Re: [PATCH 2/2] KVM: MMU: optimize pte write path if don't have protectedsp

[Avi Kivity]

On 03/28/2011 04:30 AM, Xiao Guangrong wrote:
> Simplely return from kvm_mmu_pte_write path if no shadow page is
> write-protected, than we can avoid to walk all shadow pages and hold
> mmu-lock
>
> Signed-off-by: Xiao Guangrong<xiaoguangrong@xxxxxxxxxxxxxx>
> ---
>   arch/x86/kvm/mmu.c |    8 ++++++++
>   1 files changed, 8 insertions(+), 0 deletions(-)
>
> diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c
> index 2841805..ccac6c4 100644
> --- a/arch/x86/kvm/mmu.c
> +++ b/arch/x86/kvm/mmu.c
> @@ -3219,6 +3219,11 @@ static void kvm_mmu_access_page(struct kvm_vcpu *vcpu, gfn_t gfn)
>   		set_bit(PT_ACCESSED_SHIFT, (unsigned long *)spte);
>   }
>
> +static bool need_track_pte_changed(struct kvm_vcpu *vcpu)
> +{
> +	return !tdp_enabled || mmu_is_nested(vcpu);
> +}
> +

What happens if this vcpu is nested but another vcpu is not?

Or if we have some indirect pages, but currently we're in non-nested mode?

Suggest instead to keep track of the number of indirect pages across the 
guest.  If it's zero, we can exit early.

>   void kvm_mmu_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa,
>   		       const u8 *new, int bytes,
>   		       bool guest_initiated)
> @@ -3233,6 +3238,9 @@ void kvm_mmu_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa,
>   	int level, npte, invlpg_counter, r, flooded = 0;
>   	bool remote_flush, local_flush, zap_page;
>
> +	if (!need_track_pte_changed(vcpu))
> +		return;
> +
>   	zap_page = remote_flush = local_flush = false;
>   	offset = offset_in_page(gpa);


--
error compiling committee.c: too many arguments to function

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/