2 * This file is part of the GROMACS molecular simulation package.
4 * Copyright (c) 2020,2021, 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.
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.
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.
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.
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.
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.
37 * Helper functions and constants for SYCL NBNXM kernels
39 * \ingroup module_nbnxm
41 #ifndef GMX_NBNXM_SYCL_NBNXN_SYCL_KERNEL_UTILS_H
42 #define GMX_NBNXM_SYCL_NBNXN_SYCL_KERNEL_UTILS_H
44 #include "gromacs/gpu_utils/sycl_kernel_utils.h"
45 #include "gromacs/nbnxm/pairlist.h"
46 #include "gromacs/nbnxm/pairlistparams.h"
51 #ifndef GMX_NBNXN_PRUNE_KERNEL_J4_CONCURRENCY
52 # define GMX_NBNXN_PRUNE_KERNEL_J4_CONCURRENCY 4
54 /*! \brief Macro defining default for the prune kernel's j4 processing concurrency.
56 * The GMX_NBNXN_PRUNE_KERNEL_J4_CONCURRENCY macro allows compile-time override.
58 static constexpr int c_syclPruneKernelJ4Concurrency = GMX_NBNXN_PRUNE_KERNEL_J4_CONCURRENCY;
60 /* Convenience constants */
62 // cluster size = number of atoms per cluster.
63 static constexpr int c_clSize = c_nbnxnGpuClusterSize;
64 // j-cluster size after split (4 in the current implementation).
65 static constexpr int c_splitClSize = c_clSize / c_nbnxnGpuClusterpairSplit;
66 // i-cluster interaction mask for a super-cluster with all c_nbnxnGpuNumClusterPerSupercluster=8 bits set.
67 static constexpr unsigned superClInteractionMask = ((1U << c_nbnxnGpuNumClusterPerSupercluster) - 1U);
69 // 1/sqrt(pi), same value as \c M_FLOAT_1_SQRTPI in other NB kernels.
70 static constexpr float c_OneOverSqrtPi = 0.564189583547756F;
71 // 1/6, same value as in other NB kernels.
72 static constexpr float c_oneSixth = 0.16666667F;
73 // 1/12, same value as in other NB kernels.
74 static constexpr float c_oneTwelfth = 0.08333333F;
77 /* The following functions are necessary because on some versions of Intel OpenCL RT, subgroups
78 * do not properly work (segfault or create subgroups of size 1) if used in kernels
79 * with non-1-dimensional workgroup. */
80 //! \brief Convert 3D range to 1D
81 static inline cl::sycl::range<1> flattenRange(cl::sycl::range<3> range3d)
83 return cl::sycl::range<1>(range3d.size());
86 //! \brief Convert 3D nd_range to 1D
87 static inline cl::sycl::nd_range<1> flattenNDRange(cl::sycl::nd_range<3> nd_range3d)
89 return cl::sycl::nd_range<1>(flattenRange(nd_range3d.get_global_range()),
90 flattenRange(nd_range3d.get_local_range()));
93 //! \brief Convert flattened 1D index to 3D
94 template<int rangeX, int rangeY>
95 static inline cl::sycl::id<3> unflattenId(cl::sycl::id<1> id1d)
97 constexpr unsigned rangeXY = rangeX * rangeY;
98 const unsigned id = id1d[0];
99 const unsigned z = id / rangeXY;
100 const unsigned xy = id % rangeXY;
101 return cl::sycl::id<3>(xy % rangeX, xy / rangeX, z);
106 #endif // GMX_NBNXM_SYCL_NBNXN_SYCL_KERNEL_UTILS_H