From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S269132AbUINCTn (ORCPT ); Mon, 13 Sep 2004 22:19:43 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S269121AbUINCRm (ORCPT ); Mon, 13 Sep 2004 22:17:42 -0400 Received: from holomorphy.com ([207.189.100.168]:656 "EHLO holomorphy.com") by vger.kernel.org with ESMTP id S269132AbUINCPD (ORCPT ); Mon, 13 Sep 2004 22:15:03 -0400 Date: Mon, 13 Sep 2004 19:14:56 -0700 From: William Lee Irwin III To: Andrey Savochkin Cc: Linus Torvalds , Andrew Morton , linux-kernel@vger.kernel.org Subject: Re: [PATCH] fix for fsync ignoring writing errors Message-ID: <20040914021456.GL9106@holomorphy.com> References: <20040913161255.A18665@castle.nmd.msu.ru> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20040913161255.A18665@castle.nmd.msu.ru> Organization: The Domain of Holomorphy User-Agent: Mutt/1.5.6+20040722i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Mon, Sep 13, 2004 at 04:12:55PM +0400, Andrey Savochkin wrote: > AFACS, currently metadata writing errors are ignored and not returned from > sys_fsync on ext2 and ext3 filesystems. > That is, at least ext2 and ext3. > Both ext2 and ext3 resort to sync_inode() in their ->sync_inode method, which > in turn calls ->write_inode. ->write_inode method has void type, and any IO > errors happening inside are lost. > Any objections to making ->write_inode return the error code? > Signed-off-by: Andrey Savochkin While I've not reviewed this in any detail I'm very much in favor of propagating any and all errors to fsync(), particular for the benefit of userspace applications responsible for maintaining data integrity. -- wli