From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S964787AbWFNKe1 (ORCPT ); Wed, 14 Jun 2006 06:34:27 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S932324AbWFNKe1 (ORCPT ); Wed, 14 Jun 2006 06:34:27 -0400 Received: from mail.clusterfs.com ([206.168.112.78]:19408 "EHLO mail.clusterfs.com") by vger.kernel.org with ESMTP id S932285AbWFNKe0 (ORCPT ); Wed, 14 Jun 2006 06:34:26 -0400 From: Nikita Danilov MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <17551.58643.704359.815153@gargle.gargle.HOWL> Date: Wed, 14 Jun 2006 14:29:39 +0400 To: Nathan Scott Cc: "Theodore Ts'o" , Jan Engelhardt , linux-kernel@vger.kernel.org Subject: Re: [RFC] Slimming down struct inode Newsgroups: gmane.linux.kernel In-Reply-To: <20060614084155.C888012@wobbly.melbourne.sgi.com> References: <20060613143230.A867599@wobbly.melbourne.sgi.com> <448EC51B.6040404@argo.co.il> <20060614084155.C888012@wobbly.melbourne.sgi.com> X-Mailer: VM 7.17 under 21.5 (patch 17) "chayote" (+CVS-20040321) XEmacs Lucid Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Nathan Scott writes: > 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. Sorry, but why this operation is needed? Generic code (in fs/*.c) doesn't use ->i_blksize at all. If XFS wants to provide per-inode st_blksize, all it has to do is to store preferred buffer size in its file system specific inode (struct xfs_inode), and use something different from generic_fillattr() as its ->i_op->getattr() callback (xfs_vn_getattr()). > > cheers. > > -- > Nathan Nikita.