On Mon, Nov 08, 2010 at 03:09:57PM +0000, David L. Mills wrote: > It is not an issue of limiting the slew rate to less than > 500 PPM, as that is in fact the result shown in the looopstats > trace.
No, it really is about the 500ppm limit. In loopstats is only the measured clock frequency, it doesn't contain the phase adjustments which are included in the adjtime() argument. If you add a printf call to see what is passed to adjtime() and run the test, you'll see values over 3 milliseconds. On Solaris, which apparently slews at 63000 ppm, this doesn't make any problems as it will always finish the adjustment in one second and there is no leftover. > It seems at least some kernels do not forget past programmed > offsets when presented with new ones. For that reason if no other, > the mission to measure the intrinsic clock offset with large initial > offsets is dead in the water. Which kernels? I'm pretty sure it's not the case with Linux. -- Miroslav Lichvar _______________________________________________ questions mailing list questions@lists.ntp.org http://lists.ntp.org/listinfo/questions