Tony Hoyle wrote:
> Udo van den Heuvel wrote:
> 
>> - A shared memory driver method (Address: 127.127.28.u), documented at
>> http://time.qnan.org/
>> - PPS Clock Discipline (Address: 127.127.22.u)
>> - Generic NMEA GPS Receiver (Address: 127.127.20.u)
>>
> 
> You'll need either a shared memory or PPS source because the generic
> source is only accurate to the time taken to receive/decode the serial
> stream, which is to say not very.

I see.

> I was unable to get PPS working on linux 2.6.x - the available driver
> has half of it ripped out (compared to the 2.4.x version) and has
> bitrotted somewhat anyway.. people have had some success with the shared
> memory route so I'd recommend that.

So far I was able to compile 2.6.16.11 with the LinuxPPS patch (had to
edit the source manually because of a few rejects) and the ppstest tool
appears to work OK with the ktimer test module. I could make a fresh
patch for those interested.
Can't say anything more about actual results since my GPS is still on
the way.


BTW: What is the standard for PPS signals? Rising or dropping edge?

Udo
_______________________________________________
timekeepers mailing list
[email protected]
https://fortytwo.ch/mailman/cgi-bin/listinfo/timekeepers

Reply via email to