[Bench] add SubmitKernel benchmark variant with longer kernel

[igchor]

To test OOO queues

[igchor]

I'm not sure if we want all the scenarios that I added (SubmitKernel has a lot of variants already) but it would be nice to have comparison between SYCL, UR and L0

[pbalcer]

I'm not sure if we want all the scenarios that I added (SubmitKernel has a lot of variants already) but it would be nice to have comparison between SYCL, UR and L0

Ideally we'd wait until we deal with the long load times / run changes on the dashboard.... but yeah, we shouldn't worry too much about adding new scenarios.

[igchor]

@intel/llvm-gatekeepers this is ready for merge, the failures are unrelated (this PR only modifies benchmark scripts)

[aelovikov-intel]

@intel/llvm-gatekeepers this is ready for merge, the failures are unrelated (this PR only modifies benchmark scripts)

This will be the second PR where I ask folks working on benchmarks to fix CI not to run unnecessary tasks.

[PatKamin]

@intel/llvm-gatekeepers this is ready for merge, the failures are unrelated (this PR only modifies benchmark scripts)

This will be the second PR where I ask folks working on benchmarks to fix CI not to run unnecessary tasks.

We'll mitigate this so to not to block the CI with bench PRs

[aelovikov-intel]

@intel/llvm-gatekeepers this is ready for merge, the failures are unrelated (this PR only modifies benchmark scripts)

This will be the second PR where I ask folks working on benchmarks to fix CI not to run unnecessary tasks.

We'll mitigate this so to not to block the CI with bench PRs

Ian uploaded #18732