On Thu, 2016-12-08 at 20:45 +0100, Hannes Frederic Sowa wrote: > Hi Eric, >
> This patch mostly improves situation for non-connected sockets. Do you > think it makes sense to acquire the spinlock depending on the sockets > state? Connected UDP sockets flow in on one CPU anyway? We could do that, definitely. However I could not measure any difference with a conditional test here for connected socket. Maybe it would show up if we are unlucky and multiple cpus compete on the same cache line because of the hashed spinlock array.
