src/gromacs/gmxlib/nonbonded/nb_kernel_avx_128_fma_single/nb_kernel_ElecEw_VdwCSTab_GeomW3P1_avx_128_fma_single.c

   1 /*
   2  * This file is part of the GROMACS molecular simulation package.
   3  *
   4  * Copyright (c) 2012,2013,2014, by the GROMACS development team, led by
   5  * Mark Abraham, David van der Spoel, Berk Hess, and Erik Lindahl,
   6  * and including many others, as listed in the AUTHORS file in the
   7  * top-level source directory and at http://www.gromacs.org.
   8  *
   9  * GROMACS is free software; you can redistribute it and/or
  10  * modify it under the terms of the GNU Lesser General Public License
  11  * as published by the Free Software Foundation; either version 2.1
  12  * of the License, or (at your option) any later version.
  13  *
  14  * GROMACS is distributed in the hope that it will be useful,
  15  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  16  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  17  * Lesser General Public License for more details.
  18  *
  19  * You should have received a copy of the GNU Lesser General Public
  20  * License along with GROMACS; if not, see
  21  * http://www.gnu.org/licenses, or write to the Free Software Foundation,
  22  * Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA.
  23  *
  24  * If you want to redistribute modifications to GROMACS, please
  25  * consider that scientific software is very special. Version
  26  * control is crucial - bugs must be traceable. We will be happy to
  27  * consider code for inclusion in the official distribution, but
  28  * derived work must not be called official GROMACS. Details are found
  29  * in the README & COPYING files - if they are missing, get the
  30  * official version at http://www.gromacs.org.
  31  *
  32  * To help us fund GROMACS development, we humbly ask that you cite
  33  * the research papers on the package. Check out http://www.gromacs.org.
  34  */
  35 /*
  36  * Note: this file was generated by the GROMACS avx_128_fma_single kernel generator.
  37  */
  38 #include "config.h"
  39
  40 #include <math.h>
  41
  42 #include "../nb_kernel.h"
  43 #include "types/simple.h"
  44 #include "gromacs/math/vec.h"
  45 #include "nrnb.h"
  46
  47 #include "gromacs/simd/math_x86_avx_128_fma_single.h"
  48 #include "kernelutil_x86_avx_128_fma_single.h"
  49
  50 /*
  51  * Gromacs nonbonded kernel:   nb_kernel_ElecEw_VdwCSTab_GeomW3P1_VF_avx_128_fma_single
  52  * Electrostatics interaction: Ewald
  53  * VdW interaction:            CubicSplineTable
  54  * Geometry:                   Water3-Particle
  55  * Calculate force/pot:        PotentialAndForce
  56  */
  57 void
  58 nb_kernel_ElecEw_VdwCSTab_GeomW3P1_VF_avx_128_fma_single
  59                     (t_nblist                    * gmx_restrict       nlist,
  60                      rvec                        * gmx_restrict          xx,
  61                      rvec                        * gmx_restrict          ff,
  62                      t_forcerec                  * gmx_restrict          fr,
  63                      t_mdatoms                   * gmx_restrict     mdatoms,
  64                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
  65                      t_nrnb                      * gmx_restrict        nrnb)
  66 {
  67     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
  68      * just 0 for non-waters.
  69      * Suffixes A,B,C,D refer to j loop unrolling done with AVX_128, e.g. for the four different
  70      * jnr indices corresponding to data put in the four positions in the SIMD register.
  71      */
  72     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
  73     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
  74     int              jnrA,jnrB,jnrC,jnrD;
  75     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
  76     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
  77     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  78     real             rcutoff_scalar;
  79     real             *shiftvec,*fshift,*x,*f;
  80     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
  81     real             scratch[4*DIM];
  82     __m128           fscal,rcutoff,rcutoff2,jidxall;
  83     int              vdwioffset0;
  84     __m128           ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
  85     int              vdwioffset1;
  86     __m128           ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
  87     int              vdwioffset2;
  88     __m128           ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
  89     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
  90     __m128           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  91     __m128           dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
  92     __m128           dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
  93     __m128           dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
  94     __m128           velec,felec,velecsum,facel,crf,krf,krf2;
  95     real             *charge;
  96     int              nvdwtype;
  97     __m128           rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
  98     int              *vdwtype;
  99     real             *vdwparam;
 100     __m128           one_sixth   = _mm_set1_ps(1.0/6.0);
 101     __m128           one_twelfth = _mm_set1_ps(1.0/12.0);
 102     __m128i          vfitab;
 103     __m128i          ifour       = _mm_set1_epi32(4);
 104     __m128           rt,vfeps,twovfeps,vftabscale,Y,F,G,H,Fp,VV,FF;
 105     real             *vftab;
 106     __m128i          ewitab;
 107     __m128           ewtabscale,eweps,twoeweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
 108     __m128           beta,beta2,beta3,zeta2,pmecorrF,pmecorrV,rinv3;
 109     real             *ewtab;
 110     __m128           dummy_mask,cutoff_mask;
 111     __m128           signbit = _mm_castsi128_ps( _mm_set1_epi32(0x80000000) );
 112     __m128           one     = _mm_set1_ps(1.0);
 113     __m128           two     = _mm_set1_ps(2.0);
 114     x                = xx[0];
 115     f                = ff[0];
 116
 117     nri              = nlist->nri;
 118     iinr             = nlist->iinr;
 119     jindex           = nlist->jindex;
 120     jjnr             = nlist->jjnr;
 121     shiftidx         = nlist->shift;
 122     gid              = nlist->gid;
 123     shiftvec         = fr->shift_vec[0];
 124     fshift           = fr->fshift[0];
 125     facel            = _mm_set1_ps(fr->epsfac);
 126     charge           = mdatoms->chargeA;
 127     nvdwtype         = fr->ntype;
 128     vdwparam         = fr->nbfp;
 129     vdwtype          = mdatoms->typeA;
 130
 131     vftab            = kernel_data->table_vdw->data;
 132     vftabscale       = _mm_set1_ps(kernel_data->table_vdw->scale);
 133
 134     sh_ewald         = _mm_set1_ps(fr->ic->sh_ewald);
 135     beta             = _mm_set1_ps(fr->ic->ewaldcoeff_q);
 136     beta2            = _mm_mul_ps(beta,beta);
 137     beta3            = _mm_mul_ps(beta,beta2);
 138     ewtab            = fr->ic->tabq_coul_FDV0;
 139     ewtabscale       = _mm_set1_ps(fr->ic->tabq_scale);
 140     ewtabhalfspace   = _mm_set1_ps(0.5/fr->ic->tabq_scale);
 141
 142     /* Setup water-specific parameters */
 143     inr              = nlist->iinr[0];
 144     iq0              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+0]));
 145     iq1              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+1]));
 146     iq2              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+2]));
 147     vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 148
 149     /* Avoid stupid compiler warnings */
 150     jnrA = jnrB = jnrC = jnrD = 0;
 151     j_coord_offsetA = 0;
 152     j_coord_offsetB = 0;
 153     j_coord_offsetC = 0;
 154     j_coord_offsetD = 0;
 155
 156     outeriter        = 0;
 157     inneriter        = 0;
 158
 159     for(iidx=0;iidx<4*DIM;iidx++)
 160     {
 161         scratch[iidx] = 0.0;
 162     }
 163
 164     /* Start outer loop over neighborlists */
 165     for(iidx=0; iidx<nri; iidx++)
 166     {
 167         /* Load shift vector for this list */
 168         i_shift_offset   = DIM*shiftidx[iidx];
 169
 170         /* Load limits for loop over neighbors */
 171         j_index_start    = jindex[iidx];
 172         j_index_end      = jindex[iidx+1];
 173
 174         /* Get outer coordinate index */
 175         inr              = iinr[iidx];
 176         i_coord_offset   = DIM*inr;
 177
 178         /* Load i particle coords and add shift vector */
 179         gmx_mm_load_shift_and_3rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,
 180                                                  &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2);
 181
 182         fix0             = _mm_setzero_ps();
 183         fiy0             = _mm_setzero_ps();
 184         fiz0             = _mm_setzero_ps();
 185         fix1             = _mm_setzero_ps();
 186         fiy1             = _mm_setzero_ps();
 187         fiz1             = _mm_setzero_ps();
 188         fix2             = _mm_setzero_ps();
 189         fiy2             = _mm_setzero_ps();
 190         fiz2             = _mm_setzero_ps();
 191
 192         /* Reset potential sums */
 193         velecsum         = _mm_setzero_ps();
 194         vvdwsum          = _mm_setzero_ps();
 195
 196         /* Start inner kernel loop */
 197         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 198         {
 199
 200             /* Get j neighbor index, and coordinate index */
 201             jnrA             = jjnr[jidx];
 202             jnrB             = jjnr[jidx+1];
 203             jnrC             = jjnr[jidx+2];
 204             jnrD             = jjnr[jidx+3];
 205             j_coord_offsetA  = DIM*jnrA;
 206             j_coord_offsetB  = DIM*jnrB;
 207             j_coord_offsetC  = DIM*jnrC;
 208             j_coord_offsetD  = DIM*jnrD;
 209
 210             /* load j atom coordinates */
 211             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 212                                               x+j_coord_offsetC,x+j_coord_offsetD,
 213                                               &jx0,&jy0,&jz0);
 214
 215             /* Calculate displacement vector */
 216             dx00             = _mm_sub_ps(ix0,jx0);
 217             dy00             = _mm_sub_ps(iy0,jy0);
 218             dz00             = _mm_sub_ps(iz0,jz0);
 219             dx10             = _mm_sub_ps(ix1,jx0);
 220             dy10             = _mm_sub_ps(iy1,jy0);
 221             dz10             = _mm_sub_ps(iz1,jz0);
 222             dx20             = _mm_sub_ps(ix2,jx0);
 223             dy20             = _mm_sub_ps(iy2,jy0);
 224             dz20             = _mm_sub_ps(iz2,jz0);
 225
 226             /* Calculate squared distance and things based on it */
 227             rsq00            = gmx_mm_calc_rsq_ps(dx00,dy00,dz00);
 228             rsq10            = gmx_mm_calc_rsq_ps(dx10,dy10,dz10);
 229             rsq20            = gmx_mm_calc_rsq_ps(dx20,dy20,dz20);
 230
 231             rinv00           = gmx_mm_invsqrt_ps(rsq00);
 232             rinv10           = gmx_mm_invsqrt_ps(rsq10);
 233             rinv20           = gmx_mm_invsqrt_ps(rsq20);
 234
 235             rinvsq00         = _mm_mul_ps(rinv00,rinv00);
 236             rinvsq10         = _mm_mul_ps(rinv10,rinv10);
 237             rinvsq20         = _mm_mul_ps(rinv20,rinv20);
 238
 239             /* Load parameters for j particles */
 240             jq0              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 241                                                               charge+jnrC+0,charge+jnrD+0);
 242             vdwjidx0A        = 2*vdwtype[jnrA+0];
 243             vdwjidx0B        = 2*vdwtype[jnrB+0];
 244             vdwjidx0C        = 2*vdwtype[jnrC+0];
 245             vdwjidx0D        = 2*vdwtype[jnrD+0];
 246
 247             fjx0             = _mm_setzero_ps();
 248             fjy0             = _mm_setzero_ps();
 249             fjz0             = _mm_setzero_ps();
 250
 251             /**************************
 252              * CALCULATE INTERACTIONS *
 253              **************************/
 254
 255             r00              = _mm_mul_ps(rsq00,rinv00);
 256
 257             /* Compute parameters for interactions between i and j atoms */
 258             qq00             = _mm_mul_ps(iq0,jq0);
 259             gmx_mm_load_4pair_swizzle_ps(vdwparam+vdwioffset0+vdwjidx0A,
 260                                          vdwparam+vdwioffset0+vdwjidx0B,
 261                                          vdwparam+vdwioffset0+vdwjidx0C,
 262                                          vdwparam+vdwioffset0+vdwjidx0D,
 263                                          &c6_00,&c12_00);
 264
 265             /* Calculate table index by multiplying r with table scale and truncate to integer */
 266             rt               = _mm_mul_ps(r00,vftabscale);
 267             vfitab           = _mm_cvttps_epi32(rt);
 268 #ifdef __XOP__
 269             vfeps            = _mm_frcz_ps(rt);
 270 #else
 271             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
 272 #endif
 273             twovfeps         = _mm_add_ps(vfeps,vfeps);
 274             vfitab           = _mm_slli_epi32(vfitab,3);
 275
 276             /* EWALD ELECTROSTATICS */
 277
 278             /* Analytical PME correction */
 279             zeta2            = _mm_mul_ps(beta2,rsq00);
 280             rinv3            = _mm_mul_ps(rinvsq00,rinv00);
 281             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
 282             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
 283             felec            = _mm_mul_ps(qq00,felec);
 284             pmecorrV         = gmx_mm_pmecorrV_ps(zeta2);
 285             velec            = _mm_nmacc_ps(pmecorrV,beta,rinv00);
 286             velec            = _mm_mul_ps(qq00,velec);
 287
 288             /* CUBIC SPLINE TABLE DISPERSION */
 289             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 290             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 291             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 292             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 293             _MM_TRANSPOSE4_PS(Y,F,G,H);
 294             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 295             VV               = _mm_macc_ps(vfeps,Fp,Y);
 296             vvdw6            = _mm_mul_ps(c6_00,VV);
 297             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 298             fvdw6            = _mm_mul_ps(c6_00,FF);
 299
 300             /* CUBIC SPLINE TABLE REPULSION */
 301             vfitab           = _mm_add_epi32(vfitab,ifour);
 302             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 303             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 304             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 305             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 306             _MM_TRANSPOSE4_PS(Y,F,G,H);
 307             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 308             VV               = _mm_macc_ps(vfeps,Fp,Y);
 309             vvdw12           = _mm_mul_ps(c12_00,VV);
 310             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 311             fvdw12           = _mm_mul_ps(c12_00,FF);
 312             vvdw             = _mm_add_ps(vvdw12,vvdw6);
 313             fvdw             = _mm_xor_ps(signbit,_mm_mul_ps(_mm_add_ps(fvdw6,fvdw12),_mm_mul_ps(vftabscale,rinv00)));
 314
 315             /* Update potential sum for this i atom from the interaction with this j atom. */
 316             velecsum         = _mm_add_ps(velecsum,velec);
 317             vvdwsum          = _mm_add_ps(vvdwsum,vvdw);
 318
 319             fscal            = _mm_add_ps(felec,fvdw);
 320
 321              /* Update vectorial force */
 322             fix0             = _mm_macc_ps(dx00,fscal,fix0);
 323             fiy0             = _mm_macc_ps(dy00,fscal,fiy0);
 324             fiz0             = _mm_macc_ps(dz00,fscal,fiz0);
 325
 326             fjx0             = _mm_macc_ps(dx00,fscal,fjx0);
 327             fjy0             = _mm_macc_ps(dy00,fscal,fjy0);
 328             fjz0             = _mm_macc_ps(dz00,fscal,fjz0);
 329
 330             /**************************
 331              * CALCULATE INTERACTIONS *
 332              **************************/
 333
 334             r10              = _mm_mul_ps(rsq10,rinv10);
 335
 336             /* Compute parameters for interactions between i and j atoms */
 337             qq10             = _mm_mul_ps(iq1,jq0);
 338
 339             /* EWALD ELECTROSTATICS */
 340
 341             /* Analytical PME correction */
 342             zeta2            = _mm_mul_ps(beta2,rsq10);
 343             rinv3            = _mm_mul_ps(rinvsq10,rinv10);
 344             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
 345             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
 346             felec            = _mm_mul_ps(qq10,felec);
 347             pmecorrV         = gmx_mm_pmecorrV_ps(zeta2);
 348             velec            = _mm_nmacc_ps(pmecorrV,beta,rinv10);
 349             velec            = _mm_mul_ps(qq10,velec);
 350
 351             /* Update potential sum for this i atom from the interaction with this j atom. */
 352             velecsum         = _mm_add_ps(velecsum,velec);
 353
 354             fscal            = felec;
 355
 356              /* Update vectorial force */
 357             fix1             = _mm_macc_ps(dx10,fscal,fix1);
 358             fiy1             = _mm_macc_ps(dy10,fscal,fiy1);
 359             fiz1             = _mm_macc_ps(dz10,fscal,fiz1);
 360
 361             fjx0             = _mm_macc_ps(dx10,fscal,fjx0);
 362             fjy0             = _mm_macc_ps(dy10,fscal,fjy0);
 363             fjz0             = _mm_macc_ps(dz10,fscal,fjz0);
 364
 365             /**************************
 366              * CALCULATE INTERACTIONS *
 367              **************************/
 368
 369             r20              = _mm_mul_ps(rsq20,rinv20);
 370
 371             /* Compute parameters for interactions between i and j atoms */
 372             qq20             = _mm_mul_ps(iq2,jq0);
 373
 374             /* EWALD ELECTROSTATICS */
 375
 376             /* Analytical PME correction */
 377             zeta2            = _mm_mul_ps(beta2,rsq20);
 378             rinv3            = _mm_mul_ps(rinvsq20,rinv20);
 379             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
 380             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
 381             felec            = _mm_mul_ps(qq20,felec);
 382             pmecorrV         = gmx_mm_pmecorrV_ps(zeta2);
 383             velec            = _mm_nmacc_ps(pmecorrV,beta,rinv20);
 384             velec            = _mm_mul_ps(qq20,velec);
 385
 386             /* Update potential sum for this i atom from the interaction with this j atom. */
 387             velecsum         = _mm_add_ps(velecsum,velec);
 388
 389             fscal            = felec;
 390
 391              /* Update vectorial force */
 392             fix2             = _mm_macc_ps(dx20,fscal,fix2);
 393             fiy2             = _mm_macc_ps(dy20,fscal,fiy2);
 394             fiz2             = _mm_macc_ps(dz20,fscal,fiz2);
 395
 396             fjx0             = _mm_macc_ps(dx20,fscal,fjx0);
 397             fjy0             = _mm_macc_ps(dy20,fscal,fjy0);
 398             fjz0             = _mm_macc_ps(dz20,fscal,fjz0);
 399
 400             fjptrA             = f+j_coord_offsetA;
 401             fjptrB             = f+j_coord_offsetB;
 402             fjptrC             = f+j_coord_offsetC;
 403             fjptrD             = f+j_coord_offsetD;
 404
 405             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
 406
 407             /* Inner loop uses 121 flops */
 408         }
 409
 410         if(jidx<j_index_end)
 411         {
 412
 413             /* Get j neighbor index, and coordinate index */
 414             jnrlistA         = jjnr[jidx];
 415             jnrlistB         = jjnr[jidx+1];
 416             jnrlistC         = jjnr[jidx+2];
 417             jnrlistD         = jjnr[jidx+3];
 418             /* Sign of each element will be negative for non-real atoms.
 419              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 420              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
 421              */
 422             dummy_mask = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
 423             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 424             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 425             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 426             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 427             j_coord_offsetA  = DIM*jnrA;
 428             j_coord_offsetB  = DIM*jnrB;
 429             j_coord_offsetC  = DIM*jnrC;
 430             j_coord_offsetD  = DIM*jnrD;
 431
 432             /* load j atom coordinates */
 433             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 434                                               x+j_coord_offsetC,x+j_coord_offsetD,
 435                                               &jx0,&jy0,&jz0);
 436
 437             /* Calculate displacement vector */
 438             dx00             = _mm_sub_ps(ix0,jx0);
 439             dy00             = _mm_sub_ps(iy0,jy0);
 440             dz00             = _mm_sub_ps(iz0,jz0);
 441             dx10             = _mm_sub_ps(ix1,jx0);
 442             dy10             = _mm_sub_ps(iy1,jy0);
 443             dz10             = _mm_sub_ps(iz1,jz0);
 444             dx20             = _mm_sub_ps(ix2,jx0);
 445             dy20             = _mm_sub_ps(iy2,jy0);
 446             dz20             = _mm_sub_ps(iz2,jz0);
 447
 448             /* Calculate squared distance and things based on it */
 449             rsq00            = gmx_mm_calc_rsq_ps(dx00,dy00,dz00);
 450             rsq10            = gmx_mm_calc_rsq_ps(dx10,dy10,dz10);
 451             rsq20            = gmx_mm_calc_rsq_ps(dx20,dy20,dz20);
 452
 453             rinv00           = gmx_mm_invsqrt_ps(rsq00);
 454             rinv10           = gmx_mm_invsqrt_ps(rsq10);
 455             rinv20           = gmx_mm_invsqrt_ps(rsq20);
 456
 457             rinvsq00         = _mm_mul_ps(rinv00,rinv00);
 458             rinvsq10         = _mm_mul_ps(rinv10,rinv10);
 459             rinvsq20         = _mm_mul_ps(rinv20,rinv20);
 460
 461             /* Load parameters for j particles */
 462             jq0              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 463                                                               charge+jnrC+0,charge+jnrD+0);
 464             vdwjidx0A        = 2*vdwtype[jnrA+0];
 465             vdwjidx0B        = 2*vdwtype[jnrB+0];
 466             vdwjidx0C        = 2*vdwtype[jnrC+0];
 467             vdwjidx0D        = 2*vdwtype[jnrD+0];
 468
 469             fjx0             = _mm_setzero_ps();
 470             fjy0             = _mm_setzero_ps();
 471             fjz0             = _mm_setzero_ps();
 472
 473             /**************************
 474              * CALCULATE INTERACTIONS *
 475              **************************/
 476
 477             r00              = _mm_mul_ps(rsq00,rinv00);
 478             r00              = _mm_andnot_ps(dummy_mask,r00);
 479
 480             /* Compute parameters for interactions between i and j atoms */
 481             qq00             = _mm_mul_ps(iq0,jq0);
 482             gmx_mm_load_4pair_swizzle_ps(vdwparam+vdwioffset0+vdwjidx0A,
 483                                          vdwparam+vdwioffset0+vdwjidx0B,
 484                                          vdwparam+vdwioffset0+vdwjidx0C,
 485                                          vdwparam+vdwioffset0+vdwjidx0D,
 486                                          &c6_00,&c12_00);
 487
 488             /* Calculate table index by multiplying r with table scale and truncate to integer */
 489             rt               = _mm_mul_ps(r00,vftabscale);
 490             vfitab           = _mm_cvttps_epi32(rt);
 491 #ifdef __XOP__
 492             vfeps            = _mm_frcz_ps(rt);
 493 #else
 494             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
 495 #endif
 496             twovfeps         = _mm_add_ps(vfeps,vfeps);
 497             vfitab           = _mm_slli_epi32(vfitab,3);
 498
 499             /* EWALD ELECTROSTATICS */
 500
 501             /* Analytical PME correction */
 502             zeta2            = _mm_mul_ps(beta2,rsq00);
 503             rinv3            = _mm_mul_ps(rinvsq00,rinv00);
 504             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
 505             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
 506             felec            = _mm_mul_ps(qq00,felec);
 507             pmecorrV         = gmx_mm_pmecorrV_ps(zeta2);
 508             velec            = _mm_nmacc_ps(pmecorrV,beta,rinv00);
 509             velec            = _mm_mul_ps(qq00,velec);
 510
 511             /* CUBIC SPLINE TABLE DISPERSION */
 512             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 513             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 514             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 515             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 516             _MM_TRANSPOSE4_PS(Y,F,G,H);
 517             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 518             VV               = _mm_macc_ps(vfeps,Fp,Y);
 519             vvdw6            = _mm_mul_ps(c6_00,VV);
 520             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 521             fvdw6            = _mm_mul_ps(c6_00,FF);
 522
 523             /* CUBIC SPLINE TABLE REPULSION */
 524             vfitab           = _mm_add_epi32(vfitab,ifour);
 525             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 526             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 527             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 528             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 529             _MM_TRANSPOSE4_PS(Y,F,G,H);
 530             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 531             VV               = _mm_macc_ps(vfeps,Fp,Y);
 532             vvdw12           = _mm_mul_ps(c12_00,VV);
 533             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 534             fvdw12           = _mm_mul_ps(c12_00,FF);
 535             vvdw             = _mm_add_ps(vvdw12,vvdw6);
 536             fvdw             = _mm_xor_ps(signbit,_mm_mul_ps(_mm_add_ps(fvdw6,fvdw12),_mm_mul_ps(vftabscale,rinv00)));
 537
 538             /* Update potential sum for this i atom from the interaction with this j atom. */
 539             velec            = _mm_andnot_ps(dummy_mask,velec);
 540             velecsum         = _mm_add_ps(velecsum,velec);
 541             vvdw             = _mm_andnot_ps(dummy_mask,vvdw);
 542             vvdwsum          = _mm_add_ps(vvdwsum,vvdw);
 543
 544             fscal            = _mm_add_ps(felec,fvdw);
 545
 546             fscal            = _mm_andnot_ps(dummy_mask,fscal);
 547
 548              /* Update vectorial force */
 549             fix0             = _mm_macc_ps(dx00,fscal,fix0);
 550             fiy0             = _mm_macc_ps(dy00,fscal,fiy0);
 551             fiz0             = _mm_macc_ps(dz00,fscal,fiz0);
 552
 553             fjx0             = _mm_macc_ps(dx00,fscal,fjx0);
 554             fjy0             = _mm_macc_ps(dy00,fscal,fjy0);
 555             fjz0             = _mm_macc_ps(dz00,fscal,fjz0);
 556
 557             /**************************
 558              * CALCULATE INTERACTIONS *
 559              **************************/
 560
 561             r10              = _mm_mul_ps(rsq10,rinv10);
 562             r10              = _mm_andnot_ps(dummy_mask,r10);
 563
 564             /* Compute parameters for interactions between i and j atoms */
 565             qq10             = _mm_mul_ps(iq1,jq0);
 566
 567             /* EWALD ELECTROSTATICS */
 568
 569             /* Analytical PME correction */
 570             zeta2            = _mm_mul_ps(beta2,rsq10);
 571             rinv3            = _mm_mul_ps(rinvsq10,rinv10);
 572             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
 573             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
 574             felec            = _mm_mul_ps(qq10,felec);
 575             pmecorrV         = gmx_mm_pmecorrV_ps(zeta2);
 576             velec            = _mm_nmacc_ps(pmecorrV,beta,rinv10);
 577             velec            = _mm_mul_ps(qq10,velec);
 578
 579             /* Update potential sum for this i atom from the interaction with this j atom. */
 580             velec            = _mm_andnot_ps(dummy_mask,velec);
 581             velecsum         = _mm_add_ps(velecsum,velec);
 582
 583             fscal            = felec;
 584
 585             fscal            = _mm_andnot_ps(dummy_mask,fscal);
 586
 587              /* Update vectorial force */
 588             fix1             = _mm_macc_ps(dx10,fscal,fix1);
 589             fiy1             = _mm_macc_ps(dy10,fscal,fiy1);
 590             fiz1             = _mm_macc_ps(dz10,fscal,fiz1);
 591
 592             fjx0             = _mm_macc_ps(dx10,fscal,fjx0);
 593             fjy0             = _mm_macc_ps(dy10,fscal,fjy0);
 594             fjz0             = _mm_macc_ps(dz10,fscal,fjz0);
 595
 596             /**************************
 597              * CALCULATE INTERACTIONS *
 598              **************************/
 599
 600             r20              = _mm_mul_ps(rsq20,rinv20);
 601             r20              = _mm_andnot_ps(dummy_mask,r20);
 602
 603             /* Compute parameters for interactions between i and j atoms */
 604             qq20             = _mm_mul_ps(iq2,jq0);
 605
 606             /* EWALD ELECTROSTATICS */
 607
 608             /* Analytical PME correction */
 609             zeta2            = _mm_mul_ps(beta2,rsq20);
 610             rinv3            = _mm_mul_ps(rinvsq20,rinv20);
 611             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
 612             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
 613             felec            = _mm_mul_ps(qq20,felec);
 614             pmecorrV         = gmx_mm_pmecorrV_ps(zeta2);
 615             velec            = _mm_nmacc_ps(pmecorrV,beta,rinv20);
 616             velec            = _mm_mul_ps(qq20,velec);
 617
 618             /* Update potential sum for this i atom from the interaction with this j atom. */
 619             velec            = _mm_andnot_ps(dummy_mask,velec);
 620             velecsum         = _mm_add_ps(velecsum,velec);
 621
 622             fscal            = felec;
 623
 624             fscal            = _mm_andnot_ps(dummy_mask,fscal);
 625
 626              /* Update vectorial force */
 627             fix2             = _mm_macc_ps(dx20,fscal,fix2);
 628             fiy2             = _mm_macc_ps(dy20,fscal,fiy2);
 629             fiz2             = _mm_macc_ps(dz20,fscal,fiz2);
 630
 631             fjx0             = _mm_macc_ps(dx20,fscal,fjx0);
 632             fjy0             = _mm_macc_ps(dy20,fscal,fjy0);
 633             fjz0             = _mm_macc_ps(dz20,fscal,fjz0);
 634
 635             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 636             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 637             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 638             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 639
 640             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
 641
 642             /* Inner loop uses 124 flops */
 643         }
 644
 645         /* End of innermost loop */
 646
 647         gmx_mm_update_iforce_3atom_swizzle_ps(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,
 648                                               f+i_coord_offset,fshift+i_shift_offset);
 649
 650         ggid                        = gid[iidx];
 651         /* Update potential energies */
 652         gmx_mm_update_1pot_ps(velecsum,kernel_data->energygrp_elec+ggid);
 653         gmx_mm_update_1pot_ps(vvdwsum,kernel_data->energygrp_vdw+ggid);
 654
 655         /* Increment number of inner iterations */
 656         inneriter                  += j_index_end - j_index_start;
 657
 658         /* Outer loop uses 20 flops */
 659     }
 660
 661     /* Increment number of outer iterations */
 662     outeriter        += nri;
 663
 664     /* Update outer/inner flops */
 665
 666     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W3_VF,outeriter*20 + inneriter*124);
 667 }
 668 /*
 669  * Gromacs nonbonded kernel:   nb_kernel_ElecEw_VdwCSTab_GeomW3P1_F_avx_128_fma_single
 670  * Electrostatics interaction: Ewald
 671  * VdW interaction:            CubicSplineTable
 672  * Geometry:                   Water3-Particle
 673  * Calculate force/pot:        Force
 674  */
 675 void
 676 nb_kernel_ElecEw_VdwCSTab_GeomW3P1_F_avx_128_fma_single
 677                     (t_nblist                    * gmx_restrict       nlist,
 678                      rvec                        * gmx_restrict          xx,
 679                      rvec                        * gmx_restrict          ff,
 680                      t_forcerec                  * gmx_restrict          fr,
 681                      t_mdatoms                   * gmx_restrict     mdatoms,
 682                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 683                      t_nrnb                      * gmx_restrict        nrnb)
 684 {
 685     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
 686      * just 0 for non-waters.
 687      * Suffixes A,B,C,D refer to j loop unrolling done with AVX_128, e.g. for the four different
 688      * jnr indices corresponding to data put in the four positions in the SIMD register.
 689      */
 690     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
 691     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
 692     int              jnrA,jnrB,jnrC,jnrD;
 693     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
 694     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
 695     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 696     real             rcutoff_scalar;
 697     real             *shiftvec,*fshift,*x,*f;
 698     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
 699     real             scratch[4*DIM];
 700     __m128           fscal,rcutoff,rcutoff2,jidxall;
 701     int              vdwioffset0;
 702     __m128           ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 703     int              vdwioffset1;
 704     __m128           ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
 705     int              vdwioffset2;
 706     __m128           ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
 707     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
 708     __m128           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 709     __m128           dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
 710     __m128           dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
 711     __m128           dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
 712     __m128           velec,felec,velecsum,facel,crf,krf,krf2;
 713     real             *charge;
 714     int              nvdwtype;
 715     __m128           rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 716     int              *vdwtype;
 717     real             *vdwparam;
 718     __m128           one_sixth   = _mm_set1_ps(1.0/6.0);
 719     __m128           one_twelfth = _mm_set1_ps(1.0/12.0);
 720     __m128i          vfitab;
 721     __m128i          ifour       = _mm_set1_epi32(4);
 722     __m128           rt,vfeps,twovfeps,vftabscale,Y,F,G,H,Fp,VV,FF;
 723     real             *vftab;
 724     __m128i          ewitab;
 725     __m128           ewtabscale,eweps,twoeweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
 726     __m128           beta,beta2,beta3,zeta2,pmecorrF,pmecorrV,rinv3;
 727     real             *ewtab;
 728     __m128           dummy_mask,cutoff_mask;
 729     __m128           signbit = _mm_castsi128_ps( _mm_set1_epi32(0x80000000) );
 730     __m128           one     = _mm_set1_ps(1.0);
 731     __m128           two     = _mm_set1_ps(2.0);
 732     x                = xx[0];
 733     f                = ff[0];
 734
 735     nri              = nlist->nri;
 736     iinr             = nlist->iinr;
 737     jindex           = nlist->jindex;
 738     jjnr             = nlist->jjnr;
 739     shiftidx         = nlist->shift;
 740     gid              = nlist->gid;
 741     shiftvec         = fr->shift_vec[0];
 742     fshift           = fr->fshift[0];
 743     facel            = _mm_set1_ps(fr->epsfac);
 744     charge           = mdatoms->chargeA;
 745     nvdwtype         = fr->ntype;
 746     vdwparam         = fr->nbfp;
 747     vdwtype          = mdatoms->typeA;
 748
 749     vftab            = kernel_data->table_vdw->data;
 750     vftabscale       = _mm_set1_ps(kernel_data->table_vdw->scale);
 751
 752     sh_ewald         = _mm_set1_ps(fr->ic->sh_ewald);
 753     beta             = _mm_set1_ps(fr->ic->ewaldcoeff_q);
 754     beta2            = _mm_mul_ps(beta,beta);
 755     beta3            = _mm_mul_ps(beta,beta2);
 756     ewtab            = fr->ic->tabq_coul_F;
 757     ewtabscale       = _mm_set1_ps(fr->ic->tabq_scale);
 758     ewtabhalfspace   = _mm_set1_ps(0.5/fr->ic->tabq_scale);
 759
 760     /* Setup water-specific parameters */
 761     inr              = nlist->iinr[0];
 762     iq0              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+0]));
 763     iq1              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+1]));
 764     iq2              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+2]));
 765     vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 766
 767     /* Avoid stupid compiler warnings */
 768     jnrA = jnrB = jnrC = jnrD = 0;
 769     j_coord_offsetA = 0;
 770     j_coord_offsetB = 0;
 771     j_coord_offsetC = 0;
 772     j_coord_offsetD = 0;
 773
 774     outeriter        = 0;
 775     inneriter        = 0;
 776
 777     for(iidx=0;iidx<4*DIM;iidx++)
 778     {
 779         scratch[iidx] = 0.0;
 780     }
 781
 782     /* Start outer loop over neighborlists */
 783     for(iidx=0; iidx<nri; iidx++)
 784     {
 785         /* Load shift vector for this list */
 786         i_shift_offset   = DIM*shiftidx[iidx];
 787
 788         /* Load limits for loop over neighbors */
 789         j_index_start    = jindex[iidx];
 790         j_index_end      = jindex[iidx+1];
 791
 792         /* Get outer coordinate index */
 793         inr              = iinr[iidx];
 794         i_coord_offset   = DIM*inr;
 795
 796         /* Load i particle coords and add shift vector */
 797         gmx_mm_load_shift_and_3rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,
 798                                                  &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2);
 799
 800         fix0             = _mm_setzero_ps();
 801         fiy0             = _mm_setzero_ps();
 802         fiz0             = _mm_setzero_ps();
 803         fix1             = _mm_setzero_ps();
 804         fiy1             = _mm_setzero_ps();
 805         fiz1             = _mm_setzero_ps();
 806         fix2             = _mm_setzero_ps();
 807         fiy2             = _mm_setzero_ps();
 808         fiz2             = _mm_setzero_ps();
 809
 810         /* Start inner kernel loop */
 811         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 812         {
 813
 814             /* Get j neighbor index, and coordinate index */
 815             jnrA             = jjnr[jidx];
 816             jnrB             = jjnr[jidx+1];
 817             jnrC             = jjnr[jidx+2];
 818             jnrD             = jjnr[jidx+3];
 819             j_coord_offsetA  = DIM*jnrA;
 820             j_coord_offsetB  = DIM*jnrB;
 821             j_coord_offsetC  = DIM*jnrC;
 822             j_coord_offsetD  = DIM*jnrD;
 823
 824             /* load j atom coordinates */
 825             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 826                                               x+j_coord_offsetC,x+j_coord_offsetD,
 827                                               &jx0,&jy0,&jz0);
 828
 829             /* Calculate displacement vector */
 830             dx00             = _mm_sub_ps(ix0,jx0);
 831             dy00             = _mm_sub_ps(iy0,jy0);
 832             dz00             = _mm_sub_ps(iz0,jz0);
 833             dx10             = _mm_sub_ps(ix1,jx0);
 834             dy10             = _mm_sub_ps(iy1,jy0);
 835             dz10             = _mm_sub_ps(iz1,jz0);
 836             dx20             = _mm_sub_ps(ix2,jx0);
 837             dy20             = _mm_sub_ps(iy2,jy0);
 838             dz20             = _mm_sub_ps(iz2,jz0);
 839
 840             /* Calculate squared distance and things based on it */
 841             rsq00            = gmx_mm_calc_rsq_ps(dx00,dy00,dz00);
 842             rsq10            = gmx_mm_calc_rsq_ps(dx10,dy10,dz10);
 843             rsq20            = gmx_mm_calc_rsq_ps(dx20,dy20,dz20);
 844
 845             rinv00           = gmx_mm_invsqrt_ps(rsq00);
 846             rinv10           = gmx_mm_invsqrt_ps(rsq10);
 847             rinv20           = gmx_mm_invsqrt_ps(rsq20);
 848
 849             rinvsq00         = _mm_mul_ps(rinv00,rinv00);
 850             rinvsq10         = _mm_mul_ps(rinv10,rinv10);
 851             rinvsq20         = _mm_mul_ps(rinv20,rinv20);
 852
 853             /* Load parameters for j particles */
 854             jq0              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 855                                                               charge+jnrC+0,charge+jnrD+0);
 856             vdwjidx0A        = 2*vdwtype[jnrA+0];
 857             vdwjidx0B        = 2*vdwtype[jnrB+0];
 858             vdwjidx0C        = 2*vdwtype[jnrC+0];
 859             vdwjidx0D        = 2*vdwtype[jnrD+0];
 860
 861             fjx0             = _mm_setzero_ps();
 862             fjy0             = _mm_setzero_ps();
 863             fjz0             = _mm_setzero_ps();
 864
 865             /**************************
 866              * CALCULATE INTERACTIONS *
 867              **************************/
 868
 869             r00              = _mm_mul_ps(rsq00,rinv00);
 870
 871             /* Compute parameters for interactions between i and j atoms */
 872             qq00             = _mm_mul_ps(iq0,jq0);
 873             gmx_mm_load_4pair_swizzle_ps(vdwparam+vdwioffset0+vdwjidx0A,
 874                                          vdwparam+vdwioffset0+vdwjidx0B,
 875                                          vdwparam+vdwioffset0+vdwjidx0C,
 876                                          vdwparam+vdwioffset0+vdwjidx0D,
 877                                          &c6_00,&c12_00);
 878
 879             /* Calculate table index by multiplying r with table scale and truncate to integer */
 880             rt               = _mm_mul_ps(r00,vftabscale);
 881             vfitab           = _mm_cvttps_epi32(rt);
 882 #ifdef __XOP__
 883             vfeps            = _mm_frcz_ps(rt);
 884 #else
 885             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
 886 #endif
 887             twovfeps         = _mm_add_ps(vfeps,vfeps);
 888             vfitab           = _mm_slli_epi32(vfitab,3);
 889
 890             /* EWALD ELECTROSTATICS */
 891
 892             /* Analytical PME correction */
 893             zeta2            = _mm_mul_ps(beta2,rsq00);
 894             rinv3            = _mm_mul_ps(rinvsq00,rinv00);
 895             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
 896             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
 897             felec            = _mm_mul_ps(qq00,felec);
 898
 899             /* CUBIC SPLINE TABLE DISPERSION */
 900             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 901             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 902             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 903             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 904             _MM_TRANSPOSE4_PS(Y,F,G,H);
 905             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 906             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 907             fvdw6            = _mm_mul_ps(c6_00,FF);
 908
 909             /* CUBIC SPLINE TABLE REPULSION */
 910             vfitab           = _mm_add_epi32(vfitab,ifour);
 911             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
 912             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
 913             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
 914             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
 915             _MM_TRANSPOSE4_PS(Y,F,G,H);
 916             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
 917             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
 918             fvdw12           = _mm_mul_ps(c12_00,FF);
 919             fvdw             = _mm_xor_ps(signbit,_mm_mul_ps(_mm_add_ps(fvdw6,fvdw12),_mm_mul_ps(vftabscale,rinv00)));
 920
 921             fscal            = _mm_add_ps(felec,fvdw);
 922
 923              /* Update vectorial force */
 924             fix0             = _mm_macc_ps(dx00,fscal,fix0);
 925             fiy0             = _mm_macc_ps(dy00,fscal,fiy0);
 926             fiz0             = _mm_macc_ps(dz00,fscal,fiz0);
 927
 928             fjx0             = _mm_macc_ps(dx00,fscal,fjx0);
 929             fjy0             = _mm_macc_ps(dy00,fscal,fjy0);
 930             fjz0             = _mm_macc_ps(dz00,fscal,fjz0);
 931
 932             /**************************
 933              * CALCULATE INTERACTIONS *
 934              **************************/
 935
 936             r10              = _mm_mul_ps(rsq10,rinv10);
 937
 938             /* Compute parameters for interactions between i and j atoms */
 939             qq10             = _mm_mul_ps(iq1,jq0);
 940
 941             /* EWALD ELECTROSTATICS */
 942
 943             /* Analytical PME correction */
 944             zeta2            = _mm_mul_ps(beta2,rsq10);
 945             rinv3            = _mm_mul_ps(rinvsq10,rinv10);
 946             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
 947             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
 948             felec            = _mm_mul_ps(qq10,felec);
 949
 950             fscal            = felec;
 951
 952              /* Update vectorial force */
 953             fix1             = _mm_macc_ps(dx10,fscal,fix1);
 954             fiy1             = _mm_macc_ps(dy10,fscal,fiy1);
 955             fiz1             = _mm_macc_ps(dz10,fscal,fiz1);
 956
 957             fjx0             = _mm_macc_ps(dx10,fscal,fjx0);
 958             fjy0             = _mm_macc_ps(dy10,fscal,fjy0);
 959             fjz0             = _mm_macc_ps(dz10,fscal,fjz0);
 960
 961             /**************************
 962              * CALCULATE INTERACTIONS *
 963              **************************/
 964
 965             r20              = _mm_mul_ps(rsq20,rinv20);
 966
 967             /* Compute parameters for interactions between i and j atoms */
 968             qq20             = _mm_mul_ps(iq2,jq0);
 969
 970             /* EWALD ELECTROSTATICS */
 971
 972             /* Analytical PME correction */
 973             zeta2            = _mm_mul_ps(beta2,rsq20);
 974             rinv3            = _mm_mul_ps(rinvsq20,rinv20);
 975             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
 976             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
 977             felec            = _mm_mul_ps(qq20,felec);
 978
 979             fscal            = felec;
 980
 981              /* Update vectorial force */
 982             fix2             = _mm_macc_ps(dx20,fscal,fix2);
 983             fiy2             = _mm_macc_ps(dy20,fscal,fiy2);
 984             fiz2             = _mm_macc_ps(dz20,fscal,fiz2);
 985
 986             fjx0             = _mm_macc_ps(dx20,fscal,fjx0);
 987             fjy0             = _mm_macc_ps(dy20,fscal,fjy0);
 988             fjz0             = _mm_macc_ps(dz20,fscal,fjz0);
 989
 990             fjptrA             = f+j_coord_offsetA;
 991             fjptrB             = f+j_coord_offsetB;
 992             fjptrC             = f+j_coord_offsetC;
 993             fjptrD             = f+j_coord_offsetD;
 994
 995             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
 996
 997             /* Inner loop uses 110 flops */
 998         }
 999
1000         if(jidx<j_index_end)
1001         {
1002
1003             /* Get j neighbor index, and coordinate index */
1004             jnrlistA         = jjnr[jidx];
1005             jnrlistB         = jjnr[jidx+1];
1006             jnrlistC         = jjnr[jidx+2];
1007             jnrlistD         = jjnr[jidx+3];
1008             /* Sign of each element will be negative for non-real atoms.
1009              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
1010              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
1011              */
1012             dummy_mask = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
1013             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
1014             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
1015             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
1016             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
1017             j_coord_offsetA  = DIM*jnrA;
1018             j_coord_offsetB  = DIM*jnrB;
1019             j_coord_offsetC  = DIM*jnrC;
1020             j_coord_offsetD  = DIM*jnrD;
1021
1022             /* load j atom coordinates */
1023             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
1024                                               x+j_coord_offsetC,x+j_coord_offsetD,
1025                                               &jx0,&jy0,&jz0);
1026
1027             /* Calculate displacement vector */
1028             dx00             = _mm_sub_ps(ix0,jx0);
1029             dy00             = _mm_sub_ps(iy0,jy0);
1030             dz00             = _mm_sub_ps(iz0,jz0);
1031             dx10             = _mm_sub_ps(ix1,jx0);
1032             dy10             = _mm_sub_ps(iy1,jy0);
1033             dz10             = _mm_sub_ps(iz1,jz0);
1034             dx20             = _mm_sub_ps(ix2,jx0);
1035             dy20             = _mm_sub_ps(iy2,jy0);
1036             dz20             = _mm_sub_ps(iz2,jz0);
1037
1038             /* Calculate squared distance and things based on it */
1039             rsq00            = gmx_mm_calc_rsq_ps(dx00,dy00,dz00);
1040             rsq10            = gmx_mm_calc_rsq_ps(dx10,dy10,dz10);
1041             rsq20            = gmx_mm_calc_rsq_ps(dx20,dy20,dz20);
1042
1043             rinv00           = gmx_mm_invsqrt_ps(rsq00);
1044             rinv10           = gmx_mm_invsqrt_ps(rsq10);
1045             rinv20           = gmx_mm_invsqrt_ps(rsq20);
1046
1047             rinvsq00         = _mm_mul_ps(rinv00,rinv00);
1048             rinvsq10         = _mm_mul_ps(rinv10,rinv10);
1049             rinvsq20         = _mm_mul_ps(rinv20,rinv20);
1050
1051             /* Load parameters for j particles */
1052             jq0              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
1053                                                               charge+jnrC+0,charge+jnrD+0);
1054             vdwjidx0A        = 2*vdwtype[jnrA+0];
1055             vdwjidx0B        = 2*vdwtype[jnrB+0];
1056             vdwjidx0C        = 2*vdwtype[jnrC+0];
1057             vdwjidx0D        = 2*vdwtype[jnrD+0];
1058
1059             fjx0             = _mm_setzero_ps();
1060             fjy0             = _mm_setzero_ps();
1061             fjz0             = _mm_setzero_ps();
1062
1063             /**************************
1064              * CALCULATE INTERACTIONS *
1065              **************************/
1066
1067             r00              = _mm_mul_ps(rsq00,rinv00);
1068             r00              = _mm_andnot_ps(dummy_mask,r00);
1069
1070             /* Compute parameters for interactions between i and j atoms */
1071             qq00             = _mm_mul_ps(iq0,jq0);
1072             gmx_mm_load_4pair_swizzle_ps(vdwparam+vdwioffset0+vdwjidx0A,
1073                                          vdwparam+vdwioffset0+vdwjidx0B,
1074                                          vdwparam+vdwioffset0+vdwjidx0C,
1075                                          vdwparam+vdwioffset0+vdwjidx0D,
1076                                          &c6_00,&c12_00);
1077
1078             /* Calculate table index by multiplying r with table scale and truncate to integer */
1079             rt               = _mm_mul_ps(r00,vftabscale);
1080             vfitab           = _mm_cvttps_epi32(rt);
1081 #ifdef __XOP__
1082             vfeps            = _mm_frcz_ps(rt);
1083 #else
1084             vfeps            = _mm_sub_ps(rt,_mm_round_ps(rt, _MM_FROUND_FLOOR));
1085 #endif
1086             twovfeps         = _mm_add_ps(vfeps,vfeps);
1087             vfitab           = _mm_slli_epi32(vfitab,3);
1088
1089             /* EWALD ELECTROSTATICS */
1090
1091             /* Analytical PME correction */
1092             zeta2            = _mm_mul_ps(beta2,rsq00);
1093             rinv3            = _mm_mul_ps(rinvsq00,rinv00);
1094             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
1095             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
1096             felec            = _mm_mul_ps(qq00,felec);
1097
1098             /* CUBIC SPLINE TABLE DISPERSION */
1099             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
1100             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
1101             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
1102             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
1103             _MM_TRANSPOSE4_PS(Y,F,G,H);
1104             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
1105             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
1106             fvdw6            = _mm_mul_ps(c6_00,FF);
1107
1108             /* CUBIC SPLINE TABLE REPULSION */
1109             vfitab           = _mm_add_epi32(vfitab,ifour);
1110             Y                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,0) );
1111             F                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,1) );
1112             G                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,2) );
1113             H                = _mm_load_ps( vftab + _mm_extract_epi32(vfitab,3) );
1114             _MM_TRANSPOSE4_PS(Y,F,G,H);
1115             Fp               = _mm_macc_ps(vfeps,_mm_macc_ps(H,vfeps,G),F);
1116             FF               = _mm_macc_ps(vfeps,_mm_macc_ps(twovfeps,H,G),Fp);
1117             fvdw12           = _mm_mul_ps(c12_00,FF);
1118             fvdw             = _mm_xor_ps(signbit,_mm_mul_ps(_mm_add_ps(fvdw6,fvdw12),_mm_mul_ps(vftabscale,rinv00)));
1119
1120             fscal            = _mm_add_ps(felec,fvdw);
1121
1122             fscal            = _mm_andnot_ps(dummy_mask,fscal);
1123
1124              /* Update vectorial force */
1125             fix0             = _mm_macc_ps(dx00,fscal,fix0);
1126             fiy0             = _mm_macc_ps(dy00,fscal,fiy0);
1127             fiz0             = _mm_macc_ps(dz00,fscal,fiz0);
1128
1129             fjx0             = _mm_macc_ps(dx00,fscal,fjx0);
1130             fjy0             = _mm_macc_ps(dy00,fscal,fjy0);
1131             fjz0             = _mm_macc_ps(dz00,fscal,fjz0);
1132
1133             /**************************
1134              * CALCULATE INTERACTIONS *
1135              **************************/
1136
1137             r10              = _mm_mul_ps(rsq10,rinv10);
1138             r10              = _mm_andnot_ps(dummy_mask,r10);
1139
1140             /* Compute parameters for interactions between i and j atoms */
1141             qq10             = _mm_mul_ps(iq1,jq0);
1142
1143             /* EWALD ELECTROSTATICS */
1144
1145             /* Analytical PME correction */
1146             zeta2            = _mm_mul_ps(beta2,rsq10);
1147             rinv3            = _mm_mul_ps(rinvsq10,rinv10);
1148             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
1149             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
1150             felec            = _mm_mul_ps(qq10,felec);
1151
1152             fscal            = felec;
1153
1154             fscal            = _mm_andnot_ps(dummy_mask,fscal);
1155
1156              /* Update vectorial force */
1157             fix1             = _mm_macc_ps(dx10,fscal,fix1);
1158             fiy1             = _mm_macc_ps(dy10,fscal,fiy1);
1159             fiz1             = _mm_macc_ps(dz10,fscal,fiz1);
1160
1161             fjx0             = _mm_macc_ps(dx10,fscal,fjx0);
1162             fjy0             = _mm_macc_ps(dy10,fscal,fjy0);
1163             fjz0             = _mm_macc_ps(dz10,fscal,fjz0);
1164
1165             /**************************
1166              * CALCULATE INTERACTIONS *
1167              **************************/
1168
1169             r20              = _mm_mul_ps(rsq20,rinv20);
1170             r20              = _mm_andnot_ps(dummy_mask,r20);
1171
1172             /* Compute parameters for interactions between i and j atoms */
1173             qq20             = _mm_mul_ps(iq2,jq0);
1174
1175             /* EWALD ELECTROSTATICS */
1176
1177             /* Analytical PME correction */
1178             zeta2            = _mm_mul_ps(beta2,rsq20);
1179             rinv3            = _mm_mul_ps(rinvsq20,rinv20);
1180             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
1181             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
1182             felec            = _mm_mul_ps(qq20,felec);
1183
1184             fscal            = felec;
1185
1186             fscal            = _mm_andnot_ps(dummy_mask,fscal);
1187
1188              /* Update vectorial force */
1189             fix2             = _mm_macc_ps(dx20,fscal,fix2);
1190             fiy2             = _mm_macc_ps(dy20,fscal,fiy2);
1191             fiz2             = _mm_macc_ps(dz20,fscal,fiz2);
1192
1193             fjx0             = _mm_macc_ps(dx20,fscal,fjx0);
1194             fjy0             = _mm_macc_ps(dy20,fscal,fjy0);
1195             fjz0             = _mm_macc_ps(dz20,fscal,fjz0);
1196
1197             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
1198             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
1199             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
1200             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
1201
1202             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
1203
1204             /* Inner loop uses 113 flops */
1205         }
1206
1207         /* End of innermost loop */
1208
1209         gmx_mm_update_iforce_3atom_swizzle_ps(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,
1210                                               f+i_coord_offset,fshift+i_shift_offset);
1211
1212         /* Increment number of inner iterations */
1213         inneriter                  += j_index_end - j_index_start;
1214
1215         /* Outer loop uses 18 flops */
1216     }
1217
1218     /* Increment number of outer iterations */
1219     outeriter        += nri;
1220
1221     /* Update outer/inner flops */
1222
1223     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W3_F,outeriter*18 + inneriter*113);
1224 }