cuda.bindings._v2.driver.graph_remove_dependencies_v2#

cuda.bindings._v2.driver.graph_remove_dependencies_v2( intptr_t h_graph, from_, to, edge_data, size_t num_dependencies, )#

Removes dependency edges from a graph.

The number of dependencies to be removed is defined by num_dependencies. Elements in from and to at corresponding indices define a dependency. Each node in from and to must belong to h_graph. If num_dependencies is 0, elements in from and to will be ignored. Specifying an edge that does not exist in the graph, with data matching edge_data, results in an error. edge_data is nullable, which is equivalent to passing default (zeroed) data for each edge. Dependencies cannot be removed from graphs which contain allocation or free nodes. Any attempt to do so will return an error.

Parameters:

h_graph (intptr_t) – Graph from which to remove dependencies.
from (object) –
Array of nodes that provide the dependencies. It can be:
- an int as the pointer address to the array, or
- a Python sequence of ints (as pointer addresses).
to (object) –
Array of dependent nodes. It can be:
- an int as the pointer address to the array, or
- a Python sequence of ints (as pointer addresses).
edge_data (intptr_t) – Optional array of edge data. If NULL, edge data is assumed to be default (zeroed).
num_dependencies (size_t) – Number of dependencies to be removed.