Always fully coalesce guest Rx packets into the minimum number of ring
slots.  Reducing the number of slots per packet has significant
performance benefits when receiving off-host traffic.

Results from XenServer's performance benchmarks:

                         Baseline    Full coalesce
Interhost VM receive      7.2 Gb/s   11 Gb/s
Interhost aggregate      24 Gb/s     24 Gb/s
Intrahost single stream  14 Gb/s     14 Gb/s
Intrahost aggregate      34 Gb/s     34 Gb/s

However, this can increase the number of grant ops per packet which
decreases performance of backend (dom0) to VM traffic (by ~10%)
/unless/ grant copy has been optimized for adjacent ops with the same
source or destination (see "grant-table: defer releasing pages
acquired in a grant copy"[1] expected in Xen 4.6).

[1] http://lists.xen.org/archives/html/xen-devel/2015-01/msg01118.html



Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Nicolas Dichtel says:

====================
netns: advertise netns via netlink

The first patch of the series fix a bug of the previous series (present in
net-next only).
The rest of the series adds an attribute to advertise the peer netns for
rtnetlink messages where this information is needed by userland to be able to
interpret fully the received message.
====================

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

Netlink FDB messages are sent in the link netns. The header of these messages
contains the ifindex (ndm_ifindex) of the netdevice, but this ifindex is
unusable in case of x-netns vxlan.
I named the new attribute NDA_NDM_IFINDEX_NETNSID, to avoid confusion with
NDA_IFINDEX.

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

Assign rtnl_link_ops->get_link_net() callback so that IFLA_LINK_NETNSID is
added to rtnetlink messages.

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

Assign rtnl_link_ops->get_link_net() callback so that IFLA_LINK_NETNSID is
added to rtnetlink messages.

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

Assign rtnl_link_ops->get_link_net() callback so that IFLA_LINK_NETNSID is
added to rtnetlink messages.

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

Assign rtnl_link_ops->get_link_net() callback so that IFLA_LINK_NETNSID is
added to rtnetlink messages.

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

If an error occurs when the netdevice is moved to the link netns, a full cleanup
must be done.

Fixes: 317f4810

 ("rtnl: allow to create device with IFLA_LINK_NETNSID set")
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Depending on NETFILTER is not sufficient to ensure the presence of the
'mark' field in nf_conn, also needs to depend on NF_CONNTRACK_MARK.

Fixes: 22a5dc ("net: sched: Introduce connmark action")
Cc: Felix Fietkau <nbd@openwrt.org>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge tag 'mac80211-next-for-davem-2015-01-19' of git://git.kernel.org/pub/scm/linux/kernel/git/jberg/mac80211-next



Some further updates for net-next:
 * fix network-manager which was broken by the previous changes
 * fix delete-station events, which were broken by me making the
   genlmsg_end() mistake
 * fix a timer left running during suspend in some race conditions
   that would cause an annoying (but harmless) warning
 * (less important, but in the tree already) remove 80+80 MHz rate
   reporting since the spec doesn't distinguish it from 160 MHz;
   as the bitrate they're both 160 MHz bandwidth

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

My previous patch to this file changed the code to be bug-compatible
towards userspace. Unless userspace (which I wasn't able to find)
implements the dump reader by hand in a wrong way, this isn't needed.
If it uses libnl or similar code putting multiple messages into a
single SKB is far more efficient.

Change the code to do this. While at it, also clean it up and don't
use so many variables - just store the address in the callback args
directly.

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

The commit (3d125f9c

) cause i.MX6SX sdb enet cannot work. The cause is
the commit add mdio node with un-correct phy address.

The patch just correct i.MX6sx sdb board enet phy address.

Signed-off-by: Fugang Duan <B38611@freescale.com>
Acked-by: Stefan Agner <stefan@agner.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

This tc action allows you to retrieve the connection tracking mark
This action has been used heavily by openwrt for a few years now.

There are known limitations currently:

doesn't work for initial packets, since we only query the ct table.
  Fine given use case is for returning packets

no implicit defrag.
  frags should be rare so fix later..

won't work for more complex tasks, e.g. lookup of other extensions
  since we have no means to store results

we still have a 2nd lookup later on via normal conntrack path.
This shouldn't break anything though since skb->nfct isn't altered.

V2:
remove unnecessary braces (Jiri)
change the action identifier to 14 (Jiri)
Fix some stylistic issues caught by checkpatch
V3:
Move module params to bottom (Cong)
Get rid of tcf_hashinfo_init and friends and conform to newer API (Cong)

Acked-by: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Florian Fainelli says:

====================
net: DSA fixes for bridge and ip-autoconf

These two patches address some real world use cases of the DSA master and slave
network devices.

You have already seen patch 1 previously and you rejected it since my
explanations were not good enough to provide a justification as to why it is
useful, hopefully this time my explanation is better.

Patch 2 solves a different, yet very real problem as well at the bridge layer
when using DSA network devices.
====================

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

DSA-enabled master network devices with a switch tagging protocol should
strip the protocol specific format before handing the frame over to
higher layer.

When adding such a DSA master network device as a bridge member, we go
through the following code path when receiving a frame:

__netif_receive_skb_core
	-> first ptype check against ptype_all is not returning any
	   handler for this skb

	-> check and invoke rx_handler:
		-> deliver frame to the bridge layer: br_handle_frame

DSA registers a ptype handler with the fake ETH_XDSA ethertype, which is
called *after* the bridge-layer rx_handler has run. br_handle_frame()
tries to parse the frame it received from the DSA master network device,
and will not be able to match any of its conditions and jumps straight
at the end of the end of br_handle_frame() and returns
RX_HANDLER_CONSUMED there.

Since we returned RX_HANDLER_CONSUMED, __netif_receive_skb_core() stops
RX processing for this frame and returns NET_RX_SUCCESS, so we never get
a chance to call our switch tag packet processing logic and deliver
frames to the DSA slave network devices, and so we do not get any
functional bridge members at all.

Instead of cluttering the bridge receive path with DSA-specific checks,
and rely on assumptions about how __netif_receive_skb_core() is
processing frames, we simply deny adding the DSA master network device
(conduit interface) as a bridge member, leaving only the slave DSA
network devices to be bridge members, since those will work correctly in
all circumstances.

Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The logic to configure a network interface for kernel IP
auto-configuration is very simplistic, and does not handle the case
where a device is stacked onto another such as with DSA. This causes the
kernel not to open and configure the master network device in a DSA
switch tree, and therefore slave network devices using this master
network devices as conduit device cannot be open.

This restriction comes from a check in net/dsa/slave.c, which is
basically checking the master netdev flags for IFF_UP and returns
-ENETDOWN if it is not the case.

Automatically bringing-up DSA master network devices allows DSA slave
network devices to be used as valid interfaces for e.g: NFS root booting
by allowing kernel IP autoconfiguration to succeed on these interfaces.

On the reverse path, make sure we do not attempt to close a DSA-enabled
device as this would implicitely prevent the slave DSA network device
from operating.

Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

mii_check_media() does not update the link (carrier) state or log link
changes when the link mode is forced.  Drivers using the mii library
must do this themselves, but most of them do not.

Instead of changing them all, provide a sensible default behaviour
similar to mii_check_link() when the mode is forced.

via-rhine depends on it being a no-op in this case, so make its call
to mii_check_media() conditional.

Signed-off-by: Ben Hutchings <ben.hutchings@codethink.co.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>

Praveen Madhavan says:

====================
csiostor: Remove T4 FCoE support

We found a subtle issue with FCoE on T4 very late in the game
and decided not to productize FCoE on T4 and therefore there
are no customers that will be impacted by this change. FCoE is
supported on T5 cards.

Please apply on net-next since depends on previous commits.

Changes in v2:
  - Make the commit message more clearer.
====================

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

We have decided not to productize FCoE on T4.
Hence file is removed.

Signed-off-by: Praveen Madhavan <praveenm@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

We found a subtle issue with FCoE on T4 very late in the game
and decided not to productize FCoE on T4 and therefore there
are no customers that will be impacted by this change. Hence
T4 FCoE support is removed. FCoE supported only on T5 cards.

changes in v2:
  - Make the commit message more clearer.

Signed-off-by: Praveen Madhavan <praveenm@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Murali Karicheri says:

====================
net: Add Keystone NetCP ethernet driver support

The Network Coprocessor (NetCP) is a hardware accelerator that processes
Ethernet packets. NetCP has a gigabit Ethernet (GbE) subsystem with a ethernet
switch sub-module to send and receive packets. NetCP also includes a packet
accelerator (PA) module to perform packet classification operations such as
header matching, and packet modification operations such as checksum
generation. NetCP can also optionally include a Security Accelerator(SA)
capable of performing IPSec operations on ingress/egress packets.

Keystone SoC's also have a 10 Gigabit Ethernet Subsystem (XGbE) which
includes a 3-port Ethernet switch sub-module capable of 10Gb/s and
1Gb/s rates per Ethernet port.

Both GBE and XGBE network processors supported using common driver. It
is also designed to handle future variants of NetCP.

version history
---------------
v7->v8

 - Reworked comments against v7, related to checker warning.
 - Patch 2/4 that has all of the driver code in v7 is now split into 3
   patches based on functionality so that we have 3 smaller patches
   review instead of a big patch.
 - Patch for MAINTAINER is merged to 2/4 along with netcp core driver
 - Separate patch (3/4) for 1G and  (4/4) for 10G
 - Removed big endian support for initial version (will add it later)

v6->v7
 - Fixed some minor documentation error and also modified the netcp driver
   to fix the set* functions to include correct le/be macros.

v5->v6
 - updated version after incorporating comments [6] from David Miller,
   David Laight & Geert Uytterhoeven on v5. I would like get this in
   for v3.19 merge window if the latest version is acceptable.

v4->v5
 - Sorry to spin v5 quickly but I missed few check-patch warnings which
   were pointed by Joe Perches(thanks). I folded his changes [5] along with
   few more check-patch warning fixes. I would like get this in for v3.18
   merge window if David is happy with this version.

v3->v4
 - Couple of fixes in in error path as pointed [4] out by David. Rest of
   the patches are unchanged from v3.

v2->v3
 - Update v3 after incorporating Jamal and David Miller's comment/suggestion
   from earlier versions [1] [2].  After per the discussion here [3], the
   controversial custom exports have been dropped now. And for future
   future offload support additions, we will plug into generic frameworks
   as an when they are available.
====================

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

This patch enhances the NetCP gbe driver to support 10GbE subsystem
available in Keystone NetCP. The 3-port 10GbE switch sub-module contains
the following components:- 10GbE Switch, MDIO Module, 2 PCS-R Modules
(10GBase-R) and 2 SGMII modules (10/100/1000Base-T). The GBE driver
together with netcp core driver provides support for 10G Ethernet
on Keystone SoCs.

10GbE hardware spec is available at

http://www.ti.com/general/docs/lit/getliterature.tsp?baseLiteratureNumber=spruhj5&fileType=pdf



 Cc: David Miller <davem@davemloft.net>
 Cc: Rob Herring <robh+dt@kernel.org>
 Cc: Grant Likely <grant.likely@linaro.org>
 Cc: Santosh Shilimkar <santosh.shilimkar@kernel.org>
 Cc: Pawel Moll <pawel.moll@arm.com>
 Cc: Mark Rutland <mark.rutland@arm.com>
 Cc: Ian Campbell <ijc+devicetree@hellion.org.uk>
 Cc: Kumar Gala <galak@codeaurora.org>

Signed-off-by: Wingman Kwok <w-kwok2@ti.com>
Signed-off-by: Murali Karicheri <m-karicheri2@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch add support for 1G Ethernet driver based on Keystone
NetCP hardware. The gigabit Ethernet (GbE) switch subsystem is one of the main
components of the network coprocessor (NETCP) peripheral. The purpose of the
gigabit Ethernet switch subsystem in the NETCP is to provide an interface to
transfer data between the host device and another connected device in
compliance with the Ethernet protocol. GbE consists of 5 port Ethernet Switch
module, 4 Serial Gigabit Media Independent Interface (SGMII) modules, MDIO
module and SerDes.

Driver for 5 port GbE switch and SGMII module is added in this patch. These
hardware modules along with netcp core driver provides Network driver functions
for 1G Ethernet.

Detailed hardware spec is available at

http://www.ti.com/lit/ug/sprugv9d/sprugv9d.pdf



 Cc: David Miller <davem@davemloft.net>
 Cc: Rob Herring <robh+dt@kernel.org>
 Cc: Grant Likely <grant.likely@linaro.org>
 Cc: Santosh Shilimkar <santosh.shilimkar@kernel.org>
 Cc: Pawel Moll <pawel.moll@arm.com>
 Cc: Mark Rutland <mark.rutland@arm.com>
 Cc: Ian Campbell <ijc+devicetree@hellion.org.uk>
 Cc: Kumar Gala <galak@codeaurora.org>

Signed-off-by: Wingman Kwok <w-kwok2@ti.com>
Signed-off-by: Murali Karicheri <m-karicheri2@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The network coprocessor (NetCP) is a hardware accelerator available in
Keystone SoCs that processes Ethernet packets. NetCP consists of following
hardware components

 1 Gigabit Ethernet (GbE) subsystem with a Ethernet switch sub-module to
   send and receive packets.
 2 Packet Accelerator (PA) module to perform packet classification
   operations such as header matching, and packet modification operations
   such as checksum generation.
 3 Security Accelerator(SA) capable of performing IPSec operations on
   ingress/egress packets.
 4 An optional 10 Gigabit Ethernet Subsystem (XGbE) which includes a
   3-port Ethernet switch sub-module capable of 10Gb/s and 1Gb/s rates
   per Ethernet port.
 5 Packet DMA and Queue Management Subsystem (QMSS) to enqueue and dequeue
   packets and DMA the packets between memory and NetCP hardware components
   described above.

NetCP core driver make use of the Keystone Navigator driver API to allocate
DMA channel for the Ethenet device and to handle packet queue/de-queue,
Please refer API's in include/linux/soc/ti/knav_dma.h and
drivers/soc/ti/knav_qmss.h for details.

NetCP driver consists of NetCP core driver and at a minimum Gigabit
Ethernet (GBE) module (1) driver to implement the Network device function.
Other modules (2,3) can be optionally added to achieve supported hardware
acceleration function. The initial version of the driver include NetCP
core driver and GBE driver modules.

Please refer Documentation/devicetree/bindings/net/keystone-netcp.txt
for design of the driver.

 Cc: David Miller <davem@davemloft.net>
 Cc: Rob Herring <robh+dt@kernel.org>
 Cc: Grant Likely <grant.likely@linaro.org>
 Cc: Santosh Shilimkar <santosh.shilimkar@kernel.org>
 Cc: Pawel Moll <pawel.moll@arm.com>
 Cc: Mark Rutland <mark.rutland@arm.com>
 Cc: Ian Campbell <ijc+devicetree@hellion.org.uk>
 Cc: Kumar Gala <galak@codeaurora.org>

Signed-off-by: Murali Karicheri <m-karicheri2@ti.com>
Signed-off-by: Wingman Kwok <w-kwok2@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The network coprocessor (NetCP) is a hardware accelerator that processes
Ethernet packets. NetCP has a gigabit Ethernet (GbE) subsystem with a ethernet
switch sub-module to send and receive packets. NetCP also includes a packet
accelerator (PA) module to perform packet classification operations such as
header matching, and packet modification operations such as checksum
generation. NetCP can also optionally include a Security Accelerator(SA)
capable of performing IPSec operations on ingress/egress packets.

Keystone SoC's also have a 10 Gigabit Ethernet Subsystem (XGbE) which
includes a 3-port Ethernet switch sub-module capable of 10Gb/s and
1Gb/s rates per Ethernet port.

NetCP Subsystem device tree layout looks something like below:

-----------------------------
  NetCP subsystem(10G or 1G)
-----------------------------
	|
	|-> NetCP Devices ->	|
	|			|-> GBE/XGBE Switch
	|			|
	|			|-> Packet Accelerator
	|			|
	|			|-> Security Accelerator
	|
	|
	|
	|-> NetCP Interfaces ->	|
				|-> Ethernet Port 0
				|
				|-> Ethernet Port 1
				|
				|-> Ethernet Port 2
				|
				|-> Ethernet Port 3

Common driver supports GBE as well XGBE network processors.

Cc: Rob Herring <robh+dt@kernel.org>
Cc: Grant Likely <grant.likely@linaro.org>
Cc: Pawel Moll <pawel.moll@arm.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Ian Campbell <ijc+devicetree@hellion.org.uk>
Cc: Kumar Gala <galak@codeaurora.org>
Cc: David Miller <davem@davemloft.net>
Cc: Santosh Shilimkar <santosh.shilimkar@kernel.org>

Signed-off-by: Murali Karicheri <m-karicheri2@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Commit c03abd84 (net: ethernet: cpsw: don't requests IRQs we don't
use) left one build breakage when NET_POLL_CONTROLLER is enabled.

Fix this build break by referring to the correct irqs_table array.

Fixes: c03abd84

 (net: ethernet: cpsw: don't requests IRQs we don't use)
Reported-by: kbuild test robot <fengguang.wu@intel.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge branch 'link_netns'

Nicolas Dichtel says:

====================
netns: allow to identify peer netns

The goal of this serie is to be able to multicast netlink messages with an
attribute that identify a peer netns.
This is needed by the userland to interpret some information contained in
netlink messages (like IFLA_LINK value, but also some other attributes in case
of x-netns netdevice (see also
http://thread.gmane.org/gmane.linux.network/315933/focus=316064 and
http://thread.gmane.org/gmane.linux.kernel.containers/28301/focus=4239)

).

Ids of peer netns can be set by userland via a new rtnl cmd RTM_NEWNSID. When
the kernel needs an id for a peer (for example when advertising a new x-netns
interface via netlink), if the user didn't allocate an id, one will be
automatically allocated.
These ids are stored per netns and are local (ie only valid in the netns where
they are set). To avoid allocating an int for each peer netns, I use
idr_for_each() to retrieve the id of a peer netns. Note that it will be possible
to add a table (struct net -> id) later to optimize this lookup if needed.

Patch 1/4 introduces the rtnetlink API mechanism to set and get these ids.
Patch 2/4 and 3/4 implements an example of how to use these ids when advertising
information about a x-netns interface.
And patch 4/4 shows that the netlink messages can be symetric between a GET and
a SET.

iproute2 patches are available, I can send them on demand.

Here is a small screenshot to show how it can be used by userland.

$ ip netns add foo
$ ip netns del foo
$ ip netns
$ touch /var/run/netns/init_net
$ mount --bind /proc/1/ns/net /var/run/netns/init_net
$ ip netns add foo
$ ip -n foo netns
foo
init_net
$ ip -n foo netns set init_net 0
$ ip -n foo netns set foo 1

$ ip netns
foo
init_net
$ ip -n foo netns
foo (id: 1)
init_net (id: 0)

$ ip -n foo link add ipip1 link-netnsid 0 type ipip remote 10.16.0.121 local 10.16.0.249
$ ip -n foo link ls ipip1
6: ipip1@NONE: <POINTOPOINT,NOARP> mtu 1480 qdisc noop state DOWN mode DEFAULT group default
    link/ipip 10.16.0.249 peer 10.16.0.121 link-netnsid 0

$ ip netns
foo
init_net
$ ip -n foo link add ipip2 type ipip remote 10.16.0.121 local 10.16.0.249
$ ip -n foo link set ipip2 netns init_net
$ ip link ls ipip2
7: ipip2@NONE: <POINTOPOINT,NOARP> mtu 1480 qdisc noop state DOWN mode DEFAULT group default
    link/ipip 10.16.0.249 peer 10.16.0.121 link-netnsid 0
$ ip netns
foo (id: 0)
init_net

v4 -> v5:
  use rtnetlink instead of genetlink
  allocate automatically an id if user didn't assign one
  rename include/uapi/linux/netns.h to include/uapi/linux/net_namespace.h
  add vxlan in patch #3

RFCv3 -> v4:
  rebase on net-next
  add copyright text in the new netns.h file

RFCv2 -> RFCv3:
  ids are now defined by userland (via netlink). Ids are stored in each netns
  (and they are local to this netns).
  add get_link_net support for ip6 tunnels
  netnsid is now a s32 instead of a u32

RFCv1 -> RFCv2:
  remove useless ()
  ids are now stored in the user ns. It's possible to get an id for a peer netns
  only if the current netns and the peer netns have the same user ns parent.
====================

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

This patch adds the ability to create a netdevice in a specified netns and
then move it into the final netns. In fact, it allows to have a symetry between
get and set rtnl messages.

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

Implement rtnl_link_ops->get_link_net() callback so that IFLA_LINK_NETNSID is
added to rtnetlink messages.

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

This patch adds a new attribute (IFLA_LINK_NETNSID) which contains the 'link'
netns id when this netns is different from the netns where the interface
stands (for example for x-net interfaces like ip tunnels).
With this attribute, it's possible to interpret correctly all advertised
information (like IFLA_LINK, etc.).

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

With this patch, a user can define an id for a peer netns by providing a FD or a
PID. These ids are local to the netns where it is added (ie valid only into this
netns).

The main function (ie the one exported to other module), peernet2id(), allows to
get the id of a peer netns. If no id has been assigned by the user, this
function allocates one.

These ids will be used in netlink messages to point to a peer netns, for example
in case of a x-netns interface.

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

While suspending, we destroy the authentication /
association that might be taking place. While doing so, we
forgot to delete the timer which can be firing after
local->suspended is already set, producing the warning below.

Fix that by deleting the timer.

[66722.825487] WARNING: CPU: 2 PID: 5612 at net/mac80211/util.c:755 ieee80211_can_queue_work.isra.18+0x32/0x40 [mac80211]()
[66722.825487] queueing ieee80211 work while going to suspend
[66722.825529] CPU: 2 PID: 5612 Comm: kworker/u16:69 Tainted: G        W  O  3.16.1+ #24
[66722.825537] Workqueue: events_unbound async_run_entry_fn
[66722.825545] Call Trace:
[66722.825552]  <IRQ>  [<ffffffff817edbb2>] dump_stack+0x4d/0x66
[66722.825556]  [<ffffffff81075cad>] warn_slowpath_common+0x7d/0xa0
[66722.825572]  [<ffffffffa06b5b90>] ? ieee80211_sta_bcn_mon_timer+0x50/0x50 [mac80211]
[66722.825573]  [<ffffffff81075d1c>] warn_slowpath_fmt+0x4c/0x50
[66722.825586]  [<ffffffffa06977a2>] ieee80211_can_queue_work.isra.18+0x32/0x40 [mac80211]
[66722.825598]  [<ffffffffa06977d5>] ieee80211_queue_work+0x25/0x50 [mac80211]
[66722.825611]  [<ffffffffa06b5bac>] ieee80211_sta_timer+0x1c/0x20 [mac80211]
[66722.825614]  [<ffffffff8108655a>] call_timer_fn+0x8a/0x300

Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

This reverts commit ba1debdf

.

Oliver reported that it breaks network-manager, for some reason with
this patch NM decides that the device isn't wireless but "generic"
(ethernet), sees no carrier (as expected with wifi) and fails to do
anything else with it.

Revert this to unbreak userspace.

Reported-by: Oliver Hartkopp <socketcan@hartkopp.net>
Tested-by: Oliver Hartkopp <socketcan@hartkopp.net>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

Commit 02dba438

 ("bridge: fix setlink/dellink notifications") removed usage of oflags in
both rtnl_bridge_setlink() and rtnl_bridge_dellink() methods. This patch removes this variable as it is no
longer needed.

Signed-off-by: Rami Rosen <rami.rosen@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Commit 053c095a

 ("netlink: make nlmsg_end() and genlmsg_end()
void") didn't catch all of the cases where callers were breaking out
on the return value being equal to zero, which they no longer should
when zero means success.

Fix all such cases.

Reported-by: Marcel Holtmann <marcel@holtmann.org>
Reported-by: Scott Feldman <sfeldma@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: David S. Miller <davem@davemloft.net>

In the ipip tunnel, the skb->queue_mapping is lost in ipip_rcv().
All skb will be queued to the same cell->napi_skbs.  The
gro_cell_poll is pinned to one core under load.  In production traffic,
we also see severe rx_dropped in the tunl iface and it is probably due to
this limit: skb_queue_len(&cell->napi_skbs) > netdev_max_backlog.

This patch is trying to alloc_percpu(struct gro_cell) and schedule
gro_cell_poll to process the skb in the same core.

Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add packet and byte counters for RX and TX paths.

$ ifconfig eth1
eth1: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
        inet6 fe80::5054:ff:fe12:3501  prefixlen 64  scopeid 0x20<link>
        ether 52:54:00:12:35:01  txqueuelen 1000  (Ethernet)
        RX packets 63  bytes 15813 (15.4 KiB)
        RX errors 1  dropped 0  overruns 0  frame 0
        TX packets 79  bytes 17991 (17.5 KiB)
        TX errors 7  dropped 0 overruns 0  carrier 0  collisions 0

Rx / Tx errors tested by injecting faults in qemu's hardware model for Rocker.

v2:
- moved counter locations to avoid potential use after free per Florian's comment

Signed-off-by: David Ahern <dsahern@gmail.com>
Cc: Scott Feldman <sfeldma@gmail.com>
Cc: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: Scott Feldman <sfeldma@gmail.com>
Acked-by: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: David S. Miller <davem@davemloft.net>

CPSW never uses RX_THRESHOLD or MISC interrupts. In
fact, they are always kept masked in their appropriate
IRQ Enable register.

Instead of allocating an IRQ that never fires, it's best
to remove that code altogether and let future patches
implement it if anybody needs those.

Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch is in preparation for a nicer IRQ
handling scheme where we use different IRQ
handlers for each IRQ line (as it should be).

Later, we will also drop IRQs offset 0 and 3
because they are always disabled in this driver.

Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>