[Linuxpps] Duplicated sequence numbers

Rodolfo Giometti giometti at linux.it
Tue May 30 00:02:56 CEST 2006


On Mon, May 29, 2006 at 10:37:12PM +0100, Adam James wrote:
> 
> I have a Garmin GPS25-LVC configured to output a 100ms PPS pulse and
> the standard NMEA sentences. Rodolfo's LinuxPPS implementation seemed
> much neater than the SHMPPS driver I am using at the moment, so I
> thought I'd give it a try. 

Thanks! However LinuxPPS has to be well tested since is still a at
developing stage.

> While it does seem to be working, the jitter and offset are much
> higher than when I was running the SHMPPS driver. Can anyone show me
> some graphs or `ntpq -p' output to show what they have achieved using
> the current implentation?

Here I have to ask some help to other people reading this ML since,
currently, I have no hardware to test it.

> Perhaps you could also advise as to whether the following ppstest
> output is a bug or a problem with my system (notice the duplicated
> sequence numbers):
> 
> # ./ppstest
> found PPS source #0 "pps_8250_0" on "/dev/ttyS0"
> Assert timestamp: 1148936547.001043675, sequence: 2285
> Assert timestamp: 1148936547.001043675, sequence: 2285
> Assert timestamp: 1148936548.001042675, sequence: 2286
> Assert timestamp: 1148936548.001042675, sequence: 2286
> Assert timestamp: 1148936549.001045675, sequence: 2287
> Assert timestamp: 1148936549.001045675, sequence: 2287
> 
> I'm running a patched 2.6.16.7 kernel and ntpd
> 4.2.0a at 1:4.2.0a+stable-8-r (patches taken from
> http://time.qnan.org/linux2.6-pps/).

Yes, the ppstest.c is a bit old and it could be fixed. If you read the code you'll see that
the printf() show only assert events:

                printf("Assert timestamp: %d.%09d, sequence: %ld\n",
                        infobuf.assert_timestamp.tv_sec,
                        infobuf.assert_timestamp.tv_nsec,
                        infobuf.assert_sequence);

So this program is not the best to test LinuxPPS... please, try to use
the procfs or enable the kernel debugging kernel messages to verify
that the system really gets assert and clear events.

> dmesg output:
> 
> May 29 21:10:40 evolution kernel: "pps_core": procfs enabled
> May 29 21:10:40 evolution kernel: "pps_core": LinuxPPS API ver. 1 registered May 29 21:10:40 evolution kernel: "pps_core": Software ver. 2.0.0 - Copyright 2005-2006 Rodolfo Giometti <giometti at linux.it>

As soon as you'll enable the kernel messages you should see the PPS
events here.

> Any advice would be appreciated.

Thanks for your help,

Rodolfo

-- 

GNU/Linux Solutions                  e-mail:    giometti at enneenne.com
Linux Device Driver                             giometti at gnudd.com
Embedded Systems                     		giometti at linux.it
UNIX programming                     phone:     +39 349 2432127
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 191 bytes
Desc: Digital signature
Url : http://ml.enneenne.com/pipermail/linuxpps/attachments/20060530/a7271e82/attachment.pgp


More information about the LinuxPPS mailing list