ArmNN
 26.07
TosaRefBackend Class Reference

#include <TosaRefBackend.hpp>

Inheritance diagram for TosaRefBackend:
[legend]
Collaboration diagram for TosaRefBackend:
[legend]

Public Member Functions

 TosaRefBackend ()=default
 
 ~TosaRefBackend ()=default
 
const BackendIdGetId () const override
 
IBackendInternal::IMemoryManagerUniquePtr CreateMemoryManager () const override
 
IBackendInternal::IWorkloadFactoryPtr CreateWorkloadFactory (const IBackendInternal::IMemoryManagerSharedPtr &memoryManager=nullptr) const override
 
IBackendInternal::IWorkloadFactoryPtr CreateWorkloadFactory (class TensorHandleFactoryRegistry &tensorHandleFactoryRegistry) const override
 
IBackendInternal::IBackendContextPtr CreateBackendContext (const IRuntime::CreationOptions &) const override
 Create the runtime context of the backend. More...
 
IBackendInternal::IBackendProfilingContextPtr CreateBackendProfilingContext (const IRuntime::CreationOptions &creationOptions, IBackendProfilingPtr &backendProfiling) override
 Create context specifically used for profiling interaction from backends. More...
 
IBackendInternal::ILayerSupportSharedPtr GetLayerSupport () const override
 
OptimizationViews OptimizeSubgraphView (const SubgraphView &subgraph, const ModelOptions &modelOptions) const override
 
std::vector< ITensorHandleFactory::FactoryIdGetHandleFactoryPreferences () const override
 (Optional) Returns a vector of supported TensorHandleFactory ids in preference order. More...
 
void RegisterTensorHandleFactories (class TensorHandleFactoryRegistry &registry) override
 (Optional) Register TensorHandleFactories Either this method or CreateMemoryManager() and IWorkloadFactory::CreateTensor() IWorkloadFactory::CreateSubtensor() methods must be implemented. More...
 
std::unique_ptr< ICustomAllocatorGetDefaultAllocator () const override
 Returns the default memory allocator for the backend. More...
 
BackendCapabilities GetCapabilities () const override
 Returns a BackendCapability if the backend lists the capability The BackendCapability must then be inspected to check whether or not that BackendCapability is supported Otherwise returns an EmptyOptional if the BackendCapability is unlisted. More...
 
- Public Member Functions inherited from IBackendInternal
 ~IBackendInternal () override=default
 Allow backends created by the factory function to be destroyed through IBackendInternal. More...
 
virtual IWorkloadFactoryPtr CreateWorkloadFactory (const IMemoryManagerSharedPtr &memoryManager, const ModelOptions &modelOptions) const
 
virtual IWorkloadFactoryPtr CreateWorkloadFactory (class TensorHandleFactoryRegistry &tensorHandleFactoryRegistry, const ModelOptions &modelOptions) const
 
virtual IWorkloadFactoryPtr CreateWorkloadFactory (class TensorHandleFactoryRegistry &tensorHandleFactoryRegistry, const ModelOptions &modelOptions, MemorySourceFlags inputFlags, MemorySourceFlags outputFlags) const
 
virtual IBackendSpecificModelContextPtr CreateBackendSpecificModelContext (const ModelOptions &modelOptions) const
 
virtual ILayerSupportSharedPtr GetLayerSupport (const ModelOptions &modelOptions) const
 
virtual OptimizationViews OptimizeSubgraphView (const SubgraphView &subgraph) const
 
bool SupportsTensorAllocatorAPI () const
 
ITensorHandleFactory::FactoryId GetBackwardCompatibleFavoriteHandleFactory ()
 
virtual void RegisterTensorHandleFactories (class TensorHandleFactoryRegistry &registry, MemorySourceFlags inputFlags, MemorySourceFlags outputFlags)
 (Optional) Register TensorHandleFactories Either this method or CreateMemoryManager() and IWorkloadFactory::CreateTensor() IWorkloadFactory::CreateSubtensor() methods must be implemented. More...
 
virtual bool UseCustomMemoryAllocator (std::shared_ptr< ICustomAllocator > allocator, armnn::Optional< std::string & > errMsg)
 Signals the backend to use a custom memory allocator provided by the user. More...
 
virtual unsigned int GetNumberOfCacheFiles () const
 Returns the number of files cached if backend supports caching. More...
 

Static Public Member Functions

static const BackendIdGetIdStatic ()
 
- Static Public Member Functions inherited from IBackendInternal
static constexpr BackendVersion GetApiVersion ()
 Returns the version of the Backend API. More...
 

Additional Inherited Members

- Public Types inherited from IBackendInternal
using IWorkloadFactoryPtr = std::unique_ptr< IWorkloadFactory >
 
using IBackendContextPtr = std::unique_ptr< IBackendContext >
 
using IBackendProfilingContextPtr = std::shared_ptr< arm::pipe::IBackendProfilingContext >
 This is the bridge between backend and backend profiling we'll keep it in the backend namespace. More...
 
using IBackendProfilingPtr = std::unique_ptr< arm::pipe::IBackendProfiling >
 
using ILayerSupportSharedPtr = std::shared_ptr< ILayerSupport >
 
using IBackendSpecificModelContextPtr = std::shared_ptr< IBackendModelContext >
 
using IMemoryManagerUniquePtr = std::unique_ptr< IMemoryManager >
 
using IMemoryManagerSharedPtr = std::shared_ptr< IMemoryManager >
 
- Protected Member Functions inherited from IBackendInternal
 IBackendInternal ()=default
 Creation must be done through a specific backend interface. More...
 
- Protected Member Functions inherited from IBackend
 IBackend ()
 
virtual ~IBackend ()
 

Detailed Description

Definition at line 12 of file TosaRefBackend.hpp.

Constructor & Destructor Documentation

◆ TosaRefBackend()

TosaRefBackend ( )
default

◆ ~TosaRefBackend()

~TosaRefBackend ( )
default

Member Function Documentation

◆ CreateBackendContext()

IBackendInternal::IBackendContextPtr CreateBackendContext ( const IRuntime::CreationOptions ) const
overridevirtual

Create the runtime context of the backend.

Implementations may return a default-constructed IBackendContextPtr if no context is needed at runtime. Implementations must throw BackendUnavailableException if the backend cannot be used (for example, necessary accelerator hardware is not present). The default implementation always returns a default-constructed pointer.

Reimplemented from IBackendInternal.

Definition at line 60 of file TosaRefBackend.cpp.

61 {
62  return IBackendContextPtr{};
63 }
std::unique_ptr< IBackendContext > IBackendContextPtr

◆ CreateBackendProfilingContext()

IBackendInternal::IBackendProfilingContextPtr CreateBackendProfilingContext ( const IRuntime::CreationOptions creationOptions,
IBackendProfilingPtr backendProfiling 
)
overridevirtual

Create context specifically used for profiling interaction from backends.

Reimplemented from IBackendInternal.

Definition at line 65 of file TosaRefBackend.cpp.

67 {
69 }
std::shared_ptr< arm::pipe::IBackendProfilingContext > IBackendProfilingContextPtr
This is the bridge between backend and backend profiling we'll keep it in the backend namespace.

◆ CreateMemoryManager()

IBackendInternal::IMemoryManagerUniquePtr CreateMemoryManager ( ) const
overridevirtual

Reimplemented from IBackendInternal.

Definition at line 71 of file TosaRefBackend.cpp.

72 {
73  return std::make_unique<TosaRefMemoryManager>();
74 }

◆ CreateWorkloadFactory() [1/2]

IBackendInternal::IWorkloadFactoryPtr CreateWorkloadFactory ( class TensorHandleFactoryRegistry tensorHandleFactoryRegistry) const
overridevirtual

Reimplemented from IBackendInternal.

Definition at line 44 of file TosaRefBackend.cpp.

46 {
47  auto memoryManager = std::make_shared<TosaRefMemoryManager>();
48 
49  tensorHandleFactoryRegistry.RegisterMemoryManager(memoryManager);
50 
51  auto factory = std::make_unique<TosaRefTensorHandleFactory>(memoryManager);
52  // Register copy and import factory pair
53  tensorHandleFactoryRegistry.RegisterCopyAndImportFactoryPair(factory->GetId(), factory->GetId());
54  // Register the factory
55  tensorHandleFactoryRegistry.RegisterFactory(std::move(factory));
56 
57  return std::make_unique<TosaRefWorkloadFactory>(PolymorphicPointerDowncast<TosaRefMemoryManager>(memoryManager));
58 }

References TensorHandleFactoryRegistry::RegisterCopyAndImportFactoryPair(), TensorHandleFactoryRegistry::RegisterFactory(), and TensorHandleFactoryRegistry::RegisterMemoryManager().

◆ CreateWorkloadFactory() [2/2]

IBackendInternal::IWorkloadFactoryPtr CreateWorkloadFactory ( const IBackendInternal::IMemoryManagerSharedPtr memoryManager = nullptr) const
overridevirtual

Implements IBackendInternal.

Definition at line 38 of file TosaRefBackend.cpp.

40 {
41  return std::make_unique<TosaRefWorkloadFactory>(PolymorphicPointerDowncast<TosaRefMemoryManager>(memoryManager));
42 }

◆ GetCapabilities()

BackendCapabilities GetCapabilities ( ) const
overridevirtual

Returns a BackendCapability if the backend lists the capability The BackendCapability must then be inspected to check whether or not that BackendCapability is supported Otherwise returns an EmptyOptional if the BackendCapability is unlisted.

Reimplemented from IBackendInternal.

Definition at line 213 of file TosaRefBackend.cpp.

214 {
215  return BackendCapabilities ("TosaRef",
216  {
217  {"NonConstWeights", true},
218  {"ConstantTensorsAsInputs", true}
219  });
220 }
BackendOptions BackendCapabilities

◆ GetDefaultAllocator()

std::unique_ptr< ICustomAllocator > GetDefaultAllocator ( ) const
overridevirtual

Returns the default memory allocator for the backend.

Returns
- Returns unique pointer to the Default Allocator of the Backend

Reimplemented from IBackendInternal.

Definition at line 208 of file TosaRefBackend.cpp.

209 {
210  return std::make_unique<DefaultAllocator>();
211 }

◆ GetHandleFactoryPreferences()

std::vector< ITensorHandleFactory::FactoryId > GetHandleFactoryPreferences ( ) const
overridevirtual

(Optional) Returns a vector of supported TensorHandleFactory ids in preference order.

Reimplemented from IBackendInternal.

Definition at line 189 of file TosaRefBackend.cpp.

190 {
191  return std::vector<ITensorHandleFactory::FactoryId> { TosaRefTensorHandleFactory::GetIdStatic() };
192 }

References TosaRefTensorHandleFactory::GetIdStatic().

◆ GetId()

const BackendId& GetId ( ) const
inlineoverridevirtual

Implements IBackend.

Definition at line 19 of file TosaRefBackend.hpp.

20  {
21  return GetIdStatic();
22  }
static const BackendId & GetIdStatic()

References TosaRefBackend::GetIdStatic().

Referenced by TosaRefBackend::OptimizeSubgraphView().

◆ GetIdStatic()

const BackendId & GetIdStatic ( )
static

Definition at line 32 of file TosaRefBackend.cpp.

33 {
34  static const BackendId s_Id{TosaRefBackendId()};
35  return s_Id;
36 }
constexpr const char * TosaRefBackendId()

References armnn::TosaRefBackendId().

Referenced by TosaRefBackend::GetId().

◆ GetLayerSupport()

IBackendInternal::ILayerSupportSharedPtr GetLayerSupport ( ) const
overridevirtual

Implements IBackendInternal.

Definition at line 76 of file TosaRefBackend.cpp.

77 {
78  static ILayerSupportSharedPtr layerSupport{new TosaRefLayerSupport};
79  return layerSupport;
80 }
std::shared_ptr< ILayerSupport > ILayerSupportSharedPtr

◆ OptimizeSubgraphView()

OptimizationViews OptimizeSubgraphView ( const SubgraphView subgraph,
const ModelOptions modelOptions 
) const
overridevirtual

Reimplemented from IBackendInternal.

Definition at line 82 of file TosaRefBackend.cpp.

84 {
85  OptimizationViews optimizationViews(modelOptions);
86 
87  auto handler = std::make_unique<TosaSerializationHandler>();
88 
89  std::vector<std::string> graphInputs;
90  std::vector<std::string> graphOutputs;
91  std::vector<TosaSerializationOperator*> operators;
92  std::vector<TosaSerializationTensor*> tensors;
94 
95  // These sets are created to check the duplication of tensor and input
96  std::unordered_set<std::string> graphInputsSet;
97  std::unordered_set<std::string> uniqueTensorNamesSet;
98 
99  auto it = subgraph.begin();
100  while (it != subgraph.end())
101  {
102  Layer& base = *(PolymorphicDowncast<Layer*>(*it));
103  ++it;
104 
105  if (base.GetType() == LayerType::Input ||
106  base.GetType() == LayerType::Output)
107  {
108  continue;
109  }
110 
111  TosaSerializationBasicBlock* mappings = GetTosaMappingFromLayer(&base);
112 
113  // Loop through inputs to see if there are any graph inputs, if so save them.
114  // If it's an input to the graph "input" can be found in the string.
115  for (const std::string& blockInputName : mappings->GetInputs())
116  {
117  if ((blockInputName.find("input") != std::string::npos) && !graphInputsSet.count(blockInputName))
118  {
119  graphInputs.push_back(blockInputName);
120  graphInputsSet.insert(blockInputName);
121  }
122  }
123 
124  // Loop through outputs to see if there are any graph outputs, if so save them.
125  // If it's an output to the graph "output" can be found in the string.
126  for (const std::string& blockOutputName : mappings->GetOutputs())
127  {
128  if (blockOutputName.find("output") != std::string::npos)
129  {
130  graphOutputs.push_back(blockOutputName);
131  }
132  }
133 
134  sequencer.Add(&base, mappings);
135  }
136 
137  for (auto & pair : sequencer.Finish())
138  {
139  auto blockOperators = pair.block->GetOperators();
140  operators.insert(operators.end(), blockOperators.begin(), blockOperators.end());
141 
142  auto blockTensors = pair.block->GetTensors();
143 
144  // Checking for duplicate tensors, to handle special cases where same input is
145  // added to different layer in the architecture.
146  for (const auto& tensor : blockTensors)
147  {
148  const std::string& name = tensor->GetName();
149 
150  // If the input tensor is already present, we don't need to add to the list
151  if (!uniqueTensorNamesSet.count(name))
152  {
153  uniqueTensorNamesSet.insert(name);
154  tensors.push_back(tensor);
155  }
156  }
157  }
158 
159  // Add all mappings to main block.
160  auto* block = new TosaSerializationBasicBlock("main", "main", operators, tensors, graphInputs, graphOutputs);
161 
162  std::vector<TosaSerializationBasicBlock*> blocks;
163  blocks.emplace_back(block);
164 
165  // Add blocks to the main region.
166  auto* region = new TosaSerializationRegion("main", blocks);
167  handler->GetRegions().emplace_back(region);
168 
169  auto compiledBlob =
170  std::make_unique<PreCompiledObjectPtr>(handler.release(), DeleteAsType<TosaSerializationHandler>);
171 
172  IConnectableLayer* preCompiledLayer = optimizationViews.GetINetwork()->AddPrecompiledLayer(
174  std::move(*compiledBlob),
176  "TOSA_Pre_Compiled_Layer");
177 
178  // Copy the output tensor infos from sub-graph
179  for (unsigned int i = 0; i < subgraph.GetNumOutputSlots(); i++)
180  {
181  preCompiledLayer->GetOutputSlot(i).SetTensorInfo(subgraph.GetIOutputSlot(i)->GetTensorInfo());
182  }
183 
184  optimizationViews.AddSubstitution({ std::move(subgraph), SubgraphView(preCompiledLayer) });
185  return optimizationViews;
186 }
TosaSerializationBasicBlock * GetTosaMappingFromLayer(const Layer *layer)
Interface for a layer that is connectable to other layers via InputSlots and OutputSlots.
Definition: INetwork.hpp:81
virtual const IOutputSlot & GetOutputSlot(unsigned int index) const =0
Get the const output slot handle by slot index.
virtual void SetTensorInfo(const TensorInfo &tensorInfo)=0
virtual const TensorInfo & GetTensorInfo() const =0
LayerType GetType() const override
Returns the armnn::LayerType of this layer.
Definition: Layer.hpp:286
std::list< Pair > & Finish()
void Add(LayerT *layer, BlockT *block)
The SubgraphView class represents a subgraph of a Graph.
IConnectableLayerIterator begin()
unsigned int GetNumOutputSlots() const
unsigned int GetNumInputSlots() const
const IOutputSlot * GetIOutputSlot(unsigned int index) const
IConnectableLayerIterator end()
const BackendId & GetId() const override
A PreCompiledDescriptor for the PreCompiledLayer.

References OpBlockSequencer< LayerT, BlockT >::Add(), INetwork::AddPrecompiledLayer(), OptimizationViews::AddSubstitution(), SubgraphView::begin(), SubgraphView::end(), OpBlockSequencer< LayerT, BlockT >::Finish(), TosaRefBackend::GetId(), OptimizationViews::GetINetwork(), SubgraphView::GetIOutputSlot(), SubgraphView::GetNumInputSlots(), SubgraphView::GetNumOutputSlots(), IConnectableLayer::GetOutputSlot(), IOutputSlot::GetTensorInfo(), GetTosaMappingFromLayer(), Layer::GetType(), armnn::Input, armnn::Output, and IOutputSlot::SetTensorInfo().

◆ RegisterTensorHandleFactories()

void RegisterTensorHandleFactories ( class TensorHandleFactoryRegistry )
overridevirtual

(Optional) Register TensorHandleFactories Either this method or CreateMemoryManager() and IWorkloadFactory::CreateTensor() IWorkloadFactory::CreateSubtensor() methods must be implemented.

Reimplemented from IBackendInternal.

Definition at line 194 of file TosaRefBackend.cpp.

195 {
196  auto memoryManager = std::make_shared<TosaRefMemoryManager>();
197 
198  registry.RegisterMemoryManager(memoryManager);
199 
200  auto factory = std::make_unique<TosaRefTensorHandleFactory>(memoryManager);
201 
202  // Register copy and import factory pair
203  registry.RegisterCopyAndImportFactoryPair(factory->GetId(), factory->GetId());
204  // Register the factory
205  registry.RegisterFactory(std::move(factory));
206 }

References TensorHandleFactoryRegistry::RegisterCopyAndImportFactoryPair(), TensorHandleFactoryRegistry::RegisterFactory(), and TensorHandleFactoryRegistry::RegisterMemoryManager().


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