[Kernel-packages] [Bug 1403152] Re: unregister_netdevice: waiting for lo to become free. Usage count

2015-03-27 Thread Jordan Curzon
An addition to my #18 comment above. I forgot that I had to remove logging on calls from skb_release_head_state to dst_release due to performance impact of the number of calls. In both failure and success scenarios, calls to dst_release by skb_release_head_state are also present. -- You received

[Kernel-packages] [Bug 1403152] Re: unregister_netdevice: waiting for lo to become free. Usage count

2015-03-24 Thread Jordan Curzon
To add to my comment just above, we have found a workload that is not under own our control (which limits what we can do with it, including sharing it) but which exacerbates the issue on our systems. This workload causes the issue less than 5% of the time and that gives us the chance to look at

[Kernel-packages] [Bug 1403152] Re: unregister_netdevice: waiting for lo to become free. Usage count

2015-03-23 Thread Jordan Curzon
I'm working with Rodrigo Vaz and we've found some details about our occurrence of this issue using systemtap. rt_cache_route places a dst_entry struct into a fib_nh struct as the nh_rth_input. Occasionally the reference counter on that dst is not decremented by the time free_fib_info_rcu is