[LinuxPPS] RFC considerations [was: Measuring interrupt latency with NTP]

[Rodolfo Giometti]

On Sun, Aug 13, 2006 at 05:50:37PM -0400, linux at horizon.com wrote:
> 
> The one occupying the two lines above the call to time_pps_findsource(),
> that is, lines 182-183 of the post-patch refclock_atom.c, which I cut &
> pasted from 
> 
> http://wiki.enneenne.com/index.php/LinuxPPS_support#How_to_modify_a_refclock_to_work_with_LinuxPPS
> 
> The one that I changed in my revised patch, below, to say
> 
> +	/* Find the source corresponding to the requested unit number. */
> +	/* See /proc/pps/sources for a list of the possibilities. */
> +	up->fddev = time_pps_findsource(unit, device, sizeof device,
> +		id, sizeof id);

Ah! Ok. :)

> That's fine.  But the RFC says that <sys/timepps.h> MUST define the constant
> PPS_API_VERS_1, even if the implementation doesn't support it.  The
> current linuxpps <sys/timepps.h> doesn't define it.  Ergo, it's in
> violation of RFC2781.

But doing like this is useless... if the define PPS_API_VERS_1 specify
that current code is API version 1 and if I MUST specify it in a
future release, i.e., how I can use it to distinguish between version
1 and, i.e., version 2?

LinuxPPS is not "properly compatible" with version 1 of the API so, I
think, is correct that code that uses the define PPS_API_VERS_1 return
an error.

Ciao,

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