[LinuxPPS] Debugging the pps

[Marc Leclerc]

Hi,

When the PPS system starts, how does it goes about finding PPS sources? 
We have gone right to the soc (vortex86DX) DCD pin for a signal and its 
there so the interrupt on IRQ3 (ttyS1) is getting there and is shown 
incrementing in /proc/interrupt. but beside the messages at boot there 
is no sign at all of the subsystem. even with debugging enabled... 
should there not be at least a log entry pass boot that its trying 
something? I had that working on our old board ok (arm based with a GPIO 
driver) but on this one I am clueless as of now

Thanks


On 2011-04-27 17:10, Marc Leclerc wrote:
> Hi,
>
> I did the check and we do get the interrupt on IRQ3 every second with
> the Rx/Tx deconnected from the gps. still no log pass the PPS line
> discipline registered. still no entry in /sys/class/pps/.
>
> I would hope that with debugging output enabled there would be sing of
> life... or death.
>
> Any other suggestions or next step woudl be appreciated
>
> Thanks
>
> On 04/26/2011 07:16 PM, Bernhard Schiffner wrote:
>> Am Dienstag, 26. April 2011, 21:25:16 schrieb Marc Leclerc:
>>> Hi,
>>>
>>> I have moved to a new board x86 based and compiled kernel 2.6.34.7 with
>>> pps support + line discipline. I have a gps module hooked to ttyS1 + the
>>> pps to pin 1 DCD. We checked the pps signal with a scope and it does get
>>> to the pin.
>> First: Is this real RS232 level or some "pseudo" (5V, 3.3V)?
>> Second: Is the pulse long enough to be catched by interrupt controllers
>> (>70µs)?
>>
>> Next: A 8250 (or similar) interrupts on DCD-change. Do you see "advance" of
>> the according interrupt in /proc/interrupts while RxD is disconnected?
>> ( pps / line_discipline is a level above that.)
>>
>>> when the kernel boots I do see the message that pps line
>>> discipline is registered. I did create the pps device + links for ntp.
>>> as for the gps it does work ok with ntp.
>> So this part seems ok.
>>
>>> problem is there is no sign of
>>> pps at all. the /sys/class/pps is empty and there are no messages
>>> appering in the logs (dmesg) that the pps is doing anything beside being
>>> registered.
>> What hardware / kernelmodul (*.ko) do you speak about?
>> I'am not quite sure, if pps is implemented for all "freaky" serial drivers.
>> 8250/16550 should work.
>>
>>> I have also enabled debugging support but no further sign.
>>> Also tried the patch for kernel consumer. Any clue on the steps I could
>>> take to find what's wrong.
>>>
>>> Regards
>>>
>> Hope it helps to make you sure about "hardware" first.
>>
>>
>> Bernhard
>>
>> _______________________________________________
>> LinuxPPS mailing list
>> LinuxPPS at ml.enneenne.com
>> http://ml.enneenne.com/cgi-bin/mailman/listinfo/linuxpps
>> Wiki: http://wiki.enneenne.com/index.php/LinuxPPS_support
>>
>>
>
> _______________________________________________
> LinuxPPS mailing list
> LinuxPPS at ml.enneenne.com
> http://ml.enneenne.com/cgi-bin/mailman/listinfo/linuxpps
> Wiki: http://wiki.enneenne.com/index.php/LinuxPPS_support
>
>
>