Use python 3.14 environment in GitHub workflows

[antonwolfy]

This PR updates the remaining GitHub workflows and conda environments to run under python 3.14, and bumps coveralls to 4.1.0 (which supports 3.14).

Version bumps (3.13/3.12 → 3.14)

• check-onemath.yaml — oneMath interface tests
• conda-package.yml — array API conformance tests (python-ver)
• environments/building_docs.yml — docs build
• environments/upload_cleanup_conda_pkg.yml — conda package upload/cleanup
• environments/coverage.yml — coverage job (was pinned to 3.12)

Coverage job fixes for python 3.14

Bumping the coverage environment from 3.12 to 3.14 surfaced two failures unrelated to dpnp's runtime code — both stem from how coverage instrumentation interacts with 3.14's sys.monitoring:

1. Build failure — undeclared identifier '__Pyx_MonitoringEventTypes_CyGen_count'

Coverage builds compile every extension with CYTHON_TRACE=1 (DPNP_GENERATE_COVERAGE=ON), but Cython line tracing also needs the matching # cython: linetrace=True directive at cythonize time. Three .pyx files were missing it (_slicing.pyx, _stride_utils.pyx, _types.pyx): without the directive, Cython references __Pyx_MonitoringEventTypes_CyGen_count (from the shared generator/coroutine struct) but never defines it. On python ≤3.12 the guarding CYTHON_USE_SYS_MONITORING macro is 0, so the code was unreachable — which is why the previous 3.12 pin hid this; on 3.13+ it becomes reachable and fails to compile. Added the directive to the three files. As a side effect, these files were previously producing no line-coverage data and are now measured.

2. Test failures — test_dlpack.py::TestDLPack::test_invaid_stream[scalar|dictionary|device]

coverage.py 7.9+ makes the sys.monitoring core (sysmon) the default on python 3.14. That core (a) does not support plugins such as the Cython.Coverage plugin used here, and (b) corrupts the pending exception when a C/Cython function raises — turning dpctl's correct TypeError from usm_ndarray.__dlpack__ into a SystemError, so assert_raises(TypeError, ...) fails. Pinned the measurement core back to ctrace via core = "ctrace" in [tool.coverage.run] (pyproject.toml). ctrace supports both the Cython plugin and branch coverage, so no coverage data is lost.

• 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/2922/index.html

[github-actions]

Array API standard conformance tests for dpnp=0.21.0dev2=py314h509198e_29 ran successfully.
Passed: 1372
Failed: 3
Skipped: 5

[vlad-perevezentsev]

LGTM
Thank you @antonwolfy