From: Alan Charlton (agc@weasel.demon.co.uk)
Date: 01/17/93 

From: agc@weasel.demon.co.uk (Alan Charlton)
Subject: Re: Major seagate bugfix.
Date: Sun, 17 Jan 1993 22:21:11 +0000

In article <C0z65A.Fyo@ra.nrl.navy.mil> eric@tantalus.nrl.navy.mil (Eric Youngdale) writes:

           I think that I have tracked down the bug that was responsible for
   spontaneous reboots when using certain types of disks (i.e. Syquest) with the
   Seagate ST0x SCSI controller.

I think you have....! I found 0.99.3 completely unusable because as
soon as I tried to access my SCSI disk, the entire machine would hang
with:

        Kernel panic: SCint == NULL in Seagate
        In swapper task - not syncing

I tracked the memory location at which it failed to ext.o - perhaps the
relative slowness of the extended fs causes the SCSI comand to time-out?

           I am enclosing patches which would seem to fix this (I have had one
   tentative report of success). I would like other people with the Seagate ST0x
   to try these out and report as to if this helps or not. These patches should
   be applied to a 0.99.3 kernel.

It's odd that this behaviour has only manifested itself in the 0.99.3
kernel; the 0.99.2/.1 seemed much more reliable - is this because of
tidy-ups in the original code?

Thanks very much for the patches - life seems much easier now,
although it seems that other people are having more serious problems
with this version of the kernel.....

Cheers,

Alan