Add changelog

doc/news/changes/incompatibilities/20190603MartinKronbichler

@@ -0,0 +1,12 @@
+Changed: The functions DoFHandler::n_locally_owned_dofs_per_processor(),
+DoFHandler::locally_owned_dofs_per_processor() and
+DoFHandler::locally_owned_mg_dofs_per_processor() previously return a
+reference to an internally stored array of index sets on all processors. As
+this cannot scale to large processor counts, these functions have been marked
+deprecated and only populate the internal vectors on the first demand. Use the
+new functions DoFHandler::compute_n_locally_owned_dofs_per_processor(),
+DoFHandler::compute_locally_owned_dofs_per_processor() and
+DoFHandler::compute_locally_owned_mg_dofs_per_processor() instead or, even
+better for scalability, avoid them in favor of some local communication.
+<br>
+(Martin Kronbichler, 2019/06/03)

doc/news/changes/minor/20190603MartinKronbichlerPeterMunch

@@ -0,0 +1,6 @@
+Improved: A number of places in deal.II previously called `MPI_Allgather`
+followed by some local sums to compute prefix sums. Now the more appropriate
+`MPI_Exscan` is used instead, which improves scalability of setup routines on
+more than 10k processors.
+<br>
+(Martin Kronbichler, Peter Munch, 2019/06/03)