diff options
| author |   Willem de Bruijn <willemb@google.com> | 2014-08-12 14:53:16 -0400 |
|---|---|---|
| committer |   David S. Miller <davem@davemloft.net> | 2014-08-13 20:06:06 -0700 |
| commit | 712a72213fad36cc9e6ec706b5e020d7eb6e03bc (patch) | |
| tree | 7cc2a560a0b95d500b12c172f9e8b32b1d1321d1 /net/ipv4/tcp_input.c | |
| parent | drivers/net/irda/donauboe.c: convert to module_pci_driver (diff) | |
| download | linux-dev-712a72213fad36cc9e6ec706b5e020d7eb6e03bc.tar.xz linux-dev-712a72213fad36cc9e6ec706b5e020d7eb6e03bc.zip | |
net-timestamp: fix missing ACK timestamp
ACK timestamps are generated in tcp_clean_rtx_queue. The TSO datapath
can break out early, causing the timestamp code to be skipped. Move
the code up before the break.
Reported-by: David S. Miller <davem@davemloft.net>
Also fix a boundary condition: tp->snd_una is the next unacknowledged
byte and between tests inclusive (a <= b <= c), so generate a an ACK
timestamp if (prior_snd_una <= tskey <= tp->snd_una - 1).
Signed-off-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'net/ipv4/tcp_input.c')
| -rw-r--r-- | net/ipv4/tcp_input.c | 10 |
1 files changed, 5 insertions, 5 deletions
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index a3d47af01906..1a8e89fdd331 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -3050,10 +3050,15 @@ static int tcp_clean_rtx_queue(struct sock *sk, int prior_fackets, first_ackt.v64 = 0; while ((skb = tcp_write_queue_head(sk)) && skb != tcp_send_head(sk)) { + struct skb_shared_info *shinfo = skb_shinfo(skb); struct tcp_skb_cb *scb = TCP_SKB_CB(skb); u8 sacked = scb->sacked; u32 acked_pcount; + if (unlikely(shinfo->tx_flags & SKBTX_ACK_TSTAMP) && + between(shinfo->tskey, prior_snd_una, tp->snd_una - 1)) + __skb_tstamp_tx(skb, NULL, sk, SCM_TSTAMP_ACK); + /* Determine how many packets and what bytes were acked, tso and else */ if (after(scb->end_seq, tp->snd_una)) { if (tcp_skb_pcount(skb) == 1 || @@ -3107,11 +3112,6 @@ static int tcp_clean_rtx_queue(struct sock *sk, int prior_fackets, tp->retrans_stamp = 0; } - if (unlikely(skb_shinfo(skb)->tx_flags & SKBTX_ACK_TSTAMP) && - between(skb_shinfo(skb)->tskey, prior_snd_una, - tp->snd_una + 1)) - __skb_tstamp_tx(skb, NULL, sk, SCM_TSTAMP_ACK); - if (!fully_acked) break; |