On 2017-11-16 18:21, Paolo Bonzini wrote:
> On 16/11/2017 06:06, Quan Xu wrote:
>> when vcpu thread is scheduled out, the pkru value in
>> current->thread.fpu.state may be the host pkru value, instead of
>> guest pkru value (of course, this _assumes_ that the pkru is in
>> current->thread.fpu.state as well). in this way, the pkru may be a
>> coner case.
> Rik may correct me, but I think this is not possible. Preemption is
> disabled all the time while PKRU = guest_pkru (which is only during
> vmx_vcpu_run).
refer to the following code,
vcpu_enter_guest()
{
...
preempt_disable()
...
kvm_x86_ops->run(vcpu); (actually it is vmx_vcpu_run())
...
...
preempt_enable();
}
when preempt_disable before to run kvm_x86_ops->run..
in kvm_x86_ops->run, the pkru is restored with host_pkru (IF guest_pkru
!= host_pkru).
all this happened under preempt_disable().
it's not true that preemtion is disable all the time while PKRU =
guest_pkru..
However it seems there is still some gap..
as Rik said, "at context switch time, the context switch code will save
the guest
FPU state to current->thread.fpu when the VCPU thread is scheduled out."
after preempt_enable() in vcpu_enter_guest(), the vcpu thread is
scheduled out,
in kvm_x86_ops->run, the PKRU has been restored with host_pkru (IF
guest_pkru != host_pkru)..
instead of guest_pkru..
then the PKRU is host_pkru, how to save guest_pkru current->thread.fpu?
as mentioned, all this _assumes_ that the pkru is in
current->thread.fpu.state as well.
thanks,
Quan
Alibaba Cloud
> Context switching will only happen in vcpu_enter_guest() after
> preempt_enable() for a preemptible kernel, or in vcpu_run via
> cond_resched() for a non-preemptible kernel.
>
> Thanks,
>
> Paolo
>
>> VM migration again, in case,
>> source_host_pkru_value != guest_pkru_value,
>> target_host_pkru_value == guest_pkru_value..
>>
>> the pkru status would be inconsistent..
>
From 1584225337503196489@xxx Thu Nov 16 12:23:35 +0000 2017
X-GM-THRID: 1584017174910331026
X-Gmail-Labels: Inbox,Category Forums,HistoricalUnread
On 16/11/2017 13:12, Quan Xu wrote:
> However it seems there is still some gap..
>
> as Rik said, "at context switch time, the context switch code will save
> the guest FPU state to current->thread.fpu when the VCPU thread is scheduled out."
By "guest FPU state" Rik means "guest FPU with host PKRU".
Guest PKRU is always stored in vcpu->arch.pkru rather than in the guest
FPU state, so guest PKRU will never be in current->thread.fpu.state either.
KVM_GET_XSAVE will the guest FPU state with vcpu->arch.pkru and
migration will work properly.
Thanks,
Paolo
> after preempt_enable() in vcpu_enter_guest(), the vcpu thread is
> scheduled out,
> in kvm_x86_ops->run, the PKRU has been restored with host_pkru (IF
> guest_pkru != host_pkru)..
> instead of guest_pkru..
>
> then the PKRU is host_pkru, how to save guest_pkru current->thread.fpu?
>
> as mentioned, all this _assumes_ that the pkru is in
> current->thread.fpu.state as well.
>
>
> thanks,
>
> Quan
> Alibaba Cloud
>
>> Context switching will only happen in vcpu_enter_guest() after
>> preempt_enable() for a preemptible kernel, or in vcpu_run via
>> cond_resched() for a non-preemptible kernel.
>>
>> Thanks,
>>
>> Paolo
>>
>>> VM migration again, in case,
>>> source_host_pkru_value != guest_pkru_value,
>>> target_host_pkru_value == guest_pkru_value..
>>>
>>> the pkru status would be inconsistent..
>>
>
From 1584222569117491443@xxx Thu Nov 16 11:39:35 +0000 2017
X-GM-THRID: 1584017174910331026
X-Gmail-Labels: Inbox,Category Forums,HistoricalUnread