src/gromacs/mdlib/ns.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, by the GROMACS development team, led by
   7  * Mark Abraham, David van der Spoel, Berk Hess, and Erik Lindahl,
   8  * and including many others, as listed in the AUTHORS file in the
   9  * top-level source directory and at http://www.gromacs.org.
  10  *
  11  * GROMACS is free software; you can redistribute it and/or
  12  * modify it under the terms of the GNU Lesser General Public License
  13  * as published by the Free Software Foundation; either version 2.1
  14  * of the License, or (at your option) any later version.
  15  *
  16  * GROMACS is distributed in the hope that it will be useful,
  17  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  18  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  19  * Lesser General Public License for more details.
  20  *
  21  * You should have received a copy of the GNU Lesser General Public
  22  * License along with GROMACS; if not, see
  23  * http://www.gnu.org/licenses, or write to the Free Software Foundation,
  24  * Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA.
  25  *
  26  * If you want to redistribute modifications to GROMACS, please
  27  * consider that scientific software is very special. Version
  28  * control is crucial - bugs must be traceable. We will be happy to
  29  * consider code for inclusion in the official distribution, but
  30  * derived work must not be called official GROMACS. Details are found
  31  * in the README & COPYING files - if they are missing, get the
  32  * official version at http://www.gromacs.org.
  33  *
  34  * To help us fund GROMACS development, we humbly ask that you cite
  35  * the research papers on the package. Check out http://www.gromacs.org.
  36  */
  37 #include "gmxpre.h"
  38
  39 #include "gromacs/legacyheaders/ns.h"
  40
  41 #include <math.h>
  42 #include <stdlib.h>
  43 #include <string.h>
  44
  45 #include <cmath>
  46
  47 #include <algorithm>
  48
  49 #include "gromacs/domdec/domdec.h"
  50 #include "gromacs/legacyheaders/force.h"
  51 #include "gromacs/legacyheaders/macros.h"
  52 #include "gromacs/legacyheaders/names.h"
  53 #include "gromacs/legacyheaders/network.h"
  54 #include "gromacs/legacyheaders/nonbonded.h"
  55 #include "gromacs/legacyheaders/nrnb.h"
  56 #include "gromacs/legacyheaders/nsgrid.h"
  57 #include "gromacs/legacyheaders/txtdump.h"
  58 #include "gromacs/legacyheaders/types/commrec.h"
  59 #include "gromacs/math/utilities.h"
  60 #include "gromacs/math/vec.h"
  61 #include "gromacs/pbcutil/ishift.h"
  62 #include "gromacs/pbcutil/pbc.h"
  63 #include "gromacs/topology/mtop_util.h"
  64 #include "gromacs/utility/fatalerror.h"
  65 #include "gromacs/utility/smalloc.h"
  66
  67 #include "adress.h"
  68
  69 /*
  70  *    E X C L U S I O N   H A N D L I N G
  71  */
  72
  73 #ifdef DEBUG
  74 static void SETEXCL_(t_excl e[], atom_id i, atom_id j)
  75 {
  76     e[j] = e[j] | (1<<i);
  77 }
  78 static void RMEXCL_(t_excl e[], atom_id i, atom_id j)
  79 {
  80     e[j] = e[j] & ~(1<<i);
  81 }
  82 static gmx_bool ISEXCL_(t_excl e[], atom_id i, atom_id j)
  83 {
  84     return (gmx_bool)(e[j] & (1<<i));
  85 }
  86 static gmx_bool NOTEXCL_(t_excl e[], atom_id i, atom_id j)
  87 {
  88     return !(ISEXCL(e, i, j));
  89 }
  90 #else
  91 #define SETEXCL(e, i, j) (e)[((atom_id) (j))] |= (1<<((atom_id) (i)))
  92 #define RMEXCL(e, i, j)  (e)[((atom_id) (j))] &= (~(1<<((atom_id) (i))))
  93 #define ISEXCL(e, i, j)  (gmx_bool) ((e)[((atom_id) (j))] & (1<<((atom_id) (i))))
  94 #define NOTEXCL(e, i, j) !(ISEXCL(e, i, j))
  95 #endif
  96
  97 static int
  98 round_up_to_simd_width(int length, int simd_width)
  99 {
 100     int offset;
 101
 102     offset = (simd_width > 0) ? length % simd_width : 0;
 103
 104     return (offset == 0) ? length : length-offset+simd_width;
 105 }
 106 /************************************************
 107  *
 108  *  U T I L I T I E S    F O R    N S
 109  *
 110  ************************************************/
 111
 112 void reallocate_nblist(t_nblist *nl)
 113 {
 114     if (gmx_debug_at)
 115     {
 116         fprintf(debug, "reallocating neigborlist (ielec=%d, ivdw=%d, igeometry=%d, type=%d), maxnri=%d\n",
 117                 nl->ielec, nl->ivdw, nl->igeometry, nl->type, nl->maxnri);
 118     }
 119     srenew(nl->iinr,   nl->maxnri);
 120     if (nl->igeometry == GMX_NBLIST_GEOMETRY_CG_CG)
 121     {
 122         srenew(nl->iinr_end, nl->maxnri);
 123     }
 124     srenew(nl->gid,    nl->maxnri);
 125     srenew(nl->shift,  nl->maxnri);
 126     srenew(nl->jindex, nl->maxnri+1);
 127 }
 128
 129
 130 static void init_nblist(FILE *log, t_nblist *nl_sr, t_nblist *nl_lr,
 131                         int maxsr, int maxlr,
 132                         int ivdw, int ivdwmod,
 133                         int ielec, int ielecmod,
 134                         int igeometry, int type,
 135                         gmx_bool bElecAndVdwSwitchDiffers)
 136 {
 137     t_nblist *nl;
 138     int       homenr;
 139     int       i;
 140
 141     for (i = 0; (i < 2); i++)
 142     {
 143         nl     = (i == 0) ? nl_sr : nl_lr;
 144         homenr = (i == 0) ? maxsr : maxlr;
 145
 146         if (nl == NULL)
 147         {
 148             continue;
 149         }
 150
 151
 152         /* Set coul/vdw in neighborlist, and for the normal loops we determine
 153          * an index of which one to call.
 154          */
 155         nl->ivdw        = ivdw;
 156         nl->ivdwmod     = ivdwmod;
 157         nl->ielec       = ielec;
 158         nl->ielecmod    = ielecmod;
 159         nl->type        = type;
 160         nl->igeometry   = igeometry;
 161
 162         if (nl->type == GMX_NBLIST_INTERACTION_FREE_ENERGY)
 163         {
 164             nl->igeometry  = GMX_NBLIST_GEOMETRY_PARTICLE_PARTICLE;
 165         }
 166
 167         /* This will also set the simd_padding_width field */
 168         gmx_nonbonded_set_kernel_pointers( (i == 0) ? log : NULL, nl, bElecAndVdwSwitchDiffers);
 169
 170         /* maxnri is influenced by the number of shifts (maximum is 8)
 171          * and the number of energy groups.
 172          * If it is not enough, nl memory will be reallocated during the run.
 173          * 4 seems to be a reasonable factor, which only causes reallocation
 174          * during runs with tiny and many energygroups.
 175          */
 176         nl->maxnri      = homenr*4;
 177         nl->maxnrj      = 0;
 178         nl->nri         = -1;
 179         nl->nrj         = 0;
 180         nl->iinr        = NULL;
 181         nl->gid         = NULL;
 182         nl->shift       = NULL;
 183         nl->jindex      = NULL;
 184         nl->jjnr        = NULL;
 185         nl->excl_fep    = NULL;
 186         reallocate_nblist(nl);
 187         nl->jindex[0] = 0;
 188
 189         if (debug)
 190         {
 191             fprintf(debug, "Initiating neighbourlist (ielec=%d, ivdw=%d, type=%d) for %s interactions,\nwith %d SR, %d LR atoms.\n",
 192                     nl->ielec, nl->ivdw, nl->type, gmx_nblist_geometry_names[nl->igeometry], maxsr, maxlr);
 193         }
 194     }
 195 }
 196
 197 void init_neighbor_list(FILE *log, t_forcerec *fr, int homenr)
 198 {
 199     /* Make maxlr tunable! (does not seem to be a big difference though)
 200      * This parameter determines the number of i particles in a long range
 201      * neighbourlist. Too few means many function calls, too many means
 202      * cache trashing.
 203      */
 204     int        maxsr, maxsr_wat, maxlr, maxlr_wat;
 205     int        ielec, ivdw, ielecmod, ivdwmod, type;
 206     int        igeometry_def, igeometry_w, igeometry_ww;
 207     int        i;
 208     gmx_bool   bElecAndVdwSwitchDiffers;
 209     t_nblists *nbl;
 210
 211     /* maxsr     = homenr-fr->nWatMol*3; */
 212     maxsr     = homenr;
 213
 214     if (maxsr < 0)
 215     {
 216         gmx_fatal(FARGS, "%s, %d: Negative number of short range atoms.\n"
 217                   "Call your GROMACS dealer for assistance.", __FILE__, __LINE__);
 218     }
 219     /* This is just for initial allocation, so we do not reallocate
 220      * all the nlist arrays many times in a row.
 221      * The numbers seem very accurate, but they are uncritical.
 222      */
 223     maxsr_wat = std::min(fr->nWatMol, (homenr+2)/3);
 224     if (fr->bTwinRange)
 225     {
 226         maxlr     = 50;
 227         maxlr_wat = std::min(maxsr_wat, maxlr);
 228     }
 229     else
 230     {
 231         maxlr = maxlr_wat = 0;
 232     }
 233
 234     /* Determine the values for ielec/ivdw. */
 235     ielec                    = fr->nbkernel_elec_interaction;
 236     ivdw                     = fr->nbkernel_vdw_interaction;
 237     ielecmod                 = fr->nbkernel_elec_modifier;
 238     ivdwmod                  = fr->nbkernel_vdw_modifier;
 239     type                     = GMX_NBLIST_INTERACTION_STANDARD;
 240     bElecAndVdwSwitchDiffers = ( (fr->rcoulomb_switch != fr->rvdw_switch) || (fr->rcoulomb != fr->rvdw));
 241
 242     fr->ns.bCGlist = (getenv("GMX_NBLISTCG") != 0);
 243     if (!fr->ns.bCGlist)
 244     {
 245         igeometry_def = GMX_NBLIST_GEOMETRY_PARTICLE_PARTICLE;
 246     }
 247     else
 248     {
 249         igeometry_def = GMX_NBLIST_GEOMETRY_CG_CG;
 250         if (log != NULL)
 251         {
 252             fprintf(log, "\nUsing charge-group - charge-group neighbor lists and kernels\n\n");
 253         }
 254     }
 255
 256     if (fr->solvent_opt == esolTIP4P)
 257     {
 258         igeometry_w  = GMX_NBLIST_GEOMETRY_WATER4_PARTICLE;
 259         igeometry_ww = GMX_NBLIST_GEOMETRY_WATER4_WATER4;
 260     }
 261     else
 262     {
 263         igeometry_w  = GMX_NBLIST_GEOMETRY_WATER3_PARTICLE;
 264         igeometry_ww = GMX_NBLIST_GEOMETRY_WATER3_WATER3;
 265     }
 266
 267     for (i = 0; i < fr->nnblists; i++)
 268     {
 269         nbl = &(fr->nblists[i]);
 270
 271         if ((fr->adress_type != eAdressOff) && (i >= fr->nnblists/2))
 272         {
 273             type = GMX_NBLIST_INTERACTION_ADRESS;
 274         }
 275         init_nblist(log, &nbl->nlist_sr[eNL_VDWQQ], &nbl->nlist_lr[eNL_VDWQQ],
 276                     maxsr, maxlr, ivdw, ivdwmod, ielec, ielecmod, igeometry_def, type, bElecAndVdwSwitchDiffers);
 277         init_nblist(log, &nbl->nlist_sr[eNL_VDW], &nbl->nlist_lr[eNL_VDW],
 278                     maxsr, maxlr, ivdw, ivdwmod, GMX_NBKERNEL_ELEC_NONE, eintmodNONE, igeometry_def, type, bElecAndVdwSwitchDiffers);
 279         init_nblist(log, &nbl->nlist_sr[eNL_QQ], &nbl->nlist_lr[eNL_QQ],
 280                     maxsr, maxlr, GMX_NBKERNEL_VDW_NONE, eintmodNONE, ielec, ielecmod, igeometry_def, type, bElecAndVdwSwitchDiffers);
 281         init_nblist(log, &nbl->nlist_sr[eNL_VDWQQ_WATER], &nbl->nlist_lr[eNL_VDWQQ_WATER],
 282                     maxsr_wat, maxlr_wat, ivdw, ivdwmod, ielec, ielecmod, igeometry_w, type, bElecAndVdwSwitchDiffers);
 283         init_nblist(log, &nbl->nlist_sr[eNL_QQ_WATER], &nbl->nlist_lr[eNL_QQ_WATER],
 284                     maxsr_wat, maxlr_wat, GMX_NBKERNEL_VDW_NONE, eintmodNONE, ielec, ielecmod, igeometry_w, type, bElecAndVdwSwitchDiffers);
 285         init_nblist(log, &nbl->nlist_sr[eNL_VDWQQ_WATERWATER], &nbl->nlist_lr[eNL_VDWQQ_WATERWATER],
 286                     maxsr_wat, maxlr_wat, ivdw, ivdwmod, ielec, ielecmod, igeometry_ww, type, bElecAndVdwSwitchDiffers);
 287         init_nblist(log, &nbl->nlist_sr[eNL_QQ_WATERWATER], &nbl->nlist_lr[eNL_QQ_WATERWATER],
 288                     maxsr_wat, maxlr_wat, GMX_NBKERNEL_VDW_NONE, eintmodNONE, ielec, ielecmod, igeometry_ww, type, bElecAndVdwSwitchDiffers);
 289
 290         /* Did we get the solvent loops so we can use optimized water kernels? */
 291         if (nbl->nlist_sr[eNL_VDWQQ_WATER].kernelptr_vf == NULL
 292             || nbl->nlist_sr[eNL_QQ_WATER].kernelptr_vf == NULL
 293             || nbl->nlist_sr[eNL_VDWQQ_WATERWATER].kernelptr_vf == NULL
 294             || nbl->nlist_sr[eNL_QQ_WATERWATER].kernelptr_vf == NULL)
 295         {
 296             fr->solvent_opt = esolNO;
 297             if (log != NULL)
 298             {
 299                 fprintf(log, "Note: The available nonbonded kernels do not support water optimization - disabling.\n");
 300             }
 301         }
 302
 303         if (fr->efep != efepNO)
 304         {
 305             init_nblist(log, &nbl->nlist_sr[eNL_VDWQQ_FREE], &nbl->nlist_lr[eNL_VDWQQ_FREE],
 306                         maxsr, maxlr, ivdw, ivdwmod, ielec, ielecmod, GMX_NBLIST_GEOMETRY_PARTICLE_PARTICLE, GMX_NBLIST_INTERACTION_FREE_ENERGY, bElecAndVdwSwitchDiffers);
 307             init_nblist(log, &nbl->nlist_sr[eNL_VDW_FREE], &nbl->nlist_lr[eNL_VDW_FREE],
 308                         maxsr, maxlr, ivdw, ivdwmod, GMX_NBKERNEL_ELEC_NONE, eintmodNONE, GMX_NBLIST_GEOMETRY_PARTICLE_PARTICLE, GMX_NBLIST_INTERACTION_FREE_ENERGY, bElecAndVdwSwitchDiffers);
 309             init_nblist(log, &nbl->nlist_sr[eNL_QQ_FREE], &nbl->nlist_lr[eNL_QQ_FREE],
 310                         maxsr, maxlr, GMX_NBKERNEL_VDW_NONE, eintmodNONE, ielec, ielecmod, GMX_NBLIST_GEOMETRY_PARTICLE_PARTICLE, GMX_NBLIST_INTERACTION_FREE_ENERGY, bElecAndVdwSwitchDiffers);
 311         }
 312     }
 313     /* QMMM MM list */
 314     if (fr->bQMMM && fr->qr->QMMMscheme != eQMMMschemeoniom)
 315     {
 316         init_nblist(log, &fr->QMMMlist, NULL,
 317                     maxsr, maxlr, 0, 0, ielec, ielecmod, GMX_NBLIST_GEOMETRY_PARTICLE_PARTICLE, GMX_NBLIST_INTERACTION_STANDARD, bElecAndVdwSwitchDiffers);
 318     }
 319
 320     if (log != NULL)
 321     {
 322         fprintf(log, "\n");
 323     }
 324
 325     fr->ns.nblist_initialized = TRUE;
 326 }
 327
 328 static void reset_nblist(t_nblist *nl)
 329 {
 330     nl->nri       = -1;
 331     nl->nrj       = 0;
 332     if (nl->jindex)
 333     {
 334         nl->jindex[0] = 0;
 335     }
 336 }
 337
 338 static void reset_neighbor_lists(t_forcerec *fr, gmx_bool bResetSR, gmx_bool bResetLR)
 339 {
 340     int n, i;
 341
 342     if (fr->bQMMM)
 343     {
 344         /* only reset the short-range nblist */
 345         reset_nblist(&(fr->QMMMlist));
 346     }
 347
 348     for (n = 0; n < fr->nnblists; n++)
 349     {
 350         for (i = 0; i < eNL_NR; i++)
 351         {
 352             if (bResetSR)
 353             {
 354                 reset_nblist( &(fr->nblists[n].nlist_sr[i]) );
 355             }
 356             if (bResetLR)
 357             {
 358                 reset_nblist( &(fr->nblists[n].nlist_lr[i]) );
 359             }
 360         }
 361     }
 362 }
 363
 364
 365
 366
 367 static gmx_inline void new_i_nblist(t_nblist *nlist, atom_id i_atom, int shift, int gid)
 368 {
 369     int    nri = nlist->nri;
 370
 371     /* Check whether we have to increase the i counter */
 372     if ((nri == -1) ||
 373         (nlist->iinr[nri]  != i_atom) ||
 374         (nlist->shift[nri] != shift) ||
 375         (nlist->gid[nri]   != gid))
 376     {
 377         /* This is something else. Now see if any entries have
 378          * been added in the list of the previous atom.
 379          */
 380         if ((nri == -1) ||
 381             ((nlist->jindex[nri+1] > nlist->jindex[nri]) &&
 382              (nlist->gid[nri] != -1)))
 383         {
 384             /* If so increase the counter */
 385             nlist->nri++;
 386             nri++;
 387             if (nlist->nri >= nlist->maxnri)
 388             {
 389                 nlist->maxnri += over_alloc_large(nlist->nri);
 390                 reallocate_nblist(nlist);
 391             }
 392         }
 393         /* Set the number of neighbours and the atom number */
 394         nlist->jindex[nri+1] = nlist->jindex[nri];
 395         nlist->iinr[nri]     = i_atom;
 396         nlist->gid[nri]      = gid;
 397         nlist->shift[nri]    = shift;
 398     }
 399     else
 400     {
 401         /* Adding to previous list. First remove possible previous padding */
 402         if (nlist->simd_padding_width > 1)
 403         {
 404             while (nlist->nrj > 0 && nlist->jjnr[nlist->nrj-1] < 0)
 405             {
 406                 nlist->nrj--;
 407             }
 408         }
 409     }
 410 }
 411
 412 static gmx_inline void close_i_nblist(t_nblist *nlist)
 413 {
 414     int nri = nlist->nri;
 415     int len;
 416
 417     if (nri >= 0)
 418     {
 419         /* Add elements up to padding. Since we allocate memory in units
 420          * of the simd_padding width, we do not have to check for possible
 421          * list reallocation here.
 422          */
 423         while ((nlist->nrj % nlist->simd_padding_width) != 0)
 424         {
 425             /* Use -4 here, so we can write forces for 4 atoms before real data */
 426             nlist->jjnr[nlist->nrj++] = -4;
 427         }
 428         nlist->jindex[nri+1] = nlist->nrj;
 429
 430         len = nlist->nrj -  nlist->jindex[nri];
 431         /* If there are no j-particles we have to reduce the
 432          * number of i-particles again, to prevent errors in the
 433          * kernel functions.
 434          */
 435         if ((len == 0) && (nlist->nri > 0))
 436         {
 437             nlist->nri--;
 438         }
 439     }
 440 }
 441
 442 static gmx_inline void close_nblist(t_nblist *nlist)
 443 {
 444     /* Only close this nblist when it has been initialized.
 445      * Avoid the creation of i-lists with no j-particles.
 446      */
 447     if (nlist->nrj == 0)
 448     {
 449         /* Some assembly kernels do not support empty lists,
 450          * make sure here that we don't generate any empty lists.
 451          * With the current ns code this branch is taken in two cases:
 452          * No i-particles at all: nri=-1 here
 453          * There are i-particles, but no j-particles; nri=0 here
 454          */
 455         nlist->nri = 0;
 456     }
 457     else
 458     {
 459         /* Close list number nri by incrementing the count */
 460         nlist->nri++;
 461     }
 462 }
 463
 464 static gmx_inline void close_neighbor_lists(t_forcerec *fr, gmx_bool bMakeQMMMnblist)
 465 {
 466     int n, i;
 467
 468     if (bMakeQMMMnblist)
 469     {
 470         close_nblist(&(fr->QMMMlist));
 471     }
 472
 473     for (n = 0; n < fr->nnblists; n++)
 474     {
 475         for (i = 0; (i < eNL_NR); i++)
 476         {
 477             close_nblist(&(fr->nblists[n].nlist_sr[i]));
 478             close_nblist(&(fr->nblists[n].nlist_lr[i]));
 479         }
 480     }
 481 }
 482
 483
 484 static gmx_inline void add_j_to_nblist(t_nblist *nlist, atom_id j_atom, gmx_bool bLR)
 485 {
 486     int nrj = nlist->nrj;
 487
 488     if (nlist->nrj >= nlist->maxnrj)
 489     {
 490         nlist->maxnrj = round_up_to_simd_width(over_alloc_small(nlist->nrj + 1), nlist->simd_padding_width);
 491
 492         if (gmx_debug_at)
 493         {
 494             fprintf(debug, "Increasing %s nblist (ielec=%d,ivdw=%d,type=%d,igeometry=%d) j size to %d\n",
 495                     bLR ? "LR" : "SR", nlist->ielec, nlist->ivdw, nlist->type, nlist->igeometry, nlist->maxnrj);
 496         }
 497
 498         srenew(nlist->jjnr, nlist->maxnrj);
 499     }
 500
 501     nlist->jjnr[nrj] = j_atom;
 502     nlist->nrj++;
 503 }
 504
 505 static gmx_inline void add_j_to_nblist_cg(t_nblist *nlist,
 506                                           atom_id j_start, int j_end,
 507                                           t_excl *bexcl, gmx_bool i_is_j,
 508                                           gmx_bool bLR)
 509 {
 510     int nrj = nlist->nrj;
 511     int j;
 512
 513     if (nlist->nrj >= nlist->maxnrj)
 514     {
 515         nlist->maxnrj = over_alloc_small(nlist->nrj + 1);
 516         if (gmx_debug_at)
 517         {
 518             fprintf(debug, "Increasing %s nblist (ielec=%d,ivdw=%d,type=%d,igeometry=%d) j size to %d\n",
 519                     bLR ? "LR" : "SR", nlist->ielec, nlist->ivdw, nlist->type, nlist->igeometry, nlist->maxnrj);
 520         }
 521
 522         srenew(nlist->jjnr, nlist->maxnrj);
 523         srenew(nlist->jjnr_end, nlist->maxnrj);
 524         srenew(nlist->excl, nlist->maxnrj*MAX_CGCGSIZE);
 525     }
 526
 527     nlist->jjnr[nrj]     = j_start;
 528     nlist->jjnr_end[nrj] = j_end;
 529
 530     if (j_end - j_start > MAX_CGCGSIZE)
 531     {
 532         gmx_fatal(FARGS, "The charge-group - charge-group neighborlist do not support charge groups larger than %d, found a charge group of size %d", MAX_CGCGSIZE, j_end-j_start);
 533     }
 534
 535     /* Set the exclusions */
 536     for (j = j_start; j < j_end; j++)
 537     {
 538         nlist->excl[nrj*MAX_CGCGSIZE + j - j_start] = bexcl[j];
 539     }
 540     if (i_is_j)
 541     {
 542         /* Avoid double counting of intra-cg interactions */
 543         for (j = 1; j < j_end-j_start; j++)
 544         {
 545             nlist->excl[nrj*MAX_CGCGSIZE + j] |= (1<<j) - 1;
 546         }
 547     }
 548
 549     nlist->nrj++;
 550 }
 551
 552 typedef void
 553     put_in_list_t (gmx_bool              bHaveVdW[],
 554                    int                   ngid,
 555                    t_mdatoms     *       md,
 556                    int                   icg,
 557                    int                   jgid,
 558                    int                   nj,
 559                    atom_id               jjcg[],
 560                    atom_id               index[],
 561                    t_excl                bExcl[],
 562                    int                   shift,
 563                    t_forcerec     *      fr,
 564                    gmx_bool              bLR,
 565                    gmx_bool              bDoVdW,
 566                    gmx_bool              bDoCoul,
 567                    int                   solvent_opt);
 568
 569 static void
 570 put_in_list_at(gmx_bool              bHaveVdW[],
 571                int                   ngid,
 572                t_mdatoms     *       md,
 573                int                   icg,
 574                int                   jgid,
 575                int                   nj,
 576                atom_id               jjcg[],
 577                atom_id               index[],
 578                t_excl                bExcl[],
 579                int                   shift,
 580                t_forcerec     *      fr,
 581                gmx_bool              bLR,
 582                gmx_bool              bDoVdW,
 583                gmx_bool              bDoCoul,
 584                int                   solvent_opt)
 585 {
 586     /* The a[] index has been removed,
 587      * to put it back in i_atom should be a[i0] and jj should be a[jj].
 588      */
 589     t_nblist  *   vdwc;
 590     t_nblist  *   vdw;
 591     t_nblist  *   coul;
 592     t_nblist  *   vdwc_free  = NULL;
 593     t_nblist  *   vdw_free   = NULL;
 594     t_nblist  *   coul_free  = NULL;
 595     t_nblist  *   vdwc_ww    = NULL;
 596     t_nblist  *   coul_ww    = NULL;
 597
 598     int           i, j, jcg, igid, gid, nbl_ind;
 599     atom_id       jj, jj0, jj1, i_atom;
 600     int           i0, nicg;
 601
 602     int          *cginfo;
 603     int          *type, *typeB;
 604     real         *charge, *chargeB;
 605     real          qi, qiB;
 606     gmx_bool      bFreeEnergy, bFree, bFreeJ, bNotEx, *bPert;
 607     gmx_bool      bDoVdW_i, bDoCoul_i, bDoCoul_i_sol;
 608     int           iwater, jwater;
 609     t_nblist     *nlist;
 610
 611     /* Copy some pointers */
 612     cginfo  = fr->cginfo;
 613     charge  = md->chargeA;
 614     chargeB = md->chargeB;
 615     type    = md->typeA;
 616     typeB   = md->typeB;
 617     bPert   = md->bPerturbed;
 618
 619     /* Get atom range */
 620     i0     = index[icg];
 621     nicg   = index[icg+1]-i0;
 622
 623     /* Get the i charge group info */
 624     igid   = GET_CGINFO_GID(cginfo[icg]);
 625
 626     iwater = (solvent_opt != esolNO) ? GET_CGINFO_SOLOPT(cginfo[icg]) : esolNO;
 627
 628     bFreeEnergy = FALSE;
 629     if (md->nPerturbed)
 630     {
 631         /* Check if any of the particles involved are perturbed.
 632          * If not we can do the cheaper normal put_in_list
 633          * and use more solvent optimization.
 634          */
 635         for (i = 0; i < nicg; i++)
 636         {
 637             bFreeEnergy |= bPert[i0+i];
 638         }
 639         /* Loop over the j charge groups */
 640         for (j = 0; (j < nj && !bFreeEnergy); j++)
 641         {
 642             jcg = jjcg[j];
 643             jj0 = index[jcg];
 644             jj1 = index[jcg+1];
 645             /* Finally loop over the atoms in the j-charge group */
 646             for (jj = jj0; jj < jj1; jj++)
 647             {
 648                 bFreeEnergy |= bPert[jj];
 649             }
 650         }
 651     }
 652
 653     /* Unpack pointers to neighbourlist structs */
 654     if (fr->nnblists == 1)
 655     {
 656         nbl_ind = 0;
 657     }
 658     else
 659     {
 660         nbl_ind = fr->gid2nblists[GID(igid, jgid, ngid)];
 661     }
 662     if (bLR)
 663     {
 664         nlist = fr->nblists[nbl_ind].nlist_lr;
 665     }
 666     else
 667     {
 668         nlist = fr->nblists[nbl_ind].nlist_sr;
 669     }
 670
 671     if (iwater != esolNO)
 672     {
 673         vdwc    = &nlist[eNL_VDWQQ_WATER];
 674         vdw     = &nlist[eNL_VDW];
 675         coul    = &nlist[eNL_QQ_WATER];
 676         vdwc_ww = &nlist[eNL_VDWQQ_WATERWATER];
 677         coul_ww = &nlist[eNL_QQ_WATERWATER];
 678     }
 679     else
 680     {
 681         vdwc = &nlist[eNL_VDWQQ];
 682         vdw  = &nlist[eNL_VDW];
 683         coul = &nlist[eNL_QQ];
 684     }
 685
 686     if (!bFreeEnergy)
 687     {
 688         if (iwater != esolNO)
 689         {
 690             /* Loop over the atoms in the i charge group */
 691             i_atom  = i0;
 692             gid     = GID(igid, jgid, ngid);
 693             /* Create new i_atom for each energy group */
 694             if (bDoCoul && bDoVdW)
 695             {
 696                 new_i_nblist(vdwc, i_atom, shift, gid);
 697                 new_i_nblist(vdwc_ww, i_atom, shift, gid);
 698             }
 699             if (bDoVdW)
 700             {
 701                 new_i_nblist(vdw, i_atom, shift, gid);
 702             }
 703             if (bDoCoul)
 704             {
 705                 new_i_nblist(coul, i_atom, shift, gid);
 706                 new_i_nblist(coul_ww, i_atom, shift, gid);
 707             }
 708             /* Loop over the j charge groups */
 709             for (j = 0; (j < nj); j++)
 710             {
 711                 jcg = jjcg[j];
 712
 713                 if (jcg == icg)
 714                 {
 715                     continue;
 716                 }
 717
 718                 jj0    = index[jcg];
 719                 jwater = GET_CGINFO_SOLOPT(cginfo[jcg]);
 720
 721                 if (iwater == esolSPC && jwater == esolSPC)
 722                 {
 723                     /* Interaction between two SPC molecules */
 724                     if (!bDoCoul)
 725                     {
 726                         /* VdW only - only first atoms in each water interact */
 727                         add_j_to_nblist(vdw, jj0, bLR);
 728                     }
 729                     else
 730                     {
 731                         /* One entry for the entire water-water interaction */
 732                         if (!bDoVdW)
 733                         {
 734                             add_j_to_nblist(coul_ww, jj0, bLR);
 735                         }
 736                         else
 737                         {
 738                             add_j_to_nblist(vdwc_ww, jj0, bLR);
 739                         }
 740                     }
 741                 }
 742                 else if (iwater == esolTIP4P && jwater == esolTIP4P)
 743                 {
 744                     /* Interaction between two TIP4p molecules */
 745                     if (!bDoCoul)
 746                     {
 747                         /* VdW only - only first atoms in each water interact */
 748                         add_j_to_nblist(vdw, jj0, bLR);
 749                     }
 750                     else
 751                     {
 752                         /* One entry for the entire water-water interaction */
 753                         if (!bDoVdW)
 754                         {
 755                             add_j_to_nblist(coul_ww, jj0, bLR);
 756                         }
 757                         else
 758                         {
 759                             add_j_to_nblist(vdwc_ww, jj0, bLR);
 760                         }
 761                     }
 762                 }
 763                 else
 764                 {
 765                     /* j charge group is not water, but i is.
 766                      * Add entries to the water-other_atom lists; the geometry of the water
 767                      * molecule doesn't matter - that is taken care of in the nonbonded kernel,
 768                      * so we don't care if it is SPC or TIP4P...
 769                      */
 770
 771                     jj1 = index[jcg+1];
 772
 773                     if (!bDoVdW)
 774                     {
 775                         for (jj = jj0; (jj < jj1); jj++)
 776                         {
 777                             if (charge[jj] != 0)
 778                             {
 779                                 add_j_to_nblist(coul, jj, bLR);
 780                             }
 781                         }
 782                     }
 783                     else if (!bDoCoul)
 784                     {
 785                         for (jj = jj0; (jj < jj1); jj++)
 786                         {
 787                             if (bHaveVdW[type[jj]])
 788                             {
 789                                 add_j_to_nblist(vdw, jj, bLR);
 790                             }
 791                         }
 792                     }
 793                     else
 794                     {
 795                         /* _charge_ _groups_ interact with both coulomb and LJ */
 796                         /* Check which atoms we should add to the lists!       */
 797                         for (jj = jj0; (jj < jj1); jj++)
 798                         {
 799                             if (bHaveVdW[type[jj]])
 800                             {
 801                                 if (charge[jj] != 0)
 802                                 {
 803                                     add_j_to_nblist(vdwc, jj, bLR);
 804                                 }
 805                                 else
 806                                 {
 807                                     add_j_to_nblist(vdw, jj, bLR);
 808                                 }
 809                             }
 810                             else if (charge[jj] != 0)
 811                             {
 812                                 add_j_to_nblist(coul, jj, bLR);
 813                             }
 814                         }
 815                     }
 816                 }
 817             }
 818             close_i_nblist(vdw);
 819             close_i_nblist(coul);
 820             close_i_nblist(vdwc);
 821             close_i_nblist(coul_ww);
 822             close_i_nblist(vdwc_ww);
 823         }
 824         else
 825         {
 826             /* no solvent as i charge group */
 827             /* Loop over the atoms in the i charge group */
 828             for (i = 0; i < nicg; i++)
 829             {
 830                 i_atom  = i0+i;
 831                 gid     = GID(igid, jgid, ngid);
 832                 qi      = charge[i_atom];
 833
 834                 /* Create new i_atom for each energy group */
 835                 if (bDoVdW && bDoCoul)
 836                 {
 837                     new_i_nblist(vdwc, i_atom, shift, gid);
 838                 }
 839                 if (bDoVdW)
 840                 {
 841                     new_i_nblist(vdw, i_atom, shift, gid);
 842                 }
 843                 if (bDoCoul)
 844                 {
 845                     new_i_nblist(coul, i_atom, shift, gid);
 846                 }
 847                 bDoVdW_i  = (bDoVdW  && bHaveVdW[type[i_atom]]);
 848                 bDoCoul_i = (bDoCoul && qi != 0);
 849
 850                 if (bDoVdW_i || bDoCoul_i)
 851                 {
 852                     /* Loop over the j charge groups */
 853                     for (j = 0; (j < nj); j++)
 854                     {
 855                         jcg = jjcg[j];
 856
 857                         /* Check for large charge groups */
 858                         if (jcg == icg)
 859                         {
 860                             jj0 = i0 + i + 1;
 861                         }
 862                         else
 863                         {
 864                             jj0 = index[jcg];
 865                         }
 866
 867                         jj1 = index[jcg+1];
 868                         /* Finally loop over the atoms in the j-charge group */
 869                         for (jj = jj0; jj < jj1; jj++)
 870                         {
 871                             bNotEx = NOTEXCL(bExcl, i, jj);
 872
 873                             if (bNotEx)
 874                             {
 875                                 if (!bDoVdW_i)
 876                                 {
 877                                     if (charge[jj] != 0)
 878                                     {
 879                                         add_j_to_nblist(coul, jj, bLR);
 880                                     }
 881                                 }
 882                                 else if (!bDoCoul_i)
 883                                 {
 884                                     if (bHaveVdW[type[jj]])
 885                                     {
 886                                         add_j_to_nblist(vdw, jj, bLR);
 887                                     }
 888                                 }
 889                                 else
 890                                 {
 891                                     if (bHaveVdW[type[jj]])
 892                                     {
 893                                         if (charge[jj] != 0)
 894                                         {
 895                                             add_j_to_nblist(vdwc, jj, bLR);
 896                                         }
 897                                         else
 898                                         {
 899                                             add_j_to_nblist(vdw, jj, bLR);
 900                                         }
 901                                     }
 902                                     else if (charge[jj] != 0)
 903                                     {
 904                                         add_j_to_nblist(coul, jj, bLR);
 905                                     }
 906                                 }
 907                             }
 908                         }
 909                     }
 910                 }
 911                 close_i_nblist(vdw);
 912                 close_i_nblist(coul);
 913                 close_i_nblist(vdwc);
 914             }
 915         }
 916     }
 917     else
 918     {
 919         /* we are doing free energy */
 920         vdwc_free = &nlist[eNL_VDWQQ_FREE];
 921         vdw_free  = &nlist[eNL_VDW_FREE];
 922         coul_free = &nlist[eNL_QQ_FREE];
 923         /* Loop over the atoms in the i charge group */
 924         for (i = 0; i < nicg; i++)
 925         {
 926             i_atom  = i0+i;
 927             gid     = GID(igid, jgid, ngid);
 928             qi      = charge[i_atom];
 929             qiB     = chargeB[i_atom];
 930
 931             /* Create new i_atom for each energy group */
 932             if (bDoVdW && bDoCoul)
 933             {
 934                 new_i_nblist(vdwc, i_atom, shift, gid);
 935             }
 936             if (bDoVdW)
 937             {
 938                 new_i_nblist(vdw, i_atom, shift, gid);
 939             }
 940             if (bDoCoul)
 941             {
 942                 new_i_nblist(coul, i_atom, shift, gid);
 943             }
 944
 945             new_i_nblist(vdw_free, i_atom, shift, gid);
 946             new_i_nblist(coul_free, i_atom, shift, gid);
 947             new_i_nblist(vdwc_free, i_atom, shift, gid);
 948
 949             bDoVdW_i  = (bDoVdW  &&
 950                          (bHaveVdW[type[i_atom]] || bHaveVdW[typeB[i_atom]]));
 951             bDoCoul_i = (bDoCoul && (qi != 0 || qiB != 0));
 952             /* For TIP4P the first atom does not have a charge,
 953              * but the last three do. So we should still put an atom
 954              * without LJ but with charge in the water-atom neighborlist
 955              * for a TIP4p i charge group.
 956              * For SPC type water the first atom has LJ and charge,
 957              * so there is no such problem.
 958              */
 959             if (iwater == esolNO)
 960             {
 961                 bDoCoul_i_sol = bDoCoul_i;
 962             }
 963             else
 964             {
 965                 bDoCoul_i_sol = bDoCoul;
 966             }
 967
 968             if (bDoVdW_i || bDoCoul_i_sol)
 969             {
 970                 /* Loop over the j charge groups */
 971                 for (j = 0; (j < nj); j++)
 972                 {
 973                     jcg = jjcg[j];
 974
 975                     /* Check for large charge groups */
 976                     if (jcg == icg)
 977                     {
 978                         jj0 = i0 + i + 1;
 979                     }
 980                     else
 981                     {
 982                         jj0 = index[jcg];
 983                     }
 984
 985                     jj1 = index[jcg+1];
 986                     /* Finally loop over the atoms in the j-charge group */
 987                     bFree = bPert[i_atom];
 988                     for (jj = jj0; (jj < jj1); jj++)
 989                     {
 990                         bFreeJ = bFree || bPert[jj];
 991                         /* Complicated if, because the water H's should also
 992                          * see perturbed j-particles
 993                          */
 994                         if (iwater == esolNO || i == 0 || bFreeJ)
 995                         {
 996                             bNotEx = NOTEXCL(bExcl, i, jj);
 997
 998                             if (bNotEx)
 999                             {
1000                                 if (bFreeJ)
1001                                 {
1002                                     if (!bDoVdW_i)
1003                                     {
1004                                         if (charge[jj] != 0 || chargeB[jj] != 0)
1005                                         {
1006                                             add_j_to_nblist(coul_free, jj, bLR);
1007                                         }
1008                                     }
1009                                     else if (!bDoCoul_i)
1010                                     {
1011                                         if (bHaveVdW[type[jj]] || bHaveVdW[typeB[jj]])
1012                                         {
1013                                             add_j_to_nblist(vdw_free, jj, bLR);
1014                                         }
1015                                     }
1016                                     else
1017                                     {
1018                                         if (bHaveVdW[type[jj]] || bHaveVdW[typeB[jj]])
1019                                         {
1020                                             if (charge[jj] != 0 || chargeB[jj] != 0)
1021                                             {
1022                                                 add_j_to_nblist(vdwc_free, jj, bLR);
1023                                             }
1024                                             else
1025                                             {
1026                                                 add_j_to_nblist(vdw_free, jj, bLR);
1027                                             }
1028                                         }
1029                                         else if (charge[jj] != 0 || chargeB[jj] != 0)
1030                                         {
1031                                             add_j_to_nblist(coul_free, jj, bLR);
1032                                         }
1033                                     }
1034                                 }
1035                                 else if (!bDoVdW_i)
1036                                 {
1037                                     /* This is done whether or not bWater is set */
1038                                     if (charge[jj] != 0)
1039                                     {
1040                                         add_j_to_nblist(coul, jj, bLR);
1041                                     }
1042                                 }
1043                                 else if (!bDoCoul_i_sol)
1044                                 {
1045                                     if (bHaveVdW[type[jj]])
1046                                     {
1047                                         add_j_to_nblist(vdw, jj, bLR);
1048                                     }
1049                                 }
1050                                 else
1051                                 {
1052                                     if (bHaveVdW[type[jj]])
1053                                     {
1054                                         if (charge[jj] != 0)
1055                                         {
1056                                             add_j_to_nblist(vdwc, jj, bLR);
1057                                         }
1058                                         else
1059                                         {
1060                                             add_j_to_nblist(vdw, jj, bLR);
1061                                         }
1062                                     }
1063                                     else if (charge[jj] != 0)
1064                                     {
1065                                         add_j_to_nblist(coul, jj, bLR);
1066                                     }
1067                                 }
1068                             }
1069                         }
1070                     }
1071                 }
1072             }
1073             close_i_nblist(vdw);
1074             close_i_nblist(coul);
1075             close_i_nblist(vdwc);
1076             close_i_nblist(vdw_free);
1077             close_i_nblist(coul_free);
1078             close_i_nblist(vdwc_free);
1079         }
1080     }
1081 }
1082
1083 static void
1084 put_in_list_adress(gmx_bool              bHaveVdW[],
1085                    int                   ngid,
1086                    t_mdatoms     *       md,
1087                    int                   icg,
1088                    int                   jgid,
1089                    int                   nj,
1090                    atom_id               jjcg[],
1091                    atom_id               index[],
1092                    t_excl                bExcl[],
1093                    int                   shift,
1094                    t_forcerec     *      fr,
1095                    gmx_bool              bLR,
1096                    gmx_bool              bDoVdW,
1097                    gmx_bool              bDoCoul,
1098                    int       gmx_unused  solvent_opt)
1099 {
1100     /* The a[] index has been removed,
1101      * to put it back in i_atom should be a[i0] and jj should be a[jj].
1102      */
1103     t_nblist  *   vdwc;
1104     t_nblist  *   vdw;
1105     t_nblist  *   coul;
1106     t_nblist  *   vdwc_adress  = NULL;
1107     t_nblist  *   vdw_adress   = NULL;
1108     t_nblist  *   coul_adress  = NULL;
1109
1110     int           i, j, jcg, igid, gid, nbl_ind, nbl_ind_adress;
1111     atom_id       jj, jj0, jj1, i_atom;
1112     int           i0, nicg;
1113
1114     int          *cginfo;
1115     int          *type;
1116     real         *charge;
1117     real         *wf;
1118     real          qi;
1119     gmx_bool      bNotEx;
1120     gmx_bool      bDoVdW_i, bDoCoul_i;
1121     gmx_bool      b_hybrid;
1122     t_nblist     *nlist, *nlist_adress;
1123     gmx_bool      bEnergyGroupCG;
1124
1125     /* Copy some pointers */
1126     cginfo  = fr->cginfo;
1127     charge  = md->chargeA;
1128     type    = md->typeA;
1129     wf      = md->wf;
1130
1131     /* Get atom range */
1132     i0     = index[icg];
1133     nicg   = index[icg+1]-i0;
1134
1135     /* Get the i charge group info */
1136     igid   = GET_CGINFO_GID(cginfo[icg]);
1137
1138     if (md->nPerturbed)
1139     {
1140         gmx_fatal(FARGS, "AdResS does not support free energy pertubation\n");
1141     }
1142
1143     /* Unpack pointers to neighbourlist structs */
1144     if (fr->nnblists == 2)
1145     {
1146         nbl_ind        = 0;
1147         nbl_ind_adress = 1;
1148     }
1149     else
1150     {
1151         nbl_ind        = fr->gid2nblists[GID(igid, jgid, ngid)];
1152         nbl_ind_adress = nbl_ind+fr->nnblists/2;
1153     }
1154     if (bLR)
1155     {
1156         nlist        = fr->nblists[nbl_ind].nlist_lr;
1157         nlist_adress = fr->nblists[nbl_ind_adress].nlist_lr;
1158     }
1159     else
1160     {
1161         nlist        = fr->nblists[nbl_ind].nlist_sr;
1162         nlist_adress = fr->nblists[nbl_ind_adress].nlist_sr;
1163     }
1164
1165
1166     vdwc = &nlist[eNL_VDWQQ];
1167     vdw  = &nlist[eNL_VDW];
1168     coul = &nlist[eNL_QQ];
1169
1170     vdwc_adress = &nlist_adress[eNL_VDWQQ];
1171     vdw_adress  = &nlist_adress[eNL_VDW];
1172     coul_adress = &nlist_adress[eNL_QQ];
1173
1174     /* We do not support solvent optimization with AdResS for now.
1175        For this we would need hybrid solvent-other kernels */
1176
1177     /* no solvent as i charge group */
1178     /* Loop over the atoms in the i charge group */
1179     for (i = 0; i < nicg; i++)
1180     {
1181         i_atom  = i0+i;
1182         gid     = GID(igid, jgid, ngid);
1183         qi      = charge[i_atom];
1184
1185         /* Create new i_atom for each energy group */
1186         if (bDoVdW && bDoCoul)
1187         {
1188             new_i_nblist(vdwc, i_atom, shift, gid);
1189             new_i_nblist(vdwc_adress, i_atom, shift, gid);
1190
1191         }
1192         if (bDoVdW)
1193         {
1194             new_i_nblist(vdw, i_atom, shift, gid);
1195             new_i_nblist(vdw_adress, i_atom, shift, gid);
1196
1197         }
1198         if (bDoCoul)
1199         {
1200             new_i_nblist(coul, i_atom, shift, gid);
1201             new_i_nblist(coul_adress, i_atom, shift, gid);
1202         }
1203         bDoVdW_i  = (bDoVdW  && bHaveVdW[type[i_atom]]);
1204         bDoCoul_i = (bDoCoul && qi != 0);
1205
1206         /* Here we find out whether the energy groups interaction belong to a
1207          * coarse-grained (vsite) or atomistic interaction. Note that, beacuse
1208          * interactions between coarse-grained and other (atomistic) energygroups
1209          * are excluded automatically by grompp, it is sufficient to check for
1210          * the group id of atom i (igid) */
1211         bEnergyGroupCG = !egp_explicit(fr, igid);
1212
1213         if (bDoVdW_i || bDoCoul_i)
1214         {
1215             /* Loop over the j charge groups */
1216             for (j = 0; (j < nj); j++)
1217             {
1218                 jcg = jjcg[j];
1219
1220                 /* Check for large charge groups */
1221                 if (jcg == icg)
1222                 {
1223                     jj0 = i0 + i + 1;
1224                 }
1225                 else
1226                 {
1227                     jj0 = index[jcg];
1228                 }
1229
1230                 jj1 = index[jcg+1];
1231                 /* Finally loop over the atoms in the j-charge group */
1232                 for (jj = jj0; jj < jj1; jj++)
1233                 {
1234                     bNotEx = NOTEXCL(bExcl, i, jj);
1235
1236                     /* Now we have to exclude interactions which will be zero
1237                      * anyway due to the AdResS weights (in previous implementations
1238                      * this was done in the force kernel). This is necessary as
1239                      * pure interactions (those with b_hybrid=false, i.e. w_i*w_j==1 or 0)
1240                      * are put into neighbour lists which will be passed to the
1241                      * standard (optimized) kernels for speed. The interactions with
1242                      * b_hybrid=true are placed into the _adress neighbour lists and
1243                      * processed by the generic AdResS kernel.
1244                      */
1245                     if ( (bEnergyGroupCG &&
1246                           wf[i_atom] >= 1-GMX_REAL_EPS && wf[jj] >= 1-GMX_REAL_EPS ) ||
1247                          ( !bEnergyGroupCG && wf[jj] <= GMX_REAL_EPS ) )
1248                     {
1249                         continue;
1250                     }
1251
1252                     b_hybrid = !((wf[i_atom] >= 1-GMX_REAL_EPS && wf[jj] >= 1-GMX_REAL_EPS) ||
1253                                  (wf[i_atom] <= GMX_REAL_EPS && wf[jj] <= GMX_REAL_EPS));
1254
1255                     if (bNotEx)
1256                     {
1257                         if (!bDoVdW_i)
1258                         {
1259                             if (charge[jj] != 0)
1260                             {
1261                                 if (!b_hybrid)
1262                                 {
1263                                     add_j_to_nblist(coul, jj, bLR);
1264                                 }
1265                                 else
1266                                 {
1267                                     add_j_to_nblist(coul_adress, jj, bLR);
1268                                 }
1269                             }
1270                         }
1271                         else if (!bDoCoul_i)
1272                         {
1273                             if (bHaveVdW[type[jj]])
1274                             {
1275                                 if (!b_hybrid)
1276                                 {
1277                                     add_j_to_nblist(vdw, jj, bLR);
1278                                 }
1279                                 else
1280                                 {
1281                                     add_j_to_nblist(vdw_adress, jj, bLR);
1282                                 }
1283                             }
1284                         }
1285                         else
1286                         {
1287                             if (bHaveVdW[type[jj]])
1288                             {
1289                                 if (charge[jj] != 0)
1290                                 {
1291                                     if (!b_hybrid)
1292                                     {
1293                                         add_j_to_nblist(vdwc, jj, bLR);
1294                                     }
1295                                     else
1296                                     {
1297                                         add_j_to_nblist(vdwc_adress, jj, bLR);
1298                                     }
1299                                 }
1300                                 else
1301                                 {
1302                                     if (!b_hybrid)
1303                                     {
1304                                         add_j_to_nblist(vdw, jj, bLR);
1305                                     }
1306                                     else
1307                                     {
1308                                         add_j_to_nblist(vdw_adress, jj, bLR);
1309                                     }
1310
1311                                 }
1312                             }
1313                             else if (charge[jj] != 0)
1314                             {
1315                                 if (!b_hybrid)
1316                                 {
1317                                     add_j_to_nblist(coul, jj, bLR);
1318                                 }
1319                                 else
1320                                 {
1321                                     add_j_to_nblist(coul_adress, jj, bLR);
1322                                 }
1323
1324                             }
1325                         }
1326                     }
1327                 }
1328             }
1329
1330             close_i_nblist(vdw);
1331             close_i_nblist(coul);
1332             close_i_nblist(vdwc);
1333             close_i_nblist(vdw_adress);
1334             close_i_nblist(coul_adress);
1335             close_i_nblist(vdwc_adress);
1336         }
1337     }
1338 }
1339
1340 static void
1341 put_in_list_qmmm(gmx_bool gmx_unused              bHaveVdW[],
1342                  int                              ngid,
1343                  t_mdatoms gmx_unused     *       md,
1344                  int                              icg,
1345                  int                              jgid,
1346                  int                              nj,
1347                  atom_id                          jjcg[],
1348                  atom_id                          index[],
1349                  t_excl                           bExcl[],
1350                  int                              shift,
1351                  t_forcerec                *      fr,
1352                  gmx_bool                         bLR,
1353                  gmx_bool  gmx_unused             bDoVdW,
1354                  gmx_bool  gmx_unused             bDoCoul,
1355                  int       gmx_unused             solvent_opt)
1356 {
1357     t_nblist  *   coul;
1358     int           i, j, jcg, igid, gid;
1359     atom_id       jj, jj0, jj1, i_atom;
1360     int           i0, nicg;
1361     gmx_bool      bNotEx;
1362
1363     /* Get atom range */
1364     i0     = index[icg];
1365     nicg   = index[icg+1]-i0;
1366
1367     /* Get the i charge group info */
1368     igid   = GET_CGINFO_GID(fr->cginfo[icg]);
1369
1370     coul = &fr->QMMMlist;
1371
1372     /* Loop over atoms in the ith charge group */
1373     for (i = 0; i < nicg; i++)
1374     {
1375         i_atom = i0+i;
1376         gid    = GID(igid, jgid, ngid);
1377         /* Create new i_atom for each energy group */
1378         new_i_nblist(coul, i_atom, shift, gid);
1379
1380         /* Loop over the j charge groups */
1381         for (j = 0; j < nj; j++)
1382         {
1383             jcg = jjcg[j];
1384
1385             /* Charge groups cannot have QM and MM atoms simultaneously */
1386             if (jcg != icg)
1387             {
1388                 jj0 = index[jcg];
1389                 jj1 = index[jcg+1];
1390                 /* Finally loop over the atoms in the j-charge group */
1391                 for (jj = jj0; jj < jj1; jj++)
1392                 {
1393                     bNotEx = NOTEXCL(bExcl, i, jj);
1394                     if (bNotEx)
1395                     {
1396                         add_j_to_nblist(coul, jj, bLR);
1397                     }
1398                 }
1399             }
1400         }
1401         close_i_nblist(coul);
1402     }
1403 }
1404
1405 static void
1406 put_in_list_cg(gmx_bool  gmx_unused             bHaveVdW[],
1407                int                              ngid,
1408                t_mdatoms  gmx_unused    *       md,
1409                int                              icg,
1410                int                              jgid,
1411                int                              nj,
1412                atom_id                          jjcg[],
1413                atom_id                          index[],
1414                t_excl                           bExcl[],
1415                int                              shift,
1416                t_forcerec                *      fr,
1417                gmx_bool                         bLR,
1418                gmx_bool   gmx_unused            bDoVdW,
1419                gmx_bool   gmx_unused            bDoCoul,
1420                int        gmx_unused            solvent_opt)
1421 {
1422     int          cginfo;
1423     int          igid, gid, nbl_ind;
1424     t_nblist *   vdwc;
1425     int          j, jcg;
1426
1427     cginfo = fr->cginfo[icg];
1428
1429     igid = GET_CGINFO_GID(cginfo);
1430     gid  = GID(igid, jgid, ngid);
1431
1432     /* Unpack pointers to neighbourlist structs */
1433     if (fr->nnblists == 1)
1434     {
1435         nbl_ind = 0;
1436     }
1437     else
1438     {
1439         nbl_ind = fr->gid2nblists[gid];
1440     }
1441     if (bLR)
1442     {
1443         vdwc = &fr->nblists[nbl_ind].nlist_lr[eNL_VDWQQ];
1444     }
1445     else
1446     {
1447         vdwc = &fr->nblists[nbl_ind].nlist_sr[eNL_VDWQQ];
1448     }
1449
1450     /* Make a new neighbor list for charge group icg.
1451      * Currently simply one neighbor list is made with LJ and Coulomb.
1452      * If required, zero interactions could be removed here
1453      * or in the force loop.
1454      */
1455     new_i_nblist(vdwc, index[icg], shift, gid);
1456     vdwc->iinr_end[vdwc->nri] = index[icg+1];
1457
1458     for (j = 0; (j < nj); j++)
1459     {
1460         jcg = jjcg[j];
1461         /* Skip the icg-icg pairs if all self interactions are excluded */
1462         if (!(jcg == icg && GET_CGINFO_EXCL_INTRA(cginfo)))
1463         {
1464             /* Here we add the j charge group jcg to the list,
1465              * exclusions are also added to the list.
1466              */
1467             add_j_to_nblist_cg(vdwc, index[jcg], index[jcg+1], bExcl, icg == jcg, bLR);
1468         }
1469     }
1470
1471     close_i_nblist(vdwc);
1472 }
1473
1474 static void setexcl(atom_id start, atom_id end, t_blocka *excl, gmx_bool b,
1475                     t_excl bexcl[])
1476 {
1477     atom_id i, k;
1478
1479     if (b)
1480     {
1481         for (i = start; i < end; i++)
1482         {
1483             for (k = excl->index[i]; k < excl->index[i+1]; k++)
1484             {
1485                 SETEXCL(bexcl, i-start, excl->a[k]);
1486             }
1487         }
1488     }
1489     else
1490     {
1491         for (i = start; i < end; i++)
1492         {
1493             for (k = excl->index[i]; k < excl->index[i+1]; k++)
1494             {
1495                 RMEXCL(bexcl, i-start, excl->a[k]);
1496             }
1497         }
1498     }
1499 }
1500
1501 int calc_naaj(int icg, int cgtot)
1502 {
1503     int naaj;
1504
1505     if ((cgtot % 2) == 1)
1506     {
1507         /* Odd number of charge groups, easy */
1508         naaj = 1 + (cgtot/2);
1509     }
1510     else if ((cgtot % 4) == 0)
1511     {
1512         /* Multiple of four is hard */
1513         if (icg < cgtot/2)
1514         {
1515             if ((icg % 2) == 0)
1516             {
1517                 naaj = 1+(cgtot/2);
1518             }
1519             else
1520             {
1521                 naaj = cgtot/2;
1522             }
1523         }
1524         else
1525         {
1526             if ((icg % 2) == 1)
1527             {
1528                 naaj = 1+(cgtot/2);
1529             }
1530             else
1531             {
1532                 naaj = cgtot/2;
1533             }
1534         }
1535     }
1536     else
1537     {
1538         /* cgtot/2 = odd */
1539         if ((icg % 2) == 0)
1540         {
1541             naaj = 1+(cgtot/2);
1542         }
1543         else
1544         {
1545             naaj = cgtot/2;
1546         }
1547     }
1548 #ifdef DEBUG
1549     fprintf(log, "naaj=%d\n", naaj);
1550 #endif
1551
1552     return naaj;
1553 }
1554
1555 /************************************************
1556  *
1557  *  S I M P L E      C O R E     S T U F F
1558  *
1559  ************************************************/
1560
1561 static real calc_image_tric(rvec xi, rvec xj, matrix box,
1562                             rvec b_inv, int *shift)
1563 {
1564     /* This code assumes that the cut-off is smaller than
1565      * a half times the smallest diagonal element of the box.
1566      */
1567     const real h25 = 2.5;
1568     real       dx, dy, dz;
1569     real       r2;
1570     int        tx, ty, tz;
1571
1572     /* Compute diff vector */
1573     dz = xj[ZZ] - xi[ZZ];
1574     dy = xj[YY] - xi[YY];
1575     dx = xj[XX] - xi[XX];
1576
1577     /* Perform NINT operation, using trunc operation, therefore
1578      * we first add 2.5 then subtract 2 again
1579      */
1580     tz  = static_cast<int>(dz*b_inv[ZZ] + h25);
1581     tz -= 2;
1582     dz -= tz*box[ZZ][ZZ];
1583     dy -= tz*box[ZZ][YY];
1584     dx -= tz*box[ZZ][XX];
1585
1586     ty  = static_cast<int>(dy*b_inv[YY] + h25);
1587     ty -= 2;
1588     dy -= ty*box[YY][YY];
1589     dx -= ty*box[YY][XX];
1590
1591     tx  = static_cast<int>(dx*b_inv[XX]+h25);
1592     tx -= 2;
1593     dx -= tx*box[XX][XX];
1594
1595     /* Distance squared */
1596     r2 = (dx*dx) + (dy*dy) + (dz*dz);
1597
1598     *shift = XYZ2IS(tx, ty, tz);
1599
1600     return r2;
1601 }
1602
1603 static real calc_image_rect(rvec xi, rvec xj, rvec box_size,
1604                             rvec b_inv, int *shift)
1605 {
1606     const real h15 = 1.5;
1607     real       ddx, ddy, ddz;
1608     real       dx, dy, dz;
1609     real       r2;
1610     int        tx, ty, tz;
1611
1612     /* Compute diff vector */
1613     dx = xj[XX] - xi[XX];
1614     dy = xj[YY] - xi[YY];
1615     dz = xj[ZZ] - xi[ZZ];
1616
1617     /* Perform NINT operation, using trunc operation, therefore
1618      * we first add 1.5 then subtract 1 again
1619      */
1620     tx = static_cast<int>(dx*b_inv[XX] + h15);
1621     ty = static_cast<int>(dy*b_inv[YY] + h15);
1622     tz = static_cast<int>(dz*b_inv[ZZ] + h15);
1623     tx--;
1624     ty--;
1625     tz--;
1626
1627     /* Correct diff vector for translation */
1628     ddx = tx*box_size[XX] - dx;
1629     ddy = ty*box_size[YY] - dy;
1630     ddz = tz*box_size[ZZ] - dz;
1631
1632     /* Distance squared */
1633     r2 = (ddx*ddx) + (ddy*ddy) + (ddz*ddz);
1634
1635     *shift = XYZ2IS(tx, ty, tz);
1636
1637     return r2;
1638 }
1639
1640 static void add_simple(t_ns_buf * nsbuf, int nrj, atom_id cg_j,
1641                        gmx_bool bHaveVdW[], int ngid, t_mdatoms *md,
1642                        int icg, int jgid, t_block *cgs, t_excl bexcl[],
1643                        int shift, t_forcerec *fr, put_in_list_t *put_in_list)
1644 {
1645     if (nsbuf->nj + nrj > MAX_CG)
1646     {
1647         put_in_list(bHaveVdW, ngid, md, icg, jgid, nsbuf->ncg, nsbuf->jcg,
1648                     cgs->index, bexcl, shift, fr, FALSE, TRUE, TRUE, fr->solvent_opt);
1649         /* Reset buffer contents */
1650         nsbuf->ncg = nsbuf->nj = 0;
1651     }
1652     nsbuf->jcg[nsbuf->ncg++] = cg_j;
1653     nsbuf->nj               += nrj;
1654 }
1655
1656 static void ns_inner_tric(rvec x[], int icg, int *i_egp_flags,
1657                           int njcg, atom_id jcg[],
1658                           matrix box, rvec b_inv, real rcut2,
1659                           t_block *cgs, t_ns_buf **ns_buf,
1660                           gmx_bool bHaveVdW[], int ngid, t_mdatoms *md,
1661                           t_excl bexcl[], t_forcerec *fr,
1662                           put_in_list_t *put_in_list)
1663 {
1664     int       shift;
1665     int       j, nrj, jgid;
1666     int      *cginfo = fr->cginfo;
1667     atom_id   cg_j, *cgindex;
1668
1669     cgindex = cgs->index;
1670     shift   = CENTRAL;
1671     for (j = 0; (j < njcg); j++)
1672     {
1673         cg_j   = jcg[j];
1674         nrj    = cgindex[cg_j+1]-cgindex[cg_j];
1675         if (calc_image_tric(x[icg], x[cg_j], box, b_inv, &shift) < rcut2)
1676         {
1677             jgid  = GET_CGINFO_GID(cginfo[cg_j]);
1678             if (!(i_egp_flags[jgid] & EGP_EXCL))
1679             {
1680                 add_simple(&ns_buf[jgid][shift], nrj, cg_j,
1681                            bHaveVdW, ngid, md, icg, jgid, cgs, bexcl, shift, fr,
1682                            put_in_list);
1683             }
1684         }
1685     }
1686 }
1687
1688 static void ns_inner_rect(rvec x[], int icg, int *i_egp_flags,
1689                           int njcg, atom_id jcg[],
1690                           gmx_bool bBox, rvec box_size, rvec b_inv, real rcut2,
1691                           t_block *cgs, t_ns_buf **ns_buf,
1692                           gmx_bool bHaveVdW[], int ngid, t_mdatoms *md,
1693                           t_excl bexcl[], t_forcerec *fr,
1694                           put_in_list_t *put_in_list)
1695 {
1696     int       shift;
1697     int       j, nrj, jgid;
1698     int      *cginfo = fr->cginfo;
1699     atom_id   cg_j, *cgindex;
1700
1701     cgindex = cgs->index;
1702     if (bBox)
1703     {
1704         shift = CENTRAL;
1705         for (j = 0; (j < njcg); j++)
1706         {
1707             cg_j   = jcg[j];
1708             nrj    = cgindex[cg_j+1]-cgindex[cg_j];
1709             if (calc_image_rect(x[icg], x[cg_j], box_size, b_inv, &shift) < rcut2)
1710             {
1711                 jgid  = GET_CGINFO_GID(cginfo[cg_j]);
1712                 if (!(i_egp_flags[jgid] & EGP_EXCL))
1713                 {
1714                     add_simple(&ns_buf[jgid][shift], nrj, cg_j,
1715                                bHaveVdW, ngid, md, icg, jgid, cgs, bexcl, shift, fr,
1716                                put_in_list);
1717                 }
1718             }
1719         }
1720     }
1721     else
1722     {
1723         for (j = 0; (j < njcg); j++)
1724         {
1725             cg_j   = jcg[j];
1726             nrj    = cgindex[cg_j+1]-cgindex[cg_j];
1727             if ((rcut2 == 0) || (distance2(x[icg], x[cg_j]) < rcut2))
1728             {
1729                 jgid  = GET_CGINFO_GID(cginfo[cg_j]);
1730                 if (!(i_egp_flags[jgid] & EGP_EXCL))
1731                 {
1732                     add_simple(&ns_buf[jgid][CENTRAL], nrj, cg_j,
1733                                bHaveVdW, ngid, md, icg, jgid, cgs, bexcl, CENTRAL, fr,
1734                                put_in_list);
1735                 }
1736             }
1737         }
1738     }
1739 }
1740
1741 /* ns_simple_core needs to be adapted for QMMM still 2005 */
1742
1743 static int ns_simple_core(t_forcerec *fr,
1744                           gmx_localtop_t *top,
1745                           t_mdatoms *md,
1746                           matrix box, rvec box_size,
1747                           t_excl bexcl[], atom_id *aaj,
1748                           int ngid, t_ns_buf **ns_buf,
1749                           put_in_list_t *put_in_list, gmx_bool bHaveVdW[])
1750 {
1751     int          naaj, k;
1752     real         rlist2;
1753     int          nsearch, icg, igid, nn;
1754     int         *cginfo;
1755     t_ns_buf    *nsbuf;
1756     /* atom_id  *i_atoms; */
1757     t_block     *cgs  = &(top->cgs);
1758     t_blocka    *excl = &(top->excls);
1759     rvec         b_inv;
1760     int          m;
1761     gmx_bool     bBox, bTriclinic;
1762     int         *i_egp_flags;
1763
1764     rlist2 = sqr(fr->rlist);
1765
1766     bBox = (fr->ePBC != epbcNONE);
1767     if (bBox)
1768     {
1769         for (m = 0; (m < DIM); m++)
1770         {
1771             b_inv[m] = divide_err(1.0, box_size[m]);
1772         }
1773         bTriclinic = TRICLINIC(box);
1774     }
1775     else
1776     {
1777         bTriclinic = FALSE;
1778     }
1779
1780     cginfo = fr->cginfo;
1781
1782     nsearch = 0;
1783     for (icg = fr->cg0; (icg < fr->hcg); icg++)
1784     {
1785         /*
1786            i0        = cgs->index[icg];
1787            nri       = cgs->index[icg+1]-i0;
1788            i_atoms   = &(cgs->a[i0]);
1789            i_eg_excl = fr->eg_excl + ngid*md->cENER[*i_atoms];
1790            setexcl(nri,i_atoms,excl,TRUE,bexcl);
1791          */
1792         igid        = GET_CGINFO_GID(cginfo[icg]);
1793         i_egp_flags = fr->egp_flags + ngid*igid;
1794         setexcl(cgs->index[icg], cgs->index[icg+1], excl, TRUE, bexcl);
1795
1796         naaj = calc_naaj(icg, cgs->nr);
1797         if (bTriclinic)
1798         {
1799             ns_inner_tric(fr->cg_cm, icg, i_egp_flags, naaj, &(aaj[icg]),
1800                           box, b_inv, rlist2, cgs, ns_buf,
1801                           bHaveVdW, ngid, md, bexcl, fr, put_in_list);
1802         }
1803         else
1804         {
1805             ns_inner_rect(fr->cg_cm, icg, i_egp_flags, naaj, &(aaj[icg]),
1806                           bBox, box_size, b_inv, rlist2, cgs, ns_buf,
1807                           bHaveVdW, ngid, md, bexcl, fr, put_in_list);
1808         }
1809         nsearch += naaj;
1810
1811         for (nn = 0; (nn < ngid); nn++)
1812         {
1813             for (k = 0; (k < SHIFTS); k++)
1814             {
1815                 nsbuf = &(ns_buf[nn][k]);
1816                 if (nsbuf->ncg > 0)
1817                 {
1818                     put_in_list(bHaveVdW, ngid, md, icg, nn, nsbuf->ncg, nsbuf->jcg,
1819                                 cgs->index, bexcl, k, fr, FALSE, TRUE, TRUE, fr->solvent_opt);
1820                     nsbuf->ncg = nsbuf->nj = 0;
1821                 }
1822             }
1823         }
1824         /* setexcl(nri,i_atoms,excl,FALSE,bexcl); */
1825         setexcl(cgs->index[icg], cgs->index[icg+1], excl, FALSE, bexcl);
1826     }
1827     close_neighbor_lists(fr, FALSE);
1828
1829     return nsearch;
1830 }
1831
1832 /************************************************
1833  *
1834  *    N S 5     G R I D     S T U F F
1835  *
1836  ************************************************/
1837
1838 static gmx_inline void get_dx(int Nx, real gridx, real rc2, int xgi, real x,
1839                               int *dx0, int *dx1, real *dcx2)
1840 {
1841     real dcx, tmp;
1842     int  xgi0, xgi1, i;
1843
1844     if (xgi < 0)
1845     {
1846         *dx0 = 0;
1847         xgi0 = -1;
1848         *dx1 = -1;
1849         xgi1 = 0;
1850     }
1851     else if (xgi >= Nx)
1852     {
1853         *dx0 = Nx;
1854         xgi0 = Nx-1;
1855         *dx1 = Nx-1;
1856         xgi1 = Nx;
1857     }
1858     else
1859     {
1860         dcx2[xgi] = 0;
1861         *dx0      = xgi;
1862         xgi0      = xgi-1;
1863         *dx1      = xgi;
1864         xgi1      = xgi+1;
1865     }
1866
1867     for (i = xgi0; i >= 0; i--)
1868     {
1869         dcx = (i+1)*gridx-x;
1870         tmp = dcx*dcx;
1871         if (tmp >= rc2)
1872         {
1873             break;
1874         }
1875         *dx0    = i;
1876         dcx2[i] = tmp;
1877     }
1878     for (i = xgi1; i < Nx; i++)
1879     {
1880         dcx = i*gridx-x;
1881         tmp = dcx*dcx;
1882         if (tmp >= rc2)
1883         {
1884             break;
1885         }
1886         *dx1    = i;
1887         dcx2[i] = tmp;
1888     }
1889 }
1890
1891 static gmx_inline void get_dx_dd(int Nx, real gridx, real rc2, int xgi, real x,
1892                                  int ncpddc, int shift_min, int shift_max,
1893                                  int *g0, int *g1, real *dcx2)
1894 {
1895     real dcx, tmp;
1896     int  g_min, g_max, shift_home;
1897
1898     if (xgi < 0)
1899     {
1900         g_min = 0;
1901         g_max = Nx - 1;
1902         *g0   = 0;
1903         *g1   = -1;
1904     }
1905     else if (xgi >= Nx)
1906     {
1907         g_min = 0;
1908         g_max = Nx - 1;
1909         *g0   = Nx;
1910         *g1   = Nx - 1;
1911     }
1912     else
1913     {
1914         if (ncpddc == 0)
1915         {
1916             g_min = 0;
1917             g_max = Nx - 1;
1918         }
1919         else
1920         {
1921             if (xgi < ncpddc)
1922             {
1923                 shift_home = 0;
1924             }
1925             else
1926             {
1927                 shift_home = -1;
1928             }
1929             g_min = (shift_min == shift_home ? 0          : ncpddc);
1930             g_max = (shift_max == shift_home ? ncpddc - 1 : Nx - 1);
1931         }
1932         if (shift_min > 0)
1933         {
1934             *g0 = g_min;
1935             *g1 = g_min - 1;
1936         }
1937         else if (shift_max < 0)
1938         {
1939             *g0 = g_max + 1;
1940             *g1 = g_max;
1941         }
1942         else
1943         {
1944             *g0       = xgi;
1945             *g1       = xgi;
1946             dcx2[xgi] = 0;
1947         }
1948     }
1949
1950     while (*g0 > g_min)
1951     {
1952         /* Check one grid cell down */
1953         dcx = ((*g0 - 1) + 1)*gridx - x;
1954         tmp = dcx*dcx;
1955         if (tmp >= rc2)
1956         {
1957             break;
1958         }
1959         (*g0)--;
1960         dcx2[*g0] = tmp;
1961     }
1962
1963     while (*g1 < g_max)
1964     {
1965         /* Check one grid cell up */
1966         dcx = (*g1 + 1)*gridx - x;
1967         tmp = dcx*dcx;
1968         if (tmp >= rc2)
1969         {
1970             break;
1971         }
1972         (*g1)++;
1973         dcx2[*g1] = tmp;
1974     }
1975 }
1976
1977
1978 #define sqr(x) ((x)*(x))
1979 #define calc_dx2(XI, YI, ZI, y) (sqr(XI-y[XX]) + sqr(YI-y[YY]) + sqr(ZI-y[ZZ]))
1980 #define calc_cyl_dx2(XI, YI, y) (sqr(XI-y[XX]) + sqr(YI-y[YY]))
1981 /****************************************************
1982  *
1983  *    F A S T   N E I G H B O R  S E A R C H I N G
1984  *
1985  *    Optimized neighboursearching routine using grid
1986  *    at least 1x1x1, see GROMACS manual
1987  *
1988  ****************************************************/
1989
1990
1991 static void get_cutoff2(t_forcerec *fr, gmx_bool bDoLongRange,
1992                         real *rvdw2, real *rcoul2,
1993                         real *rs2, real *rm2, real *rl2)
1994 {
1995     *rs2 = sqr(fr->rlist);
1996
1997     if (bDoLongRange && fr->bTwinRange)
1998     {
1999         /* With plain cut-off or RF we need to make the list exactly
2000          * up to the cut-off and the cut-off's can be different,
2001          * so we can not simply set them to rlistlong.
2002          * To keep this code compatible with (exotic) old cases,
2003          * we also create lists up to rvdw/rcoulomb for PME and Ewald.
2004          * The interaction check should correspond to:
2005          * !ir_vdw/coulomb_might_be_zero_at_cutoff from inputrec.c.
2006          */
2007         if (((fr->vdwtype == evdwCUT || fr->vdwtype == evdwPME) &&
2008              fr->vdw_modifier == eintmodNONE) ||
2009             fr->rvdw <= fr->rlist)
2010         {
2011             *rvdw2  = sqr(fr->rvdw);
2012         }
2013         else
2014         {
2015             *rvdw2  = sqr(fr->rlistlong);
2016         }
2017         if (((fr->eeltype == eelCUT ||
2018               (EEL_RF(fr->eeltype) && fr->eeltype != eelRF_ZERO) ||
2019               fr->eeltype == eelPME ||
2020               fr->eeltype == eelEWALD) &&
2021              fr->coulomb_modifier == eintmodNONE) ||
2022             fr->rcoulomb <= fr->rlist)
2023         {
2024             *rcoul2 = sqr(fr->rcoulomb);
2025         }
2026         else
2027         {
2028             *rcoul2 = sqr(fr->rlistlong);
2029         }
2030     }
2031     else
2032     {
2033         /* Workaround for a gcc -O3 or -ffast-math problem */
2034         *rvdw2  = *rs2;
2035         *rcoul2 = *rs2;
2036     }
2037     *rm2 = std::min(*rvdw2, *rcoul2);
2038     *rl2 = std::max(*rvdw2, *rcoul2);
2039 }
2040
2041 static void init_nsgrid_lists(t_forcerec *fr, int ngid, gmx_ns_t *ns)
2042 {
2043     real rvdw2, rcoul2, rs2, rm2, rl2;
2044     int  j;
2045
2046     get_cutoff2(fr, TRUE, &rvdw2, &rcoul2, &rs2, &rm2, &rl2);
2047
2048     /* Short range buffers */
2049     snew(ns->nl_sr, ngid);
2050     /* Counters */
2051     snew(ns->nsr, ngid);
2052     snew(ns->nlr_ljc, ngid);
2053     snew(ns->nlr_one, ngid);
2054
2055     /* Always allocate both list types, since rcoulomb might now change with PME load balancing */
2056     /* Long range VdW and Coul buffers */
2057     snew(ns->nl_lr_ljc, ngid);
2058     /* Long range VdW or Coul only buffers */
2059     snew(ns->nl_lr_one, ngid);
2060
2061     for (j = 0; (j < ngid); j++)
2062     {
2063         snew(ns->nl_sr[j], MAX_CG);
2064         snew(ns->nl_lr_ljc[j], MAX_CG);
2065         snew(ns->nl_lr_one[j], MAX_CG);
2066     }
2067     if (debug)
2068     {
2069         fprintf(debug,
2070                 "ns5_core: rs2 = %g, rm2 = %g, rl2 = %g (nm^2)\n",
2071                 rs2, rm2, rl2);
2072     }
2073 }
2074
2075 static int nsgrid_core(t_commrec *cr, t_forcerec *fr,
2076                        matrix box, int ngid,
2077                        gmx_localtop_t *top,
2078                        t_grid *grid,
2079                        t_excl bexcl[], gmx_bool *bExcludeAlleg,
2080                        t_mdatoms *md,
2081                        put_in_list_t *put_in_list,
2082                        gmx_bool bHaveVdW[],
2083                        gmx_bool bDoLongRange, gmx_bool bMakeQMMMnblist)
2084 {
2085     gmx_ns_t     *ns;
2086     atom_id     **nl_lr_ljc, **nl_lr_one, **nl_sr;
2087     int          *nlr_ljc, *nlr_one, *nsr;
2088     gmx_domdec_t *dd;
2089     t_block      *cgs    = &(top->cgs);
2090     int          *cginfo = fr->cginfo;
2091     /* atom_id *i_atoms,*cgsindex=cgs->index; */
2092     ivec          sh0, sh1, shp;
2093     int           cell_x, cell_y, cell_z;
2094     int           d, tx, ty, tz, dx, dy, dz, cj;
2095 #ifdef ALLOW_OFFDIAG_LT_HALFDIAG
2096     int           zsh_ty, zsh_tx, ysh_tx;
2097 #endif
2098     int           dx0, dx1, dy0, dy1, dz0, dz1;
2099     int           Nx, Ny, Nz, shift = -1, j, nrj, nns, nn = -1;
2100     real          gridx, gridy, gridz, grid_x, grid_y;
2101     real         *dcx2, *dcy2, *dcz2;
2102     int           zgi, ygi, xgi;
2103     int           cg0, cg1, icg = -1, cgsnr, i0, igid, naaj, max_jcg;
2104     int           jcg0, jcg1, jjcg, cgj0, jgid;
2105     int          *grida, *gridnra, *gridind;
2106     gmx_bool      rvdw_lt_rcoul, rcoul_lt_rvdw;
2107     rvec         *cgcm, grid_offset;
2108     real          r2, rs2, rvdw2, rcoul2, rm2, rl2, XI, YI, ZI, tmp1, tmp2;
2109     int          *i_egp_flags;
2110     gmx_bool      bDomDec, bTriclinicX, bTriclinicY;
2111     ivec          ncpddc;
2112
2113     ns = &fr->ns;
2114
2115     bDomDec = DOMAINDECOMP(cr);
2116     dd      = cr->dd;
2117
2118     bTriclinicX = ((YY < grid->npbcdim &&
2119                     (!bDomDec || dd->nc[YY] == 1) && box[YY][XX] != 0) ||
2120                    (ZZ < grid->npbcdim &&
2121                     (!bDomDec || dd->nc[ZZ] == 1) && box[ZZ][XX] != 0));
2122     bTriclinicY =  (ZZ < grid->npbcdim &&
2123                     (!bDomDec || dd->nc[ZZ] == 1) && box[ZZ][YY] != 0);
2124
2125     cgsnr    = cgs->nr;
2126
2127     get_cutoff2(fr, bDoLongRange, &rvdw2, &rcoul2, &rs2, &rm2, &rl2);
2128
2129     rvdw_lt_rcoul = (rvdw2 >= rcoul2);
2130     rcoul_lt_rvdw = (rcoul2 >= rvdw2);
2131
2132     if (bMakeQMMMnblist)
2133     {
2134         rm2 = rl2;
2135         rs2 = rl2;
2136     }
2137
2138     nl_sr     = ns->nl_sr;
2139     nsr       = ns->nsr;
2140     nl_lr_ljc = ns->nl_lr_ljc;
2141     nl_lr_one = ns->nl_lr_one;
2142     nlr_ljc   = ns->nlr_ljc;
2143     nlr_one   = ns->nlr_one;
2144
2145     /* Unpack arrays */
2146     cgcm    = fr->cg_cm;
2147     Nx      = grid->n[XX];
2148     Ny      = grid->n[YY];
2149     Nz      = grid->n[ZZ];
2150     grida   = grid->a;
2151     gridind = grid->index;
2152     gridnra = grid->nra;
2153     nns     = 0;
2154
2155     gridx      = grid->cell_size[XX];
2156     gridy      = grid->cell_size[YY];
2157     gridz      = grid->cell_size[ZZ];
2158     grid_x     = 1/gridx;
2159     grid_y     = 1/gridy;
2160     copy_rvec(grid->cell_offset, grid_offset);
2161     copy_ivec(grid->ncpddc, ncpddc);
2162     dcx2       = grid->dcx2;
2163     dcy2       = grid->dcy2;
2164     dcz2       = grid->dcz2;
2165
2166 #ifdef ALLOW_OFFDIAG_LT_HALFDIAG
2167     zsh_ty = floor(-box[ZZ][YY]/box[YY][YY]+0.5);
2168     zsh_tx = floor(-box[ZZ][XX]/box[XX][XX]+0.5);
2169     ysh_tx = floor(-box[YY][XX]/box[XX][XX]+0.5);
2170     if (zsh_tx != 0 && ysh_tx != 0)
2171     {
2172         /* This could happen due to rounding, when both ratios are 0.5 */
2173         ysh_tx = 0;
2174     }
2175 #endif
2176
2177     debug_gmx();
2178
2179     if (fr->n_tpi)
2180     {
2181         /* We only want a list for the test particle */
2182         cg0 = cgsnr - 1;
2183     }
2184     else
2185     {
2186         cg0 = grid->icg0;
2187     }
2188     cg1 = grid->icg1;
2189
2190     /* Set the shift range */
2191     for (d = 0; d < DIM; d++)
2192     {
2193         sh0[d] = -1;
2194         sh1[d] = 1;
2195         /* Check if we need periodicity shifts.
2196          * Without PBC or with domain decomposition we don't need them.
2197          */
2198         if (d >= ePBC2npbcdim(fr->ePBC) || (bDomDec && dd->nc[d] > 1))
2199         {
2200             shp[d] = 0;
2201         }
2202         else
2203         {
2204             if (d == XX &&
2205                 box[XX][XX] - fabs(box[YY][XX]) - fabs(box[ZZ][XX]) < std::sqrt(rl2))
2206             {
2207                 shp[d] = 2;
2208             }
2209             else
2210             {
2211                 shp[d] = 1;
2212             }
2213         }
2214     }
2215
2216     /* Loop over charge groups */
2217     for (icg = cg0; (icg < cg1); icg++)
2218     {
2219         igid = GET_CGINFO_GID(cginfo[icg]);
2220         /* Skip this charge group if all energy groups are excluded! */
2221         if (bExcludeAlleg[igid])
2222         {
2223             continue;
2224         }
2225
2226         i0   = cgs->index[icg];
2227
2228         if (bMakeQMMMnblist)
2229         {
2230             /* Skip this charge group if it is not a QM atom while making a
2231              * QM/MM neighbourlist
2232              */
2233             if (md->bQM[i0] == FALSE)
2234             {
2235                 continue; /* MM particle, go to next particle */
2236             }
2237
2238             /* Compute the number of charge groups that fall within the control
2239              * of this one (icg)
2240              */
2241             naaj    = calc_naaj(icg, cgsnr);
2242             jcg0    = icg;
2243             jcg1    = icg + naaj;
2244             max_jcg = cgsnr;
2245         }
2246         else
2247         {
2248             /* make a normal neighbourlist */
2249
2250             if (bDomDec)
2251             {
2252                 /* Get the j charge-group and dd cell shift ranges */
2253                 dd_get_ns_ranges(cr->dd, icg, &jcg0, &jcg1, sh0, sh1);
2254                 max_jcg = 0;
2255             }
2256             else
2257             {
2258                 /* Compute the number of charge groups that fall within the control
2259                  * of this one (icg)
2260                  */
2261                 naaj = calc_naaj(icg, cgsnr);
2262                 jcg0 = icg;
2263                 jcg1 = icg + naaj;
2264
2265                 if (fr->n_tpi)
2266                 {
2267                     /* The i-particle is awlways the test particle,
2268                      * so we want all j-particles
2269                      */
2270                     max_jcg = cgsnr - 1;
2271                 }
2272                 else
2273                 {
2274                     max_jcg  = jcg1 - cgsnr;
2275                 }
2276             }
2277         }
2278
2279         i_egp_flags = fr->egp_flags + igid*ngid;
2280
2281         /* Set the exclusions for the atoms in charge group icg using a bitmask */
2282         setexcl(i0, cgs->index[icg+1], &top->excls, TRUE, bexcl);
2283
2284         ci2xyz(grid, icg, &cell_x, &cell_y, &cell_z);
2285
2286         /* Changed iicg to icg, DvdS 990115
2287          * (but see consistency check above, DvdS 990330)
2288          */
2289 #ifdef NS5DB
2290         fprintf(log, "icg=%5d, naaj=%5d, cell %d %d %d\n",
2291                 icg, naaj, cell_x, cell_y, cell_z);
2292 #endif
2293         /* Loop over shift vectors in three dimensions */
2294         for (tz = -shp[ZZ]; tz <= shp[ZZ]; tz++)
2295         {
2296             ZI = cgcm[icg][ZZ]+tz*box[ZZ][ZZ];
2297             /* Calculate range of cells in Z direction that have the shift tz */
2298             zgi = cell_z + tz*Nz;
2299 #define FAST_DD_NS
2300 #ifndef FAST_DD_NS
2301             get_dx(Nz, gridz, rl2, zgi, ZI, &dz0, &dz1, dcz2);
2302 #else
2303             get_dx_dd(Nz, gridz, rl2, zgi, ZI-grid_offset[ZZ],
2304                       ncpddc[ZZ], sh0[ZZ], sh1[ZZ], &dz0, &dz1, dcz2);
2305 #endif
2306             if (dz0 > dz1)
2307             {
2308                 continue;
2309             }
2310             for (ty = -shp[YY]; ty <= shp[YY]; ty++)
2311             {
2312                 YI = cgcm[icg][YY]+ty*box[YY][YY]+tz*box[ZZ][YY];
2313                 /* Calculate range of cells in Y direction that have the shift ty */
2314                 if (bTriclinicY)
2315                 {
2316                     ygi = (int)(Ny + (YI - grid_offset[YY])*grid_y) - Ny;
2317                 }
2318                 else
2319                 {
2320                     ygi = cell_y + ty*Ny;
2321                 }
2322 #ifndef FAST_DD_NS
2323                 get_dx(Ny, gridy, rl2, ygi, YI, &dy0, &dy1, dcy2);
2324 #else
2325                 get_dx_dd(Ny, gridy, rl2, ygi, YI-grid_offset[YY],
2326                           ncpddc[YY], sh0[YY], sh1[YY], &dy0, &dy1, dcy2);
2327 #endif
2328                 if (dy0 > dy1)
2329                 {
2330                     continue;
2331                 }
2332                 for (tx = -shp[XX]; tx <= shp[XX]; tx++)
2333                 {
2334                     XI = cgcm[icg][XX]+tx*box[XX][XX]+ty*box[YY][XX]+tz*box[ZZ][XX];
2335                     /* Calculate range of cells in X direction that have the shift tx */
2336                     if (bTriclinicX)
2337                     {
2338                         xgi = (int)(Nx + (XI - grid_offset[XX])*grid_x) - Nx;
2339                     }
2340                     else
2341                     {
2342                         xgi = cell_x + tx*Nx;
2343                     }
2344 #ifndef FAST_DD_NS
2345                     get_dx(Nx, gridx, rl2, xgi*Nx, XI, &dx0, &dx1, dcx2);
2346 #else
2347                     get_dx_dd(Nx, gridx, rl2, xgi, XI-grid_offset[XX],
2348                               ncpddc[XX], sh0[XX], sh1[XX], &dx0, &dx1, dcx2);
2349 #endif
2350                     if (dx0 > dx1)
2351                     {
2352                         continue;
2353                     }
2354                     /* Adress: an explicit cg that has a weigthing function of 0 is excluded
2355                      *  from the neigbour list as it will not interact  */
2356                     if (fr->adress_type != eAdressOff)
2357                     {
2358                         if (md->wf[cgs->index[icg]] <= GMX_REAL_EPS && egp_explicit(fr, igid))
2359                         {
2360                             continue;
2361                         }
2362                     }
2363                     /* Get shift vector */
2364                     shift = XYZ2IS(tx, ty, tz);
2365 #ifdef NS5DB
2366                     range_check(shift, 0, SHIFTS);
2367 #endif
2368                     for (nn = 0; (nn < ngid); nn++)
2369                     {
2370                         nsr[nn]      = 0;
2371                         nlr_ljc[nn]  = 0;
2372                         nlr_one[nn]  = 0;
2373                     }
2374 #ifdef NS5DB
2375                     fprintf(log, "shift: %2d, dx0,1: %2d,%2d, dy0,1: %2d,%2d, dz0,1: %2d,%2d\n",
2376                             shift, dx0, dx1, dy0, dy1, dz0, dz1);
2377                     fprintf(log, "cgcm: %8.3f  %8.3f  %8.3f\n", cgcm[icg][XX],
2378                             cgcm[icg][YY], cgcm[icg][ZZ]);
2379                     fprintf(log, "xi:   %8.3f  %8.3f  %8.3f\n", XI, YI, ZI);
2380 #endif
2381                     for (dx = dx0; (dx <= dx1); dx++)
2382                     {
2383                         tmp1 = rl2 - dcx2[dx];
2384                         for (dy = dy0; (dy <= dy1); dy++)
2385                         {
2386                             tmp2 = tmp1 - dcy2[dy];
2387                             if (tmp2 > 0)
2388                             {
2389                                 for (dz = dz0; (dz <= dz1); dz++)
2390                                 {
2391                                     if (tmp2 > dcz2[dz])
2392                                     {
2393                                         /* Find grid-cell cj in which possible neighbours are */
2394                                         cj   = xyz2ci(Ny, Nz, dx, dy, dz);
2395
2396                                         /* Check out how many cgs (nrj) there in this cell */
2397                                         nrj  = gridnra[cj];
2398
2399                                         /* Find the offset in the cg list */
2400                                         cgj0 = gridind[cj];
2401
2402                                         /* Check if all j's are out of range so we
2403                                          * can skip the whole cell.
2404                                          * Should save some time, especially with DD.
2405                                          */
2406                                         if (nrj == 0 ||
2407                                             (grida[cgj0] >= max_jcg &&
2408                                              (grida[cgj0] >= jcg1 || grida[cgj0+nrj-1] < jcg0)))
2409                                         {
2410                                             continue;
2411                                         }
2412
2413                                         /* Loop over cgs */
2414                                         for (j = 0; (j < nrj); j++)
2415                                         {
2416                                             jjcg = grida[cgj0+j];
2417
2418                                             /* check whether this guy is in range! */
2419                                             if ((jjcg >= jcg0 && jjcg < jcg1) ||
2420                                                 (jjcg < max_jcg))
2421                                             {
2422                                                 r2 = calc_dx2(XI, YI, ZI, cgcm[jjcg]);
2423                                                 if (r2 < rl2)
2424                                                 {
2425                                                     /* jgid = gid[cgsatoms[cgsindex[jjcg]]]; */
2426                                                     jgid = GET_CGINFO_GID(cginfo[jjcg]);
2427                                                     /* check energy group exclusions */
2428                                                     if (!(i_egp_flags[jgid] & EGP_EXCL))
2429                                                     {
2430                                                         if (r2 < rs2)
2431                                                         {
2432                                                             if (nsr[jgid] >= MAX_CG)
2433                                                             {
2434                                                                 /* Add to short-range list */
2435                                                                 put_in_list(bHaveVdW, ngid, md, icg, jgid,
2436                                                                             nsr[jgid], nl_sr[jgid],
2437                                                                             cgs->index, /* cgsatoms, */ bexcl,
2438                                                                             shift, fr, FALSE, TRUE, TRUE, fr->solvent_opt);
2439                                                                 nsr[jgid] = 0;
2440                                                             }
2441                                                             nl_sr[jgid][nsr[jgid]++] = jjcg;
2442                                                         }
2443                                                         else if (r2 < rm2)
2444                                                         {
2445                                                             if (nlr_ljc[jgid] >= MAX_CG)
2446                                                             {
2447                                                                 /* Add to LJ+coulomb long-range list */
2448                                                                 put_in_list(bHaveVdW, ngid, md, icg, jgid,
2449                                                                             nlr_ljc[jgid], nl_lr_ljc[jgid], top->cgs.index,
2450                                                                             bexcl, shift, fr, TRUE, TRUE, TRUE, fr->solvent_opt);
2451                                                                 nlr_ljc[jgid] = 0;
2452                                                             }
2453                                                             nl_lr_ljc[jgid][nlr_ljc[jgid]++] = jjcg;
2454                                                         }
2455                                                         else
2456                                                         {
2457                                                             if (nlr_one[jgid] >= MAX_CG)
2458                                                             {
2459                                                                 /* Add to long-range list with only coul, or only LJ */
2460                                                                 put_in_list(bHaveVdW, ngid, md, icg, jgid,
2461                                                                             nlr_one[jgid], nl_lr_one[jgid], top->cgs.index,
2462                                                                             bexcl, shift, fr, TRUE, rvdw_lt_rcoul, rcoul_lt_rvdw, fr->solvent_opt);
2463                                                                 nlr_one[jgid] = 0;
2464                                                             }
2465                                                             nl_lr_one[jgid][nlr_one[jgid]++] = jjcg;
2466                                                         }
2467                                                     }
2468                                                 }
2469                                                 nns++;
2470                                             }
2471                                         }
2472                                     }
2473                                 }
2474                             }
2475                         }
2476                     }
2477                     /* CHECK whether there is anything left in the buffers */
2478                     for (nn = 0; (nn < ngid); nn++)
2479                     {
2480                         if (nsr[nn] > 0)
2481                         {
2482                             put_in_list(bHaveVdW, ngid, md, icg, nn, nsr[nn], nl_sr[nn],
2483                                         cgs->index, /* cgsatoms, */ bexcl,
2484                                         shift, fr, FALSE, TRUE, TRUE, fr->solvent_opt);
2485                         }
2486
2487                         if (nlr_ljc[nn] > 0)
2488                         {
2489                             put_in_list(bHaveVdW, ngid, md, icg, nn, nlr_ljc[nn],
2490                                         nl_lr_ljc[nn], top->cgs.index,
2491                                         bexcl, shift, fr, TRUE, TRUE, TRUE, fr->solvent_opt);
2492                         }
2493
2494                         if (nlr_one[nn] > 0)
2495                         {
2496                             put_in_list(bHaveVdW, ngid, md, icg, nn, nlr_one[nn],
2497                                         nl_lr_one[nn], top->cgs.index,
2498                                         bexcl, shift, fr, TRUE, rvdw_lt_rcoul, rcoul_lt_rvdw, fr->solvent_opt);
2499                         }
2500                     }
2501                 }
2502             }
2503         }
2504         setexcl(cgs->index[icg], cgs->index[icg+1], &top->excls, FALSE, bexcl);
2505     }
2506     /* No need to perform any left-over force calculations anymore (as we used to do here)
2507      * since we now save the proper long-range lists for later evaluation.
2508      */
2509
2510     debug_gmx();
2511
2512     /* Close neighbourlists */
2513     close_neighbor_lists(fr, bMakeQMMMnblist);
2514
2515     return nns;
2516 }
2517
2518 void ns_realloc_natoms(gmx_ns_t *ns, int natoms)
2519 {
2520     int i;
2521
2522     if (natoms > ns->nra_alloc)
2523     {
2524         ns->nra_alloc = over_alloc_dd(natoms);
2525         srenew(ns->bexcl, ns->nra_alloc);
2526         for (i = 0; i < ns->nra_alloc; i++)
2527         {
2528             ns->bexcl[i] = 0;
2529         }
2530     }
2531 }
2532
2533 void init_ns(FILE *fplog, const t_commrec *cr,
2534              gmx_ns_t *ns, t_forcerec *fr,
2535              const gmx_mtop_t *mtop)
2536 {
2537     int  mt, icg, nr_in_cg, maxcg, i, j, jcg, ngid, ncg;
2538     t_block *cgs;
2539
2540     /* Compute largest charge groups size (# atoms) */
2541     nr_in_cg = 1;
2542     for (mt = 0; mt < mtop->nmoltype; mt++)
2543     {
2544         cgs = &mtop->moltype[mt].cgs;
2545         for (icg = 0; (icg < cgs->nr); icg++)
2546         {
2547             nr_in_cg = std::max(nr_in_cg, (int)(cgs->index[icg+1]-cgs->index[icg]));
2548         }
2549     }
2550
2551     /* Verify whether largest charge group is <= max cg.
2552      * This is determined by the type of the local exclusion type
2553      * Exclusions are stored in bits. (If the type is not large
2554      * enough, enlarge it, unsigned char -> unsigned short -> unsigned long)
2555      */
2556     maxcg = sizeof(t_excl)*8;
2557     if (nr_in_cg > maxcg)
2558     {
2559         gmx_fatal(FARGS, "Max #atoms in a charge group: %d > %d\n",
2560                   nr_in_cg, maxcg);
2561     }
2562
2563     ngid = mtop->groups.grps[egcENER].nr;
2564     snew(ns->bExcludeAlleg, ngid);
2565     for (i = 0; i < ngid; i++)
2566     {
2567         ns->bExcludeAlleg[i] = TRUE;
2568         for (j = 0; j < ngid; j++)
2569         {
2570             if (!(fr->egp_flags[i*ngid+j] & EGP_EXCL))
2571             {
2572                 ns->bExcludeAlleg[i] = FALSE;
2573             }
2574         }
2575     }
2576
2577     if (fr->bGrid)
2578     {
2579         /* Grid search */
2580         ns->grid = init_grid(fplog, fr);
2581         init_nsgrid_lists(fr, ngid, ns);
2582     }
2583     else
2584     {
2585         /* Simple search */
2586         snew(ns->ns_buf, ngid);
2587         for (i = 0; (i < ngid); i++)
2588         {
2589             snew(ns->ns_buf[i], SHIFTS);
2590         }
2591         ncg = ncg_mtop(mtop);
2592         snew(ns->simple_aaj, 2*ncg);
2593         for (jcg = 0; (jcg < ncg); jcg++)
2594         {
2595             ns->simple_aaj[jcg]     = jcg;
2596             ns->simple_aaj[jcg+ncg] = jcg;
2597         }
2598     }
2599
2600     /* Create array that determines whether or not atoms have VdW */
2601     snew(ns->bHaveVdW, fr->ntype);
2602     for (i = 0; (i < fr->ntype); i++)
2603     {
2604         for (j = 0; (j < fr->ntype); j++)
2605         {
2606             ns->bHaveVdW[i] = (ns->bHaveVdW[i] ||
2607                                (fr->bBHAM ?
2608                                 ((BHAMA(fr->nbfp, fr->ntype, i, j) != 0) ||
2609                                  (BHAMB(fr->nbfp, fr->ntype, i, j) != 0) ||
2610                                  (BHAMC(fr->nbfp, fr->ntype, i, j) != 0)) :
2611                                 ((C6(fr->nbfp, fr->ntype, i, j) != 0) ||
2612                                  (C12(fr->nbfp, fr->ntype, i, j) != 0))));
2613         }
2614     }
2615     if (debug)
2616     {
2617         pr_bvec(debug, 0, "bHaveVdW", ns->bHaveVdW, fr->ntype, TRUE);
2618     }
2619
2620     ns->nra_alloc = 0;
2621     ns->bexcl     = NULL;
2622     if (!DOMAINDECOMP(cr))
2623     {
2624         ns_realloc_natoms(ns, mtop->natoms);
2625     }
2626
2627     ns->nblist_initialized = FALSE;
2628
2629     /* nbr list debug dump */
2630     {
2631         char *ptr = getenv("GMX_DUMP_NL");
2632         if (ptr)
2633         {
2634             ns->dump_nl = strtol(ptr, NULL, 10);
2635             if (fplog)
2636             {
2637                 fprintf(fplog, "GMX_DUMP_NL = %d", ns->dump_nl);
2638             }
2639         }
2640         else
2641         {
2642             ns->dump_nl = 0;
2643         }
2644     }
2645 }
2646
2647
2648 int search_neighbours(FILE *log, t_forcerec *fr,
2649                       matrix box,
2650                       gmx_localtop_t *top,
2651                       gmx_groups_t *groups,
2652                       t_commrec *cr,
2653                       t_nrnb *nrnb, t_mdatoms *md,
2654                       gmx_bool bFillGrid,
2655                       gmx_bool bDoLongRangeNS)
2656 {
2657     t_block  *cgs = &(top->cgs);
2658     rvec     box_size, grid_x0, grid_x1;
2659     int      m, ngid;
2660     real     min_size, grid_dens;
2661     int      nsearch;
2662     gmx_bool     bGrid;
2663     int      start, end;
2664     gmx_ns_t *ns;
2665     t_grid   *grid;
2666     gmx_domdec_zones_t *dd_zones;
2667     put_in_list_t *put_in_list;
2668
2669     ns = &fr->ns;
2670
2671     /* Set some local variables */
2672     bGrid = fr->bGrid;
2673     ngid  = groups->grps[egcENER].nr;
2674
2675     for (m = 0; (m < DIM); m++)
2676     {
2677         box_size[m] = box[m][m];
2678     }
2679
2680     if (fr->ePBC != epbcNONE)
2681     {
2682         if (sqr(fr->rlistlong) >= max_cutoff2(fr->ePBC, box))
2683         {
2684             gmx_fatal(FARGS, "One of the box vectors has become shorter than twice the cut-off length or box_yy-|box_zy| or box_zz has become smaller than the cut-off.");
2685         }
2686         if (!bGrid)
2687         {
2688             min_size = std::min(box_size[XX], std::min(box_size[YY], box_size[ZZ]));
2689             if (2*fr->rlistlong >= min_size)
2690             {
2691                 gmx_fatal(FARGS, "One of the box diagonal elements has become smaller than twice the cut-off length.");
2692             }
2693         }
2694     }
2695
2696     if (DOMAINDECOMP(cr))
2697     {
2698         ns_realloc_natoms(ns, cgs->index[cgs->nr]);
2699     }
2700     debug_gmx();
2701
2702     /* Reset the neighbourlists */
2703     reset_neighbor_lists(fr, TRUE, TRUE);
2704
2705     if (bGrid && bFillGrid)
2706     {
2707
2708         grid = ns->grid;
2709         if (DOMAINDECOMP(cr))
2710         {
2711             dd_zones = domdec_zones(cr->dd);
2712         }
2713         else
2714         {
2715             dd_zones = NULL;
2716
2717             get_nsgrid_boundaries(grid->nboundeddim, box, NULL, NULL, NULL, NULL,
2718                                   cgs->nr, fr->cg_cm, grid_x0, grid_x1, &grid_dens);
2719
2720             grid_first(log, grid, NULL, NULL, box, grid_x0, grid_x1,
2721                        fr->rlistlong, grid_dens);
2722         }
2723         debug_gmx();
2724
2725         start = 0;
2726         end   = cgs->nr;
2727
2728         if (DOMAINDECOMP(cr))
2729         {
2730             end = cgs->nr;
2731             fill_grid(dd_zones, grid, end, -1, end, fr->cg_cm);
2732             grid->icg0 = 0;
2733             grid->icg1 = dd_zones->izone[dd_zones->nizone-1].cg1;
2734         }
2735         else
2736         {
2737             fill_grid(NULL, grid, cgs->nr, fr->cg0, fr->hcg, fr->cg_cm);
2738             grid->icg0 = fr->cg0;
2739             grid->icg1 = fr->hcg;
2740             debug_gmx();
2741         }
2742
2743         calc_elemnr(grid, start, end, cgs->nr);
2744         calc_ptrs(grid);
2745         grid_last(grid, start, end, cgs->nr);
2746
2747         if (gmx_debug_at)
2748         {
2749             check_grid(grid);
2750             print_grid(debug, grid);
2751         }
2752     }
2753     else if (fr->n_tpi)
2754     {
2755         /* Set the grid cell index for the test particle only.
2756          * The cell to cg index is not corrected, but that does not matter.
2757          */
2758         fill_grid(NULL, ns->grid, fr->hcg, fr->hcg-1, fr->hcg, fr->cg_cm);
2759     }
2760     debug_gmx();
2761
2762     if (fr->adress_type == eAdressOff)
2763     {
2764         if (!fr->ns.bCGlist)
2765         {
2766             put_in_list = put_in_list_at;
2767         }
2768         else
2769         {
2770             put_in_list = put_in_list_cg;
2771         }
2772     }
2773     else
2774     {
2775         put_in_list = put_in_list_adress;
2776     }
2777
2778     /* Do the core! */
2779     if (bGrid)
2780     {
2781         grid    = ns->grid;
2782         nsearch = nsgrid_core(cr, fr, box, ngid, top,
2783                               grid, ns->bexcl, ns->bExcludeAlleg,
2784                               md, put_in_list, ns->bHaveVdW,
2785                               bDoLongRangeNS, FALSE);
2786
2787         /* neighbour searching withouth QMMM! QM atoms have zero charge in
2788          * the classical calculation. The charge-charge interaction
2789          * between QM and MM atoms is handled in the QMMM core calculation
2790          * (see QMMM.c). The VDW however, we'd like to compute classically
2791          * and the QM MM atom pairs have just been put in the
2792          * corresponding neighbourlists. in case of QMMM we still need to
2793          * fill a special QMMM neighbourlist that contains all neighbours
2794          * of the QM atoms. If bQMMM is true, this list will now be made:
2795          */
2796         if (fr->bQMMM && fr->qr->QMMMscheme != eQMMMschemeoniom)
2797         {
2798             nsearch += nsgrid_core(cr, fr, box, ngid, top,
2799                                    grid, ns->bexcl, ns->bExcludeAlleg,
2800                                    md, put_in_list_qmmm, ns->bHaveVdW,
2801                                    bDoLongRangeNS, TRUE);
2802         }
2803     }
2804     else
2805     {
2806         nsearch = ns_simple_core(fr, top, md, box, box_size,
2807                                  ns->bexcl, ns->simple_aaj,
2808                                  ngid, ns->ns_buf, put_in_list, ns->bHaveVdW);
2809     }
2810     debug_gmx();
2811
2812 #ifdef DEBUG
2813     pr_nsblock(log);
2814 #endif
2815
2816     inc_nrnb(nrnb, eNR_NS, nsearch);
2817     /* inc_nrnb(nrnb,eNR_LR,fr->nlr); */
2818
2819     return nsearch;
2820 }