                if(FFABS( p0 - q0 ) < (( alpha >> 2 ) + 2 )){
-                    if( FFABS( p2 - p0 ) < beta)
-                    {
-                        const int p3 = pix[-4];
-                        /* p0', p1', p2' */
-                        pix[-1] = ( p2 + 2*p1 + 2*p0 + 2*q0 + q1 + 4 ) >> 3;
-                        pix[-2] = ( p2 + p1 + p0 + q0 + 2 ) >> 2;
-                        pix[-3] = ( 2*p3 + 3*p2 + p1 + p0 + q0 + 4 ) >> 3;
-                    } else {
-                        /* p0' */
-                        pix[-1] = ( 2*p1 + p0 + q1 + 2 ) >> 2;
-                    }
-                    if( FFABS( q2 - q0 ) < beta)
-                    {
-                        const int q3 = pix[3];
-                        /* q0', q1', q2' */
-                        pix[0] = ( p1 + 2*p0 + 2*q0 + 2*q1 + q2 + 4 ) >> 3;
-                        pix[1] = ( p0 + q0 + q1 + q2 + 2 ) >> 2;
-                        pix[2] = ( 2*q3 + 3*q2 + q1 + q0 + p0 + 4 ) >> 3;
-                    } else {
-                        /* q0' */
-                        pix[0] = ( 2*q1 + q0 + p1 + 2 ) >> 2;
-                    }
-                }else{
-                    /* p0', q0' */
-                    pix[-1] = ( 2*p1 + p0 + q1 + 2 ) >> 2;
-                    pix[ 0] = ( 2*q1 + q0 + p1 + 2 ) >> 2;
-                }
-                tprintf("filter_mb_mbaff_edgev i:%d, qp:%d, indexA:%d, alpha:%d, beta:%d\n# bS:4 -> [%02x, %02x, %02x, %02x, %02x, %02x] =>[%02x, %02x, %02x, %02x, %02x, %02x]\n", i, qp[qp_index], index_a, alpha, beta, p2, p1, p0, q0, q1, q2, pix[-3], pix[-2], pix[-1], pix[0], pix[1], pix[2]);
-            }
-        }
-    }
-}
-static void filter_mb_mbaff_edgecv( H264Context *h, uint8_t *pix, int stride, int16_t bS[8], int qp[2] ) {
-    int i;
-    for( i = 0; i < 8; i++, pix += stride) {
-        int index_a;
-        int alpha;
-        int beta;
-
-        int qp_index;
-        int bS_index = i;
-
-        if( bS[bS_index] == 0 ) {
-            continue;
-        }
-
-        qp_index = MB_FIELD ? (i >> 2) : (i & 1);
-        index_a = clip( qp[qp_index] + h->slice_alpha_c0_offset, 0, 51 );
-        alpha = alpha_table[index_a];
-        beta  = beta_table[clip( qp[qp_index] + h->slice_beta_offset, 0, 51 )];
-
-        if( bS[bS_index] < 4 ) {
-            const int tc = tc0_table[index_a][bS[bS_index] - 1] + 1;
-            const int p0 = pix[-1];
-            const int p1 = pix[-2];
-            const int q0 = pix[0];
-            const int q1 = pix[1];
-
-            if( FFABS( p0 - q0 ) < alpha &&
-                FFABS( p1 - p0 ) < beta &&
-                FFABS( q1 - q0 ) < beta ) {
-                const int i_delta = clip( (((q0 - p0 ) << 2) + (p1 - q1) + 4) >> 3, -tc, tc );
-
-                pix[-1] = clip_uint8( p0 + i_delta );    /* p0' */
-                pix[0]  = clip_uint8( q0 - i_delta );    /* q0' */
-                tprintf("filter_mb_mbaff_edgecv i:%d, qp:%d, indexA:%d, alpha:%d, beta:%d, tc:%d\n# bS:%d -> [%02x, %02x, %02x, %02x, %02x, %02x] =>[%02x, %02x, %02x, %02x]\n", i, qp[qp_index], index_a, alpha, beta, tc, bS[bS_index], pix[-3], p1, p0, q0, q1, pix[2], p1, pix[-1], pix[0], q1);
-            }
-        }else{
-            const int p0 = pix[-1];
-            const int p1 = pix[-2];
-            const int q0 = pix[0];
-            const int q1 = pix[1];
-
-            if( FFABS( p0 - q0 ) < alpha &&
-                FFABS( p1 - p0 ) < beta &&
-                FFABS( q1 - q0 ) < beta ) {
-
-                pix[-1] = ( 2*p1 + p0 + q1 + 2 ) >> 2;   /* p0' */
-                pix[0]  = ( 2*q1 + q0 + p1 + 2 ) >> 2;   /* q0' */
-                tprintf("filter_mb_mbaff_edgecv i:%d\n# bS:4 -> [%02x, %02x, %02x, %02x, %02x, %02x] =>[%02x, %02x, %02x, %02x, %02x, %02x]\n", i, pix[-3], p1, p0, q0, q1, pix[2], pix[-3], pix[-2], pix[-1], pix[0], pix[1], pix[2]);
-            }
-        }
-    }
-}
-
-static void filter_mb_edgeh( H264Context *h, uint8_t *pix, int stride, int16_t bS[4], int qp ) {
-    int i, d;
-    const int index_a = clip( qp + h->slice_alpha_c0_offset, 0, 51 );
-    const int alpha = alpha_table[index_a];
-    const int beta  = beta_table[clip( qp + h->slice_beta_offset, 0, 51 )];
-    const int pix_next  = stride;
-
-    if( bS[0] < 4 ) {
-        int8_t tc[4];
-        for(i=0; i<4; i++)
-            tc[i] = bS[i] ? tc0_table[index_a][bS[i] - 1] : -1;
-        h->s.dsp.h264_v_loop_filter_luma(pix, stride, alpha, beta, tc);
-    } else {
-        /* 16px edge length, see filter_mb_edgev */
-            for( d = 0; d < 16; d++ ) {
-                const int p0 = pix[-1*pix_next];
-                const int p1 = pix[-2*pix_next];
-                const int p2 = pix[-3*pix_next];
-                const int q0 = pix[0];
-                const int q1 = pix[1*pix_next];
-                const int q2 = pix[2*pix_next];
-
-                if( FFABS( p0 - q0 ) < alpha &&
-                    FFABS( p1 - p0 ) < beta &&
-                    FFABS( q1 - q0 ) < beta ) {
-
-                    const int p3 = pix[-4*pix_next];
-                    const int q3 = pix[ 3*pix_next];
-
-                    if(FFABS( p0 - q0 ) < (( alpha >> 2 ) + 2 )){
-                        if( FFABS( p2 - p0 ) < beta) {
-                            /* p0', p1', p2' */
-                            pix[-1*pix_next] = ( p2 + 2*p1 + 2*p0 + 2*q0 + q1 + 4 ) >> 3;
-                            pix[-2*pix_next] = ( p2 + p1 + p0 + q0 + 2 ) >> 2;
-                            pix[-3*pix_next] = ( 2*p3 + 3*p2 + p1 + p0 + q0 + 4 ) >> 3;
-                        } else {
-                            /* p0' */
-                            pix[-1*pix_next] = ( 2*p1 + p0 + q1 + 2 ) >> 2;
-                        }
-                        if( FFABS( q2 - q0 ) < beta) {
-                            /* q0', q1', q2' */
-                            pix[0*pix_next] = ( p1 + 2*p0 + 2*q0 + 2*q1 + q2 + 4 ) >> 3;
-                            pix[1*pix_next] = ( p0 + q0 + q1 + q2 + 2 ) >> 2;
-                            pix[2*pix_next] = ( 2*q3 + 3*q2 + q1 + q0 + p0 + 4 ) >> 3;
-                        } else {
-                            /* q0' */
-                            pix[0*pix_next] = ( 2*q1 + q0 + p1 + 2 ) >> 2;
-                        }
-                    }else{
-                        /* p0', q0' */
-                        pix[-1*pix_next] = ( 2*p1 + p0 + q1 + 2 ) >> 2;
-                        pix[ 0*pix_next] = ( 2*q1 + q0 + p1 + 2 ) >> 2;
-                    }
-                    tprintf("filter_mb_edgeh i:%d d:%d, qp:%d, indexA:%d, alpha:%d, beta:%d\n# bS:%d -> [%02x, %02x, %02x, %02x, %02x, %02x] =>[%02x, %02x, %02x, %02x]\n", i, d, qp, index_a, alpha, beta, bS[i], p2, p1, p0, q0, q1, q2, pix[-2*pix_next], pix[-pix_next], pix[0], pix[pix_next]);
-                }
-                pix++;
-            }
-    }
-}
-
-static void filter_mb_edgech( H264Context *h, uint8_t *pix, int stride, int16_t bS[4], int qp ) {
-    int i;
-    const int index_a = clip( qp + h->slice_alpha_c0_offset, 0, 51 );
-    const int alpha = alpha_table[index_a];
-    const int beta  = beta_table[clip( qp + h->slice_beta_offset, 0, 51 )];
-
-    if( bS[0] < 4 ) {
-        int8_t tc[4];
-        for(i=0; i<4; i++)
-            tc[i] = bS[i] ? tc0_table[index_a][bS[i] - 1] + 1 : 0;
-        h->s.dsp.h264_v_loop_filter_chroma(pix, stride, alpha, beta, tc);
-    } else {
-        h->s.dsp.h264_v_loop_filter_chroma_intra(pix, stride, alpha, beta);
-    }
-}
-
-static void filter_mb_fast( H264Context *h, int mb_x, int mb_y, uint8_t *img_y, uint8_t *img_cb, uint8_t *img_cr, unsigned int linesize, unsigned int uvlinesize) {
-    MpegEncContext * const s = &h->s;
-    int mb_xy, mb_type;
-    int qp, qp0, qp1, qpc, qpc0, qpc1, qp_thresh;
-
-    if(mb_x==0 || mb_y==0 || !s->dsp.h264_loop_filter_strength) {
-        filter_mb(h, mb_x, mb_y, img_y, img_cb, img_cr, linesize, uvlinesize);
-        return;
-    }
-    assert(!FRAME_MBAFF);
-
-    mb_xy = mb_x + mb_y*s->mb_stride;
-    mb_type = s->current_picture.mb_type[mb_xy];
-    qp = s->current_picture.qscale_table[mb_xy];
-    qp0 = s->current_picture.qscale_table[mb_xy-1];
-    qp1 = s->current_picture.qscale_table[h->top_mb_xy];
-    qpc = get_chroma_qp( h->pps.chroma_qp_index_offset, qp );
-    qpc0 = get_chroma_qp( h->pps.chroma_qp_index_offset, qp0 );
-    qpc1 = get_chroma_qp( h->pps.chroma_qp_index_offset, qp1 );
-    qp0 = (qp + qp0 + 1) >> 1;
-    qp1 = (qp + qp1 + 1) >> 1;
-    qpc0 = (qpc + qpc0 + 1) >> 1;
-    qpc1 = (qpc + qpc1 + 1) >> 1;
-    qp_thresh = 15 - h->slice_alpha_c0_offset;
-    if(qp <= qp_thresh && qp0 <= qp_thresh && qp1 <= qp_thresh &&
-       qpc <= qp_thresh && qpc0 <= qp_thresh && qpc1 <= qp_thresh)
-        return;
-
-    if( IS_INTRA(mb_type) ) {
-        int16_t bS4[4] = {4,4,4,4};
-        int16_t bS3[4] = {3,3,3,3};
-        if( IS_8x8DCT(mb_type) ) {
-            filter_mb_edgev( h, &img_y[4*0], linesize, bS4, qp0 );
-            filter_mb_edgev( h, &img_y[4*2], linesize, bS3, qp );
-            filter_mb_edgeh( h, &img_y[4*0*linesize], linesize, bS4, qp1 );
-            filter_mb_edgeh( h, &img_y[4*2*linesize], linesize, bS3, qp );
-        } else {
-            filter_mb_edgev( h, &img_y[4*0], linesize, bS4, qp0 );
-            filter_mb_edgev( h, &img_y[4*1], linesize, bS3, qp );
-            filter_mb_edgev( h, &img_y[4*2], linesize, bS3, qp );
-            filter_mb_edgev( h, &img_y[4*3], linesize, bS3, qp );
-            filter_mb_edgeh( h, &img_y[4*0*linesize], linesize, bS4, qp1 );
-            filter_mb_edgeh( h, &img_y[4*1*linesize], linesize, bS3, qp );
-            filter_mb_edgeh( h, &img_y[4*2*linesize], linesize, bS3, qp );
-            filter_mb_edgeh( h, &img_y[4*3*linesize], linesize, bS3, qp );
-        }
-        filter_mb_edgecv( h, &img_cb[2*0], uvlinesize, bS4, qpc0 );
-        filter_mb_edgecv( h, &img_cb[2*2], uvlinesize, bS3, qpc );
-        filter_mb_edgecv( h, &img_cr[2*0], uvlinesize, bS4, qpc0 );
-        filter_mb_edgecv( h, &img_cr[2*2], uvlinesize, bS3, qpc );
-        filter_mb_edgech( h, &img_cb[2*0*uvlinesize], uvlinesize, bS4, qpc1 );
-        filter_mb_edgech( h, &img_cb[2*2*uvlinesize], uvlinesize, bS3, qpc );
-        filter_mb_edgech( h, &img_cr[2*0*uvlinesize], uvlinesize, bS4, qpc1 );
-        filter_mb_edgech( h, &img_cr[2*2*uvlinesize], uvlinesize, bS3, qpc );
-        return;
-    } else {
-        DECLARE_ALIGNED_8(int16_t, bS[2][4][4]);
-        uint64_t (*bSv)[4] = (uint64_t(*)[4])bS;
-        int edges;
-        if( IS_8x8DCT(mb_type) && (h->cbp&7) == 7 ) {
-            edges = 4;
-            bSv[0][0] = bSv[0][2] = bSv[1][0] = bSv[1][2] = 0x0002000200020002ULL;
-        } else {
-            int mask_edge1 = (mb_type & (MB_TYPE_16x16 | MB_TYPE_8x16)) ? 3 :
-                             (mb_type & MB_TYPE_16x8) ? 1 : 0;
-            int mask_edge0 = (mb_type & (MB_TYPE_16x16 | MB_TYPE_8x16))
-                             && (s->current_picture.mb_type[mb_xy-1] & (MB_TYPE_16x16 | MB_TYPE_8x16))
-                             ? 3 : 0;
-            int step = IS_8x8DCT(mb_type) ? 2 : 1;
-            edges = (mb_type & MB_TYPE_16x16) && !(h->cbp & 15) ? 1 : 4;
-            s->dsp.h264_loop_filter_strength( bS, h->non_zero_count_cache, h->ref_cache, h->mv_cache,
-                                              (h->slice_type == B_TYPE), edges, step, mask_edge0, mask_edge1 );
-        }
-        if( IS_INTRA(s->current_picture.mb_type[mb_xy-1]) )
-            bSv[0][0] = 0x0004000400040004ULL;
-        if( IS_INTRA(s->current_picture.mb_type[h->top_mb_xy]) )
-            bSv[1][0] = 0x0004000400040004ULL;
-
-#define FILTER(hv,dir,edge)\
-        if(bSv[dir][edge]) {\
-            filter_mb_edge##hv( h, &img_y[4*edge*(dir?linesize:1)], linesize, bS[dir][edge], edge ? qp : qp##dir );\
-            if(!(edge&1)) {\
-                filter_mb_edgec##hv( h, &img_cb[2*edge*(dir?uvlinesize:1)], uvlinesize, bS[dir][edge], edge ? qpc : qpc##dir );\
-                filter_mb_edgec##hv( h, &img_cr[2*edge*(dir?uvlinesize:1)], uvlinesize, bS[dir][edge], edge ? qpc : qpc##dir );\
-            }\
-        }
-        if( edges == 1 ) {
-            FILTER(v,0,0);
-            FILTER(h,1,0);
-        } else if( IS_8x8DCT(mb_type) ) {
-            FILTER(v,0,0);
-            FILTER(v,0,2);
-            FILTER(h,1,0);
-            FILTER(h,1,2);
-        } else {
-            FILTER(v,0,0);
-            FILTER(v,0,1);
-            FILTER(v,0,2);
-            FILTER(v,0,3);
-            FILTER(h,1,0);
-            FILTER(h,1,1);
-            FILTER(h,1,2);
-            FILTER(h,1,3);
-        }
-#undef FILTER
-    }
-}
-
-static void filter_mb( H264Context *h, int mb_x, int mb_y, uint8_t *img_y, uint8_t *img_cb, uint8_t *img_cr, unsigned int linesize, unsigned int uvlinesize) {
-    MpegEncContext * const s = &h->s;
-    const int mb_xy= mb_x + mb_y*s->mb_stride;
-    const int mb_type = s->current_picture.mb_type[mb_xy];
-    const int mvy_limit = IS_INTERLACED(mb_type) ? 2 : 4;
-    int first_vertical_edge_done = 0;
-    int dir;
-    /* FIXME: A given frame may occupy more than one position in
-     * the reference list. So ref2frm should be populated with
-     * frame numbers, not indices. */
-    static const int ref2frm[34] = {-1,-1,0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,
-                                    16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31};
-
-    //for sufficiently low qp, filtering wouldn't do anything
-    //this is a conservative estimate: could also check beta_offset and more accurate chroma_qp
-    if(!FRAME_MBAFF){
-        int qp_thresh = 15 - h->slice_alpha_c0_offset - FFMAX(0, h->pps.chroma_qp_index_offset);
-        int qp = s->current_picture.qscale_table[mb_xy];
-        if(qp <= qp_thresh
-           && (mb_x == 0 || ((qp + s->current_picture.qscale_table[mb_xy-1] + 1)>>1) <= qp_thresh)
-           && (mb_y == 0 || ((qp + s->current_picture.qscale_table[h->top_mb_xy] + 1)>>1) <= qp_thresh)){
-            return;
-        }
-    }
-
-    if (FRAME_MBAFF
-            // left mb is in picture
-            && h->slice_table[mb_xy-1] != 255
-            // and current and left pair do not have the same interlaced type
-            && (IS_INTERLACED(mb_type) != IS_INTERLACED(s->current_picture.mb_type[mb_xy-1]))
-            // and left mb is in the same slice if deblocking_filter == 2
-            && (h->deblocking_filter!=2 || h->slice_table[mb_xy-1] == h->slice_table[mb_xy])) {
-        /* First vertical edge is different in MBAFF frames
-         * There are 8 different bS to compute and 2 different Qp
-         */
-        const int pair_xy = mb_x + (mb_y&~1)*s->mb_stride;
-        const int left_mb_xy[2] = { pair_xy-1, pair_xy-1+s->mb_stride };
-        int16_t bS[8];
-        int qp[2];
-        int chroma_qp[2];
-        int mb_qp, mbn0_qp, mbn1_qp;
-        int i;
-        first_vertical_edge_done = 1;
-
-        if( IS_INTRA(mb_type) )
-            bS[0] = bS[1] = bS[2] = bS[3] = bS[4] = bS[5] = bS[6] = bS[7] = 4;
-        else {
-            for( i = 0; i < 8; i++ ) {
-                int mbn_xy = MB_FIELD ? left_mb_xy[i>>2] : left_mb_xy[i&1];
-
-                if( IS_INTRA( s->current_picture.mb_type[mbn_xy] ) )
-                    bS[i] = 4;
-                else if( h->non_zero_count_cache[12+8*(i>>1)] != 0 ||
-                         /* FIXME: with 8x8dct + cavlc, should check cbp instead of nnz */
-                         h->non_zero_count[mbn_xy][MB_FIELD ? i&3 : (i>>2)+(mb_y&1)*2] )
-                    bS[i] = 2;
-                else
-                    bS[i] = 1;
-            }
-        }
-
-        mb_qp = s->current_picture.qscale_table[mb_xy];
-        mbn0_qp = s->current_picture.qscale_table[left_mb_xy[0]];
-        mbn1_qp = s->current_picture.qscale_table[left_mb_xy[1]];
-        qp[0] = ( mb_qp + mbn0_qp + 1 ) >> 1;
-        chroma_qp[0] = ( get_chroma_qp( h->pps.chroma_qp_index_offset, mb_qp ) +
-                         get_chroma_qp( h->pps.chroma_qp_index_offset, mbn0_qp ) + 1 ) >> 1;
-        qp[1] = ( mb_qp + mbn1_qp + 1 ) >> 1;
-        chroma_qp[1] = ( get_chroma_qp( h->pps.chroma_qp_index_offset, mb_qp ) +
-                         get_chroma_qp( h->pps.chroma_qp_index_offset, mbn1_qp ) + 1 ) >> 1;
-
-        /* Filter edge */
-        tprintf("filter mb:%d/%d MBAFF, QPy:%d/%d, QPc:%d/%d ls:%d uvls:%d", mb_x, mb_y, qp[0], qp[1], chroma_qp[0], chroma_qp[1], linesize, uvlinesize);
-        { int i; for (i = 0; i < 8; i++) tprintf(" bS[%d]:%d", i, bS[i]); tprintf("\n"); }
-        filter_mb_mbaff_edgev ( h, &img_y [0], linesize,   bS, qp );
-        filter_mb_mbaff_edgecv( h, &img_cb[0], uvlinesize, bS, chroma_qp );
-        filter_mb_mbaff_edgecv( h, &img_cr[0], uvlinesize, bS, chroma_qp );
-    }
-    /* dir : 0 -> vertical edge, 1 -> horizontal edge */
-    for( dir = 0; dir < 2; dir++ )
-    {
-        int edge;
-        const int mbm_xy = dir == 0 ? mb_xy -1 : h->top_mb_xy;
-        const int mbm_type = s->current_picture.mb_type[mbm_xy];
-        int start = h->slice_table[mbm_xy] == 255 ? 1 : 0;
-
-        const int edges = (mb_type & (MB_TYPE_16x16|MB_TYPE_SKIP))
-                                  == (MB_TYPE_16x16|MB_TYPE_SKIP) ? 1 : 4;
-        // how often to recheck mv-based bS when iterating between edges
-        const int mask_edge = (mb_type & (MB_TYPE_16x16 | (MB_TYPE_16x8 << dir))) ? 3 :
-                              (mb_type & (MB_TYPE_8x16 >> dir)) ? 1 : 0;
-        // how often to recheck mv-based bS when iterating along each edge
-        const int mask_par0 = mb_type & (MB_TYPE_16x16 | (MB_TYPE_8x16 >> dir));
-
-        if (first_vertical_edge_done) {
-            start = 1;
-            first_vertical_edge_done = 0;
-        }
-
-        if (h->deblocking_filter==2 && h->slice_table[mbm_xy] != h->slice_table[mb_xy])
-            start = 1;
-
-        if (FRAME_MBAFF && (dir == 1) && ((mb_y&1) == 0) && start == 0
-            && !IS_INTERLACED(mb_type)
-            && IS_INTERLACED(mbm_type)
-            ) {
-            // This is a special case in the norm where the filtering must
-            // be done twice (one each of the field) even if we are in a
-            // frame macroblock.
-            //
-            static const int nnz_idx[4] = {4,5,6,3};
-            unsigned int tmp_linesize   = 2 *   linesize;
-            unsigned int tmp_uvlinesize = 2 * uvlinesize;
-            int mbn_xy = mb_xy - 2 * s->mb_stride;
-            int qp, chroma_qp;
-            int i, j;
-            int16_t bS[4];
-
-            for(j=0; j<2; j++, mbn_xy += s->mb_stride){
-                if( IS_INTRA(mb_type) ||
-                    IS_INTRA(s->current_picture.mb_type[mbn_xy]) ) {
-                    bS[0] = bS[1] = bS[2] = bS[3] = 3;
-                } else {
-                    const uint8_t *mbn_nnz = h->non_zero_count[mbn_xy];
-                    for( i = 0; i < 4; i++ ) {
-                        if( h->non_zero_count_cache[scan8[0]+i] != 0 ||
-                            mbn_nnz[nnz_idx[i]] != 0 )
-                            bS[i] = 2;
-                        else
-                            bS[i] = 1;
-                    }
-                }
-                // Do not use s->qscale as luma quantizer because it has not the same
-                // value in IPCM macroblocks.
-                qp = ( s->current_picture.qscale_table[mb_xy] + s->current_picture.qscale_table[mbn_xy] + 1 ) >> 1;
-                tprintf("filter mb:%d/%d dir:%d edge:%d, QPy:%d ls:%d uvls:%d", mb_x, mb_y, dir, edge, qp, tmp_linesize, tmp_uvlinesize);
-                { int i; for (i = 0; i < 4; i++) tprintf(" bS[%d]:%d", i, bS[i]); tprintf("\n"); }
-                filter_mb_edgeh( h, &img_y[j*linesize], tmp_linesize, bS, qp );
-                chroma_qp = ( h->chroma_qp +
-                              get_chroma_qp( h->pps.chroma_qp_index_offset, s->current_picture.qscale_table[mbn_xy] ) + 1 ) >> 1;
-                filter_mb_edgech( h, &img_cb[j*uvlinesize], tmp_uvlinesize, bS, chroma_qp );
-                filter_mb_edgech( h, &img_cr[j*uvlinesize], tmp_uvlinesize, bS, chroma_qp );
-            }
-
-            start = 1;
-        }
-
-        /* Calculate bS */
-        for( edge = start; edge < edges; edge++ ) {
-            /* mbn_xy: neighbor macroblock */
-            const int mbn_xy = edge > 0 ? mb_xy : mbm_xy;
-            const int mbn_type = s->current_picture.mb_type[mbn_xy];
-            int16_t bS[4];
-            int qp;
-
-            if( (edge&1) && IS_8x8DCT(mb_type) )
-                continue;
-
-            if( IS_INTRA(mb_type) ||
-                IS_INTRA(mbn_type) ) {
-                int value;
-                if (edge == 0) {
-                    if (   (!IS_INTERLACED(mb_type) && !IS_INTERLACED(mbm_type))
-                        || ((FRAME_MBAFF || (s->picture_structure != PICT_FRAME)) && (dir == 0))
-                    ) {
-                        value = 4;
-                    } else {
-                        value = 3;
-                    }
-                } else {
-                    value = 3;
-                }
-                bS[0] = bS[1] = bS[2] = bS[3] = value;
-            } else {
-                int i, l;
-                int mv_done;
-
-                if( edge & mask_edge ) {
-                    bS[0] = bS[1] = bS[2] = bS[3] = 0;
-                    mv_done = 1;
-                }
-                else if( FRAME_MBAFF && IS_INTERLACED(mb_type ^ mbn_type)) {
-                    bS[0] = bS[1] = bS[2] = bS[3] = 1;
-                    mv_done = 1;
-                }
-                else if( mask_par0 && (edge || (mbn_type & (MB_TYPE_16x16 | (MB_TYPE_8x16 >> dir)))) ) {
-                    int b_idx= 8 + 4 + edge * (dir ? 8:1);
-                    int bn_idx= b_idx - (dir ? 8:1);
-                    int v = 0;
-                    for( l = 0; !v && l < 1 + (h->slice_type == B_TYPE); l++ ) {
-                        v |= ref2frm[h->ref_cache[l][b_idx]+2] != ref2frm[h->ref_cache[l][bn_idx]+2] ||
-                             FFABS( h->mv_cache[l][b_idx][0] - h->mv_cache[l][bn_idx][0] ) >= 4 ||
-                             FFABS( h->mv_cache[l][b_idx][1] - h->mv_cache[l][bn_idx][1] ) >= mvy_limit;
-                    }
-                    bS[0] = bS[1] = bS[2] = bS[3] = v;
-                    mv_done = 1;
-                }
-                else
-                    mv_done = 0;
-
-                for( i = 0; i < 4; i++ ) {
-                    int x = dir == 0 ? edge : i;
-                    int y = dir == 0 ? i    : edge;
-                    int b_idx= 8 + 4 + x + 8*y;
-                    int bn_idx= b_idx - (dir ? 8:1);
-
-                    if( h->non_zero_count_cache[b_idx] != 0 ||
-                        h->non_zero_count_cache[bn_idx] != 0 ) {
-                        bS[i] = 2;
-                    }
-                    else if(!mv_done)
-                    {
-                        bS[i] = 0;
-                        for( l = 0; l < 1 + (h->slice_type == B_TYPE); l++ ) {
-                            if( ref2frm[h->ref_cache[l][b_idx]+2] != ref2frm[h->ref_cache[l][bn_idx]+2] ||
-                                FFABS( h->mv_cache[l][b_idx][0] - h->mv_cache[l][bn_idx][0] ) >= 4 ||
-                                FFABS( h->mv_cache[l][b_idx][1] - h->mv_cache[l][bn_idx][1] ) >= mvy_limit ) {
-                                bS[i] = 1;
-                                break;
-                            }
-                        }
-                    }
-                }
-
-                if(bS[0]+bS[1]+bS[2]+bS[3] == 0)
-                    continue;
-            }
-
-            /* Filter edge */
-            // Do not use s->qscale as luma quantizer because it has not the same
-            // value in IPCM macroblocks.
-            qp = ( s->current_picture.qscale_table[mb_xy] + s->current_picture.qscale_table[mbn_xy] + 1 ) >> 1;
-            //tprintf("filter mb:%d/%d dir:%d edge:%d, QPy:%d, QPc:%d, QPcn:%d\n", mb_x, mb_y, dir, edge, qp, h->chroma_qp, s->current_picture.qscale_table[mbn_xy]);
-            tprintf("filter mb:%d/%d dir:%d edge:%d, QPy:%d ls:%d uvls:%d", mb_x, mb_y, dir, edge, qp, linesize, uvlinesize);
-            { int i; for (i = 0; i < 4; i++) tprintf(" bS[%d]:%d", i, bS[i]); tprintf("\n"); }
-            if( dir == 0 ) {
-                filter_mb_edgev( h, &img_y[4*edge], linesize, bS, qp );
-                if( (edge&1) == 0 ) {
-                    int chroma_qp = ( h->chroma_qp +
-                                      get_chroma_qp( h->pps.chroma_qp_index_offset, s->current_picture.qscale_table[mbn_xy] ) + 1 ) >> 1;
-                    filter_mb_edgecv( h, &img_cb[2*edge], uvlinesize, bS, chroma_qp );
-                    filter_mb_edgecv( h, &img_cr[2*edge], uvlinesize, bS, chroma_qp );
-                }
-            } else {
-                filter_mb_edgeh( h, &img_y[4*edge*linesize], linesize, bS, qp );
-                if( (edge&1) == 0 ) {
-                    int chroma_qp = ( h->chroma_qp +
-                                      get_chroma_qp( h->pps.chroma_qp_index_offset, s->current_picture.qscale_table[mbn_xy] ) + 1 ) >> 1;
-                    filter_mb_edgech( h, &img_cb[2*edge*uvlinesize], uvlinesize, bS, chroma_qp );
-                    filter_mb_edgech( h, &img_cr[2*edge*uvlinesize], uvlinesize, bS, chroma_qp );
-                }
-            }
-        }
-    }
-}
-
-static int decode_slice(H264Context *h){
-    MpegEncContext * const s = &h->s;
-    const int part_mask= s->partitioned_frame ? (AC_END|AC_ERROR) : 0x7F;
-
-    s->mb_skip_run= -1;
-
-    if( h->pps.cabac ) {
-        int i;
-
-        /* realign */
-        align_get_bits( &s->gb );
-
-        /* init cabac */
-        ff_init_cabac_states( &h->cabac);
-        ff_init_cabac_decoder( &h->cabac,
-                               s->gb.buffer + get_bits_count(&s->gb)/8,
-                               ( s->gb.size_in_bits - get_bits_count(&s->gb) + 7)/8);
-        /* calculate pre-state */
-        for( i= 0; i < 460; i++ ) {
-            int pre;
-            if( h->slice_type == I_TYPE )
-                pre = clip( ((cabac_context_init_I[i][0] * s->qscale) >>4 ) + cabac_context_init_I[i][1], 1, 126 );
-            else
-                pre = clip( ((cabac_context_init_PB[h->cabac_init_idc][i][0] * s->qscale) >>4 ) + cabac_context_init_PB[h->cabac_init_idc][i][1], 1, 126 );
-
-            if( pre <= 63 )
-                h->cabac_state[i] = 2 * ( 63 - pre ) + 0;
-            else
-                h->cabac_state[i] = 2 * ( pre - 64 ) + 1;
-        }
-
-        for(;;){
-//START_TIMER
-            int ret = decode_mb_cabac(h);
-            int eos;
-//STOP_TIMER("decode_mb_cabac")
-
-            if(ret>=0) hl_decode_mb(h);
-
-            if( ret >= 0 && FRAME_MBAFF ) { //FIXME optimal? or let mb_decode decode 16x32 ?
-                s->mb_y++;
-
-                if(ret>=0) ret = decode_mb_cabac(h);
-
-                if(ret>=0) hl_decode_mb(h);
-                s->mb_y--;
-            }
-            eos = get_cabac_terminate( &h->cabac );
-
-            if( ret < 0 || h->cabac.bytestream > h->cabac.bytestream_end + 2) {
-                av_log(h->s.avctx, AV_LOG_ERROR, "error while decoding MB %d %d, bytestream (%d)\n", s->mb_x, s->mb_y, h->cabac.bytestream_end - h->cabac.bytestream);
-                ff_er_add_slice(s, s->resync_mb_x, s->resync_mb_y, s->mb_x, s->mb_y, (AC_ERROR|DC_ERROR|MV_ERROR)&part_mask);
-                return -1;
-            }
-
-            if( ++s->mb_x >= s->mb_width ) {
-                s->mb_x = 0;
-                ff_draw_horiz_band(s, 16*s->mb_y, 16);
-                ++s->mb_y;
-                if(FRAME_MBAFF) {
-                    ++s->mb_y;
-                }
-            }
-
-            if( eos || s->mb_y >= s->mb_height ) {
-                tprintf("slice end %d %d\n", get_bits_count(&s->gb), s->gb.size_in_bits);
-                ff_er_add_slice(s, s->resync_mb_x, s->resync_mb_y, s->mb_x-1, s->mb_y, (AC_END|DC_END|MV_END)&part_mask);
-                return 0;
-            }
-        }
-
-    } else {
-        for(;;){
-            int ret = decode_mb_cavlc(h);
-
-            if(ret>=0) hl_decode_mb(h);
-
-            if(ret>=0 && FRAME_MBAFF){ //FIXME optimal? or let mb_decode decode 16x32 ?
-                s->mb_y++;
-                ret = decode_mb_cavlc(h);
-
-                if(ret>=0) hl_decode_mb(h);
-                s->mb_y--;
-            }
-
-            if(ret<0){
-                av_log(h->s.avctx, AV_LOG_ERROR, "error while decoding MB %d %d\n", s->mb_x, s->mb_y);
-                ff_er_add_slice(s, s->resync_mb_x, s->resync_mb_y, s->mb_x, s->mb_y, (AC_ERROR|DC_ERROR|MV_ERROR)&part_mask);
-
-                return -1;
-            }
-
-            if(++s->mb_x >= s->mb_width){
-                s->mb_x=0;
-                ff_draw_horiz_band(s, 16*s->mb_y, 16);
-                ++s->mb_y;
-                if(FRAME_MBAFF) {
-                    ++s->mb_y;
-                }
-                if(s->mb_y >= s->mb_height){
-                    tprintf("slice end %d %d\n", get_bits_count(&s->gb), s->gb.size_in_bits);
-
-                    if(get_bits_count(&s->gb) == s->gb.size_in_bits ) {
-                        ff_er_add_slice(s, s->resync_mb_x, s->resync_mb_y, s->mb_x-1, s->mb_y, (AC_END|DC_END|MV_END)&part_mask);
-
-                        return 0;
-                    }else{
-                        ff_er_add_slice(s, s->resync_mb_x, s->resync_mb_y, s->mb_x, s->mb_y, (AC_END|DC_END|MV_END)&part_mask);
-
-                        return -1;
-                    }
-                }
-            }
-
-            if(get_bits_count(&s->gb) >= s->gb.size_in_bits && s->mb_skip_run<=0){
-                tprintf("slice end %d %d\n", get_bits_count(&s->gb), s->gb.size_in_bits);
-                if(get_bits_count(&s->gb) == s->gb.size_in_bits ){
-                    ff_er_add_slice(s, s->resync_mb_x, s->resync_mb_y, s->mb_x-1, s->mb_y, (AC_END|DC_END|MV_END)&part_mask);
-
-                    return 0;
-                }else{
-                    ff_er_add_slice(s, s->resync_mb_x, s->resync_mb_y, s->mb_x, s->mb_y, (AC_ERROR|DC_ERROR|MV_ERROR)&part_mask);
-
-                    return -1;
-                }
-            }
-        }
-    }
-
-#if 0
-    for(;s->mb_y < s->mb_height; s->mb_y++){
-        for(;s->mb_x < s->mb_width; s->mb_x++){
-            int ret= decode_mb(h);
-
-            hl_decode_mb(h);
-
-            if(ret<0){
-                av_log(s->avctx, AV_LOG_ERROR, "error while decoding MB %d %d\n", s->mb_x, s->mb_y);
-                ff_er_add_slice(s, s->resync_mb_x, s->resync_mb_y, s->mb_x, s->mb_y, (AC_ERROR|DC_ERROR|MV_ERROR)&part_mask);
-
-                return -1;
-            }
-
-            if(++s->mb_x >= s->mb_width){
-                s->mb_x=0;
-                if(++s->mb_y >= s->mb_height){
-                    if(get_bits_count(s->gb) == s->gb.size_in_bits){
-                        ff_er_add_slice(s, s->resync_mb_x, s->resync_mb_y, s->mb_x-1, s->mb_y, (AC_END|DC_END|MV_END)&part_mask);
-
-                        return 0;
-                    }else{
-                        ff_er_add_slice(s, s->resync_mb_x, s->resync_mb_y, s->mb_x, s->mb_y, (AC_END|DC_END|MV_END)&part_mask);
-
-                        return -1;
-                    }
-                }
-            }
-
-            if(get_bits_count(s->?gb) >= s->gb?.size_in_bits){
-                if(get_bits_count(s->gb) == s->gb.size_in_bits){
-                    ff_er_add_slice(s, s->resync_mb_x, s->resync_mb_y, s->mb_x-1, s->mb_y, (AC_END|DC_END|MV_END)&part_mask);
-
-                    return 0;
-                }else{
-                    ff_er_add_slice(s, s->resync_mb_x, s->resync_mb_y, s->mb_x, s->mb_y, (AC_ERROR|DC_ERROR|MV_ERROR)&part_mask);
-
-                    return -1;
-                }
-            }
-        }
-        s->mb_x=0;
-        ff_draw_horiz_band(s, 16*s->mb_y, 16);
-    }
-#endif
-    return -1; //not reached
-}
-
-static int decode_unregistered_user_data(H264Context *h, int size){
-    MpegEncContext * const s = &h->s;
-    uint8_t user_data[16+256];
-    int e, build, i;
-
-    if(size<16)
-        return -1;
-
-    for(i=0; i<sizeof(user_data)-1 && i<size; i++){
-        user_data[i]= get_bits(&s->gb, 8);
-    }
-
-    user_data[i]= 0;
-    e= sscanf(user_data+16, "x264 - core %d"/*%s - H.264/MPEG-4 AVC codec - Copyleft 2005 - http://www.videolan.org/x264.html*/, &build);
-    if(e==1 && build>=0)
-        h->x264_build= build;
-
-    if(s->avctx->debug & FF_DEBUG_BUGS)
-        av_log(s->avctx, AV_LOG_DEBUG, "user data:\"%s\"\n", user_data+16);
-
-    for(; i<size; i++)
-        skip_bits(&s->gb, 8);
-
-    return 0;
-}
-
-static int decode_sei(H264Context *h){
-    MpegEncContext * const s = &h->s;
-
-    while(get_bits_count(&s->gb) + 16 < s->gb.size_in_bits){
-        int size, type;
-
-        type=0;
-        do{
-            type+= show_bits(&s->gb, 8);
-        }while(get_bits(&s->gb, 8) == 255);
-
-        size=0;
-        do{
-            size+= show_bits(&s->gb, 8);
-        }while(get_bits(&s->gb, 8) == 255);
-
-        switch(type){
-        case 5:
-            if(decode_unregistered_user_data(h, size) < 0)
-                return -1;
-            break;
-        default:
-            skip_bits(&s->gb, 8*size);
-        }
-
-        //FIXME check bits here
-        align_get_bits(&s->gb);
-    }
-
-    return 0;
-}
-
-static inline void decode_hrd_parameters(H264Context *h, SPS *sps){
-    MpegEncContext * const s = &h->s;
-    int cpb_count, i;
-    cpb_count = get_ue_golomb(&s->gb) + 1;
-    get_bits(&s->gb, 4); /* bit_rate_scale */
-    get_bits(&s->gb, 4); /* cpb_size_scale */
-    for(i=0; i<cpb_count; i++){
-        get_ue_golomb(&s->gb); /* bit_rate_value_minus1 */
-        get_ue_golomb(&s->gb); /* cpb_size_value_minus1 */
-        get_bits1(&s->gb);     /* cbr_flag */
-    }
-    get_bits(&s->gb, 5); /* initial_cpb_removal_delay_length_minus1 */
-    get_bits(&s->gb, 5); /* cpb_removal_delay_length_minus1 */
-    get_bits(&s->gb, 5); /* dpb_output_delay_length_minus1 */
-    get_bits(&s->gb, 5); /* time_offset_length */
-}
-
-static inline int decode_vui_parameters(H264Context *h, SPS *sps){
-    MpegEncContext * const s = &h->s;
-    int aspect_ratio_info_present_flag, aspect_ratio_idc;
-    int nal_hrd_parameters_present_flag, vcl_hrd_parameters_present_flag;
-
-    aspect_ratio_info_present_flag= get_bits1(&s->gb);
-
-    if( aspect_ratio_info_present_flag ) {
-        aspect_ratio_idc= get_bits(&s->gb, 8);
-        if( aspect_ratio_idc == EXTENDED_SAR ) {
-            sps->sar.num= get_bits(&s->gb, 16);
-            sps->sar.den= get_bits(&s->gb, 16);
-        }else if(aspect_ratio_idc < 14){
-            sps->sar=  pixel_aspect[aspect_ratio_idc];
-        }else{
-            av_log(h->s.avctx, AV_LOG_ERROR, "illegal aspect ratio\n");
-            return -1;
-        }
-    }else{
-        sps->sar.num=
-        sps->sar.den= 0;
-    }
-//            s->avctx->aspect_ratio= sar_width*s->width / (float)(s->height*sar_height);
-
-    if(get_bits1(&s->gb)){      /* overscan_info_present_flag */
-        get_bits1(&s->gb);      /* overscan_appropriate_flag */
-    }
-
-    if(get_bits1(&s->gb)){      /* video_signal_type_present_flag */
-        get_bits(&s->gb, 3);    /* video_format */
-        get_bits1(&s->gb);      /* video_full_range_flag */
-        if(get_bits1(&s->gb)){  /* colour_description_present_flag */
-            get_bits(&s->gb, 8); /* colour_primaries */
-            get_bits(&s->gb, 8); /* transfer_characteristics */
-            get_bits(&s->gb, 8); /* matrix_coefficients */
-        }
-    }
-
-    if(get_bits1(&s->gb)){      /* chroma_location_info_present_flag */
-        get_ue_golomb(&s->gb);  /* chroma_sample_location_type_top_field */
-        get_ue_golomb(&s->gb);  /* chroma_sample_location_type_bottom_field */
-    }
-
-    sps->timing_info_present_flag = get_bits1(&s->gb);
-    if(sps->timing_info_present_flag){
-        sps->num_units_in_tick = get_bits_long(&s->gb, 32);
-        sps->time_scale = get_bits_long(&s->gb, 32);
-        sps->fixed_frame_rate_flag = get_bits1(&s->gb);
-    }
-
-    nal_hrd_parameters_present_flag = get_bits1(&s->gb);
-    if(nal_hrd_parameters_present_flag)
-        decode_hrd_parameters(h, sps);
-    vcl_hrd_parameters_present_flag = get_bits1(&s->gb);
-    if(vcl_hrd_parameters_present_flag)
-        decode_hrd_parameters(h, sps);
-    if(nal_hrd_parameters_present_flag || vcl_hrd_parameters_present_flag)
-        get_bits1(&s->gb);     /* low_delay_hrd_flag */
-    get_bits1(&s->gb);         /* pic_struct_present_flag */
-
-    sps->bitstream_restriction_flag = get_bits1(&s->gb);
-    if(sps->bitstream_restriction_flag){
-        get_bits1(&s->gb);     /* motion_vectors_over_pic_boundaries_flag */
-        get_ue_golomb(&s->gb); /* max_bytes_per_pic_denom */
-        get_ue_golomb(&s->gb); /* max_bits_per_mb_denom */
-        get_ue_golomb(&s->gb); /* log2_max_mv_length_horizontal */
-        get_ue_golomb(&s->gb); /* log2_max_mv_length_vertical */
-        sps->num_reorder_frames = get_ue_golomb(&s->gb);
-        get_ue_golomb(&s->gb); /* max_dec_frame_buffering */
-    }
-
-    return 0;
-}
-
-static void decode_scaling_list(H264Context *h, uint8_t *factors, int size,
-                                const uint8_t *jvt_list, const uint8_t *fallback_list){
-    MpegEncContext * const s = &h->s;
-    int i, last = 8, next = 8;
-    const uint8_t *scan = size == 16 ? zigzag_scan : zigzag_scan8x8;
-    if(!get_bits1(&s->gb)) /* matrix not written, we use the predicted one */
-        memcpy(factors, fallback_list, size*sizeof(uint8_t));
-    else
-    for(i=0;i<size;i++){
-        if(next)
-            next = (last + get_se_golomb(&s->gb)) & 0xff;
-        if(!i && !next){ /* matrix not written, we use the preset one */
-            memcpy(factors, jvt_list, size*sizeof(uint8_t));
-            break;
-        }
-        last = factors[scan[i]] = next ? next : last;
-    }
-}
-
-static void decode_scaling_matrices(H264Context *h, SPS *sps, PPS *pps, int is_sps,
-                                   uint8_t (*scaling_matrix4)[16], uint8_t (*scaling_matrix8)[64]){
-    MpegEncContext * const s = &h->s;
-    int fallback_sps = !is_sps && sps->scaling_matrix_present;
-    const uint8_t *fallback[4] = {
-        fallback_sps ? sps->scaling_matrix4[0] : default_scaling4[0],
-        fallback_sps ? sps->scaling_matrix4[3] : default_scaling4[1],
-        fallback_sps ? sps->scaling_matrix8[0] : default_scaling8[0],
-        fallback_sps ? sps->scaling_matrix8[1] : default_scaling8[1]
-    };
-    if(get_bits1(&s->gb)){
-        sps->scaling_matrix_present |= is_sps;
-        decode_scaling_list(h,scaling_matrix4[0],16,default_scaling4[0],fallback[0]); // Intra, Y
-        decode_scaling_list(h,scaling_matrix4[1],16,default_scaling4[0],scaling_matrix4[0]); // Intra, Cr
-        decode_scaling_list(h,scaling_matrix4[2],16,default_scaling4[0],scaling_matrix4[1]); // Intra, Cb
-        decode_scaling_list(h,scaling_matrix4[3],16,default_scaling4[1],fallback[1]); // Inter, Y
-        decode_scaling_list(h,scaling_matrix4[4],16,default_scaling4[1],scaling_matrix4[3]); // Inter, Cr
-        decode_scaling_list(h,scaling_matrix4[5],16,default_scaling4[1],scaling_matrix4[4]); // Inter, Cb
-        if(is_sps || pps->transform_8x8_mode){
-            decode_scaling_list(h,scaling_matrix8[0],64,default_scaling8[0],fallback[2]);  // Intra, Y
-            decode_scaling_list(h,scaling_matrix8[1],64,default_scaling8[1],fallback[3]);  // Inter, Y
-        }
-    } else if(fallback_sps) {
-        memcpy(scaling_matrix4, sps->scaling_matrix4, 6*16*sizeof(uint8_t));
-        memcpy(scaling_matrix8, sps->scaling_matrix8, 2*64*sizeof(uint8_t));
-    }
-}
-
-static inline int decode_seq_parameter_set(H264Context *h){
-    MpegEncContext * const s = &h->s;
-    int profile_idc, level_idc;
-    int sps_id, i;
-    SPS *sps;
-
-    profile_idc= get_bits(&s->gb, 8);
-    get_bits1(&s->gb);   //constraint_set0_flag
-    get_bits1(&s->gb);   //constraint_set1_flag
-    get_bits1(&s->gb);   //constraint_set2_flag
-    get_bits1(&s->gb);   //constraint_set3_flag
-    get_bits(&s->gb, 4); // reserved
-    level_idc= get_bits(&s->gb, 8);
-    sps_id= get_ue_golomb(&s->gb);
-
-    sps= &h->sps_buffer[ sps_id ];
-    sps->profile_idc= profile_idc;
-    sps->level_idc= level_idc;
-
-    if(sps->profile_idc >= 100){ //high profile
-        if(get_ue_golomb(&s->gb) == 3) //chroma_format_idc
-            get_bits1(&s->gb);  //residual_color_transform_flag
-        get_ue_golomb(&s->gb);  //bit_depth_luma_minus8
-        get_ue_golomb(&s->gb);  //bit_depth_chroma_minus8
-        sps->transform_bypass = get_bits1(&s->gb);
-        decode_scaling_matrices(h, sps, NULL, 1, sps->scaling_matrix4, sps->scaling_matrix8);
-    }else
-        sps->scaling_matrix_present = 0;
-
-    sps->log2_max_frame_num= get_ue_golomb(&s->gb) + 4;
-    sps->poc_type= get_ue_golomb(&s->gb);
-
-    if(sps->poc_type == 0){ //FIXME #define
-        sps->log2_max_poc_lsb= get_ue_golomb(&s->gb) + 4;
-    } else if(sps->poc_type == 1){//FIXME #define
-        sps->delta_pic_order_always_zero_flag= get_bits1(&s->gb);
-        sps->offset_for_non_ref_pic= get_se_golomb(&s->gb);
-        sps->offset_for_top_to_bottom_field= get_se_golomb(&s->gb);
-        sps->poc_cycle_length= get_ue_golomb(&s->gb);
-
-        for(i=0; i<sps->poc_cycle_length; i++)
-            sps->offset_for_ref_frame[i]= get_se_golomb(&s->gb);
-    }
-    if(sps->poc_type > 2){
-        av_log(h->s.avctx, AV_LOG_ERROR, "illegal POC type %d\n", sps->poc_type);
-        return -1;
-    }
-
-    sps->ref_frame_count= get_ue_golomb(&s->gb);
-    if(sps->ref_frame_count > MAX_PICTURE_COUNT-2){
-        av_log(h->s.avctx, AV_LOG_ERROR, "too many reference frames\n");
-    }
-    sps->gaps_in_frame_num_allowed_flag= get_bits1(&s->gb);
-    sps->mb_width= get_ue_golomb(&s->gb) + 1;
-    sps->mb_height= get_ue_golomb(&s->gb) + 1;
-    if((unsigned)sps->mb_width >= INT_MAX/16 || (unsigned)sps->mb_height >= INT_MAX/16 ||
-       avcodec_check_dimensions(NULL, 16*sps->mb_width, 16*sps->mb_height))
-        return -1;
-
-    sps->frame_mbs_only_flag= get_bits1(&s->gb);
-    if(!sps->frame_mbs_only_flag)
-        sps->mb_aff= get_bits1(&s->gb);
-    else
-        sps->mb_aff= 0;
-
-    sps->direct_8x8_inference_flag= get_bits1(&s->gb);
-
-#ifndef ALLOW_INTERLACE
-    if(sps->mb_aff)
-        av_log(h->s.avctx, AV_LOG_ERROR, "MBAFF support not included; enable it at compile-time.\n");
-#endif
-    if(!sps->direct_8x8_inference_flag && sps->mb_aff)
-        av_log(h->s.avctx, AV_LOG_ERROR, "MBAFF + !direct_8x8_inference is not implemented\n");
-
-    sps->crop= get_bits1(&s->gb);
-    if(sps->crop){
-        sps->crop_left  = get_ue_golomb(&s->gb);
-        sps->crop_right = get_ue_golomb(&s->gb);
-        sps->crop_top   = get_ue_golomb(&s->gb);
-        sps->crop_bottom= get_ue_golomb(&s->gb);
-        if(sps->crop_left || sps->crop_top){
-            av_log(h->s.avctx, AV_LOG_ERROR, "insane cropping not completely supported, this could look slightly wrong ...\n");
-        }
-    }else{
-        sps->crop_left  =
-        sps->crop_right =
-        sps->crop_top   =
-        sps->crop_bottom= 0;
-    }
-
-    sps->vui_parameters_present_flag= get_bits1(&s->gb);
-    if( sps->vui_parameters_present_flag )
-        decode_vui_parameters(h, sps);
-
-    if(s->avctx->debug&FF_DEBUG_PICT_INFO){
-        av_log(h->s.avctx, AV_LOG_DEBUG, "sps:%d profile:%d/%d poc:%d ref:%d %dx%d %s %s crop:%d/%d/%d/%d %s\n",
-               sps_id, sps->profile_idc, sps->level_idc,
-               sps->poc_type,
-               sps->ref_frame_count,
-               sps->mb_width, sps->mb_height,
-               sps->frame_mbs_only_flag ? "FRM" : (sps->mb_aff ? "MB-AFF" : "PIC-AFF"),
-               sps->direct_8x8_inference_flag ? "8B8" : "",
-               sps->crop_left, sps->crop_right,
-               sps->crop_top, sps->crop_bottom,
-               sps->vui_parameters_present_flag ? "VUI" : ""
-               );
-    }
-    return 0;
-}
-
-static inline int decode_picture_parameter_set(H264Context *h, int bit_length){
-    MpegEncContext * const s = &h->s;
-    int pps_id= get_ue_golomb(&s->gb);
-    PPS *pps= &h->pps_buffer[pps_id];
-
-    pps->sps_id= get_ue_golomb(&s->gb);
-    pps->cabac= get_bits1(&s->gb);
-    pps->pic_order_present= get_bits1(&s->gb);
-    pps->slice_group_count= get_ue_golomb(&s->gb) + 1;
-    if(pps->slice_group_count > 1 ){
-        pps->mb_slice_group_map_type= get_ue_golomb(&s->gb);
-        av_log(h->s.avctx, AV_LOG_ERROR, "FMO not supported\n");
-        switch(pps->mb_slice_group_map_type){
-        case 0:
-#if 0
-|   for( i = 0; i <= num_slice_groups_minus1; i++ ) |   |        |
-|    run_length[ i ]                                |1  |ue(v)   |
-#endif
-            break;
-        case 2:
-#if 0
-|   for( i = 0; i < num_slice_groups_minus1; i++ )  |   |        |
-|{                                                  |   |        |
-|    top_left_mb[ i ]                               |1  |ue(v)   |
-|    bottom_right_mb[ i ]                           |1  |ue(v)   |
-|   }                                               |   |        |
-#endif
-            break;
-        case 3:
-        case 4:
-        case 5:
-#if 0
-|   slice_group_change_direction_flag               |1  |u(1)    |
-|   slice_group_change_rate_minus1                  |1  |ue(v)   |
-#endif
-            break;
-        case 6:
-#if 0
-|   slice_group_id_cnt_minus1                       |1  |ue(v)   |
-|   for( i = 0; i <= slice_group_id_cnt_minus1; i++ |   |        |
-|)                                                  |   |        |
-|    slice_group_id[ i ]                            |1  |u(v)    |
-#endif
-            break;
-        }
-    }
-    pps->ref_count[0]= get_ue_golomb(&s->gb) + 1;
-    pps->ref_count[1]= get_ue_golomb(&s->gb) + 1;
-    if(pps->ref_count[0] > 32 || pps->ref_count[1] > 32){
-        av_log(h->s.avctx, AV_LOG_ERROR, "reference overflow (pps)\n");
-        return -1;
-    }
-
-    pps->weighted_pred= get_bits1(&s->gb);
-    pps->weighted_bipred_idc= get_bits(&s->gb, 2);
-    pps->init_qp= get_se_golomb(&s->gb) + 26;
-    pps->init_qs= get_se_golomb(&s->gb) + 26;
-    pps->chroma_qp_index_offset= get_se_golomb(&s->gb);
-    pps->deblocking_filter_parameters_present= get_bits1(&s->gb);
-    pps->constrained_intra_pred= get_bits1(&s->gb);
-    pps->redundant_pic_cnt_present = get_bits1(&s->gb);
-
-    pps->transform_8x8_mode= 0;
-    h->dequant_coeff_pps= -1; //contents of sps/pps can change even if id doesn't, so reinit
-    memset(pps->scaling_matrix4, 16, 6*16*sizeof(uint8_t));
-    memset(pps->scaling_matrix8, 16, 2*64*sizeof(uint8_t));
-
-    if(get_bits_count(&s->gb) < bit_length){
-        pps->transform_8x8_mode= get_bits1(&s->gb);
-        decode_scaling_matrices(h, &h->sps_buffer[pps->sps_id], pps, 0, pps->scaling_matrix4, pps->scaling_matrix8);
-        get_se_golomb(&s->gb);  //second_chroma_qp_index_offset
-    }
-
-    if(s->avctx->debug&FF_DEBUG_PICT_INFO){
-        av_log(h->s.avctx, AV_LOG_DEBUG, "pps:%d sps:%d %s slice_groups:%d ref:%d/%d %s qp:%d/%d/%d %s %s %s %s\n",
-               pps_id, pps->sps_id,
-               pps->cabac ? "CABAC" : "CAVLC",
-               pps->slice_group_count,
-               pps->ref_count[0], pps->ref_count[1],
-               pps->weighted_pred ? "weighted" : "",
-               pps->init_qp, pps->init_qs, pps->chroma_qp_index_offset,
-               pps->deblocking_filter_parameters_present ? "LPAR" : "",
-               pps->constrained_intra_pred ? "CONSTR" : "",
-               pps->redundant_pic_cnt_present ? "REDU" : "",
-               pps->transform_8x8_mode ? "8x8DCT" : ""
-               );
-    }
-
-    return 0;
-}
-
-/**
- * finds the end of the current frame in the bitstream.
- * @return the position of the first byte of the next frame, or -1
- */
-static int find_frame_end(H264Context *h, const uint8_t *buf, int buf_size){
-    int i;
-    uint32_t state;
-    ParseContext *pc = &(h->s.parse_context);
-//printf("first %02X%02X%02X%02X\n", buf[0], buf[1],buf[2],buf[3]);
-//    mb_addr= pc->mb_addr - 1;
-    state= pc->state;
-    for(i=0; i<=buf_size; i++){
-        if((state&0xFFFFFF1F) == 0x101 || (state&0xFFFFFF1F) == 0x102 || (state&0xFFFFFF1F) == 0x105){
-            tprintf("find_frame_end new startcode = %08x, frame_start_found = %d, pos = %d\n", state, pc->frame_start_found, i);
-            if(pc->frame_start_found){
-                // If there isn't one more byte in the buffer
-                // the test on first_mb_in_slice cannot be done yet
-                // do it at next call.
-                if (i >= buf_size) break;
-                if (buf[i] & 0x80) {
-                    // first_mb_in_slice is 0, probably the first nal of a new
-                    // slice
-                    tprintf("find_frame_end frame_end_found, state = %08x, pos = %d\n", state, i);
-                    pc->state=-1;
-                    pc->frame_start_found= 0;
-                    return i-4;
-                }
-            }
-            pc->frame_start_found = 1;
-        }
-        if((state&0xFFFFFF1F) == 0x107 || (state&0xFFFFFF1F) == 0x108 || (state&0xFFFFFF1F) == 0x109){
-           if(pc->frame_start_found){
-                pc->state=-1;
-                pc->frame_start_found= 0;
-                return i-4;
-           }
-        }
-        if (i<buf_size)
-            state= (state<<8) | buf[i];
-    }
-
-    pc->state= state;
-    return END_NOT_FOUND;
-}
-
-#ifdef CONFIG_H264_PARSER
-static int h264_parse(AVCodecParserContext *s,
-                      AVCodecContext *avctx,
-                      uint8_t **poutbuf, int *poutbuf_size,
-                      const uint8_t *buf, int buf_size)
-{
-    H264Context *h = s->priv_data;
-    ParseContext *pc = &h->s.parse_context;
-    int next;
-
-    next= find_frame_end(h, buf, buf_size);
-
-    if (ff_combine_frame(pc, next, (uint8_t **)&buf, &buf_size) < 0) {
-        *poutbuf = NULL;
-        *poutbuf_size = 0;
-        return buf_size;
-    }
-
-    *poutbuf = (uint8_t *)buf;
-    *poutbuf_size = buf_size;
-    return next;
-}
-
-static int h264_split(AVCodecContext *avctx,
-                      const uint8_t *buf, int buf_size)
-{
-    int i;
-    uint32_t state = -1;
-    int has_sps= 0;
-
-    for(i=0; i<=buf_size; i++){
-        if((state&0xFFFFFF1F) == 0x107)
-            has_sps=1;
-/*        if((state&0xFFFFFF1F) == 0x101 || (state&0xFFFFFF1F) == 0x102 || (state&0xFFFFFF1F) == 0x105){
-        }*/
-        if((state&0xFFFFFF00) == 0x100 && (state&0xFFFFFF1F) != 0x107 && (state&0xFFFFFF1F) != 0x108 && (state&0xFFFFFF1F) != 0x109){
-            if(has_sps){
-                while(i>4 && buf[i-5]==0) i--;
-                return i-4;
-            }
-        }
-        if (i<buf_size)
-            state= (state<<8) | buf[i];
-    }
-    return 0;
-}
-#endif /* CONFIG_H264_PARSER */
-
-static int decode_nal_units(H264Context *h, uint8_t *buf, int buf_size){
-    MpegEncContext * const s = &h->s;
-    AVCodecContext * const avctx= s->avctx;
-    int buf_index=0;
-#if 0
-    int i;
-    for(i=0; i<50; i++){
-        av_log(NULL, AV_LOG_ERROR,"%02X ", buf[i]);
-    }
-#endif
-    h->slice_num = 0;
-    s->current_picture_ptr= NULL;
-    for(;;){
-        int consumed;
-        int dst_length;
-        int bit_length;
-        uint8_t *ptr;
-        int i, nalsize = 0;
-
-      if(h->is_avc) {
-        if(buf_index >= buf_size) break;
-        nalsize = 0;
-        for(i = 0; i < h->nal_length_size; i++)
-            nalsize = (nalsize << 8) | buf[buf_index++];
-        if(nalsize <= 1){
-            if(nalsize == 1){
-                buf_index++;
-                continue;
-            }else{
-                av_log(h->s.avctx, AV_LOG_ERROR, "AVC: nal size %d\n", nalsize);
-                break;
-            }
-        }
-      } else {
-        // start code prefix search
-        for(; buf_index + 3 < buf_size; buf_index++){
-            // this should allways succeed in the first iteration
-            if(buf[buf_index] == 0 && buf[buf_index+1] == 0 && buf[buf_index+2] == 1)
-                break;
-        }
-
-        if(buf_index+3 >= buf_size) break;
-
-        buf_index+=3;
-      }
-
-        ptr= decode_nal(h, buf + buf_index, &dst_length, &consumed, h->is_avc ? nalsize : buf_size - buf_index);
-        while(ptr[dst_length - 1] == 0 && dst_length > 1)
-            dst_length--;
-        bit_length= 8*dst_length - decode_rbsp_trailing(ptr + dst_length - 1);
-
-        if(s->avctx->debug&FF_DEBUG_STARTCODE){
-            av_log(h->s.avctx, AV_LOG_DEBUG, "NAL %d at %d/%d length %d\n", h->nal_unit_type, buf_index, buf_size, dst_length);
-        }
-
-        if (h->is_avc && (nalsize != consumed))
-            av_log(h->s.avctx, AV_LOG_ERROR, "AVC: Consumed only %d bytes instead of %d\n", consumed, nalsize);
-
-        buf_index += consumed;
-
-        if(  (s->hurry_up == 1 && h->nal_ref_idc  == 0) //FIXME dont discard SEI id
-           ||(avctx->skip_frame >= AVDISCARD_NONREF && h->nal_ref_idc  == 0))
-            continue;
-
-        switch(h->nal_unit_type){
-        case NAL_IDR_SLICE:
-            idr(h); //FIXME ensure we don't loose some frames if there is reordering
-        case NAL_SLICE:
-            init_get_bits(&s->gb, ptr, bit_length);
-            h->intra_gb_ptr=
-            h->inter_gb_ptr= &s->gb;
-            s->data_partitioning = 0;
-
-            if(decode_slice_header(h) < 0){
-                av_log(h->s.avctx, AV_LOG_ERROR, "decode_slice_header error\n");
-                break;
-            }
-            s->current_picture_ptr->key_frame= (h->nal_unit_type == NAL_IDR_SLICE);
-            if(h->redundant_pic_count==0 && s->hurry_up < 5
-               && (avctx->skip_frame < AVDISCARD_NONREF || h->nal_ref_idc)
-               && (avctx->skip_frame < AVDISCARD_BIDIR  || h->slice_type!=B_TYPE)
-               && (avctx->skip_frame < AVDISCARD_NONKEY || h->slice_type==I_TYPE)
-               && avctx->skip_frame < AVDISCARD_ALL)
-                decode_slice(h);
-            break;
-        case NAL_DPA:
-            init_get_bits(&s->gb, ptr, bit_length);
-            h->intra_gb_ptr=
-            h->inter_gb_ptr= NULL;
-            s->data_partitioning = 1;
-
-            if(decode_slice_header(h) < 0){
-                av_log(h->s.avctx, AV_LOG_ERROR, "decode_slice_header error\n");
-            }
-            break;
-        case NAL_DPB:
-            init_get_bits(&h->intra_gb, ptr, bit_length);
-            h->intra_gb_ptr= &h->intra_gb;
-            break;
-        case NAL_DPC:
-            init_get_bits(&h->inter_gb, ptr, bit_length);
-            h->inter_gb_ptr= &h->inter_gb;
-
-            if(h->redundant_pic_count==0 && h->intra_gb_ptr && s->data_partitioning
-               && s->hurry_up < 5
-               && (avctx->skip_frame < AVDISCARD_NONREF || h->nal_ref_idc)
-               && (avctx->skip_frame < AVDISCARD_BIDIR  || h->slice_type!=B_TYPE)
-               && (avctx->skip_frame < AVDISCARD_NONKEY || h->slice_type==I_TYPE)
-               && avctx->skip_frame < AVDISCARD_ALL)
-                decode_slice(h);
-            break;
-        case NAL_SEI:
-            init_get_bits(&s->gb, ptr, bit_length);
-            decode_sei(h);
-            break;
-        case NAL_SPS:
-            init_get_bits(&s->gb, ptr, bit_length);
-            decode_seq_parameter_set(h);
-
-            if(s->flags& CODEC_FLAG_LOW_DELAY)
-                s->low_delay=1;
-
-            if(avctx->has_b_frames < 2)
-                avctx->has_b_frames= !s->low_delay;
-            break;
-        case NAL_PPS:
-            init_get_bits(&s->gb, ptr, bit_length);
-
-            decode_picture_parameter_set(h, bit_length);
-
-            break;
-        case NAL_AUD:
-        case NAL_END_SEQUENCE:
-        case NAL_END_STREAM:
-        case NAL_FILLER_DATA:
-        case NAL_SPS_EXT:
-        case NAL_AUXILIARY_SLICE:
-            break;
-        default:
-            av_log(avctx, AV_LOG_ERROR, "Unknown NAL code: %d\n", h->nal_unit_type);
-        }
-    }
-
-    if(!s->current_picture_ptr) return buf_index; //no frame
-
-    s->current_picture_ptr->qscale_type= FF_QSCALE_TYPE_H264;
-    s->current_picture_ptr->pict_type= s->pict_type;
-
-    h->prev_frame_num_offset= h->frame_num_offset;
-    h->prev_frame_num= h->frame_num;
-    if(s->current_picture_ptr->reference){
-        h->prev_poc_msb= h->poc_msb;
-        h->prev_poc_lsb= h->poc_lsb;
-    }
-    if(s->current_picture_ptr->reference)
-        execute_ref_pic_marking(h, h->mmco, h->mmco_index);
-
-    ff_er_frame_end(s);
-
-    MPV_frame_end(s);
-
-    return buf_index;
-}
-
-/**
- * returns the number of bytes consumed for building the current frame
- */
-static int get_consumed_bytes(MpegEncContext *s, int pos, int buf_size){
-    if(s->flags&CODEC_FLAG_TRUNCATED){
-        pos -= s->parse_context.last_index;
-        if(pos<0) pos=0; // FIXME remove (unneeded?)
-
-        return pos;
-    }else{
-        if(pos==0) pos=1; //avoid infinite loops (i doubt thats needed but ...)
-        if(pos+10>buf_size) pos=buf_size; // oops ;)
-
-        return pos;
-    }
-}
-
-static int decode_frame(AVCodecContext *avctx,
-                             void *data, int *data_size,
-                             uint8_t *buf, int buf_size)
-{
-    H264Context *h = avctx->priv_data;
-    MpegEncContext *s = &h->s;
-    AVFrame *pict = data;
-    int buf_index;
-
-    s->flags= avctx->flags;
-    s->flags2= avctx->flags2;
-
-   /* no supplementary picture */
-    if (buf_size == 0) {
-        return 0;
-    }
-
-    if(s->flags&CODEC_FLAG_TRUNCATED){
-        int next= find_frame_end(h, buf, buf_size);
-
-        if( ff_combine_frame(&s->parse_context, next, &buf, &buf_size) < 0 )
-            return buf_size;
-//printf("next:%d buf_size:%d last_index:%d\n", next, buf_size, s->parse_context.last_index);
-    }
-
-    if(h->is_avc && !h->got_avcC) {
-        int i, cnt, nalsize;
-        unsigned char *p = avctx->extradata;
-        if(avctx->extradata_size < 7) {
-            av_log(avctx, AV_LOG_ERROR, "avcC too short\n");
-            return -1;
-        }
-        if(*p != 1) {
-            av_log(avctx, AV_LOG_ERROR, "Unknown avcC version %d\n", *p);
-            return -1;
-        }
-        /* sps and pps in the avcC always have length coded with 2 bytes,
-           so put a fake nal_length_size = 2 while parsing them */
-        h->nal_length_size = 2;
-        // Decode sps from avcC
-        cnt = *(p+5) & 0x1f; // Number of sps
-        p += 6;
-        for (i = 0; i < cnt; i++) {
-            nalsize = BE_16(p) + 2;
-            if(decode_nal_units(h, p, nalsize) < 0) {
-                av_log(avctx, AV_LOG_ERROR, "Decoding sps %d from avcC failed\n", i);
-                return -1;
-            }
-            p += nalsize;
-        }
-        // Decode pps from avcC
-        cnt = *(p++); // Number of pps
-        for (i = 0; i < cnt; i++) {
-            nalsize = BE_16(p) + 2;
-            if(decode_nal_units(h, p, nalsize)  != nalsize) {
-                av_log(avctx, AV_LOG_ERROR, "Decoding pps %d from avcC failed\n", i);
-                return -1;
-            }
-            p += nalsize;
-        }
-        // Now store right nal length size, that will be use to parse all other nals
-        h->nal_length_size = ((*(((char*)(avctx->extradata))+4))&0x03)+1;
-        // Do not reparse avcC
-        h->got_avcC = 1;
-    }
-
-    if(!h->is_avc && s->avctx->extradata_size && s->picture_number==0){
-        if(decode_nal_units(h, s->avctx->extradata, s->avctx->extradata_size) < 0)
-            return -1;
-    }
-
-    buf_index=decode_nal_units(h, buf, buf_size);
-    if(buf_index < 0)
-        return -1;
-
-    //FIXME do something with unavailable reference frames
-
-//    if(ret==FRAME_SKIPPED) return get_consumed_bytes(s, buf_index, buf_size);
-    if(!s->current_picture_ptr){
-        av_log(h->s.avctx, AV_LOG_DEBUG, "error, NO frame\n");
-        return -1;
-    }
-
-    {
-        Picture *out = s->current_picture_ptr;
-#if 0 //decode order
-        *data_size = sizeof(AVFrame);
-#else
-        /* Sort B-frames into display order */
-        Picture *cur = s->current_picture_ptr;
-        Picture *prev = h->delayed_output_pic;
-        int i, pics, cross_idr, out_of_order, out_idx;
-
-        if(h->sps.bitstream_restriction_flag
-           && s->avctx->has_b_frames < h->sps.num_reorder_frames){
-            s->avctx->has_b_frames = h->sps.num_reorder_frames;
-            s->low_delay = 0;
-        }
-
-        pics = 0;
-        while(h->delayed_pic[pics]) pics++;
-        h->delayed_pic[pics++] = cur;
-        if(cur->reference == 0)
-            cur->reference = 1;
-
-        cross_idr = 0;
-        for(i=0; h->delayed_pic[i]; i++)
-            if(h->delayed_pic[i]->key_frame || h->delayed_pic[i]->poc==0)
-                cross_idr = 1;
-
-        out = h->delayed_pic[0];
-        out_idx = 0;
-        for(i=1; h->delayed_pic[i] && !h->delayed_pic[i]->key_frame; i++)
-            if(h->delayed_pic[i]->poc < out->poc){
-                out = h->delayed_pic[i];
-                out_idx = i;
-            }
-
-        out_of_order = !cross_idr && prev && out->poc < prev->poc;
-        if(h->sps.bitstream_restriction_flag && s->avctx->has_b_frames >= h->sps.num_reorder_frames)
-            { }
-        else if(prev && pics <= s->avctx->has_b_frames)
-            out = prev;
-        else if((out_of_order && pics-1 == s->avctx->has_b_frames && pics < 15)
-           || (s->low_delay &&
-            ((!cross_idr && prev && out->poc > prev->poc + 2)
-             || cur->pict_type == B_TYPE)))
-        {
-            s->low_delay = 0;
-            s->avctx->has_b_frames++;
-            out = prev;
-        }
-        else if(out_of_order)
-            out = prev;
-
-        if(out_of_order || pics > s->avctx->has_b_frames){
-            for(i=out_idx; h->delayed_pic[i]; i++)
-                h->delayed_pic[i] = h->delayed_pic[i+1];
-        }
-
-        if(prev == out)
-            *data_size = 0;
-        else
-            *data_size = sizeof(AVFrame);
-        if(prev && prev != out && prev->reference == 1)
-            prev->reference = 0;
-        h->delayed_output_pic = out;
-#endif
-
-        if(out)
-            *pict= *(AVFrame*)out;
-        else
-            av_log(avctx, AV_LOG_DEBUG, "no picture\n");
-    }
-
-    assert(pict->data[0] || !*data_size);
-    ff_print_debug_info(s, pict);
-//printf("out %d\n", (int)pict->data[0]);
-#if 0 //?
-
-    /* Return the Picture timestamp as the frame number */
-    /* we substract 1 because it is added on utils.c    */
-    avctx->frame_number = s->picture_number - 1;
-#endif
-    return get_consumed_bytes(s, buf_index, buf_size);
-}
-#if 0
-static inline void fill_mb_avail(H264Context *h){
-    MpegEncContext * const s = &h->s;
-    const int mb_xy= s->mb_x + s->mb_y*s->mb_stride;
-
-    if(s->mb_y){
-        h->mb_avail[0]= s->mb_x                 && h->slice_table[mb_xy - s->mb_stride - 1] == h->slice_num;
-        h->mb_avail[1]=                            h->slice_table[mb_xy - s->mb_stride    ] == h->slice_num;
-        h->mb_avail[2]= s->mb_x+1 < s->mb_width && h->slice_table[mb_xy - s->mb_stride + 1] == h->slice_num;
-    }else{
-        h->mb_avail[0]=
-        h->mb_avail[1]=
-        h->mb_avail[2]= 0;
-    }
-    h->mb_avail[3]= s->mb_x && h->slice_table[mb_xy - 1] == h->slice_num;
-    h->mb_avail[4]= 1; //FIXME move out
-    h->mb_avail[5]= 0; //FIXME move out
-}
-#endif
-
-#if 0 //selftest
-#define COUNT 8000
-#define SIZE (COUNT*40)
-int main(){
-    int i;
-    uint8_t temp[SIZE];
-    PutBitContext pb;
-    GetBitContext gb;
-//    int int_temp[10000];
-    DSPContext dsp;
-    AVCodecContext avctx;
-
-    dsputil_init(&dsp, &avctx);
-
-    init_put_bits(&pb, temp, SIZE);
-    printf("testing unsigned exp golomb\n");
-    for(i=0; i<COUNT; i++){
-        START_TIMER
-        set_ue_golomb(&pb, i);
-        STOP_TIMER("set_ue_golomb");
-    }
-    flush_put_bits(&pb);
-
-    init_get_bits(&gb, temp, 8*SIZE);
-    for(i=0; i<COUNT; i++){
-        int j, s;
-
-        s= show_bits(&gb, 24);
-
-        START_TIMER
-        j= get_ue_golomb(&gb);
-        if(j != i){
-            printf("missmatch! at %d (%d should be %d) bits:%6X\n", i, j, i, s);
-//            return -1;
-        }
-        STOP_TIMER("get_ue_golomb");
-    }
-
-
-    init_put_bits(&pb, temp, SIZE);
-    printf("testing signed exp golomb\n");
-    for(i=0; i<COUNT; i++){
-        START_TIMER
-        set_se_golomb(&pb, i - COUNT/2);
-        STOP_TIMER("set_se_golomb");
-    }
-    flush_put_bits(&pb);
-
-    init_get_bits(&gb, temp, 8*SIZE);
-    for(i=0; i<COUNT; i++){
-        int j, s;
-
-        s= show_bits(&gb, 24);
-
-        START_TIMER
-        j= get_se_golomb(&gb);
-        if(j != i - COUNT/2){
-            printf("missmatch! at %d (%d should be %d) bits:%6X\n", i, j, i, s);
-//            return -1;
-        }
-        STOP_TIMER("get_se_golomb");
-    }
-
-    printf("testing 4x4 (I)DCT\n");
-
-    DCTELEM block[16];
-    uint8_t src[16], ref[16];
-    uint64_t error= 0, max_error=0;
-
-    for(i=0; i<COUNT; i++){
-        int j;
-//        printf("%d %d %d\n", r1, r2, (r2-r1)*16);
-        for(j=0; j<16; j++){
-            ref[j]= random()%255;
-            src[j]= random()%255;
-        }
-
-        h264_diff_dct_c(block, src, ref, 4);
-
-        //normalize
-        for(j=0; j<16; j++){
-//            printf("%d ", block[j]);
-            block[j]= block[j]*4;
-            if(j&1) block[j]= (block[j]*4 + 2)/5;
-            if(j&4) block[j]= (block[j]*4 + 2)/5;
-        }
-//        printf("\n");
-
-        s->dsp.h264_idct_add(ref, block, 4);
-/*        for(j=0; j<16; j++){
-            printf("%d ", ref[j]);
-        }
-        printf("\n");*/
-
-        for(j=0; j<16; j++){
-            int diff= FFABS(src[j] - ref[j]);
-
-            error+= diff*diff;
-            max_error= FFMAX(max_error, diff);
-        }
-    }
-    printf("error=%f max_error=%d\n", ((float)error)/COUNT/16, (int)max_error );
-#if 0
-    printf("testing quantizer\n");
-    for(qp=0; qp<52; qp++){
-        for(i=0; i<16; i++)
-            src1_block[i]= src2_block[i]= random()%255;
-
-    }
-#endif
-    printf("Testing NAL layer\n");
-
-    uint8_t bitstream[COUNT];
-    uint8_t nal[COUNT*2];
-    H264Context h;
-    memset(&h, 0, sizeof(H264Context));
-
-    for(i=0; i<COUNT; i++){
-        int zeros= i;
-        int nal_length;
-        int consumed;
-        int out_length;
-        uint8_t *out;
-        int j;
-
-        for(j=0; j<COUNT; j++){
-            bitstream[j]= (random() % 255) + 1;
-        }
-
-        for(j=0; j<zeros; j++){
-            int pos= random() % COUNT;
-            while(bitstream[pos] == 0){
-                pos++;
-                pos %= COUNT;
-            }
-            bitstream[pos]=0;
-        }
-
-        START_TIMER
-
-        nal_length= encode_nal(&h, nal, bitstream, COUNT, COUNT*2);
-        if(nal_length<0){
-            printf("encoding failed\n");
-            return -1;
-        }
-
-        out= decode_nal(&h, nal, &out_length, &consumed, nal_length);
-
-        STOP_TIMER("NAL")
-
-        if(out_length != COUNT){
-            printf("incorrect length %d %d\n", out_length, COUNT);
-            return -1;
-        }
-
-        if(consumed != nal_length){
-            printf("incorrect consumed length %d %d\n", nal_length, consumed);
-            return -1;
-        }
-
-        if(memcmp(bitstream, out, COUNT)){
-            printf("missmatch\n");
-            return -1;
-        }
-    }
-
-    printf("Testing RBSP\n");
-
-
-    return 0;
-}
-#endif
-
-
-static int decode_end(AVCodecContext *avctx)
-{
-    H264Context *h = avctx->priv_data;
-    MpegEncContext *s = &h->s;
-
-    av_freep(&h->rbsp_buffer);
-    free_tables(h); //FIXME cleanup init stuff perhaps
-    MPV_common_end(s);
-
-//    memset(h, 0, sizeof(H264Context));
-
-    return 0;
-}
-
-
-AVCodec h264_decoder = {
-    "h264",
-    CODEC_TYPE_VIDEO,
-    CODEC_ID_H264,
-    sizeof(H264Context),
-    decode_init,
-    NULL,
-    decode_end,
-    decode_frame,
-    /*CODEC_CAP_DRAW_HORIZ_BAND |*/ CODEC_CAP_DR1 | CODEC_CAP_TRUNCATED | CODEC_CAP_DELAY,
-    .flush= flush_dpb,
-};
-
-#ifdef CONFIG_H264_PARSER
-AVCodecParser h264_parser = {
-    { CODEC_ID_H264 },
-    sizeof(H264Context),
-    NULL,
-    h264_parse,
-    ff_parse_close,
-    h264_split,
-};
-#endif
-
-#include "svq3.c"
--- a/src/ffmpeg/libavcodec/h264data.h	Mon Mar 12 11:37:35 2007 -0700
+++ /dev/null	Thu Jan 01 00:00:00 1970 +0000
@@ -1,1280 +0,0 @@
-/*
- * H26L/H264/AVC/JVT/14496-10/... encoder/decoder
- * Copyright (c) 2003 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 h264data.h
- * @brief
- *     H264 / AVC / MPEG4 part10 codec data table
- * @author Michael Niedermayer <michaelni@gmx.at>
- */
-
-#define VERT_PRED             0
-#define HOR_PRED              1
-#define DC_PRED               2
-#define DIAG_DOWN_LEFT_PRED   3
-#define DIAG_DOWN_RIGHT_PRED  4
-#define VERT_RIGHT_PRED       5
-#define HOR_DOWN_PRED         6
-#define VERT_LEFT_PRED        7
-#define HOR_UP_PRED           8
-
-#define LEFT_DC_PRED          9
-#define TOP_DC_PRED           10
-#define DC_128_PRED           11
-
-
-#define DC_PRED8x8            0
-#define HOR_PRED8x8           1
-#define VERT_PRED8x8          2
-#define PLANE_PRED8x8         3
-
-#define LEFT_DC_PRED8x8       4
-#define TOP_DC_PRED8x8        5
-#define DC_128_PRED8x8        6
-
-#define EXTENDED_SAR          255
-
-static const AVRational pixel_aspect[14]={
- {0, 1},
- {1, 1},
- {12, 11},
- {10, 11},
- {16, 11},
- {40, 33},
- {24, 11},
- {20, 11},
- {32, 11},
- {80, 33},
- {18, 11},
- {15, 11},
- {64, 33},
- {160,99},
-};
-
-static const uint8_t golomb_to_pict_type[5]=
-{P_TYPE, B_TYPE, I_TYPE, SP_TYPE, SI_TYPE};
-
-static const uint8_t pict_type_to_golomb[7]=
-{-1, 2, 0, 1, -1, 4, 3};
-
-static const uint8_t chroma_qp[52]={
-    0, 1, 2, 3, 4, 5, 6, 7, 8, 9,10,11,
-   12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,
-   28,29,29,30,31,32,32,33,34,34,35,35,36,36,37,37,
-   37,38,38,38,39,39,39,39
-
-};
-
-static const uint8_t golomb_to_intra4x4_cbp[48]={
- 47, 31, 15,  0, 23, 27, 29, 30,  7, 11, 13, 14, 39, 43, 45, 46,
- 16,  3,  5, 10, 12, 19, 21, 26, 28, 35, 37, 42, 44,  1,  2,  4,
-  8, 17, 18, 20, 24,  6,  9, 22, 25, 32, 33, 34, 36, 40, 38, 41
-};
-
-static const uint8_t golomb_to_inter_cbp[48]={
-  0, 16,  1,  2,  4,  8, 32,  3,  5, 10, 12, 15, 47,  7, 11, 13,
- 14,  6,  9, 31, 35, 37, 42, 44, 33, 34, 36, 40, 39, 43, 45, 46,
- 17, 18, 20, 24, 19, 21, 26, 28, 23, 27, 29, 30, 22, 25, 38, 41
-};
-
-static const uint8_t intra4x4_cbp_to_golomb[48]={
-  3, 29, 30, 17, 31, 18, 37,  8, 32, 38, 19,  9, 20, 10, 11,  2,
- 16, 33, 34, 21, 35, 22, 39,  4, 36, 40, 23,  5, 24,  6,  7,  1,
- 41, 42, 43, 25, 44, 26, 46, 12, 45, 47, 27, 13, 28, 14, 15,  0
-};
-
-static const uint8_t inter_cbp_to_golomb[48]={
-  0,  2,  3,  7,  4,  8, 17, 13,  5, 18,  9, 14, 10, 15, 16, 11,
-  1, 32, 33, 36, 34, 37, 44, 40, 35, 45, 38, 41, 39, 42, 43, 19,
-  6, 24, 25, 20, 26, 21, 46, 28, 27, 47, 22, 29, 23, 30, 31, 12
-};
-
-static const uint8_t chroma_dc_coeff_token_len[4*5]={
- 2, 0, 0, 0,
- 6, 1, 0, 0,
- 6, 6, 3, 0,
- 6, 7, 7, 6,
- 6, 8, 8, 7,
-};
-
-static const uint8_t chroma_dc_coeff_token_bits[4*5]={
- 1, 0, 0, 0,
- 7, 1, 0, 0,
- 4, 6, 1, 0,
- 3, 3, 2, 5,
- 2, 3, 2, 0,
-};
-
-static const uint8_t coeff_token_len[4][4*17]={
-{
-     1, 0, 0, 0,
-     6, 2, 0, 0,     8, 6, 3, 0,     9, 8, 7, 5,    10, 9, 8, 6,
-    11,10, 9, 7,    13,11,10, 8,    13,13,11, 9,    13,13,13,10,
-    14,14,13,11,    14,14,14,13,    15,15,14,14,    15,15,15,14,
-    16,15,15,15,    16,16,16,15,    16,16,16,16,    16,16,16,16,
-},
-{
-     2, 0, 0, 0,
-     6, 2, 0, 0,     6, 5, 3, 0,     7, 6, 6, 4,     8, 6, 6, 4,
-     8, 7, 7, 5,     9, 8, 8, 6,    11, 9, 9, 6,    11,11,11, 7,
-    12,11,11, 9,    12,12,12,11,    12,12,12,11,    13,13,13,12,
-    13,13,13,13,    13,14,13,13,    14,14,14,13,    14,14,14,14,
-},
-{
-     4, 0, 0, 0,
-     6, 4, 0, 0,     6, 5, 4, 0,     6, 5, 5, 4,     7, 5, 5, 4,
-     7, 5, 5, 4,     7, 6, 6, 4,     7, 6, 6, 4,     8, 7, 7, 5,
-     8, 8, 7, 6,     9, 8, 8, 7,     9, 9, 8, 8,     9, 9, 9, 8,
-    10, 9, 9, 9,    10,10,10,10,    10,10,10,10,    10,10,10,10,
-},
-{
-     6, 0, 0, 0,
-     6, 6, 0, 0,     6, 6, 6, 0,     6, 6, 6, 6,     6, 6, 6, 6,
-     6, 6, 6, 6,     6, 6, 6, 6,     6, 6, 6, 6,     6, 6, 6, 6,
-     6, 6, 6, 6,     6, 6, 6, 6,     6, 6, 6, 6,     6, 6, 6, 6,
-     6, 6, 6, 6,     6, 6, 6, 6,     6, 6, 6, 6,     6, 6, 6, 6,
-}
-};
-
-static const uint8_t coeff_token_bits[4][4*17]={
-{
-     1, 0, 0, 0,
-     5, 1, 0, 0,     7, 4, 1, 0,     7, 6, 5, 3,     7, 6, 5, 3,
-     7, 6, 5, 4,    15, 6, 5, 4,    11,14, 5, 4,     8,10,13, 4,
-    15,14, 9, 4,    11,10,13,12,    15,14, 9,12,    11,10,13, 8,
-    15, 1, 9,12,    11,14,13, 8,     7,10, 9,12,     4, 6, 5, 8,
-},
-{
-     3, 0, 0, 0,
-    11, 2, 0, 0,     7, 7, 3, 0,     7,10, 9, 5,     7, 6, 5, 4,
-     4, 6, 5, 6,     7, 6, 5, 8,    15, 6, 5, 4,    11,14,13, 4,
-    15,10, 9, 4,    11,14,13,12,     8,10, 9, 8,    15,14,13,12,
-    11,10, 9,12,     7,11, 6, 8,     9, 8,10, 1,     7, 6, 5, 4,
-},
-{
-    15, 0, 0, 0,
-    15,14, 0, 0,    11,15,13, 0,     8,12,14,12,    15,10,11,11,
-    11, 8, 9,10,     9,14,13, 9,     8,10, 9, 8,    15,14,13,13,
-    11,14,10,12,    15,10,13,12,    11,14, 9,12,     8,10,13, 8,
-    13, 7, 9,12,     9,12,11,10,     5, 8, 7, 6,     1, 4, 3, 2,
-},
-{
-     3, 0, 0, 0,
-     0, 1, 0, 0,     4, 5, 6, 0,     8, 9,10,11,    12,13,14,15,
-    16,17,18,19,    20,21,22,23,    24,25,26,27,    28,29,30,31,
-    32,33,34,35,    36,37,38,39,    40,41,42,43,    44,45,46,47,
-    48,49,50,51,    52,53,54,55,    56,57,58,59,    60,61,62,63,
-}
-};
-
-static const uint8_t total_zeros_len[16][16]= {
-    {1,3,3,4,4,5,5,6,6,7,7,8,8,9,9,9},
-    {3,3,3,3,3,4,4,4,4,5,5,6,6,6,6},
-    {4,3,3,3,4,4,3,3,4,5,5,6,5,6},
-    {5,3,4,4,3,3,3,4,3,4,5,5,5},
-    {4,4,4,3,3,3,3,3,4,5,4,5},
-    {6,5,3,3,3,3,3,3,4,3,6},
-    {6,5,3,3,3,2,3,4,3,6},
-    {6,4,5,3,2,2,3,3,6},
-    {6,6,4,2,2,3,2,5},
-    {5,5,3,2,2,2,4},
-    {4,4,3,3,1,3},
-    {4,4,2,1,3},
-    {3,3,1,2},
-    {2,2,1},
-    {1,1},
-};
-
-static const uint8_t total_zeros_bits[16][16]= {
-    {1,3,2,3,2,3,2,3,2,3,2,3,2,3,2,1},
-    {7,6,5,4,3,5,4,3,2,3,2,3,2,1,0},
-    {5,7,6,5,4,3,4,3,2,3,2,1,1,0},
-    {3,7,5,4,6,5,4,3,3,2,2,1,0},
-    {5,4,3,7,6,5,4,3,2,1,1,0},
-    {1,1,7,6,5,4,3,2,1,1,0},
-    {1,1,5,4,3,3,2,1,1,0},
-    {1,1,1,3,3,2,2,1,0},
-    {1,0,1,3,2,1,1,1},
-    {1,0,1,3,2,1,1},
-    {0,1,1,2,1,3},
-    {0,1,1,1,1},
-    {0,1,1,1},
-    {0,1,1},
-    {0,1},
-};
-
-static const uint8_t chroma_dc_total_zeros_len[3][4]= {
-    { 1, 2, 3, 3,},
-    { 1, 2, 2, 0,},
-    { 1, 1, 0, 0,},
-};
-
-static const uint8_t chroma_dc_total_zeros_bits[3][4]= {
-    { 1, 1, 1, 0,},
-    { 1, 1, 0, 0,},
-    { 1, 0, 0, 0,},
-};
-
-static const uint8_t run_len[7][16]={
-    {1,1},
-    {1,2,2},
-    {2,2,2,2},
-    {2,2,2,3,3},
-    {2,2,3,3,3,3},
-    {2,3,3,3,3,3,3},
-    {3,3,3,3,3,3,3,4,5,6,7,8,9,10,11},
-};
-
-static const uint8_t run_bits[7][16]={
-    {1,0},
-    {1,1,0},
-    {3,2,1,0},
-    {3,2,1,1,0},
-    {3,2,3,2,1,0},
-    {3,0,1,3,2,5,4},
-    {7,6,5,4,3,2,1,1,1,1,1,1,1,1,1},
-};
-
-/*
-o-o o-o
- / / /
-o-o o-o
- ,---'
-o-o o-o
- / / /
-o-o o-o
-*/
-
-static const uint8_t scan8[16 + 2*4]={
- 4+1*8, 5+1*8, 4+2*8, 5+2*8,
- 6+1*8, 7+1*8, 6+2*8, 7+2*8,
- 4+3*8, 5+3*8, 4+4*8, 5+4*8,
- 6+3*8, 7+3*8, 6+4*8, 7+4*8,
- 1+1*8, 2+1*8,
- 1+2*8, 2+2*8,
- 1+4*8, 2+4*8,
- 1+5*8, 2+5*8,
-};
-
-static const uint8_t zigzag_scan[16]={
- 0+0*4, 1+0*4, 0+1*4, 0+2*4,
- 1+1*4, 2+0*4, 3+0*4, 2+1*4,
- 1+2*4, 0+3*4, 1+3*4, 2+2*4,
- 3+1*4, 3+2*4, 2+3*4, 3+3*4,
-};
-
-static const uint8_t field_scan[16]={
- 0+0*4, 0+1*4, 1+0*4, 0+2*4,
- 0+3*4, 1+1*4, 1+2*4, 1+3*4,
- 2+0*4, 2+1*4, 2+2*4, 2+3*4,
- 3+0*4, 3+1*4, 3+2*4, 3+3*4,
-};
-
-static const uint8_t luma_dc_zigzag_scan[16]={
- 0*16 + 0*64, 1*16 + 0*64, 2*16 + 0*64, 0*16 + 2*64,
- 3*16 + 0*64, 0*16 + 1*64, 1*16 + 1*64, 2*16 + 1*64,
- 1*16 + 2*64, 2*16 + 2*64, 3*16 + 2*64, 0*16 + 3*64,
- 3*16 + 1*64, 1*16 + 3*64, 2*16 + 3*64, 3*16 + 3*64,
-};
-
-static const uint8_t luma_dc_field_scan[16]={
- 0*16 + 0*64, 2*16 + 0*64, 1*16 + 0*64, 0*16 + 2*64,
- 2*16 + 2*64, 3*16 + 0*64, 1*16 + 2*64, 3*16 + 2*64,
- 0*16 + 1*64, 2*16 + 1*64, 0*16 + 3*64, 2*16 + 3*64,
- 1*16 + 1*64, 3*16 + 1*64, 1*16 + 3*64, 3*16 + 3*64,
-};
-
-static const uint8_t chroma_dc_scan[4]={
- (0+0*2)*16, (1+0*2)*16,
- (0+1*2)*16, (1+1*2)*16,  //FIXME
-};
-
-static const uint8_t zigzag_scan8x8[64]={
- 0+0*8, 1+0*8, 0+1*8, 0+2*8,
- 1+1*8, 2+0*8, 3+0*8, 2+1*8,
- 1+2*8, 0+3*8, 0+4*8, 1+3*8,
- 2+2*8, 3+1*8, 4+0*8, 5+0*8,
- 4+1*8, 3+2*8, 2+3*8, 1+4*8,
- 0+5*8, 0+6*8, 1+5*8, 2+4*8,
- 3+3*8, 4+2*8, 5+1*8, 6+0*8,
- 7+0*8, 6+1*8, 5+2*8, 4+3*8,
- 3+4*8, 2+5*8, 1+6*8, 0+7*8,
- 1+7*8, 2+6*8, 3+5*8, 4+4*8,
- 5+3*8, 6+2*8, 7+1*8, 7+2*8,
- 6+3*8, 5+4*8, 4+5*8, 3+6*8,
- 2+7*8, 3+7*8, 4+6*8, 5+5*8,
- 6+4*8, 7+3*8, 7+4*8, 6+5*8,
- 5+6*8, 4+7*8, 5+7*8, 6+6*8,
- 7+5*8, 7+6*8, 6+7*8, 7+7*8,
-};
-
-// zigzag_scan8x8_cavlc[i] = zigzag_scan8x8[(i/4) + 16*(i%4)]
-static const uint8_t zigzag_scan8x8_cavlc[64]={
- 0+0*8, 1+1*8, 1+2*8, 2+2*8,
- 4+1*8, 0+5*8, 3+3*8, 7+0*8,
- 3+4*8, 1+7*8, 5+3*8, 6+3*8,
- 2+7*8, 6+4*8, 5+6*8, 7+5*8,
- 1+0*8, 2+0*8, 0+3*8, 3+1*8,
- 3+2*8, 0+6*8, 4+2*8, 6+1*8,
- 2+5*8, 2+6*8, 6+2*8, 5+4*8,
- 3+7*8, 7+3*8, 4+7*8, 7+6*8,
- 0+1*8, 3+0*8, 0+4*8, 4+0*8,
- 2+3*8, 1+5*8, 5+1*8, 5+2*8,
- 1+6*8, 3+5*8, 7+1*8, 4+5*8,
- 4+6*8, 7+4*8, 5+7*8, 6+7*8,
- 0+2*8, 2+1*8, 1+3*8, 5+0*8,
- 1+4*8, 2+4*8, 6+0*8, 4+3*8,
- 0+7*8, 4+4*8, 7+2*8, 3+6*8,
- 5+5*8, 6+5*8, 6+6*8, 7+7*8,
-};
-
-static const uint8_t field_scan8x8[64]={
- 0+0*8, 0+1*8, 0+2*8, 1+0*8,
- 1+1*8, 0+3*8, 0+4*8, 1+2*8,
- 2+0*8, 1+3*8, 0+5*8, 0+6*8,
- 0+7*8, 1+4*8, 2+1*8, 3+0*8,
- 2+2*8, 1+5*8, 1+6*8, 1+7*8,
- 2+3*8, 3+1*8, 4+0*8, 3+2*8,
- 2+4*8, 2+5*8, 2+6*8, 2+7*8,
- 3+3*8, 4+1*8, 5+0*8, 4+2*8,
- 3+4*8, 3+5*8, 3+6*8, 3+7*8,
- 4+3*8, 5+1*8, 6+0*8, 5+2*8,
- 4+4*8, 4+5*8, 4+6*8, 4+7*8,
- 5+3*8, 6+1*8, 6+2*8, 5+4*8,
- 5+5*8, 5+6*8, 5+7*8, 6+3*8,
- 7+0*8, 7+1*8, 6+4*8, 6+5*8,
- 6+6*8, 6+7*8, 7+2*8, 7+3*8,
- 7+4*8, 7+5*8, 7+6*8, 7+7*8,
-};
-
-static const uint8_t field_scan8x8_cavlc[64]={
- 0+0*8, 1+1*8, 2+0*8, 0+7*8,
- 2+2*8, 2+3*8, 2+4*8, 3+3*8,
- 3+4*8, 4+3*8, 4+4*8, 5+3*8,
- 5+5*8, 7+0*8, 6+6*8, 7+4*8,
- 0+1*8, 0+3*8, 1+3*8, 1+4*8,
- 1+5*8, 3+1*8, 2+5*8, 4+1*8,
- 3+5*8, 5+1*8, 4+5*8, 6+1*8,
- 5+6*8, 7+1*8, 6+7*8, 7+5*8,
- 0+2*8, 0+4*8, 0+5*8, 2+1*8,
- 1+6*8, 4+0*8, 2+6*8, 5+0*8,
- 3+6*8, 6+0*8, 4+6*8, 6+2*8,
- 5+7*8, 6+4*8, 7+2*8, 7+6*8,
- 1+0*8, 1+2*8, 0+6*8, 3+0*8,
- 1+7*8, 3+2*8, 2+7*8, 4+2*8,
- 3+7*8, 5+2*8, 4+7*8, 5+4*8,
- 6+3*8, 6+5*8, 7+3*8, 7+7*8,
-};
-
-#define MB_TYPE_REF0       MB_TYPE_ACPRED //dirty but it fits in 16bit
-#define MB_TYPE_8x8DCT     0x01000000
-#define IS_REF0(a)       ((a)&MB_TYPE_REF0)
-#define IS_8x8DCT(a)     ((a)&MB_TYPE_8x8DCT)
-
-
-typedef struct IMbInfo{
-    uint16_t type;
-    uint8_t pred_mode;
-    uint8_t cbp;
-} IMbInfo;
-
-static const IMbInfo i_mb_type_info[26]={
-{MB_TYPE_INTRA4x4  , -1, -1},
-{MB_TYPE_INTRA16x16,  2,  0},
-{MB_TYPE_INTRA16x16,  1,  0},
-{MB_TYPE_INTRA16x16,  0,  0},
-{MB_TYPE_INTRA16x16,  3,  0},
-{MB_TYPE_INTRA16x16,  2,  16},
-{MB_TYPE_INTRA16x16,  1,  16},
-{MB_TYPE_INTRA16x16,  0,  16},
-{MB_TYPE_INTRA16x16,  3,  16},
-{MB_TYPE_INTRA16x16,  2,  32},
-{MB_TYPE_INTRA16x16,  1,  32},
-{MB_TYPE_INTRA16x16,  0,  32},
-{MB_TYPE_INTRA16x16,  3,  32},
-{MB_TYPE_INTRA16x16,  2,  15+0},
-{MB_TYPE_INTRA16x16,  1,  15+0},
-{MB_TYPE_INTRA16x16,  0,  15+0},
-{MB_TYPE_INTRA16x16,  3,  15+0},
-{MB_TYPE_INTRA16x16,  2,  15+16},
-{MB_TYPE_INTRA16x16,  1,  15+16},
-{MB_TYPE_INTRA16x16,  0,  15+16},
-{MB_TYPE_INTRA16x16,  3,  15+16},
-{MB_TYPE_INTRA16x16,  2,  15+32},
-{MB_TYPE_INTRA16x16,  1,  15+32},
-{MB_TYPE_INTRA16x16,  0,  15+32},
-{MB_TYPE_INTRA16x16,  3,  15+32},
-{MB_TYPE_INTRA_PCM , -1, -1},
-};
-
-typedef struct PMbInfo{
-    uint16_t type;
-    uint8_t partition_count;
-} PMbInfo;
-
-static const PMbInfo p_mb_type_info[5]={
-{MB_TYPE_16x16|MB_TYPE_P0L0             , 1},
-{MB_TYPE_16x8 |MB_TYPE_P0L0|MB_TYPE_P1L0, 2},
-{MB_TYPE_8x16 |MB_TYPE_P0L0|MB_TYPE_P1L0, 2},
-{MB_TYPE_8x8  |MB_TYPE_P0L0|MB_TYPE_P1L0, 4},
-{MB_TYPE_8x8  |MB_TYPE_P0L0|MB_TYPE_P1L0|MB_TYPE_REF0, 4},
-};
-
-static const PMbInfo p_sub_mb_type_info[4]={
-{MB_TYPE_16x16|MB_TYPE_P0L0             , 1},
-{MB_TYPE_16x8 |MB_TYPE_P0L0             , 2},
-{MB_TYPE_8x16 |MB_TYPE_P0L0             , 2},
-{MB_TYPE_8x8  |MB_TYPE_P0L0             , 4},
-};
-
-static const PMbInfo b_mb_type_info[23]={
-{MB_TYPE_DIRECT2                                                   , 1, },
-{MB_TYPE_16x16|MB_TYPE_P0L0                                       , 1, },
-{MB_TYPE_16x16             |MB_TYPE_P0L1                          , 1, },
-{MB_TYPE_16x16|MB_TYPE_P0L0|MB_TYPE_P0L1                          , 1, },
-{MB_TYPE_16x8 |MB_TYPE_P0L0             |MB_TYPE_P1L0             , 2, },
-{MB_TYPE_8x16 |MB_TYPE_P0L0             |MB_TYPE_P1L0             , 2, },
-{MB_TYPE_16x8              |MB_TYPE_P0L1             |MB_TYPE_P1L1, 2, },
-{MB_TYPE_8x16              |MB_TYPE_P0L1             |MB_TYPE_P1L1, 2, },
-{MB_TYPE_16x8 |MB_TYPE_P0L0                          |MB_TYPE_P1L1, 2, },
-{MB_TYPE_8x16 |MB_TYPE_P0L0                          |MB_TYPE_P1L1, 2, },
-{MB_TYPE_16x8              |MB_TYPE_P0L1|MB_TYPE_P1L0             , 2, },
-{MB_TYPE_8x16              |MB_TYPE_P0L1|MB_TYPE_P1L0             , 2, },
-{MB_TYPE_16x8 |MB_TYPE_P0L0             |MB_TYPE_P1L0|MB_TYPE_P1L1, 2, },
-{MB_TYPE_8x16 |MB_TYPE_P0L0             |MB_TYPE_P1L0|MB_TYPE_P1L1, 2, },
-{MB_TYPE_16x8              |MB_TYPE_P0L1|MB_TYPE_P1L0|MB_TYPE_P1L1, 2, },
-{MB_TYPE_8x16              |MB_TYPE_P0L1|MB_TYPE_P1L0|MB_TYPE_P1L1, 2, },
-{MB_TYPE_16x8 |MB_TYPE_P0L0|MB_TYPE_P0L1|MB_TYPE_P1L0             , 2, },
-{MB_TYPE_8x16 |MB_TYPE_P0L0|MB_TYPE_P0L1|MB_TYPE_P1L0             , 2, },
-{MB_TYPE_16x8 |MB_TYPE_P0L0|MB_TYPE_P0L1             |MB_TYPE_P1L1, 2, },
-{MB_TYPE_8x16 |MB_TYPE_P0L0|MB_TYPE_P0L1             |MB_TYPE_P1L1, 2, },
-{MB_TYPE_16x8 |MB_TYPE_P0L0|MB_TYPE_P0L1|MB_TYPE_P1L0|MB_TYPE_P1L1, 2, },
-{MB_TYPE_8x16 |MB_TYPE_P0L0|MB_TYPE_P0L1|MB_TYPE_P1L0|MB_TYPE_P1L1, 2, },
-{MB_TYPE_8x8  |MB_TYPE_P0L0|MB_TYPE_P0L1|MB_TYPE_P1L0|MB_TYPE_P1L1, 4, },
-};
-
-static const PMbInfo b_sub_mb_type_info[13]={
-{MB_TYPE_DIRECT2                                                   , 1, },
-{MB_TYPE_16x16|MB_TYPE_P0L0                                       , 1, },
-{MB_TYPE_16x16             |MB_TYPE_P0L1                          , 1, },
-{MB_TYPE_16x16|MB_TYPE_P0L0|MB_TYPE_P0L1                          , 1, },
-{MB_TYPE_16x8 |MB_TYPE_P0L0             |MB_TYPE_P1L0             , 2, },
-{MB_TYPE_8x16 |MB_TYPE_P0L0             |MB_TYPE_P1L0             , 2, },
-{MB_TYPE_16x8              |MB_TYPE_P0L1             |MB_TYPE_P1L1, 2, },
-{MB_TYPE_8x16              |MB_TYPE_P0L1             |MB_TYPE_P1L1, 2, },
-{MB_TYPE_16x8 |MB_TYPE_P0L0|MB_TYPE_P0L1|MB_TYPE_P1L0|MB_TYPE_P1L1, 2, },
-{MB_TYPE_8x16 |MB_TYPE_P0L0|MB_TYPE_P0L1|MB_TYPE_P1L0|MB_TYPE_P1L1, 2, },
-{MB_TYPE_8x8  |MB_TYPE_P0L0             |MB_TYPE_P1L0             , 4, },
-{MB_TYPE_8x8               |MB_TYPE_P0L1             |MB_TYPE_P1L1, 4, },
-{MB_TYPE_8x8  |MB_TYPE_P0L0|MB_TYPE_P0L1|MB_TYPE_P1L0|MB_TYPE_P1L1, 4, },
-};
-
-
-static const uint8_t rem6[52]={
-0, 1, 2, 3, 4, 5, 0, 1, 2, 3, 4, 5, 0, 1, 2, 3, 4, 5, 0, 1, 2, 3, 4, 5, 0, 1, 2, 3, 4, 5, 0, 1, 2, 3, 4, 5, 0, 1, 2, 3, 4, 5, 0, 1, 2, 3, 4, 5, 0, 1, 2, 3,
-};
-
-static const uint8_t div6[52]={
-0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 3, 3, 3, 3, 3, 3, 4, 4, 4, 4, 4, 4, 5, 5, 5, 5, 5, 5, 6, 6, 6, 6, 6, 6, 7, 7, 7, 7, 7, 7, 8, 8, 8, 8,
-};
-
-static const uint8_t default_scaling4[2][16]={
-{   6,13,20,28,
-   13,20,28,32,
-   20,28,32,37,
-   28,32,37,42
-},{
-   10,14,20,24,
-   14,20,24,27,
-   20,24,27,30,
-   24,27,30,34
-}};
-
-static const uint8_t default_scaling8[2][64]={
-{   6,10,13,16,18,23,25,27,
-   10,11,16,18,23,25,27,29,
-   13,16,18,23,25,27,29,31,
-   16,18,23,25,27,29,31,33,
-   18,23,25,27,29,31,33,36,
-   23,25,27,29,31,33,36,38,
-   25,27,29,31,33,36,38,40,
-   27,29,31,33,36,38,40,42
-},{
-    9,13,15,17,19,21,22,24,
-   13,13,17,19,21,22,24,25,
-   15,17,19,21,22,24,25,27,
-   17,19,21,22,24,25,27,28,
-   19,21,22,24,25,27,28,30,
-   21,22,24,25,27,28,30,32,
-   22,24,25,27,28,30,32,33,
-   24,25,27,28,30,32,33,35
-}};
-
-static const int dequant4_coeff_init[6][3]={
-  {10,13,16},
-  {11,14,18},
-  {13,16,20},
-  {14,18,23},
-  {16,20,25},
-  {18,23,29},
-};
-
-static const int dequant8_coeff_init_scan[16] = {
-  0,3,4,3, 3,1,5,1, 4,5,2,5, 3,1,5,1
-};
-static const int dequant8_coeff_init[6][6]={
-  {20,18,32,19,25,24},
-  {22,19,35,21,28,26},
-  {26,23,42,24,33,31},
-  {28,25,45,26,35,33},
-  {32,28,51,30,40,38},
-  {36,32,58,34,46,43},
-};
-
-#define QUANT_SHIFT 22
-
-static const int quant_coeff[52][16]={
-    { 419430,258111,419430,258111,258111,167772,258111,167772,419430,258111,419430,258111,258111,167772,258111,167772,},
-    { 381300,239675,381300,239675,239675,149131,239675,149131,381300,239675,381300,239675,239675,149131,239675,149131,},
-    { 322639,209715,322639,209715,209715,134218,209715,134218,322639,209715,322639,209715,209715,134218,209715,134218,},
-    { 299593,186414,299593,186414,186414,116711,186414,116711,299593,186414,299593,186414,186414,116711,186414,116711,},
-    { 262144,167772,262144,167772,167772,107374,167772,107374,262144,167772,262144,167772,167772,107374,167772,107374,},
-    { 233017,145889,233017,145889,145889, 92564,145889, 92564,233017,145889,233017,145889,145889, 92564,145889, 92564,},
-    { 209715,129056,209715,129056,129056, 83886,129056, 83886,209715,129056,209715,129056,129056, 83886,129056, 83886,},
-    { 190650,119837,190650,119837,119837, 74565,119837, 74565,190650,119837,190650,119837,119837, 74565,119837, 74565,},
-    { 161319,104858,161319,104858,104858, 67109,104858, 67109,161319,104858,161319,104858,104858, 67109,104858, 67109,},
-    { 149797, 93207,149797, 93207, 93207, 58356, 93207, 58356,149797, 93207,149797, 93207, 93207, 58356, 93207, 58356,},
-    { 131072, 83886,131072, 83886, 83886, 53687, 83886, 53687,131072, 83886,131072, 83886, 83886, 53687, 83886, 53687,},
-    { 116508, 72944,116508, 72944, 72944, 46282, 72944, 46282,116508, 72944,116508, 72944, 72944, 46282, 72944, 46282,},
-    { 104858, 64528,104858, 64528, 64528, 41943, 64528, 41943,104858, 64528,104858, 64528, 64528, 41943, 64528, 41943,},
-    {  95325, 59919, 95325, 59919, 59919, 37283, 59919, 37283, 95325, 59919, 95325, 59919, 59919, 37283, 59919, 37283,},
-    {  80660, 52429, 80660, 52429, 52429, 33554, 52429, 33554, 80660, 52429, 80660, 52429, 52429, 33554, 52429, 33554,},
-    {  74898, 46603, 74898, 46603, 46603, 29178, 46603, 29178, 74898, 46603, 74898, 46603, 46603, 29178, 46603, 29178,},
-    {  65536, 41943, 65536, 41943, 41943, 26844, 41943, 26844, 65536, 41943, 65536, 41943, 41943, 26844, 41943, 26844,},
-    {  58254, 36472, 58254, 36472, 36472, 23141, 36472, 23141, 58254, 36472, 58254, 36472, 36472, 23141, 36472, 23141,},
-    {  52429, 32264, 52429, 32264, 32264, 20972, 32264, 20972, 52429, 32264, 52429, 32264, 32264, 20972, 32264, 20972,},
-    {  47663, 29959, 47663, 29959, 29959, 18641, 29959, 18641, 47663, 29959, 47663, 29959, 29959, 18641, 29959, 18641,},
-    {  40330, 26214, 40330, 26214, 26214, 16777, 26214, 16777, 40330, 26214, 40330, 26214, 26214, 16777, 26214, 16777,},
-    {  37449, 23302, 37449, 23302, 23302, 14589, 23302, 14589, 37449, 23302, 37449, 23302, 23302, 14589, 23302, 14589,},
-    {  32768, 20972, 32768, 20972, 20972, 13422, 20972, 13422, 32768, 20972, 32768, 20972, 20972, 13422, 20972, 13422,},
-    {  29127, 18236, 29127, 18236, 18236, 11570, 18236, 11570, 29127, 18236, 29127, 18236, 18236, 11570, 18236, 11570,},
-    {  26214, 16132, 26214, 16132, 16132, 10486, 16132, 10486, 26214, 16132, 26214, 16132, 16132, 10486, 16132, 10486,},
-    {  23831, 14980, 23831, 14980, 14980,  9321, 14980,  9321, 23831, 14980, 23831, 14980, 14980,  9321, 14980,  9321,},
-    {  20165, 13107, 20165, 13107, 13107,  8389, 13107,  8389, 20165, 13107, 20165, 13107, 13107,  8389, 13107,  8389,},
-    {  18725, 11651, 18725, 11651, 11651,  7294, 11651,  7294, 18725, 11651, 18725, 11651, 11651,  7294, 11651,  7294,},
-    {  16384, 10486, 16384, 10486, 10486,  6711, 10486,  6711, 16384, 10486, 16384, 10486, 10486,  6711, 10486,  6711,},
-    {  14564,  9118, 14564,  9118,  9118,  5785,  9118,  5785, 14564,  9118, 14564,  9118,  9118,  5785,  9118,  5785,},
-    {  13107,  8066, 13107,  8066,  8066,  5243,  8066,  5243, 13107,  8066, 13107,  8066,  8066,  5243,  8066,  5243,},
-    {  11916,  7490, 11916,  7490,  7490,  4660,  7490,  4660, 11916,  7490, 11916,  7490,  7490,  4660,  7490,  4660,},
-    {  10082,  6554, 10082,  6554,  6554,  4194,  6554,  4194, 10082,  6554, 10082,  6554,  6554,  4194,  6554,  4194,},
-    {   9362,  5825,  9362,  5825,  5825,  3647,  5825,  3647,  9362,  5825,  9362,  5825,  5825,  3647,  5825,  3647,},
-    {   8192,  5243,  8192,  5243,  5243,  3355,  5243,  3355,  8192,  5243,  8192,  5243,  5243,  3355,  5243,  3355,},
-    {   7282,  4559,  7282,  4559,  4559,  2893,  4559,  2893,  7282,  4559,  7282,  4559,  4559,  2893,  4559,  2893,},
-    {   6554,  4033,  6554,  4033,  4033,  2621,  4033,  2621,  6554,  4033,  6554,  4033,  4033,  2621,  4033,  2621,},
-    {   5958,  3745,  5958,  3745,  3745,  2330,  3745,  2330,  5958,  3745,  5958,  3745,  3745,  2330,  3745,  2330,},
-    {   5041,  3277,  5041,  3277,  3277,  2097,  3277,  2097,  5041,  3277,  5041,  3277,  3277,  2097,  3277,  2097,},
-    {   4681,  2913,  4681,  2913,  2913,  1824,  2913,  1824,  4681,  2913,  4681,  2913,  2913,  1824,  2913,  1824,},
-    {   4096,  2621,  4096,  2621,  2621,  1678,  2621,  1678,  4096,  2621,  4096,  2621,  2621,  1678,  2621,  1678,},
-    {   3641,  2280,  3641,  2280,  2280,  1446,  2280,  1446,  3641,  2280,  3641,  2280,  2280,  1446,  2280,  1446,},
-    {   3277,  2016,  3277,  2016,  2016,  1311,  2016,  1311,  3277,  2016,  3277,  2016,  2016,  1311,  2016,  1311,},
-    {   2979,  1872,  2979,  1872,  1872,  1165,  1872,  1165,  2979,  1872,  2979,  1872,  1872,  1165,  1872,  1165,},
-    {   2521,  1638,  2521,  1638,  1638,  1049,  1638,  1049,  2521,  1638,  2521,  1638,  1638,  1049,  1638,  1049,},
-    {   2341,  1456,  2341,  1456,  1456,   912,  1456,   912,  2341,  1456,  2341,  1456,  1456,   912,  1456,   912,},
-    {   2048,  1311,  2048,  1311,  1311,   839,  1311,   839,  2048,  1311,  2048,  1311,  1311,   839,  1311,   839,},
-    {   1820,  1140,  1820,  1140,  1140,   723,  1140,   723,  1820,  1140,  1820,  1140,  1140,   723,  1140,   723,},
-    {   1638,  1008,  1638,  1008,  1008,   655,  1008,   655,  1638,  1008,  1638,  1008,  1008,   655,  1008,   655,},
-    {   1489,   936,  1489,   936,   936,   583,   936,   583,  1489,   936,  1489,   936,   936,   583,   936,   583,},
-    {   1260,   819,  1260,   819,   819,   524,   819,   524,  1260,   819,  1260,   819,   819,   524,   819,   524,},
-    {   1170,   728,  1170,   728,   728,   456,   728,   456,  1170,   728,  1170,   728,   728,   456,   728,   456,},
-};
-
-
-/* Deblocking filter (p153) */
-static const int alpha_table[52] = {
-     0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
-     0,  0,  0,  0,  0,  0,  4,  4,  5,  6,
-     7,  8,  9, 10, 12, 13, 15, 17, 20, 22,
-    25, 28, 32, 36, 40, 45, 50, 56, 63, 71,
-    80, 90,101,113,127,144,162,182,203,226,
-    255, 255
-};
-static const int beta_table[52] = {
-     0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
-     0,  0,  0,  0,  0,  0,  2,  2,  2,  3,
-     3,  3,  3,  4,  4,  4,  6,  6,  7,  7,
-     8,  8,  9,  9, 10, 10, 11, 11, 12, 12,
-    13, 13, 14, 14, 15, 15, 16, 16, 17, 17,
-    18, 18
-};
-static const int tc0_table[52][3] = {
-    { 0, 0, 0 }, { 0, 0, 0 }, { 0, 0, 0 }, { 0, 0, 0 }, { 0, 0, 0 }, { 0, 0, 0 },
-    { 0, 0, 0 }, { 0, 0, 0 }, { 0, 0, 0 }, { 0, 0, 0 }, { 0, 0, 0 }, { 0, 0, 0 },
-    { 0, 0, 0 }, { 0, 0, 0 }, { 0, 0, 0 }, { 0, 0, 0 }, { 0, 0, 0 }, { 0, 0, 1 },
-    { 0, 0, 1 }, { 0, 0, 1 }, { 0, 0, 1 }, { 0, 1, 1 }, { 0, 1, 1 }, { 1, 1, 1 },
-    { 1, 1, 1 }, { 1, 1, 1 }, { 1, 1, 1 }, { 1, 1, 2 }, { 1, 1, 2 }, { 1, 1, 2 },
-    { 1, 1, 2 }, { 1, 2, 3 }, { 1, 2, 3 }, { 2, 2, 3 }, { 2, 2, 4 }, { 2, 3, 4 },
-    { 2, 3, 4 }, { 3, 3, 5 }, { 3, 4, 6 }, { 3, 4, 6 }, { 4, 5, 7 }, { 4, 5, 8 },
-    { 4, 6, 9 }, { 5, 7,10 }, { 6, 8,11 }, { 6, 8,13 }, { 7,10,14 }, { 8,11,16 },
-    { 9,12,18 }, {10,13,20 }, {11,15,23 }, {13,17,25 }
-};
-
-/* Cabac pre state table */
-
-static const int cabac_context_init_I[460][2] =
-{
-    /* 0 - 10 */
-    { 20, -15 }, {  2, 54 },  {  3,  74 }, { 20, -15 },
-    {  2,  54 }, {  3, 74 },  { -28,127 }, { -23, 104 },
-    { -6,  53 }, { -1, 54 },  {  7,  51 },
-
-    /* 11 - 23 unsused for I */
-    { 0, 0 },    { 0, 0 },    { 0, 0 },      { 0, 0 },
-    { 0, 0 },    { 0, 0 },    { 0, 0 },      { 0, 0 },
-    { 0, 0 },    { 0, 0 },    { 0, 0 },      { 0, 0 },
-    { 0, 0 },
-
-    /* 24- 39 */
-    { 0, 0 },    { 0, 0 },    { 0, 0 },      { 0, 0 },
-    { 0, 0 },    { 0, 0 },    { 0, 0 },      { 0, 0 },
-    { 0, 0 },    { 0, 0 },    { 0, 0 },      { 0, 0 },
-    { 0, 0 },    { 0, 0 },    { 0, 0 },      { 0, 0 },
-
-    /* 40 - 53 */
-    { 0, 0 },    { 0, 0 },    { 0, 0 },      { 0, 0 },
-    { 0, 0 },    { 0, 0 },    { 0, 0 },      { 0, 0 },
-    { 0, 0 },    { 0, 0 },    { 0, 0 },      { 0, 0 },
-    { 0, 0 },    { 0, 0 },
-
-    /* 54 - 59 */
-    { 0, 0 },    { 0, 0 },    { 0, 0 },      { 0, 0 },
-    { 0, 0 },    { 0, 0 },
-
-    /* 60 - 69 */
-    { 0, 41 },   { 0, 63 },   { 0, 63 },     { 0, 63 },
-    { -9, 83 },  { 4, 86 },   { 0, 97 },     { -7, 72 },
-    { 13, 41 },  { 3, 62 },
-
-    /* 70 -> 87 */
-    { 0, 11 },   { 1, 55 },   { 0, 69 },     { -17, 127 },
-    { -13, 102 },{ 0, 82 },   { -7, 74 },    { -21, 107 },
-    { -27, 127 },{ -31, 127 },{ -24, 127 },  { -18, 95 },
-    { -27, 127 },{ -21, 114 },{ -30, 127 },  { -17, 123 },
-    { -12, 115 },{ -16, 122 },
-
-    /* 88 -> 104 */
-    { -11, 115 },{ -12, 63 }, { -2, 68 },    { -15, 84 },
-    { -13, 104 },{ -3, 70 },  { -8, 93 },    { -10, 90 },
-    { -30, 127 },{ -1, 74 },  { -6, 97 },    { -7, 91 },
-    { -20, 127 },{ -4, 56 },  { -5, 82 },    { -7, 76 },
-    { -22, 125 },
-
-    /* 105 -> 135 */
-    { -7, 93 },  { -11, 87 }, { -3, 77 },    { -5, 71 },
-    { -4, 63 },  { -4, 68 },  { -12, 84 },   { -7, 62 },
-    { -7, 65 },  { 8, 61 },   { 5, 56 },     { -2, 66 },
-    { 1, 64 },   { 0, 61 },   { -2, 78 },    { 1, 50 },
-    { 7, 52 },   { 10, 35 },  { 0, 44 },     { 11, 38 },
-    { 1, 45 },   { 0, 46 },   { 5, 44 },     { 31, 17 },
-    { 1, 51 },   { 7, 50 },   { 28, 19 },    { 16, 33 },
-    { 14, 62 },  { -13, 108 },{ -15, 100 },
-
-    /* 136 -> 165 */
-    { -13, 101 },{ -13, 91 }, { -12, 94 },   { -10, 88 },
-    { -16, 84 }, { -10, 86 }, { -7, 83 },    { -13, 87 },
-    { -19, 94 }, { 1, 70 },   { 0, 72 },     { -5, 74 },
-    { 18, 59 },  { -8, 102 }, { -15, 100 },  { 0, 95 },
-    { -4, 75 },  { 2, 72 },   { -11, 75 },   { -3, 71 },
-    { 15, 46 },  { -13, 69 }, { 0, 62 },     { 0, 65 },
-    { 21, 37 },  { -15, 72 }, { 9, 57 },     { 16, 54 },
-    { 0, 62 },   { 12, 72 },
-
-    /* 166 -> 196 */
-    { 24, 0 },   { 15, 9 },   { 8, 25 },     { 13, 18 },
-    { 15, 9 },   { 13, 19 },  { 10, 37 },    { 12, 18 },
-    { 6, 29 },   { 20, 33 },  { 15, 30 },    { 4, 45 },
-    { 1, 58 },   { 0, 62 },   { 7, 61 },     { 12, 38 },
-    { 11, 45 },  { 15, 39 },  { 11, 42 },    { 13, 44 },
-    { 16, 45 },  { 12, 41 },  { 10, 49 },    { 30, 34 },
-    { 18, 42 },  { 10, 55 },  { 17, 51 },    { 17, 46 },
-    { 0, 89 },   { 26, -19 }, { 22, -17 },
-
-    /* 197 -> 226 */
-    { 26, -17 }, { 30, -25 }, { 28, -20 },   { 33, -23 },
-    { 37, -27 }, { 33, -23 }, { 40, -28 },   { 38, -17 },
-    { 33, -11 }, { 40, -15 }, { 41, -6 },    { 38, 1 },
-    { 41, 17 },  { 30, -6 },  { 27, 3 },     { 26, 22 },
-    { 37, -16 }, { 35, -4 },  { 38, -8 },    { 38, -3 },
-    { 37, 3 },   { 38, 5 },   { 42, 0 },     { 35, 16 },
-    { 39, 22 },  { 14, 48 },  { 27, 37 },    { 21, 60 },
-    { 12, 68 },  { 2, 97 },
-
-    /* 227 -> 251 */
-    { -3, 71 },  { -6, 42 },  { -5, 50 },    { -3, 54 },
-    { -2, 62 },  { 0, 58 },   { 1, 63 },     { -2, 72 },
-    { -1, 74 },  { -9, 91 },  { -5, 67 },    { -5, 27 },
-    { -3, 39 },  { -2, 44 },  { 0, 46 },     { -16, 64 },
-    { -8, 68 },  { -10, 78 }, { -6, 77 },    { -10, 86 },
-    { -12, 92 }, { -15, 55 }, { -10, 60 },   { -6, 62 },
-    { -4, 65 },
-
-    /* 252 -> 275 */
-    { -12, 73 }, { -8, 76 },  { -7, 80 },    { -9, 88 },
-    { -17, 110 },{ -11, 97 }, { -20, 84 },   { -11, 79 },
-    { -6, 73 },  { -4, 74 },  { -13, 86 },   { -13, 96 },
-    { -11, 97 }, { -19, 117 },{ -8, 78 },    { -5, 33 },
-    { -4, 48 },  { -2, 53 },  { -3, 62 },    { -13, 71 },
-    { -10, 79 }, { -12, 86 }, { -13, 90 },   { -14, 97 },
-
-    /* 276 a bit special (not used, bypass is used instead) */
-    { 0, 0 },
-
-    /* 277 -> 307 */
-    { -6, 93 },  { -6, 84 },  { -8, 79 },    { 0, 66 },
-    { -1, 71 },  { 0, 62 },   { -2, 60 },    { -2, 59 },
-    { -5, 75 },  { -3, 62 },  { -4, 58 },    { -9, 66 },
-    { -1, 79 },  { 0, 71 },   { 3, 68 },     { 10, 44 },
-    { -7, 62 },  { 15, 36 },  { 14, 40 },    { 16, 27 },
-    { 12, 29 },  { 1, 44 },   { 20, 36 },    { 18, 32 },
-    { 5, 42 },   { 1, 48 },   { 10, 62 },    { 17, 46 },
-    { 9, 64 },   { -12, 104 },{ -11, 97 },
-
-    /* 308 -> 337 */
-    { -16, 96 }, { -7, 88 },  { -8, 85 },    { -7, 85 },
-    { -9, 85 },  { -13, 88 }, { 4, 66 },     { -3, 77 },
-    { -3, 76 },  { -6, 76 },  { 10, 58 },    { -1, 76 },
-    { -1, 83 },  { -7, 99 },  { -14, 95 },   { 2, 95 },
-    { 0, 76 },   { -5, 74 },  { 0, 70 },     { -11, 75 },
-    { 1, 68 },   { 0, 65 },   { -14, 73 },   { 3, 62 },
-    { 4, 62 },   { -1, 68 },  { -13, 75 },   { 11, 55 },
-    { 5, 64 },   { 12, 70 },
-
-    /* 338 -> 368 */
-    { 15, 6 },   { 6, 19 },   { 7, 16 },     { 12, 14 },
-    { 18, 13 },  { 13, 11 },  { 13, 15 },    { 15, 16 },
-    { 12, 23 },  { 13, 23 },  { 15, 20 },    { 14, 26 },
-    { 14, 44 },  { 17, 40 },  { 17, 47 },    { 24, 17 },
-    { 21, 21 },  { 25, 22 },  { 31, 27 },    { 22, 29 },
-    { 19, 35 },  { 14, 50 },  { 10, 57 },    { 7, 63 },
-    { -2, 77 },  { -4, 82 },  { -3, 94 },    { 9, 69 },
-    { -12, 109 },{ 36, -35 }, { 36, -34 },
-
-    /* 369 -> 398 */
-    { 32, -26 }, { 37, -30 }, { 44, -32 },   { 34, -18 },
-    { 34, -15 }, { 40, -15 }, { 33, -7 },    { 35, -5 },
-    { 33, 0 },   { 38, 2 },   { 33, 13 },    { 23, 35 },
-    { 13, 58 },  { 29, -3 },  { 26, 0 },     { 22, 30 },
-    { 31, -7 },  { 35, -15 }, { 34, -3 },    { 34, 3 },
-    { 36, -1 },  { 34, 5 },   { 32, 11 },    { 35, 5 },
-    { 34, 12 },  { 39, 11 },  { 30, 29 },    { 34, 26 },
-    { 29, 39 },  { 19, 66 },
-
-    /* 399 -> 435 */
-    {  31,  21 }, {  31,  31 }, {  25,  50 },
-    { -17, 120 }, { -20, 112 }, { -18, 114 }, { -11,  85 },
-    { -15,  92 }, { -14,  89 }, { -26,  71 }, { -15,  81 },
-    { -14,  80 }, {   0,  68 }, { -14,  70 }, { -24,  56 },
-    { -23,  68 }, { -24,  50 }, { -11,  74 }, {  23, -13 },
-    {  26, -13 }, {  40, -15 }, {  49, -14 }, {  44,   3 },
-    {  45,   6 }, {  44,  34 }, {  33,  54 }, {  19,  82 },
-    {  -3,  75 }, {  -1,  23 }, {   1,  34 }, {   1,  43 },
-    {   0,  54 }, {  -2,  55 }, {   0,  61 }, {   1,  64 },
-    {   0,  68 }, {  -9,  92 },
-
-    /* 436 -> 459 */
-    { -14, 106 }, { -13,  97 }, { -15,  90 }, { -12,  90 },
-    { -18,  88 }, { -10,  73 }, {  -9,  79 }, { -14,  86 },
-    { -10,  73 }, { -10,  70 }, { -10,  69 }, {  -5,  66 },
-    {  -9,  64 }, {  -5,  58 }, {   2,  59 }, {  21, -10 },
-    {  24, -11 }, {  28,  -8 }, {  28,  -1 }, {  29,   3 },
-    {  29,   9 }, {  35,  20 }, {  29,  36 }, {  14,  67 }
-};
-
-static const int cabac_context_init_PB[3][460][2] =
-{
-    /* i_cabac_init_idc == 0 */
-    {
-        /* 0 - 10 */
-        {  20, -15 }, {   2,  54 }, {   3,  74 }, {  20, -15 },
-        {   2,  54 }, {   3,  74 }, { -28, 127 }, { -23, 104 },
-        {  -6,  53 }, {  -1,  54 }, {   7,  51 },
-
-        /* 11 - 23 */
-        {  23,  33 }, {  23,   2 }, {  21,   0 }, {   1,   9 },
-        {   0,  49 }, { -37, 118 }, {   5,  57 }, { -13,  78 },
-        { -11,  65 }, {   1,  62 }, {  12,  49 }, {  -4,  73 },
-        {  17,  50 },
-
-        /* 24 - 39 */
-        {  18,  64 }, {   9,  43 }, {  29,   0 }, {  26,  67 },
-        {  16,  90 }, {   9, 104 }, { -46, 127 }, { -20, 104 },
-        {   1,  67 }, { -13,  78 }, { -11,  65 }, {   1,  62 },
-        {  -6,  86 }, { -17,  95 }, {  -6,  61 }, {   9,  45 },
-
-        /* 40 - 53 */
-        {  -3,  69 }, {  -6,  81 }, { -11,  96 }, {   6,  55 },
-        {   7,  67 }, {  -5,  86 }, {   2,  88 }, {   0,  58 },
-        {  -3,  76 }, { -10,  94 }, {   5,  54 }, {   4,  69 },
-        {  -3,  81 }, {   0,  88 },
-
-        /* 54 - 59 */
-        {  -7,  67 }, {  -5,  74 }, {  -4,  74 }, {  -5,  80 },
-        {  -7,  72 }, {   1,  58 },
-
-        /* 60 - 69 */
-        {   0,  41 }, {   0,  63 }, {   0,  63 }, { 0, 63 },
-        {  -9,  83 }, {   4,  86 }, {   0,  97 }, { -7, 72 },
-        {  13,  41 }, {   3,  62 },
-
-        /* 70 - 87 */
-        {   0,  45 }, {  -4,  78 }, {  -3,  96 }, { -27,  126 },
-        { -28,  98 }, { -25, 101 }, { -23,  67 }, { -28,  82 },
-        { -20,  94 }, { -16,  83 }, { -22, 110 }, { -21,  91 },
-        { -18, 102 }, { -13,  93 }, { -29, 127 }, {  -7,  92 },
-        {  -5,  89 }, {  -7,  96 }, { -13, 108 }, {  -3,  46 },
-        {  -1,  65 }, {  -1,  57 }, {  -9,  93 }, {  -3,  74 },
-        {  -9,  92 }, {  -8,  87 }, { -23, 126 }, {   5,  54 },
-        {   6,  60 }, {   6,  59 }, {   6,  69 }, {  -1,  48 },
-        {   0,  68 }, {  -4,  69 }, {  -8,  88 },
-
-        /* 105 -> 165 */
-        {  -2,  85 }, {  -6,  78 }, {  -1,  75 }, {  -7,  77 },
-        {   2,  54 }, {   5,  50 }, {  -3,  68 }, {   1,  50 },
-        {   6,  42 }, {  -4,  81 }, {   1,  63 }, {  -4,  70 },
-        {   0,  67 }, {   2,  57 }, {  -2,  76 }, {  11,  35 },
-        {   4,  64 }, {   1,  61 }, {  11,  35 }, {  18,  25 },
-        {  12,  24 }, {  13,  29 }, {  13,  36 }, { -10,  93 },
-        {  -7,  73 }, {  -2,  73 }, {  13,  46 }, {   9,  49 },
-        {  -7, 100 }, {   9,  53 }, {   2,  53 }, {   5,  53 },
-        {  -2,  61 }, {   0,  56 }, {   0,  56 }, { -13,  63 },
-        {  -5,  60 }, {  -1,  62 }, {   4,  57 }, {  -6,  69 },
-        {   4,  57 }, {  14,  39 }, {   4,  51 }, {  13,  68 },
-        {   3,  64 }, {   1,  61 }, {   9,  63 }, {   7,  50 },
-        {  16,  39 }, {   5,  44 }, {   4,  52 }, {  11,  48 },
-        {  -5,  60 }, {  -1,  59 }, {   0,  59 }, {  22,  33 },
-        {   5,  44 }, {  14,  43 }, {  -1,  78 }, {   0,  60 },
-        {   9,  69 },
-
-        /* 166 - 226 */
-        {  11,  28 }, {   2,  40 }, {   3,  44 }, {   0,  49 },
-        {   0,  46 }, {   2,  44 }, {   2,  51 }, {   0,  47 },
-        {   4,  39 }, {   2,  62 }, {   6,  46 }, {   0,  54 },
-        {   3,  54 }, {   2,  58 }, {   4,  63 }, {   6,  51 },
-        {   6,  57 }, {   7,  53 }, {   6,  52 }, {   6,  55 },
-        {  11,  45 }, {  14,  36 }, {   8,  53 }, {  -1,  82 },
-        {   7,  55 }, {  -3,  78 }, {  15,  46 }, {  22,  31 },
-        {  -1,  84 }, {  25,   7 }, {  30,  -7 }, {  28,   3 },
-        {  28,   4 }, {  32,   0 }, {  34,  -1 }, {  30,   6 },
-        {  30,   6 }, {  32,   9 }, {  31,  19 }, {  26,  27 },
-        {  26,  30 }, {  37,  20 }, {  28,  34 }, {  17,  70 },
-        {   1,  67 }, {   5,  59 }, {   9,  67 }, {  16,  30 },
-        {  18,  32 }, {  18,  35 }, {  22,  29 }, {  24,  31 },
-        {  23,  38 }, {  18,  43 }, {  20,  41 }, {  11,  63 },
-        {   9,  59 }, {   9,  64 }, {  -1,  94 }, {  -2,  89 },
-        {  -9, 108 },
-
-        /* 227 - 275 */
-        {  -6,  76 }, {  -2,  44 }, {   0,  45 }, {   0,  52 },
-        {  -3,  64 }, {  -2,  59 }, {  -4,  70 }, {  -4,  75 },
-        {  -8,  82 }, { -17, 102 }, {  -9,  77 }, {   3,  24 },
-        {   0,  42 }, {   0,  48 }, {   0,  55 }, {  -6,  59 },
-        {  -7,  71 }, { -12,  83 }, { -11,  87 }, { -30, 119 },
-        {   1,  58 }, {  -3,  29 }, {  -1,  36 }, {   1,  38 },
-        {   2,  43 }, {  -6,  55 }, {   0,  58 }, {   0,  64 },
-        {  -3,  74 }, { -10,  90 }, {   0,  70 }, {  -4,  29 },
-        {   5,  31 }, {   7,  42 }, {   1,  59 }, {  -2,  58 },
-        {  -3,  72 }, {  -3,  81 }, { -11,  97 }, {   0,  58 },
-        {   8,   5 }, {  10,  14 }, {  14,  18 }, {  13,  27 },
-        {   2,  40 }, {   0,  58 }, {  -3,  70 }, {  -6,  79 },
-        {  -8,  85 },
-
-        /* 276 a bit special (not used, bypass is used instead) */
-        { 0, 0 },
-
-        /* 277 - 337 */
-        { -13, 106 }, { -16, 106 }, { -10,  87 }, { -21, 114 },
-        { -18, 110 }, { -14,  98 }, { -22, 110 }, { -21, 106 },
-        { -18, 103 }, { -21, 107 }, { -23, 108 }, { -26, 112 },
-        { -10,  96 }, { -12,  95 }, {  -5,  91 }, {  -9,  93 },
-        { -22,  94 }, {  -5,  86 }, {   9,  67 }, {  -4,  80 },
-        { -10,  85 }, {  -1,  70 }, {   7,  60 }, {   9,  58 },
-        {   5,  61 }, {  12,  50 }, {  15,  50 }, {  18,  49 },
-        {  17,  54 }, {  10,  41 }, {   7,  46 }, {  -1,  51 },
-        {   7,  49 }, {   8,  52 }, {   9,  41 }, {   6,  47 },
-        {   2,  55 }, {  13,  41 }, {  10,  44 }, {   6,  50 },
-        {   5,  53 }, {  13,  49 }, {   4,  63 }, {   6,  64 },
-        {  -2,  69 }, {  -2,  59 }, {   6,  70 }, {  10,  44 },
-        {   9,  31 }, {  12,  43 }, {   3,  53 }, {  14,  34 },
-        {  10,  38 }, {  -3,  52 }, {  13,  40 }, {  17,  32 },
-        {   7,  44 }, {   7,  38 }, {  13,  50 }, {  10,  57 },
-        {  26,  43 },
-
-        /* 338 - 398 */
-        {  14,  11 }, {  11,  14 }, {   9,  11 }, {  18,  11 },
-        {  21,   9 }, {  23,  -2 }, {  32, -15 }, {  32, -15 },
-        {  34, -21 }, {  39, -23 }, {  42, -33 }, {  41, -31 },
-        {  46, -28 }, {  38, -12 }, {  21,  29 }, {  45, -24 },
-        {  53, -45 }, {  48, -26 }, {  65, -43 }, {  43, -19 },
-        {  39, -10 }, {  30,   9 }, {  18,  26 }, {  20,  27 },
-        {   0,  57 }, { -14,  82 }, {  -5,  75 }, { -19,  97 },
-        { -35, 125 }, {  27,   0 }, {  28,   0 }, {  31,  -4 },
-        {  27,   6 }, {  34,   8 }, {  30,  10 }, {  24,  22 },
-        {  33,  19 }, {  22,  32 }, {  26,  31 }, {  21,  41 },
-        {  26,  44 }, {  23,  47 }, {  16,  65 }, {  14,  71 },
-        {   8,  60 }, {   6,  63 }, {  17,  65 }, {  21,  24 },
-        {  23,  20 }, {  26,  23 }, {  27,  32 }, {  28,  23 },
-        {  28,  24 }, {  23,  40 }, {  24,  32 }, {  28,  29 },
-        {  23,  42 }, {  19,  57 }, {  22,  53 }, {  22,  61 },
-        {  11,  86 },
-
-        /* 399 - 435 */
-        {  12,  40 }, {  11,  51 }, {  14,  59 },
-        {  -4,  79 }, {  -7,  71 }, {  -5,  69 }, {  -9,  70 },
-        {  -8,  66 }, { -10,  68 }, { -19,  73 }, { -12,  69 },
-        { -16,  70 }, { -15,  67 }, { -20,  62 }, { -19,  70 },
-        { -16,  66 }, { -22,  65 }, { -20,  63 }, {   9,  -2 },
-        {  26,  -9 }, {  33,  -9 }, {  39,  -7 }, {  41,  -2 },
-        {  45,   3 }, {  49,   9 }, {  45,  27 }, {  36,  59 },
-        {  -6,  66 }, {  -7,  35 }, {  -7,  42 }, {  -8,  45 },
-        {  -5,  48 }, { -12,  56 }, {  -6,  60 }, {  -5,  62 },
-        {  -8,  66 }, {  -8,  76 },
-
-        /* 436 - 459 */
-        {  -5,  85 }, {  -6,  81 }, { -10,  77 }, {  -7,  81 },
-        { -17,  80 }, { -18,  73 }, {  -4,  74 }, { -10,  83 },
-        {  -9,  71 }, {  -9,  67 }, {  -1,  61 }, {  -8,  66 },
-        { -14,  66 }, {   0,  59 }, {   2,  59 }, {  21, -13 },
-        {  33, -14 }, {  39,  -7 }, {  46,  -2 }, {  51,   2 },
-        {  60,   6 }, {  61,  17 }, {  55,  34 }, {  42,  62 },
-    },
-
-    /* i_cabac_init_idc == 1 */
-    {
-        /* 0 - 10 */
-        {  20, -15 }, {   2,  54 }, {   3,  74 }, {  20, -15 },
-        {   2,  54 }, {   3,  74 }, { -28, 127 }, { -23, 104 },
-        {  -6,  53 }, {  -1,  54 }, {   7,  51 },
-
-        /* 11 - 23 */
-        {  22,  25 }, {  34,   0 }, {  16,   0 }, {  -2,   9 },
-        {   4,  41 }, { -29, 118 }, {   2,  65 }, {  -6,  71 },
-        { -13,  79 }, {   5,  52 }, {   9,  50 }, {  -3,  70 },
-        {  10,  54 },
-
-        /* 24 - 39 */
-        {  26,  34 }, {  19,  22 }, {  40,   0 }, {  57,   2 },
-        {  41,  36 }, {  26,  69 }, { -45, 127 }, { -15, 101 },
-        {  -4,  76 }, {  -6,  71 }, { -13,  79 }, {   5,  52 },
-        {   6,  69 }, { -13,  90 }, {   0,  52 }, {   8,  43 },
-
-        /* 40 - 53 */
-        {  -2,  69 },{  -5,  82 },{ -10,  96 },{   2,  59 },
-        {   2,  75 },{  -3,  87 },{  -3,  100 },{   1,  56 },
-        {  -3,  74 },{  -6,  85 },{   0,  59 },{  -3,  81 },
-        {  -7,  86 },{  -5,  95 },
-
-        /* 54 - 59 */
-        {  -1,  66 },{  -1,  77 },{   1,  70 },{  -2,  86 },
-        {  -5,  72 },{   0,  61 },
-
-        /* 60 - 69 */
-        { 0, 41 },   { 0, 63 },   { 0, 63 },     { 0, 63 },
-        { -9, 83 },  { 4, 86 },   { 0, 97 },     { -7, 72 },
-        { 13, 41 },  { 3, 62 },
-
-        /* 70 - 104 */
-        {  13,  15 }, {   7,  51 }, {   2,  80 }, { -39, 127 },
-        { -18,  91 }, { -17,  96 }, { -26,  81 }, { -35,  98 },
-        { -24, 102 }, { -23,  97 }, { -27, 119 }, { -24,  99 },
-        { -21, 110 }, { -18, 102 }, { -36, 127 }, {   0,  80 },
-        {  -5,  89 }, {  -7,  94 }, {  -4,  92 }, {   0,  39 },
-        {   0,  65 }, { -15,  84 }, { -35, 127 }, {  -2,  73 },
-        { -12, 104 }, {  -9,  91 }, { -31, 127 }, {   3,  55 },
-        {   7,  56 }, {   7,  55 }, {   8,  61 }, {  -3,  53 },
-        {   0,  68 }, {  -7,  74 }, {  -9,  88 },
-
-        /* 105 -> 165 */
-        { -13, 103 }, { -13,  91 }, {  -9,  89 }, { -14,  92 },
-        {  -8,  76 }, { -12,  87 }, { -23, 110 }, { -24, 105 },
-        { -10,  78 }, { -20, 112 }, { -17,  99 }, { -78, 127 },
-        { -70, 127 }, { -50, 127 }, { -46, 127 }, {  -4,  66 },
-        {  -5,  78 }, {  -4,  71 }, {  -8,  72 }, {   2,  59 },
-        {  -1,  55 }, {  -7,  70 }, {  -6,  75 }, {  -8,  89 },
-        { -34, 119 }, {  -3,  75 }, {  32,  20 }, {  30,  22 },
-        { -44, 127 }, {   0,  54 }, {  -5,  61 }, {   0,  58 },
-        {  -1,  60 }, {  -3,  61 }, {  -8,  67 }, { -25,  84 },
-        { -14,  74 }, {  -5,  65 }, {   5,  52 }, {   2,  57 },
-        {   0,  61 }, {  -9,  69 }, { -11,  70 }, {  18,  55 },
-        {  -4,  71 }, {   0,  58 }, {   7,  61 }, {   9,  41 },
-        {  18,  25 }, {   9,  32 }, {   5,  43 }, {   9,  47 },
-        {   0,  44 }, {   0,  51 }, {   2,  46 }, {  19,  38 },
-        {  -4,  66 }, {  15,  38 }, {  12,  42 }, {   9,  34 },
-        {   0,  89 },
-
-        /* 166 - 226 */
-        {   4,  45 }, {  10,  28 }, {  10,  31 }, {  33, -11 },
-        {  52, -43 }, {  18,  15 }, {  28,   0 }, {  35, -22 },
-        {  38, -25 }, {  34,   0 }, {  39, -18 }, {  32, -12 },
-        { 102, -94 }, {   0,   0 }, {  56, -15 }, {  33,  -4 },
-        {  29,  10 }, {  37,  -5 }, {  51, -29 }, {  39,  -9 },
-        {  52, -34 }, {  69, -58 }, {  67, -63 }, {  44,  -5 },
-        {  32,   7 }, {  55, -29 }, {  32,   1 }, {   0,   0 },
-        {  27,  36 }, {  33, -25 }, {  34, -30 }, {  36, -28 },
-        {  38, -28 }, {  38, -27 }, {  34, -18 }, {  35, -16 },
-        {  34, -14 }, {  32,  -8 }, {  37,  -6 }, {  35,   0 },
-        {  30,  10 }, {  28,  18 }, {  26,  25 }, {  29,  41 },
-        {   0,  75 }, {   2,  72 }, {   8,  77 }, {  14,  35 },
-        {  18,  31 }, {  17,  35 }, {  21,  30 }, {  17,  45 },
-        {  20,  42 }, {  18,  45 }, {  27,  26 }, {  16,  54 },
-        {   7,  66 }, {  16,  56 }, {  11,  73 }, {  10,  67 },
-        { -10, 116 },
-
-        /* 227 - 275 */
-        { -23, 112 }, { -15,  71 }, {  -7,  61 }, {   0,  53 },
-        {  -5,  66 }, { -11,  77 }, {  -9,  80 }, {  -9,  84 },
-        { -10,  87 }, { -34, 127 }, { -21, 101 }, {  -3,  39 },
-        {  -5,  53 }, {  -7,  61 }, { -11,  75 }, { -15,  77 },
-        { -17,  91 }, { -25, 107 }, { -25, 111 }, { -28, 122 },
-        { -11,  76 }, { -10,  44 }, { -10,  52 }, { -10,  57 },
-        {  -9,  58 }, { -16,  72 }, {  -7,  69 }, {  -4,  69 },
-        {  -5,  74 }, {  -9,  86 }, {   2,  66 }, {  -9,  34 },
-        {   1,  32 }, {  11,  31 }, {   5,  52 }, {  -2,  55 },
-        {  -2,  67 }, {   0,  73 }, {  -8,  89 }, {   3,  52 },
-        {   7,   4 }, {  10,   8 }, {  17,   8 }, {  16,  19 },
-        {   3,  37 }, {  -1,  61 }, {  -5,  73 }, {  -1,  70 },
-        {  -4,  78 },
-
-        /* 276 a bit special (not used, bypass is used instead) */
-        { 0, 0 },
-
-        /* 277 - 337 */
-        { -21, 126 }, { -23, 124 }, { -20, 110 }, { -26, 126 },
-        { -25, 124 }, { -17, 105 }, { -27, 121 }, { -27, 117 },
-        { -17, 102 }, { -26, 117 }, { -27, 116 }, { -33, 122 },
-        { -10,  95 }, { -14, 100 }, {  -8,  95 }, { -17, 111 },
-        { -28, 114 }, {  -6,  89 }, {  -2,  80 }, {  -4,  82 },
-        {  -9,  85 }, {  -8,  81 }, {  -1,  72 }, {   5,  64 },
-        {   1,  67 }, {   9,  56 }, {   0,  69 }, {   1,  69 },
-        {   7,  69 }, {  -7,  69 }, {  -6,  67 }, { -16,  77 },
-        {  -2,  64 }, {   2,  61 }, {  -6,  67 }, {  -3,  64 },
-        {   2,  57 }, {  -3,  65 }, {  -3,  66 }, {   0,  62 },
-        {   9,  51 }, {  -1,  66 }, {  -2,  71 }, {  -2,  75 },
-        {  -1,  70 }, {  -9,  72 }, {  14,  60 }, {  16,  37 },
-        {   0,  47 }, {  18,  35 }, {  11,  37 }, {  12,  41 },
-        {  10,  41 }, {   2,  48 }, {  12,  41 }, {  13,  41 },
-        {   0,  59 }, {   3,  50 }, {  19,  40 }, {   3,  66 },
-        {  18,  50 },
-
-        /* 338 - 398 */
-        {  19,  -6 }, {  18,  -6 }, {  14,   0 }, {  26, -12 },
-        {  31, -16 }, {  33, -25 }, {  33, -22 }, {  37, -28 },
-        {  39, -30 }, {  42, -30 }, {  47, -42 }, {  45, -36 },
-        {  49, -34 }, {  41, -17 }, {  32,   9 }, {  69, -71 },
-        {  63, -63 }, {  66, -64 }, {  77, -74 }, {  54, -39 },
-        {  52, -35 }, {  41, -10 }, {  36,   0 }, {  40,  -1 },
-        {  30,  14 }, {  28,  26 }, {  23,  37 }, {  12,  55 },
-        {  11,  65 }, {  37, -33 }, {  39, -36 }, {  40, -37 },
-        {  38, -30 }, {  46, -33 }, {  42, -30 }, {  40, -24 },
-        {  49, -29 }, {  38, -12 }, {  40, -10 }, {  38,  -3 },
-        {  46,  -5 }, {  31,  20 }, {  29,  30 }, {  25,  44 },
-        {  12,  48 }, {  11,  49 }, {  26,  45 }, {  22,  22 },
-        {  23,  22 }, {  27,  21 }, {  33,  20 }, {  26,  28 },
-        {  30,  24 }, {  27,  34 }, {  18,  42 }, {  25,  39 },
-        {  18,  50 }, {  12,  70 }, {  21,  54 }, {  14,  71 },
-        {  11,  83 },
-
-        /* 399 - 435 */
-        {  25,  32 }, {  21,  49 }, {  21,  54 },
-        {  -5,  85 }, {  -6,  81 }, { -10,  77 }, {  -7,  81 },
-        { -17,  80 }, { -18,  73 }, {  -4,  74 }, { -10,  83 },
-        {  -9,  71 }, {  -9,  67 }, {  -1,  61 }, {  -8,  66 },
-        { -14,  66 }, {   0,  59 }, {   2,  59 }, {  17, -10 },
-        {  32, -13 }, {  42,  -9 }, {  49,  -5 }, {  53,   0 },
-        {  64,   3 }, {  68,  10 }, {  66,  27 }, {  47,  57 },
-        {  -5,  71 }, {   0,  24 }, {  -1,  36 }, {  -2,  42 },
-        {  -2,  52 }, {  -9,  57 }, {  -6,  63 }, {  -4,  65 },
-        {  -4,  67 }, {  -7,  82 },
-
-        /* 436 - 459 */
-        {  -3,  81 }, {  -3,  76 }, {  -7,  72 }, {  -6,  78 },
-        { -12,  72 }, { -14,  68 }, {  -3,  70 }, {  -6,  76 },
-        {  -5,  66 }, {  -5,  62 }, {   0,  57 }, {  -4,  61 },
-        {  -9,  60 }, {   1,  54 }, {   2,  58 }, {  17, -10 },
-        {  32, -13 }, {  42,  -9 }, {  49,  -5 }, {  53,   0 },
-        {  64,   3 }, {  68,  10 }, {  66,  27 }, {  47,  57 },
-    },
-
-    /* i_cabac_init_idc == 2 */
-    {
-        /* 0 - 10 */
-        {  20, -15 }, {   2,  54 }, {   3,  74 }, {  20, -15 },
-        {   2,  54 }, {   3,  74 }, { -28, 127 }, { -23, 104 },
-        {  -6,  53 }, {  -1,  54 }, {   7,  51 },
-
-        /* 11 - 23 */
-        {  29,  16 }, {  25,   0 }, {  14,   0 }, { -10,  51 },
-        {  -3,  62 }, { -27,  99 }, {  26,  16 }, {  -4,  85 },
-        { -24, 102 }, {   5,  57 }, {   6,  57 }, { -17,  73 },
-        {  14,  57 },
-
-        /* 24 - 39 */
-        {  20,  40 }, {  20,  10 }, {  29,   0 }, {  54,   0 },
-        {  37,  42 }, {  12,  97 }, { -32, 127 }, { -22, 117 },
-        {  -2,  74 }, {  -4,  85 }, { -24, 102 }, {   5,  57 },
-        {  -6,  93 }, { -14,  88 }, {  -6,  44 }, {   4,  55 },
-
-        /* 40 - 53 */
-        { -11,  89 },{ -15,  103 },{ -21,  116 },{  19,  57 },
-        {  20,  58 },{   4,  84 },{   6,  96 },{   1,  63 },
-        {  -5,  85 },{ -13,  106 },{   5,  63 },{   6,  75 },
-        {  -3,  90 },{  -1,  101 },
-
-        /* 54 - 59 */
-        {   3,  55 },{  -4,  79 },{  -2,  75 },{ -12,  97 },
-        {  -7,  50 },{   1,  60 },
-
-        /* 60 - 69 */
-        { 0, 41 },   { 0, 63 },   { 0, 63 },     { 0, 63 },
-        { -9, 83 },  { 4, 86 },   { 0, 97 },     { -7, 72 },
-        { 13, 41 },  { 3, 62 },
-
-        /* 70 - 104 */
-        {   7,  34 }, {  -9,  88 }, { -20, 127 }, { -36, 127 },
-        { -17,  91 }, { -14,  95 }, { -25,  84 }, { -25,  86 },
-        { -12,  89 }, { -17,  91 }, { -31, 127 }, { -14,  76 },
-        { -18, 103 }, { -13,  90 }, { -37, 127 }, {  11,  80 },
-        {   5,  76 }, {   2,  84 }, {   5,  78 }, {  -6,  55 },
-        {   4,  61 }, { -14,  83 }, { -37, 127 }, {  -5,  79 },
-        { -11, 104 }, { -11,  91 }, { -30, 127 }, {   0,  65 },
-        {  -2,  79 }, {   0,  72 }, {  -4,  92 }, {  -6,  56 },
-        {   3,  68 }, {  -8,  71 }, { -13,  98 },
-
-        /* 105 -> 165 */
-        {  -4,  86 }, { -12,  88 }, {  -5,  82 }, {  -3,  72 },
-        {  -4,  67 }, {  -8,  72 }, { -16,  89 }, {  -9,  69 },
-        {  -1,  59 }, {   5,  66 }, {   4,  57 }, {  -4,  71 },
-        {  -2,  71 }, {   2,  58 }, {  -1,  74 }, {  -4,  44 },
-        {  -1,  69 }, {   0,  62 }, {  -7,  51 }, {  -4,  47 },
-        {  -6,  42 }, {  -3,  41 }, {  -6,  53 }, {   8,  76 },
-        {  -9,  78 }, { -11,  83 }, {   9,  52 }, {   0,  67 },
-        {  -5,  90 }, {   1,  67 }, { -15,  72 }, {  -5,  75 },
-        {  -8,  80 }, { -21,  83 }, { -21,  64 }, { -13,  31 },
-        { -25,  64 }, { -29,  94 }, {   9,  75 }, {  17,  63 },
-        {  -8,  74 }, {  -5,  35 }, {  -2,  27 }, {  13,  91 },
-        {   3,  65 }, {  -7,  69 }, {   8,  77 }, { -10,  66 },
-        {   3,  62 }, {  -3,  68 }, { -20,  81 }, {   0,  30 },
-        {   1,   7 }, {  -3,  23 }, { -21,  74 }, {  16,  66 },
-        { -23, 124 }, {  17,  37 }, {  44, -18 }, {  50, -34 },
-        { -22, 127 },
-
-        /* 166 - 226 */
-        {   4,  39 }, {   0,  42 }, {   7,  34 }, {  11,  29 },
-        {   8,  31 }, {   6,  37 }, {   7,  42 }, {   3,  40 },
-        {   8,  33 }, {  13,  43 }, {  13,  36 }, {   4,  47 },
-        {   3,  55 }, {   2,  58 }, {   6,  60 }, {   8,  44 },
-        {  11,  44 }, {  14,  42 }, {   7,  48 }, {   4,  56 },
-        {   4,  52 }, {  13,  37 }, {   9,  49 }, {  19,  58 },
-        {  10,  48 }, {  12,  45 }, {   0,  69 }, {  20,  33 },
-        {   8,  63 }, {  35, -18 }, {  33, -25 }, {  28,  -3 },
-        {  24,  10 }, {  27,   0 }, {  34, -14 }, {  52, -44 },
-        {  39, -24 }, {  19,  17 }, {  31,  25 }, {  36,  29 },
-        {  24,  33 }, {  34,  15 }, {  30,  20 }, {  22,  73 },
-        {  20,  34 }, {  19,  31 }, {  27,  44 }, {  19,  16 },
-        {  15,  36 }, {  15,  36 }, {  21,  28 }, {  25,  21 },
-        {  30,  20 }, {  31,  12 }, {  27,  16 }, {  24,  42 },
-        {   0,  93 }, {  14,  56 }, {  15,  57 }, {  26,  38 },
-        { -24, 127 },
-
-        /* 227 - 275 */
-        { -24, 115 }, { -22,  82 }, {  -9,  62 }, {   0,  53 },
-        {   0,  59 }, { -14,  85 }, { -13,  89 }, { -13,  94 },
-        { -11,  92 }, { -29, 127 }, { -21, 100 }, { -14,  57 },
-        { -12,  67 }, { -11,  71 }, { -10,  77 }, { -21,  85 },
-        { -16,  88 }, { -23, 104 }, { -15,  98 }, { -37, 127 },
-        { -10,  82 }, {  -8,  48 }, {  -8,  61 }, {  -8,  66 },
-        {  -7,  70 }, { -14,  75 }, { -10,  79 }, {  -9,  83 },
-        { -12,  92 }, { -18, 108 }, {  -4,  79 }, { -22,  69 },
-        { -16,  75 }, {  -2,  58 }, {   1,  58 }, { -13,  78 },
-        {  -9,  83 }, {  -4,  81 }, { -13,  99 }, { -13,  81 },
-        {  -6,  38 }, { -13,  62 }, {  -6,  58 }, {  -2,  59 },
-        { -16,  73 }, { -10,  76 }, { -13,  86 }, {  -9,  83 },
-        { -10,  87 },
-
-        /* 276 a bit special (not used, bypass is used instead) */
-        { 0, 0 },
-
-        /* 277 - 337 */
-        { -22, 127 }, { -25, 127 }, { -25, 120 }, { -27, 127 },
-        { -19, 114 }, { -23, 117 }, { -25, 118 }, { -26, 117 },
-        { -24, 113 }, { -28, 118 }, { -31, 120 }, { -37, 124 },
-        { -10,  94 }, { -15, 102 }, { -10,  99 }, { -13, 106 },
-        { -50, 127 }, {  -5,  92 }, {  17,  57 }, {  -5,  86 },
-        { -13,  94 }, { -12,  91 }, {  -2,  77 }, {   0,  71 },
-        {  -1,  73 }, {   4,  64 }, {  -7,  81 }, {   5,  64 },
-        {  15,  57 }, {   1,  67 }, {   0,  68 }, { -10,  67 },
-        {   1,  68 }, {   0,  77 }, {   2,  64 }, {   0,  68 },
-        {  -5,  78 }, {   7,  55 }, {   5,  59 }, {   2,  65 },
-        {  14,  54 }, {  15,  44 }, {   5,  60 }, {   2,  70 },
-        {  -2,  76 }, { -18,  86 }, {  12,  70 }, {   5,  64 },
-        { -12,  70 }, {  11,  55 }, {   5,  56 }, {   0,  69 },
-        {   2,  65 }, {  -6,  74 }, {   5,  54 }, {   7,  54 },
-        {  -6,  76 }, { -11,  82 }, {  -2,  77 }, {  -2,  77 },
-        {  25,  42 },
-
-        /* 338 - 398 */
-        {  17, -13 }, {  16,  -9 }, {  17, -12 }, {  27, -21 },
-        {  37, -30 }, {  41, -40 }, {  42, -41 }, {  48, -47 },
-        {  39, -32 }, {  46, -40 }, {  52, -51 }, {  46, -41 },
-        {  52, -39 }, {  43, -19 }, {  32,  11 }, {  61, -55 },
-        {  56, -46 }, {  62, -50 }, {  81, -67 }, {  45, -20 },
-        {  35,  -2 }, {  28,  15 }, {  34,   1 }, {  39,   1 },
-        {  30,  17 }, {  20,  38 }, {  18,  45 }, {  15,  54 },
-        {   0,  79 }, {  36, -16 }, {  37, -14 }, {  37, -17 },
-        {  32,   1 }, {  34,  15 }, {  29,  15 }, {  24,  25 },
-        {  34,  22 }, {  31,  16 }, {  35,  18 }, {  31,  28 },
-        {  33,  41 }, {  36,  28 }, {  27,  47 }, {  21,  62 },
-        {  18,  31 }, {  19,  26 }, {  36,  24 }, {  24,  23 },
-        {  27,  16 }, {  24,  30 }, {  31,  29 }, {  22,  41 },
-        {  22,  42 }, {  16,  60 }, {  15,  52 }, {  14,  60 },
-        {   3,  78 }, { -16, 123 }, {  21,  53 }, {  22,  56 },
-        {  25,  61 },
-
-        /* 399 - 435 */
-        {  21,  33 }, {  19,  50 }, {  17,  61 },
-        {  -3,  78 }, {  -8,  74 }, {  -9,  72 }, { -10,  72 },
-        { -18,  75 }, { -12,  71 }, { -11,  63 }, {  -5,  70 },
-        { -17,  75 }, { -14,  72 }, { -16,  67 }, {  -8,  53 },
-        { -14,  59 }, {  -9,  52 }, { -11,  68 }, {   9,  -2 },
-        {  30, -10 }, {  31,  -4 }, {  33,  -1 }, {  33,   7 },
-        {  31,  12 }, {  37,  23 }, {  31,  38 }, {  20,  64 },
-        {  -9,  71 }, {  -7,  37 }, {  -8,  44 }, { -11,  49 },
-        { -10,  56 }, { -12,  59 }, {  -8,  63 }, {  -9,  67 },
-        {  -6,  68 }, { -10,  79 },
-
-        /* 436 - 459 */
-        {  -3,  78 }, {  -8,  74 }, {  -9,  72 }, { -10,  72 },
-        { -18,  75 }, { -12,  71 }, { -11,  63 }, {  -5,  70 },
-        { -17,  75 }, { -14,  72 }, { -16,  67 }, {  -8,  53 },
-        { -14,  59 }, {  -9,  52 }, { -11,  68 }, {   9,  -2 },
-        {  30, -10 }, {  31,  -4 }, {  33,  -1 }, {  33,   7 },
-        {  31,  12 }, {  37,  23 }, {  31,  38 }, {  20,  64 },
-    }
-};
--- a/src/ffmpeg/libavcodec/h264idct.c	Mon Mar 12 11:37:35 2007 -0700
+++ /dev/null	Thu Jan 01 00:00:00 1970 +0000
@@ -1,168 +0,0 @@
-/*
- * H.264 IDCT
- * Copyright (c) 2004 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 h264-idct.c
- * H.264 IDCT.
- * @author Michael Niedermayer <michaelni@gmx.at>
- */
-
-#include "dsputil.h"
-
-static always_inline void idct_internal(uint8_t *dst, DCTELEM *block, int stride, int block_stride, int shift, int add){
-    int i;
-    uint8_t *cm = cropTbl + MAX_NEG_CROP;
-
-    block[0] += 1<<(shift-1);
-
-    for(i=0; i<4; i++){
-        const int z0=  block[0 + block_stride*i]     +  block[2 + block_stride*i];
-        const int z1=  block[0 + block_stride*i]     -  block[2 + block_stride*i];
-        const int z2= (block[1 + block_stride*i]>>1) -  block[3 + block_stride*i];
-        const int z3=  block[1 + block_stride*i]     + (block[3 + block_stride*i]>>1);
-
-        block[0 + block_stride*i]= z0 + z3;
-        block[1 + block_stride*i]= z1 + z2;
-        block[2 + block_stride*i]= z1 - z2;
-        block[3 + block_stride*i]= z0 - z3;
-    }
-
-    for(i=0; i<4; i++){
-        const int z0=  block[i + block_stride*0]     +  block[i + block_stride*2];
-        const int z1=  block[i + block_stride*0]     -  block[i + block_stride*2];
-        const int z2= (block[i + block_stride*1]>>1) -  block[i + block_stride*3];
-        const int z3=  block[i + block_stride*1]     + (block[i + block_stride*3]>>1);
-
-        dst[i + 0*stride]= cm[ add*dst[i + 0*stride] + ((z0 + z3) >> shift) ];
-        dst[i + 1*stride]= cm[ add*dst[i + 1*stride] + ((z1 + z2) >> shift) ];
-        dst[i + 2*stride]= cm[ add*dst[i + 2*stride] + ((z1 - z2) >> shift) ];
-        dst[i + 3*stride]= cm[ add*dst[i + 3*stride] + ((z0 - z3) >> shift) ];
-    }
-}
-
-void ff_h264_idct_add_c(uint8_t *dst, DCTELEM *block, int stride){
-    idct_internal(dst, block, stride, 4, 6, 1);
-}
-
-void ff_h264_lowres_idct_add_c(uint8_t *dst, int stride, DCTELEM *block){
-    idct_internal(dst, block, stride, 8, 3, 1);
-}
-
-void ff_h264_lowres_idct_put_c(uint8_t *dst, int stride, DCTELEM *block){
-    idct_internal(dst, block, stride, 8, 3, 0);
-}
-
-void ff_h264_idct8_add_c(uint8_t *dst, DCTELEM *block, int stride){
-    int i;
-    DCTELEM (*src)[8] = (DCTELEM(*)[8])block;
-    uint8_t *cm = cropTbl + MAX_NEG_CROP;
-
-    block[0] += 32;
-
-    for( i = 0; i < 8; i++ )
-    {
-        const int a0 =  src[i][0] + src[i][4];
-        const int a2 =  src[i][0] - src[i][4];
-        const int a4 = (src[i][2]>>1) - src[i][6];
-        const int a6 = (src[i][6]>>1) + src[i][2];
-
-        const int b0 = a0 + a6;
-        const int b2 = a2 + a4;
-        const int b4 = a2 - a4;
-        const int b6 = a0 - a6;
-
-        const int a1 = -src[i][3] + src[i][5] - src[i][7] - (src[i][7]>>1);
-        const int a3 =  src[i][1] + src[i][7] - src[i][3] - (src[i][3]>>1);
-        const int a5 = -src[i][1] + src[i][7] + src[i][5] + (src[i][5]>>1);
-        const int a7 =  src[i][3] + src[i][5] + src[i][1] + (src[i][1]>>1);
-
-        const int b1 = (a7>>2) + a1;
-        const int b3 =  a3 + (a5>>2);
-        const int b5 = (a3>>2) - a5;
-        const int b7 =  a7 - (a1>>2);
-
-        src[i][0] = b0 + b7;
-        src[i][7] = b0 - b7;
-        src[i][1] = b2 + b5;
-        src[i][6] = b2 - b5;
-        src[i][2] = b4 + b3;
-        src[i][5] = b4 - b3;
-        src[i][3] = b6 + b1;
-        src[i][4] = b6 - b1;
-    }
-    for( i = 0; i < 8; i++ )
-    {
-        const int a0 =  src[0][i] + src[4][i];
-        const int a2 =  src[0][i] - src[4][i];
-        const int a4 = (src[2][i]>>1) - src[6][i];
-        const int a6 = (src[6][i]>>1) + src[2][i];
-
-        const int b0 = a0 + a6;
-        const int b2 = a2 + a4;
-        const int b4 = a2 - a4;
-        const int b6 = a0 - a6;
-
-        const int a1 = -src[3][i] + src[5][i] - src[7][i] - (src[7][i]>>1);
-        const int a3 =  src[1][i] + src[7][i] - src[3][i] - (src[3][i]>>1);
-        const int a5 = -src[1][i] + src[7][i] + src[5][i] + (src[5][i]>>1);
-        const int a7 =  src[3][i] + src[5][i] + src[1][i] + (src[1][i]>>1);
-
-        const int b1 = (a7>>2) + a1;
-        const int b3 =  a3 + (a5>>2);
-        const int b5 = (a3>>2) - a5;
-        const int b7 =  a7 - (a1>>2);
-
-        dst[i + 0*stride] = cm[ dst[i + 0*stride] + ((b0 + b7) >> 6) ];
-        dst[i + 1*stride] = cm[ dst[i + 1*stride] + ((b2 + b5) >> 6) ];
-        dst[i + 2*stride] = cm[ dst[i + 2*stride] + ((b4 + b3) >> 6) ];
-        dst[i + 3*stride] = cm[ dst[i + 3*stride] + ((b6 + b1) >> 6) ];
-        dst[i + 4*stride] = cm[ dst[i + 4*stride] + ((b6 - b1) >> 6) ];
-        dst[i + 5*stride] = cm[ dst[i + 5*stride] + ((b4 - b3) >> 6) ];
-        dst[i + 6*stride] = cm[ dst[i + 6*stride] + ((b2 - b5) >> 6) ];
-        dst[i + 7*stride] = cm[ dst[i + 7*stride] + ((b0 - b7) >> 6) ];
-    }
-}
-
-// assumes all AC coefs are 0
-void ff_h264_idct_dc_add_c(uint8_t *dst, DCTELEM *block, int stride){
-    int i, j;
-    uint8_t *cm = cropTbl + MAX_NEG_CROP;
-    int dc = (block[0] + 32) >> 6;
-    for( j = 0; j < 4; j++ )
-    {
-        for( i = 0; i < 4; i++ )
-            dst[i] = cm[ dst[i] + dc ];
-        dst += stride;
-    }
-}
-
-void ff_h264_idct8_dc_add_c(uint8_t *dst, DCTELEM *block, int stride){
-    int i, j;
-    uint8_t *cm = cropTbl + MAX_NEG_CROP;
-    int dc = (block[0] + 32) >> 6;
-    for( j = 0; j < 8; j++ )
-    {
-        for( i = 0; i < 8; i++ )
-            dst[i] = cm[ dst[i] + dc ];
-        dst += stride;
-    }
-}