diff options
author | Eric Dumazet <edumazet@google.com> | 2017-03-03 22:08:21 (GMT) |
---|---|---|
committer | Greg Kroah-Hartman <gregkh@linuxfoundation.org> | 2017-03-22 11:43:33 (GMT) |
commit | 07753bc6a2816c1c3b9f7bff133251f623c7bc91 (patch) | |
tree | 72b9554719075474e4a5b1c6ce96fd732a04217c /net/ipv4/tcp_ipv4.c | |
parent | 4547f03d1a624d7e92c65077b7cb5b2a33839cb3 (diff) | |
download | linux-07753bc6a2816c1c3b9f7bff133251f623c7bc91.tar.xz |
tcp: fix various issues for sockets morphing to listen state
[ Upstream commit 02b2faaf0af1d85585f6d6980e286d53612acfc2 ]
Dmitry Vyukov reported a divide by 0 triggered by syzkaller, exploiting
tcp_disconnect() path that was never really considered and/or used
before syzkaller ;)
I was not able to reproduce the bug, but it seems issues here are the
three possible actions that assumed they would never trigger on a
listener.
1) tcp_write_timer_handler
2) tcp_delack_timer_handler
3) MTU reduction
Only IPv6 MTU reduction was properly testing TCP_CLOSE and TCP_LISTEN
states from tcp_v6_mtu_reduced()
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Dmitry Vyukov <dvyukov@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Diffstat (limited to 'net/ipv4/tcp_ipv4.c')
-rw-r--r-- | net/ipv4/tcp_ipv4.c | 7 |
1 files changed, 5 insertions, 2 deletions
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 2259114..5bb809f 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -269,10 +269,13 @@ EXPORT_SYMBOL(tcp_v4_connect); */ void tcp_v4_mtu_reduced(struct sock *sk) { - struct dst_entry *dst; struct inet_sock *inet = inet_sk(sk); - u32 mtu = tcp_sk(sk)->mtu_info; + struct dst_entry *dst; + u32 mtu; + if ((1 << sk->sk_state) & (TCPF_LISTEN | TCPF_CLOSE)) + return; + mtu = tcp_sk(sk)->mtu_info; dst = inet_csk_update_pmtu(sk, mtu); if (!dst) return; |