8381549: GenShen: Global collections skip coalesce and fill when the collection set is empty by earthling-amzn · Pull Request #30543 · openjdk/jdk

earthling-amzn · 2026-04-01T23:30:50Z

When objects become unreachable in old regions they will no longer have their references updated. If a heap walk tries to follow one of these references, it may encounter garbage and crash. This step was inadvertently skipped by an erroneous code change in a recent refactoring.

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-8381549: GenShen: Global collections skip coalesce and fill when the collection set is empty (Bug - P4)

Reviewers

Kelvin Nilsen (@kdnilsen - Reviewer)
Xiaolong Peng (@pengxiaolong - Committer)
Y. Srinivas Ramakrishna (@ysramakrishna - Reviewer)

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.org/jdk.git pull/30543/head:pull/30543
$ git checkout pull/30543

Update a local copy of the PR:
$ git checkout pull/30543
$ git pull https://git.openjdk.org/jdk.git pull/30543/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 30543

View PR using the GUI difftool:
$ git pr show -t 30543

Using diff file

Download this PR as a diff file:
https://git.openjdk.org/jdk/pull/30543.diff

Using Webrev

Link to Webrev Comment

…t (even if it is empty)

bridgekeeper · 2026-04-01T23:32:48Z

👋 Welcome back wkemper! 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 · 2026-04-01T23:34:13Z

@earthling-amzn 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:

8381549: GenShen: Global collections skip coalesce and fill when the collection set is empty

Reviewed-by: kdnilsen, xpeng, ysr

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 20 new commits pushed to the master branch:

e254526: 8380958: GenShen: Regulator thread may observe inconsistent old generation state
410b3e8: 8381606: Javadoc typo in ResourceBundle example
0c98ab8: 8380936: Cleanup temp dir usage in ZipFSTester
... and 17 more: https://git.openjdk.org/jdk/compare/cddee6d6eb3e048635c380a32bd2f6ebfd2c18b5...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 · 2026-04-01T23:35:06Z

@earthling-amzn The following labels will be automatically applied to this pull request:

hotspot-gc
shenandoah

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

mlbridge · 2026-04-01T23:38:41Z

Webrevs

01: Full - Incremental (34d3c50c)
00: Full (f22fd13b)

ysramakrishna · 2026-04-02T00:09:11Z

If a heap walk tries to follow one of these references, it may encounter garbage and crash.

What was the heap walk that tried to follow one of these references? Are we talking about a dead (unmarked) object? Would that not be skipped because of using the valid (complete) marking bit map during such a walk? (Was it dirty card / remset scanning of old gen?) Something else?

Change looks good, but wanted to make sure I understood the need for the coalesce and fill necessary for a linear walk over the region.

earthling-amzn · 2026-04-02T01:26:00Z

In this case, the heap walk was indeed a remembered set scan. The remembered set scan cannot use the mark bitmap for old during old marking. The scan relies on old, no longer reachable objects, being filled in before old marking resets the bitmap.

kdnilsen

I have some questions about the intent here?

kdnilsen · 2026-04-02T18:19:24Z

src/hotspot/share/gc/shenandoah/heuristics/shenandoahGenerationalHeuristics.cpp

    // eligible for coalescing. As implemented now, this has the side effect of possibly initiating mixed-evacuations
    // after a global cycle for old regions that were not included in this collection set.
-    heap->old_generation()->prepare_for_mixed_collections_after_global_gc();
+    heap->old_generation()->update_old_regions_after_global_cycle();


I'm not sure we want to make this change. IIUC, as currently implemented, we may end a Global GC with many candidates for mixed evacuation. While this means "completion" of the Global GC is somewhat delayed, it also means the Global GC has a much smaller impact on the normal cadence of young collections.

What is the impact of this change?

Does it force us to do all mixed evacuation in a single cycle?

Does it reduce the amount of mixed evacuation that we are able to perform following an "old mark"?

Sorry for the panic. This is just a name change?

In this context, "update" does not mean update-refs. It means:

"update" the internal list of mixed-evacuation candidates?
"update" the gc log?

I guess I was a bit confused by the name of this function. Help me understand what we are trying to communicate with this name?

I am open to better names. In retrospect, I made the change to skip this function call when the collection set is empty because of the old name: prepare_for_mixed_collections_after_global_gc. The function does much more than prepare for mixed collections. Critically, if there are no collection candidates at all, it must still transition the old generation to the FILLING state so that dead objects are replaced with array primitives. Maybe transition_old_generation_after_global_gc?

Thanks. I like that name better.

kdnilsen

Thanks.

pengxiaolong

Thanks, looks good to me!

ysramakrishna

🚢

(but left a small remark re documentation comment; not essential and doesn't need a re-review.)

ysramakrishna · 2026-04-03T20:20:05Z

src/hotspot/share/gc/shenandoah/shenandoahOldGeneration.hpp


  bool entry_coalesce_and_fill();
-  void prepare_for_mixed_collections_after_global_gc();
+  void transition_old_generation_after_global_gc();


Thanks for the nice explanation in the PR discussion/convo w/Kelvin.

Might be a good idea to add a one-sentence documentation comment for the method here while you are here may be? I realize the method itself is fairly small, but ... :-)

I shall add one in a subsequent PR. Thank you!

earthling-amzn · 2026-04-03T21:04:11Z

/integrate

openjdk · 2026-04-03T21:04:55Z

Going to push as commit f21e47d.
Since your change was applied there have been 21 commits pushed to the master branch:

a333111: 8380291: AOT cache should store only unregistered classes with file: code source
e254526: 8380958: GenShen: Regulator thread may observe inconsistent old generation state
410b3e8: 8381606: Javadoc typo in ResourceBundle example
... and 18 more: https://git.openjdk.org/jdk/compare/cddee6d6eb3e048635c380a32bd2f6ebfd2c18b5...master

Your commit was automatically rebased without conflicts.

openjdk · 2026-04-03T21:05:04Z

@earthling-amzn Pushed as commit f21e47d.

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

Do not skip updating old regions after chosing a global collection se…

f22fd13

…t (even if it is empty)

openjdk bot added hotspot-gc hotspot-gc-dev@openjdk.org shenandoah shenandoah-dev@openjdk.org labels Apr 1, 2026

openjdk bot added the rfr Pull request is ready for review label Apr 1, 2026

kdnilsen reviewed Apr 2, 2026

View reviewed changes

Use method name that does not conflict with 'update-refs' concept

34d3c50

kdnilsen approved these changes Apr 3, 2026

View reviewed changes

openjdk bot added the ready Pull request is ready to be integrated label Apr 3, 2026

pengxiaolong approved these changes Apr 3, 2026

View reviewed changes

ysramakrishna approved these changes Apr 3, 2026

View reviewed changes

openjdk bot added the integrated Pull request has been integrated label Apr 3, 2026

openjdk bot closed this Apr 3, 2026

openjdk bot removed ready Pull request is ready to be integrated rfr Pull request is ready for review labels Apr 3, 2026

earthling-amzn mentioned this pull request Apr 3, 2026

8381643: Shenandoah: Remove duplicate region filtering logic #30571

Open

4 tasks

Conversation

earthling-amzn commented Apr 1, 2026 • 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 1, 2026

Uh oh!

openjdk bot commented Apr 1, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

openjdk bot commented Apr 1, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

mlbridge bot commented Apr 1, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Webrevs

Uh oh!

ysramakrishna commented Apr 2, 2026

Uh oh!

earthling-amzn commented Apr 2, 2026

Uh oh!

kdnilsen left a comment

Choose a reason for hiding this comment

Uh oh!

kdnilsen Apr 2, 2026

Choose a reason for hiding this comment

Uh oh!

kdnilsen Apr 2, 2026

Choose a reason for hiding this comment

Uh oh!

kdnilsen Apr 2, 2026

Choose a reason for hiding this comment

Uh oh!

earthling-amzn Apr 2, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

kdnilsen Apr 3, 2026

Choose a reason for hiding this comment

Uh oh!

kdnilsen left a comment

Choose a reason for hiding this comment

Uh oh!

pengxiaolong left a comment

Choose a reason for hiding this comment

Uh oh!

ysramakrishna left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

ysramakrishna Apr 3, 2026

Choose a reason for hiding this comment

Uh oh!

earthling-amzn Apr 3, 2026

Choose a reason for hiding this comment

Uh oh!

earthling-amzn commented Apr 3, 2026

Uh oh!

openjdk bot commented Apr 3, 2026

Uh oh!

openjdk bot commented Apr 3, 2026

Uh oh!

Reviewers

Assignees

Labels

Milestone

Development

Uh oh!

4 participants

earthling-amzn commented Apr 1, 2026 •

edited by openjdk bot

Loading

openjdk bot commented Apr 1, 2026 •

edited

Loading

openjdk bot commented Apr 1, 2026 •

edited

Loading

mlbridge bot commented Apr 1, 2026 •

edited

Loading

earthling-amzn Apr 2, 2026 •

edited

Loading

ysramakrishna left a comment •

edited

Loading