Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.kernel > #1571175 > unrolled thread

Re: [PATCHv2 00/11] KVM implementation of PAPR HPT resizing extension

Started byPaul Mackerras <paulus@ozlabs.org>
First post2017-01-31 23:20 +0100
Last post2017-02-01 06:10 +0100
Articles 3 — 2 participants

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.


Contents

  Re: [PATCHv2 00/11] KVM implementation of PAPR HPT resizing extension Paul Mackerras <paulus@ozlabs.org> - 2017-01-31 23:20 +0100
    Re: [PATCHv2 00/11] KVM implementation of PAPR HPT resizing extension David Gibson <david@gibson.dropbear.id.au> - 2017-02-01 06:00 +0100
      Re: [PATCHv2 00/11] KVM implementation of PAPR HPT resizing extension Paul Mackerras <paulus@ozlabs.org> - 2017-02-01 06:10 +0100

#1571175 — Re: [PATCHv2 00/11] KVM implementation of PAPR HPT resizing extension

FromPaul Mackerras <paulus@ozlabs.org>
Date2017-01-31 23:20 +0100
SubjectRe: [PATCHv2 00/11] KVM implementation of PAPR HPT resizing extension
Message-ID<t5OWm-6cJ-21@gated-at.bofh.it>
On Tue, Dec 20, 2016 at 04:48:56PM +1100, David Gibson wrote:
> Here is the KVM implementation for the proposed PAPR extension which
> allows the runtime resizing of a PAPR guest's Hashed Page Table (HPT).
> 
> Using this requires a guest kernel with support for the extension.
> Patches for guest side support in Linux were posted earlier:
>   https://lists.ozlabs.org/pipermail/linuxppc-dev/2016-December/152164.html
> 
> It also requires userspace (i.e. qemu) to intercept the HPT resizing
> hypercalls and invoke the KVM ioctl()s to implement them.  This is
> done instead of having KVM direclty intercept the hypercalls, so that
> userspace can, if useful, impose additional restrictions on resizes:
> for example it could refuse them entirely if policy for the VM
> precludes resizing, or it could limit the size of HPT the guest can
> request to meet resource limits.
> 
> Patches to implement the userspace part of HPT resizing are proposed
> for qemu-2.9, and can be found at:
>   https://github.com/dgibson/qemu/tree/hpt-resize
> 
> I'm posting these now, in the hopes that both these and the
> corresponding guest side patches can be staged and merged for the 4.11
> window.

Thanks, series applied to my kvm-ppc-next branch.

Next time, please cc kvm-ppc@vger.kernel.org so I shows up in
patchwork and I don't miss it.

Paul.

[toc] | [next] | [standalone]


#1571267

FromDavid Gibson <david@gibson.dropbear.id.au>
Date2017-02-01 06:00 +0100
Message-ID<t5Vbr-1jC-5@gated-at.bofh.it>
In reply to#1571175

[Multipart message — attachments visible in raw view] — view raw

On Wed, Feb 01, 2017 at 09:18:13AM +1100, Paul Mackerras wrote:
> On Tue, Dec 20, 2016 at 04:48:56PM +1100, David Gibson wrote:
> > Here is the KVM implementation for the proposed PAPR extension which
> > allows the runtime resizing of a PAPR guest's Hashed Page Table (HPT).
> > 
> > Using this requires a guest kernel with support for the extension.
> > Patches for guest side support in Linux were posted earlier:
> >   https://lists.ozlabs.org/pipermail/linuxppc-dev/2016-December/152164.html
> > 
> > It also requires userspace (i.e. qemu) to intercept the HPT resizing
> > hypercalls and invoke the KVM ioctl()s to implement them.  This is
> > done instead of having KVM direclty intercept the hypercalls, so that
> > userspace can, if useful, impose additional restrictions on resizes:
> > for example it could refuse them entirely if policy for the VM
> > precludes resizing, or it could limit the size of HPT the guest can
> > request to meet resource limits.
> > 
> > Patches to implement the userspace part of HPT resizing are proposed
> > for qemu-2.9, and can be found at:
> >   https://github.com/dgibson/qemu/tree/hpt-resize
> > 
> > I'm posting these now, in the hopes that both these and the
> > corresponding guest side patches can be staged and merged for the 4.11
> > window.
> 
> Thanks, series applied to my kvm-ppc-next branch.

Great.  Where can I grab that tree?

> Next time, please cc kvm-ppc@vger.kernel.org so I shows up in
> patchwork and I don't miss it.

Sorry, I hadn't realized that list existed.  Noted for future
reference.

-- 
David Gibson			| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au	| minimalist, thank you.  NOT _the_ _other_
				| _way_ _around_!
http://www.ozlabs.org/~dgibson

[toc] | [prev] | [next] | [standalone]


#1571268

FromPaul Mackerras <paulus@ozlabs.org>
Date2017-02-01 06:10 +0100
Message-ID<t5Vl7-1BL-3@gated-at.bofh.it>
In reply to#1571267
On Wed, Feb 01, 2017 at 03:56:36PM +1100, David Gibson wrote:
> On Wed, Feb 01, 2017 at 09:18:13AM +1100, Paul Mackerras wrote:
> > On Tue, Dec 20, 2016 at 04:48:56PM +1100, David Gibson wrote:
> > > Here is the KVM implementation for the proposed PAPR extension which
> > > allows the runtime resizing of a PAPR guest's Hashed Page Table (HPT).
> > > 
> > > Using this requires a guest kernel with support for the extension.
> > > Patches for guest side support in Linux were posted earlier:
> > >   https://lists.ozlabs.org/pipermail/linuxppc-dev/2016-December/152164.html
> > > 
> > > It also requires userspace (i.e. qemu) to intercept the HPT resizing
> > > hypercalls and invoke the KVM ioctl()s to implement them.  This is
> > > done instead of having KVM direclty intercept the hypercalls, so that
> > > userspace can, if useful, impose additional restrictions on resizes:
> > > for example it could refuse them entirely if policy for the VM
> > > precludes resizing, or it could limit the size of HPT the guest can
> > > request to meet resource limits.
> > > 
> > > Patches to implement the userspace part of HPT resizing are proposed
> > > for qemu-2.9, and can be found at:
> > >   https://github.com/dgibson/qemu/tree/hpt-resize
> > > 
> > > I'm posting these now, in the hopes that both these and the
> > > corresponding guest side patches can be staged and merged for the 4.11
> > > window.
> > 
> > Thanks, series applied to my kvm-ppc-next branch.
> 
> Great.  Where can I grab that tree?

git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpc

> > Next time, please cc kvm-ppc@vger.kernel.org so I shows up in
> > patchwork and I don't miss it.
> 
> Sorry, I hadn't realized that list existed.  Noted for future
> reference.

Thanks,
Paul.

[toc] | [prev] | [standalone]


Back to top | Article view | linux.kernel


csiph-web