[OpenAFS] why does osi_Panic oops ?

[Bart Banter]

It seems to me, then, that the AFS kernel code should distinguish between 
"This needs to be debugged!" errors and "Don't take another step!" errors.
In the Linux kernel, panic() is called when continuing may cause disk 
corruption. It would be good for the AFS code to make the same distinction. 
If osi_Panic() is ever used where panic() ought to be called, perhaps an 
osi_Oops() routine for the other cases would be a good idea?

--
Guy Streeter
Red Hat

>From: Jeffrey Hutzelman <jhutz@cmu.edu>
>To: Derrick J Brashear <shadow@dementia.org>
>CC: openafs-info@openafs.org
>Subject: Re: [OpenAFS] why does osi_Panic oops ?
>Date: Fri, 12 Oct 2001 09:21:47 -0400 (EDT)
>
>On Thu, 11 Oct 2001, Derrick J Brashear wrote:
>
> > On Thu, 11 Oct 2001, Bart Banter wrote:
> >
> > > Why does the osi_Panic() routine force an oops instead of calling the 
>kernel
> > > panic() routine?
> >
> > It's possible (though problematic) to use a machine on which AFS has
> > oops'd. Once it panics, you don't have that option. It is probably the
> > case that most people just want their machine to reboot and we should
> > panic() but this is the first time it's come up.
>
>Always calling panic() is not a viable option.  A machine on which AFS has
>oops'd is almost always still working well enough for the oops (and
>symbol translatios) to be logged.  That information can be critical in
>tracking down a problem, and it's the reason that Linux avoids panicing
>except when absolutely necessary.
>
>_______________________________________________
>OpenAFS-info mailing list
>OpenAFS-info@openafs.org
>https://lists.openafs.org/mailman/listinfo/openafs-info


_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp