From: =?EUC-KR?B?yKu9xSBzaGluIGhvbmc=?= Subject: a suggestion for ext2_release_file() Date: Fri, 15 May 2009 12:16:43 +0900 Message-ID: <2014bcab0905142016k7626acd1m187afb71f9ecc440@mail.gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit To: linux-ext4@vger.kernel.org Return-path: Received: from yw-out-2324.google.com ([74.125.46.29]:12182 "EHLO yw-out-2324.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753100AbZEODQn (ORCPT ); Thu, 14 May 2009 23:16:43 -0400 Received: by yw-out-2324.google.com with SMTP id 5so969748ywb.1 for ; Thu, 14 May 2009 20:16:43 -0700 (PDT) Sender: linux-ext4-owner@vger.kernel.org List-ID: Hi. I have a suggestion while I read ext2 code in recent Linux. mark_files_ro() function in fs/super.c turns off FMODE_WRITE of every file in a super block for the case of remounting. However, an inode whose file's mode is FMODE_WRITE may hold reservations. I think these reservations lost their chance to be discarded since FMODE_WRITE is off. So, I think ext2_release_file rather check EXT2_I(inode)->i_block_alloc_info->rsv_window_node. Could you check my suggestion is valid or not please? Sincerely Shin Hong