the commit cdf3464e ("ipv6: Fix dst_entry refcnt bugs in ip6_tunnel")
introduced percpu storage for ip6_tunnel dst cache, but while clearing
such cache it used raw_cpu_ptr to walk the per cpu entries, so cached
dst on non current cpu are not actually reset.

This patch replaces raw_cpu_ptr with per_cpu_ptr, properly cleaning
such storage.

Fixes: cdf3464e ("ipv6: Fix dst_entry refcnt bugs in ip6_tunnel")
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch adds support for the sgmii phy interface.
Signed-off-by: Punnaiah Choudary Kalluri <punnaia@xilinx.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch adds the PCI device ID (0xe0a1) and alx_pci_tbl entry for the
Killer E2400 Ethernet controller, modeled after the Killer E2200
controller support (0xe091) already present in the alx driver.

This patch was originally authored by Ben Pope, but it got held up by
issues in the commit message, so I'm resubmitting it on his behalf.

I've extensively used a kernel with this patch on a System76 serw9
laptop and am quite confident it works well (at least on the hardware I
have available for testing).

Note that as a favor to System76, Ubuntu has been carrying this as a
sauce patch in their 4.2 based Wily kernel, which presumably has given
it real-world testing on other E2400 equipped hardware (I don't know of
any Ubuntu kernel bugs filed about it):

https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1498633Signed-off-by: Jason Gerard DeRose <jason@system76.com>
Signed-off-by: Ben Pope <benpope81@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

If alloc_netdev() failed and return NULL, then the next instruction
would dereference it. Found by Coverity.

Compile tested only. Not sure if anyone still uses this driver
(or the whole WAN subsystem).
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

This was detected by Coverity.
The function skb_cow_head leaves skb alone on failure, so caller needs
to free.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Pull s390 fixes from Martin Schwidefsky:
 "Assorted bug fixes, the mlock2 system call gets added, and one
  improvement.  The boot from dasd devices is now possible from a wider
  range of devices"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux:
  s390: remove SALIPL loader
  s390: wire up mlock2 system call
  s390: remove g5 elf platform support
  s390: avoid cache aliasing under z/VM and KVM
  s390/sclp: _sclp_wait_int(): retain full PSW mask
  s390/zcrypt: Fix initialisation when zcrypt is built-in
  s390/zcrypt: Fix kernel crash on systems without AP bus support
  s390: add support for ipl devices in subchannel sets > 0
  s390/ipl: fix out of bounds access in scpdata_write
  s390/pci_dma: improve debugging of errors during dma map
  s390/pci_dma: handle dma table failures
  s390/pci_dma: unify label of invalid translation table entries
  s390/syscalls: remove system call number calculation
  s390/cio: simplify css_generate_pgid
  s390/diag: add a s390 prefix to the diagnose trace point
  s390/head: fix error message on unsupported hardware

Merge tag 'hwmon-for-linus-v4.4-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging

Pull hwmon fixes from Guenter Roeck:
 "Fix build issues in scpi and ina2xx drivers, update scpi driver to
  support recent firmware, and fix an uninitialized variable warning in
  applesmc driver"

* tag 'hwmon-for-linus-v4.4-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging:
  hwmon: (scpi) skip unsupported sensors properly
  hwmon: (scpi) add thermal-of dependency
  hwmon : (applesmc) Fix uninitialized variables warnings
  hwmon: (ina2xx) Fix build issue by selecting REGMAP_I2C

Pull networking fixes from David Miller:

 1) Fix list tests in netfilter ingress support, from Florian Westphal.

 2) Fix reversal of input and output interfaces in ingress hook
    invocation, from Pablo Neira Ayuso.

 3) We have a use after free in r8169, caught by Dave Jones, fixed by
    Francois Romieu.

 4) Splice use-after-free fix in AF_UNIX frmo Hannes Frederic Sowa.

 5) Three ipv6 route handling bug fixes from Martin KaFai Lau:
    a) Don't create clone routes not managed by the fib6 tree
    b) Don't forget to check expiration of DST_NOCACHE routes.
    c) Handle rt->dst.from == NULL properly.

 6) Several AF_PACKET fixes wrt transport header setting and SKB
    protocol setting, from Daniel Borkmann.

 7) Fix thunder driver crash on shutdown, from Pavel Fedin.

 8) Several Mellanox driver fixes (max MTU calculations, use of correct
    DMA unmap in TX path, etc.) from Saeed Mahameed, Tariq Toukan, Doron
    Tsur, Achiad Shochat, Eran Ben Elisha, and Noa Osherovich.

 9) Several mv88e6060 DSA driver fixes (wrong bit definitions for
    certain registers, etc.) from Neil Armstrong.

10) Make sure to disable preemption while updating per-cpu stats of ip
    tunnels, from Jason A.  Donenfeld.

11) Various ARM64 bpf JIT fixes, from Yang Shi.

12) Flush icache properly in ARM JITs, from Daniel Borkmann.

13) Fix masking of RX and TX interrupts in ravb driver, from Masaru
    Nagai.

14) Fix netdev feature propagation for devices not implementing
    ->ndo_set_features().  From Nikolay Aleksandrov.

15) Big endian fix in vmxnet3 driver, from Shrikrishna Khare.

16) RAW socket code increments incorrect SNMP counters, fix from Ben
    Cartwright-Cox.

17) IPv6 multicast SNMP counters are bumped twice, fix from Neil Horman.

18) Fix handling of VLAN headers on stacked devices when REORDER is
    disabled.  From Vlad Yasevich.

19) Fix SKB leaks and use-after-free in ipvlan and macvlan drivers, from
    Sabrina Dubroca.

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (83 commits)
  MAINTAINERS: Update Mellanox's Eth NIC driver entries
  net/core: revert "net: fix __netdev_update_features return.." and add comment
  af_unix: take receive queue lock while appending new skb
  rtnetlink: fix frame size warning in rtnl_fill_ifinfo
  net: use skb_clone to avoid alloc_pages failure.
  packet: Use PAGE_ALIGNED macro
  packet: Don't check frames_per_block against negative values
  net: phy: Use interrupts when available in NOLINK state
  phy: marvell: Add support for 88E1540 PHY
  arm64: bpf: make BPF prologue and epilogue align with ARM64 AAPCS
  macvlan: fix leak in macvlan_handle_frame
  ipvlan: fix use after free of skb
  ipvlan: fix leak in ipvlan_rcv_frame
  vlan: Do not put vlan headers back on bridge and macvlan ports
  vlan: Fix untag operations of stacked vlans with REORDER_HEADER off
  via-velocity: unconditionally drop frames with bad l2 length
  ipg: Remove ipg driver
  dl2k: Add support for IP1000A-based cards
  snmp: Remove duplicate OUTMCAST stat increment
  net: thunder: Check for driver data in nicvf_remove()
  ...

Eugenia (Jenny) Emantayev is replacing Amir Vadai as the
mlx4 Ethernet driver maintainer.

Saeed Mahameed is assigned to maintain mlx5 Eth functionality.
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This reverts commit 00ee5927 ("net: fix __netdev_update_features return
on ndo_set_features failure")
and adds a comment explaining why it's okay to return a value other than
0 upon error. Some drivers might actually change flags and return an
error so it's better to fire a spurious notification rather than miss
these.

CC: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

While possibly in future we don't necessarily need to use
sk_buff_head.lock this is a rather larger change, as it affects the
af_unix fd garbage collector, diag and socket cleanups. This is too much
for a stable patch.

For the time being grab sk_buff_head.lock without disabling bh and irqs,
so don't use locked skb_queue_tail.

Fixes: 869e7c62 ("net: af_unix: implement stream sendpage support")
Cc: Eric Dumazet <edumazet@google.com>
Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Reported-by: Eric Dumazet <edumazet@google.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fix the following warning:

  CC      net/core/rtnetlink.o
net/core/rtnetlink.c: In function ‘rtnl_fill_ifinfo’:
net/core/rtnetlink.c:1308:1: warning: the frame size of 2864 bytes is larger than 2048 bytes [-Wframe-larger-than=]
 }
 ^
by splitting up the huge rtnl_fill_ifinfo into some smaller ones, so we
don't have the huge frame allocations at the same time.

Cc: Eric Dumazet <edumazet@google.com>
Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

1. new skb only need dst and ip address(v4 or v6).
2. skb_copy may need high order pages, which is very rare on long running server.
Signed-off-by: Junwei Zhang <linggao.zjw@alibaba-inc.com>
Signed-off-by: Martin Zhang <martinbj2008@gmail.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Use PAGE_ALIGNED(...) instead of open-coding it.
Signed-off-by: Tobias Klauser <tklauser@distanz.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

rb->frames_per_block is an unsigned int, thus can never be negative.

Also fix spacing in the calculation of frames_per_block.
Signed-off-by: Tobias Klauser <tklauser@distanz.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

The NOLINK state will poll the phy once a second to see if the link
has come up. If the phy has an interrupt line, this polling can be
skipped, since the phy should interrupt when the link returns.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

The 88E1540 can be found embedded in the Marvell 88E6352 switch.  It
is compatible with the 88E1510, so add support for it, using the
88E1510 specific functions.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

Save and restore FP/LR in BPF prog prologue and epilogue, save SP to FP
in prologue in order to get the correct stack backtrace.

However, ARM64 JIT used FP (x29) as eBPF fp register, FP is subjected to
change during function call so it may cause the BPF prog stack base address
change too.

Use x25 to replace FP as BPF stack base register (fp). Since x25 is callee
saved register, so it will keep intact during function call.
It is initialized in BPF prog prologue when BPF prog is started to run
everytime. Save and restore x25/x26 in BPF prologue and epilogue to keep
them intact for the outside of BPF. Actually, x26 is unnecessary, but SP
requires 16 bytes alignment.

So, the BPF stack layout looks like:

                                 high
         original A64_SP =>   0:+-----+ BPF prologue
                                |FP/LR|
         current A64_FP =>  -16:+-----+
                                | ... | callee saved registers
                                +-----+
                                |     | x25/x26
         BPF fp register => -80:+-----+
                                |     |
                                | ... | BPF prog stack
                                |     |
                                |     |
         current A64_SP =>      +-----+
                                |     |
                                | ... | Function call stack
                                |     |
                                +-----+
                                  low

CC: Zi Shen Lim <zlim.lnx@gmail.com>
CC: Xi Wang <xi.wang@gmail.com>
Signed-off-by: Yang Shi <yang.shi@linaro.org>
Acked-by: Zi Shen Lim <zlim.lnx@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Reset pskb in macvlan_handle_frame in case skb_share_check returned a
clone.

Fixes: 8a4eb573 ("net: introduce rx_handler results and logic around that")
Signed-off-by: Sabrina Dubroca <sd@queasysnail.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

ipvlan_handle_frame is a rx_handler, and when it returns a value other
than RX_HANDLER_CONSUMED (here, NET_RX_DROP aka RX_HANDLER_ANOTHER),
__netif_receive_skb_core expects that the skb still exists and will
process it further, but we just freed it.

Fixes: 2ad7bf36 ("ipvlan: Initial check-in of the IPVLAN driver.")
Signed-off-by: Sabrina Dubroca <sd@queasysnail.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

Pass a **skb to ipvlan_rcv_frame so that if skb_share_check returns a
new skb, we actually use it during further processing.

It's safe to ignore the new skb in the ipvlan_xmit_* functions, because
they call ipvlan_rcv_frame with local == true, so that dev_forward_skb
is called and always takes ownership of the skb.

Fixes: 2ad7bf36 ("ipvlan: Initial check-in of the IPVLAN driver.")
Signed-off-by: Sabrina Dubroca <sd@queasysnail.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

Vladislav Yasevich says:

====================
Fix issues with vlans without REORDER_HEADER

A while ago Phil Sutter brought up an issue with vlans without
REORDER_HEADER and bridges.  The problem was that if a vlan
without REORDER_HEADER was a port in the bridge, the bridge ended
up forwarding corrupted packets that still contained the vlan header.
The same issue exists for bridge mode macvlan/macvtap devices.

An additional issue with vlans without REORDER_HEADER is that stacking
them also doesn't work.  The reason here is that skb_reorder_vlan_header()
function assumes that it on ETH_HLEN bytes deep into the packet.  That
is not the case, when you a vlan without REORRDER_HEADER flag set.

This series attempts to correct these 2 issues.

1) To solve the stacked vlans problem, the patch simply use
skb->mac_len as an offset to start copying mac addresses that
is part of header reordering.

2) To fix the issue with bridge/macvlan/macvtap, the second patch
simply doesn't write the vlan header back to the packet if the
vlan device is either a bridge or a macvlan port.  This ends up
being the simplest and least performance intrussive solution.

I've considered extending patch 2 to all stacked devices (essentially
checked for the presense of rx_handler), but that feels like a broader
restriction and _may_ break existing uses.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

When a vlan is configured with REORDER_HEADER set to 0, the vlan
header is put back into the packet and makes it appear that
the vlan header is still there even after it's been processed.
This posses a problem for bridge and macvlan ports.  The packets
passed to those device may be forwarded and at the time of the
forward, vlan headers end up being unexpectedly present.

With the patch, we make sure that we do not put the vlan header
back (when REORDER_HEADER is 0) if a bridge or macvlan has
been configured on top of the vlan device.
Signed-off-by: Vladislav Yasevich <vyasevic@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

When we have multiple stacked vlan devices all of which have
turned off REORDER_HEADER flag, the untag operation does not
locate the ethernet addresses correctly for nested vlans.
The reason is that in case of REORDER_HEADER flag being off,
the outer vlan headers are put back and the mac_len is adjusted
to account for the presense of the header.  Then, the subsequent
untag operation, for the next level vlan, always use VLAN_ETH_HLEN
to locate the begining of the ethernet header and that ends up
being a multiple of 4 bytes short of the actuall beginning
of the mac header (the multiple depending on the how many vlan
encapsulations ethere are).

As a reslult, if there are multiple levles of vlan devices
with REODER_HEADER being off, the recevied packets end up
being dropped.

To solve this, we use skb->mac_len as the offset.  The value
is always set on receive path and starts out as a ETH_HLEN.
The value is also updated when the vlan header manupations occur
so we know it will be correct.
Signed-off-by: Vladislav Yasevich <vyasevic@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

By default the driver allowed incorrect frames to be received. What is
worse the code does not handle very short frames correctly. The FCS
length is unconditionally subtracted, and the underflow can cause
skb_put to be called with large number after implicit cast to unsigned.
And indeed, an skb_over_panic() was observed with via-velocity.

This removes the module parameter as it does not work in it's
current state, and should be implemented via NETIF_F_RXALL if needed.
Suggested-by: Francois Romieu <romieu@fr.zoreil.com>
Signed-off-by: Timo Teräs <timo.teras@iki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>

Pull vfs fixes from Al Viro:
 "A fs-cache regression fix, and adding a warning about obnoxiou^W
  moderation of list given in MAINTAINERS"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
  MAINTAINERS: linux-cachefs@redhat.com is moderated for non-subscribers
  FS-Cache: Add missing initialization of ret in cachefiles_write_page()

Pull crypto fix from Herbert Xu:
 "This fixes a bug in the qat driver where a user-space pointer is
  dereferenced"

* 'linus' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6:
  crypto: qat - don't use userspace pointer

Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

fs/cachefiles/rdwr.c: In function ‘cachefiles_write_page’:
fs/cachefiles/rdwr.c:882: warning: ‘ret’ may be used uninitialized in
this function

If the jump to label "error" is taken, "ret" will indeed be
uninitialized, and random stack data may be printed by the debug code.

Fixes: 102f4d90 ("FS-Cache: Handle a write to the page immediately beyond the EOF marker")
Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

Now that IP1000A chips are supported by dl2k driver, the buggy ipg
driver can be removed.
Signed-off-by: Ondrej Zary <linux@rainbow-software.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add support for IP1000A chips to dl2k driver.
IP1000A chip looks like a TC9020 with integrated PHY.

This allows IP1000A chips to work reliably because the ipg driver is
buggy - it loses packets under load and then completely stops
transmitting data.

Tested with Asus NX1101 v2.0 at 10, 100 and 1000Mbps:
vendor=0x13f0 device=0x1023 (rev 0x41)
subsystem vendor=0x1043 device=0x8180

MAC address registers access needed to be changed from 8-bit to 16-bit
because 8-bit does not work on IP1000A. 8-bit access is not even
allowed in the TC9020 datasheet (although it worked). 16-bit access
works on both.

Tested that it does not break D-Link DGE-550T (DL-2000 chip, probably
a rebranded TC9020):
vendor=0x1186 device=0x4000 (rev 0x0c)
subsystem vendor=0x1186 device=0x4000
Signed-off-by: Ondrej Zary <linux@rainbow-software.org>
Reviewed-by: Andy Shevchenko <andy.shevchenko@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

the OUTMCAST stat is double incremented, getting bumped once in the mcast code
itself, and again in the common ip output path.  Remove the mcast bump, as its
not needed

Validated by the reporter, with good results
Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
Reported-by: Claus Jensen <claus.jensen@microsemi.com>
CC: Claus Jensen <claus.jensen@microsemi.com>
CC: David Miller <davem@davemloft.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

In some cases the crash is caused by nicvf_remove() being called from
outside. For example, if we try to feed the device to vfio after the
probe has failed for some reason. So, move the check to better place.
Signed-off-by: Pavel Fedin <p.fedin@samsung.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

A recent flaw in the netdev feature setting resulted in warnings
like this one from VLAN interfaces:

 WARNING: CPU: 1 PID: 4975 at net/core/dev.c:2419 skb_warn_bad_offload+0xbc/0xcb()
 : caps=(0x00000000001b5820, 0x00000000001b5829) len=2782 data_len=0 gso_size=1348 gso_type=16 ip_summed=3

The ":" is supposed to be preceded by a driver name, but in this
case it is an empty string since the device has no parent.

There are many types of network devices without a parent. The
anonymous warnings for these devices can be hard to debug.  Log
the network device name instead in these cases to assist further
debugging.

This is mostly similar to how __netdev_printk() handles orphan
devices.
Signed-off-by: Bjørn Mork <bjorn@mork.no>
Signed-off-by: David S. Miller <davem@davemloft.net>

In case multiple writes to a unix stream socket race we could end up in a
situation where we pre-allocate a new skb for use in unix_stream_sendpage
but have to free it again in the locked section because another skb
has been appended meanwhile, which we must use. Accidentally we didn't
clear the pointer after consuming it and so we touched freed memory
while appending it to the sk_receive_queue. So, clear the pointer after
consuming the skb.

This bug has been found with syzkaller
(http://github.com/google/syzkaller) by Dmitry Vyukov.

Fixes: 869e7c62 ("net: af_unix: implement stream sendpage support")
Reported-by: Dmitry Vyukov <dvyukov@google.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

rtnl_fdb_dump always expects an index to be returned by the ndo_fdb_dump op,
but when CONFIG_NET_SWITCHDEV is off, it returns an error.

Fix that by returning the given unmodified idx.

A similar fix was 0890cf6c ("switchdev: fix return value of
switchdev_port_fdb_dump in case of error") but for the CONFIG_NET_SWITCHDEV=y
case.

Fixes: 45d4122c ("switchdev: add support for fdb add/del/dump via switchdev_port_obj ops.")
Signed-off-by: Dragos Tatulea <dragos@endocode.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Commit 05cc5a39 "bnx2x: add vlan filtering offload" introduced
a regression in regard for vlans for 57710, 57711 adapters -
Loading 8021q module on a machine with such an adapter would cause
a null pointer dereference, as the driver mistakenly publishes it
has capabilities for vlan CTAG filtering.
Reported-by: Otto Sabart <osabart@redhat.com>
Signed-off-by: Yuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: Ariel Elior <Ariel.Elior@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This driver does not handle the AVB-DMAC Receive FIFO Warning interrupt
now, so the interrupt should not be enabled.
Signed-off-by: Masaru Nagai <masaru.nagai.vx@renesas.com>
Signed-off-by: Yoshihiro Kaneko <ykaneko0929@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Sending ICMP packets with raw sockets ends up in the SNMP counters
logging the type as the first byte of the IPv4 header rather than
the ICMP header. This is fixed by adding the IP Header Length to
the casting into a icmphdr struct.
Signed-off-by: Ben Cartwright-Cox <ben@benjojo.co.uk>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>