Subject: Re: Another changer, another changer problem
To: None <email@example.com>
From: der Mouse <mouse@Rodents.Montreal.QC.CA>
Date: 10/04/1998 00:27:47
> I know various PROMware, whose manufacturers (like Sun) shall go
> make various assumptions along the lines that "All disks have 8
> partitions", but this is a topic that has been discussed but not
> fully addressed. I know der Mouse has come up with his own solution
> to this problem (hiya, Mouse!).
...did somebody call?
I use principally the sparc and sun3 ports. Both of these tend to use
Sun-compatible disklabels; I'm not sure whether they *require* it or
not (except for the boot drive, where the PROMs demand it).
I found space within the Sun-style disklabels to tuck 8 more
partitions. I've got patches to the sparc and sun3 MD disklabel code
to understand and generate such labels, and have also taught my
disk-labeling program "sunlabel" about them. I've been running with
16-partition disks on NetBSD/sparc and NetBSD/sun3 for, oh, months at
least, I think at least a year.
There's a PR in the database with what were then my patches. I can
easily supply patches relative to the last source tree I upgraded to,
probably with minimal effort relative to -current. The biggest problem
is the flag day in /dev.
If and when I start using any other ports more than tangentially, I'm
going to arrange for them to support at least 16 partitions per pack
too; it's just too useful. Especially now that we have 32-bit dev_ts
and this thus doesn't restrict me to 8 packs per driver! (Not that
I've felt that limitation yet; the closest is one sun3 machine with
seven 80meg disks...and can't add more 'cause it's a -3/60 and has just
the one (narrow) SCSI controller, with no way to add more (AFAIK).
7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B