From mboxrd@z Thu Jan 1 00:00:00 1970 From: Neal Cardwell Subject: [PATCH] tcp: fix false reordering signal in tcp_shifted_skb Date: Sun, 26 Feb 2012 15:06:19 -0500 Message-ID: <1330286779-10462-1-git-send-email-ncardwell@google.com> Mime-Version: 1.0 Cc: netdev@vger.kernel.org, ilpo.jarvinen@helsinki.fi, Nandita Dukkipati , Yuchung Cheng , Tom Herbert , Vijay Subramanian , Neal Cardwell To: David Miller Return-path: Received: from mail-gx0-f202.google.com ([209.85.161.202]:41966 "EHLO mail-gx0-f202.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752701Ab2BZUMn (ORCPT ); Sun, 26 Feb 2012 15:12:43 -0500 Received: by ggnp1 with SMTP id p1so464079ggn.1 for ; Sun, 26 Feb 2012 12:12:43 -0800 (PST) Sender: netdev-owner@vger.kernel.org List-ID: When tcp_shifted_skb() shifts bytes from the skb that is currently pointed to by 'highest_sack' then the increment of TCP_SKB_CB(skb)->seq implicitly advances tcp_highest_sack_seq(). This implicit advancement, combined with the recent fix to pass the correct SACKed range into tcp_sacktag_one(), caused tcp_sacktag_one() to think that the newly SACKed range was before the tcp_highest_sack_seq(), leading to a call to tcp_update_reordering() with a degree of reordering matching the size of the newly SACKed range (typically just 1 packet, which is a NOP, but potentially larger). This commit fixes this by simply calling tcp_sacktag_one() before the TCP_SKB_CB(skb)->seq advancement that can advance our notion of the highest SACKed sequence. Correspondingly, we can simplify the code a little now that tcp_shifted_skb() should update the lost_cnt_hint in all cases where skb == tp->lost_skb_hint. Signed-off-by: Neal Cardwell --- net/ipv4/tcp_input.c | 18 ++++++++++-------- 1 files changed, 10 insertions(+), 8 deletions(-) diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 53c8ce4..ee42d42 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -1403,8 +1403,16 @@ static int tcp_shifted_skb(struct sock *sk, struct sk_buff *skb, BUG_ON(!pcount); - /* Adjust hint for FACK. Non-FACK is handled in tcp_sacktag_one(). */ - if (tcp_is_fack(tp) && (skb == tp->lost_skb_hint)) + /* Adjust counters and hints for the newly sacked sequence + * range but discard the return value since prev is already + * marked. We must tag the range first because the seq + * advancement below implicitly advances + * tcp_highest_sack_seq() when skb is highest_sack. + */ + tcp_sacktag_one(sk, state, TCP_SKB_CB(skb)->sacked, + start_seq, end_seq, dup_sack, pcount); + + if (skb == tp->lost_skb_hint) tp->lost_cnt_hint += pcount; TCP_SKB_CB(prev)->end_seq += shifted; @@ -1430,12 +1438,6 @@ static int tcp_shifted_skb(struct sock *sk, struct sk_buff *skb, skb_shinfo(skb)->gso_type = 0; } - /* Adjust counters and hints for the newly sacked sequence range but - * discard the return value since prev is already marked. - */ - tcp_sacktag_one(sk, state, TCP_SKB_CB(skb)->sacked, - start_seq, end_seq, dup_sack, pcount); - /* Difference in this won't matter, both ACKed by the same cumul. ACK */ TCP_SKB_CB(prev)->sacked |= (TCP_SKB_CB(skb)->sacked & TCPCB_EVER_RETRANS); -- 1.7.7.3