From 8a60bba0aef77015111570058d5a72f0428dc748 Mon Sep 17 00:00:00 2001 From: wm4 Date: Mon, 6 Mar 2017 11:34:20 +0100 Subject: [PATCH] avcodec: clarify some decoding/encoding API details Make it clear that there is no timing-dependent behavior. In particular, there is no state in which both input and output are denied, and where you have to wait for a while yourself to make progress (apparently some hardware decoders like to do this). Avoid wording that makes references to time. It shouldn't be mistaken for some kind of asynchronous API (like POSIX read() can return EAGAIN if there is no new input yet). It's a state machine, so try to use appropriate terms. Signed-off-by: Diego Biurrun --- libavcodec/avcodec.h | 44 +++++++++++++++++++++++++++++++++++--------- 1 file changed, 35 insertions(+), 9 deletions(-) diff --git a/libavcodec/avcodec.h b/libavcodec/avcodec.h index 489a618cbb..c03754695f 100644 --- a/libavcodec/avcodec.h +++ b/libavcodec/avcodec.h @@ -114,6 +114,12 @@ * are filled. This situation is handled transparently if you follow the steps * outlined above. * + * In theory, sending input can result in EAGAIN - this should happen only if + * not all output was received. You can use this to structure alternative decode + * or encode loops other than the one suggested above. For example, you could + * try sending new input on each iteration, and try to receive output if that + * returns EAGAIN. + * * End of stream situations. These require "flushing" (aka draining) the codec, * as the codec might buffer multiple frames or packets internally for * performance or out of necessity (consider B-frames). @@ -148,7 +154,8 @@ * Unlike with the old video decoding API, multiple frames might result from * a packet. For audio, splitting the input packet into frames by partially * decoding packets becomes transparent to the API user. You never need to - * feed an AVPacket to the API twice. + * feed an AVPacket to the API twice (unless it is rejected with EAGAIN - then + * no data was read from the packet). * Additionally, sending a flush/draining packet is required only once. * - avcodec_encode_video2()/avcodec_encode_audio2(): * Use avcodec_send_frame() to feed input to the encoder, then use @@ -161,7 +168,22 @@ * and will result in arbitrary behavior. * * Some codecs might require using the new API; using the old API will return - * an error when calling it. + * an error when calling it. All codecs support the new API. + * + * A codec is not allowed to return EAGAIN for both sending and receiving. This + * would be an invalid state, which could put the codec user into an endless + * loop. The API has no concept of time either: it cannot happen that trying to + * do avcodec_send_packet() results in EAGAIN, but a repeated call 1 second + * later accepts the packet (with no other receive/flush API calls involved). + * The API is a strict state machine, and the passage of time is not supposed + * to influence it. Some timing-dependent behavior might still be deemed + * acceptable in certain cases. But it must never result in both send/receive + * returning EAGAIN at the same time at any point. It must also absolutely be + * avoided that the current state is "unstable" and can "flip-flop" between + * the send/receive APIs allowing progress. For example, it's not allowed that + * the codec randomly decides that it actually wants to consume a packet now + * instead of returning a frame, after it just returned EAGAIN on an + * avcodec_send_packet() call. * @} */ @@ -4339,8 +4361,10 @@ int avcodec_decode_subtitle2(AVCodecContext *avctx, AVSubtitle *sub, * a flush packet. * * @return 0 on success, otherwise negative error code: - * AVERROR(EAGAIN): input is not accepted right now - the packet must be - * resent after trying to read output + * AVERROR(EAGAIN): input is not accepted in the current state - user + * must read output with avcodec_receive_frame() (once + * all output is read, the packet should be resent, and + * the call will not fail with EAGAIN). * AVERROR_EOF: the decoder has been flushed, and no new packets can * be sent to it (also returned if more than 1 flush * packet is sent) @@ -4361,7 +4385,7 @@ int avcodec_send_packet(AVCodecContext *avctx, const AVPacket *avpkt); * * @return * 0: success, a frame was returned - * AVERROR(EAGAIN): output is not available right now - user must try + * AVERROR(EAGAIN): output is not available in this state - user must try * to send new input * AVERROR_EOF: the decoder has been fully flushed, and there will be * no more output frames @@ -4394,8 +4418,10 @@ int avcodec_receive_frame(AVCodecContext *avctx, AVFrame *frame); * avctx->frame_size for all frames except the last. * The final frame may be smaller than avctx->frame_size. * @return 0 on success, otherwise negative error code: - * AVERROR(EAGAIN): input is not accepted right now - the frame must be - * resent after trying to read output packets + * AVERROR(EAGAIN): input is not accepted in the current state - user + * must read output with avcodec_receive_packet() (once + * all output is read, the packet should be resent, and + * the call will not fail with EAGAIN). * AVERROR_EOF: the encoder has been flushed, and no new frames can * be sent to it * AVERROR(EINVAL): codec not opened, refcounted_frames not set, it is a @@ -4413,8 +4439,8 @@ int avcodec_send_frame(AVCodecContext *avctx, const AVFrame *frame); * encoder. Note that the function will always call * av_frame_unref(frame) before doing anything else. * @return 0 on success, otherwise negative error code: - * AVERROR(EAGAIN): output is not available right now - user must try - * to send input + * AVERROR(EAGAIN): output is not available in the current state - user + * must try to send input * AVERROR_EOF: the encoder has been fully flushed, and there will be * no more output packets * AVERROR(EINVAL): codec not opened, or it is an encoder