• 1
  • 399
  • 400
  • 401(current)
  • 402
  • 403
  • 495
v18 LibreELEC Testbuilds for RaspberryPi (Kodi 18.0)
(2018-07-21, 14:00)tjost Wrote:  The Codecs are very cheap. about 3 Euro for booth. So invest that Money.     
What this has to do with DRM streams?
 
(2018-07-20, 17:23)popcornmix Wrote: Hardware decode of these streams is not possible due to DRM. The decoder is embedded in the widevine dl with no access to the unencrypted bitstream. 
(2018-07-21, 15:05)smp1 Wrote:
(2018-07-21, 14:00)tjost Wrote:  The Codecs are very cheap. about 3 Euro for booth. So invest that Money.     
What this has to do with DRM streams?
 
(2018-07-20, 17:23)popcornmix Wrote: Hardware decode of these streams is not possible due to DRM. The decoder is embedded in the widevine dl with no access to the unencrypted bitstream. 
i misread this, sorry.
New LibreELEC.tv Leia build #0721: RPi / RPi2
(Supercedes previous build)

SHA256 Checksum: 014dc2167cb918a42353548388ee5a7a961d8abf91c4b5bdecafb11e9fb28633 (RPi)
SHA256 Checksum: 3c7282594ccab6bbdbe588141af1598ec3ee61b1a348350013b380ea3cf80c10 (RPi2)

text:
# uname -a
Linux rpi512 4.14.56 #1 Sat Jul 21 21:04:08 BST 2018 armv6l GNU/Linux

# vcgencmd version
Jul 17 2018 13:28:45
Copyright © 2012 Broadcom
version 4da89b2ba95c8826c492b1d734296c8a1a2c7d2b (clean) (release)

# lsb_release
LibreELEC (Milhouse): devel-20180721210249-#0721-g27669a0 [Build #0721]

# Kodi version
(18.0-ALPHA3 Git:b7583e3). Platform: Linux ARM 32-bit

Based on tip of LibreELEC.tv master (27669a0, changelog) and tip of XBMC master (0b8f23a, changelog) with the following modifications: Build Highlights:
  1. [gui] Kill g_audioManager
Build Details:
  1. LibreELEC.tv:
    • gcc-linaro: update to latest version (PR:2812, 4 commits, 4 files changed)
  2. XBMC:
    • [gui] Kill g_audioManager (PR:14198, 1 commit, 15 files changed)
    • [cleanup] silence casts away qualifiers warnings (PR:14192, 18 commits, 73 files changed)
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.
(2018-07-21, 07:31)oztrained Wrote: 1. Using MMAL the audio is now correctly decoded as DD or DTS for Movies and TV Series
2. Recorded TV from Argus PVR using MMAL plays audio as PCM (this happened even in Kodi v17 using MMAL) however it is correctly played as DD audio using OMXPlayer. This is the same in #719 playing the Argus PVR recorded TS files from the Kodi menu TV > Recordingstain
3. However using MMAL and selecting the same TS file using the Kodi File Browser the Argus recorded file plays with DD audio. 

That sounds like expected behaviour. For real time streams (where audio is resampled to maintains sync between server and client) passthrough is disabled.
Using multichannel PCM is the preferred configuration (assuming you are outputting audio through HDMI to an AVR).

If multichannel PCM isn't an option (e.g. toslink/spdif) then you need to enable AC3 transcode. Audio will be decoded, resampled and encoded to AC3, then that will be passed through to receiver.
New LibreELEC.tv Leia build #0722: RPi / RPi2
(Supercedes previous build)

SHA256 Checksum: 894f4aa7b0d4a2c8208aff637ffb92dc44dfaa095f865e8ed8747ecabf6cd2b4 (RPi)
SHA256 Checksum: 2b76bcb9180fc325709e2cb6a98754d037b4c4efa4f1d55ff974b678f10c1c82 (RPi2)

text:
# uname -a
Linux rpi512 4.14.56 #1 Sun Jul 22 21:05:03 BST 2018 armv6l GNU/Linux

# vcgencmd version
Jul 17 2018 13:28:45
Copyright © 2012 Broadcom
version 4da89b2ba95c8826c492b1d734296c8a1a2c7d2b (clean) (release)

# lsb_release
LibreELEC (Milhouse): devel-20180722210247-#0722-g9677927 [Build #0722]

# Kodi version
(18.0-ALPHA3 Git:b7583e3). Platform: Linux ARM 32-bit

Based on tip of LibreELEC.tv master (9677927, changelog) and tip of XBMC master (56383d5, changelog) with the following modifications: Build Highlights:
  1. Minors
Build Details:
  1. LibreELEC.tv:
    • kodi binary addons (PR:2842, 2 commits, 2 files changed)
    • libretro-mame: update to libretro-mame-c816514 (PR:2839, 1 commit, 1 file changed)
  2. XBMC:
    • X11: some fixes to OML timings (PR:14211, 1 commit, 1 file changed)
    • FFmpeg: Bump to 4.0.2-Leia-Alpha3 (PR:14212, 1 commit, 1 file changed)
    • X11: implement GetFrameLatencyAdjustment (PR:14213, 1 commit, 5 files changed)
    • GLContextEGL: Adjust sleeptime when detecing missed msc increase (PR:14214, 1 commit, 1 file changed)
  3. Additional commits/pull requests/changes not yet merged upstream:
    • Added: [env] PR:2838 (perma): linux (Generic): update to linux-4.17.9
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.
ran a scan disk.  The SD card is good. 

Additional info,  it is a 64 G scandisk SD card 

The error starts on #717.  If this helps, I have the boot sector formatted as Fat16 and the other formatted as Ext4.
(2018-07-23, 00:46)hagertyoh Wrote: ran a scan disk.  The SD card is good. 

Additional info,  it is a 64 G scandisk SD card 

The error starts on #717.  If this helps, I have the boot sector formatted as Fat16 and the other formatted as Ext4.

Not really sure - there is a new kernel (4.14.56) in #0717, but obviously this isn't a widespread issue.

So you're able to upgrade from #0712 to #0716, but not to #0717? Odd.

Can you boot with #0717, and when you are in the debug shell, type:
text:

mount -o remount,rw /flash
cat /proc/kmsg >/flash/kmsg

You'll need to enter CTRL-C to terminate the second command. Run these commands exactly once, because the contents of the /proc/kmsg file will be cleared when it is read - if you mess it up, reboot (type reboot) and try again.

Insert the SD card in a PC and upload the contents of the kmsg file (it will be in the FAT16 partition) to a pastebin site, and post the link.

After you've run the above commands, reboot, then try running the following command in the debug shell:
text:

mount /dev/mmcblk0p2 /storage -orw,noatime
and post the results.

Obviously this is an obscure issue otherwise there'd be hundreds of complaints by now... is it possible for you to also test with a different SD card?
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.
here is the pastebin   
https://pastebin.com/j6H3HeL7

here is what it said after   mount /dev/mmcblk0p2 /storage -orw,noatime

mount: mounting /dev/mmcblk0p2 on /storage  failed: Invalid argument.

Unless others see this.  do not worry.  I may have done something when formatting the SD card.  I had a hard time when expanding.  I will go back to my 16 G sd card and see how that works.
(2018-07-23, 02:43)hagertyoh Wrote: here is the pastebin   
https://pastebin.com/j6H3HeL7

here is what it said after   mount /dev/mmcblk0p2 /storage -orw,noatime

mount: mounting /dev/mmcblk0p2 on /storage  failed: Invalid argument.

From the log, it would appear that your ext4 partition is suffering from severe corruption, to the point where it is no longer recognised as an ext4 file system.

I don't know what fsck command you ran, if any, but I'd be very surprised if it failed to find issues so I'm assuming you didn't check it. If you only "ran a scandisk" then that's of no use at all - scandisk is a Windows program and will only check the /flash partition.

I highly doubt this issue started with #0717, as it's very unlikely that any LibreELEC system - old or new - will be able to mount /storage with that level of corruption.

(2018-07-23, 02:43)hagertyoh Wrote: Unless others see this. do not worry. I may have done something when formatting the SD card. I had a hard time when expanding. I will go back to my 16 G sd card and see how that works.

I have no idea how you expanded your 16GB image to 64GB, but it might have been worth mentioning that, as it's almost certainly the problem. I'm pretty sure this is self-inflicted.

I would suggest you run a test to verify your 64GB card is not a fake using software such as h2testw.

The easiest way to transfer data from a 16GB card to a 64GB card, would be to:
  1. Backup your old data from the 16GB card using the "Backup" option in LibreELEC Settings addon
  2. Write a fresh 8.2.5 image to the 64GB
  3. Boot the 8.2.5 image, allowing it to expand the /storage partition to use all free space
  4. Restore your backup

If you wanted to go the hard(er) way, dd the contents of the 16GB card to the 64GB card, then use gparted to expand the ext4 filesystem on the 64GB card.

Edit: Just noticed your log is trying to mount an F2FS filesystem. I've no idea what you've done, as we don't support f2fs as standard, but try again from scratch. If you're using a non-ext4 filesystem then you're on your own for support. Smile
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.
@hagertyoh there were indeed several additional ext4 sanity checks added in 4.14.55: https://cdn.kernel.org/pub/linux/kernel/...og-4.14.55 (search for ext4), so maybe an older kernel would have mounted your filesystem (and you'd have file corruption problems later on), but the current mainline kernel now thinks it is corrupted and refuses to mount it.
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.
Sorry I did not mention the expand. I did it in February and all has worked. I did use gparted and remember it was not that simple on the old computer I was using. I must have butcher that and did not know.

This was the first time I did fsck so must have not completed correctly but it did not seem that hard. I received one flag that was wrong and selected the system to repair it. I must have done something wrong and did not know.

I will go back to 8.2.5 image expand that as suggested and use one of the back ups I have.

Thanks for checking.
I use the LibreELEC SD Formater downloaded the 8.2.5, write it to the sd card put it in the pi and start it.
afterwards I update to LibreELEC 9.

All Data I use outside of KODI like Roms/Videos/Audio I do have on an external Storage.
New LibreELEC.tv Leia build #0723: RPi / RPi2
(Supercedes previous build)

SHA256 Checksum: 3f70ea81a9da59098b50c6bbe32d37386a2b6454abc4a5698ca814b425d45860 (RPi)
SHA256 Checksum: db81a8cd5aeaa6b07f2f433ba44c05fa9a757bed7bc63ba2a2804e311063b79d (RPi2)

text:
# uname -a
Linux rpi512 4.14.56 #1 Mon Jul 23 21:04:03 BST 2018 armv6l GNU/Linux

# vcgencmd version
Jul 17 2018 13:28:45
Copyright © 2012 Broadcom
version 4da89b2ba95c8826c492b1d734296c8a1a2c7d2b (clean) (release)

# lsb_release
LibreELEC (Milhouse): devel-20180723210246-#0723-g9677927 [Build #0723]

# Kodi version
(18.0-ALPHA3 Git:b7583e3). Platform: Linux ARM 32-bit

Based on tip of LibreELEC.tv master (9677927, changelog) and tip of XBMC master (1067362, changelog) with the following modifications: Build Highlights:
  1. [database] Reset path hashes when cleaning up files in video database
  2. make sure not to delete a CThread object while spawn thread is in process
Build Details:
  1. XBMC:
    • Rework FindXkbcommon.cmake (PR:14216, 1 commit, 1 file changed)
    • [cmake][win10] remove required_deps hack (PR:14217, 1 commit, 1 file changed)
    • [database] Reset path hashes when cleaning up files in video database (PR:14165, 1 commit, 1 file changed)
    • [cmake] fix building with internal fmt (PR:14215, 1 commit, 2 files changed)
    • make sure not to delete a CThread object while spawn thread is in pro (PR:13867, 1 commit, 1 file changed)
    • [depends] bump python27 to 2.7.15 (PR:14218, 2 commits, 13 files changed)
  2. peripheral.joystick:
    • .travis.yml: Update for addition of -DADDONS_DEFINITION_DIR (PR:148, 1 commit, 1 file changed)
  3. pvr.vdr.vnsi:
    • add status channel for streaming (PR:17, 2 commits, 7 files changed)
  4. Additional commits/pull requests/changes not yet merged upstream:
    • Added: [env] PR:2843 (perma): distro: allow configurable settings addon
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.
(2018-05-30, 15:45)popcornmix Wrote: Our HEVC guy has reproduced the issue. He can see it when rebasing to ffmpeg 4 without any change in the Pi specific code.
Very high bitrate files have ~10% performance hit. We'll try to identify what has caused the issue. 
Looks like there is a regression in one of the recent HEVC optimization updates. I again see a bad performance with the high bitrate sample.
(2018-07-24, 20:07)smp1 Wrote: Looks like there is a regression in one of the recent HEVC optimization updates. I again see a bad performance with the high bitrate sample

Can you confirm the first build with the issue?
  • 1
  • 399
  • 400
  • 401(current)
  • 402
  • 403
  • 495

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