{"id":751,"url":"http://patchwork.ozlabs.org/api/1.0/patches/751/?format=json","project":{"id":7,"url":"http://patchwork.ozlabs.org/api/1.0/projects/7/?format=json","name":"Linux network development","link_name":"netdev","list_id":"netdev.vger.kernel.org","list_email":"netdev@vger.kernel.org","web_url":null,"scm_url":null,"webscm_url":null},"msgid":"<1221947278-16715-15-git-send-email-ilpo.jarvinen@helsinki.fi>","date":"2008-09-20T21:47:57","name":"[PATCHv2,net-next,14/15] tcp: don't clear lost_skb_hint when not necessary","commit_ref":null,"pull_url":null,"state":"accepted","archived":true,"hash":"2b9b842e8141f8bbc6b8f209e1668ce4d07a7583","submitter":{"id":255,"url":"http://patchwork.ozlabs.org/api/1.0/people/255/?format=json","name":"Ilpo Järvinen","email":"ilpo.jarvinen@helsinki.fi"},"delegate":{"id":34,"url":"http://patchwork.ozlabs.org/api/1.0/users/34/?format=json","username":"davem","first_name":"David","last_name":"Miller","email":"davem@davemloft.net"},"mbox":"http://patchwork.ozlabs.org/project/netdev/patch/1221947278-16715-15-git-send-email-ilpo.jarvinen@helsinki.fi/mbox/","series":[],"check":"pending","checks":"http://patchwork.ozlabs.org/api/patches/751/checks/","tags":{},"headers":{"Return-Path":"","X-Original-To":"patchwork-incoming@ozlabs.org","Delivered-To":"patchwork-incoming@ozlabs.org","Received":["from vger.kernel.org (vger.kernel.org [209.132.176.167])\n\tby ozlabs.org (Postfix) with ESMTP id 8115EDDEDE\n\tfor ;\n\tSun, 21 Sep 2008 07:48:49 +1000 (EST)","(majordomo@vger.kernel.org) by vger.kernel.org via listexpand\n\tid S1754123AbYITVsi (ORCPT );\n\tSat, 20 Sep 2008 17:48:38 -0400","(majordomo@vger.kernel.org) by vger.kernel.org id S1754106AbYITVsh\n\t(ORCPT ); Sat, 20 Sep 2008 17:48:37 -0400","from courier.cs.helsinki.fi ([128.214.9.1]:56256 \"EHLO\n\tmail.cs.helsinki.fi\" rhost-flags-OK-OK-OK-OK) by vger.kernel.org\n\twith ESMTP id S1751931AbYITVsI (ORCPT\n\t); Sat, 20 Sep 2008 17:48:08 -0400","from wrl-59.cs.helsinki.fi (wrl-59.cs.helsinki.fi\n\t[128.214.166.179])\n\t(AUTH: PLAIN cs-relay, TLS: TLSv1/SSLv3,256bits,AES256-SHA)\n\tby mail.cs.helsinki.fi with esmtp; Sun, 21 Sep 2008 00:47:59 +0300\n\tid 0005BEF2.48D56F8F.0000546E","by wrl-59.cs.helsinki.fi (Postfix, from userid 50795)\n\tid 197D1A00A4; Sun, 21 Sep 2008 00:47:59 +0300 (EEST)"],"From":"\"=?ISO-8859-1?Q?Ilpo_J=E4rvinen?=\" ","To":"David Miller ","Cc":"netdev@vger.kernel.org,\n\t\"=?utf-8?q?Ilpo=20J=E4rvinen?=\" ","Subject":"[PATCHv2 net-next 14/15] tcp: don't clear lost_skb_hint when not\n\tnecessary","Date":"Sun, 21 Sep 2008 00:47:57 +0300","Message-Id":"<1221947278-16715-15-git-send-email-ilpo.jarvinen@helsinki.fi>","X-Mailer":"git-send-email 1.5.4.2.156.ge3c5","In-Reply-To":"<1221947278-16715-14-git-send-email-ilpo.jarvinen@helsinki.fi>","References":"<1221947278-16715-1-git-send-email-ilpo.jarvinen@helsinki.fi>\n\t<1221947278-16715-2-git-send-email-ilpo.jarvinen@helsinki.fi>\n\t<1221947278-16715-3-git-send-email-ilpo.jarvinen@helsinki.fi>\n\t<1221947278-16715-4-git-send-email-ilpo.jarvinen@helsinki.fi>\n\t<1221947278-16715-5-git-send-email-ilpo.jarvinen@helsinki.fi>\n\t<1221947278-16715-6-git-send-email-ilpo.jarvinen@helsinki.fi>\n\t<1221947278-16715-7-git-send-email-ilpo.jarvinen@helsinki.fi>\n\t<1221947278-16715-8-git-send-email-ilpo.jarvinen@helsinki.fi>\n\t<1221947278-16715-9-git-send-email-ilpo.jarvinen@helsinki.fi>\n\t<1221947278-16715-10-git-send-email-ilpo.jarvinen@helsinki.fi>\n\t<1221947278-16715-11-git-send-email-ilpo.jarvinen@helsinki.fi>\n\t<1221947278-16715-12-git-send-email-ilpo.jarvinen@helsinki.fi>\n\t<1221947278-16715-13-git-send-email-ilpo.jarvinen@helsinki.fi>\n\t<1221947278-16715-14-git-send-email-ilpo.jarvinen@helsinki.fi>","MIME-Version":"1.0","Content-Type":"text/plain; charset=ISO-8859-1","Content-Transfer-Encoding":"8bit","Sender":"netdev-owner@vger.kernel.org","Precedence":"bulk","List-ID":"","X-Mailing-List":"netdev@vger.kernel.org"},"content":"Most importantly avoid doing it with cumulative ACK. However,\nsince we have lost_cnt_hint in the picture as well needing\nadjustments, it's not as trivial as dealing with\nretransmit_skb_hint (and cannot be done in the all place we\ncould trivially leave retransmit_skb_hint untouched).\n\nWith the previous patch, this should mostly remove O(n^2)\nbehavior while cumulative ACKs start flowing once rexmit\nafter a lossy round-trip made it through.\n\nSigned-off-by: Ilpo Järvinen \n---\n net/ipv4/tcp_input.c | 14 +++++++++++++-\n 1 files changed, 13 insertions(+), 1 deletions(-)","diff":"diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c\nindex 44a4fff..85627f8 100644\n--- a/net/ipv4/tcp_input.c\n+++ b/net/ipv4/tcp_input.c\n@@ -2844,6 +2844,7 @@ static int tcp_clean_rtx_queue(struct sock *sk, int prior_fackets)\n \tint flag = 0;\n \tu32 pkts_acked = 0;\n \tu32 reord = tp->packets_out;\n+\tu32 prior_sacked = tp->sacked_out;\n \ts32 seq_rtt = -1;\n \ts32 ca_seq_rtt = -1;\n \tktime_t last_ackt = net_invalid_timestamp();\n@@ -2925,9 +2926,11 @@ static int tcp_clean_rtx_queue(struct sock *sk, int prior_fackets)\n \n \t\ttcp_unlink_write_queue(skb, sk);\n \t\tsk_wmem_free_skb(sk, skb);\n-\t\ttcp_clear_retrans_hints_partial(tp);\n+\t\ttp->scoreboard_skb_hint = NULL;\n \t\tif (skb == tp->retransmit_skb_hint)\n \t\t\ttp->retransmit_skb_hint = NULL;\n+\t\tif (skb == tp->lost_skb_hint)\n+\t\t\ttp->lost_skb_hint = NULL;\n \t}\n \n \tif (skb && (TCP_SKB_CB(skb)->sacked & TCPCB_SACKED_ACKED))\n@@ -2946,6 +2949,15 @@ static int tcp_clean_rtx_queue(struct sock *sk, int prior_fackets)\n \t\t\t/* Non-retransmitted hole got filled? That's reordering */\n \t\t\tif (reord < prior_fackets)\n \t\t\t\ttcp_update_reordering(sk, tp->fackets_out - reord, 0);\n+\n+\t\t\t/* No need to care for underflows here because\n+\t\t\t * the lost_skb_hint gets NULLed if we're past it\n+\t\t\t * (or something non-trivial happened)\n+\t\t\t */\n+\t\t\tif (tcp_is_fack(tp))\n+\t\t\t\ttp->lost_cnt_hint -= pkts_acked;\n+\t\t\telse\n+\t\t\t\ttp->lost_cnt_hint -= prior_sacked - tp->sacked_out;\n \t\t}\n \n \t\ttp->fackets_out -= min(pkts_acked, tp->fackets_out);\n","prefixes":["PATCHv2","net-next","14/15"]}