src/gromacs/mdlib/force.cpp

   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) 2018,2019,2020,2021, 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 "force.h"
  41
  42 #include <cassert>
  43 #include <cmath>
  44 #include <cstring>
  45
  46 #include "gromacs/domdec/dlbtiming.h"
  47 #include "gromacs/domdec/domdec.h"
  48 #include "gromacs/domdec/domdec_struct.h"
  49 #include "gromacs/ewald/ewald.h"
  50 #include "gromacs/ewald/long_range_correction.h"
  51 #include "gromacs/ewald/pme.h"
  52 #include "gromacs/gmxlib/network.h"
  53 #include "gromacs/gmxlib/nrnb.h"
  54 #include "gromacs/math/vec.h"
  55 #include "gromacs/math/vecdump.h"
  56 #include "gromacs/mdlib/forcerec_threading.h"
  57 #include "gromacs/mdtypes/commrec.h"
  58 #include "gromacs/mdtypes/enerdata.h"
  59 #include "gromacs/mdtypes/forceoutput.h"
  60 #include "gromacs/mdtypes/forcerec.h"
  61 #include "gromacs/mdtypes/inputrec.h"
  62 #include "gromacs/mdtypes/interaction_const.h"
  63 #include "gromacs/mdtypes/md_enums.h"
  64 #include "gromacs/mdtypes/mdatom.h"
  65 #include "gromacs/mdtypes/simulation_workload.h"
  66 #include "gromacs/pbcutil/ishift.h"
  67 #include "gromacs/pbcutil/pbc.h"
  68 #include "gromacs/timing/wallcycle.h"
  69 #include "gromacs/utility/exceptions.h"
  70 #include "gromacs/utility/fatalerror.h"
  71 #include "gromacs/utility/smalloc.h"
  72
  73 using gmx::ArrayRef;
  74 using gmx::RVec;
  75
  76 static void clearEwaldThreadOutput(ewald_corr_thread_t* ewc_t)
  77 {
  78     ewc_t->Vcorr_q        = 0;
  79     ewc_t->Vcorr_lj       = 0;
  80     ewc_t->dvdl[efptCOUL] = 0;
  81     ewc_t->dvdl[efptVDW]  = 0;
  82     clear_mat(ewc_t->vir_q);
  83     clear_mat(ewc_t->vir_lj);
  84 }
  85
  86 static void reduceEwaldThreadOuput(int nthreads, ewald_corr_thread_t* ewc_t)
  87 {
  88     ewald_corr_thread_t& dest = ewc_t[0];
  89
  90     for (int t = 1; t < nthreads; t++)
  91     {
  92         dest.Vcorr_q += ewc_t[t].Vcorr_q;
  93         dest.Vcorr_lj += ewc_t[t].Vcorr_lj;
  94         dest.dvdl[efptCOUL] += ewc_t[t].dvdl[efptCOUL];
  95         dest.dvdl[efptVDW] += ewc_t[t].dvdl[efptVDW];
  96         m_add(dest.vir_q, ewc_t[t].vir_q, dest.vir_q);
  97         m_add(dest.vir_lj, ewc_t[t].vir_lj, dest.vir_lj);
  98     }
  99 }
 100
 101 void calculateLongRangeNonbondeds(t_forcerec*                   fr,
 102                                   const t_inputrec&             ir,
 103                                   const t_commrec*              cr,
 104                                   t_nrnb*                       nrnb,
 105                                   gmx_wallcycle_t               wcycle,
 106                                   const t_mdatoms*              md,
 107                                   gmx::ArrayRef<const RVec>     coordinates,
 108                                   gmx::ForceWithVirial*         forceWithVirial,
 109                                   gmx_enerdata_t*               enerd,
 110                                   const matrix                  box,
 111                                   const real*                   lambda,
 112                                   const rvec*                   mu_tot,
 113                                   const gmx::StepWorkload&      stepWork,
 114                                   const DDBalanceRegionHandler& ddBalanceRegionHandler)
 115 {
 116     const bool computePmeOnCpu = (EEL_PME(fr->ic->eeltype) || EVDW_PME(fr->ic->vdwtype))
 117                                  && thisRankHasDuty(cr, DUTY_PME)
 118                                  && (pme_run_mode(fr->pmedata) == PmeRunMode::CPU);
 119
 120     const bool haveEwaldSurfaceTerm = haveEwaldSurfaceContribution(ir);
 121
 122     /* Do long-range electrostatics and/or LJ-PME
 123      * and compute PME surface terms when necessary.
 124      */
 125     if ((computePmeOnCpu || fr->ic->eeltype == eelEWALD || haveEwaldSurfaceTerm)
 126         && stepWork.computeNonbondedForces)
 127     {
 128         int  status = 0;
 129         real Vlr_q = 0, Vlr_lj = 0;
 130
 131         /* We reduce all virial, dV/dlambda and energy contributions, except
 132          * for the reciprocal energies (Vlr_q, Vlr_lj) into the same struct.
 133          */
 134         ewald_corr_thread_t& ewaldOutput = fr->ewc_t[0];
 135         clearEwaldThreadOutput(&ewaldOutput);
 136
 137         if (EEL_PME_EWALD(fr->ic->eeltype) || EVDW_PME(fr->ic->vdwtype))
 138         {
 139             /* Calculate the Ewald surface force and energy contributions, when necessary */
 140             if (haveEwaldSurfaceTerm)
 141             {
 142                 wallcycle_sub_start(wcycle, ewcsEWALD_CORRECTION);
 143
 144                 int nthreads = fr->nthread_ewc;
 145 #pragma omp parallel for num_threads(nthreads) schedule(static)
 146                 for (int t = 0; t < nthreads; t++)
 147                 {
 148                     try
 149                     {
 150                         ewald_corr_thread_t& ewc_t = fr->ewc_t[t];
 151                         if (t > 0)
 152                         {
 153                             clearEwaldThreadOutput(&ewc_t);
 154                         }
 155
 156                         /* Threading is only supported with the Verlet cut-off
 157                          * scheme and then only single particle forces (no
 158                          * exclusion forces) are calculated, so we can store
 159                          * the forces in the normal, single forceWithVirial->force_ array.
 160                          */
 161                         const rvec* x = as_rvec_array(coordinates.data());
 162                         ewald_LRcorrection(md->homenr,
 163                                            cr,
 164                                            nthreads,
 165                                            t,
 166                                            *fr,
 167                                            ir,
 168                                            md->chargeA,
 169                                            md->chargeB,
 170                                            (md->nChargePerturbed != 0),
 171                                            x,
 172                                            box,
 173                                            mu_tot,
 174                                            as_rvec_array(forceWithVirial->force_.data()),
 175                                            &ewc_t.Vcorr_q,
 176                                            lambda[efptCOUL],
 177                                            &ewc_t.dvdl[efptCOUL]);
 178                     }
 179                     GMX_CATCH_ALL_AND_EXIT_WITH_FATAL_ERROR
 180                 }
 181                 if (nthreads > 1)
 182                 {
 183                     reduceEwaldThreadOuput(nthreads, fr->ewc_t);
 184                 }
 185                 wallcycle_sub_stop(wcycle, ewcsEWALD_CORRECTION);
 186             }
 187
 188             if (EEL_PME_EWALD(fr->ic->eeltype) && fr->n_tpi == 0)
 189             {
 190                 /* This is not in a subcounter because it takes a
 191                    negligible and constant-sized amount of time */
 192                 ewaldOutput.Vcorr_q += ewald_charge_correction(
 193                         cr, fr, lambda[efptCOUL], box, &ewaldOutput.dvdl[efptCOUL], ewaldOutput.vir_q);
 194             }
 195
 196             if (computePmeOnCpu)
 197             {
 198                 /* Do reciprocal PME for Coulomb and/or LJ. */
 199                 assert(fr->n_tpi >= 0);
 200                 if (fr->n_tpi == 0 || stepWork.stateChanged)
 201                 {
 202                     /* With domain decomposition we close the CPU side load
 203                      * balancing region here, because PME does global
 204                      * communication that acts as a global barrier.
 205                      */
 206                     ddBalanceRegionHandler.closeAfterForceComputationCpu();
 207
 208                     wallcycle_start(wcycle, ewcPMEMESH);
 209                     status = gmx_pme_do(
 210                             fr->pmedata,
 211                             gmx::constArrayRefFromArray(coordinates.data(), md->homenr - fr->n_tpi),
 212                             forceWithVirial->force_,
 213                             md->chargeA,
 214                             md->chargeB,
 215                             md->sqrt_c6A,
 216                             md->sqrt_c6B,
 217                             md->sigmaA,
 218                             md->sigmaB,
 219                             box,
 220                             cr,
 221                             DOMAINDECOMP(cr) ? dd_pme_maxshift_x(cr->dd) : 0,
 222                             DOMAINDECOMP(cr) ? dd_pme_maxshift_y(cr->dd) : 0,
 223                             nrnb,
 224                             wcycle,
 225                             ewaldOutput.vir_q,
 226                             ewaldOutput.vir_lj,
 227                             &Vlr_q,
 228                             &Vlr_lj,
 229                             lambda[efptCOUL],
 230                             lambda[efptVDW],
 231                             &ewaldOutput.dvdl[efptCOUL],
 232                             &ewaldOutput.dvdl[efptVDW],
 233                             stepWork);
 234                     wallcycle_stop(wcycle, ewcPMEMESH);
 235                     if (status != 0)
 236                     {
 237                         gmx_fatal(FARGS, "Error %d in reciprocal PME routine", status);
 238                     }
 239
 240                     /* We should try to do as little computation after
 241                      * this as possible, because parallel PME synchronizes
 242                      * the nodes, so we want all load imbalance of the
 243                      * rest of the force calculation to be before the PME
 244                      * call.  DD load balancing is done on the whole time
 245                      * of the force call (without PME).
 246                      */
 247                 }
 248                 if (fr->n_tpi > 0)
 249                 {
 250                     /* Determine the PME grid energy of the test molecule
 251                      * with the PME grid potential of the other charges.
 252                      */
 253                     gmx_pme_calc_energy(
 254                             fr->pmedata,
 255                             coordinates.subArray(md->homenr - fr->n_tpi, fr->n_tpi),
 256                             gmx::arrayRefFromArray(md->chargeA + md->homenr - fr->n_tpi, fr->n_tpi),
 257                             &Vlr_q);
 258                 }
 259             }
 260         }
 261
 262         if (fr->ic->eeltype == eelEWALD)
 263         {
 264             const rvec* x = as_rvec_array(coordinates.data());
 265             Vlr_q         = do_ewald(ir,
 266                              x,
 267                              as_rvec_array(forceWithVirial->force_.data()),
 268                              md->chargeA,
 269                              md->chargeB,
 270                              box,
 271                              cr,
 272                              md->homenr,
 273                              ewaldOutput.vir_q,
 274                              fr->ic->ewaldcoeff_q,
 275                              lambda[efptCOUL],
 276                              &ewaldOutput.dvdl[efptCOUL],
 277                              fr->ewald_table);
 278         }
 279
 280         /* Note that with separate PME nodes we get the real energies later */
 281         // TODO it would be simpler if we just accumulated a single
 282         // long-range virial contribution.
 283         forceWithVirial->addVirialContribution(ewaldOutput.vir_q);
 284         forceWithVirial->addVirialContribution(ewaldOutput.vir_lj);
 285         enerd->dvdl_lin[efptCOUL] += ewaldOutput.dvdl[efptCOUL];
 286         enerd->dvdl_lin[efptVDW] += ewaldOutput.dvdl[efptVDW];
 287         enerd->term[F_COUL_RECIP] = Vlr_q + ewaldOutput.Vcorr_q;
 288         enerd->term[F_LJ_RECIP]   = Vlr_lj + ewaldOutput.Vcorr_lj;
 289
 290         if (debug)
 291         {
 292             fprintf(debug,
 293                     "Vlr_q = %g, Vcorr_q = %g, Vlr_corr_q = %g\n",
 294                     Vlr_q,
 295                     ewaldOutput.Vcorr_q,
 296                     enerd->term[F_COUL_RECIP]);
 297             pr_rvecs(debug, 0, "vir_el_recip after corr", ewaldOutput.vir_q, DIM);
 298             fprintf(debug,
 299                     "Vlr_lj: %g, Vcorr_lj = %g, Vlr_corr_lj = %g\n",
 300                     Vlr_lj,
 301                     ewaldOutput.Vcorr_lj,
 302                     enerd->term[F_LJ_RECIP]);
 303             pr_rvecs(debug, 0, "vir_lj_recip after corr", ewaldOutput.vir_lj, DIM);
 304         }
 305     }
 306
 307     if (debug)
 308     {
 309         print_nrnb(debug, nrnb);
 310     }
 311 }