Guest - Testers are needed for the reworked CDateTime core component. See... https://forum.kodi.tv/showthread.php?tid=378981 (September 29) x
  • 1
  • 41
  • 42
  • 43(current)
  • 44
  • 45
  • 168
OpenELEC Testbuilds for RaspberryPi (Kodi 16.0)
(2015-08-13, 00:38)zaphod24 Wrote: What format is your HD live tv in? Looks like you're in the UK, I'm in the US and mine is 1080i60 and MPEG2. I'm using the following overclock settings and have enabled the new advanced deinterlacing with great success:

dtoverlay=sdhost,overclock_50=84
gpu_mem_1024=320
arm_freq=1050
core_freq=525
gpu_freq=350
over_voltage=4

Instead of force_turbo=1 I use "echo performance > /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor" in autostart.sh. I'm running 0810 and don't have any issues other than occasional odd messages in the system log like:

[ 544.967786] mceusb 1-1.5:1.0: Error: urb status = -71
[11904.044682] NYET/NAK/ACK/other in non-error case, 0x00000002

Those do not appear to affect video playback, network or IR receiver response that I can tell.
Yes I'm in the UK. Here the HD channels use DVB-T2 with H.264 and AAC. I believe the broadcasts are either 1080i50 or 1080p25.
My overclock is very similar to yours.

gpu_mem_1024=320
arm_freq=1050
core_freq=500
gpu_freq=400
over_voltage=4
force_turbo=1
avoid_pwm_pll=1
dtoverlay=sdhost,overclock_50=84
Leopold's Repository: Home of LibreELEC Dev Updater ...
Ok. Lots of freezes requiring pulling the power tonight. I tried various things including disabling overclock and enabling overclock but disabling the new auto turbo. The only thing that seems to work is changing deinterlacing back to auto instead of mmal-advanced. I was able to get it to freeze many times by skipping a minute at a time forward no more than 10 times. Not sure if the output of any of the previously mentioned vcgencmd or vcdbg would help. Many times my ssh session stopped too but I was able to kill it and logon with ssh again. Most times even rebooting would seem to work but then require pulling the power.

Guess it is back to bob deinterlacing for now. Be glad to run any of the above commands if reproducing the freezes isn't easy for Millhouse or popcornmix.
I've also had some playback freezes with the latest build, and running vcgencmd at the command line will hang too, which suggests (I think) that the ARM/GPU communication is being hosed.
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.
Thanks for the update. I appreciate the continued effort on advanced deinterlacing and auto turbo. I'll be glad to test future builds with advanced deinterlacing enabled if/when these latest issues get ironed out. I can also provide some log output if it would help. For now, boring bob deinterlacing gets the job done.
I have tested the new build and there is still a memory leak but I assume that you have not have time yet to look into this problem. So just for info that leak is still there.
Build #0812 seems to be working well for me. No lock ups so far...

Something I have noticed recently is that for live tv to play completely smoothly I have to disable "Sync playback to display". The issue is only really noticeable when scrolling text on screen shows up a slight judder. I thought it worth mentioning since this setting is now on by default. Unless I've misunderstood what the option does I would expect it to have no effect on playback of 25fps or 50fps video in 50Hz mode.
Leopold's Repository: Home of LibreELEC Dev Updater ...
I had a CEC related SEGV with build #0810. I don't use CEC so I've now disabled the CEC adapter.
http://sprunge.us/VDOj
Leopold's Repository: Home of LibreELEC Dev Updater ...
I have huge problems with all the latest Nightly Builds. The problems are similar to the ones illiac4 and others are describing but for me it happens also during video playback via NAS (NFS connection).
Mostly it happens when one movie has finished and the next one is loading. System (GUI) freezes. SSH connection works and kodi process runs at about 30-50% load all the time. I have to reboot PI2 via SSH all the time when this happens. System is pretty much out of the box with no overclocking. Only a custom installation of hyperion is running to drive my lightberry system.

As the system seems to hang (infinite loop) there is no debug log. What can I do to provide more info for you guys? I'm currently on #0811 and today morning system again hung.
(2015-08-13, 11:39)gandharva Wrote: I have huge problems with all the latest Nightly Builds. The problems are similar to the ones illiac4 and others are describing but for me it happens also during video playback via NAS (NFS connection).
Mostly it happens when one movie has finished and the next one is loading. System (GUI) freezes. SSH connection works and kodi process runs at about 30-50% load all the time. I have to reboot PI2 via SSH all the time when this happens. System is pretty much out of the box with no overclocking. Only a custom installation of hyperion is running to drive my lightberry system.

As the system seems to hang (infinite loop) there is no debug log. What can I do to provide more info for you guys? I'm currently on #0811 and today morning system again hung.

When it hangs, does "systemctl restart kodi" produce a working system or will only a reboot fix it?
When you say "it happens when one movie has finished and the next one is loading" are you using a playlist (or remote with queueing?)
Does the problem occur if first movie is stopped before second movie starts?

Can you describe a way of me reproducing it? E.g. from clean install, without hyperion installed and default settings does it fail?
Can you make it fail with a standard video like Big Buck Bunny? If so describe exactly how you provoked the failure.
A debug log (wiki) may be useful.
Anyone with freezes.
Does disabling advanced deinterlace (or deinterlace completely) help?
Does disable_auto_turbo=1 help?
Does adding force_turbo=1 and removing any other overclock settings help?
(2015-08-13, 12:45)popcornmix Wrote: Anyone with freezes.
Does disabling advanced deinterlace (or deinterlace completely) help?
Does disable_auto_turbo=1 help?
Does adding force_turbo=1 and removing any other overclock settings help?

Do you need anything more from my system? Do you think the memory leak is fixable at pi2. At pi1 there is no leak.
My guess is it is related to some sort of flush/memalloc/memfree in Kodi GUI.

When playing tv-shows (e.g. Scrubs) I have set Kodi to automatically play the next video. This works for some time (2-3 episodes, sometimes even more).
The freeze always triggers when one episode has finished and the "busy" icon in the bottom right corner pops up right before the next episode starts. This is the point where it freezes. Nearly all the time the "busy" icon is visible when GUI freezes. I have MMAL and OMX enabled, but disabling one or the other didn't help. Also filetypes do not matter it freezes with MPEG2 AVIs and also with x.264 MKVs.

The other more or less reliable way to reproduce is when navigating the menus. Freeze here also happens at the exact moment when some submenu should load (screen content is subject to change). Instead of entering the submenu or starting up the plugin, the GUI freezes. The GUI freezes I think, only trigger when I play some tv-shows/movies before. I think this cannot be triggered by only navigating the menu.

I will check your proposed settings and also the "systemctl restart kodi" command. Until now I always did it with the "reboot" command.
(2015-08-13, 13:28)gandharva Wrote: The other more or less reliable way to reproduce is when navigating the menus. Freeze here also happens at the exact moment when some submenu should load (screen content is subject to change). Instead of entering the submenu or starting up the plugin GUI freezes. The GUI freezes I think only triggered when I played some tv-shows/movies before. I think this cannot be triggered by only navigating the menu.

Do you get the hang while navigating with a video playing in the background, or are you just saying you played and stopped a video previously?
Are you overclocking? If so then disable that and confirm the problem still occurs.
Not with the video in background. I just played some videos before. No overclock at all, just SPI overlay activated to get my lightberry system working.

dtparam=spi=on
dtoverlay=spi-bcm2708 (need to set this too, else my lightberry randomly flickers)
Crash during video playback in latest nightly #0812.

Code:
Program terminated with signal SIGSEGV, Segmentation fault.
#0  0x0096abc4 in PLT_CtrlPoint::RenewSubscriber(NPT_Reference<PLT_EventSubscriber>) ()

Full log: http://pastebin.com/JrJruQeR

Relevant dmesg:

Code:
[    6.014547] usbcore: registered new interface driver usbtv
[    6.212337] smsc95xx 1-1.1:1.0 eth0: hardware isn't capable of remote wakeup
[    6.212661] IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready
[    7.767164] IPv6: ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready
[    7.770326] smsc95xx 1-1.1:1.0 eth0: link up, 100Mbps, full-duplex, lpa 0xCDE1
[ 1319.138876] Transfer to device 4 endpoint 0x3 frame 853 failed - FIQ reported NYET. Data may have been lost.
[ 3454.858352] vchiq: 0: ignoring ERROR from callback to service 12011
  • 1
  • 41
  • 42
  • 43(current)
  • 44
  • 45
  • 168

Logout Mark Read Team Forum Stats Members Help
OpenELEC Testbuilds for RaspberryPi (Kodi 16.0)10
This forum uses Lukasz Tkacz MyBB addons.