Re: boot floppies hang. kernel image problems
I'll help isolate the problem. As I said, I very thoroughly tested the problem,
checked on multiple machines, etc. After testing on 8 different systems with
the same CD, and finding it consistantly has lockup problems on each, and where
the problem was, it becomes clear that the WD-7000 auto-probe, while it works on
the majority of the systems I have tried, WILL lock the system, but ONLY when
the Adaptec 2940UW SCSI controller was present in the same system. This does
NOT happen in other systems with EXACTLY the same cards installed(just different
motherboards). It's a BIOS related issue from the look of it. It COULD be a
bug in the kernel, where it doesn't properly allocate memory after detecting(but
not yet fully initializing) a 2940. But that's not something I can be sure of.
Dave Bristel
targon@targonia.com
On 3 May 1999, Adam Di Carlo wrote:
> Date: 03 May 1999 18:35:15 -0400
> From: Adam Di Carlo <adam@onshore.com>
> To: David Bristel <targon@targonia.com>
> Cc: 36749@bugs.debian.org, debian-testing@lists.debian.org
> Subject: boot floppies hang. kernel image problems
>
>
> > The problem I am experiencing is with the boot floppy images on slink.
> > Problem is as follows:
>
> > On a number of systems, the bootup from the floppy images on the Slink
> > 2.1 CD(original CD, r0) will hang because of the auto-probe by the
> > WD-7000. The system contains a single Adaptec 2940UW SCSI controller,
> > which is detected. BUT, prior to SCSI device detection, the WD-7000
> > code tries to auto-probe the hardware, and cold-locks the system. I
> > have heard of similar problems from friends. Now, I need the Adaptec
> > support because this is a SCSI only system, but I can't use Debian at
> > ALL on certain motherboards because of the WD-7000 probe. Now, I see
> > three possible ways of solving this problem. 1) remove the WD7000
> > from the base kernel, and make it a module. 2) change the code so
> > modules can be loaded or removed without needing to set up a root
> > partition. 3) try going to a 2.2 kernel on the base floppies, and
> > hope that the auto-probe doesn't lock up the systems.
>
> > If the kernel had a way to disable the WD-7000 probe through a kernel
> > option, that would also solve the problem I am encountering. I have
> > removed the Adaptec from the system, and I don't experience the
> > problem, so it APPEARS that the WD7000 code is hitting the 2940 and
> > locking the system. I don't know why, and it doesn't happen on MOST
> > of the systems I run. Perhaps it's a BIOS issue, perhaps not, but who
> > uses a WD7000 SCSI controller?
>
> David, this is a slight variation on known bug with the kernels which
> ship with slink.
>
> I've been trying to motivate people to fix these problems but we don't
> seem to have many testers available. If you are willing to test this
> problem, please reply and say so (make sure to CC
> debian-testing@lists.debian.org).
>
> --
> .....Adam Di Carlo....adam@onShore.com.....<URL:http://www.onShore.com/>
>
--
To UNSUBSCRIBE, email to debian-testing-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Reply to: