loggerd: add unit tests and fix identified sync issues

[deanlee]

resolve: #34839

This PR adds unit tests specifically aimed at resolving synchronization issues between encoderd and loggerd. While existing tests thoroughly cover logger rotation, these new test cases address the previously untested logic for synchronizing encoder segment numbers with logger segment numbers.

Key Changes:

1. Refactored the code to make unit testing feasible. Due to the original design, it was nearly impossible to unit test the synchronization logic. To enable testing, LoggerdState and RemoteEncoder classes were moved to loggerd.h.
2. The synchronization logic has been encapsulated into a new method RemoteEncoder::syncSegment, which can now be unit tested.

Future Improvements:
As part of future work, loggerd.h can be split to separate configuration and logic. For instance, renaming loggerd.h to encoder_config.h and limiting loggerd.h to only contain declarations used by loggerd.cc. However, since this would introduce substantial changes, it will be addressed in a follow-up PR.

[deanlee]

Unit tests revealed three flawed assumptions in the encoder-logger synchronization logic, leading to synchronization failures and continuous queue growth, which ultimately exhausted memory (these issues will be fixed later in this PR.):

1. Incorrect Assumption: The encoder’s first packet always starts at segment 0
   Problem: The code assumes the encoder’s first packet is aligned with segment 0, resulting in incorrect segment offset calculations.
   Location:

   openpilot/system/loggerd/loggerd.cc

   Line 131 in dc9b5b7

   re.encoderd_segment_offset = idx.getSegmentNum();

   
   Fix: Dynamically adjust the encoder segment offset based on the actual segment number:

   re.encoderd_segment_offset = idx.getSegmentNum() - s.logger.segment();

2. Incorrect Assumption: loggerd's segment always starts at 0 when receiving the encoder stream

   Problem: The logic assumes that loggerd starts at segment 0 when it begins receiving the encoder stream, which causes segment tracking to be misaligned.
   Location:

   openpilot/system/loggerd/loggerd.cc

   Lines 172 to 175 in dc9b5b7

   	name.c_str(), idx.getSegmentNum(), s->logger.segment(), re.encoderd_segment_offset);
   	// free the message, it's useless. this should never happen
   	// actually, this can happen if you restart encoderd
   	re.encoderd_segment_offset = -s->logger.segment();

   
   Fix: Adjust the offset by referencing loggerd's current segment:

   re.encoderd_segment_offset = idx.getSegmentNum() - s->logger.segment();

3. Incorrect Assumption: Encoder segment numbers always increment sequentially

   Problem: The code assumes the encoder’s segment numbers increment sequentially by 1 (e.g., 5 to 6). If the encoder skips segments (e.g., 5 to 7), this creates a gap larger than 1 between the adjusted encoder segment and loggerd's segment, causing synchronization to fail.
   Location:

   openpilot/system/loggerd/loggerd.cc

   Lines 159 to 169 in dc9b5b7

   	} else if (offset_segment_num > s->logger.segment()) {
   	// encoderd packet has a newer segment, this means encoderd has rolled over
   	if (!re.marked_ready_to_rotate) {
   	re.marked_ready_to_rotate = true;
   	++s->ready_to_rotate;
   	LOGD("rotate %d -> %d ready %d/%d for %s",
   	s->logger.segment(), offset_segment_num,
   	s->ready_to_rotate.load(), s->max_waiting, name.c_str());
   	}
   	// queue up all the new segment messages, they go in after the rotate
   	re.q.push_back(msg);

   
   Fix: Update the logic to account for non-sequential segment jumps and maintain synchronization.