RFR: 8355631: The events might be generated after VM_DEATH event [v3]

[Leonid Mesnik]

On Mon, 27 Oct 2025 05:24:51 GMT, David Holmes <dholmes at openjdk.org> wrote:

>> Leonid Mesnik has updated the pull request incrementally with one additional commit since the last revision:
>> 
>>   cleanup
>
> src/hotspot/share/prims/jvmtiEventController.cpp line 38:
> 
>> 36: #include "runtime/frame.inline.hpp"
>> 37: #include "runtime/javaThread.inline.hpp"
>> 38: #include "runtime/serviceThread.hpp"
> 
> Why do we need this?

Removed leftover

> src/hotspot/share/prims/jvmtiEventController.cpp line 1230:
> 
>> 1228: 
>> 1229:   // Some events might be still in callback for daemons threads and ServiceThread.
>> 1230:   const double start = os::elapsedTime();
> 
> Given we sleep for 100ms each iteration we could simply count the iterations and break at 600, rather than tracking elapsed time. Not that it makes any real difference.

I prefer to set max waiting time explicitly.  The code is little bit longer but cleaner.

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

PR Review Comment: https://git.openjdk.org/jdk/pull/27504#discussion_r2466061054
PR Review Comment: https://git.openjdk.org/jdk/pull/27504#discussion_r2466108206