Update base for Update on "[ET Device Support] Parse device info from serialized tensor in tensor_parser"

Parse device info (device_type, device_index) from the serialized ExtraTensorInfo in .pte files into TensorImpl at runtime.
When a tensor's extra_tensor_info contains device annotations (e.g., CUDA), the tensor parser now reads and propagates them to the TensorImpl constructor. Tensors without extra_tensor_info default to CPU/0 for backward compatibility with older PTE files.、

Differential Revision: [D97199497](https://our.internmc.facebook.com/intern/diff/D97199497/)

[ghstack-poisoned]

Author: Gasoonjia

exir/passes/propagate_device_pass.py

@@ -144,7 +144,11 @@ def call(self, graph_module: torch.fx.GraphModule) -> PassResult:
             if node.op == "call_function" and node.target == executorch_call_delegate:
                 lowered_module = _get_lowered_module(graph_module, node)
                 if lowered_module is None:
-                    continue
+                    raise RuntimeError(
+                        f"executorch_call_delegate node '{node.name}' does not reference "
+                        "a valid LoweredBackendModule. The first argument must be a "
+                        "get_attr node pointing to a LoweredBackendModule attribute."
+                    )
 
                 result = _get_target_device_from_compile_specs(lowered_module)
                 if result is None: