• 1
  • 216
  • 217
  • 218(current)
  • 219
  • 220
  • 355
v18 LibreELEC Testbuilds for x86_64 (Kodi 18.0)
New LibreELEC.tv Leia build #0430: Generic
(Supercedes previous build)

SHA256 Checksum: 2399b74f27ae4400bb812fead33ae7f739d07316173c43a9edd9fabbb2ff6ac9 (Generic)

text:
# uname -a
Linux NUC 4.14.37 #1 SMP Mon Apr 30 21:26:34 BST 2018 x86_64 GNU/Linux

# lsb_release
LibreELEC (Milhouse): devel-20180430210323-#0430-gc84b177 [Build #0430]

# Kodi version
(18.0-ALPHA2 Git:04a039dd). Platform: Linux x86 64-bit

Based on tip of LibreELEC.tv master (c84b177, changelog) and tip of XBMC master (96f15e0, changelog) with the following modifications: Build Highlights:
  1. qemu: update to 2.12.0
Build Details:
  1. XBMC:
    • allow plugins to return another plugin path on CPluginDirectory::GetP (PR:13827, 1 commit, 1 file changed)
    • [linux] libinput: remove dependency to CWinEventsLinux (PR:13817, 6 commits, 33 files changed)
  2. libnfs:
    • Define MSG_NOSIGNAL in windows if it doesn't exist (PR:253, 1 commit, 1 file changed)
  3. Additional commits/pull requests/changes not yet merged upstream:
    • Updated: [env] PR:2654 (perma): kodi: Early May 2018
    • Added: [env] PR:2677 (perma): qemu: update to 2.12.0
    • Added: [pkg] PR:99 (perma): Favour non-X keyboard layouts over X (service.libreelec.settings)
Texture Cache Maintenance Utility: Preload your texture cache for optimal UI performance. Remotely manage media libraries. Purge unused artwork to free up space. Find missing media. Configurable QA check to highlight metadata issues. Aid in diagnosis of library and cache related problems.
(2018-04-30, 20:38)TheKraut Wrote: - RC not working: LE #0405, #0429 (and a couple of versions in between) when installed
- RC not working: LE #0429 when run in live mode (KODI log in my original post above; didn't bother to test any other version)
- RC not working: Ubuntu 18.04 LTS in live mode
- RC working: #0429 when booting to text-based installer

By "not", I mean more specifically: It is working for a seemingly random, double digit number of key presses after a reboot. Then, it ceases to work.

No manual config files in either configuration.

I was guessing on a kernel problem with 4.14/4.15. However, given that the installer works fine, it may well be UI related (and shouldn't be hardware related, I presume).
It would be good to know the first build where IR stopped working, so please also test builds before 0405. Several IR changes were added in build 0328, so checking builds 0327 and 0328 would be a good start. If build 0327 doesn't work either please check earlier builds as well.

so long,

Hias
(2018-05-01, 00:38)HiassofT Wrote:
(2018-04-30, 20:38)TheKraut Wrote: - RC not working: LE #0405, #0429 (and a couple of versions in between) when installed
- RC not working: LE #0429 when run in live mode (KODI log in my original post above; didn't bother to test any other version)
- RC not working: Ubuntu 18.04 LTS in live mode
- RC working: #0429 when booting to text-based installer

By "not", I mean more specifically: It is working for a seemingly random, double digit number of key presses after a reboot. Then, it ceases to work.

No manual config files in either configuration.

I was guessing on a kernel problem with 4.14/4.15. However, given that the installer works fine, it may well be UI related (and shouldn't be hardware related, I presume).
It would be good to know the first build where IR stopped working, so please also test builds before 0405. Several IR changes were added in build 0328, so checking builds 0327 and 0328 would be a good start. If build 0327 doesn't work either please check earlier builds as well.

so long,

Hias 
So, I tested #0327 and #1119, since #1120 mentions a major kernel-related IR change. Same issue for both versions, RC stops working after a couple of key presses.

The Harmony also allows to adjust a few device-specific properties: For the NUC, I now set key repetitions to 0 and slowly increased "key delay" and "device delay" to 1000 ms each. Still, the problem persists.

I'm a bit puzzled at this point. Remote still works fine for any other device, and since a reboot temporarily solves the problem, there is a strong indication for a software issue, right?
(2018-04-30, 21:13)Milhouse Wrote: Remember these are test builds, so occasional instability is to be expected - if you need stability over features then release builds would be my recommendation. That said I/we appreciate the testing you all do which probably wouldn't happen if these builds weren't being used as daily drivers so for the time being it might be best if you stick with #0421 until the Python/Profile issues are solved (I've pinged the developer). 
Will stick to #421 then and wait as you have suggested. Appreciate all your efforts ;-)
(2018-05-01, 08:41)TheKraut Wrote: So, I tested #0327 and #1119, since #1120 mentions a major kernel-related IR change. Same issue for both versions, RC stops working after a couple of key presses.
Very odd, I've never seen this before. Could you try going back to older versions and also test the latest stable LE version (8.2.5)? It looks like there is some general problem with IR handling on your system.

To be sure we're not chasing ghosts could you run memtest86 for a couple of hours? About 5 successful passes should be enough to get a bit of confidence there's no hardware fault in play (which can lead to all kinds of odd errors).

If memtest86 was successful it'd be good to get some more info and logs from your system.

Enable debug logging, install the latest build, use kodi until the remote stops working then ssh in and run the following commands / tests and post the output.

First grab some logs and general info:
Code:
dmesg | paste
journalctl -a | paste
paste .kodi/temp/kodi.log
paste /proc/bus/input/devices
lsof | grep /dev/input | paste
ir-keytable

Check if you get output from irw
Code:
irw

If you get no output from irw, stop kodi and eventlircd and then check if you get output from ir-keytable -t
Code:
systemctl stop kodi
systemctl stop eventlircd
ir-keytable -t

If that still gives nothing check with ir-ctl -r
Code:
ir-ctl -r

so long,

Hias
Hi Hias,

Here you go: full log (individual output separated by "*** BEGIN OF" / "*** END OF").

LE #0430, installed to and booted from SD card, no settings changed except debug log and SSH enabled, as well as WiFi connected.
RC stopped working again after a couple of key presses.

Neither irw, nor ir-keytable, nor ir-ctl would give any output after the RC stopped working. However, irw always provides output as expected until the RC fades away.

Memtest86 is running right now, looks good so far. Will report again if I find anything.

Unfortunately, LE 8.2.5 won't work due to a lack of Gemini Lake GPU support in the old kernel.
However, RC has worked flawlessly with my older NUC2820 for many years, up to LE 8.2.4, which was the last release that I had installed on it.

Thanks for your efforts!
(2018-05-01, 13:41)TheKraut Wrote: Hi Hias,

Here you go: full log (individual output separated by "*** BEGIN OF" / "*** END OF").

LE #0430, installed to and booted from SD card, no settings changed except debug log and SSH enabled, as well as WiFi connected.
RC stopped working again after a couple of key presses.

Neither irw, nor ir-keytable, nor ir-ctl would give any output after the RC stopped working. However, irw always provides output as expected until the RC fades away.
Thanks for the logs! Unfortunately there's nothing in there that indicates why the IR receiver stopped working. With ir-ctl -r you should see the raw IR data, actually IR signals from any remote.

I googled a bit and found this thread about pervious year's NUC models and a similar issue as yours: https://communities.intel.com/thread/116512

As you seem to get the same issue on Ubuntu 18.04 I'm wondering if it could be a general problem with the NUCs and Intel needs to fix their BIOS, firmwares etc.

A quick check with Win 10 could be worth a try. If you get the same issue there contact Intel. If it works on Win 10 but not Ubuntu it'd be best if you contacted Ubuntu about that - they have slightly larger support resources than we have so if anything needs fixing in the Linux kernel they are probably better suited for the job than us (and we'd in turn get the fix via standard kernel updates or could pick/backport it).

so long,

Hias
(2018-05-01, 18:59)HiassofT Wrote:
(2018-05-01, 13:41)TheKraut Wrote: Hi Hias,

Here you go: full log (individual output separated by "*** BEGIN OF" / "*** END OF").

LE #0430, installed to and booted from SD card, no settings changed except debug log and SSH enabled, as well as WiFi connected.
RC stopped working again after a couple of key presses.

Neither irw, nor ir-keytable, nor ir-ctl would give any output after the RC stopped working. However, irw always provides output as expected until the RC fades away.
Thanks for the logs! Unfortunately there's nothing in there that indicates why the IR receiver stopped working. With ir-ctl -r you should see the raw IR data, actually IR signals from any remote.

I googled a bit and found this thread about pervious year's NUC models and a similar issue as yours: https://communities.intel.com/thread/116512

As you seem to get the same issue on Ubuntu 18.04 I'm wondering if it could be a general problem with the NUCs and Intel needs to fix their BIOS, firmwares etc.

A quick check with Win 10 could be worth a try. If you get the same issue there contact Intel. If it works on Win 10 but not Ubuntu it'd be best if you contacted Ubuntu about that - they have slightly larger support resources than we have so if anything needs fixing in the Linux kernel they are probably better suited for the job than us (and we'd in turn get the fix via standard kernel updates or could pick/backport it).

so long,

Hias 
Hi Hias,

Thanks for your time and effort. The link is very helpful indeed.

Kinda looks like a BIOS issue at this point - which, however, wouldn't really explain everything functioning well in the LE installer, would it?!

Right now, unfortunately, I don't have the spare time (nor the spare SATA drive) to check with Windows 10. I'll wait for the next BIOS update at this point, see if that brings any progress, and report accordingly.

As a side note, it simply amazes me how Intel manages to mess up an originally great (HTPC) product such as the NUCs by poor software - be it audio passtrough in NUC6s or the RC mess in NUC7s...

Thanks again and keep up the great work!
@millhouse
 with current multi-channel receivers from Denon, Onkyo and Pioneer, the bitstream output of DTS High Resolution (DTS-HRA) audio output does not work.
This bug was fixed in the LAV filters as of version 0.71.x and tested with current multi-channel receivers from Denon, Onkyo and Pioneer.
 (Intel NUC / Windows 10 / Media Player Classic V1.7.16)
 Is it possible to fix the bug in Kodi 18?
New LibreELEC.tv Leia build #0501: Generic
(Supercedes previous build)

SHA256 Checksum: 37827d054e3546eca1f8c02ec035741dd859463b30471ad6d009c83f4b975ccd (Generic)

text:
# uname -a
Linux NUC 4.14.37 #1 SMP Tue May 1 21:06:25 BST 2018 x86_64 GNU/Linux

# lsb_release
LibreELEC (Milhouse): devel-20180501210302-#0501-gd2c4924 [Build #0501]

# Kodi version
(18.0-ALPHA2 Git:04a039dd). Platform: Linux x86 64-bit

Based on tip of LibreELEC.tv master (d2c4924, changelog) and tip of XBMC master (3a989ee, changelog) with the following modifications: Build Highlights:
  1. eventlircd: don't grab power buttons
  2. Busy dialog fixes
  3. mesa 18.0.2
Build Details:
  1. LibreELEC.tv:
    • libinput: update to 1.10.4 (PR:2650, 1 commit, 1 file changed)
    • systemd: don't power off when power button is pressed (PR:2658, 2 commits, 2 files changed)
    • systemd: update to v238 (PR:2652, 2 commits, 3 files changed)
  2. XBMC:
    • libinput: sleep 10ms to reduce cpu usage (PR:13834, 1 commit, 1 file changed)
    • typo: missing vec4 in shader (PR:13835, 1 commit, 1 file changed)
    • there can only be one busy dialog (PR:13833, 1 commit, 1 file changed)
    • LibInputHandler: Use CThread's Sleep (PR:13838, 1 commit, 1 file changed)
    • windowing: cleanup g_windowing globals (PR:13836, 1 commit, 4 files changed)
    • force closing busy dialog, there can only be one (PR:13839, 2 commits, 2 files changed)
    • [fix] throw exception if addDirectoryItem or setResolvedUrl called with nullptr (PR:13825, 1 commit, 1 file changed)
    • [PVR] CPVRSettings: All settings instances must be re-inited on profile switch (PR:13830, 1 commit, 4 files changed)
    • [aml]Register lirc in AML after #13761 (PR:13798, 1 commit, 2 files changed)
  3. pvr.dvbviewer:
    • If EDL is disabled make sure to zero number of entries (4ea67c5)
    • Bump version to 3.4.5 (098d23c)
  4. Additional commits/pull requests/changes not yet merged upstream:
    • Added: [env] PR:2678 (perma): mesa: update to mesa-18.0.2
    • Added: [env] PR:2679 (perma): eventlircd: don't grab power buttons
Texture Cache Maintenance Utility: Preload your texture cache for optimal UI performance. Remotely manage media libraries. Purge unused artwork to free up space. Find missing media. Configurable QA check to highlight metadata issues. Aid in diagnosis of library and cache related problems.
(2018-05-01, 22:37)Calimero1609 Wrote: @millhouse
 with current multi-channel receivers from Denon, Onkyo and Pioneer, the bitstream output of DTS High Resolution (DTS-HRA) audio output does not work.
This bug was fixed in the LAV filters as of version 0.71.x and tested with current multi-channel receivers from Denon, Onkyo and Pioneer.
 (Intel NUC / Windows 10 / Media Player Classic V1.7.16)
 Is it possible to fix the bug in Kodi 18?

LibreELEC is a Linux-based distribution, not Windows, so it doesn't use the LAV Filters package - are you posting in the wrong thread?

I've just tested a Skylake based NUC6i5SYH with #0501 (configured with 5.1 output) connecting over HDMI into an Onkyo TX-NR828 AVR and I'm able to bitstream DTS-HRA (96KHz sample) without any issue - the DTS-HRA lamp on the AVR is lit. DTS-HD MA is also bitstreaming without a problem.

So in LibreELEC on Linux (at least for Skylake, also my Revo 3700 with Nvidia ION2) there is no bug to be fixed as far as I can tell.
Texture Cache Maintenance Utility: Preload your texture cache for optimal UI performance. Remotely manage media libraries. Purge unused artwork to free up space. Find missing media. Configurable QA check to highlight metadata issues. Aid in diagnosis of library and cache related problems.
Hi Millhouse,

I noticed that the newer updates seem to be slower than before.  I am on 0430 currently.  I initially thought that the problem was with the skin I was running but I have since switched back to Covenant and the problem persists.  What happens is that going through the menus it will be very laggy and sometimes to the point of locking things up.  I am running Libreeelc on an x86 machine with a quad core 2.4 ghz i5 processor in it.  It should fly (and has in the past). 

Also I noticed that I have to reload the urlResolver settings often whereas before I never had to do this.  

All of these observations were going from a release in mid March to current.  

Please let me know if you would like any log files and give me their locations.  Thanks!

Cheers!
@mountbaldy please read the forum rules (wiki) that you accepted when you signed up here.

We offer no support for devices contaminated with banned addons (wiki) such as yours.
|Banned add-ons (wiki)|Forum rules (wiki)|VPN policy (wiki)|First time user (wiki)|FAQs (wiki) Troubleshooting (wiki)|Add-ons (wiki)|Free content (wiki)|Debug Log (wiki)|

Kodi Blog Posts
The forum moderators have determined that banned addons (wiki) are present on your system. To receive assistance here, these banned items must be removed. If a clean log is not submitted within 3 days, then the relevant post(s) will be removed after this time.
|Banned add-ons (wiki)|Forum rules (wiki)|VPN policy (wiki)|First time user (wiki)|FAQs (wiki) Troubleshooting (wiki)|Add-ons (wiki)|Free content (wiki)|Debug Log (wiki)|

Kodi Blog Posts
Hello

Snes emulation not working after build 330.  Tested on Super Mario World.

Thanks
  • 1
  • 216
  • 217
  • 218(current)
  • 219
  • 220
  • 355

Logout Mark Read Team Forum Stats Members Help
LibreELEC Testbuilds for x86_64 (Kodi 18.0)24