On Thu, 4 Jun 2026 04:22:07 GMT, Chen Liang <[email protected]> wrote:
>> src/hotspot/share/prims/jvmtiEnv.cpp line 2738:
>>
>>> 2736: // modifiers_ptr - pre-checked for null
>>> 2737: jvmtiError
>>> 2738: JvmtiEnv::GetClassModifiers(oop k_mirror, jint* modifiers_ptr) {
>>
>> Can you confirm that we are testing this with value classes? I didn't find
>> anything in test/hotspot/jtreg/serviceability/jvmti. The context for asking
>> is seeing that the exported APIs are consistent.
>
> I made this change, so I could [revert
> `test/hotspot/jtreg/vmTestbase/nsk/jdi/Accessible/modifiers/modifiers001.java`](https://github.com/openjdk/jdk/commit/a7ecceb9bd1a99f0674b76a852cc3ab2c25326a4#diff-cbe0d206a08bb013eee2a8cfcc05b7298f91ae960cd17448d2960346295d5c65)
> to mainline equivalent. I don't think this is being tested with value
> classes unfortunately, seems the existing tests are the ones in jdi around
> the test I last modified. However, these tests are run with preview-enabled
> VM so the "SUPER" trusting does work with preview identity classes.
>
> I don't know if core libs or hotspot (serviceability) would be in charge of
> adding such a value class test.
I've created a test, should I propose that for the lworld branch to its to this
PR when sync'ed up?
-------------
PR Review Comment: https://git.openjdk.org/jdk/pull/31123#discussion_r3353723801
