Re: [PATCH 2/4] KVM: SVM: Fixes setting V_IRQ while AVIC is still enabled

[Suravee Suthikulpanit <suravee.suthikulpanit@xxxxxxx>]

On 5/6/20 11:29 PM, Paolo Bonzini wrote:
> On 06/05/20 15:17, Suravee Suthikulpanit wrote:
> >    */
> > -void kvm_request_apicv_update(struct kvm *kvm, bool activate, ulong bit)
> > +void kvm_request_apicv_update(struct kvm *kvm, bool activate, ulong bit,
> > +			      struct kvm_vcpu *except)
> >   {
> >   	unsigned long old, new, expected;
> >   
> > @@ -8110,7 +8111,16 @@ void kvm_request_apicv_update(struct kvm *kvm, bool activate, ulong bit)
> >   	trace_kvm_apicv_update_request(activate, bit);
> >   	if (kvm_x86_ops.pre_update_apicv_exec_ctrl)
> >   		kvm_x86_ops.pre_update_apicv_exec_ctrl(kvm, activate);
> > -	kvm_make_all_cpus_request(kvm, KVM_REQ_APICV_UPDATE);
> > +
> > +	/*
> > +	 * Sending request to update APICV for all other vcpus,
> > +	 * while update the calling vcpu immediately instead of
> > +	 * waiting for another #VMEXIT to handle the request.
> > +	 */
> > +	kvm_make_all_cpus_request_except(kvm, KVM_REQ_APICV_UPDATE,
> > +					 except);
> > +	if (except)
> > +		kvm_vcpu_update_apicv(except);
>
> Can you use kvm_get_running_vcpu() instead of touching all callers?
> Also a slightly better subject would be "KVM: SVM: Disable AVIC before
> setting V_IRQ".
>
> Paolo

Right, that's better idea. I'll send out V2 separately.

Thanks,
Suravee