Re: [MiNT] libgem16: vs_color

[Eero Tamminen <oak@helsinkinet.fi>]

Hi,

On Monday 18 January 2010, Helmut Karlowski wrote:
> Am 18.01.2010, 20:12 Uhr, schrieb Odd Skancke <ozk@atari.org>:
> >   Then I would suggest fixing the problem where the problem is, and not
> > clutter up the XaAES source with #ifdefs to accomodate for faulty
>
> Guess why I spent several hours looking for the reason why vs_color is
> not working. But as for now my workaround does.

Well, changing things to static just hides it.  The variables go then to
different address & section in code and something else gets overwritten
at some other point and then there's again a similar debugging session...


> > but at first glance I cant see anything wrong. If I misunderstood your
> > answer, I'm sorry .. you do say that the corruption of stack/vdipb only
> > happens under fVDI?
>
> My new theory is that the (addr & 1) - test in fvdi is not legal under
> aranym.

Are you using JIT?  Does this happen only under Aranym?


	- Eero