All Numba-dppy tests and examples using the @njit decorator should be used with "with_context"

[reazulhoque]

We need to update all the tests and examples that use:

@njit(parallel={'offload':True})

Following the example: https://github.com/IntelPython/numba-dppy/blob/main/numba_dppy/examples/dppl_with_context.py#L21

List of files that needs to update:

• numba/dppl/examples/pa_examples/test1-3d.py
• numba/dppl/examples/pa_examples/test1.py
• numba/dppl/examples/pa_examples/test1-5d.py
• numba/dppl/examples/pa_examples/test1-4d.py
• numba/dppl/examples/pa_examples/test1-2d.py
• numba/dppl/tests/dppl/test_prange.py
• numba/dppl/tests/dppl/test_numpy_comparison_functions.py
• numba/dppl/tests/dppl/test_dpnp_functions.py
• numba/dppl/tests/dppl/test_numpy_bit_twiddling_functions.py
• numba/dppl/tests/dppl/test_numpy_trigonomteric_functions.py
• numba/dppl/tests/dppl/test_numpy_floating_functions.py
• numba/dppl/tests/dppl/test_dppl_fallback.py
• numba/dppl/tests/dppl/test_parfor_lower_message.py
• numba/dppl/tests/dppl/test_vectorize.py
• numba/dppl/tests/dppl/test_numpy_math_functions.py

@diptorupd @PokhodenkoSA

[PokhodenkoSA]

It will allow to remove offload with this PR: IntelPython/numba#105.
So fixing this issue is preliminary work for the PR.