Pablo Neira Ayuso <pablo@xxxxxxxxxxxxx> wrote: > > I plan to send a patch for nf tree to alter nf_conntrack_confirm() > > to not cache the ct -- I think its a bug too, we should call > > nf_ct_deliver_cached_events() on the ct that is assigned to skb *now*, > > not the old one. > > This is the clash resolution that is triggering this path you describe > in this note. Yes, its the clash resolution.
