Re: port-alpha/48697: Sysinst compute illegal disk geometry and newfs is bound to fail.

[David Laight <david%l8s.co.uk@localhost>]

On Wed, Apr 02, 2014 at 08:25:01PM +0000, Michael L. Hitch wrote:
>  
>     Ah, it looks like maybe sysinst is computing the size of the disk by 
>  multiplying the sectors/cylinder times cylinders - but the sectors/track 
>  in the disk label is not correct.  The drive reports 727 sectors/track. 
>  If sysinst had used the total sectors from the disklabel, it would have 
>  been ok.  I would presume that Tru64 had set the sectors/track and 
>  sectors/cylinder to those values.

It is really best to believe the size that the driver gets from
the disk itself - rather than that written to the disk.

The 727 sectors/track value read from the label would then not matter.
Indeed it might make the code know that the label is invalid.

A cursory glance and the output from netbsd's disklabel seemed to
show garbage values for the partitions - even though the checksum
must have been correct.
Maybe the structures don't quite match?

        David

-- 
David Laight: david%l8s.co.uk@localhost