Synchronize after copying in 2D reduced diagnostics (#5726)

Add `Gpu::streamSynchronize();` after copying `TableData` from either
host-to-device and device-to-host in the `ParticleHistogram2D` and
`DifferentialLuminosity2D` reduced diagnostics.

We have observed unstable behaviors with the `DifferentialLuminosity2D`
diagnostic and @RemiLehe suggested this could be a reason.

Marked as bug, unless otherwise recommended.

---------

Co-authored-by: Remi Lehe <remi.lehe@normalesup.org>

Source/Diagnostics/ReducedDiags/DifferentialLuminosity2D.cpp

@@ -315,6 +315,7 @@ void DifferentialLuminosity2D::ComputeDiags (int step)
 
         // Copy data from GPU memory
         m_h_data_2D.copy(m_d_data_2D);
+        Gpu::streamSynchronize();
 
         // reduced sum over mpi ranks
         const int size = static_cast<int> (m_d_data_2D.size());

Source/Diagnostics/ReducedDiags/ParticleHistogram2D.cpp

@@ -149,6 +149,8 @@ void ParticleHistogram2D::ComputeDiags (int step)
     }
 
     d_data_2D.copy(m_h_data_2D);
+    // Gpu::streamSynchronize() is not needed, because there is a sync in WarpXParIter constructor later
+
     auto d_table = d_data_2D.table();
 
     // get a reference to WarpX instance
@@ -244,6 +246,7 @@ void ParticleHistogram2D::ComputeDiags (int step)
 
     // Copy data from GPU memory
     m_h_data_2D.copy(d_data_2D);
+    Gpu::streamSynchronize();
 
     // reduced sum over mpi ranks
     const int size = static_cast<int> (d_data_2D.size());