diff --git a/level_zero/core/source/cmdlist/cmdlist_hw.inl b/level_zero/core/source/cmdlist/cmdlist_hw.inl
index be890d19e9..9c609dd6b0 100644
--- a/level_zero/core/source/cmdlist/cmdlist_hw.inl
+++ b/level_zero/core/source/cmdlist/cmdlist_hw.inl
@@ -2131,10 +2131,6 @@ inline ze_result_t CommandListCoreFamily<gfxCoreFamily>::addEventsToCmdList(uint
 
     if (hasInOrderDependencies) {
         CommandListCoreFamily<gfxCoreFamily>::appendWaitOnInOrderDependency(this->inOrderDependencyCounterAllocation, this->inOrderDependencyCounter, this->inOrderAllocationOffset, relaxedOrderingAllowed);
-
-        if (NEO::EncodeUserInterruptHelper::isOperationAllowed(NEO::EncodeUserInterruptHelper::afterSemaphoreMask)) {
-            NEO::EnodeUserInterrupt<GfxFamily>::encode(*commandContainer.getCommandStream());
-        }
     }
 
     if (numWaitEvents > 0) {
@@ -2210,6 +2206,10 @@ void CommandListCoreFamily<gfxCoreFamily>::appendWaitOnInOrderDependency(NEO::Gr
 
         gpuAddress += sizeof(uint64_t);
     }
+
+    if (NEO::EncodeUserInterruptHelper::isOperationAllowed(NEO::EncodeUserInterruptHelper::afterSemaphoreMask)) {
+        NEO::EnodeUserInterrupt<GfxFamily>::encode(*commandContainer.getCommandStream());
+    }
 }
 
 template <GFXCORE_FAMILY gfxCoreFamily>
@@ -2232,6 +2232,10 @@ ze_result_t CommandListCoreFamily<gfxCoreFamily>::appendWaitOnEvents(uint32_t nu
         callId = neoDevice->getRootDeviceEnvironment().tagsManager->currentCallCount;
     }
 
+    if (signalInOrderCompletion) {
+        CommandListCoreFamily<gfxCoreFamily>::appendWaitOnInOrderDependency(this->inOrderDependencyCounterAllocation, this->inOrderDependencyCounter, this->inOrderAllocationOffset, relaxedOrderingAllowed);
+    }
+
     bool dcFlushRequired = false;
 
     if (this->dcFlushSupport) {
@@ -2395,12 +2399,9 @@ ze_result_t CommandListCoreFamily<gfxCoreFamily>::appendWriteGlobalTimestamp(
     uint64_t *dstptr, ze_event_handle_t hSignalEvent,
     uint32_t numWaitEvents, ze_event_handle_t *phWaitEvents) {
 
-    if (numWaitEvents > 0) {
-        if (phWaitEvents) {
-            CommandListCoreFamily<gfxCoreFamily>::appendWaitOnEvents(numWaitEvents, phWaitEvents, false, true, false);
-        } else {
-            return ZE_RESULT_ERROR_INVALID_ARGUMENT;
-        }
+    ze_result_t ret = addEventsToCmdList(numWaitEvents, phWaitEvents, false, true);
+    if (ret != ZE_RESULT_SUCCESS) {
+        return ret;
     }
 
     Event *signalEvent = nullptr;
diff --git a/level_zero/core/test/unit_tests/sources/cmdlist/test_cmdlist_append_launch_kernel_3.cpp b/level_zero/core/test/unit_tests/sources/cmdlist/test_cmdlist_append_launch_kernel_3.cpp
index 67a066b73f..8c73943d88 100644
--- a/level_zero/core/test/unit_tests/sources/cmdlist/test_cmdlist_append_launch_kernel_3.cpp
+++ b/level_zero/core/test/unit_tests/sources/cmdlist/test_cmdlist_append_launch_kernel_3.cpp
@@ -1831,6 +1831,7 @@ HWTEST2_F(InOrderCmdListTests, givenCopyOnlyInOrderModeWhenProgrammingCopyRegion
 
 HWTEST2_F(InOrderCmdListTests, givenInOrderModeWhenProgrammingAppendWaitOnEventsThenSignalSyncAllocation, IsAtLeastXeHpCore) {
     using MI_STORE_DATA_IMM = typename FamilyType::MI_STORE_DATA_IMM;
+    using MI_SEMAPHORE_WAIT = typename FamilyType::MI_SEMAPHORE_WAIT;
 
     auto immCmdList = createImmCmdList<gfxCoreFamily>();
 
@@ -1841,6 +1842,7 @@ HWTEST2_F(InOrderCmdListTests, givenInOrderModeWhenProgrammingAppendWaitOnEvents
     auto eventHandle = events[0]->toHandle();
 
     immCmdList->appendLaunchKernel(kernel->toHandle(), &groupCount, eventHandle, 0, nullptr, launchParams, false);
+    immCmdList->appendLaunchKernel(kernel->toHandle(), &groupCount, nullptr, 0, nullptr, launchParams, false);
 
     auto offset = cmdStream->getUsed();
 
@@ -1851,14 +1853,23 @@ HWTEST2_F(InOrderCmdListTests, givenInOrderModeWhenProgrammingAppendWaitOnEvents
                                                       ptrOffset(cmdStream->getCpuBase(), offset),
                                                       (cmdStream->getUsed() - offset)));
 
-    auto sdiItor = find<MI_STORE_DATA_IMM *>(cmdList.begin(), cmdList.end());
+    auto semaphoreItor = find<MI_SEMAPHORE_WAIT *>(cmdList.begin(), cmdList.end());
+    ASSERT_NE(cmdList.end(), semaphoreItor);
+
+    auto semaphoreCmd = genCmdCast<MI_SEMAPHORE_WAIT *>(*semaphoreItor);
+    ASSERT_NE(nullptr, semaphoreCmd);
+
+    EXPECT_EQ(2u, semaphoreCmd->getSemaphoreDataDword());
+    EXPECT_EQ(immCmdList->inOrderDependencyCounterAllocation->getGpuAddress(), semaphoreCmd->getSemaphoreGraphicsAddress());
+
+    auto sdiItor = find<MI_STORE_DATA_IMM *>(semaphoreItor, cmdList.end());
     ASSERT_NE(cmdList.end(), sdiItor);
 
     auto sdiCmd = genCmdCast<MI_STORE_DATA_IMM *>(*sdiItor);
 
     EXPECT_EQ(immCmdList->inOrderDependencyCounterAllocation->getGpuAddress(), sdiCmd->getAddress());
     EXPECT_EQ(0u, sdiCmd->getStoreQword());
-    EXPECT_EQ(2u, sdiCmd->getDataDword0());
+    EXPECT_EQ(3u, sdiCmd->getDataDword0());
 }
 
 HWTEST2_F(InOrderCmdListTests, givenInOrderModeWhenProgrammingCounterWithOverflowThenHandleItCorrectly, IsAtLeastXeHpCore) {