From owner-freebsd-hackers Sat Feb 10 02:51:00 1996 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id CAA02777 for hackers-outgoing; Sat, 10 Feb 1996 02:51:00 -0800 (PST) Received: from hda.com (hda.com [199.232.40.182]) by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id CAA02771 for ; Sat, 10 Feb 1996 02:50:57 -0800 (PST) Received: (from dufault@localhost) by hda.com (8.6.11/8.6.9) id FAA09574; Sat, 10 Feb 1996 05:55:17 -0500 From: Peter Dufault Message-Id: <199602101055.FAA09574@hda.com> Subject: Re: Compressed RAM/SWAP To: luigi@labinfo.iet.unipi.it (Luigi Rizzo) Date: Sat, 10 Feb 1996 05:55:17 -0500 (EST) Cc: hackers@FreeBSD.ORG In-Reply-To: <199602101043.LAA04471@labinfo.iet.unipi.it> from "Luigi Rizzo" at Feb 10, 96 11:43:51 am X-Mailer: ELM [version 2.4 PL24] Content-Type: text Sender: owner-hackers@FreeBSD.ORG Precedence: bulk > > > If there are lots of zero'd pages that aren't used that often, > > can't be eliminated, and would benefit from compression wouldn't > > it be better to look at a ZFOD (zero fill on demand) mechanism? > > But this only works if the pages are not accessed at all. The first > time they are accessed they must be filled up. Well, yes, but for reads they can all point to the same underlying page, and for writes you're changing them anyway (and if you are sure the writes have to happen before they are accessed again you don't even need to fill them). If the problems are lots of zero pages it seems that there is a better way than compression, though I guess you can argue this is just compression optimized for zero pages. I better back off this discussion as it is outside my usual work. I'll let someone who knows this stuff comment. -- Peter Dufault Real-Time Machine Control and Simulation HD Associates, Inc. Voice: 508 433 6936 dufault@hda.com Fax: 508 433 5267