What are the benefits of DSP program architecture?

Jacktang

What are the benefits of DSP program architecture? [Copy link]

I learned DSP C2000 series and read a lot of examples and documents. I found that the program architecture is basically the following structure:

What is the benefit of this structure? Using a super loop to poll continuously, there is no need for a timer, so the CPU utilization rate is even higher. The only benefit I can think of for this structure is that the task execution has time determinism, but this requires calculating the running time of each task and reasonably allocating the timing value, but it is troublesome. . Question: Are there any other hidden benefits besides this? From the perspective of CPU utilization, polling is better. However, as you said, some tasks have real-time requirements, especially in the case of nested interrupts.

Given that you are all timer tasks, polling is indeed more appropriate; However, if you also need to consider low-power optimization, when there is nothing to do, the CPU needs to enter low-power mode in IDLE, then it can only be awakened by timer interrupts or other interrupts.

1. It is easy to control the task execution interval. In fact, the execution frequency of many tasks is not the higher the better. In addition, the execution of tasks often requires a certain execution time interval. 2. Power consumption. When there is no need to execute tasks, the system can enter sleep or low-power working state to save more energy.