Abstract:
FairThreads offers a simple framework for concurrent and parallel programming. Basically, it defines "schedulers" which are synchronization servers, to which threads can dynamically link or unlink. All threads linked to the same scheduler are executed in a cooperative way, at the same pace, and they can synchronize and communicate using broadcast events. Threads which are not linked to any scheduler are executed by the OS in a preemptive way, at their own pace. FairThreads offers programming constructs for dynamically linking and unlinking threads. FairThreads is fully compatible with the standard Pthreads library and has a precise and clear semantics for its cooperative part; in particular, systems exclusively made of threads linked to one unique scheduler are actually completely deterministic. Special threads, called "automata", are provided for short-lived small tasks or when a large number of tasks is needed. Automata do not need the full power of a native thread to execute and thus consume less resources.

Slides (.ps)