From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753767AbaAaDAd (ORCPT ); Thu, 30 Jan 2014 22:00:33 -0500 Received: from imap.thunk.org ([74.207.234.97]:52921 "EHLO imap.thunk.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751623AbaAaDAb (ORCPT ); Thu, 30 Jan 2014 22:00:31 -0500 Date: Thu, 30 Jan 2014 22:00:28 -0500 From: "Theodore Ts'o" To: Luca Ognibene Cc: linux-kernel@vger.kernel.org, linux-ext4@vger.kernel.org Subject: Re: same ext4 file system corruption on different machines Message-ID: <20140131030028.GB7118@thunk.org> Mail-Followup-To: Theodore Ts'o , Luca Ognibene , linux-kernel@vger.kernel.org, linux-ext4@vger.kernel.org References: <1390999955.2601.22.camel@snow.station> <1391000743.2601.26.camel@snow.station> <20140129173826.GA30419@thunk.org> <1391068749.2680.6.camel@snow.station> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1391068749.2680.6.camel@snow.station> User-Agent: Mutt/1.5.21 (2010-09-15) X-SA-Exim-Connect-IP: X-SA-Exim-Mail-From: tytso@thunk.org X-SA-Exim-Scanned: No (on imap.thunk.org); SAEximRunCond expanded to false Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thu, Jan 30, 2014 at 08:59:09AM +0100, Luca Ognibene wrote: > Yes it's indeed very strange.. i tend to rule out application errors > because i don't write directly to the device so i don't think i can > break a filesystem from userspace. I've checked previous and next blocks > and they seem ok, only the block 524320 is getting corrupted. Any idea > on what should i look for now? Are you willing to try 3.12.9 or 3.13.1 upstream kernel? Let's see if changing the kernel makes any difference. I don't recall any ext4 problems like this, but maybe it's device driver problem. The other thing I'd ask is whether you can swap out the hard drive interface --- can you use a USB 3.0 attached drive, or something like that? One final thing that you could try doing, depending on how easily/quickly you can reproduce the problem, is to use blktrace and see if you can catch who or what is writing to that specific block which is getting corrupted. - Ted