Adding this issue here, as in the past we saw linked issues between netvsc reload and mtu change.
The same hung task failure occurs when changing the MTU on the interface. Repro steps: ip link set dev eth0 mtu 3000 ip link set dev eth0 mtu 4500 At this point the command will hang and then run into the hung messages after 120 seconds. ** Description changed: On proposed 4.13.0-1001.1 linux-azure kernel reloading the netvsc module will cause the VM network to stop working on WS2016. Manual repro is easy, start a VM with that kernel and run: modprobe -r hv_netvsc modprobe hv_netvsc At this point the command will hang and after 2 minutes the hung task messages will appear. Test env info: - ubuntu 16.04.3 - - Affected platforms: WS2016 and WS2016 fall update - 1709 + - Affected platforms: + a) WS2016 and WS2016 fall update - 1709 + b) WS2012R2 4.11 series linux-azure are not showing this behavior. - This seems to affect only WS2016, and it is functional on WS2012R2. -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/1735546 Title: [hyper-v] reloading netvsc issue on linux-azure 4.13.0-1001.1 To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux-azure/+bug/1735546/+subscriptions -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs