[8u-dev] Request for approval: JDK-8133023: ParallelGCThreads is not calculated correctly

[Poonam Bajaj Parhar]

Hello Rob,

I discussed this with Jon Masamitsu, and he said that it is very hard to 
come up with a testcase to determine that the correct number of Parallel 
GC threads have been created for a particular platform. We would 
essentially need to duplicate the entire HotSpot code that calculates 
the number of Parallel GC threads in the testcase for all the platforms 
to be able to write this testcase and it is not worth it for this small 
backport.

I have added noreg-hard keyword. Please review and approve this change.

Thanks,
Poonam

>>> On 2/23/2016 6:07 AM, Rob McKenna wrote:
>>>> Hi Poonam,
>>>>
>>>> Why does this have a noreg-self label? (possibly a typo of 
>>>> noreg-hard?)
>>
>> Rob,
>>
>> This was probably my doing.  What does noreg-self label mean?
>>
>> Jon
>>
>>> Not sure what noreg-self means, it was already added for jdk9. I have
>>> added new label noreg-trivial as one can just run 'java
>>> +PrintFlagsFinal -version' and check the value of ParallelGCThreads to
>>> verify this fix.
>>>
>>>>
>>>> Also, please include a link to the bug in future requests.
>>> Sure, will do.
>>>
>>> Thanks,
>>> Poonam
>>>>
>>>>     -Rob
>>>>
>>>> On 23/02/16 14:03, Poonam Bajaj Parhar wrote:
>>>>> Hello,
>>>>>
>>>>> Please approve the backport of the following fix to 8u-dev:
>>>>>
>>>>> CR: JDK-8133023: ParallelGCThreads is not calculated correctly
>>>>> webrev: http://cr.openjdk.java.net/~poonam/8133023/webrev.8u/
>>>>> Code review thread:
>>>>> http://mail.openjdk.java.net/pipermail/hotspot-gc-dev/2016-February/016671.html 
>>>>>
>>>>>
>>>>>
>>>>> JDK9 changeset:
>>>>> http://hg.openjdk.java.net/jdk9/jdk9/hotspot/rev/0b22be0db834
>>>>>
>>>>> Thanks,
>>>>> Poonam
>>>>>
>>>>>
>>>>>
>>>>>
>>>
>>