From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752035AbZBPPUd (ORCPT ); Mon, 16 Feb 2009 10:20:33 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1750960AbZBPPUW (ORCPT ); Mon, 16 Feb 2009 10:20:22 -0500 Received: from thunk.org ([69.25.196.29]:45641 "EHLO thunker.thunk.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750814AbZBPPUV (ORCPT ); Mon, 16 Feb 2009 10:20:21 -0500 Date: Mon, 16 Feb 2009 10:01:56 -0500 From: Theodore Tso To: Andres Freund Cc: adilger@sun.com, LKML , linux-ext4@vger.kernel.org, Jonathan Bastien-Filiatrault , "Aneesh Kumar K.V" Subject: Re: EXT4 ENOSPC Bug Message-ID: <20090216150156.GD22619@mini-me.lan> Mail-Followup-To: Theodore Tso , Andres Freund , adilger@sun.com, LKML , linux-ext4@vger.kernel.org, Jonathan Bastien-Filiatrault , "Aneesh Kumar K.V" References: <200811291418.24672.andres@anarazel.de> <200812100108.04163.andres@anarazel.de> <49994FEF.2020908@anarazel.de> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <49994FEF.2020908@anarazel.de> User-Agent: Mutt/1.5.18 (2008-05-17) X-SA-Exim-Connect-IP: X-SA-Exim-Mail-From: tytso@thunk.org X-SA-Exim-Scanned: No (on thunker.thunk.org); SAEximRunCond expanded to false Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Mon, Feb 16, 2009 at 12:37:19PM +0100, Andres Freund wrote: > Hi All, > > "Luckily", to prove that I am not completly mad I recently found this > problem again. On a more recent kernel: > 2.6.29-rc3-andres-00498-g68e80d5 > (upstream 2.6.29 + Theodore's ext4 debug patches) > > Again I got: > open("/home/andres/tt", O_WRONLY|O_CREAT|O_NOCTTY|O_NONBLOCK, 0666) = -1 > ENOSPC (No space left on device) So /home/andres/tt does not exist, correct? Does this happen for any attempt to create a new zero-length file using the "touch" command? (Even in other directories) Can you append to a file using "cat /etc/mailcap >> some_existing_file"? If it is "yes" to the first two questions, and "no" the third, it looks like this might be an inode allocation failure as opposed to a block allocation failure. Once this triggers, does it reliably continue to fail if you reboot? Or does it go away when you reboot? - Ted