[BE]: Update ruff to 0.4.5 (#126979)

Update ruff to 0.4.5 and addresses some false negatives that have been found in the newer version.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/126979
Approved by: https://github.com/ezyang

.lintrunner.toml

@@ -2128,7 +2128,7 @@ init_command = [
     'python3',
     'tools/linter/adapters/pip_init.py',
     '--dry-run={{DRYRUN}}',
-    'ruff==0.4.4',
+    'ruff==0.4.5',
 ]
 is_formatter = true
 

pyproject.toml

@@ -151,6 +151,9 @@ select = [
 "test/torch_np/numpy_tests/**" = [
     "F821",
 ]
+"test/dynamo/test_debug_utils.py" = [
+    "UP037",
+]
 "test/jit/**" = [
     "PLR0133", # tests require this for JIT
     "PYI",
@@ -163,6 +166,9 @@ select = [
     "RUF015",
     "UP", # We don't want to modify the jit test as they test specify syntax
 ]
+"test/inductor/test_torchinductor.py" = [
+    "UP037",
+]
 # autogenerated #TODO figure out why file level noqa is ignored
 "torch/_inductor/fx_passes/serialized_patterns/**" = ["F401", "F501"]
 "torch/onnx/**" = [

torch/_dynamo/utils.py

@@ -2162,8 +2162,8 @@ class numpy_operator_wrapper:
 def defake(x):
     if not isinstance(x, FakeTensor):
         return x
-    size: "torch._prims_common.ShapeType"
-    stride: "torch._prims_common.StrideType"
+    size: torch._prims_common.ShapeType
+    stride: torch._prims_common.StrideType
     if x._has_symbolic_sizes_strides:
         size = []
         for s in x.size():
@@ -2204,7 +2204,7 @@ def build_checkpoint_variable(**options):
 
     # TODO - This is a temporary situation where we have two versions of
     # checkpointing implementation. We will converge on one and remove the other.
-    activation_checkpoint_op: "torch._ops.HigherOrderOperator" = (
+    activation_checkpoint_op: torch._ops.HigherOrderOperator = (
         higher_order_ops.tag_activation_checkpoint
     )
     if torch._functorch.config.functionalize_rng_ops:

torch/_export/utils.py

@@ -74,7 +74,7 @@ def _check_input_constraints_for_graph(
     # NOTE: export already guarantees that the same symbol is used in metadata
     # for all InputDims related by equality constraints, so we can just unify
     # symbols with given input dimension values to check equality constraints.
-    unification_map: "Dict[sympy.Symbol, Any]" = {}
+    unification_map: Dict[sympy.Symbol, Any] = {}
     for (key_path, arg), node in zip(flat_args_with_path, input_placeholders):
         node_val = node.meta.get("val")
         if isinstance(node_val, FakeTensor):

torch/_inductor/codegen/cpp.py

@@ -3412,7 +3412,7 @@ class CppKernelProxy(CppKernel):
 class OuterLoopFusedKernel(CppKernel):
     def __init__(self, kernel_group):
         super().__init__(kernel_group.args, kernel_group.ws.num_threads)
-        self.inner: List["LoopLevel"] = []
+        self.inner: List[LoopLevel] = []
 
     def decide_parallel_depth(self, max_parallel_depth, threads) -> int:
         kernels_parallel_depth = []

torch/_inductor/codegen/wrapper.py

@@ -430,7 +430,7 @@ class WrapperCodeGen(CodeGen):
         # If the generated source code is exactly the same, reuse the
         # pre-existing kernel for it
         self.src_to_kernel: Dict[str, str] = {}
-        self.kernel_numel_expr: Set[Tuple[str, "GraphLowering"]] = set()
+        self.kernel_numel_expr: Set[Tuple[str, GraphLowering]] = set()
         self.lines: List[Union[MemoryPlanningLine, LineContext]] = []
         self.declare = ""
         self.declare_maybe_reference = ""

torch/_inductor/comms.py

@@ -45,8 +45,8 @@ def raise_comms(
     which is the beginning of the forwards pass. We'll have to either do a special pass for FSDP,
     or we'll want to redo this pass with memory considerations so we handle the FSDP case in a general way.
     """
-    new_order_reversed: List["scheduler.BaseSchedulerNode"] = []
-    cur_comms: List["scheduler.BaseSchedulerNode"] = []
+    new_order_reversed: List[scheduler.BaseSchedulerNode] = []
+    cur_comms: List[scheduler.BaseSchedulerNode] = []
     for snode in reversed(snodes):
         if is_collective(snode.node):
             cur_comms.append(snode)

torch/_inductor/graph.py

@@ -394,7 +394,7 @@ class GraphLowering(torch.fx.Interpreter):
         self.aot_mode = aot_mode
         self.graph_id = graph_id
         self.post_grad_graph_id = next(_post_grad_graph_counter)
-        self.scheduler: "torch._inductor.scheduler.Scheduler" = None  # type: ignore[assignment]
+        self.scheduler: torch._inductor.scheduler.Scheduler = None  # type: ignore[assignment]
         self.nodes_prefer_channels_last = (
             self.find_nodes_prefer_channels_last() if self.layout_opt else set()
         )

torch/_inductor/scheduler.py

@@ -2488,7 +2488,7 @@ class Scheduler:
         if len(possible_fusions) == 0:
             return possible_fusions
         possible_fusions_group_by_priority: Dict[
-            int, List[Tuple["BaseSchedulerNode", "BaseSchedulerNode"]]
+            int, List[Tuple[BaseSchedulerNode, BaseSchedulerNode]]
         ] = {}
 
         for node1, node2 in possible_fusions:

torch/distributed/device_mesh.py

@@ -215,7 +215,7 @@ else:
 
             # private field to pre-generate DeviceMesh's hash
             self._flatten_mesh_list = tuple(self.mesh.flatten().tolist())
-            self._parent_mesh: Optional["DeviceMesh"] = None
+            self._parent_mesh: Optional[DeviceMesh] = None
             self._thread_id = threading.get_ident()
 
             # Skip process group initialization if xla device or init backend is False

torch/export/exported_program.py

@@ -224,7 +224,7 @@ class ExportedProgram:
 
         self._graph_signature: ExportGraphSignature = graph_signature
         self._state_dict: Dict[str, Any] = state_dict
-        self._range_constraints: "Dict[sympy.Symbol, ValueRanges]" = range_constraints
+        self._range_constraints: Dict[sympy.Symbol, ValueRanges] = range_constraints
         assert module_call_graph is not None
         self._module_call_graph: List[ModuleCallEntry] = module_call_graph
         self._example_inputs = example_inputs

torch/fx/passes/split_module.py

@@ -1,5 +1,5 @@
 import inspect
-from typing import Any, Callable, Dict, List, Optional, Set, TYPE_CHECKING
+from typing import Any, Callable, Dict, List, Optional, Set
 from collections import OrderedDict
 import logging
 
@@ -8,8 +8,6 @@ from torch.fx._compatibility import compatibility
 from torch.fx.graph_module import GraphModule
 from torch.fx.node import Node
 
-if TYPE_CHECKING:
-    import sympy  # noqa: F401
 
 __all__ = ["Partition", "split_module"]
 _LOGGER = logging.getLogger(__name__)
@@ -172,9 +170,11 @@ def split_module(
             base_mod_attrs[node.target] = attr_val  # type: ignore[index]
         return base_mod_env, base_mod_attrs
 
+    import sympy
+
     partitions: Dict[str, Partition] = {}
     orig_nodes: Dict[str, Node] = {}
-    symbol_to_node: Dict["sympy.Symbol", Node] = {}
+    symbol_to_node: Dict[sympy.Symbol, Node] = {}
 
     def record_cross_partition_use(
         def_node: Node, use_node: Optional[Node]
@@ -237,8 +237,6 @@ def split_module(
     active_grad = None
     active_autocasts = set()
 
-    import sympy  # noqa: F811
-
     for node in m.graph.nodes:
         if node.op in ["placeholder", "get_attr", "output"]:
             if (

torch/library.py

@@ -71,7 +71,7 @@ class Library:
         self.ns = ns
         self._op_defs: Set[str] = set()
         self._op_impls: Set[str] = set()
-        self._registration_handles: List["torch._library.utils.RegistrationHandle"] = []
+        self._registration_handles: List[torch._library.utils.RegistrationHandle] = []
         self.kind = kind
         self.dispatch_key = dispatch_key
         # Use a finalizer to setup the "destructor" instead of __del__.