[mythtv] OpenGL: MythRenderOpenGL is not a valid OpenGL rendering context

Mark Kendall mark.kendall at gmail.com
Fri Jan 31 10:23:09 UTC 2020


On Thu, 30 Jan 2020 at 20:40, John Pilkington <johnpilk222 at gmail.com> wrote:
>
> On 30/01/2020 17:01, Mark Kendall wrote:
> > On Thu, 30 Jan 2020 at 14:26, John Pilkington <johnpilk222 at gmail.com> wrote:
> >> Hi Mark:  It didn't make an obvious difference.  Here's a link to logs
> >> with and without the EV preamble.
> >
> > OK - thanks for that.
> >
> > Can you produce a few more logs for me? :-
> >
> > 1 - the output from running 'QT_LOGGING_RULES=qt.qpa.*=true
> > mythfrontend' - which should hopefully give an insight into what Qt is
> > doing
> > 2 - the output from 'ldd /usr/local/bin/mythfrontend' - or wherever
> > mythfrontend is installed.
> > 3 - the output from 'glxgears -info' - because I'm still baffled as to
> > why glxinfo shows intel and Qt/mythtv something else.
> >
> > thanks again,
> > Mark
>
> HTH!
>
> https://drive.google.com/open?id=1NowIZsgY4EFSSKcKAglXFkVEb-U07A3T
>

Thanks John,

Unfortunately it doesn't provide any more insights. I can't help but
feel there is something a little broken with your setup as I see no
reason why Qt uses a software OpenGL renderer - which I'm assuming is
the main problem.

I'll give it some more thought but will leave the code as is for the
moment to see if anyone else has issues.

I don't really want to disable using EGL unless I have to as a lot of
people will get poor or no VAAPI playback without it.

I'll also see if I can reproduce the crash when it fails - so at least
it properly falls back to the Qt renderer.

Thanks and regards
Mark


More information about the mythtv-dev mailing list