Re: [Xen-devel] [PATCH 2/2] vm_event: Add altp2m info to HVM events as well

[Tamas K Lengyel <tamas@xxxxxxxxxxxxx>]

On Jan 12, 2016 3:21 AM, "Jan Beulich" <JBeulich@xxxxxxxx> wrote:
>
> >>> On 06.01.16 at 12:50, <tamas@xxxxxxxxxxxxx> wrote:
> > On Wed, Jan 6, 2016 at 12:48 PM, Andrew Cooper <andrew.cooper3@xxxxxxxxxx>
> > wrote:
> >
> >> On 06/01/16 11:42, Tamas K Lengyel wrote:
> >>
> >>
> >>
> >> On Wed, Jan 6, 2016 at 12:32 PM, Jan Beulich <JBeulich@xxxxxxxx> wrote:
> >>
> >>> >>> On 23.12.15 at 15:53, < <tamas@xxxxxxxxxxxxx>tamas@xxxxxxxxxxxxx>
> >>> wrote:
> >>> > @@ -83,6 +84,12 @@ static int hvm_event_traps(uint8_t sync,
> >>> vm_event_request_t *req)
> >>> >Â Â Â Â Â vm_event_vcpu_pause(curr);
> >>> >Â Â Â }
> >>> >
> >>> > +Â Â if ( altp2m_active(currd) )
> >>> > +Â Â {
> >>> > +Â Â Â Â req->flags |= VM_EVENT_FLAG_ALTERNATE_P2M;
> >>> > +Â Â Â Â req->altp2m_idx = vcpu_altp2m(curr).p2midx;
> >>> > +Â Â }
> >>>
> >>> So far this info was provided just for MEM_ACCESS events. Now
> >>> you provide it for a few more ones, but wouldn't this then better
> >>> be supplied for all of them, namely also the other two MEM ones?
> >>>
> >>
> >> AFAIK altp2m is currently incompatible with sharing. I'm not 100% sure but
> >> I think it's also incompatible with paging.
> >>
> >>
> >> I don't think they are strictly incompatible; I don't see a technical
> >> reason preventing some development work to make them function together.
> >>
> >> Whether this happens or not is a very different matter.
> >
> > Sure, the two systems can be made to work in tandem, this work just hasn't
> > been done yet. I would very much like to get that to work in the future.
>
> Which re-raises the question: Shouldn't the information then be
> made available uniformly for all events?
>

IMHO there is no point doing so while the systems don't work together.

Tamas

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel