Re: [hatari-devel] Re: 040/060 emulation issue with BadMood

[ Thread Index | Date Index | More Archives ]

In Hatari (and WinUAE), this speed difference is handled by assuming all instructions will take only 1 cycle in 68060 mode ; this is of course not correct and will be noticeable when you try to sync with the DSP where timings matter.

Yes I'm aware of some/most of the 040/060 differences - but expected the opposite situation. i.e. Hatari clock is 32MHz even in 060 mode, and the real 060 is superscalar (effectively 0.5 cycles per clock when pairing well). So Hatari should theoretically be slower than real HW for two reasons, and less likely to lock up on DSP timing issues. However what I encountered is the opposite - lockups measured as less likely on a CT60 @ 75, than in Hatari @ 32.

Anyway, it's not a big deal for me at the moment as I'd prefer to just find and fix all timing issues by default, and just relax ones known/measured to matter, where its safe to do so. Hatari has been helpful in doing that, simply because it seems more sensitive than expected.



Mail converted by MHonArc 2.6.19+