Note: 1. The newly added document mainly explains how to monitor task scheduling performance through the SAFECheckpoints feature of SafeRTOS, in order to allocate task priorities and plan task processing times more reasonably.
2. Efficient implementation of task priority allocation and task functionality requires a significant amount of practical experience accumulation. (1) For example, the emWin task is generally set to have a priority just higher than that of the idle task. Additionally, in our upcoming epic RL-TCPnet networking tutorial, we generally hope that the network’s ping latency is low, allowing for timely responses to external communications. In this case, it is essential to use an event-driven approach, where the network data sending function and the data receiving function in the interrupt service routine notify the main network task as soon as there is data to process. There are many similar situations that require continuous experience accumulation. (2) However, relying solely on experience is not enough; we need quantification to obtain detection of task execution status at each stage. This requires the RTOS to carry corresponding functional components or to use a good detection software. Here, we recommend SEGGER’s free SystemView. Yesterday, Micrium sent an email notifying that both uCOS-II and uCOS-III have been upgraded to a minor version, both of which now support SystemView. Once the networking tutorial is released, we will take time to add SystemView support for our RTX, uCOS-II, uCOS-III, and FreeRTOS examples.