File: | gromacs/gmxlib/nonbonded/nb_kernel_c/nb_kernel_ElecEwSw_VdwLJSw_GeomW4W4_c.c |
Location: | line 942, 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_ElecEwSw_VdwLJSw_GeomW4W4_VF_c |
51 | * Electrostatics interaction: Ewald |
52 | * VdW interaction: LennardJones |
53 | * Geometry: Water4-Water4 |
54 | * Calculate force/pot: PotentialAndForce |
55 | */ |
56 | void |
57 | nb_kernel_ElecEwSw_VdwLJSw_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 ewitab; |
105 | real ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace; |
106 | real *ewtab; |
107 | real rswitch,swV3,swV4,swV5,swF2,swF3,swF4,d,d2,sw,dsw; |
108 | |
109 | x = xx[0]; |
110 | f = ff[0]; |
111 | |
112 | nri = nlist->nri; |
113 | iinr = nlist->iinr; |
114 | jindex = nlist->jindex; |
115 | jjnr = nlist->jjnr; |
116 | shiftidx = nlist->shift; |
117 | gid = nlist->gid; |
118 | shiftvec = fr->shift_vec[0]; |
119 | fshift = fr->fshift[0]; |
120 | facel = fr->epsfac; |
121 | charge = mdatoms->chargeA; |
122 | nvdwtype = fr->ntype; |
123 | vdwparam = fr->nbfp; |
124 | vdwtype = mdatoms->typeA; |
125 | |
126 | sh_ewald = fr->ic->sh_ewald; |
127 | ewtab = fr->ic->tabq_coul_FDV0; |
128 | ewtabscale = fr->ic->tabq_scale; |
129 | ewtabhalfspace = 0.5/ewtabscale; |
130 | |
131 | /* Setup water-specific parameters */ |
132 | inr = nlist->iinr[0]; |
133 | iq1 = facel*charge[inr+1]; |
134 | iq2 = facel*charge[inr+2]; |
135 | iq3 = facel*charge[inr+3]; |
136 | vdwioffset0 = 2*nvdwtype*vdwtype[inr+0]; |
137 | |
138 | jq1 = charge[inr+1]; |
139 | jq2 = charge[inr+2]; |
140 | jq3 = charge[inr+3]; |
141 | vdwjidx0 = 2*vdwtype[inr+0]; |
142 | c6_00 = vdwparam[vdwioffset0+vdwjidx0]; |
143 | c12_00 = vdwparam[vdwioffset0+vdwjidx0+1]; |
144 | qq11 = iq1*jq1; |
145 | qq12 = iq1*jq2; |
146 | qq13 = iq1*jq3; |
147 | qq21 = iq2*jq1; |
148 | qq22 = iq2*jq2; |
149 | qq23 = iq2*jq3; |
150 | qq31 = iq3*jq1; |
151 | qq32 = iq3*jq2; |
152 | qq33 = iq3*jq3; |
153 | |
154 | /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */ |
155 | rcutoff = fr->rcoulomb; |
156 | rcutoff2 = rcutoff*rcutoff; |
157 | |
158 | rswitch = fr->rcoulomb_switch; |
159 | /* Setup switch parameters */ |
160 | d = rcutoff-rswitch; |
161 | swV3 = -10.0/(d*d*d); |
162 | swV4 = 15.0/(d*d*d*d); |
163 | swV5 = -6.0/(d*d*d*d*d); |
164 | swF2 = -30.0/(d*d*d); |
165 | swF3 = 60.0/(d*d*d*d); |
166 | swF4 = -30.0/(d*d*d*d*d); |
167 | |
168 | outeriter = 0; |
169 | inneriter = 0; |
170 | |
171 | /* Start outer loop over neighborlists */ |
172 | for(iidx=0; iidx<nri; iidx++) |
173 | { |
174 | /* Load shift vector for this list */ |
175 | i_shift_offset = DIM3*shiftidx[iidx]; |
176 | shX = shiftvec[i_shift_offset+XX0]; |
177 | shY = shiftvec[i_shift_offset+YY1]; |
178 | shZ = shiftvec[i_shift_offset+ZZ2]; |
179 | |
180 | /* Load limits for loop over neighbors */ |
181 | j_index_start = jindex[iidx]; |
182 | j_index_end = jindex[iidx+1]; |
183 | |
184 | /* Get outer coordinate index */ |
185 | inr = iinr[iidx]; |
186 | i_coord_offset = DIM3*inr; |
187 | |
188 | /* Load i particle coords and add shift vector */ |
189 | ix0 = shX + x[i_coord_offset+DIM3*0+XX0]; |
190 | iy0 = shY + x[i_coord_offset+DIM3*0+YY1]; |
191 | iz0 = shZ + x[i_coord_offset+DIM3*0+ZZ2]; |
192 | ix1 = shX + x[i_coord_offset+DIM3*1+XX0]; |
193 | iy1 = shY + x[i_coord_offset+DIM3*1+YY1]; |
194 | iz1 = shZ + x[i_coord_offset+DIM3*1+ZZ2]; |
195 | ix2 = shX + x[i_coord_offset+DIM3*2+XX0]; |
196 | iy2 = shY + x[i_coord_offset+DIM3*2+YY1]; |
197 | iz2 = shZ + x[i_coord_offset+DIM3*2+ZZ2]; |
198 | ix3 = shX + x[i_coord_offset+DIM3*3+XX0]; |
199 | iy3 = shY + x[i_coord_offset+DIM3*3+YY1]; |
200 | iz3 = shZ + x[i_coord_offset+DIM3*3+ZZ2]; |
201 | |
202 | fix0 = 0.0; |
203 | fiy0 = 0.0; |
204 | fiz0 = 0.0; |
205 | fix1 = 0.0; |
206 | fiy1 = 0.0; |
207 | fiz1 = 0.0; |
208 | fix2 = 0.0; |
209 | fiy2 = 0.0; |
210 | fiz2 = 0.0; |
211 | fix3 = 0.0; |
212 | fiy3 = 0.0; |
213 | fiz3 = 0.0; |
214 | |
215 | /* Reset potential sums */ |
216 | velecsum = 0.0; |
217 | vvdwsum = 0.0; |
218 | |
219 | /* Start inner kernel loop */ |
220 | for(jidx=j_index_start; jidx<j_index_end; jidx++) |
221 | { |
222 | /* Get j neighbor index, and coordinate index */ |
223 | jnr = jjnr[jidx]; |
224 | j_coord_offset = DIM3*jnr; |
225 | |
226 | /* load j atom coordinates */ |
227 | jx0 = x[j_coord_offset+DIM3*0+XX0]; |
228 | jy0 = x[j_coord_offset+DIM3*0+YY1]; |
229 | jz0 = x[j_coord_offset+DIM3*0+ZZ2]; |
230 | jx1 = x[j_coord_offset+DIM3*1+XX0]; |
231 | jy1 = x[j_coord_offset+DIM3*1+YY1]; |
232 | jz1 = x[j_coord_offset+DIM3*1+ZZ2]; |
233 | jx2 = x[j_coord_offset+DIM3*2+XX0]; |
234 | jy2 = x[j_coord_offset+DIM3*2+YY1]; |
235 | jz2 = x[j_coord_offset+DIM3*2+ZZ2]; |
236 | jx3 = x[j_coord_offset+DIM3*3+XX0]; |
237 | jy3 = x[j_coord_offset+DIM3*3+YY1]; |
238 | jz3 = x[j_coord_offset+DIM3*3+ZZ2]; |
239 | |
240 | /* Calculate displacement vector */ |
241 | dx00 = ix0 - jx0; |
242 | dy00 = iy0 - jy0; |
243 | dz00 = iz0 - jz0; |
244 | dx11 = ix1 - jx1; |
245 | dy11 = iy1 - jy1; |
246 | dz11 = iz1 - jz1; |
247 | dx12 = ix1 - jx2; |
248 | dy12 = iy1 - jy2; |
249 | dz12 = iz1 - jz2; |
250 | dx13 = ix1 - jx3; |
251 | dy13 = iy1 - jy3; |
252 | dz13 = iz1 - jz3; |
253 | dx21 = ix2 - jx1; |
254 | dy21 = iy2 - jy1; |
255 | dz21 = iz2 - jz1; |
256 | dx22 = ix2 - jx2; |
257 | dy22 = iy2 - jy2; |
258 | dz22 = iz2 - jz2; |
259 | dx23 = ix2 - jx3; |
260 | dy23 = iy2 - jy3; |
261 | dz23 = iz2 - jz3; |
262 | dx31 = ix3 - jx1; |
263 | dy31 = iy3 - jy1; |
264 | dz31 = iz3 - jz1; |
265 | dx32 = ix3 - jx2; |
266 | dy32 = iy3 - jy2; |
267 | dz32 = iz3 - jz2; |
268 | dx33 = ix3 - jx3; |
269 | dy33 = iy3 - jy3; |
270 | dz33 = iz3 - jz3; |
271 | |
272 | /* Calculate squared distance and things based on it */ |
273 | rsq00 = dx00*dx00+dy00*dy00+dz00*dz00; |
274 | rsq11 = dx11*dx11+dy11*dy11+dz11*dz11; |
275 | rsq12 = dx12*dx12+dy12*dy12+dz12*dz12; |
276 | rsq13 = dx13*dx13+dy13*dy13+dz13*dz13; |
277 | rsq21 = dx21*dx21+dy21*dy21+dz21*dz21; |
278 | rsq22 = dx22*dx22+dy22*dy22+dz22*dz22; |
279 | rsq23 = dx23*dx23+dy23*dy23+dz23*dz23; |
280 | rsq31 = dx31*dx31+dy31*dy31+dz31*dz31; |
281 | rsq32 = dx32*dx32+dy32*dy32+dz32*dz32; |
282 | rsq33 = dx33*dx33+dy33*dy33+dz33*dz33; |
283 | |
284 | rinv00 = gmx_invsqrt(rsq00)gmx_software_invsqrt(rsq00); |
285 | rinv11 = gmx_invsqrt(rsq11)gmx_software_invsqrt(rsq11); |
286 | rinv12 = gmx_invsqrt(rsq12)gmx_software_invsqrt(rsq12); |
287 | rinv13 = gmx_invsqrt(rsq13)gmx_software_invsqrt(rsq13); |
288 | rinv21 = gmx_invsqrt(rsq21)gmx_software_invsqrt(rsq21); |
289 | rinv22 = gmx_invsqrt(rsq22)gmx_software_invsqrt(rsq22); |
290 | rinv23 = gmx_invsqrt(rsq23)gmx_software_invsqrt(rsq23); |
291 | rinv31 = gmx_invsqrt(rsq31)gmx_software_invsqrt(rsq31); |
292 | rinv32 = gmx_invsqrt(rsq32)gmx_software_invsqrt(rsq32); |
293 | rinv33 = gmx_invsqrt(rsq33)gmx_software_invsqrt(rsq33); |
294 | |
295 | rinvsq00 = rinv00*rinv00; |
296 | rinvsq11 = rinv11*rinv11; |
297 | rinvsq12 = rinv12*rinv12; |
298 | rinvsq13 = rinv13*rinv13; |
299 | rinvsq21 = rinv21*rinv21; |
300 | rinvsq22 = rinv22*rinv22; |
301 | rinvsq23 = rinv23*rinv23; |
302 | rinvsq31 = rinv31*rinv31; |
303 | rinvsq32 = rinv32*rinv32; |
304 | rinvsq33 = rinv33*rinv33; |
305 | |
306 | /************************** |
307 | * CALCULATE INTERACTIONS * |
308 | **************************/ |
309 | |
310 | if (rsq00<rcutoff2) |
311 | { |
312 | |
313 | r00 = rsq00*rinv00; |
314 | |
315 | /* LENNARD-JONES DISPERSION/REPULSION */ |
316 | |
317 | rinvsix = rinvsq00*rinvsq00*rinvsq00; |
318 | vvdw6 = c6_00*rinvsix; |
319 | vvdw12 = c12_00*rinvsix*rinvsix; |
320 | vvdw = vvdw12*(1.0/12.0) - vvdw6*(1.0/6.0); |
321 | fvdw = (vvdw12-vvdw6)*rinvsq00; |
322 | |
323 | d = r00-rswitch; |
324 | d = (d>0.0) ? d : 0.0; |
325 | d2 = d*d; |
326 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
327 | |
328 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
329 | |
330 | /* Evaluate switch function */ |
331 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
332 | fvdw = fvdw*sw - rinv00*vvdw*dsw; |
333 | vvdw *= sw; |
334 | |
335 | /* Update potential sums from outer loop */ |
336 | vvdwsum += vvdw; |
337 | |
338 | fscal = fvdw; |
339 | |
340 | /* Calculate temporary vectorial force */ |
341 | tx = fscal*dx00; |
342 | ty = fscal*dy00; |
343 | tz = fscal*dz00; |
344 | |
345 | /* Update vectorial force */ |
346 | fix0 += tx; |
347 | fiy0 += ty; |
348 | fiz0 += tz; |
349 | f[j_coord_offset+DIM3*0+XX0] -= tx; |
350 | f[j_coord_offset+DIM3*0+YY1] -= ty; |
351 | f[j_coord_offset+DIM3*0+ZZ2] -= tz; |
352 | |
353 | } |
354 | |
355 | /************************** |
356 | * CALCULATE INTERACTIONS * |
357 | **************************/ |
358 | |
359 | if (rsq11<rcutoff2) |
360 | { |
361 | |
362 | r11 = rsq11*rinv11; |
363 | |
364 | /* EWALD ELECTROSTATICS */ |
365 | |
366 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
367 | ewrt = r11*ewtabscale; |
368 | ewitab = ewrt; |
369 | eweps = ewrt-ewitab; |
370 | ewitab = 4*ewitab; |
371 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
372 | velec = qq11*(rinv11-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
373 | felec = qq11*rinv11*(rinvsq11-felec); |
374 | |
375 | d = r11-rswitch; |
376 | d = (d>0.0) ? d : 0.0; |
377 | d2 = d*d; |
378 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
379 | |
380 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
381 | |
382 | /* Evaluate switch function */ |
383 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
384 | felec = felec*sw - rinv11*velec*dsw; |
385 | velec *= sw; |
386 | |
387 | /* Update potential sums from outer loop */ |
388 | velecsum += velec; |
389 | |
390 | fscal = felec; |
391 | |
392 | /* Calculate temporary vectorial force */ |
393 | tx = fscal*dx11; |
394 | ty = fscal*dy11; |
395 | tz = fscal*dz11; |
396 | |
397 | /* Update vectorial force */ |
398 | fix1 += tx; |
399 | fiy1 += ty; |
400 | fiz1 += tz; |
401 | f[j_coord_offset+DIM3*1+XX0] -= tx; |
402 | f[j_coord_offset+DIM3*1+YY1] -= ty; |
403 | f[j_coord_offset+DIM3*1+ZZ2] -= tz; |
404 | |
405 | } |
406 | |
407 | /************************** |
408 | * CALCULATE INTERACTIONS * |
409 | **************************/ |
410 | |
411 | if (rsq12<rcutoff2) |
412 | { |
413 | |
414 | r12 = rsq12*rinv12; |
415 | |
416 | /* EWALD ELECTROSTATICS */ |
417 | |
418 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
419 | ewrt = r12*ewtabscale; |
420 | ewitab = ewrt; |
421 | eweps = ewrt-ewitab; |
422 | ewitab = 4*ewitab; |
423 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
424 | velec = qq12*(rinv12-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
425 | felec = qq12*rinv12*(rinvsq12-felec); |
426 | |
427 | d = r12-rswitch; |
428 | d = (d>0.0) ? d : 0.0; |
429 | d2 = d*d; |
430 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
431 | |
432 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
433 | |
434 | /* Evaluate switch function */ |
435 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
436 | felec = felec*sw - rinv12*velec*dsw; |
437 | velec *= sw; |
438 | |
439 | /* Update potential sums from outer loop */ |
440 | velecsum += velec; |
441 | |
442 | fscal = felec; |
443 | |
444 | /* Calculate temporary vectorial force */ |
445 | tx = fscal*dx12; |
446 | ty = fscal*dy12; |
447 | tz = fscal*dz12; |
448 | |
449 | /* Update vectorial force */ |
450 | fix1 += tx; |
451 | fiy1 += ty; |
452 | fiz1 += tz; |
453 | f[j_coord_offset+DIM3*2+XX0] -= tx; |
454 | f[j_coord_offset+DIM3*2+YY1] -= ty; |
455 | f[j_coord_offset+DIM3*2+ZZ2] -= tz; |
456 | |
457 | } |
458 | |
459 | /************************** |
460 | * CALCULATE INTERACTIONS * |
461 | **************************/ |
462 | |
463 | if (rsq13<rcutoff2) |
464 | { |
465 | |
466 | r13 = rsq13*rinv13; |
467 | |
468 | /* EWALD ELECTROSTATICS */ |
469 | |
470 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
471 | ewrt = r13*ewtabscale; |
472 | ewitab = ewrt; |
473 | eweps = ewrt-ewitab; |
474 | ewitab = 4*ewitab; |
475 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
476 | velec = qq13*(rinv13-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
477 | felec = qq13*rinv13*(rinvsq13-felec); |
478 | |
479 | d = r13-rswitch; |
480 | d = (d>0.0) ? d : 0.0; |
481 | d2 = d*d; |
482 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
483 | |
484 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
485 | |
486 | /* Evaluate switch function */ |
487 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
488 | felec = felec*sw - rinv13*velec*dsw; |
489 | velec *= sw; |
490 | |
491 | /* Update potential sums from outer loop */ |
492 | velecsum += velec; |
493 | |
494 | fscal = felec; |
495 | |
496 | /* Calculate temporary vectorial force */ |
497 | tx = fscal*dx13; |
498 | ty = fscal*dy13; |
499 | tz = fscal*dz13; |
500 | |
501 | /* Update vectorial force */ |
502 | fix1 += tx; |
503 | fiy1 += ty; |
504 | fiz1 += tz; |
505 | f[j_coord_offset+DIM3*3+XX0] -= tx; |
506 | f[j_coord_offset+DIM3*3+YY1] -= ty; |
507 | f[j_coord_offset+DIM3*3+ZZ2] -= tz; |
508 | |
509 | } |
510 | |
511 | /************************** |
512 | * CALCULATE INTERACTIONS * |
513 | **************************/ |
514 | |
515 | if (rsq21<rcutoff2) |
516 | { |
517 | |
518 | r21 = rsq21*rinv21; |
519 | |
520 | /* EWALD ELECTROSTATICS */ |
521 | |
522 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
523 | ewrt = r21*ewtabscale; |
524 | ewitab = ewrt; |
525 | eweps = ewrt-ewitab; |
526 | ewitab = 4*ewitab; |
527 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
528 | velec = qq21*(rinv21-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
529 | felec = qq21*rinv21*(rinvsq21-felec); |
530 | |
531 | d = r21-rswitch; |
532 | d = (d>0.0) ? d : 0.0; |
533 | d2 = d*d; |
534 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
535 | |
536 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
537 | |
538 | /* Evaluate switch function */ |
539 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
540 | felec = felec*sw - rinv21*velec*dsw; |
541 | velec *= sw; |
542 | |
543 | /* Update potential sums from outer loop */ |
544 | velecsum += velec; |
545 | |
546 | fscal = felec; |
547 | |
548 | /* Calculate temporary vectorial force */ |
549 | tx = fscal*dx21; |
550 | ty = fscal*dy21; |
551 | tz = fscal*dz21; |
552 | |
553 | /* Update vectorial force */ |
554 | fix2 += tx; |
555 | fiy2 += ty; |
556 | fiz2 += tz; |
557 | f[j_coord_offset+DIM3*1+XX0] -= tx; |
558 | f[j_coord_offset+DIM3*1+YY1] -= ty; |
559 | f[j_coord_offset+DIM3*1+ZZ2] -= tz; |
560 | |
561 | } |
562 | |
563 | /************************** |
564 | * CALCULATE INTERACTIONS * |
565 | **************************/ |
566 | |
567 | if (rsq22<rcutoff2) |
568 | { |
569 | |
570 | r22 = rsq22*rinv22; |
571 | |
572 | /* EWALD ELECTROSTATICS */ |
573 | |
574 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
575 | ewrt = r22*ewtabscale; |
576 | ewitab = ewrt; |
577 | eweps = ewrt-ewitab; |
578 | ewitab = 4*ewitab; |
579 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
580 | velec = qq22*(rinv22-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
581 | felec = qq22*rinv22*(rinvsq22-felec); |
582 | |
583 | d = r22-rswitch; |
584 | d = (d>0.0) ? d : 0.0; |
585 | d2 = d*d; |
586 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
587 | |
588 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
589 | |
590 | /* Evaluate switch function */ |
591 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
592 | felec = felec*sw - rinv22*velec*dsw; |
593 | velec *= sw; |
594 | |
595 | /* Update potential sums from outer loop */ |
596 | velecsum += velec; |
597 | |
598 | fscal = felec; |
599 | |
600 | /* Calculate temporary vectorial force */ |
601 | tx = fscal*dx22; |
602 | ty = fscal*dy22; |
603 | tz = fscal*dz22; |
604 | |
605 | /* Update vectorial force */ |
606 | fix2 += tx; |
607 | fiy2 += ty; |
608 | fiz2 += tz; |
609 | f[j_coord_offset+DIM3*2+XX0] -= tx; |
610 | f[j_coord_offset+DIM3*2+YY1] -= ty; |
611 | f[j_coord_offset+DIM3*2+ZZ2] -= tz; |
612 | |
613 | } |
614 | |
615 | /************************** |
616 | * CALCULATE INTERACTIONS * |
617 | **************************/ |
618 | |
619 | if (rsq23<rcutoff2) |
620 | { |
621 | |
622 | r23 = rsq23*rinv23; |
623 | |
624 | /* EWALD ELECTROSTATICS */ |
625 | |
626 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
627 | ewrt = r23*ewtabscale; |
628 | ewitab = ewrt; |
629 | eweps = ewrt-ewitab; |
630 | ewitab = 4*ewitab; |
631 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
632 | velec = qq23*(rinv23-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
633 | felec = qq23*rinv23*(rinvsq23-felec); |
634 | |
635 | d = r23-rswitch; |
636 | d = (d>0.0) ? d : 0.0; |
637 | d2 = d*d; |
638 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
639 | |
640 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
641 | |
642 | /* Evaluate switch function */ |
643 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
644 | felec = felec*sw - rinv23*velec*dsw; |
645 | velec *= sw; |
646 | |
647 | /* Update potential sums from outer loop */ |
648 | velecsum += velec; |
649 | |
650 | fscal = felec; |
651 | |
652 | /* Calculate temporary vectorial force */ |
653 | tx = fscal*dx23; |
654 | ty = fscal*dy23; |
655 | tz = fscal*dz23; |
656 | |
657 | /* Update vectorial force */ |
658 | fix2 += tx; |
659 | fiy2 += ty; |
660 | fiz2 += tz; |
661 | f[j_coord_offset+DIM3*3+XX0] -= tx; |
662 | f[j_coord_offset+DIM3*3+YY1] -= ty; |
663 | f[j_coord_offset+DIM3*3+ZZ2] -= tz; |
664 | |
665 | } |
666 | |
667 | /************************** |
668 | * CALCULATE INTERACTIONS * |
669 | **************************/ |
670 | |
671 | if (rsq31<rcutoff2) |
672 | { |
673 | |
674 | r31 = rsq31*rinv31; |
675 | |
676 | /* EWALD ELECTROSTATICS */ |
677 | |
678 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
679 | ewrt = r31*ewtabscale; |
680 | ewitab = ewrt; |
681 | eweps = ewrt-ewitab; |
682 | ewitab = 4*ewitab; |
683 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
684 | velec = qq31*(rinv31-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
685 | felec = qq31*rinv31*(rinvsq31-felec); |
686 | |
687 | d = r31-rswitch; |
688 | d = (d>0.0) ? d : 0.0; |
689 | d2 = d*d; |
690 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
691 | |
692 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
693 | |
694 | /* Evaluate switch function */ |
695 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
696 | felec = felec*sw - rinv31*velec*dsw; |
697 | velec *= sw; |
698 | |
699 | /* Update potential sums from outer loop */ |
700 | velecsum += velec; |
701 | |
702 | fscal = felec; |
703 | |
704 | /* Calculate temporary vectorial force */ |
705 | tx = fscal*dx31; |
706 | ty = fscal*dy31; |
707 | tz = fscal*dz31; |
708 | |
709 | /* Update vectorial force */ |
710 | fix3 += tx; |
711 | fiy3 += ty; |
712 | fiz3 += tz; |
713 | f[j_coord_offset+DIM3*1+XX0] -= tx; |
714 | f[j_coord_offset+DIM3*1+YY1] -= ty; |
715 | f[j_coord_offset+DIM3*1+ZZ2] -= tz; |
716 | |
717 | } |
718 | |
719 | /************************** |
720 | * CALCULATE INTERACTIONS * |
721 | **************************/ |
722 | |
723 | if (rsq32<rcutoff2) |
724 | { |
725 | |
726 | r32 = rsq32*rinv32; |
727 | |
728 | /* EWALD ELECTROSTATICS */ |
729 | |
730 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
731 | ewrt = r32*ewtabscale; |
732 | ewitab = ewrt; |
733 | eweps = ewrt-ewitab; |
734 | ewitab = 4*ewitab; |
735 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
736 | velec = qq32*(rinv32-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
737 | felec = qq32*rinv32*(rinvsq32-felec); |
738 | |
739 | d = r32-rswitch; |
740 | d = (d>0.0) ? d : 0.0; |
741 | d2 = d*d; |
742 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
743 | |
744 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
745 | |
746 | /* Evaluate switch function */ |
747 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
748 | felec = felec*sw - rinv32*velec*dsw; |
749 | velec *= sw; |
750 | |
751 | /* Update potential sums from outer loop */ |
752 | velecsum += velec; |
753 | |
754 | fscal = felec; |
755 | |
756 | /* Calculate temporary vectorial force */ |
757 | tx = fscal*dx32; |
758 | ty = fscal*dy32; |
759 | tz = fscal*dz32; |
760 | |
761 | /* Update vectorial force */ |
762 | fix3 += tx; |
763 | fiy3 += ty; |
764 | fiz3 += tz; |
765 | f[j_coord_offset+DIM3*2+XX0] -= tx; |
766 | f[j_coord_offset+DIM3*2+YY1] -= ty; |
767 | f[j_coord_offset+DIM3*2+ZZ2] -= tz; |
768 | |
769 | } |
770 | |
771 | /************************** |
772 | * CALCULATE INTERACTIONS * |
773 | **************************/ |
774 | |
775 | if (rsq33<rcutoff2) |
776 | { |
777 | |
778 | r33 = rsq33*rinv33; |
779 | |
780 | /* EWALD ELECTROSTATICS */ |
781 | |
782 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
783 | ewrt = r33*ewtabscale; |
784 | ewitab = ewrt; |
785 | eweps = ewrt-ewitab; |
786 | ewitab = 4*ewitab; |
787 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
788 | velec = qq33*(rinv33-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
789 | felec = qq33*rinv33*(rinvsq33-felec); |
790 | |
791 | d = r33-rswitch; |
792 | d = (d>0.0) ? d : 0.0; |
793 | d2 = d*d; |
794 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
795 | |
796 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
797 | |
798 | /* Evaluate switch function */ |
799 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
800 | felec = felec*sw - rinv33*velec*dsw; |
801 | velec *= sw; |
802 | |
803 | /* Update potential sums from outer loop */ |
804 | velecsum += velec; |
805 | |
806 | fscal = felec; |
807 | |
808 | /* Calculate temporary vectorial force */ |
809 | tx = fscal*dx33; |
810 | ty = fscal*dy33; |
811 | tz = fscal*dz33; |
812 | |
813 | /* Update vectorial force */ |
814 | fix3 += tx; |
815 | fiy3 += ty; |
816 | fiz3 += tz; |
817 | f[j_coord_offset+DIM3*3+XX0] -= tx; |
818 | f[j_coord_offset+DIM3*3+YY1] -= ty; |
819 | f[j_coord_offset+DIM3*3+ZZ2] -= tz; |
820 | |
821 | } |
822 | |
823 | /* Inner loop uses 575 flops */ |
824 | } |
825 | /* End of innermost loop */ |
826 | |
827 | tx = ty = tz = 0; |
828 | f[i_coord_offset+DIM3*0+XX0] += fix0; |
829 | f[i_coord_offset+DIM3*0+YY1] += fiy0; |
830 | f[i_coord_offset+DIM3*0+ZZ2] += fiz0; |
831 | tx += fix0; |
832 | ty += fiy0; |
833 | tz += fiz0; |
834 | f[i_coord_offset+DIM3*1+XX0] += fix1; |
835 | f[i_coord_offset+DIM3*1+YY1] += fiy1; |
836 | f[i_coord_offset+DIM3*1+ZZ2] += fiz1; |
837 | tx += fix1; |
838 | ty += fiy1; |
839 | tz += fiz1; |
840 | f[i_coord_offset+DIM3*2+XX0] += fix2; |
841 | f[i_coord_offset+DIM3*2+YY1] += fiy2; |
842 | f[i_coord_offset+DIM3*2+ZZ2] += fiz2; |
843 | tx += fix2; |
844 | ty += fiy2; |
845 | tz += fiz2; |
846 | f[i_coord_offset+DIM3*3+XX0] += fix3; |
847 | f[i_coord_offset+DIM3*3+YY1] += fiy3; |
848 | f[i_coord_offset+DIM3*3+ZZ2] += fiz3; |
849 | tx += fix3; |
850 | ty += fiy3; |
851 | tz += fiz3; |
852 | fshift[i_shift_offset+XX0] += tx; |
853 | fshift[i_shift_offset+YY1] += ty; |
854 | fshift[i_shift_offset+ZZ2] += tz; |
855 | |
856 | ggid = gid[iidx]; |
857 | /* Update potential energies */ |
858 | kernel_data->energygrp_elec[ggid] += velecsum; |
859 | kernel_data->energygrp_vdw[ggid] += vvdwsum; |
860 | |
861 | /* Increment number of inner iterations */ |
862 | inneriter += j_index_end - j_index_start; |
863 | |
864 | /* Outer loop uses 41 flops */ |
865 | } |
866 | |
867 | /* Increment number of outer iterations */ |
868 | outeriter += nri; |
869 | |
870 | /* Update outer/inner flops */ |
871 | |
872 | inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4W4_VF,outeriter*41 + inneriter*575)(nrnb)->n[eNR_NBKERNEL_ELEC_VDW_W4W4_VF] += outeriter*41 + inneriter*575; |
873 | } |
874 | /* |
875 | * Gromacs nonbonded kernel: nb_kernel_ElecEwSw_VdwLJSw_GeomW4W4_F_c |
876 | * Electrostatics interaction: Ewald |
877 | * VdW interaction: LennardJones |
878 | * Geometry: Water4-Water4 |
879 | * Calculate force/pot: Force |
880 | */ |
881 | void |
882 | nb_kernel_ElecEwSw_VdwLJSw_GeomW4W4_F_c |
883 | (t_nblist * gmx_restrict__restrict nlist, |
884 | rvec * gmx_restrict__restrict xx, |
885 | rvec * gmx_restrict__restrict ff, |
886 | t_forcerec * gmx_restrict__restrict fr, |
887 | t_mdatoms * gmx_restrict__restrict mdatoms, |
888 | nb_kernel_data_t gmx_unused__attribute__ ((unused)) * gmx_restrict__restrict kernel_data, |
889 | t_nrnb * gmx_restrict__restrict nrnb) |
890 | { |
891 | int i_shift_offset,i_coord_offset,j_coord_offset; |
892 | int j_index_start,j_index_end; |
893 | int nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter; |
894 | real shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2; |
895 | int *iinr,*jindex,*jjnr,*shiftidx,*gid; |
896 | real *shiftvec,*fshift,*x,*f; |
897 | int vdwioffset0; |
898 | real ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0; |
899 | int vdwioffset1; |
900 | real ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1; |
901 | int vdwioffset2; |
902 | real ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2; |
903 | int vdwioffset3; |
904 | real ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3; |
905 | int vdwjidx0; |
906 | real jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0; |
907 | int vdwjidx1; |
908 | real jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1; |
909 | int vdwjidx2; |
910 | real jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2; |
911 | int vdwjidx3; |
912 | real jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3; |
913 | real dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00; |
914 | real dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11,cexp1_11,cexp2_11; |
915 | real dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12,cexp1_12,cexp2_12; |
916 | real dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13,cexp1_13,cexp2_13; |
917 | real dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21,cexp1_21,cexp2_21; |
918 | real dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22,cexp1_22,cexp2_22; |
919 | real dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23,cexp1_23,cexp2_23; |
920 | real dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31,cexp1_31,cexp2_31; |
921 | real dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32,cexp1_32,cexp2_32; |
922 | real dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33,cexp1_33,cexp2_33; |
923 | real velec,felec,velecsum,facel,crf,krf,krf2; |
924 | real *charge; |
925 | int nvdwtype; |
926 | real rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6; |
927 | int *vdwtype; |
928 | real *vdwparam; |
929 | int ewitab; |
930 | real ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace; |
931 | real *ewtab; |
932 | real rswitch,swV3,swV4,swV5,swF2,swF3,swF4,d,d2,sw,dsw; |
933 | |
934 | x = xx[0]; |
935 | f = ff[0]; |
936 | |
937 | nri = nlist->nri; |
938 | iinr = nlist->iinr; |
939 | jindex = nlist->jindex; |
940 | jjnr = nlist->jjnr; |
941 | shiftidx = nlist->shift; |
942 | gid = nlist->gid; |
Value stored to 'gid' is never read | |
943 | shiftvec = fr->shift_vec[0]; |
944 | fshift = fr->fshift[0]; |
945 | facel = fr->epsfac; |
946 | charge = mdatoms->chargeA; |
947 | nvdwtype = fr->ntype; |
948 | vdwparam = fr->nbfp; |
949 | vdwtype = mdatoms->typeA; |
950 | |
951 | sh_ewald = fr->ic->sh_ewald; |
952 | ewtab = fr->ic->tabq_coul_FDV0; |
953 | ewtabscale = fr->ic->tabq_scale; |
954 | ewtabhalfspace = 0.5/ewtabscale; |
955 | |
956 | /* Setup water-specific parameters */ |
957 | inr = nlist->iinr[0]; |
958 | iq1 = facel*charge[inr+1]; |
959 | iq2 = facel*charge[inr+2]; |
960 | iq3 = facel*charge[inr+3]; |
961 | vdwioffset0 = 2*nvdwtype*vdwtype[inr+0]; |
962 | |
963 | jq1 = charge[inr+1]; |
964 | jq2 = charge[inr+2]; |
965 | jq3 = charge[inr+3]; |
966 | vdwjidx0 = 2*vdwtype[inr+0]; |
967 | c6_00 = vdwparam[vdwioffset0+vdwjidx0]; |
968 | c12_00 = vdwparam[vdwioffset0+vdwjidx0+1]; |
969 | qq11 = iq1*jq1; |
970 | qq12 = iq1*jq2; |
971 | qq13 = iq1*jq3; |
972 | qq21 = iq2*jq1; |
973 | qq22 = iq2*jq2; |
974 | qq23 = iq2*jq3; |
975 | qq31 = iq3*jq1; |
976 | qq32 = iq3*jq2; |
977 | qq33 = iq3*jq3; |
978 | |
979 | /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */ |
980 | rcutoff = fr->rcoulomb; |
981 | rcutoff2 = rcutoff*rcutoff; |
982 | |
983 | rswitch = fr->rcoulomb_switch; |
984 | /* Setup switch parameters */ |
985 | d = rcutoff-rswitch; |
986 | swV3 = -10.0/(d*d*d); |
987 | swV4 = 15.0/(d*d*d*d); |
988 | swV5 = -6.0/(d*d*d*d*d); |
989 | swF2 = -30.0/(d*d*d); |
990 | swF3 = 60.0/(d*d*d*d); |
991 | swF4 = -30.0/(d*d*d*d*d); |
992 | |
993 | outeriter = 0; |
994 | inneriter = 0; |
995 | |
996 | /* Start outer loop over neighborlists */ |
997 | for(iidx=0; iidx<nri; iidx++) |
998 | { |
999 | /* Load shift vector for this list */ |
1000 | i_shift_offset = DIM3*shiftidx[iidx]; |
1001 | shX = shiftvec[i_shift_offset+XX0]; |
1002 | shY = shiftvec[i_shift_offset+YY1]; |
1003 | shZ = shiftvec[i_shift_offset+ZZ2]; |
1004 | |
1005 | /* Load limits for loop over neighbors */ |
1006 | j_index_start = jindex[iidx]; |
1007 | j_index_end = jindex[iidx+1]; |
1008 | |
1009 | /* Get outer coordinate index */ |
1010 | inr = iinr[iidx]; |
1011 | i_coord_offset = DIM3*inr; |
1012 | |
1013 | /* Load i particle coords and add shift vector */ |
1014 | ix0 = shX + x[i_coord_offset+DIM3*0+XX0]; |
1015 | iy0 = shY + x[i_coord_offset+DIM3*0+YY1]; |
1016 | iz0 = shZ + x[i_coord_offset+DIM3*0+ZZ2]; |
1017 | ix1 = shX + x[i_coord_offset+DIM3*1+XX0]; |
1018 | iy1 = shY + x[i_coord_offset+DIM3*1+YY1]; |
1019 | iz1 = shZ + x[i_coord_offset+DIM3*1+ZZ2]; |
1020 | ix2 = shX + x[i_coord_offset+DIM3*2+XX0]; |
1021 | iy2 = shY + x[i_coord_offset+DIM3*2+YY1]; |
1022 | iz2 = shZ + x[i_coord_offset+DIM3*2+ZZ2]; |
1023 | ix3 = shX + x[i_coord_offset+DIM3*3+XX0]; |
1024 | iy3 = shY + x[i_coord_offset+DIM3*3+YY1]; |
1025 | iz3 = shZ + x[i_coord_offset+DIM3*3+ZZ2]; |
1026 | |
1027 | fix0 = 0.0; |
1028 | fiy0 = 0.0; |
1029 | fiz0 = 0.0; |
1030 | fix1 = 0.0; |
1031 | fiy1 = 0.0; |
1032 | fiz1 = 0.0; |
1033 | fix2 = 0.0; |
1034 | fiy2 = 0.0; |
1035 | fiz2 = 0.0; |
1036 | fix3 = 0.0; |
1037 | fiy3 = 0.0; |
1038 | fiz3 = 0.0; |
1039 | |
1040 | /* Start inner kernel loop */ |
1041 | for(jidx=j_index_start; jidx<j_index_end; jidx++) |
1042 | { |
1043 | /* Get j neighbor index, and coordinate index */ |
1044 | jnr = jjnr[jidx]; |
1045 | j_coord_offset = DIM3*jnr; |
1046 | |
1047 | /* load j atom coordinates */ |
1048 | jx0 = x[j_coord_offset+DIM3*0+XX0]; |
1049 | jy0 = x[j_coord_offset+DIM3*0+YY1]; |
1050 | jz0 = x[j_coord_offset+DIM3*0+ZZ2]; |
1051 | jx1 = x[j_coord_offset+DIM3*1+XX0]; |
1052 | jy1 = x[j_coord_offset+DIM3*1+YY1]; |
1053 | jz1 = x[j_coord_offset+DIM3*1+ZZ2]; |
1054 | jx2 = x[j_coord_offset+DIM3*2+XX0]; |
1055 | jy2 = x[j_coord_offset+DIM3*2+YY1]; |
1056 | jz2 = x[j_coord_offset+DIM3*2+ZZ2]; |
1057 | jx3 = x[j_coord_offset+DIM3*3+XX0]; |
1058 | jy3 = x[j_coord_offset+DIM3*3+YY1]; |
1059 | jz3 = x[j_coord_offset+DIM3*3+ZZ2]; |
1060 | |
1061 | /* Calculate displacement vector */ |
1062 | dx00 = ix0 - jx0; |
1063 | dy00 = iy0 - jy0; |
1064 | dz00 = iz0 - jz0; |
1065 | dx11 = ix1 - jx1; |
1066 | dy11 = iy1 - jy1; |
1067 | dz11 = iz1 - jz1; |
1068 | dx12 = ix1 - jx2; |
1069 | dy12 = iy1 - jy2; |
1070 | dz12 = iz1 - jz2; |
1071 | dx13 = ix1 - jx3; |
1072 | dy13 = iy1 - jy3; |
1073 | dz13 = iz1 - jz3; |
1074 | dx21 = ix2 - jx1; |
1075 | dy21 = iy2 - jy1; |
1076 | dz21 = iz2 - jz1; |
1077 | dx22 = ix2 - jx2; |
1078 | dy22 = iy2 - jy2; |
1079 | dz22 = iz2 - jz2; |
1080 | dx23 = ix2 - jx3; |
1081 | dy23 = iy2 - jy3; |
1082 | dz23 = iz2 - jz3; |
1083 | dx31 = ix3 - jx1; |
1084 | dy31 = iy3 - jy1; |
1085 | dz31 = iz3 - jz1; |
1086 | dx32 = ix3 - jx2; |
1087 | dy32 = iy3 - jy2; |
1088 | dz32 = iz3 - jz2; |
1089 | dx33 = ix3 - jx3; |
1090 | dy33 = iy3 - jy3; |
1091 | dz33 = iz3 - jz3; |
1092 | |
1093 | /* Calculate squared distance and things based on it */ |
1094 | rsq00 = dx00*dx00+dy00*dy00+dz00*dz00; |
1095 | rsq11 = dx11*dx11+dy11*dy11+dz11*dz11; |
1096 | rsq12 = dx12*dx12+dy12*dy12+dz12*dz12; |
1097 | rsq13 = dx13*dx13+dy13*dy13+dz13*dz13; |
1098 | rsq21 = dx21*dx21+dy21*dy21+dz21*dz21; |
1099 | rsq22 = dx22*dx22+dy22*dy22+dz22*dz22; |
1100 | rsq23 = dx23*dx23+dy23*dy23+dz23*dz23; |
1101 | rsq31 = dx31*dx31+dy31*dy31+dz31*dz31; |
1102 | rsq32 = dx32*dx32+dy32*dy32+dz32*dz32; |
1103 | rsq33 = dx33*dx33+dy33*dy33+dz33*dz33; |
1104 | |
1105 | rinv00 = gmx_invsqrt(rsq00)gmx_software_invsqrt(rsq00); |
1106 | rinv11 = gmx_invsqrt(rsq11)gmx_software_invsqrt(rsq11); |
1107 | rinv12 = gmx_invsqrt(rsq12)gmx_software_invsqrt(rsq12); |
1108 | rinv13 = gmx_invsqrt(rsq13)gmx_software_invsqrt(rsq13); |
1109 | rinv21 = gmx_invsqrt(rsq21)gmx_software_invsqrt(rsq21); |
1110 | rinv22 = gmx_invsqrt(rsq22)gmx_software_invsqrt(rsq22); |
1111 | rinv23 = gmx_invsqrt(rsq23)gmx_software_invsqrt(rsq23); |
1112 | rinv31 = gmx_invsqrt(rsq31)gmx_software_invsqrt(rsq31); |
1113 | rinv32 = gmx_invsqrt(rsq32)gmx_software_invsqrt(rsq32); |
1114 | rinv33 = gmx_invsqrt(rsq33)gmx_software_invsqrt(rsq33); |
1115 | |
1116 | rinvsq00 = rinv00*rinv00; |
1117 | rinvsq11 = rinv11*rinv11; |
1118 | rinvsq12 = rinv12*rinv12; |
1119 | rinvsq13 = rinv13*rinv13; |
1120 | rinvsq21 = rinv21*rinv21; |
1121 | rinvsq22 = rinv22*rinv22; |
1122 | rinvsq23 = rinv23*rinv23; |
1123 | rinvsq31 = rinv31*rinv31; |
1124 | rinvsq32 = rinv32*rinv32; |
1125 | rinvsq33 = rinv33*rinv33; |
1126 | |
1127 | /************************** |
1128 | * CALCULATE INTERACTIONS * |
1129 | **************************/ |
1130 | |
1131 | if (rsq00<rcutoff2) |
1132 | { |
1133 | |
1134 | r00 = rsq00*rinv00; |
1135 | |
1136 | /* LENNARD-JONES DISPERSION/REPULSION */ |
1137 | |
1138 | rinvsix = rinvsq00*rinvsq00*rinvsq00; |
1139 | vvdw6 = c6_00*rinvsix; |
1140 | vvdw12 = c12_00*rinvsix*rinvsix; |
1141 | vvdw = vvdw12*(1.0/12.0) - vvdw6*(1.0/6.0); |
1142 | fvdw = (vvdw12-vvdw6)*rinvsq00; |
1143 | |
1144 | d = r00-rswitch; |
1145 | d = (d>0.0) ? d : 0.0; |
1146 | d2 = d*d; |
1147 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
1148 | |
1149 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
1150 | |
1151 | /* Evaluate switch function */ |
1152 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
1153 | fvdw = fvdw*sw - rinv00*vvdw*dsw; |
1154 | |
1155 | fscal = fvdw; |
1156 | |
1157 | /* Calculate temporary vectorial force */ |
1158 | tx = fscal*dx00; |
1159 | ty = fscal*dy00; |
1160 | tz = fscal*dz00; |
1161 | |
1162 | /* Update vectorial force */ |
1163 | fix0 += tx; |
1164 | fiy0 += ty; |
1165 | fiz0 += tz; |
1166 | f[j_coord_offset+DIM3*0+XX0] -= tx; |
1167 | f[j_coord_offset+DIM3*0+YY1] -= ty; |
1168 | f[j_coord_offset+DIM3*0+ZZ2] -= tz; |
1169 | |
1170 | } |
1171 | |
1172 | /************************** |
1173 | * CALCULATE INTERACTIONS * |
1174 | **************************/ |
1175 | |
1176 | if (rsq11<rcutoff2) |
1177 | { |
1178 | |
1179 | r11 = rsq11*rinv11; |
1180 | |
1181 | /* EWALD ELECTROSTATICS */ |
1182 | |
1183 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
1184 | ewrt = r11*ewtabscale; |
1185 | ewitab = ewrt; |
1186 | eweps = ewrt-ewitab; |
1187 | ewitab = 4*ewitab; |
1188 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
1189 | velec = qq11*(rinv11-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
1190 | felec = qq11*rinv11*(rinvsq11-felec); |
1191 | |
1192 | d = r11-rswitch; |
1193 | d = (d>0.0) ? d : 0.0; |
1194 | d2 = d*d; |
1195 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
1196 | |
1197 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
1198 | |
1199 | /* Evaluate switch function */ |
1200 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
1201 | felec = felec*sw - rinv11*velec*dsw; |
1202 | |
1203 | fscal = felec; |
1204 | |
1205 | /* Calculate temporary vectorial force */ |
1206 | tx = fscal*dx11; |
1207 | ty = fscal*dy11; |
1208 | tz = fscal*dz11; |
1209 | |
1210 | /* Update vectorial force */ |
1211 | fix1 += tx; |
1212 | fiy1 += ty; |
1213 | fiz1 += tz; |
1214 | f[j_coord_offset+DIM3*1+XX0] -= tx; |
1215 | f[j_coord_offset+DIM3*1+YY1] -= ty; |
1216 | f[j_coord_offset+DIM3*1+ZZ2] -= tz; |
1217 | |
1218 | } |
1219 | |
1220 | /************************** |
1221 | * CALCULATE INTERACTIONS * |
1222 | **************************/ |
1223 | |
1224 | if (rsq12<rcutoff2) |
1225 | { |
1226 | |
1227 | r12 = rsq12*rinv12; |
1228 | |
1229 | /* EWALD ELECTROSTATICS */ |
1230 | |
1231 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
1232 | ewrt = r12*ewtabscale; |
1233 | ewitab = ewrt; |
1234 | eweps = ewrt-ewitab; |
1235 | ewitab = 4*ewitab; |
1236 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
1237 | velec = qq12*(rinv12-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
1238 | felec = qq12*rinv12*(rinvsq12-felec); |
1239 | |
1240 | d = r12-rswitch; |
1241 | d = (d>0.0) ? d : 0.0; |
1242 | d2 = d*d; |
1243 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
1244 | |
1245 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
1246 | |
1247 | /* Evaluate switch function */ |
1248 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
1249 | felec = felec*sw - rinv12*velec*dsw; |
1250 | |
1251 | fscal = felec; |
1252 | |
1253 | /* Calculate temporary vectorial force */ |
1254 | tx = fscal*dx12; |
1255 | ty = fscal*dy12; |
1256 | tz = fscal*dz12; |
1257 | |
1258 | /* Update vectorial force */ |
1259 | fix1 += tx; |
1260 | fiy1 += ty; |
1261 | fiz1 += tz; |
1262 | f[j_coord_offset+DIM3*2+XX0] -= tx; |
1263 | f[j_coord_offset+DIM3*2+YY1] -= ty; |
1264 | f[j_coord_offset+DIM3*2+ZZ2] -= tz; |
1265 | |
1266 | } |
1267 | |
1268 | /************************** |
1269 | * CALCULATE INTERACTIONS * |
1270 | **************************/ |
1271 | |
1272 | if (rsq13<rcutoff2) |
1273 | { |
1274 | |
1275 | r13 = rsq13*rinv13; |
1276 | |
1277 | /* EWALD ELECTROSTATICS */ |
1278 | |
1279 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
1280 | ewrt = r13*ewtabscale; |
1281 | ewitab = ewrt; |
1282 | eweps = ewrt-ewitab; |
1283 | ewitab = 4*ewitab; |
1284 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
1285 | velec = qq13*(rinv13-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
1286 | felec = qq13*rinv13*(rinvsq13-felec); |
1287 | |
1288 | d = r13-rswitch; |
1289 | d = (d>0.0) ? d : 0.0; |
1290 | d2 = d*d; |
1291 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
1292 | |
1293 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
1294 | |
1295 | /* Evaluate switch function */ |
1296 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
1297 | felec = felec*sw - rinv13*velec*dsw; |
1298 | |
1299 | fscal = felec; |
1300 | |
1301 | /* Calculate temporary vectorial force */ |
1302 | tx = fscal*dx13; |
1303 | ty = fscal*dy13; |
1304 | tz = fscal*dz13; |
1305 | |
1306 | /* Update vectorial force */ |
1307 | fix1 += tx; |
1308 | fiy1 += ty; |
1309 | fiz1 += tz; |
1310 | f[j_coord_offset+DIM3*3+XX0] -= tx; |
1311 | f[j_coord_offset+DIM3*3+YY1] -= ty; |
1312 | f[j_coord_offset+DIM3*3+ZZ2] -= tz; |
1313 | |
1314 | } |
1315 | |
1316 | /************************** |
1317 | * CALCULATE INTERACTIONS * |
1318 | **************************/ |
1319 | |
1320 | if (rsq21<rcutoff2) |
1321 | { |
1322 | |
1323 | r21 = rsq21*rinv21; |
1324 | |
1325 | /* EWALD ELECTROSTATICS */ |
1326 | |
1327 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
1328 | ewrt = r21*ewtabscale; |
1329 | ewitab = ewrt; |
1330 | eweps = ewrt-ewitab; |
1331 | ewitab = 4*ewitab; |
1332 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
1333 | velec = qq21*(rinv21-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
1334 | felec = qq21*rinv21*(rinvsq21-felec); |
1335 | |
1336 | d = r21-rswitch; |
1337 | d = (d>0.0) ? d : 0.0; |
1338 | d2 = d*d; |
1339 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
1340 | |
1341 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
1342 | |
1343 | /* Evaluate switch function */ |
1344 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
1345 | felec = felec*sw - rinv21*velec*dsw; |
1346 | |
1347 | fscal = felec; |
1348 | |
1349 | /* Calculate temporary vectorial force */ |
1350 | tx = fscal*dx21; |
1351 | ty = fscal*dy21; |
1352 | tz = fscal*dz21; |
1353 | |
1354 | /* Update vectorial force */ |
1355 | fix2 += tx; |
1356 | fiy2 += ty; |
1357 | fiz2 += tz; |
1358 | f[j_coord_offset+DIM3*1+XX0] -= tx; |
1359 | f[j_coord_offset+DIM3*1+YY1] -= ty; |
1360 | f[j_coord_offset+DIM3*1+ZZ2] -= tz; |
1361 | |
1362 | } |
1363 | |
1364 | /************************** |
1365 | * CALCULATE INTERACTIONS * |
1366 | **************************/ |
1367 | |
1368 | if (rsq22<rcutoff2) |
1369 | { |
1370 | |
1371 | r22 = rsq22*rinv22; |
1372 | |
1373 | /* EWALD ELECTROSTATICS */ |
1374 | |
1375 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
1376 | ewrt = r22*ewtabscale; |
1377 | ewitab = ewrt; |
1378 | eweps = ewrt-ewitab; |
1379 | ewitab = 4*ewitab; |
1380 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
1381 | velec = qq22*(rinv22-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
1382 | felec = qq22*rinv22*(rinvsq22-felec); |
1383 | |
1384 | d = r22-rswitch; |
1385 | d = (d>0.0) ? d : 0.0; |
1386 | d2 = d*d; |
1387 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
1388 | |
1389 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
1390 | |
1391 | /* Evaluate switch function */ |
1392 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
1393 | felec = felec*sw - rinv22*velec*dsw; |
1394 | |
1395 | fscal = felec; |
1396 | |
1397 | /* Calculate temporary vectorial force */ |
1398 | tx = fscal*dx22; |
1399 | ty = fscal*dy22; |
1400 | tz = fscal*dz22; |
1401 | |
1402 | /* Update vectorial force */ |
1403 | fix2 += tx; |
1404 | fiy2 += ty; |
1405 | fiz2 += tz; |
1406 | f[j_coord_offset+DIM3*2+XX0] -= tx; |
1407 | f[j_coord_offset+DIM3*2+YY1] -= ty; |
1408 | f[j_coord_offset+DIM3*2+ZZ2] -= tz; |
1409 | |
1410 | } |
1411 | |
1412 | /************************** |
1413 | * CALCULATE INTERACTIONS * |
1414 | **************************/ |
1415 | |
1416 | if (rsq23<rcutoff2) |
1417 | { |
1418 | |
1419 | r23 = rsq23*rinv23; |
1420 | |
1421 | /* EWALD ELECTROSTATICS */ |
1422 | |
1423 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
1424 | ewrt = r23*ewtabscale; |
1425 | ewitab = ewrt; |
1426 | eweps = ewrt-ewitab; |
1427 | ewitab = 4*ewitab; |
1428 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
1429 | velec = qq23*(rinv23-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
1430 | felec = qq23*rinv23*(rinvsq23-felec); |
1431 | |
1432 | d = r23-rswitch; |
1433 | d = (d>0.0) ? d : 0.0; |
1434 | d2 = d*d; |
1435 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
1436 | |
1437 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
1438 | |
1439 | /* Evaluate switch function */ |
1440 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
1441 | felec = felec*sw - rinv23*velec*dsw; |
1442 | |
1443 | fscal = felec; |
1444 | |
1445 | /* Calculate temporary vectorial force */ |
1446 | tx = fscal*dx23; |
1447 | ty = fscal*dy23; |
1448 | tz = fscal*dz23; |
1449 | |
1450 | /* Update vectorial force */ |
1451 | fix2 += tx; |
1452 | fiy2 += ty; |
1453 | fiz2 += tz; |
1454 | f[j_coord_offset+DIM3*3+XX0] -= tx; |
1455 | f[j_coord_offset+DIM3*3+YY1] -= ty; |
1456 | f[j_coord_offset+DIM3*3+ZZ2] -= tz; |
1457 | |
1458 | } |
1459 | |
1460 | /************************** |
1461 | * CALCULATE INTERACTIONS * |
1462 | **************************/ |
1463 | |
1464 | if (rsq31<rcutoff2) |
1465 | { |
1466 | |
1467 | r31 = rsq31*rinv31; |
1468 | |
1469 | /* EWALD ELECTROSTATICS */ |
1470 | |
1471 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
1472 | ewrt = r31*ewtabscale; |
1473 | ewitab = ewrt; |
1474 | eweps = ewrt-ewitab; |
1475 | ewitab = 4*ewitab; |
1476 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
1477 | velec = qq31*(rinv31-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
1478 | felec = qq31*rinv31*(rinvsq31-felec); |
1479 | |
1480 | d = r31-rswitch; |
1481 | d = (d>0.0) ? d : 0.0; |
1482 | d2 = d*d; |
1483 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
1484 | |
1485 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
1486 | |
1487 | /* Evaluate switch function */ |
1488 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
1489 | felec = felec*sw - rinv31*velec*dsw; |
1490 | |
1491 | fscal = felec; |
1492 | |
1493 | /* Calculate temporary vectorial force */ |
1494 | tx = fscal*dx31; |
1495 | ty = fscal*dy31; |
1496 | tz = fscal*dz31; |
1497 | |
1498 | /* Update vectorial force */ |
1499 | fix3 += tx; |
1500 | fiy3 += ty; |
1501 | fiz3 += tz; |
1502 | f[j_coord_offset+DIM3*1+XX0] -= tx; |
1503 | f[j_coord_offset+DIM3*1+YY1] -= ty; |
1504 | f[j_coord_offset+DIM3*1+ZZ2] -= tz; |
1505 | |
1506 | } |
1507 | |
1508 | /************************** |
1509 | * CALCULATE INTERACTIONS * |
1510 | **************************/ |
1511 | |
1512 | if (rsq32<rcutoff2) |
1513 | { |
1514 | |
1515 | r32 = rsq32*rinv32; |
1516 | |
1517 | /* EWALD ELECTROSTATICS */ |
1518 | |
1519 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
1520 | ewrt = r32*ewtabscale; |
1521 | ewitab = ewrt; |
1522 | eweps = ewrt-ewitab; |
1523 | ewitab = 4*ewitab; |
1524 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
1525 | velec = qq32*(rinv32-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
1526 | felec = qq32*rinv32*(rinvsq32-felec); |
1527 | |
1528 | d = r32-rswitch; |
1529 | d = (d>0.0) ? d : 0.0; |
1530 | d2 = d*d; |
1531 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
1532 | |
1533 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
1534 | |
1535 | /* Evaluate switch function */ |
1536 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
1537 | felec = felec*sw - rinv32*velec*dsw; |
1538 | |
1539 | fscal = felec; |
1540 | |
1541 | /* Calculate temporary vectorial force */ |
1542 | tx = fscal*dx32; |
1543 | ty = fscal*dy32; |
1544 | tz = fscal*dz32; |
1545 | |
1546 | /* Update vectorial force */ |
1547 | fix3 += tx; |
1548 | fiy3 += ty; |
1549 | fiz3 += tz; |
1550 | f[j_coord_offset+DIM3*2+XX0] -= tx; |
1551 | f[j_coord_offset+DIM3*2+YY1] -= ty; |
1552 | f[j_coord_offset+DIM3*2+ZZ2] -= tz; |
1553 | |
1554 | } |
1555 | |
1556 | /************************** |
1557 | * CALCULATE INTERACTIONS * |
1558 | **************************/ |
1559 | |
1560 | if (rsq33<rcutoff2) |
1561 | { |
1562 | |
1563 | r33 = rsq33*rinv33; |
1564 | |
1565 | /* EWALD ELECTROSTATICS */ |
1566 | |
1567 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
1568 | ewrt = r33*ewtabscale; |
1569 | ewitab = ewrt; |
1570 | eweps = ewrt-ewitab; |
1571 | ewitab = 4*ewitab; |
1572 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
1573 | velec = qq33*(rinv33-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
1574 | felec = qq33*rinv33*(rinvsq33-felec); |
1575 | |
1576 | d = r33-rswitch; |
1577 | d = (d>0.0) ? d : 0.0; |
1578 | d2 = d*d; |
1579 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
1580 | |
1581 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
1582 | |
1583 | /* Evaluate switch function */ |
1584 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
1585 | felec = felec*sw - rinv33*velec*dsw; |
1586 | |
1587 | fscal = felec; |
1588 | |
1589 | /* Calculate temporary vectorial force */ |
1590 | tx = fscal*dx33; |
1591 | ty = fscal*dy33; |
1592 | tz = fscal*dz33; |
1593 | |
1594 | /* Update vectorial force */ |
1595 | fix3 += tx; |
1596 | fiy3 += ty; |
1597 | fiz3 += tz; |
1598 | f[j_coord_offset+DIM3*3+XX0] -= tx; |
1599 | f[j_coord_offset+DIM3*3+YY1] -= ty; |
1600 | f[j_coord_offset+DIM3*3+ZZ2] -= tz; |
1601 | |
1602 | } |
1603 | |
1604 | /* Inner loop uses 555 flops */ |
1605 | } |
1606 | /* End of innermost loop */ |
1607 | |
1608 | tx = ty = tz = 0; |
1609 | f[i_coord_offset+DIM3*0+XX0] += fix0; |
1610 | f[i_coord_offset+DIM3*0+YY1] += fiy0; |
1611 | f[i_coord_offset+DIM3*0+ZZ2] += fiz0; |
1612 | tx += fix0; |
1613 | ty += fiy0; |
1614 | tz += fiz0; |
1615 | f[i_coord_offset+DIM3*1+XX0] += fix1; |
1616 | f[i_coord_offset+DIM3*1+YY1] += fiy1; |
1617 | f[i_coord_offset+DIM3*1+ZZ2] += fiz1; |
1618 | tx += fix1; |
1619 | ty += fiy1; |
1620 | tz += fiz1; |
1621 | f[i_coord_offset+DIM3*2+XX0] += fix2; |
1622 | f[i_coord_offset+DIM3*2+YY1] += fiy2; |
1623 | f[i_coord_offset+DIM3*2+ZZ2] += fiz2; |
1624 | tx += fix2; |
1625 | ty += fiy2; |
1626 | tz += fiz2; |
1627 | f[i_coord_offset+DIM3*3+XX0] += fix3; |
1628 | f[i_coord_offset+DIM3*3+YY1] += fiy3; |
1629 | f[i_coord_offset+DIM3*3+ZZ2] += fiz3; |
1630 | tx += fix3; |
1631 | ty += fiy3; |
1632 | tz += fiz3; |
1633 | fshift[i_shift_offset+XX0] += tx; |
1634 | fshift[i_shift_offset+YY1] += ty; |
1635 | fshift[i_shift_offset+ZZ2] += tz; |
1636 | |
1637 | /* Increment number of inner iterations */ |
1638 | inneriter += j_index_end - j_index_start; |
1639 | |
1640 | /* Outer loop uses 39 flops */ |
1641 | } |
1642 | |
1643 | /* Increment number of outer iterations */ |
1644 | outeriter += nri; |
1645 | |
1646 | /* Update outer/inner flops */ |
1647 | |
1648 | inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4W4_F,outeriter*39 + inneriter*555)(nrnb)->n[eNR_NBKERNEL_ELEC_VDW_W4W4_F] += outeriter*39 + inneriter *555; |
1649 | } |