Repository navigation
docs(debug-journal): a replay cannot test a fix whose inputs are built in fetch_sensor_data() - #5321
Closed
chalfontchubby wants to merge 1 commit into
Closed
chalfontchubby wants to merge 1 commit into
chalfontchubby wants to merge 1 commit into
Conversation
…t in fetch_sensor_data() run_single_debug(redo=True) re-runs set_rate_thresholds() on the dump's rates but never calls fetch_sensor_data(), so state built there (e.g. PR #5163's saving-minute snapshot) is empty and fallbacks run silently. This produced two false "the fix doesn't work" results on GH#5221. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
chalfontchubby
added
the
documentation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Written by Claude, on behalf of @chalfontchubby.
Adds a note to
tools/debug-journal.md, the developer and triage notes; it is not part of the published docs.run_single_debug(..., redo=True)re-runsset_rate_thresholds()and therate_scan_window()rescan on the rates stored in the dump, but it never callsfetch_sensor_data(). So any state that fetch builds alongside the rates is missing. Code that quietly falls back when that state is absent then behaves exactly like the code without the fix, and the result reads as "the fix doesn't work".That happened twice on #5221 with #5163.
rate_minmax_excluding_saving()needsrate_export_saving_minutesandrate_export_pre_saving, and only fetch fills those in. On a replay it falls back to plainrate_minmax(), so the Axle-boosted dump still gave main's 20.5p export threshold instead of the 19.9p the fix produces.The note says to check that the replay log shows the value the fix should produce, and to prove the fix through the real fetch sequence or by rebuilding the missing state by hand.
🤖 Generated with Claude Code