Compute Library
 19.08
CLGEMMLowpMatrixMultiplyKernel Class Reference

OpenCL kernel to multiply matrices. More...

#include <CLGEMMLowpMatrixMultiplyKernel.h>

Collaboration diagram for CLGEMMLowpMatrixMultiplyKernel:
[legend]

Public Member Functions

 CLGEMMLowpMatrixMultiplyKernel ()
 Default Constructor. More...
 
 CLGEMMLowpMatrixMultiplyKernel (const CLGEMMLowpMatrixMultiplyKernel &)=delete
 Prevent instances of this class from being copied (As this class contains pointers) More...
 
CLGEMMLowpMatrixMultiplyKerneloperator= (const CLGEMMLowpMatrixMultiplyKernel &)=delete
 Prevent instances of this class from being copied (As this class contains pointers) More...
 
 CLGEMMLowpMatrixMultiplyKernel (CLGEMMLowpMatrixMultiplyKernel &&)=default
 Allow instances of this class to be moved. More...
 
CLGEMMLowpMatrixMultiplyKerneloperator= (CLGEMMLowpMatrixMultiplyKernel &&)=default
 Allow instances of this class to be moved. More...
 
void configure (const ICLTensor *input0, const ICLTensor *input1, ICLTensor *output, const GEMMReshapeInfo &gemm_info=GEMMReshapeInfo())
 Initialise the kernel's input and output. 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...
 
- 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...
 
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 *input0, const ITensorInfo *input1, const ITensorInfo *output, const GEMMReshapeInfo &gemm_info=GEMMReshapeInfo())
 Static function to check if given info will lead to a valid configuration of CLGEMMLowpMatrixMultiplyKernel. 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

OpenCL kernel to multiply matrices.

Note
This kernel should be used ONLY for Midgard architectures

This kernel performs the following computation:

  1. Convert a values from int8 to int32
  2. Convert b values from int8 to int32
  3. Compute the int32 matrix product of the resulting a * b and store the result as int32

Definition at line 44 of file CLGEMMLowpMatrixMultiplyKernel.h.

Constructor & Destructor Documentation

◆ CLGEMMLowpMatrixMultiplyKernel() [1/3]

Default Constructor.

Definition at line 168 of file CLGEMMLowpMatrixMultiplyKernel.cpp.

169  : _input0(nullptr), _input1(nullptr), _output(nullptr), _slide_matrix_b(true), _reinterpret_input_as_3d(false), _reinterpret_output_as_3d(false)
170 {
171 }

◆ CLGEMMLowpMatrixMultiplyKernel() [2/3]

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

◆ CLGEMMLowpMatrixMultiplyKernel() [3/3]

Allow instances of this class to be moved.

Member Function Documentation

◆ configure()

void configure ( const ICLTensor input0,
const ICLTensor input1,
ICLTensor output,
const GEMMReshapeInfo gemm_info = GEMMReshapeInfo() 
)

Initialise the kernel's input and output.

Note
This kernel should be used ONLY for Midgard architectures
Parameters
[in]input0Input tensor containing the LHS matrix. Data type supported: QASYMM8
[in]input1Input tensor containing the RHS matrix. Data type supported: same as input0
[out]outputOutput tensor to store the result of matrix multiplication. Data type supported: S32
[in]gemm_info(Optional) GEMM information used to retrieve the original dimensions of the input matrices

Definition at line 173 of file CLGEMMLowpMatrixMultiplyKernel.cpp.

174 {
175  ARM_COMPUTE_ERROR_ON_NULLPTR(input0, input1, output);
176 
177  ARM_COMPUTE_ERROR_THROW_ON(validate_arguments(input0->info(), input1->info(), output->info(), gemm_info));
178 
179  _input0 = input0;
180  _input1 = input1;
181  _output = output;
182  _reinterpret_input_as_3d = gemm_info.reinterpret_input_as_3d();
183  _reinterpret_output_as_3d = (gemm_info.depth_output_gemm3d() != 0);
184 
185  // In case both input and output have to be reinterpreted as 3D tensors,
186  // force reinterpret_input_as_3d and reinterpret_output_as_3d to be false.
187  if(_reinterpret_input_as_3d == _reinterpret_output_as_3d)
188  {
189  _reinterpret_input_as_3d = false;
190  _reinterpret_output_as_3d = false;
191  }
192 
193  // Check if we need to slide the matrix B
194  const unsigned int num_dimensions_input0 = _reinterpret_input_as_3d ? _input0->info()->num_dimensions() - 1 : _input0->info()->num_dimensions();
195  _slide_matrix_b = (_input1->info()->num_dimensions() >= num_dimensions_input0);
196 
197  ElementsProcessed num_elements_processed{};
198 
199  // Configure kernel window
200  auto win_config = validate_and_configure_window(input0->info(), input1->info(), output->info(), gemm_info, num_elements_processed);
201  ARM_COMPUTE_ERROR_THROW_ON(win_config.first);
202  ICLKernel::configure_internal(win_config.second);
203 
204  // Create build options
205  std::string kernel_name(" ");
206  CLBuildOptions build_opts;
207  build_opts.add_option_if(_reinterpret_input_as_3d, "-DREINTERPRET_INPUT_AS_3D");
208  build_opts.add_option_if(_reinterpret_output_as_3d, "-DREINTERPRET_OUTPUT_AS_3D");
209  build_opts.add_option_if(_reinterpret_input_as_3d || _reinterpret_output_as_3d, "-DHEIGHT_GEMM3D=" + support::cpp11::to_string(output->info()->dimension(1)));
210  build_opts.add_option_if(_reinterpret_input_as_3d || _reinterpret_output_as_3d, "-DDEPTH_GEMM3D=" + support::cpp11::to_string(output->info()->dimension(2)));
211  build_opts.add_option_if(!_slide_matrix_b, "-DMATRIX_B_DEPTH=" + support::cpp11::to_string(input1->info()->dimension(2)));
212  build_opts.add_option("-DCOLS_A=" + support::cpp11::to_string(input0->info()->dimension(0)));
213  build_opts.add_option("-DNUM_ELEMS_PROCESSED_PER_THREAD_X=" + support::cpp11::to_string(num_elements_processed.x()));
214  build_opts.add_option("-DNUM_ELEMS_PROCESSED_PER_THREAD_Y=" + support::cpp11::to_string(num_elements_processed.y()));
215 
216  kernel_name = "gemmlowp_mm_midgard";
217 
218  // Create kernel
219  _kernel = static_cast<cl::Kernel>(CLKernelLibrary::get().create_kernel(kernel_name, build_opts.options()));
220 
221  // Set config_id for enabling LWS tuning
222  _config_id = kernel_name;
223  _config_id += "_";
224  _config_id += (_reinterpret_input_as_3d ? "3di_" : "");
225  _config_id += (_reinterpret_output_as_3d ? "3do_" : "");
226  _config_id += lower_string(string_from_data_type(input0->info()->data_type()));
227  _config_id += "_";
228  _config_id += support::cpp11::to_string(output->info()->dimension(1));
229  _config_id += "_";
230  _config_id += support::cpp11::to_string(output->info()->dimension(0));
231 }
virtual size_t num_dimensions() const =0
The number of dimensions of the tensor (rank)
virtual size_t dimension(size_t index) const =0
Return the size of the requested dimension.
const StringSet & options() const
Gets the current options list set.
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.
virtual DataType data_type() const =0
Data type used for each element of the tensor.
static CLKernelLibrary & get()
Access the KernelLibrary singleton.
#define ARM_COMPUTE_ERROR_THROW_ON(status)
Definition: Error.h:327
std::string lower_string(const std::string &val)
Lower a given string.
Definition: Utils.cpp:327
void add_option(std::string option)
Adds option to the existing build option list.
const std::string & string_from_data_type(DataType dt)
Convert a data type identity into a string.
Definition: Utils.cpp:144
virtual ITensorInfo * info() const =0
Interface to be implemented by the child class to return the tensor's metadata.
void add_option_if(bool cond, std::string option)
Adds option if a given condition is true;.
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
#define ARM_COMPUTE_ERROR_ON_NULLPTR(...)
Definition: Validate.h:161
int depth_output_gemm3d() const
Depth (third dimension) of the output tensor to be used with the GEMM3D kernel.
Definition: Types.h:1797
bool reinterpret_input_as_3d() const
Flag which specifies if the input tensor has to be reinterpreted as 3D.
Definition: Types.h:1805

References CLBuildOptions::add_option(), CLBuildOptions::add_option_if(), ARM_COMPUTE_ERROR_ON_NULLPTR, ARM_COMPUTE_ERROR_THROW_ON, arm_compute::create_kernel(), ITensorInfo::data_type(), GEMMReshapeInfo::depth_output_gemm3d(), ITensorInfo::dimension(), CLKernelLibrary::get(), ITensor::info(), arm_compute::lower_string(), ITensorInfo::num_dimensions(), CLBuildOptions::options(), GEMMReshapeInfo::reinterpret_input_as_3d(), arm_compute::string_from_data_type(), arm_compute::support::cpp11::to_string(), and arm_compute::validate_and_configure_window().

Referenced by CLGEMMLowpMatrixMultiplyCore::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 247 of file CLGEMMLowpMatrixMultiplyKernel.cpp.

248 {
251 
252  if(_input1->info()->num_dimensions() < 3)
253  {
254  // The stride_z for matrix B must be zero if we do not slice
255  ARM_COMPUTE_ERROR_ON(_input1->info()->strides_in_bytes()[3] != 0);
256  }
257 
259  Window slice_matrix_b = slice;
260 
261  slice_matrix_b.set(Window::DimX, Window::Dimension(0, 1, 1));
262  slice_matrix_b.set(Window::DimY, Window::Dimension(0, 1, 1));
263 
264  if(_reinterpret_input_as_3d)
265  {
266  // Pass bottom paddings to the kernel if the input has to be reinterpreted as 3D tensor
267  const unsigned int idx0 = 3 * num_arguments_per_2D_tensor() + 3;
268  const unsigned int total_cross_plane_pad = _input0->info()->padding().top + _input0->info()->padding().bottom;
269  _kernel.setArg<cl_uint>(idx0, static_cast<unsigned int>(total_cross_plane_pad));
270  }
271 
272  if(_reinterpret_output_as_3d)
273  {
274  // Pass bottom paddings to the kernel if the output has to be reinterpreted as 3D tensor
275  const unsigned int idx0 = 3 * num_arguments_per_2D_tensor() + 3 + (_reinterpret_input_as_3d ? 1 : 0);
276  const unsigned int total_cross_plane_pad = _output->info()->padding().top + _output->info()->padding().bottom;
277  _kernel.setArg<cl_uint>(idx0, static_cast<unsigned int>(total_cross_plane_pad));
278  }
279 
280  do
281  {
282  Window slice_b = slice;
283  // Don't slice matrix B along the z dimension if matrix B has just 2 dimensions and matrix A more than 2
284  // This scenario can happen when the matrix multiplication is used to perform a convolution operation
285  if(!_slide_matrix_b)
286  {
287  slice_b = slice_matrix_b;
288  }
289 
290  unsigned int idx = 0;
291  add_2D_tensor_argument(idx, _input0, slice);
292  add_2D_tensor_argument(idx, _input1, slice_b);
293  add_2D_tensor_argument(idx, _output, slice);
294  _kernel.setArg<cl_uint>(idx++, static_cast<unsigned int>(_input0->info()->strides_in_bytes()[2]));
295  _kernel.setArg<cl_uint>(idx++, static_cast<unsigned int>(_input1->info()->strides_in_bytes()[2]));
296  _kernel.setArg<cl_uint>(idx++, static_cast<unsigned int>(_output->info()->strides_in_bytes()[2]));
297  enqueue(queue, *this, slice, lws_hint());
298  }
300 }
unsigned int top
top of the border
Definition: Types.h:339
virtual size_t num_dimensions() const =0
The number of dimensions of the tensor (rank)
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
cl::NDRange lws_hint() const
Return the Local-Workgroup-Size hint.
Definition: ICLKernel.h:247
#define ARM_COMPUTE_ERROR_ON(cond)
If the condition is true then an error message is printed and an exception thrown.
Definition: Error.h:337
Describe one of the image's dimensions with a start, end and step.
Definition: Window.h:75
unsigned int bottom
bottom of the border
Definition: Types.h:341
static constexpr size_t DimX
Alias for dimension 0 also known as X dimension.
Definition: Window.h:43
virtual ITensorInfo * info() const =0
Interface to be implemented by the child class to return the tensor's metadata.
void set(size_t dimension, const Dimension &dim)
Set the values of a given dimension.
Definition: Window.inl:48
virtual PaddingSize padding() const =0
Padding of tensor.
static constexpr unsigned int num_arguments_per_2D_tensor()
Returns the number of arguments enqueued per 2D tensor object.
Definition: ICLKernel.h:192
bool slide_window_slice_3D(Window &slice) const
Slide the passed 3D window slice.
Definition: Window.h:319
static constexpr size_t DimY
Alias for dimension 1 also known as Y dimension.
Definition: Window.h:45
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.
Definition: ICLKernel.h:134
virtual const Strides & strides_in_bytes() const =0
The strides in bytes for accessing each dimension of the tensor.
Window first_slice_window_3D() const
First 3D slice of the window.
Definition: Window.h:275
#define ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW(f, s)
Definition: Validate.h:205
Describe a multidimensional execution window.
Definition: Window.h:39
#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_2D_tensor_argument(), ARM_COMPUTE_ERROR_ON, ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW, ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL, BorderSize::bottom, Window::DimX, Window::DimY, arm_compute::enqueue(), Window::first_slice_window_3D(), ITensor::info(), ICLKernel::lws_hint(), ICLKernel::num_arguments_per_2D_tensor(), ITensorInfo::num_dimensions(), ITensorInfo::padding(), Window::set(), arm_compute::test::validation::reference::slice(), Window::slide_window_slice_3D(), ITensorInfo::strides_in_bytes(), BorderSize::top, and IKernel::window().

◆ validate()

Status validate ( const ITensorInfo input0,
const ITensorInfo input1,
const ITensorInfo output,
const GEMMReshapeInfo gemm_info = GEMMReshapeInfo() 
)
static

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

Parameters
[in]input0Input tensor containing the LHS matrix. Data type supported: QASYMM8
[in]input1Input tensor containing the RHS matrix. Data type supported: same as input0
[in]outputOutput tensor to store the result of matrix multiplication. Data type supported: S32
[in]gemm_info(Optional) GEMM information used to retrieve the original dimensions of the input matrices
Returns
a status

Definition at line 233 of file CLGEMMLowpMatrixMultiplyKernel.cpp.

234 {
235  ElementsProcessed num_elements_processed{};
236  ARM_COMPUTE_RETURN_ON_ERROR(validate_arguments(input0, input1, output, gemm_info));
238  input1->clone().get(),
239  output->clone().get(),
240  gemm_info,
241  num_elements_processed)
242  .first);
243 
244  return Status{};
245 }
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
Status class.
Definition: Error.h:52
virtual std::unique_ptr< T > clone() const =0
Provide a clone of the current object of class T.

References ARM_COMPUTE_RETURN_ON_ERROR, ICloneable< T >::clone(), and arm_compute::validate_and_configure_window().

Referenced by CLGEMMLowpMatrixMultiplyCore::validate().


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