Automated Discovery of Race Condition Failures ...

Get Adobe Flash player

facebook  digg  delicious  newsvine  reddit  simpy  spurl  yahoo
Favorite  Add to Favorites     Feature  Feature This!     Inappropriate  Inappropriate     Share  Share     playlist  Add to Playlist
  • Currently 0.00/5

Rating: 0.0/5 (0 vote cast)

Type of abuse
Multi-threaded applications are commonplace in today’s software landscape. Pushing the boundaries of concurrency and parallelism, programmers are maximizing performance demanded by stakeholders. However, multi-threaded programs are challenging to test and debug. Prone to their own set of unique faults, such as race conditions, testers are turning to automated validation tools for assistance. This paper’s main contribution is a new algorithm called multi-stage novelty filtering (MSNF) that can aid in the discovery of race failures as well as handle anomaly detection in general. MSNF stresses tool simplicity requiring minimal configuration, no data preprocessing or software metrics, and the ability to adapt to its environment. The paper begins with the importance of testing tool simplicity. It then gives an overview of the support vector machine, a machine learning technique chosen to achieve adaptability. The paper then formulates the MSNF algorithm, empirical results are discussed, and challenges with the proposed solution investigated. Future research offers direction for taking MSNF from theory to practice. Finally, MSNF is compared to related works. The conclusion is preliminary results show promise, but MSNF requires more research before it can be implemented in a useful testing tool.
Added on Nov 24, 2009 by jcuzzola
Video Details
Video Responses (0)

Be the first to post a video response!

User Details
Share Details

Post Comments
Comment on this video:

Comments: (0)

More from: jcuzzola