aclnnAddmm/aclnnInplaceAddmm

接口原型

  • aclnnAddmm和aclnnInplaceAddmm实现相同的功能,接口使用区别如下,请根据自身实际场景选择合适接口。
    • aclnnAddmm:需新建一个输出张量对象存储计算结果。
    • aclnnInplaceAddmm:无需新建输出张量对象,直接在输入张量的内存中存储计算结果。
  • 每个算子分为两段,必须先调用“aclnnXxxGetWorkspaceSize”接口获取入参并根据计算流程计算所需workspace大小,再调用“aclnnXxx”接口执行计算。

功能描述

aclnnAddmmGetWorkspaceSize

aclnnAddmm

aclnnInplaceAddmmGetWorkspaceSize

aclnnInplaceAddmm

约束与限制

对于Atlas 训练系列产品,Cube单元不支持FLOAT32计算。当输入为FLOAT32,可通过设置cubeMathType=1(ALLOW_FP32_DOWN_PRECISION)来允许接口内部cast到FLOAT16进行计算。

调用示例

#include <iostream>
#include <vector>
#include "acl/acl.h"
#include "aclnnop/level2/aclnn_addmm.h"

#define CHECK_RET(cond, return_expr) \
  do {                               \
    if (!(cond)) {                   \
      return_expr;                   \
    }                                \
  } while (0)

#define LOG_PRINT(message, ...)     \
  do {                              \
    printf(message, ##__VA_ARGS__); \
  } while (0)

int64_t GetShapeSize(const std::vector<int64_t>& shape) {
  int64_t shape_size = 1;
  for (auto i : shape) {
    shape_size *= i;
  }
  return shape_size;
}

int Init(int32_t deviceId, aclrtContext* context, aclrtStream* stream) {
  // 固定写法,acl初始化
  auto ret = aclrtSetDevice(deviceId);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtSetDevice failed. ERROR: %d\n", ret); return ret);
  ret = aclrtCreateContext(context, deviceId);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtCreateContext failed. ERROR: %d\n", ret); return ret);
  ret = aclrtSetCurrentContext(*context);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtSetCurrentContext failed. ERROR: %d\n", ret); return ret);
  ret = aclrtCreateStream(stream);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtCreateStream failed. ERROR: %d\n", ret); return ret);

  ret = aclInit(nullptr);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclInit failed. ERROR: %d\n", ret); return ret);
  return 0;
}

template <typename T>
int CreateAclTensor(const std::vector<T>& hostData, const std::vector<int64_t>& shape, void** deviceAddr,
                    aclDataType dataType, aclTensor** tensor) {
  auto size = GetShapeSize(shape) * sizeof(T);
  // 调用aclrtMalloc申请device侧内存
  auto ret = aclrtMalloc(deviceAddr, size, ACL_MEM_MALLOC_HUGE_FIRST);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtMalloc failed. ERROR: %d\n", ret); return ret);

  // 调用aclrtMemcpy将host侧数据拷贝到device侧内存上
  ret = aclrtMemcpy(*deviceAddr, size, hostData.data(), size, ACL_MEMCPY_HOST_TO_DEVICE);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtMemcpy failed. ERROR: %d\n", ret); return ret);

  // 计算连续tensor的strides
  std::vector<int64_t> strides(shape.size(), 1);
  for (int64_t i = shape.size() - 2; i >= 0; i--) {
    strides[i] = shape[i + 1] * strides[i + 1];
  }

  // 调用aclCreateTensor接口创建aclTensor
  *tensor = aclCreateTensor(shape.data(), shape.size(), dataType, strides.data(), 0, aclFormat::ACL_FORMAT_ND,
                            shape.data(), shape.size(), *deviceAddr);
  return 0;
}

int main() {
  // 1. (固定写法)device/context/stream初始化, 参考acl对外接口列表
  // 根据自己的实际device填写deviceId
  int32_t deviceId = 0;
  aclrtContext context;
  aclrtStream stream;
  auto ret = Init(deviceId, &context, &stream);
  // check根据自己的需要处理
  CHECK_RET(ret == 0, LOG_PRINT("Init acl failed. ERROR: %d\n", ret); return ret);

  // 2. 构造输入与输出,需要根据API的接口自定义构造
  std::vector<int64_t> selfShape = {2, 2};
  std::vector<int64_t> mat1Shape = {2, 2};
  std::vector<int64_t> mat2Shape = {2, 2};
  std::vector<int64_t> outShape = {2, 2};
  void* selfDeviceAddr = nullptr;
  void* mat1DeviceAddr = nullptr;
  void* mat2DeviceAddr = nullptr;
  void* outDeviceAddr = nullptr;
  aclTensor* self = nullptr;
  aclTensor* mat1 = nullptr;
  aclTensor* mat2 = nullptr;
  aclScalar* alpha = nullptr;
  aclScalar* beta = nullptr;
  aclTensor* out = nullptr;
  std::vector<float> selfHostData = {1, 1, 1, 1};
  std::vector<float> mat1HostData = {1, 1, 1, 1};
  std::vector<float> mat2HostData = {1, 1, 1, 1};
  std::vector<float> outHostData = {0, 0, 0, 0};
  float alphaValue = 1.0f;
  float betaValue = 1.0f;
  int8_t cubeMathType = 1;
  // 创建self aclTensor
  ret = CreateAclTensor(selfHostData, selfShape, &selfDeviceAddr, aclDataType::ACL_FLOAT, &self);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  // 创建mat1 aclTensor
  ret = CreateAclTensor(mat1HostData, mat1Shape, &mat1DeviceAddr, aclDataType::ACL_FLOAT, &mat1);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  // 创建mat2 aclTensor
  ret = CreateAclTensor(mat2HostData, mat2Shape, &mat2DeviceAddr, aclDataType::ACL_FLOAT, &mat2);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  // 创建out aclTensor
  ret = CreateAclTensor(outHostData, outShape, &outDeviceAddr, aclDataType::ACL_FLOAT, &out);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  // 创建alpha aclScalar
  alpha = aclCreateScalar(&alphaValue, aclDataType::ACL_FLOAT);
  CHECK_RET(alpha != nullptr, return ret);
  // 创建upper aclScalar
  beta = aclCreateScalar(&betaValue, aclDataType::ACL_FLOAT);
  CHECK_RET(beta != nullptr, return ret);

  // 3. 调用CANN算子库API
  uint64_t workspaceSize = 0;
  aclOpExecutor* executor;
  // 调用aclnnAddmm第一段接口
  ret = aclnnAddmmGetWorkspaceSize(self, mat1, mat2, beta, alpha, out, cubeMathType, &workspaceSize, &executor);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnAddmmGetWorkspaceSize failed. ERROR: %d\n", ret); return ret);
  // 根据第一段接口计算出的workspaceSize申请device内存
  void* workspaceAddr = nullptr;
  if (workspaceSize > 0) {
    ret = aclrtMalloc(&workspaceAddr, workspaceSize, ACL_MEM_MALLOC_HUGE_FIRST);
    CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("allocate workspace failed. ERROR: %d\n", ret); return ret;);
  }
  // 调用aclnnAddmm第二段接口
  ret = aclnnAddmm(workspaceAddr, workspaceSize, executor, stream);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnAddmm failed. ERROR: %d\n", ret); return ret);

  // 4. (固定写法)同步等待任务执行结束
  ret = aclrtSynchronizeStream(stream);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtSynchronizeStream failed. ERROR: %d\n", ret); return ret);

  // 5. 获取输出的值,将device侧内存上的结果拷贝至host侧,需要根据具体API的接口定义修改
  auto size = GetShapeSize(outShape);
  std::vector<float> resultData(size, 0);
  ret = aclrtMemcpy(resultData.data(), resultData.size() * sizeof(resultData[0]), outDeviceAddr, size * sizeof(float),
                    ACL_MEMCPY_DEVICE_TO_HOST);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("copy result from device to host failed. ERROR: %d\n", ret); return ret);
  for (int64_t i = 0; i < size; i++) {
    LOG_PRINT("result[%ld] is: %f\n", i, resultData[i]);
  }

  // 6. 释放aclTensor和aclScalar,需要根据具体API的接口定义修改
  aclDestroyTensor(self);
  aclDestroyTensor(mat1);
  aclDestroyTensor(mat2);
  aclDestroyTensor(out);
  aclDestroyScalar(alpha);
  aclDestroyScalar(beta);
  return 0;
}