8364316: JFR: Incorrect validation of mirror fields

[egahlin]

Could I please have a review of a PR that fixes a bug in the mirror field validation?

Testing: jdk/jdk/jfr

Thanks
Erik

---

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-8364316: JFR: Incorrect validation of mirror fields (Bug - P4)

Reviewers

• Aleksey Shipilev (@shipilev - Reviewer)
• Markus Grönlund (@mgronlun - Reviewer)

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.org/jdk.git pull/26539/head:pull/26539
$ git checkout pull/26539

Update a local copy of the PR:
$ git checkout pull/26539
$ git pull https://git.openjdk.org/jdk.git pull/26539/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 26539

View PR using the GUI difftool:
$ git pr show -t 26539

Using diff file

Download this PR as a diff file:
https://git.openjdk.org/jdk/pull/26539.diff

Using Webrev

Link to Webrev Comment

[bridgekeeper]

👋 Welcome back egahlin! 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]

@egahlin 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:

8364316: JFR: Incorrect validation of mirror fields

Reviewed-by: shade, mgronlun

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 38 new commits pushed to the master branch:

• e0e8206: 8364166: Parallel: Remove the use of soft_ref_policy in Full GC
• 443afdc: 8364176: Serial: Group all class unloading logic at the end of marking phase
• fe09e93: 8364317: Explicitly document some assumptions of StringUTF16
• ... and 35 more: https://git.openjdk.org/jdk/compare/ea754316fd6d691a701dfb4bc921ea8c92dc5dd4...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]

@egahlin The following label will be automatically applied to this pull request:

• hotspot-jfr

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]

Webrevs

• 01: Full - Incremental (2947c54a)
• 00: Full (75be1d4a)

[shipilev]

Looks reasonable.

There might be another quirk in this code: fields of the same name can be hidden, so verification would miss that? See:

jshell> class C1 { int x;}
|  created class C1

jshell> class C2 extends C1 { int x;}
|  created class C2

jshell> Class cMirror = C2.class;
   ...> while (cMirror != Object.class) {
   ...>     System.out.println(cMirror.getDeclaredField("x").getName());
   ...>     cMirror = cMirror.getSuperclass();
   ...> }
cMirror ==> class C2
x
x

[egahlin]

There might be another quirk in this code: fields of the same name can be hidden, so verification would miss that?

I don't think mirror events can be made to work with a class hierarchy (due to shadowing), and there is no need to, since their only purpose is to provide metadata for events defined in java.base (which cannot depend on the jdk.jfr module).

It's fine to have all the metadata directly in the mirror event class.

So, I simplified the traversal code and, at the same time, realized we should not duplicate the logic used to discover which event fields should be included. Now there is only one method, Utils.getEventFields(...), handling this.

The reason I could remove the synthetic check in verifyMirror is that its only purpose was to exclude the startTime and duration fields in the jdk.internal.Event class. The traversal stops before that class is reached, so it's not an issue.

I understand this may be more to review than you are willing to do.

[shipilev]

This looks reasonable as well, as long as it verifies what you want to verify :)

[egahlin]

/integrate

[openjdk]

Going to push as commit 68a4396.
Since your change was applied there have been 70 commits pushed to the master branch:

• da0d959: 8364190: JFR: RemoteRecordingStream withers don't work
• b96b9c3: 8364461: JFR: Default constructor may not be first in setting control
• 3bc4497: 8364258: ThreadGroup constant pool serialization is not normalized
• ... and 67 more: https://git.openjdk.org/jdk/compare/ea754316fd6d691a701dfb4bc921ea8c92dc5dd4...master

Your commit was automatically rebased without conflicts.

[openjdk]

@egahlin Pushed as commit 68a4396.

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