[hatari-devel] Emulation reset, old-UAE vs. WinAUE core

[Nicolas Pomarède]

Le 18/06/2011 10:43, Eero Tamminen a écrit :
> Hi,
>
> I started to wonder why breakpoints didn't work anymore after emulation
> reset and noticed that while old UAE core always preserves the debugging
> flag on reset:

> So, why WinAUE always resets the debugger flag?

Hello

Well, I guess WinUAE's core (not AUE please :) ) as being directly taken 
from the Amiga version doesn't include this notion of debuger.

It's fine for me if we don't backport everything from current UAE to 
WinUAE, because I think we should first make the cpu core works before 
backporting our own additions (else, we might add side effects and makes 
it harder to fix the winuae's core in Hatari).
The less we modify WinUAE's core for now, the more we can be sure the 
remaining problems are in Hatari and we have a direction to look at.


> Is there any reason why we wouldn't want it to be preserved always, either
> by explicitly preserving it in all of these places, or by always re-setting
> it after reset, like in the attached patch?

Your patch looks good to me, I also think debuging features should be 
preserved when reseting (cold or warm). You can apply it if you like.


>
>
> Then I noticed also a problem related to WinUAE's resets.  While all
> old AUE m68k_reset() calls come from Reset_*() ->  M68000_Reset()
> which is called on specific configuration changes or on keyboard
> shortcut...
>

>
> Should uae_reset() calls also be converted to DlgAlert_Notice()
> calls so that in those cases user also knows what happened and
> can reboot the emulation at his/her convenience?


Why not ; I don't consider this a top priority over the cpu emulation + 
STF/STE integration, but if you have a patch for this, let's go for it.


Nicolas