[PT][CI] Add additional Inductor tests to the defaults for the Inductor UT WF

[vlad-penkin]

This is a follow up ticket to address recent failures

• inductor/test_kernel_benchmark.py ([Pytorch Upstream] Inductor UT: Triton raise "TypeError: dynamic_func() got multiple values for argument 'XBLOCK'" #3499)
• inductor/test_torchinductor_opinfo.py ([PyTorch Upstream] inductor/test_torchinductor_opinfo.py::TestInductorOpInfoXPU::test_comprehensive_masked_std_xpu_int64 regression #3589)
• inductor/test_torchinductor_dynamic_shapes.py ([Pytorch Upstream] python inductor/test_torchinductor_dynamic_shapes.py -k test_split_cumsum_dynamic_shapes_xpu failed #3570) and ([Pytorch Upstream][Regression] Inductor UT get triton raised "RuntimeError: PassManager::run failed" #3519)
• [DONE] inductor/test_aot_inductor.py ([PyTorch] python test/inductor/test_aot_inductor.py AOTInductorTestABICompatibleGpu.test_triton_kernel_weird_param_order_xpu failed #3553)
• inductor/test_torchinductor.py ([Pytorch Upstream][Regression] Inductor UT regression after update triton: "attributeError: 'dict' object has no attribute 'divisible_by_16'" #3520) and ([Pytorch upstream] Intel Triton raised ValueError: Cannot bitcast data-type of size 32 to data-type of size 8 #3447)
• [DONE] inductor/test_gpu_cpp_wrapper ([Pytorch Upstream] Inductor cpp wrapper UT failed after update triton. #3500)

These failures still needs to be reconciled

• 'tt.expand_dims' op inferred type(s) are incompatible with return type(s) of operation #3107

[vlad-penkin]

cc: @alexbaden and @anmyachev