Home Page Forums Archives Zinc ZiNc: DynaRec and Plug-ins

it is if its pointing to the next opcode implementation. so your 'cache' would really be a set of

cache[] = { lui(opcode), ori(opcode), lw(opcode), sw(opcode), nop(opcode)};

What about doing some form of branch prediction is that possible or practical?

Yes, thats kinda like it. For the func opcodes I just decoded as op_func & then went back and modified the value in the cache on the first call.

smf

With MIPS, I wonder how much you actually save. Decoding the opcode is pretty easy but with this you still need to decode the arguments.

At least you make the next opcode pointer predictable for the CPU (means pre-fetchable CALL) instead of a pipeline stalling/memory dependant switch().

How much does it save on an Athlon/Pentium M compared to P4 ? Something for benchmarks

Still, the BIGGEST saving for MAME or ZiNc on the CPU core would be CORRECT memory timing. Which means that you do not have to execute 33 million instructions but rather waste tons of these on fetch/load/store.

And with MAME I would really suggest patching the fetch/load/store with correct timing, and then bother with caching interpreter/dynamic recompiler if you still feel like it. Its all about accurate emulation vs. playing the games

Yes, dynamic recompilation will only happen once everything is 100% accurate and working. All the way along I've been relying on Intel/AMD bringing out faster processors to avoid having to optimise anything, maybe they'll pull another rabbit out of the hat.

Accurate timing is an alien concept to PSX emulation unfortunately. MAME doesn't even have half way accurate root counters yet, although I'm working on that.

smf

So with the rootcounters fixed, does it mean that you dont have to run as many instructions now ? Or is it still worth doing a big delay on every memory load ?

Well, 1.0 got slower than 0.9. 0.9 runs full speed the ZN games on my P3+Voodoo5 system. Not so for 1.0+. So I'm still using the older version for that system.

There's absolutely nothing different between 0.9 and 1.0 that would influence the framerate - if they're configured the same they should run the same.

Really? I'll try it out again then.

Nope, still slower than 0.9 on that system, same rendering settings. On 0.9 SFEX2P is 60 fps in game, in 1.1 it's fluctuating from 45 to 50 fps most of the time.

Something must have changed, better accuracy of emulation or sound or something, because it's certainly a tad slower.