src/gromacs/gmxlib/nonbonded/nb_kernel_c/nb_kernel_ElecNone_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 "types/simple.h"
  44 #include "gromacs/math/vec.h"
  45 #include "nrnb.h"
  46
  47 /*
  48  * Gromacs nonbonded kernel:   nb_kernel_ElecNone_VdwCSTab_GeomP1P1_VF_c
  49  * Electrostatics interaction: None
  50  * VdW interaction:            CubicSplineTable
  51  * Geometry:                   Particle-Particle
  52  * Calculate force/pot:        PotentialAndForce
  53  */
  54 void
  55 nb_kernel_ElecNone_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     int              nvdwtype;
  76     real             rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
  77     int              *vdwtype;
  78     real             *vdwparam;
  79     int              vfitab;
  80     real             rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF;
  81     real             *vftab;
  82
  83     x                = xx[0];
  84     f                = ff[0];
  85
  86     nri              = nlist->nri;
  87     iinr             = nlist->iinr;
  88     jindex           = nlist->jindex;
  89     jjnr             = nlist->jjnr;
  90     shiftidx         = nlist->shift;
  91     gid              = nlist->gid;
  92     shiftvec         = fr->shift_vec[0];
  93     fshift           = fr->fshift[0];
  94     nvdwtype         = fr->ntype;
  95     vdwparam         = fr->nbfp;
  96     vdwtype          = mdatoms->typeA;
  97
  98     vftab            = kernel_data->table_vdw->data;
  99     vftabscale       = kernel_data->table_vdw->scale;
 100
 101     outeriter        = 0;
 102     inneriter        = 0;
 103
 104     /* Start outer loop over neighborlists */
 105     for(iidx=0; iidx<nri; iidx++)
 106     {
 107         /* Load shift vector for this list */
 108         i_shift_offset   = DIM*shiftidx[iidx];
 109         shX              = shiftvec[i_shift_offset+XX];
 110         shY              = shiftvec[i_shift_offset+YY];
 111         shZ              = shiftvec[i_shift_offset+ZZ];
 112
 113         /* Load limits for loop over neighbors */
 114         j_index_start    = jindex[iidx];
 115         j_index_end      = jindex[iidx+1];
 116
 117         /* Get outer coordinate index */
 118         inr              = iinr[iidx];
 119         i_coord_offset   = DIM*inr;
 120
 121         /* Load i particle coords and add shift vector */
 122         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 123         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 124         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 125
 126         fix0             = 0.0;
 127         fiy0             = 0.0;
 128         fiz0             = 0.0;
 129
 130         /* Load parameters for i particles */
 131         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 132
 133         /* Reset potential sums */
 134         vvdwsum          = 0.0;
 135
 136         /* Start inner kernel loop */
 137         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 138         {
 139             /* Get j neighbor index, and coordinate index */
 140             jnr              = jjnr[jidx];
 141             j_coord_offset   = DIM*jnr;
 142
 143             /* load j atom coordinates */
 144             jx0              = x[j_coord_offset+DIM*0+XX];
 145             jy0              = x[j_coord_offset+DIM*0+YY];
 146             jz0              = x[j_coord_offset+DIM*0+ZZ];
 147
 148             /* Calculate displacement vector */
 149             dx00             = ix0 - jx0;
 150             dy00             = iy0 - jy0;
 151             dz00             = iz0 - jz0;
 152
 153             /* Calculate squared distance and things based on it */
 154             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 155
 156             rinv00           = gmx_invsqrt(rsq00);
 157
 158             /* Load parameters for j particles */
 159             vdwjidx0         = 2*vdwtype[jnr+0];
 160
 161             /**************************
 162              * CALCULATE INTERACTIONS *
 163              **************************/
 164
 165             r00              = rsq00*rinv00;
 166
 167             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 168             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
 169
 170             /* Calculate table index by multiplying r with table scale and truncate to integer */
 171             rt               = r00*vftabscale;
 172             vfitab           = rt;
 173             vfeps            = rt-vfitab;
 174             vfitab           = 2*4*vfitab;
 175
 176             /* CUBIC SPLINE TABLE DISPERSION */
 177             vfitab          += 0;
 178             Y                = vftab[vfitab];
 179             F                = vftab[vfitab+1];
 180             Geps             = vfeps*vftab[vfitab+2];
 181             Heps2            = vfeps*vfeps*vftab[vfitab+3];
 182             Fp               = F+Geps+Heps2;
 183             VV               = Y+vfeps*Fp;
 184             vvdw6            = c6_00*VV;
 185             FF               = Fp+Geps+2.0*Heps2;
 186             fvdw6            = c6_00*FF;
 187
 188             /* CUBIC SPLINE TABLE REPULSION */
 189             Y                = vftab[vfitab+4];
 190             F                = vftab[vfitab+5];
 191             Geps             = vfeps*vftab[vfitab+6];
 192             Heps2            = vfeps*vfeps*vftab[vfitab+7];
 193             Fp               = F+Geps+Heps2;
 194             VV               = Y+vfeps*Fp;
 195             vvdw12           = c12_00*VV;
 196             FF               = Fp+Geps+2.0*Heps2;
 197             fvdw12           = c12_00*FF;
 198             vvdw             = vvdw12+vvdw6;
 199             fvdw             = -(fvdw6+fvdw12)*vftabscale*rinv00;
 200
 201             /* Update potential sums from outer loop */
 202             vvdwsum         += vvdw;
 203
 204             fscal            = fvdw;
 205
 206             /* Calculate temporary vectorial force */
 207             tx               = fscal*dx00;
 208             ty               = fscal*dy00;
 209             tz               = fscal*dz00;
 210
 211             /* Update vectorial force */
 212             fix0            += tx;
 213             fiy0            += ty;
 214             fiz0            += tz;
 215             f[j_coord_offset+DIM*0+XX] -= tx;
 216             f[j_coord_offset+DIM*0+YY] -= ty;
 217             f[j_coord_offset+DIM*0+ZZ] -= tz;
 218
 219             /* Inner loop uses 55 flops */
 220         }
 221         /* End of innermost loop */
 222
 223         tx = ty = tz = 0;
 224         f[i_coord_offset+DIM*0+XX] += fix0;
 225         f[i_coord_offset+DIM*0+YY] += fiy0;
 226         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 227         tx                         += fix0;
 228         ty                         += fiy0;
 229         tz                         += fiz0;
 230         fshift[i_shift_offset+XX]  += tx;
 231         fshift[i_shift_offset+YY]  += ty;
 232         fshift[i_shift_offset+ZZ]  += tz;
 233
 234         ggid                        = gid[iidx];
 235         /* Update potential energies */
 236         kernel_data->energygrp_vdw[ggid] += vvdwsum;
 237
 238         /* Increment number of inner iterations */
 239         inneriter                  += j_index_end - j_index_start;
 240
 241         /* Outer loop uses 13 flops */
 242     }
 243
 244     /* Increment number of outer iterations */
 245     outeriter        += nri;
 246
 247     /* Update outer/inner flops */
 248
 249     inc_nrnb(nrnb,eNR_NBKERNEL_VDW_VF,outeriter*13 + inneriter*55);
 250 }
 251 /*
 252  * Gromacs nonbonded kernel:   nb_kernel_ElecNone_VdwCSTab_GeomP1P1_F_c
 253  * Electrostatics interaction: None
 254  * VdW interaction:            CubicSplineTable
 255  * Geometry:                   Particle-Particle
 256  * Calculate force/pot:        Force
 257  */
 258 void
 259 nb_kernel_ElecNone_VdwCSTab_GeomP1P1_F_c
 260                     (t_nblist                    * gmx_restrict       nlist,
 261                      rvec                        * gmx_restrict          xx,
 262                      rvec                        * gmx_restrict          ff,
 263                      t_forcerec                  * gmx_restrict          fr,
 264                      t_mdatoms                   * gmx_restrict     mdatoms,
 265                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 266                      t_nrnb                      * gmx_restrict        nrnb)
 267 {
 268     int              i_shift_offset,i_coord_offset,j_coord_offset;
 269     int              j_index_start,j_index_end;
 270     int              nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
 271     real             shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
 272     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 273     real             *shiftvec,*fshift,*x,*f;
 274     int              vdwioffset0;
 275     real             ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 276     int              vdwjidx0;
 277     real             jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 278     real             dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
 279     int              nvdwtype;
 280     real             rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
 281     int              *vdwtype;
 282     real             *vdwparam;
 283     int              vfitab;
 284     real             rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF;
 285     real             *vftab;
 286
 287     x                = xx[0];
 288     f                = ff[0];
 289
 290     nri              = nlist->nri;
 291     iinr             = nlist->iinr;
 292     jindex           = nlist->jindex;
 293     jjnr             = nlist->jjnr;
 294     shiftidx         = nlist->shift;
 295     gid              = nlist->gid;
 296     shiftvec         = fr->shift_vec[0];
 297     fshift           = fr->fshift[0];
 298     nvdwtype         = fr->ntype;
 299     vdwparam         = fr->nbfp;
 300     vdwtype          = mdatoms->typeA;
 301
 302     vftab            = kernel_data->table_vdw->data;
 303     vftabscale       = kernel_data->table_vdw->scale;
 304
 305     outeriter        = 0;
 306     inneriter        = 0;
 307
 308     /* Start outer loop over neighborlists */
 309     for(iidx=0; iidx<nri; iidx++)
 310     {
 311         /* Load shift vector for this list */
 312         i_shift_offset   = DIM*shiftidx[iidx];
 313         shX              = shiftvec[i_shift_offset+XX];
 314         shY              = shiftvec[i_shift_offset+YY];
 315         shZ              = shiftvec[i_shift_offset+ZZ];
 316
 317         /* Load limits for loop over neighbors */
 318         j_index_start    = jindex[iidx];
 319         j_index_end      = jindex[iidx+1];
 320
 321         /* Get outer coordinate index */
 322         inr              = iinr[iidx];
 323         i_coord_offset   = DIM*inr;
 324
 325         /* Load i particle coords and add shift vector */
 326         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 327         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 328         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 329
 330         fix0             = 0.0;
 331         fiy0             = 0.0;
 332         fiz0             = 0.0;
 333
 334         /* Load parameters for i particles */
 335         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 336
 337         /* Start inner kernel loop */
 338         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 339         {
 340             /* Get j neighbor index, and coordinate index */
 341             jnr              = jjnr[jidx];
 342             j_coord_offset   = DIM*jnr;
 343
 344             /* load j atom coordinates */
 345             jx0              = x[j_coord_offset+DIM*0+XX];
 346             jy0              = x[j_coord_offset+DIM*0+YY];
 347             jz0              = x[j_coord_offset+DIM*0+ZZ];
 348
 349             /* Calculate displacement vector */
 350             dx00             = ix0 - jx0;
 351             dy00             = iy0 - jy0;
 352             dz00             = iz0 - jz0;
 353
 354             /* Calculate squared distance and things based on it */
 355             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 356
 357             rinv00           = gmx_invsqrt(rsq00);
 358
 359             /* Load parameters for j particles */
 360             vdwjidx0         = 2*vdwtype[jnr+0];
 361
 362             /**************************
 363              * CALCULATE INTERACTIONS *
 364              **************************/
 365
 366             r00              = rsq00*rinv00;
 367
 368             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 369             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
 370
 371             /* Calculate table index by multiplying r with table scale and truncate to integer */
 372             rt               = r00*vftabscale;
 373             vfitab           = rt;
 374             vfeps            = rt-vfitab;
 375             vfitab           = 2*4*vfitab;
 376
 377             /* CUBIC SPLINE TABLE DISPERSION */
 378             vfitab          += 0;
 379             F                = vftab[vfitab+1];
 380             Geps             = vfeps*vftab[vfitab+2];
 381             Heps2            = vfeps*vfeps*vftab[vfitab+3];
 382             Fp               = F+Geps+Heps2;
 383             FF               = Fp+Geps+2.0*Heps2;
 384             fvdw6            = c6_00*FF;
 385
 386             /* CUBIC SPLINE TABLE REPULSION */
 387             F                = vftab[vfitab+5];
 388             Geps             = vfeps*vftab[vfitab+6];
 389             Heps2            = vfeps*vfeps*vftab[vfitab+7];
 390             Fp               = F+Geps+Heps2;
 391             FF               = Fp+Geps+2.0*Heps2;
 392             fvdw12           = c12_00*FF;
 393             fvdw             = -(fvdw6+fvdw12)*vftabscale*rinv00;
 394
 395             fscal            = fvdw;
 396
 397             /* Calculate temporary vectorial force */
 398             tx               = fscal*dx00;
 399             ty               = fscal*dy00;
 400             tz               = fscal*dz00;
 401
 402             /* Update vectorial force */
 403             fix0            += tx;
 404             fiy0            += ty;
 405             fiz0            += tz;
 406             f[j_coord_offset+DIM*0+XX] -= tx;
 407             f[j_coord_offset+DIM*0+YY] -= ty;
 408             f[j_coord_offset+DIM*0+ZZ] -= tz;
 409
 410             /* Inner loop uses 47 flops */
 411         }
 412         /* End of innermost loop */
 413
 414         tx = ty = tz = 0;
 415         f[i_coord_offset+DIM*0+XX] += fix0;
 416         f[i_coord_offset+DIM*0+YY] += fiy0;
 417         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 418         tx                         += fix0;
 419         ty                         += fiy0;
 420         tz                         += fiz0;
 421         fshift[i_shift_offset+XX]  += tx;
 422         fshift[i_shift_offset+YY]  += ty;
 423         fshift[i_shift_offset+ZZ]  += tz;
 424
 425         /* Increment number of inner iterations */
 426         inneriter                  += j_index_end - j_index_start;
 427
 428         /* Outer loop uses 12 flops */
 429     }
 430
 431     /* Increment number of outer iterations */
 432     outeriter        += nri;
 433
 434     /* Update outer/inner flops */
 435
 436     inc_nrnb(nrnb,eNR_NBKERNEL_VDW_F,outeriter*12 + inneriter*47);
 437 }