From owner-freebsd-current@FreeBSD.ORG Sun Oct 24 00:33:18 2004 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 11B7216A4CF for ; Sun, 24 Oct 2004 00:33:18 +0000 (GMT) Received: from aiolos.otenet.gr (aiolos.otenet.gr [195.170.0.23]) by mx1.FreeBSD.org (Postfix) with ESMTP id 3164643D4C for ; Sun, 24 Oct 2004 00:33:15 +0000 (GMT) (envelope-from keramida@freebsd.org) Received: from gothmog.gr (patr530-a027.otenet.gr [212.205.215.27]) i9O0X6YX001926; Sun, 24 Oct 2004 03:33:07 +0300 Received: from gothmog.gr (gothmog [127.0.0.1]) by gothmog.gr (8.13.1/8.13.1) with ESMTP id i9O0X0XD049246; Sun, 24 Oct 2004 03:33:00 +0300 (EEST) (envelope-from keramida@freebsd.org) Received: (from giorgos@localhost) by gothmog.gr (8.13.1/8.13.1/Submit) id i9O0X0mf049245; Sun, 24 Oct 2004 03:33:00 +0300 (EEST) (envelope-from keramida@freebsd.org) Date: Sun, 24 Oct 2004 03:33:00 +0300 From: Giorgos Keramidas To: cpghost@cordula.ws Message-ID: <20041024003300.GA42838@gothmog.gr> References: <20041022112947.GA96234@orion.daedalusnetworks.priv> <20041022133333.GA561@bsdbox.farid-hajji.net> <20041022142603.GA43820@orion.daedalusnetworks.priv> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20041022142603.GA43820@orion.daedalusnetworks.priv> cc: freebsd-current@freebsd.org Subject: Re: delayed process exit X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 24 Oct 2004 00:33:18 -0000 On 2004-10-22 17:26, Giorgos Keramidas wrote: > On 2004-10-22 15:33, cpghost@cordula.ws wrote: > > On Fri, Oct 22, 2004 at 02:29:47PM +0300, Giorgos Keramidas wrote: > > > Some times, in a seemingly unpredictable manner, exiting processes take > > > a lot of time to finish. I'm seeing this almost every time I exit > > > Emacs, because bash takes a long time to show me a shell prompt. > > > > This is a known issue, that has AFAICS not yet been solved (at least > > not MFC-ed in RELENG_5 as of Oct. 21st). > > > > The problem is not with the exiting process, but with the wakeup > > of its parent. There seems to be some bug in the scheduler that > > results in LONG (1 to 5 seconds) delays to wakeup processes that > > have been swapped out. Nearly every swapped out process can be > > affected by this bug. > > Ah, I see... Thanks for the detailed explanation! > > At least, now I'm not worried that something I did locally has caused > the problem. > > Do you perchance know if this affects both SCHED_ULE and SCHED_BSD? I've since updated my workstation at home to a more recent version of HEAD, but kept seeing these delays. After exchanging some email with Julian Elischer, whose help has been invaluable in understanding what might be the cause of this, I turned off vm.swap_idle_enabled which I had enabled a long time ago in my sysctl.conf. Now all seems to be back to normal. Thanks to all who have helped me find out what this could be caused by, including you too cpghost@cordula.ws :-)