On Thu, Nov 10, 2011 at 7:14 AM, Jan Beulich <JBeulich@xxxxxxxx> wrote:
>>>> On 10.11.11 at 16:00, Gianluca Guida <glguida@xxxxxxxxx> wrote:
>> On Thu, Nov 10, 2011 at 6:59 AM, Stefano Stabellini
>> <stefano.stabellini@xxxxxxxxxxxxx> wrote:
>>> On Thu, 10 Nov 2011, Jan Beulich wrote:
>>>> It's SLE11 SP1 guests that suffered a regression after a maintenance
>>>> update (originally shipped with 4.0.0, while that patch got backported
>>>> later into 4.0.x).
>>>
>>> Is SLES11 SP1 using HVMOP_pagetable_dying (see
>>> arch/x86/xen/mmu.c:xen_hvm_init_mmu_ops)?
>>
>> Good question. In case it does not, just changing the ">l1" check to
>> work only in case we used the hypercall might do the trick.
>
> Possibly, though I would argue that the difference between Linuxes
> shouldn't be that big (i.e. I would rather expect that pv-on-hvm
> guests also see a performance penalty here).
No, because what the patch does is to unshadow the user pagetables
(all levels) as soon as the hypercall is made (when the kernel is
destroying the process). check_for_early_unshadow(), when the
hypercall is used, is for removing the pagetables that maps some
process-specific non-user mappings.
Importing the hypercall is a good idea, but fixing the logic would
help linux guests not using the hypercall.
Gianluca
--
It was a type of people I did not know, I found them very strange and
they did not inspire confidence at all. Later I learned that I had been
introduced to electronic engineers.
E. W. Dijkstra
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel
|
Home