Added doc

git-svn-id: https://svn.dealii.org/trunk@4958 0785d39b-7218-0410-832d-ea1e28bc413d

diff --git a/deal.II/lac/include/lac/solver_bicgstab.h b/deal.II/lac/include/lac/solver_bicgstab.h
index 28d562df232d833300214042861af19e15aa7be7..4bb43692790f1198ff8eccde817575b93eb73272 100644 (file)
--- a/deal.II/lac/include/lac/solver_bicgstab.h
+++ b/deal.II/lac/include/lac/solver_bicgstab.h
@@ -43,7 +43,8 @@
  * the size of the orthogonalized residual computed during the
  * iteration may be larger by orders of magnitude than the true
  * residual. This is due to numerical instabilities related to badly
- * conditioned matrices.
+ * conditioned matrices. Since this instability results in a bad
+ * stopping criterion, the default for this parameter is @p{true}.
  *
  * The second parameter is the size of a breakdown criterion. It is
  * difficult to find a general good criterion, so if things do not
@@ -69,11 +70,11 @@ class SolverBicgstab : public Solver<VECTOR>
                                         /**
                                          * Constructor.
                                          *
-                                         * The default is no exact residual
+                                         * The default is exact residual
                                          * computation and breakdown
                                          * parameter 1e-16.
                                          */
-       AdditionalData(bool exact_residual = false,
+       AdditionalData(bool exact_residual = true,
                       double breakdown=1.e-10) :
                        exact_residual(exact_residual),
                        breakdown(breakdown)