(XS) RFR: 8131045: Obsolete JNIDetachReleasesMonitors

[David Holmes]

Thanks Dan!

Just need to wait for CCC approval before pushing.

David

On 27/10/2015 11:06 PM, Daniel D. Daugherty wrote:
> On 10/26/15 11:11 PM, David Holmes wrote:
>> bug: https://bugs.openjdk.java.net/browse/JDK-8131045
>>
>> webrev: http://cr.openjdk.java.net/~dholmes/8131045/webrev/
>
> src/share/vm/runtime/arguments.cpp
>      No comments.
>
> src/share/vm/runtime/globals.hpp
>      No comments.
>
> src/share/vm/runtime/thread.cpp
>      No comments.
>
> Thumbs up.
>
> Dan
>
>
>>
>> The JNIDetachReleasesMonitors flag was added to JDK 6 when it was
>> realized that the VM was not compliant with the JNI specification
>> regarding the release of monitors when a JNI attached thread detached
>> - see JDK-6336479. The flag allowed for restoration of the incorrect
>> behaviour if an application was somehow relying on it.
>>
>> By now all applications should themselves be JNI-spec compliant and
>> this workaround should not be needed, so we can obsolete the
>> JNIDetachReleasesMonitors flag in JDK 9 (accept it, report it, but
>> take no action) and expire (remove) it in JDK 10.
>>
>> > ./b00/linux-i586-internal/images/jdk/bin/java
>> -XX:-JNIDetachReleasesMonitors -version
>> Java HotSpot(TM) Server VM warning: Ignoring option
>> JNIDetachReleasesMonitors; support was removed in 9.0
>>
>> Testing: JPRT
>>
>> I haven't yet found any tests that involve this flag but there's bound
>> to be at least one hiding somewhere. :)
>>
>> A CCC request has been filed.
>>
>> Thanks,
>> David
>