ath5k: remove double opmode definition

opmode (operating mode) was defined in struct ath5k_hw and struct ath5k_softc.
remove it from ath5k_hw and use only from ath5k_softc (sc->opmode).

(btw: what's the meaning of opmode when we have multiple interfaces?)

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

ath5k: preserve antenna settings

save antenna settings and preserve across resets.

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

ath5k: add antenna statistics and debugfs file for antenna settings

keep statistics about which antenna was used for TX and RX. this is used only
for debugging right now, but might have other applications later.

add a new file 'antenna' in debugfs (/sys/kernel/debug/ath5k/phy0/antenna) to show
antenna use statistics and antenna diversity related register values. it can
also be used to set the antenna mode until we have proper support for that in
iw:
  - echo diversity > antenna: use default antenna mode (RX and TX diversity)
  - echo fixed-a > antenna: use fixed antenna A for RX and TX
  - echo fixed-b > antenna: use fixed antenna B for RX and TX
  - echo clear > antenna: reset antenna statistics

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

rndis_wlan: Implement cfg80211 PMKSA API

Add support for cfg80211 set_pmksa/del_pmksa/flush_pmksa. Updating PMKID
entry list is done on driver side since NDIS API requires full list update.

v2:
 - fixed to use new netdev_dbg/warn/etc instead of old devdbg/warn/etc
 - fixed false padding from struct ndis_80211_bssid_info

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

rndis_wlan: get max_num_pmkids from device

Extend rndis_wlan_get_caps() to get 802.11 capabilities and maximum
supported number of PMKIDs by device.

v2: fixed to use new netdev_dbg/warn/etc instead of old devdbg/warn/etc

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

rndis_wlan: remove unused variables from priv structure

Some variables were left unused after cfg80211 conversion. Remove those and related deadcode.

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

rndis_wlan: copy only useful data from rndis_command respond

rndis_query_oid() uses full output buffer size to copy response buffer
from rndis_command()/device. This doesn't cause problems as response buffer
is sized based on output buffer but does copy extra unset bytes.
So change rndis_query_oid() so that only meaningful bytes are being copied.

Also in case of malfunctioning device/cable/etc returned data offset from
device might be wrong so bound check memory access correctly, so add
checks for this.

v2: fixed to use new netdev_dbg/warn/etc instead of old devdbg/warn/etc

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

minstrel: make the rate control ops reusable from another rc implementation

This patch makes it possible to reuse the minstrel rate control ops
from another rate control module. This is useful in preparing for the
new 802.11n implementation of minstrel, which will reuse the old code
for legacy stations.

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

minstrel: simplify and fix debugfs code

This patch cleans up the debugfs read function for the statistics by
using simple_read_from_buffer instead of its own semi-broken hack.
Also removes a useless member of the minstrel debugfs info struct.

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

ath5k: Fix TX/RX padding for all frames

Currently, the padding position is based on
ieee80211_get_hdrlen_from_skb(). This is not correct since the HW does
padding on RX (and expect the same padding to be present on TX) at the
following position :

- management : 24 + 6 if 4-addr format
- control    : 24 + 6 if 4-addr format
- data       : 24 + 6 if 4-addr format + 2 if QoS
- invalid    : 24 + 6 if 4-addr format

whereas ieee80211_get_hdrlen_from_skb() is :

- management : 24
- control    : 16 except for ACK/CTS where it is 10
- data       : 24 + 6 if 4-addr format + 2 if QoS + 2 if QoS & order
- invalid    : 24

So, correct frames are not affected : management frames do not use
4-addr format, control frames have no body and invalid frames are ...
not valid by definition. However, in order to use monitor interface for
debugging purpose, one must be able to send/receive any frames, be it
correct or not. Such frames are affected by incorrect padding.

Moreover, since padding is added on TX, we need to remove it before
calling ieee80211_tx_status. This affect TX packets received by monitor
interfaces.

It has been tested between an ath5k based card (AR5212) and an ar9170usb
based card (netgear WNDA3100) using a frame generator and a monitor
interface for each card.

v2: Added ath5k_add_padding / ath5k_remove_padding

Signed-off-by: Benoit Papillault <benoit.papillault@free.fr>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

mac80211: give warning if building w/out rate ctrl algorithm

I discovered that if EMBEDDED=y, one can accidentally build a mac80211 stack
and drivers w/ no rate control algorithm.  For drivers like RTL8187 that don't
supply their own RC algorithms, this will cause ieee80211_register_hw to
fail (making the driver unusable).

This will tell kconfig to provide a warning if no rate control algorithms
have been selected.  That'll at least warn the user; users that know that
their drivers supply a rate control algorithm can safely ignore the
warning, and those who don't know (or who expect to be using multiple
drivers) can select a default RC algorithm.

Signed-off-by: Andres Salomon <dilinger@collabora.co.uk>
Cc: stable@kernel.org
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwmc3200wifi: typo fix and code cleanup

Fix wrong IWM_RX_TICKET_DROP_REASON_MSK macro define, typo and other
small cleanups.

Signed-off-by: Zhu Yi <yi.zhu@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwmc3200wifi: mark some pmksa functions static

This fixes the sparse warnings.

Signed-off-by: Zhu Yi <yi.zhu@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwmc3200wifi: increase concatenated buffer

Increase concatenated buffer from 8K to 32K to get better performance.

Signed-off-by: Zhu Yi <yi.zhu@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwmc3200wifi: protect rx_tickets and rx_packets[] lists

Protect rx_tickets and rx_packets[] lists with spinlocks to fix the
race condition for concurrent list operations. In iwmc3200wifi both
sdio_isr_worker and rx_worker workqueues can access the rx ticket
and packets lists at the same time under high rx load.

Signed-off-by: Zhu Yi <yi.zhu@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwmc3200wifi: add ftrace event tracing support

Add event tracer for iwmc3200wifi driver. When enabled, all the
commands and responses between the driver and firmware (also
including Tx/Rx frames) will be recorded in the ftrace ring buffer.

Signed-off-by: Zhu Yi <yi.zhu@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwmc3200wifi: remove "_safe" for some list_for_each_entry usage

Use list_for_each_entry instead of list_for_each_entry_safe in
places iteration against list entry removal is not required.

Signed-off-by: Zhu Yi <yi.zhu@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwmc3200wifi: refuse to associate on unallowed channels

We need to make sure we don't associate with APs on unallowed
channels (according to regulatory setting). This could happen
when the channel is not specified (auto-select) within the
connection request. In this case we get the AP's channel until
the firmware indicates the association succeeded later. We need
to verify the associated channel. If the channel is disabled by
regulatory, we have to disassociate with the AP.

Signed-off-by: Zhu Yi <yi.zhu@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

enhance sysfs rfkill interface

This commit introduces two new sysfs knobs.

/sys/class/rfkill/rfkill[0-9]+/blocked_hw: (ro)
hardblock kill state
/sys/class/rfkill/rfkill[0-9]+/blocked_sw: (rw)
softblock kill state

Signed-off-by: Florian Mickler <florian@mickler.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

Document the rfkill sysfs ABI

This moves sysfs ABI info from Documentation/rfkill.txt to the
ABI subfolder and reformats it.

This also schedules the deprecated sysfs parts to be removed in
2012 (claim file) and 2014 (state file).

Signed-off-by: Florian Mickler <florian@mickler.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwl3945: check ucode load error code

Check successful completion of BSM loading. Give a chance to
load BSM again. If BSM loading is unsuccessful we should exit
and not initilize NIC. Before this checking even if the BSM loading failed
driver tries to initilize the NIC which should not be the case.

Signed-off-by: Abhijeet Kolekar <abhijeet.kolekar@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>

iwlwifi: clean up queue/fifo handling

4965 hardware has 7 queues reserved and the
remaining ones used for aggregation, 5000
and higher need to have 10 reserved. This
is not very clear in the code right now,
unfortunately.

Introduce a new IWL_TX_FIFO_UNUSED constant
and make the queue/FIFO mapping arrays able
to hold that value, and change the setup
code to reserve all queues in the arrays
(the queue number is the index) and use the
new unused constant to not map those queues
to any FIFO.

Additionally, clear up the AC/queue mapping
code to be more understandable. The mapping
is the identity mapping right now, but with
the mapping function I think it's easier to
understand what happens there.

Finally, HCCA isn't implemented at all and
I think newer microcode removed it, so let's
remove all mention of it in the code, some
comments remain for 4965.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Acked-by: Shanyu Zhao <shanyu.zhao@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>

iwlwifi: add internal short scan support for 3945

Add internal short scan support for 3945 NIC, This allows 3945 NIC
to support radio reset request like the other series of NICs.

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

iwlwifi: remove unnecessary parameter in scan function

small change to remove unnecessary "ret" parameter since it not being
used.

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

iwlwifi: remove duplicate function

"internal_scan" debugfs function serves the same propose of mode "0" of
"force_reset" debugfs function. Remove it to avoid duplicate.

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

iwlwifi: remove needlessly exported symbols

iwl_apm_stop_master and iwl_internal_short_hw_scan
are only used within iwlcore and thus do not need to be exported.

Reported-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>

iwlwifi: change WEP key protection to use mutex

For later station notification support we would like WEP key setting to be
done synchronously always. Currently all places from which WEP key is set
can sleep, but the usage of sta_lock prevents it to do so. Modify the
locking to use priv->mutex instead and thus enable this call to sleep.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>

iwlwifi: remove un-necessary parameters

small code cleanup to remove un-necessary parameters

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

iwlagn: move sysfs flags and filter_flags files to debugfs

These files are incompatible with some changes
I'm making, and don't really belong into sysfs
anyway as they can only be used for debugging.
Since writing them will probably crash the
firmware, keep only the reading part and move
them into debugfs.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>

iwlwifi: remove STATUS_MODE_PENDING

Since rfkill integration, mac80211 can no
longer add an interface while the hardware
is not ready, so STATUS_MODE_PENDING can
never be set.

Also, remove another superfluous channel
sanity check and return the commit_rxon
return value in case it failed.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>

iwlwifi: remove alive start adhoc restart

This code will never trigger, because when
we call this during interface start, neither
priv->vif nor priv->iw_mode conditions will
be true, and when it happens during reset
then priv->vif is also NULL. Also, in both
cases the same code will be executed as part
of iwl_bss_info_changed(), which contains a
copy of this code.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>

iwlwifi: move 3945 clip groups to 3945 data

Only used on 3945, so it can be in the
3945-specific part of the data.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>

iwlwifi: clear up AC/FIFO debug output

Not all queues are mapped for frame transmission,
so saying "AC" is misleading. They are all mapped
to FIFOs, obviously, though, so use that term and
also say AC/CMD instead of just AC for the queue
type.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>

iwlwifi: remove sanity check

iwl_mac_beacon_update can only be called
when the operating mode is IBSS, so it
doesn't need to verify that again.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>

iwlwifi: remove IBSS channel sanity check

mac80211 will always enforce using a channel
that the driver allowed for IBSS use.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>

iwlwifi: remove priv->active_rate_basic

This variable is assigned a default value,
but then assigned zero as soon as mac80211
calls a change channel (which will happen
right after the hw is started) and after
that it never changes again.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>

iwlwifi: remove never-changing priv->rates_mask variable

It's always just IWL_RATES_MASK.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>

iwlwifi: remove frame dropping

With some of the new code in mac80211, public action
frames can be exchanged as non-injected frames even
while not associated.

Aside from that, dropping frames here is pointless
since we do deal with arbitrary frames that were
injected already, so let mac80211 make the decision
about which frames to allow or not.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>

iwlwifi: remove dead code from iwl_mac_reset_tsf

iwl_mac_reset_tsf will only ever be called in
IBSS mode, so checking for other modes is not
useful and the code that depends on that will
not be executed.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>

iwlwifi: remove superfluous channel check

mac80211 will not allow using channels not
marked as allowing IBSS for IBSS mode.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>

iwlagn: remove write-only variables

Updating the variables last_rx_rssi, last_tsf
and last_beacon_time needs a lot of code but
they are not actually used in iwlagn (only in
3945) so we can move them to the 3945 specific
data.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>

iwlwifi: move ICT data to agn part of union

Since the ICT data is all AGN specific, it can
be the first data to create the _agn part of the
device-specific union in the priv struct.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>

iwlwifi: move 3945 specific data into union

To be obviously correct, this patch only modifies
3945 files and the struct definition. A follow-up
patch will create the AGN part of the union.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>

iwlagn: move ICT code into separate file

All the ICT ISR code is iwlagn specific, and doesn't
need to be in iwlcore. So create a new iwl-agn.h
header file that will hold agn specific function
declarations etc., and move the ICT code into a new
iwl-agn-ict.c file that is linked into iwlagn. This
also gets rid of exporting those symbols.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>

b43: N-PHY: switch to chanspec ops

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

b43: N-PHY: implement chanspec setup

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

b43: N-PHY: isloate 2055 radio setup

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

b43: N-PHY: adjust gain table

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

b43: N-PHY: switch to chanspec struct

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

b43: N-PHY: update post init of 2055 radio

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

b43: N-PHY: update writing channel-specific radio registers

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

b43: N-PHY: turn radio on/off (rfkill)

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

b43: N-PHY: initialize super switch

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

b43: N-PHY: add some registers and structs definitions

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

mac80211: use different MAC addresses for virtual interfaces

Drivers can now advertise to cfg80211 that they have
multiple MAC addresses reserved for a device, but we
don't currently make use of that in mac80211.

Change that and assign different addresses to new
virtual interfaces (if addresses are available) in
order to make it easier for users to use multiple
virtual interfaces; they no longer need to always
assign a new MAC address manually.

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

mac80211: Improve software scan timing

The current software scan implemenation in mac80211 returns to the operating
channel after each scanned channel. However, in some situations (e.g. no
traffic) it would be nicer to scan a few channels in a row to speed up
the scan itself.

Hence, after scanning a channel, check if we have queued up any tx frames and
return to the operating channel in that case.

Unfortunately we don't know if the AP has buffered any frames for us. Hence,
scan only as many channels in a row as the pm_qos latency and the negotiated
listen interval allows us to.

Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k: never read from the AR_IMR_S2 register

The AR_IMR_S2 register sometimes cannot be read correctly.  Instead of a
valid value, 0xdeadbeef is returned.  The driver has been observed
writing that value back to AR_IMR_S2 after changing a few bits.

Cache the register value in ah->imrs2_reg and always write chached value
to the register.

Signed-off-by: Pavel Roskin <proski@gnu.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

mac80211: fix rates setup on IBSS merge

when an IBSS merge happened, the supported rates for the newly added station
were left empty, causing the rate control module to be initialized with only
the basic rates.

also the section of the ibss code which deals with updating supported rates for
an already existing station fails to inform the rate control module about the
new rates. as i don't know how to fix this (minstrel does not have an update
function), i have just added a comment for now.

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

wl1271: Remove tx-power level workaround

This patch removes the workaround limiting chipset TX power to 12dB, instead
using the value configured by the mac80211.

Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Reviewed-by: Teemu Paasikivi <ext-teemu.3.paasikivi@nokia.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

wl1271: Disable host TX rate control

Disable host TX rate control. The wl1271 firmware is already managing rate
control, so this eliminate unnecessary host processing.

Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Reviewed-by: Teemu Paasikivi <ext-teemu.3.paasikivi@nokia.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

wl1271: Clean up TX security sequence number handling

Instead of managing the TX security sequence number as two variables, use
one 64 bit variable. This greatly simplifies the handling of the number.

Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Reviewed-by: Teemu Paasikivi <ext-teemu.3.paasikivi@nokia.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

wl1271: Clean up firmware block allocation calculation

This patch simplifies the required block count calculation. Though it
introduces a division operator, it greatly simplifies the formula.

Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Reviewed-by: Teemu Paasikivi <ext-teemu.3.paasikivi@nokia.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

wl1271: Update TX packet life time handling with higher resolution time

This patch uses a higher precision timer to synchronize with the firmware
clock. Improved precision is needed as on some platforms a jiffy may be up to
tens of milliseconds, and the required precision is closer to TU's.

Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Reviewed-by: Teemu Paasikivi <ext-teemu.3.paasikivi@nokia.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

wl1271: Implement looped IRQ handling

This patch implements looped IRQ handling. In essence, if a new interrupt is
asserted by the FW while the host is processing the previous one, the host
will directly proceed processing the new IRQ without leaving the handling
function.

Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Reviewed-by: Teemu Paasikivi <ext-teemu.3.paasikivi@nokia.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

wl1271: Don't mask interrupts while handling interrupt

Don't mask firmware interrupts while processing interrupts. This allows the
interrupt handler looping to work efficiently thus reducing interrupt
processing latency.

Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Reviewed-by: Teemu Paasikivi <ext-teemu.3.paasikivi@nokia.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

wl1271: Aggregate RX acknowledgements to FW

This patch will ack RX frames read from the firmware in one single write,
instead of acking all the frames separately. This will reduce the amount of
required communication per frame.

Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Reviewed-by: Teemu Paasikivi <ext-teemu.3.paasikivi@nokia.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

wl1271: Remove annoying PSM entry/exit kernel traces

Remove the annoying and dmesg-flooding WLAN PSM entry/exit traces. Instead,
only output them if PSM traces are enabled.

Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Reviewed-by: Kalle Valo <kalle.valo@nokia.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

wl1271: Fix queue stopping/waking for TX path

The queue stopping/waking functionality was broken in a way that could
cause huge latencies in TX transfers and even cause the TX to stall in the
right circumstances. Correct these problems.

Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Reviewed-by: Kalle Valo <kalle.valo@nokia.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

wl1271: Fix ad-hoc mode neighborhood detection

This patch fixes a bug in ad-hoc mode preventing mac80211 from properly
detecting other ad-hoc networks with the same SSID.

Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Reviewed-by: Kalle Valo <kalle.valo@nokia.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

wl1271: Improvements to the TX path

- Fix a TX result overflow problem that was present in the TX path and visible
  with at least linksys AP's (probably any AP with high throughput capability.)
- Optimize TX by writing FW trigger for a group of TX frames instead of
  each and every frame.
- Slightly optimize the TX path code.

Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Reviewed-by: Kalle Valo <kalle.valo@nokia.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

wl1271: don't get received frames from hardware in PLT mode

Most probably patch "wl1271: add most of the normal initialization commands
to PLT mode" enabled the RX path in firmware so that now driver received
frames and passed them to mac80211, which warned about them.

Workaround this by not retrieving frames from the hardware, just ignore them.

Signed-off-by: Kalle Valo <kalle.valo@nokia.com>
Reviewed-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

wl1271: Fixed unloading of the wl1271_sdio module

Fixed two bugs causing problems when unloding wl1271 module. First was
missing sdio_set_drvdata call from the probe function, second was order
of function calls in the remove function.

Signed-off-by: Teemu Paasikivi <ext-teemu.3.paasikivi@nokia.com>
Reviewed-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

wl1271: Changed access to fw status register to use raw read

Changed access to fw status register to use raw read instead of
translated addressing. On SDIO access by translated addressing was not
working and raw access is working also on SPI.

Signed-off-by: Teemu Paasikivi <ext-teemu.3.paasikivi@nokia.com>
Reviewed-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

wl1271: Initial SDIO implementation

Added initial implementation of SDIO interfacte to the wl1271 driver.
When selected, this adds new module called "wl1271_sdio".

Signed-off-by: Teemu Paasikivi <ext-teemu.3.paasikivi@nokia.com>
Reviewed-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

wl1271: Divided driver to two separate modules

Divided wl1271 driver to wl1271 "core" and wl1271_spi modules in preparation
of integration of the SDIO implementation.

Signed-off-by: Teemu Paasikivi <ext-teemu.3.paasikivi@nokia.com>
Reviewed-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

wl1271: Removed wl1271_spi.h and made some functions static

Removed wl1271_spi.h header as there's no more need to access functions
declared there outside of wl1271_spi.c. Also made those SPI access
functions static.

Signed-off-by: Teemu Paasikivi <ext-teemu.3.paasikivi@nokia.com>
Reviewed-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

wl1271: Inlined IO functions

Changed IO functions to static inline.

Signed-off-by: Teemu Paasikivi <ext-teemu.3.paasikivi@nokia.com>
Reviewed-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

wl1271: Implemented abstraction of IO functions.

Changed the driver to use if_ops structure to abstract access to the IO
layer (SPI or SDIO).

Signed-off-by: Teemu Paasikivi <ext-teemu.3.paasikivi@nokia.com>
Reviewed-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

wl1271: Added functions to enable/disable interrupt handling

Added/moved enable and disable interrupt handling functions.

Signed-off-by: Teemu Paasikivi <ext-teemu.3.paasikivi@nokia.com>
Reviewed-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

wl1271: Moved module basics to wl1271_spi.c

Moved wl1271 drivers probe, remove etc. functions and structres to
wl1271_spi.c from wl1271_main.c in preparation of implementing SDIO interface.

Signed-off-by: Teemu Paasikivi <ext-teemu.3.paasikivi@nokia.com>
Reviewed-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

p54: Enable HW_REPORTS_TX_ACK_STATUS

P54 devices always generate a full tx_status report
(ACK, PSM, rate control, etc..) for every xmitted frame.

Therefore, I think The driver qualifies for the
REPORTS_TX_ACK_STATUS hardware feature flag.

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

ath5k: move ath5k_hw_register_timeout() into reset.c

ath5k_hw_register_timeout() was duplicated between phy.c and reset.c.
Since it is too big and too much used to be an inline function, move it
away from the ath5k.h header into reset.c.  Remove _ATH5K_RESET and
_ATH5K_PHY defines.

Signed-off-by: Pavel Roskin <proski@gnu.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath5k: remove useless "extern" from function declarations

Adjust formatting of the affected lines to satisfy checkpatch.pl.

Signed-off-by: Pavel Roskin <proski@gnu.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath5k: remove stale function declarations, make some functions static

Remove all unnecessary function declarations from ath5k.h.  Comment out
unused functions.  Remove ath5k_hw_get_tsf32(), which is too trivial to
be commented out.  Make functions static if suggested by sparse.  Make
ath5k_pm_ops static.

Signed-off-by: Pavel Roskin <proski@gnu.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

mac80211: deprecate RX status noise

The noise value as is won't be used, isn't
filled by most drivers and doesn't really
make a whole lot of sense on a per packet
basis -- proper cfg80211 survey support in
mac80211 will need to be different.

Mark the struct member as deprecated so it
will be removed from drivers.

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

rt2x00: Export rt2x00soc_probe from rt2x00soc

Export rt2x00soc_probe from rt2x00soc as it is used in rt2800pci.

Otherwise loading rt2800pci gives "rt2800pci: Unknown symbol
rt2x00soc_probe".

Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Acked-by: Gertjan van Wingerde <gwingerde@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

mac80211: Fix HT rate control configuration

Handling HT configuration changes involved setting the channel
with the new HT parameters and then issuing a rate_update()
notification to the driver.

This behavior changed after the off-channel changes. Now, the channel
is not updated with the new HT params in enable_ht() - instead, it
is now done when the scan work terminates. This results in the driver
depending on stale information, defaulting to non-HT mode always.

Fix this by passing the new channel type to the driver.

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

iwlwifi: load firmware asynchronously before mac80211 registration

At the wireless summit in Portland we discussed a way of
loading firmware asynchronously from ->probe() before
registration to mac80211, in order to register with the
wireless subsystems with complete information in cases
where firmware is required to know parameters.

This is not yet the case in iwlwifi, but for some new
features we're working on it will be the case since
those will only be supported by new firmware images.

Hence, to start with, convert iwlwifi to load firmware
asynchronously from probe, unbinding the device when
firmware loading fails, and only registering with the
wireless subsystems after firmware has been loaded
successfully.

Future patches will hook into this to register the
new firmware capabilities, depending on the firmware
API version.

Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ar9170: load firmware asynchronously

This converts ar9170 to load firmware asynchronously
out of ->probe() and only register with mac80211 when
all firmware has been loaded successfully. If, on the
other hand, any firmware fails to load, it will now
unbind from the device.

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

rt2x00: fix rt2800pci compilation with SoC

Compiling rt2800pci with CONFIG_RT2800PCI_SOC fails with "... rt2880pci.c:
error: incompatible type for argument 2 of 'rt2x00soc_probe'".

Fix this by using &rt2800pci_ops instead of rt2800pci_ops.

Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

rt2x00: Add USB ID for CEIVA adapter to rt73usb

This adds support for CEIVA USB wireless adapters to the rt73usb driver.

Signed-off-by: Bryan Polk <sainth@eidolons.org>
Acked-by: Ivo van Doorn <IvDoorn@gmail.com>
Acked-by: Gertjan van Wingerde <gwingerde@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

mac80211: Fix reassociation processing (within ESS roaming)

Commit e1dd33f60ced091114e4aacf141e0d03b88d3e13 changed cfg80211 to
allow association commands while in associated state to enable support
for roaming within an ESS. However, this was not enough to resolve all
cases with mac80211 which needs some additional handling of the
reassociation case to clear internal state with the BSS that was in use
previously.

This patch makes ieee80211_mgd_assoc() accept a valid reassociation
command and clean the association state with the previous BSS. This
fixes roaming between BSSes in an ESS when using wpa_supplicant with
-Dnl80211.

Signed-off-by: Jouni Malinen <j@w1.fi>
Cc: stable@kernel.org
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k: fix lockdep warning when unloading module

Since txq->axq_lock may be hold in softirq context, it must be
acquired with spin_lock_bh() instead of spin_lock() if softieq is
enabled.

The patch fixes the lockdep warning below when unloading ath9k modules.

=================================
[ INFO: inconsistent lock state ]
2.6.33-wl #12
---------------------------------
inconsistent {IN-SOFTIRQ-W} -> {SOFTIRQ-ON-W} usage.
rmmod/3642 [HC0[0]:SC0[0]:HE1:SE1] takes:
 (&(&txq->axq_lock)->rlock){+.?...}, at: [<ffffffffa03568c3>] ath_tx_node_cleanup+0x62/0x180 [ath9k]
{IN-SOFTIRQ-W} state was registered at:
  [<ffffffff8107577d>] __lock_acquire+0x2f6/0xd35
  [<ffffffff81076289>] lock_acquire+0xcd/0xf1
  [<ffffffff813a7486>] _raw_spin_lock_bh+0x3b/0x6e
  [<ffffffffa0356b49>] spin_lock_bh+0xe/0x10 [ath9k]
  [<ffffffffa0358ec7>] ath_tx_tasklet+0xcd/0x391 [ath9k]
  [<ffffffffa0354f5f>] ath9k_tasklet+0x70/0xc8 [ath9k]
  [<ffffffff8104e601>] tasklet_action+0x8c/0xf4
  [<ffffffff8104f459>] __do_softirq+0xf8/0x1cd
  [<ffffffff8100ab1c>] call_softirq+0x1c/0x30
  [<ffffffff8100c2cf>] do_softirq+0x4b/0xa3
  [<ffffffff8104f045>] irq_exit+0x4a/0x8c
  [<ffffffff813acccc>] do_IRQ+0xac/0xc3
  [<ffffffff813a7d53>] ret_from_intr+0x0/0x16
  [<ffffffff81302d52>] cpuidle_idle_call+0x9e/0xf8
  [<ffffffff81008be7>] cpu_idle+0x62/0x9d
  [<ffffffff81391c1a>] rest_init+0x7e/0x80
  [<ffffffff818bbd38>] start_kernel+0x3e8/0x3f3
  [<ffffffff818bb2bc>] x86_64_start_reservations+0xa7/0xab
  [<ffffffff818bb3b8>] x86_64_start_kernel+0xf8/0x107
irq event stamp: 42037
hardirqs last  enabled at (42037): [<ffffffff813a7b21>] _raw_spin_unlock_irqrestore+0x47/0x56
hardirqs last disabled at (42036): [<ffffffff813a72f4>] _raw_spin_lock_irqsave+0x2b/0x88
softirqs last  enabled at (42000): [<ffffffffa0353ea6>] spin_unlock_bh+0xe/0x10 [ath9k]
softirqs last disabled at (41998): [<ffffffff813a7463>] _raw_spin_lock_bh+0x18/0x6e

other info that might help us debug this:
4 locks held by rmmod/3642:
 #0:  (rtnl_mutex){+.+.+.}, at: [<ffffffff8132c10d>] rtnl_lock+0x17/0x19
 #1:  (&wdev->mtx){+.+.+.}, at: [<ffffffffa01e53f2>] cfg80211_netdev_notifier_call+0x28d/0x46d [cfg80211]
 #2:  (&ifmgd->mtx){+.+.+.}, at: [<ffffffffa0260834>] ieee80211_mgd_deauth+0x3f/0x17e [mac80211]
 #3:  (&local->sta_mtx){+.+.+.}, at: [<ffffffffa025a381>] sta_info_destroy_addr+0x2b/0x5e [mac80211]

stack backtrace:
Pid: 3642, comm: rmmod Not tainted 2.6.33-wl #12
Call Trace:
 [<ffffffff81074469>] valid_state+0x178/0x18b
 [<ffffffff81014f94>] ? save_stack_trace+0x2f/0x4c
 [<ffffffff81074e08>] ? check_usage_backwards+0x0/0x88
 [<ffffffff8107458f>] mark_lock+0x113/0x230
 [<ffffffff810757f1>] __lock_acquire+0x36a/0xd35
 [<ffffffff8101018d>] ? native_sched_clock+0x2d/0x5f
 [<ffffffffa03568c3>] ? ath_tx_node_cleanup+0x62/0x180 [ath9k]
 [<ffffffff81076289>] lock_acquire+0xcd/0xf1
 [<ffffffffa03568c3>] ? ath_tx_node_cleanup+0x62/0x180 [ath9k]
 [<ffffffff810732eb>] ? trace_hardirqs_off+0xd/0xf
 [<ffffffff813a7193>] _raw_spin_lock+0x36/0x69
 [<ffffffffa03568c3>] ? ath_tx_node_cleanup+0x62/0x180 [ath9k]
 [<ffffffffa03568c3>] ath_tx_node_cleanup+0x62/0x180 [ath9k]
 [<ffffffff810749ed>] ? trace_hardirqs_on+0xd/0xf
 [<ffffffffa0353950>] ath9k_sta_remove+0x22/0x26 [ath9k]
 [<ffffffffa025a08f>] __sta_info_destroy+0x1ad/0x38c [mac80211]
 [<ffffffffa025a394>] sta_info_destroy_addr+0x3e/0x5e [mac80211]
 [<ffffffffa02605d6>] ieee80211_set_disassoc+0x175/0x180 [mac80211]
 [<ffffffffa026084d>] ieee80211_mgd_deauth+0x58/0x17e [mac80211]
 [<ffffffff813a60c1>] ? __mutex_lock_common+0x37f/0x3a4
 [<ffffffffa01e53f2>] ? cfg80211_netdev_notifier_call+0x28d/0x46d [cfg80211]
 [<ffffffffa026786e>] ieee80211_deauth+0x1e/0x20 [mac80211]
 [<ffffffffa01f47f9>] __cfg80211_mlme_deauth+0x130/0x13f [cfg80211]
 [<ffffffffa01e53f2>] ? cfg80211_netdev_notifier_call+0x28d/0x46d [cfg80211]
 [<ffffffff810732eb>] ? trace_hardirqs_off+0xd/0xf
 [<ffffffffa01f7eee>] __cfg80211_disconnect+0x111/0x189 [cfg80211]
 [<ffffffffa01e5433>] cfg80211_netdev_notifier_call+0x2ce/0x46d [cfg80211]
 [<ffffffff813aa9ea>] notifier_call_chain+0x37/0x63
 [<ffffffff81068c98>] raw_notifier_call_chain+0x14/0x16
 [<ffffffff81322e97>] call_netdevice_notifiers+0x1b/0x1d
 [<ffffffff8132386d>] dev_close+0x6a/0xa6
 [<ffffffff8132395f>] rollback_registered_many+0xb6/0x2f4
 [<ffffffff81323bb8>] unregister_netdevice_many+0x1b/0x66
 [<ffffffffa026494f>] ieee80211_remove_interfaces+0xc5/0xd0 [mac80211]
 [<ffffffffa02580a2>] ieee80211_unregister_hw+0x47/0xe8 [mac80211]
 [<ffffffffa035290e>] ath9k_deinit_device+0x7a/0x9b [ath9k]
 [<ffffffffa035bc26>] ath_pci_remove+0x38/0x76 [ath9k]
 [<ffffffff8120940a>] pci_device_remove+0x2d/0x51
 [<ffffffff8129d797>] __device_release_driver+0x7b/0xd1
 [<ffffffff8129d885>] driver_detach+0x98/0xbe
 [<ffffffff8129ca7a>] bus_remove_driver+0x94/0xb7
 [<ffffffff8129ddd6>] driver_unregister+0x6c/0x74
 [<ffffffff812096d2>] pci_unregister_driver+0x46/0xad
 [<ffffffffa035bae1>] ath_pci_exit+0x15/0x17 [ath9k]
 [<ffffffffa035e1a2>] ath9k_exit+0xe/0x2f [ath9k]
 [<ffffffff8108050a>] sys_delete_module+0x1c7/0x236
 [<ffffffff813a7df5>] ? retint_swapgs+0x13/0x1b
 [<ffffffff810749b5>] ? trace_hardirqs_on_caller+0x119/0x144
 [<ffffffff8109b9f6>] ? audit_syscall_entry+0x11e/0x14a
 [<ffffffff81009bb2>] system_call_fastpath+0x16/0x1b
wlan1: deauthenticating from 00:23:cd:e1:f9:b2 by local choice (reason=3)
PM: Removing info for No Bus:wlan1
cfg80211: Calling CRDA to update world regulatory domain
PM: Removing info for No Bus:rfkill2
PM: Removing info for No Bus:phy1
ath9k 0000:16:00.0: PCI INT A disabled

Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

zd1211rw: fix potential array underflow

The first chunk fixes a debugging assert to print a warning about array underflows.
The second chunk corrects a potential array underflow.  I also removed an assert
in the second chunk because it can no longer happen.

Signed-off-by: Dan Carpenter <error27@gmail.com>
Acked-by: Benoit Papillault <benoit.papillault@free.fr>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

rfkill: Add support for KEY_RFKILL

Add support for handling KEY_RFKILL in the rfkill input module. This
simply toggles the state of all rfkill devices. The comment in rfkill.h
is also updated to reflect that RFKILL_TYPE_ALL may be used inside the
kernel.

Signed-off-by: Matthew Garrett <mjg@redhat.com>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

airo: return from set_wep_key() when key length is zero

Even if keylen == 0 is a bug and should not really happen, better avoid
possibility of passing bad value to firmware.

Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

virtio_net: remove forgotten assignment

This is no longer needed. I missed to remove this in
567ec874d15b478c8eda7e9a5d2dcb05f13f1fb5 ("net: convert multiple
drivers to use netdev_for_each_mc_addr, part6")

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

be2net: fix tx completion polling

In tx/mcc polling, napi_complete() is being incorrectly called
before reaping tx completions. This can cause tx compl processing
to be scheduled on another cpu concurrently which can result in a panic.
This if fixed by calling napi complete() after tx/mcc compl processing
but before re-enabling interrupts (via a cq notify).

Signed-off-by: Sathya Perla <sathyap@serverengines.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

sis190: fix cable detect via link status poll

Some sis190 devices don't report LinkChange, so do polling for
link status.

Fixes http://bugzilla.kernel.org/show_bug.cgi?id=11926

Signed-off-by: Herton Ronaldo Krzesinski <herton@mandriva.com.br>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: fix protocol sk_buff field

Commit e992cd9b72a18 (kmemcheck: make bitfield annotations truly no-ops
when disabled) allows us to revert a workaround we did in the past to
not add holes in sk_buff structure.

This patch partially reverts commit 14d18a81b5171
(net: fix kmemcheck annotations) so that sparse doesnt complain:

include/linux/skbuff.h:357:41: error: invalid bitfield specifier for
type restricted __be16.

Reported-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>