Re: [PATCH v5 4/7] kvm, mem-hotplug: Reload L1' apic access page on migration in vcpu_enter_guest().

[Gleb Natapov <gleb@kernel.org>]

On Thu, Sep 11, 2014 at 12:47:16PM +0200, Paolo Bonzini wrote:
> Il 11/09/2014 12:12, Gleb Natapov ha scritto:
> > On Thu, Sep 11, 2014 at 11:21:49AM +0200, Paolo Bonzini wrote:
> >> Il 11/09/2014 07:38, Tang Chen ha scritto:
> >>> diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
> >>> index 63c4c3e..da6d55d 100644
> >>> --- a/arch/x86/kvm/vmx.c
> >>> +++ b/arch/x86/kvm/vmx.c
> >>> @@ -7093,6 +7093,11 @@ static void vmx_set_virtual_x2apic_mode(struct kvm_vcpu *vcpu, bool set)
> >>>  	vmx_set_msr_bitmap(vcpu);
> >>>  }
> >>>  
> >>> +static void vmx_set_apic_access_page_addr(struct kvm *kvm, hpa_t hpa)
> >>> +{
> >>> +	vmcs_write64(APIC_ACCESS_ADDR, hpa);
> >>
> >> This has to be guarded by "if (!is_guest_mode(vcpu))".
> >>
> > We do need to write it if L1 and L2 share APIC_ACCESS_ADDR and skip
> > it otherwise, no?
> 
> Yes, but this would be handled by patch 6:
> 
>  		} else if (vm_need_virtualize_apic_accesses(vmx->vcpu.kvm)) {
> +			struct page *page = gfn_to_page(vmx->vcpu.kvm,
> +				APIC_DEFAULT_PHYS_BASE >> PAGE_SHIFT);
>  			exec_control |=
>  				SECONDARY_EXEC_VIRTUALIZE_APIC_ACCESSES;
> -			vmcs_write64(APIC_ACCESS_ADDR,
> -				page_to_phys(vcpu->kvm->arch.apic_access_page));
> +			vmcs_write64(APIC_ACCESS_ADDR, page_to_phys(page));
> +			/*
> +			 * Do not pin apic access page in memory so that memory
> +			 * hotplug process is able to migrate it.
> +			 */
> +			put_page(page);
>  		}
This code is in prepare_vmcs02() and is executed during L1->L2 vmentry. What happens
when apic access page is migrated while L2 is running? It needs to be update somewhere.

> 
> However, this is also useless code duplication because the above snippet could
> reuse vcpu_reload_apic_access_page too.
> 
> So I think you cannot do the is_guest_mode check in
> kvm_vcpu_reload_apic_access_page and also not in
> vmx_reload_apic_access_page.  But you could do something like
> 
> kvm_vcpu_reload_apic_access_page(...)
> {
> 	...
> 	kvm_x86_ops->reload_apic_access_page(...);
> }
> EXPORT_SYMBOL_GPL(kvm_vcpu_reload_apic_access_page);
> 
> /* used in vcpu_enter_guest only */
> vcpu_reload_apic_access_page(...)
> {
> 	if (!is_guest_mode(vcpu))
> 		kvm_vcpu_reload_apic_access_page(...)
> }
> 
> Paolo

--
			Gleb.