According to Thorsten Kranzkowski: While burning my CPU.
>
> Hello!
>
> > I'm workink on a 2.2.9 series kernel,
> > and everything goes well except
> > ifconfig ax0 down ...
> >
> > ...that makes kernel crash
>
> I've seen this, too. I can reproduce it with
>
> call bcsf0 db0hsk # connect to someone
> -- switch off trx -- # so no replys can be received
> ~. # disconnect - connection stays in DISC sent state
> # waiting for DM
> ifconfig bcsf0 down # oops following ....
>
> The oops is usually not immediate but with an up to 4 second delay!
>
> >
> > (redhat-6.0)
>
> Homemade system(s) here. (Alpha and i486)
> currently kernel 2.2.8, but I already noticed this some time ago.
>
> >
> > Have you any idea why?
>
> Yes :-)
> >From what I remember I found out from the trace that the oops happens
> because there was a timer left over whose handler doesn't exist any more
> when the device is down. Maybe the 'heartbeat' timer?
>
> If there is demand for more detailed info I will investigate this again.
Yes, this is a __real__ problem, i can also reproduce it here on a bpq
interface, i did;
call bpq0 pa3gcu-9
after the connection was established i did;
ifconfig bpq0 down
the only means of escape was the reset button, the oops was written to the
console but there is no trace of it in the system logs.
This renders a very big problem indeed for 2.2.x systems in remote
locations, imagen the effect _if_ an scc interface locks up, or any other
for that matter, there is then no way to restart it.
I will also look and see if i can get some information when i have more
time, i did not write anything on paper from the oops, as i expected it to
be in the log, it seems i expected to much...
>
> >
> > Max, IK0XKP (Rome, Italy)
> >
>
> 73 Thorsten
>
> --
> | Thorsten Kranzkowski Snail: Niemannsweg 30, 49201 Dissen, Germany |
> | Mobile: ++49 161 7210230 Inet: [EMAIL PROTECTED] |
> | Ampr: dl8bcu@db0nei.#nrw.deu.eu, [EMAIL PROTECTED] [44.130.8.19] |
>
--
Regards Richard.
[EMAIL PROTECTED]
