Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site lsuc.UUCP
Path: utzoo!utcs!lsuc!jimomura
From: jimomura@lsuc.UUCP (Jim Omura)
Newsgroups: net.micro.amiga
Subject: Comparative OS's AmigaDOS and OS-9
Message-ID: <1105@lsuc.UUCP>
Date: Sun, 9-Feb-86 21:43:51 EST
Article-I.D.: lsuc.1105
Posted: Sun Feb  9 21:43:51 1986
Date-Received: Mon, 10-Feb-86 00:40:41 EST
Reply-To: jimomura@lsuc.UUCP (Jim Omura)
Organization: Barrister & Solicitor, Toronto
Lines: 27
Summary: Relative Addressing limits of 68k technology


     A while back someone mentioned the limit of the 68000 relative
addressing mode (+/-32K and *longer* with index registers).

     First, I think the person who posted it misunderstands the index
register displacement addressing system.  It can use the whole 32 bits
of a register as an address (granted with the real limit in practice
being the pinouts of the DIP package).  He seems to think it's limited
to +/-128 bytes.  That figure is for an additional 8 bit variable
taken from the instruction.  It's a very complex, but I suspect slow,
form of addressing.  Nevertheless, there *is* full +/-2 Gigabyte
relative addressing in a 68k system subject to the above.

     The 68020 adds a full 32 bit Long Branch set of instructions whic
should execute much faster.  In either case though, there should be
no problem with the OS-9 design philosophy.  I am guessing here, but
I should think that a top notch C compiler system should be able to
choose the appropriate addressing mode to minimize object code (and
speed) for these situations.

                                          Cheers! -- Jim O.

-- 
James Omura, Barrister & Solicitor, Toronto
ihnp4!utzoo!lsuc!jimomura
Byte Information eXchange: jimomura
(416) 652-3880