src/gromacs/gmxlib/nonbonded/nb_kernel_sse2_single/nb_kernel_ElecRFCut_VdwNone_GeomW4P1_sse2_single.c

   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 sse2_single kernel generator.
  37  */
  38 #include "config.h"
  39
  40 #include <math.h>
  41
  42 #include "../nb_kernel.h"
  43 #include "types/simple.h"
  44 #include "gromacs/math/vec.h"
  45 #include "nrnb.h"
  46
  47 #include "gromacs/simd/math_x86_sse2_single.h"
  48 #include "kernelutil_x86_sse2_single.h"
  49
  50 /*
  51  * Gromacs nonbonded kernel:   nb_kernel_ElecRFCut_VdwNone_GeomW4P1_VF_sse2_single
  52  * Electrostatics interaction: ReactionField
  53  * VdW interaction:            None
  54  * Geometry:                   Water4-Particle
  55  * Calculate force/pot:        PotentialAndForce
  56  */
  57 void
  58 nb_kernel_ElecRFCut_VdwNone_GeomW4P1_VF_sse2_single
  59                     (t_nblist                    * gmx_restrict       nlist,
  60                      rvec                        * gmx_restrict          xx,
  61                      rvec                        * gmx_restrict          ff,
  62                      t_forcerec                  * gmx_restrict          fr,
  63                      t_mdatoms                   * gmx_restrict     mdatoms,
  64                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
  65                      t_nrnb                      * gmx_restrict        nrnb)
  66 {
  67     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
  68      * just 0 for non-waters.
  69      * Suffixes A,B,C,D refer to j loop unrolling done with SSE, e.g. for the four different
  70      * jnr indices corresponding to data put in the four positions in the SIMD register.
  71      */
  72     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
  73     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
  74     int              jnrA,jnrB,jnrC,jnrD;
  75     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
  76     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
  77     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  78     real             rcutoff_scalar;
  79     real             *shiftvec,*fshift,*x,*f;
  80     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
  81     real             scratch[4*DIM];
  82     __m128           tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
  83     int              vdwioffset1;
  84     __m128           ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
  85     int              vdwioffset2;
  86     __m128           ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
  87     int              vdwioffset3;
  88     __m128           ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
  89     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
  90     __m128           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  91     __m128           dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
  92     __m128           dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
  93     __m128           dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30;
  94     __m128           velec,felec,velecsum,facel,crf,krf,krf2;
  95     real             *charge;
  96     __m128           dummy_mask,cutoff_mask;
  97     __m128           signbit = _mm_castsi128_ps( _mm_set1_epi32(0x80000000) );
  98     __m128           one     = _mm_set1_ps(1.0);
  99     __m128           two     = _mm_set1_ps(2.0);
 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            = _mm_set1_ps(fr->epsfac);
 112     charge           = mdatoms->chargeA;
 113     krf              = _mm_set1_ps(fr->ic->k_rf);
 114     krf2             = _mm_set1_ps(fr->ic->k_rf*2.0);
 115     crf              = _mm_set1_ps(fr->ic->c_rf);
 116
 117     /* Setup water-specific parameters */
 118     inr              = nlist->iinr[0];
 119     iq1              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+1]));
 120     iq2              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+2]));
 121     iq3              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+3]));
 122
 123     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
 124     rcutoff_scalar   = fr->rcoulomb;
 125     rcutoff          = _mm_set1_ps(rcutoff_scalar);
 126     rcutoff2         = _mm_mul_ps(rcutoff,rcutoff);
 127
 128     /* Avoid stupid compiler warnings */
 129     jnrA = jnrB = jnrC = jnrD = 0;
 130     j_coord_offsetA = 0;
 131     j_coord_offsetB = 0;
 132     j_coord_offsetC = 0;
 133     j_coord_offsetD = 0;
 134
 135     outeriter        = 0;
 136     inneriter        = 0;
 137
 138     for(iidx=0;iidx<4*DIM;iidx++)
 139     {
 140         scratch[iidx] = 0.0;
 141     }
 142
 143     /* Start outer loop over neighborlists */
 144     for(iidx=0; iidx<nri; iidx++)
 145     {
 146         /* Load shift vector for this list */
 147         i_shift_offset   = DIM*shiftidx[iidx];
 148
 149         /* Load limits for loop over neighbors */
 150         j_index_start    = jindex[iidx];
 151         j_index_end      = jindex[iidx+1];
 152
 153         /* Get outer coordinate index */
 154         inr              = iinr[iidx];
 155         i_coord_offset   = DIM*inr;
 156
 157         /* Load i particle coords and add shift vector */
 158         gmx_mm_load_shift_and_3rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset+DIM,
 159                                                  &ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
 160
 161         fix1             = _mm_setzero_ps();
 162         fiy1             = _mm_setzero_ps();
 163         fiz1             = _mm_setzero_ps();
 164         fix2             = _mm_setzero_ps();
 165         fiy2             = _mm_setzero_ps();
 166         fiz2             = _mm_setzero_ps();
 167         fix3             = _mm_setzero_ps();
 168         fiy3             = _mm_setzero_ps();
 169         fiz3             = _mm_setzero_ps();
 170
 171         /* Reset potential sums */
 172         velecsum         = _mm_setzero_ps();
 173
 174         /* Start inner kernel loop */
 175         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 176         {
 177
 178             /* Get j neighbor index, and coordinate index */
 179             jnrA             = jjnr[jidx];
 180             jnrB             = jjnr[jidx+1];
 181             jnrC             = jjnr[jidx+2];
 182             jnrD             = jjnr[jidx+3];
 183             j_coord_offsetA  = DIM*jnrA;
 184             j_coord_offsetB  = DIM*jnrB;
 185             j_coord_offsetC  = DIM*jnrC;
 186             j_coord_offsetD  = DIM*jnrD;
 187
 188             /* load j atom coordinates */
 189             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 190                                               x+j_coord_offsetC,x+j_coord_offsetD,
 191                                               &jx0,&jy0,&jz0);
 192
 193             /* Calculate displacement vector */
 194             dx10             = _mm_sub_ps(ix1,jx0);
 195             dy10             = _mm_sub_ps(iy1,jy0);
 196             dz10             = _mm_sub_ps(iz1,jz0);
 197             dx20             = _mm_sub_ps(ix2,jx0);
 198             dy20             = _mm_sub_ps(iy2,jy0);
 199             dz20             = _mm_sub_ps(iz2,jz0);
 200             dx30             = _mm_sub_ps(ix3,jx0);
 201             dy30             = _mm_sub_ps(iy3,jy0);
 202             dz30             = _mm_sub_ps(iz3,jz0);
 203
 204             /* Calculate squared distance and things based on it */
 205             rsq10            = gmx_mm_calc_rsq_ps(dx10,dy10,dz10);
 206             rsq20            = gmx_mm_calc_rsq_ps(dx20,dy20,dz20);
 207             rsq30            = gmx_mm_calc_rsq_ps(dx30,dy30,dz30);
 208
 209             rinv10           = gmx_mm_invsqrt_ps(rsq10);
 210             rinv20           = gmx_mm_invsqrt_ps(rsq20);
 211             rinv30           = gmx_mm_invsqrt_ps(rsq30);
 212
 213             rinvsq10         = _mm_mul_ps(rinv10,rinv10);
 214             rinvsq20         = _mm_mul_ps(rinv20,rinv20);
 215             rinvsq30         = _mm_mul_ps(rinv30,rinv30);
 216
 217             /* Load parameters for j particles */
 218             jq0              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 219                                                               charge+jnrC+0,charge+jnrD+0);
 220
 221             fjx0             = _mm_setzero_ps();
 222             fjy0             = _mm_setzero_ps();
 223             fjz0             = _mm_setzero_ps();
 224
 225             /**************************
 226              * CALCULATE INTERACTIONS *
 227              **************************/
 228
 229             if (gmx_mm_any_lt(rsq10,rcutoff2))
 230             {
 231
 232             /* Compute parameters for interactions between i and j atoms */
 233             qq10             = _mm_mul_ps(iq1,jq0);
 234
 235             /* REACTION-FIELD ELECTROSTATICS */
 236             velec            = _mm_mul_ps(qq10,_mm_sub_ps(_mm_add_ps(rinv10,_mm_mul_ps(krf,rsq10)),crf));
 237             felec            = _mm_mul_ps(qq10,_mm_sub_ps(_mm_mul_ps(rinv10,rinvsq10),krf2));
 238
 239             cutoff_mask      = _mm_cmplt_ps(rsq10,rcutoff2);
 240
 241             /* Update potential sum for this i atom from the interaction with this j atom. */
 242             velec            = _mm_and_ps(velec,cutoff_mask);
 243             velecsum         = _mm_add_ps(velecsum,velec);
 244
 245             fscal            = felec;
 246
 247             fscal            = _mm_and_ps(fscal,cutoff_mask);
 248
 249             /* Calculate temporary vectorial force */
 250             tx               = _mm_mul_ps(fscal,dx10);
 251             ty               = _mm_mul_ps(fscal,dy10);
 252             tz               = _mm_mul_ps(fscal,dz10);
 253
 254             /* Update vectorial force */
 255             fix1             = _mm_add_ps(fix1,tx);
 256             fiy1             = _mm_add_ps(fiy1,ty);
 257             fiz1             = _mm_add_ps(fiz1,tz);
 258
 259             fjx0             = _mm_add_ps(fjx0,tx);
 260             fjy0             = _mm_add_ps(fjy0,ty);
 261             fjz0             = _mm_add_ps(fjz0,tz);
 262
 263             }
 264
 265             /**************************
 266              * CALCULATE INTERACTIONS *
 267              **************************/
 268
 269             if (gmx_mm_any_lt(rsq20,rcutoff2))
 270             {
 271
 272             /* Compute parameters for interactions between i and j atoms */
 273             qq20             = _mm_mul_ps(iq2,jq0);
 274
 275             /* REACTION-FIELD ELECTROSTATICS */
 276             velec            = _mm_mul_ps(qq20,_mm_sub_ps(_mm_add_ps(rinv20,_mm_mul_ps(krf,rsq20)),crf));
 277             felec            = _mm_mul_ps(qq20,_mm_sub_ps(_mm_mul_ps(rinv20,rinvsq20),krf2));
 278
 279             cutoff_mask      = _mm_cmplt_ps(rsq20,rcutoff2);
 280
 281             /* Update potential sum for this i atom from the interaction with this j atom. */
 282             velec            = _mm_and_ps(velec,cutoff_mask);
 283             velecsum         = _mm_add_ps(velecsum,velec);
 284
 285             fscal            = felec;
 286
 287             fscal            = _mm_and_ps(fscal,cutoff_mask);
 288
 289             /* Calculate temporary vectorial force */
 290             tx               = _mm_mul_ps(fscal,dx20);
 291             ty               = _mm_mul_ps(fscal,dy20);
 292             tz               = _mm_mul_ps(fscal,dz20);
 293
 294             /* Update vectorial force */
 295             fix2             = _mm_add_ps(fix2,tx);
 296             fiy2             = _mm_add_ps(fiy2,ty);
 297             fiz2             = _mm_add_ps(fiz2,tz);
 298
 299             fjx0             = _mm_add_ps(fjx0,tx);
 300             fjy0             = _mm_add_ps(fjy0,ty);
 301             fjz0             = _mm_add_ps(fjz0,tz);
 302
 303             }
 304
 305             /**************************
 306              * CALCULATE INTERACTIONS *
 307              **************************/
 308
 309             if (gmx_mm_any_lt(rsq30,rcutoff2))
 310             {
 311
 312             /* Compute parameters for interactions between i and j atoms */
 313             qq30             = _mm_mul_ps(iq3,jq0);
 314
 315             /* REACTION-FIELD ELECTROSTATICS */
 316             velec            = _mm_mul_ps(qq30,_mm_sub_ps(_mm_add_ps(rinv30,_mm_mul_ps(krf,rsq30)),crf));
 317             felec            = _mm_mul_ps(qq30,_mm_sub_ps(_mm_mul_ps(rinv30,rinvsq30),krf2));
 318
 319             cutoff_mask      = _mm_cmplt_ps(rsq30,rcutoff2);
 320
 321             /* Update potential sum for this i atom from the interaction with this j atom. */
 322             velec            = _mm_and_ps(velec,cutoff_mask);
 323             velecsum         = _mm_add_ps(velecsum,velec);
 324
 325             fscal            = felec;
 326
 327             fscal            = _mm_and_ps(fscal,cutoff_mask);
 328
 329             /* Calculate temporary vectorial force */
 330             tx               = _mm_mul_ps(fscal,dx30);
 331             ty               = _mm_mul_ps(fscal,dy30);
 332             tz               = _mm_mul_ps(fscal,dz30);
 333
 334             /* Update vectorial force */
 335             fix3             = _mm_add_ps(fix3,tx);
 336             fiy3             = _mm_add_ps(fiy3,ty);
 337             fiz3             = _mm_add_ps(fiz3,tz);
 338
 339             fjx0             = _mm_add_ps(fjx0,tx);
 340             fjy0             = _mm_add_ps(fjy0,ty);
 341             fjz0             = _mm_add_ps(fjz0,tz);
 342
 343             }
 344
 345             fjptrA             = f+j_coord_offsetA;
 346             fjptrB             = f+j_coord_offsetB;
 347             fjptrC             = f+j_coord_offsetC;
 348             fjptrD             = f+j_coord_offsetD;
 349
 350             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
 351
 352             /* Inner loop uses 108 flops */
 353         }
 354
 355         if(jidx<j_index_end)
 356         {
 357
 358             /* Get j neighbor index, and coordinate index */
 359             jnrlistA         = jjnr[jidx];
 360             jnrlistB         = jjnr[jidx+1];
 361             jnrlistC         = jjnr[jidx+2];
 362             jnrlistD         = jjnr[jidx+3];
 363             /* Sign of each element will be negative for non-real atoms.
 364              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 365              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
 366              */
 367             dummy_mask = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
 368             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 369             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 370             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 371             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 372             j_coord_offsetA  = DIM*jnrA;
 373             j_coord_offsetB  = DIM*jnrB;
 374             j_coord_offsetC  = DIM*jnrC;
 375             j_coord_offsetD  = DIM*jnrD;
 376
 377             /* load j atom coordinates */
 378             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 379                                               x+j_coord_offsetC,x+j_coord_offsetD,
 380                                               &jx0,&jy0,&jz0);
 381
 382             /* Calculate displacement vector */
 383             dx10             = _mm_sub_ps(ix1,jx0);
 384             dy10             = _mm_sub_ps(iy1,jy0);
 385             dz10             = _mm_sub_ps(iz1,jz0);
 386             dx20             = _mm_sub_ps(ix2,jx0);
 387             dy20             = _mm_sub_ps(iy2,jy0);
 388             dz20             = _mm_sub_ps(iz2,jz0);
 389             dx30             = _mm_sub_ps(ix3,jx0);
 390             dy30             = _mm_sub_ps(iy3,jy0);
 391             dz30             = _mm_sub_ps(iz3,jz0);
 392
 393             /* Calculate squared distance and things based on it */
 394             rsq10            = gmx_mm_calc_rsq_ps(dx10,dy10,dz10);
 395             rsq20            = gmx_mm_calc_rsq_ps(dx20,dy20,dz20);
 396             rsq30            = gmx_mm_calc_rsq_ps(dx30,dy30,dz30);
 397
 398             rinv10           = gmx_mm_invsqrt_ps(rsq10);
 399             rinv20           = gmx_mm_invsqrt_ps(rsq20);
 400             rinv30           = gmx_mm_invsqrt_ps(rsq30);
 401
 402             rinvsq10         = _mm_mul_ps(rinv10,rinv10);
 403             rinvsq20         = _mm_mul_ps(rinv20,rinv20);
 404             rinvsq30         = _mm_mul_ps(rinv30,rinv30);
 405
 406             /* Load parameters for j particles */
 407             jq0              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 408                                                               charge+jnrC+0,charge+jnrD+0);
 409
 410             fjx0             = _mm_setzero_ps();
 411             fjy0             = _mm_setzero_ps();
 412             fjz0             = _mm_setzero_ps();
 413
 414             /**************************
 415              * CALCULATE INTERACTIONS *
 416              **************************/
 417
 418             if (gmx_mm_any_lt(rsq10,rcutoff2))
 419             {
 420
 421             /* Compute parameters for interactions between i and j atoms */
 422             qq10             = _mm_mul_ps(iq1,jq0);
 423
 424             /* REACTION-FIELD ELECTROSTATICS */
 425             velec            = _mm_mul_ps(qq10,_mm_sub_ps(_mm_add_ps(rinv10,_mm_mul_ps(krf,rsq10)),crf));
 426             felec            = _mm_mul_ps(qq10,_mm_sub_ps(_mm_mul_ps(rinv10,rinvsq10),krf2));
 427
 428             cutoff_mask      = _mm_cmplt_ps(rsq10,rcutoff2);
 429
 430             /* Update potential sum for this i atom from the interaction with this j atom. */
 431             velec            = _mm_and_ps(velec,cutoff_mask);
 432             velec            = _mm_andnot_ps(dummy_mask,velec);
 433             velecsum         = _mm_add_ps(velecsum,velec);
 434
 435             fscal            = felec;
 436
 437             fscal            = _mm_and_ps(fscal,cutoff_mask);
 438
 439             fscal            = _mm_andnot_ps(dummy_mask,fscal);
 440
 441             /* Calculate temporary vectorial force */
 442             tx               = _mm_mul_ps(fscal,dx10);
 443             ty               = _mm_mul_ps(fscal,dy10);
 444             tz               = _mm_mul_ps(fscal,dz10);
 445
 446             /* Update vectorial force */
 447             fix1             = _mm_add_ps(fix1,tx);
 448             fiy1             = _mm_add_ps(fiy1,ty);
 449             fiz1             = _mm_add_ps(fiz1,tz);
 450
 451             fjx0             = _mm_add_ps(fjx0,tx);
 452             fjy0             = _mm_add_ps(fjy0,ty);
 453             fjz0             = _mm_add_ps(fjz0,tz);
 454
 455             }
 456
 457             /**************************
 458              * CALCULATE INTERACTIONS *
 459              **************************/
 460
 461             if (gmx_mm_any_lt(rsq20,rcutoff2))
 462             {
 463
 464             /* Compute parameters for interactions between i and j atoms */
 465             qq20             = _mm_mul_ps(iq2,jq0);
 466
 467             /* REACTION-FIELD ELECTROSTATICS */
 468             velec            = _mm_mul_ps(qq20,_mm_sub_ps(_mm_add_ps(rinv20,_mm_mul_ps(krf,rsq20)),crf));
 469             felec            = _mm_mul_ps(qq20,_mm_sub_ps(_mm_mul_ps(rinv20,rinvsq20),krf2));
 470
 471             cutoff_mask      = _mm_cmplt_ps(rsq20,rcutoff2);
 472
 473             /* Update potential sum for this i atom from the interaction with this j atom. */
 474             velec            = _mm_and_ps(velec,cutoff_mask);
 475             velec            = _mm_andnot_ps(dummy_mask,velec);
 476             velecsum         = _mm_add_ps(velecsum,velec);
 477
 478             fscal            = felec;
 479
 480             fscal            = _mm_and_ps(fscal,cutoff_mask);
 481
 482             fscal            = _mm_andnot_ps(dummy_mask,fscal);
 483
 484             /* Calculate temporary vectorial force */
 485             tx               = _mm_mul_ps(fscal,dx20);
 486             ty               = _mm_mul_ps(fscal,dy20);
 487             tz               = _mm_mul_ps(fscal,dz20);
 488
 489             /* Update vectorial force */
 490             fix2             = _mm_add_ps(fix2,tx);
 491             fiy2             = _mm_add_ps(fiy2,ty);
 492             fiz2             = _mm_add_ps(fiz2,tz);
 493
 494             fjx0             = _mm_add_ps(fjx0,tx);
 495             fjy0             = _mm_add_ps(fjy0,ty);
 496             fjz0             = _mm_add_ps(fjz0,tz);
 497
 498             }
 499
 500             /**************************
 501              * CALCULATE INTERACTIONS *
 502              **************************/
 503
 504             if (gmx_mm_any_lt(rsq30,rcutoff2))
 505             {
 506
 507             /* Compute parameters for interactions between i and j atoms */
 508             qq30             = _mm_mul_ps(iq3,jq0);
 509
 510             /* REACTION-FIELD ELECTROSTATICS */
 511             velec            = _mm_mul_ps(qq30,_mm_sub_ps(_mm_add_ps(rinv30,_mm_mul_ps(krf,rsq30)),crf));
 512             felec            = _mm_mul_ps(qq30,_mm_sub_ps(_mm_mul_ps(rinv30,rinvsq30),krf2));
 513
 514             cutoff_mask      = _mm_cmplt_ps(rsq30,rcutoff2);
 515
 516             /* Update potential sum for this i atom from the interaction with this j atom. */
 517             velec            = _mm_and_ps(velec,cutoff_mask);
 518             velec            = _mm_andnot_ps(dummy_mask,velec);
 519             velecsum         = _mm_add_ps(velecsum,velec);
 520
 521             fscal            = felec;
 522
 523             fscal            = _mm_and_ps(fscal,cutoff_mask);
 524
 525             fscal            = _mm_andnot_ps(dummy_mask,fscal);
 526
 527             /* Calculate temporary vectorial force */
 528             tx               = _mm_mul_ps(fscal,dx30);
 529             ty               = _mm_mul_ps(fscal,dy30);
 530             tz               = _mm_mul_ps(fscal,dz30);
 531
 532             /* Update vectorial force */
 533             fix3             = _mm_add_ps(fix3,tx);
 534             fiy3             = _mm_add_ps(fiy3,ty);
 535             fiz3             = _mm_add_ps(fiz3,tz);
 536
 537             fjx0             = _mm_add_ps(fjx0,tx);
 538             fjy0             = _mm_add_ps(fjy0,ty);
 539             fjz0             = _mm_add_ps(fjz0,tz);
 540
 541             }
 542
 543             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 544             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 545             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 546             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 547
 548             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
 549
 550             /* Inner loop uses 108 flops */
 551         }
 552
 553         /* End of innermost loop */
 554
 555         gmx_mm_update_iforce_3atom_swizzle_ps(fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
 556                                               f+i_coord_offset+DIM,fshift+i_shift_offset);
 557
 558         ggid                        = gid[iidx];
 559         /* Update potential energies */
 560         gmx_mm_update_1pot_ps(velecsum,kernel_data->energygrp_elec+ggid);
 561
 562         /* Increment number of inner iterations */
 563         inneriter                  += j_index_end - j_index_start;
 564
 565         /* Outer loop uses 19 flops */
 566     }
 567
 568     /* Increment number of outer iterations */
 569     outeriter        += nri;
 570
 571     /* Update outer/inner flops */
 572
 573     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_W4_VF,outeriter*19 + inneriter*108);
 574 }
 575 /*
 576  * Gromacs nonbonded kernel:   nb_kernel_ElecRFCut_VdwNone_GeomW4P1_F_sse2_single
 577  * Electrostatics interaction: ReactionField
 578  * VdW interaction:            None
 579  * Geometry:                   Water4-Particle
 580  * Calculate force/pot:        Force
 581  */
 582 void
 583 nb_kernel_ElecRFCut_VdwNone_GeomW4P1_F_sse2_single
 584                     (t_nblist                    * gmx_restrict       nlist,
 585                      rvec                        * gmx_restrict          xx,
 586                      rvec                        * gmx_restrict          ff,
 587                      t_forcerec                  * gmx_restrict          fr,
 588                      t_mdatoms                   * gmx_restrict     mdatoms,
 589                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 590                      t_nrnb                      * gmx_restrict        nrnb)
 591 {
 592     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
 593      * just 0 for non-waters.
 594      * Suffixes A,B,C,D refer to j loop unrolling done with SSE, e.g. for the four different
 595      * jnr indices corresponding to data put in the four positions in the SIMD register.
 596      */
 597     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
 598     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
 599     int              jnrA,jnrB,jnrC,jnrD;
 600     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
 601     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
 602     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 603     real             rcutoff_scalar;
 604     real             *shiftvec,*fshift,*x,*f;
 605     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
 606     real             scratch[4*DIM];
 607     __m128           tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
 608     int              vdwioffset1;
 609     __m128           ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
 610     int              vdwioffset2;
 611     __m128           ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
 612     int              vdwioffset3;
 613     __m128           ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
 614     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
 615     __m128           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 616     __m128           dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
 617     __m128           dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
 618     __m128           dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30;
 619     __m128           velec,felec,velecsum,facel,crf,krf,krf2;
 620     real             *charge;
 621     __m128           dummy_mask,cutoff_mask;
 622     __m128           signbit = _mm_castsi128_ps( _mm_set1_epi32(0x80000000) );
 623     __m128           one     = _mm_set1_ps(1.0);
 624     __m128           two     = _mm_set1_ps(2.0);
 625     x                = xx[0];
 626     f                = ff[0];
 627
 628     nri              = nlist->nri;
 629     iinr             = nlist->iinr;
 630     jindex           = nlist->jindex;
 631     jjnr             = nlist->jjnr;
 632     shiftidx         = nlist->shift;
 633     gid              = nlist->gid;
 634     shiftvec         = fr->shift_vec[0];
 635     fshift           = fr->fshift[0];
 636     facel            = _mm_set1_ps(fr->epsfac);
 637     charge           = mdatoms->chargeA;
 638     krf              = _mm_set1_ps(fr->ic->k_rf);
 639     krf2             = _mm_set1_ps(fr->ic->k_rf*2.0);
 640     crf              = _mm_set1_ps(fr->ic->c_rf);
 641
 642     /* Setup water-specific parameters */
 643     inr              = nlist->iinr[0];
 644     iq1              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+1]));
 645     iq2              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+2]));
 646     iq3              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+3]));
 647
 648     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
 649     rcutoff_scalar   = fr->rcoulomb;
 650     rcutoff          = _mm_set1_ps(rcutoff_scalar);
 651     rcutoff2         = _mm_mul_ps(rcutoff,rcutoff);
 652
 653     /* Avoid stupid compiler warnings */
 654     jnrA = jnrB = jnrC = jnrD = 0;
 655     j_coord_offsetA = 0;
 656     j_coord_offsetB = 0;
 657     j_coord_offsetC = 0;
 658     j_coord_offsetD = 0;
 659
 660     outeriter        = 0;
 661     inneriter        = 0;
 662
 663     for(iidx=0;iidx<4*DIM;iidx++)
 664     {
 665         scratch[iidx] = 0.0;
 666     }
 667
 668     /* Start outer loop over neighborlists */
 669     for(iidx=0; iidx<nri; iidx++)
 670     {
 671         /* Load shift vector for this list */
 672         i_shift_offset   = DIM*shiftidx[iidx];
 673
 674         /* Load limits for loop over neighbors */
 675         j_index_start    = jindex[iidx];
 676         j_index_end      = jindex[iidx+1];
 677
 678         /* Get outer coordinate index */
 679         inr              = iinr[iidx];
 680         i_coord_offset   = DIM*inr;
 681
 682         /* Load i particle coords and add shift vector */
 683         gmx_mm_load_shift_and_3rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset+DIM,
 684                                                  &ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
 685
 686         fix1             = _mm_setzero_ps();
 687         fiy1             = _mm_setzero_ps();
 688         fiz1             = _mm_setzero_ps();
 689         fix2             = _mm_setzero_ps();
 690         fiy2             = _mm_setzero_ps();
 691         fiz2             = _mm_setzero_ps();
 692         fix3             = _mm_setzero_ps();
 693         fiy3             = _mm_setzero_ps();
 694         fiz3             = _mm_setzero_ps();
 695
 696         /* Start inner kernel loop */
 697         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 698         {
 699
 700             /* Get j neighbor index, and coordinate index */
 701             jnrA             = jjnr[jidx];
 702             jnrB             = jjnr[jidx+1];
 703             jnrC             = jjnr[jidx+2];
 704             jnrD             = jjnr[jidx+3];
 705             j_coord_offsetA  = DIM*jnrA;
 706             j_coord_offsetB  = DIM*jnrB;
 707             j_coord_offsetC  = DIM*jnrC;
 708             j_coord_offsetD  = DIM*jnrD;
 709
 710             /* load j atom coordinates */
 711             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 712                                               x+j_coord_offsetC,x+j_coord_offsetD,
 713                                               &jx0,&jy0,&jz0);
 714
 715             /* Calculate displacement vector */
 716             dx10             = _mm_sub_ps(ix1,jx0);
 717             dy10             = _mm_sub_ps(iy1,jy0);
 718             dz10             = _mm_sub_ps(iz1,jz0);
 719             dx20             = _mm_sub_ps(ix2,jx0);
 720             dy20             = _mm_sub_ps(iy2,jy0);
 721             dz20             = _mm_sub_ps(iz2,jz0);
 722             dx30             = _mm_sub_ps(ix3,jx0);
 723             dy30             = _mm_sub_ps(iy3,jy0);
 724             dz30             = _mm_sub_ps(iz3,jz0);
 725
 726             /* Calculate squared distance and things based on it */
 727             rsq10            = gmx_mm_calc_rsq_ps(dx10,dy10,dz10);
 728             rsq20            = gmx_mm_calc_rsq_ps(dx20,dy20,dz20);
 729             rsq30            = gmx_mm_calc_rsq_ps(dx30,dy30,dz30);
 730
 731             rinv10           = gmx_mm_invsqrt_ps(rsq10);
 732             rinv20           = gmx_mm_invsqrt_ps(rsq20);
 733             rinv30           = gmx_mm_invsqrt_ps(rsq30);
 734
 735             rinvsq10         = _mm_mul_ps(rinv10,rinv10);
 736             rinvsq20         = _mm_mul_ps(rinv20,rinv20);
 737             rinvsq30         = _mm_mul_ps(rinv30,rinv30);
 738
 739             /* Load parameters for j particles */
 740             jq0              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 741                                                               charge+jnrC+0,charge+jnrD+0);
 742
 743             fjx0             = _mm_setzero_ps();
 744             fjy0             = _mm_setzero_ps();
 745             fjz0             = _mm_setzero_ps();
 746
 747             /**************************
 748              * CALCULATE INTERACTIONS *
 749              **************************/
 750
 751             if (gmx_mm_any_lt(rsq10,rcutoff2))
 752             {
 753
 754             /* Compute parameters for interactions between i and j atoms */
 755             qq10             = _mm_mul_ps(iq1,jq0);
 756
 757             /* REACTION-FIELD ELECTROSTATICS */
 758             felec            = _mm_mul_ps(qq10,_mm_sub_ps(_mm_mul_ps(rinv10,rinvsq10),krf2));
 759
 760             cutoff_mask      = _mm_cmplt_ps(rsq10,rcutoff2);
 761
 762             fscal            = felec;
 763
 764             fscal            = _mm_and_ps(fscal,cutoff_mask);
 765
 766             /* Calculate temporary vectorial force */
 767             tx               = _mm_mul_ps(fscal,dx10);
 768             ty               = _mm_mul_ps(fscal,dy10);
 769             tz               = _mm_mul_ps(fscal,dz10);
 770
 771             /* Update vectorial force */
 772             fix1             = _mm_add_ps(fix1,tx);
 773             fiy1             = _mm_add_ps(fiy1,ty);
 774             fiz1             = _mm_add_ps(fiz1,tz);
 775
 776             fjx0             = _mm_add_ps(fjx0,tx);
 777             fjy0             = _mm_add_ps(fjy0,ty);
 778             fjz0             = _mm_add_ps(fjz0,tz);
 779
 780             }
 781
 782             /**************************
 783              * CALCULATE INTERACTIONS *
 784              **************************/
 785
 786             if (gmx_mm_any_lt(rsq20,rcutoff2))
 787             {
 788
 789             /* Compute parameters for interactions between i and j atoms */
 790             qq20             = _mm_mul_ps(iq2,jq0);
 791
 792             /* REACTION-FIELD ELECTROSTATICS */
 793             felec            = _mm_mul_ps(qq20,_mm_sub_ps(_mm_mul_ps(rinv20,rinvsq20),krf2));
 794
 795             cutoff_mask      = _mm_cmplt_ps(rsq20,rcutoff2);
 796
 797             fscal            = felec;
 798
 799             fscal            = _mm_and_ps(fscal,cutoff_mask);
 800
 801             /* Calculate temporary vectorial force */
 802             tx               = _mm_mul_ps(fscal,dx20);
 803             ty               = _mm_mul_ps(fscal,dy20);
 804             tz               = _mm_mul_ps(fscal,dz20);
 805
 806             /* Update vectorial force */
 807             fix2             = _mm_add_ps(fix2,tx);
 808             fiy2             = _mm_add_ps(fiy2,ty);
 809             fiz2             = _mm_add_ps(fiz2,tz);
 810
 811             fjx0             = _mm_add_ps(fjx0,tx);
 812             fjy0             = _mm_add_ps(fjy0,ty);
 813             fjz0             = _mm_add_ps(fjz0,tz);
 814
 815             }
 816
 817             /**************************
 818              * CALCULATE INTERACTIONS *
 819              **************************/
 820
 821             if (gmx_mm_any_lt(rsq30,rcutoff2))
 822             {
 823
 824             /* Compute parameters for interactions between i and j atoms */
 825             qq30             = _mm_mul_ps(iq3,jq0);
 826
 827             /* REACTION-FIELD ELECTROSTATICS */
 828             felec            = _mm_mul_ps(qq30,_mm_sub_ps(_mm_mul_ps(rinv30,rinvsq30),krf2));
 829
 830             cutoff_mask      = _mm_cmplt_ps(rsq30,rcutoff2);
 831
 832             fscal            = felec;
 833
 834             fscal            = _mm_and_ps(fscal,cutoff_mask);
 835
 836             /* Calculate temporary vectorial force */
 837             tx               = _mm_mul_ps(fscal,dx30);
 838             ty               = _mm_mul_ps(fscal,dy30);
 839             tz               = _mm_mul_ps(fscal,dz30);
 840
 841             /* Update vectorial force */
 842             fix3             = _mm_add_ps(fix3,tx);
 843             fiy3             = _mm_add_ps(fiy3,ty);
 844             fiz3             = _mm_add_ps(fiz3,tz);
 845
 846             fjx0             = _mm_add_ps(fjx0,tx);
 847             fjy0             = _mm_add_ps(fjy0,ty);
 848             fjz0             = _mm_add_ps(fjz0,tz);
 849
 850             }
 851
 852             fjptrA             = f+j_coord_offsetA;
 853             fjptrB             = f+j_coord_offsetB;
 854             fjptrC             = f+j_coord_offsetC;
 855             fjptrD             = f+j_coord_offsetD;
 856
 857             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
 858
 859             /* Inner loop uses 90 flops */
 860         }
 861
 862         if(jidx<j_index_end)
 863         {
 864
 865             /* Get j neighbor index, and coordinate index */
 866             jnrlistA         = jjnr[jidx];
 867             jnrlistB         = jjnr[jidx+1];
 868             jnrlistC         = jjnr[jidx+2];
 869             jnrlistD         = jjnr[jidx+3];
 870             /* Sign of each element will be negative for non-real atoms.
 871              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 872              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
 873              */
 874             dummy_mask = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
 875             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 876             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 877             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 878             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 879             j_coord_offsetA  = DIM*jnrA;
 880             j_coord_offsetB  = DIM*jnrB;
 881             j_coord_offsetC  = DIM*jnrC;
 882             j_coord_offsetD  = DIM*jnrD;
 883
 884             /* load j atom coordinates */
 885             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 886                                               x+j_coord_offsetC,x+j_coord_offsetD,
 887                                               &jx0,&jy0,&jz0);
 888
 889             /* Calculate displacement vector */
 890             dx10             = _mm_sub_ps(ix1,jx0);
 891             dy10             = _mm_sub_ps(iy1,jy0);
 892             dz10             = _mm_sub_ps(iz1,jz0);
 893             dx20             = _mm_sub_ps(ix2,jx0);
 894             dy20             = _mm_sub_ps(iy2,jy0);
 895             dz20             = _mm_sub_ps(iz2,jz0);
 896             dx30             = _mm_sub_ps(ix3,jx0);
 897             dy30             = _mm_sub_ps(iy3,jy0);
 898             dz30             = _mm_sub_ps(iz3,jz0);
 899
 900             /* Calculate squared distance and things based on it */
 901             rsq10            = gmx_mm_calc_rsq_ps(dx10,dy10,dz10);
 902             rsq20            = gmx_mm_calc_rsq_ps(dx20,dy20,dz20);
 903             rsq30            = gmx_mm_calc_rsq_ps(dx30,dy30,dz30);
 904
 905             rinv10           = gmx_mm_invsqrt_ps(rsq10);
 906             rinv20           = gmx_mm_invsqrt_ps(rsq20);
 907             rinv30           = gmx_mm_invsqrt_ps(rsq30);
 908
 909             rinvsq10         = _mm_mul_ps(rinv10,rinv10);
 910             rinvsq20         = _mm_mul_ps(rinv20,rinv20);
 911             rinvsq30         = _mm_mul_ps(rinv30,rinv30);
 912
 913             /* Load parameters for j particles */
 914             jq0              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 915                                                               charge+jnrC+0,charge+jnrD+0);
 916
 917             fjx0             = _mm_setzero_ps();
 918             fjy0             = _mm_setzero_ps();
 919             fjz0             = _mm_setzero_ps();
 920
 921             /**************************
 922              * CALCULATE INTERACTIONS *
 923              **************************/
 924
 925             if (gmx_mm_any_lt(rsq10,rcutoff2))
 926             {
 927
 928             /* Compute parameters for interactions between i and j atoms */
 929             qq10             = _mm_mul_ps(iq1,jq0);
 930
 931             /* REACTION-FIELD ELECTROSTATICS */
 932             felec            = _mm_mul_ps(qq10,_mm_sub_ps(_mm_mul_ps(rinv10,rinvsq10),krf2));
 933
 934             cutoff_mask      = _mm_cmplt_ps(rsq10,rcutoff2);
 935
 936             fscal            = felec;
 937
 938             fscal            = _mm_and_ps(fscal,cutoff_mask);
 939
 940             fscal            = _mm_andnot_ps(dummy_mask,fscal);
 941
 942             /* Calculate temporary vectorial force */
 943             tx               = _mm_mul_ps(fscal,dx10);
 944             ty               = _mm_mul_ps(fscal,dy10);
 945             tz               = _mm_mul_ps(fscal,dz10);
 946
 947             /* Update vectorial force */
 948             fix1             = _mm_add_ps(fix1,tx);
 949             fiy1             = _mm_add_ps(fiy1,ty);
 950             fiz1             = _mm_add_ps(fiz1,tz);
 951
 952             fjx0             = _mm_add_ps(fjx0,tx);
 953             fjy0             = _mm_add_ps(fjy0,ty);
 954             fjz0             = _mm_add_ps(fjz0,tz);
 955
 956             }
 957
 958             /**************************
 959              * CALCULATE INTERACTIONS *
 960              **************************/
 961
 962             if (gmx_mm_any_lt(rsq20,rcutoff2))
 963             {
 964
 965             /* Compute parameters for interactions between i and j atoms */
 966             qq20             = _mm_mul_ps(iq2,jq0);
 967
 968             /* REACTION-FIELD ELECTROSTATICS */
 969             felec            = _mm_mul_ps(qq20,_mm_sub_ps(_mm_mul_ps(rinv20,rinvsq20),krf2));
 970
 971             cutoff_mask      = _mm_cmplt_ps(rsq20,rcutoff2);
 972
 973             fscal            = felec;
 974
 975             fscal            = _mm_and_ps(fscal,cutoff_mask);
 976
 977             fscal            = _mm_andnot_ps(dummy_mask,fscal);
 978
 979             /* Calculate temporary vectorial force */
 980             tx               = _mm_mul_ps(fscal,dx20);
 981             ty               = _mm_mul_ps(fscal,dy20);
 982             tz               = _mm_mul_ps(fscal,dz20);
 983
 984             /* Update vectorial force */
 985             fix2             = _mm_add_ps(fix2,tx);
 986             fiy2             = _mm_add_ps(fiy2,ty);
 987             fiz2             = _mm_add_ps(fiz2,tz);
 988
 989             fjx0             = _mm_add_ps(fjx0,tx);
 990             fjy0             = _mm_add_ps(fjy0,ty);
 991             fjz0             = _mm_add_ps(fjz0,tz);
 992
 993             }
 994
 995             /**************************
 996              * CALCULATE INTERACTIONS *
 997              **************************/
 998
 999             if (gmx_mm_any_lt(rsq30,rcutoff2))
1000             {
1001
1002             /* Compute parameters for interactions between i and j atoms */
1003             qq30             = _mm_mul_ps(iq3,jq0);
1004
1005             /* REACTION-FIELD ELECTROSTATICS */
1006             felec            = _mm_mul_ps(qq30,_mm_sub_ps(_mm_mul_ps(rinv30,rinvsq30),krf2));
1007
1008             cutoff_mask      = _mm_cmplt_ps(rsq30,rcutoff2);
1009
1010             fscal            = felec;
1011
1012             fscal            = _mm_and_ps(fscal,cutoff_mask);
1013
1014             fscal            = _mm_andnot_ps(dummy_mask,fscal);
1015
1016             /* Calculate temporary vectorial force */
1017             tx               = _mm_mul_ps(fscal,dx30);
1018             ty               = _mm_mul_ps(fscal,dy30);
1019             tz               = _mm_mul_ps(fscal,dz30);
1020
1021             /* Update vectorial force */
1022             fix3             = _mm_add_ps(fix3,tx);
1023             fiy3             = _mm_add_ps(fiy3,ty);
1024             fiz3             = _mm_add_ps(fiz3,tz);
1025
1026             fjx0             = _mm_add_ps(fjx0,tx);
1027             fjy0             = _mm_add_ps(fjy0,ty);
1028             fjz0             = _mm_add_ps(fjz0,tz);
1029
1030             }
1031
1032             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
1033             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
1034             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
1035             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
1036
1037             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
1038
1039             /* Inner loop uses 90 flops */
1040         }
1041
1042         /* End of innermost loop */
1043
1044         gmx_mm_update_iforce_3atom_swizzle_ps(fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
1045                                               f+i_coord_offset+DIM,fshift+i_shift_offset);
1046
1047         /* Increment number of inner iterations */
1048         inneriter                  += j_index_end - j_index_start;
1049
1050         /* Outer loop uses 18 flops */
1051     }
1052
1053     /* Increment number of outer iterations */
1054     outeriter        += nri;
1055
1056     /* Update outer/inner flops */
1057
1058     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_W4_F,outeriter*18 + inneriter*90);
1059 }