Merge master into impl_lu_solve_2D

Author: vlad-perevezentsev

CHANGELOG.md

@@ -64,6 +64,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 * Fixed `dpnp.linalg.cond` to always return a real dtype [#2547](https://github.com/IntelPython/dpnp/pull/2547)
 * Resolved the issue in `dpnp.random` functions to allow any value of `size` where each element is castable to `Py_ssize_t` type [#2578](https://github.com/IntelPython/dpnp/pull/2578)
 * Resolved `conda build --test` issue in python 3.9 environment [#2583](https://github.com/IntelPython/dpnp/pull/2583)
+* Fixed tests for the rounding functions to depend on minimum required numpy version [#2589](https://github.com/IntelPython/dpnp/pull/2589)
 
 ### Security
 

dpnp/tests/test_mathematical.py

@@ -2302,6 +2302,7 @@ def test_projection(self, dtype):
 
 @pytest.mark.parametrize("func", ["ceil", "floor", "trunc", "fix"])
 class TestRoundingFuncs:
+    @testing.with_requires("numpy>=2.1.0")
     @pytest.mark.parametrize(
         "dt", get_all_dtypes(no_none=True, no_complex=True)
     )