From owner-cvs-all Tue Mar 13 17:42:22 2001 Delivered-To: cvs-all@freebsd.org Received: from khavrinen.lcs.mit.edu (khavrinen.lcs.mit.edu [18.24.4.193]) by hub.freebsd.org (Postfix) with ESMTP id 51EBF37B719; Tue, 13 Mar 2001 17:42:10 -0800 (PST) (envelope-from wollman@khavrinen.lcs.mit.edu) Received: (from wollman@localhost) by khavrinen.lcs.mit.edu (8.9.3/8.9.3) id UAA42549; Tue, 13 Mar 2001 20:42:00 -0500 (EST) (envelope-from wollman) Date: Tue, 13 Mar 2001 20:42:00 -0500 (EST) From: Garrett Wollman Message-Id: <200103140142.UAA42549@khavrinen.lcs.mit.edu> To: "Andrew Reilly" Cc: cvs-all@FreeBSD.org, cvs-committers@FreeBSD.org Subject: Re: Core's function (was: The Project and onward [was: Re: cvscommit: src/sys/netinet ip_output.c]) In-Reply-To: <20010314104739.A50356@gurney.reilly.home> References: <20010313121002.F59348@wantadilla.lemis.com> <3794.984471257@critter> <20010313104930.C60817@daemon.ninth-circle.org> <20010313133108S.jkh@osd.bsdi.com> <3AAEA353.B31800B5@originative.co.uk> <20010314104739.A50356@gurney.reilly.home> Sender: owner-cvs-all@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG < said: > I don't think that anyone is saying that anything is "too hard". > Take the SMP-NG effort as an example. To my eyes, that project > has some seriously _hard_ things to achieve, and yet there is > progress. Not only that, but SMPng actually brings a significant body of interesting research close to hand. For example, no commercial (in the sense of non-toy) operating system has made a serious stab at doing non-blocking synchronization. FreeBSD thus far hasn't done much in that arena either, but the restructuring being done in the SMPng effort makes FreeBSD a reasonable platform for someone who wanted to work on that. (There's probably at least a good S.M. thesis to be had in that direction, if not a Ph.D.) Indeed, even the tradeoffs between subsystem-level and object-level locking have yet to be fully explored. It would be a Really Good Thing if some of the FreeBSD people and some of the BSDi people got together and published a paper on the methodology if SMPng, if for no other reason than to raise the academic stature of the project and get people interested in working on the next step. (I have a bit set that IA64 would be particularly interesting in this regard; specifically, that the data-speculation features could be used to implement DCAS.) -GAWollman To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message