It’s been a little while, but I’d like to include a paper that uses and extends LLVM for more precise and declarative dynamic taint-tracking:
Compiler-instrumented, Dynamic Secret-Redaction of Legacy Processes for Attacker Deception
Frederico Araujo, Kevin W. Hamlen
Proceedings of the 24th USENIX Security Symposium, August 2015.
I would appreciate if someone could add it to the LLVM publication list.