- 15 May, 2014 22 commits
-
-
Jakub Adam authored
-
Jakub Adam authored
STUN_BINDING requests. Candidate check pair foundation was used - a combination of local and remote candidates' foundation, which does not correspond with [MS-ICE2] 2.2.2.1 0x8054 CANDIDATE-IDENTIFIER
-
Jakub Adam authored
At least temporary, maybe it is not used at all. TURN refresh messages were not seen in the packet capture.
-
Youness Alaoui authored
-
Youness Alaoui authored
-
Youness Alaoui authored
-
Youness Alaoui authored
-
Youness Alaoui authored
-
Youness Alaoui authored
-
Youness Alaoui authored
-
Youness Alaoui authored
-
Youness Alaoui authored
-
Youness Alaoui authored
One issue with tcp-bsd is that it will queue messages when the tcp socket is not writable, but it will also drop messages when the queue is full. If we want to do proper reliable ice-tcp, we need to make sure that messages don't get dropped, also, this would affect http/socks5/pseudossl if their messages get dropped. For ice-tcp, when the socket is not writable we want to return 0, not queue the message. The change here is to allow connchecks and other important protocol messages to be sent as 'reliable' on tcp transports by queuing them and in the various socket layers and to never drop them, but all user messages will be dropped. if the tcp socket is not writable.
-
Youness Alaoui authored
-
Youness Alaoui authored
-
Youness Alaoui authored
-
Youness Alaoui authored
-
Youness Alaoui authored
-
Youness Alaoui authored
-
Youness Alaoui authored
-
Youness Alaoui authored
-
Youness Alaoui authored
-
- 12 May, 2014 1 commit
-
-
Olivier Crête authored
RFC 5389 says: All STUN messages sent over UDP SHOULD be less than the path MTU, if known. If the path MTU is unknown, messages SHOULD be the smaller of 576 bytes and the first-hop MTU for IPv4 [RFC1122] and 1280 bytes for IPv6 [RFC2460]. So sending 65536 bytes is always wrong
-
- 05 May, 2014 2 commits
-
-
Olivier Crête authored
-
Olivier Crête authored
-
- 02 May, 2014 1 commit
-
-
Olivier Crête authored
The farstream unit tests were using it.
-
- 29 Apr, 2014 3 commits
-
-
Olivier Crête authored
-
Olivier Crête authored
-
Olivier Crête authored
-
- 28 Apr, 2014 1 commit
-
-
Youness Alaoui authored
When using optimal path where recv message has only one buffer, the buffer length was set to the size of the buffer, not its length, which was causing the wrong message length to be returned causing the agent to discard STUN messages because their size doesn't match received data length.
-
- 25 Apr, 2014 10 commits
-
-
Philip Withnall authored
-
Philip Withnall authored
This introduces no functional changes, but consolidates and documents the code a bit more.
-
Philip Withnall authored
Use a NULL check to determine if upnp_mapping is an empty list (rather than counting every element). Use g_slist_free_full() where appropriate. This introduces no functional changes.
-
Philip Withnall authored
It’s in milliseconds.
-
Olivier Crête authored
-
Olivier Crête authored
-
Olivier Crête authored
The test for fe80::/10 always returned FALSE.
-
Guillaume Desmottes authored
-
Olivier Crête authored
But keep the turn connection that's currently being used if it is the selected candidate. Also clear the TURN candidate refresh.
-
Olivier Crête authored
Instead of just blindling assuming that all of the pointers are valid until the end.
-