cuda.bindings._v2.driver.graph_exec_memcpy_node_set_params#

cuda.bindings._v2.driver.graph_exec_memcpy_node_set_params( intptr_t h_graph_exec, intptr_t h_node, copy_params, intptr_t ctx, )#

Sets the parameters for a memcpy node in the given graphExec.

Updates the work represented by h_node in h_graph_exec as though h_node had contained copy_params at instantiation. h_node must remain in the graph which was used to instantiate h_graph_exec. Changed edges to and from h_node are ignored. The source and destination memory in copy_params must be allocated from the same contexts as the original source and destination memory. Both the instantiation-time memory operands and the memory operands in copy_params must be 1-dimensional. Zero-length operations are not supported. The modifications only affect future launches of h_graph_exec. Already enqueued or running launches of h_graph_exec are not affected by this call. h_node is also not modified by this call. Returns CUDA_ERROR_INVALID_VALUE if the memory operands’ mappings changed or either the original or new memory operands are multidimensional.

Parameters:

h_graph_exec (intptr_t) – The executable graph in which to set the specified node.
h_node (intptr_t) – Memcpy node from the graph which was used to instantiate graphExec.
copy_params (intptr_t) – The updated parameters to set.
ctx (intptr_t) – Context on which to run the node.