• DocumentCode
    3008444
  • Title

    Analysis and Transformations for Efficient Query-Based Debugging

  • Author

    Gorbovitski, Michael ; Tekle, K. Tuncay ; Rothamel, Tom ; Stoller, Scott ; Liu, Yanhong A.

  • Author_Institution
    Comput. Sci. Dept., State Univ. of New York at Stony Brook, Stony Brook, NY
  • fYear
    2008
  • fDate
    28-29 Sept. 2008
  • Firstpage
    174
  • Lastpage
    183
  • Abstract
    This paper describes a framework that supports powerful queries in debugging tools, and describes in particular the transformations, alias analysis, and type analysis used to make the queries efficient. The framework allows queries over the states of all objects at any point in the execution as well as over the history of states. The transformations are based on incrementally maintaining the results of expensive queries studied in previous work. The alias analysis extends the flow-sensitive intraprocedural analysis to an efficient flow-sensitive interprocedural analysis for an object-oriented language with also a form of context sensitivity. We also show the power of the framework and the effectiveness of the analyses through case studies and experiments with XML DOM tree transformations, an FTP client, and others. We were able to easily determine the sources of all injected bugs, and we also found an actual bug in the case study on the FTP client.
  • Keywords
    object-oriented languages; program debugging; program diagnostics; query languages; alias analysis; flow-sensitive interprocedural analysis; flow-sensitive intraprocedural analysis; object-oriented language; query-based debugging; type analysis; Algorithm design and analysis; Computer bugs; Computer science; Debugging; History; Iterative algorithms; Performance analysis; Watches; Writing; XML; Alias Analysis; Incremental Computation; Invariant Verification; Query-based Debugging; Type Analysis;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Source Code Analysis and Manipulation, 2008 Eighth IEEE International Working Conference on
  • Conference_Location
    Beijing
  • Print_ISBN
    978-0-7695-3353-7
  • Type

    conf

  • DOI
    10.1109/SCAM.2008.27
  • Filename
    4637550