pystencils.codegen.config.CreateKernelConfig

pystencils.codegen.config.CreateKernelConfig#

class pystencils.codegen.config.CreateKernelConfig(target=None, jit=None, function_name=None, ghost_layers=None, iteration_slice=None, index_field=None, index_dtype=None, default_dtype=None, allow_double_writes=None, skip_independence_check=None, cpu=None, gpu=None, sycl=None)#

Options for create_kernel.

Methods

__init__([target, jit, function_name, ...])

copy()

Perform a semi-deep copy of this configuration object.

get_jit()

Returns either the user-specified JIT compiler, or infers one from the target if none is given.

get_option(name)

Get the value set for the specified option, or the option's default value if none has been set.

get_target()

is_option_set(name)

override(other)

validate_default_dtype(spec)

validate_index_field(idx_field)

validate_index_type(spec)

Attributes

allow_double_writes

If True, don't check if every field is only written at a single location.

cpu

Options for CPU kernels.

default_dtype

Default numeric data type.

function_name

Name of the generated function

ghost_layers

Specifies the number of ghost layers of the iteration region.

gpu

Options for GPU Kernels.

index_dtype

Data type used for all index calculations.

index_field

Index field for a sparse kernel.

iteration_slice

Specifies the kernel's iteration slice.

jit

Just-in-time compiler used to compile and load the kernel for invocation from the current Python environment.

skip_independence_check

By default the assignment list is checked for read/write independence.

sycl

Options for SYCL kernels.

target

The code generation target.
Parameters:
Contents