Recently GRO started generating packets with frag_lists of frags.
This was not handled by GSO, thus leading to a crash.

Thankfully these packets are of a regular form and are easy to
handle.  This patch handles them in two ways.  For completely
non-linear frag_list entries, we simply continue to iterate over
the frag_list frags once we exhaust the normal frags.  For frag_list
entries with linear parts, we call pskb_trim on the first part
of the frag_list skb, and then process the rest of the frags in
the usual way.

This patch also kills a chunk of dead frag_list code that has
obviously never ever been run since it ends up generating a bogus
GSO-segmented packet with a frag_list entry.

Future work is planned to split super big packets into TSO
ones.

Fixes: 8a29111c

 ("net: gro: allow to build full sized skb")
Reported-by: Christoph Paasch <christoph.paasch@uclouvain.be>
Reported-by: Jerry Chu <hkchu@google.com>
Reported-by: Sander Eikelenboom <linux@eikelenboom.it>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Tested-by: Sander Eikelenboom <linux@eikelenboom.it>
Tested-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fix another really stupid bug - I introduced genl_set_err()
precisely to be able to adjust the group and reject invalid
ones, but then forgot to do so.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Unfortunately, I introduced a tremendously stupid bug into
genlmsg_multicast() when doing all those multicast group
changes: it adjusts the group number, but then passes it
to genlmsg_multicast_netns() which does that again.

Somehow, my tests failed to catch this, so add a warning
into genlmsg_multicast_netns() and remove the offending
group ID adjustment.

Also add a warning to the similar code in other functions
so people who misuse them are more loudly warned.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Salam reported a use after free bug in PF_PACKET that occurs when
we're sending out frames on a socket bound device and suddenly the
net device is being unregistered. It appears that commit 827d9780
introduced a possible race condition between {t,}packet_snd() and
packet_notifier(). In the case of a bound socket, packet_notifier()
can drop the last reference to the net_device and {t,}packet_snd()
might end up suddenly sending a packet over a freed net_device.

To avoid reverting 827d9780 and thus introducing a performance
regression compared to the current state of things, we decided to
hold a cached RCU protected pointer to the net device and maintain
it on write side via bind spin_lock protected register_prot_hook()
and __unregister_prot_hook() calls.

In {t,}packet_snd() path, we access this pointer under rcu_read_lock
through packet_cached_dev_get() that holds reference to the device
to prevent it from being freed through packet_notifier() while
we're in send path. This is okay to do as dev_put()/dev_hold() are
per-cpu counters, so this should not be a performance issue. Also,
the code simplifies a bit as we don't need need_rls_dev anymore.

Fixes: 827d9780

 ("af-packet: Use existing netdev reference for bound sockets.")
Reported-by: Salam Noureddine <noureddine@aristanetworks.com>
Signed-off-by: Daniel Borkmann <dborkman@redhat.com>
Signed-off-by: Salam Noureddine <noureddine@aristanetworks.com>
Cc: Ben Greear <greearb@candelatech.com>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

If the VIF thread is still running after unbinding the Tx and Rx IRQs
in xenvif_disconnect(), the thread may attempt to raise an event which
will BUG (as the irq is unbound).

Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Acked-by: Wei Liu <wei.liu2@citrix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This removes a code line that is between a "return 0;" and an error label.
This code line can never be reached.

Found by Coverity (CID: 1130529)

Signed-off-by: Michael Opdenacker <michael.opdenacker@free-electrons.com>
Acked-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

John W. Linville says:

====================
pull request: wireless 2013-11-21

Please pull this batch of fixes intended for the 3.13 stream!

For the Bluetooth bits, Gustavo says:

"A few fixes for 3.13. There is 3 fixes to the RFCOMM protocol. One
crash fix to L2CAP. A simple fix to a bad behaviour in the SMP
protocol."

On top of that...

Amitkumar Karwar sends a quintet of mwifiex fixes -- two fixes related
to failure handling, two memory leak fixes, and a NULL pointer fix.

Felix Fietkau corrects and earlier rt2x00 HT descriptor handling fix
to address a crash.

Geyslan G. Bem fixes a memory leak in brcmfmac.

Larry Finger address more pointer arithmetic errors in rtlwifi.

Luis R. Rodriguez provides a regulatory fix in the shared ath code.

Sujith Manoharan brings a couple ath9k initialization fixes.

Ujjal Roy offers one more mwifiex fix to avoid invalid memory accesses
when unloading the USB driver.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

Pablo Neira Ayuso says:

====================
netfilter fixes for net

The following patchset contains fixes for your net tree, they are:

* Remove extra quote from connlimit configuration in Kconfig, from
  Randy Dunlap.

* Fix missing mss option in syn packets sent to the backend in our
  new synproxy target, from Martin Topholm.

* Use window scale announced by client when sending the forged
  syn to the backend, from Martin Topholm.

* Fix IPv6 address comparison in ebtables, from Luís Fernando
  Cornachioni Estrozi.

* Fix wrong endianess in sequence adjustment which breaks helpers
  in NAT configurations, from Phil Oester.

* Fix the error path handling of nft_compat, from me.

* Make sure the global conntrack counter is decremented after the
  object has been released, also from me.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless into for-davem

Add auto-MDI/MDI-X capability for forced (autonegotiation disabled)
10/100 Mbps speeds on Vitesse VSC82x4 PHYs. Exported previously static
function genphy_setup_forced() required by the new config_aneg handler
in the Vitesse PHY module.

Signed-off-by: Madalin Bucur <madalin.bucur@freescale.com>
Signed-off-by: Shruti Kanetkar <Shruti@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Vitesse VSC8662 is Dual Port 10/100/1000Base-T Phy
Its register set and features are similar to other Vitesse Phys.

Signed-off-by: Sandeep Singh <Sandeep@freescale.com>
Signed-off-by: Andy Fleming <afleming@gmail.com>
Signed-off-by: Shruti Kanetkar <Shruti@Freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The VSC8574 is a quad-port Gigabit Ethernet transceiver with four SerDes
interfaces for quad-port dual media capability.

Signed-off-by: Shaohui Xie <Shaohui.Xie@freescale.com>
Signed-off-by: Andy Fleming <afleming@gmail.com>
Signed-off-by: Shruti Kanetkar <Shruti@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Vitesse VSC8234 is quad port 10/100/1000BASE-T PHY
with SGMII and SERDES MAC interfaces.

Signed-off-by: Andy Fleming <afleming@gmail.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Signed-off-by: Shruti Kanetkar <Shruti@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

In that case it is probable that kernel code overwrote part of the
stack. So we should bail out loudly here.

The BUG_ON may be removed in future if we are sure all protocols are
conformant.

Suggested-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch now always passes msg->msg_namelen as 0. recvmsg handlers must
set msg_namelen to the proper size <= sizeof(struct sockaddr_storage)
to return msg_name to the user.

This prevents numerous uninitialized memory leaks we had in the
recvmsg handlers and makes it harder for new code to accidentally leak
uninitialized memory.

Optimize for the case recvfrom is called with NULL as address. We don't
need to copy the address at all, so set it to NULL before invoking the
recvmsg handler. We can do so, because all the recvmsg handlers must
cope with the case a plain read() is called on them. read() also sets
msg_name to NULL.

Also document these changes in include/linux/net.h as suggested by David
Miller.

Changes since RFC:

Set msg->msg_name = NULL if user specified a NULL in msg_name but had a
non-null msg_namelen in verify_iovec/verify_compat_iovec. This doesn't
affect sendto as it would bail out earlier while trying to copy-in the
address. It also more naturally reflects the logic by the callers of
verify_iovec.

With this change in place I could remove "
if (!uaddr || msg_sys->msg_namelen == 0)
	msg->msg_name = NULL
".

This change does not alter the user visible error logic as we ignore
msg_namelen as long as msg_name is NULL.

Also remove two unnecessary curly brackets in ___sys_recvmsg and change
comments to netdev style.

Cc: David Miller <davem@davemloft.net>
Suggested-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

 bridge dev

When the following commands are executed:

brctl addbr br0
ifconfig br0 hw ether <addr>
rmmod bridge

The calltrace will occur:

[  563.312114] device eth1 left promiscuous mode
[  563.312188] br0: port 1(eth1) entered disabled state
[  563.468190] kmem_cache_destroy bridge_fdb_cache: Slab cache still has objects
[  563.468197] CPU: 6 PID: 6982 Comm: rmmod Tainted: G           O 3.12.0-0.7-default+ #9
[  563.468199] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2007
[  563.468200]  0000000000000880 ffff88010f111e98 ffffffff814d1c92 ffff88010f111eb8
[  563.468204]  ffffffff81148efd ffff88010f111eb8 0000000000000000 ffff88010f111ec8
[  563.468206]  ffffffffa062a270 ffff88010f111ed8 ffffffffa063ac76 ffff88010f111f78
[  563.468209] Call Trace:
[  563.468218]  [<ffffffff814d1c92>] dump_stack+0x6a/0x78
[  563.468234]  [<ffffffff81148efd>] kmem_cache_destroy+0xfd/0x100
[  563.468242]  [<ffffffffa062a270>] br_fdb_fini+0x10/0x20 [bridge]
[  563.468247]  [<ffffffffa063ac76>] br_deinit+0x4e/0x50 [bridge]
[  563.468254]  [<ffffffff810c7dc9>] SyS_delete_module+0x199/0x2b0
[  563.468259]  [<ffffffff814e0922>] system_call_fastpath+0x16/0x1b
[  570.377958] Bridge firewalling registered

--------------------------- cut here -------------------------------

The reason is that when the bridge dev's address is changed, the
br_fdb_change_mac_address() will add new address in fdb, but when
the bridge was removed, the address entry in the fdb did not free,
the bridge_fdb_cache still has objects when destroy the cache, Fix
this by flushing the bridge address entry when removing the bridge.

v2: according to the Toshiaki Makita and Vlad's suggestion, I only
    delete the vlan0 entry, it still have a leak here if the vlan id
    is other number, so I need to call fdb_delete_by_port(br, NULL, 1)
    to flush all entries whose dst is NULL for the bridge.

Suggested-by: Toshiaki Makita <toshiaki.makita1@gmail.com>
Suggested-by: Vlad Yasevich <vyasevich@gmail.com>
Signed-off-by: Ding Tianhong <dingtianhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The following commit:
    b6c40d68
    net: only invoke dev->change_rx_flags when device is UP

tried to fix a problem with VLAN devices and promiscuouse flag setting.
The issue was that VLAN device was setting a flag on an interface that
was down, thus resulting in bad promiscuity count.
This commit blocked flag propagation to any device that is currently
down.

A later commit:
    deede2fa
    vlan: Don't propagate flag changes on down interfaces

fixed VLAN code to only propagate flags when the VLAN interface is up,
thus fixing the same issue as above, only localized to VLAN.

The problem we have now is that if we have create a complex stack
involving multiple software devices like bridges, bonds, and vlans,
then it is possible that the flags would not propagate properly to
the physical devices.  A simple examle of the scenario is the
following:

  eth0----> bond0 ----> bridge0 ---> vlan50

If bond0 or eth0 happen to be down at the time bond0 is added to
the bridge, then eth0 will never have promisc mode set which is
currently required for operation as part of the bridge.  As a
result, packets with vlan50 will be dropped by the interface.

The only 2 devices that implement the special flag handling are
VLAN and DSA and they both have required code to prevent incorrect
flag propagation.  As a result we can remove the generic solution
introduced in b6c40d68

 and leave
it to the individual devices to decide whether they will block
flag propagation or not.

Reported-by: Stefan Priebe <s.priebe@profihost.ag>
Suggested-by: Veaceslav Falico <vfalico@redhat.com>
Signed-off-by: Vlad Yasevich <vyasevic@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

CPUs can ask for local route via ip_route_input_noref() concurrently.
if nh_rth_input is not cached yet, CPUs will proceed to allocate
equivalent DSTs on 'lo' and then will try to cache them in nh_rth_input
via rt_cache_route()
Most of the time they succeed, but on occasion the following two lines:
	orig = *p;
	prev = cmpxchg(p, orig, rt);
in rt_cache_route() do race and one of the cpus fails to complete cmpxchg.
But ip_route_input_slow() doesn't check the return code of rt_cache_route(),
so dst is leaking. dst_destroy() is never called and 'lo' device
refcnt doesn't go to zero, which can be seen in the logs as:
	unregister_netdevice: waiting for lo to become free. Usage count = 1
Adding mdelay() between above two lines makes it easily reproducible.
Fix it similar to nh_pcpu_rth_output case.

Fixes: d2d68ba9

 ("ipv4: Cache input routes in fib_info nexthops.")
Signed-off-by: Alexei Starovoitov <ast@plumgrid.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Hayes Wang says:

====================
r8152 bug fixes

For the patch #3, I add netif_tx_lock() before checking the
netif_queue_stopped(). Besides, I add checking the skb queue
length before waking the tx queue.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

The data from the hardware should be little endian. Correct the
declaration.

Signed-off-by: Hayes Wang <hayeswang@realtek.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The maximum packet number which a tx aggregation buffer could contain
is the tx_qlen.

	tx_qlen = buffer size / (packet size + descriptor size).

If the tx buffer is empty and the queued packets are more than the
maximum value which is defined above, stop the tx queue. Wake the
tx queue if tx queue is stopped and the queued packets are less than
tx_qlen.

Signed-off-by: Hayes Wang <hayeswang@realtek.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Remove the code for sending the packet in the rtl8152_start_xmit().
Let rtl8152_start_xmit() to queue the packet only, and schedule a
tasklet to send the queued packets. This simplify the code and make
sure all the packet would be sent by the original order.

Signed-off-by: Hayes Wang <hayeswang@realtek.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The tx/rx would access the memory which is out of the desired range.
Modify the method of checking the end of the memory to avoid it.

For r8152_tx_agg_fill(), the variable remain may become negative.
However, the declaration is unsigned, so the while loop wouldn't
break when reaching the end of the desied memory. Although to change
the declaration from unsigned to signed is enough to fix it, I also
modify the checking method for safe. Replace

		remain = rx_buf_sz - sizeof(*tx_desc) -
			 (u32)((void *)tx_data - agg->head);

with

		remain = rx_buf_sz - (int)(tx_agg_align(tx_data) - agg->head);

to make sure the variable remain is always positive. Then, the
overflow wouldn't happen.

For rx_bottom(), the rx_desc should not be used to calculate the
packet length before making sure the rx_desc is in the desired range.
Change the checking to two parts. First, check the descriptor is in
the memory. The other, using the descriptor to find out the packet
length and check if the packet is in the memory.

Signed-off-by: Hayes Wang <hayeswang@realtek.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

It appears that driver runs into a problem here if fibsize is too small
because we allocate user_srbcmd with fibsize size only but later we
access it until user_srbcmd->sg.count to copy it over to srbcmd.

It is not correct to test (fibsize < sizeof(*user_srbcmd)) because this
structure already includes one sg element and this is not needed for
commands without data.  So, we would recommend to add the following
(instead of test for fibsize == 0).

Signed-off-by: Mahesh Rajashekhara <Mahesh.Rajashekhara@pmcs.com>
Reported-by: Nico Golde <nico@ngolde.de>
Reported-by: Fabian Yamaguchi <fabs@goesec.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Pull networking fixes from David Miller:
 "Mostly these are fixes for fallout due to merge window changes, as
  well as cures for problems that have been with us for a much longer
  period of time"

 1) Johannes Berg noticed two major deficiencies in our genetlink
    registration.  Some genetlink protocols we passing in constant
    counts for their ops array rather than something like
    ARRAY_SIZE(ops) or similar.  Also, some genetlink protocols were
    using fixed IDs for their multicast groups.

    We have to retain these fixed IDs to keep existing userland tools
    working, but reserve them so that other multicast groups used by
    other protocols can not possibly conflict.

    In dealing with these two problems, we actually now use less state
    management for genetlink operations and multicast groups.

 2) When configuring interface hardware timestamping, fix several
    drivers that simply do not validate that the hwtstamp_confi...

Pull sparc fixes from David Miller:
 "Two merge window fallout build fixes"

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc:
  sparc64: merge fix
  sparc64: fix build regession

Pull ia64 fix from Tony Luck:
 "Unbreak ia64 build by avoiding circular dependency"

* tag 'please-pull-fixia64' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux:
  kernel/bounds: avoid circular dependencies in generated headers

<linux/spinlock.h> has heavy dependencies on other header files.
It triggers circular dependencies in generated headers on IA64, at
least:

  CC      kernel/bounds.s
In file included from /home/space/kas/git/public/linux/arch/ia64/include/asm/thread_info.h:9:0,
                 from include/linux/thread_info.h:54,
                 from include/asm-generic/preempt.h:4,
                 from arch/ia64/include/generated/asm/preempt.h:1,
                 from include/linux/preempt.h:18,
                 from include/linux/spinlock.h:50,
                 from kernel/bounds.c:14:
/home/space/kas/git/public/linux/arch/ia64/include/asm/asm-offsets.h:1:35: fatal error: generated/asm-offsets.h: No such file or directory
compilation terminated.

Let's replace <linux/spinlock.h> with <linux/spinlock_types.h>, it's
enough to find out size of spinlock_t.

Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Reported-and-Tested-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>

Johannes Berg says:

====================
genetlink: clean up multicast group APIs

The generic netlink multicast group registration doesn't have to
be dynamic, and can thus be simplified just like I did with the
ops. This removes some complexity in registration code.

Additionally, two users of generic netlink already use multicast
groups in a wrong way, add workarounds for those two to keep the
userspace API working, but at the same time make them not clash
with other users of multicast groups as might happen now.

While making it all a bit easier, also prevent such abuse by adding
checks to the APIs so each family can only use the groups it owns.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

Register generic netlink multicast groups as an array with
the family and give them contiguous group IDs. Then instead
of passing the global group ID to the various functions that
send messages, pass the ID relative to the family - for most
families that's just 0 because the only have one group.

This avoids the list_head and ID in each group, adding a new
field for the mcast group ID offset to the family.

At the same time, this allows us to prevent abusing groups
again like the quota and dropmon code did, since we can now
check that a family only uses a group it owns.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This doesn't really change anything, but prepares for the
next patch that will change the APIs to pass the group ID
within the family, rather than the global group ID.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add a static inline to generic netlink to wrap netlink_set_err()
to make it easier to use here - use it in openvswitch (the only
generic netlink user of netlink_set_err()).

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

There's no reason to have the family pointer there since it
can just be passed internally where needed, so remove it.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

There are no users of this API remaining, and we'll soon
change group registration to be static (like ops are now)

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

There's no need to unregister the multicast group if the
generic netlink family is registered immediately after.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The quota code is abusing the genetlink API and is using
its family ID as the multicast group ID, which is invalid
and may belong to somebody else (and likely will.)

Make the quota code use the correct API, but since this
is already used as-is by userspace, reserve a family ID
for this code and also reserve that group ID to not break
userspace assumptions.

Acked-by: Jan Kara <jack@suse.cz>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The drop monitor code is abusing the genetlink API and is
statically using the generic netlink multicast group 1, even
if that group belongs to somebody else (which it invariably
will, since it's not reserved.)

Make the drop monitor code use the proper APIs to reserve a
group ID, but also reserve the group id 1 in generic netlink
code to preserve the userspace API. Since drop monitor can
be a module, don't clear the bit for it on unregistration.

Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

As suggested by David Miller, make genl_register_family_with_ops()
a macro and pass only the array, evaluating ARRAY_SIZE() in the
macro, this is a little safer.

The openvswitch has some indirection, assing ops/n_ops directly in
that code. This might ultimately just assign the pointers in the
family initializations, saving the struct genl_family_and_ops and
code (once mcast groups are handled differently.)

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

snd_nxt must be updated synchronously with sk_send_head.  Otherwise
tp->packets_out may be updated incorrectly, what may bring a kernel panic.

Here is a kernel panic from my host.
[  103.043194] BUG: unable to handle kernel NULL pointer dereference at 0000000000000048
[  103.044025] IP: [<ffffffff815aaaaf>] tcp_rearm_rto+0xcf/0x150
...
[  146.301158] Call Trace:
[  146.301158]  [<ffffffff815ab7f0>] tcp_ack+0xcc0/0x12c0

Before this panic a tcp socket was restored. This socket had sent and
unsent data in the write queue. Sent data was restored in repair mode,
then the socket was switched from reapair mode and unsent data was
restored. After that the socket was switched back into repair mode.

In that moment we had a socket where write queue looks like this:
snd_una    snd_nxt   write_seq
   |_________|________|
             |
	  sk_send_head

After a second switching from repair mode the state of socket was
changed:

snd_una          snd_nxt, write_seq
   |_________ ________|
             |
	  sk_send_head

This state is inconsistent, because snd_nxt and sk_send_head are not
synchronized.

Bellow you can find a call trace, how packets_out can be incremented
twice for one skb, if snd_nxt and sk_send_head are not synchronized.
In this case packets_out will be always positive, even when
sk_write_queue is empty.

tcp_write_wakeup
	skb = tcp_send_head(sk);
	tcp_fragment
		if (!before(tp->snd_nxt, TCP_SKB_CB(buff)->end_seq))
			tcp_adjust_pcount(sk, skb, diff);
	tcp_event_new_data_sent
		tp->packets_out += tcp_skb_pcount(skb);

I think update of snd_nxt isn't required, when a socket is switched from
repair mode.  Because it's initialized in tcp_connect_init. Then when a
write queue is restored, snd_nxt is incremented in tcp_event_new_data_sent,
so it's always is in consistent state.

I have checked, that the bug is not reproduced with this patch and
all tests about restoring tcp connections work fine.

Cc: Pavel Emelyanov <xemul@parallels.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Alexey Kuznetsov <kuznet@ms2.inr.ac.ru>
Cc: James Morris <jmorris@namei.org>
Cc: Hideaki YOSHIFUJI <yoshfuji@linux-ipv6.org>
Cc: Patrick McHardy <kaber@trash.net>
Signed-off-by: Andrey Vagin <avagin@openvz.org>
Acked-by: Pavel Emelyanov <xemul@parallels.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

init_card() calls dev_get_by_name() to get a network deceive. But it
doesn't decrease network device reference count after the device is
used.

Signed-off-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>