test_runner: make end of work check stricter

This commit updates the logic that checks for the end of the
test run. Prior to this change, it was possible for root.run() to
be called multiple times because of the way pending subtests
were tracked. The extra calls to root.run() were harmless, but
could trigger an EventEmitter leak warning due to 'abort'
listeners being created.

lib/internal/test_runner/test.js

@@ -822,11 +822,20 @@ class Test extends AsyncResource {
         this.parent.activeSubtests--;
       }
 
+      // The call to processPendingSubtests() below can change the number of
+      // pending subtests. When detecting if we are done running tests, we want
+      // to check if there are no pending subtests both before and after
+      // calling processPendingSubtests(). Otherwise, it is possible to call
+      // root.run() multiple times (which is harmless but can trigger an
+      // EventEmitter leak warning).
+      const pendingSiblingCount = this.parent.pendingSubtests.length;
+
       this.parent.addReadySubtest(this);
       this.parent.processReadySubtestRange(false);
       this.parent.processPendingSubtests();
 
       if (this.parent === this.root &&
+          pendingSiblingCount === 0 &&
           this.root.activeSubtests === 0 &&
           this.root.pendingSubtests.length === 0 &&
           this.root.readySubtests.size === 0) {

test/parallel/test-runner-filter-warning.js

@@ -0,0 +1,11 @@
+// Flags: --test-only
+'use strict';
+const common = require('../common');
+const { test } = require('node:test');
+const { defaultMaxListeners } = require('node:events');
+
+process.on('warning', common.mustNotCall());
+
+for (let i = 0; i < defaultMaxListeners + 1; ++i) {
+  test(`test ${i + 1}`);
+}