mm/migration: fix potential invalid node access for reclaim-based migration

If we failed to setup hotplug state callbacks for mm/demotion:online in
some corner cases, node_demotion will be left uninitialized.  Invalid node
might be returned from the next_demotion_node() when doing reclaim-based
migration.  Use kcalloc to allocate node_demotion to fix the issue.

Link: https://lkml.kernel.org/r/20220318111709.60311-11-linmiaohe@huawei.com
Fixes: ac16ec83 ("mm: migrate: support multiple target nodes demotion")
Signed-off-by: NMiaohe Lin <linmiaohe@huawei.com>
Reviewed-by: N"Huang, Ying" <ying.huang@intel.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: Baolin Wang <baolin.wang@linux.alibaba.com>
Cc: Muchun Song <songmuchun@bytedance.com>
Cc: Zi Yan <ziy@nvidia.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>

mm/migrate.c

@@ -2479,9 +2479,9 @@ static int __meminit migrate_on_reclaim_callback(struct notifier_block *self,
 
 void __init migrate_on_reclaim_init(void)
 {
-	node_demotion = kmalloc_array(nr_node_ids,
-				      sizeof(struct demotion_nodes),
-				      GFP_KERNEL);
+	node_demotion = kcalloc(nr_node_ids,
+				sizeof(struct demotion_nodes),
+				GFP_KERNEL);
 	WARN_ON(!node_demotion);
 
 	hotplug_memory_notifier(migrate_on_reclaim_callback, 100);