Static analysis of non-interference in expressive low-level languages

3Citations
Citations of this article
8Readers
Mendeley users who have this article in their library.
Get full text

Abstract

Early work in implicit information flow detection applied only to flat, procedureless languages with structured control-flow (e.g., if statements, while loops). These techniques have yet to be adequately extended and generalized to expressive languages with interprocedural, exceptional and irregular control-flow behavior. We present an implicit information flow analysis suitable for languages with conditional jumps, dynamically dispatched methods, and exceptions. We implement this analysis for the Dalvik bytecode format, the substrate for Android. In order to capture information flows across interprocedural and exceptional boundaries, this analysis uses a projection of a small-step abstract interpreter’s rich state graph instead of the control-flow graph typically used for such purposes in weaker linguistic settings. We present a proof of termination-insensitive non-interference. To our knowledge, it is the first analysis capable of proving non-trivial non-interference in a language with this combination of features.

Cite

CITATION STYLE

APA

Aldous, P., & Might, M. (2015). Static analysis of non-interference in expressive low-level languages. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9291, pp. 1–17). Springer Verlag. https://doi.org/10.1007/978-3-662-48288-9_1

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free