8162795: was: Re: [8u-dev] Request for approval: backport 8152271: MemberNameTable doesn't purge stale entries

[Kevin Walls]

Hi again,

I'm back: 8152271 has been re-applied to 9.  While it did cause a 
microbenchmark regression, certain apps have a more serious regression 
between jdk7 and 8, which is remedied by 8152271.  The bug for the 
re-application is:

8162795: [REDO] MemberNameTable doesn't purge stale entries
https://bugs.openjdk.java.net/browse/JDK-8162795


So this is a request for approval to backport 8162795 into jdk8u-dev.  
The 9 changeset imports directly except for one copyright year difference.

approval mails in 9: 
http://mail.openjdk.java.net/pipermail/hotspot-runtime-dev/2017-February/022620.html
9 changeset: http://hg.openjdk.java.net/jdk9/dev/hotspot/rev/b1606e8fee8c
(which is Coleen's original change from JDK-8152271 with one copyright 
year change)

Many thanks
Kevin


On 27/01/2017 12:48, Kevin Walls wrote:
>
> Oh, thanks David.  There was me getting excited it solved all my 
> problems...
>
> Thanks
> Kevin
>
>
> On 27/01/2017 11:04, David Holmes wrote:
>> Hi Kevin,
>>
>> On 27/01/2017 8:25 PM, Kevin Walls wrote:
>>> Hi,
>>>
>>> I'd like to backport the following to 8u-dev.
>>>
>>> 8152271: MemberNameTable doesn't purge stale entries
>>>
>>> https://bugs.openjdk.java.net/browse/JDK-8152271
>>
>> That change was backed out again:
>>
>> https://bugs.openjdk.java.net/browse/JDK-8161445
>>
>> David
>> -----
>>
>>> An hg import from the 9 changeset:
>>>
>>> http://hg.openjdk.java.net/jdk9/jdk9/hotspot/rev/13b2c7ac95a5
>>>
>>> ...works and tests well in jprt and with a manual testcase that 
>>> showed a
>>> gc slowdown before the fix.
>>>
>>>
>>> Thanks
>>> Kevin
>>>
>>>
>