6.11-stable review patch. If anyone has any objections, please let me know.
------------------
From: Steve Sistare <steven.sistare@xxxxxxxxxx>
commit c56b6f3d801d7ec8965993342bdd9e2972b6cb8e upstream.
memfd_pin_folios followed by unpin_folios fails to restore free_huge_pages
if the pages were not already faulted in, because the folio refcount for
pages created by memfd_alloc_folio never goes to 0. memfd_pin_folios
needs another folio_put to undo the folio_try_get below:
memfd_alloc_folio()
alloc_hugetlb_folio_nodemask()
dequeue_hugetlb_folio_nodemask()
dequeue_hugetlb_folio_node_exact()
folio_ref_unfreeze(folio, 1); ; adds 1 refcount
folio_try_get() ; adds 1 refcount
hugetlb_add_to_page_cache() ; adds 512 refcount (on x86)
With the fix, after memfd_pin_folios + unpin_folios, the refcount for the
(unfaulted) page is 512, which is correct, as the refcount for a faulted
unpinned page is 513.
Link: https://lkml.kernel.org/r/1725373521-451395-3-git-send-email-steven.sistare@xxxxxxxxxx
Fixes: 89c1905d9c14 ("mm/gup: introduce memfd_pin_folios() for pinning memfd folios")
Signed-off-by: Steve Sistare <steven.sistare@xxxxxxxxxx>
Acked-by: Vivek Kasireddy <vivek.kasireddy@xxxxxxxxx>
Cc: David Hildenbrand <david@xxxxxxxxxx>
Cc: Jason Gunthorpe <jgg@xxxxxxxxxx>
Cc: Matthew Wilcox <willy@xxxxxxxxxxxxx>
Cc: Muchun Song <muchun.song@xxxxxxxxx>
Cc: Peter Xu <peterx@xxxxxxxxxx>
Cc: <stable@xxxxxxxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
---
mm/gup.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
--- a/mm/gup.c
+++ b/mm/gup.c
@@ -3618,7 +3618,7 @@ long memfd_pin_folios(struct file *memfd
pgoff_t start_idx, end_idx, next_idx;
struct folio *folio = NULL;
struct folio_batch fbatch;
- struct hstate *h;
+ struct hstate *h = NULL;
long ret = -EINVAL;
if (start < 0 || start > end || !max_folios)
@@ -3662,6 +3662,8 @@ long memfd_pin_folios(struct file *memfd
&fbatch);
if (folio) {
folio_put(folio);
+ if (h)
+ folio_put(folio);
folio = NULL;
}
