>1.  Is there any good way to tell if I am really clearing the interrupt
>for the card ?

It aught to "work" at that point. It's stuff like this that makes
driver writing fun :)

>My values seem a bit "strange" compared to those for the
>other cards that are in the code.

The 8*24GC card is a weird card, it has a RISC accelerator chip amongst
other things so I would imagine that the interrupt clearing code would
probably be very weird. Though clearning interrupts is usually not more than
a few assembly instructions.

>2.  NetBSD (no matter what bit depth I have the card set to) seem to
>always think that the card is in monochrome mode.  Where exactly is this
>determination made, and how do I go about fixing this ?

The depth is passed to NetBSD from the booter. If you boot in 8 bit mode
the kernel should be able to know that. You're not booting with two monitors,
are you?

>3.  On the kernels that boot, it claims that ite is not configured.  How
>do I fix this ?

Ok, this is sorta related to your problem in number 4. Are you using a machine
with a 68040 in it? I think I have a hunch. I've been looking at the ite
code and there's a weird thing going on during the match phase that might
be related to 68040 machines

>4.  On the booting kernels, I can run a minimal console fine, but once I
>try using X (using the OSFA server), X starts up, but it never appears on
>my monitor.  i.e. I can login to xdm, type my username and password, and
>then use the xterm that eventually would appear, but I am doing it all
>blindly, i.e. my screen doesn't change from the getty login prompt (though
>xdm does lock/ grab the keyboard like it is supposed to).

Ok, the base address of video is wrong. The grf driver thinks its in one
place when its really somewhere else. I think this is a long standing bug
for some machines/cards. The kernel has too many values for the base address
of the screen and then it trys to use the wrong one or one of them is
the wrong value.

Quadra internal video had a similar problem until I added code that physically
calculated the address with the help of the video controller. While that
fixed the problem I think it may have masked a bigger problem.

Oh, BTW, if we ever get 24 bit mode working on X, you probably won't be able
to use it with your card. If I remember correctly, the 8*24 cards use
interleaved bitplanes instead of chunky pixels to fit 24 bit mode in their
small video memories. Our X server does not understand interleaved bitplanes.


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