Re: [MiNT] Problem with mint 1.16 installation

[Odd Skancke <ozk@atari.org>]

tor, 16,.06.2005 kl. 20.32 -0500, skrev Evan Langlois:
> On Tue, 2005-06-14 at 00:11 +0200, Frank Naumann wrote:
> 
> > As the XaAES is a kernel module and thus part of the kernel it's for sure 
> > fully legal to manipulate caches. I aggree with Ozk that it's much better 
> > and more user friendly if no TSR is needed, e.g. the TSR is the bigger 
> > hack of the two alternatives (and will for sure lead to much more 
> > confusion).

 Thank you Frank ;-)

> 
> I think the issue here is that the AES shouldn't compensate for a
> problem in a single driver when it isn't an AES problem.  Which is worse
> - change the code for everyone that it works perfectly fine for, or have
> the relatively few people with a problem install a hack to fix it?

 This is not a problem with "a single driver", it is a problem with all
video-card drivers on the Hades. If anyone have a Hades and all possible
video-cards, patch the file emulator.prg for each card and I'll remove
the clean Ssystem() call.

> 
> I don't know the "right" answer.  I'm a bit confused as to why the cache
> isn't transparent in the first place.  Is the cache caching some memory
> mapped hardware or something?  I'm wondering if there isn't some third
> alternative.

 This is a problem which comes from the fact that the 060 in the Hades
accesses hardare too fast, or something along those lines. The cache-
mode used on hardware accesses are correctly setup when one uses
set_mmu, which I wrote.

Best Regards,
Odd Skancke