From Theory to Code: Identifying Logical Flaws in Cryptographic Implementations in C/C++

Rahaman, Sazzadur; Cai, Haipeng; Chowdhury, Omar; Yao, Danfeng

doi:10.1109/tdsc.2021.3108031

Cited by 4 publications

(4 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although static analysis is not expected to be perfect, our results suggest that false alarms in detecting cryptographic misuse can occur not only due to classic challenges, but also because of ① overly conservative misuse rules, ② imprecise modeling, and ③ implementation bugs in detectors themselves. The false alarm patterns due to ① likely also apply to other papers (e.g., [4], [6], [7], [9]) and industrial tools not considered by this work, as they share similar misuse rules.…”

Section: Introductionmentioning

confidence: 75%

“…A recent work that uses a hybrid approach to find vulnerabilities in binaries [71] also detects any constant seeds used in PRNG regardless of context, and we conjecture that some legitimate usages are also reported as vulnerabilities. Likewise, some overly conservative rules discussed in Section VII are also enforced by a recent work [9].…”

Section: Generalizability Of False Alarm Patternsmentioning

confidence: 99%

“…As cryptography plays an indispensable role in securing contemporary systems, many research efforts went into analyzing both the implementation and usage of cryptography. Specifically, a long line of work has been dedicated to the problem of designing and implementing detectors that can catch misuses of cryptographic APIs in various applications [1]- [9]. Common examples of misuses that are caught and reported include the invocation of algorithms perceived to be weak (e.g., AES-ECB), as well as the use of constants (e.g., keys and passwords) and weak parameters (e.g., short key lengths).…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Towards Precise Reporting of Cryptographic Misuses

Chen,

Liu,

et al. 2024

Proceedings 2024 Network and Distributed System Security Symposium

View full text Add to dashboard Cite

In the last decade, a series of papers were published on using static analysis to detect cryptographic API misuse. In each paper, apps are checked against a set of rules to see if violations exist. A common theme among these papers is that rule violations are plentiful, often at the scale of thousands. Interestingly, while much effort went into tackling false negatives, curiously, not much has been said on (1) whether the misuse alarms are indeed correct and meaningful, and ( 2) what can future work improve upon apart from finding more misuses.In this paper, we take a deep dive into the rule violations reported by various academic papers as well as the rules, models and implementations of their detectors, in an attempt to (1) explain the gap between their misuse alarms and actual vulnerabilities, and (2) shed light on possible directions for improving the precision and usability of misuse detectors. Results of our analysis suggest that the small-scale inspections done by previous work had some unfortunate blind-spots, leaving problems in their rules, models, and implementations unnoticed, which in turn led to unnecessary overestimation of misuses (and vulnerabilities). To facilitate future research on the topic, we distill these avoidable false alarms into high-level patterns that capture their root causes, and discuss design, evaluation and reporting strategies that can improve the precision of misuse findings. Furthermore, to demonstrate the generalizability of these false alarm patterns and improvement directions, we also investigate a popular industry detector and a dynamic detector, and discuss how some of the false alarm patterns do and do not apply to them. Our findings suggest that the problem of precisely reporting cryptographic misuses still has much room for future work to improve upon.

show abstract

Section: Introductionmentioning

confidence: 75%

Section: Generalizability Of False Alarm Patternsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Towards Precise Reporting of Cryptographic Misuses

Chen,

Liu,

et al. 2024

Proceedings 2024 Network and Distributed System Security Symposium

View full text Add to dashboard Cite

show abstract

“…Their research revealed that 52.26% of Python projects exhibit at least one instance of cryptographic misuse. Rahaman et al [21] developed a static tool named TAINTCRYPT, tailored for detecting cryptographic implementations in C/C++ programming languages. They utilized a specification language of deterministic finite automata (DFA).Rodrigues et al [22] proposed an innovative methodology that combines graph embedding techniques with machine learning models to detect instances of cryptographic abuse in source code.…”

Section: Related Workmentioning

confidence: 99%

Intelligent Detection of Cryptographic Misuse in Android Applications Based on Program Slicing andTransformer-Based Classifier

Wang

Sui

et al. 2023

Electronics

View full text Add to dashboard Cite

The utilization of cryptography in applications has assumed paramount importance with the escalating security standards for Android applications. The adept utilization of cryptographic APIs can significantly enhance application security; however, in practice, software developers frequently misuse these APIs due to their inadequate grasp of cryptography. A study reveals that a staggering 88% of Android applications exhibit some form of cryptographic misuse. Although certain tools have been proposed to detect such misuse, most of them rely on manually devised rules which are susceptible to errors and require researchers possessing an exhaustive comprehension of cryptography. In this study, we propose a research methodology founded on a neural network model to pinpoint code related to cryptography by employing program slices as a dataset. We subsequently employ active learning, rooted in clustering, to select the portion of the data harboring security issues for annotation in accordance with the Android cryptography usage guidelines. Ultimately, we feed the dataset into a transformer and multilayer perceptron (MLP) to derive the classification outcome. Comparative experiments are also conducted to assess the model’s efficacy in comparison to other existing approaches. Furthermore, planned combination tests utilizing supplementary techniques aim to validate the model’s generalizability.

show abstract