From mboxrd@z Thu Jan 1 00:00:00 1970 From: Mark Fasheh Subject: Re: [-mm PATCH] ocfs2: ->fallocate() support Date: Mon, 2 Jul 2007 16:18:12 -0700 Message-ID: <20070702231812.GA15463@ca-server1.us.oracle.com> References: <20070621190143.GC17713@ca-server1.us.oracle.com> <20070630093413.GE22354@infradead.org> Reply-To: Mark Fasheh Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: linux-fsdevel@vger.kernel.org, Andrew Morton , Amit Arora , ocfs2-devel@oss.oracle.com, xfs@oss.sgi.com To: Christoph Hellwig Return-path: Content-Disposition: inline In-Reply-To: <20070630093413.GE22354@infradead.org> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: ocfs2-devel-bounces@oss.oracle.com Errors-To: ocfs2-devel-bounces@oss.oracle.com List-Id: linux-fsdevel.vger.kernel.org On Sat, Jun 30, 2007 at 10:34:13AM +0100, Christoph Hellwig wrote: > Btw, it seems like ocfs implements the xfs preallocation ioctls. What > would people thing about moving those up to work over ->fallocate so they > can be used on all filesystems that support preallocation? That sounds like a good idea. We've already got a very good match on RESVP/UNRESVP with the various ->fallocate flags. I'm not sure whether we want to do this for ALLOCSP/FREESP - as far as I've heard those are hardly used. > While the ABI is quite ugly it has a huge userbase because it's the only > existing preallocation mechanism on Linux. Yeah, that was the primary reason I chose to implement the ioctl (along with the ->fallocate patch) --Mark -- Mark Fasheh Senior Software Developer, Oracle mark.fasheh@oracle.com