src/gromacs/gmxlib/nonbonded/nb_kernel_avx_128_fma_single/nb_kernel_ElecCSTab_VdwCSTab_GeomW3W3_avx_128_fma_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 avx_128_fma_single kernel generator.
  37  */
  38 #include "config.h"
  39
  40 #include <math.h>
  41
  42 #include "../nb_kernel.h"
  43 #include "gromacs/legacyheaders/types/simple.h"
  44 #include "gromacs/math/vec.h"
  45 #include "gromacs/legacyheaders/nrnb.h"
  46
  47 #include "gromacs/simd/math_x86_avx_128_fma_single.h"
  48 #include "kernelutil_x86_avx_128_fma_single.h"
  49
  50 /*
  51  * Gromacs nonbonded kernel:   nb_kernel_ElecCSTab_VdwCSTab_GeomW3W3_VF_avx_128_fma_single
  52  * Electrostatics interaction: CubicSplineTable
  53  * VdW interaction:            CubicSplineTable
  54  * Geometry:                   Water3-Water3
  55  * Calculate force/pot:        PotentialAndForce
  56  */
  57 void
  58 nb_kernel_ElecCSTab_VdwCSTab_GeomW3W3_VF_avx_128_fma_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 AVX_128, 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           fscal,rcutoff,rcutoff2,jidxall;
  83     int              vdwioffset0;
  84     __m128           ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
  85     int              vdwioffset1;
  86     __m128           ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
  87     int              vdwioffset2;
  88     __m128           ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
  89     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
  90     __m128           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  91     int              vdwjidx1A,vdwjidx1B,vdwjidx1C,vdwjidx1D;
  92     __m128           jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
  93     int              vdwjidx2A,vdwjidx2B,vdwjidx2C,vdwjidx2D;
  94     __m128           jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
  95     __m128           dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
  96     __m128           dx01,dy01,dz01,rsq01,rinv01,rinvsq01,r01,qq01,c6_01,c12_01;
  97     __m128           dx02,dy02,dz02,rsq02,rinv02,rinvsq02,r02,qq02,c6_02,c12_02;
  98     __m128           dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
  99     __m128           dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11;
 100     __m128           dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12;
 101     __m128           dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
 102     __m128           dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21;
 103     __m128           dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22;
 104     __m128           velec,felec,velecsum,facel,crf,krf,krf2;
 105     real             *charge;
 106     int              nvdwtype;
 107     __m128           rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 108     int              *vdwtype;
 109     real             *vdwparam;
 110     __m128           one_sixth   = _mm_set1_ps(1.0/6.0);
 111     __m128           one_twelfth = _mm_set1_ps(1.0/12.0);
 112     __m128i          vfitab;
 113     __m128i          ifour       = _mm_set1_epi32(4);
 114     __m128           rt,vfeps,twovfeps,vftabscale,Y,F,G,H,Fp,VV,FF;
 115     real             *vftab;
 116     __m128           dummy_mask,cutoff_mask;
 117     __m128           signbit = _mm_castsi128_ps( _mm_set1_epi32(0x80000000) );
 118     __m128           one     = _mm_set1_ps(1.0);
 119     __m128           two     = _mm_set1_ps(2.0);
 120     x                = xx[0];
 121     f                = ff[0];
 122
 123     nri              = nlist->nri;
 124     iinr             = nlist->iinr;
 125     jindex           = nlist->jindex;
 126     jjnr             = nlist->jjnr;
 127     shiftidx         = nlist->shift;
 128     gid              = nlist->gid;
 129     shiftvec         = fr->shift_vec[0];
 130     fshift           = fr->fshift[0];
 131     facel            = _mm_set1_ps(fr->epsfac);
 132     charge           = mdatoms->chargeA;
 133     nvdwtype         = fr->ntype;
 134     vdwparam         = fr->nbfp;
 135     vdwtype          = mdatoms->typeA;
 136
 137     vftab            = kernel_data->table_elec_vdw->data;
 138     vftabscale       = _mm_set1_ps(kernel_data->table_elec_vdw->scale);
 139
 140     /* Setup water-specific parameters */
 141     inr              = nlist->iinr[0];
 142     iq0              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+0]));
 143     iq1              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+1]));
 144     iq2              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+2]));
 145     vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 146
 147     jq0              = _mm_set1_ps(charge[inr+0]);
 148     jq1              = _mm_set1_ps(charge[inr+1]);
 149     jq2              = _mm_set1_ps(charge[inr+2]);
 150     vdwjidx0A        = 2*vdwtype[inr+0];
 151     qq00             = _mm_mul_ps(iq0,jq0);
 152     c6_00            = _mm_set1_ps(vdwparam[vdwioffset0+vdwjidx0A]);
 153     c12_00           = _mm_set1_ps(vdwparam[vdwioffset0+vdwjidx0A+1]);
 154     qq01             = _mm_mul_ps(iq0,jq1);
 155     qq02             = _mm_mul_ps(iq0,jq2);
 156     qq10             = _mm_mul_ps(iq1,jq0);
 157     qq11             = _mm_mul_ps(iq1,jq1);
 158     qq12             = _mm_mul_ps(iq1,jq2);
 159     qq20             = _mm_mul_ps(iq2,jq0);
 160     qq21             = _mm_mul_ps(iq2,jq1);
 161     qq22             = _mm_mul_ps(iq2,jq2);
 162
 163     /* Avoid stupid compiler warnings */
 164     jnrA = jnrB = jnrC = jnrD = 0;
 165     j_coord_offsetA = 0;
 166     j_coord_offsetB = 0;
 167     j_coord_offsetC = 0;
 168     j_coord_offsetD = 0;
 169
 170     outeriter        = 0;
 171     inneriter        = 0;
 172
 173     for(iidx=0;iidx<4*DIM;iidx++)
 174     {
 175         scratch[iidx] = 0.0;
 176     }
 177
 178     /* Start outer loop over neighborlists */
 179     for(iidx=0; iidx<nri; iidx++)
 180     {
 181         /* Load shift vector for this list */
 182         i_shift_offset   = DIM*shiftidx[iidx];
 183
 184         /* Load limits for loop over neighbors */
 185         j_index_start    = jindex[iidx];
 186         j_index_end      = jindex[iidx+1];
 187
 188         /* Get outer coordinate index */
 189         inr              = iinr[iidx];
 190         i_coord_offset   = DIM*inr;
 191
 192         /* Load i particle coords and add shift vector */
 193         gmx_mm_load_shift_and_3rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,
 194                                                  &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2);
 195
 196         fix0             = _mm_setzero_ps();
 197         fiy0             = _mm_setzero_ps();
 198         fiz0             = _mm_setzero_ps();
 199         fix1             = _mm_setzero_ps();
 200         fiy1             = _mm_setzero_ps();
 201         fiz1             = _mm_setzero_ps();
 202         fix2             = _mm_setzero_ps();
 203         fiy2             = _mm_setzero_ps();
 204         fiz2             = _mm_setzero_ps();
 205
 206         /* Reset potential sums */
 207         velecsum         = _mm_setzero_ps();
 208         vvdwsum          = _mm_setzero_ps();
 209
 210         /* Start inner kernel loop */
 211         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 212         {
 213
 214             /* Get j neighbor index, and coordinate index */
 215             jnrA             = jjnr[jidx];
 216             jnrB             = jjnr[jidx+1];
 217             jnrC             = jjnr[jidx+2];
 218             jnrD             = jjnr[jidx+3];
 219             j_coord_offsetA  = DIM*jnrA;
 220             j_coord_offsetB  = DIM*jnrB;
 221             j_coord_offsetC  = DIM*jnrC;
 222             j_coord_offsetD  = DIM*jnrD;
 223
 224             /* load j atom coordinates */
 225             gmx_mm_load_3rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 226                                               x+j_coord_offsetC,x+j_coord_offsetD,
 227                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,&jy2,&jz2);
 228
 229             /* Calculate displacement vector */
 230             dx00             = _mm_sub_ps(ix0,jx0);
 231             dy00             = _mm_sub_ps(iy0,jy0);
 232             dz00             = _mm_sub_ps(iz0,jz0);
 233             dx01             = _mm_sub_ps(ix0,jx1);
 234             dy01             = _mm_sub_ps(iy0,jy1);
 235             dz01             = _mm_sub_ps(iz0,jz1);
 236             dx02             = _mm_sub_ps(ix0,jx2);
 237             dy02             = _mm_sub_ps(iy0,jy2);
 238             dz02             = _mm_sub_ps(iz0,jz2);
 239             dx10             = _mm_sub_ps(ix1,jx0);
 240             dy10             = _mm_sub_ps(iy1,jy0);
 241             dz10             = _mm_sub_ps(iz1,jz0);
 242             dx11             = _mm_sub_ps(ix1,jx1);
 243             dy11             = _mm_sub_ps(iy1,jy1);
 244             dz11             = _mm_sub_ps(iz1,jz1);
 245             dx12             = _mm_sub_ps(ix1,jx2);
 246             dy12             = _mm_sub_ps(iy1,jy2);
 247             dz12             = _mm_sub_ps(iz1,jz2);
 248             dx20             = _mm_sub_ps(ix2,jx0);
 249             dy20             = _mm_sub_ps(iy2,jy0);
 250             dz20             = _mm_sub_ps(iz2,jz0);
 251             dx21             = _mm_sub_ps(ix2,jx1);
 252             dy21             = _mm_sub_ps(iy2,jy1);
 253             dz21             = _mm_sub_ps(iz2,jz1);
 254             dx22             = _mm_sub_ps(ix2,jx2);
 255             dy22             = _mm_sub_ps(iy2,jy2);
 256             dz22             = _mm_sub_ps(iz2,jz2);
 257
 258             /* Calculate squared distance and things based on it */
 259             rsq00            = gmx_mm_calc_rsq_ps(dx00,dy00,dz00);
 260             rsq01            = gmx_mm_calc_rsq_ps(dx01,dy01,dz01);
 261             rsq02            = gmx_mm_calc_rsq_ps(dx02,dy02,dz02);
 262             rsq10            = gmx_mm_calc_rsq_ps(dx10,dy10,dz10);
 263             rsq11            = gmx_mm_calc_rsq_ps(dx11,dy11,dz11);
 264             rsq12            = gmx_mm_calc_rsq_ps(dx12,dy12,dz12);
 265             rsq20            = gmx_mm_calc_rsq_ps(dx20,dy20,dz20);
 266             rsq21            = gmx_mm_calc_rsq_ps(dx21,dy21,dz21);
 267             rsq22            = gmx_mm_calc_rsq_ps(dx22,dy22,dz22);
 268
 269             rinv00           = gmx_mm_invsqrt_ps(rsq00);
 270             rinv01           = gmx_mm_invsqrt_ps(rsq01);
 271             rinv02           = gmx_mm_invsqrt_ps(rsq02);
 272             rinv10           = gmx_mm_invsqrt_ps(rsq10);
 273             rinv11           = gmx_mm_invsqrt_ps(rsq11);
 274             rinv12           = gmx_mm_invsqrt_ps(rsq12);
 275             rinv20           = gmx_mm_invsqrt_ps(rsq20);
 276             rinv21           = gmx_mm_invsqrt_ps(rsq21);
 277             rinv22           = gmx_mm_invsqrt_ps(rsq22);
 278
 279             fjx0             = _mm_setzero_ps();
 280             fjy0             = _mm_setzero_ps();
 281             fjz0             = _mm_setzero_ps();
 282             fjx1             = _mm_setzero_ps();
 283             fjy1             = _mm_setzero_ps();
 284             fjz1             = _mm_setzero_ps();
 285             fjx2             = _mm_setzero_ps();
 286             fjy2             = _mm_setzero_ps();
 287             fjz2             = _mm_setzero_ps();
 288
 289             /**************************
 290              * CALCULATE INTERACTIONS *
 291              **************************/
 292
 293             r00              = _mm_mul_ps(rsq00,rinv00);
 294
 295             /* Calculate table index by multiplying r with table scale and truncate to integer */
 296             rt               = _mm_mul_ps(r00,vftabscale);
 297             vfitab           = _mm_cvttps_epi32(rt);
 298 #ifdef __XOP__
 299             vfeps            = _mm_frcz_ps(rt);
 300 #else
 301             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
 302 #endif
 303             twovfeps         = _mm_add_ps(vfeps,vfeps);
 304             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
 305
 306             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 307             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 308             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 309             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 310             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 311             _MM_TRANSPOSE4_PS(Y,F,G,H);
 312             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 313             VV               = _mm_macc_ps(vfeps,Fp,Y);
 314             velec            = _mm_mul_ps(qq00,VV);
 315             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 316             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq00,FF),_mm_mul_ps(vftabscale,rinv00)));
 317
 318             /* CUBIC SPLINE TABLE DISPERSION */
 319             vfitab           = _mm_add_epi32(vfitab,ifour);
 320             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 321             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 322             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 323             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 324             _MM_TRANSPOSE4_PS(Y,F,G,H);
 325             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 326             VV               = _mm_macc_ps(vfeps,Fp,Y);
 327             vvdw6            = _mm_mul_ps(c6_00,VV);
 328             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 329             fvdw6            = _mm_mul_ps(c6_00,FF);
 330
 331             /* CUBIC SPLINE TABLE REPULSION */
 332             vfitab           = _mm_add_epi32(vfitab,ifour);
 333             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 334             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 335             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 336             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 337             _MM_TRANSPOSE4_PS(Y,F,G,H);
 338             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 339             VV               = _mm_macc_ps(vfeps,Fp,Y);
 340             vvdw12           = _mm_mul_ps(c12_00,VV);
 341             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 342             fvdw12           = _mm_mul_ps(c12_00,FF);
 343             vvdw             = _mm_add_ps(vvdw12,vvdw6);
 344             fvdw             = _mm_xor_ps(signbit,_mm_mul_ps(_mm_add_ps(fvdw6,fvdw12),_mm_mul_ps(vftabscale,rinv00)));
 345
 346             /* Update potential sum for this i atom from the interaction with this j atom. */
 347             velecsum         = _mm_add_ps(velecsum,velec);
 348             vvdwsum          = _mm_add_ps(vvdwsum,vvdw);
 349
 350             fscal            = _mm_add_ps(felec,fvdw);
 351
 352              /* Update vectorial force */
 353             fix0             = _mm_macc_ps(dx00,fscal,fix0);
 354             fiy0             = _mm_macc_ps(dy00,fscal,fiy0);
 355             fiz0             = _mm_macc_ps(dz00,fscal,fiz0);
 356
 357             fjx0             = _mm_macc_ps(dx00,fscal,fjx0);
 358             fjy0             = _mm_macc_ps(dy00,fscal,fjy0);
 359             fjz0             = _mm_macc_ps(dz00,fscal,fjz0);
 360
 361             /**************************
 362              * CALCULATE INTERACTIONS *
 363              **************************/
 364
 365             r01              = _mm_mul_ps(rsq01,rinv01);
 366
 367             /* Calculate table index by multiplying r with table scale and truncate to integer */
 368             rt               = _mm_mul_ps(r01,vftabscale);
 369             vfitab           = _mm_cvttps_epi32(rt);
 370 #ifdef __XOP__
 371             vfeps            = _mm_frcz_ps(rt);
 372 #else
 373             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
 374 #endif
 375             twovfeps         = _mm_add_ps(vfeps,vfeps);
 376             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
 377
 378             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 379             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 380             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 381             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 382             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 383             _MM_TRANSPOSE4_PS(Y,F,G,H);
 384             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 385             VV               = _mm_macc_ps(vfeps,Fp,Y);
 386             velec            = _mm_mul_ps(qq01,VV);
 387             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 388             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq01,FF),_mm_mul_ps(vftabscale,rinv01)));
 389
 390             /* Update potential sum for this i atom from the interaction with this j atom. */
 391             velecsum         = _mm_add_ps(velecsum,velec);
 392
 393             fscal            = felec;
 394
 395              /* Update vectorial force */
 396             fix0             = _mm_macc_ps(dx01,fscal,fix0);
 397             fiy0             = _mm_macc_ps(dy01,fscal,fiy0);
 398             fiz0             = _mm_macc_ps(dz01,fscal,fiz0);
 399
 400             fjx1             = _mm_macc_ps(dx01,fscal,fjx1);
 401             fjy1             = _mm_macc_ps(dy01,fscal,fjy1);
 402             fjz1             = _mm_macc_ps(dz01,fscal,fjz1);
 403
 404             /**************************
 405              * CALCULATE INTERACTIONS *
 406              **************************/
 407
 408             r02              = _mm_mul_ps(rsq02,rinv02);
 409
 410             /* Calculate table index by multiplying r with table scale and truncate to integer */
 411             rt               = _mm_mul_ps(r02,vftabscale);
 412             vfitab           = _mm_cvttps_epi32(rt);
 413 #ifdef __XOP__
 414             vfeps            = _mm_frcz_ps(rt);
 415 #else
 416             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
 417 #endif
 418             twovfeps         = _mm_add_ps(vfeps,vfeps);
 419             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
 420
 421             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 422             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 423             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 424             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 425             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 426             _MM_TRANSPOSE4_PS(Y,F,G,H);
 427             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 428             VV               = _mm_macc_ps(vfeps,Fp,Y);
 429             velec            = _mm_mul_ps(qq02,VV);
 430             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 431             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq02,FF),_mm_mul_ps(vftabscale,rinv02)));
 432
 433             /* Update potential sum for this i atom from the interaction with this j atom. */
 434             velecsum         = _mm_add_ps(velecsum,velec);
 435
 436             fscal            = felec;
 437
 438              /* Update vectorial force */
 439             fix0             = _mm_macc_ps(dx02,fscal,fix0);
 440             fiy0             = _mm_macc_ps(dy02,fscal,fiy0);
 441             fiz0             = _mm_macc_ps(dz02,fscal,fiz0);
 442
 443             fjx2             = _mm_macc_ps(dx02,fscal,fjx2);
 444             fjy2             = _mm_macc_ps(dy02,fscal,fjy2);
 445             fjz2             = _mm_macc_ps(dz02,fscal,fjz2);
 446
 447             /**************************
 448              * CALCULATE INTERACTIONS *
 449              **************************/
 450
 451             r10              = _mm_mul_ps(rsq10,rinv10);
 452
 453             /* Calculate table index by multiplying r with table scale and truncate to integer */
 454             rt               = _mm_mul_ps(r10,vftabscale);
 455             vfitab           = _mm_cvttps_epi32(rt);
 456 #ifdef __XOP__
 457             vfeps            = _mm_frcz_ps(rt);
 458 #else
 459             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
 460 #endif
 461             twovfeps         = _mm_add_ps(vfeps,vfeps);
 462             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
 463
 464             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 465             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 466             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 467             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 468             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 469             _MM_TRANSPOSE4_PS(Y,F,G,H);
 470             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 471             VV               = _mm_macc_ps(vfeps,Fp,Y);
 472             velec            = _mm_mul_ps(qq10,VV);
 473             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 474             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq10,FF),_mm_mul_ps(vftabscale,rinv10)));
 475
 476             /* Update potential sum for this i atom from the interaction with this j atom. */
 477             velecsum         = _mm_add_ps(velecsum,velec);
 478
 479             fscal            = felec;
 480
 481              /* Update vectorial force */
 482             fix1             = _mm_macc_ps(dx10,fscal,fix1);
 483             fiy1             = _mm_macc_ps(dy10,fscal,fiy1);
 484             fiz1             = _mm_macc_ps(dz10,fscal,fiz1);
 485
 486             fjx0             = _mm_macc_ps(dx10,fscal,fjx0);
 487             fjy0             = _mm_macc_ps(dy10,fscal,fjy0);
 488             fjz0             = _mm_macc_ps(dz10,fscal,fjz0);
 489
 490             /**************************
 491              * CALCULATE INTERACTIONS *
 492              **************************/
 493
 494             r11              = _mm_mul_ps(rsq11,rinv11);
 495
 496             /* Calculate table index by multiplying r with table scale and truncate to integer */
 497             rt               = _mm_mul_ps(r11,vftabscale);
 498             vfitab           = _mm_cvttps_epi32(rt);
 499 #ifdef __XOP__
 500             vfeps            = _mm_frcz_ps(rt);
 501 #else
 502             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
 503 #endif
 504             twovfeps         = _mm_add_ps(vfeps,vfeps);
 505             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
 506
 507             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 508             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 509             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 510             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 511             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 512             _MM_TRANSPOSE4_PS(Y,F,G,H);
 513             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 514             VV               = _mm_macc_ps(vfeps,Fp,Y);
 515             velec            = _mm_mul_ps(qq11,VV);
 516             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 517             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq11,FF),_mm_mul_ps(vftabscale,rinv11)));
 518
 519             /* Update potential sum for this i atom from the interaction with this j atom. */
 520             velecsum         = _mm_add_ps(velecsum,velec);
 521
 522             fscal            = felec;
 523
 524              /* Update vectorial force */
 525             fix1             = _mm_macc_ps(dx11,fscal,fix1);
 526             fiy1             = _mm_macc_ps(dy11,fscal,fiy1);
 527             fiz1             = _mm_macc_ps(dz11,fscal,fiz1);
 528
 529             fjx1             = _mm_macc_ps(dx11,fscal,fjx1);
 530             fjy1             = _mm_macc_ps(dy11,fscal,fjy1);
 531             fjz1             = _mm_macc_ps(dz11,fscal,fjz1);
 532
 533             /**************************
 534              * CALCULATE INTERACTIONS *
 535              **************************/
 536
 537             r12              = _mm_mul_ps(rsq12,rinv12);
 538
 539             /* Calculate table index by multiplying r with table scale and truncate to integer */
 540             rt               = _mm_mul_ps(r12,vftabscale);
 541             vfitab           = _mm_cvttps_epi32(rt);
 542 #ifdef __XOP__
 543             vfeps            = _mm_frcz_ps(rt);
 544 #else
 545             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
 546 #endif
 547             twovfeps         = _mm_add_ps(vfeps,vfeps);
 548             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
 549
 550             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 551             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 552             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 553             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 554             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 555             _MM_TRANSPOSE4_PS(Y,F,G,H);
 556             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 557             VV               = _mm_macc_ps(vfeps,Fp,Y);
 558             velec            = _mm_mul_ps(qq12,VV);
 559             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 560             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq12,FF),_mm_mul_ps(vftabscale,rinv12)));
 561
 562             /* Update potential sum for this i atom from the interaction with this j atom. */
 563             velecsum         = _mm_add_ps(velecsum,velec);
 564
 565             fscal            = felec;
 566
 567              /* Update vectorial force */
 568             fix1             = _mm_macc_ps(dx12,fscal,fix1);
 569             fiy1             = _mm_macc_ps(dy12,fscal,fiy1);
 570             fiz1             = _mm_macc_ps(dz12,fscal,fiz1);
 571
 572             fjx2             = _mm_macc_ps(dx12,fscal,fjx2);
 573             fjy2             = _mm_macc_ps(dy12,fscal,fjy2);
 574             fjz2             = _mm_macc_ps(dz12,fscal,fjz2);
 575
 576             /**************************
 577              * CALCULATE INTERACTIONS *
 578              **************************/
 579
 580             r20              = _mm_mul_ps(rsq20,rinv20);
 581
 582             /* Calculate table index by multiplying r with table scale and truncate to integer */
 583             rt               = _mm_mul_ps(r20,vftabscale);
 584             vfitab           = _mm_cvttps_epi32(rt);
 585 #ifdef __XOP__
 586             vfeps            = _mm_frcz_ps(rt);
 587 #else
 588             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
 589 #endif
 590             twovfeps         = _mm_add_ps(vfeps,vfeps);
 591             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
 592
 593             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 594             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 595             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 596             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 597             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 598             _MM_TRANSPOSE4_PS(Y,F,G,H);
 599             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 600             VV               = _mm_macc_ps(vfeps,Fp,Y);
 601             velec            = _mm_mul_ps(qq20,VV);
 602             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 603             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq20,FF),_mm_mul_ps(vftabscale,rinv20)));
 604
 605             /* Update potential sum for this i atom from the interaction with this j atom. */
 606             velecsum         = _mm_add_ps(velecsum,velec);
 607
 608             fscal            = felec;
 609
 610              /* Update vectorial force */
 611             fix2             = _mm_macc_ps(dx20,fscal,fix2);
 612             fiy2             = _mm_macc_ps(dy20,fscal,fiy2);
 613             fiz2             = _mm_macc_ps(dz20,fscal,fiz2);
 614
 615             fjx0             = _mm_macc_ps(dx20,fscal,fjx0);
 616             fjy0             = _mm_macc_ps(dy20,fscal,fjy0);
 617             fjz0             = _mm_macc_ps(dz20,fscal,fjz0);
 618
 619             /**************************
 620              * CALCULATE INTERACTIONS *
 621              **************************/
 622
 623             r21              = _mm_mul_ps(rsq21,rinv21);
 624
 625             /* Calculate table index by multiplying r with table scale and truncate to integer */
 626             rt               = _mm_mul_ps(r21,vftabscale);
 627             vfitab           = _mm_cvttps_epi32(rt);
 628 #ifdef __XOP__
 629             vfeps            = _mm_frcz_ps(rt);
 630 #else
 631             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
 632 #endif
 633             twovfeps         = _mm_add_ps(vfeps,vfeps);
 634             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
 635
 636             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 637             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 638             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 639             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 640             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 641             _MM_TRANSPOSE4_PS(Y,F,G,H);
 642             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 643             VV               = _mm_macc_ps(vfeps,Fp,Y);
 644             velec            = _mm_mul_ps(qq21,VV);
 645             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 646             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq21,FF),_mm_mul_ps(vftabscale,rinv21)));
 647
 648             /* Update potential sum for this i atom from the interaction with this j atom. */
 649             velecsum         = _mm_add_ps(velecsum,velec);
 650
 651             fscal            = felec;
 652
 653              /* Update vectorial force */
 654             fix2             = _mm_macc_ps(dx21,fscal,fix2);
 655             fiy2             = _mm_macc_ps(dy21,fscal,fiy2);
 656             fiz2             = _mm_macc_ps(dz21,fscal,fiz2);
 657
 658             fjx1             = _mm_macc_ps(dx21,fscal,fjx1);
 659             fjy1             = _mm_macc_ps(dy21,fscal,fjy1);
 660             fjz1             = _mm_macc_ps(dz21,fscal,fjz1);
 661
 662             /**************************
 663              * CALCULATE INTERACTIONS *
 664              **************************/
 665
 666             r22              = _mm_mul_ps(rsq22,rinv22);
 667
 668             /* Calculate table index by multiplying r with table scale and truncate to integer */
 669             rt               = _mm_mul_ps(r22,vftabscale);
 670             vfitab           = _mm_cvttps_epi32(rt);
 671 #ifdef __XOP__
 672             vfeps            = _mm_frcz_ps(rt);
 673 #else
 674             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
 675 #endif
 676             twovfeps         = _mm_add_ps(vfeps,vfeps);
 677             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
 678
 679             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 680             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 681             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 682             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 683             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 684             _MM_TRANSPOSE4_PS(Y,F,G,H);
 685             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 686             VV               = _mm_macc_ps(vfeps,Fp,Y);
 687             velec            = _mm_mul_ps(qq22,VV);
 688             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 689             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq22,FF),_mm_mul_ps(vftabscale,rinv22)));
 690
 691             /* Update potential sum for this i atom from the interaction with this j atom. */
 692             velecsum         = _mm_add_ps(velecsum,velec);
 693
 694             fscal            = felec;
 695
 696              /* Update vectorial force */
 697             fix2             = _mm_macc_ps(dx22,fscal,fix2);
 698             fiy2             = _mm_macc_ps(dy22,fscal,fiy2);
 699             fiz2             = _mm_macc_ps(dz22,fscal,fiz2);
 700
 701             fjx2             = _mm_macc_ps(dx22,fscal,fjx2);
 702             fjy2             = _mm_macc_ps(dy22,fscal,fjy2);
 703             fjz2             = _mm_macc_ps(dz22,fscal,fjz2);
 704
 705             fjptrA             = f+j_coord_offsetA;
 706             fjptrB             = f+j_coord_offsetB;
 707             fjptrC             = f+j_coord_offsetC;
 708             fjptrD             = f+j_coord_offsetD;
 709
 710             gmx_mm_decrement_3rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,
 711                                                    fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2);
 712
 713             /* Inner loop uses 444 flops */
 714         }
 715
 716         if(jidx<j_index_end)
 717         {
 718
 719             /* Get j neighbor index, and coordinate index */
 720             jnrlistA         = jjnr[jidx];
 721             jnrlistB         = jjnr[jidx+1];
 722             jnrlistC         = jjnr[jidx+2];
 723             jnrlistD         = jjnr[jidx+3];
 724             /* Sign of each element will be negative for non-real atoms.
 725              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 726              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
 727              */
 728             dummy_mask = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
 729             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 730             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 731             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 732             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 733             j_coord_offsetA  = DIM*jnrA;
 734             j_coord_offsetB  = DIM*jnrB;
 735             j_coord_offsetC  = DIM*jnrC;
 736             j_coord_offsetD  = DIM*jnrD;
 737
 738             /* load j atom coordinates */
 739             gmx_mm_load_3rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 740                                               x+j_coord_offsetC,x+j_coord_offsetD,
 741                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,&jy2,&jz2);
 742
 743             /* Calculate displacement vector */
 744             dx00             = _mm_sub_ps(ix0,jx0);
 745             dy00             = _mm_sub_ps(iy0,jy0);
 746             dz00             = _mm_sub_ps(iz0,jz0);
 747             dx01             = _mm_sub_ps(ix0,jx1);
 748             dy01             = _mm_sub_ps(iy0,jy1);
 749             dz01             = _mm_sub_ps(iz0,jz1);
 750             dx02             = _mm_sub_ps(ix0,jx2);
 751             dy02             = _mm_sub_ps(iy0,jy2);
 752             dz02             = _mm_sub_ps(iz0,jz2);
 753             dx10             = _mm_sub_ps(ix1,jx0);
 754             dy10             = _mm_sub_ps(iy1,jy0);
 755             dz10             = _mm_sub_ps(iz1,jz0);
 756             dx11             = _mm_sub_ps(ix1,jx1);
 757             dy11             = _mm_sub_ps(iy1,jy1);
 758             dz11             = _mm_sub_ps(iz1,jz1);
 759             dx12             = _mm_sub_ps(ix1,jx2);
 760             dy12             = _mm_sub_ps(iy1,jy2);
 761             dz12             = _mm_sub_ps(iz1,jz2);
 762             dx20             = _mm_sub_ps(ix2,jx0);
 763             dy20             = _mm_sub_ps(iy2,jy0);
 764             dz20             = _mm_sub_ps(iz2,jz0);
 765             dx21             = _mm_sub_ps(ix2,jx1);
 766             dy21             = _mm_sub_ps(iy2,jy1);
 767             dz21             = _mm_sub_ps(iz2,jz1);
 768             dx22             = _mm_sub_ps(ix2,jx2);
 769             dy22             = _mm_sub_ps(iy2,jy2);
 770             dz22             = _mm_sub_ps(iz2,jz2);
 771
 772             /* Calculate squared distance and things based on it */
 773             rsq00            = gmx_mm_calc_rsq_ps(dx00,dy00,dz00);
 774             rsq01            = gmx_mm_calc_rsq_ps(dx01,dy01,dz01);
 775             rsq02            = gmx_mm_calc_rsq_ps(dx02,dy02,dz02);
 776             rsq10            = gmx_mm_calc_rsq_ps(dx10,dy10,dz10);
 777             rsq11            = gmx_mm_calc_rsq_ps(dx11,dy11,dz11);
 778             rsq12            = gmx_mm_calc_rsq_ps(dx12,dy12,dz12);
 779             rsq20            = gmx_mm_calc_rsq_ps(dx20,dy20,dz20);
 780             rsq21            = gmx_mm_calc_rsq_ps(dx21,dy21,dz21);
 781             rsq22            = gmx_mm_calc_rsq_ps(dx22,dy22,dz22);
 782
 783             rinv00           = gmx_mm_invsqrt_ps(rsq00);
 784             rinv01           = gmx_mm_invsqrt_ps(rsq01);
 785             rinv02           = gmx_mm_invsqrt_ps(rsq02);
 786             rinv10           = gmx_mm_invsqrt_ps(rsq10);
 787             rinv11           = gmx_mm_invsqrt_ps(rsq11);
 788             rinv12           = gmx_mm_invsqrt_ps(rsq12);
 789             rinv20           = gmx_mm_invsqrt_ps(rsq20);
 790             rinv21           = gmx_mm_invsqrt_ps(rsq21);
 791             rinv22           = gmx_mm_invsqrt_ps(rsq22);
 792
 793             fjx0             = _mm_setzero_ps();
 794             fjy0             = _mm_setzero_ps();
 795             fjz0             = _mm_setzero_ps();
 796             fjx1             = _mm_setzero_ps();
 797             fjy1             = _mm_setzero_ps();
 798             fjz1             = _mm_setzero_ps();
 799             fjx2             = _mm_setzero_ps();
 800             fjy2             = _mm_setzero_ps();
 801             fjz2             = _mm_setzero_ps();
 802
 803             /**************************
 804              * CALCULATE INTERACTIONS *
 805              **************************/
 806
 807             r00              = _mm_mul_ps(rsq00,rinv00);
 808             r00              = _mm_andnot_ps(dummy_mask,r00);
 809
 810             /* Calculate table index by multiplying r with table scale and truncate to integer */
 811             rt               = _mm_mul_ps(r00,vftabscale);
 812             vfitab           = _mm_cvttps_epi32(rt);
 813 #ifdef __XOP__
 814             vfeps            = _mm_frcz_ps(rt);
 815 #else
 816             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
 817 #endif
 818             twovfeps         = _mm_add_ps(vfeps,vfeps);
 819             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
 820
 821             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 822             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 823             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 824             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 825             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 826             _MM_TRANSPOSE4_PS(Y,F,G,H);
 827             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 828             VV               = _mm_macc_ps(vfeps,Fp,Y);
 829             velec            = _mm_mul_ps(qq00,VV);
 830             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 831             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq00,FF),_mm_mul_ps(vftabscale,rinv00)));
 832
 833             /* CUBIC SPLINE TABLE DISPERSION */
 834             vfitab           = _mm_add_epi32(vfitab,ifour);
 835             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 836             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 837             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 838             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 839             _MM_TRANSPOSE4_PS(Y,F,G,H);
 840             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 841             VV               = _mm_macc_ps(vfeps,Fp,Y);
 842             vvdw6            = _mm_mul_ps(c6_00,VV);
 843             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 844             fvdw6            = _mm_mul_ps(c6_00,FF);
 845
 846             /* CUBIC SPLINE TABLE REPULSION */
 847             vfitab           = _mm_add_epi32(vfitab,ifour);
 848             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 849             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 850             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 851             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 852             _MM_TRANSPOSE4_PS(Y,F,G,H);
 853             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 854             VV               = _mm_macc_ps(vfeps,Fp,Y);
 855             vvdw12           = _mm_mul_ps(c12_00,VV);
 856             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 857             fvdw12           = _mm_mul_ps(c12_00,FF);
 858             vvdw             = _mm_add_ps(vvdw12,vvdw6);
 859             fvdw             = _mm_xor_ps(signbit,_mm_mul_ps(_mm_add_ps(fvdw6,fvdw12),_mm_mul_ps(vftabscale,rinv00)));
 860
 861             /* Update potential sum for this i atom from the interaction with this j atom. */
 862             velec            = _mm_andnot_ps(dummy_mask,velec);
 863             velecsum         = _mm_add_ps(velecsum,velec);
 864             vvdw             = _mm_andnot_ps(dummy_mask,vvdw);
 865             vvdwsum          = _mm_add_ps(vvdwsum,vvdw);
 866
 867             fscal            = _mm_add_ps(felec,fvdw);
 868
 869             fscal            = _mm_andnot_ps(dummy_mask,fscal);
 870
 871              /* Update vectorial force */
 872             fix0             = _mm_macc_ps(dx00,fscal,fix0);
 873             fiy0             = _mm_macc_ps(dy00,fscal,fiy0);
 874             fiz0             = _mm_macc_ps(dz00,fscal,fiz0);
 875
 876             fjx0             = _mm_macc_ps(dx00,fscal,fjx0);
 877             fjy0             = _mm_macc_ps(dy00,fscal,fjy0);
 878             fjz0             = _mm_macc_ps(dz00,fscal,fjz0);
 879
 880             /**************************
 881              * CALCULATE INTERACTIONS *
 882              **************************/
 883
 884             r01              = _mm_mul_ps(rsq01,rinv01);
 885             r01              = _mm_andnot_ps(dummy_mask,r01);
 886
 887             /* Calculate table index by multiplying r with table scale and truncate to integer */
 888             rt               = _mm_mul_ps(r01,vftabscale);
 889             vfitab           = _mm_cvttps_epi32(rt);
 890 #ifdef __XOP__
 891             vfeps            = _mm_frcz_ps(rt);
 892 #else
 893             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
 894 #endif
 895             twovfeps         = _mm_add_ps(vfeps,vfeps);
 896             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
 897
 898             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 899             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 900             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 901             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 902             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 903             _MM_TRANSPOSE4_PS(Y,F,G,H);
 904             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 905             VV               = _mm_macc_ps(vfeps,Fp,Y);
 906             velec            = _mm_mul_ps(qq01,VV);
 907             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 908             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq01,FF),_mm_mul_ps(vftabscale,rinv01)));
 909
 910             /* Update potential sum for this i atom from the interaction with this j atom. */
 911             velec            = _mm_andnot_ps(dummy_mask,velec);
 912             velecsum         = _mm_add_ps(velecsum,velec);
 913
 914             fscal            = felec;
 915
 916             fscal            = _mm_andnot_ps(dummy_mask,fscal);
 917
 918              /* Update vectorial force */
 919             fix0             = _mm_macc_ps(dx01,fscal,fix0);
 920             fiy0             = _mm_macc_ps(dy01,fscal,fiy0);
 921             fiz0             = _mm_macc_ps(dz01,fscal,fiz0);
 922
 923             fjx1             = _mm_macc_ps(dx01,fscal,fjx1);
 924             fjy1             = _mm_macc_ps(dy01,fscal,fjy1);
 925             fjz1             = _mm_macc_ps(dz01,fscal,fjz1);
 926
 927             /**************************
 928              * CALCULATE INTERACTIONS *
 929              **************************/
 930
 931             r02              = _mm_mul_ps(rsq02,rinv02);
 932             r02              = _mm_andnot_ps(dummy_mask,r02);
 933
 934             /* Calculate table index by multiplying r with table scale and truncate to integer */
 935             rt               = _mm_mul_ps(r02,vftabscale);
 936             vfitab           = _mm_cvttps_epi32(rt);
 937 #ifdef __XOP__
 938             vfeps            = _mm_frcz_ps(rt);
 939 #else
 940             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
 941 #endif
 942             twovfeps         = _mm_add_ps(vfeps,vfeps);
 943             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
 944
 945             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 946             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 947             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 948             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 949             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 950             _MM_TRANSPOSE4_PS(Y,F,G,H);
 951             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 952             VV               = _mm_macc_ps(vfeps,Fp,Y);
 953             velec            = _mm_mul_ps(qq02,VV);
 954             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 955             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq02,FF),_mm_mul_ps(vftabscale,rinv02)));
 956
 957             /* Update potential sum for this i atom from the interaction with this j atom. */
 958             velec            = _mm_andnot_ps(dummy_mask,velec);
 959             velecsum         = _mm_add_ps(velecsum,velec);
 960
 961             fscal            = felec;
 962
 963             fscal            = _mm_andnot_ps(dummy_mask,fscal);
 964
 965              /* Update vectorial force */
 966             fix0             = _mm_macc_ps(dx02,fscal,fix0);
 967             fiy0             = _mm_macc_ps(dy02,fscal,fiy0);
 968             fiz0             = _mm_macc_ps(dz02,fscal,fiz0);
 969
 970             fjx2             = _mm_macc_ps(dx02,fscal,fjx2);
 971             fjy2             = _mm_macc_ps(dy02,fscal,fjy2);
 972             fjz2             = _mm_macc_ps(dz02,fscal,fjz2);
 973
 974             /**************************
 975              * CALCULATE INTERACTIONS *
 976              **************************/
 977
 978             r10              = _mm_mul_ps(rsq10,rinv10);
 979             r10              = _mm_andnot_ps(dummy_mask,r10);
 980
 981             /* Calculate table index by multiplying r with table scale and truncate to integer */
 982             rt               = _mm_mul_ps(r10,vftabscale);
 983             vfitab           = _mm_cvttps_epi32(rt);
 984 #ifdef __XOP__
 985             vfeps            = _mm_frcz_ps(rt);
 986 #else
 987             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
 988 #endif
 989             twovfeps         = _mm_add_ps(vfeps,vfeps);
 990             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
 991
 992             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 993             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 994             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 995             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 996             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 997             _MM_TRANSPOSE4_PS(Y,F,G,H);
 998             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 999             VV               = _mm_macc_ps(vfeps,Fp,Y);
1000             velec            = _mm_mul_ps(qq10,VV);
1001             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
1002             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq10,FF),_mm_mul_ps(vftabscale,rinv10)));
1003
1004             /* Update potential sum for this i atom from the interaction with this j atom. */
1005             velec            = _mm_andnot_ps(dummy_mask,velec);
1006             velecsum         = _mm_add_ps(velecsum,velec);
1007
1008             fscal            = felec;
1009
1010             fscal            = _mm_andnot_ps(dummy_mask,fscal);
1011
1012              /* Update vectorial force */
1013             fix1             = _mm_macc_ps(dx10,fscal,fix1);
1014             fiy1             = _mm_macc_ps(dy10,fscal,fiy1);
1015             fiz1             = _mm_macc_ps(dz10,fscal,fiz1);
1016
1017             fjx0             = _mm_macc_ps(dx10,fscal,fjx0);
1018             fjy0             = _mm_macc_ps(dy10,fscal,fjy0);
1019             fjz0             = _mm_macc_ps(dz10,fscal,fjz0);
1020
1021             /**************************
1022              * CALCULATE INTERACTIONS *
1023              **************************/
1024
1025             r11              = _mm_mul_ps(rsq11,rinv11);
1026             r11              = _mm_andnot_ps(dummy_mask,r11);
1027
1028             /* Calculate table index by multiplying r with table scale and truncate to integer */
1029             rt               = _mm_mul_ps(r11,vftabscale);
1030             vfitab           = _mm_cvttps_epi32(rt);
1031 #ifdef __XOP__
1032             vfeps            = _mm_frcz_ps(rt);
1033 #else
1034             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
1035 #endif
1036             twovfeps         = _mm_add_ps(vfeps,vfeps);
1037             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
1038
1039             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1040             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
1041             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
1042             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
1043             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
1044             _MM_TRANSPOSE4_PS(Y,F,G,H);
1045             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
1046             VV               = _mm_macc_ps(vfeps,Fp,Y);
1047             velec            = _mm_mul_ps(qq11,VV);
1048             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
1049             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq11,FF),_mm_mul_ps(vftabscale,rinv11)));
1050
1051             /* Update potential sum for this i atom from the interaction with this j atom. */
1052             velec            = _mm_andnot_ps(dummy_mask,velec);
1053             velecsum         = _mm_add_ps(velecsum,velec);
1054
1055             fscal            = felec;
1056
1057             fscal            = _mm_andnot_ps(dummy_mask,fscal);
1058
1059              /* Update vectorial force */
1060             fix1             = _mm_macc_ps(dx11,fscal,fix1);
1061             fiy1             = _mm_macc_ps(dy11,fscal,fiy1);
1062             fiz1             = _mm_macc_ps(dz11,fscal,fiz1);
1063
1064             fjx1             = _mm_macc_ps(dx11,fscal,fjx1);
1065             fjy1             = _mm_macc_ps(dy11,fscal,fjy1);
1066             fjz1             = _mm_macc_ps(dz11,fscal,fjz1);
1067
1068             /**************************
1069              * CALCULATE INTERACTIONS *
1070              **************************/
1071
1072             r12              = _mm_mul_ps(rsq12,rinv12);
1073             r12              = _mm_andnot_ps(dummy_mask,r12);
1074
1075             /* Calculate table index by multiplying r with table scale and truncate to integer */
1076             rt               = _mm_mul_ps(r12,vftabscale);
1077             vfitab           = _mm_cvttps_epi32(rt);
1078 #ifdef __XOP__
1079             vfeps            = _mm_frcz_ps(rt);
1080 #else
1081             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
1082 #endif
1083             twovfeps         = _mm_add_ps(vfeps,vfeps);
1084             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
1085
1086             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1087             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
1088             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
1089             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
1090             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
1091             _MM_TRANSPOSE4_PS(Y,F,G,H);
1092             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
1093             VV               = _mm_macc_ps(vfeps,Fp,Y);
1094             velec            = _mm_mul_ps(qq12,VV);
1095             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
1096             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq12,FF),_mm_mul_ps(vftabscale,rinv12)));
1097
1098             /* Update potential sum for this i atom from the interaction with this j atom. */
1099             velec            = _mm_andnot_ps(dummy_mask,velec);
1100             velecsum         = _mm_add_ps(velecsum,velec);
1101
1102             fscal            = felec;
1103
1104             fscal            = _mm_andnot_ps(dummy_mask,fscal);
1105
1106              /* Update vectorial force */
1107             fix1             = _mm_macc_ps(dx12,fscal,fix1);
1108             fiy1             = _mm_macc_ps(dy12,fscal,fiy1);
1109             fiz1             = _mm_macc_ps(dz12,fscal,fiz1);
1110
1111             fjx2             = _mm_macc_ps(dx12,fscal,fjx2);
1112             fjy2             = _mm_macc_ps(dy12,fscal,fjy2);
1113             fjz2             = _mm_macc_ps(dz12,fscal,fjz2);
1114
1115             /**************************
1116              * CALCULATE INTERACTIONS *
1117              **************************/
1118
1119             r20              = _mm_mul_ps(rsq20,rinv20);
1120             r20              = _mm_andnot_ps(dummy_mask,r20);
1121
1122             /* Calculate table index by multiplying r with table scale and truncate to integer */
1123             rt               = _mm_mul_ps(r20,vftabscale);
1124             vfitab           = _mm_cvttps_epi32(rt);
1125 #ifdef __XOP__
1126             vfeps            = _mm_frcz_ps(rt);
1127 #else
1128             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
1129 #endif
1130             twovfeps         = _mm_add_ps(vfeps,vfeps);
1131             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
1132
1133             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1134             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
1135             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
1136             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
1137             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
1138             _MM_TRANSPOSE4_PS(Y,F,G,H);
1139             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
1140             VV               = _mm_macc_ps(vfeps,Fp,Y);
1141             velec            = _mm_mul_ps(qq20,VV);
1142             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
1143             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq20,FF),_mm_mul_ps(vftabscale,rinv20)));
1144
1145             /* Update potential sum for this i atom from the interaction with this j atom. */
1146             velec            = _mm_andnot_ps(dummy_mask,velec);
1147             velecsum         = _mm_add_ps(velecsum,velec);
1148
1149             fscal            = felec;
1150
1151             fscal            = _mm_andnot_ps(dummy_mask,fscal);
1152
1153              /* Update vectorial force */
1154             fix2             = _mm_macc_ps(dx20,fscal,fix2);
1155             fiy2             = _mm_macc_ps(dy20,fscal,fiy2);
1156             fiz2             = _mm_macc_ps(dz20,fscal,fiz2);
1157
1158             fjx0             = _mm_macc_ps(dx20,fscal,fjx0);
1159             fjy0             = _mm_macc_ps(dy20,fscal,fjy0);
1160             fjz0             = _mm_macc_ps(dz20,fscal,fjz0);
1161
1162             /**************************
1163              * CALCULATE INTERACTIONS *
1164              **************************/
1165
1166             r21              = _mm_mul_ps(rsq21,rinv21);
1167             r21              = _mm_andnot_ps(dummy_mask,r21);
1168
1169             /* Calculate table index by multiplying r with table scale and truncate to integer */
1170             rt               = _mm_mul_ps(r21,vftabscale);
1171             vfitab           = _mm_cvttps_epi32(rt);
1172 #ifdef __XOP__
1173             vfeps            = _mm_frcz_ps(rt);
1174 #else
1175             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
1176 #endif
1177             twovfeps         = _mm_add_ps(vfeps,vfeps);
1178             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
1179
1180             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1181             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
1182             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
1183             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
1184             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
1185             _MM_TRANSPOSE4_PS(Y,F,G,H);
1186             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
1187             VV               = _mm_macc_ps(vfeps,Fp,Y);
1188             velec            = _mm_mul_ps(qq21,VV);
1189             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
1190             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq21,FF),_mm_mul_ps(vftabscale,rinv21)));
1191
1192             /* Update potential sum for this i atom from the interaction with this j atom. */
1193             velec            = _mm_andnot_ps(dummy_mask,velec);
1194             velecsum         = _mm_add_ps(velecsum,velec);
1195
1196             fscal            = felec;
1197
1198             fscal            = _mm_andnot_ps(dummy_mask,fscal);
1199
1200              /* Update vectorial force */
1201             fix2             = _mm_macc_ps(dx21,fscal,fix2);
1202             fiy2             = _mm_macc_ps(dy21,fscal,fiy2);
1203             fiz2             = _mm_macc_ps(dz21,fscal,fiz2);
1204
1205             fjx1             = _mm_macc_ps(dx21,fscal,fjx1);
1206             fjy1             = _mm_macc_ps(dy21,fscal,fjy1);
1207             fjz1             = _mm_macc_ps(dz21,fscal,fjz1);
1208
1209             /**************************
1210              * CALCULATE INTERACTIONS *
1211              **************************/
1212
1213             r22              = _mm_mul_ps(rsq22,rinv22);
1214             r22              = _mm_andnot_ps(dummy_mask,r22);
1215
1216             /* Calculate table index by multiplying r with table scale and truncate to integer */
1217             rt               = _mm_mul_ps(r22,vftabscale);
1218             vfitab           = _mm_cvttps_epi32(rt);
1219 #ifdef __XOP__
1220             vfeps            = _mm_frcz_ps(rt);
1221 #else
1222             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
1223 #endif
1224             twovfeps         = _mm_add_ps(vfeps,vfeps);
1225             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
1226
1227             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1228             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
1229             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
1230             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
1231             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
1232             _MM_TRANSPOSE4_PS(Y,F,G,H);
1233             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
1234             VV               = _mm_macc_ps(vfeps,Fp,Y);
1235             velec            = _mm_mul_ps(qq22,VV);
1236             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
1237             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq22,FF),_mm_mul_ps(vftabscale,rinv22)));
1238
1239             /* Update potential sum for this i atom from the interaction with this j atom. */
1240             velec            = _mm_andnot_ps(dummy_mask,velec);
1241             velecsum         = _mm_add_ps(velecsum,velec);
1242
1243             fscal            = felec;
1244
1245             fscal            = _mm_andnot_ps(dummy_mask,fscal);
1246
1247              /* Update vectorial force */
1248             fix2             = _mm_macc_ps(dx22,fscal,fix2);
1249             fiy2             = _mm_macc_ps(dy22,fscal,fiy2);
1250             fiz2             = _mm_macc_ps(dz22,fscal,fiz2);
1251
1252             fjx2             = _mm_macc_ps(dx22,fscal,fjx2);
1253             fjy2             = _mm_macc_ps(dy22,fscal,fjy2);
1254             fjz2             = _mm_macc_ps(dz22,fscal,fjz2);
1255
1256             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
1257             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
1258             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
1259             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
1260
1261             gmx_mm_decrement_3rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,
1262                                                    fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2);
1263
1264             /* Inner loop uses 453 flops */
1265         }
1266
1267         /* End of innermost loop */
1268
1269         gmx_mm_update_iforce_3atom_swizzle_ps(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,
1270                                               f+i_coord_offset,fshift+i_shift_offset);
1271
1272         ggid                        = gid[iidx];
1273         /* Update potential energies */
1274         gmx_mm_update_1pot_ps(velecsum,kernel_data->energygrp_elec+ggid);
1275         gmx_mm_update_1pot_ps(vvdwsum,kernel_data->energygrp_vdw+ggid);
1276
1277         /* Increment number of inner iterations */
1278         inneriter                  += j_index_end - j_index_start;
1279
1280         /* Outer loop uses 20 flops */
1281     }
1282
1283     /* Increment number of outer iterations */
1284     outeriter        += nri;
1285
1286     /* Update outer/inner flops */
1287
1288     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W3W3_VF,outeriter*20 + inneriter*453);
1289 }
1290 /*
1291  * Gromacs nonbonded kernel:   nb_kernel_ElecCSTab_VdwCSTab_GeomW3W3_F_avx_128_fma_single
1292  * Electrostatics interaction: CubicSplineTable
1293  * VdW interaction:            CubicSplineTable
1294  * Geometry:                   Water3-Water3
1295  * Calculate force/pot:        Force
1296  */
1297 void
1298 nb_kernel_ElecCSTab_VdwCSTab_GeomW3W3_F_avx_128_fma_single
1299                     (t_nblist                    * gmx_restrict       nlist,
1300                      rvec                        * gmx_restrict          xx,
1301                      rvec                        * gmx_restrict          ff,
1302                      t_forcerec                  * gmx_restrict          fr,
1303                      t_mdatoms                   * gmx_restrict     mdatoms,
1304                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
1305                      t_nrnb                      * gmx_restrict        nrnb)
1306 {
1307     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
1308      * just 0 for non-waters.
1309      * Suffixes A,B,C,D refer to j loop unrolling done with AVX_128, e.g. for the four different
1310      * jnr indices corresponding to data put in the four positions in the SIMD register.
1311      */
1312     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
1313     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
1314     int              jnrA,jnrB,jnrC,jnrD;
1315     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
1316     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
1317     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
1318     real             rcutoff_scalar;
1319     real             *shiftvec,*fshift,*x,*f;
1320     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
1321     real             scratch[4*DIM];
1322     __m128           fscal,rcutoff,rcutoff2,jidxall;
1323     int              vdwioffset0;
1324     __m128           ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
1325     int              vdwioffset1;
1326     __m128           ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
1327     int              vdwioffset2;
1328     __m128           ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
1329     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
1330     __m128           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
1331     int              vdwjidx1A,vdwjidx1B,vdwjidx1C,vdwjidx1D;
1332     __m128           jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
1333     int              vdwjidx2A,vdwjidx2B,vdwjidx2C,vdwjidx2D;
1334     __m128           jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
1335     __m128           dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
1336     __m128           dx01,dy01,dz01,rsq01,rinv01,rinvsq01,r01,qq01,c6_01,c12_01;
1337     __m128           dx02,dy02,dz02,rsq02,rinv02,rinvsq02,r02,qq02,c6_02,c12_02;
1338     __m128           dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
1339     __m128           dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11;
1340     __m128           dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12;
1341     __m128           dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
1342     __m128           dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21;
1343     __m128           dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22;
1344     __m128           velec,felec,velecsum,facel,crf,krf,krf2;
1345     real             *charge;
1346     int              nvdwtype;
1347     __m128           rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
1348     int              *vdwtype;
1349     real             *vdwparam;
1350     __m128           one_sixth   = _mm_set1_ps(1.0/6.0);
1351     __m128           one_twelfth = _mm_set1_ps(1.0/12.0);
1352     __m128i          vfitab;
1353     __m128i          ifour       = _mm_set1_epi32(4);
1354     __m128           rt,vfeps,twovfeps,vftabscale,Y,F,G,H,Fp,VV,FF;
1355     real             *vftab;
1356     __m128           dummy_mask,cutoff_mask;
1357     __m128           signbit = _mm_castsi128_ps( _mm_set1_epi32(0x80000000) );
1358     __m128           one     = _mm_set1_ps(1.0);
1359     __m128           two     = _mm_set1_ps(2.0);
1360     x                = xx[0];
1361     f                = ff[0];
1362
1363     nri              = nlist->nri;
1364     iinr             = nlist->iinr;
1365     jindex           = nlist->jindex;
1366     jjnr             = nlist->jjnr;
1367     shiftidx         = nlist->shift;
1368     gid              = nlist->gid;
1369     shiftvec         = fr->shift_vec[0];
1370     fshift           = fr->fshift[0];
1371     facel            = _mm_set1_ps(fr->epsfac);
1372     charge           = mdatoms->chargeA;
1373     nvdwtype         = fr->ntype;
1374     vdwparam         = fr->nbfp;
1375     vdwtype          = mdatoms->typeA;
1376
1377     vftab            = kernel_data->table_elec_vdw->data;
1378     vftabscale       = _mm_set1_ps(kernel_data->table_elec_vdw->scale);
1379
1380     /* Setup water-specific parameters */
1381     inr              = nlist->iinr[0];
1382     iq0              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+0]));
1383     iq1              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+1]));
1384     iq2              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+2]));
1385     vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
1386
1387     jq0              = _mm_set1_ps(charge[inr+0]);
1388     jq1              = _mm_set1_ps(charge[inr+1]);
1389     jq2              = _mm_set1_ps(charge[inr+2]);
1390     vdwjidx0A        = 2*vdwtype[inr+0];
1391     qq00             = _mm_mul_ps(iq0,jq0);
1392     c6_00            = _mm_set1_ps(vdwparam[vdwioffset0+vdwjidx0A]);
1393     c12_00           = _mm_set1_ps(vdwparam[vdwioffset0+vdwjidx0A+1]);
1394     qq01             = _mm_mul_ps(iq0,jq1);
1395     qq02             = _mm_mul_ps(iq0,jq2);
1396     qq10             = _mm_mul_ps(iq1,jq0);
1397     qq11             = _mm_mul_ps(iq1,jq1);
1398     qq12             = _mm_mul_ps(iq1,jq2);
1399     qq20             = _mm_mul_ps(iq2,jq0);
1400     qq21             = _mm_mul_ps(iq2,jq1);
1401     qq22             = _mm_mul_ps(iq2,jq2);
1402
1403     /* Avoid stupid compiler warnings */
1404     jnrA = jnrB = jnrC = jnrD = 0;
1405     j_coord_offsetA = 0;
1406     j_coord_offsetB = 0;
1407     j_coord_offsetC = 0;
1408     j_coord_offsetD = 0;
1409
1410     outeriter        = 0;
1411     inneriter        = 0;
1412
1413     for(iidx=0;iidx<4*DIM;iidx++)
1414     {
1415         scratch[iidx] = 0.0;
1416     }
1417
1418     /* Start outer loop over neighborlists */
1419     for(iidx=0; iidx<nri; iidx++)
1420     {
1421         /* Load shift vector for this list */
1422         i_shift_offset   = DIM*shiftidx[iidx];
1423
1424         /* Load limits for loop over neighbors */
1425         j_index_start    = jindex[iidx];
1426         j_index_end      = jindex[iidx+1];
1427
1428         /* Get outer coordinate index */
1429         inr              = iinr[iidx];
1430         i_coord_offset   = DIM*inr;
1431
1432         /* Load i particle coords and add shift vector */
1433         gmx_mm_load_shift_and_3rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,
1434                                                  &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2);
1435
1436         fix0             = _mm_setzero_ps();
1437         fiy0             = _mm_setzero_ps();
1438         fiz0             = _mm_setzero_ps();
1439         fix1             = _mm_setzero_ps();
1440         fiy1             = _mm_setzero_ps();
1441         fiz1             = _mm_setzero_ps();
1442         fix2             = _mm_setzero_ps();
1443         fiy2             = _mm_setzero_ps();
1444         fiz2             = _mm_setzero_ps();
1445
1446         /* Start inner kernel loop */
1447         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
1448         {
1449
1450             /* Get j neighbor index, and coordinate index */
1451             jnrA             = jjnr[jidx];
1452             jnrB             = jjnr[jidx+1];
1453             jnrC             = jjnr[jidx+2];
1454             jnrD             = jjnr[jidx+3];
1455             j_coord_offsetA  = DIM*jnrA;
1456             j_coord_offsetB  = DIM*jnrB;
1457             j_coord_offsetC  = DIM*jnrC;
1458             j_coord_offsetD  = DIM*jnrD;
1459
1460             /* load j atom coordinates */
1461             gmx_mm_load_3rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
1462                                               x+j_coord_offsetC,x+j_coord_offsetD,
1463                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,&jy2,&jz2);
1464
1465             /* Calculate displacement vector */
1466             dx00             = _mm_sub_ps(ix0,jx0);
1467             dy00             = _mm_sub_ps(iy0,jy0);
1468             dz00             = _mm_sub_ps(iz0,jz0);
1469             dx01             = _mm_sub_ps(ix0,jx1);
1470             dy01             = _mm_sub_ps(iy0,jy1);
1471             dz01             = _mm_sub_ps(iz0,jz1);
1472             dx02             = _mm_sub_ps(ix0,jx2);
1473             dy02             = _mm_sub_ps(iy0,jy2);
1474             dz02             = _mm_sub_ps(iz0,jz2);
1475             dx10             = _mm_sub_ps(ix1,jx0);
1476             dy10             = _mm_sub_ps(iy1,jy0);
1477             dz10             = _mm_sub_ps(iz1,jz0);
1478             dx11             = _mm_sub_ps(ix1,jx1);
1479             dy11             = _mm_sub_ps(iy1,jy1);
1480             dz11             = _mm_sub_ps(iz1,jz1);
1481             dx12             = _mm_sub_ps(ix1,jx2);
1482             dy12             = _mm_sub_ps(iy1,jy2);
1483             dz12             = _mm_sub_ps(iz1,jz2);
1484             dx20             = _mm_sub_ps(ix2,jx0);
1485             dy20             = _mm_sub_ps(iy2,jy0);
1486             dz20             = _mm_sub_ps(iz2,jz0);
1487             dx21             = _mm_sub_ps(ix2,jx1);
1488             dy21             = _mm_sub_ps(iy2,jy1);
1489             dz21             = _mm_sub_ps(iz2,jz1);
1490             dx22             = _mm_sub_ps(ix2,jx2);
1491             dy22             = _mm_sub_ps(iy2,jy2);
1492             dz22             = _mm_sub_ps(iz2,jz2);
1493
1494             /* Calculate squared distance and things based on it */
1495             rsq00            = gmx_mm_calc_rsq_ps(dx00,dy00,dz00);
1496             rsq01            = gmx_mm_calc_rsq_ps(dx01,dy01,dz01);
1497             rsq02            = gmx_mm_calc_rsq_ps(dx02,dy02,dz02);
1498             rsq10            = gmx_mm_calc_rsq_ps(dx10,dy10,dz10);
1499             rsq11            = gmx_mm_calc_rsq_ps(dx11,dy11,dz11);
1500             rsq12            = gmx_mm_calc_rsq_ps(dx12,dy12,dz12);
1501             rsq20            = gmx_mm_calc_rsq_ps(dx20,dy20,dz20);
1502             rsq21            = gmx_mm_calc_rsq_ps(dx21,dy21,dz21);
1503             rsq22            = gmx_mm_calc_rsq_ps(dx22,dy22,dz22);
1504
1505             rinv00           = gmx_mm_invsqrt_ps(rsq00);
1506             rinv01           = gmx_mm_invsqrt_ps(rsq01);
1507             rinv02           = gmx_mm_invsqrt_ps(rsq02);
1508             rinv10           = gmx_mm_invsqrt_ps(rsq10);
1509             rinv11           = gmx_mm_invsqrt_ps(rsq11);
1510             rinv12           = gmx_mm_invsqrt_ps(rsq12);
1511             rinv20           = gmx_mm_invsqrt_ps(rsq20);
1512             rinv21           = gmx_mm_invsqrt_ps(rsq21);
1513             rinv22           = gmx_mm_invsqrt_ps(rsq22);
1514
1515             fjx0             = _mm_setzero_ps();
1516             fjy0             = _mm_setzero_ps();
1517             fjz0             = _mm_setzero_ps();
1518             fjx1             = _mm_setzero_ps();
1519             fjy1             = _mm_setzero_ps();
1520             fjz1             = _mm_setzero_ps();
1521             fjx2             = _mm_setzero_ps();
1522             fjy2             = _mm_setzero_ps();
1523             fjz2             = _mm_setzero_ps();
1524
1525             /**************************
1526              * CALCULATE INTERACTIONS *
1527              **************************/
1528
1529             r00              = _mm_mul_ps(rsq00,rinv00);
1530
1531             /* Calculate table index by multiplying r with table scale and truncate to integer */
1532             rt               = _mm_mul_ps(r00,vftabscale);
1533             vfitab           = _mm_cvttps_epi32(rt);
1534 #ifdef __XOP__
1535             vfeps            = _mm_frcz_ps(rt);
1536 #else
1537             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
1538 #endif
1539             twovfeps         = _mm_add_ps(vfeps,vfeps);
1540             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
1541
1542             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1543             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
1544             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
1545             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
1546             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
1547             _MM_TRANSPOSE4_PS(Y,F,G,H);
1548             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
1549             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
1550             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq00,FF),_mm_mul_ps(vftabscale,rinv00)));
1551
1552             /* CUBIC SPLINE TABLE DISPERSION */
1553             vfitab           = _mm_add_epi32(vfitab,ifour);
1554             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
1555             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
1556             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
1557             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
1558             _MM_TRANSPOSE4_PS(Y,F,G,H);
1559             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
1560             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
1561             fvdw6            = _mm_mul_ps(c6_00,FF);
1562
1563             /* CUBIC SPLINE TABLE REPULSION */
1564             vfitab           = _mm_add_epi32(vfitab,ifour);
1565             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
1566             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
1567             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
1568             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
1569             _MM_TRANSPOSE4_PS(Y,F,G,H);
1570             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
1571             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
1572             fvdw12           = _mm_mul_ps(c12_00,FF);
1573             fvdw             = _mm_xor_ps(signbit,_mm_mul_ps(_mm_add_ps(fvdw6,fvdw12),_mm_mul_ps(vftabscale,rinv00)));
1574
1575             fscal            = _mm_add_ps(felec,fvdw);
1576
1577              /* Update vectorial force */
1578             fix0             = _mm_macc_ps(dx00,fscal,fix0);
1579             fiy0             = _mm_macc_ps(dy00,fscal,fiy0);
1580             fiz0             = _mm_macc_ps(dz00,fscal,fiz0);
1581
1582             fjx0             = _mm_macc_ps(dx00,fscal,fjx0);
1583             fjy0             = _mm_macc_ps(dy00,fscal,fjy0);
1584             fjz0             = _mm_macc_ps(dz00,fscal,fjz0);
1585
1586             /**************************
1587              * CALCULATE INTERACTIONS *
1588              **************************/
1589
1590             r01              = _mm_mul_ps(rsq01,rinv01);
1591
1592             /* Calculate table index by multiplying r with table scale and truncate to integer */
1593             rt               = _mm_mul_ps(r01,vftabscale);
1594             vfitab           = _mm_cvttps_epi32(rt);
1595 #ifdef __XOP__
1596             vfeps            = _mm_frcz_ps(rt);
1597 #else
1598             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
1599 #endif
1600             twovfeps         = _mm_add_ps(vfeps,vfeps);
1601             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
1602
1603             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1604             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
1605             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
1606             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
1607             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
1608             _MM_TRANSPOSE4_PS(Y,F,G,H);
1609             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
1610             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
1611             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq01,FF),_mm_mul_ps(vftabscale,rinv01)));
1612
1613             fscal            = felec;
1614
1615              /* Update vectorial force */
1616             fix0             = _mm_macc_ps(dx01,fscal,fix0);
1617             fiy0             = _mm_macc_ps(dy01,fscal,fiy0);
1618             fiz0             = _mm_macc_ps(dz01,fscal,fiz0);
1619
1620             fjx1             = _mm_macc_ps(dx01,fscal,fjx1);
1621             fjy1             = _mm_macc_ps(dy01,fscal,fjy1);
1622             fjz1             = _mm_macc_ps(dz01,fscal,fjz1);
1623
1624             /**************************
1625              * CALCULATE INTERACTIONS *
1626              **************************/
1627
1628             r02              = _mm_mul_ps(rsq02,rinv02);
1629
1630             /* Calculate table index by multiplying r with table scale and truncate to integer */
1631             rt               = _mm_mul_ps(r02,vftabscale);
1632             vfitab           = _mm_cvttps_epi32(rt);
1633 #ifdef __XOP__
1634             vfeps            = _mm_frcz_ps(rt);
1635 #else
1636             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
1637 #endif
1638             twovfeps         = _mm_add_ps(vfeps,vfeps);
1639             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
1640
1641             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1642             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
1643             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
1644             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
1645             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
1646             _MM_TRANSPOSE4_PS(Y,F,G,H);
1647             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
1648             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
1649             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq02,FF),_mm_mul_ps(vftabscale,rinv02)));
1650
1651             fscal            = felec;
1652
1653              /* Update vectorial force */
1654             fix0             = _mm_macc_ps(dx02,fscal,fix0);
1655             fiy0             = _mm_macc_ps(dy02,fscal,fiy0);
1656             fiz0             = _mm_macc_ps(dz02,fscal,fiz0);
1657
1658             fjx2             = _mm_macc_ps(dx02,fscal,fjx2);
1659             fjy2             = _mm_macc_ps(dy02,fscal,fjy2);
1660             fjz2             = _mm_macc_ps(dz02,fscal,fjz2);
1661
1662             /**************************
1663              * CALCULATE INTERACTIONS *
1664              **************************/
1665
1666             r10              = _mm_mul_ps(rsq10,rinv10);
1667
1668             /* Calculate table index by multiplying r with table scale and truncate to integer */
1669             rt               = _mm_mul_ps(r10,vftabscale);
1670             vfitab           = _mm_cvttps_epi32(rt);
1671 #ifdef __XOP__
1672             vfeps            = _mm_frcz_ps(rt);
1673 #else
1674             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
1675 #endif
1676             twovfeps         = _mm_add_ps(vfeps,vfeps);
1677             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
1678
1679             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1680             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
1681             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
1682             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
1683             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
1684             _MM_TRANSPOSE4_PS(Y,F,G,H);
1685             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
1686             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
1687             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq10,FF),_mm_mul_ps(vftabscale,rinv10)));
1688
1689             fscal            = felec;
1690
1691              /* Update vectorial force */
1692             fix1             = _mm_macc_ps(dx10,fscal,fix1);
1693             fiy1             = _mm_macc_ps(dy10,fscal,fiy1);
1694             fiz1             = _mm_macc_ps(dz10,fscal,fiz1);
1695
1696             fjx0             = _mm_macc_ps(dx10,fscal,fjx0);
1697             fjy0             = _mm_macc_ps(dy10,fscal,fjy0);
1698             fjz0             = _mm_macc_ps(dz10,fscal,fjz0);
1699
1700             /**************************
1701              * CALCULATE INTERACTIONS *
1702              **************************/
1703
1704             r11              = _mm_mul_ps(rsq11,rinv11);
1705
1706             /* Calculate table index by multiplying r with table scale and truncate to integer */
1707             rt               = _mm_mul_ps(r11,vftabscale);
1708             vfitab           = _mm_cvttps_epi32(rt);
1709 #ifdef __XOP__
1710             vfeps            = _mm_frcz_ps(rt);
1711 #else
1712             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
1713 #endif
1714             twovfeps         = _mm_add_ps(vfeps,vfeps);
1715             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
1716
1717             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1718             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
1719             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
1720             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
1721             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
1722             _MM_TRANSPOSE4_PS(Y,F,G,H);
1723             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
1724             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
1725             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq11,FF),_mm_mul_ps(vftabscale,rinv11)));
1726
1727             fscal            = felec;
1728
1729              /* Update vectorial force */
1730             fix1             = _mm_macc_ps(dx11,fscal,fix1);
1731             fiy1             = _mm_macc_ps(dy11,fscal,fiy1);
1732             fiz1             = _mm_macc_ps(dz11,fscal,fiz1);
1733
1734             fjx1             = _mm_macc_ps(dx11,fscal,fjx1);
1735             fjy1             = _mm_macc_ps(dy11,fscal,fjy1);
1736             fjz1             = _mm_macc_ps(dz11,fscal,fjz1);
1737
1738             /**************************
1739              * CALCULATE INTERACTIONS *
1740              **************************/
1741
1742             r12              = _mm_mul_ps(rsq12,rinv12);
1743
1744             /* Calculate table index by multiplying r with table scale and truncate to integer */
1745             rt               = _mm_mul_ps(r12,vftabscale);
1746             vfitab           = _mm_cvttps_epi32(rt);
1747 #ifdef __XOP__
1748             vfeps            = _mm_frcz_ps(rt);
1749 #else
1750             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
1751 #endif
1752             twovfeps         = _mm_add_ps(vfeps,vfeps);
1753             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
1754
1755             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1756             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
1757             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
1758             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
1759             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
1760             _MM_TRANSPOSE4_PS(Y,F,G,H);
1761             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
1762             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
1763             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq12,FF),_mm_mul_ps(vftabscale,rinv12)));
1764
1765             fscal            = felec;
1766
1767              /* Update vectorial force */
1768             fix1             = _mm_macc_ps(dx12,fscal,fix1);
1769             fiy1             = _mm_macc_ps(dy12,fscal,fiy1);
1770             fiz1             = _mm_macc_ps(dz12,fscal,fiz1);
1771
1772             fjx2             = _mm_macc_ps(dx12,fscal,fjx2);
1773             fjy2             = _mm_macc_ps(dy12,fscal,fjy2);
1774             fjz2             = _mm_macc_ps(dz12,fscal,fjz2);
1775
1776             /**************************
1777              * CALCULATE INTERACTIONS *
1778              **************************/
1779
1780             r20              = _mm_mul_ps(rsq20,rinv20);
1781
1782             /* Calculate table index by multiplying r with table scale and truncate to integer */
1783             rt               = _mm_mul_ps(r20,vftabscale);
1784             vfitab           = _mm_cvttps_epi32(rt);
1785 #ifdef __XOP__
1786             vfeps            = _mm_frcz_ps(rt);
1787 #else
1788             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
1789 #endif
1790             twovfeps         = _mm_add_ps(vfeps,vfeps);
1791             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
1792
1793             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1794             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
1795             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
1796             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
1797             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
1798             _MM_TRANSPOSE4_PS(Y,F,G,H);
1799             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
1800             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
1801             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq20,FF),_mm_mul_ps(vftabscale,rinv20)));
1802
1803             fscal            = felec;
1804
1805              /* Update vectorial force */
1806             fix2             = _mm_macc_ps(dx20,fscal,fix2);
1807             fiy2             = _mm_macc_ps(dy20,fscal,fiy2);
1808             fiz2             = _mm_macc_ps(dz20,fscal,fiz2);
1809
1810             fjx0             = _mm_macc_ps(dx20,fscal,fjx0);
1811             fjy0             = _mm_macc_ps(dy20,fscal,fjy0);
1812             fjz0             = _mm_macc_ps(dz20,fscal,fjz0);
1813
1814             /**************************
1815              * CALCULATE INTERACTIONS *
1816              **************************/
1817
1818             r21              = _mm_mul_ps(rsq21,rinv21);
1819
1820             /* Calculate table index by multiplying r with table scale and truncate to integer */
1821             rt               = _mm_mul_ps(r21,vftabscale);
1822             vfitab           = _mm_cvttps_epi32(rt);
1823 #ifdef __XOP__
1824             vfeps            = _mm_frcz_ps(rt);
1825 #else
1826             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
1827 #endif
1828             twovfeps         = _mm_add_ps(vfeps,vfeps);
1829             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
1830
1831             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1832             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
1833             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
1834             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
1835             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
1836             _MM_TRANSPOSE4_PS(Y,F,G,H);
1837             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
1838             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
1839             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq21,FF),_mm_mul_ps(vftabscale,rinv21)));
1840
1841             fscal            = felec;
1842
1843              /* Update vectorial force */
1844             fix2             = _mm_macc_ps(dx21,fscal,fix2);
1845             fiy2             = _mm_macc_ps(dy21,fscal,fiy2);
1846             fiz2             = _mm_macc_ps(dz21,fscal,fiz2);
1847
1848             fjx1             = _mm_macc_ps(dx21,fscal,fjx1);
1849             fjy1             = _mm_macc_ps(dy21,fscal,fjy1);
1850             fjz1             = _mm_macc_ps(dz21,fscal,fjz1);
1851
1852             /**************************
1853              * CALCULATE INTERACTIONS *
1854              **************************/
1855
1856             r22              = _mm_mul_ps(rsq22,rinv22);
1857
1858             /* Calculate table index by multiplying r with table scale and truncate to integer */
1859             rt               = _mm_mul_ps(r22,vftabscale);
1860             vfitab           = _mm_cvttps_epi32(rt);
1861 #ifdef __XOP__
1862             vfeps            = _mm_frcz_ps(rt);
1863 #else
1864             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
1865 #endif
1866             twovfeps         = _mm_add_ps(vfeps,vfeps);
1867             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
1868
1869             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1870             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
1871             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
1872             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
1873             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
1874             _MM_TRANSPOSE4_PS(Y,F,G,H);
1875             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
1876             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
1877             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq22,FF),_mm_mul_ps(vftabscale,rinv22)));
1878
1879             fscal            = felec;
1880
1881              /* Update vectorial force */
1882             fix2             = _mm_macc_ps(dx22,fscal,fix2);
1883             fiy2             = _mm_macc_ps(dy22,fscal,fiy2);
1884             fiz2             = _mm_macc_ps(dz22,fscal,fiz2);
1885
1886             fjx2             = _mm_macc_ps(dx22,fscal,fjx2);
1887             fjy2             = _mm_macc_ps(dy22,fscal,fjy2);
1888             fjz2             = _mm_macc_ps(dz22,fscal,fjz2);
1889
1890             fjptrA             = f+j_coord_offsetA;
1891             fjptrB             = f+j_coord_offsetB;
1892             fjptrC             = f+j_coord_offsetC;
1893             fjptrD             = f+j_coord_offsetD;
1894
1895             gmx_mm_decrement_3rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,
1896                                                    fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2);
1897
1898             /* Inner loop uses 400 flops */
1899         }
1900
1901         if(jidx<j_index_end)
1902         {
1903
1904             /* Get j neighbor index, and coordinate index */
1905             jnrlistA         = jjnr[jidx];
1906             jnrlistB         = jjnr[jidx+1];
1907             jnrlistC         = jjnr[jidx+2];
1908             jnrlistD         = jjnr[jidx+3];
1909             /* Sign of each element will be negative for non-real atoms.
1910              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
1911              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
1912              */
1913             dummy_mask = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
1914             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
1915             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
1916             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
1917             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
1918             j_coord_offsetA  = DIM*jnrA;
1919             j_coord_offsetB  = DIM*jnrB;
1920             j_coord_offsetC  = DIM*jnrC;
1921             j_coord_offsetD  = DIM*jnrD;
1922
1923             /* load j atom coordinates */
1924             gmx_mm_load_3rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
1925                                               x+j_coord_offsetC,x+j_coord_offsetD,
1926                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,&jy2,&jz2);
1927
1928             /* Calculate displacement vector */
1929             dx00             = _mm_sub_ps(ix0,jx0);
1930             dy00             = _mm_sub_ps(iy0,jy0);
1931             dz00             = _mm_sub_ps(iz0,jz0);
1932             dx01             = _mm_sub_ps(ix0,jx1);
1933             dy01             = _mm_sub_ps(iy0,jy1);
1934             dz01             = _mm_sub_ps(iz0,jz1);
1935             dx02             = _mm_sub_ps(ix0,jx2);
1936             dy02             = _mm_sub_ps(iy0,jy2);
1937             dz02             = _mm_sub_ps(iz0,jz2);
1938             dx10             = _mm_sub_ps(ix1,jx0);
1939             dy10             = _mm_sub_ps(iy1,jy0);
1940             dz10             = _mm_sub_ps(iz1,jz0);
1941             dx11             = _mm_sub_ps(ix1,jx1);
1942             dy11             = _mm_sub_ps(iy1,jy1);
1943             dz11             = _mm_sub_ps(iz1,jz1);
1944             dx12             = _mm_sub_ps(ix1,jx2);
1945             dy12             = _mm_sub_ps(iy1,jy2);
1946             dz12             = _mm_sub_ps(iz1,jz2);
1947             dx20             = _mm_sub_ps(ix2,jx0);
1948             dy20             = _mm_sub_ps(iy2,jy0);
1949             dz20             = _mm_sub_ps(iz2,jz0);
1950             dx21             = _mm_sub_ps(ix2,jx1);
1951             dy21             = _mm_sub_ps(iy2,jy1);
1952             dz21             = _mm_sub_ps(iz2,jz1);
1953             dx22             = _mm_sub_ps(ix2,jx2);
1954             dy22             = _mm_sub_ps(iy2,jy2);
1955             dz22             = _mm_sub_ps(iz2,jz2);
1956
1957             /* Calculate squared distance and things based on it */
1958             rsq00            = gmx_mm_calc_rsq_ps(dx00,dy00,dz00);
1959             rsq01            = gmx_mm_calc_rsq_ps(dx01,dy01,dz01);
1960             rsq02            = gmx_mm_calc_rsq_ps(dx02,dy02,dz02);
1961             rsq10            = gmx_mm_calc_rsq_ps(dx10,dy10,dz10);
1962             rsq11            = gmx_mm_calc_rsq_ps(dx11,dy11,dz11);
1963             rsq12            = gmx_mm_calc_rsq_ps(dx12,dy12,dz12);
1964             rsq20            = gmx_mm_calc_rsq_ps(dx20,dy20,dz20);
1965             rsq21            = gmx_mm_calc_rsq_ps(dx21,dy21,dz21);
1966             rsq22            = gmx_mm_calc_rsq_ps(dx22,dy22,dz22);
1967
1968             rinv00           = gmx_mm_invsqrt_ps(rsq00);
1969             rinv01           = gmx_mm_invsqrt_ps(rsq01);
1970             rinv02           = gmx_mm_invsqrt_ps(rsq02);
1971             rinv10           = gmx_mm_invsqrt_ps(rsq10);
1972             rinv11           = gmx_mm_invsqrt_ps(rsq11);
1973             rinv12           = gmx_mm_invsqrt_ps(rsq12);
1974             rinv20           = gmx_mm_invsqrt_ps(rsq20);
1975             rinv21           = gmx_mm_invsqrt_ps(rsq21);
1976             rinv22           = gmx_mm_invsqrt_ps(rsq22);
1977
1978             fjx0             = _mm_setzero_ps();
1979             fjy0             = _mm_setzero_ps();
1980             fjz0             = _mm_setzero_ps();
1981             fjx1             = _mm_setzero_ps();
1982             fjy1             = _mm_setzero_ps();
1983             fjz1             = _mm_setzero_ps();
1984             fjx2             = _mm_setzero_ps();
1985             fjy2             = _mm_setzero_ps();
1986             fjz2             = _mm_setzero_ps();
1987
1988             /**************************
1989              * CALCULATE INTERACTIONS *
1990              **************************/
1991
1992             r00              = _mm_mul_ps(rsq00,rinv00);
1993             r00              = _mm_andnot_ps(dummy_mask,r00);
1994
1995             /* Calculate table index by multiplying r with table scale and truncate to integer */
1996             rt               = _mm_mul_ps(r00,vftabscale);
1997             vfitab           = _mm_cvttps_epi32(rt);
1998 #ifdef __XOP__
1999             vfeps            = _mm_frcz_ps(rt);
2000 #else
2001             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
2002 #endif
2003             twovfeps         = _mm_add_ps(vfeps,vfeps);
2004             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
2005
2006             /* CUBIC SPLINE TABLE ELECTROSTATICS */
2007             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
2008             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
2009             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
2010             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
2011             _MM_TRANSPOSE4_PS(Y,F,G,H);
2012             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
2013             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
2014             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq00,FF),_mm_mul_ps(vftabscale,rinv00)));
2015
2016             /* CUBIC SPLINE TABLE DISPERSION */
2017             vfitab           = _mm_add_epi32(vfitab,ifour);
2018             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
2019             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
2020             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
2021             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
2022             _MM_TRANSPOSE4_PS(Y,F,G,H);
2023             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
2024             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
2025             fvdw6            = _mm_mul_ps(c6_00,FF);
2026
2027             /* CUBIC SPLINE TABLE REPULSION */
2028             vfitab           = _mm_add_epi32(vfitab,ifour);
2029             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
2030             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
2031             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
2032             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
2033             _MM_TRANSPOSE4_PS(Y,F,G,H);
2034             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
2035             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
2036             fvdw12           = _mm_mul_ps(c12_00,FF);
2037             fvdw             = _mm_xor_ps(signbit,_mm_mul_ps(_mm_add_ps(fvdw6,fvdw12),_mm_mul_ps(vftabscale,rinv00)));
2038
2039             fscal            = _mm_add_ps(felec,fvdw);
2040
2041             fscal            = _mm_andnot_ps(dummy_mask,fscal);
2042
2043              /* Update vectorial force */
2044             fix0             = _mm_macc_ps(dx00,fscal,fix0);
2045             fiy0             = _mm_macc_ps(dy00,fscal,fiy0);
2046             fiz0             = _mm_macc_ps(dz00,fscal,fiz0);
2047
2048             fjx0             = _mm_macc_ps(dx00,fscal,fjx0);
2049             fjy0             = _mm_macc_ps(dy00,fscal,fjy0);
2050             fjz0             = _mm_macc_ps(dz00,fscal,fjz0);
2051
2052             /**************************
2053              * CALCULATE INTERACTIONS *
2054              **************************/
2055
2056             r01              = _mm_mul_ps(rsq01,rinv01);
2057             r01              = _mm_andnot_ps(dummy_mask,r01);
2058
2059             /* Calculate table index by multiplying r with table scale and truncate to integer */
2060             rt               = _mm_mul_ps(r01,vftabscale);
2061             vfitab           = _mm_cvttps_epi32(rt);
2062 #ifdef __XOP__
2063             vfeps            = _mm_frcz_ps(rt);
2064 #else
2065             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
2066 #endif
2067             twovfeps         = _mm_add_ps(vfeps,vfeps);
2068             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
2069
2070             /* CUBIC SPLINE TABLE ELECTROSTATICS */
2071             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
2072             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
2073             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
2074             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
2075             _MM_TRANSPOSE4_PS(Y,F,G,H);
2076             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
2077             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
2078             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq01,FF),_mm_mul_ps(vftabscale,rinv01)));
2079
2080             fscal            = felec;
2081
2082             fscal            = _mm_andnot_ps(dummy_mask,fscal);
2083
2084              /* Update vectorial force */
2085             fix0             = _mm_macc_ps(dx01,fscal,fix0);
2086             fiy0             = _mm_macc_ps(dy01,fscal,fiy0);
2087             fiz0             = _mm_macc_ps(dz01,fscal,fiz0);
2088
2089             fjx1             = _mm_macc_ps(dx01,fscal,fjx1);
2090             fjy1             = _mm_macc_ps(dy01,fscal,fjy1);
2091             fjz1             = _mm_macc_ps(dz01,fscal,fjz1);
2092
2093             /**************************
2094              * CALCULATE INTERACTIONS *
2095              **************************/
2096
2097             r02              = _mm_mul_ps(rsq02,rinv02);
2098             r02              = _mm_andnot_ps(dummy_mask,r02);
2099
2100             /* Calculate table index by multiplying r with table scale and truncate to integer */
2101             rt               = _mm_mul_ps(r02,vftabscale);
2102             vfitab           = _mm_cvttps_epi32(rt);
2103 #ifdef __XOP__
2104             vfeps            = _mm_frcz_ps(rt);
2105 #else
2106             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
2107 #endif
2108             twovfeps         = _mm_add_ps(vfeps,vfeps);
2109             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
2110
2111             /* CUBIC SPLINE TABLE ELECTROSTATICS */
2112             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
2113             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
2114             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
2115             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
2116             _MM_TRANSPOSE4_PS(Y,F,G,H);
2117             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
2118             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
2119             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq02,FF),_mm_mul_ps(vftabscale,rinv02)));
2120
2121             fscal            = felec;
2122
2123             fscal            = _mm_andnot_ps(dummy_mask,fscal);
2124
2125              /* Update vectorial force */
2126             fix0             = _mm_macc_ps(dx02,fscal,fix0);
2127             fiy0             = _mm_macc_ps(dy02,fscal,fiy0);
2128             fiz0             = _mm_macc_ps(dz02,fscal,fiz0);
2129
2130             fjx2             = _mm_macc_ps(dx02,fscal,fjx2);
2131             fjy2             = _mm_macc_ps(dy02,fscal,fjy2);
2132             fjz2             = _mm_macc_ps(dz02,fscal,fjz2);
2133
2134             /**************************
2135              * CALCULATE INTERACTIONS *
2136              **************************/
2137
2138             r10              = _mm_mul_ps(rsq10,rinv10);
2139             r10              = _mm_andnot_ps(dummy_mask,r10);
2140
2141             /* Calculate table index by multiplying r with table scale and truncate to integer */
2142             rt               = _mm_mul_ps(r10,vftabscale);
2143             vfitab           = _mm_cvttps_epi32(rt);
2144 #ifdef __XOP__
2145             vfeps            = _mm_frcz_ps(rt);
2146 #else
2147             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
2148 #endif
2149             twovfeps         = _mm_add_ps(vfeps,vfeps);
2150             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
2151
2152             /* CUBIC SPLINE TABLE ELECTROSTATICS */
2153             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
2154             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
2155             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
2156             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
2157             _MM_TRANSPOSE4_PS(Y,F,G,H);
2158             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
2159             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
2160             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq10,FF),_mm_mul_ps(vftabscale,rinv10)));
2161
2162             fscal            = felec;
2163
2164             fscal            = _mm_andnot_ps(dummy_mask,fscal);
2165
2166              /* Update vectorial force */
2167             fix1             = _mm_macc_ps(dx10,fscal,fix1);
2168             fiy1             = _mm_macc_ps(dy10,fscal,fiy1);
2169             fiz1             = _mm_macc_ps(dz10,fscal,fiz1);
2170
2171             fjx0             = _mm_macc_ps(dx10,fscal,fjx0);
2172             fjy0             = _mm_macc_ps(dy10,fscal,fjy0);
2173             fjz0             = _mm_macc_ps(dz10,fscal,fjz0);
2174
2175             /**************************
2176              * CALCULATE INTERACTIONS *
2177              **************************/
2178
2179             r11              = _mm_mul_ps(rsq11,rinv11);
2180             r11              = _mm_andnot_ps(dummy_mask,r11);
2181
2182             /* Calculate table index by multiplying r with table scale and truncate to integer */
2183             rt               = _mm_mul_ps(r11,vftabscale);
2184             vfitab           = _mm_cvttps_epi32(rt);
2185 #ifdef __XOP__
2186             vfeps            = _mm_frcz_ps(rt);
2187 #else
2188             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
2189 #endif
2190             twovfeps         = _mm_add_ps(vfeps,vfeps);
2191             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
2192
2193             /* CUBIC SPLINE TABLE ELECTROSTATICS */
2194             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
2195             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
2196             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
2197             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
2198             _MM_TRANSPOSE4_PS(Y,F,G,H);
2199             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
2200             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
2201             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq11,FF),_mm_mul_ps(vftabscale,rinv11)));
2202
2203             fscal            = felec;
2204
2205             fscal            = _mm_andnot_ps(dummy_mask,fscal);
2206
2207              /* Update vectorial force */
2208             fix1             = _mm_macc_ps(dx11,fscal,fix1);
2209             fiy1             = _mm_macc_ps(dy11,fscal,fiy1);
2210             fiz1             = _mm_macc_ps(dz11,fscal,fiz1);
2211
2212             fjx1             = _mm_macc_ps(dx11,fscal,fjx1);
2213             fjy1             = _mm_macc_ps(dy11,fscal,fjy1);
2214             fjz1             = _mm_macc_ps(dz11,fscal,fjz1);
2215
2216             /**************************
2217              * CALCULATE INTERACTIONS *
2218              **************************/
2219
2220             r12              = _mm_mul_ps(rsq12,rinv12);
2221             r12              = _mm_andnot_ps(dummy_mask,r12);
2222
2223             /* Calculate table index by multiplying r with table scale and truncate to integer */
2224             rt               = _mm_mul_ps(r12,vftabscale);
2225             vfitab           = _mm_cvttps_epi32(rt);
2226 #ifdef __XOP__
2227             vfeps            = _mm_frcz_ps(rt);
2228 #else
2229             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
2230 #endif
2231             twovfeps         = _mm_add_ps(vfeps,vfeps);
2232             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
2233
2234             /* CUBIC SPLINE TABLE ELECTROSTATICS */
2235             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
2236             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
2237             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
2238             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
2239             _MM_TRANSPOSE4_PS(Y,F,G,H);
2240             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
2241             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
2242             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq12,FF),_mm_mul_ps(vftabscale,rinv12)));
2243
2244             fscal            = felec;
2245
2246             fscal            = _mm_andnot_ps(dummy_mask,fscal);
2247
2248              /* Update vectorial force */
2249             fix1             = _mm_macc_ps(dx12,fscal,fix1);
2250             fiy1             = _mm_macc_ps(dy12,fscal,fiy1);
2251             fiz1             = _mm_macc_ps(dz12,fscal,fiz1);
2252
2253             fjx2             = _mm_macc_ps(dx12,fscal,fjx2);
2254             fjy2             = _mm_macc_ps(dy12,fscal,fjy2);
2255             fjz2             = _mm_macc_ps(dz12,fscal,fjz2);
2256
2257             /**************************
2258              * CALCULATE INTERACTIONS *
2259              **************************/
2260
2261             r20              = _mm_mul_ps(rsq20,rinv20);
2262             r20              = _mm_andnot_ps(dummy_mask,r20);
2263
2264             /* Calculate table index by multiplying r with table scale and truncate to integer */
2265             rt               = _mm_mul_ps(r20,vftabscale);
2266             vfitab           = _mm_cvttps_epi32(rt);
2267 #ifdef __XOP__
2268             vfeps            = _mm_frcz_ps(rt);
2269 #else
2270             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
2271 #endif
2272             twovfeps         = _mm_add_ps(vfeps,vfeps);
2273             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
2274
2275             /* CUBIC SPLINE TABLE ELECTROSTATICS */
2276             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
2277             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
2278             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
2279             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
2280             _MM_TRANSPOSE4_PS(Y,F,G,H);
2281             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
2282             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
2283             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq20,FF),_mm_mul_ps(vftabscale,rinv20)));
2284
2285             fscal            = felec;
2286
2287             fscal            = _mm_andnot_ps(dummy_mask,fscal);
2288
2289              /* Update vectorial force */
2290             fix2             = _mm_macc_ps(dx20,fscal,fix2);
2291             fiy2             = _mm_macc_ps(dy20,fscal,fiy2);
2292             fiz2             = _mm_macc_ps(dz20,fscal,fiz2);
2293
2294             fjx0             = _mm_macc_ps(dx20,fscal,fjx0);
2295             fjy0             = _mm_macc_ps(dy20,fscal,fjy0);
2296             fjz0             = _mm_macc_ps(dz20,fscal,fjz0);
2297
2298             /**************************
2299              * CALCULATE INTERACTIONS *
2300              **************************/
2301
2302             r21              = _mm_mul_ps(rsq21,rinv21);
2303             r21              = _mm_andnot_ps(dummy_mask,r21);
2304
2305             /* Calculate table index by multiplying r with table scale and truncate to integer */
2306             rt               = _mm_mul_ps(r21,vftabscale);
2307             vfitab           = _mm_cvttps_epi32(rt);
2308 #ifdef __XOP__
2309             vfeps            = _mm_frcz_ps(rt);
2310 #else
2311             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
2312 #endif
2313             twovfeps         = _mm_add_ps(vfeps,vfeps);
2314             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
2315
2316             /* CUBIC SPLINE TABLE ELECTROSTATICS */
2317             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
2318             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
2319             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
2320             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
2321             _MM_TRANSPOSE4_PS(Y,F,G,H);
2322             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
2323             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
2324             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq21,FF),_mm_mul_ps(vftabscale,rinv21)));
2325
2326             fscal            = felec;
2327
2328             fscal            = _mm_andnot_ps(dummy_mask,fscal);
2329
2330              /* Update vectorial force */
2331             fix2             = _mm_macc_ps(dx21,fscal,fix2);
2332             fiy2             = _mm_macc_ps(dy21,fscal,fiy2);
2333             fiz2             = _mm_macc_ps(dz21,fscal,fiz2);
2334
2335             fjx1             = _mm_macc_ps(dx21,fscal,fjx1);
2336             fjy1             = _mm_macc_ps(dy21,fscal,fjy1);
2337             fjz1             = _mm_macc_ps(dz21,fscal,fjz1);
2338
2339             /**************************
2340              * CALCULATE INTERACTIONS *
2341              **************************/
2342
2343             r22              = _mm_mul_ps(rsq22,rinv22);
2344             r22              = _mm_andnot_ps(dummy_mask,r22);
2345
2346             /* Calculate table index by multiplying r with table scale and truncate to integer */
2347             rt               = _mm_mul_ps(r22,vftabscale);
2348             vfitab           = _mm_cvttps_epi32(rt);
2349 #ifdef __XOP__
2350             vfeps            = _mm_frcz_ps(rt);
2351 #else
2352             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
2353 #endif
2354             twovfeps         = _mm_add_ps(vfeps,vfeps);
2355             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
2356
2357             /* CUBIC SPLINE TABLE ELECTROSTATICS */
2358             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
2359             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
2360             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
2361             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
2362             _MM_TRANSPOSE4_PS(Y,F,G,H);
2363             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
2364             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
2365             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq22,FF),_mm_mul_ps(vftabscale,rinv22)));
2366
2367             fscal            = felec;
2368
2369             fscal            = _mm_andnot_ps(dummy_mask,fscal);
2370
2371              /* Update vectorial force */
2372             fix2             = _mm_macc_ps(dx22,fscal,fix2);
2373             fiy2             = _mm_macc_ps(dy22,fscal,fiy2);
2374             fiz2             = _mm_macc_ps(dz22,fscal,fiz2);
2375
2376             fjx2             = _mm_macc_ps(dx22,fscal,fjx2);
2377             fjy2             = _mm_macc_ps(dy22,fscal,fjy2);
2378             fjz2             = _mm_macc_ps(dz22,fscal,fjz2);
2379
2380             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
2381             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
2382             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
2383             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
2384
2385             gmx_mm_decrement_3rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,
2386                                                    fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2);
2387
2388             /* Inner loop uses 409 flops */
2389         }
2390
2391         /* End of innermost loop */
2392
2393         gmx_mm_update_iforce_3atom_swizzle_ps(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,
2394                                               f+i_coord_offset,fshift+i_shift_offset);
2395
2396         /* Increment number of inner iterations */
2397         inneriter                  += j_index_end - j_index_start;
2398
2399         /* Outer loop uses 18 flops */
2400     }
2401
2402     /* Increment number of outer iterations */
2403     outeriter        += nri;
2404
2405     /* Update outer/inner flops */
2406
2407     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W3W3_F,outeriter*18 + inneriter*409);
2408 }