2 * This file is part of the GROMACS molecular simulation package.
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.
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.
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.
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.
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.
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.
36 * Note: this file was generated by the GROMACS c kernel generator.
42 #include "../nb_kernel.h"
43 #include "gromacs/legacyheaders/types/simple.h"
44 #include "gromacs/math/vec.h"
45 #include "gromacs/legacyheaders/nrnb.h"
48 * Gromacs nonbonded kernel: nb_kernel_ElecEwSh_VdwLJEwSh_GeomP1P1_VF_c
49 * Electrostatics interaction: Ewald
50 * VdW interaction: LJEwald
51 * Geometry: Particle-Particle
52 * Calculate force/pot: PotentialAndForce
55 nb_kernel_ElecEwSh_VdwLJEwSh_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)
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;
71 real ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
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;
78 real rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
82 real ewclj,ewclj2,ewclj6,ewcljrsq,poly,exponent,sh_lj_ewald;
85 real ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace;
93 jindex = nlist->jindex;
95 shiftidx = nlist->shift;
97 shiftvec = fr->shift_vec[0];
98 fshift = fr->fshift[0];
100 charge = mdatoms->chargeA;
101 nvdwtype = fr->ntype;
103 vdwtype = mdatoms->typeA;
104 vdwgridparam = fr->ljpme_c6grid;
105 ewclj = fr->ewaldcoeff_lj;
106 sh_lj_ewald = fr->ic->sh_lj_ewald;
107 ewclj2 = ewclj*ewclj;
108 ewclj6 = ewclj2*ewclj2*ewclj2;
110 sh_ewald = fr->ic->sh_ewald;
111 ewtab = fr->ic->tabq_coul_FDV0;
112 ewtabscale = fr->ic->tabq_scale;
113 ewtabhalfspace = 0.5/ewtabscale;
115 /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
116 rcutoff = fr->rcoulomb;
117 rcutoff2 = rcutoff*rcutoff;
119 sh_vdw_invrcut6 = fr->ic->sh_invrc6;
125 /* Start outer loop over neighborlists */
126 for(iidx=0; iidx<nri; iidx++)
128 /* Load shift vector for this list */
129 i_shift_offset = DIM*shiftidx[iidx];
130 shX = shiftvec[i_shift_offset+XX];
131 shY = shiftvec[i_shift_offset+YY];
132 shZ = shiftvec[i_shift_offset+ZZ];
134 /* Load limits for loop over neighbors */
135 j_index_start = jindex[iidx];
136 j_index_end = jindex[iidx+1];
138 /* Get outer coordinate index */
140 i_coord_offset = DIM*inr;
142 /* Load i particle coords and add shift vector */
143 ix0 = shX + x[i_coord_offset+DIM*0+XX];
144 iy0 = shY + x[i_coord_offset+DIM*0+YY];
145 iz0 = shZ + x[i_coord_offset+DIM*0+ZZ];
151 /* Load parameters for i particles */
152 iq0 = facel*charge[inr+0];
153 vdwioffset0 = 2*nvdwtype*vdwtype[inr+0];
155 /* Reset potential sums */
159 /* Start inner kernel loop */
160 for(jidx=j_index_start; jidx<j_index_end; jidx++)
162 /* Get j neighbor index, and coordinate index */
164 j_coord_offset = DIM*jnr;
166 /* load j atom coordinates */
167 jx0 = x[j_coord_offset+DIM*0+XX];
168 jy0 = x[j_coord_offset+DIM*0+YY];
169 jz0 = x[j_coord_offset+DIM*0+ZZ];
171 /* Calculate displacement vector */
176 /* Calculate squared distance and things based on it */
177 rsq00 = dx00*dx00+dy00*dy00+dz00*dz00;
179 rinv00 = gmx_invsqrt(rsq00);
181 rinvsq00 = rinv00*rinv00;
183 /* Load parameters for j particles */
185 vdwjidx0 = 2*vdwtype[jnr+0];
187 /**************************
188 * CALCULATE INTERACTIONS *
189 **************************/
197 c6_00 = vdwparam[vdwioffset0+vdwjidx0];
198 c12_00 = vdwparam[vdwioffset0+vdwjidx0+1];
199 c6grid_00 = vdwgridparam[vdwioffset0+vdwjidx0];
201 /* EWALD ELECTROSTATICS */
203 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
204 ewrt = r00*ewtabscale;
208 felec = ewtab[ewitab]+eweps*ewtab[ewitab+1];
209 velec = qq00*((rinv00-sh_ewald)-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec)));
210 felec = qq00*rinv00*(rinvsq00-felec);
212 rinvsix = rinvsq00*rinvsq00*rinvsq00;
213 ewcljrsq = ewclj2*rsq00;
214 exponent = exp(-ewcljrsq);
215 poly = exponent*(1.0 + ewcljrsq + ewcljrsq*ewcljrsq*0.5);
216 vvdw6 = (c6_00-c6grid_00*(1.0-poly))*rinvsix;
217 vvdw12 = c12_00*rinvsix*rinvsix;
218 vvdw = (vvdw12 - c12_00*sh_vdw_invrcut6*sh_vdw_invrcut6)*(1.0/12.0) - (vvdw6 - c6_00*sh_vdw_invrcut6 - c6grid_00*sh_lj_ewald)*(1.0/6.0);
219 fvdw = (vvdw12 - vvdw6 - c6grid_00*(1.0/6.0)*exponent*ewclj6)*rinvsq00;
221 /* Update potential sums from outer loop */
227 /* Calculate temporary vectorial force */
232 /* Update vectorial force */
236 f[j_coord_offset+DIM*0+XX] -= tx;
237 f[j_coord_offset+DIM*0+YY] -= ty;
238 f[j_coord_offset+DIM*0+ZZ] -= tz;
242 /* Inner loop uses 74 flops */
244 /* End of innermost loop */
247 f[i_coord_offset+DIM*0+XX] += fix0;
248 f[i_coord_offset+DIM*0+YY] += fiy0;
249 f[i_coord_offset+DIM*0+ZZ] += fiz0;
253 fshift[i_shift_offset+XX] += tx;
254 fshift[i_shift_offset+YY] += ty;
255 fshift[i_shift_offset+ZZ] += tz;
258 /* Update potential energies */
259 kernel_data->energygrp_elec[ggid] += velecsum;
260 kernel_data->energygrp_vdw[ggid] += vvdwsum;
262 /* Increment number of inner iterations */
263 inneriter += j_index_end - j_index_start;
265 /* Outer loop uses 15 flops */
268 /* Increment number of outer iterations */
271 /* Update outer/inner flops */
273 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_VF,outeriter*15 + inneriter*74);
276 * Gromacs nonbonded kernel: nb_kernel_ElecEwSh_VdwLJEwSh_GeomP1P1_F_c
277 * Electrostatics interaction: Ewald
278 * VdW interaction: LJEwald
279 * Geometry: Particle-Particle
280 * Calculate force/pot: Force
283 nb_kernel_ElecEwSh_VdwLJEwSh_GeomP1P1_F_c
284 (t_nblist * gmx_restrict nlist,
285 rvec * gmx_restrict xx,
286 rvec * gmx_restrict ff,
287 t_forcerec * gmx_restrict fr,
288 t_mdatoms * gmx_restrict mdatoms,
289 nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
290 t_nrnb * gmx_restrict nrnb)
292 int i_shift_offset,i_coord_offset,j_coord_offset;
293 int j_index_start,j_index_end;
294 int nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
295 real shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
296 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
297 real *shiftvec,*fshift,*x,*f;
299 real ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
301 real jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
302 real dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
303 real velec,felec,velecsum,facel,crf,krf,krf2;
306 real rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
310 real ewclj,ewclj2,ewclj6,ewcljrsq,poly,exponent,sh_lj_ewald;
313 real ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace;
321 jindex = nlist->jindex;
323 shiftidx = nlist->shift;
325 shiftvec = fr->shift_vec[0];
326 fshift = fr->fshift[0];
328 charge = mdatoms->chargeA;
329 nvdwtype = fr->ntype;
331 vdwtype = mdatoms->typeA;
332 vdwgridparam = fr->ljpme_c6grid;
333 ewclj = fr->ewaldcoeff_lj;
334 sh_lj_ewald = fr->ic->sh_lj_ewald;
335 ewclj2 = ewclj*ewclj;
336 ewclj6 = ewclj2*ewclj2*ewclj2;
338 sh_ewald = fr->ic->sh_ewald;
339 ewtab = fr->ic->tabq_coul_F;
340 ewtabscale = fr->ic->tabq_scale;
341 ewtabhalfspace = 0.5/ewtabscale;
343 /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
344 rcutoff = fr->rcoulomb;
345 rcutoff2 = rcutoff*rcutoff;
347 sh_vdw_invrcut6 = fr->ic->sh_invrc6;
353 /* Start outer loop over neighborlists */
354 for(iidx=0; iidx<nri; iidx++)
356 /* Load shift vector for this list */
357 i_shift_offset = DIM*shiftidx[iidx];
358 shX = shiftvec[i_shift_offset+XX];
359 shY = shiftvec[i_shift_offset+YY];
360 shZ = shiftvec[i_shift_offset+ZZ];
362 /* Load limits for loop over neighbors */
363 j_index_start = jindex[iidx];
364 j_index_end = jindex[iidx+1];
366 /* Get outer coordinate index */
368 i_coord_offset = DIM*inr;
370 /* Load i particle coords and add shift vector */
371 ix0 = shX + x[i_coord_offset+DIM*0+XX];
372 iy0 = shY + x[i_coord_offset+DIM*0+YY];
373 iz0 = shZ + x[i_coord_offset+DIM*0+ZZ];
379 /* Load parameters for i particles */
380 iq0 = facel*charge[inr+0];
381 vdwioffset0 = 2*nvdwtype*vdwtype[inr+0];
383 /* Start inner kernel loop */
384 for(jidx=j_index_start; jidx<j_index_end; jidx++)
386 /* Get j neighbor index, and coordinate index */
388 j_coord_offset = DIM*jnr;
390 /* load j atom coordinates */
391 jx0 = x[j_coord_offset+DIM*0+XX];
392 jy0 = x[j_coord_offset+DIM*0+YY];
393 jz0 = x[j_coord_offset+DIM*0+ZZ];
395 /* Calculate displacement vector */
400 /* Calculate squared distance and things based on it */
401 rsq00 = dx00*dx00+dy00*dy00+dz00*dz00;
403 rinv00 = gmx_invsqrt(rsq00);
405 rinvsq00 = rinv00*rinv00;
407 /* Load parameters for j particles */
409 vdwjidx0 = 2*vdwtype[jnr+0];
411 /**************************
412 * CALCULATE INTERACTIONS *
413 **************************/
421 c6_00 = vdwparam[vdwioffset0+vdwjidx0];
422 c12_00 = vdwparam[vdwioffset0+vdwjidx0+1];
423 c6grid_00 = vdwgridparam[vdwioffset0+vdwjidx0];
425 /* EWALD ELECTROSTATICS */
427 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
428 ewrt = r00*ewtabscale;
431 felec = (1.0-eweps)*ewtab[ewitab]+eweps*ewtab[ewitab+1];
432 felec = qq00*rinv00*(rinvsq00-felec);
434 rinvsix = rinvsq00*rinvsq00*rinvsq00;
435 ewcljrsq = ewclj2*rsq00;
436 exponent = exp(-ewcljrsq);
437 poly = exponent*(1.0 + ewcljrsq + ewcljrsq*ewcljrsq*0.5);
438 fvdw = (((c12_00*rinvsix - c6_00 + c6grid_00*(1.0-poly))*rinvsix) - c6grid_00*(1.0/6.0)*exponent*ewclj6)*rinvsq00;
442 /* Calculate temporary vectorial force */
447 /* Update vectorial force */
451 f[j_coord_offset+DIM*0+XX] -= tx;
452 f[j_coord_offset+DIM*0+YY] -= ty;
453 f[j_coord_offset+DIM*0+ZZ] -= tz;
457 /* Inner loop uses 55 flops */
459 /* End of innermost loop */
462 f[i_coord_offset+DIM*0+XX] += fix0;
463 f[i_coord_offset+DIM*0+YY] += fiy0;
464 f[i_coord_offset+DIM*0+ZZ] += fiz0;
468 fshift[i_shift_offset+XX] += tx;
469 fshift[i_shift_offset+YY] += ty;
470 fshift[i_shift_offset+ZZ] += tz;
472 /* Increment number of inner iterations */
473 inneriter += j_index_end - j_index_start;
475 /* Outer loop uses 13 flops */
478 /* Increment number of outer iterations */
481 /* Update outer/inner flops */
483 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_F,outeriter*13 + inneriter*55);