Abstract

The paper considers the challenge of deductively verifying Linux kernel code written in C programming language with extensive use of low-level memory operations and interactions with the highly concurrent environment. The paper presents an initial approach to specification and verification of concurrent code working with shared data by proving the code's compliance with specified synchronization discipline. The proposal is illustrated with an example specifying a user-side simplified model of the read-copy-update synchronization mechanism widely used within the Linux kernel.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call