xiangfu0 commented on PR #17082:
URL: https://github.com/apache/pinot/pull/17082#issuecomment-5066608179

   Closing as superseded by #19014 ("Upgrade Pinot to JDK 25"), which landed on 
master after this PR was last rebased.
   
   **Why this PR is no longer applicable:**
   
   1. **CI matrix change is now backwards.** This PR adds JDK 25 to the matrix 
as an *experimental* entry (`java: [ 21, 25 ]` with `continue-on-error: ${{ 
matrix.java == 25 }}`). Master now runs `java: [ 25 ]` as the sole baseline — 
JDK 21 was dropped entirely. Merging this would re-add JDK 21 and downgrade JDK 
25 back to non-blocking.
   
   2. **The stated blocker is gone.** The `continue-on-error` existed solely 
because `datasketches-memory` 3.0.2 has a hardcoded JDK allowlist rejecting 25. 
#19014 bumped `datasketches-java` 6.2.0 -> 9.0.0, which drops 
`datasketches-memory` altogether in favor of `java.lang.foreign.MemorySegment`. 
No allowlist, no sketch test failures.
   
   3. **The predownload test fix already landed.** #19014 contains the same 
change — dropping the vestigial `System.setSecurityManager()` save/restore (it 
throws `UnsupportedOperationException` unconditionally under JEP 486) and 
restoring the default `ExitHelper` action in `finally` instead.
   
   **What is left over:** three files still carry dead 
SecurityManager/AccessController code — `NamedThreadFactory`, `CleanerUtil`, 
and `BenchmarkTextMatchQueriesSSQE`. These are not JDK 25 blockers:
   
   - `pinot-common` and `pinot-segment-spi` compile at `--release 11`, where 
`System.getSecurityManager()` and `AccessController.doPrivileged()` are not yet 
deprecated, so no build warning is emitted.
   - At runtime on JDK 25, `System.getSecurityManager()` always returns `null`, 
so the ternary already falls through to 
`Thread.currentThread().getThreadGroup()` — exactly what this PR hardcodes. 
`AccessController.doPrivileged` still exists and simply invokes the action.
   - `BenchmarkTextMatchQueriesSSQE` already carries 
`@SuppressWarnings("removal")`.
   
   That residual is worth doing as dead-code cleanup, but it is a small 
standalone change rather than a rebase of this branch, which currently 
conflicts with master.
   


-- 
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