8355369: Remove setAccessible usage for setting final fields in java.util.concurrent #24821

viktorklang-ora · 2025-04-23T10:10:54Z

This Pull Request replaces the uses of Field + setAccessible to modify final fields in java.util.concurrent with Unsafe.

Progress

Change must be properly reviewed (1 review required, with at least 1 Reviewer)
Change must not contain extraneous whitespace
Commit message must refer to an issue

Issue

JDK-8355369: Remove setAccessible usage for setting final fields in java.util.concurrent (Enhancement - P4)

Reviewers

Per Minborg (@minborg - Reviewer)
Doug Lea (@DougLea - Reviewer)
Raffaello Giulietti (@rgiulietti - Reviewer)
Alan Bateman (@AlanBateman - Reviewer)

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.org/jdk.git pull/24821/head:pull/24821
$ git checkout pull/24821

Update a local copy of the PR:
$ git checkout pull/24821
$ git pull https://git.openjdk.org/jdk.git pull/24821/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 24821

View PR using the GUI difftool:
$ git pr show -t 24821

Using diff file

Download this PR as a diff file:
https://git.openjdk.org/jdk/pull/24821.diff

Using Webrev

Link to Webrev Comment

…ceArray to use Unsafe rather than Field.setAccessible

bridgekeeper · 2025-04-23T10:11:43Z

👋 Welcome back vklang! A progress list of the required criteria for merging this PR into master will be added to the body of your pull request. There are additional pull request commands available for use with this pull request.

openjdk · 2025-04-23T10:12:23Z

@viktorklang-ora This change now passes all automated pre-integration checks.

ℹ️ This project also has non-automated pre-integration requirements. Please see the file CONTRIBUTING.md for details.

After integration, the commit message for the final commit will be:

8355369: Remove setAccessible usage for setting final fields in java.util.concurrent

Reviewed-by: pminborg, dl, rgiulietti, alanb

You can use pull request commands such as /summary, /contributor and /issue to adjust it as needed.

At the time when this comment was updated there had been 40 new commits pushed to the master branch:

1ec6481: 8350582: Correct the parsing of the ssl value in javax.net.debug
d3f31b0: 8347719: [REDO] Portable implementation of FORBID_C_FUNCTION and ALLOW_C_FUNCTION
6254046: 8320909: C2: Adapt IGVN's enqueuing logic to match idealization of AndNode with LShift operand
... and 37 more: https://git.openjdk.org/jdk/compare/c2e90bcc8026fb2047b42deae6cdad738d6d01b8...master

As there are no conflicts, your changes will automatically be rebased on top of these commits when integrating. If you prefer to avoid this automatic rebasing, please check the documentation for the /integrate command for further details.

➡️ To integrate this PR with the above commit message to the master branch, type /integrate in a new comment.

openjdk · 2025-04-23T10:13:02Z

@viktorklang-ora The following label will be automatically applied to this pull request:

core-libs

When this pull request is ready to be reviewed, an "RFR" email will be sent to the corresponding mailing list. If you would like to change these labels, use the /label pull request command.

mlbridge · 2025-04-23T10:21:00Z

Webrevs

01: Full - Incremental (39dba5dc)
00: Full (e6e7ac01)

minborg · 2025-04-23T10:44:05Z

src/java.base/share/classes/java/util/concurrent/ConcurrentSkipListSet.java

+        final Unsafe U = Unsafe.getUnsafe();
+        U.putReference(
+            this,
+            U.objectFieldOffset(ConcurrentSkipListSet.class, "m"),


Would it make sense to compute the offset once and for all and put it in a static final long field?

Yeah, I originally did that, but the following patch is the "smallest change".
Given that the "original code" obtained the Field instance each call, this is still likely a performance improvement.

This code has always used getDeclaredField each time so if anyone had run into a performance issue then we should have heard by now. So I think keeping the changes simple and just moving to Unsafe is okay for now.

Fine. The main effect of this change is to nearly revert to the pre-VarHandle version of this code.

liach · 2025-04-23T14:06:27Z

Seems these are used mainly for cloning: is it possible for us to just use readResolve to resolve an alternative object?

viktorklang-ora · 2025-04-23T14:11:59Z

@liach Not sure I understand, could you elaborate a bit?

rgiulietti · 2025-04-23T14:11:59Z

I guess the intent is just to have the minimal patch to come away from setAccessible().

AlanBateman · 2025-04-23T14:14:11Z

I guess the intent is just to have the minimal patch to come away from setAccessible().

It's to move away from using Field.set to mutate non-static final fields. Viktor offered to get this done before the "prepare for final means final" changes.

liach · 2025-04-23T14:54:28Z

@liach Not sure I understand, could you elaborate a bit?

I mean that we call the setters in readObject methods. If we can instead call readResolve to provide safely constructed objects from constructors, that may be better, but since we are looking at marshalling guess we will not touch serialization.

rgiulietti · 2025-04-23T15:07:33Z

@viktorklang-ora Before approving, what is the status of the copyright notices? There seem to be no years to update...

viktorklang-ora · 2025-04-23T15:18:49Z

src/java.base/share/classes/java/util/concurrent/ConcurrentSkipListSet.java

                (ConcurrentSkipListSet<E>) super.clone();
            clone.setMap(new ConcurrentSkipListMap<E,Object>(m));
+            // Needed to ensure safe publication of setMap()
+            VarHandle.releaseFence();


@DougLea This might be interesting to you. 👍

I wonder if we could just change it to return new ConcurrentSkipListSet<>(m). COWAL could be changed to return a new object too.

The set only has a map field, and AbstractSet does not define any additional field. The map should be fine too - two fields in AbstractMap are cleared when cloning happens, so recreating a map from a constructor should have the same effect. (Note a significant field head is not cleared upon clone, but seems immediately replaced later in buildFromSorted). Both should still be fine with this new return values.

As others have mentioned, clone() needs to be re-checked wrt issuing final field fences. But can be conservatively done so here anyway.

@DougLea Yeah. I'd be surprised if clone()-invocations were specially tracked to get a trailing release fence inserted before handing out the instance.

viktorklang-ora · 2025-04-23T15:35:38Z

@rgiulietti These are the JSR166 files, so I didn't touch the copyright notices.

AlanBateman · 2025-04-23T15:37:54Z

Seems these are used mainly for cloning: is it possible for us to just use readResolve to resolve an alternative object?

Potentially AtomicReferenceArray.readObject could be replaced, the others are clone methods.

liach · 2025-04-23T16:23:00Z

Thanks. We can investigate how clone cleanups interact with truly final or strict finals in another issue.

viktorklang-ora · 2025-04-24T10:39:27Z

@liach @AlanBateman My proposal here would be to move forward with the proposed changes for now, and then once 25 is branched off there's an opportunity to take a broader look at clone().

minborg

I think these changes look good. Before we integrate, could you please raise an issue with the future improvements of clone() and friends so that we do not forget that?

DougLea · 2025-04-24T11:27:32Z

src/java.base/share/classes/java/util/concurrent/ConcurrentSkipListSet.java

                (ConcurrentSkipListSet<E>) super.clone();
            clone.setMap(new ConcurrentSkipListMap<E,Object>(m));
+            // Needed to ensure safe publication of setMap()
+            VarHandle.releaseFence();


As others have mentioned, clone() needs to be re-checked wrt issuing final field fences. But can be conservatively done so here anyway.

rgiulietti

LGTM

AlanBateman · 2025-04-24T12:30:30Z

@liach @AlanBateman My proposal here would be to move forward with the proposed changes for now, and then once 25 is branched off there's an opportunity to take a broader look at clone().

Okay with me. We can also add a non-public constructor to some of these classes and use it from readResolve so that we don't need readObject methods that mutate finals.

viktorklang-ora · 2025-04-24T14:12:25Z

/integrate

openjdk · 2025-04-24T14:14:26Z

Going to push as commit 356c4d9.
Since your change was applied there have been 43 commits pushed to the master branch:

cf96b10: 8354362: Use automatic indentation in CollectedHeap printing
0537c69: 8355492: MissedOptCastII is missing UnlockDiagnosticVMOptions flag
34c4d89: 8353190: Use "/native" Run Option for TestAvailableProcessors Execution
... and 40 more: https://git.openjdk.org/jdk/compare/c2e90bcc8026fb2047b42deae6cdad738d6d01b8...master

Your commit was automatically rebased without conflicts.

openjdk · 2025-04-24T14:14:34Z

@viktorklang-ora Pushed as commit 356c4d9.

💡 You may see a message that your pull request was closed with unmerged commits. This can be safely ignored.

Update ConcurrentSkipListSet, CopyOnWriteArrayList, and AtomicReferen…

e6e7ac0

…ceArray to use Unsafe rather than Field.setAccessible

openjdk bot added core-libs core-libs-dev@openjdk.org rfr Pull request is ready for review labels Apr 23, 2025

minborg reviewed Apr 23, 2025

View reviewed changes

Adding a releaseFence() to ConcurrentskipListSet.clone()

39dba5d

viktorklang-ora commented Apr 23, 2025

View reviewed changes

minborg approved these changes Apr 24, 2025

View reviewed changes

openjdk bot added the ready Pull request is ready to be integrated label Apr 24, 2025

DougLea approved these changes Apr 24, 2025

View reviewed changes

rgiulietti approved these changes Apr 24, 2025

View reviewed changes

AlanBateman approved these changes Apr 24, 2025

View reviewed changes

openjdk bot added the integrated Pull request has been integrated label Apr 24, 2025

openjdk bot closed this Apr 24, 2025

openjdk bot removed ready Pull request is ready to be integrated rfr Pull request is ready for review labels Apr 24, 2025

8355369: Remove setAccessible usage for setting final fields in java.util.concurrent #24821

8355369: Remove setAccessible usage for setting final fields in java.util.concurrent #24821

Uh oh!

Conversation

viktorklang-ora commented Apr 23, 2025 • edited by openjdk bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Progress

Issue

Reviewers

Reviewing

Uh oh!

bridgekeeper bot commented Apr 23, 2025

Uh oh!

openjdk bot commented Apr 23, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

openjdk bot commented Apr 23, 2025

Uh oh!

mlbridge bot commented Apr 23, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Webrevs

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

liach commented Apr 23, 2025

Uh oh!

viktorklang-ora commented Apr 23, 2025

Uh oh!

rgiulietti commented Apr 23, 2025

Uh oh!

AlanBateman commented Apr 23, 2025

Uh oh!

liach commented Apr 23, 2025

Uh oh!

rgiulietti commented Apr 23, 2025

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

viktorklang-ora commented Apr 23, 2025

Uh oh!

AlanBateman commented Apr 23, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

liach commented Apr 23, 2025

Uh oh!

viktorklang-ora commented Apr 24, 2025

Uh oh!

minborg left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

rgiulietti left a comment

Choose a reason for hiding this comment

Uh oh!

AlanBateman commented Apr 24, 2025

Uh oh!

viktorklang-ora commented Apr 24, 2025

Uh oh!

openjdk bot commented Apr 24, 2025

Uh oh!

openjdk bot commented Apr 24, 2025

Uh oh!

Reviewers

Assignees

Labels

viktorklang-ora commented Apr 23, 2025 •

edited by openjdk bot

Loading

openjdk bot commented Apr 23, 2025 •

edited

Loading

mlbridge bot commented Apr 23, 2025 •

edited

Loading

AlanBateman commented Apr 23, 2025 •

edited

Loading