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_ElecRFCut_VdwCSTab_GeomW4P1_VF_c
49 * Electrostatics interaction: ReactionField
50 * VdW interaction: CubicSplineTable
51 * Geometry: Water4-Particle
52 * Calculate force/pot: PotentialAndForce
55 nb_kernel_ElecRFCut_VdwCSTab_GeomW4P1_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 ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
75 real ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
77 real ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
79 real jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
80 real dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
81 real dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10,cexp1_10,cexp2_10;
82 real dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20,cexp1_20,cexp2_20;
83 real dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30,cexp1_30,cexp2_30;
84 real velec,felec,velecsum,facel,crf,krf,krf2;
87 real rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
91 real rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF;
99 jindex = nlist->jindex;
101 shiftidx = nlist->shift;
103 shiftvec = fr->shift_vec[0];
104 fshift = fr->fshift[0];
106 charge = mdatoms->chargeA;
110 nvdwtype = fr->ntype;
112 vdwtype = mdatoms->typeA;
114 vftab = kernel_data->table_vdw->data;
115 vftabscale = kernel_data->table_vdw->scale;
117 /* Setup water-specific parameters */
118 inr = nlist->iinr[0];
119 iq1 = facel*charge[inr+1];
120 iq2 = facel*charge[inr+2];
121 iq3 = facel*charge[inr+3];
122 vdwioffset0 = 2*nvdwtype*vdwtype[inr+0];
124 /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
125 rcutoff = fr->rcoulomb;
126 rcutoff2 = rcutoff*rcutoff;
131 /* Start outer loop over neighborlists */
132 for(iidx=0; iidx<nri; iidx++)
134 /* Load shift vector for this list */
135 i_shift_offset = DIM*shiftidx[iidx];
136 shX = shiftvec[i_shift_offset+XX];
137 shY = shiftvec[i_shift_offset+YY];
138 shZ = shiftvec[i_shift_offset+ZZ];
140 /* Load limits for loop over neighbors */
141 j_index_start = jindex[iidx];
142 j_index_end = jindex[iidx+1];
144 /* Get outer coordinate index */
146 i_coord_offset = DIM*inr;
148 /* Load i particle coords and add shift vector */
149 ix0 = shX + x[i_coord_offset+DIM*0+XX];
150 iy0 = shY + x[i_coord_offset+DIM*0+YY];
151 iz0 = shZ + x[i_coord_offset+DIM*0+ZZ];
152 ix1 = shX + x[i_coord_offset+DIM*1+XX];
153 iy1 = shY + x[i_coord_offset+DIM*1+YY];
154 iz1 = shZ + x[i_coord_offset+DIM*1+ZZ];
155 ix2 = shX + x[i_coord_offset+DIM*2+XX];
156 iy2 = shY + x[i_coord_offset+DIM*2+YY];
157 iz2 = shZ + x[i_coord_offset+DIM*2+ZZ];
158 ix3 = shX + x[i_coord_offset+DIM*3+XX];
159 iy3 = shY + x[i_coord_offset+DIM*3+YY];
160 iz3 = shZ + x[i_coord_offset+DIM*3+ZZ];
175 /* Reset potential sums */
179 /* Start inner kernel loop */
180 for(jidx=j_index_start; jidx<j_index_end; jidx++)
182 /* Get j neighbor index, and coordinate index */
184 j_coord_offset = DIM*jnr;
186 /* load j atom coordinates */
187 jx0 = x[j_coord_offset+DIM*0+XX];
188 jy0 = x[j_coord_offset+DIM*0+YY];
189 jz0 = x[j_coord_offset+DIM*0+ZZ];
191 /* Calculate displacement vector */
205 /* Calculate squared distance and things based on it */
206 rsq00 = dx00*dx00+dy00*dy00+dz00*dz00;
207 rsq10 = dx10*dx10+dy10*dy10+dz10*dz10;
208 rsq20 = dx20*dx20+dy20*dy20+dz20*dz20;
209 rsq30 = dx30*dx30+dy30*dy30+dz30*dz30;
211 rinv00 = gmx_invsqrt(rsq00);
212 rinv10 = gmx_invsqrt(rsq10);
213 rinv20 = gmx_invsqrt(rsq20);
214 rinv30 = gmx_invsqrt(rsq30);
216 rinvsq10 = rinv10*rinv10;
217 rinvsq20 = rinv20*rinv20;
218 rinvsq30 = rinv30*rinv30;
220 /* Load parameters for j particles */
222 vdwjidx0 = 2*vdwtype[jnr+0];
224 /**************************
225 * CALCULATE INTERACTIONS *
226 **************************/
230 c6_00 = vdwparam[vdwioffset0+vdwjidx0];
231 c12_00 = vdwparam[vdwioffset0+vdwjidx0+1];
233 /* Calculate table index by multiplying r with table scale and truncate to integer */
239 /* CUBIC SPLINE TABLE DISPERSION */
243 Geps = vfeps*vftab[vfitab+2];
244 Heps2 = vfeps*vfeps*vftab[vfitab+3];
248 FF = Fp+Geps+2.0*Heps2;
251 /* CUBIC SPLINE TABLE REPULSION */
254 Geps = vfeps*vftab[vfitab+6];
255 Heps2 = vfeps*vfeps*vftab[vfitab+7];
259 FF = Fp+Geps+2.0*Heps2;
262 fvdw = -(fvdw6+fvdw12)*vftabscale*rinv00;
264 /* Update potential sums from outer loop */
269 /* Calculate temporary vectorial force */
274 /* Update vectorial force */
278 f[j_coord_offset+DIM*0+XX] -= tx;
279 f[j_coord_offset+DIM*0+YY] -= ty;
280 f[j_coord_offset+DIM*0+ZZ] -= tz;
282 /**************************
283 * CALCULATE INTERACTIONS *
284 **************************/
291 /* REACTION-FIELD ELECTROSTATICS */
292 velec = qq10*(rinv10+krf*rsq10-crf);
293 felec = qq10*(rinv10*rinvsq10-krf2);
295 /* Update potential sums from outer loop */
300 /* Calculate temporary vectorial force */
305 /* Update vectorial force */
309 f[j_coord_offset+DIM*0+XX] -= tx;
310 f[j_coord_offset+DIM*0+YY] -= ty;
311 f[j_coord_offset+DIM*0+ZZ] -= tz;
315 /**************************
316 * CALCULATE INTERACTIONS *
317 **************************/
324 /* REACTION-FIELD ELECTROSTATICS */
325 velec = qq20*(rinv20+krf*rsq20-crf);
326 felec = qq20*(rinv20*rinvsq20-krf2);
328 /* Update potential sums from outer loop */
333 /* Calculate temporary vectorial force */
338 /* Update vectorial force */
342 f[j_coord_offset+DIM*0+XX] -= tx;
343 f[j_coord_offset+DIM*0+YY] -= ty;
344 f[j_coord_offset+DIM*0+ZZ] -= tz;
348 /**************************
349 * CALCULATE INTERACTIONS *
350 **************************/
357 /* REACTION-FIELD ELECTROSTATICS */
358 velec = qq30*(rinv30+krf*rsq30-crf);
359 felec = qq30*(rinv30*rinvsq30-krf2);
361 /* Update potential sums from outer loop */
366 /* Calculate temporary vectorial force */
371 /* Update vectorial force */
375 f[j_coord_offset+DIM*0+XX] -= tx;
376 f[j_coord_offset+DIM*0+YY] -= ty;
377 f[j_coord_offset+DIM*0+ZZ] -= tz;
381 /* Inner loop uses 151 flops */
383 /* End of innermost loop */
386 f[i_coord_offset+DIM*0+XX] += fix0;
387 f[i_coord_offset+DIM*0+YY] += fiy0;
388 f[i_coord_offset+DIM*0+ZZ] += fiz0;
392 f[i_coord_offset+DIM*1+XX] += fix1;
393 f[i_coord_offset+DIM*1+YY] += fiy1;
394 f[i_coord_offset+DIM*1+ZZ] += fiz1;
398 f[i_coord_offset+DIM*2+XX] += fix2;
399 f[i_coord_offset+DIM*2+YY] += fiy2;
400 f[i_coord_offset+DIM*2+ZZ] += fiz2;
404 f[i_coord_offset+DIM*3+XX] += fix3;
405 f[i_coord_offset+DIM*3+YY] += fiy3;
406 f[i_coord_offset+DIM*3+ZZ] += fiz3;
410 fshift[i_shift_offset+XX] += tx;
411 fshift[i_shift_offset+YY] += ty;
412 fshift[i_shift_offset+ZZ] += tz;
415 /* Update potential energies */
416 kernel_data->energygrp_elec[ggid] += velecsum;
417 kernel_data->energygrp_vdw[ggid] += vvdwsum;
419 /* Increment number of inner iterations */
420 inneriter += j_index_end - j_index_start;
422 /* Outer loop uses 41 flops */
425 /* Increment number of outer iterations */
428 /* Update outer/inner flops */
430 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4_VF,outeriter*41 + inneriter*151);
433 * Gromacs nonbonded kernel: nb_kernel_ElecRFCut_VdwCSTab_GeomW4P1_F_c
434 * Electrostatics interaction: ReactionField
435 * VdW interaction: CubicSplineTable
436 * Geometry: Water4-Particle
437 * Calculate force/pot: Force
440 nb_kernel_ElecRFCut_VdwCSTab_GeomW4P1_F_c
441 (t_nblist * gmx_restrict nlist,
442 rvec * gmx_restrict xx,
443 rvec * gmx_restrict ff,
444 t_forcerec * gmx_restrict fr,
445 t_mdatoms * gmx_restrict mdatoms,
446 nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
447 t_nrnb * gmx_restrict nrnb)
449 int i_shift_offset,i_coord_offset,j_coord_offset;
450 int j_index_start,j_index_end;
451 int nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
452 real shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
453 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
454 real *shiftvec,*fshift,*x,*f;
456 real ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
458 real ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
460 real ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
462 real ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
464 real jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
465 real dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
466 real dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10,cexp1_10,cexp2_10;
467 real dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20,cexp1_20,cexp2_20;
468 real dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30,cexp1_30,cexp2_30;
469 real velec,felec,velecsum,facel,crf,krf,krf2;
472 real rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
476 real rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF;
484 jindex = nlist->jindex;
486 shiftidx = nlist->shift;
488 shiftvec = fr->shift_vec[0];
489 fshift = fr->fshift[0];
491 charge = mdatoms->chargeA;
495 nvdwtype = fr->ntype;
497 vdwtype = mdatoms->typeA;
499 vftab = kernel_data->table_vdw->data;
500 vftabscale = kernel_data->table_vdw->scale;
502 /* Setup water-specific parameters */
503 inr = nlist->iinr[0];
504 iq1 = facel*charge[inr+1];
505 iq2 = facel*charge[inr+2];
506 iq3 = facel*charge[inr+3];
507 vdwioffset0 = 2*nvdwtype*vdwtype[inr+0];
509 /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
510 rcutoff = fr->rcoulomb;
511 rcutoff2 = rcutoff*rcutoff;
516 /* Start outer loop over neighborlists */
517 for(iidx=0; iidx<nri; iidx++)
519 /* Load shift vector for this list */
520 i_shift_offset = DIM*shiftidx[iidx];
521 shX = shiftvec[i_shift_offset+XX];
522 shY = shiftvec[i_shift_offset+YY];
523 shZ = shiftvec[i_shift_offset+ZZ];
525 /* Load limits for loop over neighbors */
526 j_index_start = jindex[iidx];
527 j_index_end = jindex[iidx+1];
529 /* Get outer coordinate index */
531 i_coord_offset = DIM*inr;
533 /* Load i particle coords and add shift vector */
534 ix0 = shX + x[i_coord_offset+DIM*0+XX];
535 iy0 = shY + x[i_coord_offset+DIM*0+YY];
536 iz0 = shZ + x[i_coord_offset+DIM*0+ZZ];
537 ix1 = shX + x[i_coord_offset+DIM*1+XX];
538 iy1 = shY + x[i_coord_offset+DIM*1+YY];
539 iz1 = shZ + x[i_coord_offset+DIM*1+ZZ];
540 ix2 = shX + x[i_coord_offset+DIM*2+XX];
541 iy2 = shY + x[i_coord_offset+DIM*2+YY];
542 iz2 = shZ + x[i_coord_offset+DIM*2+ZZ];
543 ix3 = shX + x[i_coord_offset+DIM*3+XX];
544 iy3 = shY + x[i_coord_offset+DIM*3+YY];
545 iz3 = shZ + x[i_coord_offset+DIM*3+ZZ];
560 /* Start inner kernel loop */
561 for(jidx=j_index_start; jidx<j_index_end; jidx++)
563 /* Get j neighbor index, and coordinate index */
565 j_coord_offset = DIM*jnr;
567 /* load j atom coordinates */
568 jx0 = x[j_coord_offset+DIM*0+XX];
569 jy0 = x[j_coord_offset+DIM*0+YY];
570 jz0 = x[j_coord_offset+DIM*0+ZZ];
572 /* Calculate displacement vector */
586 /* Calculate squared distance and things based on it */
587 rsq00 = dx00*dx00+dy00*dy00+dz00*dz00;
588 rsq10 = dx10*dx10+dy10*dy10+dz10*dz10;
589 rsq20 = dx20*dx20+dy20*dy20+dz20*dz20;
590 rsq30 = dx30*dx30+dy30*dy30+dz30*dz30;
592 rinv00 = gmx_invsqrt(rsq00);
593 rinv10 = gmx_invsqrt(rsq10);
594 rinv20 = gmx_invsqrt(rsq20);
595 rinv30 = gmx_invsqrt(rsq30);
597 rinvsq10 = rinv10*rinv10;
598 rinvsq20 = rinv20*rinv20;
599 rinvsq30 = rinv30*rinv30;
601 /* Load parameters for j particles */
603 vdwjidx0 = 2*vdwtype[jnr+0];
605 /**************************
606 * CALCULATE INTERACTIONS *
607 **************************/
611 c6_00 = vdwparam[vdwioffset0+vdwjidx0];
612 c12_00 = vdwparam[vdwioffset0+vdwjidx0+1];
614 /* Calculate table index by multiplying r with table scale and truncate to integer */
620 /* CUBIC SPLINE TABLE DISPERSION */
623 Geps = vfeps*vftab[vfitab+2];
624 Heps2 = vfeps*vfeps*vftab[vfitab+3];
626 FF = Fp+Geps+2.0*Heps2;
629 /* CUBIC SPLINE TABLE REPULSION */
631 Geps = vfeps*vftab[vfitab+6];
632 Heps2 = vfeps*vfeps*vftab[vfitab+7];
634 FF = Fp+Geps+2.0*Heps2;
636 fvdw = -(fvdw6+fvdw12)*vftabscale*rinv00;
640 /* Calculate temporary vectorial force */
645 /* Update vectorial force */
649 f[j_coord_offset+DIM*0+XX] -= tx;
650 f[j_coord_offset+DIM*0+YY] -= ty;
651 f[j_coord_offset+DIM*0+ZZ] -= tz;
653 /**************************
654 * CALCULATE INTERACTIONS *
655 **************************/
662 /* REACTION-FIELD ELECTROSTATICS */
663 felec = qq10*(rinv10*rinvsq10-krf2);
667 /* Calculate temporary vectorial force */
672 /* Update vectorial force */
676 f[j_coord_offset+DIM*0+XX] -= tx;
677 f[j_coord_offset+DIM*0+YY] -= ty;
678 f[j_coord_offset+DIM*0+ZZ] -= tz;
682 /**************************
683 * CALCULATE INTERACTIONS *
684 **************************/
691 /* REACTION-FIELD ELECTROSTATICS */
692 felec = qq20*(rinv20*rinvsq20-krf2);
696 /* Calculate temporary vectorial force */
701 /* Update vectorial force */
705 f[j_coord_offset+DIM*0+XX] -= tx;
706 f[j_coord_offset+DIM*0+YY] -= ty;
707 f[j_coord_offset+DIM*0+ZZ] -= tz;
711 /**************************
712 * CALCULATE INTERACTIONS *
713 **************************/
720 /* REACTION-FIELD ELECTROSTATICS */
721 felec = qq30*(rinv30*rinvsq30-krf2);
725 /* Calculate temporary vectorial force */
730 /* Update vectorial force */
734 f[j_coord_offset+DIM*0+XX] -= tx;
735 f[j_coord_offset+DIM*0+YY] -= ty;
736 f[j_coord_offset+DIM*0+ZZ] -= tz;
740 /* Inner loop uses 128 flops */
742 /* End of innermost loop */
745 f[i_coord_offset+DIM*0+XX] += fix0;
746 f[i_coord_offset+DIM*0+YY] += fiy0;
747 f[i_coord_offset+DIM*0+ZZ] += fiz0;
751 f[i_coord_offset+DIM*1+XX] += fix1;
752 f[i_coord_offset+DIM*1+YY] += fiy1;
753 f[i_coord_offset+DIM*1+ZZ] += fiz1;
757 f[i_coord_offset+DIM*2+XX] += fix2;
758 f[i_coord_offset+DIM*2+YY] += fiy2;
759 f[i_coord_offset+DIM*2+ZZ] += fiz2;
763 f[i_coord_offset+DIM*3+XX] += fix3;
764 f[i_coord_offset+DIM*3+YY] += fiy3;
765 f[i_coord_offset+DIM*3+ZZ] += fiz3;
769 fshift[i_shift_offset+XX] += tx;
770 fshift[i_shift_offset+YY] += ty;
771 fshift[i_shift_offset+ZZ] += tz;
773 /* Increment number of inner iterations */
774 inneriter += j_index_end - j_index_start;
776 /* Outer loop uses 39 flops */
779 /* Increment number of outer iterations */
782 /* Update outer/inner flops */
784 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4_F,outeriter*39 + inneriter*128);