Re: [hatari-devel] relative perf between 030 / DSP / CPU?

[Nicolas Pomarède <npomarede@xxxxxxxxxxxx>]

Le 26/06/2025 à 17:44, Eero Tamminen a écrit :
> Hi,
>
> On 26.6.2025 10.01, Nicolas Pomarède wrote:
> > Le 25/06/2025 à 22:34, Eero Tamminen a écrit :
> > > On 25.6.2025 10.56, Nicolas Pomarède wrote:
> ...
> > address translation, page swap in case physical page is available or 
> > not. Also I think the read/write functions themselves are not the 
> > cycle accurate "plain 68030" version when MMU is used, because MMU is 
> > not cycle accurate at the time and using CE memory access functions 
> > would just slow down everything for no real accuracy gain
>
> Enabling MMU emulation on top of CE slows Hatari noticeably, that's why 
> I though that it least _tries_ to be cycle-accurate...

in MMU mode every memory access must have its address compared to the 
TTx registers and possibly to the MMU "tree" that contains all 
physical<->logical mapping, so this will slow down emulation a lot, even 
if no translation are in place, you still need to check it on every 
access. This overhead is much higher than the CE overhead.