09.11.2017 10:33, Kirill Tkhai пишет:
>
> Looks good for me, except there is one more call of nfs_callback_down_net()
> in nfs_callback_up().
> It's not clear for me, we shouldn't protect it via the mutex too. Should we?
>
Looks like no, mount is not ready yet and there is no connection to
On 08.11.2017 20:23, Stanislav Kinsburskiy wrote:
> From: Stanislav Kinsburskiy
>
> Here is the race:
>
> CPU #0CPU#1
>
> cleanup_mnt nfs41_callback_svc (get xprt from the list)
> nfs_callback_down ...
>
From: Stanislav Kinsburskiy
Here is the race:
CPU #0 CPU#1
cleanup_mnt nfs41_callback_svc (get xprt from the list)
nfs_callback_down ...
... ...
svc_close_net ...