Modified condition/decision coverage (MC/DC) oriented compiler optimization for symbolic execution

Hong, Wooyoung; Liu, Yijun; Chen, Zhen-bang; Dong, Wei; Wang, Ji

doi:10.1631/fitee.1900213

Cited by 3 publications

(1 citation statement)

References 48 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This means that, at the current program point, the execution flow is more likely to execute reaching the goal point if the critical constraints are satisfied. Critical constraint only depends on a condition instead of a decision [15]. To identify the critical constraint, the first thing we need to do is converting a decision into one or more condition guards, for example, as shown in Figure 1, sum== 0 && m==100 is a decision while sum==0 and m==100 are condition guards respectively.…”

Section: Critical Constraint Recognitionmentioning

confidence: 99%

Vulnerability verification with the assistance of directed symbolic execution

Sun¹,

Tang²,

Wei³

et al. 2023

International Conference on Cryptography, Network Security, and Communication Technology (CNSCT 2023)

View full text Add to dashboard Cite

Static analysis is a popular approach for finding vulnerability in software. However, suffered from its imprecision, it is quite difficult and time-consuming to confirm static warnings. It also leads to a vulnerability verification problem. In this paper, we regard the verification problem as line reachability problem and utilize directed symbolic execution to verify specific bug. Existing techniques mostly relied on static heuristic metric but neglected some special dynamic metrics. For this reason, we designed a new search strategy named as DCCM which modifies others forward strategy via collecting critical constraint during the process of dynamic symbolic execution. By conduct contrast experiments on example program and 5 real-world C projects, we find out that DCCM performs well than KLEE RP, KLEE CoverNew and SDSE search strategy on many testing scenarios.

show abstract

Section: Critical Constraint Recognitionmentioning

confidence: 99%