This is an archived copy of the Xen.org mailing list, which we have preserved to ensure that existing links to archives are not broken. The live archive, which contains the latest emails, can be found at http://lists.xen.org/
Home Products Support Community News


Re: [Xen-devel] Xen 3.2.1-rc1: FATAL PAGE FAULT

To: "Christopher S. Aker" <caker@xxxxxxxxxxxx>, xen devel <xen-devel@xxxxxxxxxxxxxxxxxxx>
Subject: Re: [Xen-devel] Xen 3.2.1-rc1: FATAL PAGE FAULT
From: Keir Fraser <keir.fraser@xxxxxxxxxxxxx>
Date: Thu, 03 Apr 2008 16:55:29 +0100
Delivery-date: Thu, 03 Apr 2008 08:57:54 -0700
Envelope-to: www-data@xxxxxxxxxxxxxxxxxx
In-reply-to: <47F4E401.2080703@xxxxxxxxxxxx>
List-help: <mailto:xen-devel-request@lists.xensource.com?subject=help>
List-id: Xen developer discussion <xen-devel.lists.xensource.com>
List-post: <mailto:xen-devel@lists.xensource.com>
List-subscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-devel>, <mailto:xen-devel-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-devel>, <mailto:xen-devel-request@lists.xensource.com?subject=unsubscribe>
Sender: xen-devel-bounces@xxxxxxxxxxxxxxxxxxx
Thread-index: AciVoyOgYfnoIwGWEd2UmgAX8io7RQ==
Thread-topic: [Xen-devel] Xen 3.2.1-rc1: FATAL PAGE FAULT
User-agent: Microsoft-Entourage/
On 3/4/08 15:04, "Christopher S. Aker" <caker@xxxxxxxxxxxx> wrote:

>>> Maybe the symbols would help just a little bit?  In any case, here are
>>> the files:
>>> http://theshore.net/~caker/xen/BUGfatal_page_fault/
>> I will take a look. It might help narrow down the possibilities a bit.

My analysis is that the hypervisor crashed because one of the entries in a
dying guest's third-level page directory has the present bit (bit 0) set,
yet the physical address mapped by that entry is 0xb72dedde51000. That is a
rather large and obviously bogus number! It causes us to access way off the
end of an array indexed by physical address, resulting in a fatal page

Obviously the question is: Where did the bogus address come from?

That's going to be rather hard to answer without finding a more reliable
repro of the bug, and then adding some hypervisor tracing.

 -- Keir

Xen-devel mailing list