We should have 2 references on the device at this point, did you see a “nbd: possibly leaking a device” message before the kasan stuff? Thanks,
Josef
On 4/28/17, 12:29 AM, "Ming Lei" <ming.lei@xxxxxxxxxx> wrote:
Looks it is a typo, just fix it, otherwise the following
warning can be trigerred:
[ming@VM]$sudo rmmod nbd
[sudo] password for ming:
==================================================================
BUG: KASAN: use-after-free in nbd_cleanup+0x115/0x18e [nbd] at addr ffff88024ca539b0
Read of size 8 by task rmmod/2079
Object at ffff88024ca53900, in cache kmalloc-256 size: 256
Allocated:
PID = 1414
Freed:
PID = 2079
Memory state around the buggy address:
ffff88024ca53880: fb fb fb fb fb fb fb fb fc fc fc fc fc fc fc fc
ffff88024ca53900: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
>ffff88024ca53980: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
^
ffff88024ca53a00: fc fc fc fc fc fc fc fc fb fb fb fb fb fb fb fb
ffff88024ca53a80: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
==================================================================
Fixes: c6a4759ea0c9(nbd: add device refcounting)
Cc: Josef Bacik <jbacik@xxxxxx>
Signed-off-by: Ming Lei <ming.lei@xxxxxxxxxx>
---
drivers/block/nbd.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c
index 5583dc4ff941..fa44a6fce4cb 100644
--- a/drivers/block/nbd.c
+++ b/drivers/block/nbd.c
@@ -2110,7 +2110,6 @@ static void __exit nbd_cleanup(void)
if (refcount_read(&nbd->refs) != 2)
printk(KERN_ERR "nbd: possibly leaking a device\n");
nbd_put(nbd);
- nbd_put(nbd);
}
idr_destroy(&nbd_index_idr);
--
2.9.3
