Could read-copy-update wait/wake speed up fibers?

Fibers used [in the skynet benchmark](https://github.com/atemerev/skynet/blob/master/guile-fibers/skynet.scm) currently takes 2-4 seconds while the best contenders are at 200-400ms, so there seems to be some bottleneck.

old asked on IRC, whether the [Read Copy Update wait/wake pattern](https://github.com/urcu/userspace-rcu/blob/master/src/urcu-wait.h) could speed them up.