Stop language server on shutdown (#681)

[basilevs]

The problem (#681):
Language server is not stopped on Eclipse shutdown, as LanguageServerWrapper does not call stop() before stopping its thread pools.

Solution:

• make stop() observable
• wait for stop() to complete on Eclipse shutdown
• fix synchronization in background threads
• use owned dispatcher when applicable
• move LanguageWrapperTest to be able to call package private method

Concerns:
Test is slow locally - shutdown command times out for no apparent reason
Synchronization fixes may lead to deadlocks.

[basilevs]

I now understand, why there is LanguageServiceAcessor.shutdownAllDispatchers().
This method shuts down the dispatchers in a non-recoverable manner, without removing them from the registry. This prevents their accidental recreation during shutdown process.
Unfortunately this means, that tests testing shutdown log a bunch of exceptions.
Further work is required to reconcile the requirement of immutable termination and test cleanup.

UPDATE: done by preventing repeating calls to stopDispatchers() from calling stop().

[basilevs]

The failing Junit test seems to be unstable in master.

[rubenporras]

Hi @basilevs,

I have realized that this PR has been waiting for 9 months for my review. Sorry for that.

Would you still like to merge the PR after resolving the conflicts? I did not look at the conflicts, but might any of the conflicts have resolved the problem you wanted to fix?

Regards

[basilevs]

Failed to execute goal org.eclipse.tycho.extras:tycho-p2-extras-plugin:4.0.6:compare-version-with-baselines (compare-version-with-baseline) on project org.eclipse.lsp4e.tests.mock: Only qualifier changed for (org.eclipse.lsp4e.tests.mock/0.16.8.202403191733). Expected to have bigger x.y.z than what is available in baseline (0.16.8.202402141138) -> [Help 1]

[rubenporras]

Hi @basilevs , you should do the bump as part of this PR, not as a new PR. Regarding the error that you got in the build of #948 , I will ask for help, I do not know what causes it yet.

[basilevs]

Hi @basilevs , you should do the bump as part of this PR, not as a new PR.

But this PR would have tens of version bumps this way? Surely that is impractical?
Also, this PR has nothing to do with release management.

[rubenporras]

Yes, it is not very practical, but it is also not a big deal, and it is so far how the project has handled it. I think the reason is not to bump plugin versions unless there is a code change.
You could create also a PR with two commits, if you want to not mix all in one commit.

[basilevs]

Yes, it is not very practical, but it is also not a big deal, and it is so far how the project has handled it. I think the reason is not to bump plugin versions unless there is a code change. You could create also a PR with two commits, if you want to not mix all in one commit.

Well, anyway, a separate PR in this case is important, as it demonstrates, that builds are broken and meaningful changes in other PRs are not the cause.

[basilevs]

Tests are failing. Probably due to merge side-effects. Further work is required.

java.lang.AssertionError
	at org.junit.Assert.fail(Assert.java:87)
	at org.junit.Assert.assertTrue(Assert.java:42)
	at org.junit.Assert.assertTrue(Assert.java:53)
	at org.eclipse.lsp4e.LanguageServerWrapperTest.testConnect(LanguageServerWrapperTest.java:73)

java.lang.OutOfMemoryError: unable to create native thread: possibly out of memory or process/resource limits reached
	at java.base/java.lang.Thread.start0(Native Method)
	at java.base/java.lang.Thread.start(Thread.java:802)
	at java.base/java.util.Timer.<init>(Timer.java:188)
	at java.base/java.util.Timer.<init>(Timer.java:170)
	at org.eclipse.lsp4e.LanguageServerWrapper.<init>(LanguageServerWrapper.java:164)
	at org.eclipse.lsp4e.LanguageServerWrapper.<init>(LanguageServerWrapper.java:181)
	at org.eclipse.lsp4e.LanguageServiceAccessor.getLSWrapper(LanguageServiceAccessor.java:378)
	at org.eclipse.lsp4e.LanguageServiceAccessor.getLSWrappers(LanguageServiceAccessor.java:253)
	at org.eclipse.lsp4e.LanguageServerWrapperTest.doNotStopBusyDispatchers(LanguageServerWrapperTest.java:149)

[basilevs]

My project using LSP4E is suspended, so I'm dropping this work.