Fix TearDown issue of some App Tests

Author: Alami-Amine

src/app/tests/AppTestContext.cpp

@@ -49,6 +49,12 @@ void AppContext::SetUpTestSuite()
 
 void AppContext::TearDownTestSuite()
 {
+    // Adding a DrainAndServiceIO call fixes the teardown for some tests (TestAclAttribute and TestReportScheduler); these were
+    // triggering failures in tests that follow them when running on nRF CI (Zephyr native_posix where all Unit Tests are compiled
+    // into a single file)
+    // TODO: understand this more and solve underlying issue
+    LoopbackMessagingContext::DrainAndServiceIO();
+
     chip::DeviceLayer::PlatformMgr().Shutdown();
     LoopbackMessagingContext::TearDownTestSuite();
 }

src/app/tests/BUILD.gn

@@ -231,11 +231,12 @@ chip_test_suite_using_nltest("tests_nltest") {
 
   # The platform manager is not properly clearing queues in test teardown, which results in
   # DrainIO calls not being able to run in expected time (5seconds) if unprocessed reported engine
-  # runs are remaining, causing tests to crash in Open IoT SDK and Zephyr tests since they are
+  # runs are remaining, causing tests to crash in Open IoT SDK since they are
   # running all tests in one file. We need to figure out how to properly clean the event queues
   # before enabling this test for these platforms.
-  if (chip_device_platform != "nrfconnect" &&
-      chip_device_platform != "openiotsdk") {
+  # Update: This was fixed for Nordic Zephyr Tests by calling DrainAndServiceIO in AppContext::TearDownTestSuite()
+  # TODO: check if above solution works for openiotsdk as well
+  if (chip_device_platform != "openiotsdk") {
     test_sources += [ "TestReportScheduler.cpp" ]
   }