General question on `numa_run_on_node` about numactl HOT 2 CLOSED

First, it is clear that task refers to thread.

from numa manuals：
numa_run_on_node() runs the current task and its children on a specific node. They will not migrate to CPUs of other nodes until the node affinity is reset with a new call to numa_run_on_node_mask(). Passing -1 permits the kernel to schedule on all nodes again. On success, 0 is returned; on error -1 is returned, and errno is set to indicate the error.

1. spawn 1 thread for each NUMA domain, for each thread restrict it to a different NUMA domain as:
   numa_run_on_node(i);

• This restricts the thread and its children to run only on the CPU of the specified node.

2、From each thread spawn X threads, where X is the number of cores inside of each domain.

• The child thread inherits the attributes of its parent thread and can run only on the CPU of the specified node.

I think this will meet your requirements.

from numactl.

Thank you very much for your quick response!

from numactl.