> And, I guess the disklabel(8) man page will need to be revised:
> "On the sparc... the size of each partition must be a multiple of the
> number of sectors per cylinder..."

Yes, that probably should be updated to tell the truth.

> [...] some "conventions" that are adopted for LBA [...]
> [...] (e.g., [E]IDE).

Does NetBSD/sparc (as distinct from NetBSD/sparc64) support any
machines that even have IDE?

>> Well, the `cylinder boundaries' on which partitions must begin are
>> not actual cylinder boundaries on the disk, but rather multiples of
>> the (sec/trk) * (trk/cyl) product, where those two numbers are the
>> ones in the disklabel.
> The *BSD* disklabel, in this case.

Well, of whichever disklabel is in use.  If you're using stock tools to
set labels, the Sun-compatability label and the BSD label will have the
same numbers in them.  If you use something like my sunlabel or
bsdlabel programs to create disagreements, then it's up to the kernel
which one it believes - but the Sun label is the only one that actually
suffers from the silly "must begin on cylinder boundary" bogosity.
I've not actually tried it (I don't run NetBSD/sparc versions that deal
in BSD disklabels), but I conjecture that you could write a Sun label
that describes your boot partition and then a different BSD label that
describes everything, and expect it to work....

> Yes.   factor(1) is your friend... :>

Actually, disklabel programs that let you specify things in terms of
c/h/s are an even bigger friend. :-)  That's one reason I wrote
sunlabel and bsdlabel.

> Also, I think that the boot loader for shared drives requires the
> first complete *track* to be setaside

I've never seen this on anything but Inhell boxen, and certainly not on
SPARCs.

>> Personally, what I generally do is ignore the geometry reported by
>> the drive and set trk/cyl=32 and sec/trk=64 in the label, which
>> gives me 1MB `cylinders'.
> By contrast, I *try* to use the geometry reported by the drive *IF*
> it will also satisfy the requirements imposed by BIOS

I have only one Inhell box with its idiotic BIOS restrictions, I don't
boot it from SCSI, and I have no other IDE-capable machine (hmm,
actually, I think my macppc machine runs IDE - but I don't move drives
between them in any case).  On IDE drives, yes, I leave the geometry
settings alone so as not to upset the BIOS - but this is port-sparc,
which as I said I think doesn't do IDE.  Certainly I've never used IDE
under NetBSD/sparc.

Yeah, if you move drives between machines/OSes, you have to make sure you
satisfy all the machines'/OSes' requirements, which may preclude things
like what I suggest above....

/~\ The ASCII				der Mouse
\ / Ribbon Campaign
 X  Against HTML	       mouse@rodents.montreal.qc.ca
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B