From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756164AbZIUNLj (ORCPT ); Mon, 21 Sep 2009 09:11:39 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1756081AbZIUNLh (ORCPT ); Mon, 21 Sep 2009 09:11:37 -0400 Received: from cantor.suse.de ([195.135.220.2]:56399 "EHLO mx1.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756070AbZIUNLg (ORCPT ); Mon, 21 Sep 2009 09:11:36 -0400 Date: Mon, 21 Sep 2009 15:11:38 +0200 From: Jan Kara To: Jens Axboe Cc: Linus Torvalds , Linux Kernel , jack@suse.cz, fengguang.wu@intel.com Subject: Re: [GIT PULL] a few writeback fixes Message-ID: <20090921131138.GE1099@duck.suse.cz> References: <20090921125620.GV23126@kernel.dk> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20090921125620.GV23126@kernel.dk> User-Agent: Mutt/1.5.17 (2007-11-01) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Mon 21-09-09 14:56:20, Jens Axboe wrote: Hi Jens, > Jan, since we're not really sure on the inode wait yet, I dropped that > patch. It's definitely something to look into, as are the congestion > bits. OK. Honza > diff --git a/mm/page-writeback.c b/mm/page-writeback.c > index 1eea4fa..6bb510b 100644 > --- a/mm/page-writeback.c > +++ b/mm/page-writeback.c > @@ -485,6 +485,7 @@ static void balance_dirty_pages(struct address_space *mapping) > unsigned long bdi_thresh; > unsigned long pages_written = 0; > unsigned long write_chunk = sync_writeback_pages(); > + unsigned long pause = 1; > > struct backing_dev_info *bdi = mapping->backing_dev_info; > > @@ -561,7 +562,15 @@ static void balance_dirty_pages(struct address_space *mapping) > if (pages_written >= write_chunk) > break; /* We've done our duty */ > > - schedule_timeout(1); > + schedule_timeout_interruptible(pause); > + > + /* > + * Increase the delay for each loop, up to our previous > + * default of taking a 100ms nap. > + */ > + pause <<= 1; > + if (pause < HZ / 10) > + pause = HZ / 10; Isn't this check the other way around? > } > > if (bdi_nr_reclaimable + bdi_nr_writeback < bdi_thresh && Honza -- Jan Kara SUSE Labs, CR