On Fri, 24 Apr 2026 21:00:52 GMT, Serguei Spitsyn <[email protected]> wrote:
>> src/hotspot/share/prims/jvmtiEnvBase.cpp line 2506:
>>
>>> 2504: }
>>> 2505: if (!_set) { // ClearAllFramePops
>>> 2506: _state->clear_vthread_pending_deopts();
>>
>> The state is cleared always, even there are some other jvmti environments
>> still set FramePops. They are going to miss the events.
>>
>> Seems the clear_vthread_pending_deopts() should be called based on the
>> status of FramePop for all environments.
>
> Thank you for checking. Only one JVMTI environment can enable debugging
> features including `FramePop` events. So, this is not a problem if I do not
> miss anything.
It is not a solo capability and there is a test
test/hotspot/jtreg/vmTestbase/nsk/jvmti/scenarios/multienv/MA05/ma05t001/
that actually test that FramePop events are correctly works for different
environments.
-------------
PR Review Comment: https://git.openjdk.org/jdk/pull/28407#discussion_r3140590749
