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.
38 * Data management and kernel launch functions for nbnxm sycl.
40 * \ingroup module_nbnxm
44 #include "gromacs/nbnxm/gpu_common.h"
45 #include "gromacs/utility/exceptions.h"
47 #include "nbnxm_sycl_kernel.h"
48 #include "nbnxm_sycl_kernel_pruneonly.h"
49 #include "nbnxm_sycl_types.h"
55 * Launch asynchronously the download of nonbonded forces from the GPU
56 * (and energies/shift forces if required).
58 void gpu_launch_cpyback(NbnxmGpu* nb,
59 struct nbnxn_atomdata_t* nbatom,
60 const gmx::StepWorkload& stepWork,
61 const AtomLocality atomLocality)
63 GMX_ASSERT(nb, "Need a valid nbnxn_gpu object");
65 const InteractionLocality iloc = gpuAtomToInteractionLocality(atomLocality);
66 GMX_ASSERT(iloc == InteractionLocality::Local
67 || (iloc == InteractionLocality::NonLocal && nb->bNonLocalStreamDoneMarked == false),
68 "Non-local stream is indicating that the copy back event is enqueued at the "
69 "beginning of the copy back function.");
71 const DeviceStream& deviceStream = *nb->deviceStreams[iloc];
72 NBAtomData* adat = nb->atdat;
74 /* don't launch non-local copy-back if there was no non-local work to do */
75 if ((iloc == InteractionLocality::NonLocal) && !haveGpuShortRangeWork(*nb, iloc))
77 nb->bNonLocalStreamDoneMarked = false;
81 /* local/nonlocal offset and length used for xq and f */
82 auto atomsRange = getGpuAtomRange(adat, atomLocality);
84 // With DD the local D2H transfer can only start after the non-local kernel has finished.
85 if (iloc == InteractionLocality::Local && nb->bNonLocalStreamDoneMarked)
87 nb->nonlocal_done.waitForEvent();
88 nb->bNonLocalStreamDoneMarked = false;
92 * Skip if buffer ops / reduction is offloaded to the GPU.
94 if (!stepWork.useGpuFBufferOps)
96 GMX_ASSERT(adat->f.elementSize() == sizeof(Float3),
97 "The size of the force buffer element should be equal to the size of float3.");
98 copyFromDeviceBuffer(reinterpret_cast<Float3*>(nbatom->out[0].f.data()) + atomsRange.begin(),
103 GpuApiCallBehavior::Async,
107 /* After the non-local D2H is launched the nonlocal_done event can be
108 recorded which signals that the local D2H can proceed. This event is not
109 placed after the non-local kernel because we want the non-local data
111 if (iloc == InteractionLocality::NonLocal)
113 nb->nonlocal_done.markEvent(deviceStream);
114 nb->bNonLocalStreamDoneMarked = true;
117 /* only transfer energies in the local stream */
118 if (iloc == InteractionLocality::Local)
120 /* DtoH fshift when virial is needed */
121 if (stepWork.computeVirial)
123 GMX_ASSERT(sizeof(*nb->nbst.fShift) == adat->fShift.elementSize(),
124 "Sizes of host- and device-side shift vector elements should be the same.");
125 copyFromDeviceBuffer(
126 nb->nbst.fShift, &adat->fShift, 0, SHIFTS, deviceStream, GpuApiCallBehavior::Async, nullptr);
130 if (stepWork.computeEnergy)
132 GMX_ASSERT(sizeof(*nb->nbst.eLJ) == sizeof(float),
133 "Sizes of host- and device-side LJ energy terms should be the same.");
134 copyFromDeviceBuffer(
135 nb->nbst.eLJ, &adat->eLJ, 0, 1, deviceStream, GpuApiCallBehavior::Async, nullptr);
136 GMX_ASSERT(sizeof(*nb->nbst.eElec) == sizeof(float),
137 "Sizes of host- and device-side electrostatic energy terms should be the "
139 copyFromDeviceBuffer(
140 nb->nbst.eElec, &adat->eElec, 0, 1, deviceStream, GpuApiCallBehavior::Async, nullptr);
145 void gpu_launch_kernel_pruneonly(NbnxmGpu* nb, const InteractionLocality iloc, const int numParts)
147 gpu_plist* plist = nb->plist[iloc];
149 if (plist->haveFreshList)
151 GMX_ASSERT(numParts == 1, "With first pruning we expect 1 part");
153 /* Set rollingPruningNumParts to signal that it is not set */
154 plist->rollingPruningNumParts = 0;
155 plist->rollingPruningPart = 0;
159 if (plist->rollingPruningNumParts == 0)
161 plist->rollingPruningNumParts = numParts;
165 GMX_ASSERT(numParts == plist->rollingPruningNumParts,
166 "It is not allowed to change numParts in between list generation steps");
170 /* Use a local variable for part and update in plist, so we can return here
171 * without duplicating the part increment code.
173 const int part = plist->rollingPruningPart;
175 plist->rollingPruningPart++;
176 if (plist->rollingPruningPart >= plist->rollingPruningNumParts)
178 plist->rollingPruningPart = 0;
181 /* Compute the number of list entries to prune in this pass */
182 const int numSciInPart = (plist->nsci - part) / numParts;
184 /* Don't launch the kernel if there is no work to do */
185 if (numSciInPart <= 0)
187 plist->haveFreshList = false;
191 launchNbnxmKernelPruneOnly(nb, iloc, numParts, part, numSciInPart);
193 if (plist->haveFreshList)
195 plist->haveFreshList = false;
196 nb->didPrune[iloc] = true; // Mark that pruning has been done
200 nb->didRollingPrune[iloc] = true; // Mark that rolling pruning has been done
204 void gpu_launch_kernel(NbnxmGpu* nb, const gmx::StepWorkload& stepWork, const Nbnxm::InteractionLocality iloc)
206 const NBParamGpu* nbp = nb->nbparam;
207 gpu_plist* plist = nb->plist[iloc];
209 if (canSkipNonbondedWork(*nb, iloc))
211 plist->haveFreshList = false;
215 if (nbp->useDynamicPruning && plist->haveFreshList)
217 // Prunes for rlistOuter and rlistInner, sets plist->haveFreshList=false
218 gpu_launch_kernel_pruneonly(nb, iloc, 1);
221 if (plist->nsci == 0)
223 /* Don't launch an empty local kernel */
227 launchNbnxmKernel(nb, stepWork, iloc);