abacus-develop/source/source_cell/parallel_kpoints.cpp at develop · pplab/abacus-develop · GitHub

History

262 lines (222 loc) · 6.52 KB

Raw

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

#include "parallel_kpoints.h"

#include "source_base/parallel_common.h"

#include "source_base/parallel_global.h"

// the kpoints here are reduced after symmetry applied.

void Parallel_Kpoints::kinfo(int& nkstot_in,

const int& kpar_in,

const int& my_pool_in,

const int& rank_in_pool_in,

const int& nproc_in,

const int& nspin_in)

{

#ifdef __MPI

this->kpar = kpar_in; // number of pools

this->my_pool = my_pool_in;

this->rank_in_pool = rank_in_pool_in;

this->nproc = nproc_in;

this->nspin = nspin_in;

Parallel_Common::bcast_int(nkstot_in);

this->get_nks_pool(nkstot_in); // assign k-points to each pool

this->get_startk_pool(nkstot_in); // get the start k-point index for each pool

this->get_whichpool(nkstot_in); // get the pool index for each k-point

this->set_startpro_pool(); // get the start processor index for each pool

this->nkstot_np = nkstot_in;

this->nks_np = this->nks_pool[this->my_pool]; // number of k-points in this pool

#else

this->kpar = 1;

this->my_pool = 0;

this->rank_in_pool = 0;

this->nproc = 1;

this->nspin = nspin_in;

this->nkstot_np = nkstot_in;

this->nks_np = nkstot_in;

#endif

return;

}

#ifdef __MPI

void Parallel_Kpoints::get_whichpool(const int& nkstot)

{

this->whichpool.resize(nkstot, 0);

for (int i = 0; i < this->kpar; i++)

{

for (int ik = 0; ik < this->nks_pool[i]; ik++)

{

const int k_now = ik + startk_pool[i];

this->whichpool[k_now] = i;

}

return;

}

void Parallel_Kpoints::get_nks_pool(const int& nkstot)

{

nks_pool.resize(this->kpar, 0);

const int nks_ave = nkstot / this->kpar;

const int remain = nkstot % this->kpar;

for (int i = 0; i < this->kpar; i++)

{

this->nks_pool[i] = nks_ave;

if (i < remain)

{

nks_pool[i]++;

}

return;

}

void Parallel_Kpoints::get_startk_pool(const int& nkstot)

{

startk_pool.resize(this->kpar, 0);

startk_pool[0] = 0;

for (int i = 1; i < this->kpar; i++)

{

startk_pool[i] = startk_pool[i - 1] + nks_pool[i - 1];

}

return;

}

void Parallel_Kpoints::set_startpro_pool()

{

startpro_pool.resize(this->kpar, 0);

const int nproc_ave = this->nproc / this->kpar;

const int remain = this->nproc % this->kpar;

startpro_pool[0] = 0;

for (int i = 1; i < this->kpar; i++)

{

startpro_pool[i] = startpro_pool[i - 1] + nproc_ave;

if (i - 1 < remain)

{

startpro_pool[i]++;

}

return;

}

// gather kpoints from all processor pools, only need to be called by the first processor of each pool.

void Parallel_Kpoints::gatherkvec(const std::vector<ModuleBase::Vector3<double>>& vec_local,

std::vector<ModuleBase::Vector3<double>>& vec_global) const

{

vec_global.resize(this->nkstot_np, ModuleBase::Vector3<double>(0.0, 0.0, 0.0));

for (int i = 0; i < this->nks_np; ++i)

{

if (this->rank_in_pool == 0)

{

vec_global[i + startk_pool[this->my_pool]] = vec_local[i];

}

MPI_Allreduce(MPI_IN_PLACE, &vec_global[0], 3 * this->nkstot_np, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);

return;

}

#endif

void Parallel_Kpoints::pool_collection(double& value, const double* wk, const int& ik)

{

#ifdef __MPI

const int ik_now = ik - this->startk_pool[this->my_pool];

const int pool = this->whichpool[ik];

if (this->rank_in_pool == 0)

{

if (this->my_pool == 0)

{

if (pool == 0)

{

value = wk[ik_now];

}

else

{

MPI_Status ierror;

MPI_Recv(&value, 1, MPI_DOUBLE, this->startpro_pool[pool], ik, MPI_COMM_WORLD, &ierror);

}

else

{

if (this->my_pool == pool)

{

MPI_Send(&wk[ik_now], 1, MPI_DOUBLE, 0, ik, MPI_COMM_WORLD);

}

else

{

}

MPI_Barrier(MPI_COMM_WORLD);

#else

value = wk[ik];

#endif

return;

}

void Parallel_Kpoints::pool_collection(double* value_re,

double* value_im,

const ModuleBase::realArray& re,

const ModuleBase::realArray& im,

const int& ik)

{

const int dim2 = re.getBound2();

const int dim3 = re.getBound3();

const int dim4 = re.getBound4();

assert(re.getBound2() == im.getBound2());

assert(re.getBound3() == im.getBound3());

assert(re.getBound4() == im.getBound4());

const int dim = dim2 * dim3 * dim4;

pool_collection_aux(value_re, re, dim, ik);

pool_collection_aux(value_im, im, dim, ik);

return;

}

void Parallel_Kpoints::pool_collection(std::complex<double>* value, const ModuleBase::ComplexArray& w, const int& ik) const

{

const int dim2 = w.getBound2();

const int dim3 = w.getBound3();

const int dim4 = w.getBound4();

const int dim = dim2 * dim3 * dim4;

pool_collection_aux(value, w, dim, ik);

}

template <class T, class V>

void Parallel_Kpoints::pool_collection_aux(T* value, const V& w, const int& dim, const int& ik) const

{

#ifdef __MPI

const int ik_now = ik - this->startk_pool[this->my_pool];

const int begin = ik_now * dim;

T* p = &w.ptr[begin];

// temprary restrict kpar=1 for NSPIN=2 case for generating_orbitals

int pool = 0;

if (this->nspin != 2)

{

pool = this->whichpool[ik];

}

if (this->rank_in_pool == 0)

{

if (this->my_pool == 0)

{

if (pool == 0)

{

for (int i = 0; i < dim; i++)

{

value[i] = *p;

++p;

}

else

{

MPI_Status ierror;

MPI_Recv(value, dim, MPI_DOUBLE, this->startpro_pool[pool], ik * 2 + 0, MPI_COMM_WORLD, &ierror);

}

else

{

if (this->my_pool == pool)

{

MPI_Send(p, dim, MPI_DOUBLE, 0, ik * 2 + 0, MPI_COMM_WORLD);

}

else

{

}

MPI_Barrier(MPI_COMM_WORLD);

#else

// data transfer ends.

const int begin = ik * dim;

T* p = &w.ptr[begin];

for (int i = 0; i < dim; i++)

{

value[i] = *p;

++p;

}

// data transfer ends.

#endif

}

abacus-develop/source/source_cell/parallel_kpoints.cpp at develop · pplab/abacus-develop · GitHub

Expand file tree

Breadcrumbs

parallel_kpoints.cpp

Latest commit

History

Breadcrumbs

parallel_kpoints.cpp

File metadata and controls