Project Title: Analysis Suite for Rust's Advantages in Linux
Student: Fahad Khan
Course: BSc Hons Computer Science
	
Abstract:
Rust has been introduced into Linux kernel development to reduce long-standing memory-safety and
concurrency hazards associated with C, yet maintainers still lack disciplined module-level evidence
for deciding where migration effort is justified. This dissertation presents an analysis suite for Linux
oriented modules that assesses Rust’s practical advantage across three scenarios: C-only migration
pressure analysis, Rust-only boundary profiling, and paired C/Rust comparison. The suite combines
structural source analysis with optional compiler-backed support to examine C hazard patterns,
Rust unsafe containment, raw-binding and foreign-function-interface exposure, abstraction use, and
reviewed correspondence quality. It supports bounded Linux-oriented inputs including source files,
Cargo projects, crate roots, module trees, compile databases, and curated pair manifests, and produces
aligned command-line, Markdown, and JSON reports with explicit claim boundaries. Evaluation over
a curated corpus of 17 cases and 27 runs shows that the workflow behaves credibly and repeatably on
supported inputs. Of 109 manually reviewed high-confidence findings, 103 were confirmed, giving
a 94.5% confirmation rate on the reviewed sample. Compiler-backed support sharpened recovered
evidence in most comparison cases, while the paired lane promoted stronger comparative claims only
when reviewed coverage justified them. The dissertation therefore contributes a reusable methodology
and implementation for bounded, inspectable Rust-advantage assessment in Linux-oriented code,
rather than a general proof of Rust superiority or program correctness.