Hello, we are planning to read sensor data using an MCU and intend to utilize this data on our custom board equipped with an i.MX8MP. The data will be transferred from the MCU to the i.MX via SPI. My question is: should we use the M7 core on the i.MX to receive the data and then process it within our user application on our custom distro, or should we handle the SPI communication directly with the A core? Which method would be easier to integrate and result in lower latency?
There is latency using rpmsg, the m7 write data to shared memory(need time), then use MU to notify A core(need time), then A core read data from shared memory(need time). So why not receive data from A core SPI directly as you need lower latency
We'll have to write a driver to parse the incoming data to use in the userspace. Any suggestions for this purpose or do you see any limitations? Also Why using m7 core and using the sensor data on the user space via rpmsg is a bad idea?
There is latency using rpmsg, the m7 write data to shared memory(need time), then use MU to notify A core(need time), then A core read data from shared memory(need time). So why not receive data from A core SPI directly as you need lower latency
