From owner-freebsd-stable@FreeBSD.ORG Sat Sep 23 23:50:05 2006 Return-Path: X-Original-To: freebsd-stable@freebsd.org Delivered-To: freebsd-stable@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 107FE16A407 for ; Sat, 23 Sep 2006 23:50:05 +0000 (UTC) (envelope-from ronald-freebsd8@klop.yi.org) Received: from smtp-out2.tiscali.nl (smtp-out2.tiscali.nl [195.241.79.177]) by mx1.FreeBSD.org (Postfix) with ESMTP id 9249843D49 for ; Sat, 23 Sep 2006 23:50:04 +0000 (GMT) (envelope-from ronald-freebsd8@klop.yi.org) Received: from [82.171.39.195] (helo=guido.klop.ws) by smtp-out2.tiscali.nl with smtp (Tiscali http://www.tiscali.nl) id 1GRHFz-000083-Fq for ; Sun, 24 Sep 2006 01:50:03 +0200 Received: (qmail 23272 invoked from network); 23 Sep 2006 23:50:01 -0000 Received: from localhost.thuis.klop.ws (HELO guido.klop.ws) (127.0.0.1) by localhost.thuis.klop.ws with SMTP; 23 Sep 2006 23:50:01 -0000 Date: Sun, 24 Sep 2006 01:50:00 +0200 To: =?utf-8?Q?V=C3=A1clav_Haisman?= , freebsd-stable@freebsd.org From: "Ronald Klop" Content-Type: text/plain; format=flowed; delsp=yes; charset=utf-8 MIME-Version: 1.0 References: <45158807.8090801@sh.cvut.cz> Content-Transfer-Encoding: 8bit Message-ID: In-Reply-To: <45158807.8090801@sh.cvut.cz> User-Agent: Opera Mail/9.01 (FreeBSD) Cc: Subject: Re: (Wrong) CPU utilization reported by top X-BeenThere: freebsd-stable@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Production branch of FreeBSD source code List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 23 Sep 2006 23:50:05 -0000 On Sat, 23 Sep 2006 21:16:23 +0200, Václav Haisman wrote: > Hi, > I have noticed that top reports strange values for CPU utilization. The > problem is that while it reports this for overall utilization... > > CPU states: 40.8% user, 0.0% nice, 3.7% system, 4.5% interrupt, 50.9% > idle > > ...it doesn't show a single process that would have over 1% of WCPU or > CPU! > > PID USERNAME THR PRI NICE SIZE RES STATE TIME WCPU > COMMAND > 17980 xxxxx 1 96 0 5568K 3536K *Giant 213:05 0.00% > silc > 88370 xxxxxxxxx 1 96 0 5584K 3592K RUN 211:07 0.00% > silc > 93031 xxxxxx 5 20 0 58288K 40712K kserel 164:25 0.00% > vlc > 20223 xxx 5 20 0 238M 29572K kserel 36:04 0.00% > java > 17539 xxxxx 1 96 0 10692K 6032K select 26:51 0.00% > centericq > 30491 xxxxxx 1 96 0 7076K 4708K select 26:48 0.00% > ventrilo_srv > > I am quite positive it is the vlc process that eats most of the reported > 40% > of CPU because if I turn it off the user time goes down bellow 10%. But > despite that top shows zero in its row. > > The box is uniprocessor x86 FreeBSD 6.1. I see the same and it is always with threaded applications (libpthread). If you use libthr (see libmap.conf) it shows a lot better. Ronald. -- Ronald Klop Amsterdam, The Netherlands