1 |
mmeineke |
377 |
#ifdef IS_MPI |
2 |
chuckv |
432 |
#include <iostream> |
3 |
mmeineke |
377 |
#include <cstdlib> |
4 |
|
|
#include <cstring> |
5 |
gezelter |
416 |
#include <cmath> |
6 |
mmeineke |
377 |
#include <mpi.h> |
7 |
|
|
|
8 |
|
|
#include "mpiSimulation.hpp" |
9 |
|
|
#include "simError.h" |
10 |
|
|
#include "fortranWrappers.hpp" |
11 |
gezelter |
405 |
#include "randomSPRNG.hpp" |
12 |
mmeineke |
377 |
|
13 |
gezelter |
416 |
#define BASE_SEED 123456789 |
14 |
mmeineke |
377 |
|
15 |
|
|
mpiSimulation* mpiSim; |
16 |
|
|
|
17 |
|
|
mpiSimulation::mpiSimulation(SimInfo* the_entryPlug) |
18 |
|
|
{ |
19 |
|
|
entryPlug = the_entryPlug; |
20 |
|
|
mpiPlug = new mpiSimData; |
21 |
|
|
|
22 |
mmeineke |
447 |
MPI_Comm_size(MPI_COMM_WORLD, &(mpiPlug->numberProcessors) ); |
23 |
mmeineke |
377 |
mpiPlug->myNode = worldRank; |
24 |
gezelter |
405 |
|
25 |
|
|
MolToProcMap = new int[entryPlug->n_mol]; |
26 |
|
|
MolComponentType = new int[entryPlug->n_mol]; |
27 |
|
|
AtomToProcMap = new int[entryPlug->n_atoms]; |
28 |
|
|
|
29 |
mmeineke |
377 |
mpiSim = this; |
30 |
|
|
wrapMeSimParallel( this ); |
31 |
|
|
} |
32 |
|
|
|
33 |
|
|
|
34 |
|
|
mpiSimulation::~mpiSimulation(){ |
35 |
|
|
|
36 |
mmeineke |
418 |
delete[] MolToProcMap; |
37 |
|
|
delete[] MolComponentType; |
38 |
|
|
delete[] AtomToProcMap; |
39 |
|
|
|
40 |
mmeineke |
377 |
delete mpiPlug; |
41 |
|
|
// perhaps we should let fortran know the party is over. |
42 |
|
|
|
43 |
|
|
} |
44 |
|
|
|
45 |
|
|
int* mpiSimulation::divideLabor( void ){ |
46 |
|
|
|
47 |
|
|
int* globalIndex; |
48 |
|
|
|
49 |
|
|
int nComponents; |
50 |
|
|
MoleculeStamp** compStamps; |
51 |
gezelter |
416 |
randomSPRNG *myRandom; |
52 |
mmeineke |
377 |
int* componentsNmol; |
53 |
gezelter |
405 |
int* AtomsPerProc; |
54 |
mmeineke |
377 |
|
55 |
|
|
double numerator; |
56 |
|
|
double denominator; |
57 |
|
|
double precast; |
58 |
gezelter |
405 |
double x, y, a; |
59 |
|
|
int old_atoms, add_atoms, new_atoms; |
60 |
mmeineke |
377 |
|
61 |
|
|
int nTarget; |
62 |
|
|
int molIndex, atomIndex, compIndex, compStart; |
63 |
|
|
int done; |
64 |
|
|
int nLocal, molLocal; |
65 |
gezelter |
416 |
int i, j, loops, which_proc, nmol_local, natoms_local; |
66 |
|
|
int nmol_global, natoms_global; |
67 |
|
|
int local_index, index; |
68 |
mmeineke |
377 |
int smallDiff, bigDiff; |
69 |
tim |
708 |
int baseSeed = entryPlug->getSeed(); |
70 |
mmeineke |
377 |
|
71 |
|
|
int testSum; |
72 |
|
|
|
73 |
|
|
nComponents = entryPlug->nComponents; |
74 |
|
|
compStamps = entryPlug->compStamps; |
75 |
|
|
componentsNmol = entryPlug->componentsNmol; |
76 |
gezelter |
405 |
AtomsPerProc = new int[mpiPlug->numberProcessors]; |
77 |
|
|
|
78 |
mmeineke |
377 |
mpiPlug->nAtomsGlobal = entryPlug->n_atoms; |
79 |
|
|
mpiPlug->nBondsGlobal = entryPlug->n_bonds; |
80 |
|
|
mpiPlug->nBendsGlobal = entryPlug->n_bends; |
81 |
|
|
mpiPlug->nTorsionsGlobal = entryPlug->n_torsions; |
82 |
|
|
mpiPlug->nSRIGlobal = entryPlug->n_SRI; |
83 |
|
|
mpiPlug->nMolGlobal = entryPlug->n_mol; |
84 |
|
|
|
85 |
tim |
708 |
|
86 |
gezelter |
416 |
myRandom = new randomSPRNG( baseSeed ); |
87 |
chuckv |
404 |
|
88 |
chuckv |
434 |
a = 3.0 * (double)mpiPlug->nMolGlobal / (double)mpiPlug->nAtomsGlobal; |
89 |
chuckv |
404 |
|
90 |
gezelter |
405 |
// Initialize things that we'll send out later: |
91 |
|
|
for (i = 0; i < mpiPlug->numberProcessors; i++ ) { |
92 |
|
|
AtomsPerProc[i] = 0; |
93 |
|
|
} |
94 |
|
|
for (i = 0; i < mpiPlug->nMolGlobal; i++ ) { |
95 |
|
|
// default to an error condition: |
96 |
|
|
MolToProcMap[i] = -1; |
97 |
|
|
MolComponentType[i] = -1; |
98 |
|
|
} |
99 |
|
|
for (i = 0; i < mpiPlug->nAtomsGlobal; i++ ) { |
100 |
|
|
// default to an error condition: |
101 |
|
|
AtomToProcMap[i] = -1; |
102 |
|
|
} |
103 |
|
|
|
104 |
|
|
if (mpiPlug->myNode == 0) { |
105 |
|
|
numerator = (double) entryPlug->n_atoms; |
106 |
|
|
denominator = (double) mpiPlug->numberProcessors; |
107 |
|
|
precast = numerator / denominator; |
108 |
|
|
nTarget = (int)( precast + 0.5 ); |
109 |
chuckv |
404 |
|
110 |
gezelter |
405 |
// Build the array of molecule component types first |
111 |
|
|
molIndex = 0; |
112 |
|
|
for (i=0; i < nComponents; i++) { |
113 |
|
|
for (j=0; j < componentsNmol[i]; j++) { |
114 |
|
|
MolComponentType[molIndex] = i; |
115 |
|
|
molIndex++; |
116 |
|
|
} |
117 |
|
|
} |
118 |
chuckv |
404 |
|
119 |
gezelter |
405 |
atomIndex = 0; |
120 |
chuckv |
404 |
|
121 |
gezelter |
405 |
for (i = 0; i < molIndex; i++ ) { |
122 |
chuckv |
404 |
|
123 |
gezelter |
405 |
done = 0; |
124 |
|
|
loops = 0; |
125 |
chuckv |
404 |
|
126 |
gezelter |
405 |
while( !done ){ |
127 |
|
|
loops++; |
128 |
|
|
|
129 |
|
|
// Pick a processor at random |
130 |
chuckv |
404 |
|
131 |
gezelter |
416 |
which_proc = (int) (myRandom->getRandom() * mpiPlug->numberProcessors); |
132 |
chuckv |
404 |
|
133 |
gezelter |
405 |
// How many atoms does this processor have? |
134 |
|
|
|
135 |
|
|
old_atoms = AtomsPerProc[which_proc]; |
136 |
chuckv |
432 |
add_atoms = compStamps[MolComponentType[i]]->getNAtoms(); |
137 |
|
|
new_atoms = old_atoms + add_atoms; |
138 |
chuckv |
404 |
|
139 |
gezelter |
405 |
// If we've been through this loop too many times, we need |
140 |
|
|
// to just give up and assign the molecule to this processor |
141 |
|
|
// and be done with it. |
142 |
|
|
|
143 |
|
|
if (loops > 100) { |
144 |
|
|
sprintf( painCave.errMsg, |
145 |
|
|
"I've tried 100 times to assign molecule %d to a " |
146 |
|
|
" processor, but can't find a good spot.\n" |
147 |
|
|
"I'm assigning it at random to processor %d.\n", |
148 |
|
|
i, which_proc); |
149 |
|
|
painCave.isFatal = 0; |
150 |
|
|
simError(); |
151 |
|
|
|
152 |
|
|
MolToProcMap[i] = which_proc; |
153 |
|
|
AtomsPerProc[which_proc] += add_atoms; |
154 |
|
|
for (j = 0 ; j < add_atoms; j++ ) { |
155 |
mmeineke |
422 |
AtomToProcMap[atomIndex] = which_proc; |
156 |
|
|
atomIndex++; |
157 |
gezelter |
405 |
} |
158 |
|
|
done = 1; |
159 |
|
|
continue; |
160 |
|
|
} |
161 |
chuckv |
432 |
|
162 |
|
|
// If we can add this molecule to this processor without sending |
163 |
|
|
// it above nTarget, then go ahead and do it: |
164 |
|
|
|
165 |
|
|
if (new_atoms <= nTarget) { |
166 |
|
|
MolToProcMap[i] = which_proc; |
167 |
|
|
AtomsPerProc[which_proc] += add_atoms; |
168 |
|
|
for (j = 0 ; j < add_atoms; j++ ) { |
169 |
|
|
AtomToProcMap[atomIndex] = which_proc; |
170 |
|
|
atomIndex++; |
171 |
|
|
} |
172 |
|
|
done = 1; |
173 |
|
|
continue; |
174 |
|
|
} |
175 |
|
|
|
176 |
|
|
|
177 |
chuckv |
434 |
// The only situation left is when new_atoms > nTarget. We |
178 |
|
|
// want to accept this with some probability that dies off the |
179 |
|
|
// farther we are from nTarget |
180 |
gezelter |
405 |
|
181 |
|
|
// roughly: x = new_atoms - nTarget |
182 |
|
|
// Pacc(x) = exp(- a * x) |
183 |
chuckv |
434 |
// where a = penalty / (average atoms per molecule) |
184 |
gezelter |
405 |
|
185 |
|
|
x = (double) (new_atoms - nTarget); |
186 |
gezelter |
416 |
y = myRandom->getRandom(); |
187 |
chuckv |
434 |
|
188 |
|
|
if (y < exp(- a * x)) { |
189 |
gezelter |
405 |
MolToProcMap[i] = which_proc; |
190 |
|
|
AtomsPerProc[which_proc] += add_atoms; |
191 |
|
|
for (j = 0 ; j < add_atoms; j++ ) { |
192 |
mmeineke |
422 |
AtomToProcMap[atomIndex] = which_proc; |
193 |
|
|
atomIndex++; |
194 |
|
|
} |
195 |
gezelter |
405 |
done = 1; |
196 |
|
|
continue; |
197 |
|
|
} else { |
198 |
|
|
continue; |
199 |
|
|
} |
200 |
|
|
|
201 |
mmeineke |
377 |
} |
202 |
|
|
} |
203 |
gezelter |
405 |
|
204 |
|
|
// Spray out this nonsense to all other processors: |
205 |
|
|
|
206 |
mmeineke |
447 |
MPI_Bcast(MolToProcMap, mpiPlug->nMolGlobal, |
207 |
|
|
MPI_INT, 0, MPI_COMM_WORLD); |
208 |
gezelter |
405 |
|
209 |
mmeineke |
447 |
MPI_Bcast(AtomToProcMap, mpiPlug->nAtomsGlobal, |
210 |
|
|
MPI_INT, 0, MPI_COMM_WORLD); |
211 |
gezelter |
405 |
|
212 |
mmeineke |
447 |
MPI_Bcast(MolComponentType, mpiPlug->nMolGlobal, |
213 |
|
|
MPI_INT, 0, MPI_COMM_WORLD); |
214 |
gezelter |
405 |
|
215 |
mmeineke |
447 |
MPI_Bcast(AtomsPerProc, mpiPlug->numberProcessors, |
216 |
|
|
MPI_INT, 0, MPI_COMM_WORLD); |
217 |
gezelter |
405 |
} else { |
218 |
|
|
|
219 |
|
|
// Listen to your marching orders from processor 0: |
220 |
mmeineke |
377 |
|
221 |
mmeineke |
447 |
MPI_Bcast(MolToProcMap, mpiPlug->nMolGlobal, |
222 |
|
|
MPI_INT, 0, MPI_COMM_WORLD); |
223 |
mmeineke |
377 |
|
224 |
mmeineke |
447 |
MPI_Bcast(AtomToProcMap, mpiPlug->nAtomsGlobal, |
225 |
|
|
MPI_INT, 0, MPI_COMM_WORLD); |
226 |
gezelter |
405 |
|
227 |
mmeineke |
447 |
MPI_Bcast(MolComponentType, mpiPlug->nMolGlobal, |
228 |
|
|
MPI_INT, 0, MPI_COMM_WORLD); |
229 |
gezelter |
405 |
|
230 |
mmeineke |
447 |
MPI_Bcast(AtomsPerProc, mpiPlug->numberProcessors, |
231 |
|
|
MPI_INT, 0, MPI_COMM_WORLD); |
232 |
chuckv |
432 |
|
233 |
|
|
|
234 |
mmeineke |
377 |
} |
235 |
|
|
|
236 |
|
|
|
237 |
gezelter |
405 |
// Let's all check for sanity: |
238 |
|
|
|
239 |
|
|
nmol_local = 0; |
240 |
|
|
for (i = 0 ; i < mpiPlug->nMolGlobal; i++ ) { |
241 |
|
|
if (MolToProcMap[i] == mpiPlug->myNode) { |
242 |
|
|
nmol_local++; |
243 |
|
|
} |
244 |
mmeineke |
377 |
} |
245 |
|
|
|
246 |
gezelter |
405 |
natoms_local = 0; |
247 |
|
|
for (i = 0; i < mpiPlug->nAtomsGlobal; i++) { |
248 |
|
|
if (AtomToProcMap[i] == mpiPlug->myNode) { |
249 |
|
|
natoms_local++; |
250 |
|
|
} |
251 |
|
|
} |
252 |
mmeineke |
377 |
|
253 |
mmeineke |
447 |
MPI_Allreduce(&nmol_local,&nmol_global,1,MPI_INT,MPI_SUM, |
254 |
|
|
MPI_COMM_WORLD); |
255 |
|
|
MPI_Allreduce(&natoms_local,&natoms_global,1,MPI_INT, |
256 |
|
|
MPI_SUM, MPI_COMM_WORLD); |
257 |
mmeineke |
377 |
|
258 |
gezelter |
405 |
if( nmol_global != entryPlug->n_mol ){ |
259 |
|
|
sprintf( painCave.errMsg, |
260 |
|
|
"The sum of all nmol_local, %d, did not equal the " |
261 |
|
|
"total number of molecules, %d.\n", |
262 |
|
|
nmol_global, entryPlug->n_mol ); |
263 |
|
|
painCave.isFatal = 1; |
264 |
|
|
simError(); |
265 |
mmeineke |
377 |
} |
266 |
gezelter |
405 |
|
267 |
|
|
if( natoms_global != entryPlug->n_atoms ){ |
268 |
|
|
sprintf( painCave.errMsg, |
269 |
|
|
"The sum of all natoms_local, %d, did not equal the " |
270 |
|
|
"total number of atoms, %d.\n", |
271 |
|
|
natoms_global, entryPlug->n_atoms ); |
272 |
|
|
painCave.isFatal = 1; |
273 |
|
|
simError(); |
274 |
|
|
} |
275 |
mmeineke |
377 |
|
276 |
|
|
sprintf( checkPointMsg, |
277 |
|
|
"Successfully divided the molecules among the processors.\n" ); |
278 |
|
|
MPIcheckPoint(); |
279 |
|
|
|
280 |
gezelter |
405 |
mpiPlug->myNMol = nmol_local; |
281 |
|
|
mpiPlug->myNlocal = natoms_local; |
282 |
mmeineke |
377 |
|
283 |
|
|
globalIndex = new int[mpiPlug->myNlocal]; |
284 |
gezelter |
405 |
local_index = 0; |
285 |
|
|
for (i = 0; i < mpiPlug->nAtomsGlobal; i++) { |
286 |
|
|
if (AtomToProcMap[i] == mpiPlug->myNode) { |
287 |
mmeineke |
422 |
globalIndex[local_index] = i; |
288 |
chuckv |
406 |
local_index++; |
289 |
gezelter |
405 |
} |
290 |
mmeineke |
377 |
} |
291 |
chuckv |
432 |
|
292 |
gezelter |
416 |
return globalIndex; |
293 |
mmeineke |
377 |
} |
294 |
|
|
|
295 |
|
|
|
296 |
|
|
void mpiSimulation::mpiRefresh( void ){ |
297 |
|
|
|
298 |
|
|
int isError, i; |
299 |
|
|
int *globalIndex = new int[mpiPlug->myNlocal]; |
300 |
|
|
|
301 |
chuckv |
441 |
// Fortran indexing needs to be increased by 1 in order to get the 2 languages to |
302 |
|
|
// not barf |
303 |
mmeineke |
377 |
|
304 |
chuckv |
441 |
for(i=0; i<mpiPlug->myNlocal; i++) globalIndex[i] = entryPlug->atoms[i]->getGlobalIndex()+1; |
305 |
|
|
|
306 |
mmeineke |
377 |
|
307 |
|
|
isError = 0; |
308 |
|
|
setFsimParallel( mpiPlug, &(entryPlug->n_atoms), globalIndex, &isError ); |
309 |
|
|
if( isError ){ |
310 |
|
|
|
311 |
|
|
sprintf( painCave.errMsg, |
312 |
|
|
"mpiRefresh errror: fortran didn't like something we gave it.\n" ); |
313 |
|
|
painCave.isFatal = 1; |
314 |
|
|
simError(); |
315 |
|
|
} |
316 |
|
|
|
317 |
|
|
delete[] globalIndex; |
318 |
|
|
|
319 |
|
|
sprintf( checkPointMsg, |
320 |
|
|
" mpiRefresh successful.\n" ); |
321 |
|
|
MPIcheckPoint(); |
322 |
|
|
} |
323 |
|
|
|
324 |
|
|
|
325 |
|
|
#endif // is_mpi |