WARNING - OLD ARCHIVES

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/
   
 
 
Xen 
 
Home Products Support Community News
 
   
 

xen-ia64-devel

MagiXen (was Re: [Xen-ia64-devel] Dan Magenheimer leaving HP, and MagiX

To: xen-ia64-devel@xxxxxxxxxxxxxxxxxxx
Subject: MagiXen (was Re: [Xen-ia64-devel] Dan Magenheimer leaving HP, and MagiXen tech report)
From: Mark Williamson <mark.williamson@xxxxxxxxxxxx>
Date: Sat, 26 May 2007 23:35:10 +0100
Cc: "Magenheimer, Dan \(HP Labs Fort Collins\)" <dan.magenheimer@xxxxxx>, Partha.Ranganathan@xxxxxx
Delivery-date: Sat, 26 May 2007 15:33:17 -0700
Envelope-to: www-data@xxxxxxxxxxxxxxxxxx
In-reply-to: <B1AC58F6B4AE1B43972576AD3125882EE2ADF3@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>
List-help: <mailto:xen-ia64-devel-request@lists.xensource.com?subject=help>
List-id: Discussion of the ia64 port of Xen <xen-ia64-devel.lists.xensource.com>
List-post: <mailto:xen-ia64-devel@lists.xensource.com>
List-subscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-ia64-devel>, <mailto:xen-ia64-devel-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-ia64-devel>, <mailto:xen-ia64-devel-request@lists.xensource.com?subject=unsubscribe>
References: <B1AC58F6B4AE1B43972576AD3125882EE2ADF3@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>
Sender: xen-ia64-devel-bounces@xxxxxxxxxxxxxxxxxxx
User-agent: KMail/1.9.6
> Last year, there was a brief exchange on xen-ia64-devel asking
> if it was possible to run xen-ia32 virtual machines on xen-ia64.
> The answer was of course "No", but several people were intrigued
> by the idea.  Some of us in HP Labs decided to develop the idea,
> implemented a prototype called MagiXen (pronounced "magician"),
> and submitted a paper for Usenix 2007. Sadly, we ran out of time
> and although it worked, the initial performance results did not
> look particularly good and the paper was rejected. So we have
> instead published the paper as an HP Labs Tech Report that can
> be found here:
>
> http://www.hpl.hp.com/techreports/2007/HPL-2007-77.pdf

This is something I've wondered about doing in the past - it's quite exciting 
to see it actively persued.  I rather like the concept of doing this.  In a 
dynamic data centre with VMs being live migrated around, matching 
architectures would presumably become another optional optimisation :-)

A couple of questions arose:

* Did you try building a VHPT in the domain and having Xen walk that as it 
does for IA64 guests, as an alternative to the x86 pagetable walker?

* You mention modifying the guest to expose address spaces to improve context 
switching performance...  couldn't you do address space tagging based on 
whether value of CR3 was pinned in the validated state?  If the guest is 
Xenified, it should be notifying Xen of modified pagetable entries (either 
through traps on write protected pages, or through explicit calls), which I'd 
have thought would supply the information you need to make this work 
efficiently.

* Is there anywhere you could inline your x86-IA64 pipefitting directly into 
the cached code translations if you were able to add this feature to the 
translator?  Do you think this would be worthwhile?

Cheers,
Mark

-- 
Dave: Just a question. What use is a unicyle with no seat?  And no pedals!
Mark: To answer a question with a question: What use is a skateboard?
Dave: Skateboards have wheels.
Mark: My wheel has a wheel!

_______________________________________________
Xen-ia64-devel mailing list
Xen-ia64-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-ia64-devel

<Prev in Thread] Current Thread [Next in Thread>