Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1763026AbXIXQmf (ORCPT ); Mon, 24 Sep 2007 12:42:35 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1760916AbXIXQ2i (ORCPT ); Mon, 24 Sep 2007 12:28:38 -0400 Received: from canuck.infradead.org ([209.217.80.40]:58126 "EHLO canuck.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1761183AbXIXQ2g (ORCPT ); Mon, 24 Sep 2007 12:28:36 -0400 Date: Mon, 24 Sep 2007 09:22:38 -0700 From: Greg KH To: linux-kernel@vger.kernel.org, stable@kernel.org Cc: Justin Forbes , Zwane Mwaikambo , "Theodore Ts'o" , Randy Dunlap , Dave Jones , Chuck Wolber , Chris Wedgwood , Michael Krufky , Chuck Ebbert , Domenico Andreoli , torvalds@linux-foundation.org, akpm@linux-foundation.org, alan@lxorguk.ukuu.org.uk, bunk@kernel.org, Ilpo J?rvinen , David Miller Subject: [45/50] Handle snd_una in tcp_cwnd_down() Message-ID: <20070924162238.GT13510@kroah.com> References: <20070924161246.983665021@mini.kroah.org> MIME-Version: 1.0 Content-Type: text/plain; charset=unknown-8bit Content-Disposition: inline; filename="handle-snd_una-in-tcp_cwnd_down.patch" Content-Transfer-Encoding: 8bit In-Reply-To: <20070924161733.GA13510@kroah.com> User-Agent: Mutt/1.5.16 (2007-06-09) X-Bad-Reply: References and In-Reply-To but no 'Re:' in Subject. Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 4028 Lines: 107 From: Ilpo J?rvinen commit 6ee8009e38006da81d2a53da1aaa27365552553e in mainline Subject: [45/50] [PATCH 1/1] [TCP]: Also handle snd_una changes in tcp_cwnd_down tcp_cwnd_down must check for it too as it should be conservative in case of collapse stuff and also when receiver is trying to lie (though it wouldn't be successful anyway). Note: - Separated also is_dupack and do_lost in fast_retransalert * Much cleaner look-and-feel now * This time it really fixes cumulative ACK + many new SACK blocks recovery entry (I claimed this fixes with last patch but it wasn't). TCP will now call tcp_update_scoreboard regardless of is_dupack when in recovery as long as there is enough fackets_out. - Introduce FLAG_SND_UNA_ADVANCED * Some prior_snd_una arguments are unnecessary after it - Added helper FLAG_ANY_PROGRESS to avoid long FLAG...|FLAG... constructs This is a reduced version of a mainline patch. Signed-off-by: Ilpo J?rvinen Cc: David Miller Signed-off-by: Greg Kroah-Hartman --- net/ipv4/tcp_input.c | 18 +++++++++++------- 1 file changed, 11 insertions(+), 7 deletions(-) --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -102,11 +102,13 @@ int sysctl_tcp_abc __read_mostly; #define FLAG_DATA_LOST 0x80 /* SACK detected data lossage. */ #define FLAG_SLOWPATH 0x100 /* Do not skip RFC checks for window update.*/ #define FLAG_ONLY_ORIG_SACKED 0x200 /* SACKs only non-rexmit sent before RTO */ +#define FLAG_SND_UNA_ADVANCED 0x400 /* Snd_una was changed (!= FLAG_DATA_ACKED) */ #define FLAG_ACKED (FLAG_DATA_ACKED|FLAG_SYN_ACKED) #define FLAG_NOT_DUP (FLAG_DATA|FLAG_WIN_UPDATE|FLAG_ACKED) #define FLAG_CA_ALERT (FLAG_DATA_SACKED|FLAG_ECE) #define FLAG_FORWARD_PROGRESS (FLAG_ACKED|FLAG_DATA_SACKED) +#define FLAG_ANY_PROGRESS (FLAG_FORWARD_PROGRESS|FLAG_SND_UNA_ADVANCED) #define IsReno(tp) ((tp)->rx_opt.sack_ok == 0) #define IsFack(tp) ((tp)->rx_opt.sack_ok & 2) @@ -1856,7 +1858,7 @@ static void tcp_cwnd_down(struct sock *s struct tcp_sock *tp = tcp_sk(sk); int decr = tp->snd_cwnd_cnt + 1; - if ((flag&FLAG_FORWARD_PROGRESS) || + if ((flag&FLAG_ANY_PROGRESS) || (IsReno(tp) && !(flag&FLAG_NOT_DUP))) { tp->snd_cwnd_cnt = decr&1; decr >>= 1; @@ -2112,10 +2114,9 @@ tcp_fastretrans_alert(struct sock *sk, u { struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); - int is_dupack = (tp->snd_una == prior_snd_una && - (!(flag&FLAG_NOT_DUP) || - ((flag&FLAG_DATA_SACKED) && - (tp->fackets_out > tp->reordering)))); + int is_dupack = !(flag&(FLAG_SND_UNA_ADVANCED|FLAG_NOT_DUP)); + int do_lost = is_dupack || ((flag&FLAG_DATA_SACKED) && + (tp->fackets_out > tp->reordering)); /* Some technical things: * 1. Reno does not count dupacks (sacked_out) automatically. */ @@ -2199,7 +2200,7 @@ tcp_fastretrans_alert(struct sock *sk, u int acked = prior_packets - tp->packets_out; if (IsReno(tp)) tcp_remove_reno_sacks(sk, acked); - is_dupack = tcp_try_undo_partial(sk, acked); + do_lost = tcp_try_undo_partial(sk, acked); } break; case TCP_CA_Loss: @@ -2264,7 +2265,7 @@ tcp_fastretrans_alert(struct sock *sk, u tcp_set_ca_state(sk, TCP_CA_Recovery); } - if (is_dupack || tcp_head_timedout(sk)) + if (do_lost || tcp_head_timedout(sk)) tcp_update_scoreboard(sk); tcp_cwnd_down(sk, flag); tcp_xmit_retransmit_queue(sk); @@ -2774,6 +2775,9 @@ static int tcp_ack(struct sock *sk, stru if (before(ack, prior_snd_una)) goto old_ack; + if (after(ack, prior_snd_una)) + flag |= FLAG_SND_UNA_ADVANCED; + if (sysctl_tcp_abc) { if (icsk->icsk_ca_state < TCP_CA_CWR) tp->bytes_acked += ack - prior_snd_una; -- - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/