[vdr] Re: Exit state handling in runvdr

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

Steffen Barszus wrote:
> 
> Am Dienstag, 26. August 2003 23:22 schrieb Rene Bartsch:
> > Am Die, 2003-08-26 um 18.54 schrieb Klaus Schmidinger:
> > > > Hmmm. From your own definition, the current would be the correct.
> > > > Any exitcode other than that three I would suspect to be a bug.
> > > > Otherwise the one you mention is correct too and the recover from
> > > > an error is safer.
> > >
> > > If VDR exits in any "normal" way, it will return one of the exit
> > > states 1, 2 or 3, respectively. What I was worried about was if it
> > > exists for some unexpected reason, like a segfault, for instance.
> >
> > It seems I've lost the plot.
> >
> > If I understand well, VDR exits with 0 at normal shutdown, 1 when it
> > has to be restarted, and 2 when a config-file is broken (but you told
> > broken config-files will be ignored now ...).
> 
> 1, 2 and 3 was a typo of Klaus I guess.

Yes, sorry, that should, of course, have been 0, 1 and 2.

> 0,1 and 2 remainig valid. What
> Klaus have changed is, that he does not prove on exitcode 1 for restart
> but 0 and 2 for exit and all other for restart. If vdr can't start
> (i.e. missing plugin) exitcode 2 will be issued.

Exactly. Just check all the 'return' statements in VDR and the 'exit()'
calls in the entire program to see which exit codes are possible.
There should be only 0,1 and 2 - everything else means an unexpected
error and should trigger a restart.

Klaus


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