src/gromacs/gmxlib/nonbonded/nb_kernel_adress_c/nb_kernel211_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_kernel211_adress.h"
  28
  29
  30
  31 /*
  32  * Gromacs nonbonded kernel nb_kernel211_adress_cg
  33  * Coulomb interaction:     Reaction field
  34  * VdW interaction:         Lennard-Jones
  35  * water optimization:      SPC/TIP3P - other atoms
  36  * Calculate forces:        yes
  37  */
  38 void nb_kernel211_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         krsq;
  87     real         ix1,iy1,iz1,fix1,fiy1,fiz1;
  88     real         ix2,iy2,iz2,fix2,fiy2,fiz2;
  89     real         ix3,iy3,iz3,fix3,fiy3,fiz3;
  90     real         jx1,jy1,jz1,fjx1,fjy1,fjz1;
  91     real         dx11,dy11,dz11,rsq11,rinv11;
  92     real         dx21,dy21,dz21,rsq21,rinv21;
  93     real         dx31,dy31,dz31,rsq31,rinv31;
  94     real         qO,qH;
  95     real         c6,c12;
  96     real         weight_cg1, weight_cg2, weight_product;
  97     real         hybscal;
  98
  99     nri              = *p_nri;
 100     ntype            = *p_ntype;
 101     nthreads         = *p_nthreads;
 102     facel            = *p_facel;
 103     krf              = *p_krf;
 104     crf              = *p_crf;
 105     tabscale         = *p_tabscale;
 106     ii               = iinr[0];
 107     qO               = facel*charge[ii];
 108     qH               = facel*charge[ii+1];
 109     nti              = 2*ntype*type[ii];
 110
 111     nouter           = 0;
 112     ninner           = 0;
 113
 114     do
 115     {
 116         #ifdef GMX_THREAD_SHM_FDECOMP
 117         tMPI_Thread_mutex_lock((tMPI_Thread_mutex_t *)mtx);
 118         nn0              = *count;
 119         nn1              = nn0+(nri-nn0)/(2*nthreads)+10;
 120         *count           = nn1;
 121         tMPI_Thread_mutex_unlock((tMPI_Thread_mutex_t *)mtx);
 122         if(nn1>nri) nn1=nri;
 123         #else
 124         nn0 = 0;
 125         nn1 = nri;
 126         #endif
 127
 128         for(n=nn0; (n<nn1); n++)
 129         {
 130             is3              = 3*shift[n];
 131             shX              = shiftvec[is3];
 132             shY              = shiftvec[is3+1];
 133             shZ              = shiftvec[is3+2];
 134             nj0              = jindex[n];
 135             nj1              = jindex[n+1];
 136             ii               = iinr[n];
 137             ii3              = 3*ii;
 138             ix1              = shX + pos[ii3+0];
 139             iy1              = shY + pos[ii3+1];
 140             iz1              = shZ + pos[ii3+2];
 141             ix2              = shX + pos[ii3+3];
 142             iy2              = shY + pos[ii3+4];
 143             iz2              = shZ + pos[ii3+5];
 144             ix3              = shX + pos[ii3+6];
 145             iy3              = shY + pos[ii3+7];
 146             iz3              = shZ + pos[ii3+8];
 147             weight_cg1       = wf[ii];
 148             vctot            = 0;
 149             Vvdwtot          = 0;
 150             fix1             = 0;
 151             fiy1             = 0;
 152             fiz1             = 0;
 153             fix2             = 0;
 154             fiy2             = 0;
 155             fiz2             = 0;
 156             fix3             = 0;
 157             fiy3             = 0;
 158             fiz3             = 0;
 159
 160             for(k=nj0; (k<nj1); k++)
 161             {
 162                 jnr              = jjnr[k];
 163                 weight_cg2       = wf[jnr];
 164                 weight_product   = weight_cg1*weight_cg2;
 165                 if (weight_product < ALMOST_ZERO) {
 166                        hybscal = 1.0;
 167                 }
 168                 else if (weight_product >= ALMOST_ONE)
 169                 {
 170                   /* force is zero, skip this molecule */
 171                        continue;
 172                 }
 173                 else
 174                 {
 175                    hybscal = 1.0 - weight_product;
 176                 }
 177                 j3               = 3*jnr;
 178                 jx1              = pos[j3+0];
 179                 jy1              = pos[j3+1];
 180                 jz1              = pos[j3+2];
 181                 dx11             = ix1 - jx1;
 182                 dy11             = iy1 - jy1;
 183                 dz11             = iz1 - jz1;
 184                 rsq11            = dx11*dx11+dy11*dy11+dz11*dz11;
 185                 dx21             = ix2 - jx1;
 186                 dy21             = iy2 - jy1;
 187                 dz21             = iz2 - jz1;
 188                 rsq21            = dx21*dx21+dy21*dy21+dz21*dz21;
 189                 dx31             = ix3 - jx1;
 190                 dy31             = iy3 - jy1;
 191                 dz31             = iz3 - jz1;
 192                 rsq31            = dx31*dx31+dy31*dy31+dz31*dz31;
 193                 rinv11           = 1.0/sqrt(rsq11);
 194                 rinv21           = 1.0/sqrt(rsq21);
 195                 rinv31           = 1.0/sqrt(rsq31);
 196                 jq               = charge[jnr+0];
 197                 qq               = qO*jq;
 198                 tj               = nti+2*type[jnr];
 199                 c6               = vdwparam[tj];
 200                 c12              = vdwparam[tj+1];
 201                 rinvsq           = rinv11*rinv11;
 202                 krsq             = krf*rsq11;
 203                 vcoul            = qq*(rinv11+krsq-crf);
 204                 vctot            = vctot+vcoul;
 205                 rinvsix          = rinvsq*rinvsq*rinvsq;
 206                 Vvdw6            = c6*rinvsix;
 207                 Vvdw12           = c12*rinvsix*rinvsix;
 208                 Vvdwtot          = Vvdwtot+Vvdw12-Vvdw6;
 209                 fscal            = (qq*(rinv11-2.0*krsq)+12.0*Vvdw12-6.0*Vvdw6)*rinvsq;
 210                 fscal *= hybscal;
 211                 tx               = fscal*dx11;
 212                 ty               = fscal*dy11;
 213                 tz               = fscal*dz11;
 214                 fix1             = fix1 + tx;
 215                 fiy1             = fiy1 + ty;
 216                 fiz1             = fiz1 + tz;
 217                 fjx1             = faction[j3+0] - tx;
 218                 fjy1             = faction[j3+1] - ty;
 219                 fjz1             = faction[j3+2] - tz;
 220                 qq               = qH*jq;
 221                 rinvsq           = rinv21*rinv21;
 222                 krsq             = krf*rsq21;
 223                 vcoul            = qq*(rinv21+krsq-crf);
 224                 vctot            = vctot+vcoul;
 225                 fscal            = (qq*(rinv21-2.0*krsq))*rinvsq;
 226                 fscal *= hybscal;
 227                 tx               = fscal*dx21;
 228                 ty               = fscal*dy21;
 229                 tz               = fscal*dz21;
 230                 fix2             = fix2 + tx;
 231                 fiy2             = fiy2 + ty;
 232                 fiz2             = fiz2 + tz;
 233                 fjx1             = fjx1 - tx;
 234                 fjy1             = fjy1 - ty;
 235                 fjz1             = fjz1 - tz;
 236                 rinvsq           = rinv31*rinv31;
 237                 krsq             = krf*rsq31;
 238                 vcoul            = qq*(rinv31+krsq-crf);
 239                 vctot            = vctot+vcoul;
 240                 fscal            = (qq*(rinv31-2.0*krsq))*rinvsq;
 241                 fscal *= hybscal;
 242                 tx               = fscal*dx31;
 243                 ty               = fscal*dy31;
 244                 tz               = fscal*dz31;
 245                 fix3             = fix3 + tx;
 246                 fiy3             = fiy3 + ty;
 247                 fiz3             = fiz3 + tz;
 248                 faction[j3+0]    = fjx1 - tx;
 249                 faction[j3+1]    = fjy1 - ty;
 250                 faction[j3+2]    = fjz1 - tz;
 251             }
 252
 253             faction[ii3+0]   = faction[ii3+0] + fix1;
 254             faction[ii3+1]   = faction[ii3+1] + fiy1;
 255             faction[ii3+2]   = faction[ii3+2] + fiz1;
 256             faction[ii3+3]   = faction[ii3+3] + fix2;
 257             faction[ii3+4]   = faction[ii3+4] + fiy2;
 258             faction[ii3+5]   = faction[ii3+5] + fiz2;
 259             faction[ii3+6]   = faction[ii3+6] + fix3;
 260             faction[ii3+7]   = faction[ii3+7] + fiy3;
 261             faction[ii3+8]   = faction[ii3+8] + fiz3;
 262             fshift[is3]      = fshift[is3]+fix1+fix2+fix3;
 263             fshift[is3+1]    = fshift[is3+1]+fiy1+fiy2+fiy3;
 264             fshift[is3+2]    = fshift[is3+2]+fiz1+fiz2+fiz3;
 265             ggid             = gid[n];
 266             Vc[ggid]         = Vc[ggid] + vctot;
 267             Vvdw[ggid]       = Vvdw[ggid] + Vvdwtot;
 268             ninner           = ninner + nj1 - nj0;
 269         }
 270
 271         nouter           = nouter + nn1 - nn0;
 272     }
 273     while (nn1<nri);
 274
 275     *outeriter       = nouter;
 276     *inneriter       = ninner;
 277 }
 278
 279
 280
 281
 282
 283 /*
 284  * Gromacs nonbonded kernel nb_kernel211_adress_ex
 285  * Coulomb interaction:     Reaction field
 286  * VdW interaction:         Lennard-Jones
 287  * water optimization:      SPC/TIP3P - other atoms
 288  * Calculate forces:        yes
 289  */
 290 void nb_kernel211_adress_ex(
 291                     int *           p_nri,
 292                     int *           iinr,
 293                     int *           jindex,
 294                     int *           jjnr,
 295                     int *           shift,
 296                     real *         shiftvec,
 297                     real *         fshift,
 298                     int *           gid,
 299                     real *         pos,
 300                     real *         faction,
 301                     real *         charge,
 302                     real *         p_facel,
 303                     real *         p_krf,
 304                     real *         p_crf,
 305                     real *         Vc,
 306                     int *           type,
 307                     int *           p_ntype,
 308                     real *         vdwparam,
 309                     real *         Vvdw,
 310                     real *         p_tabscale,
 311                     real *         VFtab,
 312                     real *         invsqrta,
 313                     real *         dvda,
 314                     real *         p_gbtabscale,
 315                     real *         GBtab,
 316                     int *           p_nthreads,
 317                     int *           count,
 318                     void *          mtx,
 319                     int *           outeriter,
 320                     int *           inneriter,
 321                     real           force_cap,
 322                     real *         wf)
 323 {
 324     int           nri,ntype,nthreads;
 325     real         facel,krf,crf,tabscale,gbtabscale;
 326     int           n,ii,is3,ii3,k,nj0,nj1,jnr,j3,ggid;
 327     int           nn0,nn1,nouter,ninner;
 328     real         shX,shY,shZ;
 329     real         fscal,tx,ty,tz;
 330     real         rinvsq;
 331     real         jq;
 332     real         qq,vcoul,vctot;
 333     int           nti;
 334     int           tj;
 335     real         rinvsix;
 336     real         Vvdw6,Vvdwtot;
 337     real         Vvdw12;
 338     real         krsq;
 339     real         ix1,iy1,iz1,fix1,fiy1,fiz1;
 340     real         ix2,iy2,iz2,fix2,fiy2,fiz2;
 341     real         ix3,iy3,iz3,fix3,fiy3,fiz3;
 342     real         jx1,jy1,jz1,fjx1,fjy1,fjz1;
 343     real         dx11,dy11,dz11,rsq11,rinv11;
 344     real         dx21,dy21,dz21,rsq21,rinv21;
 345     real         dx31,dy31,dz31,rsq31,rinv31;
 346     real         qO,qH;
 347     real         c6,c12;
 348     real         weight_cg1, weight_cg2, weight_product;
 349     real         hybscal;
 350
 351     nri              = *p_nri;
 352     ntype            = *p_ntype;
 353     nthreads         = *p_nthreads;
 354     facel            = *p_facel;
 355     krf              = *p_krf;
 356     crf              = *p_crf;
 357     tabscale         = *p_tabscale;
 358     ii               = iinr[0];
 359     qO               = facel*charge[ii];
 360     qH               = facel*charge[ii+1];
 361     nti              = 2*ntype*type[ii];
 362
 363     nouter           = 0;
 364     ninner           = 0;
 365
 366     do
 367     {
 368         #ifdef GMX_THREAD_SHM_FDECOMP
 369         tMPI_Thread_mutex_lock((tMPI_Thread_mutex_t *)mtx);
 370         nn0              = *count;
 371         nn1              = nn0+(nri-nn0)/(2*nthreads)+10;
 372         *count           = nn1;
 373         tMPI_Thread_mutex_unlock((tMPI_Thread_mutex_t *)mtx);
 374         if(nn1>nri) nn1=nri;
 375         #else
 376         nn0 = 0;
 377         nn1 = nri;
 378         #endif
 379
 380         for(n=nn0; (n<nn1); n++)
 381         {
 382             is3              = 3*shift[n];
 383             shX              = shiftvec[is3];
 384             shY              = shiftvec[is3+1];
 385             shZ              = shiftvec[is3+2];
 386             nj0              = jindex[n];
 387             nj1              = jindex[n+1];
 388             ii               = iinr[n];
 389             ii3              = 3*ii;
 390             ix1              = shX + pos[ii3+0];
 391             iy1              = shY + pos[ii3+1];
 392             iz1              = shZ + pos[ii3+2];
 393             ix2              = shX + pos[ii3+3];
 394             iy2              = shY + pos[ii3+4];
 395             iz2              = shZ + pos[ii3+5];
 396             ix3              = shX + pos[ii3+6];
 397             iy3              = shY + pos[ii3+7];
 398             iz3              = shZ + pos[ii3+8];
 399             weight_cg1       = wf[ii];
 400             vctot            = 0;
 401             Vvdwtot          = 0;
 402             fix1             = 0;
 403             fiy1             = 0;
 404             fiz1             = 0;
 405             fix2             = 0;
 406             fiy2             = 0;
 407             fiz2             = 0;
 408             fix3             = 0;
 409             fiy3             = 0;
 410             fiz3             = 0;
 411
 412             for(k=nj0; (k<nj1); k++)
 413             {
 414                 jnr              = jjnr[k];
 415                 weight_cg2       = wf[jnr];
 416                 weight_product   = weight_cg1*weight_cg2;
 417                 if (weight_product < ALMOST_ZERO) {
 418                 /* force is zero, skip this molecule */
 419                  continue;
 420                 }
 421                 else if (weight_product >= ALMOST_ONE)
 422                 {
 423                        hybscal = 1.0;
 424                 }
 425                 else
 426                 {
 427                    hybscal = weight_product;
 428                 }
 429                 j3               = 3*jnr;
 430                 jx1              = pos[j3+0];
 431                 jy1              = pos[j3+1];
 432                 jz1              = pos[j3+2];
 433                 dx11             = ix1 - jx1;
 434                 dy11             = iy1 - jy1;
 435                 dz11             = iz1 - jz1;
 436                 rsq11            = dx11*dx11+dy11*dy11+dz11*dz11;
 437                 dx21             = ix2 - jx1;
 438                 dy21             = iy2 - jy1;
 439                 dz21             = iz2 - jz1;
 440                 rsq21            = dx21*dx21+dy21*dy21+dz21*dz21;
 441                 dx31             = ix3 - jx1;
 442                 dy31             = iy3 - jy1;
 443                 dz31             = iz3 - jz1;
 444                 rsq31            = dx31*dx31+dy31*dy31+dz31*dz31;
 445                 rinv11           = 1.0/sqrt(rsq11);
 446                 rinv21           = 1.0/sqrt(rsq21);
 447                 rinv31           = 1.0/sqrt(rsq31);
 448                 jq               = charge[jnr+0];
 449                 qq               = qO*jq;
 450                 tj               = nti+2*type[jnr];
 451                 c6               = vdwparam[tj];
 452                 c12              = vdwparam[tj+1];
 453                 rinvsq           = rinv11*rinv11;
 454                 krsq             = krf*rsq11;
 455                 vcoul            = qq*(rinv11+krsq-crf);
 456                 vctot            = vctot+vcoul;
 457                 rinvsix          = rinvsq*rinvsq*rinvsq;
 458                 Vvdw6            = c6*rinvsix;
 459                 Vvdw12           = c12*rinvsix*rinvsix;
 460                 Vvdwtot          = Vvdwtot+Vvdw12-Vvdw6;
 461                 fscal            = (qq*(rinv11-2.0*krsq)+12.0*Vvdw12-6.0*Vvdw6)*rinvsq;
 462                 fscal *= hybscal;
 463                 if(force_cap>0 && (fabs(fscal)> force_cap)){
 464                 fscal=force_cap*fscal/fabs(fscal);
 465                 }
 466                 tx               = fscal*dx11;
 467                 ty               = fscal*dy11;
 468                 tz               = fscal*dz11;
 469                 fix1             = fix1 + tx;
 470                 fiy1             = fiy1 + ty;
 471                 fiz1             = fiz1 + tz;
 472                 fjx1             = faction[j3+0] - tx;
 473                 fjy1             = faction[j3+1] - ty;
 474                 fjz1             = faction[j3+2] - tz;
 475                 qq               = qH*jq;
 476                 rinvsq           = rinv21*rinv21;
 477                 krsq             = krf*rsq21;
 478                 vcoul            = qq*(rinv21+krsq-crf);
 479                 vctot            = vctot+vcoul;
 480                 fscal            = (qq*(rinv21-2.0*krsq))*rinvsq;
 481                 fscal *= hybscal;
 482                 if(force_cap>0 && (fabs(fscal)> force_cap)){
 483                 fscal=force_cap*fscal/fabs(fscal);
 484                 }
 485                 tx               = fscal*dx21;
 486                 ty               = fscal*dy21;
 487                 tz               = fscal*dz21;
 488                 fix2             = fix2 + tx;
 489                 fiy2             = fiy2 + ty;
 490                 fiz2             = fiz2 + tz;
 491                 fjx1             = fjx1 - tx;
 492                 fjy1             = fjy1 - ty;
 493                 fjz1             = fjz1 - tz;
 494                 rinvsq           = rinv31*rinv31;
 495                 krsq             = krf*rsq31;
 496                 vcoul            = qq*(rinv31+krsq-crf);
 497                 vctot            = vctot+vcoul;
 498                 fscal            = (qq*(rinv31-2.0*krsq))*rinvsq;
 499                 fscal *= hybscal;
 500                 if(force_cap>0 && (fabs(fscal)> force_cap)){
 501                 fscal=force_cap*fscal/fabs(fscal);
 502                 }
 503                 tx               = fscal*dx31;
 504                 ty               = fscal*dy31;
 505                 tz               = fscal*dz31;
 506                 fix3             = fix3 + tx;
 507                 fiy3             = fiy3 + ty;
 508                 fiz3             = fiz3 + tz;
 509                 faction[j3+0]    = fjx1 - tx;
 510                 faction[j3+1]    = fjy1 - ty;
 511                 faction[j3+2]    = fjz1 - tz;
 512             }
 513
 514             faction[ii3+0]   = faction[ii3+0] + fix1;
 515             faction[ii3+1]   = faction[ii3+1] + fiy1;
 516             faction[ii3+2]   = faction[ii3+2] + fiz1;
 517             faction[ii3+3]   = faction[ii3+3] + fix2;
 518             faction[ii3+4]   = faction[ii3+4] + fiy2;
 519             faction[ii3+5]   = faction[ii3+5] + fiz2;
 520             faction[ii3+6]   = faction[ii3+6] + fix3;
 521             faction[ii3+7]   = faction[ii3+7] + fiy3;
 522             faction[ii3+8]   = faction[ii3+8] + fiz3;
 523             fshift[is3]      = fshift[is3]+fix1+fix2+fix3;
 524             fshift[is3+1]    = fshift[is3+1]+fiy1+fiy2+fiy3;
 525             fshift[is3+2]    = fshift[is3+2]+fiz1+fiz2+fiz3;
 526             ggid             = gid[n];
 527             Vc[ggid]         = Vc[ggid] + vctot;
 528             Vvdw[ggid]       = Vvdw[ggid] + Vvdwtot;
 529             ninner           = ninner + nj1 - nj0;
 530         }
 531
 532         nouter           = nouter + nn1 - nn0;
 533     }
 534     while (nn1<nri);
 535
 536     *outeriter       = nouter;
 537     *inneriter       = ninner;
 538 }
 539
 540