Posts by chris_kmn

    Good question. I honestly was not aware of that in my sources and I am not really a linux expert…


    But there are some additional files in the sources.d folder.


    What about backing up your sources and copy mine into your folder and see if it works ?


    One more thing to try is using the standard repository:


    apt-get install nvidia-xconfig

    I tried some more things and changed one setting that I have done to overcome issues with nvidia hw transcoding.


    in grub I changed the following kernel parameter:


    GRUB_CMDLINE_LINUX_DEFAULT="text acpi_osi=! \"acpi_osi=Windows 2015\" mem_sleep_default=deep pcie_aspm=off"


    As far as I understand is, that this is telling Linux it is on a windows system.



    After I changed it back to:


    GRUB_CMDLINE_LINUX_DEFAULT="text mem_sleep_default=deep"



    Everything seems to work. Autoshutdown service is active and HW transcoding is operative.....


    Does this make sense to you ?

    I don`t know how to tell....... But after suspend - and now with running autoshutdown - my hardware transcoding isn't working anymore...


    so my guess is, that there IS a conflict with the nvidia scripts from #43 :-( :-( :-(


    nvidia-smi isnt working anymore also. And to get it back working I have to reboot my system :-(

    Code
    root@MalibuNAS:~# nvidia-smi
    Unable to determine the device handle for GPU 0000:01:00.0: Unknown Error



    can't we go back to the original version (5.1.7 ?) ?

    I'll give it a try. But I already removed the plugin as you described, started from scratch - buth without purge. But after removing it again the result is:



    BUT:


    This time it seems to work.... and one more but: meanwhile I've changed my network setup from DHCP to static and also disabled the ipv6 support of my Plex server that is running on my OMV.


    This is the log:

    There might be a (major) difference in my setup. I am using nvidia hardware transcoding and therefore I had to install specific nvidia sleep and resume scripts. may be that could help to find the reason ?!


    /etc/systemd/system/nvidia-hibernate.service:



    /etc/systemd/system/nvidia-resume.service:


    /etc/systemd/system/nvidia-suspend.service:


    /usr/bin/nvidia-sleep.sh:


    and /lib/systemd/system-sleep/nvidia:

    Bash
    #!/bin/sh
    case "$1" in
    post)
    /usr/bin/nvidia-sleep.sh "resume"
    ;;
    esac



    does this make any sense ?

    I tried it, service doesn't start up after suspend:


    well... I think I misunderstood the problem with the path. I thought Dagobert did the mistake and did not realise that it is a mistake in #26.


    So I copied to the correct path and gave exec rights, suspend-resume and still error, but this time with an error in the log:


    and this is the journal log:



    but I will repeat everything one more time...

    Path is correct and I also gave exec rights to the files.


    after I've activated verbose and deleted the old log, the logfile was empty after restart from suspend (?), service was inactive.


    dmesg log shows interface is up after about 5 secs: