From mboxrd@z Thu Jan 1 00:00:00 1970 From: Pascal Terjan Subject: Re: Possible ext3 corruption with 1K block size Date: Thu, 16 Oct 2008 16:40:44 +0200 Message-ID: <1224168044.11578.15.camel@plop> References: <200810150724.42180.arvidjaar@mail.ru> <48F5E6D4.8080603@redhat.com> <200810151824.36119.arvidjaar@mail.ru> <48F601A5.1050405@redhat.com> <1224164860.11578.14.camel@plop> <48F751E9.8070705@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-15 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Andrey Borzenkov , pterjan@mandriva.org, linux-ext4@vger.kernel.org, Linux Kernel Mailing List To: Eric Sandeen Return-path: In-Reply-To: <48F751E9.8070705@redhat.com> Sender: linux-kernel-owner@vger.kernel.org List-Id: linux-ext4.vger.kernel.org Le jeudi 16 octobre 2008 =E0 09:38 -0500, Eric Sandeen a =E9crit : > Pascal Terjan wrote: > > Le mercredi 15 octobre 2008 =E0 09:43 -0500, Eric Sandeen a =E9crit= : > >> Andrey Borzenkov wrote: > >>> On Wednesday 15 October 2008, Eric Sandeen wrote: > >> > >>>> My kingdom for a testcase... does anyone have simple steps to re= produce > >>>> this? Or do they all start with "install mandriva on a 1k block= size > >>>> system?" :) > >>>> > >>> May be RH will do? :) > >> I did try a 1k-block root fs Fedora install, and didn't see any pr= oblems... > >> > >>> As indicated by last comment, Pascal has some ways to trigger it;= I > >>> forgot to Cc to him initially; doing it now. > >> Ok, good deal. > >> > >=20 > > On my test machine I reproduce it easily : rpm --rebuilddb and if t= he db > > is not detected to be corrupted yet it will be after installing a f= ew > > packages (tested again with 2.6.27). > >=20 > > If I do the rebuilddb on a 2.6.17 and then reboot on a recent kerne= l, > > then I can install/uninstall thousands of packages without any > > corruption. >=20 > so it seems to be the database rebuilding, under a recent kernel, whi= ch > causes the problem? installing under a recent kernel is ok, as long = as > the db was created on an older kernel? Yes