Bug: Misattributed Java.net 8u/11u builds

[shipilev]

Bug report

If you look at the current sdk java list here, then you would see the 11.0.10 ... 11.0.12 releases for the "Java.net" vendor:

================================================================================
Available Java Versions
================================================================================
 Vendor        | Use | Version      | Dist    | Status     | Identifier
--------------------------------------------------------------------------------
 Java.net      |     | 18.ea.27     | open    |            | 18.ea.27-open       
               |     | 18.ea.26     | open    |            | 18.ea.26-open       
               |     | 18.ea.7.lm   | open    |            | 18.ea.7.lm-open     
               |     | 17           | open    |            | 17-open             
               |     | 17.ea.3.pma  | open    |            | 17.ea.3.pma-open    
               |     | 17.0.1       | open    |            | 17.0.1-open         
               |     | 16.0.2       | open    |            | 16.0.2-open         
               | >>> | 11.0.12      | open    | installed  | 11.0.12-open        
               |     | 11.0.11      | open    |            | 11.0.11-open        
               |     | 11.0.10      | open    |            | 11.0.10-open        
               |     | 11.0.2       | open    |            | 11.0.2-open         
               |     | 8.0.302      | open    |            | 8.0.302-open        
               |     | 8.0.292      | open    |            | 8.0.292-open        
               |     | 8.0.282      | open    |            | 8.0.282-open        
               |     | 8.0.265      | open    |            | 8.0.265-open   

But where do those binaries came from? There is no 11.0.10...11.0.12 in the jdk.java.net Archive. Same goes for 8.0.265...8.0.302.

These binaries have the metadata of OpenJDK builds:

$ java -version
openjdk version "11.0.12" 2021-07-20
OpenJDK Runtime Environment 18.9 (build 11.0.12+7)
OpenJDK 64-Bit Server VM 18.9 (build 11.0.12+7, mixed mode)

$ java -Xinternalversion
OpenJDK 64-Bit Server VM (11.0.12+7) for linux-amd64 JRE (11.0.12+7), built on Jul 14 2021 12:53:10 by "openjdk" with gcc 4.4.7 20120313 (Red Hat 4.4.7-23)

...and they do not include Shenandoah GC:

$ java -XX:+UseShenandoahGC
Unrecognized VM option 'UseShenandoahGC'
Error: Could not create the Java Virtual Machine.
Error: A fatal exception has occurred. Program will exit.

...which suggests they were built by Oracle? If so, they should be labeled as such, since current "Vendor" is misleading. There is a major caveat in the ecosystem that Oracle does not build their 11u builds from upstream OpenJDK, which makes their attribution as "-open" even more misleading.

To reproduce

$ sdk list java

...as above

System info

Linux x86_64, Xubuntu 20.04

[marc0der]

I'm moving this to db migrations for now so that @eddumelendez can pick this up.

[eddumelendez]

@shipilev thanks for reporting this one and glad someone who can help us figure it out how to solve this. For more context see here. I just realized those builds are provided by Red Hat and previously there was a lot of confusion so we moved them from *-open-adpt to *-open. Make sense to move them to *-redhat? I am wondering it this is correct or we just should drop them.