pf_state_key_link_reverse() needs atomic ops -- resending

2020-04-03 Thread Alexandr Nedvedicky
Hello, my apologize to resend the same diff [1]. I'm not sure I got OK or not. It can be the case the privately received OK got lost. The change is required to allow multiple instances of pf_test() running concurrently. Without this change in, PF trips 'KASSERT(sk->reverse == NULL);' thanks and

Re: pf_state_key_link_reverse() needs atomic ops

2019-06-11 Thread Alexandr Nedvedicky
Hello, > > > > > > yes that's correct. the patch above comes from my private branch [1]. > > > mpi@ pointed out in off-line email exchange the patch unlocks local > > > inbound > > > packets too, which is coming bit early. However for forwarding case > > > things > > > seem to

Re: pf_state_key_link_reverse() needs atomic ops

2019-06-10 Thread Jonathan Matthew
On Mon, Jun 10, 2019 at 11:46:55AM -0300, Martin Pieuchot wrote: > On 10/06/19(Mon) 09:29, Alexandr Nedvedicky wrote: > > Hello, > > > > sorry for extra delay (was off-line over the weekend). > > > > On Sat, Jun 08, 2019 at 09:46:24PM +1000, Jonathan Matthew wrote: > > > On Tue, Jun 04, 2019 at 0

Re: pf_state_key_link_reverse() needs atomic ops

2019-06-10 Thread Martin Pieuchot
On 10/06/19(Mon) 09:29, Alexandr Nedvedicky wrote: > Hello, > > sorry for extra delay (was off-line over the weekend). > > On Sat, Jun 08, 2019 at 09:46:24PM +1000, Jonathan Matthew wrote: > > On Tue, Jun 04, 2019 at 01:50:51AM +0200, Alexandr Nedvedicky wrote: > > > Hello, > > > > > > I've mana

Re: pf_state_key_link_reverse() needs atomic ops

2019-06-10 Thread Alexandr Nedvedicky
Hello, sorry for extra delay (was off-line over the weekend). On Sat, Jun 08, 2019 at 09:46:24PM +1000, Jonathan Matthew wrote: > On Tue, Jun 04, 2019 at 01:50:51AM +0200, Alexandr Nedvedicky wrote: > > Hello, > > > > I've managed to get pf_test() running in parallel on forwarding path in my > >

Re: pf_state_key_link_reverse() needs atomic ops

2019-06-08 Thread Jonathan Matthew
On Tue, Jun 04, 2019 at 01:50:51AM +0200, Alexandr Nedvedicky wrote: > Hello, > > I've managed to get pf_test() running in parallel on forwarding path in my > experimental tree. And there was some fall out. PF died on ASSERT() in > pf_state_key_link_reverse() at line 7371: > > 7368 pf_state_k

pf_state_key_link_reverse() needs atomic ops

2019-06-03 Thread Alexandr Nedvedicky
Hello, I've managed to get pf_test() running in parallel on forwarding path in my experimental tree. And there was some fall out. PF died on ASSERT() in pf_state_key_link_reverse() at line 7371: 7368 pf_state_key_link_reverse(struct pf_state_key *sk, ...) 7369 { 7370 /* Note t