view bfi.c @ 11032:01bd040f8607 libavcodec

Unroll main loop so the edge==0 case is seperate. This allows many things to be simplified away. h264 decoder is overall 1% faster with a mbaff sample and 0.1% slower with the cathedral sample, probably because the slow loop filter code must be loaded into the code cache for each first MB of each row but isnt used for the following MBs.

/*
 * Brute Force & Ignorance (BFI) video decoder
 * Copyright (c) 2008 Sisir Koppaka
 *
 * 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 libavcodec/bfi.c
 * @brief Brute Force & Ignorance (.bfi) video decoder
 * @author Sisir Koppaka ( sisir.koppaka at gmail dot com )
 * @sa http://wiki.multimedia.cx/index.php?title=BFI
 */

#include "libavutil/common.h"
#include "avcodec.h"
#include "bytestream.h"

typedef struct BFIContext {
    AVCodecContext *avctx;
    AVFrame frame;
    uint8_t *dst;
} BFIContext;

static av_cold int bfi_decode_init(AVCodecContext * avctx)
{
    BFIContext *bfi = avctx->priv_data;
    avctx->pix_fmt = PIX_FMT_PAL8;
    bfi->dst = av_mallocz(avctx->width * avctx->height);
    return 0;
}

static int bfi_decode_frame(AVCodecContext * avctx, void *data,
                            int *data_size, AVPacket *avpkt)
{
    const uint8_t *buf = avpkt->data;
    int buf_size = avpkt->size;
    BFIContext *bfi = avctx->priv_data;
    uint8_t *dst = bfi->dst;
    uint8_t *src, *dst_offset, colour1, colour2;
    uint8_t *frame_end = bfi->dst + avctx->width * avctx->height;
    uint32_t *pal;
    int i, j, height = avctx->height;

    if (bfi->frame.data[0])
        avctx->release_buffer(avctx, &bfi->frame);

    bfi->frame.reference = 1;

    if (avctx->get_buffer(avctx, &bfi->frame) < 0) {
        av_log(avctx, AV_LOG_ERROR, "get_buffer() failed\n");
        return -1;
    }

    /* Set frame parameters and palette, if necessary */
    if (!avctx->frame_number) {
        bfi->frame.pict_type = FF_I_TYPE;
        bfi->frame.key_frame = 1;
        /* Setting the palette */
        if(avctx->extradata_size>768) {
            av_log(NULL, AV_LOG_ERROR, "Palette is too large.\n");
            return -1;
        }
        pal = (uint32_t *) bfi->frame.data[1];
        for (i = 0; i < avctx->extradata_size / 3; i++) {
            int shift = 16;
            *pal = 0;
            for (j = 0; j < 3; j++, shift -= 8)
                *pal +=
                    ((avctx->extradata[i * 3 + j] << 2) |
                    (avctx->extradata[i * 3 + j] >> 4)) << shift;
            pal++;
        }
        bfi->frame.palette_has_changed = 1;
    } else {
        bfi->frame.pict_type = FF_P_TYPE;
        bfi->frame.key_frame = 0;
    }

    buf += 4; //Unpacked size, not required.

    while (dst != frame_end) {
        static const uint8_t lentab[4]={0,2,0,1};
        unsigned int byte = *buf++, av_uninit(offset);
        unsigned int code = byte >> 6;
        unsigned int length = byte & ~0xC0;

        /* Get length and offset(if required) */
        if (length == 0) {
            if (code == 1) {
                length = bytestream_get_byte(&buf);
                offset = bytestream_get_le16(&buf);
            } else {
                length = bytestream_get_le16(&buf);
                if (code == 2 && length == 0)
                    break;
            }
        } else {
            if (code == 1)
                offset = bytestream_get_byte(&buf);
        }

        /* Do boundary check */
        if (dst + (length<<lentab[code]) > frame_end)
            break;

        switch (code) {

        case 0:                //Normal Chain
            bytestream_get_buffer(&buf, dst, length);
            dst += length;
            break;

        case 1:                //Back Chain
            dst_offset = dst - offset;
            length *= 4;        //Convert dwords to bytes.
            if (dst_offset < bfi->dst)
                break;
            while (length--)
                *dst++ = *dst_offset++;
            break;

        case 2:                //Skip Chain
            dst += length;
            break;

        case 3:                //Fill Chain
            colour1 = bytestream_get_byte(&buf);
            colour2 = bytestream_get_byte(&buf);
            while (length--) {
                *dst++ = colour1;
                *dst++ = colour2;
            }
            break;

        }
    }

    src = bfi->dst;
    dst = bfi->frame.data[0];
    while (height--) {
        memcpy(dst, src, avctx->width);
        src += avctx->width;
        dst += bfi->frame.linesize[0];
    }
    *data_size = sizeof(AVFrame);
    *(AVFrame *) data = bfi->frame;
    return buf_size;
}

static av_cold int bfi_decode_close(AVCodecContext * avctx)
{
    BFIContext *bfi = avctx->priv_data;
    if (bfi->frame.data[0])
        avctx->release_buffer(avctx, &bfi->frame);
    av_free(bfi->dst);
    return 0;
}

AVCodec bfi_decoder = {
    .name = "bfi",
    .type = CODEC_TYPE_VIDEO,
    .id = CODEC_ID_BFI,
    .priv_data_size = sizeof(BFIContext),
    .init = bfi_decode_init,
    .close = bfi_decode_close,
    .decode = bfi_decode_frame,
    .capabilities = CODEC_CAP_DR1,
    .long_name = NULL_IF_CONFIG_SMALL("Brute Force & Ignorance"),
};