I frequently run into this warning, and added some
debugging to see why, and got this:

b43 bad rx: 00000000: 2f d2 e2 63 cf a7 14 04 28 18 c8 5f 88 4a a2 00
                      bogus junk       | plcp            | fctl| dur

b43 bad rx: 00000010: 00 11 24 91 07 4d 00 06 25 ff 8f 78 00 06 25 ff
                      my MAC address   | BSSID           | AP MAC
b43 bad rx: 00000020: 8f 76 20 74 00 00 42 07 00 20 00 00 00 00 aa aa
                           | seq | QoS | CCMP IV               | data
                      ...

As you can see, there are 6 bogus bytes (sometimes only five) and then
the frame. I don't know why, and I don't see how to recover, so let's
just drop these frames.
Signed-off-by: NJohannes Berg <johannes@sipsolutions.net>
Signed-off-by: NMichael Buesch <mb@bu3sch.de>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

This implements proper short slot handling and adds code to
program the hardware for the correct response rates derived
from the basic rate set for the current BSS.

(port from b43)
Signed-off-by: NJohannes Berg <johannes@sipsolutions.net>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

'iwconfig eth1 channel 6' would trigger association to _something_,
which is wrong.  Changing the channel should (and does) trigger reassociation,
but only if there is an SSID to associate with.
Signed-off-by: NAlexey Fisher <bug-track@fisher-privat.net>
Acked-by: NZhu Yi <yi.zhu@intel.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

This enables beacons to come through on STA/IBSS.
It should fix sporadic connection issues. Right now
mac80211 expect beacons so give it beacons.
Signed-off-by: NLuis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

Allow APs to receive beacons to detect when it needs
to use protection to update the NAV correctly on
11b stations.
Signed-off-by: NLuis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

*On a previous patch i splitted AR5K_INT_TX to multiple different TX
interrupt flags for better handling but i forgot to unmask the new
TXDESC and TXEOL interrupts on ath5k_init and only left TXOK. However
for each queue we enable TXDESC and TXEOL interrupts, not TXOK so we don't
handle TX interrupts at all (because these interrupts remain masked on
PISR) and under load it results packet loss. Fix the problem by
unmasking TXDESC and TXEOL on ath5k_init.
Signed-Off-by: NNick Kossifidis <mickflemm@gmail.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

This patch decodes the MAC/BB version (for instance: AR5416) and the RF
part version (for instance: AR5133). It has been tested on AR5416/AR5133
which is a 2.4/5GHz 11n device. It also makes the differences between
AR5416 (PCI) and AR5418 (PCI Express). Both are named AR5416 in
the register definitions.
Signed-off-by: NBenoit Papillault <benoit.papillault@free.fr>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

On ARM alignment is done slightly different from other architectures.
struct ieee80211_tx_rate is aligned to word size, even though it only has 3
single-byte members, which triggers the BUILD_BUG_ON in
ieee80211_tx_info_clear_status

This patch marks the struct ieee80211_tx_rate as packed, so that ARM
behaves like the other architectures.
Signed-off-by: NFelix Fietkau <nbd@openwrt.org>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

Noise floor calibration occasionally fails on Atheros hardware.
This is not fatal and can happen if there's simply too much
noise on the air. Ignoring the calibration error is the right
thing to do here, because when the error is ignored, the hardware
will still work, whereas if the error causes the driver to bail out
of a bigger configuration function and does not configure the tx
queues or the IMR (as is the case in reset.c), the hw no longer
works properly until the next reset.
Signed-off-by: NFelix Fietkau <nbd@openwrt.org>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

With the new configuration handling, and more specifically
splitting the configuration of the antenna from the normal
configuration steps allowed a BUG_ON() to be triggered
in the driver because the SW_DIVERSITY was send to the
driver. This fixes that by catching the value early in
rt2x00config.c and replacing it with a sensible value.

This also fixes a problem where the antenna is not being
initialized at all when the radio is enabled. Since it
no longer is part of the mac80211 configuration the
only place where rt2x00 configured it was the SW diversity
handler. Obviously this is broken for all non-diversity
hardware and breaks SW diversity due to a broken initialization.

When the radio is enabled the antenna will be configured
once as soon as the config() callback function is called.
Signed-off-by: NIvo van Doorn <IvDoorn@gmail.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

mac80211 will call set_key() when the device is
shutting down. When the device is unplugged the
keys will be lost automatically due to the power
loss. When the device is not plugged but the module
is only unloaded the keys can remain in the device
hardware, when the module is loaded the keys will
be cleaned up during initialization.

This should prevent the problem reported by Johannes Berg,
where unplugging the device while suspended resulted in
a NULL pointer error during set_key() which was
caused because of the CSR base address being freed.
Signed-off-by: NIvo van Doorn <IvDoorn@gmail.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

For every global LED state change (register/unregister,
suspend/resume) we should force the LEDS to turn off.
This makes sure that the LEDS will always be in a sane
state after the state switch.

Note that when unregister is called but the LED class
wasn't resumed yet, we shouldn't change the LED state
since we might not have access to the device (device
was unplugged while suspended).

Also remove the checks in the activity, assoc and
radio LEDS which blocked calls to brightness_set()
when the state hasn't changed. Some of those LEDS
could be enabled by themselves when something happens
in the hardware (e.g. firmware is loaded). We already
did called rt2x00leds to switch the LED off, but those
calls were blocked.
Signed-off-by: NIvo van Doorn <IvDoorn@gmail.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

mac80211 is in charge of determining the basic rates,
so we are not using the RATE_BASIC flag anymore.
Signed-off-by: NIvo van Doorn <IvDoorn@gmail.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

This change improves the maintainability of these drivers. No functionality
is changed.
Signed-off-by: NLarry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

Initial mesh support: add Mesh Point to supported interfaces mask and allow
hwsim to send beacons in mesh mode.
Signed-off-by: NAndrey Yurovsky <andrey@cozybit.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

Mesh interfaces are currently opened with the FIF_ALLMULTI rx filter flag set,
however there is no BSSID in mesh so BSSID filtering should be disabled by
setting the FIF_OTHER_BSS flag as well.  Also explicitly call
ieee80211_configure_filter for mesh.
Signed-off-by: NAndrey Yurovsky <andrey@cozybit.com>
Signed-off-by: NJavier Cardona <javier@cozbit.com>
Acked-by: NJohannes Berg <johannes@sipsolutions.net>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

This patch is part of a larger patch series which will remove
the "char bus_id[20]" name string from struct device. The device
name is managed in the kobject anyway, and without any size
limitation, and just needlessly copied into "struct device".

To set and read the device name dev_name(dev) and dev_set_name(dev)
must be used. If your code uses static kobjects, which it shouldn't
do, "const char *init_name" can be used to statically provide the
name the registered device should have. At registration time, the
init_name field is cleared, to enforce the use of dev_name(dev) to
access the device name at a later time.

We need to get rid of all occurrences of bus_id in the entire tree
to be able to enable the new interface. Please apply this patch,
and possibly convert any remaining remaining occurrences of bus_id.

We want to submit a patch to -next, which will remove bus_id from
"struct device", to find the remaining pieces to convert, and finally
switch over to the new api, which will remove the 20 bytes array
and does no longer have a size limitation.

CC: Michael Buesch <mb@bu3sch.de>
Acked-by: NGreg Kroah-Hartman <gregkh@suse.de>
Signed-Off-By: NKay Sievers <kay.sievers@vrfy.org>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

Based on a patch from Shailendra Govardhan <shailen@marvell.com>.
This patch allows implementation of more specific wake-on-lan rules than those
of ethtool.

Please note that only firmware 5.110.22.p20 and above supports this feature.

This patch only implements the driver/firmware interface, not the
userspace/driver interface.
Signed-off-by: NAnna Neal <anna@cozybit.com>
Signed-off-by: NJavier Cardona <javier@cozybit.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

Otherwise some configs can result in missing print_ssid...
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

We have some reasons to kill netdev->priv:
1. netdev->priv is equal to netdev_priv().
2. netdev_priv() wraps the calculation of netdev->priv's offset, obviously
   netdev_priv() is more flexible than netdev->priv.
But we cann't kill netdev->priv, because so many drivers reference to it
directly.

OK, becasue Dave S. Miller said, "every direct netdev->priv usage is a bug",
and I want to kill netdev->priv later, I decided to convert all the direct
reference of netdev->priv first.

(Original patch posted by Wang Chen <wangchen@cn.fujitsu.com> w/ above
changelog but using dev->ml_priv.  That doesn't seem appropriate
to me for this driver, so I've revamped it to use netdev_priv()
instead. -- JWL)
Reviewed-by: NWang Chen <wangchen@cn.fujitsu.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

Signed-off-by: NWang Chen <wangchen@cn.fujitsu.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch adds support for SFP+ PHY in the following device ID's (10DB,
10F1, 10E1).  These SFP+ PHY's are accessed via an I2C interface so the
patch also includes functions to support this.

Another feature of note is that the PHY is pluggable and some
rearchitecting was needed to support this.
Signed-off-by: NDonald Skidmore <donald.c.skidmore@intel.com>
Signed-off-by: NPeter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Adds an interface to configure the Backward Congestion Notification
(BCN) feature.  In a BCN capabale network, congestion notifications
from congested points out in the network can cause the end station
limit the rate of a given traffic flow.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NPeter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Adds a netlink interface for Data Center Bridging (DCB) to get and set
the enable state of the Priority Flow Control (PFC) feature.
Primarily, this is a way to turn off PFC in the driver while DCB
remains enabled.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NPeter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Adds interface for Data Center Bridging (DCB) to query (and set if
supported) the number of traffic classes currently supported by the
device for the two (DCB) features: priority groups (PG) and priority
flow control (PFC).
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NPeter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Adds to the netlink interface for Data Center Bridging (DCB), allowing
the DCB capabilities supported by a device to be queried.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NPeter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This adds support for Data Center Bridging (DCB) features in the ixgbe
driver and adds an rtnetlink interface for configuring DCB to the
kernel.  The DCB feature support included are Priority Grouping (PG) -
which allows bandwidth guarantees to be allocated to groups to traffic
based on the 802.1q priority, and Priority Based Flow Control (PFC) -
which introduces a new MAC control PAUSE frame which works at
granularity of the 802.1p priority instead of the link (IEEE 802.3x).
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NPeter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Now TCP & DCCP use RCU lookups, we can convert ehash rwlocks to spinlocks.

/proc/net/tcp and other seq_file 'readers' can safely be converted to 'writers'.

This should speedup writers, since spin_lock()/spin_unlock()
only use one atomic operation instead of two for write_lock()/write_unlock()
Signed-off-by: NEric Dumazet <dada1@cosmosbay.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Convert ipgre tunnel to netdevice ops.
Signed-off-by: NStephen Hemminger <shemminger@vyatta.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Like IPV4, convert the tunnel virtual devices to use net_device_ops.
Signed-off-by: NStephen Hemminger <shemminger@vyatta.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Convert to network device ops. Needed to change to directly call
the init routine since two sides share same ops.  In the process
found by inspection a device ref count leak if register_netdevice failed.
Signed-off-by: NStephen Hemminger <shemminger@vyatta.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Convert the HIPPI infrastructure for use with net_device_ops.
Signed-off-by: NStephen Hemminger <shemminger@vyatta.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Convert this driver to net_device_ops.
Signed-off-by: NStephen Hemminger <shemminger@vyatta.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Similar to ethernet. Convert infrastructure and the one lone FDDI
driver (for the one lone user of that hardware??). Compile tested only.
Signed-off-by: NStephen Hemminger <shemminger@vyatta.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Convert to new network device ops interface.
Signed-off-by: NStephen Hemminger <shemminger@vyatta.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Convert to new network device ops interface.
Signed-off-by: NStephen Hemminger <shemminger@vyatta.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch moves neigh_setup and hard_start_xmit into the network device ops
structure. For bisection, fix all the previously converted drivers as well.
Bonding driver took the biggest hit on this.

Added a prefetch of the hard_start_xmit in the fast path to try and reduce
any impact this would have.
Signed-off-by: NStephen Hemminger <shemminger@vyatta.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Conflicts:

	drivers/net/ixgbe/ixgbe_main.c
	include/net/mac80211.h
	net/phonet/af_phonet.c

* 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6:
  [IA64] xen: fix xen_get_eflags.
  [IA64] ia64/pv_ops/pv_cpu_ops: fix _IA64_REG_IP case.
  [IA64] remove duplicate include iommu.h
  [IA64] use mprintk instead of printk, in ia64_mca_modify_original_stack
  [IA64] Rationalize kernel mode alignment checking