[mythtv] mythbackend 0.22-fixes problem with duplicate database connection name
roger at beardandsandals.co.uk
Sat Mar 13 12:47:16 UTC 2010
On 08/03/2010 12:58, Roger James wrote:
> On 03/03/2010 13:39, Roger James wrote:
>> On 02/03/2010 21:02, Jim Stichnoth wrote:
>>> On Tue, Mar 2, 2010 at 11:35 AM, Roger James
>>> <roger at beardandsandals.co.uk> wrote:
>>>> I think I will just apply your split of m_connID to my 0.22-fixes
>>>> leaving out the rest of your patch and see what happens. It looks
>>> I think you would be safe applying the entire r23057 changeset
>>> I'm not sure why it was never applied to 0.22-fixes. As an added
>>> benefit, those 5-second freeze-ups will go away.
>> I am a great believer in minimal changes. I will run with the small
>> fix for a week or so to see what happens and then try the full
> Hmm, just had a crash in mythfilldatabase. I have had some crashes
> before, but not no one reported this way. I don't know if this is
> related to the change I just made.
> credits element: not writing empty content for actor at
> /usr/share/perl5/XMLTV.pm line 2110.
> *** glibc detected *** mythfilldatabase: corrupted double-linked list:
> 0x098b50e0 ***
> QMutex::lock: mutex lock failure:
> ======= Backtrace: =========
I just had another crash in mythfilldatabase with exactly the same
*** glibc detected *** mythfilldatabase: corrupted double-linked list: 0x081d6fe0 ***
QMutex::lock: mutex lock failure:
======= Backtrace: =========
======= Memory map: ========
It looks like the m_connID fix appears to have sorted the duplicate
database connections problem, but may have introduced another problem
elsewhere. I will leave it in, as I can live with mythfilldatabase
occasionally falling over!
However, I have just realised that I made another chnage recently and
that was to run the debug version. Maybe these fails were happening all
the time and not being trapped in the non debug version.
More information about the mythtv-dev