FEther: An Extensible Definitional Interpreter for Smart-Contract Verifications in Coq

Abstract

Recently, blockchain technology, which adds records to a list using cryptographic links, has been widely applied in the financial field. Therefore, the security of blockchain smart contracts is among the most popular contemporary research topics. To improve the theorem-proving technology in this field, we are developing an extensible hybrid verification tool chain, denoted as FSPVM-E, for Ethereum smart contract verification. This hybrid system extends the Coq proof assistant, a formal proof-management system. Combining symbolic execution with higher-order theorem-proving, it solves consistency, automation, and reusability problems by standard theorem-proving approaches. This article completes the FSPVM-E by developing its proof engine. FSPVM-E is an extensible definitional interpreter based on our previous work FEther, which is totally developed in the Coq proof assistant. It supports almost all semantics of the Solidity programing language, and simultaneously executes multiple types of symbols. FEther also contains a set of automatic strategies that execute and verify the smart contracts in Coq with a high level of automation. The functional correctness of FEther was verified in Coq. The execution efficiency of FEther far exceeded that of the interpreters which are developed in Coq in accordance with the standard tutorial. To our knowledge, FEther is the first definitional interpreter of the Solidity language in Coq.Blockchain technology [1], which adds records to a list using cryptographic links, is among the most popular contemporary technologies. Ethereum is a widely adopted blockchain system that implements a general-purpose, Turing-complete programing language called Solidity [2]. Ethereum enables the development of arbitrary smart contracts that can automate blockchain transactions in a virtual runtime environment, namely, the Ethereum Virtual Machine (EVM). Here smart contracts refer to the applications and scripts (i.e., programs) that execute the blockchain. The growing use of smart contracts has necessitated increased scrutiny of their security. Smart contracts can include particular properties (i.e., bugs) that expose them to deliberate attacks causing direct economic loss. Some of the largest attacks on smart contracts are well known, such as the attacks on decentralized autonomous organizations [3] and parity wallet contracts [4]. Many classes of subtle bugs, ranging from transaction-ordering dependencies to mishandled exceptions, exist in smart contracts [5]. Therefore, the security and reliability of smart-contract programs must be verified as rigorously as possible. The properties of programs can be rigorously verified by proving higher-order logic theorems. In the standard approach, a formal model for the target software system is manually abstracted using higher-order theorem-proving assistants. Such formal verification technology provides sufficient freedom and flexibility for designing formal models based on higher-order logic theories, and can abstract and express very complex systems. Howe...

A blockchain-based decentralized booking system

“…There have been notably a few works on the source code level, for example Bhargavan et al . (2016), Yang and Lei (2019). However, they require the analysts to learn another language to reinterpret the applications; and the learning curve of these languages is steep.…”

“…In addition, the formal verification in this work only focuses on the application design; the implementation may introduce vulnerabilities; therefore, formal verification on the implementation is needed. There have been notably a few works on the source code level, for example Bhargavan et al (2016), Yang and Lei (2019). However, they require the analysts to learn another language to reinterpret the applications; and the learning curve of these languages is steep.…”

A Hybrid Formal Verification System in Coq for Ensuring the Reliability and Security of Ethereum-Based Service Smart Contracts

“…Thus, compared with other static analysis tools that must assume the correctness of the TCB, the smart contract verifications that take the proposed FSPVM-E as the TCB must only trust the TCOC. The corresponding verification details are given elsewhere [11,31,32]. At present, the core functions of FSPVM-E, which includes 74 theorems and 183 lemmas in Coq, have been completely verified.…”

“…In response to this issue, we optimized FEther [32] by applying three proposed optimization schemes, including redefining semantics, deeply embedding, and multiple pumps. This optimized version of FEther is that employed for the proposed FSPVM-E discussed in the present report.…”

A Formal Verification Framework for Security Issues of Blockchain Smart Contracts

“…The SECBIT group completed the formal verification of the ERC20 contract framework in [30]. Yang and Lei [31] developed a definitional interpreter FEther, which supports hybrid symbolic executions of Ethereum smart-contract formal verifications in Coq.…”