src/gromacs/gmxlib/nonbonded/nb_kernel_adress_c/nb_kernel133_c_adress.c

   1 /*
   2  * Copyright (c) Erik Lindahl, David van der Spoel 2003
   3  *
   4  * This file is generated automatically at compile time
   5  * by the program mknb in the Gromacs distribution.
   6  *
   7  * Options used when generation this file:
   8  * Language:         c
   9  * Precision:        single
  10  * Threads:          yes
  11  * Software invsqrt: no
  12  * PowerPC invsqrt:  no
  13  * Prefetch forces:  no
  14  * Adress kernel:  yes
  15  * Comments:         no
  16  */
  17 #ifdef HAVE_CONFIG_H
  18 #include<config.h>
  19 #endif
  20 #ifdef GMX_THREAD_SHM_FDECOMP
  21 #include<thread_mpi.h>
  22 #endif
  23 #define ALMOST_ZERO 1e-30
  24 #define ALMOST_ONE 1-(1e-30)
  25 #include<math.h>
  26
  27 #include "nb_kernel133_adress.h"
  28
  29
  30
  31 /*
  32  * Gromacs nonbonded kernel nb_kernel133_adress_cg
  33  * Coulomb interaction:     Normal Coulomb
  34  * VdW interaction:         Tabulated
  35  * water optimization:      TIP4P - other atoms
  36  * Calculate forces:        yes
  37  */
  38 void nb_kernel133_adress_cg(
  39                     int *           p_nri,
  40                     int *           iinr,
  41                     int *           jindex,
  42                     int *           jjnr,
  43                     int *           shift,
  44                     real *         shiftvec,
  45                     real *         fshift,
  46                     int *           gid,
  47                     real *         pos,
  48                     real *         faction,
  49                     real *         charge,
  50                     real *         p_facel,
  51                     real *         p_krf,
  52                     real *         p_crf,
  53                     real *         Vc,
  54                     int *           type,
  55                     int *           p_ntype,
  56                     real *         vdwparam,
  57                     real *         Vvdw,
  58                     real *         p_tabscale,
  59                     real *         VFtab,
  60                     real *         invsqrta,
  61                     real *         dvda,
  62                     real *         p_gbtabscale,
  63                     real *         GBtab,
  64                     int *           p_nthreads,
  65                     int *           count,
  66                     void *          mtx,
  67                     int *           outeriter,
  68                     int *           inneriter,
  69                     real           force_cap,
  70                     real *         wf)
  71 {
  72     int           nri,ntype,nthreads;
  73     real         facel,krf,crf,tabscale,gbtabscale;
  74     int           n,ii,is3,ii3,k,nj0,nj1,jnr,j3,ggid;
  75     int           nn0,nn1,nouter,ninner;
  76     real         shX,shY,shZ;
  77     real         fscal,tx,ty,tz;
  78     real         rinvsq;
  79     real         jq;
  80     real         qq,vcoul,vctot;
  81     int           nti;
  82     int           tj;
  83     real         Vvdw6,Vvdwtot;
  84     real         Vvdw12;
  85     real         r,rt,eps,eps2;
  86     int           n0,nnn;
  87     real         Y,F,Geps,Heps2,Fp,VV;
  88     real         FF;
  89     real         fijD,fijR;
  90     real         ix1,iy1,iz1,fix1,fiy1,fiz1;
  91     real         ix2,iy2,iz2,fix2,fiy2,fiz2;
  92     real         ix3,iy3,iz3,fix3,fiy3,fiz3;
  93     real         ix4,iy4,iz4,fix4,fiy4,fiz4;
  94     real         jx1,jy1,jz1,fjx1,fjy1,fjz1;
  95     real         dx11,dy11,dz11,rsq11,rinv11;
  96     real         dx21,dy21,dz21,rsq21,rinv21;
  97     real         dx31,dy31,dz31,rsq31,rinv31;
  98     real         dx41,dy41,dz41,rsq41,rinv41;
  99     real         qH,qM;
 100     real         c6,c12;
 101     real         weight_cg1, weight_cg2, weight_product;
 102     real         hybscal;
 103
 104     nri              = *p_nri;
 105     ntype            = *p_ntype;
 106     nthreads         = *p_nthreads;
 107     facel            = *p_facel;
 108     krf              = *p_krf;
 109     crf              = *p_crf;
 110     tabscale         = *p_tabscale;
 111     ii               = iinr[0];
 112     qH               = facel*charge[ii+1];
 113     qM               = facel*charge[ii+3];
 114     nti              = 2*ntype*type[ii];
 115
 116     nouter           = 0;
 117     ninner           = 0;
 118
 119     do
 120     {
 121         #ifdef GMX_THREAD_SHM_FDECOMP
 122         tMPI_Thread_mutex_lock((tMPI_Thread_mutex_t *)mtx);
 123         nn0              = *count;
 124         nn1              = nn0+(nri-nn0)/(2*nthreads)+10;
 125         *count           = nn1;
 126         tMPI_Thread_mutex_unlock((tMPI_Thread_mutex_t *)mtx);
 127         if(nn1>nri) nn1=nri;
 128         #else
 129         nn0 = 0;
 130         nn1 = nri;
 131         #endif
 132
 133         for(n=nn0; (n<nn1); n++)
 134         {
 135             is3              = 3*shift[n];
 136             shX              = shiftvec[is3];
 137             shY              = shiftvec[is3+1];
 138             shZ              = shiftvec[is3+2];
 139             nj0              = jindex[n];
 140             nj1              = jindex[n+1];
 141             ii               = iinr[n];
 142             ii3              = 3*ii;
 143             ix1              = shX + pos[ii3+0];
 144             iy1              = shY + pos[ii3+1];
 145             iz1              = shZ + pos[ii3+2];
 146             ix2              = shX + pos[ii3+3];
 147             iy2              = shY + pos[ii3+4];
 148             iz2              = shZ + pos[ii3+5];
 149             ix3              = shX + pos[ii3+6];
 150             iy3              = shY + pos[ii3+7];
 151             iz3              = shZ + pos[ii3+8];
 152             ix4              = shX + pos[ii3+9];
 153             iy4              = shY + pos[ii3+10];
 154             iz4              = shZ + pos[ii3+11];
 155             weight_cg1       = wf[ii];
 156             vctot            = 0;
 157             Vvdwtot          = 0;
 158             fix1             = 0;
 159             fiy1             = 0;
 160             fiz1             = 0;
 161             fix2             = 0;
 162             fiy2             = 0;
 163             fiz2             = 0;
 164             fix3             = 0;
 165             fiy3             = 0;
 166             fiz3             = 0;
 167             fix4             = 0;
 168             fiy4             = 0;
 169             fiz4             = 0;
 170
 171             for(k=nj0; (k<nj1); k++)
 172             {
 173                 jnr              = jjnr[k];
 174                 weight_cg2       = wf[jnr];
 175                 weight_product   = weight_cg1*weight_cg2;
 176                 if (weight_product < ALMOST_ZERO) {
 177                        hybscal = 1.0;
 178                 }
 179                 else if (weight_product >= ALMOST_ONE)
 180                 {
 181                   /* force is zero, skip this molecule */
 182                        continue;
 183                 }
 184                 else
 185                 {
 186                    hybscal = 1.0 - weight_product;
 187                 }
 188                 j3               = 3*jnr;
 189                 jx1              = pos[j3+0];
 190                 jy1              = pos[j3+1];
 191                 jz1              = pos[j3+2];
 192                 dx11             = ix1 - jx1;
 193                 dy11             = iy1 - jy1;
 194                 dz11             = iz1 - jz1;
 195                 rsq11            = dx11*dx11+dy11*dy11+dz11*dz11;
 196                 dx21             = ix2 - jx1;
 197                 dy21             = iy2 - jy1;
 198                 dz21             = iz2 - jz1;
 199                 rsq21            = dx21*dx21+dy21*dy21+dz21*dz21;
 200                 dx31             = ix3 - jx1;
 201                 dy31             = iy3 - jy1;
 202                 dz31             = iz3 - jz1;
 203                 rsq31            = dx31*dx31+dy31*dy31+dz31*dz31;
 204                 dx41             = ix4 - jx1;
 205                 dy41             = iy4 - jy1;
 206                 dz41             = iz4 - jz1;
 207                 rsq41            = dx41*dx41+dy41*dy41+dz41*dz41;
 208                 rinv11           = 1.0/sqrt(rsq11);
 209                 rinv21           = 1.0/sqrt(rsq21);
 210                 rinv31           = 1.0/sqrt(rsq31);
 211                 rinv41           = 1.0/sqrt(rsq41);
 212                 tj               = nti+2*type[jnr];
 213                 c6               = vdwparam[tj];
 214                 c12              = vdwparam[tj+1];
 215                 r                = rsq11*rinv11;
 216                 rt               = r*tabscale;
 217                 n0               = rt;
 218                 eps              = rt-n0;
 219                 eps2             = eps*eps;
 220                 nnn              = 8*n0;
 221                 Y                = VFtab[nnn];
 222                 F                = VFtab[nnn+1];
 223                 Geps             = eps*VFtab[nnn+2];
 224                 Heps2            = eps2*VFtab[nnn+3];
 225                 Fp               = F+Geps+Heps2;
 226                 VV               = Y+eps*Fp;
 227                 FF               = Fp+Geps+2.0*Heps2;
 228                 Vvdw6            = c6*VV;
 229                 fijD             = c6*FF;
 230                 nnn              = nnn+4;
 231                 Y                = VFtab[nnn];
 232                 F                = VFtab[nnn+1];
 233                 Geps             = eps*VFtab[nnn+2];
 234                 Heps2            = eps2*VFtab[nnn+3];
 235                 Fp               = F+Geps+Heps2;
 236                 VV               = Y+eps*Fp;
 237                 FF               = Fp+Geps+2.0*Heps2;
 238                 Vvdw12           = c12*VV;
 239                 fijR             = c12*FF;
 240                 Vvdwtot          = Vvdwtot+ Vvdw6 + Vvdw12;
 241                 fscal            = -((fijD+fijR)*tabscale)*rinv11;
 242                 fscal *= hybscal;
 243                 tx               = fscal*dx11;
 244                 ty               = fscal*dy11;
 245                 tz               = fscal*dz11;
 246                 fix1             = fix1 + tx;
 247                 fiy1             = fiy1 + ty;
 248                 fiz1             = fiz1 + tz;
 249                 fjx1             = faction[j3+0] - tx;
 250                 fjy1             = faction[j3+1] - ty;
 251                 fjz1             = faction[j3+2] - tz;
 252                 jq               = charge[jnr+0];
 253                 qq               = qH*jq;
 254                 rinvsq           = rinv21*rinv21;
 255                 vcoul            = qq*rinv21;
 256                 vctot            = vctot+vcoul;
 257                 fscal            = (vcoul)*rinvsq;
 258                 fscal *= hybscal;
 259                 tx               = fscal*dx21;
 260                 ty               = fscal*dy21;
 261                 tz               = fscal*dz21;
 262                 fix2             = fix2 + tx;
 263                 fiy2             = fiy2 + ty;
 264                 fiz2             = fiz2 + tz;
 265                 fjx1             = fjx1 - tx;
 266                 fjy1             = fjy1 - ty;
 267                 fjz1             = fjz1 - tz;
 268                 rinvsq           = rinv31*rinv31;
 269                 vcoul            = qq*rinv31;
 270                 vctot            = vctot+vcoul;
 271                 fscal            = (vcoul)*rinvsq;
 272                 fscal *= hybscal;
 273                 tx               = fscal*dx31;
 274                 ty               = fscal*dy31;
 275                 tz               = fscal*dz31;
 276                 fix3             = fix3 + tx;
 277                 fiy3             = fiy3 + ty;
 278                 fiz3             = fiz3 + tz;
 279                 fjx1             = fjx1 - tx;
 280                 fjy1             = fjy1 - ty;
 281                 fjz1             = fjz1 - tz;
 282                 qq               = qM*jq;
 283                 rinvsq           = rinv41*rinv41;
 284                 vcoul            = qq*rinv41;
 285                 vctot            = vctot+vcoul;
 286                 fscal            = (vcoul)*rinvsq;
 287                 fscal *= hybscal;
 288                 tx               = fscal*dx41;
 289                 ty               = fscal*dy41;
 290                 tz               = fscal*dz41;
 291                 fix4             = fix4 + tx;
 292                 fiy4             = fiy4 + ty;
 293                 fiz4             = fiz4 + tz;
 294                 faction[j3+0]    = fjx1 - tx;
 295                 faction[j3+1]    = fjy1 - ty;
 296                 faction[j3+2]    = fjz1 - tz;
 297             }
 298
 299             faction[ii3+0]   = faction[ii3+0] + fix1;
 300             faction[ii3+1]   = faction[ii3+1] + fiy1;
 301             faction[ii3+2]   = faction[ii3+2] + fiz1;
 302             faction[ii3+3]   = faction[ii3+3] + fix2;
 303             faction[ii3+4]   = faction[ii3+4] + fiy2;
 304             faction[ii3+5]   = faction[ii3+5] + fiz2;
 305             faction[ii3+6]   = faction[ii3+6] + fix3;
 306             faction[ii3+7]   = faction[ii3+7] + fiy3;
 307             faction[ii3+8]   = faction[ii3+8] + fiz3;
 308             faction[ii3+9]   = faction[ii3+9] + fix4;
 309             faction[ii3+10]  = faction[ii3+10] + fiy4;
 310             faction[ii3+11]  = faction[ii3+11] + fiz4;
 311             fshift[is3]      = fshift[is3]+fix1+fix2+fix3+fix4;
 312             fshift[is3+1]    = fshift[is3+1]+fiy1+fiy2+fiy3+fiy4;
 313             fshift[is3+2]    = fshift[is3+2]+fiz1+fiz2+fiz3+fiz4;
 314             ggid             = gid[n];
 315             Vc[ggid]         = Vc[ggid] + vctot;
 316             Vvdw[ggid]       = Vvdw[ggid] + Vvdwtot;
 317             ninner           = ninner + nj1 - nj0;
 318         }
 319
 320         nouter           = nouter + nn1 - nn0;
 321     }
 322     while (nn1<nri);
 323
 324     *outeriter       = nouter;
 325     *inneriter       = ninner;
 326 }
 327
 328
 329
 330
 331
 332 /*
 333  * Gromacs nonbonded kernel nb_kernel133_adress_ex
 334  * Coulomb interaction:     Normal Coulomb
 335  * VdW interaction:         Tabulated
 336  * water optimization:      TIP4P - other atoms
 337  * Calculate forces:        yes
 338  */
 339 void nb_kernel133_adress_ex(
 340                     int *           p_nri,
 341                     int *           iinr,
 342                     int *           jindex,
 343                     int *           jjnr,
 344                     int *           shift,
 345                     real *         shiftvec,
 346                     real *         fshift,
 347                     int *           gid,
 348                     real *         pos,
 349                     real *         faction,
 350                     real *         charge,
 351                     real *         p_facel,
 352                     real *         p_krf,
 353                     real *         p_crf,
 354                     real *         Vc,
 355                     int *           type,
 356                     int *           p_ntype,
 357                     real *         vdwparam,
 358                     real *         Vvdw,
 359                     real *         p_tabscale,
 360                     real *         VFtab,
 361                     real *         invsqrta,
 362                     real *         dvda,
 363                     real *         p_gbtabscale,
 364                     real *         GBtab,
 365                     int *           p_nthreads,
 366                     int *           count,
 367                     void *          mtx,
 368                     int *           outeriter,
 369                     int *           inneriter,
 370                     real           force_cap,
 371                     real *         wf)
 372 {
 373     int           nri,ntype,nthreads;
 374     real         facel,krf,crf,tabscale,gbtabscale;
 375     int           n,ii,is3,ii3,k,nj0,nj1,jnr,j3,ggid;
 376     int           nn0,nn1,nouter,ninner;
 377     real         shX,shY,shZ;
 378     real         fscal,tx,ty,tz;
 379     real         rinvsq;
 380     real         jq;
 381     real         qq,vcoul,vctot;
 382     int           nti;
 383     int           tj;
 384     real         Vvdw6,Vvdwtot;
 385     real         Vvdw12;
 386     real         r,rt,eps,eps2;
 387     int           n0,nnn;
 388     real         Y,F,Geps,Heps2,Fp,VV;
 389     real         FF;
 390     real         fijD,fijR;
 391     real         ix1,iy1,iz1,fix1,fiy1,fiz1;
 392     real         ix2,iy2,iz2,fix2,fiy2,fiz2;
 393     real         ix3,iy3,iz3,fix3,fiy3,fiz3;
 394     real         ix4,iy4,iz4,fix4,fiy4,fiz4;
 395     real         jx1,jy1,jz1,fjx1,fjy1,fjz1;
 396     real         dx11,dy11,dz11,rsq11,rinv11;
 397     real         dx21,dy21,dz21,rsq21,rinv21;
 398     real         dx31,dy31,dz31,rsq31,rinv31;
 399     real         dx41,dy41,dz41,rsq41,rinv41;
 400     real         qH,qM;
 401     real         c6,c12;
 402     real         weight_cg1, weight_cg2, weight_product;
 403     real         hybscal;
 404
 405     nri              = *p_nri;
 406     ntype            = *p_ntype;
 407     nthreads         = *p_nthreads;
 408     facel            = *p_facel;
 409     krf              = *p_krf;
 410     crf              = *p_crf;
 411     tabscale         = *p_tabscale;
 412     ii               = iinr[0];
 413     qH               = facel*charge[ii+1];
 414     qM               = facel*charge[ii+3];
 415     nti              = 2*ntype*type[ii];
 416
 417     nouter           = 0;
 418     ninner           = 0;
 419
 420     do
 421     {
 422         #ifdef GMX_THREAD_SHM_FDECOMP
 423         tMPI_Thread_mutex_lock((tMPI_Thread_mutex_t *)mtx);
 424         nn0              = *count;
 425         nn1              = nn0+(nri-nn0)/(2*nthreads)+10;
 426         *count           = nn1;
 427         tMPI_Thread_mutex_unlock((tMPI_Thread_mutex_t *)mtx);
 428         if(nn1>nri) nn1=nri;
 429         #else
 430         nn0 = 0;
 431         nn1 = nri;
 432         #endif
 433
 434         for(n=nn0; (n<nn1); n++)
 435         {
 436             is3              = 3*shift[n];
 437             shX              = shiftvec[is3];
 438             shY              = shiftvec[is3+1];
 439             shZ              = shiftvec[is3+2];
 440             nj0              = jindex[n];
 441             nj1              = jindex[n+1];
 442             ii               = iinr[n];
 443             ii3              = 3*ii;
 444             ix1              = shX + pos[ii3+0];
 445             iy1              = shY + pos[ii3+1];
 446             iz1              = shZ + pos[ii3+2];
 447             ix2              = shX + pos[ii3+3];
 448             iy2              = shY + pos[ii3+4];
 449             iz2              = shZ + pos[ii3+5];
 450             ix3              = shX + pos[ii3+6];
 451             iy3              = shY + pos[ii3+7];
 452             iz3              = shZ + pos[ii3+8];
 453             ix4              = shX + pos[ii3+9];
 454             iy4              = shY + pos[ii3+10];
 455             iz4              = shZ + pos[ii3+11];
 456             weight_cg1       = wf[ii];
 457             vctot            = 0;
 458             Vvdwtot          = 0;
 459             fix1             = 0;
 460             fiy1             = 0;
 461             fiz1             = 0;
 462             fix2             = 0;
 463             fiy2             = 0;
 464             fiz2             = 0;
 465             fix3             = 0;
 466             fiy3             = 0;
 467             fiz3             = 0;
 468             fix4             = 0;
 469             fiy4             = 0;
 470             fiz4             = 0;
 471
 472             for(k=nj0; (k<nj1); k++)
 473             {
 474                 jnr              = jjnr[k];
 475                 weight_cg2       = wf[jnr];
 476                 weight_product   = weight_cg1*weight_cg2;
 477                 if (weight_product < ALMOST_ZERO) {
 478                 /* force is zero, skip this molecule */
 479                  continue;
 480                 }
 481                 else if (weight_product >= ALMOST_ONE)
 482                 {
 483                        hybscal = 1.0;
 484                 }
 485                 else
 486                 {
 487                    hybscal = weight_product;
 488                 }
 489                 j3               = 3*jnr;
 490                 jx1              = pos[j3+0];
 491                 jy1              = pos[j3+1];
 492                 jz1              = pos[j3+2];
 493                 dx11             = ix1 - jx1;
 494                 dy11             = iy1 - jy1;
 495                 dz11             = iz1 - jz1;
 496                 rsq11            = dx11*dx11+dy11*dy11+dz11*dz11;
 497                 dx21             = ix2 - jx1;
 498                 dy21             = iy2 - jy1;
 499                 dz21             = iz2 - jz1;
 500                 rsq21            = dx21*dx21+dy21*dy21+dz21*dz21;
 501                 dx31             = ix3 - jx1;
 502                 dy31             = iy3 - jy1;
 503                 dz31             = iz3 - jz1;
 504                 rsq31            = dx31*dx31+dy31*dy31+dz31*dz31;
 505                 dx41             = ix4 - jx1;
 506                 dy41             = iy4 - jy1;
 507                 dz41             = iz4 - jz1;
 508                 rsq41            = dx41*dx41+dy41*dy41+dz41*dz41;
 509                 rinv11           = 1.0/sqrt(rsq11);
 510                 rinv21           = 1.0/sqrt(rsq21);
 511                 rinv31           = 1.0/sqrt(rsq31);
 512                 rinv41           = 1.0/sqrt(rsq41);
 513                 tj               = nti+2*type[jnr];
 514                 c6               = vdwparam[tj];
 515                 c12              = vdwparam[tj+1];
 516                 r                = rsq11*rinv11;
 517                 rt               = r*tabscale;
 518                 n0               = rt;
 519                 eps              = rt-n0;
 520                 eps2             = eps*eps;
 521                 nnn              = 8*n0;
 522                 Y                = VFtab[nnn];
 523                 F                = VFtab[nnn+1];
 524                 Geps             = eps*VFtab[nnn+2];
 525                 Heps2            = eps2*VFtab[nnn+3];
 526                 Fp               = F+Geps+Heps2;
 527                 VV               = Y+eps*Fp;
 528                 FF               = Fp+Geps+2.0*Heps2;
 529                 Vvdw6            = c6*VV;
 530                 fijD             = c6*FF;
 531                 nnn              = nnn+4;
 532                 Y                = VFtab[nnn];
 533                 F                = VFtab[nnn+1];
 534                 Geps             = eps*VFtab[nnn+2];
 535                 Heps2            = eps2*VFtab[nnn+3];
 536                 Fp               = F+Geps+Heps2;
 537                 VV               = Y+eps*Fp;
 538                 FF               = Fp+Geps+2.0*Heps2;
 539                 Vvdw12           = c12*VV;
 540                 fijR             = c12*FF;
 541                 Vvdwtot          = Vvdwtot+ Vvdw6 + Vvdw12;
 542                 fscal            = -((fijD+fijR)*tabscale)*rinv11;
 543                 fscal *= hybscal;
 544                 if(force_cap>0 && (fabs(fscal)> force_cap)){
 545                 fscal=force_cap*fscal/fabs(fscal);
 546                 }
 547                 tx               = fscal*dx11;
 548                 ty               = fscal*dy11;
 549                 tz               = fscal*dz11;
 550                 fix1             = fix1 + tx;
 551                 fiy1             = fiy1 + ty;
 552                 fiz1             = fiz1 + tz;
 553                 fjx1             = faction[j3+0] - tx;
 554                 fjy1             = faction[j3+1] - ty;
 555                 fjz1             = faction[j3+2] - tz;
 556                 jq               = charge[jnr+0];
 557                 qq               = qH*jq;
 558                 rinvsq           = rinv21*rinv21;
 559                 vcoul            = qq*rinv21;
 560                 vctot            = vctot+vcoul;
 561                 fscal            = (vcoul)*rinvsq;
 562                 fscal *= hybscal;
 563                 if(force_cap>0 && (fabs(fscal)> force_cap)){
 564                 fscal=force_cap*fscal/fabs(fscal);
 565                 }
 566                 tx               = fscal*dx21;
 567                 ty               = fscal*dy21;
 568                 tz               = fscal*dz21;
 569                 fix2             = fix2 + tx;
 570                 fiy2             = fiy2 + ty;
 571                 fiz2             = fiz2 + tz;
 572                 fjx1             = fjx1 - tx;
 573                 fjy1             = fjy1 - ty;
 574                 fjz1             = fjz1 - tz;
 575                 rinvsq           = rinv31*rinv31;
 576                 vcoul            = qq*rinv31;
 577                 vctot            = vctot+vcoul;
 578                 fscal            = (vcoul)*rinvsq;
 579                 fscal *= hybscal;
 580                 if(force_cap>0 && (fabs(fscal)> force_cap)){
 581                 fscal=force_cap*fscal/fabs(fscal);
 582                 }
 583                 tx               = fscal*dx31;
 584                 ty               = fscal*dy31;
 585                 tz               = fscal*dz31;
 586                 fix3             = fix3 + tx;
 587                 fiy3             = fiy3 + ty;
 588                 fiz3             = fiz3 + tz;
 589                 fjx1             = fjx1 - tx;
 590                 fjy1             = fjy1 - ty;
 591                 fjz1             = fjz1 - tz;
 592                 qq               = qM*jq;
 593                 rinvsq           = rinv41*rinv41;
 594                 vcoul            = qq*rinv41;
 595                 vctot            = vctot+vcoul;
 596                 fscal            = (vcoul)*rinvsq;
 597                 fscal *= hybscal;
 598                 if(force_cap>0 && (fabs(fscal)> force_cap)){
 599                 fscal=force_cap*fscal/fabs(fscal);
 600                 }
 601                 tx               = fscal*dx41;
 602                 ty               = fscal*dy41;
 603                 tz               = fscal*dz41;
 604                 fix4             = fix4 + tx;
 605                 fiy4             = fiy4 + ty;
 606                 fiz4             = fiz4 + tz;
 607                 faction[j3+0]    = fjx1 - tx;
 608                 faction[j3+1]    = fjy1 - ty;
 609                 faction[j3+2]    = fjz1 - tz;
 610             }
 611
 612             faction[ii3+0]   = faction[ii3+0] + fix1;
 613             faction[ii3+1]   = faction[ii3+1] + fiy1;
 614             faction[ii3+2]   = faction[ii3+2] + fiz1;
 615             faction[ii3+3]   = faction[ii3+3] + fix2;
 616             faction[ii3+4]   = faction[ii3+4] + fiy2;
 617             faction[ii3+5]   = faction[ii3+5] + fiz2;
 618             faction[ii3+6]   = faction[ii3+6] + fix3;
 619             faction[ii3+7]   = faction[ii3+7] + fiy3;
 620             faction[ii3+8]   = faction[ii3+8] + fiz3;
 621             faction[ii3+9]   = faction[ii3+9] + fix4;
 622             faction[ii3+10]  = faction[ii3+10] + fiy4;
 623             faction[ii3+11]  = faction[ii3+11] + fiz4;
 624             fshift[is3]      = fshift[is3]+fix1+fix2+fix3+fix4;
 625             fshift[is3+1]    = fshift[is3+1]+fiy1+fiy2+fiy3+fiy4;
 626             fshift[is3+2]    = fshift[is3+2]+fiz1+fiz2+fiz3+fiz4;
 627             ggid             = gid[n];
 628             Vc[ggid]         = Vc[ggid] + vctot;
 629             Vvdw[ggid]       = Vvdw[ggid] + Vvdwtot;
 630             ninner           = ninner + nj1 - nj0;
 631         }
 632
 633         nouter           = nouter + nn1 - nn0;
 634     }
 635     while (nn1<nri);
 636
 637     *outeriter       = nouter;
 638     *inneriter       = ninner;
 639 }
 640
 641