Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754180AbYLGERY (ORCPT ); Sat, 6 Dec 2008 23:17:24 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1753548AbYLGERQ (ORCPT ); Sat, 6 Dec 2008 23:17:16 -0500 Received: from ipmail05.adl2.internode.on.net ([203.16.214.145]:38162 "EHLO ipmail05.adl2.internode.on.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753544AbYLGERP (ORCPT ); Sat, 6 Dec 2008 23:17:15 -0500 X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: ApoEABvcOkl5LJfT/2dsb2JhbADMXIMF X-IronPort-AV: E=Sophos;i="4.33,727,1220193000"; d="scan'208";a="270850768" Date: Sun, 7 Dec 2008 15:17:11 +1100 From: Dave Chinner To: Mikulas Patocka Cc: Andi Kleen , linux-kernel@vger.kernel.org, xfs@oss.sgi.com, Alasdair G Kergon , Andi Kleen , Milan Broz Subject: Re: Device loses barrier support (was: Fixed patch for simple barriers.) Message-ID: <20081207041711.GA22525@disturbed> Mail-Followup-To: Mikulas Patocka , Andi Kleen , linux-kernel@vger.kernel.org, xfs@oss.sgi.com, Alasdair G Kergon , Andi Kleen , Milan Broz References: <20081204100050.GN6703@one.firstfloor.org> <20081204142015.GQ6703@one.firstfloor.org> <20081204145810.GR6703@one.firstfloor.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.18 (2008-05-17) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 960 Lines: 25 On Thu, Dec 04, 2008 at 11:45:44AM -0500, Mikulas Patocka wrote: > > The write block with barrier bit varies, jbd/gfs2 do it synchronously > > too and xfs does it asynchronously (with io done callbacks), but > > And how does xfs preserve write ordering, if the barrier asynchronously > fails with -EOPNOTSUPP and there are other writes submitted after the > barrier? Doesn't matter. XFS executes journal state changes during the I/O completion callbacks a layer above this resubmit code. Hence, if the I/O is resubmitted before the completion callbacks are run, it just appears that the I/O has taken longer than expected and the state change is delayed.... Cheers, Dave. -- Dave Chinner david@fromorbit.com -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/