Re: Trouble with freeMinT 1.15.0

[Konrad Kokoszkiewicz <draco@mi.com.pl>]

> > If so, I'd be rather surprised.
> > 
> > OK, the point is that -> I <- optimized the MiNT scheduler. 
> > I am rather surprised that it makes problems on 040+, because:
> 
> You apparently missed the point:
> the problems happen on plain ST machines, not 020 or better...

"Plain" ST? I heard that it was MSTE (68000 + cache) and some ST turbo
board (with cache as well).
 
> > By the way, I'd be for not restoring the previous code if possible.
> > It was very ugly and unefficient, especially the part known as
> > enter_kernel() and leave_kernel(). The code is still there (commented
> > out), so anyone can compare it with my versions (inside kernel.spp).
> 
> Which is where the problem comes from....

You can't be sure. And generally I am surprised that the previous version
of code didn't cause any problems: system trap vectors were changed at
every system call and taskswitch. This was plain nonsense and wasted a lot
of CPU time IMHO.
 
> I haven't looked at the sources, but I suspect that you
> might have used 030 instructions on code that was meant
> for 68000... Just a guess.

Well, no.

--
Konrad M.Kokoszkiewicz
|mail: draco@mi.com.pl                  | Atari Falcon030/TT030/65XE |
|http://www.orient.uw.edu.pl/~conradus/

** Ea natura multitudinis est,
** aut servit humiliter, aut superbe dominatur (Liv. XXIV,25)
*************************************************************
** U pospolstwa normalne jest, ze albo sluzy ono unizenie,
** albo bezczelnie sie panoszy.