Skip to content

Latest commit

 

History

History
60 lines (40 loc) · 3.94 KB

README.md

File metadata and controls

60 lines (40 loc) · 3.94 KB
Raw

FreeRTOS Custom Project

This repository contains a FreeRTOS-based project with several enhancements and modifications beyond the base operating system. The project was developed as part of the Master's in Cybersecurity at Politecnico di Torino during the 2023/2024 academic year. The goal of this project is to showcase FreeRTOS features and implement additional functionalities, including custom scheduling algorithms and memory allocation mechanisms.

Features

  1. FreeRTOS Scheduler:

    • The base FreeRTOS scheduler supports preemptive, priority-based, and round-robin scheduling.
    • Demonstrations of various scheduling algorithms, including Rate Monotonic (RM) and custom Early Deadline First (EDF) scheduling.

  2. Priority Inversion Problem:

    • A scenario demonstrating priority inversion and its resolution using binary and mutex semaphores for mutual exclusion. Mutex semaphores temporarily raise task priority to solve the inversion.

  3. Task Communication (Event Groups):

    • Multiple tasks communicate via event groups for task synchronization. Mutex semaphores and event flags are used for mutual exclusion and event-driven task synchronization.

  4. Rate Monotonic Scheduling:

    • Implementation of RM scheduling where tasks are executed based on fixed priorities. vTaskDelayUntil() ensures periodic task execution with constant frequency.

Modifications Over Base FreeRTOS

1. EDF Scheduling Implementation

  • EDF Scheduler: Added support for EDF scheduling, which is not available in the default FreeRTOS. EDF schedules tasks based on deadlines, running the task with the closest deadline to avoid deadline misses.
  • New Ready List: A new list (xReadyTaskListEDF) is introduced to store tasks based on their deadlines.
  • API Extensions: Introduced xTaskPeriodicCreate() for periodic task creation with a specified deadline.
  • Task Context Switching: Context switching is modified to allow tasks with closer deadlines to preempt other tasks.

2. Memory Allocation Mechanisms

  • Two new memory allocation strategies were implemented:
    • Best Fit Algorithm: Allocates the smallest available block that fits the request, optimizing memory use.
    • Buddy System Allocation: Combines small memory blocks to reduce fragmentation.
  • These are implemented in the heap_5.c file, allowing allocation from multiple memory spaces.

3. Benchmark

  • A benchmark comparing Rate Monotonic (RM) and EDF scheduling was performed. The results show:
    • EDF: Maximizes CPU utilization and handles tighter deadlines, but with increased context switches.
    • RM: Predictable, fixed priorities, but limited CPU utilization and more susceptible to sporadic tasks causing deadline misses.

Directories:

  • The FreeRTOS/Source directory contains the FreeRTOS source code, along with its own readme file.

  • The FreeRTOS/Demo directory contains a customized demo application.

For full details of the directory structure and information on locating the files, see FreeRTOS/SourceOrganization.

The easiest way to use FreeRTOS is to start with one of the pre-configured demo application projects (found in the FreeRTOS/Demo directory). This ensures that:

  • The correct FreeRTOS source files are included.
  • The appropriate include paths are configured.

Once a demo application is building and executing, you can remove the demo files and start adding your own application source files.

Quick Links:

Conclusion

This project enhances FreeRTOS with an EDF scheduler and improved memory allocation techniques, making it more suitable for complex real-time applications. These modifications expand FreeRTOS's capabilities in environments where dynamic scheduling and memory management are critical for task management.