<I know basically how microcode works. It's the one-cycle instructions li
<"gate data bus to address bus" or "gate data bus to ALU".
<One MOV can be 2-5 microinstructions.
First most microinstructions are bit coded... meaning that it can do
several things concurrently if the hardware can based on what bits are
set. That also means you cant output data to the data bus while trying
input data. Watch out as some of the vax MOVs are fairly complex
instructions.
<The large question I have is: How does the MOV translate to microcode?
<Does the VAX figure a microcode offset based on the opcode? Is there a t
<of entry points somewhere? How is this done?
The macroinstruction is a pointer to a table entry that containes the
address of a microsubroutine that will execute it. That means in hardware
there will be a rom or ram between the instruction register and the
external microcode address entry point (usually a operand to a microcode
jump.).
<Another thing, the VAX 780 was microcoded from the console RX01. Does an
<have a copy of that RX01? Can anyone get a disk image? What else was on
<(RT-11? P/OS?)
I believe it was RT-11 with special application code for starting and
monitoring the 780. The 730 and 750 had minimal microcode and used a
tu58 to load the remainder or overlay existing ucode to patch it.
Allison
Received on Tue May 19 1998 - 16:23:42 BST