[LinuxPPS] ntp-dev-4.2.5p237-RC does sync with PPS for me

Rodolfo Giometti giometti at enneenne.com
Wed Oct 28 11:32:01 CET 2009


On Wed, Oct 28, 2009 at 03:37:37AM -0600, clemens at dwf.com wrote:
> 
> The patch I received from Rodolfo solves the problem I was having with
> having set ASSERT, and receiving both ASSERT and CLEAR events.

Ok, so I can send the patch to the LKML? Can I add the line:

   Tested-by: Reg Clemens <reg at dwf.com>

?

> I have noticed two other small problems that do not affect the
> running of ntp with the ONCORE driver.
> 
> ---
> 
> If one process has set some options with time_pps_setparams, and then
> another process tries to change them, this does not seem to work.
> I would expect that if ntp sets (only) ASSERT, and then ppsapitest
> sets both ASSERT and CLEAR, that BOTH programs should be seeing
> BOTH ASSERT and CLEAR events.
> 
> In fact they BOTH see the ASSERT events that ntp set (first).
> Its as if this stuff is set once, and never changed.
> 
> For practical purposes, this is not a problem, just something I see
> when running ntp, and trying to run my ppsapitest program.

Can you please report this issue for the LinuxPPS wiki as «know bugs»
section to Udo van den Heuvel <udovdh at xs4all.nl>?

> ---
> 
> A second problem is that when I do a getcap in a first process, I see 1133, 
> fine.
> I set the mode 1003 with setparams.
> and a getparams returns 0x33.
> 
> HOWEVER, if I do a setparam of 1011 in ntp (first process)
> and then do a setparam of 1003 in ppsapitest (second process)
> I see a getparams of 1111.
> 
> Now two things, first I would expect to see the 0x1000 bit set in the
> first getparams.
> 
> And more important I never set the 0x30 bits, so they should be off.
> 
> On the 2nd processes getparams I have set ASSERT/CLEAR (0x3) but dont
> see it, and the offset bit is again on.
> 
> Bottom line, something is wrong here.

Can you please provide some code showing what exacly are you doing?

> ---
> 
> Neither of these affects the running of ntpd with the ONCORE driver, 
> but they are both nits that should be fixed.

I agree...

Ciao,

Rodolfo

-- 

GNU/Linux Solutions                  e-mail: giometti at enneenne.com
Linux Device Driver                          giometti at linux.it
Embedded Systems                     phone:  +39 349 2432127
UNIX programming                     skype:  rodolfo.giometti
Freelance ICT Italia - Consulente ICT Italia - www.consulenti-ict.it



More information about the LinuxPPS mailing list