Patrick McHardy wrote:
> Eric Dumazet wrote:
>> IPS_CONFIRMED_BIT is set under nf_conntrack_lock (in 
>> __nf_conntrack_confirm()),
>> we probably want to add a synchronisation under ct->lock as well,
>> or __nf_ct_refresh_acct() could set ct->timeout.expires to extra_jiffies,
>> while a different cpu could confirm the conntrack.
> 
> Before the conntrack is confirmed, it is exclusively handled by a
> single CPU. I agree that we need to make sure the IPS_CONFIRMED_BIT
> is visible before we add the conntrack to the hash table since the
> lookup is lockless, but simply moving the set_bit before the hash
> insertion should be fine I think.
> 

A slightly changed version which moves hash insertion to the end
and adds a comment about ordering. This make sure the timer is
actually running before the conntrack can be found be other CPUs.