src/gromacs/gmxlib/nonbonded/nb_kernel_avx_128_fma_single/nb_kernel_ElecEw_VdwLJEw_GeomP1P1_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 "gmxpre.h"
  39
  40 #include "config.h"
  41
  42 #include <math.h>
  43
  44 #include "../nb_kernel.h"
  45 #include "gromacs/legacyheaders/types/simple.h"
  46 #include "gromacs/math/vec.h"
  47 #include "gromacs/legacyheaders/nrnb.h"
  48
  49 #include "gromacs/simd/math_x86_avx_128_fma_single.h"
  50 #include "kernelutil_x86_avx_128_fma_single.h"
  51
  52 /*
  53  * Gromacs nonbonded kernel:   nb_kernel_ElecEw_VdwLJEw_GeomP1P1_VF_avx_128_fma_single
  54  * Electrostatics interaction: Ewald
  55  * VdW interaction:            LJEwald
  56  * Geometry:                   Particle-Particle
  57  * Calculate force/pot:        PotentialAndForce
  58  */
  59 void
  60 nb_kernel_ElecEw_VdwLJEw_GeomP1P1_VF_avx_128_fma_single
  61                     (t_nblist                    * gmx_restrict       nlist,
  62                      rvec                        * gmx_restrict          xx,
  63                      rvec                        * gmx_restrict          ff,
  64                      t_forcerec                  * gmx_restrict          fr,
  65                      t_mdatoms                   * gmx_restrict     mdatoms,
  66                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
  67                      t_nrnb                      * gmx_restrict        nrnb)
  68 {
  69     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
  70      * just 0 for non-waters.
  71      * Suffixes A,B,C,D refer to j loop unrolling done with AVX_128, e.g. for the four different
  72      * jnr indices corresponding to data put in the four positions in the SIMD register.
  73      */
  74     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
  75     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
  76     int              jnrA,jnrB,jnrC,jnrD;
  77     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
  78     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
  79     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  80     real             rcutoff_scalar;
  81     real             *shiftvec,*fshift,*x,*f;
  82     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
  83     real             scratch[4*DIM];
  84     __m128           fscal,rcutoff,rcutoff2,jidxall;
  85     int              vdwioffset0;
  86     __m128           ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
  87     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
  88     __m128           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  89     __m128           dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
  90     __m128           velec,felec,velecsum,facel,crf,krf,krf2;
  91     real             *charge;
  92     int              nvdwtype;
  93     __m128           rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
  94     int              *vdwtype;
  95     real             *vdwparam;
  96     __m128           one_sixth   = _mm_set1_ps(1.0/6.0);
  97     __m128           one_twelfth = _mm_set1_ps(1.0/12.0);
  98     __m128           c6grid_00;
  99     real             *vdwgridparam;
 100     __m128           ewclj,ewclj2,ewclj6,ewcljrsq,poly,exponent,f6A,f6B,sh_lj_ewald;
 101     __m128           one_half = _mm_set1_ps(0.5);
 102     __m128           minus_one = _mm_set1_ps(-1.0);
 103     __m128i          ewitab;
 104     __m128           ewtabscale,eweps,twoeweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
 105     __m128           beta,beta2,beta3,zeta2,pmecorrF,pmecorrV,rinv3;
 106     real             *ewtab;
 107     __m128           dummy_mask,cutoff_mask;
 108     __m128           signbit = _mm_castsi128_ps( _mm_set1_epi32(0x80000000) );
 109     __m128           one     = _mm_set1_ps(1.0);
 110     __m128           two     = _mm_set1_ps(2.0);
 111     x                = xx[0];
 112     f                = ff[0];
 113
 114     nri              = nlist->nri;
 115     iinr             = nlist->iinr;
 116     jindex           = nlist->jindex;
 117     jjnr             = nlist->jjnr;
 118     shiftidx         = nlist->shift;
 119     gid              = nlist->gid;
 120     shiftvec         = fr->shift_vec[0];
 121     fshift           = fr->fshift[0];
 122     facel            = _mm_set1_ps(fr->epsfac);
 123     charge           = mdatoms->chargeA;
 124     nvdwtype         = fr->ntype;
 125     vdwparam         = fr->nbfp;
 126     vdwtype          = mdatoms->typeA;
 127     vdwgridparam     = fr->ljpme_c6grid;
 128     sh_lj_ewald      = _mm_set1_ps(fr->ic->sh_lj_ewald);
 129     ewclj            = _mm_set1_ps(fr->ewaldcoeff_lj);
 130     ewclj2           = _mm_mul_ps(minus_one,_mm_mul_ps(ewclj,ewclj));
 131
 132     sh_ewald         = _mm_set1_ps(fr->ic->sh_ewald);
 133     beta             = _mm_set1_ps(fr->ic->ewaldcoeff_q);
 134     beta2            = _mm_mul_ps(beta,beta);
 135     beta3            = _mm_mul_ps(beta,beta2);
 136     ewtab            = fr->ic->tabq_coul_FDV0;
 137     ewtabscale       = _mm_set1_ps(fr->ic->tabq_scale);
 138     ewtabhalfspace   = _mm_set1_ps(0.5/fr->ic->tabq_scale);
 139
 140     /* Avoid stupid compiler warnings */
 141     jnrA = jnrB = jnrC = jnrD = 0;
 142     j_coord_offsetA = 0;
 143     j_coord_offsetB = 0;
 144     j_coord_offsetC = 0;
 145     j_coord_offsetD = 0;
 146
 147     outeriter        = 0;
 148     inneriter        = 0;
 149
 150     for(iidx=0;iidx<4*DIM;iidx++)
 151     {
 152         scratch[iidx] = 0.0;
 153     }
 154
 155     /* Start outer loop over neighborlists */
 156     for(iidx=0; iidx<nri; iidx++)
 157     {
 158         /* Load shift vector for this list */
 159         i_shift_offset   = DIM*shiftidx[iidx];
 160
 161         /* Load limits for loop over neighbors */
 162         j_index_start    = jindex[iidx];
 163         j_index_end      = jindex[iidx+1];
 164
 165         /* Get outer coordinate index */
 166         inr              = iinr[iidx];
 167         i_coord_offset   = DIM*inr;
 168
 169         /* Load i particle coords and add shift vector */
 170         gmx_mm_load_shift_and_1rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,&ix0,&iy0,&iz0);
 171
 172         fix0             = _mm_setzero_ps();
 173         fiy0             = _mm_setzero_ps();
 174         fiz0             = _mm_setzero_ps();
 175
 176         /* Load parameters for i particles */
 177         iq0              = _mm_mul_ps(facel,_mm_load1_ps(charge+inr+0));
 178         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 179
 180         /* Reset potential sums */
 181         velecsum         = _mm_setzero_ps();
 182         vvdwsum          = _mm_setzero_ps();
 183
 184         /* Start inner kernel loop */
 185         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 186         {
 187
 188             /* Get j neighbor index, and coordinate index */
 189             jnrA             = jjnr[jidx];
 190             jnrB             = jjnr[jidx+1];
 191             jnrC             = jjnr[jidx+2];
 192             jnrD             = jjnr[jidx+3];
 193             j_coord_offsetA  = DIM*jnrA;
 194             j_coord_offsetB  = DIM*jnrB;
 195             j_coord_offsetC  = DIM*jnrC;
 196             j_coord_offsetD  = DIM*jnrD;
 197
 198             /* load j atom coordinates */
 199             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 200                                               x+j_coord_offsetC,x+j_coord_offsetD,
 201                                               &jx0,&jy0,&jz0);
 202
 203             /* Calculate displacement vector */
 204             dx00             = _mm_sub_ps(ix0,jx0);
 205             dy00             = _mm_sub_ps(iy0,jy0);
 206             dz00             = _mm_sub_ps(iz0,jz0);
 207
 208             /* Calculate squared distance and things based on it */
 209             rsq00            = gmx_mm_calc_rsq_ps(dx00,dy00,dz00);
 210
 211             rinv00           = gmx_mm_invsqrt_ps(rsq00);
 212
 213             rinvsq00         = _mm_mul_ps(rinv00,rinv00);
 214
 215             /* Load parameters for j particles */
 216             jq0              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 217                                                               charge+jnrC+0,charge+jnrD+0);
 218             vdwjidx0A        = 2*vdwtype[jnrA+0];
 219             vdwjidx0B        = 2*vdwtype[jnrB+0];
 220             vdwjidx0C        = 2*vdwtype[jnrC+0];
 221             vdwjidx0D        = 2*vdwtype[jnrD+0];
 222
 223             /**************************
 224              * CALCULATE INTERACTIONS *
 225              **************************/
 226
 227             r00              = _mm_mul_ps(rsq00,rinv00);
 228
 229             /* Compute parameters for interactions between i and j atoms */
 230             qq00             = _mm_mul_ps(iq0,jq0);
 231             gmx_mm_load_4pair_swizzle_ps(vdwparam+vdwioffset0+vdwjidx0A,
 232                                          vdwparam+vdwioffset0+vdwjidx0B,
 233                                          vdwparam+vdwioffset0+vdwjidx0C,
 234                                          vdwparam+vdwioffset0+vdwjidx0D,
 235                                          &c6_00,&c12_00);
 236
 237             c6grid_00       = gmx_mm_load_4real_swizzle_ps(vdwgridparam+vdwioffset0+vdwjidx0A,
 238                                                                vdwgridparam+vdwioffset0+vdwjidx0B,
 239                                                                vdwgridparam+vdwioffset0+vdwjidx0C,
 240                                                                vdwgridparam+vdwioffset0+vdwjidx0D);
 241
 242             /* EWALD ELECTROSTATICS */
 243
 244             /* Analytical PME correction */
 245             zeta2            = _mm_mul_ps(beta2,rsq00);
 246             rinv3            = _mm_mul_ps(rinvsq00,rinv00);
 247             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
 248             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
 249             felec            = _mm_mul_ps(qq00,felec);
 250             pmecorrV         = gmx_mm_pmecorrV_ps(zeta2);
 251             velec            = _mm_nmacc_ps(pmecorrV,beta,rinv00);
 252             velec            = _mm_mul_ps(qq00,velec);
 253
 254             /* Analytical LJ-PME */
 255             rinvsix          = _mm_mul_ps(_mm_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 256             ewcljrsq         = _mm_mul_ps(ewclj2,rsq00);
 257             ewclj6           = _mm_mul_ps(ewclj2,_mm_mul_ps(ewclj2,ewclj2));
 258             exponent         = gmx_simd_exp_r(ewcljrsq);
 259             /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
 260             poly             = _mm_mul_ps(exponent,_mm_macc_ps(_mm_mul_ps(ewcljrsq,ewcljrsq),one_half,_mm_sub_ps(one,ewcljrsq)));
 261             /* vvdw6 = [C6 - C6grid * (1-poly)]/r6 */
 262             vvdw6            = _mm_mul_ps(_mm_macc_ps(-c6grid_00,_mm_sub_ps(one,poly),c6_00),rinvsix);
 263             vvdw12           = _mm_mul_ps(c12_00,_mm_mul_ps(rinvsix,rinvsix));
 264             vvdw             = _mm_msub_ps(vvdw12,one_twelfth,_mm_mul_ps(vvdw6,one_sixth));
 265             /* fvdw = vvdw12/r - (vvdw6/r + (C6grid * exponent * beta^6)/r) */
 266             fvdw             = _mm_mul_ps(_mm_add_ps(vvdw12,_mm_msub_ps(_mm_mul_ps(c6grid_00,one_sixth),_mm_mul_ps(exponent,ewclj6),vvdw6)),rinvsq00);
 267
 268             /* Update potential sum for this i atom from the interaction with this j atom. */
 269             velecsum         = _mm_add_ps(velecsum,velec);
 270             vvdwsum          = _mm_add_ps(vvdwsum,vvdw);
 271
 272             fscal            = _mm_add_ps(felec,fvdw);
 273
 274              /* Update vectorial force */
 275             fix0             = _mm_macc_ps(dx00,fscal,fix0);
 276             fiy0             = _mm_macc_ps(dy00,fscal,fiy0);
 277             fiz0             = _mm_macc_ps(dz00,fscal,fiz0);
 278
 279             fjptrA             = f+j_coord_offsetA;
 280             fjptrB             = f+j_coord_offsetB;
 281             fjptrC             = f+j_coord_offsetC;
 282             fjptrD             = f+j_coord_offsetD;
 283             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,
 284                                                    _mm_mul_ps(dx00,fscal),
 285                                                    _mm_mul_ps(dy00,fscal),
 286                                                    _mm_mul_ps(dz00,fscal));
 287
 288             /* Inner loop uses 53 flops */
 289         }
 290
 291         if(jidx<j_index_end)
 292         {
 293
 294             /* Get j neighbor index, and coordinate index */
 295             jnrlistA         = jjnr[jidx];
 296             jnrlistB         = jjnr[jidx+1];
 297             jnrlistC         = jjnr[jidx+2];
 298             jnrlistD         = jjnr[jidx+3];
 299             /* Sign of each element will be negative for non-real atoms.
 300              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 301              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
 302              */
 303             dummy_mask = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
 304             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 305             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 306             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 307             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 308             j_coord_offsetA  = DIM*jnrA;
 309             j_coord_offsetB  = DIM*jnrB;
 310             j_coord_offsetC  = DIM*jnrC;
 311             j_coord_offsetD  = DIM*jnrD;
 312
 313             /* load j atom coordinates */
 314             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 315                                               x+j_coord_offsetC,x+j_coord_offsetD,
 316                                               &jx0,&jy0,&jz0);
 317
 318             /* Calculate displacement vector */
 319             dx00             = _mm_sub_ps(ix0,jx0);
 320             dy00             = _mm_sub_ps(iy0,jy0);
 321             dz00             = _mm_sub_ps(iz0,jz0);
 322
 323             /* Calculate squared distance and things based on it */
 324             rsq00            = gmx_mm_calc_rsq_ps(dx00,dy00,dz00);
 325
 326             rinv00           = gmx_mm_invsqrt_ps(rsq00);
 327
 328             rinvsq00         = _mm_mul_ps(rinv00,rinv00);
 329
 330             /* Load parameters for j particles */
 331             jq0              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 332                                                               charge+jnrC+0,charge+jnrD+0);
 333             vdwjidx0A        = 2*vdwtype[jnrA+0];
 334             vdwjidx0B        = 2*vdwtype[jnrB+0];
 335             vdwjidx0C        = 2*vdwtype[jnrC+0];
 336             vdwjidx0D        = 2*vdwtype[jnrD+0];
 337
 338             /**************************
 339              * CALCULATE INTERACTIONS *
 340              **************************/
 341
 342             r00              = _mm_mul_ps(rsq00,rinv00);
 343             r00              = _mm_andnot_ps(dummy_mask,r00);
 344
 345             /* Compute parameters for interactions between i and j atoms */
 346             qq00             = _mm_mul_ps(iq0,jq0);
 347             gmx_mm_load_4pair_swizzle_ps(vdwparam+vdwioffset0+vdwjidx0A,
 348                                          vdwparam+vdwioffset0+vdwjidx0B,
 349                                          vdwparam+vdwioffset0+vdwjidx0C,
 350                                          vdwparam+vdwioffset0+vdwjidx0D,
 351                                          &c6_00,&c12_00);
 352
 353             c6grid_00       = gmx_mm_load_4real_swizzle_ps(vdwgridparam+vdwioffset0+vdwjidx0A,
 354                                                                vdwgridparam+vdwioffset0+vdwjidx0B,
 355                                                                vdwgridparam+vdwioffset0+vdwjidx0C,
 356                                                                vdwgridparam+vdwioffset0+vdwjidx0D);
 357
 358             /* EWALD ELECTROSTATICS */
 359
 360             /* Analytical PME correction */
 361             zeta2            = _mm_mul_ps(beta2,rsq00);
 362             rinv3            = _mm_mul_ps(rinvsq00,rinv00);
 363             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
 364             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
 365             felec            = _mm_mul_ps(qq00,felec);
 366             pmecorrV         = gmx_mm_pmecorrV_ps(zeta2);
 367             velec            = _mm_nmacc_ps(pmecorrV,beta,rinv00);
 368             velec            = _mm_mul_ps(qq00,velec);
 369
 370             /* Analytical LJ-PME */
 371             rinvsix          = _mm_mul_ps(_mm_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 372             ewcljrsq         = _mm_mul_ps(ewclj2,rsq00);
 373             ewclj6           = _mm_mul_ps(ewclj2,_mm_mul_ps(ewclj2,ewclj2));
 374             exponent         = gmx_simd_exp_r(ewcljrsq);
 375             /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
 376             poly             = _mm_mul_ps(exponent,_mm_macc_ps(_mm_mul_ps(ewcljrsq,ewcljrsq),one_half,_mm_sub_ps(one,ewcljrsq)));
 377             /* vvdw6 = [C6 - C6grid * (1-poly)]/r6 */
 378             vvdw6            = _mm_mul_ps(_mm_macc_ps(-c6grid_00,_mm_sub_ps(one,poly),c6_00),rinvsix);
 379             vvdw12           = _mm_mul_ps(c12_00,_mm_mul_ps(rinvsix,rinvsix));
 380             vvdw             = _mm_msub_ps(vvdw12,one_twelfth,_mm_mul_ps(vvdw6,one_sixth));
 381             /* fvdw = vvdw12/r - (vvdw6/r + (C6grid * exponent * beta^6)/r) */
 382             fvdw             = _mm_mul_ps(_mm_add_ps(vvdw12,_mm_msub_ps(_mm_mul_ps(c6grid_00,one_sixth),_mm_mul_ps(exponent,ewclj6),vvdw6)),rinvsq00);
 383
 384             /* Update potential sum for this i atom from the interaction with this j atom. */
 385             velec            = _mm_andnot_ps(dummy_mask,velec);
 386             velecsum         = _mm_add_ps(velecsum,velec);
 387             vvdw             = _mm_andnot_ps(dummy_mask,vvdw);
 388             vvdwsum          = _mm_add_ps(vvdwsum,vvdw);
 389
 390             fscal            = _mm_add_ps(felec,fvdw);
 391
 392             fscal            = _mm_andnot_ps(dummy_mask,fscal);
 393
 394              /* Update vectorial force */
 395             fix0             = _mm_macc_ps(dx00,fscal,fix0);
 396             fiy0             = _mm_macc_ps(dy00,fscal,fiy0);
 397             fiz0             = _mm_macc_ps(dz00,fscal,fiz0);
 398
 399             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 400             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 401             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 402             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 403             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,
 404                                                    _mm_mul_ps(dx00,fscal),
 405                                                    _mm_mul_ps(dy00,fscal),
 406                                                    _mm_mul_ps(dz00,fscal));
 407
 408             /* Inner loop uses 54 flops */
 409         }
 410
 411         /* End of innermost loop */
 412
 413         gmx_mm_update_iforce_1atom_swizzle_ps(fix0,fiy0,fiz0,
 414                                               f+i_coord_offset,fshift+i_shift_offset);
 415
 416         ggid                        = gid[iidx];
 417         /* Update potential energies */
 418         gmx_mm_update_1pot_ps(velecsum,kernel_data->energygrp_elec+ggid);
 419         gmx_mm_update_1pot_ps(vvdwsum,kernel_data->energygrp_vdw+ggid);
 420
 421         /* Increment number of inner iterations */
 422         inneriter                  += j_index_end - j_index_start;
 423
 424         /* Outer loop uses 9 flops */
 425     }
 426
 427     /* Increment number of outer iterations */
 428     outeriter        += nri;
 429
 430     /* Update outer/inner flops */
 431
 432     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_VF,outeriter*9 + inneriter*54);
 433 }
 434 /*
 435  * Gromacs nonbonded kernel:   nb_kernel_ElecEw_VdwLJEw_GeomP1P1_F_avx_128_fma_single
 436  * Electrostatics interaction: Ewald
 437  * VdW interaction:            LJEwald
 438  * Geometry:                   Particle-Particle
 439  * Calculate force/pot:        Force
 440  */
 441 void
 442 nb_kernel_ElecEw_VdwLJEw_GeomP1P1_F_avx_128_fma_single
 443                     (t_nblist                    * gmx_restrict       nlist,
 444                      rvec                        * gmx_restrict          xx,
 445                      rvec                        * gmx_restrict          ff,
 446                      t_forcerec                  * gmx_restrict          fr,
 447                      t_mdatoms                   * gmx_restrict     mdatoms,
 448                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 449                      t_nrnb                      * gmx_restrict        nrnb)
 450 {
 451     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
 452      * just 0 for non-waters.
 453      * Suffixes A,B,C,D refer to j loop unrolling done with AVX_128, e.g. for the four different
 454      * jnr indices corresponding to data put in the four positions in the SIMD register.
 455      */
 456     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
 457     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
 458     int              jnrA,jnrB,jnrC,jnrD;
 459     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
 460     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
 461     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 462     real             rcutoff_scalar;
 463     real             *shiftvec,*fshift,*x,*f;
 464     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
 465     real             scratch[4*DIM];
 466     __m128           fscal,rcutoff,rcutoff2,jidxall;
 467     int              vdwioffset0;
 468     __m128           ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 469     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
 470     __m128           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 471     __m128           dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
 472     __m128           velec,felec,velecsum,facel,crf,krf,krf2;
 473     real             *charge;
 474     int              nvdwtype;
 475     __m128           rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 476     int              *vdwtype;
 477     real             *vdwparam;
 478     __m128           one_sixth   = _mm_set1_ps(1.0/6.0);
 479     __m128           one_twelfth = _mm_set1_ps(1.0/12.0);
 480     __m128           c6grid_00;
 481     real             *vdwgridparam;
 482     __m128           ewclj,ewclj2,ewclj6,ewcljrsq,poly,exponent,f6A,f6B,sh_lj_ewald;
 483     __m128           one_half = _mm_set1_ps(0.5);
 484     __m128           minus_one = _mm_set1_ps(-1.0);
 485     __m128i          ewitab;
 486     __m128           ewtabscale,eweps,twoeweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
 487     __m128           beta,beta2,beta3,zeta2,pmecorrF,pmecorrV,rinv3;
 488     real             *ewtab;
 489     __m128           dummy_mask,cutoff_mask;
 490     __m128           signbit = _mm_castsi128_ps( _mm_set1_epi32(0x80000000) );
 491     __m128           one     = _mm_set1_ps(1.0);
 492     __m128           two     = _mm_set1_ps(2.0);
 493     x                = xx[0];
 494     f                = ff[0];
 495
 496     nri              = nlist->nri;
 497     iinr             = nlist->iinr;
 498     jindex           = nlist->jindex;
 499     jjnr             = nlist->jjnr;
 500     shiftidx         = nlist->shift;
 501     gid              = nlist->gid;
 502     shiftvec         = fr->shift_vec[0];
 503     fshift           = fr->fshift[0];
 504     facel            = _mm_set1_ps(fr->epsfac);
 505     charge           = mdatoms->chargeA;
 506     nvdwtype         = fr->ntype;
 507     vdwparam         = fr->nbfp;
 508     vdwtype          = mdatoms->typeA;
 509     vdwgridparam     = fr->ljpme_c6grid;
 510     sh_lj_ewald      = _mm_set1_ps(fr->ic->sh_lj_ewald);
 511     ewclj            = _mm_set1_ps(fr->ewaldcoeff_lj);
 512     ewclj2           = _mm_mul_ps(minus_one,_mm_mul_ps(ewclj,ewclj));
 513
 514     sh_ewald         = _mm_set1_ps(fr->ic->sh_ewald);
 515     beta             = _mm_set1_ps(fr->ic->ewaldcoeff_q);
 516     beta2            = _mm_mul_ps(beta,beta);
 517     beta3            = _mm_mul_ps(beta,beta2);
 518     ewtab            = fr->ic->tabq_coul_F;
 519     ewtabscale       = _mm_set1_ps(fr->ic->tabq_scale);
 520     ewtabhalfspace   = _mm_set1_ps(0.5/fr->ic->tabq_scale);
 521
 522     /* Avoid stupid compiler warnings */
 523     jnrA = jnrB = jnrC = jnrD = 0;
 524     j_coord_offsetA = 0;
 525     j_coord_offsetB = 0;
 526     j_coord_offsetC = 0;
 527     j_coord_offsetD = 0;
 528
 529     outeriter        = 0;
 530     inneriter        = 0;
 531
 532     for(iidx=0;iidx<4*DIM;iidx++)
 533     {
 534         scratch[iidx] = 0.0;
 535     }
 536
 537     /* Start outer loop over neighborlists */
 538     for(iidx=0; iidx<nri; iidx++)
 539     {
 540         /* Load shift vector for this list */
 541         i_shift_offset   = DIM*shiftidx[iidx];
 542
 543         /* Load limits for loop over neighbors */
 544         j_index_start    = jindex[iidx];
 545         j_index_end      = jindex[iidx+1];
 546
 547         /* Get outer coordinate index */
 548         inr              = iinr[iidx];
 549         i_coord_offset   = DIM*inr;
 550
 551         /* Load i particle coords and add shift vector */
 552         gmx_mm_load_shift_and_1rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,&ix0,&iy0,&iz0);
 553
 554         fix0             = _mm_setzero_ps();
 555         fiy0             = _mm_setzero_ps();
 556         fiz0             = _mm_setzero_ps();
 557
 558         /* Load parameters for i particles */
 559         iq0              = _mm_mul_ps(facel,_mm_load1_ps(charge+inr+0));
 560         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 561
 562         /* Start inner kernel loop */
 563         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 564         {
 565
 566             /* Get j neighbor index, and coordinate index */
 567             jnrA             = jjnr[jidx];
 568             jnrB             = jjnr[jidx+1];
 569             jnrC             = jjnr[jidx+2];
 570             jnrD             = jjnr[jidx+3];
 571             j_coord_offsetA  = DIM*jnrA;
 572             j_coord_offsetB  = DIM*jnrB;
 573             j_coord_offsetC  = DIM*jnrC;
 574             j_coord_offsetD  = DIM*jnrD;
 575
 576             /* load j atom coordinates */
 577             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 578                                               x+j_coord_offsetC,x+j_coord_offsetD,
 579                                               &jx0,&jy0,&jz0);
 580
 581             /* Calculate displacement vector */
 582             dx00             = _mm_sub_ps(ix0,jx0);
 583             dy00             = _mm_sub_ps(iy0,jy0);
 584             dz00             = _mm_sub_ps(iz0,jz0);
 585
 586             /* Calculate squared distance and things based on it */
 587             rsq00            = gmx_mm_calc_rsq_ps(dx00,dy00,dz00);
 588
 589             rinv00           = gmx_mm_invsqrt_ps(rsq00);
 590
 591             rinvsq00         = _mm_mul_ps(rinv00,rinv00);
 592
 593             /* Load parameters for j particles */
 594             jq0              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 595                                                               charge+jnrC+0,charge+jnrD+0);
 596             vdwjidx0A        = 2*vdwtype[jnrA+0];
 597             vdwjidx0B        = 2*vdwtype[jnrB+0];
 598             vdwjidx0C        = 2*vdwtype[jnrC+0];
 599             vdwjidx0D        = 2*vdwtype[jnrD+0];
 600
 601             /**************************
 602              * CALCULATE INTERACTIONS *
 603              **************************/
 604
 605             r00              = _mm_mul_ps(rsq00,rinv00);
 606
 607             /* Compute parameters for interactions between i and j atoms */
 608             qq00             = _mm_mul_ps(iq0,jq0);
 609             gmx_mm_load_4pair_swizzle_ps(vdwparam+vdwioffset0+vdwjidx0A,
 610                                          vdwparam+vdwioffset0+vdwjidx0B,
 611                                          vdwparam+vdwioffset0+vdwjidx0C,
 612                                          vdwparam+vdwioffset0+vdwjidx0D,
 613                                          &c6_00,&c12_00);
 614
 615             c6grid_00       = gmx_mm_load_4real_swizzle_ps(vdwgridparam+vdwioffset0+vdwjidx0A,
 616                                                                vdwgridparam+vdwioffset0+vdwjidx0B,
 617                                                                vdwgridparam+vdwioffset0+vdwjidx0C,
 618                                                                vdwgridparam+vdwioffset0+vdwjidx0D);
 619
 620             /* EWALD ELECTROSTATICS */
 621
 622             /* Analytical PME correction */
 623             zeta2            = _mm_mul_ps(beta2,rsq00);
 624             rinv3            = _mm_mul_ps(rinvsq00,rinv00);
 625             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
 626             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
 627             felec            = _mm_mul_ps(qq00,felec);
 628
 629             /* Analytical LJ-PME */
 630             rinvsix          = _mm_mul_ps(_mm_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 631             ewcljrsq         = _mm_mul_ps(ewclj2,rsq00);
 632             ewclj6           = _mm_mul_ps(ewclj2,_mm_mul_ps(ewclj2,ewclj2));
 633             exponent         = gmx_simd_exp_r(ewcljrsq);
 634             /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
 635             poly             = _mm_mul_ps(exponent,_mm_macc_ps(_mm_mul_ps(ewcljrsq,ewcljrsq),one_half,_mm_sub_ps(one,ewcljrsq)));
 636             /* f6A = 6 * C6grid * (1 - poly) */
 637             f6A              = _mm_mul_ps(c6grid_00,_mm_sub_ps(one,poly));
 638             /* f6B = C6grid * exponent * beta^6 */
 639             f6B              = _mm_mul_ps(_mm_mul_ps(c6grid_00,one_sixth),_mm_mul_ps(exponent,ewclj6));
 640             /* fvdw = 12*C12/r13 - ((6*C6 - f6A)/r6 + f6B)/r */
 641             fvdw              = _mm_mul_ps(_mm_macc_ps(_mm_msub_ps(c12_00,rinvsix,_mm_sub_ps(c6_00,f6A)),rinvsix,f6B),rinvsq00);
 642
 643             fscal            = _mm_add_ps(felec,fvdw);
 644
 645              /* Update vectorial force */
 646             fix0             = _mm_macc_ps(dx00,fscal,fix0);
 647             fiy0             = _mm_macc_ps(dy00,fscal,fiy0);
 648             fiz0             = _mm_macc_ps(dz00,fscal,fiz0);
 649
 650             fjptrA             = f+j_coord_offsetA;
 651             fjptrB             = f+j_coord_offsetB;
 652             fjptrC             = f+j_coord_offsetC;
 653             fjptrD             = f+j_coord_offsetD;
 654             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,
 655                                                    _mm_mul_ps(dx00,fscal),
 656                                                    _mm_mul_ps(dy00,fscal),
 657                                                    _mm_mul_ps(dz00,fscal));
 658
 659             /* Inner loop uses 49 flops */
 660         }
 661
 662         if(jidx<j_index_end)
 663         {
 664
 665             /* Get j neighbor index, and coordinate index */
 666             jnrlistA         = jjnr[jidx];
 667             jnrlistB         = jjnr[jidx+1];
 668             jnrlistC         = jjnr[jidx+2];
 669             jnrlistD         = jjnr[jidx+3];
 670             /* Sign of each element will be negative for non-real atoms.
 671              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 672              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
 673              */
 674             dummy_mask = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
 675             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 676             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 677             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 678             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 679             j_coord_offsetA  = DIM*jnrA;
 680             j_coord_offsetB  = DIM*jnrB;
 681             j_coord_offsetC  = DIM*jnrC;
 682             j_coord_offsetD  = DIM*jnrD;
 683
 684             /* load j atom coordinates */
 685             gmx_mm_load_1rvec_4ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 686                                               x+j_coord_offsetC,x+j_coord_offsetD,
 687                                               &jx0,&jy0,&jz0);
 688
 689             /* Calculate displacement vector */
 690             dx00             = _mm_sub_ps(ix0,jx0);
 691             dy00             = _mm_sub_ps(iy0,jy0);
 692             dz00             = _mm_sub_ps(iz0,jz0);
 693
 694             /* Calculate squared distance and things based on it */
 695             rsq00            = gmx_mm_calc_rsq_ps(dx00,dy00,dz00);
 696
 697             rinv00           = gmx_mm_invsqrt_ps(rsq00);
 698
 699             rinvsq00         = _mm_mul_ps(rinv00,rinv00);
 700
 701             /* Load parameters for j particles */
 702             jq0              = gmx_mm_load_4real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 703                                                               charge+jnrC+0,charge+jnrD+0);
 704             vdwjidx0A        = 2*vdwtype[jnrA+0];
 705             vdwjidx0B        = 2*vdwtype[jnrB+0];
 706             vdwjidx0C        = 2*vdwtype[jnrC+0];
 707             vdwjidx0D        = 2*vdwtype[jnrD+0];
 708
 709             /**************************
 710              * CALCULATE INTERACTIONS *
 711              **************************/
 712
 713             r00              = _mm_mul_ps(rsq00,rinv00);
 714             r00              = _mm_andnot_ps(dummy_mask,r00);
 715
 716             /* Compute parameters for interactions between i and j atoms */
 717             qq00             = _mm_mul_ps(iq0,jq0);
 718             gmx_mm_load_4pair_swizzle_ps(vdwparam+vdwioffset0+vdwjidx0A,
 719                                          vdwparam+vdwioffset0+vdwjidx0B,
 720                                          vdwparam+vdwioffset0+vdwjidx0C,
 721                                          vdwparam+vdwioffset0+vdwjidx0D,
 722                                          &c6_00,&c12_00);
 723
 724             c6grid_00       = gmx_mm_load_4real_swizzle_ps(vdwgridparam+vdwioffset0+vdwjidx0A,
 725                                                                vdwgridparam+vdwioffset0+vdwjidx0B,
 726                                                                vdwgridparam+vdwioffset0+vdwjidx0C,
 727                                                                vdwgridparam+vdwioffset0+vdwjidx0D);
 728
 729             /* EWALD ELECTROSTATICS */
 730
 731             /* Analytical PME correction */
 732             zeta2            = _mm_mul_ps(beta2,rsq00);
 733             rinv3            = _mm_mul_ps(rinvsq00,rinv00);
 734             pmecorrF         = gmx_mm_pmecorrF_ps(zeta2);
 735             felec            = _mm_macc_ps(pmecorrF,beta3,rinv3);
 736             felec            = _mm_mul_ps(qq00,felec);
 737
 738             /* Analytical LJ-PME */
 739             rinvsix          = _mm_mul_ps(_mm_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 740             ewcljrsq         = _mm_mul_ps(ewclj2,rsq00);
 741             ewclj6           = _mm_mul_ps(ewclj2,_mm_mul_ps(ewclj2,ewclj2));
 742             exponent         = gmx_simd_exp_r(ewcljrsq);
 743             /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
 744             poly             = _mm_mul_ps(exponent,_mm_macc_ps(_mm_mul_ps(ewcljrsq,ewcljrsq),one_half,_mm_sub_ps(one,ewcljrsq)));
 745             /* f6A = 6 * C6grid * (1 - poly) */
 746             f6A              = _mm_mul_ps(c6grid_00,_mm_sub_ps(one,poly));
 747             /* f6B = C6grid * exponent * beta^6 */
 748             f6B              = _mm_mul_ps(_mm_mul_ps(c6grid_00,one_sixth),_mm_mul_ps(exponent,ewclj6));
 749             /* fvdw = 12*C12/r13 - ((6*C6 - f6A)/r6 + f6B)/r */
 750             fvdw              = _mm_mul_ps(_mm_macc_ps(_mm_msub_ps(c12_00,rinvsix,_mm_sub_ps(c6_00,f6A)),rinvsix,f6B),rinvsq00);
 751
 752             fscal            = _mm_add_ps(felec,fvdw);
 753
 754             fscal            = _mm_andnot_ps(dummy_mask,fscal);
 755
 756              /* Update vectorial force */
 757             fix0             = _mm_macc_ps(dx00,fscal,fix0);
 758             fiy0             = _mm_macc_ps(dy00,fscal,fiy0);
 759             fiz0             = _mm_macc_ps(dz00,fscal,fiz0);
 760
 761             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 762             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 763             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 764             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 765             gmx_mm_decrement_1rvec_4ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,
 766                                                    _mm_mul_ps(dx00,fscal),
 767                                                    _mm_mul_ps(dy00,fscal),
 768                                                    _mm_mul_ps(dz00,fscal));
 769
 770             /* Inner loop uses 50 flops */
 771         }
 772
 773         /* End of innermost loop */
 774
 775         gmx_mm_update_iforce_1atom_swizzle_ps(fix0,fiy0,fiz0,
 776                                               f+i_coord_offset,fshift+i_shift_offset);
 777
 778         /* Increment number of inner iterations */
 779         inneriter                  += j_index_end - j_index_start;
 780
 781         /* Outer loop uses 7 flops */
 782     }
 783
 784     /* Increment number of outer iterations */
 785     outeriter        += nri;
 786
 787     /* Update outer/inner flops */
 788
 789     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_F,outeriter*7 + inneriter*50);
 790 }