The design and analysis of a new GC scheme called MOA is presented with its implementation on PLisp (Portable Lisp). MOA is "stop-and-collect" type GC and is based on a Morris's sliding compaction scheme. MOA has the excellent features such as: (1) it can perform sliding compaction with a time proportional nearly to the size of all data objects in use, (2) it requires an additional space of a small size to achieve such a time cost saving, (3) it can skip a GC process for a special cluster called an "anchor", reducing the total GC processing time considerably. MOA has been successfully implemented on PLisp which provides a large amount of storage space. MOA is superior to other GC based on conventional sliding compaction and copying collection, as shown in several experiments.
CITATION STYLE
Suzuki, M., Koide, H., & Terashima, M. (1995). Moa — a fast sliding compaction scheme for a large storage space. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 986, pp. 197–210). Springer Verlag. https://doi.org/10.1007/3-540-60368-9_25
Mendeley helps you to discover research relevant for your work.