tech-kern archive

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

Re: NetBSD 6.0/amd64 panic: ciss: dead (DL ProLiant 360P)



On Tue, Oct 23, 2012 at 05:30:39PM +0300, Aleksey Cheusov wrote:
> On Tue, Oct 23, 2012 at 5:19 PM, Michael van Elst 
> <mlelstv%serpens.de@localhost> wrote:
> > On Tue, Oct 23, 2012 at 01:12:29PM +0200, iMil wrote:
> >>
> >> >Mostly says that our driver doesn't talk "correctly" to the hardware.
> >> >The panic in this place is of course pretty silly.
> >> >
> >> >       if (hb == sc->heartbeat)
> >> >               panic("ciss: dead");    /* XX reset! */
> >> >
> >> >Linux just ignores the heartbeat counter. I'd replace the panic with
> >> >a possibly ratechecked printf() and try if that works.
> >>
> >> It works. The controller is now seen and working, I'm not really familiar
> >> with those drivers, is this check bypass safe? May I fill a PR with
> >> this "fix"?
> >
> > Yes, please open a PR.
> 
> Already opened
> 
> kern/47037 [serious/high]: ciss(4) -- kernel panic


"This is very bad, Power cycle is necessary. But this does NOT always
happen. Sometimes (probably boot after reboot ?) no problem at booting."

Could mean that the controller isn't initialized properly or just that
the heartbeat counter rate is not anticipated.

In any case, the panic is silly. If the controller is 'broken' you
should get something like an I/O error or just hang. A panic doesn't
help anyone.

Please replace the panic() with a printf() that reports the counter.
Maybe we can see what goes wrong.


Greetings,
-- 
                                Michael van Elst
Internet: mlelstv%serpens.de@localhost
                                "A potential Snark may lurk in every tree."


Home | Main Index | Thread Index | Old Index