A Parallel Task Orinted Memory Manager for Dynamic Objects


The Transactions of the Korea Information Processing Society (1994 ~ 2000), Vol. 4, No. 5, pp. 1391-1400, May. 1997
10.3745/KIPSTE.1997.4.5.1391,   PDF Download:

Abstract

When a language to produce many dynamic objects is executed on a shared memory multiprocessor, a memory management algorithm for dynamic objects has much influence on program performance. This paper presents a new memory management algorithm that enhances program performance in this environment. We presents taskwise allocation and collection policy for heap spaces. And dynamic objects are categorized by shared objects and non-shared objects among the parallel tasks. The heap space is partitioned by a shared area and a private area. This strategy makes dynamic scheduling for parallel tasks easy, a reference locality is enhanced, and frequency of collecting the garbages is reduced because of memory reuse for a private area.


Statistics
Show / Hide Statistics

Statistics (Cumulative Counts from September 1st, 2017)
Multiple requests among the same browser session are counted as one view.
If you mouse over a chart, the values of data points will be shown.


Cite this article
[IEEE Style]
K. E. Jung and B. J. Min, "A Parallel Task Orinted Memory Manager for Dynamic Objects," The Transactions of the Korea Information Processing Society (1994 ~ 2000), vol. 4, no. 5, pp. 1391-1400, 1997. DOI: 10.3745/KIPSTE.1997.4.5.1391.

[ACM Style]
Kim Eun Jung and Bae Jong Min. 1997. A Parallel Task Orinted Memory Manager for Dynamic Objects. The Transactions of the Korea Information Processing Society (1994 ~ 2000), 4, 5, (1997), 1391-1400. DOI: 10.3745/KIPSTE.1997.4.5.1391.