Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Posting-Version: version B 2.10.1 6/24/83 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.lang.c,net.micro.68k Subject: Re: Freezing 68020 Cache Message-ID: <1704@sun.uucp> Date: Tue, 25-Sep-84 00:55:35 EDT Article-I.D.: sun.1704 Posted: Tue Sep 25 00:55:35 1984 Date-Received: Sat, 29-Sep-84 08:52:30 EDT References: <3@tikal.UUCP>, <8836@watmath.UUCP>, <235@callan.UUCP> <424@watdcsu.UUCP> <139@rlgvax.UUCP> Organization: Sun Microsystems, Inc. Lines: 11 Actually I had assumed that the passage about "emulator functions" was talking about other-machine emulators. I once wrote the basic structure of a program that would emulate the IBM 370 on a 68000 -- the slowdown was about 20:1 (i.e. it took 20 68000 instructions to fetch, decode, and execute the average 370 instruction). If you could freeze the inner loop of this into the 68020 cache you might get a much better performance ratio. Kinda like 1401 emulator mode on 370's, those dinosaurs just won't die. If anybody wants a copy of the 3-page very rough paper on the 370 emulator, send me mail.