[ntp:questions] Re: ntp servers reporting leap second erroneously?

David L. Mills mills at udel.edu
Wed Oct 19 13:44:38 UTC 2005


Thanks you for using plain ASCII, which I can read much more easily than 
the formatted version.

Let's face it, the version you are using has a bug, since corrected in 
ntp-dev. Dirty rotten gmtime() numbers the days of the month from zero, 
not one. I can verify here that the kernel does get the ntp_adjtime() 
update on the correct month and day and does the right thing. I don't 
have the test setup you have to actually verify the intended behavior. 
Please report.


Martin Burnicki wrote:
> Dave,
> David L. Mills wrote:
>>No fair on the bug, which has been there probably most of the seven
>>years since the last leap.
>>Can you tell your radio to step UTC at the same time the system kernel
>>implements it after being so advised by ntpd?
> Yes. The device is not really a radio clock but it behaves like one. 
> It starts to count time at a predefined date, then announces the leap second
> before it occurs, and inserts the leap second correctly.
>>I think you misunderstand how the kernel implements the leap. The ntpd
>>does not implement the leap; the kernel does. The kernel does not step
>>the clock. Please see the white paper "The NTP Timescale and Leap
>>Seconds" on the NTP project page for a detailed explanation.
> I haven't looked at the project pages right now, but as far as I remenber
> ntpd passes the leap second announcement to the kernel (if the kernel
> supports it, unlike e.g. Windows), and the kernel inserts the leap second
> at the right time. 
> This is how it has worked 5 days ago but didn't work now with the latest
> patches.
> Martin

More information about the questions mailing list