A lightweight cyclic reference counting algorithm

Chin Yang Lin, Ting Wei Hou

Research output: Chapter in Book/Report/Conference proceedingConference contribution

7 Citations (Scopus)


This paper focuses on a major weakness of reference counting technique - the lack of collecting cyclic garbage. Most reference counted systems handle this problem by either invoking a global mark-sweep collector occasionally, or incorporating a local ("partial") tracing collector that considers only the cycle candidates (objects) but needs several traces on them. This paper proposes a "lightweight" cycle detector, which is based on the partial tracing approach but collects garbage cycles in a simpler and more efficient way. Key to the algorithm is the removal of multiple traces on the cycle candidates - It effectively reclaims garbage cycles in only one trace. We have evaluated the algorithm in the Jikes Research Virtual Machine, where a set of benchmark programs from SPECjvm98 were applied. The experiments demonstrate the efficiency and practicability of the lightweight cycle detector, compared to a modern cycle detector that requires multiple traces on objects.

Original languageEnglish
Title of host publicationAdvances in Grid and Pervasive Computing - First International Conference, GPC 2006, Proceedings
PublisherSpringer Verlag
Number of pages14
ISBN (Print)3540338098, 9783540338093
Publication statusPublished - 2006
Event1st International Conference on Grid and Pervasive Computing,GPC 2006 - Taichung, Taiwan
Duration: 2006 May 32006 May 5

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume3947 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Other1st International Conference on Grid and Pervasive Computing,GPC 2006

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'A lightweight cyclic reference counting algorithm'. Together they form a unique fingerprint.

Cite this