<br><div class="gmail_quote">On Fri, Aug 24, 2012 at 4:59 PM, Bert Haverkamp <span dir="ltr"><<a href="mailto:bert@bertenselena.net" target="_blank">bert@bertenselena.net</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class="im"><p><br>
On Aug 22, 2012 10:52 PM, "Bill Meek" <<a href="mailto:keemllib@gmail.com" target="_blank">keemllib@gmail.com</a>> wrote:<br>
><br>
> On 08/22/2012 02:53 PM, Bert Haverkamp wrote:<br>
> ...<br>
>><br>
>> Core was generated by `mythfrontend.real'.<br>
>> Program terminated with signal 11, Segmentation fault.<br>
>> #0 0x00007f03f67ee0b7 in () from /lib/x86_64-linux-gnu/libpthread.so.0</p>
</div><p>> You can search for the address (/7f03f67ee0b7<cr> in vi) in the same file to</p><div class="im"><br>
> see where the failure is (perhaps see who's calling pthread_join.)<br>
><br>
> At that point, I'd start searching on Trac:<br>
><br>
> <a href="http://code.mythtv.org/trac/search" target="_blank">http://code.mythtv.org/trac/search</a><br>
><br>
> to see if there's already a ticket open, and if not, open one.<br>
><br>
> Good luck,<br>
><br></div>
Im away from home this weekend, so it will be Monday, but just to be sure I understand: I need to open mythfrontend.real in vi? Or do you mean the coredump? Is that humanreadable in any way?<p></p>
<p>Regards,</p>
<p>Bert</p>
</blockquote></div><br>Ah, the joys of ssh;-) I was able to connect to my pvr and try what you described. But I think I must be misunderstanding you. I'm not clear what you want me to use vi on...<div>Instead I used gdb</div>
<div><div>(gdb) info symbol 0x00007f03f67ee0b7</div><div>pthread_join + 39 in section .text of /lib/x86_64-linux-gnu/libpthread.so.0</div><div>(gdb) </div></div><div>But this doesn't tell me where in mythfrontend the problem arises. Or is it a problem in libpthread?</div>
<div><br></div><div>Something else I noticed last week: there is an odd chance the frontend does not crash. It takes 3-5 tries to get it up and running. All the other times it crashes with a segmentation fault as described above.</div>
<div>This also happens when the backend is running. I have a feeling the chances are slimmer in that case, but I havent tried long enough to be able to support that claim..</div><div><br></div><div>Bert</div><div><br></div>