arm_compute::CLMultiplyScaleFactorKernel Class Reference

#include <CLMultiplyScaleFactorKernel.h>

Collaboration diagram for arm_compute::CLMultiplyScaleFactorKernel:

Public Member Functions
	CLMultiplyScaleFactorKernel ()
	CLMultiplyScaleFactorKernel (const CLMultiplyScaleFactorKernel &)=delete
CLMultiplyScaleFactorKernel &	operator= (const CLMultiplyScaleFactorKernel &)=delete
	CLMultiplyScaleFactorKernel (CLMultiplyScaleFactorKernel &&)=default
CLMultiplyScaleFactorKernel &	operator= (CLMultiplyScaleFactorKernel &&)=default
	~CLMultiplyScaleFactorKernel ()=default
void	configure (const ICLTensor *input, const ICLTensor *scale_factor, ICLTensor *output, float multiplier=1.f)
void	run (const Window &window, cl::CommandQueue &queue) override
	Enqueue the OpenCL kernel to process the given window on the passed OpenCL command queue.

Static Public Member Functions
static Status	validate (const ITensorInfo *input, const ITensorInfo *scale_factor, const ITensorInfo *output)

Detailed Description

Interface to multiply scale factor kernel.

Definition at line 51 of file CLMultiplyScaleFactorKernel.h.

Constructor & Destructor Documentation

CLMultiplyScaleFactorKernel() [1/3]

CLMultiplyScaleFactorKernel::CLMultiplyScaleFactorKernel

(

)

Default constructor

Definition at line 105 of file CLMultiplyScaleFactorKernel.cpp.

  : _input(nullptr), _scale_factor(nullptr), _output(nullptr), _multiplier(1.f)
{
}

CLMultiplyScaleFactorKernel() [2/3]

arm_compute::CLMultiplyScaleFactorKernel::CLMultiplyScaleFactorKernel ( const CLMultiplyScaleFactorKernel &  )

delete

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

CLMultiplyScaleFactorKernel() [3/3]

arm_compute::CLMultiplyScaleFactorKernel::CLMultiplyScaleFactorKernel ( CLMultiplyScaleFactorKernel &&  )

default

Default Move Constructor.

~CLMultiplyScaleFactorKernel()

arm_compute::CLMultiplyScaleFactorKernel::~CLMultiplyScaleFactorKernel ( )

default

Default destructor

References validate().

Member Function Documentation

configure()

void CLMultiplyScaleFactorKernel::configure	(	const ICLTensor *	input,
		const ICLTensor *	scale_factor,
		ICLTensor *	output,
		float	multiplier = 1.f
	)

Set input, output tensors.

Parameters

	[in/out]	input Source tensor. Data type supported: S32.
[in]	scale_factor	Scale tensor. Data type supported: F16/F32.
[out]	output	Destination tensor. Data type supported: Same as scale_factor.
[in]	multiplier	Additional scale value.

Definition at line 110 of file CLMultiplyScaleFactorKernel.cpp.

{
  ARM_COMPUTE_ERROR_ON_NULLPTR(input, output);
  ARM_COMPUTE_ERROR_THROW_ON(
    validate_arguments(input->info(), scale_factor->info(), output->info()));
 
  _input = input;
  _scale_factor = scale_factor;
  _output = output;
  _multiplier = multiplier;
 
  const int vec_size_x = 16 / output->info()->element_size();
  const int output_width_x = output->info()->tensor_shape().x();
  const bool multi_access_x = (output_width_x / vec_size_x > 0);
 
  // Create and update the window (if needed)
  Window win = calculate_max_window(*output->info());
  if (multi_access_x)
  {
    win.set(
      Window::DimX,
      Window::Dimension(win.x().start(), ceil_to_multiple(win.x().end(), vec_size_x), vec_size_x));
  }
  ICLKernel::configure_internal(win);
 
  // Create kernel
  CLBuildOptions build_opts;
  build_opts.add_option("-DVEC_SIZE=" + support::cpp11::to_string(vec_size_x));
  build_opts.add_option("-DDATA_TYPE=" + get_cl_type_from_data_type(output->info()->data_type()));
  build_opts.add_option_if(
    multi_access_x, "-DLAST_ACCESSED_X=" +
                      support::cpp11::to_string(std::max<int>(output_width_x - vec_size_x, 0)));
 
  _kernel = static_cast<cl::Kernel>(
    CLKernelLibraryEx::get().create_kernel("multiply_scale_factor", build_opts.options()));
}

References arm_compute::CLKernelLibraryEx::create_kernel(), and arm_compute::CLKernelLibraryEx::get().

Referenced by arm_compute::CLFullyConnectedHybridLayer::configure().

operator=() [1/2]

CLMultiplyScaleFactorKernel & arm_compute::CLMultiplyScaleFactorKernel::operator= ( CLMultiplyScaleFactorKernel &&  )

default

Default move assignment operator

operator=() [2/2]

CLMultiplyScaleFactorKernel & arm_compute::CLMultiplyScaleFactorKernel::operator= ( const CLMultiplyScaleFactorKernel &  )

delete

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

run()

void CLMultiplyScaleFactorKernel::run ( const Window &  window, cl::CommandQueue &  queue  )

override

Enqueue the OpenCL kernel to process the given window on the passed OpenCL command queue.

Note

The queue is not flushed by this method, and therefore the kernel will not have been executed by the time this method returns.

Parameters

[in]	window	Region on which to execute the kernel. (Must be a valid region of the window returned by window()).
[in,out]	queue	Command queue on which to enqueue the kernel.

Returns

N/A

N/A

Definition at line 158 of file CLMultiplyScaleFactorKernel.cpp.

{
  ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL(this);
  ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW(ICLKernel::window(), window);
 
  Window window_collapsed = window.collapse_if_possible(ICLKernel::window(), Window::DimZ);
  Window slice = window_collapsed.first_slice_window_2D();
 
  // Set scale_factor window
  Window win_scale = calculate_max_window(*_scale_factor->info(), Steps());
 
  do
  {
    unsigned int idx = 0;
    add_2D_tensor_argument(idx, _input, slice);
    add_1D_tensor_argument(idx, _scale_factor, win_scale);
    add_2D_tensor_argument(idx, _output, slice);
    _kernel.setArg<float>(idx++, _multiplier);
    enqueue(queue, *this, slice, lws_hint());
  } while (window_collapsed.slide_window_slice_2D(slice));
}

Referenced by package.infer.session::inference().

validate()

Status CLMultiplyScaleFactorKernel::validate ( const ITensorInfo *  input, const ITensorInfo *  scale_factor, const ITensorInfo *  output  )

static

Static function to check if given info will lead to a valid configuration of CLMultiplyScaleFactorKernel

Parameters

[in]	input	Input tensor info. Data types supported: S32.
[in]	scale_factor	Scale tensor. Data type supported: F16/F32.
[in]	output	Output tensor info. Data types supported: Same as scale_factor.
[in]	multiplier	Additional scale value.

Returns

a status

Definition at line 148 of file CLMultiplyScaleFactorKernel.cpp.

{
  ARM_COMPUTE_RETURN_ON_ERROR(validate_arguments(input, scale_factor, output));
  ARM_COMPUTE_RETURN_ON_ERROR(
    std::get<0>(validate_and_configure_window(input->clone().get(), output->clone().get())));
  return Status{};
}

Referenced by arm_compute::CLFullyConnectedHybridLayer::validate(), and ~CLMultiplyScaleFactorKernel().

---

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

• compute/ARMComputeEx/arm_compute/core/CL/kernels/CLMultiplyScaleFactorKernel.h
• compute/ARMComputeEx/src/core/CL/kernels/CLMultiplyScaleFactorKernel.cpp