1 Mar
2009
1 Mar
'09
11:52 p.m.
On 26 Feb 2009 at 7:38, Les Hildenbrandt wrote:
I am sorry I am so slow here. Which is the problem?
The value or the sequence?
ISTR it was SP == call target address that precipitated the problem.
Darn it, are you going to make me go digging through over 20 years of
notes to find the thing?
The Z80 thing is an interesting point. Before I
started running CP/M
on a V20, I allways ran on a 8085. I would get annoyed when cp/m
software required a z80, because it was not a z80 operating system. I
later did build a few z80 systems, a laptop and a 20mhz sbc system,
but all of the coding I did for them was 8080, except for using the 16
bit io address features of the z80.
ZCPR and some compression (ZIP-type) utilities are the ones that made
me wish I had a Z80 in my system and not an 8085. I've got some
NSC800 CPUs--I figure that I might be able to shoehorn one into an
8085 system with a little glue (as long as the system doesn't use RIM
and SIM). Or maybe not, but it's close.
I assumed the rabbit was a z180, you learn something
new every day. I
have done a few designes with z180's, but I cant even rememer what
language tools I used at this point.
The Rabbit Semiconductor R2000, R3000, R4000 series (still being
made) are Z80/Z180 "sort-of". Lots of "rarely used" opcodes (e.g.
DAA, IN, OUT, etc.) are replaced by other instructions. See:
http://www.rabbit.com/documentation/docs/manuals/Rabbit4000/Instructio
nReference/
How much patching do you think it would take to get CP/M to run on an
R3000?
--Chuck