src/gromacs/gmxlib/nonbonded/nb_kernel_c/nb_kernel_ElecCoul_VdwCSTab_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_ElecCoul_VdwCSTab_GeomP1P1_VF_c
  51  * Electrostatics interaction: Coulomb
  52  * VdW interaction:            CubicSplineTable
  53  * Geometry:                   Particle-Particle
  54  * Calculate force/pot:        PotentialAndForce
  55  */
  56 void
  57 nb_kernel_ElecCoul_VdwCSTab_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_vdw->data;
 105     vftabscale       = kernel_data->table_vdw->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           = 2*4*vfitab;
 187
 188             /* COULOMB ELECTROSTATICS */
 189             velec            = qq00*rinv00;
 190             felec            = velec*rinvsq00;
 191
 192             /* CUBIC SPLINE TABLE DISPERSION */
 193             vfitab          += 0;
 194             Y                = vftab[vfitab];
 195             F                = vftab[vfitab+1];
 196             Geps             = vfeps*vftab[vfitab+2];
 197             Heps2            = vfeps*vfeps*vftab[vfitab+3];
 198             Fp               = F+Geps+Heps2;
 199             VV               = Y+vfeps*Fp;
 200             vvdw6            = c6_00*VV;
 201             FF               = Fp+Geps+2.0*Heps2;
 202             fvdw6            = c6_00*FF;
 203
 204             /* CUBIC SPLINE TABLE REPULSION */
 205             Y                = vftab[vfitab+4];
 206             F                = vftab[vfitab+5];
 207             Geps             = vfeps*vftab[vfitab+6];
 208             Heps2            = vfeps*vfeps*vftab[vfitab+7];
 209             Fp               = F+Geps+Heps2;
 210             VV               = Y+vfeps*Fp;
 211             vvdw12           = c12_00*VV;
 212             FF               = Fp+Geps+2.0*Heps2;
 213             fvdw12           = c12_00*FF;
 214             vvdw             = vvdw12+vvdw6;
 215             fvdw             = -(fvdw6+fvdw12)*vftabscale*rinv00;
 216
 217             /* Update potential sums from outer loop */
 218             velecsum        += velec;
 219             vvdwsum         += vvdw;
 220
 221             fscal            = felec+fvdw;
 222
 223             /* Calculate temporary vectorial force */
 224             tx               = fscal*dx00;
 225             ty               = fscal*dy00;
 226             tz               = fscal*dz00;
 227
 228             /* Update vectorial force */
 229             fix0            += tx;
 230             fiy0            += ty;
 231             fiz0            += tz;
 232             f[j_coord_offset+DIM*0+XX] -= tx;
 233             f[j_coord_offset+DIM*0+YY] -= ty;
 234             f[j_coord_offset+DIM*0+ZZ] -= tz;
 235
 236             /* Inner loop uses 62 flops */
 237         }
 238         /* End of innermost loop */
 239
 240         tx = ty = tz = 0;
 241         f[i_coord_offset+DIM*0+XX] += fix0;
 242         f[i_coord_offset+DIM*0+YY] += fiy0;
 243         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 244         tx                         += fix0;
 245         ty                         += fiy0;
 246         tz                         += fiz0;
 247         fshift[i_shift_offset+XX]  += tx;
 248         fshift[i_shift_offset+YY]  += ty;
 249         fshift[i_shift_offset+ZZ]  += tz;
 250
 251         ggid                        = gid[iidx];
 252         /* Update potential energies */
 253         kernel_data->energygrp_elec[ggid] += velecsum;
 254         kernel_data->energygrp_vdw[ggid] += vvdwsum;
 255
 256         /* Increment number of inner iterations */
 257         inneriter                  += j_index_end - j_index_start;
 258
 259         /* Outer loop uses 15 flops */
 260     }
 261
 262     /* Increment number of outer iterations */
 263     outeriter        += nri;
 264
 265     /* Update outer/inner flops */
 266
 267     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_VF,outeriter*15 + inneriter*62);
 268 }
 269 /*
 270  * Gromacs nonbonded kernel:   nb_kernel_ElecCoul_VdwCSTab_GeomP1P1_F_c
 271  * Electrostatics interaction: Coulomb
 272  * VdW interaction:            CubicSplineTable
 273  * Geometry:                   Particle-Particle
 274  * Calculate force/pot:        Force
 275  */
 276 void
 277 nb_kernel_ElecCoul_VdwCSTab_GeomP1P1_F_c
 278                     (t_nblist                    * gmx_restrict       nlist,
 279                      rvec                        * gmx_restrict          xx,
 280                      rvec                        * gmx_restrict          ff,
 281                      t_forcerec                  * gmx_restrict          fr,
 282                      t_mdatoms                   * gmx_restrict     mdatoms,
 283                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 284                      t_nrnb                      * gmx_restrict        nrnb)
 285 {
 286     int              i_shift_offset,i_coord_offset,j_coord_offset;
 287     int              j_index_start,j_index_end;
 288     int              nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
 289     real             shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
 290     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 291     real             *shiftvec,*fshift,*x,*f;
 292     int              vdwioffset0;
 293     real             ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 294     int              vdwjidx0;
 295     real             jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 296     real             dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
 297     real             velec,felec,velecsum,facel,crf,krf,krf2;
 298     real             *charge;
 299     int              nvdwtype;
 300     real             rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
 301     int              *vdwtype;
 302     real             *vdwparam;
 303     int              vfitab;
 304     real             rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF;
 305     real             *vftab;
 306
 307     x                = xx[0];
 308     f                = ff[0];
 309
 310     nri              = nlist->nri;
 311     iinr             = nlist->iinr;
 312     jindex           = nlist->jindex;
 313     jjnr             = nlist->jjnr;
 314     shiftidx         = nlist->shift;
 315     gid              = nlist->gid;
 316     shiftvec         = fr->shift_vec[0];
 317     fshift           = fr->fshift[0];
 318     facel            = fr->epsfac;
 319     charge           = mdatoms->chargeA;
 320     nvdwtype         = fr->ntype;
 321     vdwparam         = fr->nbfp;
 322     vdwtype          = mdatoms->typeA;
 323
 324     vftab            = kernel_data->table_vdw->data;
 325     vftabscale       = kernel_data->table_vdw->scale;
 326
 327     outeriter        = 0;
 328     inneriter        = 0;
 329
 330     /* Start outer loop over neighborlists */
 331     for(iidx=0; iidx<nri; iidx++)
 332     {
 333         /* Load shift vector for this list */
 334         i_shift_offset   = DIM*shiftidx[iidx];
 335         shX              = shiftvec[i_shift_offset+XX];
 336         shY              = shiftvec[i_shift_offset+YY];
 337         shZ              = shiftvec[i_shift_offset+ZZ];
 338
 339         /* Load limits for loop over neighbors */
 340         j_index_start    = jindex[iidx];
 341         j_index_end      = jindex[iidx+1];
 342
 343         /* Get outer coordinate index */
 344         inr              = iinr[iidx];
 345         i_coord_offset   = DIM*inr;
 346
 347         /* Load i particle coords and add shift vector */
 348         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 349         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 350         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 351
 352         fix0             = 0.0;
 353         fiy0             = 0.0;
 354         fiz0             = 0.0;
 355
 356         /* Load parameters for i particles */
 357         iq0              = facel*charge[inr+0];
 358         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 359
 360         /* Start inner kernel loop */
 361         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 362         {
 363             /* Get j neighbor index, and coordinate index */
 364             jnr              = jjnr[jidx];
 365             j_coord_offset   = DIM*jnr;
 366
 367             /* load j atom coordinates */
 368             jx0              = x[j_coord_offset+DIM*0+XX];
 369             jy0              = x[j_coord_offset+DIM*0+YY];
 370             jz0              = x[j_coord_offset+DIM*0+ZZ];
 371
 372             /* Calculate displacement vector */
 373             dx00             = ix0 - jx0;
 374             dy00             = iy0 - jy0;
 375             dz00             = iz0 - jz0;
 376
 377             /* Calculate squared distance and things based on it */
 378             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 379
 380             rinv00           = gmx_invsqrt(rsq00);
 381
 382             rinvsq00         = rinv00*rinv00;
 383
 384             /* Load parameters for j particles */
 385             jq0              = charge[jnr+0];
 386             vdwjidx0         = 2*vdwtype[jnr+0];
 387
 388             /**************************
 389              * CALCULATE INTERACTIONS *
 390              **************************/
 391
 392             r00              = rsq00*rinv00;
 393
 394             qq00             = iq0*jq0;
 395             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 396             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
 397
 398             /* Calculate table index by multiplying r with table scale and truncate to integer */
 399             rt               = r00*vftabscale;
 400             vfitab           = rt;
 401             vfeps            = rt-vfitab;
 402             vfitab           = 2*4*vfitab;
 403
 404             /* COULOMB ELECTROSTATICS */
 405             velec            = qq00*rinv00;
 406             felec            = velec*rinvsq00;
 407
 408             /* CUBIC SPLINE TABLE DISPERSION */
 409             vfitab          += 0;
 410             F                = vftab[vfitab+1];
 411             Geps             = vfeps*vftab[vfitab+2];
 412             Heps2            = vfeps*vfeps*vftab[vfitab+3];
 413             Fp               = F+Geps+Heps2;
 414             FF               = Fp+Geps+2.0*Heps2;
 415             fvdw6            = c6_00*FF;
 416
 417             /* CUBIC SPLINE TABLE REPULSION */
 418             F                = vftab[vfitab+5];
 419             Geps             = vfeps*vftab[vfitab+6];
 420             Heps2            = vfeps*vfeps*vftab[vfitab+7];
 421             Fp               = F+Geps+Heps2;
 422             FF               = Fp+Geps+2.0*Heps2;
 423             fvdw12           = c12_00*FF;
 424             fvdw             = -(fvdw6+fvdw12)*vftabscale*rinv00;
 425
 426             fscal            = felec+fvdw;
 427
 428             /* Calculate temporary vectorial force */
 429             tx               = fscal*dx00;
 430             ty               = fscal*dy00;
 431             tz               = fscal*dz00;
 432
 433             /* Update vectorial force */
 434             fix0            += tx;
 435             fiy0            += ty;
 436             fiz0            += tz;
 437             f[j_coord_offset+DIM*0+XX] -= tx;
 438             f[j_coord_offset+DIM*0+YY] -= ty;
 439             f[j_coord_offset+DIM*0+ZZ] -= tz;
 440
 441             /* Inner loop uses 53 flops */
 442         }
 443         /* End of innermost loop */
 444
 445         tx = ty = tz = 0;
 446         f[i_coord_offset+DIM*0+XX] += fix0;
 447         f[i_coord_offset+DIM*0+YY] += fiy0;
 448         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 449         tx                         += fix0;
 450         ty                         += fiy0;
 451         tz                         += fiz0;
 452         fshift[i_shift_offset+XX]  += tx;
 453         fshift[i_shift_offset+YY]  += ty;
 454         fshift[i_shift_offset+ZZ]  += tz;
 455
 456         /* Increment number of inner iterations */
 457         inneriter                  += j_index_end - j_index_start;
 458
 459         /* Outer loop uses 13 flops */
 460     }
 461
 462     /* Increment number of outer iterations */
 463     outeriter        += nri;
 464
 465     /* Update outer/inner flops */
 466
 467     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_F,outeriter*13 + inneriter*53);
 468 }