Function initialize_with_tests 

pub fn initialize_with_tests(
    integrity: &[KatEntry],
    tests: &[KatEntry],
) -> Result<(), Error>

Initializes the module and runs every supplied power-up KAT.

This is the canonical entry point. It transitions PowerOff -> SelfTest, executes every entry in tests in order, and then transitions to Operational on success or to the terminal Error state on the first failure. The returned Error::SelfTestFailed carries the name of the failing test.

Concurrent calls are safe: exactly one caller wins the PowerOff -> SelfTest CAS and runs the tests on its own thread. A caller that does not win waits for the winner’s outcome and receives Error::AlreadyInitialized if it succeeded or Error::AlreadyFailed if it failed. A call on a module that has already left PowerOff runs no test and changes nothing: initialisation happens once per process, and a failed one is terminal until the process restarts. A call made from inside a self-test, on the thread running them, returns Error::NotOperational naming SelfTest.

On a platform that gives the calling thread no id the module can record, the winning call latches Error and returns Error::SelfTestFailed naming THREAD_ID_CHECK.

Note: the slice is passed by reference, not built up by magic. This repository deliberately avoids linker-section registration tricks: the full set of power-up tests is visible in source at every call site, which makes it straightforward to audit the module’s test inventory against the Security Policy.