Messages in this thread Patch in this message | | | From | Jonathan Maxwell <> | Date | Thu, 21 Oct 2021 12:48:27 +1100 | Subject | Re: [net-next] tcp: don't free a FIN sk_buff in tcp_remove_empty_skb() |
| |
On Thu, Oct 21, 2021 at 12:11 PM Eric Dumazet <edumazet@google.com> wrote: > > On Wed, Oct 20, 2021 at 4:25 PM Jon Maxwell <jmaxwell37@gmail.com> wrote: > > > > A customer reported sockets stuck in the CLOSING state. A Vmcore revealed that > > the write_queue was not empty as determined by tcp_write_queue_empty() but the > > sk_buff containing the FIN flag had been freed and the socket was zombied in > > that state. Corresponding pcaps show no FIN from the Linux kernel on the wire. > > > > Some instrumentation was added to the kernel and it was found that there is a > > timing window where tcp_sendmsg() can run after tcp_send_fin(). > > > > tcp_sendmsg() will hit an error, for example: > > > > 1269 ▹ if (sk->sk_err || (sk->sk_shutdown & SEND_SHUTDOWN))↩ > > 1270 ▹ ▹ goto do_error;↩ > > > > tcp_remove_empty_skb() will then free the FIN sk_buff as "skb->len == 0". The > > TCP socket is now wedged in the FIN-WAIT-1 state because the FIN is never sent. > > > > If the other side sends a FIN packet the socket will transition to CLOSING and > > remain that way until the system is rebooted. > > > > Fix this by checking for the FIN flag in the sk_buff and don't free it if that > > is the case. Testing confirmed that fixed the issue. > > > > Fixes: fdfc5c8594c2 ("tcp: remove empty skb from write queue in error cases") > > Signed-off-by: Jon Maxwell <jmaxwell37@gmail.com> > > --- > > net/ipv4/tcp.c | 2 +- > > 1 file changed, 1 insertion(+), 1 deletion(-) > > > > diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c > > index c2d9830136d2..d2b06d8f0c37 100644 > > --- a/net/ipv4/tcp.c > > +++ b/net/ipv4/tcp.c > > @@ -938,7 +938,7 @@ int tcp_send_mss(struct sock *sk, int *size_goal, int flags) > > */ > > void tcp_remove_empty_skb(struct sock *sk, struct sk_buff *skb) > > { > > - if (skb && !skb->len) { > > + if (skb && !skb->len && !TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN) { > > tcp_unlink_write_queue(skb, sk); > > if (tcp_write_queue_empty(sk)) > > tcp_chrono_stop(sk, TCP_CHRONO_BUSY); > > > > Very nice catch ! >
Thanks Eric.
> The FIN flag is a really special case here. > > What we need is to make sure the skb is 'empty' . > > What about using a single condition ? > > if (skb && TCP_SKB_CB(skb)->seq == TCP_SKB_CB(skb)->end_seq)
Good call as the end_seq will be +1 for a FIN. So that's better.
Let me give the customer a kernel with your idea:
--- net/ipv4/tcp.c 2021-10-20 22:50:35.836001950 +0530 +++ net/ipv4/tcp.c.patch 2021-10-21 01:42:08.493569483 +0530 @@ -955,7 +955,7 @@ */ void tcp_remove_empty_skb(struct sock *sk, struct sk_buff *skb) { - if (skb && !skb->len) { + if (skb && TCP_SKB_CB(skb)->seq == TCP_SKB_CB(skb)->end_seq) { tcp_unlink_write_queue(skb, sk); if (tcp_write_queue_empty(sk)) tcp_chrono_stop(sk, TCP_CHRONO_BUSY); I'll ask the customer to confirm that the v1 patch as above also resolves the issue. Although I expect it will.
Then I'll resubmit a v1 patch with your suggestion probably early next week.
Regards
Jon
| |