[FX qconfig] add weighted_op_qint8_dtype_config for int8 TRT and import linear config to get_tensorrt_backend_config_dict (#76877)
Summary:
- Add weighted_op_qint8_dtype_config
- import import linear config to get_tensorrt_backend_config_dict()
- Add qconfig for _get_linear_configs()
Test Plan: CI
Differential Revision: D36152319
Pull Request resolved: https://github.com/pytorch/pytorch/pull/76877
Approved by: https://github.com/jerryzh168
