Declare free-threading support in the sparse extension

[antonwolfy]

dpnp.backend.extensions.sparse._sparse_impl was the only dpnp extension module declared without py::mod_gil_not_used(). On a free-threaded Python build, importing it (through dpnp.scipy.sparse) re-enabled the GIL for the whole process, with a RuntimeWarning. Every other pybind11 extension and every Cython module in dpnp already declares free-threading support.

The extension keeps no shared mutable state beyond its dispatch tables, which are filled once at import, so it can be declared safe like the others.

A new dpnp/tests/test_free_threading.py runs on free-threaded builds only (the 3.14t CI jobs). In a subprocess it imports every compiled dpnp module and checks that the GIL stays disabled, so a future extension that forgets the declaration is caught too. It skips if a dependency has already re-enabled the GIL.

The C++ change was not built locally; the new test was checked locally only for its module discovery (it finds and imports all 25 compiled modules) and its skip paths, so the CI 3.14t jobs are the real check.

• Have you provided a meaningful PR description?
• Have you added a test, reproducer or referred to an issue with a reproducer?
• Have you tested your changes locally for CPU and GPU devices?
• Have you made sure that new changes do not introduce compiler warnings?
• Have you checked performance impact of proposed changes?
• Have you added documentation for your changes, if necessary?
• Have you added your changes to the changelog?

[github-actions]

View rendered docs @ https://intelpython.github.io/dpnp/pull/3101/index.html

[coveralls]

No base build to compare — fix-sparse-free-threading into master

[github-actions]

Array API standard conformance tests for dpnp=0.21.0dev14=np2py314h8d9cdd5_5 ran successfully.
Passed: 1376
Failed: 0
Skipped: 6