Compute Library
 19.08
CLPixelWiseMultiplicationKernel Class Reference

Interface for the pixelwise multiplication kernel. More...

#include <CLPixelWiseMultiplicationKernel.h>

Collaboration diagram for CLPixelWiseMultiplicationKernel:
[legend]

Public Member Functions

 CLPixelWiseMultiplicationKernel ()
 Default constructor. More...
 
 CLPixelWiseMultiplicationKernel (const CLPixelWiseMultiplicationKernel &)=delete
 Prevent instances of this class from being copied (As this class contains pointers) More...
 
CLPixelWiseMultiplicationKerneloperator= (const CLPixelWiseMultiplicationKernel &)=delete
 Prevent instances of this class from being copied (As this class contains pointers) More...
 
 CLPixelWiseMultiplicationKernel (CLPixelWiseMultiplicationKernel &&)=default
 Allow instances of this class to be moved. More...
 
CLPixelWiseMultiplicationKerneloperator= (CLPixelWiseMultiplicationKernel &&)=default
 Allow instances of this class to be moved. More...
 
void configure (const ICLTensor *input1, const ICLTensor *input2, ICLTensor *output, float scale, ConvertPolicy overflow_policy, RoundingPolicy rounding_policy)
 Initialise the kernel's input, output and border mode. More...
 
void run (const Window &window, cl::CommandQueue &queue) override
 Enqueue the OpenCL kernel to process the given window on the passed OpenCL command queue. More...
 
BorderSize border_size () const override
 The size of the border for that kernel. More...
 
- Public Member Functions inherited from ICLKernel
 ICLKernel ()
 Constructor. More...
 
cl::Kernel & kernel ()
 Returns a reference to the OpenCL kernel of this object. More...
 
template<typename T >
void add_1D_array_argument (unsigned int &idx, const ICLArray< T > *array, const Strides &strides, unsigned int num_dimensions, const Window &window)
 Add the passed 1D array's parameters to the object's kernel's arguments starting from the index idx. More...
 
void add_1D_tensor_argument (unsigned int &idx, const ICLTensor *tensor, const Window &window)
 Add the passed 1D tensor's parameters to the object's kernel's arguments starting from the index idx. More...
 
void add_1D_tensor_argument_if (bool cond, unsigned int &idx, const ICLTensor *tensor, const Window &window)
 Add the passed 1D tensor's parameters to the object's kernel's arguments starting from the index idx if the condition is true. More...
 
void add_2D_tensor_argument (unsigned int &idx, const ICLTensor *tensor, const Window &window)
 Add the passed 2D tensor's parameters to the object's kernel's arguments starting from the index idx. More...
 
void add_2D_tensor_argument_if (bool cond, unsigned int &idx, const ICLTensor *tensor, const Window &window)
 Add the passed 2D tensor's parameters to the object's kernel's arguments starting from the index idx if the condition is true. More...
 
void add_3D_tensor_argument (unsigned int &idx, const ICLTensor *tensor, const Window &window)
 Add the passed 3D tensor's parameters to the object's kernel's arguments starting from the index idx. More...
 
void add_4D_tensor_argument (unsigned int &idx, const ICLTensor *tensor, const Window &window)
 Add the passed 4D tensor's parameters to the object's kernel's arguments starting from the index idx. More...
 
template<typename T >
void add_argument (unsigned int &idx, T value)
 Add the passed parameters to the object's kernel's arguments starting from the index idx. More...
 
void set_lws_hint (const cl::NDRange &lws_hint)
 Set the Local-Workgroup-Size hint. More...
 
cl::NDRange lws_hint () const
 Return the Local-Workgroup-Size hint. More...
 
const std::string & config_id () const
 Get the configuration ID. More...
 
void set_target (GPUTarget target)
 Set the targeted GPU architecture. More...
 
void set_target (cl::Device &device)
 Set the targeted GPU architecture according to the CL device. More...
 
GPUTarget get_target () const
 Get the targeted GPU architecture. More...
 
size_t get_max_workgroup_size ()
 Get the maximum workgroup size for the device the CLKernelLibrary uses. More...
 
template<typename T , unsigned int dimension_size>
void add_array_argument (unsigned &idx, const ICLArray< T > *array, const Strides &strides, unsigned int num_dimensions, const Window &window)
 Add the passed array's parameters to the object's kernel's arguments starting from the index idx. More...
 
template<unsigned int dimension_size>
void add_tensor_argument (unsigned &idx, const ICLTensor *tensor, const Window &window)
 
- 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...
 
const Windowwindow () const
 The maximum window the kernel can be executed on. More...
 

Static Public Member Functions

static Status validate (const ITensorInfo *input1, const ITensorInfo *input2, const ITensorInfo *output, float scale, ConvertPolicy overflow_policy, RoundingPolicy rounding_policy)
 Static function to check if given info will lead to a valid configuration of CLPixelWiseMultiplicationKernel. More...
 
- Static Public Member Functions inherited from ICLKernel
static constexpr unsigned int num_arguments_per_1D_array ()
 Returns the number of arguments enqueued per 1D array object. More...
 
static constexpr unsigned int num_arguments_per_1D_tensor ()
 Returns the number of arguments enqueued per 1D tensor object. More...
 
static constexpr unsigned int num_arguments_per_2D_tensor ()
 Returns the number of arguments enqueued per 2D tensor object. More...
 
static constexpr unsigned int num_arguments_per_3D_tensor ()
 Returns the number of arguments enqueued per 3D tensor object. More...
 
static constexpr unsigned int num_arguments_per_4D_tensor ()
 Returns the number of arguments enqueued per 4D tensor object. More...
 
static cl::NDRange gws_from_window (const Window &window)
 Get the global work size given an execution window. More...
 

Detailed Description

Interface for the pixelwise multiplication kernel.

Definition at line 36 of file CLPixelWiseMultiplicationKernel.h.

Constructor & Destructor Documentation

◆ CLPixelWiseMultiplicationKernel() [1/3]

Default constructor.

Definition at line 127 of file CLPixelWiseMultiplicationKernel.cpp.

128  : _input1(nullptr), _input2(nullptr), _output(nullptr)
129 {
130 }

◆ CLPixelWiseMultiplicationKernel() [2/3]

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

◆ CLPixelWiseMultiplicationKernel() [3/3]

Allow instances of this class to be moved.

Member Function Documentation

◆ border_size()

BorderSize border_size ( ) const
overridevirtual

The size of the border for that kernel.

Returns
The width in number of elements of the border.

Reimplemented from IKernel.

Definition at line 285 of file CLPixelWiseMultiplicationKernel.cpp.

286 {
287  const unsigned int replicateSize = _output->info()->dimension(0) - std::min(_input1->info()->dimension(0), _input2->info()->dimension(0));
288  const unsigned int border = std::min<unsigned int>(num_elems_processed_per_iteration - 1U, replicateSize);
289  return BorderSize{ 0, border, 0, 0 };
290 }
virtual size_t dimension(size_t index) const =0
Return the size of the requested dimension.
virtual ITensorInfo * info() const =0
Interface to be implemented by the child class to return the tensor's metadata.

References ITensorInfo::dimension(), ITensor::info(), and arm_compute::U.

◆ configure()

void configure ( const ICLTensor input1,
const ICLTensor input2,
ICLTensor output,
float  scale,
ConvertPolicy  overflow_policy,
RoundingPolicy  rounding_policy 
)

Initialise the kernel's input, output and border mode.

Parameters
[in]input1An input tensor. Data types supported: U8/QASYMM8/S16/QSYMM16/F16/F32.
[in]input2An input tensor. Data types supported: same as input1.
[out]outputThe output tensor, Data types supported: same as input1. Note: U8 requires both inputs to be U8.
[in]scaleScale to apply after multiplication. Scale must be positive and its value must be either 1/255 or 1/2^n where n is between 0 and 15.
[in]overflow_policyOverflow policy. Supported overflow policies: Wrap, Saturate
[in]rounding_policyRounding policy. Supported rounding modes: to zero, to nearest even.

Definition at line 132 of file CLPixelWiseMultiplicationKernel.cpp.

134 {
135  ARM_COMPUTE_ERROR_ON_NULLPTR(input1, input2, output);
136  ARM_COMPUTE_ERROR_THROW_ON(validate_arguments(input1->info(), input2->info(), output->info(),
137  scale, overflow_policy, rounding_policy));
138 
139  // Configure kernel window
140  auto win_config = validate_and_configure_window(input1->info(), input2->info(), output->info());
141  ARM_COMPUTE_ERROR_THROW_ON(win_config.first);
142 
143  _input1 = input1;
144  _input2 = input2;
145  _output = output;
146 
147  int scale_int = -1;
148  // Extract sign, exponent and mantissa
149  int exponent = 0;
150  float normalized_mantissa = std::frexp(scale, &exponent);
151  // Use int scaling if factor is equal to 1/2^n for 0 <= n <= 15
152  // frexp returns 0.5 as mantissa which means that the exponent will be in the range of -1 <= e <= 14
153  // Moreover, it will be negative as we deal with 1/2^n
154  if((normalized_mantissa == 0.5f) && (-14 <= exponent) && (exponent <= 1))
155  {
156  // Store the positive exponent. We know that we compute 1/2^n
157  // Additionally we need to subtract 1 to compensate that frexp used a mantissa of 0.5
158  scale_int = std::abs(exponent - 1);
159  }
160 
161  std::string compute_type;
162  // Check if it has float inputs and output
163  if(is_data_type_float(input1->info()->data_type()) || is_data_type_float(input2->info()->data_type()))
164  {
165  scale_int = -1;
166  compute_type = (input1->info()->data_type() == DataType::F32 || input2->info()->data_type() == DataType::F32) ? "float" : "half";
167  }
168  else
169  {
170  if(input1->info()->data_type() == DataType::S16 || input2->info()->data_type() == DataType::S16)
171  {
172  compute_type = "int";
173  }
174  else
175  {
176  compute_type = "ushort";
177  }
178  }
179 
180  const bool is_quantized = is_data_type_quantized(input1->info()->data_type());
181 
182  // Set kernel build options
183  std::string kernel_name = "pixelwise_mul";
184  CLBuildOptions build_opts;
185  build_opts.add_option("-DDATA_TYPE_IN1=" + get_cl_type_from_data_type(input1->info()->data_type()));
186  build_opts.add_option("-DDATA_TYPE_IN2=" + get_cl_type_from_data_type(input2->info()->data_type()));
187  build_opts.add_option("-DDATA_TYPE_OUT=" + get_cl_type_from_data_type(output->info()->data_type()));
188  build_opts.add_option("-DVEC_SIZE=" + support::cpp11::to_string(num_elems_processed_per_iteration));
189  if(is_quantized)
190  {
191  const UniformQuantizationInfo iq1_info = input1->info()->quantization_info().uniform();
192  const UniformQuantizationInfo iq2_info = input2->info()->quantization_info().uniform();
193  const UniformQuantizationInfo oq_info = output->info()->quantization_info().uniform();
194 
195  build_opts.add_option_if(is_data_type_quantized_asymmetric(input1->info()->data_type()),
196  "-DOFFSET_IN1=" + support::cpp11::to_string(iq1_info.offset));
197  build_opts.add_option_if(is_data_type_quantized_asymmetric(input2->info()->data_type()),
198  "-DOFFSET_IN2=" + support::cpp11::to_string(iq2_info.offset));
199  build_opts.add_option_if(is_data_type_quantized_asymmetric(output->info()->data_type()),
200  "-DOFFSET_OUT=" + support::cpp11::to_string(oq_info.offset));
201  build_opts.add_option("-DSCALE_IN1=" + float_to_string_with_full_precision(iq1_info.scale));
202  build_opts.add_option("-DSCALE_IN2=" + float_to_string_with_full_precision(iq2_info.scale));
203  build_opts.add_option("-DSCALE_OUT=" + float_to_string_with_full_precision(oq_info.scale));
204  kernel_name += "_quantized";
205  }
206  else
207  {
208  kernel_name += (scale_int >= 0) ? "_int" : "_float";
209  build_opts.add_option_if_else(overflow_policy == ConvertPolicy::WRAP || is_data_type_float(output->info()->data_type()), "-DWRAP", "-DSATURATE");
210  build_opts.add_option_if_else(rounding_policy == RoundingPolicy::TO_ZERO, "-DROUND=_rtz", "-DROUND=_rte");
211  build_opts.add_option("-DDATA_TYPE_RES=" + compute_type);
212  }
213 
214  // Create kernel
215  _kernel = static_cast<cl::Kernel>(CLKernelLibrary::get().create_kernel(kernel_name, build_opts.options()));
216 
217  // Set scale argument
218  unsigned int idx = 3 * num_arguments_per_3D_tensor(); // Skip the inputs and output parameters
219 
220  if(scale_int >= 0 && !is_quantized)
221  {
222  _kernel.setArg(idx++, scale_int);
223  }
224  else
225  {
226  _kernel.setArg(idx++, scale);
227  }
228 
229  ICLKernel::configure_internal(win_config.second);
230 }
bool is_data_type_quantized(DataType dt)
Check if a given data type is of quantized type.
Definition: Utils.h:1010
std::pair< Status, Window > validate_and_configure_window(ITensorInfo *input, ITensorInfo *weights, ITensorInfo *biases, ITensorInfo *output, const PadStrideInfo &conv_info, unsigned int depth_multiplier, const Size2D &dilation)
std::string to_string(T &&value)
Convert integer and float values to string.
1 channel, 1 F32 per channel
static CLKernelLibrary & get()
Access the KernelLibrary singleton.
#define ARM_COMPUTE_ERROR_THROW_ON(status)
Definition: Error.h:327
static constexpr unsigned int num_arguments_per_3D_tensor()
Returns the number of arguments enqueued per 3D tensor object.
Definition: ICLKernel.h:200
std::string float_to_string_with_full_precision(float val)
Create a string with the float in full precision.
Definition: Utils.h:1066
std::string get_cl_type_from_data_type(const DataType &dt)
Translates a tensor data type to the appropriate OpenCL type.
Definition: CLHelpers.cpp:35
std::unique_ptr< Kernel > create_kernel()
Helper function to create and return a unique_ptr pointed to a CL/GLES kernel object.
Definition: Helpers.h:86
1 channel, 1 S16 per channel
bool is_data_type_quantized_asymmetric(DataType dt)
Check if a given data type is of asymmetric quantized type.
Definition: Utils.h:1030
#define ARM_COMPUTE_ERROR_ON_NULLPTR(...)
Definition: Validate.h:161
Truncates the least significant values that are lost in operations.
bool is_data_type_float(DataType dt)
Check if a given data type is of floating point type.
Definition: Utils.h:990

References CLBuildOptions::add_option(), CLBuildOptions::add_option_if(), CLBuildOptions::add_option_if_else(), ARM_COMPUTE_ERROR_ON_NULLPTR, ARM_COMPUTE_ERROR_THROW_ON, arm_compute::create_kernel(), ITensorInfo::data_type(), arm_compute::F32, arm_compute::float_to_string_with_full_precision(), CLKernelLibrary::get(), arm_compute::get_cl_type_from_data_type(), ITensor::info(), arm_compute::is_data_type_float(), arm_compute::is_data_type_quantized(), arm_compute::is_data_type_quantized_asymmetric(), ICLKernel::num_arguments_per_3D_tensor(), UniformQuantizationInfo::offset, CLBuildOptions::options(), ITensorInfo::quantization_info(), arm_compute::test::validation::rounding_policy, arm_compute::S16, UniformQuantizationInfo::scale, arm_compute::test::validation::scale, arm_compute::support::cpp11::to_string(), arm_compute::TO_ZERO, QuantizationInfo::uniform(), arm_compute::validate_and_configure_window(), and arm_compute::WRAP.

Referenced by CLLSTMLayer::configure().

◆ operator=() [1/2]

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

◆ operator=() [2/2]

Allow instances of this class to be moved.

◆ run()

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

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

Implements ICLKernel.

Definition at line 242 of file CLPixelWiseMultiplicationKernel.cpp.

243 {
246 
247  const TensorShape &in_shape1 = _input1->info()->tensor_shape();
248  const TensorShape &in_shape2 = _input2->info()->tensor_shape();
249  const TensorShape &out_shape = _output->info()->tensor_shape();
250 
251  bool can_collapse = true;
252  if(std::min(in_shape1.total_size(), in_shape2.total_size()) > 1)
253  {
254  can_collapse = (std::min(in_shape1.num_dimensions(), in_shape2.num_dimensions()) > Window::DimZ);
255  for(size_t d = Window::DimZ; can_collapse && (d < out_shape.num_dimensions()); ++d)
256  {
257  can_collapse = (in_shape1[d] == in_shape2[d]);
258  }
259  }
260 
261  bool has_collapsed = false;
262  Window collapsed = can_collapse ? window.collapse_if_possible(ICLKernel::window(), Window::DimZ, &has_collapsed) : window;
263 
264  const TensorShape &in_shape1_collapsed = has_collapsed ? in_shape1.collapsed_from(Window::DimZ) : in_shape1;
265  const TensorShape &in_shape2_collapsed = has_collapsed ? in_shape2.collapsed_from(Window::DimZ) : in_shape2;
266 
267  Window slice = collapsed.first_slice_window_3D();
268  Window slice_input1 = slice.broadcast_if_dimension_le_one(in_shape1_collapsed);
269  Window slice_input2 = slice.broadcast_if_dimension_le_one(in_shape2_collapsed);
270 
271  do
272  {
273  unsigned int idx = 0;
274  add_3D_tensor_argument(idx, _input1, slice_input1);
275  add_3D_tensor_argument(idx, _input2, slice_input2);
276  add_3D_tensor_argument(idx, _output, slice);
277  enqueue(queue, *this, slice);
278 
279  ARM_COMPUTE_UNUSED(collapsed.slide_window_slice_3D(slice_input1));
280  ARM_COMPUTE_UNUSED(collapsed.slide_window_slice_3D(slice_input2));
281  }
282  while(collapsed.slide_window_slice_3D(slice));
283 }
const Window & window() const
The maximum window the kernel can be executed on.
Definition: IKernel.cpp:28
void enqueue(cl::CommandQueue &queue, ICLKernel &kernel, const Window &window, const cl::NDRange &lws_hint=CLKernelLibrary::get().default_ndrange(), bool use_dummy_work_items=false)
Add the kernel to the command queue with the given window.
Definition: ICLKernel.cpp:39
void add_3D_tensor_argument(unsigned int &idx, const ICLTensor *tensor, const Window &window)
Add the passed 3D tensor's parameters to the object's kernel's arguments starting from the index idx.
Definition: ICLKernel.h:158
#define ARM_COMPUTE_UNUSED(...)
To avoid unused variables warnings.
Definition: Error.h:160
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:54
virtual const TensorShape & tensor_shape() const =0
Size for each dimension of the tensor.
virtual ITensorInfo * info() const =0
Interface to be implemented by the child class to return the tensor's metadata.
static constexpr size_t DimZ
Alias for dimension 2 also known as Z dimension.
Definition: Window.h:47
#define ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW(f, s)
Definition: Validate.h:205
#define ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL(k)
Definition: Validate.h:940
SimpleTensor< T > slice(const SimpleTensor< T > &src, Coordinates starts, Coordinates ends)

References ICLKernel::add_3D_tensor_argument(), ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW, ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL, ARM_COMPUTE_UNUSED, Window::collapse_if_possible(), TensorShape::collapsed_from(), Window::DimZ, arm_compute::enqueue(), Window::first_slice_window_3D(), ITensor::info(), Dimensions< T >::num_dimensions(), arm_compute::test::validation::reference::slice(), Window::slide_window_slice_3D(), ITensorInfo::tensor_shape(), TensorShape::total_size(), and IKernel::window().

◆ validate()

Status validate ( const ITensorInfo input1,
const ITensorInfo input2,
const ITensorInfo output,
float  scale,
ConvertPolicy  overflow_policy,
RoundingPolicy  rounding_policy 
)
static

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

Parameters
[in]input1An input tensor info. Data types supported: U8/QASYMM8/S16/QSYMM16/F16/F32.
[in]input2An input tensor info. Data types supported: same as input1.
[in]outputThe output tensor info, Data types supported: same as input1. Note: U8 requires both inputs to be U8.
[in]scaleScale to apply after multiplication. Scale must be positive and its value must be either 1/255 or 1/2^n where n is between 0 and 15.
[in]overflow_policyOverflow policy. Supported overflow policies: Wrap, Saturate
[in]rounding_policyRounding policy. Supported rounding modes: to zero, to nearest even.
Returns
a status

Definition at line 232 of file CLPixelWiseMultiplicationKernel.cpp.

234 {
235  ARM_COMPUTE_ERROR_ON_NULLPTR(input1, input2, output);
236  ARM_COMPUTE_RETURN_ON_ERROR(validate_arguments(input1, input2, output, scale, overflow_policy, rounding_policy));
237  ARM_COMPUTE_RETURN_ON_ERROR(validate_and_configure_window(input1->clone().get(), input2->clone().get(), output->clone().get()).first);
238 
239  return Status{};
240 }
std::pair< Status, Window > validate_and_configure_window(ITensorInfo *input, ITensorInfo *weights, ITensorInfo *biases, ITensorInfo *output, const PadStrideInfo &conv_info, unsigned int depth_multiplier, const Size2D &dilation)
#define ARM_COMPUTE_RETURN_ON_ERROR(status)
Checks if a status contains an error and returns it.
Definition: Error.h:193
#define ARM_COMPUTE_ERROR_ON_NULLPTR(...)
Definition: Validate.h:161

References ARM_COMPUTE_ERROR_ON_NULLPTR, ARM_COMPUTE_RETURN_ON_ERROR, ICloneable< T >::clone(), arm_compute::test::validation::rounding_policy, arm_compute::test::validation::scale, and arm_compute::validate_and_configure_window().

Referenced by CLPixelWiseMultiplication::validate(), and CLLSTMLayer::validate().


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