[FFmpeg-devel] Trouble with packet duplication/queue

[Michael Niedermayer]

On Wed, Mar 21, 2012 at 04:09:06PM -0600, Deron wrote:
> 
> Trying to fix some odd bugs in the decklink patch I posted awhile
> ago. The last big one seems to be related to the way I am putting
> packets into the queue. I tried using the same basic code from
> ffplay and the various player demos, but they crash a violent death
> (double free).
> 
> here is the code in question from ffplay that does not work for me:
> 
>     if (av_dup_packet(pkt) < 0)
>         return -1;
> 
>     pkt1 = (AVPacketList*)av_malloc(sizeof(AVPacketList));
>     if (!pkt1)
>         return -1;
>     pkt1->pkt = *pkt;
> 
> 
> if instead I do this in my code (more or less borrowed from
> avformat/utils.c as I recall):
> 
>     pkt1 = (AVPacketList *)av_malloc(sizeof(AVPacketList));
>     if (!pkt1)
>         return -1;
>     pkt1->pkt = *pkt;
>     pkt->destruct = NULL;             // do not free original but
> only the copy
>     /* duplicate the packet */
>     if (av_dup_packet(&pkt1->pkt) < 0) {
>         av_free(pkt1);
>         return -1;
>     }
> 
> 
> It works 99% of the time. In fact, it was working so well I thought
> all was well. Until I ran across a set of videos which have a
> _bunch_ of frame with just 75 samples. The audio is also being
> resampled from 30khz to 48khz with -ar 48000 in the ffmpeg line.
> This audio comes out as complete garbage (lot of stuttering noise),
> and sounds like the audio when I had the queue almost right. If I
> first reencode the video to 48k then is plays fine (but it also ends
> up with bigger frames of audio).
> 
> So I guess the first question is what is the proper way to stuff a
> packet into a queue inside an avdevice encoder?

either you copy them or you have to implement something else in
interleave_packet()

[...]
-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Old school: Use the lowest level language in which you can solve the problem
            conveniently.
New school: Use the highest level language in which the latest supercomputer
            can solve the problem without the user falling asleep waiting.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: Digital signature
URL: <http://ffmpeg.org/pipermail/ffmpeg-devel/attachments/20120322/d5876279/attachment.asc>