Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 30 Oct 1996 09:56:13 +0200
From:      Nadav Eiron <nadav@barcode.co.il>
To:        dwhite@resnet.uoregon.edu
Cc:        adrian@virginia.edu, FreeBSD Questions List <questions@freebsd.org>
Subject:   Re: instability with 2.1.5 kernel
Message-ID:  <32770A1D.592E@barcode.co.il>
References:  <Pine.BSI.3.94.961029222330.369W-100000@gdi.uoregon.edu>

next in thread | previous in thread | raw e-mail | index | archive | help
Doug White wrote:
> 
> On Mon, 28 Oct 1996, Nadav Eiron wrote:
> 
> > >         When shutting down my 2.1.5-RELEASE system, the kernel is failing
> > > with alarming regularity to flush all of the buffers.  I frequently see
> > > about a dozen 4's and the a "giving up" message before the system
> > > reboots.  When the systme comes up the root filesystem and sometimes
> > > others are not marked clean and must be fsck'ed.
> > >
> > >         Has anyone else seen this?
> >
> > Yes! I never bothered to post to questions (or stable) about it as I seldom
> > shutdown my machines, but almost whenever I do, I can see this symptom. There
> > are other weird things with shutdown on 2.1.5: For example, shutting down from
> > an xterm (which used to work just fine) would hang up the machine instead of
> > switching back to the console vty0 before halting.
> 
> I haven't seen any failures to flush on any of my machines, but I have
> seen "lockups" (esp. with 3.1.2S) when shutting down from X.  The system
> really hasn't locked up, the console has gotten locked out but it really
> is shutting down.

I wouldn't know what it does. But waiting a few minutes and then cycling the power
would find the root partition with its clean flag off. This might mean that the
shutdown came through, but the buffers didn't flush, or that the shutdown hanged. I
have no way to know which is right, nor any idea on how to fix it.

> 
> > >         I have also seen a 'shutdown -h +0' go directly to rebooting.
> 
> That doesn't directly reboot; it halts the system giving no guard time,
> but doesn't actually force a reboot -- it says "press any key to reboot".
> It's equivalent to 'shutdown -h now'
> 
> You may have meant 'shutdown -r now' which _does_ restart.

I suspect he means that it reboots eventhough it should just halt.

> 
> > >         Bad hardware or a bug?
> >
> > I guess it's a bug of some sortrs, as I see the same behaviour on two machines
> > (though I only shut them down for kernel rebuilds and the such, which is not
> > very often).
> 
> It depends.  It may be that your SCSI controller is trying to quit to
> quickly or the system is telling the SCSI controller to flush out too
> late/early.

No SCSI on either machines, and they both used to shutdown just fine under 2.1.0R.

> 
> Doug White                              | University of Oregon
> Internet:  dwhite@resnet.uoregon.edu    | Residence Networking Assistant
> http://gladstone.uoregon.edu/~dwhite    | Computer Science Major

Nadav



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?32770A1D.592E>