Messages in this thread Patch in this message | | | From | Jiri Slaby <> | Subject | [PATCH 3.12 084/127] tcp: fix overflow in __tcp_retransmit_skb() | Date | Fri, 25 Nov 2016 09:29:57 +0100 |
| |
From: Eric Dumazet <edumazet@google.com>
3.12-stable review patch. If anyone has any objections, please let me know.
===============
[ Upstream commit ffb4d6c8508657824bcef68a36b2a0f9d8c09d10 ]
If a TCP socket gets a large write queue, an overflow can happen in a test in __tcp_retransmit_skb() preventing all retransmits.
The flow then stalls and resets after timeouts.
Tested:
sysctl -w net.core.wmem_max=1000000000 netperf -H dest -- -s 1000000000
Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Jiri Slaby <jslaby@suse.cz> --- net/ipv4/tcp_output.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index aa72c9d604a0..f08921156be8 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -2336,7 +2336,8 @@ int __tcp_retransmit_skb(struct sock *sk, struct sk_buff *skb) * copying overhead: fragmentation, tunneling, mangling etc. */ if (atomic_read(&sk->sk_wmem_alloc) > - min(sk->sk_wmem_queued + (sk->sk_wmem_queued >> 2), sk->sk_sndbuf)) + min_t(u32, sk->sk_wmem_queued + (sk->sk_wmem_queued >> 2), + sk->sk_sndbuf)) return -EAGAIN; if (before(TCP_SKB_CB(skb)->seq, tp->snd_una)) { -- 2.10.2
| |