A Scalable Adaptive Locking Mechanism for High Performance Computing

With the rapid development of multi-core processor technology, most multi-threaded applications have performance scalability bottleneck in the multi-core system environment. That is, the applications cannot make full use of the performance advantages of multiple cores, and its performance decreases as the number of processors increases. In application scenarios with different reader/writer ratios, the performance of lock mechanism also differs, and limits the scalability of applications concurrency, restricting the scalability of the performance of applications in multi-core systems. Therefore, in a multi-core environment, it is of practical significance to implement a new type of lock mechanism with scalable performance. This paper proposes a simple, efficient, optimized, and easy-to-implement adaptive reader-writer lock mechanism which automatically adjusts the lock acquisition path according to the read/write ratio of the application scenario. There are only minor changes to the application by replacing the original reader-writer lock instance and API. Through the performance comparison test results, it can be seen that the concurrency and scalability of the reader-writer lock in the multi-core system are improved, and it has the advantages of simplicity, efficiency, optimization, and easy implementation. It can automatically adjust the way of lock acquisition based on the application scenarios of different read/write ratios with good scalability.