switch to EPSGetErrorEstimate as others might give inconsistent results

diff --git a/source/lac/slepc_solver.cc b/source/lac/slepc_solver.cc
index 5bfcd6cac18c04b3172332f4ebdd91fac2377163..0aea6ad11800a8d5aed26cb87f2731cb1c9f79c4 100644 (file)
--- a/source/lac/slepc_solver.cc
+++ b/source/lac/slepc_solver.cc
@@ -211,11 +211,9 @@ namespace SLEPcWrappers
       // only if at least one eigenvector has converged.
       if ((*n_converged)>0)
         {
-          // EPSComputeRelativeError is consistent with the one
-          // used during the solution process. Given EPS_CONV_ABS set above,
-          // this can be either the l2 norm or the mass-matrix induced norm
-          // when EPS_GHEP is set.
-          ierr = EPSComputeRelativeError (solver_data->eps, 0, &residual_norm);
+          // EPSGetErrorEstimate is consistent with the residual norm
+          // used during the solution process.
+          ierr = EPSGetErrorEstimate (solver_data->eps, 0, &residual_norm);
           AssertThrow (ierr == 0, ExcSLEPcError(ierr));
         }