Re: MOL on Amiga/APUS again.....


Subject: Re: MOL on Amiga/APUS again.....
dpates@dsdk12.net
Date: Wed Dec 20 2000 - 15:02:36 MST


Quoting Gianluca <icjtqr@tin.it>:

> Yep. You are right. You wrote "MacOS depends heavily on this chip for
> most of
> its functionality..."
> But MoL it's a MacIntosh Emulator. Completely software based. Heard of
> MAME? or
> Basilisk or ShapeShifter? Every Emulator, catch every call to low-level
> i/o
> chips and redirect evrything on a higher level. So it needs so much
> horsepower
> to emulate something relativly simpler.
> Try to think about a PowerComputing Apple Clone (quite different from a
> real
> Mac...) running MacOS...

MoL doesn't emulate anything except some peripheral hardware. Processor-level
instructions are executed directly on the PPC, through processor virtualization.

>
> The main advantages are if you are not emulating the opcode on the
> host
> processor as MoL. It doesn't waste CPU time to translate PowerPC
> Opcodes,
> because it is running on a PowerPC CPU System. It is like WinEmu for
> Linux
> Intel Based, or ArmEmu for Archimedes ARM Based RISCOS..or Emulating a
> 68k Mac
> on a 68k based Amiga (only the access to i/o chips are wrapped into
> AmigaOS
> System...Sometimes better than MacOS counterparts...).
> The main part is NOT emulating the PROCESSOR ITSELF, but only the
> chips
> addressing and operation. (Simpler than an entire system, anyway...)

Then why is the Mac IO controller mapped as part of the MoL startup phase?

>
> Have you ever heard of Alpha Processors? I saw a Emulated MacOS
> (Basilisk)
> faster than a fastest G4 on earth... running 68k software...
> Imagine if AlphaCPU can obtain the same speed from emulating the RISC
> PowerPC...

Yeah, if you have the ROMs and emulate everything else (including the Mac IO
controller). The Alpha AXP is an amazingly fast CPU, so if you have a well-honed
emulator, sure, it can be lightning fast.

If it works, hey, that's great, but I don't know that it can without the Mac IO
controller chip. Go ahead and try it, and see what happens, though, and let us
know what the result is.

Derrik Pates
dpates@dsdk12.net



This archive was generated by hypermail 2a24 : Wed Dec 20 2000 - 15:02:36 MST