[Gentle Ping?][8u] RFR: 8260589: Crash in JfrTraceIdLoadBarrier::load(_jclass*)

[Zhengyu Gu]

Looks good to me.

Thanks,

-Zhengyu

> 
> Hi,
> 
> I would appreciate it if you could review the backport of JDK-8260589 to 8u.
> This crash problem is very easy to reproduce, so I feel it is
> necessary to backport it to 8u.
> 
> Bug: https://bugs.openjdk.java.net/browse/JDK-8260589
> 11u commit: https://github.com/openjdk/jdk11u-dev/commit/1d204c554ffe969567161cc05992486ff47d346d
> 8u webrev: https://cr.openjdk.java.net/~ddong/8260589/webrev.00
> Test: jdk/test/jdk/jfr/jvm/TestPrimitiveClasses.java passed.
> 
> Due to the differences between JFR in 11u and 17, Denghui made some
> changes when backporting this fix to 11u.
> Denghui's changes also apply to 8u, so I directly quote Denghui's list
> here. (4 in total)
> 1. use MaxJfrEventId + 101 instead of LAST_TYPE_ID + 1 as the klass id
> of void.class
> 2. jdk 11 doesn't support jfr streaming, so I removed the call
> `JfrTraceIdEpoch::set_changed_tag_state()` in `load_primitive`
> 3. the Class in jdk 11 doesn't have the field 'hidden', so I removed
> `writer->write<bool>(false);` in `write_primitive`
> 4. there are many differences in the API of JfrTraceId between 11u and tip
> 
> In addition to the above differences, I also make supplementary
> explanations for the modifications I made.
> 1. The static global variable clear_artifacts in 11u was introduced by
> the bugfix https://bugs.openjdk.java.net/browse/JDK-8231081, but this
> fix has not been backported to 8u, so I removed the code related to
> clear_artifacts. In 8u, the metadata of the primitive types will be
> written into the previous chunk on every chunk rotation.
> 2. In 11u, _artifacts and _class_unload are static global variables,
> but in 8u, they are static member variables of JfrTypeSet, so I also
> made necessary changes to the functions that use these two variables.
> 
> Thanks,
> Long Yang
> 
> 
>