v16.1-20170310 update makes my T3 unstable

zebitty · Post by **zebitty** » Fri Apr 07, 2017 15:01

Since upgrading to beyonwizt3-v16.1-20170310 last week, my T3 has become quite unstable.

On different occasions, I have been doing different things, so I can't even suggest that this issue is linked to a specific feature or action.

Examples include:
* hangs while navigating the epg
* hangs while 'zap'ing to a different channel from within the epg
* hangs during playback of a previously recorded show. This has happened with shows that were recorded both before and after the date I performed the firmware update
* hangs while looking at the timers that have been set via ice-tv
* glitchy playback with lots of pauses and jumps in playback. This happened with a show that was recorded (and watched) under the old firmware and there were no playback issues, so I'm fairly confident the glitchy playback is not due to a corrupted file. Additionally, if I reboot the T3 and watch the same recording, it plays fine (after the reboot).

Each time these problems happen, the unit becomes unresponsive to the remote and the control buttons on the front of the T3 including the power button. The only way to regain control is to use the power/rocker switch on the back to power cycle the T3 and wait for the reboot to complete.

Sometimes these issues happen within a few minutes of turning on the T3. Other times I can watch TV for a few hours before a problem occurs, but I don't think I have had a single problem free day since the firmware upgrade a week ago.

I should also point out that I was not experiencing any of these issues before the firmware update last week, so it seems fairly unlikely these problems are the result of a hardware issue (eg corrupted/failing hard disk, faulty memory etc)

Is anyone else experiencing these (or other issues) after upgrading to v16.1-20170310 ?
Zeb.

MrQuade · Post by **MrQuade** » Fri Apr 07, 2017 15:10

zebitty wrote: Is anyone else experiencing these (or other issues) after upgrading to v16.1-20170310 ?

There haven't been any reports of behaviour quite like that.

You didn't restore your old settings I hope?

If you did restore your old settings, then you should re-flash and set up from scratch.

If you did set up your box from scratch after the USB update then I would advise you to flash from USB again and not restore settings to see if the problems go away.

If the problems go away then you can try restoring your settings. If the problems return then the issue is something in your settings. In which case re-flash *again* and set up from scratch.

peteru · Post by **peteru** » Fri Apr 07, 2017 17:16

It totally sounds like a failing power supply in the T3. The firmware update seems like a red herring.

Paul_oz53 · Post by **Paul_oz53** » Fri Apr 07, 2017 21:09

I have been experiencing similar behaviour on my T3 despite having had the PS repaired by Bwiz.

Have reinstalled firmware several times and setup from scratch but still experience this behavior. Tonight tried again with a full install but no joy.

I have had repeated freezes in the EPG, setup menu, live TV and media player. There is no pattern. It can be anything from under a minute from a reboot to hours later before it happens.

Looked for logs but they don't exist - the folder is always empty.

Thought it may have had something to do with tuners as per the thread discussing T3 freezes. Reseating is a dead end for me.

Although I participated in the alpha and beta stage, the problem did not arise until 20170310 firmware. It does not affect my T4 which is setup near identically.

Could there be something in that version that triggers this?

I have reinstalled 20170226 and the immediate effect is no freezes but it's early days. I'll let it run and report if it fails.

The skin is easy-aus. Standard plugin set - no additions or removals. Standard keymap.

Paul.

PS Unit responds to telnet while frozen. The remote red light flashes in the front panel though unresponsive to the remote. The freeze can be either the picture or the menu overlay with an active picture but seems to be the foreground object freezing.

zebitty · Post by **zebitty** » Fri Apr 07, 2017 22:35

MrQuade wrote:You didn't restore your old settings I hope?

Correct. Old settings were not restored. In fact, I printed the PDF of that upgrade guide (with the blue bar at the top) and followed the steps as written.

The firmware update didn't seem to do anything unusual. After the reboot, the setup wizard ran as expected.
I specified my region, scanned for services, entered my icetv login details etc. The only small hiccup was the timers from icetv wouldn't seem to update, althought the epg data did. I had to go to the icetv website and 'resend recordings', after which the timers updated correctly.

peteru wrote:It totally sounds like a failing power supply in the T3. The firmware update seems like a red herring.

I used to work in an enterprise scale IT environment as a network engineer and I'm pretty familiar with troubleshooting techniques when diagnosing problems. I've seen my share of weird coincidences and I know how these things can sometimes go, so obviously I can't definitively rule out an intermittent hardware failure such as a failing power supply, but almost 20 years of doing that kind of work (admittedly on server farms and routing/switching gear) it gives you a gut feel for these sorts of things, and I have to say, the timing is _very_ suggestive of the update being the culprit. Coupled with the knowledge that before the firmware update, my T3 was rock solid for the ~2-3 years since I purchased it, right up to _literally_ after the firmware update, when it started to exhibit all sorts of problems.

Some more background info:
It was running a fairly old version of the firmware (which was working fine for my purposes) but I knew there was a new firmware version due out in March/April, so about a month ago I did a boot loader update, to prepare for the new firmware that was due out soon.

The T3 worked without fault for a few weeks after the boot loader update so I went ahead with the firmware update, which appeared to go smoothly, and that's when all the issues started.

Also, the T3 and the network hub behind the TV are both plugged into a UPS, so there's surge protection in place.
I don't use the wifi features of the T3 at all.
I have not installed any additional addons etc onto the T3.

My money's on the firmware changes being responsible here, but it pays to be methodical, so to try and rule in/out a hardware issue, can you suggest things I might do to further test this posibility?

MrQuade · Post by **MrQuade** » Fri Apr 07, 2017 22:45

zebitty wrote: Coupled with the knowledge that before the firmware update, my T3 was rock solid for the ~2-3 years since I purchased it, right up to _literally_ after the firmware update, when it started to exhibit all sorts of problems.

The thing is, that since September or October last year, a lot of T3's have started to die all of a sudden. It is not out of the question that a series of reboots and a firmware upgrade have pushed a marginal unit over the edge.

However, there is a hint that perhaps the latest USB image may be the cause of the issue. My T3 use a slightly old version of the 16.1 image and was then updated online. I intend to load on the latest image and test my T3 to see if I can replicate any of these lockups this weekend.

zebitty · Post by **zebitty** » Fri Apr 07, 2017 22:49

Thanks. Sounds like a good start.

I'll keep an eye on this thread to hear how you got on and see if we can sort out the next step after you've done your testing.

In the mean-time, can you point me at a guide to step me thru an efficient way of archiving my recordings (maybe to an external hdd?) and I'll try formatting the disk and performing a disc scan. Might as well make sure the HDD in it is in good working order and can be ruled out as a possible cause.

peteru · Post by **peteru** » Fri Apr 07, 2017 23:08

As far as an efficient way of cloning the HDD contents over network, the firmware has ssh/rsync installed, so that's your best bet. I would not bother reformatting the HDD until there is some evidence of it being a source of issues, but there's no harm (most of the time) in taking a backup.

One way of confirming that the issue is not hardware related would be to go back to a 4.4 series firmware for a few days and if you don't see any stability issues then we have a strong case for suspecting the firmware.

If it is the firmware, then there will be some head scratching to be done. During the very long beta cycle, there were no issues reported. With the exception of the drivers and kernel, the firmware for T2/T3/T4 is built from the same source with almost identical settings. That's not to say that a firmware issue is not there, but given the reported symptoms and the recent spate of T3 power supply issues, power supply related problems seem like the prime suspect.

Next time you get a lockup, see if you can telnet/ssh into your T3 and post the output of dmesg - the most likely cause would be a driver crash or kernel oops and those won't show up in enigma2 crash logs.

Paul_oz53 · Post by **Paul_oz53** » Sat Apr 08, 2017 03:01

peteru wrote: One way of confirming that the issue is not hardware related would be to go back to a 4.4 series firmware for a few days and if you don't see any stability issues then we have a strong case for suspecting the firmware.

Hi peteru,
I have reverted to 20170226 and the bad behavior stopped immediately whereas, on 20170310, it would freeze within a few minutes of intense use of the remote. It's not the main unit and if its left alone its fine on 20170310. Recording, network access, streaming, playback across the network are all ok. Problem only arises if I use it as a TV and use the remote a lot.

I have not had a problem after a couple of hours of intense remote usage on 20170226. I have downloaded and reinstalled 20170310 three times with clean setups and each time experienced the issue before reverting to the earlier firmware.

If it is the firmware, then there will be some head scratching to be done. During the very long beta cycle, there were no issues reported. With the exception of the drivers and kernel, the firmware for T2/T3/T4 is built from the same source with almost identical settings. That's not to say that a firmware issue is not there, but given the reported symptoms and the recent spate of T3 power supply issues, power supply related problems seem like the prime suspect.

As mentioned elsewhere, my PS was repaired by Bwiz but I don't dismiss the possibility out of hand. It may still be a residual issue. However, as you say, reverting to earlier firmware should give a clue. If it stays stable for a day or two I will reinstall 20170310 and see if the problem returns.

As one of the testers I agree completely, the Beta was certainly fine right up to the official release. This only started after I did the USB install.

Next time you get a lockup, see if you can telnet/ssh into your T3 and post the output of dmesg - the most likely cause would be a driver crash or kernel oops and those won't show up in enigma2 crash logs.

Where should I look? I have been able to telnet in when it's frozen (normally to reboot) so it should be possible to help here. Is there anything I need to set to activate the dmesg logging?

I truly appreciate the enormous effort that has gone into the firmware so anything I can do to help solve this I'm happy to do.

peteru · Post by **peteru** » Sat Apr 08, 2017 03:58

Once you telnet in, just type dmesg at the prompt. This will print a lot of stuff, possibly many screens.

We are interested in anything that may be indicative of an error or failure. If you see anything interesting/suspicious, please copy everything that dmesg prints and paste it here in the forums, inside [code] tags, to preserve formatting.

zebitty · Post by **zebitty** » Sat Apr 08, 2017 13:54

peteru wrote:One way of confirming that the issue is not hardware related would be to go back to a 4.4 series firmware for a few days and if you don't see any stability issues then we have a strong case for suspecting the firmware.

I'll give that a go. I can only see the 16.1 download in the beyonwiz downloads area. Where can I find older versions of the firmware?

If it turns out that using the remote is part of what's causing the problem (as Paul theorised), I should note that if I switch back to 4.4, I don't think my current remote will work anymore.

My original remote that came with the T3 died a few months back and I purchased a Beyonwiz Multi Backlit Remote. Initially the new remote wasn't detected by the T3 and only after I applied the new firmware update could the T3 could detect the new backlit remote.

Currently the T3 is set to 'detect all' for remotes but won't reverting to 4.4 stop the T3 from detecting the new backlit remote?

MrQuade · Post by **MrQuade** » Sat Apr 08, 2017 14:24

zebitty wrote: Currently the T3 is set to 'detect all' for remotes but won't reverting to 4.4 stop the T3 from detecting the new backlit remote?

No. by default, the 4.4 series firmware also responds to all supported remotes.

prl · Post by **prl** » Sat Apr 08, 2017 16:20

zebitty wrote:... Where can I find older versions of the firmware? ...

In the topics they were published in in the first place.

To get the most recent firmware prior to 20170310, fetch firmware version 20160817 from New Firmware for all T-Series Models (T2/T3/T4) 17/08/2016

MrQuade · Post by **MrQuade** » Sat Apr 08, 2017 21:18

So far, no joy in reproducing the fault.

I have loaded from the latest USB image and restore my settings, then applied the online update

I am using the original remote and am spamming the T3 as hard as I can, zapping from the EPG and shipping round the media player.

I tried turning the DLNA server back on (I usually switch it off) and no joy.

My T3 was one of the very first batch of public units. I ordered mine and one for a mate at the same time. His has suffered from a hardware failure in the last month, but mine has been rock solid so far.

I'll keep smashing it as much as I can.

Paul_oz53 · Post by **Paul_oz53** » Sat Apr 08, 2017 23:43

MrQuade wrote:I'll keep smashing it as much as I can.

Sounds exactly like what I was doing.

Do try the main menu, setup menus and media player menu too.

Having gone back to the penultimate beta, mine is rock solid at the moment. This would appear to eliminate a hardware fault.

That some others report similar symptoms suggests it's either the firmware or the download load at fault.

The obvious mystery is why only some units are affected and not all T3's. I can accept an occasional bad download but having done it three times it seems unlikely that is the cause.

Recently I spent a small fortune upgrading the antenna installation so that is no longer an issue.

All in all, a frustrating and unexpected problem.

Thanks for the interest and effort.
Paul

prl · Post by **prl** » Sat Apr 08, 2017 23:57

Paul_oz53 wrote:... I can accept an occasional bad download but having done it three times it seems unlikely that is the cause. ...

ZIP files have a CRC32 checksum on each file in the archive, so undetected corruption of the download is unlikely.

MrQuade · Post by **MrQuade** » Sun Apr 09, 2017 02:28

Paul_oz53 wrote: Do try the main menu, setup menus and media player menu too.

Yep, been hitting the main Setup menu too.

I haven't been into too many detailed config areas though.

I am wondering if this might be getting triggered by one of those "things that Quade doesn't do with his Wiz on a regular basis".
Some sequence of actions like the old up/down bug that kept hitting people who regularly accessed the main menu while playing recordings......some users never did that so never saw the bug, yet others who liked tweaking timers while watching recordings were hit with it constantly. Obviously the issue at hand is not triggered by that particular sequence, but something like that.

Paul_oz53 · Post by **Paul_oz53** » Sun Apr 09, 2017 03:57

I believe what I was doing was pretty normal stuff. Last time it happened from live TV I was in menu-setup-tuners after checking the AV settings and system settings.

I have been testing how well my usb tuner works with a reduced signal to prevent breakup (very well in fact). I set the preferred tuner to D, the recording tuner to D and scrolled down to fallback remote receiver url.

With a tv picture playing in the background it froze on that setting line and stopped responding to the remote. The front panel red light flashed as buttons were pressed but no response and no spinner.

Telnet worked and it started a reboot but locked up at 65%. At that point I decided to reinstall the earlier firmware.

I'll reload 20170310 tomorrow and see what happens.

And yes, I was an up/down victim too!

PS Sorry no quote - too hard with this phone.

prl · Post by **prl** » Sun Apr 09, 2017 08:01

Do you have "MENU>Setup>TV>Channel selection>Enable multiple bouquets" enabled (default) or disabled?

Paul_oz53 · Post by **Paul_oz53** » Sun Apr 09, 2017 15:56

prl wrote:Do you have "MENU>Setup>TV>Channel selection>Enable multiple bouquets" enabled (default) or disabled?

Hi prl,

Multiple bouquets is Enabled. Never turn it off these days. Only setting I conciously changed was setting the info bar timeouts to 90 secs while testing the usb tuner. Currently at 20 secs.

Updated last night from 20170226 to 20170311 by internet and, so far, it's stable.

Bearing in mind the T3 was ok in the Beta phase right up until the USB update, can I check please: would this be the same firmware state it was in as the last beta update via the internet feed prior to the official release (ignoring customisations)?

You may recall I recently tinkered with keymap.xml to substitute 'mute' for a broken 'exit' key. My problems only began after that bout of tinkering. At one stage, it was locked in a bootloop so I downloaded and did the usb update to recover. Restored settings but took care not to restore the faulty keymap.xml file. I inspected the usr/share/enigma2 keymap.xml with a unix style text editor (Editpad Lite 7) to be sure it was the correct version.

Then the freezing troubles began. Redownloaded the usb update and went to the clean install routine using a different usb port and usb drive. All default settings - no customisation, no restore settings, no internet update after install. Checked it was working but the damn thing would freeze and not respond to inputs except the power switch. Didn't try telnet initially, not until about the fourth time it froze but telnet was available.

On the third software install did both the usb and the internet update but no other customisations. Still had the freezing within minutes of use of the remote. Which then led to trying a reversion to the earlier version of firmware (successfully).

I take your earlier point that the CRC check should largely prevent a corrupt unzip. The only clue to this seems to be that for some strange reason, the usb update process is not working properly for some users, sometimes and for no apparent reason. The positive side seems to be that if you persist, it comes good eventually. Today it has behaved as expected.

Given it seems a small minority of users are having a problem i'm inclined to let it go for now and monitor how it behaves. If it freezes again I'll capture dmesg and post as peteru suggested. Without something concrete to analyse, this is unlikely to progress.

Paul

PS. Checked the usb ports with a range of devices and they are all working as normal today.

prl · Post by **prl** » Sun Apr 09, 2017 16:58

Paul_oz53 wrote:...
Bearing in mind the T3 was ok in the Beta phase right up until the USB update, can I check please: would this be the same firmware state it was in as the last beta update via the internet feed prior to the official release (ignoring customisations)?

I don't know the answer to that (and it's probably difficult to answer). The USB release files for the beta and the official 20170310 releases are identical (I saved a copy of the beta version). There was a beta stream online update for 20170311, but I don't know what was in that.

Paul_oz53 wrote:... Given it seems a small minority of users are having a problem i'm inclined to let it go for now and monitor how it behaves. If it freezes again I'll capture dmesg and post as peteru suggested. Without something concrete to analyse, this is unlikely to progress.
...

Lockups like this are often due to the code running off into an infinite loop. You can check whether that's the case by logging in to the T3 and running the command "top" if logging in is possible.

By default "top" displays a dynamically updating list of the top CPU-consuming processes. If enigma2 is at the top and its %CPU is stuch at, or near 100, it's gone off into an infinite loop. It should not be using more than a few percent of the CPU when you're just in live TV.

Type "q" to exit "top".

Paul_oz53 · Post by **Paul_oz53** » Sun Apr 09, 2017 19:06

prl wrote:
Paul_oz53 wrote:...
Bearing in mind the T3 was ok in the Beta phase right up until the USB update, can I check please: would this be the same firmware state it was in as the last beta update via the internet feed prior to the official release (ignoring customisations)?
I don't know the answer to that (and it's probably difficult to answer). The USB release files for the beta and the official 20170310 releases are identical (I saved a copy of the beta version).

Thanks, I thought that was the case. Since I did the online update it's skipped straight to 20170311. Like you, I retain old copies of firmware so it's easier to go back.

Lockups like this are often due to the code running off into an infinite loop. You can check whether that's the case by logging in to the T3 and running the command "top" if logging in is possible.

By default "top" displays a dynamically updating list of the top CPU-consuming processes. If enigma2 is at the top and its %CPU is stuch at, or near 100, it's gone off into an infinite loop. It should not be using more than a few percent of the CPU when you're just in live TV.

Type "q" to exit "top".

I'll do that - it could indeed be the problem.
Paul

gawaterman · Post by **gawaterman** » Tue Apr 11, 2017 18:47

zebitty wrote:Since upgrading to beyonwizt3-v16.1-20170310 last week, my T3 has become quite unstable.

On different occasions, I have been doing different things, so I can't even suggest that this issue is linked to a specific feature or action.

Examples include:
* hangs while navigating the epg
* hangs while 'zap'ing to a different channel from within the epg
* hangs during playback of a previously recorded show. This has happened with shows that were recorded both before and after the date I performed the firmware update
* hangs while looking at the timers that have been set via ice-tv
* glitchy playback with lots of pauses and jumps in playback. This happened with a show that was recorded (and watched) under the old firmware and there were no playback issues, so I'm fairly confident the glitchy playback is not due to a corrupted file. Additionally, if I reboot the T3 and watch the same recording, it plays fine (after the reboot).

Each time these problems happen, the unit becomes unresponsive to the remote and the control buttons on the front of the T3 including the power button. The only way to regain control is to use the power/rocker switch on the back to power cycle the T3 and wait for the reboot to complete.

Sometimes these issues happen within a few minutes of turning on the T3. Other times I can watch TV for a few hours before a problem occurs, but I don't think I have had a single problem free day since the firmware upgrade a week ago.

I should also point out that I was not experiencing any of these issues before the firmware update last week, so it seems fairly unlikely these problems are the result of a hardware issue (eg corrupted/failing hard disk, faulty memory etc)

Is anyone else experiencing these (or other issues) after upgrading to v16.1-20170310 ?
Zeb.

Yes, mine has been doing exactly the same as yours and only after the new firmware upgrade. Am going to revert to previous version and see what happens.

Paul_oz53 · Post by **Paul_oz53** » Tue Apr 11, 2017 21:01

gawaterman wrote:Yes, mine has been doing exactly the same as yours and only after the new bios upgrade. Am going to revert to previous version and see what happens.

After a period of instability, mine is now quite stable on the new firmware.

Repeating the update is definitely worth trying, despite the inconvenience.
Paul

gawaterman · Post by **gawaterman** » Fri Apr 14, 2017 16:07

Paul_oz53 wrote:
gawaterman wrote:Yes, mine has been doing exactly the same as yours and only after the new bios upgrade. Am going to revert to previous version and see what happens.
After a period of instability, mine is now quite stable on the new firmware.

Repeating the update is definitely worth trying, despite the inconvenience.
Paul

Yes , have done 3 installs, 2nd one seemed to be more stable and so far no problems after 3rd install.
should mention 3rd install was off a different usb.

Paul_oz53 · Post by **Paul_oz53** » Fri Apr 14, 2017 16:23

gawaterman wrote: Yes , have done 3 installs, 2nd one seemed to be more stable and so far no problems after 3rd install.
should mention 3rd install was off a different usb.

Interesting - I too changed usb sticks for the successful install. Maybe the instability cause is a misread of the usb stick. Would explain why my initial reinstalls were unsuccessful.

This would also explain why only some users encounter the problem.

Paul.

gawaterman · Post by **gawaterman** » Mon Apr 17, 2017 11:19

Update: Yesterday had several crashes, so looks like it wasn't fixed.
Going to go back to previous firmware and see how that goes.

Paul_oz53 · Post by **Paul_oz53** » Tue Apr 18, 2017 12:50

gawaterman wrote:Update: Yesterday had several crashes, so looks like it wasn't fixed.
Going to go back to previous firmware and see how that goes.

Sorry to also report that once again the T3 froze up completely while editing in Menu->Setup->TV->Recordings. I was checking the fast forward speed settings. No response to any key on the remote and unable to access via Telnet or windows explorer.

As previously noted, it was ok until a period of heavy usage of the remote. At the time, was checking out a bug in the fast forward speed setting using Openwebif, which bug is discussed elsewhere.

P.

Paul_oz53 · Post by **Paul_oz53** » Thu Apr 20, 2017 15:53

OK last night, yet another freeze. This time playing back an edited show - used Cutlist and Execute Cuts (as implemented by Adoxa) to produce the file so a few keypresses were involved.

I did manage to Telnet in - it's 50/50 if it will work - and ran TOP and DMESG. Also retrieved the dmesg and messages files from var/logs/. Had to add an extension of '.txt' to post though - I can see why prl complains now. There was no relevant crashlog.

A screenshot of TOP is attached. Not clear from this but TOP was running 85-100% with kthread flickering up and down.

The snippet below is from dmesg > freeze_file.txt. The line "{21718}[fp_mcu_cmd_write] full!!! " repeats for many pages.

Hopefully, one of the Gurus can gleen something from these to suggest a cause or a cure. Regardless, I am going to pull out the HDD and sata cable over the weekend since I have many spares. I ran the disk check function and it completed ok but I'm getting desperate.
Thanks,
Paul

Code: Select all

{21718}[fp_mcu_cmd_write] full!!! 
{21718}[fp_mcu_cmd_write] full!!! 
{21718}[fp_mcu_cmd_write] full!!! 
{21718}[fp_mcu_cmd_write] full!!! 
{21718}[fp_mcu_cmd_write] full!!! 
{21718}[fp_mcu_cmd_write] full!!! 
{21718}[fp_mcu_cmd_write] full!!! 
{21718}[fp_mcu_cmd_write] full!!! 
{21718}[fp_mcu_cmd_write] full!!! 
{21718}[fp_mcu_cmd_write] full!!! 
BUG: scheduling while atomic: enigma2/21718/0x00000100
Modules linked in: ipv6 dvb_usb_it913x it913x_fe bcm(O) dvb(OF)
Call Trace:
[<80638df0>] dump_stack+0x8/0x34
[<80057f74>] __schedule_bug+0x48/0x58

freeze_file.txt: (9.97 KiB) Downloaded 44 times

dmesg.txt: (11.03 KiB) Downloaded 59 times

messages.txt: (139.79 KiB) Downloaded 48 times

MrQuade · Post by **MrQuade** » Thu Apr 20, 2017 16:17

Looks like it is talking about the USB tuner.

Are you able to try running without the USB tuner for a while?

Paul_oz53 · Post by **Paul_oz53** » Thu Apr 20, 2017 16:44

MrQuade wrote:Looks like it is talking about the USB tuner.

Are you able to try running without the USB tuner for a while?

Yes MrQuade, I'll unplug it now.

Actually the usb tuner has been working fine but it's not the default currently. I ran it on test for over a week as the primary tuner without a problem. Its only when I try changing something (channels or settings ) or playback that it throws a hissy fit.

I take it your basing this thought on the dmesg file where it is the last device mentioned.

Paul

MrQuade · Post by **MrQuade** » Thu Apr 20, 2017 16:49

Paul_oz53 wrote: I take it your basing this thought on the dmesg file where it is the last device mentioned.

dmesg shows the tuner in a "cold" state and is then initialised to a "warm" state.

It was more the end of the messages file that shows:

Code: Select all

BUG: scheduling while atomic: enigma2/21718/0x00000100
Modules linked in: ipv6 dvb_usb_it913x it913x_fe bcm(O) dvb(OF)

peteru · Post by **peteru** » Thu Apr 20, 2017 16:51

Is your front panel working properly? Have you updated your front panel to 20057?

Paul_oz53 · Post by **Paul_oz53** » Thu Apr 20, 2017 17:07

peteru wrote:Is your front panel working properly? Have you updated your front panel to 20057?

Yes, quite some time ago and it works properly. Note that I generally leave the machine on 24/7 so I don't use the timer functions for standby.

prl · Post by **prl** » Thu Apr 20, 2017 17:21

MrQuade wrote:
Paul_oz53 wrote: I take it your basing this thought on the dmesg file where it is the last device mentioned.
dmesg shows the tuner in a "cold" state and is then initialised to a "warm" state.

It was more the end of the messages file that shows:
Code: Select all
BUG: scheduling while atomic: enigma2/21718/0x00000100
Modules linked in: ipv6 dvb_usb_it913x it913x_fe bcm(O) dvb(OF)

I think that's just crash logging information about installed kernel modules. Like this, though crash list isn't quite the same:

Code: Select all

root@beyonwizt4:~# lsmod
Module                  Size  Used by
ipv6                  376754  29
bcm                     8191  0
dvb                 58507441  30
root@beyonwizt4:~#

As peteru's question suggests, I think that "fp_mcu" is Front Panel Microcontroller Unit. The final cause of the crash is that the kernel scheduler has got deeply confused (BUG: scheduling while atomic).

A search on Google for fp_mcu_cmd_write just brings up two pages from 2013, both in German. I couldn't find anything helpful in either.

peteru · Post by **peteru** » Thu Apr 20, 2017 17:27

Are you running any non-standard skins, plugins or modifications?

The symptoms indicate that something is hammering the front panel really hard and it can't keep up.

What's the front panel showing when you start getting these messages?

Code: Select all

Apr 20 02:06:06 beyonwizt3 user.warn kernel: {21718}[fp_mcu_cmd_write] full!!!

I see no indication at all that the USB tuner is involved in any way.

MrQuade · Post by **MrQuade** » Thu Apr 20, 2017 17:31

prl wrote: As peteru's question suggests, I think that "fp_mcu" is Front Panel Microcontroller Unit. The final cause of the crash is that the kernel scheduler has got deeply confused (BUG: scheduling while atomic).

A search on Google for fp_mcu_cmd_write just brings up two pages from 2013, both in German. I couldn't find anything helpful in either.

Ok, sounds fair enough. I also had a Google search for that message, and saw those two articles. I wasn't able to make the connection with the front panel from fp_mcu though....my bad.

it probably makes sense that the front panel is involved, if these incidents all followed high levels of remote control activity.

Paul_oz53 · Post by **Paul_oz53** » Thu Apr 20, 2017 17:49

peteru wrote:Are you running any non-standard skins, plugins or modifications?

The symptoms indicate that something is hammering the front panel really hard and it can't keep up.

What's the front panel showing when you start getting these messages?
Code: Select all
Apr 20 02:06:06 beyonwizt3 user.warn kernel: {21718}[fp_mcu_cmd_write] full!!! 
I see no indication at all that the USB tuner is involved in any way.

The skin is Overlay HD 1.61. Extra plugins are series2folder and reconstructapsc. Removed gmail and webcams.

The front panel is displaying the appropriate message relative to the context it is in. I did alter the date setting and turn down the brightness numbers at some point recently.

Paul_oz53 · Post by **Paul_oz53** » Thu Apr 20, 2017 18:40

Did some testing with the extra plugins removed.

If I reboot and Telnet in and run dmesg, the {nnn}[fp_mcu_cmd_write] full!!! line is not present. Then press Media. The message appears after a few seconds.

Repeat the reboot and Telnet/dmesg but this time press Channel Up (default is ABC HD). The {nnn}[fp_mcu_cmd_write] full!!! message appears after a few seconds.

Swapped back to easy-aus-skin. Restart GUI. Telnet in and run dmesg, the {nnn}[fp_mcu_cmd_write] full!!! line is not present. Then press Media. The message appears after a few seconds.

So, pressing either key triggered the message almost instantly after a full restart or GUI restart on the standard skin and no extra plugins.

Could it be a front panel setting at fault?

peteru · Post by **peteru** » Thu Apr 20, 2017 21:23

Looks like front panel dimming on the T3 issues too many requests as it tries to do a fade. Turn off the dimming and see if that helps.

peteru · Post by **peteru** » Thu Apr 20, 2017 22:10

I have a fix for the flood of dim commands that the code was sending. It's been pushed on the next branch, so it will be a while before it makes it into an official firmware update. I might cherry pick it for a 16.1 beta release if I have time.

This is only likely to be an issue with the dimming of the T3 front panel. T2 and T4 should be unaffected.

Paul_oz53 · Post by **Paul_oz53** » Thu Apr 20, 2017 22:19

peteru wrote:Looks like front panel dimming on the T3 issues too many requests as it tries to do a fade. Turn off the dimming and see if that helps.

I think you may be right. Did a factory reset which has set it to off. No longer getting the fp_mcu messages. It is looking a very likely culprit.

I recall an earlier log had a series of messages before the fp_mcu messages that were obviously reducing the front panel display brightness. That it was the front panel locking up also explains why heavy use of the remote was a factor for me - straw breaking the camel's back!

Thanks peteru.

Hopefully this is the cause more generally of these reports of freezing T3s - a setting to be avoided, pending the fix you propose. I'll monitor how it behaves and report whether the solution appears permanent or not...

IanSav · Post by **IanSav** » Fri Apr 21, 2017 03:48

Hi PeterU,

peteru wrote:I have a fix for the flood of dim commands that the code was sending. It's been pushed on the next branch, so it will be a while before it makes it into an official firmware update. I might cherry pick it for a 16.1 beta release if I have time.

If your main reason for halting merges from OpenViX and moving on with 16.1 is the timezone issue then why don't you simply apply the fix you proposed to OpenViX into the Beyonwiz 16.1 branch and then keep moving forward with the timezone regression tolerant 16.1 branch?

Regards,
Ian.

gawaterman · Post by **gawaterman** » Wed Apr 26, 2017 18:20

gawaterman wrote:Update: Yesterday had several crashes, so looks like it wasn't fixed.
Going to go back to previous firmware and see how that goes.

Update. Went back to previous firmware, but still had lots of lockups, so reinstalled latest version about 5 days ago and havent had a problem since, so looks like you just have to persevere

MrQuade · Post by **MrQuade** » Wed Apr 26, 2017 18:24

gawaterman wrote: Update. Went back to previous firmware, but still had lots of lockups, so reinstalled latest version about 5 days ago and havent had a problem since, so looks like you just have to persevere

What did you have your front panel Dimming setting set as (both on the old and new firmware)? That appeared to be the cause of one major source of lockups.

Paul_oz53 · Post by **Paul_oz53** » Wed Apr 26, 2017 18:43

MrQuade wrote:
gawaterman wrote: Update. Went back to previous firmware, but still had lots of lockups, so reinstalled latest version about 5 days ago and havent had a problem since, so looks like you just have to persevere
What did you have your front panel Dimming setting set as (both on the old and new firmware)? That appeared to be the cause of one major source of lockups.

I asked gawaterman that question in a PM. His reply was that he did not alter the dimming settings.

Regardless, my T3 has been stable ever since turning dimming off.

Paul

zebitty · Post by **zebitty** » Tue Jun 27, 2017 16:21

Great find with the panel dimming issue. Just wanted to add that my T3 has been rock solid for the last 3 months, however, we had a brief power outage today and after rebooting it has reverted to the unstable behaviour that it was exhibiting right after I applied the update (back in April). Three crashes/freeze-ups in 90 mins. Once with lots of remote usage, the other two with no remote usage at all.

I've turned the LCD dimming off and set the brightness for everything to the lowest setting. I'll get back with an update in a few days once I've had time to see if this has any effect.

About the Forum

Join us