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 "gmxpre.h"
  39
  40 #include "config.h"
  41
  42 #include <math.h>
  43
  44 #include "../nb_kernel.h"
  45 #include "gromacs/legacyheaders/types/simple.h"
  46 #include "gromacs/math/vec.h"
  47 #include "gromacs/legacyheaders/nrnb.h"
  48
  49 /*
  50  * Gromacs nonbonded kernel:   nb_kernel_ElecCSTab_VdwLJ_GeomP1P1_VF_c
  51  * Electrostatics interaction: CubicSplineTable
  52  * VdW interaction:            LennardJones
  53  * Geometry:                   Particle-Particle
  54  * Calculate force/pot:        PotentialAndForce
  55  */
  56 void
  57 nb_kernel_ElecCSTab_VdwLJ_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      = 2*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         = 2*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             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
 181
 182             /* Calculate table index by multiplying r with table scale and truncate to integer */
 183             rt               = r00*vftabscale;
 184             vfitab           = rt;
 185             vfeps            = rt-vfitab;
 186             vfitab           = 1*4*vfitab;
 187
 188             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 189             Y                = vftab[vfitab];
 190             F                = vftab[vfitab+1];
 191             Geps             = vfeps*vftab[vfitab+2];
 192             Heps2            = vfeps*vfeps*vftab[vfitab+3];
 193             Fp               = F+Geps+Heps2;
 194             VV               = Y+vfeps*Fp;
 195             velec            = qq00*VV;
 196             FF               = Fp+Geps+2.0*Heps2;
 197             felec            = -qq00*FF*vftabscale*rinv00;
 198
 199             /* LENNARD-JONES DISPERSION/REPULSION */
 200
 201             rinvsix          = rinvsq00*rinvsq00*rinvsq00;
 202             vvdw6            = c6_00*rinvsix;
 203             vvdw12           = c12_00*rinvsix*rinvsix;
 204             vvdw             = vvdw12*(1.0/12.0) - vvdw6*(1.0/6.0);
 205             fvdw             = (vvdw12-vvdw6)*rinvsq00;
 206
 207             /* Update potential sums from outer loop */
 208             velecsum        += velec;
 209             vvdwsum         += vvdw;
 210
 211             fscal            = felec+fvdw;
 212
 213             /* Calculate temporary vectorial force */
 214             tx               = fscal*dx00;
 215             ty               = fscal*dy00;
 216             tz               = fscal*dz00;
 217
 218             /* Update vectorial force */
 219             fix0            += tx;
 220             fiy0            += ty;
 221             fiz0            += tz;
 222             f[j_coord_offset+DIM*0+XX] -= tx;
 223             f[j_coord_offset+DIM*0+YY] -= ty;
 224             f[j_coord_offset+DIM*0+ZZ] -= tz;
 225
 226             /* Inner loop uses 55 flops */
 227         }
 228         /* End of innermost loop */
 229
 230         tx = ty = tz = 0;
 231         f[i_coord_offset+DIM*0+XX] += fix0;
 232         f[i_coord_offset+DIM*0+YY] += fiy0;
 233         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 234         tx                         += fix0;
 235         ty                         += fiy0;
 236         tz                         += fiz0;
 237         fshift[i_shift_offset+XX]  += tx;
 238         fshift[i_shift_offset+YY]  += ty;
 239         fshift[i_shift_offset+ZZ]  += tz;
 240
 241         ggid                        = gid[iidx];
 242         /* Update potential energies */
 243         kernel_data->energygrp_elec[ggid] += velecsum;
 244         kernel_data->energygrp_vdw[ggid] += vvdwsum;
 245
 246         /* Increment number of inner iterations */
 247         inneriter                  += j_index_end - j_index_start;
 248
 249         /* Outer loop uses 15 flops */
 250     }
 251
 252     /* Increment number of outer iterations */
 253     outeriter        += nri;
 254
 255     /* Update outer/inner flops */
 256
 257     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_VF,outeriter*15 + inneriter*55);
 258 }
 259 /*
 260  * Gromacs nonbonded kernel:   nb_kernel_ElecCSTab_VdwLJ_GeomP1P1_F_c
 261  * Electrostatics interaction: CubicSplineTable
 262  * VdW interaction:            LennardJones
 263  * Geometry:                   Particle-Particle
 264  * Calculate force/pot:        Force
 265  */
 266 void
 267 nb_kernel_ElecCSTab_VdwLJ_GeomP1P1_F_c
 268                     (t_nblist                    * gmx_restrict       nlist,
 269                      rvec                        * gmx_restrict          xx,
 270                      rvec                        * gmx_restrict          ff,
 271                      t_forcerec                  * gmx_restrict          fr,
 272                      t_mdatoms                   * gmx_restrict     mdatoms,
 273                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 274                      t_nrnb                      * gmx_restrict        nrnb)
 275 {
 276     int              i_shift_offset,i_coord_offset,j_coord_offset;
 277     int              j_index_start,j_index_end;
 278     int              nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
 279     real             shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
 280     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 281     real             *shiftvec,*fshift,*x,*f;
 282     int              vdwioffset0;
 283     real             ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 284     int              vdwjidx0;
 285     real             jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 286     real             dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
 287     real             velec,felec,velecsum,facel,crf,krf,krf2;
 288     real             *charge;
 289     int              nvdwtype;
 290     real             rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
 291     int              *vdwtype;
 292     real             *vdwparam;
 293     int              vfitab;
 294     real             rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF;
 295     real             *vftab;
 296
 297     x                = xx[0];
 298     f                = ff[0];
 299
 300     nri              = nlist->nri;
 301     iinr             = nlist->iinr;
 302     jindex           = nlist->jindex;
 303     jjnr             = nlist->jjnr;
 304     shiftidx         = nlist->shift;
 305     gid              = nlist->gid;
 306     shiftvec         = fr->shift_vec[0];
 307     fshift           = fr->fshift[0];
 308     facel            = fr->epsfac;
 309     charge           = mdatoms->chargeA;
 310     nvdwtype         = fr->ntype;
 311     vdwparam         = fr->nbfp;
 312     vdwtype          = mdatoms->typeA;
 313
 314     vftab            = kernel_data->table_elec->data;
 315     vftabscale       = kernel_data->table_elec->scale;
 316
 317     outeriter        = 0;
 318     inneriter        = 0;
 319
 320     /* Start outer loop over neighborlists */
 321     for(iidx=0; iidx<nri; iidx++)
 322     {
 323         /* Load shift vector for this list */
 324         i_shift_offset   = DIM*shiftidx[iidx];
 325         shX              = shiftvec[i_shift_offset+XX];
 326         shY              = shiftvec[i_shift_offset+YY];
 327         shZ              = shiftvec[i_shift_offset+ZZ];
 328
 329         /* Load limits for loop over neighbors */
 330         j_index_start    = jindex[iidx];
 331         j_index_end      = jindex[iidx+1];
 332
 333         /* Get outer coordinate index */
 334         inr              = iinr[iidx];
 335         i_coord_offset   = DIM*inr;
 336
 337         /* Load i particle coords and add shift vector */
 338         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 339         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 340         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 341
 342         fix0             = 0.0;
 343         fiy0             = 0.0;
 344         fiz0             = 0.0;
 345
 346         /* Load parameters for i particles */
 347         iq0              = facel*charge[inr+0];
 348         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 349
 350         /* Start inner kernel loop */
 351         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 352         {
 353             /* Get j neighbor index, and coordinate index */
 354             jnr              = jjnr[jidx];
 355             j_coord_offset   = DIM*jnr;
 356
 357             /* load j atom coordinates */
 358             jx0              = x[j_coord_offset+DIM*0+XX];
 359             jy0              = x[j_coord_offset+DIM*0+YY];
 360             jz0              = x[j_coord_offset+DIM*0+ZZ];
 361
 362             /* Calculate displacement vector */
 363             dx00             = ix0 - jx0;
 364             dy00             = iy0 - jy0;
 365             dz00             = iz0 - jz0;
 366
 367             /* Calculate squared distance and things based on it */
 368             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 369
 370             rinv00           = gmx_invsqrt(rsq00);
 371
 372             rinvsq00         = rinv00*rinv00;
 373
 374             /* Load parameters for j particles */
 375             jq0              = charge[jnr+0];
 376             vdwjidx0         = 2*vdwtype[jnr+0];
 377
 378             /**************************
 379              * CALCULATE INTERACTIONS *
 380              **************************/
 381
 382             r00              = rsq00*rinv00;
 383
 384             qq00             = iq0*jq0;
 385             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 386             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
 387
 388             /* Calculate table index by multiplying r with table scale and truncate to integer */
 389             rt               = r00*vftabscale;
 390             vfitab           = rt;
 391             vfeps            = rt-vfitab;
 392             vfitab           = 1*4*vfitab;
 393
 394             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 395             F                = vftab[vfitab+1];
 396             Geps             = vfeps*vftab[vfitab+2];
 397             Heps2            = vfeps*vfeps*vftab[vfitab+3];
 398             Fp               = F+Geps+Heps2;
 399             FF               = Fp+Geps+2.0*Heps2;
 400             felec            = -qq00*FF*vftabscale*rinv00;
 401
 402             /* LENNARD-JONES DISPERSION/REPULSION */
 403
 404             rinvsix          = rinvsq00*rinvsq00*rinvsq00;
 405             fvdw             = (c12_00*rinvsix-c6_00)*rinvsix*rinvsq00;
 406
 407             fscal            = felec+fvdw;
 408
 409             /* Calculate temporary vectorial force */
 410             tx               = fscal*dx00;
 411             ty               = fscal*dy00;
 412             tz               = fscal*dz00;
 413
 414             /* Update vectorial force */
 415             fix0            += tx;
 416             fiy0            += ty;
 417             fiz0            += tz;
 418             f[j_coord_offset+DIM*0+XX] -= tx;
 419             f[j_coord_offset+DIM*0+YY] -= ty;
 420             f[j_coord_offset+DIM*0+ZZ] -= tz;
 421
 422             /* Inner loop uses 46 flops */
 423         }
 424         /* End of innermost loop */
 425
 426         tx = ty = tz = 0;
 427         f[i_coord_offset+DIM*0+XX] += fix0;
 428         f[i_coord_offset+DIM*0+YY] += fiy0;
 429         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 430         tx                         += fix0;
 431         ty                         += fiy0;
 432         tz                         += fiz0;
 433         fshift[i_shift_offset+XX]  += tx;
 434         fshift[i_shift_offset+YY]  += ty;
 435         fshift[i_shift_offset+ZZ]  += tz;
 436
 437         /* Increment number of inner iterations */
 438         inneriter                  += j_index_end - j_index_start;
 439
 440         /* Outer loop uses 13 flops */
 441     }
 442
 443     /* Increment number of outer iterations */
 444     outeriter        += nri;
 445
 446     /* Update outer/inner flops */
 447
 448     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_F,outeriter*13 + inneriter*46);
 449 }