Hi,
I wrote a PRU code to interface a 500kHz spi ADC.
At present, I am using a Flip/flop buffer in the PRU shared memory, and the ARM is reading the buffer once when it is complete.
In terms of IO's cost, what is the most efficient scenario :
1/ The PRU is writing data in its own memory space, and the ARM read back the full buffer and copy it on its own space.
2/ The PRU write each data in the ARM space, using the OCP master port, and the ARM copy "locally" the buffer.
My concern is on the ARM side, as 500kHz is slow enought for the PRU.
I am not sure that writing thru the OCP port is not slowing down the ARM.
Thanks for your expectise.
Eric.
Thanks for your
