Kodi 18 - LibreELEC Testbuilds for RaspberryPi (Kodi 18.0)

  Thread Rating:
  • 6 Vote(s) - 4.83 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Post Reply
Memphiz Offline
Kodi MVP
Posts: 16,090
Joined: Feb 2011
Reputation: 227
Location: germany
Post: #1606
Well something in kodi keeps a handle inside that export. (Background images? Fanart? Whatever it is...)

AppleTV4/iPhone/iPod/iPad: HowTo find debug logs and everything else which the devs like so much: click here
HowTo setup NFS for Kodi: NFS (wiki)
HowTo configure avahi (zeroconf): Avahi_Zeroconf (wiki)
READ THE IOS FAQ!: iOS FAQ (wiki)
find quote
ramis52 Offline
Junior Member
Posts: 12
Joined: Aug 2016
Reputation: 0
Post: #1607
(2017-04-27 16:46)SpokV Wrote:  
(2017-04-27 14:44)popcornmix Wrote:  
(2017-04-22 10:51)SpokV Wrote:  Hi. I've uploaded crashlog from 420,421. Happens when streaming video via network.419 works fine.
It seems to do with the ffmpeg buffer increase. Not sure...

Can you confirm if this crash still occurs with latest build?
I think it will be resolved in tonight's build (when ffmpeg buffer increase is reverted) but there is a possible fix in last night's build we'd be interested in knowing about.

I've checked on today's build and it's still there. I'll check the tonight's build too.

+1 same problem here. this happens since #0419
(This post was last modified: 2017-04-28 18:51 by ramis52.)
find quote
popcornmix Offline
Team-Kodi Member
Posts: 7,862
Joined: Feb 2011
Reputation: 326
Post: #1608
(2017-04-28 18:50)ramis52 Wrote:  +1 same problem here. this happens since #0419

Did you test with last night's build? That reverted the commit that introduced the crash.
find quote
ramis52 Offline
Junior Member
Posts: 12
Joined: Aug 2016
Reputation: 0
Post: #1609
yes. #0427 crashed after the 1 episode i watched. rpi rebootet after a few secs
(This post was last modified: 2017-04-28 19:01 by ramis52.)
find quote
popcornmix Offline
Team-Kodi Member
Posts: 7,862
Joined: Feb 2011
Reputation: 326
Post: #1610
(2017-04-28 19:00)ramis52 Wrote:  yes. #0427 crashed after the 1 episode i watched. rpi rebootet after a few secs

Probably something else. Can you post crash log?
find quote
ramis52 Offline
Junior Member
Posts: 12
Joined: Aug 2016
Reputation: 0
Post: #1611
sure.

sorry. had to delete that link. cauz i recognised i played illegal content with original filenames...

here is the new one

https://pastebin.com/pVGrt6hi

this crash happens on all SD Content for me. HD Content works fine
(This post was last modified: 2017-04-28 19:40 by ramis52.)
find quote
Milhouse Offline
Team-Kodi Member
Posts: 12,877
Joined: Jan 2011
Reputation: 642
Post: #1612
(2017-04-28 19:24)ramis52 Wrote:  sure.

sorry. had to delete that link. cauz i recognised i played illegal content with original filenames...

here is the new one

https://pastebin.com/pVGrt6hi

I don't think that's the same problem - your crash is occurring in the webserver while calling JSON methods, not ffmpeg. Can you confirm when this crash started happening?

Edit: You said #0419... thanks. But not really seeing anything in that build which might be related. Can you confirm it really is #0419 and not an earlier build?

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.
(This post was last modified: 2017-04-28 19:46 by Milhouse.)
find quote
ramis52 Offline
Junior Member
Posts: 12
Joined: Aug 2016
Reputation: 0
Post: #1613
Ok. This makes for me a little bit sense. What we using since a few builds is the chorus webserver to update the library. Never had that Problem before. Maybe thats a reason. I disable the webfronted for now in the next few builds and tell my wife that she have to use the ipad or the remote to update the library.

I will test all builds from 0410 to 0419 over the weekend to make it more precise since when the problem startet.

edit:

ok just checked. this happens only after a library update via chorus
(This post was last modified: 2017-04-28 19:57 by ramis52.)
find quote
Milhouse Offline
Team-Kodi Member
Posts: 12,877
Joined: Jan 2011
Reputation: 642
Post: #1614
(2017-04-28 14:11)mm00 Wrote:  Milhouse, could you create a patched testbuild without this commit? https://github.com/Pulse-Eight/libcec/co...7d91f0f573

It might cause LG tvs strange behaviour in this topic: https://github.com/Pulse-Eight/libcec/issues/307

Try test build #0427l: RPi2

Same as #0427 but with PR:256 reverted.

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.
find quote
SpokV Offline
Junior Member
Posts: 44
Joined: Mar 2016
Reputation: 0
Post: #1615
(2017-04-28 18:51)popcornmix Wrote:  
(2017-04-28 18:50)ramis52 Wrote:  +1 same problem here. this happens since #0419

Did you test with last night's build? That reverted the commit that introduced the crash.

For me the issue is solved. I don't get a crash on playback. I could see it happens way back untill 420. 419 worked ok.
find quote
Milhouse Offline
Team-Kodi Member
Posts: 12,877
Joined: Jan 2011
Reputation: 642
Post: #1616
New LibreELEC.tv Leia build #0428: RPi / RPi2
(Supercedes previous build)

Code:
# uname -a
Linux rpi512 4.10.13 #1 Fri Apr 28 21:38:35 BST 2017 armv6l GNU/Linux

# vcgencmd version
Apr 27 2017 17:27:25
Copyright (c) 2012 Broadcom
version 17af5814bb19dbb7c70ccd2c845b80a160943811 (clean) (release)

# lsb_release
LibreELEC (Milhouse) - Version: devel-20170428213652-#0428-ga01c088 [Build #0428]

# Kodi version
(18.0-ALPHA1 Git:91a9066). Platform: Linux ARM 32-bit

Based on tip of LibreELEC.tv master (a01c0887, changelog) and tip of XBMC master (5e12c614, changelog) with the following modifications: Build Highlights:
  1. AESinkAUDIOTRACK: Refactor and separate PCM and Passthrough Enumeration
  2. busybox: enabling CONFIG_FEATURE_USERNAME_COMPLETION
Build Details:
  1. LibreELEC.tv:
    • busybox: enabling CONFIG_FEATURE_USERNAME_COMPLETION in busybox-targeā€¦ (PR:1572, 1 commit, 1 file changed)
  2. XBMC:
    • guilib: only update container after condition changed (PR:12016, 1 commit, 2 files changed)
    • AESinkAUDIOTRACK: Refactor and separate PCM and Passthrough Enumeration (PR:11997, 2 commits, 2 files changed)
  3. newclock5:
    • New commits in this build:
      • fixup! renderer: Allow gui updates to be skipped when playing video (7144faef)
    • Commits no longer in build:
      • guilib: only update container if after condition changed (5d9a1f94)
  4. Additional commits/pull requests/changes not yet merged upstream:
    • Added: [env] PR:1571: createlog: bring sanity to kodi log order
    • Added: [pkg] PR:65: update language from transifex (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.
find quote
skylarking Offline
Fan
Posts: 408
Joined: Jun 2013
Reputation: 13
Post: #1617
(2017-04-27 14:10)asavah Wrote:  
(2017-04-27 09:46)skylarking Wrote:  To see native functions removed and migrated to an add-ons seems OK on the surface.
And if the end result is a cleaner code base that supports aded features, it can only be a good thing.
However, when there seems to be a waning desire for the coders to actually keep the functionality that made Kodi what it is, it becomes a little concerning as to what lovers of Kodi will end up with.

And yes, 'waning desire' is speculation on my part as 'i know nothing' about how the team coordinates the tasks at hand, nor the people involved.
I know it's open source and all, but i'd have thought as part of a cleanup and add-on migration activities, the coders actually removing stuff would have at least migrated the functionality into an add-on and verified that the functionality they moved still works.

It just seems odd to me that code and thus functionality can be so easily removed, a wild west of coding as such with an obvious end result of a loss of functionality.

@Millhouse, i don't believe it's an issue of your making and do appreciate your efforts in providing these edgy LE PRi builds.
If your not too busy, maybe you can enlighten me as to why this is happening and what features we may fall from the table?

You should be aware that these excellent @Milhouse's builds are bleeding edge, and are based on current master code, which may contain bugs and regressions, that's why these builds exist - to help the devs hunt the bugs long before new code hits the rest of the users who use release versions.
Eventually new binary addons replacing built-in functionality will be fixed. AFAIK there is huge rework of addon related code incoming, so it may be inefficient to spend developer time now to fix/improve this stuff. We'll have to wait.

If you want stable non-breaking functionality - use releases, but if you want to help Kodi become better and better by testing 0day stuff you may have to suffer a breakage or two from time to time Tongue
Thanks for your comments.
Guess i was just a little confused at the process of core feature migration to add-ons, particlularly the split between those involved as it seemed a little loose from what i read earlier.
Maybe that's just how open source developement is - a little loose at times Tongue

Anyway, I do appreciation @Millhouse's efforts on these builds Nod

Hopefully i can provide some feedback but i first need to improve my processes to keep up with the latest builds.
As is, i'm already behind the game as i'm on #0426 which i updated a just a couple of days ago Blush

I'm a XBMC novice :)
find quote
mm00 Offline
Junior Member
Posts: 8
Joined: Apr 2016
Reputation: 0
Post: #1618
(2017-04-28 19:57)Milhouse Wrote:  
(2017-04-28 14:11)mm00 Wrote:  Milhouse, could you create a patched testbuild without this commit? https://github.com/Pulse-Eight/libcec/co...7d91f0f573

It might cause LG tvs strange behaviour in this topic: https://github.com/Pulse-Eight/libcec/issues/307

Try test build #0427l: RPi2

Same as #0427 but with PR:256 reverted.

Thank you! I will try it tomorrow and inform you with the result. Smile
find quote
mcelliott Offline
Fan
Posts: 392
Joined: Mar 2015
Reputation: 5
Post: #1619
Hi,

Since updating to the latest version (I haven't updated for a while) I now have a strange "%H" showing in my windows network. I fixed my samba.conf as advised, but it is still there, but not accessible.

Does anyone else see this?

Thanks,

Mark
find quote
bkuhls Offline
Junior Member
Posts: 47
Joined: May 2013
Reputation: 1
Post: #1620
(2017-04-28 10:41)popcornmix Wrote:  The patch that fixed the audio crackles for most users was PR12002.
Should this patch be backported to Krypton then?
find quote
Post Reply