[mythtv-users] Using two EPG sources - XMLTV for TV and EIT for radio on DVB

[Michael T. Dean]

On 03/13/2012 05:31 PM, Joe Nyland wrote:
> On 13 Mar 2012, at 19:33, Michael T. Dean wrote:
>
>> On 03/13/2012 04:53 AM, Joe Nyland wrote:
>>> On 11 Mar 2012, at 22:29, Joe Nyland wrote:
>>>
>>>> :-( Now I'm getting this in /var/log/messages, when I see
>>>> mythtv-setup crash:
>>>>
>>>> Mar 11 21:53:12 MythTVServer1 kernel: [382500.166847]
>>>> mythtv-setup.re[3222]: segfault at 7fab58e66db8 ip 00007fab699bb897
>>>> sp 00007fab58e66da0 error 6 in libQtCore.so.4.6.2[7fab698f9000+276000]
>>>>
>>>> Also, if I run 'taskset -c 0 /usr/bin/mythtv-setup.real' I get:
>>>> 2012-03-11 21:51:58.394 DTVMux, Warning: Invalid inversion, falling
>>>> back to 'auto'.<-- not sure if this is actually an issue though...
>> That's just informational.  Does it segfault when you run with taskset,
>> too?  What version of MythTV?
>>
>>>> Any suggestions on this?
>>>>
>>> Should this segfault be logged as a bug on Trac?
>>>
>>> If so, for which binary should I provide a backtrace - mythbackend or
>>> mythtv-setup? Can you do a backtrace on mythtv-setup, if it's that
>>> which segfaults?
>> Yes, you'll want a backtrace on mythtv-setup.
>>
>> Do it exactly like you would for mythbackend (
>> http://www.mythtv.org/wiki/Debugging ), but change the command line (and
>> "args" in your gdb script file) as appropriate for mythtv-setup.  Note
>> that if you're using 0.25beta, you will need to change the "-l myth.log"
>> to an appropriate "--logpath /path/to/dir" (we still show 0.24 args on
>> that page).
>>
>> FWIW, the taskset shouldn't be necessary with any recent version of
>> mythtv-setup (at least 0.24+, but I don't remember when that issue was
>> fixed).
> After upgrading my backend to 0.25 I don't this issue when scanning for channels now.
>
> Do you want me to revert my snapshot of my BE, so that I should be able to get this backtrace, or would you rather I left it as it is?

If it's fixed in master/0.25beta, that's good enough for me.  Any ticket 
reported against 0.24, only, is likely to be closed, anyway, at this point.

Thanks for testing master and for the followup.

Mike