performance: don't wait for paging fence on user thread

Related-To: NEO-12197

Currently for new resources user thread must wait before submitting
actual workload. With this commit, instead of waiting on user thread,
request is sent to background ULLS controller thread and additional
semaphore is programmed. ULLS controller will perform actual wait
and signal semaphore when paging fence reaches required value.

Signed-off-by: Szymon Morek <szymon.morek@intel.com>

shared/source/command_stream/submissions_aggregator.h

@@ -19,6 +19,11 @@ class Event;
 class FlushStampTracker;
 class GraphicsAllocation;
 
+struct PagingFenceSemaphoreInfo {
+    bool requiresBlockingResidencyHandling = true;
+    uint64_t pagingFenceValue = 0u;
+};
+
 struct BatchBuffer {
     BatchBuffer(GraphicsAllocation *commandBufferAllocation,
                 size_t startOffset,
@@ -53,6 +58,8 @@ struct BatchBuffer {
     void *endCmdPtr = nullptr;
     uint32_t numCsrClients = 0;
 
+    PagingFenceSemaphoreInfo pagingFenceSemInfo{};
+
     bool hasStallingCmds = false;
     bool hasRelaxedOrderingDependencies = false;
     bool disableFlatRingBuffer = false;