Real-time vs general-purpose OS in FreeRTOS - Performance Comparison
When comparing real-time and general-purpose operating systems, it is important to understand how their task handling affects execution time.
We want to see how the time to complete tasks grows as the system workload increases.
Analyze the time complexity of task scheduling in a FreeRTOS real-time system versus a general-purpose OS.
// FreeRTOS task scheduler example
void vTaskScheduler(void) {
for (;;) {
TaskHandle_t nextTask = xTaskSelectNext();
vTaskSwitchContext();
}
}
// General-purpose OS scheduler (simplified)
void scheduler(void) {
while (1) {
selectNextProcess();
contextSwitch();
}
}
This code shows the main loop of task scheduling in both systems, where tasks or processes are selected and switched.
Both schedulers run an infinite loop repeatedly selecting the next task or process.
- Primary operation: Selecting the next task or process to run.
- How many times: This happens continuously, once per scheduling cycle.
As the number of tasks or processes increases, the scheduler must check more items to pick the next one.
| Input Size (number of tasks/processes) | Approx. Operations per scheduling |
|---|---|
| 10 | 10 checks |
| 100 | 100 checks |
| 1000 | 1000 checks |
Pattern observation: The number of operations grows roughly in direct proportion to the number of tasks or processes.
Time Complexity: O(n)
This means the time to select the next task grows linearly as the number of tasks increases.
[X] Wrong: "The scheduler always picks the next task instantly, no matter how many tasks there are."
[OK] Correct: In reality, the scheduler must check each task to decide which runs next, so more tasks mean more work and longer selection time.
Understanding how task scheduling time grows helps you explain system responsiveness and design choices in real-time versus general OS environments.
"What if the scheduler used a priority queue instead of scanning all tasks? How would the time complexity change?"