GitLab

* [Refactor] Enhance TMA barrier validation and support for additional architectures

* Updated the TMA barrier validation in `inject_tma_barrier.cc` to check for non-empty `barrier_id_to_range_` before raising an error for missing `create_list_of_mbarrier`.
* Refactored architecture checks in `phase.py` to utilize a new constant `SUPPORTED_TMA_ARCHS`, allowing for easier updates and improved readability in the target architecture validation logic.

* Enhance logging in setup.py and refactor TMA architecture checks in phase.py

* Added logging configuration to setup.py, replacing print statements with logger for better traceability.
* Updated download and extraction functions to use logger for status messages.
* Refactored TMA architecture checks in phase.py to utilize the new `have_tma` function for improved clarity and maintainability.
* Introduced support for additional compute capabilities in nvcc.py, including TMA support checks.

* Update documentation for get_target_compute_version to reflect correct GPU compute capability range

* Refactor have_tma function to accept tvm.target.Target instead of compute_version

* Updated the `have_tma` function in nvcc.py to take a `target` parameter, improving clarity and usability.
* Adjusted calls to `have_tma` in phase.py to pass the target directly, enhancing maintainability and consistency in TMA support checks.