[8u] RFR: 8256682: JDK-8202343 is incomplete

[Martin Balao]

On 2/11/21 2:21 PM, Severin Gehwolf wrote:
> 
> Coming back to this, I'm not 100% sure what Martin meant with this.
> Status quo in 8u-dev is that NullHostnameCheck.java fails for runs with
> 'TLSv1' and 'TLSv1.1', passes with 'TLSv1.2' and 'TLSv1.3'.
> 
> I believe it's because this reasoning was *before* JDK-8234728 got
> backported. With that backported it now also runs for TLSv1, TLSv1.1
> and TLSv1.3.
> 

Yes, that's the reason: the comment was previous to decide backporting
JDK-8234728 to 8u.

>> I must admit it doesn't make sense to me why it would be TLS 1.2 only,
>> but have runs that pass a TLSv1, TLSv1.1 and TLSv1.3 argument to it.
> 
> See above.
> 
>> Are you seeing test failures due to the lack of this block? It would
>> be good to know why we need to revisit the original decision on
>> including this block.
> 
> I can confirm the proposed patch fixes the test. As to the reason, the
> referenced mailing list thread's argument was prior Martin's decision
> to backport JDK-8234728[i]. That backport was done rendering this
> (original) argument wrong. Martin, please correct me if I'm wrong here.
> 

You are right.

The proposed backport looks good to me.

Thanks,
Martin.-