net: forcedeth: fix compile warning of not used nv_set_tso function

Fix the below compile warning:
drivers/net/forcedeth.c:4266: warning: ‘nv_set_tso’ defined but not used

commit 569e146 converts forcedeth driver to use hw_features.
So, implement function of .set_tso is abandoned.

Signed-off-by: Shan Wei <shanwei@cn.fujitsu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge branch 'master' of git://git./linux/kernel/git/kaber/nf-next-2.6

Merge branch 'master' of /linux/kernel/git/davem/net-2.6

Conflicts:
drivers/net/bnx2x/bnx2x_ethtool.c

net: qlge: convert to hw_features

Another simple conversion.

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: vmxnet3: convert to hw_features

This also removes private feature flags that were always set to true.

You may want to move vmxnet3_set_features() to vmxnet3_drv.c as a following
cleanup.

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: vxge: convert to hw_features

Side effect: ->gro_enable is removed as napi_gro_receive() does the
fallback itself.

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: chelsio: convert to hw_features

Also remove flags that were not used or are now redundant to hw_features bits.
No device had UDP_CSUM_CAPABLE set.

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: s2io: convert to hw_features

This removes advertising HW_CSUM as driver does not support it.

Note: driver advertises TSO6 but not IPV6_CSUM - bug maybe?

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: fix section mismatches

Fix build warnings like the following:

WARNING: drivers/net/built-in.o(.data+0x12434): Section mismatch in reference from the variable madgemc_driver to the variable .init.data:madgemc_adapter_ids

And add some consts to EISA device ID tables along the way.

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>

r8169: TSO fixes.

- the MSS value is actually contained in a 11 bits wide (0x7ff) field.
  The extra bit in the former MSSMask did encompass the TSO command
  bit ("LargeSend") as well (0xfff). Oops.

- the Tx descriptor layout is not the same through the whole chipset
  family. The 8169 documentation, the 8168c documentation and Realtek's
  drivers (8.020.00, 1.019.00, 6.014.00) highlight two layouts:
  1. 8169, 8168 up to 8168b (included) and 8101
  2. {8102e, 8168c} and beyond

- notwithstanding the "first descriptor" and "last descriptor" bits, the
  same Tx descriptor content is enforced when a packet consists of several
  descriptors. The chipsets are documented to require it.

Credits go to David Dillow <dave@thedillows.org> for the original patch.

Signed-off-by: Francois Romieu <romieu@fr.zoreil.com>
Cc: Realtek <nic_swsd@realtek.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: myri10ge: convert to hw_features

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Acked-by: Jon Mason <jon.mason@myri.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: dm9000: Fix build

Commit c88fcb (net: dm9000: convert to hw_features) broke the build of
the dm9000 driver since it merged functions which use different names
for the board info structure used for I/O operations without updating
all the references to use the same name. Fix that.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bonding, ipv4, ipv6, vlan: Handle NETDEV_BONDING_FAILOVER like NETDEV_NOTIFY_PEERS

It is undesirable for the bonding driver to be poking into higher
level protocols, and notifiers provide a way to avoid that.  This does
mean removing the ability to configure reptitition of gratuitous ARPs
and unsolicited NAs.

Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

vlan: Propagate NETDEV_NOTIFY_PEERS notifier

The NETDEV_NOTIFY_PEERS notifier indicates that a device moved to a

different physical link; this also applies to any VLAN devices on top
of it.

Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

ipv6: Send unsolicited neighbour advertismements when notified

The NETDEV_NOTIFY_PEERS notifier is a request to send such
advertisements following migration to a different physical link,
e.g. virtual machine migration.

Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bnx2x: cosmetics: Using ethtool_cmd_speed() API

This updates bnx2x to use the ethtool_cmd_speed() family of functions
(see b11f8d8c in 2.6.27-rc3 aka. "ethtool: Expand ethtool_cmd.speed to
32 bits") to get and set the link speed via ethtool. This allows to
avoid manually accessing ethtool_cmd's speed_hi field.

Signed-off-by: David Decotigny <decot@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

ip6_pol_route panic: Do not allow VLAN on loopback

Several tests in the ipv6 routing code check IFF_LOOPBACK, and
allowing stacking such as VLAN'ing on top of loopback results in a
netdevice which reports IFF_LOOPBACK but really isn't the loopback
device.

Instead of spamming the ipv6 routing code with even more special tests,
simply disallow VLAN over loopback.

The result of this patch is:

# modprobe 8021q
# vconfig add lo 43
ERROR: trying to add VLAN #43 to IF -:lo:-  error: Operation not supported

Signed-off-by: Krishna Kumar <krkumar2@in.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

decnet: Don't leak entries when rebuilding zone.

As noticed by Ben Hutchings, when we move entries from
one table to another we leak all except the first entry.

Put back the "next" variable removed by commit
9bf9055eb716f85372c41b3fbc51f90bc7653740 ("decnet: Fix set-but-unused
variable.") and use it properly.

Reported-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

via-rhine: Assign random MAC address if necessary

Roger Luethi has had several reports of Rhine NICs providing
an invalid MAC address.  If so, assign a random MAC address so
the hardware can still be used.

Tested as a standalone interface, as carrier for ppp, and as a
bonding slave.

Original-patch-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

via_rhine: Use netdev_<level> and pr_<level>

Use the more current logging styles.

Add #define DEBUG to make netdev_dbg always active.

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bridge: fix accidental creation of sysfs directory

Commit bb900b27a2f49b37bc38c08e656ea13048fee13b ("bridge: allow
creating bridge devices with netlink") introduced a bug in net-next
because of a typo in notifier. Every device would have the sysfs
bridge directory (and files).

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: cxgb4{,vf}: convert to hw_features

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Acked-by: Dimitris Michailidis <dm@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bnx2x: Fix port identification problem

This patch fixes port identification on optic devices when there's no link on the port.

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

drivers/net/usb/usbnet.c: Use FIELD_SIZEOF macro in usbnet_init() function.

Signed-off-by: Thiago Farina <tfransosi@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: ibm_newemac: convert to hw_features

Side effect: allow toggling of TX offloads.

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: greth: convert to hw_features

Note: Driver modifies its struct net_device_ops. This will break if used for
multiple devices that are not all the same (if that HW config is possible).

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: niu: convert to hw_features

Side effect: allow toggling of TX offloads.

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: ehea: convert to hw_features

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: benet: convert to hw_features - fixup

Remove be_set_flags() as it's already covered by hw_features.

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Acked-by: Ajit Khaparde ajit.khaparde@emulex.com
Signed-off-by: David S. Miller <davem@davemloft.net>

net: typhoon: convert to hw_features

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: mv643xx: convert to hw_features

Side effect: don't reenable RXCSUM on every ifdown/ifup.

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: tehuti: convert to hw_features

As a side effect, make TX offloads changeable.

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>

r8169: add Realtek as maintainer.

Per Hayes's request.

Signed-off-by: Francois Romieu <romieu@fr.zoreil.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: cxgb3: convert to hw_features

This removes some of the remnants of LRO -> GRO conversion.

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: macvlan: convert to hw_features

Not much of a conversion anyway - macvlan has no way to change the offload
settings independently to its base device.

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge branch 'batman-adv/next' of git://git.open-mesh.org/ecsv/linux-merge

be2net: Fix unused-but-set variables.

The variables 'tx_min' and 'tx_max' are set but not used in
be_set_coalesce().

Similarly for 'region' in be_do_flash().

Just kill them off.

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

isdn: i4l: isdn_tty: Fix unused-but-set variables.

The variable 'fcr' is set but not used in isdn_tty_change_speed().

Just kill it off.

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

netfilter: ip6table_mangle: Fix set-but-unused variables.

The variable 'flowlabel' is set but unused in ip6t_mangle_out().

The intention here was to compare this key to the header value after
mangling, and trigger a route lookup on mismatch.

Make it so.

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

netfilter: ip6_tables: Fix set-but-unused variables.

The variable 'target' is set but unused in compat_copy_entry_from_user().

Just kill it off.

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

netfilter: nf_conntrack_standalone: Fix set-but-unused variables.

The variable 'ret' is set but unused in ct_seq_show().

This was obviously meant to be used to propagate error codes
to the caller, so make it so.

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

netfilter: nfnetlink_log: Fix set-but-unused variables.

The variable 'tmp_uint' is set but unused in __build_packet_message().

Just kill it off.

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

netlabel: Fix set-but-unused variables.

The variable 'type_str' is set but unused in netlbl_cipsov4_add().

Just kill it off.

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

l2tp: Fix set-but-unused variables.

The variable 'ret' is set but unused in l2tp_nl_register_ops().

This was obviously meant to maintain error codes which are
returned to the caller, make it so.

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

irda: irproc: Fix set-but-unused variables.

The variable 'd' is set but unused in irda_proc_register().

Just kill it off.

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

irda: irlap_event: Fix set-but-unused variables.

The variable 'ret' is set but unused in irlap_state_sclose().

Just kill it off.

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

caif: Fix set-but-unused variables.

The variable 'caifdef' is set but unused in modemcmd().

Similarly for 'net' in receive(), and 'res' in
caif_device_notify() and caif_exit_net().

Just kill them off.

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

bnx2x: Fix set-but-unused variables.

The variable 'rc' is set but unused in bnx2x_timer().

Similarly for 'hc_index_p' in bnx2x_init_sb(), and 'port' in
bnx2x_get_hwinfo().

Just kill them off.

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

bna: Fix set-but-unused variables.

The variable 'pgoff' is set but unused in bfa_nw_ioc_fwver_get()
and bfa_ioc_download_fw().

Similarly for 'cmd_h' in bna_mbox_flush_q and the entirety of
bna_rit_mod_uninit() is unused since variables are purely set but no
action is made using them.

Same for 'bna' in bna_rit_create() and 'ret' in bna_rx_create().

Just kill them off.

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

isdn: mISDN: socket: Fix set-but-unused variables.

The variable 'len' is set but unused in data_sock_getsockopt().

The code should use 'len' to validate that the user's socket option is
indeed the right size.

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

isdn: i4l: isdn_net: Fix set-but-unused variables.

The variable 'unused' is set but unused in
isdn_net_ciscohdlck_slarp_in().  Just kill it off.

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

isdn: i4l: isdn_common: Fix set-but-unused variables.

The variable 'ch' is set but unused in isdn_capi_rec_hl_msg().  Just
kill it off.

Similarly for 'chidx' in isdn_ioctl() and 'di' in
isdn_capi_rec_hl_msg().

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

isdn: teles_cs: Fix set-but-unused variables.

The variable 'dev' is set but unused in teles_cs_config().  Just
kill it off.

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

isdn: l3ni1: Fix set-but-unused variables.

The variable 'cause' is set but unused in ni1up().  Just
kill it off.

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

isdn: l3dss1: Fix set-but-unused variables.

The variable 'cause' is set but unused in dss1up().  Just
kill it off.

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

isdn: jade: Fix set-but-unused variables.

The variable 'i' is set but unused in JadeVersion().  Just
kill it off.

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

isdn: ipacx: Fix set-but-unused variables.

The variable 'cda2_cr' is set but unused in ctrl_complete().  Just
kill it off.

Keep the cs->readisac() call just in case the register read has side
effects.

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

isdn: hfc_usb: Fix set-but-unused variables.

The variable 'buf' is set but unused in
ctrl_complete().  Just kill it off.

Similarly for 'err' in hfc_usb_init().

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

isdn: elsa_ser: Fix set-but-unused variables.

The variable 'bits' is set but unused in
change_speed().  Just kill it off.

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

isdn: elsa_cs: Fix set-but-unused variables.

The variable 'dev' is set but unused in
elsa_cs_config().  Just kill it off.

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

isdn: arcofi: Fix set-but-unused variables.

The variable 'val' is set but unused in
send_arcofi().  Just kill it off.

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

isdn: hfcsusb: Fix set-but-unused variables.

The variable 'buf' is set but unused in
ctrl_complete().  Just kill it off.

Similarly for the variable 'err' in setup_hfcsusb().

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

isdn: hfcpci: Fix set-but-unused variables.

The variable 'total' is set but unused in
hfcpci_empty_bfifo().  Just kill it off.

Similarly for the variable 'val' in ph_state_nt().

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

isdn: eicon: Fix set-but-unused variables.

The variable 'best_id' is set but unused in
diva_mnt_add_xdi_adapter().  Just kill it off.

Similarly for the variable 'CIP' in connect_req(), 'Number' in
sig_ind(), 'Info' in dtmf_confirmation() mixer_command()
fax_connect_ack_command() fax_edata_ack_command()
rtp_connect_b3_res_command() and rtp_connect_b3_res_command(), and 'a'
in mixer_indication_coefs_set(),

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

isdn: gigaset: Fix set-but-unused variable.

The variable 'offset' is set but unused in
write_iso_tasklet().  Just kill it off.

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

batman-adv: Set the txqueuelen to zero when creating soft interface

Like other virtual interfaces, e.g. br0, we don't need a transmit
queue. Packets should only be queued on real interfaces which are
underneath. In practice this patch makes little difference since the
virtual interfaces can accept packets as fast as they come, but the
patch will avoid bufferbloat questions to the mailling lists in the
future.

Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Tested-by: Linus Lüssing <linus.luessing@web.de>
Signed-off-by: Marek Lindner <lindner_marek@yahoo.de>
Signed-off-by: Sven Eckelmann <sven@narfation.org>

batman-adv: protect softif_neigh by rcu

Add get/set wrapper functions for softif_neigh and
use rcu functions to manipulate the pointers.

Signed-off-by: Simon Wunderlich <siwu@hrz.tu-chemnitz.de>
Signed-off-by: Marek Lindner <lindner_marek@yahoo.de>
Signed-off-by: Sven Eckelmann <sven@narfation.org>

batman-adv: concentrate all curr_gw related rcu operations in select/deselect functions

Signed-off-by: Marek Lindner <lindner_marek@yahoo.de>
Signed-off-by: Sven Eckelmann <sven@narfation.org>

batman-adv: Protect global TQ window with a spinlock

Signed-off-by: Linus Lüssing <linus.luessing@web.de>
Signed-off-by: Marek Lindner <lindner_marek@yahoo.de>
Signed-off-by: Sven Eckelmann <sven@narfation.org>

batman-adv: Make orig_node->router an rcu protected pointer

The rcu protected macros rcu_dereference() and rcu_assign_pointer()
for the orig_node->router need to be used, as well as spin/rcu locking.
Otherwise we might end up using a router pointer pointing to already
freed memory.

Therefore this commit introduces the safe getter method
orig_node_get_router().

Signed-off-by: Linus Lüssing <linus.luessing@web.de>
Signed-off-by: Marek Lindner <lindner_marek@yahoo.de>
Signed-off-by: Sven Eckelmann <sven@narfation.org>

batman-adv: Simplify gw_check_election(), use gw_get_selected()

gw_get_selected() can get us the desired orig_node directly, therefore
reusing that function in gw_check_election().

Signed-off-by: Linus Lüssing <linus.luessing@web.de>
Signed-off-by: Marek Lindner <lindner_marek@yahoo.de>
Signed-off-by: Sven Eckelmann <sven@narfation.org>

batman-adv: Make gateway_get_selected type safe

Make the return value explicit instead of (void *).

Signed-off-by: Linus Lüssing <linus.luessing@web.de>
Signed-off-by: Marek Lindner <lindner_marek@yahoo.de>
Signed-off-by: Sven Eckelmann <sven@narfation.org>

batman-adv: Move bonding / iface alternating router search to own functions

This decreases the size of find_router() by outsourcing the router
search for the bonding and interface alternating modes to their own sub
functions. This shall make it easier to keep track of the correct
refcounting later.

Signed-off-by: Linus Lüssing <linus.luessing@web.de>
Signed-off-by: Marek Lindner <lindner_marek@yahoo.de>
Signed-off-by: Sven Eckelmann <sven@narfation.org>

atl1c: Fix set-but-unused variable.

The variable 'extra_size' is set but unused in
atl1c_configure_tx().  Just kill it off.

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

atl1e: Fix set-but-unused variable.

The variable 'tx_ring' is set but unused in
atl1e_init_ring_resources().  Just kill it off.

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

atlx: Fix set-but-unused variable.

The variable 'tpc' is set but unused in atl1_intr_tx().
Just kill it off.

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

bonding: Fix set-but-unused variable.

The variable 'vlan_dev' is set but unused in
bond_send_gratuitous_arp().  Just kill it off.

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

atm: lec: Fix set-but-unused variables.

The variable 'eth' is set but unused in lec_handle_bridge().  Also,
the variable 'priv' is set but unused in lane_module_cleanup().

Just kill them off.

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

ax25: Fix set-but-unused variable.

The variable 's' is set but unused in ax25_protocol_release().
Just kill it off.

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

econet: Fix set-but-unused variable.

#if 0'd out code for IP handling in aun_data_available() has
been commented out since the beginning, which makes the variable
"ip" set but not used.

Kill it off as well as the stub code.

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

decnet: Fix set-but-unused variable.

"next" in dn_rebuild_zone() is set but not actually used,
kill it off.

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

pktgen: Fix set-but-unused variable.

"iph" in pktgen_output_ipsec() is set but never actually
used.  Kill it off.

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

atm: solos-pci: Fix set-but-unused variable.

This is just a readback to entire completion of a register
write, keep the readback but kill the unused variable.

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

atm: idt77252: Fix set-but-unused variables.

Two cases here:

1) idt77252_rx_raw() really does not make any use of the
   extracted PTI field of the atm header.

2) idt77252_collect_stat() only uses the register values
   in code which has been compiled out by a "NOTDEF" cpp
   test for more than 10 years.  Just kill this NOTDEF
   code entirely, but keep the register reads in case
   they have side effects.

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

atm: he: Fix undefined sequence points.

GCC complains in these queue index operations because we
perform operations of the form:

x = some_operation(++x);

which is undefined.  Replace with:

x = some_operation(x + 1);

which is well defined and provides the intended operation.

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

atm: eni: Kill set-but-unused variables.

The variable eni_dev is initialized but never subsequently used in
these two functions.

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

net: gianfar: convert to hw_features

Note: I bet that gfar_set_features() don't really need a full reset.

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: mlx4: convert to hw_features

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: forcedeth: convert to hw_features

This also fixes a race around np->txrxctl_bits while changing RXCSUM offload.

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: dm9000: convert to hw_features

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: spider_net: convert to hw_features

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: minor cleanup to net_namespace.c.

Inline a small static function that's only ever called from one place.

Signed-off-by: Rob Landley <rlandley@parallels.com>
Reviewed-by: Jiri Pirko <jpirko@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

ipv6: RTA_PREFSRC support for ipv6 route source address selection

[ipv6] Add support for RTA_PREFSRC

This patch allows a user to select the preferred source address
for a specific IPv6-Route. It can be set via a netlink message
setting RTA_PREFSRC to a valid IPv6 address which must be
up on the device the route will be bound to.

Signed-off-by: Daniel Walter <dwalter@barracuda.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

ipv6: ignore looped-back NA while dad is running

[ipv6] Ignore looped-back NAs while in Duplicate Address Detection

If we send an unsolicited NA shortly after bringing up an
IPv6 address, the duplicate address detection algorithm
fails and the ip stays in tentative mode forever.
This is due a missing check if the NA is looped-back to us.

Signed-off-by: Daniel Walter <dwalter@barracuda.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

ip: ip_options_compile() resilient to NULL skb route

Scot Doyle demonstrated ip_options_compile() could be called with an skb
without an attached route, using a setup involving a bridge, netfilter,
and forged IP packets.

Let's make ip_options_compile() and ip_options_rcv_srr() a bit more
robust, instead of changing bridge/netfilter code.

With help from Hiroaki SHIMODA.

Reported-by: Scot Doyle <lkml@scotdoyle.com>
Tested-by: Scot Doyle <lkml@scotdoyle.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Stephen Hemminger <shemminger@vyatta.com>
Acked-by: Hiroaki SHIMODA <shimoda.hiroaki@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

rndis_host: Quirky devices are still 'point-to-point'

My changes in commit 4d42d417be75d750b82798922b6e775915e11bce were
written some time before the introduction of FLAG_POINTTOPOINT, so
didn't include that flag in the new driver_info.  Change the new
driver_info to be consistent.

Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>

net-bonding: Adding support for throughputs larger than 65536 Mbps

This updates the bonding driver to support v2.6.27-rc3 enhancements
(b11f8d8c aka. "ethtool: Expand ethtool_cmd.speed to 32 bits") which
allow to encode the Mbps link speed on 32-bits (Max 4 Pbps) instead of
16 (Max 65536 Mbps).

This patch also attempts to compact struct slave by reordering its
fields.

Signed-off-by: David Decotigny <decot@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net-bonding: Fix minor/cosmetic type inconsistencies

The __get_link_speed() function returns a u16 value which was stored
in a u32 local variable. This patch uses the return value directly,
thus fixing that minor type consistency.

The 'duplex' field in struct slave being encoded on 8 bits, to be more
consistent we use a u8 integer (instead of u16) whenever we copy it to
local variables.

Signed-off-by: David Decotigny <decot@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net-bonding: Fix minor sparse complaints

This gets rid of minor sparse complaints:
drivers/net/bonding/bond_main.c:4361:4: warning: do-while statement is not a compound statement
drivers/net/bonding/bond_main.c:243:12: warning: symbol 'bond_mode_name' was not declared. Should it be static?

Signed-off-by: David Decotigny <decot@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

qlge: make nic_operations struct const

The struct nic_operations is just function pointers and should be
declared const for added security.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>