When wait_event happens in dequeue_event(), it checks if event_list is empty
without acquiring client->lock. A potential race can happen as follows:
T1 T2 T3
sleep in sleep in
dequeue_event() dequeue_event()
enque_event()
wake up, check
if event_list is empty
and is preempted
device is shut down
wake up and
list_del()
try to dequeue event
from empty list
The patch moves acquiring client->lock before checking the event_list.
Found by Linux Driver Verification project (linuxtesting.org).
Signed-off-by: Alexey Khoroshilov <[email protected]>
---
drivers/firewire/core-cdev.c | 6 ++++--
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/drivers/firewire/core-cdev.c b/drivers/firewire/core-cdev.c
index 2a3973a7c441..7010dc2f02f2 100644
--- a/drivers/firewire/core-cdev.c
+++ b/drivers/firewire/core-cdev.c
@@ -314,11 +314,13 @@ static int dequeue_event(struct client *client,
if (ret < 0)
return ret;
+ spin_lock_irq(&client->lock);
if (list_empty(&client->event_list) &&
- fw_device_is_shutdown(client->device))
+ fw_device_is_shutdown(client->device)) {
+ spin_unlock_irq(&client->lock);
return -ENODEV;
+ }
- spin_lock_irq(&client->lock);
event = list_first_entry(&client->event_list, struct event, link);
list_del(&event->link);
spin_unlock_irq(&client->lock);
--
1.9.1
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
