9 May
2018
9 May
'18
5:30 a.m.
On a second thought, maybe I could even change the implementation of the future_queue to internally use e.g. the well-known boost::lockfree queues together with some semaphore to implement the blocking behaviour. I am not 100% sure yet, but I think all current features could be implemented by that. Do you think this would improve the implementation? It might be a bit simpler and maybe even more efficient in some cases.
I would love to see a portable eventcount (http://www.1024cores.net/home/lock-free-algorithms/eventcounts) implementation in Boost.Lockfree.
i'm not sure, if boost.lockfree is a good place for it (is it even possible to implement them without relying on system calls, which may acquire locks in kernel-space?)