Allow matching and setting the ct_mark field. As with ct_state and
ct_zone, these fields are populated when the CT action is executed. To
write to this field, a value and mask can be specified as a nested
attribute under the CT action. This data is stored with the conntrack
entry, and is executed after the lookup occurs for the CT action. The
conntrack entry itself must be committed using the COMMIT flag in the CT
action flags for this change to persist.
Signed-off-by: NJustin Pettit <jpettit@nicira.com>
Signed-off-by: NJoe Stringer <joestringer@nicira.com>
Acked-by: NThomas Graf <tgraf@suug.ch>
Acked-by: NPravin B Shelar <pshelar@nicira.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Expose the kernel connection tracker via OVS. Userspace components can
make use of the CT action to populate the connection state (ct_state)
field for a flow. This state can be subsequently matched.

Exposed connection states are OVS_CS_F_*:
- NEW (0x01) - Beginning of a new connection.
- ESTABLISHED (0x02) - Part of an existing connection.
- RELATED (0x04) - Related to an established connection.
- INVALID (0x20) - Could not track the connection for this packet.
- REPLY_DIR (0x40) - This packet is in the reply direction for the flow.
- TRACKED (0x80) - This packet has been sent through conntrack.

When the CT action is executed by itself, it will send the packet
through the connection tracker and populate the ct_state field with one
or more of the connection state flags above. The CT action will always
set the TRACKED bit.

When the COMMIT flag is passed to the conntrack action, this specifies
that information about the connection should be stored. This allows
subsequent packets for the same (or related) connections to be
correlated with this connection. Sending subsequent packets for the
connection through conntrack allows the connection tracker to consider
the packets as ESTABLISHED, RELATED, and/or REPLY_DIR.

The CT action may optionally take a zone to track the flow within. This
allows connections with the same 5-tuple to be kept logically separate
from connections in other zones. If the zone is specified, then the
"ct_zone" match field will be subsequently populated with the zone id.

IP fragments are handled by transparently assembling them as part of the
CT action. The maximum received unit (MRU) size is tracked so that
refragmentation can occur during output.

IP frag handling contributed by Andy Zhou.

Based on original design by Justin Pettit.
Signed-off-by: NJoe Stringer <joestringer@nicira.com>
Signed-off-by: NJustin Pettit <jpettit@nicira.com>
Signed-off-by: NAndy Zhou <azhou@nicira.com>
Acked-by: NThomas Graf <tgraf@suug.ch>
Acked-by: NPravin B Shelar <pshelar@nicira.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This variation on skb_dst_copy() doesn't require two skbs.
Signed-off-by: NJoe Stringer <joestringer@nicira.com>
Acked-by: NPravin B Shelar <pshelar@nicira.com>
Acked-by: NThomas Graf <tgraf@suug.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NJoe Stringer <joestringer@nicira.com>
Acked-by: NThomas Graf <tgraf@suug.ch>
Acked-by: NPravin B Shelar <pshelar@nicira.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This will allow the ovs-conntrack code to reuse these macros.
Signed-off-by: NJoe Stringer <joestringer@nicira.com>
Acked-by: NThomas Graf <tgraf@suug.ch>
Acked-by: NPravin B Shelar <pshelar@nicira.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Previously, we used the kernel-internal netlink actions length to
calculate the size of messages to serialize back to userspace.
However,the sw_flow_actions may not be formatted exactly the same as the
actions on the wire, so store the original actions length when
de-serializing and re-use the original length when serializing.
Signed-off-by: NJoe Stringer <joestringer@nicira.com>
Acked-by: NPravin B Shelar <pshelar@nicira.com>
Acked-by: NThomas Graf <tgraf@suug.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Broadcom buses may have more than 1 Ethernet device. This is used e.g.
to have few interfaces connected to different switch ports. So far we
saw chipsets with only 2 devices (e.g. BCM4706) but recent ones have
up to 3 (e.g. Netgear R8000 uses 3rd interface for most of switch
traffic, lower interfaces are for some kind of offloading).
Signed-off-by: NRafał Miłecki <zajec5@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Merge tag 'wireless-drivers-next-for-davem-2015-08-26' of git://git.kernel.org/pub/scm/linux/kernel/git/kvalo/wireless-drivers-next

Kalle Valo says:

====================
Major changes:

iwlwifi:

* new Tx power firmware API
* bump max firmware API to 17
* fix bug in debug prints
* static checker fix
* fix unused defines
* fix command list on newest firmware

brcmfmac:

* support NVRAM loading for bcm47xx platform
* new debugfs entry for msgbuf protocol layer used with PCIe devices

ath10k:

* add spectral scan support for qca99x0
* add qca6164 support
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Make sure to indicate to tunnel driver that key.tun_id is set,
otherwise gre won't recognize the metadata.

Fixes: d3aa45ce ("bpf: add helpers to access tunnel metadata")
Signed-off-by: NAlexei Starovoitov <ast@plumgrid.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Alexei Starovoitov says:

====================
act_bpf: remove spinlock in fast path

v1 version had a race condition in cleanup path of bpf_prog.
I tried to fix it by adding new callback 'cleanup_rcu' to 'struct tcf_common'
and call it out of act_api cleanup path, but Daniel noticed
(thanks for the idea!) that most of the classifiers already do action cleanup
out of rcu callback.
So instead this set of patches converts tcindex and rsvp classifiers to call
tcf_exts_destroy() after rcu grace period and since action cleanup logic
in __tcf_hash_release() is only called when bind and refcnt goes to zero,
it's guaranteed that cleanup() callback is called from rcu callback.
More specifically:
patches 1 and 2 - simple fixes
patches 2 and 3 - convert tcf_exts_destroy in tcindex and rsvp to call_rcu
patch 5 - removes spin_lock from act_bpf

The cleanup of actions is now universally done after rcu grace period
and in the future we can drop (now unnecessary) call_rcu from tcf_hash_destroy()
patch 5 is using synchronize_rcu() in act_bpf replacement path, since it's
very rare and alternative of dynamically allocating 'struct tcf_bpf_cfg' just
to pass it to call_rcu looks even less appealing.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Similar to act_gact/act_mirred, act_bpf can be lockless in packet processing
with extra care taken to free bpf programs after rcu grace period.
Replacement of existing act_bpf (very rare) is done with synchronize_rcu()
and final destruction is done from tc_action_ops->cleanup() callback that is
called from tcf_exts_destroy()->tcf_action_destroy()->__tcf_hash_release() when
bind and refcnt reach zero which is only possible when classifier is destroyed.
Previous two patches fixed the last two classifiers (tcindex and rsvp) to
call tcf_exts_destroy() from rcu callback.

Similar to gact/mirred there is a race between prog->filter and
prog->tcf_action. Meaning that the program being replaced may use
previous default action if it happened to return TC_ACT_UNSPEC.
act_mirred race betwen tcf_action and tcfm_dev is similar.
In all cases the race is harmless.
Long term we may want to improve the situation by replacing the whole
tc_action->priv as single pointer instead of updating inner fields one by one.
Signed-off-by: NAlexei Starovoitov <ast@plumgrid.com>
Acked-by: NDaniel Borkmann <daniel@iogearbox.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Adjust destroy path of cls_rsvp to call tcf_exts_destroy() after
rcu grace period.
Signed-off-by: NAlexei Starovoitov <ast@plumgrid.com>
Acked-by: NDaniel Borkmann <daniel@iogearbox.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Adjust destroy path of cls_tcindex to call tcf_exts_destroy() after
rcu grace period.
Signed-off-by: NAlexei Starovoitov <ast@plumgrid.com>
Acked-by: NDaniel Borkmann <daniel@iogearbox.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Fix harmless typo and avoid unnecessary copy of empty 'prog' into
unused 'strcut tcf_bpf_cfg old'.

Fixes: f4eaed28 ("act_bpf: fix memory leaks when replacing bpf programs")
Signed-off-by: NAlexei Starovoitov <ast@plumgrid.com>
Acked-by: NDaniel Borkmann <daniel@iogearbox.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

tcf_hash_destroy() used once. Make it static.
Signed-off-by: NAlexei Starovoitov <ast@plumgrid.com>
Acked-by: NDaniel Borkmann <daniel@iogearbox.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The Kconfig option AVERAGE and its implementation has been removed by
commit f4e774f5 ("average: remove out-of-line implementation").
Remove the dead build rule in lib/Makefile.
Signed-off-by: NValentin Rothberg <valentinrothberg@gmail.com>
Reviewed-by: NJohannes Berg <johannes@sipsolutions.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Major changes in ath10k:

* add spectral scan support for qca99x0
* add qca6164 support

Below compilation warnings are observed in gcc version 4.8.2.
Even though it's not seen in bit older gcc versions (for ex, 4.7.3),
It's good to fix it by changing format specifier from %d to
%zd in wmi pull phyerr functions.

wmi.c: In function 'ath10k_wmi_op_pull_phyerr_ev':
wmi.c:3567:8: warning: format '%d' expects argument of type 'int',
              but argument 4 has type 'long unsigned int' [-Wformat=]
              left_len, sizeof(*phyerr));
                        ^
wmi.c: In function 'ath10k_wmi_10_4_op_pull_phyerr_ev':
wmi.c:3612:8: warning: format '%d' expects argument of type 'int',
	      but argument 4 has type 'long unsigned int' [-Wformat=]
              left_len, sizeof(*phyerr));
                        ^
Fixes: 991adf71 ("ath10k: refactor phyerr event handlers")
Fixes: 2b0a2e0d ("ath10k: handle 10.4 firmware phyerr event")
Signed-off-by: NRaja Mani <rmani@qti.qualcomm.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

This adds additional 0x0041 PCI Device ID
definition to ath10k for QCA6164 which is a 1
spatial stream sibling of the QCA6174 (which is 2
spatial stream chip).

The QCA6164 needs a dedicated board.bin file which
is different than the one used for QCA6174. If the
board.bin is wrong the device will crash early
while trying to boot firmware. The register dump
will look like this:

 ath10k_pci 0000:02:00.0: firmware register dump:
 ath10k_pci 0000:02:00.0: [00]: 0x05010000 0x000015B3 0x000A012D 0x00955B31
 ...

Note the value 0x000A012D.

Special credit goes to Alan Liu
<alanliu@qca.qualcomm.com> for providing support
help which enabled me to come up with this patch.
Signed-off-by: NMichal Kazior <michal.kazior@tieto.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

To enable/configure spectral scan parameters in 10.4 firmware, existing
wmi spectral related functions can be reused. Link those functions in
10.4 wmi ops table.

In addition, adjust bin size (only when size is 68 bytes) before reporting
bin samples to user space. The background for this adjustment is that
qca99x0 reports bin size as 68 bytes (64 bytes + 4 bytes) in report
mode 2. First 64 bytes carries in-band tones (-32 to +31) and last 4 byte
carries band edge detection data (+32) mainly used in radar detection
purpose. Additional last 4 bytes are stripped to make bin size valid one.

This bin size adjustment will happen only for qca99x0, all other chipsets
will report proper bin sizes (64/128) without extra 4 bytes being added
at the end. The changes are validated in qca99x0 using 10.4 firmware.
Signed-off-by: NRaja Mani <rmani@qti.qualcomm.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

The function returns 1 when DMA mapping fails. The
driver would return bogus values and could
possibly confuse itself if DMA failed.

Fixes: 767d34fc ("ath10k: remove DMA mapping wrappers")
Reported-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NMichal Kazior <michal.kazior@tieto.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

Kernel would complain about leaving a held lock
after going back to userspace and would
subsequently deadlock.

Fixes: e04cafbc ("ath10k: fix peer limit enforcement")
Reported-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NMichal Kazior <michal.kazior@tieto.com>
Signed-off-by: NKalle Valo <kvalo@qca.qualcomm.com>

Florian Fainelli says:

====================
Documentation: dsa

This patch series adds some documentation about DSA as a subsystem as well
as the SF2 driver since it slightly diverges from your average DSA driver ;)
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add a document describing the Broadcom Starfigther 2 switch hardware,
its specifics, and how the driver is implemented and its specifics.
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Reviewed-by: NVivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Describe how the DSA subsystem works, its design principles,
limitations, and describe in details how to implement a DSA switch
driver.
Acked-by: NAndrew Lunn <andrew@lunn.ch>
Acked-by: NScott Feldman <sfeldma@gmail.com>
Reviewed-by: NVivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Santosh Shilimkar says:

====================
RDS: Few more fixes

As indicated in the earlier series [1], this is a follow-up series which
addresses few issues around the RDS FMR code. With [1] and the subject
series, now I can run many parallel threads with multiple sockets with
N x N traffic. The stress tests has survived overnight runs.

[1] https://lkml.org/lkml/2015/8/22/127
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Memory allocated for 'ibmr' uses kzalloc_node() which already
initialises the memory to zero. There is no need to do
memset() 0 on that memory.
Signed-off-by: NSantosh Shilimkar <ssantosh@kernel.org>
Signed-off-by: NSantosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

FMR flush is an expensive and time consuming operation. Reduce the
frequency of FMR pool flush by 50% so that more FMR work gets accumulated
for more efficient flushing.
Signed-off-by: NSantosh Shilimkar <ssantosh@kernel.org>
Signed-off-by: NSantosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

RDS FMR flush operation and also it races with connect/reconect
which happes a lot with RDS. FMR flush being on common rds_wq aggrevates
the problem. Lets push RDS FMR pool flush work to its own worker.
Signed-off-by: NSantosh Shilimkar <ssantosh@kernel.org>
Signed-off-by: NSantosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In rds_ib_flush_mr_pool(), dirty_count accounts the clean ones
which is wrong. This can lead to a negative dirty count value.

Lets fix it.
Signed-off-by: NWengang Wang <wen.gang.wang@oracle.com>
Signed-off-by: NSantosh Shilimkar <ssantosh@kernel.org>
Signed-off-by: NSantosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

rds_rdma_unuse() drops the mr reference count which it hasn't
taken. Correct way of removing mr is to remove mr from the tree
and then rdma_destroy_mr() it first, then rds_mr_put() to decrement
its reference count. Whichever thread holds last reference will free
the mr via rds_mr_put()

This bug was triggering weird null pointer crashes. One if the trace
for it is captured below.

BUG: unable to handle kernel NULL pointer dereference at
0000000000000104
IP: [<ffffffffa0899471>] rds_ib_free_mr+0x31/0x130 [rds_rdma]
PGD 4366fa067 PUD 4366f9067 PMD 0
Oops: 0000 [#1] SMP

[...]

task: ffff88046da6a000 ti: ffff88046da6c000 task.ti: ffff88046da6c000
RIP: 0010:[<ffffffffa0899471>]  [<ffffffffa0899471>]
rds_ib_free_mr+0x31/0x130 [rds_rdma]
RSP: 0018:ffff88046fa43bd8  EFLAGS: 00010286
RAX: 0000000071d38b80 RBX: 0000000000000000 RCX: 0000000000000000
RDX: 0000000000000001 RSI: 0000000000000000 RDI: ffff880079e7ff40
RBP: ffff88046fa43bf8 R08: 0000000000000000 R09: 0000000000000000
R10: ffff88046fa43ca8 R11: ffff88046a802ed8 R12: ffff880079e7fa40
R13: 0000000000000000 R14: ffff880079e7ff40 R15: 0000000000000000
FS:  0000000000000000(0000) GS:ffff88046fa40000(0000)
knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 0000000000000104 CR3: 00000004366fb000 CR4: 00000000000006e0
Stack:
 ffff880079e7fa40 ffff880671d38f08 ffff880079e7ff40 0000000000000296
 ffff88046fa43c28 ffffffffa087a38b ffff880079e7fa40 ffff880671d38f10
 0000000000000000 0000000000000292 ffff88046fa43c48 ffffffffa087a3b6
Call Trace:
 <IRQ>
 [<ffffffffa087a38b>] rds_destroy_mr+0x8b/0xa0 [rds]
 [<ffffffffa087a3b6>] __rds_put_mr_final+0x16/0x30 [rds]
 [<ffffffffa087a492>] rds_rdma_unuse+0xc2/0x120 [rds]
 [<ffffffffa08766d3>] rds_recv_incoming_exthdrs+0x83/0xa0 [rds]
 [<ffffffffa0876782>] rds_recv_incoming+0x92/0x200 [rds]
 [<ffffffffa0895269>] rds_ib_process_recv+0x259/0x320 [rds_rdma]
 [<ffffffffa08962a8>] rds_ib_recv_tasklet_fn+0x1a8/0x490 [rds_rdma]
 [<ffffffff810dcd78>] ? __remove_hrtimer+0x58/0x90
 [<ffffffff810799e1>] tasklet_action+0xb1/0xc0
 [<ffffffff81079b52>] __do_softirq+0xe2/0x290
 [<ffffffff81079df6>] irq_exit+0xa6/0xb0
 [<ffffffff81613915>] do_IRQ+0x65/0xf0
 [<ffffffff816118ab>] common_interrupt+0x6b/0x6b
Signed-off-by: NSantosh Shilimkar <ssantosh@kernel.org>
Signed-off-by: NSantosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

On rds_ib_frag_slab allocation failure, ensure rds_ib_incoming_slab
is not pointing to the detsroyed memory.
Signed-off-by: NSantosh Shilimkar <ssantosh@kernel.org>
Signed-off-by: NSantosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Antonio Quartulli says:

====================
Included changes:
- code restyling and beautification
- use int kernel types instead of C99
- update kereldoc
- prevent potential hlist double deletion of VLAN objects
- fix gw bandwidth calculation
- convert list to hlist when needed
- add lockdep_asserts calls in function with lock requirements
  described in kerneldoc
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

posted_index is RO in firmware. We need not do ioread everytime to get
posted index. Store posted index locally.
Signed-off-by: NGovindarajulu Varadarajan <_govind@gmx.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The r8169 driver collects statistical information returned by
@get_stats64 by counting them in the driver itself, even though many
(but not all) of the values are already collected by tally counters
(TCs) in the NIC.  Some of these TC values are not returned by
@get_stats64.  Especially the received multicast packages are missing
from /proc/net/dev.

Rectify this by fetching the TCs and returning them from
rtl8169_get_stats64.

The counters collected in the driver obviously disappear as soon as the
driver is unloaded so after a driver is loaded the counters always start
at 0. The TCs on the other hand are only reset by a power cycle.  Without
further considerations the values collected by the driver would not match
up against the TC values.

This patch introduces a new function rtl8169_reset_counters which
resets the TCs.  Also, since rtl8169_reset_counters shares most of
its code with rtl8169_update_counters, refactor the shared code into
two new functions  rtl8169_map_counters and rtl8169_unmap_counters.

Unfortunately chip versions prior to RTL_GIGA_MAC_VER_19 don't allow
to reset the TCs programatically.  Therefore introduce an addition to
the rtl8169_private struct and a function rtl8169_init_counter_offsets
to store the TCs at first rtl_open.  Use these values as offsets in
rtl8169_get_stats64.  Propagate a failure to reset *and* update the
counters up to rtl_open and emit a warning message, if so.
Signed-off-by: NCorinna Vinschen <vinschen@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

>> net/rds/ib_recv.c:382:28: sparse: incorrect type in initializer (different base types)
   net/rds/ib_recv.c:382:28:    expected int [signed] can_wait
   net/rds/ib_recv.c:382:28:    got restricted gfp_t
   net/rds/ib_recv.c:828:23: sparse: cast to restricted __le64
Reported-by: Nkbuild test robot <fengguang.wu@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Shreyas Bhatewara would no longer maintain the vmxnet3 driver. Taking over
the role of vmxnet3 maintainer.
Signed-off-by: NShrikrishna Khare <skhare@vmware.com>
Signed off-by: Shreyas Bhatewara <sbhatewara@vmware.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The vxlan_get_sk_family inline function was added after the last #endif,
making multiple inclusion of net/vxlan.h fail. Move it to the proper place.
Reported-by: NMark Rustad <mark.d.rustad@intel.com>
Fixes: 705cc62f ("vxlan: provide access function for vxlan socket address family")
Signed-off-by: NJiri Benc <jbenc@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add entry for new VRF device driver.
Signed-off-by: NDavid Ahern <dsa@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch fixes the following crash:

 general protection fault: 0000 [#1] SMP DEBUG_PAGEALLOC
 CPU: 1 PID: 0 Comm: swapper/1 Not tainted 4.2.0-rc7+ #166
 Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011
 task: ffff88010656d280 ti: ffff880106570000 task.ti: ffff880106570000
 RIP: 0010:[<ffffffff8182f91b>]  [<ffffffff8182f91b>] dst_destroy+0xa6/0xef
 RSP: 0018:ffff880107603e38  EFLAGS: 00010202
 RAX: 0000000000000001 RBX: ffff8800d225a000 RCX: ffffffff82250fd0
 RDX: 0000000000000001 RSI: ffffffff82250fd0 RDI: 6b6b6b6b6b6b6b6b
 RBP: ffff880107603e58 R08: 0000000000000001 R09: 0000000000000001
 R10: 000000000000b530 R11: ffff880107609000 R12: 0000000000000000
 R13: ffffffff82343c40 R14: 0000000000000000 R15: ffffffff8182fb4f
 FS:  0000000000000000(0000) GS:ffff880107600000(0000) knlGS:0000000000000000
 CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
 CR2: 00007fcabd9d3000 CR3: 00000000d7279000 CR4: 00000000000006e0
 Stack:
  ffffffff82250fd0 ffff8801077d6f00 ffffffff82253c40 ffff8800d225a000
  ffff880107603e68 ffffffff8182fb5d ffff880107603f08 ffffffff810d795e
  ffffffff810d7648 ffff880106574000 ffff88010656d280 ffff88010656d280
 Call Trace:
  <IRQ>
  [<ffffffff8182fb5d>] dst_destroy_rcu+0xe/0x1d
  [<ffffffff810d795e>] rcu_process_callbacks+0x618/0x7eb
  [<ffffffff810d7648>] ? rcu_process_callbacks+0x302/0x7eb
  [<ffffffff8182fb4f>] ? dst_gc_task+0x1eb/0x1eb
  [<ffffffff8107e11b>] __do_softirq+0x178/0x39f
  [<ffffffff8107e52e>] irq_exit+0x41/0x95
  [<ffffffff81a4f215>] smp_apic_timer_interrupt+0x34/0x40
  [<ffffffff81a4d5cd>] apic_timer_interrupt+0x6d/0x80
  <EOI>
  [<ffffffff8100b968>] ? default_idle+0x21/0x32
  [<ffffffff8100b966>] ? default_idle+0x1f/0x32
  [<ffffffff8100bf19>] arch_cpu_idle+0xf/0x11
  [<ffffffff810b0bc7>] default_idle_call+0x1f/0x21
  [<ffffffff810b0dce>] cpu_startup_entry+0x1ad/0x273
  [<ffffffff8102fe67>] start_secondary+0x135/0x156

dst is freed right before lwtstate_put(), this is not correct...

Fixes: 61adedf3 ("route: move lwtunnel state to dst_entry")
Acked-by: NJiri Benc <jbenc@redhat.com>
Signed-off-by: NCong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: NCong Wang <cwang@twopensource.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>