Compute Library
 22.05
ClDirectConv2dKernel Class Reference

Interface for the direct convolution kernel. More...

#include <ClDirectConv2dKernel.h>

Collaboration diagram for ClDirectConv2dKernel:
[legend]

Public Member Functions

 ClDirectConv2dKernel ()
 
 ARM_COMPUTE_DISALLOW_COPY_ALLOW_MOVE (ClDirectConv2dKernel)
 
void configure (const CLCompileContext &compile_context, ITensorInfo *src, ITensorInfo *weights, ITensorInfo *biases, ITensorInfo *dst, const PadStrideInfo &conv_info, const ActivationLayerInfo &act_info)
 Set the src, weights, biases and dst tensors info. More...
 
void run_op (ITensorPack &tensors, 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...
 
CLKernelType type () const
 Returns the CL kernel type. 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...
 
void add_5D_tensor_argument (unsigned int &idx, const ICLTensor *tensor, const Window &window)
 Add the passed 5D tensor's parameters to the object's kernel's arguments starting from the index idx. More...
 
void add_3d_tensor_nhw_argument (unsigned int &idx, const ICLTensor *tensor)
 Add the passed NHW 3D tensor's parameters to the object's kernel's arguments by passing strides, dimensions and the offset to the first valid element in bytes. More...
 
void add_4d_tensor_nhwc_argument (unsigned int &idx, const ICLTensor *tensor)
 Add the passed NHWC 4D tensor's parameters to the object's kernel's arguments by passing strides, dimensions and the offset to the first valid element in bytes. More...
 
virtual void run (const Window &window, cl::CommandQueue &queue)
 Enqueue the OpenCL kernel to process the given window on the passed OpenCL command queue. More...
 
virtual void run_composite_op (ITensorPack &tensors, const Window &window, cl::CommandQueue &queue, const experimental::dynamic_fusion::ClExecutionDescriptor &exec_desc)
 The execution is carried out through run_op method. But the run_op method needs to be extended to include ClExecutionDescriptor as now LWS GWS tuning will be separated from the IKernel. 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...
 
void set_wbsm_hint (const cl_int &wbsm_hint)
 Set the workgroup batch size modifier hint. More...
 
cl_int wbsm_hint () const
 Return the workgroup batch size modifier 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<unsigned int dimension_size>
void add_tensor_argument (unsigned &idx, const ICLTensor *tensor, const Window &window)
 
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...
 
- 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...
 
bool is_window_configured () const
 Function to check if the embedded window of this kernel has been configured. More...
 

Static Public Member Functions

static Status validate (const ITensorInfo *src, const ITensorInfo *weights, const ITensorInfo *biases, const ITensorInfo *dst, const PadStrideInfo &conv_info, const ActivationLayerInfo &act_info)
 Static function to check if given info will lead to a valid configuration. More...
 
- Static Public Member Functions inherited from ICLKernel
static constexpr unsigned int num_arguments_per_3d_tensor_nhw ()
 Returns the number of arguments enqueued per NHW 3D Tensor object. More...
 
static constexpr unsigned int num_arguments_per_4d_tensor_nhwc ()
 Returns the number of arguments enqueued per NHWC 4D Tensor object. More...
 
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...
 

Data Fields

DataLayout _data_layout {}
 
PadStrideInfo _conv_info {}
 

Detailed Description

Interface for the direct convolution kernel.

Definition at line 38 of file ClDirectConv2dKernel.h.

Constructor & Destructor Documentation

◆ ClDirectConv2dKernel()

Definition at line 170 of file ClDirectConv2dKernel.cpp.

References arm_compute::DIRECT.

171 {
172  _type = CLKernelType::DIRECT;
173 }

Member Function Documentation

◆ ARM_COMPUTE_DISALLOW_COPY_ALLOW_MOVE()

ARM_COMPUTE_DISALLOW_COPY_ALLOW_MOVE ( ClDirectConv2dKernel  )

◆ configure()

void configure ( const CLCompileContext compile_context,
ITensorInfo src,
ITensorInfo weights,
ITensorInfo biases,
ITensorInfo dst,
const PadStrideInfo conv_info,
const ActivationLayerInfo act_info 
)

Set the src, weights, biases and dst tensors info.

Note
: Due to set_valid_region() in NCHW, src/weights/biases cannot be const. Need to change this once the set_valid_region() is removed.
: DirectConvolution only works in the following configurations for the NCHW data layout: 1x1 convolution with stride_x = 1/2/3, stride_y = 1/2/3 3x3 convolution with stride_x = 1/2, stride_y = 1/2 5x5 convolution with stride_x = 1/2, stride_y = 1/2 9x9 convolution with stride_x = 1/2, stride_y = 1/2
Parameters
[in]compile_contextThe compile context to be used.
[in]srcThe src tensor info to convolve. 3 lower dimensions represent a single src [width, height, IFM], while every optional dimension from 4 and above represent a batch of inputs. Data types supported: QASYMM8_SIGNED/QASYMM8/F16/F32.
[in]weightsWeights tensor info. Weights are 4D tensor with dimensions [kernel_x, kernel_y, IFM, OFM]. The 3rd dimension must be the same as the src's volume 3rd dimension. Data type supported:Same as src.
[in]biasesBiases tensor info. Biases are 1D tensor with dimension [OFM]. Data type supported: Should match src data type, except for src of QASYMM8 and QASYMM8_SIGNED type where biases should be of S32 type
[out]dstOutput tensor info. The 3rd dimensions must be equal to the 4th dimension of the kernels tensor. Data types supported: Same as src.
[in]conv_infoContains padding and stride information described in PadStrideInfo.
[in]act_infoContains activaton information described in ActivationLayerInfo.

Definition at line 175 of file ClDirectConv2dKernel.cpp.

References ClDirectConv2dKernel::_conv_info, ClDirectConv2dKernel::_data_layout, ActivationLayerInfo::a(), ActivationLayerInfo::activation(), CLBuildOptions::add_option(), CLBuildOptions::add_option_if(), CLBuildOptions::add_option_if_else(), arm_compute::adjust_vec_size(), ARM_COMPUTE_ERROR_ON_NULLPTR, ARM_COMPUTE_ERROR_THROW_ON, arm_compute::auto_init_if_empty(), ActivationLayerInfo::b(), IKernel::border_size(), build_options, arm_compute::calculate_max_window(), arm_compute::quantization::calculate_quantized_multiplier(), arm_compute::CHANNEL, arm_compute::misc::shape_calculator::compute_deep_convolution_shape(), arm_compute::test::validation::conv_info, conv_stride_x, conv_stride_y, arm_compute::create_kernel(), ITensorInfo::data_layout(), arm_compute::test::validation::data_type, ITensorInfo::data_type(), ITensorInfo::dimension(), ActivationLayerInfo::enabled(), arm_compute::experimental::dynamic_fusion::export_to_cl_image_support(), arm_compute::F32, arm_compute::float_to_string_with_full_precision(), PixelValue::get(), arm_compute::get_cl_type_from_data_type(), arm_compute::get_data_layout_dimension_index(), arm_compute::get_data_size_from_data_type(), ICLKernel::get_target(), arm_compute::HEIGHT, arm_compute::is_data_type_quantized(), kernel_name, arm_compute::lower_string(), arm_compute::NCHW, arm_compute::NHWC, UniformQuantizationInfo::offset, CLBuildOptions::options(), arm_compute::test::validation::output_shape, PadStrideInfo::pad_left(), PadStrideInfo::pad_top(), ITensorInfo::quantization_info(), arm_compute::S32, UniformQuantizationInfo::scale, PadStrideInfo::stride(), arm_compute::string_from_activation_func(), arm_compute::string_from_data_layout(), arm_compute::string_from_data_type(), ITensorInfo::tensor_shape(), arm_compute::support::cpp11::to_string(), arm_compute::utils::cast::U, QuantizationInfo::uniform(), arm_compute::opencl::kernels::gemm::update_padding_for_cl_image(), arm_compute::cpu::kernels::validate_arguments(), and arm_compute::WIDTH.

177 {
179 
180  // Perform validation
181  ARM_COMPUTE_ERROR_THROW_ON(validate_arguments(src, weights, biases, dst, conv_info, act_info));
182 
183  const int conv_stride_x = std::get<0>(conv_info.stride());
184  const int conv_stride_y = std::get<1>(conv_info.stride());
185 
186  _data_layout = src->data_layout();
188 
189  const unsigned int width_idx = get_data_layout_dimension_index(_data_layout, DataLayoutDimension::WIDTH);
190  const unsigned int height_idx = get_data_layout_dimension_index(_data_layout, DataLayoutDimension::HEIGHT);
191  const unsigned int channel_idx = get_data_layout_dimension_index(_data_layout, DataLayoutDimension::CHANNEL);
192  const unsigned int kernel_size = weights->dimension(width_idx);
193  const DataType data_type = src->data_type();
194 
195  const GPUTarget gpu_target = get_target();
196  unsigned int _num_elems_processed_per_iteration = 0;
197 
198  // Get dst shape
200 
201  // Output auto inizialitation if not yet initialized
202  auto_init_if_empty(*dst, output_shape,
203  1,
204  src->data_type(),
205  src->quantization_info());
206 
207  // Configure kernel window
208  Window win;
209  if(_data_layout == DataLayout::NHWC)
210  {
211  const unsigned int vec_size = std::min(static_cast<unsigned int>(dst->tensor_shape()[0]), 4u);
212  unsigned int num_rows = 1U;
213  if(dst->tensor_shape()[0] > 16)
214  {
215  num_rows = src->data_type() == DataType::F32 ? 2U : 4U;
216  }
217 
218  // Create window and update padding
219  win = calculate_max_window(output_shape, Steps(vec_size, num_rows));
220  }
221  else if(_data_layout == DataLayout::NCHW)
222  {
223  _num_elems_processed_per_iteration = 1u;
224  win = calculate_max_window(*dst, Steps(_num_elems_processed_per_iteration));
225  }
226 
227  ICLKernel::configure_internal(win);
228 
229  std::stringstream kernel_name;
230  CLBuildOptions build_options;
231 
232  if(_data_layout == DataLayout::NHWC)
233  {
234  kernel_name << "direct_convolution_nhwc";
235 
236  const unsigned int n0 = win.x().step();
237  const unsigned int m0 = win.y().step();
238  const unsigned int k0 = adjust_vec_size(is_data_type_quantized(data_type) ? 16u : 8u, src->dimension(channel_idx));
239  const unsigned int partial_store_n0 = dst->dimension(channel_idx) % n0;
240  const unsigned int pad_left = conv_info.pad_left();
241  const unsigned int pad_top = conv_info.pad_top();
242  const bool export_to_cl_image = export_to_cl_image_support(weights, gpu_target, _data_layout);
243 
244  // Update the padding for the weights tensor if we can export to cl_image
245  if(export_to_cl_image)
246  {
248  }
249 
250  if(biases != nullptr)
251  {
252  build_options.add_option(std::string("-DHAS_BIAS"));
253  build_options.add_option(std::string("-DBIA_DATA_TYPE=" + get_cl_type_from_data_type(biases->data_type())));
254  }
255 
256  build_options.add_option("-cl-fast-relaxed-math");
257  build_options.add_option("-DSRC_TENSOR_TYPE=BUFFER");
258  build_options.add_option("-DSRC_DATA_TYPE=" + get_cl_type_from_data_type(src->data_type()));
259  build_options.add_option("-DDST_TENSOR_TYPE=BUFFER");
260  build_options.add_option("-DDST_DATA_TYPE=" + get_cl_type_from_data_type(dst->data_type()));
261  build_options.add_option_if_else(export_to_cl_image, "-DWEI_TENSOR_TYPE=IMAGE", "-DWEI_TENSOR_TYPE=BUFFER");
262  build_options.add_option("-DWEI_WIDTH=" + support::cpp11::to_string(weights->dimension(width_idx)));
263  build_options.add_option("-DWEI_HEIGHT=" + support::cpp11::to_string(weights->dimension(height_idx)));
264  build_options.add_option("-DWEI_DATA_TYPE=" + get_cl_type_from_data_type(weights->data_type()));
265  build_options.add_option("-DSTRIDE_X=" + support::cpp11::to_string(conv_stride_x));
266  build_options.add_option("-DSTRIDE_Y=" + support::cpp11::to_string(conv_stride_y));
267  build_options.add_option("-DPAD_LEFT=" + support::cpp11::to_string(pad_left));
268  build_options.add_option("-DPAD_TOP=" + support::cpp11::to_string(pad_top));
269  build_options.add_option("-DN0=" + support::cpp11::to_string(n0));
270  build_options.add_option("-DM0=" + support::cpp11::to_string(m0));
271  build_options.add_option("-DK0=" + support::cpp11::to_string(k0));
272  build_options.add_option("-DPARTIAL_N0=" + support::cpp11::to_string(partial_store_n0));
273  build_options.add_option_if((src->dimension(channel_idx) % k0) != 0, "-DLEFTOVER_LOOP");
274  build_options.add_option("-DACTIVATION_TYPE=" + lower_string(string_from_activation_func(act_info.activation())));
275 
276  if(is_data_type_quantized(data_type))
277  {
278  const UniformQuantizationInfo iqinfo = src->quantization_info().uniform();
279  const UniformQuantizationInfo wqinfo = weights->quantization_info().uniform();
280  const UniformQuantizationInfo oqinfo = dst->quantization_info().uniform();
281 
282  PixelValue zero_value = PixelValue(0, src->data_type(), src->quantization_info());
283  int zero_value_s32;
284  zero_value.get(zero_value_s32);
285 
286  float multiplier = iqinfo.scale * wqinfo.scale / oqinfo.scale;
287  int output_multiplier = 0;
288  int output_shift = 0;
289  quantization::calculate_quantized_multiplier(multiplier, &output_multiplier, &output_shift);
290  build_options.add_option("-DIS_QUANTIZED");
291  build_options.add_option("-DDST_MULTIPLIER=" + support::cpp11::to_string(output_multiplier));
292  build_options.add_option("-DDST_SHIFT=" + support::cpp11::to_string(output_shift));
293  build_options.add_option("-DSRC_OFFSET=" + support::cpp11::to_string(-iqinfo.offset));
294  build_options.add_option("-DWEI_OFFSET=" + support::cpp11::to_string(-wqinfo.offset));
295  build_options.add_option("-DDST_OFFSET=" + support::cpp11::to_string(oqinfo.offset));
296  build_options.add_option("-DZERO_VALUE=" + support::cpp11::to_string(zero_value_s32));
297  build_options.add_option("-DACC_DATA_TYPE=" + get_cl_type_from_data_type(DataType::S32));
298  }
299  else
300  {
301  build_options.add_option("-DACC_DATA_TYPE=" + get_cl_type_from_data_type(data_type));
302  build_options.add_option("-DZERO_VALUE=" + support::cpp11::to_string(0));
303  build_options.add_option("-DSRC_OFFSET=" + support::cpp11::to_string(0));
304  build_options.add_option("-DWEI_OFFSET=" + support::cpp11::to_string(0));
305  build_options.add_option("-DDST_OFFSET=" + support::cpp11::to_string(0));
306  build_options.add_option_if(act_info.enabled(), "-DA_VAL=" + float_to_string_with_full_precision(act_info.a()));
307  build_options.add_option_if(act_info.enabled(), "-DB_VAL=" + float_to_string_with_full_precision(act_info.b()));
308  }
309  }
310  else
311  {
312  kernel_name << "direct_convolution_nchw";
313  build_options.add_option_if(biases != nullptr, std::string("-DHAS_BIAS"));
314  build_options.add_option("-DSRC_WIDTH=" + support::cpp11::to_string(src->dimension(width_idx)));
315  build_options.add_option("-DSRC_HEIGHT=" + support::cpp11::to_string(src->dimension(height_idx)));
316  build_options.add_option("-DSRC_CHANNELS=" + support::cpp11::to_string(src->dimension(channel_idx)));
317  build_options.add_option("-DPAD_LEFT=" + support::cpp11::to_string(conv_info.pad_left()));
318  build_options.add_option("-DPAD_TOP=" + support::cpp11::to_string(conv_info.pad_top()));
319  build_options.add_option("-DSTRIDE_X=" + support::cpp11::to_string(conv_stride_x));
320  build_options.add_option("-DSTRIDE_Y=" + support::cpp11::to_string(conv_stride_y));
321  build_options.add_option("-DWEI_WIDTH=" + support::cpp11::to_string(weights->dimension(width_idx)));
322  build_options.add_option("-DWEI_HEIGHT=" + support::cpp11::to_string(weights->dimension(height_idx)));
323  build_options.add_option(std::string("-DDATA_TYPE=" + get_cl_type_from_data_type(data_type)));
324  build_options.add_option(std::string("-DDATA_SIZE=" + get_data_size_from_data_type(data_type)));
325  build_options.add_option(std::string("-DWEIGHTS_DEPTH=" + support::cpp11::to_string(weights->dimension(channel_idx))));
326  build_options.add_option(std::string("-DSTRIDE_X=" + support::cpp11::to_string(conv_stride_x)));
327  build_options.add_option(std::string("-DDATA_TYPE_PROMOTED=" + get_cl_type_from_data_type(data_type)));
328  build_options.add_option(std::string("-DVEC_SIZE=" + support::cpp11::to_string(_num_elems_processed_per_iteration)));
329  build_options.add_option(std::string("-DVEC_SIZE_LEFTOVER=" + support::cpp11::to_string(src->dimension(0) % _num_elems_processed_per_iteration)));
330 
331  if(is_data_type_quantized(data_type))
332  {
333  const UniformQuantizationInfo iqinfo = src->quantization_info().uniform();
334  const UniformQuantizationInfo wqinfo = weights->quantization_info().uniform();
335  const UniformQuantizationInfo oqinfo = dst->quantization_info().uniform();
336 
337  float multiplier = iqinfo.scale * wqinfo.scale / oqinfo.scale;
338  int output_multiplier = 0;
339  int output_shift = 0;
340  quantization::calculate_quantized_multiplier(multiplier, &output_multiplier, &output_shift);
341  build_options.add_option("-DIS_QUANTIZED");
342  build_options.add_option("-DOUTPUT_MULTIPLIER=" + support::cpp11::to_string(output_multiplier));
343  build_options.add_option("-DOUTPUT_SHIFT=" + support::cpp11::to_string(output_shift));
344  build_options.add_option("-DKERNEL_SIZE=" + support::cpp11::to_string(kernel_size));
345  build_options.add_option("-DINPUT_OFFSET=" + support::cpp11::to_string(-iqinfo.offset));
346  build_options.add_option("-DWEIGHTS_OFFSET=" + support::cpp11::to_string(-wqinfo.offset));
347  build_options.add_option("-DOUTPUT_OFFSET=" + support::cpp11::to_string(oqinfo.offset));
348  }
349  }
350 
351  _kernel = create_kernel(compile_context, kernel_name.str(), build_options.options());
352 
353  // Set config_id for enabling LWS tuning
354  _config_id = kernel_name.str();
355  _config_id += "_";
356  _config_id += lower_string(string_from_data_type(data_type));
357  _config_id += "_";
358  _config_id += support::cpp11::to_string(kernel_size);
359  _config_id += "_";
360  _config_id += support::cpp11::to_string(border_size().left);
361  _config_id += "_";
362  _config_id += support::cpp11::to_string(border_size().top);
363  _config_id += "_";
364  _config_id += support::cpp11::to_string(border_size().right);
365  _config_id += "_";
366  _config_id += support::cpp11::to_string(border_size().bottom);
367  _config_id += "_";
368  _config_id += support::cpp11::to_string(conv_stride_x);
369  _config_id += "_";
370  _config_id += support::cpp11::to_string(conv_stride_y);
371  _config_id += "_";
372  _config_id += support::cpp11::to_string(dst->dimension(width_idx));
373  _config_id += "_";
374  _config_id += support::cpp11::to_string(dst->dimension(height_idx));
375  _config_id += "_";
376  _config_id += lower_string(string_from_data_layout(_data_layout));
377 }
bool is_data_type_quantized(DataType dt)
Check if a given data type is of quantized type.
Definition: Utils.h:1030
Window calculate_max_window(const ValidRegion &valid_region, const Steps &steps, bool skip_border, BorderSize border_size)
bool export_to_cl_image_support(const ITensorInfo *tensor, GPUTarget gpu_target, DataLayout data_layout)
std::string to_string(T &&value)
Convert integer and float values to string.
1 channel, 1 F32 per channel
const std::string & string_from_activation_func(ActivationLayerInfo::ActivationFunction act)
Translates a given activation function to a string.
Definition: Utils.cpp:163
#define ARM_COMPUTE_ERROR_THROW_ON(status)
Definition: Error.h:455
Status calculate_quantized_multiplier(float multiplier, int32_t *quant_multiplier, int32_t *shift, bool ignore_epsilon=false)
Calculate quantized representation of multiplier.
std::string lower_string(const std::string &val)
Lower a given string.
Definition: Utils.cpp:351
Status validate_arguments(const ITensorInfo *src, const ITensorInfo *weights, const ITensorInfo *dst, const PadStrideInfo &conv_info)
std::set< std::string > build_options
void update_padding_for_cl_image(ITensorInfo *tensor)
Update padding required to export the OpenCL buffer to OpenCL image2d.
SimpleTensor< float > src
Definition: DFT.cpp:155
1 channel, 1 S32 per channel
const size_t conv_stride_y
Definition: impl.cpp:58
cl::Kernel create_kernel(const CLCompileContext &ctx, const std::string &kernel_name, const std::set< std::string > &build_opts=std::set< std::string >())
Creates an opencl kernel using a compile context.
Definition: CLHelpers.cpp:391
const std::string & string_from_data_type(DataType dt)
Convert a data type identity into a string.
Definition: Utils.cpp:135
std::string get_data_size_from_data_type(const DataType &dt)
Get the size of a data type in number of bits.
Definition: CLHelpers.cpp:193
std::string float_to_string_with_full_precision(float val)
Create a string with the float in full precision.
Definition: Utils.h:1124
GPUTarget get_target() const
Get the targeted GPU architecture.
Definition: ICLKernel.h:444
std::string get_cl_type_from_data_type(const DataType &dt)
Translates a tensor data type to the appropriate OpenCL type.
Definition: CLHelpers.cpp:39
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...
virtual BorderSize border_size() const
The size of the border for that kernel.
Definition: IKernel.cpp:46
Num samples, channels, height, width.
const std::string & string_from_data_layout(DataLayout dl)
Convert a data layout identity into a string.
Definition: Utils.cpp:123
GPUTarget
Available GPU Targets.
Definition: GPUTarget.h:34
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
Num samples, height, width, channels.
#define ARM_COMPUTE_ERROR_ON_NULLPTR(...)
Definition: Validate.h:157
unsigned int adjust_vec_size(unsigned int vec_size, size_t dim0)
Returns the adjusted vector size in case it is less than the input&#39;s first dimension, getting rounded down to its closest valid vector size.
Definition: Utils.h:1222
std::string kernel_name
DataType
Available data types.
Definition: Types.h:79
const size_t conv_stride_x
Definition: impl.cpp:57
TensorShape compute_deep_convolution_shape(const TensorShape &input_shape, DataLayout input_data_layout, const TensorShape &weights_shape, const PadStrideInfo &conv_info)
Calculate the deep convolution shape output shape of a tensor.

◆ run_op()

void run_op ( ITensorPack tensors,
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]tensorsA vector containing the tensors to operato on.
[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.

Reimplemented from ICLKernel.

Definition at line 386 of file ClDirectConv2dKernel.cpp.

References ClDirectConv2dKernel::_data_layout, arm_compute::ACL_DST, arm_compute::ACL_SRC_0, arm_compute::ACL_SRC_1, arm_compute::ACL_SRC_2, ICLKernel::add_1D_tensor_argument(), ICLKernel::add_3D_tensor_argument(), ICLKernel::add_4d_tensor_nhwc_argument(), ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW, ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL, arm_compute::ceil_to_multiple(), arm_compute::create_image2d_from_buffer(), Window::DimY, Window::DimZ, arm_compute::enqueue(), arm_compute::experimental::dynamic_fusion::export_to_cl_image_support(), Window::first_slice_window_3D(), CLKernelLibrary::get(), ITensorPack::get_const_tensor(), ICLKernel::get_target(), ITensorPack::get_tensor(), ICLKernel::lws_hint(), arm_compute::NHWC, ICLKernel::num_arguments_per_3D_tensor(), Window::set(), arm_compute::test::validation::reference::slice(), Window::slide_window_slice_3D(), Window::Dimension::step(), Window::use_tensor_dimensions(), IKernel::window(), and Window::y().

387 {
390 
391  // Get initial windows
393 
394  const auto src = utils::cast::polymorphic_downcast<const ICLTensor *>(tensors.get_const_tensor(TensorType::ACL_SRC_0));
395  const auto weights = utils::cast::polymorphic_downcast<const ICLTensor *>(tensors.get_const_tensor(TensorType::ACL_SRC_1));
396  const auto biases = utils::cast::polymorphic_downcast<const ICLTensor *>(tensors.get_const_tensor(TensorType::ACL_SRC_2));
397  auto dst = utils::cast::polymorphic_downcast<ICLTensor *>(tensors.get_tensor(TensorType::ACL_DST));
398 
400  {
401  cl::Image2D weights_cl_image;
402 
403  const size_t dim_y_collapsed = ceil_to_multiple(dst->info()->dimension(1) * dst->info()->dimension(2), slice.y().step());
404  const bool export_to_cl_image = export_to_cl_image_support(weights->info(), get_target(), _data_layout);
405 
406  slice.set(Window::DimY, Window::Dimension(0, dim_y_collapsed, slice.y().step()));
407  slice.set(Window::DimZ, Window::Dimension(0, dst->info()->dimension(3), 1));
408 
409  if(export_to_cl_image)
410  {
411  const size_t image_w = weights->info()->dimension(0) / 4;
412  const size_t image_h = weights->info()->dimension(1) * weights->info()->dimension(2) * weights->info()->dimension(3);
413  const TensorShape shape2d(image_w, image_h);
414  const size_t image_row_pitch = weights->info()->strides_in_bytes()[1];
415 
416  // Export cl_buffer to cl_image
417  weights_cl_image = create_image2d_from_buffer(CLKernelLibrary::get().context(), weights->cl_buffer(), shape2d, weights->info()->data_type(), image_row_pitch);
418  }
419 
420  unsigned int idx = 0;
422  add_4d_tensor_nhwc_argument(idx, dst);
423  if(export_to_cl_image)
424  {
425  _kernel.setArg(idx++, weights_cl_image);
426  }
427  add_4d_tensor_nhwc_argument(idx, weights);
428  if(biases != nullptr)
429  {
430  add_1D_tensor_argument(idx, biases, slice);
431  }
432  enqueue(queue, *this, slice, lws_hint());
433  }
434  else
435  {
436  unsigned int idx1 = 2 * num_arguments_per_3D_tensor();
437  add_3D_tensor_argument(idx1, weights, slice);
438 
439  if(biases != nullptr)
440  {
441  Window slice_biases;
442  slice_biases.use_tensor_dimensions(biases->info()->tensor_shape());
443  add_1D_tensor_argument(idx1, biases, slice_biases);
444  }
445 
446  _kernel.setArg(idx1++, static_cast<unsigned int>(weights->info()->strides_in_bytes()[3]));
447 
448  do
449  {
450  unsigned int idx = 0;
451  add_3D_tensor_argument(idx, src, slice);
452  add_3D_tensor_argument(idx, dst, slice);
453  enqueue(queue, *this, slice, lws_hint());
454  }
455  while(window.slide_window_slice_3D(slice));
456  }
457 }
void add_4d_tensor_nhwc_argument(unsigned int &idx, const ICLTensor *tensor)
Add the passed NHWC 4D tensor&#39;s parameters to the object&#39;s kernel&#39;s arguments by passing strides...
Definition: ICLKernel.cpp:144
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:32
bool export_to_cl_image_support(const ITensorInfo *tensor, GPUTarget gpu_target, DataLayout data_layout)
cl::NDRange lws_hint() const
Return the Local-Workgroup-Size hint.
Definition: ICLKernel.h:384
static CLKernelLibrary & get()
Access the KernelLibrary singleton.
void add_3D_tensor_argument(unsigned int &idx, const ICLTensor *tensor, const Window &window)
Add the passed 3D tensor&#39;s parameters to the object&#39;s kernel&#39;s arguments starting from the index idx...
Definition: ICLKernel.h:227
SimpleTensor< float > src
Definition: DFT.cpp:155
static constexpr unsigned int num_arguments_per_3D_tensor()
Returns the number of arguments enqueued per 3D tensor object.
Definition: ICLKernel.h:314
auto ceil_to_multiple(S value, T divisor) -> decltype(((value+divisor - 1)/divisor) *divisor)
Computes the smallest number larger or equal to value that is a multiple of divisor.
Definition: Utils.h:71
GPUTarget get_target() const
Get the targeted GPU architecture.
Definition: ICLKernel.h:444
bool slide_window_slice_3D(Window &slice) const
Slide the passed 3D window slice.
Definition: Window.h:348
#define ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL(k)
Definition: Validate.h:915
static constexpr size_t DimY
Alias for dimension 1 also known as Y dimension.
Definition: Window.h:45
static constexpr size_t DimZ
Alias for dimension 2 also known as Z dimension.
Definition: Window.h:47
Num samples, height, width, channels.
cl::Image2D create_image2d_from_buffer(const cl::Context &ctx, const cl::Buffer &buffer, const TensorShape &shape2d, DataType data_type, size_t image_row_pitch)
Create a cl::Image2D object from an OpenCL buffer.
Definition: CLUtils.cpp:35
void add_1D_tensor_argument(unsigned int &idx, const ICLTensor *tensor, const Window &window)
Add the passed 1D tensor&#39;s parameters to the object&#39;s kernel&#39;s arguments starting from the index idx...
Definition: ICLKernel.h:179
Window first_slice_window_3D() const
First 3D slice of the window.
Definition: Window.h:304
#define ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW(f, s)
Definition: Validate.h:201
SimpleTensor< T > slice(const SimpleTensor< T > &src, Coordinates starts, Coordinates ends)

◆ validate()

Status validate ( const ITensorInfo src,
const ITensorInfo weights,
const ITensorInfo biases,
const ITensorInfo dst,
const PadStrideInfo conv_info,
const ActivationLayerInfo act_info 
)
static

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

Similar to ClDirectConv2dKernel::configure()

Returns
a status

Definition at line 379 of file ClDirectConv2dKernel.cpp.

References ARM_COMPUTE_RETURN_ON_ERROR, and arm_compute::cpu::kernels::validate_arguments().

Referenced by ClDirectConv2d::validate().

381 {
382  ARM_COMPUTE_RETURN_ON_ERROR(validate_arguments(src, weights, biases, dst, conv_info, act_info));
383  return Status{};
384 }
#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 *src, const ITensorInfo *weights, const ITensorInfo *dst, const PadStrideInfo &conv_info)
SimpleTensor< float > src
Definition: DFT.cpp:155

Field Documentation

◆ _conv_info

PadStrideInfo _conv_info {}

Definition at line 82 of file ClDirectConv2dKernel.h.

Referenced by ClDirectConv2dKernel::configure().

◆ _data_layout

DataLayout _data_layout {}

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