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 "types/simple.h"
44 #include "gromacs/math/vec.h"
48 * Gromacs nonbonded kernel: nb_kernel_ElecEw_VdwCSTab_GeomW4P1_VF_c
49 * Electrostatics interaction: Ewald
50 * VdW interaction: CubicSplineTable
51 * Geometry: Water4-Particle
52 * Calculate force/pot: PotentialAndForce
55 nb_kernel_ElecEw_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;
94 real ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace;
102 jindex = nlist->jindex;
104 shiftidx = nlist->shift;
106 shiftvec = fr->shift_vec[0];
107 fshift = fr->fshift[0];
109 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 sh_ewald = fr->ic->sh_ewald;
118 ewtab = fr->ic->tabq_coul_FDV0;
119 ewtabscale = fr->ic->tabq_scale;
120 ewtabhalfspace = 0.5/ewtabscale;
122 /* Setup water-specific parameters */
123 inr = nlist->iinr[0];
124 iq1 = facel*charge[inr+1];
125 iq2 = facel*charge[inr+2];
126 iq3 = facel*charge[inr+3];
127 vdwioffset0 = 2*nvdwtype*vdwtype[inr+0];
132 /* Start outer loop over neighborlists */
133 for(iidx=0; iidx<nri; iidx++)
135 /* Load shift vector for this list */
136 i_shift_offset = DIM*shiftidx[iidx];
137 shX = shiftvec[i_shift_offset+XX];
138 shY = shiftvec[i_shift_offset+YY];
139 shZ = shiftvec[i_shift_offset+ZZ];
141 /* Load limits for loop over neighbors */
142 j_index_start = jindex[iidx];
143 j_index_end = jindex[iidx+1];
145 /* Get outer coordinate index */
147 i_coord_offset = DIM*inr;
149 /* Load i particle coords and add shift vector */
150 ix0 = shX + x[i_coord_offset+DIM*0+XX];
151 iy0 = shY + x[i_coord_offset+DIM*0+YY];
152 iz0 = shZ + x[i_coord_offset+DIM*0+ZZ];
153 ix1 = shX + x[i_coord_offset+DIM*1+XX];
154 iy1 = shY + x[i_coord_offset+DIM*1+YY];
155 iz1 = shZ + x[i_coord_offset+DIM*1+ZZ];
156 ix2 = shX + x[i_coord_offset+DIM*2+XX];
157 iy2 = shY + x[i_coord_offset+DIM*2+YY];
158 iz2 = shZ + x[i_coord_offset+DIM*2+ZZ];
159 ix3 = shX + x[i_coord_offset+DIM*3+XX];
160 iy3 = shY + x[i_coord_offset+DIM*3+YY];
161 iz3 = shZ + x[i_coord_offset+DIM*3+ZZ];
176 /* Reset potential sums */
180 /* Start inner kernel loop */
181 for(jidx=j_index_start; jidx<j_index_end; jidx++)
183 /* Get j neighbor index, and coordinate index */
185 j_coord_offset = DIM*jnr;
187 /* load j atom coordinates */
188 jx0 = x[j_coord_offset+DIM*0+XX];
189 jy0 = x[j_coord_offset+DIM*0+YY];
190 jz0 = x[j_coord_offset+DIM*0+ZZ];
192 /* Calculate displacement vector */
206 /* Calculate squared distance and things based on it */
207 rsq00 = dx00*dx00+dy00*dy00+dz00*dz00;
208 rsq10 = dx10*dx10+dy10*dy10+dz10*dz10;
209 rsq20 = dx20*dx20+dy20*dy20+dz20*dz20;
210 rsq30 = dx30*dx30+dy30*dy30+dz30*dz30;
212 rinv00 = gmx_invsqrt(rsq00);
213 rinv10 = gmx_invsqrt(rsq10);
214 rinv20 = gmx_invsqrt(rsq20);
215 rinv30 = gmx_invsqrt(rsq30);
217 rinvsq10 = rinv10*rinv10;
218 rinvsq20 = rinv20*rinv20;
219 rinvsq30 = rinv30*rinv30;
221 /* Load parameters for j particles */
223 vdwjidx0 = 2*vdwtype[jnr+0];
225 /**************************
226 * CALCULATE INTERACTIONS *
227 **************************/
231 c6_00 = vdwparam[vdwioffset0+vdwjidx0];
232 c12_00 = vdwparam[vdwioffset0+vdwjidx0+1];
234 /* Calculate table index by multiplying r with table scale and truncate to integer */
240 /* CUBIC SPLINE TABLE DISPERSION */
244 Geps = vfeps*vftab[vfitab+2];
245 Heps2 = vfeps*vfeps*vftab[vfitab+3];
249 FF = Fp+Geps+2.0*Heps2;
252 /* CUBIC SPLINE TABLE REPULSION */
255 Geps = vfeps*vftab[vfitab+6];
256 Heps2 = vfeps*vfeps*vftab[vfitab+7];
260 FF = Fp+Geps+2.0*Heps2;
263 fvdw = -(fvdw6+fvdw12)*vftabscale*rinv00;
265 /* Update potential sums from outer loop */
270 /* Calculate temporary vectorial force */
275 /* Update vectorial force */
279 f[j_coord_offset+DIM*0+XX] -= tx;
280 f[j_coord_offset+DIM*0+YY] -= ty;
281 f[j_coord_offset+DIM*0+ZZ] -= tz;
283 /**************************
284 * CALCULATE INTERACTIONS *
285 **************************/
291 /* EWALD ELECTROSTATICS */
293 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
294 ewrt = r10*ewtabscale;
298 felec = ewtab[ewitab]+eweps*ewtab[ewitab+1];
299 velec = qq10*(rinv10-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec)));
300 felec = qq10*rinv10*(rinvsq10-felec);
302 /* Update potential sums from outer loop */
307 /* Calculate temporary vectorial force */
312 /* Update vectorial force */
316 f[j_coord_offset+DIM*0+XX] -= tx;
317 f[j_coord_offset+DIM*0+YY] -= ty;
318 f[j_coord_offset+DIM*0+ZZ] -= tz;
320 /**************************
321 * CALCULATE INTERACTIONS *
322 **************************/
328 /* EWALD ELECTROSTATICS */
330 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
331 ewrt = r20*ewtabscale;
335 felec = ewtab[ewitab]+eweps*ewtab[ewitab+1];
336 velec = qq20*(rinv20-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec)));
337 felec = qq20*rinv20*(rinvsq20-felec);
339 /* Update potential sums from outer loop */
344 /* Calculate temporary vectorial force */
349 /* Update vectorial force */
353 f[j_coord_offset+DIM*0+XX] -= tx;
354 f[j_coord_offset+DIM*0+YY] -= ty;
355 f[j_coord_offset+DIM*0+ZZ] -= tz;
357 /**************************
358 * CALCULATE INTERACTIONS *
359 **************************/
365 /* EWALD ELECTROSTATICS */
367 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
368 ewrt = r30*ewtabscale;
372 felec = ewtab[ewitab]+eweps*ewtab[ewitab+1];
373 velec = qq30*(rinv30-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec)));
374 felec = qq30*rinv30*(rinvsq30-felec);
376 /* Update potential sums from outer loop */
381 /* Calculate temporary vectorial force */
386 /* Update vectorial force */
390 f[j_coord_offset+DIM*0+XX] -= tx;
391 f[j_coord_offset+DIM*0+YY] -= ty;
392 f[j_coord_offset+DIM*0+ZZ] -= tz;
394 /* Inner loop uses 178 flops */
396 /* End of innermost loop */
399 f[i_coord_offset+DIM*0+XX] += fix0;
400 f[i_coord_offset+DIM*0+YY] += fiy0;
401 f[i_coord_offset+DIM*0+ZZ] += fiz0;
405 f[i_coord_offset+DIM*1+XX] += fix1;
406 f[i_coord_offset+DIM*1+YY] += fiy1;
407 f[i_coord_offset+DIM*1+ZZ] += fiz1;
411 f[i_coord_offset+DIM*2+XX] += fix2;
412 f[i_coord_offset+DIM*2+YY] += fiy2;
413 f[i_coord_offset+DIM*2+ZZ] += fiz2;
417 f[i_coord_offset+DIM*3+XX] += fix3;
418 f[i_coord_offset+DIM*3+YY] += fiy3;
419 f[i_coord_offset+DIM*3+ZZ] += fiz3;
423 fshift[i_shift_offset+XX] += tx;
424 fshift[i_shift_offset+YY] += ty;
425 fshift[i_shift_offset+ZZ] += tz;
428 /* Update potential energies */
429 kernel_data->energygrp_elec[ggid] += velecsum;
430 kernel_data->energygrp_vdw[ggid] += vvdwsum;
432 /* Increment number of inner iterations */
433 inneriter += j_index_end - j_index_start;
435 /* Outer loop uses 41 flops */
438 /* Increment number of outer iterations */
441 /* Update outer/inner flops */
443 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4_VF,outeriter*41 + inneriter*178);
446 * Gromacs nonbonded kernel: nb_kernel_ElecEw_VdwCSTab_GeomW4P1_F_c
447 * Electrostatics interaction: Ewald
448 * VdW interaction: CubicSplineTable
449 * Geometry: Water4-Particle
450 * Calculate force/pot: Force
453 nb_kernel_ElecEw_VdwCSTab_GeomW4P1_F_c
454 (t_nblist * gmx_restrict nlist,
455 rvec * gmx_restrict xx,
456 rvec * gmx_restrict ff,
457 t_forcerec * gmx_restrict fr,
458 t_mdatoms * gmx_restrict mdatoms,
459 nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
460 t_nrnb * gmx_restrict nrnb)
462 int i_shift_offset,i_coord_offset,j_coord_offset;
463 int j_index_start,j_index_end;
464 int nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
465 real shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
466 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
467 real *shiftvec,*fshift,*x,*f;
469 real ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
471 real ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
473 real ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
475 real ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
477 real jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
478 real dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
479 real dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10,cexp1_10,cexp2_10;
480 real dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20,cexp1_20,cexp2_20;
481 real dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30,cexp1_30,cexp2_30;
482 real velec,felec,velecsum,facel,crf,krf,krf2;
485 real rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
489 real rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF;
492 real ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace;
500 jindex = nlist->jindex;
502 shiftidx = nlist->shift;
504 shiftvec = fr->shift_vec[0];
505 fshift = fr->fshift[0];
507 charge = mdatoms->chargeA;
508 nvdwtype = fr->ntype;
510 vdwtype = mdatoms->typeA;
512 vftab = kernel_data->table_vdw->data;
513 vftabscale = kernel_data->table_vdw->scale;
515 sh_ewald = fr->ic->sh_ewald;
516 ewtab = fr->ic->tabq_coul_F;
517 ewtabscale = fr->ic->tabq_scale;
518 ewtabhalfspace = 0.5/ewtabscale;
520 /* Setup water-specific parameters */
521 inr = nlist->iinr[0];
522 iq1 = facel*charge[inr+1];
523 iq2 = facel*charge[inr+2];
524 iq3 = facel*charge[inr+3];
525 vdwioffset0 = 2*nvdwtype*vdwtype[inr+0];
530 /* Start outer loop over neighborlists */
531 for(iidx=0; iidx<nri; iidx++)
533 /* Load shift vector for this list */
534 i_shift_offset = DIM*shiftidx[iidx];
535 shX = shiftvec[i_shift_offset+XX];
536 shY = shiftvec[i_shift_offset+YY];
537 shZ = shiftvec[i_shift_offset+ZZ];
539 /* Load limits for loop over neighbors */
540 j_index_start = jindex[iidx];
541 j_index_end = jindex[iidx+1];
543 /* Get outer coordinate index */
545 i_coord_offset = DIM*inr;
547 /* Load i particle coords and add shift vector */
548 ix0 = shX + x[i_coord_offset+DIM*0+XX];
549 iy0 = shY + x[i_coord_offset+DIM*0+YY];
550 iz0 = shZ + x[i_coord_offset+DIM*0+ZZ];
551 ix1 = shX + x[i_coord_offset+DIM*1+XX];
552 iy1 = shY + x[i_coord_offset+DIM*1+YY];
553 iz1 = shZ + x[i_coord_offset+DIM*1+ZZ];
554 ix2 = shX + x[i_coord_offset+DIM*2+XX];
555 iy2 = shY + x[i_coord_offset+DIM*2+YY];
556 iz2 = shZ + x[i_coord_offset+DIM*2+ZZ];
557 ix3 = shX + x[i_coord_offset+DIM*3+XX];
558 iy3 = shY + x[i_coord_offset+DIM*3+YY];
559 iz3 = shZ + x[i_coord_offset+DIM*3+ZZ];
574 /* Start inner kernel loop */
575 for(jidx=j_index_start; jidx<j_index_end; jidx++)
577 /* Get j neighbor index, and coordinate index */
579 j_coord_offset = DIM*jnr;
581 /* load j atom coordinates */
582 jx0 = x[j_coord_offset+DIM*0+XX];
583 jy0 = x[j_coord_offset+DIM*0+YY];
584 jz0 = x[j_coord_offset+DIM*0+ZZ];
586 /* Calculate displacement vector */
600 /* Calculate squared distance and things based on it */
601 rsq00 = dx00*dx00+dy00*dy00+dz00*dz00;
602 rsq10 = dx10*dx10+dy10*dy10+dz10*dz10;
603 rsq20 = dx20*dx20+dy20*dy20+dz20*dz20;
604 rsq30 = dx30*dx30+dy30*dy30+dz30*dz30;
606 rinv00 = gmx_invsqrt(rsq00);
607 rinv10 = gmx_invsqrt(rsq10);
608 rinv20 = gmx_invsqrt(rsq20);
609 rinv30 = gmx_invsqrt(rsq30);
611 rinvsq10 = rinv10*rinv10;
612 rinvsq20 = rinv20*rinv20;
613 rinvsq30 = rinv30*rinv30;
615 /* Load parameters for j particles */
617 vdwjidx0 = 2*vdwtype[jnr+0];
619 /**************************
620 * CALCULATE INTERACTIONS *
621 **************************/
625 c6_00 = vdwparam[vdwioffset0+vdwjidx0];
626 c12_00 = vdwparam[vdwioffset0+vdwjidx0+1];
628 /* Calculate table index by multiplying r with table scale and truncate to integer */
634 /* CUBIC SPLINE TABLE DISPERSION */
637 Geps = vfeps*vftab[vfitab+2];
638 Heps2 = vfeps*vfeps*vftab[vfitab+3];
640 FF = Fp+Geps+2.0*Heps2;
643 /* CUBIC SPLINE TABLE REPULSION */
645 Geps = vfeps*vftab[vfitab+6];
646 Heps2 = vfeps*vfeps*vftab[vfitab+7];
648 FF = Fp+Geps+2.0*Heps2;
650 fvdw = -(fvdw6+fvdw12)*vftabscale*rinv00;
654 /* Calculate temporary vectorial force */
659 /* Update vectorial force */
663 f[j_coord_offset+DIM*0+XX] -= tx;
664 f[j_coord_offset+DIM*0+YY] -= ty;
665 f[j_coord_offset+DIM*0+ZZ] -= tz;
667 /**************************
668 * CALCULATE INTERACTIONS *
669 **************************/
675 /* EWALD ELECTROSTATICS */
677 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
678 ewrt = r10*ewtabscale;
681 felec = (1.0-eweps)*ewtab[ewitab]+eweps*ewtab[ewitab+1];
682 felec = qq10*rinv10*(rinvsq10-felec);
686 /* Calculate temporary vectorial force */
691 /* Update vectorial force */
695 f[j_coord_offset+DIM*0+XX] -= tx;
696 f[j_coord_offset+DIM*0+YY] -= ty;
697 f[j_coord_offset+DIM*0+ZZ] -= tz;
699 /**************************
700 * CALCULATE INTERACTIONS *
701 **************************/
707 /* EWALD ELECTROSTATICS */
709 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
710 ewrt = r20*ewtabscale;
713 felec = (1.0-eweps)*ewtab[ewitab]+eweps*ewtab[ewitab+1];
714 felec = qq20*rinv20*(rinvsq20-felec);
718 /* Calculate temporary vectorial force */
723 /* Update vectorial force */
727 f[j_coord_offset+DIM*0+XX] -= tx;
728 f[j_coord_offset+DIM*0+YY] -= ty;
729 f[j_coord_offset+DIM*0+ZZ] -= tz;
731 /**************************
732 * CALCULATE INTERACTIONS *
733 **************************/
739 /* EWALD ELECTROSTATICS */
741 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
742 ewrt = r30*ewtabscale;
745 felec = (1.0-eweps)*ewtab[ewitab]+eweps*ewtab[ewitab+1];
746 felec = qq30*rinv30*(rinvsq30-felec);
750 /* Calculate temporary vectorial force */
755 /* Update vectorial force */
759 f[j_coord_offset+DIM*0+XX] -= tx;
760 f[j_coord_offset+DIM*0+YY] -= ty;
761 f[j_coord_offset+DIM*0+ZZ] -= tz;
763 /* Inner loop uses 149 flops */
765 /* End of innermost loop */
768 f[i_coord_offset+DIM*0+XX] += fix0;
769 f[i_coord_offset+DIM*0+YY] += fiy0;
770 f[i_coord_offset+DIM*0+ZZ] += fiz0;
774 f[i_coord_offset+DIM*1+XX] += fix1;
775 f[i_coord_offset+DIM*1+YY] += fiy1;
776 f[i_coord_offset+DIM*1+ZZ] += fiz1;
780 f[i_coord_offset+DIM*2+XX] += fix2;
781 f[i_coord_offset+DIM*2+YY] += fiy2;
782 f[i_coord_offset+DIM*2+ZZ] += fiz2;
786 f[i_coord_offset+DIM*3+XX] += fix3;
787 f[i_coord_offset+DIM*3+YY] += fiy3;
788 f[i_coord_offset+DIM*3+ZZ] += fiz3;
792 fshift[i_shift_offset+XX] += tx;
793 fshift[i_shift_offset+YY] += ty;
794 fshift[i_shift_offset+ZZ] += tz;
796 /* Increment number of inner iterations */
797 inneriter += j_index_end - j_index_start;
799 /* Outer loop uses 39 flops */
802 /* Increment number of outer iterations */
805 /* Update outer/inner flops */
807 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4_F,outeriter*39 + inneriter*149);