Subject: Re: tough filesystem stains
To: <>
From: Simon Burge <simonb@telstra.com.au>
List: netbsd-help
Date: 08/18/1998 08:48:24
On Mon, 17 Aug 1998 08:57:42 -0700 (MST)  Tim Rightnour wrote:

> On 13-Aug-98 Andrew Brown spoke unto us all:
> # >I don't  think this is a new problem; as I understand it, this is why
> # >the -n option exists for reboot.
> #  
> #  actually...my sunos 4.1.3 man page for reboot says:
> #  
> #       -n   Avoid the sync(1).  It can be used if  a  disk  or  the
> #            processor is on fire.
> #  
> #  although...if "a disk or the processor" were on fire, i doubt you'd
> #  actually calmly (madly?) type "reboot".
> 
> No doubt.. I'd be headed for the wall socket with a quickness.. ;)  Perhaps
> they are just trying to say "only in extreme circumstances"  But then.. who
> would type "man reboot" at a time like that?

Part of the reason for "-n" was for when fsck'ing the root filesystem
when they were mounted read/write on boot.  Using my favorite old legacy
Unix O/S (Ultrix) as an example, if you fsck the boot disk and it
changes something, you get the following messages:

	***** FILE SYSTEM WAS MODIFIED *****
	***** HALT PROCESSOR WITHOUT SYNCING DISK *****

Does anyone know if NetBSD ever mounted it's boot disk read/write right
from the word go, instead of needing the "mount -u" bit?

Simon.