RFR: Cleaner GHA build of hermetic-java-runtime branch

[Jiangli Zhou]

On Tue, 29 Apr 2025 08:24:01 GMT, Severin Gehwolf <sgehwolf at openjdk.org> wrote:

> Is this duplicate symbol issue still a problem? If so, how would I be able to reproduce? This breaks cross compiled code. For example see: https://github.com/jerboaa/leyden/actions/runs/14613421232/job/40996157052

The `Thread` duplicate symbol issue exists when statically linking with any application native code and native dependencies with a symbol also defined as `Thread`. We ran into the issue when prototyping on JDK 11. 


In file included from /home/runner/work/leyden/leyden/src/hotspot/share/cds/aotClassFilter.hpp:29,
                 from /home/runner/work/leyden/leyden/src/hotspot/share/cds/aotClassFilter.cpp:25:
/home/runner/work/leyden/leyden/src/hotspot/share/cds/aotClassFilter.cpp: In destructor ‘AOTClassFilter::FilterMark::~FilterMark()’:
/home/runner/work/leyden/leyden/src/hotspot/share/cds/aotClassFilter.cpp:42:53: error: comparison between distinct pointer types ‘Thread*’ and ‘HotspotBaseThread*’ lacks a cast
   42 |   assert(_current_mark == this && _filtering_thread == Thread::current(), "sanity");


Ok, we also ran into aotClassFilter.* related build failure when syncing with JDK mainline internally in early April. I've resolved the build issue in our code base. I'll push an update to hermetic-java-runtime branch to address this build issue shortly. Thanks for noticing it.

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

PR Review Comment: https://git.openjdk.org/leyden/pull/63#discussion_r2066880516