Each ring is an array of kvm_dirty_gfn\rĮach GFN is a state machine itself. ![]() We only need to read the ring data, no ioctl is needed.\r KVM_RESET_DIRTY_RINGS ioctl when we want to reset the collected\rĭirty pages to protected mode again (works like\r KVM_CLEAR_DIRTY_LOG interfaces, the new ring uses the new\r Interface: Instead of using the old KVM_GET_DIRTY_LOG,\r ![]() Data copy: The sync of dirty pages does not need data copy any more,\rīut instead the ring is shared between the userspace and kernel by\r Also, the dirty ring is per-vcpu, while the dirty\r Data format: The dirty data was in a ring format rather than a\rīitmap format, so dirty bits to sync for dirty logging does not\rĭepend on the size of guest memory any more, but speed of\rĭirtying. It is different from the existing dirty logging\r The new dirty ring interface is another way to collect dirty pages for\r patch 3: _x86_set_memory_region: squash another trivial change to return\rīonzini on the KVM dirty ring interface.\r s/false/null/ in gfn_to_memslot_dirty_bitmap() \r kvm_page_in_dirty_ring() use "#if" macros for KVM_DIRTY_LOG_PAGE_OFFSET to\r remove unused identity_map_pfn in init_rmode_identity_map \r rename 'r' in alloc_apic_access_page into 'hva' \r let _copy_to_user() always return -EFAULT \r set memslot->as_id in kvm_delete_memslot() \r removed kvm_dirty_ring_waitqueue() tracepoint since not used\r rebased to kvm/queue (seems the newest)\r David Alan Gilbert, peterx, Andrew Jones, ![]() ` (13 more replies) 0 siblings, 14 replies 17 messages in threadĬc: Dr. 21:49 ` KVM: Cache as_id in kvm_memory_slot Peter Xu ![]() KVM: Dirty ring interface LKML Archive on help / color / mirror / Atom feed * KVM: Dirty ring interface 21:49 Peter Xu
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |