I ran gdb on your provided example and the exact same thing happens as in 
wireshark_connector. So, I’ve rolled this up into a bug report since we can 
replicate it with baseline code (pdu_to_tagged_stream). I’m yet to be convinced 
this is application specific as it seems to be an issue with gnuradio-runtime.

http://gnuradio.org/redmine/issues/797 <http://gnuradio.org/redmine/issues/797>

PWG
> On Jun 4, 2015, at 9:20 AM, Paul Garver <[email protected]> wrote:
> 
> Hi Bastian,
> 
> Thanks for the response and the simplified example. Maybe someone who 
> understands the scheduler a bit better can chime in and let us know the scope 
> of the issue here.  I’m happy to dig into the details but it would be good to 
> know if this is a fundamental issue with the scheduler or a simple fix.  I 
> thought based on [1] this was an issue with python message-only blocks, but 
> everything here should be c++. Like you, I’ve previously implemented some 
> work-arounds such as having the file pointer location in file source 
> accessible.  But it seems this issue is not application specific and thus it 
> sure would be nice to have the fix in the baseline vs. running custom file 
> source blocks, forks of GR, etc.
> 
> [1] https://lists.gnu.org/archive/html/discuss-gnuradio/2015-04/msg00093.html 
> <https://lists.gnu.org/archive/html/discuss-gnuradio/2015-04/msg00093.html>
> 
> PWG
> 
>> On Jun 4, 2015, at 7:47 AM, Bastian Bloessl <[email protected] 
>> <mailto:[email protected]>> wrote:
>> 
>> I tried to create a minimal example of the problem:
>> 
>> http://www.ccs-labs.org/~bloessl/not_stopping.grc 
>> <http://www.ccs-labs.org/~bloessl/not_stopping.grc>
> 

_______________________________________________
Discuss-gnuradio mailing list
[email protected]
https://lists.gnu.org/mailman/listinfo/discuss-gnuradio

Reply via email to