On Mon, Jan 25, 2021 at 04:34:23PM -0800, Dave Hansen wrote:

From: Dave Hansen <dave.hansen@linux.intel.com>

Reclaim-based migration is attempting to optimize data placement in
memory based on the system topology.  If the system changes, so must
the migration ordering.

The implementation here is pretty simple and entirely unoptimized.  On
any memory or CPU hotplug events, assume that a node was added or
removed and recalculate all migration targets.  This ensures that the
node_demotion[] array is always ready to be used in case the new
reclaim mode is enabled.

This recalculation is far from optimal, most glaringly that it does
not even attempt to figure out if nodes are actually coming or going.
But, given the expected paucity of hotplug events, this should be
fine.

Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Yang Shi <redacted>
Cc: David Rientjes <rientjes@google.com>
Cc: Huang Ying <redacted>
Cc: Dan Williams <redacted>
Cc: David Hildenbrand <redacted>
Cc: osalvador <osalvador@suse.de>
---

[...]

+
+/*
+ * React to hotplug events that might affect the migration targes
+ * like events that online or offline NUMA nodes.
+ *
+ * The ordering is also currently dependent on which nodes have
+ * CPUs.  That means we need CPU on/offline notification too.
+ */
+static int migration_online_cpu(unsigned int cpu)
+{
+	set_migration_target_nodes();
+	return 0;
+}
+
+static int migration_offline_cpu(unsigned int cpu)
+{
+	set_migration_target_nodes();
+	return 0;
+}
+
+/*
+ * This leaves migrate-on-reclaim transiently disabled
+ * between the MEM_GOING_OFFLINE and MEM_OFFLINE events.
+ * This runs reclaim-based micgration is enabled or not.
+ * This ensures that the user can turn reclaim-based
+ * migration at any time without needing to recalcuate
+ * migration targets.
+ *
+ * These callbacks already hold get_online_mems().  That
+ * is why __set_migration_target_nodes() can be used as
+ * opposed to set_migration_target_nodes().
+ */
+static int __meminit migrate_on_reclaim_callback(struct notifier_block *self,
+						 unsigned long action, void *arg)
+{
+	switch (action) {
+	case MEM_GOING_OFFLINE:
+		/*
+		 * Make sure there are not transient states where
+		 * an offline node is a migration target.  This
+		 * will leave migration disabled until the offline
+		 * completes and the MEM_OFFLINE case below runs.
+		 */
+		disable_all_migrate_targets();
+		break;
+	case MEM_OFFLINE:
+	case MEM_ONLINE:
+		/*
+		 * Recalculate the target nodes once the node
+		 * reaches its final state (online or offline).
+		 */
+		__set_migration_target_nodes();
+		break;
+	case MEM_CANCEL_OFFLINE:
+		/*
+		 * MEM_GOING_OFFLINE disabled all the migration
+		 * targets.  Reenable them.
+		 */
+		__set_migration_target_nodes();
+		break;
+	case MEM_GOING_ONLINE:
+	case MEM_CANCEL_ONLINE:
+		break;
+	}
+
+	return notifier_from_errno(0);
+}

This looks good, and I kinda like it.
But in this case, all we care about is whether NUMA node does or does
not have memory, so we have to remove/added into the demotion list.
So, would make more sense to have a kinda helper in
node_states_{set,clear}_node that calls the respective functions
(disable_all_migrate_targets and __set_migration_target_nodes)?

-- 
Oscar Salvador
SUSE L3