src/gromacs/pulling/pullutil.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 "config.h"
  41
  42 #include <cassert>
  43 #include <cstdlib>
  44
  45 #include "gromacs/fileio/confio.h"
  46 #include "gromacs/gmxlib/network.h"
  47 #include "gromacs/math/functions.h"
  48 #include "gromacs/math/utilities.h"
  49 #include "gromacs/math/vec.h"
  50 #include "gromacs/mdtypes/commrec.h"
  51 #include "gromacs/mdtypes/inputrec.h"
  52 #include "gromacs/mdtypes/md_enums.h"
  53 #include "gromacs/mdtypes/mdatom.h"
  54 #include "gromacs/mdtypes/state.h"
  55 #include "gromacs/pbcutil/pbc.h"
  56 #include "gromacs/pulling/pull.h"
  57 #include "gromacs/utility/fatalerror.h"
  58 #include "gromacs/utility/futil.h"
  59 #include "gromacs/utility/gmxassert.h"
  60 #include "gromacs/utility/real.h"
  61 #include "gromacs/utility/smalloc.h"
  62
  63 #include "pull_internal.h"
  64
  65 #if GMX_MPI
  66
  67 // Helper function to deduce MPI datatype from the type of data
  68 gmx_unused static MPI_Datatype mpiDatatype(const float gmx_unused* data)
  69 {
  70     return MPI_FLOAT;
  71 }
  72
  73 // Helper function to deduce MPI datatype from the type of data
  74 gmx_unused static MPI_Datatype mpiDatatype(const double gmx_unused* data)
  75 {
  76     return MPI_DOUBLE;
  77 }
  78
  79 #endif // GMX_MPI
  80
  81 #if !GMX_DOUBLE
  82 // Helper function; note that gmx_sum(d) should actually be templated
  83 gmx_unused static void gmxAllReduce(int n, real* data, const t_commrec* cr)
  84 {
  85     gmx_sum(n, data, cr);
  86 }
  87 #endif
  88
  89 // Helper function; note that gmx_sum(d) should actually be templated
  90 gmx_unused static void gmxAllReduce(int n, double* data, const t_commrec* cr)
  91 {
  92     gmx_sumd(n, data, cr);
  93 }
  94
  95 // Reduce data of n elements over all ranks currently participating in pull
  96 template<typename T>
  97 static void pullAllReduce(const t_commrec* cr, pull_comm_t* comm, int n, T* data)
  98 {
  99     if (cr != nullptr && PAR(cr))
 100     {
 101         if (comm->bParticipateAll)
 102         {
 103             /* Sum the contributions over all DD ranks */
 104             gmxAllReduce(n, data, cr);
 105         }
 106         else
 107         {
 108             /* Separate branch because gmx_sum uses cr->mpi_comm_mygroup */
 109 #if GMX_MPI
 110             MPI_Allreduce(MPI_IN_PLACE, data, n, mpiDatatype(data), MPI_SUM, comm->mpi_comm_com);
 111 #else
 112             gmx_incons("comm->bParticipateAll=FALSE without GMX_MPI");
 113 #endif
 114         }
 115     }
 116 }
 117
 118 /* Copies the coordinates of the PBC atom of pgrp to x_pbc.
 119  * When those coordinates are not available on this rank, clears x_pbc.
 120  */
 121 static void setPbcAtomCoords(const pull_group_work_t& pgrp, const rvec* x, rvec x_pbc)
 122 {
 123     if (pgrp.pbcAtomSet != nullptr)
 124     {
 125         if (pgrp.pbcAtomSet->numAtomsLocal() > 0)
 126         {
 127             /* We have the atom locally, copy its coordinates */
 128             copy_rvec(x[pgrp.pbcAtomSet->localIndex()[0]], x_pbc);
 129         }
 130         else
 131         {
 132             /* Another rank has it, clear the coordinates for MPI_Allreduce */
 133             clear_rvec(x_pbc);
 134         }
 135     }
 136     else
 137     {
 138         copy_rvec(x[pgrp.params.pbcatom], x_pbc);
 139     }
 140 }
 141
 142 static void pull_set_pbcatoms(const t_commrec* cr, struct pull_t* pull, const rvec* x, gmx::ArrayRef<gmx::RVec> x_pbc)
 143 {
 144     int numPbcAtoms = 0;
 145     for (size_t g = 0; g < pull->group.size(); g++)
 146     {
 147         const pull_group_work_t& group = pull->group[g];
 148         if (group.needToCalcCom && (group.epgrppbc == epgrppbcREFAT || group.epgrppbc == epgrppbcPREVSTEPCOM))
 149         {
 150             setPbcAtomCoords(pull->group[g], x, x_pbc[g]);
 151             numPbcAtoms++;
 152         }
 153         else
 154         {
 155             clear_rvec(x_pbc[g]);
 156         }
 157     }
 158
 159     if (cr && PAR(cr) && numPbcAtoms > 0)
 160     {
 161         /* Sum over participating ranks to get x_pbc from the home ranks.
 162          * This can be very expensive at high parallelization, so we only
 163          * do this after each DD repartitioning.
 164          */
 165         pullAllReduce(cr, &pull->comm, pull->group.size() * DIM, static_cast<real*>(x_pbc[0]));
 166     }
 167 }
 168
 169 static void
 170 make_cyl_refgrps(const t_commrec* cr, pull_t* pull, const real* masses, t_pbc* pbc, double t, const rvec* x)
 171 {
 172     pull_comm_t* comm = &pull->comm;
 173
 174     GMX_ASSERT(comm->cylinderBuffer.size() == pull->coord.size() * c_cylinderBufferStride,
 175                "cylinderBuffer should have the correct size");
 176
 177     double inv_cyl_r2 = 1.0 / gmx::square(pull->params.cylinder_r);
 178
 179     /* loop over all groups to make a reference group for each*/
 180     for (size_t c = 0; c < pull->coord.size(); c++)
 181     {
 182         pull_coord_work_t* pcrd;
 183         double             sum_a, wmass, wwmass;
 184         dvec               radf_fac0, radf_fac1;
 185
 186         pcrd = &pull->coord[c];
 187
 188         sum_a  = 0;
 189         wmass  = 0;
 190         wwmass = 0;
 191         clear_dvec(radf_fac0);
 192         clear_dvec(radf_fac1);
 193
 194         if (pcrd->params.eGeom == PullGroupGeometry::Cylinder)
 195         {
 196             /* pref will be the same group for all pull coordinates */
 197             const pull_group_work_t& pref  = pull->group[pcrd->params.group[0]];
 198             const pull_group_work_t& pgrp  = pull->group[pcrd->params.group[1]];
 199             pull_group_work_t&       pdyna = pull->dyna[c];
 200             rvec                     direction;
 201             copy_dvec_to_rvec(pcrd->spatialData.vec, direction);
 202
 203             /* Since we have not calculated the COM of the cylinder group yet,
 204              * we calculate distances with respect to location of the pull
 205              * group minus the reference position along the vector.
 206              * here we already have the COM of the pull group. This resolves
 207              * any PBC issues and we don't need to use a PBC-atom here.
 208              */
 209             if (pcrd->params.rate != 0)
 210             {
 211                 /* With rate=0, value_ref is set initially */
 212                 pcrd->value_ref = pcrd->params.init + pcrd->params.rate * t;
 213             }
 214             rvec reference;
 215             for (int m = 0; m < DIM; m++)
 216             {
 217                 reference[m] = pgrp.x[m] - pcrd->spatialData.vec[m] * pcrd->value_ref;
 218             }
 219
 220             auto localAtomIndices = pref.atomSet.localIndex();
 221
 222             /* This actually only needs to be done at init or DD time,
 223              * but resizing with the same size does not cause much overhead.
 224              */
 225             pdyna.localWeights.resize(localAtomIndices.size());
 226             pdyna.mdw.resize(localAtomIndices.size());
 227             pdyna.dv.resize(localAtomIndices.size());
 228
 229             /* loop over all atoms in the main ref group */
 230             for (gmx::index indexInSet = 0; indexInSet < localAtomIndices.ssize(); indexInSet++)
 231             {
 232                 int  atomIndex = localAtomIndices[indexInSet];
 233                 rvec dx;
 234                 pbc_dx_aiuc(pbc, x[atomIndex], reference, dx);
 235                 double axialLocation = iprod(direction, dx);
 236                 dvec   radialLocation;
 237                 double dr2 = 0;
 238                 for (int m = 0; m < DIM; m++)
 239                 {
 240                     /* Determine the radial components */
 241                     radialLocation[m] = dx[m] - axialLocation * direction[m];
 242                     dr2 += gmx::square(radialLocation[m]);
 243                 }
 244                 double dr2_rel = dr2 * inv_cyl_r2;
 245
 246                 if (dr2_rel < 1)
 247                 {
 248                     /* add atom to sum of COM and to weight array */
 249
 250                     double mass = masses[atomIndex];
 251                     /* The radial weight function is 1-2x^2+x^4,
 252                      * where x=r/cylinder_r. Since this function depends
 253                      * on the radial component, we also get radial forces
 254                      * on both groups.
 255                      */
 256                     double weight                  = 1 + (-2 + dr2_rel) * dr2_rel;
 257                     double dweight_r               = (-4 + 4 * dr2_rel) * inv_cyl_r2;
 258                     pdyna.localWeights[indexInSet] = weight;
 259                     sum_a += mass * weight * axialLocation;
 260                     wmass += mass * weight;
 261                     wwmass += mass * weight * weight;
 262                     dvec mdw;
 263                     dsvmul(mass * dweight_r, radialLocation, mdw);
 264                     copy_dvec(mdw, pdyna.mdw[indexInSet]);
 265                     /* Currently we only have the axial component of the
 266                      * offset from the cylinder COM up to an unkown offset.
 267                      * We add this offset after the reduction needed
 268                      * for determining the COM of the cylinder group.
 269                      */
 270                     pdyna.dv[indexInSet] = axialLocation;
 271                     for (int m = 0; m < DIM; m++)
 272                     {
 273                         radf_fac0[m] += mdw[m];
 274                         radf_fac1[m] += mdw[m] * axialLocation;
 275                     }
 276                 }
 277                 else
 278                 {
 279                     pdyna.localWeights[indexInSet] = 0;
 280                 }
 281             }
 282         }
 283
 284         auto buffer = gmx::arrayRefFromArray(
 285                 comm->cylinderBuffer.data() + c * c_cylinderBufferStride, c_cylinderBufferStride);
 286
 287         buffer[0] = wmass;
 288         buffer[1] = wwmass;
 289         buffer[2] = sum_a;
 290
 291         buffer[3] = radf_fac0[XX];
 292         buffer[4] = radf_fac0[YY];
 293         buffer[5] = radf_fac0[ZZ];
 294
 295         buffer[6] = radf_fac1[XX];
 296         buffer[7] = radf_fac1[YY];
 297         buffer[8] = radf_fac1[ZZ];
 298     }
 299
 300     if (cr != nullptr && PAR(cr))
 301     {
 302         /* Sum the contributions over the ranks */
 303         pullAllReduce(cr, comm, pull->coord.size() * c_cylinderBufferStride, comm->cylinderBuffer.data());
 304     }
 305
 306     for (size_t c = 0; c < pull->coord.size(); c++)
 307     {
 308         pull_coord_work_t* pcrd;
 309
 310         pcrd = &pull->coord[c];
 311
 312         if (pcrd->params.eGeom == PullGroupGeometry::Cylinder)
 313         {
 314             pull_group_work_t*    pdyna       = &pull->dyna[c];
 315             pull_group_work_t*    pgrp        = &pull->group[pcrd->params.group[1]];
 316             PullCoordSpatialData& spatialData = pcrd->spatialData;
 317
 318             auto buffer = gmx::constArrayRefFromArray(
 319                     comm->cylinderBuffer.data() + c * c_cylinderBufferStride, c_cylinderBufferStride);
 320             double wmass   = buffer[0];
 321             double wwmass  = buffer[1];
 322             pdyna->mwscale = 1.0 / wmass;
 323             /* Cylinder pulling can't be used with constraints, but we set
 324              * wscale and invtm anyhow, in case someone would like to use them.
 325              */
 326             pdyna->wscale = wmass / wwmass;
 327             pdyna->invtm  = wwmass / (wmass * wmass);
 328
 329             /* We store the deviation of the COM from the reference location
 330              * used above, since we need it when we apply the radial forces
 331              * to the atoms in the cylinder group.
 332              */
 333             spatialData.cyl_dev = 0;
 334             for (int m = 0; m < DIM; m++)
 335             {
 336                 double reference = pgrp->x[m] - spatialData.vec[m] * pcrd->value_ref;
 337                 double dist      = -spatialData.vec[m] * buffer[2] * pdyna->mwscale;
 338                 pdyna->x[m]      = reference - dist;
 339                 spatialData.cyl_dev += dist;
 340             }
 341             /* Now we know the exact COM of the cylinder reference group,
 342              * we can determine the radial force factor (ffrad) that when
 343              * multiplied with the axial pull force will give the radial
 344              * force on the pulled (non-cylinder) group.
 345              */
 346             for (int m = 0; m < DIM; m++)
 347             {
 348                 spatialData.ffrad[m] = (buffer[6 + m] + buffer[3 + m] * spatialData.cyl_dev) / wmass;
 349             }
 350
 351             if (debug)
 352             {
 353                 fprintf(debug,
 354                         "Pull cylinder group %zu:%8.3f%8.3f%8.3f m:%8.3f\n",
 355                         c,
 356                         pdyna->x[0],
 357                         pdyna->x[1],
 358                         pdyna->x[2],
 359                         1.0 / pdyna->invtm);
 360                 fprintf(debug,
 361                         "ffrad %8.3f %8.3f %8.3f\n",
 362                         spatialData.ffrad[XX],
 363                         spatialData.ffrad[YY],
 364                         spatialData.ffrad[ZZ]);
 365             }
 366         }
 367     }
 368 }
 369
 370 static double atan2_0_2pi(double y, double x)
 371 {
 372     double a;
 373
 374     a = atan2(y, x);
 375     if (a < 0)
 376     {
 377         a += 2.0 * M_PI;
 378     }
 379     return a;
 380 }
 381
 382 static void sum_com_part(const pull_group_work_t* pgrp,
 383                          int                      ind_start,
 384                          int                      ind_end,
 385                          const rvec*              x,
 386                          const rvec*              xp,
 387                          const real*              mass,
 388                          const t_pbc*             pbc,
 389                          const rvec               x_pbc,
 390                          ComSums*                 sum_com)
 391 {
 392     double sum_wm   = 0;
 393     double sum_wwm  = 0;
 394     dvec   sum_wmx  = { 0, 0, 0 };
 395     dvec   sum_wmxp = { 0, 0, 0 };
 396
 397     auto localAtomIndices = pgrp->atomSet.localIndex();
 398     for (int i = ind_start; i < ind_end; i++)
 399     {
 400         int  ii = localAtomIndices[i];
 401         real wm;
 402         if (pgrp->localWeights.empty())
 403         {
 404             wm = mass[ii];
 405             sum_wm += wm;
 406         }
 407         else
 408         {
 409             real w;
 410
 411             w  = pgrp->localWeights[i];
 412             wm = w * mass[ii];
 413             sum_wm += wm;
 414             sum_wwm += wm * w;
 415         }
 416         if (pgrp->epgrppbc == epgrppbcNONE)
 417         {
 418             /* Plain COM: sum the coordinates */
 419             for (int d = 0; d < DIM; d++)
 420             {
 421                 sum_wmx[d] += wm * x[ii][d];
 422             }
 423             if (xp)
 424             {
 425                 for (int d = 0; d < DIM; d++)
 426                 {
 427                     sum_wmxp[d] += wm * xp[ii][d];
 428                 }
 429             }
 430         }
 431         else
 432         {
 433             rvec dx;
 434
 435             /* Sum the difference with the reference atom */
 436             pbc_dx(pbc, x[ii], x_pbc, dx);
 437             for (int d = 0; d < DIM; d++)
 438             {
 439                 sum_wmx[d] += wm * dx[d];
 440             }
 441             if (xp)
 442             {
 443                 /* For xp add the difference between xp and x to dx,
 444                  * such that we use the same periodic image,
 445                  * also when xp has a large displacement.
 446                  */
 447                 for (int d = 0; d < DIM; d++)
 448                 {
 449                     sum_wmxp[d] += wm * (dx[d] + xp[ii][d] - x[ii][d]);
 450                 }
 451             }
 452         }
 453     }
 454
 455     sum_com->sum_wm  = sum_wm;
 456     sum_com->sum_wwm = sum_wwm;
 457     copy_dvec(sum_wmx, sum_com->sum_wmx);
 458     if (xp)
 459     {
 460         copy_dvec(sum_wmxp, sum_com->sum_wmxp);
 461     }
 462 }
 463
 464 static void sum_com_part_cosweight(const pull_group_work_t* pgrp,
 465                                    int                      ind_start,
 466                                    int                      ind_end,
 467                                    int                      cosdim,
 468                                    real                     twopi_box,
 469                                    const rvec*              x,
 470                                    const rvec*              xp,
 471                                    const real*              mass,
 472                                    ComSums*                 sum_com)
 473 {
 474     /* Cosine weighting geometry */
 475     double sum_cm  = 0;
 476     double sum_sm  = 0;
 477     double sum_ccm = 0;
 478     double sum_csm = 0;
 479     double sum_ssm = 0;
 480     double sum_cmp = 0;
 481     double sum_smp = 0;
 482
 483     auto localAtomIndices = pgrp->atomSet.localIndex();
 484
 485     for (int i = ind_start; i < ind_end; i++)
 486     {
 487         int  ii = localAtomIndices[i];
 488         real m  = mass[ii];
 489         /* Determine cos and sin sums */
 490         real cw = std::cos(x[ii][cosdim] * twopi_box);
 491         real sw = std::sin(x[ii][cosdim] * twopi_box);
 492         sum_cm += static_cast<double>(cw * m);
 493         sum_sm += static_cast<double>(sw * m);
 494         sum_ccm += static_cast<double>(cw * cw * m);
 495         sum_csm += static_cast<double>(cw * sw * m);
 496         sum_ssm += static_cast<double>(sw * sw * m);
 497
 498         if (xp != nullptr)
 499         {
 500             real cw = std::cos(xp[ii][cosdim] * twopi_box);
 501             real sw = std::sin(xp[ii][cosdim] * twopi_box);
 502             sum_cmp += static_cast<double>(cw * m);
 503             sum_smp += static_cast<double>(sw * m);
 504         }
 505     }
 506
 507     sum_com->sum_cm  = sum_cm;
 508     sum_com->sum_sm  = sum_sm;
 509     sum_com->sum_ccm = sum_ccm;
 510     sum_com->sum_csm = sum_csm;
 511     sum_com->sum_ssm = sum_ssm;
 512     sum_com->sum_cmp = sum_cmp;
 513     sum_com->sum_smp = sum_smp;
 514 }
 515
 516 /* calculates center of mass of selection index from all coordinates x */
 517 void pull_calc_coms(const t_commrec* cr, pull_t* pull, const real* masses, t_pbc* pbc, double t, const rvec x[], rvec* xp)
 518 {
 519     real         twopi_box = 0;
 520     pull_comm_t* comm;
 521
 522     comm = &pull->comm;
 523
 524     GMX_ASSERT(comm->pbcAtomBuffer.size() == pull->group.size(),
 525                "pbcAtomBuffer should have size number of groups");
 526     GMX_ASSERT(comm->comBuffer.size() == pull->group.size() * c_comBufferStride,
 527                "comBuffer should have size #group*c_comBufferStride");
 528
 529     if (pull->bRefAt && pull->bSetPBCatoms)
 530     {
 531         pull_set_pbcatoms(cr, pull, x, comm->pbcAtomBuffer);
 532
 533         if (cr != nullptr && DOMAINDECOMP(cr))
 534         {
 535             /* We can keep these PBC reference coordinates fixed for nstlist
 536              * steps, since atoms won't jump over PBC.
 537              * This avoids a global reduction at the next nstlist-1 steps.
 538              * Note that the exact values of the pbc reference coordinates
 539              * are irrelevant, as long all atoms in the group are within
 540              * half a box distance of the reference coordinate.
 541              */
 542             pull->bSetPBCatoms = FALSE;
 543         }
 544     }
 545
 546     if (pull->cosdim >= 0)
 547     {
 548         int m;
 549
 550         assert(pull->npbcdim <= DIM);
 551
 552         for (m = pull->cosdim + 1; m < pull->npbcdim; m++)
 553         {
 554             if (pbc->box[m][pull->cosdim] != 0)
 555             {
 556                 gmx_fatal(FARGS, "Can not do cosine weighting for trilinic dimensions");
 557             }
 558         }
 559         twopi_box = 2.0 * M_PI / pbc->box[pull->cosdim][pull->cosdim];
 560     }
 561
 562     for (size_t g = 0; g < pull->group.size(); g++)
 563     {
 564         pull_group_work_t* pgrp = &pull->group[g];
 565
 566         /* Cosine-weighted COMs behave different from all other weighted COMs
 567          * in the sense that the weights depend on instantaneous coordinates,
 568          * not on pre-set weights. Thus we resize the local weight buffer here.
 569          */
 570         if (pgrp->epgrppbc == epgrppbcCOS)
 571         {
 572             pgrp->localWeights.resize(pgrp->atomSet.localIndex().size());
 573         }
 574
 575         auto comBuffer = gmx::arrayRefFromArray(comm->comBuffer.data() + g * c_comBufferStride,
 576                                                 c_comBufferStride);
 577
 578         if (pgrp->needToCalcCom)
 579         {
 580             if (pgrp->epgrppbc != epgrppbcCOS)
 581             {
 582                 rvec x_pbc = { 0, 0, 0 };
 583
 584                 switch (pgrp->epgrppbc)
 585                 {
 586                     case epgrppbcREFAT:
 587                         /* Set the pbc atom */
 588                         copy_rvec(comm->pbcAtomBuffer[g], x_pbc);
 589                         break;
 590                     case epgrppbcPREVSTEPCOM:
 591                         /* Set the pbc reference to the COM of the group of the last step */
 592                         copy_dvec_to_rvec(pgrp->x_prev_step, comm->pbcAtomBuffer[g]);
 593                         copy_dvec_to_rvec(pgrp->x_prev_step, x_pbc);
 594                 }
 595
 596                 /* The final sums should end up in comSums[0] */
 597                 ComSums& comSumsTotal = pull->comSums[0];
 598
 599                 /* If we have a single-atom group the mass is irrelevant, so
 600                  * we can remove the mass factor to avoid division by zero.
 601                  * Note that with constraint pulling the mass does matter, but
 602                  * in that case a check group mass != 0 has been done before.
 603                  */
 604                 if (pgrp->params.ind.size() == 1 && pgrp->atomSet.numAtomsLocal() == 1
 605                     && masses[pgrp->atomSet.localIndex()[0]] == 0)
 606                 {
 607                     GMX_ASSERT(xp == nullptr,
 608                                "We should not have groups with zero mass with constraints, i.e. "
 609                                "xp!=NULL");
 610
 611                     /* Copy the single atom coordinate */
 612                     for (int d = 0; d < DIM; d++)
 613                     {
 614                         comSumsTotal.sum_wmx[d] = x[pgrp->atomSet.localIndex()[0]][d];
 615                     }
 616                     /* Set all mass factors to 1 to get the correct COM */
 617                     comSumsTotal.sum_wm  = 1;
 618                     comSumsTotal.sum_wwm = 1;
 619                 }
 620                 else if (pgrp->atomSet.numAtomsLocal() <= c_pullMaxNumLocalAtomsSingleThreaded)
 621                 {
 622                     sum_com_part(pgrp, 0, pgrp->atomSet.numAtomsLocal(), x, xp, masses, pbc, x_pbc, &comSumsTotal);
 623                 }
 624                 else
 625                 {
 626 #pragma omp parallel for num_threads(pull->nthreads) schedule(static)
 627                     for (int t = 0; t < pull->nthreads; t++)
 628                     {
 629                         int ind_start = (pgrp->atomSet.numAtomsLocal() * (t + 0)) / pull->nthreads;
 630                         int ind_end   = (pgrp->atomSet.numAtomsLocal() * (t + 1)) / pull->nthreads;
 631                         sum_com_part(
 632                                 pgrp, ind_start, ind_end, x, xp, masses, pbc, x_pbc, &pull->comSums[t]);
 633                     }
 634
 635                     /* Reduce the thread contributions to sum_com[0] */
 636                     for (int t = 1; t < pull->nthreads; t++)
 637                     {
 638                         comSumsTotal.sum_wm += pull->comSums[t].sum_wm;
 639                         comSumsTotal.sum_wwm += pull->comSums[t].sum_wwm;
 640                         dvec_inc(comSumsTotal.sum_wmx, pull->comSums[t].sum_wmx);
 641                         dvec_inc(comSumsTotal.sum_wmxp, pull->comSums[t].sum_wmxp);
 642                     }
 643                 }
 644
 645                 if (pgrp->localWeights.empty())
 646                 {
 647                     comSumsTotal.sum_wwm = comSumsTotal.sum_wm;
 648                 }
 649
 650                 /* Copy local sums to a buffer for global summing */
 651                 copy_dvec(comSumsTotal.sum_wmx, comBuffer[0]);
 652
 653                 copy_dvec(comSumsTotal.sum_wmxp, comBuffer[1]);
 654
 655                 comBuffer[2][0] = comSumsTotal.sum_wm;
 656                 comBuffer[2][1] = comSumsTotal.sum_wwm;
 657                 comBuffer[2][2] = 0;
 658             }
 659             else
 660             {
 661                 /* Cosine weighting geometry.
 662                  * This uses a slab of the system, thus we always have many
 663                  * atoms in the pull groups. Therefore, always use threads.
 664                  */
 665 #pragma omp parallel for num_threads(pull->nthreads) schedule(static)
 666                 for (int t = 0; t < pull->nthreads; t++)
 667                 {
 668                     int ind_start = (pgrp->atomSet.numAtomsLocal() * (t + 0)) / pull->nthreads;
 669                     int ind_end   = (pgrp->atomSet.numAtomsLocal() * (t + 1)) / pull->nthreads;
 670                     sum_com_part_cosweight(
 671                             pgrp, ind_start, ind_end, pull->cosdim, twopi_box, x, xp, masses, &pull->comSums[t]);
 672                 }
 673
 674                 /* Reduce the thread contributions to comSums[0] */
 675                 ComSums& comSumsTotal = pull->comSums[0];
 676                 for (int t = 1; t < pull->nthreads; t++)
 677                 {
 678                     comSumsTotal.sum_cm += pull->comSums[t].sum_cm;
 679                     comSumsTotal.sum_sm += pull->comSums[t].sum_sm;
 680                     comSumsTotal.sum_ccm += pull->comSums[t].sum_ccm;
 681                     comSumsTotal.sum_csm += pull->comSums[t].sum_csm;
 682                     comSumsTotal.sum_ssm += pull->comSums[t].sum_ssm;
 683                     comSumsTotal.sum_cmp += pull->comSums[t].sum_cmp;
 684                     comSumsTotal.sum_smp += pull->comSums[t].sum_smp;
 685                 }
 686
 687                 /* Copy local sums to a buffer for global summing */
 688                 comBuffer[0][0] = comSumsTotal.sum_cm;
 689                 comBuffer[0][1] = comSumsTotal.sum_sm;
 690                 comBuffer[0][2] = 0;
 691                 comBuffer[1][0] = comSumsTotal.sum_ccm;
 692                 comBuffer[1][1] = comSumsTotal.sum_csm;
 693                 comBuffer[1][2] = comSumsTotal.sum_ssm;
 694                 comBuffer[2][0] = comSumsTotal.sum_cmp;
 695                 comBuffer[2][1] = comSumsTotal.sum_smp;
 696                 comBuffer[2][2] = 0;
 697             }
 698         }
 699         else
 700         {
 701             clear_dvec(comBuffer[0]);
 702             clear_dvec(comBuffer[1]);
 703             clear_dvec(comBuffer[2]);
 704         }
 705     }
 706
 707     pullAllReduce(cr,
 708                   comm,
 709                   pull->group.size() * c_comBufferStride * DIM,
 710                   static_cast<double*>(comm->comBuffer[0]));
 711
 712     for (size_t g = 0; g < pull->group.size(); g++)
 713     {
 714         pull_group_work_t* pgrp;
 715
 716         pgrp = &pull->group[g];
 717         if (pgrp->needToCalcCom)
 718         {
 719             GMX_ASSERT(!pgrp->params.ind.empty(),
 720                        "Normal pull groups should have atoms, only group 0, which should have "
 721                        "bCalcCom=FALSE has nat=0");
 722
 723             const auto comBuffer = gmx::constArrayRefFromArray(
 724                     comm->comBuffer.data() + g * c_comBufferStride, c_comBufferStride);
 725
 726             if (pgrp->epgrppbc != epgrppbcCOS)
 727             {
 728                 double wmass, wwmass;
 729                 int    m;
 730
 731                 /* Determine the inverse mass */
 732                 wmass         = comBuffer[2][0];
 733                 wwmass        = comBuffer[2][1];
 734                 pgrp->mwscale = 1.0 / wmass;
 735                 /* invtm==0 signals a frozen group, so then we should keep it zero */
 736                 if (pgrp->invtm != 0)
 737                 {
 738                     pgrp->wscale = wmass / wwmass;
 739                     pgrp->invtm  = wwmass / (wmass * wmass);
 740                 }
 741                 /* Divide by the total mass */
 742                 for (m = 0; m < DIM; m++)
 743                 {
 744                     pgrp->x[m] = comBuffer[0][m] * pgrp->mwscale;
 745                     if (xp)
 746                     {
 747                         pgrp->xp[m] = comBuffer[1][m] * pgrp->mwscale;
 748                     }
 749                     if (pgrp->epgrppbc == epgrppbcREFAT || pgrp->epgrppbc == epgrppbcPREVSTEPCOM)
 750                     {
 751                         pgrp->x[m] += comm->pbcAtomBuffer[g][m];
 752                         if (xp)
 753                         {
 754                             pgrp->xp[m] += comm->pbcAtomBuffer[g][m];
 755                         }
 756                     }
 757                 }
 758             }
 759             else
 760             {
 761                 /* Cosine weighting geometry */
 762                 double csw, snw, wmass, wwmass;
 763
 764                 /* Determine the optimal location of the cosine weight */
 765                 csw                   = comBuffer[0][0];
 766                 snw                   = comBuffer[0][1];
 767                 pgrp->x[pull->cosdim] = atan2_0_2pi(snw, csw) / twopi_box;
 768                 /* Set the weights for the local atoms */
 769                 wmass  = sqrt(csw * csw + snw * snw);
 770                 wwmass = (comBuffer[1][0] * csw * csw + comBuffer[1][1] * csw * snw
 771                           + comBuffer[1][2] * snw * snw)
 772                          / (wmass * wmass);
 773
 774                 pgrp->mwscale = 1.0 / wmass;
 775                 pgrp->wscale  = wmass / wwmass;
 776                 pgrp->invtm   = wwmass / (wmass * wmass);
 777                 /* Set the weights for the local atoms */
 778                 csw *= pgrp->invtm;
 779                 snw *= pgrp->invtm;
 780                 for (size_t i = 0; i < pgrp->atomSet.numAtomsLocal(); i++)
 781                 {
 782                     int ii                = pgrp->atomSet.localIndex()[i];
 783                     pgrp->localWeights[i] = csw * std::cos(twopi_box * x[ii][pull->cosdim])
 784                                             + snw * std::sin(twopi_box * x[ii][pull->cosdim]);
 785                 }
 786                 if (xp)
 787                 {
 788                     csw                    = comBuffer[2][0];
 789                     snw                    = comBuffer[2][1];
 790                     pgrp->xp[pull->cosdim] = atan2_0_2pi(snw, csw) / twopi_box;
 791                 }
 792             }
 793             if (debug)
 794             {
 795                 fprintf(debug, "Pull group %zu wmass %f invtm %f\n", g, 1.0 / pgrp->mwscale, pgrp->invtm);
 796             }
 797         }
 798     }
 799
 800     if (pull->bCylinder)
 801     {
 802         /* Calculate the COMs for the cyclinder reference groups */
 803         make_cyl_refgrps(cr, pull, masses, pbc, t, x);
 804     }
 805 }
 806
 807 using BoolVec = gmx::BasicVector<bool>;
 808
 809 /* Returns whether the pull group obeys the PBC restrictions */
 810 static bool pullGroupObeysPbcRestrictions(const pull_group_work_t& group,
 811                                           const BoolVec&           dimUsed,
 812                                           const rvec*              x,
 813                                           const t_pbc&             pbc,
 814                                           const gmx::RVec&         x_pbc,
 815                                           const real               pbcMargin)
 816 {
 817     /* Determine which dimensions are relevant for PBC */
 818     BoolVec dimUsesPbc       = { false, false, false };
 819     bool    pbcIsRectangular = true;
 820     for (int d = 0; d < pbc.ndim_ePBC; d++)
 821     {
 822         if (dimUsed[d])
 823         {
 824             dimUsesPbc[d] = true;
 825             /* All non-zero dimensions of vector v are involved in PBC */
 826             for (int d2 = d + 1; d2 < pbc.ndim_ePBC; d2++)
 827             {
 828                 assert(d2 < DIM);
 829                 if (pbc.box[d2][d] != 0)
 830                 {
 831                     dimUsesPbc[d2]   = true;
 832                     pbcIsRectangular = false;
 833                 }
 834             }
 835         }
 836     }
 837
 838     rvec marginPerDim    = {};
 839     real marginDistance2 = 0;
 840     if (pbcIsRectangular)
 841     {
 842         /* Use margins for dimensions independently */
 843         for (int d = 0; d < pbc.ndim_ePBC; d++)
 844         {
 845             marginPerDim[d] = pbcMargin * pbc.hbox_diag[d];
 846         }
 847     }
 848     else
 849     {
 850         /* Check the total distance along the relevant dimensions */
 851         for (int d = 0; d < pbc.ndim_ePBC; d++)
 852         {
 853             if (dimUsesPbc[d])
 854             {
 855                 marginDistance2 += pbcMargin * gmx::square(0.5) * norm2(pbc.box[d]);
 856             }
 857         }
 858     }
 859
 860     auto localAtomIndices = group.atomSet.localIndex();
 861     for (gmx::index indexInSet = 0; indexInSet < localAtomIndices.ssize(); indexInSet++)
 862     {
 863         rvec dx;
 864         pbc_dx(&pbc, x[localAtomIndices[indexInSet]], x_pbc, dx);
 865
 866         bool atomIsTooFar = false;
 867         if (pbcIsRectangular)
 868         {
 869             for (int d = 0; d < pbc.ndim_ePBC; d++)
 870             {
 871                 if (dimUsesPbc[d] && (dx[d] < -marginPerDim[d] || dx[d] > marginPerDim[d]))
 872                 {
 873                     atomIsTooFar = true;
 874                 }
 875             }
 876         }
 877         else
 878         {
 879             real pbcDistance2 = 0;
 880             for (int d = 0; d < pbc.ndim_ePBC; d++)
 881             {
 882                 if (dimUsesPbc[d])
 883                 {
 884                     pbcDistance2 += gmx::square(dx[d]);
 885                 }
 886             }
 887             atomIsTooFar = (pbcDistance2 > marginDistance2);
 888         }
 889         if (atomIsTooFar)
 890         {
 891             return false;
 892         }
 893     }
 894
 895     return true;
 896 }
 897
 898 int pullCheckPbcWithinGroups(const pull_t& pull, const rvec* x, const t_pbc& pbc, real pbcMargin)
 899 {
 900     if (pbc.pbcType == PbcType::No)
 901     {
 902         return -1;
 903     }
 904
 905     /* Determine what dimensions are used for each group by pull coordinates */
 906     std::vector<BoolVec> dimUsed(pull.group.size(), { false, false, false });
 907     for (size_t c = 0; c < pull.coord.size(); c++)
 908     {
 909         const t_pull_coord& coordParams = pull.coord[c].params;
 910         for (int groupIndex = 0; groupIndex < coordParams.ngroup; groupIndex++)
 911         {
 912             for (int d = 0; d < DIM; d++)
 913             {
 914                 if (coordParams.dim[d]
 915                     && !(coordParams.eGeom == PullGroupGeometry::Cylinder && groupIndex == 0))
 916                 {
 917                     dimUsed[coordParams.group[groupIndex]][d] = true;
 918                 }
 919             }
 920         }
 921     }
 922
 923     /* Check PBC for every group that uses a PBC reference atom treatment */
 924     for (size_t g = 0; g < pull.group.size(); g++)
 925     {
 926         const pull_group_work_t& group = pull.group[g];
 927         if ((group.epgrppbc == epgrppbcREFAT || group.epgrppbc == epgrppbcPREVSTEPCOM)
 928             && !pullGroupObeysPbcRestrictions(group, dimUsed[g], x, pbc, pull.comm.pbcAtomBuffer[g], pbcMargin))
 929         {
 930             return g;
 931         }
 932     }
 933
 934     return -1;
 935 }
 936
 937 bool pullCheckPbcWithinGroup(const pull_t&                  pull,
 938                              gmx::ArrayRef<const gmx::RVec> x,
 939                              const t_pbc&                   pbc,
 940                              int                            groupNr,
 941                              real                           pbcMargin)
 942 {
 943     if (pbc.pbcType == PbcType::No)
 944     {
 945         return true;
 946     }
 947     GMX_ASSERT(groupNr < gmx::ssize(pull.group), "groupNr is out of range");
 948
 949     /* Check PBC if the group uses a PBC reference atom treatment. */
 950     const pull_group_work_t& group = pull.group[groupNr];
 951     if (group.epgrppbc != epgrppbcREFAT && group.epgrppbc != epgrppbcPREVSTEPCOM)
 952     {
 953         return true;
 954     }
 955
 956     /* Determine what dimensions are used for each group by pull coordinates */
 957     BoolVec dimUsed = { false, false, false };
 958     for (size_t c = 0; c < pull.coord.size(); c++)
 959     {
 960         const t_pull_coord& coordParams = pull.coord[c].params;
 961         for (int groupIndex = 0; groupIndex < coordParams.ngroup; groupIndex++)
 962         {
 963             if (coordParams.group[groupIndex] == groupNr)
 964             {
 965                 for (int d = 0; d < DIM; d++)
 966                 {
 967                     if (coordParams.dim[d]
 968                         && !(coordParams.eGeom == PullGroupGeometry::Cylinder && groupIndex == 0))
 969                     {
 970                         dimUsed[d] = true;
 971                     }
 972                 }
 973             }
 974         }
 975     }
 976
 977     return (pullGroupObeysPbcRestrictions(
 978             group, dimUsed, as_rvec_array(x.data()), pbc, pull.comm.pbcAtomBuffer[groupNr], pbcMargin));
 979 }
 980
 981 void setPrevStepPullComFromState(struct pull_t* pull, const t_state* state)
 982 {
 983     for (size_t g = 0; g < pull->group.size(); g++)
 984     {
 985         for (int j = 0; j < DIM; j++)
 986         {
 987             pull->group[g].x_prev_step[j] = state->pull_com_prev_step[g * DIM + j];
 988         }
 989     }
 990 }
 991
 992 void updatePrevStepPullCom(struct pull_t* pull, t_state* state)
 993 {
 994     for (size_t g = 0; g < pull->group.size(); g++)
 995     {
 996         if (pull->group[g].needToCalcCom)
 997         {
 998             for (int j = 0; j < DIM; j++)
 999             {
1000                 pull->group[g].x_prev_step[j]          = pull->group[g].x[j];
1001                 state->pull_com_prev_step[g * DIM + j] = pull->group[g].x[j];
1002             }
1003         }
1004     }
1005 }
1006
1007 void allocStatePrevStepPullCom(t_state* state, const pull_t* pull)
1008 {
1009     if (!pull)
1010     {
1011         state->pull_com_prev_step.clear();
1012         return;
1013     }
1014     size_t ngroup = pull->group.size();
1015     if (state->pull_com_prev_step.size() / DIM != ngroup)
1016     {
1017         state->pull_com_prev_step.resize(ngroup * DIM, NAN);
1018     }
1019 }
1020
1021 void initPullComFromPrevStep(const t_commrec* cr, pull_t* pull, const real* masses, t_pbc* pbc, const rvec x[])
1022 {
1023     pull_comm_t* comm   = &pull->comm;
1024     size_t       ngroup = pull->group.size();
1025
1026     if (!comm->bParticipate)
1027     {
1028         return;
1029     }
1030
1031     GMX_ASSERT(comm->pbcAtomBuffer.size() == pull->group.size(),
1032                "pbcAtomBuffer should have size number of groups");
1033     GMX_ASSERT(comm->comBuffer.size() == pull->group.size() * c_comBufferStride,
1034                "comBuffer should have size #group*c_comBufferStride");
1035
1036     pull_set_pbcatoms(cr, pull, x, comm->pbcAtomBuffer);
1037
1038     for (size_t g = 0; g < ngroup; g++)
1039     {
1040         pull_group_work_t* pgrp;
1041
1042         pgrp = &pull->group[g];
1043
1044         if (pgrp->needToCalcCom && pgrp->epgrppbc == epgrppbcPREVSTEPCOM)
1045         {
1046             GMX_ASSERT(pgrp->params.ind.size() > 1,
1047                        "Groups with no atoms, or only one atom, should not "
1048                        "use the COM from the previous step as reference.");
1049
1050             rvec x_pbc = { 0, 0, 0 };
1051             copy_rvec(comm->pbcAtomBuffer[g], x_pbc);
1052
1053             if (debug)
1054             {
1055                 fprintf(debug, "Initialising prev step COM of pull group %zu. x_pbc =", g);
1056                 for (int m = 0; m < DIM; m++)
1057                 {
1058                     fprintf(debug, " %f", x_pbc[m]);
1059                 }
1060                 fprintf(debug, "\n");
1061             }
1062
1063             /* The following is to a large extent similar to pull_calc_coms() */
1064
1065             /* The final sums should end up in sum_com[0] */
1066             ComSums& comSumsTotal = pull->comSums[0];
1067
1068             if (pgrp->atomSet.numAtomsLocal() <= c_pullMaxNumLocalAtomsSingleThreaded)
1069             {
1070                 sum_com_part(pgrp, 0, pgrp->atomSet.numAtomsLocal(), x, nullptr, masses, pbc, x_pbc, &comSumsTotal);
1071             }
1072             else
1073             {
1074 #pragma omp parallel for num_threads(pull->nthreads) schedule(static)
1075                 for (int t = 0; t < pull->nthreads; t++)
1076                 {
1077                     int ind_start = (pgrp->atomSet.numAtomsLocal() * (t + 0)) / pull->nthreads;
1078                     int ind_end   = (pgrp->atomSet.numAtomsLocal() * (t + 1)) / pull->nthreads;
1079                     sum_com_part(
1080                             pgrp, ind_start, ind_end, x, nullptr, masses, pbc, x_pbc, &pull->comSums[t]);
1081                 }
1082
1083                 /* Reduce the thread contributions to sum_com[0] */
1084                 for (int t = 1; t < pull->nthreads; t++)
1085                 {
1086                     comSumsTotal.sum_wm += pull->comSums[t].sum_wm;
1087                     comSumsTotal.sum_wwm += pull->comSums[t].sum_wwm;
1088                     dvec_inc(comSumsTotal.sum_wmx, pull->comSums[t].sum_wmx);
1089                     dvec_inc(comSumsTotal.sum_wmxp, pull->comSums[t].sum_wmxp);
1090                 }
1091             }
1092
1093             if (pgrp->localWeights.empty())
1094             {
1095                 comSumsTotal.sum_wwm = comSumsTotal.sum_wm;
1096             }
1097
1098             /* Copy local sums to a buffer for global summing */
1099             auto localSums = gmx::arrayRefFromArray(comm->comBuffer.data() + g * c_comBufferStride,
1100                                                     c_comBufferStride);
1101
1102             localSums[0]    = comSumsTotal.sum_wmx;
1103             localSums[1]    = comSumsTotal.sum_wmxp;
1104             localSums[2][0] = comSumsTotal.sum_wm;
1105             localSums[2][1] = comSumsTotal.sum_wwm;
1106             localSums[2][2] = 0;
1107         }
1108     }
1109
1110     pullAllReduce(cr, comm, ngroup * c_comBufferStride * DIM, static_cast<double*>(comm->comBuffer[0]));
1111
1112     for (size_t g = 0; g < ngroup; g++)
1113     {
1114         pull_group_work_t* pgrp;
1115
1116         pgrp = &pull->group[g];
1117         if (pgrp->needToCalcCom)
1118         {
1119             if (pgrp->epgrppbc == epgrppbcPREVSTEPCOM)
1120             {
1121                 auto localSums = gmx::arrayRefFromArray(
1122                         comm->comBuffer.data() + g * c_comBufferStride, c_comBufferStride);
1123                 double wmass, wwmass;
1124
1125                 /* Determine the inverse mass */
1126                 wmass         = localSums[2][0];
1127                 wwmass        = localSums[2][1];
1128                 pgrp->mwscale = 1.0 / wmass;
1129                 /* invtm==0 signals a frozen group, so then we should keep it zero */
1130                 if (pgrp->invtm != 0)
1131                 {
1132                     pgrp->wscale = wmass / wwmass;
1133                     pgrp->invtm  = wwmass / (wmass * wmass);
1134                 }
1135                 /* Divide by the total mass */
1136                 for (int m = 0; m < DIM; m++)
1137                 {
1138                     pgrp->x[m] = localSums[0][m] * pgrp->mwscale;
1139                     pgrp->x[m] += comm->pbcAtomBuffer[g][m];
1140                 }
1141                 if (debug)
1142                 {
1143                     fprintf(debug, "Pull group %zu wmass %f invtm %f\n", g, 1.0 / pgrp->mwscale, pgrp->invtm);
1144                     fprintf(debug, "Initialising prev step COM of pull group %zu to", g);
1145                     for (int m = 0; m < DIM; m++)
1146                     {
1147                         fprintf(debug, " %f", pgrp->x[m]);
1148                     }
1149                     fprintf(debug, "\n");
1150                 }
1151                 copy_dvec(pgrp->x, pgrp->x_prev_step);
1152             }
1153         }
1154     }
1155 }