Re: HD probs

[Bill Leach <bleach@bellsouth.net>]

I have "played" with several instances of various combinations of drives on
Amiga A3000's under both AmigaOS and Linux.

There were some problems (related to the SCSI-II standard) with AmigaOS scsi
(in the ROM code I believe).  The specific problem was that during AutoConfig
the code failed to properly allocate buffers.  Data from drives providing 
longer SCSI-II inquiry response would be accepted but then if a SCSI-I drive
was queried the trailing (stale) data from the previous response was included.

Additionally, there probably are more drives that DO NOT correctly follow the
SCSI specification than there are those that DO!  In particular there are (or
at least were) a great many drives that claimed to support "disconnect/
reconnect" but do not.

A few years ago the problems with drives were particularly noticable on Amigas
because the Amiga SCSI controller was far and away the fastest SCSI controller
on "PC" type machines.  Indeed, for awhile MaxTor required that all of their 
drive controllers/firmware test satisfactorily on an Amiga 3000 as a routine
QA test.

I don't have any of the Amiga stuff here right now but, I would start with
setting reselection off under AmigaOS (and rebooting of course) since
possibly your problem is "setup" during AutoConfig and only manifests itself
when Linux begins "hammering" on your drive.  Admittedly I don't know anything
about how the Linux SCSI code is set up so this may not make any difference.

I think that there is a way to disable synchronous transfer by passing
parameters to the kernel during boot (that then are passed to the SCSI code).


On Wed, Feb 10, 1999 at 04:51:57PM +0000, sto0 wrote:
> *··· hi,*
> 
> I have been having a minor problem with my Quantum HD where Linux is installed. I`m not sure whether
> the HD is dodgy or not (i hope not), but BASH keeps on coming up with:
> 
> hdb: read_intr: status=0x50 { DriveReady SeekComplete }
> hdb: unexpected_intr: status=0x58 { DriveReady SeekComplete DataRequest }
> 
> and it doesn`t look particularly healthy for a drive to give out these messages. I also have an
> AmigaOS HD and have checked the Quantum drive for surface errors in two different programs, but
> found absolutely nothing wrong with it at all! 
> 
> Help! I`m stuck.
> 
> later,
> *stoO*
> -- 
> info:
>  ··· http://www.tvrr.demon.co.uk ··· stu-r@bigfoot.com
> --
> Hackers have kernel knowledge.
> 
> 
> 
> --  
> To UNSUBSCRIBE, email to debian-68k-request@lists.debian.org
> with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
> 
> 

-- 
best,
-bill
                bleach@BellSouth.net
           b.leach@usa.net  LinuxPC@Hotmail.com
from a 1996 Micro$loth ad campaign:
"The less you know about computers the more you want Micro$oft!"
         See!  They do get some things right!