Papers › Targeted Static Analysis for OCaml C Stubs: eliminating gremlins from the code
Targeted Static Analysis for OCaml C Stubs: eliminating gremlins from the code
Edwin Török
The archive published only this paper's code-link row. Authors, date and abstract are from arXiv's metadata (CC0), read from the Kaggle arXiv metadata snapshot of 2026-09-12 where its title matched the archive's; the title is the archive's.
Migration to OCaml 5 requires updating a lot of C bindings due to the removal of naked pointer support. Writing OCaml user-defined primitives in C is a necessity, but is unsafe and error-prone. It does not benefit from either OCaml's or C's type checking, and existing C static analysers are not aware of the OCaml GC safety rules, and cannot infer them from existing macros alone.The alternative is automatically generating C stubs, which requires correctly managing value lifetimes. Having a static analyser for OCaml to C interfaces is useful outside the OCaml 5 porting effort too. After some motivating examples of real bugs in C bindings a static analyser is presented that finds these known classes of bugs. The tool works on the OCaml abstract parse and typed trees, and generates a header file and a caller model. Together with a simplified model of the OCaml runtime this is used as input to a static analysis framework, Goblint. An analysis is developed that tracks dereferences of OCaml values, and together with the existing framework reports incorrect dereferences. An example is shown how to extend the analysis to cover more safety properties. The tools and runtime models are generic and could be reused with other static analysis tools.
Code
Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.
Code Syntology ran Syntology
Not run by Syntology. Nothing on this page verifies that the listed code works.
Results from the paper archive 2025-07-28
No leaderboard rows for this paper in the archive.
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections