tilelang.jit.adapter.cutedsl.adapter¶

Attributes¶

logger

類別¶

CuTeDSLKernelAdapter

Helper class that provides a standard way to create an ABC using

Module Contents¶

tilelang.jit.adapter.cutedsl.adapter.logger¶

class tilelang.jit.adapter.cutedsl.adapter.CuTeDSLKernelAdapter(params, result_idx, target, func_or_mod, host_mod=None, device_mod=None, host_kernel_source=None, device_kernel_source=None, verbose=False, pass_configs=None, compile_flags=None)¶

Bases: tilelang.jit.adapter.base.BaseKernelAdapter

Helper class that provides a standard way to create an ABC using inheritance.

參數:

params (list[tilelang.engine.param.KernelParam])
result_idx (list[int])
target (str | tvm.target.Target)
func_or_mod (tvm.tir.PrimFunc | tilelang.tvm.IRModule)
host_mod (tilelang.tvm.IRModule | None)
device_mod (tilelang.tvm.IRModule | None)
host_kernel_source (str | None)
device_kernel_source (str | None)
verbose (bool)
pass_configs (dict[str, Any] | None)
compile_flags (list[str] | None)

pymodule = None¶

params¶

result_idx¶

host_kernel_source = None¶

device_kernel_source = None¶

param_dtypes¶

param_shapes = []¶

target¶

verbose = False¶

wrapper¶

host_func¶

function_names = None¶

launcher_cpp_code¶

launcher_lib_name¶

lib_generator¶

libpath¶

kernel_global_source = None¶

classmethod from_database(params, result_idx, target, func_or_mod, host_kernel_source, device_kernel_source, kernel_lib_path, verbose=False, pass_configs=None, compile_flags=None)¶

參數:

params (list[tilelang.engine.param.KernelParam])
result_idx (list[int])
target (str)
func_or_mod (tvm.tir.PrimFunc | tilelang.tvm.IRModule)
host_kernel_source (str)
device_kernel_source (str)
kernel_lib_path (str)
verbose (bool)
pass_configs (dict[str, Any] | None)
compile_flags (list[str] | None)

get_kernel_source(kernel_only=True)¶

Get the CUDA kernel source code.

回傳:: The kernel source code, or None if not available
回傳型別:: str | None
參數:: kernel_only (bool)

cleanup()¶

Explicitly cleanup this adapter's CUDA resources.

This method can be called explicitly to immediately release CUDA resources without waiting for garbage collection. Useful in Jupyter notebooks or tests.

Note: This is safe to call multiple times as the C++ implementation is idempotent.

property prim_func: tvm.tir.PrimFunc¶

Returns the primary TIR function from the IR module.

回傳型別:: tvm.tir.PrimFunc