[99] in athena10
Re: Corruption bug affecting debathenificator
daemon@ATHENA.MIT.EDU (Greg Hudson)
Thu Feb 28 00:45:53 2008
From: Greg Hudson <ghudson@MIT.EDU>
To: athena10@mit.edu
In-Reply-To: <1204150756.5862.21.camel@error-messages.mit.edu>
Content-Type: text/plain
Date: Thu, 28 Feb 2008 00:45:23 -0500
Message-Id: <1204177523.5862.28.camel@error-messages.mit.edu>
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
On Wed, 2008-02-27 at 17:19 -0500, Greg Hudson wrote:
> A second possibility is that something schroot -b is doing is causing
> kernel memory corruption in the page cache. One way to verify this
> assessment would be to dump the page cache and see if the file remains
> corrupted.
Yup.
root@linux-build-10:~# mv /tmp/debathenify.6430.IldosQ6991 /root
root@linux-build-10:~# md5sum /root/debathenify.6430.IldosQ6991/krb5_1.6.dfsg.1.orig.tar.gz
f40e9bbc6cabe7108a1aebab6bebce4f /root/debathenify.6430.IldosQ6991/krb5_1.6.dfsg.1.orig.tar.gz
root@linux-build-10:~# reboot
root@linux-build-10:~# md5sum /root/debathenify.6430.IldosQ6991/krb5_1.6.dfsg.1.orig.tar.gz
8f8d6a494380f01a7a0a9236162afa52 /root/debathenify.6430.IldosQ6991/krb5_1.6.dfsg.1.orig.tar.gz
(The second md5sum matches the AFS copy.)
Kind of mystified why this only happens with debathenify-krb5; you'd
think a kernel memory corruption bug in LVM snapshots would mess up our
whole build system.
