[parted-devel] [PATCH 1/2] fdasd: geometry handling updated from upstream s390-tools

[Viktor Mihajlovski]

On 25.09.2015 13:51, Phil Susi wrote:
>
> How strange... why on earth does IBM make such odd disks?
>
History ... if you're truly interested:
https://en.wikipedia.org/wiki/Count_key_data

> So there is some out of band DASD hardware specific way of locating
> where the VTOC is, and to handle it being in a different location, you
> want to keep the ioctl, but in the virtio case, you just assume it's in
> the normal location since you can't get the ioctl passed through, and
> anyone who doesn't have it in the usual location is out of luck?
>
As I said a Linux-formatted disk will always have this layout.
So, yes one could apply the same heuristic for native DASD.
This here is a backport of the known-to-be working and tested 
partitioning code from s390-tools. For the sake of future 
maintainability it seems to be prudent to keep the code close
to the origin.

-- 

Mit freundlichen Grüßen/Kind Regards
    Viktor Mihajlovski

IBM Deutschland Research & Development GmbH
Vorsitzender des Aufsichtsrats: Martina Köderitz
Geschäftsführung: Dirk Wittkopp
Sitz der Gesellschaft: Böblingen
Registergericht: Amtsgericht Stuttgart, HRB 243294