Newsgroups: comp.sys.transputer
From: bailey@fix.fenris.com (Kirk Bailey)
Subject: CJ comments
Organization: Fenris Information Exchange, Corvallis, Ore.
Date: 5 Apr 1994 14:11:39 -0700
Message-ID: <2nsk6b$fh8@fenris.com>


I'm not sure about the original design reason for this although I recall
hearing that it was somehow related to the code generation strategy of an
early OCCAM backend.  Perhaps someone who worked at INMOS at the time could
comment?

	I STRONGLY agree that the instruction operates backwards with respect
to the A register and zero.  In both writing compiler back ends, and in
assembly language programming, I have often thought quite unkind things of
those who did that part of the instruction set design!  I rather like the
instruction set in general, but that is a pretty glaring wart.  When doing
assembly code I often seem to run into the situation where I would like to
carry three temp values into a code segment, but can't, Grrr!  Of course,
now days only fossils worry about how fast their code executes so perhaps
someone should just shoot me and get it over with...

	Cheers,
		Kirk Bailey


