From mboxrd@z Thu Jan 1 00:00:00 1970 From: Wu Fengguang Subject: Re: [PATCH 1/2] writeback: Improve busyloop prevention Date: Thu, 20 Oct 2011 21:39:38 +0800 Message-ID: <20111020133938.GA18058@localhost> References: <20111013142638.GB6938@localhost> <20111013143939.GA9691@localhost> <20111013201835.GD27363@quack.suse.cz> <20111014160047.GA13330@localhost> <20111014162807.GA4617@localhost> <20111018005128.GI4528@quack.suse.cz> <20111018143504.GA17818@localhost> <20111019115630.GA22266@quack.suse.cz> <20111020120909.GA8193@localhost> <20111020123300.GA12317@localhost> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: "linux-fsdevel@vger.kernel.org" , Christoph Hellwig , Dave Chinner To: Jan Kara Return-path: Received: from mga14.intel.com ([143.182.124.37]:53367 "EHLO mga14.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751122Ab1JTNjn (ORCPT ); Thu, 20 Oct 2011 09:39:43 -0400 Content-Disposition: inline In-Reply-To: <20111020123300.GA12317@localhost> Sender: linux-fsdevel-owner@vger.kernel.org List-ID: On Thu, Oct 20, 2011 at 08:33:00PM +0800, Wu Fengguang wrote: > On Thu, Oct 20, 2011 at 08:09:09PM +0800, Wu Fengguang wrote: > > Jan, > > > > I tried the below combined patch over the ioless one, and find some > > minor regressions. I studied the thresh=1G/ext3-1dd case in particular > > and find that nr_writeback and the iostat avgrq-sz drops from time to time. > > > > I'll try to bisect the changeset. This is interesting, the culprit is found to be patch 1, which is simply if (work->for_kupdate) { oldest_jif = jiffies - msecs_to_jiffies(dirty_expire_interval * 10); - work->older_than_this = &oldest_jif; - } + } else if (work->for_background) + oldest_jif = jiffies; Thanks, Fengguang