On 03/09/2011 09:43 AM, Xiao Guangrong wrote:
This patch does: - call vcpu->arch.mmu.update_pte directly - use gfn_to_pfn_atomic in update_pte path The suggestion is from Avi. - mmu_guess_page_from_pte_write(vcpu, gpa, gentry); + mmu_seq = vcpu->kvm->mmu_notifier_seq; + smp_rmb();
smp_rmb() should come before, no? but the problem was present in the original code, too.
+ spin_lock(&vcpu->kvm->mmu_lock); if (atomic_read(&vcpu->kvm->arch.invlpg_counter) != invlpg_counter) gentry = 0; @@ -3365,7 +3345,8 @@ void kvm_mmu_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa, if (gentry&& !((sp->role.word ^ vcpu->arch.mmu.base_role.word) & mask.word)) - mmu_pte_write_new_pte(vcpu, sp, spte,&gentry); + mmu_pte_write_new_pte(vcpu, sp, spte,&gentry, + mmu_seq);
Okay, we're only iterating over indirect pages, so we won't call nonpaging_update_pte().
-- error compiling committee.c: too many arguments to function -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
