search

yuanjianjiang / webrtc2sip

Automatically exported from code.google.com/p/webrtc2sip
0 stars 0 forks source link

webrtc2sip: double free or corruption (fasttop) #140

Closed GoogleCodeExporter closed 9 years ago

GoogleCodeExporter commented 9 years ago 
What steps will reproduce the problem?
1. Video call(sip) with Polycom -> Asterisk -> webrtc2sip -> sipML5
2. sipML5 ringing, after clicking the answer is webrtc2sip: double free or 
corruption (fasttop)

What is the expected output? What do you see instead?

What version of the product are you using? On what operating system?
CentOS 6    2.6.32-358.23.2.el6.i686
Asterisk 11.6
Polycom RealPresence Group 500

Please provide server logs with DEBUG level equal to INFO

Please provide browser logs

Original issue reported on code.google.com by grigoriy...@gmail.com on 7 Nov 2013 at 12:14

Attachments:

GoogleCodeExporter commented 9 years ago 
if you put a checkbox in the (Cache the media stream), the first few minutes of 
the video call goes well, then at the video there is a delay of 2 seconds or 
more, then the image starts to go bad, and the log entries appear:

*INFO: Receiving RTCP-AFB-REMB (0), exp=4, mantissa=155604, bandwidth = 2431kbps
*INFO: According to rtp-timestamps ...FPS = 30 (clipped to 30) tail_max=60, 
latency_max=30
*INFO: Time to decode frame...but some RTP packets are missing 
(missing_seq_num_start=34028, missing_seq_num_count=1, 
last_seq_num_with_mark=34026). Postpone :(
*INFO: Request re-send(34028)
*INFO: Time to decode frame...but some RTP packets are missing 
(missing_seq_num_start=34028, missing_seq_num_count=1, 
last_seq_num_with_mark=34026). Postpone :(
*INFO: Receiving RTCP-AFB-REMB (0), exp=4, mantissa=155604, bandwidth = 2431kbps
*INFO: Time to decode frame...but some RTP packets are missing 
(missing_seq_num_start=34028, missing_seq_num_count=1, 
last_seq_num_with_mark=34026). Postpone :(
*INFO: frames_count(30)>=latency_max(30)...decoding video frame even if pkts 
are missing :(
*INFO: Receiving RTCP-AFB-REMB (0), exp=4, mantissa=155604, bandwidth = 2431kbps

and

*INFO: frames_count(21)>=latency_max(20)...decoding video frame even if pkts 
are missing :(
*INFO: Receiving RTCP-AFB-REMB (0), exp=4, mantissa=155604, bandwidth = 2431kbps
*INFO: frames_count(21)>=latency_max(20)...decoding video frame even if pkts 
are missing :(
*INFO: frames_count(21)>=latency_max(20)...decoding video frame even if pkts 
are missing :(
*INFO: frames_count(21)>=latency_max(20)...decoding video frame even if pkts 
are missing :(
*INFO: frames_count(21)>=latency_max(20)...decoding video frame even if pkts 
are missing :(
*INFO: frames_count(21)>=latency_max(20)...decoding video frame even if pkts 
are missing :(
*INFO: According to rtp-timestamps ...FPS = 30 (clipped to 30) tail_max=60, 
latency_max=30
*INFO: State machine: tsip_transac_nist_Completed_2_Terminated_X_tirmerJ
*INFO: === NIST terminated ===
*INFO: *** NIST destroyed ***
*INFO: Receiving RTCP-AFB-REMB (0), exp=4, mantissa=155604, bandwidth = 2431kbps
*INFO: Receiving RTCP-AFB-REMB (0), exp=4, mantissa=155604, bandwidth = 2431kbps

Original comment by grigoriy...@gmail.com on 7 Nov 2013 at 12:29

GoogleCodeExporter commented 9 years ago 
Testing with Asterisk -> webrtc2sip -> sipML5 I was getting the same double 
free or corruption (fasttop) and backtrace stopped within tsk_object_delete 
method in tinySAK (doubango). Changing tsk_object_unref to call 
TSK_OBJECT_SAFE_FREE(self) instead of tsk_object_delete(self) seems to have 
fixed the issue for me. Modified source file is attached. See line 199.

Original comment by stephani...@tridsys.com on 6 Dec 2013 at 9:06

Attachments:

GoogleCodeExporter commented 9 years ago 
It causes memleaks. But increases stability :)

Original comment by minkin.a...@gmail.com on 12 Feb 2014 at 10:44

GoogleCodeExporter commented 9 years ago

Original comment by boss...@yahoo.fr on 15 May 2014 at 1:50