Re: [Xen-devel] [PATCH] x86/VMX: don't risk corrupting host CR4

["Jan Beulich" <JBeulich@xxxxxxxx>]

>>> On 09.03.18 at 15:35, <andrew.cooper3@xxxxxxxxxx> wrote:
> On 09/03/18 14:18, Jan Beulich wrote:
>> Instead of "syncing" the live value to what mmu_cr4_features has, make
>> sure vCPU-s run with the value most recently loaded into %cr4, such that
>> after the next VM exit we continue to run with the intended value rather
>> than a possibly stale one.
>>
>> Signed-off-by: Jan Beulich <jbeulich@xxxxxxxx>
>> ---
>> TBD: Is the conditional really worthwhile, i.e. is the VMWRITE perhaps
>>      not meaningfully more expensive than the VMREAD?
> 
> Which bits are you worried about here?  We play with %cr4 quite a bit in
> PV context, but we shouldn't be changing after boot in HVM context.

See the discussion on Jürgen's series, playing with CR4.PGE. By
doing what I do here, VMX and SVM will no longer behave
differently in regard to CR4, which I think is quite desirable. But
even beyond that it is a recipe for problems to force a control
register to some fixed value.

Jan

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel