src/gromacs/mdtypes/state_propagator_data_gpu_impl_gpu.cpp

   1 /*
   2  * This file is part of the GROMACS molecular simulation package.
   3  *
   4  * Copyright (c) 2019, 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 /*! \internal \file
  36  *
  37  * \brief Definitions of interfaces for GPU state data propagator object.
  38  *
  39  * \author Artem Zhmurov <zhmurov@gmail.com>
  40  *
  41  * \ingroup module_mdtypes
  42  */
  43 #include "gmxpre.h"
  44
  45 #include "config.h"
  46
  47 #if GMX_GPU != GMX_GPU_NONE
  48
  49 #    if GMX_GPU == GMX_GPU_CUDA
  50 #        include "gromacs/gpu_utils/cudautils.cuh"
  51 #    endif
  52 #    include "gromacs/gpu_utils/devicebuffer.h"
  53 #    if GMX_GPU == GMX_GPU_OPENCL
  54 #        include "gromacs/gpu_utils/oclutils.h"
  55 #    endif
  56 #    include "gromacs/math/vectypes.h"
  57 #    include "gromacs/mdtypes/state_propagator_data_gpu.h"
  58 #    include "gromacs/utility/classhelpers.h"
  59
  60 #    include "state_propagator_data_gpu_impl.h"
  61
  62 namespace gmx
  63 {
  64
  65 StatePropagatorDataGpu::Impl::Impl(const void*        pmeStream,
  66                                    const void*        localStream,
  67                                    const void*        nonLocalStream,
  68                                    const void*        deviceContext,
  69                                    GpuApiCallBehavior transferKind,
  70                                    int                paddingSize) :
  71     transferKind_(transferKind),
  72     paddingSize_(paddingSize)
  73 {
  74     static_assert(GMX_GPU != GMX_GPU_NONE,
  75                   "This object should only be constructed on the GPU code-paths.");
  76
  77     // TODO: Refactor when the StreamManager is introduced.
  78     if (GMX_GPU == GMX_GPU_OPENCL)
  79     {
  80         GMX_ASSERT(deviceContext != nullptr, "GPU context should be set in OpenCL builds.");
  81         GMX_ASSERT(pmeStream != nullptr, "GPU PME stream should be set in OpenCL builds.");
  82
  83         // The update stream is set to the PME stream in OpenCL, since PME stream is the only stream created in the PME context.
  84         pmeStream_     = *static_cast<const CommandStream*>(pmeStream);
  85         updateStream_  = *static_cast<const CommandStream*>(pmeStream);
  86         deviceContext_ = *static_cast<const DeviceContext*>(deviceContext);
  87         GMX_UNUSED_VALUE(localStream);
  88         GMX_UNUSED_VALUE(nonLocalStream);
  89     }
  90
  91     if (GMX_GPU == GMX_GPU_CUDA)
  92     {
  93         if (pmeStream != nullptr)
  94         {
  95             pmeStream_ = *static_cast<const CommandStream*>(pmeStream);
  96         }
  97         if (localStream != nullptr)
  98         {
  99             localStream_ = *static_cast<const CommandStream*>(localStream);
 100         }
 101         if (nonLocalStream != nullptr)
 102         {
 103             nonLocalStream_ = *static_cast<const CommandStream*>(nonLocalStream);
 104         }
 105
 106         // TODO: The update stream should be created only when it is needed.
 107 #    if (GMX_GPU == GMX_GPU_CUDA)
 108         cudaError_t stat;
 109         stat = cudaStreamCreate(&updateStream_);
 110         CU_RET_ERR(stat, "CUDA stream creation failed in StatePropagatorDataGpu");
 111 #    endif
 112         GMX_UNUSED_VALUE(deviceContext);
 113     }
 114
 115     // Map the atom locality to the stream that will be used for coordinates,
 116     // velocities and forces transfers. Same streams are used for H2D and D2H copies.
 117     // Note, that nullptr stream is used here to indicate that the copy is not supported.
 118     xCopyStreams_[AtomLocality::Local]    = updateStream_;
 119     xCopyStreams_[AtomLocality::NonLocal] = nonLocalStream_;
 120     xCopyStreams_[AtomLocality::All]      = updateStream_;
 121
 122     vCopyStreams_[AtomLocality::Local]    = updateStream_;
 123     vCopyStreams_[AtomLocality::NonLocal] = nullptr;
 124     vCopyStreams_[AtomLocality::All]      = updateStream_;
 125
 126     fCopyStreams_[AtomLocality::Local]    = localStream_;
 127     fCopyStreams_[AtomLocality::NonLocal] = nonLocalStream_;
 128     fCopyStreams_[AtomLocality::All]      = updateStream_;
 129 }
 130
 131 StatePropagatorDataGpu::Impl::Impl(const void*        pmeStream,
 132                                    const void*        deviceContext,
 133                                    GpuApiCallBehavior transferKind,
 134                                    int                paddingSize) :
 135     transferKind_(transferKind),
 136     paddingSize_(paddingSize)
 137 {
 138     static_assert(GMX_GPU != GMX_GPU_NONE,
 139                   "This object should only be constructed on the GPU code-paths.");
 140
 141     if (GMX_GPU == GMX_GPU_OPENCL)
 142     {
 143         GMX_ASSERT(deviceContext != nullptr, "GPU context should be set in OpenCL builds.");
 144         deviceContext_ = *static_cast<const DeviceContext*>(deviceContext);
 145     }
 146
 147     GMX_ASSERT(pmeStream != nullptr, "GPU PME stream should be set.");
 148     pmeStream_ = *static_cast<const CommandStream*>(pmeStream);
 149
 150     localStream_    = nullptr;
 151     nonLocalStream_ = nullptr;
 152     updateStream_   = nullptr;
 153
 154
 155     // Only local/all coordinates are allowed to be copied in PME-only rank/ PME tests.
 156     // This it temporary measure to make it safe to use this class in those cases.
 157     xCopyStreams_[AtomLocality::Local]    = pmeStream_;
 158     xCopyStreams_[AtomLocality::NonLocal] = nullptr;
 159     xCopyStreams_[AtomLocality::All]      = pmeStream_;
 160
 161     vCopyStreams_[AtomLocality::Local]    = nullptr;
 162     vCopyStreams_[AtomLocality::NonLocal] = nullptr;
 163     vCopyStreams_[AtomLocality::All]      = nullptr;
 164
 165     fCopyStreams_[AtomLocality::Local]    = nullptr;
 166     fCopyStreams_[AtomLocality::NonLocal] = nullptr;
 167     fCopyStreams_[AtomLocality::All]      = nullptr;
 168 }
 169
 170 StatePropagatorDataGpu::Impl::~Impl() {}
 171
 172 void StatePropagatorDataGpu::Impl::reinit(int numAtomsLocal, int numAtomsAll)
 173 {
 174     numAtomsLocal_ = numAtomsLocal;
 175     numAtomsAll_   = numAtomsAll;
 176
 177     int numAtomsPadded;
 178     if (paddingSize_ > 0)
 179     {
 180         numAtomsPadded = ((numAtomsAll_ + paddingSize_ - 1) / paddingSize_) * paddingSize_;
 181     }
 182     else
 183     {
 184         numAtomsPadded = numAtomsAll_;
 185     }
 186
 187     reallocateDeviceBuffer(&d_x_, DIM * numAtomsPadded, &d_xSize_, &d_xCapacity_, deviceContext_);
 188
 189     const size_t paddingAllocationSize = numAtomsPadded - numAtomsAll_;
 190     if (paddingAllocationSize > 0)
 191     {
 192         // The PME stream is used here because the padding region of d_x_ is only in the PME task.
 193         clearDeviceBufferAsync(&d_x_, DIM * numAtomsAll_, DIM * paddingAllocationSize, pmeStream_);
 194     }
 195
 196     reallocateDeviceBuffer(&d_v_, DIM * numAtomsAll_, &d_vSize_, &d_vCapacity_, deviceContext_);
 197     reallocateDeviceBuffer(&d_f_, DIM * numAtomsAll_, &d_fSize_, &d_fCapacity_, deviceContext_);
 198 }
 199
 200 std::tuple<int, int> StatePropagatorDataGpu::Impl::getAtomRangesFromAtomLocality(AtomLocality atomLocality)
 201 {
 202     int atomsStartAt   = 0;
 203     int numAtomsToCopy = 0;
 204     switch (atomLocality)
 205     {
 206         case AtomLocality::All:
 207             atomsStartAt   = 0;
 208             numAtomsToCopy = numAtomsAll_;
 209             break;
 210         case AtomLocality::Local:
 211             atomsStartAt   = 0;
 212             numAtomsToCopy = numAtomsLocal_;
 213             break;
 214         case AtomLocality::NonLocal:
 215             atomsStartAt   = numAtomsLocal_;
 216             numAtomsToCopy = numAtomsAll_ - numAtomsLocal_;
 217             break;
 218         default:
 219             GMX_RELEASE_ASSERT(false,
 220                                "Wrong range of atoms requested in GPU state data manager. Should "
 221                                "be All, Local or NonLocal.");
 222     }
 223     GMX_ASSERT(atomsStartAt >= 0,
 224                "The first elemtnt to copy has negative index. Probably, the GPU propagator state "
 225                "was not initialized.");
 226     GMX_ASSERT(numAtomsToCopy >= 0,
 227                "Number of atoms to copy is negative. Probably, the GPU propagator state was not "
 228                "initialized.");
 229     return std::make_tuple(atomsStartAt, numAtomsToCopy);
 230 }
 231
 232 void StatePropagatorDataGpu::Impl::copyToDevice(DeviceBuffer<float>                  d_data,
 233                                                 const gmx::ArrayRef<const gmx::RVec> h_data,
 234                                                 int                                  dataSize,
 235                                                 AtomLocality                         atomLocality,
 236                                                 CommandStream                        commandStream)
 237 {
 238
 239     GMX_UNUSED_VALUE(dataSize);
 240
 241     GMX_ASSERT(dataSize >= 0, "Trying to copy to device buffer before it was allocated.");
 242
 243     int atomsStartAt, numAtomsToCopy;
 244     std::tie(atomsStartAt, numAtomsToCopy) = getAtomRangesFromAtomLocality(atomLocality);
 245
 246     int elementsStartAt   = atomsStartAt * DIM;
 247     int numElementsToCopy = numAtomsToCopy * DIM;
 248
 249     if (numAtomsToCopy != 0)
 250     {
 251         GMX_ASSERT(elementsStartAt + numElementsToCopy <= dataSize,
 252                    "The device allocation is smaller than requested copy range.");
 253         GMX_ASSERT(atomsStartAt + numAtomsToCopy <= h_data.ssize(),
 254                    "The host buffer is smaller than the requested copy range.");
 255
 256         copyToDeviceBuffer(&d_data, reinterpret_cast<const float*>(&h_data.data()[atomsStartAt]),
 257                            elementsStartAt, numElementsToCopy, commandStream, transferKind_, nullptr);
 258     }
 259 }
 260
 261 void StatePropagatorDataGpu::Impl::copyFromDevice(gmx::ArrayRef<gmx::RVec> h_data,
 262                                                   DeviceBuffer<float>      d_data,
 263                                                   int                      dataSize,
 264                                                   AtomLocality             atomLocality,
 265                                                   CommandStream            commandStream)
 266 {
 267
 268     GMX_UNUSED_VALUE(dataSize);
 269
 270     GMX_ASSERT(dataSize >= 0, "Trying to copy from device buffer before it was allocated.");
 271
 272     int atomsStartAt, numAtomsToCopy;
 273     std::tie(atomsStartAt, numAtomsToCopy) = getAtomRangesFromAtomLocality(atomLocality);
 274
 275     int elementsStartAt   = atomsStartAt * DIM;
 276     int numElementsToCopy = numAtomsToCopy * DIM;
 277
 278     if (numAtomsToCopy != 0)
 279     {
 280         GMX_ASSERT(elementsStartAt + numElementsToCopy <= dataSize,
 281                    "The device allocation is smaller than requested copy range.");
 282         GMX_ASSERT(atomsStartAt + numAtomsToCopy <= h_data.ssize(),
 283                    "The host buffer is smaller than the requested copy range.");
 284
 285         copyFromDeviceBuffer(reinterpret_cast<float*>(&h_data.data()[atomsStartAt]), &d_data,
 286                              elementsStartAt, numElementsToCopy, commandStream, transferKind_, nullptr);
 287     }
 288 }
 289
 290 DeviceBuffer<float> StatePropagatorDataGpu::Impl::getCoordinates()
 291 {
 292     return d_x_;
 293 }
 294
 295 void StatePropagatorDataGpu::Impl::copyCoordinatesToGpu(const gmx::ArrayRef<const gmx::RVec> h_x,
 296                                                         AtomLocality atomLocality)
 297 {
 298     GMX_ASSERT(atomLocality < AtomLocality::Count, "Wrong atom locality.");
 299     CommandStream commandStream = xCopyStreams_[atomLocality];
 300     GMX_ASSERT(commandStream != nullptr,
 301                "No stream is valid for copying positions with given atom locality.");
 302
 303     copyToDevice(d_x_, h_x, d_xSize_, atomLocality, commandStream);
 304
 305     // markEvent is skipped in OpenCL as:
 306     //   - it's not needed, copy is done in the same stream as the only consumer task (PME)
 307     //   - we don't consume the events in OpenCL which is not allowed by GpuEventSynchronizer (would leak memory).
 308     // TODO: remove this by adding an event-mark free flavor of this function
 309     if (GMX_GPU == GMX_GPU_CUDA)
 310     {
 311         xReadyOnDevice_[atomLocality].markEvent(commandStream);
 312     }
 313 }
 314
 315 GpuEventSynchronizer*
 316 StatePropagatorDataGpu::Impl::getCoordinatesReadyOnDeviceEvent(AtomLocality atomLocality,
 317                                                                const SimulationWorkload& simulationWork,
 318                                                                const StepWorkload&       stepWork)
 319 {
 320     // The provider of the coordinates may be different for local atoms. If the update is offloaded
 321     // and this is not a neighbor search step, then the consumer needs to wait for the update
 322     // to complete. Otherwise, the coordinates are copied from the host and we need to wait for
 323     // the copy event. Non-local coordinates are always provided by the H2D copy.
 324     //
 325     // TODO: This should be reconsidered to support the halo exchange.
 326     //
 327     // In OpenCL no events are used as coordinate sync is not necessary
 328     if (GMX_GPU == GMX_GPU_OPENCL)
 329     {
 330         return nullptr;
 331     }
 332     if (atomLocality == AtomLocality::Local && simulationWork.useGpuUpdate && !stepWork.doNeighborSearch)
 333     {
 334         return &xUpdatedOnDevice_;
 335     }
 336     else
 337     {
 338         return &xReadyOnDevice_[atomLocality];
 339     }
 340 }
 341
 342 void StatePropagatorDataGpu::Impl::waitCoordinatesCopiedToDevice(AtomLocality atomLocality)
 343 {
 344     GMX_ASSERT(atomLocality < AtomLocality::Count, "Wrong atom locality.");
 345     xReadyOnDevice_[atomLocality].waitForEvent();
 346 }
 347
 348 GpuEventSynchronizer* StatePropagatorDataGpu::Impl::xUpdatedOnDevice()
 349 {
 350     return &xUpdatedOnDevice_;
 351 }
 352
 353 void StatePropagatorDataGpu::Impl::copyCoordinatesFromGpu(gmx::ArrayRef<gmx::RVec> h_x, AtomLocality atomLocality)
 354 {
 355     GMX_ASSERT(atomLocality < AtomLocality::Count, "Wrong atom locality.");
 356     CommandStream commandStream = xCopyStreams_[atomLocality];
 357     GMX_ASSERT(commandStream != nullptr,
 358                "No stream is valid for copying positions with given atom locality.");
 359
 360     copyFromDevice(h_x, d_x_, d_xSize_, atomLocality, commandStream);
 361     // Note: unlike copyCoordinatesToGpu this is not used in OpenCL, and the conditional is not needed.
 362     xReadyOnHost_[atomLocality].markEvent(commandStream);
 363 }
 364
 365 void StatePropagatorDataGpu::Impl::waitCoordinatesReadyOnHost(AtomLocality atomLocality)
 366 {
 367     xReadyOnHost_[atomLocality].waitForEvent();
 368 }
 369
 370
 371 DeviceBuffer<float> StatePropagatorDataGpu::Impl::getVelocities()
 372 {
 373     return d_v_;
 374 }
 375
 376 void StatePropagatorDataGpu::Impl::copyVelocitiesToGpu(const gmx::ArrayRef<const gmx::RVec> h_v,
 377                                                        AtomLocality atomLocality)
 378 {
 379     GMX_ASSERT(atomLocality < AtomLocality::Count, "Wrong atom locality.");
 380     CommandStream commandStream = vCopyStreams_[atomLocality];
 381     GMX_ASSERT(commandStream != nullptr,
 382                "No stream is valid for copying velocities with given atom locality.");
 383
 384     copyToDevice(d_v_, h_v, d_vSize_, atomLocality, commandStream);
 385     vReadyOnDevice_[atomLocality].markEvent(commandStream);
 386 }
 387
 388 GpuEventSynchronizer* StatePropagatorDataGpu::Impl::getVelocitiesReadyOnDeviceEvent(AtomLocality atomLocality)
 389 {
 390     return &vReadyOnDevice_[atomLocality];
 391 }
 392
 393
 394 void StatePropagatorDataGpu::Impl::copyVelocitiesFromGpu(gmx::ArrayRef<gmx::RVec> h_v, AtomLocality atomLocality)
 395 {
 396     GMX_ASSERT(atomLocality < AtomLocality::Count, "Wrong atom locality.");
 397     CommandStream commandStream = vCopyStreams_[atomLocality];
 398     GMX_ASSERT(commandStream != nullptr,
 399                "No stream is valid for copying velocities with given atom locality.");
 400
 401     copyFromDevice(h_v, d_v_, d_vSize_, atomLocality, commandStream);
 402     vReadyOnHost_[atomLocality].markEvent(commandStream);
 403 }
 404
 405 void StatePropagatorDataGpu::Impl::waitVelocitiesReadyOnHost(AtomLocality atomLocality)
 406 {
 407     vReadyOnHost_[atomLocality].waitForEvent();
 408 }
 409
 410
 411 DeviceBuffer<float> StatePropagatorDataGpu::Impl::getForces()
 412 {
 413     return d_f_;
 414 }
 415
 416 void StatePropagatorDataGpu::Impl::copyForcesToGpu(const gmx::ArrayRef<const gmx::RVec> h_f,
 417                                                    AtomLocality atomLocality)
 418 {
 419     GMX_ASSERT(atomLocality < AtomLocality::Count, "Wrong atom locality.");
 420     CommandStream commandStream = fCopyStreams_[atomLocality];
 421     GMX_ASSERT(commandStream != nullptr,
 422                "No stream is valid for copying forces with given atom locality.");
 423
 424     copyToDevice(d_f_, h_f, d_fSize_, atomLocality, commandStream);
 425     fReadyOnDevice_[atomLocality].markEvent(commandStream);
 426 }
 427
 428 GpuEventSynchronizer* StatePropagatorDataGpu::Impl::getForcesReadyOnDeviceEvent(AtomLocality atomLocality,
 429                                                                                 bool useGpuFBufferOps)
 430 {
 431     if ((atomLocality == AtomLocality::Local || atomLocality == AtomLocality::NonLocal) && useGpuFBufferOps)
 432     {
 433         return &fReducedOnDevice_;
 434     }
 435     else
 436     {
 437         return &fReadyOnDevice_[atomLocality];
 438     }
 439 }
 440
 441 GpuEventSynchronizer* StatePropagatorDataGpu::Impl::fReducedOnDevice()
 442 {
 443     return &fReducedOnDevice_;
 444 }
 445
 446 void StatePropagatorDataGpu::Impl::copyForcesFromGpu(gmx::ArrayRef<gmx::RVec> h_f, AtomLocality atomLocality)
 447 {
 448     GMX_ASSERT(atomLocality < AtomLocality::Count, "Wrong atom locality.");
 449     CommandStream commandStream = fCopyStreams_[atomLocality];
 450     GMX_ASSERT(commandStream != nullptr,
 451                "No stream is valid for copying forces with given atom locality.");
 452
 453     copyFromDevice(h_f, d_f_, d_fSize_, atomLocality, commandStream);
 454     fReadyOnHost_[atomLocality].markEvent(commandStream);
 455 }
 456
 457 void StatePropagatorDataGpu::Impl::waitForcesReadyOnHost(AtomLocality atomLocality)
 458 {
 459     fReadyOnHost_[atomLocality].waitForEvent();
 460 }
 461
 462 void* StatePropagatorDataGpu::Impl::getUpdateStream()
 463 {
 464     return &updateStream_;
 465 }
 466
 467 int StatePropagatorDataGpu::Impl::numAtomsLocal()
 468 {
 469     return numAtomsLocal_;
 470 }
 471
 472 int StatePropagatorDataGpu::Impl::numAtomsAll()
 473 {
 474     return numAtomsAll_;
 475 }
 476
 477
 478 StatePropagatorDataGpu::StatePropagatorDataGpu(const void*        pmeStream,
 479                                                const void*        localStream,
 480                                                const void*        nonLocalStream,
 481                                                const void*        deviceContext,
 482                                                GpuApiCallBehavior transferKind,
 483                                                int                paddingSize) :
 484     impl_(new Impl(pmeStream, localStream, nonLocalStream, deviceContext, transferKind, paddingSize))
 485 {
 486 }
 487
 488 StatePropagatorDataGpu::StatePropagatorDataGpu(const void*        pmeStream,
 489                                                const void*        deviceContext,
 490                                                GpuApiCallBehavior transferKind,
 491                                                int                paddingSize) :
 492     impl_(new Impl(pmeStream, deviceContext, transferKind, paddingSize))
 493 {
 494 }
 495
 496 StatePropagatorDataGpu::StatePropagatorDataGpu(StatePropagatorDataGpu&& /* other */) noexcept = default;
 497
 498 StatePropagatorDataGpu& StatePropagatorDataGpu::operator=(StatePropagatorDataGpu&& /* other */) noexcept = default;
 499
 500 StatePropagatorDataGpu::~StatePropagatorDataGpu() = default;
 501
 502
 503 void StatePropagatorDataGpu::reinit(int numAtomsLocal, int numAtomsAll)
 504 {
 505     return impl_->reinit(numAtomsLocal, numAtomsAll);
 506 }
 507
 508 std::tuple<int, int> StatePropagatorDataGpu::getAtomRangesFromAtomLocality(AtomLocality atomLocality)
 509 {
 510     return impl_->getAtomRangesFromAtomLocality(atomLocality);
 511 }
 512
 513
 514 DeviceBuffer<float> StatePropagatorDataGpu::getCoordinates()
 515 {
 516     return impl_->getCoordinates();
 517 }
 518
 519 void StatePropagatorDataGpu::copyCoordinatesToGpu(const gmx::ArrayRef<const gmx::RVec> h_x,
 520                                                   AtomLocality                         atomLocality)
 521 {
 522     return impl_->copyCoordinatesToGpu(h_x, atomLocality);
 523 }
 524
 525 GpuEventSynchronizer*
 526 StatePropagatorDataGpu::getCoordinatesReadyOnDeviceEvent(AtomLocality              atomLocality,
 527                                                          const SimulationWorkload& simulationWork,
 528                                                          const StepWorkload&       stepWork)
 529 {
 530     return impl_->getCoordinatesReadyOnDeviceEvent(atomLocality, simulationWork, stepWork);
 531 }
 532
 533 void StatePropagatorDataGpu::waitCoordinatesCopiedToDevice(AtomLocality atomLocality)
 534 {
 535     return impl_->waitCoordinatesCopiedToDevice(atomLocality);
 536 }
 537
 538 GpuEventSynchronizer* StatePropagatorDataGpu::xUpdatedOnDevice()
 539 {
 540     return impl_->xUpdatedOnDevice();
 541 }
 542
 543 void StatePropagatorDataGpu::copyCoordinatesFromGpu(gmx::ArrayRef<RVec> h_x, AtomLocality atomLocality)
 544 {
 545     return impl_->copyCoordinatesFromGpu(h_x, atomLocality);
 546 }
 547
 548 void StatePropagatorDataGpu::waitCoordinatesReadyOnHost(AtomLocality atomLocality)
 549 {
 550     return impl_->waitCoordinatesReadyOnHost(atomLocality);
 551 }
 552
 553
 554 DeviceBuffer<float> StatePropagatorDataGpu::getVelocities()
 555 {
 556     return impl_->getVelocities();
 557 }
 558
 559 void StatePropagatorDataGpu::copyVelocitiesToGpu(const gmx::ArrayRef<const gmx::RVec> h_v,
 560                                                  AtomLocality                         atomLocality)
 561 {
 562     return impl_->copyVelocitiesToGpu(h_v, atomLocality);
 563 }
 564
 565 GpuEventSynchronizer* StatePropagatorDataGpu::getVelocitiesReadyOnDeviceEvent(AtomLocality atomLocality)
 566 {
 567     return impl_->getVelocitiesReadyOnDeviceEvent(atomLocality);
 568 }
 569
 570 void StatePropagatorDataGpu::copyVelocitiesFromGpu(gmx::ArrayRef<RVec> h_v, AtomLocality atomLocality)
 571 {
 572     return impl_->copyVelocitiesFromGpu(h_v, atomLocality);
 573 }
 574
 575 void StatePropagatorDataGpu::waitVelocitiesReadyOnHost(AtomLocality atomLocality)
 576 {
 577     return impl_->waitVelocitiesReadyOnHost(atomLocality);
 578 }
 579
 580
 581 DeviceBuffer<float> StatePropagatorDataGpu::getForces()
 582 {
 583     return impl_->getForces();
 584 }
 585
 586 void StatePropagatorDataGpu::copyForcesToGpu(const gmx::ArrayRef<const gmx::RVec> h_f, AtomLocality atomLocality)
 587 {
 588     return impl_->copyForcesToGpu(h_f, atomLocality);
 589 }
 590
 591 GpuEventSynchronizer* StatePropagatorDataGpu::getForcesReadyOnDeviceEvent(AtomLocality atomLocality,
 592                                                                           bool useGpuFBufferOps)
 593 {
 594     return impl_->getForcesReadyOnDeviceEvent(atomLocality, useGpuFBufferOps);
 595 }
 596
 597 GpuEventSynchronizer* StatePropagatorDataGpu::fReducedOnDevice()
 598 {
 599     return impl_->fReducedOnDevice();
 600 }
 601
 602 void StatePropagatorDataGpu::copyForcesFromGpu(gmx::ArrayRef<RVec> h_f, AtomLocality atomLocality)
 603 {
 604     return impl_->copyForcesFromGpu(h_f, atomLocality);
 605 }
 606
 607 void StatePropagatorDataGpu::waitForcesReadyOnHost(AtomLocality atomLocality)
 608 {
 609     return impl_->waitForcesReadyOnHost(atomLocality);
 610 }
 611
 612
 613 void* StatePropagatorDataGpu::getUpdateStream()
 614 {
 615     return impl_->getUpdateStream();
 616 }
 617
 618 int StatePropagatorDataGpu::numAtomsLocal()
 619 {
 620     return impl_->numAtomsLocal();
 621 }
 622
 623 int StatePropagatorDataGpu::numAtomsAll()
 624 {
 625     return impl_->numAtomsAll();
 626 }
 627
 628 } // namespace gmx
 629
 630 #endif // GMX_GPU == GMX_GPU_NONE