kern/96644: crash after mount/umount disturbance: fatal trap 12

[Kris Kennaway]

The following reply was made to PR kern/96644; it has been noted by GNATS.

From: Kris Kennaway <kris at obsecurity.org>
To: Arne Woerner <arne_woerner at yahoo.com>
Cc: freebsd-gnats-submit at FreeBSD.org
Subject: Re: kern/96644: crash after mount/umount disturbance: fatal trap 12
Date: Tue, 2 May 2006 13:37:04 -0400

 --5/uDoXvLw7AC5HRs
 Content-Type: text/plain; charset=us-ascii
 Content-Disposition: inline
 Content-Transfer-Encoding: quoted-printable
 
 On Tue, May 02, 2006 at 06:38:19AM +0000, Arne Woerner wrote:
 >=20
 > >Number:         96644
 > >Category:       kern
 > >Synopsis:       crash after mount/umount disturbance: fatal trap 12
 > >Confidential:   no
 > >Severity:       non-critical
 > >Priority:       low
 > >Responsible:    freebsd-bugs
 > >State:          open
 > >Quarter:       =20
 > >Keywords:      =20
 > >Date-Required:
 > >Class:          sw-bug
 > >Submitter-Id:   current-users
 > >Arrival-Date:   Tue May 02 06:40:13 GMT 2006
 > >Closed-Date:
 > >Last-Modified:
 > >Originator:     Arne Woerner
 > >Release:        R6.0
 > >Organization:
 > >Environment:
 > FreeBSD neo.riddick.homeunix.org 6.0-RELEASE FreeBSD 6.0-RELEASE #10:
 > Mon Apr 10 15:10:38 UTC 2006
 > aw at neo.riddick.homeunix.org:/usr/src/sys/i386/compile/NEO  i386
 >=20
 > >Description:
 > Once upon a time I looked through my self made DVDs. When I mounted a DVD
 > twice at the same mount point and read from that DVD and unmounted the DVD
 > once and read again from it, the system crashed.
 >=20
 > This is not really bad, since I usually know, what is mounted...
 > But it might be nicer, if the kernel could cope with such things...
 > >How-To-Repeat:
 > cd /opt/tmp
 > dd if=3D/dev/zero of=3Da bs=3D1m count=3D200
 > mdconfig -a -t vnode -f a
 > newfs /dev/md0
 > mount /dev/md0 /mnt
 > dd if=3D/dev/zero of=3D/mnt/a bs=3D1m count=3D100
 > umount /mnt
 > mount -r /dev/md0 /mnt
 > mount -r /dev/md0 /mnt
 > dd if=3D/mnt/a of=3D/dev/null bs=3D1m
 > umount /mnt
 > dd if=3D/mnt/a of=3D/dev/null bs=3D1m
 > *cloink*
 >=20
 > kernel says:
 > fatal trap 12: page fault while in kernel mode
 > fault virtual address =3D 0x0
 > fault code =3D supervisor read, page not present
 
 FYI, usually more details are required when submitting reports of
 panics, e.g. a full backtrace from kgdb (see the developers handbook
 chapter on kernel debugging).
 
 In this case, it's a known bug that mounting the same filesystem twice
 as you did can lead to panics.
 
 Kris
 
 --5/uDoXvLw7AC5HRs
 Content-Type: application/pgp-signature
 Content-Disposition: inline
 
 -----BEGIN PGP SIGNATURE-----
 Version: GnuPG v1.4.3 (FreeBSD)
 
 iD8DBQFEV5jAWry0BWjoQKURAiArAKCOzs5ZdPbAQhI56gGDFI+Mp9mlhgCg18g4
 wGW5iaOaVjKYMHUOQx4ZJWY=
 =clJB
 -----END PGP SIGNATURE-----
 
 --5/uDoXvLw7AC5HRs--