[linux-dvb] Re: Problems with WinTV NOVA-CI

[Andrew de Quincey <adq_dvb@lidskialf.net>]

On Monday 13 Dec 2004 13:32, Stefan Taferner wrote:
> On Monday 13 December 2004 14:20, Andrew de Quincey wrote:
> > On Monday 13 Dec 2004 12:57, Werner Hauger wrote:
> > > > I have looked through the mailing list archives from this year, and
> > > > found 2 or 3 posts about the same problem, however there was no
> > > > resolution posted. Is anybody here using a standalone DVB-S TT budget
> > > > card with the dvb-kernel drivers and a 2.4 kernel successfully ?
> > > > (Btw. I get the same behaviour with the 2.6.7 kernel included drivers
> > > > under Knoppix).
> > >
> > > Judging from the lack of responses, everybody that has a budget card
> > > (and more specifically the exact one I have) is using it in
> > > conjunction with a FF one. Isn't that maybe the reason only I now have
> > > this problem i.e. the driver/firmware of the FF card does something to
> > > make the budget card work, which is lacking in my case.
> >
> > The problem is (I think) none of the developers have seen the issue you
> > describe.
> >
> > I have *many* TT Nova-S (stv0299) cards working 24 hours a day (they're
> > the same physical card as the TT Nova-CI, minus the CI components). The
> > machines have never had windows on them. I have never seen this issue. I
> > have used the drivers under both 2.4 and 2.6. I have never used a FF
> > card.
> >
> > I also never saw this behaviour when I was developing the budget_ci
> > driver for the TT Nova-S-CI.
> >
> > Is it possible it is a motherboard issue - ACPI or the like?
>
> How can one tell?
>
> Could it be a timing problem?  I remember vaguely that I read about some
> timing issues one year ago. I think it was some I2C issue there, but it
> might be unrelated.
>
> Is there anything that we can do to find out what makes the problems?
> What could be missing that is fixed with a reboot but is not fixed with
> multiple driver loading/unloading?
>
> Kind regards,
> Stefan
>
> PS: I am already running bleeding edge Linux kernel and DVB drivers, but it
> does not change anything.

Werner said he'd compared the saa7146 PCI memory between working and 
non-working and there wasn't really an obvious difference. On the Nova-S 
cards there isn't really anything else you can control besides the demod and 
PLL - and someone said that tuning works OK....

This sounds _extremely_ like the issue many people who use VDR with an FF and 
a budget card have - after a bit the budget card just stops delivering data, 
but it can still tune OK. (the FF card works fine though). A reboot (although 
not to windows) is required to fix it.

/proc/interrupts showed that when this issue with FF+budget occurs, the budget 
simply doesn't generate interrupts any more.

Lots of different things have been tried with the FF issue - none of them have 
worked so far.

I'm beginning to think that we're using the saa7146 in some way that makes it 
fall over occasionally.