From owner-freebsd-bugs  Tue Mar 28 23:04:15 1995
Return-Path: bugs-owner
Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id XAA10695 for bugs-outgoing; Tue, 28 Mar 1995 23:04:15 -0800
Received: from relay.philips.nl (relay.philips.nl [130.144.65.1]) by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id XAA10683; Tue, 28 Mar 1995 23:04:13 -0800
Received: from muxgw1.ms.philips.nl ([130.144.90.6]) by relay.philips.nl (8.6.9/8.6.9-950103) with SMTP id JAA27319; Wed, 29 Mar 1995 09:02:24 +0200
Received: by muxgw1.ms.philips.nl (5.57/Ultrix2.4-C)
	id AA00503; Wed, 29 Mar 95 08:45:50 +0300
Received: by mmra1.ms.philips.nl (4.1/SMI-4.1)
	id AA23134; Wed, 29 Mar 95 09:00:39 +0200
From: gvrooij@mmra1.ms.philips.nl (Guido van Rooij)
Message-Id: <9503290700.AA23134@mmra1.ms.philips.nl>
Subject: Re: kern/280: new slice manager totally confused about old slice disks
To: pst@shockwave.com (Paul Traina)
Date: Wed, 29 Mar 1995 09:00:39 +0200 (MET DST)
Cc: jkh@freefall.cdrom.com, freebsd-bugs@freefall.cdrom.com
In-Reply-To: <199503290525.VAA01190@precipice.shockwave.com> from "Paul Traina" at Mar 28, 95 09:25:09 pm
X-Mailer: ELM [version 2.4 PL21]
Content-Type: text
Content-Length: 1458      
Sender: bugs-owner@FreeBSD.org
Precedence: bulk

> 
> Aha, ok, this worked for me, I now understand a lot more about the slice
> scheme (took only 30 seconds wall clock time and 1 second CPU time).
> 

I still dont get it completely. At first I thought the slices referred
to `fdisk partition' and partitions to `BSD partitions'.
But now I'm not sure anymore, as I saw slice 5 as well. Or do we support
extended slices now?
Further: I *know* 2.0R made releases where disklable would complain about
partition c and d extending past the end of the drive (due to the total
#sec > #secpertrack * #heads * #cyl. I guess the option to use the rest
of the drive for a partition took all sectors left, while not checking
the consistency of all drive params).
How is that handled by this new scheme?

> I would suggest documenting the living hell out of this.
> 

Yep.

> The gratuitous kernel printfs when accessing the label are a bit much,
> do they only happen when booting in verbose mode?
> 
> sd0s1: start 32, end = 511999, size 511968: OK
> sd0s4: start 512000, end = 3514367, size 3002368: OK
> sd0: rejecting partition in BSD label: it isn't entirely within the slice

What exactly does this message mean? And, more important, what are its
consequences on operation?

> This is not going to be fun to support,  but I have to admit that it's
> far nicer than the old organization.  Kudos to all.
> 

It looks nice at first sight, but I haven't got the feeling that I completely
understand it (yet).

-Guido