Re: pf: percpu anchor stacks

2017-05-19 Thread Alexander Bluhm
On Fri, May 19, 2017 at 11:47:21PM +0200, Alexandr Nedvedicky wrote: > would you be able to try patch below to check if it will fix pf_forward > failures? Yes, this fixes it. OK bluhm@ > thanks a lot > and sorry for inconveniences Thanks for the quick fix. And there was no inconvenience, I

Re: pf: percpu anchor stacks

2017-05-19 Thread Alexandr Nedvedicky
Hello, would you be able to try patch below to check if it will fix pf_forward failures? thanks a lot and sorry for inconveniences regards sasha 8<---8<---8<--8< diff -r eb40d8d52679 src/sys/net/pf.c --- a/src/sys/net/pf.c Fri May 19

Re: pf: percpu anchor stacks

2017-05-19 Thread Alexandr Nedvedicky
Hello, On Fri, May 19, 2017 at 06:10:54PM +0200, Alexander Bluhm wrote: > On Mon, May 15, 2017 at 03:19:19PM +0200, Alexandr Nedvedicky wrote: > > I'm attaching updated final patch, which accepts your suggestion. > > I think this broke sys/net/pf_forward. >

Re: pf: percpu anchor stacks

2017-05-19 Thread Alexander Bluhm
On Mon, May 15, 2017 at 03:19:19PM +0200, Alexandr Nedvedicky wrote: > I'm attaching updated final patch, which accepts your suggestion. I think this broke sys/net/pf_forward. http://bluhm.genua.de/regress/results/regress.html When backing out pf.c rev 1.1024 it works again. I guess it is a

Re: pf: percpu anchor stacks

2017-05-15 Thread Mike Belopuhov
On Mon, May 15, 2017 at 15:19 +0200, Alexandr Nedvedicky wrote: > Hello, > > > Now *is* the time to commit the first step, the refactoring. Once > > that's done we can discuss the introduction of the context. > > > > Could you come up with such diff? > > first of all: I have not managed to

Re: pf: percpu anchor stacks

2017-05-15 Thread Alexandr Nedvedicky
Hello, > Now *is* the time to commit the first step, the refactoring. Once > that's done we can discuss the introduction of the context. > > Could you come up with such diff? first of all: I have not managed to finish the re-factoring step yet, work is still in progress. I stole some

Re: pf: percpu anchor stacks

2017-05-08 Thread Martin Pieuchot
On 28/03/17(Tue) 13:02, Alexandr Nedvedicky wrote: > [...] > > > > - s/test_status/action/ as it's done everywhere else? > > I've opted to test_status, because it's something different to 'action' > as we use it in current code. I agree with you for test_status. What about naming the

Re: pf: percpu anchor stacks

2017-03-28 Thread Alexandr Nedvedicky
Hello Mike, thank you for looking at my patch. I accept most of your comments. I believe the items below deserve further discussion. > - instead of checking "rv" against 0 in the "break on quick >rule or failure" I'd like to see an actual check against >PF_TEST_* values so that it's

Re: pf: percpu anchor stacks

2017-03-27 Thread Mike Belopuhov
On Fri, Mar 24, 2017 at 12:19 +0100, Alexandr Nedvedicky wrote: > Hello, > > I'm attaching patch, which removes stack-as-a-global variable. > it's updated patch [1] to current tree. > > sorry for being pushy advocating my old, rusty patch. > I think your diff is the way to go indeed. If we can

Re: pf: percpu anchor stacks

2017-03-24 Thread Alexandr Nedvedicky
Hello, I'm attaching patch, which removes stack-as-a-global variable. it's updated patch [1] to current tree. sorry for being pushy advocating my old, rusty patch. thanks and regards sasha [1]

Re: pf: percpu anchor stacks

2017-03-19 Thread Alexandr Nedvedicky
Hello, I've sent different patch [1], which was touching same functions some time ago. The old patch [1] basically splits pf_test_rule() to two functions: pf_test_rule() pf_match_rule(), which walks anchor stack recursively. the recursion depth is limited to 64. the memory foot