Re: [hatari-devel] Bug report: AHCC not working anymore with GEMDOS emulation

[Eero Tamminen <oak@xxxxxxxxxxxxxx>]

Hi,

On keskiviikko 06 kesäkuu 2012, Matthias Arndt wrote:
> Am Mittwoch, den 06.06.2012, 00:19 +0300 schrieb Eero Tamminen:
> > On perjantai 20 tammikuu 2012, Matthias Arndt wrote:
> > > Config as always 4MB STE with TOS 2.06
> > > 
> > > Any ideas?
> > 
> > Are you by any chance using WinUAE core?
> > (I was testing old UAE CPU core.)
> 
> FOr ST/STE I'm using the old core. Didn't everyone tell here, it was
> considered the stable one for ST/STe emulation?

Sure, I was the only remaining thing that I could think of
that could explain why it fails for you but not for others.

To recap, the problem is that when using *both*:
- NVDI v4
- GEMDOS HD emulation

AHCC command line compiler will either bomb when you try to
use it, or not give any output (which?).

And the problem doesn't appear with AHCC GUI compiler, or
if NVDI v4 is removed nor when using ACSI emulation instead
of GEMDOS HD emulation?


I've now tried AHCC with NVDI v2.50 and it works fine.


Next you need to try following things:

* Start Hatari with --bios-intercept and see whether it
  outputs something extra on console in addition to stuff
  you see on screen

* Try whether the issue happens with some other TOS
  version than v2.06, e.g. with latest EmuTOS:
  http://sourceforge.net/projects/emutos/files/snapshots/CVS-20120603/

* Try starting AHCC command line tool without arguments
  to see whether it outputs then anything / crash

* If it still crashes, go to debugger (AltGr+Pause)
  and use "setopt -D" to get to debugger when there are
  exceptions, then start AHCC again and check what
  happens when the crash happens:
  - Is PC within AHCC or TOS ROM
    ("info basebase" tells AHCC address)
  - What instructions ("d" output) and register values
    ("r" output) caused the crash, was it e.g. NULL
    pointer access in AHCC?


	- Eero