13 Feb
2017
13 Feb
'17
11:35 a.m.
On Mon, Feb 13, 2017 at 7:30 PM, Craig Ruff <cruff at ruffspot.net> wrote:
Tony Duell wrote:
That seems possible, yes.
Having other hardware decode instructions that appear as NOPs to the main
processor is not uncommon in HP machines.
-tony
My first thought, and it's probably wrong, is
that these instrucitons
(which differ by one bit, so might be
setting/reseting something) are NOPs to the CPU, but are interpretted
by the memory mapping hardware in those
9825s that have more than 64K or RAM and ROM total.
This seems likely. According to your schematics, gate U47 detects the pattern 0701xx.
This signal feeds into the U43c flipflop, which appears to latch the state of the low 4
bits
of the MAD bus into register U42, which sets the state of the /ForceRAM (bit 3),
/ForceROM (bit 2), /DiagRd (bit 1) and ALLROM (bit 0) signals. Thus, these two
instructions appear to toggle the state of the /ForceROM signal.
If I?ve wrapped my brain around the details it appears that 070113 deasserts the
/ForceROM signal, and 070117 asserts it?