AROS 68k EMU

[bloodline]

Crumb wrote:
Wouldn't it be possible on systems with the same endianess of the 68k to run an integrated 68k emulator like in OS4 or MOS?

Yes, exactly like MOS and OS4.

[bloodline]

xeron wrote:
Endianness doesn't matter. You could make an x86 OS with integrated 68k emulator along the same vein as MOS and OS4, if you wanted.

But the 68k code would fail when if tried to make an AROS x86 function call since the byte order in the data structures would be backwards.

Unless you have a better idea... if so I would love to have you join teh AROS team and develop your idea :-)

-Edit- You *could*make a Big Endien version of AROS x86, but that would have a performance penalty...

[bloodline]

4pLaY wrote:
The 68k version of AROS havent been updated since 97 so no :/ anyone is free to port it if they wish =) would be fun to see how AROS did on my A1200.

That's not quite true. The Palm version has been updated and boots, this uses the 68k :-)

But the Amiga version of AROS has not been developed for a while... Maybe if someone wanted to donate a working ZII network card for Johan's A3K he could get AROS booting on the Amiga :-)

[bloodline]

No chance. I'm not interested in AROS. It seems like a really nice project, but I have OS4

Without any code to support your claim, I have to say you are wrong :-)


@Crumb:

The Best solution is to have a *special* version of AROS 68k running in an Emuator. This *special* version of AROS would not implement many functions but instead pass most over to the Native AROS x86 running the machine. This would make the 68k program appear to be running nativly with the added bonus of Chipset emulation!!! :-D

[bloodline]

Well, if you mark a zone with the MMU to use it with 68k programs and other for native software you could share memory because you could know where you have to reverse words and longwords and were you haven't.

I'm not saying that it's easy but I think that it's possible. If you can notice if a program is writing on an area where you have to invert the bytes and in other were you don't and everything is aligned... it may be possible wouldn't it?
 

Even if that were possible, the performance hit would be far below The emulator option I have already suggested.

You seem concerned about performance penatys from the emulator option, but I assue you, 68k software will run much faster than any real Amiga (as it will ahve the UAE JIT) and any 68K->x86 AROS overhead would be not much greater than a byte swapping + function call....

[bloodline]

Crumb wrote:
@bloodline:
"You seem concerned about performance penatys from the emulator option, but I assue you, 68k software will run much faster than any real Amiga (as it will ahve the UAE JIT) and any 68K->x86 AROS overhead would be not much greater than a byte swapping + function call...."

In amithlon native software reverses the bytes of the word and achieves a 75% of performance. We would have to add the overhead of the 68k emu, but UAE would be hardly faster.

And you are forgetting than on big-endian systems an integrated 68k emu would be FAR faster than the approach taken for the x86 port. It would be as fast as in MOS or OS4.

Several issues here:
1. Only the actuall function call would have any overhead, the function itself would be executed at fll speed natively on the CPU.

2. The Emulator would run as fast as any JIT, since only the actual OS function calls need to be passed to the Host OS, everything else runs via the JIT (as in MOS and OS4).

3. On an x86 system, you tend to have quite a bit of raw horsepower available even on low end systems.

4. By using an Emulator we do have the ability to use Games and Demos that hit the hardware.

5. By Big endien systems, you are refering to the PPC, which will be increasing in performance and AROS would have the advantage of harware hitting software, which is less likely to be ported to the new CPU anyway :-)

[bloodline]

Crumb wrote:
The difficulty I see is that this approach would require that AROS was nearly finished (I mean, that it was 100% compatible with 3.1)

I guess that finishing the Amiga 68k port would help a little... once it's done it would be a matter of changing the "rom" 68k functions with functions that called native ones.

AROS is suppsoed to be 100% compatible to 3.1. Getting the 68K version to work is a trivial task, it just that the only persopn with the time and experience to do it has a broken Network card (ZorroII)  in his Amiga 3000 so he can't do much at the moment (how to get 100megs of sources to the A3K?).

[bloodline]

Crumb wrote:
BTW, what do you think about the idea of creating ONE distro that generated code for all amiga-like OSes hosted in OS3.x (firstly, then it would be ported to AROS, MOS, OS4, Amithlon, MacOS, Linux, windows...)?

Most of Amiga developers won't have linux or aros installed on their systems so there are more chances of seeing AROS soft appear.

In addition to that stupid wars would end because developers will be able to compile binaries for all targets changing only one flag.

Look up openamiga.org :-)

That is the place to start such an idea :-)

[bloodline]

melott wrote:
I was going to say that I'm disappointed that AROS
had abandoned the idea of Classic Amiga
compatability. But after reading this thread maybe
not.
As I see it, any 'Amiga Like' OS, to have a chance
of being anything more than just another 'Also Ran'
has carry with it Classic Amiga compatability.
Without that compatability I don't really see any
reason for it. There are already a number of other
OS's that have very little support.
The Classic Amiga is a cult, look at how many ppl
now don't even consider the A1 a real Amiga.
No Classic compatability.
Without Classic compatability AROS ( IMHO )
is a waste of time, whats the point??

Once such a system is built for AROS, I don't see why it couldn't be ported to MOS and OS4.

[bloodline]

BTW: with your idea of making using native calls from UAE etc... how would you mix AROS native and 68k libraries and devices? It sounds as it would be much more difficult than using an integrated emu.
 

That is the only drawback with the emulator system.

Native tasks could not access the 68k libraries and devices.

But it would not take much work to make the *Special* 68k AROS build maybe a speical calling library on the Native side (That traps calls and directs them to the Emualtor)... hmmm I think it could be done actually... I have to assk Fabio though.

[bloodline]

ptek wrote:
Personally, I don't see the reason for all the fuss about AROS ...

OK, I agree that's a serious project and the people involved are surelly skilled ...

But I have only one complain : Did you ever tried to run it on a 486 on the past ?

Before you laugh at me, let me say that I tried AROS some years ago when I have a 80Mhz 486 and that thing was damn slow !

Are you blaming the 486 ?
Don't.
On that same 486 I also run win95 and that was fast !
Usable.

OK, AROS is AMIGA like, now we have GHz CPU, blahblahbla, but I was totally putted by the lack of optimization of AROS ...

I really looked like a 68K emulator (which it isn't!)

Since AROS is written in C, we did not use any optimisation until recently, sicne optimisation makes bug hunting more difficult.

THe Graphics drivers you would have seen were the first ones made and they only ued the most basic VGA display (which is slow on my 3Ghz machine). The Windows95 you were using would have had speciall build accelerated GFX drivers.

Basicly AROS only looked slow :-) I tried AROS on an old P75 a year ago and it was faster than my 1200 with a BlizzPPC :-)

-Edit- I'm not blaming the 486, it's a powerful CPU, I'm blaming the lack of AROS specific gfx drivers for your old GFX card.

[bloodline]

Crumb wrote:
@Matt:

How do you install AROS?
The floppy disks of sourceforge or your web can't boot, they complain about lack of execute, copy and other commands...

The CD includes some kind of installation? I couldn't find it... a friend of mine wanted to install it on a 486 laptop but he couldn't find how to install it.

Oh do you plan to add swap memory soon?

The CD is currently required for an install. It's a very slow process, but it does actually work.

Before we include Virtual memory (Which I personally hate and have switched off on all my machines) we have to finalise a design for an mmu.library (we need some standard, platform independant way to access all the different MMUs that exist.)

[bloodline]

Crumb wrote:
"we have to finalise a design for an mmu.library (we need some standard, platform independant way to access all the different MMUs that exist.)"

If I were you I would send Thor an email to see what he thinks about a standard MMU.library, it seems he has a lot of experience with MMUs :-)
I don't know the level of portability of this library, but it would be a good idea to try it.

BTW reimplementing Prometheus.library for AROS would be great... and later aplying the openpci.library wrapper, of course.

Well, as with everything in AROS... all it takes is someone to get on and do it :-/

Brand new PCI drivers are on the way:

[bloodline]

Crumb wrote:
Cool snapshot! :-)

I don't plan to install Linux and as there's no gcc that can select different targets (AROS/68k, AROS/i386, OS3.x, WarpOS, PowerUP, MorphOS, OS4) I doubt that I'll be able to compile anything for AROS...

Well, AROS will come with GCC included for the HD install, so you could actually develop within AROS :-)