Date: Sat, 28 Feb 1998 10:25:31 +0000 (GMT) From: Terry Lambert <tlambert@primenet.com> To: root@mantar.slip.netcom.com (Manfred Antar) Cc: mike@smith.net.au, tlambert@primenet.com, current@FreeBSD.ORG Subject: Re: VM: Process hangs sleeping on vmpfw Message-ID: <199802281025.DAA10560@usr08.primenet.com> In-Reply-To: <Pine.BSF.3.96.980228001243.5503A-100000@mantar.slip.netcom.com> from "Manfred Antar" at Feb 28, 98 00:21:23 am
next in thread | previous in thread | raw e-mail | index | archive | help
> > > > Check your dmesg/messages log for the following message:
> > > >
> > > > vnode_pager: *** WARNING *** stale FS code in system.
> > ...
> > > I got that message on console last night and rebooted
> > > did fsck and have not seen it lately. do i need to do anything ?
> >
> > It would help to know what filesystems you're running, and whether you
> > are using any LKMs. Do you rebuild your LKMs when you update your
> > kernel?
> >
> > The message indicates a potentially dangerous incompatability between
> > one or more filesystems and other parts of the kernel. If this is due
> > to oversight on our part, we would very much like to know so that it
> > can be remedied ASAP.
>
> I've got NFS FFS MFS MSDOS CD9660 PROCFS compiled in kernel everything is
> mounted local on 1 4.3gig scsi drive:
> /dev/sd0a on /
> /dev/sd0s1e /var
> /dev/sd0s1f /usr
>
> I don't use any LKM's and did a make world today at noon with current
> as of then. I only saw the message once last night repeated 3 time on
> xterm console.I have not noticed any problems
This shows that you did not have anything other than FFS mounted. The
warning can not occur with FFS.
What this means is that the vnode pager was called to read in or write
out pages on an FS that doesn't have a VOP_{GET|PUT}PAGES, and didn't
have the legacy code referenced via stub functions.
In simple terms, this means:
(1) You had something other than an FFS filesystem mounted, and
you aren't telling us about it. If you tell us about it, it's
a simple fix (less than a minute).
(2) Something is erroneously attempting a paging operation using
an erroneous FS type. This is more serious, in that it means
we are now catching some very bad behaviour that we didn't
know about before. This is more complicated, because according
to what we know about FreeBSD, you aren't supposed to need to
do paging on an FS type that's not being used. It means that
someone, somewhare, is probably stomping on a vnode's operations
vector. We would need to change the warning to a panic, wait
for you to repeat the problem, and then post-mortem the system
dump to get things like a stack trace.
I'm betting it's #1, though...
Terry Lambert
terry@lambert.org
---
Any opinions in this posting are my own and not those of my present
or previous employers.
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?199802281025.DAA10560>