Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 25 Feb 2002 22:30:11 +0000
From:      Ian Dowse <iedowse@maths.tcd.ie>
To:        Kris Kennaway <kris@obsecurity.org>
Cc:        mckusick@mckusick.com, fs@FreeBSD.org, dillon@FreeBSD.org, fanf@chiark.greenend.org.uk
Subject:   Re: UFS panic on -stable 
Message-ID:   <200202252230.aa23166@salmon.maths.tcd.ie>
In-Reply-To: Your message of "Mon, 25 Feb 2002 13:17:14 PST." <20020225131714.B59373@xor.obsecurity.org> 

next in thread | previous in thread | raw e-mail | index | archive | help
In message <20020225131714.B59373@xor.obsecurity.org>, Kris Kennaway writes:
>Here you go, hope I got everything:

Mostly :-) Could you print out "*(struct proc *)0xcc2fba00", as the
`p' pointer in frame 18 got lost. I just want to see what process
was exiting to see if it offers any further hints.

The superblock claims that this is a 2MB "/dev" filesystem - is
that a real disk filesystem, or mounted from a vnconfig disk or
something?

In the vnode we have v_type set to VNON, v_mount is non-NULL, v_data
is non-NULL, i_mode was 0666. Candidates I could see for leaving
vnodes like this are getnewvnode() and the weird code in ufs_mknod().
I guess ufs_mknod is the main suspect since the filesystem is /dev.

I wonder is the new vlrureclaim() code somehow reclaiming vnodes
at a bad time. Maybe try setting kern.maxvnodes to some really large
value to see if it stops the crashes, or set it very small and see
if they occur more frequently.

Matt, can you see any way that the weird/broken ufs_mknod() code
could interact badly with vlrureclaim()?

Ian

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




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi? <200202252230.aa23166>