[vdr] Re: For VDR-1.2.0: VDR should disable broken cards

[Klaus Schmidinger <Klaus.Schmidinger@cadsoft.de>]

Reinhard Walter Buchner wrote:
> 
> Hi Klaus,
> 
> > > Having VDR restart when the primary card has a reception problem
> > > may be a good idea, but I think VDR should *not* restart, just
> > > because a background scan of the EPG fails. Rather it should simply
> > > jump over the channel which doesn't deliver anything and continue
> > > with the next one
> 
> > Are you sure that VDR resarts in such a case?
> > I can only see emergency exits during actual recordings or if the
> > main program loop hangs.
> 
> Yes ;o)) 100% sure in case of the 1.1.17. I don't know what
> you have changed since then, but if I watch a movie on one
> card and the EPG scan kicks in my tech patch nicely shows
> the signal strength for each channel scan and as soon as
> VDR reaches non available channel, signal strength goes
> down to almost zero and the logs show "panic can't sync on
> channel "xyz".

kls@hawk2:/home/kls/vdr/VDR > grep -i "panic" *.c
vdr.c:  esyslog("PANIC: watchdog timer expired - exiting!");
kls@hawk2:/home/kls/vdr/VDR > 

Now where exactly do you suppose that string comes from?
I'd say _not_ from VDR!

> Sooner or later VDR restarts. Same thing goes
> for watching a (already finished) recording. AFAIR the same
> thing happend in the 1.1.28 version (last version I had time
> to play with)
> 
> > > Pretty much the same goes if one of the TV stations decides to
> > > change any of the PIDs. Obviously in such a case, VDR cannot
> > > tune to the channel and conks out.
> 
> > I could undertsand this if it were recording, but not during the EPG
> > scan.
> 
> > > What I think VDR should do is:
> > > - on an EPG scan: simply ignore channels that don't deliver any data
> > >   (this would also prevent a current recording from becoming messed
> > >   up just because the EPG scan fails --> I'd rather have no EPG data
> > >   than a junked recording)
> 
> > Where exactly does the EPG scan cause this?
> 
> Sorry, I never investigated the true code call that actually causes
> this problem I just reconneted the antenna after VDR restarted
> 3 times in a row ;o))

A good starting point might be the 'grep' I've shown above.
I'd be surprized if it showed anything else on your system...
Of course, that's assuming you're using VDR 1.1.33 - which is the only one
I'm talking about.

Klaus
-- 
_______________________________________________________________

Klaus Schmidinger                       Phone: +49-8635-6989-10
CadSoft Computer GmbH                   Fax:   +49-8635-6989-40
Hofmark 2                               Email:   kls@cadsoft.de
D-84568 Pleiskirchen, Germany           URL:     www.cadsoft.de
_______________________________________________________________


-- 
Info:
To unsubscribe send a mail to ecartis@linuxtv.org with "unsubscribe vdr" as subject.