Encryption fix in HotSpot

[Aleksey Shipilev]

On 2/28/19 12:36 PM, Andrew Haley wrote:
> On 2/27/19 11:12 PM, Aleksey Shipilev wrote:
>> On 2/26/19 2:59 PM, Aleksey Shipilev wrote:
>>> Once 11.0.4 opens, we can start backporting these. You are welcome to do it :)
>>
>> Ah, that requires some benchmarking.
> 
> Um, why? The benchmarking, if it were necessary, should have been done
> on the upstream fix.

I'd like performance bugs to be verified during backports, if possible. Especially for
compiler/intrinsic bugs that have a high chance to interact with something else and turn out to be
the pessimization. This is not a requirement for those who do Fix Requests, but the more we know
about things we are backporting, the better. Saves time fixing the post-integration regressions.

> If you really think that special benchmarking for 11u is needed, please be specific about the
> form you believe it should take. Newcomers shouldn't have to guess.

I have already did it in my Fix Request, there is a link with performance results:
  https://bugs.openjdk.java.net/browse/JDK-8201633

-Aleksey