3.2.87-rc1 review patch. If anyone has any objections, please let me know.
------------------
From: Eric Dumazet <eduma...@google.com>
[ 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 <eduma...@google.com>
Signed-off-by: David S. Miller <da...@davemloft.net>
Signed-off-by: Ben Hutchings <b...@decadent.org.uk>
---
net/ipv4/tcp_output.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -2097,7 +2097,8 @@ int tcp_retransmit_skb(struct sock *sk,
* 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)) {
