From: Badalian Vyacheslav Subject: Re: deadlocks if use htb Date: Fri, 16 Jan 2009 09:51:37 +0300 Message-ID: <49702E79.6040100@bigtelecom.ru> References: <20090114063909.GA4234@ff.dom.local> <200901141417.58667.denys@visp.net.lb> <1231937404.14825.4.camel@laptop> <20090115065322.GA4190@ff.dom.local> <20090115075450.GB4190@ff.dom.local> <20090115094530.GG4190@ff.dom.local> <20090115121801.GH5461@ff.dom.local> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Jarek Poplawski , Thomas Gleixner , Denys Fedoryschenko , Peter Zijlstra , netdev@vger.kernel.org, linux-kernel@vger.kernel.org To: Chris Caputo Return-path: Received: from mail.bigtelecom.ru ([87.255.0.61]:60603 "EHLO mail.bigtelecom.ru" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1758610AbZAPGvl (ORCPT ); Fri, 16 Jan 2009 01:51:41 -0500 In-Reply-To: Sender: netdev-owner@vger.kernel.org List-ID: Chris Caputo =D0=BF=D0=B8=D1=88=D0=B5=D1=82: > On Thu, 15 Jan 2009, Jarek Poplawski wrote: > =20 >> On Thu, Jan 15, 2009 at 12:00:50PM +0000, Chris Caputo wrote: >> ... >> =20 >>> I wish I had a repro scenario in a non-production environment, so I= could=20 >>> help out further with this. If I did, I would test without nmi_wat= chdog=20 >>> while trying just 2.6.28 and Jarek's #4 patch. >>> =20 >> I think this watchdog doesn't matter too much yet. Probably it's mor= e >> about hardware (smp, maybe kind of hw timer), and high traffic vs. h= tb >> rules. >> =20 > > Per Thomas' comment in http://bugzilla.kernel.org/show_bug.cgi?id=3D1= 0944=20 > regarding broken hrtimers: > > -- > nmi_watchdog=3D1 disables the local apic timers and therefor highre= s/nohz > > Try to use nmi_watchdog=3D2 instead. > -- > > Chris > > > =20 Oh... after first deadlocks i add nmi_watchdog=3D1 to kernel and test w= ith it. Is this result what highres/nohz allways on and my message where i say what test bug which highres/nohz on/off is wrong?