Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!swrinde!elroy.jpl.nasa.gov!jarthur!nntp-server.caltech.edu!toddpw
From: toddpw@nntp-server.caltech.edu (Todd P. Whitesel)
Newsgroups: comp.sys.apple2
Subject: Re: sideaffect of MVN/MVP instruction
Message-ID: <1991May30.025443.428@nntp-server.caltech.edu>
Date: 30 May 91 02:54:43 GMT
References: <3644@kluge.fiu.edu>
Organization: California Institute of Technology, Pasadena
Lines: 23

acmfiu@serss0.fiu.edu (ACMFIU) writes:

>someone want to explain to me why the MVN/MVP instructions don't restore
>the data bank register? (yes, i'm moving data in different banks).

Because. No explanation was given in the 65816 data sheet, although I
suppose it could be argued that the data bank is set to the destination data
so you can use absolute addressing on it right away...

I think the real reason is that there only one temporary bank latch, so the
other one ended up being the data bank register. What's stupid is that there
"should" be a way to do it without modifying the data bank register. It has
been speculated that the MVP/MVN instructions were originally different and
somewhat useless instructions, and that Mensch hacked in the move code at
the last minute -- this would easily explain the 7 cycle/byte timing (the
instruction is fetched repeatedly for each byte copied -- blech!) and the
side effects. Besides, it took them long enough to get the original 65816
ready (the GS ROM 00 was tested with 65802's because they were all that
was available -- evidence of this is buried in the softswitch list in the
firmware ref).

Todd Whitesel
toddpw @ tybalt.caltech.edu