Re: index architecture, The Design of Design

According to EricP  <ThatWouldBeTelling@thevillage.com>:

John Levine wrote:

Some earlier IBM
model (don't remember which one) had the property that instructions
were somewhat like microcode, and some undocumented combinations of
bits would do useful things.

 
I wonder if that was the way that the 704 OR'ed the index registers.

Having three 1-bit register select fields saves a decoder for
index register specifier.
>
This feature looks like it was was just a consequence of using
a wired-OR bus and skipping the decoder on the index field.
For completeness they documented what happens if one enables multiple
index registers at once - that it OR's (as opposed to burn out).

That's what I was thinking, but as I said, I have never found anything
to say whether it was deliberate ("it's free, maybe someone will find
a way to use it") or not ("hey, boss, I found this when writing the
diagnostics.")

For that matter, the 704 was mostly sign-magnitude, but indexing did a
two's complement subtract of the index register(s) from the address in
the instruction. I've never found an explanation of why they
subtracted or why it was a two's comp subtraction.

Again, there are plenty of guesses like they thought it would let you
use the index register both as an index and a counter down to zero,
but that's just a guess, nothing documented. In fact, Fortran stored
its arrays backward to avoid having to negate values before using them
as an index, which suggests that whatever they expected didn't happen.

Since it was 70 years ago, everyone involved has since died so we're
stuck with the sparse material they left.


--
Regards,
John Levine, johnl@taugh.com, Primary Perpetrator of "The Internet for Dummies",
Please consider the environment before reading this e-mail. https://jl.ly