Abstract

Recently, two new parallel algorithms for on-the-fly model checking of LTL properties were presented at the same conference: Automated Technology for Verification and Analysis, 2011. Both approaches extend Swarmed NDFS, which runs several sequential NDFS instances in parallel. While parallel random search already speeds up detection of bugs, the workers must share some global information in order to speed up full verification of correct models. The two algorithms differ considerably in the global information shared between workers, and in the way they synchronize. Here, we provide a thorough experimental comparison between the two algorithms, by measuring the runtime of their implementations on a multi-core machine. Both algorithms were implemented in the same framework of the model checker LTSmin, using similar optimizations, and have been subjected to the full BEEM model database. Because both algorithms have complementary advantages, we constructed an algorithm that combines both ideas. This combination clearly has an improved speedup. We also compare the results with the alternative parallel algorithm for accepting cycle detection OWCTY-MAP. Finally, we study a simple statistical model for input models that do contain accepting cycles. The goal is to distinguish the speedup due to parallel random search from the speedup that can be attributed to clever work sharing schemes.
Original languageUndefined
Title of host publicationProceedings of the 10th International Workshop on Parallel and Distributed Methods in verifiCation, PDMC 2011
EditorsJ. Barnat, K. Heljanko
Place of PublicationUSA
PublisherEPTCS
Pages13-28
Number of pages16
DOIs
StatePublished - 14 Jul 2011
Event10th International Workshop on Parallel and Distributed Methods in verifiCation, PDMC 2011 - Snowbird, United States

Publication series

NameElectronic Proceedings in Theoretical Computer Science
PublisherEPTCS
Volume72
ISSN (Print)2075-2180
ISSN (Electronic)2075-2180

Workshop

Workshop10th International Workshop on Parallel and Distributed Methods in verifiCation, PDMC 2011
Abbreviated titlePDMC
CountryUnited States
CitySnowbird
Period14/07/1114/07/11
Internet address

Fingerprint

Parallel algorithms
Model checking

Keywords

  • IR-78261
  • METIS-278846
  • Parallel
  • LTL model checking
  • EWI-20618
  • Multi-Core
  • depth first search
  • swarmed
  • Nested DFS

Cite this

Laarman, A., & van de Pol, J. C. (2011). Variations on Multi-Core Nested Depth-First Search. In J. Barnat, & K. Heljanko (Eds.), Proceedings of the 10th International Workshop on Parallel and Distributed Methods in verifiCation, PDMC 2011 (pp. 13-28). (Electronic Proceedings in Theoretical Computer Science; Vol. 72). USA: EPTCS. DOI: 10.4204/EPTCS.72.2

Laarman, Alfons; van de Pol, Jan Cornelis / Variations on Multi-Core Nested Depth-First Search.

Proceedings of the 10th International Workshop on Parallel and Distributed Methods in verifiCation, PDMC 2011. ed. / J. Barnat; K. Heljanko. USA : EPTCS, 2011. p. 13-28 (Electronic Proceedings in Theoretical Computer Science; Vol. 72).

Research output: ScientificConference contribution

@inbook{7bbf3fd4d5f74a0fa953579363a8d778,
title = "Variations on Multi-Core Nested Depth-First Search",
abstract = "Recently, two new parallel algorithms for on-the-fly model checking of LTL properties were presented at the same conference: Automated Technology for Verification and Analysis, 2011. Both approaches extend Swarmed NDFS, which runs several sequential NDFS instances in parallel. While parallel random search already speeds up detection of bugs, the workers must share some global information in order to speed up full verification of correct models. The two algorithms differ considerably in the global information shared between workers, and in the way they synchronize. Here, we provide a thorough experimental comparison between the two algorithms, by measuring the runtime of their implementations on a multi-core machine. Both algorithms were implemented in the same framework of the model checker LTSmin, using similar optimizations, and have been subjected to the full BEEM model database. Because both algorithms have complementary advantages, we constructed an algorithm that combines both ideas. This combination clearly has an improved speedup. We also compare the results with the alternative parallel algorithm for accepting cycle detection OWCTY-MAP. Finally, we study a simple statistical model for input models that do contain accepting cycles. The goal is to distinguish the speedup due to parallel random search from the speedup that can be attributed to clever work sharing schemes.",
keywords = "IR-78261, METIS-278846, Parallel, LTL model checking, EWI-20618, Multi-Core, depth first search, swarmed, Nested DFS",
author = "Alfons Laarman and {van de Pol}, {Jan Cornelis}",
year = "2011",
month = "7",
doi = "10.4204/EPTCS.72.2",
series = "Electronic Proceedings in Theoretical Computer Science",
publisher = "EPTCS",
pages = "13--28",
editor = "J. Barnat and K. Heljanko",
booktitle = "Proceedings of the 10th International Workshop on Parallel and Distributed Methods in verifiCation, PDMC 2011",

}

Laarman, A & van de Pol, JC 2011, Variations on Multi-Core Nested Depth-First Search. in J Barnat & K Heljanko (eds), Proceedings of the 10th International Workshop on Parallel and Distributed Methods in verifiCation, PDMC 2011. Electronic Proceedings in Theoretical Computer Science, vol. 72, EPTCS, USA, pp. 13-28, 10th International Workshop on Parallel and Distributed Methods in verifiCation, PDMC 2011, Snowbird, United States, 14-14 July. DOI: 10.4204/EPTCS.72.2

Variations on Multi-Core Nested Depth-First Search. / Laarman, Alfons; van de Pol, Jan Cornelis.

Proceedings of the 10th International Workshop on Parallel and Distributed Methods in verifiCation, PDMC 2011. ed. / J. Barnat; K. Heljanko. USA : EPTCS, 2011. p. 13-28 (Electronic Proceedings in Theoretical Computer Science; Vol. 72).

Research output: ScientificConference contribution

TY - CHAP

T1 - Variations on Multi-Core Nested Depth-First Search

AU - Laarman,Alfons

AU - van de Pol,Jan Cornelis

PY - 2011/7/14

Y1 - 2011/7/14

N2 - Recently, two new parallel algorithms for on-the-fly model checking of LTL properties were presented at the same conference: Automated Technology for Verification and Analysis, 2011. Both approaches extend Swarmed NDFS, which runs several sequential NDFS instances in parallel. While parallel random search already speeds up detection of bugs, the workers must share some global information in order to speed up full verification of correct models. The two algorithms differ considerably in the global information shared between workers, and in the way they synchronize. Here, we provide a thorough experimental comparison between the two algorithms, by measuring the runtime of their implementations on a multi-core machine. Both algorithms were implemented in the same framework of the model checker LTSmin, using similar optimizations, and have been subjected to the full BEEM model database. Because both algorithms have complementary advantages, we constructed an algorithm that combines both ideas. This combination clearly has an improved speedup. We also compare the results with the alternative parallel algorithm for accepting cycle detection OWCTY-MAP. Finally, we study a simple statistical model for input models that do contain accepting cycles. The goal is to distinguish the speedup due to parallel random search from the speedup that can be attributed to clever work sharing schemes.

AB - Recently, two new parallel algorithms for on-the-fly model checking of LTL properties were presented at the same conference: Automated Technology for Verification and Analysis, 2011. Both approaches extend Swarmed NDFS, which runs several sequential NDFS instances in parallel. While parallel random search already speeds up detection of bugs, the workers must share some global information in order to speed up full verification of correct models. The two algorithms differ considerably in the global information shared between workers, and in the way they synchronize. Here, we provide a thorough experimental comparison between the two algorithms, by measuring the runtime of their implementations on a multi-core machine. Both algorithms were implemented in the same framework of the model checker LTSmin, using similar optimizations, and have been subjected to the full BEEM model database. Because both algorithms have complementary advantages, we constructed an algorithm that combines both ideas. This combination clearly has an improved speedup. We also compare the results with the alternative parallel algorithm for accepting cycle detection OWCTY-MAP. Finally, we study a simple statistical model for input models that do contain accepting cycles. The goal is to distinguish the speedup due to parallel random search from the speedup that can be attributed to clever work sharing schemes.

KW - IR-78261

KW - METIS-278846

KW - Parallel

KW - LTL model checking

KW - EWI-20618

KW - Multi-Core

KW - depth first search

KW - swarmed

KW - Nested DFS

U2 - 10.4204/EPTCS.72.2

DO - 10.4204/EPTCS.72.2

M3 - Conference contribution

T3 - Electronic Proceedings in Theoretical Computer Science

SP - 13

EP - 28

BT - Proceedings of the 10th International Workshop on Parallel and Distributed Methods in verifiCation, PDMC 2011

PB - EPTCS

ER -

Laarman A, van de Pol JC. Variations on Multi-Core Nested Depth-First Search. In Barnat J, Heljanko K, editors, Proceedings of the 10th International Workshop on Parallel and Distributed Methods in verifiCation, PDMC 2011. USA: EPTCS. 2011. p. 13-28. (Electronic Proceedings in Theoretical Computer Science). Available from, DOI: 10.4204/EPTCS.72.2