242 replies to this topic

[WanWizard]

Yes, I am telling you that.

 

Someone is telling you his finger hurts, and you suggest to amputate the entire arm, as you have no clue why the finger hurts, only that removing the arm will surely remove the pain!

 

So if you are serious to fix this issue, test if the problem also occurs if epgRefresh doesn't use "fake recording", and if not, fix epgRefresh. Because that seems to be the trigger, not normal recordings.

[WanWizard]

I would say first confirm if it is really related to the fake recording alias dummy streaming of the epgrefresh plugin and if so resolved that. I think for now reverting is not an option as it worked as it for years. I never get the described interlock here, but I also never used epgrefresh.

 

I do use epgRefresh on my Duo 4K, but do not use fake recording, I've got plenty of tuners. And I also have never seen this problem.

[littlesat]

Maybe the thread starter can also try another option… to confirm this and then this should be targeted.

[WanWizard]

And they other way around, if @Abu can confirm he also uses "fake recording", and can test with another option?

[ims]

No problem on any of my boxes. I am using EPGRefresh, but I do not use fake recording.

[paneologist]

Yes, I am telling you that.

 

Someone is telling you his finger hurts, and you suggest to amputate the entire arm, as you have no clue why the finger hurts, only that removing the arm will surely remove the pain!

 

So if you are serious to fix this issue, test if the problem also occurs if epgRefresh doesn't use "fake recording", and if not, fix epgRefresh. Because that seems to be the trigger, not normal recordings.

Just wanted to remind, that the problem doesn't occur only while epgrefresh is running, but also when zapping (stopping timeshift) or after regular timer recordings. sometimes.

[littlesat]

Just wanted to remind, that the problem doesn't occur only while epgrefresh is running, but also when zapping (stopping timeshift) or after regular timer recordings. sometimes.

If so can you give us some more 'instructions' how we can verify this....

 

In one of your logs I see this just before it happens...

2023-08-05T04:27:12+02:00 stb1.swabian.net enigma2: [eEPGChannelData] schedule finished(1691202432)
2023-08-05T04:27:12+02:00 stb1.swabian.net enigma2: [eEPGChannelData] stop caching events(1691202432)
2023-08-05T04:27:12+02:00 stb1.swabian.net enigma2: [eEPGChannelData] next update in 60 min

I'm aware that when something is corrupted in the EPG data that such crashes also can happen... then you need to 'clear' the epg.dat.

 

I also do recordings, streamings and timeshifts and I never experienced this for years.... This will not say it never could happen of course.

 

It is also important we do not loss the complete focus... at least we think for now that epgrefresh's fake recording is one of the triggers.

Edited by Frenske, 10 August 2023 - 11:20.
Edited the first quote.

[paneologist]

 

Just wanted to remind, that the problem doesn't occur only while epgrefresh is running, but also when zapping (stopping timeshift) or after regular timer recordings. sometimes.
[/quote[
If so can you give us some more 'instructions' how we can verify this....

unfortunately, it seems just random - after end of an recording

In one of your logs I see this just before it happens...

2023-08-05T04:27:12+02:00 stb1.swabian.net enigma2: [eEPGChannelData] schedule finished(1691202432)
2023-08-05T04:27:12+02:00 stb1.swabian.net enigma2: [eEPGChannelData] stop caching events(1691202432)
2023-08-05T04:27:12+02:00 stb1.swabian.net enigma2: [eEPGChannelData] next update in 60 min

I'm aware that when something is corrupted in the EPG data that such crashes also can happen... then you need to 'clear' the epg.dat.
 
I also do recordings, streamings and timeshifts and I never experienced this for years.... This will not say it never could happen of course.
 
It is also important we do not loss the complete focus... at least we think for now that epgrefresh's fake recording is one of the triggers.

 

 

well, changed runlevel to 4, deleted epg.dat, init 3 and now I run epgrefresh manually, and see what happens

 

Edit: btw. I am still on release image, as I didn't have enough time the last few days.

(a question a bit off-topic. But it seems many of you use epgrefresh. what is a good (and lowest practical) setting for time to stay on one channel? default is 120 seconds, which would make epgrefresh take ages. I reduced it to 30 seconds and for testing I set it to 10.)

Edited by paneologist, 10 August 2023 - 10:05.

[littlesat]

 I reduced it to 30 seconds and for testing I set it to 10

 

The two minutes might be mandatory for a 'transponder' to send all the EPG evailable.... Reducing the time may result in EPG that is partly received. All depends on the 'bandwidth' the provider uses for the EPG to send (all) the channel's EPG..And maybe also due to the 'fake recording method' getting the EPG input that is sometimes available on a different 'stream' on a transponder than the channel zapped to might be interrupted and lead to invalid data. About the last statement I made here I'm not fully sure, but this might also something that needs to be investigated.

Edited by littlesat, 10 August 2023 - 11:28.

[paneologist]

 

 I reduced it to 30 seconds and for testing I set it to 10

 

The two minutes might be mandatory for a 'transponder' to send all the EPG evailable.... Reducing the time may result in EPG that is partly received. All depends on the 'bandwidth' the provider uses for the EPG to send (all) the channel's EPG..And maybe also due to the 'fake recording method' getting the EPG input that is sometimes available on a different 'stream' on a transponder than the channel zapped to might be interrupted and lead to invalid data. About the last statement I made here I'm not fully sure, but this might also something that needs to be investigated.

 

thank you. Would be a great feature to use several tuners in parallel for epgrefresh, as i have plenty of them Jess with 2 twin-FBC tuners..... 2minutes on a channel using one tuner is way too slow imho. I already reduced cahnnels to only those bouquets i use.

using the PIP mode in epgrefresh seems to cause some issues, tune of channels I am listening to or watching sometimes fail (timeout, PID not found, tune failed) - never had this issue before. !? in the log i see these three kinds of errors:

[eDVBServicePlay] DVB service failed to tune - error 1

[eDVBServicePlay] DVB service failed to tune - error 2

[eDVBServicePlay] DVB service failed to tune - error 3

 

 

regarding the topic:
I ran epgrefresh using PIP and fakerecording. box didn't got stuck, yet. created timers with around 20 parallel recordings. no problem yet. afterwards I reboot and try again.

Edited by paneologist, 10 August 2023 - 11:55.

[paneologist]

and it happened again after a timer finished:

 stb1debuglog3.txt   24.77KB   4 downloads

tomorrow I'll install dev image

[Dimitrij]

 

2023-08-10T15:35:09+02:00 stb1.swabian.net enigma2: [eFilePushThreadRecorder] stopping thread: 1
2023-08-10T15:35:10+02:00 stb1.swabian.net enigma2: [eFilePushThreadRecorder] thread could not be stopped!

simple test

+	kill();
-	if(safeguard > 0)
-		kill();
-	else
-		eWarning("[eFilePushThreadRecorder] thread could not be stopped!");
}

 

[WanWizard]

@Dimitrij,

 

We know that will stop the hang-up from happening. But that is not why the hang-up happens.

 

Let's agree that if we can't find the issue before the release of 9.0, we'll add this to the release code as a workaround, ok?

 

@paneologist,

 

What I notice in this log, is that while the recording is running the EPG reader is started, I assume on the channel that is newly tuned for the recording as there is a ts file playing (radio?).

If the recording stops, the tuning also stops and the tuner is released, but the EPGreader only stops 14 seconds after the recording thead has given up?

2023-08-10T15:33:47+02:00 stb1.swabian.net enigma2: [eEPGChannelData] start reading events(1691674427)
2023-08-10T15:33:47+02:00 stb1.swabian.net enigma2: [huffman] read.. '/usr/share/enigma2/otv_00c00000_0001_040f.dict'
2023-08-10T15:33:47+02:00 stb1.swabian.net enigma2: [eEPGChannelData] abort non avail OpenTV EIT reading
2023-08-10T15:33:54+02:00 stb1.swabian.net enigma2: [eEPGChannelData] abort non avail schedule other reading
2023-08-10T15:33:54+02:00 stb1.swabian.net enigma2: [eEPGChannelData] abort non avail virgin nownext reading
2023-08-10T15:33:54+02:00 stb1.swabian.net enigma2: [eEPGChannelData] abort non avail virgin schedule reading
2023-08-10T15:33:54+02:00 stb1.swabian.net enigma2: [eEPGChannelData] abort non avail netmed schedule reading
2023-08-10T15:33:54+02:00 stb1.swabian.net enigma2: [eEPGChannelData] abort non avail netmed schedule other reading
2023-08-10T15:33:54+02:00 stb1.swabian.net enigma2: [eEPGChannelData] abort non avail FreeSat schedule_other reading
2023-08-10T15:33:54+02:00 stb1.swabian.net enigma2: [eEPGChannelData] abort non avail viasat reading
2023-08-10T15:33:54+02:00 stb1.swabian.net enigma2: [eEPGChannelData] abort non avail mhw reading
2023-08-10T15:33:55+02:00 stb1.swabian.net enigma2: [eEPGChannelData] nownext finished(1691674435)

2023-08-10T15:35:00+02:00 stb1.swabian.net enigma2: [eDVBServiceRecord] stop recording!

2023-08-10T15:35:24+02:00 stb1.swabian.net enigma2: [eEPGChannelData] schedule finished(1691674524)
2023-08-10T15:35:24+02:00 stb1.swabian.net enigma2: [eEPGChannelData] stop caching events(1691674524)
2023-08-10T15:35:24+02:00 stb1.swabian.net enigma2: [eEPGChannelData] next update in 60 min

What if the EPG reader locks things? We've had more issues with EPG readers and concurrency.

 

Also.

 

Eventually it gets to

2023-08-10T15:35:10+02:00 stb1.swabian.net enigma2: [eThread] Destroyed thread without joining it, this usually means your thread is now running with a halfway destroyed object

which is emitted in the thread object destructor, and which also issues a kill() to forcefully end the thread.

 

Does this mean that after this message the spinner disappeared and your box was usable again?

[littlesat]

So finally we end up at the epg reader stuff that is buggy since the beginning? So I’m afraid that is still not totally resolved?

[Abu Baniaz]

I'm really sorry, I wasn't able to do any tests yesterday. I'll try later today.

[paneologist]

@paneologist,
 
What I notice in this log, is that while the recording is running the EPG reader is started, I assume on the channel that is newly tuned for the recording as there is a ts file playing (radio?).

yes I was listening to a radio recording

Also.
 
Eventually it gets to

2023-08-10T15:35:10+02:00 stb1.swabian.net enigma2: [eThread] Destroyed thread without joining it, this usually means your thread is now running with a halfway destroyed object

which is emitted in the thread object destructor, and which also issues a kill() to forcefully end the thread.
 
Does this mean that after this message the spinner disappeared and your box was usable again?

unfortunately no. this message is shown sometimes, but spinner stays

[paneologist]

so tried to install latest nightly build.

resulted in enigma2 not starting:

2023-08-11T09:48:53+02:00 stb1.swabian.net enigma2: [gRC] main thread is non-idle! display spinner!
2023-08-11T09:48:53+02:00 stb1.swabian.net enigma2: [gRC] enabelSpinner: no spinner DC!

keeps looping.

Full log:  stb1debuglog4.txt   85.32KB   1 downloads

 

some troubles i had, and what I did:

it first booted well, after reapplying config, it didn't start up anymore.

unfortunately not all /etc config was restored, so I did manually for:

/etc/syslog-startup

/etc/fstab

/etc/inittab

/etc/ciplus/

and my change to /usr/bin/enigma2.sh for continuos debug logging to stdout

 

next time I know to manually backup those files and the root's home directory...

 

 

-- tried flashing again. same result:  stb1debuglog5.txt   77.16KB   1 downloads

(just a thought: are package dependencies enforced, when restoring? maybe it lacks something..?)

note the kernel oopses.

trying to install release. hopefully it'll work.
and there is a lot more config I need to restore... smartd, openwebif/ssl, hostname, ntp, nfs... hm :-/

Edited by paneologist, 11 August 2023 - 09:28.

[littlesat]

Both logs... something did crash after 

 

2023-08-11T09:47:49+02:00 stb1.swabian.net enigma2: [eDVBDB] ---- opening lame channel db

2023-08-11T09:47:49+02:00 stb1.swabian.net enigma2: [eDVBDB] reading services (version 4)

2023-08-11T09:47:49+02:00 stb1.swabian.net enigma2: [eDVBDB] loaded 466 channels/transponders and 6272 services

2023-08-11T09:47:49+02:00 stb1.swabian.net enigma2: eDVBSatelliteEquipmentControl::clear()

 

Looks like at startup of E2... at least settings are loaded and just after this....

2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.454283] Unable to handle kernel NULL pointer dereference at virtual address 00000020
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.462403] pgd = c0003000
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.465129] [00000020] *pgd=80000000004003, *pmd=00000000
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.470577] Internal error: Oops: 207 [#1] SMP ARM
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.475373] Modules linked in: bluetooth ipv6 brcmfb(O) dvb_bcm7444(PO) dvb_base(PO) bcm_event(PO) procmk(O) dhd(O) sit2fe(O) dvb_usb_dvbsky(O) dvb_usb_v2_media_tree(O) fuse
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.491031] CPU: 3 PID: 1535 Comm: kdvb-ad-1-fe--6 Tainted: P           O    4.1.45-1.17 #1
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.499390] Hardware name: Broadcom STB (Flattened Device Tree)
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.505316] task: d3addb00 ti: d3afa000 task.ti: d3afa000
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.510728] PC is at dvb_usb_fe_init+0x24/0x174 [dvb_usb_v2_media_tree]
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.517355] LR is at dvb_frontend_init+0x34/0xc4
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.521978] pc : [<bf01cfd4>]    lr : [<c046d544>]    psr: a00b0013
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.521978] sp : d3afbee0  ip : 00000000  fp : 00000000
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.533466] r10: 00000000  r9 : 00000000  r8 : bf020368
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.538696] r7 : 00000438  r6 : d5e92800  r5 : d5f1dc00  r4 : d5e92800
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.545231] r3 : fffffd80  r2 : 00000000  r1 : 000001da  r0 : d5e92800
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.551765] Flags: NzCv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment kernel
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.559081] Control: 30c5387d  Table: 15e3e0c0  DAC: 55555555
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.564834] Process kdvb-ad-1-fe--6 (pid: 1535, stack limit = 0xd3afa210)
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.571629] Stack: (0xd3afbee0 to 0xd3afc000)
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.575991] bee0: c0874880 d3addb00 c0874880 d5e92800 d5f1dc00 d5e92800 d5f1dd98 00000000
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.584177] bf00: 00000000 c046d544 00000000 c087868c d3afbf54 c05fe228 d5e92800 c046eff4
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.592363] bf20: 00000000 d381fd58 d381fd5c 00000003 c046efa0 d3afa000 00000000 d5e92800
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.600549] bf40: c046efa0 d4a6bcc0 00000000 d5e92800 c046efa0 00000000 00000000 00000000
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.608736] bf60: 00000000 c0040bf8 00000000 00000000 d3afbf70 d5e92800 00000000 00000000
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.616922] bf80: d3afbf80 d3afbf80 00000000 00000000 d3afbf90 d3afbf90 d3afbfac d4a6bcc0
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.625108] bfa0: c0040b0c 00000000 00000000 c000ffa8 00000000 00000000 00000000 00000000
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.633294] bfc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.641480] bfe0: 00000000 00000000 00000000 00000000 00000013 00000000 5f70e64d ef5b5f0d
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.649678] [<bf01cfd4>] (dvb_usb_fe_init [dvb_usb_v2_media_tree]) from [<c046d544>] (dvb_frontend_init+0x34/0xc4)
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.660042] [<c046d544>] (dvb_frontend_init) from [<c046eff4>] (dvb_frontend_thread+0x54/0x728)
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.668753] [<c046eff4>] (dvb_frontend_thread) from [<c0040bf8>] (kthread+0xec/0xf4)
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.676511] [<c0040bf8>] (kthread) from [<c000ffa8>] (ret_from_fork+0x14/0x2c)
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.683743] Code: e34b8f02 e3007438 e30011da e24dd00c (e5925020) 
2023-08-11T09:47:49+02:00 stb1.swabian.net kernel: [   29.689892] ---[ end trace 255c3d93e8f29106 ]---

Edited by littlesat, 11 August 2023 - 11:16.

[WanWizard]

Crash in the USB DVB driver, from the looks of it.

 

I'm still trying to find a way to reproduce it. If it is a code error somewhere, we surely must be able to reproduce it? Why do some people have this, and most don't (given the popularity of the Duo 4K /  Duo 4K SE)?

 

I have a Duo 4K here, with FBC tuners, with a Unicable setup, just like @paneologist and @abu. It does nothing but record, it uses EPG readers (including OpenTV), it uses EPGimport and EPGrefresh once a day. It is my stream server, and I use OWIF to manage it remotely.

 

And yet, stable as a rock... ???

[paneologist]

Crash in the USB DVB driver, from the looks of it.

I might try afain tomrrow with first uninstalling usb-dvb drivers.

I'm still trying to find a way to reproduce it. If it is a code error somewhere, we surely must be able to reproduce it? Why do some people have this, and most don't (given the popularity of the Duo 4K /  Duo 4K SE)?

hardware related? quite implausile..?

I have a Duo 4K here, with FBC tuners, with a Unicable setup, just like @paneologist and @abu. It does nothing but record, it uses EPG readers (including OpenTV), it uses EPGimport and EPGrefresh once a day. It is my stream server, and I use OWIF to manage it remotely.
 
And yet, stable as a rock... ???

I wish I was that lucky.