import collections import contextlib import dataclasses import functools import os import re from itertools import count from typing import Any, Dict, List, Optional, Tuple import sympy from sympy import Expr import torch from torch._dynamo.utils import counters, dynamo_timed from torch.fx.experimental.symbolic_shapes import SymTypes from torch.fx.node import _get_qualified_name from .. import codecache, config, ir from ..codecache import CudaKernelParamCache from ..utils import ( cache_on_self, get_benchmark_name, LineContext, sympy_dot, sympy_product, ) from ..virtualized import V from .common import CodeGen, DeferredLine, IndentedBuffer, PythonPrinter pexpr = PythonPrinter().doprint def buffer_reuse_key(node: ir.Buffer): size = node.get_size() stride = node.get_stride() last_element = sympy_dot([s - 1 for s in size], stride) return ( node.get_device(), node.get_dtype(), V.graph.sizevars.simplify(sympy_product(size)), # Detect gaps in tensor storage caused by strides V.graph.sizevars.size_hint(last_element), ) def is_int(s: str): try: int(s) except ValueError: return False return True def is_float(s: str): try: float(s) except ValueError: return False return True def convert_arg_type(python_type): from .cpp import CONTAINER_PYTHON_TO_CPP, PYTHON_TO_CPP if python_type == "Tensor": # Conversions rules follow https://github.com/pytorch/pytorch/tree/main/aten/src/ATen/native#func return f"at::{python_type} const&" if python_type in PYTHON_TO_CPP: return PYTHON_TO_CPP[python_type] # Convert args of container types e.g. Optional[*] for py_container, cpp_container in CONTAINER_PYTHON_TO_CPP.items(): container_match = re.findall(py_container + r"\[([a-zA-Z_]+)]", python_type) if len(container_match) == 1: contained_type = container_match[0] assert ( contained_type in PYTHON_TO_CPP ), f"unsupported {py_container} type in convert_arg_type: {contained_type}" cpp_contained_type = PYTHON_TO_CPP[contained_type] return f"{cpp_container}<{cpp_contained_type}>" raise AssertionError(f"unsupport python_type: {python_type}") def convert_return_type(python_type): # TODO: only support Tensor as func return type for now # TODO: support alias assert ( python_type == "Tensor" ), f"only support tensor output for cpp_wrapper, but receive type {python_type}" return f"at::{python_type}" def get_cpp_op_schema(kernel): # use x.real_type instead of x.type so that we get ScalarType instead of int arg_types = [repr(x.real_type) for x in kernel._schema.arguments] arg_names = [x.name for x in kernel._schema.arguments] # TODO: only support len(returns) == 1 for now. returns = [repr(x.type) for x in kernel._schema.returns] assert ( len(returns) == 1 ), f"only support 1 single output for cpp_wrapper, but {kernel.__name__} has {len(returns)} outputs" return_value = returns[0] cpp_return_value = convert_return_type(return_value) cpp_arg_type = [ f"{convert_arg_type(arg_type)} {arg_name}" for arg_type, arg_name in zip(arg_types, arg_names) ] return f"{cpp_return_value}({', '.join(cpp_arg_type)})" @dataclasses.dataclass class SymbolicCallArg: inner: Any def __str__(self): return str(self.inner) class MemoryPlanningState: def __init__(self): super().__init__() self.reuse_pool: Dict[Any, List[FreeIfNotReusedLine]] = collections.defaultdict( list ) def __contains__(self, key): return bool(self.reuse_pool.get(key, None)) def pop(self, key) -> "FreeIfNotReusedLine": item = self.reuse_pool[key].pop() assert not item.is_reused return item def push(self, key, item: "FreeIfNotReusedLine"): assert not item.is_reused self.reuse_pool[key].append(item) @dataclasses.dataclass class EnterCudaDeviceContextManagerLine: device_idx: int first_time: bool def codegen(self, code: IndentedBuffer, device_cm_stack: contextlib.ExitStack): if V.graph.cpp_wrapper: code.writeline("\n") if V.graph.aot_mode: # In AOT mode, we have a stream provided as a param. A stream is # associated with a device, so we never expect the device to change. assert self.first_time # CUDAStreamGuard sets the stream and the device. code.writeline( f"at::cuda::CUDAStreamGuard stream_guard(" f"at::cuda::getStreamFromExternal(stream, {self.device_idx}));" ) else: if self.first_time: code.writeline( f"at::cuda::CUDAGuard device_guard({self.device_idx});" ) else: code.writeline(f"device_guard.set_index({self.device_idx});") else: # Note _DeviceGuard has less overhead than device, but only accepts # integers code.writeline(f"with torch.cuda._DeviceGuard({self.device_idx}):") device_cm_stack.enter_context(code.indent()) code.writeline( f"torch.cuda.set_device({self.device_idx}) # no-op to ensure context" ) class ExitCudaDeviceContextManagerLine: def codegen(self, code: IndentedBuffer, device_cm_stack: contextlib.ExitStack): if not V.graph.cpp_wrapper: device_cm_stack.close() @dataclasses.dataclass class MemoryPlanningLine: wrapper: "WrapperCodeGen" def plan(self, state: MemoryPlanningState) -> "MemoryPlanningLine": """First pass to find reuse""" return self def codegen(self, code: IndentedBuffer): """Second pass to output code""" pass @dataclasses.dataclass class AllocateLine(MemoryPlanningLine): node: ir.Buffer can_reuse: bool = True def plan(self, state: MemoryPlanningState): if self.node.get_name() in V.graph.removed_buffers: return NullLine(self.wrapper) # try to reuse a recently freed buffer key = buffer_reuse_key(self.node) if config.allow_buffer_reuse and key in state and self.can_reuse: free_line = state.pop(key) free_line.is_reused = True return ReuseLine(self.wrapper, free_line.node, self.node) return self def codegen(self, code: IndentedBuffer): assert self.node.get_name() not in V.graph.removed_buffers line = self.wrapper.make_buffer_allocation(self.node) code.writeline(line) @dataclasses.dataclass class FreeIfNotReusedLine(MemoryPlanningLine): node: ir.Buffer is_reused: bool = False def plan(self, state: MemoryPlanningState): assert not self.is_reused if self.node.get_name() in V.graph.removed_buffers: return NullLine(self.wrapper) if config.allow_buffer_reuse: state.push(buffer_reuse_key(self.node), self) return self def codegen(self, code: IndentedBuffer): assert self.node.get_name() not in V.graph.removed_buffers if not self.is_reused: code.writeline(self.wrapper.make_buffer_free(self.node)) @dataclasses.dataclass class ReuseLine(MemoryPlanningLine): node: ir.Buffer reused_as: ir.Buffer def plan(self, state: MemoryPlanningState): if self.node.get_name() in V.graph.removed_buffers: assert self.reused_as.get_name() in V.graph.removed_buffers return NullLine(self.wrapper) assert self.reused_as.get_name() not in V.graph.removed_buffers return self def codegen(self, code: IndentedBuffer): assert self.node.get_name() not in V.graph.removed_buffers assert self.reused_as.get_name() not in V.graph.removed_buffers code.writeline( self.wrapper.make_buffer_reuse( self.node, self.reused_as, ) ) class NullLine(MemoryPlanningLine): pass class WrapperCodeGen(CodeGen): """ Generate outer wrapper in Python that calls the kernels. """ def __init__(self): super().__init__() self._names_iter = count() self.header = IndentedBuffer() self.prefix = IndentedBuffer() self.wrapper_call = IndentedBuffer() self.src_to_kernel = {} self.kenel_numel_expr = set() self.lines = [] self.declare = "" self.ending = "" self.open_bracket = "[" self.closed_bracket = "]" self.comment = "#" self.namespace = "" self.none_str = "None" self.optional_tensor_str = "None" self.size = "size()" self.stride = "stride()" self.first_device_guard = True self.supports_intermediate_hooks = True self.expr_printer = pexpr self.write_header() self.write_prefix() for name, hashed in V.graph.constant_reprs.items(): # include a hash so our code cache gives different constants different files self.write_constant(name, hashed) self.allocated = set() self.freed = set() # maps from reusing buffer to reused buffer self.reuses = dict() self.write_get_cuda_stream = functools.lru_cache(None)( # type: ignore[assignment] self.write_get_cuda_stream ) @functools.lru_cache(None) def add_import_once(line): self.header.writeline(line) self.add_import_once = add_import_once self._metas = {} def write_constant(self, name, hashed): self.header.writeline(f"{name} = None # {hashed}") def write_header(self): self.header.splice( f""" from ctypes import c_void_p, c_long import torch import math import random import os import tempfile from math import inf, nan from torch._inductor.hooks import run_intermediate_hooks from torch._inductor.utils import maybe_profile from torch import empty_strided, device from {codecache.__name__} import AsyncCompile from torch._inductor.select_algorithm import extern_kernels aten = torch.ops.aten assert_size_stride = torch._C._dynamo.guards.assert_size_stride reinterpret_tensor = torch.ops.inductor._reinterpret_tensor async_compile = AsyncCompile() """ ) @cache_on_self def write_triton_header_once(self): self.header.splice( """ import triton import triton.language as tl from torch._inductor.triton_heuristics import grid, start_graph, end_graph from torch._C import _cuda_getCurrentRawStream as get_cuda_stream """ ) def add_meta_once(self, meta): meta = repr(meta) if meta not in self._metas: var = f"meta{len(self._metas)}" self._metas[meta] = var self.header.writeline(f"{var} = {meta}") return self._metas[meta] @cache_on_self def get_output_refs(self): return [x.codegen_reference() for x in V.graph.graph_outputs] def mark_output_type(self): return def codegen_input_size_asserts(self): for name, buf in V.graph.graph_inputs.items(): if isinstance(buf, sympy.Expr): continue # comparing strides for 0 size tensor is tricky. Ignore them for now. if sympy_product(buf.get_size()) == 0: continue size = self.codegen_shape_tuple(buf.get_size()) stride = self.codegen_shape_tuple(buf.get_stride()) self.prefix.writeline(f"assert_size_stride({name}, {size}, {stride})") def write_prefix(self): self.prefix.splice( """ async_compile.wait(globals()) del async_compile def call(args): """ ) with self.prefix.indent(): if config.triton.debug_sync_graph: self.prefix.writeline("torch.cuda.synchronize()") inp_len = len(V.graph.graph_inputs.keys()) if inp_len != 0: lhs = f"{', '.join(V.graph.graph_inputs.keys())}{'' if inp_len != 1 else ','}" self.prefix.writeline(f"{lhs} = args") self.prefix.writeline("args.clear()") self.codegen_inputs(self.prefix, V.graph.graph_inputs) if config.size_asserts: self.codegen_input_size_asserts() def write_get_cuda_stream(self, index): self.write_triton_header_once() name = f"stream{index}" self.writeline(f"{name} = get_cuda_stream({index})") return name def next_kernel_suffix(self): return f"{next(self._names_iter)}" def codegen_device_guard_enter(self, device_idx): self.writeline( EnterCudaDeviceContextManagerLine(device_idx, self.first_device_guard) ) self.first_device_guard = False def codegen_device_guard_exit(self): self.writeline(ExitCudaDeviceContextManagerLine()) def generate_return(self, output_refs): if output_refs: self.wrapper_call.writeline("return (" + ", ".join(output_refs) + ", )") else: self.wrapper_call.writeline("return ()") def generate_end(self, result): return def generate_extern_kernel_alloc(self, output_name, kernel, args, origin_node): self.writeline( f"{self.declare}{output_name} = {kernel}({', '.join(args)}){self.ending}" ) if ( self.supports_intermediate_hooks and config.generate_intermediate_hooks and origin_node is not None ): counters["inductor"]["intermediate_hooks"] += 1 self.writeline( f"run_intermediate_hooks({origin_node.name!r}, {output_name})" ) def generate_extern_kernel_out(self, output_view, codegen_reference, args, kernel): if output_view: args.append(f"out={output_view.codegen_reference()}") else: args.append(f"out={codegen_reference}") self.writeline(f"{kernel}({', '.join(args)})") def generate_scatter_fallback( self, output, inputs, kernel, fn, src_is_tensor, reduce, kwargs ): line = f"{kernel}({','.join(map(str, inputs))}" if kernel == "aten.scatter_": if reduce: line += f", reduce={repr(reduce)}" else: line += ", ".join([""] + kwargs) line += f"){self.ending}" self.writeline(line) def generate_extern_kernel_alloc_and_find_schema_if_needed( self, name, kernel, codegen_args, cpp_op_schema, cpp_kernel_key, cpp_kernel_overload_name="", ): self.writeline(f"{name} = {kernel}({', '.join(codegen_args)})") @dynamo_timed def generate(self): result = IndentedBuffer() result.splice(self.header) out_names = V.graph.get_output_names() with contextlib.ExitStack() as stack: stack.enter_context(self.wrapper_call.indent()) if config.profiler_mark_wrapper_call: self.generate_profiler_mark_wrapper_call(stack) if config.profile_bandwidth: self.write_triton_header_once() self.wrapper_call.writeline("start_graph()") while ( self.lines and isinstance(self.lines[-1], MemoryPlanningLine) # TODO: this seems legit, NullLine has no node and self.lines[-1].node.name not in out_names # type: ignore[attr-defined] ): # these lines will be pointless self.lines.pop() # codegen allocations in two passes planning_state = MemoryPlanningState() for i in range(len(self.lines)): if isinstance(self.lines[i], MemoryPlanningLine): self.lines[i] = self.lines[i].plan(planning_state) device_cm_stack = contextlib.ExitStack() for line in self.lines: if isinstance(line, MemoryPlanningLine): line.codegen(self.wrapper_call) elif isinstance( line, ( EnterCudaDeviceContextManagerLine, ExitCudaDeviceContextManagerLine, ), ): line.codegen(self.wrapper_call, device_cm_stack) else: self.wrapper_call.writeline(line) output_refs = self.get_output_refs() self.mark_output_type() if config.triton.debug_sync_graph: self.wrapper_call.writeline("torch.cuda.synchronize()") if config.profile_bandwidth: self.wrapper_call.writeline("end_graph()") self.generate_return(output_refs) self.append_precomputed_sizes_to_prefix() result.splice(self.prefix) with result.indent(): result.splice(self.wrapper_call) self.generate_end(result) self.add_benchmark_harness(result) return result.getvaluewithlinemap() def codegen_inputs(self, code: IndentedBuffer, graph_inputs: Dict[str, ir.Buffer]): """Assign all symbolic shapes to locals""" @functools.lru_cache(None) def sizeof(name): code.writeline( f"{self.declare}{name}_size = {name}.{self.size}{self.ending}" ) return f"{name}_size" @functools.lru_cache(None) def strideof(name): code.writeline( f"{self.declare}{name}_stride = {name}.{self.stride}{self.ending}" ) return f"{name}_stride" # Assign all symbolic shapes needed to local variables needed = set(V.graph.sizevars.var_to_val.keys()) - set( V.graph.sizevars.replacements.keys() ) def is_expr(x): return isinstance(x[1], sympy.Expr) graph_inputs_expr = list(filter(is_expr, graph_inputs.items())) graph_inputs_tensors = list( filter(lambda x: not is_expr(x), graph_inputs.items()) ) for name, shape in graph_inputs_expr: shape = V.graph.sizevars.simplify(shape) if shape in needed: needed.remove(shape) code.writeline(f"{self.declare}{shape} = {name}{self.ending}") for name, value in graph_inputs_tensors: shapes = value.get_size() for dim, shape in enumerate(shapes): shape = V.graph.sizevars.simplify(shape) if shape in needed: needed.remove(shape) code.writeline( f"{self.declare}{shape} = {sizeof(name)}[{dim}]{self.ending}" ) for name, value in graph_inputs_tensors: shapes = value.get_stride() for dim, shape in enumerate(shapes): shape = V.graph.sizevars.simplify(shape) if shape in needed: needed.remove(shape) code.writeline( f"{self.declare}{shape} = {strideof(name)}[{dim}]{self.ending}" ) def append_precomputed_sizes_to_prefix(self): with self.prefix.indent(): for sym, expr in V.graph.sizevars.inv_precomputed_replacements.items(): self.prefix.writeline( f"{self.declare}{sym} = {self.expr_printer(expr)}{self.ending}" ) def codegen_python_sizevar(self, x: Expr) -> str: return pexpr(V.graph.sizevars.simplify(x)) def codegen_sizevar(self, x: Expr) -> str: return self.codegen_python_sizevar(x) def codegen_tuple_access(self, basename: str, index: str) -> str: return f"{basename}[{index}]" def codegen_python_shape_tuple(self, shape: Tuple[Expr, ...]) -> str: parts = list(map(self.codegen_python_sizevar, shape)) if len(parts) == 0: return "()" if len(parts) == 1: return f"({parts[0]}, )" return f"({', '.join(parts)})" def codegen_shape_tuple(self, shape: Tuple[Expr, ...]) -> str: return self.codegen_python_shape_tuple(shape) def benchmark_compiled_module(self, output): def add_fake_input(name, shape, stride, device, dtype): output.writeline( f"{name} = rand_strided(" f"{self.codegen_python_shape_tuple(shape)}, " f"{self.codegen_python_shape_tuple(stride)}, " f"device='{device}', dtype={dtype})" ) def add_expr_input(name, val): output.writeline(f"{name} = {val}") output.writelines( ["", "", "def benchmark_compiled_module(times=10, repeat=10):"] ) with output.indent(): output.splice( """ from torch._dynamo.testing import rand_strided from torch._inductor.utils import print_performance """, strip=True, ) for name, value in V.graph.constants.items(): # all the constants are global variables, that's why we need # these 'global var_name' lines output.writeline(f"global {name}") add_fake_input( name, value.size(), value.stride(), value.device, value.dtype ) for name, value in V.graph.graph_inputs.items(): if isinstance(value, sympy.Expr): # Don't need to add symbolic add_expr_input(name, V.graph.sizevars.size_hint(value)) else: shape = [V.graph.sizevars.size_hint(x) for x in value.get_size()] stride = [V.graph.sizevars.size_hint(x) for x in value.get_stride()] add_fake_input( name, shape, stride, value.get_device(), value.get_dtype() ) call_str = f"call([{', '.join(V.graph.graph_inputs.keys())}])" output.writeline( f"return print_performance(lambda: {call_str}, times=times, repeat=repeat)" ) def add_benchmark_harness(self, output): """ Append a benchmark harness to generated code for debugging """ if not config.benchmark_harness: return self.benchmark_compiled_module(output) output.writelines(["", "", 'if __name__ == "__main__":']) with output.indent(): output.writelines( [ "from torch._inductor.wrapper_benchmark import compiled_module_main", f"compiled_module_main('{get_benchmark_name()}', benchmark_compiled_module)", ] ) def define_kernel( self, name: str, kernel: str, metadata: Optional[str] = None, cuda=True ): metadata_comment = f"{metadata}\n" if metadata else "" self.header.splice(f"\n\n{metadata_comment}{name} = {kernel}") def generate_numel_expr(self, kernel_name: str, tree): expr = f"{kernel_name}_{tree.prefix}numel" if expr not in self.kenel_numel_expr: self.kenel_numel_expr.add(expr) self.writeline( f"{self.declare}{expr} = {self.expr_printer(tree.numel)}{self.ending}" ) else: self.writeline(f"{expr} = {self.expr_printer(tree.numel)}{self.ending}") # We can get symbolic expressions here, like s0*64 # It is fine to have them here, but we need to handle them correctly as their own type # This is tricky to do, so we wrap in a custom type, distinct from scalars, but also from sympy* # scalars as well. # This is handled in `generate_args_decl` which has a correct comment of: TODO: only works for # constant now, need type info. I agree, this needs type info, and while this is not true type info # it suffices as a type hint for the purposes of producing the correct code for this type. return SymbolicCallArg(expr) def wrap_kernel_call(self, name, call_args): return f"{name}({', '.join(call_args)}){self.ending}" def generate_profiler_mark_wrapper_call(self, stack): self.wrapper_call.writeline("from torch.profiler import record_function") self.wrapper_call.writeline( f"with record_function('graph_{V.graph.graph_id}_inductor_wrapper_call'):" ) stack.enter_context(self.wrapper_call.indent()) def generate_kernel_call( self, name, call_args, grid=None, device_index=None, cuda=True ): if cuda: call_args_str = ", ".join(pexpr(item) for item in call_args) grid_str = ", ".join(pexpr(item) for item in grid) stream_name = self.write_get_cuda_stream( V.graph.scheduler.current_device.index ) self.writeline( f"{name}.run({call_args_str}, grid=grid({grid_str}), stream={stream_name})" ) else: self.writeline(self.wrap_kernel_call(name, call_args)) def writeline(self, line): self.lines.append(line) def enter_context(self, ctx): self.lines.append(LineContext(ctx)) def val_to_arg_str(self, s): if isinstance(s, SymTypes): return pexpr(sympy.expand(repr(s))) elif isinstance(s, sympy.Expr): return pexpr(s) elif isinstance(s, (tuple, list)): @dataclasses.dataclass class Shim: ref: Any def __repr__(self): return self.ref return repr(type(s)(Shim(self.val_to_arg_str(a)) for a in s)) elif isinstance(s, torch._ops.OpOverload): return _get_qualified_name(s) else: return repr(s) # The following methods are for memory management def make_buffer_allocation(self, buffer): device = buffer.get_device() dtype = buffer.get_dtype() shape = tuple(buffer.get_size()) stride = tuple(buffer.get_stride()) return ( f"{buffer.get_name()} = empty_strided(" f"{self.codegen_shape_tuple(shape)}, " f"{self.codegen_shape_tuple(stride)}, " f"device='{device.type}', dtype={dtype})" ) def make_buffer_free(self, buffer): return f"del {buffer.get_name()}" def make_buffer_reuse(self, old, new): assert old.get_dtype() == new.get_dtype() del_line = "" if old.get_name() not in V.graph.get_output_names(): del_line = f"; {self.make_buffer_free(old)}" if old.get_size() == new.get_size() and old.get_stride() == new.get_stride(): return f"{self.declare}{new.get_name()} = {old.get_name()}{del_line} {self.comment} reuse" return ( f"{self.declare}{new.get_name()} = reinterpret_tensor(" f"{old.get_name()}, " f"{self.codegen_shape_tuple(new.get_size())}, " f"{self.codegen_shape_tuple(new.get_stride())}){del_line} {self.comment} reuse" ) def codegen_deferred_allocation(self, name, layout): self.writeline( DeferredLine( name, f"{self.declare}{name} = {layout.view.codegen_reference()}{self.ending} {self.comment} alias", ) ) def use_preallocated_ouput(self, buffer): # outputs are passed-in in the AOT mode return ( V.graph.aot_mode and buffer and buffer.get_name() in set(V.graph.get_output_names()) ) def codegen_allocation(self, buffer): name = buffer.get_name() if name in V.graph.removed_buffers or name in self.allocated: return self.allocated.add(name) if isinstance( buffer, (ir.ExternKernelAlloc, ir.MultiOutput), ): return layout = buffer.get_layout() if isinstance(layout, ir.MutationLayout): return if isinstance(layout, ir.AliasedLayout): assert isinstance( layout.view, ir.ReinterpretView ), f"unexpected {type(layout.view)}: {layout.view}" if not layout.maybe_guard_aligned(): V.graph.unaligned_buffers.add(name) self.codegen_allocation(layout.view.data) self.codegen_deferred_allocation(name, layout) return self.writeline( AllocateLine( self, buffer, not self.use_preallocated_ouput(buffer), ) ) def codegen_free(self, buffer): name = buffer.get_name() # can be freed but not reused if isinstance(buffer, ir.InputBuffer): self.writeline(self.make_buffer_free(buffer)) return if not self.can_reuse(buffer): return self.freed.add(name) layout = buffer.get_layout() if isinstance(layout, (ir.AliasedLayout, ir.MultiOutputLayout)): self.writeline(self.make_buffer_free(buffer)) return self.writeline(FreeIfNotReusedLine(self, buffer)) def can_reuse(self, input_buffer, output_buffer=None): name = input_buffer.get_name() if ( name in V.graph.removed_buffers or name in V.graph.graph_inputs or name in V.graph.constants or name in self.freed or self.use_preallocated_ouput(output_buffer) ): return False return True def did_reuse(self, buffer, reused_buffer): # Check whether a given buffer was reused by a possible reuser in the wrapper codegen # Can be consulted from inside ir codegen, e.g. to determine whether a copy is needed return ( buffer.get_name() in self.reuses and self.reuses[buffer.get_name()] == reused_buffer.get_name() ) def codegen_inplace_reuse(self, input_buffer, output_buffer): assert buffer_reuse_key(input_buffer) == buffer_reuse_key(output_buffer) self.codegen_allocation(input_buffer) self.freed.add(input_buffer.get_name()) self.allocated.add(output_buffer.get_name()) self.reuses[output_buffer.get_name()] = input_buffer.get_name() self.writeline(ReuseLine(self, input_buffer, output_buffer)) class CppWrapperCodeGen(WrapperCodeGen): """ Generates cpp wrapper for running on CPU and calls cpp kernels """ def __init__(self): super().__init__() from ..ir import OptionalTensor self.declare = "auto " self.ending = ";" self.open_bracket = "{" self.closed_bracket = "}" self.comment = "//" self.namespace = "at::" self.none_str = "at::Tensor()" self.optional_tensor_str = repr(OptionalTensor()) self.extern_call_ops = set() self.size = "sizes()" self.stride = "strides()" self.call_func_name = "inductor_entry_cpp" self.cuda = False self.supports_intermediate_hooks = False self.outputs_need_copy = set() self.resized_outputs = {} from .cpp import cexpr self.expr_printer = cexpr def write_constant(self, name, hashed): # include a hash so our code cache gives different constants different files self.header.writeline(f"// {name} {hashed}") def write_header(self): if V.graph.aot_mode: with open( os.path.join(os.path.dirname(__file__), "aot_inductor_interface.cpp") ) as f: self.header.splice(f.read()) else: self.header.splice( """ import torch from torch._inductor.codecache import CppWrapperCodeCache cpp_wrapper_src = ( ''' """ ) self.header.splice( """ #include #define reinterpret_tensor torch::inductor::_reinterpret_tensor """ ) def mark_output_type(self): # mark output type to unwrap tensor back to python scalar from ..ir import ShapeAsConstantBuffer output_is_tensor = dict() for idx, x in enumerate(V.graph.graph_outputs): if isinstance(x, ShapeAsConstantBuffer): output_is_tensor[idx] = False else: output_is_tensor[idx] = True self.output_is_tensor = output_is_tensor def write_prefix(self): if V.graph.aot_mode: self.prefix.writeline("namespace torch {") self.prefix.writeline("namespace aot_inductor {") def write_wrapper_decl(self): inputs_len = len(V.graph.graph_inputs.keys()) if V.graph.aot_mode: self.prefix.splice( """ void AOTInductorModel::run_impl( const std::vector& args, std::vector& outputs, cudaStream_t stream) { """ ) else: self.prefix.splice( f"""std::vector {self.call_func_name}(const std::vector& args) {{""" ) with self.prefix.indent(): if inputs_len != 0: for idx, input_key in enumerate(V.graph.graph_inputs.keys()): # unwrap input tensor back to scalar if isinstance(V.graph.graph_inputs[input_key], sympy.Expr): from ..graph import may_get_constant_buffer_dtype from .cpp import DTYPE_TO_CPP dtype = may_get_constant_buffer_dtype( V.graph.graph_inputs[input_key] ) assert ( dtype is not None ), "Fails to get the dtype of the sympy.Expr" cpp_dtype = DTYPE_TO_CPP[dtype] self.prefix.writeline( f"{cpp_dtype} {input_key} = args[{idx}].item<{cpp_dtype}>();" ) else: self.prefix.writeline(f"at::Tensor {input_key} = args[{idx}];") assert all( isinstance(v, torch.Tensor) for v in list(V.graph.constants.values()) ), "Expect all constants to be Tensor" for idx, constants_key in enumerate(V.graph.constants.keys()): constants_idx = inputs_len + idx self.prefix.writeline( f"at::Tensor {constants_key} = args[{constants_idx}];" ) self.codegen_inputs(self.prefix, V.graph.graph_inputs) self.wrapper_call.splice( """ c10::optional optional_scalar; c10::optional optional_string; c10::optional optional_layout; c10::optional optional_tensor; torch::List> optional_list; """ ) def codegen_model_constructor(self): """ // Generated code example AOTInductorModel::AOTInductorModel() : AOTInductorModelBase(4, 1) { inputs_info_[0].name = "linear.weight"; inputs_info_[0].shape.reserve(2); inputs_info_[0].shape.emplace_back(10, 10, nullptr); inputs_info_[0].shape.emplace_back(64, 64, nullptr); ... outputs_info_[0].name = "output0"; outputs_info_[0].shape.reserve(2); outputs_info_[0].shape.emplace_back(32, 32, nullptr); outputs_info_[0].shape.emplace_back(10, 10, nullptr); } """ num_inputs = len(V.graph.graph_inputs) num_outputs = len(V.graph.graph_outputs) self.prefix.splice( f""" AOTInductorModel::AOTInductorModel() : AOTInductorModelBase({num_inputs}, {num_outputs}) {{ """ ) with self.prefix.indent(): for idx, name in enumerate(V.graph.graph_inputs.keys()): # TODO: handle symbolic expressions later. assert not isinstance(V.graph.graph_inputs[name], sympy.Expr) self.prefix.writeline(f"""inputs_info_[{idx}].name = "{name}";""") self.prefix.writeline( f"""inputs_info_[{idx}].dtype = "{V.graph.graph_inputs[name].get_dtype()}";""" ) sizes = V.graph.graph_inputs[name].get_size() self.prefix.writeline( f"inputs_info_[{idx}].shape.reserve({len(sizes)});" ) for size in sizes: # FIXME: set the lower bound and the upper bound to be "size". # Later, we should specify the correct range for dynamic dimentions. self.prefix.writeline( f"inputs_info_[{idx}].shape.emplace_back({size}, {size}, nullptr);" ) for idx, output in enumerate(V.graph.graph_outputs): # TODO: handle symbolic expressions later. assert not isinstance(output, sympy.Expr) self.prefix.writeline(f"""outputs_info_[{idx}].name = "output{idx}";""") self.prefix.writeline( f"""outputs_info_[{idx}].dtype = "{output.get_dtype()}";""" ) sizes = output.get_size() self.prefix.writeline( f"outputs_info_[{idx}].shape.reserve({len(sizes)});" ) for size in sizes: # FIXME: set the lower bound and the upper bound to be "size". # Later, we should specify the correct range for dynamic dimentions. self.prefix.writeline( f"outputs_info_[{idx}].shape.emplace_back({size}, {size}, nullptr);" ) self.prefix.writeline("}") def generate(self): if V.graph.aot_mode: self.codegen_model_constructor() self.write_wrapper_decl() return super().generate() def define_kernel( self, name: str, kernel: str, metadata: Optional[str] = None, cuda=False ): self.header.splice(f"\n{kernel}\n") def generate_return(self, output_refs): # Output tensors are allocated by the AOT runtime. if V.graph.aot_mode: for idx, output in enumerate(V.graph.graph_outputs): if hasattr(output, "get_name"): name = output.get_name() if name in self.outputs_need_copy: output_as_strided = output.codegen_reference() self.wrapper_call.writeline( f"outputs[{idx}].copy_({output_as_strided});" ) resize_to = self.resized_outputs.get(name, None) if resize_to is not None: resize_to_args = ", ".join( self.expr_printer(d) for d in resize_to ) self.wrapper_call.writeline( f"outputs[{idx}].resize_({{{resize_to_args}}});" ) self.wrapper_call.writeline("\n}") else: self.wrapper_call.writeline(f"return {{{', '.join(output_refs)}}};\n}}") def generate_end(self, result): if V.graph.aot_mode: result.writeline("} // namespace aot_inductor") result.writeline("} // namespace inductor") return result.writeline("'''\n)") # get the hash of the wrapper code to name the extension wrapper_call_hash = codecache.code_hash(result.getvalue()) result.splice( f""" module = CppWrapperCodeCache.load(cpp_wrapper_src, '{self.call_func_name}', '{wrapper_call_hash}', {self.cuda}) """ ) # unwrap output tensor back to python scalar if all(x for x in self.output_is_tensor.values()): # If no ShapeAsConstantBuffer in the output, directly return the output as tensors return_str = "return f(args_tensor)" else: outputs = [ f"outputs[{i}]" if self.output_is_tensor[i] else f"outputs[{i}].item()" for i in range(len(V.graph.graph_outputs)) ] outputs_str = f"[{', '.join(outputs)}]" return_str = f""" outputs = f(args_tensor) return {outputs_str} """ args_str = "args_tensor = [arg if isinstance(arg, torch.Tensor) else torch.tensor(arg) for arg in args]" if V.graph.constants: # Append constants to the input args for cpp wrapper. # Python wrapper directly gets the value inside the wrapper call # as a global variable passed when calling exec(code, mod.__dict__, mod.__dict__). # For cpp wrapper, we need to pass this python value to the inductor_entry_cpp function explicitly. assert all( isinstance(v, torch.Tensor) for v in list(V.graph.constants.values()) ), "Expect all constants to be Tensor" constants_str = f"[{', '.join(V.graph.constants.keys())}]" args_str += f""" constants_tensor = {constants_str} args_tensor.extend(constants_tensor) """ # Wrap the func to support setting result._boxed_call = True result.splice( f""" def _wrap_func(f): def g(args): {args_str} {return_str} return g call = _wrap_func(module.{self.call_func_name}) """ ) def generate_extern_kernel_out(self, output_view, codegen_reference, args, kernel): if output_view: output_as_strided = f"{output_view.codegen_reference()}" output_name = f"{output_view.get_name()}_as_strided" self.writeline(f"auto {output_name} = {output_as_strided};") args.insert(0, output_name) else: args.insert(0, f"{codegen_reference}") self.writeline(self.wrap_kernel_call(kernel, args)) def generate_scatter_fallback( self, output, inputs, kernel, fn, src_is_tensor, reduce, kwargs ): # TODO: support other overload for cpp wrapper and remove the below assertions line = f"{kernel}({output}, {','.join(map(str, inputs))}" if fn == "aten.scatter_": if src_is_tensor: if reduce: line += f", {V.graph.wrapper_code.val_to_arg_str(reduce)}" else: assert ( reduce is None ), "Expect reduce to be None for aten.scatter_ with scalar src" else: line += f", {','.join(kwargs)}" line += f"){self.ending}" self.writeline(line) def add_benchmark_harness(self, output): if V.graph.aot_mode: return super().add_benchmark_harness(output) def codegen_sizevar(self, x: Expr) -> str: return self.expr_printer(V.graph.sizevars.simplify(x)) def codegen_tuple_access(self, basename: str, index: str) -> str: return f"std::get<{index}>({basename})" def codegen_shape_tuple(self, shape: Tuple[Expr, ...]) -> str: parts = list(map(self.codegen_sizevar, shape)) if len(parts) == 0: return "{}" if len(parts) == 1: return f"{{{parts[0]}, }}" return f"{{{', '.join(parts)}}}" def make_buffer_free(self, buffer): return ( "" if isinstance(buffer.get_layout(), ir.MultiOutputLayout) else f"{buffer.get_name()}.reset();" ) def generate_profiler_mark_wrapper_call(self, stack): self.wrapper_call.writeline( 'RECORD_FUNCTION("inductor_wrapper_call", c10::ArrayRef());' ) def codegen_device(self, device): from .cpp import DEVICE_TO_ATEN return ( f"c10::Device({DEVICE_TO_ATEN[device.type]}, {device.index})" if device.index is not None else f"{DEVICE_TO_ATEN[device.type]}" ) def codegen_tensor_option(self, device, dtype): from .cpp import DTYPE_TO_ATEN cpp_device = self.codegen_device(device) return f"at::TensorOptions({cpp_device}).dtype({DTYPE_TO_ATEN[dtype]}))" def make_buffer_allocation(self, buffer): name = buffer.get_name() # outputs are passed-in in the AOT mode if self.use_preallocated_ouput(buffer): output_idx = None output_buffer = None for idx, output in enumerate(V.graph.graph_outputs): if hasattr(output, "get_name") and name == output.get_name(): output_idx = idx output_buffer = output break assert ( output_idx is not None and output_buffer is not None ), "Unknown output index" if V.graph.sizevars.statically_known_leq( buffer.get_numel(), output_buffer.get_numel() ): buf_str = f"auto {name} = outputs[{output_idx}];" # avoid resize_output warning: # "An output with one or more elements was resized since it had..." if buffer.get_size() != output_buffer.get_size(): resize_to_args = ", ".join( self.expr_printer(d) for d in buffer.get_size() ) buf_str += f" {name}.resize_({{{resize_to_args}}});" assert name not in self.resized_outputs self.resized_outputs[name] = list(output_buffer.get_size()) return buf_str else: self.outputs_need_copy.add(name) # TODO: map layout here. device = buffer.get_device() dtype = buffer.get_dtype() shape = tuple(buffer.get_size()) stride = tuple(buffer.get_stride()) return ( f"{self.declare}{name} = {self.namespace}empty_strided(" f"{self.codegen_shape_tuple(shape)}, " f"{self.codegen_shape_tuple(stride)}, " f"{self.codegen_tensor_option(device, dtype)};" ) def generate_extern_kernel_alloc_and_find_schema_if_needed( self, name, kernel, codegen_args, cpp_op_schema, cpp_kernel_key, cpp_kernel_overload_name="", ): if cpp_kernel_key not in self.extern_call_ops: self.writeline( f"static auto op_{cpp_kernel_key} = c10::Dispatcher::singleton()" ) self.writeline( f'\t.findSchemaOrThrow("{kernel}", "{cpp_kernel_overload_name}")' ) self.writeline(f"\t.typed<{cpp_op_schema}>();") self.extern_call_ops.add(cpp_kernel_key) self.writeline( f"auto {name} = op_{cpp_kernel_key}.call({', '.join(codegen_args)});" ) def val_to_arg_str(self, val): from .cpp import DTYPE_TO_ATEN if val is None: # When None is passed as an argument, it represents an optional that does not contain a value. return self.optional_tensor_str elif isinstance(val, bool): return "true" if val else "false" elif isinstance(val, str): return f'"{val}"' elif isinstance(val, torch.device): return self.codegen_device(val) elif isinstance(val, torch.dtype): return DTYPE_TO_ATEN[val] elif isinstance(val, float) and val in [float("inf"), float("-inf")]: if val == float("inf"): return "std::numeric_limits::infinity()" else: return "-std::numeric_limits::infinity()" elif isinstance(val, (list, tuple)): return f"{{{', '.join(list(map(self.val_to_arg_str, val)))}}}" else: return repr(val) class CudaWrapperCodeGen(CppWrapperCodeGen): """ Generates cpp wrapper for running on GPU and calls CUDA kernels """ def __init__(self): super().__init__() self.kernel_callsite_id = count() self.arg_var_id = count() self.cuda = True def write_header(self): super().write_header() self.header.splice( """ #include #include #include #include #define AT_CUDA_DRIVER_CHECK_OVERRIDE(EXPR) \\ do { \\ CUresult __err = EXPR; \\ if (__err != CUDA_SUCCESS) { \\ AT_ERROR("CUDA driver error: ", static_cast(__err)); \\ } \\ } while (0) static inline CUfunction loadKernel( const std::string &filePath, const std::string &funcName, int sharedMemBytes) { CUmodule mod; CUfunction func; AT_CUDA_DRIVER_CHECK_OVERRIDE(cuModuleLoad(&mod, filePath.c_str())); AT_CUDA_DRIVER_CHECK_OVERRIDE(cuModuleGetFunction(&func, mod, funcName.c_str())); if (sharedMemBytes > 0) { AT_CUDA_DRIVER_CHECK_OVERRIDE(cuFuncSetAttribute( func, CU_FUNC_ATTRIBUTE_MAX_DYNAMIC_SHARED_SIZE_BYTES, sharedMemBytes )); } return func; } static inline void launchKernel( CUfunction func, int gridX, int gridY, int gridZ, int numWarps, int sharedMemBytes, void* args[], cudaStream_t stream) { AT_CUDA_DRIVER_CHECK_OVERRIDE(cuLaunchKernel( func, gridX, gridY, gridZ, 32*numWarps, 1, 1, sharedMemBytes, stream, args, nullptr)); } """ ) def write_get_cuda_stream(self, index): name = f"stream{index}" self.writeline( f"cudaStream_t {name} = at::cuda::getCurrentCUDAStream({index});" ) return name def define_kernel( self, name: str, kernel: str, metadata: Optional[str] = None, cuda=True ): if not cuda: return super().define_kernel(name, kernel, metadata, cuda) def generate(self): self.prefix.writeline("\n") for kernel in self.src_to_kernel.values(): self.prefix.writeline(f"static CUfunction {kernel} = nullptr;") self.prefix.writeline("\n") return super().generate() def generate_load_kernel(self, name, params): mangled_name = params.get("mangled_name", None) assert mangled_name is not None, "missing mangled_name" cubin_path = params.get("cubin_path", None) assert os.path.exists( cubin_path ), "cubin file should already exist at this moment" shared_mem = params.get("shared_mem", 0) self.writeline(f"if ({name} == nullptr) {{") self.writeline( f""" {name} = loadKernel("{cubin_path}", "{mangled_name}", {shared_mem});""" ) self.writeline("}") def generate_args_decl(self, call_args): # TODO: only works for constant now, need type info new_args = [] for arg in call_args: var_name = f"var_{next(self.arg_var_id)}" if isinstance( arg, ( sympy.Integer, sympy.Symbol, SymbolicCallArg, ), ): self.writeline(f"auto {var_name} = {arg};") elif is_int(arg): self.writeline(f"int {var_name} = {arg};") elif is_float(arg): self.writeline(f"float {var_name} = {arg};") else: self.writeline( f"CUdeviceptr {var_name} = reinterpret_cast({arg}.data_ptr());" ) new_args.append(f"&{var_name}") return ", ".join(new_args) def generate_kernel_call( self, name, call_args, grid=None, device_index=None, cuda=True ): if not cuda: return super().generate_kernel_call( name, call_args, grid, device_index, cuda ) params = CudaKernelParamCache.get(name) assert ( params is not None ), f"cuda kernel parameters for {name} should already exist at this moment" self.generate_load_kernel(name, params) call_args = self.generate_args_decl(call_args) kernel_args_var = f"kernel_args_var_{next(self.kernel_callsite_id)}" self.writeline(f"void* {kernel_args_var}[] = {{{call_args}}};") stream = ( "stream" if V.graph.aot_mode else self.write_get_cuda_stream(device_index) ) self.writeline( "launchKernel({}, {}, {}, {}, {}, {}, {}, {});".format( name, params["grid_x"], params["grid_y"], params["grid_z"], params["num_warps"], params["shared_mem"], kernel_args_var, stream, ) )