view nuv.c @ 12272:dd90555c98fd libavcodec

Split pextrw macro-spaghetti into several opt-specific macros, this will make future new optimizations (imagine a sse5) much easier. Also fix a bug where we used the direction (%2) rather than optimization (%1) to enable this, which means it wasn't ever actually used...

/*
 * 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"),
};