src/gromacs/nbnxm/pairlist.h

   1 /*
   2  * This file is part of the GROMACS molecular simulation package.
   3  *
   4  * Copyright (c) 2012,2013,2014,2015,2016 by the GROMACS development team.
   5  * Copyright (c) 2017,2018,2019,2020, by the GROMACS development team, led by
   6  * Mark Abraham, David van der Spoel, Berk Hess, and Erik Lindahl,
   7  * and including many others, as listed in the AUTHORS file in the
   8  * top-level source directory and at http://www.gromacs.org.
   9  *
  10  * GROMACS is free software; you can redistribute it and/or
  11  * modify it under the terms of the GNU Lesser General Public License
  12  * as published by the Free Software Foundation; either version 2.1
  13  * of the License, or (at your option) any later version.
  14  *
  15  * GROMACS is distributed in the hope that it will be useful,
  16  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  17  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  18  * Lesser General Public License for more details.
  19  *
  20  * You should have received a copy of the GNU Lesser General Public
  21  * License along with GROMACS; if not, see
  22  * http://www.gnu.org/licenses, or write to the Free Software Foundation,
  23  * Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA.
  24  *
  25  * If you want to redistribute modifications to GROMACS, please
  26  * consider that scientific software is very special. Version
  27  * control is crucial - bugs must be traceable. We will be happy to
  28  * consider code for inclusion in the official distribution, but
  29  * derived work must not be called official GROMACS. Details are found
  30  * in the README & COPYING files - if they are missing, get the
  31  * official version at http://www.gromacs.org.
  32  *
  33  * To help us fund GROMACS development, we humbly ask that you cite
  34  * the research papers on the package. Check out http://www.gromacs.org.
  35  */
  36
  37 #ifndef GMX_NBNXM_PAIRLIST_H
  38 #define GMX_NBNXM_PAIRLIST_H
  39
  40 #include <cstddef>
  41
  42 #include "gromacs/gpu_utils/hostallocator.h"
  43 #include "gromacs/math/vectypes.h"
  44 #include "gromacs/mdtypes/locality.h"
  45 #include "gromacs/mdtypes/nblist.h"
  46 #include "gromacs/utility/basedefinitions.h"
  47 #include "gromacs/utility/defaultinitializationallocator.h"
  48 #include "gromacs/utility/enumerationhelpers.h"
  49 #include "gromacs/utility/real.h"
  50
  51 // This file with constants is separate from this file to be able
  52 // to include it during OpenCL jitting without including config.h
  53 #include "constants.h"
  54 #include "pairlistparams.h"
  55
  56 struct NbnxnPairlistCpuWork;
  57 struct NbnxnPairlistGpuWork;
  58
  59
  60 /* Convenience type for vector with aligned memory */
  61 template<typename T>
  62 using AlignedVector = std::vector<T, gmx::AlignedAllocator<T>>;
  63
  64 /* Convenience type for vector that avoids initialization at resize() */
  65 template<typename T>
  66 using FastVector = std::vector<T, gmx::DefaultInitializationAllocator<T>>;
  67
  68 /* A buffer data structure of 64 bytes
  69  * to be placed at the beginning and end of structs
  70  * to avoid cache invalidation of the real contents
  71  * of the struct by writes to neighboring memory.
  72  */
  73 typedef struct
  74 {
  75     int dummy[16];
  76 } gmx_cache_protect_t;
  77
  78 /* This is the actual cluster-pair list j-entry.
  79  * cj is the j-cluster.
  80  * The interaction bits in excl are indexed i-major, j-minor.
  81  * The cj entries are sorted such that ones with exclusions come first.
  82  * This means that once a full mask (=NBNXN_INTERACTION_MASK_ALL)
  83  * is found, all subsequent j-entries in the i-entry also have full masks.
  84  */
  85 struct nbnxn_cj_t
  86 {
  87     int          cj;   /* The j-cluster                    */
  88     unsigned int excl; /* The exclusion (interaction) bits */
  89 };
  90
  91 /* In nbnxn_ci_t the integer shift contains the shift in the lower 7 bits.
  92  * The upper bits contain information for non-bonded kernel optimization.
  93  * Simply calculating LJ and Coulomb for all pairs in a cluster pair is fine.
  94  * But three flags can be used to skip interactions, currently only for subc=0
  95  * !(shift & NBNXN_CI_DO_LJ(subc))   => we can skip LJ for all pairs
  96  * shift & NBNXN_CI_HALF_LJ(subc)    => we can skip LJ for the second half of i
  97  * !(shift & NBNXN_CI_DO_COUL(subc)) => we can skip Coulomb for all pairs
  98  */
  99 #define NBNXN_CI_SHIFT 127
 100 #define NBNXN_CI_DO_LJ(subc) (1 << (7 + 3 * (subc)))
 101 #define NBNXN_CI_HALF_LJ(subc) (1 << (8 + 3 * (subc)))
 102 #define NBNXN_CI_DO_COUL(subc) (1 << (9 + 3 * (subc)))
 103
 104 /* Cluster-pair Interaction masks
 105  * Bit i*j-cluster-size + j tells if atom i and j interact.
 106  */
 107 // TODO: Rename according to convention when moving into Nbnxn namespace
 108 /* All interaction mask is the same for all kernels */
 109 constexpr unsigned int NBNXN_INTERACTION_MASK_ALL = 0xffffffffU;
 110 /* 4x4 kernel diagonal mask */
 111 constexpr unsigned int NBNXN_INTERACTION_MASK_DIAG = 0x08ceU;
 112 /* 4x2 kernel diagonal masks */
 113 constexpr unsigned int NBNXN_INTERACTION_MASK_DIAG_J2_0 = 0x0002U;
 114 constexpr unsigned int NBNXN_INTERACTION_MASK_DIAG_J2_1 = 0x002fU;
 115 /* 4x8 kernel diagonal masks */
 116 constexpr unsigned int NBNXN_INTERACTION_MASK_DIAG_J8_0 = 0xf0f8fcfeU;
 117 constexpr unsigned int NBNXN_INTERACTION_MASK_DIAG_J8_1 = 0x0080c0e0U;
 118
 119 /* Simple pair-list i-unit */
 120 struct nbnxn_ci_t
 121 {
 122     int ci;           /* i-cluster             */
 123     int shift;        /* Shift vector index plus possible flags, see above */
 124     int cj_ind_start; /* Start index into cj   */
 125     int cj_ind_end;   /* End index into cj     */
 126 };
 127
 128 /* Grouped pair-list i-unit */
 129 typedef struct
 130 {
 131     /* Returns the number of j-cluster groups in this entry */
 132     int numJClusterGroups() const { return cj4_ind_end - cj4_ind_start; }
 133
 134     int sci;           /* i-super-cluster       */
 135     int shift;         /* Shift vector index plus possible flags */
 136     int cj4_ind_start; /* Start index into cj4  */
 137     int cj4_ind_end;   /* End index into cj4    */
 138 } nbnxn_sci_t;
 139
 140 /* Interaction data for a j-group for one warp */
 141 struct nbnxn_im_ei_t
 142 {
 143     // The i-cluster interactions mask for 1 warp
 144     unsigned int imask = 0U;
 145     // Index into the exclusion array for 1 warp, default index 0 which means no exclusions
 146     int excl_ind = 0;
 147 };
 148
 149 typedef struct
 150 {
 151     int           cj[c_nbnxnGpuJgroupSize];         /* The 4 j-clusters */
 152     nbnxn_im_ei_t imei[c_nbnxnGpuClusterpairSplit]; /* The i-cluster mask data       for 2 warps */
 153 } nbnxn_cj4_t;
 154
 155 /* Struct for storing the atom-pair interaction bits for a cluster pair in a GPU pairlist */
 156 struct nbnxn_excl_t
 157 {
 158     /* Constructor, sets no exclusions, so all atom pairs interacting */
 159     nbnxn_excl_t()
 160     {
 161         for (unsigned int& pairEntry : pair)
 162         {
 163             pairEntry = NBNXN_INTERACTION_MASK_ALL;
 164         }
 165     }
 166
 167     /* Topology exclusion interaction bits per warp */
 168     unsigned int pair[c_nbnxnGpuExclSize];
 169 };
 170
 171 /* Cluster pairlist type for use on CPUs */
 172 struct NbnxnPairlistCpu
 173 {
 174     NbnxnPairlistCpu();
 175
 176     gmx_cache_protect_t cp0;
 177
 178     int                    na_ci;   /* The number of atoms per i-cluster        */
 179     int                    na_cj;   /* The number of atoms per j-cluster        */
 180     real                   rlist;   /* The radius for constructing the list     */
 181     FastVector<nbnxn_ci_t> ci;      /* The i-cluster list                       */
 182     FastVector<nbnxn_ci_t> ciOuter; /* The outer, unpruned i-cluster list       */
 183
 184     FastVector<nbnxn_cj_t> cj;      /* The j-cluster list, size ncj             */
 185     FastVector<nbnxn_cj_t> cjOuter; /* The outer, unpruned j-cluster list       */
 186     int                    ncjInUse; /* The number of j-clusters that are used by ci entries in this list, will be <= cj.size() */
 187
 188     int nci_tot; /* The total number of i clusters           */
 189
 190     /* Working data storage for list construction */
 191     std::unique_ptr<NbnxnPairlistCpuWork> work;
 192
 193     gmx_cache_protect_t cp1;
 194 };
 195
 196 /* Cluster pairlist type, with extra hierarchies, for on the GPU
 197  *
 198  * NOTE: for better performance when combining lists over threads,
 199  *       all vectors should use default initialization. But when
 200  *       changing this, excl should be intialized when adding entries.
 201  */
 202 struct NbnxnPairlistGpu
 203 {
 204     /* Constructor
 205      *
 206      * \param[in] pinningPolicy  Sets the pinning policy for all buffers used on the GPU
 207      */
 208     NbnxnPairlistGpu(gmx::PinningPolicy pinningPolicy);
 209
 210     gmx_cache_protect_t cp0;
 211
 212     int  na_ci; /* The number of atoms per i-cluster        */
 213     int  na_cj; /* The number of atoms per j-cluster        */
 214     int  na_sc; /* The number of atoms per super cluster    */
 215     real rlist; /* The radius for constructing the list     */
 216     // The i-super-cluster list, indexes into cj4;
 217     gmx::HostVector<nbnxn_sci_t> sci;
 218     // The list of 4*j-cluster groups
 219     gmx::HostVector<nbnxn_cj4_t> cj4;
 220     // Atom interaction bits (non-exclusions)
 221     gmx::HostVector<nbnxn_excl_t> excl;
 222     // The total number of i-clusters
 223     int nci_tot;
 224
 225     /* Working data storage for list construction */
 226     std::unique_ptr<NbnxnPairlistGpuWork> work;
 227
 228     gmx_cache_protect_t cp1;
 229 };
 230
 231 //! Initializes a free-energy pair-list
 232 void nbnxn_init_pairlist_fep(t_nblist* nl);
 233
 234 #endif