Bryan Biedenkapp
|
24ad09c5ee
|
whoops forgot to include defines;
|
3 years ago |
Bryan Biedenkapp
|
3cce6e0b68
|
implement support for TIA-102.AABF-D LC_EXPLICIT_SOURCE_ID;
|
3 years ago |
Bryan Biedenkapp
|
2a3c827cab
|
properly complete grantTG REST API implementation;
|
3 years ago |
Bryan Biedenkapp
|
ca3810fa12
|
report destination ID being rejected when for non-authoratative mode;
|
3 years ago |
Bryan Biedenkapp
|
952606b304
|
revert commit fbf8ec3 (I mistakenly added this with a group of other commits);
|
3 years ago |
Bryan Biedenkapp
|
fbf8ec32c1
|
add support to send raw CSBK or TSBKs from the network;
|
3 years ago |
Bryan Biedenkapp
|
ee36fc5927
|
fix missed constant name change;
|
3 years ago |
Bryan Biedenkapp
|
f6167360d6
|
fix missed constant name change;
|
3 years ago |
Bryan Biedenkapp
|
f42b9f597c
|
whoops forgot header file in last commit;
|
3 years ago |
Bryan Biedenkapp
|
88f67ec7f5
|
cleanup more direct frame queue calls;
|
3 years ago |
Bryan Biedenkapp
|
9dd2fc2ce6
|
replace direct calls for FrameQueue enqueue and flush;
|
3 years ago |
Bryan Biedenkapp
|
d0be614c27
|
split announcement group config into its own config var;
|
3 years ago |
Bryan Biedenkapp
|
7014d3823e
|
complete conference bridge FNE peer communication;
|
3 years ago |
Bryan Biedenkapp
|
27f73c1089
|
cleanup RTP protocal and remove vestigial text tags in most cases;
|
3 years ago |
Bryan Biedenkapp
|
054a9a5da8
|
refactor conference bridge FNE slightly; remove mostly unused functions;
|
3 years ago |
Bryan Biedenkapp
|
f2de006954
|
ensure stream IDs are maintained from source peer to destination peer;
|
3 years ago |
Bryan Biedenkapp
|
e94964ea2f
|
refactor TG hang to only apply to authoritative voice channels; correct handling of non-authoritative permitted destination IDs for NXDN and DMR; move non-authoratative handling into main processNetwork() for P25;
|
3 years ago |
Bryan Biedenkapp
|
bed079f68e
|
implement a fixed internal TG hangtimer for network traffic, this will put a hold on the last received TG for 2 seconds; refactor how P25 handles simultaneous calls from the network and how RF preemption of in-progress network calls when RF wants to transmit on a different TG works;
|
3 years ago |
Bryan Biedenkapp
|
1128d28d97
|
actually AMBT PDUs don't transit the network, so we'll ignore network PDUs for control only mode;
|
3 years ago |
Bryan Biedenkapp
|
c1639729a6
|
correct typo;
|
3 years ago |
Bryan Biedenkapp
|
7a86b4f5b8
|
allow PDUs to be processed for dedicated CCs; enhance W3AXLs code Data.cpp code to check the dedicatedControl flag too;
|
3 years ago |
Patrick W3AXL
|
74838f990a
|
initial re-commit based on latest master (#35)
|
3 years ago |
Bryan Biedenkapp
|
a2865f7052
|
be verbose on VC's about when we send the grant release REST call;
|
3 years ago |
Bryan Biedenkapp
|
6330a4c2a9
|
fix some log messages; ensure VOC is forcibly disabled if control support is disabled;
|
3 years ago |
Bryan Biedenkapp
|
ff66d6bbe2
|
ensure DataHeader's are always passed by reference;
|
3 years ago |
Bryan Biedenkapp
|
9a71eb344f
|
refactor how confirmed data blocks are handled; better handle determining src and dst LLIds during PDU processing; better handle data offset and PDU data length calculations; handle properly acknowledging confirmed packets with A/N flag set;
|
3 years ago |
Bryan Biedenkapp
|
9b43411376
|
when decoding a PDU data block ensure class variables are set to reasonable defaults (since the actual DataBlock class tends to be reused); ensure DataHeader is passed to DataBlock decode() and setFormat() by reference; implement supporton DataHeader to retreive the raw byte contents of a decoded PDU header; ensure the PDU Rx process resets the second header data, irregardless of use; if we are using a PDU second header, ensure its data contents are added to pduUserData as the second header contents are CRC-32'ed along side all the data blocks (infact only the initial header is excluded from CRC-32 all other blocks following initial header *must* be CRC-32'ed with the data blocks); refactor PDU last block detection; ensure writeRF_PDU_Buffered and writeNet_PDU_Buffered regenerate the full packet CRC-32;
|
3 years ago |
Bryan Biedenkapp
|
20882d5100
|
whoops actually fix the PDU confirmed last block problem;
|
3 years ago |
Bryan Biedenkapp
|
798ca22315
|
fix incorrect use of secondary header;
|
3 years ago |
Bryan Biedenkapp
|
8c26de644f
|
reset pdu data length at the end of handling a PDU sequence;
|
3 years ago |
Bryan Biedenkapp
|
4daac87c1a
|
refactor P25 PDU network handling code; better handle PDU confirmed extended addressing;
|
3 years ago |
Bryan Biedenkapp
|
41f590fb0f
|
better handle SIGHUP (reset g_killed to false at the top of the HostMain signal loop);
|
3 years ago |
Bryan Biedenkapp
|
99c908e475
|
fix buffer copy lengths for FNENetwork;
|
3 years ago |
Bryan Biedenkapp
|
31e6769ddc
|
don't network repeat AMBT PDUs; ensure AMBTs aren't processed if control processing is disabled;
|
3 years ago |
Bryan Biedenkapp
|
b4bbca3ca5
|
introduce RF loss watchdog to prevent a situation where a frame loss could cause a deadlock; fix message output for TDULC;
|
3 years ago |
Bryan Biedenkapp
|
c97c3f9412
|
prevent possible deadlock scenario from a failed login sequence, allow RPTL to reset to login sequence if the peer hasn't reached a running state;
|
3 years ago |
Bryan Biedenkapp
|
b1718f3e0a
|
fix payload calculation for activity log; correct order of operations for testing if the RF hangtimer is running;
|
3 years ago |
Bryan Biedenkapp
|
8cd92914ba
|
reword some warning messages to be more verbose; reorder when embedded FNE will transmit parrot frames;
|
3 years ago |
Bryan Biedenkapp
|
91a79cfc24
|
fix issue with using a reference vs copy for the talkgroup rules YAML filename;
|
3 years ago |
Bryan Biedenkapp
|
95e0ed9260
|
be more verbose in logging when the REST service network layer returns errors;
|
3 years ago |
Bryan Biedenkapp
|
ca7eb1a407
|
remove TCP_NODELAY socket option, refactor when SO_LINGER is set on the socket options; correct some file formatting;
|
3 years ago |
Bryan Biedenkapp
|
8c15cb5b50
|
don't attempt a socket tear down, the shutdown of the ioService should clean these up;
|
3 years ago |
Bryan Biedenkapp
|
220ad671ac
|
remove extraneous messages from reconnect loop; m_ioService must be the *LAST* object in the ASIO chain to be destroyed, otherwise weirdness will happen;
|
3 years ago |
Bryan Biedenkapp
|
963a9ab421
|
fix warning message when failing to update status;
|
3 years ago |
Bryan Biedenkapp
|
6573386cd7
|
perform status update 4 times before switching a node into a failed state;
|
3 years ago |
Bryan Biedenkapp
|
a020dd84d0
|
ensure socket processing loop for HTTPClient won't try to process when the the request is considered "completed"; properly set SO_LINGER and TCP_NODELAY;
|
3 years ago |
Bryan Biedenkapp
|
ee74cad81b
|
report network peer ID in status API call; display network peer ID on dvmmon;
|
3 years ago |
Bryan Biedenkapp
|
31d9a3e1c3
|
implement support for "Tab" to cycle the selected and active host;
|
3 years ago |
Bryan Biedenkapp
|
50f85538f8
|
refactor display of usage to be consistent with dvmhost;
|
3 years ago |
Bryan Biedenkapp
|
8c5d9eb2c6
|
add command line option to auto hide the log window at startup;
|
3 years ago |