aclnnInplaceRemainderTensorTensor
接口原型
每个算子有两段接口,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。两段式接口如下:
- 第一段接口:aclnnStatus aclnnInplaceRemainderTensorTensorGetWorkspaceSize(aclTensor* selfRef, const aclTensor *other, uint64_t *workspaceSize, aclOpExecutor **executor)
- 第二段接口:aclnnStatus aclnnInplaceRemainderTensorTensor(void *workspace, uint64_t workspaceSize, aclOpExecutor *executor, aclrtStream stream)
功能描述
- 算子功能:将selfRef(张量)broadcast成和other(张量)一样的shape后,将其每个元素都转换为除以other对应元素后得到的余数。该结果与除数other同符号,并且该结果的绝对值是小于other的绝对值。
- 计算公式:
- 示例:
selfRef = tensor([[-1, -2], [-3, -4]]).type(torch.float16) other = tensor([-3, -3]).type(torch.int64) selfRef = remainder(selfRef, other) # selfRef的值 # tensor([[-1., -2.], # [-0., -1.]], dtype=torch.float16) # 首先是将other broadcast成和selfRef一致的shape,成为 [[-3, -3], [-3, -3]],然后再进行计算。 # 对于元素selfRef中的-3来说,计算结果为 (-3) % (-3) = 0 # 可以看到,最终结果0的绝对值小于原来的-3的绝对值。
aclnnInplaceRemainderTensorTensorGetWorkspaceSize
- 接口定义:
aclnnStatus aclnnInplaceRemainderTensorTensorGetWorkspaceSize(aclTensor* selfRef, const aclTensor *other, uint64_t *workspaceSize, aclOpExecutor **executor)
- 参数说明:
- selfRef:Device侧的aclTensor,数据类型与other的数据类型需满足数据类型推导规则,且推导出的数据类型支持INT32、INT64、FLOAT16、FLOAT、DOUBLE,且需要是推导之后可转换为selfRef的数据类型。shape需要与other满足broadcast关系,且shape与最终broadcast后的shape一致。支持非连续的Tensor,数据格式支持ND。
- other:Device侧的aclTensor, 数据类型与selfRef的数据类型需满足数据类型推导规则,且推导出的数据类型支持INT32、INT64、FLOAT16、FLOAT、DOUBLE。shape需要与selfRef满足broadcast关系,支持非连续的Tensor,数据格式支持ND。
- workspaceSize:返回用户需要在Device侧申请的workspace大小。
- executor:返回op执行器,包含了算子计算流程。
- 返回值:
返回aclnnStatus状态码,具体参见aclnn返回码。
第一段接口完成入参校验,出现以下场景时报错:
- 返回161001(ACLNN_ERR_PARAM_NULLPTR):传入的selfRef、other是空指针。
- 返回161002(ACLNN_ERR_PARAM_INVALID):
- selfRef和other无法做数据类型推导。
- selfRef和other推导出的数据类型不属于支持的数据类型。
- selfRef和other推导出的数据类型无法转换为selfRef的类型。
- selfRef和other的shape无法做broadcast。
- selfRef和other broadcast以后的shape与selfRef的shape不一致。
- selfRef、other的维度数大于8维。
aclnnInplaceRemainderTensorTensor
- 接口定义:
aclnnStatus aclnnInplaceRemainderTensorTensor(void *workspace, uint64_t workspaceSize, aclOpExecutor *executor, aclrtStream stream)
- 参数说明:
- workspace:在Device侧申请的workspace内存起址。
- workspaceSize:在Device侧申请的workspace大小,由第一段接口aclnnInplaceRemainderTensorTensorGetWorkspaceSize获取。
- executor:op执行器,包含了算子计算流程。
- stream:指定执行任务的AscendCL stream流。
- 返回值:
返回aclnnStatus状态码,具体参见aclnn返回码。
调用示例
#include <iostream>
#include <vector>
#include "acl/acl.h"
#include "aclnnop/aclnn_remainder.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 shapeSize = 1;
for (auto i : shape) {
shapeSize *= i;
}
return shapeSize;
}
int Init(int32_t deviceId, aclrtContext* context, aclrtStream* stream) {
// 固定写法,AscendCL初始化
auto ret = aclInit(nullptr);
CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclInit failed. ERROR: %d\n", ret); return ret);
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);
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AscendCL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_RET(ret == ACL_SUCCESS, 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> selfRefShape = {3, 3};
std::vector<int64_t> otherShape = {3, 3};
void* selfRefDeviceAddr = nullptr;
void* otherDeviceAddr = nullptr;
aclTensor* selfRef = nullptr;
aclTensor* other = nullptr;
std::vector<int64_t> selfRefHostData = {1, 2, 3, 4, 5, 6, 7, 8, 9};
std::vector<int64_t> otherHostData = {0, 1, 2, 3, 4, 5, 6, 7, 8};
// 创建selfRef aclTensor
selfRef = CreateAclTensor(selfRefHostData, selfRefShape, &selfRefDeviceAddr, aclDataType::ACL_INT64, &selfRef);
CHECK_RET(selfRef != nullptr, return ret);
// 创建other aclTensor
ret = CreateAclTensor(otherHostData, otherShape, &otherDeviceAddr, aclDataType::ACL_INT64, &other);
CHECK_RET(ret == ACL_SUCCESS, return ret);
// 3. 调用CANN算子库API,需要修改为具体的API名称
uint64_t workspaceSize = 0;
aclOpExecutor* executor;
// 调用aclnnInplaceRemainderTensorTensor第一段接口
ret = aclnnInplaceRemainderTensorTensorGetWorkspaceSize(selfRef, other, &workspaceSize, &executor);
CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnInplaceRemainderTensorTensorGetWorkspaceSize 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aclnnInplaceRemainderTensorTensor第二段接口
ret = aclnnInplaceRemainderTensorTensor(workspaceAddr, workspaceSize, executor, stream);
CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnInplaceRemainderTensorTensor 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(selfRefShape);
std::vector<int64_t> resultData(size, 0);
ret = aclrtMemcpy(resultData.data(), resultData.size() * sizeof(resultData[0]), selfRefDeviceAddr,
size * sizeof(resultData[0]), 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: %ld\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(selfRef);
aclDestroyTensor(other);
return 0;
}
父主题: NN类算子接口
