📜 Description

NoOpSentryExecutorService returned new FutureTask<>(() -> null) from submit/schedule — a task that is never run and never cancelled. Callers had no way to distinguish it from a genuinely queued task: isCancelled() and isDone() both stay false forever, and get() blocks until its timeout.

It now returns an already-cancelled Future, matching what SentryExecutorService already does when its work queue is full. CancelledFuture moves out of SentryExecutorService into a shared package-private class, and ISentryExecutorService documents the requirement so future implementations honour it.

No public API change — CancelledFuture is package-private and no existing signature changed shape, so .api files are untouched.

💡 Motivation and Context

Found by cursor here. That PR's PersistingScopeObserver uses future.isCancelled() to detect a flush task the executor dropped, and releases its hasPendingFlush latch when it sees one. Against a no-op executor the latch would be set with no task to ever clear it, silently stopping scope persistence for the rest of the process.

That is not a reachable bug today because the observer is only installed after options.activate() has swapped in a real executor so this more to prevent regressions in the future. But setExecutorService is public, and the contract the caller relies on was never written down anywhere. Fixing it here, off main, keeps #5791 focused and makes the guarantee available to any other caller that wants to check whether its task was accepted.

💚 How did you test it?

Tests pass

📝 Checklist

• I added GH Issue ID & Linear ID
• I added tests to verify the changes.
• No new PII added or SDK only sends newly added PII if sendDefaultPII is enabled.
• I updated the docs if needed.
• I updated the wizard if needed.
• Review from the native team if needed.
• No breaking change or entry added to the changelog.
• No breaking change for hybrid SDKs or communicated to hybrid SDKs.

🔮 Next steps

#5791 builds on this guarantee for its scope-write batching.