On Mon 14-12-20 15:21:21, David Hildenbrand wrote: > On 14.12.20 14:36, Jason Gunthorpe wrote: > > On Sat, Dec 12, 2020 at 08:29:11AM +0100, David Hildenbrand wrote: > > > >>> Racing with another GUP in another thread is also not reasonable, so > >>> failing to isolate can't be a failure > >> > >> Having VMs with multiple vfio containers is certainly realistic, and > >> optimizing in user space to do vfio mappings concurrently doesn‘t > >> sound too crazy to me. But I haven‘t checked if vfio common code > >> already handles such concurrency. > > > > There is a lot more out there than vfio.. RDMA already does concurrent > > pin_user_pages in real apps > > I actually misread your comment. I think we both agree that temporary > isolation failures must not lead to a failure. Yes, isolation failures are ephemeral. I believe that the migration should start distinguishing between these and hard failures. -- Michal Hocko SUSE Labs
