fix: Fix L0_perf_nomodel shared memory

[kthui]

What does the PR do?

Fix L0_perf_nomodel shared memory test - system and cuda.

• Update flag name BENCHMARK_TEST_SHARED_MEMORY to TEST_SHARED_MEMORY.
• Correctly set the --output-shared-memory-size flag for perf analyzer.

Root cause:

• The CI runners were re-configured to set TEST_SHARED_MEMORY instead of BENCHMARK_TEST_SHARED_MEMORY, causing the system/CUDA shared memory tests to run as non-shared memory test, resulting in no data reported for both shared memory tests since the re-configuration.

Checklist

• PR title reflects the change and is of format <commit_type>: <Title>
• Changes are described in the pull request.
• Related issues are referenced.
• Populated github labels field
• Added test plan and verified test passes.
• Verified that the PR passes existing CI.
• Verified copyright is correct on all changed files.
• Added succinct git squash message before merging ref.
• All template sections are filled out.
• Optional: Additional screenshots for behavior/output changes with before/after.

Commit Type:

Check the conventional commit type
box here and add the label to the github PR.

• build
• ci
• docs
• feat
• fix
• perf
• refactor
• revert
• style
• test

Related PRs:

N/A

Where should the reviewer start?

N/A

Test plan:

N/A

• CI Pipeline ID: 19429907

Caveats:

N/A

Background

N/A

Related Issues: (use one of the action keywords Closes / Fixes / Resolves / Relates to)

N/A

[indrajit96]

Can you add the RCA of tests not completing either on the bug or PR description?

[kthui]

Sure, added root cause to the description.

[indrajit96]

Just a FYI, we might need to edit these test soon one the PA fixes for C API and PA Hang are done.
LGTM for now. Approving,

[kthui]

I will merge to main first. We can file additional tickets if there are breakages with the PA fixes.