feat(graph): capture CUDA Graph execution topology

[CodingInAVan]

Description

Capture the static node topology for each CUDA graph execution handle and
link graph-launch activity to that opaque handle.

• emit graph_node_definition records with node type and dependency count
• support CUDA Runtime and Driver graph instantiation paths
• re-emit cached topology for later segments without duplicating identities
• release cached definitions when a graph execution object is destroyed

Type of Change

• Bug fix
• New feature
• Documentation update

Testing

Checklist

• My code follows the style guidelines of this project
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas