Date: Sun, 04 Jan 2004 16:04:34 +0100 From: Stefan Ehmann <shoesoft@gmx.net> To: Bruce Evans <bde@zeta.org.au> Cc: current@FreeBSD.org Subject: Re: page fault panic tracked down (selwakeuppri()) Message-ID: <1073228674.1764.28.camel@shoeserv.freebsd> In-Reply-To: <20040105003929.Y631@gamplex.bde.org> References: <200401030401.i0341O7E003439@gw.catspoiler.org> <20040105003929.Y631@gamplex.bde.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, 2004-01-04 at 14:48, Bruce Evans wrote: > On Fri, 2 Jan 2004, Don Lewis wrote: > My not seeing any problems is consistent with my not using any sound > drivers. > > Did you rule out an invalid file system? I just tested -current and > was reminded that it doesn't reboot cleanly if ext2fs is mounted. > The unclean reboot doesn't affect r/o mounts of ext2fs. OTOH, r/o > mounts of dirty file systems are permitted and there is no easy way > to fsck ext2fs file systems under FreeBSD. (I use fsck.ext2fs from > the e2fsprogs port, but it still needs a block device or filesystem > image to work correctly so it hasn't been easy to use since block > devices were axed. I sometimes copy file systems to regular files > to use it.) I already thought of an invalid file system. But I think it would be very unlikely that it only causes panics with the selwakeuppri() changes. I normally mount it read only and unmount it manually before shutdown to avoid these problems. Last time when I mounted it rw and the system crashed I had big troubles repairing it again. In fact I gave up repairing it under FreeBSD (always complained about wrong superblock - ro mounting was working fine though). Finally repaired it using Knoppix which also had troubles. I only found 1 of about 10 superblock backups working.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1073228674.1764.28.camel>