src/gromacs/gmxlib/nonbonded/nb_kernel_sse4_1_single/nb_kernel_template_sse4_1_single.pre

   1 /* #if 0 */
   2 /*
   3  * This file is part of the GROMACS molecular simulation package.
   4  *
   5  * Copyright (c) 2012,2013,2014,2015,2017,2018, 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/gmxlib/nrnb.h"
  47
  48 #include "kernelutil_x86_sse4_1_single.h"
  49 /* #endif */
  50
  51 /* ## List of variables set by the generating script:                                    */
  52 /* ##                                                                                    */
  53 /* ## Setttings that apply to the entire kernel:                                         */
  54 /* ## KERNEL_ELEC:           String, choice for electrostatic interactions               */
  55 /* ## KERNEL_VDW:            String, choice for van der Waals interactions               */
  56 /* ## KERNEL_NAME:           String, name of this kernel                                 */
  57 /* ## KERNEL_VF:             String telling if we calculate potential, force, or both    */
  58 /* ## GEOMETRY_I/GEOMETRY_J: String, name of each geometry, e.g. 'Water3' or '1Particle' */
  59 /* ##                                                                                    */
  60 /* ## Setttings that apply to particles in the outer (I) or inner (J) loops:             */
  61 /* ## PARTICLES_I[]/         Arrays with lists of i/j particles to use in kernel. It is  */
  62 /* ## PARTICLES_J[]:         just [0] for particle geometry, but can be longer for water */
  63 /* ## PARTICLES_ELEC_I[]/    Arrays with lists of i/j particle that have electrostatics  */
  64 /* ## PARTICLES_ELEC_J[]:    interactions that should be calculated in this kernel.      */
  65 /* ## PARTICLES_VDW_I[]/     Arrays with the list of i/j particle that have VdW          */
  66 /* ## PARTICLES_VDW_J[]:     interactions that should be calculated in this kernel.      */
  67 /* ##                                                                                    */
  68 /* ## Setttings for pairs of interactions (e.g. 2nd i particle against 1st j particle)   */
  69 /* ## PAIRS_IJ[]:            Array with (i,j) tuples of pairs for which interactions     */
  70 /* ##                        should be calculated in this kernel. Zero-charge particles  */
  71 /* ##                        do not have interactions with particles without vdw, and    */
  72 /* ##                        Vdw-only interactions are not evaluated in a no-vdw-kernel. */
  73 /* ## INTERACTION_FLAGS[][]: 2D matrix, dimension e.g. 3*3 for water-water interactions. */
  74 /* ##                        For each i-j pair, the element [I][J] is a list of strings  */
  75 /* ##                        defining properties/flags of this interaction. Examples     */
  76 /* ##                        include 'electrostatics'/'vdw' if that type of interaction  */
  77 /* ##                        should be evaluated, 'rsq'/'rinv'/'rinvsq' if those values  */
  78 /* ##                        are needed, and 'exactcutoff' or 'shift','switch' to        */
  79 /* ##                        decide if the force/potential should be modified. This way  */
  80 /* ##                        we only calculate values absolutely needed for each case.   */
  81
  82 /* ## Calculate the size and offset for (merged/interleaved) table data */
  83
  84 /*
  85  * Gromacs nonbonded kernel:   {KERNEL_NAME}
  86  * Electrostatics interaction: {KERNEL_ELEC}
  87  * VdW interaction:            {KERNEL_VDW}
  88  * Geometry:                   {GEOMETRY_I}-{GEOMETRY_J}
  89  * Calculate force/pot:        {KERNEL_VF}
  90  */
  91 void
  92 {KERNEL_NAME}
  93                     (t_nblist                    * gmx_restrict       nlist,
  94                      rvec                        * gmx_restrict          xx,
  95                      rvec                        * gmx_restrict          ff,
  96                      struct t_forcerec           * gmx_restrict          fr,
  97                      t_mdatoms                   * gmx_restrict     mdatoms,
  98                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
  99                      t_nrnb                      * gmx_restrict        nrnb)
 100 {
 101     /* ## Not all variables are used for all kernels, but any optimizing compiler fixes that, */
 102     /* ## so there is no point in going to extremes to exclude variables that are not needed. */
 103     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
 104      * just 0 for non-waters.
 105      * Suffixes A,B,C,D refer to j loop unrolling done with SSE, e.g. for the four different
 106      * jnr indices corresponding to data put in the four positions in the SIMD register.
 107      */
 108     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
 109     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
 110     int              jnrA,jnrB,jnrC,jnrD;
 111     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
 112     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
 113     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 114     real             rcutoff_scalar;
 115     real             *shiftvec,*fshift,*x,*f;
 116     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
 117     real             scratch[4*DIM];
 118     __m128           tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
 119     /* #for I in PARTICLES_I */
 120     int              vdwioffset{I};
 121     __m128           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,vdwjidx{J}C,vdwjidx{J}D;
 125     __m128           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     __m128           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     __m128           velec,felec,velecsum,facel,crf,krf,krf2;
 132     real             *charge;
 133     /* #endif */
 134     /* #if KERNEL_VDW != 'None' */
 135     int              nvdwtype;
 136     __m128           rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 137     int              *vdwtype;
 138     real             *vdwparam;
 139     __m128           one_sixth   = _mm_set1_ps(1.0/6.0);
 140     __m128           one_twelfth = _mm_set1_ps(1.0/12.0);
 141     /* #endif */
 142     /* #if 'Table' in KERNEL_ELEC or 'Table' in KERNEL_VDW */
 143     __m128i          vfitab;
 144     __m128i          ifour       = _mm_set1_epi32(4);
 145     __m128           rt,vfeps,vftabscale,Y,F,G,H,Heps,Fp,VV,FF;
 146     real             *vftab;
 147     /* #endif */
 148     /* #if 'LJEwald' in KERNEL_VDW */
 149     /* #for I,J in PAIRS_IJ */
 150     __m128           c6grid_{I}{J};
 151     /* #endfor */
 152     __m128           ewclj,ewclj2,ewclj6,ewcljrsq,poly,exponent,f6A,f6B,sh_lj_ewald;
 153     real             *vdwgridparam;
 154     __m128           one_half  = _mm_set1_ps(0.5);
 155     __m128           minus_one = _mm_set1_ps(-1.0);
 156     /* #endif */
 157     /* #if 'Ewald' in KERNEL_ELEC */
 158     __m128i          ewitab;
 159     __m128           ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
 160     real             *ewtab;
 161     /* #endif */
 162     /* #if 'PotentialSwitch' in [KERNEL_MOD_ELEC,KERNEL_MOD_VDW] */
 163     __m128           rswitch,swV3,swV4,swV5,swF2,swF3,swF4,d,d2,sw,dsw;
 164     real             rswitch_scalar,d_scalar;
 165     /* #endif */
 166     __m128           dummy_mask,cutoff_mask;
 167     __m128           signbit = _mm_castsi128_ps( _mm_set1_epi32(0x80000000) );
 168     __m128           one     = _mm_set1_ps(1.0);
 169     __m128           two     = _mm_set1_ps(2.0);
 170     x                = xx[0];
 171     f                = ff[0];
 172
 173     nri              = nlist->nri;
 174     iinr             = nlist->iinr;
 175     jindex           = nlist->jindex;
 176     jjnr             = nlist->jjnr;
 177     shiftidx         = nlist->shift;
 178     gid              = nlist->gid;
 179     shiftvec         = fr->shift_vec[0];
 180     fshift           = fr->fshift[0];
 181     /* #if KERNEL_ELEC != 'None' */
 182     facel            = _mm_set1_ps(fr->ic->epsfac);
 183     charge           = mdatoms->chargeA;
 184     /*     #if 'ReactionField' in KERNEL_ELEC */
 185     krf              = _mm_set1_ps(fr->ic->k_rf);
 186     krf2             = _mm_set1_ps(fr->ic->k_rf*2.0);
 187     crf              = _mm_set1_ps(fr->ic->c_rf);
 188     /*     #endif */
 189     /* #endif */
 190     /* #if KERNEL_VDW != 'None' */
 191     nvdwtype         = fr->ntype;
 192     vdwparam         = fr->nbfp;
 193     vdwtype          = mdatoms->typeA;
 194     /* #endif */
 195     /* #if 'LJEwald' in KERNEL_VDW */
 196     vdwgridparam     = fr->ljpme_c6grid;
 197     sh_lj_ewald      = _mm_set1_ps(fr->ic->sh_lj_ewald);
 198     ewclj            = _mm_set1_ps(fr->ic->ewaldcoeff_lj);
 199     ewclj2           = _mm_mul_ps(minus_one,_mm_mul_ps(ewclj,ewclj));
 200     /* #endif */
 201
 202     /* #if 'Table' in KERNEL_ELEC and 'Table' in KERNEL_VDW */
 203     vftab            = kernel_data->table_elec_vdw->data;
 204     vftabscale       = _mm_set1_ps(kernel_data->table_elec_vdw->scale);
 205     /* #elif 'Table' in KERNEL_ELEC */
 206     vftab            = kernel_data->table_elec->data;
 207     vftabscale       = _mm_set1_ps(kernel_data->table_elec->scale);
 208     /* #elif 'Table' in KERNEL_VDW */
 209     vftab            = kernel_data->table_vdw->data;
 210     vftabscale       = _mm_set1_ps(kernel_data->table_vdw->scale);
 211     /* #endif */
 212
 213     /* #if 'Ewald' in KERNEL_ELEC */
 214     sh_ewald         = _mm_set1_ps(fr->ic->sh_ewald);
 215     /*     #if KERNEL_VF=='Force' and KERNEL_MOD_ELEC!='PotentialSwitch' */
 216     ewtab            = fr->ic->tabq_coul_F;
 217     ewtabscale       = _mm_set1_ps(fr->ic->tabq_scale);
 218     ewtabhalfspace   = _mm_set1_ps(0.5/fr->ic->tabq_scale);
 219     /*     #else */
 220     ewtab            = fr->ic->tabq_coul_FDV0;
 221     ewtabscale       = _mm_set1_ps(fr->ic->tabq_scale);
 222     ewtabhalfspace   = _mm_set1_ps(0.5/fr->ic->tabq_scale);
 223      /*     #endif */
 224     /* #endif */
 225
 226     /* #if 'Water' in GEOMETRY_I */
 227     /* Setup water-specific parameters */
 228     inr              = nlist->iinr[0];
 229     /*     #for I in PARTICLES_ELEC_I */
 230     iq{I}              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+{I}]));
 231     /*     #endfor */
 232     /*     #for I in PARTICLES_VDW_I */
 233     vdwioffset{I}      = 2*nvdwtype*vdwtype[inr+{I}];
 234     /*     #endfor */
 235     /* #endif */
 236
 237     /* #if 'Water' in GEOMETRY_J */
 238     /*     #for J in PARTICLES_ELEC_J */
 239     jq{J}              = _mm_set1_ps(charge[inr+{J}]);
 240     /*     #endfor */
 241     /*     #for J in PARTICLES_VDW_J */
 242     vdwjidx{J}A        = 2*vdwtype[inr+{J}];
 243     /*     #endfor */
 244     /*     #for I,J in PAIRS_IJ */
 245     /*         #if 'electrostatics' in INTERACTION_FLAGS[I][J] */
 246     qq{I}{J}             = _mm_mul_ps(iq{I},jq{J});
 247     /*         #endif */
 248     /*         #if 'vdw' in INTERACTION_FLAGS[I][J] */
 249     /*             #if 'LJEwald' in KERNEL_VDW */
 250     c6_{I}{J}            = _mm_set1_ps(vdwparam[vdwioffset{I}+vdwjidx{J}A]);
 251     c12_{I}{J}           = _mm_set1_ps(vdwparam[vdwioffset{I}+vdwjidx{J}A+1]);
 252     c6grid_{I}{J}        = _mm_set1_ps(vdwgridparam[vdwioffset{I}+vdwjidx{J}A]);
 253     /*             #else */
 254     c6_{I}{J}            = _mm_set1_ps(vdwparam[vdwioffset{I}+vdwjidx{J}A]);
 255     c12_{I}{J}           = _mm_set1_ps(vdwparam[vdwioffset{I}+vdwjidx{J}A+1]);
 256     /*             #endif */
 257     /*         #endif */
 258     /*     #endfor */
 259     /* #endif */
 260
 261     /* #if KERNEL_MOD_ELEC!='None' or KERNEL_MOD_VDW!='None' */
 262     /*     #if KERNEL_ELEC!='None' */
 263     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
 264     rcutoff_scalar   = fr->ic->rcoulomb;
 265     /*     #else */
 266     rcutoff_scalar   = fr->ic->rvdw;
 267     /*     #endif */
 268     rcutoff          = _mm_set1_ps(rcutoff_scalar);
 269     rcutoff2         = _mm_mul_ps(rcutoff,rcutoff);
 270     /* #endif */
 271
 272     /* #if KERNEL_MOD_VDW=='PotentialShift' */
 273     sh_vdw_invrcut6  = _mm_set1_ps(fr->ic->sh_invrc6);
 274     rvdw             = _mm_set1_ps(fr->ic->rvdw);
 275     /* #endif */
 276
 277     /* #if 'PotentialSwitch' in [KERNEL_MOD_ELEC,KERNEL_MOD_VDW] */
 278     /*     #if KERNEL_MOD_ELEC=='PotentialSwitch'  */
 279     rswitch_scalar   = fr->ic->rcoulomb_switch;
 280     rswitch          = _mm_set1_ps(rswitch_scalar);
 281     /*     #else */
 282     rswitch_scalar   = fr->ic->rvdw_switch;
 283     rswitch          = _mm_set1_ps(rswitch_scalar);
 284     /*     #endif */
 285     /* Setup switch parameters */
 286     d_scalar         = rcutoff_scalar-rswitch_scalar;
 287     d                = _mm_set1_ps(d_scalar);
 288     swV3             = _mm_set1_ps(-10.0/(d_scalar*d_scalar*d_scalar));
 289     swV4             = _mm_set1_ps( 15.0/(d_scalar*d_scalar*d_scalar*d_scalar));
 290     swV5             = _mm_set1_ps( -6.0/(d_scalar*d_scalar*d_scalar*d_scalar*d_scalar));
 291     /*     #if 'Force' in KERNEL_VF */
 292     swF2             = _mm_set1_ps(-30.0/(d_scalar*d_scalar*d_scalar));
 293     swF3             = _mm_set1_ps( 60.0/(d_scalar*d_scalar*d_scalar*d_scalar));
 294     swF4             = _mm_set1_ps(-30.0/(d_scalar*d_scalar*d_scalar*d_scalar*d_scalar));
 295     /*     #endif */
 296     /* #endif */
 297
 298     /* Avoid stupid compiler warnings */
 299     jnrA = jnrB = jnrC = jnrD = 0;
 300     j_coord_offsetA = 0;
 301     j_coord_offsetB = 0;
 302     j_coord_offsetC = 0;
 303     j_coord_offsetD = 0;
 304
 305     /* ## Keep track of the floating point operations we issue for reporting! */
 306     /* #define OUTERFLOPS 0 */
 307     outeriter        = 0;
 308     inneriter        = 0;
 309
 310     for(iidx=0;iidx<4*DIM;iidx++)
 311     {
 312         scratch[iidx] = 0.0;
 313     }
 314
 315     /* Start outer loop over neighborlists */
 316     for(iidx=0; iidx<nri; iidx++)
 317     {
 318         /* Load shift vector for this list */
 319         i_shift_offset   = DIM*shiftidx[iidx];
 320
 321         /* Load limits for loop over neighbors */
 322         j_index_start    = jindex[iidx];
 323         j_index_end      = jindex[iidx+1];
 324
 325         /* Get outer coordinate index */
 326         inr              = iinr[iidx];
 327         i_coord_offset   = DIM*inr;
 328
 329         /* Load i particle coords and add shift vector */
 330         /* #if GEOMETRY_I == 'Particle' */
 331         gmx_mm_load_shift_and_1rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,&ix0,&iy0,&iz0);
 332         /* #elif GEOMETRY_I == 'Water3' */
 333         gmx_mm_load_shift_and_3rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,
 334                                                  &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2);
 335         /* #elif GEOMETRY_I == 'Water4' */
 336         /*     #if 0 in PARTICLES_I                 */
 337         gmx_mm_load_shift_and_4rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,
 338                                                  &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
 339         /*     #else                                */
 340         gmx_mm_load_shift_and_3rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset+DIM,
 341                                                  &ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
 342         /*     #endif                               */
 343         /* #endif                                   */
 344
 345         /* #if 'Force' in KERNEL_VF */
 346         /*     #for I in PARTICLES_I */
 347         fix{I}             = _mm_setzero_ps();
 348         fiy{I}             = _mm_setzero_ps();
 349         fiz{I}             = _mm_setzero_ps();
 350         /*     #endfor */
 351         /* #endif */
 352
 353         /* ## For water we already preloaded parameters at the start of the kernel */
 354         /* #if not 'Water' in GEOMETRY_I */
 355         /* Load parameters for i particles */
 356         /*     #for I in PARTICLES_ELEC_I */
 357         iq{I}              = _mm_mul_ps(facel,_mm_load1_ps(charge+inr+{I}));
 358         /*         #define OUTERFLOPS OUTERFLOPS+1 */
 359         /*     #endfor */
 360         /*     #for I in PARTICLES_VDW_I */
 361         vdwioffset{I}      = 2*nvdwtype*vdwtype[inr+{I}];
 362         /*     #endfor */
 363         /* #endif */
 364
 365         /* #if 'Potential' in KERNEL_VF */
 366         /* Reset potential sums */
 367         /*     #if KERNEL_ELEC != 'None' */
 368         velecsum         = _mm_setzero_ps();
 369         /*     #endif */
 370         /*     #if KERNEL_VDW != 'None' */
 371         vvdwsum          = _mm_setzero_ps();
 372         /*     #endif */
 373         /* #endif */
 374
 375         /* #for ROUND in ['Loop','Epilogue'] */
 376
 377         /* #if ROUND =='Loop' */
 378         /* Start inner kernel loop */
 379         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 380         {
 381         /* ## First round is normal loop (next statement resets indentation) */
 382         /*     #if 0 */
 383         }
 384         /*     #endif */
 385         /* #else */
 386         if(jidx<j_index_end)
 387         {
 388         /* ## Second round is epilogue */
 389         /* #endif */
 390         /* #define INNERFLOPS 0 */
 391
 392             /* Get j neighbor index, and coordinate index */
 393             /* #if ROUND =='Loop' */
 394             jnrA             = jjnr[jidx];
 395             jnrB             = jjnr[jidx+1];
 396             jnrC             = jjnr[jidx+2];
 397             jnrD             = jjnr[jidx+3];
 398             /* #else */
 399             jnrlistA         = jjnr[jidx];
 400             jnrlistB         = jjnr[jidx+1];
 401             jnrlistC         = jjnr[jidx+2];
 402             jnrlistD         = jjnr[jidx+3];
 403             /* Sign of each element will be negative for non-real atoms.
 404              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 405              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
 406              */
 407             dummy_mask = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
 408             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 409             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 410             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 411             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 412             /* #endif */
 413             j_coord_offsetA  = DIM*jnrA;
 414             j_coord_offsetB  = DIM*jnrB;
 415             j_coord_offsetC  = DIM*jnrC;
 416             j_coord_offsetD  = DIM*jnrD;
 417
 418             /* load j atom coordinates */
 419             /* #if GEOMETRY_J == 'Particle'             */
 420             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 421                                               x+j_coord_offsetC,x+j_coord_offsetD,
 422                                               &jx0,&jy0,&jz0);
 423             /* #elif GEOMETRY_J == 'Water3'             */
 424             gmx_mm_load_3rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 425                                               x+j_coord_offsetC,x+j_coord_offsetD,
 426                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,&jy2,&jz2);
 427             /* #elif GEOMETRY_J == 'Water4'             */
 428             /*     #if 0 in PARTICLES_J                 */
 429             gmx_mm_load_4rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 430                                               x+j_coord_offsetC,x+j_coord_offsetD,
 431                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
 432                                               &jy2,&jz2,&jx3,&jy3,&jz3);
 433             /*     #else                                */
 434             gmx_mm_load_3rvec_4ptr_swizzle_ps(x+j_coord_offsetA+DIM,x+j_coord_offsetB+DIM,
 435                                               x+j_coord_offsetC+DIM,x+j_coord_offsetD+DIM,
 436                                               &jx1,&jy1,&jz1,&jx2,&jy2,&jz2,&jx3,&jy3,&jz3);
 437             /*     #endif                               */
 438             /* #endif                                   */
 439
 440             /* Calculate displacement vector */
 441             /* #for I,J in PAIRS_IJ */
 442             dx{I}{J}             = _mm_sub_ps(ix{I},jx{J});
 443             dy{I}{J}             = _mm_sub_ps(iy{I},jy{J});
 444             dz{I}{J}             = _mm_sub_ps(iz{I},jz{J});
 445             /*     #define INNERFLOPS INNERFLOPS+3 */
 446             /* #endfor */
 447
 448             /* Calculate squared distance and things based on it */
 449             /* #for I,J in PAIRS_IJ */
 450             rsq{I}{J}            = gmx_mm_calc_rsq_ps(dx{I}{J},dy{I}{J},dz{I}{J});
 451             /*     #define INNERFLOPS INNERFLOPS+5 */
 452             /* #endfor */
 453
 454             /* #for I,J in PAIRS_IJ */
 455             /*     #if 'rinv' in INTERACTION_FLAGS[I][J] */
 456             rinv{I}{J}           = sse41_invsqrt_f(rsq{I}{J});
 457             /*         #define INNERFLOPS INNERFLOPS+5 */
 458             /*     #endif */
 459             /* #endfor */
 460
 461             /* #for I,J in PAIRS_IJ */
 462             /*     #if 'rinvsq' in INTERACTION_FLAGS[I][J] */
 463             /*         # if 'rinv' not in INTERACTION_FLAGS[I][J] */
 464             rinvsq{I}{J}         = sse41_inv_f(rsq{I}{J});
 465             /*             #define INNERFLOPS INNERFLOPS+4 */
 466             /*         #else */
 467             rinvsq{I}{J}         = _mm_mul_ps(rinv{I}{J},rinv{I}{J});
 468             /*             #define INNERFLOPS INNERFLOPS+1 */
 469             /*         #endif */
 470             /*     #endif */
 471             /* #endfor */
 472
 473             /* #if not 'Water' in GEOMETRY_J */
 474             /* Load parameters for j particles */
 475             /*     #for J in PARTICLES_ELEC_J */
 476             jq{J}              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+{J},charge+jnrB+{J},
 477                                                               charge+jnrC+{J},charge+jnrD+{J});
 478             /*     #endfor */
 479             /*     #for J in PARTICLES_VDW_J */
 480             vdwjidx{J}A        = 2*vdwtype[jnrA+{J}];
 481             vdwjidx{J}B        = 2*vdwtype[jnrB+{J}];
 482             vdwjidx{J}C        = 2*vdwtype[jnrC+{J}];
 483             vdwjidx{J}D        = 2*vdwtype[jnrD+{J}];
 484             /*     #endfor */
 485             /* #endif */
 486
 487             /* #if 'Force' in KERNEL_VF and not 'Particle' in GEOMETRY_I */
 488             /*     #for J in PARTICLES_J */
 489             fjx{J}             = _mm_setzero_ps();
 490             fjy{J}             = _mm_setzero_ps();
 491             fjz{J}             = _mm_setzero_ps();
 492             /*     #endfor */
 493             /* #endif */
 494
 495             /* #for I,J in PAIRS_IJ */
 496
 497             /**************************
 498              * CALCULATE INTERACTIONS *
 499              **************************/
 500
 501             /*     ## Note special check for TIP4P-TIP4P. Since we are cutting of all hydrogen interactions we also cut the LJ-only O-O interaction */
 502             /*     #if 'exactcutoff' in INTERACTION_FLAGS[I][J] or (GEOMETRY_I=='Water4' and GEOMETRY_J=='Water4' and 'exactcutoff' in INTERACTION_FLAGS[1][1]) */
 503             /*         ## We always calculate rinv/rinvsq above to enable pipelineing in compilers (performance tested on x86) */
 504             if (gmx_mm_any_lt(rsq{I}{J},rcutoff2))
 505             {
 506                 /*     #if 0    ## this and the next two lines is a hack to maintain auto-indentation in template file */
 507             }
 508             /*         #endif */
 509             /*         #define INNERFLOPS INNERFLOPS+1 */
 510             /*     #endif */
 511
 512             /*     #if 'r' in INTERACTION_FLAGS[I][J] */
 513             r{I}{J}              = _mm_mul_ps(rsq{I}{J},rinv{I}{J});
 514             /*         #if ROUND == 'Epilogue' */
 515             r{I}{J}              = _mm_andnot_ps(dummy_mask,r{I}{J});
 516             /*             #define INNERFLOPS INNERFLOPS+1 */
 517             /*         #endif */
 518             /*         #define INNERFLOPS INNERFLOPS+1 */
 519             /*     #endif */
 520
 521             /*     ## For water geometries we already loaded parameters at the start of the kernel */
 522             /*     #if not 'Water' in GEOMETRY_J */
 523             /* Compute parameters for interactions between i and j atoms */
 524             /*         #if 'electrostatics' in INTERACTION_FLAGS[I][J] */
 525             qq{I}{J}             = _mm_mul_ps(iq{I},jq{J});
 526             /*             #define INNERFLOPS INNERFLOPS+1 */
 527             /*         #endif */
 528             /*         #if 'vdw' in INTERACTION_FLAGS[I][J] */
 529             gmx_mm_load_4pair_swizzle_ps(vdwparam+vdwioffset{I}+vdwjidx{J}A,
 530                                          vdwparam+vdwioffset{I}+vdwjidx{J}B,
 531                                          vdwparam+vdwioffset{I}+vdwjidx{J}C,
 532                                          vdwparam+vdwioffset{I}+vdwjidx{J}D,
 533                                          &c6_{I}{J},&c12_{I}{J});
 534
 535             /*         #if 'LJEwald' in KERNEL_VDW */
 536             c6grid_{I}{J}       = gmx_mm_load_4real_swizzle_ps(vdwgridparam+vdwioffset{I}+vdwjidx{J}A,
 537                                                                vdwgridparam+vdwioffset{I}+vdwjidx{J}B,
 538                                                                vdwgridparam+vdwioffset{I}+vdwjidx{J}C,
 539                                                                vdwgridparam+vdwioffset{I}+vdwjidx{J}D);
 540             /*          #endif */
 541             /*         #endif */
 542             /*     #endif */
 543
 544             /*     #if 'table' in INTERACTION_FLAGS[I][J] */
 545             /* Calculate table index by multiplying r with table scale and truncate to integer */
 546             rt               = _mm_mul_ps(r{I}{J},vftabscale);
 547             vfitab           = _mm_cvttps_epi32(rt);
 548             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
 549             /*         #define INNERFLOPS INNERFLOPS+4                          */
 550             /*         #if 'Table' in KERNEL_ELEC and 'Table' in KERNEL_VDW     */
 551             /*             ## 3 tables, 4 bytes per point: multiply index by 12 */
 552             vfitab           = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
 553             /*         #elif 'Table' in KERNEL_ELEC                             */
 554             /*             ## 1 table, 4 bytes per point: multiply index by 4   */
 555             vfitab           = _mm_slli_epi32(vfitab,2);
 556             /*         #elif 'Table' in KERNEL_VDW                              */
 557             /*             ## 2 tables, 4 bytes per point: multiply index by 8  */
 558             vfitab           = _mm_slli_epi32(vfitab,3);
 559             /*         #endif                                                   */
 560             /*     #endif */
 561
 562             /*     ## ELECTROSTATIC INTERACTIONS */
 563             /*     #if 'electrostatics' in INTERACTION_FLAGS[I][J] */
 564
 565             /*         #if KERNEL_ELEC=='Coulomb' */
 566
 567             /* COULOMB ELECTROSTATICS */
 568             velec            = _mm_mul_ps(qq{I}{J},rinv{I}{J});
 569             /*             #define INNERFLOPS INNERFLOPS+1 */
 570             /*             #if 'Force' in KERNEL_VF */
 571             felec            = _mm_mul_ps(velec,rinvsq{I}{J});
 572             /*                 #define INNERFLOPS INNERFLOPS+2 */
 573             /*             #endif */
 574
 575             /*         #elif KERNEL_ELEC=='ReactionField' */
 576
 577             /* REACTION-FIELD ELECTROSTATICS */
 578             /*             #if 'Potential' in KERNEL_VF */
 579             velec            = _mm_mul_ps(qq{I}{J},_mm_sub_ps(_mm_add_ps(rinv{I}{J},_mm_mul_ps(krf,rsq{I}{J})),crf));
 580             /*                 #define INNERFLOPS INNERFLOPS+4 */
 581             /*             #endif */
 582             /*             #if 'Force' in KERNEL_VF */
 583             felec            = _mm_mul_ps(qq{I}{J},_mm_sub_ps(_mm_mul_ps(rinv{I}{J},rinvsq{I}{J}),krf2));
 584             /*                 #define INNERFLOPS INNERFLOPS+3 */
 585             /*             #endif */
 586
 587             /*         #elif KERNEL_ELEC=='Ewald' */
 588             /* EWALD ELECTROSTATICS */
 589
 590             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 591             ewrt             = _mm_mul_ps(r{I}{J},ewtabscale);
 592             ewitab           = _mm_cvttps_epi32(ewrt);
 593             eweps            = _mm_sub_ps(ewrt,_mm_round_ps(ewrt, _MM_FROUND_FLOOR));
 594             /*             #define INNERFLOPS INNERFLOPS+4 */
 595             /*             #if 'Potential' in KERNEL_VF or KERNEL_MOD_ELEC=='PotentialSwitch' */
 596             ewitab           = _mm_slli_epi32(ewitab,2);
 597             ewtabF           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,0) );
 598             ewtabD           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,1) );
 599             ewtabV           = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,2) );
 600             ewtabFn          = _mm_load_ps( ewtab + gmx_mm_extract_epi32(ewitab,3) );
 601             _MM_TRANSPOSE4_PS(ewtabF,ewtabD,ewtabV,ewtabFn);
 602             felec            = _mm_add_ps(ewtabF,_mm_mul_ps(eweps,ewtabD));
 603             /*                 #define INNERFLOPS INNERFLOPS+2 */
 604             /*                 #if KERNEL_MOD_ELEC=='PotentialShift' */
 605             velec            = _mm_sub_ps(ewtabV,_mm_mul_ps(_mm_mul_ps(ewtabhalfspace,eweps),_mm_add_ps(ewtabF,felec)));
 606             velec            = _mm_mul_ps(qq{I}{J},_mm_sub_ps(_mm_sub_ps(rinv{I}{J},sh_ewald),velec));
 607             /*                     #define INNERFLOPS INNERFLOPS+7 */
 608             /*                 #else */
 609             velec            = _mm_sub_ps(ewtabV,_mm_mul_ps(_mm_mul_ps(ewtabhalfspace,eweps),_mm_add_ps(ewtabF,felec)));
 610             velec            = _mm_mul_ps(qq{I}{J},_mm_sub_ps(rinv{I}{J},velec));
 611             /*                     #define INNERFLOPS INNERFLOPS+6 */
 612             /*                 #endif */
 613             /*                 #if 'Force' in KERNEL_VF */
 614             felec            = _mm_mul_ps(_mm_mul_ps(qq{I}{J},rinv{I}{J}),_mm_sub_ps(rinvsq{I}{J},felec));
 615             /*                      #define INNERFLOPS INNERFLOPS+3 */
 616             /*                 #endif */
 617             /*             #elif KERNEL_VF=='Force' */
 618             gmx_mm_load_4pair_swizzle_ps(ewtab + gmx_mm_extract_epi32(ewitab,0),ewtab + gmx_mm_extract_epi32(ewitab,1),
 619                                          ewtab + gmx_mm_extract_epi32(ewitab,2),ewtab + gmx_mm_extract_epi32(ewitab,3),
 620                                          &ewtabF,&ewtabFn);
 621             felec            = _mm_add_ps(_mm_mul_ps( _mm_sub_ps(one,eweps),ewtabF),_mm_mul_ps(eweps,ewtabFn));
 622             felec            = _mm_mul_ps(_mm_mul_ps(qq{I}{J},rinv{I}{J}),_mm_sub_ps(rinvsq{I}{J},felec));
 623             /*                 #define INNERFLOPS INNERFLOPS+7 */
 624             /*             #endif */
 625
 626             /*         #elif KERNEL_ELEC=='CubicSplineTable' */
 627
 628             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 629             Y                = _mm_load_ps( vftab + gmx_mm_extract_epi32(vfitab,0) );
 630             F                = _mm_load_ps( vftab + gmx_mm_extract_epi32(vfitab,1) );
 631             G                = _mm_load_ps( vftab + gmx_mm_extract_epi32(vfitab,2) );
 632             H                = _mm_load_ps( vftab + gmx_mm_extract_epi32(vfitab,3) );
 633             _MM_TRANSPOSE4_PS(Y,F,G,H);
 634             Heps             = _mm_mul_ps(vfeps,H);
 635             Fp               = _mm_add_ps(F,_mm_mul_ps(vfeps,_mm_add_ps(G,Heps)));
 636             /*             #define INNERFLOPS INNERFLOPS+4 */
 637             /*             #if 'Potential' in KERNEL_VF */
 638             VV               = _mm_add_ps(Y,_mm_mul_ps(vfeps,Fp));
 639             velec            = _mm_mul_ps(qq{I}{J},VV);
 640             /*                 #define INNERFLOPS INNERFLOPS+3 */
 641             /*             #endif */
 642             /*             #if 'Force' in KERNEL_VF */
 643             FF               = _mm_add_ps(Fp,_mm_mul_ps(vfeps,_mm_add_ps(G,_mm_add_ps(Heps,Heps))));
 644             felec            = _mm_xor_ps(signbit,_mm_mul_ps(_mm_mul_ps(qq{I}{J},FF),_mm_mul_ps(vftabscale,rinv{I}{J})));
 645             /*                 #define INNERFLOPS INNERFLOPS+7 */
 646             /*             #endif */
 647             /*         #endif */
 648             /*         ## End of check for electrostatics interaction forms */
 649             /*     #endif */
 650             /*     ## END OF ELECTROSTATIC INTERACTION CHECK FOR PAIR I-J */
 651
 652             /*     #if 'vdw' in INTERACTION_FLAGS[I][J] */
 653
 654             /*         #if KERNEL_VDW=='LennardJones' */
 655
 656             /* LENNARD-JONES DISPERSION/REPULSION */
 657
 658             rinvsix          = _mm_mul_ps(_mm_mul_ps(rinvsq{I}{J},rinvsq{I}{J}),rinvsq{I}{J});
 659             /*             #define INNERFLOPS INNERFLOPS+2 */
 660             /*             #if 'Potential' in KERNEL_VF or KERNEL_MOD_VDW=='PotentialSwitch' */
 661             vvdw6            = _mm_mul_ps(c6_{I}{J},rinvsix);
 662             vvdw12           = _mm_mul_ps(c12_{I}{J},_mm_mul_ps(rinvsix,rinvsix));
 663             /*                 #define INNERFLOPS INNERFLOPS+3 */
 664             /*                 #if KERNEL_MOD_VDW=='PotentialShift' */
 665             vvdw             = _mm_sub_ps(_mm_mul_ps( _mm_sub_ps(vvdw12 , _mm_mul_ps(c12_{I}{J},_mm_mul_ps(sh_vdw_invrcut6,sh_vdw_invrcut6))), one_twelfth) ,
 666                                           _mm_mul_ps( _mm_sub_ps(vvdw6,_mm_mul_ps(c6_{I}{J},sh_vdw_invrcut6)),one_sixth));
 667             /*                     #define INNERFLOPS INNERFLOPS+8 */
 668             /*                 #else */
 669             vvdw             = _mm_sub_ps( _mm_mul_ps(vvdw12,one_twelfth) , _mm_mul_ps(vvdw6,one_sixth) );
 670             /*                     #define INNERFLOPS INNERFLOPS+3 */
 671             /*                 #endif */
 672             /*                 ## Check for force inside potential check, i.e. this means we already did the potential part */
 673             /*                 #if 'Force' in KERNEL_VF */
 674             fvdw             = _mm_mul_ps(_mm_sub_ps(vvdw12,vvdw6),rinvsq{I}{J});
 675             /*                     #define INNERFLOPS INNERFLOPS+2 */
 676             /*                 #endif */
 677             /*             #elif KERNEL_VF=='Force' */
 678             /*                 ## Force-only LennardJones makes it possible to save 1 flop (they do add up...) */
 679             fvdw             = _mm_mul_ps(_mm_sub_ps(_mm_mul_ps(c12_{I}{J},rinvsix),c6_{I}{J}),_mm_mul_ps(rinvsix,rinvsq{I}{J}));
 680             /*                 #define INNERFLOPS INNERFLOPS+4 */
 681             /*             #endif */
 682
 683             /*         #elif KERNEL_VDW=='CubicSplineTable' */
 684
 685             /* CUBIC SPLINE TABLE DISPERSION */
 686             /*             #if 'Table' in KERNEL_ELEC */
 687             vfitab           = _mm_add_epi32(vfitab,ifour);
 688             /*             #endif                     */
 689             Y                = _mm_load_ps( vftab + gmx_mm_extract_epi32(vfitab,0) );
 690             F                = _mm_load_ps( vftab + gmx_mm_extract_epi32(vfitab,1) );
 691             G                = _mm_load_ps( vftab + gmx_mm_extract_epi32(vfitab,2) );
 692             H                = _mm_load_ps( vftab + gmx_mm_extract_epi32(vfitab,3) );
 693             _MM_TRANSPOSE4_PS(Y,F,G,H);
 694             Heps             = _mm_mul_ps(vfeps,H);
 695             Fp               = _mm_add_ps(F,_mm_mul_ps(vfeps,_mm_add_ps(G,Heps)));
 696             /*             #define INNERFLOPS INNERFLOPS+4 */
 697             /*             #if 'Potential' in KERNEL_VF */
 698             VV               = _mm_add_ps(Y,_mm_mul_ps(vfeps,Fp));
 699             vvdw6            = _mm_mul_ps(c6_{I}{J},VV);
 700             /*                 #define INNERFLOPS INNERFLOPS+3 */
 701             /*             #endif */
 702             /*             #if 'Force' in KERNEL_VF */
 703             FF               = _mm_add_ps(Fp,_mm_mul_ps(vfeps,_mm_add_ps(G,_mm_add_ps(Heps,Heps))));
 704             fvdw6            = _mm_mul_ps(c6_{I}{J},FF);
 705             /*                 #define INNERFLOPS INNERFLOPS+4 */
 706             /*             #endif */
 707
 708             /* CUBIC SPLINE TABLE REPULSION */
 709             vfitab           = _mm_add_epi32(vfitab,ifour);
 710             Y                = _mm_load_ps( vftab + gmx_mm_extract_epi32(vfitab,0) );
 711             F                = _mm_load_ps( vftab + gmx_mm_extract_epi32(vfitab,1) );
 712             G                = _mm_load_ps( vftab + gmx_mm_extract_epi32(vfitab,2) );
 713             H                = _mm_load_ps( vftab + gmx_mm_extract_epi32(vfitab,3) );
 714             _MM_TRANSPOSE4_PS(Y,F,G,H);
 715             Heps             = _mm_mul_ps(vfeps,H);
 716             Fp               = _mm_add_ps(F,_mm_mul_ps(vfeps,_mm_add_ps(G,Heps)));
 717             /*             #define INNERFLOPS INNERFLOPS+4 */
 718             /*             #if 'Potential' in KERNEL_VF */
 719             VV               = _mm_add_ps(Y,_mm_mul_ps(vfeps,Fp));
 720             vvdw12           = _mm_mul_ps(c12_{I}{J},VV);
 721             /*                 #define INNERFLOPS INNERFLOPS+3 */
 722             /*             #endif */
 723             /*             #if 'Force' in KERNEL_VF */
 724             FF               = _mm_add_ps(Fp,_mm_mul_ps(vfeps,_mm_add_ps(G,_mm_add_ps(Heps,Heps))));
 725             fvdw12           = _mm_mul_ps(c12_{I}{J},FF);
 726             /*                 #define INNERFLOPS INNERFLOPS+5 */
 727             /*             #endif */
 728             /*             #if 'Potential' in KERNEL_VF */
 729             vvdw             = _mm_add_ps(vvdw12,vvdw6);
 730             /*                 #define INNERFLOPS INNERFLOPS+1 */
 731             /*             #endif */
 732             /*             #if 'Force' in KERNEL_VF */
 733             fvdw             = _mm_xor_ps(signbit,_mm_mul_ps(_mm_add_ps(fvdw6,fvdw12),_mm_mul_ps(vftabscale,rinv{I}{J})));
 734             /*                 #define INNERFLOPS INNERFLOPS+4 */
 735             /*             #endif */
 736
 737             /*         #elif KERNEL_VDW=='LJEwald' */
 738
 739             /* Analytical LJ-PME */
 740             rinvsix          = _mm_mul_ps(_mm_mul_ps(rinvsq{I}{J},rinvsq{I}{J}),rinvsq{I}{J});
 741             ewcljrsq         = _mm_mul_ps(ewclj2,rsq{I}{J});
 742             ewclj6           = _mm_mul_ps(ewclj2,_mm_mul_ps(ewclj2,ewclj2));
 743             exponent         = sse41_exp_f(ewcljrsq);
 744             /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
 745             poly             = _mm_mul_ps(exponent,_mm_add_ps(_mm_sub_ps(one,ewcljrsq),_mm_mul_ps(_mm_mul_ps(ewcljrsq,ewcljrsq),one_half)));
 746             /*             #define INNERFLOPS INNERFLOPS+11 */
 747             /*             #if 'Potential' in KERNEL_VF or KERNEL_MOD_VDW=='PotentialSwitch' */
 748             /* vvdw6 = [C6 - C6grid * (1-poly)]/r6 */
 749             vvdw6            = _mm_mul_ps(_mm_sub_ps(c6_{I}{J},_mm_mul_ps(c6grid_{I}{J},_mm_sub_ps(one,poly))),rinvsix);
 750             vvdw12           = _mm_mul_ps(c12_{I}{J},_mm_mul_ps(rinvsix,rinvsix));
 751             /*                 #define INNERFLOPS INNERFLOPS+6 */
 752             /*                 #if KERNEL_MOD_VDW=='PotentialShift' */
 753             vvdw             = _mm_sub_ps(_mm_mul_ps( _mm_sub_ps(vvdw12 , _mm_mul_ps(c12_{I}{J},_mm_mul_ps(sh_vdw_invrcut6,sh_vdw_invrcut6))),one_twelfth),
 754                                _mm_mul_ps( _mm_sub_ps(vvdw6,_mm_add_ps(_mm_mul_ps(c6_{I}{J},sh_vdw_invrcut6),_mm_mul_ps(c6grid_{I}{J},sh_lj_ewald))),one_sixth));
 755             /*                 #define INNERFLOPS INNERFLOPS+10 */
 756             /*                 #else */
 757             vvdw             = _mm_sub_ps(_mm_mul_ps(vvdw12,one_twelfth),_mm_mul_ps(vvdw6,one_sixth));
 758             /*                 #define INNERFLOPS INNERFLOPS+3 */
 759             /*                 #endif */
 760             /*                 ## Check for force inside potential check, i.e. this means we already did the potential part */
 761             /*                 #if 'Force' in KERNEL_VF */
 762             /* fvdw = vvdw12/r - (vvdw6/r + (C6grid * exponent * beta^6)/r) */
 763             fvdw             = _mm_mul_ps(_mm_sub_ps(vvdw12,_mm_sub_ps(vvdw6,_mm_mul_ps(_mm_mul_ps(c6grid_{I}{J},one_sixth),_mm_mul_ps(exponent,ewclj6)))),rinvsq{I}{J});
 764             /*                 #define INNERFLOPS INNERFLOPS+6 */
 765             /*                 #endif */
 766             /*             #elif KERNEL_VF=='Force' */
 767             /* f6A = 6 * C6grid * (1 - poly) */
 768             f6A              = _mm_mul_ps(c6grid_{I}{J},_mm_sub_ps(one,poly));
 769             /* f6B = C6grid * exponent * beta^6 */
 770             f6B              = _mm_mul_ps(_mm_mul_ps(c6grid_{I}{J},one_sixth),_mm_mul_ps(exponent,ewclj6));
 771             /* fvdw = 12*C12/r13 - ((6*C6 - f6A)/r6 + f6B)/r */
 772             fvdw              = _mm_mul_ps(_mm_add_ps(_mm_mul_ps(_mm_sub_ps(_mm_mul_ps(c12_{I}{J},rinvsix),_mm_sub_ps(c6_{I}{J},f6A)),rinvsix),f6B),rinvsq{I}{J});
 773             /*                 #define INNERFLOPS INNERFLOPS+11 */
 774             /*             #endif */
 775             /*         #endif */
 776             /*         ## End of check for vdw interaction forms */
 777             /*     #endif */
 778             /*     ## END OF VDW INTERACTION CHECK FOR PAIR I-J */
 779
 780             /*     #if 'switch' in INTERACTION_FLAGS[I][J] */
 781             d                = _mm_sub_ps(r{I}{J},rswitch);
 782             d                = _mm_max_ps(d,_mm_setzero_ps());
 783             d2               = _mm_mul_ps(d,d);
 784             sw               = _mm_add_ps(one,_mm_mul_ps(d2,_mm_mul_ps(d,_mm_add_ps(swV3,_mm_mul_ps(d,_mm_add_ps(swV4,_mm_mul_ps(d,swV5)))))));
 785             /*         #define INNERFLOPS INNERFLOPS+10 */
 786
 787             /*         #if 'Force' in KERNEL_VF */
 788             dsw              = _mm_mul_ps(d2,_mm_add_ps(swF2,_mm_mul_ps(d,_mm_add_ps(swF3,_mm_mul_ps(d,swF4)))));
 789             /*             #define INNERFLOPS INNERFLOPS+5 */
 790             /*         #endif */
 791
 792             /* Evaluate switch function */
 793             /*         #if 'Force' in KERNEL_VF */
 794             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 795             /*             #if 'electrostatics' in INTERACTION_FLAGS[I][J] and KERNEL_MOD_ELEC=='PotentialSwitch' */
 796             felec            = _mm_sub_ps( _mm_mul_ps(felec,sw) , _mm_mul_ps(rinv{I}{J},_mm_mul_ps(velec,dsw)) );
 797             /*                 #define INNERFLOPS INNERFLOPS+4 */
 798             /*             #endif */
 799             /*             #if 'vdw' in INTERACTION_FLAGS[I][J] and KERNEL_MOD_VDW=='PotentialSwitch' */
 800             fvdw             = _mm_sub_ps( _mm_mul_ps(fvdw,sw) , _mm_mul_ps(rinv{I}{J},_mm_mul_ps(vvdw,dsw)) );
 801             /*                 #define INNERFLOPS INNERFLOPS+4 */
 802             /*             #endif */
 803             /*         #endif */
 804             /*         #if 'Potential' in KERNEL_VF */
 805             /*             #if 'electrostatics' in INTERACTION_FLAGS[I][J] and KERNEL_MOD_ELEC=='PotentialSwitch' */
 806             velec            = _mm_mul_ps(velec,sw);
 807             /*                 #define INNERFLOPS INNERFLOPS+1 */
 808             /*             #endif */
 809             /*             #if 'vdw' in INTERACTION_FLAGS[I][J] and KERNEL_MOD_VDW=='PotentialSwitch' */
 810             vvdw             = _mm_mul_ps(vvdw,sw);
 811             /*                 #define INNERFLOPS INNERFLOPS+1 */
 812             /*             #endif */
 813             /*         #endif */
 814             /*     #endif */
 815             /*     ## Note special check for TIP4P-TIP4P. Since we are cutting of all hydrogen interactions we also cut the LJ-only O-O interaction */
 816             /*     #if 'exactcutoff' in INTERACTION_FLAGS[I][J] or (GEOMETRY_I=='Water4' and GEOMETRY_J=='Water4' and 'exactcutoff' in INTERACTION_FLAGS[1][1]) */
 817             cutoff_mask      = _mm_cmplt_ps(rsq{I}{J},rcutoff2);
 818             /*         #define INNERFLOPS INNERFLOPS+1 */
 819             /*     #endif */
 820
 821             /*     #if 'Potential' in KERNEL_VF */
 822             /* Update potential sum for this i atom from the interaction with this j atom. */
 823             /*         #if 'electrostatics' in INTERACTION_FLAGS[I][J] */
 824             /*             #if 'exactcutoff' in INTERACTION_FLAGS[I][J] */
 825             velec            = _mm_and_ps(velec,cutoff_mask);
 826             /*                 #define INNERFLOPS INNERFLOPS+1 */
 827             /*             #endif                                       */
 828             /*             #if ROUND == 'Epilogue' */
 829             velec            = _mm_andnot_ps(dummy_mask,velec);
 830             /*             #endif */
 831             velecsum         = _mm_add_ps(velecsum,velec);
 832             /*             #define INNERFLOPS INNERFLOPS+1 */
 833             /*         #endif */
 834             /*         #if 'vdw' in INTERACTION_FLAGS[I][J] */
 835             /*     ## Note special check for TIP4P-TIP4P. Since we are cutting of all hydrogen interactions we also cut the LJ-only O-O interaction */
 836             /*     #if 'exactcutoff' in INTERACTION_FLAGS[I][J] or (GEOMETRY_I=='Water4' and GEOMETRY_J=='Water4' and 'exactcutoff' in INTERACTION_FLAGS[1][1]) */
 837             vvdw             = _mm_and_ps(vvdw,cutoff_mask);
 838             /*                 #define INNERFLOPS INNERFLOPS+1 */
 839             /*             #endif                                       */
 840             /*             #if ROUND == 'Epilogue' */
 841             vvdw             = _mm_andnot_ps(dummy_mask,vvdw);
 842             /*             #endif */
 843             vvdwsum          = _mm_add_ps(vvdwsum,vvdw);
 844             /*             #define INNERFLOPS INNERFLOPS+1 */
 845             /*         #endif */
 846             /*     #endif */
 847
 848             /*     #if 'Force' in KERNEL_VF */
 849
 850             /*         #if 'electrostatics' in INTERACTION_FLAGS[I][J] and 'vdw' in INTERACTION_FLAGS[I][J] */
 851             fscal            = _mm_add_ps(felec,fvdw);
 852             /*             #define INNERFLOPS INNERFLOPS+1 */
 853             /*         #elif 'electrostatics' in INTERACTION_FLAGS[I][J] */
 854             fscal            = felec;
 855             /*         #elif 'vdw' in INTERACTION_FLAGS[I][J] */
 856             fscal            = fvdw;
 857             /*        #endif */
 858
 859             /*     ## Note special check for TIP4P-TIP4P. Since we are cutting of all hydrogen interactions we also cut the LJ-only O-O interaction */
 860             /*     #if 'exactcutoff' in INTERACTION_FLAGS[I][J] or (GEOMETRY_I=='Water4' and GEOMETRY_J=='Water4' and 'exactcutoff' in INTERACTION_FLAGS[1][1]) */
 861             fscal            = _mm_and_ps(fscal,cutoff_mask);
 862             /*                 #define INNERFLOPS INNERFLOPS+1 */
 863             /*             #endif                                       */
 864
 865             /*             #if ROUND == 'Epilogue' */
 866             fscal            = _mm_andnot_ps(dummy_mask,fscal);
 867             /*             #endif */
 868
 869             /* Calculate temporary vectorial force */
 870             tx               = _mm_mul_ps(fscal,dx{I}{J});
 871             ty               = _mm_mul_ps(fscal,dy{I}{J});
 872             tz               = _mm_mul_ps(fscal,dz{I}{J});
 873
 874             /* Update vectorial force */
 875             fix{I}             = _mm_add_ps(fix{I},tx);
 876             fiy{I}             = _mm_add_ps(fiy{I},ty);
 877             fiz{I}             = _mm_add_ps(fiz{I},tz);
 878             /*             #define INNERFLOPS INNERFLOPS+6 */
 879
 880             /* #if GEOMETRY_I == 'Particle'             */
 881             /*     #if ROUND == 'Loop' */
 882             fjptrA             = f+j_coord_offsetA;
 883             fjptrB             = f+j_coord_offsetB;
 884             fjptrC             = f+j_coord_offsetC;
 885             fjptrD             = f+j_coord_offsetD;
 886             /*     #else */
 887             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 888             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 889             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 890             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 891             /*     #endif */
 892             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,tx,ty,tz);
 893             /*     #define INNERFLOPS INNERFLOPS+3      */
 894             /* #else                                    */
 895             fjx{J}             = _mm_add_ps(fjx{J},tx);
 896             fjy{J}             = _mm_add_ps(fjy{J},ty);
 897             fjz{J}             = _mm_add_ps(fjz{J},tz);
 898             /*     #define INNERFLOPS INNERFLOPS+3      */
 899             /* #endif                                   */
 900
 901             /*     #endif */
 902
 903             /*     ## Note special check for TIP4P-TIP4P. Since we are cutting of all hydrogen interactions we also cut the LJ-only O-O interaction */
 904             /*     #if 'exactcutoff' in INTERACTION_FLAGS[I][J] or (GEOMETRY_I=='Water4' and GEOMETRY_J=='Water4' and 'exactcutoff' in INTERACTION_FLAGS[1][1]) */
 905             /*         #if 0    ## This and next two lines is a hack to maintain indentation in template file */
 906             {
 907                 /*     #endif */
 908             }
 909             /*     #endif */
 910             /*    ## End of check for the interaction being outside the cutoff */
 911
 912             /* #endfor */
 913             /* ## End of loop over i-j interaction pairs */
 914
 915             /* #if GEOMETRY_I != 'Particle' */
 916             /*     #if ROUND == 'Loop' */
 917             fjptrA             = f+j_coord_offsetA;
 918             fjptrB             = f+j_coord_offsetB;
 919             fjptrC             = f+j_coord_offsetC;
 920             fjptrD             = f+j_coord_offsetD;
 921             /*     #else */
 922             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 923             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 924             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 925             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 926             /*     #endif */
 927             /* #endif */
 928
 929             /* #if 'Water' in GEOMETRY_I and GEOMETRY_J == 'Particle' */
 930             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
 931             /* #elif GEOMETRY_J == 'Water3'               */
 932             gmx_mm_decrement_3rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,
 933                                                    fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2);
 934             /*     #define INNERFLOPS INNERFLOPS+9      */
 935             /* #elif GEOMETRY_J == 'Water4'             */
 936             /*     #if 0 in PARTICLES_J                 */
 937             gmx_mm_decrement_4rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,
 938                                                    fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,
 939                                                    fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
 940             /*     #define INNERFLOPS INNERFLOPS+12     */
 941             /*     #else                                */
 942             gmx_mm_decrement_3rvec_4ptr_swizzle_ps(fjptrA+DIM,fjptrB+DIM,fjptrC+DIM,fjptrD+DIM,
 943                                                    fjx1,fjy1,fjz1,fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
 944             /*     #define INNERFLOPS INNERFLOPS+9      */
 945             /*     #endif                               */
 946             /* #endif                                   */
 947
 948             /* Inner loop uses {INNERFLOPS} flops */
 949         }
 950
 951         /* #endfor */
 952
 953         /* End of innermost loop */
 954
 955         /* #if 'Force' in KERNEL_VF */
 956         /*     #if GEOMETRY_I == 'Particle'            */
 957         gmx_mm_update_iforce_1atom_swizzle_ps(fix0,fiy0,fiz0,
 958                                               f+i_coord_offset,fshift+i_shift_offset);
 959         /*         #define OUTERFLOPS OUTERFLOPS+6     */
 960         /*     #elif GEOMETRY_I == 'Water3'            */
 961         gmx_mm_update_iforce_3atom_swizzle_ps(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,
 962                                               f+i_coord_offset,fshift+i_shift_offset);
 963         /*         #define OUTERFLOPS OUTERFLOPS+18    */
 964         /*     #elif GEOMETRY_I == 'Water4'            */
 965         /*         #if 0 in PARTICLES_I                */
 966         gmx_mm_update_iforce_4atom_swizzle_ps(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
 967                                               f+i_coord_offset,fshift+i_shift_offset);
 968         /*             #define OUTERFLOPS OUTERFLOPS+24    */
 969         /*         #else                               */
 970         gmx_mm_update_iforce_3atom_swizzle_ps(fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
 971                                               f+i_coord_offset+DIM,fshift+i_shift_offset);
 972         /*             #define OUTERFLOPS OUTERFLOPS+18    */
 973         /*         #endif                              */
 974         /*     #endif                                  */
 975         /* #endif                                      */
 976
 977         /* #if 'Potential' in KERNEL_VF */
 978         ggid                        = gid[iidx];
 979         /* Update potential energies */
 980         /*     #if KERNEL_ELEC != 'None' */
 981         gmx_mm_update_1pot_ps(velecsum,kernel_data->energygrp_elec+ggid);
 982         /*         #define OUTERFLOPS OUTERFLOPS+1 */
 983         /*     #endif */
 984         /*     #if KERNEL_VDW != 'None' */
 985         gmx_mm_update_1pot_ps(vvdwsum,kernel_data->energygrp_vdw+ggid);
 986         /*         #define OUTERFLOPS OUTERFLOPS+1 */
 987         /*     #endif */
 988         /* #endif */
 989
 990         /* Increment number of inner iterations */
 991         inneriter                  += j_index_end - j_index_start;
 992
 993         /* Outer loop uses {OUTERFLOPS} flops */
 994     }
 995
 996     /* Increment number of outer iterations */
 997     outeriter        += nri;
 998
 999     /* Update outer/inner flops */
1000     /* ## NB: This is not important, it just affects the flopcount. However, since our preprocessor is */
1001     /* ## primitive and replaces aggressively even in strings inside these directives, we need to      */
1002     /* ## assemble the main part of the name (containing KERNEL/ELEC/VDW) directly in the source.      */
1003     /* #if GEOMETRY_I == 'Water3'            */
1004     /*     #define ISUFFIX '_W3'             */
1005     /* #elif GEOMETRY_I == 'Water4'          */
1006     /*     #define ISUFFIX '_W4'             */
1007     /* #else                                 */
1008     /*     #define ISUFFIX ''                */
1009     /* #endif                                */
1010     /* #if GEOMETRY_J == 'Water3'            */
1011     /*     #define JSUFFIX 'W3'              */
1012     /* #elif GEOMETRY_J == 'Water4'          */
1013     /*     #define JSUFFIX 'W4'              */
1014     /* #else                                 */
1015     /*     #define JSUFFIX ''                */
1016     /* #endif                                */
1017     /* #if 'PotentialAndForce' in KERNEL_VF  */
1018     /*     #define VFSUFFIX  '_VF'           */
1019     /* #elif 'Potential' in KERNEL_VF        */
1020     /*     #define VFSUFFIX '_V'             */
1021     /* #else                                 */
1022     /*     #define VFSUFFIX '_F'             */
1023     /* #endif                                */
1024
1025     /* #if KERNEL_ELEC != 'None' and KERNEL_VDW != 'None' */
1026     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW{ISUFFIX}{JSUFFIX}{VFSUFFIX},outeriter*{OUTERFLOPS} + inneriter*{INNERFLOPS});
1027     /* #elif KERNEL_ELEC != 'None' */
1028     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC{ISUFFIX}{JSUFFIX}{VFSUFFIX},outeriter*{OUTERFLOPS} + inneriter*{INNERFLOPS});
1029     /* #else */
1030     inc_nrnb(nrnb,eNR_NBKERNEL_VDW{ISUFFIX}{JSUFFIX}{VFSUFFIX},outeriter*{OUTERFLOPS} + inneriter*{INNERFLOPS});
1031     /* #endif  */
1032 }