github-actions[bot] commented on code in PR #67119:
URL: https://github.com/apache/doris/pull/67119#discussion_r3868408380


##########
.github/workflows/checkstyle.yaml:
##########
@@ -26,6 +26,9 @@ jobs:
   java-checkstyle:
     name: "CheckStyle"
     runs-on: ubuntu-latest
+    concurrency:
+      group: fe-code-style-${{ github.event.pull_request.number || 
github.run_id }}

Review Comment:
   [P1] Isolate fresh events by immutable validation target
   
   This is distinct from the historical-rerun thread and the withdrawn Code 
Review Runner case: both events here can have `run_attempt == 1`. GitHub orders 
a concurrency group by when each job starts waiting, not workflow dispatch 
time, and says actual ordering is not guaranteed. If newer head B enters this 
PR group first and delayed older head A enters second, A cancels/replaces B 
while remaining bound to A's old event/checkout; B is left without a successful 
CheckStyle run. The same current-target validation gap exists in the new Clang 
Formatter, Gitleaks, and License Check groups (Gitleaks explicitly scans the 
old payload's `HEAD_SHA`). Please prevent different immutable targets from 
sharing a cancellation group—for example, include the event's `github.sha` so 
only exact-target duplicates/reruns coalesce—or use genuinely atomic monotonic 
ownership if cross-target reclamation is required. A one-time live-head 
prerequisite is still racy because its result can go stale before gro
 up admission. See GitHub's [concurrency 
ordering](https://docs.github.com/en/actions/how-tos/write-workflows/choose-when-workflows-run/control-workflow-concurrency).



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to