PDA

View Full Version : [ViX_Misc] Software crashes when recording on any channel


Red Alert
22-05-20, 17:51
Just downloaded latest VIX image and latest Edision software. There seems to be a bug everytime i try to record any channel on satellite. The receiver crashes then restarts. Here is the log.

< 2915.8320> Traceback (most recent call last):
< 2915.8321> File "/usr/lib/enigma2/python/mytest.py", line 244, in processDelay
< 2915.8324> callback(*retval)
< 2915.8324> File "/usr/lib/enigma2/python/Screens/InfoBarGenerics.py", line 3131, in recordQuestionCallback
< 2915.8325> File "/usr/lib/enigma2/python/Screens/InfoBarGenerics.py", line 3043, in startInstantRecording
< 2915.8326> File "/usr/lib/enigma2/python/RecordTimer.py", line 191, in __init__
< 2915.8327> AttributeError: 'iPlayableServicePtr' object has no attribute 'isRecordable'
< 2915.8328> [ePyObject] (CallObject(<bound method Session.processDelay of <__main__.Session instance at 0xb188b490>>,()) failed)


Anyone know why this is happening?
Andy_Hazza
22-05-20, 18:03
What version of ViX?


Sent from my iPhone using Tapatalk
Red Alert
22-05-20, 18:07
version 5.3.032

I forgot to say this happens when you press "Add recording (indefinitely)" on any channel
ccs
22-05-20, 18:24
I can't reproduce this on freeview 5.3.032 with plenty of free tuners.

With or without test version of /usr/lib/enigma2/python/ServiceReference.py in use.
Red Alert
22-05-20, 18:29
I can't reproduce this on freeview 5.3.032 with plenty of free tuners.

Also test version of /usr/lib/enigma2/python/ServiceReference.py in use.

Did you try recording using the "Add recording (indefinitely)" option

If its a timer recording then it does not crash. so strange
ccs
22-05-20, 18:30
Did you try recording using the "Add recording (indefinitely)" option

If its a timer recording then it does not crash. so strangeYes, although I never have before! :)
Red Alert
22-05-20, 18:34
Could this be related to Edision Mio firmware rather than Vix image?
ccs
22-05-20, 18:39
Could this be related to Edision Mio firmware rather than Vix image?
No idea, did this recording sequence work in earlier versions of ViX?
Red Alert
22-05-20, 18:49
No idea, did this recording sequence work in earlier versions of ViX?

Yes, On some channels i recorded this way.
bbbuk
22-05-20, 19:06
online updated to 5.3.32 and tried several recording (indefinitely) and no crash neither.

Tried a 1: and 5002: service ref.

5002: service ref gave information message that recording isn't supported. This is expected and didn't crash. 1: worked fine.


Update:
For clarification, you tried recording (indefinitely) by pressing the dedicated Rec button on remote when watching live TV?
Red Alert
22-05-20, 19:30
Yes, thats what I do. I press the record button then select the recording option. But it dont crash if it starts on timer. Could the EPG cause this issue?
adm
22-05-20, 22:46
I can't reproduce this on freeview 5.3.032 with plenty of free tuners.

With or without test version of /usr/lib/enigma2/python/ServiceReference.py in use.

no problem seen here with satelitte Ver .32, Xtrend 10K. While watch live TV pressing record and selecting Add recording (indefinitely)
ccs
23-05-20, 10:50
I wonder if it's anything to do with timeshift?
spanner123
23-05-20, 16:39
I just tried to record an iptv (type 4097) channel and got a crash. Tried again and crashed again. Solo2 running vix 5.3.032. Will see if i have crash log.

< 98425.2097> 16:36:06.1181 callback(*retval)
< 98425.2111> 16:36:06.1195 File "/usr/lib/enigma2/python/Screens/InfoBarGenerics.py", line 3131, in recordQuestionCallback
< 98425.2127> 16:36:06.1211 File "/usr/lib/enigma2/python/Screens/InfoBarGenerics.py", line 3043, in startInstantRecording
< 98425.2142> 16:36:06.1226 File "/usr/lib/enigma2/python/RecordTimer.py", line 191, in __init__
< 98425.2157> 16:36:06.1241 AttributeError: 'iPlayableServicePtr' object has no attribute 'isRecordable'
< 98425.2171> 16:36:06.1255 [ePyObject] (CallObject(<bound method Session.processDelay of <__main__.Session instance at 0x637e4b48>>,()) failed)

Same as OP
bellejt
23-05-20, 16:47
updated on vu duo2 and working perfect.All recordings possible and no crash sofar
ccs
23-05-20, 16:48
Same as OP
... who said he was recording from satellite - maybe he needs to re-confirm what he was dong.
Andy_Hazza
23-05-20, 16:50
I just tried to record an iptv (type 1 although i tought i had it set to type 4097)) channel and got a crash. Tried again and crashed again. Solo2 running vix 5.3.032. Will see if i have crash log.

< 98425.2097> 16:36:06.1181 callback(*retval)
< 98425.2111> 16:36:06.1195 File "/usr/lib/enigma2/python/Screens/InfoBarGenerics.py", line 3131, in recordQuestionCallback
< 98425.2127> 16:36:06.1211 File "/usr/lib/enigma2/python/Screens/InfoBarGenerics.py", line 3043, in startInstantRecording
< 98425.2142> 16:36:06.1226 File "/usr/lib/enigma2/python/RecordTimer.py", line 191, in __init__
< 98425.2157> 16:36:06.1241 AttributeError: 'iPlayableServicePtr' object has no attribute 'isRecordable'
< 98425.2171> 16:36:06.1255 [ePyObject] (CallObject(<bound method Session.processDelay of <__main__.Session instance at 0x637e4b48>>,()) failed)

Same as OP

OP is recording from Satellite channels. You’re trying to record from IPTV channels.


Sent from my iPhone using Tapatalk
spanner123
23-05-20, 16:55
Yes I realise OP said satellite but strange crash log is exactly the same! By the way it was type 4097 not type 1 iptv.
abu baniaz
23-05-20, 17:21
Are you using a plugin that replaces the InfoBarGenerics.py file?
spanner123
23-05-20, 17:37
Are you using a plugin that replaces the InfoBarGenerics.py file?

Not as far as I know. Simple 1080 skin. Standard vix with a few plugins including jedimaker and jediepg. Basically the setup I have been using for a long long time.
spanner123
23-05-20, 17:42
Maybe a clue just tried an instant record again and it's ok. Difference is this channel has an epg the other one didn't!
twol
23-05-20, 18:16
Maybe a clue just tried an instant record again and it's ok. Difference is this channel has an epg the other one didn't!
Kind of makes sense - although as you are recording forever, it shouldn‘t make a difference - but I am sure somewhere it expects an epg entry
spanner123
23-05-20, 18:22
Kind of makes sense - although as you are recording forever, it shouldn‘t make a difference - but I am sure somewhere it expects an epg entry

When it crashed twice first was stop after current event although no epg so I tried it with record indefinately and it still crashed. In the past wth no epg even you you selected stop after current event it would bring up a message saying recording indefinately if there was no epg.
abu baniaz
23-05-20, 18:28
If you disable the split timer function, do you get the same crash?
spanner123
23-05-20, 18:31
If you disable the split timer function, do you get the same crash?

Yes still crashes with split timer disabled.
spanner123
23-05-20, 19:30
removed jedimaker just in case and tried again still get crash. Latest log:

<108551.9101> 19:24:52.8186 [eInputDeviceInit] 0 a7 (167) 1
<108551.9121> 19:24:52.8205 [InfoBarGenerics] KEY: 167 RECORD
<108552.7732> 19:24:53.6816 [eDVBFrontend1] close frontend
<108552.7746> 19:24:53.6831 [eDVBFrontend1] setVoltage FE_ENABLE_HIGH_LNB_VOLTAGE 0 FE_SET_VOLTAGE 2
<108552.7869> 19:24:53.6953 [eInputDeviceInit] 1 6c (108) 1
<108552.7891> 19:24:53.6975 [InfoBarGenerics] KEY: 108 DOWN
<108552.7907> 19:24:53.6991 [ActionMap] Keymap 'WizardActions' -> Action = 'down'.
<108553.0171> 19:24:53.9255 [eInputDeviceInit] 0 6c (108) 1
<108553.0193> 19:24:53.9277 [InfoBarGenerics] KEY: 108 DOWN
<108553.3017> 19:24:54.2102 [eInputDeviceInit] 1 160 (352) 1
<108553.3040> 19:24:54.2124 [InfoBarGenerics] KEY: 352 OK
<108553.3058> 19:24:54.2142 [ActionMap] Keymap 'WizardActions' -> Action = 'ok'.
<108553.3846> 19:24:54.2930 Traceback (most recent call last):
<108553.3873> 19:24:54.2957 File "/usr/lib/enigma2/python/mytest.py", line 244, in processDelay
<108553.3964> 19:24:54.3048 callback(*retval)
<108553.3983> 19:24:54.3067 File "/usr/lib/enigma2/python/Screens/InfoBarGenerics.py", line 3131, in recordQuestionCallback
<108553.4003> 19:24:54.3087 File "/usr/lib/enigma2/python/Screens/InfoBarGenerics.py", line 3043, in startInstantRecording
<108553.4024> 19:24:54.3108 File "/usr/lib/enigma2/python/RecordTimer.py", line 191, in __init__
<108553.4045> 19:24:54.3129 AttributeError: 'iPlayableServicePtr' object has no attribute 'isRecordable'
<108553.4063> 19:24:54.3147 [ePyObject] (CallObject(<bound method Session.processDelay of <__main__.Session instance at 0x634eac38>>,()) failed)
ccs
23-05-20, 19:39
Just noticed this when comparing boxes......


root@et10000:~# ls -l /usr/lib/enigma2/python/mytest.py*
-rw-r--r-- 1 root root 21709 May 21 14:02 /usr/lib/enigma2/python/mytest.py
-rw-r--r-- 1 root root 22942 May 21 14:02 /usr/lib/enigma2/python/mytest.pyo
root@et10000:~#

Is this .py file meant to be there?
spanner123
23-05-20, 19:41
Just noticed this when comparing boxes......


root@et10000:~# ls -l /usr/lib/enigma2/python/mytest.py*
-rw-r--r-- 1 root root 21709 May 21 14:02 /usr/lib/enigma2/python/mytest.py
-rw-r--r-- 1 root root 22942 May 21 14:02 /usr/lib/enigma2/python/mytest.pyo
root@et10000:~#

Is this .py file meant to be there?

I've got both! I renamed mytest.py and rebooted and it got stuck on splash screen so I guess it must be there!
bbbuk
23-05-20, 21:00
Tried recording (indefinitely) with a 4097 service ref and no issues at all - with and without new time split feature.

PS: I also have two mytest.* files
spanner123
23-05-20, 21:05
Tried recording (indefinitely) with a 4097 service ref and no issues at all - with and without new time split feature.

PS: I also have two mytest.* files

Was this with or without epg? Definately seems to be iptv without epg that is the problem.
abu baniaz
23-05-20, 21:17
the issue is the recordable attribute being called in timer edit.
ccs
23-05-20, 21:23
I've managed to crash it on freeview channel 1094 Paramount Network+1.

The grid epg is empty, and when I select it I get Service not found (SID not found in PAT).

Instant Record + Add recording (indefinitely), crashes like reported earlier.

1st record option crashes as well.
spanner123
23-05-20, 21:28
I've managed to crash it on freeview channel 1094 Paramount Network+1.

The grid epg is empty, and when I select it I get Service not found (SID not found in PAT).

Instant Record + Add recording (indefinitely), crashes like reported earlier.

1st record option crashes as well.

I tried recording a satellite channel with no epg and it was ok! All very starnge.
ccs
23-05-20, 21:32
Debug log.......


< 35504.3064> 21:13:53.3225 [eDVBDemux] open demux /dev/dvb/adapter0/demux1
< 35504.3065> 21:13:53.3225 [eDVBSectionReader] DMX_SET_FILTER pid=1013
< 35504.4875> 21:13:53.5036 [eDVBDemux] open demux /dev/dvb/adapter0/demux1
< 35504.4876> 21:13:53.5036 [eDVBSectionReader] DMX_SET_FILTER pid=1019
< 35504.5406> 21:13:53.5566 [eDVBServicePMTHandler] sdt update done!
< 35504.9803> 21:13:53.9963 [eDVBServicePlay] DVB service failed to tune - error 3
< 35504.9809> 21:13:53.9970 [Notifications] RemovePopup, id = ZapError
< 35504.9810> 21:13:53.9971 [Notifications] AddPopup, id = ZapError
< 35504.9820> 21:13:53.9981 [Skin] Processing screen 'MessageBoxSimple', position=(480, 532), size=(960 x 15) for module 'MessageBox'.
< 35504.9884> 21:13:54.0045 [GUISkin] warning, skin is missing element icon in <class 'Screens.MessageBox.MessageBox'>(Service not found!
(SID not found in PAT))< 35504.9885> 21:13:54.0046
< 35504.9886> 21:13:54.0047 [Pixmap] setPixmapNum(3) failed! defined pixmaps: []
< 35505.9811> 21:13:54.9972 [eEPGCache] start caching events(1590264834)
< 35505.9813> 21:13:54.9973 [eDVBSectionReader] DMX_SET_FILTER pid=18
< 35505.9819> 21:13:54.9980 [eDVBSectionReader] DMX_SET_FILTER pid=18
< 35505.9827> 21:13:54.9988 [eDVBSectionReader] DMX_SET_FILTER pid=18
< 35513.9738> 21:14:02.9899 [eEPGCache] nownext finished(1590264842)
< 35520.5279> 21:14:09.5440 [eInputDeviceInit] 1 a7 (167) 1
< 35520.5285> 21:14:09.5445 [InfoBarGenerics] KEY: 167 RECORD
< 35520.5287> 21:14:09.5448 [ActionMap] Keymap 'InfobarInstantRecord' -> Action = 'instantRecord'.
< 35520.5364> 21:14:09.5524 [Skin] Processing screen 'ChoiceBox', position=(0, 0), size=(1920 x 1080) for module 'ChoiceBox'.
< 35520.5429> 21:14:09.5590 [GUISkin] warning, skin is missing element description in <class 'Screens.ChoiceBox.ChoiceBox'>
< 35520.5457> 21:14:09.5618 [Skin] Processing screen 'ChoiceBox_summary' from list 'ChoiceBox_summary, SimpleSummary', position=(0, 0), size=(800 x 480) for module 'SimpleSummary'.
< 35521.0790> 21:14:10.0951 [eInputDeviceInit] 0 a7 (167) 1
< 35521.0794> 21:14:10.0954 [InfoBarGenerics] KEY: 167 RECORD
< 35522.1836> 21:14:11.1996 [eInputDeviceInit] 1 6c (108) 1
< 35522.1840> 21:14:11.2001 [InfoBarGenerics] KEY: 108 DOWN
< 35522.1843> 21:14:11.2003 [ActionMap] Keymap 'WizardActions' -> Action = 'down'.
< 35522.4151> 21:14:11.4312 [eInputDeviceInit] 0 6c (108) 1
< 35522.4156> 21:14:11.4316 [InfoBarGenerics] KEY: 108 DOWN
< 35525.2575> 21:14:14.2735 [eInputDeviceInit] 1 160 (352) 1
< 35525.2579> 21:14:14.2740 [InfoBarGenerics] KEY: 352 OK
< 35525.2582> 21:14:14.2742 [ActionMap] Keymap 'WizardActions' -> Action = 'ok'.
[MAIN] executing main
NXPL_RegisterNexusDisplayPlatform data->hwInterface 0xb3903a40
TuxTxt cache cleared
cleaning up
TuxTxt cache cleared
TuxTxt cache cleared
cleaning up
TuxTxt cache cleared
TuxTxt cache cleared
cleaning up
TuxTxt cache cleared
TuxTxt cache cleared
cleaning up
TuxTxt cache cleared
TuxTxt cache cleared
cleaning up
TuxTxt cache cleared
TuxTxt cache cleared
cleaning up
TuxTxt cache cleared
TuxTxt cache cleared
cleaning up
TuxTxt cache cleared
TuxTxt cache cleared
cleaning up
TuxTxt cache cleared
< 35525.2655> 21:14:14.2815 Traceback (most recent call last):
< 35525.2655> 21:14:14.2816 File "/usr/lib/enigma2/python/mytest.py", line 244, in processDelay
< 35525.2658> 21:14:14.2818 callback(*retval)
< 35525.2658> 21:14:14.2819 File "/usr/lib/enigma2/python/Screens/InfoBarGenerics.py", line 3131, in recordQuestionCallback
< 35525.2661> 21:14:14.2821 File "/usr/lib/enigma2/python/Screens/InfoBarGenerics.py", line 3043, in startInstantRecording
< 35525.2662> 21:14:14.2822 File "/usr/lib/enigma2/python/RecordTimer.py", line 191, in __init__
< 35525.2664> 21:14:14.2824 AttributeError: 'iPlayableServicePtr' object has no attribute 'isRecordable'
< 35525.2664> 21:14:14.2825 [ePyObject] (CallObject(<bound method Session.processDelay of <__main__.Session instance at 0x9e838788>>,()) failed)
< 35528.2079> 21:14:17.2240 [gRC] main thread is non-idle! display spinner!
bbbuk
23-05-20, 21:58
Right I put box to sleep and deleted epg.dat...

I too can confirm a crash when I had no epg for an IPTV channel.

I couldn't create a crash for a normal channel but that will be because now/next had populated so there was EPG, albeit limited to just now/next.



< 125.2843> 21:50:08.2360 File "/usr/lib/enigma2/python/mytest.py", line 244, in processDelay
< 125.2849> 21:50:08.2366 callback(*retval)
< 125.2852> 21:50:08.2368 File "/usr/lib/enigma2/python/Screens/InfoBarGenerics.py", line 3131, in recordQuestionCallback
< 125.2857> 21:50:08.2374 File "/usr/lib/enigma2/python/Screens/InfoBarGenerics.py", line 3043, in startInstantRecording
< 125.2861> 21:50:08.2378 File "/usr/lib/enigma2/python/RecordTimer.py", line 191, in __init__
< 125.2869> 21:50:08.2386 AttributeError: 'iPlayableServicePtr' object has no attribute 'isRecordable'
< 125.2873> 21:50:08.2389 [ePyObject] (CallObject(<bound method Session.processDelay of <__main__.Session instance at 0x6539e2b0>>,()) failed)
ccs
23-05-20, 22:09
Another odd thing.

I've got Skip empty services set to no, but when I'm in Grid EPG, sometimes I can select freeview channel 1094 Paramount Network+1 using the up/down arrow keys, and sometimes it skips passed it, going from 1093 to 1095.
abu baniaz
24-05-20, 01:07
That block of code has not changed since at least last year, must be section where the data is forwarded to.

https://github.com/OpenViX/enigma2/blob/7c85b99bb7bc2b06253fe802307f5e33187c74bf/RecordTimer.py#L161:L209


Another similar crash on DVB services. Disable all EPG readers, delete EPG, restart. Record button, select the enter duration option.



< 66073.2817> 23:42:29.0795 Traceback (most recent call last):
< 66073.2817> 23:42:29.0796 File "/usr/lib/enigma2/python/mytest.py", line 244, in processDelay
< 66073.2820> 23:42:29.0799 callback(*retval)
< 66073.2821> 23:42:29.0799 File "/usr/lib/enigma2/python/Screens/InfoBarGenerics.py", line 3116, in recordQuestionCallback
< 66073.2823> 23:42:29.0802 File "/usr/lib/enigma2/python/Screens/InfoBarGenerics.py", line 3014, in startInstantRecording
< 66073.2824> 23:42:29.0803 File "/usr/lib/enigma2/python/Screens/InfoBarGenerics.py", line 3002, in getProgramInfoAndEvent
< 66073.2825> 23:42:29.0804 File "/usr/lib/enigma2/python/RecordTimer.py", line 58, in parseEvent
< 66073.2827> 23:42:29.0806 AttributeError: 'eServiceReference' object has no attribute 'ref'
< 66073.2828> 23:42:29.0806 [ePyObject] (CallObject(<bound method Session.processDelay of <__main__.Session instance at 0xb15e2aa8>>,()) failed)



If I repeat the same test with the prototype service reference file, I get the same crash as everyone above.



< 16663.0877> 01:12:45.5571 [ActionMap] Keymap 'WizardActions' -> Action = 'ok'.
< 16663.0945> 01:12:45.5639 Traceback (most recent call last):
< 16663.0946> 01:12:45.5639 File "/usr/lib/enigma2/python/mytest.py", line 244, in processDelay
< 16663.0948> 01:12:45.5642 callback(*retval)
< 16663.0949> 01:12:45.5642 File "/usr/lib/enigma2/python/Screens/InfoBarGenerics.py", line 3131, in recordQuestionCallback
< 16663.0950> 01:12:45.5644 File "/usr/lib/enigma2/python/Screens/InfoBarGenerics.py", line 3043, in startInstantRecording
< 16663.0952> 01:12:45.5645 File "/usr/lib/enigma2/python/RecordTimer.py", line 191, in __init__
< 16663.0953> 01:12:45.5647 AttributeError: 'iPlayableServicePtr' object has no attribute 'isRecordable'
< 16663.0954> 01:12:45.5647 [ePyObject] (CallObject(<bound method Session.processDelay of <__main__.Session instance at 0xb16fed28>>,()) failed)
simonc
24-05-20, 09:25
I've spotted the problem. It's a horrible section of code and symtomatic of many things that are bad with dynamic languages like Python! When is an eServiceReference not an eServiceReference? When it's an instance of an abstract interface class. Shouldn't be possible, but somehow here it is.

Yet another try/except block will make it behave I suppose :rolleyes:
bbbuk
24-05-20, 10:20
I've tracked it down to something that was only changed/introduced in 5.3.032.

Was fine in 5.3.030 release and random ones earlier than that.

Could it have been the fix to instant record or the split time record feature?

Update:
Does that fit in with what you have found @simonc ?
ccs
24-05-20, 10:34
I've just tried 5.3.019 and can confirm that I can't make it crash as in #32.
simonc
24-05-20, 10:36
Yes. My fault. Here: https://github.com/OpenViX/enigma2/blob/0f2314bd2bbe6575da6ee67787962f5f225a3aa7/lib/python/Screens/InfoBarGenerics.py#L3005

Who'd have thought that a function called getCurrentService didn't return an eServiceReference. You call getCurrentlyPlayingServiceOrGroup to get that!
abu baniaz
24-05-20, 10:48
That has fixed it for me. Tried with the two versions of service reference.

Can someone with IPTV try please before I merge it?
Unzip and transfer to /usr/lib/enigma2/python/Screens
ccs
24-05-20, 10:49
Works ok for me as well following #32 sequence.
ccs
24-05-20, 10:56
Works ok for me as well following #32 sequence.

Also ok using the test version of ServiceReference.
spanner123
24-05-20, 11:05
That has fixed it for me. Tried with the two versions of service reference.

Can someone with IPTV try please before I merge it?
Unzip and transfer to /usr/lib/enigma2/python/Screens

Can confirm this has resolved the problem. many thanks!
bbbuk
24-05-20, 11:23
That has fixed it for me. Tried with the two versions of service reference.

Can someone with IPTV try please before I merge it?
Unzip and transfer to /usr/lib/enigma2/python/ScreensCan also confirm this fixes it for me on IPTV as well as normal dvb where no EPG exists.

Just a thought, is it worth being prudent and checking the split time recording still works (I'm not sure if this fix will have an unintended consequence with simon's split time recording feature)?
ccs
24-05-20, 12:01
Split Timer still looks fine to me, all the checks I did before in the other thread have produced exactly the same results.