[mythtv] backtrack

[Chad J McQuinn]

On Feb 25, 2004, at 7:39 PM, Isaac Richards wrote:

> He cut off the most important part of the backtrace, the beginning.
>
> For example:
>
> Program received signal SIGSEGV, Segmentation fault.
> [Switching to Thread 94948272 (LWP 7003)]
> 0x080817f1 in Scheduler::getAllPending(QStringList*) (this=0x0,
>     strList=0x5a8c8b0) at scheduler.cpp:575
> 575         QMutexLocker lockit(recordingList_lock);
> (gdb)

Hmmm..you're right, I did. I would not have done that on purpose so 
either I messed up or else I don't have the buffer for 'screen' set 
large enough. I don't have email on my myth box so I have to grab the 
output on another machine. I'll bump up the buffer size and make sure I 
get it all the next time it happens.


> I have observed the crash
>> he described (and posted the backtrack for).  It seems to happen with
>> quite high frequency  shortly after starting recording (sometimes
>> immediately, sometimes up to a couple of minutes after starting the
>> recordings).
>
> Which sounds completely different than the problem the original poster
> described.

Soudns like the same thing I reported; it happens after recordings 
start. Sometimes almost immediately, sometimes a few minutes in, and in 
one case about 45 minutes in.

One thing I have figured out is that it *may* be related to my slave 
backend. I have shut it down as of a day or so ago, so only the master 
backend is running, and I have not had a segfault since through several 
recordings. If I get a few more days of uptime then I can be reasonably 
sure it is related to having a slave backend running, since I have had 
that kind of uptime with this CVS build. If that's the case I'll see 
what I can find out.

-Chad