Martim Carbone, Weidong Cui, Long Lu, Wenke Lee, Marcus Peinado, and Xuxian Jiang
9 November 2009
Dynamic kernel data have become an attractive target for kernel-mode malware. However, previous solutions for checking kernel integrity either limit themselves to code and static data or can onlyinspect a fraction of dynamic data, resulting in limited protection. Our study shows that previous solutions may reach only 28% of the dynamic kernel data and thus may fail to identify function pointers manipulated by many kernel-mode malware.
To enable systematic kernel integrity checking, in this paper we present KOP, a system that can map dynamic kernel data with nearly complete coverage and nearly perfect accuracy. Unlike previous approaches, which ignore generic pointers, unions and dynamic arrays when locating dynamic kernel objects, KOP (1) applies interprocedural points-to analysis to compute all possible types for generic pointers (e.g., void*), (2) uses a pattern matching algorithm to resolve type ambiguities (e.g., unions), and (3) recognizes dynamic arrays by leveraging knowledge of kernel memory pool boundaries. We implemented a prototype of KOP and evaluated it on a Windows Vista SP1 system loaded with 63 kernel drivers. KOP was able to accurately map 99% of all the dynamic kernel data.
To demonstrate KOP’s power, we developed two tools based on it to systematically identify malicious function pointers and uncover hidden kernel objects. Our tools correctly identified all malicious function pointers and all hidden objects from nine real-world kernel-mode alware samples as well as one created by ourselves, with no false alarms.
In Proceedings of the 16th ACM Conference on Computer and Communications Security (CCS)
Publisher Association for Computing Machinery, Inc.
Copyright © 2007 by the Association for Computing Machinery, Inc. Permission to make digital or hard copies of part or all of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, to republish, to post on servers, or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from Publications Dept, ACM Inc., fax +1 (212) 869-0481, or email@example.com. The definitive version of this paper can be found at ACM’s Digital Library --http://www.acm.org/dl/.