Any Plans to Support Synchronization Primitives?

[joshuadahlunr]

If "The workers inside libfork's thread pools should never block" then by extension they shouldn't block waiting on a mutex or condition variable. Are there any plans to provide similar synchronization primitives for libfork? Or is the convention to just allow blocking in this case?

[ConorWilliams]

The problem with coroutine synchronisation primitives is that they often serialize work that is intended to be parallel. I would like to believe that fork join is powerful enough that no further synchronisation primitives are required for compute bound work, for io bound work I would like to provide some kind of io pool that task can be moved to.