2 * This file is part of the GROMACS molecular simulation package.
4 * Copyright (c) 2018,2019,2020, 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.
35 /*! \libinternal \file
36 * \brief Implements a GpuEventSynchronizer class for CUDA.
38 * \author Aleksei Iupinov <a.yupinov@gmail.com>
41 #ifndef GMX_GPU_UTILS_GPUEVENTSYNCHRONIZER_CUH
42 #define GMX_GPU_UTILS_GPUEVENTSYNCHRONIZER_CUH
44 #include "gromacs/gpu_utils/device_stream.h"
45 #include "gromacs/gpu_utils/gputraits.cuh"
46 #include "gromacs/utility/gmxassert.h"
48 /*! \libinternal \brief
49 * A class which allows for CPU thread to mark and wait for certain GPU stream execution point.
50 * The event can be put into the stream with markEvent() and then later waited on with waitForEvent().
51 * This can be repeated as necessary, but the current implementation does not allow waiting on
52 * completed event more than once, expecting only exact pairs of markEvent(stream); waitForEvent().
53 * The class generally attempts to track the correctness of its state transitions, but
54 * please note that calling waitForEvent() right after the construction will succeed with CUDA but fail with OpenCL.
56 * Another possible mode of operation can be implemented if needed:
57 * multiple calls to waitForEvent() after a single markEvent().
58 * For this, only some small alterations to gpueventsynchronizer_ocl.h need to be made.
59 * This was tested to work both with CUDA and NVidia OpenCL, but not with AMD/Intel OpenCL.
61 class GpuEventSynchronizer
64 GpuEventSynchronizer()
66 cudaError_t gmx_used_in_debug stat = cudaEventCreateWithFlags(&event_, cudaEventDisableTiming);
67 GMX_RELEASE_ASSERT(stat == cudaSuccess, "cudaEventCreate failed");
69 ~GpuEventSynchronizer()
71 cudaError_t gmx_used_in_debug stat = cudaEventDestroy(event_);
72 GMX_ASSERT(stat == cudaSuccess, "cudaEventDestroy failed");
75 GpuEventSynchronizer(const GpuEventSynchronizer&) = delete;
77 GpuEventSynchronizer& operator=(GpuEventSynchronizer&&) = delete;
78 //! Moving is disabled but can be considered in the future if needed
79 GpuEventSynchronizer(GpuEventSynchronizer&&) = delete;
81 /*! \brief Marks the synchronization point in the \p stream.
82 * Should be followed by waitForEvent().
84 inline void markEvent(const DeviceStream& deviceStream)
86 cudaError_t gmx_used_in_debug stat = cudaEventRecord(event_, deviceStream.stream());
87 GMX_ASSERT(stat == cudaSuccess, "cudaEventRecord failed");
89 /*! \brief Synchronizes the host thread on the marked event. */
90 inline void waitForEvent()
92 cudaError_t gmx_used_in_debug stat = cudaEventSynchronize(event_);
93 GMX_ASSERT(stat == cudaSuccess, "cudaEventSynchronize failed");
95 /*! \brief Enqueues a wait for the recorded event in stream \p stream */
96 inline void enqueueWaitEvent(const DeviceStream& deviceStream)
98 cudaError_t gmx_used_in_debug stat = cudaStreamWaitEvent(deviceStream.stream(), event_, 0);
99 GMX_ASSERT(stat == cudaSuccess, "cudaStreamWaitEvent failed");