From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756942Ab2I1IEg (ORCPT ); Fri, 28 Sep 2012 04:04:36 -0400 Received: from cn.fujitsu.com ([222.73.24.84]:22333 "EHLO song.cn.fujitsu.com" rhost-flags-OK-FAIL-OK-OK) by vger.kernel.org with ESMTP id S1752619Ab2I1IEd (ORCPT ); Fri, 28 Sep 2012 04:04:33 -0400 X-IronPort-AV: E=Sophos;i="4.80,501,1344182400"; d="scan'208";a="5933383" Message-ID: <50655A98.3090507@cn.fujitsu.com> Date: Fri, 28 Sep 2012 16:06:48 +0800 From: Lai Jiangshan User-Agent: Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.2.9) Gecko/20100921 Fedora/3.1.4-1.fc14 Thunderbird/3.1.4 MIME-Version: 1.0 To: Tejun Heo CC: linux-kernel@vger.kernel.org Subject: Re: [RFC PATCH 11/12] workqueue: add WQ_CPU_INTENSIVE to system_long_wq References: <1348680043-5077-1-git-send-email-laijs@cn.fujitsu.com> <1348680043-5077-12-git-send-email-laijs@cn.fujitsu.com> <20120926183833.GI12544@google.com> In-Reply-To: <20120926183833.GI12544@google.com> X-MIMETrack: Itemize by SMTP Server on mailserver/fnst(Release 8.5.3|September 15, 2011) at 2012/09/28 16:04:42, Serialize by Router on mailserver/fnst(Release 8.5.3|September 15, 2011) at 2012/09/28 16:04:42, Serialize complete at 2012/09/28 16:04:42 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=UTF-8 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi, Tejun On 09/27/2012 02:38 AM, Tejun Heo wrote: > On Thu, Sep 27, 2012 at 01:20:42AM +0800, Lai Jiangshan wrote: >> works in system_long_wq will be running long. >> add WQ_CPU_INTENSIVE to system_long_wq to avoid these kinds of works occupy >> the running wokers which delay the normal works. >> >> if system_long_wq is designed for only sleep-long works, not running-long works, >> this patch makes no sense. > > There "long" doesn't mean it's gonna consume a lot of CPU cycles, so > it should contribute to concurrency management. The plan is to remove > system_long_wq once we got rid of flushing of system_wq via > flush_scheduled_work() and directly through flush_workqueue(). I > think we're pretty close, so let's not add more usage of it. > OK. But does we need a stand-alone workqueue for work_on_cpu() as it is original introduced? (2d3854a3) Thanks, Lai