Static Information Flow Control Research Articles

A Lightweight Dynamic Enforcement of Privacy Protection for Android

Inter-process communication (IPC) provides a message passing mechanism for information exchange between applications. It has been long believed that IPCs can be abused by malware writers to launch collusive information leak using two or more applications. Much work on privacy protection focuses on the simple information leak caused by the individual applications and lacks effective approaches to preventing the collusive information leak caused by IPCs between multiple processes. In this paper, we propose a hybrid approach to prevent the collusive information leak based on information flow control. Our approach combines static information flow analysis and dynamic runtime checking together. Information leak caused by individual processes is prevented through static information flow control, and dynamic checking is done at runtime to prevent the collusive information leak. Such a combination may effectively reduce the runtime overhead of pure dynamic checking, and reduce false-alarms in pure static analysis. We develop this approach based on an abstract and simplified programming model, and formalize a novel definition of the leak-freedom property as our target security property. A simulation-based proof technique is used to prove that our approach is able to guarantee leak-freedom. All proofs are mechanized in Coq.

MAC A verified static information-flow control library

The programming language Haskell plays a unique, privileged role in information-flow control (IFC) research: it is able to enforce information security via libraries. Many state-of-the-art IFC libraries (e.g., LIO and HLIO) support a variety of advanced features like mutable data structures, exceptions, and concurrency, whose subtle interaction makes verification of security guarantees challenging. In this work, we focus on MAC, a statically-enforced IFC library for Haskell. In MAC, like other IFC libraries, computations have a well-established algebraic structure for computations (i.e., monads) responsible to manipulate labeled values—values coming from an abstract data type which associates a sensitivity label to a piece of information. In this work, we enrich labeled values with a functor structure and provide an applicative functor operator which encourages a more functional programming style and simplifies code. Furthermore, we present a full-fledged, mechanically-verified model of MAC. Specifically, we show progress-insensitive noninterference for our sequential calculus and pinpoint sufficient requirements on the scheduler to prove progress-sensitive noninterference for our concurrent calculus. For that, we study the security guarantees of MAC using term erasure, a proof technique that ensures that the same public output should be produced if secrets are erased before or after program execution. As another contribution, we extend term erasure with two-steps erasure, a flexible novel technique that greatly simplifies the noninterference proof and helps to prove many advanced features of MAC.

Dynamic security labels and static information flow control

This paper presents a language in which information flow is securely controlled by a type system, yet the security class of data can vary dynamically. Information flow policies provide the means to express strong security requirements for data confidentiality and integrity. Recent work on security-typed programming languages has shown that information flow can be analyzed statically, ensuring that programs will respect the restrictions placed on data. However, real computing systems have security policies that cannot be determined at the time of program analysis. For example, a file has associated access permissions that cannot be known with certainty until it is opened. Although one security-typed programming language has included support for dynamic security labels, there has been no demonstration that a general mechanism for dynamic labels can securely control information flow. In this paper, we present an expressive language-based mechanism for reasoning about dynamic security labels. The mechanism is formally presented in a core language based on the typed lambda calculus; any well-typed program in this language is secure because it satisfies noninterference.