Hey there,
I'm trying to make a simple PRU application there is reading an enhanced GPIO input and immidiatly sendt that value to an enhanced GPIO output. So here is what i came up with:
// INPUT = P9_29 = r31.t1 // OUTPUT = P9_30 = r30.t2 READ: qbbs WRITEHIGH, r31, 1 qbbc WRITELOW, r31, 1 WRITEHIGH: set r30.t2 jmp READ WRITELOW: clr r30.t2 jmp READ
So the application quick branches to a write state after having read the input bit where it writes the output bit and jumps back to start over.
It works fairly fine, but when I measure the input and output with a oscilloscope, I get the delay to ~35ns, which corresponds to ~7 cycles at 200MHz.
How come 35ns.!??
I read some where that the quick branches, and the jump took one cycle. If that's true i get the fastest case to 3 cycles (15ns) and the slow case to 4 cycles (20 ns). Maybe I'm wrong about branch and jump assumption, if so, does someone know or have the documentation for the cycle count for these instructions?
Does anyone know a faster way to do this data loopback?
