Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84 SMI; site sun.uucp
Path: utzoo!watmath!clyde!burl!ulysses!allegra!mit-eddie!genrad!decvax!decwrl!sun!gnu
From: gnu@sun.uucp (John Gilmore)
Newsgroups: net.unix-wizards,net.micro.68k
Subject: Re: Pre-fetch: fixed in 68020
Message-ID: <2408@sun.uucp>
Date: Wed, 10-Jul-85 22:24:58 EDT
Article-I.D.: sun.2408
Posted: Wed Jul 10 22:24:58 1985
Date-Received: Sat, 13-Jul-85 08:13:48 EDT
References: <891@bunker.UUCP>
Distribution: net
Organization: Sun Microsystems, Inc.
Lines: 11
Xref: watmath net.unix-wizards:13791 net.micro.68k:1000

>                                         ...the last two bytes of
> memory contained an RTS (return from subroutine). When this
> instruction was executed the fault occurred because the 68K prefetches
> 2 to 4 bytes ahead of where it's executing.

This is fixed in the 68020.  The processor will only take a bus error
on a prefetch if it actually uses the instruction that it tried to
fetch.  If the memory cycle that got the bus error ends up not being used,
no bus error trap occurs.

Thanks for listening, Motorola...