- Mar 06, 2022
-
-
-
-
After 'from_user' config has been removed from pjsip_endpoint config the user in FROM header was wrong. Fix it with using session->id.number instead of connected_id.number (which is wrong in this case).
-
Don't destroy the connection when the peer sub-channel is in dialing or ringback state.
-
-
There was a possibility that during attended call transfer, deadlock happened in locking bridge_channel and channel. bridge_channel_queue_deferred_frames() was waiting for brcm_indicate() to unlock the channel, while brcm_indicate() which called ast_bridge_channel_queue_frame() was waiting for bridge_channel_queue_deferred_frames() to unlock the bridge. Creating new macro to lock both channel and bridge_channel at the same time and sched_yield() -> retry if failed and use it in bridge_channel_queue_deferred_frames() avoiding the deadlock.
-
-
-
Those two lines are included in brcm_signal_dialtone() which is about to invoke.
-
-
-
Call Status for each line can be get by: ubus call asterisk call_status '{ "line" : X }', where X is line: {0, 1, 2, or 3} callStatus is set in chan_brcm according to TR-104 requirements: Idle: no ongoing calls Dialing: the process of collecting digits for the calling number in an outgoing call, and ringing in outgoing call. Delivered: N/A Connected: a call is established Alerting: ringing when an incoming call is received Disconnected: the remote party has ended a call. but the local party has not gone on-hook.
-
-
-
- asterisk.telephony -> voice.line, "line" -> "id" - asterisk.sip -> voice.sip.client, "line" -> "uri" - asterisk.mwi -> voice.mwi
-
Also remove unused parameter "data":0 in asterisk.telephony ubus events.
-
Also fix some miscellaneous compiling warnings.
-
-
-
-
-
Signed-off-by:
Grzegorz Sluja <grzegorz.sluja@iopsys.eu>
-
When chan_brcm.so module is unloading we are trying to stop the ubus thread. With pthread_cancel() called before pthread_join() we are hanging for a long time since the thread has not set cancellation type. The cancellation type 'deferred' is set by default what means that cancellation is delayed until the thread next call, that takes time. Using pthread_kill() instead of pthread_cancel() makes the signal is sent to the thread immediately causing thread exits/joins suddenly.
-
Signed-off-by:
Grzegorz Sluja <grzegorz.sluja@iopsys.eu>
-
This commit includes 3 fixes: - change place of get/set rtp statistics to have them properly updated in cdr for both incoming and outgoing call - do not create additional "hangup" call log record - save rtp stats counters from endptmngr into brcm_subchannel instead of one global structure to support concurrent calls simultaneously Signed-off-by:
Grzegorz Sluja <grzegorz.sluja@iopsys.eu>
-
-
-
-
-
-
For an unanswered inbound call, the connection has not been created in endptmngr when the call terminates so the RTP statistics are not available.
-
The bug was a regression caused by the commitment 01294e2b which changed connection creation and destroy procedures inproperly in endptmngr.
-
Fix a bug that causes asterisk crash or hang when starting 3 party conference in some cases.The issue was caused by ubus events which has been temporarily disabled by commenting.The feature will be reimplented/fixed in a seperate ticket.
-
The bug was caused by a synchronous ubus API ubus_invoke(). The asynchronous version of API ubus_invoke_aync() shall be used instead. Also remove some unused code and stop ubus thread when the chan_brcm module is unloaded.
-
There is an issue with copying RTP statistics from snapshot to cdr hence call log does not contains RTP stats for incoming calls - for outgoing call it works fine. Signed-off-by:
Grzegorz Sluja <grzegorz.sluja@iopsys.eu>
-
-
-
-
-
-