From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756461AbdIHP2K (ORCPT ); Fri, 8 Sep 2017 11:28:10 -0400 Received: from mga02.intel.com ([134.134.136.20]:58612 "EHLO mga02.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754397AbdIHP2I (ORCPT ); Fri, 8 Sep 2017 11:28:08 -0400 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.42,362,1500966000"; d="scan'208";a="1170383416" Date: Fri, 8 Sep 2017 09:28:05 -0600 From: Ross Zwisler To: Christoph Hellwig Cc: Ross Zwisler , linux-kernel@vger.kernel.org, Jan Kara , "Darrick J. Wong" , linux-nvdimm@lists.01.org, Dave Chinner , stable@vger.kernel.org, Christoph Hellwig , linux-xfs@vger.kernel.org Subject: Re: [PATCH 1/2] xfs: always use DAX if mount option is used Message-ID: <20170908152805.GA16646@linux.intel.com> References: <20170907210832.17969-1-ross.zwisler@linux.intel.com> <20170907210832.17969-2-ross.zwisler@linux.intel.com> <20170908072027.GA27725@infradead.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20170908072027.GA27725@infradead.org> User-Agent: Mutt/1.8.3 (2017-05-23) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Fri, Sep 08, 2017 at 12:20:28AM -0700, Christoph Hellwig wrote: > On Thu, Sep 07, 2017 at 03:08:31PM -0600, Ross Zwisler wrote: > > Before support for the per-inode DAX flag was disabled the XFS the code had > > an issue where the user couldn't reliably tell whether or not DAX was being > > used to service page faults and I/O when the DAX mount option was used. In > > this case each inode within the mounted filesystem started with S_DAX set > > due to the mount option, but it could be cleared if someone touched the > > individual inode flag. > > Looks good, but can you please add a testcase to xfstests for this? Hmm...aside from looking at tracepoints, I'm not sure how to detect whether or not S_DAX is actually being used for an inode. I don't see any other xfstests that try and look at the trace buffer, but I guess that could work. Do you know of a better way to detect this test failure/success? > Reviewed-by: Christoph Hellwig Thanks for the review!