RFR: 8346123: [REDO] NMT should not use ThreadCritical [v12]

[Robert Toyonaga]

On Mon, 6 Jan 2025 05:00:01 GMT, David Holmes <dholmes at openjdk.org> wrote:

> In case my comment within the existing conversations gets missed I will re-state that I don't think you need any new "is bootstrapping" logic because you can just use `Threads::number_of_threads() > 0` to tell you if you need to acquire the NMT lock. Though that assumes that the `WatcherThread` does not use NMT ... but in that case you can use `WatcherThread::watcher_thread() != nullptr` as the bootstrap condition instead.

I think that the `WatcherThread` does use NMT (at least upon starting), since `Thread::call_run() ` calls `register_thread_stack_with_NMT()`. However, it looks like `WatcherThread::stop()` is called early in `before_exit` -- after which point NMT can still  be used. So checking `WatcherThread::watcher_thread() != nullptr` may not work in this case as it could be set back to nullptr too early.

Another solution might be to introduce something like `Threads::is_single_threaded()` which checks for  `WatcherThread`, `Threads::number_of_threads()`, and any other threads that fall outside this set. I'm not sure whether that would be better than the current approach.

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

PR Comment: https://git.openjdk.org/jdk/pull/22745#issuecomment-2573756615