[965] in linux-scsi channel archive
Re: MO (Fujitsu 2513A). Was 2048 byte sectorsize...
daemon@ATHENA.MIT.EDU (Eric Youngdale)
Mon Nov 18 00:25:24 1996
Date: Sun, 17 Nov 1996 23:55:15 -0500
From: Eric Youngdale <eric@sub2317.jic.com>
To: steve@dragon.central.de
CC: reuther@os.inf.tu-dresden.de
CC: ak@centipede (none)
CC: andi@fred.mlm.extern.lrz-muenchen.de
CC: root@wopper.wupper.de
CC: rudolf@dose.muc.de, ki@boss.kretz.co.at
CC: thomas@cevis.uni-bremen.de, squid@hub-wue.franken.de
In-reply-to: <199611171118.MAA00641@elrond.central.de> (message from Steffen
Weinreich on Sun, 17 Nov 1996 12:18:05 +0100)
CC: linux-scsi@vger.rutgers.edu
OK, I had a chance to play with this for real myself. I found the
reason you were still unable to read the partition tables, and I have fixed it.
At this point you should *probably* be able to put a real partition table down,
and put real filesystems on it.
All of my testing was done by forcing the blocksize for a regular
hard disk to 2048 bytes. Unfortunately the existing partition table is such
that the first partition starts on sector 1, so I wouldn't be able to use
that one. Also, the second partition also starts on an odd sector, so I cannot
test that one either. Not without completely nuking the contents of the
disk, anyways.
The updated diffs are on ftp://sub2317.jic.com/pub/scsi/2048-1.diff.
-Eric
