[LinuxPPS] Kernel configuration and more ......

[Rodolfo Giometti]

On Mon, Oct 29, 2007 at 08:35:42PM +0000, Paul wrote:
> My GPS unit generates MNEA data and PPS output.
> I have checked that the PPS signal appears on the DCD pin (it causes the
> meter needle to kick every second).
> 
> The pulse from this unit is short (25 mS.), but this has worked on a PPS
> kernel from before the big changes (i.e 2.6.19 or so).
> 
> I know NMEA data is being generated because I see it using minicom.
> 
> I know the DCD pin is being pulsed because I use Statserial. This is a
> command line program which gives the levels on all the control pins for
> the serial port. So usually it gives a 0 on the DCD pin, but sometimes
> it gives a 1 (which is what you would expect from a short pulse.)
> 
> PPS and the 8250 are compiled in, so when I boot up the
> directory /sys/class/pps is generated, but is empty.
> 
> When I ppsctl the /dev/ttyS0, the files in /sys/class/pps/pps0 are
> generated, exactly as I would expect.
> # cat /sys/class/pps/pps0/assert
> 1193688471.903089607#599
> every time I try it
> which tells me the PPS pulses are not being registered

I not agree. I see you have got 599 events till now...

> The next bit is interesting
> ktimer is as a module so
> # modprobe ktimer
> new PPS source ktimer at ID 1
> ktimer PPS source registered at 1
> #    (now the data starts)
> PPS event at 19570298
> PPS event at 19570548
> PPS event at 19571048
> PPS event at 19571298
> PPS event at 19571548
> PPS event at 19571798
> and it just goes on and on
> 
> But
> # tree /sys/class/pps  (Difficult when the events keep writing on top)

Just use another shell... or give "echo 0 > /proc/sys/kernel/printk".

> shows another PPS, PPS1
> which looks like PPS0 except there is no device line (with only one
> serial line active in the BIOS)

ktimer is just a debugging module. A dummy PPS device (only software)
used just for debugging purpose.

> # rrmod ktimer (difficult when you can't see what you are typing.)
> stops the event output.

Right.

> I think the events are NMEA data as these come at about this rate.

Not from ktimer.

> What more can I try?

Post your kernel messages.

Post the commands you did and their output.

Post two ore three output of cat /proc/interrupts at 2 seconds delay
(after enabling the serial line and PPS pulse generation).

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