if we succeed grabbing the refcount, then
  if (err && !xfrm_pol_hold_rcu)

will evaluate to false so this hits last else branch which then
sets policy to ERR_PTR(0).

Fixes: ae33786f ("xfrm: policy: only use rcu in xfrm_sk_policy_lookup")
Reported-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Tested-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>

This is the same fix than commit a5d0dc81 ("vti: flush x-netns xfrm
cache when vti interface is removed")

This patch fixes a refcnt problem when a x-netns vti6 interface is removed:
unregister_netdevice: waiting for vti6_test to become free. Usage count = 1

Here is a script to reproduce the problem:

ip link set dev ntfp2 up
ip addr add dev ntfp2 2001::1/64
ip link add vti6_test type vti6 local 2001::1 remote 2001::2 key 1
ip netns add secure
ip link set vti6_test netns secure
ip netns exec secure ip link set vti6_test up
ip netns exec secure ip link s lo up
ip netns exec secure ip addr add dev vti6_test 2003::1/64
ip -6 xfrm policy add dir out tmpl src 2001::1 dst 2001::2 proto esp \
	   mode tunnel mark 1
ip -6 xfrm policy add dir in tmpl src 2001::2 dst 2001::1 proto esp \
	   mode tunnel mark 1
ip xfrm state add src 2001::1 dst 2001::2 proto esp spi 1 mode tunnel \
	   enc des3_ede 0x112233445566778811223344556677881122334455667788 mark 1
ip xfrm state add src 2001::2 dst 2001::1 proto esp spi 1 mode tunnel \
	   enc des3_ede 0x112233445566778811223344556677881122334455667788 mark 1
ip netns exec secure  ping6 -c 4 2003::2
ip netns del secure

CC: Lance Richardson <lrichard@redhat.com>
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Acked-by: Lance Richardson <lrichard@redhat.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>

Sriharsha Basavapatna says:

====================
be2net: patch-set

The following patch set contains a few bug fixes.
Please consider applying this to the net-next tree.
Thanks.

Patch-1 Obtains proper PF number for BEx chips
Patch-2 Fixes a FW update issue seen with BEx chips
Patch-3 Updates copyright string
Patch-4 Fixes TX stats for TSO packets
Patch-5 Enables VF link state setting for BE3
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

The VF link state setting feature now works on BE3 chips too from
FW ver 11.1.192.0 onwards.
Signed-off-by: Suresh Reddy <suresh.reddy@broadcom.com>
Signed-off-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

TX stats update does not take into account headers which get duplicated
when the TSO packet is split into segments by HW. Fix this for both
tunneled (vxlan) and non-tunneled TSO packets.
Signed-off-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch updates the year and company name in the copyright string
in be_hw.h.
Signed-off-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The driver has a check to ensure that NCSI FW section is updated only
if the current FW version in the card supports it. This FW version check
is done using memcmp() which obviously fails in some cases. Fix this by
breaking up the version string into integer version components and
comparing them.
Signed-off-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The driver gets the pf_num for Skyhawk and Lancer using
GET_FUNC_CONFIG FW command. But since that command is not
supported in BEx, we need to get it from some other command.
Otherwise TPE recovery would fail since all NIC PFs would
end up with a func num of 0. There's a pci function number
field in the response  of GET_CNTL_ATTRIBUTES command that
can be read to get the same info for BEx adapters.
Signed-off-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Johan Hedberg says:

====================
pull request: bluetooth 2016-10-08

Here are a couple of Bluetooth fixes for the 4.9 kernel:

 - Firmware download fix for Atheros controllers
 - Fixes to the content of LE scan response
 - New USB ID for a Marvell chipset

Please let me know if there are any issues pulling. Thanks.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Size used with 'dma_alloc_coherent()' and 'dma_free_coherent()' should be
consistent.
Here, the size of a pointer is used in dma_alloc... and the size of the
pointed structure is used in dma_free...

This has been spotted with coccinelle, using the following script:
////////////////////
@r@
expression x0, x1, y0, y1, z0, z1, t0, t1, ret;
@@

*   ret = dma_alloc_coherent(x0, y0, z0, t0);
    ...
*   dma_free_coherent(x1, y1, ret, t1);

@script:python@
y0 << r.y0;
y1 << r.y1;

@@
if y1.find(y0) == -1:
 print "WARNING: sizes look different:  '%s'   vs   '%s'" % (y0, y1)
////////////////////
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: David S. Miller <davem@davemloft.net>

To ensure the dev->phydev pointer is not used after becoming invalid in
mdiobus_unregister, set it to NULL. This happens when removing the macb
driver without first taking its interface down, since unregister_netdev
will end up calling macb_close.
Signed-off-by: Xander Huff <xander.huff@ni.com>
Signed-off-by: Nathan Sullivan <nathan.sullivan@ni.com>
Signed-off-by: Brad Mouring <brad.mouring@ni.com>
Reviewed-by: Moritz Fischer <moritz.fischer@ettus.com>
Acked-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

In the case when a frontend only negotiates a single queue with xen-
netback it is possible for a skbuff with a s/w hash to result in a
hash extra_info segment being sent to the frontend even when no hash
algorithm has been configured. (The ndo_select_queue() entry point makes
sure the hash is not set if no algorithm is configured, but this entry
point is not called when there is only a single queue). This can result
in a frontend that is unable to handle extra_info segments being given
such a segment, causing it to crash.

This patch fixes the problem by clearing the hash in ndo_start_xmit()
instead, which is clearly guaranteed to be called irrespective of the
number of queues.
Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
Cc: Wei Liu <wei.liu2@citrix.com>
Acked-by: Wei Liu <wei.liu2@citrix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Roundrobin runner of team driver uses 'unsigned int' variable to count
the number of sent_packets. Later it is passed to a subroutine
team_num_to_port_index(struct team *team, int num) as 'num' and when
we reach MAXINT (2**31-1), 'num' becomes negative.

This leads to using incorrect hash-bucket for port lookup
and as a result, packets are dropped. The fix consists of changing
'int num' to 'unsigned int num'. Testing of a fixed kernel shows that
there is no packet drop anymore.
Signed-off-by: Alex Sidorenko <alexandre.sidorenko@hpe.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
Cc: Wei Liu <wei.liu2@citrix.com>
Acked-by: Wei Liu <wei.liu2@citrix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This disallows setting /proc/sys/net/ipv6/conf/*/router_solicitations
to values below -1.

-1 continues to mean an unlimited number of retransmits.

Note: this depends on 'ipv6 addrconf: remove addrconf_sysctl_hop_limit()'
Signed-off-by: Maciej Żenczykowski <maze@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Move usb_autopm_get_interface() ahead of setup_on_usb() to prevent
device from sending usb control message in usb suspend mode.

The error message is as below:

[   83.944103] btusb 1-2:1.1: usb_suspend_interface: status 0
[   83.944107] btusb 1-2:1.0: usb_suspend_interface: status 0
[   83.960132] usb 1-2: usb auto-suspend, wakeup 0
[   83.976156] usb 1-2: usb_suspend_device: status 0
[   83.976162] usb 1-2: usb_suspend_both: status 0
[  298.689106] Bluetooth: hci0
[  298.689399] Bluetooth: hci0: Failed to access otp area (-113)
Signed-off-by: Ethan Hsieh <ethan.hsieh@canonical.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>

An extra entry for MDIO_XGENE got added during merging.
Delete it.
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Laura Abbott <labbott@redhat.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

If NO_DMA=y:

    drivers/built-in.o: In function `emac_probe':
    emac.c:(.text+0x3780b8): undefined reference to `bad_dma_ops'
    emac.c:(.text+0x3780e2): undefined reference to `bad_dma_ops'
    emac.c:(.text+0x378112): undefined reference to `bad_dma_ops'
    emac.c:(.text+0x378146): undefined reference to `bad_dma_ops'
    emac.c:(.text+0x37816e): undefined reference to `bad_dma_ops'
    drivers/built-in.o:emac.c:(.text+0x37819a): more undefined references to `bad_dma_ops' follow

If NO_IOMEM=y:

    drivers/net/ethernet/qualcomm/emac/emac.c: In function ‘emac_remove’:
    drivers/net/ethernet/qualcomm/emac/emac.c:736:3: error: implicit declaration of function ‘iounmap’ [-Werror=implicit-function-declaration]
       iounmap(adpt->phy.digital);
       ^

Add dependencies on HAS_DMA and HAS_IOMEM to fix this.
Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Acked-by: Timur Tabi <timur@codeaurora.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Nelson Chang says:

====================
net: ethernet: mediatek: check the hw lro capability by the chip id instead of the dtsi

The series modify to check if hw lro is supported by the chip id.

changes since v3:
- Refine mtk_is_hwlro_supported() function

changes since v2:
- Refine mtk_get_chip_id() function

changes since v1:
- Because hw lro started to be supported from MT7623, the proper way to check if the feature is capable is to judge by the chip id instead of by the dtsi.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Since the proper way to check the hw lro capability is by the chip id,
hwlro property in the device tree should be removed.
Signed-off-by: Nelson Chang <nelson.chang@mediatek.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Because hw lro started to be supported from MT7623, the proper way to check if
the feature is capable is to judge by the chip id instead of by the dtsi.
Signed-off-by: Nelson Chang <nelson.chang@mediatek.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The driver gets the chip id by ETHSYS_CHIPID0_3/ETHSYS_CHIPID4_7 registers
in mtk_probe().
Signed-off-by: Nelson Chang <nelson.chang@mediatek.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge tag 'rxrpc-rewrite-20161004' of git://git.kernel.org/pub/scm/linux/kernel/git/dhowells/linux-fs

David Howells says:

====================
rxrpc: Fixes

This set of patches contains a bunch of fixes:

 (1) Fix an oops on incoming call to a local endpoint without a bound
     service.

 (2) Only ping for a lost reply in a client call (this is inapplicable to
     service calls).

 (3) Fix maybe uninitialised variable warnings in the ACK/ABORT sending
     function by splitting it.

 (4) Fix loss of PING RESPONSE ACKs due to them being subsumed by PING ACK
     generation.

 (5) OpenAFS improperly terminates calls it makes as a client under some
     circumstances by not fully hard-ACK'ing the last DATA packets.  This
     is alleviated by a new call appearing on the same channel implicitly
     completing the previous call on that channel.  Handle this implicit
     completion.

 (6) Properly handle expiry of service calls due to the aforementioned
     improper termination with no follow up call to implicitly complete it:

     (a) The call's background processor needs to be queued to complete the
     	 call, send an abort and notify the socket.

     (b) The call's background processor needs to notify the socket (or the
     	 kernel service) when it has completed the call.

     (c) A negative error code must thence be returned to the kernel
     	 service so that it knows the call died.

     (d) The AFS filesystem must detect the fatal error and end the call.

 (7) Must produce a DELAY ACK when the actual service operation takes a
     while to process and must cancel the ACK when the reply is ready.

 (8) Don't request an ACK on the last DATA packet of the Tx phase as this
     confuses OpenAFS.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

During the conversion to the feature flags, a check against
ci->id != BCMA_CHIP_ID_BCM47162
became
bgmac->feature_flags & BGMAC_FEAT_CLKCTLS
instead of
!(bgmac->feature_flags & BGMAC_FEAT_CLKCTLS)
Reported-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: Jon Mason <jon.mason@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Since linux-3.15, netlink_dump() can use up to 16384 bytes skb
allocations.

Due to struct skb_shared_info ~320 bytes overhead, we end up using
order-3 (on x86) page allocations, that might trigger direct reclaim and
add stress.

The intent was really to attempt a large allocation but immediately
fallback to a smaller one (order-1 on x86) in case of memory stress.

On recent kernels (linux-4.4), we can remove __GFP_DIRECT_RECLAIM to
meet the goal. Old kernels would need to remove __GFP_WAIT

While we are at it, since we do an order-3 allocation, allow to use
all the allocated bytes instead of 16384 to reduce syscalls during
large dumps.

iproute2 already uses 32KB recvmsg() buffer sizes.

Alexei provided an initial patch downsizing to SKB_WITH_OVERHEAD(16384)

Fixes: 9063e21f ("netlink: autosize skb lengthes")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Alexei Starovoitov <ast@kernel.org>
Cc: Greg Thelen <gthelen@google.com>
Reviewed-by: Greg Rose <grose@lightfleet.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

If a socket has FANOUT sockopt set, a new proto_hook is registered
as part of fanout_add(). When processing a NETDEV_UNREGISTER event in
af_packet, __fanout_unlink is called for all sockets, but prot_hook which was
registered as part of fanout_add is not removed. Call fanout_release, on a
NETDEV_UNREGISTER, which removes prot_hook and removes fanout from the
fanout_list.

This fixes BUG_ON(!list_empty(&dev->ptype_specific)) in netdev_run_todo()
Signed-off-by: Anoob Soman <anoob.soman@citrix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The KSZ9031 skew registers contain an offset, the chip's default value
is "neutral" which does not add any skew. Programming a 0 into a skew
property will actually set it the maximal negative adjustment and not
to a neutral position as one would expect.

Explain this situation in the devicetree binding documentation and list
the settings that the chip considers neutral.

Changing the implementation to accept negative values would have been
a better solution, but would break existing configurations.
Signed-off-by: Mike Looijmans <mike.looijmans@topic.nl>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Wake-on-LAN (WoL) is an Ethernet networking standard that allows
a computer/device to be turned on or awakened by a network message.

VSC8531 PHY can support this feature configure by driver set function.
WoL status get by driver get function.

Tested on Beaglebone Black with VSC 8531 PHY.
Signed-off-by: Raju Lakkaraju <Raju.Lakkaraju@microsemi.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add a new compatible string for the R8A7796 (M3-W) RAVB.
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add support to enable CPSW RGMII internal delay (id mode) bits
when rgmii internal delay is configured in phy.
Signed-off-by: Mugunthan V N <mugunthanvnm@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Trival fix, dev_err messages are missing a \n, so add it. Also
fix grammer, spelling mistake and add white spaces to various
error messages.
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Trival fix, dev_dbg message is missing a \n, so add it.
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Trival fix, dev_err messages are missing a \n, so add it.
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Paul Durrant says:

====================
xen-netback: guest rx side refactor

This series refactors the guest rx side of xen-netback:

- The code is moved into its own source module.

- The prefix variant of GSO handling is retired (since it is no longer
  in common use, and alternatives exist).

- The code is then simplified and modifications made to improve
  performance.

v2:
- Rebased onto refreshed net-next
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

This allows full 64K skbuffs (with 1500 mtu ethernet, composed of 45
fragments) to be handled by netback for to-guest rx.
Signed-off-by: Ross Lagerwall <ross.lagerwall@citrix.com>
[re-based]
Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
Reviewed-by: David Vrabel <david.vrabel@citrix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Instead of flushing the copy ops when an packet is complete, complete
packets when their copy ops are done.  This improves performance by
reducing the number of grant copy hypercalls.

Latency is still limited by the relatively small size of the copy
batch.
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
[re-based]
Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Instead of only placing one skb on the guest rx ring at a time, process
a batch of up-to 64.  This improves performance by ~10% in some tests.
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
[re-based]
Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

When an skb is removed from the guest rx queue, immediately wake the
tx queue, instead of after processing them.
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
[re-based]
Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Refactor the to-guest (rx) path to:

1. Push responses for completed skbs earlier, reducing latency.

2. Reduce the per-queue memory overhead by greatly reducing the
   maximum number of grant copy ops in each hypercall (from 4352 to
   64).  Each struct xenvif_queue is now only 44 kB instead of 220 kB.

3. Make the code more maintainable.
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
[re-based]
Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

As far as I am aware only very old Windows network frontends make use of
this style of passing GSO packets from backend to frontend. These
frontends can easily be replaced by the freely available Xen Project
Windows PV network frontend, which uses the 'default' mechanism for
passing GSO packets, which is also used by all Linux frontends.

NOTE: Removal of this feature will not cause breakage in old Windows
      frontends. They simply will no longer receive GSO packets - the
      packets instead being fragmented in the backend.
Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
Reviewed-by: David Vrabel <david.vrabel@citrix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>