Reducing False Sharing on Shared Memory Multiprocessors

We have developed compiler algorithms that analyze coarse-grained, explicitly parallel programs and restructure their shared data to minimize the number of false sharing misses. The algorithms analyze the per-process data accesses to shared data, use this information to pinpoint the data structures that are prone to false sharing and choose an appropriate transformation to reduce it. The algorithms eliminated an average (across the entire workload) of 64% of false sharing misses, and in two programs more than 90%. However, how well the reduction in false sharing misses translated into improved execution time depended heavily on the memory subsystem architecture and previous programmer e orts to optimize for locality. On a multiprocessor with a large cache con guration and high cache miss penalty, the transformations improved the execution time of programmer-unoptimized applications by as much as 60%. However, on programs where previous programmer e orts to improve data locality had reduced the original amount of false sharing, and on a multiprocessor with a small cache con guration and cache miss penalty, the gains were more modest.