zd1211rw: slight optimization of addr compare

Use possibly more efficient ether_addr_equal
instead of memcmp.

Cc: Daniel Drake <dsd@gentoo.org>
Cc: Ulrich Kunitz <kune@deine-taler.de>
Cc: "John W. Linville" <linville@tuxdriver.com>
Cc: linux-wireless@vger.kernel.org
Cc: netdev@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Tan Xiaojun <tanxiaojun@huawei.com>
Signed-off-by: Ding Tianhong <dingtianhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

batman-adv: use batadv_compare_eth for concise

It is better to use batadv_compate_eth instead of memcpy for
concise style.

Cc: Marek Lindner <mareklindner@neomailbox.ch>
Cc: Simon Wunderlich <sw@simonwunderlich.de>
Cc: Antonio Quartulli <antonio@meshcoding.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: b.a.t.m.a.n@lists.open-mesh.org
Cc: netdev@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Tan Xiaojun <tanxiaojun@huawei.com>
Signed-off-by: Ding Tianhong <dingtianhong@huawei.com>
Acked-by: Antonio Quartulli <antonio@meshcoding.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

ppp: slight optimization of addr compare

Use possibly more efficient ether_addr_equal
to instead of memcmp.

Cc: Michal Ostrowski <mostrows@earthlink.net>
Cc: netdev@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Ding Tianhong <dingtianhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

hamradio: slight optimization of addr compare

Use possibly more efficient ether_addr_equal
to instead of memcmp.

Cc: Joerg Reuter <jreuter@yaina.de>
Cc: linux-hams@vger.kernel.org
Cc: netdev@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Ding Tianhong <dingtianhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: atl1: slight optimization of addr compare

Use the recently added and possibly more efficient
ether_addr_equal_unaligned to instead of memcmp.

Cc: Jay Cliburn <jcliburn@gmail.com>
Cc: Chris Snook <chris.snook@gmail.com>
Cc: netdev@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Weilong Chen <chenweilong@huawei.com>
Signed-off-by: Ding Tianhong <dingtianhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: pcnet: slight optimization of addr compare

Use possibly more efficient ether_addr_equal
to instead of memcmp.

Cc: Don Fry <pcnet32@frontier.com>
Cc: netdev@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Ding Tianhong <dingtianhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: freescale: remove unused compare_addr()

The function did not be used any more, so remove it.

Cc: Li Yang <leoli@freescale.com>
Cc: netdev@vger.kernel.org
Cc: linuxppc-dev@lists.ozlabs.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Ding Tianhong <dingtianhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

qlge: make local function static

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

mdio: unused ethtool functions

Use it or lose it.

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

hhf: make qdisc ops static

This module shouldn't be randomly exporting symbols

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

rose: cleanup checkpatch errors,spaces required

This patch add spaces to cleanup checkpatch errors.

Signed-off-by: Weilong Chen <chenweilong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

sctp: remove the never used 'return' and redundant 'break'

In switch() had do return, and never use the 'return NULL'. The
'break' after return or goto has no effect. Remove it.

v2: make it more readable as suggested by Neil.

Signed-off-by: Wang Weidong <wangweidong1@huawei.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

mac802154: fix following checkpath.pl warning Prefer pr_warn(... to pr_warning(...

This patch fixes checkpath.pl:
 WARNING: Prefer pr_warn(... to pr_warning(...
 #447: FILE: ./wpan.c:447:

Signed-off-by: Weilong Chen <chenweilong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

cxgb4: make functions static and remove dead code

Cleanup by making local functions static.

The code to load config file is unreachable in net-next, probably came
from some out of tree driver.

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

sctp: remove casting from function calls through ops structure

remove the unnecessary cast.

Signed-off-by: Wang Weidong <wangweidong1@huawei.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

netconf: add support for IPv6 proxy_ndp

Need to be able to see changes to proxy NDP status on a per
interface basis via netlink (analog to proxy_arp).

Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

netconf: rename PROXY_ARP to NEIGH_PROXY

Use same field for both IPv4 (proxy_arp) and IPv6 (proxy_ndp)
so fix it before API is set to be a common name

Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: vxlan: use custom ndo_change_mtu handler

When adding a new vxlan device to an "underlying carrier" (here:
dst->remote_ifindex), the MTU size assigned to the vxlan device
is the MTU at setup time of the carrier - needed headroom, when
adding a vxlan device w/o explicit carrier, then it defaults
to 1500.

In case of an explicit carrier that supports jumbo frames, we
currently cannot change vxlan MTU via ip(8) to > 1500 in
post-setup time, as vxlan driver uses eth_change_mtu() as default
method for manually setting MTU.

Hence, use a custom implementation that only falls back to
eth_change_mtu() in case we didn't use a dev parameter on device
setup time, and otherwise allow a max MTU setting of the carrier
incl. adjustment for headroom.

Reported-by: Shahed Shaikh <shahed.shaikh@qlogic.com>
Signed-off-by: Daniel Borkmann <dborkman@redhat.com>
Acked-by: Stephen Hemminger <stephen@networkplumber.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: use kfree_skb_list() helper

We can use kfree_skb_list() instead of open coding it.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge branch 'for-davem' of git://gitorious.org/linux-can/linux-can-next

Marc Kleine-Budde says:

====================
this is a pull request of three patches for net-next/master.

There is a patch by Oliver Hartkopp, to clean up the CAN gw code.
Alexander Shiyan adds device tree support to the mcp251x driver and a
patch by Ezequiel Garcia lets the ti_hecc driver compile on all ARM
platforms.
====================

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

ipv6: move ip6_sk_accept_pmtu from generic pmtu update path to ipv6 one

In commit 93b36cf3425b9b ("ipv6: support IPV6_PMTU_INTERFACE on sockets")
I made a horrible mistake to add ip6_sk_accept_pmtu to the generic
sctp_icmp_frag_needed path. This results in build warnings if IPv6 is
disabled which were luckily caught by Fengguang's kbuild bot. But it
also leads to a kernel panic IPv4 frag-needed packet is received.

Reported-by: kbuild test robot <fengguang.wu@intel.com>
Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Acked-by: Daniel Borkmann <dborkman@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Revert "sctp: fix missing include file"

This reverts commit ac0917f250c61ae1e395844ebd924db0e07d1f87.

Better version of this fix forthcoming.

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

can: ti_hecc: Replace platform dependency with ARM dependency

OMAP's ti_hecc driver is used to support the CAN controller on many omap2plus
SoCs (OMAP2430, OMAP3, OMAP4, OMAP5 and AM335x), so it's wrong to make this
depend on OMAP3 only. Take an extra step, to get wider build coverage, and make
the driver depend on ARM.

Signed-off-by: Ezequiel Garcia <ezequiel.garcia@free-electrons.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: mcp251x: Add device tree support

This patch adds Device Tree support to the Microchip MCP251X driver.

Signed-off-by: Alexander Shiyan <shc_work@mail.ru>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: gw: remove obsolete checks

In commit be286bafe1f4069094865264f29805854c5788bf ("can: gw: add a variable
limit for CAN frame routings") the detection of the frame routing has been
changed. The former solution required dev->header_ops to be unused (== NULL).

I missed to remove the obsolete checks in the original commit - so here it is.

Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

packet: doc: add documentation for VLAN TPID delivery

Introduce TP_STATUS_VLAN_TPID_VALID bit into the documentation.

Signed-off-by: Atzm Watanabe <atzm@stratosphere.co.jp>
Acked-by: Daniel Borkmann <dborkman@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

sctp: fix missing include file

Compile error reported by Jim Davis on netdev.

ip6_sk_accept_pmtu() needs net/ip6_route.h

Signed-off-by: Scott Feldman <sfeldma@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge branch 'b44'

Hauke Mehrtens says:

====================
b44: add support for external PHY

This adds support for an external phy connected to the mac controlled
by b44. This is used on home routers of the BCM47xx line where this MAC
core was used and is contended to an external switch core through MII.
These patches are in OpenWrt for some time and are tested by different
users with different devices.

The publicly available Programmer's Reference Guide for some of these
chips help making this patch it would be nice to have such a
documentation for more Broadcom Chips like some SoCs.
https://www.broadcom.com/collateral/pg/440X-PG02-R.pdf
====================

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

b44: use fixed PHY device if we do not find any

The ADM6996L switch and some Broadcom switches with two MII interfaces
like the BCM5325F connected to two MACs on the SoC, used on some
routers do not return a valid value when reading the PHY id register
and Linux thinks there is no PHY at all, but that is wrong.
This patch registers a fixed phy in the arch code and then searches it
when there is no other phy in the Ethernet driver code.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

b44: do not set PHY address to 30 for every ext PHY

b44_phy_reset() will fail for an external PHY and only work with the
internal PHY, this was an old workaround when the detection of an
external switch based on the PHY address failed and it is not needed
any more.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

b44: activate PHY when MAC is off

Without this patch we can not access the PHY when the MAC is switched
off. This PHY access is needed to configure the switch, which is done
through PHY registers.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

b44: add phylib support

Most of the older home routers based on the Broadcom BCM47XX SoC series
are using a MAC that is supported by b44. On most of these routers not
the internal PHY of this MAC core is used, but a switch sometimes on an
external chip or integrated into the same SoC as the Ethernet core.
For this switch a special PHY driver is needed which should not be
integrated into b44 as the same switches are also used by other
Broadcom home networking SoCs which are using different Ethernet MAC
drivers. This was tested with the b53 switch driver which is currently
on its way to mainline.

If the internal PHY is not used, b44 will now search on the MDIO bus
for a phy and use the Linux phylib subsystem to register a driver.
Support for the internal PHY must stay here, because there are some
device which are suing the internal phy.

With this patch we scan the mdio bus when the sprom or nvram says that
the PHY address is 30, if a PHY was found at this address b44 uses it.

This was tested with a BCM4704, BCM4712 and BCM5354.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

b44: rename b44_mii_{read, write} to b44_mdio_{read, write}_mii

The next patch will add these functions for phylib, and we should
rename the old ones before. This now indicates that these functions are
used for the mdio registers and on the mii interface.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

b44: abort when no PHY is available at all

When the phy address is 31, this means that there is no PHY connected
to this MAC at all, no internal and no external PHY. Reading these PHY
registers causes a system reset on some routers.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

b44: rename B44_PHY_ADDR_NO_PHY to B44_PHY_ADDR_NO_LOCAL_PHY

The PHY address 30 means there is no local PHY, but there could be an
external PHY like a switch connected via MII. This is the case on most
embedded home routers where this driver is used.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

b44: check register instead of PHY address to detect external PHY

The Ethernet core supported by b44 supports an internal PHY integrated
into the mac core, which is supported by the b44 driver and an external
PHY to which the mac core is connected. This external PHY could be a
switch connected through MII, which is often the case when this core is
used on home routers. The usage of an external PHY was assumed when the
PHY address 30 was used and an internal PHY was assumed when the PHY
address was different. To verify that b44_phy_reset() was called and
checked if it worked, otherwise PHY address 30 was assumed, an external
PHY. It is better to check the register which says which PHY is
connected to the MAC instead of checking the PHY address.
The interface to an external PHY was only activated when this register
was set.

This also changes B44_FLAG_INTERNAL_PHY to B44_FLAG_EXTERNAL_PHY, it is
easier to check.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

tcp: autocork should not hold first packet in write queue

Willem noticed a TCP_RR regression caused by TCP autocorking
on a Mellanox test bed. MLX4_EN_TX_COAL_TIME is 16 us, which can be
right above RTT between hosts.

We can receive a ACK for a packet still in NIC TX ring buffer or in a
softnet completion queue.

Fix this by always pushing the skb if it is at the head of write queue.

Also, as TX completion is lockless, it's safer to perform sk_wmem_alloc
test after setting TSQ_THROTTLED.

erd:~# MIB="MIN_LATENCY,MEAN_LATENCY,MAX_LATENCY,P99_LATENCY,STDDEV_LATENCY"
erd:~#  ./netperf -H remote -t TCP_RR -- -o $MIB | tail -n 1
(repeat 3 times)

Before patch :

18,1049.87,41004,39631,6295.47
17,239.52,40804,48,2912.79
18,348.40,40877,54,3573.39

After patch :

18,22.84,4606,38,16.39
17,21.56,2871,36,13.51
17,22.46,2705,37,11.83

Reported-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Fixes: f54b311142a9 ("tcp: auto corking")
Signed-off-by: David S. Miller <davem@davemloft.net>

net_sched: fix regression in tc_action_ops

list_for_each_entry(a, &act_base, head) doesn't
exit with a = NULL if we reached the end of the list.

tcf_unregister_action(), tc_lookup_action_n() and tc_lookup_action()
need fixes.

Remove tc_lookup_action_id() as its unused and not worth 'fixing'

Signed-off-by: Eric Dumazet <edumazet@google.com>
Fixes: 1f747c26c48b ("net_sched: convert tc_action_ops to use struct list_head")
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Reviewed-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net_sched: fix a regression in tcf_proto_lookup_ops()

list_for_each_entry(t, &tcf_proto_base, head) doesn't
exit with t = NULL if we reached the end of the list.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Fixes: 3627287463b4 ("net_sched: convert tcf_proto_ops to use struct
list_head")
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Reviewed-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net_sched: fix a regression in tc actions

This patch fixes:
1) pass mask rather than size to tcf_hashinfo_init()
2) the cleanup should be in reversed order in mirred_cleanup_module()

Reported-by: Eric Dumazet <eric.dumazet@gmail.com>
Fixes: 369ba56787d7469c0afd ("net_sched: init struct tcf_hashinfo at register time")
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: David S. Miller <davem@davemloft.net>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge branch 'bridge_cleanups'

Tan Xiaojun says:

====================
bridge: cleanup and fix checkpatch errors

Clean up and fix some checkpatch errors in bridge driver.
====================

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

bridge: change the position of '{' to the pre line

That open brace { should be on the previous line.

Signed-off-by: Tan Xiaojun <tanxiaojun@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bridge: change "foo* bar" to "foo *bar"

"foo * bar" should be "foo *bar".

Signed-off-by: Tan Xiaojun <tanxiaojun@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bridge: add space before '(/{', after ',', etc.

Spaces required before the open parenthesis '(', before the open
brace '{', after that ',' and around that '?/:'.

Signed-off-by: Tan Xiaojun <tanxiaojun@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bridge: remove unnecessary parentheses

Return is not a function, parentheses are not required.

Signed-off-by: Tan Xiaojun <tanxiaojun@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bridge: remove unnecessary condition judgment

Because err is always negative, remove unnecessary condition
judgment.

Signed-off-by: Tan Xiaojun <tanxiaojun@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bnx2x: cleanup skb_set_hash

Save pkt_hash_types in tpa structure.

Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge branch 'sh_eth'

Sergei Shtylyov says:

====================
Add PHY IRQ support to the 'sh_eth' driver

   This patchset adds support for passing PHY's IRQ to the 'sh_eth' driver from
the platform code and removes now useless code that hinders the PHY IRQ handing
by philib.
====================

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

sh_eth: do not reset PHY needlessly

There's no need anymore to call phy_init_hw() to reset/resume the PHY from the
driver, as the call chain in phylib already has reached it, and so reset/resumed
the PHY (even resuming it twice). This duplicate reset is not only  needless, it
e.g. clears the PHY's interrupt enables just setup by phylib and so prevents the
expected IRQs from the PHY.

Signed-off-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

sh_eth: add PHY IRQ to platform data

Allow the platform code to pass PHY's IRQ to the driver.  Print this IRQ along
with the other PHY datails in sh_eth_phy_init().

Signed-off-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge branch 'mlx4'

Amir Vadai says:

====================
net/mlx4: Mellanox driver update 08-12-2013

This patchset contains:
1. Support in ndo_get_phys_port_id added by Hadar.
2. Change the driver to use by default CQE/EQE size of 64 bytes done by Eyal.
   This doubles the packet-rate  of the NIC.
3. Configure the XPS queue mapping on driver load - added by Ido.
4. Fixes for some small bugs done by Jenny and Matan

Patchset was applied and tested against commit: "2372175 lib: hash: follow-up
fixups for arch hash"

Changes from V1:
- Removed Patch 10 "net/mlx4_en: Fix Supported/Advertised link mode reported by
  ethtool". This patch is needed to be rewritten from scratch and I wouldn't
  like it to block the rest of the patches in this set.
  Also, fix to Kconfig suggested by Ben will be sent in next patchset.

Changes from V0:
- Found some issues in "Reuse memory in RX flow" patch from V0.
  Removing this patch from the patchset till analyzed and fixed.
- Fix some coding style issues in patch 6 "Configure the XPS queue mapping on
  driver load"
- Changed patch 9 "Add NAPI support for transmit side" to use NAPI_POLL_WEIGHT
  instead of MLX4_EN_TX_BUDGET
====================

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

net/mlx4_core: Check port number for validity before accessing data

Need to validate port number at mlx4_promisc_qp() before use.
Since port number is extracted from gid, as a cooked or corrupted gid
could lead to a crash.

Signed-off-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Amir Vadai <amirv@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net/mlx4_en: Add NAPI support for transmit side

Add NAPI for TX side,
implement its support and provide NAPI callback.

Signed-off-by: Yevgeny Petrilin <yevgenyp@mellanox.com>
Signed-off-by: Eugenia Emantayev <eugenia@mellanox.com>
Signed-off-by: Amir Vadai <amirv@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net/mlx4_en: Ignore irrelevant hypervisor events

MLX4_DEV_EVENT_SLAVE_INIT and MLX4_DEV_EVENT_SLAVE_SHUTDOWN
events used by Hypervisor to inform the PPF IB driver that
IB para-virtualization must be initialized/destroyed for a slave.
If this event is catched by ETH VF annoying but harmless error message
is printed into dmesg. Remove dmesg prints for these events.

Signed-off-by: Eugenia Emantayev <eugenia@mellanox.co.il>
Signed-off-by: Amir Vadai <amirv@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net/mlx4_core: Set CQE/EQE size to 64B by default

To achieve out of the box performance default is to use 64 byte CQE/EQE.
In tests that we conduct in our labs, we achieved a performance
improvement of twice the message rate. For older VF/libmlx4 support,
enable_64b_cqe_eqe must be set to 0 (disabled).

Signed-off-by: Eyal Perry <eyalpe@mellanox.com>
Signed-off-by: Amir Vadai <amirv@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net/mlx4_en: Configure the XPS queue mapping on driver load

Only TX rings of User Piority 0 are mapped.
TX rings of other UP's are using UP 0 mapping.
XPS is not in use when num_tc is set.

Signed-off-by: Ido Shamay <idos@mellanox.com>
Signed-off-by: Amir Vadai <amirv@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net/mlx4_en: Implement ndo_get_phys_port_id

Use the port GUID read from the firmware to identify the physical port.
This port identifier is available via ndo_get_phys_port_id for both PF
and VF net-devices.

Signed-off-by: Hadar Hen Zion <hadarh@mellanox.com>
Signed-off-by: Amir Vadai <amirv@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net/mlx4_core: Expose physical port id as PF/VF capability

Add the infrastructure needed to support ndo_get_phys_port_id which
allows users to identify to which physical port a net-device is connected
to by reading a unique port id.
This will work for VFs and PFs.
The driver uses a new device capability - phys_port_id, The PF driver
reads the port phys_port_id from Firmware and stores it. The VF driver
reads the port phys_port_id from the PF using QUERY_FUNC_CAP command.

Signed-off-by: Hadar Hen Zion <hadarh@mellanox.com>
Signed-off-by: Amir Vadai <amirv@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net/mlx4_core: Introduce nic_info new flag in QUERY_FUNC_CAP

Set nic_info field in QUERY_FUNC_CAP, which designates
supplementary NIC information is provided by the hypervisor.
When set, the following fields are valid: nic_num_rings,
nic_indirection_tbl_sz, cur_mac and phys_port_id.

Signed-off-by: Hadar Hen Zion <hadarh@mellanox.com>
Signed-off-by: Amir Vadai <amirv@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net/mlx4_core: Rename QUERY_FUNC_CAP fields

Use correct names for QUERY_FUNC_CAP fields: flags0 and flags1.

Signed-off-by: Hadar Hen Zion <hadarh@mellanox.com>
Signed-off-by: Amir Vadai <amirv@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net/mlx4_core: Remove zeroed out of explicit QUERY_FUNC_CAP fields

All mailboxes are already zeroed by commit:
571b8b9 net/mlx4_core: Initialize all mailbox buffers to zero before use
Remove explicit zero set for force mac and force vlan fields in
mlx4_QUERY_FUNC_CAP_wrapper

Signed-off-by: Hadar Hen Zion <hadarh@mellanox.com>
Signed-off-by: Amir Vadai <amirv@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

rps: NUMA flow limit allocations

Given we allocate memory for each cpu, we can do this
using NUMA affinities, instead of using NUMA policies
of the process changing flow_limit_cpu_bitmap value.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers: net: Mark functions as static in debug.c

This patch marks the function il_clear_traffic_stats() in debug.c as
static because they are not used outside this file.

Thus, it also removes the following warnings in
wireless/iwlegacy/debug.c:

drivers/net/wireless/iwlegacy/debug.c:35:1: warning: no previous prototype for ‘il_clear_traffic_stats’ [-Wmissing-prototypes]

Signed-off-by: Rashika Kheria <rashika.kheria@gmail.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers: net: Mark functions as static in 4965-debug.c

This patch marks the function il4965_ucode_rx_stats_read(),
il4965_ucode_tx_stats_read() and il4965_ucode_general_stats_read() in
4965-debug.c as static because they are not used outside this file.

Thus, it also removes the following warnings in
wireless/iwlegacy/4965-debug.c:

drivers/net/wireless/iwlegacy/4965-debug.c:59:1: warning: no previous prototype for ‘il4965_ucode_rx_stats_read’ [-Wmissing-prototypes]
drivers/net/wireless/iwlegacy/4965-debug.c:471:1: warning: no previous prototype for ‘il4965_ucode_tx_stats_read’ [-Wmissing-prototypes]
drivers/net/wireless/iwlegacy/4965-debug.c:637:1: warning: no previous prototype for ‘il4965_ucode_general_stats_read’ [-Wmissing-prototypes]

Signed-off-by: Rashika Kheria <rashika.kheria@gmail.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers: net: Mark functions as static in 3945-debug.c

This patch marks the function il3945_ucode_rx_stats_read(),
il3945_ucode_tx_stats_read() and il3945_ucode_general_stats_read() in
3945-debug.c as static because they are not used outside this file.

Thus, it also removes the following warnings in
wireless/iwlegacy/3945-debug.c:

drivers/net/wireless/iwlegacy/3945-debug.c:52:1: warning: no previous prototype for ‘il3945_ucode_rx_stats_read’ [-Wmissing-prototypes]
drivers/net/wireless/iwlegacy/3945-debug.c:317:1: warning: no previous prototype for ‘il3945_ucode_tx_stats_read’ [-Wmissing-prototypes]
drivers/net/wireless/iwlegacy/3945-debug.c:407:1: warning: no previous prototype for ‘il3945_ucode_general_stats_read’ [-Wmissing-prototypes]

Signed-off-by: Rashika Kheria <rashika.kheria@gmail.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers: net: Include new header file in sbni.c

Create a new header file include/net/Space.h which contains
prototype declaration of sbni_probe().

Include the new header file in drivers/net/Space.c and
drivers/net/wan/sbni.c because they use this function.

This eliminates the following warning in wan/sbni.c:
drivers/net/wan/sbni.c:224:12: warning: no previous prototype for ‘sbni_probe’ [-Wmissing-prototypes]

Signed-off-by: Rashika Kheria <rashika.kheria@gmail.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers: net: Mark functions as static in stmmac_platform.c

This patch marks the function stmmac_pltfr_freeze() and
stmmac_pltfr_restore() in stmmac_platform.c as static because they are
not used outside this file.

Thus, it also removes the following warnings in
ethernet/stmicro/stmmac/stmmac_platform.c:

drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c:222:5: warning: no previous prototype for ‘stmmac_pltfr_freeze’ [-Wmissing-prototypes]
drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c:236:5: warning: no previous prototype for ‘stmmac_pltfr_restore’ [-Wmissing-prototypes]

Signed-off-by: Rashika Kheria <rashika.kheria@gmail.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers: net: Mark functions as static in qlcnic_83xx_hw.c

This patch marks the function qlcnic_83xx_clear_legacy_intr_mask(),
qlcnic_83xx_set_legacy_intr_mask() and
qlcnic_83xx_enable_legacy_msix_mbx_intr() in qlcnic_83xx_hw.c as static
because they are not used outside this file.

Thus, it also removes the following warnings
in ethernet/qlogic/qlcnic/qlcnic_83xx_hw.c:

drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.c:318:13: warning: no previous prototype for ‘qlcnic_83xx_clear_legacy_intr_mask’ [-Wmissing-prototypes]
drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.c:323:13: warning: no previous prototype for ‘qlcnic_83xx_set_legacy_intr_mask’ [-Wmissing-prototypes]
drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.c:343:13: warning: no previous prototype for ‘qlcnic_83xx_enable_legacy_msix_mbx_intr’ [-Wmissing-prototypes]

Signed-off-by: Rashika Kheria <rashika.kheria@gmail.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers: net: Mark functions as static in qlcnic_io.c

This patch marks the function qlcnic_enable_tx_intr(),
qlcnic_83xx_enable_tx_intr() and qlcnic_83xx_disable_tx_intr() in
qlcnic_io.c as static because they are not used outside this file.

Thus, it also removes the following warnings in
ethernet/qlogic/qlcnic/qlcnic_io.c:

drivers/net/ethernet/qlogic/qlcnic/qlcnic_io.c:130:13: warning: no previous prototype for ‘qlcnic_enable_tx_intr’ [-Wmissing-prototypes]
drivers/net/ethernet/qlogic/qlcnic/qlcnic_io.c:147:13: warning: no previous prototype for ‘qlcnic_83xx_enable_tx_intr’ [-Wmissing-prototypes]
drivers/net/ethernet/qlogic/qlcnic/qlcnic_io.c:153:13: warning: no previous prototype for ‘qlcnic_83xx_disable_tx_intr’ [-Wmissing-prototypes]

Signed-off-by: Rashika Kheria <rashika.kheria@gmail.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge branch 'master' of git://git./linux/kernel/git/klassert/ipsec-next

Steffen Klassert says:

====================
pull request (net-next): ipsec-next 2013-12-19

1) Use the user supplied policy index instead of a generated one
   if present. From Fan Du.

2) Make xfrm migration namespace aware. From Fan Du.

3) Make the xfrm state and policy locks namespace aware. From Fan Du.

4) Remove ancient sleeping when the SA is in acquire state,
   we now queue packets to the policy instead. This replaces the
   sleeping code.

5) Remove FLOWI_FLAG_CAN_SLEEP. This was used to notify xfrm about the
   posibility to sleep. The sleeping code is gone, so remove it.

6) Check user specified spi for IPComp. Thr spi for IPcomp is only
   16 bit wide, so check for a valid value. From Fan Du.

7) Export verify_userspi_info to check for valid user supplied spi ranges
   with pfkey and netlink. From Fan Du.

8) RFC3173 states that if the total size of a compressed payload and the IPComp
   header is not smaller than the size of the original payload, the IP datagram
   must be sent in the original non-compressed form. These packets are dropped
   by the inbound policy check because they are not transformed. Document the need
   to set 'level use' for IPcomp to receive such packets anyway. From Fan Du.

Please pull or let me know if there are problems.
====================

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

Merge branch 'bonding_nl'

Scott Feldman says:

====================
bonding: add even more netlink attributes

The following series implements five more bonding netlink attributes:

num_grat_arp|num_unsol_na
all_slaves_active
min_links
lp_interval
packets_per_slave

Tested with modified iproute2 to verify attributes can be set at bond creation
time or set later.  Verified sysfs interface to attributes continues to work.

I have one more set after this one for all of the 802.3ad attributes.  Whew.
====================

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

bonding: add packets_per_slave attribute netlink support

Add IFLA_BOND_PACKETS_PER_SLAVE to allow get/set of bonding parameter
packets_per_slave via netlink.

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

bonding: add lp_interval attribute netlink support

Add IFLA_BOND_LP_INTERVAL to allow get/set of bonding parameter
lp_interval via netlink.

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

bonding: add min_links attribute netlink support

Add IFLA_BOND_MIN_LINKS to allow get/set of bonding parameter
min_links via netlink.

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

bonding: add all_slaves_active attribute netlink support

Add IFLA_BOND_ALL_SLAVES_ACTIVE to allow get/set of bonding parameter
all_slaves_active via netlink.

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

bonding: add num_grat_arp attribute netlink support

Add IFLA_BOND_NUM_PEER_NOTIF to allow get/set of bonding parameter
num_grat_arp via netlink.  Bonding parameter num_unsol_na is
synonymous with num_grat_arp, so add only one netlink attribute
to represent both bonding parameters.

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

Merge branch 'for-davem' of git://gitorious.org/linux-can/linux-can-next

Marc Kleine-Budde says:

====================
this is a pull request of four patches for net-next/master.

There is one patch by Markus Pargmann, which speeds up the c_can
driver, a patch by John Whitmore which updates the in tree
documentation. A patch by Jeff Kirsher which replaces the FSF's address
by a link and a patch by Alexander Shiyan which converts the mcp251x
driver to make use of managed resources.
====================

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

xen-netback: add gso_segs calculation

netback already has code which parses IPv4 and v6 headers to set up checksum
offsets and these are always applied to GSO packets being sent from
frontends. It's therefore suboptimal that GSOs are being marked
SKB_GSO_DODGY to defer the gso_segs calculation when netback already has all
necessary information to hand to do the calculation. This patch adds that
calculation.

Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
Cc: Wei Liu <wei.liu2@citrix.com>
Cc: Ian Campbell <ian.campbell@citrix.com>
Cc: David Vrabel <david.vrabel@citrix.com>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Acked-by: Wei Liu <wei.liu2@citrix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

sch_cbq: remove unnecessary null pointer check

It already has a NULL pointer check of rtab in qdisc_put_rtab().
Remove the check outside of qdisc_put_rtab().

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

act_police: remove unnecessary null pointer check

It already has a NULL pointer check of rtab in qdisc_put_rtab().
Remove the check outside of qdisc_put_rtab().

Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

sfc: fix sparse non static symbol warning

Fixes the following sparse warning:

drivers/net/ethernet/sfc/falcon.c:2601:6: warning:
 symbol 'falcon_pull_nic_stats' was not declared. Should it be static?

Signed-off-by: Wei Yongjun <yongjun_wei@trendmicro.com.cn>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: skbuff: improve comment on checksumming

It can be a bit confusing when looking for checksumming flags that
the actual comment for this resides elsewhere further below in the
header file.

Thus, bring the documentation where we define these flags, and
slightly improve the doc text to make it a bit more clear/readable.

Also, whitespace-align values of the define while at it.

Signed-off-by: Daniel Borkmann <dborkman@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net-qdisc-hhf: Heavy-Hitter Filter (HHF) qdisc

This patch implements the first size-based qdisc that attempts to
differentiate between small flows and heavy-hitters.  The goal is to
catch the heavy-hitters and move them to a separate queue with less
priority so that bulk traffic does not affect the latency of critical
traffic.  Currently "less priority" means less weight (2:1 in
particular) in a Weighted Deficit Round Robin (WDRR) scheduler.

In essence, this patch addresses the "delay-bloat" problem due to
bloated buffers. In some systems, large queues may be necessary for
obtaining CPU efficiency, or due to the presence of unresponsive
traffic like UDP, or just a large number of connections with each
having a small amount of outstanding traffic. In these circumstances,
HHF aims to reduce the HoL blocking for latency sensitive traffic,
while not impacting the queues built up by bulk traffic.  HHF can also
be used in conjunction with other AQM mechanisms such as CoDel.

To capture heavy-hitters, we implement the "multi-stage filter" design
in the following paper:
C. Estan and G. Varghese, "New Directions in Traffic Measurement and
Accounting", in ACM SIGCOMM, 2002.

Some configurable qdisc settings through 'tc':
- hhf_reset_timeout: period to reset counter values in the multi-stage
                     filter (default 40ms)
- hhf_admit_bytes:   threshold to classify heavy-hitters
                     (default 128KB)
- hhf_evict_timeout: threshold to evict idle heavy-hitters
                     (default 1s)
- hhf_non_hh_weight: Weighted Deficit Round Robin (WDRR) weight for
                     non-heavy-hitters (default 2)
- hh_flows_limit:    max number of heavy-hitter flow entries
                     (default 2048)

Note that the ratio between hhf_admit_bytes and hhf_reset_timeout
reflects the bandwidth of heavy-hitters that we attempt to capture
(25Mbps with the above default settings).

The false negative rate (heavy-hitter flows getting away unclassified)
is zero by the design of the multi-stage filter algorithm.
With 100 heavy-hitter flows, using four hashes and 4000 counters yields
a false positive rate (non-heavy-hitters mistakenly classified as
heavy-hitters) of less than 1e-4.

Signed-off-by: Terry Lam <vtlam@google.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

3c59x: Add software timestamping

Added software timestamping ability. Tested with linuxptp and synchronized
clocks to an average of less than 200 microseconds on 10 megabit ethernet.

Tested on both Vortex and Boomerang models.

Signed-off-by: Matthew Whitehead <tedheadster@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

lib: hash: follow-up fixups for arch hash

This patch adds the include file to pull in __read_mostly on some
architectures e.g. ppc and also fixes up signatures in generic
asm.

Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Francesco Fusco <ffusco@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

isdn: use strlcpy() instead strcpy()

I don't think the in-kernel drivers ever hit this strcpy() so this
doesn't change how the code works.  But strlcpy() is safer.

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bridge: spelling fixes

Fix spelling errors in bridge driver.

Signed-off-by: Tan Xiaojun <tanxiaojun@huawei.com>
Acked-by: Stephen Hemminger <stephen@networkplumber.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

ipv6: support IPV6_PMTU_INTERFACE on sockets

IPV6_PMTU_INTERFACE is the same as IPV6_PMTU_PROBE for ipv6. Add it
nontheless for symmetry with IPv4 sockets. Also drop incoming MTU
information if this mode is enabled.

The additional bit in ipv6_pinfo just eats in the padding behind the
bitfield. There are no changes to the layout of the struct at all.

Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

ipv4: new ip_no_pmtu_disc mode to always discard incoming frag needed msgs

This new mode discards all incoming fragmentation-needed notifications
as I guess was originally intended with this knob. To not break backward
compatibility too much, I only added a special case for mode 2 in the
receiving path.

Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

inet: make no_pmtu_disc per namespace and kill ipv4_config

The other field in ipv4_config, log_martians, was converted to a
per-interface setting, so we can just remove the whole structure.

Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge git://git./linux/kernel/git/davem/net

Conflicts:
drivers/net/ethernet/intel/i40e/i40e_main.c
drivers/net/macvtap.c

Both minor merge hassles, simple overlapping changes.

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

net: vmxnet3 calls skb_set_hash

Drivers should call skb_set_hash to set the hash and its type
in an skbuff.

Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: sun-niu calls skb_set_hash

Drivers should call skb_set_hash to set the hash and its type
in an skbuff.

Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: sfc calls skb_set_hash

Drivers should call skb_set_hash to set the hash and its type
in an skbuff.

Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: vxge calls skb_set_hash

Drivers should call skb_set_hash to set the hash and its type
in an skbuff.

Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: mlx4 calls skb_set_hash

Drivers should call skb_set_hash to set the hash and its type
in an skbuff.

Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: sky2 calls skb_set_hash

Drivers should call skb_set_hash to set the hash and its type
in an skbuff.

Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: emulex-benet calls skb_set_hash

Drivers should call skb_set_hash to set the hash and its type
in an skbuff.

Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: cisco-enic calls skb_set_hash

Drivers should call skb_set_hash to set the hash and its type
in an skbuff.

Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: cxgb4 calls skb_set_hash

Drivers should call skb_set_hash to set the hash and its type
in an skbuff.

Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>