On multi-threaded list-processing and garbage collection

Kuechlin, Wolfgang; Nevin, Nicholas

doi:10.1109/spdp.1991.218226

Cited by 6 publications

(6 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The original S-threads memory management scheme distributes the SACLIB heap to threads as paged segments of cells, and it uses preventive garbage collection (Küchlin and Nevin, 1991). If a side-effect free functional programming style is used, then the result of a function can be copied into its parent's heap segment, and its own heap segment, containing all garbage, can be recycled in bulk.…”

Section: Parallel Symbolic Computation With Virtual S-threadsmentioning

confidence: 99%

Parallel Term Rewriting with Paredux

Bündgen¹,

Göbel²,

Küchlin³

et al. 1998

Applied Logic Series

View full text Add to dashboard Cite

Section: Parallel Symbolic Computation With Virtual S-threadsmentioning

confidence: 99%

Parallel Term Rewriting with Paredux

Bündgen¹,

Göbel²,

Küchlin³

et al. 1998

Applied Logic Series

View full text Add to dashboard Cite

“…The original S-threads memory management scheme distributes the SAC-2 heap to threads as paged segments of cells, and it uses preventive garbage collection (Küchlin and Nevin, 1991). If a side-effect free functional programming style is used, then the result of a function can be copied into its parent's heap segment, and its own heap segment, containing all garbage, can be recycled in bulk.…”

Section: Parallel Computation With Virtual S-threadsmentioning

confidence: 99%

Strategy Compliant Multi-Threaded Term Completion

Bündgen

Göbel

Küchlin

1996

Journal of Symbolic Computation

View full text Add to dashboard Cite

We report on the design, implementation, and performance, of the parallel term-rewriting system PaReDuX. We discuss the parallelization of three term completion procedures: Knuth-Bendix completion, completion modulo AC, and unfailing completion. Our parallelization is strategy-compliant, i.e., the parallel code performs exactly the same work as the sequential code, but the work load is shared by many processors. PaReDuX is designed for shared memory parallel architectures, such as multi-processor workstations, where it shows good performance on a variety of examples.

show abstract

“…) and language independent implementations (Ferreira, 1991). Mark-and-Sweep is also amenable to various forms of concurrency (Dijkstra et al, 1978;Baker, 1978;Washabaugh & Kafura, 1990;Kuechlin & Nevin, 1991;Boehm et al, 1991;Tanaka et al, 1994). Most implementations use a timestamp to mark chunks mutated during the relatively long marking phase, allowing both the mutator -and-Compact (A and C) and the collector free access to the heap.…”

Section: Mark-and-sweepmentioning

confidence: 99%

“…Preventive garbage collection, is the technique of performing garbage collection at a point in time at which the application is using minimal memory, in terms of both stack and heap (Kuechlin & Nevin, 1991). Because memory in use is at a minimum, the time taken to perform the garbage collection is minimised.…”

Section: Preventive Garbage Collectionmentioning

confidence: 99%

“…For example a text editor after closing a le, or a scheme interpreter after the completion of a program. A garbage collection at these points could be expected to reclaim signi cantly more memory and/or be completed in signi cantly less time than a garbage collection performed at a random time (Kuechlin & Nevin, 1991). Preventive garbage collection is most useful when used with a non-incremental algorithm, as this allows the entire garbage collection cycle to enjoy the optimal conditions.…”

Section: Preventive Garbage Collectionmentioning

confidence: 99%

See 1 more Smart Citation

Design patterns in garbage collection

Yeates

Champlain²

Proceedings. Technology of Object-Oriented Languages and Systems, TOOLS 25 (Cat. No.97TB100239)

View full text Add to dashboard Cite

This thesis presents an examination of design patterns within the context of garbage collection. Initially, I review garbage collection and design patterns. Four garbage collectors are then examined and the design patterns found described. Both domain speci c and generic patterns are described. The domain speci c patterns are TriColour and RootSet, the generic patterns are Adaptor, Facade, Iterator and Proxy.It is hoped that by, applying these patterns, systems designers have access to a less e cient, but simpler and more exible way of implementing and reusing garbage collectors in programming languages.The requirements analysis for a garbage collector for a real-time object-oriented microkernel is then performed, and a design prepared using the design patterns found in the other garbage collectors. The garbage collector is then implemented in Java using appropriate data structures. Due to timing di culties in the runtime environment, timing was ruled out as a method of performance analysis. Algorithmic analysis is performed to evaluate the worst-case performance of the collector, which is found to be satisfactory in all but one method of the RootSet implementation. An approach to remedying this is suggested.

show abstract

On multi-threaded list-processing and garbage collection

Cited by 6 publications

References 12 publications

Parallel Term Rewriting with Paredux

Parallel Term Rewriting with Paredux

Strategy Compliant Multi-Threaded Term Completion

Design patterns in garbage collection

Contact Info

Product

Resources

About