[Coco] Re: Nitros9 and the '86 GIME

Robert Gault robert.gault at worldnet.att.net
Thu Sep 30 11:27:26 EDT 2004


Charlie wrote:
>>From my setup:
> 
>>What are you booting, 6309 or 6809 version, LevelI, LevelII, LevelIII?
> 
> 
> 6809L2
> 
> 
>>When does the boot process die, immediately, at the NitrOS-9 Boot
>>screen, during the module boot listing, after sysgo starts? How much
>>memory, 128K or 512K?
> 
> 
> The Nitros9 boot screen, usualy after KREL Boot Kernel ... *CRASH*
> 
> 128K or 512K
> 
> Charlie
> 
> 
> 
> 

OK, what is being described by several people corresponds to a crash 
either in the models rel, osp1/krn, or boot. Since the crash occurs 
after rel boot and krn have loaded (see quote above), that means track 
35 has been loaded, rel has started and initialized the GIME. Rel, boot 
and os9p1/krn have been relocated in memory, and the code in rel has 
jumped to os9p1/krn.

Since the code works with an '87 GIME, we know that rel jumps to the 
correct location to start os9p1 running. My guess, since as described 
above very few ..... and no other module names are seen during the boot 
process, os9p1 is crashing during the setup of the GIME IRQ & SWI 
routines, before boot tries to load the os9boot file.

At one time I had an '86 GIME but have not been able to find it. That 
means I can't do any testing. The way to do the testing would be to put 
trap routines at strategic stops in rel and os9p1/krn to see exactly 
just how far the boot process gets. At some point in the sequence of 
traps, video will get corrupted and we will have found the offending 
sequence of code. If needed, boot and os9p2/krn2 can also be tested

Any volunteers?




More information about the Coco mailing list