Kodi Community Forum

Full Version: LibreELEC Testbuilds for RaspberryPi (Kodi 17.0)
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
(2016-07-11, 08:51)Pienoet Wrote: [ -> ]
(2016-07-11, 08:18)tuxen Wrote: [ -> ]just a little update:
I tried "up/downgrading" now official LE 7.90.003 alpha is released, and there the stalker pvr client works fine.

Where did you find LE 7.90.003?
Can't find it on their downloadpage
it's available via the libreelec-8 update channel in libreelec settings.
(2016-07-11, 06:33)smp1 Wrote: [ -> ]I have to say that TvHeadend 4.2 performs shockingly bad in Milhouse builds compared to the official LE 7.0.2.
Just to confirm, the DVB dongle is a USB device connected to the Pi you are watching? (i.e. frontend and backend are on the same device)
Can you test if disabling deinterlace, or switching the deinterlace method from automatic to bob makes a difference?
(2016-07-09, 21:39)Milhouse Wrote: [ -> ]
(2016-07-09, 12:45)stevegal Wrote: [ -> ]Just for information: build #708 still has lirc gpio problems. When I revert to #619 all works okay.

I've tried to re-record the remote keys, but they do not *all* replay in irw after they have been recorded by irrecord. Seems to be random as to which ones don't record/play-back properly (although KEY_OK seems to consistently not work in play back, but the code sequence is recorded in irrecord).

The repeat "key_up" bug seems to have gone though.

Let me know if I can help debug this.

Can you paste the output of:
Code:
systemctl status | paste

Hi, i also have this problem with recent builds, my output of systemctl status | paste :
http://sprunge.us/fRNU

for me its the ok button that stopped working and doesnt show up on irw anymore (quite annoying)
is this a kernel update related problem?
(2016-07-11, 12:41)popcornmix Wrote: [ -> ]Just to confirm, the DVB dongle is a USB device connected to the Pi you are watching?
Yes.
Quote:Can you test if disabling deinterlace, or switching the deinterlace method from automatic to bob makes a difference?
Just tested, no errors when deinterlace is disabled. Lots of errors with Bob and Advanced.
(2016-07-11, 16:02)smp1 Wrote: [ -> ]Just tested, no errors when deinterlace is disabled. Lots of errors with Bob and Advanced.

Jarvis defaults to bob when deinterlacing HD. Krypton defaults to advanced with HD,

Can you double check with Krypton and Jarvis using bob and advanced?

I could understand advanced giving errors on both Jarvis and Krypton.
I can't think of a reason why bob would give errors on Krypton and not on Jarvis.
(2016-07-11, 15:49)Cerv4nteZ Wrote: [ -> ]
(2016-07-09, 21:39)Milhouse Wrote: [ -> ]
(2016-07-09, 12:45)stevegal Wrote: [ -> ]Just for information: build #708 still has lirc gpio problems. When I revert to #619 all works okay.

I've tried to re-record the remote keys, but they do not *all* replay in irw after they have been recorded by irrecord. Seems to be random as to which ones don't record/play-back properly (although KEY_OK seems to consistently not work in play back, but the code sequence is recorded in irrecord).

The repeat "key_up" bug seems to have gone though.

Let me know if I can help debug this.

Can you paste the output of:
Code:
systemctl status | paste

Hi, i also have this problem with recent builds, my output of systemctl status | paste :
http://sprunge.us/fRNU

for me its the ok button that stopped working and doesnt show up on irw anymore (quite annoying)
is this a kernel update related problem?

Same here on RPI2. Anything else we can do/deliver?
Regarding the miniDV-AVIs from this post: 2372251 (post)
ffmpeg is totally capable of decoding CVSD and DVSD. Not MSDV however, maybe I was mistaken about that or that FourCC is just used for type1-AVIs.
I'm testing a standalone ffmpeg version for Windows and the video stream is detected as "dvvideo":
Code:
C:\dv>C:\prg\ffmpeg.x64\bin\ffmpeg -i sample_dvsd.avi -vcodec h264 output_dvsd.avi
ffmpeg version N-75716-g061b67f Copyright (c) 2000-2015 the FFmpeg developers
  built with gcc 5.2.0 (GCC)
  configuration: --enable-gpl --enable-version3 --disable-w32threads --enable-avisynth --enable-bzlib --enable-fontconfig --enable-frei0r --enable-gnutls --enable-iconv --enable-libass --enable-libbluray --enable-libbs2b --enable-libcaca --enable-libdcadec --enable-libfreetype --enable-libgme --enable-libgsm --enable-libilbc --enable-libmodplug --enable-libmp3lame --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenjpeg --enable-libopus --enable-librtmp --enable-libschroedinger --enable-libsoxr --enable-libspeex --enable-libtheora --enable-libtwolame --enable-libvidstab --enable-libvo-aacenc --enable-libvo-amrwbenc --enable-libvorbis --enable-libvpx --enable-libwavpack --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxavs --enable-libxvid --enable-lzma --enable-decklink --enable-zlib
  libavutil      55.  2.100 / 55.  2.100
  libavcodec     57.  4.100 / 57.  4.100
  libavformat    57.  2.102 / 57.  2.102
  libavdevice    57.  0.100 / 57.  0.100
  libavfilter     6.  9.101 /  6.  9.101
  libswscale      4.  0.100 /  4.  0.100
  libswresample   2.  0.100 /  2.  0.100
  libpostproc    54.  0.100 / 54.  0.100
Guessed Channel Layout for  Input Stream #0.1 : stereo
Input #0, avi, from 'sample_dvsd.avi':
  Duration: 00:00:10.00, start: 0.000000, bitrate: 30353 kb/s
    Stream #0:0: Video: dvvideo (dvsd / 0x64737664), yuv420p, 720x576 [SAR 16:15 DAR 4:3], 28915 kb/s, 25 fps, 25 tbr, 25 tbn, 25 tbc
    Stream #0:1: Audio: pcm_s16le ([1][0][0][0] / 0x0001), 48000 Hz, 2 channels, s16, 1536 kb/s
[libx264 @ 0000020db46a1700] using SAR=16/15
[libx264 @ 0000020db46a1700] using cpu capabilities: MMX2 SSE2Fast SSSE3 SSE4.2 AVX
[libx264 @ 0000020db46a1700] profile High, level 3.0
Output #0, avi, to 'output_dvsd.avi':
  Metadata:
    ISFT            : Lavf57.2.102
    Stream #0:0: Video: h264 (libx264) (H264 / 0x34363248), yuv420p, 720x576 [SAR 16:15 DAR 4:3], q=-1--1, 25 fps, 25 tbn, 25 tbc
    Metadata:
      encoder         : Lavc57.4.100 libx264
    Stream #0:1: Audio: mp3 (libmp3lame) (U[0][0][0] / 0x0055), 48000 Hz, stereo, s16p
    Metadata:
      encoder         : Lavc57.4.100 libmp3lame
Stream mapping:
  Stream #0:0 -> #0:0 (dvvideo (native) -> h264 (libx264))
  Stream #0:1 -> #0:1 (pcm_s16le (native) -> mp3 (libmp3lame))
Press [q] to stop, [?] for help
frame=  250 fps= 81 q=-1.0 Lsize=    7587kB time=00:00:10.00 bitrate=6210.5kbits/s
video:7405kB audio:157kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: 0.338777%
[libx264 @ 0000020db46a1700] frame I:1     Avg QP:26.76  size: 71606
[libx264 @ 0000020db46a1700] frame P:127   Avg QP:27.70  size: 46351
[libx264 @ 0000020db46a1700] frame B:122   Avg QP:30.89  size: 13315
[libx264 @ 0000020db46a1700] consecutive B-frames:  2.4% 97.6%  0.0%  0.0%
[libx264 @ 0000020db46a1700] mb I  I16..4:  0.4% 91.7%  7.8%
[libx264 @ 0000020db46a1700] mb P  I16..4:  0.3%  9.1%  0.4%  P16..4: 40.0% 28.3% 20.5%  0.0%  0.0%    skip: 1.4%
[libx264 @ 0000020db46a1700] mb B  I16..4:  0.0%  0.0%  0.0%  B16..8: 24.4%  6.9%  3.7%  direct:31.6%  skip:33.4%  L0:31.9% L1:26.2% BI:42.0%
[libx264 @ 0000020db46a1700] 8x8 transform intra:92.1% inter:63.9%
[libx264 @ 0000020db46a1700] coded y,uvDC,uvAC intra: 92.5% 70.1% 8.3% inter: 62.7% 24.0% 0.5%
[libx264 @ 0000020db46a1700] i16 v,h,dc,p: 33% 34% 15% 18%
[libx264 @ 0000020db46a1700] i8 v,h,dc,ddl,ddr,vr,hd,vl,hu:  7% 26% 29%  3%  5%  3% 10%  4% 13%
[libx264 @ 0000020db46a1700] i4 v,h,dc,ddl,ddr,vr,hd,vl,hu:  7% 45% 16%  3%  4%  3%  8%  3% 11%
[libx264 @ 0000020db46a1700] i8c dc,h,v,p: 63% 22% 12%  3%
[libx264 @ 0000020db46a1700] Weighted P-Frames: Y:56.7% UV:25.2%
[libx264 @ 0000020db46a1700] ref P L0: 50.1% 15.5% 15.1% 14.1%  5.2%
[libx264 @ 0000020db46a1700] ref B L0: 77.7% 22.3%
[libx264 @ 0000020db46a1700] kb/s:6066.10
(2016-07-11, 16:07)popcornmix Wrote: [ -> ]I can't think of a reason why bob would give errors on Krypton and not on Jarvis.
Nevermind my previous post, I did more testing and no errors in Bob mode after 6 hours of running the HD channel. Is it normal that the Advanced mode doesn't work properly with TVH on RPi3? Will it make sense to overclock the GPU? I will try "h264_freq=1000" and leave it on an HD channel overnight.
New LibreELEC.tv Krypton build #0711: RPi / RPi2
(Supercedes previous build)

Code:
# uname -a
Linux rpi512 4.7.0-rc6 #1 Mon Jul 11 22:57:12 BST 2016 armv6l GNU/Linux

# vcgencmd version
Jul  8 2016 14:34:25
Copyright (c) 2012 Broadcom
version f9e71a2b696c48a14756c1e756f3cda6d6f87a07 (clean) (release)

# lsb_release
LibreELEC (Milhouse) - Version: devel-20160711225532-#0711-g1cfefd3 [Build #0711]

# vcdbg log msg 2>&1 | grep DTOK
002472.352: Kernel trailer DTOK property says yes

# Kernel device tree status: Enabled

Based on tip of LibreELEC.tv master (1cfefd35, changelog) and tip of XBMC master (874b7bdc, changelog) with the following modifications: Build Highlights:
  1. PVR change
Build Details:
  1. XBMC:
    • [PVR] Recordings window: make 'group items' setting persistent. (PR:10101, 1 commit, 6 files changed)
(2016-07-11, 15:49)Cerv4nteZ Wrote: [ -> ]Hi, i also have this problem with recent builds, my output of systemctl status | paste :
http://sprunge.us/fRNU

for me its the ok button that stopped working and doesnt show up on irw anymore (quite annoying)
Thanks for the info!

systemctl output shows that lircd-uinput is running (this should be disabled in recent Milhouse builds). Just to be sure, could you update to the latest Milhouse build and post the systemctl output of that?

Also: what kind of remote are you using? And, as you seem to be using a custom lircd.conf, could you please upload that as well?
Code:
paste /storage/.config/lircd.conf

Quote:is this a kernel update related problem?
It seems to be a problem caused by the lircd update a few weeks ago, but not all remotes seem to be affected by it.

Therefore it's important to know which kind of remote you are using, which keys are not working and if you are using a custom lircd.conf. That info should help us to reproduce the issue and track down the root cause of it.

so long,

Hias
(2016-07-11, 16:15)mule1 Wrote: [ -> ]Same here on RPI2. Anything else we can do/deliver?
Could you please tell us which remote you are using and also post the output of "systemctl status | paste" and - if you are using a custom lircd.conf - post that file as well?

so long,

Hias
(2016-07-12, 00:43)HiassofT Wrote: [ -> ]
(2016-07-11, 16:15)mule1 Wrote: [ -> ]Same here on RPI2. Anything else we can do/deliver?
Could you please tell us which remote you are using and also post the output of "systemctl status | paste" and - if you are using a custom lircd.conf - post that file as well?

so long,

Hias
I'm using a harmony remote which has been programmed as Intel D34010WYK (should be the same as the MCE Remote). I'm not using a custom lircd.conf.
systemctl status | paste using Build #0711

Affected button is for example OK while cursor keys are not affected.
Hi.
Im using a crypto remote with a custom lircd.conf.
Using build 711. http://sprunge.us/eBYO
key OK , Numeric keys not working.
Cursor keys are ok..
(2016-07-11, 21:52)smp1 Wrote: [ -> ]Nevermind my previous post, I did more testing and no errors in Bob mode after 6 hours of running the HD channel. Is it normal that the Advanced mode doesn't work properly with TVH on RPi3? Will it make sense to overclock the GPU? I will try "h264_freq=1000" and leave it on an HD channel overnight.

h264_freq=1000 doesn't make sense. The maximum overclock is around 500.

I believe the issue is sdram bandwidth. advanced deinterlace takes enough sdram bandwidth that it can interfere with some types of dvb dongle.
polo_joe did have a similar issue, but he changed the dvb dongle to a different make which didn't have the issue.

So, options are:
use bob deinterlace
use a different dvb dongle
run the dongle/backend on a separate machine
overclock sdram
tweak the sdram priorities

If you are happy to do some experiments, there are a few controls that affect sdram priorities which I'd like to know if they can help.
The most likely one to help is v3d_limiter which defaults to 0x0880f1. The bits define as:
Code:
bit 0 is "enable"
bits 7:4 are "max_priority" (don't limit if priority is greater than this, so set to 0xf)
bits 15:8 are "increment" (add this to 8-bit counter every time we see a transaction)
bits 23:16 are "holdoff" (number of cycles to disable the bus for when 8-bit counter overflows)

So, to reduce artefacts (although at the cost of making the deinterlace take longer which could result in skips),
you want to either increase the "holdoff" (the initial 08), or increase the "increment" (the next 80).

As the increment is already quite high, the holdoff is more likely to help. Can you try with it increased to 0x10?
i.e. add v3d_limiter=0x1080f1 to config.txt and reboot and test. If that doesn't help, then try v3d_limiter=0x2080f1.
(2016-07-12, 00:43)HiassofT Wrote: [ -> ]Could you please tell us which remote you are using and also post the output of "systemctl status | paste" and - if you are using a custom lircd.conf - post that file as well?

Please see this post for the lircd.conf for one remote that isn't working, along with a list of the buttons that don't work - that remote is from the 'Top Up TV' branded Thomson box DTI6300. (This remote)

Also not working is the Xbox 360 Media Remote (this remote), and the associated lircd.conf - I haven't had the time to test this one fully, but at the very least the OK and the Info buttons aren't working.

systemctl status from build 0711

Cheers.