Re: Kernel panic with 4.99.64 kernel on an iBook SE

["David H. Gutteridge" <dhgutteridge%sympatico.ca@localhost>]

Michael Lorenz wrote:
> On Jun 6, 2008, at 00:19, David H. Gutteridge wrote:
>
> > > Hi all,
> > >
> > > I thought I'd test the latest -current kernel, per Michael's  request to
> > > check things out in anticipation of the 5.0 branch.  A kernel from  last
> > > night's releng build panics on boot during disk activity, like so:
> >
> > I'm not sure if this is much more useful, but I built a kernel with
> > some debugging enabled/added for envsys (and I commented out the
> > panic() call so I could actually save a dmesg), as below.
> >
> > So it looks like the CPU is getting picked up in the sensor framework.
>
> Ah yes, I forgot you've got a G3 with built-in sensor ;)
>
> > sme_event_register: (cpu0) event registered (sensor= snum=8907424  
> > type=110 critval=0)
> > sme_events_init: events framework initialized for 'cpu0'
> > sme_event_register: (cpu0) event registered (sensor= snum=8907424  
> > type=190 critval=0)
>
> Hmm, the sensor number looks bogus. I'll have a look at the code and  fire 
> up my beige G3 - maybe I can reproduce it.

I've PRed this as kern/38994.  Nothing new to add at my end.  I've been
meaning to check what's involved in simply not including the sensor
framework in compiles; I suspect it may be complicated.

Dave