* Separate test-linux from build-linux/cuda in GitHub Actions
* Prefer unittest when possible
Co-authored-by: Mike Drob <mdrob@apache.org>
---------
Co-authored-by: Mike Drob <mdrob@apache.org>
* Use async cuda malloc managed with cuda 13
* add pool threshold
* refactor for regular cuda malloc
* load eval gpu for cuda
* remove use of cuda pool, use cuda free async
* fix
* fix
* fix
* fix
* fix + comment