| Message ID | 20221216162917.119406-1-edumazet@google.com (mailing list archive) |
|---|---|
| State | Accepted |
| Commit | e0c8bccd40fc1c19e1d246c39bcf79e357e1ada3 |
| Delegated to: | Netdev Maintainers |
| Headers | show |
| Series | [net] net: stream: purge sk_error_queue in sk_stream_kill_queues() | expand |
Hello: This patch was applied to netdev/net.git (master) by David S. Miller <davem@davemloft.net>: On Fri, 16 Dec 2022 16:29:17 +0000 you wrote: > Changheon Lee reported TCP socket leaks, with a nice repro. > > It seems we leak TCP sockets with the following sequence: > > 1) SOF_TIMESTAMPING_TX_ACK is enabled on the socket. > > Each ACK will cook an skb put in error queue, from __skb_tstamp_tx(). > __skb_tstamp_tx() is using skb_clone(), unless > SOF_TIMESTAMPING_OPT_TSONLY was also requested. > > [...] Here is the summary with links: - [net] net: stream: purge sk_error_queue in sk_stream_kill_queues() https://git.kernel.org/netdev/net/c/e0c8bccd40fc You are awesome, thank you!
diff --git a/net/core/stream.c b/net/core/stream.c index 5b1fe2b82eac753bc8e18c02db04c5906b3a2d97..cd06750dd3297cd0e0f073057a4d85d4078f87c3 100644 --- a/net/core/stream.c +++ b/net/core/stream.c @@ -196,6 +196,12 @@ void sk_stream_kill_queues(struct sock *sk) /* First the read buffer. */ __skb_queue_purge(&sk->sk_receive_queue); + /* Next, the error queue. + * We need to use queue lock, because other threads might + * add packets to the queue without socket lock being held. + */ + skb_queue_purge(&sk->sk_error_queue); + /* Next, the write queue. */ WARN_ON_ONCE(!skb_queue_empty(&sk->sk_write_queue));
Changheon Lee reported TCP socket leaks, with a nice repro. It seems we leak TCP sockets with the following sequence: 1) SOF_TIMESTAMPING_TX_ACK is enabled on the socket. Each ACK will cook an skb put in error queue, from __skb_tstamp_tx(). __skb_tstamp_tx() is using skb_clone(), unless SOF_TIMESTAMPING_OPT_TSONLY was also requested. 2) If the application is also using MSG_ZEROCOPY, then we put in the error queue cloned skbs that had a struct ubuf_info attached to them. Whenever an struct ubuf_info is allocated, sock_zerocopy_alloc() does a sock_hold(). As long as the cloned skbs are still in sk_error_queue, socket refcount is kept elevated. 3) Application closes the socket, while error queue is not empty. Since tcp_close() no longer purges the socket error queue, we might end up with a TCP socket with at least one skb in error queue keeping the socket alive forever. This bug can be (ab)used to consume all kernel memory and freeze the host. We need to purge the error queue, with proper synchronization against concurrent writers. Fixes: 24bcbe1cc69f ("net: stream: don't purge sk_error_queue in sk_stream_kill_queues()") Reported-by: Changheon Lee <darklight2357@icloud.com> Signed-off-by: Eric Dumazet <edumazet@google.com> --- net/core/stream.c | 6 ++++++ 1 file changed, 6 insertions(+)