Re: Help needed to configure and compile gpsd to run on a Raspberry Pi 4 and NOT use systemd

[Gary E. Miller]

Yo Mick!

On Mon, 18 Nov 2024 20:06:39 +0000
Mick Durkin <mickdurkinuk@gmail.com> wrote:

> Here is a fresh attempt.

DOn't give up you are almost there.
 
> It takes a bit of time to sort as my son (Patrick) is in Berlin but I
> am in the UK.

I'm in USA, and my son is in Australia, so I got you beat there.

> We downloaded a fresh clone from gitlab then rebooted the Pi.
> Immediately after the reboot we ran "dmesg".

This is most of it:

[    0.000000] Kernel command line: coherent_pool=1M 8250.nr_uarts=1 
snd_bcm2835.enable_headphones=0 cgroup_disable=memory numa_policy=interleave 
snd_bcm2835.enable_headphones=1 snd_bcm2835.enable_hdmi=1 
snd_bcm2835.enable_hdmi=0  smsc95xx.macaddr=D8:3A:DD:B0:4A:E2 
vc_mem.mem_base=0x3ec00000 vc_mem.mem_size=0x40000000  console=tty1 
root=PARTUUID=aecb6dde-02 rootfstype=ext4 fsck.repair=yes rootwait

I see console=serial is not there.  God.

> We then printed "/boot/firmmware/config.txt",

I see your config.txt does not have all config to free up the serial
port.  But the serial port does seem to be working for you.  Odd.

I see your PPS is on pin 18. 

    [all]
    #GPS/PPS
    dtoverlay=pps-gpio,gpiopin=18
    enable_uart=1

Can you confirm that GPIO 18 is the PPS pin you used?  And confirm the
voltages on TX/RX and 18?  The Pi uses 3.3V, not 5V, and you can damage the
GPIO if you use 5V.

The dtoverlay is what loads the pps gpio driver, and dmesg shows that here:

[    5.532329] pps pps0: new PPS source pps@12.-1
[    5.532408] pps pps0: Registered IRQ 39 as PPS source

So now you GPIO 18 is connected to /dev/pps0

The enable_uart=1 is a new one, looks like it just enables the old behavior,
so that is good.

> followed by running "scons"  and "scons install".

Which look good, same comments on them as before.

> We ran the daemon, as requested using "gpsd -n -D 4 /dev/ttyS0", but
> no output was generated. I wonder if the output was going to somewhere
> other than the ssh session we were using?

The output went into your syslog.  To get the output to the console, add
the -N (no daemonize) option to gpsd.

> We then ran "gpsdebuginfo".

Wich looks good.  Except the pps part.  Since your system uses /dev/ttyS0
instead of /dev/ttyAMA0, the Magic Hat detection is broken.  And the
kernel is telling you that PPS for /ddev/ttyS0 is on /dev/pps1, but it
is not.  It is on /dev/pps0, because you did teh dtoverlay thing.

Never seen that odd behavior before.

You might try being explicit about the PPS device:

"gpsd -n -N -D 4 /dev/ttyS0 /dev/pps0"

You might also try: "ppscheck /dev/pps0"

RGDS
GARY
---------------------------------------------------------------------------
Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97703
        gem@rellim.com  Tel:+1 541 382 8588

            Veritas liberabit vos. -- Quid est veritas?
    "If you can't measure it, you can't improve it." - Lord Kelvin