Are Lock-Free Concurrent Algorithms Practically Wait-Free?

Alistarh, Dan; Censor-Hillel, Keren; Shavit, Nir

doi:10.1145/2903136

Are Lock-Free Concurrent Algorithms Practically Wait-Free?

Journal Article · Wed Sep 14 00:00:00 EDT 2016 · Journal of the Association for Computing Machinery

DOI:https://doi.org/10.1145/2903136· OSTI ID:1557574

Alistarh, Dan ^[1]; Censor-Hillel, Keren ^[2]; Shavit, Nir ^[3]

ETH Zurich, Zürich (Switzerland); Office of Scientific and Technical Information (OSTI)
Technion, Haifa (Israel)
Massachusetts Inst. of Technology (MIT), Cambridge, MA (United States); Tel-Aviv Univ., Tel Aviv (Israel)

Lock-free concurrent algorithms guarantee that some concurrent operation will always make progress in a finite number of steps. Yet programmers prefer to treat concurrent code as if it were wait-free, guaranteeing that all operations always make progress. Unfortunately, designing wait-free algorithms is generally a very complex task, and the resulting algorithms are not always efficient. Although obtaining efficient wait-free algorithms has been a long-time goal for the theory community, most nonblocking commercial code is only lock-free. This article suggests a simple solution to this problem. We show that for a large class of lock-free algorithms, under scheduling conditions that approximate those found in commercial hardware architectures, lock-free algorithms behave as if they are wait-free. In other words, programmers can continue to design simple lock-free algorithms instead of complex wait-free ones, and in practice, they will get wait-free progress. Our main contribution is a new way of analyzing a general class of lock-free algorithms under a stochastic scheduler. Our analysis relates the individual performance of processes to the global performance of the system using Markov chain lifting between a complex per-process chain and a simpler system progress chain. We show that lock-free algorithms are not only wait-free with probability 1 but that in fact a general subset of lock-free algorithms can be closely bounded in terms of the average number of steps required until an operation completes. Here, to the best of our knowledge, this is the first attempt to analyze progress conditions, typically stated in relation to a worst-case adversary, in a stochastic model capturing their expected asymptotic behavior.

View Accepted Manuscript (DOE)

Research Organization:: Massachusetts Inst. of Technology (MIT), Cambridge, MA (United States)

Sponsoring Organization:: USDOE Office of Science (SC), Advanced Scientific Computing Research (ASCR) (SC-21)

Grant/Contract Number:: SC0008923

OSTI ID:: 1557574

Journal Information:: Journal of the Association for Computing Machinery, Journal Name: Journal of the Association for Computing Machinery Journal Issue: 4 Vol. 63; ISSN 0004-5411

Publisher:: ACMCopyright Statement

Country of Publication:: United States

Language:: English

References (17)

A new solution of Dijkstra's concurrent programming problem Lamport, Leslie Communications of the ACM, Vol. 17, Issue 8 https://doi.org/10.1145/361082.361093	journal	August 1974
Practical lock-freedom Fraser, Keir Computer Laboratory, University of Cambridge https://doi.org/10.48456/tr-579	report	January 2004
Obstruction-Free Step Complexity: Lock-Free DCAS as an Example Fich, Faith Ellen; Luchangco, Victor; Moir, Mark Distributed Computing. Lecture Notes in Computer Science https://doi.org/10.1007/11561927_37	book	January 2005
Obstruction-Free Algorithms Can Be Practically Wait-Free Fich, Faith Ellen; Luchangco, Victor; Moir, Mark Distributed Computing. Lecture Notes in Computer Science https://doi.org/10.1007/11561927_8	book	January 2005
Liftings of Tree-Structured Markov Chains Hayes, Thomas P.; Sinclair, Alistair Approximation, Randomization, and Combinatorial Optimization. Algorithms and Techniques https://doi.org/10.1007/978-3-642-15369-3_45	book	January 2010
Analyzing the Performance of Lock-Free Data Structures: A Conflict-Based Model Atalar, Aras; Renaud-Goud, Paul; Tsigas, Philippas Distributed Computing. Lecture Notes in Computer Science https://doi.org/10.1007/978-3-662-48653-5_23	book	January 2015
Fast deterministic consensus in a noisy environment Aspnes, J. Journal of Algorithms, Vol. 45, Issue 1 https://doi.org/10.1016/S0196-6774(02)00220-1	journal	October 2002
Markov Chains and Mixing Times Levin, David A.; Peres, Yuval; Wilmer, Elizabeth L. https://doi.org/10.1090/mbk/058	book	January 2009
Wait-free synchronization Herlihy, Maurice ACM Transactions on Programming Languages and Systems, Vol. 13, Issue 1 https://doi.org/10.1145/114005.102808	journal	January 1991
Simple, fast, and practical non-blocking and blocking concurrent queue algorithms Michael, Maged M.; Scott, Michael L. Proceedings of the fifteenth annual ACM symposium on Principles of distributed computing - PODC '96 https://doi.org/10.1145/248052.248106	conference	January 1996
Nonblocking algorithms and scalable multicore programming Al Bahra, Samy Communications of the ACM, Vol. 56, Issue 7 https://doi.org/10.1145/2483852.2483866	journal	July 2013
Scalable statistics counters Dice, Dave; Lev, Yossi; Moir, Mark Proceedings of the 25th ACM symposium on Parallelism in algorithms and architectures - SPAA '13 https://doi.org/10.1145/2486159.2486182	conference	January 2013
Are lock-free concurrent algorithms practically wait-free? Alistarh, Dan; Censor-Hillel, Keren; Shavit, Nir Proceedings of the 46th Annual ACM Symposium on Theory of Computing - STOC '14 https://doi.org/10.1145/2591796.2591836	conference	January 2014
Lock-Free Algorithms under Stochastic Schedulers Alistarh, Dan; Sauerwald, Thomas; Vojnović, Milan Proceedings of the 2015 ACM Symposium on Principles of Distributed Computing - PODC '15 https://doi.org/10.1145/2767386.2767430	conference	January 2015
Lease/release: architectural support for scaling contended data structures Haider, Syed Kamran; Hasenplaugh, William; Alistarh, Dan Proceedings of the 21st ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming - PPoPP '16 https://doi.org/10.1145/2851141.2851155	conference	January 2016
Lifting Markov chains to speed up mixing Chen, Fang; Lovász, László; Pak, Igor Proceedings of the thirty-first annual ACM symposium on Theory of computing - STOC '99 https://doi.org/10.1145/301250.301315	conference	January 1999
The read-copy-update mechanism for supporting real-time applications on shared-memory multiprocessor systems with Linux Guniguntala, D.; McKenney, P. E.; Triplett, J. IBM Systems Journal, Vol. 47, Issue 2 https://doi.org/10.1147/sj.472.0221	journal	January 2008

Cited By (4)

How Lock-free Data Structures Perform in Dynamic Environments: Models and Analyses Atalar, Aras; Renaud-Goud, Paul; Tsigas, Philippas arXiv https://doi.org/10.48550/arxiv.1611.05793	preprint	January 2016
D2.3 Power models, energy models and libraries for energy-efficient concurrent data structures and algorithms Ha, Phuong Hoai; Tran, Vi Ngoc-Nha; Umar, Ibrahim arXiv https://doi.org/10.48550/arxiv.1801.10556	preprint	January 2018
D2.4 Report on the final prototype of programming abstractions for energy-efficient inter-process communication Ha, Phuong Hoai; Tran, Vi Ngoc-Nha; Umar, Ibrahim arXiv https://doi.org/10.48550/arxiv.1802.03013	preprint	January 2018
Recoverable Consensus in Shared Memory Golab, Wojciech arXiv https://doi.org/10.48550/arxiv.1804.10597	preprint	January 2018

Similar Records

Nonblocking algorithm for shared queues using compare-and-swap

Journal Article · Sun May 01 00:00:00 EDT 1994 · IEEE Transactions on Computers (Institute of Electrical and Electronics Engineers); (United States) · OSTI ID:6697614

Effective Software Design and Development for the New Graph Architecture HPC Machines

Technical Report · Fri Mar 16 20:00:00 EDT 2012 · OSTI ID:1039407

Highly concurrent scalar processing. Doctoral thesis

Technical Report · Tue Dec 31 23:00:00 EST 1985 · OSTI ID:5975786

Related Subjects

97 MATHEMATICS AND COMPUTING
distributed computing
lock-free algorithms
progress properties
schedulers
shared memory
wait-free algorithms

Are Lock-Free Concurrent Algorithms Practically Wait-Free?

Citation Formats

References (17)

Cited By (4)

Similar Records

Related Subjects