locked
insmod error inserting '/lib/hv_timesource.ko' after kernel upgrade on Centos RRS feed

  • Question

  • Hi everyone,

    My apoligise for the lack of advanced knowledge in Centos. I have been asked to perform a yum upgrade on a Centos 5.9 Final running as a Hyper-V machine on Windows 2008 R2.

    After running into kernel panics after the kernel upgrade, I followed this link: http://www.sudonym.com/398/kernel-panic-after-yum-update-centos-with-hyper-v-linux-integration-components/

    The information provided helped in in solving the kernel panic issue.

    However, now I am getting the following warning when Centos boots:

    insmod error inserting '/lib/hv_timesource.ko': -1 device or resource busy

    The system boots normally. However, I am worried that this might cause problems as I don't have a clue what is error is about. I tried googling but to no avail.

    I am running kernel version 2.6.18-348.3.1.el5.

    Can somebody please give me some clues on how to remove this warning and solve this issue?

    Thanks in advance

    Wednesday, April 10, 2013 1:56 PM

All replies

  • I followed the same instructions from sudonym.com and am experiencing the same issue.  Only difference is I'm on kernel 2.6.18-371.11.1.el5.

    Any info on this would be greatly appreciated!

    Edit:

    Doing a little more digging I see in /var/log/messages:

    kernel: Time: hyperv_clocksource clocksource has been installed.

    kernel: HyperV recommendations: 2c
    kernel: HyperV spin count: fff

    Then a little bit after, within the same second, I see the following which I think is resulting in the error insmod error inserting '/lib/hv_timesource.ko': -1 device or resource busy:

    kernel: Registering HyperV clock source
    kernel: register_clocksource: Cannot register hyperv_clocksource. Already registered!

    I don't fully understand Linux and it's modules, but my assumption is that the time source is already set to the Hyper-V clock and when loading the module hv_timesource.ko it detects that and refuses to load.

    This error can be recreated after boot by doing a
    insmod /lib/modules/2.6.18-371.11.1.el5/kernel/drivers/vmbus/hv_timesource.ko
    Note: you will need to change 2.6.18-371.11.1.el5 to the kernel version you are using.

    Additionally, when I go to list the current time source it shows that the server is using hyperv_clocksource which makes me assume that everything is setup correctly and that the error can be safely ignored.

    To get this information:
    cat /sys/devices/system/clocksource/clocksource0/current_clocksource

    To list all available clock sources:
    cat /sys/devices/system/clocksource/clocksource0/available_clocksource

    Finally, I don't notice my system clock drifting at all.  I have been keeping an eye on ntp using ntpq -p and so far all looks well.

    • Edited by workshad Friday, August 22, 2014 6:53 PM Added more debugging info that I gathered
    Friday, August 22, 2014 5:09 PM