PosFuzz: augmenting greybox fuzzing with effective position distribution

Cybersecurity

Research

Wei Zou¹  Nanyu Zhong¹  Ji Shi¹  Wei Huo¹  Yanyan Zou¹  Yu Zhang¹  JiaCheng Zhao² 

[1] Institute of Information Engineering, Chinese Academy of Sciences, Beijing, China;School of Cyber Security, University of Chinese Academy of Sciences, Beijing, China;Key Laboratory of Network Assessment Technology, Chinese Academy of Sciences, Beijing, China;Beijing Key Laboratory of Network Security and Protection Technology, Beijing, China;State Key Lab of Processors, Institute of Computing Technology, Chinese Academy of Sciences, Beijing, China;Zhongguancun Laboratory, Beijing, China;

关键词: Greybox fuzzing;    Mutation position;    Mutation operator;    Code coverage;    Vulnerability discovery;   

DOI  :  10.1186/s42400-023-00143-2

 received in 2022-08-06, accepted in 2023-02-10,  发布年份 2023

来源: Springer

【 摘 要 】

Mutation-based greybox fuzzing has been one of the most prevalent techniques for security vulnerability discovery and a great deal of research work has been proposed to improve both its efficiency and effectiveness. Mutation-based greybox fuzzing generates input cases by mutating the input seed, i.e., applying a sequence of mutation operators to randomly selected mutation positions of the seed. However, existing fruitful research work focuses on scheduling mutation operators, leaving the schedule of mutation positions as an overlooked aspect of fuzzing efficiency. This paper proposes a novel greybox fuzzing method, PosFuzz, that statistically schedules mutation positions based on their historical performance. PosFuzz makes use of a concept of effective position distribution to represent the semantics of the input and to guide the mutations. PosFuzz first utilizes Good-Turing frequency estimation to calculate an effective position distribution for each mutation operator. It then leverages two sampling methods in different mutating stages to select the positions from the distribution. We have implemented PosFuzz on top of AFL, AFLFast and MOPT, called Pos-AFL, -AFLFast and -MOPT respectively, and evaluated them on the UNIFUZZ benchmark (20 widely used open source programs) and LAVA-M dataset. The result shows that, under the same testing time budget, the Pos-AFL, -AFLFast and -MOPT outperform their counterparts in code coverage and vulnerability discovery ability. Compared with AFL, AFLFast, and MOPT, PosFuzz gets 21% more edge coverage and finds 133% more paths on average. It also triggers 275% more unique bugs on average.

【 授权许可】

CC BY   
© The Author(s) 2023

【 图 表 】

42004_2023_911_Article_IEq13.gif

42004_2023_911_Article_IEq10.gif

42004_2023_911_Article_IEq9.gif

42004_2023_911_Article_IEq8.gif

Fig. 2

Fig. 1

40854_2023_495_Article_IEq32.gif

Fig. 9

Fig. 1

12302_2023_750_Article_IEq9.gif

12302_2023_750_Article_IEq8.gif

Fig. 4

Fig. 6

Fig. 8

Fig. 2

40854_2023_495_Article_IEq22.gif

Fig. 1

Fig. 1

12302_2023_750_Article_IEq5.gif

42004_2023_932_Article_IEq5.gif

Fig. 3

Fig. 3

Fig. 7

40538_2023_419_Article_IEq1.gif

Fig. 2

12864_2023_9424_Article_IEq11.gif

Fig. 6

Fig. 4

Fig. 3

Fig. 2

Fig. 4

12864_2023_9407_Article_IEq46.gif

12864_2023_9424_Article_IEq2.gif

Fig. 7

Fig. 5

Fig. 1

Fig. 2

Fig. 1

Fig. 3