Re: [gpsd-dev] [ATOM (PPS) Refclock] CAPTURE ASSERT & CLEAR

[Gary E. Miller]

Yo f.hiesinger!

On Fri, 17 Apr 2015 10:31:58 +0200
"f.hiesinger" <address@hidden> wrote:

>  Hi Garry Miller,
>  while searching for a solution of my problem, I stumbled over your 
>  recent discussion (https://lkml.org/lkml/2015/4/2/700).

I wish more people stumbed over it.  I've had no luck getting it
pulled into the kernel.  Maybe you can post a reply saying
you approve of the patch?

>  I'm also working on the pps-gpio module and funnily I put the same
> Code in my patch, that you suggested above (see: 
>  http://www.linuxpps.org/pipermail/discussions/2015-April/004470.html)

Cool.

>  I don't know, what exactly you are working on, but maybe you got any 
>  advices for my problem:

I am making sure that gpsd (http://catb.org/gpsd) works with the Adafruit
Ultimate GPS HAT on the Raspberry Pi B+ and 2B running Debian Wheezy.

>  I'm sourcing a pps-signal via GPIO in a raspberry Pi. The pps0 is 
>  configured to capture "Assert" AND "Clear" (mode = 0x11f3). The
> internal echo-mechanism is used, so reception of both edges is
> required. Up to here this works fine.

I have never tried the echo mechanism.  Let me know how you use that
and why.

>  When ntp (incl. ATOM Driver 22) is starting, it instantly prevents
> the reception of either "asserting" or "clearing". fudge -> flag2 1
> inverts the behavior. Any suggestions, how to solve this?

Yeah, use gpsd to read the PPS and feed to ntpd.  Ditch ntpd totally
for chronyd if you can.

>  PS: Which maillinglist is appropriate for this? I didn't find one in
> a trice.

Cc: gpsd-dev

RGDS
GARY
---------------------------------------------------------------------------
Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97701
        address@hidden  Tel:+1(541)382-8588