GameTime: Toolkit for the Timing Analysis of Software
Sanjit Seshia, Jonathan Kotker

Citation
Sanjit Seshia, Jonathan Kotker. "GameTime: Toolkit for the Timing Analysis of Software". UC Berkeley, 20, August, 2014.

Abstract
GameTime is an open-source toolkit for the timing analysis of software. It is targeted towards embedded software, particularly for sequential, terminating programs. The core algorithm of GameTime is based on a combination of game-theoretic online learning of a platform model, and systematic testing on the target platform, leveraging satisfiability modulo theories (SMT) solvers. In contrast with many existing tools for timing analysis, GameTime can be used for a range of tasks that involve the prediction of an execution time property, including: Estimating the worst-case execution time (WCET), Estimating the distribution of execution times, Detecting timing-related bugs by generating test cases that violate program deadlines, and Predicting execution times of particular program path as part of software-in-the-loop simulations. GameTime reduces the potentially exponentially large path space of a program to a polynomial number of basis paths. This reduction enables the tool to perform only a polynomial number of measurements on the target platform, which is very useful in practice as accurate measurements can be tedious and expensive. The use of machine learning to generate a platform model enables GameTime to be easily ported to any platform on which accurate path measurements can be made.

Electronic downloads


Internal. This publication has been marked by the author for TerraSwarm-only distribution, so electronic downloads are not available without logging in.
Citation formats  
  • HTML
    Sanjit Seshia, Jonathan Kotker. <a
    href="http://www.terraswarm.org/pubs/498.html"
    ><i>GameTime: Toolkit for the Timing Analysis of
    Software</i></a>, UC Berkeley, 20, August, 2014.
  • Plain text
    Sanjit Seshia, Jonathan Kotker. "GameTime: Toolkit for
    the Timing Analysis of Software". UC Berkeley, 20,
    August, 2014.
  • BibTeX
    @software{SeshiaKotker14_GameTimeToolkitForTimingAnalysisOfSoftware,
        author = {Sanjit Seshia and Jonathan Kotker},
        title = {GameTime: Toolkit for the Timing Analysis of
                  Software},
        institution = {UC Berkeley},
        day = {20},
        month = {August},
        year = {2014},
        abstract = {GameTime is an open-source toolkit for the timing
                  analysis of software. It is targeted towards
                  embedded software, particularly for sequential,
                  terminating programs. The core algorithm of
                  GameTime is based on a combination of
                  game-theoretic online learning of a platform
                  model, and systematic testing on the target
                  platform, leveraging satisfiability modulo
                  theories (SMT) solvers. In contrast with many
                  existing tools for timing analysis, GameTime can
                  be used for a range of tasks that involve the
                  prediction of an execution time property,
                  including: Estimating the worst-case execution
                  time (WCET), Estimating the distribution of
                  execution times, Detecting timing-related bugs by
                  generating test cases that violate program
                  deadlines, and Predicting execution times of
                  particular program path as part of
                  software-in-the-loop simulations. GameTime reduces
                  the potentially exponentially large path space of
                  a program to a polynomial number of basis paths.
                  This reduction enables the tool to perform only a
                  polynomial number of measurements on the target
                  platform, which is very useful in practice as
                  accurate measurements can be tedious and
                  expensive. The use of machine learning to generate
                  a platform model enables GameTime to be easily
                  ported to any platform on which accurate path
                  measurements can be made.},
        URL = {http://terraswarm.org/pubs/498.html}
    }
    

Posted by Barb Hoversten on 10 Feb 2015.
Groups: tools

Notice: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright.