home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!mcsun!sun4nl!phigate!relay.philips.nl!philica!adrie
- From: adrie@ica.philips.nl (Adrie Koolen)
- Newsgroups: comp.unix.sysv386
- Subject: Re: sector remapping woes with 1542A/SCO Unix
- Message-ID: <1992Jul21.100207.429@ica.philips.nl>
- Date: 21 Jul 92 10:02:07 GMT
- References: <467@wurtel.hobby.nl>
- Organization: Philips Consumer Electronics, Eindhoven, The Netherlands
- Lines: 46
-
- In article <467@wurtel.hobby.nl> paul@wurtel.hobby.nl (Paul Slootman) writes:
- >I have a strange problem. I recently added a SCSI controller and
- >disk to my 386 system (Adaptec 1540A and Fujitsu M2249SA (I think)).
- >I already have an Adaptec RLL controller with two drives in my machine.
- >
- >The problem is this: if I disable the two "standard" disks in the CMOS
- >setup of the machine, the Adaptec SCSI controller properly sees the
- >SCSI disk and installs it as C:. If I put DOS on it, there's no problem,
- >I can boot from it, run apps, etc.: everything works ok. The SCSI BIOS
- >remaps the sectors to 318 cylinders of 64 heads, 32 sectors, giving me
- >318 MB, which is right.
- >
- >However, when I access it from UNIX by installing SCO Unix 3.2.2, SCO
- >tells me it's a disk with 1272 cylinders, 16 heads, 32 sectors, also
- >giving me 318MB. However, after installing everything, I can't boot from
- >it ("NO OS").
-
- I'm no expert on this, but I think that you've got a `old' 1540A with an
- updated BIOS EPROM. Old 1540A's have a BIOS which maps the disk on 16
- heads. With 1024 cylinders and 32 sectors per track, disks max-out at
- 256MB. Newer EPROMS map at 64 heads per cylinder. You can ask the
- AHA154x adapters for their type. Your 1540A obviously reports that it's
- a 1540A, which the SCO thinks that it has a 16 head BIOS. That's why
- it maps the disk that way. It talks directly with the 1540A so it has
- no 1024 cylinder limit of the PC BIOS.
-
- >Some exploratory work shows me that the PC BIOS wants to load the stuff
- >from the hard disk, but does this using the first translation (64 heads).
- >As the partition table is written while 16 heads are assumed, the stuff
- >can't be found, and the bootstrap fails.
- >
- >Why does SCO Unix think that a different sector translation is called
- >for?
-
- You've got a problem here. You could switch to the old BIOS, but I don't
- know where to get one (although a brother of mine has an 16 head 1540A,
- you probably don't want a copy of its PROM). You'd lose 62MB this way!
- You also could have the microcontroller PROM replaced so that it reports
- as a 1540A with 64 heads. I'm not sure if this PROM update exists and
- where to get it. The last possible solution is to update the SCO SCSI
- driver that it maps the disk with 64 heads per track.
-
- Note, I'm only guessing, but I think that I'm in the right direction.
-
- Adrie Koolen (adrie@ica.philips.nl)
- Philips Consumer Electronics, Eindhoven, the Netherlands
-