SWDEV-301667 - Use large signal pool

Use large signal pool if profiler is connected or profiling forced
enabled. This is needed to mitigate signal creation overhead when
profiling as signals are attached to every packet and deeper batch may
show overhead of signal allocation.

Change-Id: I8034b8a20b55328b87d593bf044f59672f9653e8

rocclr/device/rocm/rocvirtual.cpp

@@ -341,6 +341,9 @@ VirtualGPU::HwQueueTracker::~HwQueueTracker() {
 // ================================================================================================
 bool VirtualGPU::HwQueueTracker::Create() {
   uint kSignalListSize = ROC_SIGNAL_POOL_SIZE;
+  if (activity_prof::IsEnabled(OP_ID_DISPATCH) || gpu_.profiling_) {
+    kSignalListSize = !flagIsDefault(ROC_SIGNAL_POOL_SIZE) ? ROC_SIGNAL_POOL_SIZE : 4 * Ki;
+  }
   signal_list_.resize(kSignalListSize);
 
   hsa_agent_t agent = gpu_.gpu_device();

rocclr/platform/commandqueue.cpp

@@ -42,6 +42,9 @@ HostQueue::HostQueue(Context& context, Device& device, cl_command_queue_properti
       tail_(nullptr),
       isActive_(false),
       markerTsCount_(0) {
+  if (GPU_FORCE_QUEUE_PROFILING) {
+    properties().set(CL_QUEUE_PROFILING_ENABLE);
+  }
   if (AMD_DIRECT_DISPATCH) {
     // Initialize the queue
     thread_.Init(this);
@@ -52,10 +55,6 @@ HostQueue::HostQueue(Context& context, Device& device, cl_command_queue_properti
       queueLock_.wait();
     }
   }
-
-  if (GPU_FORCE_QUEUE_PROFILING) {
-    properties().set(CL_QUEUE_PROFILING_ENABLE);
-  }
 }
 
 bool HostQueue::terminate() {
@@ -239,7 +238,7 @@ void HostQueue::append(Command& command) {
 
   // Set last submitted command
   Command* prevLastEnqueueCommand = nullptr;
- 
+
   // Attach only real commands and skip internal notifications for CPU queue
   if (command.waitingEvent() == nullptr) {
     command.retain();