RFR: 8251397: Add release fence to ClassValueMap constructor [v2]

[Galder Zamarreño]

On Mon, 14 Sep 2020 12:04:30 GMT, Severin Gehwolf <sgehwolf at openjdk.org> wrote:

>> @galderz Now it's the other way round. Commit message the bot mentions
>> [here](https://github.com/openjdk/jdk/pull/94#issuecomment-691478179) should be something like this:
>> 8251397: NPE on ClassValue.ClassValueMap.cacheArray
>> 
>> Add release fence to ClassValueMap constructor.
>> 
>> * Release fence guarantees that cacheArray field will published
>>   with a non-null value.
>> * Without this fix, CacheValueMap.cacheArray can sometimes be
>>    seen as null.
>> 
>> Reviewed-by: shade, psandoz
>> 
>> The first line can be achieved by force-pushing with the changed commit message. I know David said to not force push,
>> but the review is done and no other way to achieve this except changing the bug synopsis which isn't great either. The
>> latter lines can be added with /summary e.g.
>> Add release fence to ClassValueMap constructor.
>> 
>> * Release fence guarantees that cacheArray field will published
>>   with a non-null value.
>> * Without this fix, CacheValueMap.cacheArray can sometimes be
>>    seen as null.
>> 
>> The `Reviewed-by:` line comes from the reviewers.
>
> Note that I've opened [SKARA-633](https://bugs.openjdk.java.net/browse/SKARA-633) for pre-population of the commit
> message with an appropriate summary.

Add release fence to ClassValueMap constructor.

* Release fence guarantees that cacheArray field will published
  with a non-null value.
* Without this fix, CacheValueMap.cacheArray can sometimes be
   seen as null.

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

PR: https://git.openjdk.java.net/jdk/pull/94