Phase 2: Building custom Schedulers

🧠 The Mission

Transitioning from simple Super-Loop architectures to a deterministic, Multi-threaded Preemptive Kernel. This phase focuses on the “under the hood” mechanics of how an OS manages the CPU.

---

🛠 Core Objectives

1. Context Switching

Mastering the transition between tasks by manually saving and restoring CPU registers using the PendSV exception.

2. Priority Management

• Scheduling Logic: Implementing Fixed-Priority and Round-Robin algorithms.
• Priority Inversion: Simulating the “Unbounded Priority Inversion” bug and implementing Priority Inheritance as a professional-grade fix.

3. Memory Safety & Analysis

• Stack Guard Bands: Implementing magic-number watermarking to detect stack growth.
• MPU Integration: Using the Memory Protection Unit to trigger a HardFault on stack overflows.

---

🚀 Key Milestones

1. The Custom Scheduler: A “from-scratch” implementation on the STM32L475.
2. Multicore Concepts: Introduction to SMP (Symmetric Multiprocessing) and Spinlocks.

---

Table of contents

• Folder Structure
• 01. Custom Preemptive Round-Robin Scheduler
• 02. Custom Priority-Based Preemptive Scheduler
• Lesson learned