Re: Forwarded mail....


From: Amon Ott <ao@rsbac.org>
Subject: Re: Forwarded mail....
Date: Mon, 12 Mar 2001 17:13:07 +0100

Next Article (by Author): Re: something about INIT too. Amon Ott
Previous Article (by Author): Re: v1.1.1-pre5 lookup device problems Amon Ott
Top of Thread: Forwarded mail.... Bencsath Boldizsar
Articles sorted by: [Date] [Author] [Subject]


On Mon, 12 Mär 2001 Bencsath Boldizsar wrote:
> As some members wrote too, problems with RAID, again:
> 
> I tried 2.2.18+pre4, it seems no to be fully o.k., then now I tried
> 2.4.2+pre5. 
> I have a test raid-1 mirrored disk /dev/hda5 + /dev/hda6 (just a test!),
> with reiserfs (3.6, 2.4.2+the latest patch on www.reiserfs.org)
> 
> Sometimes I used the reiserfs disk, but then I just turned it off from
> /etc/fstab. RAID is turned on in the kernel.
> 
> Problems:
> 
> -At restarting the system (after all gets killed) EINVALIDDEV and some
> other rsbac execptions on the screen.

Can you see what device is called invalid? If yes, what is mounted on it?

> -At the system start about 50% chance to get a kernel oops. If not, then
> the system is functional. Doing systematic reboots without anything other
> the oops is not deterministic,seems to be fully random. 
> 
> I just copied the kernel oops report by hand, anyone could tell me in a
> priv letter how to save it to a file for mailing in a cheaper way?
> I can send the kernel image or anything if anyone wishes...

A serial console on another system is very handy for that - if you have another
system that is.

> Below the oops, I didn't checked if everything ever is the same, but the
> eip seems alway be the same, EFLAGS is sometimes 00010004.
> 
> rsbac_init(): Setting RSBAC auto timer
> Unable to handle kernel NULL pointer dereference at virtual address 00000004

This looks like the old problem we had with SMP systems. It might be the kernel
stack size that goes over the limit if too many things happen during init.

I will see how I could reduce the stack pressure a bit more, but I am not sure
whether it will help here.

>  printing eip:
> c011c4b5

What function is there (see System.map)?

Amon.
-
To unsubscribe from the rsbac list, send a mail to
majordomo@rsbac.org with
unsubscribe rsbac
as single line in the body.

Next Article (by Author): Re: something about INIT too. Amon Ott
Previous Article (by Author): Re: v1.1.1-pre5 lookup device problems Amon Ott
Top of Thread: Forwarded mail.... Bencsath Boldizsar
Articles sorted by: [Date] [Author] [Subject]


Go to Compuniverse LWGate Home Page.