Comment 1 for bug 1792959

Thanks for your bugreport.

The 300 nanoseconds apparently comes from a usleep(3) in GO runtime (runtime/proc.go:runqgrab to be precise). I have not looked into the details why this is happening yet, the internet (https://stackoverflow.com/questions/35155119/how-to-optimize-golang-program-that-spends-most-time-in-runtime-osyield-and-runt) indicates that we might trigger too much GC for some reason. But it might also be a red-herring. The next step here is probably to run pprof to see what triggers the usleep(3).