SearcharxivSearch

arXiv subjects

Dongpeng Xu

Publications and source records attributed to Dongpeng Xu.

2 recordsLinked to original sources

HALF: Hollowing Analysis Framework for Binary Programs with Kernel Module Assistance

Binary program analysis represents a fundamental pillar of modern system security. Fine-grained methodologies like dynamic taint analysis still suffer from deployment complexity and performance overhead despite significant progress. Traditional in-process analysis tools trigger severe \textbf{address-space conflicts} that inevitably disrupt the native memory layout of the target. These conflicts frequently cause layout-sensitive exploits and evasive malware to deviate from their intended execution paths or fail entirely. This paper introduces \textbf{HALF} as a novel framework that resolves this fundamental tension while ensuring both analysis fidelity and practical performance. HALF achieves high-fidelity address-space transparency by leveraging a kernel-assisted process hollowing mechanism. This design effectively eliminates the observation artifacts that characterize traditional instrumentation tools. We further mitigate the synchronization latency of decoupled execution by implementing an exception-driven strategy via a lightweight kernel monitor. Extensive evaluation of a Windows-based prototype demonstrates that HALF maintains superior performance compared to conventional in-process baselines. HALF also provides unique capabilities for deconstructing complex, stealthy threats where existing frameworks fail to maintain execution integrity.

cs.SE

Towards Optimal Use of Exception Handling Information for Function Detection

Function entry detection is critical for security of binary code. Conventional methods heavily rely on patterns, inevitably missing true functions and introducing errors. Recently, call frames have been used in exception-handling for function start detection. However, existing methods have two problems. First, they combine call frames with heuristic-based approaches, which often brings error and uncertain benefits. Second, they trust the fidelity of call frames, without handling the errors that are introduced by call frames. In this paper, we first study the coverage and accuracy of existing approaches in detecting function starts using call frames. We found that recursive disassembly with call frames can maximize coverage, and using extra heuristic-based approaches does not improve coverage and actually hurts accuracy. Second, we unveil call-frame errors and develop the first approach to fix them, making their use more reliable.

cs.CR