Search results

I don't think that would help. This is either some race condition somewhere, or something unhandled in general in the emulation. Or some wrong opcode emulation. I'm currently checking things all over the place in the hope to find more details. But i desperately still need somebody skilled in...

I by now also experimented with that one. So waiting on the respective MPE until a DMA is finished up, similar to the real HW, but with no positive change. Instead, obviously the emulation just got slower overall. :/

Just a small comment on that: As also the Jaguar emu BigPEmu was brought up somewhere else, and how well it performs (and its great, too, btw :)), note that the Nuon is a completely different beast, even compared to the already complicated Jaguar. So to put that in perspective: First of all...

Just a heads up: 0.6.5 has been released: https://github.com/andkrau/NuanceResurrection/releases/tag/0.6.5 btw: Still searching for help on the T3K hang issues. Somebody skilled at reading/interpreting T3Ks audio code and how it syncs would be much appreciated..

From the video available on Steam, the newest game is T2K (and a gridrunner remake).. EDIT: also the game listing on the publisher page indicates that

New release with minor fixes: https://github.com/andkrau/NuanceResurrection/releases/tag/0.6.4

Could you make a PR for this? Cause i don't know exactly which code/comments you mean. Thanks!

New version: https://github.com/andkrau/NuanceResurrection/releases/tag/0.6.3

Its not that straight forward though. One would have to buffer things that should be in flight whereas at the moment things are 'just done'. But maybe its the only solution, who knows. (Jeffs source code could tell ;))

Then back to the T3K deadlock issue.. One theory i have is the following: Could it be that due to the way the 'instantaneous' DMAs/Comms/etc work in Nuance, that T3K actually relies on these operations to take more cycles than 1? So that the operations are executed too fast and thus the T3K...

I just tweaked some more things in your commit, could you please doublecheck/review?

Thanks! This was also already available on archive.org, but isn't this just a demo? Or the full thing?

The other ones were flagged, too. There is also a issue on github about it. I guess its due to the way the HLE works (generating code, etc).

In principle i agree, but at the moment the major focus is on making things work first. ;) But if anyone has a patch to change the UI, i'll happily merge it.

Neat! Thanks for testing! The vector test necessity is the one that still confuses me most, especially as its also only necessary when dynamic recompilation is enabled. This means that there is at least one bug with that one left. As for the music test/gameplay: If you can play the game and...

Great! And good to know. Be sure to compile 'Release' and 'x86'/32bit, as the 64bit version does (at least) not support the dynamic recompilation.

Can you compile the Nuance proj on your own?

Changes are commited now. Performance should also be roughly the same as before (0.6.2) again. Sometimes (peaks) its a bit slower, most of the time even a tiny bit faster. ..oh, and to the coders in here: Please review my changes if something is now fishy..

And Merlin Racing adventure mode now also works :)

Good news! I found one more source for the hangs/memory corruptions, so now one can finally enter (and finish!) Merlin Racings Arcade mode. This was also one of the sources that makes T3K hang, BUT unfortunately not the most prominent one, so T3K is still very buggy. :/ Downside: This makes...