Re: Hurd Latencies -- Questions...
Brent Fulgham <bfulgham@xpsystems.com> writes:
> I was wondering if any profiling had been done to determine
> where the time is being spent. Is it due to the overhead
> of RPC? Is it because of poor swap memory allocation?
> Is it something else entirely?
There are several things.
First, no serious time has been spent in profiling. It would be
helpful to profile some of the more important servers, the C library,
and the kernel. Doing the C library and the kernel is a little
trickier than just doing gprof, but the servers should be gprofable
without much effort. Then we can know where to look.
Mach VM might be behaving pessimally, and we really don't know yet.
Basically to find out we need to devise and implement appropriate
measurement tools.
Thomas
Reply to: