Sessions are already removed by the proto ->destroy() handlers, and
since commit f3c66d4e144a ("l2tp: prevent creation of sessions on terminated
tunnels"),
we're guaranteed that no new session can be created afterwards.
Furthermore, l2tp_tunnel_closeall() can sleep when there are sessions
left to close. So we really shouldn't call it in a ->sk_destruct()
handler, as it can be used from atomic context.
Signed-off-by: Guillaume Nault <g.na...@alphalink.fr>
---
net/l2tp/l2tp_core.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/net/l2tp/l2tp_core.c b/net/l2tp/l2tp_core.c
index 7c8d1eb757a5..350fcd39ebd8 100644
--- a/net/l2tp/l2tp_core.c
+++ b/net/l2tp/l2tp_core.c
@@ -1246,8 +1246,6 @@ static void l2tp_tunnel_destruct(struct sock *sk)
list_del_rcu(&tunnel->list);
spin_unlock_bh(&pn->l2tp_tunnel_list_lock);
- l2tp_tunnel_closeall(tunnel);
-
tunnel->sock = NULL;
l2tp_tunnel_dec_refcount(tunnel);
--
2.11.0
