From: Jean noel Cordenner Subject: Re: [PATCH 0/6] inode version updated patches Date: Fri, 15 Jun 2007 16:28:49 +0200 Message-ID: <4672A221.2050002@bull.net> References: <1181836571.7779.42.camel@garfield> <1181854139.3883.7.camel@dyn9047017103.beaverton.ibm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Kalpak Shah , linux-ext4 , Andreas Dilger To: cmm@us.ibm.com Return-path: Received: from ecfrec.frec.bull.fr ([129.183.4.8]:43737 "EHLO ecfrec.frec.bull.fr" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751289AbXFOOeq (ORCPT ); Fri, 15 Jun 2007 10:34:46 -0400 In-Reply-To: <1181854139.3883.7.camel@dyn9047017103.beaverton.ibm.com> Sender: linux-ext4-owner@vger.kernel.org List-Id: linux-ext4.vger.kernel.org Mingming Cao a =E9crit : > On Thu, 2007-06-14 at 21:26 +0530, Kalpak Shah wrote: >> Hi, >> >> I have updated the inode version patches and below are the details: >> >> 64-bit-i_version.patch which only replaces the 32-bit i_version fiel= d in >> the generic inode with a 64-bit i_version field. The VFS-level updat= es >> of the inode version are still in the i_version_update_vfs.patch and >> this patch has been moved to the bottom of the patch series. >> >> ext4_i_version_hi_2.patch was still using ei->i_fs_version instead o= f >> inode->i_version. I have corrected this patch. >> >> Also there were whitespace problems in i_version_hi.patch. >> >> I am also attaching a ext4_no_version.patch which adds a "noversion" >> mount option to disable inode version updates. >> >> I am not sure why i_version needs to be updated in VFS since it is b= eing >> updated in ext4_mark_iloc_dirty(). >> >=20 > Yes I think we can remove the inode version update in VFS patch. So I= am > going to drop the patch 5/6. Jean Noel, if you disagree, please let = me > know. >=20 I agree, the update of the i_version in the vfs are done in simple_rename() , simple_link() , simple_unlink() , which are mainly=20 used by fs/ramfs and fs/hugetlbfs. So I think if each fs can update the= =20 i_version, that will be good. Jean noel