firewire in 2.4.21-ben2

[Benjamin Herrenschmidt]

On Sat, 2003-07-26 at 10:39, Calum Selkirk wrote:
> * Calum Selkirk [cselkirk at xs4all.nl] [2003-07-17 12:11 +0200]:
>
> > * Benjamin Herrenschmidt [benh at kernel.crashing.org] [2003-07-15 11:30 +0200]:
> >
> > [snip]
> > > >   After I reboot back in 2.4.20-ben10 everything works fine...
> > >
> > > That's a change in the firewire layer, you have to trigger a scsi
> > > probe yourself now. Either using the rescan-scsi-bus.sh script, or the
> > > scsiadd command for example
> >
> > just FYI, I have had the kernel oops on one occasion when rescaning the
> > scsi bus this way (with a 2.4.21-ben2). Sorry, wasn't able to get a
> > backtrace, though I will attempt to if it happens again.
>
> OK .. attatched is a backtrace (hopefully accuratly transcribed) and
> snippit from /var/log/messages.
>
> As i said previously I haven't been able to reproduce this, it's only
> happened to me on one occassion (the above backtrace is from another
> user/machine). Might this be related to the other ieee1394 bug, where
> unloading ohci1394 after sleep will oops?

A backtrace with just numbers is useless, you should either provide
insmod -m output or run the oops through ksymoops.


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/