57 constexpr
unsigned int num_elems_read_per_iteration = 16;
58 constexpr
unsigned int num_rows_read_pes_iteration = 3;
69 ICLKernel::configure_internal(win);
unsigned int top
top of the border
Window calculate_max_window(const ValidRegion &valid_region, const Steps &steps, bool skip_border, BorderSize border_size)
Container for 2D border size.
1 channel, 1 U8 per channel
BorderSize border_size() const override
The size of the border for that kernel.
static CLKernelLibrary & get()
Access the KernelLibrary singleton.
Copyright (c) 2017-2021 Arm Limited.
virtual ValidRegion valid_region() const =0
Valid region of the tensor.
Implementation of a rectangular access pattern.
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.
bool update_window_and_padding(Window &win, Ts &&... patterns)
Update window and padding size for each of the access patterns.
Class to describe a number of elements in each dimension.
Implementation of a row access pattern.
virtual ITensorInfo * info() const =0
Interface to be implemented by the child class to return the tensor's metadata.
unsigned int left
left of the border
#define ARM_COMPUTE_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(t, c,...)
void configure(const ICLTensor *input, ICLTensor *output, bool border_undefined)
Initialise the kernel's input and output.
Interface for OpenCL tensor.
Manages all the OpenCL kernels compilation and caching, provides accessors for the OpenCL Context...
Wrapper to configure the Khronos OpenCL C++ header.
unsigned int num_elems_processed_per_iteration
Describe a multidimensional execution window.