src/gromacs/gmxlib/nonbonded/nb_kernel_avx_128_fma_single/nb_kernel_ElecEwSh_VdwLJSh_GeomW4P1_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_ElecEwSh_VdwLJSh_GeomW4P1_VF_avx_128_fma_single
  52  * Electrostatics interaction: Ewald
  53  * VdW interaction:            LennardJones
  54  * Geometry:                   Water4-Particle
  55  * Calculate force/pot:        PotentialAndForce
  56  */
  57 void
  58 nb_kernel_ElecEwSh_VdwLJSh_GeomW4P1_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              vdwioffset3;
  90     __m128           ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
  91     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
  92     __m128           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  93     __m128           dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
  94     __m128           dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
  95     __m128           dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
  96     __m128           dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30;
  97     __m128           velec,felec,velecsum,facel,crf,krf,krf2;
  98     real             *charge;
  99     int              nvdwtype;
 100     __m128           rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 101     int              *vdwtype;
 102     real             *vdwparam;
 103     __m128           one_sixth   = _mm_set1_ps(1.0/6.0);
 104     __m128           one_twelfth = _mm_set1_ps(1.0/12.0);
 105     __m128i          ewitab;
 106     __m128           ewtabscale,eweps,twoeweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
 107     __m128           beta,beta2,beta3,zeta2,pmecorrF,pmecorrV,rinv3;
 108     real             *ewtab;
 109     __m128           dummy_mask,cutoff_mask;
 110     __m128           signbit = _mm_castsi128_ps( _mm_set1_epi32(0x80000000) );
 111     __m128           one     = _mm_set1_ps(1.0);
 112     __m128           two     = _mm_set1_ps(2.0);
 113     x                = xx[0];
 114     f                = ff[0];
 115
 116     nri              = nlist->nri;
 117     iinr             = nlist->iinr;
 118     jindex           = nlist->jindex;
 119     jjnr             = nlist->jjnr;
 120     shiftidx         = nlist->shift;
 121     gid              = nlist->gid;
 122     shiftvec         = fr->shift_vec[0];
 123     fshift           = fr->fshift[0];
 124     facel            = _mm_set1_ps(fr->epsfac);
 125     charge           = mdatoms->chargeA;
 126     nvdwtype         = fr->ntype;
 127     vdwparam         = fr->nbfp;
 128     vdwtype          = mdatoms->typeA;
 129
 130     sh_ewald         = _mm_set1_ps(fr->ic->sh_ewald);
 131     beta             = _mm_set1_ps(fr->ic->ewaldcoeff_q);
 132     beta2            = _mm_mul_ps(beta,beta);
 133     beta3            = _mm_mul_ps(beta,beta2);
 134     ewtab            = fr->ic->tabq_coul_FDV0;
 135     ewtabscale       = _mm_set1_ps(fr->ic->tabq_scale);
 136     ewtabhalfspace   = _mm_set1_ps(0.5/fr->ic->tabq_scale);
 137
 138     /* Setup water-specific parameters */
 139     inr              = nlist->iinr[0];
 140     iq1              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+1]));
 141     iq2              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+2]));
 142     iq3              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+3]));
 143     vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 144
 145     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
 146     rcutoff_scalar   = fr->rcoulomb;
 147     rcutoff          = _mm_set1_ps(rcutoff_scalar);
 148     rcutoff2         = _mm_mul_ps(rcutoff,rcutoff);
 149
 150     sh_vdw_invrcut6  = _mm_set1_ps(fr->ic->sh_invrc6);
 151     rvdw             = _mm_set1_ps(fr->rvdw);
 152
 153     /* Avoid stupid compiler warnings */
 154     jnrA = jnrB = jnrC = jnrD = 0;
 155     j_coord_offsetA = 0;
 156     j_coord_offsetB = 0;
 157     j_coord_offsetC = 0;
 158     j_coord_offsetD = 0;
 159
 160     outeriter        = 0;
 161     inneriter        = 0;
 162
 163     for(iidx=0;iidx<4*DIM;iidx++)
 164     {
 165         scratch[iidx] = 0.0;
 166     }
 167
 168     /* Start outer loop over neighborlists */
 169     for(iidx=0; iidx<nri; iidx++)
 170     {
 171         /* Load shift vector for this list */
 172         i_shift_offset   = DIM*shiftidx[iidx];
 173
 174         /* Load limits for loop over neighbors */
 175         j_index_start    = jindex[iidx];
 176         j_index_end      = jindex[iidx+1];
 177
 178         /* Get outer coordinate index */
 179         inr              = iinr[iidx];
 180         i_coord_offset   = DIM*inr;
 181
 182         /* Load i particle coords and add shift vector */
 183         gmx_mm_load_shift_and_4rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,
 184                                                  &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
 185
 186         fix0             = _mm_setzero_ps();
 187         fiy0             = _mm_setzero_ps();
 188         fiz0             = _mm_setzero_ps();
 189         fix1             = _mm_setzero_ps();
 190         fiy1             = _mm_setzero_ps();
 191         fiz1             = _mm_setzero_ps();
 192         fix2             = _mm_setzero_ps();
 193         fiy2             = _mm_setzero_ps();
 194         fiz2             = _mm_setzero_ps();
 195         fix3             = _mm_setzero_ps();
 196         fiy3             = _mm_setzero_ps();
 197         fiz3             = _mm_setzero_ps();
 198
 199         /* Reset potential sums */
 200         velecsum         = _mm_setzero_ps();
 201         vvdwsum          = _mm_setzero_ps();
 202
 203         /* Start inner kernel loop */
 204         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 205         {
 206
 207             /* Get j neighbor index, and coordinate index */
 208             jnrA             = jjnr[jidx];
 209             jnrB             = jjnr[jidx+1];
 210             jnrC             = jjnr[jidx+2];
 211             jnrD             = jjnr[jidx+3];
 212             j_coord_offsetA  = DIM*jnrA;
 213             j_coord_offsetB  = DIM*jnrB;
 214             j_coord_offsetC  = DIM*jnrC;
 215             j_coord_offsetD  = DIM*jnrD;
 216
 217             /* load j atom coordinates */
 218             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 219                                               x+j_coord_offsetC,x+j_coord_offsetD,
 220                                               &jx0,&jy0,&jz0);
 221
 222             /* Calculate displacement vector */
 223             dx00             = _mm_sub_ps(ix0,jx0);
 224             dy00             = _mm_sub_ps(iy0,jy0);
 225             dz00             = _mm_sub_ps(iz0,jz0);
 226             dx10             = _mm_sub_ps(ix1,jx0);
 227             dy10             = _mm_sub_ps(iy1,jy0);
 228             dz10             = _mm_sub_ps(iz1,jz0);
 229             dx20             = _mm_sub_ps(ix2,jx0);
 230             dy20             = _mm_sub_ps(iy2,jy0);
 231             dz20             = _mm_sub_ps(iz2,jz0);
 232             dx30             = _mm_sub_ps(ix3,jx0);
 233             dy30             = _mm_sub_ps(iy3,jy0);
 234             dz30             = _mm_sub_ps(iz3,jz0);
 235
 236             /* Calculate squared distance and things based on it */
 237             rsq00            = gmx_mm_calc_rsq_ps(dx00,dy00,dz00);
 238             rsq10            = gmx_mm_calc_rsq_ps(dx10,dy10,dz10);
 239             rsq20            = gmx_mm_calc_rsq_ps(dx20,dy20,dz20);
 240             rsq30            = gmx_mm_calc_rsq_ps(dx30,dy30,dz30);
 241
 242             rinv10           = gmx_mm_invsqrt_ps(rsq10);
 243             rinv20           = gmx_mm_invsqrt_ps(rsq20);
 244             rinv30           = gmx_mm_invsqrt_ps(rsq30);
 245
 246             rinvsq00         = gmx_mm_inv_ps(rsq00);
 247             rinvsq10         = _mm_mul_ps(rinv10,rinv10);
 248             rinvsq20         = _mm_mul_ps(rinv20,rinv20);
 249             rinvsq30         = _mm_mul_ps(rinv30,rinv30);
 250
 251             /* Load parameters for j particles */
 252             jq0              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 253                                                               charge+jnrC+0,charge+jnrD+0);
 254             vdwjidx0A        = 2*vdwtype[jnrA+0];
 255             vdwjidx0B        = 2*vdwtype[jnrB+0];
 256             vdwjidx0C        = 2*vdwtype[jnrC+0];
 257             vdwjidx0D        = 2*vdwtype[jnrD+0];
 258
 259             fjx0             = _mm_setzero_ps();
 260             fjy0             = _mm_setzero_ps();
 261             fjz0             = _mm_setzero_ps();
 262
 263             /**************************
 264              * CALCULATE INTERACTIONS *
 265              **************************/
 266
 267             if (gmx_mm_any_lt(rsq00,rcutoff2))
 268             {
 269
 270             /* Compute parameters for interactions between i and j atoms */
 271             gmx_mm_load_4pair_swizzle_ps(vdwparam+vdwioffset0+vdwjidx0A,
 272                                          vdwparam+vdwioffset0+vdwjidx0B,
 273                                          vdwparam+vdwioffset0+vdwjidx0C,
 274                                          vdwparam+vdwioffset0+vdwjidx0D,
 275                                          &c6_00,&c12_00);
 276
 277             /* LENNARD-JONES DISPERSION/REPULSION */
 278
 279             rinvsix          = _mm_mul_ps(_mm_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 280             vvdw6            = _mm_mul_ps(c6_00,rinvsix);
 281             vvdw12           = _mm_mul_ps(c12_00,_mm_mul_ps(rinvsix,rinvsix));
 282             vvdw             = _mm_msub_ps(_mm_nmacc_ps(c12_00,_mm_mul_ps(sh_vdw_invrcut6,sh_vdw_invrcut6),vvdw12),one_twelfth,
 283                                           _mm_mul_ps( _mm_nmacc_ps(c6_00,sh_vdw_invrcut6,vvdw6),one_sixth));
 284             fvdw             = _mm_mul_ps(_mm_sub_ps(vvdw12,vvdw6),rinvsq00);
 285
 286             cutoff_mask      = _mm_cmplt_ps(rsq00,rcutoff2);
 287
 288             /* Update potential sum for this i atom from the interaction with this j atom. */
 289             vvdw             = _mm_and_ps(vvdw,cutoff_mask);
 290             vvdwsum          = _mm_add_ps(vvdwsum,vvdw);
 291
 292             fscal            = fvdw;
 293
 294             fscal            = _mm_and_ps(fscal,cutoff_mask);
 295
 296              /* Update vectorial force */
 297             fix0             = _mm_macc_ps(dx00,fscal,fix0);
 298             fiy0             = _mm_macc_ps(dy00,fscal,fiy0);
 299             fiz0             = _mm_macc_ps(dz00,fscal,fiz0);
 300
 301             fjx0             = _mm_macc_ps(dx00,fscal,fjx0);
 302             fjy0             = _mm_macc_ps(dy00,fscal,fjy0);
 303             fjz0             = _mm_macc_ps(dz00,fscal,fjz0);
 304
 305             }
 306
 307             /**************************
 308              * CALCULATE INTERACTIONS *
 309              **************************/
 310
 311             if (gmx_mm_any_lt(rsq10,rcutoff2))
 312             {
 313
 314             r10              = _mm_mul_ps(rsq10,rinv10);
 315
 316             /* Compute parameters for interactions between i and j atoms */
 317             qq10             = _mm_mul_ps(iq1,jq0);
 318
 319             /* EWALD ELECTROSTATICS */
 320
 321             /* Analytical PME correction */
 322             zeta2            = _mm_mul_ps(beta2,rsq10);
 323             rinv3            = _mm_mul_ps(rinvsq10,rinv10);
 324             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
 325             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
 326             felec            = _mm_mul_ps(qq10,felec);
 327             pmecorrV         = gmx_mm_pmecorrV_ps(zeta2);
 328             velec            = _mm_nmacc_ps(pmecorrV,beta,_mm_sub_ps(rinv10,sh_ewald));
 329             velec            = _mm_mul_ps(qq10,velec);
 330
 331             cutoff_mask      = _mm_cmplt_ps(rsq10,rcutoff2);
 332
 333             /* Update potential sum for this i atom from the interaction with this j atom. */
 334             velec            = _mm_and_ps(velec,cutoff_mask);
 335             velecsum         = _mm_add_ps(velecsum,velec);
 336
 337             fscal            = felec;
 338
 339             fscal            = _mm_and_ps(fscal,cutoff_mask);
 340
 341              /* Update vectorial force */
 342             fix1             = _mm_macc_ps(dx10,fscal,fix1);
 343             fiy1             = _mm_macc_ps(dy10,fscal,fiy1);
 344             fiz1             = _mm_macc_ps(dz10,fscal,fiz1);
 345
 346             fjx0             = _mm_macc_ps(dx10,fscal,fjx0);
 347             fjy0             = _mm_macc_ps(dy10,fscal,fjy0);
 348             fjz0             = _mm_macc_ps(dz10,fscal,fjz0);
 349
 350             }
 351
 352             /**************************
 353              * CALCULATE INTERACTIONS *
 354              **************************/
 355
 356             if (gmx_mm_any_lt(rsq20,rcutoff2))
 357             {
 358
 359             r20              = _mm_mul_ps(rsq20,rinv20);
 360
 361             /* Compute parameters for interactions between i and j atoms */
 362             qq20             = _mm_mul_ps(iq2,jq0);
 363
 364             /* EWALD ELECTROSTATICS */
 365
 366             /* Analytical PME correction */
 367             zeta2            = _mm_mul_ps(beta2,rsq20);
 368             rinv3            = _mm_mul_ps(rinvsq20,rinv20);
 369             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
 370             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
 371             felec            = _mm_mul_ps(qq20,felec);
 372             pmecorrV         = gmx_mm_pmecorrV_ps(zeta2);
 373             velec            = _mm_nmacc_ps(pmecorrV,beta,_mm_sub_ps(rinv20,sh_ewald));
 374             velec            = _mm_mul_ps(qq20,velec);
 375
 376             cutoff_mask      = _mm_cmplt_ps(rsq20,rcutoff2);
 377
 378             /* Update potential sum for this i atom from the interaction with this j atom. */
 379             velec            = _mm_and_ps(velec,cutoff_mask);
 380             velecsum         = _mm_add_ps(velecsum,velec);
 381
 382             fscal            = felec;
 383
 384             fscal            = _mm_and_ps(fscal,cutoff_mask);
 385
 386              /* Update vectorial force */
 387             fix2             = _mm_macc_ps(dx20,fscal,fix2);
 388             fiy2             = _mm_macc_ps(dy20,fscal,fiy2);
 389             fiz2             = _mm_macc_ps(dz20,fscal,fiz2);
 390
 391             fjx0             = _mm_macc_ps(dx20,fscal,fjx0);
 392             fjy0             = _mm_macc_ps(dy20,fscal,fjy0);
 393             fjz0             = _mm_macc_ps(dz20,fscal,fjz0);
 394
 395             }
 396
 397             /**************************
 398              * CALCULATE INTERACTIONS *
 399              **************************/
 400
 401             if (gmx_mm_any_lt(rsq30,rcutoff2))
 402             {
 403
 404             r30              = _mm_mul_ps(rsq30,rinv30);
 405
 406             /* Compute parameters for interactions between i and j atoms */
 407             qq30             = _mm_mul_ps(iq3,jq0);
 408
 409             /* EWALD ELECTROSTATICS */
 410
 411             /* Analytical PME correction */
 412             zeta2            = _mm_mul_ps(beta2,rsq30);
 413             rinv3            = _mm_mul_ps(rinvsq30,rinv30);
 414             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
 415             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
 416             felec            = _mm_mul_ps(qq30,felec);
 417             pmecorrV         = gmx_mm_pmecorrV_ps(zeta2);
 418             velec            = _mm_nmacc_ps(pmecorrV,beta,_mm_sub_ps(rinv30,sh_ewald));
 419             velec            = _mm_mul_ps(qq30,velec);
 420
 421             cutoff_mask      = _mm_cmplt_ps(rsq30,rcutoff2);
 422
 423             /* Update potential sum for this i atom from the interaction with this j atom. */
 424             velec            = _mm_and_ps(velec,cutoff_mask);
 425             velecsum         = _mm_add_ps(velecsum,velec);
 426
 427             fscal            = felec;
 428
 429             fscal            = _mm_and_ps(fscal,cutoff_mask);
 430
 431              /* Update vectorial force */
 432             fix3             = _mm_macc_ps(dx30,fscal,fix3);
 433             fiy3             = _mm_macc_ps(dy30,fscal,fiy3);
 434             fiz3             = _mm_macc_ps(dz30,fscal,fiz3);
 435
 436             fjx0             = _mm_macc_ps(dx30,fscal,fjx0);
 437             fjy0             = _mm_macc_ps(dy30,fscal,fjy0);
 438             fjz0             = _mm_macc_ps(dz30,fscal,fjz0);
 439
 440             }
 441
 442             fjptrA             = f+j_coord_offsetA;
 443             fjptrB             = f+j_coord_offsetB;
 444             fjptrC             = f+j_coord_offsetC;
 445             fjptrD             = f+j_coord_offsetD;
 446
 447             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
 448
 449             /* Inner loop uses 143 flops */
 450         }
 451
 452         if(jidx<j_index_end)
 453         {
 454
 455             /* Get j neighbor index, and coordinate index */
 456             jnrlistA         = jjnr[jidx];
 457             jnrlistB         = jjnr[jidx+1];
 458             jnrlistC         = jjnr[jidx+2];
 459             jnrlistD         = jjnr[jidx+3];
 460             /* Sign of each element will be negative for non-real atoms.
 461              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 462              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
 463              */
 464             dummy_mask = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
 465             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 466             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 467             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 468             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 469             j_coord_offsetA  = DIM*jnrA;
 470             j_coord_offsetB  = DIM*jnrB;
 471             j_coord_offsetC  = DIM*jnrC;
 472             j_coord_offsetD  = DIM*jnrD;
 473
 474             /* load j atom coordinates */
 475             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 476                                               x+j_coord_offsetC,x+j_coord_offsetD,
 477                                               &jx0,&jy0,&jz0);
 478
 479             /* Calculate displacement vector */
 480             dx00             = _mm_sub_ps(ix0,jx0);
 481             dy00             = _mm_sub_ps(iy0,jy0);
 482             dz00             = _mm_sub_ps(iz0,jz0);
 483             dx10             = _mm_sub_ps(ix1,jx0);
 484             dy10             = _mm_sub_ps(iy1,jy0);
 485             dz10             = _mm_sub_ps(iz1,jz0);
 486             dx20             = _mm_sub_ps(ix2,jx0);
 487             dy20             = _mm_sub_ps(iy2,jy0);
 488             dz20             = _mm_sub_ps(iz2,jz0);
 489             dx30             = _mm_sub_ps(ix3,jx0);
 490             dy30             = _mm_sub_ps(iy3,jy0);
 491             dz30             = _mm_sub_ps(iz3,jz0);
 492
 493             /* Calculate squared distance and things based on it */
 494             rsq00            = gmx_mm_calc_rsq_ps(dx00,dy00,dz00);
 495             rsq10            = gmx_mm_calc_rsq_ps(dx10,dy10,dz10);
 496             rsq20            = gmx_mm_calc_rsq_ps(dx20,dy20,dz20);
 497             rsq30            = gmx_mm_calc_rsq_ps(dx30,dy30,dz30);
 498
 499             rinv10           = gmx_mm_invsqrt_ps(rsq10);
 500             rinv20           = gmx_mm_invsqrt_ps(rsq20);
 501             rinv30           = gmx_mm_invsqrt_ps(rsq30);
 502
 503             rinvsq00         = gmx_mm_inv_ps(rsq00);
 504             rinvsq10         = _mm_mul_ps(rinv10,rinv10);
 505             rinvsq20         = _mm_mul_ps(rinv20,rinv20);
 506             rinvsq30         = _mm_mul_ps(rinv30,rinv30);
 507
 508             /* Load parameters for j particles */
 509             jq0              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 510                                                               charge+jnrC+0,charge+jnrD+0);
 511             vdwjidx0A        = 2*vdwtype[jnrA+0];
 512             vdwjidx0B        = 2*vdwtype[jnrB+0];
 513             vdwjidx0C        = 2*vdwtype[jnrC+0];
 514             vdwjidx0D        = 2*vdwtype[jnrD+0];
 515
 516             fjx0             = _mm_setzero_ps();
 517             fjy0             = _mm_setzero_ps();
 518             fjz0             = _mm_setzero_ps();
 519
 520             /**************************
 521              * CALCULATE INTERACTIONS *
 522              **************************/
 523
 524             if (gmx_mm_any_lt(rsq00,rcutoff2))
 525             {
 526
 527             /* Compute parameters for interactions between i and j atoms */
 528             gmx_mm_load_4pair_swizzle_ps(vdwparam+vdwioffset0+vdwjidx0A,
 529                                          vdwparam+vdwioffset0+vdwjidx0B,
 530                                          vdwparam+vdwioffset0+vdwjidx0C,
 531                                          vdwparam+vdwioffset0+vdwjidx0D,
 532                                          &c6_00,&c12_00);
 533
 534             /* LENNARD-JONES DISPERSION/REPULSION */
 535
 536             rinvsix          = _mm_mul_ps(_mm_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 537             vvdw6            = _mm_mul_ps(c6_00,rinvsix);
 538             vvdw12           = _mm_mul_ps(c12_00,_mm_mul_ps(rinvsix,rinvsix));
 539             vvdw             = _mm_msub_ps(_mm_nmacc_ps(c12_00,_mm_mul_ps(sh_vdw_invrcut6,sh_vdw_invrcut6),vvdw12),one_twelfth,
 540                                           _mm_mul_ps( _mm_nmacc_ps(c6_00,sh_vdw_invrcut6,vvdw6),one_sixth));
 541             fvdw             = _mm_mul_ps(_mm_sub_ps(vvdw12,vvdw6),rinvsq00);
 542
 543             cutoff_mask      = _mm_cmplt_ps(rsq00,rcutoff2);
 544
 545             /* Update potential sum for this i atom from the interaction with this j atom. */
 546             vvdw             = _mm_and_ps(vvdw,cutoff_mask);
 547             vvdw             = _mm_andnot_ps(dummy_mask,vvdw);
 548             vvdwsum          = _mm_add_ps(vvdwsum,vvdw);
 549
 550             fscal            = fvdw;
 551
 552             fscal            = _mm_and_ps(fscal,cutoff_mask);
 553
 554             fscal            = _mm_andnot_ps(dummy_mask,fscal);
 555
 556              /* Update vectorial force */
 557             fix0             = _mm_macc_ps(dx00,fscal,fix0);
 558             fiy0             = _mm_macc_ps(dy00,fscal,fiy0);
 559             fiz0             = _mm_macc_ps(dz00,fscal,fiz0);
 560
 561             fjx0             = _mm_macc_ps(dx00,fscal,fjx0);
 562             fjy0             = _mm_macc_ps(dy00,fscal,fjy0);
 563             fjz0             = _mm_macc_ps(dz00,fscal,fjz0);
 564
 565             }
 566
 567             /**************************
 568              * CALCULATE INTERACTIONS *
 569              **************************/
 570
 571             if (gmx_mm_any_lt(rsq10,rcutoff2))
 572             {
 573
 574             r10              = _mm_mul_ps(rsq10,rinv10);
 575             r10              = _mm_andnot_ps(dummy_mask,r10);
 576
 577             /* Compute parameters for interactions between i and j atoms */
 578             qq10             = _mm_mul_ps(iq1,jq0);
 579
 580             /* EWALD ELECTROSTATICS */
 581
 582             /* Analytical PME correction */
 583             zeta2            = _mm_mul_ps(beta2,rsq10);
 584             rinv3            = _mm_mul_ps(rinvsq10,rinv10);
 585             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
 586             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
 587             felec            = _mm_mul_ps(qq10,felec);
 588             pmecorrV         = gmx_mm_pmecorrV_ps(zeta2);
 589             velec            = _mm_nmacc_ps(pmecorrV,beta,_mm_sub_ps(rinv10,sh_ewald));
 590             velec            = _mm_mul_ps(qq10,velec);
 591
 592             cutoff_mask      = _mm_cmplt_ps(rsq10,rcutoff2);
 593
 594             /* Update potential sum for this i atom from the interaction with this j atom. */
 595             velec            = _mm_and_ps(velec,cutoff_mask);
 596             velec            = _mm_andnot_ps(dummy_mask,velec);
 597             velecsum         = _mm_add_ps(velecsum,velec);
 598
 599             fscal            = felec;
 600
 601             fscal            = _mm_and_ps(fscal,cutoff_mask);
 602
 603             fscal            = _mm_andnot_ps(dummy_mask,fscal);
 604
 605              /* Update vectorial force */
 606             fix1             = _mm_macc_ps(dx10,fscal,fix1);
 607             fiy1             = _mm_macc_ps(dy10,fscal,fiy1);
 608             fiz1             = _mm_macc_ps(dz10,fscal,fiz1);
 609
 610             fjx0             = _mm_macc_ps(dx10,fscal,fjx0);
 611             fjy0             = _mm_macc_ps(dy10,fscal,fjy0);
 612             fjz0             = _mm_macc_ps(dz10,fscal,fjz0);
 613
 614             }
 615
 616             /**************************
 617              * CALCULATE INTERACTIONS *
 618              **************************/
 619
 620             if (gmx_mm_any_lt(rsq20,rcutoff2))
 621             {
 622
 623             r20              = _mm_mul_ps(rsq20,rinv20);
 624             r20              = _mm_andnot_ps(dummy_mask,r20);
 625
 626             /* Compute parameters for interactions between i and j atoms */
 627             qq20             = _mm_mul_ps(iq2,jq0);
 628
 629             /* EWALD ELECTROSTATICS */
 630
 631             /* Analytical PME correction */
 632             zeta2            = _mm_mul_ps(beta2,rsq20);
 633             rinv3            = _mm_mul_ps(rinvsq20,rinv20);
 634             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
 635             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
 636             felec            = _mm_mul_ps(qq20,felec);
 637             pmecorrV         = gmx_mm_pmecorrV_ps(zeta2);
 638             velec            = _mm_nmacc_ps(pmecorrV,beta,_mm_sub_ps(rinv20,sh_ewald));
 639             velec            = _mm_mul_ps(qq20,velec);
 640
 641             cutoff_mask      = _mm_cmplt_ps(rsq20,rcutoff2);
 642
 643             /* Update potential sum for this i atom from the interaction with this j atom. */
 644             velec            = _mm_and_ps(velec,cutoff_mask);
 645             velec            = _mm_andnot_ps(dummy_mask,velec);
 646             velecsum         = _mm_add_ps(velecsum,velec);
 647
 648             fscal            = felec;
 649
 650             fscal            = _mm_and_ps(fscal,cutoff_mask);
 651
 652             fscal            = _mm_andnot_ps(dummy_mask,fscal);
 653
 654              /* Update vectorial force */
 655             fix2             = _mm_macc_ps(dx20,fscal,fix2);
 656             fiy2             = _mm_macc_ps(dy20,fscal,fiy2);
 657             fiz2             = _mm_macc_ps(dz20,fscal,fiz2);
 658
 659             fjx0             = _mm_macc_ps(dx20,fscal,fjx0);
 660             fjy0             = _mm_macc_ps(dy20,fscal,fjy0);
 661             fjz0             = _mm_macc_ps(dz20,fscal,fjz0);
 662
 663             }
 664
 665             /**************************
 666              * CALCULATE INTERACTIONS *
 667              **************************/
 668
 669             if (gmx_mm_any_lt(rsq30,rcutoff2))
 670             {
 671
 672             r30              = _mm_mul_ps(rsq30,rinv30);
 673             r30              = _mm_andnot_ps(dummy_mask,r30);
 674
 675             /* Compute parameters for interactions between i and j atoms */
 676             qq30             = _mm_mul_ps(iq3,jq0);
 677
 678             /* EWALD ELECTROSTATICS */
 679
 680             /* Analytical PME correction */
 681             zeta2            = _mm_mul_ps(beta2,rsq30);
 682             rinv3            = _mm_mul_ps(rinvsq30,rinv30);
 683             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
 684             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
 685             felec            = _mm_mul_ps(qq30,felec);
 686             pmecorrV         = gmx_mm_pmecorrV_ps(zeta2);
 687             velec            = _mm_nmacc_ps(pmecorrV,beta,_mm_sub_ps(rinv30,sh_ewald));
 688             velec            = _mm_mul_ps(qq30,velec);
 689
 690             cutoff_mask      = _mm_cmplt_ps(rsq30,rcutoff2);
 691
 692             /* Update potential sum for this i atom from the interaction with this j atom. */
 693             velec            = _mm_and_ps(velec,cutoff_mask);
 694             velec            = _mm_andnot_ps(dummy_mask,velec);
 695             velecsum         = _mm_add_ps(velecsum,velec);
 696
 697             fscal            = felec;
 698
 699             fscal            = _mm_and_ps(fscal,cutoff_mask);
 700
 701             fscal            = _mm_andnot_ps(dummy_mask,fscal);
 702
 703              /* Update vectorial force */
 704             fix3             = _mm_macc_ps(dx30,fscal,fix3);
 705             fiy3             = _mm_macc_ps(dy30,fscal,fiy3);
 706             fiz3             = _mm_macc_ps(dz30,fscal,fiz3);
 707
 708             fjx0             = _mm_macc_ps(dx30,fscal,fjx0);
 709             fjy0             = _mm_macc_ps(dy30,fscal,fjy0);
 710             fjz0             = _mm_macc_ps(dz30,fscal,fjz0);
 711
 712             }
 713
 714             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 715             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 716             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 717             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 718
 719             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
 720
 721             /* Inner loop uses 146 flops */
 722         }
 723
 724         /* End of innermost loop */
 725
 726         gmx_mm_update_iforce_4atom_swizzle_ps(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
 727                                               f+i_coord_offset,fshift+i_shift_offset);
 728
 729         ggid                        = gid[iidx];
 730         /* Update potential energies */
 731         gmx_mm_update_1pot_ps(velecsum,kernel_data->energygrp_elec+ggid);
 732         gmx_mm_update_1pot_ps(vvdwsum,kernel_data->energygrp_vdw+ggid);
 733
 734         /* Increment number of inner iterations */
 735         inneriter                  += j_index_end - j_index_start;
 736
 737         /* Outer loop uses 26 flops */
 738     }
 739
 740     /* Increment number of outer iterations */
 741     outeriter        += nri;
 742
 743     /* Update outer/inner flops */
 744
 745     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4_VF,outeriter*26 + inneriter*146);
 746 }
 747 /*
 748  * Gromacs nonbonded kernel:   nb_kernel_ElecEwSh_VdwLJSh_GeomW4P1_F_avx_128_fma_single
 749  * Electrostatics interaction: Ewald
 750  * VdW interaction:            LennardJones
 751  * Geometry:                   Water4-Particle
 752  * Calculate force/pot:        Force
 753  */
 754 void
 755 nb_kernel_ElecEwSh_VdwLJSh_GeomW4P1_F_avx_128_fma_single
 756                     (t_nblist                    * gmx_restrict       nlist,
 757                      rvec                        * gmx_restrict          xx,
 758                      rvec                        * gmx_restrict          ff,
 759                      t_forcerec                  * gmx_restrict          fr,
 760                      t_mdatoms                   * gmx_restrict     mdatoms,
 761                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 762                      t_nrnb                      * gmx_restrict        nrnb)
 763 {
 764     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
 765      * just 0 for non-waters.
 766      * Suffixes A,B,C,D refer to j loop unrolling done with AVX_128, e.g. for the four different
 767      * jnr indices corresponding to data put in the four positions in the SIMD register.
 768      */
 769     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
 770     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
 771     int              jnrA,jnrB,jnrC,jnrD;
 772     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
 773     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
 774     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 775     real             rcutoff_scalar;
 776     real             *shiftvec,*fshift,*x,*f;
 777     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
 778     real             scratch[4*DIM];
 779     __m128           fscal,rcutoff,rcutoff2,jidxall;
 780     int              vdwioffset0;
 781     __m128           ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 782     int              vdwioffset1;
 783     __m128           ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
 784     int              vdwioffset2;
 785     __m128           ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
 786     int              vdwioffset3;
 787     __m128           ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
 788     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
 789     __m128           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 790     __m128           dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
 791     __m128           dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
 792     __m128           dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
 793     __m128           dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30;
 794     __m128           velec,felec,velecsum,facel,crf,krf,krf2;
 795     real             *charge;
 796     int              nvdwtype;
 797     __m128           rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 798     int              *vdwtype;
 799     real             *vdwparam;
 800     __m128           one_sixth   = _mm_set1_ps(1.0/6.0);
 801     __m128           one_twelfth = _mm_set1_ps(1.0/12.0);
 802     __m128i          ewitab;
 803     __m128           ewtabscale,eweps,twoeweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
 804     __m128           beta,beta2,beta3,zeta2,pmecorrF,pmecorrV,rinv3;
 805     real             *ewtab;
 806     __m128           dummy_mask,cutoff_mask;
 807     __m128           signbit = _mm_castsi128_ps( _mm_set1_epi32(0x80000000) );
 808     __m128           one     = _mm_set1_ps(1.0);
 809     __m128           two     = _mm_set1_ps(2.0);
 810     x                = xx[0];
 811     f                = ff[0];
 812
 813     nri              = nlist->nri;
 814     iinr             = nlist->iinr;
 815     jindex           = nlist->jindex;
 816     jjnr             = nlist->jjnr;
 817     shiftidx         = nlist->shift;
 818     gid              = nlist->gid;
 819     shiftvec         = fr->shift_vec[0];
 820     fshift           = fr->fshift[0];
 821     facel            = _mm_set1_ps(fr->epsfac);
 822     charge           = mdatoms->chargeA;
 823     nvdwtype         = fr->ntype;
 824     vdwparam         = fr->nbfp;
 825     vdwtype          = mdatoms->typeA;
 826
 827     sh_ewald         = _mm_set1_ps(fr->ic->sh_ewald);
 828     beta             = _mm_set1_ps(fr->ic->ewaldcoeff_q);
 829     beta2            = _mm_mul_ps(beta,beta);
 830     beta3            = _mm_mul_ps(beta,beta2);
 831     ewtab            = fr->ic->tabq_coul_F;
 832     ewtabscale       = _mm_set1_ps(fr->ic->tabq_scale);
 833     ewtabhalfspace   = _mm_set1_ps(0.5/fr->ic->tabq_scale);
 834
 835     /* Setup water-specific parameters */
 836     inr              = nlist->iinr[0];
 837     iq1              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+1]));
 838     iq2              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+2]));
 839     iq3              = _mm_mul_ps(facel,_mm_set1_ps(charge[inr+3]));
 840     vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 841
 842     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
 843     rcutoff_scalar   = fr->rcoulomb;
 844     rcutoff          = _mm_set1_ps(rcutoff_scalar);
 845     rcutoff2         = _mm_mul_ps(rcutoff,rcutoff);
 846
 847     sh_vdw_invrcut6  = _mm_set1_ps(fr->ic->sh_invrc6);
 848     rvdw             = _mm_set1_ps(fr->rvdw);
 849
 850     /* Avoid stupid compiler warnings */
 851     jnrA = jnrB = jnrC = jnrD = 0;
 852     j_coord_offsetA = 0;
 853     j_coord_offsetB = 0;
 854     j_coord_offsetC = 0;
 855     j_coord_offsetD = 0;
 856
 857     outeriter        = 0;
 858     inneriter        = 0;
 859
 860     for(iidx=0;iidx<4*DIM;iidx++)
 861     {
 862         scratch[iidx] = 0.0;
 863     }
 864
 865     /* Start outer loop over neighborlists */
 866     for(iidx=0; iidx<nri; iidx++)
 867     {
 868         /* Load shift vector for this list */
 869         i_shift_offset   = DIM*shiftidx[iidx];
 870
 871         /* Load limits for loop over neighbors */
 872         j_index_start    = jindex[iidx];
 873         j_index_end      = jindex[iidx+1];
 874
 875         /* Get outer coordinate index */
 876         inr              = iinr[iidx];
 877         i_coord_offset   = DIM*inr;
 878
 879         /* Load i particle coords and add shift vector */
 880         gmx_mm_load_shift_and_4rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,
 881                                                  &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
 882
 883         fix0             = _mm_setzero_ps();
 884         fiy0             = _mm_setzero_ps();
 885         fiz0             = _mm_setzero_ps();
 886         fix1             = _mm_setzero_ps();
 887         fiy1             = _mm_setzero_ps();
 888         fiz1             = _mm_setzero_ps();
 889         fix2             = _mm_setzero_ps();
 890         fiy2             = _mm_setzero_ps();
 891         fiz2             = _mm_setzero_ps();
 892         fix3             = _mm_setzero_ps();
 893         fiy3             = _mm_setzero_ps();
 894         fiz3             = _mm_setzero_ps();
 895
 896         /* Start inner kernel loop */
 897         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 898         {
 899
 900             /* Get j neighbor index, and coordinate index */
 901             jnrA             = jjnr[jidx];
 902             jnrB             = jjnr[jidx+1];
 903             jnrC             = jjnr[jidx+2];
 904             jnrD             = jjnr[jidx+3];
 905             j_coord_offsetA  = DIM*jnrA;
 906             j_coord_offsetB  = DIM*jnrB;
 907             j_coord_offsetC  = DIM*jnrC;
 908             j_coord_offsetD  = DIM*jnrD;
 909
 910             /* load j atom coordinates */
 911             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 912                                               x+j_coord_offsetC,x+j_coord_offsetD,
 913                                               &jx0,&jy0,&jz0);
 914
 915             /* Calculate displacement vector */
 916             dx00             = _mm_sub_ps(ix0,jx0);
 917             dy00             = _mm_sub_ps(iy0,jy0);
 918             dz00             = _mm_sub_ps(iz0,jz0);
 919             dx10             = _mm_sub_ps(ix1,jx0);
 920             dy10             = _mm_sub_ps(iy1,jy0);
 921             dz10             = _mm_sub_ps(iz1,jz0);
 922             dx20             = _mm_sub_ps(ix2,jx0);
 923             dy20             = _mm_sub_ps(iy2,jy0);
 924             dz20             = _mm_sub_ps(iz2,jz0);
 925             dx30             = _mm_sub_ps(ix3,jx0);
 926             dy30             = _mm_sub_ps(iy3,jy0);
 927             dz30             = _mm_sub_ps(iz3,jz0);
 928
 929             /* Calculate squared distance and things based on it */
 930             rsq00            = gmx_mm_calc_rsq_ps(dx00,dy00,dz00);
 931             rsq10            = gmx_mm_calc_rsq_ps(dx10,dy10,dz10);
 932             rsq20            = gmx_mm_calc_rsq_ps(dx20,dy20,dz20);
 933             rsq30            = gmx_mm_calc_rsq_ps(dx30,dy30,dz30);
 934
 935             rinv10           = gmx_mm_invsqrt_ps(rsq10);
 936             rinv20           = gmx_mm_invsqrt_ps(rsq20);
 937             rinv30           = gmx_mm_invsqrt_ps(rsq30);
 938
 939             rinvsq00         = gmx_mm_inv_ps(rsq00);
 940             rinvsq10         = _mm_mul_ps(rinv10,rinv10);
 941             rinvsq20         = _mm_mul_ps(rinv20,rinv20);
 942             rinvsq30         = _mm_mul_ps(rinv30,rinv30);
 943
 944             /* Load parameters for j particles */
 945             jq0              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 946                                                               charge+jnrC+0,charge+jnrD+0);
 947             vdwjidx0A        = 2*vdwtype[jnrA+0];
 948             vdwjidx0B        = 2*vdwtype[jnrB+0];
 949             vdwjidx0C        = 2*vdwtype[jnrC+0];
 950             vdwjidx0D        = 2*vdwtype[jnrD+0];
 951
 952             fjx0             = _mm_setzero_ps();
 953             fjy0             = _mm_setzero_ps();
 954             fjz0             = _mm_setzero_ps();
 955
 956             /**************************
 957              * CALCULATE INTERACTIONS *
 958              **************************/
 959
 960             if (gmx_mm_any_lt(rsq00,rcutoff2))
 961             {
 962
 963             /* Compute parameters for interactions between i and j atoms */
 964             gmx_mm_load_4pair_swizzle_ps(vdwparam+vdwioffset0+vdwjidx0A,
 965                                          vdwparam+vdwioffset0+vdwjidx0B,
 966                                          vdwparam+vdwioffset0+vdwjidx0C,
 967                                          vdwparam+vdwioffset0+vdwjidx0D,
 968                                          &c6_00,&c12_00);
 969
 970             /* LENNARD-JONES DISPERSION/REPULSION */
 971
 972             rinvsix          = _mm_mul_ps(_mm_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 973             fvdw             = _mm_mul_ps(_mm_msub_ps(c12_00,rinvsix,c6_00),_mm_mul_ps(rinvsix,rinvsq00));
 974
 975             cutoff_mask      = _mm_cmplt_ps(rsq00,rcutoff2);
 976
 977             fscal            = fvdw;
 978
 979             fscal            = _mm_and_ps(fscal,cutoff_mask);
 980
 981              /* Update vectorial force */
 982             fix0             = _mm_macc_ps(dx00,fscal,fix0);
 983             fiy0             = _mm_macc_ps(dy00,fscal,fiy0);
 984             fiz0             = _mm_macc_ps(dz00,fscal,fiz0);
 985
 986             fjx0             = _mm_macc_ps(dx00,fscal,fjx0);
 987             fjy0             = _mm_macc_ps(dy00,fscal,fjy0);
 988             fjz0             = _mm_macc_ps(dz00,fscal,fjz0);
 989
 990             }
 991
 992             /**************************
 993              * CALCULATE INTERACTIONS *
 994              **************************/
 995
 996             if (gmx_mm_any_lt(rsq10,rcutoff2))
 997             {
 998
 999             r10              = _mm_mul_ps(rsq10,rinv10);
1000
1001             /* Compute parameters for interactions between i and j atoms */
1002             qq10             = _mm_mul_ps(iq1,jq0);
1003
1004             /* EWALD ELECTROSTATICS */
1005
1006             /* Analytical PME correction */
1007             zeta2            = _mm_mul_ps(beta2,rsq10);
1008             rinv3            = _mm_mul_ps(rinvsq10,rinv10);
1009             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
1010             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
1011             felec            = _mm_mul_ps(qq10,felec);
1012
1013             cutoff_mask      = _mm_cmplt_ps(rsq10,rcutoff2);
1014
1015             fscal            = felec;
1016
1017             fscal            = _mm_and_ps(fscal,cutoff_mask);
1018
1019              /* Update vectorial force */
1020             fix1             = _mm_macc_ps(dx10,fscal,fix1);
1021             fiy1             = _mm_macc_ps(dy10,fscal,fiy1);
1022             fiz1             = _mm_macc_ps(dz10,fscal,fiz1);
1023
1024             fjx0             = _mm_macc_ps(dx10,fscal,fjx0);
1025             fjy0             = _mm_macc_ps(dy10,fscal,fjy0);
1026             fjz0             = _mm_macc_ps(dz10,fscal,fjz0);
1027
1028             }
1029
1030             /**************************
1031              * CALCULATE INTERACTIONS *
1032              **************************/
1033
1034             if (gmx_mm_any_lt(rsq20,rcutoff2))
1035             {
1036
1037             r20              = _mm_mul_ps(rsq20,rinv20);
1038
1039             /* Compute parameters for interactions between i and j atoms */
1040             qq20             = _mm_mul_ps(iq2,jq0);
1041
1042             /* EWALD ELECTROSTATICS */
1043
1044             /* Analytical PME correction */
1045             zeta2            = _mm_mul_ps(beta2,rsq20);
1046             rinv3            = _mm_mul_ps(rinvsq20,rinv20);
1047             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
1048             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
1049             felec            = _mm_mul_ps(qq20,felec);
1050
1051             cutoff_mask      = _mm_cmplt_ps(rsq20,rcutoff2);
1052
1053             fscal            = felec;
1054
1055             fscal            = _mm_and_ps(fscal,cutoff_mask);
1056
1057              /* Update vectorial force */
1058             fix2             = _mm_macc_ps(dx20,fscal,fix2);
1059             fiy2             = _mm_macc_ps(dy20,fscal,fiy2);
1060             fiz2             = _mm_macc_ps(dz20,fscal,fiz2);
1061
1062             fjx0             = _mm_macc_ps(dx20,fscal,fjx0);
1063             fjy0             = _mm_macc_ps(dy20,fscal,fjy0);
1064             fjz0             = _mm_macc_ps(dz20,fscal,fjz0);
1065
1066             }
1067
1068             /**************************
1069              * CALCULATE INTERACTIONS *
1070              **************************/
1071
1072             if (gmx_mm_any_lt(rsq30,rcutoff2))
1073             {
1074
1075             r30              = _mm_mul_ps(rsq30,rinv30);
1076
1077             /* Compute parameters for interactions between i and j atoms */
1078             qq30             = _mm_mul_ps(iq3,jq0);
1079
1080             /* EWALD ELECTROSTATICS */
1081
1082             /* Analytical PME correction */
1083             zeta2            = _mm_mul_ps(beta2,rsq30);
1084             rinv3            = _mm_mul_ps(rinvsq30,rinv30);
1085             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
1086             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
1087             felec            = _mm_mul_ps(qq30,felec);
1088
1089             cutoff_mask      = _mm_cmplt_ps(rsq30,rcutoff2);
1090
1091             fscal            = felec;
1092
1093             fscal            = _mm_and_ps(fscal,cutoff_mask);
1094
1095              /* Update vectorial force */
1096             fix3             = _mm_macc_ps(dx30,fscal,fix3);
1097             fiy3             = _mm_macc_ps(dy30,fscal,fiy3);
1098             fiz3             = _mm_macc_ps(dz30,fscal,fiz3);
1099
1100             fjx0             = _mm_macc_ps(dx30,fscal,fjx0);
1101             fjy0             = _mm_macc_ps(dy30,fscal,fjy0);
1102             fjz0             = _mm_macc_ps(dz30,fscal,fjz0);
1103
1104             }
1105
1106             fjptrA             = f+j_coord_offsetA;
1107             fjptrB             = f+j_coord_offsetB;
1108             fjptrC             = f+j_coord_offsetC;
1109             fjptrD             = f+j_coord_offsetD;
1110
1111             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
1112
1113             /* Inner loop uses 126 flops */
1114         }
1115
1116         if(jidx<j_index_end)
1117         {
1118
1119             /* Get j neighbor index, and coordinate index */
1120             jnrlistA         = jjnr[jidx];
1121             jnrlistB         = jjnr[jidx+1];
1122             jnrlistC         = jjnr[jidx+2];
1123             jnrlistD         = jjnr[jidx+3];
1124             /* Sign of each element will be negative for non-real atoms.
1125              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
1126              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
1127              */
1128             dummy_mask = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
1129             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
1130             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
1131             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
1132             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
1133             j_coord_offsetA  = DIM*jnrA;
1134             j_coord_offsetB  = DIM*jnrB;
1135             j_coord_offsetC  = DIM*jnrC;
1136             j_coord_offsetD  = DIM*jnrD;
1137
1138             /* load j atom coordinates */
1139             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
1140                                               x+j_coord_offsetC,x+j_coord_offsetD,
1141                                               &jx0,&jy0,&jz0);
1142
1143             /* Calculate displacement vector */
1144             dx00             = _mm_sub_ps(ix0,jx0);
1145             dy00             = _mm_sub_ps(iy0,jy0);
1146             dz00             = _mm_sub_ps(iz0,jz0);
1147             dx10             = _mm_sub_ps(ix1,jx0);
1148             dy10             = _mm_sub_ps(iy1,jy0);
1149             dz10             = _mm_sub_ps(iz1,jz0);
1150             dx20             = _mm_sub_ps(ix2,jx0);
1151             dy20             = _mm_sub_ps(iy2,jy0);
1152             dz20             = _mm_sub_ps(iz2,jz0);
1153             dx30             = _mm_sub_ps(ix3,jx0);
1154             dy30             = _mm_sub_ps(iy3,jy0);
1155             dz30             = _mm_sub_ps(iz3,jz0);
1156
1157             /* Calculate squared distance and things based on it */
1158             rsq00            = gmx_mm_calc_rsq_ps(dx00,dy00,dz00);
1159             rsq10            = gmx_mm_calc_rsq_ps(dx10,dy10,dz10);
1160             rsq20            = gmx_mm_calc_rsq_ps(dx20,dy20,dz20);
1161             rsq30            = gmx_mm_calc_rsq_ps(dx30,dy30,dz30);
1162
1163             rinv10           = gmx_mm_invsqrt_ps(rsq10);
1164             rinv20           = gmx_mm_invsqrt_ps(rsq20);
1165             rinv30           = gmx_mm_invsqrt_ps(rsq30);
1166
1167             rinvsq00         = gmx_mm_inv_ps(rsq00);
1168             rinvsq10         = _mm_mul_ps(rinv10,rinv10);
1169             rinvsq20         = _mm_mul_ps(rinv20,rinv20);
1170             rinvsq30         = _mm_mul_ps(rinv30,rinv30);
1171
1172             /* Load parameters for j particles */
1173             jq0              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
1174                                                               charge+jnrC+0,charge+jnrD+0);
1175             vdwjidx0A        = 2*vdwtype[jnrA+0];
1176             vdwjidx0B        = 2*vdwtype[jnrB+0];
1177             vdwjidx0C        = 2*vdwtype[jnrC+0];
1178             vdwjidx0D        = 2*vdwtype[jnrD+0];
1179
1180             fjx0             = _mm_setzero_ps();
1181             fjy0             = _mm_setzero_ps();
1182             fjz0             = _mm_setzero_ps();
1183
1184             /**************************
1185              * CALCULATE INTERACTIONS *
1186              **************************/
1187
1188             if (gmx_mm_any_lt(rsq00,rcutoff2))
1189             {
1190
1191             /* Compute parameters for interactions between i and j atoms */
1192             gmx_mm_load_4pair_swizzle_ps(vdwparam+vdwioffset0+vdwjidx0A,
1193                                          vdwparam+vdwioffset0+vdwjidx0B,
1194                                          vdwparam+vdwioffset0+vdwjidx0C,
1195                                          vdwparam+vdwioffset0+vdwjidx0D,
1196                                          &c6_00,&c12_00);
1197
1198             /* LENNARD-JONES DISPERSION/REPULSION */
1199
1200             rinvsix          = _mm_mul_ps(_mm_mul_ps(rinvsq00,rinvsq00),rinvsq00);
1201             fvdw             = _mm_mul_ps(_mm_msub_ps(c12_00,rinvsix,c6_00),_mm_mul_ps(rinvsix,rinvsq00));
1202
1203             cutoff_mask      = _mm_cmplt_ps(rsq00,rcutoff2);
1204
1205             fscal            = fvdw;
1206
1207             fscal            = _mm_and_ps(fscal,cutoff_mask);
1208
1209             fscal            = _mm_andnot_ps(dummy_mask,fscal);
1210
1211              /* Update vectorial force */
1212             fix0             = _mm_macc_ps(dx00,fscal,fix0);
1213             fiy0             = _mm_macc_ps(dy00,fscal,fiy0);
1214             fiz0             = _mm_macc_ps(dz00,fscal,fiz0);
1215
1216             fjx0             = _mm_macc_ps(dx00,fscal,fjx0);
1217             fjy0             = _mm_macc_ps(dy00,fscal,fjy0);
1218             fjz0             = _mm_macc_ps(dz00,fscal,fjz0);
1219
1220             }
1221
1222             /**************************
1223              * CALCULATE INTERACTIONS *
1224              **************************/
1225
1226             if (gmx_mm_any_lt(rsq10,rcutoff2))
1227             {
1228
1229             r10              = _mm_mul_ps(rsq10,rinv10);
1230             r10              = _mm_andnot_ps(dummy_mask,r10);
1231
1232             /* Compute parameters for interactions between i and j atoms */
1233             qq10             = _mm_mul_ps(iq1,jq0);
1234
1235             /* EWALD ELECTROSTATICS */
1236
1237             /* Analytical PME correction */
1238             zeta2            = _mm_mul_ps(beta2,rsq10);
1239             rinv3            = _mm_mul_ps(rinvsq10,rinv10);
1240             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
1241             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
1242             felec            = _mm_mul_ps(qq10,felec);
1243
1244             cutoff_mask      = _mm_cmplt_ps(rsq10,rcutoff2);
1245
1246             fscal            = felec;
1247
1248             fscal            = _mm_and_ps(fscal,cutoff_mask);
1249
1250             fscal            = _mm_andnot_ps(dummy_mask,fscal);
1251
1252              /* Update vectorial force */
1253             fix1             = _mm_macc_ps(dx10,fscal,fix1);
1254             fiy1             = _mm_macc_ps(dy10,fscal,fiy1);
1255             fiz1             = _mm_macc_ps(dz10,fscal,fiz1);
1256
1257             fjx0             = _mm_macc_ps(dx10,fscal,fjx0);
1258             fjy0             = _mm_macc_ps(dy10,fscal,fjy0);
1259             fjz0             = _mm_macc_ps(dz10,fscal,fjz0);
1260
1261             }
1262
1263             /**************************
1264              * CALCULATE INTERACTIONS *
1265              **************************/
1266
1267             if (gmx_mm_any_lt(rsq20,rcutoff2))
1268             {
1269
1270             r20              = _mm_mul_ps(rsq20,rinv20);
1271             r20              = _mm_andnot_ps(dummy_mask,r20);
1272
1273             /* Compute parameters for interactions between i and j atoms */
1274             qq20             = _mm_mul_ps(iq2,jq0);
1275
1276             /* EWALD ELECTROSTATICS */
1277
1278             /* Analytical PME correction */
1279             zeta2            = _mm_mul_ps(beta2,rsq20);
1280             rinv3            = _mm_mul_ps(rinvsq20,rinv20);
1281             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
1282             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
1283             felec            = _mm_mul_ps(qq20,felec);
1284
1285             cutoff_mask      = _mm_cmplt_ps(rsq20,rcutoff2);
1286
1287             fscal            = felec;
1288
1289             fscal            = _mm_and_ps(fscal,cutoff_mask);
1290
1291             fscal            = _mm_andnot_ps(dummy_mask,fscal);
1292
1293              /* Update vectorial force */
1294             fix2             = _mm_macc_ps(dx20,fscal,fix2);
1295             fiy2             = _mm_macc_ps(dy20,fscal,fiy2);
1296             fiz2             = _mm_macc_ps(dz20,fscal,fiz2);
1297
1298             fjx0             = _mm_macc_ps(dx20,fscal,fjx0);
1299             fjy0             = _mm_macc_ps(dy20,fscal,fjy0);
1300             fjz0             = _mm_macc_ps(dz20,fscal,fjz0);
1301
1302             }
1303
1304             /**************************
1305              * CALCULATE INTERACTIONS *
1306              **************************/
1307
1308             if (gmx_mm_any_lt(rsq30,rcutoff2))
1309             {
1310
1311             r30              = _mm_mul_ps(rsq30,rinv30);
1312             r30              = _mm_andnot_ps(dummy_mask,r30);
1313
1314             /* Compute parameters for interactions between i and j atoms */
1315             qq30             = _mm_mul_ps(iq3,jq0);
1316
1317             /* EWALD ELECTROSTATICS */
1318
1319             /* Analytical PME correction */
1320             zeta2            = _mm_mul_ps(beta2,rsq30);
1321             rinv3            = _mm_mul_ps(rinvsq30,rinv30);
1322             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
1323             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
1324             felec            = _mm_mul_ps(qq30,felec);
1325
1326             cutoff_mask      = _mm_cmplt_ps(rsq30,rcutoff2);
1327
1328             fscal            = felec;
1329
1330             fscal            = _mm_and_ps(fscal,cutoff_mask);
1331
1332             fscal            = _mm_andnot_ps(dummy_mask,fscal);
1333
1334              /* Update vectorial force */
1335             fix3             = _mm_macc_ps(dx30,fscal,fix3);
1336             fiy3             = _mm_macc_ps(dy30,fscal,fiy3);
1337             fiz3             = _mm_macc_ps(dz30,fscal,fiz3);
1338
1339             fjx0             = _mm_macc_ps(dx30,fscal,fjx0);
1340             fjy0             = _mm_macc_ps(dy30,fscal,fjy0);
1341             fjz0             = _mm_macc_ps(dz30,fscal,fjz0);
1342
1343             }
1344
1345             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
1346             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
1347             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
1348             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
1349
1350             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
1351
1352             /* Inner loop uses 129 flops */
1353         }
1354
1355         /* End of innermost loop */
1356
1357         gmx_mm_update_iforce_4atom_swizzle_ps(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
1358                                               f+i_coord_offset,fshift+i_shift_offset);
1359
1360         /* Increment number of inner iterations */
1361         inneriter                  += j_index_end - j_index_start;
1362
1363         /* Outer loop uses 24 flops */
1364     }
1365
1366     /* Increment number of outer iterations */
1367     outeriter        += nri;
1368
1369     /* Update outer/inner flops */
1370
1371     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4_F,outeriter*24 + inneriter*129);
1372 }