home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.arch
- Path: sparky!uunet!ukma!darwin.sura.net!sgiblab!sgigate!odin!nubbins.wpd.sgi.com!martyi
- From: martyi@nubbins.wpd.sgi.com (Marty Itzkowitz)
- Subject: Re: MINIMUM instruction set
- Message-ID: <1992Nov13.171253.14060@odin.corp.sgi.com>
- Sender: news@odin.corp.sgi.com (Net News)
- Nntp-Posting-Host: nubbins.wpd.sgi.com
- Reply-To: martyi@sgi.com
- Organization: Silicon Graphics, System Software Division
- References: <1992Nov10.235849.19192@fcom.cc.utah.edu> <pete.180021.12Nov92@sst.icl.co.uk> <1durqhINN3di@rave.larc.nasa.gov>
- Date: Fri, 13 Nov 1992 17:12:53 GMT
- Lines: 17
-
- >Well, it's certainly practical. The PDP-8, for example, had a total of
- >seven instructions, and while it wasn't any speed demon and wasn't the
- >best machine to program that I can think of, it worked and plenty of them
- >were sold.
-
- The PDP-8 had seven OPCODES, not seven instructions. Two of them,
- OPR and IOT were microcoded, and so it had many more instructions than
- that. By the above logic all RISC machines have only one instruction because
- the instructions are, after all, just microcoded words.
-
- As a humorous aside, judicious use of the OPR microcde on some of the PDP-8
- machines, could even generate a CISCy "set accumulator to 2" instruction =
- CLA CLL CML RTL
- This worked on a PDP8/S in run mode, but, oddly enough, not in single step. I
- never tried it on the other PDP-8 flavors.
-
- Marty Itzkowitz
-