I couldn't get my EZ Drive to settle down enough to get a kernel compiled.
It seems that whenever the kernel prints the messages about "Target requested
too much data" it munges the inodes to the point where fsck can't recover
them. I think what the SCSI driver should do when it gets something like this
is resync the bus and run a re-try on at least the block that's getting
toasted. It might also be a hardware problem with me but haven't other people
complained about this drive wigging out?

Anyway, in my attempts to fix an unrelated problem I hosed most of my SCSI
setup. Nothing too serious, I just foobar'd a cable. My data's OK.
Unfortunately, I no-longer have access to my NetBSD drives. :(

I'm fed up and I'm going to get a gig drive in a couple of weeks and give
500 Megs to BSD. No more problems :) Until then I can only work on
the internal video driver in an abstract and non-compileable manner.
Besides, I'd like wait for a few more of the bugs to be worked out in the
NuBus code so that I have something stable to pattern off of.

There is one slight problem with internal video in that it doesn't look like
it has a nice, generalized solution. Unless machines share similar drivers
each machine is going to have to have a slightly different video controller
base mapped in. I'm also trying to figure out if the base address used by
the driver is a tride and true physical address or not. I need a util that
can dump out the MMU's page tables on the mac side. Does anybody happen
to know of one off hand?

Thanks.

_______________________________________________________________________
 Michael Zucca - mrz5149@rit.cs.rit.edu - http://www.rit.edu/~mrz5149/
 "I will choose a path that's clear. I will choose Freewill. "
  --Rush, Freewill
_______________________________________________________________________