Can someone with Pi0 or Pi1 do some testing?
No early Pis here but works on my Pi2
I’m getting the same Kodi startup error as in post #12: corrupted double-linked list
osmc@osmc:~$ sudo systemctl start mediacenter osmc@osmc:~$ sudo systemctl status mediacenter ● mediacenter.service - media center application Loaded: loaded (/lib/systemd/system/mediacenter.service; disabled) Active: active (running) since Wed 2017-05-31 15:08:01 BST; 11s ago Main PID: 622 (mediacenter) CGroup: /system.slice/mediacenter.service └─622 /bin/bash /usr/bin/mediacenter May 31 15:08:06 osmc mediacenter: ERROR: Unable to create application. Exiting May 31 15:08:06 osmc mediacenter: *** Error in `/usr/lib/kodi/kodi.bin': corrupted double-linked list: 0x038d9ef0 *** May 31 15:08:06 osmc mediacenter: Kodi exited with return code 134 after 0 hours, 0 minutes and 5 seconds
So either kodi.bin really is corrupt or there’s some other legacy from running with the bad kernel.
Anyone care to give me an md5 of kodi.bin on a Pi 2/3?
osmc@osmc:~$ md5sum /usr/lib/kodi/kodi.bin
Try sudo apt-get install --reinstall rbp2-mediacenter-osmc
Thanks, that matches what I have, so not kodi.bin itself.
Reinstall the next step… It seems to have worked. I’ll test it when I get home.
Everything I installed from the App Store works. The Youtube addon I installed works.
Do you want us to test something specific @sam_nazarko?
EDIT: Adding the same disk attached via usb on the Vero 4k hung the rpi3, and it became very hot.
Logs here: https://paste.osmc.tv/zuxicujihe
Since I know my rpi3 might be faulty (notice the “usb 1-1-port2: over-current change” spam) I wonder if it can still be a valid candidate to test the new kernel without giving false positives.
Pi1 (B+) seems to work just fine too.
I’ve only done the most basic or testing so far but it boots, I can update my library and scrape the content and I can play back movies as well as home videos and pictures.
imho push it, stable enough
You realise you’ll be press-ganged into full-time support if it all goes bad on us.
atleast for rpi is good enough been pushing in my usual style doing my modifications using peripherals not included in the OSMC Distro and its all holding.
I tried to give the 4.9 kernel a good workout today by compiling mediacenter-osmc but it failed:
<snip> Updating sources Sources updated successfully Package armv7-toolchain-osmc is already installed in the environment Installing /etc/resolv.conf Installing /etc/network/interfaces chroot: failed to run command ‘/usr/bin/make’: No such file or directory Makefile:8: recipe for target 'rbp2' failed make: *** [rbp2] Error 127 osmc@osmc:/media/29GB/osmc/package/mediacenter-osmc$
That was for a
make rbp2. Since I’ve never compiled mediacenter-osmc on the Pi before, I let it go and tried to compile connman, which has worked in the past.
<snip> Updating sources Sources updated successfully Package armv7-toolchain-osmc is already installed in the environment Installing /etc/resolv.conf Installing /etc/network/interfaces chroot: failed to run command ‘/usr/bin/make’: No such file or directory Makefile:8: recipe for target 'armv7' failed make: *** [armv7] Error 127 osmc@osmc:/media/29GB/osmc/package/connman-osmc$
The same error, this time for
make armv7. Looks like it could be something related to chroot.
I know I I’ve changed a few things on the box, including running the 2017.04-2 update, since I last compiled connman but the biggest change has been the 4.9 kernel.
Edit: I forgot to mention that this was a completely new git clone…
seems that its missing make in chroot, could be a script error in anycase thats kinda offtopic since this is about the package not the source
Sam knows his build environment. The build script worked a few weeks ago, so the 4.9 kernel cannot be ruled out.
sudo apt-get install --reinstall armv7-toolchain-osmc
You probably got corruption when we had a dodgy kernel
Unfortunately, it’s likely that there’ll be more iffy files, since I didn’t do anything remotely connected to compilations while the dodgy kernel was being used. I sense a reinstall looming on the horizon.
Yup, best way to go forward unfortunately.
It’s the risks of testing.
That’s why I do nightly backups
Interestingly enough, so do I (thanks to your script, BTW), so less of the smug, please.
But since files that I’ve not touched seem to have been corrupted on an apparently random basis, it’s going to be quite a job figuring out which ones to restore. Plus… (pauses for effect), I only have 7 days’ worth of backups - and the kernel was installed 10 days ago. And who’s to say that the USB pendrive with the backups wasn’t blitzed, as well? Nope, a full reinstall is definitely on the cards.
Opportunity for a good spring clean, then.