src/gromacs/gmxlib/nonbonded/nb_kernel_avx_128_fma_double/nb_kernel_template_avx_128_fma_double.pre

   1 /* #if 0 */
   2 /*
   3  * This file is part of the GROMACS molecular simulation package.
   4  *
   5  * Copyright (c) 2012,2013,2014, by the GROMACS development team, led by
   6  * Mark Abraham, David van der Spoel, Berk Hess, and Erik Lindahl,
   7  * and including many others, as listed in the AUTHORS file in the
   8  * top-level source directory and at http://www.gromacs.org.
   9  *
  10  * GROMACS is free software; you can redistribute it and/or
  11  * modify it under the terms of the GNU Lesser General Public License
  12  * as published by the Free Software Foundation; either version 2.1
  13  * of the License, or (at your option) any later version.
  14  *
  15  * GROMACS is distributed in the hope that it will be useful,
  16  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  17  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  18  * Lesser General Public License for more details.
  19  *
  20  * You should have received a copy of the GNU Lesser General Public
  21  * License along with GROMACS; if not, see
  22  * http://www.gnu.org/licenses, or write to the Free Software Foundation,
  23  * Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA.
  24  *
  25  * If you want to redistribute modifications to GROMACS, please
  26  * consider that scientific software is very special. Version
  27  * control is crucial - bugs must be traceable. We will be happy to
  28  * consider code for inclusion in the official distribution, but
  29  * derived work must not be called official GROMACS. Details are found
  30  * in the README & COPYING files - if they are missing, get the
  31  * official version at http://www.gromacs.org.
  32  *
  33  * To help us fund GROMACS development, we humbly ask that you cite
  34  * the research papers on the package. Check out http://www.gromacs.org.
  35  */
  36 #error This file must be processed with the Gromacs pre-preprocessor
  37 /* #endif */
  38 /* #if INCLUDE_HEADER */
  39 #include "gmxpre.h"
  40
  41 #include "config.h"
  42
  43 #include <math.h>
  44
  45 #include "../nb_kernel.h"
  46 #include "gromacs/legacyheaders/types/simple.h"
  47 #include "gromacs/math/vec.h"
  48 #include "gromacs/legacyheaders/nrnb.h"
  49
  50 #include "gromacs/simd/math_x86_avx_128_fma_double.h"
  51 #include "kernelutil_x86_avx_128_fma_double.h"
  52 /* #endif */
  53
  54 /* ## List of variables set by the generating script:                                    */
  55 /* ##                                                                                    */
  56 /* ## Setttings that apply to the entire kernel:                                         */
  57 /* ## KERNEL_ELEC:           String, choice for electrostatic interactions               */
  58 /* ## KERNEL_VDW:            String, choice for van der Waals interactions               */
  59 /* ## KERNEL_NAME:           String, name of this kernel                                 */
  60 /* ## KERNEL_VF:             String telling if we calculate potential, force, or both    */
  61 /* ## GEOMETRY_I/GEOMETRY_J: String, name of each geometry, e.g. 'Water3' or '1Particle' */
  62 /* ##                                                                                    */
  63 /* ## Setttings that apply to particles in the outer (I) or inner (J) loops:             */
  64 /* ## PARTICLES_I[]/         Arrays with lists of i/j particles to use in kernel. It is  */
  65 /* ## PARTICLES_J[]:         just [0] for particle geometry, but can be longer for water */
  66 /* ## PARTICLES_ELEC_I[]/    Arrays with lists of i/j particle that have electrostatics  */
  67 /* ## PARTICLES_ELEC_J[]:    interactions that should be calculated in this kernel.      */
  68 /* ## PARTICLES_VDW_I[]/     Arrays with the list of i/j particle that have VdW          */
  69 /* ## PARTICLES_VDW_J[]:     interactions that should be calculated in this kernel.      */
  70 /* ##                                                                                    */
  71 /* ## Setttings for pairs of interactions (e.g. 2nd i particle against 1st j particle)   */
  72 /* ## PAIRS_IJ[]:            Array with (i,j) tuples of pairs for which interactions     */
  73 /* ##                        should be calculated in this kernel. Zero-charge particles  */
  74 /* ##                        do not have interactions with particles without vdw, and    */
  75 /* ##                        Vdw-only interactions are not evaluated in a no-vdw-kernel. */
  76 /* ## INTERACTION_FLAGS[][]: 2D matrix, dimension e.g. 3*3 for water-water interactions. */
  77 /* ##                        For each i-j pair, the element [I][J] is a list of strings  */
  78 /* ##                        defining properties/flags of this interaction. Examples     */
  79 /* ##                        include 'electrostatics'/'vdw' if that type of interaction  */
  80 /* ##                        should be evaluated, 'rsq'/'rinv'/'rinvsq' if those values  */
  81 /* ##                        are needed, and 'exactcutoff' or 'shift','switch' to        */
  82 /* ##                        decide if the force/potential should be modified. This way  */
  83 /* ##                        we only calculate values absolutely needed for each case.   */
  84
  85 /* ## Calculate the size and offset for (merged/interleaved) table data */
  86
  87 /*
  88  * Gromacs nonbonded kernel:   {KERNEL_NAME}
  89  * Electrostatics interaction: {KERNEL_ELEC}
  90  * VdW interaction:            {KERNEL_VDW}
  91  * Geometry:                   {GEOMETRY_I}-{GEOMETRY_J}
  92  * Calculate force/pot:        {KERNEL_VF}
  93  */
  94 void
  95 {KERNEL_NAME}
  96                     (t_nblist                    * gmx_restrict       nlist,
  97                      rvec                        * gmx_restrict          xx,
  98                      rvec                        * gmx_restrict          ff,
  99                      t_forcerec                  * gmx_restrict          fr,
 100                      t_mdatoms                   * gmx_restrict     mdatoms,
 101                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 102                      t_nrnb                      * gmx_restrict        nrnb)
 103 {
 104     /* ## Not all variables are used for all kernels, but any optimizing compiler fixes that, */
 105     /* ## so there is no point in going to extremes to exclude variables that are not needed. */
 106     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
 107      * just 0 for non-waters.
 108      * Suffixes A,B refer to j loop unrolling done with SSE double precision, e.g. for the two different
 109      * jnr indices corresponding to data put in the four positions in the SIMD register.
 110      */
 111     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
 112     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
 113     int              jnrA,jnrB;
 114     int              j_coord_offsetA,j_coord_offsetB;
 115     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 116     real             rcutoff_scalar;
 117     real             *shiftvec,*fshift,*x,*f;
 118     __m128d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
 119     /* #for I in PARTICLES_I */
 120     int              vdwioffset{I};
 121     __m128d          ix{I},iy{I},iz{I},fix{I},fiy{I},fiz{I},iq{I},isai{I};
 122     /* #endfor */
 123     /* #for J in PARTICLES_J */
 124     int              vdwjidx{J}A,vdwjidx{J}B;
 125     __m128d          jx{J},jy{J},jz{J},fjx{J},fjy{J},fjz{J},jq{J},isaj{J};
 126     /* #endfor */
 127     /* #for I,J in PAIRS_IJ */
 128     __m128d          dx{I}{J},dy{I}{J},dz{I}{J},rsq{I}{J},rinv{I}{J},rinvsq{I}{J},r{I}{J},qq{I}{J},c6_{I}{J},c12_{I}{J};
 129     /* #endfor */
 130     /* #if KERNEL_ELEC != 'None' */
 131     __m128d          velec,felec,velecsum,facel,crf,krf,krf2;
 132     real             *charge;
 133     /* #endif */
 134     /* #if 'GeneralizedBorn' in KERNEL_ELEC */
 135     __m128i          gbitab;
 136     __m128d          vgb,fgb,vgbsum,dvdasum,gbscale,gbtabscale,isaprod,gbqqfactor,gbinvepsdiff,dvdaj,gbeps,twogbeps,dvdatmp;
 137     __m128d          minushalf = _mm_set1_pd(-0.5);
 138     real             *invsqrta,*dvda,*gbtab;
 139     /* #endif */
 140     /* #if KERNEL_VDW != 'None' */
 141     int              nvdwtype;
 142     __m128d          rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 143     int              *vdwtype;
 144     real             *vdwparam;
 145     __m128d          one_sixth   = _mm_set1_pd(1.0/6.0);
 146     __m128d          one_twelfth = _mm_set1_pd(1.0/12.0);
 147     /* #endif */
 148     /* #if 'Table' in KERNEL_ELEC or 'GeneralizedBorn' in KERNEL_ELEC or 'Table' in KERNEL_VDW */
 149     __m128i          vfitab;
 150     __m128i          ifour       = _mm_set1_epi32(4);
 151     __m128d          rt,vfeps,vftabscale,Y,F,G,H,Heps,Fp,VV,FF,twovfeps;
 152     real             *vftab;
 153     /* #endif */
 154     /* #if 'LJEwald' in KERNEL_VDW */
 155     /* #for I,J in PAIRS_IJ */
 156     __m128d           c6grid_{I}{J};
 157     /* #endfor */
 158     real             *vdwgridparam;
 159     __m128d           ewclj,ewclj2,ewclj6,ewcljrsq,poly,exponent,f6A,f6B,sh_lj_ewald;
 160     __m128d           one_half  = _mm_set1_pd(0.5);
 161     __m128d           minus_one = _mm_set1_pd(-1.0);
 162     /* #endif */
 163     /* #if 'Ewald' in KERNEL_ELEC */
 164     __m128i          ewitab;
 165     __m128d          ewtabscale,eweps,twoeweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
 166     real             *ewtab;
 167     /* #endif */
 168     /* #if 'PotentialSwitch' in [KERNEL_MOD_ELEC,KERNEL_MOD_VDW] */
 169     __m128d          rswitch,swV3,swV4,swV5,swF2,swF3,swF4,d,d2,sw,dsw;
 170     real             rswitch_scalar,d_scalar;
 171     /* #endif */
 172     __m128d          dummy_mask,cutoff_mask;
 173     __m128d          signbit   = gmx_mm_castsi128_pd( _mm_set_epi32(0x80000000,0x00000000,0x80000000,0x00000000) );
 174     __m128d          one     = _mm_set1_pd(1.0);
 175     __m128d          two     = _mm_set1_pd(2.0);
 176     x                = xx[0];
 177     f                = ff[0];
 178
 179     nri              = nlist->nri;
 180     iinr             = nlist->iinr;
 181     jindex           = nlist->jindex;
 182     jjnr             = nlist->jjnr;
 183     shiftidx         = nlist->shift;
 184     gid              = nlist->gid;
 185     shiftvec         = fr->shift_vec[0];
 186     fshift           = fr->fshift[0];
 187     /* #if KERNEL_ELEC != 'None' */
 188     facel            = _mm_set1_pd(fr->epsfac);
 189     charge           = mdatoms->chargeA;
 190     /*     #if 'ReactionField' in KERNEL_ELEC */
 191     krf              = _mm_set1_pd(fr->ic->k_rf);
 192     krf2             = _mm_set1_pd(fr->ic->k_rf*2.0);
 193     crf              = _mm_set1_pd(fr->ic->c_rf);
 194     /*     #endif */
 195     /* #endif */
 196     /* #if KERNEL_VDW != 'None' */
 197     nvdwtype         = fr->ntype;
 198     vdwparam         = fr->nbfp;
 199     vdwtype          = mdatoms->typeA;
 200     /* #endif */
 201     /* #if 'LJEwald' in KERNEL_VDW */
 202     vdwgridparam     = fr->ljpme_c6grid;
 203     sh_lj_ewald      = _mm_set1_pd(fr->ic->sh_lj_ewald);
 204     ewclj            = _mm_set1_pd(fr->ewaldcoeff_lj);
 205     ewclj2           = _mm_mul_pd(minus_one,_mm_mul_pd(ewclj,ewclj));
 206     /* #endif */
 207
 208     /* #if 'Table' in KERNEL_ELEC and 'Table' in KERNEL_VDW */
 209     vftab            = kernel_data->table_elec_vdw->data;
 210     vftabscale       = _mm_set1_pd(kernel_data->table_elec_vdw->scale);
 211     /* #elif 'Table' in KERNEL_ELEC */
 212     vftab            = kernel_data->table_elec->data;
 213     vftabscale       = _mm_set1_pd(kernel_data->table_elec->scale);
 214     /* #elif 'Table' in KERNEL_VDW */
 215     vftab            = kernel_data->table_vdw->data;
 216     vftabscale       = _mm_set1_pd(kernel_data->table_vdw->scale);
 217     /* #endif */
 218
 219     /* #if 'Ewald' in KERNEL_ELEC */
 220     sh_ewald         = _mm_set1_pd(fr->ic->sh_ewald);
 221     /*     #if KERNEL_VF=='Force' and KERNEL_MOD_ELEC!='PotentialSwitch' */
 222     ewtab            = fr->ic->tabq_coul_F;
 223     ewtabscale       = _mm_set1_pd(fr->ic->tabq_scale);
 224     ewtabhalfspace   = _mm_set1_pd(0.5/fr->ic->tabq_scale);
 225     /*     #else */
 226     ewtab            = fr->ic->tabq_coul_FDV0;
 227     ewtabscale       = _mm_set1_pd(fr->ic->tabq_scale);
 228     ewtabhalfspace   = _mm_set1_pd(0.5/fr->ic->tabq_scale);
 229      /*     #endif */
 230     /* #endif */
 231
 232     /* #if KERNEL_ELEC=='GeneralizedBorn' */
 233     invsqrta         = fr->invsqrta;
 234     dvda             = fr->dvda;
 235     gbtabscale       = _mm_set1_pd(fr->gbtab.scale);
 236     gbtab            = fr->gbtab.data;
 237     gbinvepsdiff     = _mm_set1_pd((1.0/fr->epsilon_r) - (1.0/fr->gb_epsilon_solvent));
 238     /* #endif */
 239
 240     /* #if 'Water' in GEOMETRY_I */
 241     /* Setup water-specific parameters */
 242     inr              = nlist->iinr[0];
 243     /*     #for I in PARTICLES_ELEC_I */
 244     iq{I}              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+{I}]));
 245     /*     #endfor */
 246     /*     #for I in PARTICLES_VDW_I */
 247     vdwioffset{I}      = 2*nvdwtype*vdwtype[inr+{I}];
 248     /*     #endfor */
 249     /* #endif */
 250
 251     /* #if 'Water' in GEOMETRY_J */
 252     /*     #for J in PARTICLES_ELEC_J */
 253     jq{J}              = _mm_set1_pd(charge[inr+{J}]);
 254     /*     #endfor */
 255     /*     #for J in PARTICLES_VDW_J */
 256     vdwjidx{J}A        = 2*vdwtype[inr+{J}];
 257     /*     #endfor */
 258     /*     #for I,J in PAIRS_IJ */
 259     /*         #if 'electrostatics' in INTERACTION_FLAGS[I][J] */
 260     qq{I}{J}             = _mm_mul_pd(iq{I},jq{J});
 261     /*         #endif */
 262     /*         #if 'vdw' in INTERACTION_FLAGS[I][J] */
 263     /*             #if 'LJEwald' in KERNEL_VDW */
 264     c6_{I}{J}            = _mm_set1_pd(vdwparam[vdwioffset{I}+vdwjidx{J}A]);
 265     c12_{I}{J}           = _mm_set1_pd(vdwparam[vdwioffset{I}+vdwjidx{J}A+1]);
 266     c6grid_{I}{J}        = _mm_set1_pd(vdwgridparam[vdwioffset{I}+vdwjidx{J}A]);
 267     /*             #else */
 268     c6_{I}{J}            = _mm_set1_pd(vdwparam[vdwioffset{I}+vdwjidx{J}A]);
 269     c12_{I}{J}           = _mm_set1_pd(vdwparam[vdwioffset{I}+vdwjidx{J}A+1]);
 270     /*             #endif */
 271     /*         #endif */
 272     /*     #endfor */
 273     /* #endif */
 274
 275     /* #if KERNEL_MOD_ELEC!='None' or KERNEL_MOD_VDW!='None' */
 276     /*     #if KERNEL_ELEC!='None' */
 277     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
 278     rcutoff_scalar   = fr->rcoulomb;
 279     /*     #else */
 280     rcutoff_scalar   = fr->rvdw;
 281     /*     #endif */
 282     rcutoff          = _mm_set1_pd(rcutoff_scalar);
 283     rcutoff2         = _mm_mul_pd(rcutoff,rcutoff);
 284     /* #endif */
 285
 286     /* #if KERNEL_MOD_VDW=='PotentialShift' */
 287     sh_vdw_invrcut6  = _mm_set1_pd(fr->ic->sh_invrc6);
 288     rvdw             = _mm_set1_pd(fr->rvdw);
 289     /* #endif */
 290
 291     /* #if 'PotentialSwitch' in [KERNEL_MOD_ELEC,KERNEL_MOD_VDW] */
 292     /*     #if KERNEL_MOD_ELEC=='PotentialSwitch'  */
 293     rswitch_scalar   = fr->rcoulomb_switch;
 294     rswitch          = _mm_set1_pd(rswitch_scalar);
 295     /*     #else */
 296     rswitch_scalar   = fr->rvdw_switch;
 297     rswitch          = _mm_set1_pd(rswitch_scalar);
 298     /*     #endif */
 299     /* Setup switch parameters */
 300     d_scalar         = rcutoff_scalar-rswitch_scalar;
 301     d                = _mm_set1_pd(d_scalar);
 302     swV3             = _mm_set1_pd(-10.0/(d_scalar*d_scalar*d_scalar));
 303     swV4             = _mm_set1_pd( 15.0/(d_scalar*d_scalar*d_scalar*d_scalar));
 304     swV5             = _mm_set1_pd( -6.0/(d_scalar*d_scalar*d_scalar*d_scalar*d_scalar));
 305     /*     #if 'Force' in KERNEL_VF */
 306     swF2             = _mm_set1_pd(-30.0/(d_scalar*d_scalar*d_scalar));
 307     swF3             = _mm_set1_pd( 60.0/(d_scalar*d_scalar*d_scalar*d_scalar));
 308     swF4             = _mm_set1_pd(-30.0/(d_scalar*d_scalar*d_scalar*d_scalar*d_scalar));
 309     /*     #endif */
 310     /* #endif */
 311
 312     /* Avoid stupid compiler warnings */
 313     jnrA = jnrB = 0;
 314     j_coord_offsetA = 0;
 315     j_coord_offsetB = 0;
 316
 317     /* ## Keep track of the floating point operations we issue for reporting! */
 318     /* #define OUTERFLOPS 0 */
 319     outeriter        = 0;
 320     inneriter        = 0;
 321
 322     /* Start outer loop over neighborlists */
 323     for(iidx=0; iidx<nri; iidx++)
 324     {
 325         /* Load shift vector for this list */
 326         i_shift_offset   = DIM*shiftidx[iidx];
 327
 328         /* Load limits for loop over neighbors */
 329         j_index_start    = jindex[iidx];
 330         j_index_end      = jindex[iidx+1];
 331
 332         /* Get outer coordinate index */
 333         inr              = iinr[iidx];
 334         i_coord_offset   = DIM*inr;
 335
 336         /* Load i particle coords and add shift vector */
 337         /* #if GEOMETRY_I == 'Particle' */
 338         gmx_mm_load_shift_and_1rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,&ix0,&iy0,&iz0);
 339         /* #elif GEOMETRY_I == 'Water3' */
 340         gmx_mm_load_shift_and_3rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
 341                                                  &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2);
 342         /* #elif GEOMETRY_I == 'Water4' */
 343         /*     #if 0 in PARTICLES_I                 */
 344         gmx_mm_load_shift_and_4rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
 345                                                  &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
 346         /*     #else                                */
 347         gmx_mm_load_shift_and_3rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset+DIM,
 348                                                  &ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
 349         /*     #endif                               */
 350         /* #endif                                   */
 351
 352         /* #if 'Force' in KERNEL_VF */
 353         /*     #for I in PARTICLES_I */
 354         fix{I}             = _mm_setzero_pd();
 355         fiy{I}             = _mm_setzero_pd();
 356         fiz{I}             = _mm_setzero_pd();
 357         /*     #endfor */
 358         /* #endif */
 359
 360         /* ## For water we already preloaded parameters at the start of the kernel */
 361         /* #if not 'Water' in GEOMETRY_I */
 362         /* Load parameters for i particles */
 363         /*     #for I in PARTICLES_ELEC_I */
 364         iq{I}              = _mm_mul_pd(facel,_mm_load1_pd(charge+inr+{I}));
 365         /*         #define OUTERFLOPS OUTERFLOPS+1 */
 366         /*         #if KERNEL_ELEC=='GeneralizedBorn' */
 367         isai{I}            = _mm_load1_pd(invsqrta+inr+{I});
 368         /*         #endif */
 369         /*     #endfor */
 370         /*     #for I in PARTICLES_VDW_I */
 371         vdwioffset{I}      = 2*nvdwtype*vdwtype[inr+{I}];
 372         /*     #endfor */
 373         /* #endif */
 374
 375         /* #if 'Potential' in KERNEL_VF */
 376         /* Reset potential sums */
 377         /*     #if KERNEL_ELEC != 'None' */
 378         velecsum         = _mm_setzero_pd();
 379         /*     #endif */
 380         /*     #if 'GeneralizedBorn' in KERNEL_ELEC */
 381         vgbsum           = _mm_setzero_pd();
 382         /*     #endif */
 383         /*     #if KERNEL_VDW != 'None' */
 384         vvdwsum          = _mm_setzero_pd();
 385         /*     #endif */
 386         /* #endif */
 387         /*     #if 'GeneralizedBorn' in KERNEL_ELEC and 'Force' in KERNEL_VF */
 388         dvdasum          = _mm_setzero_pd();
 389         /*     #endif */
 390
 391         /* #for ROUND in ['Loop','Epilogue'] */
 392
 393         /* #if ROUND =='Loop' */
 394         /* Start inner kernel loop */
 395         for(jidx=j_index_start; jidx<j_index_end-1; jidx+=2)
 396         {
 397         /* ## First round is normal loop (next statement resets indentation) */
 398         /*     #if 0 */
 399         }
 400         /*     #endif */
 401         /* #else */
 402         if(jidx<j_index_end)
 403         {
 404         /* ## Second round is epilogue */
 405         /* #endif */
 406         /* #define INNERFLOPS 0 */
 407
 408             /* #if ROUND =='Loop' */
 409             /* Get j neighbor index, and coordinate index */
 410             jnrA             = jjnr[jidx];
 411             jnrB             = jjnr[jidx+1];
 412             j_coord_offsetA  = DIM*jnrA;
 413             j_coord_offsetB  = DIM*jnrB;
 414
 415             /* load j atom coordinates */
 416             /*     #if GEOMETRY_J == 'Particle'             */
 417             gmx_mm_load_1rvec_2ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 418                                               &jx0,&jy0,&jz0);
 419             /*     #elif GEOMETRY_J == 'Water3'             */
 420             gmx_mm_load_3rvec_2ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 421                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,&jy2,&jz2);
 422             /*     #elif GEOMETRY_J == 'Water4'             */
 423             /*         #if 0 in PARTICLES_J                 */
 424             gmx_mm_load_4rvec_2ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 425                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
 426                                               &jy2,&jz2,&jx3,&jy3,&jz3);
 427             /*         #else                                */
 428             gmx_mm_load_3rvec_2ptr_swizzle_pd(x+j_coord_offsetA+DIM,x+j_coord_offsetB+DIM,
 429                                               &jx1,&jy1,&jz1,&jx2,&jy2,&jz2,&jx3,&jy3,&jz3);
 430             /*         #endif                               */
 431             /*     #endif                                   */
 432             /* #else */
 433             jnrA             = jjnr[jidx];
 434             j_coord_offsetA  = DIM*jnrA;
 435
 436             /* load j atom coordinates */
 437             /*     #if GEOMETRY_J == 'Particle'             */
 438             gmx_mm_load_1rvec_1ptr_swizzle_pd(x+j_coord_offsetA,
 439                                               &jx0,&jy0,&jz0);
 440             /*     #elif GEOMETRY_J == 'Water3'             */
 441             gmx_mm_load_3rvec_1ptr_swizzle_pd(x+j_coord_offsetA,
 442                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,&jy2,&jz2);
 443             /*     #elif GEOMETRY_J == 'Water4'             */
 444             /*         #if 0 in PARTICLES_J                 */
 445             gmx_mm_load_4rvec_1ptr_swizzle_pd(x+j_coord_offsetA,
 446                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
 447                                               &jy2,&jz2,&jx3,&jy3,&jz3);
 448             /*         #else                                */
 449             gmx_mm_load_3rvec_1ptr_swizzle_pd(x+j_coord_offsetA+DIM,
 450                                               &jx1,&jy1,&jz1,&jx2,&jy2,&jz2,&jx3,&jy3,&jz3);
 451             /*         #endif                               */
 452             /*     #endif                                   */
 453             /* #endif */
 454
 455             /* Calculate displacement vector */
 456             /* #for I,J in PAIRS_IJ */
 457             dx{I}{J}             = _mm_sub_pd(ix{I},jx{J});
 458             dy{I}{J}             = _mm_sub_pd(iy{I},jy{J});
 459             dz{I}{J}             = _mm_sub_pd(iz{I},jz{J});
 460             /*     #define INNERFLOPS INNERFLOPS+3 */
 461             /* #endfor */
 462
 463             /* Calculate squared distance and things based on it */
 464             /* #for I,J in PAIRS_IJ */
 465             rsq{I}{J}            = gmx_mm_calc_rsq_pd(dx{I}{J},dy{I}{J},dz{I}{J});
 466             /*     #define INNERFLOPS INNERFLOPS+5 */
 467             /* #endfor */
 468
 469             /* #for I,J in PAIRS_IJ */
 470             /*     #if 'rinv' in INTERACTION_FLAGS[I][J] */
 471             rinv{I}{J}           = gmx_mm_invsqrt_pd(rsq{I}{J});
 472             /*         #define INNERFLOPS INNERFLOPS+5 */
 473             /*     #endif */
 474             /* #endfor */
 475
 476             /* #for I,J in PAIRS_IJ */
 477             /*     #if 'rinvsq' in INTERACTION_FLAGS[I][J] */
 478             /*         # if 'rinv' not in INTERACTION_FLAGS[I][J] */
 479             rinvsq{I}{J}         = gmx_mm_inv_pd(rsq{I}{J});
 480             /*             #define INNERFLOPS INNERFLOPS+4 */
 481             /*         #else */
 482             rinvsq{I}{J}         = _mm_mul_pd(rinv{I}{J},rinv{I}{J});
 483             /*             #define INNERFLOPS INNERFLOPS+1 */
 484             /*         #endif */
 485             /*     #endif */
 486             /* #endfor */
 487
 488             /* #if not 'Water' in GEOMETRY_J */
 489             /* Load parameters for j particles */
 490             /*     #for J in PARTICLES_ELEC_J */
 491             /*         #if ROUND =='Loop' */
 492             jq{J}              = gmx_mm_load_2real_swizzle_pd(charge+jnrA+{J},charge+jnrB+{J});
 493             /*         #else */
 494             jq{J}              = _mm_load_sd(charge+jnrA+{J});
 495             /*         #endif */
 496             /*         #if KERNEL_ELEC=='GeneralizedBorn' */
 497             /*             #if ROUND =='Loop' */
 498             isaj{J}            = gmx_mm_load_2real_swizzle_pd(invsqrta+jnrA+{J},invsqrta+jnrB+{J});
 499             /*             #else */
 500             isaj{J}            = _mm_load_sd(invsqrta+jnrA+{J});
 501             /*             #endif */
 502             /*         #endif */
 503             /*     #endfor */
 504             /*     #for J in PARTICLES_VDW_J */
 505             vdwjidx{J}A        = 2*vdwtype[jnrA+{J}];
 506             /*         #if ROUND =='Loop' */
 507             vdwjidx{J}B        = 2*vdwtype[jnrB+{J}];
 508             /*         #endif */
 509             /*     #endfor */
 510             /* #endif */
 511
 512             /* #if 'Force' in KERNEL_VF and not 'Particle' in GEOMETRY_I */
 513             /*     #for J in PARTICLES_J */
 514             fjx{J}             = _mm_setzero_pd();
 515             fjy{J}             = _mm_setzero_pd();
 516             fjz{J}             = _mm_setzero_pd();
 517             /*     #endfor */
 518             /* #endif */
 519
 520             /* #for I,J in PAIRS_IJ */
 521
 522             /**************************
 523              * CALCULATE INTERACTIONS *
 524              **************************/
 525
 526             /*     #if 'exactcutoff' in INTERACTION_FLAGS[I][J] */
 527             /*         ## We always calculate rinv/rinvsq above to enable pipelineing in compilers (performance tested on x86) */
 528             if (gmx_mm_any_lt(rsq{I}{J},rcutoff2))
 529             {
 530                 /*     #if 0    ## this and the next two lines is a hack to maintain auto-indentation in template file */
 531             }
 532             /*         #endif */
 533             /*         #define INNERFLOPS INNERFLOPS+1 */
 534             /*     #endif */
 535
 536             /*     #if 'r' in INTERACTION_FLAGS[I][J] */
 537             r{I}{J}              = _mm_mul_pd(rsq{I}{J},rinv{I}{J});
 538              /*         #define INNERFLOPS INNERFLOPS+1 */
 539             /*     #endif */
 540
 541             /*     ## For water geometries we already loaded parameters at the start of the kernel */
 542             /*     #if not 'Water' in GEOMETRY_J */
 543             /* Compute parameters for interactions between i and j atoms */
 544             /*         #if 'electrostatics' in INTERACTION_FLAGS[I][J] */
 545             qq{I}{J}             = _mm_mul_pd(iq{I},jq{J});
 546             /*             #define INNERFLOPS INNERFLOPS+1 */
 547             /*         #endif */
 548             /*         #if 'vdw' in INTERACTION_FLAGS[I][J] */
 549             /*             #if ROUND == 'Loop' */
 550             gmx_mm_load_2pair_swizzle_pd(vdwparam+vdwioffset{I}+vdwjidx{J}A,
 551                                          vdwparam+vdwioffset{I}+vdwjidx{J}B,&c6_{I}{J},&c12_{I}{J});
 552             /*                 #if 'LJEwald' in KERNEL_VDW */
 553             c6grid_{I}{J}       = gmx_mm_load_2real_swizzle_pd(vdwgridparam+vdwioffset{I}+vdwjidx{J}A,
 554                                                                vdwgridparam+vdwioffset{I}+vdwjidx{J}B);
 555             /*                 #endif */
 556             /*             #else */
 557             gmx_mm_load_1pair_swizzle_pd(vdwparam+vdwioffset{I}+vdwjidx{J}A,&c6_{I}{J},&c12_{I}{J});
 558             /*                 #if 'LJEwald' in KERNEL_VDW */
 559             c6grid_{I}{J}       = gmx_mm_load_1real_pd(vdwgridparam+vdwioffset{I}+vdwjidx{J}A);
 560             /*                 #endif */
 561             /*             #endif */
 562             /*         #endif */
 563             /*     #endif */
 564
 565             /*     #if 'table' in INTERACTION_FLAGS[I][J] */
 566             /* Calculate table index by multiplying r with table scale and truncate to integer */
 567             rt               = _mm_mul_pd(r{I}{J},vftabscale);
 568             vfitab           = _mm_cvttpd_epi32(rt);
 569 #ifdef __XOP__
 570             vfeps            = _mm_frcz_pd(rt);
 571 #else
 572             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
 573 #endif
 574             twovfeps         = _mm_add_pd(vfeps,vfeps);
 575             /*         #define INNERFLOPS INNERFLOPS+4                          */
 576             /*         #if 'Table' in KERNEL_ELEC and 'Table' in KERNEL_VDW     */
 577             /*             ## 3 tables, 4 data per point: multiply index by 12 */
 578             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
 579             /*         #elif 'Table' in KERNEL_ELEC                             */
 580             /*             ## 1 table, 4 data per point: multiply index by 4   */
 581             vfitab           = _mm_slli_epi32(vfitab,2);
 582             /*         #elif 'Table' in KERNEL_VDW                              */
 583             /*             ## 2 tables, 4 data per point: multiply index by 8  */
 584             vfitab           = _mm_slli_epi32(vfitab,3);
 585             /*         #endif                                                   */
 586             /*     #endif */
 587
 588             /*     ## ELECTROSTATIC INTERACTIONS */
 589             /*     #if 'electrostatics' in INTERACTION_FLAGS[I][J] */
 590
 591             /*         #if KERNEL_ELEC=='Coulomb' */
 592
 593             /* COULOMB ELECTROSTATICS */
 594             velec            = _mm_mul_pd(qq{I}{J},rinv{I}{J});
 595             /*             #define INNERFLOPS INNERFLOPS+1 */
 596             /*             #if 'Force' in KERNEL_VF */
 597             felec            = _mm_mul_pd(velec,rinvsq{I}{J});
 598             /*                 #define INNERFLOPS INNERFLOPS+2 */
 599             /*             #endif */
 600
 601             /*         #elif KERNEL_ELEC=='ReactionField' */
 602
 603             /* REACTION-FIELD ELECTROSTATICS */
 604             /*             #if 'Potential' in KERNEL_VF */
 605             velec            = _mm_mul_pd(qq{I}{J},_mm_sub_pd(_mm_macc_pd(krf,rsq{I}{J},rinv{I}{J}),crf));
 606             /*                 #define INNERFLOPS INNERFLOPS+4 */
 607             /*             #endif */
 608             /*             #if 'Force' in KERNEL_VF */
 609             felec            = _mm_mul_pd(qq{I}{J},_mm_msub_pd(rinv{I}{J},rinvsq{I}{J},krf2));
 610             /*                 #define INNERFLOPS INNERFLOPS+3 */
 611             /*             #endif */
 612
 613             /*         #elif KERNEL_ELEC=='GeneralizedBorn' */
 614
 615             /* GENERALIZED BORN AND COULOMB ELECTROSTATICS */
 616             isaprod          = _mm_mul_pd(isai{I},isaj{J});
 617             gbqqfactor       = _mm_xor_pd(signbit,_mm_mul_pd(qq{I}{J},_mm_mul_pd(isaprod,gbinvepsdiff)));
 618             gbscale          = _mm_mul_pd(isaprod,gbtabscale);
 619             /*             #define INNERFLOPS INNERFLOPS+5 */
 620
 621             /* Calculate generalized born table index - this is a separate table from the normal one,
 622              * but we use the same procedure by multiplying r with scale and truncating to integer.
 623              */
 624             rt               = _mm_mul_pd(r{I}{J},gbscale);
 625             gbitab           = _mm_cvttpd_epi32(rt);
 626 #ifdef __XOP__
 627             gbeps            = _mm_frcz_pd(rt);
 628 #else
 629             gbeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
 630 #endif
 631             gbitab           = _mm_slli_epi32(gbitab,2);
 632
 633             Y                = _mm_load_pd( gbtab + _mm_extract_epi32(gbitab,0) );
 634             /*             #if ROUND == 'Loop' */
 635             F                = _mm_load_pd( gbtab + _mm_extract_epi32(gbitab,1) );
 636             /*             #else */
 637             F                = _mm_setzero_pd();
 638             /*             #endif */
 639             GMX_MM_TRANSPOSE2_PD(Y,F);
 640             G                = _mm_load_pd( gbtab + _mm_extract_epi32(gbitab,0) +2);
 641             /*             #if ROUND == 'Loop' */
 642             H                = _mm_load_pd( gbtab + _mm_extract_epi32(gbitab,1) +2);
 643             /*             #else */
 644             H                = _mm_setzero_pd();
 645             /*             #endif */
 646             GMX_MM_TRANSPOSE2_PD(G,H);
 647             Fp               = _mm_macc_pd(gbeps,_mm_macc_pd(gbeps,H,G),F);
 648             VV               = _mm_macc_pd(gbeps,Fp,Y);
 649             vgb              = _mm_mul_pd(gbqqfactor,VV);
 650             /*             #define INNERFLOPS INNERFLOPS+10 */
 651
 652             /*             #if 'Force' in KERNEL_VF */
 653             twogbeps         = _mm_add_pd(gbeps,gbeps);
 654             FF               = _mm_macc_pd(_mm_macc_pd(twogbeps,H,G),gbeps,Fp);
 655             fgb              = _mm_mul_pd(gbqqfactor,_mm_mul_pd(FF,gbscale));
 656             dvdatmp          = _mm_mul_pd(minushalf,_mm_macc_pd(fgb,r{I}{J},vgb));
 657             /*                 #if ROUND == 'Epilogue' */
 658             dvdatmp          = _mm_unpacklo_pd(dvdatmp,_mm_setzero_pd());
 659             /*                 #endif */
 660             dvdasum          = _mm_add_pd(dvdasum,dvdatmp);
 661             /*             #if ROUND == 'Loop' */
 662             gmx_mm_increment_2real_swizzle_pd(dvda+jnrA,dvda+jnrB,_mm_mul_pd(dvdatmp,_mm_mul_pd(isaj{J},isaj{J})));
 663             /*             #else */
 664             gmx_mm_increment_1real_pd(dvda+jnrA,_mm_mul_pd(dvdatmp,_mm_mul_pd(isaj{J},isaj{J})));
 665             /*             #endif */
 666             /*                 #define INNERFLOPS INNERFLOPS+13 */
 667             /*             #endif */
 668             velec            = _mm_mul_pd(qq{I}{J},rinv{I}{J});
 669             /*                 #define INNERFLOPS INNERFLOPS+1 */
 670             /*             #if 'Force' in KERNEL_VF */
 671             felec            = _mm_mul_pd(_mm_msub_pd(velec,rinv{I}{J},fgb),rinv{I}{J});
 672             /*                 #define INNERFLOPS INNERFLOPS+3 */
 673             /*             #endif */
 674
 675             /*         #elif KERNEL_ELEC=='Ewald' */
 676             /* EWALD ELECTROSTATICS */
 677
 678             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 679             ewrt             = _mm_mul_pd(r{I}{J},ewtabscale);
 680             ewitab           = _mm_cvttpd_epi32(ewrt);
 681 #ifdef __XOP__
 682             eweps            = _mm_frcz_pd(ewrt);
 683 #else
 684             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 685 #endif
 686             twoeweps         = _mm_add_pd(eweps,eweps);
 687             /*             #define INNERFLOPS INNERFLOPS+4 */
 688             /*             #if 'Potential' in KERNEL_VF or KERNEL_MOD_ELEC=='PotentialSwitch' */
 689             ewitab           = _mm_slli_epi32(ewitab,2);
 690             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 691             /*                 #if ROUND == 'Loop' */
 692             ewtabD           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 693             /*                 #else */
 694             ewtabD           = _mm_setzero_pd();
 695             /*                 #endif */
 696             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 697             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 698             /*                 #if ROUND == 'Loop' */
 699             ewtabFn          = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,1) +2);
 700             /*                 #else */
 701             ewtabFn          = _mm_setzero_pd();
 702             /*                 #endif */
 703             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 704             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 705             /*                 #define INNERFLOPS INNERFLOPS+2 */
 706             /*                 #if KERNEL_MOD_ELEC=='PotentialShift' */
 707             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 708             velec            = _mm_mul_pd(qq{I}{J},_mm_sub_pd(_mm_sub_pd(rinv{I}{J},sh_ewald),velec));
 709             /*                     #define INNERFLOPS INNERFLOPS+7 */
 710             /*                 #else */
 711             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 712             velec            = _mm_mul_pd(qq{I}{J},_mm_sub_pd(rinv{I}{J},velec));
 713             /*                     #define INNERFLOPS INNERFLOPS+6 */
 714             /*                 #endif */
 715             /*                 #if 'Force' in KERNEL_VF */
 716             felec            = _mm_mul_pd(_mm_mul_pd(qq{I}{J},rinv{I}{J}),_mm_sub_pd(rinvsq{I}{J},felec));
 717             /*                      #define INNERFLOPS INNERFLOPS+3 */
 718             /*                 #endif */
 719             /*             #elif KERNEL_VF=='Force' */
 720             /*                 #if ROUND == 'Loop' */
 721             gmx_mm_load_2pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),ewtab+_mm_extract_epi32(ewitab,1),
 722                                          &ewtabF,&ewtabFn);
 723             /*                 #else */
 724             gmx_mm_load_1pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),&ewtabF,&ewtabFn);
 725             /*                 #endif */
 726             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
 727             felec            = _mm_mul_pd(_mm_mul_pd(qq{I}{J},rinv{I}{J}),_mm_sub_pd(rinvsq{I}{J},felec));
 728             /*                 #define INNERFLOPS INNERFLOPS+7 */
 729             /*             #endif */
 730
 731             /*         #elif KERNEL_ELEC=='CubicSplineTable' */
 732
 733             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 734             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 735             /*             #if ROUND == 'Loop' */
 736             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 737             /*             #else */
 738             F                = _mm_setzero_pd();
 739             /*             #endif */
 740             GMX_MM_TRANSPOSE2_PD(Y,F);
 741             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 742             /*             #if ROUND == 'Loop' */
 743             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
 744             /*             #else */
 745             H                = _mm_setzero_pd();
 746             /*             #endif */
 747             GMX_MM_TRANSPOSE2_PD(G,H);
 748             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(vfeps,H,G),F);
 749             /*             #define INNERFLOPS INNERFLOPS+4 */
 750             /*             #if 'Potential' in KERNEL_VF */
 751             VV               = _mm_macc_pd(vfeps,Fp,Y);
 752             velec            = _mm_mul_pd(qq{I}{J},VV);
 753             /*                 #define INNERFLOPS INNERFLOPS+3 */
 754             /*             #endif */
 755             /*             #if 'Force' in KERNEL_VF */
 756             FF               = _mm_macc_pd(_mm_macc_pd(twovfeps,H,G),vfeps,Fp);
 757             felec            = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq{I}{J},FF),_mm_mul_pd(vftabscale,rinv{I}{J})));
 758             /*                 #define INNERFLOPS INNERFLOPS+7 */
 759             /*             #endif */
 760             /*         #endif */
 761             /*         ## End of check for electrostatics interaction forms */
 762             /*     #endif */
 763             /*     ## END OF ELECTROSTATIC INTERACTION CHECK FOR PAIR I-J */
 764
 765             /*     #if 'vdw' in INTERACTION_FLAGS[I][J] */
 766
 767             /*         #if KERNEL_VDW=='LennardJones' */
 768
 769             /* LENNARD-JONES DISPERSION/REPULSION */
 770
 771             rinvsix          = _mm_mul_pd(_mm_mul_pd(rinvsq{I}{J},rinvsq{I}{J}),rinvsq{I}{J});
 772             /*             #define INNERFLOPS INNERFLOPS+2 */
 773             /*             #if 'Potential' in KERNEL_VF or KERNEL_MOD_VDW=='PotentialSwitch' */
 774             vvdw6            = _mm_mul_pd(c6_{I}{J},rinvsix);
 775             vvdw12           = _mm_mul_pd(c12_{I}{J},_mm_mul_pd(rinvsix,rinvsix));
 776             /*                 #define INNERFLOPS INNERFLOPS+3 */
 777             /*                 #if KERNEL_MOD_VDW=='PotentialShift' */
 778             vvdw             = _mm_msub_pd(_mm_nmacc_pd(c12_{I}{J},_mm_mul_pd(sh_vdw_invrcut6,sh_vdw_invrcut6),vvdw12),one_twelfth,
 779                                            _mm_mul_pd(_mm_nmacc_pd( c6_{I}{J},sh_vdw_invrcut6,vvdw6),one_sixth));
 780             /*                     #define INNERFLOPS INNERFLOPS+8 */
 781             /*                 #else */
 782             vvdw             = _mm_msub_pd( vvdw12,one_twelfth, _mm_mul_pd(vvdw6,one_sixth) );
 783             /*                     #define INNERFLOPS INNERFLOPS+3 */
 784             /*                 #endif */
 785             /*                 ## Check for force inside potential check, i.e. this means we already did the potential part */
 786             /*                 #if 'Force' in KERNEL_VF */
 787             fvdw             = _mm_mul_pd(_mm_sub_pd(vvdw12,vvdw6),rinvsq{I}{J});
 788             /*                     #define INNERFLOPS INNERFLOPS+2 */
 789             /*                 #endif */
 790             /*             #elif KERNEL_VF=='Force' */
 791             /*                 ## Force-only LennardJones makes it possible to save 1 flop (they do add up...) */
 792             fvdw             = _mm_mul_pd(_mm_msub_pd(c12_{I}{J},rinvsix,c6_{I}{J}),_mm_mul_pd(rinvsix,rinvsq{I}{J}));
 793             /*                 #define INNERFLOPS INNERFLOPS+4 */
 794             /*             #endif */
 795
 796             /*         #elif KERNEL_VDW=='CubicSplineTable' */
 797
 798             /* CUBIC SPLINE TABLE DISPERSION */
 799             /*             #if 'Table' in KERNEL_ELEC */
 800             vfitab           = _mm_add_epi32(vfitab,ifour);
 801             /*             #endif                     */
 802             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 803             /*             #if ROUND == 'Loop' */
 804             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 805             /*             #else */
 806             F                = _mm_setzero_pd();
 807             /*             #endif */
 808             GMX_MM_TRANSPOSE2_PD(Y,F);
 809             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 810             /*             #if ROUND == 'Loop' */
 811             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
 812             /*             #else */
 813             H                = _mm_setzero_pd();
 814             /*             #endif */
 815             GMX_MM_TRANSPOSE2_PD(G,H);
 816             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(H,vfeps,G),F);
 817             /*             #define INNERFLOPS INNERFLOPS+4 */
 818             /*             #if 'Potential' in KERNEL_VF */
 819             VV               = _mm_macc_pd(vfeps,Fp,Y);
 820             vvdw6            = _mm_mul_pd(c6_{I}{J},VV);
 821             /*                 #define INNERFLOPS INNERFLOPS+3 */
 822             /*             #endif */
 823             /*             #if 'Force' in KERNEL_VF */
 824             FF               = _mm_macc_pd(vfeps,_mm_macc_pd(twovfeps,H,G),Fp);
 825             fvdw6            = _mm_mul_pd(c6_{I}{J},FF);
 826             /*                 #define INNERFLOPS INNERFLOPS+4 */
 827             /*             #endif */
 828
 829             /* CUBIC SPLINE TABLE REPULSION */
 830             vfitab           = _mm_add_epi32(vfitab,ifour);
 831             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 832             /*             #if ROUND == 'Loop' */
 833             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 834             /*             #else */
 835             F                = _mm_setzero_pd();
 836             /*             #endif */
 837             GMX_MM_TRANSPOSE2_PD(Y,F);
 838             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 839             /*             #if ROUND == 'Loop' */
 840             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
 841             /*             #else */
 842             H                = _mm_setzero_pd();
 843             /*             #endif */
 844             GMX_MM_TRANSPOSE2_PD(G,H);
 845             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(H,vfeps,G),F);
 846             /*             #define INNERFLOPS INNERFLOPS+4 */
 847             /*             #if 'Potential' in KERNEL_VF */
 848             VV               = _mm_macc_pd(vfeps,Fp,Y);
 849             vvdw12           = _mm_mul_pd(c12_{I}{J},VV);
 850             /*                 #define INNERFLOPS INNERFLOPS+3 */
 851             /*             #endif */
 852             /*             #if 'Force' in KERNEL_VF */
 853             FF               = _mm_macc_pd(vfeps,_mm_macc_pd(twovfeps,H,G),Fp);
 854             fvdw12           = _mm_mul_pd(c12_{I}{J},FF);
 855             /*                 #define INNERFLOPS INNERFLOPS+5 */
 856             /*             #endif */
 857             /*             #if 'Potential' in KERNEL_VF */
 858             vvdw             = _mm_add_pd(vvdw12,vvdw6);
 859             /*                 #define INNERFLOPS INNERFLOPS+1 */
 860             /*             #endif */
 861             /*             #if 'Force' in KERNEL_VF */
 862             fvdw             = _mm_xor_pd(signbit,_mm_mul_pd(_mm_add_pd(fvdw6,fvdw12),_mm_mul_pd(vftabscale,rinv{I}{J})));
 863             /*                 #define INNERFLOPS INNERFLOPS+4 */
 864             /*             #endif */
 865
 866             /*         #elif KERNEL_VDW=='LJEwald' */
 867
 868             /* Analytical LJ-PME */
 869             rinvsix          = _mm_mul_pd(_mm_mul_pd(rinvsq{I}{J},rinvsq{I}{J}),rinvsq{I}{J});
 870             ewcljrsq         = _mm_mul_pd(ewclj2,rsq{I}{J});
 871             ewclj6           = _mm_mul_pd(ewclj2,_mm_mul_pd(ewclj2,ewclj2));
 872             exponent         = gmx_simd_exp_d(ewcljrsq);
 873             /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
 874             poly             = _mm_mul_pd(exponent,_mm_macc_pd(_mm_mul_pd(ewcljrsq,ewcljrsq),one_half,_mm_sub_pd(one,ewcljrsq)));
 875             /*                 #define INNERFLOPS INNERFLOPS+10 */
 876             /*             #if 'Potential' in KERNEL_VF or KERNEL_MOD_VDW=='PotentialSwitch' */
 877             /* vvdw6 = [C6 - C6grid * (1-poly)]/r6 */
 878             vvdw6            = _mm_mul_pd(_mm_macc_pd(-c6grid_{I}{J},_mm_sub_pd(one,poly),c6_{I}{J}),rinvsix);
 879             vvdw12           = _mm_mul_pd(c12_{I}{J},_mm_mul_pd(rinvsix,rinvsix));
 880             /*                     #define INNERFLOPS INNERFLOPS+5 */
 881             /*                 #if KERNEL_MOD_VDW=='PotentialShift' */
 882             vvdw             = _mm_msub_pd(_mm_nmacc_pd(c12_{I}{J},_mm_mul_pd(sh_vdw_invrcut6,sh_vdw_invrcut6),vvdw12),one_twelfth,
 883                                _mm_mul_pd(_mm_sub_pd(vvdw6,_mm_macc_pd(c6grid_{I}{J},sh_lj_ewald,_mm_mul_pd(c6_{I}{J},sh_vdw_invrcut6))),one_sixth));
 884             /*                     #define INNERFLOPS INNERFLOPS+6 */
 885             /*                 #else */
 886             vvdw             = _mm_msub_pd(vvdw12,one_twelfth,_mm_mul_pd(vvdw6,one_sixth));
 887             /*                     #define INNERFLOPS INNERFLOPS+2 */
 888             /*                 #endif */
 889             /*                  ## Check for force inside potential check, i.e. this means we already did the potential part */
 890             /*                  #if 'Force' in KERNEL_VF */
 891             /* fvdw = vvdw12/r - (vvdw6/r + (C6grid * exponent * beta^6)/r) */
 892             fvdw             = _mm_mul_pd(_mm_add_pd(vvdw12,_mm_msub_pd(_mm_mul_pd(c6grid_{I}{J},one_sixth),_mm_mul_pd(exponent,ewclj6),vvdw6)),rinvsq{I}{J});
 893             /*                 #define INNERFLOPS INNERFLOPS+5 */
 894             /*                  #endif */
 895             /*              #elif KERNEL_VF=='Force' */
 896             /* f6A = 6 * C6grid * (1 - poly) */
 897             f6A              = _mm_mul_pd(c6grid_{I}{J},_mm_sub_pd(one,poly));
 898             /* f6B = C6grid * exponent * beta^6 */
 899             f6B              = _mm_mul_pd(_mm_mul_pd(c6grid_{I}{J},one_sixth),_mm_mul_pd(exponent,ewclj6));
 900             /* fvdw = 12*C12/r13 - ((6*C6 - f6A)/r6 + f6B)/r */
 901             fvdw              = _mm_mul_pd(_mm_macc_pd(_mm_msub_pd(c12_{I}{J},rinvsix,_mm_sub_pd(c6_{I}{J},f6A)),rinvsix,f6B),rinvsq{I}{J});
 902             /*                 #define INNERFLOPS INNERFLOPS+10 */
 903             /*              #endif */
 904             /*         #endif */
 905             /*         ## End of check for vdw interaction forms */
 906             /*     #endif */
 907             /*     ## END OF VDW INTERACTION CHECK FOR PAIR I-J */
 908
 909             /*     #if 'switch' in INTERACTION_FLAGS[I][J] */
 910             d                = _mm_sub_pd(r{I}{J},rswitch);
 911             d                = _mm_max_pd(d,_mm_setzero_pd());
 912             d2               = _mm_mul_pd(d,d);
 913             sw               = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_macc_pd(d,_mm_macc_pd(d,swV5,swV4),swV3))));
 914             /*         #define INNERFLOPS INNERFLOPS+10 */
 915
 916             /*         #if 'Force' in KERNEL_VF */
 917             dsw              = _mm_mul_pd(d2,_mm_macc_pd(d,_mm_macc_pd(d,swF4,swF3),swF2));
 918             /*             #define INNERFLOPS INNERFLOPS+5 */
 919             /*         #endif */
 920
 921             /* Evaluate switch function */
 922             /*         #if 'Force' in KERNEL_VF */
 923             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 924             /*             #if 'electrostatics' in INTERACTION_FLAGS[I][J] and KERNEL_MOD_ELEC=='PotentialSwitch' */
 925             felec            = _mm_msub_pd( felec,sw , _mm_mul_pd(rinv{I}{J},_mm_mul_pd(velec,dsw)) );
 926             /*                 #define INNERFLOPS INNERFLOPS+4 */
 927             /*             #endif */
 928             /*             #if 'vdw' in INTERACTION_FLAGS[I][J] and KERNEL_MOD_VDW=='PotentialSwitch' */
 929             fvdw             = _mm_msub_pd( fvdw,sw , _mm_mul_pd(rinv{I}{J},_mm_mul_pd(vvdw,dsw)) );
 930             /*                 #define INNERFLOPS INNERFLOPS+4 */
 931             /*             #endif */
 932             /*         #endif */
 933             /*         #if 'Potential' in KERNEL_VF */
 934             /*             #if 'electrostatics' in INTERACTION_FLAGS[I][J] and KERNEL_MOD_ELEC=='PotentialSwitch' */
 935             velec            = _mm_mul_pd(velec,sw);
 936             /*                 #define INNERFLOPS INNERFLOPS+1 */
 937             /*             #endif */
 938             /*             #if 'vdw' in INTERACTION_FLAGS[I][J] and KERNEL_MOD_VDW=='PotentialSwitch' */
 939             vvdw             = _mm_mul_pd(vvdw,sw);
 940             /*                 #define INNERFLOPS INNERFLOPS+1 */
 941             /*             #endif */
 942             /*         #endif */
 943             /*     #endif */
 944             /*     #if 'exactcutoff' in INTERACTION_FLAGS[I][J] */
 945             cutoff_mask      = _mm_cmplt_pd(rsq{I}{J},rcutoff2);
 946             /*         #define INNERFLOPS INNERFLOPS+1 */
 947             /*     #endif */
 948
 949             /*     #if 'Potential' in KERNEL_VF */
 950             /* Update potential sum for this i atom from the interaction with this j atom. */
 951             /*         #if 'electrostatics' in INTERACTION_FLAGS[I][J] */
 952             /*             #if 'exactcutoff' in INTERACTION_FLAGS[I][J] */
 953             velec            = _mm_and_pd(velec,cutoff_mask);
 954             /*                 #define INNERFLOPS INNERFLOPS+1 */
 955             /*             #endif                                       */
 956             /*             #if ROUND == 'Epilogue' */
 957             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
 958             /*             #endif */
 959             velecsum         = _mm_add_pd(velecsum,velec);
 960             /*             #define INNERFLOPS INNERFLOPS+1 */
 961             /*             #if KERNEL_ELEC=='GeneralizedBorn' */
 962             /*             #if 'exactcutoff' in INTERACTION_FLAGS[I][J] */
 963             vgb              = _mm_and_pd(vgb,cutoff_mask);
 964             /*                 #define INNERFLOPS INNERFLOPS+1 */
 965             /*             #endif                                       */
 966             /*             #if ROUND == 'Epilogue' */
 967             vgb              = _mm_unpacklo_pd(vgb,_mm_setzero_pd());
 968             /*             #endif */
 969             vgbsum           = _mm_add_pd(vgbsum,vgb);
 970             /*                 #define INNERFLOPS INNERFLOPS+1 */
 971             /*             #endif */
 972             /*         #endif */
 973             /*         #if 'vdw' in INTERACTION_FLAGS[I][J] */
 974             /*             #if 'exactcutoff' in INTERACTION_FLAGS[I][J] */
 975             vvdw             = _mm_and_pd(vvdw,cutoff_mask);
 976             /*                 #define INNERFLOPS INNERFLOPS+1 */
 977             /*             #endif                                       */
 978             /*             #if ROUND == 'Epilogue' */
 979             vvdw             = _mm_unpacklo_pd(vvdw,_mm_setzero_pd());
 980             /*             #endif */
 981             vvdwsum          = _mm_add_pd(vvdwsum,vvdw);
 982             /*             #define INNERFLOPS INNERFLOPS+1 */
 983             /*         #endif */
 984             /*     #endif */
 985
 986             /*     #if 'Force' in KERNEL_VF */
 987
 988             /*         #if 'electrostatics' in INTERACTION_FLAGS[I][J] and 'vdw' in INTERACTION_FLAGS[I][J] */
 989             fscal            = _mm_add_pd(felec,fvdw);
 990             /*             #define INNERFLOPS INNERFLOPS+1 */
 991             /*         #elif 'electrostatics' in INTERACTION_FLAGS[I][J] */
 992             fscal            = felec;
 993             /*         #elif 'vdw' in INTERACTION_FLAGS[I][J] */
 994             fscal            = fvdw;
 995             /*        #endif */
 996
 997             /*             #if 'exactcutoff' in INTERACTION_FLAGS[I][J] */
 998             fscal            = _mm_and_pd(fscal,cutoff_mask);
 999             /*                 #define INNERFLOPS INNERFLOPS+1 */
1000             /*             #endif                                       */
1001
1002             /*             #if ROUND == 'Epilogue' */
1003             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1004             /*             #endif */
1005
1006             /* ## Construction of vectorial force built into FMA instructions now */
1007             /* #define INNERFLOPS INNERFLOPS+3      */
1008
1009             /* Update vectorial force */
1010             fix{I}             = _mm_macc_pd(dx{I}{J},fscal,fix{I});
1011             fiy{I}             = _mm_macc_pd(dy{I}{J},fscal,fiy{I});
1012             fiz{I}             = _mm_macc_pd(dz{I}{J},fscal,fiz{I});
1013             /*             #define INNERFLOPS INNERFLOPS+6 */
1014
1015             /* #if GEOMETRY_I == 'Particle'             */
1016             /*     #if ROUND == 'Loop' */
1017             gmx_mm_decrement_1rvec_2ptr_swizzle_pd(f+j_coord_offsetA,f+j_coord_offsetB,
1018                                                    _mm_mul_pd(dx{I}{J},fscal),
1019                                                    _mm_mul_pd(dy{I}{J},fscal),
1020                                                    _mm_mul_pd(dz{I}{J},fscal));
1021             /*     #else */
1022             gmx_mm_decrement_1rvec_1ptr_swizzle_pd(f+j_coord_offsetA,
1023                                                    _mm_mul_pd(dx{I}{J},fscal),
1024                                                    _mm_mul_pd(dy{I}{J},fscal),
1025                                                    _mm_mul_pd(dz{I}{J},fscal));
1026             /*     #endif */
1027             /*     #define INNERFLOPS INNERFLOPS+3      */
1028             /* #else                                    */
1029             fjx{J}             = _mm_macc_pd(dx{I}{J},fscal,fjx{J});
1030             fjy{J}             = _mm_macc_pd(dy{I}{J},fscal,fjy{J});
1031             fjz{J}             = _mm_macc_pd(dz{I}{J},fscal,fjz{J});
1032             /*     #define INNERFLOPS INNERFLOPS+3      */
1033             /* #endif                                   */
1034
1035             /*     #endif */
1036
1037             /*     #if 'exactcutoff' in INTERACTION_FLAGS[I][J] */
1038             /*         #if 0    ## This and next two lines is a hack to maintain indentation in template file */
1039             {
1040                 /*     #endif */
1041             }
1042             /*     #endif */
1043             /*    ## End of check for the interaction being outside the cutoff */
1044
1045             /* #endfor */
1046             /* ## End of loop over i-j interaction pairs */
1047
1048             /* #if 'Water' in GEOMETRY_I and GEOMETRY_J == 'Particle' */
1049             /*     #if ROUND == 'Loop' */
1050             gmx_mm_decrement_1rvec_2ptr_swizzle_pd(f+j_coord_offsetA,f+j_coord_offsetB,fjx0,fjy0,fjz0);
1051             /*     #else */
1052             gmx_mm_decrement_1rvec_1ptr_swizzle_pd(f+j_coord_offsetA,fjx0,fjy0,fjz0);
1053             /*     #endif */
1054             /*     #define INNERFLOPS INNERFLOPS+3      */
1055             /* #elif GEOMETRY_J == 'Water3'             */
1056             /*     #if ROUND == 'Loop' */
1057             gmx_mm_decrement_3rvec_2ptr_swizzle_pd(f+j_coord_offsetA,f+j_coord_offsetB,fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2);
1058             /*     #else */
1059             gmx_mm_decrement_3rvec_1ptr_swizzle_pd(f+j_coord_offsetA,fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2);
1060             /*     #endif */
1061             /*     #define INNERFLOPS INNERFLOPS+9      */
1062             /* #elif GEOMETRY_J == 'Water4'             */
1063             /*     #if 0 in PARTICLES_J                 */
1064             /*         #if ROUND == 'Loop' */
1065             gmx_mm_decrement_4rvec_2ptr_swizzle_pd(f+j_coord_offsetA,f+j_coord_offsetB,fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
1066             /*         #else */
1067             gmx_mm_decrement_4rvec_1ptr_swizzle_pd(f+j_coord_offsetA,fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
1068             /*         #endif */
1069             /*         #define INNERFLOPS INNERFLOPS+12 */
1070             /*     #else                                */
1071             /*         #if ROUND == 'Loop' */
1072             gmx_mm_decrement_3rvec_2ptr_swizzle_pd(f+j_coord_offsetA+DIM,f+j_coord_offsetB+DIM,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
1073             /*         #else */
1074             gmx_mm_decrement_3rvec_1ptr_swizzle_pd(f+j_coord_offsetA+DIM,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
1075             /*         #endif */
1076             /*         #define INNERFLOPS INNERFLOPS+9  */
1077             /*     #endif                               */
1078             /* #endif                                   */
1079
1080             /* Inner loop uses {INNERFLOPS} flops */
1081         }
1082
1083         /* #endfor */
1084
1085         /* End of innermost loop */
1086
1087         /* #if 'Force' in KERNEL_VF */
1088         /*     #if GEOMETRY_I == 'Particle'            */
1089         gmx_mm_update_iforce_1atom_swizzle_pd(fix0,fiy0,fiz0,
1090                                               f+i_coord_offset,fshift+i_shift_offset);
1091         /*         #define OUTERFLOPS OUTERFLOPS+6     */
1092         /*     #elif GEOMETRY_I == 'Water3'            */
1093         gmx_mm_update_iforce_3atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,
1094                                               f+i_coord_offset,fshift+i_shift_offset);
1095         /*         #define OUTERFLOPS OUTERFLOPS+18    */
1096         /*     #elif GEOMETRY_I == 'Water4'            */
1097         /*         #if 0 in PARTICLES_I                */
1098         gmx_mm_update_iforce_4atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
1099                                               f+i_coord_offset,fshift+i_shift_offset);
1100         /*             #define OUTERFLOPS OUTERFLOPS+24    */
1101         /*         #else                               */
1102         gmx_mm_update_iforce_3atom_swizzle_pd(fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
1103                                               f+i_coord_offset+DIM,fshift+i_shift_offset);
1104         /*             #define OUTERFLOPS OUTERFLOPS+18    */
1105         /*         #endif                              */
1106         /*     #endif                                  */
1107         /* #endif                                      */
1108
1109         /* #if 'Potential' in KERNEL_VF */
1110         ggid                        = gid[iidx];
1111         /* Update potential energies */
1112         /*     #if KERNEL_ELEC != 'None' */
1113         gmx_mm_update_1pot_pd(velecsum,kernel_data->energygrp_elec+ggid);
1114         /*         #define OUTERFLOPS OUTERFLOPS+1 */
1115         /*     #endif */
1116         /*     #if 'GeneralizedBorn' in KERNEL_ELEC */
1117         gmx_mm_update_1pot_pd(vgbsum,kernel_data->energygrp_polarization+ggid);
1118         /*         #define OUTERFLOPS OUTERFLOPS+1 */
1119         /*     #endif */
1120         /*     #if KERNEL_VDW != 'None' */
1121         gmx_mm_update_1pot_pd(vvdwsum,kernel_data->energygrp_vdw+ggid);
1122         /*         #define OUTERFLOPS OUTERFLOPS+1 */
1123         /*     #endif */
1124         /* #endif */
1125         /*     #if 'GeneralizedBorn' in KERNEL_ELEC and 'Force' in KERNEL_VF */
1126         dvdasum = _mm_mul_pd(dvdasum, _mm_mul_pd(isai{I},isai{I}));
1127         gmx_mm_update_1pot_pd(dvdasum,dvda+inr);
1128         /*     #endif */
1129
1130         /* Increment number of inner iterations */
1131         inneriter                  += j_index_end - j_index_start;
1132
1133         /* Outer loop uses {OUTERFLOPS} flops */
1134     }
1135
1136     /* Increment number of outer iterations */
1137     outeriter        += nri;
1138
1139     /* Update outer/inner flops */
1140     /* ## NB: This is not important, it just affects the flopcount. However, since our preprocessor is */
1141     /* ## primitive and replaces aggressively even in strings inside these directives, we need to      */
1142     /* ## assemble the main part of the name (containing KERNEL/ELEC/VDW) directly in the source.      */
1143     /* #if GEOMETRY_I == 'Water3'            */
1144     /*     #define ISUFFIX '_W3'             */
1145     /* #elif GEOMETRY_I == 'Water4'          */
1146     /*     #define ISUFFIX '_W4'             */
1147     /* #else                                 */
1148     /*     #define ISUFFIX ''                */
1149     /* #endif                                */
1150     /* #if GEOMETRY_J == 'Water3'            */
1151     /*     #define JSUFFIX 'W3'              */
1152     /* #elif GEOMETRY_J == 'Water4'          */
1153     /*     #define JSUFFIX 'W4'              */
1154     /* #else                                 */
1155     /*     #define JSUFFIX ''                */
1156     /* #endif                                */
1157     /* #if 'PotentialAndForce' in KERNEL_VF  */
1158     /*     #define VFSUFFIX  '_VF'           */
1159     /* #elif 'Potential' in KERNEL_VF        */
1160     /*     #define VFSUFFIX '_V'             */
1161     /* #else                                 */
1162     /*     #define VFSUFFIX '_F'             */
1163     /* #endif                                */
1164
1165     /* #if KERNEL_ELEC != 'None' and KERNEL_VDW != 'None' */
1166     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW{ISUFFIX}{JSUFFIX}{VFSUFFIX},outeriter*{OUTERFLOPS} + inneriter*{INNERFLOPS});
1167     /* #elif KERNEL_ELEC != 'None' */
1168     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC{ISUFFIX}{JSUFFIX}{VFSUFFIX},outeriter*{OUTERFLOPS} + inneriter*{INNERFLOPS});
1169     /* #else */
1170     inc_nrnb(nrnb,eNR_NBKERNEL_VDW{ISUFFIX}{JSUFFIX}{VFSUFFIX},outeriter*{OUTERFLOPS} + inneriter*{INNERFLOPS});
1171     /* #endif  */
1172 }