Re: [PATCH v2] Enable SMEP CPU feature support for XEN hypervisor

[Keir Fraser <keir.xen@gmail.com>]

On 05/06/2011 09:39, "Li, Xin" <xin.li@intel.com> wrote:

>> I mean, I know we may as well just hide the feature from PV 64b guests
>> totally. That's obvious. Let's stop talking about PV 64b guests already! The
>> question is: what to do about PV 32b guests?
> 
>> Quite obviously we ought to allow 32-bit pv guests to control this for
>> themselves (and hence see the feature).
> 
> That needs
> 1) inject SMEP faults back to the 32-bit pv guest.
> 2) let the guest see SMEP thru CPUID and config it in CR4 (actually it's
> already set, but just to let guest see it).
> 
> Anything else?

I thought about this myself and realised that we can't let PV guests control
this feature if we want Xen to benefit from it. There's little point in a
feature to protect Xen from guests, if an untrusted guest can turn it off!

Hence I think we probably have to leave the feature always on for PV guests.
Unless we find some guests are incompatible with that.

 -- Keir

>> Besides that, assuming Xin verified it's working, your latest patch
>> looks great to me.
> 
> Yeah, verified, the system crashed from a SMEP fault from 64-bit pv kernel.
> Thanks!
> -Xin