From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from relay.sgi.com (relay1.corp.sgi.com [137.38.102.111]) by oss.sgi.com (Postfix) with ESMTP id C04027F9C for ; Thu, 20 Feb 2014 17:23:13 -0600 (CST) Received: from cuda.sgi.com (cuda1.sgi.com [192.48.157.11]) by relay1.corp.sgi.com (Postfix) with ESMTP id A37748F8040 for ; Thu, 20 Feb 2014 15:23:10 -0800 (PST) Received: from ipmail06.adl2.internode.on.net (ipmail06.adl2.internode.on.net [150.101.137.129]) by cuda.sgi.com with ESMTP id liEK5Cepnlx0s6rm for ; Thu, 20 Feb 2014 15:23:08 -0800 (PST) Date: Fri, 21 Feb 2014 10:23:05 +1100 From: Dave Chinner Subject: Re: testing flink (aka linkat(..., AT_EMPTY_PATH) Message-ID: <20140220232305.GT4916@dastard> References: <20140220221245.GA32288@infradead.org> MIME-Version: 1.0 Content-Disposition: inline In-Reply-To: <20140220221245.GA32288@infradead.org> List-Id: XFS Filesystem from SGI List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: xfs-bounces@oss.sgi.com Sender: xfs-bounces@oss.sgi.com To: Christoph Hellwig Cc: xfs@oss.sgi.com On Thu, Feb 20, 2014 at 02:12:45PM -0800, Christoph Hellwig wrote: > Turns out so far we aren't really do metadata operation in xfs_io > and don't even have link/unlink. Should we starting adding these or > where else should I start looking into adding support for exercising > flink? I think we need a tool that aggregates such functionality, otherwise we just end up with a prolification of small helper utilities in xfstests/src directory. There's the new renameat2 code that would dearly love a helper function, reflink support woul be useful, etc. And for O_TMPFILE, is just doesn't make sense to duplicate what xfs_io does just so that flinkat() support can be tested. Hence I'm happy to see such metadata operations added to xfs_io from a functionality point of view. I also can't see any problems with being able to integrate them into the command structure or the source layout structure that xfs_io uses, so I'm not finding an reason why we shouldn't be adding things like this... Cheers, Dave. -- Dave Chinner david@fromorbit.com _______________________________________________ xfs mailing list xfs@oss.sgi.com http://oss.sgi.com/mailman/listinfo/xfs