From owner-freebsd-hackers Wed Aug 5 13:42:51 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id NAA21071 for freebsd-hackers-outgoing; Wed, 5 Aug 1998 13:42:51 -0700 (PDT) (envelope-from owner-freebsd-hackers@FreeBSD.ORG) Received: from echonyc.com (echonyc.com [198.67.15.2]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id NAA21066 for ; Wed, 5 Aug 1998 13:42:49 -0700 (PDT) (envelope-from benedict@echonyc.com) Received: from localhost (benedict@localhost) by echonyc.com (8.8.7/8.8.7) with SMTP id QAA11473; Wed, 5 Aug 1998 16:42:21 -0400 (EDT) Date: Wed, 5 Aug 1998 16:42:21 -0400 (EDT) From: Snob Art Genre Reply-To: ben@rosengart.com To: Stefan Eggers cc: freebsd-hackers@FreeBSD.ORG Subject: Re: executables over NFS In-Reply-To: <199808052022.WAA10737@semyam.dinoco.de> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG On Wed, 5 Aug 1998, Stefan Eggers wrote: > > Why would the [NFS executable] file be paged out to local swap? > > Being clean, wouldn't the pages be discarded once selected by the > > page-replacement algorithm? > > It seems - to me - to be easier to just get a few selected blocks for > which one knows the block numbers already (i.e. swap space) than to go > through the vnode, have to do the translation from offset in file to > block number and then get the blocks. The 4.4 book says that Mach 2.0 used the vnode pager as the swap pager, but the filesystem had not been able to deliver enough bandwidth. It doesn't specify whether the problems were with write bandwidth or read or both. The reason I asked in the first place was that it seemed odd to me to allow pages to be discarded if there's swap available and the backing store is significantly slower than local disk I/O (be it swap or filesystem). Ben "You have your mind on computers, it seems." To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message