We have a strange problem in using a time critial interrupt (samplin input) in conjunction with the freeRTOS.
my first question is, how long a yield is, which uses the critical section and disables the global interrupts.
So this could be the reason for blocking the interrupt.
I would expect, that the yield should not block that long the interupt or i am i something missing?
Any ideas are welcome. I have to check, if the RTOS itself is the problem for the issues in our application.
If you are referring to the context switch time then you can find some
information for a Cortex-M3 on the link below. An M0 will be a bit
slower due to its more limited instruction set.
Also unlike an M3, the M0 will globally disable interrupts in a critical
section - do you have long critical sections?