[FFmpeg-devel] [Fwd: Re: [PATCH] libavcodec/mmaldec.c: add interlaced_frame and format struct to AVFrame for deinterlacing]

[Jens Ziller]

Hello Rodger and wm4,

for interlaced frames I need AVFrame->interlaced_frame. For invoke MMAL
components like deinterlacer and renderer added a pointer data[2] to
existing MMAL_ES_FORMAT_T. I don't have find a better solution to give
a pointer to user app. I have added a patch as suggestion. Please help 
me that I can release my code.

regards Jens


-------- Weitergeleitete Nachricht --------
Von: Michael Niedermayer <michael at niedermayer.cc>
Reply-to: FFmpeg development discussions and patches <ffmpeg-
devel at ffmpeg.org>
An: FFmpeg development discussions and patches <ffmpeg-devel at ffmpeg.org
>
Betreff: Re: [FFmpeg-devel] [PATCH] libavcodec/mmaldec.c: add
interlaced_frame and format struct to AVFrame for deinterlacing
Datum: Sat, 16 Jul 2016 14:41:50 +0200

On Sat, Jul 16, 2016 at 11:08:12AM +0200, Jens Ziller wrote:
> 
> Am Sonntag, den 03.07.2016, 19:36 +0200 schrieb Jens Ziller:
> > 
> > Am Sonntag, den 03.07.2016, 18:05 +0200 schrieb Moritz Barsnick:
> > > 
> > > 
> > > On Sun, Jul 03, 2016 at 17:20:41 +0200, Jens Ziller wrote:
> > > > 
> > > > 
> > > > 
> > > > Am Samstag, den 02.07.2016, 17:54 +0200 schrieb Moritz
> > > > Barsnick:
> > > > > 
> > > > > 
> > > > > 
> > > > > On Sun, Jun 26, 2016 at 17:12:14 +0200, Jens Ziller wrote:
> > > > > > 
> > > > > > 
> > > > > > 
> > > > > > 
> > > > > > +        ctx->interlaced_frame = !(interlace_type.eMode ==
> > > > > > MMAL_InterlaceProgressive);
> > > > > What's wrong with using the "!=" operator instead?
> > > > "!=" is a comparing. "= !()" assign with a negate. Here is "=
> > > > !()"
> > > > needed.
> > > I meant the comparison, not the assignment, so replacing:
> > >   ctx->interlaced_frame = !(interlace_type.eMode ==
> > > MMAL_InterlaceProgressive)
> > > with
> > >   ctx->interlaced_frame = (interlace_type.eMode !=
> > > MMAL_InterlaceProgressive)
> > > 
> > > The former is rather ... convoluted.
> > > (Brackets optional, but probably better for readability.)
> > > 
> > > Moritz
> > Oh, sorry! I'am so blind. Yes, that's not really smart. I changed
> > that.
> > The new patch is attached.
> > 
> > Regards Jens
> > 
> Hello Michael and list,
> 
> this patch was discussed and improved on the ML. What can I still do
> that my app get interlaced_frame and a pointer to the existing
> 
> MMAL_ES_FORMAT_T from ffmpeg? mmaldec.c have unfortunately no
> maintainer.
But it has authors,
you can ask rodger combs and wm4

If they do not reply then please explain why you need this structure
in data[2]
Also what is the lifetime of this structure and what is the liftime
of the AVFrame that contains it. Its neccessary that the struct stays
valid as long as the AVFrame could be

[...]

-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Observe your enemies, for they first find out your faults. --
Antisthenes
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel at ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-v4-fill-AVFrame-interlaced_frame-with-MMAL_PARAMETER.patch
Type: text/x-patch
Size: 3113 bytes
Desc: not available
URL: <http://ffmpeg.org/pipermail/ffmpeg-devel/attachments/20160716/e5f89294/attachment.bin>