The following exercises cover the various topics from this chapter. Most of the exercises require implementing new concurrent data structures using atomic variables and the CAS instruction. These data structures can also be solved using the synchronized
statement, so it is helpful to contrast the advantages of the two approaches.
Implement a custom
ExecutionContext
class calledPiggybackContext
, which executesRunnable
objects on the same thread that callsexecute
. Ensure that aRunnable
object executing on thePiggybackContext
can also callexecute
and that exceptions are properly reported.Implement a
TreiberStack
class, which implements a concurrent stack abstraction:class TreiberStack[T] { def push(x: T): Unit = ??? def pop(): T = ??? }
Use an atomic reference variable that points to a linked list of nodes that were previously pushed to the stack. Make sure that your implementation is lock-free and not susceptible to the ABA problem.
Implement a
ConcurrentSortedList
class,...