[LinuxPPS] [PATCHv3 12/16] ntp: add hardpps implementation

[Felix Joussein]

Update:

Even without that flag3  the bouncing occurs:

ntptime:

ntp_gettime() returns code 0 (OK)
   time d05569c2.5be13294  Tue, Oct  5 2010  9:09:22.358, (.358905736),
   maximum error 73807 us, estimated error 10381 us
ntp_adjtime() returns code 0 (OK)
   modes 0x0 (),
   offset 360.614 us, frequency -36.293 ppm, interval 4 s,
   maximum error 73807 us, estimated error 10381 us,
   status 0x2001 (PLL,NANO),
   time constant 4, precision 0.001 us, tolerance 500 ppm,
   pps frequency 0.000 ppm, stability 0.000 ppm, jitter 0.000 us,
   intervals 0, jitter exceeded 0, stability exceeded 0, errors 0.

Felix



Am 2010-10-04 18:02, schrieb Reg Clemens:
> Just a question about your compiles w/o the pps.h .
>
> All of the programs
> 	ppsapitest8.c (the source I gave you)
> 	refclock_atom.c
> 	refclock_jupiter.c
> 	refclock_mx4200.c
> 	refclock_nmea.c
> 	refclock_oncore.c
> 	refclock_parse.c
> 	refclock_ripencc.c
> 	refclock_wwvb.c
>
> have an include ppsapi_timepps.h in them.
> This is just a silly stack of #ifdef's to look several places for the actual
> timepps.h .
>
> But then timepps.h includes pps.h from /usr/include/linux/pps.h .
>
> So, I dont understand how you could have had a clean compile of either my
> ppsapitest8.c
> nor atom.c / parse.c without it.
>
> The 'new' pps.h adds a define for PPS_KC_BIND , which is needed in the
> static __inline at the very end of timepps.h .
>
> Just a question, not a solution to your problem.
>
>