Re: [Live-devel] openRTSP segfault on max osx

2010-06-22 Thread Ross Finlayson
running openRTSP -q -w 704 -h 576 
rtsp://192.168.0.101/axis-media/media.amp test.mov  
and sending kill -HUP sometimes ends with a segfault. This is with 
live555 2010.06.18 but I didn't see this with 2010.04.09


Aha!  You've found a bug in the openRTSP code.  We're closing the 
output file(s) too soon.  (This bug turned out to be a problem only 
with the new RTSPClient implementation, because - in this new 
implementation - we go to the event loop (potentially handling 
incoming RTP data) while awaiting the response to the RTSP TEARDOWN 
command.)


Please try the following, which should fix this problem:

In testProgs/playCommon.cpp, line 1104, move the statement
closeMediaSinks();
down 5 lines, so that it now takes place after
Medium::close(session);
--

Ross Finlayson
Live Networks, Inc.
http://www.live555.com/
___
live-devel mailing list
live-devel@lists.live555.com
http://lists.live555.com/mailman/listinfo/live-devel


Re: [Live-devel] openRTSP segfault on max osx

2010-06-22 Thread Ross Finlayson

Please try the following, which should fix this problem:

In testProgs/playCommon.cpp, line 1104, move the statement
closeMediaSinks();
down 5 lines, so that it now takes place after
Medium::close(session);


It turns out that I didn't have this quite right.  The correct 
sequence of statements - starting at line 1104 - should be:


tearDownStreams();
closeMediaSinks();
Medium::close(session);

I have now installed a new version (2010.06.22) of the code that 
fixes this bug (as well as Sebastien Escudier's latest RTSPClient 
bug).

--

Ross Finlayson
Live Networks, Inc.
http://www.live555.com/
___
live-devel mailing list
live-devel@lists.live555.com
http://lists.live555.com/mailman/listinfo/live-devel


Re: [Live-devel] openRTSP segfault on max osx

2010-06-22 Thread Kenneth Ljungh
22 jun 2010 kl. 11.26 skrev Ross Finlayson:

 Please try the following, which should fix this problem:
 
 In testProgs/playCommon.cpp, line 1104, move the statement
  closeMediaSinks();
 down 5 lines, so that it now takes place after
  Medium::close(session);
 
 It turns out that I didn't have this quite right.  The correct sequence of 
 statements - starting at line 1104 - should be:
 
   tearDownStreams();
   closeMediaSinks();
   Medium::close(session);
 
 I have now installed a new version (2010.06.22) of the code that fixes this 
 bug (as well as Sebastien Escudier's latest RTSPClient bug).
 -- 
 
 Ross Finlayson
 Live Networks, Inc.
 http://www.live555.com/
 ___
 live-devel mailing list
 live-devel@lists.live555.com
 http://lists.live555.com/mailman/listinfo/live-devel

Works perfectly. Thanks for your quick fix and effort, it's great software.

Kenneth


___
live-devel mailing list
live-devel@lists.live555.com
http://lists.live555.com/mailman/listinfo/live-devel