[145] in linux-scsi channel archive
Re: NCR 53c810 lockups
daemon@ATHENA.MIT.EDU (Richard Gooch)
Fri Apr 14 23:01:06 1995
From: Richard Gooch <rgooch@rp.CSIRO.AU>
Date: Sat, 15 Apr 1995 12:23:57 +1000
To: Bogdan Urma <bau1@crux2.cit.cornell.edu>
Cc: linux-scsi@vger.rutgers.edu
In-Reply-To: <199504141842.OAA00240@mustafa.com>
Bogdan Urma writes:
>
> Ever since I got a Seagate 31230N SCSI II drive and plugged it into
> the NCR 53c810 controler on my 486/PCI motherboard, I've been experiencing
> random lockups with Linux. The SCSI disk seems to work fine, but at
> random intervals(hours or days) the system just locks up. Since I never
> had these problems before when using an IDE drive, I'm attributing these
> lockups to the NCR controller and the SCSI drive. One thing I came up with is
> that the NCR is using IRQ 9, the ONLY one I can use with it. Is using
> IRQ 9 a problem with the NCR? Is anybody else doing it successfully?
> What else can be the problem? I'm including the boot up messages. BTW, when
> the system locks up, there are no messages reported by syslog, even
> though I enabled verbose SCSI in the kernel.
It would have been more helpful to post the messages you see when
the system locks up (you won't be able to run X of course).
If you are having the same problem as I had, then going into the BIOS
and enabling "LGNT# synchronous with LCLK" should fix the problem.
Don't ask me what it means: it just works for me.
A good test I used for "encouraging" the problem is:
1) cold boot
2) log in as root and:
# dd if=/dev/hda of=/dev/null bs=1024k count=<size of disc in MBytes)
When I was having my problem, this would cause a SCSI failure every
time. Once I fixed the problem, I never got a failure.
Regards,
Richard....
