8379863: [lworld] C2: assert(!value->obj_is_scalar_replaced() || realloc_failures) failed: object should be reallocated already

[benoitmaillard]

This PR prevents hitting an assert in vframeArrayElement::fill_in when escape analysis and autobox elimination are disabled.

With value classes, scalarization can take place independently of escape analysis. When deoptimizing, this implies that we must always rematerialize objects when EliminateAllocations is true, even if escape analysis is disabled.
Prior to this change, we hit an assert in vframeArrayElement::fill_in because it is expected that scalarized objects are already reallocated there. We have to make sure that rematerialize_objects is run before that.

Testing

• tier1-3, plus some internal testing

Thank you for reviewing!

---

Progress

• Change must not contain extraneous whitespace
• Commit message must refer to an issue
• Change must be properly reviewed (1 review required, with at least 1 Committer)

Issue

• JDK-8379863: [lworld] C2: assert(!value->obj_is_scalar_replaced() || realloc_failures) failed: object should be reallocated already (Bug - P4)

Reviewers

• Marc Chevalier (@marc-chevalier - Committer)
• Tobias Hartmann (@TobiHartmann - Committer)

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.org/valhalla.git pull/2240/head:pull/2240
$ git checkout pull/2240

Update a local copy of the PR:
$ git checkout pull/2240
$ git pull https://git.openjdk.org/valhalla.git pull/2240/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 2240

View PR using the GUI difftool:
$ git pr show -t 2240

Using diff file

Download this PR as a diff file:
https://git.openjdk.org/valhalla/pull/2240.diff

Using Webrev

Link to Webrev Comment

[bridgekeeper]

👋 Welcome back bmaillard! A progress list of the required criteria for merging this PR into lworld will be added to the body of your pull request. There are additional pull request commands available for use with this pull request.

[openjdk]

@benoitmaillard 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:

8379863: [lworld] C2: assert(!value->obj_is_scalar_replaced() || realloc_failures) failed: object should be reallocated already

Reviewed-by: mchevalier, thartmann

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 12 new commits pushed to the lworld branch:

• 0ca8597: 8343835: [lworld] C2: assert fails with "no mismatched stores, except on raw memory: StoreD StoreI"
• 632fc4f: 8379333: [lworld] [REDO] JNI NewWeakGlobalRef should throw IdentityException for value types
• 0436584: 8380395: [lworld] ProblemList 4 sub-tests of compiler/valhalla/inlinetypes/TestCallingConvention.java
• ... and 9 more: https://git.openjdk.org/valhalla/compare/329029d25f15d069c2de4ce3c3860a91a354e458...lworld

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 (@marc-chevalier, @TobiHartmann) 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).

[mlbridge]

Webrevs

• 02: Full - Incremental (74ddbfde)
• 01: Full - Incremental (e1aea2be)
• 00: Full (87f23134)

[marc-chevalier]

Why was DoEscapeAnalysis necessary before, when EliminateAllocations is true?

[benoitmaillard]

On mainline scalarization is dependent on escape analysis, with valhalla inline types can be scalarized in InlineTypeNode::make_scalar_in_safepoints, independently of escape analysis @marc-chevalier.

[TobiHartmann]

Could you please add a regression test (or add a configuration to an existing test)?

[benoitmaillard]

I have added a regression test @TobiHartmann, sorry for not doing this earlier.

[marc-chevalier]

I like this version better.

[TobiHartmann]

Thanks, that looks good to me!

[benoitmaillard]

Thank you for your reviews @marc-chevalier @TobiHartmann!
/integrate

[openjdk]

@benoitmaillard
Your change (at version 74ddbfd) is now ready to be sponsored by a Committer.

[TobiHartmann]

/sponsor

[openjdk]

Going to push as commit 19fe8ba.
Since your change was applied there have been 13 commits pushed to the lworld branch:

• 2d66adb: 8380015: [lworld] runtime/valhalla/inlinetypes/field_layout/ValueRandomLayoutTest.java divides by zero during statistic printing
• 0ca8597: 8343835: [lworld] C2: assert fails with "no mismatched stores, except on raw memory: StoreD StoreI"
• 632fc4f: 8379333: [lworld] [REDO] JNI NewWeakGlobalRef should throw IdentityException for value types
• ... and 10 more: https://git.openjdk.org/valhalla/compare/329029d25f15d069c2de4ce3c3860a91a354e458...lworld

Your commit was automatically rebased without conflicts.

[openjdk]

@TobiHartmann @benoitmaillard Pushed as commit 19fe8ba.

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