November 24, 2006
View c628e43aa0

- usrloc hash fix (performace): switched to a numeric string optimized hash function (which also has very good distribution for strings) - added the new hash to core's hashes.h & changes hashes.h license to openbsd style (so that anybody can use the hash functions without being forced to gpl the code)

Andrei Pelinescu-Onciul authored on 24/11/2006 17:20:08
November 20, 2006
View 609b1abdc0

- tm fix: uri backup/restore print_uac_request bug - tm: new_uri is no longer saved/restored across add_uac(...)/print_uac_request calls since print_uac_request takes care of this internally

Andrei Pelinescu-Onciul authored on 20/11/2006 19:16:49
November 15, 2006
View 2db960ff4a

- get_max_procs() will log a bug and abort() if used too early (e.g. from mod_init())

Andrei Pelinescu-Onciul authored on 15/11/2006 19:57:42
November 4, 2006
View c890787095

- make with -Wall even whem compiling in debug mode, patch from Bogdan Pintea <>

Andrei Pelinescu-Onciul authored on 04/11/2006 16:17:29
View d8b11bbcd5

- tcp fixes: tcpconn_timeout(); expire now timeout; switched to "raw" ticks (the conversion to s was plagued by truncation errors => in some corner cases one could end up with tcp connections that never expire) - if tcp_con_lifetime is set to a negative value, use the maximum possible value instead

Andrei Pelinescu-Onciul authored on 04/11/2006 15:55:08
November 2, 2006
View be7401cc34

- tcp fixes/atomic_t switch

Andrei Pelinescu-Onciul authored on 02/11/2006 22:55:56
October 25, 2006
View f2e1aa5002

- init_childs(PROC_MAIN) moved before starting tcp_main to allow possible tcp usage from module started processes - fork_process() sanity checks & debugging: fail & log a bug if called with make_sock==1 from a process != main or if called after tcp was started; reset is_main for children - fork_tcp_process() sanity checks as above - init_pt(): initialize tcp comm. fds to -1, fix description writing for main/attendant.

Andrei Pelinescu-Onciul authored on 25/10/2006 18:44:36
View 96d091075b

- log messages in children signal handlers off if NO_SIG_DEBUG is defined (safer, but not default for now) - new config param.: exit_timeout (how much time ser will wait for its shutdown to complete, when it expires it will kill everything) - missing timeout added to SIGTERM triggered shutdown

Andrei Pelinescu-Onciul authored on 25/10/2006 16:44:21
October 16, 2006
View 3ed13e7767

- tm: aggregate challenges (WWW/Proxy-Authenticate) from all the 401 & 407 if the final reply is a 401/407 (required by rfc3261). To turn this off and fall back to the old behaviour set tm aggregate_challenges parameter to 0 (see tm docs for more details) - core: header parse support for WWW-Authenticate and Proxy-Authenticate

Andrei Pelinescu-Onciul authored on 16/10/2006 18:44:01
October 13, 2006
View 96227c65df

- added STUN keep-alive functionality in accordance with draft-ietf-behave-rfc3489bis-04.txt

Vladimir Marek authored on 13/10/2006 13:08:42
October 12, 2006
View 9d8edd2980

- dns failover when a 503 reply is received (if the dns name resolves to more ips/srvs try the next one) - if the final reply is a relayed 503 replace it with a 500 (generate a "fake" 500). If the 503 is locally generated, let it pass. - tm_rpc_stats: removed duplicate "waiting" stats

Andrei Pelinescu-Onciul authored on 12/10/2006 19:15:33
October 11, 2006
View 81f72c6ca1

- stop creating new branches (forking)if a 6xx or a CANCEL was received - cancel branch fix: if a branch was canceled stop possible dns failover forking on it - stop retransmission timer fix for cancel for non-invites

Andrei Pelinescu-Onciul authored on 11/10/2006 17:45:50
October 10, 2006
View 65e998d19c

- tm 6xx fixes: when a 6xx is received send a CANCEL on all the branches for which a provisional response was received and then wait for the branches to finish (timeout, 487 from the CANCEL or another final response which "raced" the CANCEL). Branches that did not receive any response (in fact any response >=100) will be "terminated" immediately (fake 487). - tm which_cancel() 0-the cancel bitmap fix - tm which_cancel(), should_cancel(), cancel_branch() enhancements/changes - tm rpc_cancel fixes (it works now)

Andrei Pelinescu-Onciul authored on 10/10/2006 19:21:59
October 6, 2006
View b0cfdf5fa6

- tm better final reply selection: 6xx is preferred over other negative replies; from several 4xx prefer 401, 407, 415, 420, 484 (in this order). For all the other cases, return the lowest code (as before).

Andrei Pelinescu-Onciul authored on 06/10/2006 11:32:41
October 2, 2006
View 144c394f42

- fork_process & fork_tcp_process fixes - reverted to the old fork()-in-parallel behaviour (uncomment FORK_DONT_WAIT for the "serial" fork()).

Andrei Pelinescu-Onciul authored on 02/10/2006 17:29:23