Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

VMI and kmap_atomic_pte paravirt_op

0 views
Skip to first unread message

Ian Campbell

unread,
Feb 25, 2010, 3:00:02 PM2/25/10
to
Hi Alok,

How critical is the vmi_kmap_atomic_pte pv_mmu_op hook to VMI? Is it
required for correct functionality or is it simply an optimisation?

We think we may be able to do away with the Xen version of the hook and
if VMI (which is deprecated) can live without it and you agree then I'll
cook up a patch to remove it altogether.

Thanks,
Ian.


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majo...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

Jeremy Fitzhardinge

unread,
Feb 25, 2010, 3:20:02 PM2/25/10
to
On 02/25/2010 11:52 AM, Ian Campbell wrote:
> How critical is the vmi_kmap_atomic_pte pv_mmu_op hook to VMI? Is it
> required for correct functionality or is it simply an optimisation?
>

Or disable high ptes for VMI too.

J

Alok Kataria

unread,
Feb 25, 2010, 3:40:03 PM2/25/10
to
Hi Ian,

On Thu, 2010-02-25 at 11:52 -0800, Ian Campbell wrote:
> Hi Alok,
>
> How critical is the vmi_kmap_atomic_pte pv_mmu_op hook to VMI? Is it
> required for correct functionality or is it simply an optimisation?
>
> We think we may be able to do away with the Xen version of the hook and
> if VMI (which is deprecated) can live without it and you agree then I'll
> cook up a patch to remove it altogether.

As Jeremy suggests we will be fine with disabling high ptes for VMI.
Thanks for doing this.

Alok

Ian Campbell

unread,
Feb 26, 2010, 4:10:02 AM2/26/10
to
On Thu, 2010-02-25 at 20:32 +0000, Alok Kataria wrote:
> Hi Ian,
>
> On Thu, 2010-02-25 at 11:52 -0800, Ian Campbell wrote:
> > Hi Alok,
> >
> > How critical is the vmi_kmap_atomic_pte pv_mmu_op hook to VMI? Is it
> > required for correct functionality or is it simply an optimisation?
> >
> > We think we may be able to do away with the Xen version of the hook and
> > if VMI (which is deprecated) can live without it and you agree then I'll
> > cook up a patch to remove it altogether.
>
> As Jeremy suggests we will be fine with disabling high ptes for VMI.
> Thanks for doing this.

OK, I'll do that.

0 new messages