Re: [hatari-devel] New WinUAE core issue with Bad Mood?

["Roger Burrows" <rfburrows@xxxxxxxxx>]

On 9 Dec 2014 at 0:31, Nicolas Pomarède wrote:
>
> Regarding 6ec030, this is what Tony told me regarding TC and his
> experience with 68ec030 on Amiga :
>
> "Enforcer does use full MMU and it can work with some EC030's, at least
> partially. I had one, Enforcer worked, most other programs that use MMU
> simply hung the system immediately.
>
> As I said previously, there are 680EC030's that have partially broken
> MMU, most operations work but some don't. Probably 68030's that got
> relabeled as EC because some MMU-only factory check/test failed."
>
>
> So, it seems checking TC is not 100% foolproof to ensure you have a
> working MMU. If the cpu was made as a 68030, but failed at some quality
> checks in the factory and was relabeled/sold as a 68ec030, TC register
> will exist, but CPU might give undefined results when really using MMU
> translation's tables.
>
> Well, I'm not sure some 68ec030 were used in some Atari or TT/Falcon
> extension boards, so this is rather theoretical in the end (unless you
> want EmuTOS to run on Amiga :) )
>
I think it does actually ;-).  I suppose to handle the broken 68030 case, there
should be an option in EmuTOS for the user to override the automatic detection
.... but we'll leave that for later :-).

Thanks for all the great work on Hatari,
Roger