From owner-freebsd-hackers Fri Dec 8 22:49:41 2000 From owner-freebsd-hackers@FreeBSD.ORG Fri Dec 8 22:49:38 2000 Return-Path: Delivered-To: freebsd-hackers@freebsd.org Received: from mta5.snfc21.pbi.net (mta5.snfc21.pbi.net [206.13.28.241]) by hub.freebsd.org (Postfix) with ESMTP id E01D637B400 for ; Fri, 8 Dec 2000 22:49:38 -0800 (PST) Received: from alumni.caltech.edu ([63.201.176.178]) by mta5.snfc21.pbi.net (Sun Internet Mail Server sims.3.5.2000.01.05.12.18.p9) with ESMTP id <0G5A008AEFUJMD@mta5.snfc21.pbi.net> for hackers@FreeBSD.ORG; Fri, 8 Dec 2000 22:39:56 -0800 (PST) Date: Fri, 08 Dec 2000 22:42:25 -0800 From: Dan Kegel Subject: Re: kqueue microbenchmark results Sender: dank@mta5.snfc21.pbi.net To: Wes Peters Cc: "Daniel C. Sobral" , hackers@FreeBSD.ORG Reply-To: dank@alumni.caltech.edu Message-id: <3A31D451.CD333721@alumni.caltech.edu> MIME-version: 1.0 X-Mailer: Mozilla 4.73 [en] (X11; U; Linux 2.2.14-5.0 i686) Content-type: text/plain; charset=us-ascii Content-transfer-encoding: 7bit X-Accept-Language: en References: <20001024225637.A54554@prism.flugsvamp.com> <39F6655A.353FD236@alumni.caltech.edu> <20001025115457.X28123@fw.wintelcom.net> <20001025170117.C87091@prism.flugsvamp.com> <20001207154925.A25785@walton.maths.tcd.ie> <3A30E115.CF7C76E8@newsguy.com> <3A31120E.3536F07D@alumni.caltech.edu> <3A31CBCF.B8E21E1B@softweyr.com> Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Wes Peters wrote: > Dan Kegel wrote: > > "Daniel C. Sobral" wrote: > > > Why is it that I get the feeling more and more nowadays that Linus is > > > suffering from a worsening case of NIH when it comes to things > > > originated on BSD? > > > > Don't jump to conclusions. He's honestly trying to > > understand what the optimal interface would be. > > Let him catch up. Help him understand the requirements > > which motivated the kqueue design and why his proposed > > system call does not meet them. > > > > His role right now is to keep the kernel as simple as possible. > > So the major advancements of pushing file servers and web servers into > the kernel fit into this role how? Regardless of whether those complicate the kernel - and I suspect khttpd and tux don't complicate the kernel much - what Linus is doing here is different: he's doing a reductionist analysis of what it takes to do poll() right. I've done the same thing before, and yes, the people whose favorite interface I seemed to be ignoring were pissed off. But it was the only way for me to understand the true requirements. In the end, I usually add back part of the stuff I initially stripped out, once I understood what it was for. That said, I like kqueue, and I don't like the interface Linus proposed. But I'm still not quite sure how to demonstrate that his interface won't do the job. (Wish I had time.) - Dan To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message