Summary:	ASTERISK-00681: [patch] hangup during initiation of 3-way call
Reporter:	derek (derek)	Labels:
Date Opened:	2003-12-18 16:26:14.000-0600	Date Closed:	2004-09-25 02:52:20
Priority:	Blocker	Regression?	No
Status:	Closed/Complete	Components:	Core/General
Versions:		Frequency of Occurrence
Related Issues:
Environment:		Attachments:	( 0) chan_zap.c.patch
Description:	During the initiation of a 3-way call if the person making the 3-way call hangs up their phone with another person on hold then asterisk endlessly phones back that person until the asterisk process is killed with a kill -9. Here are the step to reproduce the problem: 1) I make an outbound call 2) Called person answers phone 3) I hit the "flashhook" to initiate a 3-way call 4) I hear dial tone and called person is on hold 5) I hang up my phone 6) called person hangs up their phone 7) my phone starts ringing 8) I answer and no one is there, I hang up 9) endless loop between step 7 & 8 happens after this happens and this endless ringing loop begins asterisk cannot be stopped from within the console but must be killed with kill -9.
Comments:	By: Brian West (bkw918) 2003-12-19 10:55:24.000-0600 Are you using the latests cvs and what distro and card config? By: Brian West (bkw918) 2003-12-19 11:05:55.000-0600 I see what you are doing.  Use call callparking if you want to put a call on hold.  Granted this is a condition that needs to be fixed.  Flash Hook is not hold.  Use call parking. By: Brian West (bkw918) 2003-12-19 11:14:34.000-0600 What signalling are you using because we can't reproduce this in the digium lab. By: derek (derek) 2003-12-19 12:48:26.000-0600 Yes, we are using the latest cvs with Redhat Linux 9.0 and an Wildcard TE410P. By: derek (derek) 2003-12-19 12:49:36.000-0600 We don't want to put a call on hold, we want to do a 3-way call.  It just so happens that when you have someone on the line and want to do a three-way call that when you push the flash button it puts them on hold. By: Brian West (bkw918) 2003-12-19 12:51:06.000-0600 ok angler and I tried to get this to happen and we couldn't.... we shall try again. By: derek (derek) 2003-12-19 12:56:05.000-0600 We are using FXS signalling with an Adtran channel bank and analog phones. By: Brian West (bkw918) 2003-12-19 14:35:50.000-0600 ks,ls or gs FXS signalling? By: derek (derek) 2003-12-19 15:55:36.000-0600 we are using ks FXS By: Brian West (bkw918) 2003-12-19 16:53:36.000-0600 Have you tried gs or ls? By: Brian West (bkw918) 2004-01-07 00:14:01.000-0600 derek any input on this? By: Brian West (bkw918) 2004-01-10 21:24:38.000-0600 hey see if ASTERISK-576 fixes this for you. By: derek (derek) 2004-01-13 16:21:40.000-0600 thanks brian, we'll test that patch and see if it helps.  We actually tried duplicating the bug with both gs & ls and were able to duplicate crash with both signalling types.  However, we did patch chan_zap.c and will upload a diff shortly. By: derek (derek) 2004-01-13 16:41:25.000-0600 just uploaded a patch that fixed the problem on our end, chan_zap.c.patch. hopefully this will give more insight into the problem. By: Mark Spencer (markster) 2004-02-27 01:58:01.000-0600 I can't duplcaite this problem either without the patch.  Can i login to your system to look and see it happen?  Is there a particular mix of zap and non-zap channels you're using? By: heison (heison) 2004-03-21 17:18:44.000-0600 derek: is this still a problem? By: Mark Spencer (markster) 2004-03-31 03:21:38.000-0600 I'm resolving this as "unable to duplicate" until we hear from the person.