src/gromacs/gmxlib/nonbonded/nb_kernel_sse2_single/nb_kernel_ElecEwSw_VdwLJSw_GeomW4P1_sse2_single.c

   1 /*
   2  * This file is part of the GROMACS molecular simulation package.
   3  *
   4  * Copyright (c) 2012,2013, 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 #ifdef HAVE_CONFIG_H
  39 #include <config.h>
  40 #endif
  41
  42 #include <math.h>
  43
  44 #include "../nb_kernel.h"
  45 #include "types/simple.h"
  46 #include "vec.h"
  47 #include "nrnb.h"
  48
  49 #include "gmx_math_x86_sse2_single.h"
  50 #include "kernelutil_x86_sse2_single.h"
  51
  52 /*
  53  * Gromacs nonbonded kernel:   nb_kernel_ElecEwSw_VdwLJSw_GeomW4P1_VF_sse2_single
  54  * Electrostatics interaction: Ewald
  55  * VdW interaction:            LennardJones
  56  * Geometry:                   Water4-Particle
  57  * Calculate force/pot:        PotentialAndForce
  58  */
  59 void
  60 nb_kernel_ElecEwSw_VdwLJSw_GeomW4P1_VF_sse2_single
  61                     (t_nblist                    * gmx_restrict       nlist,
  62                      rvec                        * gmx_restrict          xx,
  63                      rvec                        * gmx_restrict          ff,
  64                      t_forcerec                  * gmx_restrict          fr,
  65                      t_mdatoms                   * gmx_restrict     mdatoms,
  66                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
  67                      t_nrnb                      * gmx_restrict        nrnb)
  68 {
  69     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
  70      * just 0 for non-waters.
  71      * Suffixes A,B,C,D refer to j loop unrolling done with SSE, e.g. for the four different
  72      * jnr indices corresponding to data put in the four positions in the SIMD register.
  73      */
  74     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
  75     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
  76     int              jnrA,jnrB,jnrC,jnrD;
  77     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
  78     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
  79     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  80     real             rcutoff_scalar;
  81     real             *shiftvec,*fshift,*x,*f;
  82     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
  83     real             scratch[4*DIM];
  84     __m128           tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
  85     int              vdwioffset0;
  86     __m128           ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
  87     int              vdwioffset1;
  88     __m128           ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
  89     int              vdwioffset2;
  90     __m128           ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
  91     int              vdwioffset3;
  92     __m128           ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
  93     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
  94     __m128           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  95     __m128           dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
  96     __m128           dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
  97     __m128           dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
  98     __m128           dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30;
  99     __m128           velec,felec,velecsum,facel,crf,krf,krf2;
 100     real             *charge;
 101     int              nvdwtype;
 102     __m128           rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 103     int              *vdwtype;
 104     real             *vdwparam;
 105     __m128           one_sixth   = _mm_set1_ps(1.0/6.0);
 106     __m128           one_twelfth = _mm_set1_ps(1.0/12.0);
 107     __m128i          ewitab;
 108     __m128           ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
 109     real             *ewtab;
 110     __m128           rswitch,swV3,swV4,swV5,swF2,swF3,swF4,d,d2,sw,dsw;
 111     real             rswitch_scalar,d_scalar;
 112     __m128           dummy_mask,cutoff_mask;
 113     __m128           signbit = _mm_castsi128_ps( _mm_set1_epi32(0x80000000) );
 114     __m128           one     = _mm_set1_ps(1.0);
 115     __m128           two     = _mm_set1_ps(2.0);
 116     x                = xx[0];
 117     f                = ff[0];
 118
 119     nri              = nlist->nri;
 120     iinr             = nlist->iinr;
 121     jindex           = nlist->jindex;
 122     jjnr             = nlist->jjnr;
 123     shiftidx         = nlist->shift;
 124     gid              = nlist->gid;
 125     shiftvec         = fr->shift_vec[0];
 126     fshift           = fr->fshift[0];
 127     facel            = _mm_set1_ps(fr->epsfac);
 128     charge           = mdatoms->chargeA;
 129     nvdwtype         = fr->ntype;
 130     vdwparam         = fr->nbfp;
 131     vdwtype          = mdatoms->typeA;
 132
 133     sh_ewald         = _mm_set1_ps(fr->ic->sh_ewald);
 134     ewtab            = fr->ic->tabq_coul_FDV0;
 135     ewtabscale       = _mm_set1_ps(fr->ic->tabq_scale);
 136     ewtabhalfspace   = _mm_set1_ps(0.5/fr->ic->tabq_scale);
 137
 138     /* Setup water-specific parameters */
 139     inr              = nlist->iinr[0];
 140     iq1              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+1]));
 141     iq2              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+2]));
 142     iq3              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+3]));
 143     vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 144
 145     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
 146     rcutoff_scalar   = fr->rcoulomb;
 147     rcutoff          = _mm_set1_ps(rcutoff_scalar);
 148     rcutoff2         = _mm_mul_ps(rcutoff,rcutoff);
 149
 150     rswitch_scalar   = fr->rcoulomb_switch;
 151     rswitch          = _mm_set1_ps(rswitch_scalar);
 152     /* Setup switch parameters */
 153     d_scalar         = rcutoff_scalar-rswitch_scalar;
 154     d                = _mm_set1_ps(d_scalar);
 155     swV3             = _mm_set1_ps(-10.0/(d_scalar*d_scalar*d_scalar));
 156     swV4             = _mm_set1_ps( 15.0/(d_scalar*d_scalar*d_scalar*d_scalar));
 157     swV5             = _mm_set1_ps( -6.0/(d_scalar*d_scalar*d_scalar*d_scalar*d_scalar));
 158     swF2             = _mm_set1_ps(-30.0/(d_scalar*d_scalar*d_scalar));
 159     swF3             = _mm_set1_ps( 60.0/(d_scalar*d_scalar*d_scalar*d_scalar));
 160     swF4             = _mm_set1_ps(-30.0/(d_scalar*d_scalar*d_scalar*d_scalar*d_scalar));
 161
 162     /* Avoid stupid compiler warnings */
 163     jnrA = jnrB = jnrC = jnrD = 0;
 164     j_coord_offsetA = 0;
 165     j_coord_offsetB = 0;
 166     j_coord_offsetC = 0;
 167     j_coord_offsetD = 0;
 168
 169     outeriter        = 0;
 170     inneriter        = 0;
 171
 172     for(iidx=0;iidx<4*DIM;iidx++)
 173     {
 174         scratch[iidx] = 0.0;
 175     }
 176
 177     /* Start outer loop over neighborlists */
 178     for(iidx=0; iidx<nri; iidx++)
 179     {
 180         /* Load shift vector for this list */
 181         i_shift_offset   = DIM*shiftidx[iidx];
 182
 183         /* Load limits for loop over neighbors */
 184         j_index_start    = jindex[iidx];
 185         j_index_end      = jindex[iidx+1];
 186
 187         /* Get outer coordinate index */
 188         inr              = iinr[iidx];
 189         i_coord_offset   = DIM*inr;
 190
 191         /* Load i particle coords and add shift vector */
 192         gmx_mm_load_shift_and_4rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,
 193                                                  &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
 194
 195         fix0             = _mm_setzero_ps();
 196         fiy0             = _mm_setzero_ps();
 197         fiz0             = _mm_setzero_ps();
 198         fix1             = _mm_setzero_ps();
 199         fiy1             = _mm_setzero_ps();
 200         fiz1             = _mm_setzero_ps();
 201         fix2             = _mm_setzero_ps();
 202         fiy2             = _mm_setzero_ps();
 203         fiz2             = _mm_setzero_ps();
 204         fix3             = _mm_setzero_ps();
 205         fiy3             = _mm_setzero_ps();
 206         fiz3             = _mm_setzero_ps();
 207
 208         /* Reset potential sums */
 209         velecsum         = _mm_setzero_ps();
 210         vvdwsum          = _mm_setzero_ps();
 211
 212         /* Start inner kernel loop */
 213         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 214         {
 215
 216             /* Get j neighbor index, and coordinate index */
 217             jnrA             = jjnr[jidx];
 218             jnrB             = jjnr[jidx+1];
 219             jnrC             = jjnr[jidx+2];
 220             jnrD             = jjnr[jidx+3];
 221             j_coord_offsetA  = DIM*jnrA;
 222             j_coord_offsetB  = DIM*jnrB;
 223             j_coord_offsetC  = DIM*jnrC;
 224             j_coord_offsetD  = DIM*jnrD;
 225
 226             /* load j atom coordinates */
 227             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 228                                               x+j_coord_offsetC,x+j_coord_offsetD,
 229                                               &jx0,&jy0,&jz0);
 230
 231             /* Calculate displacement vector */
 232             dx00             = _mm_sub_ps(ix0,jx0);
 233             dy00             = _mm_sub_ps(iy0,jy0);
 234             dz00             = _mm_sub_ps(iz0,jz0);
 235             dx10             = _mm_sub_ps(ix1,jx0);
 236             dy10             = _mm_sub_ps(iy1,jy0);
 237             dz10             = _mm_sub_ps(iz1,jz0);
 238             dx20             = _mm_sub_ps(ix2,jx0);
 239             dy20             = _mm_sub_ps(iy2,jy0);
 240             dz20             = _mm_sub_ps(iz2,jz0);
 241             dx30             = _mm_sub_ps(ix3,jx0);
 242             dy30             = _mm_sub_ps(iy3,jy0);
 243             dz30             = _mm_sub_ps(iz3,jz0);
 244
 245             /* Calculate squared distance and things based on it */
 246             rsq00            = gmx_mm_calc_rsq_ps(dx00,dy00,dz00);
 247             rsq10            = gmx_mm_calc_rsq_ps(dx10,dy10,dz10);
 248             rsq20            = gmx_mm_calc_rsq_ps(dx20,dy20,dz20);
 249             rsq30            = gmx_mm_calc_rsq_ps(dx30,dy30,dz30);
 250
 251             rinv00           = gmx_mm_invsqrt_ps(rsq00);
 252             rinv10           = gmx_mm_invsqrt_ps(rsq10);
 253             rinv20           = gmx_mm_invsqrt_ps(rsq20);
 254             rinv30           = gmx_mm_invsqrt_ps(rsq30);
 255
 256             rinvsq00         = _mm_mul_ps(rinv00,rinv00);
 257             rinvsq10         = _mm_mul_ps(rinv10,rinv10);
 258             rinvsq20         = _mm_mul_ps(rinv20,rinv20);
 259             rinvsq30         = _mm_mul_ps(rinv30,rinv30);
 260
 261             /* Load parameters for j particles */
 262             jq0              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 263                                                               charge+jnrC+0,charge+jnrD+0);
 264             vdwjidx0A        = 2*vdwtype[jnrA+0];
 265             vdwjidx0B        = 2*vdwtype[jnrB+0];
 266             vdwjidx0C        = 2*vdwtype[jnrC+0];
 267             vdwjidx0D        = 2*vdwtype[jnrD+0];
 268
 269             fjx0             = _mm_setzero_ps();
 270             fjy0             = _mm_setzero_ps();
 271             fjz0             = _mm_setzero_ps();
 272
 273             /**************************
 274              * CALCULATE INTERACTIONS *
 275              **************************/
 276
 277             if (gmx_mm_any_lt(rsq00,rcutoff2))
 278             {
 279
 280             r00              = _mm_mul_ps(rsq00,rinv00);
 281
 282             /* Compute parameters for interactions between i and j atoms */
 283             gmx_mm_load_4pair_swizzle_ps(vdwparam+vdwioffset0+vdwjidx0A,
 284                                          vdwparam+vdwioffset0+vdwjidx0B,
 285                                          vdwparam+vdwioffset0+vdwjidx0C,
 286                                          vdwparam+vdwioffset0+vdwjidx0D,
 287                                          &c6_00,&c12_00);
 288
 289             /* LENNARD-JONES DISPERSION/REPULSION */
 290
 291             rinvsix          = _mm_mul_ps(_mm_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 292             vvdw6            = _mm_mul_ps(c6_00,rinvsix);
 293             vvdw12           = _mm_mul_ps(c12_00,_mm_mul_ps(rinvsix,rinvsix));
 294             vvdw             = _mm_sub_ps( _mm_mul_ps(vvdw12,one_twelfth) , _mm_mul_ps(vvdw6,one_sixth) );
 295             fvdw             = _mm_mul_ps(_mm_sub_ps(vvdw12,vvdw6),rinvsq00);
 296
 297             d                = _mm_sub_ps(r00,rswitch);
 298             d                = _mm_max_ps(d,_mm_setzero_ps());
 299             d2               = _mm_mul_ps(d,d);
 300             sw               = _mm_add_ps(one,_mm_mul_ps(d2,_mm_mul_ps(d,_mm_add_ps(swV3,_mm_mul_ps(d,_mm_add_ps(swV4,_mm_mul_ps(d,swV5)))))));
 301
 302             dsw              = _mm_mul_ps(d2,_mm_add_ps(swF2,_mm_mul_ps(d,_mm_add_ps(swF3,_mm_mul_ps(d,swF4)))));
 303
 304             /* Evaluate switch function */
 305             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 306             fvdw             = _mm_sub_ps( _mm_mul_ps(fvdw,sw) , _mm_mul_ps(rinv00,_mm_mul_ps(vvdw,dsw)) );
 307             vvdw             = _mm_mul_ps(vvdw,sw);
 308             cutoff_mask      = _mm_cmplt_ps(rsq00,rcutoff2);
 309
 310             /* Update potential sum for this i atom from the interaction with this j atom. */
 311             vvdw             = _mm_and_ps(vvdw,cutoff_mask);
 312             vvdwsum          = _mm_add_ps(vvdwsum,vvdw);
 313
 314             fscal            = fvdw;
 315
 316             fscal            = _mm_and_ps(fscal,cutoff_mask);
 317
 318             /* Calculate temporary vectorial force */
 319             tx               = _mm_mul_ps(fscal,dx00);
 320             ty               = _mm_mul_ps(fscal,dy00);
 321             tz               = _mm_mul_ps(fscal,dz00);
 322
 323             /* Update vectorial force */
 324             fix0             = _mm_add_ps(fix0,tx);
 325             fiy0             = _mm_add_ps(fiy0,ty);
 326             fiz0             = _mm_add_ps(fiz0,tz);
 327
 328             fjx0             = _mm_add_ps(fjx0,tx);
 329             fjy0             = _mm_add_ps(fjy0,ty);
 330             fjz0             = _mm_add_ps(fjz0,tz);
 331
 332             }
 333
 334             /**************************
 335              * CALCULATE INTERACTIONS *
 336              **************************/
 337
 338             if (gmx_mm_any_lt(rsq10,rcutoff2))
 339             {
 340
 341             r10              = _mm_mul_ps(rsq10,rinv10);
 342
 343             /* Compute parameters for interactions between i and j atoms */
 344             qq10             = _mm_mul_ps(iq1,jq0);
 345
 346             /* EWALD ELECTROSTATICS */
 347
 348             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 349             ewrt             = _mm_mul_ps(r10,ewtabscale);
 350             ewitab           = _mm_cvttps_epi32(ewrt);
 351             eweps            = _mm_sub_ps(ewrt,_mm_cvtepi32_ps(ewitab));
 352             ewitab           = _mm_slli_epi32(ewitab,2);
 353             ewtabF           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,0) );
 354             ewtabD           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,1) );
 355             ewtabV           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,2) );
 356             ewtabFn          = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,3) );
 357             _MM_TRANSPOSE4_PS(ewtabF,ewtabD,ewtabV,ewtabFn);
 358             felec            = _mm_add_ps(ewtabF,_mm_mul_ps(eweps,ewtabD));
 359             velec            = _mm_sub_ps(ewtabV,_mm_mul_ps(_mm_mul_ps(ewtabhalfspace,eweps),_mm_add_ps(ewtabF,felec)));
 360             velec            = _mm_mul_ps(qq10,_mm_sub_ps(rinv10,velec));
 361             felec            = _mm_mul_ps(_mm_mul_ps(qq10,rinv10),_mm_sub_ps(rinvsq10,felec));
 362
 363             d                = _mm_sub_ps(r10,rswitch);
 364             d                = _mm_max_ps(d,_mm_setzero_ps());
 365             d2               = _mm_mul_ps(d,d);
 366             sw               = _mm_add_ps(one,_mm_mul_ps(d2,_mm_mul_ps(d,_mm_add_ps(swV3,_mm_mul_ps(d,_mm_add_ps(swV4,_mm_mul_ps(d,swV5)))))));
 367
 368             dsw              = _mm_mul_ps(d2,_mm_add_ps(swF2,_mm_mul_ps(d,_mm_add_ps(swF3,_mm_mul_ps(d,swF4)))));
 369
 370             /* Evaluate switch function */
 371             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 372             felec            = _mm_sub_ps( _mm_mul_ps(felec,sw) , _mm_mul_ps(rinv10,_mm_mul_ps(velec,dsw)) );
 373             velec            = _mm_mul_ps(velec,sw);
 374             cutoff_mask      = _mm_cmplt_ps(rsq10,rcutoff2);
 375
 376             /* Update potential sum for this i atom from the interaction with this j atom. */
 377             velec            = _mm_and_ps(velec,cutoff_mask);
 378             velecsum         = _mm_add_ps(velecsum,velec);
 379
 380             fscal            = felec;
 381
 382             fscal            = _mm_and_ps(fscal,cutoff_mask);
 383
 384             /* Calculate temporary vectorial force */
 385             tx               = _mm_mul_ps(fscal,dx10);
 386             ty               = _mm_mul_ps(fscal,dy10);
 387             tz               = _mm_mul_ps(fscal,dz10);
 388
 389             /* Update vectorial force */
 390             fix1             = _mm_add_ps(fix1,tx);
 391             fiy1             = _mm_add_ps(fiy1,ty);
 392             fiz1             = _mm_add_ps(fiz1,tz);
 393
 394             fjx0             = _mm_add_ps(fjx0,tx);
 395             fjy0             = _mm_add_ps(fjy0,ty);
 396             fjz0             = _mm_add_ps(fjz0,tz);
 397
 398             }
 399
 400             /**************************
 401              * CALCULATE INTERACTIONS *
 402              **************************/
 403
 404             if (gmx_mm_any_lt(rsq20,rcutoff2))
 405             {
 406
 407             r20              = _mm_mul_ps(rsq20,rinv20);
 408
 409             /* Compute parameters for interactions between i and j atoms */
 410             qq20             = _mm_mul_ps(iq2,jq0);
 411
 412             /* EWALD ELECTROSTATICS */
 413
 414             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 415             ewrt             = _mm_mul_ps(r20,ewtabscale);
 416             ewitab           = _mm_cvttps_epi32(ewrt);
 417             eweps            = _mm_sub_ps(ewrt,_mm_cvtepi32_ps(ewitab));
 418             ewitab           = _mm_slli_epi32(ewitab,2);
 419             ewtabF           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,0) );
 420             ewtabD           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,1) );
 421             ewtabV           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,2) );
 422             ewtabFn          = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,3) );
 423             _MM_TRANSPOSE4_PS(ewtabF,ewtabD,ewtabV,ewtabFn);
 424             felec            = _mm_add_ps(ewtabF,_mm_mul_ps(eweps,ewtabD));
 425             velec            = _mm_sub_ps(ewtabV,_mm_mul_ps(_mm_mul_ps(ewtabhalfspace,eweps),_mm_add_ps(ewtabF,felec)));
 426             velec            = _mm_mul_ps(qq20,_mm_sub_ps(rinv20,velec));
 427             felec            = _mm_mul_ps(_mm_mul_ps(qq20,rinv20),_mm_sub_ps(rinvsq20,felec));
 428
 429             d                = _mm_sub_ps(r20,rswitch);
 430             d                = _mm_max_ps(d,_mm_setzero_ps());
 431             d2               = _mm_mul_ps(d,d);
 432             sw               = _mm_add_ps(one,_mm_mul_ps(d2,_mm_mul_ps(d,_mm_add_ps(swV3,_mm_mul_ps(d,_mm_add_ps(swV4,_mm_mul_ps(d,swV5)))))));
 433
 434             dsw              = _mm_mul_ps(d2,_mm_add_ps(swF2,_mm_mul_ps(d,_mm_add_ps(swF3,_mm_mul_ps(d,swF4)))));
 435
 436             /* Evaluate switch function */
 437             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 438             felec            = _mm_sub_ps( _mm_mul_ps(felec,sw) , _mm_mul_ps(rinv20,_mm_mul_ps(velec,dsw)) );
 439             velec            = _mm_mul_ps(velec,sw);
 440             cutoff_mask      = _mm_cmplt_ps(rsq20,rcutoff2);
 441
 442             /* Update potential sum for this i atom from the interaction with this j atom. */
 443             velec            = _mm_and_ps(velec,cutoff_mask);
 444             velecsum         = _mm_add_ps(velecsum,velec);
 445
 446             fscal            = felec;
 447
 448             fscal            = _mm_and_ps(fscal,cutoff_mask);
 449
 450             /* Calculate temporary vectorial force */
 451             tx               = _mm_mul_ps(fscal,dx20);
 452             ty               = _mm_mul_ps(fscal,dy20);
 453             tz               = _mm_mul_ps(fscal,dz20);
 454
 455             /* Update vectorial force */
 456             fix2             = _mm_add_ps(fix2,tx);
 457             fiy2             = _mm_add_ps(fiy2,ty);
 458             fiz2             = _mm_add_ps(fiz2,tz);
 459
 460             fjx0             = _mm_add_ps(fjx0,tx);
 461             fjy0             = _mm_add_ps(fjy0,ty);
 462             fjz0             = _mm_add_ps(fjz0,tz);
 463
 464             }
 465
 466             /**************************
 467              * CALCULATE INTERACTIONS *
 468              **************************/
 469
 470             if (gmx_mm_any_lt(rsq30,rcutoff2))
 471             {
 472
 473             r30              = _mm_mul_ps(rsq30,rinv30);
 474
 475             /* Compute parameters for interactions between i and j atoms */
 476             qq30             = _mm_mul_ps(iq3,jq0);
 477
 478             /* EWALD ELECTROSTATICS */
 479
 480             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 481             ewrt             = _mm_mul_ps(r30,ewtabscale);
 482             ewitab           = _mm_cvttps_epi32(ewrt);
 483             eweps            = _mm_sub_ps(ewrt,_mm_cvtepi32_ps(ewitab));
 484             ewitab           = _mm_slli_epi32(ewitab,2);
 485             ewtabF           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,0) );
 486             ewtabD           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,1) );
 487             ewtabV           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,2) );
 488             ewtabFn          = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,3) );
 489             _MM_TRANSPOSE4_PS(ewtabF,ewtabD,ewtabV,ewtabFn);
 490             felec            = _mm_add_ps(ewtabF,_mm_mul_ps(eweps,ewtabD));
 491             velec            = _mm_sub_ps(ewtabV,_mm_mul_ps(_mm_mul_ps(ewtabhalfspace,eweps),_mm_add_ps(ewtabF,felec)));
 492             velec            = _mm_mul_ps(qq30,_mm_sub_ps(rinv30,velec));
 493             felec            = _mm_mul_ps(_mm_mul_ps(qq30,rinv30),_mm_sub_ps(rinvsq30,felec));
 494
 495             d                = _mm_sub_ps(r30,rswitch);
 496             d                = _mm_max_ps(d,_mm_setzero_ps());
 497             d2               = _mm_mul_ps(d,d);
 498             sw               = _mm_add_ps(one,_mm_mul_ps(d2,_mm_mul_ps(d,_mm_add_ps(swV3,_mm_mul_ps(d,_mm_add_ps(swV4,_mm_mul_ps(d,swV5)))))));
 499
 500             dsw              = _mm_mul_ps(d2,_mm_add_ps(swF2,_mm_mul_ps(d,_mm_add_ps(swF3,_mm_mul_ps(d,swF4)))));
 501
 502             /* Evaluate switch function */
 503             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 504             felec            = _mm_sub_ps( _mm_mul_ps(felec,sw) , _mm_mul_ps(rinv30,_mm_mul_ps(velec,dsw)) );
 505             velec            = _mm_mul_ps(velec,sw);
 506             cutoff_mask      = _mm_cmplt_ps(rsq30,rcutoff2);
 507
 508             /* Update potential sum for this i atom from the interaction with this j atom. */
 509             velec            = _mm_and_ps(velec,cutoff_mask);
 510             velecsum         = _mm_add_ps(velecsum,velec);
 511
 512             fscal            = felec;
 513
 514             fscal            = _mm_and_ps(fscal,cutoff_mask);
 515
 516             /* Calculate temporary vectorial force */
 517             tx               = _mm_mul_ps(fscal,dx30);
 518             ty               = _mm_mul_ps(fscal,dy30);
 519             tz               = _mm_mul_ps(fscal,dz30);
 520
 521             /* Update vectorial force */
 522             fix3             = _mm_add_ps(fix3,tx);
 523             fiy3             = _mm_add_ps(fiy3,ty);
 524             fiz3             = _mm_add_ps(fiz3,tz);
 525
 526             fjx0             = _mm_add_ps(fjx0,tx);
 527             fjy0             = _mm_add_ps(fjy0,ty);
 528             fjz0             = _mm_add_ps(fjz0,tz);
 529
 530             }
 531
 532             fjptrA             = f+j_coord_offsetA;
 533             fjptrB             = f+j_coord_offsetB;
 534             fjptrC             = f+j_coord_offsetC;
 535             fjptrD             = f+j_coord_offsetD;
 536
 537             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
 538
 539             /* Inner loop uses 254 flops */
 540         }
 541
 542         if(jidx<j_index_end)
 543         {
 544
 545             /* Get j neighbor index, and coordinate index */
 546             jnrlistA         = jjnr[jidx];
 547             jnrlistB         = jjnr[jidx+1];
 548             jnrlistC         = jjnr[jidx+2];
 549             jnrlistD         = jjnr[jidx+3];
 550             /* Sign of each element will be negative for non-real atoms.
 551              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 552              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
 553              */
 554             dummy_mask = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
 555             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 556             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 557             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 558             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 559             j_coord_offsetA  = DIM*jnrA;
 560             j_coord_offsetB  = DIM*jnrB;
 561             j_coord_offsetC  = DIM*jnrC;
 562             j_coord_offsetD  = DIM*jnrD;
 563
 564             /* load j atom coordinates */
 565             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 566                                               x+j_coord_offsetC,x+j_coord_offsetD,
 567                                               &jx0,&jy0,&jz0);
 568
 569             /* Calculate displacement vector */
 570             dx00             = _mm_sub_ps(ix0,jx0);
 571             dy00             = _mm_sub_ps(iy0,jy0);
 572             dz00             = _mm_sub_ps(iz0,jz0);
 573             dx10             = _mm_sub_ps(ix1,jx0);
 574             dy10             = _mm_sub_ps(iy1,jy0);
 575             dz10             = _mm_sub_ps(iz1,jz0);
 576             dx20             = _mm_sub_ps(ix2,jx0);
 577             dy20             = _mm_sub_ps(iy2,jy0);
 578             dz20             = _mm_sub_ps(iz2,jz0);
 579             dx30             = _mm_sub_ps(ix3,jx0);
 580             dy30             = _mm_sub_ps(iy3,jy0);
 581             dz30             = _mm_sub_ps(iz3,jz0);
 582
 583             /* Calculate squared distance and things based on it */
 584             rsq00            = gmx_mm_calc_rsq_ps(dx00,dy00,dz00);
 585             rsq10            = gmx_mm_calc_rsq_ps(dx10,dy10,dz10);
 586             rsq20            = gmx_mm_calc_rsq_ps(dx20,dy20,dz20);
 587             rsq30            = gmx_mm_calc_rsq_ps(dx30,dy30,dz30);
 588
 589             rinv00           = gmx_mm_invsqrt_ps(rsq00);
 590             rinv10           = gmx_mm_invsqrt_ps(rsq10);
 591             rinv20           = gmx_mm_invsqrt_ps(rsq20);
 592             rinv30           = gmx_mm_invsqrt_ps(rsq30);
 593
 594             rinvsq00         = _mm_mul_ps(rinv00,rinv00);
 595             rinvsq10         = _mm_mul_ps(rinv10,rinv10);
 596             rinvsq20         = _mm_mul_ps(rinv20,rinv20);
 597             rinvsq30         = _mm_mul_ps(rinv30,rinv30);
 598
 599             /* Load parameters for j particles */
 600             jq0              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 601                                                               charge+jnrC+0,charge+jnrD+0);
 602             vdwjidx0A        = 2*vdwtype[jnrA+0];
 603             vdwjidx0B        = 2*vdwtype[jnrB+0];
 604             vdwjidx0C        = 2*vdwtype[jnrC+0];
 605             vdwjidx0D        = 2*vdwtype[jnrD+0];
 606
 607             fjx0             = _mm_setzero_ps();
 608             fjy0             = _mm_setzero_ps();
 609             fjz0             = _mm_setzero_ps();
 610
 611             /**************************
 612              * CALCULATE INTERACTIONS *
 613              **************************/
 614
 615             if (gmx_mm_any_lt(rsq00,rcutoff2))
 616             {
 617
 618             r00              = _mm_mul_ps(rsq00,rinv00);
 619             r00              = _mm_andnot_ps(dummy_mask,r00);
 620
 621             /* Compute parameters for interactions between i and j atoms */
 622             gmx_mm_load_4pair_swizzle_ps(vdwparam+vdwioffset0+vdwjidx0A,
 623                                          vdwparam+vdwioffset0+vdwjidx0B,
 624                                          vdwparam+vdwioffset0+vdwjidx0C,
 625                                          vdwparam+vdwioffset0+vdwjidx0D,
 626                                          &c6_00,&c12_00);
 627
 628             /* LENNARD-JONES DISPERSION/REPULSION */
 629
 630             rinvsix          = _mm_mul_ps(_mm_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 631             vvdw6            = _mm_mul_ps(c6_00,rinvsix);
 632             vvdw12           = _mm_mul_ps(c12_00,_mm_mul_ps(rinvsix,rinvsix));
 633             vvdw             = _mm_sub_ps( _mm_mul_ps(vvdw12,one_twelfth) , _mm_mul_ps(vvdw6,one_sixth) );
 634             fvdw             = _mm_mul_ps(_mm_sub_ps(vvdw12,vvdw6),rinvsq00);
 635
 636             d                = _mm_sub_ps(r00,rswitch);
 637             d                = _mm_max_ps(d,_mm_setzero_ps());
 638             d2               = _mm_mul_ps(d,d);
 639             sw               = _mm_add_ps(one,_mm_mul_ps(d2,_mm_mul_ps(d,_mm_add_ps(swV3,_mm_mul_ps(d,_mm_add_ps(swV4,_mm_mul_ps(d,swV5)))))));
 640
 641             dsw              = _mm_mul_ps(d2,_mm_add_ps(swF2,_mm_mul_ps(d,_mm_add_ps(swF3,_mm_mul_ps(d,swF4)))));
 642
 643             /* Evaluate switch function */
 644             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 645             fvdw             = _mm_sub_ps( _mm_mul_ps(fvdw,sw) , _mm_mul_ps(rinv00,_mm_mul_ps(vvdw,dsw)) );
 646             vvdw             = _mm_mul_ps(vvdw,sw);
 647             cutoff_mask      = _mm_cmplt_ps(rsq00,rcutoff2);
 648
 649             /* Update potential sum for this i atom from the interaction with this j atom. */
 650             vvdw             = _mm_and_ps(vvdw,cutoff_mask);
 651             vvdw             = _mm_andnot_ps(dummy_mask,vvdw);
 652             vvdwsum          = _mm_add_ps(vvdwsum,vvdw);
 653
 654             fscal            = fvdw;
 655
 656             fscal            = _mm_and_ps(fscal,cutoff_mask);
 657
 658             fscal            = _mm_andnot_ps(dummy_mask,fscal);
 659
 660             /* Calculate temporary vectorial force */
 661             tx               = _mm_mul_ps(fscal,dx00);
 662             ty               = _mm_mul_ps(fscal,dy00);
 663             tz               = _mm_mul_ps(fscal,dz00);
 664
 665             /* Update vectorial force */
 666             fix0             = _mm_add_ps(fix0,tx);
 667             fiy0             = _mm_add_ps(fiy0,ty);
 668             fiz0             = _mm_add_ps(fiz0,tz);
 669
 670             fjx0             = _mm_add_ps(fjx0,tx);
 671             fjy0             = _mm_add_ps(fjy0,ty);
 672             fjz0             = _mm_add_ps(fjz0,tz);
 673
 674             }
 675
 676             /**************************
 677              * CALCULATE INTERACTIONS *
 678              **************************/
 679
 680             if (gmx_mm_any_lt(rsq10,rcutoff2))
 681             {
 682
 683             r10              = _mm_mul_ps(rsq10,rinv10);
 684             r10              = _mm_andnot_ps(dummy_mask,r10);
 685
 686             /* Compute parameters for interactions between i and j atoms */
 687             qq10             = _mm_mul_ps(iq1,jq0);
 688
 689             /* EWALD ELECTROSTATICS */
 690
 691             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 692             ewrt             = _mm_mul_ps(r10,ewtabscale);
 693             ewitab           = _mm_cvttps_epi32(ewrt);
 694             eweps            = _mm_sub_ps(ewrt,_mm_cvtepi32_ps(ewitab));
 695             ewitab           = _mm_slli_epi32(ewitab,2);
 696             ewtabF           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,0) );
 697             ewtabD           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,1) );
 698             ewtabV           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,2) );
 699             ewtabFn          = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,3) );
 700             _MM_TRANSPOSE4_PS(ewtabF,ewtabD,ewtabV,ewtabFn);
 701             felec            = _mm_add_ps(ewtabF,_mm_mul_ps(eweps,ewtabD));
 702             velec            = _mm_sub_ps(ewtabV,_mm_mul_ps(_mm_mul_ps(ewtabhalfspace,eweps),_mm_add_ps(ewtabF,felec)));
 703             velec            = _mm_mul_ps(qq10,_mm_sub_ps(rinv10,velec));
 704             felec            = _mm_mul_ps(_mm_mul_ps(qq10,rinv10),_mm_sub_ps(rinvsq10,felec));
 705
 706             d                = _mm_sub_ps(r10,rswitch);
 707             d                = _mm_max_ps(d,_mm_setzero_ps());
 708             d2               = _mm_mul_ps(d,d);
 709             sw               = _mm_add_ps(one,_mm_mul_ps(d2,_mm_mul_ps(d,_mm_add_ps(swV3,_mm_mul_ps(d,_mm_add_ps(swV4,_mm_mul_ps(d,swV5)))))));
 710
 711             dsw              = _mm_mul_ps(d2,_mm_add_ps(swF2,_mm_mul_ps(d,_mm_add_ps(swF3,_mm_mul_ps(d,swF4)))));
 712
 713             /* Evaluate switch function */
 714             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 715             felec            = _mm_sub_ps( _mm_mul_ps(felec,sw) , _mm_mul_ps(rinv10,_mm_mul_ps(velec,dsw)) );
 716             velec            = _mm_mul_ps(velec,sw);
 717             cutoff_mask      = _mm_cmplt_ps(rsq10,rcutoff2);
 718
 719             /* Update potential sum for this i atom from the interaction with this j atom. */
 720             velec            = _mm_and_ps(velec,cutoff_mask);
 721             velec            = _mm_andnot_ps(dummy_mask,velec);
 722             velecsum         = _mm_add_ps(velecsum,velec);
 723
 724             fscal            = felec;
 725
 726             fscal            = _mm_and_ps(fscal,cutoff_mask);
 727
 728             fscal            = _mm_andnot_ps(dummy_mask,fscal);
 729
 730             /* Calculate temporary vectorial force */
 731             tx               = _mm_mul_ps(fscal,dx10);
 732             ty               = _mm_mul_ps(fscal,dy10);
 733             tz               = _mm_mul_ps(fscal,dz10);
 734
 735             /* Update vectorial force */
 736             fix1             = _mm_add_ps(fix1,tx);
 737             fiy1             = _mm_add_ps(fiy1,ty);
 738             fiz1             = _mm_add_ps(fiz1,tz);
 739
 740             fjx0             = _mm_add_ps(fjx0,tx);
 741             fjy0             = _mm_add_ps(fjy0,ty);
 742             fjz0             = _mm_add_ps(fjz0,tz);
 743
 744             }
 745
 746             /**************************
 747              * CALCULATE INTERACTIONS *
 748              **************************/
 749
 750             if (gmx_mm_any_lt(rsq20,rcutoff2))
 751             {
 752
 753             r20              = _mm_mul_ps(rsq20,rinv20);
 754             r20              = _mm_andnot_ps(dummy_mask,r20);
 755
 756             /* Compute parameters for interactions between i and j atoms */
 757             qq20             = _mm_mul_ps(iq2,jq0);
 758
 759             /* EWALD ELECTROSTATICS */
 760
 761             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 762             ewrt             = _mm_mul_ps(r20,ewtabscale);
 763             ewitab           = _mm_cvttps_epi32(ewrt);
 764             eweps            = _mm_sub_ps(ewrt,_mm_cvtepi32_ps(ewitab));
 765             ewitab           = _mm_slli_epi32(ewitab,2);
 766             ewtabF           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,0) );
 767             ewtabD           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,1) );
 768             ewtabV           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,2) );
 769             ewtabFn          = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,3) );
 770             _MM_TRANSPOSE4_PS(ewtabF,ewtabD,ewtabV,ewtabFn);
 771             felec            = _mm_add_ps(ewtabF,_mm_mul_ps(eweps,ewtabD));
 772             velec            = _mm_sub_ps(ewtabV,_mm_mul_ps(_mm_mul_ps(ewtabhalfspace,eweps),_mm_add_ps(ewtabF,felec)));
 773             velec            = _mm_mul_ps(qq20,_mm_sub_ps(rinv20,velec));
 774             felec            = _mm_mul_ps(_mm_mul_ps(qq20,rinv20),_mm_sub_ps(rinvsq20,felec));
 775
 776             d                = _mm_sub_ps(r20,rswitch);
 777             d                = _mm_max_ps(d,_mm_setzero_ps());
 778             d2               = _mm_mul_ps(d,d);
 779             sw               = _mm_add_ps(one,_mm_mul_ps(d2,_mm_mul_ps(d,_mm_add_ps(swV3,_mm_mul_ps(d,_mm_add_ps(swV4,_mm_mul_ps(d,swV5)))))));
 780
 781             dsw              = _mm_mul_ps(d2,_mm_add_ps(swF2,_mm_mul_ps(d,_mm_add_ps(swF3,_mm_mul_ps(d,swF4)))));
 782
 783             /* Evaluate switch function */
 784             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 785             felec            = _mm_sub_ps( _mm_mul_ps(felec,sw) , _mm_mul_ps(rinv20,_mm_mul_ps(velec,dsw)) );
 786             velec            = _mm_mul_ps(velec,sw);
 787             cutoff_mask      = _mm_cmplt_ps(rsq20,rcutoff2);
 788
 789             /* Update potential sum for this i atom from the interaction with this j atom. */
 790             velec            = _mm_and_ps(velec,cutoff_mask);
 791             velec            = _mm_andnot_ps(dummy_mask,velec);
 792             velecsum         = _mm_add_ps(velecsum,velec);
 793
 794             fscal            = felec;
 795
 796             fscal            = _mm_and_ps(fscal,cutoff_mask);
 797
 798             fscal            = _mm_andnot_ps(dummy_mask,fscal);
 799
 800             /* Calculate temporary vectorial force */
 801             tx               = _mm_mul_ps(fscal,dx20);
 802             ty               = _mm_mul_ps(fscal,dy20);
 803             tz               = _mm_mul_ps(fscal,dz20);
 804
 805             /* Update vectorial force */
 806             fix2             = _mm_add_ps(fix2,tx);
 807             fiy2             = _mm_add_ps(fiy2,ty);
 808             fiz2             = _mm_add_ps(fiz2,tz);
 809
 810             fjx0             = _mm_add_ps(fjx0,tx);
 811             fjy0             = _mm_add_ps(fjy0,ty);
 812             fjz0             = _mm_add_ps(fjz0,tz);
 813
 814             }
 815
 816             /**************************
 817              * CALCULATE INTERACTIONS *
 818              **************************/
 819
 820             if (gmx_mm_any_lt(rsq30,rcutoff2))
 821             {
 822
 823             r30              = _mm_mul_ps(rsq30,rinv30);
 824             r30              = _mm_andnot_ps(dummy_mask,r30);
 825
 826             /* Compute parameters for interactions between i and j atoms */
 827             qq30             = _mm_mul_ps(iq3,jq0);
 828
 829             /* EWALD ELECTROSTATICS */
 830
 831             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 832             ewrt             = _mm_mul_ps(r30,ewtabscale);
 833             ewitab           = _mm_cvttps_epi32(ewrt);
 834             eweps            = _mm_sub_ps(ewrt,_mm_cvtepi32_ps(ewitab));
 835             ewitab           = _mm_slli_epi32(ewitab,2);
 836             ewtabF           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,0) );
 837             ewtabD           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,1) );
 838             ewtabV           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,2) );
 839             ewtabFn          = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,3) );
 840             _MM_TRANSPOSE4_PS(ewtabF,ewtabD,ewtabV,ewtabFn);
 841             felec            = _mm_add_ps(ewtabF,_mm_mul_ps(eweps,ewtabD));
 842             velec            = _mm_sub_ps(ewtabV,_mm_mul_ps(_mm_mul_ps(ewtabhalfspace,eweps),_mm_add_ps(ewtabF,felec)));
 843             velec            = _mm_mul_ps(qq30,_mm_sub_ps(rinv30,velec));
 844             felec            = _mm_mul_ps(_mm_mul_ps(qq30,rinv30),_mm_sub_ps(rinvsq30,felec));
 845
 846             d                = _mm_sub_ps(r30,rswitch);
 847             d                = _mm_max_ps(d,_mm_setzero_ps());
 848             d2               = _mm_mul_ps(d,d);
 849             sw               = _mm_add_ps(one,_mm_mul_ps(d2,_mm_mul_ps(d,_mm_add_ps(swV3,_mm_mul_ps(d,_mm_add_ps(swV4,_mm_mul_ps(d,swV5)))))));
 850
 851             dsw              = _mm_mul_ps(d2,_mm_add_ps(swF2,_mm_mul_ps(d,_mm_add_ps(swF3,_mm_mul_ps(d,swF4)))));
 852
 853             /* Evaluate switch function */
 854             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 855             felec            = _mm_sub_ps( _mm_mul_ps(felec,sw) , _mm_mul_ps(rinv30,_mm_mul_ps(velec,dsw)) );
 856             velec            = _mm_mul_ps(velec,sw);
 857             cutoff_mask      = _mm_cmplt_ps(rsq30,rcutoff2);
 858
 859             /* Update potential sum for this i atom from the interaction with this j atom. */
 860             velec            = _mm_and_ps(velec,cutoff_mask);
 861             velec            = _mm_andnot_ps(dummy_mask,velec);
 862             velecsum         = _mm_add_ps(velecsum,velec);
 863
 864             fscal            = felec;
 865
 866             fscal            = _mm_and_ps(fscal,cutoff_mask);
 867
 868             fscal            = _mm_andnot_ps(dummy_mask,fscal);
 869
 870             /* Calculate temporary vectorial force */
 871             tx               = _mm_mul_ps(fscal,dx30);
 872             ty               = _mm_mul_ps(fscal,dy30);
 873             tz               = _mm_mul_ps(fscal,dz30);
 874
 875             /* Update vectorial force */
 876             fix3             = _mm_add_ps(fix3,tx);
 877             fiy3             = _mm_add_ps(fiy3,ty);
 878             fiz3             = _mm_add_ps(fiz3,tz);
 879
 880             fjx0             = _mm_add_ps(fjx0,tx);
 881             fjy0             = _mm_add_ps(fjy0,ty);
 882             fjz0             = _mm_add_ps(fjz0,tz);
 883
 884             }
 885
 886             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 887             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 888             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 889             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 890
 891             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
 892
 893             /* Inner loop uses 258 flops */
 894         }
 895
 896         /* End of innermost loop */
 897
 898         gmx_mm_update_iforce_4atom_swizzle_ps(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
 899                                               f+i_coord_offset,fshift+i_shift_offset);
 900
 901         ggid                        = gid[iidx];
 902         /* Update potential energies */
 903         gmx_mm_update_1pot_ps(velecsum,kernel_data->energygrp_elec+ggid);
 904         gmx_mm_update_1pot_ps(vvdwsum,kernel_data->energygrp_vdw+ggid);
 905
 906         /* Increment number of inner iterations */
 907         inneriter                  += j_index_end - j_index_start;
 908
 909         /* Outer loop uses 26 flops */
 910     }
 911
 912     /* Increment number of outer iterations */
 913     outeriter        += nri;
 914
 915     /* Update outer/inner flops */
 916
 917     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4_VF,outeriter*26 + inneriter*258);
 918 }
 919 /*
 920  * Gromacs nonbonded kernel:   nb_kernel_ElecEwSw_VdwLJSw_GeomW4P1_F_sse2_single
 921  * Electrostatics interaction: Ewald
 922  * VdW interaction:            LennardJones
 923  * Geometry:                   Water4-Particle
 924  * Calculate force/pot:        Force
 925  */
 926 void
 927 nb_kernel_ElecEwSw_VdwLJSw_GeomW4P1_F_sse2_single
 928                     (t_nblist                    * gmx_restrict       nlist,
 929                      rvec                        * gmx_restrict          xx,
 930                      rvec                        * gmx_restrict          ff,
 931                      t_forcerec                  * gmx_restrict          fr,
 932                      t_mdatoms                   * gmx_restrict     mdatoms,
 933                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 934                      t_nrnb                      * gmx_restrict        nrnb)
 935 {
 936     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
 937      * just 0 for non-waters.
 938      * Suffixes A,B,C,D refer to j loop unrolling done with SSE, e.g. for the four different
 939      * jnr indices corresponding to data put in the four positions in the SIMD register.
 940      */
 941     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
 942     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
 943     int              jnrA,jnrB,jnrC,jnrD;
 944     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
 945     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
 946     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 947     real             rcutoff_scalar;
 948     real             *shiftvec,*fshift,*x,*f;
 949     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
 950     real             scratch[4*DIM];
 951     __m128           tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
 952     int              vdwioffset0;
 953     __m128           ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 954     int              vdwioffset1;
 955     __m128           ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
 956     int              vdwioffset2;
 957     __m128           ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
 958     int              vdwioffset3;
 959     __m128           ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
 960     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
 961     __m128           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 962     __m128           dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
 963     __m128           dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
 964     __m128           dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
 965     __m128           dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30;
 966     __m128           velec,felec,velecsum,facel,crf,krf,krf2;
 967     real             *charge;
 968     int              nvdwtype;
 969     __m128           rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 970     int              *vdwtype;
 971     real             *vdwparam;
 972     __m128           one_sixth   = _mm_set1_ps(1.0/6.0);
 973     __m128           one_twelfth = _mm_set1_ps(1.0/12.0);
 974     __m128i          ewitab;
 975     __m128           ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
 976     real             *ewtab;
 977     __m128           rswitch,swV3,swV4,swV5,swF2,swF3,swF4,d,d2,sw,dsw;
 978     real             rswitch_scalar,d_scalar;
 979     __m128           dummy_mask,cutoff_mask;
 980     __m128           signbit = _mm_castsi128_ps( _mm_set1_epi32(0x80000000) );
 981     __m128           one     = _mm_set1_ps(1.0);
 982     __m128           two     = _mm_set1_ps(2.0);
 983     x                = xx[0];
 984     f                = ff[0];
 985
 986     nri              = nlist->nri;
 987     iinr             = nlist->iinr;
 988     jindex           = nlist->jindex;
 989     jjnr             = nlist->jjnr;
 990     shiftidx         = nlist->shift;
 991     gid              = nlist->gid;
 992     shiftvec         = fr->shift_vec[0];
 993     fshift           = fr->fshift[0];
 994     facel            = _mm_set1_ps(fr->epsfac);
 995     charge           = mdatoms->chargeA;
 996     nvdwtype         = fr->ntype;
 997     vdwparam         = fr->nbfp;
 998     vdwtype          = mdatoms->typeA;
 999
1000     sh_ewald         = _mm_set1_ps(fr->ic->sh_ewald);
1001     ewtab            = fr->ic->tabq_coul_FDV0;
1002     ewtabscale       = _mm_set1_ps(fr->ic->tabq_scale);
1003     ewtabhalfspace   = _mm_set1_ps(0.5/fr->ic->tabq_scale);
1004
1005     /* Setup water-specific parameters */
1006     inr              = nlist->iinr[0];
1007     iq1              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+1]));
1008     iq2              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+2]));
1009     iq3              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+3]));
1010     vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
1011
1012     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
1013     rcutoff_scalar   = fr->rcoulomb;
1014     rcutoff          = _mm_set1_ps(rcutoff_scalar);
1015     rcutoff2         = _mm_mul_ps(rcutoff,rcutoff);
1016
1017     rswitch_scalar   = fr->rcoulomb_switch;
1018     rswitch          = _mm_set1_ps(rswitch_scalar);
1019     /* Setup switch parameters */
1020     d_scalar         = rcutoff_scalar-rswitch_scalar;
1021     d                = _mm_set1_ps(d_scalar);
1022     swV3             = _mm_set1_ps(-10.0/(d_scalar*d_scalar*d_scalar));
1023     swV4             = _mm_set1_ps( 15.0/(d_scalar*d_scalar*d_scalar*d_scalar));
1024     swV5             = _mm_set1_ps( -6.0/(d_scalar*d_scalar*d_scalar*d_scalar*d_scalar));
1025     swF2             = _mm_set1_ps(-30.0/(d_scalar*d_scalar*d_scalar));
1026     swF3             = _mm_set1_ps( 60.0/(d_scalar*d_scalar*d_scalar*d_scalar));
1027     swF4             = _mm_set1_ps(-30.0/(d_scalar*d_scalar*d_scalar*d_scalar*d_scalar));
1028
1029     /* Avoid stupid compiler warnings */
1030     jnrA = jnrB = jnrC = jnrD = 0;
1031     j_coord_offsetA = 0;
1032     j_coord_offsetB = 0;
1033     j_coord_offsetC = 0;
1034     j_coord_offsetD = 0;
1035
1036     outeriter        = 0;
1037     inneriter        = 0;
1038
1039     for(iidx=0;iidx<4*DIM;iidx++)
1040     {
1041         scratch[iidx] = 0.0;
1042     }
1043
1044     /* Start outer loop over neighborlists */
1045     for(iidx=0; iidx<nri; iidx++)
1046     {
1047         /* Load shift vector for this list */
1048         i_shift_offset   = DIM*shiftidx[iidx];
1049
1050         /* Load limits for loop over neighbors */
1051         j_index_start    = jindex[iidx];
1052         j_index_end      = jindex[iidx+1];
1053
1054         /* Get outer coordinate index */
1055         inr              = iinr[iidx];
1056         i_coord_offset   = DIM*inr;
1057
1058         /* Load i particle coords and add shift vector */
1059         gmx_mm_load_shift_and_4rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,
1060                                                  &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
1061
1062         fix0             = _mm_setzero_ps();
1063         fiy0             = _mm_setzero_ps();
1064         fiz0             = _mm_setzero_ps();
1065         fix1             = _mm_setzero_ps();
1066         fiy1             = _mm_setzero_ps();
1067         fiz1             = _mm_setzero_ps();
1068         fix2             = _mm_setzero_ps();
1069         fiy2             = _mm_setzero_ps();
1070         fiz2             = _mm_setzero_ps();
1071         fix3             = _mm_setzero_ps();
1072         fiy3             = _mm_setzero_ps();
1073         fiz3             = _mm_setzero_ps();
1074
1075         /* Start inner kernel loop */
1076         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
1077         {
1078
1079             /* Get j neighbor index, and coordinate index */
1080             jnrA             = jjnr[jidx];
1081             jnrB             = jjnr[jidx+1];
1082             jnrC             = jjnr[jidx+2];
1083             jnrD             = jjnr[jidx+3];
1084             j_coord_offsetA  = DIM*jnrA;
1085             j_coord_offsetB  = DIM*jnrB;
1086             j_coord_offsetC  = DIM*jnrC;
1087             j_coord_offsetD  = DIM*jnrD;
1088
1089             /* load j atom coordinates */
1090             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
1091                                               x+j_coord_offsetC,x+j_coord_offsetD,
1092                                               &jx0,&jy0,&jz0);
1093
1094             /* Calculate displacement vector */
1095             dx00             = _mm_sub_ps(ix0,jx0);
1096             dy00             = _mm_sub_ps(iy0,jy0);
1097             dz00             = _mm_sub_ps(iz0,jz0);
1098             dx10             = _mm_sub_ps(ix1,jx0);
1099             dy10             = _mm_sub_ps(iy1,jy0);
1100             dz10             = _mm_sub_ps(iz1,jz0);
1101             dx20             = _mm_sub_ps(ix2,jx0);
1102             dy20             = _mm_sub_ps(iy2,jy0);
1103             dz20             = _mm_sub_ps(iz2,jz0);
1104             dx30             = _mm_sub_ps(ix3,jx0);
1105             dy30             = _mm_sub_ps(iy3,jy0);
1106             dz30             = _mm_sub_ps(iz3,jz0);
1107
1108             /* Calculate squared distance and things based on it */
1109             rsq00            = gmx_mm_calc_rsq_ps(dx00,dy00,dz00);
1110             rsq10            = gmx_mm_calc_rsq_ps(dx10,dy10,dz10);
1111             rsq20            = gmx_mm_calc_rsq_ps(dx20,dy20,dz20);
1112             rsq30            = gmx_mm_calc_rsq_ps(dx30,dy30,dz30);
1113
1114             rinv00           = gmx_mm_invsqrt_ps(rsq00);
1115             rinv10           = gmx_mm_invsqrt_ps(rsq10);
1116             rinv20           = gmx_mm_invsqrt_ps(rsq20);
1117             rinv30           = gmx_mm_invsqrt_ps(rsq30);
1118
1119             rinvsq00         = _mm_mul_ps(rinv00,rinv00);
1120             rinvsq10         = _mm_mul_ps(rinv10,rinv10);
1121             rinvsq20         = _mm_mul_ps(rinv20,rinv20);
1122             rinvsq30         = _mm_mul_ps(rinv30,rinv30);
1123
1124             /* Load parameters for j particles */
1125             jq0              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
1126                                                               charge+jnrC+0,charge+jnrD+0);
1127             vdwjidx0A        = 2*vdwtype[jnrA+0];
1128             vdwjidx0B        = 2*vdwtype[jnrB+0];
1129             vdwjidx0C        = 2*vdwtype[jnrC+0];
1130             vdwjidx0D        = 2*vdwtype[jnrD+0];
1131
1132             fjx0             = _mm_setzero_ps();
1133             fjy0             = _mm_setzero_ps();
1134             fjz0             = _mm_setzero_ps();
1135
1136             /**************************
1137              * CALCULATE INTERACTIONS *
1138              **************************/
1139
1140             if (gmx_mm_any_lt(rsq00,rcutoff2))
1141             {
1142
1143             r00              = _mm_mul_ps(rsq00,rinv00);
1144
1145             /* Compute parameters for interactions between i and j atoms */
1146             gmx_mm_load_4pair_swizzle_ps(vdwparam+vdwioffset0+vdwjidx0A,
1147                                          vdwparam+vdwioffset0+vdwjidx0B,
1148                                          vdwparam+vdwioffset0+vdwjidx0C,
1149                                          vdwparam+vdwioffset0+vdwjidx0D,
1150                                          &c6_00,&c12_00);
1151
1152             /* LENNARD-JONES DISPERSION/REPULSION */
1153
1154             rinvsix          = _mm_mul_ps(_mm_mul_ps(rinvsq00,rinvsq00),rinvsq00);
1155             vvdw6            = _mm_mul_ps(c6_00,rinvsix);
1156             vvdw12           = _mm_mul_ps(c12_00,_mm_mul_ps(rinvsix,rinvsix));
1157             vvdw             = _mm_sub_ps( _mm_mul_ps(vvdw12,one_twelfth) , _mm_mul_ps(vvdw6,one_sixth) );
1158             fvdw             = _mm_mul_ps(_mm_sub_ps(vvdw12,vvdw6),rinvsq00);
1159
1160             d                = _mm_sub_ps(r00,rswitch);
1161             d                = _mm_max_ps(d,_mm_setzero_ps());
1162             d2               = _mm_mul_ps(d,d);
1163             sw               = _mm_add_ps(one,_mm_mul_ps(d2,_mm_mul_ps(d,_mm_add_ps(swV3,_mm_mul_ps(d,_mm_add_ps(swV4,_mm_mul_ps(d,swV5)))))));
1164
1165             dsw              = _mm_mul_ps(d2,_mm_add_ps(swF2,_mm_mul_ps(d,_mm_add_ps(swF3,_mm_mul_ps(d,swF4)))));
1166
1167             /* Evaluate switch function */
1168             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1169             fvdw             = _mm_sub_ps( _mm_mul_ps(fvdw,sw) , _mm_mul_ps(rinv00,_mm_mul_ps(vvdw,dsw)) );
1170             cutoff_mask      = _mm_cmplt_ps(rsq00,rcutoff2);
1171
1172             fscal            = fvdw;
1173
1174             fscal            = _mm_and_ps(fscal,cutoff_mask);
1175
1176             /* Calculate temporary vectorial force */
1177             tx               = _mm_mul_ps(fscal,dx00);
1178             ty               = _mm_mul_ps(fscal,dy00);
1179             tz               = _mm_mul_ps(fscal,dz00);
1180
1181             /* Update vectorial force */
1182             fix0             = _mm_add_ps(fix0,tx);
1183             fiy0             = _mm_add_ps(fiy0,ty);
1184             fiz0             = _mm_add_ps(fiz0,tz);
1185
1186             fjx0             = _mm_add_ps(fjx0,tx);
1187             fjy0             = _mm_add_ps(fjy0,ty);
1188             fjz0             = _mm_add_ps(fjz0,tz);
1189
1190             }
1191
1192             /**************************
1193              * CALCULATE INTERACTIONS *
1194              **************************/
1195
1196             if (gmx_mm_any_lt(rsq10,rcutoff2))
1197             {
1198
1199             r10              = _mm_mul_ps(rsq10,rinv10);
1200
1201             /* Compute parameters for interactions between i and j atoms */
1202             qq10             = _mm_mul_ps(iq1,jq0);
1203
1204             /* EWALD ELECTROSTATICS */
1205
1206             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1207             ewrt             = _mm_mul_ps(r10,ewtabscale);
1208             ewitab           = _mm_cvttps_epi32(ewrt);
1209             eweps            = _mm_sub_ps(ewrt,_mm_cvtepi32_ps(ewitab));
1210             ewitab           = _mm_slli_epi32(ewitab,2);
1211             ewtabF           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,0) );
1212             ewtabD           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,1) );
1213             ewtabV           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,2) );
1214             ewtabFn          = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,3) );
1215             _MM_TRANSPOSE4_PS(ewtabF,ewtabD,ewtabV,ewtabFn);
1216             felec            = _mm_add_ps(ewtabF,_mm_mul_ps(eweps,ewtabD));
1217             velec            = _mm_sub_ps(ewtabV,_mm_mul_ps(_mm_mul_ps(ewtabhalfspace,eweps),_mm_add_ps(ewtabF,felec)));
1218             velec            = _mm_mul_ps(qq10,_mm_sub_ps(rinv10,velec));
1219             felec            = _mm_mul_ps(_mm_mul_ps(qq10,rinv10),_mm_sub_ps(rinvsq10,felec));
1220
1221             d                = _mm_sub_ps(r10,rswitch);
1222             d                = _mm_max_ps(d,_mm_setzero_ps());
1223             d2               = _mm_mul_ps(d,d);
1224             sw               = _mm_add_ps(one,_mm_mul_ps(d2,_mm_mul_ps(d,_mm_add_ps(swV3,_mm_mul_ps(d,_mm_add_ps(swV4,_mm_mul_ps(d,swV5)))))));
1225
1226             dsw              = _mm_mul_ps(d2,_mm_add_ps(swF2,_mm_mul_ps(d,_mm_add_ps(swF3,_mm_mul_ps(d,swF4)))));
1227
1228             /* Evaluate switch function */
1229             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1230             felec            = _mm_sub_ps( _mm_mul_ps(felec,sw) , _mm_mul_ps(rinv10,_mm_mul_ps(velec,dsw)) );
1231             cutoff_mask      = _mm_cmplt_ps(rsq10,rcutoff2);
1232
1233             fscal            = felec;
1234
1235             fscal            = _mm_and_ps(fscal,cutoff_mask);
1236
1237             /* Calculate temporary vectorial force */
1238             tx               = _mm_mul_ps(fscal,dx10);
1239             ty               = _mm_mul_ps(fscal,dy10);
1240             tz               = _mm_mul_ps(fscal,dz10);
1241
1242             /* Update vectorial force */
1243             fix1             = _mm_add_ps(fix1,tx);
1244             fiy1             = _mm_add_ps(fiy1,ty);
1245             fiz1             = _mm_add_ps(fiz1,tz);
1246
1247             fjx0             = _mm_add_ps(fjx0,tx);
1248             fjy0             = _mm_add_ps(fjy0,ty);
1249             fjz0             = _mm_add_ps(fjz0,tz);
1250
1251             }
1252
1253             /**************************
1254              * CALCULATE INTERACTIONS *
1255              **************************/
1256
1257             if (gmx_mm_any_lt(rsq20,rcutoff2))
1258             {
1259
1260             r20              = _mm_mul_ps(rsq20,rinv20);
1261
1262             /* Compute parameters for interactions between i and j atoms */
1263             qq20             = _mm_mul_ps(iq2,jq0);
1264
1265             /* EWALD ELECTROSTATICS */
1266
1267             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1268             ewrt             = _mm_mul_ps(r20,ewtabscale);
1269             ewitab           = _mm_cvttps_epi32(ewrt);
1270             eweps            = _mm_sub_ps(ewrt,_mm_cvtepi32_ps(ewitab));
1271             ewitab           = _mm_slli_epi32(ewitab,2);
1272             ewtabF           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,0) );
1273             ewtabD           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,1) );
1274             ewtabV           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,2) );
1275             ewtabFn          = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,3) );
1276             _MM_TRANSPOSE4_PS(ewtabF,ewtabD,ewtabV,ewtabFn);
1277             felec            = _mm_add_ps(ewtabF,_mm_mul_ps(eweps,ewtabD));
1278             velec            = _mm_sub_ps(ewtabV,_mm_mul_ps(_mm_mul_ps(ewtabhalfspace,eweps),_mm_add_ps(ewtabF,felec)));
1279             velec            = _mm_mul_ps(qq20,_mm_sub_ps(rinv20,velec));
1280             felec            = _mm_mul_ps(_mm_mul_ps(qq20,rinv20),_mm_sub_ps(rinvsq20,felec));
1281
1282             d                = _mm_sub_ps(r20,rswitch);
1283             d                = _mm_max_ps(d,_mm_setzero_ps());
1284             d2               = _mm_mul_ps(d,d);
1285             sw               = _mm_add_ps(one,_mm_mul_ps(d2,_mm_mul_ps(d,_mm_add_ps(swV3,_mm_mul_ps(d,_mm_add_ps(swV4,_mm_mul_ps(d,swV5)))))));
1286
1287             dsw              = _mm_mul_ps(d2,_mm_add_ps(swF2,_mm_mul_ps(d,_mm_add_ps(swF3,_mm_mul_ps(d,swF4)))));
1288
1289             /* Evaluate switch function */
1290             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1291             felec            = _mm_sub_ps( _mm_mul_ps(felec,sw) , _mm_mul_ps(rinv20,_mm_mul_ps(velec,dsw)) );
1292             cutoff_mask      = _mm_cmplt_ps(rsq20,rcutoff2);
1293
1294             fscal            = felec;
1295
1296             fscal            = _mm_and_ps(fscal,cutoff_mask);
1297
1298             /* Calculate temporary vectorial force */
1299             tx               = _mm_mul_ps(fscal,dx20);
1300             ty               = _mm_mul_ps(fscal,dy20);
1301             tz               = _mm_mul_ps(fscal,dz20);
1302
1303             /* Update vectorial force */
1304             fix2             = _mm_add_ps(fix2,tx);
1305             fiy2             = _mm_add_ps(fiy2,ty);
1306             fiz2             = _mm_add_ps(fiz2,tz);
1307
1308             fjx0             = _mm_add_ps(fjx0,tx);
1309             fjy0             = _mm_add_ps(fjy0,ty);
1310             fjz0             = _mm_add_ps(fjz0,tz);
1311
1312             }
1313
1314             /**************************
1315              * CALCULATE INTERACTIONS *
1316              **************************/
1317
1318             if (gmx_mm_any_lt(rsq30,rcutoff2))
1319             {
1320
1321             r30              = _mm_mul_ps(rsq30,rinv30);
1322
1323             /* Compute parameters for interactions between i and j atoms */
1324             qq30             = _mm_mul_ps(iq3,jq0);
1325
1326             /* EWALD ELECTROSTATICS */
1327
1328             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1329             ewrt             = _mm_mul_ps(r30,ewtabscale);
1330             ewitab           = _mm_cvttps_epi32(ewrt);
1331             eweps            = _mm_sub_ps(ewrt,_mm_cvtepi32_ps(ewitab));
1332             ewitab           = _mm_slli_epi32(ewitab,2);
1333             ewtabF           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,0) );
1334             ewtabD           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,1) );
1335             ewtabV           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,2) );
1336             ewtabFn          = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,3) );
1337             _MM_TRANSPOSE4_PS(ewtabF,ewtabD,ewtabV,ewtabFn);
1338             felec            = _mm_add_ps(ewtabF,_mm_mul_ps(eweps,ewtabD));
1339             velec            = _mm_sub_ps(ewtabV,_mm_mul_ps(_mm_mul_ps(ewtabhalfspace,eweps),_mm_add_ps(ewtabF,felec)));
1340             velec            = _mm_mul_ps(qq30,_mm_sub_ps(rinv30,velec));
1341             felec            = _mm_mul_ps(_mm_mul_ps(qq30,rinv30),_mm_sub_ps(rinvsq30,felec));
1342
1343             d                = _mm_sub_ps(r30,rswitch);
1344             d                = _mm_max_ps(d,_mm_setzero_ps());
1345             d2               = _mm_mul_ps(d,d);
1346             sw               = _mm_add_ps(one,_mm_mul_ps(d2,_mm_mul_ps(d,_mm_add_ps(swV3,_mm_mul_ps(d,_mm_add_ps(swV4,_mm_mul_ps(d,swV5)))))));
1347
1348             dsw              = _mm_mul_ps(d2,_mm_add_ps(swF2,_mm_mul_ps(d,_mm_add_ps(swF3,_mm_mul_ps(d,swF4)))));
1349
1350             /* Evaluate switch function */
1351             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1352             felec            = _mm_sub_ps( _mm_mul_ps(felec,sw) , _mm_mul_ps(rinv30,_mm_mul_ps(velec,dsw)) );
1353             cutoff_mask      = _mm_cmplt_ps(rsq30,rcutoff2);
1354
1355             fscal            = felec;
1356
1357             fscal            = _mm_and_ps(fscal,cutoff_mask);
1358
1359             /* Calculate temporary vectorial force */
1360             tx               = _mm_mul_ps(fscal,dx30);
1361             ty               = _mm_mul_ps(fscal,dy30);
1362             tz               = _mm_mul_ps(fscal,dz30);
1363
1364             /* Update vectorial force */
1365             fix3             = _mm_add_ps(fix3,tx);
1366             fiy3             = _mm_add_ps(fiy3,ty);
1367             fiz3             = _mm_add_ps(fiz3,tz);
1368
1369             fjx0             = _mm_add_ps(fjx0,tx);
1370             fjy0             = _mm_add_ps(fjy0,ty);
1371             fjz0             = _mm_add_ps(fjz0,tz);
1372
1373             }
1374
1375             fjptrA             = f+j_coord_offsetA;
1376             fjptrB             = f+j_coord_offsetB;
1377             fjptrC             = f+j_coord_offsetC;
1378             fjptrD             = f+j_coord_offsetD;
1379
1380             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
1381
1382             /* Inner loop uses 242 flops */
1383         }
1384
1385         if(jidx<j_index_end)
1386         {
1387
1388             /* Get j neighbor index, and coordinate index */
1389             jnrlistA         = jjnr[jidx];
1390             jnrlistB         = jjnr[jidx+1];
1391             jnrlistC         = jjnr[jidx+2];
1392             jnrlistD         = jjnr[jidx+3];
1393             /* Sign of each element will be negative for non-real atoms.
1394              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
1395              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
1396              */
1397             dummy_mask = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
1398             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
1399             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
1400             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
1401             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
1402             j_coord_offsetA  = DIM*jnrA;
1403             j_coord_offsetB  = DIM*jnrB;
1404             j_coord_offsetC  = DIM*jnrC;
1405             j_coord_offsetD  = DIM*jnrD;
1406
1407             /* load j atom coordinates */
1408             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
1409                                               x+j_coord_offsetC,x+j_coord_offsetD,
1410                                               &jx0,&jy0,&jz0);
1411
1412             /* Calculate displacement vector */
1413             dx00             = _mm_sub_ps(ix0,jx0);
1414             dy00             = _mm_sub_ps(iy0,jy0);
1415             dz00             = _mm_sub_ps(iz0,jz0);
1416             dx10             = _mm_sub_ps(ix1,jx0);
1417             dy10             = _mm_sub_ps(iy1,jy0);
1418             dz10             = _mm_sub_ps(iz1,jz0);
1419             dx20             = _mm_sub_ps(ix2,jx0);
1420             dy20             = _mm_sub_ps(iy2,jy0);
1421             dz20             = _mm_sub_ps(iz2,jz0);
1422             dx30             = _mm_sub_ps(ix3,jx0);
1423             dy30             = _mm_sub_ps(iy3,jy0);
1424             dz30             = _mm_sub_ps(iz3,jz0);
1425
1426             /* Calculate squared distance and things based on it */
1427             rsq00            = gmx_mm_calc_rsq_ps(dx00,dy00,dz00);
1428             rsq10            = gmx_mm_calc_rsq_ps(dx10,dy10,dz10);
1429             rsq20            = gmx_mm_calc_rsq_ps(dx20,dy20,dz20);
1430             rsq30            = gmx_mm_calc_rsq_ps(dx30,dy30,dz30);
1431
1432             rinv00           = gmx_mm_invsqrt_ps(rsq00);
1433             rinv10           = gmx_mm_invsqrt_ps(rsq10);
1434             rinv20           = gmx_mm_invsqrt_ps(rsq20);
1435             rinv30           = gmx_mm_invsqrt_ps(rsq30);
1436
1437             rinvsq00         = _mm_mul_ps(rinv00,rinv00);
1438             rinvsq10         = _mm_mul_ps(rinv10,rinv10);
1439             rinvsq20         = _mm_mul_ps(rinv20,rinv20);
1440             rinvsq30         = _mm_mul_ps(rinv30,rinv30);
1441
1442             /* Load parameters for j particles */
1443             jq0              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
1444                                                               charge+jnrC+0,charge+jnrD+0);
1445             vdwjidx0A        = 2*vdwtype[jnrA+0];
1446             vdwjidx0B        = 2*vdwtype[jnrB+0];
1447             vdwjidx0C        = 2*vdwtype[jnrC+0];
1448             vdwjidx0D        = 2*vdwtype[jnrD+0];
1449
1450             fjx0             = _mm_setzero_ps();
1451             fjy0             = _mm_setzero_ps();
1452             fjz0             = _mm_setzero_ps();
1453
1454             /**************************
1455              * CALCULATE INTERACTIONS *
1456              **************************/
1457
1458             if (gmx_mm_any_lt(rsq00,rcutoff2))
1459             {
1460
1461             r00              = _mm_mul_ps(rsq00,rinv00);
1462             r00              = _mm_andnot_ps(dummy_mask,r00);
1463
1464             /* Compute parameters for interactions between i and j atoms */
1465             gmx_mm_load_4pair_swizzle_ps(vdwparam+vdwioffset0+vdwjidx0A,
1466                                          vdwparam+vdwioffset0+vdwjidx0B,
1467                                          vdwparam+vdwioffset0+vdwjidx0C,
1468                                          vdwparam+vdwioffset0+vdwjidx0D,
1469                                          &c6_00,&c12_00);
1470
1471             /* LENNARD-JONES DISPERSION/REPULSION */
1472
1473             rinvsix          = _mm_mul_ps(_mm_mul_ps(rinvsq00,rinvsq00),rinvsq00);
1474             vvdw6            = _mm_mul_ps(c6_00,rinvsix);
1475             vvdw12           = _mm_mul_ps(c12_00,_mm_mul_ps(rinvsix,rinvsix));
1476             vvdw             = _mm_sub_ps( _mm_mul_ps(vvdw12,one_twelfth) , _mm_mul_ps(vvdw6,one_sixth) );
1477             fvdw             = _mm_mul_ps(_mm_sub_ps(vvdw12,vvdw6),rinvsq00);
1478
1479             d                = _mm_sub_ps(r00,rswitch);
1480             d                = _mm_max_ps(d,_mm_setzero_ps());
1481             d2               = _mm_mul_ps(d,d);
1482             sw               = _mm_add_ps(one,_mm_mul_ps(d2,_mm_mul_ps(d,_mm_add_ps(swV3,_mm_mul_ps(d,_mm_add_ps(swV4,_mm_mul_ps(d,swV5)))))));
1483
1484             dsw              = _mm_mul_ps(d2,_mm_add_ps(swF2,_mm_mul_ps(d,_mm_add_ps(swF3,_mm_mul_ps(d,swF4)))));
1485
1486             /* Evaluate switch function */
1487             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1488             fvdw             = _mm_sub_ps( _mm_mul_ps(fvdw,sw) , _mm_mul_ps(rinv00,_mm_mul_ps(vvdw,dsw)) );
1489             cutoff_mask      = _mm_cmplt_ps(rsq00,rcutoff2);
1490
1491             fscal            = fvdw;
1492
1493             fscal            = _mm_and_ps(fscal,cutoff_mask);
1494
1495             fscal            = _mm_andnot_ps(dummy_mask,fscal);
1496
1497             /* Calculate temporary vectorial force */
1498             tx               = _mm_mul_ps(fscal,dx00);
1499             ty               = _mm_mul_ps(fscal,dy00);
1500             tz               = _mm_mul_ps(fscal,dz00);
1501
1502             /* Update vectorial force */
1503             fix0             = _mm_add_ps(fix0,tx);
1504             fiy0             = _mm_add_ps(fiy0,ty);
1505             fiz0             = _mm_add_ps(fiz0,tz);
1506
1507             fjx0             = _mm_add_ps(fjx0,tx);
1508             fjy0             = _mm_add_ps(fjy0,ty);
1509             fjz0             = _mm_add_ps(fjz0,tz);
1510
1511             }
1512
1513             /**************************
1514              * CALCULATE INTERACTIONS *
1515              **************************/
1516
1517             if (gmx_mm_any_lt(rsq10,rcutoff2))
1518             {
1519
1520             r10              = _mm_mul_ps(rsq10,rinv10);
1521             r10              = _mm_andnot_ps(dummy_mask,r10);
1522
1523             /* Compute parameters for interactions between i and j atoms */
1524             qq10             = _mm_mul_ps(iq1,jq0);
1525
1526             /* EWALD ELECTROSTATICS */
1527
1528             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1529             ewrt             = _mm_mul_ps(r10,ewtabscale);
1530             ewitab           = _mm_cvttps_epi32(ewrt);
1531             eweps            = _mm_sub_ps(ewrt,_mm_cvtepi32_ps(ewitab));
1532             ewitab           = _mm_slli_epi32(ewitab,2);
1533             ewtabF           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,0) );
1534             ewtabD           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,1) );
1535             ewtabV           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,2) );
1536             ewtabFn          = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,3) );
1537             _MM_TRANSPOSE4_PS(ewtabF,ewtabD,ewtabV,ewtabFn);
1538             felec            = _mm_add_ps(ewtabF,_mm_mul_ps(eweps,ewtabD));
1539             velec            = _mm_sub_ps(ewtabV,_mm_mul_ps(_mm_mul_ps(ewtabhalfspace,eweps),_mm_add_ps(ewtabF,felec)));
1540             velec            = _mm_mul_ps(qq10,_mm_sub_ps(rinv10,velec));
1541             felec            = _mm_mul_ps(_mm_mul_ps(qq10,rinv10),_mm_sub_ps(rinvsq10,felec));
1542
1543             d                = _mm_sub_ps(r10,rswitch);
1544             d                = _mm_max_ps(d,_mm_setzero_ps());
1545             d2               = _mm_mul_ps(d,d);
1546             sw               = _mm_add_ps(one,_mm_mul_ps(d2,_mm_mul_ps(d,_mm_add_ps(swV3,_mm_mul_ps(d,_mm_add_ps(swV4,_mm_mul_ps(d,swV5)))))));
1547
1548             dsw              = _mm_mul_ps(d2,_mm_add_ps(swF2,_mm_mul_ps(d,_mm_add_ps(swF3,_mm_mul_ps(d,swF4)))));
1549
1550             /* Evaluate switch function */
1551             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1552             felec            = _mm_sub_ps( _mm_mul_ps(felec,sw) , _mm_mul_ps(rinv10,_mm_mul_ps(velec,dsw)) );
1553             cutoff_mask      = _mm_cmplt_ps(rsq10,rcutoff2);
1554
1555             fscal            = felec;
1556
1557             fscal            = _mm_and_ps(fscal,cutoff_mask);
1558
1559             fscal            = _mm_andnot_ps(dummy_mask,fscal);
1560
1561             /* Calculate temporary vectorial force */
1562             tx               = _mm_mul_ps(fscal,dx10);
1563             ty               = _mm_mul_ps(fscal,dy10);
1564             tz               = _mm_mul_ps(fscal,dz10);
1565
1566             /* Update vectorial force */
1567             fix1             = _mm_add_ps(fix1,tx);
1568             fiy1             = _mm_add_ps(fiy1,ty);
1569             fiz1             = _mm_add_ps(fiz1,tz);
1570
1571             fjx0             = _mm_add_ps(fjx0,tx);
1572             fjy0             = _mm_add_ps(fjy0,ty);
1573             fjz0             = _mm_add_ps(fjz0,tz);
1574
1575             }
1576
1577             /**************************
1578              * CALCULATE INTERACTIONS *
1579              **************************/
1580
1581             if (gmx_mm_any_lt(rsq20,rcutoff2))
1582             {
1583
1584             r20              = _mm_mul_ps(rsq20,rinv20);
1585             r20              = _mm_andnot_ps(dummy_mask,r20);
1586
1587             /* Compute parameters for interactions between i and j atoms */
1588             qq20             = _mm_mul_ps(iq2,jq0);
1589
1590             /* EWALD ELECTROSTATICS */
1591
1592             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1593             ewrt             = _mm_mul_ps(r20,ewtabscale);
1594             ewitab           = _mm_cvttps_epi32(ewrt);
1595             eweps            = _mm_sub_ps(ewrt,_mm_cvtepi32_ps(ewitab));
1596             ewitab           = _mm_slli_epi32(ewitab,2);
1597             ewtabF           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,0) );
1598             ewtabD           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,1) );
1599             ewtabV           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,2) );
1600             ewtabFn          = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,3) );
1601             _MM_TRANSPOSE4_PS(ewtabF,ewtabD,ewtabV,ewtabFn);
1602             felec            = _mm_add_ps(ewtabF,_mm_mul_ps(eweps,ewtabD));
1603             velec            = _mm_sub_ps(ewtabV,_mm_mul_ps(_mm_mul_ps(ewtabhalfspace,eweps),_mm_add_ps(ewtabF,felec)));
1604             velec            = _mm_mul_ps(qq20,_mm_sub_ps(rinv20,velec));
1605             felec            = _mm_mul_ps(_mm_mul_ps(qq20,rinv20),_mm_sub_ps(rinvsq20,felec));
1606
1607             d                = _mm_sub_ps(r20,rswitch);
1608             d                = _mm_max_ps(d,_mm_setzero_ps());
1609             d2               = _mm_mul_ps(d,d);
1610             sw               = _mm_add_ps(one,_mm_mul_ps(d2,_mm_mul_ps(d,_mm_add_ps(swV3,_mm_mul_ps(d,_mm_add_ps(swV4,_mm_mul_ps(d,swV5)))))));
1611
1612             dsw              = _mm_mul_ps(d2,_mm_add_ps(swF2,_mm_mul_ps(d,_mm_add_ps(swF3,_mm_mul_ps(d,swF4)))));
1613
1614             /* Evaluate switch function */
1615             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1616             felec            = _mm_sub_ps( _mm_mul_ps(felec,sw) , _mm_mul_ps(rinv20,_mm_mul_ps(velec,dsw)) );
1617             cutoff_mask      = _mm_cmplt_ps(rsq20,rcutoff2);
1618
1619             fscal            = felec;
1620
1621             fscal            = _mm_and_ps(fscal,cutoff_mask);
1622
1623             fscal            = _mm_andnot_ps(dummy_mask,fscal);
1624
1625             /* Calculate temporary vectorial force */
1626             tx               = _mm_mul_ps(fscal,dx20);
1627             ty               = _mm_mul_ps(fscal,dy20);
1628             tz               = _mm_mul_ps(fscal,dz20);
1629
1630             /* Update vectorial force */
1631             fix2             = _mm_add_ps(fix2,tx);
1632             fiy2             = _mm_add_ps(fiy2,ty);
1633             fiz2             = _mm_add_ps(fiz2,tz);
1634
1635             fjx0             = _mm_add_ps(fjx0,tx);
1636             fjy0             = _mm_add_ps(fjy0,ty);
1637             fjz0             = _mm_add_ps(fjz0,tz);
1638
1639             }
1640
1641             /**************************
1642              * CALCULATE INTERACTIONS *
1643              **************************/
1644
1645             if (gmx_mm_any_lt(rsq30,rcutoff2))
1646             {
1647
1648             r30              = _mm_mul_ps(rsq30,rinv30);
1649             r30              = _mm_andnot_ps(dummy_mask,r30);
1650
1651             /* Compute parameters for interactions between i and j atoms */
1652             qq30             = _mm_mul_ps(iq3,jq0);
1653
1654             /* EWALD ELECTROSTATICS */
1655
1656             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1657             ewrt             = _mm_mul_ps(r30,ewtabscale);
1658             ewitab           = _mm_cvttps_epi32(ewrt);
1659             eweps            = _mm_sub_ps(ewrt,_mm_cvtepi32_ps(ewitab));
1660             ewitab           = _mm_slli_epi32(ewitab,2);
1661             ewtabF           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,0) );
1662             ewtabD           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,1) );
1663             ewtabV           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,2) );
1664             ewtabFn          = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,3) );
1665             _MM_TRANSPOSE4_PS(ewtabF,ewtabD,ewtabV,ewtabFn);
1666             felec            = _mm_add_ps(ewtabF,_mm_mul_ps(eweps,ewtabD));
1667             velec            = _mm_sub_ps(ewtabV,_mm_mul_ps(_mm_mul_ps(ewtabhalfspace,eweps),_mm_add_ps(ewtabF,felec)));
1668             velec            = _mm_mul_ps(qq30,_mm_sub_ps(rinv30,velec));
1669             felec            = _mm_mul_ps(_mm_mul_ps(qq30,rinv30),_mm_sub_ps(rinvsq30,felec));
1670
1671             d                = _mm_sub_ps(r30,rswitch);
1672             d                = _mm_max_ps(d,_mm_setzero_ps());
1673             d2               = _mm_mul_ps(d,d);
1674             sw               = _mm_add_ps(one,_mm_mul_ps(d2,_mm_mul_ps(d,_mm_add_ps(swV3,_mm_mul_ps(d,_mm_add_ps(swV4,_mm_mul_ps(d,swV5)))))));
1675
1676             dsw              = _mm_mul_ps(d2,_mm_add_ps(swF2,_mm_mul_ps(d,_mm_add_ps(swF3,_mm_mul_ps(d,swF4)))));
1677
1678             /* Evaluate switch function */
1679             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1680             felec            = _mm_sub_ps( _mm_mul_ps(felec,sw) , _mm_mul_ps(rinv30,_mm_mul_ps(velec,dsw)) );
1681             cutoff_mask      = _mm_cmplt_ps(rsq30,rcutoff2);
1682
1683             fscal            = felec;
1684
1685             fscal            = _mm_and_ps(fscal,cutoff_mask);
1686
1687             fscal            = _mm_andnot_ps(dummy_mask,fscal);
1688
1689             /* Calculate temporary vectorial force */
1690             tx               = _mm_mul_ps(fscal,dx30);
1691             ty               = _mm_mul_ps(fscal,dy30);
1692             tz               = _mm_mul_ps(fscal,dz30);
1693
1694             /* Update vectorial force */
1695             fix3             = _mm_add_ps(fix3,tx);
1696             fiy3             = _mm_add_ps(fiy3,ty);
1697             fiz3             = _mm_add_ps(fiz3,tz);
1698
1699             fjx0             = _mm_add_ps(fjx0,tx);
1700             fjy0             = _mm_add_ps(fjy0,ty);
1701             fjz0             = _mm_add_ps(fjz0,tz);
1702
1703             }
1704
1705             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
1706             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
1707             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
1708             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
1709
1710             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
1711
1712             /* Inner loop uses 246 flops */
1713         }
1714
1715         /* End of innermost loop */
1716
1717         gmx_mm_update_iforce_4atom_swizzle_ps(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
1718                                               f+i_coord_offset,fshift+i_shift_offset);
1719
1720         /* Increment number of inner iterations */
1721         inneriter                  += j_index_end - j_index_start;
1722
1723         /* Outer loop uses 24 flops */
1724     }
1725
1726     /* Increment number of outer iterations */
1727     outeriter        += nri;
1728
1729     /* Update outer/inner flops */
1730
1731     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4_F,outeriter*24 + inneriter*246);
1732 }