summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorYuchung Cheng <ycheng@google.com>2019-01-16 15:05:30 -0800
committerDavid S. Miller <davem@davemloft.net>2019-01-17 15:12:26 -0800
commit7ae189759cc48cf8b54beebff566e9fd2d4e7d7c (patch)
tree352db05710fad943494d48107a6fb4d7555d0a70
parent7f12422c4873e9b274bc151ea59cb0cdf9415cf1 (diff)
tcp: always set retrans_stamp on recovery
Previously TCP socket's retrans_stamp is not set if the retransmission has failed to send. As a result if a socket is experiencing local issues to retransmit packets, determining when to abort a socket is complicated w/o knowning the starting time of the recovery since retrans_stamp may remain zero. This complication causes sub-optimal behavior that TCP may use the latest, instead of the first, retransmission time to compute the elapsed time of a stalling connection due to local issues. Then TCP may disrecard TCP retries settings and keep retrying until it finally succeed: not a good idea when the local host is already strained. The simple fix is to always timestamp the start of a recovery. It's worth noting that retrans_stamp is also used to compare echo timestamp values to detect spurious recovery. This patch does not break that because retrans_stamp is still later than when the original packet was sent. Signed-off-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Reviewed-by: Neal Cardwell <ncardwell@google.com> Reviewed-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
-rw-r--r--net/ipv4/tcp_output.c9
-rw-r--r--net/ipv4/tcp_timer.c23
2 files changed, 7 insertions, 25 deletions
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index 57a56e205070..d2d494c74811 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -2963,13 +2963,12 @@ int tcp_retransmit_skb(struct sock *sk, struct sk_buff *skb, int segs)
#endif
TCP_SKB_CB(skb)->sacked |= TCPCB_RETRANS;
tp->retrans_out += tcp_skb_pcount(skb);
-
- /* Save stamp of the first retransmit. */
- if (!tp->retrans_stamp)
- tp->retrans_stamp = tcp_skb_timestamp(skb);
-
}
+ /* Save stamp of the first (attempted) retransmit. */
+ if (!tp->retrans_stamp)
+ tp->retrans_stamp = tcp_skb_timestamp(skb);
+
if (tp->undo_retrans < 0)
tp->undo_retrans = 0;
tp->undo_retrans += tcp_skb_pcount(skb);
diff --git a/net/ipv4/tcp_timer.c b/net/ipv4/tcp_timer.c
index e7d09e3705b8..1e61f0bd6e24 100644
--- a/net/ipv4/tcp_timer.c
+++ b/net/ipv4/tcp_timer.c
@@ -22,28 +22,14 @@
#include <linux/gfp.h>
#include <net/tcp.h>
-static u32 tcp_retransmit_stamp(const struct sock *sk)
-{
- u32 start_ts = tcp_sk(sk)->retrans_stamp;
-
- if (unlikely(!start_ts)) {
- struct sk_buff *head = tcp_rtx_queue_head(sk);
-
- if (!head)
- return 0;
- start_ts = tcp_skb_timestamp(head);
- }
- return start_ts;
-}
-
static u32 tcp_clamp_rto_to_user_timeout(const struct sock *sk)
{
struct inet_connection_sock *icsk = inet_csk(sk);
u32 elapsed, start_ts;
s32 remaining;
- start_ts = tcp_retransmit_stamp(sk);
- if (!icsk->icsk_user_timeout || !start_ts)
+ start_ts = tcp_sk(sk)->retrans_stamp;
+ if (!icsk->icsk_user_timeout)
return icsk->icsk_rto;
elapsed = tcp_time_stamp(tcp_sk(sk)) - start_ts;
remaining = icsk->icsk_user_timeout - elapsed;
@@ -197,10 +183,7 @@ static bool retransmits_timed_out(struct sock *sk,
if (!inet_csk(sk)->icsk_retransmits)
return false;
- start_ts = tcp_retransmit_stamp(sk);
- if (!start_ts)
- return false;
-
+ start_ts = tcp_sk(sk)->retrans_stamp;
if (likely(timeout == 0)) {
linear_backoff_thresh = ilog2(TCP_RTO_MAX/rto_base);