From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S964789AbWFMWmf (ORCPT ); Tue, 13 Jun 2006 18:42:35 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S964786AbWFMWmf (ORCPT ); Tue, 13 Jun 2006 18:42:35 -0400 Received: from omx2-ext.sgi.com ([192.48.171.19]:7849 "EHLO omx2.sgi.com") by vger.kernel.org with ESMTP id S964789AbWFMWme (ORCPT ); Tue, 13 Jun 2006 18:42:34 -0400 Date: Wed, 14 Jun 2006 08:41:55 +1000 From: Nathan Scott To: Avi Kivity Cc: "Theodore Ts'o" , Jan Engelhardt , linux-kernel@vger.kernel.org Subject: Re: [RFC] Slimming down struct inode Message-ID: <20060614084155.C888012@wobbly.melbourne.sgi.com> References: <20060613143230.A867599@wobbly.melbourne.sgi.com> <448EC51B.6040404@argo.co.il> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <448EC51B.6040404@argo.co.il>; from avi@argo.co.il on Tue, Jun 13, 2006 at 05:00:59PM +0300 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Tue, Jun 13, 2006 at 05:00:59PM +0300, Avi Kivity wrote: > Nathan Scott wrote: > > Such a change would would indeed break XFS, in exactly the way you Oh, I should clarify - the suggestion of using sb->s_blocksize/ s_blocksize_bits was what I meant by "would break XFS". > > suggest Jan - the realtime subvolume does typically use a different > > blocksize from the data subvolume (the realtime extent size is used, > > and this can be set per-inode too), and there would now be no way to > > distinguish this preferred IO size difference. > > It can be made into an inode operation: *nod* - that'd work fine for our needs here. cheers. -- Nathan