Plugin class for user-implemented layers. More...

Inheritance diagram for nvinfer1::IPluginV2IOExt:

[legend]

Collaboration diagram for nvinfer1::IPluginV2IOExt:

[legend]

Public Member Functions
virtual void	configurePlugin (const PluginTensorDesc in, int32_t nbInput, const PluginTensorDesc out, int32_t nbOutput)=0
	Configure the layer. More...

virtual bool	supportsFormatCombination (int32_t pos, const PluginTensorDesc *inOut, int32_t nbInputs, int32_t nbOutputs) const =0
	Return true if plugin supports the format and datatype for the input/output indexed by pos. More...

virtual nvinfer1::DataType	getOutputDataType (int32_t index, const nvinfer1::DataType *inputTypes, int32_t nbInputs) const =0
	Return the DataType of the plugin output at the requested index. More...

virtual bool	isOutputBroadcastAcrossBatch (int32_t outputIndex, const bool *inputIsBroadcasted, int32_t nbInputs) const =0
	Return true if output tensor is broadcast across a batch. More...

virtual bool	canBroadcastInputAcrossBatch (int32_t inputIndex) const =0
	Return true if plugin can use input that is broadcast across batch without replication. More...

virtual void	configurePlugin (const Dims inputDims, int32_t nbInputs, const Dims outputDims, int32_t nbOutputs, const DataType inputTypes, const DataType outputTypes, const bool inputIsBroadcast, const bool outputIsBroadcast, PluginFormat floatFormat, int32_t maxBatchSize)=0
	Configure the layer with input and output data types. More...

virtual void	attachToContext (cudnnContext , cublasContext , IGpuAllocator *)
	Attach the plugin object to an execution context and grant the plugin the access to some context resource. More...

virtual void	detachFromContext ()
	Detach the plugin object from its execution context. More...

virtual IPluginV2Ext *	clone () const =0
	Clone the plugin object. More...

virtual const char *	getPluginType () const =0
	Return the plugin type. More...

virtual const char *	getPluginVersion () const =0
	Return the plugin version. More...

virtual int32_t	getNbOutputs () const =0
	Get the number of outputs from the layer. More...

virtual Dims	getOutputDimensions (int32_t index, const Dims *inputs, int32_t nbInputDims)=0
	Get the dimension of an output tensor. More...

virtual bool	supportsFormat (DataType type, PluginFormat format) const =0
	Check format support. More...

virtual int32_t	initialize ()=0
	Initialize the layer for execution. More...

virtual void	terminate ()=0
	Release resources acquired during plugin layer initialization. More...

virtual size_t	getWorkspaceSize (int32_t maxBatchSize) const =0
	Find the workspace size required by the layer. More...

virtual int32_t	enqueue (int32_t batchSize, const void const inputs, void *outputs, void workspace, cudaStream_t stream)=0
	Execute the layer. More...

virtual size_t	getSerializationSize () const =0
	Find the size of the serialization buffer required. More...

virtual void	serialize (void *buffer) const =0
	Serialize the layer. More...

virtual void	destroy ()=0
	Destroy the plugin object. More...

virtual void	setPluginNamespace (const char *pluginNamespace)=0
	Set the namespace that this plugin object belongs to. More...

virtual const char *	getPluginNamespace () const =0
	Return the namespace of the plugin object. More...

Protected Member Functions
	__attribute__ ((deprecated)) int32_t getTensorRTVersion() const
	Return the API version with which this plugin was built. More...

	__attribute__ ((deprecated)) void configureWithFormat(const Dims *
	Deprecated interface inheriting from base class. More...

int32_t	getTensorRTVersion () const
	Return the API version with which this plugin was built. More...

void	configureWithFormat (const Dims , int32_t, const Dims , int32_t, DataType, PluginFormat, int32_t)
	Derived classes should not implement this. More...

Protected Attributes
	int32_t

const Dims	int32_t

const Dims	DataType

const Dims	PluginFormat

Detailed Description

Plugin class for user-implemented layers.

Plugins are a mechanism for applications to implement custom layers. This interface provides additional capabilities to the IPluginV2Ext interface by extending different I/O data types and tensor formats.

See also: IPluginV2Ext

Member Function Documentation

◆ configurePlugin() [1/2]

virtual void nvinfer1::IPluginV2IOExt::configurePlugin	(	const PluginTensorDesc *	in,
		int32_t	nbInput,
		const PluginTensorDesc *	out,
		int32_t	nbOutput
	)

pure virtual

Configure the layer.

This function is called by the builder prior to initialize(). It provides an opportunity for the layer to make algorithm choices on the basis of I/O PluginTensorDesc and the maximum batch size.

Parameters

in	The input tensors attributes that are used for configuration.
nbInput	Number of input tensors.
out	The output tensors attributes that are used for configuration.
nbOutput	Number of output tensors.

◆ supportsFormatCombination()

virtual bool nvinfer1::IPluginV2IOExt::supportsFormatCombination	(	int32_t	pos,
		const PluginTensorDesc *	inOut,
		int32_t	nbInputs,
		int32_t	nbOutputs
	)		const

pure virtual

Return true if plugin supports the format and datatype for the input/output indexed by pos.

For this method inputs are numbered 0..(nbInputs-1) and outputs are numbered nbInputs..(nbInputs+nbOutputs-1). Using this numbering, pos is an index into InOut, where 0 <= pos < nbInputs+nbOutputs-1.

TensorRT invokes this method to ask if the input/output indexed by pos supports the format/datatype specified by inOut[pos].format and inOut[pos].type. The override should return true if that format/datatype at inOut[pos] are supported by the plugin. If support is conditional on other input/output formats/datatypes, the plugin can make its result conditional on the formats/datatypes in inOut[0..pos-1], which will be set to values that the plugin supports. The override should not inspect inOut[pos+1..nbInputs+nbOutputs-1], which will have invalid values. In other words, the decision for pos must be based on inOut[0..pos] only.

Some examples:

A definition for a plugin that supports only FP16 NCHW:

  return inOut.format[pos] == TensorFormat::kLINEAR && inOut.type[pos] == DataType::kHALF;

A definition for a plugin that supports only FP16 NCHW for its two inputs, and FP32 NCHW for its single output:

  return inOut.format[pos] == TensorFormat::kLINEAR && (inOut.type[pos] == pos < 2 ?  DataType::kHALF :
  DataType::kFLOAT);

A definition for a "polymorphic" plugin with two inputs and one output that supports any format or type, but the inputs and output must have the same format and type:
```
  return pos == 0 || (inOut.format[pos] == inOut.format[0] && inOut.type[pos] == inOut.type[0]);
```

Warning: TensorRT will stop asking for formats once it finds kFORMAT_COMBINATION_LIMIT on combinations.

◆ attribute() [1/2]

nvinfer1::IPluginV2IOExt::__attribute__ ( (deprecated) ) const

inlineprotected

Return the API version with which this plugin was built.

The upper byte is reserved by TensorRT and is used to differentiate this from IPlguinV2 and IPluginV2Ext.

Do not override this method as it is used by the TensorRT library to maintain backwards-compatibility with plugins.

Deprecated:: Deprecated interface will be removed in TensorRT 8.0.

◆ attribute() [2/2]

nvinfer1::IPluginV2IOExt::__attribute__ ( (deprecated) ) const

protected

Deprecated interface inheriting from base class.

Derived classes should not implement this. In a C++11 API it would be override final.

Deprecated:: Deprecated interface will be removed in TensorRT 8.0.

◆ getOutputDataType()

virtual nvinfer1::DataType nvinfer1::IPluginV2Ext::getOutputDataType	(	int32_t	index,
		const nvinfer1::DataType *	inputTypes,
		int32_t	nbInputs
	)		const

pure virtualinherited

Return the DataType of the plugin output at the requested index.

The default behavior should be to return the type of the first input, or DataType::kFLOAT if the layer has no inputs. The returned data type must have a format that is supported by the plugin.

See also: supportsFormat()

Warning: DataType:kBOOL not supported.

◆ isOutputBroadcastAcrossBatch()

virtual bool nvinfer1::IPluginV2Ext::isOutputBroadcastAcrossBatch	(	int32_t	outputIndex,
		const bool *	inputIsBroadcasted,
		int32_t	nbInputs
	)		const

pure virtualinherited

Return true if output tensor is broadcast across a batch.

Parameters

outputIndex	The index of the output
inputIsBroadcasted	The ith element is true if the tensor for the ith input is broadcast across a batch.
nbInputs	The number of inputs

The values in inputIsBroadcasted refer to broadcasting at the semantic level, i.e. are unaffected by whether method canBroadcastInputAcrossBatch requests physical replication of the values.

◆ canBroadcastInputAcrossBatch()

virtual bool nvinfer1::IPluginV2Ext::canBroadcastInputAcrossBatch ( int32_t inputIndex ) const

pure virtualinherited

Return true if plugin can use input that is broadcast across batch without replication.

Parameters

inputIndex Index of input that could be broadcast.

For each input whose tensor is semantically broadcast across a batch, TensorRT calls this method before calling configurePlugin. If canBroadcastInputAcrossBatch returns true, TensorRT will not replicate the input tensor; i.e., there will be a single copy that the plugin should share across the batch. If it returns false, TensorRT will replicate the input tensor so that it appears like a non-broadcasted tensor.

This method is called only for inputs that can be broadcast.

◆ configurePlugin() [2/2]

virtual void nvinfer1::IPluginV2Ext::configurePlugin	(	const Dims *	inputDims,
		int32_t	nbInputs,
		const Dims *	outputDims,
		int32_t	nbOutputs,
		const DataType *	inputTypes,
		const DataType *	outputTypes,
		const bool *	inputIsBroadcast,
		const bool *	outputIsBroadcast,
		PluginFormat	floatFormat,
		int32_t	maxBatchSize
	)

pure virtualinherited

Configure the layer with input and output data types.

This function is called by the builder prior to initialize(). It provides an opportunity for the layer to make algorithm choices on the basis of its weights, dimensions, data types and maximum batch size.

Parameters

inputDims	The input tensor dimensions.
nbInputs	The number of inputs.
outputDims	The output tensor dimensions.
nbOutputs	The number of outputs.
inputTypes	The data types selected for the plugin inputs.
outputTypes	The data types selected for the plugin outputs.
inputIsBroadcast	True for each input that the plugin must broadcast across the batch.
outputIsBroadcast	True for each output that TensorRT will broadcast across the batch.
floatFormat	The format selected for the engine for the floating point inputs/outputs.
maxBatchSize	The maximum batch size.

The dimensions passed here do not include the outermost batch size (i.e. for 2-D image networks, they will be 3-dimensional CHW dimensions). When inputIsBroadcast or outputIsBroadcast is true, the outermost batch size for that input or output should be treated as if it is one. inputIsBroadcast[i] is true only if the input is semantically broadcast across the batch and canBroadcastInputAcrossBatch(i) returned true. outputIsBroadcast[i] is true only if isOutputBroadcastAcrossBatch(i) returned true.

Warning: for the floatFormat field, the values PluginFormat::kCHW4, PluginFormat::kCHW16, and PluginFormat::kCHW32 will not be passed in, this is to keep backward compatibility with TensorRT 5.x series. Use PluginV2IOExt or PluginV2DynamicExt for other PluginFormats.

◆ attachToContext()

virtual void nvinfer1::IPluginV2Ext::attachToContext	(	cudnnContext *	,
		cublasContext *	,
		IGpuAllocator *
	)

inlinevirtualinherited

Attach the plugin object to an execution context and grant the plugin the access to some context resource.

Parameters

cudnn	The cudnn context handle of the execution context
cublas	The cublas context handle of the execution context
allocator	The allocator used by the execution context

This function is called automatically for each plugin when a new execution context is created. If the plugin needs per-context resource, it can be allocated here. The plugin can also get context-owned CUDNN and CUBLAS context here.

◆ detachFromContext()

virtual void nvinfer1::IPluginV2Ext::detachFromContext ( )

inlinevirtualinherited

Detach the plugin object from its execution context.

This function is called automatically for each plugin when a execution context is destroyed. If the plugin owns per-context resource, it can be released here.

◆ clone()

virtual IPluginV2Ext* nvinfer1::IPluginV2Ext::clone ( ) const

pure virtualinherited

Clone the plugin object.

This copies over internal plugin parameters as well and returns a new plugin object with these parameters. If the source plugin is pre-configured with configurePlugin(), the returned object should also be pre-configured. The returned object should allow attachToContext() with a new execution context. Cloned plugin objects can share the same per-engine immutable resource (e.g. weights) with the source object (e.g. via ref-counting) to avoid duplication.

Implements nvinfer1::IPluginV2.

◆ getTensorRTVersion()

int32_t nvinfer1::IPluginV2Ext::getTensorRTVersion ( ) const

inlineprotectedvirtualinherited

Return the API version with which this plugin was built.

The upper byte reserved by TensorRT and is used to differentiate this from IPlguinV2.

Do not override this method as it is used by the TensorRT library to maintain backwards-compatibility with plugins.

Reimplemented from nvinfer1::IPluginV2.

◆ configureWithFormat()

void nvinfer1::IPluginV2Ext::configureWithFormat	(	const Dims *	,
		int32_t	,
		const Dims *	,
		int32_t	,
		DataType	,
		PluginFormat	,
		int32_t
	)

inlineprotectedvirtualinherited

Derived classes should not implement this.

In a C++11 API it would be override final.

Implements nvinfer1::IPluginV2.

◆ getPluginType()

virtual const char* nvinfer1::IPluginV2::getPluginType ( ) const

pure virtualinherited

Return the plugin type.

Should match the plugin name returned by the corresponding plugin creator

See also: IPluginCreator::getPluginName()

◆ getPluginVersion()

virtual const char* nvinfer1::IPluginV2::getPluginVersion ( ) const

pure virtualinherited

Return the plugin version.

Should match the plugin version returned by the corresponding plugin creator

See also: IPluginCreator::getPluginVersion()

◆ getNbOutputs()

virtual int32_t nvinfer1::IPluginV2::getNbOutputs ( ) const

pure virtualinherited

Get the number of outputs from the layer.

Returns: The number of outputs.

This function is called by the implementations of INetworkDefinition and IBuilder. In particular, it is called prior to any call to initialize().

Here is the caller graph for this function:

◆ getOutputDimensions()

virtual Dims nvinfer1::IPluginV2::getOutputDimensions	(	int32_t	index,
		const Dims *	inputs,
		int32_t	nbInputDims
	)

pure virtualinherited

Get the dimension of an output tensor.

Parameters

index	The index of the output tensor.
inputs	The input tensors.
nbInputDims	The number of input tensors.

This function is called by the implementations of INetworkDefinition and IBuilder. In particular, it is called prior to any call to initialize().

◆ supportsFormat()

virtual bool nvinfer1::IPluginV2::supportsFormat	(	DataType	type,
		PluginFormat	format
	)		const

pure virtualinherited

Check format support.

Parameters

type	DataType requested.
format	PluginFormat requested.

Returns: true if the plugin supports the type-format combination.

This function is called by the implementations of INetworkDefinition, IBuilder, and safe::ICudaEngine/ICudaEngine. In particular, it is called when creating an engine and when deserializing an engine.

Warning: for the format field, the values PluginFormat::kCHW4, PluginFormat::kCHW16, and PluginFormat::kCHW32 will not be passed in, this is to keep backward compatibility with TensorRT 5.x series. Use PluginV2IOExt or PluginV2DynamicExt for other PluginFormats.; DataType:kBOOL not supported.

◆ initialize()

virtual int32_t nvinfer1::IPluginV2::initialize ( )

pure virtualinherited

Initialize the layer for execution.

This is called when the engine is created.

Returns: 0 for success, else non-zero (which will cause engine termination).

◆ terminate()

virtual void nvinfer1::IPluginV2::terminate ( )

pure virtualinherited

Release resources acquired during plugin layer initialization.

This is called when the engine is destroyed.

See also: initialize()

◆ getWorkspaceSize()

virtual size_t nvinfer1::IPluginV2::getWorkspaceSize ( int32_t maxBatchSize ) const

pure virtualinherited

Find the workspace size required by the layer.

This function is called during engine startup, after initialize(). The workspace size returned should be sufficient for any batch size up to the maximum.

Returns: The workspace size.

◆ enqueue()

virtual int32_t nvinfer1::IPluginV2::enqueue	(	int32_t	batchSize,
		const void const	inputs,
		void **	outputs,
		void *	workspace,
		cudaStream_t	stream
	)

pure virtualinherited

Execute the layer.

Parameters

batchSize	The number of inputs in the batch.
inputs	The memory for the input tensors.
outputs	The memory for the output tensors.
workspace	Workspace for execution.
stream	The stream in which to execute the kernels.

Returns: 0 for success, else non-zero (which will cause engine termination).

◆ getSerializationSize()

virtual size_t nvinfer1::IPluginV2::getSerializationSize ( ) const

pure virtualinherited

Find the size of the serialization buffer required.

Returns: The size of the serialization buffer.

◆ serialize()

virtual void nvinfer1::IPluginV2::serialize ( void * buffer ) const

pure virtualinherited

Serialize the layer.

Parameters

buffer A pointer to a buffer to serialize data. Size of buffer must be equal to value returned by getSerializationSize.

See also: getSerializationSize()

◆ destroy()

virtual void nvinfer1::IPluginV2::destroy ( )

pure virtualinherited

Destroy the plugin object.

This will be called when the network, builder or engine is destroyed.

◆ setPluginNamespace()

virtual void nvinfer1::IPluginV2::setPluginNamespace ( const char * pluginNamespace )

pure virtualinherited

Set the namespace that this plugin object belongs to.

Ideally, all plugin objects from the same plugin library should have the same namespace.

Here is the caller graph for this function:

◆ getPluginNamespace()

virtual const char* nvinfer1::IPluginV2::getPluginNamespace ( ) const

pure virtualinherited

Return the namespace of the plugin object.

Member Data Documentation

◆ int32_t [1/2]

nvinfer1::IPluginV2IOExt::int32_t

protected

◆ int32_t [2/2]

const Dims nvinfer1::IPluginV2IOExt::int32_t

protected

◆ DataType

const Dims nvinfer1::IPluginV2IOExt::DataType

protected

◆ PluginFormat

const Dims nvinfer1::IPluginV2IOExt::PluginFormat

protected

The documentation for this class was generated from the following file:

NvInferRuntimeCommon.h

Public Member Functions

Protected Member Functions

Protected Attributes

Detailed Description

Member Function Documentation

◆ configurePlugin() [1/2]

◆ supportsFormatCombination()

◆ __attribute__() [1/2]

◆ __attribute__() [2/2]

◆ getOutputDataType()

◆ isOutputBroadcastAcrossBatch()

◆ canBroadcastInputAcrossBatch()

◆ configurePlugin() [2/2]

◆ attachToContext()

◆ detachFromContext()

◆ clone()

◆ getTensorRTVersion()

◆ configureWithFormat()

◆ getPluginType()

◆ getPluginVersion()

◆ getNbOutputs()

◆ getOutputDimensions()

◆ supportsFormat()

◆ initialize()

◆ terminate()

◆ getWorkspaceSize()

◆ enqueue()

◆ getSerializationSize()

◆ serialize()

◆ destroy()

◆ setPluginNamespace()

◆ getPluginNamespace()

Member Data Documentation

◆ int32_t [1/2]

◆ int32_t [2/2]

◆ DataType

◆ PluginFormat

◆ attribute() [1/2]

◆ attribute() [2/2]