src/gromacs/gmxlib/nonbonded/nb_kernel_c/nb_kernel_ElecCSTab_VdwCSTab_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_VdwCSTab_GeomP1P1_VF_c
  49  * Electrostatics interaction: CubicSplineTable
  50  * VdW interaction:            CubicSplineTable
  51  * Geometry:                   Particle-Particle
  52  * Calculate force/pot:        PotentialAndForce
  53  */
  54 void
  55 nb_kernel_ElecCSTab_VdwCSTab_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_vdw->data;
 103     vftabscale       = kernel_data->table_elec_vdw->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             /* Load parameters for j particles */
 165             jq0              = charge[jnr+0];
 166             vdwjidx0         = 2*vdwtype[jnr+0];
 167
 168             /**************************
 169              * CALCULATE INTERACTIONS *
 170              **************************/
 171
 172             r00              = rsq00*rinv00;
 173
 174             qq00             = iq0*jq0;
 175             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 176             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
 177
 178             /* Calculate table index by multiplying r with table scale and truncate to integer */
 179             rt               = r00*vftabscale;
 180             vfitab           = rt;
 181             vfeps            = rt-vfitab;
 182             vfitab           = 3*4*vfitab;
 183
 184             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 185             Y                = vftab[vfitab];
 186             F                = vftab[vfitab+1];
 187             Geps             = vfeps*vftab[vfitab+2];
 188             Heps2            = vfeps*vfeps*vftab[vfitab+3];
 189             Fp               = F+Geps+Heps2;
 190             VV               = Y+vfeps*Fp;
 191             velec            = qq00*VV;
 192             FF               = Fp+Geps+2.0*Heps2;
 193             felec            = -qq00*FF*vftabscale*rinv00;
 194
 195             /* CUBIC SPLINE TABLE DISPERSION */
 196             vfitab          += 4;
 197             Y                = vftab[vfitab];
 198             F                = vftab[vfitab+1];
 199             Geps             = vfeps*vftab[vfitab+2];
 200             Heps2            = vfeps*vfeps*vftab[vfitab+3];
 201             Fp               = F+Geps+Heps2;
 202             VV               = Y+vfeps*Fp;
 203             vvdw6            = c6_00*VV;
 204             FF               = Fp+Geps+2.0*Heps2;
 205             fvdw6            = c6_00*FF;
 206
 207             /* CUBIC SPLINE TABLE REPULSION */
 208             Y                = vftab[vfitab+4];
 209             F                = vftab[vfitab+5];
 210             Geps             = vfeps*vftab[vfitab+6];
 211             Heps2            = vfeps*vfeps*vftab[vfitab+7];
 212             Fp               = F+Geps+Heps2;
 213             VV               = Y+vfeps*Fp;
 214             vvdw12           = c12_00*VV;
 215             FF               = Fp+Geps+2.0*Heps2;
 216             fvdw12           = c12_00*FF;
 217             vvdw             = vvdw12+vvdw6;
 218             fvdw             = -(fvdw6+fvdw12)*vftabscale*rinv00;
 219
 220             /* Update potential sums from outer loop */
 221             velecsum        += velec;
 222             vvdwsum         += vvdw;
 223
 224             fscal            = felec+fvdw;
 225
 226             /* Calculate temporary vectorial force */
 227             tx               = fscal*dx00;
 228             ty               = fscal*dy00;
 229             tz               = fscal*dz00;
 230
 231             /* Update vectorial force */
 232             fix0            += tx;
 233             fiy0            += ty;
 234             fiz0            += tz;
 235             f[j_coord_offset+DIM*0+XX] -= tx;
 236             f[j_coord_offset+DIM*0+YY] -= ty;
 237             f[j_coord_offset+DIM*0+ZZ] -= tz;
 238
 239             /* Inner loop uses 73 flops */
 240         }
 241         /* End of innermost loop */
 242
 243         tx = ty = tz = 0;
 244         f[i_coord_offset+DIM*0+XX] += fix0;
 245         f[i_coord_offset+DIM*0+YY] += fiy0;
 246         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 247         tx                         += fix0;
 248         ty                         += fiy0;
 249         tz                         += fiz0;
 250         fshift[i_shift_offset+XX]  += tx;
 251         fshift[i_shift_offset+YY]  += ty;
 252         fshift[i_shift_offset+ZZ]  += tz;
 253
 254         ggid                        = gid[iidx];
 255         /* Update potential energies */
 256         kernel_data->energygrp_elec[ggid] += velecsum;
 257         kernel_data->energygrp_vdw[ggid] += vvdwsum;
 258
 259         /* Increment number of inner iterations */
 260         inneriter                  += j_index_end - j_index_start;
 261
 262         /* Outer loop uses 15 flops */
 263     }
 264
 265     /* Increment number of outer iterations */
 266     outeriter        += nri;
 267
 268     /* Update outer/inner flops */
 269
 270     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_VF,outeriter*15 + inneriter*73);
 271 }
 272 /*
 273  * Gromacs nonbonded kernel:   nb_kernel_ElecCSTab_VdwCSTab_GeomP1P1_F_c
 274  * Electrostatics interaction: CubicSplineTable
 275  * VdW interaction:            CubicSplineTable
 276  * Geometry:                   Particle-Particle
 277  * Calculate force/pot:        Force
 278  */
 279 void
 280 nb_kernel_ElecCSTab_VdwCSTab_GeomP1P1_F_c
 281                     (t_nblist                    * gmx_restrict       nlist,
 282                      rvec                        * gmx_restrict          xx,
 283                      rvec                        * gmx_restrict          ff,
 284                      t_forcerec                  * gmx_restrict          fr,
 285                      t_mdatoms                   * gmx_restrict     mdatoms,
 286                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 287                      t_nrnb                      * gmx_restrict        nrnb)
 288 {
 289     int              i_shift_offset,i_coord_offset,j_coord_offset;
 290     int              j_index_start,j_index_end;
 291     int              nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
 292     real             shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
 293     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 294     real             *shiftvec,*fshift,*x,*f;
 295     int              vdwioffset0;
 296     real             ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 297     int              vdwjidx0;
 298     real             jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 299     real             dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
 300     real             velec,felec,velecsum,facel,crf,krf,krf2;
 301     real             *charge;
 302     int              nvdwtype;
 303     real             rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
 304     int              *vdwtype;
 305     real             *vdwparam;
 306     int              vfitab;
 307     real             rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF;
 308     real             *vftab;
 309
 310     x                = xx[0];
 311     f                = ff[0];
 312
 313     nri              = nlist->nri;
 314     iinr             = nlist->iinr;
 315     jindex           = nlist->jindex;
 316     jjnr             = nlist->jjnr;
 317     shiftidx         = nlist->shift;
 318     gid              = nlist->gid;
 319     shiftvec         = fr->shift_vec[0];
 320     fshift           = fr->fshift[0];
 321     facel            = fr->epsfac;
 322     charge           = mdatoms->chargeA;
 323     nvdwtype         = fr->ntype;
 324     vdwparam         = fr->nbfp;
 325     vdwtype          = mdatoms->typeA;
 326
 327     vftab            = kernel_data->table_elec_vdw->data;
 328     vftabscale       = kernel_data->table_elec_vdw->scale;
 329
 330     outeriter        = 0;
 331     inneriter        = 0;
 332
 333     /* Start outer loop over neighborlists */
 334     for(iidx=0; iidx<nri; iidx++)
 335     {
 336         /* Load shift vector for this list */
 337         i_shift_offset   = DIM*shiftidx[iidx];
 338         shX              = shiftvec[i_shift_offset+XX];
 339         shY              = shiftvec[i_shift_offset+YY];
 340         shZ              = shiftvec[i_shift_offset+ZZ];
 341
 342         /* Load limits for loop over neighbors */
 343         j_index_start    = jindex[iidx];
 344         j_index_end      = jindex[iidx+1];
 345
 346         /* Get outer coordinate index */
 347         inr              = iinr[iidx];
 348         i_coord_offset   = DIM*inr;
 349
 350         /* Load i particle coords and add shift vector */
 351         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 352         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 353         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 354
 355         fix0             = 0.0;
 356         fiy0             = 0.0;
 357         fiz0             = 0.0;
 358
 359         /* Load parameters for i particles */
 360         iq0              = facel*charge[inr+0];
 361         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 362
 363         /* Start inner kernel loop */
 364         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 365         {
 366             /* Get j neighbor index, and coordinate index */
 367             jnr              = jjnr[jidx];
 368             j_coord_offset   = DIM*jnr;
 369
 370             /* load j atom coordinates */
 371             jx0              = x[j_coord_offset+DIM*0+XX];
 372             jy0              = x[j_coord_offset+DIM*0+YY];
 373             jz0              = x[j_coord_offset+DIM*0+ZZ];
 374
 375             /* Calculate displacement vector */
 376             dx00             = ix0 - jx0;
 377             dy00             = iy0 - jy0;
 378             dz00             = iz0 - jz0;
 379
 380             /* Calculate squared distance and things based on it */
 381             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 382
 383             rinv00           = gmx_invsqrt(rsq00);
 384
 385             /* Load parameters for j particles */
 386             jq0              = charge[jnr+0];
 387             vdwjidx0         = 2*vdwtype[jnr+0];
 388
 389             /**************************
 390              * CALCULATE INTERACTIONS *
 391              **************************/
 392
 393             r00              = rsq00*rinv00;
 394
 395             qq00             = iq0*jq0;
 396             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 397             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
 398
 399             /* Calculate table index by multiplying r with table scale and truncate to integer */
 400             rt               = r00*vftabscale;
 401             vfitab           = rt;
 402             vfeps            = rt-vfitab;
 403             vfitab           = 3*4*vfitab;
 404
 405             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 406             F                = vftab[vfitab+1];
 407             Geps             = vfeps*vftab[vfitab+2];
 408             Heps2            = vfeps*vfeps*vftab[vfitab+3];
 409             Fp               = F+Geps+Heps2;
 410             FF               = Fp+Geps+2.0*Heps2;
 411             felec            = -qq00*FF*vftabscale*rinv00;
 412
 413             /* CUBIC SPLINE TABLE DISPERSION */
 414             vfitab          += 4;
 415             F                = vftab[vfitab+1];
 416             Geps             = vfeps*vftab[vfitab+2];
 417             Heps2            = vfeps*vfeps*vftab[vfitab+3];
 418             Fp               = F+Geps+Heps2;
 419             FF               = Fp+Geps+2.0*Heps2;
 420             fvdw6            = c6_00*FF;
 421
 422             /* CUBIC SPLINE TABLE REPULSION */
 423             F                = vftab[vfitab+5];
 424             Geps             = vfeps*vftab[vfitab+6];
 425             Heps2            = vfeps*vfeps*vftab[vfitab+7];
 426             Fp               = F+Geps+Heps2;
 427             FF               = Fp+Geps+2.0*Heps2;
 428             fvdw12           = c12_00*FF;
 429             fvdw             = -(fvdw6+fvdw12)*vftabscale*rinv00;
 430
 431             fscal            = felec+fvdw;
 432
 433             /* Calculate temporary vectorial force */
 434             tx               = fscal*dx00;
 435             ty               = fscal*dy00;
 436             tz               = fscal*dz00;
 437
 438             /* Update vectorial force */
 439             fix0            += tx;
 440             fiy0            += ty;
 441             fiz0            += tz;
 442             f[j_coord_offset+DIM*0+XX] -= tx;
 443             f[j_coord_offset+DIM*0+YY] -= ty;
 444             f[j_coord_offset+DIM*0+ZZ] -= tz;
 445
 446             /* Inner loop uses 61 flops */
 447         }
 448         /* End of innermost loop */
 449
 450         tx = ty = tz = 0;
 451         f[i_coord_offset+DIM*0+XX] += fix0;
 452         f[i_coord_offset+DIM*0+YY] += fiy0;
 453         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 454         tx                         += fix0;
 455         ty                         += fiy0;
 456         tz                         += fiz0;
 457         fshift[i_shift_offset+XX]  += tx;
 458         fshift[i_shift_offset+YY]  += ty;
 459         fshift[i_shift_offset+ZZ]  += tz;
 460
 461         /* Increment number of inner iterations */
 462         inneriter                  += j_index_end - j_index_start;
 463
 464         /* Outer loop uses 13 flops */
 465     }
 466
 467     /* Increment number of outer iterations */
 468     outeriter        += nri;
 469
 470     /* Update outer/inner flops */
 471
 472     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_F,outeriter*13 + inneriter*61);
 473 }