[OperationalCredentials]: Reporting change for the attributes not managed by the attribute store.

src/app/clusters/operational-credentials-server/operational-credentials-server.cpp

@@ -28,6 +28,7 @@
 #include <app/AttributeAccessInterface.h>
 #include <app/CommandHandler.h>
 #include <app/ConcreteCommandPath.h>
+#include <app/reporting/reporting.h>
 #include <app/server/Dnssd.h>
 #include <app/server/Server.h>
 #include <app/util/af.h>
@@ -263,6 +264,10 @@ CHIP_ERROR writeFabricsIntoFabricsListAttribute()
         err = CHIP_ERROR_PERSISTED_STORAGE_FAILED;
     }
 
+    // Currently, we only manage FabricsList attribute in endpoint 0, OperationalCredentials cluster is always required to be on
+    // EP0.
+    MatterReportingAttributeChangeCallback(0, OperationalCredentials::Id, OperationalCredentials::Attributes::FabricsList::Id);
+
     return err;
 }
 
@@ -785,6 +790,11 @@ bool emberAfOperationalCredentialsClusterAddTrustedRootCertificateCallback(
         gFabricBeingCommissioned.Reset();
         emberAfPrintln(EMBER_AF_PRINT_DEBUG, "OpCreds: Failed AddTrustedRootCert request.");
     }
+    else
+    {
+        MatterReportingAttributeChangeCallback(commandPath.mEndpointId, OperationalCredentials::Id,
+                                               OperationalCredentials::Attributes::TrustedRootCertificates::Id);
+    }
 
     return true;
 }
@@ -795,5 +805,9 @@ bool emberAfOperationalCredentialsClusterRemoveTrustedRootCertificateCallback(
 {
     EmberAfStatus status = EMBER_ZCL_STATUS_FAILURE;
     emberAfSendImmediateDefaultResponse(status);
+
+    MatterReportingAttributeChangeCallback(commandPath.mEndpointId, OperationalCredentials::Id,
+                                           OperationalCredentials::Attributes::TrustedRootCertificates::Id);
+
     return true;
 }

src/app/reporting/Engine.cpp

@@ -334,8 +334,14 @@ void Engine::Run(System::Layer * aSystemLayer, void * apAppState)
 
 CHIP_ERROR Engine::ScheduleRun()
 {
+    if (mScheduleRunPending)
+    {
+        return CHIP_NO_ERROR;
+    }
+
     if (InteractionModelEngine::GetInstance()->GetExchangeManager() != nullptr)
     {
+        mScheduleRunPending = true;
         return InteractionModelEngine::GetInstance()->GetExchangeManager()->GetSessionManager()->SystemLayer()->ScheduleWork(Run,
                                                                                                                              this);
     }
@@ -352,6 +358,8 @@ void Engine::Run()
     InteractionModelEngine * imEngine = InteractionModelEngine::GetInstance();
     ReadHandler * readHandler         = imEngine->mReadHandlers + mCurReadHandlerIdx;
 
+    mScheduleRunPending = false;
+
     while ((mNumReportsInFlight < CHIP_IM_MAX_REPORTS_IN_FLIGHT) && (numReadHandled < CHIP_IM_MAX_NUM_READ_HANDLER))
     {
         if (readHandler->IsReportable())

src/app/reporting/Engine.h

@@ -121,6 +121,13 @@ class Engine
      */
     bool mMoreChunkedMessages = false;
 
+    /**
+     * Boolean to indicate if ScheduleRun is pending. This flag is used to prevent calling ScheduleRun multiple times
+     * within the same excution context.
+     *
+     */
+    bool mScheduleRunPending = false;
+
     /**
      * The number of report date request in flight
      *

src/app/util/ember-compatibility-functions.cpp

@@ -504,4 +504,9 @@ void MatterReportingAttributeChangeCallback(EndpointId endpoint, ClusterId clust
     info.mFlags.Set(ClusterInfo::Flags::kFieldIdValid);
 
     InteractionModelEngine::GetInstance()->GetReportingEngine().SetDirty(info);
+
+    // Schedule a run work asynchronously on the CHIP thread, the scheduled work won't execute until the current execution context
+    // has completed. This ensures that we can 'gather up' multiple attribute changes that have occurred in the same execution
+    // context without requiring any explicit 'start' or 'end' change calls into the engine to book-end the change.
+    InteractionModelEngine::GetInstance()->GetReportingEngine().ScheduleRun();
 }