Re: Re[2]: GREAT NEWS! ! ! !

From: Jo Even Skarstein <joska_at_nuts.edu>
Date: Fri, 12 Sep 1997 16:59:26 +0200

On 12 Sep 1997 Geir.Validalo_at_softnet.no wrote:

> The biggest problem is that we need to initialise both processors,
> which actually means two boots! If we could control the 030's reset
> signal and control the address from which it boots, then it all would
> be much easier. But we'll see what I can do...

This is what the Atari Compendium says about this:

"Upon a cold or warm boot, microprocessors in the 680x0 series load
the initial supervisor stack pointer from the first longword in memory
($0) and begin execution at the PC found in the second longword ($4).
The location this points to is the base initialization point for Atari
computers."

If it was possible to change these values things would be very easy:
Boot the 040, change $0 and $4 and allow the 030 to use ST-RAM when
the ST-bus is idle. But since the two first longwords in memory are
located in ROM, we can't just change these. $4 also points to ROM so
it's impossible to change the actual boot-code :-(

I guess the only realistic way to work around this is to patch the
ROM.

To actually *use* the 030 we could add another Pexec() which
explicitly loads a binary to ST-ram and initialises the 030. With MiNT
then *true* multiprocessing should be possible :-)


/*
** Jo Even Skarstein http://www.stud.ntnu.no/~josk/
**
** beer - maria mckee - atari falcon - babylon 5
*/
Received on fr. sep. 12 1997 - 16:59:26 CEST

This archive was generated by hypermail 2.3.0 : ti. nov. 03 2015 - 20:07:53 CET