Inductor cpp wrapper: support Constant in input (#103496)
## Description
Fix cpp wrapper for models which have constants in the graph inputs.
Python wrapper directly gets the value inside the wrapper call as a global variable passed when calling:
https://github.com/pytorch/pytorch/blob/4081e924a8d701d5201db3e4b4ed2da60b072d30/torch/_inductor/codecache.py#L757
The constants value has been saved in `mod.__dict__` in
https://github.com/pytorch/pytorch/blob/4081e924a8d701d5201db3e4b4ed2da60b072d30/torch/_inductor/graph.py#L874-L875
For cpp wrapper, we need to append constants to the input args, so as to pass this python value to the `inductor_entry_cpp` function explicitly.
### Example
Example of output code for dlrm in TorchBench with this fix:
```py
module = CppWrapperCodeCache.load(cpp_wrapper_src, 'inductor_entry_cpp', 'cfkc6c36t7cggi6mnokrdm5jhesnunjg5xysv3o3x3vaqmzmpe6r', False)
def _wrap_func(f):
def g(args):
args_tensor = [arg if isinstance(arg, torch.Tensor) else torch.tensor(arg) for arg in args]
constants_tensor = [constant0, constant1]
args_tensor.extend(constants_tensor)
return f(args_tensor)
return g
call = _wrap_func(module.inductor_entry_cpp)
```
Pull Request resolved: https://github.com/pytorch/pytorch/pull/103496
Approved by: https://github.com/jgong5, https://github.com/jansel, https://github.com/desertfire
