On Mon 28-08-23 14:18:30, Jean Delvare wrote: [...] > > > It would likely be better to use refcount_t instead of atomic_t. > > > > Patches welcomed. > > Michal, do I understand correctly that this would prevent the case our > customer had (too many gets), but won't make a difference for actual > too-many-puts situations? yes, refcount_t cannot protect from too-many-puts because there is not real way to protect from those AFAICS. At a certain moment you just drop to 0 and lose your object and all following that is a UAF. But I do not think this is actually the interesting case at all. -- Michal Hocko SUSE Labs
