src/gromacs/gmxlib/nonbonded/nb_kernel_adress_c/nb_kernel313_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_kernel313_adress.h"
  28
  29
  30
  31 /*
  32  * Gromacs nonbonded kernel nb_kernel313_adress_cg
  33  * Coulomb interaction:     Tabulated
  34  * VdW interaction:         Lennard-Jones
  35  * water optimization:      TIP4P - other atoms
  36  * Calculate forces:        yes
  37  */
  38 void nb_kernel313_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         rinvsix;
  84     real         Vvdw6,Vvdwtot;
  85     real         Vvdw12;
  86     real         r,rt,eps,eps2;
  87     int           n0,nnn;
  88     real         Y,F,Geps,Heps2,Fp,VV;
  89     real         FF;
  90     real         fijC;
  91     real         ix1,iy1,iz1,fix1,fiy1,fiz1;
  92     real         ix2,iy2,iz2,fix2,fiy2,fiz2;
  93     real         ix3,iy3,iz3,fix3,fiy3,fiz3;
  94     real         ix4,iy4,iz4,fix4,fiy4,fiz4;
  95     real         jx1,jy1,jz1,fjx1,fjy1,fjz1;
  96     real         dx11,dy11,dz11,rsq11;
  97     real         dx21,dy21,dz21,rsq21,rinv21;
  98     real         dx31,dy31,dz31,rsq31,rinv31;
  99     real         dx41,dy41,dz41,rsq41,rinv41;
 100     real         qH,qM;
 101     real         c6,c12;
 102     real         weight_cg1, weight_cg2, weight_product;
 103     real         hybscal;
 104
 105     nri              = *p_nri;
 106     ntype            = *p_ntype;
 107     nthreads         = *p_nthreads;
 108     facel            = *p_facel;
 109     krf              = *p_krf;
 110     crf              = *p_crf;
 111     tabscale         = *p_tabscale;
 112     ii               = iinr[0];
 113     qH               = facel*charge[ii+1];
 114     qM               = facel*charge[ii+3];
 115     nti              = 2*ntype*type[ii];
 116
 117     nouter           = 0;
 118     ninner           = 0;
 119
 120     do
 121     {
 122         #ifdef GMX_THREAD_SHM_FDECOMP
 123         tMPI_Thread_mutex_lock((tMPI_Thread_mutex_t *)mtx);
 124         nn0              = *count;
 125         nn1              = nn0+(nri-nn0)/(2*nthreads)+10;
 126         *count           = nn1;
 127         tMPI_Thread_mutex_unlock((tMPI_Thread_mutex_t *)mtx);
 128         if(nn1>nri) nn1=nri;
 129         #else
 130         nn0 = 0;
 131         nn1 = nri;
 132         #endif
 133
 134         for(n=nn0; (n<nn1); n++)
 135         {
 136             is3              = 3*shift[n];
 137             shX              = shiftvec[is3];
 138             shY              = shiftvec[is3+1];
 139             shZ              = shiftvec[is3+2];
 140             nj0              = jindex[n];
 141             nj1              = jindex[n+1];
 142             ii               = iinr[n];
 143             ii3              = 3*ii;
 144             ix1              = shX + pos[ii3+0];
 145             iy1              = shY + pos[ii3+1];
 146             iz1              = shZ + pos[ii3+2];
 147             ix2              = shX + pos[ii3+3];
 148             iy2              = shY + pos[ii3+4];
 149             iz2              = shZ + pos[ii3+5];
 150             ix3              = shX + pos[ii3+6];
 151             iy3              = shY + pos[ii3+7];
 152             iz3              = shZ + pos[ii3+8];
 153             ix4              = shX + pos[ii3+9];
 154             iy4              = shY + pos[ii3+10];
 155             iz4              = shZ + pos[ii3+11];
 156             weight_cg1       = wf[ii];
 157             vctot            = 0;
 158             Vvdwtot          = 0;
 159             fix1             = 0;
 160             fiy1             = 0;
 161             fiz1             = 0;
 162             fix2             = 0;
 163             fiy2             = 0;
 164             fiz2             = 0;
 165             fix3             = 0;
 166             fiy3             = 0;
 167             fiz3             = 0;
 168             fix4             = 0;
 169             fiy4             = 0;
 170             fiz4             = 0;
 171
 172             for(k=nj0; (k<nj1); k++)
 173             {
 174                 jnr              = jjnr[k];
 175                 weight_cg2       = wf[jnr];
 176                 weight_product   = weight_cg1*weight_cg2;
 177                 if (weight_product < ALMOST_ZERO) {
 178                        hybscal = 1.0;
 179                 }
 180                 else if (weight_product >= ALMOST_ONE)
 181                 {
 182                   /* force is zero, skip this molecule */
 183                        continue;
 184                 }
 185                 else
 186                 {
 187                    hybscal = 1.0 - weight_product;
 188                 }
 189                 j3               = 3*jnr;
 190                 jx1              = pos[j3+0];
 191                 jy1              = pos[j3+1];
 192                 jz1              = pos[j3+2];
 193                 dx11             = ix1 - jx1;
 194                 dy11             = iy1 - jy1;
 195                 dz11             = iz1 - jz1;
 196                 rsq11            = dx11*dx11+dy11*dy11+dz11*dz11;
 197                 dx21             = ix2 - jx1;
 198                 dy21             = iy2 - jy1;
 199                 dz21             = iz2 - jz1;
 200                 rsq21            = dx21*dx21+dy21*dy21+dz21*dz21;
 201                 dx31             = ix3 - jx1;
 202                 dy31             = iy3 - jy1;
 203                 dz31             = iz3 - jz1;
 204                 rsq31            = dx31*dx31+dy31*dy31+dz31*dz31;
 205                 dx41             = ix4 - jx1;
 206                 dy41             = iy4 - jy1;
 207                 dz41             = iz4 - jz1;
 208                 rsq41            = dx41*dx41+dy41*dy41+dz41*dz41;
 209                 rinvsq           = 1.0/rsq11;
 210                 rinv21           = 1.0/sqrt(rsq21);
 211                 rinv31           = 1.0/sqrt(rsq31);
 212                 rinv41           = 1.0/sqrt(rsq41);
 213                 tj               = nti+2*type[jnr];
 214                 c6               = vdwparam[tj];
 215                 c12              = vdwparam[tj+1];
 216                 rinvsix          = rinvsq*rinvsq*rinvsq;
 217                 Vvdw6            = c6*rinvsix;
 218                 Vvdw12           = c12*rinvsix*rinvsix;
 219                 Vvdwtot          = Vvdwtot+Vvdw12-Vvdw6;
 220                 fscal            = (12.0*Vvdw12-6.0*Vvdw6)*rinvsq;
 221                 fscal *= hybscal;
 222                 tx               = fscal*dx11;
 223                 ty               = fscal*dy11;
 224                 tz               = fscal*dz11;
 225                 fix1             = fix1 + tx;
 226                 fiy1             = fiy1 + ty;
 227                 fiz1             = fiz1 + tz;
 228                 fjx1             = faction[j3+0] - tx;
 229                 fjy1             = faction[j3+1] - ty;
 230                 fjz1             = faction[j3+2] - tz;
 231                 jq               = charge[jnr+0];
 232                 qq               = qH*jq;
 233                 r                = rsq21*rinv21;
 234                 rt               = r*tabscale;
 235                 n0               = rt;
 236                 eps              = rt-n0;
 237                 eps2             = eps*eps;
 238                 nnn              = 4*n0;
 239                 Y                = VFtab[nnn];
 240                 F                = VFtab[nnn+1];
 241                 Geps             = eps*VFtab[nnn+2];
 242                 Heps2            = eps2*VFtab[nnn+3];
 243                 Fp               = F+Geps+Heps2;
 244                 VV               = Y+eps*Fp;
 245                 FF               = Fp+Geps+2.0*Heps2;
 246                 vcoul            = qq*VV;
 247                 fijC             = qq*FF;
 248                 vctot            = vctot + vcoul;
 249                 fscal            = -((fijC)*tabscale)*rinv21;
 250                 fscal *= hybscal;
 251                 tx               = fscal*dx21;
 252                 ty               = fscal*dy21;
 253                 tz               = fscal*dz21;
 254                 fix2             = fix2 + tx;
 255                 fiy2             = fiy2 + ty;
 256                 fiz2             = fiz2 + tz;
 257                 fjx1             = fjx1 - tx;
 258                 fjy1             = fjy1 - ty;
 259                 fjz1             = fjz1 - tz;
 260                 r                = rsq31*rinv31;
 261                 rt               = r*tabscale;
 262                 n0               = rt;
 263                 eps              = rt-n0;
 264                 eps2             = eps*eps;
 265                 nnn              = 4*n0;
 266                 Y                = VFtab[nnn];
 267                 F                = VFtab[nnn+1];
 268                 Geps             = eps*VFtab[nnn+2];
 269                 Heps2            = eps2*VFtab[nnn+3];
 270                 Fp               = F+Geps+Heps2;
 271                 VV               = Y+eps*Fp;
 272                 FF               = Fp+Geps+2.0*Heps2;
 273                 vcoul            = qq*VV;
 274                 fijC             = qq*FF;
 275                 vctot            = vctot + vcoul;
 276                 fscal            = -((fijC)*tabscale)*rinv31;
 277                 fscal *= hybscal;
 278                 tx               = fscal*dx31;
 279                 ty               = fscal*dy31;
 280                 tz               = fscal*dz31;
 281                 fix3             = fix3 + tx;
 282                 fiy3             = fiy3 + ty;
 283                 fiz3             = fiz3 + tz;
 284                 fjx1             = fjx1 - tx;
 285                 fjy1             = fjy1 - ty;
 286                 fjz1             = fjz1 - tz;
 287                 qq               = qM*jq;
 288                 r                = rsq41*rinv41;
 289                 rt               = r*tabscale;
 290                 n0               = rt;
 291                 eps              = rt-n0;
 292                 eps2             = eps*eps;
 293                 nnn              = 4*n0;
 294                 Y                = VFtab[nnn];
 295                 F                = VFtab[nnn+1];
 296                 Geps             = eps*VFtab[nnn+2];
 297                 Heps2            = eps2*VFtab[nnn+3];
 298                 Fp               = F+Geps+Heps2;
 299                 VV               = Y+eps*Fp;
 300                 FF               = Fp+Geps+2.0*Heps2;
 301                 vcoul            = qq*VV;
 302                 fijC             = qq*FF;
 303                 vctot            = vctot + vcoul;
 304                 fscal            = -((fijC)*tabscale)*rinv41;
 305                 fscal *= hybscal;
 306                 tx               = fscal*dx41;
 307                 ty               = fscal*dy41;
 308                 tz               = fscal*dz41;
 309                 fix4             = fix4 + tx;
 310                 fiy4             = fiy4 + ty;
 311                 fiz4             = fiz4 + tz;
 312                 faction[j3+0]    = fjx1 - tx;
 313                 faction[j3+1]    = fjy1 - ty;
 314                 faction[j3+2]    = fjz1 - tz;
 315             }
 316
 317             faction[ii3+0]   = faction[ii3+0] + fix1;
 318             faction[ii3+1]   = faction[ii3+1] + fiy1;
 319             faction[ii3+2]   = faction[ii3+2] + fiz1;
 320             faction[ii3+3]   = faction[ii3+3] + fix2;
 321             faction[ii3+4]   = faction[ii3+4] + fiy2;
 322             faction[ii3+5]   = faction[ii3+5] + fiz2;
 323             faction[ii3+6]   = faction[ii3+6] + fix3;
 324             faction[ii3+7]   = faction[ii3+7] + fiy3;
 325             faction[ii3+8]   = faction[ii3+8] + fiz3;
 326             faction[ii3+9]   = faction[ii3+9] + fix4;
 327             faction[ii3+10]  = faction[ii3+10] + fiy4;
 328             faction[ii3+11]  = faction[ii3+11] + fiz4;
 329             fshift[is3]      = fshift[is3]+fix1+fix2+fix3+fix4;
 330             fshift[is3+1]    = fshift[is3+1]+fiy1+fiy2+fiy3+fiy4;
 331             fshift[is3+2]    = fshift[is3+2]+fiz1+fiz2+fiz3+fiz4;
 332             ggid             = gid[n];
 333             Vc[ggid]         = Vc[ggid] + vctot;
 334             Vvdw[ggid]       = Vvdw[ggid] + Vvdwtot;
 335             ninner           = ninner + nj1 - nj0;
 336         }
 337
 338         nouter           = nouter + nn1 - nn0;
 339     }
 340     while (nn1<nri);
 341
 342     *outeriter       = nouter;
 343     *inneriter       = ninner;
 344 }
 345
 346
 347
 348
 349
 350 /*
 351  * Gromacs nonbonded kernel nb_kernel313_adress_ex
 352  * Coulomb interaction:     Tabulated
 353  * VdW interaction:         Lennard-Jones
 354  * water optimization:      TIP4P - other atoms
 355  * Calculate forces:        yes
 356  */
 357 void nb_kernel313_adress_ex(
 358                     int *           p_nri,
 359                     int *           iinr,
 360                     int *           jindex,
 361                     int *           jjnr,
 362                     int *           shift,
 363                     real *         shiftvec,
 364                     real *         fshift,
 365                     int *           gid,
 366                     real *         pos,
 367                     real *         faction,
 368                     real *         charge,
 369                     real *         p_facel,
 370                     real *         p_krf,
 371                     real *         p_crf,
 372                     real *         Vc,
 373                     int *           type,
 374                     int *           p_ntype,
 375                     real *         vdwparam,
 376                     real *         Vvdw,
 377                     real *         p_tabscale,
 378                     real *         VFtab,
 379                     real *         invsqrta,
 380                     real *         dvda,
 381                     real *         p_gbtabscale,
 382                     real *         GBtab,
 383                     int *           p_nthreads,
 384                     int *           count,
 385                     void *          mtx,
 386                     int *           outeriter,
 387                     int *           inneriter,
 388                     real           force_cap,
 389                     real *         wf)
 390 {
 391     int           nri,ntype,nthreads;
 392     real         facel,krf,crf,tabscale,gbtabscale;
 393     int           n,ii,is3,ii3,k,nj0,nj1,jnr,j3,ggid;
 394     int           nn0,nn1,nouter,ninner;
 395     real         shX,shY,shZ;
 396     real         fscal,tx,ty,tz;
 397     real         rinvsq;
 398     real         jq;
 399     real         qq,vcoul,vctot;
 400     int           nti;
 401     int           tj;
 402     real         rinvsix;
 403     real         Vvdw6,Vvdwtot;
 404     real         Vvdw12;
 405     real         r,rt,eps,eps2;
 406     int           n0,nnn;
 407     real         Y,F,Geps,Heps2,Fp,VV;
 408     real         FF;
 409     real         fijC;
 410     real         ix1,iy1,iz1,fix1,fiy1,fiz1;
 411     real         ix2,iy2,iz2,fix2,fiy2,fiz2;
 412     real         ix3,iy3,iz3,fix3,fiy3,fiz3;
 413     real         ix4,iy4,iz4,fix4,fiy4,fiz4;
 414     real         jx1,jy1,jz1,fjx1,fjy1,fjz1;
 415     real         dx11,dy11,dz11,rsq11;
 416     real         dx21,dy21,dz21,rsq21,rinv21;
 417     real         dx31,dy31,dz31,rsq31,rinv31;
 418     real         dx41,dy41,dz41,rsq41,rinv41;
 419     real         qH,qM;
 420     real         c6,c12;
 421     real         weight_cg1, weight_cg2, weight_product;
 422     real         hybscal;
 423
 424     nri              = *p_nri;
 425     ntype            = *p_ntype;
 426     nthreads         = *p_nthreads;
 427     facel            = *p_facel;
 428     krf              = *p_krf;
 429     crf              = *p_crf;
 430     tabscale         = *p_tabscale;
 431     ii               = iinr[0];
 432     qH               = facel*charge[ii+1];
 433     qM               = facel*charge[ii+3];
 434     nti              = 2*ntype*type[ii];
 435
 436     nouter           = 0;
 437     ninner           = 0;
 438
 439     do
 440     {
 441         #ifdef GMX_THREAD_SHM_FDECOMP
 442         tMPI_Thread_mutex_lock((tMPI_Thread_mutex_t *)mtx);
 443         nn0              = *count;
 444         nn1              = nn0+(nri-nn0)/(2*nthreads)+10;
 445         *count           = nn1;
 446         tMPI_Thread_mutex_unlock((tMPI_Thread_mutex_t *)mtx);
 447         if(nn1>nri) nn1=nri;
 448         #else
 449         nn0 = 0;
 450         nn1 = nri;
 451         #endif
 452
 453         for(n=nn0; (n<nn1); n++)
 454         {
 455             is3              = 3*shift[n];
 456             shX              = shiftvec[is3];
 457             shY              = shiftvec[is3+1];
 458             shZ              = shiftvec[is3+2];
 459             nj0              = jindex[n];
 460             nj1              = jindex[n+1];
 461             ii               = iinr[n];
 462             ii3              = 3*ii;
 463             ix1              = shX + pos[ii3+0];
 464             iy1              = shY + pos[ii3+1];
 465             iz1              = shZ + pos[ii3+2];
 466             ix2              = shX + pos[ii3+3];
 467             iy2              = shY + pos[ii3+4];
 468             iz2              = shZ + pos[ii3+5];
 469             ix3              = shX + pos[ii3+6];
 470             iy3              = shY + pos[ii3+7];
 471             iz3              = shZ + pos[ii3+8];
 472             ix4              = shX + pos[ii3+9];
 473             iy4              = shY + pos[ii3+10];
 474             iz4              = shZ + pos[ii3+11];
 475             weight_cg1       = wf[ii];
 476             vctot            = 0;
 477             Vvdwtot          = 0;
 478             fix1             = 0;
 479             fiy1             = 0;
 480             fiz1             = 0;
 481             fix2             = 0;
 482             fiy2             = 0;
 483             fiz2             = 0;
 484             fix3             = 0;
 485             fiy3             = 0;
 486             fiz3             = 0;
 487             fix4             = 0;
 488             fiy4             = 0;
 489             fiz4             = 0;
 490
 491             for(k=nj0; (k<nj1); k++)
 492             {
 493                 jnr              = jjnr[k];
 494                 weight_cg2       = wf[jnr];
 495                 weight_product   = weight_cg1*weight_cg2;
 496                 if (weight_product < ALMOST_ZERO) {
 497                 /* force is zero, skip this molecule */
 498                  continue;
 499                 }
 500                 else if (weight_product >= ALMOST_ONE)
 501                 {
 502                        hybscal = 1.0;
 503                 }
 504                 else
 505                 {
 506                    hybscal = weight_product;
 507                 }
 508                 j3               = 3*jnr;
 509                 jx1              = pos[j3+0];
 510                 jy1              = pos[j3+1];
 511                 jz1              = pos[j3+2];
 512                 dx11             = ix1 - jx1;
 513                 dy11             = iy1 - jy1;
 514                 dz11             = iz1 - jz1;
 515                 rsq11            = dx11*dx11+dy11*dy11+dz11*dz11;
 516                 dx21             = ix2 - jx1;
 517                 dy21             = iy2 - jy1;
 518                 dz21             = iz2 - jz1;
 519                 rsq21            = dx21*dx21+dy21*dy21+dz21*dz21;
 520                 dx31             = ix3 - jx1;
 521                 dy31             = iy3 - jy1;
 522                 dz31             = iz3 - jz1;
 523                 rsq31            = dx31*dx31+dy31*dy31+dz31*dz31;
 524                 dx41             = ix4 - jx1;
 525                 dy41             = iy4 - jy1;
 526                 dz41             = iz4 - jz1;
 527                 rsq41            = dx41*dx41+dy41*dy41+dz41*dz41;
 528                 rinvsq           = 1.0/rsq11;
 529                 rinv21           = 1.0/sqrt(rsq21);
 530                 rinv31           = 1.0/sqrt(rsq31);
 531                 rinv41           = 1.0/sqrt(rsq41);
 532                 tj               = nti+2*type[jnr];
 533                 c6               = vdwparam[tj];
 534                 c12              = vdwparam[tj+1];
 535                 rinvsix          = rinvsq*rinvsq*rinvsq;
 536                 Vvdw6            = c6*rinvsix;
 537                 Vvdw12           = c12*rinvsix*rinvsix;
 538                 Vvdwtot          = Vvdwtot+Vvdw12-Vvdw6;
 539                 fscal            = (12.0*Vvdw12-6.0*Vvdw6)*rinvsq;
 540                 fscal *= hybscal;
 541                 if(force_cap>0 && (fabs(fscal)> force_cap)){
 542                 fscal=force_cap*fscal/fabs(fscal);
 543                 }
 544                 tx               = fscal*dx11;
 545                 ty               = fscal*dy11;
 546                 tz               = fscal*dz11;
 547                 fix1             = fix1 + tx;
 548                 fiy1             = fiy1 + ty;
 549                 fiz1             = fiz1 + tz;
 550                 fjx1             = faction[j3+0] - tx;
 551                 fjy1             = faction[j3+1] - ty;
 552                 fjz1             = faction[j3+2] - tz;
 553                 jq               = charge[jnr+0];
 554                 qq               = qH*jq;
 555                 r                = rsq21*rinv21;
 556                 rt               = r*tabscale;
 557                 n0               = rt;
 558                 eps              = rt-n0;
 559                 eps2             = eps*eps;
 560                 nnn              = 4*n0;
 561                 Y                = VFtab[nnn];
 562                 F                = VFtab[nnn+1];
 563                 Geps             = eps*VFtab[nnn+2];
 564                 Heps2            = eps2*VFtab[nnn+3];
 565                 Fp               = F+Geps+Heps2;
 566                 VV               = Y+eps*Fp;
 567                 FF               = Fp+Geps+2.0*Heps2;
 568                 vcoul            = qq*VV;
 569                 fijC             = qq*FF;
 570                 vctot            = vctot + vcoul;
 571                 fscal            = -((fijC)*tabscale)*rinv21;
 572                 fscal *= hybscal;
 573                 if(force_cap>0 && (fabs(fscal)> force_cap)){
 574                 fscal=force_cap*fscal/fabs(fscal);
 575                 }
 576                 tx               = fscal*dx21;
 577                 ty               = fscal*dy21;
 578                 tz               = fscal*dz21;
 579                 fix2             = fix2 + tx;
 580                 fiy2             = fiy2 + ty;
 581                 fiz2             = fiz2 + tz;
 582                 fjx1             = fjx1 - tx;
 583                 fjy1             = fjy1 - ty;
 584                 fjz1             = fjz1 - tz;
 585                 r                = rsq31*rinv31;
 586                 rt               = r*tabscale;
 587                 n0               = rt;
 588                 eps              = rt-n0;
 589                 eps2             = eps*eps;
 590                 nnn              = 4*n0;
 591                 Y                = VFtab[nnn];
 592                 F                = VFtab[nnn+1];
 593                 Geps             = eps*VFtab[nnn+2];
 594                 Heps2            = eps2*VFtab[nnn+3];
 595                 Fp               = F+Geps+Heps2;
 596                 VV               = Y+eps*Fp;
 597                 FF               = Fp+Geps+2.0*Heps2;
 598                 vcoul            = qq*VV;
 599                 fijC             = qq*FF;
 600                 vctot            = vctot + vcoul;
 601                 fscal            = -((fijC)*tabscale)*rinv31;
 602                 fscal *= hybscal;
 603                 if(force_cap>0 && (fabs(fscal)> force_cap)){
 604                 fscal=force_cap*fscal/fabs(fscal);
 605                 }
 606                 tx               = fscal*dx31;
 607                 ty               = fscal*dy31;
 608                 tz               = fscal*dz31;
 609                 fix3             = fix3 + tx;
 610                 fiy3             = fiy3 + ty;
 611                 fiz3             = fiz3 + tz;
 612                 fjx1             = fjx1 - tx;
 613                 fjy1             = fjy1 - ty;
 614                 fjz1             = fjz1 - tz;
 615                 qq               = qM*jq;
 616                 r                = rsq41*rinv41;
 617                 rt               = r*tabscale;
 618                 n0               = rt;
 619                 eps              = rt-n0;
 620                 eps2             = eps*eps;
 621                 nnn              = 4*n0;
 622                 Y                = VFtab[nnn];
 623                 F                = VFtab[nnn+1];
 624                 Geps             = eps*VFtab[nnn+2];
 625                 Heps2            = eps2*VFtab[nnn+3];
 626                 Fp               = F+Geps+Heps2;
 627                 VV               = Y+eps*Fp;
 628                 FF               = Fp+Geps+2.0*Heps2;
 629                 vcoul            = qq*VV;
 630                 fijC             = qq*FF;
 631                 vctot            = vctot + vcoul;
 632                 fscal            = -((fijC)*tabscale)*rinv41;
 633                 fscal *= hybscal;
 634                 if(force_cap>0 && (fabs(fscal)> force_cap)){
 635                 fscal=force_cap*fscal/fabs(fscal);
 636                 }
 637                 tx               = fscal*dx41;
 638                 ty               = fscal*dy41;
 639                 tz               = fscal*dz41;
 640                 fix4             = fix4 + tx;
 641                 fiy4             = fiy4 + ty;
 642                 fiz4             = fiz4 + tz;
 643                 faction[j3+0]    = fjx1 - tx;
 644                 faction[j3+1]    = fjy1 - ty;
 645                 faction[j3+2]    = fjz1 - tz;
 646             }
 647
 648             faction[ii3+0]   = faction[ii3+0] + fix1;
 649             faction[ii3+1]   = faction[ii3+1] + fiy1;
 650             faction[ii3+2]   = faction[ii3+2] + fiz1;
 651             faction[ii3+3]   = faction[ii3+3] + fix2;
 652             faction[ii3+4]   = faction[ii3+4] + fiy2;
 653             faction[ii3+5]   = faction[ii3+5] + fiz2;
 654             faction[ii3+6]   = faction[ii3+6] + fix3;
 655             faction[ii3+7]   = faction[ii3+7] + fiy3;
 656             faction[ii3+8]   = faction[ii3+8] + fiz3;
 657             faction[ii3+9]   = faction[ii3+9] + fix4;
 658             faction[ii3+10]  = faction[ii3+10] + fiy4;
 659             faction[ii3+11]  = faction[ii3+11] + fiz4;
 660             fshift[is3]      = fshift[is3]+fix1+fix2+fix3+fix4;
 661             fshift[is3+1]    = fshift[is3+1]+fiy1+fiy2+fiy3+fiy4;
 662             fshift[is3+2]    = fshift[is3+2]+fiz1+fiz2+fiz3+fiz4;
 663             ggid             = gid[n];
 664             Vc[ggid]         = Vc[ggid] + vctot;
 665             Vvdw[ggid]       = Vvdw[ggid] + Vvdwtot;
 666             ninner           = ninner + nj1 - nj0;
 667         }
 668
 669         nouter           = nouter + nn1 - nn0;
 670     }
 671     while (nn1<nri);
 672
 673     *outeriter       = nouter;
 674     *inneriter       = ninner;
 675 }
 676
 677