Automatic Optimization of Software Transactional Memory Through Linear Regression and Decision Tree

  • Yang Xiao
  • Zhen Li
  • Ehsan AtoofianEmail author
  • Ali Jannesari
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 9531)


Software Transactional Memory (STM) is a promising paradigm that facilitates programming for shared memory multiprocessors. In STM, synchronization of accesses to the shared memory locations is fully handled by STM library and does not require any intervention by programmers. While STM eases parallel programming, it results in run-time overhead which increases execution time of certain applications. In this paper, we focus on overhead of STM and propose optimization techniques to enhance speed of STM applications. In particular, we focus on size of transaction, read-set, and write-set and show that execution time of applications significantly changes by varying these parameters. Optimizing these parameters manually is a time consuming process and requires significant labor work. We exploit Linear Regression (LR) and propose an optimization technique that decides on these parameters automatically. We further enhance this technique by using decision tree. The decision tree improves accuracy of predictions by selecting appropriate LR model for a given transaction. We evaluate our optimization techniques using a set of benchmarks from NAS and DiscoPoP benchmark suites. Our experimental results reveal that LR and decision tree together are able to improve performance of STM programs up to 54.8 %.


Software Transactional Memory Linear Regression Decision tree Performance 


  1. 1.
    Herlihy, M., Moss, J.E.B.: Transactional memory: Architectural support for lock-free data structures. In: Proceedings of the Twentieth Annual International Symposium on Computer Architecture (ISCA), pp. 289–300, May 1993Google Scholar
  2. 2.
    Dice, D., Shalev, O., Shavit, N.N.: Transactional locking II. In: Dolev, S. (ed.) DISC 2006. LNCS, vol. 4167, pp. 194–208. Springer, Heidelberg (2006)CrossRefGoogle Scholar
  3. 3.
    Abadi, M., Harris, T., Mehrara, M.: Transactional memory with strong atomicity using off-the-shelf memory protection hardware. In: Proceedings of the 14th ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, Raleigh, NC, February 2009Google Scholar
  4. 4.
    Felber, P., Fetzer, C., Riegel, T.: Dynamic performance tuning of word-based software transactional memory. In: Proceedings of the 13th ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, Salt Lake City, UT, February 2008Google Scholar
  5. 5.
    Marathe, V.J., Scherer III, W.N., Scott, M.L.: Adaptive software transactional memory. In: Fraigniaud, P. (ed.) DISC 2005. LNCS, vol. 3724, pp. 354–368. Springer, Heidelberg (2005)CrossRefGoogle Scholar
  6. 6.
    Dice, D., Lev, Y., Moir, M., Nussbaum, D.: Early experience with a commercial hardware transactional memory implementation. In: Proceedings of the 14th International Conference on Architectural Support for Programming Languages and Operating Systems, Washington, DC, March 2009Google Scholar
  7. 7.
    Shavit, N., Touitou, D.: Software transactional memory. In: Proceedings of the Fourteenth Annual Symposium on Principles of Distributed Computing (PODC), pp. 204–213, August 1995Google Scholar
  8. 8.
    Goldberger, S.: Best linear unbiased prediction in the generalized linear regression model. J. Am. Stat. Assoc. 57(298), 369–375 (1962)MathSciNetCrossRefzbMATHGoogle Scholar
  9. 9.
    Bailey, D., Barszcz, E., Barton, J., Browning, D., Carter, R., Dagum, L., Fatoohi, R., Fineberg, S., Frederickson, P., Lasinski, T., Schreiber, R., Simon, H., Venkatakrishnan, V., Weeratunga, S.: The NAS parallel Benchmarks. RNR Technical Report RNR-94-007, March 1994Google Scholar
  10. 10.
    Li, Z., Jannesari, A., Wolf, F.: Discovery of potential parallelism in sequential programs. In: Proceedings of the 42nd International Conference on Parallel Processing Workshops (ICPPW), Workshop on Parallel Software Tools and Tool Infrastructures (PSTI), Lyon, France, pp. 1004–1013, October 2013Google Scholar
  11. 11.
    Didona, D., Felber, P., Harmanci, D., Romano, P., Schenker, J.: Identifying the optimal level of parallelism in transactional memory applications. In: Gramoli, V., Guerraoui, R. (eds.) NETYS 2013. LNCS, vol. 7853, pp. 233–247. Springer, Heidelberg (2013)CrossRefGoogle Scholar
  12. 12.
    Castro, M., Góes, L.F.W., Fernandes, L.G., Méhaut, J.-F.: Dynamic thread mapping based on machine learning for transactional memory applications. In: Kaklamanis, C., Papatheodorou, T., Spirakis, P.G. (eds.) Euro-Par 2012. LNCS, vol. 7484, pp. 465–476. Springer, Heidelberg (2012)CrossRefGoogle Scholar
  13. 13.
    Google Inc. Quantifying Movie Magic with GoogleSearch. June 2013Google Scholar
  14. 14.
    Wang, C., Chen, W.Y., Wu, Y., et al.: Code generation and optimization for transactional memory constructs in an unmanaged language. In: The Proceedings of the International Symposium on Code Generation and Optimization, pp. 34–48. IEEE Computer Society (2007)Google Scholar
  15. 15.
    Quinlan, J.R.: C4.5: Programs for Machine Learning. Morgan Kaufmann Publishers, San Francisco (1993)Google Scholar
  16. 16.
    SPSS Inc. 2007. SPSS 16.0 Command Syntax Reference. Chicago, IL: SPSS IncGoogle Scholar
  17. 17.
    Quinlan, J.R.: Induction of decision tree. Mach. Learn. 1(1), 81–106 (1986)Google Scholar
  18. 18.
    Asanovic, K., Bodik, R., Catanzaro, B.C., Gebis, J.J., Husbands, P., Keutzer, K., Yelick, K.A.: The landscape of parallel computing research: A view from berkeley. Technical Report UCB/EECS-2006-183, EECS Department, University of California, Berkeley (2006)Google Scholar

Copyright information

© Springer International Publishing Switzerland 2015

Authors and Affiliations

  • Yang Xiao
    • 1
  • Zhen Li
    • 2
  • Ehsan Atoofian
    • 1
    Email author
  • Ali Jannesari
    • 2
  1. 1.Electrical Engineering DepartmentLakehead UniversityThunder BayCanada
  2. 2.Technical University of DarmstadtDarmstadtGermany

Personalised recommendations