On Fri, 19 Jun 2026 07:44:28 GMT, Jaikiran Pai <[email protected]> wrote:

> Can I please get a review of this change which proposes to improve the 
> exception message in `IllegalMonitorStateException` when it gets thrown for 
> value objects?
> 
> With this change, we now report the following exception message when value 
> objects are used to invoke `Object.wait()` and related methods:
> 
> 
> jshell> o.wait()
> |  Exception java.lang.IllegalMonitorStateException: current thread is not 
> owner
> |        at Object.wait0 (Native Method)
> |        at Object.wait (Object.java:433)
> |        at Object.wait (Object.java:387)
> |        at (#2:1)
> 
> There were some thoughts about using an exception message of the form 
> `Exception java.lang.IllegalMonitorStateException: object's type 
> java.lang.Integer is not an identity class`. But since value objects cannot 
> be synchronized upon, we decided it would be fine (and accurate) to reuse the 
> same "current thread is not owner" exception message that we use for identity 
> objects.
> 
> Given the nature of this change, I haven't added any new jtreg tests for 
> this. Existing tests in tier1, tier2 and tier3 continue to pass after this 
> change.
> 
> 
> ---------
> - [x] I confirm that I make this contribution in accordance with the [OpenJDK 
> Interim AI Policy](https://openjdk.org/legal/ai).

src/hotspot/share/runtime/synchronizer.cpp line 323:

> 321:   if ((obj)->mark().is_inline_type()) {  \
> 322:     JavaThread* THREAD = current;           \
> 323:     ResourceMark rm(THREAD);                \

Is the RM still needed?

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

PR Review Comment: 
https://git.openjdk.org/valhalla/pull/2568#discussion_r3441114485

Reply via email to