Fix build env used for testing oneMath interface

[antonwolfy]

The PR updates the file with the conda environment used to create build conda env for testing oneMath interfaces:

• forced to use DPC++ and MKL from the latest release 2025.3 (no the latest compiler packages on conda-forge yet)
• forced to install TBB from Intel channel, because MKL is not compatible with the latest tbb from the conda-forge when cmake option MKL_THREADING=tbb_thread used

Alternatively there might an approach to create the env with strict channel priority, now the default one=flexible used, which can't guarantee the packages from high priority channel is installing.

• 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/2646/index.html

[github-actions]

Array API standard conformance tests for dpnp=0.20.0dev0=py313h509198e_29 ran successfully.
Passed: 1227
Failed: 0
Skipped: 9

[coveralls]

coverage: 81.45%. remained the same
when pulling 648d59c on fix-gh-action-onemath
into 574e1ca on master.