Logger

[celery6]

minor adjustments to hopefully increase certainty that logger won't overflow any buffers or get stuck.

• give back logwritemutex before exiting logGeneric (I thought josh fixed this already but I couldn't find that change?)
• get rid of counting the log header string separately from the log msg. It's all under MAX_MSG_LENGTH now for simplicity and less confusing/potentially incorrect logic.
• change fullBuffersQueue to length NUM_LOG_BUFFERS instead of NUM_LOG_BUFFERS - 1. reasoning described in comment
• replace stupid source enums with source strings

[celery6]

testing code:

• enabled controller, log, flightphase tasks.
• Added log calls in controller and flightphase.
• Added uart print in log task.

Verify that the logs are in fact printing

[Joe-Joe-Joe-Joe]

Nice I can probably find time to test this in person tmr

[celery6]

@Lucifersan just realized the test print uses uarttransmit not printf so this does actually print the entire buffer together.

This issue is actually just the hal_uart_transmit timeout of 500 ticks being too low since unsurprisingly it takes longer than 0.5s to transmit a book-length novel of logs :skull:

increased timeout to 3000 and it works as expected. (log buffer itself is actually fine.) we can ignore this issue

[celery6]

ill clean up the testing code then this should be good to merge!

[celery6]

removed testing code, also cleaned up commits. lgtm?

[Lucifersan]

mommy what. im gonna change that and test it lmao

[celery6]

wtf I literally built and it built how tf 😭😭