feat(aorta): disaggregated multi-node torchrun launch (AIMVT-173)

[speriaswamy-amd]

Stack 4/6 — splits #171. Base: #328. Implements AIMVT-173.

Why

Run the Aorta benchmark across N nodes from a single cluster.json instead of requiring N single-node cluster files, mirroring the disaggregated pattern the PyTorch xDiT and SGLang multi-node suites already use.

What changed

run() branches on the resolved launch mode:

• script — today's behavior verbatim (experiment_script on the head node).
• torchrun — a rank-group on every node in parallel, rendezvous-ing on the head.
• auto (default) — script for 1-node clusters, torchrun above that. Existing single-node configs are unaffected.

New helpers in cvs/runners/aorta.py:

• _resolve_launch_mode() — script with >1 node is a config error rather than a silent single-node run.
• _resolve_master_addr() — explicit override → head node's VPC/RDMA address → plain node id. The mgmt/SSH address is frequently orchestrator-only and unreachable from peer nodes.
• _pick_free_port_on() / _pick_master_port() — binds and releases a port over SSH on the head node, which is the host that actually holds the rendezvous socket. The snippet is passed as one shlex-quoted argument; as separate argv elements the remote shell re-splits it on its own semicolons.
• _build_torchrun_command(), _run_single_node().

TraceLens/GEMM analysis now explicitly target the head node's container, and training-log discovery iterates the node list, rather than relying on locals left over from the single-node path.

Test

ruff clean. Unit tests 603 → 623 (20 new: mode resolution, port/addr selection, command construction, validate_config).

[speriaswamy-amd]

Reopening — this was auto-closed by GitHub when a rebase mistake on my end briefly collapsed this branch to be identical to its base (0 commits/0 diff). The branch has since been corrected and force-pushed with the real commit restored; verified diff is +564/-42 across 2 files as intended.

[speriaswamy-amd]

Superseded by #390 (same branches, same content) — this PR got stuck in an auto-closed state that GitHub wouldn't let me reopen after a rebase mistake temporarily zeroed its diff. No content was lost.