Compute Library
 21.02
NEReorgLayerKernel Class Reference

Interface for the kernel to perform tensor re-organization. More...

#include <NEReorgLayerKernel.h>

Collaboration diagram for NEReorgLayerKernel:
[legend]

Public Member Functions

const char * name () const override
 Name of the kernel. More...
 
 NEReorgLayerKernel ()
 Default constructor. More...
 
 NEReorgLayerKernel (const NEReorgLayerKernel &)=delete
 Prevent instances of this class from being copied (As this class contains pointers) More...
 
NEReorgLayerKerneloperator= (const NEReorgLayerKernel &)=delete
 Prevent instances of this class from being copied (As this class contains pointers) More...
 
 NEReorgLayerKernel (NEReorgLayerKernel &&)=default
 Default Move Constructor. More...
 
NEReorgLayerKerneloperator= (NEReorgLayerKernel &&)=default
 Default move assignment operator. More...
 
 ~NEReorgLayerKernel ()=default
 Default destructor. More...
 
void configure (const ITensor *input, ITensor *output, int32_t stride)
 Set the input and output of the kernel. More...
 
void run (const Window &window, const ThreadInfo &info) override
 Execute the kernel on the passed window. More...
 
- Public Member Functions inherited from ICPPKernel
virtual ~ICPPKernel ()=default
 Default destructor. More...
 
virtual void run_nd (const Window &window, const ThreadInfo &info, const Window &thread_locator)
 legacy compatibility layer for implemantions which do not support thread_locator In these cases we simply narrow the interface down the legacy version More...
 
virtual void run_op (ITensorPack &tensors, const Window &window, const ThreadInfo &info)
 Execute the kernel on the passed window. More...
 
- Public Member Functions inherited from IKernel
 IKernel ()
 Constructor. More...
 
virtual ~IKernel ()=default
 Destructor. More...
 
virtual bool is_parallelisable () const
 Indicates whether or not the kernel is parallelisable. More...
 
virtual BorderSize border_size () const
 The size of the border for that kernel. More...
 
const Windowwindow () const
 The maximum window the kernel can be executed on. More...
 

Static Public Member Functions

static Status validate (const ITensorInfo *input, const ITensorInfo *output, int32_t stride)
 Static function to check if given info will lead to a valid configuration of cpu::kernels::CpuReshapeKernel. More...
 

Detailed Description

Interface for the kernel to perform tensor re-organization.

Definition at line 35 of file NEReorgLayerKernel.h.

Constructor & Destructor Documentation

◆ NEReorgLayerKernel() [1/3]

Default constructor.

Definition at line 68 of file NEReorgLayerKernel.cpp.

Referenced by NEReorgLayerKernel::name().

69  : _input(nullptr), _output(nullptr), _stride(1)
70 {
71 }

◆ NEReorgLayerKernel() [2/3]

NEReorgLayerKernel ( const NEReorgLayerKernel )
delete

Prevent instances of this class from being copied (As this class contains pointers)

◆ NEReorgLayerKernel() [3/3]

Default Move Constructor.

◆ ~NEReorgLayerKernel()

~NEReorgLayerKernel ( )
default

Default destructor.

Referenced by NEReorgLayerKernel::name().

Member Function Documentation

◆ configure()

void configure ( const ITensor input,
ITensor output,
int32_t  stride 
)

Set the input and output of the kernel.

Parameters
[in]inputSource tensor. Data type supported: All
[out]outputDestination tensor. Data type supported: Same as input
[in]strideStride to be used during data re-organization. It defines the spatial distance between 2 consecutive pixels in the x and y direction

Definition at line 73 of file NEReorgLayerKernel.cpp.

References ARM_COMPUTE_ERROR_ON_NULLPTR, ARM_COMPUTE_ERROR_THROW_ON, arm_compute::auto_init_if_empty(), arm_compute::calculate_max_window(), ICloneable< T >::clone(), arm_compute::misc::shape_calculator::compute_reorg_output_shape(), ITensor::info(), arm_compute::test::validation::input, arm_compute::test::validation::output_shape, ITensorInfo::set_valid_region(), ITensorInfo::tensor_shape(), and arm_compute::validate_arguments().

Referenced by NEReorgLayerKernel::name().

74 {
76 
77  // Output auto inizialitation if not yet initialized
78  const TensorShape output_shape = misc::shape_calculator::compute_reorg_output_shape(*input->info(), stride);
79  auto_init_if_empty(*output->info(), input->info()->clone()->set_tensor_shape(output_shape));
80 
81  // Perform validation step
82  ARM_COMPUTE_ERROR_THROW_ON(validate_arguments(input->info(), output->info(), stride));
83 
84  _input = input;
85  _output = output;
86  _stride = stride;
87 
88  // The NEReorgLayerKernel doesn't need padding so update_window_and_padding() can be skipped
89  output->info()->set_valid_region(ValidRegion(Coordinates(), output->info()->tensor_shape()));
90 
91  // Configure kernel window
92  Window win = calculate_max_window(*output->info(), Steps());
93 
94  ICPPKernel::configure(win);
95 }
Window calculate_max_window(const ValidRegion &valid_region, const Steps &steps, bool skip_border, BorderSize border_size)
#define ARM_COMPUTE_ERROR_THROW_ON(status)
Definition: Error.h:455
TensorShape compute_reorg_output_shape(const ITensorInfo &input, int32_t stride)
Calculate the output shape of the reorg layer given a stride.
bool auto_init_if_empty(ITensorInfo &info, const TensorShape &shape, int num_channels, DataType data_type, QuantizationInfo quantization_info=QuantizationInfo())
Auto initialize the tensor info (shape, number of channels and data type) if the current assignment i...
Status validate_arguments(const ITensorInfo *input, const ITensorInfo *bias, const ITensorInfo *output, const GEMMLowpOutputStageInfo *output_stage)
#define ARM_COMPUTE_ERROR_ON_NULLPTR(...)
Definition: Validate.h:161

◆ name()

◆ operator=() [1/2]

NEReorgLayerKernel& operator= ( const NEReorgLayerKernel )
delete

Prevent instances of this class from being copied (As this class contains pointers)

Referenced by NEReorgLayerKernel::name().

◆ operator=() [2/2]

NEReorgLayerKernel& operator= ( NEReorgLayerKernel &&  )
default

Default move assignment operator.

◆ run()

void run ( const Window window,
const ThreadInfo info 
)
overridevirtual

Execute the kernel on the passed window.

Warning
If is_parallelisable() returns false then the passed window must be equal to window()
Note
The window has to be a region within the window returned by the window() method
The width of the window has to be a multiple of num_elems_processed_per_iteration().
Parameters
[in]windowRegion on which to execute the kernel. (Must be a region of the window returned by window())
[in]infoInfo about executing thread and CPU.

Reimplemented from ICPPKernel.

Definition at line 103 of file NEReorgLayerKernel.cpp.

References ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW, ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL, ARM_COMPUTE_UNUSED, ITensor::buffer(), arm_compute::CHANNEL, Window::collapse_if_possible(), arm_compute::test::validation::data_layout, ITensorInfo::data_layout(), ITensorInfo::element_size(), arm_compute::execute_window_loop(), arm_compute::get_data_layout_dimension_index(), arm_compute::HEIGHT, ITensor::info(), offset(), ITensorInfo::offset_element_in_bytes(), Iterator::ptr(), Dimensions< T >::set(), ITensorInfo::tensor_shape(), arm_compute::test::validation::w, arm_compute::WIDTH, and IKernel::window().

Referenced by NEReorgLayerKernel::name().

104 {
108 
109  const DataLayout data_layout = _input->info()->data_layout();
110  const size_t idx_w = get_data_layout_dimension_index(data_layout, DataLayoutDimension::WIDTH);
111  const size_t idx_h = get_data_layout_dimension_index(data_layout, DataLayoutDimension::HEIGHT);
112  const size_t idx_c = get_data_layout_dimension_index(data_layout, DataLayoutDimension::CHANNEL);
113 
114  const unsigned int stride = _stride;
115  const unsigned int out_c = _output->info()->tensor_shape()[idx_c] / (stride * stride);
116  const uint8_t *in_ptr = _input->buffer();
117 
118  // Collapse
119  Window collapsed_window = window.collapse_if_possible(window, 4);
120 
121  // Create Iterator
122  Iterator out(_output, collapsed_window);
123 
124  // Perform reorg
125  execute_window_loop(collapsed_window, [&](const Coordinates & id)
126  {
127  // Get spatial coords and channels
128  const unsigned int w = id[idx_w];
129  const unsigned int h = id[idx_h];
130  const unsigned int c = id[idx_c];
131 
132  // Calculate mapping
133  const unsigned int offset = c / out_c;
134  Coordinates map_coords = id;
135  map_coords.set(idx_w, w * stride + offset % stride);
136  map_coords.set(idx_h, h * stride + offset / stride);
137  map_coords.set(idx_c, c % out_c);
138 
139  // Perform mapping
140  std::memcpy(out.ptr(), in_ptr + _input->info()->offset_element_in_bytes(map_coords), _input->info()->element_size());
141  },
142  out);
143 }
__global uchar * offset(const Image *img, int x, int y)
Get the pointer position of a Image.
Definition: helpers.h:846
SimpleTensor< float > w
Definition: DFT.cpp:156
const Window & window() const
The maximum window the kernel can be executed on.
Definition: IKernel.cpp:28
virtual int32_t offset_element_in_bytes(const Coordinates &pos) const =0
The offset in bytes from the beginning of the memory allocation to access the element at position (x...
const DataLayout data_layout
Definition: Im2Col.cpp:151
#define ARM_COMPUTE_UNUSED(...)
To avoid unused variables warnings.
Definition: Error.h:152
Window collapse_if_possible(const Window &full_window, size_t first, size_t last, bool *has_collapsed=nullptr) const
Collapse the dimensions between first and last if possible.
Definition: Window.inl:68
virtual const TensorShape & tensor_shape() const =0
Size for each dimension of the tensor.
virtual uint8_t * buffer() const =0
Interface to be implemented by the child class to return a pointer to CPU memory. ...
virtual ITensorInfo * info() const =0
Interface to be implemented by the child class to return the tensor&#39;s metadata.
virtual size_t element_size() const =0
Element size in bytes calculated as data_size() * num_channels()
#define ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL(k)
Definition: Validate.h:941
ScaleKernelInfo info(interpolation_policy, default_border_mode, PixelValue(), sampling_policy, false)
void execute_window_loop(const Window &w, L &&lambda_function, Ts &&... iterators)
Iterate through the passed window, automatically adjusting the iterators and calling the lambda_funct...
Definition: Helpers.inl:77
size_t get_data_layout_dimension_index(const DataLayout data_layout, const DataLayoutDimension data_layout_dimension)
Get the index of the given dimension.
Definition: Helpers.inl:193
DataLayout
[DataLayout enum definition]
Definition: Types.h:120
#define ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW(f, s)
Definition: Validate.h:205
virtual DataLayout data_layout() const =0
Get the data layout of the tensor.

◆ validate()

Status validate ( const ITensorInfo input,
const ITensorInfo output,
int32_t  stride 
)
static

Static function to check if given info will lead to a valid configuration of cpu::kernels::CpuReshapeKernel.

Parameters
[in]inputSource tensor info. Data type supported: All
[in]outputDestination tensor info. Data type supported: Same as input
[in]strideStride to be used during data re-organization It defines the spatial distance between 2 consecutive pixels in the x and y direction
Returns
a status

Definition at line 97 of file NEReorgLayerKernel.cpp.

References ARM_COMPUTE_RETURN_ON_ERROR, and arm_compute::validate_arguments().

Referenced by NEReorgLayerKernel::name(), and NEReorgLayer::validate().

98 {
100  return Status{};
101 }
#define ARM_COMPUTE_RETURN_ON_ERROR(status)
Checks if a status contains an error and returns it.
Definition: Error.h:204
Status validate_arguments(const ITensorInfo *input, const ITensorInfo *bias, const ITensorInfo *output, const GEMMLowpOutputStageInfo *output_stage)

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