device.h

Go to the documentation of this file.

/*  This file is part of the OpenLB library
 *
 *  Copyright (C) 2022 Adrian Kummerlaender
 *
 *  E-mail contact: info@openlb.net
 *  The most recent release of OpenLB can be downloaded at
 *  <http://www.openlb.net/>
 *
 *  This program is free software; you can redistribute it and/or
 *  modify it under the terms of the GNU General Public License
 *  as published by the Free Software Foundation; either version 2
 *  of the License, or (at your option) any later version.
 *
 *  This program 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 General Public License for more details.
 *
 *  You should have received a copy of the GNU General Public
 *  License along with this program; if not, write to the Free
 *  Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
 *  Boston, MA  02110-1301, USA.
*/
 
#ifndef GPU_CUDA_DEVICE_H
#define GPU_CUDA_DEVICE_H
 
#include <cuda_runtime_api.h>
 
namespace olb {
 
namespace gpu {
 
namespace cuda {
 
namespace device {
 
int getCount();
 
void check();
 
void synchronize();
 
int get();
 
void copyToHost(void* src, void* dst, std::size_t count);
void copyToDevice(void* src, void* dst, std::size_t count);
 
template <typename T>
T* malloc(std::size_t size);
 
std::size_t getDevicePageSize();
template <typename T>
std::size_t getPageAlignedCount(std::size_t count)
{
  const std::size_t page_size = getDevicePageSize();
  const std::size_t size = ((count * sizeof(T) - 1) / page_size + 1) * page_size;
  const std::size_t volume = size / sizeof(T);
 
  if (size % page_size != 0) {
    throw std::invalid_argument("Buffer size must be multiple of PAGE_SIZE");
  }
 
  return volume;
};
 
template <typename T>
class unique_ptr {
private:
  T* _ptr;
 
public:
  unique_ptr():
    _ptr{nullptr} { }
  unique_ptr(T* ptr):
    _ptr{ptr} { }
  unique_ptr(unique_ptr&& rhs):
    _ptr{rhs.release()} { }
 
  ~unique_ptr();
 
  unique_ptr& operator=(unique_ptr&& rhs);
 
  void reset(T* ptr) {
    operator=(unique_ptr<T>(ptr));
  }
 
  unique_ptr& operator=(T* ptr) {
    reset(ptr);
    return *this;
  }
 
  T* release() {
    T* ptr = _ptr;
    _ptr = nullptr;
    return ptr;
  }
 
  const T* get() const {
    return _ptr;
  }
  T* get() {
    return _ptr;
  }
 
};
 
 
class Stream {
private:
  cudaStream_t _stream;
 
public:
  Stream(unsigned int flags);
  ~Stream();
 
  cudaStream_t& get() {
    return _stream;
  }
 
  void synchronize();
 
};
 
void asyncCopyToHost(Stream& stream, void* src, void* dst, std::size_t count);
void asyncCopyToDevice(Stream& stream, void* src, void* dst, std::size_t count);
 
}
 
}
 
}
 
}
 
#endif