Skip to content

Commit

Permalink
mm: Clear page->private when splitting or migrating a page
Browse files Browse the repository at this point in the history
In our efforts to remove uses of PG_private, we have found folios with
the private flag clear and folio->private not-NULL.  That is the root
cause behind 642d51f ("ceph: check folio PG_private bit instead
of folio->private").  It can also affect a few other filesystems that
haven't yet reported a problem.

compaction_alloc() can return a page with uninitialised page->private,
and rather than checking all the callers of migrate_pages(), just zero
page->private after calling get_new_page().  Similarly, the tail pages
from split_huge_page() may also have an uninitialised page->private.

Reported-by: Xiubo Li <xiubli@redhat.com>
Tested-by: Xiubo Li <xiubli@redhat.com>
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
  • Loading branch information
Matthew Wilcox (Oracle) committed Jun 23, 2022
1 parent cb995f4 commit b653db7
Show file tree
Hide file tree
Showing 2 changed files with 2 additions and 0 deletions.
1 change: 1 addition & 0 deletions mm/huge_memory.c
Expand Up @@ -2377,6 +2377,7 @@ static void __split_huge_page_tail(struct page *head, int tail,
page_tail);
page_tail->mapping = head->mapping;
page_tail->index = head->index + tail;
page_tail->private = 0;

/* Page flags must be visible before we make the page non-compound. */
smp_wmb();
Expand Down
1 change: 1 addition & 0 deletions mm/migrate.c
Expand Up @@ -1106,6 +1106,7 @@ static int unmap_and_move(new_page_t get_new_page,
if (!newpage)
return -ENOMEM;

newpage->private = 0;
rc = __unmap_and_move(page, newpage, force, mode);
if (rc == MIGRATEPAGE_SUCCESS)
set_page_owner_migrate_reason(newpage, reason);
Expand Down

0 comments on commit b653db7

Please sign in to comment.