## Adaptive mutual exclusion with local spinning (2000)

### Cached

### Download Links

- [cs.unc.edu]
- [www.cs.unc.edu]
- [www.cs.unc.edu]
- [www.cs.unc.edu]
- [www.cs.unc.edu]
- [cs.unc.edu]
- DBLP

### Other Repositories/Bibliography

Venue: | In Proceedings of the 14th International Symposium on Distributed Computing |

Citations: | 44 - 11 self |

### BibTeX

@INPROCEEDINGS{Anderson00adaptivemutual,

author = {James H. Anderson and Yong-jik Kim},

title = {Adaptive mutual exclusion with local spinning},

booktitle = {In Proceedings of the 14th International Symposium on Distributed Computing},

year = {2000},

pages = {29--43}

}

### Years of Citing Articles

### OpenURL

### Abstract

Abstract We present an adaptive algorithm for N-process mutual exclusion under read/write atomicity in which all busy waiting is by local spinning. In our algorithm, each process p performs O(k) remote memory references to enter and exit its critical section, where k is the maximum "point contention " experienced by p. The space complexity of our algorithm is \Theta (N), which is clearly optimal. Our algorithm is the first mutual exclusion algorithm under read/write atomicity that is adaptive when time complexity is measured by counting remote memory references. All previous so-called adaptive mutual exclusion algorithms employ busy-waiting loops that can generate an unbounded number of remote memory references. Thus, they have unbounded time complexity under this measure.

### Citations

498 | Algorithms for Scalable Synchronization on Shared-Memory Multiprocessors
- Mellor-Crummey, Scott
- 1991
(Show Context)
Citation Context ...me [5]. Recent work on scalable local-spin mutual exclusion algorithms has shown that the most crucial factor in determining an algorithm's performance is the amount of interconnect trac it generates =-=[4, 8, 10, 14-=-]. In light of this, we dene the time complexity of a mutual exclusion algorithm to be the worstcase number of remote memory references by one process in order to enter and then exit its critical sect... |

296 |
Solution of a problem in concurrent programming control
- Dijkstra
- 1965
(Show Context)
Citation Context ...nt contention" experienced by p. The space complexity of our algorithm is (N ), which is clearly optimal. 1 Introduction In this paper, we consider adaptive solutions to the mutual exclusion prob=-=lem [7]-=- under read/write atomicity. A mutual exclusion algorithm is adaptive if its time complexity is a function of the number of contending processes [6, 11, 13]. Two notions of contention have been consid... |

221 |
The Performance of Spin Lock Alternatives for Shared-Memory Multiprocessors
- Anderson
- 1990
(Show Context)
Citation Context ...me [5]. Recent work on scalable local-spin mutual exclusion algorithms has shown that the most crucial factor in determining an algorithm's performance is the amount of interconnect trac it generates =-=[4, 8, 10, 14-=-]. In light of this, we dene the time complexity of a mutual exclusion algorithm to be the worstcase number of remote memory references by one process in order to enter and then exit its critical sect... |

220 | A fast mutual exclusion algorithm
- Lamport
- 1987
(Show Context)
Citation Context ... algorithm is the splitter element from the grid-based longlived renaming algorithm of Moir and Anderson [12]. This splitter element was actuallysrst used in Lamport's fast mutual exclusion algorithm =-=[9-=-]. The splitter element is dened by the code fragment shown in Fig. 1. (In this and subsequentsgures, we assume that each labeled sequence of statements is atomic; in eachsgure, each labeled sequence ... |

95 |
Synchronization algorithms for shared-memory multipro- cessors
- Graunke, Thakkar
- 1990
(Show Context)
Citation Context ...me [5]. Recent work on scalable local-spin mutual exclusion algorithms has shown that the most crucial factor in determining an algorithm's performance is the amount of interconnect trac it generates =-=[4, 8, 10, 14-=-]. In light of this, we dene the time complexity of a mutual exclusion algorithm to be the worstcase number of remote memory references by one process in order to enter and then exit its critical sect... |

71 | Wait-free algorithms for fast, long-lived renaming
- Moir, Anderson
- 1995
(Show Context)
Citation Context ...algorithm can be seen as an extension of the fast-path algorithm of Anderson and Kim [3]. This algorithm was devised by thinking about connections between fast-path mechanisms and long-lived renaming =-=[12]. Lo-=-ng-lived renaming algorithms are used to \shrink" the size of the name space from which process identiers are taken. The problem is to design operations that processes may inn n-1 n-1 1 stop righ... |

53 |
Adaptive solutions to the mutual exclusion problem
- Choy, Singh
- 1994
(Show Context)
Citation Context ... adaptive solutions to the mutual exclusion problem [7] under read/write atomicity. A mutual exclusion algorithm is adaptive if its time complexity is a function of the number of contending processes =-=[6, 11, 13]. Two not-=-ions of contention have been considered in the literature: \interval contention " and \point contention" [1]. These two notions are dened with respect to a history H . The interval contentio... |

46 |
Long-lived renaming made adaptive
- Afek, Attiya, et al.
- 1999
(Show Context)
Citation Context ...if its time complexity is a function of the number of contending processes [6, 11, 13]. Two notions of contention have been considered in the literature: \interval contention " and \point content=-=ion" [1-=-]. These two notions are dened with respect to a history H . The interval contention over H is the number of processes that are active in H , i.e., that execute outside of their noncritical sections i... |

45 | A fast, scalable mutual exclusion algorithm
- Yang, Anderson
- 1995
(Show Context)
Citation Context |

30 |
Improving Fast Mutual Exclusion
- Styer
- 1992
(Show Context)
Citation Context ... adaptive solutions to the mutual exclusion problem [7] under read/write atomicity. A mutual exclusion algorithm is adaptive if its time complexity is a function of the number of contending processes =-=[6, 11, 13]. Two not-=-ions of contention have been considered in the literature: \interval contention " and \point contention" [1]. These two notions are dened with respect to a history H . The interval contentio... |

28 | Speeding Lamport's Fast Mutual Exclusion Algorithm
- Merritt, Taubenfeld
- 1993
(Show Context)
Citation Context ... adaptive solutions to the mutual exclusion problem [7] under read/write atomicity. A mutual exclusion algorithm is adaptive if its time complexity is a function of the number of contending processes =-=[6, 11, 13]. Two not-=-ions of contention have been considered in the literature: \interval contention " and \point contention" [1]. These two notions are dened with respect to a history H . The interval contentio... |

16 | Fast and scalable mutual exclusion
- Anderson, Kim
- 1999
(Show Context)
Citation Context ... presented a new fast-path mechanism that results in with O(1) time complexity in the absence of contention and (log N) time complexity under contention, when used with Yang and Anderson's algorithm [=-=3]-=-. All of the previously-cited adaptive algorithms are not local-spin algorithms, and thus they have unbounded time complexity under the remote-memoryreferences time measure. One could argue that for a... |

1 |
Adaptive and ecient mutual exclusion. To appear
- Attiya, Bortnikov
(Show Context)
Citation Context ...) response time ? Work supported by NSF grants CCR 9732916 and CCR 9972211. [6]. More recently, Attiya and Bortnikov presented an algorithm with O(k) remote step complexity and O(log k) response time =-=[5]-=-. Recent work on scalable local-spin mutual exclusion algorithms has shown that the most crucial factor in determining an algorithm's performance is the amount of interconnect trac it generates [4, 8,... |