Re: [hatari-devel] Basepage and proc_lives debugging from the debugger

[Matthias Arndt <marndt@xxxxxxxxxxxxxx>]

Am Sun, 17 Nov 2019 13:09:35 +0100
schrieb Christian Zietz <czietz@xxxxxxx>:

> Matthias Arndt schrieb:
> 
> > Feel free to edit my patch for anything that is not to common Hatari
> > standards. It has been a long time for a code contribution on my
> > end.  
> 
> I have one minor objection: Because of the way the data seems to be
> saved in proc_lives, you only display the bottom 24 bits of the PC.
> 
> Please check what Atari TOS does when the PC is larger than that. If
> it really only saves 24 bits, I'd suggest to make this somehow
> visible to the user of "info postmortem". Otherwise, the user will be
> confused: "But my program was running from TT RAM. Why did the
> exception occur in ST RAM??"

The TOS documentation says the exception number is stored in the top
byte of the PC in this case.

My guess is this is not handled specifically for TT or Falcon in
official TOS versions as tens of MB of RAM was too pricy back then to
take their address range into special consideration.

I don't know how this is handled for TT or Falcon. Feel free to clean
up so that it works in all cases.

See
http://info-coach.fr/atari/documents/_mydoc/Hitchhikers-Guide-V1.1.pdf
and the relevant chapter in the Atari ST Profibuch 1988 edition as my
sources of information. 

Best regards,
Matthias

-- 
http://final-memory.org/