8376185: NoSuchFieldError thrown after a record with type annotation retransformed

[jpbempel]

Hi all,

This pull request contains a backport of commit 0c60c9ca from the openjdk/jdk repository.

The commit being backported was authored by Jean-Philippe Bempel on 25 Mar 2026 and was reviewed by Serguei Spitsyn and David Holmes.

Thanks!

---

• I confirm that I make this contribution in accordance with the OpenJDK Interim AI Policy.

---

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-8376185 needs maintainer approval

Issue

• JDK-8376185: NoSuchFieldError thrown after a record with type annotation retransformed (Bug - P4 - Requested)

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.org/jdk21u-dev.git pull/2900/head:pull/2900
$ git checkout pull/2900

Update a local copy of the PR:
$ git checkout pull/2900
$ git pull https://git.openjdk.org/jdk21u-dev.git pull/2900/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 2900

View PR using the GUI difftool:
$ git pr show -t 2900

Using diff file

Download this PR as a diff file:
https://git.openjdk.org/jdk21u-dev/pull/2900.diff

Using Webrev

Link to Webrev Comment

[bridgekeeper]

👋 Welcome back jpbempel! 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]

❗ This change is not yet ready to be integrated.
See the Progress checklist in the description for automated requirements.

[openjdk]

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

[openjdk]

⚠️ @jpbempel 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.

[jpbempel]

/approval request retransformation on a record with type annotation on a record component leads to a corrupted state in the JVM memory. Best cases we have an exception, worst cases a JVM crash. It is common to use type annotation like @NotNull on the record components.

[openjdk]

@jpbempel
8376185: The approval request has been created successfully.

[mlbridge]

Webrevs

• 01: Full - Incremental (5c88a9e3)
• 00: Full (400cfd9c)

[GoeLin]

We see java/lang/instrument/RetransformRecordTypeAnn/TestRetransformRecord.java failing in 25 now.

There is a bug open for this, please do a ProblemList change for this and backport that change.

By the way, how did you test this if the test is failing? I.e. without noticeing?

[jpbempel]

We see java/lang/instrument/RetransformRecordTypeAnn/TestRetransformRecord.java failing in 25 now.

There is a bug open for this, please do a ProblemList change for this and backport that change.

By the way, how did you test this if the test is failing? I.e. without noticeing?

I was using latest jtreg, even for old JDK.
How I know which version of jtreg I need to use for a specific openjdk version/repo?

[GoeLin]

First, there was a bug open in JBS. Before you backport, you should check all bugs that are linked to the JBS issue.
If needed, you should backport them with your change. If not yet fixed, you should not backport the change.

https://bugs.openjdk.org/browse/JDK-8369488 is one of the changes that determines the jtreg to be used.
You find the previous and next jtreg bump linked, along with the information where it was pushed or backported.
Also you can look at the TEST.ROOT files in your repo which tell the minimum version.
Most tests work with any jtreg version, but some have problems either with older or younger. I usually use the minimum jtreg version, as does GHA. But GHA only runs tier1.