Potential bug in rangefinder_teraranger_evo.c

[mart1npetroff]

Current Behavior

Due to limited documentation regarding TeraRanger Evo rangefinder support in INAV, I reviewed the driver implementation to understand what is currently supported.

While inspecting the code (with some assistance from ChatGPT for static analysis), I noticed two potential issues in rangefinder_teraranger_evo.c.
Since I am not deeply familiar with all design decisions in the codebase, I would appreciate confirmation from devs (or @error414 ) whether these are real bugs or intentional behavior.

Potential Issue 1 – Zero-length buffer used with busWriteBuf()

uint8_t triggerValue[0];
busWriteBuf(rangefinder->busDev, TERARANGER_EVO_I2C_REGISTRY_TRIGGER_READING, triggerValue, 1);

A zero-length array is declared, but busWriteBuf() is instructed to send 1 byte.
This results in an out-of-bounds read and undefined behavior — effectively transmitting a random byte from memory over I²C.

Potential Issue 2 – Incorrect register argument in periodic trigger

In teraRangerUpdate():

busWriteBuf(rangefinder->busDev, TERARANGER_EVO_I2C_ADDRESS, triggerValue, 1);

The second parameter of busWriteBuf() is a register/command, not the I²C device address.
Here TERARANGER_EVO_I2C_ADDRESS (0x31) is used instead of TERARANGER_EVO_I2C_REGISTRY_TRIGGER_READING (0x00).
This may result in invalid command writes to the sensor.

Suggested solution(s)

For Issue 1:
Replace the zero-length array with a valid trigger byte:

static const uint8_t triggerValue = 0x00;
busWriteBuf(rangefinder->busDev, TERARANGER_EVO_I2C_REGISTRY_TRIGGER_READING, &triggerValue, 1);

For Issue 2:
Use the correct trigger register:

busWriteBuf(rangefinder->busDev, TERARANGER_EVO_I2C_REGISTRY_TRIGGER_READING, &triggerValue, 1);

Additional context

I would also appreciate confirmation from devs (or @error414 ) regarding which exact TeraRanger Evo models/firmware variants are officially supported by this driver (e.g. Evo 60m single-point with I²C/UART backboard).

Documentation around Evo support is currently sparse, so clarification would be very helpful for users.

[error414]

You are right, there are two potential bugs. I will look at that. I used Terra Ranger Evo 60m. this one https://www.mouser.ch/new/terabee/terabee-teraranger-evo/?srsltid=AfmBOorc79LKggYKtJqdwUkHHGgxTRMdLcWW4gqXu9w3ssbIArS90Wa5

I found datasheet: https://www.mouser.lt/datasheet/2/944/User_Manual_for_TeraRanger_Evo_single_point_distan-1729051.pdf?srsltid=AfmBOork4pNRMVMmnZ12zIk3yt79aeE066Hq5VpcBHv4wAnv-FSBLAKc
But it seems terra ranger Evo is no longer produced :(.

What model do you use?

I tested on plane only few times, I was not satisfied with performance, I use NRA15 radar now.
https://www.youtube.com/watch?v=3SvQbgpmemA

[mart1npetroff]

I’ve been looking at this sensor: https://tribotix.com/product/teraranger-evo-60m/

At the moment, I’m still in the research phase and searching for a high-quality, reliable sensor.

Is it possible that the performance issues you experienced were caused by the topic-related problems?

Are you satisfied with the NRA15?

[error414]

It's exactly sensor what I have. I don't think performance issue is caused by bugs in I2C. Even in document from terabee were test how reflections of surface can reduce max distance.

Mud, grass can reduce max distance to 1/2. Snow or water is no way, in the best case max 7m. NRA15 is much beter, max distance is 120m (if you use right FW) in all conditions. It's much better than any optical system. Just check if 24Ghz radar is allowed in your country, many old card uses 24Ghz radar so I would say it's ok in Europe.

Personally I would not recomend to buy any optical system to for fixed wing UAV.

BTW: if you would need more ditance you can buy NRA24 (200m). And biy UART version, not CAN version, only UART version is supported in Inav.

NRA15 on T2 Cruza: https://youtu.be/2juFhekefxI?si=Unw5hr_cGEgEKvka&t=82 number above "FLAPS" text

[error414]

Driver is fixed, but old one works as well.

[mart1npetroff]

I’m genuinely impressed by the speed of the response and the fix — thank you for addressing this so quickly 🙌

Also, thanks for the radar recommendations and the additional sensor guidance. We’ll move forward with the suggested setup (radar for high alt + optical flow for small altitudes and good POSH) and test it in VTOL.

Everything looks good from my side, so I’m closing this issue as resolved.

Thanks again for the great support and for maintaining INAV! 🚀