Re: [hatari-devel] 68030 MMU work

[ Thread Index | Date Index | More Archives ]

Am Thu, 15 Nov 2012 11:47:51 +0200
schrieb Eero Tamminen <oak@xxxxxxxxxxxxxx>:

> On torstai 15 marraskuu 2012, Thomas Huth wrote:
> > 
> > FYI: X-Tasie is working now!
> Verified.  (It's ugly and sounds awful, but works)
> > It's the first working Falcon program that
> > is using the MMU for page swapping :-)
> Why I need to specify "--mmu true", shouldn't MMU be enabled
> by default for Falcon (with WinUAE)?

MMU is still experimental, and it might also be slower than normal
emulation, so I would leave it disabled by default for now.

> > There is still a glitch in the first screen, you should not move the
> > mouse there before pressing the button, but that's not related to
> > the MMU but to Videl emulation.
> Is this Hatari glitch or bug in the demo itself?

Looks like the demo writes negative values to the "vertical display
begin" register... I don't know whether this is legal on the real
hardware, but it caused a segmentation fault on Hatari. I've added a
hack yesterday so that the segfault does not happen anymore, but the
screen still gets messed up in that case. I think this needs some more
investigation by a Videl expert one day...

> (I wonder why you left the "minor" issue color to the X-tasie
> entry in compatibility list, and the version should be set to
> "1.6.2+", to indicate that it doesn't work yet in 1.6.2.)

Oops, I've forgot to bump the version, ... I've fixed that now.
But the "minor" issue color has been done on purpose since there is
still this Videl glitch left.


Mail converted by MHonArc 2.6.19+