2 * This file is part of the GROMACS molecular simulation package.
4 * Copyright (c) 2005,2006,2007,2008,2009,2010,2011,2012,2013,2014,2015,2016,2017,2018, by the GROMACS development team, led by
5 * Mark Abraham, David van der Spoel, Berk Hess, and Erik Lindahl,
6 * and including many others, as listed in the AUTHORS file in the
7 * top-level source directory and at http://www.gromacs.org.
9 * GROMACS is free software; you can redistribute it and/or
10 * modify it under the terms of the GNU Lesser General Public License
11 * as published by the Free Software Foundation; either version 2.1
12 * of the License, or (at your option) any later version.
14 * GROMACS is distributed in the hope that it will be useful,
15 * but WITHOUT ANY WARRANTY; without even the implied warranty of
16 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
17 * Lesser General Public License for more details.
19 * You should have received a copy of the GNU Lesser General Public
20 * License along with GROMACS; if not, see
21 * http://www.gnu.org/licenses, or write to the Free Software Foundation,
22 * Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
24 * If you want to redistribute modifications to GROMACS, please
25 * consider that scientific software is very special. Version
26 * control is crucial - bugs must be traceable. We will be happy to
27 * consider code for inclusion in the official distribution, but
28 * derived work must not be called official GROMACS. Details are found
29 * in the README & COPYING files - if they are missing, get the
30 * official version at http://www.gromacs.org.
32 * To help us fund GROMACS development, we humbly ask that you cite
33 * the research papers on the package. Check out http://www.gromacs.org.
52 #include "gromacs/compat/make_unique.h"
53 #include "gromacs/domdec/collect.h"
54 #include "gromacs/domdec/dlb.h"
55 #include "gromacs/domdec/dlbtiming.h"
56 #include "gromacs/domdec/domdec_network.h"
57 #include "gromacs/domdec/ga2la.h"
58 #include "gromacs/domdec/partition.h"
59 #include "gromacs/gmxlib/network.h"
60 #include "gromacs/gmxlib/nrnb.h"
61 #include "gromacs/gpu_utils/gpu_utils.h"
62 #include "gromacs/hardware/hw_info.h"
63 #include "gromacs/listed-forces/manage-threading.h"
64 #include "gromacs/math/vec.h"
65 #include "gromacs/math/vectypes.h"
66 #include "gromacs/mdlib/calc_verletbuf.h"
67 #include "gromacs/mdlib/constr.h"
68 #include "gromacs/mdlib/constraintrange.h"
69 #include "gromacs/mdlib/mdrun.h"
70 #include "gromacs/mdlib/updategroups.h"
71 #include "gromacs/mdlib/vsite.h"
72 #include "gromacs/mdtypes/commrec.h"
73 #include "gromacs/mdtypes/inputrec.h"
74 #include "gromacs/mdtypes/state.h"
75 #include "gromacs/pbcutil/ishift.h"
76 #include "gromacs/pbcutil/pbc.h"
77 #include "gromacs/pulling/pull.h"
78 #include "gromacs/timing/wallcycle.h"
79 #include "gromacs/topology/block.h"
80 #include "gromacs/topology/idef.h"
81 #include "gromacs/topology/ifunc.h"
82 #include "gromacs/topology/mtop_lookup.h"
83 #include "gromacs/topology/mtop_util.h"
84 #include "gromacs/topology/topology.h"
85 #include "gromacs/utility/basedefinitions.h"
86 #include "gromacs/utility/basenetwork.h"
87 #include "gromacs/utility/cstringutil.h"
88 #include "gromacs/utility/exceptions.h"
89 #include "gromacs/utility/fatalerror.h"
90 #include "gromacs/utility/gmxmpi.h"
91 #include "gromacs/utility/logger.h"
92 #include "gromacs/utility/real.h"
93 #include "gromacs/utility/smalloc.h"
94 #include "gromacs/utility/strconvert.h"
95 #include "gromacs/utility/stringstream.h"
96 #include "gromacs/utility/stringutil.h"
97 #include "gromacs/utility/textwriter.h"
99 #include "atomdistribution.h"
101 #include "cellsizes.h"
102 #include "distribute.h"
103 #include "domdec_constraints.h"
104 #include "domdec_internal.h"
105 #include "domdec_vsite.h"
106 #include "redistribute.h"
109 static const char *edlbs_names[int(DlbState::nr)] = { "off", "auto", "locked", "on", "on" };
111 /* The size per charge group of the cggl_flag buffer in gmx_domdec_comm_t */
114 /* The flags for the cggl_flag buffer in gmx_domdec_comm_t */
115 #define DD_FLAG_NRCG 65535
116 #define DD_FLAG_FW(d) (1<<(16+(d)*2))
117 #define DD_FLAG_BW(d) (1<<(16+(d)*2+1))
119 /* The DD zone order */
120 static const ivec dd_zo[DD_MAXZONE] =
121 {{0, 0, 0}, {1, 0, 0}, {1, 1, 0}, {0, 1, 0}, {0, 1, 1}, {0, 0, 1}, {1, 0, 1}, {1, 1, 1}};
123 /* The non-bonded zone-pair setup for domain decomposition
124 * The first number is the i-zone, the second number the first j-zone seen by
125 * this i-zone, the third number the last+1 j-zone seen by this i-zone.
126 * As is, this is for 3D decomposition, where there are 4 i-zones.
127 * With 2D decomposition use only the first 2 i-zones and a last+1 j-zone of 4.
128 * With 1D decomposition use only the first i-zone and a last+1 j-zone of 2.
131 ddNonbondedZonePairRanges[DD_MAXIZONE][3] = {{0, 0, 8},
138 #define dd_index(n,i) ((((i)[ZZ]*(n)[YY] + (i)[YY])*(n)[XX]) + (i)[XX])
140 static void index2xyz(ivec nc,int ind,ivec xyz)
142 xyz[XX] = ind % nc[XX];
143 xyz[YY] = (ind / nc[XX]) % nc[YY];
144 xyz[ZZ] = ind / (nc[YY]*nc[XX]);
148 static void ddindex2xyz(const ivec nc, int ind, ivec xyz)
150 xyz[XX] = ind / (nc[YY]*nc[ZZ]);
151 xyz[YY] = (ind / nc[ZZ]) % nc[YY];
152 xyz[ZZ] = ind % nc[ZZ];
155 static int ddcoord2ddnodeid(gmx_domdec_t *dd, ivec c)
160 ddindex = dd_index(dd->nc, c);
161 if (dd->comm->bCartesianPP_PME)
163 ddnodeid = dd->comm->ddindex2ddnodeid[ddindex];
165 else if (dd->comm->bCartesianPP)
168 MPI_Cart_rank(dd->mpi_comm_all, c, &ddnodeid);
179 int ddglatnr(const gmx_domdec_t *dd, int i)
189 if (i >= dd->comm->atomRanges.numAtomsTotal())
191 gmx_fatal(FARGS, "glatnr called with %d, which is larger than the local number of atoms (%d)", i, dd->comm->atomRanges.numAtomsTotal());
193 atnr = dd->globalAtomIndices[i] + 1;
199 t_block *dd_charge_groups_global(gmx_domdec_t *dd)
201 return &dd->comm->cgs_gl;
204 void dd_store_state(gmx_domdec_t *dd, t_state *state)
208 if (state->ddp_count != dd->ddp_count)
210 gmx_incons("The MD state does not match the domain decomposition state");
213 state->cg_gl.resize(dd->ncg_home);
214 for (i = 0; i < dd->ncg_home; i++)
216 state->cg_gl[i] = dd->globalAtomGroupIndices[i];
219 state->ddp_count_cg_gl = dd->ddp_count;
222 gmx_domdec_zones_t *domdec_zones(gmx_domdec_t *dd)
224 return &dd->comm->zones;
227 void dd_get_ns_ranges(const gmx_domdec_t *dd, int icg,
228 int *jcg0, int *jcg1, ivec shift0, ivec shift1)
230 gmx_domdec_zones_t *zones;
233 zones = &dd->comm->zones;
236 while (icg >= zones->izone[izone].cg1)
245 else if (izone < zones->nizone)
247 *jcg0 = zones->izone[izone].jcg0;
251 gmx_fatal(FARGS, "DD icg %d out of range: izone (%d) >= nizone (%d)",
252 icg, izone, zones->nizone);
255 *jcg1 = zones->izone[izone].jcg1;
257 for (d = 0; d < dd->ndim; d++)
260 shift0[dim] = zones->izone[izone].shift0[dim];
261 shift1[dim] = zones->izone[izone].shift1[dim];
262 if (dd->comm->tric_dir[dim] || (isDlbOn(dd->comm) && d > 0))
264 /* A conservative approach, this can be optimized */
271 int dd_numHomeAtoms(const gmx_domdec_t &dd)
273 return dd.comm->atomRanges.numHomeAtoms();
276 int dd_natoms_mdatoms(const gmx_domdec_t *dd)
278 /* We currently set mdatoms entries for all atoms:
279 * local + non-local + communicated for vsite + constraints
282 return dd->comm->atomRanges.numAtomsTotal();
285 int dd_natoms_vsite(const gmx_domdec_t *dd)
287 return dd->comm->atomRanges.end(DDAtomRanges::Type::Vsites);
290 void dd_get_constraint_range(const gmx_domdec_t *dd, int *at_start, int *at_end)
292 *at_start = dd->comm->atomRanges.start(DDAtomRanges::Type::Constraints);
293 *at_end = dd->comm->atomRanges.end(DDAtomRanges::Type::Constraints);
296 void dd_move_x(gmx_domdec_t *dd,
298 gmx::ArrayRef<gmx::RVec> x,
299 gmx_wallcycle *wcycle)
301 wallcycle_start(wcycle, ewcMOVEX);
304 gmx_domdec_comm_t *comm;
305 gmx_domdec_comm_dim_t *cd;
306 rvec shift = {0, 0, 0};
307 gmx_bool bPBC, bScrew;
311 const gmx::RangePartitioning &atomGrouping = dd->atomGrouping();
314 nat_tot = comm->atomRanges.numHomeAtoms();
315 for (int d = 0; d < dd->ndim; d++)
317 bPBC = (dd->ci[dd->dim[d]] == 0);
318 bScrew = (bPBC && dd->bScrewPBC && dd->dim[d] == XX);
321 copy_rvec(box[dd->dim[d]], shift);
324 for (const gmx_domdec_ind_t &ind : cd->ind)
326 DDBufferAccess<gmx::RVec> sendBufferAccess(comm->rvecBuffer, ind.nsend[nzone + 1]);
327 gmx::ArrayRef<gmx::RVec> &sendBuffer = sendBufferAccess.buffer;
331 for (int g : ind.index)
333 for (int j : atomGrouping.block(g))
335 sendBuffer[n] = x[j];
342 for (int g : ind.index)
344 for (int j : atomGrouping.block(g))
346 /* We need to shift the coordinates */
347 for (int d = 0; d < DIM; d++)
349 sendBuffer[n][d] = x[j][d] + shift[d];
357 for (int g : ind.index)
359 for (int j : atomGrouping.block(g))
362 sendBuffer[n][XX] = x[j][XX] + shift[XX];
364 * This operation requires a special shift force
365 * treatment, which is performed in calc_vir.
367 sendBuffer[n][YY] = box[YY][YY] - x[j][YY];
368 sendBuffer[n][ZZ] = box[ZZ][ZZ] - x[j][ZZ];
374 DDBufferAccess<gmx::RVec> receiveBufferAccess(comm->rvecBuffer2, cd->receiveInPlace ? 0 : ind.nrecv[nzone + 1]);
376 gmx::ArrayRef<gmx::RVec> receiveBuffer;
377 if (cd->receiveInPlace)
379 receiveBuffer = gmx::arrayRefFromArray(x.data() + nat_tot, ind.nrecv[nzone + 1]);
383 receiveBuffer = receiveBufferAccess.buffer;
385 /* Send and receive the coordinates */
386 ddSendrecv(dd, d, dddirBackward,
387 sendBuffer, receiveBuffer);
389 if (!cd->receiveInPlace)
392 for (int zone = 0; zone < nzone; zone++)
394 for (int i = ind.cell2at0[zone]; i < ind.cell2at1[zone]; i++)
396 x[i] = receiveBuffer[j++];
400 nat_tot += ind.nrecv[nzone+1];
405 wallcycle_stop(wcycle, ewcMOVEX);
408 void dd_move_f(gmx_domdec_t *dd,
409 gmx::ArrayRef<gmx::RVec> f,
411 gmx_wallcycle *wcycle)
413 wallcycle_start(wcycle, ewcMOVEF);
416 gmx_domdec_comm_t *comm;
417 gmx_domdec_comm_dim_t *cd;
420 gmx_bool bShiftForcesNeedPbc, bScrew;
424 const gmx::RangePartitioning &atomGrouping = dd->atomGrouping();
426 nzone = comm->zones.n/2;
427 nat_tot = comm->atomRanges.end(DDAtomRanges::Type::Zones);
428 for (int d = dd->ndim-1; d >= 0; d--)
430 /* Only forces in domains near the PBC boundaries need to
431 consider PBC in the treatment of fshift */
432 bShiftForcesNeedPbc = (dd->ci[dd->dim[d]] == 0);
433 bScrew = (bShiftForcesNeedPbc && dd->bScrewPBC && dd->dim[d] == XX);
434 if (fshift == nullptr && !bScrew)
436 bShiftForcesNeedPbc = FALSE;
438 /* Determine which shift vector we need */
444 for (int p = cd->numPulses() - 1; p >= 0; p--)
446 const gmx_domdec_ind_t &ind = cd->ind[p];
447 DDBufferAccess<gmx::RVec> receiveBufferAccess(comm->rvecBuffer, ind.nsend[nzone + 1]);
448 gmx::ArrayRef<gmx::RVec> &receiveBuffer = receiveBufferAccess.buffer;
450 nat_tot -= ind.nrecv[nzone+1];
452 DDBufferAccess<gmx::RVec> sendBufferAccess(comm->rvecBuffer2, cd->receiveInPlace ? 0 : ind.nrecv[nzone + 1]);
454 gmx::ArrayRef<gmx::RVec> sendBuffer;
455 if (cd->receiveInPlace)
457 sendBuffer = gmx::arrayRefFromArray(f.data() + nat_tot, ind.nrecv[nzone + 1]);
461 sendBuffer = sendBufferAccess.buffer;
463 for (int zone = 0; zone < nzone; zone++)
465 for (int i = ind.cell2at0[zone]; i < ind.cell2at1[zone]; i++)
467 sendBuffer[j++] = f[i];
471 /* Communicate the forces */
472 ddSendrecv(dd, d, dddirForward,
473 sendBuffer, receiveBuffer);
474 /* Add the received forces */
476 if (!bShiftForcesNeedPbc)
478 for (int g : ind.index)
480 for (int j : atomGrouping.block(g))
482 for (int d = 0; d < DIM; d++)
484 f[j][d] += receiveBuffer[n][d];
492 /* fshift should always be defined if this function is
493 * called when bShiftForcesNeedPbc is true */
494 assert(nullptr != fshift);
495 for (int g : ind.index)
497 for (int j : atomGrouping.block(g))
499 for (int d = 0; d < DIM; d++)
501 f[j][d] += receiveBuffer[n][d];
503 /* Add this force to the shift force */
504 for (int d = 0; d < DIM; d++)
506 fshift[is][d] += receiveBuffer[n][d];
514 for (int g : ind.index)
516 for (int j : atomGrouping.block(g))
518 /* Rotate the force */
519 f[j][XX] += receiveBuffer[n][XX];
520 f[j][YY] -= receiveBuffer[n][YY];
521 f[j][ZZ] -= receiveBuffer[n][ZZ];
524 /* Add this force to the shift force */
525 for (int d = 0; d < DIM; d++)
527 fshift[is][d] += receiveBuffer[n][d];
537 wallcycle_stop(wcycle, ewcMOVEF);
540 /* Convenience function for extracting a real buffer from an rvec buffer
542 * To reduce the number of temporary communication buffers and avoid
543 * cache polution, we reuse gmx::RVec buffers for storing reals.
544 * This functions return a real buffer reference with the same number
545 * of elements as the gmx::RVec buffer (so 1/3 of the size in bytes).
547 static gmx::ArrayRef<real>
548 realArrayRefFromRvecArrayRef(gmx::ArrayRef<gmx::RVec> arrayRef)
550 return gmx::arrayRefFromArray(as_rvec_array(arrayRef.data())[0],
554 void dd_atom_spread_real(gmx_domdec_t *dd, real v[])
557 gmx_domdec_comm_t *comm;
558 gmx_domdec_comm_dim_t *cd;
562 const gmx::RangePartitioning &atomGrouping = dd->atomGrouping();
565 nat_tot = comm->atomRanges.numHomeAtoms();
566 for (int d = 0; d < dd->ndim; d++)
569 for (const gmx_domdec_ind_t &ind : cd->ind)
571 /* Note: We provision for RVec instead of real, so a factor of 3
572 * more than needed. The buffer actually already has this size
573 * and we pass a plain pointer below, so this does not matter.
575 DDBufferAccess<gmx::RVec> sendBufferAccess(comm->rvecBuffer, ind.nsend[nzone + 1]);
576 gmx::ArrayRef<real> sendBuffer = realArrayRefFromRvecArrayRef(sendBufferAccess.buffer);
578 for (int g : ind.index)
580 for (int j : atomGrouping.block(g))
582 sendBuffer[n++] = v[j];
586 DDBufferAccess<gmx::RVec> receiveBufferAccess(comm->rvecBuffer2, cd->receiveInPlace ? 0 : ind.nrecv[nzone + 1]);
588 gmx::ArrayRef<real> receiveBuffer;
589 if (cd->receiveInPlace)
591 receiveBuffer = gmx::arrayRefFromArray(v + nat_tot, ind.nrecv[nzone + 1]);
595 receiveBuffer = realArrayRefFromRvecArrayRef(receiveBufferAccess.buffer);
597 /* Send and receive the data */
598 ddSendrecv(dd, d, dddirBackward,
599 sendBuffer, receiveBuffer);
600 if (!cd->receiveInPlace)
603 for (int zone = 0; zone < nzone; zone++)
605 for (int i = ind.cell2at0[zone]; i < ind.cell2at1[zone]; i++)
607 v[i] = receiveBuffer[j++];
611 nat_tot += ind.nrecv[nzone+1];
617 void dd_atom_sum_real(gmx_domdec_t *dd, real v[])
620 gmx_domdec_comm_t *comm;
621 gmx_domdec_comm_dim_t *cd;
625 const gmx::RangePartitioning &atomGrouping = dd->atomGrouping();
627 nzone = comm->zones.n/2;
628 nat_tot = comm->atomRanges.end(DDAtomRanges::Type::Zones);
629 for (int d = dd->ndim-1; d >= 0; d--)
632 for (int p = cd->numPulses() - 1; p >= 0; p--)
634 const gmx_domdec_ind_t &ind = cd->ind[p];
636 /* Note: We provision for RVec instead of real, so a factor of 3
637 * more than needed. The buffer actually already has this size
638 * and we typecast, so this works as intended.
640 DDBufferAccess<gmx::RVec> receiveBufferAccess(comm->rvecBuffer, ind.nsend[nzone + 1]);
641 gmx::ArrayRef<real> receiveBuffer = realArrayRefFromRvecArrayRef(receiveBufferAccess.buffer);
642 nat_tot -= ind.nrecv[nzone + 1];
644 DDBufferAccess<gmx::RVec> sendBufferAccess(comm->rvecBuffer2, cd->receiveInPlace ? 0 : ind.nrecv[nzone + 1]);
646 gmx::ArrayRef<real> sendBuffer;
647 if (cd->receiveInPlace)
649 sendBuffer = gmx::arrayRefFromArray(v + nat_tot, ind.nrecv[nzone + 1]);
653 sendBuffer = realArrayRefFromRvecArrayRef(sendBufferAccess.buffer);
655 for (int zone = 0; zone < nzone; zone++)
657 for (int i = ind.cell2at0[zone]; i < ind.cell2at1[zone]; i++)
659 sendBuffer[j++] = v[i];
663 /* Communicate the forces */
664 ddSendrecv(dd, d, dddirForward,
665 sendBuffer, receiveBuffer);
666 /* Add the received forces */
668 for (int g : ind.index)
670 for (int j : atomGrouping.block(g))
672 v[j] += receiveBuffer[n];
681 real dd_cutoff_multibody(const gmx_domdec_t *dd)
683 gmx_domdec_comm_t *comm;
690 if (comm->bInterCGBondeds)
692 if (comm->cutoff_mbody > 0)
694 r = comm->cutoff_mbody;
698 /* cutoff_mbody=0 means we do not have DLB */
699 r = comm->cellsize_min[dd->dim[0]];
700 for (di = 1; di < dd->ndim; di++)
702 r = std::min(r, comm->cellsize_min[dd->dim[di]]);
706 r = std::max(r, comm->cutoff_mbody);
710 r = std::min(r, comm->cutoff);
718 real dd_cutoff_twobody(const gmx_domdec_t *dd)
722 r_mb = dd_cutoff_multibody(dd);
724 return std::max(dd->comm->cutoff, r_mb);
728 static void dd_cart_coord2pmecoord(const gmx_domdec_t *dd, const ivec coord,
733 nc = dd->nc[dd->comm->cartpmedim];
734 ntot = dd->comm->ntot[dd->comm->cartpmedim];
735 copy_ivec(coord, coord_pme);
736 coord_pme[dd->comm->cartpmedim] =
737 nc + (coord[dd->comm->cartpmedim]*(ntot - nc) + (ntot - nc)/2)/nc;
740 static int ddindex2pmeindex(const gmx_domdec_t *dd, int ddindex)
745 npme = dd->comm->npmenodes;
747 /* Here we assign a PME node to communicate with this DD node
748 * by assuming that the major index of both is x.
749 * We add cr->npmenodes/2 to obtain an even distribution.
751 return (ddindex*npme + npme/2)/npp;
754 static int *dd_interleaved_pme_ranks(const gmx_domdec_t *dd)
759 snew(pme_rank, dd->comm->npmenodes);
761 for (i = 0; i < dd->nnodes; i++)
763 p0 = ddindex2pmeindex(dd, i);
764 p1 = ddindex2pmeindex(dd, i+1);
765 if (i+1 == dd->nnodes || p1 > p0)
769 fprintf(debug, "pme_rank[%d] = %d\n", n, i+1+n);
771 pme_rank[n] = i + 1 + n;
779 static int gmx_ddcoord2pmeindex(const t_commrec *cr, int x, int y, int z)
787 if (dd->comm->bCartesian) {
788 gmx_ddindex2xyz(dd->nc,ddindex,coords);
789 dd_coords2pmecoords(dd,coords,coords_pme);
790 copy_ivec(dd->ntot,nc);
791 nc[dd->cartpmedim] -= dd->nc[dd->cartpmedim];
792 coords_pme[dd->cartpmedim] -= dd->nc[dd->cartpmedim];
794 slab = (coords_pme[XX]*nc[YY] + coords_pme[YY])*nc[ZZ] + coords_pme[ZZ];
796 slab = (ddindex*cr->npmenodes + cr->npmenodes/2)/dd->nnodes;
802 slab = ddindex2pmeindex(dd, dd_index(dd->nc, coords));
807 static int ddcoord2simnodeid(const t_commrec *cr, int x, int y, int z)
809 gmx_domdec_comm_t *comm;
811 int ddindex, nodeid = -1;
818 if (comm->bCartesianPP_PME)
821 MPI_Cart_rank(cr->mpi_comm_mysim, coords, &nodeid);
826 ddindex = dd_index(cr->dd->nc, coords);
827 if (comm->bCartesianPP)
829 nodeid = comm->ddindex2simnodeid[ddindex];
835 nodeid = ddindex + gmx_ddcoord2pmeindex(cr, x, y, z);
847 static int dd_simnode2pmenode(const gmx_domdec_t *dd,
848 const t_commrec gmx_unused *cr,
853 const gmx_domdec_comm_t *comm = dd->comm;
855 /* This assumes a uniform x domain decomposition grid cell size */
856 if (comm->bCartesianPP_PME)
859 ivec coord, coord_pme;
860 MPI_Cart_coords(cr->mpi_comm_mysim, sim_nodeid, DIM, coord);
861 if (coord[comm->cartpmedim] < dd->nc[comm->cartpmedim])
863 /* This is a PP node */
864 dd_cart_coord2pmecoord(dd, coord, coord_pme);
865 MPI_Cart_rank(cr->mpi_comm_mysim, coord_pme, &pmenode);
869 else if (comm->bCartesianPP)
871 if (sim_nodeid < dd->nnodes)
873 pmenode = dd->nnodes + ddindex2pmeindex(dd, sim_nodeid);
878 /* This assumes DD cells with identical x coordinates
879 * are numbered sequentially.
881 if (dd->comm->pmenodes == nullptr)
883 if (sim_nodeid < dd->nnodes)
885 /* The DD index equals the nodeid */
886 pmenode = dd->nnodes + ddindex2pmeindex(dd, sim_nodeid);
892 while (sim_nodeid > dd->comm->pmenodes[i])
896 if (sim_nodeid < dd->comm->pmenodes[i])
898 pmenode = dd->comm->pmenodes[i];
906 NumPmeDomains getNumPmeDomains(const gmx_domdec_t *dd)
911 dd->comm->npmenodes_x, dd->comm->npmenodes_y
922 std::vector<int> get_pme_ddranks(const t_commrec *cr, int pmenodeid)
926 ivec coord, coord_pme;
930 std::vector<int> ddranks;
931 ddranks.reserve((dd->nnodes+cr->npmenodes-1)/cr->npmenodes);
933 for (x = 0; x < dd->nc[XX]; x++)
935 for (y = 0; y < dd->nc[YY]; y++)
937 for (z = 0; z < dd->nc[ZZ]; z++)
939 if (dd->comm->bCartesianPP_PME)
944 dd_cart_coord2pmecoord(dd, coord, coord_pme);
945 if (dd->ci[XX] == coord_pme[XX] &&
946 dd->ci[YY] == coord_pme[YY] &&
947 dd->ci[ZZ] == coord_pme[ZZ])
949 ddranks.push_back(ddcoord2simnodeid(cr, x, y, z));
954 /* The slab corresponds to the nodeid in the PME group */
955 if (gmx_ddcoord2pmeindex(cr, x, y, z) == pmenodeid)
957 ddranks.push_back(ddcoord2simnodeid(cr, x, y, z));
966 static gmx_bool receive_vir_ener(const gmx_domdec_t *dd, const t_commrec *cr)
968 gmx_bool bReceive = TRUE;
970 if (cr->npmenodes < dd->nnodes)
972 gmx_domdec_comm_t *comm = dd->comm;
973 if (comm->bCartesianPP_PME)
976 int pmenode = dd_simnode2pmenode(dd, cr, cr->sim_nodeid);
978 MPI_Cart_coords(cr->mpi_comm_mysim, cr->sim_nodeid, DIM, coords);
979 coords[comm->cartpmedim]++;
980 if (coords[comm->cartpmedim] < dd->nc[comm->cartpmedim])
983 MPI_Cart_rank(cr->mpi_comm_mysim, coords, &rank);
984 if (dd_simnode2pmenode(dd, cr, rank) == pmenode)
986 /* This is not the last PP node for pmenode */
991 GMX_RELEASE_ASSERT(false, "Without MPI we should not have Cartesian PP-PME with #PMEnodes < #DDnodes");
996 int pmenode = dd_simnode2pmenode(dd, cr, cr->sim_nodeid);
997 if (cr->sim_nodeid+1 < cr->nnodes &&
998 dd_simnode2pmenode(dd, cr, cr->sim_nodeid+1) == pmenode)
1000 /* This is not the last PP node for pmenode */
1009 static void set_slb_pme_dim_f(gmx_domdec_t *dd, int dim, real **dim_f)
1011 gmx_domdec_comm_t *comm;
1016 snew(*dim_f, dd->nc[dim]+1);
1018 for (i = 1; i < dd->nc[dim]; i++)
1020 if (comm->slb_frac[dim])
1022 (*dim_f)[i] = (*dim_f)[i-1] + comm->slb_frac[dim][i-1];
1026 (*dim_f)[i] = static_cast<real>(i)/static_cast<real>(dd->nc[dim]);
1029 (*dim_f)[dd->nc[dim]] = 1;
1032 static void init_ddpme(gmx_domdec_t *dd, gmx_ddpme_t *ddpme, int dimind)
1034 int pmeindex, slab, nso, i;
1037 if (dimind == 0 && dd->dim[0] == YY && dd->comm->npmenodes_x == 1)
1043 ddpme->dim = dimind;
1045 ddpme->dim_match = (ddpme->dim == dd->dim[dimind]);
1047 ddpme->nslab = (ddpme->dim == 0 ?
1048 dd->comm->npmenodes_x :
1049 dd->comm->npmenodes_y);
1051 if (ddpme->nslab <= 1)
1056 nso = dd->comm->npmenodes/ddpme->nslab;
1057 /* Determine for each PME slab the PP location range for dimension dim */
1058 snew(ddpme->pp_min, ddpme->nslab);
1059 snew(ddpme->pp_max, ddpme->nslab);
1060 for (slab = 0; slab < ddpme->nslab; slab++)
1062 ddpme->pp_min[slab] = dd->nc[dd->dim[dimind]] - 1;
1063 ddpme->pp_max[slab] = 0;
1065 for (i = 0; i < dd->nnodes; i++)
1067 ddindex2xyz(dd->nc, i, xyz);
1068 /* For y only use our y/z slab.
1069 * This assumes that the PME x grid size matches the DD grid size.
1071 if (dimind == 0 || xyz[XX] == dd->ci[XX])
1073 pmeindex = ddindex2pmeindex(dd, i);
1076 slab = pmeindex/nso;
1080 slab = pmeindex % ddpme->nslab;
1082 ddpme->pp_min[slab] = std::min(ddpme->pp_min[slab], xyz[dimind]);
1083 ddpme->pp_max[slab] = std::max(ddpme->pp_max[slab], xyz[dimind]);
1087 set_slb_pme_dim_f(dd, ddpme->dim, &ddpme->slb_dim_f);
1090 int dd_pme_maxshift_x(const gmx_domdec_t *dd)
1092 if (dd->comm->ddpme[0].dim == XX)
1094 return dd->comm->ddpme[0].maxshift;
1102 int dd_pme_maxshift_y(const gmx_domdec_t *dd)
1104 if (dd->comm->ddpme[0].dim == YY)
1106 return dd->comm->ddpme[0].maxshift;
1108 else if (dd->comm->npmedecompdim >= 2 && dd->comm->ddpme[1].dim == YY)
1110 return dd->comm->ddpme[1].maxshift;
1118 void dd_cycles_add(const gmx_domdec_t *dd, float cycles, int ddCycl)
1120 /* Note that the cycles value can be incorrect, either 0 or some
1121 * extremely large value, when our thread migrated to another core
1122 * with an unsynchronized cycle counter. If this happens less often
1123 * that once per nstlist steps, this will not cause issues, since
1124 * we later subtract the maximum value from the sum over nstlist steps.
1125 * A zero count will slightly lower the total, but that's a small effect.
1126 * Note that the main purpose of the subtraction of the maximum value
1127 * is to avoid throwing off the load balancing when stalls occur due
1128 * e.g. system activity or network congestion.
1130 dd->comm->cycl[ddCycl] += cycles;
1131 dd->comm->cycl_n[ddCycl]++;
1132 if (cycles > dd->comm->cycl_max[ddCycl])
1134 dd->comm->cycl_max[ddCycl] = cycles;
1139 static void make_load_communicator(gmx_domdec_t *dd, int dim_ind, ivec loc)
1144 gmx_bool bPartOfGroup = FALSE;
1146 dim = dd->dim[dim_ind];
1147 copy_ivec(loc, loc_c);
1148 for (i = 0; i < dd->nc[dim]; i++)
1151 rank = dd_index(dd->nc, loc_c);
1152 if (rank == dd->rank)
1154 /* This process is part of the group */
1155 bPartOfGroup = TRUE;
1158 MPI_Comm_split(dd->mpi_comm_all, bPartOfGroup ? 0 : MPI_UNDEFINED, dd->rank,
1162 dd->comm->mpi_comm_load[dim_ind] = c_row;
1163 if (!isDlbDisabled(dd->comm))
1165 DDCellsizesWithDlb &cellsizes = dd->comm->cellsizesWithDlb[dim_ind];
1167 if (dd->ci[dim] == dd->master_ci[dim])
1169 /* This is the root process of this row */
1170 cellsizes.rowMaster = gmx::compat::make_unique<RowMaster>();
1172 RowMaster &rowMaster = *cellsizes.rowMaster;
1173 rowMaster.cellFrac.resize(ddCellFractionBufferSize(dd, dim_ind));
1174 rowMaster.oldCellFrac.resize(dd->nc[dim] + 1);
1175 rowMaster.isCellMin.resize(dd->nc[dim]);
1178 rowMaster.bounds.resize(dd->nc[dim]);
1180 rowMaster.buf_ncd.resize(dd->nc[dim]);
1184 /* This is not a root process, we only need to receive cell_f */
1185 cellsizes.fracRow.resize(ddCellFractionBufferSize(dd, dim_ind));
1188 if (dd->ci[dim] == dd->master_ci[dim])
1190 snew(dd->comm->load[dim_ind].load, dd->nc[dim]*DD_NLOAD_MAX);
1196 void dd_setup_dlb_resource_sharing(t_commrec *cr,
1200 int physicalnode_id_hash;
1202 MPI_Comm mpi_comm_pp_physicalnode;
1204 if (!thisRankHasDuty(cr, DUTY_PP) || gpu_id < 0)
1206 /* Only ranks with short-ranged tasks (currently) use GPUs.
1207 * If we don't have GPUs assigned, there are no resources to share.
1212 physicalnode_id_hash = gmx_physicalnode_id_hash();
1218 fprintf(debug, "dd_setup_dd_dlb_gpu_sharing:\n");
1219 fprintf(debug, "DD PP rank %d physical node hash %d gpu_id %d\n",
1220 dd->rank, physicalnode_id_hash, gpu_id);
1222 /* Split the PP communicator over the physical nodes */
1223 /* TODO: See if we should store this (before), as it's also used for
1224 * for the nodecomm summation.
1226 // TODO PhysicalNodeCommunicator could be extended/used to handle
1227 // the need for per-node per-group communicators.
1228 MPI_Comm_split(dd->mpi_comm_all, physicalnode_id_hash, dd->rank,
1229 &mpi_comm_pp_physicalnode);
1230 MPI_Comm_split(mpi_comm_pp_physicalnode, gpu_id, dd->rank,
1231 &dd->comm->mpi_comm_gpu_shared);
1232 MPI_Comm_free(&mpi_comm_pp_physicalnode);
1233 MPI_Comm_size(dd->comm->mpi_comm_gpu_shared, &dd->comm->nrank_gpu_shared);
1237 fprintf(debug, "nrank_gpu_shared %d\n", dd->comm->nrank_gpu_shared);
1240 /* Note that some ranks could share a GPU, while others don't */
1242 if (dd->comm->nrank_gpu_shared == 1)
1244 MPI_Comm_free(&dd->comm->mpi_comm_gpu_shared);
1247 GMX_UNUSED_VALUE(cr);
1248 GMX_UNUSED_VALUE(gpu_id);
1252 static void make_load_communicators(gmx_domdec_t gmx_unused *dd)
1255 int dim0, dim1, i, j;
1260 fprintf(debug, "Making load communicators\n");
1263 snew(dd->comm->load, std::max(dd->ndim, 1));
1264 snew(dd->comm->mpi_comm_load, std::max(dd->ndim, 1));
1272 make_load_communicator(dd, 0, loc);
1276 for (i = 0; i < dd->nc[dim0]; i++)
1279 make_load_communicator(dd, 1, loc);
1285 for (i = 0; i < dd->nc[dim0]; i++)
1289 for (j = 0; j < dd->nc[dim1]; j++)
1292 make_load_communicator(dd, 2, loc);
1299 fprintf(debug, "Finished making load communicators\n");
1304 /*! \brief Sets up the relation between neighboring domains and zones */
1305 static void setup_neighbor_relations(gmx_domdec_t *dd)
1307 int d, dim, i, j, m;
1309 gmx_domdec_zones_t *zones;
1310 gmx_domdec_ns_ranges_t *izone;
1312 for (d = 0; d < dd->ndim; d++)
1315 copy_ivec(dd->ci, tmp);
1316 tmp[dim] = (tmp[dim] + 1) % dd->nc[dim];
1317 dd->neighbor[d][0] = ddcoord2ddnodeid(dd, tmp);
1318 copy_ivec(dd->ci, tmp);
1319 tmp[dim] = (tmp[dim] - 1 + dd->nc[dim]) % dd->nc[dim];
1320 dd->neighbor[d][1] = ddcoord2ddnodeid(dd, tmp);
1323 fprintf(debug, "DD rank %d neighbor ranks in dir %d are + %d - %d\n",
1326 dd->neighbor[d][1]);
1330 int nzone = (1 << dd->ndim);
1331 int nizone = (1 << std::max(dd->ndim - 1, 0));
1332 assert(nizone >= 1 && nizone <= DD_MAXIZONE);
1334 zones = &dd->comm->zones;
1336 for (i = 0; i < nzone; i++)
1339 clear_ivec(zones->shift[i]);
1340 for (d = 0; d < dd->ndim; d++)
1342 zones->shift[i][dd->dim[d]] = dd_zo[i][m++];
1347 for (i = 0; i < nzone; i++)
1349 for (d = 0; d < DIM; d++)
1351 s[d] = dd->ci[d] - zones->shift[i][d];
1356 else if (s[d] >= dd->nc[d])
1362 zones->nizone = nizone;
1363 for (i = 0; i < zones->nizone; i++)
1365 assert(ddNonbondedZonePairRanges[i][0] == i);
1367 izone = &zones->izone[i];
1368 /* dd_zp3 is for 3D decomposition, for fewer dimensions use only
1369 * j-zones up to nzone.
1371 izone->j0 = std::min(ddNonbondedZonePairRanges[i][1], nzone);
1372 izone->j1 = std::min(ddNonbondedZonePairRanges[i][2], nzone);
1373 for (dim = 0; dim < DIM; dim++)
1375 if (dd->nc[dim] == 1)
1377 /* All shifts should be allowed */
1378 izone->shift0[dim] = -1;
1379 izone->shift1[dim] = 1;
1383 /* Determine the min/max j-zone shift wrt the i-zone */
1384 izone->shift0[dim] = 1;
1385 izone->shift1[dim] = -1;
1386 for (j = izone->j0; j < izone->j1; j++)
1388 int shift_diff = zones->shift[j][dim] - zones->shift[i][dim];
1389 if (shift_diff < izone->shift0[dim])
1391 izone->shift0[dim] = shift_diff;
1393 if (shift_diff > izone->shift1[dim])
1395 izone->shift1[dim] = shift_diff;
1402 if (!isDlbDisabled(dd->comm))
1404 dd->comm->cellsizesWithDlb.resize(dd->ndim);
1407 if (dd->comm->bRecordLoad)
1409 make_load_communicators(dd);
1413 static void make_pp_communicator(const gmx::MDLogger &mdlog,
1415 t_commrec gmx_unused *cr,
1416 bool gmx_unused reorder)
1419 gmx_domdec_comm_t *comm;
1426 if (comm->bCartesianPP)
1428 /* Set up cartesian communication for the particle-particle part */
1429 GMX_LOG(mdlog.info).appendTextFormatted(
1430 "Will use a Cartesian communicator: %d x %d x %d",
1431 dd->nc[XX], dd->nc[YY], dd->nc[ZZ]);
1433 for (int i = 0; i < DIM; i++)
1437 MPI_Cart_create(cr->mpi_comm_mygroup, DIM, dd->nc, periods, static_cast<int>(reorder),
1439 /* We overwrite the old communicator with the new cartesian one */
1440 cr->mpi_comm_mygroup = comm_cart;
1443 dd->mpi_comm_all = cr->mpi_comm_mygroup;
1444 MPI_Comm_rank(dd->mpi_comm_all, &dd->rank);
1446 if (comm->bCartesianPP_PME)
1448 /* Since we want to use the original cartesian setup for sim,
1449 * and not the one after split, we need to make an index.
1451 snew(comm->ddindex2ddnodeid, dd->nnodes);
1452 comm->ddindex2ddnodeid[dd_index(dd->nc, dd->ci)] = dd->rank;
1453 gmx_sumi(dd->nnodes, comm->ddindex2ddnodeid, cr);
1454 /* Get the rank of the DD master,
1455 * above we made sure that the master node is a PP node.
1465 MPI_Allreduce(&rank, &dd->masterrank, 1, MPI_INT, MPI_SUM, dd->mpi_comm_all);
1467 else if (comm->bCartesianPP)
1469 if (cr->npmenodes == 0)
1471 /* The PP communicator is also
1472 * the communicator for this simulation
1474 cr->mpi_comm_mysim = cr->mpi_comm_mygroup;
1476 cr->nodeid = dd->rank;
1478 MPI_Cart_coords(dd->mpi_comm_all, dd->rank, DIM, dd->ci);
1480 /* We need to make an index to go from the coordinates
1481 * to the nodeid of this simulation.
1483 snew(comm->ddindex2simnodeid, dd->nnodes);
1484 snew(buf, dd->nnodes);
1485 if (thisRankHasDuty(cr, DUTY_PP))
1487 buf[dd_index(dd->nc, dd->ci)] = cr->sim_nodeid;
1489 /* Communicate the ddindex to simulation nodeid index */
1490 MPI_Allreduce(buf, comm->ddindex2simnodeid, dd->nnodes, MPI_INT, MPI_SUM,
1491 cr->mpi_comm_mysim);
1494 /* Determine the master coordinates and rank.
1495 * The DD master should be the same node as the master of this sim.
1497 for (int i = 0; i < dd->nnodes; i++)
1499 if (comm->ddindex2simnodeid[i] == 0)
1501 ddindex2xyz(dd->nc, i, dd->master_ci);
1502 MPI_Cart_rank(dd->mpi_comm_all, dd->master_ci, &dd->masterrank);
1507 fprintf(debug, "The master rank is %d\n", dd->masterrank);
1512 /* No Cartesian communicators */
1513 /* We use the rank in dd->comm->all as DD index */
1514 ddindex2xyz(dd->nc, dd->rank, dd->ci);
1515 /* The simulation master nodeid is 0, so the DD master rank is also 0 */
1517 clear_ivec(dd->master_ci);
1521 GMX_LOG(mdlog.info).appendTextFormatted(
1522 "Domain decomposition rank %d, coordinates %d %d %d\n",
1523 dd->rank, dd->ci[XX], dd->ci[YY], dd->ci[ZZ]);
1527 "Domain decomposition rank %d, coordinates %d %d %d\n\n",
1528 dd->rank, dd->ci[XX], dd->ci[YY], dd->ci[ZZ]);
1532 static void receive_ddindex2simnodeid(gmx_domdec_t *dd,
1536 gmx_domdec_comm_t *comm = dd->comm;
1538 if (!comm->bCartesianPP_PME && comm->bCartesianPP)
1541 snew(comm->ddindex2simnodeid, dd->nnodes);
1542 snew(buf, dd->nnodes);
1543 if (thisRankHasDuty(cr, DUTY_PP))
1545 buf[dd_index(dd->nc, dd->ci)] = cr->sim_nodeid;
1547 /* Communicate the ddindex to simulation nodeid index */
1548 MPI_Allreduce(buf, comm->ddindex2simnodeid, dd->nnodes, MPI_INT, MPI_SUM,
1549 cr->mpi_comm_mysim);
1553 GMX_UNUSED_VALUE(dd);
1554 GMX_UNUSED_VALUE(cr);
1558 static void split_communicator(const gmx::MDLogger &mdlog,
1559 t_commrec *cr, gmx_domdec_t *dd,
1560 DdRankOrder gmx_unused rankOrder,
1561 bool gmx_unused reorder)
1563 gmx_domdec_comm_t *comm;
1572 if (comm->bCartesianPP)
1574 for (i = 1; i < DIM; i++)
1576 bDiv[i] = ((cr->npmenodes*dd->nc[i]) % (dd->nnodes) == 0);
1578 if (bDiv[YY] || bDiv[ZZ])
1580 comm->bCartesianPP_PME = TRUE;
1581 /* If we have 2D PME decomposition, which is always in x+y,
1582 * we stack the PME only nodes in z.
1583 * Otherwise we choose the direction that provides the thinnest slab
1584 * of PME only nodes as this will have the least effect
1585 * on the PP communication.
1586 * But for the PME communication the opposite might be better.
1588 if (bDiv[ZZ] && (comm->npmenodes_y > 1 ||
1590 dd->nc[YY] > dd->nc[ZZ]))
1592 comm->cartpmedim = ZZ;
1596 comm->cartpmedim = YY;
1598 comm->ntot[comm->cartpmedim]
1599 += (cr->npmenodes*dd->nc[comm->cartpmedim])/dd->nnodes;
1603 GMX_LOG(mdlog.info).appendTextFormatted(
1604 "Number of PME-only ranks (%d) is not a multiple of nx*ny (%d*%d) or nx*nz (%d*%d)",
1605 cr->npmenodes, dd->nc[XX], dd->nc[YY], dd->nc[XX], dd->nc[ZZ]);
1606 GMX_LOG(mdlog.info).appendText("Will not use a Cartesian communicator for PP <-> PME\n");
1610 if (comm->bCartesianPP_PME)
1616 GMX_LOG(mdlog.info).appendTextFormatted(
1617 "Will use a Cartesian communicator for PP <-> PME: %d x %d x %d",
1618 comm->ntot[XX], comm->ntot[YY], comm->ntot[ZZ]);
1620 for (i = 0; i < DIM; i++)
1624 MPI_Cart_create(cr->mpi_comm_mysim, DIM, comm->ntot, periods, static_cast<int>(reorder),
1626 MPI_Comm_rank(comm_cart, &rank);
1627 if (MASTER(cr) && rank != 0)
1629 gmx_fatal(FARGS, "MPI rank 0 was renumbered by MPI_Cart_create, we do not allow this");
1632 /* With this assigment we loose the link to the original communicator
1633 * which will usually be MPI_COMM_WORLD, unless have multisim.
1635 cr->mpi_comm_mysim = comm_cart;
1636 cr->sim_nodeid = rank;
1638 MPI_Cart_coords(cr->mpi_comm_mysim, cr->sim_nodeid, DIM, dd->ci);
1640 GMX_LOG(mdlog.info).appendTextFormatted(
1641 "Cartesian rank %d, coordinates %d %d %d\n",
1642 cr->sim_nodeid, dd->ci[XX], dd->ci[YY], dd->ci[ZZ]);
1644 if (dd->ci[comm->cartpmedim] < dd->nc[comm->cartpmedim])
1648 if (cr->npmenodes == 0 ||
1649 dd->ci[comm->cartpmedim] >= dd->nc[comm->cartpmedim])
1651 cr->duty = DUTY_PME;
1654 /* Split the sim communicator into PP and PME only nodes */
1655 MPI_Comm_split(cr->mpi_comm_mysim,
1656 getThisRankDuties(cr),
1657 dd_index(comm->ntot, dd->ci),
1658 &cr->mpi_comm_mygroup);
1665 case DdRankOrder::pp_pme:
1666 GMX_LOG(mdlog.info).appendText("Order of the ranks: PP first, PME last");
1668 case DdRankOrder::interleave:
1669 /* Interleave the PP-only and PME-only ranks */
1670 GMX_LOG(mdlog.info).appendText("Interleaving PP and PME ranks");
1671 comm->pmenodes = dd_interleaved_pme_ranks(dd);
1673 case DdRankOrder::cartesian:
1676 gmx_fatal(FARGS, "Invalid ddRankOrder=%d", static_cast<int>(rankOrder));
1679 if (dd_simnode2pmenode(dd, cr, cr->sim_nodeid) == -1)
1681 cr->duty = DUTY_PME;
1688 /* Split the sim communicator into PP and PME only nodes */
1689 MPI_Comm_split(cr->mpi_comm_mysim,
1690 getThisRankDuties(cr),
1692 &cr->mpi_comm_mygroup);
1693 MPI_Comm_rank(cr->mpi_comm_mygroup, &cr->nodeid);
1697 GMX_LOG(mdlog.info).appendTextFormatted(
1698 "This rank does only %s work.\n",
1699 thisRankHasDuty(cr, DUTY_PP) ? "particle-particle" : "PME-mesh");
1702 /*! \brief Generates the MPI communicators for domain decomposition */
1703 static void make_dd_communicators(const gmx::MDLogger &mdlog,
1705 gmx_domdec_t *dd, DdRankOrder ddRankOrder)
1707 gmx_domdec_comm_t *comm;
1712 copy_ivec(dd->nc, comm->ntot);
1714 comm->bCartesianPP = (ddRankOrder == DdRankOrder::cartesian);
1715 comm->bCartesianPP_PME = FALSE;
1717 /* Reorder the nodes by default. This might change the MPI ranks.
1718 * Real reordering is only supported on very few architectures,
1719 * Blue Gene is one of them.
1721 CartReorder = getenv("GMX_NO_CART_REORDER") == nullptr;
1723 if (cr->npmenodes > 0)
1725 /* Split the communicator into a PP and PME part */
1726 split_communicator(mdlog, cr, dd, ddRankOrder, CartReorder);
1727 if (comm->bCartesianPP_PME)
1729 /* We (possibly) reordered the nodes in split_communicator,
1730 * so it is no longer required in make_pp_communicator.
1732 CartReorder = FALSE;
1737 /* All nodes do PP and PME */
1739 /* We do not require separate communicators */
1740 cr->mpi_comm_mygroup = cr->mpi_comm_mysim;
1744 if (thisRankHasDuty(cr, DUTY_PP))
1746 /* Copy or make a new PP communicator */
1747 make_pp_communicator(mdlog, dd, cr, CartReorder);
1751 receive_ddindex2simnodeid(dd, cr);
1754 if (!thisRankHasDuty(cr, DUTY_PME))
1756 /* Set up the commnuication to our PME node */
1757 dd->pme_nodeid = dd_simnode2pmenode(dd, cr, cr->sim_nodeid);
1758 dd->pme_receive_vir_ener = receive_vir_ener(dd, cr);
1761 fprintf(debug, "My pme_nodeid %d receive ener %s\n",
1762 dd->pme_nodeid, gmx::boolToString(dd->pme_receive_vir_ener));
1767 dd->pme_nodeid = -1;
1772 dd->ma = gmx::compat::make_unique<AtomDistribution>(dd->nc,
1774 comm->cgs_gl.index[comm->cgs_gl.nr]);
1778 static real *get_slb_frac(const gmx::MDLogger &mdlog,
1779 const char *dir, int nc, const char *size_string)
1781 real *slb_frac, tot;
1786 if (nc > 1 && size_string != nullptr)
1788 GMX_LOG(mdlog.info).appendTextFormatted(
1789 "Using static load balancing for the %s direction", dir);
1792 for (i = 0; i < nc; i++)
1795 sscanf(size_string, "%20lf%n", &dbl, &n);
1798 gmx_fatal(FARGS, "Incorrect or not enough DD cell size entries for direction %s: '%s'", dir, size_string);
1805 std::string relativeCellSizes = "Relative cell sizes:";
1806 for (i = 0; i < nc; i++)
1809 relativeCellSizes += gmx::formatString(" %5.3f", slb_frac[i]);
1811 GMX_LOG(mdlog.info).appendText(relativeCellSizes);
1817 static int multi_body_bondeds_count(const gmx_mtop_t *mtop)
1820 gmx_mtop_ilistloop_t iloop = gmx_mtop_ilistloop_init(mtop);
1822 while (const InteractionLists *ilists = gmx_mtop_ilistloop_next(iloop, &nmol))
1824 for (auto &ilist : extractILists(*ilists, IF_BOND))
1826 if (NRAL(ilist.functionType) > 2)
1828 n += nmol*(ilist.iatoms.size()/ilistStride(ilist));
1836 static int dd_getenv(const gmx::MDLogger &mdlog,
1837 const char *env_var, int def)
1843 val = getenv(env_var);
1846 if (sscanf(val, "%20d", &nst) <= 0)
1850 GMX_LOG(mdlog.info).appendTextFormatted(
1851 "Found env.var. %s = %s, using value %d",
1858 static void check_dd_restrictions(const gmx_domdec_t *dd,
1859 const t_inputrec *ir,
1860 const gmx::MDLogger &mdlog)
1862 if (ir->ePBC == epbcSCREW &&
1863 (dd->nc[XX] == 1 || dd->nc[YY] > 1 || dd->nc[ZZ] > 1))
1865 gmx_fatal(FARGS, "With pbc=%s can only do domain decomposition in the x-direction", epbc_names[ir->ePBC]);
1868 if (ir->ns_type == ensSIMPLE)
1870 gmx_fatal(FARGS, "Domain decomposition does not support simple neighbor searching, use grid searching or run with one MPI rank");
1873 if (ir->nstlist == 0)
1875 gmx_fatal(FARGS, "Domain decomposition does not work with nstlist=0");
1878 if (ir->comm_mode == ecmANGULAR && ir->ePBC != epbcNONE)
1880 GMX_LOG(mdlog.warning).appendText("comm-mode angular will give incorrect results when the comm group partially crosses a periodic boundary");
1884 static real average_cellsize_min(gmx_domdec_t *dd, gmx_ddbox_t *ddbox)
1889 r = ddbox->box_size[XX];
1890 for (di = 0; di < dd->ndim; di++)
1893 /* Check using the initial average cell size */
1894 r = std::min(r, ddbox->box_size[d]*ddbox->skew_fac[d]/dd->nc[d]);
1900 /*! \brief Depending on the DLB initial value return the DLB switched off state or issue an error.
1902 static DlbState forceDlbOffOrBail(DlbState cmdlineDlbState,
1903 const std::string &reasonStr,
1904 const gmx::MDLogger &mdlog)
1906 std::string dlbNotSupportedErr = "Dynamic load balancing requested, but ";
1907 std::string dlbDisableNote = "NOTE: disabling dynamic load balancing as ";
1909 if (cmdlineDlbState == DlbState::onUser)
1911 gmx_fatal(FARGS, "%s", (dlbNotSupportedErr + reasonStr).c_str());
1913 else if (cmdlineDlbState == DlbState::offCanTurnOn)
1915 GMX_LOG(mdlog.info).appendText(dlbDisableNote + reasonStr);
1917 return DlbState::offForever;
1920 /*! \brief Return the dynamic load balancer's initial state based on initial conditions and user inputs.
1922 * This function parses the parameters of "-dlb" command line option setting
1923 * corresponding state values. Then it checks the consistency of the determined
1924 * state with other run parameters and settings. As a result, the initial state
1925 * may be altered or an error may be thrown if incompatibility of options is detected.
1927 * \param [in] mdlog Logger.
1928 * \param [in] dlbOption Enum value for the DLB option.
1929 * \param [in] bRecordLoad True if the load balancer is recording load information.
1930 * \param [in] mdrunOptions Options for mdrun.
1931 * \param [in] ir Pointer mdrun to input parameters.
1932 * \returns DLB initial/startup state.
1934 static DlbState determineInitialDlbState(const gmx::MDLogger &mdlog,
1935 DlbOption dlbOption, gmx_bool bRecordLoad,
1936 const MdrunOptions &mdrunOptions,
1937 const t_inputrec *ir)
1939 DlbState dlbState = DlbState::offCanTurnOn;
1943 case DlbOption::turnOnWhenUseful: dlbState = DlbState::offCanTurnOn; break;
1944 case DlbOption::no: dlbState = DlbState::offUser; break;
1945 case DlbOption::yes: dlbState = DlbState::onUser; break;
1946 default: gmx_incons("Invalid dlbOption enum value");
1949 /* Reruns don't support DLB: bail or override auto mode */
1950 if (mdrunOptions.rerun)
1952 std::string reasonStr = "it is not supported in reruns.";
1953 return forceDlbOffOrBail(dlbState, reasonStr, mdlog);
1956 /* Unsupported integrators */
1957 if (!EI_DYNAMICS(ir->eI))
1959 auto reasonStr = gmx::formatString("it is only supported with dynamics, not with integrator '%s'.", EI(ir->eI));
1960 return forceDlbOffOrBail(dlbState, reasonStr, mdlog);
1963 /* Without cycle counters we can't time work to balance on */
1966 std::string reasonStr = "cycle counters unsupported or not enabled in the operating system kernel.";
1967 return forceDlbOffOrBail(dlbState, reasonStr, mdlog);
1970 if (mdrunOptions.reproducible)
1972 std::string reasonStr = "you started a reproducible run.";
1975 case DlbState::offUser:
1977 case DlbState::offForever:
1978 GMX_RELEASE_ASSERT(false, "DlbState::offForever is not a valid initial state");
1980 case DlbState::offCanTurnOn:
1981 return forceDlbOffOrBail(dlbState, reasonStr, mdlog);
1982 case DlbState::onCanTurnOff:
1983 GMX_RELEASE_ASSERT(false, "DlbState::offCanTurnOff is not a valid initial state");
1985 case DlbState::onUser:
1986 return forceDlbOffOrBail(dlbState, reasonStr + " In load balanced runs binary reproducibility cannot be ensured.", mdlog);
1988 gmx_fatal(FARGS, "Death horror: undefined case (%d) for load balancing choice", static_cast<int>(dlbState));
1995 static void set_dd_dim(const gmx::MDLogger &mdlog, gmx_domdec_t *dd)
1998 if (getenv("GMX_DD_ORDER_ZYX") != nullptr)
2000 /* Decomposition order z,y,x */
2001 GMX_LOG(mdlog.info).appendText("Using domain decomposition order z, y, x");
2002 for (int dim = DIM-1; dim >= 0; dim--)
2004 if (dd->nc[dim] > 1)
2006 dd->dim[dd->ndim++] = dim;
2012 /* Decomposition order x,y,z */
2013 for (int dim = 0; dim < DIM; dim++)
2015 if (dd->nc[dim] > 1)
2017 dd->dim[dd->ndim++] = dim;
2024 /* Set dim[0] to avoid extra checks on ndim in several places */
2029 static gmx_domdec_comm_t *init_dd_comm()
2031 gmx_domdec_comm_t *comm = new gmx_domdec_comm_t;
2033 comm->n_load_have = 0;
2034 comm->n_load_collect = 0;
2036 comm->haveTurnedOffDlb = false;
2038 for (int i = 0; i < static_cast<int>(DDAtomRanges::Type::Number); i++)
2040 comm->sum_nat[i] = 0;
2044 comm->load_step = 0;
2047 clear_ivec(comm->load_lim);
2051 /* This should be replaced by a unique pointer */
2052 comm->balanceRegion = ddBalanceRegionAllocate();
2057 /* Returns whether mtop contains constraints and/or vsites */
2058 static bool systemHasConstraintsOrVsites(const gmx_mtop_t &mtop)
2060 auto ilistLoop = gmx_mtop_ilistloop_init(mtop);
2062 while (const InteractionLists *ilists = gmx_mtop_ilistloop_next(ilistLoop, &nmol))
2064 if (!extractILists(*ilists, IF_CONSTRAINT | IF_VSITE).empty())
2073 static void setupUpdateGroups(const gmx::MDLogger &mdlog,
2074 const gmx_mtop_t &mtop,
2075 const t_inputrec &inputrec,
2077 int numMpiRanksTotal,
2078 gmx_domdec_comm_t *comm)
2080 /* When we have constraints and/or vsites, it is beneficial to use
2081 * update groups (when possible) to allow independent update of groups.
2083 if (!systemHasConstraintsOrVsites(mtop))
2085 /* No constraints or vsites, atoms can be updated independently */
2089 comm->updateGroupingPerMoleculetype = gmx::makeUpdateGroups(mtop);
2090 comm->useUpdateGroups = !comm->updateGroupingPerMoleculetype.empty();
2092 if (comm->useUpdateGroups)
2094 int numUpdateGroups = 0;
2095 for (const auto &molblock : mtop.molblock)
2097 numUpdateGroups += molblock.nmol*comm->updateGroupingPerMoleculetype[molblock.type].numBlocks();
2100 /* Note: We would like to use dd->nnodes for the atom count estimate,
2101 * but that is not yet available here. But this anyhow only
2102 * affect performance up to the second dd_partition_system call.
2104 int homeAtomCountEstimate = mtop.natoms/numMpiRanksTotal;
2105 comm->updateGroupsCog =
2106 gmx::compat::make_unique<gmx::UpdateGroupsCog>(mtop,
2107 comm->updateGroupingPerMoleculetype,
2108 maxReferenceTemperature(inputrec),
2109 homeAtomCountEstimate);
2111 /* To use update groups, the large domain-to-domain cutoff distance
2112 * should be compatible with the box size.
2114 comm->useUpdateGroups = (atomToAtomIntoDomainToDomainCutoff(*comm, 0) < cutoffMargin);
2115 /* TODO: Enable update groups when all infrastructure is present */
2116 comm->useUpdateGroups = false;
2118 if (comm->useUpdateGroups)
2120 GMX_LOG(mdlog.info).appendTextFormatted(
2121 "Using update groups, nr %d, average size %.1f atoms, max. radius %.3f nm\n",
2123 mtop.natoms/static_cast<double>(numUpdateGroups),
2124 comm->updateGroupsCog->maxUpdateGroupRadius());
2128 /* TODO: Remove this comment when enabling update groups
2129 * GMX_LOG(mdlog.info).appendTextFormatted("The combination of rlist and box size prohibits the use of update groups\n");
2131 comm->updateGroupingPerMoleculetype.clear();
2132 comm->updateGroupsCog.reset(nullptr);
2137 /*! \brief Set the cell size and interaction limits, as well as the DD grid */
2138 static void set_dd_limits_and_grid(const gmx::MDLogger &mdlog,
2139 t_commrec *cr, gmx_domdec_t *dd,
2140 const DomdecOptions &options,
2141 const MdrunOptions &mdrunOptions,
2142 const gmx_mtop_t *mtop,
2143 const t_inputrec *ir,
2145 gmx::ArrayRef<const gmx::RVec> xGlobal,
2149 real r_bonded_limit = -1;
2150 const real tenPercentMargin = 1.1;
2151 gmx_domdec_comm_t *comm = dd->comm;
2153 dd->npbcdim = ePBC2npbcdim(ir->ePBC);
2154 dd->numBoundedDimensions = inputrec2nboundeddim(ir);
2155 dd->haveDynamicBox = inputrecDynamicBox(ir);
2156 dd->bScrewPBC = (ir->ePBC == epbcSCREW);
2158 dd->pme_recv_f_alloc = 0;
2159 dd->pme_recv_f_buf = nullptr;
2161 /* Initialize to GPU share count to 0, might change later */
2162 comm->nrank_gpu_shared = 0;
2164 comm->dlbState = determineInitialDlbState(mdlog, options.dlbOption, comm->bRecordLoad, mdrunOptions, ir);
2165 dd_dlb_set_should_check_whether_to_turn_dlb_on(dd, TRUE);
2166 /* To consider turning DLB on after 2*nstlist steps we need to check
2167 * at partitioning count 3. Thus we need to increase the first count by 2.
2169 comm->ddPartioningCountFirstDlbOff += 2;
2171 GMX_LOG(mdlog.info).appendTextFormatted(
2172 "Dynamic load balancing: %s", edlbs_names[int(comm->dlbState)]);
2174 comm->bPMELoadBalDLBLimits = FALSE;
2176 /* Allocate the charge group/atom sorting struct */
2177 comm->sort = gmx::compat::make_unique<gmx_domdec_sort_t>();
2179 comm->bCGs = (ncg_mtop(mtop) < mtop->natoms);
2181 /* We need to decide on update groups early, as this affects communication distances */
2182 comm->useUpdateGroups = false;
2183 if (ir->cutoff_scheme == ecutsVERLET)
2185 real cutoffMargin = std::sqrt(max_cutoff2(ir->ePBC, box)) - ir->rlist;
2186 setupUpdateGroups(mdlog, *mtop, *ir, cutoffMargin, cr->nnodes, comm);
2189 comm->bInterCGBondeds = ((ncg_mtop(mtop) > gmx_mtop_num_molecules(*mtop)) ||
2190 mtop->bIntermolecularInteractions);
2191 if (comm->bInterCGBondeds)
2193 comm->bInterCGMultiBody = (multi_body_bondeds_count(mtop) > 0);
2197 comm->bInterCGMultiBody = FALSE;
2200 if (comm->useUpdateGroups)
2202 dd->splitConstraints = false;
2203 dd->splitSettles = false;
2207 dd->splitConstraints = gmx::inter_charge_group_constraints(*mtop);
2208 dd->splitSettles = gmx::inter_charge_group_settles(*mtop);
2213 /* Set the cut-off to some very large value,
2214 * so we don't need if statements everywhere in the code.
2215 * We use sqrt, since the cut-off is squared in some places.
2217 comm->cutoff = GMX_CUTOFF_INF;
2221 comm->cutoff = atomToAtomIntoDomainToDomainCutoff(*comm, ir->rlist);
2223 comm->cutoff_mbody = 0;
2225 /* Determine the minimum cell size limit, affected by many factors */
2226 comm->cellsize_limit = 0;
2227 comm->bBondComm = FALSE;
2229 /* We do not allow home atoms to move beyond the neighboring domain
2230 * between domain decomposition steps, which limits the cell size.
2231 * Get an estimate of cell size limit due to atom displacement.
2232 * In most cases this is a large overestimate, because it assumes
2233 * non-interaction atoms.
2234 * We set the chance to 1 in a trillion steps.
2236 constexpr real c_chanceThatAtomMovesBeyondDomain = 1e-12;
2237 const real limitForAtomDisplacement =
2238 minCellSizeForAtomDisplacement(*mtop, *ir,
2239 comm->updateGroupingPerMoleculetype,
2240 c_chanceThatAtomMovesBeyondDomain);
2241 GMX_LOG(mdlog.info).appendTextFormatted(
2242 "Minimum cell size due to atom displacement: %.3f nm",
2243 limitForAtomDisplacement);
2245 comm->cellsize_limit = std::max(comm->cellsize_limit,
2246 limitForAtomDisplacement);
2248 /* TODO: PME decomposition currently requires atoms not to be more than
2249 * 2/3 of comm->cutoff, which is >=rlist, outside of their domain.
2250 * In nearly all cases, limitForAtomDisplacement will be smaller
2251 * than 2/3*rlist, so the PME requirement is satisfied.
2252 * But it would be better for both correctness and performance
2253 * to use limitForAtomDisplacement instead of 2/3*comm->cutoff.
2254 * Note that we would need to improve the pairlist buffer case.
2257 if (comm->bInterCGBondeds)
2259 if (options.minimumCommunicationRange > 0)
2261 comm->cutoff_mbody = atomToAtomIntoDomainToDomainCutoff(*comm, options.minimumCommunicationRange);
2262 if (options.useBondedCommunication)
2264 comm->bBondComm = (comm->cutoff_mbody > comm->cutoff);
2268 comm->cutoff = std::max(comm->cutoff, comm->cutoff_mbody);
2270 r_bonded_limit = comm->cutoff_mbody;
2272 else if (ir->bPeriodicMols)
2274 /* Can not easily determine the required cut-off */
2275 GMX_LOG(mdlog.warning).appendText("NOTE: Periodic molecules are present in this system. Because of this, the domain decomposition algorithm cannot easily determine the minimum cell size that it requires for treating bonded interactions. Instead, domain decomposition will assume that half the non-bonded cut-off will be a suitable lower bound.");
2276 comm->cutoff_mbody = comm->cutoff/2;
2277 r_bonded_limit = comm->cutoff_mbody;
2285 dd_bonded_cg_distance(mdlog, mtop, ir, as_rvec_array(xGlobal.data()), box,
2286 options.checkBondedInteractions,
2289 gmx_bcast(sizeof(r_2b), &r_2b, cr);
2290 gmx_bcast(sizeof(r_mb), &r_mb, cr);
2292 /* We use an initial margin of 10% for the minimum cell size,
2293 * except when we are just below the non-bonded cut-off.
2295 if (options.useBondedCommunication)
2297 if (std::max(r_2b, r_mb) > comm->cutoff)
2299 r_bonded = std::max(r_2b, r_mb);
2300 r_bonded_limit = tenPercentMargin*r_bonded;
2301 comm->bBondComm = TRUE;
2306 r_bonded_limit = std::min(tenPercentMargin*r_bonded, comm->cutoff);
2308 /* We determine cutoff_mbody later */
2312 /* No special bonded communication,
2313 * simply increase the DD cut-off.
2315 r_bonded_limit = tenPercentMargin*std::max(r_2b, r_mb);
2316 comm->cutoff_mbody = r_bonded_limit;
2317 comm->cutoff = std::max(comm->cutoff, comm->cutoff_mbody);
2320 GMX_LOG(mdlog.info).appendTextFormatted(
2321 "Minimum cell size due to bonded interactions: %.3f nm",
2324 comm->cellsize_limit = std::max(comm->cellsize_limit, r_bonded_limit);
2328 if (dd->splitConstraints && options.constraintCommunicationRange <= 0)
2330 /* There is a cell size limit due to the constraints (P-LINCS) */
2331 rconstr = gmx::constr_r_max(mdlog, mtop, ir);
2332 GMX_LOG(mdlog.info).appendTextFormatted(
2333 "Estimated maximum distance required for P-LINCS: %.3f nm",
2335 if (rconstr > comm->cellsize_limit)
2337 GMX_LOG(mdlog.info).appendText("This distance will limit the DD cell size, you can override this with -rcon");
2340 else if (options.constraintCommunicationRange > 0)
2342 /* Here we do not check for dd->splitConstraints.
2343 * because one can also set a cell size limit for virtual sites only
2344 * and at this point we don't know yet if there are intercg v-sites.
2346 GMX_LOG(mdlog.info).appendTextFormatted(
2347 "User supplied maximum distance required for P-LINCS: %.3f nm",
2348 options.constraintCommunicationRange);
2349 rconstr = options.constraintCommunicationRange;
2351 comm->cellsize_limit = std::max(comm->cellsize_limit, rconstr);
2353 comm->cgs_gl = gmx_mtop_global_cgs(mtop);
2355 if (options.numCells[XX] > 0)
2357 copy_ivec(options.numCells, dd->nc);
2358 set_dd_dim(mdlog, dd);
2359 set_ddbox_cr(*cr, &dd->nc, *ir, box, xGlobal, ddbox);
2361 if (options.numPmeRanks >= 0)
2363 cr->npmenodes = options.numPmeRanks;
2367 /* When the DD grid is set explicitly and -npme is set to auto,
2368 * don't use PME ranks. We check later if the DD grid is
2369 * compatible with the total number of ranks.
2374 real acs = average_cellsize_min(dd, ddbox);
2375 if (acs < comm->cellsize_limit)
2377 gmx_fatal_collective(FARGS, cr->mpi_comm_mysim, MASTER(cr),
2378 "The initial cell size (%f) is smaller than the cell size limit (%f), change options -dd, -rdd or -rcon, see the log file for details",
2379 acs, comm->cellsize_limit);
2384 set_ddbox_cr(*cr, nullptr, *ir, box, xGlobal, ddbox);
2386 /* We need to choose the optimal DD grid and possibly PME nodes */
2388 dd_choose_grid(mdlog, cr, dd, ir, mtop, box, ddbox,
2389 options.numPmeRanks,
2390 !isDlbDisabled(comm),
2392 comm->cellsize_limit, comm->cutoff,
2393 comm->bInterCGBondeds);
2395 if (dd->nc[XX] == 0)
2398 gmx_bool bC = (dd->splitConstraints && rconstr > r_bonded_limit);
2399 sprintf(buf, "Change the number of ranks or mdrun option %s%s%s",
2400 !bC ? "-rdd" : "-rcon",
2401 comm->dlbState != DlbState::offUser ? " or -dds" : "",
2402 bC ? " or your LINCS settings" : "");
2404 gmx_fatal_collective(FARGS, cr->mpi_comm_mysim, MASTER(cr),
2405 "There is no domain decomposition for %d ranks that is compatible with the given box and a minimum cell size of %g nm\n"
2407 "Look in the log file for details on the domain decomposition",
2408 cr->nnodes-cr->npmenodes, limit, buf);
2410 set_dd_dim(mdlog, dd);
2413 GMX_LOG(mdlog.info).appendTextFormatted(
2414 "Domain decomposition grid %d x %d x %d, separate PME ranks %d",
2415 dd->nc[XX], dd->nc[YY], dd->nc[ZZ], cr->npmenodes);
2417 dd->nnodes = dd->nc[XX]*dd->nc[YY]*dd->nc[ZZ];
2418 if (cr->nnodes - dd->nnodes != cr->npmenodes)
2420 gmx_fatal_collective(FARGS, cr->mpi_comm_mysim, MASTER(cr),
2421 "The size of the domain decomposition grid (%d) does not match the number of ranks (%d). The total number of ranks is %d",
2422 dd->nnodes, cr->nnodes - cr->npmenodes, cr->nnodes);
2424 if (cr->npmenodes > dd->nnodes)
2426 gmx_fatal_collective(FARGS, cr->mpi_comm_mysim, MASTER(cr),
2427 "The number of separate PME ranks (%d) is larger than the number of PP ranks (%d), this is not supported.", cr->npmenodes, dd->nnodes);
2429 if (cr->npmenodes > 0)
2431 comm->npmenodes = cr->npmenodes;
2435 comm->npmenodes = dd->nnodes;
2438 if (EEL_PME(ir->coulombtype) || EVDW_PME(ir->vdwtype))
2440 /* The following choices should match those
2441 * in comm_cost_est in domdec_setup.c.
2442 * Note that here the checks have to take into account
2443 * that the decomposition might occur in a different order than xyz
2444 * (for instance through the env.var. GMX_DD_ORDER_ZYX),
2445 * in which case they will not match those in comm_cost_est,
2446 * but since that is mainly for testing purposes that's fine.
2448 if (dd->ndim >= 2 && dd->dim[0] == XX && dd->dim[1] == YY &&
2449 comm->npmenodes > dd->nc[XX] && comm->npmenodes % dd->nc[XX] == 0 &&
2450 getenv("GMX_PMEONEDD") == nullptr)
2452 comm->npmedecompdim = 2;
2453 comm->npmenodes_x = dd->nc[XX];
2454 comm->npmenodes_y = comm->npmenodes/comm->npmenodes_x;
2458 /* In case nc is 1 in both x and y we could still choose to
2459 * decompose pme in y instead of x, but we use x for simplicity.
2461 comm->npmedecompdim = 1;
2462 if (dd->dim[0] == YY)
2464 comm->npmenodes_x = 1;
2465 comm->npmenodes_y = comm->npmenodes;
2469 comm->npmenodes_x = comm->npmenodes;
2470 comm->npmenodes_y = 1;
2473 GMX_LOG(mdlog.info).appendTextFormatted(
2474 "PME domain decomposition: %d x %d x %d",
2475 comm->npmenodes_x, comm->npmenodes_y, 1);
2479 comm->npmedecompdim = 0;
2480 comm->npmenodes_x = 0;
2481 comm->npmenodes_y = 0;
2484 snew(comm->slb_frac, DIM);
2485 if (isDlbDisabled(comm))
2487 comm->slb_frac[XX] = get_slb_frac(mdlog, "x", dd->nc[XX], options.cellSizeX);
2488 comm->slb_frac[YY] = get_slb_frac(mdlog, "y", dd->nc[YY], options.cellSizeY);
2489 comm->slb_frac[ZZ] = get_slb_frac(mdlog, "z", dd->nc[ZZ], options.cellSizeZ);
2492 if (comm->bInterCGBondeds && comm->cutoff_mbody == 0)
2494 if (comm->bBondComm || !isDlbDisabled(comm))
2496 /* Set the bonded communication distance to halfway
2497 * the minimum and the maximum,
2498 * since the extra communication cost is nearly zero.
2500 real acs = average_cellsize_min(dd, ddbox);
2501 comm->cutoff_mbody = 0.5*(r_bonded + acs);
2502 if (!isDlbDisabled(comm))
2504 /* Check if this does not limit the scaling */
2505 comm->cutoff_mbody = std::min(comm->cutoff_mbody,
2506 options.dlbScaling*acs);
2508 if (!comm->bBondComm)
2510 /* Without bBondComm do not go beyond the n.b. cut-off */
2511 comm->cutoff_mbody = std::min(comm->cutoff_mbody, comm->cutoff);
2512 if (comm->cellsize_limit >= comm->cutoff)
2514 /* We don't loose a lot of efficieny
2515 * when increasing it to the n.b. cut-off.
2516 * It can even be slightly faster, because we need
2517 * less checks for the communication setup.
2519 comm->cutoff_mbody = comm->cutoff;
2522 /* Check if we did not end up below our original limit */
2523 comm->cutoff_mbody = std::max(comm->cutoff_mbody, r_bonded_limit);
2525 if (comm->cutoff_mbody > comm->cellsize_limit)
2527 comm->cellsize_limit = comm->cutoff_mbody;
2530 /* Without DLB and cutoff_mbody<cutoff, cutoff_mbody is dynamic */
2535 fprintf(debug, "Bonded atom communication beyond the cut-off: %s\n"
2536 "cellsize limit %f\n",
2537 gmx::boolToString(comm->bBondComm), comm->cellsize_limit);
2542 check_dd_restrictions(dd, ir, mdlog);
2546 static char *init_bLocalCG(const gmx_mtop_t *mtop)
2551 ncg = ncg_mtop(mtop);
2552 snew(bLocalCG, ncg);
2553 for (cg = 0; cg < ncg; cg++)
2555 bLocalCG[cg] = FALSE;
2561 void dd_init_bondeds(FILE *fplog,
2563 const gmx_mtop_t *mtop,
2564 const gmx_vsite_t *vsite,
2565 const t_inputrec *ir,
2566 gmx_bool bBCheck, cginfo_mb_t *cginfo_mb)
2568 gmx_domdec_comm_t *comm;
2570 dd_make_reverse_top(fplog, dd, mtop, vsite, ir, bBCheck);
2574 if (comm->bBondComm)
2576 /* Communicate atoms beyond the cut-off for bonded interactions */
2579 comm->cglink = make_charge_group_links(mtop, dd, cginfo_mb);
2581 comm->bLocalCG = init_bLocalCG(mtop);
2585 /* Only communicate atoms based on cut-off */
2586 comm->cglink = nullptr;
2587 comm->bLocalCG = nullptr;
2591 static void writeSettings(gmx::TextWriter *log,
2593 const gmx_mtop_t *mtop,
2594 const t_inputrec *ir,
2595 gmx_bool bDynLoadBal,
2597 const gmx_ddbox_t *ddbox)
2599 gmx_domdec_comm_t *comm;
2608 log->writeString("The maximum number of communication pulses is:");
2609 for (d = 0; d < dd->ndim; d++)
2611 log->writeStringFormatted(" %c %d", dim2char(dd->dim[d]), comm->cd[d].np_dlb);
2613 log->ensureLineBreak();
2614 log->writeLineFormatted("The minimum size for domain decomposition cells is %.3f nm", comm->cellsize_limit);
2615 log->writeLineFormatted("The requested allowed shrink of DD cells (option -dds) is: %.2f", dlb_scale);
2616 log->writeString("The allowed shrink of domain decomposition cells is:");
2617 for (d = 0; d < DIM; d++)
2621 if (d >= ddbox->npbcdim && dd->nc[d] == 2)
2628 comm->cellsize_min_dlb[d]/
2629 (ddbox->box_size[d]*ddbox->skew_fac[d]/dd->nc[d]);
2631 log->writeStringFormatted(" %c %.2f", dim2char(d), shrink);
2634 log->ensureLineBreak();
2638 set_dd_cell_sizes_slb(dd, ddbox, setcellsizeslbPULSE_ONLY, np);
2639 log->writeString("The initial number of communication pulses is:");
2640 for (d = 0; d < dd->ndim; d++)
2642 log->writeStringFormatted(" %c %d", dim2char(dd->dim[d]), np[dd->dim[d]]);
2644 log->ensureLineBreak();
2645 log->writeString("The initial domain decomposition cell size is:");
2646 for (d = 0; d < DIM; d++)
2650 log->writeStringFormatted(" %c %.2f nm",
2651 dim2char(d), dd->comm->cellsize_min[d]);
2654 log->ensureLineBreak();
2658 gmx_bool bInterCGVsites = count_intercg_vsites(mtop) != 0;
2660 if (comm->bInterCGBondeds ||
2662 dd->splitConstraints || dd->splitSettles)
2664 log->writeLine("The maximum allowed distance for charge groups involved in interactions is:");
2665 log->writeLineFormatted("%40s %-7s %6.3f nm", "non-bonded interactions", "", comm->cutoff);
2669 limit = dd->comm->cellsize_limit;
2673 if (dynamic_dd_box(*dd))
2675 log->writeLine("(the following are initial values, they could change due to box deformation)");
2677 limit = dd->comm->cellsize_min[XX];
2678 for (d = 1; d < DIM; d++)
2680 limit = std::min(limit, dd->comm->cellsize_min[d]);
2684 if (comm->bInterCGBondeds)
2686 log->writeLineFormatted("%40s %-7s %6.3f nm",
2687 "two-body bonded interactions", "(-rdd)",
2688 std::max(comm->cutoff, comm->cutoff_mbody));
2689 log->writeLineFormatted("%40s %-7s %6.3f nm",
2690 "multi-body bonded interactions", "(-rdd)",
2691 (comm->bBondComm || isDlbOn(dd->comm)) ? comm->cutoff_mbody : std::min(comm->cutoff, limit));
2695 log->writeLineFormatted("%40s %-7s %6.3f nm",
2696 "virtual site constructions", "(-rcon)", limit);
2698 if (dd->splitConstraints || dd->splitSettles)
2700 std::string separation = gmx::formatString("atoms separated by up to %d constraints",
2702 log->writeLineFormatted("%40s %-7s %6.3f nm\n",
2703 separation.c_str(), "(-rcon)", limit);
2705 log->ensureLineBreak();
2709 static void logSettings(const gmx::MDLogger &mdlog,
2711 const gmx_mtop_t *mtop,
2712 const t_inputrec *ir,
2714 const gmx_ddbox_t *ddbox)
2716 gmx::StringOutputStream stream;
2717 gmx::TextWriter log(&stream);
2718 writeSettings(&log, dd, mtop, ir, isDlbOn(dd->comm), dlb_scale, ddbox);
2719 if (dd->comm->dlbState == DlbState::offCanTurnOn)
2722 log.ensureEmptyLine();
2723 log.writeLine("When dynamic load balancing gets turned on, these settings will change to:");
2725 writeSettings(&log, dd, mtop, ir, true, dlb_scale, ddbox);
2727 GMX_LOG(mdlog.info).asParagraph().appendText(stream.toString());
2730 static void set_cell_limits_dlb(const gmx::MDLogger &mdlog,
2733 const t_inputrec *ir,
2734 const gmx_ddbox_t *ddbox)
2736 gmx_domdec_comm_t *comm;
2737 int d, dim, npulse, npulse_d_max, npulse_d;
2742 bNoCutOff = (ir->rvdw == 0 || ir->rcoulomb == 0);
2744 /* Determine the maximum number of comm. pulses in one dimension */
2746 comm->cellsize_limit = std::max(comm->cellsize_limit, comm->cutoff_mbody);
2748 /* Determine the maximum required number of grid pulses */
2749 if (comm->cellsize_limit >= comm->cutoff)
2751 /* Only a single pulse is required */
2754 else if (!bNoCutOff && comm->cellsize_limit > 0)
2756 /* We round down slightly here to avoid overhead due to the latency
2757 * of extra communication calls when the cut-off
2758 * would be only slightly longer than the cell size.
2759 * Later cellsize_limit is redetermined,
2760 * so we can not miss interactions due to this rounding.
2762 npulse = static_cast<int>(0.96 + comm->cutoff/comm->cellsize_limit);
2766 /* There is no cell size limit */
2767 npulse = std::max(dd->nc[XX]-1, std::max(dd->nc[YY]-1, dd->nc[ZZ]-1));
2770 if (!bNoCutOff && npulse > 1)
2772 /* See if we can do with less pulses, based on dlb_scale */
2774 for (d = 0; d < dd->ndim; d++)
2777 npulse_d = static_cast<int>(1 + dd->nc[dim]*comm->cutoff
2778 /(ddbox->box_size[dim]*ddbox->skew_fac[dim]*dlb_scale));
2779 npulse_d_max = std::max(npulse_d_max, npulse_d);
2781 npulse = std::min(npulse, npulse_d_max);
2784 /* This env var can override npulse */
2785 d = dd_getenv(mdlog, "GMX_DD_NPULSE", 0);
2792 comm->bVacDLBNoLimit = (ir->ePBC == epbcNONE);
2793 for (d = 0; d < dd->ndim; d++)
2795 comm->cd[d].np_dlb = std::min(npulse, dd->nc[dd->dim[d]]-1);
2796 comm->maxpulse = std::max(comm->maxpulse, comm->cd[d].np_dlb);
2797 if (comm->cd[d].np_dlb < dd->nc[dd->dim[d]]-1)
2799 comm->bVacDLBNoLimit = FALSE;
2803 /* cellsize_limit is set for LINCS in init_domain_decomposition */
2804 if (!comm->bVacDLBNoLimit)
2806 comm->cellsize_limit = std::max(comm->cellsize_limit,
2807 comm->cutoff/comm->maxpulse);
2809 comm->cellsize_limit = std::max(comm->cellsize_limit, comm->cutoff_mbody);
2810 /* Set the minimum cell size for each DD dimension */
2811 for (d = 0; d < dd->ndim; d++)
2813 if (comm->bVacDLBNoLimit ||
2814 comm->cd[d].np_dlb*comm->cellsize_limit >= comm->cutoff)
2816 comm->cellsize_min_dlb[dd->dim[d]] = comm->cellsize_limit;
2820 comm->cellsize_min_dlb[dd->dim[d]] =
2821 comm->cutoff/comm->cd[d].np_dlb;
2824 if (comm->cutoff_mbody <= 0)
2826 comm->cutoff_mbody = std::min(comm->cutoff, comm->cellsize_limit);
2834 gmx_bool dd_bonded_molpbc(const gmx_domdec_t *dd, int ePBC)
2836 /* If each molecule is a single charge group
2837 * or we use domain decomposition for each periodic dimension,
2838 * we do not need to take pbc into account for the bonded interactions.
2840 return (ePBC != epbcNONE && dd->comm->bInterCGBondeds &&
2843 (dd->nc[ZZ] > 1 || ePBC == epbcXY)));
2846 /*! \brief Sets grid size limits and PP-PME setup, prints settings to log */
2847 static void set_ddgrid_parameters(const gmx::MDLogger &mdlog,
2848 gmx_domdec_t *dd, real dlb_scale,
2849 const gmx_mtop_t *mtop, const t_inputrec *ir,
2850 const gmx_ddbox_t *ddbox)
2852 gmx_domdec_comm_t *comm;
2858 if (EEL_PME(ir->coulombtype) || EVDW_PME(ir->vdwtype))
2860 init_ddpme(dd, &comm->ddpme[0], 0);
2861 if (comm->npmedecompdim >= 2)
2863 init_ddpme(dd, &comm->ddpme[1], 1);
2868 comm->npmenodes = 0;
2869 if (dd->pme_nodeid >= 0)
2871 gmx_fatal_collective(FARGS, dd->mpi_comm_all, DDMASTER(dd),
2872 "Can not have separate PME ranks without PME electrostatics");
2878 fprintf(debug, "The DD cut-off is %f\n", comm->cutoff);
2880 if (!isDlbDisabled(comm))
2882 set_cell_limits_dlb(mdlog, dd, dlb_scale, ir, ddbox);
2885 logSettings(mdlog, dd, mtop, ir, dlb_scale, ddbox);
2887 if (ir->ePBC == epbcNONE)
2889 vol_frac = 1 - 1/static_cast<double>(dd->nnodes);
2894 (1 + comm_box_frac(dd->nc, comm->cutoff, ddbox))/static_cast<double>(dd->nnodes);
2898 fprintf(debug, "Volume fraction for all DD zones: %f\n", vol_frac);
2900 natoms_tot = comm->cgs_gl.index[comm->cgs_gl.nr];
2902 dd->ga2la = new gmx_ga2la_t(natoms_tot,
2903 static_cast<int>(vol_frac*natoms_tot));
2906 /*! \brief Set some important DD parameters that can be modified by env.vars */
2907 static void set_dd_envvar_options(const gmx::MDLogger &mdlog,
2908 gmx_domdec_t *dd, int rank_mysim)
2910 gmx_domdec_comm_t *comm = dd->comm;
2912 dd->bSendRecv2 = (dd_getenv(mdlog, "GMX_DD_USE_SENDRECV2", 0) != 0);
2913 comm->dlb_scale_lim = dd_getenv(mdlog, "GMX_DLB_MAX_BOX_SCALING", 10);
2914 comm->eFlop = dd_getenv(mdlog, "GMX_DLB_BASED_ON_FLOPS", 0);
2915 int recload = dd_getenv(mdlog, "GMX_DD_RECORD_LOAD", 1);
2916 comm->nstDDDump = dd_getenv(mdlog, "GMX_DD_NST_DUMP", 0);
2917 comm->nstDDDumpGrid = dd_getenv(mdlog, "GMX_DD_NST_DUMP_GRID", 0);
2918 comm->DD_debug = dd_getenv(mdlog, "GMX_DD_DEBUG", 0);
2922 GMX_LOG(mdlog.info).appendText("Will use two sequential MPI_Sendrecv calls instead of two simultaneous non-blocking MPI_Irecv and MPI_Isend pairs for constraint and vsite communication");
2927 GMX_LOG(mdlog.info).appendText("Will load balance based on FLOP count");
2928 if (comm->eFlop > 1)
2930 srand(1 + rank_mysim);
2932 comm->bRecordLoad = TRUE;
2936 comm->bRecordLoad = (wallcycle_have_counter() && recload > 0);
2940 gmx_domdec_t *init_domain_decomposition(const gmx::MDLogger &mdlog,
2942 const DomdecOptions &options,
2943 const MdrunOptions &mdrunOptions,
2944 const gmx_mtop_t *mtop,
2945 const t_inputrec *ir,
2947 gmx::ArrayRef<const gmx::RVec> xGlobal,
2948 gmx::LocalAtomSetManager *atomSets)
2952 GMX_LOG(mdlog.info).appendTextFormatted(
2953 "\nInitializing Domain Decomposition on %d ranks", cr->nnodes);
2955 dd = new gmx_domdec_t;
2957 dd->comm = init_dd_comm();
2959 /* Initialize DD paritioning counters */
2960 dd->comm->partition_step = INT_MIN;
2963 set_dd_envvar_options(mdlog, dd, cr->nodeid);
2965 gmx_ddbox_t ddbox = {0};
2966 set_dd_limits_and_grid(mdlog, cr, dd, options, mdrunOptions,
2971 make_dd_communicators(mdlog, cr, dd, options.rankOrder);
2973 if (thisRankHasDuty(cr, DUTY_PP))
2975 set_ddgrid_parameters(mdlog, dd, options.dlbScaling, mtop, ir, &ddbox);
2977 setup_neighbor_relations(dd);
2980 /* Set overallocation to avoid frequent reallocation of arrays */
2981 set_over_alloc_dd(TRUE);
2983 clear_dd_cycle_counts(dd);
2985 dd->atomSets = atomSets;
2990 static gmx_bool test_dd_cutoff(t_commrec *cr,
2991 const t_state &state,
2992 real cutoffRequested)
3002 set_ddbox(*dd, false, state.box, true, state.x, &ddbox);
3006 for (d = 0; d < dd->ndim; d++)
3010 inv_cell_size = DD_CELL_MARGIN*dd->nc[dim]/ddbox.box_size[dim];
3011 if (dynamic_dd_box(*dd))
3013 inv_cell_size *= DD_PRES_SCALE_MARGIN;
3016 np = 1 + static_cast<int>(cutoffRequested*inv_cell_size*ddbox.skew_fac[dim]);
3018 if (!isDlbDisabled(dd->comm) && (dim < ddbox.npbcdim) && (dd->comm->cd[d].np_dlb > 0))
3020 if (np > dd->comm->cd[d].np_dlb)
3025 /* If a current local cell size is smaller than the requested
3026 * cut-off, we could still fix it, but this gets very complicated.
3027 * Without fixing here, we might actually need more checks.
3029 real cellSizeAlongDim = (dd->comm->cell_x1[dim] - dd->comm->cell_x0[dim])*ddbox.skew_fac[dim];
3030 if (cellSizeAlongDim*dd->comm->cd[d].np_dlb < cutoffRequested)
3037 if (!isDlbDisabled(dd->comm))
3039 /* If DLB is not active yet, we don't need to check the grid jumps.
3040 * Actually we shouldn't, because then the grid jump data is not set.
3042 if (isDlbOn(dd->comm) &&
3043 check_grid_jump(0, dd, cutoffRequested, &ddbox, FALSE))
3048 gmx_sumi(1, &LocallyLimited, cr);
3050 if (LocallyLimited > 0)
3059 gmx_bool change_dd_cutoff(t_commrec *cr,
3060 const t_state &state,
3061 real cutoffRequested)
3063 gmx_bool bCutoffAllowed;
3065 bCutoffAllowed = test_dd_cutoff(cr, state, cutoffRequested);
3069 cr->dd->comm->cutoff = cutoffRequested;
3072 return bCutoffAllowed;