void TrilinosVector::create(Common::PE::CommPattern& cp, Uint neq)
{
// if built
if (m_is_created) destroy();
// get global ids vector
int *gid=(int*)cp.gid()->pack();
// prepare intermediate data
int nmyglobalelements=0;
std::vector<int> myglobalelements(0);
for (int i=0; i<(const int)cp.isUpdatable().size(); i++)
if (cp.isUpdatable()[i])
{
myglobalelements.push_back((int)gid[i]);
++nmyglobalelements;
}
// process local to matrix local numbering mapper
int iupd=0;
int ighost=nmyglobalelements;
m_p2m.resize(0);
m_p2m.reserve(cp.isUpdatable().size());
for (int i=0; i<(const int)cp.isUpdatable().size(); i++)
{
if (cp.isUpdatable()[i])
{
m_p2m.push_back(iupd++);
}
else
{
myglobalelements.push_back((int)gid[i]);
m_p2m.push_back(ighost++);
}
}
// map (its actually blockmap insteady of rowmap, to involve ghosts)
Epetra_BlockMap map(-1,cp.isUpdatable().size(),&myglobalelements[0],neq,0,m_comm);
// create matrix
m_vec=Teuchos::rcp(new Epetra_Vector(map));
/*must be a bug in Trilinos, Epetra_FEVbrMatrix constructor is in Copy mode but it hangs up anyway
more funny, when it gets out of scope and gets dealloc'd, everything survives according to memcheck
map.~Epetra_BlockMap();
*/
// set private data
delete[] gid;
m_neq=neq;
m_blockrow_size=cp.isUpdatable().size();
m_is_created=true;
}
void create_map_data(common::PE::CommPattern& cp, const VariablesDescriptor& variables, std::vector< int >& p2m, std::vector< int >& my_global_elements, int& num_my_elements)
{
// get global ids vector
int *gid=(int*)cp.gid()->pack();
num_my_elements = 0;
const Uint nb_vars = variables.nb_vars();
const Uint total_nb_eq = variables.size();
const Uint nb_nodes_for_rank = cp.isUpdatable().size();
my_global_elements.reserve(nb_nodes_for_rank*total_nb_eq);
// Get the maximum gid, for per-equation blocked storage
int local_max_gid = 0;
int global_nb_gid = 0;
for(Uint i = 0; i != nb_nodes_for_rank; ++i)
local_max_gid = gid[i] > local_max_gid ? gid[i] : local_max_gid;
common::PE::Comm::instance().all_reduce(common::PE::max(), &local_max_gid, 1, &global_nb_gid);
++global_nb_gid; // number of GIDs is the maximum + 1
CFdebug << "Number of GIDs: " << global_nb_gid << CFendl;
for(Uint var_idx = 0; var_idx != nb_vars; ++var_idx)
{
const Uint neq = variables.var_length(var_idx);
const Uint var_offset = variables.offset(var_idx);
const int var_start_gid = var_offset * global_nb_gid;
for (int i=0; i<nb_nodes_for_rank; i++)
{
if (cp.isUpdatable()[i])
{
num_my_elements += neq;
const int start_gid = var_start_gid + gid[i]*neq;
for(int j = 0; j != neq; ++j)
{
my_global_elements.push_back(start_gid+j);
}
}
}
}
// process local to matrix local numbering mapper
const int nb_local_nodes = num_my_elements / total_nb_eq;
const int nb_ghosts = nb_nodes_for_rank - nb_local_nodes;
p2m.resize(nb_nodes_for_rank*total_nb_eq);
for(Uint var_idx = 0; var_idx != nb_vars; ++var_idx)
{
const Uint neq = variables.var_length(var_idx);
const Uint var_offset = variables.offset(var_idx);
int iupd=nb_local_nodes*var_offset;
int ighost=num_my_elements + nb_ghosts*var_offset;
int p_idx = 0;
for (int i=0; i<nb_nodes_for_rank; ++i)
{
const int p_start = i*total_nb_eq+var_offset;
if (cp.isUpdatable()[i])
{
for(Uint j = 0; j != neq; ++j)
p2m[p_start + j] = iupd++;
}
else
{
for(Uint j = 0; j != neq; ++j)
p2m[p_start + j] = ighost++;
}
}
}
// append the ghosts at the end of the element list
for(Uint var_idx = 0; var_idx != nb_vars; ++var_idx)
{
const Uint neq = variables.var_length(var_idx);
const Uint var_offset = variables.offset(var_idx);
const int var_start_gid = var_offset * global_nb_gid;
for (int i=0; i<nb_nodes_for_rank; i++)
{
if (!cp.isUpdatable()[i])
{
const int start_gid = var_start_gid + gid[i]*neq;
for(int j = 0; j != neq; ++j)
my_global_elements.push_back(start_gid+j);
}
}
}
delete[] gid;
}