Re: /etc/disktab on vax

[Anders Magnusson <ragge%ludd.ltu.se@localhost>]

On 06/23/2012 10:51 PM, David Brownlee wrote:
> On 22 June 2012 20:55, Anders Magnusson <ragge%ludd.ltu.se@localhost> wrote:
> > On 06/22/2012 09:17 PM, Jochen Kunz wrote:
> > > On Fri, 22 Jun 2012 15:11:46 +0200
> > > Anders Magnusson <ragge%ludd.ltu.se@localhost> wrote:
> > >
> > > > In reality, I assume that there are
> > > > very few people in the world (besides me) that still has Massbus disks
> > > > around :-)
> > > Well. No Massbus nor Eagles. But assorted 9" SMD drives from NEC,
> > > Hitachi and Seagate as well as a RL02 and a RK05. :-)
> > Are the SMD disks connected via a Massbus emulating card?
> > In that case you are a typical example of the problem that the disktab info
> > solves :-)
> So would it make sense to keep the entries but just update them to a
> single large partition? That retains the key ability to safely label
> such disks, without the (of only historical relevance) partition
> tables?
Yes, maybe, but there is one other thing that is influencing (which also 
is only of historical interest):
The vax massbus disks uses DEC 144 bad block forwarding, which means 
that a bunch of blocks at the end of the c partition are unallocated and 
used to replace bad blocks in other places on the disk. This is managed 
by the bad144 program.
The disktab entries are avoiding these blocks, but it must be possible 
to access them via the c partition to allocate them.

The current hp driver do not make use of dec144 but I have dream that 
someday the 4.4 encumbered source code will be free so that the massbus 
drivers could be included in NetBSD.  BTW anyone know the state of the 
license right now? :-)

-- Ragge