Lock-free Atomic Shared Pointers Without a Split Reference Count - Implementation Strategies

Lock-free Atomic Shared Pointers Without a Split Reference Count - Implementation Strategies

Save Big on Coursera Plus. 7,000+ courses at $160 off. Limited Time Only! Grab it Explore a groundbreaking approach to implementing lock-free atomic shared pointers without using a split reference count in this CppCon 2023 talk. Delve into the challenges of using std::shared_ptr in concurrent programming and learn about a novel strategy that achieves excellent performance without relying on double-word compare-exchange operations or pointer bit manipulation. Discover how this elegant solution leverages hazard pointers and deferred reclamation, making it timely and simple to implement with nearly-standard C++. Gain insights from Daniel Anderson, an assistant teaching professor at Carnegie Mellon University, as he presents this innovative technique for managing dynamic memory in highly concurrent C++ applications.