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.
44 #include "../nb_kernel.h"
45 #include "gromacs/legacyheaders/types/simple.h"
46 #include "gromacs/math/vec.h"
47 #include "gromacs/legacyheaders/nrnb.h"
50 * Gromacs nonbonded kernel: nb_kernel_ElecGB_VdwLJ_GeomP1P1_VF_c
51 * Electrostatics interaction: GeneralizedBorn
52 * VdW interaction: LennardJones
53 * Geometry: Particle-Particle
54 * Calculate force/pot: PotentialAndForce
57 nb_kernel_ElecGB_VdwLJ_GeomP1P1_VF_c
58 (t_nblist * gmx_restrict nlist,
59 rvec * gmx_restrict xx,
60 rvec * gmx_restrict ff,
61 t_forcerec * gmx_restrict fr,
62 t_mdatoms * gmx_restrict mdatoms,
63 nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
64 t_nrnb * gmx_restrict nrnb)
66 int i_shift_offset,i_coord_offset,j_coord_offset;
67 int j_index_start,j_index_end;
68 int nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
69 real shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
70 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
71 real *shiftvec,*fshift,*x,*f;
73 real ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
75 real jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
76 real dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
77 real velec,felec,velecsum,facel,crf,krf,krf2;
80 real vgb,fgb,vgbsum,dvdasum,gbscale,gbtabscale,isaprod,gbqqfactor,gbinvepsdiff,dvdaj,gbeps,dvdatmp;
81 real *invsqrta,*dvda,*gbtab;
83 real rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
87 real rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF;
95 jindex = nlist->jindex;
97 shiftidx = nlist->shift;
99 shiftvec = fr->shift_vec[0];
100 fshift = fr->fshift[0];
102 charge = mdatoms->chargeA;
103 nvdwtype = fr->ntype;
105 vdwtype = mdatoms->typeA;
107 invsqrta = fr->invsqrta;
109 gbtabscale = fr->gbtab.scale;
110 gbtab = fr->gbtab.data;
111 gbinvepsdiff = (1.0/fr->epsilon_r) - (1.0/fr->gb_epsilon_solvent);
116 /* Start outer loop over neighborlists */
117 for(iidx=0; iidx<nri; iidx++)
119 /* Load shift vector for this list */
120 i_shift_offset = DIM*shiftidx[iidx];
121 shX = shiftvec[i_shift_offset+XX];
122 shY = shiftvec[i_shift_offset+YY];
123 shZ = shiftvec[i_shift_offset+ZZ];
125 /* Load limits for loop over neighbors */
126 j_index_start = jindex[iidx];
127 j_index_end = jindex[iidx+1];
129 /* Get outer coordinate index */
131 i_coord_offset = DIM*inr;
133 /* Load i particle coords and add shift vector */
134 ix0 = shX + x[i_coord_offset+DIM*0+XX];
135 iy0 = shY + x[i_coord_offset+DIM*0+YY];
136 iz0 = shZ + x[i_coord_offset+DIM*0+ZZ];
142 /* Load parameters for i particles */
143 iq0 = facel*charge[inr+0];
144 isai0 = invsqrta[inr+0];
145 vdwioffset0 = 2*nvdwtype*vdwtype[inr+0];
147 /* Reset potential sums */
153 /* Start inner kernel loop */
154 for(jidx=j_index_start; jidx<j_index_end; jidx++)
156 /* Get j neighbor index, and coordinate index */
158 j_coord_offset = DIM*jnr;
160 /* load j atom coordinates */
161 jx0 = x[j_coord_offset+DIM*0+XX];
162 jy0 = x[j_coord_offset+DIM*0+YY];
163 jz0 = x[j_coord_offset+DIM*0+ZZ];
165 /* Calculate displacement vector */
170 /* Calculate squared distance and things based on it */
171 rsq00 = dx00*dx00+dy00*dy00+dz00*dz00;
173 rinv00 = gmx_invsqrt(rsq00);
175 rinvsq00 = rinv00*rinv00;
177 /* Load parameters for j particles */
179 isaj0 = invsqrta[jnr+0];
180 vdwjidx0 = 2*vdwtype[jnr+0];
182 /**************************
183 * CALCULATE INTERACTIONS *
184 **************************/
189 c6_00 = vdwparam[vdwioffset0+vdwjidx0];
190 c12_00 = vdwparam[vdwioffset0+vdwjidx0+1];
192 /* GENERALIZED BORN AND COULOMB ELECTROSTATICS */
193 isaprod = isai0*isaj0;
194 gbqqfactor = isaprod*(-qq00)*gbinvepsdiff;
195 gbscale = isaprod*gbtabscale;
198 /* Calculate generalized born table index - this is a separate table from the normal one,
199 * but we use the same procedure by multiplying r with scale and truncating to integer.
208 Geps = gbeps*gbtab[gbitab+2];
209 Heps2 = gbeps*gbeps*gbtab[gbitab+3];
214 FF = Fp+Geps+2.0*Heps2;
215 fgb = gbqqfactor*FF*gbscale;
216 dvdatmp = -0.5*(vgb+fgb*r00);
217 dvdasum = dvdasum + dvdatmp;
218 dvda[jnr] = dvdaj+dvdatmp*isaj0*isaj0;
220 felec = (velec*rinv00-fgb)*rinv00;
222 /* LENNARD-JONES DISPERSION/REPULSION */
224 rinvsix = rinvsq00*rinvsq00*rinvsq00;
225 vvdw6 = c6_00*rinvsix;
226 vvdw12 = c12_00*rinvsix*rinvsix;
227 vvdw = vvdw12*(1.0/12.0) - vvdw6*(1.0/6.0);
228 fvdw = (vvdw12-vvdw6)*rinvsq00;
230 /* Update potential sums from outer loop */
237 /* Calculate temporary vectorial force */
242 /* Update vectorial force */
246 f[j_coord_offset+DIM*0+XX] -= tx;
247 f[j_coord_offset+DIM*0+YY] -= ty;
248 f[j_coord_offset+DIM*0+ZZ] -= tz;
250 /* Inner loop uses 71 flops */
252 /* End of innermost loop */
255 f[i_coord_offset+DIM*0+XX] += fix0;
256 f[i_coord_offset+DIM*0+YY] += fiy0;
257 f[i_coord_offset+DIM*0+ZZ] += fiz0;
261 fshift[i_shift_offset+XX] += tx;
262 fshift[i_shift_offset+YY] += ty;
263 fshift[i_shift_offset+ZZ] += tz;
266 /* Update potential energies */
267 kernel_data->energygrp_elec[ggid] += velecsum;
268 kernel_data->energygrp_polarization[ggid] += vgbsum;
269 kernel_data->energygrp_vdw[ggid] += vvdwsum;
270 dvda[inr] = dvda[inr] + dvdasum*isai0*isai0;
272 /* Increment number of inner iterations */
273 inneriter += j_index_end - j_index_start;
275 /* Outer loop uses 16 flops */
278 /* Increment number of outer iterations */
281 /* Update outer/inner flops */
283 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_VF,outeriter*16 + inneriter*71);
286 * Gromacs nonbonded kernel: nb_kernel_ElecGB_VdwLJ_GeomP1P1_F_c
287 * Electrostatics interaction: GeneralizedBorn
288 * VdW interaction: LennardJones
289 * Geometry: Particle-Particle
290 * Calculate force/pot: Force
293 nb_kernel_ElecGB_VdwLJ_GeomP1P1_F_c
294 (t_nblist * gmx_restrict nlist,
295 rvec * gmx_restrict xx,
296 rvec * gmx_restrict ff,
297 t_forcerec * gmx_restrict fr,
298 t_mdatoms * gmx_restrict mdatoms,
299 nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
300 t_nrnb * gmx_restrict nrnb)
302 int i_shift_offset,i_coord_offset,j_coord_offset;
303 int j_index_start,j_index_end;
304 int nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
305 real shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
306 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
307 real *shiftvec,*fshift,*x,*f;
309 real ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
311 real jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
312 real dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
313 real velec,felec,velecsum,facel,crf,krf,krf2;
316 real vgb,fgb,vgbsum,dvdasum,gbscale,gbtabscale,isaprod,gbqqfactor,gbinvepsdiff,dvdaj,gbeps,dvdatmp;
317 real *invsqrta,*dvda,*gbtab;
319 real rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
323 real rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF;
331 jindex = nlist->jindex;
333 shiftidx = nlist->shift;
335 shiftvec = fr->shift_vec[0];
336 fshift = fr->fshift[0];
338 charge = mdatoms->chargeA;
339 nvdwtype = fr->ntype;
341 vdwtype = mdatoms->typeA;
343 invsqrta = fr->invsqrta;
345 gbtabscale = fr->gbtab.scale;
346 gbtab = fr->gbtab.data;
347 gbinvepsdiff = (1.0/fr->epsilon_r) - (1.0/fr->gb_epsilon_solvent);
352 /* Start outer loop over neighborlists */
353 for(iidx=0; iidx<nri; iidx++)
355 /* Load shift vector for this list */
356 i_shift_offset = DIM*shiftidx[iidx];
357 shX = shiftvec[i_shift_offset+XX];
358 shY = shiftvec[i_shift_offset+YY];
359 shZ = shiftvec[i_shift_offset+ZZ];
361 /* Load limits for loop over neighbors */
362 j_index_start = jindex[iidx];
363 j_index_end = jindex[iidx+1];
365 /* Get outer coordinate index */
367 i_coord_offset = DIM*inr;
369 /* Load i particle coords and add shift vector */
370 ix0 = shX + x[i_coord_offset+DIM*0+XX];
371 iy0 = shY + x[i_coord_offset+DIM*0+YY];
372 iz0 = shZ + x[i_coord_offset+DIM*0+ZZ];
378 /* Load parameters for i particles */
379 iq0 = facel*charge[inr+0];
380 isai0 = invsqrta[inr+0];
381 vdwioffset0 = 2*nvdwtype*vdwtype[inr+0];
385 /* Start inner kernel loop */
386 for(jidx=j_index_start; jidx<j_index_end; jidx++)
388 /* Get j neighbor index, and coordinate index */
390 j_coord_offset = DIM*jnr;
392 /* load j atom coordinates */
393 jx0 = x[j_coord_offset+DIM*0+XX];
394 jy0 = x[j_coord_offset+DIM*0+YY];
395 jz0 = x[j_coord_offset+DIM*0+ZZ];
397 /* Calculate displacement vector */
402 /* Calculate squared distance and things based on it */
403 rsq00 = dx00*dx00+dy00*dy00+dz00*dz00;
405 rinv00 = gmx_invsqrt(rsq00);
407 rinvsq00 = rinv00*rinv00;
409 /* Load parameters for j particles */
411 isaj0 = invsqrta[jnr+0];
412 vdwjidx0 = 2*vdwtype[jnr+0];
414 /**************************
415 * CALCULATE INTERACTIONS *
416 **************************/
421 c6_00 = vdwparam[vdwioffset0+vdwjidx0];
422 c12_00 = vdwparam[vdwioffset0+vdwjidx0+1];
424 /* GENERALIZED BORN AND COULOMB ELECTROSTATICS */
425 isaprod = isai0*isaj0;
426 gbqqfactor = isaprod*(-qq00)*gbinvepsdiff;
427 gbscale = isaprod*gbtabscale;
430 /* Calculate generalized born table index - this is a separate table from the normal one,
431 * but we use the same procedure by multiplying r with scale and truncating to integer.
440 Geps = gbeps*gbtab[gbitab+2];
441 Heps2 = gbeps*gbeps*gbtab[gbitab+3];
446 FF = Fp+Geps+2.0*Heps2;
447 fgb = gbqqfactor*FF*gbscale;
448 dvdatmp = -0.5*(vgb+fgb*r00);
449 dvdasum = dvdasum + dvdatmp;
450 dvda[jnr] = dvdaj+dvdatmp*isaj0*isaj0;
452 felec = (velec*rinv00-fgb)*rinv00;
454 /* LENNARD-JONES DISPERSION/REPULSION */
456 rinvsix = rinvsq00*rinvsq00*rinvsq00;
457 fvdw = (c12_00*rinvsix-c6_00)*rinvsix*rinvsq00;
461 /* Calculate temporary vectorial force */
466 /* Update vectorial force */
470 f[j_coord_offset+DIM*0+XX] -= tx;
471 f[j_coord_offset+DIM*0+YY] -= ty;
472 f[j_coord_offset+DIM*0+ZZ] -= tz;
474 /* Inner loop uses 64 flops */
476 /* End of innermost loop */
479 f[i_coord_offset+DIM*0+XX] += fix0;
480 f[i_coord_offset+DIM*0+YY] += fiy0;
481 f[i_coord_offset+DIM*0+ZZ] += fiz0;
485 fshift[i_shift_offset+XX] += tx;
486 fshift[i_shift_offset+YY] += ty;
487 fshift[i_shift_offset+ZZ] += tz;
489 dvda[inr] = dvda[inr] + dvdasum*isai0*isai0;
491 /* Increment number of inner iterations */
492 inneriter += j_index_end - j_index_start;
494 /* Outer loop uses 13 flops */
497 /* Increment number of outer iterations */
500 /* Update outer/inner flops */
502 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_F,outeriter*13 + inneriter*64);