[jdk11u-dev] RFR: 8179503: Java should support GET OCSP calls [v2]

[Christoph Langer]

On Wed, 17 May 2023 06:48:31 GMT, Christoph Langer <clanger at openjdk.org> wrote:

>>> This issue causes a regression, if the OCSP server is not returning the `Content-Length` header which at least happens now and then on a DigiCert server. In that case the `contentLength` is set to `Integer.MAX_VALUE` and `IOUtils.readExactlyNBytes` fails with an `EOFException` because it can't read `Integer.MAX_VALUE` bytes
>>> 
>>> See also [phax/phase4#124](https://github.com/phax/phase4/issues/124) for my original analysis on the Java 17 problem
>> 
>> Thanks a lot, Philip! I will keep it on mind.
>
>> This issue causes a regression, if the OCSP server is not returning the `Content-Length` header which at least happens now and then on a DigiCert server. In that case the `contentLength` is set to `Integer.MAX_VALUE` and `IOUtils.readExactlyNBytes` fails with an `EOFException` because it can't read `Integer.MAX_VALUE` bytes
>> 
>> See also [phax/phase4#124](https://github.com/phax/phase4/issues/124) for my original analysis on the Java 17 problem
> 
> This is recorded in JBS now as [JDK-8308255](https://bugs.openjdk.org/browse/JDK-8308255).

> @RealCLanger
> 
> Hi, Chris! As I see you've marked dependent [JDK-8274471](https://bugs.openjdk.org/browse/JDK-8274471) with `jdk11u-todo`. Does it mean I should keep working on this and create depended PR(s)? Previously this one was rejected with `jdk11u-fix-request-no` as inappropriate for jdk11

Hi @apavlyutkin,
I have removed jdk11u-fix-no from [JDK-8179503](https://bugs.openjdk.org/browse/JDK-8179503) now, as I think it was backported for Oracle as well (although the backport is not public) and there is justification for backporting this change as well as [JDK-8274471](https://bugs.openjdk.org/browse/JDK-8274471).
You need not care about the jdk11u-todo label. It is just some flagging that helps us keep the JBS filters for open backports lean. You could remove the label once a backport to 11u was merged. But you can also keep the label around for us to clean up.

So, please go ahead with your efforts to backport these two items. As for this PR, I think a merge with master would clean it up and also trigger GHA, hoping for all green results.

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

PR Comment: https://git.openjdk.org/jdk11u-dev/pull/847#issuecomment-1550853092