LINS, R. D. A multi-processor shared memory architecture for parallel cyclic reference counting. Microprocessing and Microprogramming 35, 1--5 (Sept. 1992), 563--568. Proceedings of the 18th EUROMICRO Conference (Paris, France).  Home/Search   Document Details and Download   Summary   Related Articles   Check

....nevertheless, it provides a basis for comparing the inherent overhead of the two approaches in terms of overall work performed. 8. RELATED WORK While numerous concurrent, multiprocessor garbage collectors for general purpose programming languages have been described in the literature [12, 14, 16, 18, 19, 21, 22, 24, 29, 30], the number that have been implemented is quite small and of these, only a few actually run on a multiprocessor [2, 12, 18, 16, 17, 27] DeTreville s work on garbage collectors for Modula 2 on the DEC Firefly workstation [12] is the only comparative evaluation of multiprocessor garbage ....

....in a number of important respects: its complexity is linear rather than quadratic; it avoids placing a root in the root buffer more than once per epoch; and it greatly reduces overhead by not considering inherently acyclic structures. Lins has presented a concurrent cycle collection algorithm [24] based on his synchronous algorithm. Unlike the Recycler, Lins does not use a separate reference count for the cycle collector; instead he relies on processor supported asymmetric locking primitives to prevent concurrent mutation to the graph. His scheme has, to our knowledge, never been ....

LINS, R. D. A multi-processor shared memory architecture for parallel cyclic reference counting. Microprocessing and Microprogramming 35, 1--5 (Sept. 1992), 563--568. Proceedings of the 18th EUROMICRO Conference (Paris, France).

.... is described in greater detail and compared quantitatively to a parallel mark and sweep collector by Bacon et al. [3] 7 Related Work on Concurrent Collection While numerous concurrent, multiprocessor collectors for general purpose programming languages have been described in the literature [8, 10, 11, 14, 15, 18, 19, 21, 26, 27], the number that have been implemented is quite small and of these, only a few actually run on a multiprocessor [2, 8, 14, 11, 13, 24] DeTreville s work on garbage collectors for Modula 2 on the DEC Firefly workstation [8] is the only comparative evaluation of multiprocessor garbage collection ....

....multiprocessors their barrier would require synchronization on every access to a mutable object, so it is not clear that the algorithm is practical either for imperative languages or for the current generation of multiprocessor machines. Lins has presented a concurrent cycle collection algorithm [21] based on his synchronous algorithm. Unlike the Recycler, Lins does not use a separate reference count for the cycle collector; instead he relies on processor supported asymmetric locking primitives to prevent concurrent mutation to the graph. His scheme has, to our knowledge, never been ....

LINS, R. D. A multi-processor shared memory architecture for parallel cyclic reference counting. Microprocessing and Microprogramming 35, 1--5 (Sept. 1992), 563--568. Proceedings of the 18th EUROMICRO Conference (Paris, France).

....environment; nevertheless, it provides a basis for comparing the inherent overhead of the two approaches in terms of overall work performed. 8. RELATED WORK While numerous concurrent, multiprocessor collectors for general purpose programming languages have been described in the literature [3, 6, 8, 10, 12, 13, 15, 16, 18, 21, 22], the number that have been implemented is quite small and of these, only a few actually run on a multiprocessor [2, 6, 12, 10, 11, 20] DeTreville s work on garbage collectors for Modula 2 on the DEC Fire y workstation [6] is the only comparative evaluation of multiprocessor garbage collection ....

....in a number of important respects: its complexity is linear rather than quadratic; it avoids placing a root in the root bu er more than once per epoch; and it greatly reduces overhead by not considering inherently acyclic structures. Lins has presented a concurrent cycle collection algorithm [18] based on his synchronous algorithm. Unlike the Recycler, Lins does not use a separate reference count for the cycle collector; instead he relies on processor supported asymmetric locking primitives to prevent concurrent mutation to the graph. His scheme has, to our knowledge, never been ....

Lins, R. D. A multi-processor shared memory architecture for parallel cyclic reference counting. Microprocessing and Microprogramming 35, 1-5 (Sept. 1992), 563-568. Proceedings of the 18th EUROMICRO Conference (Paris, France).

....is described in greater detail and compared quantitatively to a parallel mark and sweep collector by Bacon et al. [3] 6. 3 Related Work on Concurrent Collection While numerous concurrent, multiprocessor collectors for general purpose programming languages have been described in the literature [6, 8, 10, 12, 13, 15, 16, 18, 22, 23], the number that have been implemented is quite small and of these, only a few actually run on a multiprocessor [2, 6, 12, 10, 11, 21] DeTreville s work on garbage collectors for Modula 2 on the DEC Fire y workstation [6] is the only comparative evaluation of multiprocessor garbage collection ....

....multiprocessors their barrier would require synchronization on every access to a mutable object, so it is not clear that the algorithm is practical either for imperative languages or for the current generation of multiprocessor machines. Lins has presented a concurrent cycle collection algorithm [18] based on his synchronous algorithm. Unlike the Recycler, Lins does not use a separate reference count for the cycle collector; instead he relies on processor supported asymmetric locking primitives to prevent concurrent mutation to the graph. His scheme has, to our knowledge, never been ....

Lins, R. D. A multi-processor shared memory architecture for parallel cyclic reference counting. Microprocessing and Microprogramming 35, 1-5 (Sept. 1992), 563-568. Proceedings of the 18th EUROMICRO Conference (Paris, France).

....would be allowed to call markred before the node currently garbage collecting had completed collectblue. This could easily be implemented by imposing an order on nodes in the network and using a token passing mechanism. A less restrictive method is to synchronise the phases of garbage collection [20]. In the markred phase, any processor that wishes to do so is allowed to initiate marking; no processor is allowed to start either scanning or collecting until there are no processors marking red. Notice that only a subset of processing elements need be involved in this garbage collection at any ....

Rafael D. Lins. A multi-processor shared memory architecture for parallel cyclic reference counting. Microprocessing and Microprogramming, 38:31--35, September 1992.

....to the mark scan. With minimal overhead one can also check for cycles during marking. This allows a more efficient scan phase, saving one pass through the subgraph under analysis. The algorithm presented can be easy and advantageously incorporated to the shared memory architectures described in [4, 5]. Acknowledgements Research reported herein has been sponsored jointly by the British Council, CAPES (Brazil) grant CBE 4875 91, and C.N.Pq. Brazil) grants No 40.9110 88.4. and 46.0782 89.4. ....

R.D.Lins. A multi-processor shared memory architecture for parallel cyclic reference counting. Microprocessing and Microprogramming, 35:563--568, NorthHolland, August 1992.

....free list is empty or the queue is full is the local mark scan required. Our performance figures show that lazy mark scan is far more efficient than local mark scan. The algorithm presented here is the kernel of the shared memory architectures for parallel cyclic reference counting described in [2, 3]. The Local Mark Scan Algorithm The algorithm presented in [5] performs a local mark scan whenever a pointer to a shared structure is deleted. It works in three phases. In the first phase, the graph below the deleted pointer is traversed, counts due to internal references are decremented and ....

R.D.Lins. A multi-processor shared memory architecture for parallel cyclic reference counting. Microprocessing and Microprogramming, 35:563-568, North-Holland, August 1992.

No context found.

LINS, R. D. A multi-processor shared memory architecture for parallel cyclic reference counting. Microprocessing and Microprogramming 35, 1--5 (Sept. 1992), 563--568. Proceedings of the 18th EUROMICRO Conference (Paris, France).

No context found.

Rafael D. Lins. A multi-processor shared memory architecture for parallel cyclic reference counting. Microprocessing and Microprogramming, 35:563--568, September 1992.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC