Re: architecture, The Design of Design

John Levine <johnl@taugh.com> writes:

According to Tim Rentsch  <tr.17687@z991.linuxsc.com>:
>

John Levine <johnl@taugh.com> writes:
>

According to MitchAlsup1 <mitchalsup@aol.com>:
>

For a modern ISA, the architect should specify that various bits
of the general format "must be zero"* when those bits are not used
in the instruction.

>
That was another innovation of the 360.  It specifically said that
unused bits (of which there were a few) and unused unstructions (of
which there were a lot) were reserved.  The unused bits had to be
zero, and the instructions all trapped.

>
I would describe this not so much as an innovation but just as
applying a lesson learned from earlier experience.

>
Well, yes, but another 360 innovation was the whole idea of computer
architecture, as well as the term.  It was the first time that the
programmer's view of the computer was described independently of any
implementation.

I don't buy it.  An architecture is just a description of system
behavior, and surely there were descriptions of system behavior
before System/360.  Even in the 1950s companies must have changed
implementations of a given model while still conforming to its
earlier description.  As for the word architecture, it seems like
an obvious and natural word choice, given the hundreds (or more)
of years of experience with blueprints and buildings.

The Forbidden City in China was designed and built in a very
short time (less than 15 years) in the early 1400s, and is
still there today.  The intellectual history of architecture
is well established;  it doesn't seem like any great leap to
use the word "architecture" for something that is very much
like a blueprint.

I grant you that the idea of having a single architecture for a
line of computers covering a large range of performance was a new
idea, and a revolutionary one.  Certainly IBM deserves credit for
that.  Furthermore that one idea is responsible for much if not
most of the success of System/360, and well worth recognizing as
such.  However that idea is much more than just the notion of
describing system behavior.

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.
There were three of them, numbered 1, 2, and 4, so if your index field
was 5, it OR'ed registers 1 and 4.  It subtracted the index (or the
OR'ed combination of indexes) from the base address, so it would have
taken some really tricky programming to make use of that.  But someone
must have since they documented it and it continued to work on the
709, 7090, and 7094 until they provided 7 index registers and a mode
bit to switch between the old OR and the new 7 registers.

My memory (fuzzy and unreliable though it may be) is that the
property I mentioned had nothing to do with index registers, but
rather was about what operation (or combination of operations)
would be performed.  Apparently there was very little encoding of
the "opcode" bits, so undocumented combinations of bits would
have a different, and sometimes useful, effect.  Like I said my
memory is less than 100% reliable so feel to apply any number of
grains of salt.

I have never found anything that says whether it was deliberate or an
accident of the 704's implementation, and I have looked pretty hard.

Another dim memory from ages past is that the choice in early
versions of FORTRAN to limit arrays to three dimensions was due
to an early IBM model having three index registers.  Probably an
interested person could track that down if they wanted to, but
for myself I am content to let the question fade into the mists
of time.