February 23, 2005
View 0bfbf63f83

- changed parse_headers api, HDR_xxx split into HDR_xxx_F (hdr_flags_t) and HDR_xxx_T (hdr_types_t), for more info read doc/parse_headers.txt. The header flags were extended top 64 bits. Be carefull not to mix types with flags, they won't work!

Andrei Pelinescu-Onciul authored on 23/02/2005 17:16:01
February 15, 2005
View 120d03ba6f

Fix a bug in the extract_mediaport routine.

Maxim Sobolev authored on 15/02/2005 18:50:56
View bf142520e8

Use L_ERR log level for reporting errors instead of L_DBG.

Maxim Sobolev authored on 15/02/2005 00:56:05
February 3, 2005
View a11c608742

Improve extract_mediaport() function, so that it takes into consideration transport name (i.e. udp, udptl etc), not media name (i.e. audio, application, etc) when deciding if this stream can be relayed by rtpproxy.

Maxim Sobolev authored on 03/02/2005 17:46:03
View f3cf1ec01c

Add new flag for fix_nated_sdp(), which simply adds nortpproxy flag, usually automatically inserted by functions that force rtp proxy to prevent forcing media session to be relayed more that one rtp proxy in a chain. In some cases this can be useful.

Maxim Sobolev authored on 03/02/2005 14:41:05
View 828db534ff

Reduce diff to Porta's own internal version of the module. NO FUNCTIONAL CHANGES!

Maxim Sobolev authored on 03/02/2005 14:21:11
View 8b12f2c1be

o Add missing space to the warning message;

Maxim Sobolev authored on 03/02/2005 14:08:26
View e585dfe44e

Add support for `application' media type. This makes things like application sharing in Nortel softphone working.

Maxim Sobolev authored on 03/02/2005 14:06:24
View eeb7349395

Initialize mypid module global variable early, so that it is already available for those functions that may need it (i.e. rtpp_test()).

Maxim Sobolev authored on 03/02/2005 13:54:38
View 6be946711d

Don't assume that call-id is already parsed, since it may not be be true in some cases (depending on the order of calling various routines in the routing script), resulting in nathelper functions failing without any reason.

Maxim Sobolev authored on 03/02/2005 13:22:14
February 2, 2005
View e2bba28137

Fix a very bad typo in a comment - option `r' has been called `d' there.

Maxim Sobolev authored on 02/02/2005 18:29:49
December 3, 2004
View a4ff81aac8

- lots of fixes (alignments, casts, warnings) - it compiles cleanly now wit sun cc and intel icc. - slight api change for param_func_t (it takes a void* now instead of a param_func_param_t)

Andrei Pelinescu-Onciul authored on 03/12/2004 19:09:32
November 9, 2004
View f4d1b4fcf0

- force_send_socket per message API support => changedinterface to get_send_socket; it takes now also the sip msg as param (updated the calls to it in all the modules)

Andrei Pelinescu-Onciul authored on 09/11/2004 15:15:11
October 10, 2004
View 9faaea4e26

Changed the way destination is determined for some packets if mhomed is turned on. That's particularly helpful in scenarios in which multiple IPs are used on the same interface and you wish to know in advance which IP will be used without knowledge of IP routing tables. - nathelper allows to bind to a specific IP address; that's a plus if you have multiple IPs on the same interface and wish to guarantee that an IP will be used which is not default route (this overrides mhomed algorithm) - sl and tm send replies always back from socket where request were received (now regardless of mhomed option)

Jiri Kuthan authored on 10/10/2004 21:25:01
October 5, 2004
View 6b884701d8

- nathelper received port test (flag 16 for nat_uac_test): compares the source port of the message with sip port in the first Via and returns true if different (for more info see http://mail.iptel.org/pipermail/serdev/2004-September/002922.html)

Andrei Pelinescu-Onciul authored on 05/10/2004 10:18:53