src/mdlib/sim_util.c

   1 /* -*- mode: c; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4; c-file-style: "stroustrup"; -*-
   2  *
   3  *
   4  *                This source code is part of
   5  *
   6  *                 G   R   O   M   A   C   S
   7  *
   8  *          GROningen MAchine for Chemical Simulations
   9  *
  10  *                        VERSION 3.2.0
  11  * Written by David van der Spoel, Erik Lindahl, Berk Hess, and others.
  12  * Copyright (c) 1991-2000, University of Groningen, The Netherlands.
  13  * Copyright (c) 2001-2004, The GROMACS development team,
  14  * check out http://www.gromacs.org for more information.
  15
  16  * This program is free software; you can redistribute it and/or
  17  * modify it under the terms of the GNU General Public License
  18  * as published by the Free Software Foundation; either version 2
  19  * of the License, or (at your option) any later version.
  20  *
  21  * If you want to redistribute modifications, please consider that
  22  * scientific software is very special. Version control is crucial -
  23  * bugs must be traceable. We will be happy to consider code for
  24  * inclusion in the official distribution, but derived work must not
  25  * be called official GROMACS. Details are found in the README & COPYING
  26  * files - if they are missing, get the official version at www.gromacs.org.
  27  *
  28  * To help us fund GROMACS development, we humbly ask that you cite
  29  * the papers on the package - you can find them in the top README file.
  30  *
  31  * For more info, check our website at http://www.gromacs.org
  32  *
  33  * And Hey:
  34  * GROwing Monsters And Cloning Shrimps
  35  */
  36 #ifdef HAVE_CONFIG_H
  37 #include <config.h>
  38 #endif
  39
  40 #ifdef GMX_CRAY_XT3
  41 #include<catamount/dclock.h>
  42 #endif
  43
  44
  45 #include <stdio.h>
  46 #include <time.h>
  47 #ifdef HAVE_SYS_TIME_H
  48 #include <sys/time.h>
  49 #endif
  50 #include <math.h>
  51 #include "typedefs.h"
  52 #include "string2.h"
  53 #include "gmxfio.h"
  54 #include "smalloc.h"
  55 #include "names.h"
  56 #include "confio.h"
  57 #include "mvdata.h"
  58 #include "txtdump.h"
  59 #include "pbc.h"
  60 #include "chargegroup.h"
  61 #include "vec.h"
  62 #include <time.h>
  63 #include "nrnb.h"
  64 #include "mshift.h"
  65 #include "mdrun.h"
  66 #include "update.h"
  67 #include "physics.h"
  68 #include "main.h"
  69 #include "mdatoms.h"
  70 #include "force.h"
  71 #include "bondf.h"
  72 #include "pme.h"
  73 #include "disre.h"
  74 #include "orires.h"
  75 #include "network.h"
  76 #include "calcmu.h"
  77 #include "constr.h"
  78 #include "xvgr.h"
  79 #include "trnio.h"
  80 #include "xtcio.h"
  81 #include "copyrite.h"
  82 #include "pull_rotation.h"
  83 #include "mpelogging.h"
  84 #include "domdec.h"
  85 #include "partdec.h"
  86 #include "gmx_wallcycle.h"
  87 #include "genborn.h"
  88
  89 #ifdef GMX_LIB_MPI
  90 #include <mpi.h>
  91 #endif
  92 #ifdef GMX_THREAD_MPI
  93 #include "tmpi.h"
  94 #endif
  95
  96 #include "adress.h"
  97 #include "qmmm.h"
  98
  99 #if 0
 100 typedef struct gmx_timeprint {
 101
 102 } t_gmx_timeprint;
 103 #endif
 104
 105 /* Portable version of ctime_r implemented in src/gmxlib/string2.c, but we do not want it declared in public installed headers */
 106 char *
 107 gmx_ctime_r(const time_t *clock,char *buf, int n);
 108
 109
 110 double
 111 gmx_gettime()
 112 {
 113 #ifdef HAVE_GETTIMEOFDAY
 114         struct timeval t;
 115         double seconds;
 116
 117         gettimeofday(&t,NULL);
 118
 119         seconds = (double) t.tv_sec + 1e-6*(double)t.tv_usec;
 120
 121         return seconds;
 122 #else
 123         double  seconds;
 124
 125         seconds = time(NULL);
 126
 127         return seconds;
 128 #endif
 129 }
 130
 131
 132 #define difftime(end,start) ((double)(end)-(double)(start))
 133
 134 void print_time(FILE *out,gmx_runtime_t *runtime,gmx_large_int_t step,
 135                 t_inputrec *ir, t_commrec *cr)
 136 {
 137     time_t finish;
 138     char   timebuf[STRLEN];
 139     double dt;
 140     char buf[48];
 141
 142 #ifndef GMX_THREAD_MPI
 143     if (!PAR(cr))
 144 #endif
 145     {
 146         fprintf(out,"\r");
 147     }
 148     fprintf(out,"step %s",gmx_step_str(step,buf));
 149     if ((step >= ir->nstlist))
 150     {
 151         if ((ir->nstlist == 0) || ((step % ir->nstlist) == 0))
 152         {
 153             /* We have done a full cycle let's update time_per_step */
 154             runtime->last = gmx_gettime();
 155             dt = difftime(runtime->last,runtime->real);
 156             runtime->time_per_step = dt/(step - ir->init_step + 1);
 157         }
 158         dt = (ir->nsteps + ir->init_step - step)*runtime->time_per_step;
 159
 160         if (ir->nsteps >= 0)
 161         {
 162             if (dt >= 300)
 163             {
 164                 finish = (time_t) (runtime->last + dt);
 165                 gmx_ctime_r(&finish,timebuf,STRLEN);
 166                 sprintf(buf,"%s",timebuf);
 167                 buf[strlen(buf)-1]='\0';
 168                 fprintf(out,", will finish %s",buf);
 169             }
 170             else
 171                 fprintf(out,", remaining runtime: %5d s          ",(int)dt);
 172         }
 173         else
 174         {
 175             fprintf(out," performance: %.1f ns/day    ",
 176                     ir->delta_t/1000*24*60*60/runtime->time_per_step);
 177         }
 178     }
 179 #ifndef GMX_THREAD_MPI
 180     if (PAR(cr))
 181     {
 182         fprintf(out,"\n");
 183     }
 184 #endif
 185
 186     fflush(out);
 187 }
 188
 189 #ifdef NO_CLOCK
 190 #define clock() -1
 191 #endif
 192
 193 static double set_proctime(gmx_runtime_t *runtime)
 194 {
 195     double diff;
 196 #ifdef GMX_CRAY_XT3
 197     double prev;
 198
 199     prev = runtime->proc;
 200     runtime->proc = dclock();
 201
 202     diff = runtime->proc - prev;
 203 #else
 204     clock_t prev;
 205
 206     prev = runtime->proc;
 207     runtime->proc = clock();
 208
 209     diff = (double)(runtime->proc - prev)/(double)CLOCKS_PER_SEC;
 210 #endif
 211     if (diff < 0)
 212     {
 213         /* The counter has probably looped, ignore this data */
 214         diff = 0;
 215     }
 216
 217     return diff;
 218 }
 219
 220 void runtime_start(gmx_runtime_t *runtime)
 221 {
 222     runtime->real = gmx_gettime();
 223     runtime->proc          = 0;
 224     set_proctime(runtime);
 225     runtime->realtime      = 0;
 226     runtime->proctime      = 0;
 227     runtime->last          = 0;
 228     runtime->time_per_step = 0;
 229 }
 230
 231 void runtime_end(gmx_runtime_t *runtime)
 232 {
 233     double now;
 234
 235     now = gmx_gettime();
 236
 237     runtime->proctime += set_proctime(runtime);
 238     runtime->realtime  = now - runtime->real;
 239     runtime->real      = now;
 240 }
 241
 242 void runtime_upd_proc(gmx_runtime_t *runtime)
 243 {
 244     runtime->proctime += set_proctime(runtime);
 245 }
 246
 247 void print_date_and_time(FILE *fplog,int nodeid,const char *title,
 248                          const gmx_runtime_t *runtime)
 249 {
 250     int i;
 251     char timebuf[STRLEN];
 252     char time_string[STRLEN];
 253     time_t tmptime;
 254
 255     if (fplog)
 256     {
 257         if (runtime != NULL)
 258         {
 259             tmptime = (time_t) runtime->real;
 260             gmx_ctime_r(&tmptime,timebuf,STRLEN);
 261         }
 262         else
 263         {
 264             tmptime = (time_t) gmx_gettime();
 265             gmx_ctime_r(&tmptime,timebuf,STRLEN);
 266         }
 267         for(i=0; timebuf[i]>=' '; i++)
 268         {
 269             time_string[i]=timebuf[i];
 270         }
 271         time_string[i]='\0';
 272
 273         fprintf(fplog,"%s on node %d %s\n",title,nodeid,time_string);
 274     }
 275 }
 276
 277 static void sum_forces(int start,int end,rvec f[],rvec flr[])
 278 {
 279   int i;
 280
 281   if (gmx_debug_at) {
 282     pr_rvecs(debug,0,"fsr",f+start,end-start);
 283     pr_rvecs(debug,0,"flr",flr+start,end-start);
 284   }
 285   for(i=start; (i<end); i++)
 286     rvec_inc(f[i],flr[i]);
 287 }
 288
 289 /*
 290  * calc_f_el calculates forces due to an electric field.
 291  *
 292  * force is kJ mol^-1 nm^-1 = e * kJ mol^-1 nm^-1 / e
 293  *
 294  * Et[] contains the parameters for the time dependent
 295  * part of the field (not yet used).
 296  * Ex[] contains the parameters for
 297  * the spatial dependent part of the field. You can have cool periodic
 298  * fields in principle, but only a constant field is supported
 299  * now.
 300  * The function should return the energy due to the electric field
 301  * (if any) but for now returns 0.
 302  *
 303  * WARNING:
 304  * There can be problems with the virial.
 305  * Since the field is not self-consistent this is unavoidable.
 306  * For neutral molecules the virial is correct within this approximation.
 307  * For neutral systems with many charged molecules the error is small.
 308  * But for systems with a net charge or a few charged molecules
 309  * the error can be significant when the field is high.
 310  * Solution: implement a self-consitent electric field into PME.
 311  */
 312 static void calc_f_el(FILE *fp,int  start,int homenr,
 313                       real charge[],rvec x[],rvec f[],
 314                       t_cosines Ex[],t_cosines Et[],double t)
 315 {
 316     rvec Ext;
 317     real t0;
 318     int  i,m;
 319
 320     for(m=0; (m<DIM); m++)
 321     {
 322         if (Et[m].n > 0)
 323         {
 324             if (Et[m].n == 3)
 325             {
 326                 t0 = Et[m].a[1];
 327                 Ext[m] = cos(Et[m].a[0]*(t-t0))*exp(-sqr(t-t0)/(2.0*sqr(Et[m].a[2])));
 328             }
 329             else
 330             {
 331                 Ext[m] = cos(Et[m].a[0]*t);
 332             }
 333         }
 334         else
 335         {
 336             Ext[m] = 1.0;
 337         }
 338         if (Ex[m].n > 0)
 339         {
 340             /* Convert the field strength from V/nm to MD-units */
 341             Ext[m] *= Ex[m].a[0]*FIELDFAC;
 342             for(i=start; (i<start+homenr); i++)
 343                 f[i][m] += charge[i]*Ext[m];
 344         }
 345         else
 346         {
 347             Ext[m] = 0;
 348         }
 349     }
 350     if (fp != NULL)
 351     {
 352         fprintf(fp,"%10g  %10g  %10g  %10g #FIELD\n",t,
 353                 Ext[XX]/FIELDFAC,Ext[YY]/FIELDFAC,Ext[ZZ]/FIELDFAC);
 354     }
 355 }
 356
 357 static void calc_virial(FILE *fplog,int start,int homenr,rvec x[],rvec f[],
 358                         tensor vir_part,t_graph *graph,matrix box,
 359                         t_nrnb *nrnb,const t_forcerec *fr,int ePBC)
 360 {
 361   int i,j;
 362   tensor virtest;
 363
 364   /* The short-range virial from surrounding boxes */
 365   clear_mat(vir_part);
 366   calc_vir(fplog,SHIFTS,fr->shift_vec,fr->fshift,vir_part,ePBC==epbcSCREW,box);
 367   inc_nrnb(nrnb,eNR_VIRIAL,SHIFTS);
 368
 369   /* Calculate partial virial, for local atoms only, based on short range.
 370    * Total virial is computed in global_stat, called from do_md
 371    */
 372   f_calc_vir(fplog,start,start+homenr,x,f,vir_part,graph,box);
 373   inc_nrnb(nrnb,eNR_VIRIAL,homenr);
 374
 375   /* Add position restraint contribution */
 376   for(i=0; i<DIM; i++) {
 377     vir_part[i][i] += fr->vir_diag_posres[i];
 378   }
 379
 380   /* Add wall contribution */
 381   for(i=0; i<DIM; i++) {
 382     vir_part[i][ZZ] += fr->vir_wall_z[i];
 383   }
 384
 385   if (debug)
 386     pr_rvecs(debug,0,"vir_part",vir_part,DIM);
 387 }
 388
 389 static void print_large_forces(FILE *fp,t_mdatoms *md,t_commrec *cr,
 390                                gmx_large_int_t step,real pforce,rvec *x,rvec *f)
 391 {
 392   int  i;
 393   real pf2,fn2;
 394   char buf[STEPSTRSIZE];
 395
 396   pf2 = sqr(pforce);
 397   for(i=md->start; i<md->start+md->homenr; i++) {
 398     fn2 = norm2(f[i]);
 399     /* We also catch NAN, if the compiler does not optimize this away. */
 400     if (fn2 >= pf2 || fn2 != fn2) {
 401       fprintf(fp,"step %s  atom %6d  x %8.3f %8.3f %8.3f  force %12.5e\n",
 402               gmx_step_str(step,buf),
 403               ddglatnr(cr->dd,i),x[i][XX],x[i][YY],x[i][ZZ],sqrt(fn2));
 404     }
 405   }
 406 }
 407
 408 void do_force(FILE *fplog,t_commrec *cr,
 409               t_inputrec *inputrec,
 410               gmx_large_int_t step,t_nrnb *nrnb,gmx_wallcycle_t wcycle,
 411               gmx_localtop_t *top,
 412               gmx_mtop_t *mtop,
 413               gmx_groups_t *groups,
 414               matrix box,rvec x[],history_t *hist,
 415               rvec f[],
 416               tensor vir_force,
 417               t_mdatoms *mdatoms,
 418               gmx_enerdata_t *enerd,t_fcdata *fcd,
 419               real lambda,t_graph *graph,
 420               t_forcerec *fr,gmx_vsite_t *vsite,rvec mu_tot,
 421               double t,FILE *field,gmx_edsam_t ed,
 422               gmx_bool bBornRadii,
 423               int flags)
 424 {
 425     int    cg0,cg1,i,j;
 426     int    start,homenr;
 427     double mu[2*DIM];
 428     gmx_bool   bSepDVDL,bStateChanged,bNS,bFillGrid,bCalcCGCM,bBS;
 429     gmx_bool   bDoLongRange,bDoForces,bSepLRF;
 430     gmx_bool   bDoAdressWF;
 431     matrix boxs;
 432     real   e,v,dvdl;
 433     t_pbc  pbc;
 434     float  cycles_ppdpme,cycles_pme,cycles_seppme,cycles_force;
 435
 436     start  = mdatoms->start;
 437     homenr = mdatoms->homenr;
 438
 439     bSepDVDL = (fr->bSepDVDL && do_per_step(step,inputrec->nstlog));
 440
 441     clear_mat(vir_force);
 442
 443     if (PARTDECOMP(cr))
 444     {
 445         pd_cg_range(cr,&cg0,&cg1);
 446     }
 447     else
 448     {
 449         cg0 = 0;
 450         if (DOMAINDECOMP(cr))
 451         {
 452             cg1 = cr->dd->ncg_tot;
 453         }
 454         else
 455         {
 456             cg1 = top->cgs.nr;
 457         }
 458         if (fr->n_tpi > 0)
 459         {
 460             cg1--;
 461         }
 462     }
 463
 464     bStateChanged = (flags & GMX_FORCE_STATECHANGED);
 465     bNS           = (flags & GMX_FORCE_NS) && (fr->bAllvsAll==FALSE);
 466     bFillGrid     = (bNS && bStateChanged);
 467     bCalcCGCM     = (bFillGrid && !DOMAINDECOMP(cr));
 468     bDoLongRange  = (fr->bTwinRange && bNS && (flags & GMX_FORCE_DOLR));
 469     bDoForces     = (flags & GMX_FORCE_FORCES);
 470     bSepLRF       = (bDoLongRange && bDoForces && (flags & GMX_FORCE_SEPLRF));
 471     /* should probably move this to the forcerec since it doesn't change */
 472     bDoAdressWF   = ((fr->adress_type!=eAdressOff));
 473
 474     if (bStateChanged)
 475     {
 476         update_forcerec(fplog,fr,box);
 477
 478         /* Calculate total (local) dipole moment in a temporary common array.
 479          * This makes it possible to sum them over nodes faster.
 480          */
 481         calc_mu(start,homenr,
 482                 x,mdatoms->chargeA,mdatoms->chargeB,mdatoms->nChargePerturbed,
 483                 mu,mu+DIM);
 484     }
 485
 486   if (fr->ePBC != epbcNONE) {
 487     /* Compute shift vectors every step,
 488      * because of pressure coupling or box deformation!
 489      */
 490     if ((flags & GMX_FORCE_DYNAMICBOX) && bStateChanged)
 491       calc_shifts(box,fr->shift_vec);
 492
 493     if (bCalcCGCM) {
 494       put_charge_groups_in_box(fplog,cg0,cg1,fr->ePBC,box,
 495                                &(top->cgs),x,fr->cg_cm);
 496       inc_nrnb(nrnb,eNR_CGCM,homenr);
 497       inc_nrnb(nrnb,eNR_RESETX,cg1-cg0);
 498     }
 499     else if (EI_ENERGY_MINIMIZATION(inputrec->eI) && graph) {
 500       unshift_self(graph,box,x);
 501     }
 502   }
 503   else if (bCalcCGCM) {
 504     calc_cgcm(fplog,cg0,cg1,&(top->cgs),x,fr->cg_cm);
 505     inc_nrnb(nrnb,eNR_CGCM,homenr);
 506   }
 507
 508   if (bCalcCGCM) {
 509     if (PAR(cr)) {
 510       move_cgcm(fplog,cr,fr->cg_cm);
 511     }
 512     if (gmx_debug_at)
 513       pr_rvecs(debug,0,"cgcm",fr->cg_cm,top->cgs.nr);
 514   }
 515
 516 #ifdef GMX_MPI
 517   if (!(cr->duty & DUTY_PME)) {
 518     /* Send particle coordinates to the pme nodes.
 519      * Since this is only implemented for domain decomposition
 520      * and domain decomposition does not use the graph,
 521      * we do not need to worry about shifting.
 522      */
 523
 524     wallcycle_start(wcycle,ewcPP_PMESENDX);
 525     GMX_MPE_LOG(ev_send_coordinates_start);
 526
 527     bBS = (inputrec->nwall == 2);
 528     if (bBS) {
 529       copy_mat(box,boxs);
 530       svmul(inputrec->wall_ewald_zfac,boxs[ZZ],boxs[ZZ]);
 531     }
 532
 533     gmx_pme_send_x(cr,bBS ? boxs : box,x,
 534                    mdatoms->nChargePerturbed,lambda,
 535                    ( flags & GMX_FORCE_VIRIAL),step);
 536
 537     GMX_MPE_LOG(ev_send_coordinates_finish);
 538     wallcycle_stop(wcycle,ewcPP_PMESENDX);
 539   }
 540 #endif /* GMX_MPI */
 541
 542     /* Communicate coordinates and sum dipole if necessary */
 543     if (PAR(cr))
 544     {
 545         wallcycle_start(wcycle,ewcMOVEX);
 546         if (DOMAINDECOMP(cr))
 547         {
 548             dd_move_x(cr->dd,box,x);
 549         }
 550         else
 551         {
 552             move_x(fplog,cr,GMX_LEFT,GMX_RIGHT,x,nrnb);
 553         }
 554         /* When we don't need the total dipole we sum it in global_stat */
 555         if (bStateChanged && NEED_MUTOT(*inputrec))
 556         {
 557             gmx_sumd(2*DIM,mu,cr);
 558         }
 559         wallcycle_stop(wcycle,ewcMOVEX);
 560     }
 561     if (bStateChanged)
 562     {
 563
 564         /* update adress weight beforehand */
 565         if(bDoAdressWF)
 566         {
 567             /* need pbc for adress weight calculation with pbc_dx */
 568             set_pbc(&pbc,inputrec->ePBC,box);
 569             if(fr->adress_site == eAdressSITEcog)
 570             {
 571                 update_adress_weights_cog(top->idef.iparams,top->idef.il,x,fr,mdatoms,
 572                                           inputrec->ePBC==epbcNONE ? NULL : &pbc);
 573             }
 574             else if (fr->adress_site == eAdressSITEcom)
 575             {
 576                 update_adress_weights_com(fplog,cg0,cg1,&(top->cgs),x,fr,mdatoms,
 577                                           inputrec->ePBC==epbcNONE ? NULL : &pbc);
 578             }
 579             else if (fr->adress_site == eAdressSITEatomatom){
 580                 update_adress_weights_atom_per_atom(cg0,cg1,&(top->cgs),x,fr,mdatoms,
 581                                           inputrec->ePBC==epbcNONE ? NULL : &pbc);
 582             }
 583             else
 584             {
 585                 update_adress_weights_atom(cg0,cg1,&(top->cgs),x,fr,mdatoms,
 586                                            inputrec->ePBC==epbcNONE ? NULL : &pbc);
 587             }
 588         }
 589
 590         for(i=0; i<2; i++)
 591         {
 592             for(j=0;j<DIM;j++)
 593             {
 594                 fr->mu_tot[i][j] = mu[i*DIM + j];
 595             }
 596         }
 597     }
 598     if (fr->efep == efepNO)
 599     {
 600         copy_rvec(fr->mu_tot[0],mu_tot);
 601     }
 602     else
 603     {
 604         for(j=0; j<DIM; j++)
 605         {
 606             mu_tot[j] =
 607                 (1.0 - lambda)*fr->mu_tot[0][j] + lambda*fr->mu_tot[1][j];
 608         }
 609     }
 610
 611     /* Reset energies */
 612     reset_enerdata(&(inputrec->opts),fr,bNS,enerd,MASTER(cr));
 613     clear_rvecs(SHIFTS,fr->fshift);
 614
 615     if (bNS)
 616     {
 617         wallcycle_start(wcycle,ewcNS);
 618
 619         if (graph && bStateChanged)
 620         {
 621             /* Calculate intramolecular shift vectors to make molecules whole */
 622             mk_mshift(fplog,graph,fr->ePBC,box,x);
 623         }
 624
 625         /* Reset long range forces if necessary */
 626         if (fr->bTwinRange)
 627         {
 628             /* Reset the (long-range) forces if necessary */
 629             clear_rvecs(fr->natoms_force_constr,bSepLRF ? fr->f_twin : f);
 630         }
 631
 632         /* Do the actual neighbour searching and if twin range electrostatics
 633          * also do the calculation of long range forces and energies.
 634          */
 635         dvdl = 0;
 636         ns(fplog,fr,x,box,
 637            groups,&(inputrec->opts),top,mdatoms,
 638            cr,nrnb,lambda,&dvdl,&enerd->grpp,bFillGrid,
 639            bDoLongRange,bDoForces,bSepLRF ? fr->f_twin : f);
 640         if (bSepDVDL)
 641         {
 642             fprintf(fplog,sepdvdlformat,"LR non-bonded",0.0,dvdl);
 643         }
 644         enerd->dvdl_lin += dvdl;
 645
 646         wallcycle_stop(wcycle,ewcNS);
 647     }
 648
 649     if (inputrec->implicit_solvent && bNS)
 650     {
 651         make_gb_nblist(cr,inputrec->gb_algorithm,inputrec->rlist,
 652                        x,box,fr,&top->idef,graph,fr->born);
 653     }
 654
 655     if (DOMAINDECOMP(cr))
 656     {
 657         if (!(cr->duty & DUTY_PME))
 658         {
 659             wallcycle_start(wcycle,ewcPPDURINGPME);
 660             dd_force_flop_start(cr->dd,nrnb);
 661         }
 662     }
 663
 664     if (inputrec->bRot)
 665     {
 666         /* Enforced rotation has its own cycle counter that starts after the collective
 667          * coordinates have been communicated. It is added to ddCyclF to allow
 668          * for proper load-balancing */
 669         wallcycle_start(wcycle,ewcROT);
 670         do_rotation(cr,inputrec,box,x,t,step,wcycle,bNS);
 671         wallcycle_stop(wcycle,ewcROT);
 672     }
 673
 674     /* Start the force cycle counter.
 675      * This counter is stopped in do_forcelow_level.
 676      * No parallel communication should occur while this counter is running,
 677      * since that will interfere with the dynamic load balancing.
 678      */
 679     wallcycle_start(wcycle,ewcFORCE);
 680
 681     if (bDoForces)
 682     {
 683         /* Reset forces for which the virial is calculated separately:
 684          * PME/Ewald forces if necessary */
 685         if (fr->bF_NoVirSum)
 686         {
 687             if (flags & GMX_FORCE_VIRIAL)
 688             {
 689                 fr->f_novirsum = fr->f_novirsum_alloc;
 690                 GMX_BARRIER(cr->mpi_comm_mygroup);
 691                 if (fr->bDomDec)
 692                 {
 693                     clear_rvecs(fr->f_novirsum_n,fr->f_novirsum);
 694                 }
 695                 else
 696                 {
 697                     clear_rvecs(homenr,fr->f_novirsum+start);
 698                 }
 699                 GMX_BARRIER(cr->mpi_comm_mygroup);
 700             }
 701             else
 702             {
 703                 /* We are not calculating the pressure so we do not need
 704                  * a separate array for forces that do not contribute
 705                  * to the pressure.
 706                  */
 707                 fr->f_novirsum = f;
 708             }
 709         }
 710
 711         if (bSepLRF)
 712         {
 713             /* Add the long range forces to the short range forces */
 714             for(i=0; i<fr->natoms_force_constr; i++)
 715             {
 716                 copy_rvec(fr->f_twin[i],f[i]);
 717             }
 718         }
 719         else if (!(fr->bTwinRange && bNS))
 720         {
 721             /* Clear the short-range forces */
 722             clear_rvecs(fr->natoms_force_constr,f);
 723         }
 724
 725         clear_rvec(fr->vir_diag_posres);
 726
 727         GMX_BARRIER(cr->mpi_comm_mygroup);
 728     }
 729     if (inputrec->ePull == epullCONSTRAINT)
 730     {
 731         clear_pull_forces(inputrec->pull);
 732     }
 733
 734     /* update QMMMrec, if necessary */
 735     if(fr->bQMMM)
 736     {
 737         update_QMMMrec(cr,fr,x,mdatoms,box,top);
 738     }
 739
 740     if ((flags & GMX_FORCE_BONDED) && top->idef.il[F_POSRES].nr > 0)
 741     {
 742         /* Position restraints always require full pbc. Check if we already did it for Adress */
 743         if(!(bStateChanged && bDoAdressWF))
 744         {
 745             set_pbc(&pbc,inputrec->ePBC,box);
 746         }
 747         v = posres(top->idef.il[F_POSRES].nr,top->idef.il[F_POSRES].iatoms,
 748                    top->idef.iparams_posres,
 749                    (const rvec*)x,fr->f_novirsum,fr->vir_diag_posres,
 750                    inputrec->ePBC==epbcNONE ? NULL : &pbc,lambda,&dvdl,
 751                    fr->rc_scaling,fr->ePBC,fr->posres_com,fr->posres_comB);
 752         if (bSepDVDL)
 753         {
 754             fprintf(fplog,sepdvdlformat,
 755                     interaction_function[F_POSRES].longname,v,dvdl);
 756         }
 757         enerd->term[F_POSRES] += v;
 758         /* This linear lambda dependence assumption is only correct
 759          * when only k depends on lambda,
 760          * not when the reference position depends on lambda.
 761          * grompp checks for this.
 762          */
 763         enerd->dvdl_lin += dvdl;
 764         inc_nrnb(nrnb,eNR_POSRES,top->idef.il[F_POSRES].nr/2);
 765     }
 766
 767     /* Compute the bonded and non-bonded energies and optionally forces */
 768     do_force_lowlevel(fplog,step,fr,inputrec,&(top->idef),
 769                       cr,nrnb,wcycle,mdatoms,&(inputrec->opts),
 770                       x,hist,f,enerd,fcd,mtop,top,fr->born,
 771                       &(top->atomtypes),bBornRadii,box,
 772                       lambda,graph,&(top->excls),fr->mu_tot,
 773                       flags,&cycles_pme);
 774
 775     cycles_force = wallcycle_stop(wcycle,ewcFORCE);
 776     GMX_BARRIER(cr->mpi_comm_mygroup);
 777
 778     if (ed)
 779     {
 780         do_flood(fplog,cr,x,f,ed,box,step);
 781     }
 782
 783     if (DOMAINDECOMP(cr))
 784     {
 785         dd_force_flop_stop(cr->dd,nrnb);
 786         if (wcycle)
 787         {
 788             dd_cycles_add(cr->dd,cycles_force-cycles_pme,ddCyclF);
 789         }
 790     }
 791
 792     if (bDoForces)
 793     {
 794         if (IR_ELEC_FIELD(*inputrec))
 795         {
 796             /* Compute forces due to electric field */
 797             calc_f_el(MASTER(cr) ? field : NULL,
 798                       start,homenr,mdatoms->chargeA,x,fr->f_novirsum,
 799                       inputrec->ex,inputrec->et,t);
 800         }
 801
 802         if (bDoAdressWF && fr->adress_icor == eAdressICThermoForce)
 803         {
 804             /* Compute thermodynamic force in hybrid AdResS region */
 805             adress_thermo_force(start,homenr,&(top->cgs),x,fr->f_novirsum,fr,mdatoms,
 806                                 inputrec->ePBC==epbcNONE ? NULL : &pbc);
 807         }
 808
 809         /* Communicate the forces */
 810         if (PAR(cr))
 811         {
 812             wallcycle_start(wcycle,ewcMOVEF);
 813             if (DOMAINDECOMP(cr))
 814             {
 815                 dd_move_f(cr->dd,f,fr->fshift);
 816                 /* Do we need to communicate the separate force array
 817                  * for terms that do not contribute to the single sum virial?
 818                  * Position restraints and electric fields do not introduce
 819                  * inter-cg forces, only full electrostatics methods do.
 820                  * When we do not calculate the virial, fr->f_novirsum = f,
 821                  * so we have already communicated these forces.
 822                  */
 823                 if (EEL_FULL(fr->eeltype) && cr->dd->n_intercg_excl &&
 824                     (flags & GMX_FORCE_VIRIAL))
 825                 {
 826                     dd_move_f(cr->dd,fr->f_novirsum,NULL);
 827                 }
 828                 if (bSepLRF)
 829                 {
 830                     /* We should not update the shift forces here,
 831                      * since f_twin is already included in f.
 832                      */
 833                     dd_move_f(cr->dd,fr->f_twin,NULL);
 834                 }
 835             }
 836             else
 837             {
 838                 pd_move_f(cr,f,nrnb);
 839                 if (bSepLRF)
 840                 {
 841                     pd_move_f(cr,fr->f_twin,nrnb);
 842                 }
 843             }
 844             wallcycle_stop(wcycle,ewcMOVEF);
 845         }
 846
 847         /* If we have NoVirSum forces, but we do not calculate the virial,
 848          * we sum fr->f_novirum=f later.
 849          */
 850         if (vsite && !(fr->bF_NoVirSum && !(flags & GMX_FORCE_VIRIAL)))
 851         {
 852             wallcycle_start(wcycle,ewcVSITESPREAD);
 853             spread_vsite_f(fplog,vsite,x,f,fr->fshift,FALSE,NULL,nrnb,
 854                            &top->idef,fr->ePBC,fr->bMolPBC,graph,box,cr);
 855             wallcycle_stop(wcycle,ewcVSITESPREAD);
 856
 857             if (bSepLRF)
 858             {
 859                 wallcycle_start(wcycle,ewcVSITESPREAD);
 860                 spread_vsite_f(fplog,vsite,x,fr->f_twin,NULL,FALSE,NULL,
 861                                nrnb,
 862                                &top->idef,fr->ePBC,fr->bMolPBC,graph,box,cr);
 863                 wallcycle_stop(wcycle,ewcVSITESPREAD);
 864             }
 865         }
 866
 867         if (flags & GMX_FORCE_VIRIAL)
 868         {
 869             /* Calculation of the virial must be done after vsites! */
 870             calc_virial(fplog,mdatoms->start,mdatoms->homenr,x,f,
 871                         vir_force,graph,box,nrnb,fr,inputrec->ePBC);
 872         }
 873     }
 874
 875     enerd->term[F_COM_PULL] = 0;
 876     if (inputrec->ePull == epullUMBRELLA || inputrec->ePull == epullCONST_F)
 877     {
 878         /* Calculate the center of mass forces, this requires communication,
 879          * which is why pull_potential is called close to other communication.
 880          * The virial contribution is calculated directly,
 881          * which is why we call pull_potential after calc_virial.
 882          */
 883         set_pbc(&pbc,inputrec->ePBC,box);
 884         dvdl = 0;
 885         enerd->term[F_COM_PULL] +=
 886             pull_potential(inputrec->ePull,inputrec->pull,mdatoms,&pbc,
 887                            cr,t,lambda,x,f,vir_force,&dvdl);
 888         if (bSepDVDL)
 889         {
 890             fprintf(fplog,sepdvdlformat,"Com pull",enerd->term[F_COM_PULL],dvdl);
 891         }
 892         enerd->dvdl_lin += dvdl;
 893     }
 894
 895     /* Add the forces from enforced rotation potentials (if any) */
 896     if (inputrec->bRot)
 897     {
 898         wallcycle_start(wcycle,ewcROTadd);
 899         enerd->term[F_COM_PULL] += add_rot_forces(inputrec->rot, f, cr,step,t);
 900         wallcycle_stop(wcycle,ewcROTadd);
 901     }
 902
 903     if (PAR(cr) && !(cr->duty & DUTY_PME))
 904     {
 905         cycles_ppdpme = wallcycle_stop(wcycle,ewcPPDURINGPME);
 906         dd_cycles_add(cr->dd,cycles_ppdpme,ddCyclPPduringPME);
 907
 908         /* In case of node-splitting, the PP nodes receive the long-range
 909          * forces, virial and energy from the PME nodes here.
 910          */
 911         wallcycle_start(wcycle,ewcPP_PMEWAITRECVF);
 912         dvdl = 0;
 913         gmx_pme_receive_f(cr,fr->f_novirsum,fr->vir_el_recip,&e,&dvdl,
 914                           &cycles_seppme);
 915         if (bSepDVDL)
 916         {
 917             fprintf(fplog,sepdvdlformat,"PME mesh",e,dvdl);
 918         }
 919         enerd->term[F_COUL_RECIP] += e;
 920         enerd->dvdl_lin += dvdl;
 921         if (wcycle)
 922         {
 923             dd_cycles_add(cr->dd,cycles_seppme,ddCyclPME);
 924         }
 925         wallcycle_stop(wcycle,ewcPP_PMEWAITRECVF);
 926     }
 927
 928     if (bDoForces && fr->bF_NoVirSum)
 929     {
 930         if (vsite)
 931         {
 932             /* Spread the mesh force on virtual sites to the other particles...
 933              * This is parallellized. MPI communication is performed
 934              * if the constructing atoms aren't local.
 935              */
 936             wallcycle_start(wcycle,ewcVSITESPREAD);
 937             spread_vsite_f(fplog,vsite,x,fr->f_novirsum,NULL,
 938                            (flags & GMX_FORCE_VIRIAL),fr->vir_el_recip,
 939                            nrnb,
 940                            &top->idef,fr->ePBC,fr->bMolPBC,graph,box,cr);
 941             wallcycle_stop(wcycle,ewcVSITESPREAD);
 942         }
 943         if (flags & GMX_FORCE_VIRIAL)
 944         {
 945             /* Now add the forces, this is local */
 946             if (fr->bDomDec)
 947             {
 948                 sum_forces(0,fr->f_novirsum_n,f,fr->f_novirsum);
 949             }
 950             else
 951             {
 952                 sum_forces(start,start+homenr,f,fr->f_novirsum);
 953             }
 954             if (EEL_FULL(fr->eeltype))
 955             {
 956                 /* Add the mesh contribution to the virial */
 957                 m_add(vir_force,fr->vir_el_recip,vir_force);
 958             }
 959             if (debug)
 960             {
 961                 pr_rvecs(debug,0,"vir_force",vir_force,DIM);
 962             }
 963         }
 964     }
 965
 966     /* Sum the potential energy terms from group contributions */
 967     sum_epot(&(inputrec->opts),enerd);
 968
 969     if (fr->print_force >= 0 && bDoForces)
 970     {
 971         print_large_forces(stderr,mdatoms,cr,step,fr->print_force,x,f);
 972     }
 973 }
 974
 975 void do_constrain_first(FILE *fplog,gmx_constr_t constr,
 976                         t_inputrec *ir,t_mdatoms *md,
 977                         t_state *state,rvec *f,
 978                         t_graph *graph,t_commrec *cr,t_nrnb *nrnb,
 979                         t_forcerec *fr, gmx_localtop_t *top, tensor shake_vir)
 980 {
 981     int    i,m,start,end;
 982     gmx_large_int_t step;
 983     real   dt=ir->delta_t;
 984     real   dvdlambda;
 985     rvec   *savex;
 986
 987     snew(savex,state->natoms);
 988
 989     start = md->start;
 990     end   = md->homenr + start;
 991
 992     if (debug)
 993         fprintf(debug,"vcm: start=%d, homenr=%d, end=%d\n",
 994                 start,md->homenr,end);
 995     /* Do a first constrain to reset particles... */
 996     step = ir->init_step;
 997     if (fplog)
 998     {
 999         char buf[STEPSTRSIZE];
1000         fprintf(fplog,"\nConstraining the starting coordinates (step %s)\n",
1001                 gmx_step_str(step,buf));
1002     }
1003     dvdlambda = 0;
1004
1005     /* constrain the current position */
1006     constrain(NULL,TRUE,FALSE,constr,&(top->idef),
1007               ir,NULL,cr,step,0,md,
1008               state->x,state->x,NULL,
1009               state->box,state->lambda,&dvdlambda,
1010               NULL,NULL,nrnb,econqCoord,ir->epc==epcMTTK,state->veta,state->veta);
1011     if (EI_VV(ir->eI))
1012     {
1013         /* constrain the inital velocity, and save it */
1014         /* also may be useful if we need the ekin from the halfstep for velocity verlet */
1015         /* might not yet treat veta correctly */
1016         constrain(NULL,TRUE,FALSE,constr,&(top->idef),
1017                   ir,NULL,cr,step,0,md,
1018                   state->x,state->v,state->v,
1019                   state->box,state->lambda,&dvdlambda,
1020                   NULL,NULL,nrnb,econqVeloc,ir->epc==epcMTTK,state->veta,state->veta);
1021     }
1022     /* constrain the inital velocities at t-dt/2 */
1023     if (EI_STATE_VELOCITY(ir->eI) && ir->eI!=eiVV)
1024     {
1025         for(i=start; (i<end); i++)
1026         {
1027             for(m=0; (m<DIM); m++)
1028             {
1029                 /* Reverse the velocity */
1030                 state->v[i][m] = -state->v[i][m];
1031                 /* Store the position at t-dt in buf */
1032                 savex[i][m] = state->x[i][m] + dt*state->v[i][m];
1033             }
1034         }
1035     /* Shake the positions at t=-dt with the positions at t=0
1036      * as reference coordinates.
1037          */
1038         if (fplog)
1039         {
1040             char buf[STEPSTRSIZE];
1041             fprintf(fplog,"\nConstraining the coordinates at t0-dt (step %s)\n",
1042                     gmx_step_str(step,buf));
1043         }
1044         dvdlambda = 0;
1045         constrain(NULL,TRUE,FALSE,constr,&(top->idef),
1046                   ir,NULL,cr,step,-1,md,
1047                   state->x,savex,NULL,
1048                   state->box,state->lambda,&dvdlambda,
1049                   state->v,NULL,nrnb,econqCoord,ir->epc==epcMTTK,state->veta,state->veta);
1050
1051         for(i=start; i<end; i++) {
1052             for(m=0; m<DIM; m++) {
1053                 /* Re-reverse the velocities */
1054                 state->v[i][m] = -state->v[i][m];
1055             }
1056         }
1057     }
1058
1059     sfree(savex);
1060 }
1061
1062 void calc_enervirdiff(FILE *fplog,int eDispCorr,t_forcerec *fr)
1063 {
1064   double eners[2],virs[2],enersum,virsum,y0,f,g,h;
1065   double r0,r1,r,rc3,rc9,ea,eb,ec,pa,pb,pc,pd;
1066   double invscale,invscale2,invscale3;
1067   int    ri0,ri1,ri,i,offstart,offset;
1068   real   scale,*vdwtab;
1069
1070   fr->enershiftsix = 0;
1071   fr->enershifttwelve = 0;
1072   fr->enerdiffsix = 0;
1073   fr->enerdifftwelve = 0;
1074   fr->virdiffsix = 0;
1075   fr->virdifftwelve = 0;
1076
1077   if (eDispCorr != edispcNO) {
1078     for(i=0; i<2; i++) {
1079       eners[i] = 0;
1080       virs[i]  = 0;
1081     }
1082     if ((fr->vdwtype == evdwSWITCH) || (fr->vdwtype == evdwSHIFT)) {
1083       if (fr->rvdw_switch == 0)
1084         gmx_fatal(FARGS,
1085                   "With dispersion correction rvdw-switch can not be zero "
1086                   "for vdw-type = %s",evdw_names[fr->vdwtype]);
1087
1088       scale  = fr->nblists[0].tab.scale;
1089       vdwtab = fr->nblists[0].vdwtab;
1090
1091       /* Round the cut-offs to exact table values for precision */
1092       ri0 = floor(fr->rvdw_switch*scale);
1093       ri1 = ceil(fr->rvdw*scale);
1094       r0  = ri0/scale;
1095       r1  = ri1/scale;
1096       rc3 = r0*r0*r0;
1097       rc9  = rc3*rc3*rc3;
1098
1099       if (fr->vdwtype == evdwSHIFT) {
1100         /* Determine the constant energy shift below rvdw_switch */
1101         fr->enershiftsix    = (real)(-1.0/(rc3*rc3)) - vdwtab[8*ri0];
1102         fr->enershifttwelve = (real)( 1.0/(rc9*rc3)) - vdwtab[8*ri0 + 4];
1103       }
1104       /* Add the constant part from 0 to rvdw_switch.
1105        * This integration from 0 to rvdw_switch overcounts the number
1106        * of interactions by 1, as it also counts the self interaction.
1107        * We will correct for this later.
1108        */
1109       eners[0] += 4.0*M_PI*fr->enershiftsix*rc3/3.0;
1110       eners[1] += 4.0*M_PI*fr->enershifttwelve*rc3/3.0;
1111
1112       invscale = 1.0/(scale);
1113       invscale2 = invscale*invscale;
1114       invscale3 = invscale*invscale2;
1115
1116       /* following summation derived from cubic spline definition,
1117         Numerical Recipies in C, second edition, p. 113-116.  Exact
1118         for the cubic spline.  We first calculate the negative of
1119         the energy from rvdw to rvdw_switch, assuming that g(r)=1,
1120         and then add the more standard, abrupt cutoff correction to
1121         that result, yielding the long-range correction for a
1122         switched function.  We perform both the pressure and energy
1123         loops at the same time for simplicity, as the computational
1124         cost is low. */
1125
1126       for (i=0;i<2;i++) {
1127         enersum = 0.0; virsum = 0.0;
1128         if (i==0)
1129           offstart = 0;
1130         else
1131           offstart = 4;
1132         for (ri=ri0; ri<ri1; ri++) {
1133           r = ri*invscale;
1134           ea = invscale3;
1135           eb = 2.0*invscale2*r;
1136           ec = invscale*r*r;
1137
1138           pa = invscale3;
1139           pb = 3.0*invscale2*r;
1140           pc = 3.0*invscale*r*r;
1141           pd = r*r*r;
1142
1143           /* this "8" is from the packing in the vdwtab array - perhaps
1144             should be #define'ed? */
1145           offset = 8*ri + offstart;
1146           y0 = vdwtab[offset];
1147           f = vdwtab[offset+1];
1148           g = vdwtab[offset+2];
1149           h = vdwtab[offset+3];
1150
1151           enersum += y0*(ea/3 + eb/2 + ec) + f*(ea/4 + eb/3 + ec/2)+
1152             g*(ea/5 + eb/4 + ec/3) + h*(ea/6 + eb/5 + ec/4);
1153           virsum  +=  f*(pa/4 + pb/3 + pc/2 + pd) +
1154             2*g*(pa/5 + pb/4 + pc/3 + pd/2) + 3*h*(pa/6 + pb/5 + pc/4 + pd/3);
1155
1156         }
1157         enersum *= 4.0*M_PI;
1158         virsum  *= 4.0*M_PI;
1159         eners[i] -= enersum;
1160         virs[i]  -= virsum;
1161       }
1162
1163       /* now add the correction for rvdw_switch to infinity */
1164       eners[0] += -4.0*M_PI/(3.0*rc3);
1165       eners[1] +=  4.0*M_PI/(9.0*rc9);
1166       virs[0]  +=  8.0*M_PI/rc3;
1167       virs[1]  += -16.0*M_PI/(3.0*rc9);
1168     }
1169     else if ((fr->vdwtype == evdwCUT) || (fr->vdwtype == evdwUSER)) {
1170       if (fr->vdwtype == evdwUSER && fplog)
1171         fprintf(fplog,
1172                 "WARNING: using dispersion correction with user tables\n");
1173       rc3  = fr->rvdw*fr->rvdw*fr->rvdw;
1174       rc9  = rc3*rc3*rc3;
1175       eners[0] += -4.0*M_PI/(3.0*rc3);
1176       eners[1] +=  4.0*M_PI/(9.0*rc9);
1177       virs[0]  +=  8.0*M_PI/rc3;
1178       virs[1]  += -16.0*M_PI/(3.0*rc9);
1179     } else {
1180       gmx_fatal(FARGS,
1181                 "Dispersion correction is not implemented for vdw-type = %s",
1182                 evdw_names[fr->vdwtype]);
1183     }
1184     fr->enerdiffsix    = eners[0];
1185     fr->enerdifftwelve = eners[1];
1186     /* The 0.5 is due to the Gromacs definition of the virial */
1187     fr->virdiffsix     = 0.5*virs[0];
1188     fr->virdifftwelve  = 0.5*virs[1];
1189   }
1190 }
1191
1192 void calc_dispcorr(FILE *fplog,t_inputrec *ir,t_forcerec *fr,
1193                    gmx_large_int_t step,int natoms,
1194                    matrix box,real lambda,tensor pres,tensor virial,
1195                    real *prescorr, real *enercorr, real *dvdlcorr)
1196 {
1197     gmx_bool bCorrAll,bCorrPres;
1198     real dvdlambda,invvol,dens,ninter,avcsix,avctwelve,enerdiff,svir=0,spres=0;
1199     int  m;
1200
1201     *prescorr = 0;
1202     *enercorr = 0;
1203     *dvdlcorr = 0;
1204
1205     clear_mat(virial);
1206     clear_mat(pres);
1207
1208     if (ir->eDispCorr != edispcNO) {
1209         bCorrAll  = (ir->eDispCorr == edispcAllEner ||
1210                      ir->eDispCorr == edispcAllEnerPres);
1211         bCorrPres = (ir->eDispCorr == edispcEnerPres ||
1212                      ir->eDispCorr == edispcAllEnerPres);
1213
1214         invvol = 1/det(box);
1215         if (fr->n_tpi)
1216         {
1217             /* Only correct for the interactions with the inserted molecule */
1218             dens = (natoms - fr->n_tpi)*invvol;
1219             ninter = fr->n_tpi;
1220         }
1221         else
1222         {
1223             dens = natoms*invvol;
1224             ninter = 0.5*natoms;
1225         }
1226
1227         if (ir->efep == efepNO)
1228         {
1229             avcsix    = fr->avcsix[0];
1230             avctwelve = fr->avctwelve[0];
1231         }
1232         else
1233         {
1234             avcsix    = (1 - lambda)*fr->avcsix[0]    + lambda*fr->avcsix[1];
1235             avctwelve = (1 - lambda)*fr->avctwelve[0] + lambda*fr->avctwelve[1];
1236         }
1237
1238         enerdiff = ninter*(dens*fr->enerdiffsix - fr->enershiftsix);
1239         *enercorr += avcsix*enerdiff;
1240         dvdlambda = 0.0;
1241         if (ir->efep != efepNO)
1242         {
1243             dvdlambda += (fr->avcsix[1] - fr->avcsix[0])*enerdiff;
1244         }
1245         if (bCorrAll)
1246         {
1247             enerdiff = ninter*(dens*fr->enerdifftwelve - fr->enershifttwelve);
1248             *enercorr += avctwelve*enerdiff;
1249             if (fr->efep != efepNO)
1250             {
1251                 dvdlambda += (fr->avctwelve[1] - fr->avctwelve[0])*enerdiff;
1252             }
1253         }
1254
1255         if (bCorrPres)
1256         {
1257             svir = ninter*dens*avcsix*fr->virdiffsix/3.0;
1258             if (ir->eDispCorr == edispcAllEnerPres)
1259             {
1260                 svir += ninter*dens*avctwelve*fr->virdifftwelve/3.0;
1261             }
1262             /* The factor 2 is because of the Gromacs virial definition */
1263             spres = -2.0*invvol*svir*PRESFAC;
1264
1265             for(m=0; m<DIM; m++) {
1266                 virial[m][m] += svir;
1267                 pres[m][m] += spres;
1268             }
1269             *prescorr += spres;
1270         }
1271
1272         /* Can't currently control when it prints, for now, just print when degugging */
1273         if (debug)
1274         {
1275             if (bCorrAll) {
1276                 fprintf(debug,"Long Range LJ corr.: <C6> %10.4e, <C12> %10.4e\n",
1277                         avcsix,avctwelve);
1278             }
1279             if (bCorrPres)
1280             {
1281                 fprintf(debug,
1282                         "Long Range LJ corr.: Epot %10g, Pres: %10g, Vir: %10g\n",
1283                         *enercorr,spres,svir);
1284             }
1285             else
1286             {
1287                 fprintf(debug,"Long Range LJ corr.: Epot %10g\n",*enercorr);
1288             }
1289         }
1290
1291         if (fr->bSepDVDL && do_per_step(step,ir->nstlog))
1292         {
1293             fprintf(fplog,sepdvdlformat,"Dispersion correction",
1294                     *enercorr,dvdlambda);
1295         }
1296         if (fr->efep != efepNO)
1297         {
1298             *dvdlcorr += dvdlambda;
1299         }
1300     }
1301 }
1302
1303 void do_pbc_first(FILE *fplog,matrix box,t_forcerec *fr,
1304                   t_graph *graph,rvec x[])
1305 {
1306   if (fplog)
1307     fprintf(fplog,"Removing pbc first time\n");
1308   calc_shifts(box,fr->shift_vec);
1309   if (graph) {
1310     mk_mshift(fplog,graph,fr->ePBC,box,x);
1311     if (gmx_debug_at)
1312       p_graph(debug,"do_pbc_first 1",graph);
1313     shift_self(graph,box,x);
1314     /* By doing an extra mk_mshift the molecules that are broken
1315      * because they were e.g. imported from another software
1316      * will be made whole again. Such are the healing powers
1317      * of GROMACS.
1318      */
1319     mk_mshift(fplog,graph,fr->ePBC,box,x);
1320     if (gmx_debug_at)
1321       p_graph(debug,"do_pbc_first 2",graph);
1322   }
1323   if (fplog)
1324     fprintf(fplog,"Done rmpbc\n");
1325 }
1326
1327 static void low_do_pbc_mtop(FILE *fplog,int ePBC,matrix box,
1328                             gmx_mtop_t *mtop,rvec x[],
1329                             gmx_bool bFirst)
1330 {
1331   t_graph *graph;
1332   int mb,as,mol;
1333   gmx_molblock_t *molb;
1334
1335   if (bFirst && fplog)
1336     fprintf(fplog,"Removing pbc first time\n");
1337
1338   snew(graph,1);
1339   as = 0;
1340   for(mb=0; mb<mtop->nmolblock; mb++) {
1341     molb = &mtop->molblock[mb];
1342     if (molb->natoms_mol == 1 ||
1343         (!bFirst && mtop->moltype[molb->type].cgs.nr == 1)) {
1344       /* Just one atom or charge group in the molecule, no PBC required */
1345       as += molb->nmol*molb->natoms_mol;
1346     } else {
1347       /* Pass NULL iso fplog to avoid graph prints for each molecule type */
1348       mk_graph_ilist(NULL,mtop->moltype[molb->type].ilist,
1349                      0,molb->natoms_mol,FALSE,FALSE,graph);
1350
1351       for(mol=0; mol<molb->nmol; mol++) {
1352         mk_mshift(fplog,graph,ePBC,box,x+as);
1353
1354         shift_self(graph,box,x+as);
1355         /* The molecule is whole now.
1356          * We don't need the second mk_mshift call as in do_pbc_first,
1357          * since we no longer need this graph.
1358          */
1359
1360         as += molb->natoms_mol;
1361       }
1362       done_graph(graph);
1363     }
1364   }
1365   sfree(graph);
1366 }
1367
1368 void do_pbc_first_mtop(FILE *fplog,int ePBC,matrix box,
1369                        gmx_mtop_t *mtop,rvec x[])
1370 {
1371   low_do_pbc_mtop(fplog,ePBC,box,mtop,x,TRUE);
1372 }
1373
1374 void do_pbc_mtop(FILE *fplog,int ePBC,matrix box,
1375                  gmx_mtop_t *mtop,rvec x[])
1376 {
1377   low_do_pbc_mtop(fplog,ePBC,box,mtop,x,FALSE);
1378 }
1379
1380 void finish_run(FILE *fplog,t_commrec *cr,const char *confout,
1381                 t_inputrec *inputrec,
1382                 t_nrnb nrnb[],gmx_wallcycle_t wcycle,
1383                 gmx_runtime_t *runtime,
1384                 gmx_bool bWriteStat)
1385 {
1386   int    i,j;
1387   t_nrnb *nrnb_tot=NULL;
1388   real   delta_t;
1389   double nbfs,mflop;
1390   double cycles[ewcNR];
1391
1392   wallcycle_sum(cr,wcycle,cycles);
1393
1394   if (cr->nnodes > 1) {
1395     if (SIMMASTER(cr))
1396       snew(nrnb_tot,1);
1397 #ifdef GMX_MPI
1398     MPI_Reduce(nrnb->n,nrnb_tot->n,eNRNB,MPI_DOUBLE,MPI_SUM,
1399                MASTERRANK(cr),cr->mpi_comm_mysim);
1400 #endif
1401   } else {
1402     nrnb_tot = nrnb;
1403   }
1404
1405   if (SIMMASTER(cr)) {
1406     print_flop(fplog,nrnb_tot,&nbfs,&mflop);
1407     if (cr->nnodes > 1) {
1408       sfree(nrnb_tot);
1409     }
1410   }
1411
1412   if ((cr->duty & DUTY_PP) && DOMAINDECOMP(cr)) {
1413     print_dd_statistics(cr,inputrec,fplog);
1414   }
1415
1416 #ifdef GMX_MPI
1417     if (PARTDECOMP(cr))
1418     {
1419         if (MASTER(cr))
1420         {
1421             t_nrnb     *nrnb_all;
1422             int        s;
1423             MPI_Status stat;
1424
1425             snew(nrnb_all,cr->nnodes);
1426             nrnb_all[0] = *nrnb;
1427             for(s=1; s<cr->nnodes; s++)
1428             {
1429                 MPI_Recv(nrnb_all[s].n,eNRNB,MPI_DOUBLE,s,0,
1430                          cr->mpi_comm_mysim,&stat);
1431             }
1432             pr_load(fplog,cr,nrnb_all);
1433             sfree(nrnb_all);
1434         }
1435         else
1436         {
1437             MPI_Send(nrnb->n,eNRNB,MPI_DOUBLE,MASTERRANK(cr),0,
1438                      cr->mpi_comm_mysim);
1439         }
1440     }
1441 #endif
1442
1443   if (SIMMASTER(cr)) {
1444     wallcycle_print(fplog,cr->nnodes,cr->npmenodes,runtime->realtime,
1445                     wcycle,cycles);
1446
1447     if (EI_DYNAMICS(inputrec->eI)) {
1448       delta_t = inputrec->delta_t;
1449     } else {
1450       delta_t = 0;
1451     }
1452
1453     if (fplog) {
1454         print_perf(fplog,runtime->proctime,runtime->realtime,
1455                    cr->nnodes-cr->npmenodes,
1456                    runtime->nsteps_done,delta_t,nbfs,mflop);
1457     }
1458     if (bWriteStat) {
1459         print_perf(stderr,runtime->proctime,runtime->realtime,
1460                    cr->nnodes-cr->npmenodes,
1461                    runtime->nsteps_done,delta_t,nbfs,mflop);
1462     }
1463
1464     /*
1465     runtime=inputrec->nsteps*inputrec->delta_t;
1466     if (bWriteStat) {
1467       if (cr->nnodes == 1)
1468         fprintf(stderr,"\n\n");
1469       print_perf(stderr,nodetime,realtime,runtime,&ntot,
1470                  cr->nnodes-cr->npmenodes,FALSE);
1471     }
1472     wallcycle_print(fplog,cr->nnodes,cr->npmenodes,realtime,wcycle,cycles);
1473     print_perf(fplog,nodetime,realtime,runtime,&ntot,cr->nnodes-cr->npmenodes,
1474                TRUE);
1475     if (PARTDECOMP(cr))
1476       pr_load(fplog,cr,nrnb_all);
1477     if (cr->nnodes > 1)
1478       sfree(nrnb_all);
1479     */
1480   }
1481 }
1482
1483 void init_md(FILE *fplog,
1484              t_commrec *cr,t_inputrec *ir,const output_env_t oenv,
1485              double *t,double *t0,
1486              real *lambda,double *lam0,
1487              t_nrnb *nrnb,gmx_mtop_t *mtop,
1488              gmx_update_t *upd,
1489              int nfile,const t_filenm fnm[],
1490              gmx_mdoutf_t **outf,t_mdebin **mdebin,
1491              tensor force_vir,tensor shake_vir,rvec mu_tot,
1492              gmx_bool *bSimAnn,t_vcm **vcm, t_state *state, unsigned long Flags)
1493 {
1494     int  i,j,n;
1495     real tmpt,mod;
1496
1497     /* Initial values */
1498     *t = *t0       = ir->init_t;
1499     if (ir->efep != efepNO)
1500     {
1501         *lam0 = ir->init_lambda;
1502         *lambda = *lam0 + ir->init_step*ir->delta_lambda;
1503     }
1504     else
1505     {
1506         *lambda = *lam0   = 0.0;
1507     }
1508
1509     *bSimAnn=FALSE;
1510     for(i=0;i<ir->opts.ngtc;i++)
1511     {
1512         /* set bSimAnn if any group is being annealed */
1513         if(ir->opts.annealing[i]!=eannNO)
1514         {
1515             *bSimAnn = TRUE;
1516         }
1517     }
1518     if (*bSimAnn)
1519     {
1520         update_annealing_target_temp(&(ir->opts),ir->init_t);
1521     }
1522
1523     if (upd)
1524     {
1525         *upd = init_update(fplog,ir);
1526     }
1527
1528     if (vcm != NULL)
1529     {
1530         *vcm = init_vcm(fplog,&mtop->groups,ir);
1531     }
1532
1533     if (EI_DYNAMICS(ir->eI) && !(Flags & MD_APPENDFILES))
1534     {
1535         if (ir->etc == etcBERENDSEN)
1536         {
1537             please_cite(fplog,"Berendsen84a");
1538         }
1539         if (ir->etc == etcVRESCALE)
1540         {
1541             please_cite(fplog,"Bussi2007a");
1542         }
1543     }
1544
1545     init_nrnb(nrnb);
1546
1547     if (nfile != -1)
1548     {
1549         *outf = init_mdoutf(nfile,fnm,Flags,cr,ir,oenv);
1550
1551         *mdebin = init_mdebin((Flags & MD_APPENDFILES) ? NULL : (*outf)->fp_ene,
1552                               mtop,ir, (*outf)->fp_dhdl);
1553     }
1554
1555     if (ir->bAdress)
1556     {
1557       please_cite(fplog,"Fritsch12");
1558       please_cite(fplog,"Junghans10");
1559     }
1560     /* Initiate variables */
1561     clear_mat(force_vir);
1562     clear_mat(shake_vir);
1563     clear_rvec(mu_tot);
1564
1565     debug_gmx();
1566 }
1567