Git Inbox Mirror of the ffmpeg-devel mailing list - see https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
 help / color / mirror / Atom feed
From: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
To: ffmpeg-devel@ffmpeg.org
Subject: Re: [FFmpeg-devel] [PATCH 2/2 v2] avformat/demux: extract extradata from packets when context update is requested
Date: Sun, 21 Apr 2024 18:48:10 +0200
Message-ID: <AS8P250MB0744D640566BE823894028AD8F132@AS8P250MB0744.EURP250.PROD.OUTLOOK.COM> (raw)
In-Reply-To: <d85790d2-e530-4ba3-8f5c-7f9c4323a0cc@gmail.com>

James Almer:
> On 4/21/2024 12:32 PM, Andreas Rheinhardt wrote:
>> James Almer:
>>> If the demuxer doesn't set extradata in the stream's codecpar, a
>>> need_context_update request will delete the previously extracted
>>> extradata in
>>> the stream's internal AVCodecContext.
>>> As we can't ensure the old extradata is valid for the stream in its post
>>> context update request state, try to get extradata from the new
>>> packet instead
>>> of attempting to preserve the old in some form.
>>>
>>> Signed-off-by: James Almer <jamrial@gmail.com>
>>> ---
>>>   libavformat/demux.c | 12 +++++++++++-
>>>   1 file changed, 11 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/libavformat/demux.c b/libavformat/demux.c
>>> index abfd5fee7d..39aa1cd4e3 100644
>>> --- a/libavformat/demux.c
>>> +++ b/libavformat/demux.c
>>> @@ -1319,6 +1319,8 @@ fail:
>>>       return ret;
>>>   }
>>>   +static int extract_extradata(FFFormatContext *si, AVStream *st,
>>> const AVPacket *pkt);
>>> +
>>>   static int read_frame_internal(AVFormatContext *s, AVPacket *pkt)
>>>   {
>>>       FFFormatContext *const si = ffformatcontext(s);
>>> @@ -1373,6 +1375,11 @@ static int read_frame_internal(AVFormatContext
>>> *s, AVPacket *pkt)
>>>                   return ret;
>>>               }
>>>   +            if (!sti->avctx->extradata &&
>>> +                (ret = extract_extradata(si, st, pkt)) < 0) {
>>> +                av_packet_unref(pkt);
>>> +                return ret;
>>> +            }
>>>               sti->codec_desc =
>>> avcodec_descriptor_get(sti->avctx->codec_id);
>>>                 sti->need_context_update = 0;
>>> @@ -2470,6 +2477,8 @@ static int extract_extradata(FFFormatContext
>>> *si, AVStream *st, const AVPacket *
>>>       if (ret < 0)
>>>           return ret;
>>>   +    av_bsf_flush(sti->extract_extradata.bsf);
>>> +
>>>       ret = av_bsf_send_packet(sti->extract_extradata.bsf, pkt_ref);
>>>       if (ret < 0) {
>>>           av_packet_unref(pkt_ref);
>>> @@ -3090,7 +3099,8 @@ find_stream_info_err:
>>>           err = codec_close(sti);
>>>           if (err < 0 && ret >= 0)
>>>               ret = err;
>>> -        av_bsf_free(&sti->extract_extradata.bsf);
>>> +        if (sti->extract_extradata.bsf)
>>> +            av_bsf_flush(sti->extract_extradata.bsf);
>>>       }
>>>       if (ic->pb) {
>>>           FFIOContext *const ctx = ffiocontext(ic->pb);
>>
>> This will keep the BSF around for longer than necessary, although only a
>> tiny minority of demuxers ever set need_context_update at all.
> 
> It's not exactly a massive context, so i don't think it's a problem. But
> if you prefer, i could add an FF_INFMT_FLAG_ for this.

Actually, it is worse: It is legal for the codec id to change upon
need_context_update and then the BSF would still be configured for the
old BSF.
Anyway, an FF_INFMT_FLAG flag is overkill for this: Just reset the
extract_extradata.inited flag.

- Andreas

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

  reply	other threads:[~2024-04-21 16:48 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-04-20 23:43 [FFmpeg-devel] [PATCH 1/2] avcodec/codec_par: always clear extradata_size in avcodec_parameters_to_context() James Almer
2024-04-20 23:43 ` [FFmpeg-devel] [PATCH 2/2] avformat/demux: extract extradata from packets when context update is requested James Almer
2024-04-21 15:18   ` [FFmpeg-devel] [PATCH 2/2 v2] " James Almer
2024-04-21 15:32     ` Andreas Rheinhardt
2024-04-21 15:52       ` James Almer
2024-04-21 16:48         ` Andreas Rheinhardt [this message]
2024-04-22  3:01           ` [FFmpeg-devel] [PATCH 2/2 v3] " James Almer
2024-04-23 13:57 ` [FFmpeg-devel] [PATCH 1/2] avcodec/codec_par: always clear extradata_size in avcodec_parameters_to_context() James Almer

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=AS8P250MB0744D640566BE823894028AD8F132@AS8P250MB0744.EURP250.PROD.OUTLOOK.COM \
    --to=andreas.rheinhardt@outlook.com \
    --cc=ffmpeg-devel@ffmpeg.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

Git Inbox Mirror of the ffmpeg-devel mailing list - see https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

This inbox may be cloned and mirrored by anyone:

	git clone --mirror https://master.gitmailbox.com/ffmpegdev/0 ffmpegdev/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 ffmpegdev ffmpegdev/ https://master.gitmailbox.com/ffmpegdev \
		ffmpegdev@gitmailbox.com
	public-inbox-index ffmpegdev

Example config snippet for mirrors.


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git