Kodi Community Forum
v18 LibreELEC Testbuilds for x86_64 (Kodi 18.0) - Printable Version

+- Kodi Community Forum (https://forum.kodi.tv)
+-- Forum: Support (https://forum.kodi.tv/forumdisplay.php?fid=33)
+--- Forum: General Support (https://forum.kodi.tv/forumdisplay.php?fid=111)
+---- Forum: Linux (https://forum.kodi.tv/forumdisplay.php?fid=52)
+---- Thread: v18 LibreELEC Testbuilds for x86_64 (Kodi 18.0) (/showthread.php?tid=298462)



RE: LibreELEC Testbuilds for x86_64 (Kodi 18.0) - Milhouse - 2016-12-16

(2016-12-16, 21:21)kekomat Wrote: Anyone else having a problem booting after installing #1215?? "Failed to start Xorg Server" ...

Please connect to your machine using Samba and from the Logfiles folder find the latest zip file and upload it somewhere (dropbox etc.) and paste the link.

(2016-12-16, 21:21)kekomat Wrote: Running an ION v1 machine. Any way to fix the installation without wiping the system?? No current backup available Sad

Just ssh in to the box and wget #1214 into the .update folder, or connect using Samba and copy #1214 into the Update folder.


RE: LibreELEC Testbuilds for x86_64 (Kodi 18.0) - Milhouse - 2016-12-16

(2016-12-16, 21:21)kekomat Wrote: Anyone else having a problem booting after installing #1215?? "Failed to start Xorg Server" ...

(2016-12-16, 21:40)blokkie Wrote: Same probleem here, with Nvidia ION2 hardware...

This is my mistake - I thought I'd tested this new driver but apparently not. Getting the same issue here on ION2. It's the same driver issue that this driver was meant to hopefully fix:
Code:
[    9.826931] vgaarb: device changed decodes: PCI:0000:01:00.0,olddecodes=io+mem,decodes=none:owns=io+mem
[    9.827779] Error: Driver 'nvidia' is already registered, aborting...
[    9.827783] NVRM: DRM init failed
I'm afraid nvidia-legacy users will have to stick on #1214 for the time being until we have a fix.

Edit: What I had in fact done is test 340.101 with the 4.8.13 kernel. D'oh.


RE: LibreELEC Testbuilds for x86_64 (Kodi 18.0) - Milhouse - 2016-12-16

New LibreELEC.tv Krypton build #1216: Generic
(Supercedes previous build)

Code:
# uname -a
Linux NUC 4.9.0 #1 SMP Fri Dec 16 21:05:44 GMT 2016 x86_64 GNU/Linux

# lsb_release
LibreELEC (Milhouse) - Version: devel-20161216210423-#1216-g7ab3814 [Build #1216]

# Kodi version
(18.0-ALPHA1 Git:17.0-beta7-7581c4a). Platform: Linux x86 64-bit

Based on tip of LibreELEC.tv master (7ab38141, changelog) and tip of XBMC master (921d7f36, changelog) with the following modifications:
  • Includes latest kodi-platform master (c8188d82)
  • Includes latest libcec master (12fa2750, ahead +2)
  • Includes latest libnfs master (a2b37349, ahead +2)
  • Includes latest p8-platform master (2d90f986, ahead +9)
  • Includes latest addons: inputstream.adaptive (fc862638), inputstream.rtmp (55437ac7), peripheral.joystick (987a818e), pvr.argustv (4702a9e7, +3), pvr.demo (99db6bf5, +3), pvr.dvblink (4e6992d9, +3), pvr.dvbviewer (4c8df55e, +9), pvr.filmon (a23db0c5, +3), pvr.hdhomerun (baeafbde, +3), pvr.hts (117d4020, +3), pvr.iptvsimple (b19850dd, +3), pvr.mediaportal.tvserver (f00ccfd3, +15), pvr.mythtv (087235ce, +41), pvr.nextpvr (48a7796a, +3), pvr.njoy (d1ad569b, +3), pvr.octonet (8549a31c), pvr.pctv (b8286a23, +3), pvr.stalker (dfda4c51, +3), pvr.vbox (bfa938c7, +8), pvr.vdr.vnsi (3efe8682, +6), pvr.vuplus (acd75dc5, +3), pvr.wmc (35be514e, +3)
  • Include [env] patch: Don't build ADSP addons
  • Include [env] patch: kodi: fix intel gpu detection patch after kodi/10848
  • Include [env] patch: linux: use in-tree rtl8192cu for RPi/RPi2
  • Include [env] patch: repository.retroplayer.libreelec.addons: add temp repo for testing
  • Include [env] patch: Bump included addon versions to prevent online updates
  • Include [env] patch: Add kodi binary addons (pvr, adsp, other)
  • Include [env] patch: Add experimental splash video for RPi
  • Include [env] patch: HACK: Disable multiple PVR addons during migration. Always enable inputstream.* and os.*
  • Include [env] PR:968: busybox: add getedid script
  • Include [env] PR:1025: scripts/create_addon: use PKG_REV as part of the binary addon version
  • Include [env] PR:1052: Package updates
  • Include [env] PR:1058: libdvbcsa: AArch64 support
  • Include [env] PR:1063: kodi: update to 17.0-beta7-7581c4a
  • Include [pkg] patch: drm/i915: Fix setting of boost freq tunable (linux)
  • Include [pkg] PR:4: add Geniatech T230C fw (dvb-firmware)
  • Include [pkg] PR:5: add all TBS firmwares (dvb-firmware)
  • Include [pkg] PR:39: use prettyname_regex to show shortname in update menu (service.libreelec.settings)
  • Include [pkg] PR:44: update infobools (service.libreelec.settings)
  • Include [pkg] PR:249: Fix ass_set_cache_limits() to work with composite cache (libass)
  • Include [pkg] PR:11143: [PVR] trac17040: fix crashes on pvr manager deinit/reinit.
  • Include [pkg] PR:11144: CMake: Fix TexturePacker installation directory
  • Include [pkg] PR:11163: Fix silent failure when game add-on can't be installed
Build Highlights:
  1. Minors
  2. nvidia-legacy hardware not supported (sorry)
Build Details:
  1. XBMC:
    • [linux] Remove deprecated OnlyShowIn from desktop file (PR:11189, 1 commit, 1 file changed)
    • [cmake] fix build after 8ec984f0fa5bd8e75026e5558870b96071e54288 (PR:11191, 1 commit, 1 file changed)
    • libdvd: fix setting compiler when cross-compiling (PR:11185, 1 commit, 1 file changed)
    • [addon] remove not needed 'GetInfo(...)' call for ScreenSaver (PR:11190, 1 commit, 4 files changed)
    • VideoPlayer: Don't add timestamp offsets to DVD_NOPTS_VALUE (PR:11178, 1 commit, 1 file changed)
    • Remove the special windows free_string_val (PR:11192, 1 commit, 2 files changed)
  2. Additional commits/pull requests/changes not yet merged upstream:
    • Added: [env] PR:1058: libdvbcsa: AArch64 support
    • Added: [env] PR:1063: kodi: update to 17.0-beta7-7581c4a



RE: LibreELEC Testbuilds for x86_64 (Kodi 18.0) - swaggerwocky - 2016-12-17

Anyone have issues with immediate wakeups after suspending? I can solve this by disabling XHC in ACPI wakeup, but then I can't wake the system with Pulse-Eight CEC. I don't have this issue in Ubuntu, the suspend and wake work flawlessly. I'm on a Zotac CI520 with an Intel i3 4020y CPU.


RE: LibreELEC Testbuilds for x86_64 (Kodi 18.0) - Milhouse - 2016-12-17

(2016-12-17, 00:24)swaggerwocky Wrote: Anyone have issues with immediate wakeups after suspending? I can solve this by disabling XHC in ACPI wakeup, but then I can't wake the system with Pulse-Eight CEC. I don't have this issue in Ubuntu, the suspend and wake work flawlessly. I'm on a Zotac CI520 with an Intel i3 4020y CPU.

Is this a new issue with #1216? Which of these Milhouse test builds were you using that had working suspend/resume, and Pulse-Eight?

There's a known issue with the Pulse-Eight adapter since libcec-4.0, but that version has been in use since #1026. So either you've not upgraded your test build in a long time, or you have never used these test builds.

Or it's a completely different issue...


RE: LibreELEC Testbuilds for x86_64 (Kodi 18.0) - kekomat - 2016-12-17

(2016-12-16, 22:24)Milhouse Wrote:
(2016-12-16, 21:21)kekomat Wrote: Running an ION v1 machine. Any way to fix the installation without wiping the system?? No current backup available Sad

Just ssh in to the box and wget #1214 into the .update folder, or connect using Samba and copy #1214 into the Update folder.

Thanks for the advise, already hoped the update mechanism would save the day.


RE: LibreELEC Testbuilds for x86_64 (Kodi 18.0) - swaggerwocky - 2016-12-17

(2016-12-17, 01:18)Milhouse Wrote:
(2016-12-17, 00:24)swaggerwocky Wrote: Anyone have issues with immediate wakeups after suspending? I can solve this by disabling XHC in ACPI wakeup, but then I can't wake the system with Pulse-Eight CEC. I don't have this issue in Ubuntu, the suspend and wake work flawlessly. I'm on a Zotac CI520 with an Intel i3 4020y CPU.

Is this a new issue with #1216? Which of these Milhouse test builds were you using that had working suspend/resume, and Pulse-Eight?

There's a known issue with the Pulse-Eight adapter since libcec-4.0, but that version has been in use since #1026. So either you've not upgraded your test build in a long time, or you have never used these test builds.

Or it's a completely different issue...

I believe build #1215 was the first that I used, I've been using Windows 10 prior to this and wanted something more streamlined. After I couldn't figure out how to fix the suspend/wake issue with LibreELEC, I confirmed it wasn't something specific to my hardware configuration that couldn't be fixed by checking out a live install of Ubuntu. I moved over to Ubuntu and followed fritsch's guide to setup Intel VAAPI and the latest unstable Krypton build. One thing I didn't try was check to see if suspend/wake functioned normally on the latest stable LibreELEC. I might give that a go over the weekend when I find some time.


RE: LibreELEC Testbuilds for x86_64 (Kodi 18.0) - Illumination19 - 2016-12-17

(2016-12-17, 00:24)swaggerwocky Wrote: I believe build #1215 was the first that I used, I've been using Windows 10 prior to this and wanted something more streamlined. After I couldn't figure out how to fix the suspend/wake issue with LibreELEC, I confirmed it wasn't something specific to my hardware configuration that couldn't be fixed by checking out a live install of Ubuntu. I moved over to Ubuntu and followed fritsch's guide to setup Intel VAAPI and the latest unstable Krypton build. One thing I didn't try was check to see if suspend/wake functioned normally on the latest stable LibreELEC. I might give that a go over the weekend when I find some time.

Did you try adding
echo XHC > /proc/acpi/wakeup
echo EHC1 > /proc/acpi/wakeup
to your autostart.sh in Libreelec? It solved my suspend/wake issues.


RE: LibreELEC Testbuilds for x86_64 (Kodi 18.0) - swaggerwocky - 2016-12-17

(2016-12-17, 09:22)Illumination19 Wrote:
(2016-12-17, 00:24)swaggerwocky Wrote: I believe build #1215 was the first that I used, I've been using Windows 10 prior to this and wanted something more streamlined. After I couldn't figure out how to fix the suspend/wake issue with LibreELEC, I confirmed it wasn't something specific to my hardware configuration that couldn't be fixed by checking out a live install of Ubuntu. I moved over to Ubuntu and followed fritsch's guide to setup Intel VAAPI and the latest unstable Krypton build. One thing I didn't try was check to see if suspend/wake functioned normally on the latest stable LibreELEC. I might give that a go over the weekend when I find some time.

Did you try adding
echo XHC > /proc/acpi/wakeup
echo EHC1 > /proc/acpi/wakeup
to your autostart.sh in Libreelec? It solved my suspend/wake issues.

I prefer to keep XHC enabled since I use Pulse-Eight CEC to wake my system with my TV remote. I won't be able to do so, otherwise.


RE: LibreELEC Testbuilds for x86_64 (Kodi 18.0) - Milhouse - 2016-12-17

Good news, the issue with nvidia-legacy 340.101 and linux-4.9.0 will be fixed in tonight's build. Smile


RE: LibreELEC Testbuilds for x86_64 (Kodi 18.0) - cabal2k - 2016-12-17

(2016-12-16, 19:19)Milhouse Wrote: @cabal2k & @sw4y: Any change in DAZN behaviour with inputstream.adaptive 1.0.6 and removal of the (curl) header filter (build #1215)? The removal of the header filter will be dropped in the next build as it doesn't look like a popular change, but would be good to know if it has any effect on the issue you are both seeing.

No change with build #1215:

Code:
11:13:45.468 T:140052262594304   ERROR: CCurlFile::FillBuffer - Failed: SSL connect error(35)
11:13:45.468 T:140052262594304   ERROR: CCurlFile::Open failed with code 0 for https://widevine.entitlement.theplatform.eu/wv/web/ModularDrm/getWidevineLicense?form=json&schema=1.0&token=P1ltrElWMaJOfVDE6VUoYWByEJDEQFC0&account=http%3a%2f%2faccess.auth.theplatform.com%2fdata%2f{...}vj7z88ynqV7Q%3D|Expect&acceptencoding=gzip%2C%20deflate&seekable=0
11:13:45.468 T:140052262594304   ERROR: AddOnLog: InputStream Adaptive: License server returned failure
11:13:45.469 T:140052262594304   ERROR: AddOnLog: InputStream Adaptive: Unable to generate a license



LibreELEC Testbuilds for x86_64 (Kodi 18.0) - blokkie - 2016-12-17

(2016-12-17, 11:22)Milhouse Wrote: Good news, the issue with nvidia-legacy 340.101 and linux-4.9.0 will be fixed in tonight's build. Smile


Hey Milhouse good news, after release I'll give you an update [emoji106]


Verzonden vanaf mijn iPad met Tapatalk


RE: LibreELEC Testbuilds for x86_64 (Kodi 18.0) - tomtomclub - 2016-12-17

(2016-12-17, 09:51)swaggerwocky Wrote:
(2016-12-17, 09:22)Illumination19 Wrote:
(2016-12-17, 00:24)swaggerwocky Wrote: I believe build #1215 was the first that I used, I've been using Windows 10 prior to this and wanted something more streamlined. After I couldn't figure out how to fix the suspend/wake issue with LibreELEC, I confirmed it wasn't something specific to my hardware configuration that couldn't be fixed by checking out a live install of Ubuntu. I moved over to Ubuntu and followed fritsch's guide to setup Intel VAAPI and the latest unstable Krypton build. One thing I didn't try was check to see if suspend/wake functioned normally on the latest stable LibreELEC. I might give that a go over the weekend when I find some time.

Did you try adding
echo XHC > /proc/acpi/wakeup
echo EHC1 > /proc/acpi/wakeup
to your autostart.sh in Libreelec? It solved my suspend/wake issues.

I prefer to keep XHC enabled since I use Pulse-Eight CEC to wake my system with my TV remote. I won't be able to do so, otherwise.

yes mi too


RE: LibreELEC Testbuilds for x86_64 (Kodi 18.0) - Milhouse - 2016-12-17

(2016-12-16, 19:20)Devil-Strike Wrote:
(2016-12-16, 18:52)flyingernst Wrote: with #1215 file transfer still works with 113mb/sec right now (J4205 ITX). Clocksettings work normally.

Strange here wont get more than 16MB/s tested with download files/smb/ftp, and it always stays at 799Mhz.

Meaby bios bug?

It would be best if you can file a bug on bugs.freedesktop.org, something along the lines of "CPU Frequency Scaling not working on Apollo Lake". Provide the contents of "cat /proc/cpuinfo" while software decoding.


RE: LibreELEC Testbuilds for x86_64 (Kodi 18.0) - fritsch - 2016-12-17

And don't forget the dmesg output