* nta: fixing old resolver bugs
Ignore-this: 6e9bb9dadc0b08e6436655b6d4ea322
- Resolving NAPTR records for SIPS (sf.net #1292657)
- Selecting single transport when resolving
git-svn-id: http://svn.freeswitch.org/svn/freeswitch/trunk@14173 d0543943-73ff-0310-b7d9-9358b9ac24b2
* nta.c: handling of forked transactions getting CANCELed and timing out
Ignore-this: 30a92f1c8721f9d3fa22b9611ac9e7b1
Fixes freeswitch.org bug FSCORE-338.
Thanks for Gabriel Kuri for reporting the problem.
git-svn-id: http://svn.freeswitch.org/svn/freeswitch/trunk@14167 d0543943-73ff-0310-b7d9-9358b9ac24b2
* nta: do not free transaction in reliable_timeout() if caused by failure to send final response
Ignore-this: 8e32f83925b6f2a3629afbb8b20ba693
Issue detected by Coverity.
git-svn-id: http://svn.freeswitch.org/svn/freeswitch/trunk@13322 d0543943-73ff-0310-b7d9-9358b9ac24b2
* nta.c: nta_leg_tag(leg, NULL) now always returns the tag (old or new)
Ignore-this: f5a7d67ed90e2c284f6696d6b5b89326
git-svn-id: http://svn.freeswitch.org/svn/freeswitch/trunk@12739 d0543943-73ff-0310-b7d9-9358b9ac24b2
* nta: fix timer N3
Ignore-this: 4018f3a32723692153389c2109b72296
Try UDP instead of TCP if TCP server does not respond quick enough.
git-svn-id: http://svn.freeswitch.org/svn/freeswitch/trunk@12417 d0543943-73ff-0310-b7d9-9358b9ac24b2
* nta: calculate timer D correctly
Ignore-this: f43d9dcc11fb6a19c3674a6da35ee7ec
Timer D is either 32 seconds or 64 times T1, whichever is larger.
git-svn-id: http://svn.freeswitch.org/svn/freeswitch/trunk@12387 d0543943-73ff-0310-b7d9-9358b9ac24b2
* test_nta.c: timer C now CANCELs INVITE transaction
Ignore-this: 8de2e5747e7ae0e62e2d8ad9aeb0cbed
408 response is not generated (unless CANCEL times out, too.)
git-svn-id: http://svn.freeswitch.org/svn/freeswitch/trunk@12383 d0543943-73ff-0310-b7d9-9358b9ac24b2
* nta: timeout CANCELed INVITE transactions properly
Ignore-this: 4e7fdc56065dba617352443a9310bb28
Use timer D (instead of timer C) to timeout CANCELed INVITE transactions.
Also, generate 408 Request Timeout to all forks that have not received a
final response.
git-svn-id: http://svn.freeswitch.org/svn/freeswitch/trunk@12381 d0543943-73ff-0310-b7d9-9358b9ac24b2
* nta.c: ACK and BYE orphan 200 OK matching with destroyed client transaction
Ignore-this: c4f2b165dd2f197f129044e5c653cd8
ETSI Test cases SIP_CC_OE_CE_V_019, SIP_CC_OE_CE_TI_011_012
git-svn-id: http://svn.freeswitch.org/svn/freeswitch/trunk@11873 d0543943-73ff-0310-b7d9-9358b9ac24b2
* nta: added nta_leg_client_reroute()
nta_leg_client_route() did not catch actually RFC 3261 section 12.1.2 feel
and spirit, not to speak about letter.
git-svn-id: http://svn.freeswitch.org/svn/freeswitch/trunk@11833 d0543943-73ff-0310-b7d9-9358b9ac24b2
* nta.c: when sending ACK to [3456]XX, copy the To header as-is from the response
Problem reported by Paulo Pizarro.
git-svn-id: http://svn.freeswitch.org/svn/freeswitch/trunk@11831 d0543943-73ff-0310-b7d9-9358b9ac24b2
* nta_tag.c: updated tag documentation
NTATAG_GRAYLIST added in 1.12,8, NTATAG_MAX_PROCEEDING() in 1.12.9,
NTATAG_TLS_RPORT() in 1.12.10
git-svn-id: http://svn.freeswitch.org/svn/freeswitch/trunk@11825 d0543943-73ff-0310-b7d9-9358b9ac24b2
* nta.c: do not restart timer E when provisional response is received
According to the (informational) RFC 4321 section 1.2 Timer E is not altered
during the transition to Proceeding.
git-svn-id: http://svn.freeswitch.org/svn/freeswitch/trunk@11808 d0543943-73ff-0310-b7d9-9358b9ac24b2
* nta: NULL host and port in user Via are filled automaticaly
NULL host or port in user-supplied Via header will be filled
automaticaly by NTA, just like branch and rport params.
Added related test case to test_nta_api.c.
git-svn-id: http://svn.freeswitch.org/svn/freeswitch/trunk@10826 d0543943-73ff-0310-b7d9-9358b9ac24b2
Michael Jerris
Anthony Minessale