X-Original-To: pgsql-performance-postgresql.org@localhost.postgresql.org Received: from localhost (unknown [200.46.204.2]) by svr1.postgresql.org (Postfix) with ESMTP id 3914BD1D25B for ; Thu, 15 Apr 2004 17:04:31 -0300 (ADT) Received: from svr1.postgresql.org ([200.46.204.71]) by localhost (neptune.hub.org [200.46.204.2]) (amavisd-new, port 10024) with ESMTP id 88416-09 for ; Thu, 15 Apr 2004 17:04:30 -0300 (ADT) Received: from mailout11.sul.t-online.com (mailout11.sul.t-online.com [194.25.134.85]) by svr1.postgresql.org (Postfix) with ESMTP id 03E24D1D083 for ; Thu, 15 Apr 2004 17:04:24 -0300 (ADT) Received: from fwd03.aul.t-online.de by mailout11.sul.t-online.com with smtp id 1BED5M-00010G-06; Thu, 15 Apr 2004 22:03:44 +0200 Received: from aeccom.com (ZknwbMZrZe57WMjZzTNOinFhX4rdoiy+yeQm4pJ8W4cxP3xVE-aZ8n@[217.231.158.140]) by fwd03.sul.t-online.com with esmtp id 1BED5I-1IbFnE0; Thu, 15 Apr 2004 22:03:40 +0200 Message-ID: <407EEA9B.7010807@aeccom.com> Date: Thu, 15 Apr 2004 22:03:39 +0200 From: Dirk.Lutzebaeck@t-online.de (Dirk Lutzebaeck) Reply-To: lutzeb@aeccom.com User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.4.1) Gecko/20031030 X-Accept-Language: en-us, en MIME-Version: 1.0 To: Tom Lane Cc: Joe Conway , pgsql-performance@postgresql.org Subject: Re: Toooo many context switches (maybe SLES8?) References: <407E7B99.3050306@aeccom.com> <407EB1E5.4030600@joeconway.com> <407EB86B.6080801@aeccom.com> <407EBA89.6080007@joeconway.com> <20498.1082057841@sss.pgh.pa.us> In-Reply-To: <20498.1082057841@sss.pgh.pa.us> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Seen: false X-ID: ZknwbMZrZe57WMjZzTNOinFhX4rdoiy+yeQm4pJ8W4cxP3xVE-aZ8n X-Virus-Scanned: by amavisd-new at postgresql.org X-Archive-Number: 200404/298 X-Sequence-Number: 6598 Could this be related to the O(1) scheduler backpatches from 2.6 to 2.4 kernel on newer 2.4er distros (RedHat, SuSE)? Tom Lane wrote: >Joe Conway writes: > > >>>Improve spinlock code for recent x86 processors: insert a PAUSE >>>instruction in the s_lock() wait loop, and use test before test-and-set >>>in TAS() macro to avoid unnecessary bus traffic. Patch from Manfred >>>Spraul, reworked a bit by Tom. >>> >>> > > > >>I thought this had been committed to the 7.4 stable branch as well, but >>it appears not. >> >> > >I am currently chasing what seems to be the same issue: massive context >swapping on a dual Xeon system. I tried back-patching the above-mentioned >patch ... it helps a little but by no means solves the problem ... > > regards, tom lane > > >