23 Aug
2005
23 Aug
'05
9:53 p.m.
Wow, what a great collection of hardware! I'd really love to find one
of those machines some day. I wrote part of the console firmware for
the 6200's and 6300's.
Selftest runs initially using RAM in the System Support Chip on the CPU
board. At some point though, the CPU's need to communicate through
memory to run extended selftest. It sounds like that's where you're
seeing the failure. I don't remember what those LED codes
mean...hopefully you can find some documentation.
Just to be sure. you might check the memory interleave configuration
from the console. With 3 modules, you might have to
set them all
non-interleaved.
It might be difficult getting NetBSD running on that system since DEC
was particularly paranoid about giving out the programming details for
hardware in those days. At least you have DEBNI's (which were reliable,
unlike the earlier DEBNA's).
Oh, and just to tie this into another thread...much of the VAX 6000
series console firmware was written in BLISS-32 :-) BLISS could do
JSB/RSB subroutine linkage, which was useful, since the firmware had
only 1 Kbyte of RAM available for variables and stack. There were also
some truly amazing macros that generated the parse tables for the
command language.
Good luck!
Gary