Osmc on rpi3 freezes almost daily

Hi there :),

My RPi3 with the latest version of OSMC freezes almost daily now.
It used to happen sometimes while I was using it which I could reset via command line with ‘sudo systemctl stop/start mediacenter’.
However, now it also happens when I am not using it, after which I cannot login with ssh anymore either.
Only solution seems to be a hard reset (dis/connect power cable), after which it works fine again.

My logs are posted here: https://paste.osmc.tv/okiyavisoq
After it froze, I applied the hard reset just before 16:00 I believe.
Don’t know if/where logs from before hard resetare stored?

Have been using rpi3+osmc with great pleasure for the last few years, and did not change anything significant before this started to happen.

Thanks in advance for your help,


22:37:51.450 T:1959002112  NOTICE: Config:

Some of your clockings are concerning. You are actually underclocking arm for a Pi3 and overclocking core which wouldn’t be surprising to cause such instability as you are reporting. You should just use the MyOSMC addon to set to normal overclock to get standard clockspeeds for a Pi3. Make sure you reboot afterward.

Thank you for your quick reply. :slight_smile:

I had indeed at one point changed overclocking to the default Turbo profile as my RPi3 was struggling to playback some 1080p H265 files (to no avail ;)).
The normal overclock default is 900/450/450 hz, while turbo is 1000/500/500 hz.
However, I found out from a different thread here that if you set it to normal it will actually be ignored and use the rpi3’s native cpu frequency.
Is this true?

I have set it to normal now and rebooted, let’s hope this helps. :slight_smile:

Ah, it states that it is running at 1200hz now at settings → system info → hardware.
Also, it runs a lot faster again. :slight_smile:
I will post back here if it does not freeze in the next week.

Should the standard overclocking profiles not be updated for the RPi3 then?
As the current ones seem rather confusing (and created my problem)…


It is still running stable now, so this issue is solved and can be closed. :slight_smile: