On Wed, Jul 31, 2024 at 6:01 PM Paolo Bonzini <pbonzini@xxxxxxxxxx> wrote: > > On 7/31/24 11:09, Hao Peng wrote: > >> Yep. And kvm_zap_obsolete_pages() is a relatively cheap nop if there are no > >> pages on active_mmu_pages. E.g. we could check kvm_memslots_have_rmaps(), but I > >> don't see any point in doing so, as the existing code should be blazing fast > >> relative to the total cost of the zap. > > Here can be optimized by judging whether active_mmu_pages is empty, > > just like kvm_zap_obsolete_pages. > > Regardless of L0 kvm or L1 kvm, when tdp_mmu is enabled, the > > active_mmu_pages list will not be used. > > When ept=0 , the probability that active_mmu_pages is empty is also > > high, not every time > > kvm_zap_obsolete_pages is called. > > So if anything you could check list_empty(&kvm->arch.active_mmu_pages) > before the loop of kvm_zap_obsolete_pages(), similar to what is done in > kvm_mmu_zap_oldest_mmu_pages(). I doubt it can have any practical > benefit, though. > > Paolo > I did some tests, when ept=0, kvm_zap_obsolete_pages was called 42 times, and only 17 times active_mmu_page list was not empty. When tdp_mmu was enabled, active_mmu_page list was always empty.
