[FFmpeg-devel] [PATCH] avdevice/decklink_dec: extracting and outputing klv from vanc

[Zivkovic, Milos]

I'll submit the patch with the proposed changes once that build issue is
resolved.

I'm not having any build issues on multiple compilers, and there's one
really weird error in the log that you shared.

In file included from /usr/include/c++/7/ext/new_allocator.h:33:0,
>                  from
> /usr/include/c++/7/x86_64-suse-linux/bits/c++allocator.h:33,
>                  from /usr/include/c++/7/bits/allocator.h:46,
>                  from /usr/include/c++/7/vector:61,
>                  from libavdevice/decklink_dec.cpp:25:
> ./new:1:1: error: expected unqualified-id before ‘[’ token
>  [PACKET]
>  ^
>

I'd say that basically everything goes to hell due to that error and
invalid token [PACKET] located in /usr/include/c++/7/new, on the first
line.
Do you mind checking the file and confirming that it actually contains it?
Those kinds of tokens shouldn't really ever be in the standard headers.

I created a very simple sample on Compiler Explorer (with GCC 7.5) where
token [PACKET] is above the includes.
Basically everything after it is treated as an error, just like in the
shared log.
You can check it out here: https://godbolt.org/z/zTJn_9

Thanks,
Milos



On Mon, Jun 15, 2020 at 9:20 PM Marton Balint <cus at passwd.hu> wrote:

>
>
> On Mon, 15 Jun 2020, Zivkovic, Milos wrote:
>
> > Hello,
> >
> > I appreciate the comments.
> >
> > Missing docs for the new option.
> >>
> >
> > Does this refer to .texi files? I'll update the docs.
>
> Yes.
>
> >
> >
> >> Avoid these if possible. Yes, this is C++ code but only because the
> >> decklink
> >> API requires it. Also it gives me a compile error, and I am not familiar
> >> enough
> >> with C++ to figure out what is going on
> >>
> >
> > It makes everything much easier (just like using std::atomic for ref
> > counting does).
> > It also doesn't change the needed version of C++ since std::atomic is
> > available in C++11 and above.
> > However, that error message is concerning and I'd like to fix it. Which
> > compiler is it?
>
> GCC 7.5.0, full compiler log is attached.
>
> >
> > You only use this to avoid calling ->Release() right? If that is the
> case,
> >> then
> >> I'd avoid this as well, I am not sure about the compiler support of such
> >> templating.
> >>
> >
> > Yes, it's used to avoid manually releasing COM pointers which can become
> > quite a hassle especially if multiple COM objects are involved.
> > Error handling is simplified as well since there's no need to worry which
> > COM object(s) should be released.
> > As far as compiler support goes, this is the most basic C++ template
> that's
> > supported by basically every C++ compiler in the last 20+ years.
> > I'm willing to remove it if you still feel strongly that it's
> unnecessary.
>
> Yes please.
>
> >
> > This interface is only available from DeckLink SDK 10.10 or so, but you
> >> have not
> >> bumped the version requirement in configure.
> >>
> >
> > Wasn't sure which files should be edited for this.
> > I'll get the exact SDK version that introduces it and bump it in
> configure.
> >
> > Is the sorting really needed? Aren't they supposed to be sorted in the
> >> VANC?
> >>
> >
> > Well, they should be, but it's not guaranteed so this is just a safety
> net.
> > If it was guaranteed there would be no need for PSC fields.
> >
> > Maybe avoid the error message and increase ctx->dropped like other
> similar
> >> code
> >> does it.
> >>
> >
> > Will do.
> >
> > Mixed code and declaration.
> >>
> >
> > I'll move that block below the declaration.
> >
> > Lose the error message for ENOMEM.
> >>
> >
> > I was just following what the other code that adds streams does.
> >
> > AV_OPT_TYPE_BOOL
> >>
> >
> > I'll change that asap.
> >
> > Thanks,
> > Milos
>
> Thanks,
> Marton_______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel at ffmpeg.org
> https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>
> To unsubscribe, visit link above, or email
> ffmpeg-devel-request at ffmpeg.org with subject "unsubscribe".