Runner: make MPI bind-to configurable

[one (alephpiece)]

Description

Make MPI process binding configurable in SuperBench runner modes so benchmarks with their own topology-aware affinity logic can disable the default -bind-to numa behavior when needed.

Major Revision

• Add a new optional bind_to field for mpi modes in the runner
• Keep backward compatibility by defaulting bind_to to numa
• Update MPI command generation to use the configured bind_to value instead of a hard-coded numa

[one (alephpiece)]

@microsoft-github-policy-service agree

[Hongtao Zhang (polarG)]

Overall: The feature is well-motivated and docs are solid, but the PR needs fixes before merge. See inline comments for details.

#	Severity	Finding
1	BLOCKER	Existing MPI tests break — missing bind_to key on test mode dicts
2	SHOULD-FIX	No input validation on bind_to — injection risk + invalid values
3	SHOULD-FIX	No new tests for the bind_to feature

Replaced with inline file comments.

[Hongtao Zhang (polarG)]

/azp run

[azure-pipelines]

Azure Pipelines successfully started running 3 pipeline(s).

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 85.70%. Comparing base (3c95714) to head (a5fb2be).
⚠️ Report is 1 commits behind head on main.

Additional details and impacted files

@@ Coverage Diff @@## main #797 +/- ##
==========================================
+ Coverage 85.69% 85.70% +0.01% 
==========================================
Files 103 103 Lines 7890 7897 +7 ==========================================
+ Hits 6761 6768 +7 
Misses 1129 1129 

Flag	Coverage Δ
cpu-python3.10-unit-test	70.45% <100.00%> (+0.02%)	⬆️
cpu-python3.12-unit-test	70.45% <100.00%> (+0.02%)	⬆️
cpu-python3.7-unit-test	69.87% <100.00%> (+0.02%)	⬆️
cuda-unit-test	83.61% <100.00%> (+0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.