Lines Matching refs:nodes

1355  * between these nodes are slowed down, to allow things to settle down.
1364 /* Handle placement on systems where not all nodes are directly connected. */
1371      * All nodes are directly connected, and the same distance
1380      * which should be ok given the number of nodes rarely exceeds 8.
1387          * The furthest away nodes in the system are not interesting
1396          * of nodes, and move tasks towards the group with the most
1397          * memory accesses. When comparing two nodes at distance
1398          * "hoplimit", only nodes closer by than "hoplimit" are part
1399          * of each group. Skip other nodes.
1405         /* Add up the faults from nearby nodes. */
1414          * no fixed "groups of nodes". Instead, nodes that are not
1416          * nodes; a numa_group can occupy any set of nodes.
1435  * evenly spread out between numa nodes.
2101      * Look at other nodes in these cases:
2104      *   multiple NUMA nodes; in order to better consolidate the group,
2121             /* Only consider nodes where both task and groups benefit */
2136      * If the task is part of a workload that spans multiple NUMA nodes,
2137      * and is migrating into one of the workload's active nodes, remember
2205  * Find out how many nodes on the workload is actively running on. Do this by
2206  * tracking the nodes from which NUMA hinting faults are triggered. This can
2207  * be different from the set of nodes where the workload's memory is currently
2360     nodemask_t nodes;
2363     /* Direct connections between all NUMA nodes. */
2370      * scores nodes according to the number of NUMA hinting faults on
2371      * both the node itself, and on nearby nodes.
2396      * inside the highest scoring group of nodes. The nodemask tricks
2399     nodes = node_online_map;
2405         /* Are there nodes at this distance from each other? */
2410         for_each_node_mask(a, nodes)
2417             for_each_node_mask(b, nodes)
2422                     node_clear(b, nodes);
2438         /* Next round, evaluate the nodes within max_group. */
2442         nodes = max_group;
2775      * If a workload spans multiple NUMA nodes, a shared fault that
2776      * occurs wholly within the set of nodes that the workload is