Hi,everyone
When I was using the gpio interrupt, I found that the interrupt service function was not entered immediately after the interrupt was triggered. This period of time is not fixed, sometimes very long, sometimes very short. This situation will be directly imaged in my data collection. Is there any good solution?
The yellow line is the data line, and the blue line is the interrupt pin.
I found that when the same version of the program runs on different boards, the resources occupied by sys on Linux are different. Some boards take up a lot of resources, while some are very small. The above situation will occur if the resources are large, and there is no such problem if the resources are small. I wonder if there is a difference in the performance of the chip? How to measure the performance of the chip?
Hi
in general linux is not real time operating system, it does not guarantee interrupt
latencies. One can look at linux RT patches or nxp Real-Time Edge Software
https://wiki.linuxfoundation.org/realtime/documentation/howto/applications/preemptrt_setup
Best regards
igor
Hi!
I know that Linux is not a real-time operating system. This is because the CPU usage is too high and the system freezes. When I tested it, I found that some boards would be very lagging, but some were very good. The code is also the same version, will it be related to the chip itself?