[ntp:questions] NTP and Leap Second testing (update)

Dave Hart hart at ntp.org
Wed May 23 22:01:51 UTC 2012


On Wed, May 23, 2012 at 8:10 PM, Phil Fisher <Phil.Fisher at ipaccess.com> wrote:
> Now, having looked at the kernel code, I understand why this can occur
> concerned about why it is propagated into the second day at all.  Why is
> the LI from ntpd not cleared once the insertion time is passed?

Based on your earlier testing, where we saw that once you manually
forced on the kernel leap bit, it was not cleared by subsequent
updates from ntpd via adjtimex/ntp_adjtime, I'm guessing the kernel in
question is failing to clear the leap indication despite ntpd having
done so.

If this reproduces with 4.2.6p5 or the latest 4.2.7, I'd love to fix
it.  There may be a need to specially handle the case where upstream
LI is seen but UTC midnight has passed since the sample in question.

Cheers,
Dave Hart


More information about the questions mailing list