From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from 0122700014.0.fullrate.dk ([95.166.99.235]:41915 "EHLO kernel.dk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753187Ab0FZUvM (ORCPT ); Sat, 26 Jun 2010 16:51:12 -0400 Message-ID: <4C26683D.8010006@fusionio.com> Date: Sat, 26 Jun 2010 22:51:09 +0200 From: Jens Axboe MIME-Version: 1.0 Subject: Re: [PATCH] gracefully handle full /tmp file system References: <20100625152419.GA29790@w-gedwards.lhn.com> In-Reply-To: <20100625152419.GA29790@w-gedwards.lhn.com> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Sender: fio-owner@vger.kernel.org List-Id: fio@vger.kernel.org To: Greg Edwards Cc: fio@vger.kernel.org On 25/06/10 17.24, Greg Edwards wrote: > If /tmp happens to be full, fio gives you a very unhelpful error: > > # ./fio rand-write-256k-q256.job > Bus error > > Use posix_fallocate() to gracefully handle this condition. > > Signed-off-by: Greg Edwards Thanks, that looks sane. As far as I remember, if the fs doesn't do fallocate, then glibc will zero fill it. If that is the case, then it should be safe. I will apply it. -- Jens Axboe