8358534: Bailout in Conv2B::Ideal when type of cmp input is not supported

[shipilev]

Clean bailout from the optimization added in JDK 21. While the patch is fairly recent, it does not come with elevated risk, as it just skips the optimization in the case where the alternative is compiler crash. This crash is seen in real world: adoptium/adoptium-support#1191

Additional testing:

• Linux x86_64 server fastdebug, all

---

Progress

• Change must not contain extraneous whitespace
• Commit message must refer to an issue
• JDK-8358534 needs maintainer approval

Issue

• JDK-8358534: Bailout in Conv2B::Ideal when type of cmp input is not supported (Bug - P3 - Approved)

Reviewers

• Cesar Soares Lucas (@JohnTortugo - no project role)

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.org/jdk21u-dev.git pull/1904/head:pull/1904
$ git checkout pull/1904

Update a local copy of the PR:
$ git checkout pull/1904
$ git pull https://git.openjdk.org/jdk21u-dev.git pull/1904/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 1904

View PR using the GUI difftool:
$ git pr show -t 1904

Using diff file

Download this PR as a diff file:
https://git.openjdk.org/jdk21u-dev/pull/1904.diff

Using Webrev

Link to Webrev Comment

[bridgekeeper]

👋 Welcome back shade! 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]

@shipilev 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:

8358534: Bailout in Conv2B::Ideal when type of cmp input is not supported

Reviewed-by: cslucas

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 15 new commits pushed to the master branch:

• 6f05078: 8261242: [Linux] OSContainer::is_containerized() returns true when run outside a container
• 12edcff: 8344671: Few JFR streaming tests fail with application not alive error on MacOS 15
• dfa0f73: 8345471: Clean up compiler/intrinsics/sha/cli tests
• ... and 12 more: https://git.openjdk.org/jdk21u-dev/compare/9a6ef52b82e66dfd5779f66113c28371ff2fd776...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]

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

[openjdk]

⚠️ @shipilev 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.

[mlbridge]

Webrevs

• 00: Full (2e9f6bd8)

[shipilev]

/approval request Fixes the regression caused by the optimization added in JDK 21. While the patch is fairly recent, its risk is low, as it just skips the optimization in the case where the alternative is compiler crash. This crash is seen in real world. Applies cleanly. All testing passes.

[openjdk]

@shipilev
8358534: The approval request has been created successfully.

[JohnTortugo]

LGTM

[jerboaa]

/approve yes

[openjdk]

@jerboaa
8358534: The approval request has been approved.

[shipilev]

Push approval granted. Here goes.

/integrate

[openjdk]

Going to push as commit f55ff01.
Since your change was applied there have been 15 commits pushed to the master branch:

• 6f05078: 8261242: [Linux] OSContainer::is_containerized() returns true when run outside a container
• 12edcff: 8344671: Few JFR streaming tests fail with application not alive error on MacOS 15
• dfa0f73: 8345471: Clean up compiler/intrinsics/sha/cli tests
• ... and 12 more: https://git.openjdk.org/jdk21u-dev/compare/9a6ef52b82e66dfd5779f66113c28371ff2fd776...master

Your commit was automatically rebased without conflicts.

[openjdk]

@shipilev Pushed as commit f55ff01.

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

[victornoel]

Hi @shipilev, sorry to bother you but I see that since this PR was merged, jdk-21.0.8+9 was released, but I don't see the corresponding commit in that tag. I feel like I'm missing something but shouldn't it have been included in it?

Thanks for your help!

[shipilev]

@victornoel -- I cannot see your reply, bots asked you something :)

[victornoel]

Done, thx for the heads up because I didn't get notified 😅

[jerboaa]

I feel like I'm missing something but shouldn't it have been included in it?

Yes. The bug is fixed in 21.0.9 (upcoming release in October) only. See backports section in: https://bugs.openjdk.org/browse/JDK-8358534

[shipilev]

Ah, OK, I see the comment now. So the answer is simple: the fix went into 21u on June 23, this is way past the stabilization cutoff for 21.0.8 (July 2025) release. The fix would normally be in 21.0.9 (Oct 2025). In retrospect, we could have requested a critical approval for July 2025 release, but there is not a solid chance it would have been accepted.

[victornoel]

Ah ! Thank you both for the explanation, I better understand now (should have thought about double checking the bug itself!). Thanks again for the fix, I will patiently wait then :)

Cheers!