force flush of command lists & GPU sync when doing non-partial replay

* In some cases there was an unpredictable crash when starting a replay
  loop, doing a full frame replay immediately after a previous replay.
  There were no debug messages, and the debug layer fixed it - as did
  enabling single flush validate. It seems to me some timing/overloading
  issue, and doing this sync should be relatively harmless so we use
  this to avoid the crash.

Author: baldurk

renderdoc/driver/d3d12/d3d12_device.cpp

@@ -2493,6 +2493,8 @@ void WrappedID3D12Device::ReplayLog(uint32_t startEventID, uint32_t endEventID,
     // signalled or waited. So instead we just signal a dummy fence each new 'frame'
     for(size_t i = 0; i < m_Queues.size(); i++)
       m_Queues[i]->Signal(m_QueueFences[i], m_GPUSyncCounter);
+
+    FlushLists(true);
   }
 
   D3D12ChunkType header = (D3D12ChunkType)m_pSerialiser->PushContext(NULL, NULL, 1, false);