Reason about deadlocks in async Rust

[ple1n]

How does its model compare to Petri Net which some tool uses to prove the absence of deadlock for Rust code ?

magmide/magmide#29

It may be problematic if it can not cover an entire codebase, too. Maybe a silly question, but I couldn't find handy enough tooling/solution about this problem.