[GHA] Uplift Linux IGC Dev RT version to igc-dev-b456477

[bb-sycl]

Scheduled igc dev drivers uplift

[aelovikov-intel]

CUDA failures:

Failed Tests (2):
SYCL :: Graph/Update/Explicit/whole_update_double_buffer.cpp
SYCL :: Graph/Update/RecordReplay/whole_update_double_buffer.cpp

Dev IGC update couldn't be directly guilty, but the implicit image rebuild might have caused that if third party base image has changed. @intel/llvm-gatekeepers , @intel/llvm-reviewers-cuda , @intel/sycl-graphs-reviewers , did anybody see those elsewhere?

[EwanC]

CUDA failures:

Failed Tests (2):
SYCL :: Graph/Update/Explicit/whole_update_double_buffer.cpp
SYCL :: Graph/Update/RecordReplay/whole_update_double_buffer.cpp

Dev IGC update couldn't be directly guilty, but the implicit image rebuild might have caused that if third party base image has changed. @intel/llvm-gatekeepers , @intel/llvm-reviewers-cuda , @intel/sycl-graphs-reviewers , did anybody see those elsewhere?

@JackAKirk reported a bug with these 2 tests internally to Codeplay SYCL-Graph team that was spotted on Windows with CUDA 12.3/12.4

[aelovikov-intel]

CUDA failures:

Failed Tests (2):
SYCL :: Graph/Update/Explicit/whole_update_double_buffer.cpp
SYCL :: Graph/Update/RecordReplay/whole_update_double_buffer.cpp

Dev IGC update couldn't be directly guilty, but the implicit image rebuild might have caused that if third party base image has changed. @intel/llvm-gatekeepers , @intel/llvm-reviewers-cuda , @intel/sycl-graphs-reviewers , did anybody see those elsewhere?

@JackAKirk reported a bug with these 2 tests internally to Codeplay SYCL-Graph team that was spotted on Windows with CUDA 12.3/12.4

Thinking more about this, AWS CUDA task is run in the target context, not in the pull_request context, so the image shouldn't have been different from any other PR. I'd expect these might just be flaky. I'm going to merge this PR.