10/100 MACB does not have the same statistics possibilities as GEM. Separate
macb_ethtool_ops to make a new GEM-specific struct with the new statistics
functions included.
Signed-off-by: Xander Huff <xander.huff@ni.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Spaces should surround add, multiply, and bitshift operators.
Signed-off-by: Xander Huff <xander.huff@ni.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Change comments to not exceed 80 characters per line.
Update block comments in macb.h to start on the line after /*.
Signed-off-by: Xander Huff <xander.huff@ni.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Gao Zhenyu <gzhenyu@vmware.com>
Signed-off-by: Shrikrishna Khare <skhare@vmware.com>
Reviewed-by: Shreyas N Bhatewara <sbhatewara@vmware.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This is just a cleanup, because in the current code MDBA_SET_ENTRY_MAX ==
MDBA_SET_ENTRY.
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add Atari specific code to the smc91x Ethernet driver. This code is used
on the EtherNAT adapter card for the Atari Falcon extension port.
Signed-off-by: Michael Schmitz <schmitz@debian.org>
Tested-by: Christian Steigies <cts@debian.org>
[geert: Sort Kconfig entries, split in hard and soft dependencies]
Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Acked-by: Nicolas Pitre <nico@fluxnic.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge tag 'mac80211-next-for-davem-2015-01-15' of git://git.kernel.org/pub/scm/linux/kernel/git/jberg/mac80211-next

Here's a big pile of changes for this round.

We have
 * a lot of regulatory code changes to deal with the
   way newer Intel devices handle this
 * a change to drop packets while disconnecting from
   an AP instead of trying to wait for them
 * a new attempt at improving the tailroom accounting
   to not kick in too much for performance reasons
 * improvements in wireless link statistics
 * many other small improvements and small fixes that
   didn't seem necessary for 3.19 (e.g. in hwsim which
   is testing only code)

Conflicts:
	drivers/staging/rtl8723au/os_dep/ioctl_cfg80211.c

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

Anish Bhatt says:

====================
cxgb4/cxgb4i : Update & use ipv6 handling api

This patch series consolidates and updates the ipv6 api, as well as exports
it for use by upper level drivers dependent on cxgb4

v2: Fix formatting issues in clip_tbl.c
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Get a reference on every ipv6 address we offload to hardware so that it cannot
be prematurely cleared out before cleanup.
Signed-off-by: Anish Bhatt <anish@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch improves on previously added support for ipv6 addresses. The code
is consolidated to a single file and adds an api for use by dependent upper
level drivers such as cxgb4i/iw_cxgb4 etc.
Signed-off-by: Anish Bhatt <anish@chelsio.com>
Signed-off-by: Deepak Singh <deepak.s@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

RAW sockets with hdrinc suffer from contention on rt_uncached_lock
spinlock.

One solution is to use percpu lists, since most routes are destroyed
by the cpu that created them.

It is unclear why we even have to put these routes in uncached_list,
as all outgoing packets should be freed when a device is dismantled.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Fixes: caacf05e ("ipv4: Properly purge netdev references on uncached routes.")
Signed-off-by: David S. Miller <davem@davemloft.net>

Replace "unsigned long" by "u32" when handling 32-bit register data.

A.o., this saves 11 * 36 bytes in the static sh_eth_cpu_data structures
when compiling for 64-bit.
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: David S. Miller <davem@davemloft.net>

In boards, the dm9000 chip's power and reset can be controlled by gpio.

It makes sense to add them to the dm9000 driver and let dt be used to
enable power and reset the phy.
Signed-off-by: Zubair Lutfullah Kakakhel <Zubair.Kakakhel@imgtec.com>
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Better to use available helpers.
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Pablo Neira Ayuso says:

====================
netfilter updates for net-next

The following patchset contains netfilter updates for net-next, just a
bunch of cleanups and small enhancement to selectively flush conntracks
in ctnetlink, more specifically the patches are:

1) Rise default number of buckets in conntrack from 16384 to 65536 in
   systems with >= 4GBytes, patch from Marcelo Leitner.

2) Small refactor to save one level on indentation in xt_osf, from
   Joe Perches.

3) Remove unnecessary sizeof(char) in nf_log, from Fabian Frederick.

4) Another small cleanup to remove redundant variable in nfnetlink,
   from Duan Jiong.

5) Fix compilation warning in nfnetlink_cthelper on parisc, from
   Chen Gang.

6) Fix wrong format in debugging for ctseqadj, from Gao feng.

7) Selective conntrack flushing through the mark for ctnetlink, patch
   from Kristian Evensen.

8) Remove nf_ct_conntrack_flush_report() exported symbol now that is
   not required anymore after the selective flushing patch, again from
   Kristian.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Thomas Graf says:

====================
VXLAN Group Policy Extension

Implements supports for the Group Policy VXLAN extension [0] to provide
a lightweight and simple security label mechanism across network peers
based on VXLAN. The security context and associated metadata is mapped
to/from skb->mark. This allows further mapping to a SELinux context
using SECMARK, to implement ACLs directly with nftables, iptables, OVS,
tc, etc.

The extension is disabled by default and should be run on a distinct
port in mixed Linux VXLAN VTEP environments. Liberal VXLAN VTEPs
which ignore unknown reserved bits will be able to receive VXLAN-GBP
frames.

Simple usage example:

10.1.1.1:
   # ip link add vxlan0 type vxlan id 10 remote 10.1.1.2 gbp
   # iptables -I OUTPUT -m owner --uid-owner 101 -j MARK --set-mark 0x200

10.1.1.2:
   # ip link add vxlan0 type vxlan id 10 remote 10.1.1.1 gbp
   # iptables -I INPUT -m mark --mark 0x200 -j DROP

iproute2 [1] and OVS [2] support will be provided in separate patches.

[0] https://tools.ietf.org/html/draft-smith-vxlan-group-policy
[1] https://github.com/tgraf/iproute2/tree/vxlan-gbp
[2] https://github.com/tgraf/ovs/tree/vxlan-gbp
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Introduces support for the group policy extension to the VXLAN virtual
port. The extension is disabled by default and only enabled if the user
has provided the respective configuration.

  ovs-vsctl add-port br0 vxlan0 -- \
     set Interface vxlan0 type=vxlan options:exts=gbp

The configuration interface to enable the extension is based on a new
attribute OVS_VXLAN_EXT_GBP nested inside OVS_TUNNEL_ATTR_EXTENSION
which can carry additional extensions as needed in the future.

The group policy metadata is stored as binary blob (struct ovs_vxlan_opts)
internally just like Geneve options but transported as nested Netlink
attributes to user space.

Renames the existing TUNNEL_OPTIONS_PRESENT to TUNNEL_GENEVE_OPT with the
binary value kept intact, a new flag TUNNEL_VXLAN_OPT is introduced.

The attributes OVS_TUNNEL_KEY_ATTR_VXLAN_OPTS and existing
OVS_TUNNEL_KEY_ATTR_GENEVE_OPTS are implemented mutually exclusive.
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

nlattr_set() is currently hardcoded to two levels of nesting. This change
introduces struct ovs_len_tbl to define minimal length requirements plus
next level nesting tables to traverse the key attributes to arbitrary depth.
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

Also factors out Geneve validation code into a new separate function
validate_and_copy_geneve_opts().

A subsequent patch will introduce VXLAN options. Rename the existing
GENEVE_TUN_OPTS() to reflect its extended purpose of carrying generic
tunnel metadata options.
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

A VXLAN net_device looking for an appropriate socket may only consider
a socket which has a matching set of flags/extensions enabled. If
incompatible flags are enabled, return a conflict to have the caller
create a distinct socket with distinct port.

The OVS VXLAN port is kept unaware of extensions at this point.
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

Implements supports for the Group Policy VXLAN extension [0] to provide
a lightweight and simple security label mechanism across network peers
based on VXLAN. The security context and associated metadata is mapped
to/from skb->mark. This allows further mapping to a SELinux context
using SECMARK, to implement ACLs directly with nftables, iptables, OVS,
tc, etc.

The group membership is defined by the lower 16 bits of skb->mark, the
upper 16 bits are used for flags.

SELinux allows to manage label to secure local resources. However,
distributed applications require ACLs to implemented across hosts. This
is typically achieved by matching on L2-L4 fields to identify the
original sending host and process on the receiver. On top of that,
netlabel and specifically CIPSO [1] allow to map security contexts to
universal labels.  However, netlabel and CIPSO are relatively complex.
This patch provides a lightweight alternative for overlay network
environments with a trusted underlay. No additional control protocol
is required.

           Host 1:                       Host 2:

      Group A        Group B        Group B     Group A
      +-----+   +-------------+    +-------+   +-----+
      | lxc |   | SELinux CTX |    | httpd |   | VM  |
      +--+--+   +--+----------+    +---+---+   +--+--+
	  \---+---/                     \----+---/
	      |                              |
	  +---+---+                      +---+---+
	  | vxlan |                      | vxlan |
	  +---+---+                      +---+---+
	      +------------------------------+

Backwards compatibility:
A VXLAN-GBP socket can receive standard VXLAN frames and will assign
the default group 0x0000 to such frames. A Linux VXLAN socket will
drop VXLAN-GBP  frames. The extension is therefore disabled by default
and needs to be specifically enabled:

   ip link add [...] type vxlan [...] gbp

In a mixed environment with VXLAN and VXLAN-GBP sockets, the GBP socket
must run on a separate port number.

Examples:
 iptables:
  host1# iptables -I OUTPUT -m owner --uid-owner 101 -j MARK --set-mark 0x200
  host2# iptables -I INPUT -m mark --mark 0x200 -j DROP

 OVS:
  # ovs-ofctl add-flow br0 'in_port=1,actions=load:0x200->NXM_NX_TUN_GBP_ID[],NORMAL'
  # ovs-ofctl add-flow br0 'in_port=2,tun_gbp_id=0x200,actions=drop'

[0] https://tools.ietf.org/html/draft-smith-vxlan-group-policy
[1] http://lwn.net/Articles/204905/Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

Conflicts:
	drivers/net/xen-netfront.c

Minor overlapping changes in xen-netfront.c, mostly to do
with some buffer management changes alongside the split
of stats into TX and RX.
Signed-off-by: David S. Miller <davem@davemloft.net>

Pull networking fixes from David Miller:

 1) Don't use uninitialized data in IPVS, from Dan Carpenter.

 2) conntrack race fixes from Pablo Neira Ayuso.

 3) Fix TX hangs with i40e, from Jesse Brandeburg.

 4) Fix budget return from poll calls in dnet and alx, from Eric
    Dumazet.

 5) Fix bugus "if (unlikely(x) < 0)" test in AF_PACKET, from Christoph
    Jaeger.

 6) Fix bug introduced by conversion to list_head in TIPC retransmit
    code, from Jon Paul Maloy.

 7) Don't use GFP_NOIO under spinlock in USB kaweth driver, from Alexey
    Khoroshilov.

 8) Fix bridge build with INET disabled, from Arnd Bergmann.

 9) Fix netlink array overrun for PROBE attributes in openvswitch, from
    Thomas Graf.

10) Don't hold spinlock across synchronize_irq() in tg3 driver, from
    Prashant Sreedharan.

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (44 commits)
  tg3: Release tp->lock before invoking synchronize_irq()
  tg3: tg3_reset_task() needs to use rtnl_lock to synchronize
  tg3: tg3_timer() should grab tp->lock before checking for tp->irq_sync
  team: avoid possible underflow of count_pending value for notify_peers and mcast_rejoin
  openvswitch: packet messages need their own probe attribtue
  i40e: adds FCoE configure option
  cxgb4vf: Fix queue allocation for 40G adapter
  netdevice: Add missing parentheses in macro
  bridge: only provide proxy ARP when CONFIG_INET is enabled
  neighbour: fix base_reachable_time(_ms) not effective immediatly when changed
  net: fec: fix MDIO bus assignement for dual fec SoC's
  xen-netfront: use different locks for Rx and Tx stats
  drivers: net: cpsw: fix multicast flush in dual emac mode
  cxgb4vf: Initialize mdio_addr before using it
  net: Corrected the comment describing the ndo operations to reflect the actual prototype for couple of operations
  usb/kaweth: use GFP_ATOMIC under spin_lock in usb_start_wait_urb()
  MAINTAINERS: add me as ibmveth maintainer
  tipc: fix bug in broadcast retransmit code
  update ip-sysctl.txt documentation (v2)
  net/at91_ether: prepare and unprepare clock
  ...

Prashant Sreedharan says:

====================
tg3: synchronize_irq() should be called without taking locks

v2: Added Reported-by, Tested-by fields and reference to the thread that
    reported the problem

This series addresses the problem reported by Peter Hurley in mail thread
https://lkml.org/lkml/2015/1/12/1082
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

synchronize_irq() can sleep waiting, for pending IRQ handlers so driver
should release the tp->lock spin lock before invoking synchronize_irq()
Reported-by: Peter Hurley <peter@hurleysoftware.com>
Tested-by: Peter Hurley <peter@hurleysoftware.com>
Signed-off-by: Prashant Sreedharan <prashant@broadcom.com>
Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Currently tg3_reset_task() uses only tp->lock for synchronizing with code
paths like tg3_open() etc. But since tp->lock is released before doing
synchronize_irq(), rtnl_lock should be taken in tg3_reset_task() to
synchronize it with other code paths.
Reported-by: Peter Hurley <peter@hurleysoftware.com>
Tested-by: Peter Hurley <peter@hurleysoftware.com>
Signed-off-by: Prashant Sreedharan <prashant@broadcom.com>
Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This is to avoid the race between tg3_timer() and the execution paths
which does not invoke tg3_timer_stop() and releases tp->lock before
calling synchronize_irq()
Reported-by: Peter Hurley <peter@hurleysoftware.com>
Tested-by: Peter Hurley <peter@hurleysoftware.com>
Signed-off-by: Prashant Sreedharan <prashant@broadcom.com>
Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Pull kvm fixes from Paolo Bonzini:
 "Two bugfixes for arm64.  I will have another pull request next week,
  but otherwise things are calm"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
  arm64: KVM: Fix HCR setting for 32bit guests
  arm64: KVM: Fix TLB invalidation by IPA/VMID

This patch is fixing a race condition that may cause setting
count_pending to -1, which results in unwanted big bulk of arp messages
(in case of "notify peers").

Consider following scenario:

count_pending == 2
   CPU0                                           CPU1
					team_notify_peers_work
					  atomic_dec_and_test (dec count_pending to 1)
					  schedule_delayed_work
 team_notify_peers
   atomic_add (adding 1 to count_pending)
					team_notify_peers_work
					  atomic_dec_and_test (dec count_pending to 1)
					  schedule_delayed_work
					team_notify_peers_work
					  atomic_dec_and_test (dec count_pending to 0)
   schedule_delayed_work
					team_notify_peers_work
					  atomic_dec_and_test (dec count_pending to -1)

Fix this race by using atomic_dec_if_positive - that will prevent
count_pending running under 0.

Fixes: fc423ff0 ("team: add peer notification")
Fixes: 492b200e  ("team: add support for sending multicast rejoins")
Signed-off-by: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: Jiri Benc <jbenc@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Pull s390 fixes from Martin Schwidefsky:
 "Two small performance tweaks, the plumbing for the execveat system
  call and a couple of bug fixes"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux:
  s390/uprobes: fix user space PER events
  s390/bpf: Fix JMP_JGE_X (A > X) and JMP_JGT_X (A >= X)
  s390/bpf: Fix ALU_NEG (A = -A)
  s390/mm: avoid using pmd_to_page for !USE_SPLIT_PMD_PTLOCKS
  s390/timex: fix get_tod_clock_ext() inline assembly
  s390: wire up execveat syscall
  s390/kernel: use stnsm 255 instead of stosm 0
  s390/vtime: Get rid of redundant WARN_ON
  s390/zcrypt: kernel oops at insmod of the z90crypt device driver

User space is currently sending a OVS_FLOW_ATTR_PROBE for both flow
and packet messages. This leads to an out-of-bounds access in
ovs_packet_cmd_execute() because OVS_FLOW_ATTR_PROBE >
OVS_PACKET_ATTR_MAX.

Introduce a new OVS_PACKET_ATTR_PROBE with the same numeric value
as OVS_FLOW_ATTR_PROBE to grow the range of accepted packet attributes
while maintaining to be binary compatible with existing OVS binaries.

Fixes: 05da5898 ("openvswitch: Add support for OVS_FLOW_ATTR_PROBE.")
Reported-by: Sander Eikelenboom <linux@eikelenboom.it>
Tracked-down-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Reviewed-by: Jesse Gross <jesse@nicira.com>
Acked-by: Pravin B Shelar <pshelar@nicira.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Adds FCoE config option I40E_FCOE, so that FCoE can be enabled
as needed but otherwise have it disabled by default.

This also eliminate multiple FCoE config checks, instead now just
one config check for CONFIG_I40E_FCOE.

The I40E FCoE was added with 3.17 kernel and therefore this patch
shall be applied to stable 3.17 kernel also.

CC: <stable@vger.kernel.org>
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
Tested-by: Jim Young <jamesx.m.young@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Hariprasad Shenai <hariprasad@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Pull file locking fix from Jeff Layton:
 "Just a simple bugfix for a regression that I introduced into v3.18
  with the internal lease API overhaul -- mea culpa.  Kudos to Linda and
  Neil for tracking this down and fixing it"

* tag 'locks-v3.19-1' of git://git.samba.org/jlayton/linux:
  locks: fix NULL-deref in generic_delete_lease

There are too many brackets. Maybe only one bracket is enough.
Signed-off-by: Zhu Yanjun <Yanjun.Zhu@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

For example, one could conceivably call
	for_each_netdev_in_bond_rcu(condition ? bond1 : bond2, slave)
and get an unexpected result.
Signed-off-by: Benjamin Poirier <bpoirier@suse.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

Introduce ovs_tunnel_route_lookup to consolidate route lookup
shared by vxlan, gre, and geneve ports.
Signed-off-by: Fan Du <fan.du@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Pull block layer fixes from Jens Axboe:
 "The major part is an update to the NVMe driver, fixing various issues
  around surprise removal and hung controllers.  Most of that is from
  Keith, and parts are simple blk-mq fixes or exports/additions of minor
  functions to aid this effort, and parts are changes directly to the
  NVMe driver.

  Apart from the above, this contains:

   - Small blk-mq change from me, killing an unused member of the
     hardware queue structure.

   - Small fix from Ming Lei, fixing up a few drivers that didn't
     properly check for ERR_PTR() returns from blk_mq_init_queue()"

* 'for-linus' of git://git.kernel.dk/linux-block:
  NVMe: Fix locking on abort handling
  NVMe: Start and stop h/w queues on reset
  NVMe: Command abort handling fixes
  NVMe: Admin queue removal handling
  NVMe: Reference count admin queue usage
  NVMe: Start all requests
  blk-mq: End unstarted requests on a dying queue
  blk-mq: Allow requests to never expire
  blk-mq: Add helper to abort requeued requests
  blk-mq: Let drivers cancel requeue_work
  blk-mq: Export if requests were started
  blk-mq: Wake tasks entering queue on dying
  blk-mq: get rid of ->cmd_size in the hardware queue
  block: fix checking return value of blk_mq_init_queue
  block: wake up waiters when a queue is marked dying
  NVMe: Fix double free irq
  blk-mq: Export freeze/unfreeze functions
  blk-mq: Exit queue on alloc failure

Tom Herbert says:

====================
net: Remote checksum offload for VXLAN

This patch set adds support for remote checksum offload in VXLAN.

The remote checksum offload is generalized by creating a common
function (remcsum_adjust) that does the work of modifying the
checksum in remote checksum offload. This function can be called
from normal or GRO path. GUE was modified to use this function.

To support RCO is VXLAN we use the 9th bit in the reserved
flags to indicated remote checksum offload. The start and offset
values are encoded n a compressed form in the low order (reserved)
byte of the vni field.

Remote checksum offload is described in
https://tools.ietf.org/html/draft-herbert-remotecsumoffload-01

Changes in v2:
  - Add udp_offload_callbacks which has GRO functions that take a
    udp_offload pointer argument. This argument can be used to retrieve
    a per port structure of the encapsulation for use in gro processing
    (mostly by doing container_of on the structure).
  - Use the 10th bit in VXLAN flags for RCO which does not seem to
    conflict with other proposals at this time (ie. VXLAN-GPE and
    VXLAN-GPB)
  - Require that RCO must be explicitly enabled on the receiver
    as well as the sender.

Tested by running 200 TCP_STREAM connections with VXLAN (over IPv4).

With UDP checksums and Remote Checksum Offload
  IPv4
      Client
        11.84% CPU utilization
      Server
        12.96% CPU utilization
      9197 Mbps
  IPv6
      Client
        12.46% CPU utilization
      Server
        14.48% CPU utilization
      8963 Mbps

With UDP checksums, no remote checksum offload
  IPv4
      Client
        15.67% CPU utilization
      Server
        14.83% CPU utilization
      9094 Mbps
  IPv6
      Client
        16.21% CPU utilization
      Server
        14.32% CPU utilization
      9058 Mbps

No UDP checksums
  IPv4
      Client
        15.03% CPU utilization
      Server
        23.09% CPU utilization
      9089 Mbps
  IPv6
      Client
        16.18% CPU utilization
      Server
        26.57% CPU utilization
       8954 Mbps
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Add support for remote checksum offload in VXLAN. This uses a
reserved bit to indicate that RCO is being done, and uses the low order
reserved eight bits of the VNI to hold the start and offset values in a
compressed manner.

Start is encoded in the low order seven bits of VNI. This is start >> 1
so that the checksum start offset is 0-254 using even values only.
Checksum offset (transport checksum field) is indicated in the high
order bit in the low order byte of the VNI. If the bit is set, the
checksum field is for UDP (so offset = start + 6), else checksum
field is for TCP (so offset = start + 16). Only TCP and UDP are
supported in this implementation.

Remote checksum offload for VXLAN is described in:

https://tools.ietf.org/html/draft-herbert-vxlan-rco-00

Tested by running 200 TCP_STREAM connections with VXLAN (over IPv4).

With UDP checksums and Remote Checksum Offload
  IPv4
      Client
        11.84% CPU utilization
      Server
        12.96% CPU utilization
      9197 Mbps
  IPv6
      Client
        12.46% CPU utilization
      Server
        14.48% CPU utilization
      8963 Mbps

With UDP checksums, no remote checksum offload
  IPv4
      Client
        15.67% CPU utilization
      Server
        14.83% CPU utilization
      9094 Mbps
  IPv6
      Client
        16.21% CPU utilization
      Server
        14.32% CPU utilization
      9058 Mbps

No UDP checksums
  IPv4
      Client
        15.03% CPU utilization
      Server
        23.09% CPU utilization
      9089 Mbps
  IPv6
      Client
        16.18% CPU utilization
      Server
        26.57% CPU utilization
       8954 Mbps
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>