aclnnLeTensor&aclnnInplaceLeTensor
支持的产品型号
- Atlas 推理系列产品。
- Atlas 训练系列产品。
- Atlas A2训练系列产品/Atlas 800I A2推理产品。
接口原型
aclnnLeTensor和aclnnInplaceLeTensor实现相同的功能,使用区别如下,请根据自身实际场景选择合适的算子。
- aclnnLeTensor:需新建一个输出张量对象存储计算结果。
- aclnnInplaceLeTensor: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。
每个算子分为两段式接口,必须先调用“aclnnLeTensorGetWorkspaceSize”或者“aclnnInplaceLeTensorGetWorkspaceSize”接口获取计算所需workspace大小以及包含了算子计算流程的执行器,再调用“aclnnLeTensor”或者“aclnnInplaceLeTensor”接口执行计算。
aclnnStatus aclnnLeTensorGetWorkspaceSize(const aclTensor *self, const aclTensor *other, aclTensor *out, uint64_t *workspaceSize, aclOpExecutor **executor)
aclnnStatus aclnnLeTensor(void *workspace, uint64_t workspaceSize, aclOpExecutor *executor, aclrtStream stream)
aclnnStatus aclnnInplaceLeTensorGetWorkspaceSize(aclTensor *selfRef, const aclTensor *other, uint64_t *workspaceSize, aclOpExecutor **executor)
aclnnStatus aclnnInplaceLeTensor(void *workspace, uint64_t workspaceSize, aclOpExecutor *executor, aclrtStream stream)
功能描述
- 算子功能:计算self中的元素值是否小于等于other的值,将self每个元素与other值的比较结果写入out中
- 计算公式:
aclnnLeTensorGetWorkspaceSize
参数说明:
- self(const aclTensor*, 计算输入):公式中的输入self,数据类型支持INT8、UINT8、INT16、INT32、INT64、FLOAT16、BFLOAT16(仅Atlas A2训练系列产品/Atlas 800I A2推理产品支持)、UINT16、FLOAT、DOUBLE、BOOL。且数据类型需要与other满足数据类型推导规则(参见互推导关系),shape需要与other满足broadcast关系。支持非连续的Tensor,数据格式支持ND。
- other(const aclTensor*, 计算输入):公式中的输入other,数据类型支持INT8、UINT8、INT16、INT32、INT64、FLOAT16、BFLOAT16(仅Atlas A2训练系列产品/Atlas 800I A2推理产品支持)、UINT16、FLOAT、DOUBLE、BOOL。且数据类型需要与self满足数据类型推导规则(参见互推导关系)。shape需要与self的shape满足broadcast关系。支持非连续的Tensor,数据格式支持ND。
- out(aclTensor*, 计算输出):公式中的out,数据类型支持INT8、UINT8、INT16、INT32、INT64、FLOAT16、BFLOAT16(仅Atlas A2训练系列产品/Atlas 800I A2推理产品支持)、UINT16、FLOAT、DOUBLE、BOOL,且数据类型与self的数据类型需满足数据类型推导规则(参见互推导关系),shape需要是self与other broadcast之后的shape(./common/broadcast关系.md),支持非连续的Tensor,数据格式支持ND。
- workspaceSize(出参):返回需要在Device侧申请的workspace大小。
- executor(出参):返回op执行器,包含了算子计算流程。
返回值:
aclnnStatus:返回状态码,具体参见aclnn返回码。
161001 (ACLNN_ERR_PARAM_NULLPTR): 1. 传入的self、other、out是空指针。 161002 (ACLNN_ERR_PARAM_INVALID): 1. self或other的数据类型不在支持的范围之内。 2. self和other无法做数据类型推导。 3. 推导出的数据类型无法转换为指定输出out的类型。 4. self和other的shape无法做broadcast。
aclnnLeTensor
参数说明:
- workspace(void *, 入参):在Device侧申请的workspace内存地址。
- workspaceSize(uint64_t, 入参):在Device侧申请的workspace大小,由第一段接口aclnnLeTensorGetWorkspaceSize获取。
- executor(aclOpExecutor *, 入参):op执行器,包含了算子计算流程。
- stream(aclrtStream, 入参):指定执行任务的 AscendCL Stream流。
返回值:
aclnnStatus:返回状态码,具体参见aclnn返回码。
aclnnInplaceLeTensorGetWorkspaceSize
参数说明:
- selfRef(aclTensor*,计算输入|计算输出):输入输出tensor,即公式中的self与out。数据类型支持FLOAT、FLOAT16、INT32、INT64、INT16、INT8、UINT8、DOUBLE、UINT16、BOOL、BFLOAT16(仅Atlas A2训练系列产品/Atlas 800I A2推理产品支持),且数据类型需要与other满足数据类型推导规则(参见互推导关系),shape需要与other满足broadcast关系,且broadcast后的shape需要与selfef的shape一致。支持非连续的Tensor,数据格式支持ND。
- other(aclTensor*,计算输入):数据类型支持FLOAT、FLOAT16、INT32、INT64、INT16、INT8、UINT8、DOUBLE、UINT16、BOOL、BFLOAT16(仅Atlas A2训练系列产品/Atlas 800I A2推理产品支持),且数据类型需要与selfRef满足数据类型推导规则(参见互推导关系),shape需要与selfRef满足broadcast关系,且broadcast后的shape需要与selfRef的shape一致。支持非连续的Tensor,数据格式支持ND。
- workspaceSize(出参):返回需要在Device侧申请的workspace大小。
- executor(出参):返回op执行器,包含了算子计算流程。
返回值:
aclnnStatus:返回状态码,具体参见aclnn返回码。
161001(ACLNN_ERR_PARAM_NULLPTR): 1. 传入的selfRef、other是空指针。 161002(ACLNN_ERR_PARAM_INVALID): 1. selfRef或other的数据类型和数据格式不在支持的范围之内。 2. selfRef和other的dtype不满足数据类型推导规则。 3. 推导出的数据类型无法转换为指定输出selfRef的类型。 4. selfRef和other的shape无法做broadcast,或者broadcast后的shape与selfRef的shape不一致。
aclnnInplaceLeTensor
参数说明:
- workspace(void *, 入参):在Device侧申请的workspace内存地址。
- workspaceSize(uint64_t, 入参):在Device侧申请的workspace大小,由第一段接口aclnnInplaceLeTensorGetWorkspaceSize获取。
- executor(aclOpExecutor *, 入参):op执行器,包含了算子计算流程。
- stream(aclrtStream, 入参):指定执行任务的 AscendCL Stream流。
返回值:
aclnnStatus:返回状态码,具体参见aclnn返回码。
约束与限制
无。
调用示例
示例代码如下,仅供参考,具体编译和执行过程请参考编译与运行样例。
aclnnLeTensor示例代码:
#include <iostream>
#include <vector>
#include "acl/acl.h"
#include "aclnnop/aclnn_le_tensor.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, 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 = 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/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;
aclrtStream stream;
auto ret = Init(deviceId, &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> selfShape = {4, 2};
std::vector<int64_t> otherShape = {4, 2};
std::vector<int64_t> outShape = {4, 2};
void* selfDeviceAddr = nullptr;
void* otherDeviceAddr = nullptr;
void* outDeviceAddr = nullptr;
aclTensor* self = nullptr;
aclTensor* other = nullptr;
aclTensor* out = nullptr;
std::vector<float> selfHostData = {0, 1, 2, 3, 4, 5, 6, 7};
std::vector<float> otherHostData = {1, 1, 1, 2, 2, 2, 3, 3};
std::vector<float> outHostData = {0, 0, 0, 0, 0, 0, 0, 0};
// 创建self aclTensor
ret = CreateAclTensor(selfHostData, selfShape, &selfDeviceAddr, aclDataType::ACL_FLOAT, &self);
CHECK_RET(ret == ACL_SUCCESS, return ret);
// 创建other aclTensor
ret = CreateAclTensor(otherHostData, otherShape, &otherDeviceAddr, aclDataType::ACL_FLOAT, &other);
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);
// 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aclnnLeTensor第一段接口
ret = aclnnLeTensorGetWorkspaceSize(self, other, out, &workspaceSize, &executor);
CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnLeTensorGetWorkspaceSize 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aclnnLeTensor第二段接口
ret = aclnnLeTensor(workspaceAddr, workspaceSize, executor, stream);
CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnLeTensor 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(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: %f\n", i, resultData[i]);
}
// 6. 释放aclTensor,需要根据具体API的接口定义修改
aclDestroyTensor(self);
aclDestroyTensor(other);
aclDestroyTensor(out);
// 7. 释放device资源,需要根据具体API的接口定义修改
aclrtFree(selfDeviceAddr);
aclrtFree(otherDeviceAddr);
aclrtFree(outDeviceAddr);
if (workspaceSize > 0) {
aclrtFree(workspaceAddr);
}
aclrtDestroyStream(stream);
aclrtResetDevice(deviceId);
aclFinalize();
return 0;
}
aclnnInplaceLeTensor示例代码:
#include <iostream>
#include <vector>
#include "acl/acl.h"
#include "aclnnop/aclnn_le_tensor.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, 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 = 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侧引擎
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() {
int32_t deviceId = 0;
aclrtStream stream;
auto ret = Init(deviceId, &stream);
CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("Init acl failed. ERROR: %d\n", ret); return ret);
std::vector<int64_t> selfShape = {4, 2};
std::vector<int64_t> otherShape = {4, 2};
void* selfDeviceAddr = nullptr;
void* otherDeviceAddr = nullptr;
aclTensor* self = nullptr;
aclTensor* other = nullptr;
std::vector<float> selfHostData = {0, 1, 2, 3, 4, 5, 6, 7};
std::vector<int> otherHostData = {1, 1, 1, 1, 0, 0, 0, 0};
ret = CreateAclTensor(selfHostData, selfShape, &selfDeviceAddr, aclDataType::ACL_FLOAT, &self);
CHECK_RET(ret == ACL_SUCCESS, return ret);
ret = CreateAclTensor(otherHostData, otherShape, &otherDeviceAddr, aclDataType::ACL_INT32, &other);
CHECK_RET(ret == ACL_SUCCESS, return ret);
uint64_t workspaceSize = 0;
aclOpExecutor* executor;
ret = aclnnInplaceLeTensorGetWorkspaceSize(self, other, &workspaceSize, &executor);
CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnInplaceLeTensorGetWorkspaceSize failed. ERROR: %d\n", ret); return ret);
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);
}
ret = aclnnInplaceLeTensor(workspaceAddr, workspaceSize, executor, stream);
CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnInplaceLeTensor failed. ERROR: %d\n", ret); return ret);
ret = aclrtSynchronizeStream(stream);
CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtSynchronizeStream failed. ERROR: %d\n", ret); return ret);
auto size = GetShapeSize(selfShape);
std::vector<float> resultData(size, 0);
ret = aclrtMemcpy(resultData.data(), resultData.size() * sizeof(resultData[0]), selfDeviceAddr, 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]);
}
// 释放aclTensor,需要根据具体API的接口定义修改
aclDestroyTensor(self);
aclDestroyTensor(other);
// 释放device资源,需要根据具体API的接口定义修改
aclrtFree(selfDeviceAddr);
aclrtFree(otherDeviceAddr);
if (workspaceSize > 0) {
aclrtFree(workspaceAddr);
}
aclrtDestroyStream(stream);
aclrtResetDevice(deviceId);
aclFinalize();
return 0;
}