[FFmpeg-devel] [PATCH] Fixed memory leak when using avformat_new_stream

Michael Niedermayer michaelni at gmx.at
Wed Jul 24 12:58:45 CEST 2013


On Wed, Jul 24, 2013 at 12:30:46PM +0530, Anshul wrote:
> On 07/23/2013 11:09 PM, Michael Niedermayer wrote:
> >On Tue, Jul 23, 2013 at 06:37:17PM +0200, Jean-Baptiste Kempf wrote:
> >>On 23 Jul, Michael Niedermayer wrote :
> >[...]
> >>>Iam cc-ing jb as he might want to know about this though it can be
> >>>argued about if theres a bug currently or not. I am not aware of any
> >>>user vissible issue, only if this close call is added do problems
> >>>occur as more locking is than needed which would be automagically
> >>>be provided through a lock manager if one was set ...
> >>vlc 2.0 branch does not support lavc 55, because of the constant
> >>breakages of API, ABI and behaviours...
> >>
> >>vlc 2.1-git branch and trunk branch work perfectly fine with lavc 55,
> >>since we plan to release vlc 2.1 with this version.
> >>
> >>No idea what this is about...
> >Its about multiple things :)
> >
> >the reason why i cc-ed you was just that i wanted to make sure that
> >you are aware of that vlc doesnt seem to use av_lockmgr_register()
> >and this could potentially be a bad decission but i dont know why
> >its not used so i could easily be missing something here.
> >
> >[...]
> >
> >
> >
> >_______________________________________________
> >ffmpeg-devel mailing list
> >ffmpeg-devel at ffmpeg.org
> >http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
> Hi Michael
> 
> I have attached a patch that document, avformat_new_stream() that
> avcodec_close()
> is required to free the whole context.
> mean While I was looking for a better way, so that vlc do not crash.

> If you think looking at issue created by old patch is waste of time,
> and of no use to community
> then attached patch is enough to close the bug.

if you or someone else posts a patch that fixes the leak without
the user having to explicitly close the codec that is surely
interresting, until then doc patch applied

Thanks

[...]

-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

No great genius has ever existed without some touch of madness. -- Aristotle
-------------- 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/20130724/f9a6d439/attachment.asc>


More information about the ffmpeg-devel mailing list