[CI][Deps] Uplift CPU/FPGAEMU RT version to 2024.17.3.0.09

[yanfeng3721]

Uplift oclcpu/oclfpgaemu from 2023.16.10.0.17 to 2024.17.3.0.09

[yanfeng3721]

For 3 regressions

Failed Tests (3):
  SYCL :: Assert/assert_in_kernels.cpp
  SYCL :: Basic/built-ins/marray_geometric.cpp
  SYCL :: KernelCompiler/kernel_compiler_spirv.cpp

Assert/assert_in_kernels.cpp is under investigation, Basic/built-ins/marray_geometric.cpp & KernelCompiler/kernel_compiler_spirv.cpp is expected cpurt version failures for cpurt 2024.17.3.0.09.
@aelovikov-intel, should we add "// XFAIL: cpu" for those expected failures? We can remove it for the next version cpurt uplift.

[aelovikov-intel]

We can remove it for the next version cpurt uplift.

Is the fix ready? If so, why can't we update directly to it? If not, what makes you think the fix will be available during the next uplift and when is it expected to happen?

Or, in other words, I don't see any benefits to this project in uplifting CPURT to a newer version if it has new bugs and doesn't fix a substantially more important bug(s).

[yanfeng3721]

Hi @aelovikov-intel , there are 4 tests can pass and 2 tests fail with oclcpurt 2024.17.3.0.09.

[YuriPlyakhin]

LGTM

[uditagarwal97]

LGTM! Thanks for making the changes.

[dkhaldi]

LGTM

[bader]

Post-commit failure:

e2e-lin (Intel Arc A-Series Graphics with Level Zero, ["Linux", "arc"], --param matrix-xmx8=True ... / Intel Arc A-Series Graphics with Level Zero summary
Failed Tests (1):
SYCL :: ESIMD/api/functional/functions/functions_select_2d_fp_extra.cpp

It doesn't seem to be related to the changes in this PR. The post-commit results for this configuration are quite unstable.

[againull]

This also caused sycl-cts/math_builtin_float_half failure in nightly testing https://github.com/intel/llvm/actions/runs/8715677722/job/23908235101

Reported in internal tracker.