[LinuxPPS] ntpd

[Udo van den Heuvel]

Folkert van Heusden wrote:
>>> [42996225.890000] drivers/pps/pps.c: New message from PID 16224 (flags 0)
>>> [42996225.890000] drivers/pps/pps.c: PPS_FETCH: source 1
>>> [42996225.910000] drivers/pps/kapi.c: capture clear seq #1812 for source 7
>>> [42996225.910000] drivers/serial/8250.c: serial8250: PPS clear event at 4655295
>>> [42996226.710000] drivers/pps/kapi.c: capture assert seq #1816 for source 7
>>> [42996226.710000] drivers/serial/8250.c: serial8250: PPS assert event at 4655375
>>> [42996226.890000] drivers/pps/pps.c: New message from PID 16224 (flags 0)
>>> [42996226.890000] drivers/pps/pps.c: PPS_FETCH: source 1
>>> [42996226.910000] drivers/pps/kapi.c: capture clear seq #1813 for source 7
>>> [42996226.910000] drivers/serial/8250.c: serial8250: PPS clear event at 4655395
>>> Ah ok, so it now works, right?
>> Yes! :)
> 
> I have a feeling that neither the atom nor the nmea ntpd driver use the
> PPS functionality.

About the nmea driver: read the docs, my dear friend:

http://www.eecis.udel.edu/~mills/ntp/html/drivers/driver20.html says:

If the Operating System supports the PPSAPI, RFC-2783, it will be used.


And how else can the NMEA driver reach such precision on my tiny
firewall when you got such bad performance with NMEA-only?