| Message ID | a8af26748dc6168c8363495b05cf8eb0db001347.1474911775.git.pfeiner@google.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
2016-09-26 10:45-0700, Peter Feiner: > The MMU notifier sequence number keeps GPA->HPA mappings in sync when > GPA->HPA lookups are done outside of the MMU lock (e.g., in > tdp_page_fault). Since kvm_age_hva doesn't change GPA->HPA, it's > unnecessary to increment the sequence number. > > Signed-off-by: Peter Feiner <pfeiner@google.com> > --- Applied to kvm/queue, thanks. -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c index 3d4cc8cc..dc6d1e8 100644 --- a/arch/x86/kvm/mmu.c +++ b/arch/x86/kvm/mmu.c @@ -1660,17 +1660,9 @@ int kvm_age_hva(struct kvm *kvm, unsigned long start, unsigned long end) * This has some overhead, but not as much as the cost of swapping * out actively used pages or breaking up actively used hugepages. */ - if (!shadow_accessed_mask) { - /* - * We are holding the kvm->mmu_lock, and we are blowing up - * shadow PTEs. MMU notifier consumers need to be kept at bay. - * This is correct as long as we don't decouple the mmu_lock - * protected regions (like invalidate_range_start|end does). - */ - kvm->mmu_notifier_seq++; + if (!shadow_accessed_mask) return kvm_handle_hva_range(kvm, start, end, 0, kvm_unmap_rmapp); - } return kvm_handle_hva_range(kvm, start, end, 0, kvm_age_rmapp); }
The MMU notifier sequence number keeps GPA->HPA mappings in sync when GPA->HPA lookups are done outside of the MMU lock (e.g., in tdp_page_fault). Since kvm_age_hva doesn't change GPA->HPA, it's unnecessary to increment the sequence number. Signed-off-by: Peter Feiner <pfeiner@google.com> --- arch/x86/kvm/mmu.c | 10 +--------- 1 file changed, 1 insertion(+), 9 deletions(-)