[1220] in linux-scsi channel archive
Re: Adaptec AHA2940
daemon@ATHENA.MIT.EDU (Edward Welbon)
Sat Jan 4 00:15:13 1997
Date: Fri, 3 Jan 1997 23:07:11 -0600 (CST)
From: Edward Welbon <welbon@bga.com>
Reply-To: Edward Welbon <welbon@bga.com>
To: "Leonard N. Zubkoff" <lnz@dandelion.com>
cc: lm@neteng.engr.sgi.com, linux-scsi@vger.rutgers.edu,
submit-linux-dev-kernel@ratatosk.yggdrasil.com
In-Reply-To: <199701031905.LAA07023@dandelion.com>
On Fri, 3 Jan 1997, Leonard N. Zubkoff wrote:
...
> the NCR/Symbios BIOS
> follows the CAM specification which tried to optimize the usage of the
> disk space by computing a C/H/S translation that's optimal for each disk.
> This causes disks partitioned on a Symbios host adapter to have problems
> on host adapters that used the fixed schemes, such as Adaptec and
> BusLogic. However, the Symbios BIOS will adopt a translation that's
> already present in the partition table, so you can generally move a disk
> from Adaptec or BusLogic to Symbios.
>
> Adaptec uses either 64/32 or 255/63 translation, based on the size of the
> disk being > 1GB and the extended translation option being selected.
This seems reasonable, but...
> Here's a table of what host adapter changes should and should not work, based
> on the host adapter that originally partitioned the disk and whether it had
> extended translation enabled:
>
> Old Host Adapter New Host Adapter Compatible
> ======================== ======================== ==========
...
> Symbios/NCR Adaptec no
Hmm... I must be missing some part of the picture. I have a seagate
baracuda 15150W (4G) disk that I partitioned using an NCR 825 card. At
one time I was having problems with the NCR-Seagate combination (later
proven to be temperature related); suspecting the NCR card (wrongly), I
tried an adaptec 2940 (wide but not ultra). Apparently, I let the disk
cool just enough that I was able to boot and use the disk (though later
when the disk was allowed to heat up it failed again). I do seem to
remember that fdisk did not like the partition boundaries which seems to
agree with your basic points, but I was able to boot OK. I am fairly sure
that fdisk complained only when the disk was too hot.
I have never formatted this disk, the format is the original from seagate.
In a system using an NCR 53c825 with a Symbios 4.0 bios (rather than the
NCR 3.07 - I have two of these cards each with a different bios - the bios
initialization messages are markedly different), fdisk claims this seagate
is mapped as 133 heads, 62 sectors, 1017 cylinders of size 8246*512 bytes.
On problem is that I am not sure what Symbios/NCR bios level I did the
partition with. I am pretty certain that it was from the older level bios
(marked as S1365 V3.07 57F700). Anyway the disk works with every scsi
controller I have tried (including aha1542) provided it is not too warm.
Ed Welbon; welbon@bga.com;
If you pick up a starving dog and make him prosperous, he will not bite
you. This is the principal difference between a dog and a man.
-- Mark Twain
