Re: [XenARM] RE: xen-arm mailing list archive

["George G. Davis" <gdavis@xxxxxxxxxx>]

On Thu, Sep 13, 2007 at 02:20:03PM +0200, ROSSIER Daniel wrote:
> >-----Original Message-----
> >From: Toni Nikkanen [mailto:toni.nikkanen@xxxxxxxxxxx]
> >Sent: jeudi 13 septembre 2007 13:37
> >To: ROSSIER Daniel
> >Subject: Re: xen-arm mailing list archive
> >
> >
> >Actually I was trying to read the xen-arm archives because I am trying
> >to find out, what is the
> >status of Xen on the ARM architecture.

> >I am aware of the presentation
> >given by Samsung at the Xen
> >Summit of this year and it looks very interesting, however there is no
> >mention if they are going to
> >release their ARM port to general use or not.

FWIW, I also asked about the Samsung port a few months back and heard no
replies from anyone:

http://lists.xensource.com/archives/html/xen-devel/2007-07/msg00136.html


> >I have also found two
> >university lab projects that
> >have reached various levels of completion in porting Xen to ARM - and
> >then I found your
> >announcement on xen-devel. 
> 
> Indeed, we also sent several mails to Samsung but never got any answer.
> Maybe, they still hesitate
> to release their code to the open source community.
> 
> I am especially interested in Xen on ARM11
> >MPCore processors, for
> >running Linux guests. I would be grateful if you could fill me in on
> the
> >current situation and
> >especially if you're aware of anyone interested in supporting the ARM11
> >MPCore. Thank you!
> 
> I've no experience with the ARM11 core, but I guess we should start
> investigations on these series in
> the upcoming months.

We're also interested in working on Xen support for ARM11 (UP for
now) and ARM in general.


> At the moment, we still focus on ARM9 or Xscale
> series.

We plan to focus on ARM1136 but will also use ARM926 for development.


> Daniel Ferstay, who graduated at the University of British Columbia did
> a great job; he ported the hypervisor on a StrongARM-based platform
> enabling the support of one guest OS. I recommend the reading of his
> Master thesis which may significantly help in understanding the major
> issues regarding the port, especially the memory management:
> http://www.cs.ubc.ca/grads/resources/thesis/May06/Ferstay_Daniel.pdf

Although we were aware of the Ferstay et al paper, we were not aware
that they had that much progress.  It's nice to here that they have
already completed their port and hope we all get to see the source soon.


> Another port was ongoing at UIUC (University of Illinois at
> Urbana-Champaign) by Philip Reames, but I've been informed
> that they stopped their work and shifted toward the development of
> avirtualization system based on L4 and some of the new features of the
> ARMv6 architecture.
> 
> I kindly asked Daniel Ferstay if I was authorized to publish his code
> over the mailing list and now waiting for his answer. 

It would be fantastic if Ferstay et al would publish their code!  Perhaps
the folks at Xensource would host a Xen ARM HG repository once we have
that code available (although I would personally prefer a GIT repository
myself  : ).  That may also then encourage more active development and
contributions from others lurking on the list.


> (Or maybe he can do it himself directly :-))

Would it be impolite to add Daniel Ferstay to the dist list or did
you already BCC him?


Meanwhile, we're also interested in porting Xen to ARM and would like
to participate in the community porting effort to help get this moving
along out in the open...

> Daniel

-- 
Regards,
George

_______________________________________________
Xen-arm mailing list
Xen-arm@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-arm