diff --git a/projects/clr/rocclr/device/rocm/rocblit.cpp b/projects/clr/rocclr/device/rocm/rocblit.cpp
index d25e8e2f4f..807ba4f28f 100644
--- a/projects/clr/rocclr/device/rocm/rocblit.cpp
+++ b/projects/clr/rocclr/device/rocm/rocblit.cpp
@@ -456,6 +456,10 @@ bool DmaBlitManager::copyBufferRect(device::Memory& srcMemory, device::Memory& d
       hsa_signal_t active = gpu().Barriers().ActiveSignal(kInitSignalValueOne, gpu().timestamp());
 
       // Copy memory line by line
+      ClPrint(amd::LOG_DEBUG, amd::LOG_COPY,
+              "[%zx]!\t HSA Asycn Copy Rect  wait_event=0x%zx, completion_signal=0x%zx\n",
+              std::this_thread::get_id(), (wait_event != nullptr) ? wait_event->handle : 0,
+              active.handle);
       hsa_status_t status = hsa_amd_memory_async_copy_rect(&dstMem, &offset,
           &srcMem, &offset, &dim, agent, direction, num_wait_events, wait_event, active);
       if (status != HSA_STATUS_SUCCESS) {
@@ -474,6 +478,10 @@ bool DmaBlitManager::copyBufferRect(device::Memory& srcMemory, device::Memory& d
           size_t dstOffset = dstRect.offset(0, y, z);
 
           // Copy memory line by line
+          ClPrint(amd::LOG_DEBUG, amd::LOG_COPY,
+                  "[%zx]!\t HSA Asycn Copy wait_event=0x%zx, completion_signal=0x%zx\n",
+                  std::this_thread::get_id(), (wait_event != nullptr) ? wait_event->handle : 0,
+                  active.handle);
           hsa_status_t status = hsa_amd_memory_async_copy(
               (reinterpret_cast<address>(dst) + dstOffset), dstAgent,
               (reinterpret_cast<const_address>(src) + srcOffset), srcAgent,
@@ -664,6 +672,11 @@ bool DmaBlitManager::hsaCopy(const Memory& srcMemory, const Memory& dstMemory,
   hsa_signal_t      active = gpu().Barriers().ActiveSignal(kInitSignalValueOne, gpu().timestamp());
 
   // Use SDMA to transfer the data
+  ClPrint(amd::LOG_DEBUG, amd::LOG_COPY,
+          "[%zx]!\t HSA Asycn Copy wait_event=0x%zx, completion_signal=0x%zx\n",
+          std::this_thread::get_id(), (wait_event != nullptr) ? wait_event->handle : 0,
+          active.handle);
+
   status = hsa_amd_memory_async_copy(dst, dstAgent, src, srcAgent,
       size[0], num_wait_events, wait_event, active);
   if (status == HSA_STATUS_SUCCESS) {
@@ -719,6 +732,9 @@ bool DmaBlitManager::hsaCopyStaged(const_address hostSrc, address hostDst, size_
       hsa_signal_t active = gpu().Barriers().ActiveSignal(kInitSignalValueOne, gpu().timestamp());
 
       memcpy(hsaBuffer, hostSrc + offset, size);
+      ClPrint(amd::LOG_DEBUG, amd::LOG_COPY,
+              "[%zx]!\t HSA Async Copy completion_signal=0x%zx\n",
+              std::this_thread::get_id(), active.handle);
       status = hsa_amd_memory_async_copy(hostDst + offset, dev().getBackendDevice(), hsaBuffer,
                                          srcAgent, size, 0, nullptr, active);
       if (status != HSA_STATUS_SUCCESS) {
@@ -748,6 +764,9 @@ bool DmaBlitManager::hsaCopyStaged(const_address hostSrc, address hostDst, size_
     hsa_signal_t active = gpu().Barriers().ActiveSignal(kInitSignalValueOne, gpu().timestamp());
 
     // Copy data from Device to Host
+    ClPrint(amd::LOG_DEBUG, amd::LOG_COPY,
+            "[%zx]!\t HSA Async Copy completion_signal=0x%zx\n",
+            std::this_thread::get_id(), active.handle);
     status = hsa_amd_memory_async_copy(hsaBuffer, dstAgent, hostSrc + offset,
         dev().getBackendDevice(), size, 0, nullptr, active);
     if (status == HSA_STATUS_SUCCESS) {
diff --git a/projects/clr/rocclr/device/rocm/rocvirtual.cpp b/projects/clr/rocclr/device/rocm/rocvirtual.cpp
index a4a6257bb2..31950d3a70 100644
--- a/projects/clr/rocclr/device/rocm/rocvirtual.cpp
+++ b/projects/clr/rocclr/device/rocm/rocvirtual.cpp
@@ -367,6 +367,8 @@ bool VirtualGPU::HwQueueTracker::CpuWaitForSignal(ProfilingSignal* signal) {
     if (signal->ts_ != nullptr) {
       signal->ts_->checkGpuTime();
     } else {
+      ClPrint(amd::LOG_DEBUG, amd::LOG_COPY, "[%zx]!\t Host wait on completion_signal=0x%zx\n",
+              std::this_thread::get_id(), signal->signal_.handle);
       if (!WaitForSignal(signal->signal_)) {
         LogPrintfError("Failed signal [0x%lx] wait", signal->signal_);
         return false;
@@ -518,8 +520,9 @@ bool VirtualGPU::processMemObjects(const amd::Kernel& kernel, const_address para
           gpuMem = static_cast<Memory*>(mem->getDeviceMemory(dev()));
 
           const void* globalAddress = *reinterpret_cast<const void* const*>(params + desc.offset_);
-          ClPrint(amd::LOG_INFO, amd::LOG_KERN, "!\targ%d: %s %s = ptr:%p obj:[%p-%p] threadId : %zx\n", index,
-            desc.typeName_.c_str(), desc.name_.c_str(),
+          ClPrint(amd::LOG_INFO, amd::LOG_KERN,
+            "!\targ%d: %s %s = ptr:%p obj:[%p-%p] threadId : %zx\n",
+            index, desc.typeName_.c_str(), desc.name_.c_str(),
             globalAddress, gpuMem->getDeviceMemory(),
             reinterpret_cast<address>(gpuMem->getDeviceMemory()) + mem->getSize(),
             std::this_thread::get_id());
@@ -668,7 +671,8 @@ bool VirtualGPU::dispatchGenericAqlPacket(
       packet_store_release(reinterpret_cast<uint32_t*>(aql_loc), header, rest);
     }
     ClPrint(amd::LOG_DEBUG, amd::LOG_AQL,
-            "[%zx] HWq=0x%zx, Dispatch Header = 0x%x (type=%d, barrier=%d, acquire=%d, release=%d), "
+            "[%zx] HWq=0x%zx, Dispatch Header = "
+            "0x%x (type=%d, barrier=%d, acquire=%d, release=%d), "
             "setup=%d, grid=[%zu, %zu, %zu], workgroup=[%zu, %zu, %zu], private_seg_size=%zu, "
             "group_seg_size=%zu, kernel_obj=0x%zx, kernarg_address=0x%zx, completion_signal=0x%zx",
             std::this_thread::get_id(), gpu_queue_,
@@ -2361,7 +2365,8 @@ bool VirtualGPU::submitKernelInternal(const amd::NDRangeContainer& sizes, const
       return false;
     }
 
-    ClPrint(amd::LOG_INFO, amd::LOG_KERN, "[%zx]!\tShaderName : %s\n", std::this_thread::get_id(), gpuKernel.name().c_str());
+    ClPrint(amd::LOG_INFO, amd::LOG_KERN, "[%zx]!\tShaderName : %s\n",
+            std::this_thread::get_id(), gpuKernel.name().c_str());
 
     // Check if runtime has to setup hidden arguments
     for (uint32_t i = signature.numParameters(); i < signature.numParametersAll(); ++i) {