I hesitated to respond with what is probably a red herring (because I
can't explain why 2.0.36 works and 2.0.35 doesn't) but I've seen similar
behaviour related to initializing keyboards and the BIOS.
I have a Jumptec DIMM-PC which requires that you go into
drivers/char/keyboard.c and change '#ifndef __i386__' to '#ifndef NOTDEF'
(around line 46) so that the keyboard gets initialized. Without the
initialization, the kernel simply reboots. I've had to do this on 2.0.36
kernels.
Apparently, an i386 BIOS is expected (by the kernel) to initialize the
keyboard, but other architectures are not. But the BIOS I have doesn't do
it. (I _think_ its an AMI BIOS, but I can't remember for sure.)
The behaviour sounds the same as what you are seeing, so I thought
I'd throw it out there.
I'd submit a patch to the kernel maintainers making keyboard init a config
option, but I'm not running 2.2 anywhere, and I doubt they're too
interested in silly patches to 2.0 :)
--Jeremy
On Mon, 12 Apr 1999, Andy Waddell wrote:
> I've seen similar posts about this issue back in January, so you might look
> there. I've got the same board and am having luck but only with one of the
> two BIOS's shipped with the board for evaluation. I have some rev 2 and
> some rev 3 boards, and the rev 2 ship with BIOS's from two different vendors
> (SystemSoft and Phoenix?). If I boot off the SystemSoft one, it reboots in
> approximately the same place, but if I use the other BIOS, it works fine.
> If you're on a rev 3 board, you'll have to re-program the ROM with the other
> BIOS if you want to give it a try. Good luck.
>
> --andy
Jeremy Impson
Network Engineer
Advanced Technologies Department
Lockheed Martin Federal Systems
[EMAIL PROTECTED]
