netfilter: ipset: make possible to hash some part of the data element only

Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Signed-off-by: Patrick McHardy <kaber@trash.net>

netfilter: nfnetlink_queue: batch verdict support

Introduces a new nfnetlink type that applies a given
verdict to all queued packets with an id <= the id in the verdict
message.

If a mark is provided it is applied to all matched packets.

This reduces the number of verdicts that have to be sent.
Applications that make use of this feature need to maintain
a timeout to send a batchverdict periodically to avoid starvation.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Patrick McHardy <kaber@trash.net>

netfilter: nfnetlink_queue: assert monotonic packet ids

Packet identifier is currently setup in nfqnl_build_packet_message(),
using one atomic_inc_return().

Problem is that since several cpus might concurrently call
nfqnl_enqueue_packet() for the same queue, we can deliver packets to
consumer in non monotonic way (packet N+1 being delivered after packet
N)

This patch moves the packet id setup from nfqnl_build_packet_message()
to nfqnl_enqueue_packet() to guarantee correct delivery order.

This also removes one atomic operation.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
CC: Florian Westphal <fw@strlen.de>
CC: Pablo Neira Ayuso <pablo@netfilter.org>
CC: Eric Leblond <eric@regit.org>
Signed-off-by: Patrick McHardy <kaber@trash.net>

netfilter: nfnetlink_queue: provide rcu enabled callbacks

nenetlink_queue operations on SMP are not efficent if several queues are
used, because of nfnl_mutex contention when applications give packet
verdict.

Use new call_rcu field in struct nfnl_callback to advertize a callback
that is called under rcu_read_lock instead of nfnl_mutex.

On my 2x4x2 machine, I was able to reach 2.000.000 pps going through
user land returning NF_ACCEPT verdicts without losses, instead of less
than 500.000 pps before patch.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
CC: Florian Westphal <fw@strlen.de>
CC: Eric Leblond <eric@regit.org>
Signed-off-by: Patrick McHardy <kaber@trash.net>

netfilter: nfnetlink: add RCU in nfnetlink_rcv_msg()

Goal of this patch is to permit nfnetlink providers not mandate
nfnl_mutex being held while nfnetlink_rcv_msg() calls them.

If struct nfnl_callback contains a non NULL call_rcu(), then
nfnetlink_rcv_msg() will use it instead of call() field, holding
rcu_read_lock instead of nfnl_mutex

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
CC: Florian Westphal <fw@strlen.de>
CC: Eric Leblond <eric@regit.org>
Signed-off-by: Patrick McHardy <kaber@trash.net>

netfilter: add SELinux context support to AUDIT target

In this revision the conversion of secid to SELinux context and adding it
to the audit log is moved from xt_AUDIT.c to audit.c with the aid of a
separate helper function - audit_log_secctx - which does both the conversion
and logging of SELinux context, thus also preventing internal secid number
being leaked to userspace. If conversion is not successful an error is raised.

With the introduction of this helper function the work done in xt_AUDIT.c is
much more simplified. It also opens the possibility of this helper function
being used by other modules (including auditd itself), if desired. With this
addition, typical (raw auditd) output after applying the patch would be:

type=NETFILTER_PKT msg=audit(1305852240.082:31012): action=0 hook=1 len=52 inif=? outif=eth0 saddr=10.1.1.7 daddr=10.1.2.1 ipid=16312 proto=6 sport=56150 dport=22 obj=system_u:object_r:ssh_client_packet_t:s0
type=NETFILTER_PKT msg=audit(1306772064.079:56): action=0 hook=3 len=48 inif=eth0 outif=? smac=00:05:5d:7c:27:0b dmac=00:02:b3:0a:7f:81 macproto=0x0800 saddr=10.1.2.1 daddr=10.1.1.7 ipid=462 proto=6 sport=22 dport=3561 obj=system_u:object_r:ssh_server_packet_t:s0

Acked-by: Eric Paris <eparis@redhat.com>
Signed-off-by: Mr Dash Four <mr.dash.four@googlemail.com>
Signed-off-by: Patrick McHardy <kaber@trash.net>

netfilter: ipset: whitespace and coding fixes detected by checkpatch.pl

Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Signed-off-by: Patrick McHardy <kaber@trash.net>

netfilter: ipset: hash:net,iface type introduced

The hash:net,iface type makes possible to store network address and
interface name pairs in a set. It's mostly suitable for egress
and ingress filtering. Examples:

        # ipset create test hash:net,iface
        # ipset add test 192.168.0.0/16,eth0
        # ipset add test 192.168.0.0/24,eth1

Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Signed-off-by: Patrick McHardy <kaber@trash.net>

netfilter: ipset: use the stored first cidr value instead of '1'

Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Signed-off-by: Patrick McHardy <kaber@trash.net>

netfilter: ipset: fix return code for destroy when sets are in use

Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Signed-off-by: Patrick McHardy <kaber@trash.net>

netfilter: ipset: add xt_action_param to the variant level kadt functions, ipset API change

With the change the sets can use any parameter available for the match
and target extensions, like input/output interface. It's required for
the hash:net,iface set type.

Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Signed-off-by: Patrick McHardy <kaber@trash.net>

netfilter: ipset: use unified from/to address masking and check the usage

Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Signed-off-by: Patrick McHardy <kaber@trash.net>

netfilter: ipset: take into account cidr value for the from address when creating the set

When creating a set from a range expressed as a network like
10.1.1.172/29, the from address was taken as the IP address part and
not masked with the netmask from the cidr.

Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Signed-off-by: Patrick McHardy <kaber@trash.net>

netfilter: ipset: adding ranges to hash types with timeout could still fail, fixed

The patch "Fix adding ranges to hash types" had got a mistypeing
in the timeout variant of the hash types, which actually made
the patch ineffective. Fixed!

Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Signed-off-by: Patrick McHardy <kaber@trash.net>

netfilter: ipset: support range for IPv4 at adding/deleting elements for hash:*net* types

The range internally is converted to the network(s) equal to the range.
Example:

# ipset new test hash:net
# ipset add test 10.2.0.0-10.2.1.12
# ipset list test
Name: test
Type: hash:net
Header: family inet hashsize 1024 maxelem 65536
Size in memory: 16888
References: 0
Members:
10.2.1.12
10.2.1.0/29
10.2.0.0/24
10.2.1.8/30

Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Signed-off-by: Patrick McHardy <kaber@trash.net>

netfilter: ipset: set type support with multiple revisions added

A set type may have multiple revisions, for example when syntax is
extended. Support continuous revision ranges in set types.

Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Signed-off-by: Patrick McHardy <kaber@trash.net>

netfilter: ipset: fix adding ranges to hash types

When ranges are added to hash types, the elements may trigger rehashing
the set. However, the last successfully added element was not kept track
so the adding started again with the first element after the rehashing.

Bug reported by Mr Dash Four.

Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Signed-off-by: Patrick McHardy <kaber@trash.net>

netfilter: ipset: support listing setnames and headers too

Current listing makes possible to list sets with full content only.
The patch adds support partial listings, i.e. listing just
the existing setnames or listing set headers, without set members.

Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Signed-off-by: Patrick McHardy <kaber@trash.net>

netfilter: ipset: options and flags support added to the kernel API

The support makes possible to specify the timeout value for
the SET target and a flag to reset the timeout for already existing
entries.

Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Signed-off-by: Patrick McHardy <kaber@trash.net>

netfilter: ipset: whitespace fixes: some space before tab slipped in

Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Signed-off-by: Patrick McHardy <kaber@trash.net>

netfilter: ipset: timeout can be modified for already added elements

When an element to a set with timeout added, one can change the timeout
by "readding" the element with the "-exist" flag. That means the timeout
value is reset to the specified one (or to the default from the set
specification if the "timeout n" option is not used). Example

ipset add foo 1.2.3.4 timeout 10
ipset add foo 1.2.3.4 timeout 600 -exist

Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Signed-off-by: Patrick McHardy <kaber@trash.net>

Merge branch 'master' of git://git./linux/kernel/git/horms/ipvs-next-2.6

Merge branch 'master' of /repos/git/net-next-2.6

tg3: Create funcs for power source switching

The power source switching code is about to get a little more complex.
This patch seeks to simplify future power source switching patches by
clarifying the existing code.

Signed-off-by: Matt Carlson <mcarlson@broadcom.com>
Reviewed-by: Michael Chan <mchan@broadcom.com>
Reviewed-by: Benjamin Li <benli@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

tg3: Migrate phy preprocessor defs to system defs

This patch changes to code to use some of the preprocessor
definitions from mii.h over its homegrown equivalents.

Signed-off-by: Matt Carlson <mcarlson@broadcom.com>
Reviewed-by: Michael Chan <mchan@broadcom.com>
Reviewed-by: Benjamin Li <benli@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

tg3: Show flowctrl settings through get_settings()

This patch adds code to present the flow control advertisements through
the ethtool get_settings callback.

Signed-off-by: Matt Carlson <mcarlson@broadcom.com>
Reviewed-by: Michael Chan <mchan@broadcom.com>
Reviewed-by: Benjamin Li <benli@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

tg3: Fix EEE debounce timer values

This patch fixes the EEE debounce timer values.

Signed-off-by: Matt Carlson <mcarlson@broadcom.com>
Reviewed-by: Michael Chan <mchan@broadcom.com>
Reviewed-by: Benjamin Li <benli@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

tg3: Add more selfboot formats to NVRAM selftest

This patch adds more selfboot formats to the NVRAM selftest.  It also
changes the code to return an error on an unsupported NVRAM format.

Signed-off-by: Matt Carlson <mcarlson@broadcom.com>
Reviewed-by: Michael Chan <mchan@broadcom.com>
Reviewed-by: Benjamin Li <benli@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

tg3: Remove 4G_DMA_BNDRY_BUG flag

Now that all chips have this bug, the flag checks become useless code.
This patch removes the flag.

Signed-off-by: Matt Carlson <mcarlson@broadcom.com>
Reviewed-by: Michael Chan <mchan@broadcom.com>
Reviewed-by: Benjamin Li <benli@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

tg3: Remove 40BIT_DMA_LIMIT_BUG

This patch removes the 40BIT_DMA_LIMIT_BUG flag.  There already exists a
flag for this purpose (TG3_FLAG_40BIT_DMA_BUG) and was already being
used in the correct spot.

Signed-off-by: Matt Carlson <mcarlson@broadcom.com>
Reviewed-by: Michael Chan <mchan@broadcom.com>
Reviewed-by: Benjamin Li <benli@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

tg3: Workaround tagged status update bug

On rare occasions, writing the tag to the interrupt mailbox does not
reenable interrupts.  This patch fixes the problem by reissuing the
mailbox update.

Signed-off-by: Matt Carlson <mcarlson@broadcom.com>
Signed-off-by: Michael Chan <mchan@broadcom.com>
Reviewed-by: Benjamin Li <benli@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

bnx2x: Update date to 2011/06/13 and version to 1.70.00-0

Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

bnx2x: PFC support for 578xx

Add supoprt for 3 COSes for 578xx devices.
Fix HW configuration for PFC feature according to new HSI in link layer.

Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

bnx2x: Rename LASI registers to definitions in mdio.h

Signed-off-by: Yaniv Rosner <yanivr@broadcom.com>
Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

bnx2x: Add a periodic task for link PHY events

Signed-off-by: Yaniv Rosner <yanivr@broadcom.com>
Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

bnx2x: Adjust BCM84833 to BCM578xx

Signed-off-by: Yaniv Rosner <yanivr@broadcom.com>
Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

bnx2x: Adjust ETS to 578xx

Signed-off-by: Yaniv Rosner <yanivr@broadcom.com>
Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

bnx2x: Add new PHY 54616s

Signed-off-by: Yaniv Rosner <yanivr@broadcom.com>
Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

bnx2x: Add Warpcore support for 578xx

Signed-off-by: Yaniv Rosner <yanivr@broadcom.com>
Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

bnx2x: Add new MAC support for 578xx

Signed-off-by: Yaniv Rosner <yanivr@broadcom.com>
Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

bnx2x: Cosmetic changes.

Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

bnx2x: update DCB data during PMF migration

Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

bnx2x: 57712 parity handling

- Added support for a parity error handling for a 57712 chip.
 - Changed the parity recovery scheme from per-chip to per-engine.

Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

New 7.0 FW: bnx2x, cnic, bnx2i, bnx2fc

New FW/HSI (7.0):
 - Added support to 578xx chips
 - Improved HSI - much less driver's direct access to the FW internal
   memory needed.

New implementation of the HSI handling layer in the bnx2x (bnx2x_sp.c):
 - Introduced chip dependent objects that have chip independent interfaces
   for configuration of MACs, multicast addresses, Rx mode, indirection table,
   fast path queues and function initialization/cleanup.
 - Objects functionality is based on the private function pointers, which
   allows not only a per-chip but also PF/VF differentiation while still
   preserving the same interface towards the driver.
 - Objects interface is not influenced by the HSI changes which do not require
   providing new parameters keeping the code outside the bnx2x_sp.c invariant
   with regard to such HSI chnages.

Changes in a CNIC, bnx2fc and bnx2i modules due to the new HSI.

Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: Bhanu Prakash Gollapudi <bprakash@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

bnx2x: Created bnx2x_sp

Moved the HSI dependent slow path code to a separate file.
Currently it contains the implementation of MACs, Rx mode,
multicast addresses, indirection table, fast path queue and function
configuration code.

Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

bnx2x: removed unused variables

Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

bnx2x: use bnx2x_reload_if_running

Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

bnx2x: dump FW memory when appropriate msglvl is raised

Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

bnx2x: do not call link update without HW notification

Signed-off-by: Yaniv Rosner <yanivr@broadcom.com>
Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

bnx2x: disable fairness if ETS is enabled

Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

bnx2x: avoid release of unrequested irqs

Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

bnx2x: put start bd csum in separate function

Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

bnx2x: remove references to intr_sem

It's not needed any more since device always operates in interrupt-driven mode

Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

bnx2x: do not allocate FCoE ring if disabled

Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

cnic: Move indexing function pointers to struct kcq_info

The hardware indexing scheme for the FCoE kcq will change in the upcoming
firmware.  This patch will cope with the change easily.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

IPVS: remove unused init and cleanup functions.

After restructuring, there is some unused or empty functions
left to be removed.

Signed-off-by: Hans Schillstrom <hans.schillstrom@ericsson.com>
Signed-off-by: Simon Horman <horms@verge.net.au>

IPVS: labels at pos 0

Put goto labels at the beginig of row
acording to coding style example.

Signed-off-by: Hans Schillstrom <hans.schillstrom@ericsson.com>
Signed-off-by: Simon Horman <horms@verge.net.au>

vmxnet3: remove unused variable

While doing some backporting I noticed that vmxnet3 had a variable that was set
but never used.  Get rid of it, and stop the compiler from griping

Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
CC: "VMware, Inc." <pv-drivers@vmware.com>
CC: "David S. Miller" <davem@davemloft.net>
Signed-off-by: Bhavesh Davda <bhavesh@vmware.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

bonding:delete a dereference before check

Dan Carpenter found that there was a dereference before a check,
added in 56d00c677de0(bonding:delete lacp_fast from ad_bond_info).

Signed-off-by: Weiping Pan <panweiping3@gmail.com>
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

l2tp: fix l2tp_ip_sendmsg() route handling

l2tp_ip_sendmsg() in non connected mode incorrectly calls
sk_setup_caps(). Subsequent send() calls send data to wrong destination.

We can also avoid changing dst refcount in connected mode, using
appropriate rcu locking. Once output route lookups can also be done
under rcu, sendto() calls wont change dst refcounts too.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
CC: James Chapman <jchapman@katalix.com>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

smsc9420: enable transmit time stamping.

This patch enables software (and phy device) transmit time stamping
for the smsc9420. Compile tested only.

Cc: Steve Glendinning <steve.glendinning@smsc.com>
Signed-off-by: Richard Cochran <richard.cochran@omicron.at>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

stmmac: enable transmit time stamping.

This patch enables software (and phy device) transmit time stamping
for the STMicroelectronics Ethernet driver. Compile tested only.

Cc: Giuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: Richard Cochran <richard.cochran@omicron.at>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

r6040: enable transmit time stamping.

This patch enables software (and phy device) transmit time stamping
for the RDC R6040 Fast Ethernet MAC. Compile tested only.

Cc: Florian Fainelli <florian@openwrt.org>
Signed-off-by: Richard Cochran <richard.cochran@omicron.at>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

ethoc: enable transmit time stamping.

This patch enables software (and phy device) transmit time stamping
for the OpenCores 10/100 MAC driver. Compile tested only.

Cc: Thierry Reding <thierry.reding@avionic-design.de>
Signed-off-by: Richard Cochran <richard.cochran@omicron.at>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

dnet: enable transmit time stamping.

This patch enables software (and phy device) transmit time stamping
in the "Dave ethernet interface." Compile tested only.

Cc: Ilya Yanok <yanok@emcraft.com>
Signed-off-by: Richard Cochran <richard.cochran@omicron.at>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

tg3: enable transmit time stamping.

This patch enables software (and phy device) transmit time stamping
for the TIGON3 driver. Compile tested only.

Cc: Matt Carlson <mcarlson@broadcom.com>
Cc: Michael Chan <mchan@broadcom.com>
Signed-off-by: Richard Cochran <richard.cochran@omicron.at>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

davinci_emac: enable transmit time stamping.

This patch enables software (and phy device) transmit time stamping
for the DaVinci EMAC driver. Tested together with the dp83640 PHY.

Cc: Anant Gole <anantgole@ti.com>
Cc: Kevin Hilman <khilman@deeprootsystems.com>
Signed-off-by: Richard Cochran <richard.cochran@omicron.at>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

davinci_emac: pass ioctls through to phy device.

The DaVinci EMAC driver does not implement any ioctls, but still it can
pass them through to the phy device. This makes it possible for a phy
to offer PHC capabilities.

Cc: Anant Gole <anantgole@ti.com>
Cc: Kevin Hilman <khilman@deeprootsystems.com>
Signed-off-by: Richard Cochran <richard.cochran@omicron.at>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

fec: enable transmit and receive time stamping.

This patch has been tested on the Freescale M5234BCC, which includes the
National Semiconductor DP83640 with IEEE 1588 support.

Cc: Greg Ungerer <gerg@uclinux.org>
Cc: "Uwe Kleine-König" <u.kleine-koenig@pengutronix.de>
Cc: Shawn Guo <shawn.guo@freescale.com>
Signed-off-by: Richard Cochran <richard.cochran@omicron.at>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

net: export time stamp utility function for Ethernet MAC drivers

The network stack provides the function, skb_clone_tx_timestamp().
Ethernet MAC drivers can call this via the transmit time stamping
hook, skb_tx_timestamp(). This commit exports the clone function so
that drivers using it can be compiled as modules.

Signed-off-by: Richard Cochran <richard.cochran@omicron.at>
Signed-off-by: David S. Miller <davem@conan.davemloft.net>

IPVS: rename of netns init and cleanup functions.

Make it more clear what the functions does,
on request by Julian.

Signed-off-by: Hans Schillstrom <hans.schillstrom@ericsson.com>
Signed-off-by: Hans Schillstrom <hans@schillstrom.com>
Signed-off-by: Simon Horman <horms@verge.net.au>

IPVS remove unused var from migration to netns

Remove variable ctl_key from struct netns_ipvs,
it's a leftover from early netns work.

Signed-off-by: Hans Schillstrom <hans.schillstrom@ericsson.com>
Signed-off-by: Simon Horman <horms@verge.net.au>

ipvs: support more FTP PASV responses

Change the parsing of FTP commands and responses to
support skip character. It allows to detect variations in
the 227 PASV response.

Signed-off-by: Julian Anastasov <ja@ssi.bg>
Signed-off-by: Simon Horman <horms@verge.net.au>

snmp: reduce percpu needs by 50%

SNMP mibs use two percpu arrays, one used in BH context, another in USER
context. With increasing number of cpus in machines, and fact that ipv6
uses per network device ipstats_mib, this is consuming a lot of memory
if many network devices are registered.

commit be281e554e2a (ipv6: reduce per device ICMP mib sizes) shrinked
percpu needs for ipv6, but we can reduce memory use a bit more.

With recent percpu infrastructure (irqsafe_cpu_inc() ...), we no longer
need this BH/USER separation since we can update counters in a single
x86 instruction, regardless of the BH/USER context.

Other arches than x86 might need to disable irq in their
irqsafe_cpu_inc() implementation : If this happens to be a problem, we
can make SNMP_ARRAY_SZ arch dependent, but a previous poll
( https://lkml.org/lkml/2011/3/17/174 ) to arch maintainers did not
raise strong opposition.

Only on 32bit arches, we need to disable BH for 64bit counters updates
done from USER context (currently used for IP MIB)

This also reduces vmlinux size :

1) x86_64 build
$ size vmlinux.before vmlinux.after
   text    data     bss     dec     hex filename
7853650 1293772 1896448 11043870  a8841e vmlinux.before
7850578 1293772 1896448 11040798  a8781e vmlinux.after

2) i386  build
$ size vmlinux.before vmlinux.afterpatch
   text    data     bss     dec     hex filename
6039335  635076 3670016 10344427  9dd7eb vmlinux.before
6037342  635076 3670016 10342434  9dd022 vmlinux.afterpatch

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
CC: Andi Kleen <andi@firstfloor.org>
CC: Ingo Molnar <mingo@elte.hu>
CC: Tejun Heo <tj@kernel.org>
CC: Christoph Lameter <cl@linux-foundation.org>
CC: Benjamin Herrenschmidt <benh@kernel.crashing.org
CC: linux-arch@vger.kernel.org
Signed-off-by: David S. Miller <davem@davemloft.net>

bonding: clean up bond_del_vlan()

1) the setting of NETIF_F_VLAN_CHALLENGED in bond_del_vlan() is
useless since commit b2a103e6 because bond_fix_features() now
sets NETIF_F_VLAN_CHALLENGED whenever the last slave is being
removed.

2) the code never triggers anyway as vlan_list is never empty
since ad1afb00.

Signed-off-by: Jiri Bohac <jbohac@suse.cz>
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: fix MIPS fallout from "net: remove interrupt.h inclusion from netdevice.h"

Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

be2net: Enable NETIF_F_TSO6 for VLAN traffic for BE

NETIF_F_TSO6 for VLAN packets was not enabled for BE adapters.
Enabling it.

Signed-off-by: Padmanabh Ratnakar <padmanabh.ratnakar@emulex.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

virtio_net: introduce VIRTIO_NET_HDR_F_DATA_VALID

There's no need for the guest to validate the checksum if it have been
validated by host nics. So this patch introduces a new flag -
VIRTIO_NET_HDR_F_DATA_VALID which is used to bypass the checksum
examing in guest. The backend (tap/macvtap) may set this flag when
met skbs with CHECKSUM_UNNECESSARY to save cpu utilization.

No feature negotiation is needed as old driver just ignore this flag.

Iperf shows 12%-30% performance improvement for UDP traffic. For TCP,
when gro is on no difference as it produces skb with partial
checksum. But when gro is disabled, 20% or even higher improvement
could be measured by netperf.

Signed-off-by: Jason Wang <jasowang@redhat.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: DM9000: Add support for byte EEPROM access

Given many versions of ethtool's reluctance to do anything other than
byte accesses to the EEPROM interface, it is easier to update the driver
to support byte accesses so that all the ethtool versions that have been
observed in Debian can write the EEPROM.

Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Reviewed-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

sctp: kzalloc() error handling on deleting last address

Signed-off-by: Michio Honda <micchie@sfc.wide.ad.jp>
Acked-by: Wei Yongjun <yjwei@cn.fujitsu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: add Faraday FTGMAC100 Gigabit Ethernet driver

FTGMAC100 Ethernet Media Access Controller supports 10/100/1000 Mbps
and MII/GMII.  This driver has been working on some ARM/NDS32 SoC's
including Faraday A369 and Andes AG102.

Signed-off-by: Po-Yu Chuang <ratbert@faraday-tech.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net/m68k: Include <linux/interrupt.h> where needed

arch/m68k/emu/nfeth.c: In function ‘nfeth_init’:
arch/m68k/emu/nfeth.c:243: error: implicit declaration of function ‘request_irq’
arch/m68k/emu/nfeth.c:243: error: ‘IRQF_SHARED’ undeclared (first use in this function)
arch/m68k/emu/nfeth.c:243: error: (Each undeclared identifier is reported only once
arch/m68k/emu/nfeth.c:243: error: for each function it appears in.)
arch/m68k/emu/nfeth.c: In function ‘nfeth_cleanup’:
arch/m68k/emu/nfeth.c:266: error: implicit declaration of function ‘free_irq’
drivers/net/apne.c: In function ‘apne_probe’:
drivers/net/apne.c:189: error: implicit declaration of function ‘free_irq’
drivers/net/apne.c: In function ‘apne_probe1’:
drivers/net/apne.c:317: error: implicit declaration of function ‘request_irq’
drivers/net/apne.c:317: error: ‘IRQF_SHARED’ undeclared (first use in this function)
drivers/net/apne.c:317: error: (Each undeclared identifier is reported only once
drivers/net/apne.c:317: error: for each function it appears in.)

Introduced by commit a6b7a407865a ("net: remove interrupt.h inclusion from
netdevice.h").

Include <linux/interrupt.h> in the individual drivers to fix the build.

Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

ixgbevf: Update the driver string

Signed-off-by: Greg Rose <gregory.v.rose@intel.com>
Tested-by: Evan Swanson <evan.swanson@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

rtnetlink: Compute and store minimum ifinfo dump size

The message size allocated for rtnl ifinfo dumps was limited to
a single page.  This is not enough for additional interface info
available with devices that support SR-IOV and caused a bug in
which VF info would not be displayed if more than approximately
40 VFs were created per interface.

Implement a new function pointer for the rtnl_register service that will
calculate the amount of data required for the ifinfo dump and allocate
enough data to satisfy the request.

Signed-off-by: Greg Rose <gregory.v.rose@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

igb: Change version to remove number after -k in kernel versions.

This patch changes the way versioning is done for igb in the kernel by
removing the number after the "k."  It has been determined that just the
"k" is sufficient to identify a kernel version and the following number
was used in an inconsistent manner.

Signed-off-by: Carolyn Wyborny <carolyn.wyborny@intel.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

igbvf: update version number

Update the version number to match version conventions. Bump the major
version to indicate that new hardware support (i350) has been added.

Signed-off-by: Mitch Williams <mitch.a.williams@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

e1000e: update driver version

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

e1000e: remove redundant reverse dependency on CRC32

Commit 5d03078a6804bf4c7f943c5b68bef80468c0717f added a redundant 'select
CRC32'; remove it.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

e1000e: Clear host wakeup bit on 82577/8 without touching PHY page 800

The Host Wakeup Active bit in the PHY Port General Configuration register
(page 769 register 17) must be cleared after every PHY reset to prevent an
unexpected wake signal from the PHY. Originally, this was accomplished by
simply reading the PHY Wakeup Control register on page 800 which clears the
Host Wakeup Active bit as a side-effect. Unfortunately, a hardware bug on
the 82577 and 82578 PHY can cause unexpected behavior when registers on
page 800 are accessed while in gigabit mode.

This patch changes the remaining instances when the Host Wakeup Active bit
needs to be cleared while possibly in gigabit mode by accessing the Port
General Configuration register directly instead of accessing any register
on page 800.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

e1000e: access multiple PHY registers on same page at the same time

Doing a PHY page select can take a long time, relatively speaking. This
can cause a significant delay when updating a number of PHY registers on
the same page by unnecessarily setting the page for each PHY access. For
example when going to Sx, all the PHY wakeup registers (WUC, RAR[], MTA[],
SHRAR[], IP4AT[], IP6AT[], etc.) on 82577/8/9 need to be updated which
takes a long time which can cause issues when suspending.

This patch introduces new PHY ops function pointers to allow callers to
set the page directly and do any number of PHY accesses on that page.
This feature is currently only implemented for 82577, 82578 and 82579
PHYs for both the normally addressed registers as well as the special-
case addressing of the PHY wakeup registers on page 800. For the latter
registers, the existing function for accessing the wakeup registers has
been divided up into three- 1) enable access to the wakeup register page,
2) perform the register access and 3) disable access to the wakeup register
page. The two functions that enable/disable access to the wakeup register
page are necessarily available to the caller so that the caller can restore
the value of the Port Control (a.k.a. Wakeup Enable) register after the
wakeup register accesses are done.

All instances of writing to multiple PHY registers on the same page are
updated to use this new method and to acquire any PHY locking mechanism
before setting the page and performing the register accesses, and release
the locking mechanism afterward.

Some affiliated magic number cleanup is done as well.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

e1000e: do not schedule the Tx queue until ready

Start the Tx queue when the interface is brought up in e1000e_up() but do
not schedule the queue until link is up as detected in the watchdog task
which sets netif_carrier_on.

Also flush the descriptors and clean the Tx and Rx rings before resetting
the hardware when bringing the interface down otherwise there is a small
window where the watchdog task can be triggered with netif_carrier_off
and the Tx ring not yet empty which causes an additional and unnecessary
reset.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

e1000e: log when swflag is cleared unexpectedly on ICH/PCH devices

Since EXTCNF_CTRL.SWFLAG (used in the ownership arbitration of shared
resources, e.g. the PHY shared between the s/w, f/w, and h/w clients)
can be cleared by any of those clients, log a debug message when
software attempts to clear it and it is already cleared unexpectedly.
And since the swflag is cleared by a hardware reset, the driver does
not need to do that, but the mutex acquired when the bit is set must
still be cleared.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

e1000e: 82579 intermittently disabled during S0->Sx

When repeatedly cycling Sx->S0 states with the network cable unplugged,
the 82579 PHY may not initialize as expected and may require a full power
cycle to recover functionality to the device.  Workaround this by testing
access of the PHY registers after resuming; if that returns unexpected
results toggle the LANPHYPC signal to power cycle the PHY.

This is implemented in the new function e1000_resume_workarounds_pchlan()
which calls another new function, e1000_toggle_lanphypc_value_ich8lan(),
which has been created to reduce code duplication (same functionality
required by a previous workaround).  Also, e1000e_disable_gig_wol_ich8lan
is now e1000_suspend_workarounds_ich8lan to better reflect what it does.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

e1000e: disable far-end loopback mode on ESB2

The ESB2 LAN includes a debug feature that enables far-end loopback (FELB)
of the SerDes/Kumeran interface.  This feature is activated when receiving
a sequence of symbols that includes a reserved codeword.  On a perfect
link, FELB would never be activated.  In the presence of bit errors, there
is a very small, but non-zero, probability of FELB being activated.

If the FELB is activated, the SerDes link becomes non-functional and must
be reset.  It could also corrupt the switching tables in the switch since
the ESB2 is transmitting packets with a different source MAC address.

This patch disables the FELB feature.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

bonding: delete unused arp_mon_pt

Now all received packets are handled by bond_handle_frame,
and arp_mon_pt isn't used any more.

Reviewed-by: WANG Cong <xiyou.wangcong@gmail.com>
Signed-off-by: Weiping Pan <panweiping3@gmail.com>
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bonding: delete unused ad_timer

Now we use agg_select_timer and ad_work.

Reviewed-by: WANG Cong <xiyou.wangcong@gmail.com>
Signed-off-by: Weiping Pan <panweiping3@gmail.com>
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bonding:delete agg_select_mode from ad_bond_info

bond_params->ad_select and ad_bond_info->agg_select_mode have the same
meaning, they are duplicate and need extra synchronization.

__get_agg_selection_mode() get ad_select from bond_params directly.

Signed-off-by: Weiping Pan <panweiping3@gmail.com>
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bonding:delete lacp_fast from ad_bond_info

These is also a bug, that if you modify lacp_rate via sysfs,
and add new slaves in bonding, new slaves won't use the latest lacp_rate,
since ad_bond_info->lacp_fast is initialized only once,
in bond_3ad_initialize().

Since both struct bond_params and ad_bond_info have lacp_fast,
they are duplicate and need extra synchronization.

bond_3ad_bind_slave() can use bond_params->lacp_fast to initialize port.
So we can just remove lacp_fast from struct ad_bond_info.

Signed-off-by: Weiping Pan <panweiping3@gmail.com>
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bonding: make 802.3ad use latest lacp_rate

There is bug that when you modify lacp_rate via sysfs,
802.3ad won't use the new value of lacp_rate to transmit packets.
This is because port->actor_oper_port_state isn't changed.

Signed-off-by: Weiping Pan <panweiping3@gmail.com>
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

enic: Bug Fix: Fix hardware transmit queue indexing in enic_poll_controller

Signed-off-by: Christian Benvenuti <benve@cisco.com>
Signed-off-by: Danny Guo <dannguo@cisco.com>
Signed-off-by: Vasanthy Kolluri <vkolluri@cisco.com>
Signed-off-by: Roopa Prabhu <roprabhu@cisco.com>
Signed-off-by: David Wang <dwang2@cisco.com>
Signed-off-by: David S. Miller <davem@davemloft.net>