Summary:	ASTERISK-09793: Asterisk Crashes with: ERROR[13628] /usr/src/asterisk-1.4.5/include/asterisk/lock.h: channel.c line 2522 (ast_indicate_data):
Reporter:	furiousgeorge (furiousgeorge)	Labels:
Date Opened:	2007-07-02 17:25:02	Date Closed:	2007-07-06 20:53:03
Priority:	Critical	Regression?	No
Status:	Closed/Complete	Components:	Addons/General
Versions:		Frequency of Occurrence
Related Issues:
Environment:		Attachments:
Description:	I was experiencing deadlocks with asterisk 1.4.4, so on good advice I upgraded to 1.4.5.  This week's crash, unlike in the past, resulted in asterisk completely dieing.  Usually, I could still connect to the CLI but it would misbehave, and asterisk would have to be kill -9'ed, and all the usual deadlock symptoms.  Not this time. Complicating matters, despite having built asterisk with DONT_OPTIMIZE and DEBUG_THREADS I see no core dumped.   The last two lines of my CLI read as follows: [Jul  2 17:33:19] ERROR[13628] /usr/src/asterisk-1.4.5/include/asterisk/lock.h: channel.c line 2522 (ast_indicate_data): Error obtaining mutex: No such file or directory [Jul  2 17:33:19] ERROR[13628] /usr/src/asterisk-1.4.5/include/asterisk/lock.h: channel.c line 2525 (ast_indicate_data): attempted unlock mutex '&chan->lock' without owning it! ****** ADDITIONAL INFORMATION ****** I've planted a seed in my user's head that our phones' (Snom 360) "ParkOrbit" feature may be the cause of this.  Today they told me that the last three times it happened, it was directly after parking a call in this fashion. By assigning one of my hotkeys to "ParkOrbit" the phones will complete the functional equivalent of a blind transfer to that extension (thereby parking the call).  Users are content to do this rather than perform a transfer to park, and use the BLF on chan_local in version 1.4.X to indicate where the call was parked. I notice that I get a Maximum Retries Exceeded error every time we park in this way.  Looking at the SIP debug output (which I barely comprehend at all), I see that for some reason the phone is sending a 487 cancellation request after the call has been parked. I don't know if this has anything to do with anything.
Comments:	By: Jason Parker (jparker) 2007-07-03 13:07:24 Can you post a backtrace of this, against 1.4.6? By: Joshua C. Colp (jcolp) 2007-07-05 08:22:11 I would also like to see the sip debug and console output from this ParkOrbit thing.