--- /dev/null
+/*
+ * This file is part of the GROMACS molecular simulation package.
+ *
+ * Copyright (c) 2019, by the GROMACS development team, led by
+ * Mark Abraham, David van der Spoel, Berk Hess, and Erik Lindahl,
+ * and including many others, as listed in the AUTHORS file in the
+ * top-level source directory and at http://www.gromacs.org.
+ *
+ * GROMACS is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public License
+ * as published by the Free Software Foundation; either version 2.1
+ * of the License, or (at your option) any later version.
+ *
+ * GROMACS is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with GROMACS; if not, see
+ * http://www.gnu.org/licenses, or write to the Free Software Foundation,
+ * Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
+ *
+ * If you want to redistribute modifications to GROMACS, please
+ * consider that scientific software is very special. Version
+ * control is crucial - bugs must be traceable. We will be happy to
+ * consider code for inclusion in the official distribution, but
+ * derived work must not be called official GROMACS. Details are found
+ * in the README & COPYING files - if they are missing, get the
+ * official version at http://www.gromacs.org.
+ *
+ * To help us fund GROMACS development, we humbly ask that you cite
+ * the research papers on the package. Check out http://www.gromacs.org.
+ */
+
+/*! \internal \file
+ *
+ * \brief
+ * Declares the ClusterDistanceKernelType enum
+ *
+ * \author Berk Hess <hess@kth.se>
+ * \ingroup module_nbnxm
+ */
+
+#ifndef GMX_NBNXM_CLUSTERDISTANCEKERNELTYPE_H
+#define GMX_NBNXM_CLUSTERDISTANCEKERNELTYPE_H
+
+#include "gromacs/nbnxm/atomdata.h"
+#include "gromacs/simd/simd.h"
+#include "gromacs/utility/gmxassert.h"
+
+#include "pairlistparams.h"
+
+//! The types of kernel for calculating the distance between pairs of atom clusters
+enum class ClusterDistanceKernelType : int
+{
+ CpuPlainC, //!< Plain-C for CPU list
+ CpuSimd_4xM, //!< SIMD for CPU list for j-cluster size matching the SIMD width
+ CpuSimd_2xMM, //!< SIMD for CPU list for j-cluster size matching half the SIMD width
+ Gpu //!< For GPU list, can be either plain-C or SIMD
+};
+
+//! Return the cluster distance kernel type given the pairlist type and atomdata
+static inline ClusterDistanceKernelType
+getClusterDistanceKernelType(const PairlistType pairlistType,
+ const nbnxn_atomdata_t &atomdata)
+{
+ if (pairlistType == PairlistType::HierarchicalNxN)
+ {
+ return ClusterDistanceKernelType::Gpu;
+ }
+ else if (atomdata.XFormat == nbatXYZ)
+ {
+ return ClusterDistanceKernelType::CpuPlainC;
+ }
+ else if (pairlistType == PairlistType::Simple4x2)
+ {
+#if GMX_SIMD && GMX_SIMD_REAL_WIDTH == 2
+ return ClusterDistanceKernelType::CpuSimd_4xM;
+#else
+ GMX_RELEASE_ASSERT(false, "Expect 2-wide SIMD with 4x2 list and nbat SIMD layout");
+#endif
+ }
+ else if (pairlistType == PairlistType::Simple4x4)
+ {
+#if GMX_SIMD && GMX_SIMD_REAL_WIDTH == 4
+ return ClusterDistanceKernelType::CpuSimd_4xM;
+#elif GMX_SIMD && GMX_SIMD_REAL_WIDTH == 8
+ return ClusterDistanceKernelType::CpuSimd_2xMM;
+#else
+ GMX_RELEASE_ASSERT(false, "Expect 4-wide or 8-wide SIMD with 4x4 list and nbat SIMD layout");
+#endif
+ }
+ else
+ {
+ GMX_ASSERT(pairlistType == PairlistType::Simple4x8, "Unhandled pairlist type");
+#if GMX_SIMD && GMX_SIMD_REAL_WIDTH == 8
+ return ClusterDistanceKernelType::CpuSimd_4xM;
+#elif GMX_SIMD && GMX_SIMD_REAL_WIDTH == 16
+ return ClusterDistanceKernelType::CpuSimd_2xMM;
+#else
+ GMX_RELEASE_ASSERT(false, "Expect 8-wide or 16-wide SIMD with 4x4 list and nbat SIMD layout");
+#endif
+ }
+
+ GMX_RELEASE_ASSERT(false, "We should have returned before getting here");
+ return ClusterDistanceKernelType::CpuPlainC;
+}
+
+#endif
#include "gromacs/simd/simd.h"
#include "gromacs/simd/vector_operations.h"
+#include "pairlistparams.h"
+
namespace Nbnxm
{
#include "gromacs/mdtypes/md_enums.h"
#include "gromacs/nbnxm/atomdata.h"
#include "gromacs/nbnxm/gpu_data_mgmt.h"
-#include "gromacs/nbnxm/nbnxm.h"
#include "gromacs/nbnxm/nbnxm_geometry.h"
#include "gromacs/nbnxm/nbnxm_simd.h"
#include "gromacs/pbcutil/ishift.h"
#include "gromacs/utility/gmxomp.h"
#include "gromacs/utility/smalloc.h"
+#include "clusterdistancekerneltype.h"
#include "gridset.h"
#include "pairlistset.h"
#include "pairlistsets.h"
static void icell_set_x(int ci,
real shx, real shy, real shz,
int stride, const real *x,
- const Nbnxm::KernelType kernelType,
+ const ClusterDistanceKernelType kernelType,
NbnxnPairlistCpuWork *work)
{
switch (kernelType)
{
#if GMX_SIMD
#ifdef GMX_NBNXN_SIMD_4XN
- case Nbnxm::KernelType::Cpu4xN_Simd_4xN:
+ case ClusterDistanceKernelType::CpuSimd_4xM:
icell_set_x_simd_4xn(ci, shx, shy, shz, stride, x, work);
break;
#endif
#ifdef GMX_NBNXN_SIMD_2XNN
- case Nbnxm::KernelType::Cpu4xN_Simd_2xNN:
+ case ClusterDistanceKernelType::CpuSimd_2xMM:
icell_set_x_simd_2xnn(ci, shx, shy, shz, stride, x, work);
break;
#endif
#endif
- case Nbnxm::KernelType::Cpu4x4_PlainC:
+ case ClusterDistanceKernelType::CpuPlainC:
icell_set_x_simple(ci, shx, shy, shz, stride, x, &work->iClusterData);
break;
default:
static void icell_set_x(int ci,
real shx, real shy, real shz,
int stride, const real *x,
- Nbnxm::KernelType gmx_unused kernelType,
+ ClusterDistanceKernelType gmx_unused kernelType,
NbnxnPairlistGpuWork *work)
{
#if !GMX_SIMD4_HAVE_REAL
return false;
}
-static void makeClusterListWrapper(NbnxnPairlistCpu *nbl,
- const Grid gmx_unused &iGrid,
- const int ci,
- const Grid &jGrid,
- const int firstCell,
- const int lastCell,
- const bool excludeSubDiagonal,
- const nbnxn_atomdata_t *nbat,
- const real rlist2,
- const real rbb2,
- const Nbnxm::KernelType kernelType,
- int *numDistanceChecks)
+static void
+makeClusterListWrapper(NbnxnPairlistCpu *nbl,
+ const Grid gmx_unused &iGrid,
+ const int ci,
+ const Grid &jGrid,
+ const int firstCell,
+ const int lastCell,
+ const bool excludeSubDiagonal,
+ const nbnxn_atomdata_t *nbat,
+ const real rlist2,
+ const real rbb2,
+ const ClusterDistanceKernelType kernelType,
+ int *numDistanceChecks)
{
switch (kernelType)
{
- case Nbnxm::KernelType::Cpu4x4_PlainC:
+ case ClusterDistanceKernelType::CpuPlainC:
makeClusterListSimple(jGrid,
nbl, ci, firstCell, lastCell,
excludeSubDiagonal,
numDistanceChecks);
break;
#ifdef GMX_NBNXN_SIMD_4XN
- case Nbnxm::KernelType::Cpu4xN_Simd_4xN:
+ case ClusterDistanceKernelType::CpuSimd_4xM:
makeClusterListSimd4xn(jGrid,
nbl, ci, firstCell, lastCell,
excludeSubDiagonal,
break;
#endif
#ifdef GMX_NBNXN_SIMD_2XNN
- case Nbnxm::KernelType::Cpu4xN_Simd_2xNN:
+ case ClusterDistanceKernelType::CpuSimd_2xMM:
makeClusterListSimd2xnn(jGrid,
nbl, ci, firstCell, lastCell,
excludeSubDiagonal,
}
}
-static void makeClusterListWrapper(NbnxnPairlistGpu *nbl,
- const Grid &gmx_unused iGrid,
- const int ci,
- const Grid &jGrid,
- const int firstCell,
- const int lastCell,
- const bool excludeSubDiagonal,
- const nbnxn_atomdata_t *nbat,
- const real rlist2,
- const real rbb2,
- Nbnxm::KernelType gmx_unused kernelType,
- int *numDistanceChecks)
+static void
+makeClusterListWrapper(NbnxnPairlistGpu *nbl,
+ const Grid &gmx_unused iGrid,
+ const int ci,
+ const Grid &jGrid,
+ const int firstCell,
+ const int lastCell,
+ const bool excludeSubDiagonal,
+ const nbnxn_atomdata_t *nbat,
+ const real rlist2,
+ const real rbb2,
+ ClusterDistanceKernelType gmx_unused kernelType,
+ int *numDistanceChecks)
{
for (int cj = firstCell; cj <= lastCell; cj++)
{
const nbnxn_atomdata_t *nbat,
const t_blocka &exclusions,
real rlist,
- const Nbnxm::KernelType kernelType,
+ const PairlistType pairlistType,
int ci_block,
gmx_bool bFBufferFlag,
int nsubpair_max,
sync_work(nbl);
GMX_ASSERT(nbl->na_ci == jGrid.geometry().numAtomsICluster,
"The cluster sizes in the list and grid should match");
- nbl->na_cj = Nbnxm::JClusterSizePerKernelType[kernelType];
+ nbl->na_cj = JClusterSizePerListType[pairlistType];
na_cj_2log = get_2log(nbl->na_cj);
nbl->rlist = rlist;
const real rlist2 = nbl->rlist*nbl->rlist;
+ // Select the cluster pair distance kernel type
+ const ClusterDistanceKernelType kernelType =
+ getClusterDistanceKernelType(pairlistType, *nbat);
+
if (haveFep && !pairlistIsSimple(*nbl))
{
/* Determine an atom-pair list cut-off distance for FEP atom pairs.
gmx::ArrayRef<PairsearchWork> searchWork,
nbnxn_atomdata_t *nbat,
const t_blocka *excl,
- const Nbnxm::KernelType kernelType,
const int minimumIlistCountForGpuBalancing,
t_nrnb *nrnb,
SearchCycleCounting *searchCycleCounting)
nbnxn_make_pairlist_part(gridSet, iGrid, jGrid,
&work, nbat, *excl,
rlist,
- kernelType,
+ params_.pairlistType,
ci_block,
nbat->bUseBufferFlags,
nsubpair_target,
nbnxn_make_pairlist_part(gridSet, iGrid, jGrid,
&work, nbat, *excl,
rlist,
- kernelType,
+ params_.pairlistType,
ci_block,
nbat->bUseBufferFlags,
nsubpair_target,
PairSearch *pairSearch,
nbnxn_atomdata_t *nbat,
const t_blocka *excl,
- const Nbnxm::KernelType kernelType,
const int64_t step,
t_nrnb *nrnb)
{
pairlistSet(iLocality).constructPairlists(pairSearch->gridSet(), pairSearch->work(),
- nbat, excl, kernelType, minimumIlistCountForGpuBalancing_,
+ nbat, excl, minimumIlistCountForGpuBalancing_,
nrnb, &pairSearch->cycleCounting_);
if (iLocality == Nbnxm::InteractionLocality::Local)
t_nrnb *nrnb)
{
pairlistSets_->construct(iLocality, pairSearch_.get(), nbat.get(), excl,
- kernelSetup_.kernelType,
step, nrnb);
if (useGpu())
#ifndef GMX_NBNXM_PAIRLIST_H
#define GMX_NBNXM_PAIRLIST_H
-#include "config.h"
-
#include <cstddef>
#include "gromacs/gpu_utils/hostallocator.h"
#include "gromacs/nbnxm/constants.h"
#include "locality.h"
+#include "pairlistparams.h"
struct NbnxnPairlistCpuWork;
struct NbnxnPairlistGpuWork;
template<typename T>
using FastVector = std::vector < T, gmx::DefaultInitializationAllocator < T>>;
-/* With CPU kernels the i-cluster size is always 4 atoms. */
-static constexpr int c_nbnxnCpuIClusterSize = 4;
-
-/* With GPU kernels the i and j cluster size is 8 atoms for CUDA and can be set at compile time for OpenCL */
-#if GMX_GPU == GMX_GPU_OPENCL
-static constexpr int c_nbnxnGpuClusterSize = GMX_OPENCL_NB_CLUSTER_SIZE;
-#else
-static constexpr int c_nbnxnGpuClusterSize = 8;
-#endif
-
-/* The number of clusters in a pair-search cell, used for GPU */
-static constexpr int c_gpuNumClusterPerCellZ = 2;
-static constexpr int c_gpuNumClusterPerCellY = 2;
-static constexpr int c_gpuNumClusterPerCellX = 2;
-static constexpr int c_gpuNumClusterPerCell = c_gpuNumClusterPerCellZ*c_gpuNumClusterPerCellY*c_gpuNumClusterPerCellX;
-
-
-/* In CUDA the number of threads in a warp is 32 and we have cluster pairs
- * of 8*8=64 atoms, so it's convenient to store data for cluster pair halves.
- */
-static constexpr int c_nbnxnGpuClusterpairSplit = 2;
-
-/* The fixed size of the exclusion mask array for a half cluster pair */
-static constexpr int c_nbnxnGpuExclSize = c_nbnxnGpuClusterSize*c_nbnxnGpuClusterSize/c_nbnxnGpuClusterpairSplit;
-
-//! The available pair list types
-enum class PairlistType : int
-{
- Simple4x2,
- Simple4x4,
- Simple4x8,
- HierarchicalNxN,
- Count
-};
-
-//! Gives the i-cluster size for each pairlist type
-static constexpr gmx::EnumerationArray<PairlistType, int> IClusterSizePerListType =
-{ {
- c_nbnxnCpuIClusterSize,
- c_nbnxnCpuIClusterSize,
- c_nbnxnCpuIClusterSize,
- c_nbnxnGpuClusterSize
- } };
-//! Gives the j-cluster size for each pairlist type
-static constexpr gmx::EnumerationArray<PairlistType, int> JClusterSizePerListType =
-{ {
- 2,
- 4,
- 8,
- c_nbnxnGpuClusterSize
- } };
-
/* A buffer data structure of 64 bytes
* to be placed at the beginning and end of structs
* to avoid cache invalidation of the real contents
/*! \internal \file
*
* \brief
- * Declares the PairlistParams class
- *
- * This class holds the Nbnxm pairlist parameters.
+ * Declares the PairlistType enum and PairlistParams class
*
* \author Berk Hess <hess@kth.se>
* \ingroup module_nbnxm
#ifndef GMX_NBNXM_PAIRLISTPARAMS_H
#define GMX_NBNXM_PAIRLISTPARAMS_H
+#include "config.h"
+
+#include "gromacs/utility/enumerationhelpers.h"
#include "gromacs/utility/real.h"
#include "locality.h"
-enum class PairlistType;
-
namespace Nbnxm
{
enum class KernelType;
}
+//! The i-cluster size for CPU kernels, always 4 atoms
+static constexpr int c_nbnxnCpuIClusterSize = 4;
+
+//! The i- and j-cluster size for GPU lists, 8 atoms for CUDA, set at compile time for OpenCL
+#if GMX_GPU == GMX_GPU_OPENCL
+static constexpr int c_nbnxnGpuClusterSize = GMX_OPENCL_NB_CLUSTER_SIZE;
+#else
+static constexpr int c_nbnxnGpuClusterSize = 8;
+#endif
+
+//! The number of clusters along Z in a pair-search grid cell for GPU lists
+static constexpr int c_gpuNumClusterPerCellZ = 2;
+//! The number of clusters along Y in a pair-search grid cell for GPU lists
+static constexpr int c_gpuNumClusterPerCellY = 2;
+//! The number of clusters along X in a pair-search grid cell for GPU lists
+static constexpr int c_gpuNumClusterPerCellX = 2;
+//! The number of clusters in a pair-search grid cell for GPU lists
+static constexpr int c_gpuNumClusterPerCell = c_gpuNumClusterPerCellZ*c_gpuNumClusterPerCellY*c_gpuNumClusterPerCellX;
+
+
+/*! \brief The number of sub-parts used for data storage for a GPU cluster pair
+ *
+ * In CUDA the number of threads in a warp is 32 and we have cluster pairs
+ * of 8*8=64 atoms, so it's convenient to store data for cluster pair halves.
+ */
+static constexpr int c_nbnxnGpuClusterpairSplit = 2;
+
+//! The fixed size of the exclusion mask array for a half GPU cluster pair
+static constexpr int c_nbnxnGpuExclSize = c_nbnxnGpuClusterSize*c_nbnxnGpuClusterSize/c_nbnxnGpuClusterpairSplit;
+
+//! The available pair list types
+enum class PairlistType : int
+{
+ Simple4x2,
+ Simple4x4,
+ Simple4x8,
+ HierarchicalNxN,
+ Count
+};
+
+//! Gives the i-cluster size for each pairlist type
+static constexpr gmx::EnumerationArray<PairlistType, int> IClusterSizePerListType =
+{ {
+ c_nbnxnCpuIClusterSize,
+ c_nbnxnCpuIClusterSize,
+ c_nbnxnCpuIClusterSize,
+ c_nbnxnGpuClusterSize
+ } };
+//! Gives the j-cluster size for each pairlist type
+static constexpr gmx::EnumerationArray<PairlistType, int> JClusterSizePerListType =
+{ {
+ 2,
+ 4,
+ 8,
+ c_nbnxnGpuClusterSize
+ } };
/*! \internal
* \brief The setup for generating and pruning the nbnxn pair list.
#define GMX_NBNXM_PAIRLISTSET_H
#include "gromacs/math/vectypes.h"
-#include "gromacs/nbnxm/nbnxm.h"
#include "gromacs/nbnxm/pairlist.h"
#include "gromacs/utility/basedefinitions.h"
#include "gromacs/utility/real.h"
#include "locality.h"
+struct nbnxn_atomdata_t;
struct PairlistParams;
struct PairsearchWork;
struct SearchCycleCounting;
+struct t_blocka;
struct t_nrnb;
namespace Nbnxm
gmx::ArrayRef<PairsearchWork> searchWork,
nbnxn_atomdata_t *nbat,
const t_blocka *excl,
- Nbnxm::KernelType kernelType,
int minimumIlistCountForGpuBalancing,
t_nrnb *nrnb,
SearchCycleCounting *searchCycleCounting);
//! Dispatch the kernel for dynamic pairlist pruning
void dispatchPruneKernel(const nbnxn_atomdata_t *nbat,
- const rvec *shift_vec,
- Nbnxm::KernelType kernelType);
+ const rvec *shift_vec);
//! Returns the locality
Nbnxm::InteractionLocality locality() const
struct t_blocka;
struct t_nrnb;
-namespace Nbnxm
-{
-enum class KernelType;
-}
-
class PairlistSets
{
PairSearch *pairSearch,
nbnxn_atomdata_t *nbat,
const t_blocka *excl,
- Nbnxm::KernelType kernelbType,
int64_t step,
t_nrnb *nrnb);
//! Dispatches the dynamic pruning kernel for the given locality
void dispatchPruneKernel(Nbnxm::InteractionLocality iLocality,
const nbnxn_atomdata_t *nbat,
- const rvec *shift_vec,
- Nbnxm::KernelType kernelType);
+ const rvec *shift_vec);
//! Returns the pair list parameters
const PairlistParams ¶ms() const
#include "gromacs/nbnxm/nbnxm.h"
#include "gromacs/utility/gmxassert.h"
+#include "clusterdistancekerneltype.h"
#include "pairlistset.h"
#include "pairlistsets.h"
#include "kernels_reference/kernel_ref_prune.h"
void
PairlistSets::dispatchPruneKernel(const Nbnxm::InteractionLocality iLocality,
const nbnxn_atomdata_t *nbat,
- const rvec *shift_vec,
- const Nbnxm::KernelType kernelType)
+ const rvec *shift_vec)
{
- pairlistSet(iLocality).dispatchPruneKernel(nbat, shift_vec, kernelType);
+ pairlistSet(iLocality).dispatchPruneKernel(nbat, shift_vec);
}
void
PairlistSet::dispatchPruneKernel(const nbnxn_atomdata_t *nbat,
- const rvec *shift_vec,
- const Nbnxm::KernelType kernelType)
+ const rvec *shift_vec)
{
const real rlistInner = params_.rlistInner;
{
NbnxnPairlistCpu *nbl = &cpuLists_[i];
- switch (kernelType)
+ switch (getClusterDistanceKernelType(params_.pairlistType, *nbat))
{
- case Nbnxm::KernelType::Cpu4xN_Simd_4xN:
+ case ClusterDistanceKernelType::CpuSimd_4xM:
nbnxn_kernel_prune_4xn(nbl, nbat, shift_vec, rlistInner);
break;
- case Nbnxm::KernelType::Cpu4xN_Simd_2xNN:
+ case ClusterDistanceKernelType::CpuSimd_2xMM:
nbnxn_kernel_prune_2xnn(nbl, nbat, shift_vec, rlistInner);
break;
- case Nbnxm::KernelType::Cpu4x4_PlainC:
+ case ClusterDistanceKernelType::CpuPlainC:
nbnxn_kernel_prune_ref(nbl, nbat, shift_vec, rlistInner);
break;
default:
nonbonded_verlet_t::dispatchPruneKernelCpu(const Nbnxm::InteractionLocality iLocality,
const rvec *shift_vec)
{
- pairlistSets_->dispatchPruneKernel(iLocality, nbat.get(), shift_vec, kernelSetup_.kernelType);
+ pairlistSets_->dispatchPruneKernel(iLocality, nbat.get(), shift_vec);
}
void nonbonded_verlet_t::dispatchPruneKernelGpu(int64_t step)
biod.pnpi.spb.ru / alexxy / gromacs.git / commitdiff