I don't see the need for a user space program to have accurate real time
unless it is a real time process (because there is no way for a non real
time process to know how long its time slice will last and that's there
no page fault or something which delays it for seconds).
For real time processes, you don't want to have clock slowing (which isn't
real clock slowing) the way it is implemented on Intel anyway cause it gets
your worst-case latency up by up to 9 ms or something.
> debugging instead that's a different matter. I'm sure it's not hard to
Debugging / profiling is the only real use for the cycle counter right now.
Philipp Rumpf
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/