On Wed, 13 Mar 2002, Andrew Doran wrote:

> Ok. I've fixed some of the reported problems. To my knowledge this leaves:
>
> - It's necessary to intercept and bail out Ctrl-Alt-<X
> - "chatty key stroke" symptom
> - SCSI hang experienced by Chris (this is most likely not wscons related)

	Indeed it appears not to be. I lack the HD50 SCSI terminator for
the back of the case, so I hooked up a CDROM with terminator, loaded 1.5Y
snapshot and it booted the 1.5ZB kernel Tohru posted on ftp.netbsd.org to
single user with the sfb. Man it seems a lot faster than before :) Anyway
just so you know it wasn't a kernel problem.

	Also, can we take a bigger whack at the Acess.Bus while we're here
and the code surrounding it is fresh in your minds? I have one of the
systems that 9 times out of 10, the mouse
attaches to 0x6a, or 0x6c I forget which, but the addresses are reversed
from the static entries in dtop.c so my mouse becomes the keyboard and the
keyboard becomes the mouse. Is there any way to probe the PROM for what
device is at which address? I could care less about the hot-swapping
ability if it would just work well enough to have a keyboard and mouse. I
have the Access.Bus documents and hardbound manuals on I2C from Phillips,
but not the programming know how :) (still).


	Chris