File: | gromacs/gmxlib/nonbonded/nb_kernel_c/nb_kernel_ElecCSTab_VdwLJ_GeomW4W4_c.c |
Location: | line 796, 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_VdwLJ_GeomW4W4_VF_c |
51 | * Electrostatics interaction: CubicSplineTable |
52 | * VdW interaction: LennardJones |
53 | * Geometry: Water4-Water4 |
54 | * Calculate force/pot: PotentialAndForce |
55 | */ |
56 | void |
57 | nb_kernel_ElecCSTab_VdwLJ_GeomW4W4_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 vdwioffset3; |
79 | real ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3; |
80 | int vdwjidx0; |
81 | real jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0; |
82 | int vdwjidx1; |
83 | real jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1; |
84 | int vdwjidx2; |
85 | real jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2; |
86 | int vdwjidx3; |
87 | real jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3; |
88 | real dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00; |
89 | real dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11,cexp1_11,cexp2_11; |
90 | real dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12,cexp1_12,cexp2_12; |
91 | real dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13,cexp1_13,cexp2_13; |
92 | real dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21,cexp1_21,cexp2_21; |
93 | real dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22,cexp1_22,cexp2_22; |
94 | real dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23,cexp1_23,cexp2_23; |
95 | real dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31,cexp1_31,cexp2_31; |
96 | real dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32,cexp1_32,cexp2_32; |
97 | real dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33,cexp1_33,cexp2_33; |
98 | real velec,felec,velecsum,facel,crf,krf,krf2; |
99 | real *charge; |
100 | int nvdwtype; |
101 | real rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6; |
102 | int *vdwtype; |
103 | real *vdwparam; |
104 | int vfitab; |
105 | real rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF; |
106 | real *vftab; |
107 | |
108 | x = xx[0]; |
109 | f = ff[0]; |
110 | |
111 | nri = nlist->nri; |
112 | iinr = nlist->iinr; |
113 | jindex = nlist->jindex; |
114 | jjnr = nlist->jjnr; |
115 | shiftidx = nlist->shift; |
116 | gid = nlist->gid; |
117 | shiftvec = fr->shift_vec[0]; |
118 | fshift = fr->fshift[0]; |
119 | facel = fr->epsfac; |
120 | charge = mdatoms->chargeA; |
121 | nvdwtype = fr->ntype; |
122 | vdwparam = fr->nbfp; |
123 | vdwtype = mdatoms->typeA; |
124 | |
125 | vftab = kernel_data->table_elec->data; |
126 | vftabscale = kernel_data->table_elec->scale; |
127 | |
128 | /* Setup water-specific parameters */ |
129 | inr = nlist->iinr[0]; |
130 | iq1 = facel*charge[inr+1]; |
131 | iq2 = facel*charge[inr+2]; |
132 | iq3 = facel*charge[inr+3]; |
133 | vdwioffset0 = 2*nvdwtype*vdwtype[inr+0]; |
134 | |
135 | jq1 = charge[inr+1]; |
136 | jq2 = charge[inr+2]; |
137 | jq3 = charge[inr+3]; |
138 | vdwjidx0 = 2*vdwtype[inr+0]; |
139 | c6_00 = vdwparam[vdwioffset0+vdwjidx0]; |
140 | c12_00 = vdwparam[vdwioffset0+vdwjidx0+1]; |
141 | qq11 = iq1*jq1; |
142 | qq12 = iq1*jq2; |
143 | qq13 = iq1*jq3; |
144 | qq21 = iq2*jq1; |
145 | qq22 = iq2*jq2; |
146 | qq23 = iq2*jq3; |
147 | qq31 = iq3*jq1; |
148 | qq32 = iq3*jq2; |
149 | qq33 = iq3*jq3; |
150 | |
151 | outeriter = 0; |
152 | inneriter = 0; |
153 | |
154 | /* Start outer loop over neighborlists */ |
155 | for(iidx=0; iidx<nri; iidx++) |
156 | { |
157 | /* Load shift vector for this list */ |
158 | i_shift_offset = DIM3*shiftidx[iidx]; |
159 | shX = shiftvec[i_shift_offset+XX0]; |
160 | shY = shiftvec[i_shift_offset+YY1]; |
161 | shZ = shiftvec[i_shift_offset+ZZ2]; |
162 | |
163 | /* Load limits for loop over neighbors */ |
164 | j_index_start = jindex[iidx]; |
165 | j_index_end = jindex[iidx+1]; |
166 | |
167 | /* Get outer coordinate index */ |
168 | inr = iinr[iidx]; |
169 | i_coord_offset = DIM3*inr; |
170 | |
171 | /* Load i particle coords and add shift vector */ |
172 | ix0 = shX + x[i_coord_offset+DIM3*0+XX0]; |
173 | iy0 = shY + x[i_coord_offset+DIM3*0+YY1]; |
174 | iz0 = shZ + x[i_coord_offset+DIM3*0+ZZ2]; |
175 | ix1 = shX + x[i_coord_offset+DIM3*1+XX0]; |
176 | iy1 = shY + x[i_coord_offset+DIM3*1+YY1]; |
177 | iz1 = shZ + x[i_coord_offset+DIM3*1+ZZ2]; |
178 | ix2 = shX + x[i_coord_offset+DIM3*2+XX0]; |
179 | iy2 = shY + x[i_coord_offset+DIM3*2+YY1]; |
180 | iz2 = shZ + x[i_coord_offset+DIM3*2+ZZ2]; |
181 | ix3 = shX + x[i_coord_offset+DIM3*3+XX0]; |
182 | iy3 = shY + x[i_coord_offset+DIM3*3+YY1]; |
183 | iz3 = shZ + x[i_coord_offset+DIM3*3+ZZ2]; |
184 | |
185 | fix0 = 0.0; |
186 | fiy0 = 0.0; |
187 | fiz0 = 0.0; |
188 | fix1 = 0.0; |
189 | fiy1 = 0.0; |
190 | fiz1 = 0.0; |
191 | fix2 = 0.0; |
192 | fiy2 = 0.0; |
193 | fiz2 = 0.0; |
194 | fix3 = 0.0; |
195 | fiy3 = 0.0; |
196 | fiz3 = 0.0; |
197 | |
198 | /* Reset potential sums */ |
199 | velecsum = 0.0; |
200 | vvdwsum = 0.0; |
201 | |
202 | /* Start inner kernel loop */ |
203 | for(jidx=j_index_start; jidx<j_index_end; jidx++) |
204 | { |
205 | /* Get j neighbor index, and coordinate index */ |
206 | jnr = jjnr[jidx]; |
207 | j_coord_offset = DIM3*jnr; |
208 | |
209 | /* load j atom coordinates */ |
210 | jx0 = x[j_coord_offset+DIM3*0+XX0]; |
211 | jy0 = x[j_coord_offset+DIM3*0+YY1]; |
212 | jz0 = x[j_coord_offset+DIM3*0+ZZ2]; |
213 | jx1 = x[j_coord_offset+DIM3*1+XX0]; |
214 | jy1 = x[j_coord_offset+DIM3*1+YY1]; |
215 | jz1 = x[j_coord_offset+DIM3*1+ZZ2]; |
216 | jx2 = x[j_coord_offset+DIM3*2+XX0]; |
217 | jy2 = x[j_coord_offset+DIM3*2+YY1]; |
218 | jz2 = x[j_coord_offset+DIM3*2+ZZ2]; |
219 | jx3 = x[j_coord_offset+DIM3*3+XX0]; |
220 | jy3 = x[j_coord_offset+DIM3*3+YY1]; |
221 | jz3 = x[j_coord_offset+DIM3*3+ZZ2]; |
222 | |
223 | /* Calculate displacement vector */ |
224 | dx00 = ix0 - jx0; |
225 | dy00 = iy0 - jy0; |
226 | dz00 = iz0 - jz0; |
227 | dx11 = ix1 - jx1; |
228 | dy11 = iy1 - jy1; |
229 | dz11 = iz1 - jz1; |
230 | dx12 = ix1 - jx2; |
231 | dy12 = iy1 - jy2; |
232 | dz12 = iz1 - jz2; |
233 | dx13 = ix1 - jx3; |
234 | dy13 = iy1 - jy3; |
235 | dz13 = iz1 - jz3; |
236 | dx21 = ix2 - jx1; |
237 | dy21 = iy2 - jy1; |
238 | dz21 = iz2 - jz1; |
239 | dx22 = ix2 - jx2; |
240 | dy22 = iy2 - jy2; |
241 | dz22 = iz2 - jz2; |
242 | dx23 = ix2 - jx3; |
243 | dy23 = iy2 - jy3; |
244 | dz23 = iz2 - jz3; |
245 | dx31 = ix3 - jx1; |
246 | dy31 = iy3 - jy1; |
247 | dz31 = iz3 - jz1; |
248 | dx32 = ix3 - jx2; |
249 | dy32 = iy3 - jy2; |
250 | dz32 = iz3 - jz2; |
251 | dx33 = ix3 - jx3; |
252 | dy33 = iy3 - jy3; |
253 | dz33 = iz3 - jz3; |
254 | |
255 | /* Calculate squared distance and things based on it */ |
256 | rsq00 = dx00*dx00+dy00*dy00+dz00*dz00; |
257 | rsq11 = dx11*dx11+dy11*dy11+dz11*dz11; |
258 | rsq12 = dx12*dx12+dy12*dy12+dz12*dz12; |
259 | rsq13 = dx13*dx13+dy13*dy13+dz13*dz13; |
260 | rsq21 = dx21*dx21+dy21*dy21+dz21*dz21; |
261 | rsq22 = dx22*dx22+dy22*dy22+dz22*dz22; |
262 | rsq23 = dx23*dx23+dy23*dy23+dz23*dz23; |
263 | rsq31 = dx31*dx31+dy31*dy31+dz31*dz31; |
264 | rsq32 = dx32*dx32+dy32*dy32+dz32*dz32; |
265 | rsq33 = dx33*dx33+dy33*dy33+dz33*dz33; |
266 | |
267 | rinv11 = gmx_invsqrt(rsq11)gmx_software_invsqrt(rsq11); |
268 | rinv12 = gmx_invsqrt(rsq12)gmx_software_invsqrt(rsq12); |
269 | rinv13 = gmx_invsqrt(rsq13)gmx_software_invsqrt(rsq13); |
270 | rinv21 = gmx_invsqrt(rsq21)gmx_software_invsqrt(rsq21); |
271 | rinv22 = gmx_invsqrt(rsq22)gmx_software_invsqrt(rsq22); |
272 | rinv23 = gmx_invsqrt(rsq23)gmx_software_invsqrt(rsq23); |
273 | rinv31 = gmx_invsqrt(rsq31)gmx_software_invsqrt(rsq31); |
274 | rinv32 = gmx_invsqrt(rsq32)gmx_software_invsqrt(rsq32); |
275 | rinv33 = gmx_invsqrt(rsq33)gmx_software_invsqrt(rsq33); |
276 | |
277 | rinvsq00 = 1.0/rsq00; |
278 | |
279 | /************************** |
280 | * CALCULATE INTERACTIONS * |
281 | **************************/ |
282 | |
283 | /* LENNARD-JONES DISPERSION/REPULSION */ |
284 | |
285 | rinvsix = rinvsq00*rinvsq00*rinvsq00; |
286 | vvdw6 = c6_00*rinvsix; |
287 | vvdw12 = c12_00*rinvsix*rinvsix; |
288 | vvdw = vvdw12*(1.0/12.0) - vvdw6*(1.0/6.0); |
289 | fvdw = (vvdw12-vvdw6)*rinvsq00; |
290 | |
291 | /* Update potential sums from outer loop */ |
292 | vvdwsum += vvdw; |
293 | |
294 | fscal = fvdw; |
295 | |
296 | /* Calculate temporary vectorial force */ |
297 | tx = fscal*dx00; |
298 | ty = fscal*dy00; |
299 | tz = fscal*dz00; |
300 | |
301 | /* Update vectorial force */ |
302 | fix0 += tx; |
303 | fiy0 += ty; |
304 | fiz0 += tz; |
305 | f[j_coord_offset+DIM3*0+XX0] -= tx; |
306 | f[j_coord_offset+DIM3*0+YY1] -= ty; |
307 | f[j_coord_offset+DIM3*0+ZZ2] -= tz; |
308 | |
309 | /************************** |
310 | * CALCULATE INTERACTIONS * |
311 | **************************/ |
312 | |
313 | r11 = rsq11*rinv11; |
314 | |
315 | /* Calculate table index by multiplying r with table scale and truncate to integer */ |
316 | rt = r11*vftabscale; |
317 | vfitab = rt; |
318 | vfeps = rt-vfitab; |
319 | vfitab = 1*4*vfitab; |
320 | |
321 | /* CUBIC SPLINE TABLE ELECTROSTATICS */ |
322 | Y = vftab[vfitab]; |
323 | F = vftab[vfitab+1]; |
324 | Geps = vfeps*vftab[vfitab+2]; |
325 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
326 | Fp = F+Geps+Heps2; |
327 | VV = Y+vfeps*Fp; |
328 | velec = qq11*VV; |
329 | FF = Fp+Geps+2.0*Heps2; |
330 | felec = -qq11*FF*vftabscale*rinv11; |
331 | |
332 | /* Update potential sums from outer loop */ |
333 | velecsum += velec; |
334 | |
335 | fscal = felec; |
336 | |
337 | /* Calculate temporary vectorial force */ |
338 | tx = fscal*dx11; |
339 | ty = fscal*dy11; |
340 | tz = fscal*dz11; |
341 | |
342 | /* Update vectorial force */ |
343 | fix1 += tx; |
344 | fiy1 += ty; |
345 | fiz1 += tz; |
346 | f[j_coord_offset+DIM3*1+XX0] -= tx; |
347 | f[j_coord_offset+DIM3*1+YY1] -= ty; |
348 | f[j_coord_offset+DIM3*1+ZZ2] -= tz; |
349 | |
350 | /************************** |
351 | * CALCULATE INTERACTIONS * |
352 | **************************/ |
353 | |
354 | r12 = rsq12*rinv12; |
355 | |
356 | /* Calculate table index by multiplying r with table scale and truncate to integer */ |
357 | rt = r12*vftabscale; |
358 | vfitab = rt; |
359 | vfeps = rt-vfitab; |
360 | vfitab = 1*4*vfitab; |
361 | |
362 | /* CUBIC SPLINE TABLE ELECTROSTATICS */ |
363 | Y = vftab[vfitab]; |
364 | F = vftab[vfitab+1]; |
365 | Geps = vfeps*vftab[vfitab+2]; |
366 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
367 | Fp = F+Geps+Heps2; |
368 | VV = Y+vfeps*Fp; |
369 | velec = qq12*VV; |
370 | FF = Fp+Geps+2.0*Heps2; |
371 | felec = -qq12*FF*vftabscale*rinv12; |
372 | |
373 | /* Update potential sums from outer loop */ |
374 | velecsum += velec; |
375 | |
376 | fscal = felec; |
377 | |
378 | /* Calculate temporary vectorial force */ |
379 | tx = fscal*dx12; |
380 | ty = fscal*dy12; |
381 | tz = fscal*dz12; |
382 | |
383 | /* Update vectorial force */ |
384 | fix1 += tx; |
385 | fiy1 += ty; |
386 | fiz1 += tz; |
387 | f[j_coord_offset+DIM3*2+XX0] -= tx; |
388 | f[j_coord_offset+DIM3*2+YY1] -= ty; |
389 | f[j_coord_offset+DIM3*2+ZZ2] -= tz; |
390 | |
391 | /************************** |
392 | * CALCULATE INTERACTIONS * |
393 | **************************/ |
394 | |
395 | r13 = rsq13*rinv13; |
396 | |
397 | /* Calculate table index by multiplying r with table scale and truncate to integer */ |
398 | rt = r13*vftabscale; |
399 | vfitab = rt; |
400 | vfeps = rt-vfitab; |
401 | vfitab = 1*4*vfitab; |
402 | |
403 | /* CUBIC SPLINE TABLE ELECTROSTATICS */ |
404 | Y = vftab[vfitab]; |
405 | F = vftab[vfitab+1]; |
406 | Geps = vfeps*vftab[vfitab+2]; |
407 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
408 | Fp = F+Geps+Heps2; |
409 | VV = Y+vfeps*Fp; |
410 | velec = qq13*VV; |
411 | FF = Fp+Geps+2.0*Heps2; |
412 | felec = -qq13*FF*vftabscale*rinv13; |
413 | |
414 | /* Update potential sums from outer loop */ |
415 | velecsum += velec; |
416 | |
417 | fscal = felec; |
418 | |
419 | /* Calculate temporary vectorial force */ |
420 | tx = fscal*dx13; |
421 | ty = fscal*dy13; |
422 | tz = fscal*dz13; |
423 | |
424 | /* Update vectorial force */ |
425 | fix1 += tx; |
426 | fiy1 += ty; |
427 | fiz1 += tz; |
428 | f[j_coord_offset+DIM3*3+XX0] -= tx; |
429 | f[j_coord_offset+DIM3*3+YY1] -= ty; |
430 | f[j_coord_offset+DIM3*3+ZZ2] -= tz; |
431 | |
432 | /************************** |
433 | * CALCULATE INTERACTIONS * |
434 | **************************/ |
435 | |
436 | r21 = rsq21*rinv21; |
437 | |
438 | /* Calculate table index by multiplying r with table scale and truncate to integer */ |
439 | rt = r21*vftabscale; |
440 | vfitab = rt; |
441 | vfeps = rt-vfitab; |
442 | vfitab = 1*4*vfitab; |
443 | |
444 | /* CUBIC SPLINE TABLE ELECTROSTATICS */ |
445 | Y = vftab[vfitab]; |
446 | F = vftab[vfitab+1]; |
447 | Geps = vfeps*vftab[vfitab+2]; |
448 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
449 | Fp = F+Geps+Heps2; |
450 | VV = Y+vfeps*Fp; |
451 | velec = qq21*VV; |
452 | FF = Fp+Geps+2.0*Heps2; |
453 | felec = -qq21*FF*vftabscale*rinv21; |
454 | |
455 | /* Update potential sums from outer loop */ |
456 | velecsum += velec; |
457 | |
458 | fscal = felec; |
459 | |
460 | /* Calculate temporary vectorial force */ |
461 | tx = fscal*dx21; |
462 | ty = fscal*dy21; |
463 | tz = fscal*dz21; |
464 | |
465 | /* Update vectorial force */ |
466 | fix2 += tx; |
467 | fiy2 += ty; |
468 | fiz2 += tz; |
469 | f[j_coord_offset+DIM3*1+XX0] -= tx; |
470 | f[j_coord_offset+DIM3*1+YY1] -= ty; |
471 | f[j_coord_offset+DIM3*1+ZZ2] -= tz; |
472 | |
473 | /************************** |
474 | * CALCULATE INTERACTIONS * |
475 | **************************/ |
476 | |
477 | r22 = rsq22*rinv22; |
478 | |
479 | /* Calculate table index by multiplying r with table scale and truncate to integer */ |
480 | rt = r22*vftabscale; |
481 | vfitab = rt; |
482 | vfeps = rt-vfitab; |
483 | vfitab = 1*4*vfitab; |
484 | |
485 | /* CUBIC SPLINE TABLE ELECTROSTATICS */ |
486 | Y = vftab[vfitab]; |
487 | F = vftab[vfitab+1]; |
488 | Geps = vfeps*vftab[vfitab+2]; |
489 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
490 | Fp = F+Geps+Heps2; |
491 | VV = Y+vfeps*Fp; |
492 | velec = qq22*VV; |
493 | FF = Fp+Geps+2.0*Heps2; |
494 | felec = -qq22*FF*vftabscale*rinv22; |
495 | |
496 | /* Update potential sums from outer loop */ |
497 | velecsum += velec; |
498 | |
499 | fscal = felec; |
500 | |
501 | /* Calculate temporary vectorial force */ |
502 | tx = fscal*dx22; |
503 | ty = fscal*dy22; |
504 | tz = fscal*dz22; |
505 | |
506 | /* Update vectorial force */ |
507 | fix2 += tx; |
508 | fiy2 += ty; |
509 | fiz2 += tz; |
510 | f[j_coord_offset+DIM3*2+XX0] -= tx; |
511 | f[j_coord_offset+DIM3*2+YY1] -= ty; |
512 | f[j_coord_offset+DIM3*2+ZZ2] -= tz; |
513 | |
514 | /************************** |
515 | * CALCULATE INTERACTIONS * |
516 | **************************/ |
517 | |
518 | r23 = rsq23*rinv23; |
519 | |
520 | /* Calculate table index by multiplying r with table scale and truncate to integer */ |
521 | rt = r23*vftabscale; |
522 | vfitab = rt; |
523 | vfeps = rt-vfitab; |
524 | vfitab = 1*4*vfitab; |
525 | |
526 | /* CUBIC SPLINE TABLE ELECTROSTATICS */ |
527 | Y = vftab[vfitab]; |
528 | F = vftab[vfitab+1]; |
529 | Geps = vfeps*vftab[vfitab+2]; |
530 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
531 | Fp = F+Geps+Heps2; |
532 | VV = Y+vfeps*Fp; |
533 | velec = qq23*VV; |
534 | FF = Fp+Geps+2.0*Heps2; |
535 | felec = -qq23*FF*vftabscale*rinv23; |
536 | |
537 | /* Update potential sums from outer loop */ |
538 | velecsum += velec; |
539 | |
540 | fscal = felec; |
541 | |
542 | /* Calculate temporary vectorial force */ |
543 | tx = fscal*dx23; |
544 | ty = fscal*dy23; |
545 | tz = fscal*dz23; |
546 | |
547 | /* Update vectorial force */ |
548 | fix2 += tx; |
549 | fiy2 += ty; |
550 | fiz2 += tz; |
551 | f[j_coord_offset+DIM3*3+XX0] -= tx; |
552 | f[j_coord_offset+DIM3*3+YY1] -= ty; |
553 | f[j_coord_offset+DIM3*3+ZZ2] -= tz; |
554 | |
555 | /************************** |
556 | * CALCULATE INTERACTIONS * |
557 | **************************/ |
558 | |
559 | r31 = rsq31*rinv31; |
560 | |
561 | /* Calculate table index by multiplying r with table scale and truncate to integer */ |
562 | rt = r31*vftabscale; |
563 | vfitab = rt; |
564 | vfeps = rt-vfitab; |
565 | vfitab = 1*4*vfitab; |
566 | |
567 | /* CUBIC SPLINE TABLE ELECTROSTATICS */ |
568 | Y = vftab[vfitab]; |
569 | F = vftab[vfitab+1]; |
570 | Geps = vfeps*vftab[vfitab+2]; |
571 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
572 | Fp = F+Geps+Heps2; |
573 | VV = Y+vfeps*Fp; |
574 | velec = qq31*VV; |
575 | FF = Fp+Geps+2.0*Heps2; |
576 | felec = -qq31*FF*vftabscale*rinv31; |
577 | |
578 | /* Update potential sums from outer loop */ |
579 | velecsum += velec; |
580 | |
581 | fscal = felec; |
582 | |
583 | /* Calculate temporary vectorial force */ |
584 | tx = fscal*dx31; |
585 | ty = fscal*dy31; |
586 | tz = fscal*dz31; |
587 | |
588 | /* Update vectorial force */ |
589 | fix3 += tx; |
590 | fiy3 += ty; |
591 | fiz3 += tz; |
592 | f[j_coord_offset+DIM3*1+XX0] -= tx; |
593 | f[j_coord_offset+DIM3*1+YY1] -= ty; |
594 | f[j_coord_offset+DIM3*1+ZZ2] -= tz; |
595 | |
596 | /************************** |
597 | * CALCULATE INTERACTIONS * |
598 | **************************/ |
599 | |
600 | r32 = rsq32*rinv32; |
601 | |
602 | /* Calculate table index by multiplying r with table scale and truncate to integer */ |
603 | rt = r32*vftabscale; |
604 | vfitab = rt; |
605 | vfeps = rt-vfitab; |
606 | vfitab = 1*4*vfitab; |
607 | |
608 | /* CUBIC SPLINE TABLE ELECTROSTATICS */ |
609 | Y = vftab[vfitab]; |
610 | F = vftab[vfitab+1]; |
611 | Geps = vfeps*vftab[vfitab+2]; |
612 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
613 | Fp = F+Geps+Heps2; |
614 | VV = Y+vfeps*Fp; |
615 | velec = qq32*VV; |
616 | FF = Fp+Geps+2.0*Heps2; |
617 | felec = -qq32*FF*vftabscale*rinv32; |
618 | |
619 | /* Update potential sums from outer loop */ |
620 | velecsum += velec; |
621 | |
622 | fscal = felec; |
623 | |
624 | /* Calculate temporary vectorial force */ |
625 | tx = fscal*dx32; |
626 | ty = fscal*dy32; |
627 | tz = fscal*dz32; |
628 | |
629 | /* Update vectorial force */ |
630 | fix3 += tx; |
631 | fiy3 += ty; |
632 | fiz3 += tz; |
633 | f[j_coord_offset+DIM3*2+XX0] -= tx; |
634 | f[j_coord_offset+DIM3*2+YY1] -= ty; |
635 | f[j_coord_offset+DIM3*2+ZZ2] -= tz; |
636 | |
637 | /************************** |
638 | * CALCULATE INTERACTIONS * |
639 | **************************/ |
640 | |
641 | r33 = rsq33*rinv33; |
642 | |
643 | /* Calculate table index by multiplying r with table scale and truncate to integer */ |
644 | rt = r33*vftabscale; |
645 | vfitab = rt; |
646 | vfeps = rt-vfitab; |
647 | vfitab = 1*4*vfitab; |
648 | |
649 | /* CUBIC SPLINE TABLE ELECTROSTATICS */ |
650 | Y = vftab[vfitab]; |
651 | F = vftab[vfitab+1]; |
652 | Geps = vfeps*vftab[vfitab+2]; |
653 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
654 | Fp = F+Geps+Heps2; |
655 | VV = Y+vfeps*Fp; |
656 | velec = qq33*VV; |
657 | FF = Fp+Geps+2.0*Heps2; |
658 | felec = -qq33*FF*vftabscale*rinv33; |
659 | |
660 | /* Update potential sums from outer loop */ |
661 | velecsum += velec; |
662 | |
663 | fscal = felec; |
664 | |
665 | /* Calculate temporary vectorial force */ |
666 | tx = fscal*dx33; |
667 | ty = fscal*dy33; |
668 | tz = fscal*dz33; |
669 | |
670 | /* Update vectorial force */ |
671 | fix3 += tx; |
672 | fiy3 += ty; |
673 | fiz3 += tz; |
674 | f[j_coord_offset+DIM3*3+XX0] -= tx; |
675 | f[j_coord_offset+DIM3*3+YY1] -= ty; |
676 | f[j_coord_offset+DIM3*3+ZZ2] -= tz; |
677 | |
678 | /* Inner loop uses 401 flops */ |
679 | } |
680 | /* End of innermost loop */ |
681 | |
682 | tx = ty = tz = 0; |
683 | f[i_coord_offset+DIM3*0+XX0] += fix0; |
684 | f[i_coord_offset+DIM3*0+YY1] += fiy0; |
685 | f[i_coord_offset+DIM3*0+ZZ2] += fiz0; |
686 | tx += fix0; |
687 | ty += fiy0; |
688 | tz += fiz0; |
689 | f[i_coord_offset+DIM3*1+XX0] += fix1; |
690 | f[i_coord_offset+DIM3*1+YY1] += fiy1; |
691 | f[i_coord_offset+DIM3*1+ZZ2] += fiz1; |
692 | tx += fix1; |
693 | ty += fiy1; |
694 | tz += fiz1; |
695 | f[i_coord_offset+DIM3*2+XX0] += fix2; |
696 | f[i_coord_offset+DIM3*2+YY1] += fiy2; |
697 | f[i_coord_offset+DIM3*2+ZZ2] += fiz2; |
698 | tx += fix2; |
699 | ty += fiy2; |
700 | tz += fiz2; |
701 | f[i_coord_offset+DIM3*3+XX0] += fix3; |
702 | f[i_coord_offset+DIM3*3+YY1] += fiy3; |
703 | f[i_coord_offset+DIM3*3+ZZ2] += fiz3; |
704 | tx += fix3; |
705 | ty += fiy3; |
706 | tz += fiz3; |
707 | fshift[i_shift_offset+XX0] += tx; |
708 | fshift[i_shift_offset+YY1] += ty; |
709 | fshift[i_shift_offset+ZZ2] += tz; |
710 | |
711 | ggid = gid[iidx]; |
712 | /* Update potential energies */ |
713 | kernel_data->energygrp_elec[ggid] += velecsum; |
714 | kernel_data->energygrp_vdw[ggid] += vvdwsum; |
715 | |
716 | /* Increment number of inner iterations */ |
717 | inneriter += j_index_end - j_index_start; |
718 | |
719 | /* Outer loop uses 41 flops */ |
720 | } |
721 | |
722 | /* Increment number of outer iterations */ |
723 | outeriter += nri; |
724 | |
725 | /* Update outer/inner flops */ |
726 | |
727 | inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4W4_VF,outeriter*41 + inneriter*401)(nrnb)->n[eNR_NBKERNEL_ELEC_VDW_W4W4_VF] += outeriter*41 + inneriter*401; |
728 | } |
729 | /* |
730 | * Gromacs nonbonded kernel: nb_kernel_ElecCSTab_VdwLJ_GeomW4W4_F_c |
731 | * Electrostatics interaction: CubicSplineTable |
732 | * VdW interaction: LennardJones |
733 | * Geometry: Water4-Water4 |
734 | * Calculate force/pot: Force |
735 | */ |
736 | void |
737 | nb_kernel_ElecCSTab_VdwLJ_GeomW4W4_F_c |
738 | (t_nblist * gmx_restrict__restrict nlist, |
739 | rvec * gmx_restrict__restrict xx, |
740 | rvec * gmx_restrict__restrict ff, |
741 | t_forcerec * gmx_restrict__restrict fr, |
742 | t_mdatoms * gmx_restrict__restrict mdatoms, |
743 | nb_kernel_data_t gmx_unused__attribute__ ((unused)) * gmx_restrict__restrict kernel_data, |
744 | t_nrnb * gmx_restrict__restrict nrnb) |
745 | { |
746 | int i_shift_offset,i_coord_offset,j_coord_offset; |
747 | int j_index_start,j_index_end; |
748 | int nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter; |
749 | real shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2; |
750 | int *iinr,*jindex,*jjnr,*shiftidx,*gid; |
751 | real *shiftvec,*fshift,*x,*f; |
752 | int vdwioffset0; |
753 | real ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0; |
754 | int vdwioffset1; |
755 | real ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1; |
756 | int vdwioffset2; |
757 | real ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2; |
758 | int vdwioffset3; |
759 | real ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3; |
760 | int vdwjidx0; |
761 | real jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0; |
762 | int vdwjidx1; |
763 | real jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1; |
764 | int vdwjidx2; |
765 | real jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2; |
766 | int vdwjidx3; |
767 | real jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3; |
768 | real dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00; |
769 | real dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11,cexp1_11,cexp2_11; |
770 | real dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12,cexp1_12,cexp2_12; |
771 | real dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13,cexp1_13,cexp2_13; |
772 | real dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21,cexp1_21,cexp2_21; |
773 | real dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22,cexp1_22,cexp2_22; |
774 | real dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23,cexp1_23,cexp2_23; |
775 | real dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31,cexp1_31,cexp2_31; |
776 | real dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32,cexp1_32,cexp2_32; |
777 | real dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33,cexp1_33,cexp2_33; |
778 | real velec,felec,velecsum,facel,crf,krf,krf2; |
779 | real *charge; |
780 | int nvdwtype; |
781 | real rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6; |
782 | int *vdwtype; |
783 | real *vdwparam; |
784 | int vfitab; |
785 | real rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF; |
786 | real *vftab; |
787 | |
788 | x = xx[0]; |
789 | f = ff[0]; |
790 | |
791 | nri = nlist->nri; |
792 | iinr = nlist->iinr; |
793 | jindex = nlist->jindex; |
794 | jjnr = nlist->jjnr; |
795 | shiftidx = nlist->shift; |
796 | gid = nlist->gid; |
Value stored to 'gid' is never read | |
797 | shiftvec = fr->shift_vec[0]; |
798 | fshift = fr->fshift[0]; |
799 | facel = fr->epsfac; |
800 | charge = mdatoms->chargeA; |
801 | nvdwtype = fr->ntype; |
802 | vdwparam = fr->nbfp; |
803 | vdwtype = mdatoms->typeA; |
804 | |
805 | vftab = kernel_data->table_elec->data; |
806 | vftabscale = kernel_data->table_elec->scale; |
807 | |
808 | /* Setup water-specific parameters */ |
809 | inr = nlist->iinr[0]; |
810 | iq1 = facel*charge[inr+1]; |
811 | iq2 = facel*charge[inr+2]; |
812 | iq3 = facel*charge[inr+3]; |
813 | vdwioffset0 = 2*nvdwtype*vdwtype[inr+0]; |
814 | |
815 | jq1 = charge[inr+1]; |
816 | jq2 = charge[inr+2]; |
817 | jq3 = charge[inr+3]; |
818 | vdwjidx0 = 2*vdwtype[inr+0]; |
819 | c6_00 = vdwparam[vdwioffset0+vdwjidx0]; |
820 | c12_00 = vdwparam[vdwioffset0+vdwjidx0+1]; |
821 | qq11 = iq1*jq1; |
822 | qq12 = iq1*jq2; |
823 | qq13 = iq1*jq3; |
824 | qq21 = iq2*jq1; |
825 | qq22 = iq2*jq2; |
826 | qq23 = iq2*jq3; |
827 | qq31 = iq3*jq1; |
828 | qq32 = iq3*jq2; |
829 | qq33 = iq3*jq3; |
830 | |
831 | outeriter = 0; |
832 | inneriter = 0; |
833 | |
834 | /* Start outer loop over neighborlists */ |
835 | for(iidx=0; iidx<nri; iidx++) |
836 | { |
837 | /* Load shift vector for this list */ |
838 | i_shift_offset = DIM3*shiftidx[iidx]; |
839 | shX = shiftvec[i_shift_offset+XX0]; |
840 | shY = shiftvec[i_shift_offset+YY1]; |
841 | shZ = shiftvec[i_shift_offset+ZZ2]; |
842 | |
843 | /* Load limits for loop over neighbors */ |
844 | j_index_start = jindex[iidx]; |
845 | j_index_end = jindex[iidx+1]; |
846 | |
847 | /* Get outer coordinate index */ |
848 | inr = iinr[iidx]; |
849 | i_coord_offset = DIM3*inr; |
850 | |
851 | /* Load i particle coords and add shift vector */ |
852 | ix0 = shX + x[i_coord_offset+DIM3*0+XX0]; |
853 | iy0 = shY + x[i_coord_offset+DIM3*0+YY1]; |
854 | iz0 = shZ + x[i_coord_offset+DIM3*0+ZZ2]; |
855 | ix1 = shX + x[i_coord_offset+DIM3*1+XX0]; |
856 | iy1 = shY + x[i_coord_offset+DIM3*1+YY1]; |
857 | iz1 = shZ + x[i_coord_offset+DIM3*1+ZZ2]; |
858 | ix2 = shX + x[i_coord_offset+DIM3*2+XX0]; |
859 | iy2 = shY + x[i_coord_offset+DIM3*2+YY1]; |
860 | iz2 = shZ + x[i_coord_offset+DIM3*2+ZZ2]; |
861 | ix3 = shX + x[i_coord_offset+DIM3*3+XX0]; |
862 | iy3 = shY + x[i_coord_offset+DIM3*3+YY1]; |
863 | iz3 = shZ + x[i_coord_offset+DIM3*3+ZZ2]; |
864 | |
865 | fix0 = 0.0; |
866 | fiy0 = 0.0; |
867 | fiz0 = 0.0; |
868 | fix1 = 0.0; |
869 | fiy1 = 0.0; |
870 | fiz1 = 0.0; |
871 | fix2 = 0.0; |
872 | fiy2 = 0.0; |
873 | fiz2 = 0.0; |
874 | fix3 = 0.0; |
875 | fiy3 = 0.0; |
876 | fiz3 = 0.0; |
877 | |
878 | /* Start inner kernel loop */ |
879 | for(jidx=j_index_start; jidx<j_index_end; jidx++) |
880 | { |
881 | /* Get j neighbor index, and coordinate index */ |
882 | jnr = jjnr[jidx]; |
883 | j_coord_offset = DIM3*jnr; |
884 | |
885 | /* load j atom coordinates */ |
886 | jx0 = x[j_coord_offset+DIM3*0+XX0]; |
887 | jy0 = x[j_coord_offset+DIM3*0+YY1]; |
888 | jz0 = x[j_coord_offset+DIM3*0+ZZ2]; |
889 | jx1 = x[j_coord_offset+DIM3*1+XX0]; |
890 | jy1 = x[j_coord_offset+DIM3*1+YY1]; |
891 | jz1 = x[j_coord_offset+DIM3*1+ZZ2]; |
892 | jx2 = x[j_coord_offset+DIM3*2+XX0]; |
893 | jy2 = x[j_coord_offset+DIM3*2+YY1]; |
894 | jz2 = x[j_coord_offset+DIM3*2+ZZ2]; |
895 | jx3 = x[j_coord_offset+DIM3*3+XX0]; |
896 | jy3 = x[j_coord_offset+DIM3*3+YY1]; |
897 | jz3 = x[j_coord_offset+DIM3*3+ZZ2]; |
898 | |
899 | /* Calculate displacement vector */ |
900 | dx00 = ix0 - jx0; |
901 | dy00 = iy0 - jy0; |
902 | dz00 = iz0 - jz0; |
903 | dx11 = ix1 - jx1; |
904 | dy11 = iy1 - jy1; |
905 | dz11 = iz1 - jz1; |
906 | dx12 = ix1 - jx2; |
907 | dy12 = iy1 - jy2; |
908 | dz12 = iz1 - jz2; |
909 | dx13 = ix1 - jx3; |
910 | dy13 = iy1 - jy3; |
911 | dz13 = iz1 - jz3; |
912 | dx21 = ix2 - jx1; |
913 | dy21 = iy2 - jy1; |
914 | dz21 = iz2 - jz1; |
915 | dx22 = ix2 - jx2; |
916 | dy22 = iy2 - jy2; |
917 | dz22 = iz2 - jz2; |
918 | dx23 = ix2 - jx3; |
919 | dy23 = iy2 - jy3; |
920 | dz23 = iz2 - jz3; |
921 | dx31 = ix3 - jx1; |
922 | dy31 = iy3 - jy1; |
923 | dz31 = iz3 - jz1; |
924 | dx32 = ix3 - jx2; |
925 | dy32 = iy3 - jy2; |
926 | dz32 = iz3 - jz2; |
927 | dx33 = ix3 - jx3; |
928 | dy33 = iy3 - jy3; |
929 | dz33 = iz3 - jz3; |
930 | |
931 | /* Calculate squared distance and things based on it */ |
932 | rsq00 = dx00*dx00+dy00*dy00+dz00*dz00; |
933 | rsq11 = dx11*dx11+dy11*dy11+dz11*dz11; |
934 | rsq12 = dx12*dx12+dy12*dy12+dz12*dz12; |
935 | rsq13 = dx13*dx13+dy13*dy13+dz13*dz13; |
936 | rsq21 = dx21*dx21+dy21*dy21+dz21*dz21; |
937 | rsq22 = dx22*dx22+dy22*dy22+dz22*dz22; |
938 | rsq23 = dx23*dx23+dy23*dy23+dz23*dz23; |
939 | rsq31 = dx31*dx31+dy31*dy31+dz31*dz31; |
940 | rsq32 = dx32*dx32+dy32*dy32+dz32*dz32; |
941 | rsq33 = dx33*dx33+dy33*dy33+dz33*dz33; |
942 | |
943 | rinv11 = gmx_invsqrt(rsq11)gmx_software_invsqrt(rsq11); |
944 | rinv12 = gmx_invsqrt(rsq12)gmx_software_invsqrt(rsq12); |
945 | rinv13 = gmx_invsqrt(rsq13)gmx_software_invsqrt(rsq13); |
946 | rinv21 = gmx_invsqrt(rsq21)gmx_software_invsqrt(rsq21); |
947 | rinv22 = gmx_invsqrt(rsq22)gmx_software_invsqrt(rsq22); |
948 | rinv23 = gmx_invsqrt(rsq23)gmx_software_invsqrt(rsq23); |
949 | rinv31 = gmx_invsqrt(rsq31)gmx_software_invsqrt(rsq31); |
950 | rinv32 = gmx_invsqrt(rsq32)gmx_software_invsqrt(rsq32); |
951 | rinv33 = gmx_invsqrt(rsq33)gmx_software_invsqrt(rsq33); |
952 | |
953 | rinvsq00 = 1.0/rsq00; |
954 | |
955 | /************************** |
956 | * CALCULATE INTERACTIONS * |
957 | **************************/ |
958 | |
959 | /* LENNARD-JONES DISPERSION/REPULSION */ |
960 | |
961 | rinvsix = rinvsq00*rinvsq00*rinvsq00; |
962 | fvdw = (c12_00*rinvsix-c6_00)*rinvsix*rinvsq00; |
963 | |
964 | fscal = fvdw; |
965 | |
966 | /* Calculate temporary vectorial force */ |
967 | tx = fscal*dx00; |
968 | ty = fscal*dy00; |
969 | tz = fscal*dz00; |
970 | |
971 | /* Update vectorial force */ |
972 | fix0 += tx; |
973 | fiy0 += ty; |
974 | fiz0 += tz; |
975 | f[j_coord_offset+DIM3*0+XX0] -= tx; |
976 | f[j_coord_offset+DIM3*0+YY1] -= ty; |
977 | f[j_coord_offset+DIM3*0+ZZ2] -= tz; |
978 | |
979 | /************************** |
980 | * CALCULATE INTERACTIONS * |
981 | **************************/ |
982 | |
983 | r11 = rsq11*rinv11; |
984 | |
985 | /* Calculate table index by multiplying r with table scale and truncate to integer */ |
986 | rt = r11*vftabscale; |
987 | vfitab = rt; |
988 | vfeps = rt-vfitab; |
989 | vfitab = 1*4*vfitab; |
990 | |
991 | /* CUBIC SPLINE TABLE ELECTROSTATICS */ |
992 | F = vftab[vfitab+1]; |
993 | Geps = vfeps*vftab[vfitab+2]; |
994 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
995 | Fp = F+Geps+Heps2; |
996 | FF = Fp+Geps+2.0*Heps2; |
997 | felec = -qq11*FF*vftabscale*rinv11; |
998 | |
999 | fscal = felec; |
1000 | |
1001 | /* Calculate temporary vectorial force */ |
1002 | tx = fscal*dx11; |
1003 | ty = fscal*dy11; |
1004 | tz = fscal*dz11; |
1005 | |
1006 | /* Update vectorial force */ |
1007 | fix1 += tx; |
1008 | fiy1 += ty; |
1009 | fiz1 += tz; |
1010 | f[j_coord_offset+DIM3*1+XX0] -= tx; |
1011 | f[j_coord_offset+DIM3*1+YY1] -= ty; |
1012 | f[j_coord_offset+DIM3*1+ZZ2] -= tz; |
1013 | |
1014 | /************************** |
1015 | * CALCULATE INTERACTIONS * |
1016 | **************************/ |
1017 | |
1018 | r12 = rsq12*rinv12; |
1019 | |
1020 | /* Calculate table index by multiplying r with table scale and truncate to integer */ |
1021 | rt = r12*vftabscale; |
1022 | vfitab = rt; |
1023 | vfeps = rt-vfitab; |
1024 | vfitab = 1*4*vfitab; |
1025 | |
1026 | /* CUBIC SPLINE TABLE ELECTROSTATICS */ |
1027 | F = vftab[vfitab+1]; |
1028 | Geps = vfeps*vftab[vfitab+2]; |
1029 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
1030 | Fp = F+Geps+Heps2; |
1031 | FF = Fp+Geps+2.0*Heps2; |
1032 | felec = -qq12*FF*vftabscale*rinv12; |
1033 | |
1034 | fscal = felec; |
1035 | |
1036 | /* Calculate temporary vectorial force */ |
1037 | tx = fscal*dx12; |
1038 | ty = fscal*dy12; |
1039 | tz = fscal*dz12; |
1040 | |
1041 | /* Update vectorial force */ |
1042 | fix1 += tx; |
1043 | fiy1 += ty; |
1044 | fiz1 += tz; |
1045 | f[j_coord_offset+DIM3*2+XX0] -= tx; |
1046 | f[j_coord_offset+DIM3*2+YY1] -= ty; |
1047 | f[j_coord_offset+DIM3*2+ZZ2] -= tz; |
1048 | |
1049 | /************************** |
1050 | * CALCULATE INTERACTIONS * |
1051 | **************************/ |
1052 | |
1053 | r13 = rsq13*rinv13; |
1054 | |
1055 | /* Calculate table index by multiplying r with table scale and truncate to integer */ |
1056 | rt = r13*vftabscale; |
1057 | vfitab = rt; |
1058 | vfeps = rt-vfitab; |
1059 | vfitab = 1*4*vfitab; |
1060 | |
1061 | /* CUBIC SPLINE TABLE ELECTROSTATICS */ |
1062 | F = vftab[vfitab+1]; |
1063 | Geps = vfeps*vftab[vfitab+2]; |
1064 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
1065 | Fp = F+Geps+Heps2; |
1066 | FF = Fp+Geps+2.0*Heps2; |
1067 | felec = -qq13*FF*vftabscale*rinv13; |
1068 | |
1069 | fscal = felec; |
1070 | |
1071 | /* Calculate temporary vectorial force */ |
1072 | tx = fscal*dx13; |
1073 | ty = fscal*dy13; |
1074 | tz = fscal*dz13; |
1075 | |
1076 | /* Update vectorial force */ |
1077 | fix1 += tx; |
1078 | fiy1 += ty; |
1079 | fiz1 += tz; |
1080 | f[j_coord_offset+DIM3*3+XX0] -= tx; |
1081 | f[j_coord_offset+DIM3*3+YY1] -= ty; |
1082 | f[j_coord_offset+DIM3*3+ZZ2] -= tz; |
1083 | |
1084 | /************************** |
1085 | * CALCULATE INTERACTIONS * |
1086 | **************************/ |
1087 | |
1088 | r21 = rsq21*rinv21; |
1089 | |
1090 | /* Calculate table index by multiplying r with table scale and truncate to integer */ |
1091 | rt = r21*vftabscale; |
1092 | vfitab = rt; |
1093 | vfeps = rt-vfitab; |
1094 | vfitab = 1*4*vfitab; |
1095 | |
1096 | /* CUBIC SPLINE TABLE ELECTROSTATICS */ |
1097 | F = vftab[vfitab+1]; |
1098 | Geps = vfeps*vftab[vfitab+2]; |
1099 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
1100 | Fp = F+Geps+Heps2; |
1101 | FF = Fp+Geps+2.0*Heps2; |
1102 | felec = -qq21*FF*vftabscale*rinv21; |
1103 | |
1104 | fscal = felec; |
1105 | |
1106 | /* Calculate temporary vectorial force */ |
1107 | tx = fscal*dx21; |
1108 | ty = fscal*dy21; |
1109 | tz = fscal*dz21; |
1110 | |
1111 | /* Update vectorial force */ |
1112 | fix2 += tx; |
1113 | fiy2 += ty; |
1114 | fiz2 += tz; |
1115 | f[j_coord_offset+DIM3*1+XX0] -= tx; |
1116 | f[j_coord_offset+DIM3*1+YY1] -= ty; |
1117 | f[j_coord_offset+DIM3*1+ZZ2] -= tz; |
1118 | |
1119 | /************************** |
1120 | * CALCULATE INTERACTIONS * |
1121 | **************************/ |
1122 | |
1123 | r22 = rsq22*rinv22; |
1124 | |
1125 | /* Calculate table index by multiplying r with table scale and truncate to integer */ |
1126 | rt = r22*vftabscale; |
1127 | vfitab = rt; |
1128 | vfeps = rt-vfitab; |
1129 | vfitab = 1*4*vfitab; |
1130 | |
1131 | /* CUBIC SPLINE TABLE ELECTROSTATICS */ |
1132 | F = vftab[vfitab+1]; |
1133 | Geps = vfeps*vftab[vfitab+2]; |
1134 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
1135 | Fp = F+Geps+Heps2; |
1136 | FF = Fp+Geps+2.0*Heps2; |
1137 | felec = -qq22*FF*vftabscale*rinv22; |
1138 | |
1139 | fscal = felec; |
1140 | |
1141 | /* Calculate temporary vectorial force */ |
1142 | tx = fscal*dx22; |
1143 | ty = fscal*dy22; |
1144 | tz = fscal*dz22; |
1145 | |
1146 | /* Update vectorial force */ |
1147 | fix2 += tx; |
1148 | fiy2 += ty; |
1149 | fiz2 += tz; |
1150 | f[j_coord_offset+DIM3*2+XX0] -= tx; |
1151 | f[j_coord_offset+DIM3*2+YY1] -= ty; |
1152 | f[j_coord_offset+DIM3*2+ZZ2] -= tz; |
1153 | |
1154 | /************************** |
1155 | * CALCULATE INTERACTIONS * |
1156 | **************************/ |
1157 | |
1158 | r23 = rsq23*rinv23; |
1159 | |
1160 | /* Calculate table index by multiplying r with table scale and truncate to integer */ |
1161 | rt = r23*vftabscale; |
1162 | vfitab = rt; |
1163 | vfeps = rt-vfitab; |
1164 | vfitab = 1*4*vfitab; |
1165 | |
1166 | /* CUBIC SPLINE TABLE ELECTROSTATICS */ |
1167 | F = vftab[vfitab+1]; |
1168 | Geps = vfeps*vftab[vfitab+2]; |
1169 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
1170 | Fp = F+Geps+Heps2; |
1171 | FF = Fp+Geps+2.0*Heps2; |
1172 | felec = -qq23*FF*vftabscale*rinv23; |
1173 | |
1174 | fscal = felec; |
1175 | |
1176 | /* Calculate temporary vectorial force */ |
1177 | tx = fscal*dx23; |
1178 | ty = fscal*dy23; |
1179 | tz = fscal*dz23; |
1180 | |
1181 | /* Update vectorial force */ |
1182 | fix2 += tx; |
1183 | fiy2 += ty; |
1184 | fiz2 += tz; |
1185 | f[j_coord_offset+DIM3*3+XX0] -= tx; |
1186 | f[j_coord_offset+DIM3*3+YY1] -= ty; |
1187 | f[j_coord_offset+DIM3*3+ZZ2] -= tz; |
1188 | |
1189 | /************************** |
1190 | * CALCULATE INTERACTIONS * |
1191 | **************************/ |
1192 | |
1193 | r31 = rsq31*rinv31; |
1194 | |
1195 | /* Calculate table index by multiplying r with table scale and truncate to integer */ |
1196 | rt = r31*vftabscale; |
1197 | vfitab = rt; |
1198 | vfeps = rt-vfitab; |
1199 | vfitab = 1*4*vfitab; |
1200 | |
1201 | /* CUBIC SPLINE TABLE ELECTROSTATICS */ |
1202 | F = vftab[vfitab+1]; |
1203 | Geps = vfeps*vftab[vfitab+2]; |
1204 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
1205 | Fp = F+Geps+Heps2; |
1206 | FF = Fp+Geps+2.0*Heps2; |
1207 | felec = -qq31*FF*vftabscale*rinv31; |
1208 | |
1209 | fscal = felec; |
1210 | |
1211 | /* Calculate temporary vectorial force */ |
1212 | tx = fscal*dx31; |
1213 | ty = fscal*dy31; |
1214 | tz = fscal*dz31; |
1215 | |
1216 | /* Update vectorial force */ |
1217 | fix3 += tx; |
1218 | fiy3 += ty; |
1219 | fiz3 += tz; |
1220 | f[j_coord_offset+DIM3*1+XX0] -= tx; |
1221 | f[j_coord_offset+DIM3*1+YY1] -= ty; |
1222 | f[j_coord_offset+DIM3*1+ZZ2] -= tz; |
1223 | |
1224 | /************************** |
1225 | * CALCULATE INTERACTIONS * |
1226 | **************************/ |
1227 | |
1228 | r32 = rsq32*rinv32; |
1229 | |
1230 | /* Calculate table index by multiplying r with table scale and truncate to integer */ |
1231 | rt = r32*vftabscale; |
1232 | vfitab = rt; |
1233 | vfeps = rt-vfitab; |
1234 | vfitab = 1*4*vfitab; |
1235 | |
1236 | /* CUBIC SPLINE TABLE ELECTROSTATICS */ |
1237 | F = vftab[vfitab+1]; |
1238 | Geps = vfeps*vftab[vfitab+2]; |
1239 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
1240 | Fp = F+Geps+Heps2; |
1241 | FF = Fp+Geps+2.0*Heps2; |
1242 | felec = -qq32*FF*vftabscale*rinv32; |
1243 | |
1244 | fscal = felec; |
1245 | |
1246 | /* Calculate temporary vectorial force */ |
1247 | tx = fscal*dx32; |
1248 | ty = fscal*dy32; |
1249 | tz = fscal*dz32; |
1250 | |
1251 | /* Update vectorial force */ |
1252 | fix3 += tx; |
1253 | fiy3 += ty; |
1254 | fiz3 += tz; |
1255 | f[j_coord_offset+DIM3*2+XX0] -= tx; |
1256 | f[j_coord_offset+DIM3*2+YY1] -= ty; |
1257 | f[j_coord_offset+DIM3*2+ZZ2] -= tz; |
1258 | |
1259 | /************************** |
1260 | * CALCULATE INTERACTIONS * |
1261 | **************************/ |
1262 | |
1263 | r33 = rsq33*rinv33; |
1264 | |
1265 | /* Calculate table index by multiplying r with table scale and truncate to integer */ |
1266 | rt = r33*vftabscale; |
1267 | vfitab = rt; |
1268 | vfeps = rt-vfitab; |
1269 | vfitab = 1*4*vfitab; |
1270 | |
1271 | /* CUBIC SPLINE TABLE ELECTROSTATICS */ |
1272 | F = vftab[vfitab+1]; |
1273 | Geps = vfeps*vftab[vfitab+2]; |
1274 | Heps2 = vfeps*vfeps*vftab[vfitab+3]; |
1275 | Fp = F+Geps+Heps2; |
1276 | FF = Fp+Geps+2.0*Heps2; |
1277 | felec = -qq33*FF*vftabscale*rinv33; |
1278 | |
1279 | fscal = felec; |
1280 | |
1281 | /* Calculate temporary vectorial force */ |
1282 | tx = fscal*dx33; |
1283 | ty = fscal*dy33; |
1284 | tz = fscal*dz33; |
1285 | |
1286 | /* Update vectorial force */ |
1287 | fix3 += tx; |
1288 | fiy3 += ty; |
1289 | fiz3 += tz; |
1290 | f[j_coord_offset+DIM3*3+XX0] -= tx; |
1291 | f[j_coord_offset+DIM3*3+YY1] -= ty; |
1292 | f[j_coord_offset+DIM3*3+ZZ2] -= tz; |
1293 | |
1294 | /* Inner loop uses 360 flops */ |
1295 | } |
1296 | /* End of innermost loop */ |
1297 | |
1298 | tx = ty = tz = 0; |
1299 | f[i_coord_offset+DIM3*0+XX0] += fix0; |
1300 | f[i_coord_offset+DIM3*0+YY1] += fiy0; |
1301 | f[i_coord_offset+DIM3*0+ZZ2] += fiz0; |
1302 | tx += fix0; |
1303 | ty += fiy0; |
1304 | tz += fiz0; |
1305 | f[i_coord_offset+DIM3*1+XX0] += fix1; |
1306 | f[i_coord_offset+DIM3*1+YY1] += fiy1; |
1307 | f[i_coord_offset+DIM3*1+ZZ2] += fiz1; |
1308 | tx += fix1; |
1309 | ty += fiy1; |
1310 | tz += fiz1; |
1311 | f[i_coord_offset+DIM3*2+XX0] += fix2; |
1312 | f[i_coord_offset+DIM3*2+YY1] += fiy2; |
1313 | f[i_coord_offset+DIM3*2+ZZ2] += fiz2; |
1314 | tx += fix2; |
1315 | ty += fiy2; |
1316 | tz += fiz2; |
1317 | f[i_coord_offset+DIM3*3+XX0] += fix3; |
1318 | f[i_coord_offset+DIM3*3+YY1] += fiy3; |
1319 | f[i_coord_offset+DIM3*3+ZZ2] += fiz3; |
1320 | tx += fix3; |
1321 | ty += fiy3; |
1322 | tz += fiz3; |
1323 | fshift[i_shift_offset+XX0] += tx; |
1324 | fshift[i_shift_offset+YY1] += ty; |
1325 | fshift[i_shift_offset+ZZ2] += tz; |
1326 | |
1327 | /* Increment number of inner iterations */ |
1328 | inneriter += j_index_end - j_index_start; |
1329 | |
1330 | /* Outer loop uses 39 flops */ |
1331 | } |
1332 | |
1333 | /* Increment number of outer iterations */ |
1334 | outeriter += nri; |
1335 | |
1336 | /* Update outer/inner flops */ |
1337 | |
1338 | inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4W4_F,outeriter*39 + inneriter*360)(nrnb)->n[eNR_NBKERNEL_ELEC_VDW_W4W4_F] += outeriter*39 + inneriter *360; |
1339 | } |