Dynamic memory management (DMM) has been a high cost component in amny software systems. Especial-ly, the use of object-orientation often results in an inten-sive use of dynamic memory, making th edynamic memory performance problem worse. This paper presents a pro-file-based strategy to improve the performance of DMM. The performance improvement comes from a segregated strategy without splitting and coalescing cost. This modifi-cation is made feasible by preallocating the free-list based on the profile data of heap memory usage. In this research, the empirical study shows that the maximum number of alive objects of each size is indepen-dent of the input; this data provides a profile and estimate for the amount of memory the application will need to run and can be preallocated to give great improvement in per-formance. Compare to the average performance of well-known algorithms, the profile-based approach is about 3.9 times to 6.5 times faster.
展开▼
