> I had unhappiness when I made partitions which were not whole
> cylinders; my sparc build system refused to boot.  However, I'd bet a
> lot that it's a self-consistency issue, i.e. a partition's notion of
> number of blocks in a cylinder needs to match the disklabel, not
> necessarily what the disk reports as its geometry in the SCSI
> IDENTIFY command (or whatever).

I think I don't quite understand.  In what sense does a partition
*have* a notion of blks/cyl?

In a Sun-style disklabel, there are only 64 bits of per-partition
information: the starting cylinder and the size in blocks (each stored
as a 32-bit value).

*Filesystems* have geometry info.  But the PROMs don't know anything
about the filesystem; they just know how to load the bootblock from the
beginning of the partition.  Right?

I'd be very interested to see the first 8K of that non-bootable disk
you managed to create (but I suppose by now it's long since been
overwritten)....

					der Mouse

			       mouse@rodents.montreal.qc.ca
		     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B