[PATCH] mm: Ignore data-race in __swap_writepage

[Pei Li]

Syzbot reported a possible data race:

BUG: KCSAN: data-race in __swap_writepage / scan_swap_map_slots

read-write to 0xffff888102fca610 of 8 bytes by task 7106 on cpu 1.
read to 0xffff888102fca610 of 8 bytes by task 7080 on cpu 0.

While we are in __swap_writepage to read sis->flags, scan_swap_map_slots
is trying to update it with SWP_SCANNING.

value changed: 0x0000000000008083 -> 0x0000000000004083.

While this can be updated non-atomicially, this won't affect
SWP_SYNCHRONOUS_IO, so we consider this data-race safe.

This is possibly introduced by commit 3222d8c2a7f8 ("block: remove
->rw_page"), where this if branch is introduced.

Reported-by: syzbot+da25887cc13da6bf3b8c@xxxxxxxxxxxxxxxxxxxxxxxxx
Closes: https://syzkaller.appspot.com/bug?extid=da25887cc13da6bf3b8c
Fixes: 3222d8c2a7f8 ("block: remove ->rw_page")
Signed-off-by: Pei Li <peili.dev@xxxxxxxxx>
---
Syzbot reported a possible data race:

BUG: KCSAN: data-race in __swap_writepage / scan_swap_map_slots

read-write to 0xffff888102fca610 of 8 bytes by task 7106 on cpu 1.
read to 0xffff888102fca610 of 8 bytes by task 7080 on cpu 0.

While we are in __swap_writepage to read sis->flags, scan_swap_map_slots
is trying to update it with SWP_SCANNING.

value changed: 0x0000000000008083 -> 0x0000000000004083.

While this can be updated non-atomicially, this won't affect
SWP_SYNCHRONOUS_IO, so we consider this data-race safe.

This is possibly introduced by commit 3222d8c2a7f8 ("block: remove
->rw_page"), where this if branch is introduced.
---
 mm/page_io.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/mm/page_io.c b/mm/page_io.c
index 0a150c240bf4..15be678259a5 100644
--- a/mm/page_io.c
+++ b/mm/page_io.c
@@ -384,7 +384,12 @@ void __swap_writepage(struct folio *folio, struct writeback_control *wbc)
 	 */
 	if (data_race(sis->flags & SWP_FS_OPS))
 		swap_writepage_fs(folio, wbc);
-	else if (sis->flags & SWP_SYNCHRONOUS_IO)
+	/*
+	 * ->flags can be updated non-atomicially (scan_swap_map_slots),
+	 * but that will never affect SWP_SYNCHRONOUS_IO, so the data_race
+	 * is safe.
+	 */
+	else if (data_race(sis->flags & SWP_SYNCHRONOUS_IO))
 		swap_writepage_bdev_sync(folio, wbc, sis);
 	else
 		swap_writepage_bdev_async(folio, wbc, sis);

---
base-commit: 256abd8e550ce977b728be79a74e1729438b4948
change-id: 20240711-bug13-634ff418d160

Best regards,
-- 
Pei Li <peili.dev@xxxxxxxxx>