[mythtv-users] Finding cause of segmentation fault

[Roger Siddons]

On Fri, 24 Apr 2015 22:01:58 +0100, dennis <deg at outlook.com> wrote:

>
>
> On 04/24/2015 03:17 PM, Stefan D wrote:
>> I've had some problems with audio sync and stuttering tv playback the 
>> last few days so i thought i would try the latest master.
>> But since i was running a pretty old Ubuntu i found it tedious to tryto  
>> install Qt5, so i decided to upgrade my Ubuntu to 14.04 LTS, wasrunning  
>> 12.04 LTS before.
>>
>> I know that i would get in some trouble to get everything runningagain,  
>> but this frontend segmentation fault was a surprise. Don't havea clue  
>> how to find the problem. I know it is master so maybe it is notin a  
>> working condition? (but it usually is) Or there is some otherthing that  
>> got left out during the upgrade that makes the frontendcrash.
>>
>> So how could i find out why i get a segmentation fault from the 
>> frontend? What debug level should i use so you guys would be able to 
>> help me?
>>
>> I do see in the backend log that the frontend connects.
>
>
> I seen similar issues when I first started building the master branch, 
> seg fault in about the same spot. My issue was related to havingmultiple  
> versions of libraries installed and I was picking up the wronglibraries.  
> Do you still have the old libraries installed anywhere?
>
> I did a build a on the 21st with a pull from master and its working fine 
> for me. You might try adding "--loglevel debug" to see if it gives you 
> additional information.
>
> Dennis
> _______________________________________________


And clear out your old plugins. It looks like it's detecting an old  
MythBrowser.

Logging isn't that helpful for segfaults. Easiest way is to run the  
frontend in a debugger to pinpoint the problem

https://www.mythtv.org/wiki/Debugging#Getting_a_Backtrace
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mythtv.org/pipermail/mythtv-users/attachments/20150424/94336016/attachment.html>