Repository navigation
DiskIO: lock the once-only failure log against sampler/marker race - #243
Merged
Merged
Conversation
Owner
Author
|
@greptile review |
Contributor
[Low impact] The PR appears safe to merge. SummaryProtects
Reviews (1) · Last reviewed commit: "DiskIO: lock the once-only failure log a..." · Reviewed by Greptile |
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Follow-up to #240 (merged at 5ebff91, before this fix landed).
Greptile round 2 flagged a P2:
logFailureOnceread/wroteloggedFailurewithout a lock, and the sampler callsDiskIO.sample()outside the trace lock, so it can race the stop marker's sample — worst case, two diagnosis lines inext-diag.loginstead of one. Fixed with a dedicatedNSLock(check-and-set inside, log call outside).Diff is 6 lines; full context and the 5/5 review it came from are on #240.