[Fabric] Test code updates and minor fixes

[aagarwalTT]

Ticket

Link to Github Issue

Problem description

The test code selects worker cores at random for running traffic b/w chips. This could impact benchmarking as the performance could be impacted by the physical distance of the cores from the routers.

What's changed

• The test code will now consistently pick the optimum worker cores for any given router. This will minimize any variations in the performance numbers because of worker core selection.
• New options:
  • Benchmark mode (--benchmark): Bundles together all the settings and configurations needed to run perf benchmarks.
  • Packet size in KB (--packet_size_kb)
• Removed options:
  • Max packet size in words (--max_packet_size_words): Use --packet_size_kb instead.

Checklist

• All post commit CI passes (https://github.com/tenstorrent/tt-metal/actions/runs/13146015390)
• Blackhole Post commit CI passes (if applicable)
• Model regression CI passes (if applicable)
• Device performance regression CI passes (if applicable)
• (For models and ops writers) Full new models tests CI passes (if applicable)
• New/Existing tests provide coverage for changes