Logging last 30 seconds of data using file ping pong buffer

[JustinDMorrison]

Summary

• Check for the previous max boot number used, if there is a previous use boot number + 1 to not overwrite data
• Create first logging file and set it to active file pointer
• Store a base time based on the last modification time of the first logging file (same as creation time)
• Enter infinite loop

Idle/AirborneState

• Get current time and check if the difference between the 2 times is greater than 30 seconds
  • If difference greater than 30 seconds, create a second logging file and set it as the active file pointer
    • Set base time as the the current time
• Write data to the active file

Landed State

• Open extraction file
• Copy over data from active file to extraction file
• Set flight state to Idle
• Teardown (close all files)
• Exit thread

[JustinDMorrison]

There are a few things left to do or check, indicated by TODOs in the code. I also think error handling could be improved a bit, right now it's mostly just printing if the debug option is defined.

[linguini1]

Can you resolve the merge conflicts and work with @AngusJull to test this on Josh?

[AngusJull]

I want to guess there's a way to work in a more general way to write this func where you just pass active_storage_file and alternate_storage_file (or whatever name you want). Then if the alternate storage file is NULL open it. Then if we ever want to change how the rest of our code works this has a better chance of still working as needed

[AngusJull]

Looking good! Logger is getting pretty big now so if you see blocks of code with a singular purpose you could potentially make them into functions. Like if we factored things out enough, the high-level overview would kind of be spelled out through the names of the functions we're calling in logger's body.

We can give it a try on Josh, but I forget if you had the simulator set up. If not, we could try running the code on that and see if the buffers fill up and overwrites happen as expected. We can even use GDB to check when the ping pong buffers get switched to verify it happens when we expect it to. So far in my testing I've found the simulator to be accurate to NuttX on Josh so if it works on there, we can be confident it'll do the same on the real system.