Add an EffectiveSan (Effective Type Sanitizer) CI job to detect sub-object overflows and type errors #23393

Add an EffectiveSan (Effective Type Sanitizer) CI job to detect sub-object overflows (bounds errors within the same object) and type errors (accessing memory using the wrong type).

• EffectiveSan tool
• EffectiveSan paper

Q: Why do we need EffectiveSan when we already have AddressSanitizer?

AddressSanitizer is a popular tool for detecting memory errors such as bounds overflows and use-after-free errors. EffectiveSan can also detect these kinds of errors, as well as other classes of error that AddressSanitizer cannot detect, such as:

• Sub-object overflows: Bounds errors within the same object. AddressSanitizer can only detect overflows that escape the bounds of the allocation; and
• Type errors: Accessing memory using the wrong type.

Unfortunately I'm swamped with way too much non-Bitcoin non-crypto development work for the foreseeable future so realistically I won't have time to add it myself, but I thought I could add an issue for it at least and hope that someone with more Bitcoin Core time can run with the idea :)

The linked project, whilst potentially looking interesting, has not had any updates in over 4 years, and has only a 0.1-alpha release which appears to point to the 4.0 release of llvm.

Along with this issue itself having no activity in 5 years, I think we can close this for now due to lack of interest.

Let me know if you think it should be re-opened, or feel free to open a new issue if this is still important.