FFmpeg在libavcodec模块，旧版本提供avcodec_decode_video2()作为视频解码函数，avcodec_decode_audio4()作为音频解码函数。在FFmpeg 3.1版本新增avcodec_send_packet()与avcodec_receive_frame()作为音视频解码函数。后来，在3.4版本把avcodec_decode_video2()和avcodec_decode_audio4()标记为过时API。版本变更描述如下：

FFmpeg 3.1

2016-04-21 - 7fc329e - lavc 57.37.100 - avcodec.h
Add a new audio/video encoding and decoding API with decoupled input
and output -- avcodec_send_packet(), avcodec_receive_frame(),
avcodec_send_frame() and avcodec_receive_packet().

FFmpeg 3.4

2017-09-26 - b1cf151c4d - lavc 57.106.102 - avcodec.h
Deprecate AVCodecContext.refcounted_frames. This was useful for deprecated
API only (avcodec_decode_video2/avcodec_decode_audio4). The new decode APIs
(avcodec_send_packet/avcodec_receive_frame) always work with reference
counted frames.

avcodec_send_packet()和avcodec_receive_frame()函数声明位于libavcodec/avcodec.h。我们来看看函数声明介绍：

/**
 * Supply raw packet data as input to a decoder.
 *
 * @param avctx codec context
 * @param[in] avpkt The input AVPacket. Usually, this will be a single video
 *                  frame, or several complete audio frames.
 *
 * @return 0 on success, otherwise negative error code:
 *      AVERROR(EAGAIN):   input is not accepted in the current state - user
 *                         must read output with avcodec_receive_frame()
 *      AVERROR_EOF:       the decoder has been flushed, and no new packets
 *      AVERROR(EINVAL):   codec not opened, it is an encoder, or requires flush
 *      AVERROR(ENOMEM):   failed to add packet to internal queue, or similar
 */
int avcodec_send_packet(AVCodecContext *avctx, const AVPacket *avpkt);


/**
 * Return decoded output data from a decoder.
 *
 * @param avctx codec context
 * @param frame This will be set to a reference-counted video or audio
 *              frame (depending on the decoder type) allocated by the
 *              decoder. Note that the function will always call
 *              av_frame_unref(frame) before doing anything else.
 *
 * @return
 *      0:                 success, a frame was returned
 *      AVERROR(EAGAIN):   output is not available in this state
 *      AVERROR_EOF:       the decoder has been fully flushed
 *      AVERROR(EINVAL):   codec not opened, or it is an encoder
 *      AVERROR_INPUT_CHANGED: current decoded frame has changed parameters
 */
int avcodec_receive_frame(AVCodecContext *avctx, AVFrame *frame);

由描述可知，avcodec_send_packet()负责把AVpacket数据包发送给解码器，avcodec_receive_frame()则从解码器取出一帧AVFrame数据。返回0，代表解码成功；返回EAGAIN，代表当前状态无可输出的数据；返回EOF，代表到达文件流结尾；返回INVAL，代表解码器未打开或者当前打开的是编码器；返回INPUT_CHANGED，代表输入参数发生变化，比如width和height改变。

avcodec_send_packet()和avcodec_receive_frame()的解码流程图如下：

一、avcodec_send_packet发送AVPacket

1、avcodec_send_packet

avcodec_send_packet()函数位于libavcodec/decode.c，具体如下：

int avcodec_send_packet(AVCodecContext *avctx, const AVPacket *avpkt)
{
    AVCodecInternal *avci = avctx->internal;
    int ret;
    // 判断解码器是否打开，是否为解码器
    if (!avcodec_is_open(avctx) || !av_codec_is_decoder(avctx->codec))
        return AVERROR(EINVAL);
    if (avctx->internal->draining)
        return AVERROR_EOF;
    if (avpkt && !avpkt->size && avpkt->data)
        return AVERROR(EINVAL);

    av_packet_unref(avci->buffer_pkt);
    if (avpkt && (avpkt->data || avpkt->side_data_elems)) {
        ret = av_packet_ref(avci->buffer_pkt, avpkt);
        if (ret < 0)
            return ret;
    }
    // 发送packet到bitstream滤波器
    ret = av_bsf_send_packet(avci->bsf, avci->buffer_pkt);
    if (ret < 0) {
        av_packet_unref(avci->buffer_pkt);
        return ret;
    }
    if (!avci->buffer_frame->buf[0]) {
        ret = decode_receive_frame_internal(avctx, avci->buffer_frame);
        if (ret < 0 && ret != AVERROR(EAGAIN) && ret != AVERROR_EOF)
            return ret;
    }

    return 0;
}

2、av_bsf_send_packet

由此可见，内部是调用av_bsf_send_packet()函数把packet发送给bitstream滤波器，代码位于libavcodec/bsf.c，具体如下：

int av_bsf_send_packet(AVBSFContext *ctx, AVPacket *pkt)
{
    AVBSFInternal *bsfi = ctx->internal;
    int ret;

    if (!pkt || IS_EMPTY(pkt)) {
        bsfi->eof = 1;
        return 0;
    }
    if (bsfi->eof) {
        return AVERROR(EINVAL);
    }
    if (!IS_EMPTY(bsfi->buffer_pkt))
        return AVERROR(EAGAIN);
    // 申请AVPacket内存，并拷贝数据
    ret = av_packet_make_refcounted(pkt);
    if (ret < 0)
        return ret;
	// 把pkt指针赋值给bsfi->buffer_pkt
    av_packet_move_ref(bsfi->buffer_pkt, pkt);

    return 0;
}

av_bsf_send_packet()函数内部调用av_packet_make_refcounted()来申请AVPacket内存，并拷贝数据。然后调用av_packet_move_ref()函数把pkt指针赋值给bsfi->buffer_pkt。

二、avcodec_receive_frame接收AVFrame

1、avcodec_receive_frame

avcodec_receive_frame()函数同样位于libavcodec/decode.c，主要是调用内部函数decode_receive_frame_internal()实现解码，具体代码如下：

int avcodec_receive_frame(AVCodecContext *avctx, AVFrame *frame)
{
    AVCodecInternal *avci = avctx->internal;
    int ret, changed;

    av_frame_unref(frame);
    // 判断解码器是否打开，是否为解码器
    if (!avcodec_is_open(avctx) || !av_codec_is_decoder(avctx->codec))
        return AVERROR(EINVAL);
    if (avci->buffer_frame->buf[0]) {
        av_frame_move_ref(frame, avci->buffer_frame);
    } else {
        ret = decode_receive_frame_internal(avctx, frame);
        if (ret < 0)
            return ret;
    }
    if (avctx->codec_type == AVMEDIA_TYPE_VIDEO) {
        ret = apply_cropping(avctx, frame);
        if (ret < 0) {
            av_frame_unref(frame);
            return ret;
        }
    }

    ......
	
    return 0;
}

2、decode_receive_frame_internal

decode_receive_frame_internal()函数判断是否支持avctx->codec->receive_frame，如果支持就调用avctx->codec->receive_frame()函数进行解码，否则调用decode_simple_receive_frame()函数进行解码。具体如下：

static int decode_receive_frame_internal(AVCodecContext *avctx, AVFrame *frame)
{
    AVCodecInternal *avci = avctx->internal;
    int ret;

    if (avctx->codec->receive_frame) {
        ret = avctx->codec->receive_frame(avctx, frame);
        if (ret != AVERROR(EAGAIN))
            av_packet_unref(avci->last_pkt_props);
    } else {
        ret = decode_simple_receive_frame(avctx, frame);
	}
    ......
	
    /* free the per-frame decode data */
    av_buffer_unref(&frame->private_ref);
    return ret;
}

3、decode_simple_receive_frame

decode_simple_receive_frame()函数又调用decode_simple_internal()内部函数进行解码：

static int decode_simple_receive_frame(AVCodecContext *avctx, AVFrame *frame)
{
    int ret;
    int64_t discarded_samples = 0;

    while (!frame->buf[0]) {
        if (discarded_samples > avctx->max_samples)
            return AVERROR(EAGAIN);
        ret = decode_simple_internal(avctx, frame, &discarded_samples);
        if (ret < 0)
            return ret;
    }

    return 0;
}

4、decode_simple_internal

由描述可知，decode_simple_internal()函数是解码器的核心封装函数。我们应该循环调用avcodec_receive_frame()函数，直到返回EAGAIN。通过判断是否需要特定线程进行解码，如果需要就调用ff_thread_decode_frame()函数，否则调用avctx->codec->decode指向的解码函数。具体调用过程如下：

/*
 * The core of the receive_frame_wrapper for the decoders implementing
 * the simple API. Certain decoders might consume partial packets without
 * returning any output, so this function needs to be called in a loop until it
 * returns EAGAIN.
 **/
static inline int decode_simple_internal(AVCodecContext *avctx, 
										 AVFrame *frame, 
										 int64_t *discarded_samples)
{
    AVCodecInternal   *avci = avctx->internal;
    DecodeSimpleContext *ds = &avci->ds;
    AVPacket           *pkt = ds->in_pkt;
    int got_frame, actual_got_frame;
    int ret;

    if (!pkt->data && !avci->draining) {
        av_packet_unref(pkt);
        ret = ff_decode_get_packet(avctx, pkt);
        if (ret < 0 && ret != AVERROR_EOF)
            return ret;
    }
    if (avci->draining_done)
        return AVERROR_EOF;

    if (!pkt->data &&
        !(avctx->codec->capabilities & AV_CODEC_CAP_DELAY ||
          avctx->active_thread_type & FF_THREAD_FRAME))
        return AVERROR_EOF;

    got_frame = 0;

    if (HAVE_THREADS && avctx->active_thread_type & FF_THREAD_FRAME) {
        ret = ff_thread_decode_frame(avctx, frame, &got_frame, pkt);
    } else {
        ret = avctx->codec->decode(avctx, frame, &got_frame, pkt);

        ......
    }
    
	......

    return ret < 0 ? ret : 0;
}

5、ff_thread_decode_frame

ff_thread_decode_frame()函数位于libavcodec/pthread_frame.c。首先是调用submit_packet()函数提交packet到下一个解码线程，然后调用数组中最前面线程进行解码。相关代码如下：

int ff_thread_decode_frame(AVCodecContext *avctx,
                           AVFrame *picture, 
						   int *got_picture_ptr,
                           AVPacket *avpkt)
{
    FrameThreadContext *fctx = avctx->internal->thread_ctx;
    int finished = fctx->next_finished;
    PerThreadContext *p;
    int err;
    async_unlock(fctx);

    /*
     * Submit a packet to the next decoding thread.
     */
    p = &fctx->threads[fctx->next_decoding];
    err = submit_packet(p, avctx, avpkt);
    if (err)
        goto finish;

    ......

    /* Return the next available frame from the oldest thread.*/
    do {
        p = &fctx->threads[finished++];
        if (atomic_load(&p->state) != STATE_INPUT_READY) {
            pthread_mutex_lock(&p->progress_mutex);
            while (atomic_load_explicit(&p->state, memory_order_relaxed) != STATE_INPUT_READY)
                pthread_cond_wait(&p->output_cond, &p->progress_mutex);
            pthread_mutex_unlock(&p->progress_mutex);
        }
        av_frame_move_ref(picture, p->frame);
        *got_picture_ptr = p->got_frame;
        picture->pkt_dts = p->avpkt->dts;
        err = p->result;
        p->got_frame = 0;
        p->result = 0;

        if (finished >= avctx->thread_count) finished = 0;
    } while (!avpkt->size && !*got_picture_ptr && err >= 0 && finished != fctx->next_finished);

    update_context_from_thread(avctx, p->avctx, 1);
    if (fctx->next_decoding >= avctx->thread_count)
		fctx->next_decoding = 0;
    fctx->next_finished = finished;
    /* return the size of the consumed packet if no error occurred */
    if (err >= 0)
        err = avpkt->size;
finish:
    async_lock(fctx);
    return err;
}

6、h264_decode_frame

以h264解码器为例，位于libavcodec/h264dec.c，对应的AVCodec如下：

AVCodec ff_h264_decoder = {
    .name                  = "h264",
    .long_name             = NULL_IF_CONFIG_SMALL("H.264/AVC/MPEG-4 AVC/MPEG-4 part10"),
    .type                  = AVMEDIA_TYPE_VIDEO,
    .id                    = AV_CODEC_ID_H264,
    .priv_data_size        = sizeof(H264Context),
    .init                  = h264_decode_init,
    .close                 = h264_decode_end,
    .decode                = h264_decode_frame,
    .capabilities          = AV_CODEC_CAP_DR1 |
                             AV_CODEC_CAP_DELAY | AV_CODEC_CAP_SLICE_THREADS |
                             AV_CODEC_CAP_FRAME_THREADS,
    .flush                 = h264_decode_flush,
    .update_thread_context = ONLY_IF_THREADS_ENABLED(ff_h264_update_thread_context),
    .profiles              = NULL_IF_CONFIG_SMALL(ff_h264_profiles),
    .priv_class            = &h264_class,
};

此时， avctx->codec->decode函数指针指向的是h264_decode_frame。如果存在extradata就调用ff_h264_decode_extradata()函数解析，最关键是调用decode_nal_units()函数来解码NAL单元。相关代码如下：

static int h264_decode_frame(AVCodecContext *avctx, void *data,
                             int *got_frame, AVPacket *avpkt)
{
    ......

    ff_h264_unref_picture(h, &h->last_pic_for_ec);
    /* end of stream, output what is still in the buffers */
    if (buf_size == 0)
        return send_next_delayed_frame(h, pict, got_frame, 0);

    if (av_packet_get_side_data(avpkt, AV_PKT_DATA_NEW_EXTRADATA, NULL)) {
        buffer_size_t side_size;
        uint8_t *side = av_packet_get_side_data(avpkt, AV_PKT_DATA_NEW_EXTRADATA, &side_size);
        ff_h264_decode_extradata(side, side_size,
                                 &h->ps, &h->is_avc, &h->nal_length_size,
                                 avctx->err_recognition, avctx);
    }
    if (h->is_avc && buf_size >= 9 && buf[0]==1 && buf[2]==0 && (buf[4]&0xFC)==0xFC) {
		// 解析avcc对应的extradata
        if (is_avcc_extradata(buf, buf_size))
            return ff_h264_decode_extradata(buf, buf_size,
                                            &h->ps, &h->is_avc, &h->nal_length_size,
                                            avctx->err_recognition, avctx);
    }
	// 解码NAL单元
    buf_index = decode_nal_units(h, buf, buf_size);
    if (buf_index < 0)
        return AVERROR_INVALIDDATA;
    if (!h->cur_pic_ptr && h->nal_unit_type == H264_NAL_END_SEQUENCE) {
        av_assert0(buf_index <= buf_size);
        return send_next_delayed_frame(h, pict, got_frame, buf_index);
    }
    if (!(avctx->flags2 & AV_CODEC_FLAG2_CHUNKS) && (!h->cur_pic_ptr || !h->has_slice)) {
        if (avctx->skip_frame >= AVDISCARD_NONREF ||
            buf_size >= 4 && !memcmp("Q264", buf, 4))
            return buf_size;
        return AVERROR_INVALIDDATA;
    }
    if (!(avctx->flags2 & AV_CODEC_FLAG2_CHUNKS) ||
        (h->mb_y >= h->mb_height && h->mb_height)) {
        if ((ret = ff_h264_field_end(h, &h->slice_ctx[0], 0)) < 0)
            return ret;

        /* Wait for second field. */
        if (h->next_output_pic) {
            ret = finalize_frame(h, pict, h->next_output_pic, got_frame);
            if (ret < 0)
                return ret;
        }
    }

    ff_h264_unref_picture(h, &h->last_pic_for_ec);
    return get_consumed_bytes(buf_index, buf_size);
}

三、avcodec_decode_video2视频解码

由于avcodec_decode_video2()函数已经过时，所以内部提供compat_decode()来兼容旧版本，具体代码如下：

int avcodec_decode_video2(AVCodecContext *avctx, AVFrame *picture,
                                              int *got_picture_ptr,
                                              const AVPacket *avpkt)
{
    return compat_decode(avctx, picture, got_picture_ptr, avpkt);
}

我们来看看compat_decode()函数源码，其实内部也是调用avcodec_send_packet()和avcodec_receive_frame()进行解码，具体如下：

static int compat_decode(AVCodecContext *avctx, AVFrame *frame,
                         int *got_frame, const AVPacket *pkt)
{
    AVCodecInternal *avci = avctx->internal;
    int ret = 0;
    *got_frame = 0;
	
    if (avci->draining_done && pkt && pkt->size != 0) {
        avcodec_flush_buffers(avctx);
    }
    if (avci->compat_decode_partial_size > 0 &&
        avci->compat_decode_partial_size != pkt->size) {
        ret = AVERROR(EINVAL);
        goto finish;
    }
    if (!avci->compat_decode_partial_size) {
		// 调用avcodec_send_packet发送packet
        ret = avcodec_send_packet(avctx, pkt);
        if (ret == AVERROR_EOF)
            ret = 0;
        else if (ret == AVERROR(EAGAIN)) {
            /* we fully drain all the output in each decode call, so this should not
             * ever happen */
            ret = AVERROR_BUG;
            goto finish;
        } else if (ret < 0)
            goto finish;
    }

    while (ret >= 0) {
		// 循环调用avcodec_receive_frame接收frame
        ret = avcodec_receive_frame(avctx, frame);
        if (ret < 0) {
            if (ret == AVERROR(EAGAIN) || ret == AVERROR_EOF)
                ret = 0;
            goto finish;
        }

        if (frame != avci->compat_decode_frame) {
            if (!avctx->refcounted_frames) {
                ret = unrefcount_frame(avci, frame);
                if (ret < 0)
                    goto finish;
            }

            *got_frame = 1;
            frame = avci->compat_decode_frame;
        } else {
            if (!avci->compat_decode_warned) {
                avci->compat_decode_warned = 1;
            }
        }

        if (avci->draining || (!avctx->codec->bsfs 
			&& avci->compat_decode_consumed < pkt->size)) {
            break;
		}
    }

finish:
    if (ret == 0) {
        /* if there are any bsfs then assume full packet is always consumed */
        if (avctx->codec->bsfs)
            ret = pkt->size;
        else
            ret = FFMIN(avci->compat_decode_consumed, pkt->size);
    }
    avci->compat_decode_consumed = 0;
    avci->compat_decode_partial_size = (ret >= 0) ? pkt->size - ret : 0;

    return ret;
}

四、avcodec_decode_audio4音频解码

和avcodec_decode_video2()函数一样，avcodec_decode_audio4()函数已经过时，内部也是提供compat_decode()来兼容旧版本，具体代码如下：

int avcodec_decode_audio4(AVCodecContext *avctx,
                                              AVFrame *frame,
                                              int *got_frame_ptr,
                                              const AVPacket *avpkt)
{
    return compat_decode(avctx, frame, got_frame_ptr, avpkt);
}

至此，avcodec_send_packet()和avcodec_receive_frame()组成的解码函数已经分析完毕。

学习FFmpeg与代码实践，可参考：FFmpegAndroid