2 * This file is part of the GROMACS molecular simulation package.
4 * Copyright (c) 2016,2017,2018, 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 * Implements common routines for PME tests.
39 * \author Aleksei Iupinov <a.yupinov@gmail.com>
40 * \ingroup module_ewald
44 #include "pmetestcommon.h"
48 #include "gromacs/domdec/domdec.h"
49 #include "gromacs/ewald/pme-gather.h"
50 #include "gromacs/ewald/pme-gpu-internal.h"
51 #include "gromacs/ewald/pme-grid.h"
52 #include "gromacs/ewald/pme-internal.h"
53 #include "gromacs/ewald/pme-solve.h"
54 #include "gromacs/ewald/pme-spread.h"
55 #include "gromacs/fft/parallel_3dfft.h"
56 #include "gromacs/gpu_utils/gpu_utils.h"
57 #include "gromacs/math/invertmatrix.h"
58 #include "gromacs/mdtypes/commrec.h"
59 #include "gromacs/pbcutil/pbc.h"
60 #include "gromacs/topology/topology.h"
61 #include "gromacs/utility/exceptions.h"
62 #include "gromacs/utility/gmxassert.h"
63 #include "gromacs/utility/logger.h"
64 #include "gromacs/utility/stringutil.h"
66 #include "testutils/testasserts.h"
73 bool pmeSupportsInputForMode(const gmx_hw_info_t &hwinfo,
74 const t_inputrec *inputRec,
86 implemented = (pme_gpu_supports_build(hwinfo, nullptr) &&
87 pme_gpu_supports_input(*inputRec, mtop, nullptr));
91 GMX_THROW(InternalError("Test not implemented for this mode"));
96 uint64_t getSplineModuliDoublePrecisionUlps(int splineOrder)
98 /* Arbitrary ulp tolerance for sine/cosine implementation. It's
99 * hard to know what to pick without testing lots of
100 * implementations. */
101 const uint64_t sineUlps = 10;
102 return 4 * (splineOrder - 2) + 2 * sineUlps * splineOrder;
105 //! PME initialization - internal
106 static PmeSafePointer pmeInitInternal(const t_inputrec *inputRec,
108 const gmx_device_info_t *gpuInfo,
109 PmeGpuProgramHandle pmeGpuProgram,
111 const Matrix3x3 &box,
112 real ewaldCoeff_q = 1.0f,
113 real ewaldCoeff_lj = 1.0f
116 const MDLogger dummyLogger;
117 const auto runMode = (mode == CodePath::CPU) ? PmeRunMode::CPU : PmeRunMode::Mixed;
118 t_commrec dummyCommrec = {0};
119 NumPmeDomains numPmeDomains = { 1, 1 };
120 gmx_pme_t *pmeDataRaw = gmx_pme_init(&dummyCommrec, numPmeDomains, inputRec, atomCount, false, false, true,
121 ewaldCoeff_q, ewaldCoeff_lj, 1, runMode, nullptr, gpuInfo, pmeGpuProgram, dummyLogger);
122 PmeSafePointer pme(pmeDataRaw); // taking ownership
124 // TODO get rid of this with proper matrix type
126 for (int i = 0; i < DIM; i++)
128 for (int j = 0; j < DIM; j++)
130 boxTemp[i][j] = box[i * DIM + j];
133 const char *boxError = check_box(-1, boxTemp);
134 GMX_RELEASE_ASSERT(boxError == nullptr, boxError);
139 invertBoxMatrix(boxTemp, pme->recipbox);
143 pme_gpu_set_testing(pme->gpu, true);
144 pme_gpu_update_input_box(pme->gpu, boxTemp);
148 GMX_THROW(InternalError("Test not implemented for this mode"));
154 //! Simple PME initialization based on input, no atom data
155 PmeSafePointer pmeInitEmpty(const t_inputrec *inputRec,
157 const gmx_device_info_t *gpuInfo,
158 PmeGpuProgramHandle pmeGpuProgram,
159 const Matrix3x3 &box,
164 return pmeInitInternal(inputRec, mode, gpuInfo, pmeGpuProgram, 0, box, ewaldCoeff_q, ewaldCoeff_lj);
165 // hiding the fact that PME actually needs to know the number of atoms in advance
168 //! PME initialization with atom data
169 PmeSafePointer pmeInitAtoms(const t_inputrec *inputRec,
171 const gmx_device_info_t *gpuInfo,
172 PmeGpuProgramHandle pmeGpuProgram,
173 const CoordinatesVector &coordinates,
174 const ChargesVector &charges,
178 const index atomCount = coordinates.size();
179 GMX_RELEASE_ASSERT(atomCount == charges.size(), "Mismatch in atom data");
180 PmeSafePointer pmeSafe = pmeInitInternal(inputRec, mode, gpuInfo, pmeGpuProgram, atomCount, box);
181 pme_atomcomm_t *atc = nullptr;
186 atc = &(pmeSafe->atc[0]);
187 atc->x = const_cast<rvec *>(as_rvec_array(coordinates.data()));
188 atc->coefficient = const_cast<real *>(charges.data());
189 /* With decomposition there would be more boilerplate atc code here, e.g. do_redist_pos_coeffs */
193 gmx_pme_reinit_atoms(pmeSafe.get(), atomCount, charges.data());
194 pme_gpu_copy_input_coordinates(pmeSafe->gpu, as_rvec_array(coordinates.data()));
198 GMX_THROW(InternalError("Test not implemented for this mode"));
204 //! Getting local PME real grid pointer for test I/O
205 static real *pmeGetRealGridInternal(const gmx_pme_t *pme)
207 const size_t gridIndex = 0;
208 return pme->fftgrid[gridIndex];
211 //! Getting local PME real grid dimensions
212 static void pmeGetRealGridSizesInternal(const gmx_pme_t *pme,
214 IVec &gridSize, //NOLINT(google-runtime-references)
215 IVec &paddedGridSize) //NOLINT(google-runtime-references)
217 const size_t gridIndex = 0;
218 IVec gridOffsetUnused;
222 gmx_parallel_3dfft_real_limits(pme->pfft_setup[gridIndex], gridSize, gridOffsetUnused, paddedGridSize);
226 pme_gpu_get_real_grid_sizes(pme->gpu, &gridSize, &paddedGridSize);
230 GMX_THROW(InternalError("Test not implemented for this mode"));
234 //! Getting local PME complex grid pointer for test I/O
235 static t_complex *pmeGetComplexGridInternal(const gmx_pme_t *pme)
237 const size_t gridIndex = 0;
238 return pme->cfftgrid[gridIndex];
241 //! Getting local PME complex grid dimensions
242 static void pmeGetComplexGridSizesInternal(const gmx_pme_t *pme,
243 IVec &gridSize, //NOLINT(google-runtime-references)
244 IVec &paddedGridSize) //NOLINT(google-runtime-references)
246 const size_t gridIndex = 0;
247 IVec gridOffsetUnused, complexOrderUnused;
248 gmx_parallel_3dfft_complex_limits(pme->pfft_setup[gridIndex], complexOrderUnused, gridSize, gridOffsetUnused, paddedGridSize); //TODO: what about YZX ordering?
251 //! Getting the PME grid memory buffer and its sizes - template definition
252 template<typename ValueType> static void pmeGetGridAndSizesInternal(const gmx_pme_t * /*unused*/, CodePath /*unused*/, ValueType * & /*unused*/, IVec & /*unused*/, IVec & /*unused*/) //NOLINT(google-runtime-references)
254 GMX_THROW(InternalError("Deleted function call"));
255 // explicitly deleting general template does not compile in clang/icc, see https://llvm.org/bugs/show_bug.cgi?id=17537
258 //! Getting the PME real grid memory buffer and its sizes
259 template<> void pmeGetGridAndSizesInternal<real>(const gmx_pme_t *pme, CodePath mode, real * &grid, IVec &gridSize, IVec &paddedGridSize)
261 grid = pmeGetRealGridInternal(pme);
262 pmeGetRealGridSizesInternal(pme, mode, gridSize, paddedGridSize);
265 //! Getting the PME complex grid memory buffer and its sizes
266 template<> void pmeGetGridAndSizesInternal<t_complex>(const gmx_pme_t *pme, CodePath /*unused*/, t_complex * &grid, IVec &gridSize, IVec &paddedGridSize)
268 grid = pmeGetComplexGridInternal(pme);
269 pmeGetComplexGridSizesInternal(pme, gridSize, paddedGridSize);
272 //! PME spline calculation and charge spreading
273 void pmePerformSplineAndSpread(gmx_pme_t *pme, CodePath mode, // TODO const qualifiers elsewhere
274 bool computeSplines, bool spreadCharges)
276 GMX_RELEASE_ASSERT(pme != nullptr, "PME data is not initialized");
277 pme_atomcomm_t *atc = &(pme->atc[0]);
278 const size_t gridIndex = 0;
279 const bool computeSplinesForZeroCharges = true;
280 real *fftgrid = spreadCharges ? pme->fftgrid[gridIndex] : nullptr;
281 real *pmegrid = pme->pmegrid[gridIndex].grid.grid;
286 spread_on_grid(pme, atc, &pme->pmegrid[gridIndex], computeSplines, spreadCharges,
287 fftgrid, computeSplinesForZeroCharges, gridIndex);
288 if (spreadCharges && !pme->bUseThreads)
290 wrap_periodic_pmegrid(pme, pmegrid);
291 copy_pmegrid_to_fftgrid(pme, pmegrid, fftgrid, gridIndex);
296 pme_gpu_spread(pme->gpu, gridIndex, fftgrid, computeSplines, spreadCharges);
300 GMX_THROW(InternalError("Test not implemented for this mode"));
304 //! Getting the internal spline data buffer pointer
305 static real *pmeGetSplineDataInternal(const gmx_pme_t *pme, PmeSplineDataType type, int dimIndex)
307 GMX_ASSERT((0 <= dimIndex) && (dimIndex < DIM), "Invalid dimension index");
308 const pme_atomcomm_t *atc = &(pme->atc[0]);
309 const size_t threadIndex = 0;
310 real *splineBuffer = nullptr;
313 case PmeSplineDataType::Values:
314 splineBuffer = atc->spline[threadIndex].theta[dimIndex];
317 case PmeSplineDataType::Derivatives:
318 splineBuffer = atc->spline[threadIndex].dtheta[dimIndex];
322 GMX_THROW(InternalError("Unknown spline data type"));
328 void pmePerformSolve(const gmx_pme_t *pme, CodePath mode,
329 PmeSolveAlgorithm method, real cellVolume,
330 GridOrdering gridOrdering, bool computeEnergyAndVirial)
332 t_complex *h_grid = pmeGetComplexGridInternal(pme);
333 const bool useLorentzBerthelot = false;
334 const size_t threadIndex = 0;
338 if (gridOrdering != GridOrdering::YZX)
340 GMX_THROW(InternalError("Test not implemented for this mode"));
344 case PmeSolveAlgorithm::Coulomb:
345 solve_pme_yzx(pme, h_grid, cellVolume,
346 computeEnergyAndVirial, pme->nthread, threadIndex);
349 case PmeSolveAlgorithm::LennardJones:
350 solve_pme_lj_yzx(pme, &h_grid, useLorentzBerthelot,
351 cellVolume, computeEnergyAndVirial, pme->nthread, threadIndex);
355 GMX_THROW(InternalError("Test not implemented for this mode"));
362 case PmeSolveAlgorithm::Coulomb:
363 pme_gpu_solve(pme->gpu, h_grid, gridOrdering, computeEnergyAndVirial);
367 GMX_THROW(InternalError("Test not implemented for this mode"));
372 GMX_THROW(InternalError("Test not implemented for this mode"));
376 //! PME force gathering
377 void pmePerformGather(gmx_pme_t *pme, CodePath mode,
378 PmeForceOutputHandling inputTreatment, ForcesVector &forces)
380 pme_atomcomm_t *atc = &(pme->atc[0]);
381 const index atomCount = atc->n;
382 GMX_RELEASE_ASSERT(forces.size() == atomCount, "Invalid force buffer size");
383 const bool forceReductionWithInput = (inputTreatment == PmeForceOutputHandling::ReduceWithInput);
384 const real scale = 1.0;
385 const size_t threadIndex = 0;
386 const size_t gridIndex = 0;
387 real *pmegrid = pme->pmegrid[gridIndex].grid.grid;
388 real *fftgrid = pme->fftgrid[gridIndex];
393 atc->f = as_rvec_array(forces.begin());
394 if (atc->nthread == 1)
396 // something which is normally done in serial spline computation (make_thread_local_ind())
397 atc->spline[threadIndex].n = atomCount;
399 copy_fftgrid_to_pmegrid(pme, fftgrid, pmegrid, gridIndex, pme->nthread, threadIndex);
400 unwrap_periodic_pmegrid(pme, pmegrid);
401 gather_f_bsplines(pme, pmegrid, !forceReductionWithInput, atc, &atc->spline[threadIndex], scale);
406 // Variable initialization needs a non-switch scope
407 auto stagingForces = pme_gpu_get_forces(pme->gpu);
408 GMX_ASSERT(forces.size() == stagingForces.size(), "Size of force buffers did not match");
409 if (forceReductionWithInput)
411 for (index i = 0; i != forces.size(); ++i)
413 stagingForces[i] = forces[i];
416 pme_gpu_gather(pme->gpu, inputTreatment, reinterpret_cast<float *>(fftgrid));
417 for (index i = 0; i != forces.size(); ++i)
419 forces[i] = stagingForces[i];
425 GMX_THROW(InternalError("Test not implemented for this mode"));
429 //! PME test finalization before fetching the outputs
430 void pmeFinalizeTest(const gmx_pme_t *pme, CodePath mode)
438 pme_gpu_synchronize(pme->gpu);
442 GMX_THROW(InternalError("Test not implemented for this mode"));
446 //! Setting atom spline values/derivatives to be used in spread/gather
447 void pmeSetSplineData(const gmx_pme_t *pme, CodePath mode,
448 const SplineParamsDimVector &splineValues, PmeSplineDataType type, int dimIndex)
450 const pme_atomcomm_t *atc = &(pme->atc[0]);
451 const index atomCount = atc->n;
452 const index pmeOrder = pme->pme_order;
453 const index dimSize = pmeOrder * atomCount;
454 GMX_RELEASE_ASSERT(dimSize == splineValues.size(), "Mismatch in spline data");
455 real *splineBuffer = pmeGetSplineDataInternal(pme, type, dimIndex);
460 std::copy(splineValues.begin(), splineValues.end(), splineBuffer);
464 std::copy(splineValues.begin(), splineValues.end(), splineBuffer);
465 pme_gpu_transform_spline_atom_data(pme->gpu, atc, type, dimIndex, PmeLayoutTransform::HostToGpu);
469 GMX_THROW(InternalError("Test not implemented for this mode"));
473 //! Setting gridline indices to be used in spread/gather
474 void pmeSetGridLineIndices(const gmx_pme_t *pme, CodePath mode,
475 const GridLineIndicesVector &gridLineIndices)
477 const pme_atomcomm_t *atc = &(pme->atc[0]);
478 const index atomCount = atc->n;
479 GMX_RELEASE_ASSERT(atomCount == gridLineIndices.size(), "Mismatch in gridline indices size");
481 IVec paddedGridSizeUnused, gridSize(0, 0, 0);
482 pmeGetRealGridSizesInternal(pme, mode, gridSize, paddedGridSizeUnused);
484 for (const auto &index : gridLineIndices)
486 for (int i = 0; i < DIM; i++)
488 GMX_RELEASE_ASSERT((0 <= index[i]) && (index[i] < gridSize[i]), "Invalid gridline index");
495 memcpy(pme->gpu->staging.h_gridlineIndices, gridLineIndices.data(), atomCount * sizeof(gridLineIndices[0]));
499 // incompatible IVec and ivec assignment?
500 //std::copy(gridLineIndices.begin(), gridLineIndices.end(), atc->idx);
501 memcpy(atc->idx, gridLineIndices.data(), atomCount * sizeof(gridLineIndices[0]));
505 GMX_THROW(InternalError("Test not implemented for this mode"));
509 //! Getting plain index into the complex 3d grid
510 inline size_t pmeGetGridPlainIndexInternal(const IVec &index, const IVec &paddedGridSize, GridOrdering gridOrdering)
513 switch (gridOrdering)
515 case GridOrdering::YZX:
516 result = (index[YY] * paddedGridSize[ZZ] + index[ZZ]) * paddedGridSize[XX] + index[XX];
519 case GridOrdering::XYZ:
520 result = (index[XX] * paddedGridSize[YY] + index[YY]) * paddedGridSize[ZZ] + index[ZZ];
524 GMX_THROW(InternalError("Test not implemented for this mode"));
529 //! Setting real or complex grid
530 template<typename ValueType>
531 static void pmeSetGridInternal(const gmx_pme_t *pme, CodePath mode,
532 GridOrdering gridOrdering,
533 const SparseGridValuesInput<ValueType> &gridValues)
535 IVec gridSize(0, 0, 0), paddedGridSize(0, 0, 0);
537 pmeGetGridAndSizesInternal<ValueType>(pme, mode, grid, gridSize, paddedGridSize);
541 case CodePath::GPU: // intentional absence of break, the grid will be copied from the host buffer in testing mode
543 std::memset(grid, 0, paddedGridSize[XX] * paddedGridSize[YY] * paddedGridSize[ZZ] * sizeof(ValueType));
544 for (const auto &gridValue : gridValues)
546 for (int i = 0; i < DIM; i++)
548 GMX_RELEASE_ASSERT((0 <= gridValue.first[i]) && (gridValue.first[i] < gridSize[i]), "Invalid grid value index");
550 const size_t gridValueIndex = pmeGetGridPlainIndexInternal(gridValue.first, paddedGridSize, gridOrdering);
551 grid[gridValueIndex] = gridValue.second;
556 GMX_THROW(InternalError("Test not implemented for this mode"));
560 //! Setting real grid to be used in gather
561 void pmeSetRealGrid(const gmx_pme_t *pme, CodePath mode,
562 const SparseRealGridValuesInput &gridValues)
564 pmeSetGridInternal<real>(pme, mode, GridOrdering::XYZ, gridValues);
567 //! Setting complex grid to be used in solve
568 void pmeSetComplexGrid(const gmx_pme_t *pme, CodePath mode,
569 GridOrdering gridOrdering,
570 const SparseComplexGridValuesInput &gridValues)
572 pmeSetGridInternal<t_complex>(pme, mode, gridOrdering, gridValues);
575 //! Getting the single dimension's spline values or derivatives
576 SplineParamsDimVector pmeGetSplineData(const gmx_pme_t *pme, CodePath mode,
577 PmeSplineDataType type, int dimIndex)
579 GMX_RELEASE_ASSERT(pme != nullptr, "PME data is not initialized");
580 const pme_atomcomm_t *atc = &(pme->atc[0]);
581 const size_t atomCount = atc->n;
582 const size_t pmeOrder = pme->pme_order;
583 const size_t dimSize = pmeOrder * atomCount;
585 real *sourceBuffer = pmeGetSplineDataInternal(pme, type, dimIndex);
586 SplineParamsDimVector result;
590 pme_gpu_transform_spline_atom_data(pme->gpu, atc, type, dimIndex, PmeLayoutTransform::GpuToHost);
591 result = arrayRefFromArray(sourceBuffer, dimSize);
595 result = arrayRefFromArray(sourceBuffer, dimSize);
599 GMX_THROW(InternalError("Test not implemented for this mode"));
604 //! Getting the gridline indices
605 GridLineIndicesVector pmeGetGridlineIndices(const gmx_pme_t *pme, CodePath mode)
607 GMX_RELEASE_ASSERT(pme != nullptr, "PME data is not initialized");
608 const pme_atomcomm_t *atc = &(pme->atc[0]);
609 const size_t atomCount = atc->n;
611 GridLineIndicesVector gridLineIndices;
615 gridLineIndices = arrayRefFromArray(reinterpret_cast<IVec *>(pme->gpu->staging.h_gridlineIndices), atomCount);
619 gridLineIndices = arrayRefFromArray(reinterpret_cast<IVec *>(atc->idx), atomCount);
623 GMX_THROW(InternalError("Test not implemented for this mode"));
625 return gridLineIndices;
628 //! Getting real or complex grid - only non zero values
629 template<typename ValueType>
630 static SparseGridValuesOutput<ValueType> pmeGetGridInternal(const gmx_pme_t *pme, CodePath mode, GridOrdering gridOrdering)
632 IVec gridSize(0, 0, 0), paddedGridSize(0, 0, 0);
634 pmeGetGridAndSizesInternal<ValueType>(pme, mode, grid, gridSize, paddedGridSize);
635 SparseGridValuesOutput<ValueType> gridValues;
638 case CodePath::GPU: // intentional absence of break
641 for (int ix = 0; ix < gridSize[XX]; ix++)
643 for (int iy = 0; iy < gridSize[YY]; iy++)
645 for (int iz = 0; iz < gridSize[ZZ]; iz++)
647 IVec temp(ix, iy, iz);
648 const size_t gridValueIndex = pmeGetGridPlainIndexInternal(temp, paddedGridSize, gridOrdering);
649 const ValueType value = grid[gridValueIndex];
650 if (value != ValueType {})
652 auto key = formatString("Cell %d %d %d", ix, iy, iz);
653 gridValues[key] = value;
661 GMX_THROW(InternalError("Test not implemented for this mode"));
666 //! Getting the real grid (spreading output of pmePerformSplineAndSpread())
667 SparseRealGridValuesOutput pmeGetRealGrid(const gmx_pme_t *pme, CodePath mode)
669 return pmeGetGridInternal<real>(pme, mode, GridOrdering::XYZ);
672 //! Getting the complex grid output of pmePerformSolve()
673 SparseComplexGridValuesOutput pmeGetComplexGrid(const gmx_pme_t *pme, CodePath mode,
674 GridOrdering gridOrdering)
676 return pmeGetGridInternal<t_complex>(pme, mode, gridOrdering);
679 //! Getting the reciprocal energy and virial
680 PmeSolveOutput pmeGetReciprocalEnergyAndVirial(const gmx_pme_t *pme, CodePath mode,
681 PmeSolveAlgorithm method)
685 matrix virialTemp = {{0}}; //TODO get rid of
691 case PmeSolveAlgorithm::Coulomb:
692 get_pme_ener_vir_q(pme->solve_work, pme->nthread, &energy, virialTemp);
695 case PmeSolveAlgorithm::LennardJones:
696 get_pme_ener_vir_lj(pme->solve_work, pme->nthread, &energy, virialTemp);
700 GMX_THROW(InternalError("Test not implemented for this mode"));
706 case PmeSolveAlgorithm::Coulomb:
707 pme_gpu_get_energy_virial(pme->gpu, &energy, virialTemp);
711 GMX_THROW(InternalError("Test not implemented for this mode"));
716 GMX_THROW(InternalError("Test not implemented for this mode"));
718 for (int i = 0; i < DIM; i++)
720 for (int j = 0; j < DIM; j++)
722 virial[i * DIM + j] = virialTemp[i][j];
725 return std::make_tuple(energy, virial);