1 /*
2 * This file is part of the GROMACS molecular simulation package.
3 *
4 * Copyright (c) 1991-2000, University of Groningen, The Netherlands.
5 * Copyright (c) 2001-2004, The GROMACS development team.
6 * Copyright (c) 2013,2014,2015,2016,2017 by the GROMACS development team.
7 * Copyright (c) 2018,2019,2020, by the GROMACS development team, led by
8 * Mark Abraham, David van der Spoel, Berk Hess, and Erik Lindahl,
9 * and including many others, as listed in the AUTHORS file in the
10 * top-level source directory and at http://www.gromacs.org.
11 *
12 * GROMACS is free software; you can redistribute it and/or
13 * modify it under the terms of the GNU Lesser General Public License
14 * as published by the Free Software Foundation; either version 2.1
15 * of the License, or (at your option) any later version.
16 *
17 * GROMACS is distributed in the hope that it will be useful,
18 * but WITHOUT ANY WARRANTY; without even the implied warranty of
19 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
20 * Lesser General Public License for more details.
21 *
22 * You should have received a copy of the GNU Lesser General Public
23 * License along with GROMACS; if not, see
24 * http://www.gnu.org/licenses, or write to the Free Software Foundation,
25 * Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
26 *
27 * If you want to redistribute modifications to GROMACS, please
28 * consider that scientific software is very special. Version
29 * control is crucial - bugs must be traceable. We will be happy to
30 * consider code for inclusion in the official distribution, but
31 * derived work must not be called official GROMACS. Details are found
32 * in the README & COPYING files - if they are missing, get the
33 * official version at http://www.gromacs.org.
34 *
35 * To help us fund GROMACS development, we humbly ask that you cite
36 * the research papers on the package. Check out http://www.gromacs.org.
37 */
38 #include "gmxpre.h"
39
40 #include "network.h"
41
42 #include "config.h"
43
44 #include <cctype>
45 #include <cstdarg>
46 #include <cstdlib>
47 #include <cstring>
48
49 #include "gromacs/commandline/filenm.h"
50 #include "gromacs/mdtypes/commrec.h"
51 #include "gromacs/utility/basenetwork.h"
52 #include "gromacs/utility/cstringutil.h"
53 #include "gromacs/utility/fatalerror.h"
54 #include "gromacs/utility/futil.h"
55 #include "gromacs/utility/gmxassert.h"
56 #include "gromacs/utility/gmxmpi.h"
57 #include "gromacs/utility/real.h"
58 #include "gromacs/utility/smalloc.h"
59
60 /* The source code in this file should be thread-safe.
61 Please keep it that way. */
62
init_commrec(MPI_Comm communicator)63 CommrecHandle init_commrec(MPI_Comm communicator)
64 {
65 CommrecHandle handle;
66 t_commrec* cr;
67
68 snew(cr, 1);
69 handle.reset(cr);
70
71 int rankInCommunicator, sizeOfCommunicator;
72 #if GMX_MPI
73 # if GMX_LIB_MPI
74 GMX_RELEASE_ASSERT(gmx_mpi_initialized(), "Must have initialized MPI before building commrec");
75 # endif
76 MPI_Comm_rank(communicator, &rankInCommunicator);
77 MPI_Comm_size(communicator, &sizeOfCommunicator);
78 #else
79 GMX_UNUSED_VALUE(communicator);
80 rankInCommunicator = 0;
81 sizeOfCommunicator = 1;
82 #endif
83
84 cr->mpiDefaultCommunicator = communicator;
85 cr->sizeOfDefaultCommunicator = sizeOfCommunicator;
86 cr->rankInDefaultCommunicator = rankInCommunicator;
87
88 // For now, we want things to go horribly wrong if this is used too early...
89 // TODO: Remove when communicators are removed from commrec (#2395)
90 cr->nnodes = -1;
91 cr->nodeid = -1;
92 cr->sim_nodeid = -1;
93 cr->mpi_comm_mysim = MPI_COMM_NULL;
94 cr->mpi_comm_mygroup = MPI_COMM_NULL;
95
96 // TODO cr->duty should not be initialized here
97 cr->duty = (DUTY_PP | DUTY_PME);
98
99 return handle;
100 }
101
done_commrec(t_commrec * cr)102 void done_commrec(t_commrec* cr)
103 {
104 if (MASTER(cr))
105 {
106 if (nullptr != cr->dd)
107 {
108 // TODO: implement
109 // done_domdec(cr->dd);
110 }
111 }
112 #if GMX_MPI
113 // TODO We need to be able to free communicators, but the
114 // structure of the commrec and domdec initialization code makes
115 // it hard to avoid both leaks and double frees.
116 bool mySimIsMyGroup = (cr->mpi_comm_mysim == cr->mpi_comm_mygroup);
117 if (cr->mpi_comm_mysim != MPI_COMM_NULL && cr->mpi_comm_mysim != MPI_COMM_WORLD)
118 {
119 // TODO see above
120 // MPI_Comm_free(&cr->mpi_comm_mysim);
121 }
122 if (!mySimIsMyGroup && cr->mpi_comm_mygroup != MPI_COMM_NULL && cr->mpi_comm_mygroup != MPI_COMM_WORLD)
123 {
124 // TODO see above
125 // MPI_Comm_free(&cr->mpi_comm_mygroup);
126 }
127 #endif
128 sfree(cr);
129 }
130
gmx_setup_nodecomm(FILE gmx_unused * fplog,t_commrec * cr)131 void gmx_setup_nodecomm(FILE gmx_unused* fplog, t_commrec* cr)
132 {
133 gmx_nodecomm_t* nc;
134
135 /* Many MPI implementations do not optimize MPI_Allreduce
136 * (and probably also other global communication calls)
137 * for multi-core nodes connected by a network.
138 * We can optimize such communication by using one MPI call
139 * within each node and one between the nodes.
140 * For MVAPICH2 and Intel MPI this reduces the time for
141 * the global_stat communication by 25%
142 * for 2x2-core 3 GHz Woodcrest connected by mixed DDR/SDR Infiniband.
143 * B. Hess, November 2007
144 */
145
146 nc = &cr->nc;
147
148 nc->bUse = FALSE;
149 #if !GMX_THREAD_MPI
150 # if GMX_MPI
151 int n, rank;
152
153 // TODO PhysicalNodeCommunicator could be extended/used to handle
154 // the need for per-node per-group communicators.
155 MPI_Comm_size(cr->mpi_comm_mygroup, &n);
156 MPI_Comm_rank(cr->mpi_comm_mygroup, &rank);
157
158 int nodehash = gmx_physicalnode_id_hash();
159
160 if (debug)
161 {
162 fprintf(debug, "In gmx_setup_nodecomm: splitting communicator of size %d\n", n);
163 }
164
165
166 /* The intra-node communicator, split on node number */
167 MPI_Comm_split(cr->mpi_comm_mygroup, nodehash, rank, &nc->comm_intra);
168 MPI_Comm_rank(nc->comm_intra, &nc->rank_intra);
169 if (debug)
170 {
171 fprintf(debug, "In gmx_setup_nodecomm: node ID %d rank within node %d\n", rank, nc->rank_intra);
172 }
173 /* The inter-node communicator, split on rank_intra.
174 * We actually only need the one for rank=0,
175 * but it is easier to create them all.
176 */
177 MPI_Comm_split(cr->mpi_comm_mygroup, nc->rank_intra, rank, &nc->comm_inter);
178 /* Check if this really created two step communication */
179 int ng, ni;
180
181 MPI_Comm_size(nc->comm_inter, &ng);
182 MPI_Comm_size(nc->comm_intra, &ni);
183 if (debug)
184 {
185 fprintf(debug, "In gmx_setup_nodecomm: groups %d, my group size %d\n", ng, ni);
186 }
187
188 if (getenv("GMX_NO_NODECOMM") == nullptr && ((ng > 1 && ng < n) || (ni > 1 && ni < n)))
189 {
190 nc->bUse = TRUE;
191 if (fplog)
192 {
193 fprintf(fplog, "Using two step summing over %d groups of on average %.1f ranks\n\n", ng,
194 (real)n / (real)ng);
195 }
196 if (nc->rank_intra > 0)
197 {
198 MPI_Comm_free(&nc->comm_inter);
199 }
200 }
201 else
202 {
203 /* One group or all processes in a separate group, use normal summing */
204 MPI_Comm_free(&nc->comm_inter);
205 MPI_Comm_free(&nc->comm_intra);
206 if (debug)
207 {
208 fprintf(debug,
209 "In gmx_setup_nodecomm: not unsing separate inter- and intra-node "
210 "communicators.\n");
211 }
212 }
213 # endif
214 #else
215 /* tMPI runs only on a single node so just use the nodeid */
216 nc->rank_intra = cr->nodeid;
217 #endif
218 }
219
gmx_barrier(MPI_Comm gmx_unused communicator)220 void gmx_barrier(MPI_Comm gmx_unused communicator)
221 {
222 #if !GMX_MPI
223 GMX_RELEASE_ASSERT(false, "Invalid call to gmx_barrier");
224 #else
225 MPI_Barrier(communicator);
226 #endif
227 }
228
gmx_bcast(int gmx_unused nbytes,void gmx_unused * b,MPI_Comm gmx_unused communicator)229 void gmx_bcast(int gmx_unused nbytes, void gmx_unused* b, MPI_Comm gmx_unused communicator)
230 {
231 #if !GMX_MPI
232 GMX_RELEASE_ASSERT(false, "Invalid call to gmx_bcast");
233 #else
234 MPI_Bcast(b, nbytes, MPI_BYTE, 0, communicator);
235 #endif
236 }
237
gmx_sumd(int gmx_unused nr,double gmx_unused r[],const t_commrec gmx_unused * cr)238 void gmx_sumd(int gmx_unused nr, double gmx_unused r[], const t_commrec gmx_unused* cr)
239 {
240 #if !GMX_MPI
241 GMX_RELEASE_ASSERT(false, "Invalid call to gmx_sumd");
242 #else
243 # if MPI_IN_PLACE_EXISTS
244 if (cr->nc.bUse)
245 {
246 if (cr->nc.rank_intra == 0)
247 {
248 /* Use two step summing. */
249 MPI_Reduce(MPI_IN_PLACE, r, nr, MPI_DOUBLE, MPI_SUM, 0, cr->nc.comm_intra);
250 /* Sum the roots of the internal (intra) buffers. */
251 MPI_Allreduce(MPI_IN_PLACE, r, nr, MPI_DOUBLE, MPI_SUM, cr->nc.comm_inter);
252 }
253 else
254 {
255 /* This is here because of the silly MPI specification
256 that MPI_IN_PLACE should be put in sendbuf instead of recvbuf */
257 MPI_Reduce(r, nullptr, nr, MPI_DOUBLE, MPI_SUM, 0, cr->nc.comm_intra);
258 }
259 MPI_Bcast(r, nr, MPI_DOUBLE, 0, cr->nc.comm_intra);
260 }
261 else
262 {
263 MPI_Allreduce(MPI_IN_PLACE, r, nr, MPI_DOUBLE, MPI_SUM, cr->mpi_comm_mygroup);
264 }
265 # else
266 int i;
267
268 if (nr > cr->mpb->dbuf_alloc)
269 {
270 cr->mpb->dbuf_alloc = nr;
271 srenew(cr->mpb->dbuf, cr->mpb->dbuf_alloc);
272 }
273 if (cr->nc.bUse)
274 {
275 /* Use two step summing */
276 MPI_Allreduce(r, cr->mpb->dbuf, nr, MPI_DOUBLE, MPI_SUM, cr->nc.comm_intra);
277 if (cr->nc.rank_intra == 0)
278 {
279 /* Sum with the buffers reversed */
280 MPI_Allreduce(cr->mpb->dbuf, r, nr, MPI_DOUBLE, MPI_SUM, cr->nc.comm_inter);
281 }
282 MPI_Bcast(r, nr, MPI_DOUBLE, 0, cr->nc.comm_intra);
283 }
284 else
285 {
286 MPI_Allreduce(r, cr->mpb->dbuf, nr, MPI_DOUBLE, MPI_SUM, cr->mpi_comm_mygroup);
287 for (i = 0; i < nr; i++)
288 {
289 r[i] = cr->mpb->dbuf[i];
290 }
291 }
292 # endif
293 #endif
294 }
295
gmx_sumf(int gmx_unused nr,float gmx_unused r[],const t_commrec gmx_unused * cr)296 void gmx_sumf(int gmx_unused nr, float gmx_unused r[], const t_commrec gmx_unused* cr)
297 {
298 #if !GMX_MPI
299 GMX_RELEASE_ASSERT(false, "Invalid call to gmx_sumf");
300 #else
301 # if MPI_IN_PLACE_EXISTS
302 if (cr->nc.bUse)
303 {
304 /* Use two step summing. */
305 if (cr->nc.rank_intra == 0)
306 {
307 MPI_Reduce(MPI_IN_PLACE, r, nr, MPI_FLOAT, MPI_SUM, 0, cr->nc.comm_intra);
308 /* Sum the roots of the internal (intra) buffers */
309 MPI_Allreduce(MPI_IN_PLACE, r, nr, MPI_FLOAT, MPI_SUM, cr->nc.comm_inter);
310 }
311 else
312 {
313 /* This is here because of the silly MPI specification
314 that MPI_IN_PLACE should be put in sendbuf instead of recvbuf */
315 MPI_Reduce(r, nullptr, nr, MPI_FLOAT, MPI_SUM, 0, cr->nc.comm_intra);
316 }
317 MPI_Bcast(r, nr, MPI_FLOAT, 0, cr->nc.comm_intra);
318 }
319 else
320 {
321 MPI_Allreduce(MPI_IN_PLACE, r, nr, MPI_FLOAT, MPI_SUM, cr->mpi_comm_mygroup);
322 }
323 # else
324 int i;
325
326 if (nr > cr->mpb->fbuf_alloc)
327 {
328 cr->mpb->fbuf_alloc = nr;
329 srenew(cr->mpb->fbuf, cr->mpb->fbuf_alloc);
330 }
331 if (cr->nc.bUse)
332 {
333 /* Use two step summing */
334 MPI_Allreduce(r, cr->mpb->fbuf, nr, MPI_FLOAT, MPI_SUM, cr->nc.comm_intra);
335 if (cr->nc.rank_intra == 0)
336 {
337 /* Sum with the buffers reversed */
338 MPI_Allreduce(cr->mpb->fbuf, r, nr, MPI_FLOAT, MPI_SUM, cr->nc.comm_inter);
339 }
340 MPI_Bcast(r, nr, MPI_FLOAT, 0, cr->nc.comm_intra);
341 }
342 else
343 {
344 MPI_Allreduce(r, cr->mpb->fbuf, nr, MPI_FLOAT, MPI_SUM, cr->mpi_comm_mygroup);
345 for (i = 0; i < nr; i++)
346 {
347 r[i] = cr->mpb->fbuf[i];
348 }
349 }
350 # endif
351 #endif
352 }
353
gmx_sumi(int gmx_unused nr,int gmx_unused r[],const t_commrec gmx_unused * cr)354 void gmx_sumi(int gmx_unused nr, int gmx_unused r[], const t_commrec gmx_unused* cr)
355 {
356 #if !GMX_MPI
357 GMX_RELEASE_ASSERT(false, "Invalid call to gmx_sumi");
358 #else
359 # if MPI_IN_PLACE_EXISTS
360 if (cr->nc.bUse)
361 {
362 /* Use two step summing */
363 if (cr->nc.rank_intra == 0)
364 {
365 MPI_Reduce(MPI_IN_PLACE, r, nr, MPI_INT, MPI_SUM, 0, cr->nc.comm_intra);
366 /* Sum with the buffers reversed */
367 MPI_Allreduce(MPI_IN_PLACE, r, nr, MPI_INT, MPI_SUM, cr->nc.comm_inter);
368 }
369 else
370 {
371 /* This is here because of the silly MPI specification
372 that MPI_IN_PLACE should be put in sendbuf instead of recvbuf */
373 MPI_Reduce(r, nullptr, nr, MPI_INT, MPI_SUM, 0, cr->nc.comm_intra);
374 }
375 MPI_Bcast(r, nr, MPI_INT, 0, cr->nc.comm_intra);
376 }
377 else
378 {
379 MPI_Allreduce(MPI_IN_PLACE, r, nr, MPI_INT, MPI_SUM, cr->mpi_comm_mygroup);
380 }
381 # else
382 int i;
383
384 if (nr > cr->mpb->ibuf_alloc)
385 {
386 cr->mpb->ibuf_alloc = nr;
387 srenew(cr->mpb->ibuf, cr->mpb->ibuf_alloc);
388 }
389 if (cr->nc.bUse)
390 {
391 /* Use two step summing */
392 MPI_Allreduce(r, cr->mpb->ibuf, nr, MPI_INT, MPI_SUM, cr->nc.comm_intra);
393 if (cr->nc.rank_intra == 0)
394 {
395 /* Sum with the buffers reversed */
396 MPI_Allreduce(cr->mpb->ibuf, r, nr, MPI_INT, MPI_SUM, cr->nc.comm_inter);
397 }
398 MPI_Bcast(r, nr, MPI_INT, 0, cr->nc.comm_intra);
399 }
400 else
401 {
402 MPI_Allreduce(r, cr->mpb->ibuf, nr, MPI_INT, MPI_SUM, cr->mpi_comm_mygroup);
403 for (i = 0; i < nr; i++)
404 {
405 r[i] = cr->mpb->ibuf[i];
406 }
407 }
408 # endif
409 #endif
410 }
411
gmx_sumli(int gmx_unused nr,int64_t gmx_unused r[],const t_commrec gmx_unused * cr)412 void gmx_sumli(int gmx_unused nr, int64_t gmx_unused r[], const t_commrec gmx_unused* cr)
413 {
414 #if !GMX_MPI
415 GMX_RELEASE_ASSERT(false, "Invalid call to gmx_sumli");
416 #else
417 # if MPI_IN_PLACE_EXISTS
418 if (cr->nc.bUse)
419 {
420 /* Use two step summing */
421 if (cr->nc.rank_intra == 0)
422 {
423 MPI_Reduce(MPI_IN_PLACE, r, nr, MPI_INT64_T, MPI_SUM, 0, cr->nc.comm_intra);
424 /* Sum with the buffers reversed */
425 MPI_Allreduce(MPI_IN_PLACE, r, nr, MPI_INT64_T, MPI_SUM, cr->nc.comm_inter);
426 }
427 else
428 {
429 /* This is here because of the silly MPI specification
430 that MPI_IN_PLACE should be put in sendbuf instead of recvbuf */
431 MPI_Reduce(r, nullptr, nr, MPI_INT64_T, MPI_SUM, 0, cr->nc.comm_intra);
432 }
433 MPI_Bcast(r, nr, MPI_INT64_T, 0, cr->nc.comm_intra);
434 }
435 else
436 {
437 MPI_Allreduce(MPI_IN_PLACE, r, nr, MPI_INT64_T, MPI_SUM, cr->mpi_comm_mygroup);
438 }
439 # else
440 int i;
441
442 if (nr > cr->mpb->libuf_alloc)
443 {
444 cr->mpb->libuf_alloc = nr;
445 srenew(cr->mpb->libuf, cr->mpb->libuf_alloc);
446 }
447 if (cr->nc.bUse)
448 {
449 /* Use two step summing */
450 MPI_Allreduce(r, cr->mpb->libuf, nr, MPI_INT64_T, MPI_SUM, cr->nc.comm_intra);
451 if (cr->nc.rank_intra == 0)
452 {
453 /* Sum with the buffers reversed */
454 MPI_Allreduce(cr->mpb->libuf, r, nr, MPI_INT64_T, MPI_SUM, cr->nc.comm_inter);
455 }
456 MPI_Bcast(r, nr, MPI_INT64_T, 0, cr->nc.comm_intra);
457 }
458 else
459 {
460 MPI_Allreduce(r, cr->mpb->libuf, nr, MPI_INT64_T, MPI_SUM, cr->mpi_comm_mygroup);
461 for (i = 0; i < nr; i++)
462 {
463 r[i] = cr->mpb->libuf[i];
464 }
465 }
466 # endif
467 #endif
468 }
469
opt2fn_master(const char * opt,int nfile,const t_filenm fnm[],t_commrec * cr)470 const char* opt2fn_master(const char* opt, int nfile, const t_filenm fnm[], t_commrec* cr)
471 {
472 return SIMMASTER(cr) ? opt2fn(opt, nfile, fnm) : nullptr;
473 }
474
gmx_fatal_collective(int f_errno,const char * file,int line,MPI_Comm comm,gmx_bool bMaster,gmx_fmtstr const char * fmt,...)475 void gmx_fatal_collective(int f_errno,
476 const char* file,
477 int line,
478 MPI_Comm comm,
479 gmx_bool bMaster,
480 gmx_fmtstr const char* fmt,
481 ...)
482 {
483 va_list ap;
484 gmx_bool bFinalize;
485 #if GMX_MPI
486 int result;
487 /* Check if we are calling on all processes in MPI_COMM_WORLD */
488 MPI_Comm_compare(comm, MPI_COMM_WORLD, &result);
489 /* Any result except MPI_UNEQUAL allows us to call MPI_Finalize */
490 bFinalize = (result != MPI_UNEQUAL);
491 #else
492 GMX_UNUSED_VALUE(comm);
493 bFinalize = TRUE;
494 #endif
495
496 va_start(ap, fmt);
497 gmx_fatal_mpi_va(f_errno, file, line, bMaster, bFinalize, fmt, ap);
498 va_end(ap);
499 }
500