src/gromacs/gmxlib/nonbonded/nb_kernel_c/nb_kernel_ElecCSTab_VdwLJ_GeomP1P1_c.c

   1 /*
   2  * This file is part of the GROMACS molecular simulation package.
   3  *
   4  * Copyright (c) 2012,2013,2014, 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 #include "config.h"
  39
  40 #include <math.h>
  41
  42 #include "../nb_kernel.h"
  43 #include "gromacs/legacyheaders/types/simple.h"
  44 #include "gromacs/math/vec.h"
  45 #include "gromacs/legacyheaders/nrnb.h"
  46
  47 /*
  48  * Gromacs nonbonded kernel:   nb_kernel_ElecCSTab_VdwLJ_GeomP1P1_VF_c
  49  * Electrostatics interaction: CubicSplineTable
  50  * VdW interaction:            LennardJones
  51  * Geometry:                   Particle-Particle
  52  * Calculate force/pot:        PotentialAndForce
  53  */
  54 void
  55 nb_kernel_ElecCSTab_VdwLJ_GeomP1P1_VF_c
  56                     (t_nblist                    * gmx_restrict       nlist,
  57                      rvec                        * gmx_restrict          xx,
  58                      rvec                        * gmx_restrict          ff,
  59                      t_forcerec                  * gmx_restrict          fr,
  60                      t_mdatoms                   * gmx_restrict     mdatoms,
  61                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
  62                      t_nrnb                      * gmx_restrict        nrnb)
  63 {
  64     int              i_shift_offset,i_coord_offset,j_coord_offset;
  65     int              j_index_start,j_index_end;
  66     int              nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
  67     real             shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
  68     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  69     real             *shiftvec,*fshift,*x,*f;
  70     int              vdwioffset0;
  71     real             ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
  72     int              vdwjidx0;
  73     real             jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  74     real             dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
  75     real             velec,felec,velecsum,facel,crf,krf,krf2;
  76     real             *charge;
  77     int              nvdwtype;
  78     real             rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
  79     int              *vdwtype;
  80     real             *vdwparam;
  81     int              vfitab;
  82     real             rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF;
  83     real             *vftab;
  84
  85     x                = xx[0];
  86     f                = ff[0];
  87
  88     nri              = nlist->nri;
  89     iinr             = nlist->iinr;
  90     jindex           = nlist->jindex;
  91     jjnr             = nlist->jjnr;
  92     shiftidx         = nlist->shift;
  93     gid              = nlist->gid;
  94     shiftvec         = fr->shift_vec[0];
  95     fshift           = fr->fshift[0];
  96     facel            = fr->epsfac;
  97     charge           = mdatoms->chargeA;
  98     nvdwtype         = fr->ntype;
  99     vdwparam         = fr->nbfp;
 100     vdwtype          = mdatoms->typeA;
 101
 102     vftab            = kernel_data->table_elec->data;
 103     vftabscale       = kernel_data->table_elec->scale;
 104
 105     outeriter        = 0;
 106     inneriter        = 0;
 107
 108     /* Start outer loop over neighborlists */
 109     for(iidx=0; iidx<nri; iidx++)
 110     {
 111         /* Load shift vector for this list */
 112         i_shift_offset   = DIM*shiftidx[iidx];
 113         shX              = shiftvec[i_shift_offset+XX];
 114         shY              = shiftvec[i_shift_offset+YY];
 115         shZ              = shiftvec[i_shift_offset+ZZ];
 116
 117         /* Load limits for loop over neighbors */
 118         j_index_start    = jindex[iidx];
 119         j_index_end      = jindex[iidx+1];
 120
 121         /* Get outer coordinate index */
 122         inr              = iinr[iidx];
 123         i_coord_offset   = DIM*inr;
 124
 125         /* Load i particle coords and add shift vector */
 126         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 127         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 128         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 129
 130         fix0             = 0.0;
 131         fiy0             = 0.0;
 132         fiz0             = 0.0;
 133
 134         /* Load parameters for i particles */
 135         iq0              = facel*charge[inr+0];
 136         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 137
 138         /* Reset potential sums */
 139         velecsum         = 0.0;
 140         vvdwsum          = 0.0;
 141
 142         /* Start inner kernel loop */
 143         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 144         {
 145             /* Get j neighbor index, and coordinate index */
 146             jnr              = jjnr[jidx];
 147             j_coord_offset   = DIM*jnr;
 148
 149             /* load j atom coordinates */
 150             jx0              = x[j_coord_offset+DIM*0+XX];
 151             jy0              = x[j_coord_offset+DIM*0+YY];
 152             jz0              = x[j_coord_offset+DIM*0+ZZ];
 153
 154             /* Calculate displacement vector */
 155             dx00             = ix0 - jx0;
 156             dy00             = iy0 - jy0;
 157             dz00             = iz0 - jz0;
 158
 159             /* Calculate squared distance and things based on it */
 160             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 161
 162             rinv00           = gmx_invsqrt(rsq00);
 163
 164             rinvsq00         = rinv00*rinv00;
 165
 166             /* Load parameters for j particles */
 167             jq0              = charge[jnr+0];
 168             vdwjidx0         = 2*vdwtype[jnr+0];
 169
 170             /**************************
 171              * CALCULATE INTERACTIONS *
 172              **************************/
 173
 174             r00              = rsq00*rinv00;
 175
 176             qq00             = iq0*jq0;
 177             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 178             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
 179
 180             /* Calculate table index by multiplying r with table scale and truncate to integer */
 181             rt               = r00*vftabscale;
 182             vfitab           = rt;
 183             vfeps            = rt-vfitab;
 184             vfitab           = 1*4*vfitab;
 185
 186             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 187             Y                = vftab[vfitab];
 188             F                = vftab[vfitab+1];
 189             Geps             = vfeps*vftab[vfitab+2];
 190             Heps2            = vfeps*vfeps*vftab[vfitab+3];
 191             Fp               = F+Geps+Heps2;
 192             VV               = Y+vfeps*Fp;
 193             velec            = qq00*VV;
 194             FF               = Fp+Geps+2.0*Heps2;
 195             felec            = -qq00*FF*vftabscale*rinv00;
 196
 197             /* LENNARD-JONES DISPERSION/REPULSION */
 198
 199             rinvsix          = rinvsq00*rinvsq00*rinvsq00;
 200             vvdw6            = c6_00*rinvsix;
 201             vvdw12           = c12_00*rinvsix*rinvsix;
 202             vvdw             = vvdw12*(1.0/12.0) - vvdw6*(1.0/6.0);
 203             fvdw             = (vvdw12-vvdw6)*rinvsq00;
 204
 205             /* Update potential sums from outer loop */
 206             velecsum        += velec;
 207             vvdwsum         += vvdw;
 208
 209             fscal            = felec+fvdw;
 210
 211             /* Calculate temporary vectorial force */
 212             tx               = fscal*dx00;
 213             ty               = fscal*dy00;
 214             tz               = fscal*dz00;
 215
 216             /* Update vectorial force */
 217             fix0            += tx;
 218             fiy0            += ty;
 219             fiz0            += tz;
 220             f[j_coord_offset+DIM*0+XX] -= tx;
 221             f[j_coord_offset+DIM*0+YY] -= ty;
 222             f[j_coord_offset+DIM*0+ZZ] -= tz;
 223
 224             /* Inner loop uses 55 flops */
 225         }
 226         /* End of innermost loop */
 227
 228         tx = ty = tz = 0;
 229         f[i_coord_offset+DIM*0+XX] += fix0;
 230         f[i_coord_offset+DIM*0+YY] += fiy0;
 231         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 232         tx                         += fix0;
 233         ty                         += fiy0;
 234         tz                         += fiz0;
 235         fshift[i_shift_offset+XX]  += tx;
 236         fshift[i_shift_offset+YY]  += ty;
 237         fshift[i_shift_offset+ZZ]  += tz;
 238
 239         ggid                        = gid[iidx];
 240         /* Update potential energies */
 241         kernel_data->energygrp_elec[ggid] += velecsum;
 242         kernel_data->energygrp_vdw[ggid] += vvdwsum;
 243
 244         /* Increment number of inner iterations */
 245         inneriter                  += j_index_end - j_index_start;
 246
 247         /* Outer loop uses 15 flops */
 248     }
 249
 250     /* Increment number of outer iterations */
 251     outeriter        += nri;
 252
 253     /* Update outer/inner flops */
 254
 255     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_VF,outeriter*15 + inneriter*55);
 256 }
 257 /*
 258  * Gromacs nonbonded kernel:   nb_kernel_ElecCSTab_VdwLJ_GeomP1P1_F_c
 259  * Electrostatics interaction: CubicSplineTable
 260  * VdW interaction:            LennardJones
 261  * Geometry:                   Particle-Particle
 262  * Calculate force/pot:        Force
 263  */
 264 void
 265 nb_kernel_ElecCSTab_VdwLJ_GeomP1P1_F_c
 266                     (t_nblist                    * gmx_restrict       nlist,
 267                      rvec                        * gmx_restrict          xx,
 268                      rvec                        * gmx_restrict          ff,
 269                      t_forcerec                  * gmx_restrict          fr,
 270                      t_mdatoms                   * gmx_restrict     mdatoms,
 271                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 272                      t_nrnb                      * gmx_restrict        nrnb)
 273 {
 274     int              i_shift_offset,i_coord_offset,j_coord_offset;
 275     int              j_index_start,j_index_end;
 276     int              nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
 277     real             shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
 278     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 279     real             *shiftvec,*fshift,*x,*f;
 280     int              vdwioffset0;
 281     real             ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 282     int              vdwjidx0;
 283     real             jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 284     real             dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
 285     real             velec,felec,velecsum,facel,crf,krf,krf2;
 286     real             *charge;
 287     int              nvdwtype;
 288     real             rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
 289     int              *vdwtype;
 290     real             *vdwparam;
 291     int              vfitab;
 292     real             rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF;
 293     real             *vftab;
 294
 295     x                = xx[0];
 296     f                = ff[0];
 297
 298     nri              = nlist->nri;
 299     iinr             = nlist->iinr;
 300     jindex           = nlist->jindex;
 301     jjnr             = nlist->jjnr;
 302     shiftidx         = nlist->shift;
 303     gid              = nlist->gid;
 304     shiftvec         = fr->shift_vec[0];
 305     fshift           = fr->fshift[0];
 306     facel            = fr->epsfac;
 307     charge           = mdatoms->chargeA;
 308     nvdwtype         = fr->ntype;
 309     vdwparam         = fr->nbfp;
 310     vdwtype          = mdatoms->typeA;
 311
 312     vftab            = kernel_data->table_elec->data;
 313     vftabscale       = kernel_data->table_elec->scale;
 314
 315     outeriter        = 0;
 316     inneriter        = 0;
 317
 318     /* Start outer loop over neighborlists */
 319     for(iidx=0; iidx<nri; iidx++)
 320     {
 321         /* Load shift vector for this list */
 322         i_shift_offset   = DIM*shiftidx[iidx];
 323         shX              = shiftvec[i_shift_offset+XX];
 324         shY              = shiftvec[i_shift_offset+YY];
 325         shZ              = shiftvec[i_shift_offset+ZZ];
 326
 327         /* Load limits for loop over neighbors */
 328         j_index_start    = jindex[iidx];
 329         j_index_end      = jindex[iidx+1];
 330
 331         /* Get outer coordinate index */
 332         inr              = iinr[iidx];
 333         i_coord_offset   = DIM*inr;
 334
 335         /* Load i particle coords and add shift vector */
 336         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 337         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 338         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 339
 340         fix0             = 0.0;
 341         fiy0             = 0.0;
 342         fiz0             = 0.0;
 343
 344         /* Load parameters for i particles */
 345         iq0              = facel*charge[inr+0];
 346         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 347
 348         /* Start inner kernel loop */
 349         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 350         {
 351             /* Get j neighbor index, and coordinate index */
 352             jnr              = jjnr[jidx];
 353             j_coord_offset   = DIM*jnr;
 354
 355             /* load j atom coordinates */
 356             jx0              = x[j_coord_offset+DIM*0+XX];
 357             jy0              = x[j_coord_offset+DIM*0+YY];
 358             jz0              = x[j_coord_offset+DIM*0+ZZ];
 359
 360             /* Calculate displacement vector */
 361             dx00             = ix0 - jx0;
 362             dy00             = iy0 - jy0;
 363             dz00             = iz0 - jz0;
 364
 365             /* Calculate squared distance and things based on it */
 366             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 367
 368             rinv00           = gmx_invsqrt(rsq00);
 369
 370             rinvsq00         = rinv00*rinv00;
 371
 372             /* Load parameters for j particles */
 373             jq0              = charge[jnr+0];
 374             vdwjidx0         = 2*vdwtype[jnr+0];
 375
 376             /**************************
 377              * CALCULATE INTERACTIONS *
 378              **************************/
 379
 380             r00              = rsq00*rinv00;
 381
 382             qq00             = iq0*jq0;
 383             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 384             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
 385
 386             /* Calculate table index by multiplying r with table scale and truncate to integer */
 387             rt               = r00*vftabscale;
 388             vfitab           = rt;
 389             vfeps            = rt-vfitab;
 390             vfitab           = 1*4*vfitab;
 391
 392             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 393             F                = vftab[vfitab+1];
 394             Geps             = vfeps*vftab[vfitab+2];
 395             Heps2            = vfeps*vfeps*vftab[vfitab+3];
 396             Fp               = F+Geps+Heps2;
 397             FF               = Fp+Geps+2.0*Heps2;
 398             felec            = -qq00*FF*vftabscale*rinv00;
 399
 400             /* LENNARD-JONES DISPERSION/REPULSION */
 401
 402             rinvsix          = rinvsq00*rinvsq00*rinvsq00;
 403             fvdw             = (c12_00*rinvsix-c6_00)*rinvsix*rinvsq00;
 404
 405             fscal            = felec+fvdw;
 406
 407             /* Calculate temporary vectorial force */
 408             tx               = fscal*dx00;
 409             ty               = fscal*dy00;
 410             tz               = fscal*dz00;
 411
 412             /* Update vectorial force */
 413             fix0            += tx;
 414             fiy0            += ty;
 415             fiz0            += tz;
 416             f[j_coord_offset+DIM*0+XX] -= tx;
 417             f[j_coord_offset+DIM*0+YY] -= ty;
 418             f[j_coord_offset+DIM*0+ZZ] -= tz;
 419
 420             /* Inner loop uses 46 flops */
 421         }
 422         /* End of innermost loop */
 423
 424         tx = ty = tz = 0;
 425         f[i_coord_offset+DIM*0+XX] += fix0;
 426         f[i_coord_offset+DIM*0+YY] += fiy0;
 427         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 428         tx                         += fix0;
 429         ty                         += fiy0;
 430         tz                         += fiz0;
 431         fshift[i_shift_offset+XX]  += tx;
 432         fshift[i_shift_offset+YY]  += ty;
 433         fshift[i_shift_offset+ZZ]  += tz;
 434
 435         /* Increment number of inner iterations */
 436         inneriter                  += j_index_end - j_index_start;
 437
 438         /* Outer loop uses 13 flops */
 439     }
 440
 441     /* Increment number of outer iterations */
 442     outeriter        += nri;
 443
 444     /* Update outer/inner flops */
 445
 446     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_F,outeriter*13 + inneriter*46);
 447 }