From: teunis <teunis@computersupportcentre.com>
  To  : ggi-develop@eskimo.com
  Date: Fri, 11 Sep 1998 19:27:54 -0700 (MST)

Re: ViRGE/TRIO clock driver

On Fri, 11 Sep 1998, Stefan Mars wrote:

> Well, I am about to finish up a version to put in CVS (look for it in a
> week or two I would assume (busy at the university)), but I have a few
> questions that I would like opinions on.
> 
> Considering that this is an integrated clock, I think it would be stupid
> of me to try to detect it, since the chipset driver does that, and the
> chipset driver is initialized _before_ the clock driver. This, however,
> does create a small problem...
> 
> I can't use configuration in IOspace, because the chipset driver might
> (and even should have) put the ViRGE/TRIO chip in MMIO mode already.
> Hence, I will have to assume it is in MMIO mode. Teunis is currently
> reworking the chipset driver, will this be a problem for you Teunis?

No probby....

> Then I somehow has to figure out the base MMIO address... hmm, might need
> some thinking..

Both that (and whether the driver is in MMIO mode) is/should be in the
chipset info..... *grin*

ZUT! it's in the [private data] section.

Any suggestions?

Is there a public part of GGI that has this info?

> Anyway, the second matter is this: This is a complete rewrite of the clock
> driver, but currently it lives under the same name as the old one,
> s3idac.c. Shall I simply replace it and have it over with, or should it go
> in under a new name so we can have both (for compatibility issues) for a
> while.
> 
> Personally I would prefer just replacing the damn thing, the old driver
> was very very very bugged anyway, and I am not at all amazed that people
> have experienced problems with ViRGE/TRIOs.

I can't vote...  (don't really care as long as it works :)

G'day, eh? :)
	- Teunis