8298061: vmTestbase/nsk/sysdict/vm/stress/btree/btree012/btree012.java failed with "fatal error: refcount has gone to zero" #3328

dchuyko · 2025-03-05T19:05:34Z

Please review the backport of the fix for JDK-8298061 to jdk17u. The bug leads to rare crashes. For instance, they cause the Spring Boot build to fail under certain circumstances. In particular on AMD EPYC 7R13 and Ubuntu 24.04.1.

Original patch doesn't apply cleanly. First, technically clear_supername() was only added by JDK-8292286 in JDK 20 and Symbol::maybe_increment/decrement_refcount were added by JDK-8291457 also in JDK 20, all before the original patch.
Second, PlaceholderTable::find_and_remove() actually needs what clear_supername() call does in 20 to make the fix complete (and to match find_and_add()). That second part was later reworked in JDK 21 (JDK-8302108) and after.

PlaceholderEntry::set_supername() was modified to perform same work as before under the same condition as in the original patch.

The early version of the original change by @coleenp had an implementation of the second part which is suitable for 17u so it was made a part of this backport. See openjdk/jdk#11726

Without that second part part we did observe assert failures.

Testing: jtreg tier 1-3 on linux-amd64, dedicated stress test of the Spring Boot build.

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
JDK-8298061 needs maintainer approval

Issue

JDK-8298061: vmTestbase/nsk/sysdict/vm/stress/btree/btree012/btree012.java failed with "fatal error: refcount has gone to zero" (Bug - P3 - Approved)

Reviewers

Coleen Phillimore (@coleenp - Reviewer)

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.org/jdk17u-dev.git pull/3328/head:pull/3328
$ git checkout pull/3328

Update a local copy of the PR:
$ git checkout pull/3328
$ git pull https://git.openjdk.org/jdk17u-dev.git pull/3328/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 3328

View PR using the GUI difftool:
$ git pr show -t 3328

Using diff file

Download this PR as a diff file:
https://git.openjdk.org/jdk17u-dev/pull/3328.diff

Using Webrev

Link to Webrev Comment

bridgekeeper · 2025-03-05T19:06:20Z

👋 Welcome back dchuyko! 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 · 2025-03-05T19:07:13Z

@dchuyko 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:

8298061: vmTestbase/nsk/sysdict/vm/stress/btree/btree012/btree012.java failed with "fatal error: refcount has gone to zero"

Reviewed-by: coleenp

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 95 new commits pushed to the master branch:

59468ba: 8349039: Adjust exception No type named in database
00c53b7: 8347267: [macOS]: UnixOperatingSystem.c:67:40: runtime error: division by zero
9631398: 8347576: Error output in libjsound has non matching format strings
96e8c56: 8277983: Remove unused fields from sun.net.www.protocol.jar.JarURLConnection
260f1c0: 8348107: test/jdk/java/net/httpclient/HttpsTunnelAuthTest.java fails intermittently
8cf4622: 8345134: Test sun/security/tools/jarsigner/ConciseJarsigner.java failed: unable to find valid certification path to requested target
a064c2f: 8343529: serviceability/sa/ClhsdbWhere.java fails AssertionFailure: Corrupted constant pool
9a5b72f: 8341373: Open source closed frame tests # 4
5d02873: 8341000: Open source some of the AWT Window tests
90b1bdd: 8340228: Open source couple more miscellaneous AWT tests
... and 85 more: https://git.openjdk.org/jdk17u-dev/compare/6e37df10401b55a1944fb5c1007dea979f696e1c...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.

As you do not have Committer status in this project an existing Committer must agree to sponsor your change. Possible candidates are the reviewers of this PR (@coleenp) but any other Committer may sponsor as well.

➡️ To flag this PR as ready for integration with the above commit message, type /integrate in a new comment. (Afterwards, your sponsor types /sponsor in a new comment to perform the integration).

openjdk · 2025-03-05T19:07:32Z

This backport pull request has now been updated with issue from the original commit.

mlbridge · 2025-03-05T19:11:12Z

Webrevs

00: Full (044f41d6)

coleenp · 2025-03-19T19:45:01Z

src/hotspot/share/classfile/placeholders.hpp

@@ -142,8 +142,12 @@ class PlaceholderEntry : public HashtableEntry<Symbol*, mtClass> {

  Symbol*            supername()           const { return _supername; }
  void               set_supername(Symbol* supername) {
-    _supername = supername;
-    if (_supername != NULL) _supername->increment_refcount();
+    if (supername != _supername) {


Why isn't there an if (_supername == NULL) _supername->decrement_refcount(); before assigning the new supername?

Did you mean if (_supername != NULL)?
Later versions call Symbol::maybe_decrement_refcount(_supername); that acts like if (s != nullptr) s->decrement_refcount();.

Also it turns out that checking for NULL may be not enough. PlaceholderEntry has no constructor, so members can contain garbage.

I see. This is the old hashtable implementation.

Coleen, thanks for the review.

coleenp

This looks good. Thank you for testing out this backport.

coleenp · 2025-03-25T18:40:33Z

src/hotspot/share/classfile/placeholders.hpp

@@ -142,8 +142,12 @@ class PlaceholderEntry : public HashtableEntry<Symbol*, mtClass> {

  Symbol*            supername()           const { return _supername; }
  void               set_supername(Symbol* supername) {
-    _supername = supername;
-    if (_supername != NULL) _supername->increment_refcount();
+    if (supername != _supername) {


I see. This is the old hashtable implementation.

openjdk · 2025-03-25T18:42:03Z

⚠️ @dchuyko This change is now ready for you to apply for maintainer approval. This can be done directly in each associated issue or by using the /approval command.

dchuyko · 2025-03-25T19:05:04Z

/approval request This is a backport of a small reference counting fix in PlaceholderEntry::set_supername(). The bug leads to rare crashes. In particular, it causes the Spring Boot build to fail under certain circumstances. The original patch doesn't apply cleanly and is not enough (because of changes in 20 and later), it is slightly modified/extended. We basically add guards t prevent null/uninitialized data access so the risk is low.

Testing: jtreg tier 1-3 on linux-amd64, dedicated stress test of the Spring Boot build.

openjdk · 2025-03-25T19:06:02Z

@dchuyko
8298061: The approval request has been created successfully.

dchuyko · 2025-03-31T12:44:09Z

/integrate

openjdk · 2025-03-31T12:44:48Z

@dchuyko
Your change (at version 044f41d) is now ready to be sponsored by a Committer.

phohensee · 2025-03-31T15:35:18Z

/sponsor

openjdk · 2025-03-31T15:35:40Z

Going to push as commit 41b3541.
Since your change was applied there have been 95 commits pushed to the master branch:

59468ba: 8349039: Adjust exception No type named in database
00c53b7: 8347267: [macOS]: UnixOperatingSystem.c:67:40: runtime error: division by zero
9631398: 8347576: Error output in libjsound has non matching format strings
96e8c56: 8277983: Remove unused fields from sun.net.www.protocol.jar.JarURLConnection
260f1c0: 8348107: test/jdk/java/net/httpclient/HttpsTunnelAuthTest.java fails intermittently
8cf4622: 8345134: Test sun/security/tools/jarsigner/ConciseJarsigner.java failed: unable to find valid certification path to requested target
a064c2f: 8343529: serviceability/sa/ClhsdbWhere.java fails AssertionFailure: Corrupted constant pool
9a5b72f: 8341373: Open source closed frame tests # 4
5d02873: 8341000: Open source some of the AWT Window tests
90b1bdd: 8340228: Open source couple more miscellaneous AWT tests
... and 85 more: https://git.openjdk.org/jdk17u-dev/compare/6e37df10401b55a1944fb5c1007dea979f696e1c...master

Your commit was automatically rebased without conflicts.

openjdk · 2025-03-31T15:35:54Z

@phohensee @dchuyko Pushed as commit 41b3541.

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

Backport 03d992728e27bd3dcd00d1af8a7b7179281e626f

044f41d

openjdk bot changed the title ~~Backport 03d992728e27bd3dcd00d1af8a7b7179281e626f~~ 8298061: vmTestbase/nsk/sysdict/vm/stress/btree/btree012/btree012.java failed with "fatal error: refcount has gone to zero" Mar 5, 2025

openjdk bot added backport rfr Pull request is ready for review labels Mar 5, 2025

coleenp reviewed Mar 19, 2025

View reviewed changes

coleenp approved these changes Mar 25, 2025

View reviewed changes

openjdk bot added approval ready Pull request is ready to be integrated and removed approval labels Mar 25, 2025

openjdk bot added the sponsor Pull request is ready to be sponsored label Mar 31, 2025

openjdk bot added the integrated Pull request has been integrated label Mar 31, 2025

openjdk bot closed this Mar 31, 2025

openjdk bot removed ready Pull request is ready to be integrated rfr Pull request is ready for review sponsor Pull request is ready to be sponsored labels Mar 31, 2025

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

8298061: vmTestbase/nsk/sysdict/vm/stress/btree/btree012/btree012.java failed with "fatal error: refcount has gone to zero" #3328

8298061: vmTestbase/nsk/sysdict/vm/stress/btree/btree012/btree012.java failed with "fatal error: refcount has gone to zero" #3328

dchuyko commented Mar 5, 2025 •

edited by openjdk bot

Loading

bridgekeeper bot commented Mar 5, 2025

openjdk bot commented Mar 5, 2025 •

edited

Loading

openjdk bot commented Mar 5, 2025

mlbridge bot commented Mar 5, 2025

coleenp Mar 19, 2025

dchuyko Mar 20, 2025

dchuyko Mar 24, 2025 •

edited

Loading

coleenp Mar 25, 2025

dchuyko Mar 25, 2025

coleenp left a comment

coleenp Mar 25, 2025

openjdk bot commented Mar 25, 2025

dchuyko commented Mar 25, 2025

openjdk bot commented Mar 25, 2025

dchuyko commented Mar 31, 2025

openjdk bot commented Mar 31, 2025

phohensee commented Mar 31, 2025

openjdk bot commented Mar 31, 2025

openjdk bot commented Mar 31, 2025

8298061: vmTestbase/nsk/sysdict/vm/stress/btree/btree012/btree012.java failed with "fatal error: refcount has gone to zero" #3328

8298061: vmTestbase/nsk/sysdict/vm/stress/btree/btree012/btree012.java failed with "fatal error: refcount has gone to zero" #3328

Conversation

dchuyko commented Mar 5, 2025 • edited by openjdk bot Loading

Progress

Issue

Reviewers

Reviewing

bridgekeeper bot commented Mar 5, 2025

openjdk bot commented Mar 5, 2025 • edited Loading

openjdk bot commented Mar 5, 2025

mlbridge bot commented Mar 5, 2025

Webrevs

coleenp Mar 19, 2025

Choose a reason for hiding this comment

dchuyko Mar 20, 2025

Choose a reason for hiding this comment

dchuyko Mar 24, 2025 • edited Loading

Choose a reason for hiding this comment

coleenp Mar 25, 2025

Choose a reason for hiding this comment

dchuyko Mar 25, 2025

Choose a reason for hiding this comment

coleenp left a comment

Choose a reason for hiding this comment

coleenp Mar 25, 2025

Choose a reason for hiding this comment

openjdk bot commented Mar 25, 2025

dchuyko commented Mar 25, 2025

openjdk bot commented Mar 25, 2025

dchuyko commented Mar 31, 2025

openjdk bot commented Mar 31, 2025

phohensee commented Mar 31, 2025

openjdk bot commented Mar 31, 2025

openjdk bot commented Mar 31, 2025

dchuyko commented Mar 5, 2025 •

edited by openjdk bot

Loading

openjdk bot commented Mar 5, 2025 •

edited

Loading

dchuyko Mar 24, 2025 •

edited

Loading