Green Screen of Death

[Timewave]

Learning new box going through the menus and seeing what they do, when I do Menu-Setup-System-Customize, I get a Green Screen of Death and the box reboots.

Anyone else or just me?

[el bandido]

I would say using the term "Green Screen of Death" is being a bit dramatic because nothing actually dies or completely quits working. The box simply sees an error, flags the error by creating a report, then reboots.
At any rate, no matter what you call it, I appreciate you listing it so the error can be fixed. Thanks.

[el bandido]

The problem has been fixed. You may do an online software update: Menu--->Setup--->Software update, and follow onscreen instructions. Internet connection to the receiver is needed for online updates.

Let me know if you have any more issues.

[Timewave]

Wow,
That was fast, I must have posted in the right place.
All new to FTA and having fun playing.

Thank You!

[MikeB]

Learning new box going through the menus and seeing what they do, when I do Menu-Setup-System-Customize, I get a Green Screen of Death and the box reboots.

Anyone else or just me?

@el bandido I woke up this morning with the same issue, on my replacement OS Mio. It was flashed with TNAP 4.0 mid-January by Brian before he sent it to me. Looks like I'm going to have to reflash it from scratch. This happened completely a random, I have NO idea why: 17186

P.S. Where is the newest flash image to reflash from USB? All I see on site is dated back from 12/21/2020. [Only registered and activated users can see links]

[el bandido]

Flash multiboot image online to the slot of your choice.
See post #17 in this thread: [Only registered and activated users can see links]

Let me know if that doesn't fix it and we will do something else.

[MikeB]

Flash multiboot image online to the slot of your choice.
See post #17 in this thread: [Only registered and activated users can see links]

Let me know if that doesn't fix it and we will do something else.

Ok, I flashed the online multiboot image to Slot 1. It seems to be ok, though its skin is different than what I'm used to. Here's what I have, it was the only choice going from the working TNAP 3.2 in Slot 2, does this look like the latest?

17187

Ok, wait, now that I'm at least on a working version of TNAP 4.0, I checked online images again, and now it sees the 4/29/21 version. Flashing that now...

Ok, now I have THIS:

17188


Whoops, also ran a standard "Software Update" and now I have THIS:

17189

I HATE that the image sometimes gets corrupted, since it's happened at random a couple times now. However, I LOVE the EASE of re-flashing a fresh image, and fixing it from online, PROVIDED there's at least ONE working image in another slot!

Let this be a warning to others once again, you should ALWAYS have a minimum of TWO slots loaded with working images. If ONE goes South for whatever reason, it's so easy to switch to another slot and fix the bad one.

[el bandido]

It should have restored it back exactly the way you had it before. When you select a slot that has "with backup" or something similar when flashing, then the receiver is supposed to make a copy of what you have installed and restore it automatically, providing what you need can be downloaded from the internet. There is a yes or no question that comes up during flashing which asks if you want to restore. If you do nothing when this question appears, then everything will be restored providing what is needed can be downloaded.

[MikeB]

It should have restored it back exactly the way you had it before. When you select a slot that has "with backup" or something similar when flashing, then the receiver is supposed to make a copy of what you have installed and restore it automatically, providing what you need can be downloaded from the internet. There is a yes or no question that comes up during flashing which asks if you want to restore. If you do nothing when this question appears, then everything will be restored providing what is needed can be downloaded.

Well, I did do a restore with backup, and when it asked if I wanted to restore, I clicked YES. Previously, I had the GlamourAuraFHD skin loaded, and that did NOT restore.

Whoops, I went into the menu and switched to the GlamourAuraFHD screen, and I'm back to the cycling Green Screen of Death... I guess I'll have to reflash it yet again... Unless there's a way to get to that slot from another slot, and switch skins somehow? Apparently, that skin got corrupted somehow and caused this issue in the first place.

[skink]

My box did the same thing this morning on start-up.I just let it cool down,unplug the ethernet and she fired back up.Are you blaming the GlamourAura skin for the problem?Thanks

[MikeB]

My box did the same thing this morning on start-up.I just let it cool down,unplug the ethernet and she fired back up.Are you blaming the GlamourAura skin for the problem?Thanks

Yes, in my case it appears that after I had it all restored it was fine. THEN, I went into the skin selection and selected the GlamourAuraFHD skin. It rebooted, and went immediately back to the green screen of death. I've now restored it again, and I am NOT going to switch to the GlamourAura skin again. Something on that skin must be corrupted now, or maybe not completely compatible with the latest updates. Beats me.

Anyway, I also checked my settings, and then did a full backup. I also did another flash to Slot 3, and I now have 3 slots loaded. 2 with TNAP 4.0, and 1 with TNAP 3.2.

[skink]

Yup I saw the same thing,as soon as I plugged in to the web.Thanks

[MikeB]

Another poster at Satguys with a Mio+ just reported that his also just went Green screen of death last night, same as my regular Mio. This is really odd to have two issues exactly the same.
[Only registered and activated users can see links]

I posted the instructions for him to recover as I did. Hopefully he has a working image in another slot.

[el bandido]

Would be nice if someone would post the crash log. That information might help. The image is supposed to have a restart timeout and should not continuously reboot forever.

Anyway, I will try to replicate the problem if possible then see what can be done to fix it.

[el bandido]

Here is what I found using the current TNAP:
When the Glamour skin is being used, the receiver will crash and bootloop continuously. ( I now have a crash log.)
Installing an older version of Glamour skin will cause the same bootloop issue.
There seem to be no problems with the current TNAP and the OpenPLi skin.

We will have to see what needs to be done to fix this asap.

[MikeB]

Would be nice if someone would post the crash log. That information might help. The image is supposed to have a restart timeout and should not continuously reboot forever.

Anyway, I will try to replicate the problem if possible then see what can be done to fix it.

I would be more than happy to post the Crash log when an issue like this happens. However, HOW can you get to the Crash log when you can't access the receiver?

Also, I've been running the GlamourAura skin since Brian sent the new Mio back to me in January. He loaded 4.0 then, and YES, I have loaded some updates since then. However, I haven't loaded any sort of updates for at least 2 months? At least, I didn't initiate them. I wonder why the GlamourAura skin has been working fine, and just all of a sudden decides to take a dump? Not only that, but suddenly with more than one person AND on the SAME night?

The receiver doesn't do Push Updates automatically, does it?

P.S. If this helps any, my receiver was set with a timer to kick off at 12:50AM that night, and record a movie on 101w Movies channel. The timer kicked off, and the receiver must have corrupted right then and there. The reason I know this, is because once I restored a working image and checked for that recording, there was a ton of small garbage files named for the movie that kept piling up after every reboot, when the timer attempted to restart.

[el bandido]

The crash log would be in /tmp or usually is in /tmp. All you have t6o do is use FileZilla or something else to ftp the receiver and get the crash file from /tmp. That's what I did.

As to what is wrong, I do not know. We don't push updates. We will remove almost everything from feeds until the problem is fixed so the issue cannot spread.

[skink]

My receiver has been running fine up to that point and I don't record or set the timer for events.My antivirus was constantly alerting me of the internet connect to the box was vulnerable to attack the days leading up to this.Now that the box is disconnected,the warnings have stopped.Consequences?

[el bandido]

Feeds and files for the TNAP images are on a dedicated server that is not shared with anything else. Images are built from various sources that are downloaded and compiled from the various places on the Internet. It is doubtful that any virus or attack was sent to the receivers running TNAP.

Antivirus programs alert to threats that are real and are imagined. Your particular antivirus program may be pointing to the receiver because it is not password protected or lacks some other protection the antivirus program has decided it needs.

[skink]

I wasn't implying anything nefarious.I thank you for pointing out the possible deficiency.

[MikeB]

Just got a report at Satguys of another person that owns 2 regular Mio 4k receivers, and they just both crashed with this error. I've asked them to report over here, but that's up to them.

[Titanium]

I just received a similar report from a OS Mio 4K eBay customer that received a software error screen last evening and is rebooting. Asked him to post the STB log.

[eelstrebor]

Same problem here on both of my osmio-4k units (not the plus model) but I don't remember which skin I was using. After re-installing the TNAP4 image to slot one it set the skin to PLi-FullNightHD. Lost my bouquets and had to reset the TZ. Apparently some stuff didn't get backed up.

[Only registered and activated users can see links]-4619777

[Megatron817]

It seems possibly to be the weatherMsn plugin causing the crash. I removed the weather plugin with remove plugins menu and also deleted the weatherMsn folder from usr/lib/enigma2/python/plugins/Extensions. No problem now with GlamourAuraFHD skin.

[MikeB]

User at Satguys reports that simply removing the internet cable allows him to boot up normally. IF he plugs in the ethernet cable, it immediately crashes with the greenscreen again.

So, a person could remove internet access, switch to a different skin, and/or remove the weatherMsn plugin, and perhaps avoid having to reflash from scratch.

At least until we have a fix.

[Megatron817]

User at Satguys reports that simply removing the internet cable allows him to boot up normally. IF he plugs in the ethernet cable, it immediately crashes with the greenscreen again.

So, a person could remove internet access, switch to a different skin, and/or remove the weatherMsn plugin, and perhaps avoid having to reflash from scratch.

At least until we have a fix.

Confirmed That works. When crash occurs remove internet then unplug power and re power up. boots normal.

[Titanium]

Deleted only the MSN Weather plugin and connected to ethernet. After a while, the error occurred again.

With the MSN Weather plugin and Glamour Aura skin removed then connected to ethernet, the STB remains stable.

[Megatron817]

Deleted only the MSN Weather plugin and connected to ethernet. After a while, the error occurred again.

With the MSN Weather plugin and Glamour Aura skin removed then connected to ethernet, the STB remains stable.

Yeah the weatherMsn folder at usr/lib/enigma2/python/plugins/Extensions has to be deleted. Seems just uninstalling with remove plugin (red button) didn't really uninstall it. I used dream-explorer plugin to delete that folder.

[el bandido]

Currently the feeds and all online images have been deleted. Doing this was probably over-kill, but it was done before a lot of information about this problem existed.

The last updates for TNAP would have been a couple of weeks ago. Nothing has been added or changed to the public TNAP since that time.
The Glamour skin was last updated in April 2021 according to the information at Github. So the image was updated a couple weeks ago, and the Glamour skin around a month ago. But the rebooting or bootlooping started around 24 hours ago.... Go figure...

I have said this before but it needs to be known: We do not push updates or anonymously send things to the receiver. On the surface, it looks like a bad update, but the image was last modified some two weeks ago, and the Glamour skin modified a couple weeks before that (April).

It would be a good idea to look at what feeds from the internet and what other plugins (bitrate, weather, etc.) the Glamour skin uses. Nothing is certain, but it seems to be something that is feeding or getting information from the Internet.

Something that is easy to do will be to build test images with some items missing such as the weather plugin. The problem needs to be narrowed down, then pinpointed and fixed. Unfortunately it might take some time to do this.

[el bandido]

For anyone interested, here is the error:

IOError: [Errno 22] Invalid argument
Traceback (most recent call last):
File "/usr/lib/enigma2/python/mytest.py", line 211, in processDelay
callback(*retval)
File "/usr/lib/enigma2/python/Tools/BoundFunction.py", line 9, in __call__
File "/usr/lib/enigma2/python/mytest.py", line 464, in runNextScreen
session.openWithCallback(boundFunction(runNextScre en, session, screensToRun[1:]), screen, *args)
File "/usr/lib/enigma2/python/mytest.py", line 290, in openWithCallback
dlg = self.open(screen, *arguments, **kwargs)
File "/usr/lib/enigma2/python/mytest.py", line 300, in open
dlg = self.current_dialog = self.instantiateDialog(screen, *arguments, **kwargs)
File "/usr/lib/enigma2/python/mytest.py", line 243, in instantiateDialog
return self.doInstantiateDialog(screen, arguments, kwargs, self.desktop)
File "/usr/lib/enigma2/python/mytest.py", line 260, in doInstantiateDialog
dlg = screen(self, *arguments, **kwargs)
File "/usr/lib/enigma2/python/Screens/InfoBar.py", line 66, in __init__
File "/usr/lib/enigma2/python/Screens/InfoBarGenerics.py", line 268, in __init__
File "/usr/lib/enigma2/python/Screens/Screen.py", line 131, in show
File "/usr/lib/enigma2/python/Components/Renderer/Renderer.py", line 10, in onShow
File "/usr/lib/enigma2/python/Components/Element.py", line 99, in setSuspend
File "/usr/lib/enigma2/python/Components/Element.py", line 104, in checkSuspend
File "/usr/lib/enigma2/python/Components/Element.py", line 94, in setSuspend
File "/usr/lib/enigma2/python/Components/Converter/Poll.py", line 32, in doSuspend
File "/usr/lib/enigma2/python/Components/Converter/Poll.py", line 25, in poll
File "/usr/lib/enigma2/python/Components/Converter/GlamMSNWeather.py", line 2315, in changed
Converter.changed(self, (self.CHANGED_POLL,))
File "/usr/lib/enigma2/python/Components/Element.py", line 86, in changed
File "/usr/lib/enigma2/python/Tools/CList.py", line 7, in __call__
File "/usr/lib/enigma2/python/Components/Renderer/GlamPiconUni.py", line 39, in changed
sname = self.source.text
File "/usr/lib/enigma2/python/Components/Element.py", line 15, in wrapper
File "/usr/lib/enigma2/python/Components/Converter/GlamMSNWeather.py", line 827, in getText
BETA = math.acos((math.cos(90.35 * DEG2RAD) - math.sin(DEC * DEG2RAD) * math.sin(lat * DEG2RAD)) // (math.cos(DEC * DEG2RAD) * math.cos(lat * DEG2RAD))) * RAD2DEG # hour angle
ValueError: math domain error


It looks like the program is computing a value for BETA and it is coming back as 0 or as a negative number. So it throws the IOError and tries to restart.
The million dollar question is why did the BETA calculation suddenly start returning an illegal value and what can we do to go around it or fix it?

[el bandido]

Looks like a possible fix. (See screencap)

17191

Should be able to put some test images out tomorrow. If it works for me, then hopefully it will work for everyone.

[el bandido]

TNAP images were uploaded to this site at the end of 2020.
The 2020 TNAP4 MIO+ images are here:[Only registered and activated users can see links]
These 2020 images have the exact same bootloop issue for me. So it looks like the data that the weather plugin grabs changed which in turn caused the IOError.

I will leave the Test Glamour skin running for the rest of the night and will check it in the AM. Should be pretty obvious if it is not working.

[MikeB]

How about just removing the Weather plugin from the images entirely? I really don't see any need to check weather on my FTA satellite receiver. I get mine from Accuweather, and 1Weather on my phone.

[el bandido]

The weather is in the skins because some users wanted it. Personally I use PLi HD night skin for simplicity. Different users have different preferences.

If we remove the weather plugin, then there will be holes or blank spaces in more than one skin. So to remove the weather plugin correctly, then we would need to modify every skin that uses the weather plugin and remove the blank spaces for the weather displays.

Then what do we do the next time there is a problem with a plugin or even a part of the image? Do we just remove it too???

This weather plugin issue will be fixed.

BTW, This is not just a TNAP problem. Other images that use the same converter and display the weather are affected by this.

[MCelliotG]

A hopefully permanent fix will be applied tonight on my Github!
And please next time I will really appreciate it if anyone with a problem contacts me on my email, it will certainly help issues get fixed sooner!

[MCelliotG]

There you go!
[Only registered and activated users can see links]-a07769c3a1eca4212a90f87b223a5d339e5f5a3a5f0cd8d61b 873b403d784b03
[Only registered and activated users can see links]

[el bandido]

The feeds are up with a Glamour skin update that seems to solve the issue.

Installing a single .ipk file may be easier, and does not require an internet connection to the receiver.

Any TNAP image that is less than 6 months old may have this issue. The feeds will be updated and rebuilt. New images will be made available, but these things take time.

For tonight and until things can be completly corrected, you may update online using the receiver menus OR use telnet with this one line command: opkg update && opkg upgrade && reboot

As an alternative, you can install the attached Glamour skin plugin which is an .ipk file.
To install the attached plugin, download the plugin and install it on a Known To Be Good usb stick, insert the usb stick into the receiver while watching a tv channel (with or without a signal), then follow onscreen instructions.

[MikeB]

I just downloaded the GlamourAura fixed skin, and re-installed the WeatherMSN plugin. I just started another full backup right now, and then I'll switch back to the GlamourAura skin, and see what happens. I'll re-post results.


Ok, backup finished, I switched back to the GlamourAura skin after previously downloading the fix through the feeds, and it rebooted. Everything now appears to be working perfectly once again!

I love that particular skin, and I give kudos and thanks to MCelliotG for a quick fix, and just keeping this maintained in general. I'm sure it's a LOT of work to do so. Also, kudos to El Bandido, as I also really appreciate his expertise, and help in general.