wehub-resource-sync
132 lines
5.0 KiB
C++
132 lines
5.0 KiB
C++
/*
|
|
* SPDX-FileCopyrightText: Copyright (c) 1993-2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
|
|
* SPDX-License-Identifier: Apache-2.0
|
|
*
|
|
* Licensed under the Apache License, Version 2.0 (the "License");
|
|
* you may not use this file except in compliance with the License.
|
|
* You may obtain a copy of the License at
|
|
*
|
|
* http://www.apache.org/licenses/LICENSE-2.0
|
|
*
|
|
* Unless required by applicable law or agreed to in writing, software
|
|
* distributed under the License is distributed on an "AS IS" BASIS,
|
|
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
* See the License for the specific language governing permissions and
|
|
* limitations under the License.
|
|
*/
|
|
|
|
#ifndef TRT_HARDMAX_PLUGIN_H
|
|
#define TRT_HARDMAX_PLUGIN_H
|
|
|
|
#include "NvInferPlugin.h"
|
|
#include <cublas_v2.h>
|
|
#include <string>
|
|
#include <vector>
|
|
|
|
// This sample demonstrates how to implement a TensorRT plugin using the IPluginV3 interface.
|
|
|
|
class HardmaxPlugin final : public nvinfer1::IPluginV3,
|
|
public nvinfer1::IPluginV3OneCore,
|
|
public nvinfer1::IPluginV3OneBuild,
|
|
public nvinfer1::IPluginV3OneRuntime
|
|
{
|
|
public:
|
|
HardmaxPlugin() = delete;
|
|
HardmaxPlugin(int32_t axis);
|
|
HardmaxPlugin(HardmaxPlugin const& other);
|
|
~HardmaxPlugin() override;
|
|
|
|
// IPluginV3 methods
|
|
nvinfer1::IPluginCapability* getCapabilityInterface(nvinfer1::PluginCapabilityType type) noexcept override;
|
|
nvinfer1::IPluginV3* clone() noexcept override;
|
|
|
|
// IPluginV3OneCore methods
|
|
char const* getPluginName() const noexcept override;
|
|
char const* getPluginVersion() const noexcept override;
|
|
char const* getPluginNamespace() const noexcept override;
|
|
|
|
// IPluginV3OneBuild methods
|
|
int32_t getNbOutputs() const noexcept override;
|
|
|
|
int32_t configurePlugin(nvinfer1::DynamicPluginTensorDesc const* in, int32_t nbInputs,
|
|
nvinfer1::DynamicPluginTensorDesc const* out, int32_t nbOutputs) noexcept override;
|
|
|
|
bool supportsFormatCombination(int32_t pos, nvinfer1::DynamicPluginTensorDesc const* inOut, int32_t nbInputs,
|
|
int32_t nbOutputs) noexcept override;
|
|
|
|
int32_t getOutputDataTypes(nvinfer1::DataType* outputTypes, int32_t nbOutputs, nvinfer1::DataType const* inputTypes,
|
|
int32_t nbInputs) const noexcept override;
|
|
|
|
int32_t getOutputShapes(nvinfer1::DimsExprs const* inputs, int32_t nbInputs, nvinfer1::DimsExprs const* shapeInputs,
|
|
int32_t nbShapeInputs, nvinfer1::DimsExprs* outputs, int32_t nbOutputs,
|
|
nvinfer1::IExprBuilder& exprBuilder) noexcept override;
|
|
|
|
size_t getWorkspaceSize(nvinfer1::DynamicPluginTensorDesc const* inputs, int32_t nbInputs,
|
|
nvinfer1::DynamicPluginTensorDesc const* outputs, int32_t nbOutputs) const noexcept override;
|
|
|
|
// IPluginV3OneRuntime methods
|
|
int32_t enqueue(nvinfer1::PluginTensorDesc const* inputDesc, nvinfer1::PluginTensorDesc const* outputDesc,
|
|
void const* const* inputs, void* const* outputs, void* workspace, cudaStream_t stream) noexcept override;
|
|
|
|
int32_t onShapeChange(nvinfer1::PluginTensorDesc const* in, int32_t nbInputs, nvinfer1::PluginTensorDesc const* out,
|
|
int32_t nbOutputs) noexcept override;
|
|
|
|
nvinfer1::IPluginV3* attachToContext(nvinfer1::IPluginResourceContext* context) noexcept override;
|
|
|
|
nvinfer1::PluginFieldCollection const* getFieldsToSerialize() noexcept override;
|
|
|
|
void setPluginNamespace(char const* pluginNamespace) noexcept;
|
|
|
|
private:
|
|
std::string mNamespace;
|
|
|
|
// Number of elements in the axis along which hardmax is performed.
|
|
int32_t mAxisSize{0};
|
|
|
|
// Product of dimensions before and after mAxis.
|
|
// For example, if the input dimensions are [3, 4, 5, 6, 7] and mAxis = 2,
|
|
// then mDimProductOuter = 12 and mDimProductInner = 42.
|
|
int32_t mDimProductOuter{1};
|
|
int32_t mDimProductInner{1};
|
|
|
|
cublasHandle_t mCublas{nullptr};
|
|
|
|
// Attributes
|
|
// Axis along which to perform hardmax.
|
|
// Can be negative initially, but once configurePlugin() is called it will
|
|
// be converted to a positive axis.
|
|
int32_t mAxis{-1};
|
|
|
|
// Serialization helpers
|
|
std::vector<nvinfer1::PluginField> mDataToSerialize;
|
|
nvinfer1::PluginFieldCollection mFCToSerialize;
|
|
};
|
|
|
|
class HardmaxPluginCreator : public nvinfer1::IPluginCreatorV3One
|
|
{
|
|
public:
|
|
HardmaxPluginCreator();
|
|
|
|
~HardmaxPluginCreator() override = default;
|
|
|
|
char const* getPluginName() const noexcept override;
|
|
|
|
char const* getPluginVersion() const noexcept override;
|
|
|
|
nvinfer1::PluginFieldCollection const* getFieldNames() noexcept override;
|
|
|
|
nvinfer1::IPluginV3* createPlugin(
|
|
char const* name, nvinfer1::PluginFieldCollection const* fc, nvinfer1::TensorRTPhase phase) noexcept override;
|
|
|
|
void setPluginNamespace(char const* pluginNamespace) noexcept;
|
|
|
|
char const* getPluginNamespace() const noexcept override;
|
|
|
|
private:
|
|
nvinfer1::PluginFieldCollection mFC;
|
|
std::vector<nvinfer1::PluginField> mPluginAttributes;
|
|
std::string mNamespace;
|
|
};
|
|
|
|
#endif // TRT_HARDMAX_PLUGIN_H
|