tcp: fix false reordering signal in tcp_shifted_skb

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 <ncardwell@google.com>
---
 net/ipv4/tcp_input.c |   18 ++++++++++--------
 1 files changed, 10 insertions(+), 8 deletions(-)

From: Neal Cardwell <ncardwell@google.com>
Date: Sun, 26 Feb 2012 15:06:19 -0500

> 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 <ncardwell@google.com>

Applied and queued up for -stable, thanks Neal.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

On Sun, Feb 26, 2012 at 12:06 PM, Neal Cardwell <ncardwell@google.com> wrote:
> 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 <ncardwell@google.com>
> ---
>  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
>
Acked-by: Yuchung Cheng <ycheng@google.com>
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html