aclnnGridSampler2DBackward
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 aclnnGridSampler2DBackwardGetWorkspaceSize(const aclTensor *gradOutput, const aclTensor *input, const aclTensor *grid, int64_t interpolationMode, int64_t paddingMode, bool alignCorners, aclTensor *inputGrad, aclTensor *gridGrad, uint64_t *workspaceSize, aclOpExecutor **executor)
- 第二段接口：aclnnStatus aclnnGridSampler2DBackward(void *workspace, uint64_t workspaceSize, aclOpExecutor *executor, const aclrtStream stream)
功能描述
算子功能：算子GridSampler2D（aclnnGridSampler2D）的反向计算。
aclnnGridSampler2DBackwardGetWorkspaceSize
- 接口定义：
aclnnStatus aclnnGridSampler2DBackwardGetWorkspaceSize(const aclTensor *gradOutput, const aclTensor *input, const aclTensor *grid, int64_t interpolationMode, int64_t paddingMode, bool alignCorners, aclTensor *inputGrad, aclTensor *gridGrad, uint64_t *workspaceSize, aclOpExecutor **executor)
- 参数说明：
- gradOutput：Device侧的aclTensor，数据类型支持FLOAT16、FLOAT、DOUBLE，需要和input类型保持一致，支持非连续的Tensor，数据格式支持ND。
- input：Device侧的aclTensor，数据类型支持FLOAT16、FLOAT、DOUBLE，支持非连续的Tensor，数据格式支持ND。
- grid：Device侧的aclTensor，数据类型支持FLOAT16、FLOAT、DOUBLE，需要和input类型保持一致，支持非连续的Tensor，数据格式支持ND。
- interpolationMode：Host侧的int64_t，表示插值模式，分别是0：bilinear（双线性插值）、1：nearest（最邻近插值）。
- paddingMode：Host侧的string，表示填充模式。当（x, y）取值超过输入特征图采样范围，返回一个特定值，有0：zeros、1：border、2：reflection三种模式。
- alignCorners：Host侧的bool，表示设定特征图坐标与特征值的对应方式。设为True时，特征值位于像素中心。
- inputGrad：Device侧的aclTensor，数据类型支持FLOAT16、FLOAT、DOUBLE，需要和input类型保持一致，支持非连续的Tensor，数据格式支持ND。
- gridGrad：Device侧的aclTensor，数据类型支持FLOAT16、FLOAT、DOUBLE，需要和grid类型保持一致，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gradOutput、input、grid、inputGrad、gridGrad是空指针。
- 返回161002（ACLNN_ERR_PARAM_INVALID）：
- gradOutput、input、grid、inputGrad、gridGrad的数据类型或数据格式不在支持的范围内或数据类型不一致。
- interpolationMode和paddingMode的值不在支持范围内。
- gradOutput、input、grid的维度关系不匹配。
- input最后两维为空。
aclnnGridSampler2DBackward
- 接口定义：
aclnnStatus aclnnGridSampler2DBackward(void *workspace, uint64_t workspaceSize, aclOpExecutor *executor, const 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aclnnGridSampler2DBackwardGetWorkspaceSize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_grid_sampler2d_backward.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
int64_t interpolationMode = 0;
int64_t paddingMode = 0;
bool alignCorners = false;
aclBoolArray* outputMask = nullptr;
std::vector<int64_t> gradOutputShape = {1, 1, 3, 3};
std::vector<int64_t> inputShape = {1, 1, 5, 8};
std::vector<int64_t> gridShape = {1, 3, 3, 2};
std::vector<int64_t> inputGradShape = {1, 1, 5, 8};
std::vector<int64_t> gridGradShape = {1, 3, 3, 2};
void* gradOutputDeviceAddr = nullptr;
void* inputDeviceAddr = nullptr;
void* gridDeviceAddr = nullptr;
void* inputGradDeviceAddr = nullptr;
void* gridGradDeviceAddr = nullptr;
aclTensor* gradOutput = nullptr;
aclTensor* input = nullptr;
aclTensor* grid = nullptr;
aclTensor* inputGrad = nullptr;
aclTensor* gridGrad = nullptr;
std::vector<float> gradOutputHostData = {1, 1, 1, 1, 1, 1, 1, 1, 1};
std::vector<float> inputHostData = {1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23,
24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40};
std::vector<float> gridHostData = {-1, -1, 0, -1, 1, -1, -1, 0, 0, 0, 1, 0, -1, 1, 0, 1, 1, 1};
std::vector<float> inputGradHostData = {0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0};
std::vector<float> gridGradHostData = {0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0};
bool maskValue[2] = {true, true};
outputMask = aclCreateBoolArray(&(maskValue[0]), 2);
// 创建gradOutput aclTensor
ret = CreateAclTensor(gradOutputHostData, gradOutputShape, &gradOutputDeviceAddr, aclDataType::ACL_FLOAT, &gradOutput);
CHECK_RET(ret == ACL_SUCCESS, return ret);
// 创建input aclTensor
ret = CreateAclTensor(inputHostData, inputShape, &inputDeviceAddr, aclDataType::ACL_FLOAT, &input);
CHECK_RET(ret == ACL_SUCCESS, return ret);
// 创建grid aclTensor
ret = CreateAclTensor(gridHostData, gridShape, &gridDeviceAddr, aclDataType::ACL_FLOAT, &grid);
CHECK_RET(ret == ACL_SUCCESS, return ret);
// 创建inputGrad aclTensor
ret = CreateAclTensor(inputGradHostData, inputGradShape, &inputGradDeviceAddr, aclDataType::ACL_FLOAT, &inputGrad);
CHECK_RET(ret == ACL_SUCCESS, return ret);
// 创建gridGrad aclTensor
ret = CreateAclTensor(gridGradHostData, gridGradShape, &gridGradDeviceAddr, aclDataType::ACL_FLOAT, &gridGrad);
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aclnnGridSampler2DBackward第一段接口
ret = aclnnGridSampler2DBackwardGetWorkspaceSize(gradOutput, input, grid, interpolationMode, paddingMode,
alignCorners, outputMask, inputGrad, gridGrad,
&workspaceSize, &executor);
CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnGridSampler2DBackwardGetWorkspaceSize 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aclnnGridSampler2DBackward第二段接口
ret = aclnnGridSampler2DBackward(workspaceAddr, workspaceSize, executor, stream);
CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnGridSampler2DBackward 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 inputGradSize = GetShapeSize(inputGradShape);
std::vector<float> inputGradResultData(inputGradSize, 0);
ret = aclrtMemcpy(inputGradResultData.data(), inputGradResultData.size() * sizeof(inputGradResultData[0]),
inputGradDeviceAddr, inputGradSize * sizeof(inputGradResultData[0]), ACL_MEMCPY_DEVICE_TO_HOST);
CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("copy inputGradResultData from device to host failed. ERROR: %d\n", ret);
return ret);
for (int64_t i = 0; i < inputGradSize; i++) {
LOG_PRINT("inputGradResultData[%ld] is: %f\n", i, inputGradResultData[i]);
}
auto gridGradSize = GetShapeSize(gridGradShape);
std::vector<float> gridGradResultData(gridGradSize, 0);
ret = aclrtMemcpy(gridGradResultData.data(), gridGradResultData.size() * sizeof(gridGradResultData[0]),
gridGradDeviceAddr, gridGradSize * sizeof(gridGradResultData[0]), ACL_MEMCPY_DEVICE_TO_HOST);
CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("copy gridGradResultData from device to host failed. ERROR: %d\n", ret);
return ret);
for (int64_t i = 0; i < gridGradSize; i++) {
LOG_PRINT("gridGradResultData[%ld] is: %f\n", i, gridGradResultData[i]);
}
// 6. 释放aclTensor和aclBoolArray，需要根据具体API的接口定义修改
aclDestroyTensor(gradOutput);
aclDestroyTensor(input);
aclDestroyTensor(grid);
aclDestroyTensor(inputGrad);
aclDestroyTensor(gridGrad);
aclDestroyBoolArray(outputMask);
return 0;
}
父主题： NN类算子接口