What is exactly the purpose of this new functionality? To allow/deny traffic from those IP addresses or for doing/skipping topology hiding?

I am somehow not very pleased with importing a lot of code similar to what permissions does in regards to address matching. I would rather do another approach that might bring more flexibility as well, respectively to run an event route where one can run functions from other modules and set some internal variable in order to control what topoh should do after execution of the event block. This should not be something big in terms of new code, I can do it in the next days if we agree on some concept that fits your needs.

We have the source code, we can change if something helps substantially -- currently, I see two options that can be controlled via a mod param (or just choose one that it is considered better):

* give a fake request, like it is done with other event_route blocks (e.g., htable:mod-init) -- this way should be very fast * parse the outgoing buffer -- might be slightly slower, as it will parse even when not doing topoh

Giving a fake request is very fast, like done at:

* https://github.com/kamailio/kamailio/blob/master/src/modules/htable/htable.c...

It is no parsing apart of the first time. This way is safe, because some functions need the msg structure.

I did an implementation to execute an event_route before the topoh processing for messages to be sent out. Documentation will follow soon, meanwhile you can see the comments of the commit f7d1d23 . Basically, $sndto(ip) should point to the ip address of the destination and calling `drop` should skip topoh processing. I did no test with the new version, so feedback and improvements are welcome.

Thank you @miconda , unfortunately there is bug that I couldnt catch.

6 days ago I have implemented same way in my repository:

https://github.com/adubovikov/kamailio/commit/ae15a717e9a5849b4ac295c5864cd6...

but your and my patches makes same results:

/usr/sbin/kamailio[1260]: ERROR: sl [sl_funcs.c:362]: sl_reply_error(): stateless error reply used: I'm terribly sorry, server error occurred (6/SL) this is after execution of event route. I have tested/removed some code and found that this happens once you call " fmsg = faked_msg_next();"

also in your patch:

https://github.com/kamailio/kamailio/commit/f7d1d234a42198fc10e40c33e92cff5e...

should be: if(_th_eventrt_outgoing<0 **&&** keng==NULL) {

any suggestions got to avoid this SL error ?

Wbr, Alexandr

no, the fake msg is not null, but once you call the faked_msg_next(), after you will always get "I'm terribly sorry, server error occurred (6/SL)", even if you didn't execute "run_top_route".

I will check a bit more, but the issue it can be with pointer for re-transmission.

bingo! :-)

now another issue:

/usr/sbin/kamailio[1798]: ERROR: topoh [topoh_mod.c:560]: th_execute_event_route(): exit due to 'drop' in event route /usr/sbin/kamailio[1798]: : <core> [core/mem/q_malloc.c:469]: qm_free(): BUG: qm_free: bad pointer 0x7fa3461b9000 (out of memory block!) called from core: core/parser/msg_parser.c: reset_new_uri(720) - aborting CRITICAL: <core> [core/pass_fd.c:277]: receive_fd(): EOF on 16

and I don't have anything in core:

(gdb) bt full #0 0x00007f69d1059067 in ?? () No symbol table info available. #1 0x00007f69d105a448 in ?? () No symbol table info available. #2 0x0000000000000020 in ?? () No symbol table info available. #3 0x0000000000000000 in ?? () No symbol table info available. (gdb) quit

So #1155 is fixing also the introduced bug with bad pointer?

Thanks, closing it then!