Callback chain in Security Manager

It would be useful if (at least) onLinkSecured could have multiple callbacks, since different modules could depend on knowing the security status of the connection.

Start by locating the Security Manager implementation and the onLinkSecured callback. Trace how the current callback is registered and invoked, then determine how multiple modules could receive the connection security status. Done means independent callbacks can be registered and all are invoked reliably when the link is secured.

Written by the indexing model from the issue text.