[PATCH 0 of 6] PV on HVM Xen

[PATCH 0 of 6] PV on HVM Xen

[Stefano Stabellini]

Hi all,
this is an incremental update on my last PV on HVM series, including
many bug fixes.
In particular multivcpu guests, suspend\resume, the unplug of
emulate devices and receving evtchns as interrupts from the platform pci
device should all work correctly now.
However the new pv clocksource is still missing, and also I left the
xen_guest_init code in enlighten.c for the moment.

This series is based on Jeremy's pvops kernel tree, branch name
stable-2.6.33.x.
The patch series consists of 6 patches, each patch comes with a detailed
description.

I want to thanks Sheng's for his very valuable contribution to this
series, that certainly wouldn't be at this stage without his help.

Any comment, critic or suggestion is very welcome.

Cheers,

Stefano

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

Re: [PATCH 0 of 6] PV on HVM Xen

[Jeremy Fitzhardinge]

On 04/22/2010 08:15 AM, Stefano Stabellini wrote:
> Hi all,
> this is an incremental update on my last PV on HVM series, including
> many bug fixes.
>   

Thanks!

> In particular multivcpu guests, suspend\resume, the unplug of
> emulate devices and receving evtchns as interrupts from the platform pci
> device should all work correctly now.
> However the new pv clocksource is still missing, and also I left the
> xen_guest_init code in enlighten.c for the moment.
>
> This series is based on Jeremy's pvops kernel tree, branch name
> stable-2.6.33.x.
>   

I still prefer patches to be based on an upstream kernel version if at
all possible (ideally 2.6.32).  If you depend on something that's not in
mainline, then merge the specific branch for that feature into your
pvhvm branch (IOW, it would help if you used a git-based workflow).

I'll try and fix this up in the meantime and see how it looks.

Thanks,
    J

Re: [PATCH 0 of 6] PV on HVM Xen

[Stefano Stabellini]

On Fri, 23 Apr 2010, Jeremy Fitzhardinge wrote:
> On 04/22/2010 08:15 AM, Stefano Stabellini wrote:
> > Hi all,
> > this is an incremental update on my last PV on HVM series, including
> > many bug fixes.
> >   
> 
> Thanks!
> 
> > In particular multivcpu guests, suspend\resume, the unplug of
> > emulate devices and receving evtchns as interrupts from the platform pci
> > device should all work correctly now.
> > However the new pv clocksource is still missing, and also I left the
> > xen_guest_init code in enlighten.c for the moment.
> >
> > This series is based on Jeremy's pvops kernel tree, branch name
> > stable-2.6.33.x.
> >   
> 
> I still prefer patches to be based on an upstream kernel version if at
> all possible (ideally 2.6.32).  If you depend on something that's not in
> mainline, then merge the specific branch for that feature into your
> pvhvm branch (IOW, it would help if you used a git-based workflow).
> 

it doesn't depend on anything in pvops; I have rebased my series on 2.6.32,
I setup a git tree here:

git://xenbits.xen.org/people/sstabellini/linux-pvhvm.git

branch name 2.6.32-pvhvm

Re: [PATCH 0 of 6] PV on HVM Xen

[Jeremy Fitzhardinge]

On 04/23/2010 10:41 AM, Stefano Stabellini wrote:
> it doesn't depend on anything in pvops; I have rebased my series on 2.6.32,
> I setup a git tree here:
>
> git://xenbits.xen.org/people/sstabellini/linux-pvhvm.git
>   

Great!  Thanks very much.

    J