From owner-freebsd-hackers Wed Feb 20 2:57:12 2002 Delivered-To: freebsd-hackers@freebsd.org Received: from rly-ip02.mx.aol.com (rly-ip02.mx.aol.com [152.163.225.160]) by hub.freebsd.org (Postfix) with ESMTP id E478C37B405 for ; Wed, 20 Feb 2002 02:57:05 -0800 (PST) Received: from logs-wc.proxy.aol.com (logs-wc.proxy.aol.com [205.188.193.5]) by rly-ip02.mx.aol.com (8.8.8/8.8.8/AOL-5.0.0) with ESMTP id FAA04154 for ; Wed, 20 Feb 2002 05:56:13 -0500 (EST) Received: from blah (AC80D3ED.ipt.aol.com [172.128.211.237]) by logs-wc.proxy.aol.com (8.10.0/8.10.0) with SMTP id g1KApOu346887 for ; Wed, 20 Feb 2002 05:51:24 -0500 (EST) Message-Id: <200202201051.g1KApOu346887@logs-wc.proxy.aol.com> Date: Wed, 20 Feb 2002 11:21:51 +0100 To: hackers@freebsd.org From: eberkut Subject: Re: Task activation delays: FreeBSD versus Linux? Organization: CNS / Minithins X-Mailer: Opera 5.11 build 904b X-Priority: 3 (Normal) Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" X-Apparently-From: SinkSuffering@aol.com Content-Transfer-Encoding: 8bit Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG 20/02/02 10:57:50, Terry Lambert a écrit: >What is it? > >You left off the URL. Looking on the site, it's not >posted to the front, and the search function does not >locate the article. > >To answer your questions, you'll have to provide more >information. A Linux patch number doesn't cut it. > >I expect that this is talking about exec pre-forking; if >so, then yes, it can yield some minor speedup, but you >should not expect much in common applications which fork >and do not exec (e.g. Apache). There are a number of >micro-benchmarks that could make this approach look good, >though... No, the andrew morton's low latency patch (and the robert love's preempt patch) try to make the kernel himself preemptible to reduce latency. There is two different approaches : "Robert Love: The model we use is to allow the kernel to be preempted at any time when it is not locked. Under this design, when an event occurs that causes a higher priority task to become runnable, the system will preempt the current task and run the higher priority task." http://kerneltrap.org/node.php?id=1 http://www.tech9.net/rml/linux/ "Andrew Morton: The approach taken by these patches is basically cooperative multitasking. The developer identifies sections of long-running kernel code and changes them so that they will yield the CPU to another task if the scheduler says that it's time to do that. Most of the complexity here is in being able to back out of any locking before yielding, and in cleanly reacquiring locking state when the interrupted task resumes." http://kerneltrap.org/node.php?id=10 http://www.zip.com.au/~akpm/linux/schedlat.html So I think that the final question is : Is the FreeBSD kernel preemptible ? Does he compete on latency ? Is it yet another gruik hack from linux ? ;) --eberkut ex diffinientium cognitione diffiniti resultat cognitio . Prelude : http://prelude.sf.net . CNS : http://minithins.net . SpeKa : http://www.speka.net To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message