hugetlbfs: fix issue of preallocation of gigantic pages can't work

Preallocation of gigantic pages can't work bacause of commit
b538908 ("hugetlbfs: extend the definition of hugepages parameter
to support node allocation").  When nid is NUMA_NO_NODE(-1),
alloc_bootmem_huge_page will always return without doing allocation.
Fix this by adding more check.

Link: https://lkml.kernel.org/r/[email protected]
Fixes: b538908 ("hugetlbfs: extend the definition of hugepages parameter to support node allocation")
Signed-off-by: Zhenguo Yao <[email protected]>
Reviewed-by: Mike Kravetz <[email protected]>
Tested-by: Maxim Levitsky <[email protected]>
Reviewed-by: Muchun Song <[email protected]>
Reviewed-by: Baolin Wang <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>

Author: ZhenguoYao1

mm/hugetlb.c

@@ -2973,7 +2973,7 @@ int __alloc_bootmem_huge_page(struct hstate *h, int nid)
 	struct huge_bootmem_page *m = NULL; /* initialize for clang */
 	int nr_nodes, node;
 
-	if (nid >= nr_online_nodes)
+	if (nid != NUMA_NO_NODE && nid >= nr_online_nodes)
 		return 0;
 	/* do node specific alloc */
 	if (nid != NUMA_NO_NODE) {