fix issue with lift_fresh_copy when using export + compile (#108243)
Fixes https://github.com/pytorch/pytorch/issues/105327. The problem is that `lift_fresh_copy()`'s functionalization implementation currently assumes that the input is always not functional. This is apparently too limiting: when you have "user" code like this (which can potentially come from exporting a model and then running compile on the resulting graph):
```
tensor_constant0 = torch.tensor(2)
lift_fresh = torch.ops.aten.lift_fresh_copy.default(tensor_constant0)
```
When we run this through AOTAutograd, the first call (torch.tensor(2)) will **already** be lifted into a functional tensor wrapper - so the `lift_fresh_copy` call doesn't need to do any "lifting" anymore - it just needs to do a clone.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/108243
Approved by: https://github.com/albanD
ghstack dependencies: #108081, #108235
