File: | gromacs/gmxlib/nonbonded/nb_kernel_c/nb_kernel_ElecCSTab_VdwCSTab_GeomW3P1_c.c |
Location: | line 455, column 5 |
Description: | Value stored to 'gid' is never read |
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 | #ifdef HAVE_CONFIG_H1 |
39 | #include <config.h> |
40 | #endif |
41 | |
42 | #include <math.h> |
43 | |
44 | #include "../nb_kernel.h" |
45 | #include "types/simple.h" |
46 | #include "gromacs/math/vec.h" |
47 | #include "nrnb.h" |
48 | |
49 | /* |
50 | * Gromacs nonbonded kernel: nb_kernel_ElecCSTab_VdwCSTab_GeomW3P1_VF_c |
51 | * Electrostatics interaction: CubicSplineTable |
52 | * VdW interaction: CubicSplineTable |
53 | * Geometry: Water3-Particle |
54 | * Calculate force/pot: PotentialAndForce |
55 | */ |
56 | void |
57 | nb_kernel_ElecCSTab_VdwCSTab_GeomW3P1_VF_c |
58 | (t_nblist * gmx_restrict__restrict nlist, |
59 | rvec * gmx_restrict__restrict xx, |
60 | rvec * gmx_restrict__restrict ff, |
61 | t_forcerec * gmx_restrict__restrict fr, |
62 | t_mdatoms * gmx_restrict__restrict mdatoms, |
63 | nb_kernel_data_t gmx_unused__attribute__ ((unused)) * gmx_restrict__restrict kernel_data, |
64 | t_nrnb * gmx_restrict__restrict nrnb) |
65 | { |
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; |
72 | int vdwioffset0; |
73 | real ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0; |
74 | int vdwioffset1; |
75 | real ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1; |
76 | int vdwioffset2; |
77 | real ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2; |
78 | int vdwjidx0; |
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 velec,felec,velecsum,facel,crf,krf,krf2; |
84 | real *charge; |
85 | int nvdwtype; |
86 | real rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6; |
87 | int *vdwtype; |
88 | real *vdwparam; |
89 | int vfitab; |
90 | real rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF; |
91 | real *vftab; |
92 | |
93 | x = xx[0]; |
94 | f = ff[0]; |
95 | |
96 | nri = nlist->nri; |
97 | iinr = nlist->iinr; |
98 | jindex = nlist->jindex; |
99 | jjnr = nlist->jjnr; |
100 | shiftidx = nlist->shift; |
101 | gid = nlist->gid; |
102 | shiftvec = fr->shift_vec[0]; |
103 | fshift = fr->fshift[0]; |
104 | facel = fr->epsfac; |
105 | charge = mdatoms->chargeA; |
106 | nvdwtype = fr->ntype; |
107 | vdwparam = fr->nbfp; |
108 | vdwtype = mdatoms->typeA; |
109 | |
110 | vftab = kernel_data->table_elec_vdw->data; |
111 | vftabscale = kernel_data->table_elec_vdw->scale; |
112 | |
113 | /* Setup water-specific parameters */ |
114 | inr = nlist->iinr[0]; |
115 | iq0 = facel*charge[inr+0]; |
116 | iq1 = facel*charge[inr+1]; |
117 | iq2 = facel*charge[inr+2]; |
118 | vdwioffset0 = 2*nvdwtype*vdwtype[inr+0]; |
119 | |
120 | outeriter = 0; |
121 | inneriter = 0; |
122 | |
123 | /* Start outer loop over neighborlists */ |
124 | for(iidx=0; iidx<nri; iidx++) |
125 | { |
126 | /* Load shift vector for this list */ |
127 | i_shift_offset = DIM3*shiftidx[iidx]; |
128 | shX = shiftvec[i_shift_offset+XX0]; |
129 | shY = shiftvec[i_shift_offset+YY1]; |
130 | shZ = shiftvec[i_shift_offset+ZZ2]; |
131 | |
132 | /* Load limits for loop over neighbors */ |
133 | j_index_start = jindex[iidx]; |
134 | j_index_end = jindex[iidx+1]; |
135 | |
136 | /* Get outer coordinate index */ |
137 | inr = iinr[iidx]; |
138 | i_coord_offset = DIM3*inr; |
139 | |
140 | /* Load i particle coords and add shift vector */ |
141 | ix0 = shX + x[i_coord_offset+DIM3*0+XX0]; |
142 | iy0 = shY + x[i_coord_offset+DIM3*0+YY1]; |
143 | iz0 = shZ + x[i_coord_offset+DIM3*0+ZZ2]; |
144 | ix1 = shX + x[i_coord_offset+DIM3*1+XX0]; |
145 | iy1 = shY + x[i_coord_offset+DIM3*1+YY1]; |
146 | iz1 = shZ + x[i_coord_offset+DIM3*1+ZZ2]; |
147 | ix2 = shX + x[i_coord_offset+DIM3*2+XX0]; |
148 | iy2 = shY + x[i_coord_offset+DIM3*2+YY1]; |
149 | iz2 = shZ + x[i_coord_offset+DIM3*2+ZZ2]; |
150 | |
151 | fix0 = 0.0; |
152 | fiy0 = 0.0; |
153 | fiz0 = 0.0; |
154 | fix1 = 0.0; |
155 | fiy1 = 0.0; |
156 | fiz1 = 0.0; |
157 | fix2 = 0.0; |
158 | fiy2 = 0.0; |
159 | fiz2 = 0.0; |
160 | |
161 | /* Reset potential sums */ |
162 | velecsum = 0.0; |
163 | vvdwsum = 0.0; |
164 | |
165 | /* Start inner kernel loop */ |
166 | for(jidx=j_index_start; jidx<j_index_end; jidx++) |
167 | { |
168 | /* Get j neighbor index, and coordinate index */ |
169 | jnr = jjnr[jidx]; |
170 | j_coord_offset = DIM3*jnr; |
171 | |
172 | /* load j atom coordinates */ |
173 | jx0 = x[j_coord_offset+DIM3*0+XX0]; |
174 | jy0 = x[j_coord_offset+DIM3*0+YY1]; |
175 | jz0 = x[j_coord_offset+DIM3*0+ZZ2]; |
176 | |
177 | /* Calculate displacement vector */ |
178 | dx00 = ix0 - jx0; |
179 | dy00 = iy0 - jy0; |
180 | dz00 = iz0 - jz0; |
181 | dx10 = ix1 - jx0; |
182 | dy10 = iy1 - jy0; |
183 | dz10 = iz1 - jz0; |
184 | dx20 = ix2 - jx0; |
185 | dy20 = iy2 - jy0; |
186 | dz20 = iz2 - jz0; |
187 | |
188 | /* Calculate squared distance and things based on it */ |
189 | rsq00 = dx00*dx00+dy00*dy00+dz00*dz00; |
190 | rsq10 = dx10*dx10+dy10*dy10+dz10*dz10; |
191 | rsq20 = dx20*dx20+dy20*dy20+dz20*dz20; |
192 | |
193 | rinv00 = gmx_invsqrt(rsq00)gmx_software_invsqrt(rsq00); |
194 | rinv10 = gmx_invsqrt(rsq10)gmx_software_invsqrt(rsq10); |
195 | rinv20 = gmx_invsqrt(rsq20)gmx_software_invsqrt(rsq20); |
196 | |
197 | /* Load parameters for j particles */ |
198 | jq0 = charge[jnr+0]; |
199 | vdwjidx0 = 2*vdwtype[jnr+0]; |
200 | |
201 | /************************** |
202 | * CALCULATE INTERACTIONS * |
203 | **************************/ |
204 | |
205 | r00 = rsq00*rinv00; |
206 | |
207 | qq00 = iq0*jq0; |
208 | c6_00 = vdwparam[vdwioffset0+vdwjidx0]; |
209 | c12_00 = vdwparam[vdwioffset0+vdwjidx0+1]; |
210 | |
211 | /* Calculate table index by multiplying r with table scale and truncate to integer */ |
212 | rt = r00*vftabscale; |
213 | vfitab = rt; |
214 | vfeps = rt-vfitab; |
215 | vfitab = 3*4*vfitab; |
216 | |
217 | /* CUBIC SPLINE TABLE ELECTROSTATICS */ |
218 | Y = vftab[vfitab]; |
219 | F = vftab[vfitab+1]; |
220 | Geps = vfeps*vftab[vfitab+2]; |
221 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
222 | Fp = F+Geps+Heps2; |
223 | VV = Y+vfeps*Fp; |
224 | velec = qq00*VV; |
225 | FF = Fp+Geps+2.0*Heps2; |
226 | felec = -qq00*FF*vftabscale*rinv00; |
227 | |
228 | /* CUBIC SPLINE TABLE DISPERSION */ |
229 | vfitab += 4; |
230 | Y = vftab[vfitab]; |
231 | F = vftab[vfitab+1]; |
232 | Geps = vfeps*vftab[vfitab+2]; |
233 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
234 | Fp = F+Geps+Heps2; |
235 | VV = Y+vfeps*Fp; |
236 | vvdw6 = c6_00*VV; |
237 | FF = Fp+Geps+2.0*Heps2; |
238 | fvdw6 = c6_00*FF; |
239 | |
240 | /* CUBIC SPLINE TABLE REPULSION */ |
241 | Y = vftab[vfitab+4]; |
242 | F = vftab[vfitab+5]; |
243 | Geps = vfeps*vftab[vfitab+6]; |
244 | Heps2 = vfeps*vfeps*vftab[vfitab+7]; |
245 | Fp = F+Geps+Heps2; |
246 | VV = Y+vfeps*Fp; |
247 | vvdw12 = c12_00*VV; |
248 | FF = Fp+Geps+2.0*Heps2; |
249 | fvdw12 = c12_00*FF; |
250 | vvdw = vvdw12+vvdw6; |
251 | fvdw = -(fvdw6+fvdw12)*vftabscale*rinv00; |
252 | |
253 | /* Update potential sums from outer loop */ |
254 | velecsum += velec; |
255 | vvdwsum += vvdw; |
256 | |
257 | fscal = felec+fvdw; |
258 | |
259 | /* Calculate temporary vectorial force */ |
260 | tx = fscal*dx00; |
261 | ty = fscal*dy00; |
262 | tz = fscal*dz00; |
263 | |
264 | /* Update vectorial force */ |
265 | fix0 += tx; |
266 | fiy0 += ty; |
267 | fiz0 += tz; |
268 | f[j_coord_offset+DIM3*0+XX0] -= tx; |
269 | f[j_coord_offset+DIM3*0+YY1] -= ty; |
270 | f[j_coord_offset+DIM3*0+ZZ2] -= tz; |
271 | |
272 | /************************** |
273 | * CALCULATE INTERACTIONS * |
274 | **************************/ |
275 | |
276 | r10 = rsq10*rinv10; |
277 | |
278 | qq10 = iq1*jq0; |
279 | |
280 | /* Calculate table index by multiplying r with table scale and truncate to integer */ |
281 | rt = r10*vftabscale; |
282 | vfitab = rt; |
283 | vfeps = rt-vfitab; |
284 | vfitab = 3*4*vfitab; |
285 | |
286 | /* CUBIC SPLINE TABLE ELECTROSTATICS */ |
287 | Y = vftab[vfitab]; |
288 | F = vftab[vfitab+1]; |
289 | Geps = vfeps*vftab[vfitab+2]; |
290 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
291 | Fp = F+Geps+Heps2; |
292 | VV = Y+vfeps*Fp; |
293 | velec = qq10*VV; |
294 | FF = Fp+Geps+2.0*Heps2; |
295 | felec = -qq10*FF*vftabscale*rinv10; |
296 | |
297 | /* Update potential sums from outer loop */ |
298 | velecsum += velec; |
299 | |
300 | fscal = felec; |
301 | |
302 | /* Calculate temporary vectorial force */ |
303 | tx = fscal*dx10; |
304 | ty = fscal*dy10; |
305 | tz = fscal*dz10; |
306 | |
307 | /* Update vectorial force */ |
308 | fix1 += tx; |
309 | fiy1 += ty; |
310 | fiz1 += tz; |
311 | f[j_coord_offset+DIM3*0+XX0] -= tx; |
312 | f[j_coord_offset+DIM3*0+YY1] -= ty; |
313 | f[j_coord_offset+DIM3*0+ZZ2] -= tz; |
314 | |
315 | /************************** |
316 | * CALCULATE INTERACTIONS * |
317 | **************************/ |
318 | |
319 | r20 = rsq20*rinv20; |
320 | |
321 | qq20 = iq2*jq0; |
322 | |
323 | /* Calculate table index by multiplying r with table scale and truncate to integer */ |
324 | rt = r20*vftabscale; |
325 | vfitab = rt; |
326 | vfeps = rt-vfitab; |
327 | vfitab = 3*4*vfitab; |
328 | |
329 | /* CUBIC SPLINE TABLE ELECTROSTATICS */ |
330 | Y = vftab[vfitab]; |
331 | F = vftab[vfitab+1]; |
332 | Geps = vfeps*vftab[vfitab+2]; |
333 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
334 | Fp = F+Geps+Heps2; |
335 | VV = Y+vfeps*Fp; |
336 | velec = qq20*VV; |
337 | FF = Fp+Geps+2.0*Heps2; |
338 | felec = -qq20*FF*vftabscale*rinv20; |
339 | |
340 | /* Update potential sums from outer loop */ |
341 | velecsum += velec; |
342 | |
343 | fscal = felec; |
344 | |
345 | /* Calculate temporary vectorial force */ |
346 | tx = fscal*dx20; |
347 | ty = fscal*dy20; |
348 | tz = fscal*dz20; |
349 | |
350 | /* Update vectorial force */ |
351 | fix2 += tx; |
352 | fiy2 += ty; |
353 | fiz2 += tz; |
354 | f[j_coord_offset+DIM3*0+XX0] -= tx; |
355 | f[j_coord_offset+DIM3*0+YY1] -= ty; |
356 | f[j_coord_offset+DIM3*0+ZZ2] -= tz; |
357 | |
358 | /* Inner loop uses 157 flops */ |
359 | } |
360 | /* End of innermost loop */ |
361 | |
362 | tx = ty = tz = 0; |
363 | f[i_coord_offset+DIM3*0+XX0] += fix0; |
364 | f[i_coord_offset+DIM3*0+YY1] += fiy0; |
365 | f[i_coord_offset+DIM3*0+ZZ2] += fiz0; |
366 | tx += fix0; |
367 | ty += fiy0; |
368 | tz += fiz0; |
369 | f[i_coord_offset+DIM3*1+XX0] += fix1; |
370 | f[i_coord_offset+DIM3*1+YY1] += fiy1; |
371 | f[i_coord_offset+DIM3*1+ZZ2] += fiz1; |
372 | tx += fix1; |
373 | ty += fiy1; |
374 | tz += fiz1; |
375 | f[i_coord_offset+DIM3*2+XX0] += fix2; |
376 | f[i_coord_offset+DIM3*2+YY1] += fiy2; |
377 | f[i_coord_offset+DIM3*2+ZZ2] += fiz2; |
378 | tx += fix2; |
379 | ty += fiy2; |
380 | tz += fiz2; |
381 | fshift[i_shift_offset+XX0] += tx; |
382 | fshift[i_shift_offset+YY1] += ty; |
383 | fshift[i_shift_offset+ZZ2] += tz; |
384 | |
385 | ggid = gid[iidx]; |
386 | /* Update potential energies */ |
387 | kernel_data->energygrp_elec[ggid] += velecsum; |
388 | kernel_data->energygrp_vdw[ggid] += vvdwsum; |
389 | |
390 | /* Increment number of inner iterations */ |
391 | inneriter += j_index_end - j_index_start; |
392 | |
393 | /* Outer loop uses 32 flops */ |
394 | } |
395 | |
396 | /* Increment number of outer iterations */ |
397 | outeriter += nri; |
398 | |
399 | /* Update outer/inner flops */ |
400 | |
401 | inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W3_VF,outeriter*32 + inneriter*157)(nrnb)->n[eNR_NBKERNEL_ELEC_VDW_W3_VF] += outeriter*32 + inneriter *157; |
402 | } |
403 | /* |
404 | * Gromacs nonbonded kernel: nb_kernel_ElecCSTab_VdwCSTab_GeomW3P1_F_c |
405 | * Electrostatics interaction: CubicSplineTable |
406 | * VdW interaction: CubicSplineTable |
407 | * Geometry: Water3-Particle |
408 | * Calculate force/pot: Force |
409 | */ |
410 | void |
411 | nb_kernel_ElecCSTab_VdwCSTab_GeomW3P1_F_c |
412 | (t_nblist * gmx_restrict__restrict nlist, |
413 | rvec * gmx_restrict__restrict xx, |
414 | rvec * gmx_restrict__restrict ff, |
415 | t_forcerec * gmx_restrict__restrict fr, |
416 | t_mdatoms * gmx_restrict__restrict mdatoms, |
417 | nb_kernel_data_t gmx_unused__attribute__ ((unused)) * gmx_restrict__restrict kernel_data, |
418 | t_nrnb * gmx_restrict__restrict nrnb) |
419 | { |
420 | int i_shift_offset,i_coord_offset,j_coord_offset; |
421 | int j_index_start,j_index_end; |
422 | int nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter; |
423 | real shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2; |
424 | int *iinr,*jindex,*jjnr,*shiftidx,*gid; |
425 | real *shiftvec,*fshift,*x,*f; |
426 | int vdwioffset0; |
427 | real ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0; |
428 | int vdwioffset1; |
429 | real ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1; |
430 | int vdwioffset2; |
431 | real ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2; |
432 | int vdwjidx0; |
433 | real jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0; |
434 | real dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00; |
435 | real dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10,cexp1_10,cexp2_10; |
436 | real dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20,cexp1_20,cexp2_20; |
437 | real velec,felec,velecsum,facel,crf,krf,krf2; |
438 | real *charge; |
439 | int nvdwtype; |
440 | real rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6; |
441 | int *vdwtype; |
442 | real *vdwparam; |
443 | int vfitab; |
444 | real rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF; |
445 | real *vftab; |
446 | |
447 | x = xx[0]; |
448 | f = ff[0]; |
449 | |
450 | nri = nlist->nri; |
451 | iinr = nlist->iinr; |
452 | jindex = nlist->jindex; |
453 | jjnr = nlist->jjnr; |
454 | shiftidx = nlist->shift; |
455 | gid = nlist->gid; |
Value stored to 'gid' is never read | |
456 | shiftvec = fr->shift_vec[0]; |
457 | fshift = fr->fshift[0]; |
458 | facel = fr->epsfac; |
459 | charge = mdatoms->chargeA; |
460 | nvdwtype = fr->ntype; |
461 | vdwparam = fr->nbfp; |
462 | vdwtype = mdatoms->typeA; |
463 | |
464 | vftab = kernel_data->table_elec_vdw->data; |
465 | vftabscale = kernel_data->table_elec_vdw->scale; |
466 | |
467 | /* Setup water-specific parameters */ |
468 | inr = nlist->iinr[0]; |
469 | iq0 = facel*charge[inr+0]; |
470 | iq1 = facel*charge[inr+1]; |
471 | iq2 = facel*charge[inr+2]; |
472 | vdwioffset0 = 2*nvdwtype*vdwtype[inr+0]; |
473 | |
474 | outeriter = 0; |
475 | inneriter = 0; |
476 | |
477 | /* Start outer loop over neighborlists */ |
478 | for(iidx=0; iidx<nri; iidx++) |
479 | { |
480 | /* Load shift vector for this list */ |
481 | i_shift_offset = DIM3*shiftidx[iidx]; |
482 | shX = shiftvec[i_shift_offset+XX0]; |
483 | shY = shiftvec[i_shift_offset+YY1]; |
484 | shZ = shiftvec[i_shift_offset+ZZ2]; |
485 | |
486 | /* Load limits for loop over neighbors */ |
487 | j_index_start = jindex[iidx]; |
488 | j_index_end = jindex[iidx+1]; |
489 | |
490 | /* Get outer coordinate index */ |
491 | inr = iinr[iidx]; |
492 | i_coord_offset = DIM3*inr; |
493 | |
494 | /* Load i particle coords and add shift vector */ |
495 | ix0 = shX + x[i_coord_offset+DIM3*0+XX0]; |
496 | iy0 = shY + x[i_coord_offset+DIM3*0+YY1]; |
497 | iz0 = shZ + x[i_coord_offset+DIM3*0+ZZ2]; |
498 | ix1 = shX + x[i_coord_offset+DIM3*1+XX0]; |
499 | iy1 = shY + x[i_coord_offset+DIM3*1+YY1]; |
500 | iz1 = shZ + x[i_coord_offset+DIM3*1+ZZ2]; |
501 | ix2 = shX + x[i_coord_offset+DIM3*2+XX0]; |
502 | iy2 = shY + x[i_coord_offset+DIM3*2+YY1]; |
503 | iz2 = shZ + x[i_coord_offset+DIM3*2+ZZ2]; |
504 | |
505 | fix0 = 0.0; |
506 | fiy0 = 0.0; |
507 | fiz0 = 0.0; |
508 | fix1 = 0.0; |
509 | fiy1 = 0.0; |
510 | fiz1 = 0.0; |
511 | fix2 = 0.0; |
512 | fiy2 = 0.0; |
513 | fiz2 = 0.0; |
514 | |
515 | /* Start inner kernel loop */ |
516 | for(jidx=j_index_start; jidx<j_index_end; jidx++) |
517 | { |
518 | /* Get j neighbor index, and coordinate index */ |
519 | jnr = jjnr[jidx]; |
520 | j_coord_offset = DIM3*jnr; |
521 | |
522 | /* load j atom coordinates */ |
523 | jx0 = x[j_coord_offset+DIM3*0+XX0]; |
524 | jy0 = x[j_coord_offset+DIM3*0+YY1]; |
525 | jz0 = x[j_coord_offset+DIM3*0+ZZ2]; |
526 | |
527 | /* Calculate displacement vector */ |
528 | dx00 = ix0 - jx0; |
529 | dy00 = iy0 - jy0; |
530 | dz00 = iz0 - jz0; |
531 | dx10 = ix1 - jx0; |
532 | dy10 = iy1 - jy0; |
533 | dz10 = iz1 - jz0; |
534 | dx20 = ix2 - jx0; |
535 | dy20 = iy2 - jy0; |
536 | dz20 = iz2 - jz0; |
537 | |
538 | /* Calculate squared distance and things based on it */ |
539 | rsq00 = dx00*dx00+dy00*dy00+dz00*dz00; |
540 | rsq10 = dx10*dx10+dy10*dy10+dz10*dz10; |
541 | rsq20 = dx20*dx20+dy20*dy20+dz20*dz20; |
542 | |
543 | rinv00 = gmx_invsqrt(rsq00)gmx_software_invsqrt(rsq00); |
544 | rinv10 = gmx_invsqrt(rsq10)gmx_software_invsqrt(rsq10); |
545 | rinv20 = gmx_invsqrt(rsq20)gmx_software_invsqrt(rsq20); |
546 | |
547 | /* Load parameters for j particles */ |
548 | jq0 = charge[jnr+0]; |
549 | vdwjidx0 = 2*vdwtype[jnr+0]; |
550 | |
551 | /************************** |
552 | * CALCULATE INTERACTIONS * |
553 | **************************/ |
554 | |
555 | r00 = rsq00*rinv00; |
556 | |
557 | qq00 = iq0*jq0; |
558 | c6_00 = vdwparam[vdwioffset0+vdwjidx0]; |
559 | c12_00 = vdwparam[vdwioffset0+vdwjidx0+1]; |
560 | |
561 | /* Calculate table index by multiplying r with table scale and truncate to integer */ |
562 | rt = r00*vftabscale; |
563 | vfitab = rt; |
564 | vfeps = rt-vfitab; |
565 | vfitab = 3*4*vfitab; |
566 | |
567 | /* CUBIC SPLINE TABLE ELECTROSTATICS */ |
568 | F = vftab[vfitab+1]; |
569 | Geps = vfeps*vftab[vfitab+2]; |
570 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
571 | Fp = F+Geps+Heps2; |
572 | FF = Fp+Geps+2.0*Heps2; |
573 | felec = -qq00*FF*vftabscale*rinv00; |
574 | |
575 | /* CUBIC SPLINE TABLE DISPERSION */ |
576 | vfitab += 4; |
577 | F = vftab[vfitab+1]; |
578 | Geps = vfeps*vftab[vfitab+2]; |
579 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
580 | Fp = F+Geps+Heps2; |
581 | FF = Fp+Geps+2.0*Heps2; |
582 | fvdw6 = c6_00*FF; |
583 | |
584 | /* CUBIC SPLINE TABLE REPULSION */ |
585 | F = vftab[vfitab+5]; |
586 | Geps = vfeps*vftab[vfitab+6]; |
587 | Heps2 = vfeps*vfeps*vftab[vfitab+7]; |
588 | Fp = F+Geps+Heps2; |
589 | FF = Fp+Geps+2.0*Heps2; |
590 | fvdw12 = c12_00*FF; |
591 | fvdw = -(fvdw6+fvdw12)*vftabscale*rinv00; |
592 | |
593 | fscal = felec+fvdw; |
594 | |
595 | /* Calculate temporary vectorial force */ |
596 | tx = fscal*dx00; |
597 | ty = fscal*dy00; |
598 | tz = fscal*dz00; |
599 | |
600 | /* Update vectorial force */ |
601 | fix0 += tx; |
602 | fiy0 += ty; |
603 | fiz0 += tz; |
604 | f[j_coord_offset+DIM3*0+XX0] -= tx; |
605 | f[j_coord_offset+DIM3*0+YY1] -= ty; |
606 | f[j_coord_offset+DIM3*0+ZZ2] -= tz; |
607 | |
608 | /************************** |
609 | * CALCULATE INTERACTIONS * |
610 | **************************/ |
611 | |
612 | r10 = rsq10*rinv10; |
613 | |
614 | qq10 = iq1*jq0; |
615 | |
616 | /* Calculate table index by multiplying r with table scale and truncate to integer */ |
617 | rt = r10*vftabscale; |
618 | vfitab = rt; |
619 | vfeps = rt-vfitab; |
620 | vfitab = 3*4*vfitab; |
621 | |
622 | /* CUBIC SPLINE TABLE ELECTROSTATICS */ |
623 | F = vftab[vfitab+1]; |
624 | Geps = vfeps*vftab[vfitab+2]; |
625 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
626 | Fp = F+Geps+Heps2; |
627 | FF = Fp+Geps+2.0*Heps2; |
628 | felec = -qq10*FF*vftabscale*rinv10; |
629 | |
630 | fscal = felec; |
631 | |
632 | /* Calculate temporary vectorial force */ |
633 | tx = fscal*dx10; |
634 | ty = fscal*dy10; |
635 | tz = fscal*dz10; |
636 | |
637 | /* Update vectorial force */ |
638 | fix1 += tx; |
639 | fiy1 += ty; |
640 | fiz1 += tz; |
641 | f[j_coord_offset+DIM3*0+XX0] -= tx; |
642 | f[j_coord_offset+DIM3*0+YY1] -= ty; |
643 | f[j_coord_offset+DIM3*0+ZZ2] -= tz; |
644 | |
645 | /************************** |
646 | * CALCULATE INTERACTIONS * |
647 | **************************/ |
648 | |
649 | r20 = rsq20*rinv20; |
650 | |
651 | qq20 = iq2*jq0; |
652 | |
653 | /* Calculate table index by multiplying r with table scale and truncate to integer */ |
654 | rt = r20*vftabscale; |
655 | vfitab = rt; |
656 | vfeps = rt-vfitab; |
657 | vfitab = 3*4*vfitab; |
658 | |
659 | /* CUBIC SPLINE TABLE ELECTROSTATICS */ |
660 | F = vftab[vfitab+1]; |
661 | Geps = vfeps*vftab[vfitab+2]; |
662 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
663 | Fp = F+Geps+Heps2; |
664 | FF = Fp+Geps+2.0*Heps2; |
665 | felec = -qq20*FF*vftabscale*rinv20; |
666 | |
667 | fscal = felec; |
668 | |
669 | /* Calculate temporary vectorial force */ |
670 | tx = fscal*dx20; |
671 | ty = fscal*dy20; |
672 | tz = fscal*dz20; |
673 | |
674 | /* Update vectorial force */ |
675 | fix2 += tx; |
676 | fiy2 += ty; |
677 | fiz2 += tz; |
678 | f[j_coord_offset+DIM3*0+XX0] -= tx; |
679 | f[j_coord_offset+DIM3*0+YY1] -= ty; |
680 | f[j_coord_offset+DIM3*0+ZZ2] -= tz; |
681 | |
682 | /* Inner loop uses 137 flops */ |
683 | } |
684 | /* End of innermost loop */ |
685 | |
686 | tx = ty = tz = 0; |
687 | f[i_coord_offset+DIM3*0+XX0] += fix0; |
688 | f[i_coord_offset+DIM3*0+YY1] += fiy0; |
689 | f[i_coord_offset+DIM3*0+ZZ2] += fiz0; |
690 | tx += fix0; |
691 | ty += fiy0; |
692 | tz += fiz0; |
693 | f[i_coord_offset+DIM3*1+XX0] += fix1; |
694 | f[i_coord_offset+DIM3*1+YY1] += fiy1; |
695 | f[i_coord_offset+DIM3*1+ZZ2] += fiz1; |
696 | tx += fix1; |
697 | ty += fiy1; |
698 | tz += fiz1; |
699 | f[i_coord_offset+DIM3*2+XX0] += fix2; |
700 | f[i_coord_offset+DIM3*2+YY1] += fiy2; |
701 | f[i_coord_offset+DIM3*2+ZZ2] += fiz2; |
702 | tx += fix2; |
703 | ty += fiy2; |
704 | tz += fiz2; |
705 | fshift[i_shift_offset+XX0] += tx; |
706 | fshift[i_shift_offset+YY1] += ty; |
707 | fshift[i_shift_offset+ZZ2] += tz; |
708 | |
709 | /* Increment number of inner iterations */ |
710 | inneriter += j_index_end - j_index_start; |
711 | |
712 | /* Outer loop uses 30 flops */ |
713 | } |
714 | |
715 | /* Increment number of outer iterations */ |
716 | outeriter += nri; |
717 | |
718 | /* Update outer/inner flops */ |
719 | |
720 | inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W3_F,outeriter*30 + inneriter*137)(nrnb)->n[eNR_NBKERNEL_ELEC_VDW_W3_F] += outeriter*30 + inneriter *137; |
721 | } |