| File: | gromacs/gmxlib/nonbonded/nb_kernel_c/nb_kernel_ElecEwSw_VdwNone_GeomW4W4_c.c |
| Location: | line 114, column 5 |
| Description: | Value stored to 'sh_ewald' 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_VdwNone_GeomW4W4_VF_c |
| 51 | * Electrostatics interaction: Ewald |
| 52 | * VdW interaction: None |
| 53 | * Geometry: Water4-Water4 |
| 54 | * Calculate force/pot: PotentialAndForce |
| 55 | */ |
| 56 | void |
| 57 | nb_kernel_ElecEwSw_VdwNone_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 vdwioffset1; |
| 73 | real ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1; |
| 74 | int vdwioffset2; |
| 75 | real ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2; |
| 76 | int vdwioffset3; |
| 77 | real ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3; |
| 78 | int vdwjidx1; |
| 79 | real jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1; |
| 80 | int vdwjidx2; |
| 81 | real jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2; |
| 82 | int vdwjidx3; |
| 83 | real jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3; |
| 84 | real dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11,cexp1_11,cexp2_11; |
| 85 | real dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12,cexp1_12,cexp2_12; |
| 86 | real dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13,cexp1_13,cexp2_13; |
| 87 | real dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21,cexp1_21,cexp2_21; |
| 88 | real dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22,cexp1_22,cexp2_22; |
| 89 | real dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23,cexp1_23,cexp2_23; |
| 90 | real dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31,cexp1_31,cexp2_31; |
| 91 | real dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32,cexp1_32,cexp2_32; |
| 92 | real dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33,cexp1_33,cexp2_33; |
| 93 | real velec,felec,velecsum,facel,crf,krf,krf2; |
| 94 | real *charge; |
| 95 | int ewitab; |
| 96 | real ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace; |
| 97 | real *ewtab; |
| 98 | real rswitch,swV3,swV4,swV5,swF2,swF3,swF4,d,d2,sw,dsw; |
| 99 | |
| 100 | x = xx[0]; |
| 101 | f = ff[0]; |
| 102 | |
| 103 | nri = nlist->nri; |
| 104 | iinr = nlist->iinr; |
| 105 | jindex = nlist->jindex; |
| 106 | jjnr = nlist->jjnr; |
| 107 | shiftidx = nlist->shift; |
| 108 | gid = nlist->gid; |
| 109 | shiftvec = fr->shift_vec[0]; |
| 110 | fshift = fr->fshift[0]; |
| 111 | facel = fr->epsfac; |
| 112 | charge = mdatoms->chargeA; |
| 113 | |
| 114 | sh_ewald = fr->ic->sh_ewald; |
Value stored to 'sh_ewald' is never read | |
| 115 | ewtab = fr->ic->tabq_coul_FDV0; |
| 116 | ewtabscale = fr->ic->tabq_scale; |
| 117 | ewtabhalfspace = 0.5/ewtabscale; |
| 118 | |
| 119 | /* Setup water-specific parameters */ |
| 120 | inr = nlist->iinr[0]; |
| 121 | iq1 = facel*charge[inr+1]; |
| 122 | iq2 = facel*charge[inr+2]; |
| 123 | iq3 = facel*charge[inr+3]; |
| 124 | |
| 125 | jq1 = charge[inr+1]; |
| 126 | jq2 = charge[inr+2]; |
| 127 | jq3 = charge[inr+3]; |
| 128 | qq11 = iq1*jq1; |
| 129 | qq12 = iq1*jq2; |
| 130 | qq13 = iq1*jq3; |
| 131 | qq21 = iq2*jq1; |
| 132 | qq22 = iq2*jq2; |
| 133 | qq23 = iq2*jq3; |
| 134 | qq31 = iq3*jq1; |
| 135 | qq32 = iq3*jq2; |
| 136 | qq33 = iq3*jq3; |
| 137 | |
| 138 | /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */ |
| 139 | rcutoff = fr->rcoulomb; |
| 140 | rcutoff2 = rcutoff*rcutoff; |
| 141 | |
| 142 | rswitch = fr->rcoulomb_switch; |
| 143 | /* Setup switch parameters */ |
| 144 | d = rcutoff-rswitch; |
| 145 | swV3 = -10.0/(d*d*d); |
| 146 | swV4 = 15.0/(d*d*d*d); |
| 147 | swV5 = -6.0/(d*d*d*d*d); |
| 148 | swF2 = -30.0/(d*d*d); |
| 149 | swF3 = 60.0/(d*d*d*d); |
| 150 | swF4 = -30.0/(d*d*d*d*d); |
| 151 | |
| 152 | outeriter = 0; |
| 153 | inneriter = 0; |
| 154 | |
| 155 | /* Start outer loop over neighborlists */ |
| 156 | for(iidx=0; iidx<nri; iidx++) |
| 157 | { |
| 158 | /* Load shift vector for this list */ |
| 159 | i_shift_offset = DIM3*shiftidx[iidx]; |
| 160 | shX = shiftvec[i_shift_offset+XX0]; |
| 161 | shY = shiftvec[i_shift_offset+YY1]; |
| 162 | shZ = shiftvec[i_shift_offset+ZZ2]; |
| 163 | |
| 164 | /* Load limits for loop over neighbors */ |
| 165 | j_index_start = jindex[iidx]; |
| 166 | j_index_end = jindex[iidx+1]; |
| 167 | |
| 168 | /* Get outer coordinate index */ |
| 169 | inr = iinr[iidx]; |
| 170 | i_coord_offset = DIM3*inr; |
| 171 | |
| 172 | /* Load i particle coords and add shift vector */ |
| 173 | ix1 = shX + x[i_coord_offset+DIM3*1+XX0]; |
| 174 | iy1 = shY + x[i_coord_offset+DIM3*1+YY1]; |
| 175 | iz1 = shZ + x[i_coord_offset+DIM3*1+ZZ2]; |
| 176 | ix2 = shX + x[i_coord_offset+DIM3*2+XX0]; |
| 177 | iy2 = shY + x[i_coord_offset+DIM3*2+YY1]; |
| 178 | iz2 = shZ + x[i_coord_offset+DIM3*2+ZZ2]; |
| 179 | ix3 = shX + x[i_coord_offset+DIM3*3+XX0]; |
| 180 | iy3 = shY + x[i_coord_offset+DIM3*3+YY1]; |
| 181 | iz3 = shZ + x[i_coord_offset+DIM3*3+ZZ2]; |
| 182 | |
| 183 | fix1 = 0.0; |
| 184 | fiy1 = 0.0; |
| 185 | fiz1 = 0.0; |
| 186 | fix2 = 0.0; |
| 187 | fiy2 = 0.0; |
| 188 | fiz2 = 0.0; |
| 189 | fix3 = 0.0; |
| 190 | fiy3 = 0.0; |
| 191 | fiz3 = 0.0; |
| 192 | |
| 193 | /* Reset potential sums */ |
| 194 | velecsum = 0.0; |
| 195 | |
| 196 | /* Start inner kernel loop */ |
| 197 | for(jidx=j_index_start; jidx<j_index_end; jidx++) |
| 198 | { |
| 199 | /* Get j neighbor index, and coordinate index */ |
| 200 | jnr = jjnr[jidx]; |
| 201 | j_coord_offset = DIM3*jnr; |
| 202 | |
| 203 | /* load j atom coordinates */ |
| 204 | jx1 = x[j_coord_offset+DIM3*1+XX0]; |
| 205 | jy1 = x[j_coord_offset+DIM3*1+YY1]; |
| 206 | jz1 = x[j_coord_offset+DIM3*1+ZZ2]; |
| 207 | jx2 = x[j_coord_offset+DIM3*2+XX0]; |
| 208 | jy2 = x[j_coord_offset+DIM3*2+YY1]; |
| 209 | jz2 = x[j_coord_offset+DIM3*2+ZZ2]; |
| 210 | jx3 = x[j_coord_offset+DIM3*3+XX0]; |
| 211 | jy3 = x[j_coord_offset+DIM3*3+YY1]; |
| 212 | jz3 = x[j_coord_offset+DIM3*3+ZZ2]; |
| 213 | |
| 214 | /* Calculate displacement vector */ |
| 215 | dx11 = ix1 - jx1; |
| 216 | dy11 = iy1 - jy1; |
| 217 | dz11 = iz1 - jz1; |
| 218 | dx12 = ix1 - jx2; |
| 219 | dy12 = iy1 - jy2; |
| 220 | dz12 = iz1 - jz2; |
| 221 | dx13 = ix1 - jx3; |
| 222 | dy13 = iy1 - jy3; |
| 223 | dz13 = iz1 - jz3; |
| 224 | dx21 = ix2 - jx1; |
| 225 | dy21 = iy2 - jy1; |
| 226 | dz21 = iz2 - jz1; |
| 227 | dx22 = ix2 - jx2; |
| 228 | dy22 = iy2 - jy2; |
| 229 | dz22 = iz2 - jz2; |
| 230 | dx23 = ix2 - jx3; |
| 231 | dy23 = iy2 - jy3; |
| 232 | dz23 = iz2 - jz3; |
| 233 | dx31 = ix3 - jx1; |
| 234 | dy31 = iy3 - jy1; |
| 235 | dz31 = iz3 - jz1; |
| 236 | dx32 = ix3 - jx2; |
| 237 | dy32 = iy3 - jy2; |
| 238 | dz32 = iz3 - jz2; |
| 239 | dx33 = ix3 - jx3; |
| 240 | dy33 = iy3 - jy3; |
| 241 | dz33 = iz3 - jz3; |
| 242 | |
| 243 | /* Calculate squared distance and things based on it */ |
| 244 | rsq11 = dx11*dx11+dy11*dy11+dz11*dz11; |
| 245 | rsq12 = dx12*dx12+dy12*dy12+dz12*dz12; |
| 246 | rsq13 = dx13*dx13+dy13*dy13+dz13*dz13; |
| 247 | rsq21 = dx21*dx21+dy21*dy21+dz21*dz21; |
| 248 | rsq22 = dx22*dx22+dy22*dy22+dz22*dz22; |
| 249 | rsq23 = dx23*dx23+dy23*dy23+dz23*dz23; |
| 250 | rsq31 = dx31*dx31+dy31*dy31+dz31*dz31; |
| 251 | rsq32 = dx32*dx32+dy32*dy32+dz32*dz32; |
| 252 | rsq33 = dx33*dx33+dy33*dy33+dz33*dz33; |
| 253 | |
| 254 | rinv11 = gmx_invsqrt(rsq11)gmx_software_invsqrt(rsq11); |
| 255 | rinv12 = gmx_invsqrt(rsq12)gmx_software_invsqrt(rsq12); |
| 256 | rinv13 = gmx_invsqrt(rsq13)gmx_software_invsqrt(rsq13); |
| 257 | rinv21 = gmx_invsqrt(rsq21)gmx_software_invsqrt(rsq21); |
| 258 | rinv22 = gmx_invsqrt(rsq22)gmx_software_invsqrt(rsq22); |
| 259 | rinv23 = gmx_invsqrt(rsq23)gmx_software_invsqrt(rsq23); |
| 260 | rinv31 = gmx_invsqrt(rsq31)gmx_software_invsqrt(rsq31); |
| 261 | rinv32 = gmx_invsqrt(rsq32)gmx_software_invsqrt(rsq32); |
| 262 | rinv33 = gmx_invsqrt(rsq33)gmx_software_invsqrt(rsq33); |
| 263 | |
| 264 | rinvsq11 = rinv11*rinv11; |
| 265 | rinvsq12 = rinv12*rinv12; |
| 266 | rinvsq13 = rinv13*rinv13; |
| 267 | rinvsq21 = rinv21*rinv21; |
| 268 | rinvsq22 = rinv22*rinv22; |
| 269 | rinvsq23 = rinv23*rinv23; |
| 270 | rinvsq31 = rinv31*rinv31; |
| 271 | rinvsq32 = rinv32*rinv32; |
| 272 | rinvsq33 = rinv33*rinv33; |
| 273 | |
| 274 | /************************** |
| 275 | * CALCULATE INTERACTIONS * |
| 276 | **************************/ |
| 277 | |
| 278 | if (rsq11<rcutoff2) |
| 279 | { |
| 280 | |
| 281 | r11 = rsq11*rinv11; |
| 282 | |
| 283 | /* EWALD ELECTROSTATICS */ |
| 284 | |
| 285 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
| 286 | ewrt = r11*ewtabscale; |
| 287 | ewitab = ewrt; |
| 288 | eweps = ewrt-ewitab; |
| 289 | ewitab = 4*ewitab; |
| 290 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
| 291 | velec = qq11*(rinv11-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
| 292 | felec = qq11*rinv11*(rinvsq11-felec); |
| 293 | |
| 294 | d = r11-rswitch; |
| 295 | d = (d>0.0) ? d : 0.0; |
| 296 | d2 = d*d; |
| 297 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
| 298 | |
| 299 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
| 300 | |
| 301 | /* Evaluate switch function */ |
| 302 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
| 303 | felec = felec*sw - rinv11*velec*dsw; |
| 304 | velec *= sw; |
| 305 | |
| 306 | /* Update potential sums from outer loop */ |
| 307 | velecsum += velec; |
| 308 | |
| 309 | fscal = felec; |
| 310 | |
| 311 | /* Calculate temporary vectorial force */ |
| 312 | tx = fscal*dx11; |
| 313 | ty = fscal*dy11; |
| 314 | tz = fscal*dz11; |
| 315 | |
| 316 | /* Update vectorial force */ |
| 317 | fix1 += tx; |
| 318 | fiy1 += ty; |
| 319 | fiz1 += tz; |
| 320 | f[j_coord_offset+DIM3*1+XX0] -= tx; |
| 321 | f[j_coord_offset+DIM3*1+YY1] -= ty; |
| 322 | f[j_coord_offset+DIM3*1+ZZ2] -= tz; |
| 323 | |
| 324 | } |
| 325 | |
| 326 | /************************** |
| 327 | * CALCULATE INTERACTIONS * |
| 328 | **************************/ |
| 329 | |
| 330 | if (rsq12<rcutoff2) |
| 331 | { |
| 332 | |
| 333 | r12 = rsq12*rinv12; |
| 334 | |
| 335 | /* EWALD ELECTROSTATICS */ |
| 336 | |
| 337 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
| 338 | ewrt = r12*ewtabscale; |
| 339 | ewitab = ewrt; |
| 340 | eweps = ewrt-ewitab; |
| 341 | ewitab = 4*ewitab; |
| 342 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
| 343 | velec = qq12*(rinv12-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
| 344 | felec = qq12*rinv12*(rinvsq12-felec); |
| 345 | |
| 346 | d = r12-rswitch; |
| 347 | d = (d>0.0) ? d : 0.0; |
| 348 | d2 = d*d; |
| 349 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
| 350 | |
| 351 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
| 352 | |
| 353 | /* Evaluate switch function */ |
| 354 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
| 355 | felec = felec*sw - rinv12*velec*dsw; |
| 356 | velec *= sw; |
| 357 | |
| 358 | /* Update potential sums from outer loop */ |
| 359 | velecsum += velec; |
| 360 | |
| 361 | fscal = felec; |
| 362 | |
| 363 | /* Calculate temporary vectorial force */ |
| 364 | tx = fscal*dx12; |
| 365 | ty = fscal*dy12; |
| 366 | tz = fscal*dz12; |
| 367 | |
| 368 | /* Update vectorial force */ |
| 369 | fix1 += tx; |
| 370 | fiy1 += ty; |
| 371 | fiz1 += tz; |
| 372 | f[j_coord_offset+DIM3*2+XX0] -= tx; |
| 373 | f[j_coord_offset+DIM3*2+YY1] -= ty; |
| 374 | f[j_coord_offset+DIM3*2+ZZ2] -= tz; |
| 375 | |
| 376 | } |
| 377 | |
| 378 | /************************** |
| 379 | * CALCULATE INTERACTIONS * |
| 380 | **************************/ |
| 381 | |
| 382 | if (rsq13<rcutoff2) |
| 383 | { |
| 384 | |
| 385 | r13 = rsq13*rinv13; |
| 386 | |
| 387 | /* EWALD ELECTROSTATICS */ |
| 388 | |
| 389 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
| 390 | ewrt = r13*ewtabscale; |
| 391 | ewitab = ewrt; |
| 392 | eweps = ewrt-ewitab; |
| 393 | ewitab = 4*ewitab; |
| 394 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
| 395 | velec = qq13*(rinv13-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
| 396 | felec = qq13*rinv13*(rinvsq13-felec); |
| 397 | |
| 398 | d = r13-rswitch; |
| 399 | d = (d>0.0) ? d : 0.0; |
| 400 | d2 = d*d; |
| 401 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
| 402 | |
| 403 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
| 404 | |
| 405 | /* Evaluate switch function */ |
| 406 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
| 407 | felec = felec*sw - rinv13*velec*dsw; |
| 408 | velec *= sw; |
| 409 | |
| 410 | /* Update potential sums from outer loop */ |
| 411 | velecsum += velec; |
| 412 | |
| 413 | fscal = felec; |
| 414 | |
| 415 | /* Calculate temporary vectorial force */ |
| 416 | tx = fscal*dx13; |
| 417 | ty = fscal*dy13; |
| 418 | tz = fscal*dz13; |
| 419 | |
| 420 | /* Update vectorial force */ |
| 421 | fix1 += tx; |
| 422 | fiy1 += ty; |
| 423 | fiz1 += tz; |
| 424 | f[j_coord_offset+DIM3*3+XX0] -= tx; |
| 425 | f[j_coord_offset+DIM3*3+YY1] -= ty; |
| 426 | f[j_coord_offset+DIM3*3+ZZ2] -= tz; |
| 427 | |
| 428 | } |
| 429 | |
| 430 | /************************** |
| 431 | * CALCULATE INTERACTIONS * |
| 432 | **************************/ |
| 433 | |
| 434 | if (rsq21<rcutoff2) |
| 435 | { |
| 436 | |
| 437 | r21 = rsq21*rinv21; |
| 438 | |
| 439 | /* EWALD ELECTROSTATICS */ |
| 440 | |
| 441 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
| 442 | ewrt = r21*ewtabscale; |
| 443 | ewitab = ewrt; |
| 444 | eweps = ewrt-ewitab; |
| 445 | ewitab = 4*ewitab; |
| 446 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
| 447 | velec = qq21*(rinv21-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
| 448 | felec = qq21*rinv21*(rinvsq21-felec); |
| 449 | |
| 450 | d = r21-rswitch; |
| 451 | d = (d>0.0) ? d : 0.0; |
| 452 | d2 = d*d; |
| 453 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
| 454 | |
| 455 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
| 456 | |
| 457 | /* Evaluate switch function */ |
| 458 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
| 459 | felec = felec*sw - rinv21*velec*dsw; |
| 460 | velec *= sw; |
| 461 | |
| 462 | /* Update potential sums from outer loop */ |
| 463 | velecsum += velec; |
| 464 | |
| 465 | fscal = felec; |
| 466 | |
| 467 | /* Calculate temporary vectorial force */ |
| 468 | tx = fscal*dx21; |
| 469 | ty = fscal*dy21; |
| 470 | tz = fscal*dz21; |
| 471 | |
| 472 | /* Update vectorial force */ |
| 473 | fix2 += tx; |
| 474 | fiy2 += ty; |
| 475 | fiz2 += tz; |
| 476 | f[j_coord_offset+DIM3*1+XX0] -= tx; |
| 477 | f[j_coord_offset+DIM3*1+YY1] -= ty; |
| 478 | f[j_coord_offset+DIM3*1+ZZ2] -= tz; |
| 479 | |
| 480 | } |
| 481 | |
| 482 | /************************** |
| 483 | * CALCULATE INTERACTIONS * |
| 484 | **************************/ |
| 485 | |
| 486 | if (rsq22<rcutoff2) |
| 487 | { |
| 488 | |
| 489 | r22 = rsq22*rinv22; |
| 490 | |
| 491 | /* EWALD ELECTROSTATICS */ |
| 492 | |
| 493 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
| 494 | ewrt = r22*ewtabscale; |
| 495 | ewitab = ewrt; |
| 496 | eweps = ewrt-ewitab; |
| 497 | ewitab = 4*ewitab; |
| 498 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
| 499 | velec = qq22*(rinv22-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
| 500 | felec = qq22*rinv22*(rinvsq22-felec); |
| 501 | |
| 502 | d = r22-rswitch; |
| 503 | d = (d>0.0) ? d : 0.0; |
| 504 | d2 = d*d; |
| 505 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
| 506 | |
| 507 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
| 508 | |
| 509 | /* Evaluate switch function */ |
| 510 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
| 511 | felec = felec*sw - rinv22*velec*dsw; |
| 512 | velec *= sw; |
| 513 | |
| 514 | /* Update potential sums from outer loop */ |
| 515 | velecsum += velec; |
| 516 | |
| 517 | fscal = felec; |
| 518 | |
| 519 | /* Calculate temporary vectorial force */ |
| 520 | tx = fscal*dx22; |
| 521 | ty = fscal*dy22; |
| 522 | tz = fscal*dz22; |
| 523 | |
| 524 | /* Update vectorial force */ |
| 525 | fix2 += tx; |
| 526 | fiy2 += ty; |
| 527 | fiz2 += tz; |
| 528 | f[j_coord_offset+DIM3*2+XX0] -= tx; |
| 529 | f[j_coord_offset+DIM3*2+YY1] -= ty; |
| 530 | f[j_coord_offset+DIM3*2+ZZ2] -= tz; |
| 531 | |
| 532 | } |
| 533 | |
| 534 | /************************** |
| 535 | * CALCULATE INTERACTIONS * |
| 536 | **************************/ |
| 537 | |
| 538 | if (rsq23<rcutoff2) |
| 539 | { |
| 540 | |
| 541 | r23 = rsq23*rinv23; |
| 542 | |
| 543 | /* EWALD ELECTROSTATICS */ |
| 544 | |
| 545 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
| 546 | ewrt = r23*ewtabscale; |
| 547 | ewitab = ewrt; |
| 548 | eweps = ewrt-ewitab; |
| 549 | ewitab = 4*ewitab; |
| 550 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
| 551 | velec = qq23*(rinv23-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
| 552 | felec = qq23*rinv23*(rinvsq23-felec); |
| 553 | |
| 554 | d = r23-rswitch; |
| 555 | d = (d>0.0) ? d : 0.0; |
| 556 | d2 = d*d; |
| 557 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
| 558 | |
| 559 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
| 560 | |
| 561 | /* Evaluate switch function */ |
| 562 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
| 563 | felec = felec*sw - rinv23*velec*dsw; |
| 564 | velec *= sw; |
| 565 | |
| 566 | /* Update potential sums from outer loop */ |
| 567 | velecsum += velec; |
| 568 | |
| 569 | fscal = felec; |
| 570 | |
| 571 | /* Calculate temporary vectorial force */ |
| 572 | tx = fscal*dx23; |
| 573 | ty = fscal*dy23; |
| 574 | tz = fscal*dz23; |
| 575 | |
| 576 | /* Update vectorial force */ |
| 577 | fix2 += tx; |
| 578 | fiy2 += ty; |
| 579 | fiz2 += tz; |
| 580 | f[j_coord_offset+DIM3*3+XX0] -= tx; |
| 581 | f[j_coord_offset+DIM3*3+YY1] -= ty; |
| 582 | f[j_coord_offset+DIM3*3+ZZ2] -= tz; |
| 583 | |
| 584 | } |
| 585 | |
| 586 | /************************** |
| 587 | * CALCULATE INTERACTIONS * |
| 588 | **************************/ |
| 589 | |
| 590 | if (rsq31<rcutoff2) |
| 591 | { |
| 592 | |
| 593 | r31 = rsq31*rinv31; |
| 594 | |
| 595 | /* EWALD ELECTROSTATICS */ |
| 596 | |
| 597 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
| 598 | ewrt = r31*ewtabscale; |
| 599 | ewitab = ewrt; |
| 600 | eweps = ewrt-ewitab; |
| 601 | ewitab = 4*ewitab; |
| 602 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
| 603 | velec = qq31*(rinv31-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
| 604 | felec = qq31*rinv31*(rinvsq31-felec); |
| 605 | |
| 606 | d = r31-rswitch; |
| 607 | d = (d>0.0) ? d : 0.0; |
| 608 | d2 = d*d; |
| 609 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
| 610 | |
| 611 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
| 612 | |
| 613 | /* Evaluate switch function */ |
| 614 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
| 615 | felec = felec*sw - rinv31*velec*dsw; |
| 616 | velec *= sw; |
| 617 | |
| 618 | /* Update potential sums from outer loop */ |
| 619 | velecsum += velec; |
| 620 | |
| 621 | fscal = felec; |
| 622 | |
| 623 | /* Calculate temporary vectorial force */ |
| 624 | tx = fscal*dx31; |
| 625 | ty = fscal*dy31; |
| 626 | tz = fscal*dz31; |
| 627 | |
| 628 | /* Update vectorial force */ |
| 629 | fix3 += tx; |
| 630 | fiy3 += ty; |
| 631 | fiz3 += tz; |
| 632 | f[j_coord_offset+DIM3*1+XX0] -= tx; |
| 633 | f[j_coord_offset+DIM3*1+YY1] -= ty; |
| 634 | f[j_coord_offset+DIM3*1+ZZ2] -= tz; |
| 635 | |
| 636 | } |
| 637 | |
| 638 | /************************** |
| 639 | * CALCULATE INTERACTIONS * |
| 640 | **************************/ |
| 641 | |
| 642 | if (rsq32<rcutoff2) |
| 643 | { |
| 644 | |
| 645 | r32 = rsq32*rinv32; |
| 646 | |
| 647 | /* EWALD ELECTROSTATICS */ |
| 648 | |
| 649 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
| 650 | ewrt = r32*ewtabscale; |
| 651 | ewitab = ewrt; |
| 652 | eweps = ewrt-ewitab; |
| 653 | ewitab = 4*ewitab; |
| 654 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
| 655 | velec = qq32*(rinv32-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
| 656 | felec = qq32*rinv32*(rinvsq32-felec); |
| 657 | |
| 658 | d = r32-rswitch; |
| 659 | d = (d>0.0) ? d : 0.0; |
| 660 | d2 = d*d; |
| 661 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
| 662 | |
| 663 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
| 664 | |
| 665 | /* Evaluate switch function */ |
| 666 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
| 667 | felec = felec*sw - rinv32*velec*dsw; |
| 668 | velec *= sw; |
| 669 | |
| 670 | /* Update potential sums from outer loop */ |
| 671 | velecsum += velec; |
| 672 | |
| 673 | fscal = felec; |
| 674 | |
| 675 | /* Calculate temporary vectorial force */ |
| 676 | tx = fscal*dx32; |
| 677 | ty = fscal*dy32; |
| 678 | tz = fscal*dz32; |
| 679 | |
| 680 | /* Update vectorial force */ |
| 681 | fix3 += tx; |
| 682 | fiy3 += ty; |
| 683 | fiz3 += tz; |
| 684 | f[j_coord_offset+DIM3*2+XX0] -= tx; |
| 685 | f[j_coord_offset+DIM3*2+YY1] -= ty; |
| 686 | f[j_coord_offset+DIM3*2+ZZ2] -= tz; |
| 687 | |
| 688 | } |
| 689 | |
| 690 | /************************** |
| 691 | * CALCULATE INTERACTIONS * |
| 692 | **************************/ |
| 693 | |
| 694 | if (rsq33<rcutoff2) |
| 695 | { |
| 696 | |
| 697 | r33 = rsq33*rinv33; |
| 698 | |
| 699 | /* EWALD ELECTROSTATICS */ |
| 700 | |
| 701 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
| 702 | ewrt = r33*ewtabscale; |
| 703 | ewitab = ewrt; |
| 704 | eweps = ewrt-ewitab; |
| 705 | ewitab = 4*ewitab; |
| 706 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
| 707 | velec = qq33*(rinv33-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
| 708 | felec = qq33*rinv33*(rinvsq33-felec); |
| 709 | |
| 710 | d = r33-rswitch; |
| 711 | d = (d>0.0) ? d : 0.0; |
| 712 | d2 = d*d; |
| 713 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
| 714 | |
| 715 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
| 716 | |
| 717 | /* Evaluate switch function */ |
| 718 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
| 719 | felec = felec*sw - rinv33*velec*dsw; |
| 720 | velec *= sw; |
| 721 | |
| 722 | /* Update potential sums from outer loop */ |
| 723 | velecsum += velec; |
| 724 | |
| 725 | fscal = felec; |
| 726 | |
| 727 | /* Calculate temporary vectorial force */ |
| 728 | tx = fscal*dx33; |
| 729 | ty = fscal*dy33; |
| 730 | tz = fscal*dz33; |
| 731 | |
| 732 | /* Update vectorial force */ |
| 733 | fix3 += tx; |
| 734 | fiy3 += ty; |
| 735 | fiz3 += tz; |
| 736 | f[j_coord_offset+DIM3*3+XX0] -= tx; |
| 737 | f[j_coord_offset+DIM3*3+YY1] -= ty; |
| 738 | f[j_coord_offset+DIM3*3+ZZ2] -= tz; |
| 739 | |
| 740 | } |
| 741 | |
| 742 | /* Inner loop uses 522 flops */ |
| 743 | } |
| 744 | /* End of innermost loop */ |
| 745 | |
| 746 | tx = ty = tz = 0; |
| 747 | f[i_coord_offset+DIM3*1+XX0] += fix1; |
| 748 | f[i_coord_offset+DIM3*1+YY1] += fiy1; |
| 749 | f[i_coord_offset+DIM3*1+ZZ2] += fiz1; |
| 750 | tx += fix1; |
| 751 | ty += fiy1; |
| 752 | tz += fiz1; |
| 753 | f[i_coord_offset+DIM3*2+XX0] += fix2; |
| 754 | f[i_coord_offset+DIM3*2+YY1] += fiy2; |
| 755 | f[i_coord_offset+DIM3*2+ZZ2] += fiz2; |
| 756 | tx += fix2; |
| 757 | ty += fiy2; |
| 758 | tz += fiz2; |
| 759 | f[i_coord_offset+DIM3*3+XX0] += fix3; |
| 760 | f[i_coord_offset+DIM3*3+YY1] += fiy3; |
| 761 | f[i_coord_offset+DIM3*3+ZZ2] += fiz3; |
| 762 | tx += fix3; |
| 763 | ty += fiy3; |
| 764 | tz += fiz3; |
| 765 | fshift[i_shift_offset+XX0] += tx; |
| 766 | fshift[i_shift_offset+YY1] += ty; |
| 767 | fshift[i_shift_offset+ZZ2] += tz; |
| 768 | |
| 769 | ggid = gid[iidx]; |
| 770 | /* Update potential energies */ |
| 771 | kernel_data->energygrp_elec[ggid] += velecsum; |
| 772 | |
| 773 | /* Increment number of inner iterations */ |
| 774 | inneriter += j_index_end - j_index_start; |
| 775 | |
| 776 | /* Outer loop uses 31 flops */ |
| 777 | } |
| 778 | |
| 779 | /* Increment number of outer iterations */ |
| 780 | outeriter += nri; |
| 781 | |
| 782 | /* Update outer/inner flops */ |
| 783 | |
| 784 | inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_W4W4_VF,outeriter*31 + inneriter*522)(nrnb)->n[eNR_NBKERNEL_ELEC_W4W4_VF] += outeriter*31 + inneriter *522; |
| 785 | } |
| 786 | /* |
| 787 | * Gromacs nonbonded kernel: nb_kernel_ElecEwSw_VdwNone_GeomW4W4_F_c |
| 788 | * Electrostatics interaction: Ewald |
| 789 | * VdW interaction: None |
| 790 | * Geometry: Water4-Water4 |
| 791 | * Calculate force/pot: Force |
| 792 | */ |
| 793 | void |
| 794 | nb_kernel_ElecEwSw_VdwNone_GeomW4W4_F_c |
| 795 | (t_nblist * gmx_restrict__restrict nlist, |
| 796 | rvec * gmx_restrict__restrict xx, |
| 797 | rvec * gmx_restrict__restrict ff, |
| 798 | t_forcerec * gmx_restrict__restrict fr, |
| 799 | t_mdatoms * gmx_restrict__restrict mdatoms, |
| 800 | nb_kernel_data_t gmx_unused__attribute__ ((unused)) * gmx_restrict__restrict kernel_data, |
| 801 | t_nrnb * gmx_restrict__restrict nrnb) |
| 802 | { |
| 803 | int i_shift_offset,i_coord_offset,j_coord_offset; |
| 804 | int j_index_start,j_index_end; |
| 805 | int nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter; |
| 806 | real shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2; |
| 807 | int *iinr,*jindex,*jjnr,*shiftidx,*gid; |
| 808 | real *shiftvec,*fshift,*x,*f; |
| 809 | int vdwioffset1; |
| 810 | real ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1; |
| 811 | int vdwioffset2; |
| 812 | real ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2; |
| 813 | int vdwioffset3; |
| 814 | real ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3; |
| 815 | int vdwjidx1; |
| 816 | real jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1; |
| 817 | int vdwjidx2; |
| 818 | real jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2; |
| 819 | int vdwjidx3; |
| 820 | real jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3; |
| 821 | real dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11,cexp1_11,cexp2_11; |
| 822 | real dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12,cexp1_12,cexp2_12; |
| 823 | real dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13,cexp1_13,cexp2_13; |
| 824 | real dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21,cexp1_21,cexp2_21; |
| 825 | real dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22,cexp1_22,cexp2_22; |
| 826 | real dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23,cexp1_23,cexp2_23; |
| 827 | real dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31,cexp1_31,cexp2_31; |
| 828 | real dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32,cexp1_32,cexp2_32; |
| 829 | real dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33,cexp1_33,cexp2_33; |
| 830 | real velec,felec,velecsum,facel,crf,krf,krf2; |
| 831 | real *charge; |
| 832 | int ewitab; |
| 833 | real ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace; |
| 834 | real *ewtab; |
| 835 | real rswitch,swV3,swV4,swV5,swF2,swF3,swF4,d,d2,sw,dsw; |
| 836 | |
| 837 | x = xx[0]; |
| 838 | f = ff[0]; |
| 839 | |
| 840 | nri = nlist->nri; |
| 841 | iinr = nlist->iinr; |
| 842 | jindex = nlist->jindex; |
| 843 | jjnr = nlist->jjnr; |
| 844 | shiftidx = nlist->shift; |
| 845 | gid = nlist->gid; |
| 846 | shiftvec = fr->shift_vec[0]; |
| 847 | fshift = fr->fshift[0]; |
| 848 | facel = fr->epsfac; |
| 849 | charge = mdatoms->chargeA; |
| 850 | |
| 851 | sh_ewald = fr->ic->sh_ewald; |
| 852 | ewtab = fr->ic->tabq_coul_FDV0; |
| 853 | ewtabscale = fr->ic->tabq_scale; |
| 854 | ewtabhalfspace = 0.5/ewtabscale; |
| 855 | |
| 856 | /* Setup water-specific parameters */ |
| 857 | inr = nlist->iinr[0]; |
| 858 | iq1 = facel*charge[inr+1]; |
| 859 | iq2 = facel*charge[inr+2]; |
| 860 | iq3 = facel*charge[inr+3]; |
| 861 | |
| 862 | jq1 = charge[inr+1]; |
| 863 | jq2 = charge[inr+2]; |
| 864 | jq3 = charge[inr+3]; |
| 865 | qq11 = iq1*jq1; |
| 866 | qq12 = iq1*jq2; |
| 867 | qq13 = iq1*jq3; |
| 868 | qq21 = iq2*jq1; |
| 869 | qq22 = iq2*jq2; |
| 870 | qq23 = iq2*jq3; |
| 871 | qq31 = iq3*jq1; |
| 872 | qq32 = iq3*jq2; |
| 873 | qq33 = iq3*jq3; |
| 874 | |
| 875 | /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */ |
| 876 | rcutoff = fr->rcoulomb; |
| 877 | rcutoff2 = rcutoff*rcutoff; |
| 878 | |
| 879 | rswitch = fr->rcoulomb_switch; |
| 880 | /* Setup switch parameters */ |
| 881 | d = rcutoff-rswitch; |
| 882 | swV3 = -10.0/(d*d*d); |
| 883 | swV4 = 15.0/(d*d*d*d); |
| 884 | swV5 = -6.0/(d*d*d*d*d); |
| 885 | swF2 = -30.0/(d*d*d); |
| 886 | swF3 = 60.0/(d*d*d*d); |
| 887 | swF4 = -30.0/(d*d*d*d*d); |
| 888 | |
| 889 | outeriter = 0; |
| 890 | inneriter = 0; |
| 891 | |
| 892 | /* Start outer loop over neighborlists */ |
| 893 | for(iidx=0; iidx<nri; iidx++) |
| 894 | { |
| 895 | /* Load shift vector for this list */ |
| 896 | i_shift_offset = DIM3*shiftidx[iidx]; |
| 897 | shX = shiftvec[i_shift_offset+XX0]; |
| 898 | shY = shiftvec[i_shift_offset+YY1]; |
| 899 | shZ = shiftvec[i_shift_offset+ZZ2]; |
| 900 | |
| 901 | /* Load limits for loop over neighbors */ |
| 902 | j_index_start = jindex[iidx]; |
| 903 | j_index_end = jindex[iidx+1]; |
| 904 | |
| 905 | /* Get outer coordinate index */ |
| 906 | inr = iinr[iidx]; |
| 907 | i_coord_offset = DIM3*inr; |
| 908 | |
| 909 | /* Load i particle coords and add shift vector */ |
| 910 | ix1 = shX + x[i_coord_offset+DIM3*1+XX0]; |
| 911 | iy1 = shY + x[i_coord_offset+DIM3*1+YY1]; |
| 912 | iz1 = shZ + x[i_coord_offset+DIM3*1+ZZ2]; |
| 913 | ix2 = shX + x[i_coord_offset+DIM3*2+XX0]; |
| 914 | iy2 = shY + x[i_coord_offset+DIM3*2+YY1]; |
| 915 | iz2 = shZ + x[i_coord_offset+DIM3*2+ZZ2]; |
| 916 | ix3 = shX + x[i_coord_offset+DIM3*3+XX0]; |
| 917 | iy3 = shY + x[i_coord_offset+DIM3*3+YY1]; |
| 918 | iz3 = shZ + x[i_coord_offset+DIM3*3+ZZ2]; |
| 919 | |
| 920 | fix1 = 0.0; |
| 921 | fiy1 = 0.0; |
| 922 | fiz1 = 0.0; |
| 923 | fix2 = 0.0; |
| 924 | fiy2 = 0.0; |
| 925 | fiz2 = 0.0; |
| 926 | fix3 = 0.0; |
| 927 | fiy3 = 0.0; |
| 928 | fiz3 = 0.0; |
| 929 | |
| 930 | /* Start inner kernel loop */ |
| 931 | for(jidx=j_index_start; jidx<j_index_end; jidx++) |
| 932 | { |
| 933 | /* Get j neighbor index, and coordinate index */ |
| 934 | jnr = jjnr[jidx]; |
| 935 | j_coord_offset = DIM3*jnr; |
| 936 | |
| 937 | /* load j atom coordinates */ |
| 938 | jx1 = x[j_coord_offset+DIM3*1+XX0]; |
| 939 | jy1 = x[j_coord_offset+DIM3*1+YY1]; |
| 940 | jz1 = x[j_coord_offset+DIM3*1+ZZ2]; |
| 941 | jx2 = x[j_coord_offset+DIM3*2+XX0]; |
| 942 | jy2 = x[j_coord_offset+DIM3*2+YY1]; |
| 943 | jz2 = x[j_coord_offset+DIM3*2+ZZ2]; |
| 944 | jx3 = x[j_coord_offset+DIM3*3+XX0]; |
| 945 | jy3 = x[j_coord_offset+DIM3*3+YY1]; |
| 946 | jz3 = x[j_coord_offset+DIM3*3+ZZ2]; |
| 947 | |
| 948 | /* Calculate displacement vector */ |
| 949 | dx11 = ix1 - jx1; |
| 950 | dy11 = iy1 - jy1; |
| 951 | dz11 = iz1 - jz1; |
| 952 | dx12 = ix1 - jx2; |
| 953 | dy12 = iy1 - jy2; |
| 954 | dz12 = iz1 - jz2; |
| 955 | dx13 = ix1 - jx3; |
| 956 | dy13 = iy1 - jy3; |
| 957 | dz13 = iz1 - jz3; |
| 958 | dx21 = ix2 - jx1; |
| 959 | dy21 = iy2 - jy1; |
| 960 | dz21 = iz2 - jz1; |
| 961 | dx22 = ix2 - jx2; |
| 962 | dy22 = iy2 - jy2; |
| 963 | dz22 = iz2 - jz2; |
| 964 | dx23 = ix2 - jx3; |
| 965 | dy23 = iy2 - jy3; |
| 966 | dz23 = iz2 - jz3; |
| 967 | dx31 = ix3 - jx1; |
| 968 | dy31 = iy3 - jy1; |
| 969 | dz31 = iz3 - jz1; |
| 970 | dx32 = ix3 - jx2; |
| 971 | dy32 = iy3 - jy2; |
| 972 | dz32 = iz3 - jz2; |
| 973 | dx33 = ix3 - jx3; |
| 974 | dy33 = iy3 - jy3; |
| 975 | dz33 = iz3 - jz3; |
| 976 | |
| 977 | /* Calculate squared distance and things based on it */ |
| 978 | rsq11 = dx11*dx11+dy11*dy11+dz11*dz11; |
| 979 | rsq12 = dx12*dx12+dy12*dy12+dz12*dz12; |
| 980 | rsq13 = dx13*dx13+dy13*dy13+dz13*dz13; |
| 981 | rsq21 = dx21*dx21+dy21*dy21+dz21*dz21; |
| 982 | rsq22 = dx22*dx22+dy22*dy22+dz22*dz22; |
| 983 | rsq23 = dx23*dx23+dy23*dy23+dz23*dz23; |
| 984 | rsq31 = dx31*dx31+dy31*dy31+dz31*dz31; |
| 985 | rsq32 = dx32*dx32+dy32*dy32+dz32*dz32; |
| 986 | rsq33 = dx33*dx33+dy33*dy33+dz33*dz33; |
| 987 | |
| 988 | rinv11 = gmx_invsqrt(rsq11)gmx_software_invsqrt(rsq11); |
| 989 | rinv12 = gmx_invsqrt(rsq12)gmx_software_invsqrt(rsq12); |
| 990 | rinv13 = gmx_invsqrt(rsq13)gmx_software_invsqrt(rsq13); |
| 991 | rinv21 = gmx_invsqrt(rsq21)gmx_software_invsqrt(rsq21); |
| 992 | rinv22 = gmx_invsqrt(rsq22)gmx_software_invsqrt(rsq22); |
| 993 | rinv23 = gmx_invsqrt(rsq23)gmx_software_invsqrt(rsq23); |
| 994 | rinv31 = gmx_invsqrt(rsq31)gmx_software_invsqrt(rsq31); |
| 995 | rinv32 = gmx_invsqrt(rsq32)gmx_software_invsqrt(rsq32); |
| 996 | rinv33 = gmx_invsqrt(rsq33)gmx_software_invsqrt(rsq33); |
| 997 | |
| 998 | rinvsq11 = rinv11*rinv11; |
| 999 | rinvsq12 = rinv12*rinv12; |
| 1000 | rinvsq13 = rinv13*rinv13; |
| 1001 | rinvsq21 = rinv21*rinv21; |
| 1002 | rinvsq22 = rinv22*rinv22; |
| 1003 | rinvsq23 = rinv23*rinv23; |
| 1004 | rinvsq31 = rinv31*rinv31; |
| 1005 | rinvsq32 = rinv32*rinv32; |
| 1006 | rinvsq33 = rinv33*rinv33; |
| 1007 | |
| 1008 | /************************** |
| 1009 | * CALCULATE INTERACTIONS * |
| 1010 | **************************/ |
| 1011 | |
| 1012 | if (rsq11<rcutoff2) |
| 1013 | { |
| 1014 | |
| 1015 | r11 = rsq11*rinv11; |
| 1016 | |
| 1017 | /* EWALD ELECTROSTATICS */ |
| 1018 | |
| 1019 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
| 1020 | ewrt = r11*ewtabscale; |
| 1021 | ewitab = ewrt; |
| 1022 | eweps = ewrt-ewitab; |
| 1023 | ewitab = 4*ewitab; |
| 1024 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
| 1025 | velec = qq11*(rinv11-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
| 1026 | felec = qq11*rinv11*(rinvsq11-felec); |
| 1027 | |
| 1028 | d = r11-rswitch; |
| 1029 | d = (d>0.0) ? d : 0.0; |
| 1030 | d2 = d*d; |
| 1031 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
| 1032 | |
| 1033 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
| 1034 | |
| 1035 | /* Evaluate switch function */ |
| 1036 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
| 1037 | felec = felec*sw - rinv11*velec*dsw; |
| 1038 | |
| 1039 | fscal = felec; |
| 1040 | |
| 1041 | /* Calculate temporary vectorial force */ |
| 1042 | tx = fscal*dx11; |
| 1043 | ty = fscal*dy11; |
| 1044 | tz = fscal*dz11; |
| 1045 | |
| 1046 | /* Update vectorial force */ |
| 1047 | fix1 += tx; |
| 1048 | fiy1 += ty; |
| 1049 | fiz1 += tz; |
| 1050 | f[j_coord_offset+DIM3*1+XX0] -= tx; |
| 1051 | f[j_coord_offset+DIM3*1+YY1] -= ty; |
| 1052 | f[j_coord_offset+DIM3*1+ZZ2] -= tz; |
| 1053 | |
| 1054 | } |
| 1055 | |
| 1056 | /************************** |
| 1057 | * CALCULATE INTERACTIONS * |
| 1058 | **************************/ |
| 1059 | |
| 1060 | if (rsq12<rcutoff2) |
| 1061 | { |
| 1062 | |
| 1063 | r12 = rsq12*rinv12; |
| 1064 | |
| 1065 | /* EWALD ELECTROSTATICS */ |
| 1066 | |
| 1067 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
| 1068 | ewrt = r12*ewtabscale; |
| 1069 | ewitab = ewrt; |
| 1070 | eweps = ewrt-ewitab; |
| 1071 | ewitab = 4*ewitab; |
| 1072 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
| 1073 | velec = qq12*(rinv12-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
| 1074 | felec = qq12*rinv12*(rinvsq12-felec); |
| 1075 | |
| 1076 | d = r12-rswitch; |
| 1077 | d = (d>0.0) ? d : 0.0; |
| 1078 | d2 = d*d; |
| 1079 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
| 1080 | |
| 1081 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
| 1082 | |
| 1083 | /* Evaluate switch function */ |
| 1084 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
| 1085 | felec = felec*sw - rinv12*velec*dsw; |
| 1086 | |
| 1087 | fscal = felec; |
| 1088 | |
| 1089 | /* Calculate temporary vectorial force */ |
| 1090 | tx = fscal*dx12; |
| 1091 | ty = fscal*dy12; |
| 1092 | tz = fscal*dz12; |
| 1093 | |
| 1094 | /* Update vectorial force */ |
| 1095 | fix1 += tx; |
| 1096 | fiy1 += ty; |
| 1097 | fiz1 += tz; |
| 1098 | f[j_coord_offset+DIM3*2+XX0] -= tx; |
| 1099 | f[j_coord_offset+DIM3*2+YY1] -= ty; |
| 1100 | f[j_coord_offset+DIM3*2+ZZ2] -= tz; |
| 1101 | |
| 1102 | } |
| 1103 | |
| 1104 | /************************** |
| 1105 | * CALCULATE INTERACTIONS * |
| 1106 | **************************/ |
| 1107 | |
| 1108 | if (rsq13<rcutoff2) |
| 1109 | { |
| 1110 | |
| 1111 | r13 = rsq13*rinv13; |
| 1112 | |
| 1113 | /* EWALD ELECTROSTATICS */ |
| 1114 | |
| 1115 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
| 1116 | ewrt = r13*ewtabscale; |
| 1117 | ewitab = ewrt; |
| 1118 | eweps = ewrt-ewitab; |
| 1119 | ewitab = 4*ewitab; |
| 1120 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
| 1121 | velec = qq13*(rinv13-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
| 1122 | felec = qq13*rinv13*(rinvsq13-felec); |
| 1123 | |
| 1124 | d = r13-rswitch; |
| 1125 | d = (d>0.0) ? d : 0.0; |
| 1126 | d2 = d*d; |
| 1127 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
| 1128 | |
| 1129 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
| 1130 | |
| 1131 | /* Evaluate switch function */ |
| 1132 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
| 1133 | felec = felec*sw - rinv13*velec*dsw; |
| 1134 | |
| 1135 | fscal = felec; |
| 1136 | |
| 1137 | /* Calculate temporary vectorial force */ |
| 1138 | tx = fscal*dx13; |
| 1139 | ty = fscal*dy13; |
| 1140 | tz = fscal*dz13; |
| 1141 | |
| 1142 | /* Update vectorial force */ |
| 1143 | fix1 += tx; |
| 1144 | fiy1 += ty; |
| 1145 | fiz1 += tz; |
| 1146 | f[j_coord_offset+DIM3*3+XX0] -= tx; |
| 1147 | f[j_coord_offset+DIM3*3+YY1] -= ty; |
| 1148 | f[j_coord_offset+DIM3*3+ZZ2] -= tz; |
| 1149 | |
| 1150 | } |
| 1151 | |
| 1152 | /************************** |
| 1153 | * CALCULATE INTERACTIONS * |
| 1154 | **************************/ |
| 1155 | |
| 1156 | if (rsq21<rcutoff2) |
| 1157 | { |
| 1158 | |
| 1159 | r21 = rsq21*rinv21; |
| 1160 | |
| 1161 | /* EWALD ELECTROSTATICS */ |
| 1162 | |
| 1163 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
| 1164 | ewrt = r21*ewtabscale; |
| 1165 | ewitab = ewrt; |
| 1166 | eweps = ewrt-ewitab; |
| 1167 | ewitab = 4*ewitab; |
| 1168 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
| 1169 | velec = qq21*(rinv21-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
| 1170 | felec = qq21*rinv21*(rinvsq21-felec); |
| 1171 | |
| 1172 | d = r21-rswitch; |
| 1173 | d = (d>0.0) ? d : 0.0; |
| 1174 | d2 = d*d; |
| 1175 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
| 1176 | |
| 1177 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
| 1178 | |
| 1179 | /* Evaluate switch function */ |
| 1180 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
| 1181 | felec = felec*sw - rinv21*velec*dsw; |
| 1182 | |
| 1183 | fscal = felec; |
| 1184 | |
| 1185 | /* Calculate temporary vectorial force */ |
| 1186 | tx = fscal*dx21; |
| 1187 | ty = fscal*dy21; |
| 1188 | tz = fscal*dz21; |
| 1189 | |
| 1190 | /* Update vectorial force */ |
| 1191 | fix2 += tx; |
| 1192 | fiy2 += ty; |
| 1193 | fiz2 += tz; |
| 1194 | f[j_coord_offset+DIM3*1+XX0] -= tx; |
| 1195 | f[j_coord_offset+DIM3*1+YY1] -= ty; |
| 1196 | f[j_coord_offset+DIM3*1+ZZ2] -= tz; |
| 1197 | |
| 1198 | } |
| 1199 | |
| 1200 | /************************** |
| 1201 | * CALCULATE INTERACTIONS * |
| 1202 | **************************/ |
| 1203 | |
| 1204 | if (rsq22<rcutoff2) |
| 1205 | { |
| 1206 | |
| 1207 | r22 = rsq22*rinv22; |
| 1208 | |
| 1209 | /* EWALD ELECTROSTATICS */ |
| 1210 | |
| 1211 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
| 1212 | ewrt = r22*ewtabscale; |
| 1213 | ewitab = ewrt; |
| 1214 | eweps = ewrt-ewitab; |
| 1215 | ewitab = 4*ewitab; |
| 1216 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
| 1217 | velec = qq22*(rinv22-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
| 1218 | felec = qq22*rinv22*(rinvsq22-felec); |
| 1219 | |
| 1220 | d = r22-rswitch; |
| 1221 | d = (d>0.0) ? d : 0.0; |
| 1222 | d2 = d*d; |
| 1223 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
| 1224 | |
| 1225 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
| 1226 | |
| 1227 | /* Evaluate switch function */ |
| 1228 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
| 1229 | felec = felec*sw - rinv22*velec*dsw; |
| 1230 | |
| 1231 | fscal = felec; |
| 1232 | |
| 1233 | /* Calculate temporary vectorial force */ |
| 1234 | tx = fscal*dx22; |
| 1235 | ty = fscal*dy22; |
| 1236 | tz = fscal*dz22; |
| 1237 | |
| 1238 | /* Update vectorial force */ |
| 1239 | fix2 += tx; |
| 1240 | fiy2 += ty; |
| 1241 | fiz2 += tz; |
| 1242 | f[j_coord_offset+DIM3*2+XX0] -= tx; |
| 1243 | f[j_coord_offset+DIM3*2+YY1] -= ty; |
| 1244 | f[j_coord_offset+DIM3*2+ZZ2] -= tz; |
| 1245 | |
| 1246 | } |
| 1247 | |
| 1248 | /************************** |
| 1249 | * CALCULATE INTERACTIONS * |
| 1250 | **************************/ |
| 1251 | |
| 1252 | if (rsq23<rcutoff2) |
| 1253 | { |
| 1254 | |
| 1255 | r23 = rsq23*rinv23; |
| 1256 | |
| 1257 | /* EWALD ELECTROSTATICS */ |
| 1258 | |
| 1259 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
| 1260 | ewrt = r23*ewtabscale; |
| 1261 | ewitab = ewrt; |
| 1262 | eweps = ewrt-ewitab; |
| 1263 | ewitab = 4*ewitab; |
| 1264 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
| 1265 | velec = qq23*(rinv23-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
| 1266 | felec = qq23*rinv23*(rinvsq23-felec); |
| 1267 | |
| 1268 | d = r23-rswitch; |
| 1269 | d = (d>0.0) ? d : 0.0; |
| 1270 | d2 = d*d; |
| 1271 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
| 1272 | |
| 1273 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
| 1274 | |
| 1275 | /* Evaluate switch function */ |
| 1276 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
| 1277 | felec = felec*sw - rinv23*velec*dsw; |
| 1278 | |
| 1279 | fscal = felec; |
| 1280 | |
| 1281 | /* Calculate temporary vectorial force */ |
| 1282 | tx = fscal*dx23; |
| 1283 | ty = fscal*dy23; |
| 1284 | tz = fscal*dz23; |
| 1285 | |
| 1286 | /* Update vectorial force */ |
| 1287 | fix2 += tx; |
| 1288 | fiy2 += ty; |
| 1289 | fiz2 += 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 | } |
| 1295 | |
| 1296 | /************************** |
| 1297 | * CALCULATE INTERACTIONS * |
| 1298 | **************************/ |
| 1299 | |
| 1300 | if (rsq31<rcutoff2) |
| 1301 | { |
| 1302 | |
| 1303 | r31 = rsq31*rinv31; |
| 1304 | |
| 1305 | /* EWALD ELECTROSTATICS */ |
| 1306 | |
| 1307 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
| 1308 | ewrt = r31*ewtabscale; |
| 1309 | ewitab = ewrt; |
| 1310 | eweps = ewrt-ewitab; |
| 1311 | ewitab = 4*ewitab; |
| 1312 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
| 1313 | velec = qq31*(rinv31-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
| 1314 | felec = qq31*rinv31*(rinvsq31-felec); |
| 1315 | |
| 1316 | d = r31-rswitch; |
| 1317 | d = (d>0.0) ? d : 0.0; |
| 1318 | d2 = d*d; |
| 1319 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
| 1320 | |
| 1321 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
| 1322 | |
| 1323 | /* Evaluate switch function */ |
| 1324 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
| 1325 | felec = felec*sw - rinv31*velec*dsw; |
| 1326 | |
| 1327 | fscal = felec; |
| 1328 | |
| 1329 | /* Calculate temporary vectorial force */ |
| 1330 | tx = fscal*dx31; |
| 1331 | ty = fscal*dy31; |
| 1332 | tz = fscal*dz31; |
| 1333 | |
| 1334 | /* Update vectorial force */ |
| 1335 | fix3 += tx; |
| 1336 | fiy3 += ty; |
| 1337 | fiz3 += tz; |
| 1338 | f[j_coord_offset+DIM3*1+XX0] -= tx; |
| 1339 | f[j_coord_offset+DIM3*1+YY1] -= ty; |
| 1340 | f[j_coord_offset+DIM3*1+ZZ2] -= tz; |
| 1341 | |
| 1342 | } |
| 1343 | |
| 1344 | /************************** |
| 1345 | * CALCULATE INTERACTIONS * |
| 1346 | **************************/ |
| 1347 | |
| 1348 | if (rsq32<rcutoff2) |
| 1349 | { |
| 1350 | |
| 1351 | r32 = rsq32*rinv32; |
| 1352 | |
| 1353 | /* EWALD ELECTROSTATICS */ |
| 1354 | |
| 1355 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
| 1356 | ewrt = r32*ewtabscale; |
| 1357 | ewitab = ewrt; |
| 1358 | eweps = ewrt-ewitab; |
| 1359 | ewitab = 4*ewitab; |
| 1360 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
| 1361 | velec = qq32*(rinv32-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
| 1362 | felec = qq32*rinv32*(rinvsq32-felec); |
| 1363 | |
| 1364 | d = r32-rswitch; |
| 1365 | d = (d>0.0) ? d : 0.0; |
| 1366 | d2 = d*d; |
| 1367 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
| 1368 | |
| 1369 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
| 1370 | |
| 1371 | /* Evaluate switch function */ |
| 1372 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
| 1373 | felec = felec*sw - rinv32*velec*dsw; |
| 1374 | |
| 1375 | fscal = felec; |
| 1376 | |
| 1377 | /* Calculate temporary vectorial force */ |
| 1378 | tx = fscal*dx32; |
| 1379 | ty = fscal*dy32; |
| 1380 | tz = fscal*dz32; |
| 1381 | |
| 1382 | /* Update vectorial force */ |
| 1383 | fix3 += tx; |
| 1384 | fiy3 += ty; |
| 1385 | fiz3 += tz; |
| 1386 | f[j_coord_offset+DIM3*2+XX0] -= tx; |
| 1387 | f[j_coord_offset+DIM3*2+YY1] -= ty; |
| 1388 | f[j_coord_offset+DIM3*2+ZZ2] -= tz; |
| 1389 | |
| 1390 | } |
| 1391 | |
| 1392 | /************************** |
| 1393 | * CALCULATE INTERACTIONS * |
| 1394 | **************************/ |
| 1395 | |
| 1396 | if (rsq33<rcutoff2) |
| 1397 | { |
| 1398 | |
| 1399 | r33 = rsq33*rinv33; |
| 1400 | |
| 1401 | /* EWALD ELECTROSTATICS */ |
| 1402 | |
| 1403 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
| 1404 | ewrt = r33*ewtabscale; |
| 1405 | ewitab = ewrt; |
| 1406 | eweps = ewrt-ewitab; |
| 1407 | ewitab = 4*ewitab; |
| 1408 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
| 1409 | velec = qq33*(rinv33-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
| 1410 | felec = qq33*rinv33*(rinvsq33-felec); |
| 1411 | |
| 1412 | d = r33-rswitch; |
| 1413 | d = (d>0.0) ? d : 0.0; |
| 1414 | d2 = d*d; |
| 1415 | sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5)); |
| 1416 | |
| 1417 | dsw = d2*(swF2+d*(swF3+d*swF4)); |
| 1418 | |
| 1419 | /* Evaluate switch function */ |
| 1420 | /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */ |
| 1421 | felec = felec*sw - rinv33*velec*dsw; |
| 1422 | |
| 1423 | fscal = felec; |
| 1424 | |
| 1425 | /* Calculate temporary vectorial force */ |
| 1426 | tx = fscal*dx33; |
| 1427 | ty = fscal*dy33; |
| 1428 | tz = fscal*dz33; |
| 1429 | |
| 1430 | /* Update vectorial force */ |
| 1431 | fix3 += tx; |
| 1432 | fiy3 += ty; |
| 1433 | fiz3 += tz; |
| 1434 | f[j_coord_offset+DIM3*3+XX0] -= tx; |
| 1435 | f[j_coord_offset+DIM3*3+YY1] -= ty; |
| 1436 | f[j_coord_offset+DIM3*3+ZZ2] -= tz; |
| 1437 | |
| 1438 | } |
| 1439 | |
| 1440 | /* Inner loop uses 504 flops */ |
| 1441 | } |
| 1442 | /* End of innermost loop */ |
| 1443 | |
| 1444 | tx = ty = tz = 0; |
| 1445 | f[i_coord_offset+DIM3*1+XX0] += fix1; |
| 1446 | f[i_coord_offset+DIM3*1+YY1] += fiy1; |
| 1447 | f[i_coord_offset+DIM3*1+ZZ2] += fiz1; |
| 1448 | tx += fix1; |
| 1449 | ty += fiy1; |
| 1450 | tz += fiz1; |
| 1451 | f[i_coord_offset+DIM3*2+XX0] += fix2; |
| 1452 | f[i_coord_offset+DIM3*2+YY1] += fiy2; |
| 1453 | f[i_coord_offset+DIM3*2+ZZ2] += fiz2; |
| 1454 | tx += fix2; |
| 1455 | ty += fiy2; |
| 1456 | tz += fiz2; |
| 1457 | f[i_coord_offset+DIM3*3+XX0] += fix3; |
| 1458 | f[i_coord_offset+DIM3*3+YY1] += fiy3; |
| 1459 | f[i_coord_offset+DIM3*3+ZZ2] += fiz3; |
| 1460 | tx += fix3; |
| 1461 | ty += fiy3; |
| 1462 | tz += fiz3; |
| 1463 | fshift[i_shift_offset+XX0] += tx; |
| 1464 | fshift[i_shift_offset+YY1] += ty; |
| 1465 | fshift[i_shift_offset+ZZ2] += tz; |
| 1466 | |
| 1467 | /* Increment number of inner iterations */ |
| 1468 | inneriter += j_index_end - j_index_start; |
| 1469 | |
| 1470 | /* Outer loop uses 30 flops */ |
| 1471 | } |
| 1472 | |
| 1473 | /* Increment number of outer iterations */ |
| 1474 | outeriter += nri; |
| 1475 | |
| 1476 | /* Update outer/inner flops */ |
| 1477 | |
| 1478 | inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_W4W4_F,outeriter*30 + inneriter*504)(nrnb)->n[eNR_NBKERNEL_ELEC_W4W4_F] += outeriter*30 + inneriter *504; |
| 1479 | } |