On Mon, 1 Jun 2026 09:03:34 GMT, David Simms <[email protected]> wrote:

>> This is a "*sub-review pull request*" for the first 
>> [preview](https://openjdk.org/jeps/12) of [JEP 401: Value Classes and 
>> Objects](https://openjdk.org/jeps/401), specifically 
>> [JDK-8317278](https://bugs.openjdk.org/browse/JDK-8317278): JVM 
>> implementation of value classes and objects.
>> 
>>> [!NOTE]
>>> This pull request and the other sub-review pull requests listed below are 
>>> based on the "*master pull request*" 
>>> (https://github.com/openjdk/jdk/pull/31120). It contains the same full set 
>>> of code changes as the "*master pull request*" to preserve the full 
>>> implementation context; the language compiler, JVM, and standard library 
>>> changes are intertwined. This separate pull requests exist only to 
>>> subdivide the review and related discussion by area.
>> 
>> Other areas for review:
>> 
>> - [JDK-8317277](https://bugs.openjdk.org/browse/JDK-8317277): Java language 
>> implementation of value classes and objects
>>   - https://github.com/openjdk/jdk/pull/31121
>> - [JDK-8317279](https://bugs.openjdk.org/browse/JDK-8317279): Standard 
>> library implementation of value classes and objects
>>   - https://github.com/openjdk/jdk/pull/31123
>> 
>> Code changes resulting from the review process should be made in 
>> [`valhalla/lworld`](https://github.com/openjdk/valhalla/).
>> 
>> `valhalla/lworld` is currently updated from `jdk/master` whenever a weekly 
>> [`jdk` tag](https://github.com/openjdk/jdk/tags) is created. At that time, 
>> code changes from `valhalla/lworld` will be propagated to the master pull 
>> request and to all sub-review pull requests, including this one.
>> 
>> Ultimately, review sign-off will be recorded on the "*master pull request*", 
>> and this pull request will be closed without integration.
>> 
>> This pull request has a large code surface area and often conflicts with 
>> `jdk/master` on a daily basis. Refer to 
>> [`valhalla/lworld`](https://github.com/openjdk/valhalla/) for the latest 
>> state of the project code, keeping in mind that it may lag several days 
>> behind `jdk/master`. Both repositories may be needed as references during 
>> review.
>> 
>> ---------
>> - [x] I confirm that I make this contribution in accordance with the 
>> [OpenJDK Interim AI Policy](https://openjdk.org/legal/ai).
>
> David Simms has updated the pull request with a new target base due to a 
> merge or a rebase. The pull request now contains 2754 commits:
> 
>  - Merge remote-tracking branch 'valhalla/lworld' into 
> jep401_sub_review_8317278
>  - Merge
>    
>    Merge jdk-27+24
>  - 8385674: [lworld] TestNullableInlineTypes.java fails after JDK-8325632
>    
>    Reviewed-by: mchevalier
>  - 8385652: [lworld] RedefineClasses should use stack map frame name
>    
>    Reviewed-by: fparain
>  - 8384107: [lworld] Update runtime/contended tests to run the same testing 
> for value classes
>    
>    Reviewed-by: fparain, lmesnik
>  - 8385600: [lworld] DA/DU issues with strict fields
>    
>    Reviewed-by: vromero
>  - 8384897: [lworld] this.staticField should be restricted in early 
> construction context
>    
>    Reviewed-by: liach, vromero
>  - 8385601: [lworld] Update testing documentation for the ValueClassPlugin 
> jtreg option
>    
>    Reviewed-by: lmesnik
>  - 8385569: [lworld] Apply JDK-8343767 to Valhalla specific StubRoutines
>    
>    Reviewed-by: fparain, vlivanov
>  - 8385581: [lworld] Remove the experimental JVMCI feature
>    8382708: [lworld] JVMCI support for Value Objects
>    8372605: [lworld] 
> compiler/jvmci/jdk.vm.ci.runtime.test/src/jdk/vm/ci/runtime/test/TestResolvedJava*.java
>  fail with --enable-preview
>    
>    Reviewed-by: thartmann
>  - ... and 2744 more: https://git.openjdk.org/jdk/compare/2c7efc08...e0250a38

src/hotspot/share/classfile/fieldLayoutBuilder.cpp line 1095:

> 1093: /* Computation of inline classes has a slightly different strategy than 
> for
> 1094:  * regular classes. Regular classes have their oop fields allocated at 
> the end
> 1095:  * of the layout to increase GC performances. Unfortunately, this 
> strategy

Suggestion:

 * of the layout to increase GC performance. Unfortunately, this strategy

src/hotspot/share/classfile/fieldLayoutBuilder.cpp line 1146:

> 1144:     // PADDING is inserted if needed to ensure the correct alignment of 
> the payload.
> 1145:     if (_is_abstract_value && _has_nonstatic_fields) {
> 1146:       // non-static fields of the abstract class must be laid out 
> without knowing

Suggestion:

      // Non-static fields of the abstract class must be laid out without 
knowing

src/hotspot/share/classfile/fieldLayoutBuilder.cpp line 1167:

> 1165:       // is computed. Super abstract value classes might have been too 
> conservative
> 1166:       // regarding alignment constraints, but now that the full set of 
> non-static fields is
> 1167:       // known, compute which alignment to use, then set first allowed 
> field offset

Suggestion:

      // known, compute which alignment to use, then set first allowed field 
offset.

src/hotspot/share/classfile/fieldLayoutBuilder.cpp line 1218:

> 1216:   // At this point, the characteristics of the raw layout (used in 
> standalone instances) are known.
> 1217:   // From this, additional layouts will be computed: atomic and 
> nullable layouts
> 1218:   // Once those additional layouts are computed, the raw layout might 
> need some adjustments

Suggestion:

  // From this, additional layouts will be computed: atomic and nullable 
layouts.
  // Once those additional layouts are computed, the raw layout might need some 
adjustments.

src/hotspot/share/classfile/fieldLayoutBuilder.cpp line 1237:

> 1235:     }
> 1236: 
> 1237:     // Next step is the nullable layouts: they must include a null 
> marker

Suggestion:

    // Next step is the nullable layouts: they must include a null marker.

src/hotspot/share/classfile/fieldLayoutBuilder.cpp line 1301:

> 1299:     // we want the raw layout to have the same alignment as those 
> atomic layouts so access codes
> 1300:     // could remain simple (single instruction without intermediate 
> copy). This might required
> 1301:     // to shift all fields in the raw layout, but this operation is 
> possible only if the class

Suggestion:

    // could remain simple (single instruction without intermediate copy). This 
might require
    // shifting all fields in the raw layout, but this operation is possible 
only if the class

src/hotspot/share/classfile/fieldLayoutBuilder.cpp line 1348:

> 1346:     // if the inline class has a null-free atomic layout, the the 
> layout used in heap allocated standalone
> 1347:     // instances must have at least equal to the atomic layout to allow 
> safe read/write atomic
> 1348:     // operation

Suggestion:

    // If the inline class has a null-free atomic layout, then the layout used 
in heap allocated standalone
    // instances must have at least equal to the atomic layout to allow safe 
read/write atomic
    // operation.

src/hotspot/share/classfile/fieldLayoutBuilder.cpp line 1502:

> 1500:             // null marker is zero (field is null), then all the fields 
> of the flat field are also
> 1501:             // zeroed. So, nullable flat field are not encoded 
> different than null-free flat fields,
> 1502:             // all fields are included in the map, plus the null marker

Suggestion:

            // all fields are included in the map, plus the null marker.

src/hotspot/share/classfile/fieldLayoutBuilder.cpp line 1506:

> 1504:             // require a dedicated section in the acmp map, and be 
> handled differently: null_marker
> 1505:             // comparison first, and if null markers are identical and 
> non-zero, then conditional
> 1506:             // comparison of the other fields

Suggestion:

            // comparison of the other fields.

src/hotspot/share/classfile/fieldLayoutBuilder.cpp line 1609:

> 1607:   // which prints the details of LayoutRawBlocks used to compute the 
> layout.
> 1608:   // The code below checks that offsets in the _field_info meta-data 
> match offsets
> 1609:   // in the LayoutRawBlocks

Suggestion:

  // in the LayoutRawBlocks.

-------------

PR Review Comment: https://git.openjdk.org/jdk/pull/31122#discussion_r3338199177
PR Review Comment: https://git.openjdk.org/jdk/pull/31122#discussion_r3338202987
PR Review Comment: https://git.openjdk.org/jdk/pull/31122#discussion_r3338204181
PR Review Comment: https://git.openjdk.org/jdk/pull/31122#discussion_r3338205762
PR Review Comment: https://git.openjdk.org/jdk/pull/31122#discussion_r3338207097
PR Review Comment: https://git.openjdk.org/jdk/pull/31122#discussion_r3338214786
PR Review Comment: https://git.openjdk.org/jdk/pull/31122#discussion_r3338219591
PR Review Comment: https://git.openjdk.org/jdk/pull/31122#discussion_r3338224149
PR Review Comment: https://git.openjdk.org/jdk/pull/31122#discussion_r3338225269
PR Review Comment: https://git.openjdk.org/jdk/pull/31122#discussion_r3338226906

Reply via email to