On Tue, Oct 2, 2018 at 7:28 AM Dmitry Vyukov <dvyu...@google.com> wrote: > > On Tue, Oct 2, 2018 at 4:04 PM, Eric Dumazet <eduma...@google.com> wrote: > > On Tue, Oct 2, 2018 at 6:46 AM Dmitry Vyukov <dvyu...@google.com> wrote: > >> > >> On Tue, Oct 2, 2018 at 3:16 PM, Eric Dumazet <eduma...@google.com> wrote: > >> > On Tue, Oct 2, 2018 at 1:19 AM Dmitry Vyukov <dvyu...@google.com> wrote: > >> >> > >> >> On Tue, Oct 2, 2018 at 7:49 AM, Eric Dumazet <eduma...@google.com> > >> >> wrote: > >> >> > >> >> > >> >> Does inet_frag_kill() hold fq->lock? I am missing how inet_frag_kill() > >> >> and inet_frags_exit_net() are synchronized. > >> >> Since you use smp_store_release()/READ_ONCE() they seem to run in > >> >> parallel. But then isn't it possible that inet_frag_kill() reads > >> >> nf->dead == 0, then inet_frags_exit_net() sets nf->dead, and then we > >> >> have the same race on concurrent removal? Or, isn't it possible that > >> >> inet_frag_kill() reads nf->dead == 1, but does not set > >> >> INET_FRAG_HASH_DEAD yet, and then inet_frags_free_cb() misses the > >> >> INET_FRAG_HASH_DEAD flag? > >> >> > >> > > >> > Yes this is kind of implied in my patch. > >> > I put the smp_store_release() and READ_ONCE exactly to document the > >> > possible races. > >> > This was the reason for my attempt in V1, doing a walk, but Herbert > >> > said walk was not designed for doing deletes. > >> > > >> > Proper synch will need a synchronize_rcu(), and thus a future > >> > conversion in net-next because we can not really > >> > add new synchronize_rcu() calls in an (struct > >> > pernet_operations.)exit() without considerable performance hit of > >> > netns dismantles. > >> > > >> > So this will require a conversion of all inet_frags_exit_net() callers > >> > to .exit_batch() to mitigate the cost. > >> > > >> > I thought of synchronize_rcu_bh() but this beast is going away soon > >> > anyway. > >> > >> But if this patch allows all the same races and corruptions, then > >> what's the point? > > > > Not really. The current races can last dozen of seconds, if youu have > > one million frags. > > > > With the fix, the race is in the order of one usec on typical hosts. > > Ah, I see. A known bug probably worth a comment in the code.
Plan is to use proper synchronize_rcu() in net-next For net tree, I prefer not having to deal with backports hassle since netns dismantles got many changes last 6 months.
