Consider integrating Fuzz Introspector into our continuous fuzzing setup

[tschneidereit]

Feature

Fuzz Introspector helps to gain insights into the effectiveness of a fuzzing setup and to identify blockers that keep parts of the code base uncovered.

Benefit

Wasmtime is sufficiently complex that it's challenging to reason about the effectiveness of the various fuzzing harnesses we employ. That's the very problem Fuzz Introspector was created to help address.

Implementation

Rust support was added to Fuzz Introspector very recently, so it's not yet fully documented. I think that PR contains sufficient bread crumbs to find the right path to integration with our OSS-Fuzz setup, though.

Alternatives

There might be alternative initiatives with similar goals that'd be more valuable to integrate. If so, I'm not aware of them.

[github-actions]

Subscribe to Label Action

cc @fitzgen

This issue or pull request has been labeled: "fuzzing"

Thus the following users have been cc'd because of the following labels:

• fitzgen: fuzzing

To subscribe or unsubscribe from this label, edit the .github/subscribe-to-label.json configuration file.

Learn more.

[tschneidereit]

One potentially interesting benefit is that we could experiment with Google's LLM-based fuzz target generation, which relies in Fuzz Introspector. "Potentially interesting" because I absolutely don't know if that would be valuable for a project like Wasmtime.

[fitzgen]

This looks like it would give us really insightful feedback!