Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 6 Jan 1997 11:10:33 -0700 (MST)
From:      Terry Lambert <terry@lambert.org>
To:        seanb012@sunbeach.net (Sean Batson)
Cc:        hackers@freebsd.org
Subject:   Re: Shutdown tip
Message-ID:  <199701061810.LAA12205@phaeton.artisoft.com>
In-Reply-To: <32C87FC1.41C67EA6@sunbeach.net> from "Sean Batson" at Dec 30, 96 10:51:45 pm

next in thread | previous in thread | raw e-mail | index | archive | help
> The next time you guys are about to shutdown try this:
> 
> shutdown -h now;exit
> 
> the above command should be done once all of the users
> are logged off. The end results would always be this:
> 
> syncing ...disk
> press anykey to reboot.
> no buffers to rewrite to the filesystem.
> 
> Just do an ordernary shutdown -h now, you'll get the
> usual syncing ..3 4 disk ,...
> 
> How if someone where modify the code for the shutdown
> command to do a root user logout then fork or spawn
> the rest of the code for the shutdown thus avoiding
> the site of the buffer related numbers each time on 
> a shutdown.

The problem is in the use of a non-"/" current directory actig as an
open file reference for the processes that are still live.

The correct thing to do is to orderly shutdown the processes.

You would probably ss the same (correct) behaviour if you forced a
shutdown to occur only in "/" (like SVR4 does) and modified the root/chroot
fd code to always exist in the process instead of conditionally existing,
and then simplifying the namei() code to unconditionally reference the
process value.


					Terry Lambert
					terry@lambert.org
---
Any opinions in this posting are my own and not those of my present
or previous employers.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199701061810.LAA12205>