CANN/atvoss算子开发指南

算子开发指南【免费下载链接】atvossATVOSSAscend C Templates for Vector Operator Subroutines是一套基于Ascend C开发的Vector算子库致力于为昇腾硬件上的Vector类融合算子提供极简、高效、高性能、高拓展的编程方式。项目地址: https://gitcode.com/cann/atvoss本指南以muls矩阵标量乘法算子为例详细介绍在 Atvoss 框架中开发自定义算子的完整流程。通过该示例您将学会如何实现一个二元运算算子。1. 算子需求1.1 功能描述muls算子实现矩阵标量乘法功能将输入矩阵的每个元素与一个标量值相乘。数学表达式 $$ Y X \times scalar $$其中$X$输入矩阵任意维度$scalar$标量乘数$Y$输出矩阵与输入矩阵形状相同1.2 支持的数据类型x输入类型scalar输入类型输出类型floatfloatfloathalfhalfhalf2. 设计规格2.1 核心组件组件类型说明MulsComputeStruct计算逻辑TileShapeShape32分块大小blockPolicyDefaultBlockPolicy默认分块策略kernelPolicyDefaultKernelPolicy默认分核策略2.2 表达式定义out in * scalar3. 编写核心代码逻辑3.1 导入头文件#include kernel_operator.h // 必须AscendC 底层 API 定义 #include atvoss.h // 必须Atvoss 框架头文件 #include example_common.h // 可选示例通用工具ACL 初始化、内存管理等通用代码3.2 配置结构体定义建议所有算子定义相关的配置都定义在一个Struct中便于后续实例化使用。// 定义TileShape参数配置 static constexpr int32_t WIDTH 32; // 算子配置结构体 template typename TensorDtype, typename ScalarDtype struct MulsConfig { using TileShape Atvoss::ShapeWIDTH; // ... };3.2.1 Compute 结构体在MulsConfig中增加Compute定义Compute()接口必须符合以下固定格式struct MulsCompute { template template typename class Tensor __host_aicore__ constexpr auto Compute() const { // 1. 定义输入输出参数描述 auto in Atvoss::PlaceHolder1, TensorTensorDtype, Atvoss::ParamUsage::IN(); auto scalar Atvoss::PlaceHolder2, ScalarDtype, Atvoss::ParamUsage::IN(); auto out Atvoss::PlaceHolder3, TensorTensorDtype, Atvoss::ParamUsage::OUT(); // 2. 返回计算表达式 return (out in * scalar); }; };接口说明先定义输入输出参数再定义基于输入输出的计算表达式使用逗号分隔多个计算步骤更多运算API接口参考API3.2.2 策略配置在MulsConfig中增加策略配置// 块级策略指定分块形状 static constexpr Atvoss::Ele::DefaultBlockPolicyTileShape blockPolicy{TileShape{}}; // 核策略使用均匀分段 static constexpr Atvoss::Ele::DefaultKernelPolicy kernelPolicy{Atvoss::Ele::DefaultSegmentPolicy::UniformSegment}; // 指定目标架构 using ArchTag Atvoss::Arch::DAV_3510;3.2.3 算子Op定义在MulsConfig中增加算子Op定义// 定义BlockOp using BlockOp Atvoss::Ele::BlockBuilder MulsCompute, ArchTag, blockPolicy, Atvoss::Ele::DefaultBlockConfig, Atvoss::Ele::DefaultBlockSchedule; // 定义KernelOp using KernelOp Atvoss::Ele::KernelBuilder BlockOp, kernelPolicy, Atvoss::Ele::DefaultKernelConfig, Atvoss::Ele::DefaultKernelSchedule; // 定义DeviceOp using DeviceOp Atvoss::DeviceAdapterKernelOp;3.3 运行时执行逻辑template typename TensorDtype, typename ScalarDtype static void ProcessMuls(std::vectorTensorDtype hostInput, ScalarDtype scalar, std::vectorTensorDtype hostOutput, const std::vectorint shape) { // Step 1: ACL 初始化 CHECK_ACL_RET(aclInit(nullptr)); auto finalizeGuard ReleaseSource([]() { aclFinalize(); }); // Step 2: 设置 device const int32_t deviceId 0; CHECK_ACL_RET(aclrtSetDevice(deviceId)); auto deviceResetGuard ReleaseSource([deviceId]() { aclrtResetDevice(deviceId); }); // Step 3: 创建 Context aclrtContext context nullptr; CHECK_ACL_RET(aclrtCreateContext(context, deviceId)); auto contextDestroyGuard ReleaseSource([context]() { aclrtDestroyContext(context); }); // Step 4: 创建 Stream aclrtStream stream nullptr; CHECK_ACL_RET(aclrtCreateStream(stream)); auto streamDestroyGuard ReleaseSource([stream]() { aclrtDestroyStream(stream); }); // Step 5: 准备数据 const size_t shapeSize std::accumulate(shape.begin(), shape.end(), size_t{1}, std::multiplies{}); const size_t inputSize shapeSize * sizeof(TensorDtype); const size_t outputSize inputSize; // Step 6: 分配设备内存 void* rawInput nullptr; CHECK_ACL_RET(aclrtMalloc(rawInput, inputSize, ACL_MEM_MALLOC_HUGE_FIRST)); auto inputFreeGuard ReleaseSource([rawInput]() { aclrtFree(rawInput); }); TensorDtype* deviceInput static_castTensorDtype*(rawInput); void* rawOutput nullptr; CHECK_ACL_RET(aclrtMalloc(rawOutput, outputSize, ACL_MEM_MALLOC_HUGE_FIRST)); auto outputFreeGuard ReleaseSource([rawOutput]() { aclrtFree(rawOutput); }); ScalarDtype* deviceOutput static_castScalarDtype*(rawOutput); // Step 7: Host → Device 数据拷贝 CHECK_ACL_RET(aclrtMemcpy(deviceInput, inputSize, hostInput.data(), inputSize, ACL_MEMCPY_HOST_TO_DEVICE)); // Step 8: 构造 Tensor 并执行算子 uint32_t shapeArray[8] {0}; std::copy(shape.begin(), shape.end(), shapeArray); Atvoss::TensorTensorDtype in(deviceInput, shapeArray, shape.size()); Atvoss::TensorScalarDtype out(deviceOutput, shapeArray, shape.size()); auto arguments Atvoss::ArgumentsBuilder{}.inputOutput(in, scalar, out).build(); // 根据输入类型选择算子 if constexpr (std::is_same_vTensorDtype, ScalarDtype) { using DeviceOp typename MulsConfigTensorDtype, ScalarDtype::DeviceOp; DeviceOp deviceOp; deviceOp.Run(arguments, stream); } else { static_assert(The data types of x and scalar must be the same.); } // Step 9: 同步并拷回结果 CHECK_ACL_RET(aclrtSynchronizeStream(stream)); CHECK_ACL_RET(aclrtMemcpy(hostOutput.data(), outputSize, deviceOutput, outputSize, ACL_MEMCPY_DEVICE_TO_HOST)); }3.4 主函数int main(int argc, char const* argv[]) { std::cout Start muls std::endl; // 定义输入数据 const std::vectorint shape {32, 32}; std::vectorfloat hostInput(shape[0] * shape[1], 3.0f); std::vectorfloat hostOutput(shape[0] * shape[1]); float scalar 3.0f; // 调用模板函数 ProcessMulsfloat, float(hostInput, scalar, hostOutput, shape); // 验证结果 std::vectorfloat golden(shape[0] * shape[1], 9.0f); if (!VerifyResults(golden, hostOutput)) { std::cout Accuracy verification failed. std::endl; } else { std::cout Accuracy verification passed. std::endl; } return 0; }4. 编写 CMake 文件文件位置CMakeLists.txtcmake_minimum_required(VERSION 3.16) project(test_muls) # 这部分需要根据用户自己环境的路径进行配置 set(ATVOSS_INCLUDE_DIRS /mnt/workspace/gitCode/cann/atvoss/include) set(ASCEND_DIR /home/developer/Ascend/cann-9.0.0) set(ASCEND_INCLUDE_DIRS ${ASCEND_DIR}/include) set(ATVOSS_EXAMPLES_COMMON_SOURCE_DIR /mnt/workspace/gitCode/cann/atvoss/examples/common) set(NPU_ARCH dav-3510) set(INCLUDE_DIRECTORIES ${ASCEND_INCLUDE_DIRS} ${ATVOSS_INCLUDE_DIRS} ${ATVOSS_EXAMPLES_COMMON_SOURCE_DIR} ) set(LINK_DIRECTORIES ${ASCEND_DIR}/lib64 ) set(LINK_LIBRARIES ascendcl platform register tiling_api runtime ) set(COMPILE_OPTIONS -O3 -fdiagnostics-coloralways -w ) set(BISHENG ${ASCEND_DIR}/bin/bisheng) unset(CMAKE_CXX_FLAGS) set(CMAKE_EXPORT_COMPILE_COMMANDS ON) set(CMAKE_CXX_STANDARD 17) set(CMAKE_CXX_STANDARD_REQUIRED ON) set(CMAKE_POSITION_INDEPENDENT_CODE ON) set(CMAKE_C_COMPILER ${BISHENG}) set(CMAKE_CXX_COMPILER ${BISHENG}) set(CMAKE_LINKER ${BISHENG}) set(COMPILE_FLAGS --npu-arch${NPU_ARCH} -xasc ) set_source_files_properties( test_muls.cpp PROPERTIES LANGUAGE CXX COMPILE_FLAGS ${COMPILE_FLAGS} ) add_executable(test_muls test_muls.cpp) set_target_properties(test_muls PROPERTIES LINK_DEPENDS_NO_SHARED ON POSITION_INDEPENDENT_CODE ON ) target_compile_options(test_muls PRIVATE ${COMPILE_OPTIONS}) target_include_directories(test_muls PRIVATE ${INCLUDE_DIRECTORIES}) target_link_directories(test_muls PRIVATE ${LINK_DIRECTORIES}) target_link_libraries(test_muls PRIVATE ${LINK_LIBRARIES})5. 编译与执行5.1 环境准备确保以下环境变量已设置# 默认路径安装以root用户为例非root用户将/usr/local替换为${HOME} source /usr/local/Ascend/cann/set_env.sh # 指定路径安装 source ${install_path}/cann/set_env.sh5.2 编译命令# 编译 muls 算子 mkdir build cd build cmake .. make5.3 编译输出成功编译后可执行文件位于build/test_muls5.4 执行命令# 执行算子用例 ./build/test_muls6. 查看输出结果6.1 正常输出Start muls Accuracy verification passed.7. 参考文档ATVOSS接口参考API 运算API接口参考API【免费下载链接】atvossATVOSSAscend C Templates for Vector Operator Subroutines是一套基于Ascend C开发的Vector算子库致力于为昇腾硬件上的Vector类融合算子提供极简、高效、高性能、高拓展的编程方式。项目地址: https://gitcode.com/cann/atvoss创作声明：本文部分内容由AI辅助生成（AIGC），仅供参考