Orchestration: resiliency state machine

[rusty1968]

This PR introduces the state machine component of the Orchestrator. It is pure transition logic only, no platform I/O.

The SM is the resiliency policy authority: it defines what happens, in what order, and under what conditions. Policy is expressed as a Hierarchical State Machine (HSM) that enforces correct sequencing through state transitions and entry/exit actions — not through ad-hoc logic scattered across a platform implementation.

[LeeTroy]

The original ASPEED-PFR state machine has a super state concept, it helps to control BMC/Host's reset when entering or leaving T-1 (Preboot) vs T0(Runtime). It seems the current design doesn't have this in mind.

For the run action, there are three handler function for each state, i.e. enter/run/exit, and two hander function for super state, i.e. enter/exit, for example:

stateDiagram-v2
    [*] --> A: e0

    state Tmin1 {
        A --> B: e1
        C --> A: e4
    }

    state Tzero { 
        B --> B: e2
        B --> C: e3
    }

Loading

1. Event e1 will trigger A.exit(), Tmin1.exit(), Tzero.enter(), B.enter(), B.run()
2. Event e2 will trigger B.run()
3. Event e3 will trigger B.exit(), C.enter(), C.run()
4. Event e4 will trigger C.exit(), Tzero.exit(), Tmin1.enter(), A.enter(), A.run()