sources/libs/nvll_osd/memory.hpp

Go to the documentation of this file.

/*
 * SPDX-FileCopyrightText: Copyright (c) 2023-2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 * SPDX-License-Identifier: Apache-2.0
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 * http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */
 
#ifndef MEMORY_HPP
#define MEMORY_HPP
 
#include <cuda_runtime.h>
#include <stdio.h>
 
#define checkRuntime(call)  check_runtime(call, #call, __LINE__, __FILE__)
 
#define CUOSD_PRINT_E(f_, ...) \
  fprintf(stderr, "[cuOSD Error] at %s:%d : " f_, (const char*)__FILE__, __LINE__, ##__VA_ARGS__)
 
#define CUOSD_PRINT_W(f_, ...) \
  printf("[cuOSD Warning] at %s:%d : " f_, (const char*)__FILE__, __LINE__, ##__VA_ARGS__)
 
static bool inline check_runtime(cudaError_t e, const char* call, int line, const char *file) {
    if (e != cudaSuccess) {
        fprintf(stderr, "CUDA Runtime error %s # %s, code = %s [ %d ] in file %s:%d\n", call, cudaGetErrorString(e), cudaGetErrorName(e), e, file, line);
        return false;
    }
    return true;
}
 
template<typename T>
class Memory{
public:
    T* host()   const{return host_;}
    T* device() const{return device_;}
    size_t size() const{return size_;}
    size_t bytes() const{return size_ * sizeof(T);}
 
    virtual ~Memory() {
        free_memory();
    }
 
    void copy_host_to_device(cudaStream_t stream=nullptr) {
        checkRuntime(cudaMemcpyAsync(device_, host_, bytes(), cudaMemcpyHostToDevice, stream));
    }
    void copy_device_to_host(cudaStream_t stream=nullptr) {
        checkRuntime(cudaMemcpyAsync(host_, device_, bytes(), cudaMemcpyDeviceToHost, stream));
    }
 
    void alloc_or_resize_to(size_t size) {
        if (capacity_ < size) {
            free_memory();
 
            checkRuntime(cudaMallocHost(&host_, size * sizeof(T)));
            checkRuntime(cudaMalloc(&device_, size * sizeof(T)));
            capacity_ = size;
        }
        size_ = size;
    }
 
    void free_memory() {
        if (host_ || device_) {
            checkRuntime(cudaFreeHost(host_));
            checkRuntime(cudaFree(device_));
            host_     = nullptr;
            device_   = nullptr;
            capacity_ = 0;
            size_     = 0;
        }
    }
 
private:
    T* host_         = nullptr;
    T* device_       = nullptr;
    size_t size_     = 0;
    size_t capacity_ = 0;
};
 
 
 
#endif // MEMORY_HPP