On 07/06/2018 12:37, David Hildenbrand wrote: > > I have a related requirement, which would be to atomically grow a > memory regions. So instead of region_del(old)+region_add(new), I would > have to do it in one shot (atomically). > > AFAICS an atomic replace of the memory map would work for this, too. > However I am not sure how we want to handle all kinds of tracking data > that is connected to e.g. x86 memory slots (e.g. rmap, dirty bitmap ...).
The dirty bitmap would be synced in kvm_region_del (so it's not true that kvm_region_del would disappear, but almost :)). The rmap is more interesting. Perhaps it can be just rebuilt on every KVM_SET_USER_MEMORY_REGIONS call. Paolo
