src/gromacs/gmxlib/nonbonded/nb_kernel_c/nb_kernel_ElecCSTab_VdwBham_GeomP1P1_c.c

   1 /*
   2  * This file is part of the GROMACS molecular simulation package.
   3  *
   4  * Copyright (c) 2012,2013, by the GROMACS development team, led by
   5  * Mark Abraham, David van der Spoel, Berk Hess, and Erik Lindahl,
   6  * and including many others, as listed in the AUTHORS file in the
   7  * top-level source directory and at http://www.gromacs.org.
   8  *
   9  * GROMACS is free software; you can redistribute it and/or
  10  * modify it under the terms of the GNU Lesser General Public License
  11  * as published by the Free Software Foundation; either version 2.1
  12  * of the License, or (at your option) any later version.
  13  *
  14  * GROMACS is distributed in the hope that it will be useful,
  15  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  16  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  17  * Lesser General Public License for more details.
  18  *
  19  * You should have received a copy of the GNU Lesser General Public
  20  * License along with GROMACS; if not, see
  21  * http://www.gnu.org/licenses, or write to the Free Software Foundation,
  22  * Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA.
  23  *
  24  * If you want to redistribute modifications to GROMACS, please
  25  * consider that scientific software is very special. Version
  26  * control is crucial - bugs must be traceable. We will be happy to
  27  * consider code for inclusion in the official distribution, but
  28  * derived work must not be called official GROMACS. Details are found
  29  * in the README & COPYING files - if they are missing, get the
  30  * official version at http://www.gromacs.org.
  31  *
  32  * To help us fund GROMACS development, we humbly ask that you cite
  33  * the research papers on the package. Check out http://www.gromacs.org.
  34  */
  35 /*
  36  * Note: this file was generated by the GROMACS c kernel generator.
  37  */
  38 #ifdef HAVE_CONFIG_H
  39 #include <config.h>
  40 #endif
  41
  42 #include <math.h>
  43
  44 #include "../nb_kernel.h"
  45 #include "types/simple.h"
  46 #include "vec.h"
  47 #include "nrnb.h"
  48
  49 /*
  50  * Gromacs nonbonded kernel:   nb_kernel_ElecCSTab_VdwBham_GeomP1P1_VF_c
  51  * Electrostatics interaction: CubicSplineTable
  52  * VdW interaction:            Buckingham
  53  * Geometry:                   Particle-Particle
  54  * Calculate force/pot:        PotentialAndForce
  55  */
  56 void
  57 nb_kernel_ElecCSTab_VdwBham_GeomP1P1_VF_c
  58                     (t_nblist                    * gmx_restrict       nlist,
  59                      rvec                        * gmx_restrict          xx,
  60                      rvec                        * gmx_restrict          ff,
  61                      t_forcerec                  * gmx_restrict          fr,
  62                      t_mdatoms                   * gmx_restrict     mdatoms,
  63                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
  64                      t_nrnb                      * gmx_restrict        nrnb)
  65 {
  66     int              i_shift_offset,i_coord_offset,j_coord_offset;
  67     int              j_index_start,j_index_end;
  68     int              nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
  69     real             shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
  70     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  71     real             *shiftvec,*fshift,*x,*f;
  72     int              vdwioffset0;
  73     real             ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
  74     int              vdwjidx0;
  75     real             jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  76     real             dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
  77     real             velec,felec,velecsum,facel,crf,krf,krf2;
  78     real             *charge;
  79     int              nvdwtype;
  80     real             rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
  81     int              *vdwtype;
  82     real             *vdwparam;
  83     int              vfitab;
  84     real             rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF;
  85     real             *vftab;
  86
  87     x                = xx[0];
  88     f                = ff[0];
  89
  90     nri              = nlist->nri;
  91     iinr             = nlist->iinr;
  92     jindex           = nlist->jindex;
  93     jjnr             = nlist->jjnr;
  94     shiftidx         = nlist->shift;
  95     gid              = nlist->gid;
  96     shiftvec         = fr->shift_vec[0];
  97     fshift           = fr->fshift[0];
  98     facel            = fr->epsfac;
  99     charge           = mdatoms->chargeA;
 100     nvdwtype         = fr->ntype;
 101     vdwparam         = fr->nbfp;
 102     vdwtype          = mdatoms->typeA;
 103
 104     vftab            = kernel_data->table_elec->data;
 105     vftabscale       = kernel_data->table_elec->scale;
 106
 107     outeriter        = 0;
 108     inneriter        = 0;
 109
 110     /* Start outer loop over neighborlists */
 111     for(iidx=0; iidx<nri; iidx++)
 112     {
 113         /* Load shift vector for this list */
 114         i_shift_offset   = DIM*shiftidx[iidx];
 115         shX              = shiftvec[i_shift_offset+XX];
 116         shY              = shiftvec[i_shift_offset+YY];
 117         shZ              = shiftvec[i_shift_offset+ZZ];
 118
 119         /* Load limits for loop over neighbors */
 120         j_index_start    = jindex[iidx];
 121         j_index_end      = jindex[iidx+1];
 122
 123         /* Get outer coordinate index */
 124         inr              = iinr[iidx];
 125         i_coord_offset   = DIM*inr;
 126
 127         /* Load i particle coords and add shift vector */
 128         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 129         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 130         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 131
 132         fix0             = 0.0;
 133         fiy0             = 0.0;
 134         fiz0             = 0.0;
 135
 136         /* Load parameters for i particles */
 137         iq0              = facel*charge[inr+0];
 138         vdwioffset0      = 3*nvdwtype*vdwtype[inr+0];
 139
 140         /* Reset potential sums */
 141         velecsum         = 0.0;
 142         vvdwsum          = 0.0;
 143
 144         /* Start inner kernel loop */
 145         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 146         {
 147             /* Get j neighbor index, and coordinate index */
 148             jnr              = jjnr[jidx];
 149             j_coord_offset   = DIM*jnr;
 150
 151             /* load j atom coordinates */
 152             jx0              = x[j_coord_offset+DIM*0+XX];
 153             jy0              = x[j_coord_offset+DIM*0+YY];
 154             jz0              = x[j_coord_offset+DIM*0+ZZ];
 155
 156             /* Calculate displacement vector */
 157             dx00             = ix0 - jx0;
 158             dy00             = iy0 - jy0;
 159             dz00             = iz0 - jz0;
 160
 161             /* Calculate squared distance and things based on it */
 162             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 163
 164             rinv00           = gmx_invsqrt(rsq00);
 165
 166             rinvsq00         = rinv00*rinv00;
 167
 168             /* Load parameters for j particles */
 169             jq0              = charge[jnr+0];
 170             vdwjidx0         = 3*vdwtype[jnr+0];
 171
 172             /**************************
 173              * CALCULATE INTERACTIONS *
 174              **************************/
 175
 176             r00              = rsq00*rinv00;
 177
 178             qq00             = iq0*jq0;
 179             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 180             cexp1_00         = vdwparam[vdwioffset0+vdwjidx0+1];
 181             cexp2_00         = vdwparam[vdwioffset0+vdwjidx0+2];
 182
 183             /* Calculate table index by multiplying r with table scale and truncate to integer */
 184             rt               = r00*vftabscale;
 185             vfitab           = rt;
 186             vfeps            = rt-vfitab;
 187             vfitab           = 1*4*vfitab;
 188
 189             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 190             Y                = vftab[vfitab];
 191             F                = vftab[vfitab+1];
 192             Geps             = vfeps*vftab[vfitab+2];
 193             Heps2            = vfeps*vfeps*vftab[vfitab+3];
 194             Fp               = F+Geps+Heps2;
 195             VV               = Y+vfeps*Fp;
 196             velec            = qq00*VV;
 197             FF               = Fp+Geps+2.0*Heps2;
 198             felec            = -qq00*FF*vftabscale*rinv00;
 199
 200             /* BUCKINGHAM DISPERSION/REPULSION */
 201             rinvsix          = rinvsq00*rinvsq00*rinvsq00;
 202             vvdw6            = c6_00*rinvsix;
 203             br               = cexp2_00*r00;
 204             vvdwexp          = cexp1_00*exp(-br);
 205             vvdw             = vvdwexp - vvdw6*(1.0/6.0);
 206             fvdw             = (br*vvdwexp-vvdw6)*rinvsq00;
 207
 208             /* Update potential sums from outer loop */
 209             velecsum        += velec;
 210             vvdwsum         += vvdw;
 211
 212             fscal            = felec+fvdw;
 213
 214             /* Calculate temporary vectorial force */
 215             tx               = fscal*dx00;
 216             ty               = fscal*dy00;
 217             tz               = fscal*dz00;
 218
 219             /* Update vectorial force */
 220             fix0            += tx;
 221             fiy0            += ty;
 222             fiz0            += tz;
 223             f[j_coord_offset+DIM*0+XX] -= tx;
 224             f[j_coord_offset+DIM*0+YY] -= ty;
 225             f[j_coord_offset+DIM*0+ZZ] -= tz;
 226
 227             /* Inner loop uses 81 flops */
 228         }
 229         /* End of innermost loop */
 230
 231         tx = ty = tz = 0;
 232         f[i_coord_offset+DIM*0+XX] += fix0;
 233         f[i_coord_offset+DIM*0+YY] += fiy0;
 234         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 235         tx                         += fix0;
 236         ty                         += fiy0;
 237         tz                         += fiz0;
 238         fshift[i_shift_offset+XX]  += tx;
 239         fshift[i_shift_offset+YY]  += ty;
 240         fshift[i_shift_offset+ZZ]  += tz;
 241
 242         ggid                        = gid[iidx];
 243         /* Update potential energies */
 244         kernel_data->energygrp_elec[ggid] += velecsum;
 245         kernel_data->energygrp_vdw[ggid] += vvdwsum;
 246
 247         /* Increment number of inner iterations */
 248         inneriter                  += j_index_end - j_index_start;
 249
 250         /* Outer loop uses 15 flops */
 251     }
 252
 253     /* Increment number of outer iterations */
 254     outeriter        += nri;
 255
 256     /* Update outer/inner flops */
 257
 258     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_VF,outeriter*15 + inneriter*81);
 259 }
 260 /*
 261  * Gromacs nonbonded kernel:   nb_kernel_ElecCSTab_VdwBham_GeomP1P1_F_c
 262  * Electrostatics interaction: CubicSplineTable
 263  * VdW interaction:            Buckingham
 264  * Geometry:                   Particle-Particle
 265  * Calculate force/pot:        Force
 266  */
 267 void
 268 nb_kernel_ElecCSTab_VdwBham_GeomP1P1_F_c
 269                     (t_nblist                    * gmx_restrict       nlist,
 270                      rvec                        * gmx_restrict          xx,
 271                      rvec                        * gmx_restrict          ff,
 272                      t_forcerec                  * gmx_restrict          fr,
 273                      t_mdatoms                   * gmx_restrict     mdatoms,
 274                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 275                      t_nrnb                      * gmx_restrict        nrnb)
 276 {
 277     int              i_shift_offset,i_coord_offset,j_coord_offset;
 278     int              j_index_start,j_index_end;
 279     int              nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
 280     real             shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
 281     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 282     real             *shiftvec,*fshift,*x,*f;
 283     int              vdwioffset0;
 284     real             ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 285     int              vdwjidx0;
 286     real             jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 287     real             dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
 288     real             velec,felec,velecsum,facel,crf,krf,krf2;
 289     real             *charge;
 290     int              nvdwtype;
 291     real             rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
 292     int              *vdwtype;
 293     real             *vdwparam;
 294     int              vfitab;
 295     real             rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF;
 296     real             *vftab;
 297
 298     x                = xx[0];
 299     f                = ff[0];
 300
 301     nri              = nlist->nri;
 302     iinr             = nlist->iinr;
 303     jindex           = nlist->jindex;
 304     jjnr             = nlist->jjnr;
 305     shiftidx         = nlist->shift;
 306     gid              = nlist->gid;
 307     shiftvec         = fr->shift_vec[0];
 308     fshift           = fr->fshift[0];
 309     facel            = fr->epsfac;
 310     charge           = mdatoms->chargeA;
 311     nvdwtype         = fr->ntype;
 312     vdwparam         = fr->nbfp;
 313     vdwtype          = mdatoms->typeA;
 314
 315     vftab            = kernel_data->table_elec->data;
 316     vftabscale       = kernel_data->table_elec->scale;
 317
 318     outeriter        = 0;
 319     inneriter        = 0;
 320
 321     /* Start outer loop over neighborlists */
 322     for(iidx=0; iidx<nri; iidx++)
 323     {
 324         /* Load shift vector for this list */
 325         i_shift_offset   = DIM*shiftidx[iidx];
 326         shX              = shiftvec[i_shift_offset+XX];
 327         shY              = shiftvec[i_shift_offset+YY];
 328         shZ              = shiftvec[i_shift_offset+ZZ];
 329
 330         /* Load limits for loop over neighbors */
 331         j_index_start    = jindex[iidx];
 332         j_index_end      = jindex[iidx+1];
 333
 334         /* Get outer coordinate index */
 335         inr              = iinr[iidx];
 336         i_coord_offset   = DIM*inr;
 337
 338         /* Load i particle coords and add shift vector */
 339         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 340         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 341         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 342
 343         fix0             = 0.0;
 344         fiy0             = 0.0;
 345         fiz0             = 0.0;
 346
 347         /* Load parameters for i particles */
 348         iq0              = facel*charge[inr+0];
 349         vdwioffset0      = 3*nvdwtype*vdwtype[inr+0];
 350
 351         /* Start inner kernel loop */
 352         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 353         {
 354             /* Get j neighbor index, and coordinate index */
 355             jnr              = jjnr[jidx];
 356             j_coord_offset   = DIM*jnr;
 357
 358             /* load j atom coordinates */
 359             jx0              = x[j_coord_offset+DIM*0+XX];
 360             jy0              = x[j_coord_offset+DIM*0+YY];
 361             jz0              = x[j_coord_offset+DIM*0+ZZ];
 362
 363             /* Calculate displacement vector */
 364             dx00             = ix0 - jx0;
 365             dy00             = iy0 - jy0;
 366             dz00             = iz0 - jz0;
 367
 368             /* Calculate squared distance and things based on it */
 369             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 370
 371             rinv00           = gmx_invsqrt(rsq00);
 372
 373             rinvsq00         = rinv00*rinv00;
 374
 375             /* Load parameters for j particles */
 376             jq0              = charge[jnr+0];
 377             vdwjidx0         = 3*vdwtype[jnr+0];
 378
 379             /**************************
 380              * CALCULATE INTERACTIONS *
 381              **************************/
 382
 383             r00              = rsq00*rinv00;
 384
 385             qq00             = iq0*jq0;
 386             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 387             cexp1_00         = vdwparam[vdwioffset0+vdwjidx0+1];
 388             cexp2_00         = vdwparam[vdwioffset0+vdwjidx0+2];
 389
 390             /* Calculate table index by multiplying r with table scale and truncate to integer */
 391             rt               = r00*vftabscale;
 392             vfitab           = rt;
 393             vfeps            = rt-vfitab;
 394             vfitab           = 1*4*vfitab;
 395
 396             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 397             F                = vftab[vfitab+1];
 398             Geps             = vfeps*vftab[vfitab+2];
 399             Heps2            = vfeps*vfeps*vftab[vfitab+3];
 400             Fp               = F+Geps+Heps2;
 401             FF               = Fp+Geps+2.0*Heps2;
 402             felec            = -qq00*FF*vftabscale*rinv00;
 403
 404             /* BUCKINGHAM DISPERSION/REPULSION */
 405             rinvsix          = rinvsq00*rinvsq00*rinvsq00;
 406             vvdw6            = c6_00*rinvsix;
 407             br               = cexp2_00*r00;
 408             vvdwexp          = cexp1_00*exp(-br);
 409             fvdw             = (br*vvdwexp-vvdw6)*rinvsq00;
 410
 411             fscal            = felec+fvdw;
 412
 413             /* Calculate temporary vectorial force */
 414             tx               = fscal*dx00;
 415             ty               = fscal*dy00;
 416             tz               = fscal*dz00;
 417
 418             /* Update vectorial force */
 419             fix0            += tx;
 420             fiy0            += ty;
 421             fiz0            += tz;
 422             f[j_coord_offset+DIM*0+XX] -= tx;
 423             f[j_coord_offset+DIM*0+YY] -= ty;
 424             f[j_coord_offset+DIM*0+ZZ] -= tz;
 425
 426             /* Inner loop uses 74 flops */
 427         }
 428         /* End of innermost loop */
 429
 430         tx = ty = tz = 0;
 431         f[i_coord_offset+DIM*0+XX] += fix0;
 432         f[i_coord_offset+DIM*0+YY] += fiy0;
 433         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 434         tx                         += fix0;
 435         ty                         += fiy0;
 436         tz                         += fiz0;
 437         fshift[i_shift_offset+XX]  += tx;
 438         fshift[i_shift_offset+YY]  += ty;
 439         fshift[i_shift_offset+ZZ]  += tz;
 440
 441         /* Increment number of inner iterations */
 442         inneriter                  += j_index_end - j_index_start;
 443
 444         /* Outer loop uses 13 flops */
 445     }
 446
 447     /* Increment number of outer iterations */
 448     outeriter        += nri;
 449
 450     /* Update outer/inner flops */
 451
 452     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_F,outeriter*13 + inneriter*74);
 453 }