Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 19 Nov 1998 12:57:26 +1030
From:      Greg Lehey <grog@lemis.com>
To:        Harold Gutch <logix@foobar.franken.de>, freebsd-current@FreeBSD.ORG
Subject:   Re: page fault when rm'ing on ccd-partition
Message-ID:  <19981119125726.I467@freebie.lemis.com>
In-Reply-To: <19981118113805.A5521@foobar.franken.de>; from Harold Gutch on Wed, Nov 18, 1998 at 11:38:05AM %2B0100
References:  <19981118113805.A5521@foobar.franken.de>

next in thread | previous in thread | raw e-mail | index | archive | help
On Wednesday, 18 November 1998 at 11:38:05 +0100, Harold Gutch wrote:
> Hi,
>
> I'm experiencing problems on a (non-local) 3.0-19980916-SNAP
> non-SMP System. When rm'ing certain files from a ccd'd disk the
> machine panics. I can't really localize the problem any further
> as:
>
> - I don't know which files exactly cause the problem, or what
>   they have in common,
>      find -mtime 7 | xargs rm
>   over one of the ccd-partitions reproducebly panics the machine
>   though.
>
> - A person local to the machine wrote down the following
>   panic-output:
>
> Fatal Trap 12: page fault while in kernel mode
>
> fault virtual address = 0x0
> fault code = supervisor read, page not present
> instruction pointer = 0x8:0xf01152c7
> stack pointer = 0x10:0xf0211f7c
> frame pointer = 0x10:0xf0211f94
> code segment = base 0x0, limit 0xfffff, type 0x1b
>              = DPL0, pres 1, def32 1, gran 1
> processor flags = interrupt enabled, resume, IOPL=0
> current process = Idle
> interrupt mask = cam
> trap number = 12
> panic: page fault
>
> $ nm -aout /kernel | grep f01152
> f0115288 t _ahc_handle_target_cmd

You really need a little more information than this.  Build a debug
kernel and use gdb to show the lines that correspond to the IP where
it fails.  You can use /dev/mem as the dump file.

>   from /etc/ccd.conf:
> ccd1            65536   0       /dev/sd4g /dev/sd5g /dev/sd6g /dev/sd7g
>
>   So if it's the SCSI-driver which causes the panics, trying to
>   dump to a SCSI-disk isn't very likely to succeed :).
>
> - The kernel even refuses to dump to a newly plugged in IDE-disk,
>   a partition of which is configured as swap- and dumpdevice, but
>   after booting the output "savecore: no core dump" can be seen.

What happens before the boot, when it's supposed to be dumping?

> As the machine is about an hour's ride away from me, I'd rather
> be able to debug the problem remotely.
> I asked for a serial cable to a linux-box standing next to the
> FreeBSD-machine, I guess I then could use it as console, get it to
> fall to ddb when panicing and then get a stacktrace - or am I
> missing something that wouldn't work ?

No, that should work.  It's liable to be a painful way of doing
things, though.  You'd be better off with a FreeBSD box with a copy of
the sources and kernel gdb, which you could then use to debug the
system relatively well.

> If i got the kernel do dump to the IDE-disk, that too would of
> course help... any ideas why this is failing ?

I could only guess, and guesswork doesn't help in this kind of
situation.

Greg
--
See complete headers for address, home page and phone numbers
finger grog@lemis.com for PGP public key

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-current" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19981119125726.I467>