A hardware coprocessor for Zynq-based Dynamic Information Flow Tracking - Archive ouverte HAL Access content directly
Conference Papers Year : 2016

A hardware coprocessor for Zynq-based Dynamic Information Flow Tracking

(1, 2) , (2, 1) , (1, 2)
1
2

Abstract

This talk introduces an efficient and portable approach for DIFT (Dynamic Information Flow Tracking) implementations on recon gurable chips. DIFT aims to track the application control flow by adding metadata (also known as tags) to information containers (e.g. registers, memory addresses, ...), propagating and checking it at runtime. These approaches have been successfully used against a wide range of attacks including bu er over ow, SQL injections and so on. Existing DIFT solutions are either hardly portable or bring unsatisfactory time overheads. For example, DIFT in software brings an overhead of at least 300% and can rise up to 3700%. Our chosen approach conists of using a dedicated coprocessor to decouple main computation (application) from tags computation. To efficiently retrieve information on executed instructions by the main CPU, ARM Coresight components are used to export CPU trace towards FPGA part of Zynq SoC. In addition to trace, static analysis needs to be done to get dependencies between information containers. This talk will present our chosen approach, ongoing work on first prototype and rst results. Existing DIFT approaches and ARM Coresight components will also be discussed.
Vignette du fichier
09-Wahab.pdf (1.41 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-01341830 , version 1 (11-07-2016)

Identifiers

  • HAL Id : hal-01341830 , version 1

Cite

Muhammad Abdul Wahab, Christophe Moy, Pascal Cotret. A hardware coprocessor for Zynq-based Dynamic Information Flow Tracking. 14th International Workshop on Cryptographic architectures embedded in logic devices (CryptArchi), Jun 2016, La Grande Motte, France. ⟨hal-01341830⟩
260 View
109 Download

Share

Gmail Facebook Twitter LinkedIn More