Tony Duell wrote:
Yes, that's correct. I must admit that when I
posted the last message
last night, I was somewhat tired and didn't feel like either pulling the
CPU out of the rack or getting the printset.
So I couldn't remember where the M8116 actually goes...
Mine is easy to get at... I don't have nearly enough computers :-(
From this, I
think it's fair to assume that the M8107 is the address path
(relocation adders,
etc) and the M8108 is the control logic. It might be
possible to kludge up a replacement for the M8108 that effectively
disables the MMU and turns the M8107 into just a SJB board, but it's not
going to be trivial.
Possibly... (bearing in mind that I don't know what I'm doing). There is
a single bit in the status register SR0 (which is on the M8108) that
enables memory management. Once I send Michael the M8108 print, it might
be practical for him to completely trace its consequences.
In particular there is a mux on the M8107 to select the physical address
from either the MMU-modified address or the raw address
based on this
bit. (Michael, this is the four 74157s beside the B finger.) Another
uses
this bit to select the two high address bits, but as I don't have a
signal index it's not immediately obvious to me from where. And of course
other parts of the M8107 may need to be driven appropriately so as not to
cause faults or other problems.
--
Kevin Schoedel
schoedel(a)kw.igs.net