OpenSource for Amiga.

[bloodline]

Atheist wrote:
Isn't micro-code rom circuits? Unalterable.

AmigaOne! 1GHz 68060 + 500 MHz AGA? Wish upon a star....

Well often it is!!! But when I was playing around with Open BIOS (in an effort to Replace the PC BIOS with AROS), I discovered that the PIII requries a microcode update during POST.
I imagine the intel chip is so buggy, that they had to allow updates after the unit has been shipped. :-o

[Ogun]

Facinating thought... A quick look at the Pentium III documentation show that the microcode* can be updated by the BIOS during the POST...

Yes, that is true. It can also be done at runtime through fiddling with MSRs.

Some with expereince in PentiumIII microcode code could, I'm sure recode the PIII to become 68k compatible...

Sorry to bust your dream here, but this will never happen.
a) Only very few instructions in the P6 family is implemented with microcode, mainly things that are slow and seldom used (divisions etc.)
b) If you know Intel P6 microcode, get a highpaid job at Intel at once, they'll love to have you onboard I'm sure. Reverse engineering CPUs is no easy feat

So no, this will not happen. Even the Transmeta Crusoe is known to be geared towards x86 emulation, and Linus T has said that the native assembly is hairer than mr. Stallmans beard.

So no go here.

[bloodline]

Ogun wrote:

Facinating thought... A quick look at the Pentium III documentation show that the microcode* can be updated by the BIOS during the POST...

Yes, that is true. It can also be done at runtime through fiddling with MSRs.

Some with expereince in PentiumIII microcode code could, I'm sure recode the PIII to become 68k compatible...

Sorry to bust your dream here, but this will never happen.
a) Only very few instructions in the P6 family is implemented with microcode, mainly things that are slow and seldom used (divisions etc.)
b) If you know Intel P6 microcode, get a highpaid job at Intel at once, they'll love to have you onboard I'm sure. Reverse engineering CPUs is no easy feat

So no, this will not happen. Even the Transmeta Crusoe is known to be geared towards x86 emulation, and Linus T has said that the native assembly is hairer than mr. Stallmans beard.

So no go here.

Ahh!!!! Well that clears that up :-D

Why didn't you pipe up when I was ranting about this on the Mailing list, last year  ;-)

Cheers

[Atheist]

Hi bloodline,

Do any CPUs after that (486, P1, etc.) have that "feature"?

If you just reverse the binary instruction (to fix endian prob.), I guess it's not that simple, it never is, is it?

AmigaOne! closer and closer.

[bloodline]

Atheist wrote:
Hi bloodline,

Do any CPUs after that (486, P1, etc.) have that "feature"?

If you just reverse the binary instruction (to fix endian prob.), I guess it's not that simple, it never is, is it?

AmigaOne! closer and closer.

As Johan, pointed out only certain instructions are microcoded (as such instructions tend to be slower than raw instrucitons.), a much better solution would probaby be the Transmetta Crusoe... I have Spoken to a Transmetta engineer, and he says that it would very simple to write 68K firmware for the chip, but no one has any need to do it :-( so it won't be done.
If only someone with a lot of money could afford to hire the Entire Alpha... I mean Athlon design team adn amke a 68K version :-/

[bbrv]

Edited by Off Topic: Argo




P.S. We were serious! :-)

[DaveP]

PS: We don't care.

[bloodline]

I thought I had more to say than I did... It turns out that I didn't and thus this message is just wasting band width...  :-D


-Edit- Why not start a new thread then? Possible Opensource futures for AmigaOS...

[bbrv]

We understand how you feel Dave...:-p

R&B
_________________
Declared lack of interest:
AmigaONE XE purchasers.
Owner of AmigaBooks.org community authorship programme.
Undeclared lack of interest:
   Loads.  :-o

[DaveP]

Change that little p to a capital P and you get the icon version. :-)

[bbrv]

It was just a little "p" ...


:-D