Use new API to register custom ops for llama model (#2840)

Summary:
Using the following 2 APIs:

* `EXECUTORCH_LIBRARY` to replace the need of a yaml file. With this macro we can directly register a custom kernel into ExecuTorch runtime.
* `WRAP_TO_ATEN` allows custom op authors to use the same kernel for ExecuTorch and PyTorch. This can be helpful during debugging.

Pull Request resolved: https://github.com/pytorch/executorch/pull/2840

Test Plan: Rely on the new CI job `test_llama` with `xnnpack+kv+custom` option.

Reviewed By: kimishpatel

Differential Revision: D55713944

Pulled By: larryliu0820

fbshipit-source-id: eff68926c16a1b2a39f36c73eab1da4e983a25c6