Information Flow Tracking for Side-Effectful Libraries

Abstract

Dynamic information flow control is a promising technique for ensuring confidentiality and integrity of applications that manipulate sensitive information. While much progress has been made on increasingly powerful programming languages ranging from low-level machine languages to high-level languages for distributed systems, surprisingly little attention has been devoted to libraries and APIs. The state of the art is largely an all-or-nothing choice: either a shallow or deep library modeling approach. Seeking to break out of this restrictive choice, we formalize a general mechanism that tracks information flow for a language that includes higher-order functions, structured data types and references.A key feature of our approach is the model heap, a part of the memory, where security information is kept to enable the interaction between the labeled program and the unlabeled library. We provide a proof-of-concept implementation and report on experiments with a file system library. The system has been proved correct using Coq.

Platform-Independent Dynamic Taint Analysis for JavaScript

“…This involves the following steps: (i) pushing the taint value corresponding to the receiver expression a onto the stack (line 24), (ii) pushing the taint value corresponding to the index expression 1 onto the stack (line 25), (iii) computing the taint value that is to be written and pushing it onto the stack (lines 26-32), (iv) generating a writeprop instruction to associate this value with the specified array element (line 33) without removing it from the stack (see instruction Ins writeprop in Figure 8), and (v) discarding the value computed by the assignment (line 34) 8 . Here, step (iii) reflects the evaluation of the expression process.argv [2], which involves retrieving the taint value associated with property argv of object process (lines [26][27][28], and retrieving the taint value associated with element 2 of the array (lines [29][30]. At this point, taint is introduced by discarding the previously read taint value (line 31) and pushing a taint value '(example.js:3:8)' (line 32).…”

“…Currently, JSFlow does not run on the Node.js applications that we consider due the absence of models for Node.js native functions. Later work by Hedin et al [29] and Sjösten et al [30] is concerned with developing concise models for tracking information flow in libraries, focusing on a small functional language in each case.…”