[EXP][CMDBUF] Add adapters code for Prefetch and Advise commands

[mfrancepillois]

Note: This PR also includes changes from the main branch (commit 0d3c1fc) which add these commands to the UR specification, but are not present in the adapter branch yet.

• Adds adapters code support for prefetch and advise memory hints for level_zero backend.
• Adds entry points for prefetch and advise memory hints for CUDA backend.

[mfrancepillois]

Linked DPC++ PR: intel/llvm#11474

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (fe2735a) 15.79% compared to head (01cd56d) 15.73%.
Report is 63 commits behind head on main.

❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #937      +/-   ##
==========================================
- Coverage   15.79%   15.73%   -0.06%     
==========================================
  Files         223      223              
  Lines       31351    31465     +114     
  Branches     3511     3556      +45     
==========================================
+ Hits         4951     4952       +1     
- Misses      26349    26462     +113     
  Partials       51       51              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[mfrancepillois]

This PR was rebased after the main branch was merged with adapter branch.
We also added the support for CUDA backend. Since Graph CUDA backend does not handle prefetch and mem-advise hints, these operations are implemented as empty nodes ("nop").

[fabiomestre]

I have updated the target branch of this PR from the adapters branch to the main branch.
Development in UR is moving back to main. The adapters branch will soon be deleted.

[EwanC]

@jandres742 Could you review this again please, it's out of draft

[EwanC]

@jandres742 Could you review this again please, it's out of draft

@nrspruit could you review on behalf of @oneapi-src/unified-runtime-level-zero-write

[martygrant]

OpenCL LGTM

[EwanC]

ping @oneapi-src/unified-runtime-cuda-write & @oneapi-src/unified-runtime-hip-write for review

[GeorgeWeb]

CUDA and HIP changes look good to me.

[nrspruit]

Approved for level zero.