stable.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	stable@vger.kernel.org, Eric Dumazet <edumazet@google.com>,
	Alexei Starovoitov <ast@plumgrid.com>,
	"David S. Miller" <davem@davemloft.net>
Subject: [PATCH 3.12 07/19] tcp: gso: fix truesize tracking
Date: Mon, 18 Nov 2013 10:37:20 -0800	[thread overview]
Message-ID: <20131118183646.248381899@linuxfoundation.org> (raw)
In-Reply-To: <20131118183644.030738675@linuxfoundation.org>

3.12-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Eric Dumazet <edumazet@google.com>

[ Upstream commit 0d08c42cf9a71530fef5ebcfe368f38f2dd0476f ]

commit 6ff50cd55545 ("tcp: gso: do not generate out of order packets")
had an heuristic that can trigger a warning in skb_try_coalesce(),
because skb->truesize of the gso segments were exactly set to mss.

This breaks the requirement that

skb->truesize >= skb->len + truesizeof(struct sk_buff);

It can trivially be reproduced by :

ifconfig lo mtu 1500
ethtool -K lo tso off
netperf

As the skbs are looped into the TCP networking stack, skb_try_coalesce()
warns us of these skb under-estimating their truesize.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Alexei Starovoitov <ast@plumgrid.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 net/ipv4/tcp_offload.c |   13 +++++--------
 1 file changed, 5 insertions(+), 8 deletions(-)

--- a/net/ipv4/tcp_offload.c
+++ b/net/ipv4/tcp_offload.c
@@ -18,6 +18,7 @@ struct sk_buff *tcp_tso_segment(struct s
 				netdev_features_t features)
 {
 	struct sk_buff *segs = ERR_PTR(-EINVAL);
+	unsigned int sum_truesize = 0;
 	struct tcphdr *th;
 	unsigned int thlen;
 	unsigned int seq;
@@ -102,13 +103,7 @@ struct sk_buff *tcp_tso_segment(struct s
 		if (copy_destructor) {
 			skb->destructor = gso_skb->destructor;
 			skb->sk = gso_skb->sk;
-			/* {tcp|sock}_wfree() use exact truesize accounting :
-			 * sum(skb->truesize) MUST be exactly be gso_skb->truesize
-			 * So we account mss bytes of 'true size' for each segment.
-			 * The last segment will contain the remaining.
-			 */
-			skb->truesize = mss;
-			gso_skb->truesize -= mss;
+			sum_truesize += skb->truesize;
 		}
 		skb = skb->next;
 		th = tcp_hdr(skb);
@@ -125,7 +120,9 @@ struct sk_buff *tcp_tso_segment(struct s
 	if (copy_destructor) {
 		swap(gso_skb->sk, skb->sk);
 		swap(gso_skb->destructor, skb->destructor);
-		swap(gso_skb->truesize, skb->truesize);
+		sum_truesize += skb->truesize;
+		atomic_add(sum_truesize - gso_skb->truesize,
+			   &skb->sk->sk_wmem_alloc);
 	}
 
 	delta = htonl(oldlen + (skb_tail_pointer(skb) -



  parent reply	other threads:[~2013-11-18 18:37 UTC|newest]

Thread overview: 25+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-11-18 18:37 [PATCH 3.12 00/19] 3.12.1-stable review Greg Kroah-Hartman
2013-11-18 18:37 ` [PATCH 3.12 01/19] net/mlx4_core: Fix call to __mlx4_unregister_mac Greg Kroah-Hartman
2013-11-18 18:37 ` [PATCH 3.12 02/19] net: sctp: do not trigger BUG_ON in sctp_cmd_delete_tcb Greg Kroah-Hartman
2013-11-18 18:37 ` [PATCH 3.12 03/19] net: flow_dissector: fail on evil iph->ihl Greg Kroah-Hartman
2013-11-18 18:37 ` [PATCH 3.12 04/19] virtio-net: correctly handle cpu hotplug notifier during resuming Greg Kroah-Hartman
2013-11-18 18:37 ` [PATCH 3.12 05/19] xen-netback: use jiffies_64 value to calculate credit timeout Greg Kroah-Hartman
2013-11-18 18:37 ` [PATCH 3.12 06/19] cxgb3: Fix length calculation in write_ofld_wr() on 32-bit architectures Greg Kroah-Hartman
2013-11-18 18:37 ` Greg Kroah-Hartman [this message]
2013-11-28  5:25   ` [PATCH 3.12 07/19] tcp: gso: fix truesize tracking Ben Hutchings
2013-12-05 21:31     ` David Miller
2013-11-18 18:37 ` [PATCH 3.12 08/19] tcp: fix SYNACK RTT estimation in Fast Open Greg Kroah-Hartman
2013-11-18 18:37 ` [PATCH 3.12 09/19] tcp: only take RTT from timestamps if new data is acked Greg Kroah-Hartman
2013-11-18 18:37 ` [PATCH 3.12 10/19] tcp: do not rearm RTO when future data are sacked Greg Kroah-Hartman
2013-11-18 18:37 ` [PATCH 3.12 11/19] ipv6: ip6_dst_check needs to check for expired dst_entries Greg Kroah-Hartman
2013-11-18 18:37 ` [PATCH 3.12 12/19] ipv6: reset dst.expires value when clearing expire flag Greg Kroah-Hartman
2013-11-18 18:37 ` [PATCH 3.12 13/19] hyperv-fb: add pci stub Greg Kroah-Hartman
2013-11-18 18:37 ` [PATCH 3.12 14/19] USB: add new zte 3g-dongles pid to option.c Greg Kroah-Hartman
2013-11-18 18:37 ` [PATCH 3.12 15/19] ALSA: hda - hdmi: Fix reported channel map on common default layouts Greg Kroah-Hartman
2013-11-18 18:37 ` [PATCH 3.12 16/19] tracing: Fix potential out-of-bounds in trace_get_user() Greg Kroah-Hartman
2013-11-18 18:37 ` [PATCH 3.12 17/19] misc: atmel_pwm: add deferred-probing support Greg Kroah-Hartman
2013-11-18 18:37 ` [PATCH 3.12 18/19] backlight: atmel-pwm-bl: fix deferred probe from __init Greg Kroah-Hartman
2013-11-18 18:37 ` [PATCH 3.12 19/19] usbcore: set lpm_capable field for LPM capable root hubs Greg Kroah-Hartman
2013-11-19  3:11 ` [PATCH 3.12 00/19] 3.12.1-stable review Guenter Roeck
2013-11-20 11:07 ` Satoru Takeuchi
2013-11-20 15:25 ` Shuah Khan

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20131118183646.248381899@linuxfoundation.org \
    --to=gregkh@linuxfoundation.org \
    --cc=ast@plumgrid.com \
    --cc=davem@davemloft.net \
    --cc=edumazet@google.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=stable@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).