Backport changelog for 0.19.1 release (#2707)

The PR backports the changelog for the `0.19.1` release to the master
branch.

Author: antonwolfy

CHANGELOG.md

@@ -17,7 +17,6 @@ Also, that release drops support for Python 3.9, making Python 3.10 the minimum
 * Added `dpnp.exceptions` submodule to aggregate the generic exceptions used by dpnp [#2616](https://github.com/IntelPython/dpnp/pull/2616)
 * Added implementation of `dpnp.scipy.special.erfcx` [#2596](https://github.com/IntelPython/dpnp/pull/2596)
 * Added implementation of `dpnp.scipy.special.erfinv` and `dpnp.scipy.special.erfcinv` [#2624](https://github.com/IntelPython/dpnp/pull/2624)
-* Enabled support of Python 3.14 [#2631](https://github.com/IntelPython/dpnp/pull/2631)
 * Added implementation of `dpnp.ndarray.tolist` method [#2652](https://github.com/IntelPython/dpnp/pull/2652)
 * Added implementation of `dpnp.frexp` [#2635](https://github.com/IntelPython/dpnp/pull/2635)
 * Added implementation of `dpnp.ndarray.tofile` method [#2635](https://github.com/IntelPython/dpnp/pull/2635)
@@ -55,6 +54,25 @@ Also, that release drops support for Python 3.9, making Python 3.10 the minimum
 ### Security
 
 
+## [0.19.1] - 2025-11-27
+
+This release achieves `dpnp` compatibility with Python 3.14 and enables distributing `dpnp` packages with the latest Python version.
+This release is compatible with NumPy 2.3.5.
+
+### Added
+
+* Enabled support of Python 3.14 [#2631](https://github.com/IntelPython/dpnp/pull/2631)
+
+### Changed
+
+* Bumped oneMath version up to `v0.9` [#2655](https://github.com/IntelPython/dpnp/pull/2655)
+* Updated CMake config files for DPC++ compiler, oneMKL, oneDPL and TBB components with `2025.3` release changes [#2648](https://github.com/IntelPython/dpnp/pull/2648)
+
+### Fixed
+
+* Updated tests for erf functions to pass with `scipy>=0.16.0` for all integer dtypes [#2669](https://github.com/IntelPython/dpnp/pull/2669)
+
+
 ## [0.19.0] - 2025-10-06
 
 This release introduces a set of new `dpnp.ndarray` methods and SciPy-compatible functions to improve CuPy compatibility.