1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
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
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
|
/* Copyright 2019 Andrew Myers, Ann Almgren, Axel Huebl
* David Grote, Maxence Thevenet, Michael Rowan
* Remi Lehe, Weiqun Zhang, levinem
*
* This file is part of WarpX.
*
* License: BSD-3-Clause-LBNL
*/
#include "WarpX.H"
#include "Utils/WarpXAlgorithmSelection.H"
#include <AMReX_BLProfiler.H>
using namespace amrex;
void
WarpX::LoadBalance ()
{
WARPX_PROFILE_REGION("LoadBalance");
WARPX_PROFILE("WarpX::LoadBalance()");
AMREX_ALWAYS_ASSERT(costs[0] != nullptr);
#ifdef AMREX_USE_MPI
if (WarpX::load_balance_costs_update_algo == LoadBalanceCostsUpdateAlgo::Heuristic)
{
// compute the costs on a per-rank basis
WarpX::ComputeCostsHeuristic(costs);
}
// By default, do not do a redistribute; this toggles to true if RemakeLevel
// is called for any level
int doLoadBalance = 0;
const int nLevels = finestLevel();
for (int lev = 0; lev <= nLevels; ++lev)
{
// Compute the new distribution mapping
DistributionMapping newdm;
const amrex::Real nboxes = costs[lev]->size();
const amrex::Real nprocs = ParallelContext::NProcsSub();
const int nmax = static_cast<int>(std::ceil(nboxes/nprocs*load_balance_knapsack_factor));
// These store efficiency (meaning, the average 'cost' over all ranks,
// normalized to max cost) for current and proposed distribution mappings
amrex::Real currentEfficiency = 0.0;
amrex::Real proposedEfficiency = 0.0;
newdm = (load_balance_with_sfc)
? DistributionMapping::makeSFC(*costs[lev],
currentEfficiency, proposedEfficiency,
false,
ParallelDescriptor::IOProcessorNumber())
: DistributionMapping::makeKnapSack(*costs[lev],
currentEfficiency, proposedEfficiency,
nmax,
false,
ParallelDescriptor::IOProcessorNumber());
// As specified in the above calls to makeSFC and makeKnapSack, the new
// distribution mapping is NOT communicated to all ranks; the loadbalanced
// dm is up-to-date only on root, and we can decide whether to broadcast
if ((load_balance_efficiency_ratio_threshold > 0.0)
&& (ParallelDescriptor::MyProc() == ParallelDescriptor::IOProcessorNumber()))
{
doLoadBalance = (proposedEfficiency > load_balance_efficiency_ratio_threshold*currentEfficiency);
}
ParallelDescriptor::Bcast(&doLoadBalance, 1,
ParallelDescriptor::IOProcessorNumber());
if (doLoadBalance)
{
Vector<int> pmap;
if (ParallelDescriptor::MyProc() == ParallelDescriptor::IOProcessorNumber())
{
pmap = newdm.ProcessorMap();
} else
{
pmap.resize(nboxes);
}
ParallelDescriptor::Bcast(&pmap[0], pmap.size(), ParallelDescriptor::IOProcessorNumber());
if (ParallelDescriptor::MyProc() != ParallelDescriptor::IOProcessorNumber())
{
newdm = DistributionMapping(pmap);
}
RemakeLevel(lev, t_new[lev], boxArray(lev), newdm);
}
}
if (doLoadBalance)
{
mypc->Redistribute();
}
#endif
}
void
WarpX::RemakeLevel (int lev, Real /*time*/, const BoxArray& ba, const DistributionMapping& dm)
{
if (ba == boxArray(lev))
{
if (ParallelDescriptor::NProcs() == 1) return;
// Fine patch
for (int idim=0; idim < 3; ++idim)
{
{
const IntVect& ng = Bfield_fp[lev][idim]->nGrowVect();
auto pmf = std::unique_ptr<MultiFab>(new MultiFab(Bfield_fp[lev][idim]->boxArray(),
dm, Bfield_fp[lev][idim]->nComp(), ng));
pmf->Redistribute(*Bfield_fp[lev][idim], 0, 0, Bfield_fp[lev][idim]->nComp(), ng);
Bfield_fp[lev][idim] = std::move(pmf);
}
{
const IntVect& ng = Efield_fp[lev][idim]->nGrowVect();
auto pmf = std::unique_ptr<MultiFab>(new MultiFab(Efield_fp[lev][idim]->boxArray(),
dm, Efield_fp[lev][idim]->nComp(), ng));
pmf->Redistribute(*Efield_fp[lev][idim], 0, 0, Efield_fp[lev][idim]->nComp(), ng);
Efield_fp[lev][idim] = std::move(pmf);
}
{
const IntVect& ng = current_fp[lev][idim]->nGrowVect();
auto pmf = std::unique_ptr<MultiFab>(new MultiFab(current_fp[lev][idim]->boxArray(),
dm, current_fp[lev][idim]->nComp(), ng));
current_fp[lev][idim] = std::move(pmf);
}
if (current_store[lev][idim])
{
const IntVect& ng = current_store[lev][idim]->nGrowVect();
auto pmf = std::unique_ptr<MultiFab>(new MultiFab(current_store[lev][idim]->boxArray(),
dm, current_store[lev][idim]->nComp(), ng));
// no need to redistribute
current_store[lev][idim] = std::move(pmf);
}
}
if (F_fp[lev] != nullptr) {
const IntVect& ng = F_fp[lev]->nGrowVect();
auto pmf = std::unique_ptr<MultiFab>(new MultiFab(F_fp[lev]->boxArray(),
dm, F_fp[lev]->nComp(), ng));
pmf->Redistribute(*F_fp[lev], 0, 0, F_fp[lev]->nComp(), ng);
F_fp[lev] = std::move(pmf);
}
if (rho_fp[lev] != nullptr) {
const int nc = rho_fp[lev]->nComp();
const IntVect& ng = rho_fp[lev]->nGrowVect();
auto pmf = std::unique_ptr<MultiFab>(new MultiFab(rho_fp[lev]->boxArray(),
dm, nc, ng));
rho_fp[lev] = std::move(pmf);
}
// Aux patch
if (lev == 0 && Bfield_aux[0][0]->ixType() == Bfield_fp[0][0]->ixType())
{
for (int idim = 0; idim < 3; ++idim) {
Bfield_aux[lev][idim].reset(new MultiFab(*Bfield_fp[lev][idim], amrex::make_alias, 0, Bfield_aux[lev][idim]->nComp()));
Efield_aux[lev][idim].reset(new MultiFab(*Efield_fp[lev][idim], amrex::make_alias, 0, Efield_aux[lev][idim]->nComp()));
}
} else {
for (int idim=0; idim < 3; ++idim)
{
{
const IntVect& ng = Bfield_aux[lev][idim]->nGrowVect();
auto pmf = std::unique_ptr<MultiFab>(new MultiFab(Bfield_aux[lev][idim]->boxArray(),
dm, Bfield_aux[lev][idim]->nComp(), ng));
// pmf->Redistribute(*Bfield_aux[lev][idim], 0, 0, Bfield_aux[lev][idim]->nComp(), ng);
Bfield_aux[lev][idim] = std::move(pmf);
}
{
const IntVect& ng = Efield_aux[lev][idim]->nGrowVect();
auto pmf = std::unique_ptr<MultiFab>(new MultiFab(Efield_aux[lev][idim]->boxArray(),
dm, Efield_aux[lev][idim]->nComp(), ng));
// pmf->Redistribute(*Efield_aux[lev][idim], 0, 0, Efield_aux[lev][idim]->nComp(), ng);
Efield_aux[lev][idim] = std::move(pmf);
}
}
}
// Coarse patch
if (lev > 0) {
for (int idim=0; idim < 3; ++idim)
{
{
const IntVect& ng = Bfield_cp[lev][idim]->nGrowVect();
auto pmf = std::unique_ptr<MultiFab>(new MultiFab(Bfield_cp[lev][idim]->boxArray(),
dm, Bfield_cp[lev][idim]->nComp(), ng));
pmf->Redistribute(*Bfield_cp[lev][idim], 0, 0, Bfield_cp[lev][idim]->nComp(), ng);
Bfield_cp[lev][idim] = std::move(pmf);
}
{
const IntVect& ng = Efield_cp[lev][idim]->nGrowVect();
auto pmf = std::unique_ptr<MultiFab>(new MultiFab(Efield_cp[lev][idim]->boxArray(),
dm, Efield_cp[lev][idim]->nComp(), ng));
pmf->Redistribute(*Efield_cp[lev][idim], 0, 0, Efield_cp[lev][idim]->nComp(), ng);
Efield_cp[lev][idim] = std::move(pmf);
}
{
const IntVect& ng = current_cp[lev][idim]->nGrowVect();
auto pmf = std::unique_ptr<MultiFab>( new MultiFab(current_cp[lev][idim]->boxArray(),
dm, current_cp[lev][idim]->nComp(), ng));
current_cp[lev][idim] = std::move(pmf);
}
}
if (F_cp[lev] != nullptr) {
const IntVect& ng = F_cp[lev]->nGrowVect();
auto pmf = std::unique_ptr<MultiFab>(new MultiFab(F_cp[lev]->boxArray(),
dm, F_cp[lev]->nComp(), ng));
pmf->Redistribute(*F_cp[lev], 0, 0, F_cp[lev]->nComp(), ng);
F_cp[lev] = std::move(pmf);
}
if (rho_cp[lev] != nullptr) {
const int nc = rho_cp[lev]->nComp();
const IntVect& ng = rho_cp[lev]->nGrowVect();
auto pmf = std::unique_ptr<MultiFab>(new MultiFab(rho_cp[lev]->boxArray(),
dm, nc, ng));
rho_cp[lev] = std::move(pmf);
}
}
if (lev > 0 && (n_field_gather_buffer > 0 || n_current_deposition_buffer > 0)) {
for (int idim=0; idim < 3; ++idim)
{
if (Bfield_cax[lev][idim])
{
const IntVect& ng = Bfield_cax[lev][idim]->nGrowVect();
auto pmf = std::unique_ptr<MultiFab>(new MultiFab(Bfield_cax[lev][idim]->boxArray(),
dm, Bfield_cax[lev][idim]->nComp(), ng));
// pmf->ParallelCopy(*Bfield_cax[lev][idim], 0, 0, Bfield_cax[lev][idim]->nComp(), ng, ng);
Bfield_cax[lev][idim] = std::move(pmf);
}
if (Efield_cax[lev][idim])
{
const IntVect& ng = Efield_cax[lev][idim]->nGrowVect();
auto pmf = std::unique_ptr<MultiFab>(new MultiFab(Efield_cax[lev][idim]->boxArray(),
dm, Efield_cax[lev][idim]->nComp(), ng));
// pmf->ParallelCopy(*Efield_cax[lev][idim], 0, 0, Efield_cax[lev][idim]->nComp(), ng, ng);
Efield_cax[lev][idim] = std::move(pmf);
}
if (current_buf[lev][idim])
{
const IntVect& ng = current_buf[lev][idim]->nGrowVect();
auto pmf = std::unique_ptr<MultiFab>(new MultiFab(current_buf[lev][idim]->boxArray(),
dm, current_buf[lev][idim]->nComp(), ng));
// pmf->ParallelCopy(*current_buf[lev][idim], 0, 0, current_buf[lev][idim]->nComp(), ng, ng);
current_buf[lev][idim] = std::move(pmf);
}
}
if (charge_buf[lev])
{
const IntVect& ng = charge_buf[lev]->nGrowVect();
auto pmf = std::unique_ptr<MultiFab>(new MultiFab(charge_buf[lev]->boxArray(),
dm, charge_buf[lev]->nComp(), ng));
// pmf->ParallelCopy(*charge_buf[lev][idim], 0, 0, charge_buf[lev]->nComp(), ng, ng);
charge_buf[lev] = std::move(pmf);
}
if (current_buffer_masks[lev])
{
const IntVect& ng = current_buffer_masks[lev]->nGrowVect();
auto pmf = std::unique_ptr<iMultiFab>(new iMultiFab(current_buffer_masks[lev]->boxArray(),
dm, current_buffer_masks[lev]->nComp(), ng));
// pmf->ParallelCopy(*current_buffer_masks[lev], 0, 0, current_buffer_masks[lev]->nComp(), ng, ng);
current_buffer_masks[lev] = std::move(pmf);
}
if (gather_buffer_masks[lev])
{
const IntVect& ng = gather_buffer_masks[lev]->nGrowVect();
auto pmf = std::unique_ptr<iMultiFab>(new iMultiFab(gather_buffer_masks[lev]->boxArray(),
dm, gather_buffer_masks[lev]->nComp(), ng));
// pmf->ParallelCopy(*gather_buffer_masks[lev], 0, 0, gather_buffer_masks[lev]->nComp(), ng, ng);
gather_buffer_masks[lev] = std::move(pmf);
}
}
if (costs[lev] != nullptr)
{
costs[lev].reset(new amrex::LayoutData<Real>(ba, dm));
for (int i : costs[lev]->IndexArray())
{
(*costs[lev])[i] = 0.0;
}
}
SetDistributionMap(lev, dm);
} else
{
amrex::Abort("RemakeLevel: to be implemented");
}
// Re-initialize diagnostic functors that stores pointers to the user-requested fields at level, lev.
multi_diags->InitializeFieldFunctors( lev );
}
void
WarpX::ComputeCostsHeuristic (amrex::Vector<std::unique_ptr<amrex::LayoutData<amrex::Real> > >& a_costs)
{
for (int lev = 0; lev <= finest_level; ++lev)
{
auto & mypc_ref = WarpX::GetInstance().GetPartContainer();
auto nSpecies = mypc_ref.nSpecies();
// Species loop
for (int i_s = 0; i_s < nSpecies; ++i_s)
{
auto & myspc = mypc_ref.GetParticleContainer(i_s);
// Particle loop
for (WarpXParIter pti(myspc, lev); pti.isValid(); ++pti)
{
(*a_costs[lev])[pti.index()] += costs_heuristic_particles_wt*pti.numParticles();
}
}
//Cell loop
MultiFab* Ex = Efield_fp[lev][0].get();
for (MFIter mfi(*Ex, false); mfi.isValid(); ++mfi)
{
const Box& gbx = mfi.growntilebox();
(*a_costs[lev])[mfi.index()] += costs_heuristic_cells_wt*gbx.numPts();
}
}
}
void
WarpX::ResetCosts ()
{
for (int lev = 0; lev <= finest_level; ++lev)
{
for (int i : costs[lev]->IndexArray())
{
(*costs[lev])[i] = 0.0;
}
}
}
|
