Re: 2.2.9-ac1 dies during iozone

Thierry Danis (danis@sagem.fr)
Mon, 31 May 1999 21:18:21 +0200

On Mon, May 31, 1999 at 04:30:31AM +0100, Alan Cox wrote:
> > I have tried 2.2.9-ac1 with iozone over NFS today and the client just crashes
> > in a few seconds, without any OOPSES or whatever outputs.
>
> In such cases play with right alt and (print-scr, scrll-lock and pause)
> one of which gives you a register dump. Get about 10 register dumps and
> look the EIP value up in System.map see where it is looping
>
>
> -

OK, I have tried again, and it is 100% reproducible at work
as well.

At the time it dies, the kernel loops in the procedure :
truncate_inode_pages.

It still responds to ping, but does not more.

Here, at work, it happens when I log as danis, issue then
a 'su' to get root access, and type 'exit'. Nothing more. Boom.

My home is NFS mounted from a Solaris 2.5.1 box. At home,
it happens against a 2.0.36 UNFSD server. I think it did
happen as well with 2.2.7-ac4 (I have done a report on it
too few days ago).

One more thing : as I typed repetitive ALT-SCROLL to check
the registers, I got a bunch of addresses printed on the
screens and a :
Code: c7 05 00 00 00 00 00 00 00 00 83 c4 08 5b 5e 5f 5d 83 c4 08
Aiee, killing interrupt handler
Unable to handle kernel NULL pointer derefere,ce at virtual address 00000001

I think it is due to my repetitive ALT-SCROLL, and I would not
bother that much (bad people would simply hit the reset button
however).

-- 
	Thierry Danis
	Poste : 53 53	danis@spmo.sagem.fr


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/