numba_dpex.compiler module

class numba_dpex.compiler.DPPYCompiler(typingctx, targetctx, library, args, return_type, flags, locals)

Bases: numba.core.compiler.CompilerBase

DPPY Compiler

define_pipelines(): Child classes override this to customize the pipelines in use.

class numba_dpex.compiler.DPPYFunction(cres): Bases: object

class numba_dpex.compiler.DPPYFunctionTemplate(pyfunc, debug=None)

Bases: object

Unmaterialized dppy function

compile(args)

Compile the function for the given argument types.

Each signature is compiled once by caching the compiled function inside this object.

class numba_dpex.compiler.DPPYKernel(context, sycl_queue, llvm_module, name, argtypes, ordered_arg_access_types=None)

Bases: numba_dpex.compiler.DPPYKernelBase

A OCL kernel object

check_for_invalid_access_type(access_type)

class numba_dpex.compiler.DPPYKernelBase

Bases: object

Define interface for configurable kernels

configure(sycl_queue, global_size, local_size=None): Configure the OpenCL kernel. The local_size can be None

copy()

class numba_dpex.compiler.JitDPPYKernel(func, debug, access_types)

Bases: numba_dpex.compiler.DPPYKernelBase

specialize(argtypes, queue)

numba_dpex.compiler.compile_dppy_func(pyfunc, return_type, args, debug=None)

numba_dpex.compiler.compile_dppy_func_template(pyfunc, debug=None): Compile a DPPYFunctionTemplate

numba_dpex.compiler.compile_kernel(sycl_queue, pyfunc, args, access_types, debug=None)

numba_dpex.compiler.compile_kernel_parfor(sycl_queue, func_ir, args, args_with_addrspaces, debug=None)

numba_dpex.compiler.compile_with_dppy(pyfunc, return_type, args, is_kernel, debug=None)

Compiles with numba-dpex’s pipeline and returns the compiled result.

Args:

pyfunc: The Python function to be compiled. return_type: The Numba type of the return value. args: The list of arguments sent to the Python function. is_kernel (bool): Indicates whether the function is decorated

with @dppy.kernel or not.

debug (bool): Flag to turn debug mode ON/OFF.

Returns:

cres: Compiled result.

Raises:

TypeError: @dppy.kernel does not allow users to return any: value. TypeError is raised when users do.

numba_dpex.compiler.get_ordered_arg_access_types(pyfunc, access_types)