Re: Linux on Game Cube

Hi Jimmi! I'm Cc:ing this to the mailing list and I would suggest that 
you also join it ( 
http://lists.sourceforge.net/lists/listinfo/gc-linux-devel ), okay?

On 20.01.2004, at 22:08, [email protected] wrote:
> I've just seen the new GC-Linux porting project, after looking from 
> time to time
> on the Xbox Linux website ... These projects are both hi-tech and fun, 
> it's great !

:-)

> I'm dropping these lines to you, because i think there are already 
> avaiable code
> that may help you on this project.
>
> The Game Cube main CPU is a custom chip (ASIC) made for Nintendo, 
> including the
> PowerPC 405 CPU and various peripherals. Besides the PPC 405 GP from 
> IBM,
> another chip includes this processor core : the Virtex II Pro from 
> Xilinx, which
> is a FPGA (you know, a programmable logic device). And there is 
> already a Linux
> port on this chip :-)
>
> There are even two : Mind Linux Solutions (www.mind.be) did it for 
> standard
> Linux kernel, and Montavista Linux (www.mvista.com) did it for the 
> real-time
> version of the Kernel, plus a complete distribution. Both these ports 
> are GPL,
> so you could get PPC405 code without any legal headache, and grab some 
> boot
> code, drivers for Ethernet, RTC, or X11 port, with minimal changes 
> required.

Hmmm... The documentation on the internet that I found, said "PowerPC 
750CXe" ( http://www.anandtech.com/showdoc.html?i=1566&p=2 ), and 
that's PPC7xx, not PPC4xx... (what's the difference anyway?)

Well, Linux exists for both PPC7xx and PPC4xx, i.e. all code that is 
CPU specific is in the standard kernel tree. And I think neither of 
Ethernet, RTC and video of the GameCube are compatible with these 
devices, as they are all implemented outside the CPU. :-(

> Are there already shcematics, or at least block-diagrams of the GC 
> arhitecture,
> available on the web ? It may be possible to map VHDL models of the 
> peripherals
> in Virtex II Pro FPGA, to get a developpment system for GC without the
> limitations of the GC (memory size, CD drive ...). Finding some JTAG
> connector/test points on the board could also give access to the debug 
> port of
> the PPC405, allowing easy trace and debug, esp for the boot process.

That's the way a hardware person thinks. :-) As I software person, I 
think quite differently - I would just use the real thing as use some 
debug code (that I know that works) to find out whether my code is 
still alive at a certain point. It is my impression that building 
custom development hardware might be too much work, as so much is known 
about the GC already.

> I'm sorry i can not help you much for the moment (i'm already busy 
> trying to
> hack some Linux on Xilinx chips) and i don't even have a CG (despite 
> it's
> getting rather cheap here, around 100€). May be i could help on some 
> french
> translation ?

Certainly. Anyone interested in doing translations, please join the 
"docs" mailing list ( 
http://lists.sourceforge.net/lists/listinfo/gc-linux-docs ) and 
announce it there. But since you seem to be a very experienced hardware 
person, I am sure you can help in other ways - a bit?

> Good luck, and have fun with you Tux-customized consoles !
> Jimmy

   Michael



-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration
See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn