On 2015/08/17 19:55, Mark Kettenis wrote: > The diff below is a first step towards running the em(4) interrupt > handler without grabbing the kernel lock. It runs the rx completion > path without the lock, which is the important bit to be able to test > the work that has been going on to make the network stack run on > multiple CPUs. But the handler still grabs the kernel lock for tx > completions, rx ring refills and everything else. Which means that it > will still grab the lock every time it runs, which may affect how it > interacts with the network stack.
I'm running it here without problems so far, and with snapshots being built again it seems a good time for it to go in. OK with me.
