Mailing List archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[linux-dvb] Re: again problems with ALPS TDLB7



On Saturday 15 November 2003 15:05, Juri Haberland wrote:
> Juergen Peitz schrieb:
> > On Friday 14 November 2003 20:54, Juri Haberland wrote:
> >> With the new version 1.14, which - according to the CVS log - includes
> >> some changes by Jürgen, I have a tuner lock-up ever second day.
> >
> > Hmm, I use this version a longer time already without any problems.
>
> I used it now for ~ two weeks, and had lock-ups every second day. I'm
> now back to a patched 1.13 version.
>
> >> Also femon reports only zeros for BER and UNC whereas with the old
> >> version I had varying readings.
> >
> > Is BER and UNC always zero or only after a lockup?
>
> Hmm, I retested with CVS HEAD after I rebooted the computer and now it
> works. So it might be related to a crash.
>
> >> So these changes made things worse or did at least nothing, but they did
> >> not improve the situation. What about Jürgens patch that he send in
> >> August? Why wan't it included?
> >
> > It's already included in version 1.14, so it should work like the patched
> > version 1.13.
> >
> > Sorry, have no idea what could cause this problems, anything in the
> > logfiles?
>
> kernel: sp8870_set_frontend: firmware crash!!!!!!
>
> this often happens when VDR does it's EPG scan...
>
> Cheers,
> Juri

Hi Juri,

did some tests the last days. First I let the card do more than 10000 channel 
switches (one every 3 seconds by a script) but didn't get any firmware crash.
Then I compared version 1.14 with patched version 1.13 and made 500 channel 
switches with both versions and debug enabled:

	patched ver 1.13:
	Nov 21 20:57:44 kernel: sp8870_set_frontend: switches = 500 lockups = 11

	ver 1.14:
	Nov 21 21:47:05 kernel: sp8870_set_frontend: switches = 500 lockups = 11

so exactly the same result and no firmware crash.

Also tested with VDR and EPG scan every hour. Again no firmware crash.

So I really see no difference in stability between a patched version 1.13 and 
the latest CVS driver here.

The only real difference between patched version 1.13 and version 1.14 
relating channel switching is that version 1.13 always used auto detection 
for frontend parameters, but 1.14 uses the values from the channels.conf. So 
is there maybe an error somewhere in your channels.conf? (especially for 
Modulation, Hierachy or Coderate). Maybe with the settings of an channel you 
don't watch so often? That would explain why the crashes mostly happen during 
the EPG scan.  

Juergen




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



Home | Main Index | Thread Index