[jdk17u-ri] RFR: 8297878: KEM: Implementation [v4]

[Weijun Wang]

On Thu, 25 Apr 2024 23:58:43 GMT, Andrew John Hughes <andrew at openjdk.org> wrote:

>> Weijun Wang has updated the pull request incrementally with one additional commit since the last revision:
>> 
>>   update tags of inner types
>
> It would be easier to review if [JDK-8322971](https://bugs.openjdk.org/browse/JDK-8322971) and [JDK-8305846](https://bugs.openjdk.org/browse/JDK-8305846): "Support compilation in Proc test utility" were backported separately.  Also, 8322971 seems to be missing from 21u as well.
> 
> The changes look correct as far as I can tell. I'm not sure why different copyright years are being used from the original patch, particularly for `Provider.java` where this is the only difference from 8297878.
> 
> The other differences (21->17, removal of the implementation, SSL & ECC changes) make sense. I presume the test differences with `getKemImpl` are due to the implementation removal making `KEM.getInstance` unusable?
> 
> So Skara correctly recognises this as a backport, the title should be "Backport 6b90b0519e89429300838fa598b2ea9ffda984a2"

@gnu-andrew I backported the 3 issues in a single PR because [JDK-8305846](https://bugs.openjdk.org/browse/JDK-8305846) was originally created to support the RSA-KEM interop test. I understand the change for it looks a little strange without any usage of it here. I hope grouping them together will make further backport - if exists - simpler.

For the year of `Provider.java`, I just updated it to *now*.

Yes, `getKemImpl` exists because after JDK-8322971 one cannot load a KEM impl from an unsigned security provider in Oracle JDK. This is a little different from OpenJDK.

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

PR Comment: https://git.openjdk.org/jdk17u-ri/pull/1#issuecomment-2082661511