Re: [MiNT] Wrong basepage pointer?

[Eero Tamminen <oak@helsinkinet.fi>]

Hi,

On keskiviikko 02 tammikuu 2013, Miro Kropáček wrote:
> > Basepage size is 256 bytes and code is supposed to normally start
> > at position basepage + 256.  Process command line is from offset
> > 128 to 255.
> 
> The thing is that *a* code starts there. It's just it's not *my* code but
> some artifical mintlib hack code which jumps (if I'm not mistaken) to my
> code.

My point was that if you got some stuff from your program, other than
its command line at offset 228, your basepage content is corrupted.

If you run your program (using PE_LOADGO) with a command line that is
e.g. 120 characters long (long enough to go over basepage + 228, short
enough not to go into extended args in environment), and you print
the command line from the executed program, does it look fine?


> > Where you're running this?  This sounds awfully like the Hatari
> > GEMDOS HD emulation bug I debugged & mentioned a while ago on
> > this list and which was fixed in Hatari almost two months ago
> > by Thomas.
> 
> Well, I run this on Aranym+FreeMiNT. When I think of it again, I guess
> it's not possible to fix since there's not only FreeMiNT but also TOS,
> EmuTOS, ... they have no clue about some a.out hacks in TOS header...

Does this issue happen also on real device, not just under Aranym?

(There may be some other things than Aranym that could affect it,
like Meta/BetaDOS driver.)


	- Eero