Ingénierie Du Logiciel

  • Full Screen
  • Wide Screen
  • Narrow Screen
  • Increase font size
  • Default font size
  • Decrease font size

Séminaire MOCS 22/04/2014

Imprimer PDF

Intervenant : Mickael Delahaye, Chercheur Postdoc, CEA.
Date & Lieu : Mardi 22 Avril 2014, 10h-11h, Salle 117-B, 1er étage, Bâtiment LC, site AES, UBO.
Titre :  Combining analysis to improve software reliability: the case of infeasible path generalization. 

Résumé : 

In this talk, I am giving an overview of how different kinds of analyses may learn from one another and be put to use together in order to improve software quality. This talk includes an in-depth presentation of infeasible path generalization for symbolic execution, where static analysis and constraint analysis is used to speed up test generation.

Infeasible paths in the control flow graph are a hinder to path-oriented white-box automated test generation (PathCrawler, Pex, etc.). This method proposes to take opportunity of a single infeasible path to find a whole family of infeasible paths so as to prune the test generation's search space.  This talk ends with an outline of other ways to combine different analyses both through examples of work already done (e.g. executable specification language, etc.) and perspectives.

You are here: