Fix for Loop Relaxing
Two of the "tight" loops should have been relaxed when no packets were available for processing to reduce pointless CPU burn, with the small price of delaying some packets 10ms or so. Running without this change on an EC2 t3a.nano instance would burn 75% of the CPU, versus a fraction of that with this fix.
Full Changelog: 2.1.0...2.1.1