Skip to main content

compute_runtime_plan

Function compute_runtime_plan 

Source
pub fn compute_runtime_plan(graph: &CuGraph) -> CuResult<CuExecutionLoop>
Expand description

Compute the default (Linearity) execution plan for graph.

The plan is now pluggable: this splits into the shared order + check_order + plan_from_order pipeline in planner, kept here so direct callers (tests, tooling) keep a one-call entry point.