Subject: Re: disklabel working?
To: Christopher R. Bowman <crb@glue.umd.edu>
From: Hauke Fath <hauke@Espresso.Rhein-Neckar.DE>
List: port-mac68k
Date: 08/02/1997 18:23:33
A while back, in the context of floppy driver development, I hacked the
MacBSD disklabel code some and was able to label a 230MB MO disk just fine.
Don't know if I still have the changed files around here.

Actually, the mac68k disklabel code is pretty old and even has some i386
compatibility stuff in it; not sure if that treats byte ordering correctly.

NetBSD disklabels become interesting when you think about moving disks to
other machines. Even if we had native disklabels in MacBSD we'd still be
limited to big endian machines as long as the ffs stores its data in host
byte order. Until then tar(1) is your friend...  ;)

As it is, given you have setup your partitions correctly under MacOS, you
can just go and newfs(8) them and the kernel will create a disklabel for
you.

You might want to look for "disklabel" in the port-mac68k mailing-list
archive; I recall there was a longish discussion about a year ago.

	hauke


At 3:36 Uhr +0200 02.08.1997, Christopher R. Bowman wrote:

>>> I'm adding an old, small SCSI HD to my SE/30, so that I can get some
>>> experience installing disks. I want to do the complete job from within
>>> NetBSD, but I'm having some problems.
>>
>>We (mac68k) don't support writing disklabels right now. It's one of the
>>things I'd like to work on, but I'm trying to get LocalTalk first. Likewise
>>others are trying to get color X, LC040's, and working SCSI drivers (no more
>>SBC vs ncrscsi) going. Though not ideal, the current setup works ok.
>>
>>The thing I really want to do is make it so that you can write any sort of
>>disklabel on a disk. And read it. So mac68k could read a MacOS disk,
>>a DOS-partitioned disk (I forgot the proper name), an AmigaOS disk, a
>>SunOS disk, a NetBSD/sun disk, etc. To do it right, we need to be able
>>to specify what kind of disklabel we want to deal with, and have all
>>the label readers around.
>
>I find this an interesting problem, is there anyone in the know who would
>like to explain why we don't allow disklabeling, and what would be required
>to implement it?  Any one want to discuss this.


--
"It's never straight up and down"     (DEVO)