Embedded Systems · Cluster
Real-Time Operating Systems
Tasks, priority inversion, and deadlines.
- 01 How a Preemptive RTOS Scheduler Picks the Next Task The ready-list, priority comparison, and context switch path a fixed-priority kernel follows on every tick, and why the runnable set is the only structure that matters.
- 02 Blocking Primitives: How a Task Waits Without Spinning The blocked-state bookkeeping behind a semaphore take, how a woken task reaches the ready set, and what the wake-up path costs in latency and stack.
- 03 Priority Inversion and How Bounded It Can Be Why a low-priority task holding a shared lock delays a high-priority one, how inheritance and ceiling protocols change the blocking term, and what each still leaves unbounded.
- 04 Rate-Monotonic Priority Assignment and Its Utilisation Bound Why assigning priorities by period is optimal among fixed-priority schemes, what the 69 percent bound actually guarantees, and where the harmonic case recovers full utilisation.
- 05 Response-Time Analysis for Fixed-Priority Task Sets The recurrence that computes exact worst-case response time, why it converges, what the blocking term adds, and where the critical instant assumption costs pessimism.
- 06 Deferred Interrupt Work: Splitting Handlers from Processing Why the top half of a handler should only acknowledge and wake, how the bottom half runs at a chosen priority, and what the split costs in latency.
- 07 Earliest-Deadline-First and What Happens When the System Overloads Why dynamic deadline ordering is optimal on one processor, how the utilisation test differs from the fixed-priority one, and how the schedule degrades after a bounded overload.
- 08 Interrupt Latency, Tick Overhead, and the Kernel Cost Budget How to decompose the delay from an external event to a task running into measurable pieces, and which kernel configuration choices move each piece.
- 09 Stack and Memory Budgeting for a Set of Preemptive Tasks Why each task needs its own stack, how worst-case stack depth is computed from the call graph plus exception frames, and where a shared stack changes the accounting.
- 10 Deadline Misses: Detection, Recovery, and Watchdog Design How a kernel learns that a deadline was missed, what recovery actions are safe after a miss, and how a watchdog is configured so that a stuck task resets cleanly.