public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org, stable@vger.kernel.org
Cc: Greg KH <gregkh@linuxfoundation.org>,
	torvalds@linux-foundation.org, akpm@linux-foundation.org,
	alan@lxorguk.ukuu.org.uk, Eric Dumazet <edumazet@google.com>,
	Tom Herbert <therbert@google.com>,
	John Fastabend <john.r.fastabend@intel.com>,
	Roland Dreier <roland@kernel.org>,
	Neil Horman <nhorman@tuxdriver.com>,
	"David S. Miller" <davem@davemloft.net>
Subject: [ 11/68] bonding: Fix corrupted queue_mapping
Date: Thu, 12 Jul 2012 16:01:45 -0700	[thread overview]
Message-ID: <20120712175036.405029842@linuxfoundation.org> (raw)
In-Reply-To: <20120712175035.530652872@linuxfoundation.org>

From: Greg KH <gregkh@linuxfoundation.org>

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

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


From: Eric Dumazet <edumazet@google.com>

[ Upstream commit 5ee31c6898ea5537fcea160999d60dc63bc0c305 ]

In the transmit path of the bonding driver, skb->cb is used to
stash the skb->queue_mapping so that the bonding device can set its
own queue mapping.  This value becomes corrupted since the skb->cb is
also used in __dev_xmit_skb.

When transmitting through bonding driver, bond_select_queue is
called from dev_queue_xmit.  In bond_select_queue the original
skb->queue_mapping is copied into skb->cb (via bond_queue_mapping)
and skb->queue_mapping is overwritten with the bond driver queue.

Subsequently in dev_queue_xmit, __dev_xmit_skb is called which writes
the packet length into skb->cb, thereby overwriting the stashed
queue mappping.  In bond_dev_queue_xmit (called from hard_start_xmit),
the queue mapping for the skb is set to the stashed value which is now
the skb length and hence is an invalid queue for the slave device.

If we want to save skb->queue_mapping into skb->cb[], best place is to
add a field in struct qdisc_skb_cb, to make sure it wont conflict with
other layers (eg : Qdiscc, Infiniband...)

This patchs also makes sure (struct qdisc_skb_cb)->data is aligned on 8
bytes :

netem qdisc for example assumes it can store an u64 in it, without
misalignment penalty.

Note : we only have 20 bytes left in (struct qdisc_skb_cb)->data[].
The largest user is CHOKe and it fills it.

Based on a previous patch from Tom Herbert.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Tom Herbert <therbert@google.com>
Cc: John Fastabend <john.r.fastabend@intel.com>
Cc: Roland Dreier <roland@kernel.org>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 drivers/net/bonding/bond_main.c |    9 +++++----
 include/net/sch_generic.h       |    7 +++++--
 2 files changed, 10 insertions(+), 6 deletions(-)

--- a/drivers/net/bonding/bond_main.c
+++ b/drivers/net/bonding/bond_main.c
@@ -77,6 +77,7 @@
 #include <net/route.h>
 #include <net/net_namespace.h>
 #include <net/netns/generic.h>
+#include <net/pkt_sched.h>
 #include "bonding.h"
 #include "bond_3ad.h"
 #include "bond_alb.h"
@@ -388,8 +389,6 @@ struct vlan_entry *bond_next_vlan(struct
 	return next;
 }
 
-#define bond_queue_mapping(skb) (*(u16 *)((skb)->cb))
-
 /**
  * bond_dev_queue_xmit - Prepare skb for xmit.
  *
@@ -403,7 +402,9 @@ int bond_dev_queue_xmit(struct bonding *
 	skb->dev = slave_dev;
 	skb->priority = 1;
 
-	skb->queue_mapping = bond_queue_mapping(skb);
+	BUILD_BUG_ON(sizeof(skb->queue_mapping) !=
+		     sizeof(qdisc_skb_cb(skb)->bond_queue_mapping));
+	skb->queue_mapping = qdisc_skb_cb(skb)->bond_queue_mapping;
 
 	if (unlikely(netpoll_tx_running(slave_dev)))
 		bond_netpoll_send_skb(bond_get_slave_by_dev(bond, slave_dev), skb);
@@ -4240,7 +4241,7 @@ static u16 bond_select_queue(struct net_
 	/*
 	 * Save the original txq to restore before passing to the driver
 	 */
-	bond_queue_mapping(skb) = skb->queue_mapping;
+	qdisc_skb_cb(skb)->bond_queue_mapping = skb->queue_mapping;
 
 	if (unlikely(txq >= dev->real_num_tx_queues)) {
 		do {
--- a/include/net/sch_generic.h
+++ b/include/net/sch_generic.h
@@ -219,13 +219,16 @@ struct tcf_proto {
 
 struct qdisc_skb_cb {
 	unsigned int		pkt_len;
-	unsigned char		data[24];
+	u16			bond_queue_mapping;
+	u16			_pad;
+	unsigned char		data[20];
 };
 
 static inline void qdisc_cb_private_validate(const struct sk_buff *skb, int sz)
 {
 	struct qdisc_skb_cb *qcb;
-	BUILD_BUG_ON(sizeof(skb->cb) < sizeof(unsigned int) + sz);
+
+	BUILD_BUG_ON(sizeof(skb->cb) < offsetof(struct qdisc_skb_cb, data) + sz);
 	BUILD_BUG_ON(sizeof(qcb->data) < sz);
 }
 



  parent reply	other threads:[~2012-07-12 23:23 UTC|newest]

Thread overview: 77+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-07-12 17:50 [ 00/68] 3.0.37-stable review Greg KH
2012-07-12 23:01 ` [ 01/68] ALSA: hda - Add Realtek ALC280 codec support Greg Kroah-Hartman
2012-07-12 23:01   ` [ 02/68] powerpc/xmon: Use cpumask iterator to avoid warning Greg Kroah-Hartman
2012-07-12 23:01   ` [ 03/68] media: smsusb: add autodetection support for USB ID 2040:f5a0 Greg Kroah-Hartman
2012-07-12 23:01   ` [ 04/68] ARM: fix rcu stalls on SMP platforms Greg Kroah-Hartman
2012-07-12 23:01   ` [ 05/68] net: sock: validate data_len before allocating skb in sock_alloc_send_pskb() Greg Kroah-Hartman
2012-07-12 23:01   ` [ 06/68] cipso: handle CIPSO options correctly when NetLabel is disabled Greg Kroah-Hartman
2012-07-12 23:01   ` [ 07/68] net: l2tp_eth: fix kernel panic on rmmod l2tp_eth Greg Kroah-Hartman
2012-07-12 23:01   ` [ 08/68] dummy: fix rcu_sched self-detected stalls Greg Kroah-Hartman
2012-07-13 14:09     ` Herton Ronaldo Krzesinski
2012-07-16 15:46       ` Greg Kroah-Hartman
2012-07-12 23:01   ` [ 09/68] ethtool: allow ETHTOOL_GSSET_INFO for users Greg Kroah-Hartman
2012-07-12 23:01   ` [ 10/68] bridge: Assign rtnl_link_ops to bridge devices created via ioctl (v2) Greg Kroah-Hartman
2012-07-12 23:01   ` Greg Kroah-Hartman [this message]
2012-07-12 23:01   ` [ 12/68] ipv6: Move ipv6 proc file registration to end of init order Greg Kroah-Hartman
2012-07-12 23:01   ` [ 13/68] sky2: fix checksum bit management on some chips Greg Kroah-Hartman
2012-07-12 23:01   ` [ 14/68] be2net: fix a race in be_xmit() Greg Kroah-Hartman
2012-07-12 23:01   ` [ 15/68] netpoll: fix netpoll_send_udp() bugs Greg Kroah-Hartman
2012-07-12 23:01   ` [ 16/68] hwmon: (applesmc) Limit key length in warning messages Greg Kroah-Hartman
2012-07-12 23:01   ` [ 17/68] nilfs2: ensure proper cache clearing for gc-inodes Greg Kroah-Hartman
2012-07-12 23:01   ` [ 18/68] udf: Use ret instead of abusing i in udf_load_logicalvol() Greg Kroah-Hartman
2012-07-12 23:01   ` [ 19/68] udf: Avoid run away loop when partition table length is corrupted Greg Kroah-Hartman
2012-07-12 23:01   ` [ 20/68] udf: Fortify loading of sparing table Greg Kroah-Hartman
2012-07-12 23:01   ` [ 21/68] ath9k: Fix softlockup in AR9485 Greg Kroah-Hartman
2012-07-12 23:01   ` [ 22/68] ath9k_hw: avoid possible infinite loop in ar9003_get_pll_sqsum_dvc Greg Kroah-Hartman
2012-07-13  1:38     ` Herton Ronaldo Krzesinski
2012-07-13  1:49       ` Greg Kroah-Hartman
2012-07-12 23:01   ` [ 23/68] ath9k: enable serialize_regmode for non-PCIE AR9287 Greg Kroah-Hartman
2012-07-12 23:01   ` [ 24/68] ASoC: tlv320aic3x: Fix codec pll configure bug Greg Kroah-Hartman
2012-07-12 23:01   ` [ 25/68] Btrfs: run delayed directory updates during log replay Greg Kroah-Hartman
2012-07-12 23:02   ` [ 26/68] drm/edid: dont return stack garbage from supports_rb Greg Kroah-Hartman
2012-07-12 23:02   ` [ 27/68] drm/nouveau/fbcon: using nv_two_heads is not a good idea Greg Kroah-Hartman
2012-07-12 23:02   ` [ 28/68] drm/i915: Fix eDP blank screen after S3 resume on HP desktops Greg Kroah-Hartman
2012-07-12 23:02   ` [ 29/68] acpi_pad: fix power_saving thread deadlock Greg Kroah-Hartman
2012-07-12 23:02   ` [ 30/68] ACPI: Add a quirk for "AMILO PRO V2030" to ignore the timer overriding Greg Kroah-Hartman
2012-07-12 23:02   ` [ 31/68] ACPI, x86: fix Dell M6600 ACPI reboot regression via DMI Greg Kroah-Hartman
2012-07-12 23:02   ` [ 32/68] ACPI sysfs.c strlen fix Greg Kroah-Hartman
2012-07-12 23:02   ` [ 33/68] stable: Allow merging of backports for serious user-visible performance issues Greg Kroah-Hartman
2012-07-12 23:02   ` [ 34/68] USB: option: add id for Cellient MEN-200 Greg Kroah-Hartman
2012-07-12 23:02   ` [ 35/68] USB: option: Add USB ID for Novatel Ovation MC551 Greg Kroah-Hartman
2012-07-12 23:02   ` [ 36/68] USB: CP210x Add 10 Device IDs Greg Kroah-Hartman
2012-07-12 23:02   ` [ 37/68] cfg80211: fix potential deadlock in regulatory Greg Kroah-Hartman
2012-07-12 23:02   ` [ 38/68] can: c_can: precedence error in c_can_chip_config() Greg Kroah-Hartman
2012-07-12 23:02   ` [ 39/68] oprofile: perf: use NR_CPUS instead or nr_cpumask_bits for static array Greg Kroah-Hartman
2012-07-12 23:02   ` [ 40/68] mac80211: correct behaviour on unrecognised action frames Greg Kroah-Hartman
2012-07-12 23:02   ` [ 41/68] mwifiex: fix 11n rx packet drop issue Greg Kroah-Hartman
2012-07-12 23:02   ` [ 42/68] vfs: make O_PATH file descriptors usable for fchdir() Greg Kroah-Hartman
2012-07-12 22:40     ` ольга крыжановская
2012-07-12 23:02   ` [ 43/68] mtd: cafe_nand: fix an & vs | mistake Greg Kroah-Hartman
2012-07-12 23:02   ` [ 44/68] tcm_fc: Resolve suspicious RCU usage warnings Greg Kroah-Hartman
2012-07-12 23:02   ` [ 45/68] eCryptfs: Gracefully refuse miscdev file ops on inherited/passed files Greg Kroah-Hartman
2012-07-12 23:02   ` [ 46/68] eCryptfs: Fix lockdep warning in miscdev operations Greg Kroah-Hartman
2012-07-12 23:02   ` [ 47/68] eCryptfs: Properly check for O_RDONLY flag before doing privileged open Greg Kroah-Hartman
2012-07-12 23:02   ` [ 48/68] USB: cdc-wdm: fix lockup on error in wdm_read Greg Kroah-Hartman
2012-07-12 23:02   ` [ 49/68] USB: option: add ZTE MF60 Greg Kroah-Hartman
2012-07-12 23:02   ` [ 50/68] USB: option: Add MEDIATEK product ids Greg Kroah-Hartman
2012-07-12 23:02   ` [ 51/68] PCI: EHCI: fix crash during suspend on ASUS computers Greg Kroah-Hartman
2012-07-13  1:42     ` Herton Ronaldo Krzesinski
2012-07-12 23:02   ` [ 52/68] xhci: Avoid dead ports when CONFIG_USB_XHCI_HCD=n Greg Kroah-Hartman
2012-07-12 23:02   ` [ 53/68] ipheth: add support for iPad Greg Kroah-Hartman
2012-07-12 23:02   ` [ 54/68] tracing: change CPU ring buffer state from tracing_cpumask Greg Kroah-Hartman
2012-07-13  1:47     ` Herton Ronaldo Krzesinski
2012-07-12 23:02   ` [ 55/68] vhost: dont forget to schedule() Greg Kroah-Hartman
2012-07-12 23:02   ` [ 56/68] raid5: delayed stripe fix Greg Kroah-Hartman
2012-07-12 23:02   ` [ 57/68] rtl8187: ->brightness_set can not sleep Greg Kroah-Hartman
2012-07-12 23:02   ` [ 58/68] umem: fix up unplugging Greg Kroah-Hartman
2012-07-12 23:02   ` [ 59/68] x86, cpufeature: Rename X86_FEATURE_DTS to X86_FEATURE_DTHERM Greg Kroah-Hartman
2012-07-12 23:02   ` [ 60/68] md/raid5: Do not add data_offset before call to is_badblock Greg Kroah-Hartman
2012-07-12 23:02   ` [ 61/68] md/raid10: Dont try to recovery unmatched (and unused) chunks Greg Kroah-Hartman
2012-07-12 23:02   ` [ 62/68] memory hotplug: fix invalid memory access caused by stale kswapd pointer Greg Kroah-Hartman
2012-07-12 23:02   ` [ 63/68] drivers/rtc/rtc-mxc.c: fix irq enabled interrupts warning Greg Kroah-Hartman
2012-07-12 23:02   ` [ 64/68] mm, thp: abort compaction if migration page cannot be charged to memcg Greg Kroah-Hartman
2012-07-12 23:02   ` [ 65/68] fs: ramfs: file-nommu: add SetPageUptodate() Greg Kroah-Hartman
2012-07-12 23:02   ` [ 66/68] mm: Hold a file reference in madvise_remove Greg Kroah-Hartman
2012-07-13  1:49     ` Herton Ronaldo Krzesinski
2012-07-12 23:02   ` [ 67/68] ACPI: Make acpi_skip_timer_override cover all source_irq==0 cases Greg Kroah-Hartman
2012-07-12 23:02   ` [ 68/68] ACPI: Remove one board specific WARN when ignoring timer overriding Greg Kroah-Hartman

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=20120712175036.405029842@linuxfoundation.org \
    --to=gregkh@linuxfoundation.org \
    --cc=akpm@linux-foundation.org \
    --cc=alan@lxorguk.ukuu.org.uk \
    --cc=davem@davemloft.net \
    --cc=edumazet@google.com \
    --cc=john.r.fastabend@intel.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=nhorman@tuxdriver.com \
    --cc=roland@kernel.org \
    --cc=stable@vger.kernel.org \
    --cc=therbert@google.com \
    --cc=torvalds@linux-foundation.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