ath9k: Add support for Adaptive Power Management

This feature is to mitigate the problem of certain 3
stream chips that exceed the PCIe power requirements.An EEPROM flag
controls which chips have APM enabled which is basically read from
miscellaneous configuration element of the EEPROM header.

This workaround will reduce power consumption by using 2 Tx chains for
Single and Double stream rates (5 GHz only).All self generated frames
(regardless of rate) are sent on 2 chains when this feature is
enabled(Chip Limitation).

Cc: Paul Shaw <paul.shaw@atheros.com>
Signed-off-by: Mohammed Shafi Shajakhan <mshajakhan@atheros.com>
Tested-by: Mohammed Shafi Shajakhan <mshajakhan@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

mac80211: fix RX aggregation locking

The RX aggregation locking documentation was
wrong, which led Christian to also code the
timer timeout handling for it somewhat wrongly.

Fix the documentation, the two places that
need to hold the reorder lock across accesses
to the structure, and the debugfs code that
should just use RCU.

Also, remove acquiring the sta->lock across
reorder timeouts since it isn't necessary, and
change a few places to GFP_KERNEL because the
code path here doesn't need atomic allocations
as I noticed when reviewing all this.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Acked-by: Christian Lamparter <chunkeey@googlemail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k: fix software retry counter tracking

The recent tx path cleanups moved the software retry count tracking
from the ath_buf to the skb cb, however the actual counter update
referred to the wrong location, confusing block-ack window tracking.
Fix this by using the retries counter in the struct ath_frame_info.

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

Revert "ath9k_htc: Handle monitor mode properly for HTC devices"

This reverts commit 446fad5a5b6be765c8ec39bfdbbc6c7aa63fbcbb.

The change had broken the packet injection on monitoring mode.

Signed-off-by: Rajkumar Manoharan <rmanoharan@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k: Remove code which enables btcoex based on subsys id

Signed-off-by: Vasanthakumar Thiagarajan <vasanth@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k: Reintroduce modparam to enable btcoex

It is not ideal to enable btcoex based on subsys id as it is
not unique, they are so random. It is also a pain keeping
all of them in a table to enable btcoex for a particular
hw. Going back to the old idea.

Signed-off-by: Vasanthakumar Thiagarajan <vasanth@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

mac80211: implement off-channel mgmt TX

This implements the new off-channel TX API
in mac80211 with a new work item type. The
operation doesn't add a new work item when
we're on the right channel and there's no
wait time so that for example p2p probe
responses will be transmitted without delay.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

nl80211/cfg80211: extend mgmt-tx API for off-channel

With p2p, it is sometimes necessary to transmit
a frame (typically an action frame) on another
channel than the current channel. Enable this
through the CMD_FRAME API, and allow it to wait
for a response. A new command allows that wait
to be aborted.

However, allow userspace to specify whether or
not it wants to allow off-channel TX, it may
actually want to use the same channel only.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwlagn: minor change in bt coex normal LUT

Minor changes in LUT

Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>

iwlagn: use different kill mask when SCO active

use different kill_ack_mask and kill_cts_mask when detect SCO
is active.

Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>

iwlagn: Offical name for 100/130 device

Change to offical name for 100 devices:
   "Intel(R) Centrino(R) Wireless-N 100"

Change to offical name for 130 devices:
   "Intel(R) Centrino(R) Wireless-N 130"

Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>

iwlagn: Offical name for 6000g2b device

Change to offical name for 6000g2b devices:
   "Intel(R) Centrino(R) Wireless-N 1030"
   "Intel(R) Centrino(R) Advanced-N 6230"

Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>

iwlagn: Offical name for 6000g2a device

Change to offical name for 6000g2a devices:
   "Intel(R) Centrino(R) Advanced-N 6205"

Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>

iwlagn: Offical name for 6050g2 device

Change to offical name for 6050g2 devices:
   "Intel(R) Centrino(R) Wireless-N 6150"

Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>

iwlwifi: use mac80211 AC defines

Instead of hardcoding the numbers that must
match mac80211, use the constants. Not that
this means we could change the constants,
but at least this way it's clearer.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>

iwlagn: advertise reliable TX status

Our hardware has reliable TX status, but we're
not currently advertising that to mac80211.
Since the packet loss monitoring will depend
on it, advertise it.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>

iwlagn: fix station powersave accounting for aggregation

Since aggregation queues are station-specific, the
device will not reject packets in them but rather
will stop the appropriate aggregation queues when
a station goes to sleep. I forgot to account for
this in the driver, so if a station went to sleep
that had aggregation enabled, traffic would stop
indefinitely.

Fix this by only accounting frames queued on the
normal AC queues for associated station.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>

iwlagn: remove powersave warning

Through races, a packet may be enqueued for
transmission to a station while that station
is going to sleep, in which case the warning
here triggers. Instead of warning, check the
condition -- if this packet is not a PS-poll
response then we still enqueue it but it will
be rejected by the device since the station
is marked as asleep.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>

iwlwifi: use antenna information in EEPROM

The valid tx/rx antenna information is part of EEPROM, so use it
to configure the device.

For few cases, the EEPROM did not reflect the correct antenna, but
it is too late to modify the EEPROM, so overwrite with .cfg parameters

Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>

iwlwifi: remove unused define

was not used, remove it

Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>

ath9k_hw: remove ath9k_hw_stoppcurecv

It is no longer used anywhere

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k_htc: Identify devices using driver_info

Categorize AR7010 & AR9287 devices based on driver_info
of usb_device_id, instead of PIDs. This avoids per-device cases
and minimize code changes for new device addition.

Signed-off-by: Rajkumar Manoharan <rmanoharan@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k_hw: Fix eeprom offset for AR9287 devices (PCI/USB)

AR9287 devices (PCI/USB) use different eeprom start location
to read nvram. New devices might endup with same devid. So use
driver_info to set offset, instead of devid. driver_info is
valid for HTC devices alone which is filled in usb_device_id.

Signed-off-by: Rajkumar Manoharan <rmanoharan@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k_htc: Add driver_info in usb device list

Added driver_info to identify AR7010, R9287 HTC devices.

Signed-off-by: Rajkumar Manoharan <rmanoharan@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath: Add a driver_info bitmask field

The driver_info stores the device category information which
is used to load appropriate device firmware, select firmware offset
and eeprom starting location. The driver_info is accessed across
ath9k_htc and ath9k_hw. Hence placed under common structure.

Signed-off-by: Rajkumar Manoharan <rmanoharan@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwlwifi: advance power management support

For 6000g2b and up, adding advance power management support
for better power consumption

Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwlwifi: power management checking for shadow register

If shadow register is enable, modify the power management
command to inform uCode

Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwlwifi: consider BT for power management

Check the BT PSPoll flag when fill PM command to uCode

Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwlwifi: add more power management flags

Adding additional power management option available for the device.

Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwlagn: name change for BT config flag

Bit 7 of BT config flag is used to enable/disable PSPoll sync.
Make the name to match it.

Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

mac80211: implement packet loss notification

For drivers that have accurate TX status reporting
we can report the number of consecutive lost packets
to userspace using the new cfg80211 CQM event. The
threshold is fixed right now, this may need to be
improved in the future.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

cfg80211: allow using CQM event to notify packet loss

This adds the ability for drivers to use CQM events
to notify about packet loss for specific stations
(which could be the AP for the managed mode case).
Since the threshold might be determined by the
driver (it isn't passed in right now) it will be
passed out of the driver to userspace in the event.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

cfg80211: Add documentation for antenna ops

Signed-off-by: Bruno Randolf <br1@einfach.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

mac80211: avoid aggregation for VO traffic

This should help with latency issues which can happen when
using aggregation.

Cc: Felix Fietkau <nbd@openwrt.org>
Cc: Matt Smith <matt.smith@atheros.com>
Cc: Senthil Balasubramanian <senthilkumar@atheros.com>
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k: avoid aggregation for VO traffic

This should help with latency issues which can happen when
using aggregation.

Cc: Matt Smith <matt.smith@atheros.com>
Cc: Senthil Balasubramanian <senthilkumar@atheros.com>
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

mac80211: reduce the number of retries for nullfunc probing

Since nullfunc frames are transmitted as unicast frames, they're more
reliable than the broadcast probe requests, so we need fewer retries
to figure out whether the AP is really gone.

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

mac80211: use nullfunc instead of probe request for connection monitoring

nullfunc frames are better for connection monitoring, because probe requests
are answered even if the AP has already dropped the connection, whereas
nullfunc frames from an unassociated station will trigger a disassoc/deauth
frame from the AP (WLAN_REASON_CLASS3_FRAME_FROM_NONASSOC_STA), which allows
the station to reconnect immediately instead of waiting until it attempts to
transmit the next unicast frame.

This only works on hardware with reliable tx ACK reporting, any other hardware
needs to fall back to the probe request method.

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

cfg80211/mac80211: improve ad-hoc multicast rate handling

- store the multicast rate as an index instead of the rate value
  (reduces cpu overhead in a hotpath)
- validate the rate values (must match a bitrate in at least one sband)

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

mac80211: probe the AP when resuming

Check the connection by probing the AP (either using nullfunc or a
probe request). If nullfunc probing is supported and the assoc is no
longer valid, the AP will send a disassoc/deauth immediately.

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

mac80211: calculate beacon loss time accurately

Instead of using a fixed 2 second timeout, calculate beacon loss interval
from the advertised beacon interval and a frame count.  With this beacon
loss happens after N (default 7) consecutive frames are missed which
for a typical setup (100TU beacon interval) is ~700ms (or ~1/3 previous).

Signed-off-by: Sam Leffler <sleffler@chromium.org>
Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

mac80211: restart beacon miss timer on system resume from suspend

Signed-off-by: Paul Stewart <pstew@google.com>
Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

b43: N-PHY: little cleanups

Remove some typos, warnings, initialize some values to follow wl's code path.

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

b43: N-PHY: rev2: save and restore PHY regs on RSSI poll

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

b43: N-PHY: fix BPHY init

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

b43: N-PHY: init BPHY when needed

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

b43: N-PHY: fix some typos, conditions, set gain_boost

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

b43: N-PHY: minor fixes to match specs

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

b43: N-PHY: fix values for PHY regs in channel tables of 2055 radio

Additional comment by Larry Finger <Larry.Finger@lwfinger.net>:

This change deserves a bit more explanation. You might include something like
"These tables came from reverse engineering the 5.10.56.46 version of the
Broadcom driver. Trace comparisons between b43 and the current Broadcom driver
(5.10.120.0) show byte reversals for the PHY register writes."

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

net/wireless: Use pr_<level> and netdev_<level>

No change in output for pr_<level> prefixes.
netdev_<level> output is different, arguably improved.

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

Merge branch 'master' of git://git./linux/kernel/git/linville/wireless-2.6

Revert "nl80211/mac80211: Report signal average"

This reverts commit 86107fd170bc379869250eb7e1bd393a3a70e8ae.

This patch inadvertantly changed the userland ABI.

Signed-off-by: John W. Linville <linville@tuxdriver.com>

wireless: b43: fix error path in SDIO

Fix unbalanced call to sdio_release_host() on the error path.

Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Acked-by: Larry Finger <Larry.Finger@lwfinger.net>
Cc: stable@kernel.org
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k_htc: fix eeprom access

wireless-testing commit a05b5d45049d60a06a1b12976150572304a51928
ath9k: add support for reading eeprom from platform data on PCI devices

This change moved the initialization of the AH_USE_EEPROM flag from ath9k_hw
to ath9k. This needs to be added to ath9k_htc as well

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k_hw: fix A-MPDU key search issues on AR9003

Under load, a large number of frames can produce decryption errors, even when
no key cache update is being done. Performing a key search for every single
frame in an A-MPDU improves reliability.

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

zd1211rw: Use const

Mark arrays const that are unmodified after initializations.

   text    data     bss     dec     hex filename
  19291      56    4136   23483    5bbb drivers/net/wireless/zd1211rw/zd_chip.o.old
  19291      56    4136   23483    5bbb drivers/net/wireless/zd1211rw/zd_chip.o.new

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

rt2x00: Use static const

Using static const generally increases object text and decreases data size.
It also generally decreases overall object size.

   text    data     bss     dec     hex filename
  40197      56    8336   48589    bdcd drivers/net/wireless/rt2x00/rt2800lib.o.new
  40205      56    8336   48597    bdd5 drivers/net/wireless/rt2x00/rt2800lib.o.old

Signed-off-by: Joe Perches <joe@perches.com>
Acked-by: Gertjan van Wingerde <gwingerde@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

rndis_wlan: Use static const

Using static const generally increases object text and decreases data size.
It also generally decreases overall object size.

   text    data     bss     dec     hex filename
  41757    2205    9896   53858    d262 drivers/net/wireless/rndis_wlan.o.old
  41653    2205    9880   53738    d1ea drivers/net/wireless/rndis_wlan.o.new

Changed functions rndis_set_oid and set_bssid to take const *'s.

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ray_cs: Use static const

Using static const generally increases object text and decreases data size.
It also generally decreases overall object size.

   text    data     bss     dec     hex filename
  42607    3581    8536   54724    d5c4 drivers/net/wireless/ray_cs.o.new
  42603    3585    8536   54724    d5c4 drivers/net/wireless/ray_cs.o.old

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

libertas: Use static const

Using static const generally increases object text and decreases data size.
It also generally decreases overall object size.

   text    data     bss     dec     hex filename
   3650      56     704    4410    113a drivers/net/wireless/libertas/rx.o.new
   3695      56     704    4455    1167 drivers/net/wireless/libertas/rx.o.old
  27328     964    5240   33532    82fc drivers/net/wireless/libertas/cfg.o.new
  27328     964    5240   33532    82fc drivers/net/wireless/libertas/cfg.o.old

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwlwifi: Use static const

Using static const generally increases object text and decreases data size.
It also generally decreases overall object size.

   text    data     bss     dec     hex filename
  48644      57   12120   60821    ed95 drivers/net/wireless/b43/phy_n.o.new
  48661      57   12120   60838    eda6 drivers/net/wireless/b43/phy_n.o.old
  37906      86    7904   45896    b348 drivers/net/wireless/iwlwifi/iwl-agn-lib.o.new
  37937      86    7904   45927    b367 drivers/net/wireless/iwlwifi/iwl-agn-lib.o.old
  37781     523    6752   45056    b000 drivers/net/wireless/iwlwifi/iwl-3945.o.new
  37781     523    6752   45056    b000 drivers/net/wireless/iwlwifi/iwl-3945.o.old

Changed b43_nphy_write_clip_detection to take a const u16 *

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

b43: Use static const

Using static const generally increases object text and decreases data size.
It also generally decreases overall object size.

   text    data     bss     dec     hex filename
   5502      56    1336    6894    1aee drivers/net/wireless/b43/phy_common.o.new
   5511      56    1336    6903    1af7 drivers/net/wireless/b43/phy_common.o.old

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

atmel: Use static const

Using static const generally increases object text and decreases data size.
It also generally decreases overall object size.

   text    data     bss     dec     hex filename
  42578     720    8528   51826    ca72 drivers/net/wireless/atmel.o.old
  42578     720    8528   51826    ca72 drivers/net/wireless/atmel.o.new

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

carl9170: Use static const

Using static const generally increases object text and decreases data size.
It also generally decreases overall object size.

   text    data     bss     dec     hex filename
   1897      56     672    2625     a41 drivers/net/wireless/ath/carl9170/cmd.o.new
   1897      56     672    2625     a41 drivers/net/wireless/ath/carl9170/cmd.o.old

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k: Use static const

Using static const generally increases object text and decreases data size.
It also generally decreases overall object size.

   text    data     bss     dec     hex filename
  11161      56    2136   13353    3429 drivers/net/wireless/ath/ath9k/ar9003_paprd.o.new
  11167      56    2136   13359    342f drivers/net/wireless/ath/ath9k/ar9003_paprd.o.old
  15428      56    3056   18540    486c drivers/net/wireless/ath/ath9k/eeprom_4k.o.old
  15451      56    3056   18563    4883 drivers/net/wireless/ath/ath9k/eeprom_4k.o.new
  14087      56    2560   16703    413f drivers/net/wireless/ath/ath9k/eeprom_9287.o.old
  14036      56    2560   16652    410c drivers/net/wireless/ath/ath9k/eeprom_9287.o.new
  10041      56    2384   12481    30c1 drivers/net/wireless/ath/ath9k/ani.o.new
  10088      56    2384   12528    30f0 drivers/net/wireless/ath/ath9k/ani.o.old
   9316    1580    2304   13200    3390 drivers/net/wireless/ath/ath9k/htc_drv_init.o.new
   9316    1580    2304   13200    3390 drivers/net/wireless/ath/ath9k/htc_drv_init.o.old
  16483      56    3432   19971    4e03 drivers/net/wireless/ath/ath9k/ar9003_phy.o.new
  16517      56    3432   20005    4e25 drivers/net/wireless/ath/ath9k/ar9003_phy.o.old
  18221     104    2960   21285    5325 drivers/net/wireless/ath/ath9k/rc.o.old
  18203     104    2960   21267    5313 drivers/net/wireless/ath/ath9k/rc.o.new
  19985      56    4288   24329    5f09 drivers/net/wireless/ath/ath9k/eeprom_def.o.new
  20040      56    4288   24384    5f40 drivers/net/wireless/ath/ath9k/eeprom_def.o.old
  23997      56    4984   29037    716d drivers/net/wireless/ath/ath9k/ar5008_phy.o.old
  23846      56    4984   28886    70d6 drivers/net/wireless/ath/ath9k/ar5008_phy.o.new
  24285      56    3184   27525    6b85 drivers/net/wireless/ath/ath9k/ar9003_eeprom.o.old
  24101      56    3184   27341    6acd drivers/net/wireless/ath/ath9k/ar9003_eeprom.o.new
   6834      56    1032    7922    1ef2 drivers/net/wireless/ath/ath9k/ar9002_phy.o.old
   6780      56    1032    7868    1ebc drivers/net/wireless/ath/ath9k/ar9002_phy.o.new
  36211      64    8624   44899    af63 drivers/net/wireless/ath/ath9k/hw.o.new
  36401      64    8624   45089    b021 drivers/net/wireless/ath/ath9k/hw.o.old
   9281      56    1496   10833    2a51 drivers/net/wireless/ath/ath9k/ar9003_calib.o.old
   9150      56    1496   10702    29ce drivers/net/wireless/ath/ath9k/ar9003_calib.o.new

Use ARRAY_SIZE instead of a magic number.

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath5k: Use static const

Using static const generally increases object text and decreases data size.
It also generally decreases overall object size.

   text    data     bss     dec     hex filename
  11266      56    2464   13786    35da drivers/net/wireless/ath/ath5k/ani.o.old
  11181      56    2464   13701    3585 drivers/net/wireless/ath/ath5k/ani.o.new

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ar9170: Use const

Mark an array const.

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

carl9170: fix init-self regression

The commit: "carl9170: tx path review" introduced a regression.

gcc (with -Winit-self):
tx.c:1264: warning: ‘super’ is used uninitialized in this function

Signed-off-by: Christian Lamparter <chunkeey@googlemail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

libertas: remove duplicated #include

Remove duplicated #include('s) in
  drivers/net/wireless/libertas/cfg.c

Signed-off-by: Huang Weiyi <weiyi.huang@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k: fix recursive locking in the tx flush path

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Tested-by: Ben Greear <greearb@candelatech.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwlwifi: fix modular 3945 only build

If only 3945 is selected, and is a module, build
fails because iwl-legacy.c won't be compiled. Fix
this by adding it to the build correctly.

This doesn't happen for 4965 because it is a bool
option, not tristate, since it's built into the
AGN module.

Reported-by: Randy Dunlap <randy.dunlap@oracle.com>
Tested-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

mac80211: Disable hw crypto for GTKs on AP VLAN interfaces

When using AP VLAN interfaces, each VLAN interface should be in its own
broadcast domain. Hostapd achieves this by assigning different GTKs to
different AP VLAN interfaces.

However, mac80211 drivers are not aware of AP VLAN interfaces and as
such mac80211 sends the GTK to the driver in the context of the base AP
mode interface. This causes problems when multiple AP VLAN interfaces
are used since the driver will use the same key slot for the different
GTKs (there's no way for the driver to distinguish the different GTKs
from different AP VLAN interfaces). Thus, only the clients associated
to one AP VLAN interface (the one that was created last) can actually
use broadcast traffic.

Fix this by not programming any GTKs for AP VLAN interfaces into the hw
but fall back to using software crypto. The GTK for the underlying AP
interface is still sent to the driver.

That means, broadcast traffic to stations associated to an AP VLAN
interface is encrypted in software whereas broadcast traffic to
stations associated to the non-VLAN AP interface is encrypted in
hardware.

Cc: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

cfg80211: Fix regulatory bug with multiple cards and delays

When two cards are connected with the same regulatory domain
if CRDA had a delayed response then cfg80211's own set regulatory
domain would still be the world regulatory domain. There was a bug
on cfg80211's logic such that it assumed that once you pegged a
request as the last request it was already the currently set
regulatory domain. This would mean we would race setting a stale
regulatory domain to secondary cards which had the same regulatory
domain since the alpha2 would match.

We fix this by processing each regulatory request atomically,
and only move on to the next one once we get it fully processed.
In the case CRDA is not present we will simply world roam.

This issue is only present when you have a slow system and the
CRDA processing is delayed. Because of this it is not a known
regression.

Without this fix when a delay is present with CRDA the second card
would end up with an intersected regulatory domain and not allow it
to use the channels it really is designed for. When two cards with
two different regulatory domains were inserted you'd end up
rejecting the second card's regulatory domain request.
This fails with mac80211_hswim's regtest=2 (two requests, same alpha2)
and regtest=3 (two requests, different alpha2) module parameter
options.

This was reproduced and tested against mac80211_hwsim using this
CRDA delayer:

       #!/bin/bash
       echo $COUNTRY >> /tmp/log
       sleep 2
       /sbin/crda.orig

And these regulatory tests:

       modprobe mac80211_hwsim regtest=2
       modprobe mac80211_hwsim regtest=3

Reported-by: Mark Mentovai <mark@moxienet.com>
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Tested-by: Mark Mentovai <mark@moxienet.com>
Tested-by: Bruno Randolf <br1@einfach.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

cfg80211: move mutex locking to reg_process_pending_hints()

This will be required in the next patch and it makes the
next patch easier to review.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Tested-by: Mark Mentovai <mark@moxienet.com>
Tested-by: Bruno Randolf <br1@einfach.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

cfg80211: move reg_work and reg_todo above

These will be used earlier in the next few patches.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Tested-by: Mark Mentovai <mark@moxienet.com>
Tested-by: Bruno Randolf <br1@einfach.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

cfg80211: put core regulatory request into queue

This will simplify the synchronization for pending requests.
Without this we have a race between the core and when we
restore regulatory settings, although this is unlikely
its best to just avoid that race altogether.

Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Tested-by: Mark Mentovai <mark@moxienet.com>
Tested-by: Bruno Randolf <br1@einfach.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

b43: rfkill: use HI enabled bit for all devices

Devices which use LO enabled bit are covered by b43legacy

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

b43: N-PHY: add 2056 radio channels tables

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

Merge branch 'wireless-next-2.6' of git://git./linux/kernel/git/iwlwifi/iwlwifi-2.6

carl9170: fix virtual interface setup crash

This patch fixes a faulty bound check which caused a
crash when too many virtual interface were brought up.

BUG: unable to handle kernel NULL pointer dereference at 00000004
IP: [<f8125f67>] carl9170_op_add_interface+0x1d7/0x2c0 [carl9170]
*pde = 00000000
Oops: 0002 [#1] PREEMPT
Modules linked in: carl9170 [...]
Pid: 4720, comm: wpa_supplicant Not tainted 2.6.37-rc2-wl+
EIP: 0060:[<f8125f67>] EFLAGS: 00210206 CPU: 0
EIP is at carl9170_op_add_interface+0x1d7/0x2c0 [carl9170]
EAX: 00000000 ...
Process wpa_supplicant
Stack:
 f4f88f34 fffffff4 ..
Call Trace:
 [<f8f4e666>] ? ieee80211_do_open+0x406/0x5c0 [mac80211]
 [...]
Code: <89> 42 04 ...
EIP: [<f8125f67>] carl9170_op_add_interface+0x1d7/0x2c0 [carl9170]
CR2: 0000000000000004

Signed-off-by: Christian Lamparter <chunkeey@googlemail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ssb: b43-pci-bridge: Add new vendor for BCM4318

Add new vendor for Broadcom 4318.

Signed-off-by: Daniel Klaffenbach <danielklaffenbach@gmail.com>
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Cc: Stable <stable@kernel.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k: fix timeout on stopping rx dma

It seems that using ath9k_hw_stoppcurecv to stop rx dma is not enough.
When it's time to stop DMA, the PCU is still busy, so the rx enable
bit never clears.
Using ath9k_hw_abortpcurecv helps with getting rx stopped much faster,
with this change, I cannot reproduce the rx stop related WARN_ON anymore.

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Cc: stable@kernel.org
Signed-off-by: John W. Linville <linville@tuxdriver.com>

wl12xx: Unset bssid filter, ssid and bssid from firmware on disassoc

On the disassociation event from the mac80211, the wl12xx driver does not
clear the chipset configuration related to the AP - i.e. it does not perform
a DISCONNECT and then a JOIN with zero SSID and dummy BSSID. Also, it does not
unset the BSSID filter.

Often this is not a problem, as the above is performed upon entering idle
state. But if a scenario arises where a new association is attempted without
cycling through idle state, the new association will fail.

Fix this by resetting the firmware state on disassociation.

Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Reviewed-by: Luciano Coelho <luciano.coelho@nokia.com>
Signed-off-by: Luciano Coelho <luciano.coelho@nokia.com>

wl1271: add support for HW TX fragmentation

Indicate to mac80211 we support HW fragmentation.
Support updates of the fragmentation threshold via the
set_frag_threshold callback.

Signed-off-by: Arik Nemtsov <arik@wizery.com>
Signed-off-by: Luciano Coelho <luciano.coelho@nokia.com>

Revert "wl1271: Change supported channel order for a more optimal scan"

This reverts commit fa21c7a9e4be439e217fe72edbd39b643b643791.

The reverted patch caused more harm than benefit.

Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Reviewed-by: Luciano Coelho <luciano.coelho@nokia.com>
Tested-by: Tuomas Katila <ext-tuomas.2.katila@nokia.com>
Signed-off-by: Luciano Coelho <luciano.coelho@nokia.com>

wl1271: Prevent ad-hoc and active scanning on 11a DFS frequencies

The wl1271 does not support radar detection. Hence, prevent ad-hoc and
active scanning on frequencies requiring DFS.

Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Tested-by: Tuomas Katila <ext-tuomas.2.katila@nokia.com>
Reviewed-by: Luciano Coelho <luciano.coelho@nokia.com>
Signed-off-by: Luciano Coelho <luciano.coelho@nokia.com>

wl1271: cleanup unused code of calibration structures

The cleanup unused code for calibration procedures.

Signed-off-by: Gery Kahn <geryk@ti.com>
Reviewed-by: Luciano Coelho <luciano.coelho@nokia.com>
Signed-off-by: Luciano Coelho <luciano.coelho@nokia.com>

wl1271: Change wl12xx Files Names

All files name prefix removed due to the fact that wl12xx driver supports
wl1271 and wl1273.
Also the definition in Kconfig and header files changed respectively.

Signed-off-by: Shahar Levi <shahar_levi@ti.com>
Signed-off-by: Luciano Coelho <luciano.coelho@nokia.com>

ssb: drop BCM4328 hack for SPROM revision

This hacks leads to incorrect SPROM parsing for me and reading for example MAC
as: 00:00:00:54:00:00. Michael G. who introduced this confirmed it is not
needed anymore.

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Tested-by: Michael Gerdau <mgd@qata.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k_hw: support reading calibration data from flash on AR9003

Embedded boards do not have compressed EEPROM data, they use the
struct ar9003_eeprom layout, with little endian fields, so copying
the raw data to the eeprom buffer is enough.

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k: add support for reading eeprom from platform data on PCI devices

Some embedded boards store platform data for connected PCIe AR92xx
chips in the system flash instead of a separate EEPROM chip.

Signed-off-by: Gabor Juhos <juhosg@openwrt.org>
Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

rndis_wlan: avoid uninitialized var warning in rndis_wlan_craft_connected_bss

  CC [M]  drivers/net/wireless/rndis_wlan.o
drivers/net/wireless/rndis_wlan.c: In function ‘rndis_wlan_craft_connected_bss’:
drivers/net/wireless/rndis_wlan.c:2542:2: warning: ‘ret’ may be used uninitialized in this function

Signed-off-by: John W. Linville <linville@tuxdriver.com>
Acked-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>

iwmc3200wifi: clarify potentially undefined operation in iwm_scan_ssids

  CC [M]  drivers/net/wireless/iwmc3200wifi/commands.o
drivers/net/wireless/iwmc3200wifi/commands.c: In function ‘iwm_scan_ssids’:
drivers/net/wireless/iwmc3200wifi/commands.c:911:15: warning: operation on ‘iwm->scan_id’ may be undefined

Signed-off-by: John W. Linville <linville@tuxdriver.com>

nl80211/mac80211: Report signal average

Extend nl80211 to report an exponential weighted moving average (EWMA) of the
signal value. Since the signal value usually fluctuates between different
packets, an average can be more useful than the value of the last packet.

This uses the recently added generic EWMA library function.

Signed-off-by: Bruno Randolf <br1@einfach.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath5k: Use generic EWMA library

Remove ath5k's private moving average implementation in favour of the generic
library version.

Signed-off-by: Bruno Randolf <br1@einfach.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

lib: Add generic exponentially weighted moving average (EWMA) function

This adds generic functions for calculating Exponentially Weighted Moving
Averages (EWMA). This implementation makes use of a structure which keeps the
EWMA parameters and a scaled up internal representation to reduce rounding
errors.

The original idea for this implementation came from the rt2x00 driver
(rt2x00link.c). I would like to use it in several places in the mac80211 and
ath5k code and I hope it can be useful in many other places in the kernel code.

Signed-off-by: Bruno Randolf <br1@einfach.org>
Reviewed-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k_htc: Avoid setting QoS control for non-QoS frames

Setting tid information in the TX header is required only for QoS
frames. Not handling this case causes severe data loss with some APs.

Cc: stable@kernel.org
Signed-off-by: Rajkumar Manoharan <rmanoharan@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

cfg80211: fix can_beacon_sec_chan, reenable HT40

This follows wireless-testing 9236d838c920e90708570d9bbd7bb82d30a38130
("cfg80211: fix extension channel checks to initiate communication") and
fixes accidental case fall-through. Without this fix, HT40 is entirely
blocked.

Signed-off-by: Mark Mentovai <mark@moxienet.com>
Cc: stable@kernel.org
Acked-by: Luis R. Rodriguez <lrodriguez@atheros.com
Signed-off-by: John W. Linville <linville@tuxdriver.com>

mac80211: fix powersaving clients races

The code to handle powersaving stations has a race:
when the powersave flag is lifted from a station,
we could transmit a packet that is being processed
for TX at the same time right away, even if there
are other frames queued for it. This would cause
frame reordering. To fix this, lift the flag only
under the appropriate lock that blocks TX.

Additionally, the code to allow drivers to block a
station while frames for it are on the HW queue is
never re-enabled the station, so traffic would get
stuck indefinitely. Fix this by clearing the flag
for this appropriately.

Finally, as an optimisation, don't do anything if
the driver unblocks an already unblocked station.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

mac80211: defines for AC numbers

In many places we've just hardcoded the
AC numbers -- which is a relic from the
original mac80211 (d80211). Add constants
for them so we know what we're talking
about.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k_hw: add support for reading EEPROM data from the internal OTP ROM

Some of the new AR9003 cards do not come with an external EEPROM chip
anymore. Calibration data on these cards is stored in the OTP ROM on
the chip.
This patch adds support for reading this data, and also adds support
for different EEPROM chip sizes (512 bytes instead of 1K).

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>