view nuv.c @ 12105:d6e87496883b libavcodec

ARM: set section to .text in 'function' macro This ensures code always goes into the .text section and avoids the need to specify it explicitly after changing sections.
author mru
date Wed, 07 Jul 2010 20:09:41 +0000
parents 8a4984c5cacc
children 914f484bb476
line wrap: on
line source

/*
 * NuppelVideo decoder
 * Copyright (c) 2006 Reimar Doeffinger
 *
 * 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
 */
#include <stdio.h>
#include <stdlib.h>

#include "libavutil/bswap.h"
#include "libavutil/lzo.h"
#include "avcodec.h"
#include "dsputil.h"
#include "rtjpeg.h"

typedef struct {
    AVFrame pic;
    int codec_frameheader;
    int quality;
    int width, height;
    unsigned int decomp_size;
    unsigned char* decomp_buf;
    uint32_t lq[64], cq[64];
    RTJpegContext rtj;
    DSPContext dsp;
} NuvContext;

static const uint8_t fallback_lquant[] = {
    16,  11,  10,  16,  24,  40,  51,  61,
    12,  12,  14,  19,  26,  58,  60,  55,
    14,  13,  16,  24,  40,  57,  69,  56,
    14,  17,  22,  29,  51,  87,  80,  62,
    18,  22,  37,  56,  68, 109, 103,  77,
    24,  35,  55,  64,  81, 104, 113,  92,
    49,  64,  78,  87, 103, 121, 120, 101,
    72,  92,  95,  98, 112, 100, 103,  99
};

static const uint8_t fallback_cquant[] = {
    17, 18, 24, 47, 99, 99, 99, 99,
    18, 21, 26, 66, 99, 99, 99, 99,
    24, 26, 56, 99, 99, 99, 99, 99,
    47, 66, 99, 99, 99, 99, 99, 99,
    99, 99, 99, 99, 99, 99, 99, 99,
    99, 99, 99, 99, 99, 99, 99, 99,
    99, 99, 99, 99, 99, 99, 99, 99,
    99, 99, 99, 99, 99, 99, 99, 99
};

/**
 * \brief copy frame data from buffer to AVFrame, handling stride.
 * \param f destination AVFrame
 * \param src source buffer, does not use any line-stride
 * \param width width of the video frame
 * \param height height of the video frame
 */
static void copy_frame(AVFrame *f, const uint8_t *src,
                       int width, int height) {
    AVPicture pic;
    avpicture_fill(&pic, src, PIX_FMT_YUV420P, width, height);
    av_picture_copy((AVPicture *)f, &pic, PIX_FMT_YUV420P, width, height);
}

/**
 * \brief extract quantization tables from codec data into our context
 */
static int get_quant(AVCodecContext *avctx, NuvContext *c,
                     const uint8_t *buf, int size) {
    int i;
    if (size < 2 * 64 * 4) {
        av_log(avctx, AV_LOG_ERROR, "insufficient rtjpeg quant data\n");
        return -1;
    }
    for (i = 0; i < 64; i++, buf += 4)
        c->lq[i] = AV_RL32(buf);
    for (i = 0; i < 64; i++, buf += 4)
        c->cq[i] = AV_RL32(buf);
    return 0;
}

/**
 * \brief set quantization tables from a quality value
 */
static void get_quant_quality(NuvContext *c, int quality) {
    int i;
    quality = FFMAX(quality, 1);
    for (i = 0; i < 64; i++) {
        c->lq[i] = (fallback_lquant[i] << 7) / quality;
        c->cq[i] = (fallback_cquant[i] << 7) / quality;
    }
}

static int codec_reinit(AVCodecContext *avctx, int width, int height, int quality) {
    NuvContext *c = avctx->priv_data;
    width = (width + 1) & ~1;
    height = (height + 1) & ~1;
    if (quality >= 0)
        get_quant_quality(c, quality);
    if (width != c->width || height != c->height) {
        if (avcodec_check_dimensions(avctx, height, width) < 0)
            return 0;
        avctx->width = c->width = width;
        avctx->height = c->height = height;
        c->decomp_size = c->height * c->width * 3 / 2;
        c->decomp_buf = av_realloc(c->decomp_buf, c->decomp_size + AV_LZO_OUTPUT_PADDING);
        if (!c->decomp_buf) {
            av_log(avctx, AV_LOG_ERROR, "Can't allocate decompression buffer.\n");
            return 0;
        }
        rtjpeg_decode_init(&c->rtj, &c->dsp, c->width, c->height, c->lq, c->cq);
    } else if (quality != c->quality)
        rtjpeg_decode_init(&c->rtj, &c->dsp, c->width, c->height, c->lq, c->cq);
    return 1;
}

static int decode_frame(AVCodecContext *avctx, void *data, int *data_size,
                        AVPacket *avpkt) {
    const uint8_t *buf = avpkt->data;
    int buf_size = avpkt->size;
    NuvContext *c = avctx->priv_data;
    AVFrame *picture = data;
    int orig_size = buf_size;
    int keyframe;
    int result;
    enum {NUV_UNCOMPRESSED = '0', NUV_RTJPEG = '1',
          NUV_RTJPEG_IN_LZO = '2', NUV_LZO = '3',
          NUV_BLACK = 'N', NUV_COPY_LAST = 'L'} comptype;

    if (buf_size < 12) {
        av_log(avctx, AV_LOG_ERROR, "coded frame too small\n");
        return -1;
    }

    // codec data (rtjpeg quant tables)
    if (buf[0] == 'D' && buf[1] == 'R') {
        int ret;
        // skip rest of the frameheader.
        buf = &buf[12];
        buf_size -= 12;
        ret = get_quant(avctx, c, buf, buf_size);
        if (ret < 0)
            return ret;
        rtjpeg_decode_init(&c->rtj, &c->dsp, c->width, c->height, c->lq, c->cq);
        return orig_size;
    }

    if (buf[0] != 'V' || buf_size < 12) {
        av_log(avctx, AV_LOG_ERROR, "not a nuv video frame\n");
        return -1;
    }
    comptype = buf[1];
    switch (comptype) {
        case NUV_RTJPEG_IN_LZO:
        case NUV_RTJPEG:
            keyframe = !buf[2]; break;
        case NUV_COPY_LAST:
            keyframe = 0; break;
        default:
            keyframe = 1; break;
    }
    // skip rest of the frameheader.
    buf = &buf[12];
    buf_size -= 12;
    if (comptype == NUV_RTJPEG_IN_LZO || comptype == NUV_LZO) {
        int outlen = c->decomp_size, inlen = buf_size;
        if (av_lzo1x_decode(c->decomp_buf, &outlen, buf, &inlen))
            av_log(avctx, AV_LOG_ERROR, "error during lzo decompression\n");
        buf = c->decomp_buf;
        buf_size = c->decomp_size;
    }
    if (c->codec_frameheader) {
        int w, h, q;
        if (buf_size < 12) {
            av_log(avctx, AV_LOG_ERROR, "invalid nuv video frame\n");
            return -1;
        }
        w = AV_RL16(&buf[6]);
        h = AV_RL16(&buf[8]);
        q = buf[10];
        if (!codec_reinit(avctx, w, h, q))
            return -1;
        buf = &buf[12];
        buf_size -= 12;
    }

    if (keyframe && c->pic.data[0])
        avctx->release_buffer(avctx, &c->pic);
    c->pic.reference = 3;
    c->pic.buffer_hints = FF_BUFFER_HINTS_VALID | FF_BUFFER_HINTS_READABLE |
                          FF_BUFFER_HINTS_PRESERVE | FF_BUFFER_HINTS_REUSABLE;
    result = avctx->reget_buffer(avctx, &c->pic);
    if (result < 0) {
        av_log(avctx, AV_LOG_ERROR, "get_buffer() failed\n");
        return -1;
    }

    c->pic.pict_type = keyframe ? FF_I_TYPE : FF_P_TYPE;
    c->pic.key_frame = keyframe;
    // decompress/copy/whatever data
    switch (comptype) {
        case NUV_LZO:
        case NUV_UNCOMPRESSED: {
            int height = c->height;
            if (buf_size < c->width * height * 3 / 2) {
                av_log(avctx, AV_LOG_ERROR, "uncompressed frame too short\n");
                height = buf_size / c->width / 3 * 2;
            }
            copy_frame(&c->pic, buf, c->width, height);
            break;
        }
        case NUV_RTJPEG_IN_LZO:
        case NUV_RTJPEG: {
            rtjpeg_decode_frame_yuv420(&c->rtj, &c->pic, buf, buf_size);
            break;
        }
        case NUV_BLACK: {
            memset(c->pic.data[0], 0, c->width * c->height);
            memset(c->pic.data[1], 128, c->width * c->height / 4);
            memset(c->pic.data[2], 128, c->width * c->height / 4);
            break;
        }
        case NUV_COPY_LAST: {
            /* nothing more to do here */
            break;
        }
        default:
            av_log(avctx, AV_LOG_ERROR, "unknown compression\n");
            return -1;
    }

    *picture = c->pic;
    *data_size = sizeof(AVFrame);
    return orig_size;
}

static av_cold int decode_init(AVCodecContext *avctx) {
    NuvContext *c = avctx->priv_data;
    avctx->pix_fmt = PIX_FMT_YUV420P;
    c->pic.data[0] = NULL;
    c->decomp_buf = NULL;
    c->quality = -1;
    c->width = 0;
    c->height = 0;
    c->codec_frameheader = avctx->codec_tag == MKTAG('R', 'J', 'P', 'G');
    if (avctx->extradata_size)
        get_quant(avctx, c, avctx->extradata, avctx->extradata_size);
    dsputil_init(&c->dsp, avctx);
    if (!codec_reinit(avctx, avctx->width, avctx->height, -1))
        return 1;
    return 0;
}

static av_cold int decode_end(AVCodecContext *avctx) {
    NuvContext *c = avctx->priv_data;
    av_freep(&c->decomp_buf);
    if (c->pic.data[0])
        avctx->release_buffer(avctx, &c->pic);
    return 0;
}

AVCodec nuv_decoder = {
    "nuv",
    AVMEDIA_TYPE_VIDEO,
    CODEC_ID_NUV,
    sizeof(NuvContext),
    decode_init,
    NULL,
    decode_end,
    decode_frame,
    CODEC_CAP_DR1,
    .long_name = NULL_IF_CONFIG_SMALL("NuppelVideo/RTJPEG"),
};