mm, slab: restore NUMA policy support for large kmalloc

The slab allocator observes the task's NUMA policy in various places
such as allocating slab pages. Large kmalloc() allocations used to do
that too, until an unintended change by c4cab55 ("mm/slab_common:
cleanup kmalloc_large()") resulted in ignoring mempolicy and just
preferring the local node. Restore the NUMA policy support.

Fixes: c4cab55 ("mm/slab_common: cleanup kmalloc_large()")
Cc: <[email protected]>
Acked-by: Christoph Lameter (Ampere) <[email protected]>
Acked-by: Roman Gushchin <[email protected]>
Reviewed-by: Harry Yoo <[email protected]>
Signed-off-by: Vlastimil Babka <[email protected]>

Author: tehcaster

mm/slub.c

@@ -4269,7 +4269,12 @@ static void *___kmalloc_large_node(size_t size, gfp_t flags, int node)
 		flags = kmalloc_fix_flags(flags);
 
 	flags |= __GFP_COMP;
-	folio = (struct folio *)alloc_pages_node_noprof(node, flags, order);
+
+	if (node == NUMA_NO_NODE)
+		folio = (struct folio *)alloc_pages_noprof(flags, order);
+	else
+		folio = (struct folio *)__alloc_pages_noprof(flags, order, node, NULL);
+
 	if (folio) {
 		ptr = folio_address(folio);
 		lruvec_stat_mod_folio(folio, NR_SLAB_UNRECLAIMABLE_B,