From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1161767AbbEEVHD (ORCPT ); Tue, 5 May 2015 17:07:03 -0400 Received: from youngberry.canonical.com ([91.189.89.112]:36582 "EHLO youngberry.canonical.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1161125AbbEEU4A (ORCPT ); Tue, 5 May 2015 16:56:00 -0400 From: Kamal Mostafa To: linux-kernel@vger.kernel.org, stable@vger.kernel.org, kernel-team@lists.ubuntu.com Cc: Neal Cardwell , Yuchung Cheng , "David S. Miller" , Kamal Mostafa Subject: [PATCH 3.13.y-ckt 60/90] tcp: fix FRTO undo on cumulative ACK of SACKed range Date: Tue, 5 May 2015 13:54:32 -0700 Message-Id: <1430859302-5886-61-git-send-email-kamal@canonical.com> X-Mailer: git-send-email 1.9.1 In-Reply-To: <1430859302-5886-1-git-send-email-kamal@canonical.com> References: <1430859302-5886-1-git-send-email-kamal@canonical.com> X-Extended-Stable: 3.13 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org 3.13.11-ckt20 -stable review patch. If anyone has any objections, please let me know. ------------------ From: Neal Cardwell [ Upstream commit 666b805150efd62f05810ff0db08f44a2370c937 ] On processing cumulative ACKs, the FRTO code was not checking the SACKed bit, meaning that there could be a spurious FRTO undo on a cumulative ACK of a previously SACKed skb. The FRTO code should only consider a cumulative ACK to indicate that an original/unretransmitted skb is newly ACKed if the skb was not yet SACKed. The effect of the spurious FRTO undo would typically be to make the connection think that all previously-sent packets were in flight when they really weren't, leading to a stall and an RTO. Signed-off-by: Neal Cardwell Signed-off-by: Yuchung Cheng Fixes: e33099f96d99c ("tcp: implement RFC5682 F-RTO") Signed-off-by: David S. Miller Signed-off-by: Kamal Mostafa --- net/ipv4/tcp_input.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 2ab6b821..2cc1313 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -3060,10 +3060,11 @@ static int tcp_clean_rtx_queue(struct sock *sk, int prior_fackets, if (seq_rtt < 0) { seq_rtt = ca_seq_rtt; } - if (!(sacked & TCPCB_SACKED_ACKED)) + if (!(sacked & TCPCB_SACKED_ACKED)) { reord = min(pkts_acked, reord); - if (!after(scb->end_seq, tp->high_seq)) - flag |= FLAG_ORIG_SACK_ACKED; + if (!after(scb->end_seq, tp->high_seq)) + flag |= FLAG_ORIG_SACK_ACKED; + } } if (sacked & TCPCB_SACKED_ACKED) -- 1.9.1