Cliff Spradlin wrote: > If you're using ptp4l as the grandmaster, it has a hardcoded UTC/TAI > offset (which can be overridden via a config option). This offset > doesn't automatically change after leap seconds are applied to the > Linux system clock, and ptp4l on the grandmaster does not query the > kernel for the offset between UTC and TAI. If ptp4l is the > grandmaster, the offset will never change unless you change it via > pmc or restarting ptp4l with new options.
What happens at the moment that the leap second is supposed to occur? Is leap seconds not something which is handled gracefully by PTP? With NTP for example the NTP clients know that a leap second will be inserted (or theoretically removed), and when it should occur, and modify the clocks accordingly at the correct time. Is that not something which ptp4l can handle? I guess phc2sys would be the agent to handle that, since the ptp protocol just transmits a monotonic clock, correct? -- Chris Caudle _______________________________________________ Linuxptp-users mailing list Linuxptp-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/linuxptp-users
