perf(nebula): optimize AT128 decoder performance

[mojomex]

PR Type

• Improvement

Related Links

--

Description

This PR improves performance of the Hesai AT128 decoder by approx. 23% without changing the logical output.
To achieve this, three changes have been made:

• Decoding points directly to scan_pc_ and overflow_pc_ instead of temporary buffers and re-using the buffers (~ 11%point improvement)
• Removing unused copy operation for incoming scan packets (~ 9%point improvement)
• Combining azimuth and elevation sin/cos lookup tables as they were exact duplicates (~ 4%point improvement)

Points were decoded in convert(block_id) to a temporary PointCloud, the contents of which were then copied to scan_pc_ or overflow_pc_ respectively. This caused avoidable latency. Now, depending on whether the current block being decoded is in a new frame/scan or not, convert(block_id, pcl) decodes those points directly into the right buffer.

Further, overflow_pc_ was reallocated on each full scan, causing avoidable heap allocations.
Because each incoming PandarScan message always contains a full scan (albeit not necessarily aligned to our phase setting), scan_decoder->get_pointcloud() will be called exactly once in ConvertScanToPointCloud. Thus, it is guaranteed that one of scan_pc_ and overflow_pc_ can always be used for decoding, and the other as the output buffer.
Once a scan is complete, the buffers are swapped and the overflow_pc_ is cleared (not reallocated).

No changes have been made to the base class, the new function is an overloaded version of the existing one (which has been changed to throw a "not implemented" error).

	Before	Optimized
duration AVG	7.972725 ms	6.147384 ms
duration STD	0.759695	0.613904
duration AVG % rel. to before	100.000000	77.105188
duration STD % rel. to before	100.000000	80.809372

All measurements have been made on an Intel Core i7-12700H with

• TurboBoost disabled
• All cores set to 2.3 GHz
• Thread pair 2/3 isolated
• Nebula taskset to core 2

Measurements were repeated 3 times and ran for 37s each (~370 iterations).

Review Procedure

To confirm logical equality with the previous implementation, a rosbag of recorded AT128 packets can be used.
To confirm performance improvement, I provide the scripts and logging that yielded the above charts:

scripts/
    test_runner.bash  # runs the driver for 3 iterations on a given rosbag
    plot_times.py     # takes the logs from test_runner.bash and creates a plot like the one above

The test runner locks/unlocks core frequencies and tasksets nebula onto the specified core(s). Ideally, the taskset cores are isolated (cf. Autoware performance measurement guide)

Remarks

Pre-Review Checklist for the PR Author

PR Author should check the checkboxes below when creating the PR.

• Assign PR to reviewer

Checklist for the PR Reviewer

Reviewers should check the checkboxes below before approval.

• Commits are properly organized and messages are according to the guideline
• (Optional) Unit tests have been written for new behavior
• PR title describes the changes

Post-Review Checklist for the PR Author

PR Author should check the checkboxes below before merging.

• All open points are addressed and tracked via issues or tickets

CI Checks

• Build and test for PR: Required to pass before the merge.

[codecov-commenter]

Codecov Report

Patch coverage: 57.14% and project coverage change: +8.17% 🎉

Comparison is base (cb25896) 13.38% compared to head (e591f46) 21.55%.

❗ Your organization is not using the GitHub App Integration. As a result you may experience degraded service beginning May 15th. Please install the Github App Integration for your organization. Read more.

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #37      +/-   ##
==========================================
+ Coverage   13.38%   21.55%   +8.17%     
==========================================
  Files         111       48      -63     
  Lines       10991     6313    -4678     
  Branches     1727     1664      -63     
==========================================
- Hits         1471     1361     -110     
+ Misses       8339     3800    -4539     
+ Partials     1181     1152      -29     

Flag	Coverage Δ
differential	21.55% <57.14%> (?)
total	?

Flags with carried forward coverage won't be shown. Click here to find out more.

Files Changed	Coverage Δ
nebula_ros/src/hesai/hesai_decoder_ros_wrapper.cpp	0.00% <0.00%> (ø)
...bula_decoders_hesai/decoders/pandar_at_decoder.cpp	69.04% <66.66%> (+0.65%)	⬆️

... and 63 files with indirect coverage changes

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[amc-nu]

Other than my comment above, the PR works great!
Thank you. @mojomex
cc @drwnz

[amc-nu]

@mojomex can you please attend to the spelling errors on the spell-check job?

[mojomex]

@amc-nu Most of the warnings were the names of Bash commands, added them to the dictionary. stds for standard deviations is also commonly used in Python/Pandas, added to the dictionary as well.