Ignore madvise(MADV_WILLNEED) for hugetlbfs-backed regions

madvise(MADV_WILLNEED) forces page cache readahead on a range of memory backed by a file. The assumption is made that the page required is order-0 and "normal" page cache. On hugetlbfs, this assumption is not true and order-0 pages are allocated and inserted into the hugetlbfs page cache. This leaks hugetlbfs page reservations and can cause BUGs to trigger related to corrupted page tables. This patch causes MADV_WILLNEED to be ignored for hugetlbfs-backed regions. Signed-off-by: Mel Gorman <mel@csn.ul.ie> Cc: stable@kernel.org Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Ignore madvise(MADV_WILLNEED) for hugetlbfs-backed regions
madvise(MADV_WILLNEED) forces page cache readahead on a range of memory backed by a file. The assumption is made that the page required is order-0 and "normal" page cache. On hugetlbfs, this assumption is not true and order-0 pages are allocated and inserted into the hugetlbfs page cache. This leaks hugetlbfs page reservations and can cause BUGs to trigger related to corrupted page tables. This patch causes MADV_WILLNEED to be ignored for hugetlbfs-backed regions. Signed-off-by: Mel Gorman <mel@csn.ul.ie> Cc: stable@kernel.org Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Mel Gorman · Linus Torvalds
1 parent 99ee12973e
Showing 1 changed file with 8 additions and 0 deletions Side-by-side Diff
mm/madvise.c
@@ -112,6 +112,14 @@
 	if (!file)
 		return -EBADF;
  
+	/*
+	 * Page cache readahead assumes page cache pages are order-0 which
+	 * is not the case for hugetlbfs. Do not give a bad return value
+	 * but ignore the advice.
+	 */
+	if (vma->vm_flags & VM_HUGETLB)
+		return 0;
+
 	if (file->f_mapping->a_ops->get_xip_mem) {
 		/* no bad return value, but ignore advice */
 		return 0;
...	...	@@ -112,6 +112,14 @@
112	112	if (!file)
113	113	return -EBADF;
114	114
	115	+ /*
	116	+ * Page cache readahead assumes page cache pages are order-0 which
	117	+ * is not the case for hugetlbfs. Do not give a bad return value
	118	+ * but ignore the advice.
	119	+ */
	120	+ if (vma->vm_flags & VM_HUGETLB)
	121	+ return 0;
	122	+
115	123	if (file->f_mapping->a_ops->get_xip_mem) {
116	124	/* no bad return value, but ignore advice */
117	125	return 0;