From: Wolfgang Bangerth Date: Thu, 22 Jun 2017 15:34:23 +0000 (-0600) Subject: Don't just store references, but SmartPointer objects. X-Git-Tag: v9.0.0-rc1~1475^2~5 X-Git-Url: https://gitweb.dealii.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=b228dbd2fa2d222d494b7d73d76cfe87baff1e59;p=dealii.git Don't just store references, but SmartPointer objects. This ensures that we get an error if the referenced object is deleted first. --- diff --git a/include/deal.II/dofs/dof_handler_policy.h b/include/deal.II/dofs/dof_handler_policy.h index 633cc7c7c8..f9c31bbb6c 100644 --- a/include/deal.II/dofs/dof_handler_policy.h +++ b/include/deal.II/dofs/dof_handler_policy.h @@ -131,7 +131,7 @@ namespace internal /** * The DoFHandler object on which this policy object works. */ - dealii::DoFHandler &dof_handler; + SmartPointer > dof_handler; }; /** @@ -219,7 +219,7 @@ namespace internal /** * The DoFHandler object on which this policy object works. */ - dealii::DoFHandler &dof_handler; + SmartPointer > dof_handler; }; } } diff --git a/source/dofs/dof_handler_policy.cc b/source/dofs/dof_handler_policy.cc index c18993589a..bf23ce3dd7 100644 --- a/source/dofs/dof_handler_policy.cc +++ b/source/dofs/dof_handler_policy.cc @@ -916,7 +916,7 @@ namespace internal Sequential:: Sequential (dealii::DoFHandler &dof_handler) : - dof_handler (dof_handler) + dof_handler (&dof_handler) {} @@ -929,7 +929,7 @@ namespace internal const types::global_dof_index n_dofs = Implementation::distribute_dofs (0, numbers::invalid_subdomain_id, - dof_handler); + *dof_handler); // now set the elements of the // number cache appropriately @@ -962,12 +962,13 @@ namespace internal { std::vector user_flags; - dof_handler.get_triangulation().save_user_flags (user_flags); - const_cast&>(dof_handler.get_triangulation()).clear_user_flags (); + dof_handler->get_triangulation().save_user_flags (user_flags); + const_cast&>(dof_handler->get_triangulation()).clear_user_flags (); - for (unsigned int level = 0; level < dof_handler.get_triangulation().n_levels(); ++level) + for (unsigned int level = 0; level < dof_handler->get_triangulation().n_levels(); ++level) { - types::global_dof_index next_free_dof = Implementation::distribute_dofs_on_level(0, numbers::invalid_subdomain_id, dof_handler, level); + types::global_dof_index next_free_dof = Implementation::distribute_dofs_on_level(0, numbers::invalid_subdomain_id, + *dof_handler, level); number_caches[level].n_global_dofs = next_free_dof; number_caches[level].n_locally_owned_dofs = next_free_dof; @@ -977,7 +978,7 @@ namespace internal number_caches[level].n_locally_owned_dofs_per_processor.resize(1); number_caches[level].n_locally_owned_dofs_per_processor[0] = next_free_dof; } - const_cast&>(dof_handler.get_triangulation()).load_user_flags (user_flags); + const_cast&>(dof_handler->get_triangulation()).load_user_flags (user_flags); } @@ -989,12 +990,12 @@ namespace internal NumberCache &number_cache_current) const { Implementation::renumber_dofs (new_numbers, IndexSet(0), - dof_handler, true); + *dof_handler, true); // in the sequential case, the number cache should not have changed but we // have to set the elements of the structure appropriately anyway NumberCache number_cache; - number_cache.n_global_dofs = dof_handler.n_dofs(); + number_cache.n_global_dofs = dof_handler->n_dofs(); number_cache.n_locally_owned_dofs = number_cache.n_global_dofs; number_cache.locally_owned_dofs @@ -1038,11 +1039,11 @@ namespace internal // Namely, we first restore the original partition (without artificial cells) // and then turn artificial cells on at the end of this function. const parallel::shared::Triangulation *tr = - (dynamic_cast*> (&this->dof_handler.get_triangulation())); + (dynamic_cast*> (&this->dof_handler->get_triangulation())); Assert(tr != nullptr, ExcInternalError()); typename parallel::shared::Triangulation::active_cell_iterator - cell = this->dof_handler.get_triangulation().begin_active(), - endc = this->dof_handler.get_triangulation().end(); + cell = this->dof_handler->get_triangulation().begin_active(), + endc = this->dof_handler->get_triangulation().end(); std::vector current_subdomain_ids(tr->n_active_cells()); const std::vector &true_subdomain_ids = tr->get_true_subdomain_ids_of_cells(); if (tr->with_artificial_cells()) @@ -1055,7 +1056,7 @@ namespace internal // let the sequential algorithm do its magic, then sort DoF indices // by subdomain this->Sequential::distribute_dofs (number_cache); - DoFRenumbering::subdomain_wise (this->dof_handler); + DoFRenumbering::subdomain_wise (*this->dof_handler); // dofrenumbering will reset subdomains, this is ugly but we need to do it again: cell = tr->begin_active(); @@ -1063,12 +1064,12 @@ namespace internal for (unsigned int index=0; cell != endc; cell++, index++) cell->set_subdomain_id(true_subdomain_ids[index]); - number_cache.locally_owned_dofs_per_processor = DoFTools::locally_owned_dofs_per_subdomain (this->dof_handler); - number_cache.locally_owned_dofs = number_cache.locally_owned_dofs_per_processor[this->dof_handler.get_triangulation().locally_owned_subdomain()]; + number_cache.locally_owned_dofs_per_processor = DoFTools::locally_owned_dofs_per_subdomain (*this->dof_handler); + number_cache.locally_owned_dofs = number_cache.locally_owned_dofs_per_processor[this->dof_handler->get_triangulation().locally_owned_subdomain()]; number_cache.n_locally_owned_dofs_per_processor.resize (number_cache.locally_owned_dofs_per_processor.size()); for (unsigned int i = 0; i < number_cache.n_locally_owned_dofs_per_processor.size(); i++) number_cache.n_locally_owned_dofs_per_processor[i] = number_cache.locally_owned_dofs_per_processor[i].n_elements(); - number_cache.n_locally_owned_dofs = number_cache.n_locally_owned_dofs_per_processor[this->dof_handler.get_triangulation().locally_owned_subdomain()]; + number_cache.n_locally_owned_dofs = number_cache.n_locally_owned_dofs_per_processor[this->dof_handler->get_triangulation().locally_owned_subdomain()]; // restore current subdomain ids cell = tr->begin_active(); @@ -1107,12 +1108,12 @@ namespace internal // Similar to distribute_dofs() we need to have a special treatment in // case artificial cells are present. const parallel::shared::Triangulation *tr = - (dynamic_cast*> (&this->dof_handler.get_triangulation())); + (dynamic_cast*> (&this->dof_handler->get_triangulation())); Assert(tr != nullptr, ExcInternalError()); typename parallel::shared::Triangulation::active_cell_iterator - cell = this->dof_handler.get_triangulation().begin_active(), - endc = this->dof_handler.get_triangulation().end(); + cell = this->dof_handler->get_triangulation().begin_active(), + endc = this->dof_handler->get_triangulation().end(); std::vector current_subdomain_ids(tr->n_active_cells()); const std::vector &true_subdomain_ids = tr->get_true_subdomain_ids_of_cells(); if (tr->with_artificial_cells()) @@ -1122,22 +1123,22 @@ namespace internal cell->set_subdomain_id(true_subdomain_ids[index]); } - std::vector global_gathered_numbers (this->dof_handler.n_dofs (), 0); - // as we call DoFRenumbering::subdomain_wise (dof_handler) from distribute_dofs(), + std::vector global_gathered_numbers (this->dof_handler->n_dofs (), 0); + // as we call DoFRenumbering::subdomain_wise (*dof_handler) from distribute_dofs(), // we need to support sequential-like input. // Distributed-like input from, for example, component_wise renumbering is also supported. - if (new_numbers.size () == this->dof_handler.n_dofs ()) + if (new_numbers.size () == this->dof_handler->n_dofs ()) { global_gathered_numbers = new_numbers; } else { - Assert(new_numbers.size() == this->dof_handler.locally_owned_dofs().n_elements(), + Assert(new_numbers.size() == this->dof_handler->locally_owned_dofs().n_elements(), ExcInternalError()); const unsigned int n_cpu = Utilities::MPI::n_mpi_processes (tr->get_communicator ()); - std::vector gathered_new_numbers (this->dof_handler.n_dofs (), 0); + std::vector gathered_new_numbers (this->dof_handler->n_dofs (), 0); Assert(Utilities::MPI::this_mpi_process (tr->get_communicator ()) == - this->dof_handler.get_triangulation().locally_owned_subdomain (), + this->dof_handler->get_triangulation().locally_owned_subdomain (), ExcInternalError()) // gather new numbers among processors into one vector @@ -1180,8 +1181,8 @@ namespace internal // flag_1 and flag_2 are // used to control that there is a // one-to-one relation between old and new DoFs. - std::vector flag_1 (this->dof_handler.n_dofs (), 0); - std::vector flag_2 (this->dof_handler.n_dofs (), 0); + std::vector flag_1 (this->dof_handler->n_dofs (), 0); + std::vector flag_2 (this->dof_handler->n_dofs (), 0); for (unsigned int i = 0; i < n_cpu; i++) { const IndexSet &iset = @@ -1191,8 +1192,8 @@ namespace internal { const types::global_dof_index target = iset.nth_index_in_set (ind); const types::global_dof_index value = gathered_new_numbers[shift + ind]; - Assert(target < this->dof_handler.n_dofs(), ExcInternalError()); - Assert(value < this->dof_handler.n_dofs(), ExcInternalError()); + Assert(target < this->dof_handler->n_dofs(), ExcInternalError()); + Assert(value < this->dof_handler->n_dofs(), ExcInternalError()); global_gathered_numbers[target] = value; flag_1[target]++; flag_2[value]++; @@ -1213,9 +1214,9 @@ namespace internal this->Sequential::renumber_dofs (global_gathered_numbers, number_cache); // correct number_cache: number_cache.locally_owned_dofs_per_processor = - DoFTools::locally_owned_dofs_per_subdomain (this->dof_handler); + DoFTools::locally_owned_dofs_per_subdomain (*this->dof_handler); number_cache.locally_owned_dofs = - number_cache.locally_owned_dofs_per_processor[this->dof_handler.get_triangulation().locally_owned_subdomain ()]; + number_cache.locally_owned_dofs_per_processor[this->dof_handler->get_triangulation().locally_owned_subdomain ()]; // sequential renumbering returns a vector of size 1 here, // correct this: number_cache.n_locally_owned_dofs_per_processor.resize(number_cache.locally_owned_dofs_per_processor.size()); @@ -1224,7 +1225,7 @@ namespace internal number_cache.n_locally_owned_dofs_per_processor[i] = number_cache.locally_owned_dofs_per_processor[i].n_elements (); number_cache.n_locally_owned_dofs = - number_cache.n_locally_owned_dofs_per_processor[this->dof_handler.get_triangulation().locally_owned_subdomain ()]; + number_cache.n_locally_owned_dofs_per_processor[this->dof_handler->get_triangulation().locally_owned_subdomain ()]; // restore artificial cells cell = tr->begin_active(); @@ -2078,7 +2079,7 @@ namespace internal ParallelDistributed:: ParallelDistributed (dealii::DoFHandler &dof_handler) : - dof_handler (dof_handler) + dof_handler (&dof_handler) {} @@ -2100,7 +2101,7 @@ namespace internal parallel::distributed::Triangulation< dim, spacedim > *tr = (dynamic_cast*> (const_cast*> - (&dof_handler.get_triangulation()))); + (&dof_handler->get_triangulation()))); Assert (tr != nullptr, ExcInternalError()); const unsigned int @@ -2109,7 +2110,7 @@ namespace internal //* 1. distribute on own subdomain const dealii::types::global_dof_index n_initial_local_dofs = Implementation::distribute_dofs (0, tr->locally_owned_subdomain(), - dof_handler); + *dof_handler); //* 2. iterate over ghostcells and kill dofs that are not // owned by us @@ -2121,8 +2122,8 @@ namespace internal std::vector local_dof_indices; typename DoFHandler::active_cell_iterator - cell = dof_handler.begin_active(), - endc = dof_handler.end(); + cell = dof_handler->begin_active(), + endc = dof_handler->end(); for (; cell != endc; ++cell) if (cell->is_ghost() && @@ -2178,7 +2179,7 @@ namespace internal // numbering form. we renumber some dofs as invalid, so // choose the nocheck-version. Implementation::renumber_dofs (renumbering, IndexSet(0), - dof_handler, false); + *dof_handler, false); // now a little bit of housekeeping number_cache.n_global_dofs @@ -2230,8 +2231,8 @@ namespace internal tr->clear_user_flags (); // mark all own cells for transfer - for (typename DoFHandler::active_cell_iterator cell = dof_handler.begin_active(); - cell != dof_handler.end(); ++cell) + for (typename DoFHandler::active_cell_iterator cell = dof_handler->begin_active(); + cell != dof_handler->end(); ++cell) if (!cell->is_artificial()) cell->set_user_flag(); @@ -2246,12 +2247,12 @@ namespace internal // Send and receive cells. After this, only the local cells // are marked, that received new data. This has to be // communicated in a second communication step. - communicate_dof_indices_on_marked_cells (dof_handler, + communicate_dof_indices_on_marked_cells (*dof_handler, vertices_with_ghost_neighbors, tr->coarse_cell_to_p4est_tree_permutation, tr->p4est_tree_to_coarse_cell_permutation); - communicate_dof_indices_on_marked_cells (dof_handler, + communicate_dof_indices_on_marked_cells (*dof_handler, vertices_with_ghost_neighbors, tr->coarse_cell_to_p4est_tree_permutation, tr->p4est_tree_to_coarse_cell_permutation); @@ -2263,8 +2264,8 @@ namespace internal { std::vector local_dof_indices; - for (typename DoFHandler::active_cell_iterator cell = dof_handler.begin_active(); - cell != dof_handler.end(); ++cell) + for (typename DoFHandler::active_cell_iterator cell = dof_handler->begin_active(); + cell != dof_handler->end(); ++cell) if (!cell->is_artificial()) { local_dof_indices.resize (cell->get_fe().dofs_per_cell); @@ -2306,7 +2307,7 @@ namespace internal parallel::distributed::Triangulation< dim, spacedim > *tr = (dynamic_cast*> (const_cast*> - (&dof_handler.get_triangulation()))); + (&dof_handler->get_triangulation()))); Assert (tr != nullptr, ExcInternalError()); AssertThrow( @@ -2333,7 +2334,7 @@ namespace internal const unsigned int n_initial_local_dofs = Implementation::distribute_dofs_on_level(0, tr->locally_owned_subdomain(), - dof_handler, + *dof_handler, level); //* 2. iterate over ghostcells and kill dofs that are not @@ -2347,8 +2348,8 @@ namespace internal std::vector local_dof_indices; typename DoFHandler::level_cell_iterator - cell = dof_handler.begin(level), - endc = dof_handler.end(level); + cell = dof_handler->begin(level), + endc = dof_handler->end(level); for (; cell != endc; ++cell) if (cell->level_subdomain_id()!=numbers::artificial_subdomain_id && @@ -2411,7 +2412,7 @@ namespace internal // in that case if (level < tr->n_levels()) Implementation::renumber_mg_dofs (renumbering, IndexSet(0), - dof_handler, level, + *dof_handler, level, false); // now a little bit of housekeeping @@ -2471,8 +2472,8 @@ namespace internal // mark all ghost cells for transfer { typename DoFHandler::level_cell_iterator - cell, endc = dof_handler.end(); - for (cell = dof_handler.begin(); cell != endc; ++cell) + cell, endc = dof_handler->end(); + for (cell = dof_handler->begin(); cell != endc; ++cell) if (cell->level_subdomain_id() != dealii::numbers::artificial_subdomain_id && !cell->is_locally_owned_on_level()) cell->set_user_flag(); @@ -2482,7 +2483,7 @@ namespace internal // managed to get every DoF, remove the user_flag, otherwise we // will request them again in the step below. communicate_mg_ghost_cells(*tr, - dof_handler, + *dof_handler, tr->coarse_cell_to_p4est_tree_permutation, tr->p4est_tree_to_coarse_cell_permutation); @@ -2494,7 +2495,7 @@ namespace internal // Phase 2, only request the cells that were not completed // in Phase 1. communicate_mg_ghost_cells(*tr, - dof_handler, + *dof_handler, tr->coarse_cell_to_p4est_tree_permutation, tr->p4est_tree_to_coarse_cell_permutation); @@ -2502,8 +2503,8 @@ namespace internal // make sure we have removed all flags: { typename DoFHandler::level_cell_iterator - cell, endc = dof_handler.end(); - for (cell = dof_handler.begin(); cell != endc; ++cell) + cell, endc = dof_handler->end(); + for (cell = dof_handler->begin(); cell != endc; ++cell) if (cell->level_subdomain_id() != dealii::numbers::artificial_subdomain_id && !cell->is_locally_owned_on_level()) Assert(cell->user_flag_set()==false, ExcInternalError()); @@ -2521,9 +2522,9 @@ namespace internal { std::vector local_dof_indices; typename DoFHandler::level_cell_iterator - cell, endc = dof_handler.end(); + cell, endc = dof_handler->end(); - for (cell = dof_handler.begin(); cell != endc; ++cell) + for (cell = dof_handler->begin(); cell != endc; ++cell) if (cell->level_subdomain_id() != dealii::numbers::artificial_subdomain_id) { local_dof_indices.resize (cell->get_fe().dofs_per_cell); @@ -2551,7 +2552,7 @@ namespace internal { (void)new_numbers; - Assert (new_numbers.size() == dof_handler.locally_owned_dofs().n_elements(), + Assert (new_numbers.size() == dof_handler->locally_owned_dofs().n_elements(), ExcInternalError()); NumberCache number_cache; @@ -2567,13 +2568,13 @@ namespace internal // we still need to go through this function, but we can skip this // calculation. - number_cache.locally_owned_dofs = IndexSet (dof_handler.n_dofs()); - if (dof_handler.locally_owned_dofs().n_elements()>0) + number_cache.locally_owned_dofs = IndexSet (dof_handler->n_dofs()); + if (dof_handler->locally_owned_dofs().n_elements()>0) { std::vector new_numbers_sorted (new_numbers); std::sort(new_numbers_sorted.begin(), new_numbers_sorted.end()); std::vector::const_iterator it = new_numbers_sorted.begin(); - const unsigned int n_blocks = dof_handler.get_fe().n_blocks(); + const unsigned int n_blocks = dof_handler->get_fe().n_blocks(); std::vector > block_indices(n_blocks); block_indices[0].first = *it++; block_indices[0].second = 1; @@ -2641,19 +2642,19 @@ namespace internal // also check with the number of locally owned degrees of freedom that // the DoFHandler object still stores Assert (number_cache.locally_owned_dofs.n_elements() == - dof_handler.n_locally_owned_dofs(), + dof_handler->n_locally_owned_dofs(), ExcInternalError()); // then also set this number in our own copy - number_cache.n_locally_owned_dofs = dof_handler.n_locally_owned_dofs(); + number_cache.n_locally_owned_dofs = dof_handler->n_locally_owned_dofs(); // mark not locally active DoFs as invalid { std::vector local_dof_indices; typename DoFHandler::active_cell_iterator - cell = dof_handler.begin_active(), - endc = dof_handler.end(); + cell = dof_handler->begin_active(), + endc = dof_handler->end(); for (; cell != endc; ++cell) if (!cell->is_artificial()) @@ -2665,7 +2666,7 @@ namespace internal if (local_dof_indices[i] == numbers::invalid_dof_index) continue; - if (!dof_handler.locally_owned_dofs().is_element(local_dof_indices[i])) + if (!dof_handler->locally_owned_dofs().is_element(local_dof_indices[i])) { //this DoF is not owned by us, so set it to invalid. local_dof_indices[i] @@ -2679,10 +2680,10 @@ namespace internal // renumber. Skip when there is nothing to do because we own no DoF. - if (dof_handler.locally_owned_dofs().n_elements() > 0) + if (dof_handler->locally_owned_dofs().n_elements() > 0) Implementation::renumber_dofs (new_numbers, - dof_handler.locally_owned_dofs(), - dof_handler, + dof_handler->locally_owned_dofs(), + *dof_handler, false); // communication @@ -2690,7 +2691,7 @@ namespace internal parallel::distributed::Triangulation< dim, spacedim > *tr = (dynamic_cast*> (const_cast*> - (&dof_handler.get_triangulation()))); + (&dof_handler->get_triangulation()))); Assert (tr != nullptr, ExcInternalError()); std::vector user_flags; @@ -2699,8 +2700,8 @@ namespace internal // mark all own cells for transfer typename DoFHandler::active_cell_iterator - cell, endc = dof_handler.end(); - for (cell = dof_handler.begin_active(); cell != endc; ++cell) + cell, endc = dof_handler->end(); + for (cell = dof_handler->begin_active(); cell != endc; ++cell) if (!cell->is_artificial()) cell->set_user_flag(); @@ -2714,12 +2715,12 @@ namespace internal // Send and receive cells. After this, only the local cells are // marked, that received new data. This has to be communicated in a // second communication step. - communicate_dof_indices_on_marked_cells (dof_handler, + communicate_dof_indices_on_marked_cells (*dof_handler, vertices_with_ghost_neighbors, tr->coarse_cell_to_p4est_tree_permutation, tr->p4est_tree_to_coarse_cell_permutation); - communicate_dof_indices_on_marked_cells (dof_handler, + communicate_dof_indices_on_marked_cells (*dof_handler, vertices_with_ghost_neighbors, tr->coarse_cell_to_p4est_tree_permutation, tr->p4est_tree_to_coarse_cell_permutation);