[Zgemma H7] Fairly frequent crash when no free tuner and coming out of watching a recording

[smipx]

What type of support thread are you creating?

Possible bug

What OpenViX Image build number are you using?

6.7.020

Have you tried re-flashing WITHOUT settings restore?

NO

Have you tried re-flashing WITH a settings restore?

NO

Hi all,

I have a fairly common scenario (for me) and I get a crash once a week or so. I have 2x s2 and 1xT2 tuners. At a point in time there are no active recordings taking place and I am on a T2 channel watching live TV (e.g. BBC1 HD). Go into Movie List and play a file (say an hour or so long). In the meantime a recording starts on the T2 tuner on another channel (e.g. ITV HD).

When I come out of my recording and the system goes back to Live TV I get a crash and the box reboots. I would expect it to just give a warning that no T2 tuner is available and simple show a black screen until I change to a channel with a free tuner (e.g. one of the S2 tuners) or switch to the recording T2 channel - but it doesn't - it crashes.

I have attached the latest crash log in the hope that it will help the people "in the know" create a fix.

This has been going on since at least 6.7.xxx - now on 6.7.020 and was recently on 6.7.011

Many thanks
Paul

22:06:17.2285 [EventInfo] gotEvent, type: 'next', what: 'evStart', is timed repeat: False
22:06:17.2311 [Notifications] RemovePopup, id = ZapError
22:06:17.2312 [eDVBResourceManager] allocate channel.. 07fe:0002
22:06:17.2312 [eDVBResourceManager] available channel.. 07dc:0002
22:06:17.2312 [eDVBResourceManager] available channel.. 07e6:0002
22:06:17.2313 [eDVBServicePMTHandler] allocate Channel: res -6
22:06:17.2313 [eDVBServicePlay] DVB service failed to tune - error 0
22:06:17.2797 Traceback (most recent call last):
22:06:17.2799   File "/usr/lib/enigma2/python/StartEnigma.py", line 67, in processDelay
22:06:17.2831     self.current_dialog.doClose()
22:06:17.2835   File "/usr/lib/enigma2/python/Screens/Screen.py", line 108, in doClose
22:06:17.2875   File "/usr/lib/enigma2/python/Screens/InfoBar.py", line 323, in __onClose
22:06:17.2879   File "/usr/lib/enigma2/python/Components/Sources/CurrentService.py", line 61, in newService
22:06:17.2890   File "/usr/lib/enigma2/python/Components/Element.py", line 92, in changed
22:06:17.2893   File "/usr/lib/enigma2/python/Tools/CList.py", line 7, in __call__
22:06:17.2900   File "/usr/lib/enigma2/python/Components/Converter/ServiceName.py", line 127, in changed
22:06:17.2907   File "/usr/lib/enigma2/python/Components/Element.py", line 92, in changed
22:06:17.2911   File "/usr/lib/enigma2/python/Tools/CList.py", line 7, in __call__
22:06:17.2914   File "/usr/lib/enigma2/python/Components/Renderer/Picon.py", line 134, in changed
22:06:17.2922   File "/usr/lib/enigma2/python/Components/Renderer/Picon.py", line 91, in getPiconName
22:06:17.2924   File "/usr/lib/enigma2/python/enigma.py", line 256, in __init__
22:06:17.2928 ValueError: invalid null reference in method 'new_eServiceReference', argument 1 of type 'std::string const &'
22:06:17.2929 [ePyObject] (CallObject(<bound method Session.processDelay of <__main__.Session object at 0xb0c1a5a0>>,()) failed)

 

[Huevos]

Attachment goes in "/usr/lib/enigma2/python/Components/Renderer"... switch on debug logs... then restart.

Try to simulate the crash again.

Post the debug log.

Thanks.

 

[smipx]

Will do that and get back to you when I can get a crash. Thanks. Paul

 

[smipx]

OK here are some logs. What I did was setup 2 x s2 recordings on different transponders (to tie them up) and 1x T2 recording on a certain MUX (all for 13:45 with 2 minute padding). I did this at 13:30 and at the same time switched to another T2 channel on a differnet MUX which was fine as the recording has not started yet.

I then went into the movielist and started to play an old recording from some days ago. Let it run until 13:45 and then stopped the days old recording I was watching.

System went back to movielist with a black PIG .... so far so normal and expected.

Then I backed out of the movielist to normal TV (which at this point is on the 2nd T2 channel - remembering that the tuner is taken up with one of the recordings on T2 on a different MUX).

Then it crashed.

Hope that is of use?

Paul

 

[Huevos]

Can you try again with this attachment please.

 

[smipx]

Hi Huevos,

new py file copied and similar recording and steps as last crash. Crash time was around 22:30hrs.

Thanks
Paul

 

[twol]

Hi Huevos,

new py file copied and similar recording and steps as last crash. Crash time was around 22:30hrs.

Thanks
Paul

this has same debug lines which were removed from picon.py in post #45
So can you try again ........ delete old Picon .py in /usr/lib/enigma2/python/Components/Renderer and copy new one across, restart and try again !

 

[Joe_90]

... and don't forget to delete the newer .py file once you have restarted (which will create a .pyc file). Otherwise, the same re-creation of the .pyc file will happen on each boot.

 

[smipx]

Doh! Forgot to reboot.
Will simulate a crash later and post resuts.

Sorry!

Paul

 

[smipx]

Hi, Did the same thing again with the correct new new py file and I could not make it crash. When I exited to movielist screen instead of a crash I just got a big black screen and from there I was able to go navigate the menus, select one of the recording channels or open the movielist to choose another recording to watch. No crashes.

I'm assuming the latest py file included a fix to what the problem was. Please confirm this is the case as otherwise I did not set up my recordings properly to make the crash happen and I will have to try again.

Thanks
Paul

 

[twol]

Hi, Did the same thing again with the correct new new py file and I could not make it crash. When I exited to movielist screen instead of a crash I just got a big black screen and from there I was able to go navigate the menus, select one of the recording channels or open the movielist to choose another recording to watch. No crashes.

I'm assuming the latest py file included a fix to what the problem was. Please confirm this is the case as otherwise I did not set up my recordings properly to make the crash happen and I will have to try again.

Thanks
Paul

Huevos added a fix…. So thats a good outcome

 

[smipx]

Perfect. Thanks to Huevos & Twol!!

Cheers, Paul

 

[adennisa15]

Apologies for a sort of meaningless "me too" post, but I updated to released build 6.7.020 on Sunday and since then I've had three of these crashes.

I had a look in the logs and they're all this NPE/ValueError. A well known search engine led me here.

I've loaded on the modified Picon.py from comment #2 / https://github.com/OpenViX/enigma2/commit/2ffac2285f4fb5ba29c57716b73d66b164c3e551 (and removed the .pyc, rebooted and confirmed a new, larger, .pyc has appeared).

Fingers crossed, that's the issue fixed , it looks as though it should be.

Thank the deities for crash logs, a working search engine, open source software and guys like @Huevos ! Thanks.

That is all. As you were...