GPS support in firmware

[MoridinBG]

This PR adds support for GPS tracking.
I am using a Quectel LC76G over UART with NMEA0183 messages as the concrete hardware implementation

The flow is as follows:

• Start a recording
  • If there is GPS configured and available, it checks for a fix
    • If no good fix found, wait for it, shows GPS... on the screen with num of satelites and EPE (estimated position error) underneath. "Good" fix means >= 6 satelites visible, 3D error < 6m, 10 consecutive good samples
      • If Left is pressed again, while waiting for a fix, starts a session without GPS and powers it down
    • If good fix found, shows GPS OK and waits for Left to start recording.
  • If no GPS found or configured, star recording normally.

It disables stdio over uart, as it uses uart0 over GPO 0 & 1, but with stdio over usb that shouldn't be too big of a deal.

I have source dropped a fork of lwgps for parsing & validating NMEA messages while under dev, but ideally it would be a submodule.

[MoridinBG]

Initial testing looks promising. I have only walked around the house, but compared to my Garmin Enduro 3, the track appears to be much closer to the actual path taken. Better than iPhone 15 Pro as well.

Using an inline USB power meter, I am getting 0.1W/0.02A consumption while idle, both with GPS connected and idle and with power lines disconnected.
With GPS active and tracking it goes up to 0.06A/0.26-0.31W when all constellations are enabled (GPS, GLONASS, Galilelo, BeiDou) and down to 0.04A/0.21-0.26W with only GPS enabled. 1Hz vs 10Hz doesn't seem to make a difference at the (low) resolution my power meter has.

I have added a tool at test_utils/sst_to_gpx.py that will create a GPX file out of a SST file with GPS chunks.